Influence of dust and oil accumulation on effectiveness of photocatalytic concrete surfaces

1 Influence of dust and oil accumulation on the effectiveness of 1 photocatalytic concrete surfaces 2 3 Miren Etxeberriab*#, Ming-Zhi Guoa#, Anibal Maury-Ramirezc, Chi Sun Poona 4 5 Assoc. Prof. Miren Etxeberria*# 6 b Department of Civil and Enviromental Engineering, Jordi Girona 1-3, Universitat 7 Politecnica de Catalunya.BarcelonaTECH, 08034 Barcelona, Spain. 8 E-mail: [email protected]du 9 10 MSc. PhD. Candidate Ming-Zhi Guo# 11 a Department of Civil and Environmental Engineering (CEE), The Hong Kong 12 Polytechnic University, Hung Hom, Kowloon, Hong Kong. 13 E-mail: [email protected]u.hk 14 15 Assoc. Prof. Anibal Maury-Ramireza,c 16 c Department of Civil and Industrial Engineering, Pontificia Universidad Javeriana Cali, 17 Calle 18 # 118-250, Av. Cañas Gordas, Cali, Colombia. 18 E-mail: [email protected] 19 20 Prof. Chi Sun Poona 21 a Department of Civil and Environmental Engineering (CEE), The Hong Kong 22 Polytechnic University, Hung Hom, Kowloon, Hong Kong. 23 E-mail: [email protected]* 24 25 *Corresponding author 26 #These authors contributed equally to the paper. 27 28 Abstract 29 The day to day negative effect of contaminating substances on the NOx removal 30 effectiveness of photocatalytic surfaces and their recovery capacity after undergoing 31 cleansing is of real academic and practical interest. This study deals with the analysis of 32 the NOx removal effectiveness of two different types of photocatalytic concrete surface 33 layers incorporating nano-TiO2 particles. Both types of surfaces were examined before 34 and after being subjected to dust accumulation and oil impregnation. In addition, the 35 recovery of their NOx removal capacity after undergoing various cleansing processes 36 was also evaluated. The results were compared to those of their respective reference 37 samples. The results showed that the high NOx removal capacity of the spray coated 38 samples was partially maintained even after a high concentration of dust accumulation. 39 However, the water cleansing process employed was sufficient enough to recover the 40 Manuscript Click here to download Manuscript Manuscript_last_final.doc 2 lost NOx removal capacity. It was also discovered that the nano-TiO2 particles lost 41 nearly all their photocatalytic capacity after being subjected to oil impregnation, and 42 that the cleansing processes, either by an alkaline detergent or using a n-hexane solvent, 43 failed to recover the initial NOx removal capacity. 44 45 Keywords: concrete surface; nano-TiO2; photocatalytic, NOx removal; dust 46 accumulation; oil impregnation 47 48 49 1. INTRODUCTION 50 The use of photocatalytic building materials to reduce atmospheric pollutants, which 51 have a negative effect on health, particularly amongst children and the elderly, is 52 becoming increasingly more necessary. Vehicle traffic emissions, as well as 53 contaminants caused by housing and industry, are the main sources of atmospheric 54 pollution in dense urban areas. Many organic compounds and air pollutants including 55 nitrogen oxides and sulfur dioxide can be decomposed by ultraviolet (UV) radiation. 56 However, this process is extremely slow. Photocatalytic materials such as nano-titanium 57 dioxide (TiO2), in the presence of UV light (sunlight), can accelerate this process 58 through the decomposition of organic and inorganic substances, including air pollutants 59 such as NOx, SO2, and VOC’s (Hodgson et al., 2007; Hassan, 2009). 