Pd Subnano-Clusters on TiO2 for Solar-Light Removal of NO - ACS

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article 2

Pd subnano-clusters on TiO for solar-light removal of NO Kakeru Fujiwara, U. Müller, and Sotiris E Pratsinis ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b02685 • Publication Date (Web): 08 Feb 2016 Downloaded from http://pubs.acs.org on February 9, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Pd subnano-clusters on TiO2 for solar-light removal of NO by Kakeru Fujiwara,1 Ulrich Müller2 & Sotiris E. Pratsinis1* 1

Particle Technology Laboratory, Institute of Process Engineering, Department of Mechanical and Process Engineering, ETH Zurich, Sonneggstrasse 3, CH-8092 Zurich, Switzerland 2

Empa, Swiss Federal Laboratories for Materials Science and Technology, Laboratory for Nanoscale Materials Science, Überlandstrasse 129, CH-8600 Dübendorf, Switzerland

Submitted to ACS Catalysis November, 2015 and revised February 2016

*Corresponding author: [email protected]

1

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

Abstract Palladium subnano-clusters (< 1 nm) on TiO2 nanoparticles are prepared in one step by flame aerosol technology. Under solar light irradiation, these materials remove NOx 3 or 7 times faster than commercial TiO2 (P25, Evonik) with or without photodeposited Pd on it. X-ray photoelectron spectroscopy (XPS) reveals that such photodeposited Pd consists of metallic Pd along with several Pd oxidation states. In contrast, flame-made Pd subnano-clusters on TiO2 dominantly consist of an intermediate Pd oxidation state between metallic Pd and PdO. In that intermediate state, the Pd subnano-clusters are stable up to, at least, 600 oC for 2 hours in air. However, a fraction of them is reduced into relatively large (> 1 nm) metallic Pd nanoparticles by annealing in N2 at 400 oC for 2 hours, as elucidated by XPS and scanning transmission electron microscopy. The Pd subnano-clusters interact with oxygen defects on the TiO2 surface as shown by Raman spectroscopy. This interaction suppresses CO adsorption on Pd as observed by diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), analogous to strong metal support interactions (SMSI) of nano-sized noble metals on TiO2.

Keywords: Photocatalyst, NOx removal, Pd cluster, TiO2, Flame synthesis, Solar photocatalysis,

2


Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Introduction Typical nitrogen oxides, NOx (NO and NO2) removal processes from combustion off-gases such as the three-way catalyst1 or selective catalytic reduction2 require high temperature (~ 300 oC) and reducing agents (e.g. NH3) making difficult the emission of NOx at sub-ppm level at ambient temperature. So solar photocatalytic NOx removal that requires only sunlight is quite attractive.3 Titanium oxide is the most widely used photocatalyst for air cleaning4 including environmental NOx removal.5 Under light excitation, TiO2 converts NOx into nitrate3 but some NO is converted also into unfavorable NO2. So high conversion and selectivity for nitrate formation are essential. The activity under sunlight can be enhanced by narrowing the TiO2 band gap by S/N-doping6 and plasma-treatment.7 For the selectivity, TiO2 composition and preparation are crucial: Pure anatase TiO2 and flame-made anatase-rutile TiO2 composites exhibit high selectivity.8 Furthermore, selectivity can be improved by combining NO2 or nitrate storage compounds such as active carbon3 or CaO,9 respectively and metal co-catalysts (Ag10, Pd,11 Cu12). Sub-nanometer sized metal clusters induce high conversion and high selectivity minimizing the use of expensive noble metals on such catalysts.13 At small sizes on metal oxide supports, the surface free energy increases by the quantum size effect14 and low coordination number,15 enhancing performance16 as well as metal-support interactions17 that prevent subnano-cluster sintering18. Strong metal-support interactions (SMSI19) decrease chemisorption of H2 and CO on noble metals because either the surface of noble metal is covered by reduced support (e.g. Ti-suboxides20) or the noble metal oxidation state is modified. In SMSI, formation of oxygen vacancies in the support is the key process18 and, analogously, subnano-sized noble metals are often anchored at such surface defects.21 Here Pd subnano-clusters on TiO2 were prepared by flame spray pyrolysis (FSP)22 which can be 3


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 24

scaled up23 to yield production rates of several kg/h. Particle composition and size of FSP- and wet-made Pd/TiO2 photocatalysts were investigated by X-ray diffraction (XRD) and N2 adsorption. The characteristics of these Pd clusters were investigated by scanning transmission electron microscopy (STEM) as well as X-ray photoelectron (XPS), Raman and diffuse reflectance infrared Fourier transform (DRIFTS) spectroscopies. Furthermore, the thermal stability of Pd subnano-clusters in air and N2 was examined. The photocatalytic NOx conversion by such Pd/TiO2 under solar light was compared to that of commercial TiO2 (P25, Evonik) with or without photodeposited Pd on it.24

