Metal-based catalysts for the non-oxidative dehydrogenation of light alkanes to light olefins

Abstract

Catalytic non-oxidative dehydrogenation of light alkanes is one of the most promising routes for the production of value-added olefins because of high selectivity to the target alkenes and increased supply of natural and shale gas. In this review, we emphasize the recent advances in the development of metal-based catalysts for the direct dehydrogenation of light alkanes to olefins. Metal-based catalysts including Pt-, Pd-, Rh- and Ni-based catalysts, particularly Pt-based bimetallic catalysts, are summarized and exemplified. A detailed discussion of the structure–performance correlations of these catalysts is presented. We systematically discussed the geometric and electronic effects of the promoters on these metal-based catalysts. Moreover, we highlighted recent advanced strategies for improving the dehydrogenation performance and the stability of metal-based catalysts. In light of the existing contributions, a perspective was provided to evaluate some of the potential avenues into future research. We hope that this review will inspire the rational design of metal-based catalysts to realize efficient direct dehydrogenation of light alkanes for olefins production.

---

1. Introduction

Light olefins, such as ethylene, propylene and butene, are bulk commodities and platform building blocks for the production of value-adding polymers, oxygenates and important chemical intermediates.^1–8 The demand for these useful olefins has increased rapidly.^9–11 Currently, steam cracking of crude-oil-derived naphtha and fluid catalytic cracking of heavy oil are the main technologies for the production of light olefins.¹² However, these two existing routes suffer from high-energy demands, enormous CO₂ emissions and low selectivity towards light olefins. In addition, these routes right now cannot produce sufficient light olefins to satisfy the increasing market demand, which motivates an interest in seeking other means for producing light olefins. Catalytic non-oxidative dehydrogenation of light alkanes including ethane, propane and butane offers an alternative, practical and environmentally friendly approach for the production of light olefins,¹³ because cheap light alkanes separated from abundant shale gas or natural gas can be readily used as the feedstock in this process, and high olefin selectivity can be obtained.⁷

The dehydrogenation of light alkanes involves the breaking of two carbon–hydrogen bonds with simultaneous release of a corresponding alkene together with a molecular hydrogen. This reaction is highly endothermic and typically operates at high temperature (400–750 °C), which makes it difficult to control the selectivity to the target olefins. Metal-based catalysts, Pt-based catalysts in particular, have been widely used for alkane dehydrogenation due to their excellent activity and high selectivity to olefins.^5,7,14 The addition of a promoter to metal-based catalysts has been shown to enhance the catalytic performance. Tin (Sn) is the most frequently used modifier, which might be due to its positive effects on the selectivity towards the target alkene products and suppression of catalytic deactivation. The first industrial alkane dehydrogenation technology (the Pacol process) on the basis of an alumina-supported platinum catalyst was commercialized by UOP for producing linear alkenes for detergents in 1968.¹⁵ Later, an alkalized-Al₂O₃-supported Pt–Sn catalyst was commercialized by UOP and utilized in their Oleflex processes to produce light olefins.^1,7 In addition, Pt–Sn bimetallic-based catalysts are also employed in two other industrial processes, the STAR process and the Linde–BASF process, for the direct dehydrogenation of light alkanes.^1,7

The catalytic properties of metal-based catalysts strongly depend on their structures.¹⁴ In general, all the surface-accessible metal atoms can serve as the active sites for catalyzing the dehydrogenation of light alkanes, and thus the dehydrogenation is structure insensitive. In striking contrast, the accompanying side reactions, such as hydrogenolysis, isomerization and coke formation, are significantly affected by the structure of metal particles as well as the acid sites on the support. The deactivation of metal-based catalysts is primarily due to coke deposition and metal sintering. Therefore, careful catalyst design that ensures high metal dispersity, suppresses the surface sites responsible for hydrogenolysis, over-dehydrogenation and coking, and stabilizes the metal particles is the key to develop efficient catalysts for dehydrogenation.

Oxidative dehydrogenation of light alkanes is another promising route for the production of olefins. Several interesting reviews have already been published in this field.^2–4,16,17 Compared with non-oxidative dehydrogenation, it suffers from deep oxidation of light alkanes and alkenes to CO_x, which results in loss of olefin selectivity and yield. In this regard, this review will focus on the non-oxidative dehydrogenation of light alkanes. We summarize the most recent and impactful progress in understanding the metal-catalyzed non-oxidative dehydrogenation of light alkanes, with a primary focus on Pt-, Pd-, Rh- and Ni-based catalysts. The structure–performance correlations of these catalysts are discussed. The roles of promoter addition are addressed. Some of the recent strategies, such as support engineering, advanced preparation methods and catalyst structural innovation to improve the dehydrogenation performance and the catalyst stability are also highlighted. It is noted that the dehydrogenation of cyclic alkanes and long-chain alkanes for the production of pure hydrogen and detergents, respectively, are also important topics in catalysis and the chemical industry. Several literature reviews dealing with these topics have been published.^15,18–20 However, these topics are out of the scope of this review and will not be addressed here.

2. The reaction

The dehydrogenation of light alkanes (eqn (1)–(7)) is thermodynamically limited, strongly endothermic and prone to volume expansion. Consequently, a high reaction temperature (500–700 °C) and/or low alkane partial pressure are required to achieve economically relevant conversions of 30–50%. Under a certain condition, the dehydrogenation equilibrium conversions for the light alkanes increase with the carbon numbers and the degrees of branching (C₂ < C₃ < C₄ < i-C₄).

Co-feeding hydrogen with light alkanes is frequently employed in the reaction systems. According to Le Chatelier's principle, the equilibrium conversion of light alkanes would decrease with co-feeding hydrogen. However, Bell et al. claimed that the addition of an optimal amount of hydrogen to the feed can enhance the formation of olefins by promoting the dehydrogenation of adsorbed alkyl species on the catalyst surface under kinetic conditions.^21,22 In addition, DFT calculations revealed that co-feeding hydrogen decreases the propylene adsorption strength while it increases the energy barriers for further dehydrogenation of propylene.²³ This could provide more free sites for the dehydrogenation and thus induce higher catalytic activity. In addition, co-feeding hydrogen can suppress coke formation through inhibiting the formation of coke precursors.

In general, the C–H bonds of light alkanes and olefins are more reactive than C–C bonds and a catalyst that possesses a higher activity of C–H scission over C–C cleavage is desirable. On the other hand, when the alkane molecules are effectively activated, the formed intermediates and the olefins are much more reactive than the alkanes and can easily undergo undesirable side reactions such as deep dehydrogenation, cracking, isomerization, and polymerization. These side reactions may result in low selectivity to alkenes and coke formation, which is a reason for catalyst deactivation. Fortunately, the active sites for the side reactions are not required for dehydrogenation. This makes it possible to potentially eliminate side reactions and exclusively enable C–H bond activation via rational catalyst design.

3. The catalysts

3.1 Pt-based catalysts

Pt is the most commonly utilized noble metal for the direct dehydrogenation of light alkanes, thanks to its facile C–H bond activation over C–C bond scission. While all surface-accessible Pt atoms can facilitate the dehydrogenation reaction in virtually the same manner, Pt–Pt ensembles with large enough domain sizes are active for side reactions (i.e. cracking, over-dehydrogenation, etc.). As a result, substantial effort has gone into the modification of the Pt surface geometric/electronic structures using metal/non-metal promoters, with bimetallic alloying being the predominate candidate developed in past decades. Table 1 lists the recently reported Pt-based catalysts for the non-oxidative dehydrogenation of propane.

Table 1 Comparison of the catalytic performance of direct propane dehydrogenation over various Pt-based catalysts

Catalysts	Metal loading (%)	Temp (°C)	WHSV (h^−1)	Feed composition	Conversion (%)	Selectivity (%)	Operation time (h)	Specific activity^a (s^−1)	Deactivation constant^b (h^−1)	Ref.
a Specific activity is defined as the moles of C3H6 formation per mole Pt atoms per second. b , where ConFinal and ConInitial are the final and initial conversions of the PDH reaction, respectively; t is the duration of a PDH reaction in hours; and kd is the deactivation constant (h^−1). The higher kd is, the faster the inactivation rate will be.
PtSn/Al2O3-sheet	0.36	590	9.4	C3H8/H2/N2 = 1/1.25/4	48.7–44.6	99.1	24	1.58	0.007	48
Pt–Sn/γ-Al2O3	0.5	590	3.0	C3H8/H2 = 4/1	29.4–22.7	76	6	0.21	0.058	41
0.5Pt1.5Sn/Al2O3	0.5	600	3.2	Pure C3H8	35.6	88.5	6	0.248	—	34
0.5Pt0.9Sn/Al2O3	0.9	590	5.2	C3H8/H2/He = 1/1.25/4	48.5–42.5	97	20	0.609	0.012	49
Pt–Sn/θ-Al2O3	0.5	550	11.8	C3H8/H2/He = 1/1/8	41.1–37.6	97.7–98.3	72	1.14	0.002	36
0.3Pt–0.2Sn–0.5K/θ-Al2O3	0.3	600	4	C3H8/H2 = 1/0.5	39.9–38.2	92.0–95.5	25	0.602	0.0029	52
Pt–Sn/mesoporous Al2O3	0.5	590	3.0	C3H8/H2 = 4/1	29.8–24.6	92	6	0.22	0.044	41
Pt/Mg(Sn)(Al)O@Al2O3	0.5	600	14	C3H8 : H2 : Ar = 1 : 0.5 : 2	48.3–43.0	86.4–98.1	47	1.46	0.0045	46
PtSn/Mg(Al)O	0.3	550	1.13	Pure C3H8	27.9–27.7	95.6–96.5	5	0.12	0.0015	55
Pt–Sn/MgAl2O4	0.39	580	2.2	C3H8/H2/He = 1/1/8	45.0–38.9	99.5	90	0.34	0.0028	145
Pt–Sn/CeO2	1	680	2.159	C3H8/He = 2/10	39.5	84.5	—	0.0887	—	66
K–PtSn@MFI	0.4	600	29.5	C3H8/He = 24/76	38.7–31.9	>97	25	3.5	0.012	147
K–PtSn@MFI	0.42	600	1.7	C3H8/N2 = 5/16	71–42	88–95	65	0.311	0.019	146
Pt–Sn/ZSM-5	0.5	590	3.0	C3H8/H2 = 4/1	33.1–26.3	47.7	6	0.24	0.054	41
Pt/Sn-ZSM-5	0.3	600	5.2	C3H8/N2 = 1/5.5	68–43	95.5–99.7	25	1.36	0.041	149
Pt–Na/Sn-ZSM-5	0.5	590	3	C3H8/H2 = 4/1	41.7–39.0	95.3–93.8	9	0.294	0.012	59
0.3Pt/0.5Sn–Si-Beta	0.3	550	1	C3H8/N2 = 5 : 95	27.5–25.2	99.1–99.9	24	0.13	0.007	62
Pt/Sn2.00-Beta	0.5	550	140	C3H8/H2/He = 1/1/8	51.1–46.1	95.4–98.6	24	16.45	0.0083	58
PtSnNa/Ce-ZSM-5	0.5	590	3	C3H8/H2 = 4/1	43.5–40.7	84.5–96.9	6.92	0.291	0.017	63
PtSnNa@SUZ-4	0.5	590	3	C3H8/H2 = 1/3	24.7–20.2	90–94.7	9	0.164	0.029	65
0.5Pt2Sn–1Na/SUZ-4	0.5	590	3	C3H8/H2 = 1/3	24.6–20.6	81.3–89.7	10	0.14	0.0229	64
PtSn/TS-1	0.5	590	3	C3H8/H2/N2 = 1/1/4	53.5–47.7	92.5	7	0.40	0.033	61
Pt–Sn/SBA-15	0.5	590	3.0	C3H8/H2 = 4/1	11.0–6.0	80	6	0.08	0.110	41
Pt/0.5Sn-SBA-15	1	580	8.25	C3H8/Ar = 7/3	43.8–38.2	97.2–97.3	6	0.42	0.0386	57
PtZn4@S-1-H	0.72	550	3.6	C3H8/N2 = 1/3	47.7–40.4	93.2–99.2	216.7	0.25	0.001	74
PtZn4@S-1-H	0.72	550	108	C3H8/N2 = 1/3	21.2–11.8	98.3–98.6	10.33	3.55	0.072	74
0.3Pt0.5Zn@S-1	0.3	550	6.5	C3H8/N2 = 11/19	45–42	99.0–99.9	60	1.53	0.002	75
Pt^0Zn^δ+/SiO2	3	550	32	C3H8/Ar = 1/4	35.3–26.6	97.6–96.3	30	0.45	0.014	77
Pt^0Zn^δ+/SiO2	3	550	75	C3H8/Ar = 1/4	30.2–16.1	98.1–95.0	30	0.91	0.027	77
PtNa/Zn-ZSM-5	0.5	590	3	C3H8/H2 = 4/1	40.6–37.6	93–97	10	0.279	0.013	70
PtGa–Pb/SiO2	3	600	30.7	C3H8/H2/He = 3.9/5/40	30.0–28.4	99.6	96	0.38	0.001	88
Pt/HT-10Ga	3	600	8.8	C3H8/He = 1/4	8–6	99	1.83	0.257	0.169	85
Ga^δ+Pt^0/SiO2	1.55	550	98.3	C3H8/Ar = 1 : 4	31.9–18.2	99	20	1.36	0.038	87
Ga^δ+Pt^0/SiO2	1.55	550	43.3	C3H8/Ar = 1 : 4	36.5–26.9	90.9	20	0.76	0.023	87
Pt/Mg(Ga)(Al)O	0.9	600	2.6	C3H8/H2/He = 20/25/55	16–11.4	99.2	2	0.056	0.196	84
PtGa/CeO2–Al2O3	1	600	10	C3H8/H2/Ar = 26/26/48	41.1–32.2	99.6	14.17	0.504	0.026	82
Pt/Mg(In)(Al)O	0.7	600	2.6	C3H8/H2/He = 20/25/55	20.4–16.3	98	2	0.183	0.14	90
Pt3In/SiO2	0.3	600	3	C3H8/H2/N2 = 14/14/72	17.5–16.5	96–97.5	2	0.207	0.073	93
PtIn/Mg(Al)O-600	0.6	620	3.3	C3H8/H2/Ar = 8/7/35	69–50	98	41.0	0.45	0.027	96
0.1Pt10Cu/Al2O3	0.1	550	4	C3H8/H2/H2 = 8/8/34	24.1–21.6	88–90	11.67	0.95	0.012	99
Pt–Cu/MgAl2O4	1	590	6.8	C3H8/H2/He = 19/19/2	25.7–21.0	87.5	24	0.24	0.011	98
PtFe@Pt/SBA-15	0.75	600	3.48	C3H8/H2/N2 = 26/26/48	35–31	85–87	1.33	0.170	0.136	106
Pt/0.5Mn-DMSN	0.5	590	2.4	C3H8/N2 = 1/2	56.1–40.1	95.6–99.3	100	0.31	0.0065	104
Pt–Ag/MgAl2O4	1	590	6.8	C3H8/H2/He = 19/19/2	30.6–16.7	95.1	24	0.28	0.034	98
Pt–Au/MgAl2O4	1	590	6.8	C3H8/H2/He = 19/19/2	33.7–16.6	60	24	0.31	0.040	98
Pt/ND@G	0.5	600	1.62	C3H8/N2 = 1/19	—	88	—	0.064	—	31
Pt/TiO2–Al2O3	1	600	10	C3H8/H2/N2 = 26/26/48	43.7–25.9	77.5–89.5	9.67	0.45	0.0976	29

