Download as pdf or txt
Download as pdf or txt
You are on page 1of 32

Accepted Manuscript

Comparative performance of copper and iron functionalized


hydroxyapatite catalysts in NH3-SCR

Sebastiano Campisi, Melissa Greta Galloni, Filippo Bossola,


Antonella Gervasini

PII: S1566-7367(19)30043-3
DOI: https://doi.org/10.1016/j.catcom.2019.02.008
Reference: CATCOM 5619
To appear in: Catalysis Communications
Received date: 22 November 2018
Revised date: 8 February 2019
Accepted date: 9 February 2019

Please cite this article as: S. Campisi, M.G. Galloni, F. Bossola, et al., Comparative
performance of copper and iron functionalized hydroxyapatite catalysts in NH3-SCR,
Catalysis Communications, https://doi.org/10.1016/j.catcom.2019.02.008

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.
ACCEPTED MANUSCRIPT

Comparative performance of copper and iron functionalized


hydroxyapatite catalysts in NH3-SCR

PT
Sebastiano Campisi, a Melissa Greta Galloni,a Filippo Bossolab and Antonella
Gervasini*a,b

RI
a
SC
Dipartimento di Chimica, Università degli Studi di Milano, via Camillo Golgi 19, 20133 Milano,
NU
Italy. E-mail: antonella.gervasini@unimi.it
b
CNR (Consiglio Nazionale delle Ricerche) - ISTM (Istituto di Scienze e Tecnologie Molecolari),
MA

Via Venezian 21, 20133 Milano, Italy.

# Electronic supplementary information (ESI) available.


T ED
C EP
AC
ACCEPTED MANUSCRIPT
Abstract

Iron-functionalized hydroxyapatite (Fe/HAP, 2 < Fe wt.%<7) was prepared by ionic exchange and

tested in the NO x reduction by ammonia (NH3 -SCR). Iron dispersion, structure and acidity of the

samples were investigated by UV-vis-DR spectroscopy, XRPD and ammonia titration. On Fe/HAP

catalysts, NO x conversion started at higher temperatures and was always lower than on the

PT
reference Cu/HAP catalysts. Selectivity to N 2 was remarkable (> 95%) up to the attainment of

maximum NO x conversion; for higher temperatures, selectivity declined due to the formation of

RI
N2 O from NH3 over-oxidation. TPR-TPO cycles allowed to relate the redox properties of samples

SC
with their catalytic performances.
NU
MA

Keywords: Hydroxyapatite; Iron functionalized hydroxyapatite; NH3 -SCR Catalysts.


T ED
C EP
AC
ACCEPTED MANUSCRIPT
1. Introduction

In the wide range of strategies for NO x emission control [1], the selective catalytic reduction of

NO x to N 2 by the use of ammonia (NH3 -SCR) still remains one of the most effective solutions for

both mobile (e.g. car engines) [2] and stationary (e.g. power plants, nitric acid plants, fiber-

producing plants, etc.) [3] applications. The SCR process is essentially based on three redox

reactions:

PT
4NH3 + 4NO + O 2 → 4N 2 +6H2 O Standard SCR (1)

RI
2NH3 + NO + NO 2 → 2N 2 +3H2 O Fast SCR (2)

SC
8NH3 + 6NO 2 → 7N 2 +12H2 O NO2 -SCR (3)
NU
Unfortunately, NH3 can be also oxidized by O 2 present in large excess in typical flue gas or exhaust
MA

gas mixtures, thus resulting in the undesired production of N 2 O and NO x . However, when the

process is carried out in the presence of an efficient SCR catalyst, NH3 oxidation by NO x (Eq. (1-3))
ED

should be favoured over the undesired oxidation by O 2 [4].

The formulation of the catalysts for the SCR process has evolved through the years in response
T
EP

to stringent requirements in terms of emission ceiling values, technological issues and operational

costs. Nowadays, vanadia-based materials [5] and metal-exchanged zeolites [6] represent the two
C

main categories of catalysts commonly used in commercial SCR systems. In particular, during the
AC

years, Fe- and Cu-exchanged zeolites have demonstrated to possess high activity and N 2 selectivity

[7], while some problems of stability have emerged. Several classical zeolites such as

silicoaluminates(ZSM-5, Beta and SSZ-13), and small-pore molecular sieves such as

silicoaluminophosphates (SAPO or aluminophosphate, ALPO) with different topology, porosity and

composition (Si/Al ratio) have been studied [8]. Among these structures, materials with chabazite

topology (e.g. SSZ-13, SAPO-34, and ALPO-34) have distinguished themselves due to their high

hydrothermal stability [8–10]. In all these structures, the metal phase (typically Cu or Fe) has been
ACCEPTED MANUSCRIPT
introduced by classical post-synthetic ion-exchange procedures, which assured high metal

dispersion [11]. However, the final location and the speciation of the metal species depend on

several features (e.g. acidity, porosity, etc). A large number of studies appeared in the last decades

devoted to the identification of the active sites with emphasis on the valence state and location of

the metal species, and the role of the local environment on them [12–14]. In fact, the

comprehension of the reaction mechanism as well as the nature of active sites should be considered

PT
fundamental to overcome the drawbacks, which still limit the effective use of zeolite-based catalysts

(e.g. low sulphur resistance and poor hydrothermal stability). Beside the improvement of these

RI
consolidated systems, the development of new materials is a viable way that can lead to new

SC
interesting catalytic materials able to overcome the above-mentioned limitations and to accomplish

the normative requirements.


NU
Hydroxyapatites (Ca10-x (PO 4 )6-x (HPO 4 )x (OH)2-x with 0 < x  1) are natural inorganic materials,
MA

which can be derived and mined from mineral rocks or mammal bones or, alternatively, synthesized

from cheap precursors under mild conditions; stoichiometric calcium hydroxyapatite (HAP,
ED

Ca10 (PO 4 )6 (OH)2 ) is the best known material among this family [15]. Similar to the zeolite

framework, the hexagonal HAP crystal lattice is susceptible to undergoing ion exchange reaction
T
EP

[16]. In particular, metal ions can be introduced into the HAP structure by replacing Ca2+ ions. The

latter are extremely mobile and occupy two crystallographic distinct sites in the HAP framework:
C

Ca(I) nona-coordinated sites and Ca(II) epta-coordinated sites. A recent review [17] reports on the
AC

successful use of HAP-based catalysts in various reactions, such as cross-coupling , condensation

(Knoevenagel reaction, Claisen-Schmidt reaction, Michael addition), oxidation (alcohol oxidation,

epoxidation, oxidative dehydrogenation of ethane and propane), as well as several other type of

reactions such as alkylation, hydration of nitriles, transesterification, hydroformylation,

hydrogenolysis and hydrogenation reactions.