Experimental Catalyst preparation Titanium dioxide was prepared in one step by FSP with or without 1 wt.% of Pd (f-TiO2 or f-Pd/TiO2).22 In a 1:1 volumetric mixture of 2-ethylhexanoic acid (Aldrich, purity > 99 %):acetonitrile (Aldrich, purity > 99.5 %), 159 mM of titanium isopropoxide (Aldrich, purity > 97 %) and 0.6 mM of Palladium acetylacetonate (Aldrich, purity 99 %) were dissolved, as needed. This precursor solution was fed through the FSP capillary nozzle at 8 mL/min, dispersed to a fine spray by 5 L/min O2 (Pan Gas, purity >99%) through the adjacent burner ring and ignited/sustained by a support premixed methane/O2 flame (CH4 = 1.5 L/min, O2 = 3.2 L/min)22. So high purity Pd/TiO2 particles are produced and collected on a glass-fiber filter (Whatman GF, 24.7 cm effective diameter) 77 cm above the FSP nozzle by a gas pump (Busch Mink MM 1202 AV). For photocatalysis tests, such particles were collected for 4 min resulting in 170 ± 19 mg of homogeneous particle layer on the filter. Also, 1 wt.% of Pd was photodeposited24 on commercially available (P25, Evonik) titania (c-TiO2) and FSP-made (f-TiO2) that will be referred to as w-Pd/c-TiO2 and w-Pd/f-TiO2, respectively. 4


Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Such particles were prepared24 by dispersing 1 g of c-TiO2 (or f-TiO2) in 0.3 L of distilled water and sonication (Sandelin DT106, 480 W, 35 kHz) for 10 min. Then, 118 mg of Pd(NO3)2 solution (Alfa Aesar, Pd cont. 8.5 wt.%) were added and the pH was adjusted to 7.5 - 8.0 by a 0.8M NaOH solution. Furthermore, 0.1 L of ethanol (Aldrich, purity > 99.8 %) was added to this suspension, stirred magneticlly at 600 rpm for 30 min under dark and, subsequently, irradiated by UV light (Konrad Benda, UV-Lamp UV-8 S) for 12 hours. Afterwards, the particles were separated by centrifugation at 104 rpm for 10 min and dried at 50 oC in a vacuum chamber (10 mbar) for 12 hours. For photocatalysis tests, 10 g/L of every powder were dispersed by ultrasonication (Sonics Vibra-Cell 500 W, 100 kJ, 30/1 s on/off pulse) in a 1:1 ethanol:water solution. Five ml/min of that suspension were fed through the same FSP capillary nozzle and dispersed to a fine spray by 5 L/min oxygen but without the support flame resulting in homogeneous collection of 170 ± 10 mg) particles on the glass-fiber filter 62 cm above the nozzle. The photocatalytic performance of this particle layer is within 2-4% to that made by direct particle deposition by FSP as described above (Table S1).

Catalyst characterization The XRD patterns of all powders, with and without 20 wt.% of NiO (Aldrich, size: -325 mesh) as an internal standard25 were obtained by a AXS D8 Advance diffractometer (Bruker, Cu Kα, 40 kV, 40 mA). The crystalline and amorphous mass fractions of TiO2, f(x), were calculated from the integrated peak area of the XRD patterns at 25.5°, 27.5° and 42.5° for anatase, rutile and NiO, respectively:25 (anatase/rutile) =

/ ∙

and (amorphous) = 1 − (anatase) − (rutile).

The average anatase (ISDC: 663711) and rutile (ISDC: 663710) sizes were obtained by software Topas 4.2 while an average crystalline TiO2 size, s, was calculated accounting for the mass fractions of anatase 5


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 24

and rutile. The specific surface area (SSA) of all powders was obtained by N2 adsorption at 77 K, using the Brunauer-Emmett-Teller (BET) method (Micromeritics, Tristar II PULS). The BET equivalent particle diameter, d, was calculated assuming the weighted density of the particles (ρ(anatase), ρ(rutile) and ρ(amorphous) = 3.8, 4.2 and 3.8 g/cm3) as follow:26 d = 6000⁄∑((*) ∙ +(*)) [nm]. Images by STEM were recorded with a high-angle annular dark-filed detector (Hitachi, HD 2700-Cs, 200 kV) Raman spectroscopy was performed on a Renishaw InVia Reflex spectrometer with an Ar-ion laser as excitation source focused with a microscope (Leica, magnification 50×). These samples were pressed into tablets and placed on a glass plate. The spectra were recorded for 10 s with 20 accumulations to obtain sufficient signal-to-noise ratio with a CCD camera after being diffracted by a prism (1800 lines per millimeter) using 2.5 mW laser energy. The Raman spectra were fitted using spline-B interpolation27 and Origin 8.6 to find the peak position. XPS was performed using an X-ray photoemission spectrometer (Physical Electronics, Quantum 2000) with monochromatic Al Kα radiation (hν = 1486.7 eV). The Au 4f7/2 peak at 83.96 eV is used for the energy calibration of the instrument. The electron take-off angle was 45o and the diameter of the measured spot was ~150 µm. The samples were mounted on a stainless steel sample holder with an adhesive carbon tape. To compensate for eventual surface charging, built-in electron and argon ion neutralizers were used. The base pressure of the system was below 5 × 10−7 Pa. For the analysis the C 1s peak at 284.6 eV was used for charge correction and for the background subtraction the Shirley method was used. The Pd 3d5/2 peak (334-339 eV) was deconvoluted by four different oxidation states of Pd: metallic28,29 Pd, intermediate Pd, PdO28,30, ionic30 Pd2+ using Origin 8.6 and the Pd 3d3/2 peak (339-344 eV) was fitted keeping the full-width at half-maximum (FWHM) and setting the heights of the deconvoluted spectra of the 3d5/2 peak to 60%. 6


Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Diffuse reflectance infrared Fourier transform (DRIFTS) of the FSP-, wet-made and commercial powders were performed using a Vertex 70v spectrometer (Bruker Optics). Spectra were obtained by averaging 100 scans (100 scan/min) at 4 cm-1 resolution. The powders were placed in an in situ DRIFTS cell31 that was heated to 40 °C in He (Pangas, purity > 99.999 %) for 10 min. Afterwards 10% of CO in He was passed over the sample and the spectra were recorded continuously (at least for 20 min) until the spectrum reached steady state.

Photocatalytic evaluation for NO removal under solar-light The glass-fiber filters coated with the catalysts (as described above) were cut in 5 cm × 10 cm rectangular sheets that were held in a standard continuous-flow reactor (ISO 22197-1:2007) and uniformly irradiated through a 5 mm thick Pyrex glass by 100 mW/cm2 of solar light (Solarsim 150, Solaronix). An air stream of 3 L/min at 50 % relative humidity containing 1 ppm of NO flowed over the catalyst sheet while NO and NO2 concentrations were monitored every 3 s by a NOx detector (CLD 882 S, Eco Physics) at the reactor outlet. After the test for 5 hours, the filter coated with the catalysts was immersed in DI-water (10 mL), at least, for 12 hours to leach any nitrate remaining on the catalyst surface and the amount of leached nitrate, 23456 , was measured by a nitrate ion meter (LAQUAtwin B-742, Horiba). The average NO and NOx (NO + NO2) removals (ηNO and ηNOx) and NO2 and nitrate formations (ηNO2 and ηNO3-) under solar-light for ti = 5 hours were defined as follows: 934: [ppm] ?@ 934 [ppm] > ; ;

734 = 8 734A 734C

; JKL 9 [ppm]?@ H 34 22.4 [L/mol] >

Results and Discussion Figure 1 shows the evolution of NO (filled symbols) and NO2 (open symbols) concentrations from solar light photocatalysis of NO (ISO 22197-1:2007) using commercial TiO2 (c-TiO2, squares) and FSP-made TiO2 with 1 wt.% of Pd (f-Pd/TiO2, circles). For the first 40 min under dark, both NO and NO2 are at 1 and 0 ppm, respectively, indicating no conversion of NO and removal of catalytic NOx, as expected. When solar light is on, NO and NO2 concentrations rapidly reach steady state by both photocatalysts during 5 h of irradiation (40-340 min). The net difference, 934 KL − 934 MN; − 934A MN; , corresponds to the formation3 of nitrate as NOx removal. Under light irradiation, more NO is removed by f-Pd/TiO2 than c-TiO2 due to Pd presence32. On the other hand, c-TiO2 forms slightly more NO2 than f-Pd/TiO2, while formation of N2O by these powders was not detected by FTIR. Table 1 shows TiO2 mass fractions (anatase, rutile and amorphous), average crystal size, s, and BET-equivalent particle diameter, d, of c-TiO2 and f-TiO2 as well as with photodeposited 1wt.% Pd on them (w-Pd/c-TiO2 and w-Pd/f-TiO2) and FSP-made Pd/TiO2 as prepared (f-Pd/TiO2) as well as annealed for 2 hours in air at 600 oC (f-Pd/TiO2/air) and in N2 at 400 oC (f-Pd/TiO2/N2) along with their average efficiency of NO and NOx removals (ηNO and ηNOx) as well as NO2 formation (ηNO2) for 5 hours under artificial solar-light as in Figure 1. Regardless of preparation method, no detectable Pd and PdO peaks appear by XRD in all powders containing Pd (Fig. S1). Furthermore, the low Pd presence (1 wt.%) does not affect their crystal size, particle diameter and phase composition as seen with trace Au22 and Pt26 on TiO2. Anatase is the dominant crystalline phase for all powders here and the mass ratio of anatase and

