Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in low-temperature selective ammonia oxidation into nitrogen and water vapour

Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in low-temperature selective ammonia oxidation into nitrogen and water vapour

Accepted Manuscript Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in low-temperature selective ammonia oxidat...

5MB Sizes 0 Downloads 3 Views

Accepted Manuscript Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in low-temperature selective ammonia oxidation into nitrogen and water vapour Magdalena Jabłońska, Wirawan Ciptonugroho, Kinga Góra-Marek, Mohammad G. AlShaal, Regina Palkovits PII:

S1387-1811(17)30132-4

DOI:

10.1016/j.micromeso.2017.02.070

Reference:

MICMAT 8180

To appear in:

Microporous and Mesoporous Materials

Received Date: 27 October 2016 Revised Date:

6 February 2017

Accepted Date: 23 February 2017

Please cite this article as: M. Jabłońska, W. Ciptonugroho, K. Góra-Marek, M.G. Al-Shaal, R. Palkovits, Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in lowtemperature selective ammonia oxidation into nitrogen and water vapour, Microporous and Mesoporous Materials (2017), doi: 10.1016/j.micromeso.2017.02.070. 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.

RI PT

ACCEPTED MANUSCRIPT

NH3 + O2

SC

100

N2, NO, N2O

1.5% Ag/mesoTiO2

75

NH3 conversion NH3 conversion (H2-pretreated) N2 selectivity

50

NO selectivity N2O selectivity

25

EP AC C

Ag0 Ag+ and Agn+ clusters

TE D

in situ H2-pretreatment

Conversion/Selecivity [%]

mesoporous TiO2

M AN U

Ag+ Ag0 Ag0 Ag+

0 100

150

200

250

300 0

Temperature [ C]

350

400

ACCEPTED MANUSCRIPT

Preparation, characterization and catalytic performance of Ag-modified mesoporous TiO2 in lowtemperature selective ammonia oxidation into nitrogen and water vapour

RI PT

Magdalena Jabłońskaa*, Wirawan Ciptonugrohoa, Kinga Góra-Marekb, Mohammad G. Al-Shaalc, Regina Palkovitsa* a

Institut für Technische und Makromolekulare Chemie, Chair of Heterogeneous Catalysis and Chemical

b

Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30-060 Kraków, Poland

Max-Planck-Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, D-45470 Mülheim an der Ruhr, Germany

M AN U

c

SC

Technology, RWTH Aachen University, Worringerweg 2, 52074 Aachen, Germany

*Corresponding author. Tel: +49 241 80 26497; Fax: +49 241 80 22177. E-mail address: [email protected]

Abstract

TE D

aachen.de (R. Palkovits)

Mesoporous TiO2 was prepared by evaporation induced self-assembly (EISA) and used as a support for silver with

EP

loadings of 1.5 or 10 wt.%, respectively. Silver deposited on commercial TiO2 (anatase) and γ-Al2O3 served as reference catalysts. The mesoporous TiO2 were calcined at 500, 600, 700 and 800°C in order to obtain anatase

AC C

and/or rutile phases. The materials were characterized with respect to structural (XRD, TEM, UV-vis-DRS), textural parameters (N2 adsorption-desorption), acidic properties (NH3-TPD, FTIR studies), redox properties (H2TPR), and were applied as catalysts for the selective ammonia oxidation into nitrogen and water vapour (NH3SCO). The catalytic performance (activity and N2 selectivity) was favoured over (1.5 wt.%)Ag-doped mesoporous TiO2 calcined at 600°C. Thus, mesoporous TiO2 with the predominant anatase phase, as a support guaranteed the formation of easily reducible highly dispersed oxidized silver species. Furthermore, these species were converted 1

ACCEPTED MANUSCRIPT into Ag0 and possibly Agnδ+ clusters through in situ H2-pretreatment of catalysts. While mainly metallic silver catalysed the decomposition of N2O into N2 and surface oxygen species, leading to higher N2 selectivity in NH3-

RI PT

SCO.

Keywords: mesoporous titania, evaporation induced self-assembly, silver, selective ammonia oxidation

SC

1. Introduction

Silver-based catalysts present a class of highly active and N2 selective catalysts for low temperature (≤300°C)

M AN U

selective catalytic ammonia oxidation into nitrogen and water vapour (NH3-SCO) [1–3]. Il’chenko et al. [4,5] found a lower specific catalytic activity of metallic Ag at 300°C than that of Pt or Pd. Among silver-based catalysts, Ag/Al2O3 was widely investigated to gain insights into the influence of the valance state of Ag species and particle size on catalytic activity [3,6], and to elucidate the reaction mechanism [1,7]. Qu et al. [1] found silver

TE D

deposited on alumina as the most active catalysts among Ag-modified Al2O3, SiO2, NaY or TiO2, respectively. Mainly metallic Ag (based on XRD and UV-vis-DRS analyses) has been proposed as an active species in NH3SCO. The application of different supports enabled preparation of a broad range of materials with different Ag0

EP

particle sizes reaching from 5.3 nm for Ag/Al2O3 up to 20.7 nm for Ag/TiO2, respectively. Silver supported on Al2O3 showed the highest catalytic activity with full conversion at 180°C, while Ag/TiO2 revealed superior

AC C

selectivity (above 85% up to 200°C). On the other hand, for other metals the application of mesoporous supports facilitated selective ammonia oxidation due to enhanced transport of the reactants to the active sites. However, no studies focused on silver deposited on mesoporous supports and only few investigations refer to mesoporous catalysts in NH3-SCO at all [8]. E.g. mesoporous CuFe3O4 synthesized via KIT-6 replication significantly enhanced ammonia oxidation compared to the conventional spinel. The formation of mesostructures provided an appreciably high specific surface area and potentially enhanced access for NH3 molecules to reach the active sites 2

ACCEPTED MANUSCRIPT

[8]. KIT-6 was also replicated to produce mesostructured CuO/RuO2, which showed full NH3 conversion at 180°C and a N2 selectivity of 97% [9]. Mesoporous Ce-Ti-MoOx [10] or SBA-15 decorated with Fe and Mo [11] were successfully applied in the selective NOx reduction with NH3 (NH3-SCR). In addition to hard templating,

RI PT

mesostructured materials can be obtained by applying organic molecules as a template, so called soft templating. Evaporation induced self-assembly (EISA) is a robust method to synthesize mesostructured materials in which organic templates are employed to direct the formation of mesopores. Volatile solvents are usually involved to

SC

carry out EISA because the slow evaporation rate at mild temperatures gradually increases the micelle concentration facilitating the further formation of a liquid-crystal mesophase through a self-assembly mechanism

M AN U

[12]. Slow evaporation at milder temperatures provides sufficient time for the self-assembly of template and metal oxide precursor to occur allowing the formation of a better mesostructure. EISA has been applied for the preparation of mesoporous TiO2 [13,14], Nb2O5 [15], or ZrO2 [16], etc. Considering its simplicity combined with exceptional mesoporosity motivated us to apply EISA to synthesize mesostructured TiO2 for deposition of silver.

TE D

We explored the effects of different silver loadings (1.5 and 10 wt.%) and different calcination temperatures (500, 600, 700, 800°C) as well as pretreatment conditions (pure Ar or 5.0 vol.% H2 diluted in Ar). We investigated the relation between the physicochemical properties and the catalytic activity in the selective ammonia oxidation into

EP

nitrogen and water vapour over Ag-doped mesoporous TiO2. Moreover, we applied different in situ pretreatment conditions (Ar or H2/Ar) of the catalysts combined with microreactor tests of NH3-SCO. Our work aimed at

AC C

understanding the role of structure, catalysts acidity and reducibility for catalytic activity and selectivity in ammonia oxidation.

2. Experimental details

2.1. Catalyst preparation

3

ACCEPTED MANUSCRIPT

Mesoporous TiO2 was prepared by evaporation induced self-assembly (EISA). 2.43 g F127 and 1.53 g citric acid (Sigma-Aldrich) were dissolved in 73 ml ethanol (Chemsolute) followed by vigorous stirring for 2 h. Subsequently, the mixture was firstly acidified with 3.65 g HCl (Chemsolute) before the addition of 3.50 g Ti-

RI PT

isopropoxide (Sigma-Aldrich). The sol was then agitated for 3 h and poured in a set of Petri dish keeping the same liquid level. The sol was aged for 48 h to evaporate the solvent at room temperature before being stored in the oven at 100°C overnight for solidification. Furthermore, the dried sample was peeled off before being calcined at 600°C

SC

for 6 h with ramping rate 1°C/min.

Mesoporous TiO2 was doped with Ag by wet impregnation using an aqueous solution of AgNO3 (Sigma-Aldrich).