ACCEPTED MANUSCRIPT
The extreme compositional flexibility of HAP has important implications also for the surface

properties of the material. HAP is an amphoteric material, whose surface acid-base properties of

moderate strength can be controlled by tuning the Ca/P ratio and can be influenced by the local

environment (e.g. water content, carbon dioxide presence, etc) [18,19]..Hence, it has been used in

reactions that require bifunctional solid catalysts, such as Friedel–Crafts [20] and Guerbet reactions

[26,28].

PT
The possibility to easily functionalize HAP with metal species together with the co-presence of

RI
acidic and basic sites on HAP surface contributed to the recent success of this material in important

reactions of environmental interest [22–24]. Recently, Ag- and Cu-loaded HAP demonstrated to be

SC
promising catalysts in the NH3 -SCR reaction [25,26]. It has been also reported [27], that the
NU
preparation method as well as the metal precursor used for the deposition of copper phase

determined the final Cu speciation and sitting on HAP, and consequently influenced the catalytic
MA

performance in the NH3 -SCR process.

Herein, we present the synthesis, characterization and catalytic behaviour of Fe-exchanged


ED

hydroxyapatite samples prepared by a modification of the conventional ion-exchange procedure.


T

The Fe-containing samples have been compared with the already studied Cu-hydroxyapatite
EP

catalysts [27], choosing the best sample of the series. To the best of our knowledge, this is the first

report describing the SCR activity of such systems.


C
AC

2. Experimental

2.1 Materials and catalyst preparation

Hydroxyapatite used in this work (HAP) was kindly supplied by Solvay (Soda Ash &

Derivatives, Brussels). The preparation procedure, composition and main properties of HAP were
ACCEPTED MANUSCRIPT
previously reported [27,28]. Copper deposition on HAP (Cu/HAP) was carried out by conventional

ion exchange (I.E.) procedure [27].In a typical preparation, 6 g of dried HAP powder (120°C

overnight) have been added to 100 mL of ca. 0.06 M copper nitrate solution (Cu(NO 3 )2 ·3H2 O, 99%,

Carlo Erba, Italy),under vigorous magnetic stirring at 40°C and pH 7; the obtained suspension has

been maintained under stirring for 20 h. After filtration, thorough washing with hot water (40°C),

drying at 120°C for 16 h, and calcination at 450°C for 4 h, a light-blue coloured sample has been

PT
recovered. Cu-loading was checked by measuring the absorbance (at 750 nm) of the filtrate waters

to compute the residual copper concentration. The Cu-concentration of the sample was ca. 6 wt.%;

RI
the sample is labelled as Cu6/HAP.

SC
The iron-loaded samples were prepared by a modified ion exchange procedure, in which very
NU
short contact time of HAP in the Fe-solution took place. The preparation was called flash ion

exchange. In a typical preparation, 250 mL of iron (III) nitrate (Fe(NO 3 )3 ∙9H2 O) solution, with
MA

concentration in the range 0.005-0.035 M, were thermostated at 40°C. The pH was adjusted to  3

by HNO 3 addition for preventing the precipitation of iron (III) hydroxide. HAP powder ( 6 g,
ED

previously dried at 120°C, overnight) was added to the iron solution and the suspension was

maintained under stirring at 40°C for 15 min. Samples were then filtered, thoroughly washed, dried
T
EP

at 120°C overnight, and finally calcined at 500°C for 1 h. The collected samples were denoted as

FeX/HAP, where X is the Fe loading (wt.%), between 2 and 7 wt.% Fe; Fe-concentration was
C

doubly checked by inductively coupled plasma atomic spectroscopy (ICP) both on the prepared
AC

samples and on filtrate solutions.

2.2 Characterization methods

BET specific surface area and porosity (pore volume, pore size and pore size distribution) have

been determined by N 2 adsorption/desorption isotherms conducted at -196 °C using an automatic

analyser of surface area (Sorptomatic 1990 Series, Carlo Erba Instruments). Pore size distribution
ACCEPTED MANUSCRIPT
(PSD) was computed by using the Barrett-Joyner-Halenda (BJH) model equation from the

desorption branch of the recorded isotherms. Prior to the analysis, fresh samples (0.15 to 0.30 g)

have been outgassed at 150°C for 4 h under vacuum.

Diffuse reflectance spectra (DRS) of HAP samples in powder form were measured on a double

beam UV–vis–NIR scanning spectrophotometer (Shimadzu UV-3600 plus, Japan) equipped with a

diffuse reflectance accessory (integrating sphere from BIS-603). A given amount of the powder

PT
sample, finely grinded, was uniformly pressed in a circular disk (E.D., ca. 4 cm) included in the

RI
sample-holder; the latter was inserted in a special quartz cuvette and then put on a window of the

integrating sphere for the reflectance measurements. The measured reflectance spectra (R,%) were

SC
converted to absorbance (Abs) using Eq. (4): NU
Abs = Log (1/R/100) (4)
MA

Structural properties of the samples were investigated using an X-ray powder diffractometer

equipped with a PW 1830 generator, monochromator in graphite, copper pipe, Cu Kα radiation at

40 kV × 40 mA. Powder X-ray diffractograms were collected according to the following method;
ED

type of scan: continuous; start angle: 10° 2; end angle: 60° 2; step size: 0.05° 2; time for step: 5
T

s; scan speed: 0.010 2/s; number of step: 1000; total time: 1h 23 min. The pattern of a typical
EP

crystalline hydroxyapatite phase can be found in JCPDS: 00-009-0432.


C

Acidity of bare and metal-loaded HAP samples was evaluated by an NH3 dynamic adsorption
AC

experiment in flowing gas conditions. Typically, the dried and weighted sample was put on a

porous septum of a quartz reactor and activated at 120°C under flowing air for 30 min. The NH3 /N2

gas mixture (500 ppm NH3 concentration) was then passed at 6 NL h-1 through the porous septum

and entered in a gas cell (path length 2.4 m multiple reflection gas cell) in the beam of an FT-IR

spectrometer (Bio-Rad with DTGS detector). NH3 was completely adsorbed for a given measured

time, as observed from the trace of the NH3 line at 966 cm-1 , that was recorded as a function of

time, on every sample. When saturation of the acid sites of the solid by NH3 was attained, the NH3
ACCEPTED MANUSCRIPT
signal was restored at the same value of its starting concentration. From the evaluation of the time

during which the NH3 -signal has remained to zero value, the amount of acid sites could be

evaluated (Eq. (5):

𝑚𝑜𝑙𝑒𝑠 𝑁 𝐻 [𝑁𝐻3 ]𝑓𝑒𝑑 ∙𝐹 ∙𝑡∙𝑃


3(𝑎𝑑𝑠)
= (5)
𝑔𝑠𝑎 𝑚𝑝𝑙𝑒 𝑅𝑇 ∙𝑚𝑠𝑎𝑚 𝑝𝑙𝑒

where [NH3 ] fed is the NH3 concentration (ppm) in the adsorption gas stream; F is the total volume

PT
flow rate (NL·h-1 ) of the NH3 /He gas mixture; t is the time (min) during which NH3 was completely

adsorbed; P is the pressure (atm); and msample (g) is the sample mass. Assuming a 1:1 stoichiometry

RI
for the NH3 adsorption on the surface acid sites, the amount of acid sites per sample mass (in

SC
mol·g-1 ) was evaluated. Measurements were replicated and in all cases a percent relative
NU
uncertainty of less than 3% was obtained.

