Solar Photocatalysis for Gas-Phase Air Cleaning: from lab to pilot plant studies

Solar Photocatalysis for Gas-Phase Air Cleaning: from lab to pilot plant studies Thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Environmental Engineering, at the Faculty of Engineering, University of Porto Ricardo André Ribeiro Monteiro Supervisor: Dr. Vítor Jorge Pais Vilar Co-Supervisors: Dr. Adrián Manuel Tavares da Silva Dr. Rui Alfredo Rocha Boaventura Associate Laboratory LSRE-LCM Department of Chemical Engineering Faculty of Engineering University of Porto October, 2014 Acknowledgements The four years of this doctoral program symbolize a long and tough path, especially in the beginning. It was only possible because of all the people and institutions supporting me in many ways to whom I would like to acknowledge. First of all, I would like to deeply thank my supervisor Dr. Vítor Vilar, for the opportunity of developing my scientific work fulfilling all the required conditions in the research group (LSRE-FEUP). His scientific supervision and suggestions were essential to bring this dissertation towards fruition. To Dr. Rui Boaventura I thank him for his time in proofreading, for giving useful comments and suggestions to improve this manuscript. A very special thanks is addressed to Dr. Adrián Silva for being supportive at all levels. His ideas, advises and contributions were and will always be invaluable. A mention must be made to the following institutions that supported this work: the Foundation for Science and Technology (FCT) for the doctoral grant: SFRH/BD/69323/2010 and project PTDC/EQU-EQU/100554/2008); the Associate Laboratory of Separation and Reaction Engineering (LSRE) and Catalysis and Materials (LCM). The projects PEstC/EQB/LA0020/2013 and PEst-OE/QUI/UI0616/2014, financed by FCT and FEDER through COMPETE, and by QREN, ON2 (North Portugal Regional Operational Programme) and FEDER through projects NORTE-07-0162-FEDER-000050, NORTE-07-0162-FEDER-000015. I consider myself very fortunate to have had the opportunity to collaborate with Prof. Adélio Mendes and Prof. Joaquim Faria and their research staff. Their contributions have broadened and enriched the contents of this dissertation. I gratefully acknowledge Yas Yamin, Marlène Rouhet, Anne Roemer, Nicolas Keller and Valerie Keller from ICPEES– Institut de Chimie et Procédés pour l’Energie, l’Environnement et la Santé, CNRS and University of Strasbourg, France, for the human and scientific knowledge shared, as well as for providing experimental facilities. Merci beaucoup! A word for those who became my friends: João Pereira, Ariana Pintor, Livia Xerez, Bianca Souza, Tânia Silva, André Fonseca, Caio Rodrigues-Silva, Catarina Ferreira, Filipe Lopes and Joana Pereira. Thank you for the odd lunches, for the invigorating morning coffees, for the ramblings about the real and the absurd. Thank you for everything. To my family, I express my sincere gratitude for being there for me. Even not knowing the how or when they are my lifeline at all times. Finally, to Catarina, my inexhaustible source of love, friendship, complicity and solidarity at all moments, even in those I did not deserve. This is for you! Thank you so much! With love to my parents, brother and Catarina. Abstract VII Abstract Heterogeneous photocatalysis or photocatalytic oxidation (PCO) has been recognized over the recent decades as one of the most promising technologies for air treatment. The present thesis aims to evaluate the efficiency of solar gas-phase PCO as well as to provide fundamental understanding of different titanium dioxide (TiO2)-based photocatalysts toward the elimination of air pollutants such as volatile organic compounds (VOCs). The following objectives were addressed: (i) synthesis and characterization of TiO2 nanotubes as well as nitrogen modified TiO2 nanotubes and nanoparticles; (ii) preparation of TiO2-based materials and their immobilization as thin films on different supports using the dip-coating technique; (iii) evaluation of photocatalytic activity of the prepared photocatalysts and its dependence on different operational conditions (flow rate – Qfeed; pollutant concentration – Cvoc, feed; relative humidity – RH, irradiance – I; presence/absence adsorbed molecular oxygen) using a lab-scale single-pass continuous-flow annular photoreactor; (iv) description of PCO mechanisms; (v) evaluation of photocatalytic activity of selected materials in a pilot-scale photoreactor for continuous removal of VOCs. Gas-phase PCO of perchloroethylene (PCE) (574 – 2442 ppm) was carried out in a labscale annular photoreactor (r = 23.2 mm) under catalytic (UV TiO2) conditions. The photoreactor was assembled with a catalytic bed made of glass spheres packed with the benchmark TiO2 P25 catalyst (Evonik®) filling the voids between the spheres. A UVC lamp (λ = 253.7 nm) was located inside an inner tube (quartz or glass) of the photoreactor. The PCO of PCE was evaluated upon three parameters and employing the photoreactor with the glass inner tube (mimicking UV fraction of solar radiation); briefly: (1) for a 4.25 fold increase in CPCE, feed (from 574 ppm to 2442 ppm), a reduction of ~76 % (from 63 % to 35 %) on the PCE conversion was observed; (2) increasing the Qfeed (from 59 cm3 min-1 to 300 cm3 min-1) led to lower PCE conversions (from 81 % to 33 %); (3) for a 3.33 fold reduction in RH (from 40 % to 12 %), the PCO efficiency merely decreased ~1.2 times. Although 97 % of initial PCE was converted into CO2 and water through pure photolysis using the quartz inner tube, only 51 % of mineralization was attained through PCO when the glass inner tube was used. A mathematical model was developed taking into consideration the following main assumptions: (a) steady-state operation; (b) isothermal and isobaric conditions; (c) ideal gas behaviour; (d) axial symmetry; (e) constant porosity of the bed (uniform packing shape and distribution); (f) axially dispersed plug flow; (g) no heat transfer resistance and no thermal and UV radiation gradients; (h) no mass, velocity, and UV radiation gradients in the radial direction. Assuming that only PCE and H2O are the major species as well as that intermediates and/or reaction products do not influence PCE degradation kinetics, it was Abstract VIII shown that PCE and H2O molecules may have to be considered in association with different specific active sites of the TiO2 surface since Langmuir-Hinshelwood bimolecular non-competitive two types of sites model described better the experimental data. PCO of gas-phase PCE and n-decane was also assessed employing the lab-scale annular photoreactor under simulated solar radiation. Cellulose acetate monolithic structures coated with two different TiO2 photocatalytic films (prepared from aqueous suspensions of P25 from Evonik® and PC500 from Cristal®) were employed as catalytic bed (Lcatalytic bed = 160 mm). The reactor was equipped with a compound parabolic collector (CPC) allowing the whole reactor and bed illumination. PC500 film provided higher conversion and mineralization of PCE and n-decane than those obtained with P25 film, most likely due to the higher specific surface area of TiO2 PC500. Conversions of both pollutants (Cdec, feed = 71 ppm and CPCE, feed = 1095 ppm) were close to 100 % using PC500 film when Qfeed = 150 cm3 min-1, I = 38.4 WUV m-2 and RH = 20 %. Competitive adsorption between the pollutants and water molecules on the PC500 film surface was found above 20 % of RH. Results showed that gas-phase molecular oxygen has a fundamental role in the PCO reaction, as the conversion of pollutants is drastically impaired in its absence, suggesting that photocatalytic mechanism consists of: (i) oxidation reactions promoted by reactive species formed from the adsorbed molecular oxygen (O2 ⦁−, HOO⦁); (ii) contribution of the oxygen from the lattice of TiO2; (iii) hydroxyl radical (HO⦁) formation on the TiO2 surface. PCO mechanism of chlorinated compounds such as PCE, may also include a series of reactions involving Cl• radicals. The best performing photocatalyst (TiO2 PC500) was then incorporated into an exterior (water based) vinyl paint, supported on the cellulose acetate monolithic structure and tested under simulated solar radiation. 60 % of the PCE (CPCE, feed = 1100 ppm) and 98 % of the n-decane (Cdec, feed = 41 ppm) feed concentrations were converted when using Qfeed = 75 cm3 min-1, RH = 40 % and I = 38.4 WUV m-2. In addition, the results revealed an optimum surface density (~0.87 mg cm-2) to achieve the highest performance; also, increasing the exposed area to radiation by removing the outer wall of the monolithic structure, the PCO of PCE was enhanced by, approximately, 58 % (from 38 % to 60 %) under the same experimental conditions. Considering that TiO2 is a wide bandgap semiconductor, only active under UV radiation, TiO2 modification with nitrogen was studied to enhance its performance for solar applications. TiO2 nanotubes were synthesized by hydrothermal treatment of TiO2 P25 and chemical modification was performed by grinding urea (nitrogen source) with the TiO2 nanotubes, followed by thermal treatment. The same procedure was applied to bare TiO2 P25 nanoparticles. The results revealed that nitrogen-modification of TiO2 P25 decreased the photocatalytic activity Abstract IX towards PCO of gas-phase PCE under simulated solar radiation: bare TiO2 P25 presented the highest activity (67 % of PCE conversion) whereas only 35 % of conversion was attained over N0.50P25-380 (material with a urea:TiO2 weight ratio of 1:2 and calcined at 380 ºC). However, when these particular materials were tested in aqueous-phase, N0.50P25-380 showed the highest photocatalytic activity for PCO of diphenhydramine (emerging water pollutant of pharmaceutical origin) and inactivation of Escherichia coli bacteria under visible (λ > 430 nm) and UVA (λ = 365 nm) radiation, respectively. The PCO over nitrogen-modified TiO2 nanotubes was also investigated for the conversion of methylethylketone (MEK) and hydrogen sulfide (H2S). MEK showed high resistance to photocatalytic degradation over these materials, but a high photocatalytic activity towards H2S degradation under UVA (λ = 365 nm) radiation and moderate photocatalytic activity under solar light radiation were observed. Finally, based on lab-scale experimental data, modelling simulations and predictions of PCO of pure-targeted VOCs, a pilot-scale annular photoreactor (r = 32.8 mm, Lcatalitic bed = 144 cm) was designed and manufactured. The photoreactor features a CPC to capture both direct and diffuse solar radiation and/or UVA lamps, in order to work continuously day and night. The PCO of n-decane (Cdec, feed = 10 ppm, Qfeed = 2 L min-1, τ = 44 s) over cellulose acetate monolithic structures coated with different TiO2-based photocatalytic films (P25, PC500 and photocatalytic paint) was studied under solar and artificial UVA radiation. Conversions up to 100 %, were attained using P25 or PC500 films under solar irradiances of 15 WUV m-2 (morning, increasing temperature) and 3 WUV m-2 (afternoon, decreasing temperature). The photocatalytic paint film promoted up to 45 % of n-decane conversion under 48 WUV m-2 in both periods of the day. The excess of photons reaching the photocatalytic bed seems to favour the direct reaction pathway of CO2 production. The PCO of n-decane under artificial UVA radiation was 29 % higher using the PC500 film in comparison with the P25 film (resulting in 100 % of conversion), while over the photocatalytic paint film no more than 25 % of n-decane was converted. Results suggest that a 24 h continuous PCO process towards the removal of n-decane can be accomplished by combining both radiation sources (artificial UVA and solar). Table of Contents XVI 2.4.1. UV photolysis of PCE ........................................................................................... 78 2.4.2. UV-TiO2 photocatalytic conversion of PCE ......................................................... 79 2.4.2.1. Influence of operation parameters ................................................................. 79 2.4.2.2. Mathematical modelling................................................................................ 82 2.4.3. Photooxidation reaction of PCE: product analysis and reaction mechanism ........ 83 2.5. Conclusions ................................................................................................................... 89 2.6. References ..................................................................................................................... 91 3. Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor ................... 93 3.1. Introduction ................................................................................................................... 95 3.2. Experimental .................................................................................................................. 96 3.2.1. Materials and methods........................................................................................... 96 3.2.2. Experimental setup and photocatalytic experiments ............................................. 98 3.3. Results and discussion ................................................................................................. 100 3.3.1. VOCs photolysis ................................................................................................. 100 3.3.2. Catalytic film performances ................................................................................ 100 3.3.2.1. Influence of the feed flow rate and VOC .................................................... 100 3.3.2.2. Effect of the RH on PCO of n-decane and PCE over PC500 films ............. 107 3.3.2.3. Effect of oxygen on n-decane and PCE PCO over PC500 film .................. 109 3.4. Conclusions ................................................................................................................. 111 3.5. References ................................................................................................................... 113 4. Gas-phase solar photocatalytic oxidation of PCE over TiO2 based paint .... 119 4.1. Introduction ................................................................................................................. 121 4.2. Experimental ................................................................................................................ 123 4.2.1. Materials and chemicals ...................................................................................... 123 4.2.2. Photocatalytic films preparation and characterization ........................................ 124 4.2.3. Experimental setup and photocatalytic experiments ........................................... 126 4.3. Results and discussion ................................................................................................. 127 4.3.1. Photocatalytic oxidation of PCE ......................................................................... 127 4.3.1.1. Influence of the number of photocatalytic paint coating layers and substrate configuration ............................................................................................... 127 4.3.1.2. Operating parameters affecting PCE photodegradation .............................. 129 4.3.2. PCE PCO reaction intermediated and pathway ................................................... 134 4.4. Conclusions ................................................................................................................. 139 4.5. References ................................................................................................................... 141 5. Gas-phase solar photocatalytic oxidation of n-decane over TiO2 based paint .............................................................................................................. 151 5.1. Introduction ................................................................................................................. 147 5.2. Experimental ................................................................................................................ 149 5.2.1. Materials and chemicals ...................................................................................... 149 5.2.2. Photocatalytic films preparation and characterization ........................................ 149 5.2.3. Experimental setup and photocatalytic experiments ........................................... 150 5.3. Results and discussion ................................................................................................. 151 5.3.1. Photocatalytic oxidation of n-decane .................................................................. 151 Table of Contents XVII 5.3.1.1. Surface characterization of photo-TiO2 powder and PC samples ............... 151 5.3.1.2. Operating parameters effect on n-decane photoconversion ........................ 153 5.3.1.3. Simulation and predictive studies of n-decane kinetics through PCO ........ 158 5.3.2. Reaction mechanism for the PCO of n-decane ................................................... 161 5.4. Conclusions ................................................................................................................. 164 5.5. References ................................................................................................................... 167 6. N-modified TiO2 photocatalytic activity towards organics degradation ......173 6.1. Introduction ................................................................................................................. 175 6.2. Experimental ............................................................................................................... 177 6.2.1. Chemicals and materials ..................................................................................... 177 6.2.2. Catalyst preparation and characterization ........................................................... 178 6.2.3. Photocatalytic experiments ................................................................................. 179 6.2.3.1. Gas-phase photocatalytic experiments ....................................................... 179 6.2.3.2. Diphenhydramine photocatalytic degradation ............................................ 180 6.2.3.3. Bacterial inactivation tests .......................................................................... 182 6.3. Results and discussion ................................................................................................ 183 6.3.1. Characterization .................................................................................................. 183 6.3.2. Gas-phase PCO of PCE under simulated solar light ........................................... 187 6.3.3. Diphenhydramine photocatalytic degradation under visible light ...................... 188 6.3.4. Escherichia coli inactivation ............................................................................... 189 6.4. Conclusions ................................................................................................................. 191 6.5. References ................................................................................................................... 193 7. Visible-light-driven photocatalytic properties of N-modified titania nanotubes toward air purification ..................................................................197 7.1. Introduction ................................................................................................................. 199 7.2. Experimental ............................................................................................................... 200 7.2.1. Hydrothermal synthesis of titanate nanotubes (TiNT) ........................................ 200 7.2.2. Nitrogen-modified TiO2 nanotubes (NTiNT) ..................................................... 200 7.2.3. Characterization of N-modified TiO2 nanotubes ................................................ 200 7.2.4. Gas-phase photocatalytic experiments ................................................................ 201 7.3. Results and discussion ................................................................................................ 203 7.3.1. Material characterization .................................................................................... 203 7.3.2. Photocatalytic tests ............................................................................................. 206 7.4. Conclusions ................................................................................................................. 209 7.5. References ................................................................................................................... 211 8. Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane ...........................................................................................217 8.1. Introduction ................................................................................................................. 219 8.2. Experimental ............................................................................................................... 221 8.2.1. Photocatalytic films preparation ......................................................................... 221 8.2.2. Solar/UV pilot-scale experimental unit .............................................................. 222 8.2.2.1. Feed generation........................................................................................... 222 8.2.2.2. Pilot-scale photoreactor .............................................................................. 223 8.2.2.3. Compound parabolic collector (CPC) and radiation sources ...................... 223 Table of Contents XVIII 8.2.2.4. Photoreactor feed and exit streams analysis ................................................ 224 8.2.3. Photocatalytic experiments .................................................................................. 224 8.3. Results and discussion ................................................................................................. 225 8.3.1. Solar/artificial UVA photolysis of n-decane ....................................................... 225 8.3.2. Solar/artificial UVA PCO of n-decane ................................................................ 226 8.4. Conclusions ................................................................................................................. 232 8.5. References ................................................................................................................... 235 9. Final remarks and suggestions for future work ............................................ 237 9.1. Final Remarks .............................................................................................................. 239 9.1.1. Photolysis ............................................................................................................ 239 9.1.2. Photocatalysis using a lab-scale photoreactor ..................................................... 240 9.1.2.1. TiO2 photocatalytic properties enhancement .............................................. 244 9.1.3. Photocatalysis using a pilot-scale photoreactor ................................................... 245 9.2. Suggestions for future work ........................................................................................ 246 A. Master gas chromatography (MGC) .............................................................. 251 A.1. MGC calibration curves ................................................................................................. 251 A.2. MGC data treatment ....................................................................................................... 253 A.3. References ...................................................................................................................... 257 Table of Figures XIX Table of Figures Page Figure 1.1. Steps of a photocatalytic reaction in a solid semiconductor particle: (1) light photons of energy h  matching or exceeding the semiconductor band-gap energy; (2) excited electron, ecb -, migrates from valence band to conduction band leaving a hole, hvb +, in the valence band; (3) hvb + migrates to surface and initiates oxidation reactions; (4) ecb - migrates to surface and initiates reduction reactions; (5) charge carrier recombination liberating heat. ............................................................................................................................. 11 Figure 1.2. Band positions (top of valence band and bottom of conduction band) for several common semiconductors together with the band-gap energy [95]. ............................................. 16 Figure 1.3. TiO2 crystallographic phases of anatase (a), rutile (b) and brookite (c) (adapted from Carp et al. [112])................................................................................................................. 17 Figure 1.4. Photoexcitation and charge transfer in a metal-modified TiO2 photocatalyst (adapted from Serpone et al. [143]). ........................................................................................... 20 Figure 1.5. Photoexcitation and charge transfer between two light active semiconductors: (a) injection of an ein the TiO2 conduction band and injection of a h+ in the valence band of the other semiconductor; (b) injection of an ein the conduction band of the other semiconductor and injection of a h+ in the TiO2 valence band (adapted from Serpone et al. [152]). .......................................................................................................................................... 21 Figure 1.6. Different morphologies of TiO2 nanostructures. ....................................................... 22 Figure 1.7. Mechanisms proposals for the changes that may occur to the bandgap electronic structure of N-doped TiO2: (a) undoped TiO2; (b) bandgap narrowing as a result of VB broadening; (c) introduction of localized states above VB or below CB; (d) electronic transitions from localized states near VB to the corresponding excited states for Ti3+ (JahnTeller split 2T2  2E) and F+ (equivalent to a single electron associated with the O vacancy) centres; (e) sensitization by compounds containing nitrogen species (adapted from Serpone [163]). .......................................................................................................................................... 23 Figure 1.8. Photoexcitation and charge transfer of TiO2 using dye molecule sensitizer (adapted from [197]). .................................................................................................................. 24 Figure 1.9. Schematic representation of TNTs synthesis by hydrothermal method. ................... 28 Figure 1.10. Cellulose acetate polymer: (a) commercially available monolith-like structure – TIMAX CA50-9/S, Wacotech GmbH & Co. K.G.; (b) transmittance in the UV-Vis range. ... 35 Figure 1.11. Schematic representation of a FPR (adapted from Águia et al. [367]) ................... 38 Figure 1.12. Schematic representation of a multi-parallel FPR (adapted from Leung et al. [314]). .......................................................................................................................................... 38 Figure 1.13. Schematic representation of a FBR (adapted from Dibble and Raupp [300]. ......... 39 Figure 1.14. Schematic representation of a PBR with glass spheres coated with the photocatalyst (adapted from Tsoukleris et al. [372]). ................................................................. 40 Table of Figures XX Figure 1.15. Schematic diagram of a PDC reactor packed with photocatalytic pellets (adapted from Kim et al. [377]). .................................................................................................. 41 Figure 1.16. Schematic representation of a tubular reactor filled with a monolithic solid structure. ...................................................................................................................................... 42 Figure 1.17. Schematic representation of an OFR filled with photocatalyst-coated fibres.......... 42 Figure 1.18. Schematic representation of a CWR. ....................................................................... 43 Figure 1.19. Schematic representation of an annular reactor: a) side and b) cross section point of view. ............................................................................................................................... 44 Figure 1.20. Schematic representation of a multi-annular reactor (adapted from Imoberdorf et al. [398])................................................................................................................................... 44 Figure 1.21. Schematic representation of three non-concentrating reflective collectors’ geometries. ................................................................................................................................... 45 Figure 2.1. Schematic representation of the lab-scale experimental unit used for the study of decontamination of air contaminated with PCE: a) lab-scale facility used for the generation of air streams containing PCE and water vapour; b) single-pass continuous-flow annular UV-photoreactor; c) master gas chromatographic analysis system used for the analysis of the photoreactor feed and exit streams....................................................................... 69 Figure 2.2. LabView routine designed to control/monitor the mass flow controllers and thermocouples throughout the experimental time. ....................................................................... 70 Figure 2.3. Detailed schematic representation of single-pass continuous flow packed bed annular photoreactor employed in the study of decontamination of air contaminated with PCE: a) side view; b) frontal view. .............................................................................................. 71 Figure 2.4. a) Transmissivity of glass (-----) and quartz (─ ─) inner tubes compared to the solar spectrum (──, [30]). b) PCE photolysis using different photoreactor inner tubes: glass (──) and quartz (----); CPCE, feed = 1221 ppm, Qfeed* = 150 cm3 min-1, RH = 40 %, and T = 298 K; operation conditions reported in Table 2.2 * measured at 298 K and 1 bar. .............. 79 Figure 2.5. Effect on PCE conversion fraction through PCO (CPCE, exit / CPCE, feed, at steadystate conditions) for air feed streams contaminated with different concentrations of PCE [CPCE, feed]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); Qfeeda = 150 cm3 min-1, RHa = 40 %, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2 (runs 1-6) [a measured at 298 K and 1 bar]. ............. 80 Figure 2.6. Effect on PCE conversion fraction through PCO (CPCE, Exit / CPCE, Feed, at steadystate conditions) for air contaminated with PCE applying different feed flow rates [Qfeed]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); CPCE, feed = 1221 ppm, RHa = 40 %, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2 (runs 1, 7-10) [a measured at 298 K and 1 bar]. ....................... 81 Figure 2.7. Effect on PCE conversion fraction through PCO (CPCE, exit / CPCE, feed, at steadystate conditions) for air contaminated with PCE applying different humidity contents [RH]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); CPCE, feed = 1221 ppm, Qfeeda = 150 cm3 min-1, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2(runs 1, 11-13) [a measured at 298 K and 1 bar]. ...... 