Mobility of NH 3-Solvated Cu II Ions in Cu-SSZ-13 and Cu-ZSM-5 NH 3-SCR Catalysts : A Comparative Impedance Spectroscopy Study

The mobility of NH3-solvated Cu ions within the zeolite framework has been recently identified as a key factor for the kinetics of the selective catalytic reduction of NOx with NH3 (NH3-SCR) over Cu-zeolite catalysts at low temperatures. Here, we utilize in situ impedance spectroscopy to explore the mobility of NH3-solvated CuII ions, i.e., Cu(NH3)n, in Cu-SSZ-13 and Cu-ZSM-5 zeolites with varied Cu ion exchange levels, and observed that both the zeolite framework (CHA or MFI) and the Cu exchange level influence the high-frequency dielectric relaxation processes that are associated with the short-range (local) motion of Cu(NH3)n. Our results suggest that the local motion of Cu(NH3)n species is favored within the CHA framework due to the unique cage structure, and thereby contribute to the overall ion conductivity at high frequencies, which, on the contrary, is not observed for ZSM-5, where NH3-solvated Cu2+ ions do not experience a comparable constrained space for local motion. This study sheds new light on the mobility of Cu active sites under NH3-SCR related reaction conditions and may contribute to an advanced understanding of the underlying mechanism.


Introduction
Nitric oxide and nitrogen dioxide (NO and NO 2 , commonly indicated as "NO x ") are harmful atmospheric pollutants and are byproducts of anthropogenic high-temperature combustions [1,2]."Three-way" catalysts, although being successfully employed for NO x abatement in gasoline-powered vehicles, are not sufficiently effective in the O 2 -rich exhaust environment typically in "lean-burn" engines (e.g., diesel engines).In this case, one of the leading strategies for NO x emission control (DeNO x ) is the selective catalytic reduction with ammonia as reducing agent (NH 3 -SCR) [3].In NH 3 -SCR systems, an aqueous urea solution (AdBlue TM , VDA, Berlin, Germany) is injected into the hot exhaust gas mixture and it subsequently decomposes to form gaseous NH 3 [3][4][5].Depending on the NO to NO 2 ratio, different SCR reactions may occur and the most important one is the so-called "standard SCR" [2,4,6]: Cu-and Fe-exchanged zeolites are among the most common catalysts for the SCR reaction in diesel-powered automobiles because of their high activity and stability under the respective reaction conditions [7].Particularly, Cu-zeolites show outstanding low-temperature performance and low sensitivity to NO 2 concentration in the exhaust flow [4,5].Due to their better performance (in terms of both activity and hydrothermal stability) when compared to large-and medium-pore zeolite materials, Cu-exchanged small-pore chabazite (CHA) zeolites (e.g., Cu-SSZ-13 and Cu-SAPO-34) are now the most common choice for NH 3 -SCR [2][3][4][5][6]8,9].Nevertheless, a deeper mechanistic understanding of Cu-zeolite catalyzed NH 3 -SCR is still needed in order to develop even more efficient catalysts for meeting the increasingly stringent NO x emission legislation [1,3,4].
Zeolites are microporous aluminosilicates, in which the ordered porous structure is generated from the crystalline structure of the material itself: the tetrahedral TO 4 primary building units (T being usually Si or Al), sharing corners, form channels and cages of molecular sizes.The inequivalent substitution of a Si T-atom with a similar-size Al T-atom results in a net negative framework charge, which requires being balanced by extra-framework cations, e.g., H + in proton-form zeolites or metallic cations in metal-exchanged zeolites.For this reason, even if zeolites are electrically non-conductive, they are ionically conductive due to the motion of extra-framework cations [10].
In Cu-zeolites, Cu 2+ cations are introduced into the zeolite framework acting as active sites for the catalytic conversion of NO x .Under SCR conditions, the Cu 2+ centers are solvated by NH 3 , leading to the formation of Cu-NH 3 complexes (hereon indicated as "Cu II (NH 3 ) n species").Due to the solvation effect of NH 3 , the electrostatic interaction between Cu 2+ and the negatively charged zeolite framework is weakened, increasing the mobility of the Cu ions within the framework [11].Further interaction of Cu II (NH 3 ) n with NO leads to the reduction of Cu II to Cu I , together with the conversion of NO and one of NH 3 molecules to N 2 and H 2 O [5].Recently, several research groups [12,13] identified that the transient formation of [Cu I (NH 3 ) 2 ] + -O 2 -[Cu I (NH 3 ) 2 ] + intermediates, which are necessary to activate O 2 , is a rate-limiting step for the Cu I → Cu II re-oxidation half-cycle.In particular, for Cu-SSZ-13 with low Cu density, the mobility of [Cu I (NH 3 ) 2 ] + complexes was found to determine the pairing of two [Cu I (NH 3 ) 2 ] + in neighboring CHA cages, and thus plays a vital role in the Cu redox cycle, and consequently, the overall NH 3 -SCR reaction.Due to a weaker electrostatic interaction, Cu I species are considered to be more mobile than Cu II species [11].
Despite its importance in SCR catalysis, direct experimental investigation of the Cu ion mobility remains a challenge.In this context, in situ impedance spectroscopy (IS) becomes a reliable technique, since it allows for studying the frequency-dependent conduction and the polarization processes of zeolites, in which the solvated Cu species are involved.By performing in situ IS under NH 3 -SCR related conditions, we have already investigated the electrical properties of a series of commercially relevant zeolite catalysts (e.g., H-ZSM-5, Cu-ZSM-5, Fe-ZSM-5, Cu-SAPO-34) [10,[14][15][16].Moreover, a novel analytical approach combining IS with diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was developed and allows for understanding the NH 3 -induced electrical properties (e.g., proton transport) of zeolite catalysts at a molecular level [17][18][19][20][21][22].A very recent study of our group has shown how IS combined with density-functional theory calculations enables to analyze the local movement of [Cu I (NH 3 ) 2 ] + under SCR conditions [23].
In view of these findings, here, we designed a series of in situ IS experiments to probe the mobility of Cu II (NH 3 ) n species, which are formed after NH 3 -loading on the Cu-exchanged zeolite catalysts.Cu-SSZ-13 and Cu-ZSM-5 catalysts with varied Cu loadings were compared, in order to probe framework-related effects.Even though further experiments and theoretical support are needed, this proof-of-concept study confirms that in situ IS is capable of probing the mobility of Cu II intermediates within the zeolite framework under NH 3 -SCR related reaction conditions.

