Noble metal modified TiO2 microspheres: Surface properties and photocatalytic activity under UV–vis and visible light

Accepted Manuscript Title: Noble metal modified TiO2 microspheres: surface properties and photocatalytic activity under UV–vis and visible light Author: E. Grabowska M. Marchelek T. Klimczuk G. Trykowski A. Zaleska-Medynska PII: DOI: Reference:

S1381-1169(16)30243-6 http://dx.doi.org/doi:10.1016/j.molcata.2016.06.021 MOLCAA 9927

To appear in:

Journal of Molecular Catalysis A: Chemical

Received date: Revised date: Accepted date:

5-1-2016 22-6-2016 23-6-2016

Please cite this article as: E.Grabowska, M.Marchelek, T.Klimczuk, G.Trykowski, A.Zaleska-Medynska, Noble metal modified TiO2 microspheres: surface properties and photocatalytic activity under UV–vis and visible light, Journal of Molecular Catalysis A: Chemical http://dx.doi.org/10.1016/j.molcata.2016.06.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Noble metal modified TiO2 microspheres: surface properties and photocatalytic activity under UV-Vis and visible light E. Grabowskaa, M. Marcheleka, T. Klimczukb, G. Trykowskic, A. Zaleska-Medynskaa a

Department of Environmental Technology, Faculty of Chemistry, University of Gdansk, Wita Stwosza 63, PL 80-308 Gdansk, Poland

b

Department of Solid State Physics, Faculty of Applied Physics and Mathematics, Gdansk University of Technology, G. Narutowicza str. 11/12, 80-233 Gdansk, Poland c

Faculty of Chemistry, Nicolaus Copernicus University, 7 Gagarina str, 87-100 Toruń, Poland

1

decreasing activity in the gas phase reaction

fi tric Elec

eld

ble Visi t ligh

E CB le Nob l a met

decreasing activity in the aqueous phase reaction

E VB

`

2

Highlights 

Ag, Au, Pt or Pd-TiO2 is active in toluene removal from the gas phase using LEDs



Decrease in NPs metal loading resulted in increase of M-TiO2 photoactivity



Vis light photoreactivity changed in order: Ag-TiO2  Pd-TiO2 > Pt-TiO2 >> Au-TiO2

3

Abstract Composite photocatalysts that consist of TiO2 and noble metal nanostructures have been considered to be the promising and pivotal material for accessible enhancement of the efficiency in the photocatalytic process carried out in the aqueous and gas phases. In this work we fabricated porous TiO2 microspheres through a hydrothermal process followed by photochemical reduction of noble metal nanoparticles at the TiO2 surface. The morphology and structure of M-TiO2 spheres (M= Ag, Au, Pt and Pd) were studied with the use of various techniques, including transmission electron microscopy (TEM), X-ray powder diffraction analysis (XRD), photoluminescence (PL) and UV–vis diffuse-reflectance spectroscopy (DRS). The effect of metal amount (from 0.1 to 1 wt.%.) on the photocatalytic activity during toluene degradation in gas phase and phenol degradation in aqueous phase was investigated. Additionally, the photocatalytic activity of the M-TiO2 samples was evaluated by measuring the formation rate of photo-induced hydroxyl radicals (•OH) under UV-vis light irradiation using coumarin as a probe. The obtained results indicated that toluene could be mostly removed from the air over TiO2 microspheres modified with Ag, Au, Pt, and Pd nanoparticles. UV-mediated photoreactivity was almost similar for all samples obtained by loading metals from solutions consisting of 0.1 and 1 wt.% of metal precursors. Under visible light, except for Au, in gas phase toluene oxidation, the optimized loading of the metals was 0.1 wt-% (photoreactivity changed in order: Ag-TiO2 Pd-TiO2> Pt-TiO2>> Au-TiO2). In case of phenol degradation in the aqueous phase, in the presence of UV irradiation the highest amount of metal (1.5 wt.%) was profitable, while under the Vis light reaction the medium amount of metal (0.5 wt.%) was beneficial. Additionally, it was noticed that phenol was degraded not only via oxidation by •OH radicals but probably also in direct reaction with the photogenerated carriers (e-/h+), particulary in the presence of TiO2 spheres loaded with Au and Ag nanoparticles. Keywords: TiO2 porous microspheres, synthesis, photoactivity, noble metals 1.

Introduction

Heterogeneous photocatalysis has been proposed as a promising process to remove pollutants from air and water streams due to possibility of organic pollutants oxidation to CO2 and H2O in the presence of semiconductors [1-10]. TiO2 is one of the most promising materials for environmental remediation due to its non-toxicity, photo-stability and 4

relatively low cost [11]. However, the wide band gap (3.2 eV for anatase) and the rapid recombination of photogenerated charge carriers mainly limit further applications of anatase [12]. To improve the photocatalytic reactivity of TiO2 and to extend its light absorption into the visible region, substitution of non-metal atoms, transition metal doping, reduced forms of TiOx, coupling some semiconductors with narrow bandgap were proposed [12-17]. Among these techniques, noble metal nanoparticles (especially Au, Pt, Pd and Ag) modification of TiO2 is also used to inhibit recombination of electron- hole recombination process and to extend absorption properties of TiO2 into visible region due to surface plasmon resonance(SPR) [18, 19]. Noble metal nanoparticles absorb and scatter light in the visible range as a result of SPR, where the resonant wavelength strongly depends on the particle size, shape, assembly state and surrounding dielectric environment [20-22]. Furthermore, it has been found that various TiO2 with different morphology and size, such as nanosheet, nanofibers, nanotube and microspheres, possess good adsorptive and photocatalytic properties [23-25]. The fabrication of TiO2 3D structures, such as microspheres has recently attracted enormous attention due to their low density, high surface-to-volume ratio, high surface area, good surface permeability and high photocatalytic activity [26-29]. Thus, it could be expected that higher energy conversion efficiency and photocatalytic activity could be achieved by using TiO2 microspheres as photocatalysts [29]. Controlling of volatile organic compounds in the atmosphere is a major environmental problem now. Toluene is a volatile liquid (22 mm Hg at 20°C) and it is released into atmosphere by industrial and consumer uses [30]. The largest sources of toluene effluents are the evaporation from gasoline, paints, paint thinners, fingernail polish, lacquers, adhesives, rubber, cigarette smoke and some printing and leather tanning processes [31]. Toluene is an irritant to the skin and mucous membrane, and can act as an anesthetic to the central nervous system [32, 33]. Thus, one of the challenges in the field of photocatalytic air treatment systems is the development of photoactive materials activated by low powered and low cost irradiation sources (such as LEDs) to remove VOCs, such as toluene. Toluene from the gas phase was efficiently removed over LED-irradiated TiO2 nanotube arrays, obtained by anodic oxidation of titanium foil in ethylene glycol-based electrolytes [1]. It was found that the preparation conditions affected the TiO 2 nanotubes morphology (length of tubes, top-opened or clogged and wall smoothness) as well as their photocatalytic 5

activity in the air purification process. The highest photoactivity in toluene degradation reaction was observed for longer (4.3-μm- and 5.9-μm-long nanotube arrays) obtained by Tifoil anodization in ethylene glycol-based electrolyte. 30-min of irradiation by twenty five UVLEDs (λmax = 375 nm, 63 mW per diode) was enough to complete the removal of toluene (C0 = 100 ppm) from the gas phase in the presence of TiO2 nanotube [1]. Cao et al. found, that toluene could be selectively photooxidized to benzaldehyde in the presence of TiO2 hollow spheres [34]. They observed that TiO2 hollow spheres exhibited higher photo-efficiency than commercial TiO2 Degussa P-25 and the conversion of toluene gradually increased from 9.0% to 21%, when the hydrothermal time of the synthesis increased from 20 min to 6 h. It was attributed to higher UV absorbance and lower recombination of free carriers of these hollow TiO2 spheres with exposed {001} facets due to their unique high surface crystallinity [34]. Phenols are generally considered to be one of the important organic pollutants discharged into the environment causing unpleasant taste and odor of drinking water [4]. Recently, the TiO2 microspheres have an attractive advantage in water treatment due to the high lightharvesting capacity and easy mass transportation [27, 35, 36]. Yang et al prepared WO3/TiO2 hollow microspheres by a spray drying method and observed that WO3/TiO2 photocatalyst shows higher photocatalytic activity than pure TiO2 in photodegradation of phenol [35]. Photocatalytic activity of TiO2 microspheres was measured by Wang et al in the photodegradation process of sulfosalicylic acid (SSA), phenol, and 2,4-Dichlorophenoxyacetic acid (2,4-D) [27]. The results showed that TiO2 microspheres had strong adsorption ability, which significantly contributed to the overall degradation rate of all three organic compounds [27]. Herein, we modified the surface of TiO2 microspheres (obtained via hydrothermal route) with noble metal nanoparticles using the photodeposition method. The M-TiO2 photocatalysts (M= Ag, Au, Pt and Pd) were studied for photodegradation of toluene in the gas phase using low-powered irradiation source (light-emitting diodes, λmax = 375 and 415 nm) and phenol in aqueous phase under UV-Vis and visible irradiation. Moreover, in the second part of our research hydroxyl radical formations were investigated. Fluorescence of irradiated coumarin solution was used as a method of •OH radical detection. Coumarin readily reacts with generated hydroxyl radicals forming hydroxycoumarins. Although the major hydroxylation product is 5-hydroxycoumarin, only 7-hydroxyproduct of coumarin 6

hydroxylation emits fluorescent light [37, 38]. Thus, this method was used only for hydroxyl radical detection, but not for determining concentration of hydroxyl radicals. To our best knowledge, correlation between surface properties of such new composites TiO2 microspheres decorated with noble metal nanoparticles- and photoactivity, including: (a) toluene degradation in the gas phase, activated by light-emitting diodes (LEDs), (b) phenol degradation in the aqueous phase, and (c) •OH radicals formation was done for the first time. The effect of the type and size of noble metal nanoparticles on the photoactivity of M-TiO2 was also discussed. Because of the fact that LEDs are low-powered and low-cost irradiation sources, there exists a promising possibility of reducing power consumption and costs of photocatalytic process. 2. 2.1

Experimental Chemicals and materials

All chemicals were analytical grade and used without further purification. The Degussa P-25 TiO2 was supplied by Evonik Industries, and used as obtained. Titanium (IV) butoxide (TBT) (97%, Sigma-Aldrich) was used as titanium source for the preparation of TiO 2 microspheres. KAuCl4 (98%), Pd(C5H7O2)2 (99%), H2PtCl6 (99%) and AgNO3 (99%) from Sigma-Aldrich were used as a metal source in the preparation procedure. Coumarin was purchased from Sigma– Aldrich Co. (Germany) and phenol was obtained from POCh S.A. (Poland). 2.2