82 Table of Figures XXI Figure 3.1. Schematic representation of the lab-scale experimental set-up and the continuous-flow photoreactor: a) generation of air streams containing n-decane and water vapour; b) sunlight simulator containing the photoreactor: b1) side view and b2) frontal view; c) master gas chromatograph analytic system used for the analysis of the photoreactor feed and exit gas streams. ....................................................................................... 99 Figure 3.2. Effect of feed flow rate [Qfeed*] on the conversion of n-decane (a) and PCE (c) and on the photocatalytic reaction rate, rVOC ((b) and (d), respectively). Experimental points for incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ), measured within the spectral range of 280 – 400 nm, at steady-state conditions; Cdec, feed = 71 ppm CPCE, feed = 1095 ppm, RH* = 40 % and 21 % oxygen. Blue columns represent TiO2 PC500 ( , , ) and orange columns P25 ( , , ). * measured at 298 K and 1 bar. .......................................................................................................................................... 102 Figure 3.3. Effect of VOC feed concentration (Cdec, feed, and CPCE, feed) on the conversion of n-decane (a) and PCE (c) and on the photocatalytic reaction rate, rVOC ((b) and (d), respectively). Experimental points for incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ) measured within the spectral range of 280 - 400 nm, at steady-state conditions; Qfeed* = 150 cm3 min-1, RH* = 40 % and 21 % oxygen. Blue columns represent the activity of PC500 ( , , ) and orange columns P25 ( , , ) films. * measured at 298 K and 1 bar. ............................................................................... 104 Figure 3.4. Mineralization yields of PCE and n-decane over PC500 and P25 films. Experimental points for incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ) measured within the spectral range of 280 - 400 nm, at steady-state conditions; CPCE, feed* = 1095 ppm, Cdec, feed* = 71 ppm Qfeed* = 150 cm3 min-1, RH* = 40 % and 21 % oxygen; Blue columns represent the activity of PC500 ( , , ) and orange columns P25 ( , , ) films. * measured at 298 K and 1 bar. ................................................... 105 Figure 3.5. Effect of water content [RH*] on the conversion over PC500 film, at steadystate conditions, of: (a) n-decane for Cdec, feed = 71 ppm ( ), Cdec, feed = 142 ppm ( ) and Cdec, feed = 284 ppm ( ); (b) PCE (CPCE, feed = 1095 ppm) for Qfeed* = 150 cm3 min-1 ( ) and Qfeed* = 300 cm3 min-1 ( ); I = 38.4 WUV m-2, measured within the spectral range of 280 - 400 nm. * measured at 298 K and 1 bar. .......................................................................... 108 Figure 3.6. Effect of water content [RH*] on: (a) n-decane conversion at steady-state conditions for Cdec, feed = 71 ppm and Qfeed* = 150 cm3 min-1; (b) PCE conversion at steadystate conditions for CPCE, feed = 1095 ppm and Qfeed* = 150 cm3 min-1. Blue columns ( , , ) and orange columns ( , , ) represent, respectively, the presence and absence of oxygen; Incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ) were measured within the spectral range of 280 – 400 nm. * measured at 298 K and 1 bar. .... 110 Figure 4.1. Annular photoreactor schematic representation: a) side view; b) frontal view of the two configurations used (PC1 and PC2). ............................................................................. 124 Figure 4.2. Influence on photocatalytic conversion, at steady-state conditions, of the (a) number of layers under xPC1 configuration and (b) structure configuration; I of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ) and 18.9 WUV m-2 ( ), measured within 280 - 400 nm; CPCE, feed = 1100 ppm, Qfeed* = 75 cm3 min-1, and RH* = 40 %; experimental conditions reported in Table 4.3. * measured at 298 K and 1 bar. .............................................................. 128 Figure 4.3. Effect of feed flow rate [Qfeed*] on PCE conversion (a) and on the reaction rate, rPCE (b) at steady-state conditions: experimental points for I of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ), measured within 280 - 400 nm; CPCE, feed = 1100 ppm, RH* = 40 %; experimental conditions reported in Table 4.3. Blue Table of Figures XXII columns represent 5PC2 first use ( , , ) and orange columns represent 50 h under simulated solar radiation and continuous feeding ( , , ). * measured at 298 K and 1 bar. ..... 130 Figure 4.4. Effect of different concentrations of PCE [CPCE, feed] on PCE conversion and on the reaction rate, rPCE, at steady-state conditions: experimental points for incident irradiances of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ), and 18.9 WUV m-2 ( ), measured within 280 - 400 nm; Qfeed* = 150 cm3 min-1 and RH* = 40 %; experimental conditions reported in Table 4.3. * measured at 298 K and 1 bar. ................................................................................. 131 Figure 4.5. Effect of water content [RH*] on PCE conversion at steady-state conditions in the presence of oxygen (blue columns coloured column) and in the absence of oxygen (orange columns): experimental points for incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ), measured within 280 - 400 nm; Qfeed* = 150 cm3 min-1 and CPCE, feed = 1100 ppm; experimental conditions reported in Table 3. * measured at 298 K and 1 bar. .............................................................................................. 133 Figure 5.1. Schematic representation of the continuous-flow photoreactor: a) from a side point of view and b) from a frontal point of view. ..................................................................... 150 Figure 5.2. SEM micrograph (a) and EDX spectrum (b) of photo-TiO2 PC500 powder. .......... 151 Figure 5.3. SEM micrographs (a-d) and EDX spectra (e, f) of PC before (left-side images) and after 50 h+ of use (right-side images) in PCO of n-decane. ................................................ 152 Figure 5.4. Photographs (a-b) and SEM micrographs (c-d) of PC used in PCO of n-decane. ... 153 Figure 5.5. Influence of feed flow rate (Qfeed) on n-decane conversion fraction (Cdec, exit / Cdec, feed, at steady-state conditions): experimental points for incident irradiance measured within 280 – 400 nm (sunlight UV fraction) of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ), and 18.9 WUV m-2 ( ), and RE-1 (- - -), RE-2 (─ ∙ ─), and RE-3 (──); Cdec, feed = 73 ppm, RH* = 40 %, and T = 298 K; operation conditions reported in Table 5.2; * measured at 298 K and 1 bar. .................................................................................................. 154 Figure 5.6. Influence of feed flow rate (Qfeed) on n-decane photocatalytic reaction rate (rdec), at steady-state conditions: experimental points for incident irradiance measured within 280 – 400 nm (sunlight UV fraction) of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ), and 18.9 WUV m-2 ( ); Cdec, feed = 73 ppm, RH* = 40 %, and T = 298 K; operation conditions reported in Table 5.2; * measured at 298 K and 1 bar. ............................................................... 155 Figure 5.7. Influence of the feed concentration (Cdec, feed) on n-decane conversion fraction (Cdec, exit / Cdec, feed, at steady-state conditions): experimental points for incident irradiance measured within 280 – 400 nm (sunlight UV fraction) of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ), and 18.9 WUV m-2 ( ), and RE-1 (- - -), RE-2 (─ ∙ ─), and RE-3 (──); Qfeed* = 150 cm3 min-1, RH* = 40 %, and T = 298 K; operation conditions reported in Table 5.2; * measured at 298 K and 1 bar. ........................................................................................... 156 Figure 5.8. Influence of feed concentration (Cdec, feed) on the photocatalytic reaction rate (rdec), at steady-state conditions: experimental points for incident irradiance measured within 280 – 400 nm (sunlight UV fraction) of 38.4 WUV m-2 ( ), 29.1 WUV m-2 ( ), and 18.9 WUV m-2 ( ); Cdec, feed = 73 ppm, RH* = 40 %, and T = 298 K; operation conditions reported in Table 5.2; * measured at 298 K and 1 bar. ............................................................... 157 Figure 5.9. Influence of the feed relative humidity (RH) on n-decane conversion fraction (Cdec, exit / Cdec, feed, at steady-state conditions): experimental points for incident irradiance measured within 280 – 400 nm (sunlight UV fraction) of 38.4 WUV m-2 ( ), 29.1 WUV m-2 Table of Figures XXIII ( ), and 18.9 WUV m-2 ( ), and RE-1 (- - -), RE-2 (─ ∙ ─), and RE-3 (──); Qfeed* = 150 cm3·min-1, Cdec, feed = 73 ppm, and T = 298 K; operation conditions reported in Table 5.2; * measured at 298 K and 1 bar. ................................................................................. 158 Figure 5.10. n-Decane photoconversion fraction profiles (Cdec, exit/Cdec, feed, at steady-state conditions) for photoreactors with different lengths [LR]: 0.16 (—), 0.24 (– – –), 0.32 (- - -), and 0.48 m (· · ·); (a) Cdec, feed = 73 ppm, Qfeed* = 150 cm3 min−1, and I = 18.9 WUV m−2 (measured within 280 – 400 nm: sunlight UV fraction); (b) Cdec, feed = 73 ppm, Qfeed* = 300 cm3 min−1, and I = 38.4 WUV m−2 (measured within 280 – 400 nm: sunlight UV fraction); RH* = 30 %, and T = 298 K; operation conditions reported in Table 5.2 (runs 3 and 7, respectively) (* measured at 298 K and 1 bar); experimental data (points); mathematical modelling with RE-3 (lines). ............................................................................... 161 Figure 6.1. Schematic representation of the experimental unit used in DP photocatalytic degradation experiments (adapted from Gomes da Silva [80]). ................................................ 180 Figure 6.2. Schematic representation of the experimental unit used for the E. coli inactivation experiments. ........................................................................................................... 182 Figure 6.3. XRD patterns of the samples synthesized with different (a) urea contents and (b) calcination temperatures. XRD pattern of TiO2 P25 is also shown as reference. A: anatase; R: rutile. ....................................................................................................................... 183 Figure 6.4. N2 adsorption-desorption isotherms and pore size distribution (inset) of: a) N0.75P25-380, b) N0.50P25-380 and c) N0.25P25-380. ................................................................. 185 Figure 6.5. UV-Vis absorption spectra of: a) N0.25P25-380, N0.50P25-380, N0.75P25-380 and bare P25 and b) N0.50P25-340, N0.50P25-380, N0.50P25-420 and bare TiO2 P25. ....................... 185 Figure 6.6. (a-b) SEM micrographs at different magnifications, and (c) EDS spectrum of N0.50P25-380. ............................................................................................................................. 186 Figure 6.7. X-ray photoelectron spectra of N0.50P25-380 and bare TiO2 P25 samples: (a) N1s, (b) Ti2p and (c) O1s core levels. ....................................................................................... 186 Figure 6.8. Gas-phase PCO of PCE (CPCE, feed = 1100 ppm; Qfeed* = 300 cm3 min-1; RH* = 40 %) at steady-state conditions, under simulated solar radiation (I = 38.4 WUV m-2, measured within the spectral range of 280 – 400 nm) for materials prepared with different urea content and bare P25 * measured at 298 K and 1 bar. ....................................................... 187 Figure 6.9. Photocatalytic degradation of DP (10 mg L-1) under visible light illumination for (a)-(b) materials prepared with different urea contents and (c) different temperatures (340, 380 and 420 ºC). Catalyst load = 1.0 g L-1. Curves represent the fitting of the pseudofirst order equation to the experimental data. ............................................................................ 188 Figure 6.10. E. coli growth under UVA for photocatalyst loads of (a) 0.500 mg mL-1 and (b) 0.125 mg mL-1. Results are mean values (n = 3) and the error bars represent the standard deviation. ................................................................................................................................... 190 Figure 7.1. Lab-scale facility employed in MEK or H2S degradation experiments comprising a photocatalytic annular concentric reactor ............................................................ 202 Figure 7.2. BET surface area (SBET) calculated by Brunauer-Emmet-Teller (BJH) formula (a), N2 adsorption-desorption isotherms (b), and median pore diameter (dp, BJH) calculated Table of Figures XXIV by Barret-Joyner-Halenda (BJH) formula (c) of N1TiNT to N4TiNT samples calcined at 380 and 400 ºC. .......................................................................................................................... 203 Figure 7.3. UV-Vis absorption spectra of N4TiNT and N1TiNT calcined at 380 ºC and 400 ºC, TiNT and bare TiO2 P25 without thermal treatment. ........................................................... 204 Figure 7.4. TG curves of TiNT, urea, and N4TiNT samples without thermal treatment. .......... 205 Figure 7.5. N1s X-ray photoelectron spectral details (a) and N/Ti and O/Ti area ratios during argon ion bombardment (b) of N2TiNT-380, N3TiNT-380, N1TiNT-400, N2TiNT400, and N3TiNT-400 samples. .................................................................................................. 206 Figure 7.6. TEM images of N4TiNT-380 (left side) and N4TiNT-400 (right side). ................... 206 Figure 7.7. Gas-phase photooxidation under UVA or solar radiation of: a) MEK with N1TiNT-380 photocatalyst (ρA = 1.67 mg cm-2, CMEK, feed = 500 ppm, RH* = 50 %, Qfeed* = 300 cm3 min-1); b) H2S with N1TiNT-400 (ρA = 0.21 mg·cm-2, CH2S, feed = 15 ppm, RH* = 0%, Qfeed* = 500 cm3 min-1. [* - measured at 298 K and 1 atm]. ..................................... 207 Figure 7.8. Schematic representation of reaction pathways for the photocatalytic degradation under UVA or solar radiation of MEK [58, 59] (a) and H2S [63-66] (b). .............. 208 Figure 8.1. Schematic representation of the pilot unit: a) feed generator of air streams containing n-decane and water vapour; b) pilot scale continuous-flow annular photoreactor placed at the top of a CPC (at local latitude of 41º facing south); c) monitoring system used for the analysis of the photoreactor feed and exit streams. ........................................................ 222 Figure 8.2. Time evolution of n-decane PCO over P25 (a), PC500 (b) and PCP (c) films under solar radiation (irradiance measured within 280 - 400 nm). Solar radiation collected with a CPC and UV irradiance measured on the outer tube of the photoreactor. UV irradiance measured within 280 - 400 nm. Operating conditions: Cdec, feed = 10 ppm; Qfeed* = 2 L min-1 (τ = 44 s) as reported in Table 8.2. ................................................................ 228 Figure 8.3. n-Decane conversion (, morning, increasing irradiance; , afternoon, decreasing irradiance) and mineralization (, morning, increasing irradiance; , afternoon, decreasing irradiance) over P25 (a) PC500 (b) and PCP (c) films under solar radiation during the fourth day of experiment. UV irradiance measured within 280 - 400 nm. Operating conditions: Cdec, feed = 10 ppm; Qfeed* = 2 L min-1 (τ = 44 s) as reported in Table 8.2. ............................................................................................................................................. 229 Figure 8.4. n-Decane PCO reaction rate (, ,  morning, increasing irradiance; , ,  afternoon, decreasing irradiance) over P25 (, ) PC500 (, ) and PCP (, ) films under solar radiation during the fourth day of experiment. UV irradiance measured within 280 - 400 nm. Operating conditions: Cdec, feed = 10 ppm; Qfeed* = 2 L min-1 (τ = 44 s) as reported in Table 8.2. ................................................................................................................. 230 Figure 8.5. Time evolution of n-decane PCO over P25 (a), PC500 (b) and PCP (c) films under artificial UVA radiation. UV irradiance measured within 280 - 400 nm and facing the inner quartz tube of the photoreactor. Operating conditions: Cdec, feed = 10 ppm; Qfeed* = 2 L min-1 (τ = 44 s) as reported in Table 8.2. ................................................................ 231 Figure A.1. GC calibration curves for PCE and n-decane ......................................................... 252 Table of Figures XXV Figure A.2. Effect on component (i)-relative vapour pressure (φi = pi/pisat) applying different air flow rates (Qi): experimental data obtained for H2O (points) determined at 281.15 K [a measured at 293.15 K and 1 atm]; lines for eq. A.7 ............................................... 254 Notation XXXII k reaction rate constant [μmol m-2 s-1 (W-1 m2)n] KH2O H2O adsorption equilibrium constant in a single site [μM-1] KH2O,1 H2O adsorption equilibrium constant on two types of sites (site 1) [M-1] KH2O,2 H2O adsorption equilibrium constant on two types of sites (site 2) [M-1] Kdec n-decane adsorption equilibrium constant in a single site [M-1] Kdec,1 n-decane adsorption equilibrium constant on two types of sites (site 1) [M-1] Kdec,2 n-decane adsorption equilibrium constant on two types of sites (site 2) [M-1] KPCE PCE adsorption equilibrium constant in a single site [M-1] KPCE,1 PCE adsorption equilibrium constant on two types of sites (site 1) [M-1] KPCE,2 PCE adsorption equilibrium constant on two types of sites (site 2) [M-1] LC length of cellulose acetate monolithic structure channel [mm] Lcc length of the capillary column [m] Lin length of the glass and quartz inner tubes [mm] Lot length of the Pirex-glass outer tube [mm] LR length of the photocatalytic bed [mm] m mass [g] M(C) molecular weight of a carbon atom [g mol-1] MGC Master Gas Chromatographer/Chromatography Mi molecular weight of compound i [g mol-1] Mj molecular weight of compound j [g mol-1] n incident irradiance exponential order constant n(C) number of carbon atoms of each component i N number of components in the gas stream mixture p total pressure of the contaminated air stream [atm] P or PCP photocatalytic paint immobilized P25 TiO2 from Evonik® PC500 TiO2 from Cristal® PCE perchloroethylene Qfeed total feed flow rate [cm3 min-1] RH relative humidity of the feed stream (water vapour content) [%] r photocatalytic reaction rate [mol min-1] R2 squared correlation coefficient [-] S2R sum of squared residuals between experimental and calculated rates [μmol2 m-4 s-2] SBET BET specific surface area [m2 g-1] t time [s] T temperature [K] Notation XXXIII feed 0, u inlet superficial velocity [m·s-1] ch0, u Superficial velocity (in the cross-section of each channel) [m·s-1] UV ultraviolet Vis visible VGS volume of glass spheres [mm3] VOC(s) Volatile Organic Compound(s) VR volume of photocatalytic bed [mm3] z partition of the photocatalytic bed length (LR) [m] Greek letters α exponential parameter for PCE [-] α1 kinetic coefficient [μmol m s-1 μM-1] α2 kinetic coefficient [W-1 m2] β exponential parameter for H2O [-] ε photocatalytic bed porosity (Σv)i sum of the diffusion volume for component i (Σv)j sum of the diffusion volume for component j ηmin mineralization efficiency [%] λ UV/visible wavelength [nm] νi stoichiometric coefficient of compound i in the overall reaction ρ density [g cm−3] ρA surface density [g cm−2]  frequency energy [s-1] φ wavelength averaged quantum efficiency [-] Part I Chapter 1. Introduction 1. Introduction This first chapter presents an overview of the problematic of airborne volatile organic compounds (VOCs) present in both indoor and outdoor atmospheres as well as of current and potential removal methods. Photocatalysis as an advanced oxidation process for removing VOCs is herein described. Crystalline structural and morphological properties of TiO2 semiconductor-based catalyst and their effect on photocatalytic activity are discussed. Methods to enhance TiO2 properties are also explained. An overview regarding photoreactors is presented in this chapter. The objective and the thesis outline are provided at the end of the chapter. Introduction 5 1.1. History of air pollution Air pollution has been a major recognised problem for centuries. Since middle ages, the burning of coal, mainly in cities, has released increasing amounts of smoke and sulphur dioxide to the atmosphere. Well documented air pollution related works can be found as early as the late 16th century. However, it was since the middle of 18th century, with the British Industrial Revolution (ca. 1760 to 1840), that an escalation in pollutant emissions related to the use of coal in households and industry led to problematic levels of urban air pollution. Such historical event marks a major turning point in almost every aspect of daily life. The transition from handproduction methods to machines, the increasing use of steam power and the development of machine tools as well as the change from wood and other bio-fuels to coal and the use of new chemical manufacturing and iron production processes are probably the pinnacle of the Industrial Revolution features. Thus, side by side with technological and industrial development, the continuous growth of pollutant emissions was a reality. Culminating in the year of 1952, in its upmost catastrophic effect known as the Great London Smog [1], air pollution was responsible for around 4000 deaths in four days and further 8000 deaths in the following weeks. In a strategic attempt of reducing air pollution, the British Government introduced in 1956 its first Clean Air Act (CAA) with the aim to control domestic sources of smoke pollution by introducing smokeless zones. Though, it was the introduction of cleaner coals, the increased use of electricity and gas and the relocation of power stations to more rural areas that promoted a dramatically decreased in air pollution in cities. In 1968, assuming that higher chimneys would improve the dispersal of the air pollution, another Clean Air Act was introduced aiming the use of tall chimneys for burning coal, liquid or gaseous fuels industries. In 1963, the United States government also deliberated the Clean Air Act (CAA) federal law to control air pollution on a national level and to protect the general public from airborne contaminants known to be hazardous to human health. The 1963 version of the legislation established a research program, further expanded in 1967. Major amendments to the law, requiring regulatory controls for air pollution, passed in 1970, 1977 and 1990. In 1969 it was also created the environmental law - National Environmental Policy Act (NEPA) [2], containing three sections: 1. the declaration of national environmental policies and goals; 2. the establishment of action-forcing provision for federal agencies to enforce those policies and goals; 3. the establishment of a Council on Environmental Quality (CEQ) in the Executive Office of the President. United Kingdom also introduced further regulations after Clean Air Acts such as the 1974 Control of Air Pollution Act which included regulations for the composition of motor fuel Chapter 1 6 and limits for the sulphur content of industrial fuel oils. Nowadays, pollution from motor vehicles has become the most recognised air quality issue. The number of cars around the world is now steadily increasing, and a speed up in environmental friendly technological development is required to tackle the pollution problem. 1.2. Indoor air: exposure to Volatile Organic Compounds (VOCs) Although much attention has been directed towards poor air quality outdoors, people spend more than 90% of their time in an indoor environment such as home, office, car, and shopping centre [3]. Due to poor Indoor Air Quality (IAQ) and related health problems, the European Environmental Agency (EEA) identified IAQ as one of the worldwide priority concerns in children’s health [4, 5]. Common indoor air pollutants include particles (such as dust and smoke), biological agents (moulds, spores), radon, asbestos, and gaseous contaminants such as CO, CO2, NOx, SOx, aldehydes and volatile organic compounds (VOCs). These pollutants have a close relation with the sick building syndrome (SBS), which is one of many terms used by occupants to describe symptoms of reduced comfort or health (e.g. headache, fatigue, skin and eye irritations or respiratory illness) [6, 7]. It is worth noting that according to World Health Organization (WHO) fact sheet nr. 313/2014 [8] indoor air pollution causes 3.3 million deaths per year, while outdoors pollution induced mortality is 2.6 million. IAQ is also an important factor in work productivity as shown in Wargocki [9] work, where individuals exposed to a typical indoor pollution source (plastic carpet) typed 6.5 % less than a control group. Furthermore, empirical studies have shown that the use of ventilation rates lower than 25 L s-1 per person in commercial and institutional buildings was correlated with an increase in the number of short-term sick leaves [10, 11]. According to the definition of the WHO, VOCs are referred as all carbon and hydrogencontaining chemicals in the boiling point range of 50 - 260 °C, excluding pesticides. Thus, there are thousands of different VOCs produced and present in our daily lives including those coming from cleaning and degreasing products, air fresheners, toilet bowl deodorants, tobacco smoke, furniture and building materials (e.g. wood products, adhesives, carpeting, paints, varnishes, vinyl floors, newspaper, upholstery, fabrics, sealing caulks), cosmetics, fuel oil, vehicle exhaust, cooking, photocopying, etc. Common VOCs include acetone, benzene, ethylene glycol, formaldehyde, methylene chloride, perchloroethylene, toluene, xylene, and 1,3-butadiene [3, 12, 13]. The concentration of indoor VOCs varies according to the total space volume, the Introduction 7 pollutant production and removal rates (the air exchange rate with the outside atmosphere) and the outdoor VOC concentrations. In a study conducted by Salthammer [14], 150 VOCs (mainly aliphatic and aromatic aldehydes, aromatic hydrocarbons, ketones, esters and glycols) from furniture coatings were identified in a test chamber air under dynamic condition. The Total VOC (TVOC) concentrations ranged from 4 μg m-3 up to 1288 μg m-3 being TVOC emission rates as high as 22280 µg m-2 h-1 [14] . Moulds and bacteria can also contribute significantly to the presence of particles (spores) and VOCs in indoor pollution [15]. Microbial development in buildings can also be found in places where humidity accumulates, such as defective heating and air conditioning systems, garbage disposal, bathrooms and water leaks, and they are responsible for toxic and allergenic responses [16]. Although the concentration of each contaminant is usually low (µg m-3), several hundred contaminants can be found at the same time, resulting in significant TVOC levels. In a study conducted by Kostiainen [17], the individual concentrations of selected pollutants were up to 1000 times higher in 38 finish sick-houses (where people experienced symptoms associated with SBS) than their individual mean concentrations found in 50 normal houses used as reference, with over 200 VOCs being simultaneously detected in 26 houses. The highest TVOC concentration found was 9538 µg m-3 in one sick house compared to the mean concentration of 121 µg m−3 for the normal houses. Daisey et al. [18] reported indoor TVOC concentrations of 230 – 7000 µg m-3 (geometric mean of 510 µg m-3) in 12 Californian office buildings. Although it is not an easy task to correlate the TVOC concentration with health effects, experience of eye, nose or mouth irritation has been reported for the range of 5000 - 25000 µgTVOC m-3 [19]. 1.3. Elimination of Volatile Organic Compounds (VOCs) Generally, methods to improve the indoor air quality include a combination of actions such as removing/controlling the VOCs sources, increasing the ventilation rates and cleaning the indoor air. Source control/removal is only possible when the pollutants are known and when their control is technically or economically feasible, which scarcely happens. New substances are constantly detected and classified as hazardous, many sources can release compounds for years and many other air pollutants are yet to be discovered [20, 21]. Increasing ventilation rates by means of leaving doors or windows opened, is the easiest procedure to ensure indoor air quality above satisfactory levels. However due to outdoor weather, external pollution conditions or even issues related to security, safety in high buildings, climate control, or noise that is not always possible [18, 22]. Consequently, installing a forced ventilation system is one of the most common procedures used for air treatment [23]. Side by side with forced ventilation is the inherent energy consumption which has been strongly encouraged in the EU to be reduced. In this sense, purifying Chapter 1 14 and carbonates as final products. In another study, Martyanov and Klabunde [72] reported the PCO of gaseous 2-chloroethyl ethyl sulfide, the main component of mustard gas. Oxidation of this compound resulted in the formation of intermediates of different toxicity levels with potential negative impact in human health. Nonetheless, such intermediates can undergo further oxidation leading to less toxic compounds. Smith [73] mentioned the possible incorporation of photocatalytic surfaces on protective garments and masks for the protection of emergency responders and Grandcolas et al. [59] reported the impregnation of photocatalyst in military textiles for CWA removal. The above-mentioned ability to inactivate and kill bioaerosols such as bacteria, viruses, and spores makes PCO attractive in the mitigation of bioterrorism activities. For instances, Kau et al. [60] studied the inactivation of anthrax spores through PCO in mice over nitrogen or carbon doped TiO2 under visible light illumination. Based on the obtained results, Kau et al. [60] suggested that PCO could directly inactivate the lethal toxin, the major virulence factor of B. anthracis. PCO can also be employed as air cleaning technology for livestock buildings maintaining good air quality and preventing contamination and microbial spoilage of food and dairy products [74, 75]. In a study conducted by Cho et al. [74], PCO was used in the non-thermal disinfection of fresh vegetables through inactivation of foodborne pathogenic bacteria as E. Coli, Salmonella Typhimurium and B cereus in fresh carrots. Maneerat and Hayata [61] also employed PCO towards the inactivation of fungal activity of P. expansum, an organism responsible for postharvest rot development in fruits such as tomatoes and lemons. The incorporation of photocatalysts, specially TiO2 (by far the most used photocatalyst [76]) into construction materials has also been an object of intense research not only for decontamination and de-polluting purposes but also for the development of self-cleaning and selfdisinfecting materials. These construction materials include cement mortar [62, 77-79], tiles [63, 77, 80, 81], paving blocks [82-84], window glasses [85-87], composite sheets [88, 89], wall papers [64, 88, 90, 91] and paints [65-67, 92]. A pioneering work on cementitious materials with photocatalytic activity was performed by Murata et al. [79] evaluating the NOx degradation over TiO2 loaded concrete blocks. Later, in a similar study, Cassar [62] reported synergetic effects towards photoabatement of NOx of the combined use of TiO2 and cement: approximately 40 % of NO was converted with 70 % of NO2 selectivity. Introduction 15 Marcos et al. [63] reported for the first time the deposition of TiO2 layers on common ceramic glazed tiles using the screen-printing process, which is a low cost and common technique used for decoration in the ceramic industry. However, despite the good results achieved in terms of photocatalytic degradation of Orange II solution, the material’s surface had an unpleasant appearance, a high degree of roughness and it was hard to clean. Earlier, in 1995, Matsubara et al. [64] developed a TiO2-containing paper presenting photocatalytic properties. These authors investigated its photocatalytic activity by measuring the decomposition of gaseous acetaldehyde under a weak UV light radiation. The highest quantum yield obtained with the TiO2-containing paper was 90 %, ca. two times larger than that of commercial TiO2. However, the cellulosic fibre matrix of the paper can be easily damaged by the process. Considering the fact that almost all surfaces in urban areas can be painted, paint coatings are especially attractive as support for photocatalysts. For instances, Bygott et al. [65] reported a field trial in London, close to a school children playground, where an area of 300 m2 of walls was painted with a silicate-based paint incorporating 7.5 wt.