The temperature programmed reduction (TPR) and oxidation (TPO) analyses were carried out
MA

with a home-made apparatus described elsewhere [29]. A mass spectrometer (Hiden Analytical,

HPR20) was used to monitor the change in gas flow composition during the analyses. The samples
ED

(ca. 300 mg) were loaded in a quartz U-shaped reactor and diluted with silica sand (ca. 200 mg).

For each sample, the series of analysis consisted of a first TPR (TPR-1) followed by a TPO, and
T

then by another TPR analysis (TPR-2). Before starting the series of analysis, the samples were
EP

purged at room temperature with Ar flow (40 ml/min) for 10 min, followed by a treatment at 150 °C
C

for 30 min; after this treatment, the samples were cooled to room temperature under Ar flow (40
AC

ml/min). For the TPR analysis, 5 vol.% H2 /Ar was used, while for the TPO analysis, 10 vol.%

O2 /Ar. The analysis procedure was as follows. Prior to the start of heating ramp, the mass

spectrometer signals (m/z=2, 32, 18 and 44 for H2 , O 2 , H2 O, CO 2 , respectively) were stabilized by

flowing the analysis gas (40 ml/min) at room temperature. The heating was then initiated with a

ramp rate of 10 °C/min to 350°C or to 500 °C, holding the final temperature for 1 h. After each

analysis, the samples were cooled down under pure Ar flow (40 ml/min), and the procedure was
ACCEPTED MANUSCRIPT
repeated by changing the analysis gas. The data were collected by recording the mass signals of H 2

(m/z = 2) and O 2 (m/z = 32) for the TPR and TPO analyses, respectively.

2.3 Catalytic tests (NH3 -SCR)

The tests of the NH3 -SCR reaction have been performed in a continuous reaction line equipped

with a set of mass flow controllers (Bronkhorst, Hi-Tec and Brooks Instruments), a tubular vertical

PT
electric oven (maximum temperature of 1000°C, Eurotherm Controller-Programmer type 818), a

RI
glass tubular catalytic microreactor (5 mm i.d.), and an on line FT-IR spectrophotometer (Bio-Rad

with DTGS detector) for the qualitative and quantitative determination of the fed and vented

SC
gaseous species. The catalyst sample (ca. 0.20 g), sieved
NU in the 0.35-0.25 mm particle size and

dried at 120°C in an oven overnight, was pre-treated in situ under O 2 /N2 flow (20% v/v) at 120 °C

for 30 min. The catalyst activity was studied as a function of temperature (120-400°C) maintaining
MA

constant the concentration of the feed gas mixture and total flow rate at 6 NL/h (corresponding to a

contact time of 0.12 s). Each temperature was maintained at least for 60 min in order to allow the
ED

attainment of steady-state reaction conditions. The fed gas mixture was prepared after mixing ca.

500 ppm of NO, 500 ppm of NH3 and 10,000 ppm of O 2 . After mixing, the effective fed mixture
T
EP

contains also NO 2 (ca. 450 ppm of NO, 50 ppm of NO 2 , 500 ppm of NH3 , and ca. 10,000 ppm of

O2 ). This gas-mixture passed over the catalyst particles at 120°C until the catalyst surface was
C

saturated with NH3 . Additional NH3 -SCR measurements have been also performed by varying the
AC

fed gas NH3 concentration (from 280 to 800 ppm) at constant NO concentration (500 ppm) and at

fixed temperature (250°C) and contact time (ca. 0.12 s).

The effluent gas from the reactor was monitored every3 min by FT-IR equipped with a multiple

reflection gas cell (with 2.4 m path length; resolution, 2 cm-1 ; sensibility, 1.5, 92 scans per 180 s)

for quantifying the gaseous species (unconverted reagents and/or products). The total absorbance of

all the IR active gaseous species (Gram–Schmidt) exiting the reactor was continuously recorded as
ACCEPTED MANUSCRIPT
a function of time, while reaction temperature was changing. The NO, NO 2 , and NH3 and other

formed species (if any) were quantified from the peak height of a selected absorbance line on the

basis of the measured calibration factors. N2 concentration was determined from the difference

between the fed NH3 and NO x (NO plus NO2 ) concentration and the sum of the unreacted NH3 plus

NO+N 2 O+NO 2 concentrations, eventually formed. Computational details can be found in

Supplementary Material.

PT
3. Results and discussion

RI
SC
3.1 Preparation and characterisation of catalystsNU
The functionalization of the HAP surface by Cu2+ or Fe3+ species has been easily accomplished

by exploiting the exchange properties of the HAP lattice, in which Ca2+ ions (both the so called
MA

Ca(I) and Ca(II) sites) can be substitued with monovalent, bivalent, or trivalent cations. The nature

of the metal precursor, pH of solution, and contact time of HAP in the metal solution determine the
ED

final metal dispersion and sitting on the HAP surface. Concerning the preparation of copper
T

containing HAP samples, Tounsi et al. [33] working with copper nitrate precursor at pH 4.5 for long
EP

contact time (ca. 50 h), have observed a first very rapid ion exchange of the surface Ca 2+ ions with

Cu2+ followed by slow reaction leading to the growth of libethenite phase [Cu2 (PO 4 )(OH)]) on HAP
C

surface. Schiavoni et al. [27] working with copper acetate, copper chloride, or copper nitrate at pH 7
AC

for ca. 20 h contact time, have observed only a highly Cu-dispersed phase on HAP surface up to ca.

6 wt.% of copper loading. Among this sample series, Cu6/HAP prepared from copper nitrate has

been taken as a representative Cu-sample in this work (Table 1).

Concerning the Fe/HAP sample preparation, a modification of the classical ion exchange

procedure was developed aimed at achieving high dispersion of Fe-phase on HAP. At first, acidic

Fe-solutions were prepared (ca. pH 3) to prevent iron aggregation and precipitation from the bulk
ACCEPTED MANUSCRIPT
Fe-nitrate present at concentrations from 0.005 to 0.035 M. Contact time of HAP in the acidic Fe-

solutions was very short (15 min) to prevent morphological and structural damages to HAP, which

is known to be sensitive to amorphization or dissolution in acidic environment [30]. By this

procedure, Fe/HAP samples with Fe-loading from ca. 2 to 7 wt.% have been prepared (Table 1).

Concerning the surface texture of the Fe/HAP samples, they have all similar values for specific

surface area and pore volume, independently of the Fe-loading (Table 1). This behaviour is an index

PT
of the good dispersion of the Fe-phase in all the samples. The specific surface area and porosity of

RI
the Cu/HAP series samples [27] are slightly higher, as shown for the Cu6/HAP sample (Table 1).