Preparation of TiO2 microspheres

TiO2 microspheres were prepared according to the procedure described by Zheng et al [39]. The hydrothermal technique is an important tool to obtain advanced nanostructural materials. Moreover, by using this method it is possible to synthesized TiO 2 with homogeneity, high purity, crystal symmetry, metastable compounds with unique properties and narrow particle size distributions [40]. In the next step, 70 cm3 of 99.8 % ethanol (alcohol was used as hole scavenger [41]), solution containing TiO2 microspheres (1 g) and metal precursor (0.1, 0.5 or 1 % m/m Ag, Au, Pt, and Pd, respectively) was sonicated for 10 min, stirring in the dark for 10 min, degassed with nitrogen in the dark for 20 min and finally illuminated by 250 W Xe lamp (light flux 30.8 mW/cm2) used as an irradiation source for 1 h. Initial concentrations of noble metal precursors in ethanol were: 7.33; 3.67 and 71.45x10 -4 mol/dm3 for KAuCl4, 0.14; 6.78 and 1.34x10 -4 mol/dm3 for Pd(C5H7O2)2, 7.40; 3.70 and 73.31x10 -4 mol/dm3 for H2PtCl6 and 0.13; 6.65 and 1.32x10 -4 mol/dm3 for AgNO3. 7

Obtained samples were rinsed with deionized water and dried at 40°C, without calcination. On the basis of literature data and our own experience it is known that the photodeposition method is a powerful technique to obtain metal nanoparticles of controlled size and shape in solution and deposited at the surface of different matrix [42, 43]. Under UV light irradiation, the illuminated anatase TiO2 generates in aqueous medium photo-excited electrons and positive holes. Afterwards, noble metal ions adsorbed on the surface of TiO 2 particles can react with the photogenerated e- to form Ag0, Au0, Pt0 and Pd0. The description of the asprepared photocatalysts is shown in Table 1. The amount of silver, gold, platinum and palladium precursors taken for photocatalyst preparation was calculated on the assumption that the content of Ag, Au, Pt, and Pd in the photocatalysts after synthesis should be equal to 0.1 - 1 % m/m of the photocatalyst dry mass. 2.3

Characterization of TiO2

The purity of the samples was confirmed with powder X-ray diffraction PXRD, X’Pert Pro MPD Philips diffractometer, with Cu Kα radiation λ = 1.5418 Ǻ. The measurements were performed on the 2θ range of 20 to 80 degrees, with the scan speed 20 deg./hour. The lattice parameters were estimated by the LeBail method using FullProf package [44]. To characterize the light-absorption properties of modified photocatalysts, diffuse reflectance (DRS) spectra in the scan range 200–900 nm were recorded. The measurements were carried out on a UV-Vis spectrophotometer (Evolution 220, Thermo Scientific) equipped with an integrating sphere and BaSO4 was used as the reference. The surface area of the bare TiO2 microspheres were evaluated from the adsorptiondesorption isotherms of liquid nitrogen (77 K) detected using a Micrimeritics Gemini V (model 2365). A TiO2 sample was dried and degassed in a sample cell at 200 °C for at last 2 h before the adsorption. The specific surface areas of the photocatalyst was determined by Brunauer-Emmett-Teller (BET) method in the relative pressure (p/p0) range of 0.05–0.3. Particle size, shape, dispersion uniformity and chemical composition of samples have been analyzed by a transmission electron microscope with energy-dispersive X-ray spectrometer (TEM-EDX, Tecnai F20 X-Twin, FEI and EDAX spectrometer) and a scanning electron microscope (SEM, Quanta 3D FEG, FEI). To study the recombination of electrons–holes in the photocatalysts the photoluminescence emission spectra (PL) were measured at room temperature on an LS-50B Luminescence

8

Spectrophotometer equipped with a Xenon discharge lamp as an excitation source and an R928 photomultiplier as a detector. The obtained samples were excited with 310 nm. The excitation radiation was falling on the sample surface at an angle of 90°. Additionally, the emission filter was applied to block UV radiation above 390 nm. 2.4

Measurement of photocatalytic activity in the gas phase

The photocatalytic activity of TiO2 microspheres modified with Ag, Au, Pt, and Pd NPs in the gas phase reaction was determined by the toluene degradation process. The concentration of toluene in a gas mixture was about 200 ppm. The photocatalyst activity tests were carried out in a flat stainless steel reactor described in our previous study [1]. The obtained TiO2coated support was placed at the bottom side of the photoreactor. Subsequently, the gaseous mixture was passed through the reaction space for 2 min. After closing the valves, the reactor was kept in the dark for 10 min in order to achieve equilibrium. An array of 25 LEDs (λmax = 375 nm or λmax = 415 nm) was used as an irradiation source. Measured light flux for LEDs λmax = 375 nm and LEDs λmax = 415 nm was 14.2 and 1.2 mW/cm2, respectively. The analysis of toluene concentration in the gas phase was performed with the use of gas chromatograph (Termo Scientific Trace 1300) equipped with a flame ionization detector (FID). 2.5

Measurement of photocatalytic activity in the aqueous phase

The photoactivity of the obtained samples was estimated by measuring the rate of phenol decomposition in an aqueous solution in the presence of visible and UV-Vis irradiation described in our previous study [42]. Phenol was selected as a model pollutant because it is a non-volatile and common contaminant present in industrial wastewaters. The mechanism of phenol decomposition is also well established both under UV and visible-light irradiation [4]. 2.5.1 Photocatalytic activity test in the presence of UV-Vis light For photocatalytic reaction in an aqueous phase, TiO2 powder (7.5 mg) was suspended in phenol solution (Co = 500 mg/dm3, 15.0 ml). The photocatalyst samples were placed in pyrex tubes. The optical path included a glass filter which cut-off wavelengths shorter than 350 nm. After 30 min in the dark, the suspension was photoirradiated with an Hg medium pressure lamp (Heraeus Noblelight GmbH, light flux 5.5 mW/cm2) under magnetic stirring (500 rpm). The temperature of the suspension during photoirradiation was maintained at 10 ˚C±0.5 using a thermostatically controlled water bath. Aqueous suspension (1.5 ml) was 9

collected at regular time periods during irradiation and filtered through syringe filters (Ø = 0.1 μm) to the remove photocatalyst. Phenol concentration was estimated by colorimetric method (λ= 480 nm) after derivatization with diazo-p-nitroaniline using UV–vis spectrophotometer. 2.5.2 Photocatalytic activity test in the presence of Vis light The aqueous phase contained 125 mg of the photocatalyst, 24 cm3 of deionized water and 1 cm3 of phenol (Co = 500 mg/dm3). The photocatalyst loading was 5 g/dm3. The prepared suspension was stirred and aerated (V = 5 dm3/h) for 30 min in the dark to obtain equilibrium and then the content of the reactor was photoirradiated with a 1000 W Xenon lamp (Oriel, light flux 0.5 mW/cm2) which emitted both UV and Vis irradiation. The photoreactor (V = 25 cm3, i.d. 37 mm, length 30 mm) was equipped with a quartz window and exposure layer thickness was 3 cm. The optical path included a water filter and a glass filter (GG 420) which cut-off wavelengths shorter than 420 nm. The temperature of the aqueous phase during irradiation was kept at 10◦C using water bath. During the irradiation the suspension (1 cm3) was collected, and filtered through syringe filters (Ø = 0.2 µm) to remove the photocatalyst particles. Phenol concentration was estimated by means of the colorimetric method (λ= 480 nm) after derivatization with diazo-p-nitroaniline using a UV–vis spectrophotometer (Thermo Evolution 220). Photocatalytic degradation runs were proceeded by blind test in the absence of a photocatalyst or illumination. No degradation of phenol was observed in the absence of either the photocatalyst or illumination. 2.6

Quantum yield of the photocatalyst

The efficiency of the photocatalytic process is controlled by the system’s light absorption characteristics and can be measured as quantum yield (QY), which is defined as the number of events occurring per photon absorbed or the reaction rate divided by the photonic flux. The calculation of QY can only be correct at the beginning of the reaction when only a small amount of byproducts have been generated at concentrations which will not significantly alter the solution’s initial absorbance, so the QY was calculated after 5 min of toluene degradation in the first measurement cycle. Quantum yield for the degradation of toluene under UV and Vis LEDs was calculate as a amount of reactant consumed or product formed in the bulk phase, to the amount of photons at wavelength λ absorbed by the photocatalyst

10

[45]. For both kind of LEDs (λmax = 375 nm or λmax = 415 nm) based on the irradiation spectra for the calculations of QY it was assumed that they emit monochromatic light.

2.7

Formation of hydroxyl radicals

The formation of hydroxyl radicals in a suspension of TiO2 under UV-vis irradiation was evaluated with a photoluminescence (PL) technique using coumarin as a probe molecule, which readily reacted with •OH radicals to produce the highly fluorescent product, 7hydroxycoumarin. The formation of hydroxyl radicals was estimated by using the same experimental set-up as it was the case with measuring decomposition rate of phenol under UV-vis light. The concentration of TiO2 and coumarin were 0.5 g/dm3 and 10-3 M, respectively. After irradiation, the solution was filtered and analyzed on an LS-50B Luminescence Spectrophotometer equipped with a Xenon discharge lamp as an excitation source and an R928 photomultiplier as a detector by excitation with radiation of a wavelength of 330 nm. 3.

Results and discussion

3.1

Crystal structure

Figure 1 shows normalized X-ray diffraction patterns (PXRD) for metal loaded TiO2 microspheres. All tested samples reveal the spectra expected for the anatase form of TiO 2 (upper tick marks) with the lattice parameters consistent with the reference TiO2 sample. The lower set of tick marks represents a corresponding noble metal. The PXRD patterns for TiO2 microspheres modified with Pd (Figure 1A) and Ag (Figure 1B) do not show traces of the used noble metals. In contrast, a small diffraction peak of Pt and Au is visible and marked by an arrow – see insets in Figure 1C and D. This can be explained by a relatively small amount of the noble metal and 70% greater atomic form factor of Pt and Au comparing with an X-ray form factor of Pd and Ag. The average TiO2 crystallite size was calculated with the Scherrer equation. For these calculations the full-width at half-maximum of the four PXRD peaks (011), (020), (015) and (121) were taken into consideration. The estimated crystallite size varies from 4.9 to 5.5 nm for Pt-TiO2 and Au-TiO2, respectively. This value is in good agreement with the crystallite size estimated for the reference TiO2 sample (6.0 nm). 11

