In-Exchanged CHA Zeolites for Selective Dehydrogenation of Ethane: Characterization and E ﬀ ect of Zeolite Framework Type

: In this study, the characterization of In-exchanged CHA zeolite (In-CHA (SiO 2 / Al 2 O 3 = 22.3)) was conducted by in-situ X-ray diffraction (XRD) and ammonia temperature-programmed desorption (NH 3 -TPD). We also prepared other In-exchanged zeolites with different zeolite structures (In-MFI (SiO 2 / Al 2 O 3 = 22.3), In-MOR (SiO 2 / Al 2 O 3 = 20), and In-BEA (SiO 2 / Al 2 O 3 = 25)) and different SiO 2 / Al 2 O 3 ratios (In-CHA(Al-rich) (SiO 2 / Al 2 O 3 = 13.7)). Their catalytic activities in nonoxidative ethane dehydrogenation were compared. Among the tested catalysts, In-CHA(Al-rich) provided the highest conversion. From kinetic experiments and in-situ Fourier transform infrared (FTIR) spectroscopy, [InH 2 ] + ions are formed regardless of SiO 2 / Al 2 O 3 ratio, serving as the active sites. ﬀ ect of zeolite framework and SiO 2 / Al 2 O 3 to achieve higher conversion compared to In-CHA. In-MFI, In-MOR, and In-BEA exhibit inferior activities compared to that of In-CHA, whereas a higher conversion value was achieved using In-CHA with a higher Al content (In-CHA(Al-rich)). Kinetic study and in-situ Fourier transform infrared (FTIR) spectroscopy indicated that [InH 2 ] + ions are formed in CHA zeolites regardless of SiO 2 / Al 2 O 3 ratio, and that they serve as catalytically active sites for the dehydrogenation of ethane. loading amount. In the catalytic dehydrogenation of ethane, In-CHA(Al-rich) exhibits the highest conversion value among the In-exchanged zeolites, including In-MFI, In-MOR, and In-BEA. The combined results of kinetic experiments and in-situ FTIR spectroscopy suggest that [InH 2 ] + ions are plausible active In-hydrides for In-CHA(Al-rich) and that the active In species and / or reaction mechanism for other In-exchanged zeolites are di ﬀ erent from those of In-CHA. This study indicates that the small pores in CHA zeolites play an important role in their unique catalytic activity. We are now investigating the e ﬀ ect of CHA pores on the formation of In hydrides and the reaction mechanism of ethane dehydrogenation.


Introduction
The valorization of ethane is highly demanded in chemical industry owing to increasing availability of inexpensive ethane from shale gas [1,2]. The most frequently used method for this purpose is the steam cracking to obtain ethylene which is an important feedstock for production of various fundamental chemicals, including polyethylene, styrene, ethylene oxide, and acetic acid. However, this method requires high reaction temperature (1073-1173 K) to achieve sufficient ethylene yields [3]. Moreover, the rapid cooling of the outlet gas is necessary for suppression of ethylene polymerization. The steam cracking of ethane is an energy-intensive process. Catalytic dehydrogenation of ethane to ethylene is a promising alternative to decrease the reaction temperature and to conserve energy consumption [4][5][6]. Various catalyst systems using platinum group metal-based alloys [7][8][9][10], Cr [11][12][13], and Ga [14,15], have been developed. However, the reported systems often suffer from low ethylene selectivity and catalysts deactivation due to coke formation. In addition, the applicability of those processes under industrially meaningful conditions, i.e., high ethane concentration regime, has rarely been investigated [11]. From the practical viewpoint, the development of selective and durable catalyst systems that are workable even under the high concentration conditions is highly demanded.