This small difference is likely due to the bare HAP used for Cu- and Fe-deposition, which belonged

SC
to the two different batches of preparation in Solvay.
NU
The surface of HAP is known to posesses both acidic and basic centres [18,23,31], thus the

addition of dispersed metal centres (as Fe3+) should increase the surface acidity of HAP. To prove
MA

this point, solid-gas acid-base titrations of the Cu6/HAP and Fe/HAP samples were carried out by

using ammonia as a base probe molecule. The obtained results are reported in Table 1 in terms of
ED

the amount of ammonia adsorbed (molNH3 /gcat ). The obtained results indicate that despite the
T

almost regular increase of the iron amount in the samples (0.371, 0.854, and 1.22 mmolFe/gHAP for
EP

Fe2/HAP, Fe5/HAP, and Fe7/HAP, respectively), there is not a regular increase of acidity. The first

addition of Fe on HAP (Fe2/HAP) led to a very high increase of the acid sites (Table 1), while a
C
AC

very lower increase of acid sites was observed following the successive Fe-additions (Fe5/HAP and

Fe7/HAP). This behaviour suggests that at first Fe3+ could be accomodated on the HAP surface and

in the exchange Ca(I) or/and Ca(II) positions of HAP, then aggregation of Fe phase could occur

with further Fe-addition. Cu6/HAP has a well-developed acidity that can be associated with high

copper phase dispersion, as already presented and discussed [27].

It is known that the incorporation of metal ions in the HAP framework can proceed according to

three different mechanisms: surface complexation, ion exchange and dissolution precipitation [16].
ACCEPTED MANUSCRIPT
The occurrence and the eventual predominance of one of these mechanisms depends on the metal

properties (e.g. the charge/radius ratio) and the operational conditions (pH, temperature, contact

time)thatdetermine the final state and speciation of the metal ion on the HAP surface. In the case of

Fe/HAP samples, to assess the presence of a new crystalline phase, structural PXRD analysis was

performed. The obtained X-rays diffraction patterns of the Fe/HAP samples and that of Cu6/HAP

are reported in Fig.S1a-b, while reference diffraction patterns of HAP (Fig. S1c) and of the main

PT
Cu- (Fig. S1d) and Fe-containing (Fig.S1e) phases are reported for comparison. In all the cases,

only the characteristic diffraction peaks of the hexagonal crystal HAP phase have been detected,

RI
thus suggesting that HAP is the only crystal phase present in Cu- and Fe-loaded samples. The

SC
absence of additional formed crystal phases cannot allow excluding that metal agglomeration might

had been occurred to some extent, resulting in the formation of amorphous, nanosized aggregates,
NU
which cannot be detected by PXRDanalysis. Consequently, the speciation and state of the metal
MA

species of the Cu6/HAP and Fe/HAP samples were furtherly investigated by UV-vis diffuse

reflectance spectroscopy in the 200-1200 nm range and recorded spectra are shown in Fig. 1.
ED

The spectrum of Cu6/HAP (Fig. 1a) presents two distinct bands: an intense band in the 210-350 nm

region, which is ascribed to Cu2+←O ligand-to-metal charge transfer transition, and a broader band
T
EP

in the region 600-850 nm, which is typically associated with d-d transitions (T2g → Eg) of Cu2+ in

an axially octahedral environment of oxygen-containing ligands. The absence of additional bands or


C

shoulders in the 400-550 nm region allows to rule out the presence ofCuOx three-dimensional
AC

clusters. A more complex feature characterized the DR spectra of Fe/HAP samples (Fig.1b), where

three bands can be observed in the region 200-550 nm, along with a very broader band in the 800-

1000 nm region (Fig. S2, Fe7/HAP). The most intense signal at 200-300 nm can be attributed to

isolated Fe3+centres, while the broad band at higher wavelength (485-550 nm) is typical of very

small Fex O y aggregates. More precisely, the bands at 200-230 nm and 270-290 nm correspond to

isolated Fe3+ centres in tetrahedral coordination (change transfer transition, 1T2u → 2t2q) and higher
ACCEPTED MANUSCRIPT
(octahedral) coordination (change transfer transition, 6 A1 → 4 T1 ), respectively. The band at 485-550

nm is usually generated by ligand field d-d transition, 2(6 A1 ) → 2(4 T1 ) in Fe-O-Fe oligomers. In the

DR spectra of Fig. 1b-d, this band appears more clearly in the case of Fe7/HAP, while it can only be

guessed in the spectrum of Fe2/HAP. Nevertheless, it is possible to conclude that isolated Fe

centres together with Fe-O-Fe oligomers are the main species present in all Fe/HAP samples

regardless of the Fe loading. Concerning the band around 800-1000 nm, it can be assigned to α-

PT
Fe2 O3 aggregates of different sizes (A1 → T1 transition) [32],without any possibility to make

quantitative evaluations on the Fe-species.

RI
The influence of metal speciation on the red-ox properties of the studied Cu- and Fe-catalysts

SC
has been evaluated by realizing H2 -TPR (TPR-1), a successive O2 -TPO, and a second H2 -TPR
NU
(TPR-2) experiment(Fig. 2). In these cycles, the maximum temperature attained is350°C for the

Cu6/HAP and 500°C for the Fe/HAP samples; it is noted that very high temperatures could cause
MA

modification of the samples and metal phase segregation. TPR profiles of the studied samples are

shown in Fig. 2; bare HAP presents three reduction peaks at high temperature (>450°C) [33], which
ED

could interfere with the quantification of the amount of hydrogen consumption from reduction of

copper and iron species. Based on the literature on Cu-containing zeolites and oxides[32],the
T

reduction of isolated Cu2+species has been proposed to occur by a two-step mechanism: Cu2+ Cu+
EP

and Cu+ Cu(0). Fig. 2a shows the TPR-1 profile of Cu6/HAP with two peaks centered at 243°C
C

and 282°C, which is an indication of the existence of more than one copper species highly dispersed
AC

on the Ca-HAP surface. The amount of H2 consumed was  50% of the total amount of copper in

Cu6/HAP. This suggests that the observed peaks are related only to the first step reduction (Cu2+

Cu+),and Cu+ species are formed from the TPR-1 analysis (carried out up to 350°C).On the other

hand, Tounsi et al. [26], on a series of copper-hydroxyapatite samples, have observed broad peaks

at 440-450°Cascribed to the reduction of Cu+ to Cu(0),formed from the reduction of isolated

Cu2+present on the HAP surface.On Cu6/HAP, a second reduction is repeated (TPR-2) after TPO
ACCEPTED MANUSCRIPT
analysis. The collected TPR-2 profile isnot superposed to the first one; a unique reduction peak

isobserved, centered at 237°C, at lower temperatures than the peaks of the TPR-1 analysis. The

hydrogen consumption in the second reduction (TPR-2) is very similar to that estimated in the TPR-

1 analysis. This suggests that Cu+(obtained from TPR-1 analyis) is completely re-oxidized to

Cu2+(during the TPO analysis) and then reduced again to Cu+.during TPR-2. Concerning the

different peak shapesbetween TPR-1 and TPR-2, some degree of copper aggregation could be

PT
guessed after the TPR-1 and TPO analyses.