60 Over the last twenty years there have been numerous studies dealing with ways of 61 improving the quality of urban air through the incorporation of nano-TiO2 into concrete 62 block production (Murata et al., 1999; Poon and Cheung, 2007; Chen and Poon, 2009a; 63 Chen and Poon, 2009b; Hüsken et al., 2009; Ballari et al., 2010; Guo and Poon, 2013; 64 Ballari et al., 2011) as well as the manufacture of surface treated stone and rendering 65 materials (Franzoni et al., 2014; Munafo et al. 2015). However, as yet there has been no 66 in depth evaluation of the durability of the nano-TiO2 surface coating. According to 67 Hassan et al. (2010), there was a small decrease in the NO removal efficiency during the 68 wearing down of the samples that had 5% TiO2 incorporation. Results of the SEM-EDS 69 analysis also confirmed that when compared with the original samples the relative 70 concentration of Ti on the worn out specimens did not substantially change. However, 71 Guo et al. (2013) found that TiO2-dip-coated mortars almost totally lost the bactericidal 72 ability after being subjected to a harsh weathering process. 73 3 Olabarrieta et al. (2012) presented an analysis of the effect of chemical and physical 74 transformations associated with water flow on two intrinsically different photocatalytic 75 coatings. Their work highlighted the deactivation of the active sites of TiO2 upon 76 prolonged immersion in water, which was found to reduce the performance of the 77 photocatalytic materials. They hypothesized that sodium chloride and UV light 78 somehow modified the TiO2–TiO2 and TiO2–glass substrate interaction, resulting in 79 facilitating the release of the TiO2 nanoparticles into the surrounding water. According 80 to De Melo et al. (2012), external agents such as dirt and abrasion are determinant in the 81 efficiency loss of the NOx removal capacity on the concrete block surfaces. Rainwater 82 or the simple water cleansing (without pressure) of the block surfaces is not sufficient to 83 ensure recovery of the photocatalytic activity, and thus periodic maintenance cleansing 84 of the surface is required by means of employing a water jet. 85 In this research work, the influence of dust and oil accumulation on the NOx removal 86 capacity of the photocatalytic nano-TiO2 based concrete surfaces as well as their 87 recovery to NOx removal after cleaning process are presented. Two separate types of 88 photocatalytic concrete surface layers were produced. One set of samples was produced 89 using TiO2 as an addition (in 5% by cementitious materials weight), whereas, the other 90 set was spray coated with a nano-TiO2 solution. The effectiveness of the NOx removal 91 capacity of the two types of samples before and after being subjected to dust 92 accumulation and oil impregnation was determined. The recovery of their respective 93 NOx removal capacity after treatment by a myriad of cleansing processes was 94 subsequently evaluated. 95 96 2. MATERIALS AND SAMPLE PREPARATION 97 2.1 Materials 98 A commercially available nano-TiO2 powder (P25, Degussa) was used as the 99 photocatalyst in all the experiments. The particle size of the TiO2 was 20-50 nm, with a 100 specific BET surface area of 50±15 m2·g-1. ASTM Type I Ordinary Portland cement 101 (OPC, Green Island Cement Limited, Hong Kong) and fly ash (FA) were used as the 102 cementitious materials. FA was added as a pozzolana to control the alkali-silica reaction 103 (ASR) between the silica-rich glass cullet (used as aggregate in preparing the photo- 104 catalytic cement mortar) and the alkali in the cement. Their physical and chemical 105 properties are listed in Table 1. The nano-TiO2-based concrete surface layers were 106 prepared by employing fine aggregates obtained from crushed recycled glass (RG) 107 4 cullet derived from post-consumer beverage glass bottles. The post-consumer beverage 108 glass used was sourced from a local Hong Kong based eco-construction material 109 company. Prior to their experimental use, all the discarded glass bottles were washed 110 and then crushed by a mechanical crusher before being finally sieved to obtain the 111 desired particle size. The particle size distribution of the RG used in the experiment is 112 shown in Table 2. 113 114 2.2 Mix proportions and production process 115 The mix proportion (by weight) for the two types of concrete surface layers was 116 0.75:0.25:3.0:0.3 (OPC: FA: RG: water). 117 The procedures for the preparation of the concrete surface layers were as follows. First, 118 all the proportioned materials were mixed uniformly for approximately 5 minutes using 119 a mechanical mixer (nano-TiO2 particles were added in the Intermixed-type 1 mixture). 