---

3.1.1 Pt monometallic catalysts. The active sites existing on Pt nanoparticles are coordinatively saturated Pt sites at terraces and coordinatively unsaturated Pt sites at steps, corners and edges.^7,24,25 All these sites can catalyze dehydrogenation reactions, while the coordinatively unsaturated Pt sites are active for the side reactions.²⁴ In addition, the sizes and structures of Pt particles affect the intrinsic activity, selectivity and stability.^26–28 The reactions with different alkanes display distinctive behaviors at varying platinum particle sizes.^26–28 Chen et al. studied the effects of platinum cluster sizes (1–21 nm) supported on Mg(Al)O (ref. 27) or SBA-15 (ref. 26) on propane dehydrogenation. They found that the activity (TOF) decreases with increasing Pt particle size, while the selectivity to propylene and the stability are related in an inverse way. Contrarily, Bell et al. revealed that the TOFs for ethane dehydrogenation increase with Pt particle size, whereas the selectivity to ethylene is nearly constant.²⁸ The reason for such large discrepancies is unclear.

A typical catalyst for light alkane dehydrogenation is Pt/Al₂O₃.⁷ However, certain characteristics, such as electron-deficient Pt particles, weak interaction between Pt and Al₂O₃, and strong acidity of Al₂O₃ could lead to low selectivity to olefins, sintering and coking.²⁴ Therefore, engineering of the supports is implemented in the development of high-performance Pt monometallic catalysts to improve the metal–support interaction and induce the electronic effects. Studies on support innovation have been focused on TiO₂–Al₂O₃,²⁹ zeolites,³⁰ and nanocarbon materials,^31–33 especially a nanocarbon (ND@G) composed of a nanodiamond core and a defective, ultrathin graphene nanoshell.³¹ Up to 90% alkene selectivity and enhanced catalyst stability have been reported on these catalysts with novel supports.^29–33

3.1.2 Pt-based bimetallic catalysts. Despite the aforementioned effort, the overall performance of monometallic Pt catalysts is unsatisfactory, especially in terms of selectivity and stability, for light alkane dehydrogenation reactions. This is due to the large ensembles of Pt surface atoms, which are active for deep dehydrogenation of the desired olefins, leading to coke formation and hydrogenolysis. The addition of inactive or weak active metal promoters such as Sn, Zn, Ga, In, Cu, Mn, Fe, Co, Ni, Ag, Ge, Cr, Bi, Sb, La, Ce and Y has been reported to be an effective way for the modification of the structural and electronic properties of Pt, resulting in the enhancement of catalytic performances.
3.1.2.1 Pt–Sn catalysts. Tin (Sn) is the most commonly used promoter of Pt-based bimetallic catalysts for light alkane dehydrogenation. All the known platinum-based catalysts that have been used on an industrial scale or a pilot scale contain Sn as a modifier.⁷ The addition of Sn greatly improves the selectivity towards target alkene products and slows down catalytic deactivation.⁷ The promotional role of Sn in Pt-based catalysts has been ascribed to both geometric and electronic effects, which will be discussed in the following sections. In addition, the presence of Sn atoms plays a positive effect on the regeneration of deactivated Pt–Sn/Al₂O₃ catalysts by providing nucleation sites for mobilized Pt atoms during oxidative regeneration.³⁴ The catalyst structure, dispersion and the formation of an active phase significantly depend on the preparation methods,^35,36 pretreatment conditions,^37–40 support type^41,42 and Sn content.^28,43 The extent of alloy formation and the location of Sn in the catalysts have a great impact on the catalytic performances.^28,39 Upon reduction, several alloy phases such as PtSn, Pt₃Sn, Pt₂Sn₃, and PtSn₄ alloys could be formed and the Pt₃Sn alloy is widely recognized as the most active one for the dehydrogenation.⁴⁴ In some cases, Sn could serve as a partition agent⁴⁵ or a stabilizer for Pt clusters⁴⁶ but does not alloy with Pt. A number of supports including Al₂O₃,^47–49 alkalized Al₂O₃,^50–52 rare-earth-oxide doped Al₂O₃,⁵³ Mg(Al)O hydrotalcite,^{21,28,46,54,55} MgAl₂O₄ spinel,³⁶ ZnAl₂O₄,^7,56 SiO₂,^39,57 zeolites,^58–65 CeO₂⁶⁶ and carbon materials⁴⁵ have been employed for light alkane dehydrogenation. Among them, alkalized θ-Al₂O₃, ZnAl₂O₄ and ZrO₂ were used in the UOP, Uhde and BASF/Linde dehydrogenation processes, respectively.⁷ Novel supports, such as Al₂O₃ nanosheets,⁴⁸ hydrotalcites,⁴⁶ CeO₂,⁶⁶ siliceous zeolites^58,62 and carbon materials, have been developed and engineered to improve the dispersion and provide specific sites for anchoring Pt or Pt–Sn clusters in recent years. The obtained Pt–Sn catalysts with engineered supports showed remarkable catalytic performances, particularly in terms of stability.
3.1.2.2 Pt–Zn catalysts. Zinc has also been considered to be an efficient promoter in Pt-based bimetallic catalysts for light alkane dehydrogenation to olefins. After reduction with hydrogen, a Pt–Zn alloy would be formed and brings about remarkable changes in the structural, geometric, and electronic properties of Pt catalysts. In addition, the unalloyed Zn species on the support, such as framework Zn in the zeolites or single site Zn on the silica support, play an essential role in the stabilization of the active Pt–Zn alloy. Consequently, the catalytic performances could be significantly improved.

In some of the pioneer work, workers at British Petroleum (BP) firstly disclose a series of catalysts composed of Pt on a zincosilicate molecular sieve support with MFI and MEL structures for the dehydrogenation of ethane, propane and isobutane.⁶⁷ These catalysts possess a high dispersion of active sites, which might be Pt–Zn alloy, and give high activity, selectivity and coke resistance for light alkane dehydrogenation. Remarkably, the activity of the catalyst can be maintained even after 796 h time-on-stream. After that, a number of Pt–Zn alloys supported on various zeolites, including CIT-6,⁶⁸ HZSM-5,^69,70 Na-ZSM-5,^70,71 Na-Beta,⁷¹ Na-mordenite,⁷¹ Na-MCM-22,⁷² Na–Y,⁷¹ KL^71,73 and silicate-1 (ref. 74 and 75) zeolites, were reported. In some cases, alkali metal ions including K⁺ and Na⁺ were added to poison the Brønsted acid sites on zeolites in order to avoid cracking, oligomerization and coke formation.⁷² The addition of Zn can effectively increase the Pt dispersion.⁷⁰ The selectivity to propene with silica–aluminum zeolites is not as high as that with silicious zeolites, apparently due to the adverse effects of the acidity on silica–aluminum zeolites.^69–75 Remarkably, the Pt–Zn bimetallic particles encaged in silicate-1 zeolites exhibit extraordinary catalytic stability.^74,75 Other supports, such as SiO₂,^76–78 Mg(Zn)AlO_x hydrotalcites⁷⁹ and titanate ETS-2,⁸⁰ are also investigated in a number of studies. The Pt–Zn/SiO₂ catalysts prepared by advanced preparation methods, such as atomic layer deposition (ALD)⁷⁸ and surface organometallic chemistry (SOMC),⁷⁷ showed excellent performance, particularly in terms of productivity.

3.1.2.3 Pt–Ga catalysts. Gallium is another frequently used promoter in Pt-based bimetallic catalysts for light alkane dehydrogenation to alkenes. Upon reduction at high temperatures (e.g., 500 °C), Pt–Ga alloys with variable compositions were formed. The introduction of Ga results in a remarkable enhancement of selectivity toward the desired products and catalytic stability, which is also attributed to the combination of geometric and electronic effects.

Scelza et al. used a series of Pt–Ga/Al₂O₃ catalysts for direct propane dehydrogenation.⁸¹ With the introduction of Ga, the selectivity to propene was enhanced and the catalyst deactivation and the carbon formation are suppressed. Given that the addition of Ga slightly modified the acidity of the Al₂O₃ support, they proposed that the geometric and electronic effects of the Pt–Ga alloy are the main reasons for the enhanced catalytic performance. Catalyst stability can be further improved by using CeO₂–Al₂O₃ as a support.⁸² In this case, the incorporation of Ga³⁺ cations in the cubic fluorite structure of CeO₂ increases both lattice oxygen storage capacity and surface oxygen mobility and thus eliminates coke deposition. Bell and his co-workers investigated Pt/Mg(Ga)(Al)O catalysts with various Ga/Pt molar ratios for ethane and propane dehydrogenation.^22,83,84 They observed the formation of Pt–Ga alloyed nanoparticles at high reduction temperatures (>500 °C). The catalytic activity and selectivity towards olefins were dramatically enhanced with the addition of Ga. The selectivity could reach up to 100% at Ga/Pt = 5.4. The activity exhibited a volcano correlation to the Ga content with a maximum rate obtained at Ga/Pt = 1.4–5.4, presumably due to the dilution of the Pt ensemble by Ga. Further, Marin et al. found that Pt–Ga catalysts showed reversible alloy/segregation looping when exposed to a redox cycle alternating between 5%H₂/He and 20%O₂/N₂ at 600 °C.^85,86 This reversibility ensures catalyst stability during the reaction/regeneration process. Recently, a series of well-defined Pt–Ga/SiO₂,⁸⁷ PtGa–Pb/SiO₂,⁸⁸ and Ga₃₇Pt/Al₂O₃ (ref. 89) catalysts have been carefully prepared. All the catalysts showed remarkable productivity and stability, which is primarily due to the high dispersion and specific structure.

