The quality and chemistry of the Earth's atmosphere are critical to the future of human and mammalian life. Since the beginning of humankind's industrial activity, the chemical composition of the atmosphere has changed due to the release of volatile pollutants and greenhouse gases (Fowler, Reference Fowler, Brimblecombe, Burrows, Heal, Grennfelt and Stevenson2020). Ammonia (NH3), an irritating, malodorous and colourless gas, is one of these pollutants. Ammonia is used as an ingredient in many commercial cleaning and pharmaceutical products, as a hydrogen carrier and as a fertilizer (Kobayashi, Reference Kobayashi, Hayakawa, Somarathne and Okafor2019; Sun et al., Reference Sun, Hong, Cai, Zhang, Kan and Zhao2021) and for selective catalytic reduction of NOx (Li et al., Reference Li, Chang, Ma, Hao and Yang2011; Wang et al., Reference Wang, Zhao, Haller and Li2017). Large amounts of ammonia are released into the atmosphere from livestock farming and agricultural activities (Ciahotný, Reference Ciahotný, Melenová, Jirglová, Boldiš and Kočiřík2002). This ammonia release can be taken up by atmospheric moisture and surface water and also accumulate in plants and soil (Renard, Reference Renard, Calidonna and Henley2004). Changes in atmospheric ammonia concentrations are known to have adverse effects on the environment (Amon, Reference Amon, Dobeic, Sneath, Phillips, Misselbrook and Pain1997). In addition, exposure to certain levels of ammonia can be extremely harmful to human health. Inhaled ammonia is mainly absorbed by water in human tissues, denaturing proteins and eventually destroying cell membranes (Sun et al., Reference Sun, Hong, Cai, Zhang, Kan and Zhao2021). This can cause nausea, coughing, dizziness, pulmonary oedema and weakening of the immune system (Lindgren, Reference Lindgren2010; Sun et al., Reference Sun, Hong, Cai, Zhang, Kan and Zhao2021). Therefore, indoor ammonia concentrations can also pose a threat to the health of workers. For example, workers who work in an environment with high levels of ammonia are at risk of developing chronic respiratory diseases such as bronchial asthma (Ballal, Reference Ballal, Ali, Albar, Ahmed and Al-Hasan1998). Ammonia can also affect the reproductive functioning of female workers (Sun et al., Reference Sun, Hong, Cai, Zhang, Kan and Zhao2021). Animals in livestock buildings are also affected by the presence of ammonia in the environment. Hence, inflammatory responses are observed in the respiratory system of pigs exposed to ammonia concentrations of 100 and 150 ppm (Drummond, Reference Drummond, Curtis, Simon and Norton1980). It is therefore clear that indoor ammonia levels also need to be controlled.
Many porous adsorbents such as metal–organic frameworks, covalent organic frameworks, hydrogen-bonded organic frameworks, porous organic polymers and their composite materials have been studied for their ability to remove ammonia (Won Kang et al., Reference Won Kang, Eungyung Ju, Won Kim, Kang, Kim and Seop Hong2020). One such adsorbent is zeolite, an Al-silicate mineral found in nature or synthesized in the laboratory. The framework of the zeolites is formed by the combination of tetrahedral silicate [SiO4]4– units. Combinations of these units form channels or networks. During the formation of zeolites, the isomorphic substitution of a trivalent cation, namely Al3+ or Ga3+ for Si4+, creates a negative charge that is balanced by the presence of exchangeable cations such as Na+, Mg2+ and Ca2+, amongst others (Gottardi & Galli, Reference Gottardi and Galli1985). Clinoptilolite (CLN), a natural zeolite, is a member of the heulandite (HEU) group. The general formula of CLN is (Na,K)6(Al6Si30O72)⋅20H2O. Its framework structure of monoclinic C2/m symmetry with the unit cell parameters a = 17.62 Å, b = 17.91 Å, c = 7.39 Å and β = 116°16′ is almost identical to that of