Group 13 metal-exchanged zeolites have attracted much attention as catalysts for the transformation of light alkanes. Ga-exchanged MFI zeolites (Ga-MFI) have been most widely studied since the 1990's [38][39][40][41]. Ono and coworkers prepared Ga-MFI using liquid-phase ion-exchange and then investigated the mechanism of the dehydrocyclodimerization (DHCD) of propane to aromatics [42]. Improving the activity of Ga-MFI by loading larger amounts of Ga species through reductive solid-state ion-exchange (RSSIE) was investigated [43]. Several cationic Ga species, such as Ga + , [GaH] 2+ , and [GaH 2 ] + ions, have been reported as catalytically active species for the dehydrogenation of propane [44][45][46]. Recently, Bell and coworkers reported that [GaH] 2+ rather than [GaH 2 ] + ions are active species using a combination of kinetic, spectroscopic, and theoretical studies [47], whereas Lewis-Brønsted acid pairs (Ga + and H + ) have been proposed as active sites by the group of Lercher based on their investigation of the Ga/Al effect and acidity characterization along with theoretical investigation [48]. Various Ga-exchanged zeolites including Ga-MFI have been applied to the transformation of other light alkanes to aromatic compounds, the dehydrogenation of light alkanes, and reactions of methane with light alkenes to produce higher hydrocarbons [49][50][51].
In contrast to that by Ga-exchanged zeolites, studies on dehydrogenative transformation of light alkanes by In-exchanged zeolites have been rarely reported. Their preparation by RSSIE between In 2 O 3 and proton-type zeolites has been studied by several groups [52]. The reaction is described in Equation (1).
where Z − denotes a zeolite anion. In the context of catalysis, Hert and coworkers prepared In-exchanged MFI (In-MFI) by RSSIE and compared its catalytic activity for propane DHCD with that of Ga-MFI [53,54]. They reported that In-MFI exhibits poor durability due to the reduction of active In + cations to In metal during the reaction. The reaction of methane with ethylene/benzene was investigated by the group of Baba using In-MFI, for which the activation of methane at In + cations coordinated to zeolite anions was proposed [55]. Recently, our research group has reported that In-exchanged CHA zeolite (In-CHA) catalyzes the selective dehydrogenation of ethane and exhibits high durability [56]. A combination of experimental and theoretical studies revealed that [InH 2 ] + ions, which are formed in situ through the reaction of In + cations with H 2 , serve as catalytically active sites. However, the detailed investigation including effects of zeolite framework and SiO 2 /Al 2 O 3 on ethane dehydrogenation were not conducted. As a continuation of our previous study, we herein report the detailed characterization of In-CHA by in-situ X-ray diffraction (XRD) and ammonia temperature-programmed desorption (NH 3 -TPD). We also investigated the effect of zeolite framework and SiO 2 /Al 2 O 3 to achieve higher conversion compared to In-CHA. In-MFI, In-MOR, and In-BEA exhibit inferior activities compared to that of In-CHA, whereas a higher conversion value was achieved using In-CHA with a higher Al content (In-CHA(Al-rich)). Kinetic study and in-situ Fourier transform infrared (FTIR) spectroscopy indicated that [InH 2 ] + ions are formed in CHA zeolites regardless of SiO 2 /Al 2 O 3 ratio, and that they serve as catalytically active sites for the dehydrogenation of ethane.

XRD
The RSSIE reactions of In 2 O 3 and zeolites under H 2 at high temperature have been investigated by H 2 temperature-programmed reduction (H 2 -TPR) and/or thermogravimetric analysis in several previous studies [52,54,57]. However, in-situ characterization of In 2 O 3 and zeolites has not been reported yet. We have reported the preparation of In-CHA through RSSIE of CHA-supported In 2 O 3 obtained by impregnation (In 2 O 3 /CHA), revealing the formation of highly dispersed In species by spectroscopic and microscopic analysis [58]. In this study, in-situ XRD was conducted for RSSIE of In 2 O 3 /CHA under 5% H 2 /N 2 flow. Figure 1a shows a series of XRD patterns obtained at different temperatures. The characteristic peak at 2θ = 30.6 • derived from In 2 O 3 (222) [59] and the typical diffraction pattern of CHA zeolites [60] were observed at 303 and 473 K, respectively. The intensity of the peak derived from In 2 O 3 was slightly decreased at 573 K and the peak was no longer observed at 773 K ( Figure 1b). After cooling to 303 K, no peaks derived from In 2 O 3 were observed. The sharp peaks in the zeolite diffraction patterns at 2θ = 9.6 • and 20.8 • decreased in intensity above 573 K due to temperature-induced disordering and then returned to their initial intensities upon cooling from 773 to 303 K [61]. These observations indicate the occurrence of RSSIE reaction at temperatures from 573 to 773 K, which is consistent with the results of H 2 -TPR measurement in our previous study [56]. A series of In-exchanged zeolites including In-MFI, In-MOR, In-BEA, and In-CHA(Al-rich) were also prepared through similar RRSIE reactions of In 2 O 3 supported on the corresponding zeolites. The XRD measurements showed that the peaks assignable to In 2 O 3 (222), (400), and (440) disappeared after RSSIE in each case (Figure 2), indicating the high dispersion of In species.