120 Then, the steel moulds (an internal dimension of 200×100×5 mm) were filled to 121 maximum capacity (about 250 g) and the resulting mixture was then hand compacted 122 before undergoing two further mechanical compacting stages of 500 kN and 600 kN at a 123 rate of 500 kN·min-1 124 The TiO2 sprayed coated samples (SP-Type 2) were produced via the use of a 125 suspension of methanol and P25 (30 g·L-1) which was sprayed on to the mortar surface. 126 A detailed explanation of the procedure employed is as follows. The mortar mixtures 127 (without TiO2) were first placed in the prepared steel moulds where they were hand 128 compacted before being spray coated 10 times with the TiO2-solution. The mixture was 129 then subjected to mechanical compacting similar to that previously mentioned, before 130 undergoing a further 10 spray coatings. All spray coating was carried out within the first 131 20 minutes after the surface layer material was introduced into the moulds. 132 The prepared surface layers were removed 24 hours later from their moulds and then 133 tightly wrapped in plastic film during curing until further testing. 134 135 2.3 Properties of concrete surface layers 136 The physical and mechanical properties of the two types of concrete surface layers 137 produced (intermixed-Type 1 and SP-Type 2) were determined using cubic test 138 specimens of 70×70×70 mm according to ASTM C642-06. The specimens were 139 produced by compacting the concrete specimens in three layers. The first two layers 140 were compacted manually using a wooden rod. The third layer was compacted 141 5 mechanically via a press in the same manner as described in Section 2.2. Table 3 shows 142 the properties of intermixed-Type 1 and SP-Type 2 samples. It is well-known that the 143 use of nano-TiO2 particles can have an impact on reducing the fluidity of the 144 intermixed-Type 1 mixture and consequently reducing also the homogeneity of the 145 prepared samples, thus leading to slightly reduced compressive strengths, a fact also 146 detected by other researchers (Lucas et al., 2013; Jimenez-Relinque et al., 2015). 147 148 3. METHODOLOGY 149 3.1 Photocatalytic conversion of NOx 150 The photocalytic conversion of NOx was tested through the use of a continuous flow 151 reactor which was constructed in the laboratory in accordance with JIS R1701-1 152 specifications. The reactor (length: 300 mm, width: 150 mm and height: 100 mm) was 153 completely sealed and free of all leakage. The test samples were placed on a rack in the 154 centre of the reactor. A schematic diagram of the experimental setup is shown in Figure 155 1. Detailed information on the reactor and the required equipment employed to carry out 156 the photocatalytic NOx conversion test can be found in our previous studies (Guo et al., 157 2015). The entire experiment was carried out at the ambient temperature of 25±3 ℃. 158 The flow of the testing gas (1000 ppb NO) was adjusted by two flow controllers at a 159 rate of 3 L·min-1 and the relative humidity (RH) was controlled at 30±5%. The UV 160 intensity at the centre of the reactor was 2 W·m-2. All photocatalytic conversion 161 processes were carried out by first introducing the testing gas stream into the reactor in 162 the absence of UV radiation for a minimum of half an hour. The reason for this was to 163 ensure that both the desired RH and the gas–solid adsorption–desorption equilibrium 164 was reached, after which the UV lamps were turned on for the photocatalytic process to 165 begin. The NO removal test took 1 hour per sample, during this time period the 166 concentration changes of NO and NO2 at the outlet were recorded. Every sample was 167 tested three times and the average value together with the standard deviation was noted. 