HEU. However, CLN has a higher Si/Al ratio (≥4) and is more thermally stable than HEU (Mumpton, Reference Mumpton1960; Ward & McKague, Reference Ward and McKague1994). CLN has a two-dimensional channel network (10-membered A and 8-membered B channels run along the a-axis whilst 8-membered C channels intersect them along the c-axis; Ambrozova, Reference Ambrozova, Kynicky, Urubek and Nguyen2017). In addition to its catalytic (Dziedzicka et al., Reference Dziedzicka, Sulikowski and Ruggiero-Mikołajczyk2016) and medical (Mastinu et al., Reference Mastinu, Kumar, Maccarinelli, Bonini, Premoli and Aria2019) applications, CLN is used for the removal of heavy metal ions (Zendelska et al., Reference Zendelska, Golomeova, Jakupi, Lisičkov, Kuvendžiev and Marinkovski2018; Elboughdiri, Reference Elboughdiri2020; El-Arish et al., Reference El-Arish, Zaki, Miskan, Setiabudi and Jaafar2022), of environmental pollutants from wastewater (Shamshiri et al., Reference Shamshiri, Alimohammadi, Sedighi, Jabbari and Mohammadi2022) and of toxic gases from air (Macala et al., Reference Macala, Pandova and Panda2009; Karousos et al., Reference Karousos, Sapalidis, Kouvelos, Romanos and Kanellopoulos2016; Ghahri et al., Reference Ghahri, Golbabaei, Vafajoo, Mireskandari, Yaseri and Shahtaheri2017; Senila, Reference Senila, Neag, Cadar, Hoaghia, Roman and Moldovan2022).
It is common to apply chemical processes such as treatment with acid (H2SO4, H3PO4 and HNO3) and salt (KNO3, NaNO3 and Mg(NO3)2) solutions to improve the physicochemical and gas adsorption properties of natural zeolites (Christidis et al., Reference Christidis, Moraetis, Keheyan, Akhalbedashvili, Kekelidze and Gevorkyan2003; Ciahotný et al., Reference Ciahotný, Melenová, Jirglová, Pachtová, Kočiřík and Eić2006; Erdoğan Alver & Sakızcı Reference Erdoğan Alver and Sakızcı2019). However, even at low molarities, dealumination during hydrochloric acid treatment causes the crystal structure to collapse rapidly (Christidis et al., Reference Christidis, Moraetis, Keheyan, Akhalbedashvili, Kekelidze and Gevorkyan2003; Garcia-Basabe et al., Reference Garcia-Basabe, Rodriguez-Iznaga, de Menorval, Llewellyn, Maurin and Lewis2010). An alternative method to modify the structure of CLN is calcination after ammonium (NH4+) exchange. In this process, raw CLN is treated with ammonium salt solution, and the obtained product (or sample) is calcined at temperatures of 400–600°C for 2–8 h. In this way, ammonium ions adsorbed from the salt solution decompose into ammonia and hydrogen ions. As a result, H+-CLN with a higher surface area than the raw material can be obtained without disrupting the structure (Rožić et al., Reference Rožić, Cerjan-Stefanović, Kurajica, Maěefat, Margeta and Farkaš2005; Allen et al., Reference Allen, Ivanova and Koumanova2009). In the literature, there are many studies in which H+ forms of CLN were obtained using different calcination temperatures and various molarities of ammonium salt solutions (Kurama et al., Reference Kurama, Zimmer and Reschetilowski2002; Elysabeth et al., Reference Elysabeth, Zulnovri, Setiadi and Slamet2019; Liao et al., Reference Liao, Zhang, Fan, Chang and Bao2019; Hieu et al., Reference Hieu, Kosslick, Riaz, Schulz, Springer and Frank2022). However, it remains to be investigated how the ammonia adsorption properties of CLN in the H+ form obtained after calcination change after modification with different cations. Therefore, the main objective of this study is to determine the ammonia adsorption capacity of H+-CLN synthesized by calcination after direct and indirect ammonium nitrate exchange and to select the most suitable parent sample in terms of ammonia retention. As a second objective of this study, the effect of doping the selected parent sample with K+, Na+, Mg2+ and Ca2+ cations on the ammonia adsorption efficiency was investigated.
