lable at ScienceDirect

Building and Environment 46 (2011) 1808e1816

Contents lists avai

Building and Environment

journal homepage: www.elsevier .com/locate/bui ldenv

Degradation of NO using photocatalytic coatings applied to different substrates

Thomas Martinez a,b, Alexandra Bertron a,*, Erick Ringot a,b, Gilles Escadeillas a

aUniversité de Toulouse, UPS, INSA, LMDC (Laboratoire Matériaux et Durabilité des Constructions) 135, avenue de Rangueil, F-31 077 Toulouse Cedex 04, Franceb LRVision SARL, Zi de Vic, 13 rue du Développement, 31320 Castanet-Tolosan, France

a r t i c l e i n f o

Article history:Received 20 December 2010Received in revised form4 March 2011Accepted 6 March 2011

Keywords:PollutionNOx

PhotocatalysisTitanium dioxideCoatingLasure

* Corresponding author. Tel.: þ33 5 61 55 99 31; faE-mail address: bertron@insa-toulouse.fr (A. Bertr

0360-1323/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.buildenv.2011.03.001

a b s t r a c t

This article deals with the degradation of NO present in the air by means of a photocatalytic oxidationprocess based on TiO2 nanoparticles incorporated in a polymer-matrix-based coating. The experimentalset-up consisted of a flow type reactor adapted from the ISO 22197-1 standard. NO2 in the gas phase, andnitrate ions adsorbed on the photocatalytic surface were detected as finals products. Various parametersinfluencing the NO degradation efficiency were studied: the coating composition, the substrate nature,the initial concentration of NO, the polluted air flow rate and the humidity. Compared to glass, the use ofmortar as the substrate enhanced the photocatalytic performance of coatings by reducing the generationof gaseous NO2 as a by-product.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

In an urban environment, air is strongly polluted by nitrogenousoxides NOx (NOx ¼ NO þ NO2) produced by intensive humanactivity, notably transport. In living areas, NOx are produced bydomestic combustion devices such as gas burners for cooking andby the infiltration of outdoor pollution. Actually, in urban areas,pollution concentration levels are very similar inside and outsideand can reach up to one ppm [1,2].

The aim of this paper was to investigate a promising way oflimiting the NOx levels in air. The main idea was to exploit thephotocatalytic properties of titanium dioxide (in the form ofanatase) under simulated solar illumination. Photocatalysis is acti-vated by irradiation of semiconductormaterials, here TiO2 particles,with high energy photons (hn) that raise electrons e- from thevalence band (vb) to the conduction band (cb), thus leaving electronholes hþ (reaction (1)). The pairs of mobile charges produced canreach the surface of the semiconductor particle and initiatea reductioneoxidation process. Moreover, through reactions withthe adsorbed oxygen and water coming from the surrounding air,reactive oxygen species such as HO, and O2

,� are created (reactions(2) and (3)) and act as strong oxidants with the potential todecompose or mineralize a wide range of compounds [3,4].

TiO2��!hn TiO2 þ hþvb þ e�cb (1)

x: þ33 5 61 55 99 49.on).

All rights reserved.

H2Oads þ hþ/Hþ þHO, (2)

ðO2Þadsþe�/O,�2 (3)

The reactivity of the oxygen species generated leads to theoxidation of NO to NO2 which, in turn, produces nitrite and nitrateions NO2

�/NO3� [5e7]. Basically, the degradation of nitrogen oxides

by photocatalysis can be described as follows:

NO/HNO2/NO2/HNO3 (4)

As the band gap energy of unmodified TiO2 is around 3.2 eV, thephotocatalysis activation requires UV light (l < 388 nm), which is,unfortunately, scarce in indoor environments. For indoor air puri-fication, two methods are employed to make the photocatalystactive under visible light. One involves chemical modifications ofthe UV active photocatalyst in order to enlarge the photoadsorptionto the visible region of the spectrum and to make it efficient inindoor environments. Other studies focus on the development ofphotocatalysts active under visible light. For example, carbon-doped TiO2, BiOBr or PbWO4 have shown NOx purification abilitiesunder visible light [8e10]. In this paper, commercially availableTiO2 active under UV light is used to study the NOx purificationunder outdoor conditions.

Different processes can be used to bind the photocatalyst to thesurface. Today, most photocatalysis applications for buildingmaterials involve the mixing of TiO2 inside concrete or mortars.Because photocatalysis is a surface phenomenon, solutions such ascoatings are also interesting because they can be applied to existing

T. Martinez et al. / Building and Environment 46 (2011) 1808e1816 1809

buildings and, at the same time, result in lower TiO2 consumption.Among coatings, one can distinguish thick coatings, paints andlasures. Lasures are a kind of light varnish, with a compositionsimilar to a paint, enabling the initial aspect of the substrate(concrete, wood, stone or brick, for example) to be at least partiallypreserved (colour, texture, glossiness).

