Research Progress of Catalysts and Initiators for Promoting the Cracking of Endothermic Hydrocarbon Fuels

Catalytic/initiated cracking of endothermic hydrocarbon fuels is an effective technology for cooling a hypersonic aircraft with a high Mach number (over 5). Catalysts and initiators can promote fuel cracking at low temperatures, increase fuel conversion and the heat sink capacity, and suppress coke deposition, thereby reducing waste heat. Catalysts mainly include metal oxide catalysts, noble metal catalysts and metal nanoparticles, zeolite catalysts, nanozeolite catalysts, and coating catalysts. Moreover, initiators roughly include nitrogenous compounds, oxygenated compounds, and hyperbranched polymer initiators. In this review, we aim to summarize the catalysts and initiators for cracking endothermic hydrocarbon fuels and their mechanisms for promoting cracking. This review will facilitate the development of the synthesis and exploration of catalysts and initiators.


Introduction
Endothermic hydrocarbon fuel [1][2][3][4], a type of functional fuel designed to solve the thermal barrier problem caused by increasing flight speed, can act as a propellant and coolant. Kerosene-type jet fuel with a carbon number of C 12 -C 16 [5] is the main component of endothermic hydrocarbon fuel, and its typical representatives include Jet A-1, JP-5, JP-7, JP-8, JP-9, RP-3, etc. The typical fuel contains approximately 40-50% n-/iso-alkanes and 20-40% naphthenes [6]. In addition, compared with the above fuels, JP-10 is a high-energy endothermic hydrocarbon fuel with a relatively low carbon number, and its composition is simply exo-tetrahydrodicyclopentadiene. Endothermic hydrocarbon fuel flows through the outer wall of the engine and absorbs heat through conduction, phase transitions, and cracking reactions in advanced cooling technology [7][8][9][10]. It has great potential to solve the thermal barrier problem of high-speed aircraft. Currently, endothermic hydrocarbon fuels are developing toward a high density, low freezing point, and high heat sink capacity, aiming to balance fuel grade and service performance [11][12][13][14].
Endothermic hydrocarbon fuel cracking occurs when the fuel critical temperature and pressure are reached. The fuel molecules will decompose into short-chain or even gaseous hydrocarbons for combustion in the engine, and a small part will be converted to aromatic hydrocarbons and polymeric substances as precursors of carbon deposition [15][16][17]. The degree of cracking will depend on the heat absorption capacity of the fuel, the deposition condition of the pipeline, and the ignition and combustion efficiency of the engine. Heat sink capacity, cracking efficiency, and product distribution and deposition are the main factors for evaluating a cracking reaction, considering the cooling and combustion functions of endothermic hydrocarbon fuels. Without interventions, thermal cracking does not react completely at relatively low temperatures below 430 °C [6,18,19] and is accompanied by a series of problems, such as carbon deposition and a limited chemical heat sink. Concise and effective measures must Yiyao Liu and Ran Chen have contribute equally to this work.
* Jie Liu jie.liu@tju.edu.cn be taken to achieve efficient cracking with milder reaction conditions, a higher heat sink capacity, more product distribution of gaseous hydrocarbons, and fewer carbon deposits. Adding catalysts or initiators is a simple and effective method for improving the cracking efficiency of endothermic hydrocarbon fuels. Catalysts and initiators activate the fuel molecules or the initiator itself to generate different effective intermediate-carbenium ions or radicals, respectively. With the addition of catalysts, the intermediate steps, such as C-C bond scission, hydrogen extraction, and protonation in catalytic cracking, are affected by catalysis following the carbenium ion mechanism [20][21][22]. With the addition of initiators, the initiation and propagation steps of thermal cracking are promoted by the free radicals generated from the homolytic cleavage of the initiators following the radical mechanism [22,23]. Herein, we will review the catalysts and initiators for promoting cracking of endothermic fuels by different categories of active components. The various types and mechanisms of catalysts (including metal and metal oxide catalysts and zeolite molecular sieve catalysts) and initiators (including amine and nitrogenous compounds, oxygenated compounds, and hyperbranched polymer initiators) are summarized class-by-class to clarify the screening and design ideas of catalysts and initiators.