M AN U

For comparative purposes commercial TiO2 (Sigma-Aldrich) and γ-Al2O3 (Merck) were prepared. All materials were impregnated with 1.5 and 10 wt.% of silver, according to literature indications [1,3,17]. All prepared samples were dried overnight and subsequently calcined in static air at 600°C for 6 h with a ramping rate of 1°C/min. The selected Ag-modified mesoporous TiO2 were calcined at 500, 700 or 800°C in order to obtain anatase and/or rutile

TE D

phases, respectively. For catalytic experiments, a sieved fraction of particles with a size of 0.250-0.50 mm was used. The weight ratio of metals was measured with respect to the mass of the supports.

EP

2.2. Catalyst characterization

The X-ray diffraction (XRD) patterns of the samples were recorded with a D5000 Siemens diffractometer using Cu

AC C

Kα radiation (λ = 1.54056 Å, 45 kV, 40 mA). Small-angle X-ray scattering (SAXS) patterns of the mesoporous TiO2 was recorded applying an Anton Paar SAXSess (width of detection area: 3 mm; exposure time: 1000s; number of frames: 10 s).

The specific surface area (SBET), pore size distributions, total pore volumes, and sorption isotherms of the samples were determined by low-temperature (-196°C) N2 physisorption using a Micrometrics ASAP 2000 device. Prior to nitrogen adsorption, the samples were outgassed at 250°C under vacuum. 4

ACCEPTED MANUSCRIPT

The temperature programmed desorption of ammonia (NH3-TPD) of the samples was performed in a flow microreactor system equipped with QMS MKS, Cirrus 2 detector. Prior to ammonia sorption, the sample (100 mg) was outgassed in a flow of pure Ar at 550ºC for 1 h. Subsequently, the microreactor was cooled down to 70ºC and

RI PT

the sample was saturated in a flow (20 cm3/min) of gas mixture containing 1.0 vol.% of NH3 diluted in Ar for about 2 h. Then, the sample was purged in a flow of pure Ar until a constant baseline level was attained (about 2 h). In the next step, the temperature of the reactor was raised in the range of 70-550°C with a linear heating rate of

SC

5°C/min in a flow of pure Ar (20 cm3/min).

The Fourier transform infrared spectroscopy (FTIR) with probe molecules – Py (pyridine) and CO – were

M AN U

performed using a Bruker Vertex spectrometer equipped with an MCT detector with a resolution of 2 cm-1. The samples were pressed into the form of self-supporting discs (ca. 5 mg/cm2) and evacuated in a quartz IR cell at 350oC under vacuum for 1 h. Pyridine (Py) was adsorbed at 130°C, and the physisorbed molecules were next removed by evacuation at the same temperature. Sorption of CO as probe molecule was performed at -100°C. The

TE D

Ag-doped samples due to their black color were diluted in SiO2; 10 mg of catalyst diluted with 20 mg of SiO2 (Avantor Performance Materials Poland). Sorption of pyridine was preformed also on pure SiO2 used for dilution of the studied materials. No Brönsted acidity was detected for SiO2. For the Ag-doped catalysts, a negligible

EP

concentration of Lewis acid sites present in SiO2 was considered. The concentration of Brönsted and Lewis acid sites was determined in quantitative IR studies of pyridine sorption, according to the protocol described in the

AC C

Supplementary Information.

The transmission electron microscopy (TEM) measurements were performed using a JEOL 2100F electron microscope operating at 200 kV, with Field Emission Gun (FEG), EDX analysis and STEM detectors for bright and dark mode. For selected materials (mesoporous TiO2, and Ag-modified mesoporous TiO2 calcined at 500, 700 and 800°C) measurements were preformed using a Hitachi HF2000 electron microscope operated at 200 kV and equipped with a cold filed emitter. 5

ACCEPTED MANUSCRIPT

The diffuse reflectance UV-vis (UV-vis-DR) spectra of the samples were recorded using a Perkin-Elmer Lambda 7 UV-VIS spectrophotometer. The measurements were performed in the range of 200-900 nm with a resolution of 1 nm. The spectra were recorded under ambient conditions and the data transformed according to the Kubelka-

RI PT

Munk equation. The temperature-programmed reduction (H2-TPR) experiments of the samples (100 mg) were performed using a Quantachrome ChemBET Pulsar TPR/TPD instrument. H2-TPR runs were carried out starting from room

SC

temperature to 1000°C, with a linear heating rate of 10°C/min and in a flow (25 cm3/min) of 5.0 vol.% H2 diluted in Ar. Water vapour was removed from effluent gas by means of a cold trap placed in an ice-water bath. The H2

M AN U

consumption was detected and recorded by a TCD detector.

2.3. Catalytic tests

The catalytic experiments of selective ammonia oxidation (NH3-SCO) were performed under atmospheric pressure in a fixed-bed flow microreactor (i.d., 6 mm; l., 320 mm). The reactant concentrations (NH3 (m/z 16), N2 (m/z 28),

TE D

NO (m/z 30), N2O (m/z 44), NO2 (m/z 46), H2O (m/z 18)) were continuously monitored using a QMS MKS, Cirrus 2 detector directly connected to the reactor outlet using a heated capillary. Prior to the test, the catalyst (100 mg)

EP

was outgassed in situ at 600°C for 1 h in a flow of: (i) pure Ar or (ii) 5.0 vol.% H2 diluted in Ar (20 cm3/min). The composition of the gas mixture at the reactor inlet consisted of [NH3] = 0.5 vol.%, [O2] = 2.5 vol.%, Ar balance.

AC C

The total flow rate of the reaction mixture was 40 cm3/min, while the weight hourly space velocity (WHSV) was about 24,000 cm3/h⋅g. Studies were performed in the temperature range of 100-400°C with a linear heating rate of 5°C/min. The signal of the argon line (m/z = 40) served as the internal standard to compensate small fluctuations of the operating pressure. The sensitivity factors of the analysed lines were calibrated using commercial mixtures of gases. The conversion of ammonia (  ) was determined using the following equation: 6

ACCEPTED MANUSCRIPT

  =

  −   ∙ 100%  

where:   and   – concentration of NH3 in the inlet gas, and concentration of NH3 in the outlet gas.

RI PT

The desired reaction product is nitrogen, while undesired by-products are N2O, NO, and NO2.

equation:   =

  ∙ 100% 1 1        +   +  +   2 2

SC

The selectivity to N2 ( ) was calculated taking into account all possible by-products based on the following

M AN U

where:  ,  , , and   – concentrations of N2, N2O, NO, and NO2, respectively, in the outlet gases. The selectivity towards other possible by-products was obtained in an analogues way.

3. Results and discussion

Figures 1-2 present powder XRD diffraction patterns of the obtained materials. Commercial Al2O3 and TiO2

TE D

showed the characteristic diffraction peaks corresponding to reflections of γ-Al2O3 (2θ at 19.9, 32.6, 37.4, 39.6, 45.9, 61.0 and 67.1° [18]), and anatase TiO2 (2θ at 25.2, 37.8, 48.0, 53.8, 55.1, 62.6, 68.7 70.2 and 75.0° [19]),

EP

respectively. The mesoporous TiO2 revealed diffraction peaks corresponding to reflections of both anatase (2θ at 25.4, 38.0, 48.1, 55.2, 62.8, and 75.2°) and rutile (2θ at 27.5, 36.2, 41.4, 44.4, 64.5, 70.4°), as well as 54.1 and

AC C

69.0° (superposition of reflections characteristic of anatase and rutile) [20]). Moreover, figure S1 (Supplementary Information) shows the pore structure of mesoporous TiO2 characterized by small-angle X-ray scattering (SAXS). Three weak scattering peaks located at 1.4, 2.3 and 3.1° suggested the presence of a lamellar mesophase [21]. However, the acquired weak intensities of the scattering peaks implied that most of lamellar structure collapsed during calcination. For the commercial supports – γ-Al2O3 and (anatase) TiO2 – modified with 1.5 wt.% of silver, no changes in the structure of the oxides were observed after their impregnation with silver and subsequent

7

ACCEPTED MANUSCRIPT

calcination. However, reflections characteristic of Ag2O (2θ at 33.9, 38.0 and 64.4°) and metallic Ag (2θ at 44.3 and 77.3°) [22], appeared in the diffractograms recorded for samples with 10 wt.% of silver. Those reflections exhibited weaker intensity for the Ag-modified mesoporous TiO2, possibly due to partial superposition of

RI PT

reflections characteristic of TiO2 (rutile) and AgOx (2θ at 44.3 and 64.4°). For mesoporous TiO2 modified with 1.5 wt.% of silver, reflection characteristic for rutile had stronger XRD peak intensity than for 10% Ag/mesoTiO2. On the other hand, for samples modified with 10 wt.% of silver, reflections characteristic of rutile were small and

SC

broad. Thus, possibly a higher amount of the AgOx clusters covered the surface of mesoporous TiO2, and prevented the transformation of anatase to rutile for 10% Ag/mesoTiO2. Among three most commonly encountered

M AN U

crystalline polymorphs of titanium oxide: anatase (tetragonal, D4h19), brookite (orthorhombic, D2h15), and rutile (tetragonal, D4h14) [23], the first one is stable at low temperatures. The sintering of the material at above 700°C is accompanied by the transformation of anatase to rutile [24]. However, in case of our mesoporous TiO2, a weak reflection characteristic of the rutile phase (2θ at 27.5°) appeared after calcination at 500°C. The mesoporous

TE D

TiO2 modified with 1.5 wt.% of silver showed well-developed anatase and/or rutile structures after calcination at 500, 700 and 800°C, respectively. Moreover, additional peak related to Ag2O appeared at 38.0° for material

Fig. 1. and Fig. 2

AC C

EP

modified with silver and calcined at temperature as high as 800°C.

