1. Introduction
Zeolites have unique structural and chemical properties that are vital for a wide range of industries. Given their economic effect, there is a powerful incentive for the intelligent design of new materials with enhanced functionality in ion exchange and catalysis [
1]. Cation-exchanged zeolites are a prospective class of catalysts and their popularity has steadily increased over the last years. Following the study of copper-containing zeolites as promising materials for the removal of nitrogen oxides from exhaust gases (de-No
x catalysts) [
2,
3,
4,
5], new processes based on ion-exchanged zeolites have recently been investigated. Zeolites, as the most important solid catalysts used in traditional petrochemical industries, also find prospective applications in many sustainable processes. In addition, they can be used to capture and convert CO
2 in fuel cells, for biomass conversion, air, and water purification, etc. [
6,
7,
8].
Among the many known zeolites, mordenite is one of the most interesting, from our point of view. It is one of the most used zeolites in the chemical industry and plays a key role in novel applications. The reason for this is that this zeolite belongs to the so-called “big five” zeolites, which are extensively applied in the industry as catalysts [
9]. Based on this, the constant interest in a detailed study of this material is quite understandable.
To focus on one of the possible systems for research, choosing the metal to modify mordenite was necessary. Among the proposed metals, for several reasons, we considered copper to be one of the most interesting. First, from an economic point of view, it has an acceptable price. Secondly, copper, an element with a variable valence and with a very interesting, although not fully investigated, chemistry on the surface of zeolitic pores [
10], has repeatedly shown very fascinating results in various processes. Numerous studies have shown that the catalytic properties of copper-exchanged zeolites are determined by the valence state of copper ions and their location and coordination in the zeolite lattice, as well as by the copper content [
10,
11]. According to Iwamoto [
12], the so-called over-exchanged zeolites (i.e., zeolites containing more polyvalent cations than is necessary for formal charge equilibrium) are more efficient in catalytic reactions. From this point of view, it is very important to find an effective method for controlling the copper content in the zeolite matrix.
It has been recently discovered [
13] that extra-framework aluminum species can also affect the catalytic activity of Cu-mordenite in the conversion of methane to methanol. Moreover, the materials obtained from a protonated form have been found to be more active. As it has been shown, there are two key properties that determine the activity of prepared materials in methane conversion: the parent counterion and the copper ion-exchange procedure [
13].
It is well known that the performance of a Cu-zeolite catalyst strongly depends on the chemical state of copper. In turn, it is accepted that the ion exchange method affects the appearance of Cu species; the matrix topology and counterions presented in the starting zeolite also influence this appearance [
14]. The recent development of microwave-assisted treatment methods has introduced a new direction in the development of chemistry of materials [
15,
16,
17,
18,
19,
20]. Being a simple, inexpensive, and efficient nonconventional heating method, the microwave (MW) technique is well established in the field of alternative methods for the synthesis of emerging materials, including zeolites [
17,
19,
20,
21,
22]. A comparison of several Ni-mordenites obtained by ion exchange of Na-mordenite in the liquid phase under irradiation by microwaves and by the conventional method showed that after the first exchange an analogous amount of exchangeable Ni
2+ ions, with and without MW assistance, was achieved [
21], but was faster with MW irradiation. Interestingly, microwaves strongly affected the acidity of the samples obtained, the acidity of which increased with longer treatment with microwaves due to the appearance of new weak acid sites. These new centers were not observed for the sample exchanged in the conventional way [
21]. Differences in the acidity of samples prepared by conventional and microwave-assisted treatments were also observed in [
22].
Finally, the effect of counter-cations in the initial zeolites can change the final properties of the obtained samples. In [
23], copper-ammonium exchange was studied for the ammonium form of SAPO zeolites obtained by various methods. A number of NH
4-SAPO-34 samples prepared by NH
3 adsorption were prepared and used for comparison with NH
4-SAPO-34 obtained by conventional exchange from a liquid. Even if an analogous amount of the NH
4+ ion was contained in the sample obtained by gas adsorption and in the sample obtained by the liquid ion exchange method, there was still a difference between the coordination of NH
4+ species with the bridging hydroxyl groups of Si(OH)Al. This difference resulted in a subsequent exchange of Cu, leading to various forms of cationic copper complexes [
23].