A more difficult situation has been found on the interpretation of the redox cycles of Fe/HAP

RI
samples (Fig. 2 b). Only the initial part of the TPR-1 profiles of the Fe/HAP samples has been

SC
recorded; the hydrogen consumption starts at temperature higher than 350°C (Tonset of 390°C,

365°C, and 370°C for Fe2/HAP, Fe5/HAP, and Fe7/HAP, respectively) and the temperature of
NU
maximum rate of reduction is not achieved at 500°C. It is known [32,33] that reduction of Fe3+
MA

species and/or small Fe3 O4 aggregates to FeO represents the first step of the reducing process that

continues with the reduction of Fe2+ to Fe(0). Moreover, Kachani et al. [34] reported H2 -TPR
ED

profiles for several Fe/HAP samples with different Fe loading prepared by conventional ion

exchange procedures. For all the cataysts, one or two overlapped peaks were detected at very high
T

temperatures (>550°C), and on the basis of H2 consuption and reported works, the authors attributed
EP

these signals to the reduction of Fe3+ to Fe2+.


C

It is interesting to compare the initial reduction profiles of TPR-1 with those of TPR-2, after the
AC

TPO experiment. TPR-2 profiles start at lower temperatures (ca. 345°C) and they all attain a clear

maximum of reduction at ca. 475°C. As observed for the Cu2+ species (redox cycle of Cu6/HAP), it

seems that the Fe3+ species possess red-ox properties and they are quite mobile in the HAP

structure, as a significant shift of the temperatures of TPR-1 and TPR-2 profiles is observed, which

causes to some extent Fe-aggregation with the formation of a more easily reducible Fe-phase.

The overall H2 -TPR results comparing the Cu6/HAP and Fe/HAP samples, indicate that the

copper species are more easily reducible than the iron species, since they start to get reduced to
ACCEPTED MANUSCRIPT
much lower temperatures (T<200°C). This has very important implications for the SCR reaction,

which will be discussed below.

3.2 Sample performance in NH3 -SCR

The catalytic performance of Cu6/HAP and Fe/HAP samples was evaluated in the lab-scale

reaction line described in Section 2.3. In a typical catalytic test, the fed reaction gas mixture,

PT
containing NO, NO 2 , NH3 , and O 2 diluted in N2 , flows through the catalytic bed and a temperature

RI
program composed of several isothermal steps is applied to study the catalytic behaviour as a

function of reaction temperature. For each isothermal step and after the achievement of steady state

SC
reaction condition, the decomposition of the Gram–Schmidt signal provided the infrared spectrum
NU
of each species, thus obtaining the composition of the exiting gas mixture from the reactor. Fig.S3

shows typical temperature-resolved FT-IR spectra obtained on Cu6/HAP (Fig. S3 a) and Fe5/HAP
MA

(Fig. S3 b)as an example. From Fig.S3a-b, it is possible to follow the disappearance and the

evolution of different species as a function of temperature. Concerning ammonia, by monitoring the


ED

signal centred at 950 cm-1 , associated with the A1 symmetric deformation of N–H bonds of the NH3

molecule, it appears clear that NH3 is progressively consumed by increasing the temperature up to a
T

total conversion, in any case. Conversely, the signal of NO stretching (1875 cm-1 ) decreases up to
EP

ca. 350°C, indicating that NO is reduced by NH3 , whereas at higher temperatures, an inverse trend
C

is observed, where the NO signal increases, indicating the occurrence of non-selective NH3
AC

oxidationreaction by oxygen. This is confirmed by the appearance of a signal at ca. 2224 cm-1 ,

typical of σ+ N–N stretching of N 2 O. From the absorbance of the different detected species, the

concentration profiles as a function of time/temperature are computed and the catalytic parameters

(conversion and selectivity) are calculated, according to equations reported in Table S1.

NO x conversion at 250°C over the Cu6/HAP and Fe/HAP samples is comparatively presented in

Fig. 3, for the three different NH3 /NO feed ratios, while all the data of concentration at each
ACCEPTED MANUSCRIPT
reaction temperature and of NH3 /NO ratio are listed in Tables S2-S10 for the Fe/HAP samples (data

for Cu6/HAP can be found in Ref. [27]). As a general trend, NO x conversion curves as a function of

reaction temperature have a volcano-shape with maximum values of conversion at ca. 350-400°C

at any NH3 /NO feed ratio, while NH3 conversion curves regularly increase with temperature starting

from 250 or 300°C, depending on the NH3 /NO feed ratiowith a quite linear trend.

Fig. 3 clearly shows that Cu6/HAP exhibits higher NOx conversion at 250°C compared with the

PT
Fe/HAP samples at any tested NH3 /NO feed ratio. Nevertheless, for the Fe/HAP samples, the

RI
temperature of maximum NO x conversion is in the interval 350-400°C, and higher NO x

conversions can be attained at such temperatures (e.g. 64-65% on Fe5/HAP with NH3 /NO feed

SC
ratio of 1 and 2). As reported by A. Sultana et al. [35], the easiness of reducibility of the metal
NU
species determines the extent of NO conversion at low temperature; the easier the reduction of

metal species is, the higher is the NO oxidation ability in the NH3 -SCR reaction. Thus, the more
MA

easily reducible copper than iron species allows oxidizing NO to NO 2 and nitrates, which are rate-

controlling intermediate species, responsible for the enhanced low-temperature NO x conversion.


ED

It is interesting to observe the effect of the NH3 /NO feed gas ratio on the NO x conversion and
T

N2 selectivityfor Cu6/HAP and Fe/HAP samples(Fig. 3). In the case of Cu6/HAP, NO x conversion
EP

regularly increases with increasing NH3 /NO ratio, while the selectivity to N 2 is maintained

constant and at a very high value. The same trend is not observed in the Fe/HAP samples; NO x
C

conversion increases when the NH3 /NO feed gas ratio increases from 0.6 to 1, but then it decreases
AC

when the NH3 /NO ratio is 2 for any sample. This trend ismaintained for the Fe/HAP samples also

for T>250°C. These trends can be interpretedconsidering possible reaction pathways (e.g. two

oxidants, NO x and O 2 are in competition for NH3 oxidation), and the different oxidizing/reducing

properties of copper and iron species. The increase of NH3 /NO feed gas ratio implies working with

a higher substrate (NH3 ) concentration available to be oxidized either by NO x or O2 . An higher

ammonia concentration has positive effect on the reaction rates of both the selective (NO x +NH3 )
ACCEPTED MANUSCRIPT
and unselective (NH3 +O 2 ) reactions.The catalyst which is able to better exploit the higher ammonia

concentration to improve the selectivity of the desired (NO x +NH3 ) reaction will be the most

selective SCR catalyst on which an increase in NO x conversion will be observed. Based on this

argument, Cu6/HAp is a more selective catalyst for the SCR reaction than Fe/HAP. In agreement

with this interpretation, the selectivity to N 2 was found to have decreased slightly at NH3 /NO feed

gas ratio of 2, as NH3 is oxidized to N 2 O (and NO x ) by O 2 (see Tables S2-S10). The latter Tables

PT
report the values of NO, NO 2 , N 2 O and NH3 concentrations at several temperatures in the 120-

500°C range and at each NH3 /NO feed gas ratio from which the relevant percent conversion can be

RI
computed; Tables S2-S10 report also the concentration values of NO x and N 2 from which the

SC
percent selectivity can be deduced.