Figure S2 presents N2 physisorption isotherms for the Al2O3, TiO2 (anatase) and mesoporous TiO2 supports. The γAl2O3 and mesoporous TiO2 supports present hysteresis loops, which are characteristic of type IV isotherms with H2 hysteresis indicating the existence of mesopores with the ink-bottle effect. On the other hand, the a very narrow loop hysteresis did not appear for at high relative pressure was observed for TiO2 (anatase) which belongs to is a characteristic of a type II a type IV isotherm with H3 hysteresis isotherm for non-porous materials. This type of 8

ACCEPTED MANUSCRIPT

hysteresis indicated a very broad pore size distribution denoting the presence of plate like particles or large macropores which were not yet completely filled with N2 condensate [25]. Figure 3 presents selected N2 physisorption isotherms for Ag-modified materials, while Table 1 lists the textural properties for the obtained

RI PT

materials. Generally, γ-Al2O3 possessed a relatively high specific surface area (134 m2/g) compared to both TiO2 supports with 12 and 46 m2/g for TiO2 (anatase) and mesoporous TiO2, respectively. The resulting specific surface area of γ-Al2O3 decreased to 125 and 109 m2/g following the impregnation with 1.5 and 10 wt.% of silver. For both

SC

materials, the hysteresis loop covered a broad range of relative pressure corresponding to type H2. For Agmodified TiO2 (anatase), the hysteresis represented a type II isotherm for non-porous materials. The addition of Ag

M AN U

did not really alter the final specific surface areas, which remained around 11 m2/g. These low specific surface areas were possibly related to the non-porous feature of TiO2 (anatase). Regarding Ag supported on mesoporous TiO2, the resulting specific surface areas were reduced from 29 to 17 m2/g after impregnation of 1.5 and 10 wt.% Ag, respectively. The shape of hysteresis for both materials corresponded to type H2. Noteworthy, the mesoporous

TE D

TiO2 support exhibited a relatively large specific surface area in comparison to the non-porous anatase TiO2. Furthermore, the materials’ pore size distributions (PSD) were evaluated via the BJH method applied to the desorption branch. Bare (Figure S2) and Ag-doped Al2O3 displayed narrow PSD centred at 4.3 – 4.6 nm. For pure

EP

mesoporous TiO2, the pore size distribution was relatively narrow centered at 8.8 nm, while the Ag-doped mesoporous TiO2 showed slightly broader PSD with maxima at 13 and 16 nm for 1.5 or 10 wt.% of silver,

AC C

respectively. The increase in the pore size distribution for the material with highest silver loading indicated a potential pore wall collapse. Regardless of the Ag loading, PSD of Ag/TiO2 exhibited featureless profiles, suggesting a non-porous characteristic of the materials. For all the supports, the higher doping with Ag resulted in a decrease of the pore volume, indicating that the pore space accommodated more silver particles Ag particles existed inside the pores or blocked entrances of pore channels. Furthermore, the influence of calcination temperatures on the materials’ textural properties was investigated for mesoTiO2 and 1.5 wt.% Ag/mesoTiO2. For 9

ACCEPTED MANUSCRIPT

both materials, higher calcination temperatures led to lower specific surface areas and pore volumes caused by collapse of the mesostructure.

chemisorbed NH3 (  ) of the materials. Vpore [cm3/g] 0.26 0.21 0.19 0.07 0.07 0.03 0.12 0.12 0.08 0.15 0.10 0.10 0.16 0.10 0.08

SC

Dpore [nm] 4.3 4.5 4.6 8.8 13 16 6.2 14 6.4 -

M AN U

Al2O3 1.5% Ag/Al2O3 10% Ag/Al2O3 TiO2 1.5% Ag/TiO2 10% Ag/TiO2 mesoTiO2 1.5% Ag/mesoTiO2 10% Ag/mesoTiO2 mesoTiO2(500) mesoTiO2(700) mesoTiO2(800) 1.5% Ag/mesoTiO2(500) 1.5% Ag/mesoTiO2(700) 1.5% Ag/mesoTiO2(800)

SBET [m2/g] 134 125 109 12 11 10 46 29 17 72 20 12 66 10 7

TE D

Sample codes

RI PT

Table 1. Specific surface area (SBET), average pore diameter (Dpore), total pore volume (Vpore), and concentration of

 [µmol/g] 321 265 167 24 48 46 156 124 72 342 72 23 246 -

EP

Fig. 3. and Table 1.

AC C

Figure 4 presents selected NH3-TPD profiles, while Table 1 lists the values of the chemisorbed ammonia for the obtained materials. NH3 was released over a broad temperature range for both 1.5 and 10 wt.% of silver loading, indicating that different acid strengths exist on the surface of the supports. Surface acidity of various supports declined according to the following order: γ-Al2O3 > mesoTiO2 > TiO2 (anatase). The Al2O3 support exhibited a substantial amount of chemisorbed NH3 as high as 321 µmol/g because of the presence of unsaturated Al3+ acting as strong Lewis acid sites dispersed over a high specific surface area [27]. For titania supports, mesoporous TiO2 with higher specific surface area exhibited higher surface acidity than TiO2 (anatase). Additionally, the larger pore 10

ACCEPTED MANUSCRIPT

opening of mesoporous TiO2 allowed more active sites to be exposed for NH3 chemisorption [28]. The surface acidity dropped from 342 to 23 µmol/g for mesoporous supports calcined at 500 and 800°C, respectively. The Ag/Al2O3 materials with 1.5 and 10 wt.% of silver loading revealed significantly lower amounts of

RI PT

chemisorbed NH3 of 265 and 167 µmol/g, respectively, than the Al2O3 support. A higher Ag-doping depleted the surface acidity because of the formation of higher AgOx coverage on the support surfaces. A similar trend appeared for Ag-doped mesoporous supports. The values significantly decreased after impregnation for materials calcined at

SC

700 and 800°C, thus, the estimation of the amount of the chemisorbed ammonia was impossible. Interestingly, the surface acidity initially increased for TiO2 (anatase) after Ag doping, indicating generation of additional acid sites

M AN U

on the TiO2 (anatase) framework. These results are in agreement with NH3-TPD data reported for TiO2 (anatase) after deposition of transition metals [29,30]. Furthermore, Table S1 (Supporting Information) summarized concentrations and strengths of Brönsted and Lewis acid sites for the obtained materials. Al2O3 together with Agmodified Al2O3 materials showed the highest contribution of Lewis sites. The Al-OH group did not evidence any

TE D

protonic acidity. Among titania supports, mesoporous TiO2 revealed higher Lewis acidity than TiO2 (anatase). Moreover, no Brönsted sites were detected on the surface of mesoporous TiO2, since the Ti-OH groups are unable to protonate probe molecules as basic as pyridine or ammonia and acetonitrile. TiO2 (anatase) was also free of

EP

protonic sites, while the Lewis acidity was strongly reduced. Deposition of 1.5 wt.% of silver on the TiO2 supports preserved or even enhanced the overall acidity; however, the strength of sites was slightly reduced. This

AC C

phenomenon appeared possibly due to the generation of additional acid sites, probably oxidized silver species after deposition of Ag on the TiO2 supports. Contrary to the Ag-doped TiO2 supports, deposition of Ag on Al2O3 did not result in enhanced Lewis acidity; however, acid strength decreased. Interestingly, independently of the support, the presence of 10 wt.% silver caused the reduction of Lewis acidity to the half of the initial amount. Acid sites on alumina and titania surfaces were nearly of the same strength. Thus, the acidity offered by materials modified with

11

ACCEPTED MANUSCRIPT

10 wt.% of Ag originated mainly from oxidized silver species, while the impact of the support surface to overall acidity was marginal. Moreover, sorption of CO evidenced the difference in electron acceptor properties of silver species deposited on

RI PT

supports of different acidity. Figure S4 presents the spectra of CO adsorbed on Ag/Al2O3, Ag/TiO2 and Ag/mesoTiO2 modified with 1.5 wt.% of silver. The Ag sites on Al2O3 were characterized by the lowest electron acceptor properties (the lowest frequency of CO band), while the Ag sites on mesoporous TiO2 were the sites of the

M AN U

Fig. 4.