% of photocatalytic TiO2. The results showed a daily NOx abatement of ca. 4.5 g in about 10000 m3 of air around the school children playground [65]. Maggos et al. [66] reported NOx depollution tests in an artificially closed parking area, which was polluted by a car exhaust during the testing period; they observed a reduction of 19 % and 20 % for NO and NO2, respectively. Salthammer and Fuhrmannh [92] studied the photocatalytic efficiency of two different types of commercially available wall paints in a 1 m3 test chamber, with and without air exchange, using artificial daylight. The results showed that formaldehyde was photo-oxidised under static conditions, while for typical VOCs, under dynamic conditions, no significant photocatalytic activity was observed. In another realoutdoor conditions study, Ângelo et al. [67] determined the NO photoabatement efficiency of photocatalytic paints when irradiated by sunlight attaining ca. 95 % of NO conversion for an initial NO concentration of 100 ppbv. 1.6. Common semiconductors used as photocatalysts Semiconductors as metal oxides (TiO2, ZnO, WO3, SnO2, Fe2O3) or metal chalcogenides (CdS, ZnS, WSe2) are often employed as heterogeneous photocatalysts in photocatalytic processes [38, 93]. A semiconductor should fulfil the following characteristics to be considered efficient: i. Able to be activated by sunlight, i.e., visible radiation or UVA; Chapter 1 16 ii. Biologically and chemically inert; iii. Not prone to photocorrosion and resistant to the attack of reagents; iv. Insoluble in water; v. Low toxicity; vi. Availability at low cost. Moreover, the redox potential of the photogenerated valence band hole must be positive enough to generate HO⦁ radicals in order to oxidize the contaminant and the redox potential of the conductance band electron should be negative enough to reduce oxygen to peroxide or other species [94]. The redox potential of some common semiconductors together with the band-gap energy is represented in Figure 1.2. Figure 1.2. Band positions (top of valence band and bottom of conduction band) for several common semiconductors together with the band-gap energy [95]. However, some semiconductors present specific issues impairing their use in photocatalytic applications. For example, CdS and CdSe undergo photoanodic corrosion in aqueous media and show toxicity [36]. Hematite (α-Fe2O3) absorbs in the visible region, but shows much lower photocatalytic activity when compared to TiO2 or ZnO, due to corrosion or formation of short-lived metal-to-ligand or ligand-to-metal charge transfer states [36]. ZnO and TiO2 present similar bandgap energies; however, the former becomes unstable when illuminated in aqueous solutions, yielding Zn(OH)2 on the particle surface and leading to the deactivation of the catalyst [96]. WO3 has also been investigated as photocatalyst but due to relatively low conduction band level (+ 0.5 V/SHE, standard hydrogen electrode) is generally less photocatalytic active than TiO2 [97, 98]. TiO2 (also known as titania), stands out as one of the most photoactive semiconductor-based catalyst, fulfilling all the referred criteria. -3 -2 -1 0 1 2 3 4 5 Redox potential [V] TiO2 3.1 eV SnO2 3.9 eV ZnO 2.8 eV ZnS 3.7 eV WO3 2.8 eV CdS 2.5 eV GaP 2.2 eV CdSe 1.7 eV O2/ ⦁OH Introduction 17 1.6.1. The preferred semiconductor: titanium dioxide (TiO2) TiO2 exists in at least eleven different polymorphic forms, three of which are the well-known and abundant minerals rutile (auburn, bronze and black), anatase (black, blue, red, etc.) and brookite (bronze and black). Rutile (tetragonal, space group P42/mnm) is a crystallization product of igneous rocks whereas anatase (tetragonal, space group I41/amd) and brookite (orthorhombic, space group Pbca) occur as accessory minerals in alteration assemblages and sedimentary and metamorphic rocks [99-103]. There are also five high-pressure TiO2 form materials: an orthorhombic α-PbO2–like form TiO2 (II) and a monoclinic baddeleyite-like form, both found at the Ries crater in Bavaria [104, 105]; the other three are orthorhombic cotunnite (PbCl2)-like form TiO2-OII [106], orthorhombic TiO2-OI [107] and cubic TiO2 form [108]. Other three forms that are metastable phases and can be produced synthetically are: monoclinic TiO2(B) [109], tetragonal hollandite-like form TiO2(H) [110] and orthorhombic ramsdellite-like form TiO2(R) [111]. The base unit of the TiO2 crystalline structure is an oxygen atoms-based octahedron centred on a titanium atom resulting in the molecular formula of (TiO6)2as it is represented in Figure 1.3. Figure 1.3. TiO2 crystallographic phases of anatase (a), rutile (b) and brookite (c) (adapted from Carp et al. [112]). Anatase and rutile share the same crystalline system (tetragonal system) but differ in the arrangement of the octahedrons and distortion. In anatase the octahedrons are connected by their vertices while in rutile the edges are connected. In both structures, Ti4+ species are surrounded in its first coordination sphere by six O2species in a distorted octahedral coordination. In anatase, Ti-Ti distances are longer than in rutile, whereas Ti-O distances are lower for anatase [113, 114]. These differences reflect on the electronic structure of both allotropic forms: the forbidden bandwidth of rutile is 2.8 - 3.0 eV while anatase is 3.2 eV. Rutile is the thermodynamically stable a) b) c) Chapter 1 18 form of TiO2 at any temperature and above 650 ºC brookite and anatase forms are converted into rutile [115]. Below 600 ºC the formation of anatase is kinetically favoured [116]. Notwithstanding, Zhang and Banfield [117] showed the superior stability of anatase form in comparison to rutile when the crystal size is smaller than 14 nm. Brookite [118] has an orthorhombic structure which is more complex than anatase and rutile forms structures. The comparison of the volume occupied by the base unit of the TiO2 crystalline structure of the three allotropic forms shows that the compactness of the TiO2 structures decreases following the order: rutile > brookite > anatase. Moreover, both vertices and edges are connected. Aside from the influence of the crystallographic structure of TiO2 on the photocatalytic efficiency, other factors such as TiO2 specific surface area, crystallinity and crystallite size may contribute to the photocatalytic properties of TiO2 [119]. For instance, higher specific surface areas promote higher photocatalytic Likewise, higher crystallinity suggests the possibility of lower density in surface defects increasing the charge recombination time. These two desired features are difficult to be obtained simultaneously since increasing crystallinity (as a result of, e.g., higher heat treatment) is often achieved with a subsequent loss of specific surface area. In this sense, the anatase phase has usually the best photocatalytic performances, while the rutile phase is the most stable phase with crystallite sizes higher than 14 nm [117, 120]. There are already a number of commercial photocatalytic TiO2 powders available on the market but that possessing better photoactivity in many reaction systems is TiO2 Degussa P25 (commercial name Aeroxide® TiO2 P25, after Evonik Industries taking over, in 2007, to be its Chemicals Business Area). This commercial TiO2 material consists in a highly dispersed powder manufactured according to the patented Aerosil® process. It is composed by a mixture of 80 % of anatase and 20 % of rutile, with a specific surface area of 50 m2 g-1 and an average particle size of 21 nm [121]. TiO2 P25 has been used as a standard photocatalyst in several research works [122, 123]. The activity obtained for this type of material, in comparison to that of pure anatase, has been extensively studied [124-126]. Despite the fact that it is still subject of discussion, it is believed that its activity comes from the formation of heterojunctions between the two types of crystalline forms, anatase and rutile [127] when employed as photocatalytic films. Indeed, rutile, owing to its lower conduction band, may slightly absorb light in the visible range and thus serving as a photosensitizer to the TiO2 particles of anatase structure. Also, the respective positions and the difference between the higher energy levels of the conduction bands of the two phases may produce a transfer of electrons from the anatase towards rutile. Due to their spatial charge separation, the transfer phenomenon of the photogenerated electrons from anatase conduction band towards the rutile conduction band of TiO2 will prevent charge recombination [128]. Introduction 19 Regarding the synthesis to the preparation conditions, TiO2 can be prepared via two types of methods [129]: gas-phase and liquid-phase. Gas-phase methods include flame hydrolysis [130] or oxidation at high temperatures of TiCl4 [131] and decomposition of titanium alkoxides (Ti(OR)4) [132]. As example, TiO2 P25 has been prepared by mixing gasified TiCl4 with reactants gases (hydrogen and oxygen) and burning at high temperatures. The reaction is described in eq. 1.9: TiCl4(g) + 2H2(g) + O2(g) → TiO2(s) + 4HCl(g) 1.9 Concerning liquid-phase methods, hydrolysis of TiCl4 in liquid-phase has been explored to manufacture TiO2 in a fine particulate form. Solvothermal and hydrothermal methods [133, 134], sol-gel [135, 136], water-in-oil microemulsion [137, 138], combustion and electrochemical synthesis [139, 140] and precipitation and co-precipitation [141, 142] have been the most popular techniques to synthesize TiO2 materials. 1.6.2. Strategies to enhance the TiO2 photocatalytic properties TiO2 has been for the last decades the most used and most efficient semiconductor in photocatalytic applications, particularly in the degradation of VOCs. However, this material presents three main limitations: (i) the threshold of absorption band does not allow visible light absorption, using only UV radiation which corresponds to 3 – 5 % of the solar spectrum reaching the Earth; (ii) the degradation process has a low quantum yield; (iii) the fast recombination of the charge carriers is in the order of nanoseconds. Chemical and physical modification of the TiO2 by doping metal and non-metal ions into the TiO2 lattice, deposition of transition metals, dye photosensitization or coupling with other semiconductors have been used over the recent years to overcome TiO2 limitations. 1.6.2.1. Increasing the life time of charge carriers The modification of TiO2 surface with metals, the coupling of the TiO2 with other semiconductors or the change over the morphology of the TiO2, are always to enhance the transfer and separation of photogenerated charges in TiO2 and subsequently, the life time of charge carriers is increased. Metal deposition onto TiO2 surface: Schottky barrier Chapter 1 20 The addition of a metal to a semiconductor changes the semiconductor surface properties by creating a barrier between the metal and the semiconductor junction, known as Schottky barrier (Figure 1.4). Figure 1.4. Photoexcitation and charge transfer in a metalmodified TiO2 photocatalyst (adapted from Serpone et al. [143]). The Schottky barrier promotes a better transfer of photogenerated electrons at TiO2 surface towards electron acceptor species, such as oxygen leading to a decrease in the electronhole recombination and, subsequently, to a more efficient charge separation [37]. Thus, the improved separation of charges by the presence of the metal increases the photocatalytic process efficiency [144-147], and ultimately increases the quantum yield. The positive effect of metal modification is highly dependent on the employed metal: the work function must be higher than the electron affinity of the semiconductor. Species like Ag, Pt, Pd, Au, Cr, among others are pointed out as very good doping metals for this purpose [142, 148-151]. The enhancement in TiO2 reactivity was first observed for the photoconversion of H2O to H2 and O2 through Pt/TiO2 [148]. However, this type of modification has also shown negative effects such as the reduction of the TiO2 bandgap energy resulting in a decrease in the energy threshold required to activate the photocatalyst [149]. It has also been reported an optimum metal content above which the metal acts as a charge recombination centre resulting in a reduction of the overall photocatalytic efficiency [37]. Coupling with other semiconductors: heterojunction The coupling between TiO2 and another semiconductor with different energy levels for their corresponding conduction and valence bands, leads to the formation of an interface between two semiconductor materials, called heterojunction. The existence of such interface will increase the photocatalytic process efficiency, reducing the probability of recombination of photogenerated charges through an irreversible spatial separation of the charge carriers. Besides the increase of the life time of the charge carriers, the structure of the interface (i.e. geometry of the particles, the TiO2 eh+ h+ eConduction band (CB) Valence band (VB) Schottky barrier Metal Introduction 21 surface contact between particles and the particles size [152]) also represents an important feature to enhance the efficiency of the interfacial electron transfer. In this sense, several coupled semiconductors have been intensively used for the degradation of pollutants such as TiO2/CdS, TiO2/SnO2, TiO2/ZnO, Bi2S3/TiO2 and TiO2/WO3 [152-155]. The evidence of a hole transfer between coupled semiconductors, together with a vectorial one-way displacement of both electrons and holes from one semiconductor to another, were first reported by Serpone et al. [152]. During the photocatalytic oxidation studies of several phenol-like molecules by using TiO2/CdS coupled semiconductors, these authors concluded that the enhancement of the interfacial charge transfer to adsorbed substrates occurs via two mechanisms: in the first one, a photogenerated electron of the illuminated semiconductor with a more negative conduction band is injected into the less negative conduction band of the other semiconductor while the hole remains in the former; in the second mechanism both semiconductors are activated resulting in a simultaneous electron transfer from the first semiconductor to the second as well as hole transfer from the second semiconductor to the first one (Figure 1.5). As a result, electrons are accumulated at the lower lying conduction band of one semiconductor, while the holes accumulate at the higher valence band of the other semiconductor [152]. Figure 1.5. Photoexcitation and charge transfer between two light active semiconductors: (a) injection of an ein the TiO2 conduction band and injection of a h+ in the valence band of the other semiconductor; (b) injection of an ein the conduction band of the other semiconductor and injection of a h+ in the TiO2 valence band (adapted from Serpone et al. [152]). Modifying TiO2 morphology: specific surface area As mentioned above the surface area of the photocatalyst is an important factor to be considered in the design of efficient photocatalysts. Successful attempts were made at synthesizing different forms of TiO2-based nanostructures with increased surface areas [118, 156, 157]. These include one dimensional nanotubes, nanorods, and two dimensional nanosheets, as h+ Conduction band ehν1 Pure TiO2 eValence band ehν2 < hν1 Semiconductor h+ h+ h+ Conduction band ehν1 Pure TiO2 Valence band h+ hν2 < hν1 Semiconductor eeh+ a) b) Chapter 1 22 shown in Figure 1.6. Titania nanotubes are one of the widely studied forms of titanium dioxide and have shown promising results over the past few years [158-162]. A detailed overviewing regarding this particular topic will be made further ahead in this introduction. Figure 1.6. Different morphologies of TiO2 nanostructures. 1.6.2.2. Red shifting of TiO2 absorption: towards solar applications Due to the rather large bandgap energy of TiO2 (~3.2 eV), photocatalytic-related processes are limited to the UV radiation (λ < 400 nm). Considering the natural sunlight, the UV radiation fraction represents only 3 – 5 % of the solar spectrum [163]. Shifting the TiO2 absorption spectrum into the visible light range, where the exploitable fraction of the solar spectrum is near 42 % seems to be advantageous from an energetic and economic point of view. In fact, over the recent years, several approaches have been presented in the literature to shift the absorption onset of TiO2 to the visible region: metal [164-182] and non-metal doping of TiO2 [183-194] , reduction of TiO2 with plasma treatments [195, 196], and sensitization with organic dyes [197, 198]. Metal and non-metal ion doping In recent years an extensive research focused on visible-light induced photocatalysis by metal ion-doped semiconductors since some of these materials have shown a stretched absorption spectra into visible-light region. This phenomenon has been explained by the excitation of electrons of dopant ions to the conduction band of the semiconductor, i.e., a metal to conduction band charge transfer. Alkaline Earth metal ions (magnesium and barium) [164-166], transition metal ions (e.g. chromium, iron, cobalt, nickel, niobium, vanadium, ruthenium, platinum and gold) [167-173], post transition metal ions (tin, bismuth, indium, …) [174-177] and rare earth metal ions (e.g. lanthanum, cerium, ytterbium, praseodymium, …) [178-182], have been nanosheets nanotubes nanorods Introduction 23 investigated as potential dopants for visible-light induced photocatalysis. However, metal ion dopant can also serve as a recombination centre, resulting in decreased photocatalytic activities as already explained previously [37, 149, 169, 175]. First reports of anion-doped TiO2 began to appear since the mid-1980s to early 1990s [183, 184, 199], but it was only in 2001 that anion-doping was considered a prelude to produce second generation materials in the study reported by Asahi et al. [190] dealing with visible-light active nitrogen (N) -doped TiO2. N-doping has the potential to increase the photoactivity under UV radiation and, more importantly, under visible-light in comparison to the pure TiO2 (Figure 1.7a). Subsequent studies reported several other visible-light active N-doped TiO2 materials [185189], together with carbon (C) -doped TiO2 [190-192] and sulphur (S) -doped TiO2 [193, 194]. It was originally proposed that N-doping of TiO2 can shift its photo-response into the visible region by mixing p states of nitrogen with 2p states of lattice oxygen and, thus, increasing the photocatalytic activity by narrowing the TiO2 bandgap as shown in (Figure 1.7b) [190, 193, 200]. Figure 1.7. Mechanisms proposals for the changes that may occur to the bandgap electronic structure of N-doped TiO2: (a) undoped TiO2; (b) bandgap narrowing as a result of VB broadening; (c) introduction of localized states above VB or below CB; (d) electronic transitions from localized states near VB to the corresponding excited states for Ti3+ (Jahn-Teller split 2T2  2E) and F+ (equivalent to a single electron associated with the O vacancy) centres; (e) sensitization by compounds containing nitrogen species (adapted from Serpone [163]). However, other studies proposed different ideas for the visible light absorption of Ndoped TiO2. One is that the nitrogen species originate localized N2p states above the valence band and the electronic transitions from localized N2p state to the conduction band (Figure 1.7c) [187]. Another hypothesis is that the visible-light absorption band of N-doped TiO2 is ascribed to the localized states of the oxygen deficiencies caused by nitrogen doping instead of the dopant itself (Figure 1.7c) [201]. In 2006, Serpone [163] proposed that the visible-light absorption band was originated by the formation of colour centres in the bandgap, resulting from the heat treatment or the photostimulation process (Figure 1.7d). In addition, Mitoraj and Kisch [202] suggested a CB VB Eg Pure TiO2 a) b) c) CB VB CB VB CB VB F+* F+ 2T2 }2E d) CB VB Melamine condensation products acting as sensitizers e) Chapter 1 30 Different investigations oriented to post-synthesis thermal treatments have been reported as these treatments have a strong influence on crystallinity, morphology, specific surface area and porosity of the nanotubes. According to the temperature of calcination, it has been demonstrated the formation of different crystal phases including brookite, anatase, rutile, Na2Ti3O7, Na2Ti6O13. Suzuki and Yoshikawa [262] and Armstrong et al. [263] found the existence of monoclinic TiO2 (B) free of anatase, respectively, after the thermal treatment at 800 ºC of TNTs and 400 – 600 ºC of nanowires. Poudel et al. [264] indicated that the rutile phase begins to crystallize at 800 ºC, well below the transformation temperature of 925 ºC for bulk anatase TiO2 nanopowder but well above the transformation temperature of 580 ºC reported by Varghese et al. [265] for TNTs prepared by anodization. Also, a change from nanotubes to nanowire morphology was observed at 650 ºC of annealing temperature. It is also for this range of calcination temperature that was observed the total transformation of the tubular structure in TiO2 nanoparticles, although some authors reported a similar phenomenon at temperatures of 300 ºC [266]. This deterioration of the morphology was attributed to removal of hydroxyl groups with the inter-sheets temperature. Despite the general beneficial effects of TiO2 nanotubes thermal treatment on the photocatalytic efficiencies, it is also reported that such treatment may impair physical aspects as BET surface area and pore volume. In this sense, other post-synthesis treatments have been explored to increase the TiO2 nanotubes activity without the undesirable effects of pore blockage, elimination of surface OH groups and of unstable tube morphology. Bavykin et al. [246] reported the formation of a stable rutile phase TiO2 nanotubes after a series of acid-immersed treatment periods. They ascribed this phenomenon to the low rate phase change. Nian and Teng [267] performed a similar study finding that only anatase phase appears at pH 2.2 while anatase along with brookite can be observed at pH 8.2; at pH 5.6 rod morphology was obtained. Yu et al. [268] revealed the formation of fibre-like structures with anatase phase after a similar posthydrothermal treatment. 1.7.3. Enhancing titania nanotube photocatalysts Despite the excellent morphologic properties of TNT, they share the same limitations of TiO2 nanoparticles. Several approaches similar to those found for TiO2 nanoparticles have been attempted to limit the recombination of the photogenerated electron-hole pairs and to broaden the photocatalytic activity from the ultra-violet region to the visible region. Metal and non-metal ion doping, coupling with other semiconductors are among other approaches used for enhancing TNTs attributes. Kukovecz’s group [269, 270] and Hsu et al. [271] reported the synthesis of CdS/TNTs by modifying the precursor as a mixing solution of Na2S/NaOH. They stated that the uniform particle size and extensive tube coverage by CdS nanoparticles attributed to the Introduction 31 homogeneous solution phase of the Cd–EDTA complex [270]. Colmenares et al. [272], in previous studies, used a series of metal and non-metal dopants, including chromium (Cr), iron (Fe), manganese (Mn), nickel (Ni), platinum (Pt) and vanadium (V), to dope TNTs for photocatalysis applications. The results showed that the new materials have higher activity and efficiency in the photooxidation of pollutants under visible light, but only up to certain level of doping. Beyond the optimum doping level, the photocatalytic activity of TNTs decreased [272]. Kim et al. [273] produced open-ended Ni-doped titanate nanotubes by hydrothermal treatment with several hundred nanometres of length and 5 – 6 nm of inner diameter and 1011 nm of outer diameter. Hsieh et al. [274] reported the photocatalytic degradation of basic violet 10 (BV10) using cobalt-doped titanate nanotubes. They also pointed out that the band gap of Co-TNTs (2.14 eV) is much lower than that of the commercial Degussa P25 photocatalyst (3.20 eV). Dong et al. [275] synthesized highly ordered N-doped TNTs by annealing anodized TNTs with ammonia at 500 °C. The nanotubes diameter ranged between 60 - 80 nm and the wall thickness was around 20 nm. Detailed analysis of the nanotubes structure revealed that the phase transformation temperature from anatase to rutile decreases after nitrogen doping. 1.7.4. Applications Some recent research works where TNTs and their derived materials were used in photocatalytic applications are listed in Table 1.3. Although the development of TNTs have been mostly directed towards photocatalytic applications, several other applications can be found in literature surveys. Supported TNTs for water-shift gas (WGS) reactions and CO2 hydrogenation, ion-exchange/adsorption, DSSCs, among others, have been the focus of several researchers. The first successful photoactivation of TNTs was accomplish by Hodos et al. [269] through supporting CdS particles in the these TNTs. Idakieva et al. [287] used gold particles supported on TNTs and performed the water–gas shift (WGS) reaction. This reaction was four times more efficient over Au/TNTs than over Au/Al2O3. Pt/Au nanosized particles supported on TNTs were also used in a study conducted by Chien et al. [288] dealing with CO2 hydrogenation and CO oxidation. TNTs impregnated with copper were employed in a study where NO conversion was evaluated [289]. Wang et al. [290] disclosed that after dispersing a monolayer of benzoic acid molecules onto the TNTs surface, i.e. acting as a carrier agent, these molecules can react with the hydroxyl groups of TNTs forming carboxylic species. Chapter 1 32 Table 1.3. Photocatalytic applications of TNTs and their derived materials. Authors Photocatalyst Model pollutant (physical state) Photocatalytic performance (Type of radiation) Zhang et al. [230] TNTs Propylene (gaseous) TNTs < P25 TiO2 (UVA) Nakahira et al. [248] TNTs Formaldehyde (aqueous) TNTs > P25 TiO2 (UV) Štengl et al. [276] TNTs 4-Chlorophenol (aqueous) TNTs < P25 TiO2 (UVA) Jing et al. [277] TNTs Escherichia Coli (aqueous) TNTs < TiO2 (UV) Xu et al. [278] Zn-TNTs Methyl orange (aqueous) Zn-TNTs > TiO2 (UV) Hsieh et al. [274] Co-TNTs Basic violet 10 (aqueous) Co-TNTs > P25 TiO2 (Visible) Zhao et al. [279] Au or Pt-TNTs Methyl orange (aqueous) Au (or Pt)-TNTs > P25 TiO2 (UVA) Song et al. [280] ITNTs Phenol (aqueous) I-TNTs > I-TiO2, TNTs, P25 TiO2 (Solar) Asapu et al. [281] P-TNTs Rhodamine B (aqueous) P-TNTs > TNTs and TiO2 P25 (UV) Zhang et al. [282] S-TNTs Glyphosate (aqueous) S-TNTs > TNTs (UV) Geng et al. [283] N-TNTs Methylene blue (aqueous) N-TNTs > TNTs (Solar) Shen et al. [284] N-TNTs Methylene blue Sulfosalicylic acid (aqueous) N-TNTs > P25 TiO2 (Visible) Xu et al. [285] N-TNTs Phenol Methyl orange (aqueous) N-TNTs > P25 TiO2 (Visible). Yamin et al. [286] WO3-TNTs Butanone (gaseous) WO3-TNTs > TNT (UVA) WO3-TNTs > TNT (Solar) Sun and Li [291] were the first authors reporting the influence of the intercalation of transition metals in metal-substituted TNTs for ion-exchange. This phenomenon was ascribed to the electrostatic interactions between the negatively charged host lattice of the nanotube and the positively charged cationic ions. Also, Umek et al. [292] indicated that Na atoms along with the hydrolysed surface of nanoribbons can transform adsorbed NO2 into NO3 and NO. The anatase phase TNTs can also be applied in lithium-ion batteries, as reported by Zhang’s group [293, 294]. During their investigations on TNTs electrochemical properties, these authors found that such structure provides an excellent cycling stability for lithium intercalation. In another study, Yu and Zhang [295] determined the electrochemical capacitance and voltammetric current of vanadium oxide/titanate composite nanorods, revealing better results with the composites than with pure V2O5. Furthermore, in a study conducted by Hu et al. [296], it was demonstrated that carbonized TNTs could significantly improve the performance of Pd-based electrocatalysts for ethanol oxidation in alkaline media. They have suggested that the open Introduction 33 mesoporous lattices of TNTs can make the Pd/TiO2C electrocatalyst structure loose and porous increasing the surface active sites and, consequently, the activity for ethanol oxidation. Another promising application was reported by Kim et al. [297], where TNTs films were fabricated on F–SnO2 coated glass (FTO) via electrophoretic deposition. The photocurrent densities of the DSSCs gradually increased with the annealing temperature (450 – 500 ºC). Above 500 ºC the photocurrent densities are lower due to the thermal limitation of the FTO substrate, as well as due to the decrease of the TNTs surface area. In the field of biocompatibility studies, Kubota et al. [298] applied TNTs to bone repair in filling defective areas of bones. They observed a newly formed bone of Ca–TNTs after a week of being implanted in the femur of a Wistar rat. The study also revealed that Ca–TNTs induced a fast acquisition and development of osteoblast and bone tissues and a better bone regeneration ability in comparison with other clinically-used biomaterials, like hydroxyapatite and β– tricalcium phosphate. As regards smart windows, displays, and optical memories, Miao et al. [299] used modified Ag/AgCl–TNTs as photochromism materials exhibiting multicolour photochromism corresponding to that of incident light. 