3.1.2.4 Pt–In catalysts. Indium is a promising candidate to promote Pt-based catalysts for the dehydrogenation of light alkanes. Compared with Pt–Sn and Pt–Ga catalysts, Pt–In catalysts were found to be more homogeneous in composition, leading to a less sensitive performance to the preparation methods.^90,91 Upon reduction, Pt and In can form seven intermetallic alloys, including Pt₃In, Pt₂In, Pt₁₃In₉, PtIn, Pt₂In₃, PtIn₂ and Pt₃In₇, the relative shares of which depend in part on the global In content.⁹² DFT calculations indicate that Pt₃In has a comparable activity to that of pure Pt and Pt₃Sn, while it can achieve higher selectivity towards propene than that of pure Pt.⁹³ Miller and his co-workers prepared a couple of silica-supported Pt–In intermetallic catalysts for ethane dehydrogenation.⁹² They proposed that the Pt₃In phase is more active than the PtIn₂ phase. Both phases showed higher intrinsic activity, selectivity to ethene (nearly 100%) and catalyst stability than those over monometallic Pt, which is due to the geometric and electronic effects of the Pt–In alloy.⁹²

Bell et al. applied a series of Pt/Mg(In)(Al)O catalysts in light alkane dehydrogenation.^90,94 They observed the formation of a Pt–In alloy with reduction in H₂ at temperatures above 400 °C. Furthermore, the content of In in the Pt–In alloy particles increases as the reduction temperature and bulk In/Pt molar ratio increase. X-ray absorption near edge structure (XANES) showed an increasing electron transfer from In to Pt with increasing In content in the Pt–In alloy, which ensures enhancement of catalytic performances. When the In/Pt ratio is 0.48, the Pt/Mg(In)(Al)O catalyst exhibited the highest activity for ethane and propane dehydrogenation, with an alkene selectivity approaching 100%.⁹⁰ The optimal ratio of In/Pt for n-butane dehydrogenation was between 0.33 and 0.88.⁹⁴ Compared with Pt/Mg(Sn)(Al)O and Pt/Mg(Ga)(Al)O catalysts, Pt/Mg(In)(Al)O exhibits the best performance for light alkane dehydrogenation in terms of activity and catalyst stability.⁹⁰ In addition, Guo et al. found that the Mg/Al molar ratio and calcination temperature significantly affect the structure and catalytic performance of Pt–In/Mg(Al)O catalysts.^95–97 With the optimal Mg/Al molar ratio (4 : 1) and calcination temperature (600 °C), the catalysts showed the lowest fraction of strong acid, largest surface area and highest dispersion of metal particles, which result in the best catalytic performance. However, the exact structure of the Pt–In alloy was not determined.

3.1.2.5 Other Pt-based bimetallic catalysts. Besides Sn, Zn, Ga and In as promoters, Cu,^98–102 Mn,^103,104 Fe,^105,106 Co,^106,107 Ni,^106,108 Ag,^98,99 Ge,^109,110 Cr,¹¹¹ Bi,¹¹² Sb,¹¹³ Ce,¹¹⁴ Y (ref. 114) and La (ref. 114) have been found to be good promoters for Pt-based catalysts in recent years. With the addition of these metals, the corresponding bimetallic alloys can be formed with various structures, such as random alloys, single-atom alloys (SAAs) and intermetallic compounds. Carefully controlling the surface and subsurface compositions can tune the catalytic performance. The geometric and electronic features of these alloys are responsible for the boosted catalytic performance.

Jung et al.¹⁰² and Gong et al.¹⁰¹ investigated Cu-modified Pt/Al₂O₃ for n-butane dehydrogenation and PDH, respectively. Copper modification enhanced alkene selectivity as well as catalyst stability. They attribute the improved performance to the electronic effects of the Pt–Cu alloy, which weakened the adsorption of propylene. Further, Pt–Cu/SiO₂ (ref. 100) and Pt–Cu/Al₂O₃ (ref. 99) single-atom alloy catalysts, in which Pt atoms were isolated by Cu, could activate C–H efficiently and improve the selectivity to alkene and coke resistance in propane and n-butane dehydrogenation. Ren et al. studied the promotion role of group IB metals (Cu, Ag, Au) in Pt/MgAl₂O₄ catalysts for PDH.⁹⁸ They found that the addition of Cu and Ag could significantly enhance the propane conversion and propylene selectivity while the introduction of Au does not, which can be explained by the dilution and electron donation of Cu and Ag to Pt. In addition, Pt–Cu/MgAl₂O₄ exhibited excellent regeneration stability, likely owing to the formation of thermally stable intermetallic compounds.

Gong's group investigated the promotion effect of 3d transition metals (Ni, Co, Fe) on Pt-based catalysts with controlled surface and subsurface compositions in PDH.¹⁰⁶ They found that the well-defined Pt-skin catalysts with different 3d transition metal atoms in the subsurface showed notable enhancement of selectivity to propene in the order Pt < PtNi@Pt < PtCo@Pt < PtFe@Pt, which corresponds to the calculated trend of d-band center shifting. Miller's group reported that the addition of Mn to Pt catalysts with different subsurface compositions could alter the catalytic and adsorption properties of Pt-based catalysts. They found that the catalyst with a Pt₃Mn intermetallic subsurface significantly improved C–H bond activation, leading to high selectivity to propene during PDH.¹⁰³ In addition, this group also reported several Co,¹⁰⁷ Fe,¹⁰⁵ Cr,¹¹¹ Bi,¹¹² and Sb (ref. 113) modified Pt-based catalysts with various compositions and intermetallic structures that are promising for light alkane dehydrogenation.

3.1.3 Pt–B catalysts. Besides metallic promoters for Pt-based dehydrogenation catalysts, boron is a promising non-metallic alternative, which has also been studied in conjunction with Pt for non-oxidative dehydrogenation of propane and n-butane.^115–119 The nature of boron, as an alloy, an amorphous oxide or both, depends on the catalytic system and the preparation method. Alexandrova et al. evaluated the effect of boron promotion on the stability of Pt nano-clusters via experimental and DFT methods in the ethylene desorption/reaction.^117–119 These studies suggest that the introduction of boron in the Pt cluster increases the sintering energy for a Pt atom to dissociate off the clusters by alloying with Pt, enhances metal cluster and support interaction via the formation of a B–O bond, substantially reduces the alkene affinity of the platinum clusters, and retards dehydrogenation leading to carbon deposition. Teplyakov and his co-workers studied the effect of boron on the catalytic activity for non-oxidative dehydrogenation of n-butane over a Pt–B/SiO₂ catalyst.¹¹⁶ They proposed that the incorporation of boron facilitates the migration of alkyl groups from Pt sites to tetrahedrally coordinated boron sites, thus enhancing the catalytic activity (3-fold) and stability. However, further characterization is necessary to elucidate the structure of the catalyst. More recently, a series of Pt–B/Al₂O₃ catalysts with different boron loadings with various impregnation sequences were studied for the dehydrogenation of propane.¹¹⁵ They only observed the promotional effect of boron when boron was introduced after Pt. They indicated that amorphous boron oxide was finely dispersed on the support and covered the platinum particles, based on the catalyst characterization and thermodynamic DFT calculations. The addition of boron decreases the amount of medium and strong acid sites on the support, which are responsible for coke formation. As a result, the selectivity to propene and the catalytic stability were improved.

3.2 Pd-based bimetallic catalysts

Pd-based bimetallic catalytic systems could be an alternative to Pt-based bimetallic catalysts for light alkane dehydrogenation. The supported monometallic Pd catalysts for the dehydrogenation of light alkanes have the same problems mentioned for Pt catalysts. Compared with Pt, Pd metal shows higher intrinsic activity towards C–C bond cracking, which results in poor selectivity to target olefins but high yield of methane. In addition, monometallic Pd catalysts rapidly deactivate due to coke formation. To overcome these shortcomings, recent studies focus on modifying Pd catalysts with the addition of promoters including Ga, Zn, In, Fe, Sn and Sb. The addition of these inert second metals can significantly enhance the selectivity as well as stability. Analogous to Pt-based bimetallic catalysts, the geometric and electronic effects of these promoters on Pd catalysts are responsible for the enhanced catalytic performances.

Arteaga et al. tested a series of Ga-modified Pd/Al₂O₃ catalysts for the dehydrogenation of n-butane.¹²⁰ They have shown that the selectively to butene and catalyst stability are improved significantly in the presence of Ga. Low molar ratios of Ga/Pd (<1.68) are critical to achieve these enhancements. However, the addition of Ga reduced the initial activity. In contrast, Wasserscheid et al. observed that the intrinsic activity can be remarkably increased (4-fold) at high molar atomic ratios of Ga/Pd (>10).¹²¹ Coke formation can be minimized. They suggested that the superior catalytic performance resulted from the unique liquid metallic nature on the support, where the atomically isolated Pd active sites were dynamically formed.

Miller and his co-workers demonstrated a series of supported Pd-based intermetallic alloy catalysts including Pd–Zn,¹²² Pd–In,¹²³ Pd–Fe,¹²⁴ Pd–Sn (ref. 125) and Pd–Sb¹²⁵ for ethane and propane dehydrogenation. All of these catalysts showed higher activity for dehydrogenation, selectivity to olefins and stability than Pd monometallic catalysts. They suggested that the geometrically isolated Pd catalytic sites neighboring the inactive metallic promoters on these Pd-based intermetallic compounds are responsible for the high olefin selectivity, while the weaker adsorbate bond energies and lower reactant surface coverages due to the decrement in energy of the filled Pd 4d orbital lead to the increase in intrinsic activity and stability.^122–125

Although Pd-based bimetallic catalytic systems showed some promising performance in the dehydrogenation of light alkanes, the activity is still lower than that over Pt-based bimetallic catalysts. In addition, the price of Pd is much higher than that of Pt,¹²⁶ making it unappealing in large-scale applications.

3.3 Other bimetallic catalysts

Apart from Pt-based and Pd-based catalysts for the selective dehydrogenation, several recent studies report that effective catalysts composed of some other metals (e.g. Rh, Ni) as active sites were reported. The Ga promoter was often added to these catalyst systems to form either a single-atom diluted alloy or intermetallic compounds. These specific catalyst structures enable unusual metal sites active in C–H bond dissociation.

Wasserscheid et al. developed supported catalytically active liquid metal solution (SCALMS) catalysts Ga–Rh/Al₂O₃ (Ga/Rh >80), which possess dynamic single Rh atoms at the surface during reaction, for propane dehydrogenation.^127,128 Due to the unique structure during reaction, these Ga–Rh/Al₂O₃ SCALMS showed high catalytic activity, high propylene selectivity (∼92%) and moderate catalyst stability.

