Physicochemical Features and NH 3 -SCR Catalytic Performance of Natural Zeolite Modiﬁed with Iron—The Effect of Fe Loading

: In modern dual-pressure nitric acid plants, the tail gas temperature usually exceeds 300 ◦ C. The NH 3 -SCR catalyst used in this temperature range must be resistant to thermal deactivation, so commercial vanadium-based systems, such as V 2 O 5 -WO 3 (MoO 3 )-TiO 2 , are most commonly used. However, selectivity of this material signiﬁcantly decreases above 350 ◦ C due to the increase in the rate of side reactions, such as oxidation of ammonia to NO and formation of N 2 O. Moreover, vanadium compounds are toxic for the environment. Thus, management of the used catalyst is complicated. One of the alternatives to commercial V 2 O 5 -TiO 2 catalysts are natural zeolites. These materials are abundant in the environment and are thus relatively cheap and easily accessible. Therefore, the aim of the study was to design a novel iron-modiﬁed zeolite catalyst for the reduction of NO x emission from dual-pressure nitric acid plants via NH 3 -SCR. The aim of the study was to determine the inﬂuence of iron loading in the natural zeolite-supported catalyst on its catalytic performance in NO x conversion. The investigated support was ﬁrstly formed into pellets and then impregnated with various contents of Fe precursor. Physicochemical characteristics of the catalyst were determined by XRF, XRD, low-temperature N 2 sorption, FT-IR, and UV–Vis. The catalytic performance of the catalyst formed into pellets was tested on a laboratory scale within the range of 250–450 ◦ C using tail gases from a pilot nitric acid plant. The results of this study indicated that the presence of various iron species, including natural isolated Fe 3+ and the introduced FexOy oligomers, contributed to efﬁcient NO x reduction, especially in the high-temperature range, where the NO x conversion rate exceeded 90%.