SC

highest electron acceptor property. Thus, highly dispersed oxidized silver species existed on 1.5% Ag/mesoTiO2.

Figure S3 presents HRTEM images of mesoporous TiO2 calcined at 600°C. A compact aggregation of polydisperse particles ranging between 20 to 30 nm, appeared due to high surface mobility, which significantly increased after exposure at high temperature. Furthermore, the interspaces between aggregated particles led to

TE D

formation of pore channels. Figure 5 shows STEM images of the obtained materials together with EDX element mapping for 1.5 wt.% Ag/TiO2 and 1.5 wt.% of Ag/mesoTiO2. The distribution of silver particles strongly depended on the applied support; however, rather large AgOx particles tended to form on all applied supports.

EP

Thus, the difference in the specific surface areas of γ-Al2O3 or both TiO2 did not significantly impact the dispersion of silver species, as indicated by Zhang et al. [3] or Qu et al. [1]. Moreover, no significant changes were observed

AC C

along with the increase of the Ag loading from 1.5 to 10 wt.%. XRD diffraction patterns revealed high dispersion of Ag species for materials with 1.5 wt.% of silver; the Ag particles were irregular and scarcely dispersed over the supports with particle sizes between 4 to 24 nm. The detailed comparison of AgOx deposited on TiO2 (anatase) and mesoporous TiO2 with 1.5 wt.% of silver loading confirmed a larger particle size accompanied by an inhomogeneous distribution over Ag/mesoTiO2. Not only clusters and larger particles existed on the mesoporous TiO2 support surface but also fine distributed silver particles. The high crystallinity of TiO2 (anatase) possibly 12

ACCEPTED MANUSCRIPT

facilitated the formation of well dispersed AgOx particles. In contrast, the deposition of silver on more or less amorphous mesoporous TiO2 – through wet impregnation and subsequent calcination at 600°C for 6 h – promoted the agglomeration and formation of AgOx clusters. Figure S3 shows HRTEM images of Ag/mesoTiO2 calcined at

RI PT

500 and 800°C. The silver particles size and their distribution significantly varied for the obtained materials. Only a few larger particles (0.6-0.8 nm) appeared on Ag/mesoTiO2(500), while a relatively uniform distribution of silver

decreased for Ag/mesoTiO2(800).

M AN U

Fig. 5.

SC

particles existed on materials calcined at 700°C. The silver particles increased up to 5-10 nm, while their dispersion

Figure 6 shows the UV-vis-DR spectra for the obtained materials. The commercial γ-Al2O revealed its optical transparency (results not shown) [28]. Ag/Al2O3 catalysts with 1.5 and 10 wt.% of silver loading revealed two

TE D

bands at about 220-248 and 354-358 nm. The first band appeared due to the dispersed Ag+ [31,32]. This band shifted to higher wavelength (248 nm) with higher silver loading, indicating a decrease in the dispersion of Ag2O particles. A shoulder at about 354-358 nm corresponded to metallic Agn clusters [33]. The position of the bands

EP

revealed a rather large AgOx particles size, in accordance with STEM and H2-TPR (below) results. The XRD analysis showed that silver species appeared in the form of Ag2O and metallic Ag for Ag/Al2O3 with 10 wt.%

AC C

of silver loading. However, the intensity of the reflections characteristic of Ag2O was significantly higher than that of metallic Ag. Thus, also the intensity of the bands characteristic of Ag0 remained lower than the bands related to Ag+ species. The TiO2 support showed a characteristic strong absorption at 200-340 nm [34], and an absorption threshold at above 400 nm. A band centred at about 235 and 310 nm for TiO2 (anatase) could be assigned to ligand-to-metal charge transfers (LMCT) from O2- to Ti4+ due to the presence of tetrahedrally and octahedrally coordinated Ti4+ [35], respectively. The sample bands were found in the UV-vis-DR spectra for 13

ACCEPTED MANUSCRIPT

materials modified with 1.5 or 10 wt.% of silver. However, for the Ag-modified materials additional bands appeared at about 485 nm due to the presence of metallic silver [3,30]. The mesoporous TiO2 – a mixture of anatase and rutile, consisting of octahedrally coordinated Ti4+ – gave rise to the spectra with absorption bands

RI PT

differing in the positions of 311 and 355 nm, respectively. The same bands appeared in the spectra obtained for Ag-modified materials; however, the contribution of the rutile phase in samples modified with 1.5 wt.% of silver caused a sharp increase toward the second absorption band at about 343 nm. Additionally, the bands

SC

characteristic of Ag0 appeared also in the spectra of Ag/mesoTiO2 at about 540 and 520 nm for samples with 1.5 and 10 wt.% loading of silver, respectively. Our XRD spectra showed that AgOx were dispersed on titania for a

M AN U

silver loading of 1.5 wt.%. In contrast, for materials with higher Ag amount the silver species existed in the form of Ag2O and metallic Ag for both TiO2 (anatase) and mesoporous TiO2. However, confirming the assignment of bands related to Ag+ in Ag-modified titania materials was no possible due to overlapping bands of AgxO species and the TiO2 supports.

TE D

UV-vis-DRS analyses were also applied to investigate the morphological changes of Ag species on the surface of Ag/Al2O3 catalysts due to in situ H2-pretreatemant. Accordingly, the samples after catalytic tests were quickly cooled down in a flow of pure Ar. Figure 7 presents UV-vis-DR spectra of the Ag/Al2O3 catalysts after in situ

EP

pretreatment in H2 and subsequent catalytic tests. Materials with 1.5 and 10 wt.% of silver revealed a broad band centred at about 481-495 nm due to appearance of metallic Agn clusters on the surface of the γ-Al2O3 support

AC C

[3,31]. While the band located at 281-284 nm arose due to oxidized silver clusters Agnδ+ (2 ≤ n ≤ 4) [31,34]. The absorption bands of mesoporous TiO2 overlapped with those of Agnδ+. Thus, the formation of Agnδ+ on the surface of mesoporous TiO2 was unclear, however, it could proceed in an analogues way as for Ag/Al2O3 catalysts. Nevertheless, such hypothesis certainly requires further confirmation via separate in situ studies of the silver oxidation states under catalytic tests conditions.

14

ACCEPTED MANUSCRIPT

Fig. 7.

Figure 8 presents H2-TPR profiles of the obtained materials. The TPR profile for γ-Al2O3 did not show any H2

RI PT

uptake [28]. While the profile obtained for TiO2 (anatase) revealed a limited H2 consumption with a main broad peak centred at about 950°C. Mesoporous TiO2 calcined at 500 up to 800°C possessed more complex H2-TPR profiles. The materials calcined at 500 and 600°C showed relatively high H2 uptakes. The peaks centred at about

SC

365 and 654°C for material calcined at 500°C, and analogues at 355 and 684°C for samples calcined at 600°C. Additionally, for mesoTiO2(600) a small peak appeared at about 575°C. Materials calcined at higher temperatures

M AN U

– 700 and 800°C – presented lower H2 uptakes; however, the reduction took place at similar temperatures (main peak at 681-699°C) as for mesoporous TiO2 calcined at 500-600°C. Additionally, for all materials a broad band appeared above 800°C. Dewan et al. [37] reported the phase changes during H2-TPR of TiO2 at different temperatures as follows: (i) Ti8O15 and unreduced TiO2 (915°C), (ii) Ti4O7 (975°C), (iii) Ti4O7 and Ti3O5

TE D

(1035°C), (iv) Ti3O5 and Ti2O3 (1165°C). Although, the differences in the H2-TPR profiles of TiO2 (anatase) and mesoporous TiO2 showed that the reduction of titania could proceed in different ways, the four reduction steps (iiv) seem to be present in the H2-TPR profiles of mesoporous TiO2. Thus, we assumed that the main reduction peak

EP

at around 654-699°C appeared possibly due to reduction of titania into a mixture of Ti4O7 and Ti3O5. The H2-TPR