Catalyst Characterization
For each zeolite framework, we synthesized three samples with different Cu/Al ratios via aqueous ion-exchange, following previously established protocols [19].The Cu-exchanged zeolites will be hereon named as "Cu(x)-SSZ-13" and "Cu(x)-ZSM-5", respectively, where x refers to the Cu/Al ratio determined by inductively coupled plasma optical-emission spectroscopy (ICP-OES) (Table 1).The obtained SSZ-13 and ZSM-5 samples show typical X-ray diffraction (XRD) patterns, which are associated to CHA and MFI crystal structures, respectively (Figure 1a,b) [24].No visible copper oxide (CuO x ) reflection was noticed in the XRD patterns for the Cu-exchange zeolite samples, implying the absence of large CuO x agglomerates.Nevertheless, the presence of highly dispersed CuO x clusters or small particles, which may be below the detection limit of XRD, cannot be fully excluded, especially in samples with high Cu/Al ratios, i.e., Cu(0.24)-SSZ-13 and Cu(0.16)-ZSM-5.Scanning electron microscopy (SEM) was applied to study the surface morphology of the samples and the homogeneity of the zeolite film on the inter-digital electrodes (IDE) that were used in IS measurements.No significant morphologic change was observed after ion exchange and the zeolite particles in the thick film on IDE are densely packed and homogeneously dispersed (Figure 2a-d).Brønsted sites.The Cu-exchanged zeolite samples present an additional peak (ca.240 and 230 • C for SSZ-13 and ZSM-5, respectively) that is overlapping partially with the high-and low-temperature ones, which is attributed to NH 3 desorption from Cu sites [5,18,25].As compared to the respective H-form zeolite, the Cu-form zeolite shows a reduced desorption of NH 3 from Brønsted sites, as a result of the replacement of H + on Brønsted sites with Cu 2+ after ion exchange [18,26].