Earlier, we analyzed the influence of various factors in the preparation of copper exchange mordenites on the properties of the materials obtained [
24,
25,
26,
27,
28]. In [
27], we reported the results of our comprehensive study of two sets of copper-exchanged samples of mordenites prepared from the sodium form of this zeolite and obtained by two different methods: conventional and MW assisted. It was found that the MW-assisted procedure resulted in a deeper and even over-exchange. The main goal of present study was to prove the ability and efficacy of MW-assisted treatment, as compared with the conventional procedure, in synthesis of copper-exchanged mordenite from ammonia and protonated forms.
3. Discussion
One of the processes studied in the present work was the copper accumulation in mordenite due to exchange procedures done in two different ways. As seen from
Table 1, after the first ion exchange treatment by both methods, copper ions did not substitute the initial compensating cations completely: the copper-exchange level did not exceed 36% for the two methods and both systems. Nevertheless, after multiple ion exchange procedures, the MW treatment proved itself to be more effective, especially for CuH-MOR-10-M
Y series. However, in contrast to the previously studied copper-exchanged mordenites obtained from Na-MOR-6.5 [
27], even six subsequent MW-assisted ion exchange procedures did not lead to an over-exchange. This discrepancy between the ammonia and proton forms of mordenite, unlike sodium, is an interesting fact, but its cause is not yet clear; one of the reasons, as we can assume, is due to the different Si/Al ratio and its effect on the exchange properties of mordenite. As expected, zeolites with the same framework topology, but with different framework charges, demonstrate quantitative differences in the exchange behavior [
29]. Less obvious, the charge density of the zeolite framework also affects ion exchange selectivity [
30]. These effects are especially pronounced for uni–divalent cations exchange. Since the charge of divalent cations is balanced by two negative framework sites, the selectivity and maximum level of exchange in zeolites with low aluminum content depend mainly on the separation distances of the framework charge sites [
30]. However, a detailed study on similar dependencies of the role of various counter-cations on the exchange of the target cation, and the role of the composition of zeolite in such processes, is beyond the scope of this present work and will be done separately. Another interesting observation was that for all the systems studied in this work, obtained both by conventional and MW-assisted exchange methods, the copper exchange stopped at approximately 70%, as was previously detected for CuNa-MOR-6.5-C
Y, prepared by the conventional method in [
27].
As it follows from XRD analysis, the copper loading led to increasing the
a to
b lattice parameters ratio that reflects the compression of the main mordenite channel. The
a/
b values versus the copper-exchange level for all the studied samples are plotted in
Figure 4 together with the data for CuNa-MOR-6.5-
XY series from [
27]. As one can see from
Figure 4, for both CuNH
4-MOR-10-
XY and CuH-MOR-10-
XY, the copper exchange led to increasing the
a/
b ratio that can be almost perfectly fitted by linear dependence. Moreover, neither the nature of the initial cation nor the preparation method has any significant effect on the slope of the line, which is essentially lower in comparison with that one for CuNa-MOR-6.5-C
Y. The weakening of the effect of Cu imbedding on the deformation of the mordenite lattice was related to the lower absolute copper content at the same copper-exchange level (due to higher Si/Al ratio). We note that the change in the slope after 70% of the copper-exchange level observed in CuNa-MOR-6.5-M
Y series can be attributed to the creation of [Cu-O-Cu]
2+ complexes formed due to MW irradiation [
27].
XRD patterns presented in
Figure 1 and
Figures S1 and S2 show that the ammonium- or proton-to-copper exchange led to some changes in relative intensities of diffraction peaks.