To lower the cost of the noble-metal-based catalysts, several Ni-based bimetallic catalysts have been developed. Ni-based catalysts possess appealing ability in the activation of alkene molecules. However, unmodified Ni-based catalysts exhibit a high intrinsic activity towards C–C hydrogenolysis, leading to methane, coke and hydrogen. This is due to the ensembles of Ni metal atoms on the catalyst surface. To improve olefin selectivity, several metal promoters such as Au,¹²⁹ Sn (ref. 130) and Ga^131,132 have been introduced to dilute the large nickel ensembles and thereby suppress the hydrogenolysis. When a Ni/SiO₂ catalyst is incorporated with Au for propane dehydrogenation, the selectivity to propene remarkably increased from 0% to 46%, while methane selectivity dramatically decreased from 99.9% to 27%.¹²⁹ The addition of Sn and Ga is more effective. Li et al. prepared a NiSn/SiO₂ catalyst for isobutane dehydrogenation.¹³⁰ Characterization indicated that the formation of a Ni_2.67Sn₂ alloy significantly breaks the large Ni ensembles. In addition, DFT calculations confirmed the electron transfer from Sn to Ni. The electron-rich Ni surface facilitates the desorption of iso-butene and thus inhibits undesired side reactions. Combined with geometric and electronic effects, 90.2% selectivity to iso-butene and high stability can be achieved on this catalyst. The Laursen group developed several non-noble Ni–Ga-based intermetallic compound catalysts for the dehydrogenation of ethane¹³¹ and propane.¹³² The catalysts have a core–shell structure, where the Ni₃Ga intermetallic phase is surrounded by a shell with varied Ni and Ga compositions. The Ni atoms on the surface are requisite to drive the reaction while the Ga at the surface sustains the selectivity and catalyst stability. The optimal surface composition is critically important for achieving high performance. Conversion decreased and selectivity increased as Ga loading was increased from 3 : 1 to 1 : 2 Ni : Ga molar ratio (Fig. 1a and b). DFT calculations revealed that unselective dehydrogenation reactions, which lead to coke, C–C cracking and methane formation, are significantly inhibited with increasing amount of Ga. The catalyst with a 1 : 1 Ni : Ga ratio ((1 : 1 Ni : Ga)@/Ni₃Ga/Al₂O₃) presented the best balance between activity and selectivity in both dehydrogenation of propane and ethane. In a propane dehydrogenation reaction prolonged to 82 h, this catalyst showed reasonably stable activity (∼13% to 9%) and a high steady-state selectivity towards olefins (propene and ethylene). The selectivity to propene decreased moderately from ∼94% to 81%, while the selectivity to ethylene increased slightly (Fig. 1c). The catalyst can be regenerated by in situ O₂ and H₂ treatment. Compared with commercial and in-house produced Pt–Sn catalysts, this catalyst showed comparable or higher selectivity and stability.

Fig. 1 (a and b) Catalytic performance of propane dehydrogenation over Ni₃Ga/Al₂O₃ with different Ni : Ga actual loadings. (c) Catalytic activity, 82 h stability, and regenerability of (1 : 1 Ni : Ga)@/Ni₃Ga/Al₂O₃. Reproduced from ref. 125 with permission from the American Chemical Society, copyright 2018.

4. The functions of the promoters

To explain the modification of the catalytic properties by addition of the promoter in bimetallic catalysts, geometric effects and electronic effects have been discussed. It is very difficult to identify the contribution of each in the dehydrogenation reactions. Usually, the combination of both effects leads to significant changes in the catalytic properties.

4.1 Geometric effect

The introduction of the promoters to the supported metal-based catalysts significantly influence the active metal structures through alloy formation and surface modification. The accompanying geometric effect leads to remarkable changes in the catalytic properties. Two kinds of explanations have been suggested to understand the geometric effect of the promoters. Both of these two geometric effects are able to suppress the hydrogenolysis and deep dehydrogenation reactions, thus reducing the production of shorter hydrocarbons and inhibiting coke formation.

It is believed that only a small ensemble of active metal sites or even a single metal atom could catalyze the structure-insensitive dehydrogenation reactions. The structure-sensitive side reactions such as hydrogenolysis, deep dehydrogenation and coke formation can only occur on the large ensembles of the active metal. The formation of an alloy and/or the partial covering of the active metal particles by promoter species break the large ensembles of active metal and create small active metal surface ensembles or even isolated atoms. This ensemble effect favors the main alkane dehydrogenation and inhibits side reactions. Indeed, the deeper dehydrogenated intermediates require more active metal sites to bind due to the highly unsaturated nature of their C atoms. DFT studies also revealed that intermediates such as CHCHCH₃ and CH₃C binding on the PtGa₂(111)⁸² and PtSn(111)¹³³ without three-fold Pt hollow sites is much weaker than those on Pt(111) with three-fold Pt sites. Hence, the selectivity to the alkene is enhanced and coking is suppressed.

It is also suggested that side reactions are facile on the low-coordinated sites, such as edge sites and corner sites. These defect sites can be selectively covered by the addition of the promoters through alloy formation and surface modification, which can be confirmed by FT-IR spectroscopy with CO adsorption.^25,39 The absence of low-coordinated defect metal sites contributes to the suppression of side reactions. In addition, DFT calculations by Honkala et al.¹³⁴ revealed that the lack of highly active Pt step sites on Pt₃Sn(211) weakens propene adsorption and suppresses further dehydrogenation of propene, thus improving catalyst selectivity and stability.

4.2 Electronic effect

Modification of the electronic properties of the active metal is the other role of the promoters. The electronic effect on the active metal is still under debate. Two different electronic effects have been proposed, which are electron donation from the promoter to the active metal and energy change of the Pt 5d orbitals or the filled Pd 4d orbitals.

The electron transfer from the promoter to the 5d band of the Pt atoms is the most commonly accepted electronic promotion in Pt-based bimetallic catalysts. The electron transfer results in an increase in the electron density of Pt and can be identified by FT-IR spectroscopy with CO adsorption,^39,48 XAS,^90,112 XPS⁷⁵ and DFT.¹³³ In the case of FT-IR spectroscopy with CO adsorption, the absorption bands for CO linearly adsorbed onto Pt-based bimetallic catalysts, such as Pt₃Sn,³⁹ PtSn³⁹ and PtZn (ref. 75) alloys, present at lower wavenumbers than that on monometallic Pt terraces. The redshifted CO vibration frequency demonstrates the electron donation from the promoter to Pt, which is due to a higher electron density available for back-donation to the 2π* orbitals in the CO bonding to Pt. Deng et al. analyzed the electronic property of Pt by calculating Pt d-band vacancy based on white line intensity in Pt L_III and L_II XANES spectra. The lower the d-band vacancy, the higher the electron density.³⁹ They revealed that the addition of Sn makes Pt electron-rich and the Pt₃Sn alloy showed the highest electron density. The change in the electronic state of Pt modified with the promoters affects the adsorption fashion of intermediates and products. DFT calculations revealed that the presence of the promoters weakened the binding energies of alkene and coke precursors.¹³³ The electron-enriched Pt surface thereby facilitates alkene formation and coke precursor desorption, concomitantly suppressing deep dehydrogenation, hydrogenolysis and coke formation. Consequently, the electron-enriched Pt surfaces improve the selectivity to target olefins and enhance the catalytic stability against coke deposits dramatically.

In contrast with electron donation, Miller and co-workers argued that the electronic effect of the promoter on the active metal is the energy change of the Pt 5d orbitals or the Pd 4d orbitals. They studied the electronic structure of Pt in Pt₁Zn₁,⁷⁶ PtFe (ref. 105) and PtIn (Pt₃In and PtIn₂)⁹² intermetallic compounds by using resonant inelastic X-ray scattering, Pt L_III XANES and DFT calculations. They found that the energy of the filled Pt valence 5d orbitals decreased and the energy of the unfilled Pt 5d orbitals increased upon bonding with Zn, Fe or In. The lowered energy of the filled Pt 5d valence orbitals results in the decrease of Pt adsorbate bond energy, leading to faster alkene desorption and higher intrinsic reaction activity. Analogously, the addition of Zn,¹²² In,¹²³ Fe (ref. 124) or Sb (ref. 125) decreases the energy of the filled Pd 4d valence orbitals in these Pd-based intermetallic compounds. This electronic promotion effect on these alloys is expected to weaken the Pd–C and Pd–H interactions and to enhance the catalytic activity but not the selectivity.

5. Reaction mechanism

The reverse Horiuti–Polanyi mechanism is the most commonly used mechanism to explain the dehydrogenation of alkanes over metal-based catalysts.¹³⁵ The mechanism consists of four main steps (Scheme 1): (i) cleavage of the first C–H bond in a dissociative adsorption of the alkane, resulting in an adsorbed alkyl group; (ii) cleavage of the second C–H bond, resulting in a π- or di-σ-bonded olefin; (iii) desorption of the olefin molecule; and (iv) combinative desorption of the hydrogen molecule. The mechanism follows the Langmuir–Hinshelwood kinetics, where all the surface sites of the catalyst are considered identical. Either the dissociative adsorption of an alkane molecule^136,137 or the second C–H cleavage step^44,82,99 has been suggested as the rate-limiting step of the dehydrogenation reaction.

Scheme 1 Proposed propane dehydrogenation mechanism.

Several studies have been performed on Pt-based catalysts to gain a detailed understanding of the mechanism of the dehydrogenation and the effect of the promoter. Koel and co-workers studied alkene adsorption modes and desorption manner on Pt(111) and Pt–Sn surface alloys.^138,139 They observed that ethylene binds in a di-σ ethylene binding geometry preferentially on Pt(111) and all the Pt–Sn alloy surface. Hook et al. also found that the di-σ ethylene binding fashion was preferred for both extended Pt and Pt–Sn surfaces by using DFT methods.¹³³ As Sn content increased, the desorption temperature for the alkene decreased substantially, which facilitates alkene desorption. Gong and his co-workers revealed that the di-σ-bound propene adsorbed on the Pt(111) and Pt₃Cu(111) surface while only π-bound propene existed on the Pt/Cu SAA surface based on DFT calculations.⁹⁹ The binding strength of π-bound propene is weaker than that of the di-σ adsorbed ones, suggesting facile propylene desorption from Pt/Cu SAA. Furukawa et al. observed that PtGa–Pt₁(111) with a single Pt atom on the surface binds propene in a π fashion as well.⁸⁸ The alkene adsorption energy is below the barrier for dehydrogenation of alkene with the aid of promoters, making further dehydrogenation and cracking reactions less likely. In the cases of supported nanoparticles, the binding mode of alkene is totally different from the observation on extended metal surfaces. Dumesic et al. studied the adsorption geometries and energies of ethylene to silica-supported Pt and PtSn catalysts using microcalorimetry and IR spectroscopy at low temperatures.¹⁴⁰ During the adsorption of ethylene, di-σ- and π-bound ethylene and ethylidyne are detected over both of the catalysts. The incorporation of Sn resulted in an increase in the fraction of π-bound ethylene and a decrease in the heat of adsorption. Similarly, Al₂O₃-supported Pt nanoparticles also resulted in three ethylene adsorption modes at 180 K.¹⁴¹ The di-σ-bound ethylene was converted to ethylidyne and hydrogen while all π-bound ethylene was found to desorb intact at room temperature. Anderson et al. observed similar adsorption and transformation modes of ethylene over the Pt₄/SiO₂ catalyst.¹⁴² However, they found that Pt₄Sn₃/SiO₂ binds ethylene only in the π mode. Ha and Baxter et al. examined changes in ethylene binding/desorption over Pt_n/Al₂O₃ and Pt_nB_x/Al₂O₃ and found that the addition of boron suppressed di-σ-binding of ethylene and substantially lowered the desorption energy.^117,118 Note that the di-σ-bonded alkene, adsorbed on the Pt surface strongly, is suspected to be an intermediate in further dehydrogenation while the π-bonded one is ready to desorb¹⁴² (Scheme 1). Moreover, ethylidyne is suggested to be an intermediate in the ethylene cracking to methane and coke during ethane dehydrogenation based on isotopic labeling experiments (Scheme 1).^21,22 Although the adsorption mode of alkene over Pt-based catalysts is still under debate, one thing is apparent: that the addition of promoter always weakens the strength of alkene adsorption and favors desorption of alkene over further dehydrogenation, leading to high selectivity towards olefins and low carbon deposition. This is rationalized by a combination of geometric and electronic effects.¹⁴² On the other hand, DFT calculations showed that the addition of promoter increases the barrier for dissociative adsorption of light alkanes and lowers the rate of dehydrogenation of light alkanes.^44,82,88,104 Therefore, the effect of the promoter on the overall reaction activity depends on the trade-off between the rates of dissociative adsorption of alkane and desorption of formed alkene.