Experimental
Materials and methods
Gördes CLN (Esenli & Sirkecioğlu, Reference Esenli and Sirkecioğlu2005; Esenli et al., Reference Esenli, Ekinci Şans, Erdoğan and Sirkecioğlu2023) was sieved with a sieving machine (Retsch, Germany) to obtain <125 μm fraction and split to 5.0 g aliquots. To remove soluble impurities, each zeolite aliquot was kept in 100 mL deionized water at 80°C for 4 h. All samples were then separated and washed several times with hot distilled water. Two different methods were used to synthesize the H+-zeolites (Fig. 1). In Method 1, by which H+ forms were obtained directly, the samples were modified with 100 mL of 0.5 and 1.0 M NH4NO3 solutions at 80°C for 4 and 8 h, respectively. The H+ forms obtained by calcining these samples at 400°C for 6 h were labelled as 0.5-4h-CLN, 0.5-8h-CLN, 1.0-4h-CLN and 1.0-8h-CLN (Fig. 1).
Schematic diagram of the sample preparation procedure.
In Method 2, by which H+ forms were obtained indirectly, CLN samples were first modified with 100 mL of 1.0 M NaNO3 solution at 80°C for 4 h. All samples were then filtered, washed several times with hot distilled water and dried at room temperature. These samples were then modified with 100 mL of 0.5 and 1.0 M NH4NO3 solutions for 4 and 8 h and calcined at 400°C for 6 h to indirectly obtain the H+-zeolites. These samples were labelled as Na-0.5-4h-CLN, Na-0.5-8h-CLN, Na-1.0-4h-CLN, and Na-1.0-8h-CLN (Fig. 1). For all cation-exchange procedures, 5.0 g of the CLN sample was used per 100 mL of solution. Ammonia adsorption measurements of these different H+-CLN samples obtained using both methods were carried out at 298 K. The Na-1.0-8h-CLN sample with the highest ammonia adsorption capacity among the H-forms (prepared using Method 2) was selected as the main CLN for further experiments. In the second step after the selection of the parent sample, 5.0 g of each parent CLN sample was exchanged with 0.5 and 1.0 M NaNO3, KNO3, Ca(NO3)2 and Mg(NO3)2 solutions at 80°C for 4 h to investigate the effect of doping the structure of this sample with different cations on the ammonia adsorption capacity. The cation-exchanged samples of the parent sample were then separated and washed several times with deionized water at boiling point, and the dried samples were kept in an oven at 100°C for 12 h and stored in a desiccator. Finally, the cation-exchanged samples of the parent sample were labelled as 0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN, 0.5-Mg-CLN, 1.0-Na-CLN, 1.0-K-CLN, 1.0-Ca-CLN or 1.0 Mg-CLN depending on the salt solution used in the cation-exchange process.
Instrumentation
The original and cation-exchanged CLN samples were characterized by X-ray fluorescence (XRF), X-ray diffraction (XRD), scanning electron microscopy (SEM) and N2 adsorption techniques. Elemental analyses of the samples were performed using a Rigaku ZSX Primus instrument. Loss on ignition (LOI) was determined by mass measurement after heating at 1000°C at a heating rate of 10°C min–1, before being allowed to stand for 1 h and then cooled to room temperature at the same rate. Powder XRD traces were obtained on a Bruker D8 Advance instrument using Cu-Kα (λ = 1.54 Å) radiation at 40 kV and 40 mA in the range 5–40°2θ, with a scanning step of 0.02°2θ. SEM images were recorded with a Zeiss Ultra Plus field emission scanning electron microscope (FE-SEM) at a 5 kV acceleration voltage. All samples were gold coated prior to analysis. Specific surface areas and micropore data were obtained from the N2 adsorption isotherms. Ammonia adsorption isotherms were measured at 298 K to 100 kPa for all samples. The N2 and NH3 adsorption analyses of the CLNs were performed using a 3Flex volumetric apparatus (Micromeritics) after degassing at 300°C for 8 h.