1.8. Substrates for immobilization of photocatalyst powders Photocatalyst suspended powders are often used in aqueous media photocatalysis owing to an efficient contact between the catalyst and the organic pollutant. Immobilization of photocatalysts on an adequate substrate is mandatory in air treatment due to the problems associated with catalyst fluidization [300] and separation/recovery [301-303]. However, increased difficulty in structured catalyst preparation and decreased photocatalytic efficiency, mainly due to a less exposed area per catalyst weight unit, are typical resulting disadvantages [304]. Various substrates already used as photocatalyst supports for air and water treatment are listed in Table 1.4, together with the respective configurations. Chapter 1 34 Table 1.4. Substrates and configurations used as photocatalyst supports. Substrate material Configuration Reference Glass Deposited on reactor wall/lamp *# [305-307] Raschig rings/Beads # [50, 308-310] Fibre mesh *# [308, 311, 312] Plate * [313, 314] Quartz Plate # [315, 316] Sand # [317, 318] Optical fibre *# [319, 320] Silica Monolith * [321] Gel # [322, 323] Plate # [324] Silicon Plate # [324, 325] Wafer # [326] Metal Stainless steel plate # [315, 327] Anodized iron plate # [328] Foam# [329] Polymer Optical fibre* [330] Poly(ethylene terephtalate) (PET), acrylic resin (AC), polycarbonate (PC) wafers# [325] Tedlar®, parylene and PE (polyethylene) stripes# [331] Cellulose microspheres# [332] PS (polystyrene) micro-capsules# [333] Polyurethane foam *# [334, 335] Monoliths * [336, 337] Paper Sheets *# [51, 64] Cotton Fabric # [59, 338] Ceramic Monoliths * [339-341] Foam # [342] Membranes # [343, 344] Zeolites # [168, 345] Activated carbon Powder # [346] Fibres # [347, 348] * for air treatment applications # for water treatment applications Although several investigations have been conducted over the recent years, the pursuit for an optimum substrate to support the photocatalyst has not met yet an end. Choosing the substrate is not a trivial task in photocatalysis and the material structure, dimension and optical properties are important features in order to achieve high photocatalytic activity. Based on Pozzo et al. [349] concepts, the ideal substrate to support a given photocatalyst should have the following properties: i. promote good photocatalyst adherence; ii. chemically inert, photo, chemical and mechanical-resistant; iii. transparent to UV radiation; Introduction 35 iv. low pressure drop, promoting a turbulent regime avoiding the formation of dead volumes and the limitations to mass transfer; v. high surface area; vi. light weight and easy handling; vii. low cost. Considering the above-mentioned features, quartz/glass and polymers have stood out as supporting materials for photocatalytic applications. Affixing the powders to the reactor internal walls [306] or to flat plates [313] are interesting procedures at lab-scale, but low mass transfer limits the flow rates. Raschig rings [50, 309], small glass tube pieces, and massive glass beads are the most common materials used for supporting the photocatalysts. Possessing high area/volume ratio and random distribution, Raschig rings and glass beads ensure a turbulent flow regime and subsequently a good contact between the gas stream and the photocatalyst. However, high pressure drop and dead volumes are the main disadvantages that are faced by this type of supports. On the other hand, monolith-like form polymers (Figure 1.10) cause low pressure drop, do not promote the occurrence of dead volumes and are UV-transparent contrarily to the ceramic or metal ones which are opaque to radiation [340, 341]. Moreover, this type of material is cheap, light-weighted and easy to handle, attracting a great interest for photocatalytic applications. Thinwalled honeycomb or monolithic structures of polyethylene terephthalate (PET) and cellulose acetate (CA), which are commercially available, are some alternatives. High thermal sensitivity, low adherence and low resistance to photooxidation are the main disadvantages of these materials. Figure 1.10. Cellulose acetate polymer: (a) commercially available monolith-like structure – TIMAX CA50-9/S, Wacotech GmbH & Co. K.G.; (b) transmittance in the UV-Vis range. 1.8.1. Routes for photocatalyst immobilization Two main routes have been explored to immobilize the photocatalyst on any supporting substrate [308, 349]: i) manipulating a previously made photocatalyst powder in a suitable 0 20 40 60 80 100 240 340 440 540 Transmitance [%] Wavelenght [nm] 18 mm Chapter 1 36 dispersing agent [307, 308, 330] or ii) in situ synthesis of the photocatalyst during the deposition process. The former is the simplest way to obtain a support coating and the latter is the result of a combined series of physical and chemical transformations of a precursor, such as a titanium salt (usually an alkoxide), in an adequate solvent and/or controlling acid-base conditions. The main techniques are: i) Chemical vapour deposition (CVD) [350, 351] – the substrate is exposed to volatile precursors which react at the substrate surface producing a thin film of the photocatalyst. This process can be performed under atmospheric pressure (APCVD), low pressure (LPCVD) or ultrahigh vacuum (UHCVD); ii) Physical vapour deposition (PVD) [352-355] – a vaporized form of the photocatalyst is condensated at the surface of the substrate under high-temperature vacuum or plasma sputter bombardment; As the sputtered material is not in its thermodynamic equilibrium state, tend to affix on the substrate material; iii) Spray pyrolysis deposition [356, 357] – a thin photocatalytic film is deposited by spraying a solution on a heated surface. The mechanism behind this technique is similar to that of CVD; iv) Liquid-phase sol-gel, hydrothermal, solvothermal, etc., methods. Other techniques based on vapour-phase deposition have also been used, such as ion implantation [358], molecular beam epitaxy [359] and dynamic ion beam mixing [360]. However, despite of providing an efficient control in the film growth (resulting in high purity materials), the very high energy consumption of such techniques is considered a great disadvantage from the economical point of view. Therefore, simple routes of synthesis such as sol-gel and hydrothermal/solvothermal methods solution routes are the most commonly employed techniques not only for the materials synthesis but also for affixing the catalyst onto the substrate surface or matrix. In this sense, different methods can be carried out to accomplish the deposition of the catalyst onto the surface of a substrate such as: i) Dip-coating - the substrate is immersed and emerged into a liquid solution containing the coating precursor at a constant rate; ii) Spin-coating – a small amount of the aqueous-phase coating material is added on the centre of the substrate while the substrate is rotated at high speed in order to spread the coating material by centrifugal force; iii) Spray – the coating material is sprayed onto the substrate surface; iv) Electrophoresis – developed for ceramic coatings, this technique can be used if the substrate is conductive: a potential is applied between a counter-electrode and the Introduction 37 substrate (both with the same size, shape and inside the coating material) forming a thin film at the substrate surface. In this case, the substrate acts as a cathode. 1.9. Photocatalytic reactors Over the years, different reactor geometries and configurations have been used in photocatalytic applications. A categorization based on the modus operandi (batch, with or without recirculation, and one-pass continuous flow) or based on their geometry (the most representative are the flat plate, tubular and annular) will help distinguishing the reactors employed for air treatment [361, 362]. They can also be categorized based on the radiation source or by the way the photocatalyst is introduced into the reactor. Choosing a reactor predicates a hard task since entails several considerations that have to be taken into account in order to enhance the photocatalytic efficiency, and at the same time, to obtain comparable and repeatable measurements. Batch reactors are an interesting choice for lab-scale applications when seeking repeatable results in controlled conditions, of relevance to study the fundamentals of the process. Most of the batch reactors operate with recirculation of gas flow providing, in this way, an uniform concentration of reactants and products and, thus, enabling kinetic studies and comparison of newly produced photocatalysts [58, 363]. Alternatively, one-pass continuous flow reactors are suitable for applications where a continuous source of pollutants is available, specially indicated for industrial facilities where large amounts of gases have to be handled. Albeit the application on industrial context, continuous flow reactors can also be employed at labsale where gaseous inlet and outlet streams are constantly monitored by techniques such as CGMS, GC-FID, GC-TCD or FT-IR. [364, 365]. 1.9.1. Flat plate reactors (FPRs) Flat plate reactors are suitable for lab-scale applications and, in particular, for kinetic studies [313, 366]. Consisting of a thin layer of photocatalyst-coated on a flat substrate (usually made of glass or metal), the polluted gas stream flows parallel and on the top of the photocatalyst (Figure 1.11). Chapter 1 38 Figure 1.11. Schematic representation of a FPR (adapted from Águia et al. [367]) The light source can be either placed within the embodiment of the reactor, or outside the reactor where the photocatalyst is irradiated through a glass window. However, a high ratio between the photocatalyst surface and the flow rate is mandatory in order to avoid mass transfer limitations resulting from a faulty contact between the photocatalyst and the pollutants. A possible and interesting approach to overcome this limitation is to employ a multi-flat plate reactor as described by Leung et al. [314] – Figure 1.12. Figure 1.12. Schematic representation of a multi-parallel FPR (adapted from Leung et al. [314]). 1.9.2. Tubular reactors Due to the simplicity of assembling and handling, tubular reactors are probably the most common in photocatalytic processes. Consisting of one – single tubular reactor – or more than one glass tube – multitubular reactor – the gas stream flows along the axis, the photocatalyst can Photocatalytic film Mixture chambers Inlet gas Outlet gas Reaction chamber Radiation source Photocatalyst coated plates Inlet gas Outlet gas Window Introduction 39 be employed under several forms such as thin coated film on its inner wall, fluidized particles, or coated on supporting substrate. The light sources are located outside the tube, in a parallel configuration relative to its axis. Powder layered tubular reactor (PLR) One of the first types of photocatalytic reactors employed for air treatment was the powder layer tubular reactor (PLR), in the seventies and beginning of the eighties, by Formenti et al. [368] for partial oxidation of isobutene. Albeit its simplicity and easy catalyst replacement, avoiding mass transfer limitations, Lewandowski and Ollis [369] pointed out that the PLR was not suitable for full scale commercial applications. Due to high flow rates usually employed in industrial/commercial applications the catalyst particles may be displaced into the gas stream. Fluidized bed reactor (FBR) The fluidized bed tubular reactor (FBR) was also one of the first reactors to be used in air treatment [300], besides the liquid-solid slurry reactors for water treatment (Figure 1.13). They are characterized by an upward stream which will lift and suspend the photocatalyst particles, enhancing the contact between the surface of the photocatalyst and the pollutant. Figure 1.13. Schematic representation of a FBR (adapted from Dibble and Raupp [300]). Glass frit Pinch clamp Catalytic bed Glass frit O-ring Inlet gas Outlet gas Outlet gas Radiation source Chapter 1 46 photonic yield, low efficiency under visible light (380 < λ < 740 nm) and the requirement of a substrate to support the photoactive catalyst are the main challenges that this work will try to tackle and overcome. The main objective of this thesis is to evaluate the efficiency of solar gas-phase heterogeneous photocatalysis towards the elimination of pollutants such as VOCs present mainly in indoor air atmospheres. For this purpose, a laband a pilot-scale, single-pass, continuous-flow annular photocatalytic reactors are proposed. To accomplish the main objective, several partial objectives must be addressed: i. Synthesis and characterization of TiO2 nanotubes as well as nitrogen modified TiO2 nanotubes and nanoparticles; ii. Preparation of TiO2-based materials and their immobilization as thin films on different supports using the dip-coating technique; iii. Evaluation of photocatalytic activity of the prepared photocatalysts and its dependence on different operational conditions (flow rate – Qfeed; pollutant concentration – Cvoc, feed; relative humidity – RH, irradiance – I; presence/absence oxygen concentration) using a lab-scale single-pass continuous-flow annular photoreactor; iv. Description of the photocatalytic oxidation mechanisms; v. Evaluation of the photocatalytic activity of selected photocatalysts in a pilot-scale photoreactor for continuous removal of volatile organic compounds. Considering the extent of the objectives, the present thesis was organized in five parts. Part I includes this chapter which is an introductory section, wherein the problematic of volatile organic compounds present in both indoor and outdoor air, as well as current and potential decontamination procedures, are discussed. The main concepts of gas-phase heterogeneous photocatalysis are presented and complemented with an extensive literature survey. Part II contains Chapter 2 and 3 where it is reported the photocatalytic conversion of perchloroethylene and n-decane over different commercially available TiO2. Chapter 4 and 5 deal with construction materials (paint) as TiO2 supporting matrix for the removal of PCE and n-decane and are both included in Part III. Part IV reports the photocatalytic activity of titania nanoparticles (Chapter 6) and titania nanotubes (Chapter 7) both modified with nitrogen towards the elimination of pollutants whether present in water or in air. Finally Part V deals with the application of heterogeneous photocatalysis in a pilot-scale towards a continuous photo-oxidation of VOCs (Chapter 8) and suggestions of future work and conclusions (Chapter 9). A detailed description of each chapter is given in the following. Introduction 47 In Chapter 2 is reported the rate of PCE photocatalytic oxidation with UV-TiO2 using a single-pass continuous flow annular reactor under UVC radiation. The effects of PCE concentration, feed flow rate, and water vapour content on the rate of PCO were investigated and several kinetic models were tested. Also, by-products formation was weighed; according to the isolated intermediates, a reaction mechanism was proposed for PCE gas-phase photocatalytic oxidation. In Chapter 3 the influence of the feed flow rate, pollutant concentration, relative humidity, presence or absence of gas-phase molecular oxygen and incident irradiance on photocatalytic conversions of PCE and n-decane was evaluated. Within this chapter the continuous-flow annular photoreactor was assembled with monolithic structure made of cellulose acetate coated with different commercially available photoactive powders and exposed to simulated solar radiation. A comparison in terms of photocatalytic efficiency between the photoactive powders tested was also assessed. In Chapter 4 the photooxidation of PCE was studied in the annular photoreactor under simulated solar radiation employing two different configurations of a monolithic structure of cellulose acetate coated with an active TiO2-based paint. The influence of the structure configuration and different experimental conditions, namely feed flow rate, PCE concentration, relative humidity in the system, absence of oxygen and incident irradiance on the PCE conversion was evaluated. Also, based on the intermediates identified, a reaction mechanism was proposed for PCE gas-phase photooxidation. Chapter 5 reports the n-decane photocatalytic oxidation carried out in the annular photoreactor under simulated solar irradiation and employing the cellulose acetate monoliths coated with a water-based exterior vinyl paint with photocatalytic properties as the catalytic bed. The influence of the feed flow rate, n-decane concentration, relative humidity, and incident irradiance on the n-decane degradation kinetics was assessed. Also, a phenomenological reaction rate model of the n-decane photocatalytic oxidation was proposed and assessed. Finally, reaction by-products were identified and, based on those compounds, a reaction mechanism was formulated. Chapter 6 reports a study where nitrogen modified TiO2 powders were modified with different urea amounts and calcined at different temperatures, in order to optimize the method for the synthesis of an active photocatalyst. Several characterizations techniques were performed on the prepared materials. The photocatalytic activity of these nitrogen (N) modified TiO2 materials was evaluated in the gas-phase perchloroethylene PCO under simulated solar radiation. The photocatalytic activity of these N modified TiO2 materials was also evaluated in the degradation Chapter 1 48 of diphenhydramine, an emerging water pollutant under visible light illumination (λ = 430 nm) and in the inactivation of Escherichia coli bacteria under UVA radiation. In Chapter 7 high aspect ratio titanate nanotubes (TiNT) synthesized with an alkaline hydrothermal treatment and the modification of the TiNT with nitrogen are reported. The effect of calcination temperature and nitrogen content on the structure of the nanotubes was also assessed. The photocatalytic efficiency of the prepared materials was evaluated for degradation of methylethylketone (MEK) and hydrogen sulfide (H2S) under UVA and solar light radiation. These gas-phase PCO reactions were performed in a single-pass continuous annular photoreactor. Chapter 8 reports the design and construction of a low-cost pilot-scale continuous-flow annular photoreactor. The photochemical and photocatalytic oxidation of n-decane was evaluated as a function of the radiation source: sunlight or UVA illumination. Chapter 9 is dedicated to the final remarks where the main conclusions are discussed and subsequent suggestions for future work are presented. Introduction 49 1.11. References [1] M.L. Bell, D.L. Davis, T. Fletcher, Environ. Health Persp. 112 (2004) 6-8. [2] C.H. Eccleston, NEPA and Environmental Planning: Tools, Techniques, and Approaches for Practitioners, CRC Press, Richland, Washington, USA, 2008. [3] S. Wang, H.M. Ang, M.O. Tade, Environ. Int. 33 (2007) 694-705. [4] M.S. Zuraimi, C.-A. Roulet, K.W. Tham, S.C. Sekhar, K.W. David Cheong, N.H. Wong, K.H. Lee, Build. Environ. 41 (2006) 316-329. [5] J. Zhang, K.R. Smith, Br. Med. J. 68 (2003) 209-225. [6] P.S. Burge, Occup. Environ. Med. 61 (2004) 185-190. [7] L.A. Wallace, Annu. Rev. Energ. Env. 26 (2001) 269-301. [8] WHO, Ambient (outdoor) air quality and health - http://www.who.int/mediacentre/factsheets/fs313/en/ Last access 14th April 2014. [9] P. Wargocki, Indoor Air 9 (1999) 165-179. [10] J. Sundell, B. Anderson, K. Anderson, T. Lindvall, Indoor Air 3 (1993) 82-93. [11] J. Sundell, Indoor Air 14 (2004) 51-58. [12] P. Wolkoff, G.D. Nielsen, Atmos. Environ. 35 (2001) 4407-4417. [13] C. Yu, D. Crump, Build. Environ. 33 (1998) 357-374. [14] T. Salthammer, Indoor Air 7 (1997) 189-197. [15] H. Schleibinger, R. Keller, H. Rüden, The handbook of environmental chemistry - indoor air pollution by microorganisms and their metabolites (2004) 149-177. [16] B. Guieysse, C. Hort, V. Platel, R. Munoz, M. Ondarts, S. Revah, Biotechnol. Adv. 26 (2008) 398410. [17] R. Kostiainen, Atmos. Environ. 29 (1995) 693-702. [18] J.M. Daisey, A.T. Hodgson, W.J. Fisk, M.J. Mendell, J. Ten Brinke, Atmos. Environ. 28 (1994) 3557-3562. [19] K. Andersson, Indoor Air 7 (1997) 78-91. [20] T. Otake, J. Yoshinaga, Y. Yanagisawa, Environ. Sci. Technol. 35 (2001) 3099-3102. [21] D.C.G. Muir, P.H. Howard, Environ. Sci. Technol. 40 (2006) 7157-7166. [22] L.E. Ekberg, Atmos. Environ. 28 (1994) 3571-3575. [23] P. Wargocki, J. Sundell, W. Bischof, G. Brundrett, P.O. Fanger, F. Gyntelberg, S.O. Hanssen, P. Harrison, A. Pickering, O. Seppänen, P. Wouters, Indoor Air 12 (2002) 113-128. [24] B.C. Wolverton, R. McDonald, E.A. Watkins, Econ. Bot. 38 (1984) 224-228. [25] R. Orwell, R. Wood, J. Tarran, F. Torpy, M. Burchett, Water Air Soil Pollut. 157 (2004) 193-207. [26] R.L. Orwell, R.A. Wood, M.D. Burchett, J. Tarran, F. Torpy, Water Air Soil Pollut. 177 (2006) 5980. [27] R. Wood, M. Burchett, R. Alquezar, R. Orwell, J. Tarran, F. Torpy, Water Air Soil Pollut. 175 (2006) 163-180. [28] R. Muñoz, S. Villaverde, B. Guieysse, S. Revah, Biotechnol. Adv. 25 (2007) 410-422. [29] D.W. Park, S.W. Chun, J.Y. Jang, H.S. Kim, H.C. Woo, J.S. Chung, Catal. Today 44 (1998) 73-79. Chapter 1 50 [30] J. Hermia, S. Vigneron, Catal. Today 17 (1993) 349-358. [31] W.H. Glaze, J.-W. Kang, D.H. Chapin, Ozone-Sci. Eng. 9 (1987) 335-352. [32] J. Zhao, X. Yang, Build. Environ. 38 (2003) 645-654. [33] W.J. Kowalski, Immune building systems technology, McGraw-Hill, Michigan University 2003. [34] W.A. Jacoby, D.M. Blake, J.A. Fennell, J.E. Boulter, L.M. Vargo, M.C. George, S.K. Dolberg, J. Air Waste Manage. Assoc. 46 (1996) 891-898. [35] M. Schiavello, Electrochim. Acta 38 (1993) 11-14. [36] M.A. Fox, M.T. Dulay, Chem. Rev. 93 (1993) 341-357. [37] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735-758. [38] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69-96. [39] A. Furube, T. Asahi, H. Masuhara, H. Yamashita, M. Anpo, Chem. Phys. Lett. 336 (2001) 424-430. [40] V. Augugliaro, S. Coluccia, V. Loddo, L. Marchese, G. Martra, L. Palmisano, M. Schiavello, Appl. Catal. B: Environ. 20 (1999) 15-27. [41] E. Pelizzetti, C. Minero, Electrochim. Acta 38 (1993) 47-55. [42] Y.V. Kolen'ko, K.A. Kovnir, A.I. Gavrilov, A.V. Garshev, P.E. Meskin, B.R. Churagulov, M. Bouchard, C. Colbeau-Justin, O.I. Lebedev, G. Van Tendeloo, M. Yoshimura, J. Phys. Chem. B 109 (2005) 20303-20309. [43] F. Benoit-Marquié, U. Wilkenhöner, V. Simon, A.M. Braun, E. Oliveros, M.-T. Maurette, J. Photochem. Photobiol. A 132 (2000) 225-232. [44] L. Cermenati, P. Pichat, C. Guillard, A. Albini, J. Phys. Chem. B 101 (1997) 2650-2658. [45] D.Y. Goswami, J. Sol. Energy Eng. 119 (1997) 101-107. [46] D.S. Bhatkhande, V.G. Pangarkar, A.A.C.M. Beenackers, J. Chem. Technol. Biotechnol. 77 (2002) 102-116. [47] C. McCullagh, J. Robertson, D. Bahnemann, P. Robertson, Res. Chem. Intermed. 33 (2007) 359375. [48] D. Bahnemann, Sol. Energy 77 (2004) 445-459. [49] J.-M. Herrmann, Top. Catal. 34 (2005) 49-65. [50] M. Hernandez-Alonso, I. Tejedor-Tejedor, J. Coronado, J. Soria, M. Anderson, Thin Solid Films 502 (2006) 125-131. [51] C. Guillard, J. Disdier, C. Monnet, J. Dussaud, S. Malato, J. Blanco, M.I. Maldonado, J.-M. Herrmann, Appl. Catal. B: Environ. 46 (2003) 319-332. [52] M.L. Satuf, R.J. Brandi, A.E. Cassano, O.M. Alfano, Appl. Catal. B: Environ. 82 (2008) 37-49. [53] M.C. Canela, R.M. Alberici, W.F. Jardim, J. Photochem. Photobiol. A 112 (1998) 73-80. [54] A. Ginestet, D. Pugnet, J. Rowley, K. Bull, H. Yeomans, Indoor Air 15 (2005) 326-334. [55] T. Matsunaga, R. Tomoda, T. Nakajima, H. Wake, FEMS Microbiology Letters 29 (1985) 211-214. [56] P. Maness, S. Smolinski, D.M. Blake, Z. Huang, E.J. Wolfrum, W.A. Jacoby, Appl. Environ. Microbiol. 65 (1999) 4094-4098. [57] D.Y. Goswami, D.M. Trivedi, S.S. Block, Journal of Solar Energy Engineering, Transactions of the ASME 119 (1997) 92-96. [58] D.V. Kozlov, A.V. Vorontsov, P.G. Smirniotis, E.N. Savinov, Appl. Catal. B: Environ. 42 (2003) Introduction 51 77-87. [59] M. Grandcolas, L. Sinault, F. Mosset, A. Louvet, N. Keller, V. Keller, Appl. Catal. A: Gen. 391 (2011) 455-467. [60] J.-H. Kau, D.-S. Sun, H.-H. Huang, M.-S. Wong, H.-C. Lin, H.-H. Chang, PLoS ONE 4 (2009) e4167. [61] C. Maneerat, Y. Hayata, Int. J. Food Microbiol. 107 (2006) 99-103. [62] L. Cassar, MRS Bull. 29 (2004) 328-331. [63] P.S. Marcos, J. Marto, T. Trindade, J.A. Labrincha, J. Photochem. Photobiol. A 197 (2008) 125-131. [64] H. Matsubara, M. Takada, S. Koyama, K. Hashimoto, A. Fujishima, Chem. Lett. 24 (1995) 767-768. [65] C.E. Bygott, J.E. Maltby, J.L. Stratton, R. McIntyre, Photocatalytic coatings for the construction industry International RILEM Symposium on Photocatalysis, Environment and Construction Materials, 8–9 October, Florence, Italy, 2007, pp. 251-258. [66] T. Maggos, J.G. Bartzis, M. Liakou, C. Gobin, J. Hazard. Mater. 146 (2007) 668-673. [67] J. Ângelo, L. Andrade, A. Mendes, Manuscript submitted for publication (2014). [68] M. Anpo, Pure Appl. Chem. 72 (2000) 1265-1270. [69] M.C. Canela, R.M. Alberici, R.C.R. Sofia, M.N. Eberlin, W.F. Jardim, Environ. Sci. Technol. 33 (1999) 2788-2792. [70] J. Hall, J.J. Sangiovanni, H.H. Hollick, T.N. Obee, S.O. Hay, in: N.L. Nagda (Ed.), Air Quality and Comfort in Airliner Cabins, ASTM STP 1393, ASTM, West Conshohocken, USA, 2000, p. 294. [71] Z. Huang, P.C. Maness, D.M. Blake, E.J. Wolfrum, S.L. Smolinski, W.A. Jacoby, J. Photochem. Photobiol. A 130 (2000) 163-170. [72] I.N. Martyanov, K.J. Klabunde, Environ. Sci. Technol. 37 (2003) 3448-3453. [73] B.M. Smith, Chem. Soc. Rev. 37 (2008) 470-478. [74] M. Cho, Y. Choi, H. Park, K. Kim, G.J. Woo, J. Park, J. Food Protect. 70 (2007) 97-101. [75] Y. Kim, Y. Choi, S. Kim, J. Park, M. Chung, K.B. Song, I. Hwang, K. Kwon, J. Park, J. Food Protect. 72 (2009) 1916-1922. [76] C.O. Robichaud, A.E. Uyar, M.R. Darby, L.G. Zucker, M.R. Wiesner, Environ. Sci. Technol. 43 (2009) 4227-4233. [77] K. Demeestere, J. Dewulf, B. De Witte, A. Beeldens, H. Van Langenhove, Build. Environ. 43 (2008) 406-414. [78] A.M. Ramirez, K. Demeestere, N. De Belie, T. Mäntylä, E. Levänen, Build. Environ. 45 (2010) 832838. [79] Y. Murata, H. Tawara, H. Obata, K. Takeuchi, J. Adv. Oxid. Technol. 4 (1999) 227-230. [80] K. Murugan, R. Subasri, T.N. Rao, A.S. Gandhi, B.S. Murty, Prog. Org. Coat. 76 (2013) 1756-1760. [81] M.P. Seabra, R.R. Pires, J.A. Labrincha, Chem. Eng. J. 171 (2011) 692-702. [82] C.S. Poon, E. Cheung, Constr. Build. Mater. 21 (2007) 1746-1753. [83] C.S. Poon, E. Cheung, Performance of photo-catalytic paving blocks made from waste, Proceedings of the ICE - Waste and Resource Management, 2006, pp. 165-171. [84] A. Beeldens, Air purification by pavement blocks: final results of the research at the BRRC, Transport Research Arena Europe, Ljubljana, 2008. Chapter 1 52 [85] J.M. Langridge, R.J. Gustafsson, P.T. Griffiths, R.A. Cox, R.M. Lambert, R.L. Jones, Atmos. Environ. 43 (2009) 5128-5131. [86] D. Byun, Y. Jin, B. Kim, J. Kee Lee, D. Park, J. Hazard. Mater. 73 (2000) 199-206. [87] Y. Paz, Z. Luo, L. Rabenberg, A. Heller, J. Mater. Res. 10 (1995) 2842-2848. [88] H. Ichiura, T. Kitaoka, H. Tanaka, Chemosphere 50 (2003) 79-83. [89] H. Ichiura, T. Kitaoka, H. Tanaka, J. Mater. Sci. 37 (2002) 2937-2941. [90] J. Zhang, W. Liu, P. Wang, K. Qian, J. Environ. Chem. Eng. 1 (2013) 175-182. [91] H. Taoda, M. Fukaya, E. Watanabe, K. Tanaka, Mater. Sci. Forum 510-511 (2006) 22-25. [92] T. Salthammer, F. Fuhrmann, Environ. Sci. Technol. 41 (2007) 6573-6578. [93] P. Pichat, J.M. Herrmann, J. Disdier, M.N. Mozzanega, J. Phys. Chem. 83 (1979) 3122-3126. [94] A. Mills, R.H. Davies, D. Worsley, Chem. Soc. Rev. 22 (1993) 417-425. [95] M. Gratzel, Nature 414 (2001) 338-344. [96] D.W. Bahnemann, C. Kormann, M.R. Hoffmann, J. Phys. Chem. 91 (1987) 3789-3798. [97] I.M. Szilágyi, B. Fórizs, O. Rosseler, Á. Szegedi, P. Németh, P. Király, G. Tárkányi, B. Vajna, K. Varga-Josepovits, K. László, A.L. Tóth, P. Baranyai, M. Leskelä, J. Catal. 294 (2012) 119-127. [98] M.S. Bazarjani, M. Hojamberdiev, K. Morita, G. Zhu, G. Cherkashinin, C. Fasel, T. Herrmann, H. Breitzke, A. Gurlo, R. Riedel, J. Am. Chem. Soc. 135 (2013) 4467-4475. [99] J.F. Banfield, D.R. Veblen, D.J. Smith, Am. Mineral. 76 (1991) 343-353. [100] J.F. Banfield, D.R. Veblen, Am. Mineral. 76 (1991) 113-127. [101] J.F. Banfield, D.R. Veblen, Am. Mineral. 77 (1992) 545-557. [102] S. Bakardjieva, V. Stengl, L. Szatmary, J. Subrt, J. Lukac, N. Murafa, D. Niznansky, K. Cizek, J. Jirkovsky, N. Petrova, J. Mater. Chem. 16 (2006) 1709-1716. [103] Z. Zheng, H. Liu, J. Ye, J. Zhao, E.R. Waclawik, H. Zhu, J. Mol. Catal. A-Chem. 316 (2010) 75-82. [104] A.E. Goresy, M. Chen, L. Dubrovinsky, P. Gillet, G. Graup, Science 293 (2001) 1467-1470. [105] A.E. Goresy, M. Chen, P. Gillet, L. Dubrovinsky, G. Graup, R. Ahuja, Earth Planet. Sc. Lett. 192 (2001) 485-495. [106] L.S. Dubrovinsky, N.A. Dubrovinskaia, V. Swamy, J. Muscat, N.M. Harrison, R. Ahuja, B. Holm, B. Johansson, Nature 410 (2001). [107] N.A. Dubrovinskaia, L.S. Dubrovinsky, R. Ahuja, V.B. Prokopenko, V. Dmitriev, H.P. Weber, J.M. Osorio-Guillen, B. Johansson, Phys. Rev. Lett. 87 (2001) 275501. [108] M. Mattesini, J.S. de Almeida, L. Dubrovinsky, N. Dubrovinskaia, B. Johansson, R. Ahuja, Phys. Rev. B 70 (2004) 212101. [109] R. Marchand, L. Brohan, M. Tournoux, Mater. Res. Bull. 15 (1980) 1129-1133. [110] M. Latroche, L. Brohan, R. Marchand, M. Tournoux, J. Solid State Chem. 81 (1989) 78-82. [111] J. Akimoto, Y. Gotoh, Y. Oosawa, N. Nonose, T. Kumagai, K. Aoki, H. Takei, J. Solid State Chem. 113 (1994) 27-36. [112] O. Carp, C.L. Huisman, A. Reller, Prog. Solid State Ch. 32 (2004) 33-177. [113] D.T. Cromer, K. Herrington, J. Am. Chem. Soc. 77 (1955) 4708-4709. [114] S. Andersson, A.D. Wadsley, Nature 211 (1966) 581. [115] S. Sankar, K.G. Gopchandran, P. Kuppusami, S. Murugesan, Ceram. Int. 37 (2011) 3307-3315. Introduction 53 [116] F. Thevenet, O. Guaïtella, J.M. Herrmann, A. Rousseau, C. Guillard, Appl. Catal. B: Environ. 61 (2005) 58-68. [117] H. Zhang, J.F. Banfield, J. Mater. Chem. 8 (1998) 2073-2076. [118] U. Diebold, Surf. Sci. Rep. 48 (2003) 53-229. [119] X. Wang, L. Sø, R. Su, S. Wendt, P. Hald, A. Mamakhel, C. Yang, Y. Huang, B.B. Iversen, F. Besenbacher, J. Catal. 310 (2014) 100-108. [120] K. Tanaka, M.F.V. Capule, T. Hisanaga, Chem. Phys. Lett. 187 (1991) 73-76. [121] Evonik Industries (www.aerosil.com, acessed February 2014). [122] M. Hussain, R. Ceccarelli, D.L. Marchisio, D. Fino, N. Russo, F. Geobaldo, Chem. Eng. J. 157 (2010) 45-51. [123] B. Abramović, D. Šojić, V. Despotović, D. Vione, M. Pazzi, J. Csanádi, Appl. Catal. B: Environ. 105 (2011) 191-198. [124] A.G. Agrios, K.A. Gray, E. Weitz, Langmuir 19 (2003) 1402-1409. [125] A.G. Agrios, K.A. Gray, E. Weitz, Langmuir 20 (2004) 5911-5917. [126] D.C. Hurum, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 108 (2004) 16483-16487. [127] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 107 (2003) 45454549. [128] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem. 92 (1991) 178-190. [129] J. Mo, Y. Zhang, Q. Xu, J.J. Lamson, R. Zhao, Atmos. Environ. 43 (2009) 2229-2246. [130] A.J. Maira, K.L. Yeung, C.Y. Lee, P.L. Yue, C.K. Chan, J. Catal. 192 (2000) 185-196. [131] L.y. Zeatoun, D. Feke, Part. Part. Syst. Char. 22 (2005) 276-281. [132] J. Rubio, J. Oteo, M. Villegas, P. Duran, J. Mater. Sci. 32 (1997) 643-652. [133] H. Cheng, J. Ma, Z. Zhao, L. Qi, Chem. Mater. 7 (1995) 663-671. [134] R.-C. Xie, J. Shang, J. Mater. Sci. 42 (2007) 6583-6589. [135] J. Yu, X. Zhao, J. Du, W. Chen, J. Sol-Gel Sci. Technol. 17 (2000) 163-171. [136] Y.V. Kolen’ko, A.V. Garshev, B.R. Churagulov, S. Boujday, P. Portes, C. Colbeau-Justin, J. Photochem. Photobiol. A 172 (2005) 19-26. [137] P. Monnoyer, A. Fonseca, J.B. Nagy, Colloids and Surfaces A: Physicochemical and Engineering Aspects 100 (1995) 233-243. [138] M. Andersson, A. Kiselev, L. Österlund, A.E.C. Palmqvist, J. Phys. Chem. 111 (2007) 6789-6797. [139] B.R. Sankapal, S.D. Sartale, M.C. Lux-Steiner, A. Ennaoui, C. R. Chim. 9 (2006) 702-707. [140] Y. Kitamura, N. Okinaka, T. Shibayama, O.O.P. Mahaney, D. Kusano, B. Ohtani, T. Akiyama, Powder Technol. 