AC C

profiles changed after deposition of 1.5 wt.% of silver. Moreover, the supports calcined at different temperatures significantly influenced formation of silver species with different reducibility. Again, the materials calcined at 500 or 600°C showed similar H2-TPR profiles, with reduction of Ag+ species at 124 and 119°C, respectively. While the peak belonging to the mixture of Ti4O7 and Ti3O5 split into two maxima indicating partial reduction at lower temperatures (530-534 and 664-755°C). Interestingly, the support calcined at 700°C revealed a significantly higher reduction of oxidized silver species centred at 134°C than the other materials calcined at 500-800°C. For 1.5% Ag/mesoTiO2(800) the silver species – mainly in the form of Ag2O (according to XRD analysis) – was reduced 15

ACCEPTED MANUSCRIPT

around 144°C. The H2-TPR profiles varied also for different supports (γ-Al2O3, TiO2 (anatase), mesoporous TiO2), and changed with different silver loading (1.5 or 10 wt.%). The low-temperature peak appeared in the H2-TPR profiles of Ag-modified samples, indicating reduction of Ag+ species. Jabłońska et al. [38] showed two peaks (at

RI PT

110 and 385°C) in the H2-TPR of analogues Ag/Al2O3 samples with 1 wt.% of silver loading. While our H2-TPR profile of 1.5% Ag/Al2O3 showed a main peak located around 123°C, indicating bulk Ag2O. Both Ag-modified TiO2 (anatase) and mesoporous TiO2 possessed a lower H2 uptake than Ag/Al2O3 samples. Thus, the silver species

SC

deposited on more acidic γ-Al2O3 oxidized significantly easier than those deposited on both titania supports [2]. However, the reduction temperature of silver species deposited on the mesoporous support remained significantly

M AN U

lower compared to the commercial TiO2 (119 versus 128°C). Further comparison of the materials with 10 wt.% of silver loading revealed also significant difference of the H2 uptakes of the different supports. γ-Al2O3 and mesoporous TiO2 facilitated formation of Ag+ species. While TiO2 (anatase) enhanced the formation of Ag0. These results are in agreement with our XRD data. The H2-TPR profile of Ag/Al2O3 showed a peak at 119°C with a

TE D

shoulder at 150°C. The oxidized silver species deposited on both titania supports possessed significantly higher reduction temperatures. For Ag/TiO2 the main reduction peak appeared at 140°C with a small shoulder at about 103°C. The Ag/mesoTiO2 revealed the main peak at 133°C together with a shoulder centred at 220 and a broad

EP

peak at 400°C. The low-temperature peaks appeared due to reduction of dispersed Ag2O [39,40], while the

AC C

shoulders and high temperature peaks (>200°C) arose due to reduction of large AgOx clusters [40]. Thus, AgOx clusters were formed preferentially on mesoporous TiO2, in good agreement with our STEM/EDX analysis. Moreover, the reduction of titania caused a peak located at 547°C.

Fig. 8.

16

ACCEPTED MANUSCRIPT

Figures 9-10 present the results of the catalytic tests for the obtained materials, while Table 2 gathers the corresponding data together with the data of the catalytic activities over (10 wt.%)Ag/A2O3 and (10 wt.%)Ag/TiO2 presented in literature. If not provided, we roughly estimated the catalytic performance based on NH3 conversions-

RI PT

temperature profiles. During reaction, no other N-containing compounds besides N2, NO and N2O were detected in the tested temperature range. However, a big difference in selectivity occurred for the tested catalysts. Ammonia oxidation over Ag-based catalysts with different supports revealed a significant contribution of N2O. Figure 9

SC

presents the results of catalytic tests for materials pretreated in pure Ar. Silver deposited on TiO2 (anatase) did not facilitate significant catalytic performances (96% conversion with 66% of N2 selectivity at 400°C). In addition, the

M AN U

catalyst with 10 wt.% silver loading exhibited an even lower activity, while N2 selectivity significantly increased possibly due to the presence of AgOx clusters with lower reducibility. Ag-modified mesoporous TiO2 enabled a superior catalytic performance; 1.5 wt.% Ag on mesoporous TiO2 facilitated full ammonia conversion at 375°C with 74% N2 selectivity at this temperature. A potential explanation relates to higher activity of silver species

TE D

supported on easily reducible mesoporous TiO2 instead of TiO2 (anatase). The effect of calcination temperature (500, 700, 800°C) on catalytic performance in NH3-SCO was also investigated for mesoporous TiO2 with 1.5 wt.% loading of silver. Surprisingly, Ag catalysts utilizing these supports reached significantly lower catalytic activity at

EP

400°C (6, 52 and 7% for catalysts calcined at 500, 700 and 800°C, respectively). Both XRD and H2-TPR analysis revealed that different calcination temperatures possessed a strong effect on the surface composition as well as on

AC C

the state of the silver species on the support surface. Thus, silver deposited on a support containing crystalline phases of both anatase and rutile – 1.5% Ag/mesoTiO2(600) – led to optimum catalytic performance for NH3-SCO. Similar to 10% Ag/TiO2, N2 selectivity for Ag/mesoTiO2 increased, while activity remained nearly unaltered upon increasing silver loading. Consequently, large AgOx clusters inhibited the catalytic activity, and enhanced N2 selectivity. Oxidized silver species with higher reducibility enhanced the activity of Ag/mesoTiO2 compared to Ag/TiO2 (with 10 wt.% of silver loading), but led to lower N2 selectivity above 350°C. Overall, the comparison 17

ACCEPTED MANUSCRIPT

between titania supports allows drawing the conclusion that surface composition, surface acidity and reducibility significantly influence catalytic activity and selectivity, respectively. As evidenced by XRD analysis, the crystalline structure of mesoporous TiO2 appeared to be a mixture of anatase and rutile. H2-TPR analysis

RI PT

confirmed that easily reducible support and oxidized silver species of Ag/mesoTiO2 led to enhanced activity together with a drop in N2 selectivity at higher temperatures. In contrast, materials obtained after in situ pretreatment in H2 possessed a significantly different catalytic behaviour.

SC

In situ H2-pretreated Ag/Al2O3 catalysts were tested for comparative purposes with literature. Qu et al. [1] found full conversion with a N2 selectivity of 89% at 180°C over H2-pretreated (10 wt.%)Ag/Al2O3. Analogous catalysts

M AN U

based on TiO2 showed a clearly lower catalytic performance (full conversion with 65% of N2 selectivity at 260°C). The authors concluded that smaller Ag0 particle size significantly enhanced activity of Ag/Al2O3 in NH3-SCO. Ag+ and Agnδ+ species improved N2 selectivity for temperatures above 140°C by promotion NO (intermediate in i-SCR mechanism) reduction by NH3. In general, H2-pretreated catalysts reached higher activity at lower temperatures

TE D

(<200°C) compared to fresh catalysts. For H2-pretreated catalysts, the selectivity to N2 remained higher in the studied temperature range (140-220°C). Thus, gathered results clearly indicated that H2-pretreatment before reaction influenced NH3-SCO. Gang et al. [6] found a similar catalytic performance with full conversion and 82%

EP

N2 selectivity at 160°C for (10 wt.%)Ag/Al2O3 after H2-pretreatment at 200°C (without any pre-calcination). As a final point, Zhang et al. [3] studied H2-pretreated (10 wt.%)Ag/Al2O3 prepared by different preparation methods:

AC C

impregnation, incipient wetness impregnation, and sol-gel. The H2-pretreated catalysts (obtained through impregnation) facilitated significantly higher activity than samples without pretreatment. However, the authors did not present or discuss changes in the selectivities over both materials. Although, Ag-modified Al2O3 catalysts have been studied comprehensively in ammonia oxidation, a general understanding of the active sites and the role of the H2-pretreatment is still missing. Zhang et al. [3] claimed Ag0 as the major active species of H2-pretreated Ag/Al2O3 (<140°C), while Ag+ presents a potential active species above 140°C. Noteworthy, although Agnδ+ was evidenced 18

ACCEPTED MANUSCRIPT by UV-vis-DRS analysis in addition to Ag0 and Ag+, the authors did not specify the role of the different silver sites in ammonia oxidation. Yu et al. [41] and More et al. [42] stressed the formation of more Agnδ+ clusters for H2pretreated Ag-modified Al2O3. The Ag+ species could convert to Agnδ+ clusters in the presence of H2 and O2 [43].