Results and discussion
Elemental analysis
The results of the XRF analysis of the raw CLN, parent CLN and the cation-exchanged forms (0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN, 0.5-Mg-CLN, 1.0-Na-CLN, 1.0-K-CLN, 1.0-Ca-CLN and 1.0-Mg-CLN) are listed in Table 1. Raw CLN is rich in potassium and calcium and has a SiO2/Al2O3 ratio of 5.7. Compared to the raw material, the parent sample prepared using Method 2 showed significant changes in its chemical composition and significant removal of exchangeable cations from the structure without damaging the crystal structure, as was confirmed by the XRD data. Prior to calcination, the cation exchange of extra-framework cations (Na+, K+, Ca2+ and Mg2+) with NH4+ resulted in significant reductions in the CaO and K2O components.
Chemical composition (wt.%) for raw CLN and parent CLN and its cation-exchanged forms.
The cation-exchanged forms of the parent CLN had higher contents of those oxides according to the salt solutions and the increasing molarity of this solution (Table 1). As Method 2, which was used to obtain the parent sample, did not damage the crystal structure and did not cause dealumination, there were no significant changes in the SiO2/Al2O3 ratios of the parent sample and those of cation-exchanged forms compared to the raw CLN.
XRD analysis
The powder XRD traces of the raw CLN, parent CLN, 0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN, 0.5-Mg Na-CLN, 1.0-CLN, 1.0-K-CLN, 1.0-Ca-CLN and 1.0-Mg-CLN samples are shown in Fig. 2. The characteristic peaks of CLN for the raw CLN sample were observed at 9.88°, 11.17°, 22.50° and 32.01°2θ, corresponding to d = 8.95, 7.91, 3.96 and 2.79 Å, with hkl indices of (020), (200), (131) and (530), respectively (Moore & Reynolds Jr, Reference Moore and Reynolds1997). In addition to CLN (80–85%), small amounts of feldspar (3%), opal-A (5–10%) and illite (2%) are also present in the raw CLN sample (quantitative analysis according to Esenli & Sirkecioğlu, Reference Esenli and Sirkecioğlu2005). The positions of the main CLN peaks did not changed significantly after cation exchange (Fig. 2). Compared to raw CLN, the relative peak intensity of (200) for the parent CLN and other cation-exchanged forms decreased relative to the peak intensity of (020), except for the 0.5-K-CLN and 1.0-K-CLN samples. The changes in the absolute and relative intensities of the characteristic peaks of the CLN samples can be attributed to the changes in the atomic positions and atomic densities in the structure and in the pore size and pore shape of the CLN (Galli et al., Reference Galli, Gottardi, Mayer, Preisinger and Passaglia1983; Castaldi et al., Reference Castaldi, Santona, Enzo and Melis2008; Kennedy & Tezel, Reference Kennedy and Tezel2018; Rodríguez-Iznaga et al., Reference Rodríguez-Iznaga, Shelyapina and Petranovskii2022). The relative changes in the peak intensities from our XRD results can be attributed to changes in the exchangeable cation ratios, as shown in the XRF data (Table 1), as the intensity of the (020) peak is highly dependent on the Na/K ratio of the CLN samples (Kitsopoulos, Reference Kitsopoulos2001). Furthermore, the absence of a broad hump between 19° and 30°2θ for the CLN and cation-exchanged forms indicates that the method used in this study did not damage the CLN structure, unlike other modification methods such as acid treatment (Arcoya et al., Reference Arcoya, González, Travieso and Seoane1994; Christidis et al., Reference Christidis, Moraetis, Keheyan, Akhalbedashvili, Kekelidze and Gevorkyan2003; Garcia-Basabe et al., Reference Garcia-Basabe, Rodriguez-Iznaga, de Menorval, Llewellyn, Maurin and Lewis2010; Kennedy & Tezel, Reference Kennedy and Tezel2018).
XRD traces of the zeolites. (a) Raw-CLN, (b) parent-CLN, (c) 0.5-Na-CLN, (d) 0.5-K-CLN, (e) 0.5-Ca-CLN, (f) 0.5-Mg-CLN, (g) 1.0-Na-CLN, (h) 1.0-K-CLN, (i) 1.0-Ca-CLN, (j) 1.0-Mg-CLN. * = CLN, • = illite, ⊕ = opal A, ♦ = feldspar.