Figure 4b–d show the changes in intensities of selected peaks, namely (310), (150), and (202), that took place with copper loading. To analyze their intensities, they were normalized to the intensity of the peak (330), which varied much less than others from sample to sample (see
Figure 4). Besides these, for both forms, ammonium and proton, it can be seen that the exchange for copper changed the intensity of the peaks (110), (020), and (200). Schematic representation of these planes is given in
Supplementary Materials, Figure S8. The changes are fairly constant with increasing copper concentration and does not do beyond the general trend. This is quite expected for ion-exchange zeolites, and the effect is mainly due to differences in nature, number, and position of extra-framework cations. This indicates that Cu
2+ ions occupy similar cationic positions in the channels of mordenite, regardless of the exchange method used. A more detailed study of the structure is undoubtedly of interest, as it would provide valuable information on the local structure rearrangement caused by ion-exchange procedure, as it was observed in Ag- and Fe-exchanged mordenite [
31], and its results will be published elsewhere.
In mordenite, counter-cations can occupy different positions. From general considerations, cations tend to be located at positions that are more stable or have minimal energy, which includes optimal values for cation/framework distances, the distance between cation and water molecules, between water molecules and the framework, minimal repulsion between guest species, etc. Referring to the studied systems, and taking into account the results of previous studies, the following possible cationic sites, shown in
Figure 5, were proposed in [
32]: A—inside the small mordenite channel; B— in the side pocket; and C—inside the main channel near the side pocket. Here we did not proceed to a deeper description since the methods used in the present study did not allow us to distinguish counter cations localized, for example, in different places within the main channel; only a rough analysis can be done. Thus, despite the fact that copper ions do not have exact crystallographic positions, the behavior of the intensities of the XRD peaks that reflects the influence of the copper atoms can be summarized as follows. The decrease in the intensity of the peak (202) (as well as that one of (020) and (200)) occurred due to the displacement of [NH
4]
+ (or H
+) cations, originally located in small mordenite channels (site A in
Figure 5) that occur under the action of the ion-exchange procedure. The above-mentioned cations leave their sites, but when entering Cu
2+ cations do not replace them, preferring to settle in the main channel near the entrance to the side pocket, the C-site. Note, that the planes (202) and (310) intersect in the side pocket entrance area. That is, the localization of the cation in this site should have a strong influence.
The difference in behavior of (150) peak intensity between CuNH
4-MOR-10-
XY and CuH-MOR-10-
XY series and the data from [
27] for the CuNa-MOR-6.5-
XY series means that the exact position of the copper ions in the main channel of mordenite with Si/Al = 10 and Si/Al = 6.5 were different. As Al can occupy T3 and T4 sites marked by orange in
Figure 5, with Si/Al increasing, the occupancy of these two sites by Al decreases and the total number of cationic sites decreases.
A slight discrepancy in the peak intensities observed for CuNH
4-MOR-10-
XY and CuH-MOR-10-
XY samples prepared by different methods may be due to the fact that copper ions can occupy different sites in the mordenite lattice, which corresponds to the further discussion of the results of the thermogravimetric analysis (TGA). The latter was also consistent with the
27Al MAS NMR study. The introduction of copper into the studied zeolite did not shift the main L1 line, but led to its broadening, which is a typical effect of paramagnetic ions and implicitly confirms that copper ions are in the Cu
2+ state. The right insert in
Figure 2 shows the half width at half maximum (HWHM) of the L1 line plotted versus the copper-exchange level. This parameter does not carry direct information about the preparation method. It can be fairly well fitted with a straight line; a noticeable deviation occurred only when the copper-exchange level exceeded 60%, which may indicate a redistribution of copper ions as compared with other ion exchange positions, and confirms, once again, the presence of a threshold exchange level before and after which the behavior of cations in the zeolite changes.
For CuNa-MOR-6.5-M
Y, where the MW treatment possibly led to the creation of [Cu-O-Cu]
2+ species [
28], changes in the intensities of diffraction peaks at high copper loading (>60%) reflect that these complex cations occupy new sites and/or, perhaps, their appearance led to a redistribution of Cu
2+.
Valuable information on differences in the accessibility of ion-exchanged sites in the CuNH
4-MOR-10-
XY and CuH-MOR-10-
XY series can be provided by TGA analysis. Compared with the data of CuNa-MOR-6.5-
XY [
27], in the ammonia and protonated series, firstly, the high temperature peak of mass loss always presents and is rather important; secondly, the temperature of the first step is essentially shifted towards a lower temperature.