In general, working with the NH3 /NO feed ratio of 0.6, low NO x conversion (around 38-45%)
NU
and not complete NH3 conversion (80-85%) were observed on all the three Fe/HAP samples. By
MA

increasing the value of NH3 /NO to 1 and 2, NO x and NH3 conversions increased. At 400°C,

Fe2/HAP attains maximum 48-50% NO x conversion; Fe7/HAP attains 56-58% NOx conversion
ED

and Fe5/HAP 64-65% NO x conversion. Concerning NH3 ,at 500°C on all the Fe/HAP samples,

conversion attained 100% with NH3 /NO feed ratio of unity, and 85-93% with NH3 /NO ratio of 2.
T
EP

Among the Fe/HAP samples, the superior activity of Fe5/HAP is obtained at any NH3 /NO feed

gas ratio (Fig. 3). Likely, this superior activity could arise from a balanced presence at the surface
C

of Fe5/HAP, of slight Fe-aggregates, like dimers and α-Fe2 O 3 nano-aggregates (as observed from
AC

UV-vis-DRS). The same behavior holds in the case of Cu/HAP samples, where samples prepared

by the wet impregnation (surface CuO x nano-aggregates) show superior SCR activity than those

prepared by ionic-exchange (high dispersion of Cu2+).

The complete trends of NO x conversion as a function of temperature over Cu6/HAP and

Fe5/HAP are comparatively shown in Fig. 4. The temperature of activity onset on Cu6/HAP is

lower than 150°C, while it is 250°C on Fe5/HAP; the maximum conversion of NO x is attained at
ACCEPTED MANUSCRIPT
250°C on Cu6/HAP and at 375°C on Fe5/HAP. In any case, lower NO x conversion above the

maximum can be explained by the higher reaction rate of the undesired NH3 oxidation with O 2 that

leads to a decrease in NH3 concentration needed for the SCR reaction. Moreover, at temperatures

higher than those of maximum NO x conversion, the concentration of NO x and N 2 O increased as

they are products of the undesired NH3 +O 2 reaction.

PT
4. Conclusions

RI
Hydroxyapatite functionalized with Cu2+ and Fe3+ species has given rise to materials that have

SC
demonstrated an interesting potential to be used as catalysts in the NH3 -SCR reaction. Addition of

Cu2+ and Fe3+ can be easily realized by exploiting the exchange properties of hydroxyapatite.
NU
Presence of Cu2+ species completely dispersed over HAP surface can be obtained, while concerning
MA

iron phase, Fe3+ dispersed ions can be accompanied by the formation of dimers and some amount

of -Fe2 O3 aggregates, depending on the total Fe-concentration.


ED

The Cu/HAP and Fe/HAP catalysts are active in the SCR reaction in different temperature

ranges, and this difference is justified by the different reduction properties of Cu2+ and Fe3+, which
T

occur at significantly lower temperatures (about 250° C) and higher temperatures (about 450° C),
EP

respectively. Reduction of NO x occurred with high selectivity to N 2 up to the attainment of the


C

maximum NO x conversion, whereas at higher temperatures the unselective reaction (NH3 +O2 )
AC

predominated with the formation of NO x and N 2 O.

Although Cu/HAP catalysts showed higher activity than Fe/HAP catalysts, the latter were able

to maintain good activity over a wider temperature range compared to Cu/HAP. This is an

important aspect from a practical point of view, since it allows to work in a broader operating

temperature window. In addition, the difference in activity between Cu/HAP and Fe/HAP catalysts

is not so great, therefore, from a cost-efficiency point of view, the use of Fe/HAP could be preferred
ACCEPTED MANUSCRIPT
over Cu/HAP owing to the not negligible cost of copper.This representsa stimulus to improve and

optimize Fe/HAP catalysts.

It should be interesting to further study hydroxyapatite materials for the SCR reaction on which

the iron phase can be deposited by other different preparative methods for which little modification

of the structure of the metal species seems to be required in order to obtain well active and selective

catalysts.

PT
Conflicts of interest

RI
There are no conflicts to declare.

SC
NU
Acknowledgements

Solvay is gratefully acknowledged for supplying hydroxyapatite in powder form.


MA

The use of instrumentation purchased through the SmartMatLab Project of Dipartimento di Chimica

(Cariplo Foundation, project 2013-1776) is gratefully acknowledged, in particular Dr. Daniele


ED

Marinotto of Consiglio Nazionale delle Ricerche (CNR), Istituto di Scienze e Tecnologie

Molecolari (ISTM).
T
EP

Dr. Marco Schiavoni and Mrs. Iolanda Biraghi are acknowledged for their experimental support.
C
AC
ACCEPTED MANUSCRIPT
References

[1] K. Skalska, J.S. Miller, S. Ledakowicz, Trends in NOx abatement: A review, Sci. Total

Environ. 408 (2010) 3976–3989. doi:10.1016/j.scitotenv.2010.06.001.

[2] G. Yang, J. Ran, X. Du, X. Wang, Y. Chen, Microporous and Mesoporous Materials Di ff

erent copper species as active sites for NH 3 -SCR reaction over Cu-SAPO-34 catalyst and

reaction pathways : A periodic DFT study, Microporous Mesoporous Mater. 266 (2018) 223–

PT
231. doi:10.1016/j.micromeso.2018.01.034.

[3] S. Roy, M.S. Hegde, G. Madras, Catalysis for NO x abatement, Appl. Energy. 86 (2009)

RI
2283–2297. doi:10.1016/j.apenergy.2009.03.022.

SC
[4] J.L. Sorrels, D.D. Randall, K.S. Schaffner, C.R. Fry, Chapter 2 Selective Catalytic

Reduction, in: EPA Air Pollut. Control Cost Man., 2016: pp. 1–108.
NU
https://www3.epa.gov/ttn/ecas/docs/SCRCostManualchapter7thEdition_2016.pdf.
MA

[5] G. Ramis, L. Yi, G. Busca, Ammonia activation over catalysts for the selective catalytic

reduction of NO , and the selective catalytic oxidation of NH 3 . An FT-IR study, Catal.


ED

Today. 28 (1996) 373–380. doi:10.1016/j.cattod.2007.10.024.