6. Recent advanced strategies for the development of bimetallic catalysts

6.1 Novel support materials

Good support materials for light alkane dehydrogenation should have a high surface area, high thermal stability, and excellent mechanical strength. The properties of the supports affects the dispersion, stability and structure of bimetallic catalysts significantly. Thus, modulating the support is an alternative strategy to enhance catalytic performances. Besides alkalized θ-Al₂O₃, ZnAl₂O₄ and ZrO₂ used in industrial and pilot scale, a number of novel supports including Al₂O₃ nanosheets, hydrotalcites, CeO₂, zeolites and carbon materials have been developed and engineered, aiming to improve the dispersion and anchor the active sites.

Alumina is a typical support for commercial Pt–Sn bimetallic catalysts. To improve metal dispersion and the metal–support interaction, Lu et al. synthesized γ-Al₂O₃ nanosheets rich in pentacoordinate Al³⁺ ions as a support for Pt–Sn bimetallic catalysts in propane direct dehydrogenation.⁴⁸ The coordinatively unsaturated Al atoms act as anchoring sites to disperse and immobilize raft-like Pt–Sn clusters with mean diameters of 1.3 nm (Fig. 2A and B). The electronic density of Pt sites was increased with the addition of Sn and the electronic interaction between Pt and Sn on this unique nanosheet support was enhanced, leading to a lower desorption temperature of propene than that on conventional γ-Al₂O₃ supports (Fig. 2C). In addition, this unique nanosheet structure facilitates mass transfer and improves diffusion kinetics. Consequently, the catalyst displays a high productivity of propene with a selectivity of >99% and a low deactivation rate of 0.007 h⁻¹ at 590 °C, which are much better than those over traditional Pt–Sn/γ-Al₂O₃ catalysts (Fig. 2D). To further improve the stability of Pt–Sn bimetallic catalysts, a Mg(Sn)(Al)O@Al₂O₃ support was used for Pt–Sn catalysts.⁴⁶ The surface basicity of hydrotalcite is expected to suppress coke formation. The hydrotalcite support interacts strongly with Sn and stabilizes it in an oxidized state. The confined SnO_x in the Mg(Al)O lattice can provide the anchoring sites for immobilizing highly dispersed Pt clusters by strong interactions and thus afford a catalyst with a lower deactivation rate of 0.0045 h⁻¹ at 600 °C.

Fig. 2 (A) TEM image and (B) HAADF-STEM image of PtSn/Al₂O₃ sheet. (C) Propylene-TPD profiles of PtSn/Al₂O₃ ref. and PtSn/Al₂O₃ sheet. (D) Specific rate of C₃H₆ formation versus WHSV for PtSn/Al₂O₃ ref. and PtSn/Al₂O₃ sheet. Reproduced from ref. 41 with permission from John Wiley and Sons, copyright 2015.

CeO₂ as an alternative support for Pt–Sn catalysts was explored.⁶⁶ The Pt–Sn/CeO₂ catalysts showed a reversible alloy/segregation process in reaction/oxidation cycling. The CeO₂ support can provide the trapping sites for atomically dispersing and re-dispersing Pt while Sn can act as a nucleation site for forming active Pt–Sn clusters during reaction. Additionally, the redox properties of CeO₂ minimize coke formation by a co-feeding stream. In consequence, Pt–Sn/CeO₂ showed remarkable stability and regenerability, alleviating the requirement of complex oxychlorination regeneration as practiced in industry.

Nanocarbon materials are appealing supports for metal catalysts, as they can provide abundant surface sites to host metal atoms. Liu et al. used a nanocarbon (ND@G) composed of a nanodiamond core and a defective, ultrathin graphene nanoshell as a support to prepare a Pt–Sn/ND@G catalyst for n-butane dehydrogenation.⁴⁵ Due to the strong interaction with the support (Pt–C bonding) and the partition of monodispersed Sn species, the catalyst contains atomically dispersed Pt clusters with three Pt atoms on average (Fig. 3A). In contrast, without Sn, both Pt clusters and crystalline Pt nanoparticles coexisted on Pt/ND@G.³¹ In n-butane dehydrogenation, the bimetallic catalysts showed higher reactivity and selectivity to C₄ olefins (>98%) than Pt/ND@G and Pt₃Sn/Al₂O₃, which is due to the fully exposed Pt atom and easy desorption of butene (Fig. 2B). However, coke formation is still unavoidable in this dehydrogenation reaction, and the regeneration of carbon-supported catalysts is a critical issue in practical applications.

Fig. 3 (A) HAADF-STEM image of the Pt–Sn/ND@G catalyst; the atomic dispersion of Pt and Sn, scale bar = 0.5 nm. (B) Direct dehydrogenation of n-butane over different catalysts. Reproduced from ref. 45 with permission from the American Chemical Society, copyright 2019.

Zeolites and mesoporous materials containing promoter atoms in the framework have been studied as supports in recent years. The promoter atoms in the framework provide anchoring sites to disperse and immobilize Pt. In addition, some of the promoter atoms in the framework might migrate and alloy with Pt during calcination and reduction processes. Yuan and coworkers used Sn–Si-Beta as a support to prepare a highly efficient Pt/Sn–Si-Beta catalyst for propane dehydrogenation (Scheme 2).⁶² The isolated Sn sites in the zeolite framework could interact strongly with Pt and stabilize ultra-small Pt nanoclusters. As a result, the catalyst gave a high propene formation rate and good stability. Jing et al. employed Sn-incorporated SBA-15 mesoporous materials as supports for Pt/Sn-SBA-15 in propane dehydrogenation.⁵⁷ The catalyst activity and selectivity to propene were enhanced due to the promoted efficiency of Sn in the support. Recently, Zhao et al. synthesized a highly dispersed MnO_x decorated dendritic mesoporous silica support by a simple in situ emulsion method.¹⁰⁴ The support stabilizes the high dispersion of Pt clusters through the strong electrostatic interaction between the support and the Pt precursor. A Pt–Mn alloy was formed after reduction and gave excellent catalytic activity and stability owing to its balanced abilities of C–H activation and propene desorption.

Scheme 2 Illustration of the preparation of the Pt/Sn–Si-Beta catalyst. Reproduced from ref. 55 with permission from the Royal Society of Chemistry, copyright 2019.

6.2 Novel preparation methods

The catalyst structure, dispersion and formation of the active phase significantly depend on the catalyst preparation and pretreatment methods. The impregnation method is the most commonly used approach for the preparation of bimetallic catalysts. However, with this method, it is very difficult to obtain homogeneously distributed nanoparticles with ultra-high dispersion and to result in satisfactory catalytic performances. To address this, several new strategies for the synthesis of supported metal catalysts have been developed.

6.2.1 Surface organometallic chemistry (SOMC). In order to improve the metal dispersion and obtain homogeneously distributed bimetallic particles on the support, a new controllable approach on the basis of surface organometallic chemistry (SOMC) on the oxide support has been developed in recent years.^143,144 With this method, Pt and its promoter, such as Sn, Ga, Zn, organometallic complex are sequentially grafted on the surface of the oxide supports (Al₂O₃, MgAl₂O₄, SiO₂ and zeolites). Upon reduction with hydrogen, uniform ultra-small bimetallic clusters were formed, which showed excellent performance in propane direct dehydrogenation.

Basset and Zhu et al. prepared a Pt–Sn/MgAl₂O₄ catalyst with surface-enriched Sn by using K₂PtCl₄ and Bu₃SH as precursors and LiB(C₂H₅)₃H as a reduction agent.¹⁴⁵ The catalyst exhibited high selectivity and long-term stability in propane dehydrogenation to propylene due to the dilution effect of the high concentration of inactive Sn on the surface of Pt nanoparticles. After that, they synthesized Pt–Sn/θ-Al₂O₃ catalysts with ultra-small bimetallic Pt–Sn clusters of around 0.75 nm by using Pt(COD)Me₂ and HSnPh₃ as precursors.³⁶ The obtained catalysts showed high activity (TOF of 102 s⁻¹), high selectivity (99%) and excellent stability in propane dehydrogenation. To further improve the activity, a completely dealuminated Beta zeolite was used as a support for Pt–Sn bimetallic catalysts.⁵⁸ After SOMC and reduction with hydrogen, homogeneously dispersed Pt–Sn alloy clusters located at the Sn single site in the zeolite framework were formed. The catalyst exhibits slightly higher activity with a TOF of 114 s⁻¹ than Pt–Sn/θ-Al₂O₃. The strong interaction between Pt–Sn bimetallic clusters and isolated Sn sites in the zeolite framework Sn conferred high catalytic stability and good regenerability.

Copéret's group prepared a uniform nanometric Pt–Ga alloy supported on silica containing Ga single sites for PDH by the SOMC method (Fig. 4a).⁸⁷ The loading amount of Pt is up to 3 wt%, which is much higher than that of typical Pt-based bimetallic catalysts (∼0.5 wt%). XPS and in situ XAS revealed that the reduced catalysts possess Ga_xPt (0.5 < x < 0.9) alloy. The presence of Lewis acidic Ga single sites remaining at the surface could favor the nucleation and stabilization of nanometric particles, leading to highly dispersed nanometric alloyed particles with an average particle size of around 1.0 nm (Fig. 4b). Consequently, the catalyst exhibited a great propene productivity (1050 g_C3H6 g_cat⁻¹ h⁻¹) at high WHSV (98 h⁻¹). To further improve the productivity, a Pt⁰–Zn^δ+/SiO₂ catalyst with a similar structure to that of Pt⁰–Ga^δ+/SiO₂ has been developed by the same group recently.⁷⁷ The catalyst gave a super high productivity (up to 703 g_C3H6 g_Pt⁻¹ h⁻¹), which might be due to the smaller average Pt–Zn alloy particle size of around 0.8 nm. However, the stability of both catalysts is not good owing to coke formation as well as sintering. In addition, the catalytic activity cannot be fully recovered (ca. 75% of the initial activity) after oxidation/reduction regeneration, which might be ascribed to the structural change during the regeneration process. In contrast, the Pt–Sn/Beta can be fully restored after regeneration, which might be associated with robust Sn single sites in the zeolite framework.

Fig. 4 (a) Synthesis of the bimetallic Pt⁰–Ga^δ+/SiO₂ catalyst. (b) Representative TEM image and particle size distribution of Pt⁰–Ga^δ+/SiO₂ catalyst. Reproduced from ref. 80 with permission from the American Chemical Society, copyright 2018.

6.2.2 Encapsulation of metal in zeolites. As stated before, dehydrogenation reaction requires a high temperature (500–700 °C) to reach a reasonable conversion level. The metal-based catalysts thus need to be stable against sintering. The encapsulation strategy is an efficient method to protect metal nanoparticles from agglomeration and sintering through spatial confinement. Zeolites with ordered microporous structures and excellent thermal stability have been proved to be ideal hosts for accommodating small metal nanoparticles. Research on such a spatial confinement in zeolites for light alkane dehydrogenation has grown exceptionally over the past few years.

An in situ ligand-stabilized method has been proven to be an effective method to encapsulate metal species in pure siliceous zeolite. Ethylenediamine is often used as a ligand to protect the metal species from premature colloidal precipitation during zeolite crystallization. Using this method, Corma and Liu et al. have synthesized K–PtSn@Silicate-1 catalysts for propane dehydrogenation.¹⁴⁶ The loading amount of Pt is about 0.4 wt%. The introduction of K⁺ could compensate for the silanol groups and further stabilize the subnanometer Pt species. High-resolution high-angle annular dark-field scanning transmission electron microscopy (HR HAADF-STEM) and integrated differential phase contrast (iDPC) imaging techniques evidenced that subnanometric PtSn clusters encapsulated in sinusoidal channels. With the confinement of zeolites, the ultra-small size of PtSn clusters and the electronic interaction between Pt and Sn, the pretreated K–PtSn@Silicate-1 showed exceptional activity, selectivity and stability. Catalyst pretreatment with hydrogen at a long reduction time (>12 h), allowing Sn to incorporate with Pt, is essential to achieve high stability.¹⁴⁷ To further improve the propylene yield, they prepared another K–PtSn@Silicate-1 with a high loading of Pt (∼1.4 wt%).¹⁴⁸ They found that high-temperature calcination in air after hydrothermal synthesis is critical to disintegrate the Pt particles on the external surface into subnanometric Pt species and stabilize Pt in the sinusoidal channels, while high-temperature reduction treatment at 650 °C favors the migration of Sn species from the surface region to the internal region. The resultant high dense subnanometric bimetallic PtSn clusters in the sinusoidal 10 MR channels showed excellent catalytic performance for propane dehydrogenation, especially in terms of propene yield.