XRD
The RSSIE reactions of In2O3 and zeolites under H2 at high temperature have been investigated by H2 temperature-programmed reduction (H2-TPR) and/or thermogravimetric analysis in several previous studies [52,54,57]. However, in-situ characterization of In2O3 and zeolites has not been reported yet. We have reported the preparation of In-CHA through RSSIE of CHA-supported In2O3 obtained by impregnation (In2O3/CHA), revealing the formation of highly dispersed In species by spectroscopic and microscopic analysis [58]. In this study, in-situ XRD was conducted for RSSIE of In2O3/CHA under 5% H2/N2 flow. Figure 1a shows a series of XRD patterns obtained at different temperatures. The characteristic peak at 2θ = 30.6° derived from In2O3 (222) [59] and the typical diffraction pattern of CHA zeolites [60] were observed at 303 and 473 K, respectively. The intensity of the peak derived from In2O3 was slightly decreased at 573 K and the peak was no longer observed at 773 K ( Figure 1b). After cooling to 303 K, no peaks derived from In2O3 were observed. The sharp peaks in the zeolite diffraction patterns at 2θ = 9.6° and 20.8° decreased in intensity above 573 K due to temperature-induced disordering and then returned to their initial intensities upon cooling from 773 to 303 K [61]. These observations indicate the occurrence of RSSIE reaction at temperatures from 573 to 773 K, which is consistent with the results of H2-TPR measurement in our previous study [56]. A series of In-exchanged zeolites including In-MFI, In-MOR, In-BEA, and In-CHA(Al-rich) were also prepared through similar RRSIE reactions of In2O3 supported on the corresponding zeolites. The XRD measurements showed that the peaks assignable to In2O3 (222), (400), and (440) disappeared after RSSIE in each case (Figure 2), indicating the high dispersion of In species.

NH3-TPD
The acidity profiles of In-CHA samples with different In/Al values (In/Al = 0.2, 0.4, and 0.8) were investigated by NH3-TPD. In each case, the catalyst was exposed to NH3/He at 473 K followed by purging with He. Thereafter, the temperature was increased from 473 to 973 K under He flow while the desorbed NH3 (m/z = 16) was monitored. The TPD profiles of proton-type CHA (H-CHA) and a series of In-CHA samples are shown in Figure 3. The profile of H-CHA showed a wide peak around 700-800 K due to NH3 desorbed from Brønsted acid sites. The intensity of the peak decreased with the increase of In/Al. This trend is similar to that observed in the NH3 adsorption experiment [56]. In addition, a desorption peak around 600 K appeared upon In loading and its intensity increased with In/Al ratio. This peak might be due to NH3 weakly adsorbed on In + cations [62]. The results of NH3-TPD are consistent with the occurrence of RSSIE reactions as described in Equation (1)

NH 3 -TPD
The acidity profiles of In-CHA samples with different In/Al values (In/Al = 0.2, 0.4, and 0.8) were investigated by NH 3 -TPD. In each case, the catalyst was exposed to NH 3 /He at 473 K followed by purging with He. Thereafter, the temperature was increased from 473 to 973 K under He flow while the desorbed NH 3 (m/z = 16) was monitored. The TPD profiles of proton-type CHA (H-CHA) and a series of In-CHA samples are shown in Figure 3. The profile of H-CHA showed a wide peak around 700-800 K due to NH 3 desorbed from Brønsted acid sites. The intensity of the peak decreased with the increase of In/Al. This trend is similar to that observed in the NH 3 adsorption experiment [56]. In addition, a desorption peak around 600 K appeared upon In loading and its intensity increased with In/Al ratio. This peak might be due to NH 3 weakly adsorbed on In + cations [62]. The results of NH 3 -TPD are consistent with the occurrence of RSSIE reactions as described in Equation (1).

