3.1 Phase composition
XRD was used to determine the crystalline structure of all synthesized catalysts. The diffraction peak intensity of each catalyst is similar and without other peaks, indicating the high purity of synthesized catalysts. Five typical diffraction peaks of MFI characteristics are 7.96°, 8.82°, 23.08°, 23.97° and 24.43°, which corresponding to the MFI skeleton structures of (011), (200), (501), (033) and (133) appear in all catalysts [31]. According to previous report [32] and the XRD result proved that neither the load nor the encapsulate Pt significantly changed the MFI structure of ZSM-5 under the preparation of Pt-containing zeolite catalysts. As shown in Fig. 2a and b, Pt/ZSM-5 series catalysts and Pt@ZSM-5 series catalysts can still maintain the structural stability after high temperature treatment in different atmosphere. Although the diffraction peaks corresponding to Pt or oxide crystal structures not were observed, ICP and EDS analysis exhibited that the existence of Pt element in the as-synthesized samples, indicating the low load amount or high dispersion of Pt in each catalyst [33].
3.2 TEM
TEM images revealed that the regular morphology of as-prepared catalysts, and size about 100 nm-200 nm (Fig. 3a-f). The size of the synthesized catalysts are smaller than reported in the literature [34], which is attributed to two stages of crystallization method. Specifically, first stage is the crystallization at 100 ℃ for 48 h to form zeolite crystal nucleus, and later crystallization at 170 ℃ make it grow up [30, 35].
For Pt/ZSM-5 catalyst, Fig. 3a showed that most Pt nanoparticles with size of 3.8 nm were dispersed on the surface and edge of the ZSM-5 heterogeneously. Pt increased to 6.7 nm after hydrogen at 400 ℃ for Pt/ZSM-5 (4R), which result from lower binding energy between metals and support, promoting it agglomeration under the reduction process [36]. The Pt nanoparticles size of Pt/ZSM-5 (4R, 5O) further agglomerated to 10.7 nm due to Ostwald Ripening after oxygen treatment at 500 ℃ [37, 38]. Finally, Fig. 3d demonstrates that the Pt@ZSM-5 retains about 2.1 nm particles, which is much smaller than that of Pt/ZSM-5. In contrast, same hydrogen treatment of Pt nanoparticles encapsulated within Pt@ZSM-5 (4R), resulted in slight sintering, which arose from Brownian motion of metal was inhibited by spatial confining effect of ZSM-5 channels. Remarkably, Pt@ZSM-5 (4R, 5O) retains Pt nanoparticles with about 2.4 nm, indicating that redispersion was achieved after oxygen treatment at 500 ℃. This demonstrates that the conditions of pretreatment are different for whether achieving the redispersion of Pt over Pt/ZSM-5 and Pt@ZSM-5. In addition, the EDS images proved that Si, Al and Pt elements were homogeneously distributed over the entire Pt@ZSM-5. These results illustrated that the spatial confining effect of ZSM-5 plays a pivotal role in stabilizing and dispersing Pt nanoparticles [39, 40].
3.3 Texture characteristics of catalyst
The texture characteristics and other parameters of the different catalysts were listed in Table 1. ICP-AES was used to determine the Pt content and Si/Al ratio of the catalysts. Similar content of Pt and Si/Al = 52 was obtained for Pt/ZSM-5 and Pt@ZSM-5 catalysts. Next, the reasons for Pt redispersion will be discussed.