[6] J. Li, H. Chang, L. Ma, J. Hao, R.T. Yang, Low-temperature selective catalytic reduction of
T

NOx with NH3 over metal oxide and zeolite catalysts—A review, Catal. Today. 175 (2011)
EP

147–156. doi:10.1016/j.cattod.2011.03.034.
C

[7] M. Colombo, I. Nova, E. Tronconi, A comparative study of the NH3-SCR reactions over a
AC

Cu-zeolite and a Fe-zeolite catalyst, Catal. Today. 151 (2010) 223–230.

doi:10.1016/j.cattod.2010.01.010.

[8] D.W. Fickel, E.D. Addio, J.A. Lauterbach, R.F. Lobo, Applied Catalysis B : Environmental

The ammonia selective catalytic reduction activity of copper-exchanged small-pore zeolites,

"Applied Catal. B, Environ. 102 (2011) 441–448. doi:10.1016/j.apcatb.2010.12.022.

[9] R. Martínez-franco, M. Moliner, C. Franch, A. Kustov, A. Corma, Applied Catalysis B :

Environmental Rational direct synthesis methodology of very active and hydrothermally


ACCEPTED MANUSCRIPT
stable Cu-SAPO-34 molecular sieves for the SCR of NO x, "Applied Catal. B, Environ. 127

(2012) 273–280. doi:10.1016/j.apcatb.2012.08.034.

[10] M. Schiavoni, S. Campisi, A. Gervasini, Effect of Cu deposition method on silico

aluminophosphate catalysts in NH3-SCR and NH3-SCO reactions, Appl. Catal. A Gen. 543

(2017) 162–172. doi:https://doi.org/10.1016/j.apcata.2017.06.034.

[11] L.O. Öhman, B. Ganemi, E. Björnbom, K. Rahkamaa, R.L. Keiski, J. Paul, Catalyst

PT
preparation through ion-exchange of zeolite Cu-, Ni-, Pd-, CuNi- and CuPd-ZSM-5, Mater.

Chem. Phys. 73 (2002) 263–267. doi:10.1016/S0254-0584(01)00391-1.

RI
[12] S.A. Bates, W.N. Delgass, F.H. Ribeiro, J.T. Miller, R. Gounder, Methods for NH 3 titration

SC
of Brønsted acid sites in Cu-zeolites that catalyze the selective catalytic reduction of NO x

with NH 3, J. Catal. 312 (2014) 26–36. doi:10.1016/j.jcat.2013.12.020.


NU
[13] C. Paolucci, A.A. Verma, S.A. Bates, V.F. Kispersky, J.T. Miller, R. Gounder, W.N.
MA

Delgass, F.H. Ribeiro, W.F. Schneider, Isolation of the copper redox steps in the standard

selective catalytic reduction on Cu-SSZ-13, Angew. Chemie - Int. Ed. 53 (2014) 11828–
ED

11833. doi:10.1002/anie.201407030.

[14] T.V.W. Janssens, H. Falsig, L.F. Lundegaard, P.N.R. Vennestrøm, S.B. Rasmussen, P.G.
T

Moses, F. Giordanino, E. Borfecchia, K.A. Lomachenko, C. Lamberti, S. Bordiga, A.


EP

Godiksen, S. Mossin, P. Beato, A consistent reaction scheme for the selective catalytic
C

reduction of nitrogen oxides with ammonia, ACS Catal. 5 (2015) 2832–2845.


AC

doi:10.1021/cs501673g.

[15] J.C. Elliott, Structure and Chemistry of the Apatites and Other Calcium Orthophosphates,

Elsevier Science, 2013. https://books.google.it/books?id=dksXBQAAQBAJ.

[16] S. Campisi, C. Castellano, A. Gervasini, Tailoring the structural and morphological

properties of hydroxyapatite materials to enhance the capture efficiency towards copper(ii)

and lead(ii) ions, New J. Chem. 42 (2018) 4520–4530. doi:10.1039/C8NJ00468D.

[17] A. Fihri, C. Len, R.S. Varma, A. Solhy, Hydroxyapatite: A review of syntheses, structure and
ACCEPTED MANUSCRIPT
applications in heterogeneous catalysis, Coord. Chem. Rev. 347 (2017) 48–76.

doi:10.1016/j.ccr.2017.06.009.

[18] L. Silvester, J.-F. Lamonier, R.-N. Vannier, C. Lamonier, M. Capron, A.-S. Mamede, F.

Pourpoint, A. Gervasini, F. Dumeignil, Structural, textural and acid–base properties of

carbonate-containing hydroxyapatites, J. Mater. Chem. A. 2 (2014) 11073–11090.

doi:10.1039/C4TA01628A.

PT
[19] S. Diallo-Garcia, M. Ben Osman, J.-M. Krafft, S. Casale, C. Thomas, J. Kubo, G. Costentin,

Identification of Surface Basic Sites and Acid–Base Pairs of Hydroxyapatite, J. Phys. Chem.

RI
C. 118 (2014) 12744–12757. doi:10.1021/jp500469x.

SC
[20] R. Tahir, R. Nazih, Comparison of different Lewis acid supported on hydroxyapatite as new

catalysts of Friedel – Crafts alkylation, Analysis. 218 (2001) 25–30.


NU
[21] C.R. Ho, S. Shylesh, A.T. Bell, Mechanism and Kinetics of Ethanol Coupling to Butanol
MA

over Hydroxyapatite, ACS Catal. 6 (2016) 939–948. doi:10.1021/acscatal.5b02672.

[22] L. Silvester, J.-F. Lamonier, J. Faye, M. Capron, R.-N. Vannier, C. Lamonier, J.-L. Dubois,
ED

J.-L. Couturier, C. Calais, F. Dumeignil, Reactivity of ethanol over hydroxyapatite-based Ca-

enriched catalysts with various carbonate contents, Catal. Sci. Technol. 5 (2015) 2994–3006.
T

doi:10.1039/C5CY00327J.
EP

[23] M. Ben Osman, S. Diallo Garcia, J.-M. Krafft, C. Methivier, J. Blanchard, T. Yoshioka, J.
C

Kubo, G. Costentin, Control of calcium accessibility over hydroxyapatite by post-


AC

precipitation steps: influence on the catalytic reactivity toward alcohols, Phys. Chem. Chem.

Phys. 18 (2016) 27837–27847. doi:10.1039/C6CP05294K.

[24] D. Chlala, J.-M. Giraudon, N. Nuns, M. Labaki, J.-F. Lamonier, Highly Active Noble-Metal-

Free Copper Hydroxyapatite Catalysts for the Total Oxidation of Toluene, ChemCatChem. 9

(2017) 2275–2283. doi:10.1002/cctc.201601714.

[25] P.A. Kumar, M.P. Reddy, L.K. Ju, H.H. Phil, Novel Silver Loaded Hydroxyapatite Catalyst

for the Selective Catalytic Reduction of NOx by Propene, Catal. Letters. 126 (2008) 78–83.
ACCEPTED MANUSCRIPT
doi:10.1007/s10562-008-9561-y.