Introduction
NO x emitted from stationary (power plants, nitric acid, or adipic acid production) and mobile sources are treated as a serious environmental problem. They contribute to the formation of acid rain and photochemical smog and cause deterioration of water and soil quality [1]. Therefore, it is highly necessary to reduce industrial NO x emissions. The method of NO x abatement is usually correlated with the emission origin. In the plants, which produce nitric acid, high-efficiency absorption, non-selective catalytic reduction (NSCR), selective catalytic reduction (SCR), and absorption in sodium hydroxide solution can be used [2]. Among them, selective catalytic reduction with ammonia (NH 3 -SCR) is Ammonia, used as the reducing agent, is easily available in nitric acid plants since it is a substrate in the production of HNO 3 . Typically, the SCR reactor is installed at the end of the technological line and does not significantly affect the production of acid. Therefore, NH 3 -SCR can be used in most existing nitric acid plants. The catalyst used in the NH 3 -SCR process is required to exhibit high activity in low-and high-temperature regions, satisfactory selectivity to N 2 , and good thermal stability. In fact, these requirements are met by metal oxide systems, such as the commercial catalyst V 2 O 5 -WO 3 (MoO 3 )-TiO 2 [3]. However, the material is not free from some important drawbacks, such as the toxicity of vanadium compounds. Moreover, selectivity of the catalysts above 350 • C is limited by the side reactions described by Equations (4) and (5): Due to the above-mentioned problems, a number of materials have been investigated as the alternative catalysts of NH 3 -SCR [4][5][6][7]. According to the study reported by Kobayashi et al. [8], application of TiO 2 does not provide sufficient dispersion of the active phase, surface acidity, and thermal stability of the catalyst. Therefore, further research has shifted to alternative supports of the novel catalyst. Among them, natural zeolites were found to be very promising precursors of the novel catalysts [9,10]. The great advantage of these materials is their abundance in the environment and thus their relatively low price, which is very beneficial for industrial applications. The representative of natural zeolites is clinoptilolite, belonging to the heulandite (HEU) family [11]. The material shows strongly acidic character, determined by its Si/Al molar ratio of ca. 4, its well-developed pore system, and its thermal stability. According to the Eley-Rideal mechanism, NH 3 -SCR assumes simultaneous adsorption of alkaline NH 3 and neutral NO and their interaction on the catalyst surface [12]. Therefore, high concentration of acid centers delivered by clinoptilolite improves ammonia adsorption capacity and NH 3 -SCR reaction rate. Moreover, clinoptilolite provides good ion exchange capacity, and as a consequence, its acidic character can be easily elevated by acid pretreatment [13]. Additionally, the presence of micro-and mesopores in the zeolitic structure facilitates the diffusion of gas molecules through the catalyst's pores and easy access to active centers. Lastly, clinoptilolite belongs to the residual materials, usually stored on heaps. Therefore, its recycling is in agreement with the assumptions of circular economy. All in all, the above-mentioned properties make clinoptilolite a promising candidate for the precursor of a new catalyst of NH 3 -SCR [13][14][15]. To date, research on the application of clinoptilolite was mostly limited to SCR with hydrocarbons as reducing agents. Ghasemian et al. [16] proved that protonated clinoptilolite is a promising precursor of a new catalyst of SCR with methane as a reducing agent. Another study conducted by the authors [15,17] concerned clinoptilolite as a possible support for the catalyst of SCR with propane. However, only few studies have explored zeolite as a support for SCR with ammonia [13,18].
Another important issue in the design of a novel NH 3 -SCR catalyst is the active phase. Over recent years, the focus of researchers has shifted to systems with transition metals, especially iron [5,12,19,20]. The choice of Fe was motivated by its environmentally benign characteristics, low price, and prominent thermal stability. Additionally, iron catalysts exhibit excellent medium-and high-temperature activity and satisfactory selectivity to N 2 . Moreover, the facile redox equilibrium, Fe 3+ ↔ Fe 3 O 4 ↔ Fe 2+ , contributes to high oxygen storage capacity, which is very beneficial for SCR catalysts.
Highly satisfactory activity of iron-modified clinoptilolite in SCR with ammonia was confirmed in the previous study [18]. It was found that raw clinoptilolite in a form of fine grain showed 30% of NO conversion in the range of 350-450 • C. The high efficiency of the material in NH 3 -SCR with the gas mixture reflecting the industrial composition was also confirmed. It was observed that at 400 • C, NO x conversion for Fe-clinoptilolite exceeded 80%, and high selectivity to N 2 was preserved in the entire temperature range. Additionally, 82% NO conversion was obtained for the previously shaped iron-modified clinoptilolite. In conditions similar to industrial ones, the highest catalytic activity was obtained above 400 • C, and these temperatures also maintained very favorable selectivity towards N 2 . Importantly, no formation of N 2 O was observed during the catalytic reaction.
In this work, the aim was to investigate the influence of iron loading on the lowand high-temperature catalytic performance of Fe-modified clinoptilolite formed into pellets. In this research, iron was considered the active phase since this transition metal exhibits outstanding redox properties and, at the same time, neutrality to the environment. Therefore, the experiments will contribute to the development of more ecologically friendly catalysts of the NH 3 -SCR process. Moreover, in the experiments, a real tail gas mixture, which normally enters SCR reactors in nitric acid plants, was used. To the best of our knowledge, no one so far has investigated the catalytic performance of such material under near-industrial conditions. Thus, this work makes a significant contribution to the field of low-price and nontoxic industrial catalysts of NH 3 -SCR.  Table 1. The crystalline structure of the materials was analyzed using XRD, and the obtained patterns are shown in Figure 1. As presented in Table 1, the raw clinoptilolite consisted mainly of SiO 2 and Al 2 O 3 and contained some additives of other alkaline metal oxides. Additionally, the analysis provided strong evidence of the presence of iron oxide in the natural zeolite. After protonation, the percentage contribution of SiO 2 increased with a simultaneous slight decrease of Al 2 O 3 content. This result is in line with that obtained by Burris and Juenger [21], who ascribed the decrease in aluminum content to partial dealumination of the material or its dissolution in acidic medium. However, since the XRD pattern of H-Clin corresponded to that of Clin, the degrading influence of the acid can be excluded. After the deposition of iron, the detected content of Fe 2 O 3 significantly increased, proving efficient incorporation of various iron species into the zeolite structure. However, it can be also observed that after the third impregnation, the amount of Fe 2 O 3 was very close to that obtained after the second impregnation. Additionally, the catalysts contained considerable amounts of SO 3 as the result of using FeSO 4 as the precursor of iron. Therefore, the applied calcination temperature was probably insufficient to provide effective decomposition of the salt deposited on the zeolite matrix.

Results and Discussion
Intensity (arb. u.)

Textural Properties of the Materials
Low-temperature N2 adsorption-desorption isotherms obtained for raw clinoptilolite (Clin), protonated clinoptilolite (H-Clin), and iron-modified zeolite (Fe-Clin-1, Fe-Clin-2, Fe-Clin-3) are presented in Figure 2. Furthermore, pore volume distribution is shown in Figure 3, while the textural and structural parameters of the samples are summarized in Table 2. Raw clinoptilolite demonstrated IV(a) type isotherm with the According to the results of XRD, the analyzed sample consisted mainly of heulandite/clinoptilolite, confirmed by the diffraction maxima at 2θ of 9.8, 11.4, 12.9, 16.8, 17.4, 20.7, 22.6, 30.0, 32.0, 32.9, 35.5, 36.7, and 50.3 • . The reflection at 2θ of 26.6 • corresponds to SiO 2 , while those at 21.9 and 28.1 • are due to the presence of cristobalite impurities in the solid [13]. The observed diffraction maxima are in good agreement with those reported in the literature [18,22]. The comparative analysis of the materials showed that protonation by acid treatment did not result in any noticeable structural changes. However, some of the diffraction maxima exhibited lower intensity or completely disappeared, indicating decreased crystallinity of the catalysts compared to the raw zeolite.
After modification with iron, the positions of diffraction maxima characteristic of the clinoptilolite phase remained unchanged. Thus, deposition of the active phase did not cause any significant damage to the structure. However, the intensity of the reflections was the lowest for the material with the highest concentration of iron. Larger aggregates of Fe 2 O 3 on the zeolite surface can potentially be present at 2θ of 42.0, 45.8, 60.2, and 68.2 • [23]. However, apart from bigger particles, iron also isomorphously substituted for aluminum in the zeolite framework and thus was impossible to be detected by XRD technique. The replacement of Al by Fe can be also confirmed by lower intensities of the structural diffraction maxima of clinoptilolite. A similar effect was obtained by Kessouri et al. [24] after the deposition of iron into an MFI framework. Hence, the noticeably decreased intensity of the reflections for Fe-Clin-3 can be explained by the highest rate of isomorphous substitution or deposition of bulky species of Fe 2 O 3 on its surface.