8


Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

rutile (= the fraction in total crystalline content in Table S2) in FSP-made powders is consistent with that in similarly made26 Pt/TiO2. Most interestingly, the amorphous fraction in commercial TiO2 (c-TiO2) is 9.1 wt%, in agreement with Ohtani et al.,25 but it is significantly smaller than that in FSP-made powders, ~33 wt.%, as indicated by their non-straight XRD base lines compared to those of c-TiO2. In c-TiO2 and w-Pd/c-TiO2, both crystal, s, and primary particle, d, sizes are similar, as expected. In FSP-made powders, the BET size, d, is slightly smaller than the crystalline size, s, indicating the presence of small amorphous TiO2 particles. This is in agreement with similarly flame-made TiO2 using propionic acid as solvent with low enthalpy.33 As shown in Fig. 1 and Table 1, the ηNO of f-Pd/TiO2 is much higher than that of c-TiO2 but their ηNO2 are similar resulting in 7 times higher ηNOx (NO removal as NOx) for f-Pd/TiO2 than c-TiO2. For all the catalysts, ηNOx is comparable with ηNO3- indicating that most of converted NO becomes NO2 and nitrate, which is consistent with photocatalytic NO removal by pure TiO2.3 Only the ηNOx of f-Pd/TiO2 (as- prepared and annealed in air and N2) is slightly higher than ηNO3- implying some N2 formation34 which is hardly detected because of the low concentration. This activity of f-Pd/TiO2 is higher than that of N-doped6 TiO2 that was about 1.4 and 2 times higher than c-TiO2 in UV and visible light, respectively.6 Furthermore, the activity of plasma-treated7 TiO2 and Ti4+ complex35 were only about 2 times higher than that of commercial pure TiO2 (ST-017,35, Ishihara). Most importantly, the ηNO and ηNO2 of f-Pd/TiO2 is higher and lower, respectively, than those of wet-made (w-Pd/c-TiO2 and w-Pd/f-TiO2) indicating that the high selectivity of f-Pd/TiO2 for nitrate formation. As a result, the ηNOx of f-Pd/TiO2 is about 3 times higher than w-Pd/c-TiO2 as Pd photodeposition doubles the ηNOx of c-TiO2, in agreement with Wu et al.32 but hardly affects that of f-TiO2. Also, the ηNOx of f-TiO2 is higher than that of c-TiO2, probably due to the higher SSA and low 9


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 24

rutile TiO2 content of the former.8 As shown in Table 1, both Pd deposition methods do not influence the size and composition of TiO2. Interestingly, annealing f-Pd/TiO2 in air at 600 oC (f-Pd/TiO2/air) slightly improves the ηNO and ηNOx because of crystallization of amorphous TiO2 that is not photocatalytically active25. On the other hand, annealing in N2 at 400 oC degrades the ηNO and ηNOx, although its size and crystalline composition of TiO2 remain almost the same. As a result, the characteristics of Pd (size, oxidation state, etc.) play a crucial role on NOx removal efficiency. Figure 2 shows representative images of photodeposited Pd particles (bright spots) on (a) commercial and (b) FSP-made TiO2 (gray) along with two magnifications (c, d) of as-prepared f-Pd/TiO2 as well as annealed for 2 hours in (e) air at 600 oC and (f) N2 at 400 oC. The size of photodeposited Pd (1-10 nm) on c-TiO2 (a) is slightly larger than that (1-5 nm) on f-TiO2 (b) probably due to smaller SSA of c-TiO2, a difference that does not affect their photocatalyst activity (Table 1). In f-Pd/TiO2 (c), however, there are no distinguishable Pd particles (> 1 nm) because of the extremely small size of Pd clusters that can be seen only (circles) at the higher magnification (Fig. 2d). Most interestingly, after annealing in air at 600 oC, the Pd size in f-Pd/TiO2 is still in the subnano range (e) even though such Pd clusters (< 1 nm) should have melted at this temperature36. On the other hand, annealing that powder under N2 converted most of the Pd clusters into nanoparticles (1-5 nm) as shown in Fig. 2f. Even though some Pd clusters could still remain, these are hardly detected because of the comparable z-contrast of Pd and TiO2. Therefore, the presence of Pd subnano-clusters that are stable in air up to 600 oC for 2 hours seems crucial for the superior photocatalytic NOx removal. Conversion of some subnano-clusters into Pd nanoparticles by N2 annealing at 400 oC decreases this activity (Table 1). In Pd clusters on TiO2, theoretical calculations suggest that Pd subnano-cluster bonds to several terminated O and Ti on TiO2 surface forming PdOx-TiO2 structure.21 Similar bonding structure is 10


Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

observed also with atomically dispersed Pt and Au on CeO2 and the bonding oxidizes such noble metals.16 Figure 3 shows XPS spectra of photodeposited Pd on FSP-made (a), and commercial TiO2 (b) along with as-prepared f-Pd/TiO2 (c) and annealed for 2 hours in air at 600 oC (d, f-Pd/TiO2/air) and in N2 at 400 oC (e, f-Pd/TiO2/N2). These spectra are deconvoluted by four different states of Pd: metallic Pd (red)28,29, intermediate Pd oxidation (green), PdO (blue)28,30, ionic Pd2+ (orange)30 along with their cumulative fitting (black). The obtained spectra and deconvolutions are reproducible as shown with two samples prepared in different batches (Fig. S2). The reasonably good match in the 3d3/2 peak (339-344 eV) using the fraction of peak area of the 3d5/2 peak (334-339 eV) indicates appropriate baseline correction and peak fitting. Both photodeposited Pd on f-TiO2 (a) and c-TiO2 (b) consist of mainly metallic Pd and different Pd oxidation states, in agreement with literature29. The peak position of metallic Pd (~334.6 eV) is slightly lower than known values (335.0 eV)28, as observed also with photodeposited Pd on TiO2 in aqueous-alcohol solutions29. This shift could be attributed to charge injection37 by intermediate products of ethanol oxidation whose redox potential is higher than the Fermi level of metallic Pd and the conduction band of TiO2. In contrast, for Pd on FSP-made f-Pd/TiO2 (Fig. 3c) an intermediate Pd oxidation state dominates. Its peak (green, ~335.6 eV) lies between metallic Pd (red, ~334.6 eV) and PdO (blue, ~336.6 eV). The absence of a separate metallic Pd peak is attributed to the weak signal intensity at 334.6 eV in contrast to Fig. 3a,b. With Pd clusters, PdOx-TiO2 interaction21 and/or sub-stoichiometric PdOx (x < 1) could induce different intermediate oxidation states broadening that peak, consistent with the broader FWHM of the intermediate Pd oxidation states than that (= 1.1 eV) of the other Pd forms (Fig. S2d). The peak intensity of PdO (blue, ~336.6 eV) is weaker than that of the intermediate Pd 11


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 24

oxidation state indicating some presence of PdO subnano-clusters and/or particles. When the size of FSP-made Pd on TiO2, however, is above 10 nm, PdO is the primary state as detected by XRD.38 Therefore, the intermediate oxidation state is a characteristic feature of the Pd subnano-clusters on oxides. Regardless of Pd preparation method, the fractions of ionic Pd2+ and PdO are rather minor and nearly equivalent indicating that these Pd states are probably irrelevant. As a consequence, the intermediate Pd oxidation state (~ 335.6 eV) seems crucial for the remarkable NO removal activity of FSP-made Pd/TiO2 (Fig. 1). After annealing f-Pd/TiO2 in air for 2 hours at 600 oC, its Pd oxidation spectra (Fig. 3d) are still identical to those of as-prepared f-Pd/TiO2 (Fig. 3c). This implies that the Pd subnano-clusters remain largely intact by annealing as seen also by microscopy (Fig. 2e). On the other hand, annealing f-Pd/TiO2 in N2 decreases drastically the intermediate Pd oxidation state spectrum while the metallic Pd spectrum becomes stronger (Fig. 3e). Metallic Pd is created mainly from intermediate Pd and to a lesser extent by PdO39 (Fig. S2) resulting in metallic Pd particles from their clusters (Fig. 2f). Nevertheless, a certain fraction of intermediate Pd oxidation state is still preserved after annealing in N2 indicating the presence of some subnano-Pd clusters. Figure 4 shows FTIR spectra of CO adsorption on the above powders. The peaks around 2130 and 2180 cm-1 that are assigned to CO adsorption40 on TiO2 appear in all spectra. The CO-Pd peaks41,42 around 1800-2000 and 2090 cm-1 that correspond to multiple-coordinated CO on Pd and linear CO-Pd, respectively, are notably lower for f-Pd/TiO2 (blue spectrum) than photodeposited and annealed in N2 powders. Furthermore, the weaker CO adsorption on that powder is maintained even after annealing it in air (purple spectrum) indicating the presence of PdOx (x < 1) and PdOx-TiO2 interaction. To form and stabilize such subnano-clusters, the number of oxygen defects in the oxide support 12


Page 13 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

could be the critical parameter16,43 as more reducible oxides provide more anchoring sites18. The importance of defects to stabilize Pd clusters on TiO2 is also predicted by density functional theory.21 The presence of oxygen defects in TiO2 and their stability were investigated by Raman spectroscopy to correlate the stability of the anchoring sites with catalyst performance. Figure 5 shows the Raman Eg peak of anatase TiO2 of the above powders. Its position correlates with the number of defects in the anatase lattice.44,45 The peak position of c-TiO2 (purple spectrum) is almost identical to the stoichiometric anatase TiO2 (dashed line) while that from f-TiO2 (blue spectrum) slightly shifts to larger wavenumbers, indicating the presence of more defects than in c-TiO2.44 Photodeposition of Pd to either commercial (w-Pd/c-TiO2, orange spectrum) or FSP-made TiO2 (w-Pd/f-TiO2, black spectrum) shifts further that peak, indicating the formation of oxygen defects44 that could be attributed to charge injection by intermediate products of ethanol oxidation.37 The peak position of FSP-made Pd on TiO2 (f-Pd/TiO2, red spectrum) also shifts to larger wavenumbers but it does not change by annealing for 2 hours under air at 600 oC (f-Pd/TiO2/air, purple spectrum) and N2 at 400 oC (f-Pd/TiO2/N2, green spectrum). The peak position of f-Pd/TiO2/N2, however, shifts to small wavenumbers when it is annealed in air indicating removal of oxygen defects (Fig.S3).44 This is attributed to the growth of Pd subnano-clusters into particles (Fig. 2f) that spatially segregates Pd clusters from TiO2 oxygen defects ceasing such PdOx-TiO2 interaction.