3.2

Morphology

Figure 2A shows the representative SEM image of the as-prepared TiO2 sample, which appear as microspheres. The particle size distribution of the TiO2 microspheres was estimated by measuring dimensions of TiO2 microsphere particles across a large number of SEM micrographs (the average value calculated was based on measurements of 50 particles). It can be noticed that as-prepared TiO2 microspheres had diameters in the range of 1-18 μm, but most of the spheres (about 40 %) are within the range of 6-9 μm (see details in Figure 2C). Moreover, observation at high magnification shows that the surface of the microspheres is fairly smooth, indicating that they are composed of ultrafine particles. Also Tang et al. obtained TiO2 mesoporous microspheres by a facile template-free solvothermal method of approximately 2–4 μm in size [46], while, Wei et al. and Zheng et al. synthesized TiO2 microspheres by the hydrothermal method and obtained samples with diameters in the range of 5–10 μm and 3–6 μm respectively [47, 48]. To confirm that the obtained structure is not destroyed during the photodeposition of noble metals, the SEM image of a sample containing silver NPs on the TiO2 surface was made (Figure 2B). It can be clearly seen that the spherical structure was maintained. Compared with unloaded TiO2 microspheres, the surface of noble metal NPs loaded samples become rough as shown in Figure 3, which indicates that metallic nanoparticles were formed at the TiO2 microspheres surface during photodeposition process. The bright spots visible at the surface of TiO2, indicate the presence of metal nanoparticles. The SEM image of 0.5 Ag_TiO2 shows that Ag NPs are deposited homogenously on TiO2 surface (Figure 3A). In other cases, the metal nanoparticles are unevenly and randomly distributed on the surface of the sample (Figure 3B-D). TEM images clearly show the presence of spherical metal nanoparticles lying on the curved surface of TiO2 sphere, as shown in Figure 4. Metal nanoparticles differed in size depending on the type of metal. The presence of metallic Ag, Au, Pd and Pt was confirmed by the EDX analysis taken during the TEM measurements at various regions of NPs-modified TiO2 microspheres (Figure 4). The size of photoreduced noble metal NPs were found to be on average from 10 to 30 nm in diameter. It was observed that photoirradiation of solution containing 0.5 M of Pt(IV) or Pd(II) ions, resulted in the formation of smaller nanoparticles (~10 nm) than photoirradiation of solution consisting the same amount of Au(III) or Ag(I) ions (nanoparticles ranged from 20 to 30 nm). This finding is consistent with our previous 12

study dealing with photoreduction of noble metal NPs at the surface of TiO 2 nanosheets with exposed {001} facets [42]. Also according to literature reports, Pt and Pd nanoparticles formed by UV reduction, chemical reduction or radiolysis are usually smaller than those made of Au or Ag. Galhenageet al. reported nucleation and growth of Co, Au, Ni and Pt on reduced and oxidized TiO2 via metal evaporation [49]. They concluded, that the resulting different metal cluster size and density at the surface of titania could be attributed to different binding energy between metal and TiO2 surface. Au clusters revealed the largest average cluster heights of the four metals and the lower cluster density, while in the case of Pt, the average cluster height was twice lower and the cluster density was more than triple than that of Au [49]. Due to the intrinsic properties of Au, its integration with most metal oxides is relatively weak, in most cases weaker than the Au-Au bond. Based on the temperature programmed desorption of Au and theoretical calculations, it was revealed that the binding energy of Au to an oxide support is much smaller than the Au-Au bonds [50]. These relative energy difference lead to facile sintering of Au nanoparticles as a function of reaction time, i.e. small highly dispersed particles eventually convert to thermodynamically preferred larger particles. Both experimental data [50, 51] and DFT calculation [49] demonstrated that the metal atoms bind preferentially at the oxygen vacancies and it has generally been observed that for growth on the oxide surfaces defects play a crucial role in the nucleation of the metal clusters. Pt binds more strongly to the bridging oxygen vacancy (6.05 eV) than Au (-3.89 eV). Thus, observed variation in metal nanoparticles size could be attributed to the relative difference in binding energy between noble metal and TiO 2 surface. In case of silver, Harada et al. proposed plausible mechanism, composed of a few sequential stages, such as autocatalytic reduction-nucleation, nucleation-growth, Ostwald ripening, and dynamic coalescence [52-54]. In the rapid nucleation step, small nuclei or particles (~2.5 nm) are formed through autocatalytic reduction. Subsequently, the diffusion-limited Ostwald ripening-based growth appeared, and large particles with a mean radius of ~11.5 nm are built by dynamic coalescence in the subsequent stage of growth [53]. In case of palladium, it was demonstrated that the photoreduction process of Pd (II) aqua chlorocomplexes was a quite fast process. Thus, the Pd (I) complexes displayed no induction period, since the reduction of Pd (II) occurred immediately [52]. Hence, a different size of nanoparticles - built

13

of gold, silver, palladium or platinum - could be attributed to different mechanisms of formation or eventually a different rate of metal diffusion at the surface of TiO 2. 3.3

Absorption properties and BET surface area of TiO2 microspheres

The optical properties of pure and metal loaded TiO2 microspheres with different loading amounts of the noble metal nanoparticles were probed by UV–vis diffuse reflectance spectroscopy (DRS) which are shown in Figure 5. Compared with the bare TiO2 microspheres, all noble metal loaded TiO2 microspheres show an absorption in the visible region. The existence of the wide absorption band may be attributed to the surface plasmon resonance of noble metal nanoparticles. The max of the resonant spectra band for TiO2 modified with Ag and Au nanoparticles is given in Table 1. The visible absorption band is strongly influenced by particle size, particle shape and particle size distribution but it is also known that the main determining factor of localized surface plasmon resonance (LSPR) is the electronic structure of the metal [55]. As it was mentioned by Sellappan et al the interband transition threshold of Ag is around 3.9 eV (320 nm), which makes it possible to tune the plasmon resonance close to the near UV region, and in case of gold, the NPs transition threshold of Au is around 2.3 eV (520 nm), which restricts (due to damping of resonance by interband transitions) the upper energy limit to 2.3 eV [56]. Moreover SPR effect of the Ag nanoparticles can be employed to improve the solar energy conversion efficiency by enlarging light absorption to much longer wavelength, and increasing light scattering and motivating photogenerated carriers in the semiconductor by transferring the plasmonic energy from the Ag0 to the TiO2 semiconductor [57]. The metal nanoparticles often act as electron scavengers improving charge separation within the semiconductor–metal photocatalyst system [58]. However, in a few cases, the metal NPs which are present on the surface can act as recombination centers [59]. It has been observed that pure TiO2 microspheres exhibited a photo-response in the UV region with wavelengths below 380 nm and a weak absorption in the visible region, which can be attributed to intrinsic band gap of TiO2. The photodeposited Ag NPs cause significant changes to the absorption spectrum of TiO2 microspheres resulting in high absorbance from 400 nm to the entire visible region, which is characteristic of surface plasmon absorption of spatially confined electrons in Ag NPs (Figure 5A). The absorbance in the visible region for the Ag_TiO2 photocatalysts system increased from 400 to 500 nm and then decreased 14

gradually. Silver nanoparticles are known to exhibit a plasmon band with a maximum at around 400 nm in water. This plasmon band is sensitive to the environment and can be shifted depending on the stabilizer or on the substrate. Because of the coupling between Ag nanoparticles and TiO2 support having a high reflective index (the absorption coefficient and refractive index are for anatase phase 90 cm–1 and 2.19 at a wavelength of 380 nm, respectively), the plasmon band is blue-shifted. It is consistent with our previously study where we obtained a series of TiO2 photocatalysts (commercial compounds P25 and ST01) modified with Ag nanoparticles induced by radiolysis [60]. It has been reported that that LSPR absorption appears in near-UV, vis, and near- or midinfrared (IR) regions depending on the size and shape of the gold nanoparticles [61-66]. After depositing Au-NPs, the Au_TiO2 photocatalysts exhibited a visible-light absorption in the region ranging from 400 nm to 700 nm, with the peak positioned in the range of 545-560 nm due to localized surface plasmon resonance (LSPR) of Au-NPs, corresponding to the observed color of obtained powders (Figure 5B). Kowalska et al. proved that the size and shape of gold nanoparticles determine the absorption ranges and that the activity depend strongly on the properties of gold and TiO2, such as particle size and shape, surface area, and crystalline form[67]. They observed two main absorption maxima of smaller (ca. 10–25 nm) and larger (ca. 30–60 nm) gold nanoparticles at 530 and 560–610 nm, respectively. It can be observed that for both Ag_TiO2 and Au_TiO2 photocatalysts, the LSPR depends on the amount of NPs loaded on TiO2 surface. In our case, it means that enhancement of the initial amount of silver and gold ions in the solution during preparation route resulted in a red shift of the LSPR band in the absorption spectrum of Au- and Ag-TiO2, respectively. Modification with noble metal NPs, such as platinum and palladium can also result in enhancements of the photoconversion quantum yield and allows the extension of the light absorption of wide band gap semiconductors to the visible light. Diffuse reflectance spectra of platinum and palladium modified TiO2 microspheres are shown in Figure 5C and 5D. The spectra of the loaded samples show a slight shift in the band gap transition to longer wavelengths for both kinds of surface-modified photocatalysts. For all Pt_TiO2 samples sharp absorbance edges occur at the wavelength of about 400 nm. Chen et al observed that the presence of platinum on the TiO2 surface could significantly enhance the absorption of light from the visible region [68]. Two different phenomena could be responsible for this observation: (1) the presence of Ti3+, resulted from the platinum 15

deposited on the TiO2 surface; and (2) the formation of the Schottky barrier between platinum and TiO2 preventing the recombination of electric holes and electrons [68]. For Pd– TiO2 samples the absorption band edge is strongly related to the amount of nanoparticles loaded on the surface of microspheres. It can be seen that for sample 0.1Pd_TiO2 sharp absorbance edges occur at the wavelength of about 400 nm, for samples 0.5Pd_TiO2 - about 430 nm and for sample 1Pd_TiO2 - about 450 nm. Cybula et al observed that for sample containing 0.5 mol% palladium on TiO2 nanoparticles obtained by water-in-oil microemulsion system sharp absorbance edges occur at the wavelength of about 450 nm [10]. Moreover, it is known from the literature, that in case of Pd NPs, the d-band overlaps with the s- and pbands in the whole range of interest, which results in broad optical resonances that are assigned to the so-called localized surface plasmon resonances (LSPRs) [69]. The band at ~440 nm (2.84 eV) could also be attributed to the d-d transition of PdO particles, that is, to the presence of PdO bulk phase [70]. Moreover, in case of TiO2 loaded with Pt or Pd nanoparticles, observed enhancement in absorption intensity could be a result of the ‘grayscale’ of the sample as compared to BaSO4, which increases with loading. To characterize the specific surface areas and porosity of TiO2 microspheres, nitrogen adsorption–desorption experiment was employed. The average pore volume was 0.116 cm3/g. In addition, the specific surface area of TiO2 microspheres is 235 m2/g calculated by the Brunauer –Emmett–Teller (BET) method, which may be due to the hollow structure which can greatly increase the internal surface. Based on the measurments of surface area and micropores size of TiO2 hollow spheres synthesized under different reaction conditions, Shang at al concluded that formation of the TiO2 spheres goes through the formation of poorly crystallized solid spheres and the transformation of the solid spheres to hollow ones is based on the Ostwald ripening process. During the transformation to the hollow spheres, the inner part of the solid spheres gradually dissolved and recrystallized on the shell of the spheres to finally form hollow spheres [36]. 3.4