Table 1
Texture characteristics of catalysts
Sample
|
Pt Loading (wt%)a
|
Pt Particle size(nm) b
|
H2 desorption quantity(µg)c
|
Dispersion(%)d
|
Amount of carbon(%)e
|
Pt/ZSM-5
|
0.32
|
3.8
|
0.317
|
19.3
|
0.67
|
Pt/ZSM-5(4R)
|
6.7
|
0.136
|
8.3
|
0.74
|
Pt/ZSM-5(4R, 5O)
|
10.7
|
0.064
|
3.9
|
0.86
|
Pt@ZSM-5
|
0.35
|
2.1
|
1.228
|
68.4
|
0.43
|
Pt@ZSM-5(4R)
|
3.6
|
0.619
|
34.5
|
0.52
|
Pt@ZSM-5(4R, 5O)
|
2.4
|
1.048
|
58.4
|
0.47
|
a) Quality of the Pt measured by ICP characterization. |
b) Several TEM was taken for each catalyst, and about 100 Pt nanoparticles were found through the TEM for particle size statistics. Mean cluster diameter (dTEM) was estimated from TEM analysis, dTEM = Σnidi3 / Σnidi2, where ni is the number of crystallites having a diameter di. |
c) H2 desorption quality was assessed by H2 Adsorption-Desorption. |
d) Pt dispersion was calculated by the D\(=\frac{2\bullet V\bullet {M}_{Pt}}{22414mL\bullet {mol}^{-1}\bullet {m}_{cat}\bullet wt\%}\), where V, MPt, mcat, wt% represent the desorption volume of H2, the atomic mass of Pt, catalyst quality and metal load, respectively. |
e) The amount of carbon deposition in the spent catalysts was measured by TG. |
3.4 H2–TPR
Generally, PtOx is reduced at 100 ℃-300 ℃, but the reduction of Pt2+ and Pt4+ coordinated with -(-O-Si≡)y of ZSM-5 requires about 450 ℃ and 550 ℃ [41, 42]. However, the reduction peak of pure ZSM-5 did not appear in the H2-TPR process. Figure 4a exhibited that the large reduction peak at 100 ℃-300 ℃ and small reduction peak at 400 ℃-600 ℃ for the Pt/ZSM-5, which indicates that more oxidized Pt on Pt/ZSM-5 surface is formed after the removal of Pt precursor during the preparation process. For H2-TPR spectra of Pt@ZSM-5, the reduction peak at high temperature indicates that a large number of Ptδ+ is formed, which has strong interaction with -(-O-Si≡)y of ZSM-5. Surprisingly, the high temperature reduction peak is higher than reported in the literature. Most likely, the introduce basic Na+ into the catalysts at the synthesis process, enhancing the strong metal-support interaction [35, 43].
H2-TPR spectra of the Pt@ZSM-5 series catalysts were compared in Fig. 4b. The reduction peaks at 100 ℃-300 ℃ of Pt@ZSM-5, Pt@ZSM-5 (4R) and Pt@ZSM-5 (4R, 5O) were very small and no obvious change after high temperature pretreatment in different atmosphere. In contrast, high temperature reduction peaks existed quite discrepancy. Among them, the reduction peak area of Pt@ZSM-5 (4R) appeared, but smaller than others, indicating that only part of Pt were reduced and agglomerated under hydrogen at 400 ℃ condition. XPS result of Pt@ZSM-5 (4R) demonstrated that part of Pt still maintained oxidation state (Pt2+), which also indicated the high stability of Pt encapsulated within ZSM-5. However, Pt@ZSM-5 (4R, 5O) increased not only the peak area but also the reduction temperature indicated that the dispersion of Pt was restored and metal-support interaction was strengthened. This may be caused by changing the electronic structure and coordination mode between Pt and support under oxygen high temperature treatment. Hence, we further inferred that Pt@ZSM-5 achieved redispersion in this condition.
3.5 XPS
According to the XPS results of Pt 4f in Fig. 5, the valence states and interaction of Pt over Pt/ZSM-5 and Pt@ZSM-5 series catalysts were researched. It has been reported that the B.E. position of Al 2p is 74.4 eV-74.8 eV, this overlaps with the 4f of partial Pt [44]. However, Pt 4f7/2, Pt2+ 4f7/2 and Pt4+ 4f5/2 correspond to 71.2 eV, 72.4 eV and 77.8 eV, respectively, which have certain band gap with Al 2p and could be easily divided [45]. The doublet due to Pt 4f7/2 and Pt 4f5/2 should be with the constrain: spin orbit separation = 3.3 eV and area ratio of Pt 4f7/2 : Pt 4f5/2 = 4:3, therefore, we can appropriately determine the peak of Pt 4f at the overlap with Al 2p [46]. Similar to most supported catalysts, the fresh Pt/ZSM-5 includes abundant of Pt0 and Ptδ+ (the ratio of Ptδ+/Pt0 = 1.12) [47]. But for the Pt@ZSM-5, the Ptδ+ is more than that of Pt0 (the ratio of Ptδ+/Pt0 = 3.45) under the same as Pt/ZSM-5 preparation conditions. This may be one of the reasons why the subsequent high temperature oxygen pretreatment generates in different dispersion results (Pt@ZSM-5 redispersion and Pt/ZSM-5 agglomeration).