Studies on photocatalytic paints have shown their abilities todegrade NOx. In an environmental chamber (30 m3), Maggos et al.[11] measured a significant NO and NO2 photocatalytic degradationby two photocatalytic paints (between 74 and 91% for NO andbetween 71 and 27% for NO2, 6 h of illumination, static mode). Ina real-scale study, Maggos et al. [12] measured near to 20% removalof NOx by a photocatalytic paint in a car park. Some laboratorystudies conducted on flow type reactor [11,13e16] showed rapidphotocatalytic degradation of NO and NO2 by photocatalytic coat-ings. This type of experimental apparatus allowed the performanceof photocatalytic samples to be compared and showed the influ-ence of simulated environmental parameters, such as air humidity,flow rate, illumination and initial pollutant concentration. Somenormative references describe the test method and the geometry ofthe photoreactor [17,18].

Studies based on this type of reactor have reached a consensusregarding the influence of flow rate, initial pollutant concentrationand light intensity on the efficiency of NO degradation. However,the influence of moisture on photocatalytic NO degradation gaverise to different observations. Its influence seems to be related tothe photocatalytic material and to the experimental conditions[5,8,11,19,20]. Furthermore, Obee and Brown showed that thephotocatalytic degradation of VOCs is dependent on both the airhumidity and the initial concentration of pollutant [21]. In the caseof NOx, the effect of humidity at different initial NO concentrationsis still unclear.

The formulation parameters of photocatalytic paints influencingthe photocatalytic degradation of NO has not aroused great interestyet, although Allen et al. showed that the paint matrix gives rise toa barrier effect as the NOx reduction is clearly better in the presenceof the photocatalyst alone [22]. To our knowledge, the influence ofthe formulation of the photocatalytic paint (parameters such as theamount of photocatalyst and of binder) has not been muchinvestigated.

NOx reduction can be enhanced by the immobilization tech-nique of the photocatalyst [19,23,24]. Measurements of the pho-tocatalytic degradation of NO by TiO2 immobilized on an activatedcarbon filter show that this material improves the degradation ofNO and decrease the associated generation of NO2 compared toTiO2 immobilized on glass fibre filter [19]. The improvement in the

Fig. 1. Schematic illustration of the a

degradation of pollutant can be observed because of the largeadsorption capacity of activated carbon. Pollutants adsorbed on thesupporting material are diffused to the photocatalyst for photo-degradation and the competition between humidity and pollutantson the photoactive sites is lowered. Consequently, in the case of theapplication of a photocatalytic paint or lasure, porous buildingmaterials could enhance the photocatalytic reaction through theiradsorption capacity.

In this article, the NOx abatement efficiency of laboratory-madephotocatalytic lasures is explored. Mortar and glass substrates wereused. The first was a porous building material representative of thesurface of concrete blocks, concrete walls or mortar coatings, andthe second was a non-adsorptive material often used as a substratein photocatalysis experiments [11,25]. An experimental procedureand a set-up based on the ISO 22197-1 standard [17] were imple-mented. Artificial light with a spectrum close to that of daylight wasused as the light source and NOx abatement was monitored usingchemiluminescence. The influence of some environmentalparameters, such as humidity, flow rate, and initial pollutantconcentration, and some formulation parameters, such as theamounts of photocatalyst and binder and the type of material usedas a substrate, on photocatalytic efficiency were investigated. Theamounts of nitrite and nitrate trapped on test pieces were titratedusing ion chromatography.

2. Materials and methods

2.1. Coating formulations

Lasuresweremadeusinganacrylicbinder (HexionD2040).Waterwas used as the solvent to limit the use of hazardous chemicalproducts. Additives such as thickeners andwetting agentswere usedto obtain a coating that was easy to apply (good wettability, unifor-mity of aspect). The photocatalyst (anatase-TiO2) was a commercialslurry solution available from Evonik (Aerodisp�W740X).

2.2. Coating method

The coatings were primarily formulated for the surface treat-ment of building materials. To identify the possible influence of thesubstrate nature on the photocatalytic efficiency, various types ofsubstrates were tested. The coatings were applied to (i) mortars, (ii)glass plates and (iii) a non-absorbent cardboard material (Lenetastandardized contrast cards intended for the coating industry). Theformulation and making of the mortars complied with standard NFEN 196-1 [26].

utomatic film applicator set-up.

Fig. 2. Schematic diagram of the experimental apparatus for NO degradation. (1) zeroair generator; (2) NO source; (3) thermal mass flow controller; (4) gas washing bottlehumidifier; (5) mixing chamber; (6) temperature and relative humidity probes; (7) reactor cell; (8) NOx analyser; (9) bypass; (10) illuminant.

T. Martinez et al. / Building and Environment 46 (2011) 1808e18161810

Coatings of a wet thickness of 40 mm were applied to thecontrast cards using an automatic film applicator (ELCOMETER,Fig. 1). A few millilitres of coating were deposited on the substratein front of a wire bar. The wire bar was attached to a motorisedsystem allowing the product to be spread uniformly on thesubstrate. The thickness of the wet coating applied was controlledby the area of the groove between the coils of wire. The coating wasapplied to mortars and glass substrates using a brush. The amountof coating deposited (about 40 g m�2 e determined by weighing)was similar to that used on non-absorbent substrates.