Effect of Zeolite Host upon Ethane Dehydrogenation Catalyzed by In-Exchanged Zeolites
We previously found that In-CHA with SiO2/Al2O3 = 22.3 and In/Al = 0.8 exhibits high selectivity for dehydrogenation of ethane to ethylene and carbon balance at 933 K. In this work, In-MFI, In-MOR, and In-BEA having similar SiO2/Al2O3 ratios (22.3, 20.0, and 25.0, respectively) and In-CHA(Alrich) (SiO2/Al2O3 = 13.7) were prepared using the same In/Al value of 0.8 and compared with In-CHA in ethane dehydrogenation (Table 1). In-MFI, In-MOR, and In-BEA showed lower conversions of 12.9%, 14.9%, and 13.5%, respectively, than that for In-CHA (25.9%) (entries 1-3 vs. entry 5) despite their larger zeolite pores (5.6 Å × 5.3 Å, 7.0 Å × 6.5 Å, and 7.6 Å × 7.3 Å, respectively) compared to that of CHA (3.8 Å × 3.6 Å). In contrast, In-CHA(Al-rich) exhibited a superior conversion (37.9%) while maintaining the high selectivity of 96.1% (entry 5). For all the In-exchanged zeolites, high selectivity for ethylene (>96%) and carbon balance (>93%) were observed while main byproduct was methane. The ethane dehydrogenation was conducted under severe reaction condition (80% C 2 H 6 /N 2 , 993 K) using 0.1 g of In-CHA(Al-rich) ( Figure 4). The conversion and selectivity values at 1 h were 32.7% and 90.7%, respectively. Although the conversion value decreased to 30.9% during the 40 h reaction, good selectivity was maintained. This catalytic performance is comparable with those of other reported catalyst systems for nonoxidative ethane dehydrogenation [4]. The catalyst was reusable after regeneration treatment (50% O 2 /N 2 for 60 min and 10% H 2 /N 2 for 30 min at 993 K) without significant loss of activity or selectivity (32.5% conversion and 90.8% selectivity at 1 h) ( Figure 4). The diffraction pattern derived from CHA framework was maintained after the reaction, as confirmed by XRD measurement (Figure 2), which is consistent with the high reusability of In-CHA(Al-rich).
[a] The determination was performed using a gas chromatography (GC).
[b] The data were previously reported in our paper [56].
The ethane dehydrogenation was conducted under severe reaction condition (80% C2H6/N2, 993 K) using 0.1 g of In-CHA(Al-rich) (Figure 4). The conversion and selectivity values at 1 h were 32.7% and 90.7%, respectively. Although the conversion value decreased to 30.9% during the 40 h reaction, good selectivity was maintained. This catalytic performance is comparable with those of other reported catalyst systems for nonoxidative ethane dehydrogenation [4]. The catalyst was reusable after regeneration treatment (50% O2/N2 for 60 min and 10% H2/N2 for 30 min at 993 K) without significant loss of activity or selectivity (32.5% conversion and 90.8% selectivity at 1 h) (Figure 4). The diffraction pattern derived from CHA framework was maintained after the reaction, as confirmed by XRD measurement (Figure 2), which is consistent with the high reusability of In-CHA(Al-rich).    (Figure 5a). The was more than five-time higher tion were also investigated. The values. The ΔH ⧧ of In-CHA(Aln-CHA, while In-MFI, In-MOR, ol, respectively). Comparison of es ([InH2] + or [InH] 2+ ions) are that 2) the active species and/or r In-exchanged zeolites.