In Situ IS: Modulus Spectra
Impedance spectroscopy (IS) is an electric perturbation technique, in which an alternating voltage U, with angular frequency ω (=2πν) and amplitude U 0 , is applied to a system in thermodynamic equilibrium.As for zeolites, the as-such generated electric field induces motion of the extra-framework cations, which is macroscopically measured as a current I(ω), inversely proportional to the complex impedance Z(ω): Complex impedance can be expressed by its real (Z ) and imaginary (Z") parts as: The so-called Argand plot (also known as Nyquist plot), which displays Z" against Z , is dominated by the low-frequency tail that is related to phenomena such as sample/electrode interface polarization.For the study of high-frequency processes, modulus plot (M" vs. log ν) has been proved to be more suitable.In M" plots, the experimentally collected impedance values (Z) are transformed according to the equation: where ν is the perturbing frequency, C 0 is the capacitance of the empty capacitor, and Z the real part of the complex impedance [10,17,27].In the modulus plot, the position of a local maximum corresponds to a resonance frequency (ν res ) of the system [17,18].The relaxation time (τ) of the involved ion movement phenomenon can be thus derived according to The modulus spectra obtained at 200 • C in the presence of NH 3 are reported in Figure 4a,b for the whole sets of SSZ-13 and ZSM-5 zeolites, respectively.At this temperature, a minimum amount of weakly bound NH 3 is adsorbed on the zeolites, while the adsorption of NH 3 on Cu sites is widely unaffected (see NH 3 -TPD profiles in Figure 3a,b).Therefore, the effect of Cu II (NH 3 ) n species on the ion movement within zeolites is expected to be most pronounced at ca. 200 • C. Each modulus spectrum presents two resonance peaks, corresponding to two dielectric relaxation modes.Previous studies proved that the low-frequency maximum (LF, i.e., 1-100 Hz) is related to long-range ion transport, i.e., translational ion motion, while at high frequencies (HF, i.e., 10 4 -10 6 Hz), short-range (local) ion motion is displayed [10,17,27].The LF ion transport processes have been extensively studied in our previous investigations [19,28].The main focus of this study will be the HF resonance peak, to which the short-range movement of Cu II (NH 3 ) n species is expected to contribute [12].As displayed in Figure 4a, for SSZ-13 the HF resonance peak undergoes a significant shift to higher frequencies with increasing Cu loading (Cu/Al ratio) in the samples, from the starting value log ν res = 4.13 (ca.1.36 × 10 4 Hz) in H-SSZ-13 to the value log ν res = 5.33 (ca.2.15 × 10 5 Hz) for Cu(0.24)-SSZ-13.On the contrary, in the case of ZSM-5 zeolites, a slight shift of HF to lower frequencies was observed.The H-ZSM-5 presents the HF peak at log ν res = 5.27 (ca.1.84 × 10 5 Hz), while the HF peak for the sample with the highest Cu-loading was shifted to log ν res = 4.93 (ca.8.58 × 10 4 Hz).
Notably, even for Cu-exchange zeolites, due to the fact that only a fraction of Brønsted sites is occupied with Cu ions, there is still the presence of protons in the zeolite framework, and NH 4 + ions will be formed in the respective temperature range and undergo local motion [15,29].Therefore, the HF resonance peak is supposed to result from a superposition of the local movement of both NH 3 -solvated H + and Cu 2+ ions.The shift to higher frequencies of the HF peak in SSZ-13 with increasing Cu/Al implies that the formed Cu II (NH 3 ) n species leads to a shorter relaxation time (Table 2), and thus becomes the frequency-determining species for the HF process.The fact that an opposite trend was observed for ZSM-5, i.e., an increase of the HF relaxation time with increasing Cu loading (Table 2), indicates that the crystalline framework plays a decisive role in the local motion of Cu II (NH 3 ) n .
It is however ruled out that non-solvated Cu 2+ ions can get similarly mobile, since a shift to higher frequencies was not observed under pure N 2 (Figure S1).On the contrary, the HF ν res for pristine SSZ-13 zeolite in N 2 was significantly shifted to lower frequencies by the presence of non-solvated Cu 2+ .This suggests that the occupancy of Brønsted sites by the bivalent metal ions decreases the amount of H + that is available for local and translational charge transport.