Conclusions Solar-light photocatalysts consisting of palladium subnano-clusters (< 1 nm) on nanostructured TiO2 (f-Pd/TiO2) were prepared by flame aerosol technology. Unser solar light they exhibited 3 or 7 times higher NO removal efficiency as NOx than commercial TiO2 (c-TiO2) with or without 13


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 24

photodeposited Pd on it. The Pd subnano-clusters are interacting strongly with oxygen defects on the TiO2 surface creating an intermediate oxidation state of Pd as proven by XPS that hinders CO chemisorption on Pd. Such strong PdOx-TiO2 interactions hold to annealing up to 600 oC in air for 2 hours but partially cease upon annealing at 400 oC in N2 converting most Pd subnano-clusters into Pd nanoparticles as detected by CO chemisorption and seen by STEM. This particle growth deteriorates the NOx removal activity of flame-made Pd/TiO2 indicating that Pd subnano-clusters are critical for the remarkable photocatalytic NOx removal under solar light.

Acknowledgements The research leading to these results has received funding from the Swiss National Science Foundation (grant no. 200021_149144). The help of Severin Hahn in particle characterization is appreciated. TEM work was performed at the Electron Microscopy Center of ETH Zurich (EMEZ).

Notes The authors declare no competing financial interests.

Supporting Information Available Additional XRD patterns, XPS fitting data and Raman peak positions of presented powders as well as NOx removals performance of f-TiO2 and f-Pd/TiO2 particle film prepared by different deposition methods are available. This material is available free of charge via the Internet at http://pubs.acs.org.

References 14


Page 15 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1. Kobylinski, T. P.; Taylor, B. W., J. Catal. 1974, 33, 376-384. 2. Stark, W. J.; Wegner, K.; Pratsinis, S. E.; Baiker, A., J. Catal. 2001, 197, 182-191. 3. Ibusuki, T.; Takeuchi, K., J. Mol. Catal. 1994, 88, 93-102. 4. Hashimoto, K.; Irie, H.; Fujishima, A., Jpn. J. Appl. Phys. 2005, 44, 8269. 5. Maggos, T.; Plassais, A.; Bartzis, J.; Vasilakos, C.; Moussiopoulos, N.; Bonafous, L., Environ. Monit. Assess. 2008, 136, 35-44. 6. Todorova, N.; Vaimakis, T.; Petrakis, D.; Hishita, S.; Boukos, N.; Giannakopoulou, T.; Giannouri, M.; Antiohos, S.; Papageorgiou, D.; Chaniotakis, E.; Trapalis, C., Catal. Today 2013, 209, 41-46. 7. Nakamura, I.; Negishi, N.; Kutsuna, S.; Ihara, T.; Sugihara, S.; Takeuchi, K., J. Mol. Catal. A: Chem. 2000, 161, 205-212. 8. Bloh, J. Z.; Folli, A.; Macphee, D. E., RSC Advances 2014, 4, 45726-45734. 9. Ichiura, H.; Kitaoka, T.; Tanaka, H., Chemosphere 2003, 51, 855-860. 10. Bowering, N.; Croston, D.; Harrison, P. G.; Walker, G. S., Int. J. Photoenergy 2007, 2007, 1-8. 11. Hsiao, Y.-C.; Tseng, Y.-H., Micro & Nano Letters 2010, 5, 317-320. 12. Anpo, M.; Nomura, T.; Kitao, T.; Giamello, E.; Murphy, D.; Che, M.; Fox, M., Res. Chem. Intermed. 1991, 15, 225-237. 13. Flytzani-Stephanopoulos, M.; Gates, B. C., Annu. Rev. Chem. Biomol. Eng. 2012, 3, 545-574. 14. Kaden, W. E.; Wu, T.; Kunkel, W. A.; Anderson, S. L., Science 2009, 326, 826-829. 15. Lopez, N.; Janssens, T. V. W.; Clausen, B. S.; Xu, Y.; Mavrikakis, M.; Bligaard, T.; Nørskov, J. K., J. Catal. 2004, 223, 232-235. 16. Fu, Q.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Science 2003, 301, 935-938. 17. Tsud, N.; Johánek, V.; Stará, I.; Veltruská, K.; Matolıń , V., Surf. Sci. 2000, 467, 169-176. 15