Similarly, Yu's group developed an ethylenediamine-protected direct hydrogen reduction method for encaging Pt–Zn bimetallic clusters into silicalite-1 zeolite (Fig. 5a).⁷⁴ Cs-corrected STEM-HAADF images and EXAFS results of PtZn@Silicate-1 confirmed that subnanometric Pt–Zn clusters are encapsulated in the five/six-membered rings (MRs) of zeolites (Fig. 5b–g). They claimed that direct hydrogen reduction after hydrothermal synthesis is essential to obtain subnanometric Pt–Zn clusters. At almost the same time, Yuan et al. also reported a Pt–Zn@Silicate-1 catalyst prepared by a similar ethylenediamine-protected method. Unlike Yu's method, Yuan et al. applied calcination followed by hydrogen reduction.⁷⁵ Ultra-small Pt–Zn clusters with an unknown particle size confined in silicate-1 could be obtained as well. In addition, the Zn species increase the electron density of Pt, as confirmed by the shift of the CO adsorption band, leading to facile desorption of propene on Pt–Zn clusters. Meanwhile, the small size of the Pt–Zn clusters could enhance the electron interaction between Pt and Zn. Consequently, combined with small Pt–Zn clusters caged in zeolites and facile desorption of propene, both of the catalysts showed superior performances in terms of activity, selectivity, stability and regenerability in propane dehydrogenation. Yu et al. indicated that the introduction of Cs⁺ ions located inside the zeolite channels is critical for maintaining the regeneration stability, as Cs⁺ ions could effectively suppress metal migrations to the outer surface of zeolite crystals and thereby inhibit the aggregation of the metal species. However, the addition of Cs⁺ was not needed in the catalyst made by Yuan et al. Remarkably, both subnanometric Pt–Sn and Pt–Zn bimetallic clusters (<1 nm) encapsulated in zeolites exhibit high activity at high weight hour space velocity (up to 108 h⁻¹).^74,147,148

Fig. 5 (a) Schematic representation of the synthesis of zeolite-encaged subnanometer Pt–Zn cluster catalysts. (b–d) HAADF-STEM images of PtZn4@S-1-H. (e) ABF-STEM image of PtZn4@S-1-H along the [010] projection with the model superimposed. (f and g) High-resolution HAADF-STEM images of PtZn4@S-1-H. Reproduced from ref. 67 with permission from John Wiley and Sons, copyright 2020.

To prepare the catalysts more efficiently, Okubo and his co-workers have recently developed an ultrafast in situ encapsulation method to prepare PtSn@ZSM-5.¹⁴⁹ They proposed this method based on ultrafast synthesis of zeolites developed previously. By passing a zeolite precursor solution after aging and seeding through a tubular reactor featuring fast heating, they are able to obtain crystallized zeolites in a few minutes. This ultrafast synthesis of zeolites could coordinate with the fast formation of metal nanoclusters to synthesize nanocluster-encapsulated zeolites. As a result, the ligand is not needed in this method. The synthesized PtSn@ZSM-5 catalysts showed excellent activity, selectivity and stability in the dehydrogenation of propane to propene.

6.3 Single-atom alloy catalysts for dehydrogenation

As discussed in the Geometric effect section, dehydrogenation reactions are structure–insensitive reactions, which means that even a single metal atom can drive the reaction. In contrast, undesired side reactions, such as hydrogenolysis and coke formation, are typically structure–sensitive reactions and require large ensembles of metal atoms. Based on these, single atom alloys (SAAs), composed of catalytically active noble metal atoms atomically dispersed in inert host metals, have been developed recently for light alkane dehydrogenation to eliminate the side reactions and improve the catalytic performance.

Marcinkowski et al. studied both a model Pt–Cu SAA surface and real particles prepared by the galvanic replacement method for n-butane dehydrogenation.¹⁰⁰ Gong et al. also prepared a Pt–Cu/Al₂O₃ SAA catalyst through a co-impregnation method for PDH (Fig. 6a–c).⁹⁹ They both found that Pt–Cu SAAs can activate C–H bonds efficiently, improve the selectivity to alkene and retard coke formation. They proposed that the single Pt atoms diluted by Cu can drastically facilitate the desorption of propylene and suppress its further dehydrogenation to form coke. Consequently, the activity of the Pt–Cu SAA catalyst can be maintained in 120 h time-on-stream duration at 520 °C (Fig. 6d). However, the copper particles sintered quickly at temperatures higher than 600 °C, resulting in quick deactivation. To further improve the stability, a PtGa–Pb/SiO₂ catalyst with a thermally stable intermetallic Pd–Ga alloy was prepared by Furukawa et al. via a simple impregnation method.⁸⁸ Intermetallic PtGa displays three-fold Pt ensembles and single Pt atoms isolated by Ga on the surface. The former is selectively blocked and disabled by Pb deposition, which enables single-atom Pt in intermetallic PtGa to catalyze the dehydrogenation. The single-atom-like PtGa–Pb/SiO₂ catalyst showed high selectivity (99.6%) towards propene and exceptional stability in a long-term operation at high temperature (600 °C, 96 h). Conversely, the PtGa/SiO₂ without Pb modification showed lower selectivity and stability, suggesting the superior catalytic properties of such single-atom-like isolated Pt atoms in PtGa–Pb/SiO₂. The performance is almost identical after oxidation/reduction regeneration, which is due to thermally stable intermetallic Pt–Ga. DFT calculations revealed that the isolated Pt atoms catalyze the dissociation of the first and the second C–H bonds, while they effectively inhibit the third one, which minimizes coke formation and thus drastically improves the selectivity and stability.

Fig. 6 (a) HAADF-STEM images of a typical region of the reduced 0.1Pt10Cu/Al₂O₃ catalyst, showing Pt atoms individually dispersed on Cu(111). Pt atoms are highlighted by red arrows. (b) The enlarged image and (c) the colored intensity map from the selected region in (a). d) Long-term stability test for 0.1Pt10Cu/Al₂O₃ at 520 °C for 120 h. Reproduced from ref. 92 with permission from Nature, copyright 2018.

Wasserscheid and co-workers first developed supported catalytically active liquid metal solution (SCALMS) catalysts Pt–Ga/Al₂O₃,⁸⁹ Pd–Ga/glass,¹²¹ and Rh–Ga/Al₂O₃ (ref. 127 and 128) for propane and n-butane dehydrogenation. Unlike previous PtGa intermetallic compounds, these SCALMS are rich in Ga and possess lower melting points. They exhibit a liquid metal phase on the support and possess dynamic single Pt or Rh atoms at the surface under the dehydrogenation reaction. The porous support is important for dispersing the gallium droplets and preventing agglomeration. All these catalysts showed high activity, selectivity to alkenes and resistance to coke formation. The noble metal and Ga catalyze dehydrogenation synergistically. The isolated noble metal atoms activate alkane and the alkyl migrates to the Ga surface. Then, two hydrides recombined to H₂ at the single noble metal site. In addition, Ga promotes the diffusion of the target alkene and thus results in high selectivity and catalytic stability.

7. Conclusion and perspectives

In this critical review, we presented the progress on metal-based catalysts for the non-oxidative dehydrogenation of light alkanes to corresponding olefins. These olefins can be used as chemical building blocks for the production of polymers, oxygenates and important chemical intermediates. We summarized and exemplified the metal catalysts including Pt-, Pd-, Rh- and Ni-based catalysts, especially in combination with the promoters (e.g., Sn, Zn, Ga, In, B), in the direct dehydrogenation of light alkanes. Among them, Pt-based catalysts are the most widely studied. The introduction of the promoters can remarkably improve their catalytic performances, especially in terms of selectivity to target olefins and catalytic stability. Next, we discussed the roles of the promoters, including geometric and electronic effects, in the modification of catalytic properties. The enhanced catalytic performances always rise from the combination of both effects. A brief reaction mechanism was also presented. Lastly, we highlighted the most recent advances in the development of novel high-performance dehydrogenation catalysts through three primary approaches: support engineering, advanced preparation methods and catalyst structure innovation.

Although improved catalytic performances in the direct dehydrogenation of light alkanes have been obtained over metal-based catalysts in the past decades, significant efforts are still needed to develop more efficient novel catalysts and understand the mechanisms. Based on the current state of the art in the development of metal-based catalyst, especially Pt-based catalysts, it can be found that subnanometric metal catalysts and single-atom catalysts exhibit promising performance in the direct dehydrogenation of light alkanes. However, the loading amount of active metals is quite low (∼0.5 wt%), resulting in low catalytic productivity. In this regard, high-density subnanometric metal catalysts and single-atom catalysts are expected to improve the catalytic productivity. Although subnanometric metal catalysts (such as Pt–Zn/SiO₂ (ref. 77) and Pt–Ga/SiO₂ (ref. 87)) with high loading (3 wt%) have been prepared by SOMC methods and demonstrate significantly high propene productivity, the stability of these catalysts is still unsatisfactory at high reaction temperatures. Thus, developing novel economically viable synthesis methods, such as encapsulation of active sites in porous shells, to obtain stable subnanometric metal catalysts and single-atom catalysts with high metal loading is encouraged for practical applications. In addition, to lower the cost of expensive noble-metal dehydrogenation catalysts, it is urgent to develop many other efficient and durable catalysts using non-noble metals. In fact, a Ni–Ga-based intermetallic compound catalyst has been prepared by Laursen's group and showed comparable or higher selectivity and stability than commercial and in-house prepared Pt–Sn catalysts.^131,132 Selecting the optimal combination of the promoters and non-noble metals and carefully controlling the catalyst structure could be the key criteria for the development of non-noble metal-based dehydrogenation catalysts. The reaction mechanisms and the roles of the promoters are still under debate. Therefore, future studies with the aid of advanced in situ and operando spectroscopy techniques and theoretical calculations are necessary to understand the catalytically active sites, the reaction mechanisms and the roles of the promoters. These will give insights into the rational design of highly efficient dehydrogenation catalysts.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Financial support by the National Natural Science Foundation of China (21522605 and 21776210) and the Tianjin Natural Science Foundation (18JCJQJC46800) is gratefully acknowledged.