Textural Properties of the Materials
Low-temperature N 2 adsorption-desorption isotherms obtained for raw clinoptilolite (Clin), protonated clinoptilolite (H-Clin), and iron-modified zeolite (Fe-Clin-1, Fe-Clin-2, Fe-Clin-3) are presented in Figure 2. Furthermore, pore volume distribution is shown in Figure 3, while the textural and structural parameters of the samples are summarized in Table 2. Raw clinoptilolite demonstrated IV(a) type isotherm with the hysteresis loop H3, according to the IUPAC classification [25]. This isotherm is characteristic of materials with wedge-shaped mesopores and nonrigid aggregates of platelike particles [25,26]. The specific surface area of the nonmodified clinoptilolite is in range of 16-30 m 2 ·g −1 , which is typical for clinoptilolite [27]. Vassileva and Voikova [26] reported that the relatively low values of specific surface area and pore volume exhibited by nonmodified clinoptilolite are caused by the limited access of N 2 molecules to the internal structure of the zeolite. As a result, the adsorbate was deposited mainly on the external surface of the material. The results of the experiment performed after NH 3 -SCR tests showed that the specific surface area was preserved, even in the case of the catalytic reaction being conducted under severe conditions. After the dealumination procedure, the volume of mesopores in clinoptilolite significantly increased, suggesting the formation of a mesopore system. The isotherms obtained for iron-modified clinoptilolite are characterized by the isotherms of type IV(a), confirming their mesoporous nature. However, the introduction of iron resulted in a change in the shape of the hysteresis loop from H3 to H4 [28]. This result suggests the transformation of wedge-shaped mesopores into slit-shaped ones. Additionally, as presented in Table 2, after modification with iron, the specific surface area, the volume of mesopores, and the average pore diameter decreased due to pore blockage probably caused by the deposition of iron oxide species [29]. Catalysts of Fe-Clin-X series characteristically possess similar pore distribution ( Figure 3). For Fe-Clin-X series, a wide bimodal pore distribution-pores with diameters ranging from 10 to 1000 Å and pores with diameters from 1500 to 7000 Å-was observed. However, the porous structure was definitely dominated by pores with diameters ranging from 10 to 1000 Å (mesopores with D meso in the range of 208-223 Å). Interestingly, multiple impregnations with FeSO 4 did not result in considerable differences between the D meso values. Nevertheless, the gradual decline of V meso with the increasing iron content suggested that iron species were effectively deposited in the inner structure of clinoptilolite.
hysteresis loop H3, according to the IUPAC classification [25]. This isotherm is charac istic of materials with wedge-shaped mesopores and nonrigid aggregates of platelike p ticles [25,26]. The specific surface area of the nonmodified clinoptilolite is in range of 30 m 2 ·g −1 , which is typical for clinoptilolite [27]. Vassileva and Voikova [26] reported t the relatively low values of specific surface area and pore volume exhibited by nonmo fied clinoptilolite are caused by the limited access of N2 molecules to the internal struct of the zeolite. As a result, the adsorbate was deposited mainly on the external surface the material. The results of the experiment performed after NH3-SCR tests showed t the specific surface area was preserved, even in the case of the catalytic reaction be conducted under severe conditions. After the dealumination procedure, the volume mesopores in clinoptilolite significantly increased, suggesting the formation of a me pore system. The isotherms obtained for iron-modified clinoptilolite are characterized the isotherms of type IV(a), confirming their mesoporous nature. However, the introd tion of iron resulted in a change in the shape of the hysteresis loop from H3 to H4 [2 This result suggests the transformation of wedge-shaped mesopores into slit-shaped on Additionally, as presented in Table 2, after modification with iron, the specific surf area, the volume of mesopores, and the average pore diameter decreased due to p blockage probably caused by the deposition of iron oxide species [29]. Catalysts of Clin-X series characteristically possess similar pore distribution ( Figure 3). For Fe-Clin series, a wide bimodal pore distribution-pores with diameters ranging from 10 to 1 Å and pores with diameters from 1500 to 7000 Å-was observed. However, the poro structure was definitely dominated by pores with diameters ranging from 10 to 1000 (mesopores with Dmeso in the range of 208-223 Å). Interestingly, multiple impregnati with FeSO4 did not result in considerable differences between the Dmeso values. Nevert less, the gradual decline of Vmeso with the increasing iron content suggested that iron s cies were effectively deposited in the inner structure of clinoptilolite.    a Specific surface area determined using the BET method; b external surface area determined using the t-plot method; c average mesopore volume and diameter determined using the BJH method.