In Situ IS: HF Arrhenius Plots
The observed ion conducting processes are temperature-dependent, and their activation energy (E a ) can be derived from the equation where Y is the real part of the admittance (i.e., Y = 1/Z ) at the temperature-dependent ν res , σ is the specific conductivity of the zeolite, A is the pre-exponential factor (which depends on the charge and number of mobile species, their on-site oscillation frequencies and the hopping distances), T is temperature in Kelvin and k B is the Boltzmann constant [17,18,27].Here, ν res is always intended as the resonance frequency associated to the HF maximum (i.e., for the short-range ion motion) in the corresponding modulus plot at the selected temperature.Arrhenius-like plots (i.e., ln Y T vs. 1/T) are therefore helpful in the study of the temperature-dependent ion motion processes in zeolites [10,17,27].Figures 5a-d and 6a-d illustrate the Arrhenius plots for the SSZ-13 and ZSM-5 catalysts, respectively.A linear increase of ln (Y T) with decreasing 1/T was observed in measurements that were performed over pristine H-SSZ-13 and H-ZSM-5 in N 2 (Figures 5a and 6a, black squares).In both cases, the measured conductivity can only be associated to proton mobility, since no other cations are present and no solvent molecule (e.g., H 2 O, NH 3 ) can be supportive [15].The observed mechanism is characterized by a higher E a in H-SSZ-13 (E a = 94.1 ± 0.7 kJ•mol −1 ), as compared with H-ZSM-5 (E a = 58.7 ± 0.6 kJ•mol −1 ).We assume that this difference is related to the structural differences between the two zeolites, which is consistent with the finding that the local mobility of protons is lower in CHA zeolites than in other zeolites [30].The NH 3 -loaded samples show, in general, higher conductivity than the pristine counterparts, thanks to the contribution of NH 3 to form NH 4 + that exhibits a higher mobility in zeolites.The non-linearity of the profiles below 350 • C reflects a combined effect of NH 3 solvation and thermal activation, as revealed in our previous studies on proton transport in zeolites [15,29].Above 350 • C, NH 3 is completely desorbed and the ion mobility follows the same mechanism as observed in the pristine zeolites.A linear increase of conductivity with temperature under N 2 is also observed for the pristine Cu-exchanged zeolites, even though the effect of the Cu exchange level is different for the two crystalline frameworks (Figures 5b-d and 6b-d ).In both catalyst systems, the increased Cu loading leads to a progressive increase of the E a for the short-range H + motion (Figure 7).We therefore assume in a first approximation that the occupation of Brønsted sites by Cu 2+ , independently whether single-or double-bounded, affect the total background charge to the zeolite lattice in a way that the proton motion gets less likely with increasing Cu content and thus requires a higher activation energy.
The analysis of NH 3 -loaded zeolites is even more complex, since the temperature ranges for desorption of solvent molecules from Cu 2+ sites and for the removal of weakly bound NH 3 species are not clearly separated.Nevertheless, it is possible to individuate a "temperature window", between 190 and 250 • C, in which the impedance measurements are supposed to be most sensitive to Cu II (NH 3 ) n species, whereas still the contribution of NH 4 + has to be taken into account because of its co-existence in the respective temperature range (see NH 3 -TPD profiles in Figure 3).The E a associated to the local ion motion in the respective zeolite was calculated and is displayed in Figure 7.
After NH 3 adsorption, a significant decrease of the E a value was observed for both Cu-SSZ-13 and Cu-ZSM-5, as compared to that for the corresponding pristine zeolite.Cu-ZSM-5 shows an increase of E a with increasing Cu/Al, which is similar to the trend that was observed for the pristine zeolites in the NH 3 -free state.On the contrary, for Cu-SSZ-13, the E a is progressively lowered by the increase of the Cu amount, suggesting a favorable short range motion of Cu II (NH 3 ) n at a higher Cu loading.Combining the information that was obtained from the modulus plot analysis and the temperature-dependent measurements, it emerges that the presence of Cu II (NH 3 ) n is affecting the dipolar relaxation processes that take place in zeolites at high frequencies.In both CHA and MFI frameworks, while the ion motion properties changed with the availability of Cu II (NH 3 ) n species, a substantial difference was observed.For ZSM-5, the formation of Cu II (NH 3 ) n species led to an increase in the relaxation time and in the overall activation energy that was associated to the ion motion processes in the selected temperature window (180-240 • C).SSZ-13 shows different properties: the increase of Cu amount favors an ion motion process with a shorter relaxation time and lower activation energy, pointing to enhanced local ion motion at a higher Cu loading.
The different local environments of Cu ion in ZSM-5 and SSZ-13 may be responsible for their distinct ion motion properties.Theoretical calculations [31] proved that Cu 2+ in ZSM-5 is usually coordinated to an Al site in a six-member ring (6-MR) that is connected directly with a 10-MR channel, while a second Al required for charge balance is located in an adjacent 5-MR, which is part of a smaller channel.The larger dimensions of the MFI principal channels do not allow for Cu II (NH 3 ) n to find intermediate position of minimized potential energy.As a result, the transport of Cu II (NH 3 ) n to another couple of Al atoms facilitates long-range rather than local motion of the species within ZSM-5 zeolites.In this way, we assume that Cu II (NH 3 ) n species do not sensibly contribute to the HF conductivity.Besides, due to the Cu ion exchange, the amount of H + /NH 4 + that could contribute to the HF mobility decreased with the increase of Cu/Al ratio.On the contrary, for SSZ-13 zeolite, the movement of Cu II (NH 3 ) n is confined inside the CHA cage.When considering this, the shift of HF maximum to higher frequencies, as well as the lowering of E a with increasing Cu-loading is consistent with an increasing influence of the intra-cage mobility of the Cu II (NH 3 ) n .It has been repeatedly reported that NH 3 -solvated and -mobilized Cu ions are the active sites for low-temperature NH 3 -SCR reactions (below 250 • C) over Cu-SSZ-13 [11][12][13].Our observations now document that the unique CHA cage plays a crucial role to sustain high local mobility of the NH 3 -solvated Cu ions.It is thus of high interest to understand further the potential correlation between the NH 3 -SCR reactivity of Cu species and their local mobility in zeolite catalysts, as experimentally accessed in this work.
The H-form zeolites were converted to Cu-zeolites via wet ion-exchange for 24 h in aqueous solutions of Cu(NO 3 ) 2 •3H 2 O (≥99.5%, Merck, Darmstadt, Germany).In the case of SSZ-13, the temperature for the Cu ion exchange was set to 80 • C, in order to facilitate the ion diffusion in the small-pore zeolite framework [32].After ion exchange, the zeolite sample was recovered, washed three times with ultra-pure water, and subsequently, dried overnight at 100 • C. The Cu-exchange process was repeated up to three times, to obtain samples with increasing Cu 2+ contents.The dried products were then ground and calcined at 500 • C for 2 h.For comparison, the same calcination treatment was carried out over the H-form zeolites as well.