Photocatalytic activity in toluene oxidation

Toluene photocatalytic degradation experiments were carried out according to the procedure described above. Photocatalyst was kept in the dark in the toluene-filled reactor for 10 min to estimate the equilibrium state and then irradiated by emitting diodes (λmax = 375 nm or λmax = 415 nm) as a low powered irradiation source. 16

Fig. 6 shows the toluene photodegradation efficiency as the function of irradiation time under UV LEDs irradiation in the presence was Ag-TiO2, Pd-TiO2, Pt-TiO2 and Au-TiO2 microspheres. It was found that all as-prepared TiO2 samples loaded with monometallic nanoparticles exhibited higher photocatalytic activity under UV light in the gas phase than pristine TiO2 microspheres (about 52%). The obtained results showed that in case of samples modified with Pt and Pd NPs almost no influence of precursor amount on photocatalytic efficiency was indicated (Fig. 6A and 6B). In both cases toluene degradation efficiency reached ~100% after 20 min of irradiation. It is in a good agreement with literature data, because it was investigated that Pd–TiO2 and Pt–TiO2 photocatalysts can be used for the UV (365 nm) photocatalytic degradation of toluene, benzene, trichloroethylene and vinyl chloride [71-75]. Moreover, the degradation of VOCs may depend significantly on their chemical bond strength and molecule stability [76]. Toluene is an aromatic compound with a ring of carbon atoms held together by alternating single and double bonds between the carbon atoms, which imparts an unusual stability to the structure. Kim et al investigated the effects of palladium nanoparticles deposited onto a TiO2 on the removal of toluene under UV irradiation [75]. They noticed that Pd–TiO2 photocatalyst significantly increased the degradation of toluene using a short retention time (5.7 s) and increased CO2 yield and decreased CO yield during toluene conversion. To explore the effect of metal precursor amount (in case of Pt and Pd) on the activity of Me-TiO2, further experiments with higher toluene concentration (C0 = 400 ppm) were performed and shown in Fig. 6E and 6F. These experiments clearly showed, that application of the smallest amount of Pt and Pd precursors during synthesis was favorable for activity of TiO2 decorated with metal nanoparticles. Toluene degradation efficiency reached 96% and 80% after 30 min of irradiation for the 0.1Pt_TiO2 and 0.1Pd_TiO2 sample, respectively. Zhang et al explained that decreasing effect of photoactivity with high noble metal loading at 1.0 wt% could be due to that excess NPs have occupied too many active sites on the TiO2 surface to reduce the overall efficiency and might cause a light shielding effect. This process may overtake the positive charge separation effect. Moreover, it is possible that at loadings beyond the optimum, the NPs deposits can behave as recombination sites due to the accumulation of excessive negative charges to attract positive holes, reducing the efficiency of charge separation[77]. Experimental data clearly indicate that in the presence of photocatalysts modified with Ag and Au NPs we can observe decrease in the efficiency of toluene degradation under UV LEDs 17

when 0.5Au_TiO2 and 0.5Ag_TiO2 samples were used during the process (Figure 6C and 6D). Toluene degradation efficiency reached about 82% after 30 min of irradiation, while in the presence of samples modified with 0.1 and 1 wt.% of silver and gold NPs 100% of toluene was degraded after 20 min. Silver-modified TiO2 with various Ag/Ti molar ratios were prepared by Li et al [78]. A 28 W ultraviolet lamp with wavelength of 254 nm was used as an irradiation source. They concluded that with the increase of Ag/Ti molar ratio from 0 to 1% photocatalytic activity enhanced, and then decreased when the amount of silver increase to 2%. They found that the optimal Ag/Ti molar ratio for the photocatalytic activity of Ag-TiO2 was 1% [78]. Fig. 7 shows the toluene photodegradation efficiency as the function of irradiation time under VIS LEDs irradiation in the presence was Ag-TiO2, Pd-TiO2, Pt-TiO2 and Au-TiO2 microspheres. It can be observed that under visible light photocatalytic properties of metal modified TiO2 depend not only on the kind of noble metal NPs but also on the metal content. Generally, it was observed that a lower amount of metal used during photodeposition resulted in photocatalysts showing highest activity under visible light. The order obtained was Ag-TiO2 Pd-TiO2> Pt-TiO2>> Au-TiO2 under the same condition and at the same loading amount (0.1 wt.%). The highest photoactivity was observed for 0.1Ag_TiO2 sample (Fig. 7). Toluene degradation efficiency reached 100% after 30 min of irradiation. Studies showed a slight decrease in photoactivity in the presence of 0.5Ag_TiO2 sample (98% of toluene was degraded), but in case of 1Ag_TiO2 sample the average degradation efficiency was 3.3 times lower as compared to 0.1Ag_TiO2 sample. As we know from the literature Ag can act as an electron trap and promote the interfacial charge transfer processes in the composite systems, which reduces the recombination of the photoinduced electron–hole pairs, thus improving the photocatalytic activity of TiO2 [79-81]. Additionally, photocatalytic activity of silver modified TiO2 depends also on the uniformity of Ag dispersion on the TiO 2 surface. In case of all the obtained samples, Ag_TiO2 photocatalysts had the highest homogeneous distribution of noble metal NPs on the titanium surface. A significant decrease in photocatalytic performance of TiO2 loaded with Au, Pt and Pd NPs during toluene photodegradation was observed with an increase in metal content on the TiO2 surface. It was observed that the 1Pt_TiO2 revealed more than 2-times lower average degradation efficiency (37% degradation of toluene) compared to 0.1Pt_TiO 2 (90% degradation of toluene) (Fig. 7A). The largest decrease in photoactivity was observed for the 18

sample modified with 1 wt.% of Pd NPs. After 30 min of irradiation only 4% of toluene was degraded, while it amounted to 96% in the presence of the sample modified with 0.1 wt.% of Pd NPs (Fig. 7B). In case of gold modified samples 0.1Au_TiO2 and 0.5Au_TiO2 (0.1 wt% and 0.5 wt% of Au NPs, respectively) photocatalytic activity was similar –50% and 58% of toluene was degraded, respectively. A significant decrease in photoactivity was observed only for 1 Au_TiO2 (12% after 30 min of irradiation). Zhang et al explained that a decreasing effect of photoactivity with high noble metal loading at 1.0 wt% could be due to the excess of NPs which have occupied too many active sites on the TiO2 surface to reduce the overall efficiency and might cause a light shielding effect [77]. This process may overtake the positive charge separation effect. Moreover, it is possible that at loadings beyond the optimum, the NPs deposits can behave as recombination sites due to the accumulation of excessive negative charges to attract positive holes, reducing the efficiency of charge separation [77]. The values of calculated quantum yields for all photocatalysts under UV and VIS LEDs irradiation are presented in Table 1. In our case, the apparent quantum yield obtained for toluene degradation over M-TiO2 samples were in the range of 0.21 % to 0.86 % and 0.12 % to 10.19 % for UV and VIS LEDs, respectively. Little information exists in the literature on quantum yields for the photocatalytic degradation of organic pollutants. Thus, it is very difficult to compare the present results with the other reported studies because the photocatalytic quantum yields appear to vary according to the reaction conditions and the measurement methods. 3.5

Photocatalytic activity in phenol decomposition

The photocatalytic activity of as-prepared samples was also evaluated in the aqueous phase model reaction using phenol as a model pollutant. Phenol was chosen as a model contamination, due to its toxicity towards higher organisms [4]. Moreover Ohtani mentioned that usage of organic dyes as a model compound for photocatalytic reaction is not recommended because dye molecules absorb light, especially in the visible light range, which indicates that a photoreaction might be induced by the light photoabsorption (dye sensitization) as well as by the photocatalyst photoabsorption [82]. Using ICP-MS analysis we confirmed, that only trace amount of noble metal NPs were eluted from the TiO 2 surface to the phenol solution (data not included). The photocatalytic efficiency of TiO2 microspheres 19

modified with Ag, Au, Pt, and Pd nanoparticles is presented in Table 1. The obtained results indicated that both: the type and the amount of noble metal NPs have influence on the photocatalytic activity under UV-vis or visible light. The rate of phenol degradation in the presence of pure TiO2 microspheres under UV−vis light was estimated to be 1.01 μmol dm−3 min−1 (26% after 60-min. irradiation) and was very close to activity reported by Shang et al. [36] for anatase TiO2 hollow spheres prepared via a template-free solvothermal method. They observed that in the presence of hollow spheres, phenol was degraded from 20 to 30% after 60 min. of UV irradiation depending on spheres size [36]. The experimental data clearly indicate the correlation between the amount of metal NPs on the TiO2 surface and photoactivity under UV−vis irradiation – the photocatalytic activity increases with an increase in the amount of metal precursor. Notwithstanding, photocatalytic activity under UV−vis illumination of almost all samples modified with noble metal was suppressed relative to the activity of pristine TiO2 spheres. Only sample loaded with 1.0 wt.% of Ag showed higher photocatalytic efficiency. In the presence of 1 Ag_TiO2 photodegradation efficiency under UV−vis light increased up to 1.16 μmol dm−3 min−1. Compared to pure TiO2 microspheres, TiO2 modified with noble metal NPs exhibited a significant increase in phenol degradation reaction driven by visible light. The efficiency of phenol degradation increased with increasing of metal content up to 0.5 wt% and then decreased. The highest photoactivity under visible light was observed for the sample with 0.5 wt.% of Ag (see details in Table 1). The average rate of phenol decomposition was 0.99 μmol dm−3 min−1 and was 4.5-times higher compared to pure TiO2 (0.22 μmol dm−3 min−1), while for 0.5 Au_TiO2, 0.5 Pt_TiO2 and 0.5 Pd_TiO2 was 0.79; 0.80 and 0.77 μmol dm−3 min−1, respectively. The silver nanoparticles are thought to enhance the TiO2 photoactivity by lowering the rate of recombination of photo-excited charge carriers by acting as electron traps and also by inducing visible light absorption through the surface plasmon resonance effect and subsequent electron transfer to TiO 2, resulting in charge separation [42, 80, 83]. It is in a good agreement with our previous observation. We obtained a series of TiO2 nanosheets with exposed {0 0 1} facets loaded with Pt, Pd, Ag and Au nanoparticles [42]. It was observed that TiO2 nanosheets loaded with Ag NPs exhibit the highest photocatalytic activity under visible light irradiation (λ > 420 nm). Silver-modified TiO2 microspheres synthesized via a novel in-situ Ag-loading method was previously reported by Liu et al [84]. The photocatalytic properties of Ag-TiO2 were 20