Catalysts for Cracking Endothermic Hydrocarbon Fuels
Catalysis can moderate the reaction conditions, improve the reaction conversion rate, and control the product selectivity, as in endothermic hydrocarbon fuel cracking. Fuel cracking mainly includes two types of reactions: decomposition and dehydrogenation. In the petrochemical industry, these reactions are catalyzed by metals and zeolite catalysts of great reference. Numerous studies have screened and designed fuel cracking catalysts with high catalytic activity, high selectivity, and high stability.  [24][25][26][27], show bifunctional catalytic properties with dehydrogenation because of noble metals and pyrolysis. Additionally, metal oxide catalysts can be coated on the wall of stainless-steel microchannels for the catalytic cracking of fuels [25,26].

Metal Oxide Catalysts
Jiao et al. [25] investigated the influence of a range of Ptloaded ZrO 2 -TiO 2 -Al 2 O 3 (ZTA) catalysts (denoted as Cat1 through Cat5) on the catalytic cracking of RP-3 jet fuel from 550 to 750 °C at 3.5 MPa. The results showed that the gas yield of RP-3 was substantially increased by 240% over Cat3 at 550 °C, and the heat sink capacity increased by 0.42 MJ/ kg (17.9%) over Cat3 at 650 °C, compared with thermal cracking. Moreover, Jiao et al. [28] designed Pt-loaded MgAl 2 O 4 catalysts as shown in Fig. 1, and they found that the high heat sink (4.05 MJ/kg) and stability of Pt-loaded MgAl 2 O 4 considerably outperformed the common Pt-based catalysts. This work provided an effective way to improve catalytic activity and stability for hydrocarbon cracking.

Noble Metal Catalysts and Metal Nanoparticles
Noble metal catalysts mainly include platinum (Pt), palladium (Pd), ruthenium (Ru), rhenium (Re), iridium (Ir), and mixtures of noble metals [29]. In general, heterogeneous metal oxide catalysts, noble metal catalysts, and zeolite catalysts (in Sect. 2.2.1) are packed into a column or installed into the reactor, and catalytic cracking is carried out with these catalysts in heterogeneous phase systems, which have small contact areas and often need large amounts of catalysts. This type of packed catalytic cracking is not suitable for hypersonic aircraft. The limitations to coating catalysts include rapid deactivation, easy scratching, and high heat resistance. Now, developing nanometer catalysts is a novel strategy that may overcome the disadvantages of catalyst coatings. The nanoparticles with large specific surface areas and abundant active sites can disperse evenly in endothermic hydrocarbon fuels, obtaining a stable suspension that is considered a quasi-homogeneous system [30][31][32]. Quasihomogeneous catalytic cracking of hydrocarbon fuels is an effective way to improve the cooling capacity of future aircraft. The stability and dispersity of nanoparticles in fuels are crucial and can be achieved by modifying the surface of the nanoparticles with water-oil amphoteric reagents.
Noble metal nanoparticles are often used in pseudohomogeneous catalytic cracking. Zhang et al. [30] synthesized Pt and Pd nanoparticles (NPs) using oleylamine as the protecting ligand, and the NPs were highly dispersed in jet fuel JP-10, as seen in Fig. 2. Pt and Pd NPs (50 ppm) exhibited a markedly enhanced cracking performance, with cracking conversions and gas yields at 680 °C that were, respectively, 4.5-, 4.4-and 1.3-, 3.1-fold that of pure JP-10. In particular, Pt NPs decreased the initial temperature of cracking from 650 to 600 °C. This work demonstrated the potential of fuel-dispersible NPs in hypersonic applications.