168 The details of the calculation of the amount of NOx removal have been described 169 previously (Guo et al., 2015) and are expressed as a subtraction of the NO2 generated 170 from the NO removed. The calculation of the amount of NOx removal, following the 171 instructions detailed in JIS R 1701-1, is shown below: 172 173 6 Eq. 1 174 where QNOx (μmol·m-2·h-1) is the amount of nitric oxides removed by the test sample, 175 [NO]0 and [NO2]0 (ppm) are the inlet concentration of nitrogen monoxide and nitrogen 176 dioxide, respectively, [NO] and [NO2] (ppm) are the outlet concentration of nitrogen 177 monoxide and nitrogen dioxide, respectively, t (min) is the time of removal operation, f 178 (L·min-1) is the flow rate converted into that at the standard state (0 ℃, 1.013 kPa), A 179 (m2) is the surface area of cement paste samples, T (1 h for all experiments) is the 180 duration of the photocatalytic process, and 22.4 represents the volume of 1 mole ideal 181 gas at the standard state, which is 22.4 L (ideal gas law). 182 183 3.2 Worsening and recovery of photocatalytic activities 184 The effectiveness of the NOx removal capacity of the intermixed-Type 1 and SP-Type 2 185 samples after the surfaces were subjected to dust accumulation and oil impregnation 186 was determined. After which, the samples were subjected to several cleansing processes, 187 and the recovery of their NOx removal capacity was then evaluated. 188 189 190 3.2.1 Dust accumulation process and its assessment 191 Dust characterization 192 A fine dust (ISO 12103-1 A2) provided by Powder Technology Inc. was employed in 193 order to simulate the effect of natural dust. Table 4 describes the grading size 194 distribution of those dust particles (the data was provided by Powder Technology Inc.). 195 The specific weight of the dust was 2.65 g·cm-3, and SiO2 and Al2O3 were the main 196 chemical components. The compositions of the dust (by weight) are given in Table 5. 197 198 Dust accumulation and cleansing process 199 The photocatalytic effectiveness of concrete surface layers could be affected by dust 200 accumulation on their surface. Although it is not easy to simulate the real situation 201 during application in the laboratory, in this work a chamber designed and constructed 202 for that purpose was used. The test chamber (1×0.6×0.6 m) was made with ESD acrylic 203 plates which were both transparent and antistatic. The test dust was injected by the dust 204 generator (RBG 1000, PALAS) into the test chamber and mixed with air via means of a 205 7 fan. Dust was deposited slowly and naturally on the surface of the samples, which were 206 positioned in the centre of the chamber (Figure 2). 207 The amounts of dust accumulated on both the surfaces of the intermixed-Type 1 and 208 SP-Type 2 samples were 0, 10, 18 and 37 gr·m-2. Figure 3a shows several samples 209 subjected to the dust accumulation process. Figure 3b shows, the samples (1) and (2) 210 with 10 and 37 gr·m-2 of dust accumulation on the surface, respectively. The NOx 211 removal capacity of the intermixed-Type 1 and SP-Type 2 samples with different 212 weight of dust accumulation on their surfaces was immediately evaluated. Then, the 213 cleansing processes were carried out in order to evaluate the recovery of the NOx 214 removal ability of the samples. Two cleansing processes were applied on each sample. 215 First, slow flowing tap water, with an approximately flow rate of 6.25 cm3/s, was 216 applied to the surface of the samples. Second, besides the water cleansing (under the 217 same conditions), an electric brush was also employed to clean the samples’ surfaces. 218 Immediately after the cleansing process, all the samples were oven-dried (60 ℃ for 24 h) 219 until NOx removal testing. The photocatalytic activity of each sample was evaluated and 220 the obtained results were compared to those of the reference samples (without dust 221 accumulation). 