According to the XRD analysis, in NH
4-MOR-10, ammonium ions were in a small mordenite channel and left it during substitution with copper. This means that before the copper is loaded, the main channel is almost empty: water molecules are weakly held by the zeolite framework due to a high Si/Al ratio (and, therefore, low polarization). Copper, entering into the main channel, is coordinated by water that results in an increase in mass loss due to water release at the first step; see
Figure 3 and
Table 3. In the CuH-MOR-10-
XY series, an additional decrease in the water content could be related with partial destruction of the zeolite framework.
Turning to a comparison of ammoniated and protonated forms of mordenite used for copper exchange, it should first be emphasized that the initial step of water desorption was also accompanied by ammonia release due to thermal decomposition of [NH
4]
+, which ends at about 250 °C. Unfortunately, the close mass values of NH
3 and H
2O make them difficult to distinguish. Nevertheless, the quantity of ammonia emission can be roughly estimated from the difference between the TGA curves for NH
4-MOR-10 and H-MOR-10 at the first step of mass loss, see
Figure 3a. The second step of mass loss at temperatures above
T0, which was between 258 and 364 °C, depending on the sample, see
Table 2, was caused by dehydroxylation processes. From this perspective, the second step of the TGA curves for NH
4-MOR-10 and H-MOR-10 should be the same. Indeed, the profiles looks quite similar; the difference in mass loss, that is the difference in the number of hydroxyl groups, could be caused by the fact that the preparation of H-MOR-10 is a static calcination process (two hours at 300 °C), whereas during the TGA study, the NH
4-MOR-10 sample experienced a dynamic thermal exposure. In other words, the main difference between the ammonia and protonated series was the amount (and possibly the number of types) of hydroxyl groups that should influence ion exchange since hydroxyls in the protonated form may occupy several ion-exchangeable sites making others accessible, which were initially less energetically favorable. The presence of the so-called “five”-coordinated Al may also lead to a redistribution between more and less favorable ion-exchanged sites.
In the ammonia samples, after the completion of the ion-exchange procedure, part of [NH4]+ was replaced by Cu2+, but the remaining ammonium cations thermally decomposed at temperatures below 300 °C, forming new proton centers. These former CuNH4-MOR-10-XY samples at temperatures above 300 °C are converted into “CuH-MOR-10-XY”, but with different numbers (and types) of hydroxyl groups, which are now determined by the previous location of the residual ammonium cations in the mordenite structure after ion exchange.
Another important point, as can be seen from
Figure 3, is that the presence of copper affects the formation of a certain form of hydroxyls; as the concentration of copper increased, hydroxyl groups that decompose at high temperatures disappeared first, and the relative number of low temperature ones increased. This implicitly points out that the high temperature hydroxyl groups in H-MOR-10 after the ion-exchange procedure may have been replaced by copper cations.
4. Materials and Methods
Copper-exchanged zeolites were prepared from NH
4+- and H
+-mordenites with nominal Si/Al atomic ratios (AR) equal to 10. The NH
4+-mordenite was supplied by Zeolist Int. (Product CBV21A). The proton form was obtained by the calcination of NH
4+ mordenite at 300 °C for 2 h. The calcination temperature was chosen on the basis of thermogravimetric analysis of the NH
4+ mordenite: the temperature at which the deammoniation process had finished but the dehydroxylation had not started yet, see
Section 2.3. The ammonia and proton forms were labeled as NH
4-MOR-10 and H-MOR-10, respectively.