2.3. Experimental apparatus

2.3.1. PrincipleA schematic diagram of the experimental set-up is shown in

Fig. 2. Two ultra-pure dry air (or zero-air) streams were supplied byan air generator (Environnement SA,model ZAG7001) using filteredambient air. One of these purified air streams was humidified bypassing through a gas washing bottle. Then, the total air streamwasmixed with a pollutant stream containing 12 ppm of NO balancedwith nitrogen (Air Liquide France). The final humidity, the concen-tration of NO and the total flow ratewere adjusted using threemassflowcontrollers (Bronkhorst). To study the influenceof oxygen in air,the air generator was punctually replaced by a nitrogen cylinder.

Fig. 3. Reac

2.3.2. ReactorA cylindrical borosilicate-glass reactor (diameter ¼ 60 mm,

length¼ 300mm)was used for its high transparency toUV-A and itslow adsorption capacity (Fig. 3). Coated and control samples(100�50mm2)wereplaced in themedianplaneof the reactor usinga PTFEholder, so that the gas circulated through the semi-cylindricalspace between the test piece and the upper part of the reactor.

2.3.3. IlluminantIn the case of NOx [5,8,20] and more generally [27], the degra-

dation rate of polluting gas increases with the light intensity in thephotocatalyst activation region of the spectrum. In outdoor condi-tions, Fujishima and Zhang reported a UV illumination of20e30 W m�2 in direct sunlight in Japan [28]. Husken et al.measured a UV-A irradiance of 7e10 W m�2 on a cloudy day. Theyalso conducted an outside measurement using a reactor celladapted from the ISO 22197-1 standard [17]. The results showedthat the photocatalytic reaction was ensured even with the low UVillumination of a cloudy day [20]. Occasional outdoor measure-ments performed using a UV-A radiometer on a summer day in thesouth of France enabled light intensities over 30 W m�2 to berecorded (original data). This data is in accordancewith the value of35 W m�2 obtained by Blöß and Elfenthal during measurementsmade in Central Europe on a summer day [29].

tor cell.

Fig. 5. Effect of the amounts of photocatalyst and of binder on the NO degradationefficiency measured on contrast card (Initial NO concentration ¼ 400 ppb,Q ¼ 1.5 l min-1, H ¼ 6 g kg-1).

T. Martinez et al. / Building and Environment 46 (2011) 1808e1816 1811

In our apparatus, samples were illuminated using a 300-WOSRAMUltravitalux bulbwith an emission spectrum close to that ofdaylight. The light intensity measured using a UV-A radiometer(GigahertzeOptik) was 5.8 W m�2 whereas the test proceduredescribed in ISO22197-1 [17] requires a light intensityof 10Wm�2 inthe UV-A.

2.3.4. NOx analysisThe measurements of NO and NOx concentrations were per-

formed using an AC32M analyser (Environnement SA France) witha detection limit of 0.4 ppb and a continuous sampling rate of 0.7litre per minute. NO and NOx concentrations were measured insuccessive, 5-s steps. The NO2 concentrationwas obtained from thedifference between the NOx and NO concentrations.

2.3.5. Experimental procedureThe first step of the experimental procedure was performed in

darkness. The gas was made to pass through the bypass (Fig. 2)directly to the analyser in order to adjust the concentration of NO tothe target value using the mass flow controller. Once the measuredconcentrationwasstable, gaswasallowedtopass through the reactor.The concentration decreased immediately because of the filling timeof the cell and adsorption on surfaces (Fig. 4). After saturation, theNOconcentration returned to the initial value and photocatalytic reac-tions were then initiated by switching on the lamp.

Fig. 4 shows the typical evolution of NO, NOx and NO2 concen-trations during a test intended to assess the photocatalytic effi-ciency of a coating. The degradation of nitrogen oxides byphotocatalysis leads to the oxidation of NO to NO2, which, in turn,produces nitrite and nitrate ions (reaction (4)).

The photocatalytic performance of the various samples wasassessed by three degradation rates calculated as follows

NOdegx ð%Þ ¼ ½NOx�in�½NOx�out

½NOx�in� 100 (5)

NOdegð%Þ ¼ ½NO�in�½NO�out½NOx�in

� 100 (6)

0

50

100

150

200

250

300

350

400

450

0 10 20 30 40

Co

nc

en

tra

tio

n (

pp

b)

Time (min

Adsorption

Illuminatio

NO2 generation

Fig. 4. Typical variation of NOx concentration in the reactor during a test

NOdeg2 ð%Þ ¼ ½NO2�in�½NO2�out

½NOx�out� 100 (7)

Where [NOx]in, [NO]in and [NO2]in are the inlet concentrations ofthe experimental procedure and [NOx]out, [NO]out, [NO2]out are theaverage outlet concentrations measured in the last 30 min of a 1-hillumination period.

A negative degradation rate meant that intermediate productshas formed (NO2

deg(%)) whereas a positive value revealeda decrease of the concentration. NOx

deg(%) represents the overallefficiency of the photocatalytic material.