Characteristic Chemical Groups in the Materials
The FT-IR spectra obtained for the raw and protonated clinoptilolite and the zeolites with various loadings of iron are presented in In the first of the above-mentioned regions, 3800-3400 cm −1 , the shape of the spectra was similar for all of the materials except Fe-Clin-3. The peak at 3650 cm −1 suggested the presence of Brönsted sites provided by the acidic hydroxyl Si-O(H)-Al. In the case of Fe-Clin-3, it was not as sharp as for the other materials; thus, multiple repetition of Fe deposition resulted in the removal of OH groups bonded to the zeolitic structure. The band at 3400 cm −1 , broad for raw and protonated clinoptilolite and sharper for Fe-modified zeolite, corresponded to the vibrations of O-H⋯O bonds [30].
In the second region of the spectra, 1700-700 cm −1 , the peak at 1650 cm −1 , attributed to deformation vibrations of physisorbed water molecules, showed a similar shape for all materials [31]. However, the characteristic bands at 1200 cm −1 and 1050 cm −1 were almost absent for clinoptilolite modified with Fe. Both of the peaks were related to Al-O or Si-O asymmetric stretching vibrations; thus, incorporation of iron resulted in structural interruptions, such as the removal of charge-balancing Ca 2+ and Mg 2+ . A similar effect was observed by Cobzaru et al. [32] after the modification of natural clinoptilolite with nitric acid. Since FeSO4 is regarded as a strongly acidic medium, our results are in line with this research. Moreover, the band at 1150 cm −1 , intense for Clin and H-Clin, partially disappeared after the introduction of iron. Since the peak corresponds to three-dimensional networks of amorphous Si-O-Si units, the modification procedure could partially remove this phase from natural and protonated zeolite. Additionally, a small peak at 1385 cm −1 , appearing  a Specific surface area determined using the BET method; b external surface area determined using the t-plot method; c average mesopore volume and diameter determined using the BJH method.

Characteristic Chemical Groups in the Materials
The FT-IR spectra obtained for the raw and protonated clinoptilolite and the zeolites with various loadings of iron are presented in only for Fe-Clin-3 and thus with the highest concentration of iron species, was probably ascribed to sulfate groups bonded to iron ions deposited in the zeolitic structure [33]. The characteristic peaks detected in the third analyzed region, 700-450 cm −1 , evidenced partial removal of amorphous silica. This effect can be confirmed by the presence of the sharp peaks at 800 cm −1 in the spectra of all the materials. However, after modification with iron, these peaks were noticeably separated. The new small peak at 780 cm −1 , formed through this division, raised from the stretching vibrations of [SiO4] tetrahedra from the zeolitic framework. Therefore, removal of the amorphous silica could enhance the detection of structural peaks of the materials. Another difference in the spectra of ironmodified clinoptilolite compared to the raw or protonated form was the presence of lowintense peaks at 585 cm −1 , which corresponded to the symmetric stretching vibrations of [AlO4] tetrahedra [34]. Two bands at 600 cm −1 and 475 cm −1 were related to O-Al-O or O-Si-O bending vibrations and Si-O stretching vibrations, respectively [35]. The comparative UV-Vis spectra of the raw clinoptilolite and the catalysts with various contents of iron are presented in Figure 5.  In the first of the above-mentioned regions, 3800-3400 cm −1 , the shape of the spectra was similar for all of the materials except Fe-Clin-3. The peak at 3650 cm −1 suggested the presence of Brönsted sites provided by the acidic hydroxyl Si-O(H)-Al. In the case of Fe-Clin-3, it was not as sharp as for the other materials; thus, multiple repetition of Fe deposition resulted in the removal of OH groups bonded to the zeolitic structure. The band at 3400 cm −1 , broad for raw and protonated clinoptilolite and sharper for Fe-modified zeolite, corresponded to the vibrations of O-H· · · O bonds [30].
In the second region of the spectra, 1700-700 cm −1 , the peak at 1650 cm −1 , attributed to deformation vibrations of physisorbed water molecules, showed a similar shape for all materials [31]. However, the characteristic bands at 1200 cm −1 and 1050 cm −1 were almost absent for clinoptilolite modified with Fe. Both of the peaks were related to Al-O or Si-O asymmetric stretching vibrations; thus, incorporation of iron resulted in structural interruptions, such as the removal of charge-balancing Ca 2+ and Mg 2+ . A similar effect was observed by Cobzaru et al. [32] after the modification of natural clinoptilolite with nitric acid. Since FeSO 4 is regarded as a strongly acidic medium, our results are in line with this research. Moreover, the band at 1150 cm −1 , intense for Clin and H-Clin, partially disappeared after the introduction of iron. Since the peak corresponds to three-dimensional networks of amorphous Si-O-Si units, the modification procedure could partially remove this phase from natural and protonated zeolite. Additionally, a small peak at 1385 cm −1 , appearing only for Fe-Clin-3 and thus with the highest concentration of iron species, was probably ascribed to sulfate groups bonded to iron ions deposited in the zeolitic structure [33].
The characteristic peaks detected in the third analyzed region, 700-450 cm −1 , evidenced partial removal of amorphous silica. This effect can be confirmed by the presence of the sharp peaks at 800 cm −1 in the spectra of all the materials. However, after modification with iron, these peaks were noticeably separated. The new small peak at 780 cm −1 , formed through this division, raised from the stretching vibrations of [SiO 4 ] tetrahedra from the zeolitic framework. Therefore, removal of the amorphous silica could enhance the detection of structural peaks of the materials. Another difference in the spectra of iron-modified clinoptilolite compared to the raw or protonated form was the presence of low-intense peaks at 585 cm −1 , which corresponded to the symmetric stretching vibrations of [AlO 4 ] tetrahedra [34]. Two bands at 600 cm