222 223 3.2.2 Oil accumulation process and its assessment 224 Oil accumulation and cleansing process 225 A synthetic motor oil of 15W-40 viscosity was employed to impregnate the surfaces of 226 the intermixed-Type 1 and SP-Type 2 samples. The surfaces of the concrete layers were 227 impregnated by separately using two types of oil-solvent mixtures. Firstly, a low 228 viscosity oil-solvent mixture consisting of 50% weight of oil and 50% weight of ethanol 229 (with ratio of 1:1) was prepared. This low viscosity mixture was then applied on the 230 samples in a dose of 200 g per m2 of concrete surface. Secondly, 400 g of undiluted 231 motor oil per m2 of concrete surface was also used to contaminate the surface of the 232 samples. A 3 mL pipette dropper was employed, the oil-solvents being distributed drop 233 by drop in order to guarantee their homogeneous accumulation on the surface. 5 pipettes 234 were employed to disperse all the oil-solvents on each concrete sample over a fixed area 235 of 20×10 cm2. The capacity of degradation of NOx via the intermixed-Type 1 and SP- 236 Type 2 samples after oil accumulation was determined. Several cleansing processes 237 were used in order to remove the oil from the samples. The oil contaminated samples 238 were separately cleaned by adding two oil removing products to the surfaces; an 239 8 alkaline industrial detergent and an organic solvent (n-hexane). The spray SP-Type 2 240 samples were cleaned by means of adding the n-hexane solvent to the surface as well as 241 submerging the samples in hexane for another 12 hours (named as cleaned–hexane-2). 242 243 4. RESULTS 244 4.1 Photocatalytic NOx removal of intermixed-Type 1 and spray SP-Type 2 245 samples 246 Figure 4 illustrates the NOx concentration profiles when the reference intermixed-Type 247 1 and SP-Type 2 samples after undergoing the photocatalytic NOx removal test. The SP- 248 Type 2 samples showed a greater reduction of NOx capacity. This finding proved to be 249 consistent with our previous work (Guo et al., n.d. ). The elimination rate of NOx by the 250 intermixed-Type 1 and SP-Type 2 samples were 136.5 µmol·h-1·m-2 and 255.0 µmol·h- 251 1·m-2, respectively. Thus, the SP-Type 2 sample proved to be 80% more effective than 252 the intermixed-Type 1 samples. This far superior NOx photocatalytic removal 253 performance of the SP-Type 2 samples can be attributed to a much higher amount of 254 TiO2 particles on the sample’s surface (supported by the following SEM-EDX analysis). 255 As a result, most of the TiO2 surface particles were able to gain easy access to the UV 256 light irradiation. In contrast, a high percentage of nano-TiO2 surface particles of the 257 intermixed-Type 1 samples were completely enveloped by the hydration products, and 258 hence were not in direct contact with the UV-light, as was previously described (Guo et 259 al., 2013). 260 SEM and EDX were employed to analyse the two types of samples. Figure 5a and 5b 261 clearly illustrate the morphological differences of the intermixed-Type1 and SP-Type2 262 samples studied, respectively. And the EDX analysis revealed that the percentage of 263 titanium element presented on the surface of intermixed-Type 1 and SP-type 2 samples 264 was 3.3% and 26.4%, respectively. Figure 5a and 5b also illustrate that some nano-TiO2 265 particles in the intermixed-Type 1 samples were completely enveloped by the 266 surrounding cement hydrated products, which more than likely impaired the NOx 267 removal capacity. Whereas, most of the spray-coated TiO2 particles were free from the 268 accumulation of the hydration products, and thus could be fully exposed to UV light. 269 Therefore, the SEM-EDX results are well validated by the aforementioned explanations 270 with respect to the observed differences in the photocatalytic NOx removal. 271 272 9 4.2 Influence of dust accumulation on NOx removal capacity and recovery by two 273 cleansing processes 274 4.2.1 Influence of dust accumulation 275 Figure 6 shows the NOx removal profiles of the intermixed-Type 1 and SP-Type 2 276 samples after being subjected to different amounts of dust accumulation on the surface. 