Using the procedure described above, the influence of differentsubstrates and different lasure formulations on photocatalysis wasassessed by varying the amounts of binder and photocatalyst. Theexperimental conditions were as follows: flow rate 1.5 l min�1;initial NO concentration 400 ppb; humidity 6 g kg�1 (correspondingto 31% RH at 25 �C). Several measurements in the conditions fixedabove showed a standard deviation of the NOx degradation rate of0.9% on seven different coatedmortar samples and of 3.6% on sevendifferent coated glass samples. In order to investigate a possiblephotolysis phenomenon (degradation of NO caused directly by the

50 60 70 80 90

ute)

NOxNONO2

n

(Initial NO concentration ¼ 400 ppb, Q ¼ 1.5 l min-1, H ¼ 6 g kg�1).

Fig. 6. Effect of flow rates on degradation rates measured on mortar substrates (InitialNO concentration ¼ 400 ppb, H ¼ 6 g kg-1).

T. Martinez et al. / Building and Environment 46 (2011) 1808e18161812

light), control experiments were performed using mortar substratewithout photocatalytic coating exposed to the illumination used inthewhole study. The removal rate ofNO in this conditionwas almostzero, which proved that no photolysis of NO occurred.

To study the influence of environmental parameters on thephotocatalytic process, flow rate, NO initial concentration, andhumidity were varied between: 1 and 5 l min�1; 100 and 2000 ppb,and 0 and 14.5 g kg�1 (0 and 74% at 25 �C), respectively. In this case,the most efficient laboratory lasures were used (2.3% dry mass ofacrylic binder, 15% TiO2).

Finally, to investigate the phenomenology of the NO degrada-tion, nitrite and nitrate ions trapped on the photocatalysis surfacewere analysed. Immediately after pollution abatement tests, coatedglass substrates were extracted from the reactor cell and immersedinto deionized water for 1 h. The solution was then analysed todetermine the NO3

� and NO2� concentrations using ion chroma-

tography (Dionex ICS-3000) equipped with a Dionex AS15 column.The eluent contained 38 mM KOH and the detection of anion wasperformed using the conductivity detection method. A secondimmersionwas carried out to ensure complete desorption of nitrateand nitrite from the test piece.

3. Results

3.1. Effect of amount of photocatalyst and binder

Photocatalytic lasures were formulated using three types ofcomponents: a solvent (water), a binder (acrylic) and a photo-catalyst suspension.

Fig. 7. Degradation of NO by photocatalytic coating immobilized on mortar (left) and glass

Degradation rates obtained for different amounts of binders andvarious concentrations of photocatalyst are shown in Fig. 5. First,the NO degradation rate increased with TiO2 content and thenremained constant. This behaviour is in accordance with otherstudies [30,31]. The optimal TiO2 content corresponded to themaximal photocatalytic surface area accessible to pollutingsubstances. Higher photocatalyst quantities were not photoactive,probably because of insufficient contact with light and pollutant.

An increase in the amount of binder decreased the coating’sperformance, probably because of a masking effect. This corre-sponded to excessive covering of the photocatalyst by the binder,which drastically reduced its contact with the gas phase. Conse-quently, it is reasonable to observe that increasing the bindercontent in the coating leads to an increase in the optimal catalystcontent.

3.2. Effect of flow rate

Gas flow rate in the reactor has a direct influence on the resi-dence time of pollutant on the photoactive surface. Fig. 6 shows theinfluence of flow rates varying from 1.0 to 5.0 l min�1. The exper-iment was performed with coated mortar samples. As in otherstudies [8,20], low flow rates enhanced the NO degradationperformance, probably because of an increase of the residence timeof the gas on the photocatalytic surface.

3.3. Impact of the substrate

The photocatalytic degradation of NO occurs in two stages: (1)oxidation of NO to NO2; (2) oxidation of NO2 to nitrate ions NO3

�.Fig. 7 shows measurements of the NOx degradation capacities of

the same coating applied to either mortar or glass substrates.For the mortar substrate, the rate of NOx degradation remained

the same throughout test. In the case of glass, this efficiencydecreased with time of exposure, revealing deactivation of thephotocatalyst. It may be assumed that, in the case of glass, nitrateions (NO3

�) formed by the degradation of NO2 [5e7] were adsorbedon the surface and progressively occupied the adsorptive sites of thecoating whereas, in the case of mortars, the nitrate ions were alsocreated but the large adsorption capacities of thematerial limited, orat least, retarded the deactivation of the photocatalysis reaction.

On glass, although a slight decrease of the NO degradation rateoccurred, the decrease of the NOx degradation rate was mainly duetoNO2generation. Bothphenomenawere certainly the consequenceof the competition for adsorptive sites between NO and NO2 on theone hand and the finally formed nitrate ions NO3

� on the other.

(right) substrates (initial NO concentration ¼ 400 ppb, Q ¼ 1.5 l min�1, H ¼ 6 g kg�1).

Fig. 10. Effect of water vapour on the degradation rate of NO2 on glass substrate underdifferent initial NO concentrations (Q ¼ 1.5 l min�1).

Fig. 8. Effect of water vapour on the degradation rate of NO on mortar substrate underdifferent initial NO concentrations (Q ¼ 1.5 l min�1).