ized to amount of In obtained ne dehydrogenation (873−933 (Al-rich) (red). The plots were e data for In-CHA (blue) were In-CHA(Al-rich) to determine K, the band derived from the ( Figure 6a) cations is much higher (ca. 400 kJ/mol). The experimental ΔH ⧧ value for ethane dehydrogenation based on Eyring plot was determined as 236 kJ/mol, suggesting that the catalytically active sites are In hydrides. To obtain insight into the effect of the zeolite, the turnover frequency (TOF) values normalized to the total amount of In species were compared (Figure 5a). The TOF value for In-CHA(Al-rich) was similar to that for In-CHA and was more than five-time higher than those for other In-zeolites. Eyring plots for ethane dehydrogenation were also investigated. The obtained plots are shown in Figure 5b with estimated apparent ΔH ⧧ values. The ΔH ⧧ of In-CHA(Alrich) was determined to be 258 kJ/mol, which is similar to that for In-CHA, while In-MFI, In-MOR, and In-BEA showed much higher ΔH ⧧ values (364, 300, and 322 kJ/mol, respectively). Comparison of TOF and ΔH ⧧ values among In-zeolites indicates that 1) In hydrides ([InH2] + or [InH] 2+ ions) are possible active sites for In-CHA(Al-rich) rather than In + cations and that 2) the active species and/or reaction mechanism are different between In-CHA zeolites and other In-exchanged zeolites. Furthermore, in-situ FTIR spectroscopy was conducted for In-CHA(Al-rich) to determine possible In hydrides in In-CHA(Al-rich). After H2 treatment at 773 K, the band derived from the stretching vibration of In−H (ν(In−H)) appeared around 1700 cm −1 (Figure 6a). This band can be deconvoluted into two peaks assignable to ν(In−H) (1777 and 1715 cm −1 ) with other peaks (1661 and 1610 cm −1 ) (Figure 6b). Based on our previous vibration analysis by density functional theory calculations [56] value for ethane dehydrogenation based on Eyring plot was determined as 236 kJ/mol, suggesting that the catalytically active sites are In hydrides. To obtain insight into the effect of the zeolite, the turnover frequency (TOF) values normalized to the total amount of In species were compared (Figure 5a). The TOF value for In-CHA(Al-rich) was similar to that for In-CHA and was more than five-time higher than those for other In-zeolites. Eyring plots for ethane dehydrogenation were also investigated. The obtained plots are shown in Figure 5b with (Figure 5a). The In-CHA(Al-rich) was similar to that for In-CHA and was more than five-time higher ther In-zeolites. Eyring plots for ethane dehydrogenation were also investigated. The are shown in Figure 5b  ions) are sites for In-CHA(Al-rich) rather than In + cations and that 2) the active species and/or nism are different between In-CHA zeolites and other In-exchanged zeolites.
) Effect of zeolite host on turnover frequency (TOF) normalized to amount of In obtained ) Eyring plots and apparent activation enthalpies for the ethane dehydrogenation (873−933 -MFI (pink), In-MOR (green), In-BEA (black), and In-CHA(Al-rich) (red). The plots were der conditions obtaining ethane conversions below 15%. The data for In-CHA (blue) were reported in our previous paper [56]. re, in-situ FTIR spectroscopy was conducted for In-CHA(Al-rich) to determine rides in In-CHA(Al-rich). After H2 treatment at 773 K, the band derived from the ation of In−H (ν(In−H)) appeared around 1700 cm −1 (Figure 6a). This band can be nto two peaks assignable to ν(In−H) (1777 and 1715 cm −1 ) with other peaks (1661 and ure 6b). Based on our previous vibration analysis by density functional theory ], the highest peak at 1715 cm − (Figure 5a). The TOF value for In-CHA(Al-rich) was similar to that for In-CHA and was more than five-time higher than those for other In-zeolites. Eyring plots for ethane dehydrogenation were also investigated. The obtained plots are shown in Figure 5b with estimated apparent ΔH ⧧ values. The ΔH ⧧ of In-CHA(Alrich) was determined to be 258 kJ/mol, which is similar to that for In-CHA, while In-MFI, In-MOR, and In-BEA showed much higher ΔH ⧧ values (364, 300, and 322 kJ/mol, respectively). Comparison of TOF and ΔH ⧧ values among In-zeolites indicates that 1) In hydrides ([InH2] + or [InH] 2+ ions) are possible active sites for In-CHA(Al-rich) rather than In + cations and that 2) the active species and/or reaction mechanism are different between In-CHA zeolites and other In-exchanged zeolites. Furthermore, in-situ FTIR spectroscopy was conducted for In-CHA(Al-rich) to determine possible In hydrides in In-CHA(Al-rich). After H2 treatment at 773 K, the band derived from the stretching vibration of In−H (ν(In−H)) appeared around 1700 cm −1 (Figure 6a). This band can be deconvoluted into two peaks assignable to ν(In−H) (1777 and 1715 cm −1 ) with other peaks (1661 and 1610 cm −1 ) (Figure 6b). Based on our previous vibration analysis by density functional theory calculations [56], the highest peak at 1715 cm − t was determined as 236 kJ/mol, suggesting that the tain insight into the effect of the zeolite, the turnover amount of In species were compared (Figure 5a). The that for In-CHA and was more than five-time higher ethane dehydrogenation were also investigated. The imated apparent ΔH ⧧ values. The ΔH ⧧ of In-CHA(Alis similar to that for In-CHA, while In-MFI, In-MOR, 64, 300, and 322 kJ/mol, respectively). Comparison of ates that 1) In hydrides ([InH2] + or [InH] 2+ ions) are than In + cations and that 2) the active species and/or HA zeolites and other In-exchanged zeolites. requency (TOF) normalized to amount of In obtained n enthalpies for the ethane dehydrogenation (873−933 A (black), and In-CHA(Al-rich) (red). The plots were versions below 15%. The data for In-CHA (blue) were .