Speciation of the Active Phase
The comparative UV-Vis spectra of the raw clinoptilolite and the catalysts with various contents of iron are presented in Figure 5.
In general, for iron-modified zeolites, three main regions in UV-Vis spectra are expected: (1) bands below 300 nm, corresponding to the oxygen-to-metal charge transfer (CT), assigned to isolated framework and extra framework pseudotetrahedral Fe 3+ species; (2) bands in the range of 300-500 nm, related to oligomeric Fe x O y or Fe 2 O 3 nanoparticles; and (3) bands detected above 500 nm, assigned to Fe 2 O 3 clusters on the external surface of the support [19].
The speciation of iron in the analyzed materials was strongly correlated with the metal loading. As presented in Figure 5, all the investigated samples, including nonmodified clinoptilolite, showed absorption bands at 230 and 260 nm. This result confirmed that iron was originally present (Clin) or isomorphously deposited (Fe-Clin-1, 2, and 3) in the zeolite structure in the form of extraframework cations with octahedral coordination [36]. Furthermore, the band at 350 nm, detected only for the zeolite modified with iron, corresponded to small, oligonuclear clusters of iron oxide [37]. The bands at 475 nm, characteristic of bigger particles of Fe 2 O 3 , were observed for the samples with increased iron content (Fe-Clin-2, Fe-Clin-3). Therefore, the extraframework phase of Fe 2 O 3 became dominant as a result of the increase in iron loading due to the agglomeration of the species into bigger particles. (Fe-Clin-2, Fe-Clin-3). Therefore, the extraframework phase of Fe2O3 became dominant as a result of the increase in iron loading due to the agglomeration of the species into bigger particles.

NH3-SCR Catalytic Tests Performed with Industrial Gas Mixture
NH3-SCR catalytic tests over protonated clinoptilolite were conducted under the conditions reflecting that of industrial nitric acid plant (regarding catalytic bed loading and temperature range). The obtained results are presented in Figure 6. The tests were carried out at two catalytic bed loads. It was observed that in both cases, NOx conversion exceeded 50% in the entire temperature range. The maximum conversion of more than 90%, was reached above 400 °C, and higher NOx conversion was achieved for the lower catalytic bed loading. In the case of GHSV = 4500 h −1 (tail gas flow 0.15 Nm 3 ·h −1 ), 93% of NOx conversion was obtained at 400 °C. On the other hand, for GHSV = 9000 h −1 (tail gas flow 0.3 Nm 3 ·h −1 ), the material exhibited 80% of NOx conversion at 450 °C.
During the test, N2O concentration upstream and downstream of the catalytic bed was measured as well. The dash line in Figure 7 represents the ratio of the N2O concentration downstream to the inlet concentration of N2O. It was clearly indicated that higher loading of the catalytic bed resulted in lower N2O concentration downstream of the bed compared to the inlet concentration of N2O. This effect was observed over almost the entire investigated temperature range. Moreover, regardless of the catalytic bed load, the highest selectivity was observed at 450 °C.