T. Martinez et al. / Building and Environment 46 (2011) 1808e1816 1813

3.4. Effect of humidity

Humidity has a complex influence on the photocatalytic reac-tion. Different behaviour was observed depending on the experi-mental conditions and the materials used. Maggos et al. [11] and Aoand Lee [19] observed a decrease of photocatalytic efficiency withincreasing humidity at an initial NO concentration of 220 and200 ppb. At ppm level, Husken et al. [20] observed a linear decreaseof the degradation rates of NO with increasing humidity in the10e70% RH range. In contrast, with the same experimental appa-ratus but at an initial NO concentration of 500 ppb, Yu andBrouwers noted an improvement of photocatalytic performance onNOwith moisture [8]. The latter phenomenonwas also observed byDevahasdin et al. with a NO concentration of 40 ppm [5].

The generally accepted hypothesis explaining the decrease ofphotocatalytic efficiency with the increase of relative humidity isthe competition between water and pollutant on the adsorptionsites. Conversely, humidity can be regarded as a reactant in thephotocatalytic reaction because it allows the creation of hydroxylradicals. According to these phenomena, the conversion rate can bedependant on both humidity and initial pollutant concentration asshown by Obee and Brown for the photocatalytic oxidation oftoluene, formaldehyde and 1,3-butadiene [21].

Fig. 8 shows how the NO degradation rate measured on mortarsubstrate depends on humidity (0e14.5 g kg�1 matching 0-74 HR %at 25 �C) and on NO concentration (400, 1000, 1500 and 2000 ppb).At initial NO concentration of 400 and 1000 ppb, no significantinfluence of humidity was observed on the NO degradation rate.The experiments conducted at 1500 and 2000 ppb showed

Fig. 9. Effect of water vapour on the degradation rate of NO on glass substrate underdifferent initial NO concentrations (Q ¼ 1.5 l min�1).

a significant decrease in the degradation rates with decreasinghumidity. The major trend was an increase of the positive slope ofthe NO degradation rate with the increase of initial NO concen-tration at humidity levels <4 g kg�1. For humidity levels higherthan 4 g kg�1, the degradation rates for initial NO concentrations of1500 and 2000 ppb were constant. This leads us think that, for theexperiments conducted on mortar substrate, (i) the photocatalyticreaction is not limited by the competition with water on the pho-toactive sites at high humidity rates, and (ii) a lack of oxygenreactive species due to insufficient humidity in the air hinders thephotocatalytic degradation of NO at high pollutant concentration.

The influence of initial NO concentration and humidity on theNO degradation rate is similar on glass and on mortar substrate(Figs. 8 and 9). Humidity had no significant influence on NOdegradation rates when the test was conducted at initial NOconcentrations of 400 and 1000 ppb. For higher pollutant levels, anoptimal humidity level for the degradation of NO seemed to appeararound 6 g kg�1 on glass substrates. This indicates that, for lowhumidity levels, the photocatalytic reaction is limited by a lack ofradicals whereas, for high humidity levels, the NO degradationrates decrease because of the competition between humidity andpollutant. Onmortar, no optimal humidity level could be identified,certainly because the excess water was adsorbed by the substratesand did not compete with pollutant on the adsorptive site. More-over, on glass, the decrease in NOx degradation rate in the presenceof water vapour was mainly due to a competition between the by-product NO2 and water on adsorptive sites (Fig. 10 and Fig. 11). Theconsequence was the presence of a large amount of NO2 in thesampled gas.

Fig. 11. Effect of water vapour on the degradation rate of NOx on glass substrate underdifferent initial NO concentrations (Q ¼ 1.5 l min�1).

Fig. 12. Measurement of NO and NO2 concentration in nitrogen and air under dry conditions (mortar substrate, initial NO concentration ¼ 400 ppb, Q ¼ 1.5 l min�1).

T. Martinez et al. / Building and Environment 46 (2011) 1808e18161814

The enhancement of the photocatalytic degradation bysuppression of the generated NO2 when the coating was applied tomortar was similar to the observations reported by Ao et al. whocompared NO degradation (concentration level of 200 ppb) by TiO2coated on a glass fibre filter and on an activated carbon filter [19].Moreover, in the experiment by Ohko et al. investigating the pho-tocatalytic degradation of NO by TiO2 thin films prepared on glasspyrex substrates, the NO2 generated was equivalent to the decreaseof NO [25].

Moreover, whatever the initial NO concentration, the presenceof humidity in the air does not seem to be necessary for the pho-tocatalytic reaction of NO. To study the influence of oxygen in air,the air stream was punctually replaced by nitrogen stream duringan abatement test conducted on mortar substrate. In nitrogen, thephotocatalytic decomposition of NO can be observed but a rapiddeactivation occurs (Fig. 12). Such behaviour can be explained bya residual hydroxylation of the photocatalytic surface at thebeginning of the test. When the nitrogen stream is replaced by dry

Fig. 13. Effect of initial NO concentration on degradation rate on mortar substrate(Q ¼ 1.5 l min�1, H ¼ 6 g kg�1).

air, superoxide radicals are created and the photocatalytic reactionreaches a steady state after 4 min. This observation is in accordancewith Laufs et al. [16] and suggests that the hydroxyl OH, is probablynot required for the photocatalytic conversion of NO (OH, is theproduct of the reaction between H2O and electron holes) and thatthe radical superoxide O2

�, is the main oxidant in the photo-catalytic reaction of NO.