277 In general, the samples with a higher volume of accumulated dust displayed less ability 278 to remove NOx, regardless of the sample types. For example, the 37 g·m-2 of dust 279 accumulation significantly reduced the NOx removal ability of both types of samples 280 tested (from 136.5 to 39.5 µmol·h-1·m-2 for intermixed-Type 1 and from 254.9 to 148.7 281 µmol·h-1·m-2 for SP-Type 2 samples, respectively). However, it must be pointed out that 282 even after 37 g·m-2 of dust accumulation, the SP-Type 2 samples were still able to 283 deliver a high photocatalytic NOx removal ability. This can be explained by the fact that 284 the dust accumulated on the surface was not densely compacted (no external force was 285 applied to the dust covered surface), thus the light blocking effect was not so 286 significant. Moreover, taking into consideration that the size of nano-TiO2 particles used 287 in this study was far smaller than that of the dust particles (see Table 4), the UV light 288 was still able to penetrate through those dust particles to reach the TiO2 particles 289 underneath, some of which were still able to produce the photocatalytic activity. 290 More interestingly, it was clearly observed that the dust accumulation noted had less 291 influence on the SP-Type 2 samples than on those of the intermixed-Type 1 samples. 292 This fact was consistent with the previous NOx removal results of the reference samples 293 (section 4.1). For example (see Figure 6), the SP-Type 2 samples still retained a 294 relatively higher capacity of NOx removal after exposure to 37 g·m-2 of dust 295 accumulation, (above 145 µmol·h-1·m-2), which proved to be even higher than that of 296 the intermixed-Type 1 samples without any dust accumulation (136.5 µmol·h-1·m-2). 297 298 4.2.2 Rejuvenation of NOx removal by the cleansing processes 299 Figure 7 shows the surface appearance of the samples after they were subjected to 300 different cleansing processes. The surfaces of all the intermixed-Type 1 samples were 301 similar in appearance to those of the reference sample after the application of the two 302 different cleansing processes. However, for the SP-Type 2 samples, the intensity of the 303 white colour on the surface (an indication of the TiO2 amount) was slightly decreased 304 after the cleansing process due to the washing away of some nano-TiO2 particles from 305 the surface. 306 1 LIST OF TABLES Table 1. Chemical compositions and physical properties of OPC and FA Table 2. Particle size distributions of recycled glass cullet Table 3. Properties of intermixed-Type 1 and SP-Type 2 concrete surface layers after 28 days curing Table 4. Volume dust Particle distribution Table 5. Chemical compositions of dust Table 1. Chemical compositions and physical properties of OPC and FA OPC FA Chemical composition SiO2 22.18 48.14 Al2O3 5.95 17.81 Fe2O3 3.37 11.63 CaO 62.46 12.66 MgO 1.56 4.59 K2O 0.48 1.26 Na2O 0.24 - TiO2 0.37 0.98 Sulphur content as SO3 2.25 2.64 Loss in ignition 0.64 3.90 Physical properties Specific surface (cm2 g-1) 3660 3960 Table 2. Particle size distributions of recycled glass cullet Sieve size (mm) Recycled glass (% passing) 5.0 99.8 2.36 90 1.18 57.9 0.6 32.1 0.3 10.6 0.15 0.075 7.5 4.7 Table 3. Properties of intermixed-Type 1 and SP-Type 2 concrete surface layers after 28 days curing Type of samples Dry density (g/cm3) DS Accessible pores (%) DS Compressive strength (MPa) DS Intermixed-Type 1 (5% TiO2 by weight of binder) 2.05 0.11 18.30 5.5 28.2 7.2 SP-Type 2 (without TiO2 addition) 2.11 0.05 14.62 3.5 35.7 6.2 Table Click here to download Table Tables_last.docx 2 Table 4. Volume dust Particle distribution Dust size (µm) 1 2 3 4 5 7 10 20 40 80 100 Vol(%) 3.5 9 9.5 7.5 6.5 10 8 20 17 9 0 Table 5. Chemical compositions of dust Component % in weight Component % in weight SiO2 68-76 CaO 2-5 Al2O3 10-15 MgO 1-2 Fe2O3 2-5 TiO2 0.5-1 Na2O 2-4 K2O 2-5 1 LIST OF FIGURES Figure 1. Equipment used. Schematic diagram of NOx removal experimental set-up Figure 2. Equipment used. Dust chamber Figure 3. a)Dust accumulation on the surface of samples in the chamber, b) (1) 10 g and (2) 37 g of dust were accumulated per m2 of surface Figure 4. Photocatalytic NOx removal profiles of intermixed-Type 1 and SP-Type 2 samples Figure 5. (a) and (b) SEM morphology of the intermixed-type 1 and SP-type 2 of concrete