Yue et al. [5] synthesized stable nanofluids containing decalin and palladium nanoparticles (Pd NPs), which were modified by octadecanethiol and/or octadecylamine, and they denoted the three types of Pd NPs as Pd@S, Pd@N, and Pd@S&N. In addition, the nanofluids could be considered pseudo-homogeneous phase systems. Under the supercritical condition of 750 °C, 3.5 MPa, and a mass flow rate of 1 g/s, the cracking of each nanofluid consisting of decalin and 500 ppm Pd NPs was performed. It was found that, compared with thermal cracking of decalin, the presence of Pd@S and Pd@S&N increased the cracking conversion of decalin by 2.90 wt% and 10.23 wt%. In addition, the presence of Pd@N increased the conversion by 21.54 wt% and the heat sink capacity by 0.29 MJ/kg, which showed the highest performance. In addition to amine and mercaptan, which can be used to prepare noble metal nanoparticles in pseudo-homogeneous systems, Fang et al. [31,33] has created a new strategy by preparing nanofluids consisting of noble metal nanoparticles capped by hyperbranched polymers and assembled three types of hyperbranched polymerencapsulated metal nanoparticles (HEMNs), Pd/Pt/Au@ CPAMAM (CPAMAM refers to a hydrocarbon-soluble amphiphilic hyperbranched polyamide), as catalysts for the supercritical cracking of decalin, as shown in Fig. 3. Pt@ CPAMAM (50 ppm) exhibited the best cracking behavior at 675 °C, with the conversion and heat sink values of decalin   [30]. Copyright © 2014 American Chemical Society increased by 28.4% and 0.44 MJ/kg, respectively. In addition, the experiments also proved that quasi-homogeneous catalytic cracking with HEMNs has a higher conversion, gas yield, and endothermic capacity than thermal cracking.
Transition metal nanoparticles and gold nanoparticles [34,35] can also be used for quasi-homogeneous catalytic cracking of endothermic hydrocarbon fuels. The transition metal Ni has excellent catalytic performance and a low price [36], but it can be easily reduced because of the coke formation at high temperatures. The Ni-B amorphous alloy has been widely used in the hydrogenation/ dehydrogenation process for better catalytic activity. Guo et al. [36] fabricated resorcinarene-encapsulated Ni-B nanoparticles, which can be stably and highly dispersed in JP-10. Compared with thermal pyrolysis, quasi-homogeneous catalytic pyrolysis of JP-10 with Ni-B nanoparticles showed remarkable improvement in conversion. Guo et al. [37] successfully prepared resorcinarene-encapsulated nickel nanoparticles with an average diameter of 35 nm, which can be well dispersed in JP-10. Compared with the thermal cracking of JP-10, the catalytic cracking of these nickel nanoparticles showed obviously enhanced conversion. Li et al. [35] successfully fabricated hydrophobic gold nanoparticles using butanediyl-1,4-bis(dimethylcetyl ammonium bromide) (16-4-16) and N,N-dimethylhexdecylamine (C16DMA) as the stabilizers and surfacemodified reagents. Then, the thermal cracking of JP-10 and the catalytic cracking of JP-10 on the obtained gold nanoparticles were studied, and the results indicated that these gold nanoparticles could increase the conversion of JP-10.
Different metal species will lead to different catalytic cracking effects. The first transition series metal elements, such as Fe and Ni, on the tube wall of the cracking reactors usually catalyze the dehydrogenation reaction to produce aromatic hydrocarbons, which are the precursors of coke and eventually lead to fibrous coke formation. However, noble metal catalysts, such as Pt and Pd, follow different catalytic mechanisms, inhibiting the aromatization of fuel molecules and promoting the deepening of cracking reactions. Taking Pt as an example, this behavior is usually due to the accelerated occurrence of hydrogenation reactions under noble metal catalyzation and the limitation of the carbocation reaction path [38], thus ultimately affecting the cracking efficiency and the product distribution.
Finally, the advantages and disadvantages of metal oxide catalysts, noble metal catalysts, and metal nanoparticle catalysts are summarized in Table 1.
Although zeolite catalysts show relatively high catalytic activity and selectivity but poor high-temperature stability, they deactivate quickly in reactions with hydrocarbons as a result of coke deposition, tend to collapse with a porous structure [43], and require regeneration and replacement. Furthermore, Süer et al. [43] found that compared with subcritical conditions (475 °C, 6.7/16.5 bar), commercial Y-type zeolite catalysts retained higher catalytic activity and longer life under supercritical conditions (475 °C, 41.6 bar), indicating that supercritical conditions benefit zeolite catalysts.