NH 3 -SCR Catalytic Tests Performed with Industrial Gas Mixture
NH 3 -SCR catalytic tests over protonated clinoptilolite were conducted under the conditions reflecting that of industrial nitric acid plant (regarding catalytic bed loading and temperature range). The obtained results are presented in Figure 6. The tests were carried out at two catalytic bed loads. It was observed that in both cases, NO x conversion exceeded 50% in the entire temperature range. The maximum conversion of more than 90%, was reached above 400 • C, and higher NO x conversion was achieved for the lower catalytic bed loading. In the case of GHSV = 4500 h −1 (tail gas flow 0.15 Nm 3 ·h −1 ), 93% of NO x conversion was obtained at 400 • C. On the other hand, for GHSV = 9000 h −1 (tail gas flow 0.3 Nm 3 ·h −1 ), the material exhibited 80% of NO x conversion at 450 • C.  Figure 8 shows the results of the catalytic tests obtained for iron-modified samples and the protonated clinoptilolite, while the selectivity of the materials to N2 is listed in Table 3. In all cases, NOx conversion of iron-modified zeolite was higher than that of H-Clin. Above 350 °C, regardless of the iron content in the sample, NOx conversion of over 90% was achieved. The highest activity in the entire temperature range was exhibited by Fe-Clin-2. Additionally, selectivity of Fe-clinoptilolite catalysts to N2 was in the range of 93-100%, confirming the negligible contribution of the side reactions to the whole mechanism of NH3-SCR performed on the materials.  Figure 8 shows the results of the catalytic tests obtained for iron-modified samples and the protonated clinoptilolite, while the selectivity of the materials to N 2 is listed in Table 3. In all cases, NO x conversion of iron-modified zeolite was higher than that of H-Clin. Above 350 • C, regardless of the iron content in the sample, NO x conversion of over 90% was achieved. The highest activity in the entire temperature range was exhibited by Fe-Clin-2. Additionally, selectivity of Fe-clinoptilolite catalysts to N 2 was in the range of 93-100%, confirming the negligible contribution of the side reactions to the whole mechanism of NH 3 -SCR performed on the materials.    The N 2 O concentrations measured during the experiments are shown in Figure 9. In the case of protonated clinoptilolite, the N 2 O concentration increased above the inlet value only at 350 • C. For iron-modified samples, the courses of the curves are similar to each other. Up to the temperature of 400 • C, the concentration of N 2 O behind the bed slightly increased in relation to the initial concentration (N 2 O/N 2 O (in) > 1), and then, a sharp decrease in the concentration of N 2 O at the temperature of 450 • C was noted. The greatest decrease was obtained for the Fe-Clin-1 and Fe-Clin-2 samples. Overall, satisfactory catalytic performance exhibited by the investigated catalysts confirmed that one or two iron impregnations of clinoptilolite are sufficient to obtain an effective NH 3 -SCR catalyst.

Catalysts Preparation
The precursor of the investigated catalysts was raw zeolite with a high content of clinoptilolite phase. Firstly, the material was dealuminated using 5% HNO3 solution. The operation was repeated three times in order to increase the dealumination rate. After each dealumination step, the precursor was washed with demineralized water until pH was <6 and dried at 105-110 °C. Subsequently, the zeolite was fractioned into 0.3-0.8 mm grains and formed into pellets of 5.0 × 4.8 mm dimensions, illustrated in Figure 10. Afterwards, the materials were calcined at 450 °C for 2 h. Iron-modified materials were prepared using the wet impregnation method using an aqueous solution of 1 M FeSO4 as Fe precursor. The samples were left in contact with the solution at 50 °C for 1 h, then dried at 105-110 °C and calcined at 500 °C for 2 h. The impregnation procedure was performed one, two, or three times in order to obtained catalysts with various Fe loadings The precursors were dried and calcined before each impregnation treatment. The preparation procedure is schematically illustrated in Figure 10.

Catalysts Preparation
The precursor of the investigated catalysts was raw zeolite with a high content of clinoptilolite phase. Firstly, the material was dealuminated using 5% HNO 3 solution. The operation was repeated three times in order to increase the dealumination rate. After each dealumination step, the precursor was washed with demineralized water until pH was <6 and dried at 105-110 • C. Subsequently, the zeolite was fractioned into 0.3-0.8 mm grains and formed into pellets of 5.0 × 4.8 mm dimensions, illustrated in Figure 10. Afterwards, the materials were calcined at 450 • C for 2 h. Iron-modified materials were prepared using the wet impregnation method using an aqueous solution of 1 M FeSO 4 as Fe precursor. The samples were left in contact with the solution at 50 • C for 1 h, then dried at 105-110 • C and calcined at 500 • C for 2 h. The impregnation procedure was performed one, two, or three times in order to obtained catalysts with various Fe loadings The precursors were dried and calcined before each impregnation treatment. The preparation procedure is schematically illustrated in Figure 10.
the materials were calcined at 450 °C for 2 h. Iron-modified materials were prepared usin the wet impregnation method using an aqueous solution of 1 M FeSO4 as Fe precurso The samples were left in contact with the solution at 50 °C for 1 h, then dried at 105-1 °C and calcined at 500 °C for 2 h. The impregnation procedure was performed one, tw or three times in order to obtained catalysts with various Fe loadings The precursors we dried and calcined before each impregnation treatment. The preparation procedure schematically illustrated in Figure 10. The formed samples of protonated clinoptilolite and Fe-clinoptilolite catalysts, pr pared on a laboratory scale, are presented in Figure 11A,B, respectively. The codes of th samples with the corresponding descriptions are listed in Table 4. The formed samples of protonated clinoptilolite and Fe-clinoptilolite catalysts, prepared on a laboratory scale, are presented in Figure 11A,B, respectively. The codes of the samples with the corresponding descriptions are listed in Table 4.  The catalyst obtained by single impregnation with Fe precursor Fe-Clin-2 The catalyst obtained by dual impregnation with Fe precursor Fe-Clin-3 The catalyst obtained by triple impregnation with Fe precursor