3.5. Effect of initial NO concentration

The usual amount of NO in urban polluted air is of severalhundred ppb and it can reach 1 ppm. Experiments performed onflow-type photoreactors showed a significant decrease of thedegradation rate with increasing initial NO concentration [8,20].

An increase in the initial NO concentration led to a decrease inthe degradation rates on both mortar and glass substrates (Fig. 13and Fig. 14). Degradation rates of NO were very similar on the

Fig. 14. Effect of initial NO concentration on degradation rate on glass substrate(Q ¼ 1.5 l min�1, H ¼ 6 g kg�1).

Table 1Experimental parameters and concentrations of nitrate and nitrite (glass substrate, Q ¼ 1.5 l min-1, initial NO concentration ¼ 2040 ppb).

Humidity (g/kg) Time of UV illumination(hours)

NOx in(mmol)

NOx out(mmol)

NOx in e NOx out(mmol)

Reaction product exp (mmol) (Reaction product exp)/(NOx in e NOx out)Total

(NO3� þ NO2

�)NO3

� NO2�

Control 0 0 0 0 0 0 0Control 0 0 0 0 0 0 00 4 33.06 26.18 6.88 6.21 6.04 0.17 0.900 2 16.53 12.07 4.46 4.28 4.18 0.10 0.970 1.06 8.88 5.28 3.60 3.32 3.32 0 0.920 1.03 8.61 6.12 2.49 2.25 2.22 0.03 0.900 1 8.33 6.64 1.69 1.63 1.48 0.16 0.970 0.5 4.51 2.77 1.74 1.63 1.63 0 0.896 2 16.53 13.94 2.59 2.16 2.16 0 0.846 1 8.33 6.53 1,81 1,29 1,29 0 0.716 1 8.33 6.59 1.74 1.61 1.59 0.02 0.936 1 8.33 5.97 2.36 2.27 2.25 0.03 0.966 0.56 4.64 3.94 0.7 0.54 0.54 0 0.7712 1.03 8.61 6.46 2.15 1.58 1.53 0.05 0.7312 1 8.33 6.47 1.86 1.66 1.64 0.02 0.9012 1.03 8.61 6.48 2.12 1.97 1.93 0.04 0.93

T. Martinez et al. / Building and Environment 46 (2011) 1808e1816 1815

two substrates. However, the influence of initial NO concentrationwas greater on glass substrate, with a larger amount of non-adsorbed generated NO2.

3.6. Nitrate analysis

What happens to the NO that is degraded by the photocatalyticcoating is an interesting question. According to the most wide-spread hypothesis [5], this part of the polluting gas is transformedinto nitrite and nitrate ions. In order to assess the amount ofproduct adsorbed on the test piece, nitrite and nitrate ions weretitrated by ion chromatography.

The following conditions were set up: glass substrates, 2040 ppbof NOx, flow rate 1.5 l min�1, 0.5e4 h of illumination, humidity0e12 g kg�1. Table 1 shows the experimental results.

When no photocatalytic reaction was carried out (no artificialUV light, kept in room), no nitrite or nitrate ions were detected(control samples). According to other studies, traces of nitrite(NO2

�) were detected on samples after the photocatalytic reactionof NO [5,25,32] whereas the amount of nitrate (NO3

�) was of thesame order of magnitude as the difference between the inletnumber of moles of NOx and the outlet number of moles of NOxdetected by the analyser (molar values are obtained by integrationof the concentration measurement on a minute basis in the UVillumination time). However, the ratio between the theoreticalnitrogen ion recovery and the experimental measurement wasbetween 0.71 and 0.97. No influence of the humidity or the UVillumination time on this ratio was observed. A release in the gasphase of undetected products like HONO, or the presence of othertrapped species, could explain the difference between the observedvalues and the theoretical value of 1.

4. Conclusion

This article focuses on the abatement of NOx in air by hetero-geneous photocatalytic oxidation using TiO2 mixed in coatings. Anexperimental set-up suitable for assessing the de-NOx abilities oflab-produced coatings was used. The parameters that affected thephotocatalytic degradation rate of NOwere investigated. Onmortarand glass substrates, the influence of increasing humidity ondegradation rates was dependent on the substrate and on the initialNO concentration. When the test was conducted at an initial NOconcentrations of 400 and 1000 ppb, no significant influence ofhumidity was observed. On the other hand, a significant decrease of

the degradation rates with the decrease of the humidity wasobserved at higher initial NO concentration (1500 and 2000 ppb).The beneficial impact of the presence of humidity at high initial NOconcentration indicates that the photocatalytic reaction is notlimited by the competition between water and pollutant on thephotoactive sites. Conversely, at low humidity rates, the lack ofphotogenerated oxygen radicals hinders the photocatalytic degra-dation of NO. On mortar, the generation of the by-product NO2 wasvery low, probably because of the good adsorption capacities of thesupporting materials. On glass, NOx degradation rates decreasedstrongly by generation of NO2, probably due to competitionbetween pollutant and humidity on the adsorptive site.