surface layers, respectively Figure 6. The photocatalytic NOx removal of samples with dust accumulation on its surface. Figure 7. Intermixed-Type 1 and SP-Type 2 samples cleaned by different processes Figure 8. Photocatalytic NOx removal of samples after subjected to cleansing with a) water and b) water and brush process. Figure 9. Influence of oil accumulation on photocatalytic NOx removal a) intermixed- Type 1 and b) SP-Type 2 samples Figure 10. a) Photocatalytic NOx removal of the samples impregnated with motor oil; b) the intermixed-Type 1 and SP-Type 2 samples cleaned by different cleansing methods. Figure 11. a) SEM-EDX analysis of SP-Type 2 reference sample before mobile oil contamination, (b) SEM-EDX analysis of the SP-Type 2 reference sample after exposure to mobile oil contamination Figure Click here to download Figure Figures_last_final.docx 2 Figure 1. Equipment used. Schematic diagram of NOx removal experimental set-up Figure 2. Equipment used. Dust chamber 3 (a) (b) Figure 3. a)Dust accumulation on the surface of samples in the chamber, b) (1) 10 g and (2) 37 g of dust were accumulated per m2 of surface Figure 4. Photocatalytic NOx removal profiles of intermixed-Type 1 and SP-Type 2 samples 0 200 400 600 800 1000 1200 010 20 30 40 50 60 70 NOx concentration (ppm) Time (min) Intermixed-Type 1 SP-Type2 (1) (2) 4 INTERMIXED-TYPE 1 reference sample (a) SP-TYPE 2 reference sample (b) Figure 5. (a) and (b) SEM morphology of the intermixed-type 1 and SP-type 2 of concrete surface layers, respectively Figure 6. The photocatalytic NOx removal of samples with dust accumulation on its surface. 0 50 100 150 200 250 300 10 g/m2 18 g/m2 37 g/m2 10 g/m2 18 g/m2 37 g/m2 NOx reduction (μmol h-1 m-2) Intermixed-Type 1 SP-Type 2 Reference-SP-Type 2, SD:12 Reference-Intermixed-Type 1, SD: 13 5 Intermixed-type 1 SP-type 2 Reference Cleaned with water after 37 g/m2 of dust Cleaned with water plus brush after 37 g/m2 of dust Reference Cleaned with water after 37 g/m2 of dust Cleaned with water plus brush after 37 g/m2 of dust Figure 7. Intermixed-Type 1 and SP-Type 2 samples cleaned by different processes (a) (b) Figure 8. Photocatalytic NOx removal of samples after subjected to cleansing with a) water and b) water and brush process. 0 50 100 150 200 250 300 10 g/m2 18 g/m2 37 g/m2 10 gr/m2 18 g/m2 37 g/m2 NOx reduction (µmol h-1 m-2) Dust accumulation Intermixed-Type 1 samples SP-Type 2 samples Reference, Intermixed- Type 1, SD: 13 Reference, SP-Type 2, SD:12 Intermixed-Type 1 samples SP-Type 2 samples Reference, Intermixed- Type 1, SD: 13 0 50 100 150 200 250 300 10 g/m2 18 g/m2 37 g/m2 10 gr/m2 18 g/m2 37 g/m2 NOx reduction (µmol h-1 m-2) Dust accumulation Intermixed-Type 1 samples SP-Type 2 samples Reference, Intermixed- Type 1, SD: 13 Reference, SP-Type 2, SD:12 6 (a) (b) Figure 9. Influence of oil accumulation on photocatalytic NOx removal a) intermixed- Type 1 and b) SP-Type 2 samples (a) (b) Figure 10. a) Photocatalytic NOx removal of the samples impregnated with motor oil; b) the intermixed-Type 1 and SP-Type 2 samples cleaned by different cleansing methods. 0 200 400 600 800 1000 1200 010 20 30 40 50 60 70 NOx concentration (ppb) Time (min) Reference sample Oil solution, 200 g/m2 Oil solution 400 g/m2 10% of sample area 0 200 400 600 800 1000 1200 010 20 30 40 50 60 70 NOx concentration (ppb) Time (min) reference-Type2 Motor Oil 400 g /m2 10% of the sample area 0 50 100 150 200 250 300 TYPE 1 (200g/m2) TYPE1 (400 g/m2) TYPE 2 (400 g/m2) NOx reduction (mol h-1 m-2) oil accumulated samples Intermixed-Type 1 (Reference), SD: 13 SP-Type 2 (Reference), SD:12 0 50 100 150 200 250 300 TYPE 1 (200g/m2)- cleaned detergent TYPE1 (200 g/m2)- cleaned n-hexane TYPE 2 (400 g/m2)-- cleaned n-hexane TYPE 2 (400 g/m2)-- cleaned n-hexane -2 NOx reduction (mol h-1 m-2) Intermixed-Type 1 (Reference), SD:13 SP-Type 2 (Reference), SD:12 7 (a) (b) Figure 11. a) SEM-EDX analysis of SP-Type 2 reference sample before mobile oil contamination, (b) SEM-EDX analysis of the SP-Type 2 reference sample after exposure to mobile oil contamination