Notes and references

1. J. C. Bricker, Top. Catal., 2012, 55, 1309–1314.
2. C. A. Carrero, R. Schloegl, I. E. Wachs and R. Schomaecker, ACS Catal., 2014, 4, 3357–3380.
3. F. Cavani, N. Ballarini and A. Cericola, Catal. Today, 2007, 127, 113–131.
4. J. T. Grant, J. M. Venegas, W. P. McDermott and I. Hermans, Chem. Rev., 2018, 118, 2769–2815.
5. Z.-P. Hu, D. Yang, Z. Wang and Z.-Y. Yuan, Chin. J. Catal., 2019, 40, 1233–1254.
6. Z. Nawaz, Rev. Chem. Eng., 2015, 31, 413–436.
7. J. J. Sattler, J. Ruiz-Martinez, E. Santillan-Jimenez and B. M. Weckhuysen, Chem. Rev., 2014, 114, 10613–10653.
8. G. Wang, X. Zhu and C. Li, Chem. Rec., 2020, 20, 604–616.
9. https://www.thebusinessresearchcompany.com/report/ethylene-global-market-report, 2020.
10. https://www.thebusinessresearchcompany.com/report/propylene-global-market-report, 2020.
11. https://www.globalinforesearch.com/Global-1-Butene_p482674.html, 2020.
12. A. Farshi, F. Shaiyegh, S. H. Burogerdi and A. Dehgan, Pet. Sci. Technol., 2011, 29, 875–885.
13. E. McFarland, Science, 2012, 338, 340–342.
14. S. Chen, C. Pei, G. Sun, Z.-J. Zhao and J. Gong, Acc. Mater. Res., 2020, 1(1), 30–40.
15. M. M. Bhasin, J. H. McCain, B. V. Vora, T. Imai and P. R. Pujadó, Appl. Catal., A, 2001, 221, 397–419.
16. M. A. Atanga, F. Rezaei, A. Jawad, M. Fitch and A. A. Rownaghi, Appl. Catal., B, 2018, 220, 429–445.
17. E. Gomez, B. Yan, S. Kattel and J. G. Chen, Nat. Rev. Chem., 2019, 3, 638–649.
18. E. Gianotti, M. Taillades-Jacquin, J. Rozière and D. J. Jones, ACS Catal., 2018, 8, 4660–4680.
19. R. B. Biniwale, S. Rayalu, S. Devotta and M. Ichikawa, Int. J. Hydrogen Energy, 2008, 33, 360–365.
20. O. O. James, S. Mandal, N. Alele, B. Chowdhury and S. Maity, Fuel Process. Technol., 2016, 149, 239–255.
21. V. Galvita, G. Siddiqi, P. Sun and A. T. Bell, J. Catal., 2010, 271, 209–219.
22. G. Siddiqi, P. Sun, V. Galvita and A. T. Bell, J. Catal., 2010, 274, 200–206.
23. S. Saerens, M. K. Sabbe, V. V. Galvita, E. A. Redekop, M.-F. Reyniers and G. B. Marin, ACS Catal., 2017, 7, 7495–7508.
24. H. Saito and Y. Sekine, RSC Adv., 2020, 10, 21427–21453.
25. A. Virnovskaia, S. Morandi, E. Rytter, G. Ghiotti and U. Olsbye, J. Phys. Chem. C, 2007, 111, 14732–14742.
26. M. Santhosh Kumar, D. Chen, J. C. Walmsley and A. Holmen, Catal. Commun., 2008, 9, 747–750.
27. J. Zhu, M.-L. Yang, Y. Yu, Y.-A. Zhu, Z.-J. Sui, X.-G. Zhou, A. Holmen and D. Chen, ACS Catal., 2015, 5, 6310–6319.
28. J. Wu, Z. Peng and A. T. Bell, J. Catal., 2014, 311, 161–168.
29. F. Jiang, L. Zeng, S. Li, G. Liu, S. Wang and J. Gong, ACS Catal., 2014, 5, 438–447.
30. M. Santhosh Kumar, A. Holmen and D. Chen, Microporous Mesoporous Mater., 2009, 126, 152–158.
31. J. Liu, Y. Yue, H. Liu, Z. Da, C. Liu, A. Ma, J. Rong, D. Su, X. Bao and H. Zheng, ACS Catal., 2017, 7, 3349–3355.
32. A. Volynkin, M. Rønning and E. A. Blekkan, Top. Catal., 2015, 58, 854–865.
33. J. Liu, J. Li, J. Rong, C. Liu, Z. Dai, J. Bao, Z. Da and H. Zheng, Appl. Surf. Sci., 2019, 464, 146–152.
34. H. N. Pham, J. J. H. B. Sattler, B. M. Weckhuysen and A. K. Datye, ACS Catal., 2016, 6, 2257–2264.
35. N. Prakash, M.-H. Lee, S. Yoon and K.-D. Jung, Catal. Today, 2017, 293–294, 33–41.
36. Z. Xu, R. Xu, Y. Yue, P. Yuan, X. Bao, E. Abou-Hamad, J.-M. Basset and H. Zhu, J. Catal., 2019, 374, 391–400.
37. L. Deng, H. Miura, T. Shishido, S. Hosokawa, K. Teramura and T. Tanaka, ChemCatChem, 2014, 6, 2680–2691.
38. L. Deng, T. Shishido, K. Teramura and T. Tanaka, Catal. Today, 2014, 232, 33–39.
39. L. Deng, H. Miura, T. Shishido, Z. Wang, S. Hosokawa, K. Teramura and T. Tanaka, J. Catal., 2018, 365, 277–291.
40. L. Deng, H. Miura, T. Ohkubo, T. Shishido, Z. Wang, S. Hosokawa, K. Teramura and T. Tanaka, Catal. Sci. Technol., 2019, 9, 947–956.
41. Y. Zhang, Y. Zhou, J. Shi, S. Zhou, X. Sheng, Z. Zhang and S. Xiang, J. Mol. Catal. A: Chem., 2014, 381, 138–147.
42. N. Kaylor and R. J. Davis, J. Catal., 2018, 367, 181–193.
43. M.-H. Lee, B. M. Nagaraja, K. Y. Lee and K.-D. Jung, Catal. Today, 2014, 232, 53–62.
44. M.-L. Yang, Y.-A. Zhu, X.-G. Zhou, Z.-J. Sui and D. Chen, ACS Catal., 2012, 2, 1247–1258.
45. J. Zhang, Y. Deng, X. Cai, Y. Chen, M. Peng, Z. Jia, Z. Jiang, P. Ren, S. Yao, J. Xie, D. Xiao, X. Wen, N. Wang, H. Liu and D. Ma, ACS Catal., 2019, 9, 5998–6005.
46. Y. Zhu, Z. An, H. Song, X. Xiang, W. Yan and J. He, ACS Catal., 2017, 7, 6973–6978.
47. H. Liu, X. Dong, J. Xia, X. Li, Y. Sun and Q. Cai, Trans. Tianjin Univ., 2020, 26, 362–372.
48. L. Shi, G. M. Deng, W. C. Li, S. Miao, Q. N. Wang, W. P. Zhang and A. H. Lu, Angew. Chem., Int. Ed., 2015, 54, 13994–13998.
49. E. J. Jang, J. Lee, H. Y. Jeong and J. H. Kwak, Appl. Catal., A, 2019, 572, 1–8.
50. B. M. Nagaraja, H. Jung, D. R. Yang and K.-D. Jung, Catal. Today, 2014, 232, 40–52.
51. Y. Zhang, Y. Zhou, J. Shi, S. Zhou, Z. Zhang, S. Zhang and M. Guo, Fuel Process. Technol., 2013, 111, 94–104.
52. Y. Shi, X. Li, X. Rong, B. Gu, H. Wei and C. Sun, RSC Adv., 2017, 7, 19841–19848.
53. B. K. Vu, M. B. Song, I. Y. Ahn, Y.-W. Suh, D. J. Suh, W.-I. L. Kim, H.-L. Koh, Y. G. Choi and E. W. Shin, Catal. Today, 2011, 164, 214–220.
54. Q. Zhang, K. Zhang, S. Zhang, Q. Liu, L. Chen, X. Li, C. Wang and L. Ma, J. Catal., 2018, 368, 79–88.
55. T. Srisakwattana, K. Suriye, P. Praserthdam and J. Panpranot, Catal. Today, 2019 DOI:10.1016/j.cattod.2019.06.058.
56. G. Aguilar-Ríos, M. Valenzuela, P. Salas, H. Armendáriz, P. Bosch, G. Del Toro, R. Silva, V. Bertín, S. Castillo, A. Ramírez-Solís and I. Schifter, Appl. Catal., A, 1995, 127, 65–75.
57. B. Li, Z. Xu, W. Chu, S. Luo and F. Jing, Chin. J. Catal., 2017, 38, 726–735.
58. Z. Xu, Y. Yue, X. Bao, Z. Xie and H. Zhu, ACS Catal., 2019, 10, 818–828.
59. Y. Zhang, Y. Zhou, L. Huang, M. Xue and S. Zhang, Ind. Eng. Chem. Res., 2011, 50, 7896–7902.
60. Z. Nawaz and F. Wei, J. Ind. Eng. Chem., 2010, 16, 774–784.
61. J. Li, J. Li, Z. Zhao, X. Fan, J. Liu, Y. Wei, A. Duan, Z. Xie and Q. Liu, J. Catal., 2017, 352, 361–370.
62. Y. Wang, Z.-P. Hu, W. Tian, L. Gao, Z. Wang and Z.-Y. Yuan, Catal. Sci. Technol., 2019, 9, 6993–7002.
63. Y. Zhang, M. Xue, Y. Zhou, H. Zhang, W. Wang, Q. Wang and X. Sheng, RSC Adv., 2016, 6, 29410–29422.
64. H. Zhou, J. Gong, B. Xu, S. Deng, Y. Ding, L. Yu and Y. Fan, Chin. J. Catal., 2017, 38, 529–536.
65. H. Zhou, J. Gong, B. Xu, L. Yu and Y. Fan, Appl. Catal., A, 2016, 527, 30–35.
66. H. Xiong, S. Lin, J. Goetze, P. Pletcher, H. Guo, L. Kovarik, K. Artyushkova, B. M. Weckhuysen and A. K. Datye, Angew. Chem., Int. Ed., 2017, 56, 8986–8991.
67. S. A. I. Barri and R. Tahir, US5208201, 1993.
68. P. Andy and M. E. Davis, Ind. Eng. Chem. Res., 2004, 43, 2922–2928.
69. C. Chen, M. Sun, Z. Hu, J. Ren, S. Zhang and Z.-Y. Yuan, Catal. Sci. Technol., 2019, 9, 1979–1988.
70. Y. Zhang, Y. Zhou, L. Huang, S. Zhou, X. Sheng, Q. Wang and C. Zhang, Chem. Eng. Sci., 2015, 270, 352–361.
71. P. L. De Cola, R. Gläser and J. Weitkamp, Appl. Catal., A, 2006, 306, 85–97.
72. I. Kley and Y. Traa, Microporous Mesoporous Mater., 2012, 164, 145–147.
73. B. D. Alexander, J. George and A. Huff, US5453558, 1995.
74. Q. Sun, N. Wang, Q. Fan, L. Zeng, A. Mayoral, S. Miao, R. Yang, Z. Jiang, W. Zhou, J. Zhang, T. Zhang, J. Xu, P. Zhang, J. Cheng, D. C. Yang, R. Jia, L. Li, Q. Zhang, Y. Wang, O. Terasaki and J. Yu, Angew. Chem., Int. Ed., 2020, 59(44), 19450–19459.
75. Y. Wang, Z.-P. Hu, X. Lv, L. Chen and Z.-Y. Yuan, J. Catal., 2020, 385, 61–69.
76. V. J. Cybulskis, B. C. Bukowski, H.-T. Tseng, J. R. Gallagher, Z. Wu, E. Wegener, A. J. Kropf, B. Ravel, F. H. Ribeiro, J. Greeley and J. T. Miller, ACS Catal., 2017, 7, 4173–4181.
77. L. Rochlitz, K. Searles, J. Alfke, D. Zemlyanov, O. V. Safonova and C. Copéret, Chem. Sci., 2020, 11, 1549–1555.
78. J. Camacho-Bunquin, M. S. Ferrandon, H. Sohn, A. J. Kropf, C. Yang, J. Wen, R. A. Hackler, C. Liu, G. Celik, C. L. Marshall, P. C. Stair and M. Delferro, ACS Catal., 2018, 8, 10058–10063.
79. O. B. Belskaya, L. N. Stepanova, T. I. Gulyaeva, S. B. Erenburg, S. V. Trubina, K. Kvashnina, A. I. Nizovskii, A. V. Kalinkin, V. I. Zaikovskii, V. I. Bukhtiyarov and V. A. Likholobov, J. Catal., 2016, 341, 13–23.
80. Z. Yu, J. A. Sawada, W. An and S. M. Kuznicki, AIChE J., 2015, 61, 4367–4376.