Cooper et al. [44] investigated the effects of HY, USY, and β zeolites on the catalytic cracking of JP-10 at 500 °C, 100 kPa, and a flow rate of 2.3 g/h, obtaining conversions of 34%, 11%, and 70%, respectively, compared with the thermal cracking conversion of 3.15%. However, during catalytic cracking, HY, USY, and β zeolites were observed with a rapid deactivation by coke deposits. Sicard et al. [45] chose powder Y and ZSM-5 zeolite catalysts for catalytic cracking of n-dodecane. They found that the catalytic cracking conversion of Y and ZSM-5 zeolites at 500 °C increased by 2% and 4%, respectively. Nevertheless, the conversion was higher with ZSM-5 than with Y, so ZSM-5 showed higher activity. Moreover, the products of thermal cracking were mainly n-alkanes (C 1 -C 9 ) and 1-alkenes (C 2 -C 10 ), while the products of catalytic cracking were mainly n-alkanes (C 1 -C 6 ) and alkenes (C 1 -C 6 ). At 425 °C, the difference in the conversion ratio and product distribution between catalytic cracking and thermal cracking is slight, which may be due to the rapid deactivation of the catalyst due to coking at this temperature.
To heighten the catalytic activity, stability, and light alkene selectivity, as well as the lifetime, zeolite catalysts can be modified by changing the Si/Al ratio, adjusting the pore size (introducing mesoporous structure [46]), introducing acidic or alkaline substances (such as metal ions, phosphorus [46,47], fluorine [48], gallium [49], and gadolinium [50]) to enhance acidity and other methods. This is because only their intensive, strong acid sites promote C-C cracking, not coke formation, and their pore size may cause diffusion resistance and coke deposition. Metal ion-containing zeolite catalysts were widely studied, and transition metal ions (such as Ag + [51], Cu 2+ , Co 2+ , Ni 2+ , and La 3+ ) can effectively regulate the acidity. Zhang et al. [52] found that the Ag-or La-modified USY and ZSM-5 catalysts brought higher selectivity of olefins, a longer catalyst life, and better stability than conventional USY and ZSM-5 catalysts. In addition, Shang et al. [48] enhanced the catalytic activity and endothermic capability of HZSM-5 by fluoride modification with NH 4 F and tested the effect of the fluoride-modified HZSM-5 catalyst on catalytic cracking of n-decane. Consequently, the conversion and heat sink of n-decane were apparently improved, and coke deposition was decreased. Furthermore, Long et al. [53] tried a novel method for modifying traditional catalysts by combining 2D transition metal oxide nanosheets with HZSM-5 catalysts. 2D Co 3 O 4 nanosheetwrapped HZSM-5 (Co 3 O 4 nanosheets@ZSM-5) was created, and its catalytic cracking of n-decane was estimated. This reaction with 0.1 wt% Co 3 O 4 nanosheets@ZSM-5 gave a heat sink as high as 4.64 MJ/kg at 758 °C, which is much higher than those of bare ZSM-5 (2.99 MJ/kg at 687 °C), and suppressed the coke deposition.
A mesoporous structure is essential for enhancing the diffusion rate of fuels and offers high and stable catalytic activity and a long lifetime, while desilication is a simple and useful approach to introducing mesopores [42,54]. Zhang et al. [54] prepared a series of hierarchical-structured HZSM-5 zeolite-processing mesopores (in Fig. 4) with desilication and then evaluated their catalytic performance for catalytic cracking of supercritical n-dodecane. They found that the conversion of n-dodecane with hierarchicalstructured HZSM-5 increased by 40% compared with the parent HZSM-5, while the deactivation rate was 1/7th that of the parent HZSM-5.

Nanozeolite Catalysts
As mentioned in Sect. 2.1.2, with increasing temperature, organic protectants can easily be removed from the metal surface of metal nanoparticles, so the catalytic activity of organic-encapsulated metal nanoparticles will be observed to gradually decrease because of rapid agglomeration. Second, the microchannel of the heat exchanger prevents selecting the conventional loading forms of zeolites (such as a fixed bed, fluidized bed, and wall-coated catalysts) [55] because of the limitations of deactivation, heat resistance, and flow resistance. Hence, the quasi-homogeneous catalytic cracking of endothermic hydrocarbon fuel with oil-soluble dispersed zeolites is considered.