Moreover, increasing initial NO concentration led to lowerdegradation rates on both substrates. Additionally, the magnitudeof the influence of initial NO concentration was very high on glasssubstrate with low degradation rates. Such behaviour may beexplained by the adsorption capacities of the substrate. It wasassumed that humidity and pollutant were in competition on theactive surface of the photocatalyst sites only on materials with lowadsorption capacities, such as glass. Photocatalytic performancecould be improved by applying the photocatalyst to materials withhigh adsorption capacity, such as cementitious materials.

Optimal contents of photocatalyst and binder were difficult todetermine since performance decreased because of the covering ofTiO2 by the binder or because of an insufficient amount of photo-catalyst. Moreover, the formulation taskwas also governed by otherfactors, such as adhesion and durability. Special attention should bepaid to this last property in the case of photocatalytic coatings as theoxidative character of photocatalysis can degrade polymeric binder.In this context, laboratory experiments are useful to compare and tooptimize the air purification performance of photocatalytic mate-rials. An accelerated weathering test can be used to assess thedurability of the photocatalytic coating with the simulation ofsunlight and rain. This will be the purpose of a further study.

Considering the results described above, the degradation of NOx

through TiO2 nanoparticles incorporated in a polymer-matrix-based coating is suitable for outdoor applications, at least as far asshort term performance is considered. However, quantification ofthe air purification properties and the influence of the ageing of theproduct must also be determined at real scale. The controlledatmosphere of the lab scale test may have resulted in an over-estimation of the capacities of the photocatalytic material for airpurification [12] notably because the contact between the pollutantand the reactive surface was optimized. Moreover, in a real

T. Martinez et al. / Building and Environment 46 (2011) 1808e18161816

atmosphere, a multitude of pollutants coexist and can compete onthe adsorptive sites of the photocatalytic surface. Organic pollut-ants such as BTEX can have an inhibiting effect on the capacities ofthe photocatalytic coating to degrade NOx [33]. Moreover, a moni-toring of the photocatalytic efficiency in a real atmosphere couldhelp to quantify the durability of the photocatalytic material. Theseaspects will be investigated in further steps of the study.

Acknowledgements

The authors thank the ANRT (French Association for Researchand Technology), OSEO (French organization supporting researchand development) and the Guard Industrie company for theirinterest and financial support.

References

[1] Lawrence AJ, Masih A, Taneja A. Indoor/outdoor relationships of carbonmonoxide and oxides of nitrogen in domestic homes with roadside, urban andrural locations in a central Indian region. Indoor Air 2005;15:76e82.

[2] Blondeau P, Iordache V, Poupard O, Genin D, Allard F. Relationship betweenoutdoor and indoor air quality in eight French schools. IndoorAir 2005;15:2e12.

[3] Hoffmann MR, Martin ST, Choi W, Bahnemann DW. Environmental applica-tions of semiconductor photocatalysis. Chemical Reviews 1995;95:69e96.

[4] Blake D. Bibliography of work on the photocatalytic removal of hazardouscompounds from water and air. Available on line at, http://www.nrel.gov/docs/fy02osti/31319.pdf; 2001.

[5] Devahasdin S, Fan C, Li K, Chen DH. TiO2 photocatalytic oxidation of nitricoxide: transient behavior and reaction kinetics. Journal of Photochemistry andPhotobiology A: Chemistry 2003;156:161e70.

[6] Hashimoto K, Wasada K, Toukai N, Kominami H, Kera Y. Chemistry: photo-catalytic oxidation of nitrogen monoxide over titanium(IV) oxide nanocrystalslarge size areas. Journal of Photochemistry and Photobiology A 2000;136:103e9.

[7] Dalton JS, Janes PA, Jones NG, Nicholson JA, Hallam KR, Allen GC. Photo-catalytic oxidation of NOx gases using TiO2: a surface spectroscopic approach.Environmental Pollution 2002;120:415e22.

[8] Yu Q, Brouwers H. Indoor air purification using heterogeneous photocatalyticoxidation. Part I: experimental study. Applied Catalysis B: Environmental2009;92:454e61.

[9] Ai Z, Ho W, Lee S, Zhang L. Efficient photocatalytic removal of NO in indoor airwith hierarchical bismuth oxybromide nanoplate microspheres under visiblelight. Environmental Science & Technology 2009;43:4143e50.

[10] Dong F, Huang Y, Zou S, Liu J, Lee SC. Ultrasonic spray pyrolysis fabrication ofsolid and hollow PbWO4 spheres with structure-directed photocatalyticactivity. The Journal of Physical Chemistry C 2011;115:241e7.

[11] Maggos T, Bartzis J, Leva P, Kotzias D. Application of photocatalytic technologyfor NOx removal. Applied Physics A: Materials Science & Processing 2007;89:81e4.