was conducted for In-CHA(Al-rich) to determine r H2 treatment at 773 K, the band derived from the ed around 1700 cm −1 (Figure 6a). This band can be −H) (1777 and 1715  In-CHA(Al-rich) to determine K, the band derived from the ( Figure 6a). This band can be ions) are possible active sites for In-CHA(Al-rich) rather than In + cations and that (2) the active species and/or reaction mechanism are different between In-CHA zeolites and other In-exchanged zeolites. Furthermore, in-situ FTIR spectroscopy was conducted for In-CHA(Al-rich) to determine possible In hydrides in In-CHA(Al-rich). After H2 treatment at 773 K, the band derived from the stretching vibration of In−H (ν(In−H)) appeared around 1700 cm −1 (Figure 6a). This band can be deconvoluted into two peaks assignable to ν(In−H) (1777 and 1715 cm −1 ) with other peaks (1661 and 1610 cm −1 ) (Figure 6b). Based on our previous vibration analysis by density functional theory calculations [56], the highest peak at 1715 cm −1 is derived from the symmetric ν(In−H) of [InH2] + ions, while the smaller peak at 1777 cm −1 is assignable to the asymmetric ν(In−H) of [InH2] + ions or the ν(In−H) of [InH] 2+ ions. These results indicate that the main In hydrides are [InH2] + rather than [InH] 2+ in In-CHA(Al-rich). The other peaks might be derived from different-type adsorbed water molecules [63,64]. From the above combined results, [InH2] + ions are likely formed in In-CHA zeolites regardless of SiO2/Al2O3 ratio and are likely to be the active sites. using In-MFI (pink), In-MOR (green), In-BEA (black), and In-CHA(Al-rich) (red). The plots were obtained under conditions obtaining ethane conversions below 15%. The data for In-CHA (blue) were previously reported in our previous paper [56].
Furthermore, in-situ FTIR spectroscopy was conducted for In-CHA(Al-rich) to determine possible In hydrides in In-CHA(Al-rich). After H 2 treatment at 773 K, the band derived from the stretching vibration of In−H (ν(In−H)) appeared around 1700 cm −1 (Figure 6a). This band can be deconvoluted into two peaks assignable to ν(In−H) (1777 and 1715 cm −1 ) with other peaks (1661 and 1610 cm −1 ) (Figure 6b). Based on our previous vibration analysis by density functional theory calculations [56], the highest peak at 1715 cm −1 is derived from the symmetric ν(In−H) of [InH 2 ] + ions, while the smaller peak at 1777 cm −1 is assignable to the asymmetric ν(In−H) of [InH 2 ] + ions or the ν(In−H) of [InH] 2+ ions. These results indicate that the main In hydrides are [InH 2 ] + rather than [InH] 2+ in In-CHA(Al-rich). The other peaks might be derived from different-type adsorbed water molecules [63,64]. From the above combined results, [InH 2 ] + ions are likely formed in In-CHA zeolites regardless of SiO 2 /Al 2 O 3 ratio and are likely to be the active sites.
Catalysts 2020, 10, 807 7 of 11 1610 cm −1 ) (Figure 6b). Based on our previous vibration analysis by density functional theory calculations [56], the highest peak at 1715 cm −1 is derived from the symmetric ν(In−H) of [InH2] + ions, while the smaller peak at 1777 cm −1 is assignable to the asymmetric ν(In−H) of [InH2] + ions or the ν(In−H) of [InH] 2+ ions. These results indicate that the main In hydrides are [InH2] + rather than [InH] 2+ in In-CHA(Al-rich). The other peaks might be derived from different-type adsorbed water molecules [63,64]. From the above combined results, [InH2] + ions are likely formed in In-CHA zeolites regardless of SiO2/Al2O3 ratio and are likely to be the active sites.