RI PT

In the present study, the H2-pretreatmend caused the formation of Ag0. Subsequently, O2 of the reaction feed could be adsorbed dissociatively accompanied by the formation of Ag-OH species. Finally, in such surface sites a transfer of charge from Ag to OH proceeds by back donation from OH to Ag (Ag+) with formation of Agnδ+

SC

clusters [40,41]. UV-vis-DRS analysis of Ag/Al2O3 (after catalytic tests) emphasised the formation of Agnδ+ clusters. As stated before, we were not able to confirm the formation of such species over Ag-doped mesoporous

M AN U

titania due to overlapping absorption bands of silver species and titania. However, the formation of Agnδ+ clusters over Ag/mesoTiO2 catalysts appears possible. Noteworthy, we did not obtain as high activity for Ag/Al2O3 catalysts for in situ H2-pretreated materials as reported in literature (full conversion of ammonia from the reaction mixture at about 325 and 225°C for materials with 1.5 and 10 wt.% of silver loading). The observed differences are

TE D

potentially related to different preparation methods as well as reaction conditions. Remarkably, as presented in figure 11, we found higher activity over in situ H2-pretreated Ag/Al2O3 catalysts than for Ag/mesoTiO2 or Ag/TiO2. Ag/TiO2 showed still poor catalytic performances with 98 and 99% of ammonia conversion and 87 and

EP

82% N2 selectivity at 400°C for materials with 1.5 and 10 wt.% of silver loading, respectively. The silver modified mesoporous TiO2 – (1.5 wt.%)Ag/mesoTiO2 – facilitated higher selectivity to N2 than the other catalysts. In

AC C

comparison to the formation of N2 and N2O, NO was only formed in minor amount on this catalyst. Increasing silver loading facilitated a higher conversion; however, 10% Ag/mesoTiO2 possessed a significantly lower N2 selectivity with full conversion at 275°C and 78% N2 selectivity. Interestingly, Ag/mesoTiO2 with 1.5 or 10 wt.% silver loading displayed distinct differences with respect to their tendency toward N2O formation. 10% Ag/mesoTiO2 caused a higher selectivity to N2O with above 25% up to 250°C, which afterwards progressively decreased. We assume that Ag0 catalyzed the decomposition of N2O into N2 and surface oxygen species at higher 19

ACCEPTED MANUSCRIPT

temperatures [6,44,45]. Gang et al. [6,46] reported that adsorbed oxygen species appeared on the reduced Ag-based catalysts. They suggested the dissociation of O2 as rate-controlling step for NH3 oxidation. Adsorbed NH3 is activated to form -NH intermediates (Eq. 1) in the presence of oxygen atoms over Ag0. Subsequently, -NH species

RI PT

interact with oxygen species accompanied by the formation of NO (intermediate). Finally, NHx species reduce NO into N2 or N2O, according to an internal selective catalytic reduction (i-SCR) mechanism (Eq. 2-4) [7,45]. The NO

NH + O → NO + OH

(Eq. 2)

NH2 + NO → N2 + H2O

(Eq. 3)

NH + NO → N2O + H

(Eq. 4)

H + OH → H2O

(Eq. 5)

M AN U

(Eq. 1)

TE D

NH3 + O → NH + H2O

SC

reduction to N2 proceeds preferentially over oxidized silver species (Ag+ and Agnδ+) [1].

The selectivity to N2 over Ag/mesoTiO2 catalysts was governed by different surface compositions (anatase and/or rutile). Note, that we observed a similar trend of catalytic performances for Ag-supported on TiO2 (anatase) with

EP

1.5 or 10% of silver. In addition, a similar trend but with higher catalytic performances was reached over 10% Ag/mesoTiO2 with strong contribution of the anatase phase compared to rutile (based on XRD analysis). Important

AC C

to note, a higher catalytic performance was obtained over 1.5% Ag/mesoTiO2, indicating that a proper mixture of anatase and rutile let to enhanced performances in NH3-SCO.

Fig. 9-11., Table 2.

20

ACCEPTED MANUSCRIPT

Table 2. Overview of literature data related to preparation methods, pretreatment conditions and catalytic performances in NH3-SCO of Ag-modified Al2O3 or TiO2.

Ag/Al2O3 Ag/TiO2

Silver loading/Preparation method/ Pretreatment conditions 1.5 wt.% 10 wt.% 1.5 wt.% 10 wt.% 1.5 wt.%

Ref. this work

SC

Ag/mesoTiO2

Conversion/Selectivity/Temperature [%/%/°C] *100/78/325 *100/94/225 96/66/400 *98/87/400 90/93/400 *99/82/400 100/74/375 *100/87/350 100/87/350 *100/78/275

RI PT

Catalyst codes

100/89/180 100/65/260 100/82/160

[1] [6] [3]

100/36/150 100/54/150 100/96/300

EP

4. Conclusions

M AN U

Ag/Al2O3

TE D

Ag/Al2O3 Ag/TiO2 Ag/Al2O3

10 wt.% /wet impregnation/600°C, 6 h, static air 600°C, 1 h, Ar *600°C, 1 h, H2/Ar 10 wt.%/wet impregnation/ 600°C, 3 h, static air; 300°C, 2 h, H2/N2 10 wt.%/wet impregnation/ 200°C, 2 h, H2/He 10 wt.% impregnation/600°C, 3 h, static air incipient wetness impregnation/600°C, 3 h , static air sol-gel/600°C, 6 h, static air 400°C, 2 h, H2/N2

In this study, we prepared mesoporous TiO2 by evaporation induced self-assembly (EISA) using different silver

AC C

loadings (1.5 or 10 wt.%). Silver supported on commercial TiO2 (anatase) and γ-Al2O3 were applied as reference materials. We could emphasise that surface composition, surface acidity and reducibility significantly influenced the catalytic performances of supported catalysts in NH3-SCO. Overall, we found an optimum performance (full conversion with 75% at 375°C) for Ag-doped mesoporous TiO2 with 1.5 wt.% silver loading. Thus, mesoporous TiO2 with the predominant anatase phase, as a support guaranteed formation of easily reducible highly dispersed oxidized silver species. Furthermore, the activity and N2 selectivity improved after in situ H2-pretreatment (full 21

ACCEPTED MANUSCRIPT conversion with 87% N2 selectivity at 350°C) due to formation of Ag0 active in ammonia oxidation and N2O decomposition. Thus, proposed approach is essential for designing active and selective catalytic materials preventing N2O formation in diesel exhaust aftertreatment systems. Future studies are underway to understand the

RI PT

effect of thermal stability tests on the stability of Ag/mesoTiO2.

SC

Acknowledgement

Funded by the Excellence Initiative of the German federal and state governments in the frame of the Center for

M AN U

Automotive Catalytic Systems Aachen (ACA) at RWTH Aachen University. The support of Indonesian government through Indonesian Ministry of National Education (DGHE-DIKTI scholarship) together with European Regional Development Fund (ERDF) and the state of NorthRhine Westphalia, Germany, under the operational program ‘‘Regional Competitiveness and Employment” Project ‘‘Sustainable Chemical Synthesis” for

TE D

the fellowship of W. Ciptonugroho are also acknowledged. The authors thank Dr. Vladimir Girman (Pavol Jozef Safarik University of Kosice, Slovakia) for his collaboration in STEM experiments.

References

Z. Qu, H. Wang, S. Wang, H. Cheng, Y. Qin, Z. Wang, Appl. Surf. Sci. 316 (2014) 373–379.

[2]

K. Gora-Marek, K.A. Tarach, Z. Piwowarska, M. Laniecki, L. Chmielarz, Catal. Sci. Technol. 6 (2016)

AC C

EP

[1]

1651–1660. [3]

L. Zhang, C. Zhang, H. He, J. Catal. 261 (2009) 101–109.

[4]

N.I. ILCHENKO, G.I. Golodets, I.M. Avilova, Kinet. Catal. 16 (1975) 1264–1268.

[5]

N.I. Il’chenko, Russ. Chem. Rev. 45 (1976) 1119–1134.

[6]

L. Gang, B.G. Anderson, J. van Grondelle, R.A. van Santen, Appl. Catal. B Environ. 40 (2003) 101–110.

[7]

L. Zhang, H. He, J. Catal. 268 (2009) 18–25. 22

ACCEPTED MANUSCRIPT

W. Yue, R. Zhang, N. Liu, B. Chen, Chinese Sci. Bull. 59 (2014) 3980–3986.

[9]

X. Cui, J. Zhou, Z. Ye, H. Chen, L. Li, M. Ruan, J. Shi, J. Catal. 270 (2010) 310–317.

[10]

Q. Zhang, X. Liu, P. Ning, Z. Song, H. Li, J. Gu, Catal. Sci. Technol. 5 (2015) 2260–2269.

[11]

H. Zhang, C. Tang, C. Sun, L. Qi, F. Gao, L. Dong, Y. Chen, Microporous Mesoporous Mater. 151 (2012)

RI PT

[8]

44–55.

B.C.J. Brinker, Y. Lu, A. Sellinger, H. Fan, Adv. Mater. 11 (1999) 579–585.