estimated for the photodegradation of rhodamine B (Rh B) under visible light irradiation. Agloaded TiO2 photocatalysts showed higher visible light-driven photocatalytic activity than the pure TiO2. Liu et al explained that the highest photoactivity was mainly caused by two factors. Firstly, in-situ loading of silver NPs caused the formation of heterojunctions at the interface and electrons diffused from TiO2 to the interface of Ag–TiO2 heterostructure, and then were transferred to silver species until their Fermi levels aligned. Separated electrons and holes were trapped by reactive oxygen and reductive species, which led to promoting the charge transfer rate and retard the recombination of photogenerated electron–holes pairs of TiO2-based composites. Secondly, TiO2 microspheres exhibited a high photoresponse, absorption and relatively narrow band gap due to the surface plasmon resonance (SPR) effect of the silver nanoparticles [84]. Also Zhao et al investigated photocatalytic properties of the Ag modified hollow SiO2/TiO2 spheres during the degradation of rhodamine B (RhB) [85]. The obtained hollow spheres exhibited high photoactivity for RhB photodegradation under both UV and visible light irradiation. Ag deposition on the SiO 2/TiO2 spheres significantly enhanced the RhB photodegradation under visible light irradiation due of the Schottky barrier formed at the Ag–TiO2 interface, which served as the electron trap, facilitating the separation of electrons and holes [85]. Zheng et al prepared M@TiO2 microspheres (M = Au, Pt, Ag) by the alcohothermal method [48]. Photocatalytic activity under visible light (λ ≥ 400 nm) was evaluated by the oxidation of benzene in aqueous phenol. Among all the noble-metal@TiO2 composites, Au@TiO2 exhibits a high yield (63%) and selectivity (91%) for the catalytic oxidation of benzene to phenol in aqueous phenol under visible light. They explained that the probable mechanism of this reaction is based on the electron transfer from the Au NPs to TiO2 particles, and the electron-depleted Au oxidizes phenoxy anions to form phenoxy radicals that oxidize benzene to phenol [48]. Concluding, it is well known that preparation of photoactive plasmonic photocatalysts requires homogeneous distribution of noble-metal NPs on the TiO2 surface as well as a good contact between the NPs and TiO2, because the properties of noble metal – semiconductor nanocomposites depend strongly on the size and dispersion of noble-metal NPs and on the extent of the metal–semiconductor contact at the interface [56, 86-88]. When noble metal contacts with TiO2, electrons transfer from TiO2 to NPs is observed, which enhances the electron–hole separation and the transfer of the trapped electron to the adsorbed O 2 which 21

acts as an electron acceptor. Consequently, the photocatalytic activity of TiO 2 will be enhanced. In this study, for TiO2 microspheres with lower amount of Ag, Au, Pt and Pd NPs the decomposition rate increases with the increasing amount of metal deposited (up to 0.5 wt.%). However, the loading of too much metal resulted in a decrease of photoactivity. By increasing the amount of noble metal, the chance of the photoelectron meeting the hole also increased. Moreover, electrons could accumulate, thus inducing an electric field to attract holes. The noble metal NPs then became the recombination center of electrons and holes induced by light. 3.6

Formation of hydroxyl radicals

Hydroxyl radicals formed in UV-illuminated TiO2 suspensions are thought to be the main oxidative species which are responsible for the degradation of organic pollutants [89]. Therefore, the formation rate of hydroxyl radicals in solution was used to compare the relative photocatalytic activity of noble-metal loaded TiO2 microspheres. Because its short lifetime (10−9 s) hinders its direct detection, fluorescence of irradiated coumarin solution was used as an indirect method of •OH radical detection. Coumarin reacts with generated hydroxyl radicals forming hydroxycoumarins. Although the major hydroxylation product is 5hydroxycoumarin, only 7-hydroxyproduct of coumarin hydroxylation emits fluorescent light [37, 38, 90]. While titanium based photocatalyst are irradiated with UV light, the valenceband electrons of TiO2 are excited to conduction band, leading to the formation of photogenerated electrons (e−) and holes (h+) with high activity. Then, dissolved oxygen adsorbed on TiO2 surface captures photogenerated electrons to produce highly active superoxide radical-anion, meanwhile •OH radicals are generated by the reaction between photogenerated holes and OH− (or water). Coumarin is a poorly fluorescent molecule that has been known to form the fluorescent 7-hydroxycoumarin (7OHC) by reaction with hydroxyl radicals in aqueous solutions, which fluoresces in the visible region, with a fluorescence maximum around 456 nm (see Scheme 1).

Scheme 1. Formation of 7-hydroxycoumarin in the reaction of coumarin with hydroxyl radicals.

22

Fig. 8 shows the influence of the type and amount of noble metal on typical PL spectral changes observed after 60 min of illumination. Graphs which show the time course of the PL intensity of 7-hydroxycoumarin at 456 nm during the irradiation of the photocatalysts can be found in the Supporting Information. It is clearly seen that the PL intensity at 450 nm increases linearly against the irradiation time. The linear relationship between fluorescence intensity and irradiation time confirms the stability of prepared TiO 2 samples, which leads to a conclusion that the generation of fluorescent 7-hydroxycoumarin is linearly proportional to illumination time, obeying a pseudo-zero order reaction rate equation in kinetics. On the basis of the data presented in Fig. 8, it is can be observed that the PL intensity of photo-generated 7-hydroxycoumarin at 450 nm (excited at 300 nm) depends on both the amount and type of NPs loaded on TiO2 surface. The lowest formation rate of hydroxyl radicals was observed for pristine TiO2 microspheres, while the highest formation rate was noticed for the 1Pd_TiO2 sample. At the same time it could be observed, that in the presence of samples modified with Pt and Pd NPs, the formation rate of •OH increased with the increase in noble metal amount, but in case of Ag and Au TiO2 samples these trend was reversed. The highest formation rate of hydroxyl radicals was observed for 0.1Ag_TiO2 and 0.1Au_TiO2. In case of TiO2 microspheres modified with silver NPs, lower intensity of photo-generated 7hydroxycoumarin can be associated with the highest electron-hole recombination compared to samples modified with Au, Pt and Pd NPs (see Fig. 9A). In this case the rates of electrons and holes recombination are faster than those of the reaction of holes and OH −/H2O, so smaller amount of •OH radicals are observed. The highest formation rate of •OH radicals observed in the presence of 1Pd_TiO2 sample can be related to the lowest electron-hole recombination, because Pd nanoparticles are an effective electron acceptor (see Fig. 9B). The production of •OH radicals can be assigned to the fact that the excited electrons from the valence band to the conduction band can migrate to Pd nanoparticles, while accumulation of holes at the valence band of anatase TiO 2 microspheres leads to the production of surface hydroxyl radical •OH. Photogenerated electrons are effectively accumulated on Pd nanocrystal particles without recombining with holes.

23

3.7

PL emission spectra and the recombination of electron–hole

The photoluminescence (PL) technique is useful to investigate the structure and properties of the active sites on the surface of metal oxides, because of its high sensitivity and nondestructive character. Moreover, the PL emission shows the efficiency of charge carrier trapping, immigration, and transfers, which allows to understand the fate of electron–hole pairs in semiconductor particles and to study electronic structure, optical and photochemical properties of semiconductor materials (surface oxygen vacancies and defects). With electron–hole pair recombination after the irradiation of a photocatalyst, photons are emitted, which results in photoluminescence signal [91, 92]. The enhanced photocatalytic performance of M-TiO2 can be partly attributed to the capability of trapping electrons by Ag, Au, Pd, or Pt, which is verified by the photoluminescence (PL) spectra analysis as shown in Fig. 9. In this study, the PL emission spectra of all samples were examined in the wavelength range of 350–600 nm. It is know from the literature that PL spectra of anatase TiO2 materials are attributed to three kinds of physical origins: self-trapped excitons [93, 94], oxygen vacancies [95] and surface states (defects) [96]. Pure TiO2 showed the highest PL intensity which mean its low separation efficiency. A peak at about 398 nm (3.12 eV) is attributed to the emission of band gap transition; which originates from the recombination of photoexcited electron–hole pair; with the energy of light approximately equal to the band gap energy of anatase (387.5 nm). The violet emission peak (420 nm, ∼2.95 eV) arising from the indirect band edge allowed transitions and self-trapped excitons localized in TiO6 octahedra [93]. The self-trapped exciton is caused by the interaction of conduction band electrons localized on Ti 3d orbital with holes in the O 2p orbital of TiO2 [97]. The emission bands at 420; 440 and 455 nm assigned to shallow-trap state near absorption band edge emission, correspond to the presence of O2- vacancies [98-100]. Oxygen vacancy is a kind of intrinsic defect in nanostructured oxide materials lattice which can easily trap the electrons/holes and create the intermediate energy states in the forbidden gap. These intrinsic defect states act as active centers in luminescence processes [92]. The blue-green emission at 480, 502 and 522 nm, which are equivalent to ∼2.58, 2.47 and 2.34 eV, respectively correspond to the deep-trap states far below the band edge emissions and collectively are called surface state emissions. These charge carriers are generally trapped by oxygen vacancies and surface hydroxyl groups, which contribute in their visible luminescence [101]. The blue-green 24

emission at 480 nm (∼2.58 eV) has been observed for all prepared samples which can be attributed to the charge transfer from Ti3+ to oxygen anion in a TiO68− complex associated with oxygen vacancies at the surface. This phenomena indicate that the band is originating from the intrinsic state rather than the surface state, so the 480 nm band can be assigned to self-trapped excitons localized on TiO6 octahedral [94]. Although the noble metal cannot result in new PL phenomena, it makes the excitonic PL intensity decrease, which is mainly attributed to the capture of noble metal ions. Thus, the photoelectrons and holes can be efficiently separated. Therefore, during the noble metaldoped system, the weaker the excitonic PL spectrum, the higher the separation rate of photo-induced charge carriers. After the deposition of noble metal NPs onto the TiO2 surface quench, the PL intensity depends on the kind of metal deposits, where Pt, Au and Pd loading leads to a maximum reduction in PL intensity in contrast to Ag deposition and bare TiO 2 microspheres (see Fig. 9). Along with the increase in the metal content, the PL intensity dropped down and achieved its lowest level at 1 wt% noble metal NPs content, which meant that the surface modification could suppress the recombination process of photo-generated carriers in TiO2. It is known that Ag and Au NPs absorb light in the visible spectrum because of localized surface plasmon resonance (LSPRs). When the noble metal is illuminated by the spectrum, the charge density is redistributed and thus establishes a strong Coulombic restoring force, which then resultes in oscillation of charge density like a harmonic oscillator in phase with the incident light [87, 102, 103]. The decrease in the intensity of PL spectra for Ag–TiO2 and Au–TiO2 might indicate that the photocatalysts modified with noble metals had a lower recombination rate of electrons and holes under light irradiation. Yu et al explained that in the first step electrons were excited from the VB to the CB and then migrated to Ag and Au clusters, which prevented the direct recombination of electrons and holes. Secondly, Ag– TiO2 and Au–TiO2 might act as rapid separation sites for the photogenerated electrons and holes due to the difference in the energy levels of their CB and VB [104]. Fig. 9A shows the studied PL spectrum of silver modified microspheres. It can be found that the PL intensity of TiO2 samples is weakened after the deposition of Ag NPs onto the surface of TiO2. PL emission spectra of Au-TiO2 microspheres are presented in Fig. 9B. There was a significant decrease in the intensity of photoluminescence spectra for gold modified samples. This could be attributed to effective shuttling of photogenerated charge carriers 25