Zeolites Characterization
The crystalline phase of the samples was determined by XRD using a STOE STADI P diffractometer (STOE, Darmstadt, Germany) with a Cu Kα radiation source (λ = 1.54059Å).The morphology of the catalysts was investigated by SEM with a Zeiss DSM 982 Gemini microscope (Zeiss, Oberkochen, Germany).The Si, Al and Cu contents of each sample were calculated from the elemental analysis obtained via ICP-OES.Before measuring, the zeolite samples were dissolved in HF.
NH 3 -TPD was used to study the acidity of the zeolites and the identity of the NH 3 adsorption sites.An amount of ca.50 mg of catalyst powder was introduced in a tubular quartz reactor, which was blocked with quartz wool and heated in a tube furnace (Carbolite, TZF, Sheffield, UK).Before each measurement, the sample was preheated for 1 h at 500 • C (ramping rate = 7.5 • C/min) in O 2 (flow rate = 50 sccm).The sample was then cooled to 50 • C in O 2 and it was hold for another 5 h at the same temperature in the N 2 atmosphere.The adsorption of NH 3 was carried out at 50 • C in flowing NH 3 atmosphere (1000 ppm in N 2 , 50 sccm) for 2 h.In order to remove the physisorbed NH 3 , the reactor was flushed with N 2 for 3 h at 50 • C. The temperature was subsequently increased from 50 to 700 • C (ramping rate = 2 • C/min) and the desorption of NH 3 was measured by means of an UV gas analyzer (LIMAS11 UV, ABB, Zürich, Switzerland).

IS Set-Up
To perform in situ IS measurements, alumina chips that were equipped with screen-printed gold inter-digital electrodes (IDEs) and backside integrated heaters were employed (Figure 8).A thick film of zeolite catalyst was deposited on the IDE and was subsequently calcined at 480 • C for 10 h.The homogeneity of the film was examined by means of SEM (Figure 2a-d).The chip was inserted in a stainless-steel chamber (inner volume = 30 cm 3 ) and was connected by electrical contact to an impedance analyzer (SI 1260, Solartron, Bognor Regis, UK) and a digital multimeter (Keithley 2400, Solon, OH, USA) for the power supply of the integrated heater.A ZnSe-window was used to cap the chamber and allows for calibrating the chip temperature by means of a pyrometer (KT 19.82, Heitronics, Wiesbaden, Germany).The gas atmosphere in the chamber is regulated by mass flow controllers (MFCs; MKS 1179A, MKS 1259C and MKS-647C, MKS Instruments, Andover, MA, USA).