The unit-cell parameters (a, b, c and β) and volumes of the CLN samples obtained from the h, k and l dimensions in the monoclinic crystal structure are listed in Table 2. A decrease in the unit-cell volume value of the parent sample compared to the raw sample was observed. This decrease due to calcination is similar that observed in other studies focused on CLN (Kudoh & Takéuchi, Reference Kudoh and Takéuchi1983; Bish, Reference Bish1984; Tomazović et al., Reference Tomazović, Ćeranić and Sijarić1996a, Reference Tomazović, Ćeranić and Sijarić1996b). In addition, the unit-cell volume values of all cation-doped forms were larger than the unit-cell volume of the parent sample.
Unit-cell parameters and volumes for raw CLN and parent CLN and its cation-exchanged forms.
SEM observations
SEM images taken at magnifications of at 5000× and 14 000× for the raw CLN, parent CLN, 0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN, 0.5-Mg-CLN, 1.0-Na-CLN, 1.0-K-CLN, 1.0-Ca-CLN and 1.0-Mg-CLN samples are shown in Figs 3 & 4. The CLN crystals form euhedral and subhedral plates, as well as coffin-shaped forms, in all samples (Figs 3 & 4a–e), similar to what has been observed in previous studies (Brundu & Cerri, Reference Brundu and Cerri2015; Favvas et al., Reference Favvas, Tsanaktsidis, Sapalidis, Tzilantonis, Papageorgiou and Mitropoulos2016; Fajdek-Bieda et al., Reference Fajdek-Bieda, Wróblewska, Miądlicki, Tołpa and Michalkiewicz2021). HEU-type crystals have platy, tabular or coffin habits (Elaiopoulos et al., Reference Elaiopoulos, Perraki and Grigoropoulou2010). In addition, some submicron and irregularly shaped CLN grains are also seen in all SEM images. The size of the crystals is consistent with that of the mineral obtained from the same region (Ünaldi et al., Reference Ünaldi, Mizrak and Kadir2013).
SEM images of (a) raw-CLN, (b) 0.5-Na-CLN, (c) 0.5-K-CLN, (d) 0.5-Ca-CLN and (e) 0.5-Mg-CLN at a magnification of 5000×. Letter labels with a * symbol corresponds to 14,000× magnification.
SEM images of (a) parent-CLN, (b) 1.0-Na-CLN, (c) 1.0-K-CLN, (d) 1.0-Ca-CLN and (e) 1.0-Mg-CLN at a magnification of 5000×. Letter labels with a * symbol corresponds to 14,000× magnification.
N2 adsorption
The N2 adsorption isotherms measured at 77 K are shown in Figs 5 & 6. The specific surface areas, micropore areas and micropore volumes of all of the samples were determined using the Brunauer–Emmett–Teller (BET) and the t-plot methods, respectively (Table 3). All isotherms are of type II according to the International Union of Pure and Applied Chemistry (IUPAC; Lowell et al., Reference Lowell, Shields, Thomas and Thommes2004). The knee portion of the isotherm indicates the stage when the coverage of the monolayer is complete and multilayer adsorption begins to take place (Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001; Lowell et al., Reference Lowell, Shields, Thomas and Thommes2004). The parent CLN has a higher specific surface area (106.93 m2 g–1), micropore area (81.27 m2 g–1) and micropore volume (0.0317 cm3 g–1) than the raw CLN sample (Table 3). This can be explained by the removal of exchangeable cations (K+, Na+, Mg2+ and Ca2+) from the structure due to the method used in the study, the preservation of the H+ form without damaging the crystal structure and the easier diffusion of nitrogen. Similar increases in the BET values have been observed for an NH4NO3-exchanged CLN from Germany (Hieu et al., Reference Hieu, Kosslick, Riaz, Schulz, Springer and Frank2022) and an NH4Cl-exchanged CLN from Bigadiç, Türkiye (Kurama et al., Reference Kurama, Zimmer and Reschetilowski2002). In addition, the BET surface areas of the cation-exchanged samples showed wide variation, in the range of 30.72–198.98 m2 g–1. The maximum specific surface area of the 1.0-Mg-CLN sample can be explained by the replacement of exchangeable cations such as Ca2+ and K+ by Mg2+ (which is smaller in size), as is confirmed by the XRF data (Table 1).