81. E. L. Jablonski, A. A. Castro, O. A. Scelza and S. R. de Miguel, Appl. Catal., A, 1999, 183, 189–198.
82. T. Wang, F. Jiang, G. Liu, L. Zeng, Z.-J. Zhao and J. Gong, AIChE J., 2016, 62, 4365–4376.
83. M. Filez, E. A. Redekop, V. V. Galvita, H. Poelman, M. Meledina, S. Turner, G. Van Tendeloo, A. T. Bell and G. B. Marin, Phys. Chem. Chem. Phys., 2016, 18, 3234–3243.
84. P. Sun, G. Siddiqi, M. Chi and A. T. Bell, J. Catal., 2010, 274, 192–199.
85. M. Filez, E. A. Redekop, H. Poelman, V. V. Galvita, R. K. Ramachandran, J. Dendooven, C. Detavernier and G. B. Marin, Chem. Mater., 2014, 26, 5936–5949.
86. E. A. Redekop, V. V. Galvita, H. Poelman, V. Bliznuk, C. Detavernier and G. B. Marin, ACS Catal., 2014, 4, 1812–1824.
87. K. Searles, K. W. Chan, J. A. Mendes Burak, D. Zemlyanov, O. Safonova and C. Copéret, J. Am. Chem. Soc., 2018, 140, 11674–11679.
88. Y. Nakaya, J. Hirayama, S. Yamazoe, K. I. Shimizu and S. Furukawa, Nat. Commun., 2020, 11, 2838.
89. T. Bauer, S. Maisel, D. Blaumeiser, J. Vecchietti, N. Taccardi, P. Wasserscheid, A. Bonivardi, A. Gorling and J. Libuda, ACS Catal., 2019, 9, 2842–2853.
90. P. Sun, G. Siddiqi, W. C. Vining, M. Chi and A. T. Bell, J. Catal., 2011, 282, 165–174.
91. M. C. Román-Martínez, J. A. Maciá-Agulló, I. M. J. Vilella, D. Cazorla-Amorós and H. Yamashita, J. Phys. Chem. C, 2007, 111, 4710–4716.
92. E. C. Wegener, Z. Wu, H.-T. Tseng, J. R. Gallagher, Y. Ren, R. E. Diaz, F. H. Ribeiro and J. T. Miller, Catal. Today, 2018, 299, 146–153.
93. S. Zha, G. Sun, T. Wu, J. Zhao, Z. J. Zhao and J. Gong, Chem. Sci., 2018, 9, 3925–3931.
94. J. Wu, Z. Peng, P. Sun and A. T. Bell, Appl. Catal., A, 2014, 470, 208–214.
95. K. Xia, W.-Z. Lang, P.-P. Li, L.-L. Long, X. Yan and Y.-J. Guo, Chem. Eng. Sci., 2016, 284, 1068–1079.
96. L.-L. Shen, K. Xia, W.-Z. Lang, L.-F. Chu, X. Yan and Y.-J. Guo, Chem. Eng. Sci., 2017, 324, 336–346.
97. K. Xia, W.-Z. Lang, P.-P. Li, X. Yan and Y.-J. Guo, RSC Adv., 2015, 5, 64689–64695.
98. G.-Q. Ren, G.-X. Pei, Y.-J. Ren, K.-P. Liu, Z.-Q. Chen, J.-Y. Yang, Y. Su, X.-Y. Liu, W.-Z. Li and T. Zhang, J. Catal., 2018, 366, 115–126.
99. G. Sun, Z. J. Zhao, R. Mu, S. Zha, L. Li, S. Chen, K. Zang, J. Luo, Z. Li, S. C. Purdy, A. J. Kropf, J. T. Miller, L. Zeng and J. Gong, Nat. Commun., 2018, 9, 4454.
100. M. D. Marcinkowski, M. T. Darby, J. Liu, J. M. Wimble, F. R. Lucci, S. Lee, A. Michaelides, M. Flytzani-Stephanopoulos, M. Stamatakis and E. C. H. Sykes, Nat. Chem., 2018, 10, 325–332.
101. Z. Han, S. Li, F. Jiang, T. Wang, X. Ma and J. Gong, Nanoscale, 2014, 6, 10000–10008.
102. S. Veldurthi, C.-H. Shin, O.-S. Joo and K.-D. Jung, Catal. Today, 2012, 185, 88–93.
103. Z. Wu, B. C. Bukowski, Z. Li, C. Milligan, L. Zhou, T. Ma, Y. Wu, Y. Ren, F. H. Ribeiro, W. N. Delgass, J. Greeley, G. Zhang and J. T. Miller, J. Am. Chem. Soc., 2018, 140, 14870–14877.
104. X. Fan, D. Liu, X. Sun, X. Yu, D. Li, Y. Yang, H. Liu, J. Diao, Z. Xie, L. Kong, X. Xiao and Z. Zhao, J. Catal., 2020, 389, 450–460.
105. E. C. Wegener, B. C. Bukowski, D. Yang, Z. Wu, A. J. Kropf, W. N. Delgass, J. Greeley, G. Zhang and J. T. Miller, ChemCatChem, 2020, 12, 1325–1333.
106. W. Cai, R. Mu, S. Zha, G. Sun, S. Chen, Z.-J. Zhao, H. Li, H. Tian, Y. Tang, F. Tao, L. Zeng and J. Gong, Sci. Adv., 2018, 4, eaar5418.
107. L. G. Cesar, C. Yang, Z. Lu, Y. Ren, G. Zhang and J. T. Miller, ACS Catal., 2019, 9, 5231–5244.
108. X. Yang, G. Liu, Y. Li, L. Zhang, X. Wang and Y. Liu, Trans. Tianjin Univ., 2019, 25, 245–257.
109. S. Rimaz, L. Chen, S. Kawi and A. Borgna, Appl. Catal., A, 2019, 588, 117266.
110. E. Jimenez-Izal, J.-Y. Liu and A. N. Alexandrova, J. Catal., 2019, 374, 93–100.
111. N. J. LiBretto, C. Yang, Y. Ren, G. Zhang and J. T. Miller, Chem. Mater., 2019, 31, 1597–1609.
112. J. Zhu Chen, Z. Wu, X. Zhang, S. Choi, Y. Xiao, A. Varma, W. Liu, G. Zhang and J. T. Miller, Catal. Sci. Technol., 2019, 9, 1349–1356.
113. C. Ye, Z. Wu, W. Liu, Y. Ren, G. Zhang and J. T. Miller, Chem. Mater., 2018, 30, 4503–4507.
114. R. Ryoo, J. Kim, C. Jo, S. W. Han, J. C. Kim, H. Park, J. Han, H. S. Shin and J. W. Shin, Nature, 2020, 585, 221–224.
115. M. Aly, E. L. Fornero, A. R. Leon-Garzon, V. V. Galvita and M. Saeys, ACS Catal., 2020, 10, 5208–5216.
116. C. Byron, S. Bai, G. Celik, M. S. Ferrandon, C. Liu, C. Ni, A. Mehdad, M. Delferro, R. F. Lobo and A. V. Teplyakov, ACS Catal., 2019, 10, 1500–1510.
117. E. T. Baxter, M.-A. Ha, A. C. Cass, H. Zhai, A. N. Alexandrova and S. L. Anderson, J. Phys. Chem. C, 2018, 122, 1631–1644.
118. M. A. Ha, E. T. Baxter, A. C. Cass, S. L. Anderson and A. N. Alexandrova, J. Am. Chem. Soc., 2017, 139, 11568–11575.
119. J. Dadras, E. Jimenez-Izal and A. N. Alexandrova, ACS Catal., 2015, 5, 5719–5727.
120. L. Rodríguez, D. Romero, D. Rodríguez, J. Sánchez, F. Domínguez and G. Arteaga, Appl. Catal., A, 2010, 373, 66–70.
121. N. Taccardi, M. Grabau, J. Debuschewitz, M. Distaso, M. Brandl, R. Hock, F. Maier, C. Papp, J. Erhard, C. Neiss, W. Peukert, A. Gorling, H. P. Steinruck and P. Wasserscheid, Nat. Chem., 2017, 9, 862–867.
122. J. R. Gallagher, D. J. Childers, H. Zhao, R. E. Winans, R. J. Meyer and J. T. Miller, Phys. Chem. Chem. Phys., 2015, 17, 28144–28153.
123. Z. Wu, E. C. Wegener, H.-T. Tseng, J. R. Gallagher, J. W. Harris, R. E. Diaz, Y. Ren, F. H. Ribeiro and J. T. Miller, Catal. Sci. Technol., 2016, 6, 6965–6976.
124. C. Yang, Z. Wu, G. Zhang, H. Sheng, J. Tian, Z. Duan, H. Sohn, A. J. Kropf, T. Wu, T. R. Krause and J. T. Miller, Catal. Today, 2019, 323, 123–128.
125. G. Zhang, C. Ye, W. Liu, X. Zhang, D. Su, X. Yang, J. Z. Chen, Z. Wu and J. T. Miller, Nano Lett., 2019, 19, 4380–4383.
126. https://apps.catalysts.basf.com/apps/eibprices/mp/defaultmain.aspx.
127. M. Wolf, N. Raman, N. Taccardi, M. Haumann and P. Wasserscheid, ChemCatChem, 2020, 12, 1085–1094.
128. N. Raman, S. Maisel, M. Grabau, N. Taccardi, J. Debuschewitz, M. Wolf, H. Wittkamper, T. Bauer, M. Wu, M. Haumann, C. Papp, A. Gorling, E. Spiecker, J. Libuda, H. P. Steinruck and P. Wasserscheid, ACS Catal., 2019, 9, 9499–9507.
129. Z. Yan and D. W. Goodman, Catal. Lett., 2012, 142, 517–520.
130. G. Wang, H. Wang, H. Zhang, Q. Zhu, C. Li and H. Shan, ChemCatChem, 2016, 8, 3137–3145.
131. Y. He, Y. Song and S. Laursen, ACS Catal., 2019, 9, 10464–10468.
132. Y. He, Y. Song, D. A. Cullen and S. Laursen, J. Am. Chem. Soc., 2018, 140, 14010–14014.
133. A. Hook, J. D. Massa and F. E. Celik, J. Phys. Chem. C, 2016, 120, 27307–27318.
134. L. Nykänen and K. Honkala, ACS Catal., 2013, 3, 3026–3030.
135. I. Horiuti and M. Polanyi, Trans. Faraday Soc., 1934, 30, 1164–1172.
136. R. D. Cortright, P. E. Levin and J. A. Dumesic, Ind. Eng. Chem. Res., 1998, 37, 1717–1723.
137. J. M. Hill, R. D. Cortright and J. A. Dumesic, Appl. Catal., A, 1998, 168, 9–21.
138. M. T. Paffett, S. C. Gebhard, R. G. Windham and B. E. Koel, Surf. Sci., 1989, 223, 449–464.
139. Y.-L. Tsai, C. Xu and B. E. Koel, Surf. Sci., 1997, 385, 37–59.
140. J. Shen, J. M. Hill, R. M. Watwe, B. E. Spiewak and J. A. Dumesic, J. Phys. Chem. B, 1999, 103, 3923–3934.
141. S. B. Mohsin, M. Trenary and H. J. Robota, J. Phys. Chem., 1988, 92, 5229–5233.
142. T. J. Gorey, B. Zandkarimi, G. Li, E. T. Baxter, A. N. Alexandrova and S. L. Anderson, ACS Catal., 2020, 10, 4543–4558.
143. C. Coperet, Nat. Energy, 2019, 4, 1018–1024.
144. M. K. Samantaray, V. D'Elia, E. Pump, L. Falivene, M. Harb, S. Ould Chikh, L. Cavallo and J.-M. Basset, Chem. Rev., 2020, 120, 734–813.
145. H. Zhu, D. H. Anjum, Q. Wang, E. Abou-Hamad, L. Emsley, H. Dong, P. Laveille, L. Li, A. K. Samal and J.-M. Basset, J. Catal., 2014, 320, 52–62.
146. L. Liu, M. Lopez-Haro, C. W. Lopes, C. Li, P. Concepcion, L. Simonelli, J. J. Calvino and A. Corma, Nat. Mater., 2019, 18, 866–873.
147. L. Liu, M. Lopez-Haro, C. W. Lopes, S. Rojas-Buzo, P. Concepcion, R. Manzorro, L. Simonelli, A. Sattler, P. Serna, J. J. Calvino and A. Corma, Nat. Catal., 2020, 3, 628–638.
148. L. Liu, M. Lopez-Haro, C. W. Lopes, D. M. Meira, P. Concepcion, J. J. Calvino and A. Corma, J. Catal., 2020, 391, 11–24.
149. J. Zhu, R. Osuga, R. Ishikawa, N. Shibata, Y. Ikuhara, J. N. Kondo, M. Ogura, J. Yu, T. Wakihara, Z. Liu and T. Okubo, Angew. Chem., Int. Ed., 2020, 59(44), 19669–19674.

---

Footnote

† These authors contributed equally to this work.

---