176 (2007) 93-98. [141] J.H. Lee, Y.S. Yang, Mater. Chem. Phys. 93 (2005) 237-242. [142] M. Bellardita, M. Addamo, A. Di Paola, L. Palmisano, Chem. Phys. 339 (2007) 94-103. [143] N. Serpone, A.V. Emeline, S. Horikoshi, V.N. Kuznetsov, V.K. Ryabchuk, Photochem. Photobiol. Sci. 11 (2012) 1121-1150. [144] C.M. Wang, A. Heller, H. Gerischer, J. Am. Chem. Soc. 114 (1992) 5230-5234. [145] H. Kobayashi, K. Kishimoto, Y. Nakato, Surf. Sci. 306 (1994) 393-405. Chapter 1 54 [146] V. Subramanian, E. Wolf, P.V. Kamat, J. Phys. Chem. B 105 (2001) 11439-11446. [147] W. Chen, J.S. Zhang, Build. Environ. 43 (2008) 246-252. [148] S. Sato, J.M. White, Chem. Phys. Lett. 72 (1980) 83-86. [149] A. Di Paola, G. Marcì, L. Palmisano, M. Schiavello, K. Uosaki, S. Ikeda, B. Ohtani, J. Phys. Chem. B 106 (2001) 637-645. [150] J. Araña, J. Mol. Catal. A-Chem. 215 (2004) 153-160. [151] J.C. Colmenares, A. Magdziarz, D. Łomot, O. Chernyayeva, D. Lisovytskiy, Appl. Catal. B: Environ. 147 (2014) 624-632. [152] N. Serpone, P. Maruthamuthu, P. Pichat, E. Pelizzetti, H. Hidaka, J. Photochem. Photobiol. A 85 (1995) 247-255. [153] Y. Bessekhouad, D. Robert, J.V. Weber, J. Photochem. Photobiol. A 163 (2004) 569-580. [154] Y. Bessekhouad, N. Chaoui, M. Trzpit, N. Ghazzal, D. Robert, J.V. Weber, J. Photochem. Photobiol. A 183 (2006) 218-224. [155] C.F. Lin, C.H. Wu, Z.N. Onn, J. Hazard. Mater. 154 (2008) 1033-1039. [156] T. Rajh, J.M. Nedeljkovic, L.X. Chen, O. Poluektov, M.C. Thurnauer, J. Phys. Chem. B 103 (1999) 3515-3519. [157] T. Tachikawa, M. Fujitsuka, T. Majima, J. Phys. Chem. 111 (2007) 5259-5275. [158] T. Kasuga, M. Hiramatsu, A. Hoson, T. Sekino, K. Niihara, Langmuir 14 (1998) 3160-3163. [159] B.D. Yao, Y.F. Chan, X.Y. Zhang, W.F. Zhang, Z.Y. Yang, N. Wang, Appl. Phys. Lett. 82 (2003) 281-283. [160] H. Ou, S. Lo, Sep. Purif. Technol. 58 (2007) 179-191. [161] M. Qamar, C.R. Yoon, H.J. Oh, N.H. Lee, K. Park, D.H. Kim, K.S. Lee, W.J. Lee, S.J. Kim, Catal. Today 131 (2008) 3-14. [162] C.L. Wong, Y.N. Tan, A.R. Mohamed, J. Environ. Manage. 92 (2011) 1669-1680. [163] N. Serpone, J. Phys. Chem. B 110 (2006) 24287-24293. [164] M. Zamora, T. López, R. Gómez, M. Asomoza, R. Melendrez, Catal. Today 107–108 (2005) 289293. [165] J.E. Son, J. Chattopadhyay, D. Pak, Int. J. Hydrogen Energy 35 (2010) 420-427. [166] H. Feng, L.E. Yu, M.-H. Zhang, Mater. Res. Bull. 48 (2013) 672-681. [167] E. Borgarello, N. Serpone, G. Emo, R. Harris, E. Pelizzetti, C. Minero, Inorg. Chem. 25 (1986) 4499-4503. [168] L. Davydov, E.P. Reddy, P. France, P.G. Smirniotis, J. Catal. 203 (2001) 157-167. [169] F.B. Li, X.Z. Li, Appl. Catal. A: Gen. 228 (2002) 15-27. [170] Z. Du, C. Feng, Q. Li, Y. Zhao, X. Tai, Colloids and Surfaces A: Physicochemical and Engineering Aspects 315 (2008) 254-258. [171] Y.-H. Peng, G.-F. Huang, W.-Q. Huang, Adv. Powder Technol. 23 (2012) 8-12. [172] J.C. Rooke, T. Barakat, M.F. Finol, P. Billemont, G. De Weireld, Y. Li, R. Cousin, J.M. Giraudon, S. Siffert, J.F. Lamonier, B.L. Su, Appl. Catal. B: Environ. 142–143 (2013) 149-160. [173] S.N.R. Inturi, T. Boningari, M. Suidan, P.G. Smirniotis, Appl. Catal. B: Environ. 144 (2014) 333342. Introduction 55 [174] W.-J. Hong, M. Kang, Mater. Lett. 60 (2006) 1296-1305. [175] J. Yu, S. Liu, Z. Xiu, W. Yu, G. Feng, J. Alloys Compd. 461 (2008) L17-L19. [176] Y. Duan, N. Fu, Q. Zhang, Y. Fang, X. Zhou, Y. Lin, Electrochim. Acta 107 (2013) 473-480. [177] C.S. Chua, O.K. Tan, M.S. Tse, X. Ding, Thin Solid Films 544 (2013) 571-575. [178] S. Zhang, Z. Zheng, J. Wang, J. Chen, Chemosphere 65 (2006) 2282-2288. [179] S. Forissier, H. Roussel, C. Jimenez, O. Chaix, A. Pereira, A. Bensalah-Ledoux, J.L. Deschanvres, B. Moine, Energy Procedia 10 (2011) 192-196. [180] M. Boutinguiza, J. del Val, A. Riveiro, F. Lusquiños, F. Quintero, R. Comesaña, J. Pou, Phys. Procedia 41 (2013) 787-793. [181] C. Zhan, F. Chen, J. Yang, D. Dai, X. Cao, M. Zhong, J. Hazard. Mater. 267 (2014) 88-97. [182] L. Zheng, R. Lisiecki, W. Ryba-Romanowski, G. Aka, J. Di, D. Li, X. Xu, J. Xu, J. Lumin. 145 (2014) 547-552. [183] S. Sato, Chem. Phys. Lett. 123 (1986) 126-128. [184] N.C. Saha, H.G. Tompkins, J. Appl. Phys. 72 (1992) 3072-3079. [185] T. Morikawa, R. Asahi, T. Ohwaki, K. Aoki, Y. Taga, Jpn. J. Appl. Phys. 40 (2001) L561-L563. [186] T. Lindgren, J.M. Mwabora, E. Avendaño, J. Jonsson, A. Hoel, C.-G. Granqvist, S.-E. Lindquist, J. Phys. Chem. B 107 (2003) 5709-5716. [187] H. Irie, Y. Watanabe, K. Hashimoto, J. Phys. Chem. B 107 (2003) 5483-5486. [188] C. Di Valentin, G. Pacchioni, A. Selloni, Phys. Rev. B 70 (2004) 085116. [189] O. Diwald, T.L. Thompson, T. Zubkov, S.D. Walck, J.T. Yates, J. Phys. Chem. B 108 (2004) 60046008. [190] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001) 269-271. [191] S. Sakthivel, H. Kisch, Angew. Chem., Int. Ed. 42 (2003) 4908-4911. [192] H. Irie, Y. Watanabe, K. Hashimoto, Chem. Lett. 32 (2003) 772-773. [193] T. Umebayashi, T. Yamaki, H. Itoh, K. Asai, Appl. Phys. Lett. 81 (2002) 454. [194] T. Umebayashi, T. Yamaki, S. Yamamoto, A. Miyashita, S. Tanaka, T. Sumita, K. Asai, J. Appl. Phys. 93 (2003) 5156-5160. [195] K. Takeuchi, I. Nakamura, O. Matsumoto, S. Sugihara, M. Ando, T. Ihara, Chem. Lett. 29 (2000) 1354-1355. [196] T. Ihara, M. Miyoshi, M. Ando, S. Sugihara, Y. Iriyama, J. Mater. Sci. 36 (2001) 4201-4207. [197] J. Moon, C.Y. Yun, K.-W. Chung, M.-S. Kang, J. Yi, Catal. Today 87 (2003) 77-86. [198] C. Li, X. Yang, R. Chen, J. Pan, H. Tian, H. Zhu, X. Wang, A. Hagfeldt, L. Sun, Sol. Energ. Mat. Sol. C. 91 (2007) 1863-1871. [199] D.H. Lee, Y.S. Cho, W.I. Yi, T.S. Kim, J.K. Lee, H. Jin Jung, Appl. Phys. Lett. 66 (1995) 815-816. [200] T. Umebayashi, T. Yamaki, S. Tanaka, K. Asai, Chem. Lett. 32 (2003) 330. [201] T. Ihara, M. Miyoshi, Y. Iriyama, O. Matsumoto, S. Sugihara, Appl. Catal. B: Environ. 42 (2003) 403-409. [202] D. Mitoraj, H. Kisch, Angew. Chem., Int. Ed. 47 (2008) 9975-9978. [203] C. Pastravanu, I.F. Alexa, I. Cretescu, E. Popovici, Photocatalytic properties of N-doped TiO2. The effect of the synthesis procedure, 2010, pp. 533-536. Chapter 1 62 [377] H.-H. Kim, S.-M. Oh, A. Ogata, S. Futamura, Appl. Catal. B: Environ. 56 (2005) 213-220. [378] A.M. Vandenbroucke, R. Morent, N. De Geyter, C. Leys, J. Hazard. Mater. 195 (2011) 30-54. [379] F. Thevenet, O. Guaitella, E. Puzenat, J.M. Herrmann, A. Rousseau, C. Guillard, Catal. Today 122 (2007) 186-194. [380] N. Doucet, F. Bocquillon, O. Zahraa, M. Bouchy, Chemosphere 65 (2006) 1188-1196. [381] C. Nicolella, M. Rovatti, Chem. Eng. J. 69 (1998) 119-126. [382] M.R. Nimlos, W.A. Jacoby, D.M. Blake, T.A. Milne, Environ. Sci. Technol. 27 (1993) 732-740. [383] T. Tsuru, T. Kan-no, T. Yoshioka, M. Asaeda, J. Membrane Sci. 280 (2006) 156-162. [384] M.L. Sauer, D.F. Ollis, J. Catal. 149 (1994) 81-91. [385] T. Tsuru, T. Kan-no, T. Yoshioka, M. Asaeda, Catal. Today 82 (2003) 41-48. [386] S.S. Chin, K. Chiang, A.G. Fane, J. Membrane Sci. 275 (2006) 202-211. [387] R.E. Marinangeli, D.F. Ollis, AIChE J. 23 (1977) 415-426. [388] N.J. Peill, M.R. Hoffmann, Environ. Sci. Technol. 30 (1996) 2806-2812. [389] N.J. Peill, M.R. Hoffmann, Environ. Sci. Technol. 32 (1998) 398-404. [390] A.V. Vorontsov, V.P. Dubovitskaya, J. Catal. 221 (2004) 102-109. [391] W.A. Jacoby, D. Blake, R.D. Noble, C.A. Koval, J. Catal. 157 (1995) 87-96. [392] T.H.K. Lim, S. D. , Chemosphere 54 (2004) 305-312. [393] A. Collin, K.E. Wirth, M. Stroeder, Powder Technol. 190 (2009) 31-35. [394] G.B. Raupp, J.A. Nico, S. Annangi, R. Changrani, R. Annapragada, AIChE J. 43 (1997) 792-801. [395] J.-F. Wu, C.-H. Hung, C.-S. Yuan, J. Photochem. Photobiol. A 170 (2005) 299-306. [396] A.A. Assadi, A. Bouzaza, C. Vallet, D. Wolbert, Chem. Eng. J. [397] H.Q. Trinh, Y.S. Mok, Chem. Eng. J. 251 (2014) 199-206. [398] G. Imoberdorf, A. Cassano, H. Irazoqui, O. Alfano, Catal. Today 129 (2007) 118-126. [399] D. Bockelmann, D. Weichgrebe, R. Goslich, D. Bahnemann, Sol. Energ. Mat. Sol. C. 38 (1995) 441451. [400] D. Curcó, S. Malato, J. Blanco, J. Giménez, P. Marco, Sol. Energy 56 (1996) 387-400. [401] J. Blanco, S. Malato, P. Fernandez, A. Vidal, A. Morales, P. Trincado, J.C. Oliveira, C. Minero, M. Musci, C. Casalle, M. Brunotte, S. Tratzky, N. Dischinger, K.-H. Funken, C. Sattler, M. Vincent, M. Collares-Pereira, J.F. Mendes, C.M. Rangel, Sol. Energy 67 (1999) 317-330. [402] J.I. Ajona, A. Vidal, Sol. Energy 68 (2000) 109-120. [403] S.M. Rodrıguez, J.B. Gálvez, M.I.M. Rubio, P.F. Ibáñez, D.A. Padilla, M.C. Pereira, J.F. Mendes, J.C.d. Oliveira, Sol. Energy 77 (2004) 513-524. [404] E.R. Bandala, C.A. Arancibia-Bulnes, S.L. Orozco, C.A. Estrada, Sol. Energy 77 (2004) 503-512. [405] E.R. Bandala, C. Estrada, J. Sol. Energy Eng. 129 (2005) 22-26. Part II Chapter 2. Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor Chapter 3. Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 2. Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor In this chapter, gas-phase photocatalytic oxidation (PCO) of PCE (574 – 2442 ppm) was carried out in a single-pass continuousflow annular packed bed photoreactor under non-catalytic (UV) and catalytic (UV-TiO2) conditions. Under UVC radiation, PCE conversion was dependent on the photoreactor inner tube: conversion was negligible (~ 0 %) with a concentric soda-lime glass inner tube placed inside the photoreactor outer tube, or almost complete (98 %) with a concentric quartz filter. The PCE PCO under UV-TiO2 was studied using the photoreactor with the glass inner tube (mimicking solar radiation). Effects of PCE concentration, feed flow rate, and water vapour content on the PCE conversion through PCO were investigated. A complete mathematical model able to describe the effects of these operating parameters on the process performance was disclosed; six different kinetic rate equations were tested, suggesting that PCE and H2O molecules have to be considered in association with different specific active sites of the surface. Under steadystate conditions, by-products formation was also weighed; according to the isolated intermediates, a reaction mechanism was proposed for PCE gas-phase PCO. This chapter is based on the research article “Lopes, F.V.S., Monteiro, R.A.R., Silva, A.M.T., Silva, G.V., Faria, J.L., Mendes, A.M., Vilar, V.J.P., Boaventura, R.A.R., Insights into UV TiO2 Photocatalytic Degradation of PCE for Air Decontamination Systems, Chemical Engineering Journal, 204 (2012) 244-257, DOI: 10.1016/j.cej.2012.07.079”. Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 67 2.1. Introduction Several chlorinated volatile organic compounds (VOCs), such as perchloroethylene (PCE), have been widely used as solvents at industrial scale in dry cleaning factories, metal degreasing facilities, and plastic and fumigant manufacturing industries [1-3]. These extensive uses lead to water and air pollution, particularly in indoor environments [4, 5]. Due to its high toxicity and volatility and since it is suspected to be carcinogenic to humans and extremely persistent in the environment, PCE present in water and/or air must be removed. Instead of conventional techniques, such as adsorption or air stripping processes which only transfer the pollutants between phases, photocatalytic oxidation (PCO) in gas phase has become a wellestablished process for air decontamination [1, 2, 6-9], capable of complete PCE conversion/mineralization. Many reactor configurations have been used for the treatment of PCE in air streams by UV-TiO2 photocatalytic processes. Among the several types of reactors designed, honeycomb, monolith, plate, fluidized bed, packed bed, and annular tube flow reactors are the most representative [6, 8]. For the last decade, it has become clear that the reactor design should be consistent with the use of a photocatalyst of high specific surface area, and consider small passthrough channels, low air velocity, and direct irradiance of UV light on catalyst surface, allowing a large contact area and reducing mass transfer limitations [8]. However, the photocatalytic oxidation reaction can be influenced, not only by the catalyst nature, structure and morphology, reactor configuration, and radiation source but also by temperature, oxygen and pollutant concentration, water vapour content, flow rate (residence time) and surface velocity [4, 6, 8, 10, 11]. Within this chapter, a wide set of experimental conditions using a lab-scale facility consisting of a single-pass continuous-flow packed-bed annular photoreactor for air decontamination systems is tested in photochemical and photocatalytic oxidation of PCE. Ideally, for the analysis of an effective industrial application process, high volumes of polluted air must be treated, at different continuous feed stream conditions, and for fast and efficient gas-phase treatment (with a contact time of seconds) [12]. Therefore, PCE conversion through PCO was studied for the same reaction temperature as a function of three parameters: (a) pollutant concentration, (b) flow rate (residence time/surface velocity), and (c) humidity content. High concentrations of PCE (between 574 and 2442 ppm) were employed to test the efficiency of the photocatalytic material in extreme conditions of pollutant concentration. Moreover, the effect of these parameters on PCE conversion was assessed for a large range of air stream flow rates contaminated with PCE (59 - 300 cm3 min-1, measured at 298 K and 1 bar) as well as for different Chapter 2 68 water vapour contents (12 – 40 %, measured at 298 K and 1 bar). A complete mathematical model for air decontamination systems using a single-pass continuous-flow annular UV-photoreactor (glass spheres and TiO2 photocatalytic packed bed) was presented and different kinetic rate expressions were combined with the mathematical model to study the kinetics through PCO. It should be mentioned that, prior to the PCO experiments, the photochemical reaction of PCE was estimated under UVC radiation. In addition, two different inner tubes were tested (glass and quartz) with the aim of evaluating how important is the effect of the higher transmittance of quartz, when compared with glass, on the global process. By-products of VOC type formed on PCE conversion were analysed by GC/MSD. Finally, a reaction mechanism covering the formation of all identified by-products was proposed for both UV and UV-TiO2 conversion processes of PCE. All the operating parameters assessed within this work as well as the theoretical/experimental considerations for kinetic modelling and reaction mechanism will contribute for the better understanding of the photocatalytic PCE conversion process. 2.2. Experimental 2.2.1. Materials and chemicals The benchmark TiO2 photocatalyst (TiO2 P25 from Evonik®) was employed in the gas-phase photocatalytic experiments. This material consists of 80 % anatase and 20 % rutile crystalline phases, an average particle size of 30 nm, a BET specific surface area of 50 m2 g-1, and a band-gap energy of ~3.00 - 3.15 eV. The perchloroethylene analytical reagent (PCE; 99.5 %) used for the generation of contaminated air streams was purchased from Panreac Química S.A.U. and used without further purification (HCl: < 0.005 %; H2O: < 0.02 %; non-volatile matter: < 0.005 %). Deionized water was used for the generation of the water vapour content in the feed stream. All gases employed in this work were provided by Air Liquide: helium N50, nitrogen N50, and air K N50 (O2: 20 ± 1 %; H2O: < 3 ppm; CnHm: < 0.1 ppm; CO2: < 1 ppm; CO: < 1 ppm), each one with a minimum total purity of 99.999 %. 2.2.2. Photocatalytic experimental apparatus A schematic representation of the experimental unit used for this study is shown in Figure 2.1 and entails three main parts: feed generation (Figure 2.1a), the photoreactor (Figure 2.1b and Figure 2.3) and the photoreactor feed and exit stream analytic system (Figure 2.1c). Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 69 2.1.1.1. Feed generation The lab-scale apparatus used for generation of air streams containing PCE and water vapour is schematically represented in Figure 2.1a Figure 2.1. Schematic representation of the lab-scale experimental unit used for the study of decontamination of air contaminated with PCE: a) lab-scale facility used for the generation of air streams containing PCE and water vapour; b) single-pass continuous-flow annular UV-photoreactor; c) master gas chromatographic analysis system used for the analysis of the photoreactor feed and exit streams. The lab-scale apparatus comprises three mass flow controllers (El-Flow, Bronkhorst High-Tech B.V.) that allow the generation of air contaminated with different PCE concentrations and with different relative humidities. For that, an air stream saturated with water vapour is generated by flowing air through two Woulff bottles (supplied by Normax, Lda), one after the other (the second for trapping). To control the relative humidity of this stream, a pure air stream is used to dilute it in accordance to the needs/requirements. The relative humidity and temperature of the mixed stream are read using a relative humidity/temperature probe (9735, Testo), placed in a stainless steel cylinder connected throughout the line. This probe measures air relative humidity and temperature in the ranges of 253 - 343 K (± 0.3 K) and 0 - 100 % (± 2 %), respectively (connected to a data logger: 635-2, Testo). At the end, the air stream (with the desired water vapour content) can be mixed with any pollutant (for this study, PCE) in gas phase. Thus, another air stream passes through two extra Woulff bottles, consecutively, the first filled with a pure PCE liquid solution and the second for trapping. The PCE vapour content was fixed as a result of controlling the flow rate of the air stream and its temperature. Note that, the temperature of MGC MFC V0 V2 HTP Air (compressed air line) HTS Woulff bottles Woulff bottles MFC MFC V0 V2 FID – flame ionized detector FIH – flexible insulated heater HCS – heating/cooling system HTP – humidity/temperature probe HTS – humidity/temperature sensor H2G – hydrogen generator MFC – mass flow controller MGC – master gas chromatograph PR – photoreactor UVL – UV lamp SL – sampling loop TC – thermocouple V0 – one-way valve V2 – two-way valve V6 – six-way valve PP – peristaltic pump He N2 H2G Air V6 SL PP FID Exhaust CT TC HCS Computer (H2O) (VOC) a) V0 V0 FIH FIH CT TC FIH 606060 c) PR b) UV radiation UV radiation UVL Chapter 2 70 Woulff bottles was kept constant using a temperature controlling system (thermostatic bath GD100 R2, Grant Instruments). All tubing, fittings, connectors, adapters, and valves are made of stainless steel (Swagelok Company) to prevent unit deterioration with the highly VOCsconcentrated air streams employed in the experiments; the 1/4" stainless steel tubing and components are covered with a flexible insulated heater to avoid pollutant and water condensation. A routine in LabVIEW environment (NI Corporation) was designed to control/monitor the mass flow controllers and thermocouples throughout the experimental time. The work environment developed for that purpose can be seen in Figure 2.2. Figure 2.2. LabVIEW routine designed to control/monitor the mass flow controllers and thermocouples throughout the experimental time. 2.1.1.2. Lab-scale photoreactor A lab-scale single-pass continuous-flow annular UV-photoreactor prototype was designed and manufactured for the study of air decontamination systems. Figure 2.1b shows the schematic representation of the UV-photoreactor. It comprises two concentric tubes with a photocatalytic bed that fills the void in-between. The outer tube is a Pyrex-glass cylinder (Duran borosilicate glass 3.3, Schott-Rorhglas GmbH) and the inner tube, centred in the axial position along the bed, can be either a glass cylinder (Soda-lime glass, Linex) or a quartz cylinder (Quarzglas-Rohr, Quarzglastechnik, GmbH & Co KG). The catalytic bed is made of glass spheres (1401/7, Karl Hecht Assistent GmbH) packed with TiO2 catalyst filling the voids between the spheres. To avoid the release of the small catalyst particles into the atmosphere, in both sides of the photocatalytic bed, a layer of ~15 mm of glass wool was added. A detailed representation of Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 71 the photoreactor is illustrated in Figure 2.3. Figure 2.3. Detailed schematic representation of single-pass continuous flow packed bed annular photoreactor employed in the study of decontamination of air contaminated with PCE: a) side view; b) frontal view. Tubes dimensions and catalytic bed characteristics are described in Table 2.1. It should be mentioned that, each cap of the photoreactor (both inlet and outlet sides) has four equidistant inlets to ensure a better distribution of the feed stream throughout the reactor. Table 2.1. Photoreactor tubes dimensions employed in the gas-phase photooxidation of PCE under non-catalytic (UV) and catalytic (UV-TiO2) conditions; catalytic bed characteristics used in the PCE photocatalytic reaction. Photoreactor tubes Outer tube (Pyrex-glass tube) Lot [cm] 30.0 dot,e [cm] 5.00 dot,i [cm] 4.64 Inner tube (Glass or quartz tube) Lin [cm] 30.0 din,e [cm] 2.00 din,i [cm] 1.64 Catalytic bed Photocatalytic bed LR [cm] 20.0 VR [cm3] 275 ε 0.406 TiO2 P25 (Evonik®) mP25 [g] 15.3 ρP25 [g cm-3] 3.8 SBET [m2 g-1] 50 Crystal structure 80 % A/20 % R* Average particle size [nm] 21 Glass spheres dGS [cm] 0.71 VGS [cm3] 124 * A stands for Anatase and R stands for Rutile An UV lamp (TUV TL Mini 8W, Koninklijke Philips Electronics N.V.) was used as radiation source. This lamp is placed inside the inner tube; it features short-wave UV radiation with a peak centred at 253.7 nm (UVC). Prior to completely mount the reactor, the intensity of the Inner quartz/glass tube Outer Pyrex glass tube Exit stream UV lamp Feed stream Outer Pyrex glass tube Glass sphere Feed / Exit stream a) b) Glass spheres TiO2 powder filling the voids TiO2 powder filling the voids Glass wool Chapter 2 78                        1 1 1 1OH OHOHPCEPCE2 2 OHOHPCEPCE OHPCE1 PCE 2 22 22 2 C CKCK I CKCK CC r  2.11 where two kinetic coefficients are considered, α1 and α2 (expressed μmol m s-1 μM-1 and W-1 m2, respectively), and φ is the wavelength averaged quantum efficiency. All reported models assume that intermediates and/or reaction products do not influence PCE kinetics of conversion through PCO, considering that only PCE and H2O are the major species; the only exception is M-2 that disregards H2O adsorption. The mathematical model combined with each suggested rate expression was solved numerically in gPROMS environment (Process System Enterprise, London, UK), using the orthogonal collocation on finite elements method. The number of elements used was 90 with third order polynomials (two interior collocation points); an absolute and relative tolerance of 1 × 10-5 was considered. The estimation of the unknown kinetic and equilibrium parameters resulting from the application of models M-1 to M-6, eqs. (2.6) - (2.11), in the complete mathematical model (see eqs. (2.1) - (2.5), in Theoretical Section 2.3.1) were performed through a sequential quadratic programming algorithm (gPROMS, Process System Enterprise, London, UK). This algorithm is based on the Newton method for unconstrained optimization, employing the first order derivative to determine its search direction. 2.4. Results and discussion 2.4.1. UV photolysis of PCE Direct photolysis of PCE was evaluated in a lab-scale prototype (Figure 2.1). For the same experimental conditions, direct photolysis of PCE was studied employing the photoreactor with a quartz inner tube and without catalyst. VOCs in contaminated air streams (in this particular case, PCE) are able to absorb light over a wide range of wavelengths; however, the absorption is often more stronger at shorter wavelengths [30]. Figure 2.4a shows the transmissivity of both photoreactor glass and quartz inner tubes in comparison to the solar spectrum (solar spectrum reported by Malato et al. (2002) used as reference [30]). Thus, if a glass tube is used as inner jacket of the reactor, it will absorb most of the UV radiation (see Figure 2.4a). On the other hand, the quartz tube is transparent to UV short wavelengths (high energy), as shown in Figure 2.4a. Thus, using the latter, direct photolysis of PCE with UVC leads to ionization (Figure 2.4b), resulting in removal of chlorine Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 79 radicals from PCE molecules (~98 % of conversion). When the glass inner tube is used the shorter wavelengths of the radiation are absorbed by the glass tube, and no conversion of PCE is observed (Figure 2.4b); however, the apparatus with the glass inner tube (where the amount of UVC radiation that passes through the glass tube mimics the UV sunlight radiation; see Figure 2.4a) is particularly more interesting for real case applications: an UV source significantly increases the operation cost of the process [31]. Figure 2.4. a) Transmissivity of glass (-----) and quartz (─ ─) inner tubes compared to the solar spectrum at sea level (──, [30]). b) PCE photolysis using different photoreactor inner tubes: glass (──) and quartz (----); CPCE, feed = 1221 ppm, Qfeed* = 150 cm3 min-1, RH = 40 %, and T = 298 K; operation conditions reported in Table 2.2 * measured at 298 K and 1 bar. 2.4.2. UV-TiO2 photocatalytic conversion of PCE 2.4.2.1. Influence of operation parameters Besides the catalyst optical properties, PCE gas-phase photooxidation also depends on the optical properties of the materials used in the reactor construction. As mentioned before, prior to enter in the reactor, a large fraction of the UV radiation is absorbed by the glass inner tube of the photoreactor; it was measured an incident irradiance (I) of 0.8 W m-2, when the radiometer was placed on the outside of the glass inner tube and in contact with it. Afterwards, part of the remaining UV radiation is absorbed by the glass spheres; the rest is absorbed by the TiO2 P25 catalyst powder, transmitted through the voids between the catalyst particles and/or glass spheres, as well as reflected on the surface of both materials. Since an irradiance of ~ 0.4 W m-2 was measured outside of the photoreactor and considering both “shading effects” and the fact that a gradient in UV intensity exists in the radial direction due to a simple geometric 2 × (dot,i - din,e)-1 factor [13-17], an averaged irradiance of 0.5 W m-2 was assumed throughout the photocatalytic Chapter 2 80 bed. UV-TiO2 photocatalytic conversion tests of PCE were carried out in the lab-scale photoreactor previously described in section 2.1.1.2. Note that no changes in PCE level were observed in the absence of radiation throughout all experiments. PCE conversion in PCO (CPCE, Exit / CPCE, Feed) was determined under steady-state conditions as a function of three parameters: PCE concentration (574 – 2442 ppm), feed flow rate (59 - 300 cm3 min-1) and humidity content (12 – 40 %), all measured at 298 K and 1 bar. Figure 2.5 discloses the effects PCE conversion through PCO for air feed streams contaminated with different concentrations of PCE (Table 2.2: runs 1 - 6), Figure 2.6 shows PCE contaminated air applying different feed flow rates (Table 2.2: runs 1, 7 - 10), and Figure 2.7 when applying different humidity contents (Table 2.2: runs 1, 11 - 13). Figure 2.5 shows that the PCE conversion decreases with the feed concentration of PCE (CPCE, feed). Figure 2.5. Effect on PCE conversion fraction through PCO (CPCE, exit / CPCE, feed, at steady-state conditions) for air feed streams contaminated with different concentrations of PCE [CPCE, feed]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); Qfeeda = 150 cm3 min-1, RHa = 40 %, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2 (runs 1-6) [a measured at 298 K and 1 bar]. For a 4.25 fold increase in PCE concentration feed, the PCE conversion decreases about 1.8 times. This is due to the fact that more PCE molecules passes through the catalytic bed 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 500 1000 1500 2000 2500 CPCE, exit / CPCE, feed CPCE, feed [ppm] Exp. M-1 M-2 M-3 M-4 M-5 M-6 Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 81 without being decomposed, either by the hydroxyl or chlorine radicals, or by direct surface oxidation (higher amount of molecules passing through “shadowed” zones) [26]. This “shading effect” can be minimized: e.g., employing transparent monoliths for supporting a thin catalytic film [13-17]. According to Figure 2.6, increasing the feed flow rate (Qfeed), higher surface velocities are achieved while the residence time decreases. A substantial reduction of the residence time (~5 fold) will appreciatively reduce the contact time between PCE molecules both with bulk hydroxyl radicals and surface electron-hole pairs, leading to poor degrees of conversion. Competitive adsorption between H2O and PCE molecules to the available hydroxyl adsorption sites, may affect the process efficiency [10]. Therefore, the effect on PCE conversion in PCO for PCE contaminated air applying different humidity contents was also evaluated. Figure 2.6. Effect on PCE conversion fraction through PCO (CPCE, Exit / CPCE, Feed, at steady-state conditions) for air contaminated with PCE applying different feed flow rates [Qfeed]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); CPCE, feed = 1221 ppm, RHa = 40 %, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2 (runs 1, 7-10) [a measured at 298 K and 1 bar]. Figure 2.7 shows that reducing the relative humidity (RH), the PCE conversion merely decreased ~1.2 times. This observation was also reported by Yamazaki et al. (2001) [32], after stating that, if mole fraction of H2O is 50 - 60 times higher than that of PCE, the conversion decreases due to PCE/H2O competition for the same TiO2 surface adsorption sites. In our case the 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 50 100 150 200 250 300 CPCE, exit / CPCE, feed Qfeed [cm3min-1]a Exp. M-1 M-2 M-3 M-4 M-5 M-6 Chapter 2 82 mole fraction of H2O at 40 % of RH is approximately 7 times higher than that of PCE. Clearly, water molecules are a key factor for the formation of hydroxyl radicals but, within the RH range studied, this parameter plays a secondary role in comparison to other parameters. Furthermore, it must be pointed out that, since the PCE conversion by UV photolysis (using this apparatus scheme: reactor inner tube of glass) was negligible, the photoactivity of the TiO2 P25 catalyst promoted the gas-phase photooxidation of PCE (see Figures 2.5 to 2.7). Figure 2.7. Effect on PCE conversion fraction through PCO (CPCE, exit / CPCE, feed, at steady-state conditions) for air contaminated with PCE applying different humidity contents [RH]: experimental points () and M-1 (——), M-2 (······), M-3 (-----), M-4 (‒ ‒ ‒), M-5 (──), M-6 (─ · ·); CPCE, feed = 1221 ppm, Qfeeda = 150 cm3 min-1, T = 298 K, and I = 0.8 W m-2; operation conditions reported in Table 2.2 (runs 1, 11-13) [a measured at 298 K and 1 bar]. 