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 24

18. Shinjoh, H.; Hatanaka, M.; Nagai, Y.; Tanabe, T.; Takahashi, N.; Yoshida, T.; Miyake, Y., Top. Catal. 2009, 52, 1967-1971. 19. Tauster, S. J., Acc. Chem. Res. 1987, 20, 389-394. 20. Fujiwara, K.; Deligiannakis, Y.; Skoutelis, C. G.; Pratsinis, S. E., Appl. Catal. B-Environ. 2014, 154-155, 9-15. 21. Zhang, J.; Zhang, M.; Han, Y.; Li, W.; Meng, X.; Zong, B., J. Phys. Chem. C 2008, 112, 19506-19515. 22. Mädler, L.; Stark, W. J.; Pratsinis, S. E., J. Mater. Res. 2003, 18, 115-120. 23. Mueller, R.; Madler, L.; Pratsinis, S. E., Chem. Eng. Sci. 2003, 58, 1969-1976. 24. Bard, A. J., J. Photochem. 1979, 10, 59-75. 25. Ohtani, B.; Prieto-Mahaney, O. O.; Li, D.; Abe, R., J. Photochem. Photobiol. A: Chem. 2010, 216, 179-182. 26. Teoh, W. Y.; Mädler, L.; Beydoun, D.; Pratsinis, S. E.; Amal, R., Chem. Eng. Sci. 2005, 60, 5852-5861. 27. Ohsaka, T.; Izumi, F.; Fujiki, Y., J. Raman Spectrosc. 1978, 7, 321-324. 28. Brun, M.; Berthet, A.; Bertolini, J. C., J. Electron. Spectrosc. Relat. Phenom. 1999, 104, 55-60. 29. Yui, T.; Kan, A.; Saitoh, C.; Koike, K.; Ibusuki, T.; Ishitani, O., ACS Appl. Mater. Interfaces 2011, 3, 2594-2600. 30. Roy, S.; Hegde, M. S.; Ravishankar, N.; Madras, G., J. Phys. Chem. C 2007, 111, 8153-8160. 31. Hoxha, F.; Schimmoeller, B.; Cakl, Z.; Urakawa, A.; Mallat, T.; Pratsinis, S. E.; Baiker, A., J. Catal. 2010, 271, 115-124. 32. Wu, Z.; Sheng, Z.; Wang, H.; Liu, Y., Chemosphere 2009, 77, 264-268. 16


Page 17 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

33. Chiarello, G. L.; Selli, E.; Forni, L., Appl. Catal. B-Environ. 2008, 84, 332-339. 34. Wu, Q.; van de Krol, R., J. Am. Chem. Soc. 2012, 134, 9369-9375. 35. Sano, T.; Negishi, N.; Koike, K.; Takeuchi, K.; Matsuzawa, S., J. Mater. Chem. 2004, 14, 380-384. 36. Guisbiers, G.; Abudukelimu, G.; Hourlier, D., Nanoscale Res. Lett. 2011, 6, 1-5. 37. Gomes, W. P.; Freund, T.; Morrison, S. R., J. Electrochem. Soc. 1968, 115, 818-823. 38. Mekasuwandumrong, O.; Phothakwanpracha, S.; Jongsomjit, B.; Shotipruk, A.; Panpranot, J., Powder Technol. 2011, 210, 328-331. 39. Peuckert, M., J. Phys. Chem. 1985, 89, 2481-2486. 40. Liao, L.-F.; Lien, C.-F.; Shieh, D.-L.; Chen, M.-T.; Lin, J.-L., J. Phys. Chem. B 2002, 106, 11240-11245. 41. Rainer, D.; Wu, M. C.; Mahon, D.; Goodman, D., J. Vac. Sci. Technol. 1996, 14, 1184-1188. 42. Unterhalt, H.; Rupprechter, G.; Freund, H.-J., J. Phys. Chem. B 2002, 106, 356-367. 43. Fu, Q.; Deng, W.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Appl. Catal. B-Environ. 2005, 56, 57-68. 44. Parker, J. C.; Siegel, R. W., Appl. Phys. Lett. 1990, 57, 943-945. 45. Parker, J. C.; Siegel, R. W., J. Mater. Res. 1990, 5, 1246-1252.