In order to prepare copper-exchange samples, the starting material (NH
4-MOR-10 or H-MOR-10) was treated in 0.05 M CuSO
4 aqueous solution, taking the volume with the twofold excess of copper amount in the solution to the estimated ion-exchange capacity of mordenite. The conventional copper ion exchange was carried out under stirring at room temperature for 1 day. For the MW-assisted ion-exchange procedure, the mixture of zeolite and CuSO
4 aqueous solution (with the same concentration of solution and the same zeolite-solution ratio) was heated at 100 °C in a Synthos 3000 Anton Paar microwave oven at 1400 W for 2 h. The exchange process was repeated several times (up to six) to increase the copper content. The samples obtained by both methods were filtered, thoroughly washed, and dried at room temperature overnight. The samples obtained from the ammonia and proton forms were labeled as CuNH
4-MOR-10-
XY and CuH-MOR-10-
XY, respectively, with
X = C or M for the conventional (C) or MW (M) ion-exchange preparation method, correspondingly;
Y = number of copper-exchange procedures; digit 10 corresponds to the Si/Al ratio. The main information concerning the exposure, which the studied samples underwent during preparation treatment, is summarized in
Table 4.
XRD analysis was done using a Bruker D8 DISCOVER (Bruker Corporation, Billerica, MA, USA) diffractometer with CuKα long-focused X-ray sealed tube (scan range: from 5.0017 to 60.0041 2-theta degree, step width: 0.0202 degree). Quantification was completed using the DIFFRAC.SUITE software package.
The chemical composition of the samples before and after the ion-exchange procedures was explored by ICP-OES using Shimadzu ICPE-9000 (Shimadzu Corporation, Kyoto, Japan). The valence state of copper ions was probed by XPS analysis carried out using a Combined Auger, X-ray and Ultraviolet Photoelectron spectrometer Thermo Fisher Scientific ESCAlab 250Xi with monochromatic AlKα radiation (photon energy 1486.6 eV (Thermo Fisher Scientific, Waltham, MA, USA) at room temperature in ultra-high vacuum with pressure of the order of 1 × 10-9 mbar.
The water content and products of the samples decomposition were determined through the use of thermal analysis using a Netzsch STA 449 F1 Jupiter coupled with a quadrupole mass spectrometer QMS 403 Aëolos (Netzsch-Gruppe, Selb, Germany). The analysis was done in the temperature range from 40 to 700 °C at a heating rate of 10 °C/min in argon stream at the rate of 90 mL/min.
To probe the regularity of the zeolite framework the magic angle spinning nuclear magnetic resonance (MAS NMR) method was used. 27Al MAS NMR spectra were recorded by applying the zg30 pulse sequence at 104.29 MHz with rotation frequency 8 kHz. Chemical shifts were determined relative to the AlCl3 solution as an external standard. NMR measurements were carried out at room temperature using Bruker WB Avance III (400 MHz) spectrometer (Bruker Corporation, Billerica, MA, USA).
5. Conclusions
The main conclusions of the work can be formulated as follows:
For both of the studied initial forms of mordenite, neither MW-assisted nor conventional methods destroy the zeolite framework. As compared with the sodium form of mordenite, the exchange of copper in ammonia and protonated forms of mordenite behaves differently but all copper is in Cu2+ form. The applied ion-exchange processes do not lead to an over-exchange, as the level of copper exchange is limited to 70%.
The exchange of the initial counterions for Cu2+ ions leads to a linear increase in the a/b ratio with an increase in the copper exchange level. This means that all copper cations are ion-exchangeable and enter the main mordenite channel. For CuH-MOR-10-M6, with the greatest amount of copper reaching 70% of the degree of exchange, changes in the width of the 27Al MAS NMR line are observed. This effect can be interpreted as the possible appearance of a new form of copper, or the redistribution of these copper ions along ion-exchangeable sites when the threshold value is reached.
When heated, there is a loss of water, which is accompanied by the simultaneous release of NH3 for the CuNH4-MOR-10-XY samples. This process ends in the range of 260–360 °C. With further increases in temperature, the mass loss is due to dehydroxylation that occurs in a complex way. In copper-exchanged samples, protons (hydroxyl groups) with a higher temperature of thermal decomposition are selectively replaced. There is a correlation between the energy of the exchange for copper and the energy of decomposition of various ammonium groups and, accordingly, the dehydroxylation of proton groups formed from non-exchanged ammonium cations.
Summing up the above, one can conclude that MW-assisted treatment has proven itself as an efficacious method for the synthesis of copper-exchanged mordenites, which not only significantly reduces preparation time but leads to a systematically higher Cu-exchange level.