[13]

E.L. Crepaldi, G.J. de A.A. Soler-Illia, D. Grosso, F. Cagnol, F. Ribot, C. Sanchez, J. Am. Chem. Soc. 125

SC

[12]

(2003) 9770–9786.

G.J.D.A.A. Soler-Illia, A. Louis, C. Sanchez, Chem. Mater. 14 (2002) 750–759.

[15]

X. Chen, T. Yu, X. Fan, H. Zhang, Z. Li, J. Ye, Z. Zou, Appl. Surf. Sci. 253 (2007) 8500–8506.

[16]

N. Pal, A. Bhaumik, Dalt. Trans. 41 (2012) 9161–9169.

[17]

M. Jabłońska, A. Król, E. Kukulska-Zajac, K. Tarach, L. Chmielarz, K. Góra-Marek, J. Catal. 316 (2014)

M AN U

[14]

TE D

36–46. [18]

H. Wan, D. Li, Y. Dai, Y. Hu, B. Liu, L. Dong, J. Mol. Catal. A Chem. 332 (2010) 32–44.

[19]

M. Hussain, M. Ahmad, A. Nisar, H. Sun, S. Karim, M. Khan, S.D. Khan, M. Iqbal, S.Z. Hussain, New J.

EP

Chem. 38 (2014) 1424–1432.

J. Yan, G. Wu, N. Guan, L. Li, Z. Li, X. Cao, Phys. Chem. Chem. Phys. 15 (2013) 10978–10988.

[21]

(n.d.).

[22]

G. Wang, X. Ma, B. Huang, H. Cheng, Z. Wang, J. Zhan, X. Qin, X. Zhang, Y. Dai, J. Mater. Chem. 22

AC C

[20]

(2012) 21189–21194. [23]

J.-G. Li, T. Ishigaki, X. Sun, J. Phys. Chem. C 111 (2007) 4969–4976.

[24]

S. Matsuda, A. Kato, Appl. Catal. 8 (1983) 149–165.

[25]

M. Thommes, K. Kaneko, A. V Neimark, J.P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol, K.S.W. Sing, Pure Appl. Chem. 87 (2015) 1051–1069. 23

ACCEPTED MANUSCRIPT

[26]

M. Che, J.C. Védrine, Characterization of Solid Materials and Heterogeneous Catalysts: From Structure to Surface Reactivity, John Wiley & Sons, 2012. H. Bai, N. Cochet, A. Drelich, A. Pauss, E. Lamy, RSC Adv. 6 (2016) 14602–14614.

[28]

Chem. Pap. 69 (2015) 1141.

[29]

D.A. Peña, B.S. Uphade, P.G. Smirniotis, J. Catal. 221 (2004) 421–431.

[30]

H.-J. Choi, J.-S. Kim, M.-S. Kang, Bull. Korean Chem. Soc. 28 (2007) 581–588.

[31]

J. Lu, J.J. Bravo-Suárez, M. Haruta, S.T. Oyama, Appl. Catal. A Gen. 302 (2006) 283–295.

[32]

N. Bogdanchikova, F.C. Meunier, M. Avalos-Borja, J.P. Breen, A. Pestryakov, Appl. Catal. B Environ. 36

SC

RI PT

[27]

M AN U

(2002) 287–297.

M.K. Kim, P.S. Kim, J.H. Baik, I.-S. Nam, B.K. Cho, S.H. Oh, Appl. Catal. B Environ. 105 (2011) 1–14.

[34]

K.M. Parida, N. Sahu, P. Mohapatra, M.S. Scurrell, J. Mol. Catal. A Chem. 319 (2010) 92–97.

[35]

S. Budhi, C.-M. Wu, D. Zhao, R.T. Koodali, Catalysts 5 (2015) 1603–1621.

[36]

J. Shibata, K. Shimizu, Y. Takada, A. Shichi, H. Yoshida, S. Satokawa, A. Satsuma, T. Hattori, J. Catal. 227

TE D

[33]

(2004) 367–374.

M.A.R. Dewan, G. Zhang, O. Ostrovski, Metall. Mater. Trans. B 40 (2009) 62–69.

[38]

M. Jabłońska, M. Nocuń, E. Bidzińska, Catal. Letters 146 (2016) 937–944.

[39]

M. Luo, X. Yuan, X. Zheng, Appl. Catal. A Gen. 175 (1998) 121–129.

[40]

A. Musi, P. Massiani, D. Brouri, J.-M. Trichard, P. Da Costa, Catal. Letters 128 (2009) 25–30.

[41]

L. Yu, Q. Zhong, S. Zhang, Phys. Chem. Chem. Phys. 16 (2014) 12560–12566.

[42]

P.M. More, D.L. Nguyen, P. Granger, C. Dujardin, M.K. Dongare, S.B. Umbarkar, Appl. Catal. B Environ.

AC C

EP

[37]

174–175 (2015) 145–156. [43]

P.S. Kim, M.K. Kim, B.K. Cho, I.-S. Nam, S.H. Oh, J. Catal. 301 (2013) 65–76.

[44]

V.A. Kondratenko, U. Bentrup, M. Richter, T.W. Hansen, E. V Kondratenko, Appl. Catal. B Environ. 84 (2008) 497–504. 24

ACCEPTED MANUSCRIPT

R. Burch, S.T. Daniells, J.P. Breen, P. Hu, Catal. Letters 94 (2004) 103–108.

[46]

L. Gang, B.G. Anderson, J. van Grondelle, R.A. van Santen, J. Catal. 199 (2001) 107–114.

[47]

W. Chen, Y. Ma, Z. Qu, Q. Liu, W. Huang, X. Hu, N. Yan, Environ. Sci. Technol. 48 (2014) 12199–12205.

AC C

EP

TE D

M AN U

SC

RI PT

[45]

25

Intensity [a.u.]

200

A

A1

ACCEPTED MANUSCRIPT

150

G A1

A2

G

G

G

G

10% Ag/Al2O 3 A2

G A1

100

1.5% Ag/Al 2O 3

RI PT

50

Al 2O 3

0 10

20

2000

30

50

60

70

80

90

B

A1 T TT

T

10% Ag/TiO 2

TT

T T A1

TT T

TT

T

M AN U

A2

SC

T

1500

Inten sity [a.u.]

40

1000

1.5% Ag/TiO 2

500

0 20

30

40

TE D

10

50

60

70

TiO 2

80

90

C

T

350

EP R

250

T R

A1

T T+R

T

10% Ag/mesoTiO2

T+R

T

R R

R

T A 2

T

T

200

AC C

Intensity [a.u.]

300

150 100

1.5% Ag/m esoTiO2

50

mesoTiO 2

0 10

20

30

40

50

60

70

80

90

0



Fig. 1. XRD patterns of γ-Al2O3 (A), TiO2 (anatase) (B), mesoporous TiO2 (C) modified with 1.5 wt.% or 10 wt.% of silver; G – γ-Al2O3, A1 – Ag2O, A2 – metallic Ag, T – anatase TiO2, R – rutile TiO2.

200

A

ACCEPTED MANUSCRIPT R

R

150

R R RR

R

R

R R

m esoTiO2(800) RR

100 mesoTiO 2(700)

50

T

T

T

R

m esoTiO2(500)

T

T

T

T

0 10

20

30

40

50

60

70

T

80

90

B

SC

R

400 R

1.5% Ag/mesoTiO2(800)

R

300

R

R R

M AN U

Intensity [a.u.]

RI PT

Intensity [a.u.]

R

AR

R

R

RR

1

200

RR

1.5% Ag/m esoTiO2(700)

100

T

1.5% Ag/mesoTiO2(500)

T

R

T

T

T

T

10

TE D

0 20

30

40

50

60

70

T

T

80

90

0



AC C

anatase TiO2, R – rutile TiO2.

EP

Fig. 2. XRD patterns of meso/TiO2 (A) and 1.5% Ag/mesoTiO2 (B) calcined at 500 up to 800°C; A1 – Ag2O, T –

250

A

1.5% Ag/Al2O3

10% Ag/Al2O3

1.5% Ag/TiO2

10% Ag/TiO2

1.5% Ag/mesoTiO2

10% Ag/mesoTiO2

150

100

50

0 0.2

0.4

0.6

0.8

1.0

0.2

0.4

p/p°

RI PT

3

Adsorbed N2 [cm /g]

200

B

ACCEPTED MANUSCRIPT

0.6

0.8

1.0

SC

p/p° D

10

TE D

3

-1

-1

dV/dlogDpore [cm nm g ]

M AN U

C

20

30

EP

Dpore [nm]

40

50

10

20

30

40

50

Dpore [nm]

Fig. 3. N2 physisorption isotherms (A,B) and pore size distribution (C,D) of Ag/Al2O3, Ag/TiO2 and Ag/mesoTiO2

AC C

modified with 1.5 wt.% or 10 wt.% of silver.