For XPS of Pt@ZSM-5 series catalysts in Fig. 5b, the content of the Pt valence states changed with different atmosphere treatment. The oxidated Pt (Pt2+) still existed in Pt@ZSM-5 (4R), which has not been completely reduced, this explains why there was still a high temperature reduction peak of H2-TPR. However, the valence states and content of Pt within the Pt@ZSM-5 (4R, 5O) were restored to quite level of the Pt@ZSM-5 catalyst (the ratio of Ptδ+/Pt0 = 3.2). Therefore, it is merited to infer that more Ptδ+ within ZSM-5 were formed under oxygen high temperature pretreatment, which migrated and may been anchored by support to realize Pt redispersion.
3.6 FT-IR
FT-IR spectrums of ZSM-5, Pt/ZSM-5 and Pt@ZSM-5 in Fig. 6. The absorbance of the above samples at low wavenumber (< 2000 cm− 1) almost unvaried, which correspond to the stretching vibration of MFI framework structure of the ZSM-5 [35, 48]. Consisting with the structure of ZSM-5 almost undamaged after loading or encapsulating Pt in XRD patterns. It is generally believed that about 3450 cm− 1 (internal Si-OH nests), and 3600 cm− 1 corresponds to Brønsted acid -OH groups by skeleton aluminum [49]. ZSM-5 as a reference, the absorbance at 3320 cm− 1-3580 cm− 1 like a bulging bag of Pt/ZSM-5 decreased slightly. In contrast, the absorbance of Pt@ZSM-5 reduced significantly. This illustrates that most Ptδ+ combined with the internal -(-O-Si≡)y nests of the ZSM-5, which proved the anchor sites of Pt within ZSM-5. Hence, we further deduced that the Ptδ+ species within ZSM-5 pore migrated at high temperature oxygen treatment, then the internal -(-O-Si≡)y nests anchored the migrated Ptδ+ through the strong metal supports interaction to realize redispersion.
3.7 TG
It is generally believed that the weight loss below 150℃ is the weight of water, and the high temperature weight loss can be considered that the carbon deposition and oxygen generate carbon dioxide [50]. Therefore, according to the TG results of Pt/ZSM-5 and Pt@ZSM-5 catalysts in Fig. 7a, which have weight loss in the low temperature range (< 150 ℃) indicates the water absorption of ZSM-5. In this temperature range, the weight loss of Pt/ZSM-5 more than that of Pt@ZSM-5. Most likely, the Pt encapsulated within ZSM-5 occupied a certain space, which lead to a relatively small amount of water absorption. Comparing with the carbon deposition of catalysts at the same starting temperature (150 ℃), Pt/ZSM-5 had a larger carbon deposition than Pt@ZSM-5. However, for Fig. 7b, there were no obvious discrepancy in the TG of used Pt@ZSM-5 series catalysts. The reason we thought the similar carbon deposition was that the Pt particles size were limited by ZSM-5, indicating that Pt@ZSM-5 series catalysts have excellent stability and anti-carbon deposition ability under severe reaction conditions.
3.8 Evaluation of catalysts applied to POM
The catalytic performance of the Pt/ZSM-5 and Pt@ZSM-5 series catalysts when applied to POM were evaluated. As shown in Fig. 8, Pt/ZSM-5 catalyst indeed showed inferior methane conversion rate, which was only about 52%. Unfortunately, the selectivity of CO and H2 were 40% and 53%, respectively. In contrast, the methane conversion rate of the Pt@ZSM-5 catalyst reached 64%, and the selectivity of CO and H2 of 49% and 56%, respectively. With the increased of Pt particles size, the lower methane conversion rate of Pt@ZSM-5 (4R) catalyst reached 56%, and the selectivity of CO and H2 were also very ordinary. However, the Pt@ZSM-5 (4R, 5O) catalyst recovered the high methane conversion rate of 62%, more satisfactory that its CO and H2 selectivity were the largest among the four catalysts, which reached 58% and 61%. Generally believed that product selectivity is more significant in industrial catalysis. The catalytic effect is more attractive if the TOF value of the catalyst is considered. Smaller Pt particles size of the catalyst with higher activity, while encapsulated catalysts are superior to supported catalysts for durability and stability from the experimental results.