[26] H. Tounsi, S. Djemal, C. Petitto, G. Delahay, Applied Catalysis B : Environmental Copper

loaded hydroxyapatite catalyst for selective catalytic reduction of nitric oxide with ammonia,

"Applied Catal. B, Environ. 107 (2011) 158–163. doi:10.1016/j.apcatb.2011.07.009.

[27] M. Schiavoni, S. Campisi, P. Carniti, A. Gervasini, T. Delplanche, Focus on the catalytic

performances of Cu-functionalized hydroxyapatites in NH3-SCR reaction, Appl. Catal. A

PT
Gen. 563 (2018) 43–53. doi:https://doi.org/10.1016/j.apcata.2018.06.020.

[28] P. Perrin, O.J.F.J.G. Bodson, T. Delplanche, D. BREUGELMANS, Process for producing a

RI
calcium phosphate reactant, reactant obtained and use thereof in the purification of liquid

SC
effluents, (2015). http://www.google.com/patents/WO2015173437A1?cl=en.

[29] V. Dal Santo, C. Dossi, A. Fusi, R. Psaro, C. Mondelli, S. Recchia, Fast transient infrared
NU
studies in material science: Development of a novel low dead-volume, high temperature
MA

DRIFTS cell, Talanta. 66 (2005) 674–682. doi:10.1016/j.talanta.2004.12.015.

[30] M. Ferri, S. Campisi, M. Scavini, C. Evangelisti, P. Carniti, A. Gervasini, In-depth study of


ED

the mechanism of heavy metal trapping on the surface of hydroxyapatite, Appl. Surf. Sci. 475

(2019) 397–409. doi:S0169433218336134.


T

[31] P. Carniti, A. Gervasini, C. Tiozzo, M. Guidotti, Niobium-Containing Hydroxyapatites as


EP

Amphoteric Catalysts: Synthesis, Properties, and Activity, ACS Catal. 4 (2014) 469–479.
C

doi:10.1021/cs4010453.
AC

[32] H. Jouini, I. Mejri, J. Martinez-Ortigosa, J.L. Cerrillo, M. Mhamdi, A.E. Palomares, G.

Delahay, T. Blasco, Selective catalytic reduction of nitric oxide with ammonia over Fe-Cu

modified highly silicated zeolites, Solid State Sci. 84 (2018) 75–85.

doi:https://doi.org/10.1016/j.solidstatesciences.2018.08.008.

[33] C. Messi, P. Carniti, A. Gervasini, Kinetics of reduction of supported nanoparticles of iron

oxide, J. Therm. Anal. Calorim. 91 (2008) 93–100. doi:10.1007/s10973-007-8427-7.

[34] M. Khachani, M. Kacimi, A. Ensuque, J.Y. Piquemal, C. Connan, F. Bozon-Verduraz, M.


ACCEPTED MANUSCRIPT
Ziyad, Iron-calcium-hydroxyapatite catalysts: Iron speciation and comparative performances

in butan-2-ol conversion and propane oxidative dehydrogenation, Appl. Catal. A Gen. 388

(2010) 113–123. doi:10.1016/j.apcata.2010.08.043.

[35] A. Sultana, M. Sasaki, K. Suzuki, H. Hamada, Tuning the NOx conversion of Cu-Fe/ZSM-5

catalyst in NH3-SCR, Catal. Commun. 41 (2013) 21–25. doi:10.1016/j.catcom.2013.06.028.

PT
RI
SC
NU
MA
T ED
C EP
AC
ACCEPTED MANUSCRIPT

Table and Figures

Table 1. Main properties of Cu-HAP and Fe-HAP catalysts.

Metal loading Surface Area a


Pore Volume b
Acidity
P T
Code
(wt.%) (m2 /g) (cm3 /g) (μmolNH3 /gcat )
R I
Main metal species e

Fe2/HAP 2.07 66.75 0.276 198.08 (1.56)c


S C
Isolated Fe3+ and Fe-O-Fe

Fe5/HAP 4.77 64.09 0.279 204.86 (1.62)c


N U Isolated Fe3+ and Fe-O-Fe

Fe7/HAP 6.83 65.68 0.247


M A
217.97 (1.72) c Isolated Fe3+ and Fe-O-Fe

Cu6/HAP 6.59 81.4 0.312

E D (Ref. 27)d Isolated Cu2+

a
determined from 2-parameters BET equation; b
T
determined at P/Po= 0.99; c as observed from PSD (BJH model from
P
desorption branch of the isotherm); cAcidity ratio of the sample compared with that of bare HAP (126.62 ±

C E
6.94μmolNH3 /gcat );d 327.3 μmolNH3 /gcat and205 μmolNH3 /gHAP ; eas determined from XRPD and UV-vis-DRS.

A C
ACCEPTED MANUSCRIPT

PT
RI
SC
Fig.1. UV-vis-DR spectra of the Cu6/HAP (a) and Fe/HAP (b) samples. Dotted lines represent UV-
NU
DRS spectra of bare HAP.
MA
T ED
EP

(a)
C
AC
ACCEPTED MANUSCRIPT

(b)

PT
RI
SC
NU
MA

Fig. 2. TPR-TPO cycles over Cu6/HAP (a) and Fe/HAP samples (b) recorded at 10°Cmin-1 to
evaluate the reduction properties of Cu- and Fe-phases (TPO profiles are not shown).
T ED
C EP
AC
ACCEPTED MANUSCRIPT

PT
RI
SC
NU
MA

Fig.3. Effect of NH3 /NO feed ratio (variable concentration of NH3 , fixed NO concentration) on the
ED

catalytic performance(NO x conversion, bars; and N 2 selectivity, round marks) of Cu6/HAP and
Fe/HAP catalysts in the NH3 -SCR reaction in oxidant atmosphere at 250°C for Cu6/HAP and at
T

300°C for Fe/HAP samples.


C EP
AC
ACCEPTED MANUSCRIPT

PT
RI
SC
NU
Fig.4.Comparison of total NO x conversion as a function of reaction temperature over Cu6/HAP and
MA

Fe5/HAP catalysts. Conditions: NO/NH3 feed ratio: 1, contact time: 0.12 s, 120 °C < T < 500°C.
T ED
C EP
AC
ACCEPTED MANUSCRIPT

PT
RI
SC
NU
MA
ED

Graphical abstract
T
C EP
AC
ACCEPTED MANUSCRIPT

Highlights
 Fe-functionalization of HAP by ion exchange produces interesting NH3 -SCR catalysts.

 High dispersion of Fe3+ species assures acid and redox centres.

 Fe/HAP catalysts require higher temperatures compared to reference Cu/HAP.

 Selectivity to N 2 is high up to the attainment of the maximum NO x conversion.

PT
RI
SC
NU
MA
T ED
C EP
AC

You might also like