Catalyst Preparation
The preparation of In-CHA was conducted according to our previous study [58].

XRD
In-situ XRD measurements during RSSIE of In 2 O 3 /CHA was conducted using a Cu Kα radiation source (Rigaku Ultima IV, Rigaku Corporation, Japan). XRD patterns were obtained while heating In 2 O 3 /CHA from 303 to 773 K at 20 K/min under 5% H 2 /N 2 flow (100 mL/min). For ex-situ XRD measurements, a Rigaku MiniFlex II/AP diffractometer with Cu Kα radiation (Rigaku Corporation, Japan) was used.

NH 3 -TPD
NH 3 -TPD experiments were investigated using a BELCAT (MicrotracBEL, Japan). The In-CHA with different In/Al ratio (0.2, 0.4, and 0.8) was prepared in situ under 5% H 2 /Ar at 773 K and then treated with 10% NH 3 /He at 473 K followed by He purge. After these treatments, the sample was heated from 473 to 973 K under He flow while the desorbed NH 3 (m/z = 16) was monitored by a mass spectrometer (BELMass (MicrotracBEL, Japan)).

Catalytic Tests
Non-oxidative dehydrogenation of ethane was performed in a fixed-bed continuous flow system at 933 K. The catalyst was prepared in situ by treatment of a 0.1 g of the In 2 O 3 immobilized on the corresponding zeolite under 10% H 2 /He flow (50 mL/min) at 773 K followed by He purge before the reaction. Thereafter, the catalyst was heated to 933 K, and 10% C 2 H 6 /He (10 mL/min) was introduced. For determination of conversions, yields, and selectivities, gas chromatography (GC) analysis was conducted using Shimadzu GC-14B (Shimadzu Corporation, Japan) combined with a SHINCARBON ST or a Unipack S column (Shinwa Chemical Industry Ltd., Japan). For dehydrogenation under a high concentration of ethane, the in-situ preparation of In-CHA(Al-rich) was carried out under 80% H 2 /N 2 (10 mL/min) at 993 K for 30 min, and the reaction was investigated at the same temperature under 10 mL/min of 80% C 2 H 6 /N 2 . The GC analysis was performed using an Agilent 490 micro-GC with a thermal conductivity detector. The conversion, selectivity, carbon balance, and TOF were calculated as follows. 3.5. In-Situ FTIR Spectroscopy FTIR spectroscopy study was conducted using a home-made in-situ cell and FT/IR-4100 (JASCO, Japan) with a mercury cadmium telluride detector. Prior to measurement, In-CHA was prepared in situ by H 2 treatment of a self-supported disk of In 2 O 3 /CHA (ca. 40 mg) at 773 K and was then kept under vacuum at 773 K for 2 h. After taking a background spectrum below 153 K cooling the cell using liquid N 2 , the sample was treated with H 2 flow at 773 K for 2 h. Then, the temperature was decreased to below 153 K again and the FTIR spectra were recorded without exposure to air.

Conclusions
In this work, a detailed characterization of In-CHA and the effect of zeolite on ethane dehydrogenation were described. In-situ XRD revealed that the RSSIE of In 2 O 3 /CHA occurs without decomposition of the zeolite framework. NH 3 -TPD measurement showed that the amount of acid sites decreases with the increase of In loading amount. In the catalytic dehydrogenation of ethane, In-CHA(Al-rich) exhibits the highest conversion value among the In-exchanged zeolites, including In-MFI, In-MOR, and In-BEA. The combined results of kinetic experiments and in-situ FTIR spectroscopy suggest that [InH 2 ] + ions are plausible active In-hydrides for In-CHA(Al-rich) and that the active In species and/or reaction mechanism for other In-exchanged zeolites are different from those of In-CHA. This study indicates that the small pores in CHA zeolites play an important role in their unique catalytic activity. We are now investigating the effect of CHA pores on the formation of In hydrides and the reaction mechanism of ethane dehydrogenation.