Multi-Frequency In Situ IS
The in situ IS measurements were performed by applying an alternating voltage of 0.1 V rms, in frequency range between 0.1 Hz and 1 MHz.For each modulus spectrum at a selected temperature, 106 measuring points at equidistant logarithmic steps were recorded, with a total measurement duration of ca. 30 min.
Impedance values for modulus plots were collected at 200 • C in pure N 2 , and subsequently, in 100 ppm NH 3 in N 2 (flow rate = 100 sccm).In both cases, the IS measurement was initialized after the chip had been flushed for 40 min in the desired atmosphere.Before each measurement, the sample was pretreated for 1.5 h in pure O 2 at 300 • C.
To obtain the Arrhenius plots, a set of modulus spectra were registered in the temperature range between 150 and 450 • C, at a step size of 10 • C. In each step, the temperature was stabilized for 10 min before collecting IS data.Prior to each experiment, the measured sample was pretreated for 1.5 h in pure O 2 at 450 • C.

Conclusions
In the present study, in situ impedance spectroscopy was applied to investigate the formation and local mobility of Cu II (NH 3 ) n species resulting from the solvation and mobilization of Cu cations in Cu-exchanged zeolite catalysts during NH 3 -SCR.A combination of the local movement of H + /NH 4 + and Cu II (NH 3 ) n species led to a resonance peak at high frequencies (i.e., 10 4 -10 6 Hz) in the modulus plots of the IS data.With the increasing of Cu loading, the HF peak of SSZ-13 shifted significantly to higher frequencies, while a gradual shift to lower frequencies was observed for ZSM-5.This contrary effect of Cu loading on the local ion mobility is believed to result from a structural effect of zeolite: the CHA framework, because of its unique cage structure, seems to facilitate the local motion of the Cu II (NH 3 ) n species, so that they can contribute to the overall ion conductivity at high frequencies; in contrast, after NH 3 solvation, Cu 2+ ions in ZSM-5 appear to move across the zeolite lattice rather than locally.This assumption was corroborated by temperature-dependent IS measurements, which evidenced a decreasing activation energy for the high-frequency ion conduction process with increasing Cu loading, specifically for SSZ-13.These results show that Cu II species experience enhanced mobility in SSZ-13 under SCR-like conditions.This sheds new light on the participation of Cu II in the full reaction cycle and the specific properties of CHA zeolites in NH 3 -SCR.As the mobility of Cu II is a new parameter to be considered in the further refinement of the reaction mechanism, the impedance data reported here may be a valuable complementation to the established experimental and theoretical methods that generated the existing knowledge of the nature of NH 3 -SCR.

Figure 3 .
Figure 3. NH 3 desorption profiles obtained for SSZ-13 (a) and ZSM-5 (b) catalysts by NH 3 -TPD, with a ramping rate of 2 • C/min.Before TPD measurements, the NH 3 -loaded samples were flushed with N 2 for 3 h at 50 • C to remove physically adsorbed surface NH 3 species.The three desorption ranges are indicated by the colored areas: (i) weakly-bound NH 3 (yellow); (ii) NH 3 adsorbed on Cu sites (orange); and, (iii) NH 3 adsorbed on Brønsted sites (red).

Figure 4 .
Figure 4. Modulus plots of in situ IS results collected in NH 3 (100 ppm in N 2 ) at 200 • C over H-form and Cu-exchanged (a) SSZ-13 and (b) ZSM-5 zeolites.

Figure 7 .
Figure 7. Activation energy E a for the local motion of ions in the pristine zeolite (empty symbols) and in the NH 3 -loaded zeolites (full symbols) as a function of Cu/Al ratio for SSZ-13 (red symbols) and ZSM-5 (green symbols) catalysts.

Figure 8 .
Figure 8. Schematic representations of the IS measurement chamber (a) and the IDE chip (b) (adapted with permission from [16]); and, a picture of a typical IDE chip with zeolite thick film (c).

Table 1 .
Cu loadings and Cu/Al ratios determined by ICP-OES analysis.

Table 2 .
Resonance frequencies (ν res ) and corresponding relaxation times (τ) for local ion motion within the studied zeolites.The ν res values were obtained from the modulus plots of the in situ IS experiments under NH 3 (100 ppm in N 2 ) at 200 • C.