[12] Maggos T, Bartzis J, Liakou M, Gobin C. Photocatalytic degradation of NOxgases using TiO2-containing paint: a real scale study. Journal of HazardousMaterials 2007;146:668e73.

[13] Salthammer T, Fuhrmann F. Photocatalytic surface reactions on indoor wallpaint. Environmental Science and Technology 2007;41:6573e8.

[14] Salthammer T, Fuhrmann F, Schulz N, Siwinski N. Removal of indoorcontaminants by photocatalytic reaction. CD-ROM: Lisboa, 4-8 June 2006.Espoo: ISIAQ. In: Oliveira Fernandes E, editor. de: HB 2006 healthy buildings.

Creating a healthy indoor environment for people. Proceedings; 2006. p.249e54.

[15] Allen NS, Edge M, Sandoval G, Verran J, Stratton J, Maltby J. Photocatalyticcoatings for environmental applications. Photochemistry and Photobiology2005;81:279e90.

[16] Laufs S, Burgeth G, Duttlinger W, Kurtenbach R, Maban M, Thomas C, et al.Conversion of nitrogen oxides on commercial photocatalytic dispersionpaints. Atmospheric Environment 2010;44:2341e9.

[17] ISO 22197-1:2007eFine ceramics (advanced ceramics, advanced technicalceramics) e Test method for air-purification performance of semiconductingphotocatalytic materials e Part 1: Removal of nitric oxide; 2007.

[18] XP B44e011 Photocatalyse eMéthode d’essai pour l’évaluation des matériauxphotocatalytiques vis-à-vis de la dégradation des NOx e Méthode à un seulpassage en mode tangentiel. French standard; 2009.

[19] Ao CH, Lee SC. Enhancement effect of TiO2 immobilized on activated carbonfilter for the photodegradation of pollutants at typical indoor air level. AppliedCatalysis B: Environmental 2003;44:191e205.

[20] Hüsken G, Hunger M, Brouwers H. Experimental study of photocatalyticconcrete products for air purification. Building and Environment 2009;44:2463e74.

[21] Obee TN, Brown RT. TiO2 photocatalysis for indoor air applications: effects ofhumidity and trace contaminant levels on the oxidation rates of formalde-hyde, toluene, and 1,3-Butadiene. Environmental Science and Technology1995;29:1223e31.

[22] Allen NS, Edge M, Verran J, Stratton J, Maltby J, Bygott C. Photocatalytic titaniabased surfaces: environmental benefits. Polymer Degradation and Stability2008;93:1632e46.

[23] Aprile C, Corma A, Garcia H. Enhancement of the photocatalytic activity ofTiO2 through spatial structuring and particle size control: from subnanometricto submillimetric length scale. Physical Chemistry Chemical Physics 2008;10:769e83.

[24] Ao CH, Lee SC. Combination effect of activated carbon with TiO2 for thephotodegradation of binary pollutants at typical indoor air level. Journal ofPhotochemistry and Photobiology A: Chemistry 2004;161:131e40.

[25] Ohko Y, Nakamura Y, Negishi N, Matsuzawa S, Takeuchi K. Unexpected releaseof HNO3 and related species from UV-illuminated TiO2 surface into air inphotocatalytic oxidation of NO2. Environmental Chemistry Letters 2009;8:289e94.

[26] NF EN 196e1, Méthodes d’essais des ciments e Partie 1: détermination desrésistances mécaniques. French Standard; 2006.

[27] Herrmann J. Heterogeneous photocatalysis: state of the art and presentapplications. Topics in Catalysis 2005;34:49e65. In honor of Pr. R.L. Burwell Jr.(1912e2003), Former Head of Ipatieff Laboratories, Northwestern University,Evanston (Ill).

[28] Fujishima A, Zhang X. Titanium dioxide photocatalysis: present situation andfuture approaches. Comptes Rendus Chimie 2006;9:750e60.

[29] Blöß S, Elfenthal L. Doped titanium dioxide as a photocatalyst for UV andvisiblelight. In: Baglioni P, Cassar L, editors. Proceedings international RILEMsymposium on photocatalysis, environment and construction materials.Florence, Italy, Bagneux, France: RILEM Publications; 2007. p. 187e94.

[30] Herrmann J. Photocatalysis fundamentals revisited to avoid several miscon-ceptions. Applied Catalysis B: Environmental 2010;99:461e8.

[31] Toma FL, Bertrand G, Klein D, Coddet C. Photocatalytic removal of nitrogenoxides via titanium dioxide. Environmental Chemistry Letters 2004;2:117e21.

[32] Ohko Y, Nakamura Y, Fukuda A, Matsuzawa S, Takeuchi K. Photocatalyticoxidation of nitrogen dioxide with TiO2 thin films under continuous UV-Lightillumination. The Journal of Physical Chemistry C 2008;112:10502e8.

[33] Ao CH, Lee SC, Mak CL, Chan LY. Photodegradation of volatile organiccompounds (VOCs) and NO for indoor air purification using TiO2: promotionversus inhibition effect of NO. Applied Catalysis B: Environmental 2003;42:119e29.