2.4.2.2. Mathematical modelling Within this work, six kinetics expressions were employed in a complete mathematical model for simulating the PCE kinetics of conversion through PCO (see eqs. (2.1)-(2.11)). Table 2.3 reports the kinetic and adsorption equilibrium parameters resulting from model fitting to experimental data in the UV-TiO2 PCO of PCE. Our measurements on the PCE conversion, when applying feed streams of air with different pollutant concentrations, are qualitatively well fitted by all models, except M-1 and M-2 (see Figure 2.5). However, M-3 (and also M-1) fails to produce an acceptable fit when different feed flow rates are used (Figure 2.6). In addition, M-3 completely fails to fit the change in humidity (Figure 2.7). According to Table 2.3 and Figures 2.5 to 2.7, 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 10 17,5 25 32,5 40 CPCE, exit / CPCE, feed RH [%]a Exp. M-1 M-2 M-3 M-4 M-5 M-6 Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 83 M-4, M-5 and M-6 produce the best fit results, suggesting that regardless the competition of PCE and H2O molecules to the surface, they must be considered independent and targeting different active sites of the photocatalyst. From a purely statistical stand point, the M-5 model produces the best fit for the existing data; however, for the sake of simplicity, due to the least number of required parameters the L-H bimolecular non-competitive two types of sites model (M-4) is more appropriate to describe the experimental data. Table 2.3. Kinetic and adsorption equilibrium parameters resulting from the application of rate expressions M-1 to M-6 in the complete mathematical model, including the statistical parameter squared correlation coefficient (R2) and sum of squared residuals between experimental and calculated rates (S2R). Kinetic parameters Model Par. Values Units R2 S2R × 106 [mol2 m-4 s-2] M-1 k’ 1.16 × 10-3 mol m-2 s-1 W-1 m2 M0.76 0.978 6.06 α 0.67 – β 0.09 – M-2 k 1.44 × 10-3 mol m-2 s-1 W-1 m2 0.982 4.46 KPCE 12.1 M-1 M-3 k 2.63 × 10-3 mol m-2 s-1 W-1 m2 0.938 17.0 KPCE 1.06 × 10-3 M-1 KH2O 1.57 × 10-2 M-1 M-4 k 2.33 × 10-2 mol m-2 s-1 W-1 m2 0.988 3.34 KPCE 0.492 M-1 KH2O 2.49 M-1 M-5 k 1.10 × 10-2 mol m-2 s-1 W-1 m2 0.991 3.22 KPCE,1 1.21 M-1 KH2O,1 1.01 × 10-7 M-1 KPCE,2 9.98 × 10-5 M-1 KH2O,2 68.1 M-1 M-6 α1 4.21 mol m s-1 M-1 0.984 4.94 α2 × φ 8.57 × 10-3 W-1 m2 KPCE 14.0 M-1 KH2O 9.97 × 10-9 M-1 2.4.3. PCO reaction of PCE: product analysis and reaction mechanism Several studies have demonstrated that the photochemical or photocatalytic conversion of PCE can occur through two initiation steps: hydroxyl radical (HO•) and chlorine radical (Cl•) pathways [2, 32-34]. Under UV radiation, the main pathways for the gas-phase photooxidation of PCE are: the photochemical or photocatalytic HO• and Cl• formation, the complete mineralization of PCE and its by-products, and the termination reactions (see Scheme 2.1) [2]. Chapter 2 84 Scheme 2.1 Main pathways for the gas-phase photooxidation of PCE under UV radiation; adapted from Petit et al. (2007) [2]. Yamazaki et al. (2001, 2004) [32, 33] reported theoretical calculations which have demonstrated that HO• addition to PCE is more exothermic than the addition of Cl• (43.8 and 9.2 kcal mol-1, respectively); the authors also stated that Cl• initial reaction rarely occurs on the catalyst surface [32, 33]. On the other hand, after comparison of kinetic data published in literature [35-46], Nicovich et al. (1996) [47] reported that the reaction rate coefficient for Cl• addition to PCE is several hundred times faster than the reaction rate coefficient for addition of HO• to PCE [2, 47]; additionally, Thüner et al. (1999) [48] demonstrated that it is 2.35 fold faster than HO• addition (at 298 K and 1 atm). Consequently, the addition of HO• to PCE can be neglected. Product analysis by GC/MSD for PCE photochemical reaction (quartz inner tube) and PCE photocatalytic reaction (glass inner tube) isolated the intermediates described in Tables 3.4 and 3.5, respectively (experimental conditions detailed in Table 2.2). Taking into consideration the identified by-products, we propose the reaction mechanism described in Scheme 2.2 for both direct and photocatalytic conversion of PCE under UV radiation. Numerous pathways (A1 to E4) were proposed in order to cover all by-products formed (intermediates in brackets: 1 to 9). Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 85 Table 2.4. Product analysis by GC/MSD for PCE photochemical reaction (reactor with a quartz inner tube); experimental conditions described in Table 2.2 (UV photolysis, run 1). Gas-phase photooxidation of PCE under UV photolysis (quartz inner tube) Compound (i) Molecular CAS no. Ci [ppm] a Ci, C-PCE [ppm] c Formula Structure Perchloroethylene b C2Cl4 127-18-4 22.3 3.23 Methyl chloroformate C2H3ClO2 79-22-1 1.31 0.33 Butanal C4H8O 123-72-8 0.06 0.04 Chloroform CHCl3 67-66-3 0.85 0.09 Carbon tetrachloride CCl4 56-23-5 0.26 0.02 Chloroacetone C3H5ClO 78-95-5 --- --- Trichloroethylene C2HCl3 79-01-6 0.03 0.005 Trichloroacetyl chloride C2Cl4O 76-02-8 --- --- Phosgene CCl2O 75-44-5 --- --- 1,1-dichloroacetone C3H4Cl2O 513-88-2 0.004 0.001 Methyl trichloroacetate C3H3Cl3O2 598-99-2 0.37 0.08 Pentachloroethane C2HCl5 76-01-7 0.002 0.0002 Hexachlorethane C2Cl6 67-72-1 0.02 0.002 a VOC concentration of the major compounds was calculated using the response factor of toluene, except for PCE and trichloroethylene, for which specific response factors were used (calibrations solutions). b PCE concentration of the feed stream was analysed by MGC: CPCE, feed = 1221 ppm; CC-PCE, feed = 177 ppm, where CC-PCE, feed is the carbon atoms concentration of PCE on the feed stream. c Ci, C-PCE stands for carbon atoms concentration of compound i formed by PCE conversion (ppm). Chapter 2 86 Table 2.5. Product analysis by GC/MSD for PCE photocatalytic reaction (reactor with a glass inner tube); experimental conditions described in Table 2.2 (UV-TiO2 photocatalysis, run 1). Gas-phase photooxidation of PCE under UV-TiO2 photolysis (glass inner tube) Compound (i) Molecular CAS no. Ci a [ppm] Ci, C-PCE c [ppm] Formula Structure Perchloroethylene b C2Cl4 127-18-4 594 86 Methyl chloroformate C2H3ClO2 79-22-1 0.006 0.002 Butanal C4H8O 123-72-8 0.02 0.02 Chloroform CHCl3 67-66-3 0.53 0.05 Carbon tetrachloride CCl4 56-23-5 1.2 0.09 Chloroacetone C3H5ClO 78-95-5 --- --- Trichloroethylene C2HCl3 79-01-6 0.007 0.001 Trichloroacetyl chloride C2Cl4O 76-02-8 --- --- Phosgene CCl2O 75-44-5 --- --- 1,1-dichloroacetone C3H4Cl2O 513-88-2 0.03 0.008 Methyl trichloroacetate C3H3Cl3O2 598-99-2 0.006 0.001 Pentachloroethane C2HCl5 76-01-7 --- --- Hexachlorethane C2Cl6 67-72-1 0.1 0.01 a VOC concentration of the major compounds was calculated using the response factor of toluene, except for PCE and trichloroethylene, for which specific response factors were used (calibrations solutions). b PCE concentration of feed and exit streams were analysed by MGC: CPCE, feed = 1221 ppm; CC-PCE, feed = 177 ppm where CC-PCE, feed is the carbon atoms concentration of PCE on the feed stream. c Ci, C-PCE stands for carbon atoms concentration of compound i formed by PCE conversion (ppm). Gas-phase UV TiO2 photocatalysis of PCE using a lab-scale packed bed annular photoreactor 87 According to Scheme 2.2, PCE can suffer a reductive dechlorination under UV radiation (A1), leading to the formation of intermediate 1 (trichloroethenyl radical); its reaction with water (A2) generates the by-product trichloroethylene (C2HCl3). Nevertheless, PCE can also react with Cl• (A3), leading to the formation of intermediate 2 (pentachloroethyl radical); then, 2 can react with O2 (B1) generating 3 (pentachloroethaneperoxyl radical), followed by its dimerization (C1), decomposing into 4 (pentachloroethoxyl radical) and oxygen. The cleavage of the C–C bond of the intermediate 4 (C2), yields to 5 (trichloromethyl radical) and phosgene (CCl2O) formation [9, 32, 49, 50]. The trichloromethyl radicals (5) can react with Cl• (D1) as well as with water (D2), yielding to the by-products carbon tetrachloride (CCl4) and chloroform (CHCl3), respectively. As reported by Suárez et al. (2011) [51], the generation of highly toxic phosgene through PCO of a similar halogenated organic compound (trichloroethylene) is unclear: some authors observed the COCl2 production [51-54], in contrast to many others that stated that humidification promotes the phosgene hydrolysis reaction [32, 50, 55-57]. In this case, phosgene could not be detected as byproduct of both reactions (photochemical and photocatalytic) due to its reaction with water (D3), forming the mineralized final products carbon dioxide and hydrochloric acid [32, 50]. According to Montgomery and Rollefson (1934) [58] and other researchers [59-61], it was also proposed a sequential decomposition mechanism (D4) via formation of the intermediate 6 (chloroformyl radical). This thermally instable radical (6) is known as an intermediate in photochemical induced formation of phosgene (D5) [62]; but, the chloroformyl radicals (6) can also react between each other and, in the presence of water followed by Cl• elimination (D6), the by-product methyl chloroformate (C2H3ClO2) and hydroxyl radicals are formed. Regarding intermediate 2 (pentachloroethyl radical), it can also react with Cl• in the gas phase (B2), leading to formation of by-product hexachlorethane (C2Cl6); or, if it reacts with water (B3), it generates another identified by-product, pentachloroethane (C2HCl5) and HO• [9, 47]. Intermediate 4 (pentachloroethoxyl radical) can generate by-product trichloroacetyl chloride (C2Cl4O), after a chlorine radical subtraction (C3) [49]. Another Cl• elimination may yields to intermediates 7 (1,1,2-trichloro-2-oxoethyl radical) or 8 (2,2,2-trichloro-1-oxoethyl radical) (C4 and C5, respectively). When intermediate 7 reacts with intermediate 2, in the presence of water followed by Cl• elimination (B4), the identified by-product butanal (C4H8O) and hydroxyl radicals are formed. Besides, it is possible to occur in the gas phase the reaction between intermediate 8 (2,2,2-trichloro-1-oxoethyl radical) and phosgene; thus, in the presence of water followed by Cl• subtraction (G6), by-product methyl trichloroacetate (C3H3Cl3O2) and hydroxyl radicals are obtained. On the other hand, if intermediate 8 (2,2,2-trichloro-1-oxoethyl radical), reacts with a trichloromethyl radical (5), the unstable intermediate 9 (2-propanone, hexachloro-) is generated Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 95 3.1. Introduction VOCs represent a very active group of air pollutants discharged into the atmosphere by household and industrial activities, such as fuel combustion, residential cleaning agents, among several others. For example, Teixeira et al. [1] reported a large concentration of VOCs (including PCE and n-decane) in the indoor air of different stages of a wastewater treatment plant (WWTP), mainly associated with the aeration and mechanical agitation processes, as well as with the different sludge treatment stages. PCE, in particular, is widely used as model target pollutant due to its toxicity and carcinogenic potential to humans. Since the pioneer work by Fujishima and Honda in 1972 [2], the interest on photocatalytic processes applied to environmental issues has been growing every year, not only regarding photocatalytic processes (PCO) for water/wastewater treatment applications, but also for air purification [3-5]. The degradation of air pollutants by PCO is an attractive and efficient route when compared to other conventional techniques, such as adsorption on activated carbon, since the photocatalytic semiconductors are able to mineralize the pollutants, instead of a simple and more common phase transfer. Several semiconductors are often employed as photocatalysts [6-11] and among them, TiO2 (also known as titania) stands out as one of the most photoactive [12] under UV radiation. A number of commercial photocatalytic TiO2 powders are available on the market, being TiO2 Aeroxide P25 from Evonik® the most commonly employed [13]. It is generally accepted that the high activity of P25 comes from the formation of heterojunctions between the two types of crystalline phases, anatase (80%) and rutile (20%), which may also explain the low activity of P25 under visible light in some particular cases [14]. Specifically, rutile, owing to its lower conduction band, may absorb some light in the visible range (red) and thus serve as a photosensitizer of the anatase phase. The respective positions and the difference between the higher energy levels of the conduction bands of the two phases may cause a transfer of electrons from anatase towards rutile, preventing at the same time charge recombination [15]. Although the actuating mechanism of pure anatase is well established, as regards P25 it has been over the years a matter of debate [16, 17]. In a study conducted by Hajaghazadeh et al. [18], PC500 (consisting of 100 % anatase) yielded a higher reaction rate than P25 in the degradation of gaseous methylethylketone under UVA light. The superior photocatalytic activity of PC500 was attributed to its higher specific surface area (SSA). Taranto et al. [19] also reported a slight higher photocatalytic activity of P25 over PC500 in the degradation of methanol and noctane as gas-phase model pollutants. In another study conducted by Águia et al. [13], ten distinct commercially available photocatalysts were incorporated into a water-based exterior paint aiming Chapter 3 96 NO photoabatement. The highest yields were obtained with the catalysts consisting of 100% anatase, such as PC500. Reactor geometry is also a key factor in gas-phase photocatalysis due to its influence on optimizing the catalyst exposure to both radiation and reactants [20]. The most common photoreactors are tubular, annular and flat plate types [20-22]. Regarding the photocatalyst structural configuration, thin-film powder layer and/or fluidized bed [21, 23], coated wall-parallel [24, 25] and honeycomb/foam monolithic reactors [26-28] are probably the most representative. This chapter presents the results from a study on gas-phase solar photo-oxidation of two VOCs: n-decane and PCE. Two different commercially available TiO2 photocatalysts (P25 from Evonik® and PC500 from Cristal®) were deposited onto the surface of a monolithic and transparent structure of cellulose acetate. Applying such structure into a annular reactor allows a high surface-area-to-volume ratio and a low pressure drop, typical of honeycomb reactors [29]. The photocatalytic efficiency of both photocatalytic films, using a continuous-flow annular photoreactor equipped with a compound parabolic collector (CPC), were compared. This configuration allows the illumination of the whole reactor perimeter and catalytic bed, enhancing therefore the photonic efficiency [30, 31]. The PCO of n-decane and PCE over PC500 and P25 films was studied for different operating conditions, namely the feed flow rate (Qfeed), feed VOC concentration (Cdec, feed or CPCE, feed), and incident irradiance (I). The influence of relative humidity (RH), and the presence or absence of oxygen, in the photodegradation of such compounds was also assessed using PC500 film. To the best of our knowledge, few papers have compared the efficiency of PC500 and P25 films in the gas-phase PCO of VOCs. 3.2. Experimental 3.2.1. Materials and methods TiO2 P25 and PC500 powders were supplied by Evonik® and Cristal®, respectively, and used as delivered, without further modification or purification. Some characteristics of PC500 and P25 powders provided by the suppliers are given in Table 3.1. Cellulose acetate monolithic structures (TIMax CA50-9/S – LC = 80 mm, dch 2 = 9 mm × 9 mm, ew,ch = 0.1 mm; Wacotech GmbH & Co. KG.) were used as substrate to affix the powders. The reagents used for the generation of humidified air streams contaminated with PCE as well as the gases provided by Air Liquide are described in sub-section 2.2.1 of Chapter 2. Additionally, n-decane (≥94 %; CAS no. 124-18-5; Merck), was used without previous Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 97 purification. Cellulose acetate monolithic structures were evenly coated using aqueous suspensions of P25 and PC500 by the dip-coating method (Dip-Coater RDC21-K, Bungard Elektronik GmbH & Co. KG.). Before coating, the honeycomb structures were soaked for 1 h with distilled water and alkaline detergent (Derquim LM 01, Panreac Química, S.A.U.), subsequently washed with Milli-Q water, and dried at 323 K. The photocatalysts aqueous suspensions (2 % wt.) were sonicated for 10 min at 50 kHz in order to better disperse the particles. Based on Lopes et al. [32] work, where the influence of the number of layers upon the PCE photocatalytic conversion was assessed, nine layers of P25 or PC500 powder were deposited at a withdrawal rate of 0.8 mm s-1 assuring a thin and uniform film of 0.850 to 1.150 µm of thickness (data not shown) on each substrate surface. It is worth noting that samples were dried at 323 K for 30 min between each layer deposition. Finally, the coated monolithic structures were assembled into the annular photocatalytic reactor (see section 2.2) for the PCO study of n-decane and PCE. The catalytic bed properties are also detailed in Table 3.1. Table 3.1. Catalysts, catalytic bed properties and dimensions of the photoreactor employed in the PCO of n-decane and PCE under simulated solar radiation. TiO2 Catalysts PC500 (Cristal®) PCO of n-decane PCO of PCE Crystal structure > 99 % Anatase Crystal size [nm] 5-10 Surface area [m2 g-1] 345 mPC500 [mg] 52.3 35.3 ρA, PC500 [mg cm-2] 6.48 × 10-2 4.38 × 10-2 P25 (Evonik®) PCO of n-decane PCO of PCE Crystal structure 80 % Anatase, 20 % Rutile Crystal size [nm] 21 Surface area [m2 g-1] 50 mP25 [mg] 74.7 100.4 ρA, P25 [mg cm-2] 9.26 × 10-2 1.24 × 10-1 Photoreactor Outer tube (Pyrex-glass) dot, e [cm] 5.00 dot, i [cm] 4.64 Inner tube (quartz) din, e [cm] 2.00 din, i [cm] 1.64 Photoreactor LR [cm] 16.0 VR [cm3] 220 Chapter 3 98 3.2.2. Experimental setup and photocatalytic experiments The schematic representation of the experimental setup is shown in Figure 3.1 which is carefully described in the Chapter 2 sub-section 2.2.2. However, a few issues regarding the experimental unit were updated at this stage of the work and are highlighted (dark) in Figure 3.1:  Additional oxygen-free experiments were performed by replacing the flowing air by nitrogen and removing dissolved oxygen in water by replacing deionized water by a 10 g L-1 Na2SO3 solution in the Woulff bottle.  The UV lamp located inside the inner tube of the photoreactor used in Chapter 2 was replaced by the following devices (Figure 3.1b): i) a solar light simulator (Atlas, model Suntest XLS+) with 0.110 m2 of working area illuminated by a 1700 W aircooled Xenon arc lamp and an infrared coated quartz glass daylight filter reproducing the solar light spectrum within 300 < λ < 800 nm; ii) an electropolished anodized aluminium CPC with 0.023 m2 of irradiated area able to use both direct and diffuse radiation [30, 31].  The CO2 concentration of the photoreactor feed/exit stream was monitored online using an Indoor Air IAQ-Calc™ quality meter 7545 (TSI, Inc.). Table 3.1 summarizes the tube dimensions of the photoreactor and Figure 3.1b1 and b2 schematically represent the side and frontal views of the lab-scale annular photoreactor, respectively. Gas-phase PCE and n-decane concentration histories were monitored using the gas chromatograph (and method) described in sub-section 2.2.2 of Chapter 2. For n-decane analysis the oven temperature initially starts at 343 K for 1.2 min, after which raises up to 351 K using a 2 K min-1 heating ramp followed by a heating ramp of 30 K min-1 until reaching 433 K. Appendix A presents a detailed description of the data treatment as well as the calibration curves for PCE and n-decane. All experiments were carried out inside the chamber of the solar light simulator (Figure 3.1b) being the incident irradiance measured by a broadband UV radiometer (CUV 5, Kipp & Zonen B.V.), placed outside the outer tube and at the same height, within a spectral range of 280 – 400 nm corresponding to the UV fraction of the solar radiation. Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 99 Before PCO studies, the photolytic oxidation of PCE and n-decane was assessed (i.e. photoreactor assembled with the cellulose acetate supporting substrate without TiO2-photocatalytic film). Figure 3.1. Schematic representation of the lab-scale experimental set-up and the continuous-flow photoreactor: a) generation of air streams containing n-decane and water vapour; b) sunlight simulator containing the photoreactor: b1) side view and b2) frontal view; c) master gas chromatograph analytic system used for the analysis of the photoreactor feed and exit gas streams. The efficiencies of the process over both TiO2 photocatalysts toward n-decane and PCE photocatalytic degradation were compared in similar operational conditions and calculated as described in eq. 3.1. Conversion (%) = (1 - CVOC, exit CVOC, feed) × 100 3.1 At steady-state, VOC conversion (where CVOC, feed and CVOC, exit are the pollutant MGC MFC V0 V2 HTP Air (compressed air line) HTS Woulff bottles Woulff bottles MFC MFC V0 V2 CPC – compound parabolic collector FID – flame ionized detector FIH – flexible insulated heater HTP – humidity/temperature probe HTS – humidity/temperature sensor H2 Gen – hydrogen generator MGC – master gas chromatograph CO2S – Carbon dioxide sensor CO2P – Carbon dioxide probe PR – photoreactor SL – sampling loop SS – solar simulator TC – thermocouple V0 – one-way valve V2 – two-way valve V6 – six-way valve PP – peristaltic pump TB – thermostatic bath MFC – mass flow controller He N2 H2 Gen Air V6 SL PP FID Exhaust CT TC TB Computer (H2O) (VOC) a) V0 V0 FIH FIH CT TC FIH 606060 CPC PR SOLAR Suntest (SS) b) c) Outer Pyrex glass tube feed stream Photocatalytic bed exit stream Inner quartz tube Feed/Exit stream Inner quartz tube Photocatalytic bed Outer Pyrex glass tube CO2SCO2P Chapter 3 100 concentration in the feed and exit gas streams, respectively, expressed in ppm) were studied for different experimental conditions as described in sub-section 2.2.3 of Chapter 2; briefly: Qfeed (75 – 300 cm3 min-1, measured at 1 bar and 298 K), CVOC, feed (Cdec, feed = 71 - 284 ppm and CPCE, feed = 548 - 2738 ppm), RH of the feed gas stream in the presence and absence of oxygen (3 - 40 %, measured at 1 bar and 298 K), and I (18.9 - 38.4 WUV m-2, measured in the spectral range 280 - 400 nm: UV fraction of the incident sunlight). The mineralization yield of n-decane and PCE due to PCO was also evaluated and calculated as follows: Mineralization (%) = (CCO2, exit CVOC, feed ×1 n) × 100 3.2 where n is the number of carbon atoms of each pollutant molecular structure and CCO2, exit.is the CO2 exit stream concentration. 3.3. Results and discussion 3.3.1. VOCs photolysis Blank experiments (i.e. photoreactor without TiO2-photocatalytic film) were performed under three different incident irradiances, i.e., 18.9, 29.1, and 38.4 WUV m-2 (reference values to the sunlight UV fraction on the 280 – 400 nm range) in order to establish the effect of direct photolysis on the conversion of PCE and n-decane. No degradation of PCE or n-decane was observed for the range of irradiances tested (data not shown). 3.3.2. Catalytic film performances 3.3.2.1. Influence of the feed flow rate and VOC concentration Figure 3.2 illustrates the effect of the feed flow rate (Qfeed) on the n-decane (Figure 3.2a) and PCE (Figure 3.2c) conversion over PC500 (blue columns) and P25 (orange columns) films. Increasing Qfeed, the conversion of n-decane and PCE reduced regardless the employed irradiance (i.e. 18.9, 29.1, and 38.4 WUV m-2) or the type of catalytic film. More than 99 % of the initial n-decane (Cdec, feed = 71 ppm) and PCE (CPCE, feed = 1095 ppm) were converted over PC500 under I = 38.4 W m-2 when the feed flow rates of 75 cm3 min-1 and 150 cm3 min-1 were set; on the other hand, 93 % and 86 % of n-decane and 99 % of PCE conversions were attained over P25 film for the same experimental conditions. Increasing Qfeed to 300 cm3 min-1, n-decane conversion Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 101 decreased to approximately 65 and 91 % while PCE conversion decreased to 67 % and 92 %, respectively over P25 and PC500 films. In order to better understand the extent of the catalytic film activity toward the conversion of n-decane and PCE, the effect of Qfeed on the pollutant photocatalytic reaction rate, rVOC, was also assessed as shown in Figure 3.2b and 3.2d, respectively. The upmost values of n-decane reaction rate over PC500 (blue columns) and P25 (orange columns) films – 1.64 ×10-5 mol min-1 and 1.17 ×10-5 mol min-1, respectively – were attained when using the highest Qfeed (300 cm3 min-1) under I = 38.4 W m-2. Although the highest conversion was observed at Qfeed = 75 cm3 min-1 and I = 38.4 W m-2 (> 99 %), rdec was the lowest (4.43 ×10-6 mol min-1 and 4.15 ×10-6 mol min-1 respectively over PC500 and P25 films). Similarly, Figure 3.2d shows the PCE reaction rate, rPCE, over PC500 and P25 films. As seen before for n-decane, higher Qfeed (300 cm3 min-1) promoted higher reaction rates of PCE (2.20 × 10-4 mol min-1 and 1.59 ×10-4 mol min-1 respectively over PC500 and P25 films) under I = 38.4 W m-2. In opposition, setting the lowest Qfeed (75 cm3 min-1) both TiO2 photocatalytic films promoted the highest PCE conversions (99 %) and the lowest PCE reaction rates for each employed irradiance. In both cases, the results suggest a double antagonistic effect as the feed flow rate increases [34, 35]: i) a decrease in the residence time inside the reactor decreases the adsorption of the pollutant molecules on the photocatalytic film surface which impairs the efficiency of the PCO process; ii) higher organic load entering the photoreactor will enhance the mass transfer between the pollutant molecules and the catalytic film surface resulting in higher PCO reaction rates. Regarding the efficiency of both catalytic films, it is worth noting that P25 film provided lower values of photocatalytic reaction rate than PC500 film in the PCO of n-decane and PCE: P25 film provided values of n-decane reaction rate 6 % up to 41 % lower than PC500 film depending on I and Qfeed as well as, the differences in reaction rates between both catalytic films become greater as the Qfeed increases and I decreases. For example, fixing Qfeed at 75 cm3 min-1 the photocatalytic reaction rates of n-decane over P25 film 6, 8 and 13 % under 38.4, 29.1 and 18.9 W m-2 whereas for Qfeed = 300 cm3 min-1 the reaction rate over P25 film is 29, 34 and 41 % lower than over PC500 film under the same irradiances. The same trend can be observed for the PCO of PCE but, in this case, P25 film provides up to 34 % lower values of reaction rate of PCE than PC500 film employing the above described experimental conditions. Chapter 3 102 Figure 3.2. Effect of feed flow rate [Qfeed*] on the conversion of n-decane (a) and PCE (c) and on the photocatalytic reaction rate, rVOC ((b) and (d), respectively). Experimental points for incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ), measured within the spectral range of 280 – 400 nm, at steady-state conditions; Cdec, feed = 71 ppm CPCE, feed = 1095 ppm, RH* = 40 % and 21 % oxygen. Blue columns represent TiO2 PC500 ( , , ) and orange columns P25 ( , , ). * measured at 298 K and 1 bar. Figure 3.3 shows the influence of the feed concentration of n-decane (Cdec, feed) and PCE (CPCE, feed) on the conversion and on photocatalytic reaction rate of n-decane and PCE. For the same residence time in the photoreactor (τ = 88 s; Qfeed = 150 cm3 min-1) the conversion of n-decane (Figure 3.3a) and PCE (Figure 3.3c) decreased as the feed concentration increased, regardless the catalytic film used. Employing PC500 film more than 98 % of n-decane conversion was attained for the three irradiances employed (i.e. 18.9, 29.1, and 38.4 WUV m-2) when the initial n-decane concentration was 71 ppm, whereas P25 film yielded 74 up to 86 % of conversion, depending on the irradiance (Figure 3.3a). Also, increasing n-decane feed concentration (Cdec, feed = 142 ppm) the conversion of n-decane over PC500 film decreased in the range of 5 – 25 % for an irradiance reduction from 38.4 to 18.9 W m-2; over P25 film the Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 103 reduction was around 20 -30 % for the same irradiance reduction. It is interesting to note that, at this concentration and fixing Qfeed = 150 cm3 min-1 and RH = 40 %, the n-decane conversion over P25 film is closer to that found when the photoreactor was packed with PC500 film and fed with twice the n-decane feed concentration, i.e. 283 ppm. In particular, 50, 62 and 69 % of n-decane (Cdec, feed = 142 ppm) was converted over P25 film, while 37, 51 and 62 % was converted over PC500 film (Cdec, feed = 283 ppm) under 18.9, 29.1 and 38.4 WUV m-2. The activity of both photocatalytic films towards PCE conversion as a function of the PCE feed concentration (CPCE, feed) is shown in Figure 3.3c. As seen before for n-decane conversion over P25 and PC500 films, PCE conversion follows the same trend, i.e., the efficiency of the process decreases as the PCE feed concentration increases. In the range of 549 – 2738 ppm of CPCE, feed) a 42 % reduction in the PCE conversion (from 99 to 57 %) can be observed under I = 18.9 WUV m-2 (Figure 3.3c). It is worth noting that under higher irradiance (38.4 WUV m-2) no efficiency loss was observed in the same PCE concentration range. Although P25 film promoted PCE conversions of approximately 99 % (Qfeed = 150 cm3 min-1, CPCE, feed = 1095 ppm and RH = 40 %) under 38.4 WUV m-2, it is under lower irradiances that the differences between both catalytic film activities become greater: 87 and 75 % of PCE conversion were attained under 29.1 and 18.9 WUV m-2 of irradiance over P25 film, while 98 and 97 % were converted over PC500 film under the same operational conditions. Increasing the PCE feed concentration from 1095 ppm to 1643 ppm, the P25 film promoted 13, 27 and 56 % less PCE conversion than over PC500 respectively under 38.4, 29.1 and 18.9 WUV m-2. A similar approach regarding the photocatalytic reaction rate of n-decane can be followed as plotted in Figure 3.3b and 3.3d: in these cases, for the three incident irradiances employed (38.4 WUV m-2, 29.1 WUV m-2 and 18.9 WUV m-2) and over PC500 and P25 films the reaction rate of n-decane and PCE increases as the feed concentration increases. In other words, higher Cdec, feed and CPCE, feed for a same flow rate (Qfeed = 150 cm3 min-1), enhances the mass transfer between the feed gas stream and the catalyst surface, increasing the reaction rate of the pollutant. The results show that P25 film provides lower pollutant reaction rates in comparison to PC500 film. The former provides 13 up to 32 % lower reaction rates of n-decane than the latter depending on Cdec,feed and I: for example, fixing Qfeed = 150 cm3 min-1, Cdec,feed = 142 ppm and I = 38.4 W m-2 the n-decane reaction rate over P25 and PC500 films were respectively 2.27 × 10-5 mol min-1 and 3.39 × 10-5 mol min-1.The PCO of PCE over PC500 film also revealed higher reaction rates than over P25 film as the latter promotes reaction rates up to 40 % lower than over PC500 in the range of PCE feed concentration tested (1095 ppm and 1643 ppm) and depending on the incident irradiance. Chapter 3 110 they observed that when gas-phase oxygen was present the oxidation rate of formic acid was greatly improved, suggesting the important role of adsorbed oxygen. Likewise, El-Maazawi et al. [57] proposed that the photocatalytic reaction could take place in the absence of oxygen due to the TiO2 lattice oxygen. The oxygen from TiO2 lattice is depleted during the conversion of gaseous acetone being replenished by the oxygen from the feed gas stream. Figure 3.6. Effect of water content [RH*] on: (a) n-decane conversion at steady-state conditions for Cdec, feed = 71 ppm and Qfeed* = 150 cm3 min-1; (b) PCE conversion at steady-state conditions for CPCE, feed = 1095 ppm and Qfeed* = 150 cm3 min-1. Blue columns ( , , ) and orange columns ( , , ) represent, respectively, the presence and absence of oxygen; Incident irradiances of 38.4 WUV m-2 ( , ), 29.1 WUV m-2 ( , ), and 18.9 WUV m-2 ( , ) were measured within the spectral range of 280 – 400 nm. * measured at 298 K and 1 bar. Considering that the conversion of the two pollutants (n-decane and PCE) was drastically hindered after removing the gas-phase molecular oxygen from the gas stream, two oxygen-related mechanisms may be involved in the photocatalytic conversion [30, 54-57]. The first is based on the photodissociation of gas-phase molecular oxygen from the feed stream at the TiO2 surface into O-, which spontaneously reacts with O2 forming O3 -. These species may greatly contribute to the conversion of pollutants being at the same time the limiting step of the photocatalytic reaction. The second mechanism, where no gas-phase oxygen is present, may be related to the oxygen existing in the TiO2 lattice. Assuming the limited availability of surface lattice oxygen, it would be expected that the photocatalytic process would be considerably impaired, as it was, in fact, observed (see Figure 3.6). Regardless of the presence or absence of gas-phase molecular oxygen and the surface density of TiO2 on the substrate, it is possible to observe in both Figures 3.5 and 3.6 that increasing the RH the conversion of the pollutant also increases as it was already previously discussed. The contribution of the classic photocatalytic mechanism where HO• radicals initiate the photocatalytic process certainly explain such behaviour. Another hypothesis Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 111 for the role of HO• radicals is based on the ability of such species to act as effective traps for the holes preventing electron-hole recombination [58, 59]. In this way the reduced titanium centres (produced by reductive reaction between 𝑒CB -(TiO2) and Ti4+ centre of the TiO2) will enhance the space charge layer resulting in a longer lifetime, which ultimately promotes the oxidation of more gas-phase molecular oxygen [58, 59]. Several authors have followed another meaningful approach to clarify the mechanism of PCE degradation by PCO [45-48, 60-62]. However, agreement is yet to be found. On one hand, different authors [45-47] claimed that Cl• radical addition to PCE occurs several times faster than HO• radical addition, neglecting therefore the role of HO• radical. Such conclusion was further substantiated by Lu et al. [61] and Fan and Yates [62] after finding that the surface hydroxyl groups were inactive in the oxidation of methylchloride and trichloroethylene. On the other hand, Yamazaki et al. [48, 60], for example, stated that the PCO of PCE could occur via HO• radical or Cl• radical, but they concluded that Cl• radical initial reaction rarely occurs on the catalyst surface since the reaction with HO• radical is thermodynamically favourable. 