from TiO2 surface to deposited noble metal NPs that prevent the recombination and hence quench the PL emission. Fig. 9C shows the photoluminescence properties of Pt-TiO2 microspheres. It was noticed that PL intensity is almost the same for samples loaded with 0.5 and 1 %wt. of platinum NPs, suggesting that the excessive Pt might contribute as new recombination centers and decreased the separation efficiency. A similar phenomenon was observed by Hu et al. [100]. They obtained platinum doped TiO2 with Pt/Ti molar ratio from 0.2 to 1.0 % and observed that with an increase of Pt content, the PL intensity dropped down and achieved its lowest level at 0.4% Pt content and increased when the Pt content continued to increase [100]. It is evident from the PL spectra that samples with palladium loaded on the TiO2 surface showed the highest decrease in the PL intensity (Fig. 9D). It is know from the literature, that palladium plays an important role in the interfacial charge transfer and in a decrease in the rate of electron-hole recombination, because Pd NPs act as an effective electron scavenger to trap the photo induced electrons and holes of TiO 2 leading to the reduction of electron–hole recombination [88]. Leong et al synthesized a series of PdTiO2 photocatalysts (0.5; 1.0 and 3.0 wt %) and noticed the lowest emission peaks for 1.0 wt% Pd/TiO2 [88]. 3.8 The effect of the type and amount of noble metal nanoparticles on photoactivity of M-TiO2 To correlate the effect of metal type and amount on the activity of M-TiO2, the efficiency of model compounds degradation in the gas and aqueous phases was related to the amount of noble metals precursor used during composited synthesis, as presented in Fig. 10. It could be clearly stated that for the UV-mediated gas phase reaction (toluene degradation), among all tested noble metals, palladium seems to be the best element to modify titanium dioxide. Application from 0.1 to 1.0 wt.% of palladium precursor during the preparation route resulted in the formation of highly active Pd-TiO2 composite. About 95% of toluene was removed from the air after only 15 min. of exposure despite the use of such low-powered irradiation source as LEDs (see Fig. 6b). In the case of visible light mediated gas phase reaction, the application of palladium, platinum or silver in a small amount (0.1 wt.%) resulted in similar effectiveness of toluene removal from the air. In this case, gold nanoparticles proved the less promising agent for TiO2 modification. As it was mentioned before, a generally lower amount of noble metal used during photodeposition resulted in photocatalysts showing highest activity under visible light. The order obtained was Ag-TiO2 26

Pd-TiO2> Pt-TiO2>> Au-TiO2 under the same condition and at the same loading amount (0.1 wt.%). Quite a different situation was observed for the phenol degradation in the aqueous phase. Generally it could be observed that in the UV-mediated reaction the highest amount of metal (1 wt.%) was profitable, since in the Vis-induced reaction the medium amount of metal (0.5 wt.%) was beneficial. Although phenol was degraded the most efficiently in the presence of higher amount of metal nanoparticles seating at the surface of TiO 2 surface, the efficiency of phenol removal is not always proportional to the amount of •OH radical generated under UV light. In the case of platinum and palladium nanoparticles, the highest amount of metal precursors used for M-TiO2 preparation resulted both in high efficiency of phenol degradation and •OH radical formation. However, in the case of Au and Ag nanoparticles the same amount of metal precursor (1 wt.%) used during synthesis, caused highly efficient removal of phenol but only moderate efficiency for •OH radical formation. Thus, in the presence of TiO2 spheres loaded with Au and Ag nanoparticles, phenol is degraded not only via oxidation by •OH radicals but probably also in direct reaction with the photogenerated carriers (e-/h+). On the other hand it could be observed that higher amount of noble metal nanoparticles efficiently suppressed the charge carriers’ recombination process. The role of noble metal seating at the surface of TiO2 was precisely described by Bumajdad and Madkour [105]. Under UV irradiation, the dominant effect is charge separation mechanism caused by the transfer of electrons from the conduction band of TiO2 to the noble metal nanoparticles. The mechanism based on the interband transitions in the noble metal nanoparticles from fully occupied d-bands below the Fermi energy to the half filled sp band was also proposed. Noble metal NPs show UV light absorption due to the interband transition of 5 d electrons to the 6 sp and, 4 d electrons to the 5 sp band and 5 d electrons to the 6sp band for Au, Ag and Pt NPs, respectively [106-108]. Furthermore, a few different mechanisms were proposed to explain visible light driven photoactivity of TiO2 loaded with noble metals NPs, such as: (1) charge transfer mechanism (excited electrons from NPs are transferred to the conduction band of TiO2), (2) local electric field enhancement at the metal-semiconductor interface, (3) induction of electromagnetic field resulting in plasmon resonance energy transfer (PRET), and (4) efficient scattering

27

mediated by LSPR leading to a longer optical path length for photons in TiO2 and finally raising the excitation of e- -h+ pairs [105]. For a better understanding of the effect of metal particles size on the photoactivity, based on TEM images the size distribution of silver for different loading of AgNO 3 precursor was investigated and presented in Fig. 11. As a lower amount of silver precursor was used (0.1 wt.%), the main fraction of Ag NPs are in the range from 10 to 20 nm (65%). However, the TEM analysis revealed also the presence of silver nanoparticles in the range of 20-30 nm (5%), 30-40 nm (15%) and 40-50 nm (15%). A higher amount of AgNO3 used during photodeposition (0.5 wt.%) resulted in the formation of mainly larger Ag nanoparticles (2030 nm: 60% and 30-40 nm: 30%). The remaining 10% of Ag NPs had the size in the range of 10-20 and 40-50 nm. The highest amount of silver ions used during the preparation step (1 wt.%) contributed to the development of the smallest silver nanoparticles. Larger silver clusters appeared with the size of 10-20 nm (65%), and only 35% of the clusters were below 10 nm. Thus, it could be stated that the 1Ag_TiO2 sample, showing the highest activity under UV light (both in gas and aqueous phase model reactions) contains very small Ag nanoparticles and small Ag NPs are responsible for the enhancement of UV induced photocatalytic activity and those types of metal nanoparticles are probably responsible for suppressing of the charge recombination effect, which is in good agreement with photoluminescence measurements (see Fig. 9A). Visible light induced photoactivity could be attributed to the presence of larger nanoparticles of silver. Samples showing highest photoactivity under visible light (0.5Ag_TiO2 and 0.1Ag_TiO2) contain also nanoparticles in the range of 20-30, 30-40 and 40-50 nm. Thus, an increase in the photoactivity of Ag-TiO2 could be caused by an increase in the adsorbed photon flux by larger Ag nanoparticles followed by the transfer of excited electrons from NPs to the conduction band of TiO2. Concluding, it could be stated that the dosage and the size of noble metal nanoparticles are the main factors controlling the photoactivity of modified TiO2. As a general rule, upon reducing the metal dosage and reducing the metal nanoparticles’ size, the photocatalytic activity is enhanced. It was proposed that the Fermi energy of the noble metal nanoparticles grow with the decrease in the particle size due to the quantum size effect [109]. Thus, noble metal NPs with a pronounced size could posses its energy level within the conduction band

28

of TiO2. On the other hand, lower photoactivity for samples loaded with higher dosage of metal nanoparticles could be attributed to the so called “screening effect”. 4.

Conclusions

Summing up, we reported for the first time the preparation and characterization of TiO2 microspheres loaded with noble metal NPs (Au, Ag, Pt and Pd) using the photodeposition method. The influence of different amounts of metal precursors on the photoactivity of toluene degradation in the gas phase, activated by light-emitting diodes (LEDs) (λmax = 375 nm or λmax = 415 nm) and phenol degradation in the aqueous

phase

source

were

investigated.

The

as-obtained

photocatalysts

microspheres had diameters in the range of 1-18 μm. Typical diffraction peaks corresponding to anatase were observed in all the samples. Owing to their unique structural features, TiO2 microspheres loaded with noble metal NPs exhibited excellent photocatalytic activity in toluene degradation under UV and visible light irradiation, compared to pure TiO2. Generally, samples decorated with smaller amount and probably smaller in size metal NPs possess higher photoactivity both under UV and visible light. The highest activity under visible light irradiation was observed for sample loaded with 0.1 wt.% of Ag. Toluene degradation efficiency reached 100% after 30 min of irradiation. At the same time it was noticed that for the phenol degradation process under UV light, the highest amount of noble metal NPs (1.0 wt.%) was profitable and in the Vis-induced reaction the medium amount of metal (0.5 wt.%) was beneficial. Thus, porous TiO2 microspheres decorated with fine noble nanoparticles seem to be highly active photocatalyst to remove toluene from the gas phase and phenol from the aqueous phase. Moreover, it could be stated that the dosage and the size of noble metal nanoparticles are the main factors controlling photoactivity of modified TiO2. Acknowledgements The research leading to these results has received funding from the National Centre for Research and Development (PHOTOAIR, Pol-Nor/207686/18/2013). Also the