Catalysts Characterization
X-ray fluorescence (XRF) was used to determine the chemical composition of the samples using Energy Dispersive X-ray Fluorescence EDXRF Spectrometer, Epsilon 3XLE PANalytical Company. The crystalline structure of the samples was analyzed using an Xray diffraction (XRD) technique. X-ray diffraction patterns were obtained using an Empyrean diffractometer (Panalytical) equipped with a copper-based anode (Cu-Kα LFF HR, λ = 0.154059 nm). The measurement was conducted in the 2θ range of 2.0-70.0° (2θ step scans of 0.02° and the counting time of 1 s per step). The specific surface area, total pore volume, and mesopore volume were determined using an ASAP ® 2050 Xtended Pressure sorption analyzer (Micromeritics Instrument Co., Norcross, GA, USA) based on N2 adsorption-desorption isotherms at −196 °C using the BET adsorption model (Brunauer-Emmett-Teller) and the BJH transformation (Barret-Joyner-Halenda). Fourier transform infrared spectroscopy studies (FT-IR) were conducted using a Perkin Elmer Frontier FT-IR spectrometer. The spectra were obtained in the wavelength range of 4000-400 cm −1 with a resolution of 4 cm −1 . Before each measurement, the sample was mixed with KBr in a ratio of 1: 100 and pressed into a disk. Coordination and aggregation of iron species were determined by UV-Vis spectroscopy at a wavelength range of 200-900 nm with a resolution of 1 nm using a Perkin Elmer Lambda 35 UV-Vis spectrophotometer.

Catalytic Tests in Real Gas Conditions
The activity and selectivity of the catalysts in the NH3-SCR process were tested in the laboratory installation in the flow of the tail gases stream derived from the pilot ammonia oxidation plant. The laboratory installation consisted of a reactor (R) with a diameter of 25 mm and heat exchangers (HEx and HExNH3) used for preheating tail gases and ammonia. In the tests, the height of the catalyst layer was 70 mm. The installation is schematically presented in Figure 12. The heated tail gases were mixed with ammonia and turned into the catalytic bed. The composition of the tail gases was similar to the tail gases emitted  The catalyst obtained by single impregnation with Fe precursor Fe-Clin-2 The catalyst obtained by dual impregnation with Fe precursor Fe-Clin-3 The catalyst obtained by triple impregnation with Fe precursor

Catalysts Characterization
X-ray fluorescence (XRF) was used to determine the chemical composition of the samples using Energy Dispersive X-ray Fluorescence EDXRF Spectrometer, Epsilon 3XLE PANalytical Company. The crystalline structure of the samples was analyzed using an X-ray diffraction (XRD) technique. X-ray diffraction patterns were obtained using an Empyrean diffractometer (Panalytical) equipped with a copper-based anode (Cu-Kα LFF HR, λ = 0.154059 nm). The measurement was conducted in the 2θ range of 2.0-70.0 • (2θ step scans of 0.02 • and the counting time of 1 s per step). The specific surface area, total pore volume, and mesopore volume were determined using an ASAP ® 2050 Xtended Pressure sorption analyzer (Micromeritics Instrument Co., Norcross, GA, USA) based on N 2 adsorption-desorption isotherms at −196 • C using the BET adsorption model (Brunauer-Emmett-Teller) and the BJH transformation (Barret-Joyner-Halenda). Fourier transform infrared spectroscopy studies (FT-IR) were conducted using a Perkin Elmer Frontier FT-IR spectrometer. The spectra were obtained in the wavelength range of 4000-400 cm −1 with a resolution of 4 cm −1 . Before each measurement, the sample was mixed with KBr in a ratio of 1:100 and pressed into a disk. Coordination and aggregation of iron species were determined by UV-Vis spectroscopy at a wavelength range of 200-900 nm with a resolution of 1 nm using a Perkin Elmer Lambda 35 UV-Vis spectrophotometer.