0,0004

1.5% Ag/Al 2O3

A

10% Ag/Al2O3

B

1.5% Ag/mesoTiO 2

10% Ag/mesoTiO2

1.5% Ag/TiO 2

10% Ag/TiO 2

ACCEPTED MANUSCRIPT

150

Intensity [a.u.]

0,0003

147

0,0002

145

0,0001 137

0,0000 100

200

300

400

500

0

200

300

400

500

0

Temperature [ C]

SC

Temperature [ C]

100

RI PT

121

136

Fig. 4. NH3-TPD profiles of Ag/Al2O3, Ag/TiO2 and Ag/mesoTiO2 modified with 1.5 wt.% (A) or 10 wt.% (B) of

M AN U

silver; experimental conditions: mass of catalyst = 100 mg, sorption: [NH3] = 1.0 vol.%, [Ar] = 99.0 vol.%; desorption:

AC C

EP

TE D

[Ar] = 100.0 vol. %; total flow rate = 20 cm3/min, linear heating of 5°C/min.

ACCEPTED MANUSCRIPT

1.5% Ag/Al2O3

C

10% Ag/TiO2

10% Ag/Al2O3

B

AC C

EP

TE D

M AN U

SC

RI PT

A A

D

10% Ag/mesoTiO2

1.5% Ag/TiO2

E

RI PT

ACCEPTED MANUSCRIPT

O

SC

Ti

M AN U

Ag

F

AC C

EP

TE D

1.5% Ag/mesoTiO2

Ag

Ti

O

Fig. 5. STEM images and EDX analysis of Ag/Al2O3 (A,B), Ag/TiO2 (C,E) and Ag/mesoTiO2 (D,F).

A

ACCEPTED MANUSCRIPT

248

3,0

2,5

354

2,0 1,5 10% Ag/Al2O3

1,0 0,5

220 358

0,0 200

300

1.5% Ag/Al 2O3

400

500

600

700

800

900

B

310

60

SC

237

50

10% Ag/TiO2

485

40

310

M AN U

K ubelka-M unk function [a.u.]

70

235

30

485

20

310 235

10 0

25 228

20

300

400

1.5% Ag/TiO 2

500

TE D

200

600

700

TiO2

800

900

C

311 355

EP

520

232

540

10

228

10% Ag/mesoTiO 2

343

15

AC C

K ubelka-M unk function [a.u.]

RI PT

K ubelka-M unk fun ction [a.u.]

3,5

1.5% Ag/mesoTiO 2 311

355

5

mesoTiO 2

0

200

300

400

500

600

700

Wavelength [nm]

Fig. 6. UV-vis-DRS of Ag/Al2O3 (A), Ag/TiO2 (B), Ag/mesoTiO2 (C).

800

900

ACCEPTED MANUSCRIPT

A

481

15 10% Ag/Al2O3 281

10

495

1.5% Ag/Al2O3

5 284

0 200

300

400

500

600

Wavelength [nm]

700

RI PT

K ubelka-M unk function [a.u.]

20

800

900

AC C

EP

TE D

M AN U

SC

Fig. 7. UV-vis-DRS of Ag/Al2O3 after in situ pretreatment in 5.0 vol.% H2 diluted in Ar and catalytic tests (A).

ACCEPTED MANUSCRIPT

60

A

123

950

551

302

681

40 mesoTiO2(700)

775

389 684

20

mesoTiO2(600)

575 355

0

mesoTiO2(500)

200

365

20

1.5% Ag/Al2O3

1.5% Ag/TiO2

119

534 755

1.5% Ag/mesoTiO2

0

400

600

800

1000

B

60 144

200

400

600

800

200

1000

D

TE D

119

1.5% Ag/mesoTiO2(800)

134

40

EP

1.5% Ag/mesoTiO2(700) 119

1.5% Ag/mesoTiO2(600)

534

20

530

124

755

AC C

H2 c ons um pti on [ a.u. ]

128

M AN U

654

40

RI PT

699

mesoTiO2(800)

SC

H2 c ons um pti on [ a.u. ]

TiO2

664

0 200

400

600

Temperature [0C]

H2 c on sum pti on [a.u .]

H2 c ons um pti on [ a.u. ]

60

C

10% Ag/Al2O3

100 140 103

10% Ag/TiO2 133

50

1.5% Ag/mesoTiO2(500)

800

150

150

220 400

547

10% Ag/mesoTiO2

0 1000

200

400

600

800

1000

Temperature [0C]

Fig. 8. H2-TPR profiles of TiO2 and meso/TiO2 (A), Ag/mesoTiO2 (B) calcined at 500 up to 800°C, and Ag/Al2O3, Ag/TiO2 and Ag/mesoTiO2 modified with 1.5 wt.% (C) or 10 wt.% (D) of silver; experimental conditions: mass of the catalysts = 100 mg; [H2] = 5.0 vol.%, Ar balance, flow rate = 25 cm3/min, linear heating of 10°C/min.

100 ACCEPTED MANUSCRIPT C

1 .5% Ag/TiO2

75

NH3 conversion N2 selectivity NO selectivity N2O selectivity

50

25

25

150

200

250

300

350

400

100 100

C onversion/S electivity [% ]

B 1 .5% Ag/mesoTiO2

75

50

25

0 100

150

D 1 0% Ag/mesoTiO2

75

50

M AN U

C onversion/S electivity [% ]

50

0

100 100

75

200

250

300

350

400

300

350

400

SC

0

1 0% Ag/TiO2

RI PT

A

C onversion/S electivity [% ]

C onversion/S electivity [% ]

100

25

0

150

200

250

300

350

0

TE D

Temperature [ C]

400

100

150

200

250 0

Temperature [ C]

Fig. 9. Results of catalytic tests for NH3-SCO performed over Ag/TiO2 (A,C), Ag/mesoTiO2 (B,D); experimental conditions: pretreatment in pure Ar, total flow rate = 20 cm3/min, linear heating of 10°C/min; mass of catalyst = 100

AC C

EP

mg, [NH3] = 0.5 vol.%, [O2] = 2.5 vol.%, Ar balance, total flow rate = 40 cm3/min, linear heating of 5°C/min.

100 ACCEPTED MANUSCRIPT C

1 .5% Ag/TiO2

75

NH3 conversion N2 selectivity NO selectivity N2O sele ctivity

50

25

50

25

0

100

150

200

250

300

350

400

100 100

B C onversion/S electivity [% ]

100

1 .5% Ag/meso TiO2

75

50

25

0 100

150

D 1 0% Ag/mesoTiO2

75

50

M AN U

C onversion/S electivity [% ]

75

200

250

300

350

400

300

350

400

SC

0

1 0% Ag/TiO2

RI PT

A

C onversion/S electivity [% ]

C onversion/S electivity [% ]

100

25

0

150

200

250

300

350

0

100

150

200

250 0

Temperature [ C]

TE D

Temperature [ C]

400

Fig. 10. Results of catalytic tests for NH3-SCO performed over Ag/TiO2 (A,C), Ag/mesoTiO2 (B,D); experimental

EP

conditions: pretreatment in 5.0 vol.% H2 diluted in Ar, total flow rate = 20 cm3/min, linear heating of 10°C/min; mass of catalyst = 100 mg, [NH3] = 0.5 vol.%, [O2] = 2.5 vol.%, Ar balance, total flow rate = 40 cm3/min, linear

AC C

heating of 5°C/min.

ACCEPTED MANUSCRIPT 100

A

50

25

0 100

200

250

300

B

400

M AN U

75

C onver sion [% ]

350

SC

100

150

RI PT

C onversion [% ]

75

50

1.5% Ag/Al2O3 10% Ag/Al2O3

25

1.5% Ag/TiO2 10% Ag/TiO 2 1.5% Ag/meso TiO2 10% Ag/mesoTiO2

0 150

200

TE D

100

250

300

350

400

0

Temperature [ C]

Fig. 11. Comparison of results of catalytic tests for NH3-SCO performed over Ag/Al2O3, Ag/TiO2, Ag/mesoTiO2

AC C

EP

after pretreatment in pure Ar (A), and 5.0 vol.% H2 diluted in Ar (B).

ACCEPTED MANUSCRIPT

♦ Formation of mesoporous TiO2 support by evaporation induced self-assembly (EISA). ♦ High catalytic activity and N2 selectivity in NH3-SCO over Ag-doped mesoporous TiO2. ♦ Full conversion with 74% N2 selectivity at 375°C over 1.5 wt.% Ag/mesoTiO2.

AC C

EP

TE D

M AN U

SC

RI PT

♦ Boosting effect of in situ H2-pretreatmnet on N2O decomposition.