17


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 18 of 24

Table 1 Mass fraction, average crystal size, s, and BET-equivalent particle diameter, d, of commercial TiO2 (c-TiO2), FSP-made (f-TiO2 and f-Pd/TiO2) and wet-photodeposited Pd on TiO2 (w-Pd/c-TiO2 and w-Pd/f-TiO2) and their average efficiencies of NO and NOx removals (ηNO and ηNOx) as well as NO2 formation (ηNO2) for 5 hours under artificial solar-light. Mass fraction wt.% Size/diameter, nm Performance, % Powders Anatase Rutile Amorphous s d ηNO ηNO2 ηNOx ηNO3c-TiO2 82.6 8.4 9.1 31.2 31.3 38.0 33.4 4.6 5.1 f-TiO2 73.4 3.6 23.0 12.9 10.6 53.0 40.1 12.9 11.9 w-Pd/c-TiO2 82.1 8.5 8.4 31.3 33.8 55.0 45.4 9.6 10.9 w-Pd/f-TiO2 72.6 3.3 23.9 12.1 10.4 52.0 40.1 11.9 12.2 f-Pd/TiO2 73.8 1.9 22.5 13.1 11.3 63.5 31.7 31.8 30.5 f-Pd/TiO2/air 86.1 8.0 5.9 17.5 18.4 67.0 32.1 34.9 29.7 f-Pd/TiO2/N2 74.1 2.5 23.4 13.4 12.5 52.7 27.4 25.3 21.4

18


Page 19 of 24

1.0

Light irradiation (5 hours) Dark

0.8 Dark

Concentration, ppm

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

0.6 0.4 0.2

c-TiO2 f-Pd/TiO2 Filled: NO Open: NO2

0.0 0

50

100

150

200

250

300

350

Time, min Figure 1. Evolution of NO and NO2 concentrations during NOx removal test (ISO 22197-1:2007) with commercial TiO2 (c-TiO2, squares) and FSP-made TiO2 with 1 wt.% of Pd (f-Pd/TiO2, circles) under artificial solar-light (100 W/cm2). The initial NO and NO2 concentrations is 1 and 0 ppm, respectively.

19


ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

a

30 nm

Page 20 of 24

b

20 nm

c

d

10 nm

6 nm

e

f

10 nm

4 nm

Figure 2. STEM images of (a) w-Pd/c-TiO2, (b) w-Pd/f-TiO2, f-Pd/TiO2, (c,d) as-prepared at two magnifications, annealed for 2 hours in (e) air at 600 oC and (f) N2 at 400 oC. 20


Page 21 of 24

metallic Pd ionic Pd2+

a:w-Pd/f-TiO

c:f-Pd/TiO

intermediate Pd Cumulative fitting

PdO

2

b:w-Pd/c-TiO Counts, a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

2

2

d:f-Pd/TiO /air 2

e:f-Pd/TiO /N 2

346

344

2

342

340 338 Binding Energy, eV

336

334

Figure 3. XPS spectra (circles) of a) w-Pd/c-TiO2, b) w-Pd/f-TiO2 and f-Pd/TiO2 c) as prepared and annealed for 2 hours in d) air at 600 oC (f-Pd/TiO2/air) and e) N2 at 400 oC (f-Pd/TiO2/N2) for 2 hours along with their deconvolution by four oxidation states of Pd: metallic Pd (red), intermediate Pd (green), PdO (blue), ionic Pd2+ (orange) and cumulative fitting (black).

21


ACS Catalysis

w-Pd/c-TiO2

Absorption, a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 24

w-Pd/f-TiO2 f-Pd/TiO2/N2 f-Pd/TiO2/air

2200

2100

2000

1900

f-Pd/TiO2 f-TiO2 c-TiO2 1800 1700

Wavenumber, cm-1 Figure 4. DRIFTS of CO adsorption on c-TiO2, photodeposited Pd on TiO2 (w-Pd/c-TiO2 and w-Pd/f-TiO2) and FSP-made powders as-prepared (f-TiO2 and f-Pd/TiO2) as well as annealed for 2 hours under air at 600 oC (f-Pd/TiO2/air) and N2 at 400 oC (f-Pd/TiO2/r-N2).

22


Page 23 of 24

Stoichiometric anatase TiO2

Intensity, a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

w-Pd/f-TiO2 w-Pd/c-TiO2 f-Pd/TiO2/N2 f-Pd/TiO2/air f-Pd/TiO2 f-TiO2 c-TiO2 100

120

140

160

180

Raman Shift, cm-1 Figure 5. Raman spectra of c-TiO2, photodeposited Pd on TiO2 (w-Pd/c-TiO2 and w-Pd/f-TiO2) and FSP-made powders as-prepared (f-TiO2 and f-Pd/TiO2) as well as annealed for 2 hours under air at 600 o C (f-Pd/TiO2/air) and N2 at 400 oC (f-Pd/TiO2/r-N2).. The vertical dashed line indicates the peak position of stoichiometric (defect-free) anatase44 TiO2 at 143 cm-1.

23


ACS Catalysis

Table of contents

NOx (NO + NO2) concentration, ppm

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 24 of 24

Solar light (100 mW/cm2)

1.0

TiO2 (P25, Evonik) 0.8

Pd/TiO2 (P25) (photodepositon) Pd/TiO2 (flame-made)

0.6

0

2

4

6

Time, hours

24


Pd Subnano-Clusters on TiO2 for Solar-Light Removal of NO - ACS

Recommend Documents