N2 adsorption isotherms of raw CLN, 0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN and 0.5-Mg-CLN samples at 77 K. STP = standard temperature and pressure.
N2 adsorption isotherms of parent CLN, 1.0-Na-CLN, 1.0-K-CLN, 1.0-Ca-CLN and 1.0-Mg-CLN samples at 77 K. STP = standard temperature and pressure.
N2 adsorption data of the raw CLN and parent CLN and its cation-exchanged forms.
Ammonia adsorption
Direct or indirect ion exchange with NH4NO3 followed by employing the calcination method can be applied to zeolite-type materials because this process removes the exchangeable cations whilst leaving the structure unaffected, unlike the methods involving solutions of HCl or H2SO4, which cause structure decomposition. In this study, the direct and indirect ammonium nitrate treatment methods before calcination were carried out using multiple molarities (0.5 and 1.0 M). The 0.5-4h-CLN (3.95 mmol g–1), 0.5-8h-CLN (4.02 mmol g–1), 1.0-4h-CLN (3.99 mmol g–1) and 1.0-8h-CLN (3.92 mmol g–1) samples in which the H+ forms were obtained directly without conversion to the Na+ form, defined as Method 1, adsorbed less ammonia than the Na-0.5-4h-CLN (4.19 mmol g–1), Na-0.5-8h-CLN (4.07 mmol g–1), Na-1.0-4h-CLN (4.10 mmol g–1) and Na-1.0-8h-CLN (4.46 mmol g–1) samples obtained using Method 2. In these H+ forms obtained by the two methods, the removal of exchangeable Mg2+, Ca2+ and Na+ cations by direct or indirect ion exchange with NH4NO3, followed by calcination, caused a general decrease in the ammonia adsorption capacities compared to the raw sample. For this reason, Na-1.0-8h, which has the highest ammonia adsorption capacity amongst the H+ forms, was selected as the parent sample. The next step was to determine the ammonia adsorption capacities by doping the parent sample with different cations. The ammonia adsorption isotherms of the raw CLN, the parent sample and that of cation-exchanged forms at 298 K up to a pressure of 100 kPa are shown in Figs 7 & 8.
NH3 adsorption isotherms of raw CLN, 0.5-Na-CLN, 0.5-K-CLN, 0.5-Ca-CLN and 0.5-Mg-CLN samples at 298 K.
NH3 adsorption isotherms of parent CLN, 1.0-Na-CLN, 1.0-K-CLN, 1.0-Ca-CLN and 1.0-Mg-CLN samples at 298 K.
The ammonia uptake of the CLN samples ranged from 3.61 to 4.93 mmol g–1 and increased in the following order: 1.0-K-CLN < 0.5-K-CLN < 1.0-Mg-CLN < 1.0-Na-CLN < 0.5-Mg-CLN < raw CLN < parent CLN < 1.0-Ca-CLN < 0.5-Ca-CLN < 0.5-Na-CLN (Table 4). When the molarity of the salt solutions used was two times higher (from 0.5 to 1.0 M), the ammonia adsorption capacity decreased. The raw CLN sample showed lower ammonia uptake (4.41 mmol g–1) compared to CLN from Mud Hills, USA (5.90 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001) but a higher uptake than the Slovakian CLN (0.71 mmol g–1; Ciahotný et al., Reference Ciahotný, Melenová, Jirglová, Pachtová, Kočiřík and Eić2006) due to its different mineralogical and chemical composition. Although the specific surface area (106.93 m2 g–1) and micropore surface area (81.27 m2 g–1) values of the parent CLN were higher than those of the cation-exchanged forms (30.72–93.15 m2 g–1 and 8.46–67.75 m2 g–1) with 0.5 and 1.0 M NaNO3, KNO3 and Ca(NO3)2 solutions, respectively (Table 4), it showed an average ammonia adsorption capacity. This can be attributed to the significant removal of exchangeable cations in the parent sample, in one of the H-forms obtained using Method 2. This also clearly demonstrates the influence of the extra-framework cations on ammonia adsorption and the interactions of the permanent dipole moment (1.47 Debye) of the ammonia molecule with the electric field generated by these cations. Amongst the cation-modified forms of the parent sample, 0.5-Na-CLN (4.93 mmol g–1) showed the highest uptake. This result for the 0.5-Na-CLN sample showed that it is beneficial to dope the parent sample with Na+ cation in a second step. As shown in Fig. 9, in CLN, Na+, Ca2+ and K+ cations prefer to occupy sites M(1) (in channel A), M(2) (in channel B) and M(3) (in channel C), respectively, whereas the Mg2+ cation is located at site M(4) (in channel A; Koyama & Takeuchi, Reference Koyama and Takeuchi1977). The exchange of Na+ cations with Mg2+, Ca2+ and K+ cations and the presence of these smaller Na+ cations in the M(1) site of channel A resulted in there being a larger area within the channels. On the other hand, the 1.0-K-CLN sample had both the lowest specific surface area (30.72 m2 g–1) and the lowest ammonia adsorption capacity (3.61 mmol g–1) due to the size and position of the K+ (largest) cation and partial pore blocking that occurred. The percentage change in the ammonia adsorption capacity of the parent and the cation-exchanged forms compared to the natural sample ranged between 0.45 and 18.14.