29

author EG is grateful for the financial support of the Ministry of Science and Higher Education (contract No.: UMO-0132/IP2/2015/73). References [1] M. Nischk, P. Mazierski, M. Gazda, A. Zaleska, Appl. Catal. B: Environ. 144 (2014) 674-685. [2] A. Zaleska, A. Hanel, M. Nischk, Recent Pat. Eng. 4 (2010) 200-216. [3] J. Mo, Y. Zhang, Q. Xu, J.J. Lamson, R. Zhao, Atmos. Environ. 43 (2009) 2229-2246. [4] E. Grabowska, J. Reszczyoska, A. Zaleska, Water Res. 46 (2012) 5453-5471. [5] N.R. Neti, G.R. Parmar, S. Bakardjieva, J. Subrt, Chem. Eng. J. 163 (2010) 219-229. [6] J. Taranto, D. Frochot, P. Pichat, Separ. Purif. Technol. 67 (2009) 187-193. [7] A.C. Affam, M. Chaudhuri, J. Environ. Manage. 130 (2013) 160-165. [8] C. Zhao, M. Pelaez, D.D. Dionysiou, S.C. Pillai, J.A. Byrne, K.E. O'Shea, Catal. Today 224 (2014) 70-76. [9] R. Su, R. Tiruvalam, Q. He, N. Dimitratos, L. Kesavan, C. Hammond, J.A. Lopez-Sanchez, R. Bechstein, C.J. Kiely, G.J. Hutchings, ACS Nano 6 (2012) 6284-6292. [10] A. Cybula, G. Nowaczyk, M. Jarek, A. Zaleska, J. Nanomater. 2014 (2014). [11] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69-96. [12] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001) 269-271. [13] E. Grabowska, J.W. Sobczak, M. Gazda, A. Zaleska, Appl. Catal. B: Environ. 117 (2012) 351-359. [14] H. Yamashita, M. Harada, J. Misaka, M. Takeuchi, Y. Ichihashi, F. Goto, M. Ishida, T. Sasaki, M. Anpo, J. Synchrotron Radiat. 8 (2001) 569-571. [15] I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihara, K. Takeuchi, J. Mol. Catal. A: Chem. 161 (2000) 205-212. [16] H. Zhang, M. Guashuai, M. Xingping, W. Bei, Int. J. Photoenergy 2013 (2013). [17] A. Zaleska, Recent Pat. Eng. 2 (2008) 157-164. [18] P.K. Jain, M.A. El-Sayed, J. Phys. Chem. C 111 (2007) 17451-17454. [19] F. Chen, R.L. Johnston, Plasmonics 4 (2009) 147-152. [20] S. Eustis, M.A. El-Sayed, Chem. Society Rev. 35 (2006) 209-217. [21] T. Pradeep, Thin solid films, 517 (2009) 6441-6478. [22] S. Xie, S.-I. Choi, X. Xia, Y. Xia, Curr. Opin. Chem. Eng. 2 (2013) 142-150. [23] C.-K. Lee, S.-S. Liu, L.-C. Juang, C.-C. Wang, M.-D. Lyu, S.-H. Hung, J. Hazard. Mater. 148 (2007) 756-760. [24] D.J. Yang, Z.F. Zheng, H.Y. Zhu, H.W. Liu, X.P. Gao, Adv. Mater. 20 (2008) 2777-2781. [25] F. Chen, Z. Liu, Y. Liu, P. Fang, Y. Dai, Chem. Eng. J. 221 (2013) 283-291. [26] Y. Zheng, J. Cai, K. Lv, J. Sun, H. Ye, M. Li, Appl. Catal. B: Environ. 147 (2014) 789-795. [27] C. Wang, H. Liu, Y. Liu, G.a. He, C. Jiang, Appl. Surf. Sci. (2014). [28] R. Wang, X. Cai, F. Shen, Ceram. Int. 39 (2013) 9465-9470. [29] Z. Zheng, B. Huang, X. Qin, X. Zhang, Y. Dai, Chem.-Eur. J 16 (2010) 11266-11270. [30] H. Shirkhanloo, M. Osanloo, O.Q. Dadras, Inter. J. Occup. Hyg. 6 (2014) 1-5. [31] S. Futamura, A. Gurusamy, J. Electrostat. 63 (2005) 949-954. [32] T. Zhu, J. Li, W. Liang, Y. Jin, J. Hazard. Mater. 165 (2009) 1258-1260. [33] T. Zhu, J. Li, Y. Jin, Y. Liang, G. Ma, Inter. J. Environ. Sci. Technol. 6 (2009) 141-148. [34] F.-L. Cao, J.-G. Wang, F.-J. Lv, D.-Q. Zhang, Y.-N. Huo, G.-S. Li, H.-X. Li, J. Zhu, Catal. Commun. 12 (2011) 946-950. 30

[35] L. Yang, Z. Si, D. Weng, Y. Yao, Appl. Surf. Sci. 313 (2014) 470-478. [36] S. Shang, X. Jiao, D. Chen, ACS Appl. Mater. Interfaces, 4 (2012) 860-865. [37] G. Louit, S. Foley, J. Cabillic, H. Coffigny, F. Taran, A. Valleix, J.P. Renault, S. Pin, Radiat. Phys. Chem. 72 (2005) 119-124. [38] P. Górska, A. Zaleska, J. Hupka, Separ. Purif. Technol. 68 (2009) 90-96. [39] Z. Zheng, B. Huang, X. Qin, X. Zhang, Y. Dai, J. Colloid Interface Sci. 358 (2011) 68-72. [40] Y. Wang, Y. He, Q. Lai, M. Fan, J. Environ. Sci. 26 (2014) 2139-2177. [41] J. Belloni, Catal. Today, 113 (2006) 141-156. [42] M. Diak, E. Grabowska, A. Zaleska, Appl. Surf. Sci. 347 (2015) 275-285. [43] V. Iliev, D. Tomova, R. Todorovska, D. Oliver, L. Petrov, D. Todorovsky, M. UzunovaBujnova, Appl. Catal. A: Gen. 313 (2006) 115-121. [44] J. Rodríguez-Carvajal, Phys. B: Cond. Mat. 192 (1993) 55-69. [45] N. Serpone, A. Salinaro, Pure Appl. Chem. 71 (1999) 303-320. [46] G. Tang, D. Zhang, L. Zhao, M. Zhang, H. Tang, H. Huang, C. Li, Mater. Lett. 118 (2014) 192-195. [47] P. Wei, J. Liu, Z. Li, Ceram. Int. 39 (2013) 5387-5391. [48] Z. Zheng, B. Huang, X. Qin, X. Zhang, Y. Dai, M.-H. Whangbo, J. Mat. Chem. 21 (2011) 9079-9087. [49] R.P. Galhenage, H. Yan, S.A. Tenney, N. Park, G. Henkelman, P. Albrecht, D.R. Mullins, D.A. Chen, J. Phys. Chem. C 117 (2013) 7191-7201. [50] M. Chen, D.W. Goodman, Chem. Soc. Rev. 37 (2008) 1860-1870. [51] H. Iddir, S. Öğüt, N.D. Browning, M.M. Disko, Phys. Rev. B 72 (2005) 081407. [52] M. Harada, Y. Inada, Langmuir, 25 (2009) 6049-6061. [53] M. Harada, E. Katagiri, Langmuir, 26 (2010) 17896-17905. [54] M. Harada, Y. Kamigaito, Langmuir, 28 (2012) 2415-2428. [55] U. Kreibig, M. Vollmer, Springer Ser. Mat. Sci 25 (1995). [56] R. Sellappan, M.G. Nielsen, F. González-Posada, P.C. Vesborg, I. Chorkendorff, D. Chakarov, J. Catal. 307 (2013) 214-221. [57] X. Fan, J. Fan, X. Hu, E. Liu, L. Kang, C. Tang, Y. Ma, H. Wu, Y. Li, Ceram. Int. 40 (2014) 15907–15917. [58] P.V. Kamat, J. Phys. Chem. C 111 (2007) 2834-2860. [59] C.A. Emilio, M.I. Litter, M. Kunst, M. Bouchard, C. Colbeau-Justin, Langmuir, 22 (2006) 3606-3613. [60] E. Grabowska, A. Zaleska, S. Sorgues, M. Kunst, A. Etcheberry, C. Colbeau-Justin, H. Remita, J. Phys. Chem. C 117 (2013) 1955-1962. [61] S.K. Dondapati, T.K. Sau, C. Hrelescu, T.A. Klar, F.D. Stefani, J. Feldmann, Acs Nano, 4 (2010) 6318-6322. [62] N.R. Jana, L. Gearheart, C.J. Murphy, J. Phys. Chem. B 105 (2001) 4065-4067. [63] F. Kim, J.H. Song, P. Yang, J. Am. Chem. Soc. 124 (2002) 14316-14317. [64] F. Kim, S. Connor, H. Song, T. Kuykendall, P. Yang, Angew. Chem. 116 (2004) 3759-3763. [65] S.K. Ghosh, S. Nath, S. Kundu, K. Esumi, T. Pal, J. Phys. Chem. B 108 (2004) 13963-13971. [66] P.S. Kumar, I. Pastoriza-Santos, B. Rodriguez-Gonzalez, F.J.G. de Abajo, L.M. Liz-Marzan, Nanotechnology, 19 (2008) 015606. [67] E. Kowalska, O.O.P. Mahaney, R. Abe, B. Ohtani, Phys. Chem. Chem. Phys. 12 (2010) 2344-2355. [68] H.-W. Chen, Y. Ku, Y.-L. Kuo, Water Res. 41 (2007) 2069-2078. [69] T. Pakizeh, C. Langhammer, I. Zoric, P. Apell, M. K ll, Nano ett. 9 (2009) 882-886. 31

[70] K. Kočí, . Matějová, M. Reli, . Čapek, V. Matějka, Z. acný, P. Kustrowski, . Obalová, Catal. Today 230 (2014) 20-26. [71] J.B. Zhong, Y. Lu, W.D. Jiang, Q.M. Meng, X.Y. He, J.Z. Li, Y.Q. Chen, J. Hazard. Mater. 168 (2009) 1632-1635. [72] S.L. Lee, J. Scott, K. Chiang, R. Amal, J. Nanopart. Res. 11 (2009) 209-219. [73] H.-H. Ou, S.-L. Lo, J. Mol. Catal. A-Chem. 275 (2007) 200-205. [74] T. Sano, S. Kutsuna, N. Negishi, K. Takeuchi, J. Mol. Catal. A-Chem. 189 (2002) 263-270. [75] J. Kim, P. Zhang, J. Li, J. Wang, P. Fu, Chem. Eng. J. 252 (2014) 337-345. [76] R.M. Alberici, W.F. Jardim, Appl. Catal. B: Environ. 14 (1997) 55-68. [77] L. Zhang, M.S. Tse, O.K. Tan, J. Environ. Chem. Eng. 2 (2014) 1214-1220. [78] X. Li, L. Wang, X. Lu, J. Hazard. Mater. 177 (2010) 639-647. [79] B. Xin, L. Jing, Z. Ren, B. Wang, H. Fu, J. Phys. Chem. B 109 (2005) 2805-2809. [80] H.M. Sung-Suh, J.R. Choi, H.J. Hah, S.M. Koo, Y.C. Bae, J J. Photochem. Photobio. A 163 (2004) 37-44. [81] X. Li, F. Li, Environ. Sci. Technol. 35 (2001) 2381-2387. [82] B. Ohtani, Chem. Lett. 37 (2008) 216-229. [83] M.K. Seery, R. George, P. Floris, S.C. Pillai, J. Photochem. Photobio. A 189 (2007) 258263. [84] H. Liu, X. Dong, T. Liu, Z. Zhu, Sol. Energ. Mat. Sol. C 132 (2015) 86-93. [85] W. Zhao, L. Feng, R. Yang, J. Zheng, X. Li, Appl. Catal. B: Environ. 103 (2011) 181-189. [86] S.T. Kochuveedu, D.-P. Kim, D.H. Kim, The J. Phys.l Chem. C 116 (2012) 2500-2506. [87] X. Zhou, G. Liu, J. Yu, W. Fan, J. Mat. Chem. 22 (2012) 21337-21354. [88] K.H. Leong, H.Y. Chu, S. Ibrahim, P. Saravanan, Beilstein J. Nanotechnol. 6 (2015) 428437. [89] K. Lv, J. Yu, K. Deng, X. Li, M. Li, J. Phys. Chem. Solids 71 (2010) 519-522. [90] H. Czili, A. Horváth, Appl. Catal. B: Environ. 81 (2008) 295-302. [91] J. Liqiang, Q. Yichun, W. Baiqi, L. Shudan, J. Baojiang, Y. Libin, F. Wei, F. Honggang, S. Jiazhong, Sol. Energ. Mat. Sol. C 90 (2006) 1773-1787. [92] L. Kernazhitsky, V. Shymanovska, T. Gavrilko, V. Naumov, L. Fedorenko, V. Kshnyakin, J. Baran, J. Lumin. 146 (2014) 199-204. [93] H. Tang, H. Berger, P.E. Schmid, F. Lévy, G. Burri, Solid State Commun. 87 (1993) 847850. [94] L. Saraf, S. Patil, S. Ogale, S. Sainkar, S. Kshirsager, Int. J. Mod. Phys. B 12 (1998) 26352647. [95] N. Serpone, D. Lawless, R. Khairutdinov, J. Phys. Chem. 99 (1995) 16646-16654. [96] L. Forss, M. Schubnell, Appl. Phys. B 56 (1993) 363-366. [97] S.T. Nishanthi, S. Iyyapushpam, B. Sundarakannan, E. Subramanian, D. Pathinettam Padiyan, Separ. Purif. Technol. 131 (2014) 102-107. [98] J. Kaur, R. Singh, B. Pal, J. Mol. Catal. A-Chem. 397 (2015) 99-105. [99] Q. Xiang, K. Lv, J. Yu, Appl. Catal. B: Environ. 96 (2010) 557-564. [100] Y. Hu, X. Song, S. Jiang, C. Wei, Chem. Eng. J. 274 (2015) 102-112. [101] S. Mathew, A. kumar Prasad, T. Benoy, P. Rakesh, M. Hari, T. Libish, P. Radhakrishnan, V. Nampoori, C. Vallabhan, J. Fluoresc. 22 (2012) 1563-1569. [102] S.T. Kochuveedu, Y.H. Jang, D.H. Kim, Chem. Soc. Rev. 42 (2013) 8467-8493. [103] X. Zhang, Y.L. Chen, R.-S. Liu, D.P. Tsai, Rep. Prog. Phys. 76 (2013) 046401. [104] J. Yu, J. Xiong, B. Cheng, S. Liu, Appl. Catal. B: Environ. 60 (2005) 211-221. [105] A. Bumajdad, M. Madkour, Phys. Chem. Chem. Phys. 16 (2014) 7146-7158. 32