Catalytic Tests in Real Gas Conditions
The activity and selectivity of the catalysts in the NH 3 -SCR process were tested in the laboratory installation in the flow of the tail gases stream derived from the pilot ammonia oxidation plant. The laboratory installation consisted of a reactor (R) with a diameter of 25 mm and heat exchangers (HEx and HExNH 3 ) used for preheating tail gases and ammonia. In the tests, the height of the catalyst layer was 70 mm. The installation is schematically presented in Figure 12. The heated tail gases were mixed with ammonia and turned into the catalytic bed. The composition of the tail gases was similar to the tail gases emitted from industrial nitric acid plants; consisted of NO, NO 2 , N 2 O, O 2 , N 2 , and H 2 O; and contained approximately 900-1100 ppm of NO x , (NO/NO 2 = 2-2.6) 400-600 ppm of N 2 O, 2-3 vol.%. of O 2 , and 0.3-0.5 vol.% of H 2 O. The amount of NH 3 used in the reaction was increased and optimized to the level providing maximum NO x conversion with minimal NH 3 slip (less than 10 ppm). Thus, the NH 3 concentration was maintained at 0.14-0.15 vol.%, depending on the inlet NO x concentration.
Catalysts 2022, 12, x FOR PEER REVIEW 13 of 16 increased and optimized to the level providing maximum NOx conversion with minimal NH3 slip (less than 10 ppm). Thus, the NH3 concentration was maintained at 0.14-0.15 vol.%, depending on the inlet NOx concentration. In each test, 30 g of the catalyst in a form of pellet (d = 5.0 × 4.8 mm) was placed into the reactor. The activity studies were performed at 250, 300, 350, 400, and 450 °C. The temperature inside the reactor was controlled by a thermocouple installed at the gas outlet from the bed. The research was conducted in GHSV = 4500 and 9000 h −1 (tail gas flow 0.15 and 0.3 Nm 3 ·h −1 , respectively). Temperature, GHSV, and shape of catalyst were selected to be as close as possible to conditions prevailing in industrial plants. The measurements of the inlet and outlet concentrations of NO, NO2, N2O, and NH3 were conducted at each temperature after stabilizing the equilibrium conditions and operating parameters. The concentrations of unreacted NO, NO2, and N2O were analyzed downstream of the reactor by a GASMET FT-IR analyzer (Vantaa, Finland). NOx reduction was important in this study; thus, NO and NO2 concentrations were not considered separately. NOx conversion was calculated according to Equation (6): where X -NOx conversion, NO in -inlet concentration of NOx, while NO -NOx concentration in the gas after catalytic reaction. In each test, 30 g of the catalyst in a form of pellet (d = 5.0 × 4.8 mm) was placed into the reactor. The activity studies were performed at 250, 300, 350, 400, and 450 • C. The temperature inside the reactor was controlled by a thermocouple installed at the gas outlet from the bed. The research was conducted in GHSV = 4500 and 9000 h −1 (tail gas flow 0.15 and 0.3 Nm 3 ·h −1 , respectively). Temperature, GHSV, and shape of catalyst were selected to be as close as possible to conditions prevailing in industrial plants. The measurements of the inlet and outlet concentrations of NO, NO 2 , N 2 O, and NH 3 were conducted at each temperature after stabilizing the equilibrium conditions and operating parameters. The concentrations of unreacted NO, NO 2 , and N 2 O were analyzed downstream of the reactor by a GASMET FT-IR analyzer (Vantaa, Finland). NO x reduction was important in this study; thus, NO and NO 2 concentrations were not considered separately. NO x conversion was calculated according to Equation (6):

Conclusions
where X NO x -NO x conversion, NO x (in)-inlet concentration of NO x , while NO x -NO x concentration in the gas after catalytic reaction.

Conclusions
This paper has demonstrated the catalytic potential of protonated or protonated and iron-modified clinoptilolite in the form of pellets in NH 3 -SCR within the range of 250-450 • C. Pretreatment with HNO 3 and deposition of iron changed the shape of mesopores and resulted in the formation of secondary porosity. Additionally, deposition of iron caused some interruptions in the order of the zeolite framework. Nevertheless, the crystallinity was not affected by the performed modifications. Catalytic tests were conducted using a gas mixture which reflected industrial conditions. For H-Clin, the maximum conversion of NO x of over 90% was achieved above 400 • C and GHSV = 4500 h −1 . At the load of 9000 h −1 , the conversion of NO x reached more than 60% in the entire temperature range. Satisfactory results obtained for the protonated zeolite without the addition of the active phase can be explained by the natural presence of iron species in the clinoptilolite structure. Regardless of iron loading, NO x conversion obtained for the catalysts was higher than that of the H-Clin. In the case of Fe-Clin-1 and Fe-Clin-2, NO x conversion exceeded 90% above 350 • C. Slightly lower NO x reduction was recorded for Fe-Clin-3. In summary, it was demonstrated that even a single impregnation of natural zeolite (Fe-Clin-1) resulted in the satisfactory catalytic performance, since more than 90% of NO x conversion was achieved between 350-450 • C. Additionally, it was noted that N 2 O concentration decreased by 20% compared to the initial concentration. The strength and significance of our work lies especially in the minimization of the catalyst preparation steps, which is highly beneficial from technological and economical points of view. In summary, it was demonstrated that Fe-clinoptilolite catalysts are advantageous, low-cost, and easy-to-prepare materials that exhibit satisfactory features in the NH 3 -SCR process.