3.4. Conclusions Different photocatalytic oxidation (PCO) reactions under simulated solar radiation showed that films using the commercial TiO2 photocatalyst PC500 provide higher conversions of PCE and n-decane than those obtained with P25. It was also found that PC500 film provides higher mineralization of n-decane than P25 film. Although the smaller crystallite size suggests the possibility of higher density in surface defects impairing the charge carriers and, therefore the photocatalytic efficiency, the higher surface area of PC500 catalyst particles may justify tge superior performance of the film towards the conversion of n-decane and PCE in comparison to that of P25 under steady state conditions. Regarding PC500 film, it was evidenced the effect of the relative humidity (RH) on the photocatalytic conversion of both pollutants. The results suggest that for RH in the range 3 - 20 % the competitive adsorption between water and pollutant molecules is unlikely to occur since the pollutants conversion increases with RH. On the other hand, at 40 % of RH the pollutant conversion over of PC500 film decreases, which means competitive adsorption between the above-mentioned molecules. The conversion of n-decane and PCE by PCO was drastically impaired or even not observable in the absence of gas-phase molecular oxygen indicating the key role of oxygen in photocatalysis. Three major mechanisms may be implicit in the effect of the oxygen content. One is the formation of reactive species from the adsorbed gas-phase molecular oxygen, which will oxidize the pollutants. The second involves the action of the oxygen from the Chapter 3 112 lattice of TiO2. The third mechanism involves the classical hydroxyl radical formation on the TiO2 surface. Acting directly over the pollutant molecules or as a trap for holes delaying the charge recombination and therefore promoting the formation of other oxidant species, HO• radicals cannot be excluded from the reaction. Chlorine radicals, Cl•, chain propagation reactions may also be involved in the PCO reaction mechanism of chlorinated compounds such as PCE. Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 113 3.5. References [1] J.V. Teixeira, S.M. Miranda, R.A.R. Monteiro, F.V.S. Lopes, J. Madureira, G.V. Silva, N. Pestana, E. Pinto, V.J.P. Vilar, R.A.R. Boaventura, Environ. Monit. Assess. 185 (2013) 59-72. [2] A. Fujishima, K. Honda, Nature 238 (1972) 37-38. [3] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69-96. [4] D.S. Bhatkhande, V.G. Pangarkar, A.A.C.M. Beenackers, J. Chem. Technol. Biotechnol. 77 (2002) 102-116. [5] J.-M. Herrmann, Top. Catal. 34 (2005) 49-65. [6] M. Anpo, Pure Appl. Chem. 72 (2000) 1787-1792. [7] C.F. Lin, C.H. Wu, Z.N. Onn, J. Hazard. Mater. 154 (2008) 1033-1039. [8] T. Ohno, T. Mitsui, M. Matsumura, Chem. Lett. 32 (2003) 364-365. [9] Q. Xiao, L. Ouyang, J. Phys. Chem. Solids 72 (2011) 39-44. [10] J.C. Colmenares, A. Magdziarz, D. Łomot, O. Chernyayeva, D. Lisovytskiy, Appl. Catal. B: Environ. 147 (2014) 624-632. [11] C. Li, X. Yang, R. Chen, J. Pan, H. Tian, H. Zhu, X. Wang, A. Hagfeldt, L. Sun, Sol. Energ. Mat. Sol. C. 91 (2007) 1863-1871. [12] R.L. Pozzo, M.A. Baltanás, A.E. Cassano, Catal. Today 39 (1997) 219-231. [13] C. Águia, J. Ângelo, L.M. Madeira, A. Mendes, J. Environ. Manage. 92 (2011) 1724-1732. [14] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 107 (2003) 45454549. [15] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem. 92 (1991) 178-190. [16] J. Moon, C.Y. Yun, K.-W. Chung, M.-S. Kang, J. Yi, Catal. Today 87 (2003) 77-86. [1] J.V. Teixeira, S.M. Miranda, R.A.R. Monteiro, F.V.S. Lopes, J. Madureira, G.V. Silva, N. Pestana, E. Pinto, V.J.P. Vilar, R.A.R. Boaventura, Environ. Monit. Assess. 185 (2013) 59-72. [2] A. Fujishima, K. Honda, Nature 238 (1972) 37-38. [3] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69-96. [4] D.S. Bhatkhande, V.G. Pangarkar, A.A.C.M. Beenackers, J. Chem. Technol. Biotechnol. 77 (2002) 102-116. [5] J.-M. Herrmann, Top. Catal. 34 (2005) 49-65. [6] M. Anpo, Pure Appl. Chem. 72 (2000) 1787-1792. [7] C.F. Lin, C.H. Wu, Z.N. Onn, J. Hazard. Mater. 154 (2008) 1033-1039. [8] T. Ohno, T. Mitsui, M. Matsumura, Chem. Lett. 32 (2003) 364-365. [9] Q. Xiao, L. Ouyang, J. Phys. Chem. Solids 72 (2011) 39-44. [10] J.C. Colmenares, A. Magdziarz, D. Łomot, O. Chernyayeva, D. Lisovytskiy, Appl. Catal. B: Environ. 147 (2014) 624-632. [11] C. Li, X. Yang, R. Chen, J. Pan, H. Tian, H. Zhu, X. Wang, A. Hagfeldt, L. Sun, Sol. Energ. Mat. Sol. C. 91 (2007) 1863-1871. [12] R.L. Pozzo, M.A. Baltanás, A.E. Cassano, Catal. Today 39 (1997) 219-231. Chapter 3 114 [13] C. Águia, J. Ângelo, L.M. Madeira, A. Mendes, J. Environ. Manage. 92 (2011) 1724-1732. [14] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 107 (2003) 45454549. [15] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem. 92 (1991) 178-190. [16] A.G. Agrios, K.A. Gray, E. Weitz, Langmuir 20 (2004) 5911-5917. [17] D.C. Hurum, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 108 (2004) 16483-16487. [18] M. Hajaghazadeh, V. Vaiano, D. Sannino, H. Kakooei, R. Sotudeh-Gharebagh, P. Ciambelli, Catal. Today 230 (2014) 79-84. [19] J. Taranto, D. Frochot, P. Pichat, Ind. Eng. Chem. Res. 48 (2009) 6229-6236. [20] Y. Paz, in: I.d.L. Hugo, R. Benito Serrano (Eds.), Advances in Chemical Engineering Photocatalytic Technologies, Academic Press, 2009, pp. 289-336. [21] M. Lewandowski, D.F. Ollis, Appl. Catal. B: Environ. 43 (2003) 309-327. [22] C. Águia, J. Ângelo, L.M. Madeira, A. Mendes, Catal. Today 151 (2010) 77-83. [23] L.A. Dibble, G.B. Raupp, Environ. Sci. Technol. 26 (1992) 492-495. [24] J. Araña, A. Peña Alonso, J.M. Doña Rodríguez, J.A. Herrera Melián, O. González Díaz, J. Pérez Peña, Appl. Catal. B: Environ. 78 (2008) 355-363. [25] Y. Yamin, N. Keller, V. Keller, J. Photochem. Photobiol. A 245 (2012) 43-57. [26] R. Portela, S. Suárez, R.F. Tessinari, M.D. Hernández-Alonso, M.C. Canela, B. Sánchez, Appl. Catal. B: Environ. 105 (2011) 95-102. [27] G.B. Raupp, A. Alexiadis, M.M. Hossain, R. Changrani, Catal. Today 69 (2001) 41-49. [28] A.N. Kouamé, R. Masson, D. Robert, N. Keller, V. Keller, Catal. Today 209 (2013) 13-20. [29] R.E. Hayes, S.T. Kolaczkowski, W.J. Thomas, Comput. Chem. Eng. 16 (1992) 645-657. [30] S.A. Larson, J.A. Widegren, J.L. Falconer, J. Catal. 157 (1995) 611-625. [31] T.H.K. Lim, S. D. , Chemosphere 54 (2004) 305-312. [32] F.V.S. Lopes, S.M. Miranda, R.A.R. Monteiro, S.D.S. Martins, A.M.T. Silva, J.L. Faria, R.A.R. Boaventura, V.J.P. Vilar, Appl. Catal. B: Environ. 140-141 (2013) 444-456. [33] F.V.S. Lopes, R.A.R. Monteiro, A.M.T. Silva, G.V. Silva, J.L. Faria, A.M. Mendes, V.J.P. Vilar, R.A.R. Boaventura, Chem. Eng. J. 204–206 (2012) 244-257. [34] M. Hussain, N. Russo, G. Saracco, Chem. Eng. J. 166 (2011) 138-149. [35] Z. Pengyi, L. Fuyan, Y. Gang, C. Qing, Z. Wanpeng, J. Photochem. Photobiol. A 156 (2003) 189194. [36] S.M. Miranda, F.V.S. Lopes, C. Rodrigues-Silva, S.D.S. Martins, A.M.T. Silva, J.L. Faria, R.A.R. Boaventura, V.J.P. Vilar, Environ. Sci. Pollut. Res. (2014) 1-13. [37] R.A.R. Monteiro, F.V.S. Lopes, A.M.T. Silva, J. Ângelo, G.V. Silva, A.M. Mendes, R.A.R. Boaventura, V.J.P. Vilar, Appl. Catal. B: Environ. 147 (2014) 988-999. [38] T. Ohno, K. Sarukawa, K. Tokieda, M. Matsumura, J. Catal. 203 (2001) 82-86. [39] B. Sun, A.V. Vorontsov, P.G. Smirniotis, Langmuir 19 (2003) 3151-3156. [40] T. Ohno, K. Tokieda, S. Higashida, M. Matsumura, Appl. Catal. A: Gen. 244 (2003) 383-391. [41] S. Sankar, K.G. Gopchandran, P. Kuppusami, S. Murugesan, Ceram. Int. 37 (2011) 3307-3315. Gas-phase solar photocatalysis of PCE and n-decane over different TiO2 photocatalysts using a lab-scale fixed bed annular photoreactor 115 [42] X. Wang, L. Sø, R. Su, S. Wendt, P. Hald, A. Mamakhel, C. Yang, Y. Huang, B.B. Iversen, F. Besenbacher, J. Catal. 310 (2014) 100-108. [43] V. Puddu, H. Choi, D.D. Dionysiou, G.L. Puma, Appl. Catal. B: Environ. 94 (2010) 211-218. [44] M. Hussain, R. Ceccarelli, D.L. Marchisio, D. Fino, N. Russo, F. Geobaldo, Chem. Eng. J. 157 (2010) 45-51. [45] J.M. Nicovich, S. Wang, M.L. McKee, P.H. Wine, J. Phys. Chem. 100 (1996) 680-688. [46] N. Petit, A. Bouzaza, D. Wolbert, P. Petit, J. Dussaud, Catal. Today 124 (2007) 266-272. [47] L.P. Thüner, I. Barnes, K.H. Becker, T.J. Wallington, L.K. Christensen, J.J. Orlando, B. Ramacher, J. Phys. Chem. A 103 (1999) 8657-8663. [48] S. Yamazaki, H. Tsukamoto, K. Araki, T. Tanimura, I. Tejedor-Tejedor, M.A. Anderson, Appl. Catal. B: Environ. 33 (2001) 109-117. [49] T.N. Obee, R.T. Brown, Environ. Sci. Technol. 29 (1995) 1223-1231. [50] L. Cao, A. Huang, F.-J. Spiess, S.L. Suib, J. Catal. 188 (1999) 48-57. [51] W.K. Jo, K.H. Park, Chemosphere 57 (2004) 555-565. [52] J. Zhao, X. Yang, Build. Environ. 38 (2003) 645-654. [53] P. Zhang, J. Liu, J. Photochem. Photobiol. A 167 (2004) 87-94. [54] D.S. Muggli, J.L. Falconer, J. Catal. 191 (2000) 318-325. [55] D.S. Muggli, J.L. Falconer, J. Catal. 187 (1999) 230-237. [56] D.S. Muggli, S.A. Keyser, J.L. Falconer, Catal. Lett. 55 (1998) 129-135. [57] M. El-Maazawi, A.N. Finken, A.B. Nair, V.H. Grassian, J. Catal. 191 (2000) 138-146. [58] T. Ohno, K. Sarukawa, M. Matsumura, J. Phys. Chem. B 105 (2001) 2417-2420. [59] A. Hagfeldt, M. Graetzel, Chem. Rev. 95 (1995) 49-68. [60] S. Yamazaki, T. Tanimura, A. Yoshida, K. Hori, J. Phys. Chem. A 108 (2004) 5183-5188. [61] G. Lu, A. Linsebigler, J.T. Yates, J. Phys. Chem. 99 (1995) 7626-7631. [62] J. Fan, J.T. Yates, J. Am. Chem. Soc. 118 (1996) 4686-4692. [63] J. Ângelo, L. Andrade, A. Mendes, Manuscript submitted for publication (2014). [64] R.A.R. Monteiro, S.M. Miranda, V.J.P. Vilar, L.M. Pastrana-Martínez, P.B. Tavares, R.A.R. Boaventura, J.L. Faria, E. Pinto, A.M.T. Silva, Appl. Catal. B: Environ. 162 (2015) 66-74. Part III Chapter 4. Gas-phase solar photocatalytic oxidation of PCE over TiO2 based paint Chapter 5. Gas-phase solar photocatalytic oxidation of n-decane over TiO2 based paint 4. Gas-phase solar photocatalytic oxidation of PCE over TiO2 based paint This chapter presents a study of the photooxidation of PCE in an annular photoreactor under simulated solar radiation employing two different configurations of a monolithic structure of cellulose acetate coated with an active TiO2-based paint. The influence of the configuration and different experimental conditions, namely feed flow rate (Qfeed), initial concentration (CPCE, feed), relative humidity (RH) in the system, absence of oxygen and incident irradiance on the PCE conversion was evaluated. Under the best experimental conditions (i.e. CPCE, feed = 1100 ppm, Qfeed = 75 cm3 min-1 (τ = 176 s), RH = 40 % and I = 38.4 WUV m-2 in the presence of oxygen). 60 % of the initial PCE concentration was converted photocatalytically. The results showed that depending on the configuration of the structure, photocatalytic degradation of PCE can be enhanced by approximately 58 %. It was also possible to observe that chlorine radicals play an important role in the degradation of PCE whereas hydroxyl radicals cannot be excluded from the photocatalytic mechanism and in the absence of oxygen the photoreaction can still take place. This chapter is based on the research article “Monteiro, R.A.R., Silva, A.M.T., Ângelo, J.R.M., Silva, G.V., Mendes, A.M., Boaventura, R.A.R., Vilar, V.J.P., Photocatalytic Oxidation of Gaseous Perchloroethylene over TiO2 Based Paint, submitted for publication, 2014”. Chapter 8 222 8.2.2. Solar/UV pilot-scale experimental unit A pilot scale prototype was designed and manufactured for the study of air detoxification systems (see Figure 8.1). It is based on an optimized compound parabolic collector (CPC) that uses solar radiation or on ultraviolet light lamps (UVA-Lamps) placed inside the photoreactor to study the possibility of working continuously day and night. This facility contains three main sections: (i) feed generator of humid air streams contaminated with VOCs; (ii) a 1.5 m long continuous-flow annular photoreactor placed at the top of a parabolic reflective surface (named CPC); (iii) integrated analysis system for both in situ monitoring and sampling of the photoreactor feed and exit streams. These sections are described in the following general description and schematically represented in Figure 8.1a to 8.1c. Figure 8.1. Schematic representation of the pilot unit: a) feed generator of air streams containing n-decane and water vapour; b) pilot scale continuous-flow annular photoreactor placed at the top of a CPC (at local latitude of 41º facing south); c) monitoring system used for the analysis of the photoreactor feed and exit streams. 8.2.2.1. Feed generation The experimental set-up (Figure 8.1a) used for generation of humid air streams containing a selected VOC mainly comprises an air compressor (Air Compressor CEVIK PRO 50 VX, 3 HP, Cevik S.A.), rotameters (FL-2010-SS and FL-2010-SS; Omega Engineering Ltd, UK), and Woulff bottles (supplied by Normax, Lda) filled with liquid VOC and H2O, respectively, and V0 V2 Feed generator Air compressor Woulff bottles Woulff bottles V0 CPC – compound parabolic collector CUV5 – CUV5 radiometer dot,e – external diameter of the outer tube FIH – flexible insulated heater HCS – heating/cooling system CT TC HCS (H2O) (n-decane) a) V0 FIH RM FI FI FI RM RM CPC UV-Lamps SUNLIGHT PR Photoreactor c) Monitoring system UV / Vis UV / Vis CT TC NV CT MS V2 CT CUV5 b) PR – photoreactor MS – multi-function sensor NV – needle valve SR – Sample resevoir RM – rotameter V0 – one-way valve V2 – two-way valve south oriented fixed platform tilted 41º 41.0° PR CPC πdot,e dot,e N SR Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane 223 placed in a temperature controlling system (thermostatic bath GD100 R2, Grant Instruments). The selected VOC and humidity contents were fixed as a result of controlling the flow rate of each air stream, and its temperature and pressure. All 1/4" stainless steel tubing, fittings, connectors, adapters, and valves are Swagelok products (Swagelok, USA). For the generation of VOCcontaining humid air streams n-decane (dec; ≥94 %, Merck S.A.) without further purification, ultrapure water and deionized water were used. 8.2.2.2. Pilot-scale photoreactor The 1.5 m long continuous-flow annular photocatalytic reactor (see Figure 8.1b) is composed by an outer Pyrex-glass tube (Duran borosilicate glass 3.3, Schott-Rorhglas GmbH) and, centred in the axial position along the bed, a concentric inner quartz tube (Quarzglas-Rohr, Quarzglastechnik, GmbH & Co KG); these tubes allow the penetration of solar/UV radiation through their structures. The void between these tubes was packed with photocatalyst-coated cellulose acetate monolithic structures. The photocatalytic bed dimensions and characteristics are shown in Table 8.1. 8.2.2.3. Compound parabolic collector (CPC) and radiation sources The photoreactor is placed at the top of a CPC of 0.220 × 1.50 m2 (0.330 m2 of static collectors made of stainless steel, showing high reflectivity in the UV range) mounted on a south oriented fixed platform tilted 41º (local latitude). This involute reflective surface around the photoreactor enables almost all the UV radiation (both direct and diffuse) to be collected for processing, without the need for solar tracking [22]; the reflected light is distributed around the photoreactors maximizing the exposure area and making the irradiance on the catalytic bed uniform. The intensity of the incident solar radiation (only the UV fraction of the incident solar light) is measured within a spectral range of 280 – 400 nm with a broadband UV radiometer (CUV 5, Kipp & Zonen B.V.), mounted on the pilot plant at the same angle as the CPC; a handheld display unit (Meteon, Kipp & Zonen B.V.) is used to produce readings in terms of incident irradiance (WUV m-2). Three 8 W UVA (29 WUV m-2) blacklight lamps (spectral peak centred at 365 nm) are placed inside the inner tube, centred in the axial position along the bed. Therefore, as soon as the sunlight intensity drops below a threshold value (in cloudy days or during the night), UVA-Lamps can illuminate the photocatalytic bed allowing the continuous operation of the pilot scale unit with low power consumption. The intensity of the incident UVA-Lamps radiation was measured with the broadband UV radiometer placed on the outside of the inner tube and in contact with it. Chapter 8 224 8.2.2.4. Photoreactor feed and exit streams analysis The photoreactor feed and exit streams are monitored online through an automatic portable VelociCalc® multifunction 9565-P ventilation meter equipped with 986 VOCs probe (TSI® incorporated) (see Figure 8.1c). The TSI probe uses a photoionization detector (PID) to measure total VOCs; it is also used to monitor the temperature, relative humidity and the CO2 concentration of the feed / exit gas stream of the photoreactor throughout the experimental time. 8.2.3. Photocatalytic experiments First, the continuous-flow annular photoreactor was assembled with uncoated cellulose acetate samples in order to study the n-decane photochemical oxidation under solar or UVA radiation. Then, after packing the reactor with the coated supporting substrates, the performances of the TiO2-based films toward n-decane photocatalytic conversion were compared in similar operational conditions. All experimental conditions are reported in Table 8.2. Prior to the solar photochemical and photocatalytic experiments and in the total absence of radiation (photoreactor initially covered by a blackout filter or the inner UVA lamp turned off), the photoreactor was continuously fed with the polluted air stream overnight, to ensure a constant volatilization of n-decane and, consequently, a constant feed composition (operating conditions reported in Table 8.2). Therefore, several samples (one sample after each five minutes) of the exit stream were analysed to confirm a constant feed composition. The efficiency of the PCO process was expressed in terms of n-decane conversion and mineralization yield, calculated as follow: Conversion (%) = (1 - Cdec, exit Cdec, feed) × 100 8.1 Mineralization (%) = (CCO2, exit Cdec, feed ×1 10) × 100 8.2 where Cdec, feed and Cdec, exit are the n-decane concentrations on the feed and exit streams, respectively and CCO2, exit is the CO2 exit stream concentration. All concentrations are expressed in ppm. For the solar PCO experiments, the time evolution of the incident irradiance (measured within 280 – 400 nm) was measured. Relative humidity and temperature of the photoreactor exit stream were also measured for the solar and artificial UVA PCO experiments. Note that, before any PCO experiment, activation and degassing of the photocatalytic bed was carried out in the absence of radiation and using a water-containing air stream at a total flow Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane 225 rate of 0.500 L min-1 (measured at 1 bar and 298 K). Solar PCO of n-decane experiments were carried out during a full week period. Then, for each photocatalytic film, only artificial UVA radiation was employed to explore the behaviour of the photoreactor without the uncertainty of the solar radiation variability. The UVA PCO of n-decane experiments were performed during 8 h. Table 8.2. Experimental conditions employed in gas-phase PCO of n-decane carried under solar or artificial UVA radiation. Film Qfeed* [L min-1] (τ [s]) Cdec, feed [ppm] RH* [%] Tfeed [ºC] Radiation source I# [WUV m-2] None 2 (44) 10 ± 1 16.3 – 31.3 21.9 – 33.3 Sunlight 0.0 – 57.6 11 ± 1 14.0 – 26.4 23.6 – 38.6 UVA 29 ± 5 P25 10 ± 1 12.5 – 34.1 17.5 – 31.0 Sunlight 0.0 – 57.2 11 ± 1 15.2 – 65.3 26.7 – 32.0 UVA 29 ± 5 PC500 10 ± 1 9.1 – 39.1 12.6 – 40.9 Sunlight 0.0 – 58.9 11 ± 1 13.2 – 36.5 27.0 – 37.3 UVA 29 ± 5 PCP 10 ± 1 12.4 – 31.7 20.1 – 35.7 Sunlight 0.0 – 64.1 11 ± 1 16.6 – 58.8 23.4 – 33.3 UVA 29 ± 5 P = 1 bar * measured at 298 K and 1 bar. # measured within 280 – 400 nm. 8.3. Results and discussion 8.3.1. Solar/artificial UVA photolysis of n-decane Control tests (photoreactor packed with uncoated cellulose acetate supporting substrates) were performed in order to establish the effect of the radiation on the conversion and mineralization of n-decane. This pollutant in contaminated air streams is able to absorb light over a wide range of wavelengths (absorption is often stronger at shorter wavelengths) [17]. Thus, if a Pyrex-glass outer tube is used as outer tube of the photoreactor, it will absorb most of the shorter wavelengths of the solar radiation (for λ < 285 nm). For this reason, direct photolysis of n-decane with sunlight do not yield measurable ionization of n-decane (i.e. no conversion of n-decane is observed). Under artificial UVA radiation, no measurable n-decane conversion or mineralization was observed. Although the inner tube is made of quartz UVA radiation (peak centred at λ = 365 nm) is not energetic enough to ionize n-decane molecules. Considering the negligible effect of the radiation (solar and artificial UVA) the evolution over time of conversion and mineralization of n-decane was not plotted. Chapter 8 226 8.3.2. Solar/artificial UVA PCO of n-decane In Figure 8.2 is illustrated the PCO of n-decane over P25 (Figure 8.2a), PC500 (Figure 8.2b) and PCP (Figure 8.2c) photocatalytic films under solar radiation. In all cases is clearly shown the dependence of the conversion and mineralization on the irradiance. Photocatalytic activity was observed between 7:00 and 20:00 approximately reaching complete n-decane conversion over P25 and PC500 films and 45 % of conversion over PCP film at around 9:00 (15 WUV m-2 of irradiance) and 14:00 (48 WUV m-2), respectively; remarkably, over P25 film, the recorded values of Cdec, exit at night were below the Cdec, feed (Figure 8.2a) revealing some adsorption capacity of P25. On the contrary, over PC500 and PCP films, the values of Cdec, exit during the night were the same which means that both photocatalytic films presented low adsorption capacity. Then, with the sunrise, the photocatalytic activity of the films was resumed and the same cyclic behaviour was observed in each day of experiment. The CO2 concentration was also measured during the experiments in order to assess the mineralization yield which is described by the following equation: C10H22(g) + 31 2 O2(g) h  →intermediates h  → 10 CO2(g) + 11 H2O(g) 8.3 As expected, the mineralization of n-decane also increased with the solar irradiance being such trend observed for all three experiments. Generally, all the n-decane converted is mineralized into CO2 and water over P25 and PC500 films. In opposition, under low irradiance conditions (but enough to attain 100 % of n-decane conversion), the mineralization is reduced: for example, on the 6th and 7th days of n-decane PCO over PC500 film (Figure 8.2b) 100 % of n-decane was converted however not all into CO2 and water. This means that the excess of photons (in high irradiance conditions) favours the direct reaction pathway to produce CO2 and water. The same behaviour was observed over PCP film (Figure 8.2c) for the first three days and the last day of experiment. By-products such as those already mentioned in Chapter 5 are, most likely, being formed and released to the atmosphere as final products of n-decane PCO as no formation of CO2 was observed. During the 4th and the 5th days of experiment over PCP film around 40 % of n-decane was mineralized into CO2 and water. Once again, high irradiance conditions favour the direct production of CO2. In opposition, low irradiance conditions (sunrise and sunset) impair the mineralization as it drops to 0 %. It is also worth noting that the PCO of n-decane over the three photocatalytic films did not evidence any catalyst deactivation during the experiments. Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane 227 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 100 16:47 4:47 16:47 4:47 16:47 4:47 16:47 4:47 16:47 4:47 16:47 4:47 T[oC] / RH [%] / I[W m-2] n-decane conversion / mineralization [%] May Mineralization Conversion Temperature RH Irradiance 678 9 10 11 12 a) 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 90 100 18:41 6:41 18:41 6:41 18:41 6:41 18:41 6:41 18:41 6:41 18:41 6:41 T[oC] / RH [%] / I[W m-2] n-decane conversion / mineralization [%] May Mineralization Conversion Temperature RH Irradiance 14 15 16 17 18 19 20 b) Chapter 8 228 Figure 8.2. Time evolution of n-decane PCO over P25 (a), PC500 (b) and PCP (c) films under solar radiation (irradiance measured within 280 - 400 nm). Solar radiation collected with a CPC and UV irradiance measured on the outer tube of the photoreactor. Operating conditions: Cdec, feed = 10 ppm; Qfeed = 2 L min-1 (τ = 44 s) as reported in Table 8.2. In Figure 8.3 the conversion and the mineralization obtained during the fourth day of operation are plotted as a function of irradiance respectively for P25 (Figure 8.3a), PC500 (Figure 8.3b) and PCP (Figure 8.3c). During the morning and due to the raising temperatures, 100 % of n-decane conversion was attained at around 15 WUV m-2 whereas in the afternoon the total conversion was maintained up to 3 WUV m-2. This hysteresis which was already described by Sánchez group [13, 14] repeated in every experimental day is a result of the difference in adsorption-desorption phenomena between morning and afternoon. When the sun rises, the temperature at the photocatalytic film surface increases and the readjustment of the adsorption equilibrium results in enhanced n-decane desorption; this way the conversion of the pollutant is reduced in comparison to that when the sun goes down, where the opposite effect is observed (i.e., when the temperature decreases). Over PCP film the conversion of n-decane started after 20 WUV m-2 of irradiance was reached and the topmost value of n-decane conversion was 45 % under 48 WUV m-2 (around 14:00). On the other hand, during the afternoon period the photocatalytic activity started to decrease below 48 WUV m-2 until the sunset. In this case, the threshold to PCP photocatalytic film initiate activity is around 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 80 90 100 9:11 21:11 9:11 21:11 9:11 21:11 9:11 21:11 9:11 21:11 9:11 21:11 9:11 T[oC] / RH [%] / I[W m-2] n-decane conversion / mineralization [%] July Mineralization Conversion Temperature RH Irradiance 18 19 20 21 22 23 24 c) Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane 229 20 WUV m-2. It is worth noting that according to Figure 8.3b no mineralization of n-decane over PC500 film is observed up to 3 WUV m-2 (in the morning) despite the increasing conversion. This is probably due to the fact that at night the PC500 photocatalytic film saturates in n-decane while CO2 is being desorbed. The higher adsorption of CO2 is responsible for the delay in CO2 appearance in the morning as well as for the high mineralization values observed at the sunset. A similar behaviour is also observed in Figure 8.3c: although the n-decane conversion started after 20 WUV m-2 of irradiance the mineralization was only observed at around 30 WUV m-2 (in the morning period); in the afternoon, mineralization was kept until n-decane conversion was no longer observed. Figure 8.3. n-Decane conversion ( morning, increasing irradiance;  afternoon, decreasing irradiance) and mineralization ( morning, increasing irradiance;  afternoon, decreasing irradiance) over P25 (a) PC500 (b) and PCP (c) films under solar radiation during the fourth day of experiment. UV irradiance measured within 280 - 400 nm. Operating conditions: Cdec, feed = 10 ppm; Qfeed = 2 L min-1 (τ = 44 s) as reported in Table 8.2. Figure 8.4 illustrates the reaction rate obtained as a function of irradiance over P25, PC500 and PCP photocatalytic films also for the fourth day of each experiment. Over P25 and 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 n-decane conversion / mineralization [%] I[WUV m-2] morning, increasing irradiance afternoon, decreasing irradiance morning, increasing irradiance afternoon, decrasing irradiance Conversion Mineralization a) 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 n-cecane conversion / mineralization [%] I[WUV m-2] morning, increasing irradiance afternoon, decreasing irradiance morning, increasing irradiance afternoon, decrasing irradiance Conversion Mineralization b) 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 n-decane conversion / mineralization [%] I[WUV m-2] morning, increasing irradiance afternoon, decreasing irradiance morning, increasing irradiance afternoon, decrasing irradiance Conversion Mineralization c) Chapter 8 230 PC500 films at low radiation levels (< 5 WUV m-2 in the morning) the reaction rate increases linearly while between 5 and around 15 WUV m-2 a power mode trend of the reaction rate is observed. Above 15 WUV m-2 in the morning or 3 WUV m-2 in the afternoon the reaction rate was controlled by transport phenomena and independent of the irradiance. The same behaviour was observed over PCP film above 45 WUV m-2. Between around 20 WUV m-2 and 45 WUV m-2, in the morning period, the reaction rate over PCP film seems to follow a linear trend. The higher reaction rate attained in the afternoon for the three cases are related to the adsorption effect on conversion and mineralization values previously described. Figure 8.4. n-Decane PCO reaction rate (, ,  morning, increasing irradiance; , ,  afternoon, decreasing irradiance) over P25 (, ) PC500 (, ) and PCP (, ) films under solar radiation during the fourth day of experiment. UV irradiance measured within 280 - 400 nm. Operating conditions: Cdec, feed = 10 ppm; Qfeed = 2 L min-1 (τ = 44 s) as reported in Table 8.2. The same experiments were repeated using artificial UVA radiation instead of using solar radiation in order to explore the behaviour of the reactor without the uncertainty of the solar radiation variability, as illustrated in Figure 8.5. Over P25 film, around 71 % of n-decane (Cdec, feed = 10 ppm) was converted 15 min after turning on the UVA lamps (29 WUV m-2). The photocatalytic activity was kept until the end of the experiment with no catalyst deactivation nor formation of by-products as all the converted n-decane was mineralized into CO2 and water. Regarding the PC500 photocatalytic film, similar behaviour under the same operational conditions can be found in Figure 8.5b. However, in this case, 100 % of the initial n-decane was converted into CO2 and water which represents an 0 10 20 30 40 50 0 1 2 3 4 5 6 0 10 20 30 40 50 60 70 0 5 10 15 20 25 I[WUV m-2] rdec 107[mol min-1] rdec 107[mol min-1] I[WUV m-2] Evaluation of a solar/UV annular pilot scale reactor for 24 h continuous PCO of n-decane 231 improvement of 29 % in conversion in comparison to that of P25 photocatalytic film. Finally, over PCP film only around 25 % of the initial n-decane was converted. This result is in accordance to that obtained under solar radiation: under approximately 29 WUV m-2 (in the morning) around 10 % is converted while ca. 25 % is converted in the afternoon under the same irradiance. Besides, the up most value of mineralization was around 52 % which means that 13 % is being converted in by-products while 75 % of n-decane remained unreacted. However, according to the mineralization profile, after 3.5 h of experiment this value starts to decrease which may suggest that photons are no longer in excess impairing the direct reaction pathway to produce CO2 and water. Figure 8.5. Time evolution of n-decane PCO over P25 (a), PC500 (b) and PCP (c) films under artificial UVA radiation. UV irradiance measured within 280 - 400 nm and facing the inner quartz tube of the photoreactor. Operating conditions: Cdec, feed = 10 ppm; Qfeed = 2 L min-1 (τ = 44 s) as reported in Table 8.2. The results obtained in this chapter are in line with those obtained in chapter 3 and chapter 5 regarding the efficiency of the photocatalytic films towards the conversion of n-decane by PCO process. In this sense, PC500 photocatalytic film presented the highest conversion of n-decane even considering the lower mass of photocatalytic film which may be ascribed to the 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 T[oC] / RH [%] n-decane conversion / mineralization [%] Time [h] Mineralization Conversion Temperature RH a) 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 T[oC] / RH [%] n-decane conversion / mineralization [%] Time [h] Mineralization Conversion Temperature RH c) 0 5 10 15 20 25 30 35 40 0 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 T[oC] / RH [%] n-decane conversion / mineralization [%] Time [h] Mineralization Conversion Temperature RH b) Final remarks and suggestions for future work 239 9.1. Final Remarks The main objective of this thesis was to evaluate the efficiency of solar gas-phase heterogeneous photocatalysis towards the elimination of pollutants such as VOCs present mainly in indoor air atmospheres. The process was performed in laband pilot-scale experimental units equipped with single-pass, continuous-flow, annular photoreactor featuring Compound Parabolic Collectors (CPCs) and/or UV lamps. The results showed that the use of heterogeneous photocatalysis as an alternative technology for air treatment is an interesting and feasible option. 9.1.1. Photolysis Gas-phase solar and UV photolytic experiments as well as aqueous-phase visible lightdriven photolytic experiments were performed in order to assess the influence of incident radiation on the conversion/inactivation of the target pollutants/bacteria. The UVC photolysis demonstrated that PCE conversion was negligible (~ 0 %) or almost complete (98 %), depending on the material of the inner tube of the lab-scale annular photoreactor (glass or quartz, respectively). Product analysis by GC/MSD for PCE photolysis, under steadystate conditions and using the quartz inner tube, isolated the main following intermediates: methyl chloroformate; butanal; chloroform; carbon tetrachloride; methyl trichloroacetate. PCE and n-decane solar photolytic experiments performed in the lab-scale annular photoreactor (Lr = 16.0 cm and Vr = 220 cm3) under three different incident irradiances, i.e., 18.9, 29.1, and 38.4 WUV m-2 (sunlight UV fraction: 280 – 400 nm range) showed no measurable conversion of both pollutants (CPCE, feed = 1095 ppm, Cdec, feed = 71 ppm, Qfeed = 150 cm3 min-1, τ = 88 s and RH = 40 %). Diphenhydramine, DP (10 mg L-1, adsorption at λ < 280 nm) showed to be a very recalcitrant pollutant in the absence of a photocatalyst, since the conversion observed was less than 3 % in 60 min under visible light (λ > 430 nm). Negligible E. coli inactivation was also observed in the experiments performed under UVA light radiation (25 W m-2, spectral peak of 365 nm) without any photocatalyst. MEK and H2S UVA (45.3 W m-2, spectral peak of 365 nm) and solar (0.7 WUV m-2) photolytic experiments demonstrated that no reaction occurred in the absence of photocatalyst. n-Decane photolytic experiments (Qfeed = 2 L min-1, τ = 44 s, and Cdec, feed = 10 ppm) carried out in pilot-scale photoreactor (LR = 144 cm and Vr = 4300 cm3) under natural solar and Chapter 9 240 artificial UVA radiation (29 W m-2, spectral peak of 365 nm) showed to be inefficient to convert this pollutant. 9.1.2. Photocatalysis using a lab-scale photoreactor PCE conversion by PCO over TiO2 P25 under UVC radiation (0.8 W m-2, spectral peak centred at 253.7 nm) and conducted in a lab-scale packed bed (glass spheres) annular photoreactor (Lr = 20.0 cm and Vr = 275 cm3) was evaluated upon different CPCE, feed and Qfeed: (1) for a 4.25 fold increase in PCE concentration feed (574 – 2442 ppm), a reduction of ~44% (from 63 to 35 %) on the PCE conversion was observed; (2) increasing the feed flow rate (from 59 to 300 cm3 min-1), a ~5 fold reduction of the residence time was observed, leading to poor degrees of conversion (~33 %). On the other hand, in terms of PCE PCO reaction rate, i.e., organic load converted, the trend followed an opposite direction: (1) for the same 4.25 fold increase in PCE concentration feed (574 – 2442 ppm), an increase of 2.4 times (from 3.93 × 10-5 to 9.28 × 10-5 mol min-1) on the photocatalytic reaction rate was observed; (2) increasing the feed flow rate (from 59 to 300 cm3 min-1), the PCO reaction rate achieved 8.69 × 10-5 mol cm-3. Under steadystate conditions, product analysis by GC/MSD for PCE photocatalytic reaction showed butanal, chloroform, and carbon tetrachloride as the main isolated intermediates. The photocatalytic approach led to a 51 % PCE conversion with complete mineralization Different PCO reactions were carried out in another lab-scale annular photoreactor (Lr = 16.0 cm and Vr = 220 cm3) packed with a transparent cellulose acetate monolithic structure coated with TiO2-based films (by dip-coating technique). This reactor configuration in comparison to the previous one ensured a low pressure drop, high catalyst surface-area-to-volume ratio, much less amount of photocatalyst and minimized “shading effects”. This photoreactor featured a CPC in order to illuminate the whole reactor perimeter and catalytic bed. The results showed that films using the commercial TiO2 photocatalyst PC500 provided higher conversion and mineralization values of PCE and n-decane than those with P25 film, under simulated solar radiation. Photocatalytic conversions close to 100 % were obtained for n-decane and PCE (Cdec, feed = 71 ppm and CPCE, feed = 1095 ppm, respectively) when I = 38.4 WUV m-2, Qfeed = 150 cm3 min-1 (τ = 88 s), and RH = 20 %. The mineralization of PCE (CPCE, feed = 1095 ppm) showed no differences between both photocatalytic films. Over PC500 film the complete mineralization of n-decane (Cdec, feed = 71 ppm) was observed while over P25 film only 69 % was mineralized in the following conditions Qfeed = 150 cm3 min-1 (τ = 88 s), I = 38.4 WUV m-2 and RH = 40 %. Although the smaller crystallite size suggests the possibility of higher density in surface defects impairing the charge carriers and, therefore the photocatalytic efficiency, the higher surface area of PC500 catalyst may justify superior performance of PC500 Final remarks and suggestions for future work 241 photocatalytic film towards the conversion of n-decane and PCE in comparison to the film prepared from P25 catalyst under steady state conditions. In terms of n-decane and PCE PCO reaction rates over PC500 photocatalytic film, the upmost values were attained when their conversion was maximized, i.e., when the following sets of conditions were used: Cdec, feed = 71 ppm, CPCE, feed = 1095 ppm and Qfeed = 300 cm3 min-1 or Cdec, feed = 284 ppm, CPCE, feed = 2738 ppm and Qfeed = 150 cm3 min-1; on the other hand, the lowest n-decane and PCE PCO reaction rates over PC500 photocatalytic film were attained when the following operational conditions Cdec, feed = 71 ppm, CPCE, feed = 1095 ppm and Qfeed = 75 cm3 min-1 or Cdec, feed = 71 ppm, CPCE, feed = 548 ppm and Qfeed = 150 cm3 min-1 were used. It is worth noting that the highest conversions were attained using these conditions. In all the above cases, it was demonstrated that increasing Qfeed and/or CVOC, feed, the organic load entering the reactor will be increased enhancing the mass transfer between the pollutant molecules and the catalyst surface and, therefore, the PCO reaction rate increased. On the other, for high Qfeed or CVOC, feed conditions, a reduction in the conversion of both pollutants was observed. This was due to an insufficient residence time inside the reactor (higher Qfeed) or an excessive presence of pollutant molecules in comparison to the number of photons/hydroxyl radicals (higher CVOC, feed) which impairs the photocatalytic conversion of the gas-phase pollutants. The PCO of PCE over a TiO2-based paint using the same lab-scale annular photoreactor (Lr = 16 cm and Vr = 220 cm3) was studied under simulated solar radiation, using the CPC to increase the photonic efficiency and cellulose acetate monolithic structures as substrate for the photocatalytic paint. The influence of substrate configuration (one structure with closed channels and another with open channels) was demonstrated in PCE PCO experiments: removing the outer walls of the substrate (configuration with open channels), further exposing more surface area to radiation, the PCE conversion through PCO enhanced up to 58 %, depending on the incident irradiance (CPCE, feed = 1100 ppm, Qfeed = 75 cm3 min-1, I = 18.9, 29.1 or 38.4 WUV m-2). Using such configuration, it was observed that PCE conversion was greatly affected by the flow rate of the inlet gas stream and initial PCE concentration: increasing four times the feed flow rate (from 75 to 300 cm3 min-1), PCE conversion decreased 73 % under 38.4 WUV m-2 of irradiance; in terms of photocatalytic reaction rate it was observed that, the highest value was attained for Qfeed = 150 cm3 min-1 (CPCE, feed = 1100 ppm and I = 38.4 WUV m-2) and increasing Qfeed to 300 cm3 min-1 the reaction rate of PCE decreased. Such behaviour was ascribed to a reduced mass transfer between the PCE molecules and the photocatalytic paint surface above 150 cm3 min-1 impairing the reaction rate most likely due an insufficient residence time inside the photoreactor. For a 3.7-fold PCE feed concentration increase (from 600 to 2200 ppm) under 38.4 WUV m-2 the Chapter 9 242 PCE reaction rate also increased although PCE conversion reduced 43 %. In this case, the mass transfer between the PCE molecules and the photocatalytic paint surface was increased as CPCE, feed increased for the range of CPCE, feed employed. The PCO of n-decane over a TiO2-based paint coated on cellulose acetate monolithic structures (configuration with open channels) using the lab-scale annular photoreactor (Lr = 16 cm and Vr = 220 cm3) was also studied under simulated solar radiation. The gas-phase photocatalytic experiments showed that highest n-decane conversion (98 %) was attained at the lowest Qfeed (75 cm3 min-1) and Cdec, feed (41 ppm), and highest RH (40 %) and I (38.4 WUV m-2). Feeding the photoreactor with the double flow rate, the n-decane conversion decreases from 96 to 62 %. Alternatively, when the n-decane concentration on the feed stream was doubled, the remaining unreacted n-decane fed increased from 4 to 46 %. PCO efficiency is a result of the combination of pollutant concentration and RH depending on the relative adsorption affinity of the photocatalyst for the pollutant and water molecules and on the mechanism of the hydroxyl radical attack (which also depends on the pollutant). Despite the acceptance that water is essential for photocatalysis as it is responsible for the formation of HO• radicals, each case requires a specific study which can hardly be extrapolated. Therefore, within this dissertation the effect of RH on the n-decane and PCE PCO was studied individually as follows: - The effect of RH on the PCE PCO over TiO2 P25 under UVC radiation (0.8 W m-2, spectral peak centred at 253.7 nm) and conducted in a lab-scale packed bed (glass spheres) annular photoreactor (Lr = 20.0 cm and Vr = 275 cm3) was evaluated in the range 12 – 40 %. The conversion of PCE increased ca. 16 % as RH increased (CPCE, feed = 1221 ppm, Qfeed = 150 cm3 min-1 and I = 0.8 W m-2); - In the PCO reactions of n-decane and PCE carried out in the lab-scale annular photoreactor (Lr = 16.0 cm and Vr = 220 cm3) packed with a cellulose acetate monolithic structure coated with PC500 photocatalytic film, the effect of RH was assessed in the range 3 – 40 %. The results demonstrated that in the range 3 - 20 % of RH competitive adsorption between water and pollutant molecules is unlikely to occur since the pollutants conversion increases with RH (except for Cdec, feed = 71 ppm at Qfeed = 150 cm3 min-1 and CPCE, feed = 1095 ppm at Qfeed = 150 cm3 min-1 where negligible effect of RH on n-decane or PCE was, respectively, observed). On the other hand, at 40 % of RH the pollutant conversion over of PC500 film decreases, which means competitive adsorption between the above-mentioned molecules (n-decane: Cdec, feed = 142 and 284 ppm, Qfeed = 150 cm3 min-1 and I = 38.4 WUV m-2; PCE: CPCE, feed = 1095 ppm, Final remarks and suggestions for future work 243 Qfeed = 300 cm3 min-1 and I = 38.4 WUV m-2). The results also showed that the effect of RH on the n-decane and PCE conversion is dependent on the organic load entering the reactor. - In the PCO over a TiO2-based paint using the same lab-scale annular photoreactor (Lr = 16 cm and Vr = 220 cm3) it was observed that the conversion of PCE decreased ca. 31 % (from 43 to 30 % when Qfeed* = 150 cm3 min-1, CPCE, feed = 1100 ppm and I = 38.4 WUV m-2) as RH decreased from 40 % to 3 % while n-decane conversion only decreased ca. 1 % (from 96 to 95 % when Qfeed = 150 cm3 min-1, Cdec, feed = 73 ppm and I = 38.4 WUV m-2) when RH decreased in the same range. The effect of irradiance was also assessed in the PCO experiments carried out in the labscale annular photoreactor towards the conversion of n-decane and PCE over photocatalytic films made by TiO2 powders or by photocatalytic paint. The radiation source is responsible for providing enough energy to generate electron-hole pairs on the photocatalyst surface, leading to the PCO of pollutants. Higher irradiance values will produce higher amount of electron-hole pairs that will participate in the redox reaction steps during the process and thus increasing the conversion of the pollutants. Therefore and as expected, in all PCO experiments where irradiance was a variable operational condition (n-decane and PCE PCO conducted in the lab-scale annular photoreactor packed with cellulose acetate monolithic structures coated with PC500, P25 or TiO2-based paint photocatalytic films), increasing the irradiance from 18.9 W m-2 to 38.4 W m-2 the conversion of n-decane or PCE also increased. Different kinetic rate expressions employed in the developed mathematical model for simulating the PCE kinetics through PCO over TiO2 P25 under UVC radiation (0.8 W m-2, spectral peak centred at 253.7 nm) and conducted in the lab-scale packed bed (glass spheres) annular photoreactor (Lr = 20.0 cm and Vr = 275 cm3) indicated that PCE and H2O molecules may have to be considered in association with different specific active sites of the surface of the catalyst as the Langmuir-Hinshelwood bimolecular non-competitive two types of sites model (M-4) described better the experimental data. However, in the n-decane PCO over TiO2-based paint coated in cellulose acetate monolithic structure and using the same lab-scale annular photoreactor (Lr = 16 cm and Vr = 220 cm3) a Langmuir-Hinshelwood mechanism where both species (water and n-decane molecules) compete for adsorption within different specific active sites (type 1 and 2) of the photocatalytic paint surface was able to describe the PCO of n-decane Regarding the effect of gas-phase molecular oxygen, the conversion of n-decane and PCE by PCO (PCO experiments conducted in the lab-scale annular photoreactor packed with cellulose acetate monolithic structure coated with PC500 photocatalytic film) was drastically impaired or Chapter 9 244 even not observable in the absence of such species indicating the key role of oxygen in photocatalysis (Qfeed = 150 cm3 min-1, Cdec, feed = 71 ppm, CPCE, feed = 1100 ppm and I = 18.9 - 38.4 WUV m-2). In the PCO of PCE over TiO2-based paint coated on cellulose acetate monolithic structures (lab-scale photoreactor - Lr = 16 cm and Vr = 220 cm3) removing oxygen from the feed, only 15 % of PCE was converted corresponding to an up to 56 % reduction in the PCO efficiency depending on the incident irradiance (Qfeed = 150 cm3 min-1, CPCE, feed = 1095 ppm and I = 18.9 - 38.4 WUV m-2). Based on the intermediate analysis by GC/MSD and on the experimental results three major reaction mechanisms may be implicit in PCE and n-decane PCO. One is the formation of reactive species from the adsorbed gas-phase molecular oxygen, which will oxidize the pollutants. The second involves the action of the oxygen from the lattice of TiO2. The third mechanism involves the classical hydroxyl radical formation on the TiO2 surface which acts directly over the pollutant molecules or as a trap for holes delaying the charge recombination and therefore promoting the formation of other oxidant species. In the case of PCE, chlorine radicals, Cl•, chain propagation reactions are also involved in the PCO reaction mechanism. Through chain propagation reactions, Cl• may be essential to maintain the process while the action of HO• radicals is probably related to the initiation of the PCO process (necessary step to generate Cl• radicals). 9.1.2.1. TiO2 photocatalytic properties enhancement Effective visible light active nitrogen modified TiO2 P25 photocatalysts were synthesized by a simple and low-cost preparation method, well suited for scale-up mass production. XPS analyses showed that anion-like nitrogen (N-) was present in the structure of TiO2, as O–Ti–N and Ti–O–N linkages. The catalytic activity of N-modified TiO2 materials for the degradation of an indoor air pollutant, PCE, under simulated solar radiation, and for the inactivation of E. coli bacteria in aqueous solution, under UVA radiation was found to depend on the amount of urea used in the preparation method. The calcination temperature also influences the catalytic activity as observed in the aqueous-phase degradation experiments of the pharmaceutical pollutant, diphenhydramine (DP), under visible light illumination. Nitrogen-modification of TiO2 P25 reduced the photocatalytic activity towards the degradation of gas-phase PCE under simulated solar radiation. While bare TiO2 P25 presented the highest activity (67 % of PCE conversion) only 35 % of conversion was attained over Final remarks and suggestions for future work 245 N0.50P25-380 (material with a urea:TiO2 weight ratio of 1:2 and calcined at 380 ºC). Regarding the aqueous-phase experiments over the same materials, N0.50P25-380 exhibited the highest photocatalytic efficiency for DP degradation and completely inactivated E. coli bacteria in 10 min; TiO2 P25 needed almost 40 min to completely inactivate E. coli under the same experimental conditions. However, further tests are needed in order to clarify the overall photocatalytic mechanism of nitrogen modified TiO2 samples towards the elimination of harmful organic molecules and microorganisms inactivation whether present in air or water media are still not fully understand. N-modified TiO2 nanotubes exhibiting visible-light photocatalytic properties were synthesized at low temperature by the hydrothermal treatment method owning significant absorbance in the visible-light range. Transmission electron microscopy evidenced the unidimensional nature of the TiO2 materials; however, their structure seemed to be sensible to a high calcination temperature (400 ºC). From BET surface area results, it was suggested that the closest to one was the TiNT/urea ratio, the higher the specific surface area of the nitrogenmodified TiNT. This fact suggests a commitment between the amount of doping and the treatment temperature in order to define the optimal doping conditions of the catalyst. X-ray photoelectron spectroscopy indicates that the chemical states of the nitrogen in TiO2 nanotubes may coexist in the form of N-Ti-O and Ti-O-N. Although for higher N/Ti ratio nitrogen seems to be concentrated at surface being more easily sputtered through Ar+ ion bombardment. Synthesised materials poorly degraded MEK and its mineralization was not efficient, generating intermediates: probably ethane, ethylene, formic acid, and/or formaldehyde. These compounds remained at the surface of the photocatalyst promoting its strong deactivation. Nevertheless, it was found that nitrogenmodified TiNT material is active photocatalyst for H2S degradation under UV radiation. A moderate photocatalytic activity was also observed under solar light. In both cases, SO2 was generated (with low concentration under UV radiation) and should be kept as low as possible since it still is a hazardous gas. 9.1.3. Photocatalysis using a pilot-scale photoreactor A vertical and south-oriented pilot-scale, single-pass continuous-flow photocatalytic reactor (LR = 144 cm and Vr = 4300 cm3) featuring a CPC and inner UVA lamps was proposed for air depollution purposes. This innovative reactor may operate 24 h a day using both types of radiation, solar and artificial. The PCO of gas-phase n-decane yields conversions up to 100 %, when using P25 or PC500 photocatalytic films for Qfeed = 2 L min-1 (τ = 44 s) and Cdec, feed = 10 ppm, under solar Chapter 9 246 irradiances of 15 WUV m-2 in the morning and 3 WUV m-2 in the afternoon. This behaviour was due the differences in adsorption-desorption phenomena between both periods which are caused by the raising temperatures in the morning and decreasing temperatures in the afternoon. Regarding the film made by photocatalytic paint (PCP film), up to 45 % of the initial n-decane (Qfeed = 2 L min-1, τ = 44 s and Cdec, feed = 10 ppm) was converted under 48 WUV m-2 in the morning reducing from this point until the sunset. Measuring the CO2 formation and subsequently the mineralization of n-decane, P25 and PC500 films enabled complete conversion into CO2 and water in all irradiance range while over PCP film 40 % of n-decane was mineralized under high irradiance conditions (above 45 WUV m-2). An important effect of the UV radiation on the mineralization at n-decane conversion was observed, suggesting that at high irradiance conditions the excess of photons favours the direct reaction pathway to the formation of CO2 and water. Under artificial UVA radiation, P25 film enabled 71 % conversion of the initial n-decane while 100 % was achieved over PC500 film. Over PCP film 52 % out of 25 % of the initial n-decane was mineralized into CO2 and water. As expected, PC500 film promoted higher conversion of n-decane by PCO most likely due to the higher specific surface area in comparison to that of P25. On the other hand, n-decane PCO over PCP film resulted in the lowest conversion values within the operational conditions of this study. However, considering the wide range of paint applications, results can be seen as promising for air treatment. For this reason it is mandatory to keep improving the performance of photocatalytic paints as well as understand the phenomena behind their photoactivity. 9.2. Suggestions for future work As suggestions for future work, it would be interesting to develop new photocatalysts with high activity in the visible light region beside those described in Chapters 6 and 7. Moreover, considering the possible (eco)toxicity of nanoscale catalysts (< 100 nm) research focused in modifying the morphology of TiO2 to nanorods, nanosheets or nano whiskers or even forcing irreversible agglomeration of TiO2 nanoparticles would also greatly benefit gas-phase photocatalysis as a clean and safe technology. Finding new and long-lasting substrates to support the photocatalysts will also add value to such technology. Regarding the photocatalytic reaction mechanisms it would be interesting coupling a mass spectrometer to the gas chromatographer used in this work which would enable the prompt identification of intermediates or by-products of pollutants. In this way, a deeper knowledge about photocatalysis mechanism of reaction would be acquired. This concept can also be applied to the paint used within this work: several authors claim that photocatalytic paints release VOCs, such Final remarks and suggestions for future work 247 as acetaldehyde and formaldehyde and free radicals; therefore, more and deeper studies must be conducted focused on the influence of paints components, interfering pollutants, as NO2, on the photocatalytic reaction mechanisms. The development of photocatalytic membrane reactors allows the immobilization of the photocatalyst in the form of a porous membrane being capable of not only selective permeation but also a photocatalytic reaction. In this way, the reactants are forced by convection to the TiO2 membranes while the PCO occurs on the outer and inner surface of the porous TiO2 membranes where high concentrations of HO⦁ radicals are expected. Finally, the photocatalytic process can be enhanced by using Microreactors/Static mixers and through breakthrough optical system designs for the optimal transfer of light radiation (either natural/simulated sunlight or artificial) between the source and the reactor (non-imaging optics techniques). Computational fluid dynamics tool (CFD-tool) can also be employed providing data related to the hydrodynamic conditions inside the reactor as also to the reaction at the photocatalyst surface. Appendix 254 sat = i i ip p  A.6 where pisat [atm] is the saturation pressure of component i at saturation temperature (Tisat in ºC), and φi is the relative vapour pressure. The relative vapour pressure depends on the flow rate that passes through the liquid solution (Qi in cm3 min-1). The generation of contaminated air streams in the lab-scale unit has shown that both H2O and VOCs relative vapour pressures vary with the flow rate (see Figure A.2): 13 13-32-5 mincm25.9590= mincm25593.0102.1104.3=     ii iiii Q QQQ   A.7 for Qi measured at 1 atm and 298 K. Figure A. 2. Effect on component (i)-relative vapour pressure (φi = pi/pisat) applying different air flow rates (Qi): experimental data obtained for H2O (points) determined at 281.15 K [a measured at 293.15 K and 1 atm]; lines for eq. A.7. The saturation pressure (pisat in mmHg) and temperature (Tisat in ºC) are correlated through the Antoine equation:   sat sat =log ii i ii TC B Ap   A.8 where Ai, Bi, and Ci are the component(i)-specific Antoine coefficients (Table A.2). Master Gas Chromatography (MGC) 255 Table A.3. Component(i)-specific Antoine coefficients: PCE, n-decane and water [2]. Component (i) Specific Antoine coefficients* Temp. range [ºC]# Ai Bi Ci Tmin Tmax PCE 7.06832 1458.45 226.986 -22.35 346.85 n-decane 7.21745 1693.93 216.459 -29.66 345.30 water (H2O) 8.05573 1723.64 233.076 0.01 373.98 Antoine equation: log pi sat= Ai − Bi Ci + Ti sat , with pisat in mmHg and Tisat in ºC * values based on regression of experimental data and researched estimates; # Tmin and Tmax describe the temperature range for which the equation is valid. Based on a derivation of the ideal gas equation of state, the generated component i molar flow rate (Fi in μmol·min-1) can be determined by: MFC TRFQp giii  A.9 where Qi [cm3 min-1] is the flow rate that passes through the VOC/H2O liquid solution, Rg is the gas constant (~8.206 × 10−5 cm3 atm K-1 μmol-1), and TMFC [K] is the temperature at mass flow controlling time. This stream can be diluted so, the total molar flow rate (FT in μmol·min-1) can be determined by MFC TRFQp gTTT  A.10 where pT [atm] and QT [cm3 min-1] are the total pressure and total flow rate of the contaminated air stream, respectively. The flow rate depends on the temperature so, considering the flow rate exiting the mass flow controller (measured at 298 K and 1 atm), it should be used TMFC = 298 K. Moreover, since the contaminated air stream is at atmospheric pressure, its component(i)-vapour mole fraction (yi) is given by: T i i T i in n yor F F y A.11 Thus, with the ideal gas equation of state: looplooploop TRnVp gT  A.12 From eq. A.12 is now possible determine the loop-collected total moles numbers in the vapour phase (nT in μmol), and then, the theoretical component(i)-moles number in the vapour phase (ni in mol). Finally, the theoretical component(i)-concentration of the stream feeding the reactor (Ci, feed in μmol cm-3) can be calculated using eq. A.12, for CVOC, feed = Ci, feed and nVOC, feed = ni, feed. Master Gas Chromatography (MGC) 257 A.3. References [1] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry, Wiley, New York, 1984. [2] C.L. Yaws, P.K. Narasimhan, C. Gabbula, Yaws' handbook of Antoine coefficients for vapor pressure, 2nd electronic ed., Knovel, 2009.