[106] H. Zhu, X. Chen, Z. Zheng, X. Ke, E. Jaatinen, J. Zhao, C. Guo, T. Xie, D. Wang, Chem. Commun. (2009) 7524-7526. [107] X. Chen, Z. Zheng, X. Ke, E. Jaatinen, T. Xie, D. Wang, C. Guo, J. Zhao, H. Zhu, Green Chem. 12 (2010) 414-419. [108] M. Arce, P. Quaino, E. Santos, Catal. Today 202 (2013) 120-127. [109] B. Cojocaru, Ş. Neaţu, E. Sacaliuc-Pârvulescu, F. Levy, V.I. Pârvulescu, H. Garcia, Appl. Catal. B: Environ. 107 (2011) 140-149.

33

Fig. 1 LeBail refinement of room temperature powder x-ray diffraction data for: (A) 0.5Pd_TiO2, (B) 0.5Ag_TiO2 (C) 0.5Pt_TiO2 and (D) 0.5Au_TiO2. Upper part – crosses, observed data, solid red lines calculated intensities. The black tick marks correspond to TiO2 (anatase), whereas the blue tick marks to the noble metal. The lower part shows the differences between the observed and calculated pattern

34

C Particles number [%]

50

TiO2 microspheres

40 30

20 10 0 1-3

3-6

6-9

9-12

12-15

15-18

Particles diameter [µm] Fig. 2 SEM microscopy images of: (A) pure TiO2 microspheres and (B) microspheres modified with Ag nanoparticles (0.5Ag_TiO2 sample), and (C) particle size distribution for pure TiO2 microspheres (average value calculated based on measurements of 50 particles)

35

Fig. 3 SEM microscopy images of TiO2 microspheres modified with 0.5 wt.% of noble metal NPs: (A) Ag, (B) Au, (C) Pd and (D) Pt. The lighter dots (indicated by black arrow marks) corresponds to the presence of noble metal nanoparticles

36

Fig. 4 TEM-EDX analysis of (A) 0.5Ag_TiO2, (B) 0.5Au_TiO2, (C) 0.5Pd_TiO2 and (D) 0.5Pt_TiO2 microspheres

37

Absorbance (a.u.)

A

1Ag_TiO2 0.5Ag_TiO2 0.1Ag_TiO2 P25

300

TiO2_sph 400

500

600

700

Wavelength (nm)

Absorbance (a.u.)

B 1Au_TiO2 0.5Au_TiO2 0.1Au_TiO2

P25

300

TiO2_sph 400

500

600

700

Wavelength (nm)

38

Absorbance (a.u.)

C 1Pt_TiO2 0.5Pt_TiO2 0.1Pt_TiO2

P25 TiO2_sph

300

400

500

600

700

Wavelength (nm)

Absorbance (a.u.)

D 1Pd_TiO2 0.5Pd_TiO2 0.1Pd_TiO2

P25 TiO2_sph

300

400

500

600

700

Wavelength (nm) Fig. 5 UV–vis absorption spectra of: (A) Ag, (B) Au, (C) Pt and (D) Pd loaded TiO2 microspheres

39

100

TiO2 sph 0.1 Pt_TiO2

80

0.5 Pt_TiO2 1 Pt_TiO2

60

40

20

Toluene concentration, C/C0 (%)

Toluene concentration, C/C0 (%)

100

A

0

0

TiO2 sph

0.1 Pd_TiO2 80

0.5 Pd_TiO2 1 Pd_TiO2

60

40

20

B

0

10

20

30

0

10

Irradiation time (min) 100

TiO2 sph

0.1 Au_TiO2

80

0.5 Au_TiO2 1 Au_TiO2

60

40

20

Toluene concentration, C/C0 (%)

Toluene concentration, C/C0 (%)

100

C

0

0

0.1 Ag_TiO2 80

0.5 Ag_TiO2 1 Ag_TiO2

60

40

20

D

20

30

0

10

Irradiation time (min) 100

0.1 Pt_TiO2 0.5 Pt_TiO2

80

1 Pt_TiO2

60

40

20

E

0

0

10

20

30

0.1 Pd_TiO2 0.5 Pd_TiO2

80

1 Pd_TiO2

60

40

20

F

0

Irradiation time (min)

20

Irradiation time (min)

Toluene concentration, C/C0 (%)

Toluene concentration, C/C0 (%)

100

30

TiO2 sph

0

10

20

Irradiation time (min)

30

0

10

20

30

Irradiation time (min)

Fig.6 Toluene photodegradation efficiency as the function of irradiation time under UV LEDs irradiation in the presence of metal modified TiO2 microspheres loaded with: (A) Pt, (B) Pd, (C) Au and (D) Ag NPs (C0 = 200 ppm) and (E) Pt and (Pd) NPs (C0 = 400 ppm)

40

TiO2 sph 0.1 Pt_TiO2 0.5 Pt_TiO2 1 Pt_TiO2

80

60

40

20

100

Toluene concentration, C/C0 (%)

Toluene concentration, C/C0 (%)

100

A

0

0

80

60 TiO2 sph 0.1 Pd_TiO2 0.5 Pd_TiO2 1 Pd_TiO2

40

20

B

0

10

20

0

30

10

100

Toluene concentration, C/C0 (%)

Toluene concentration, C/C0 (%)

100

80

60

40 TiO2 sph 0.1 Au_TiO2 0.5 Au_TiO2 1 Au_TiO2

20

C

0

0

10

20

30

TiO2 sph 0.1 Ag_TiO2 0.5 Ag_TiO2 1 Ag_TiO2

80

60

40

20

D

0

Irradiation time (min)

20

Irradiation time (min)

Irradiation time (min)

30

0

10

20

30

Irradiation time (min)

Fig.7 Toluene photodegradation efficiency (C0 = 200 ppm) as the function of irradiation time under VIS LEDs irradiation in the presence of metal modified TiO2 microspheres loaded with: (A) Pt, (B) Pd, (C) Au and (D) Ag NPs

41

42

-3

Fig. 8 Fluorescence spectra of coumarin solution (10 M) after 60 min light illumination (A) Ag, (B) Au, (C) Pt and (D) Pd loaded TiO2 microspheres

43

440 nm 420 nm 455 nm 398 nm 480 nm 502 nm 522 nm

44

Fig. 9 Photoluminescence spectra of (A) Ag, (B) Au, (C) Pt and (D) Pd loaded TiO2 microspheres

45

A

100

B Efficiency of phenol degradation in the aqueous phase, Co/C (%)

Efficiency of toluene degradation in the gas phase, Co/C (%)

90 80 70 60 50 40 30

20 10 0 0

0,25

0,5

0,75

Amount of noble metal precursor (wt. %)

40

35

30

Ag-TiO2 (Vis) Au-TiO2 (Vis) Pt-TiO2 (Vis) Pd-TiO2 (Vis) Ag-TiO2 (UV) 200ppm Au-TiO2 (UV) 200ppm Pt-TiO2 (UV) 200ppm Pd-TiO2 (UV) 200ppm Pt-TiO2 (UV) 400ppm Pd-TiO2 (UV) 400ppm

25

20

15

10

5

0 10

0,25

0,5

0,75

Amount of noble metal precursor (wt. %)

Fig. 10 The effect of the amount of noble metal precursor on the photoactivity of M-TiO2 composites (M = Ag, Au, Pt and Pd): (a) toward toluene degradation in the gas phase after 30 min irradiation using UV or Vis light, and (b) toward phenol degradation in the aqueous phase after 60 min irradiation using UV or Vis light

Fig. 11. The effect of silver precursor amount used during photodeposition on the size distribution of Ag nanoparticles seated at the surface of TiO2 spheres (based on TEM analysis)

46

1

Table 1 Sample label and preparation method of bare and metal modified TiO2 microspheres

Type and amount of noble metal precursor used to TiO2 modification

label Type of precursor

TiO2

Sample color

Phenol degradation rate (µmol/dm3/min)

Quantum y toluene deg

(%)

LSPR*

Amount of precursor [wt.%]

UV-Vis

Vis (λ > 420 nm)

UV

0.1

white

430

1.08

0.58

0.57

0.5

pale beige

460

0.62

0.99

0.21

iO2

1

beige

400

1.16

0.37

0.86

TiO2

0.1

grey

545

0.54

0.66

0.51

0.5

violet

540

0.57

0.79

0.37

iO2

1

deep grey

560

0.64

0.12

0.53

TiO2

0.1

pale grey

-

0.89

0.58

0.72

0.5

grey

-

0.69

0.80

0.50

1

grey

-

1.04

0.41

0.50

TiO2

TiO2

TiO2

AgNO3

KAuCl4

H2PtCl6

iO2

TiO2

PdCl2

0.1

pale grey

-

0.44

0.55

0.59

TiO2

0.5

grey

-

0.81

0.77

0.42

iO2

(5 wt% solution in 10 wt% HCl)

1

deep grey

-

0.93

0.17

0.31

sph

-

-

white

-

1.01

0.22

0.17

* the main absorption peak (nm)

47