Comparison of the removal of ammonia at 298 K in this study with materials from previous studies.
Views of the CLN framework and cation sites along the c-axis and a-axis.
The ammonia adsorption capacity of the 0.5-Na-CLN sample (4.93 mmol g–1) is lower than that of MOF-177 (12.2 mmol g–1; Saha & Deng, Reference Saha and Deng2010b), 13X (9.33 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001), 4A (8.71 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001) and mesoporous carbon (6.39 mmol g–1; Saha & Deng, Reference Saha and Deng2010a) but higher than those of activated alumina (2.53 mmol g–1; Saha & Deng, Reference Saha and Deng2010c), dealuminated pentasil (2.34 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001), activated carbon (4.19 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001), dealuminated faujasite (1.77 mmol g–1; Helminen et al., Reference Helminen, Helenius, Paatero and Turunen2001) and Cu-MOF-74 (3.4 mmol g–1; Katz et al., Reference Katz, Howarth, Moghadam, DeCoste, Snurr, Hupp and Farha2016; Table 4). Comparing the ammonia adsorption data obtained at the same temperature (Table 4), it is clear that the structural and textural properties of these synthetic materials are completely different from natural zeolite of the CLN type. Although in general the synthetic zeolites, due to their uniform structure, have higher gas adsorption capacities than natural zeolites, they are more expensive. The abundance of CLN-type natural zeolite, its low cost and its high capacity to adsorb harmful gases such as ammonia lead to its widespread use in industrial applications. As a result, the 0.5-Na-CLN sample with the highest ammonia adsorption capacity is recommended as an effective adsorbent in environments where ammonia gas needs to be removed, such as poultry houses.
Conclusions
In this study, the structural properties and ammonia adsorption capacities of the parent CLN and its forms doped with Na+, K+, Ca2+ and Mg2+ cations were compared. The XRD data of the CLN samples showed that the NH4NO3 modification and the calcination process to obtain the H+ forms prior to the cation exchange did not cause any significant damage to the crystal structure. It was also found that the morphology of the modified samples was not affected by calcining. A more than threefold increase in BET surface area (106.93 m2 g–1) was observed for the parent sample compared to the raw CLN (30.08 m2 g–1). CLN samples in which the H+ forms were obtained by first converting to the Na+ form (as a result of ammonium nitrate and calcination), as defined by Method 2, adsorbed more ammonia than directly obtained H+ forms, as defined by Method 1. A wide variation in ammonia adsorption was observed in the cation-exchanged CLNs, being more dependent on the size, amount and location of the exchanged cation than on the BET surface areas. Consequently, 0.5-Na-CLN, which has the highest ammonia adsorption capacity amongst the samples used in this study, can be suggested as a potential material for ammonia removal applications.
Acknowledgements
The authors thank Ceramic Research Center (SAM, Eskisehir/Turkey) for the XRF analysis of all samples.
Financial support
Financial support from Eskisehir Technical University Scientific Research Committee under grant number 22ADP355 is gratefully acknowledged.
Conflicts of interest
The authors declare none.