Bao et al. [56] proposed an original pseudo-homogeneous catalysis method for hydrocarbon fuel with highly dispersed nano-HZSM-5 catalysts, as depicted in Fig. 5. They equipped nano-HZSM-5 zeolite with organosilane containing ethyltrichlorosilane, dodecyltrichlorosilane, and so on. Then, the catalytic cracking of n-dodecane was carried out with the surface-silanized nano-HZSM-5 catalyst, as mentioned above. This work revealed that surface silylation could increase the hydrophobicity of nano-HZSM-5 and protect against aggregation, giving rise to stable dispersions in n-dodecane. By analyzing the liquid and products, the presence of the nano-HZSM-5 catalyst substantially improved the conversion of n-dodecane. Sun et al. [57] prepared high hydrocarbon dispersible beta nanozeolites (HD-NZs) using silanizing seeds with phenylaminopropyltrimethoxysilane in the organic medium (HDZ-O), leading to good dispersibility in JP-10. The quasi-homogeneous catalytic cracking of JP-10 with HDZ-O at 700 °C and 4 MPa showed excellent performance such that the conversion of JP-10 was threefold higher than that of thermal cracking.
As catalysts for the catalytic cracking of JP-10, Tian et al. [58] applied synthesized HZSM-5 nanosheets with Si/Al molar ratios of 25 and 50 (ZNS-25 and ZNS-50, respectively) and thicknesses of approximately 2.0 nm. Consequently, the conversion of the catalytic cracking of JP-10 with HZSM-5 nanosheets was 19.79% higher than that of JP-10 with a conventional HZSM-5 catalyst. This result is due to HZSM-5 nanosheets having more Brønsted acid sites with better accessibility than HZSM-5 zeolites. Composite molecular sieves also can exhibit excellent catalytic cracking Fig. 4 Schematic of the preparation of hierarchical HZSM-5 coatings. Graphical abstract. Reproduced with permission from Ref. [54]. Copyright © 2013 Elsevier

Fig. 5
Schematic of the preparation of hydrocarbon dispersible nanozeolites. Reproduced with permission from Ref. [57]. Copyright © 2014 Elsevier activity. Sang et al. [41] prepared a series of HZSM-5/ MCM-41 composite molecular sieves (HZM-Ns(x)), using nano-ZSM-5 zeolites as the source. Catalytic cracking of n-decane with HZM-Ns was studied, and it was found that the HZSM-5 nanoparticles uniformly dispersed in the MCM-41 matrix provided suitable acidity and high specific surface areas, resulting in high catalytic activity.
Radical and carbenium mechanisms are commonly used to explain fuel cracking reactions. Thermal cracking is commonly demonstrated by a radical mechanism [23], in which radical formation in the initiation stage by homolytic cleavage of a C-H or a C-C bond is the regular process. During catalytic cracking, the chain reaction is driven by carbenium ion formation on the catalytic surface. The acid sites on the surface of zeolite catalysts are the key to affecting catalytic cracking in terms of forming carbenium ions and the direction of subsequent reactions. During the initiation stage, the acid sites on the zeolite surface may decide whether the carbocation generation mechanism is the bimolecular mechanism or the monomolecular mechanism [38]. The Brønsted acid sites usually traverse bimolecular activation pathways involving alkyl activation and hydride transfer and β-scission processes. In addition, the surface acid sites also affect the direction of subsequent reactions. Haag [59] demonstrated that alkanes can implement protonic dehydrogenation and cracking reactions on the zeolite Brønsted acid. Especially on superacid catalysts, carbenium ions have a strong tendency to undergo rearrangement reactions, thus rapidly decomposing/isomerizing to tertiary carbocations, such as (CH 3 ) 3 C + and (CH 3 ) 2 (C 2 H 5 )C + [60]. This behavior directly affects the bond-breaking position of β-scission, and the cracking reaction rate is accelerated compared with pyrolysis, which is more consistent with free radical theory.
In addition, the zeolite pore size will change the product distribution affected by the subsequent isomerization reaction to a certain extent. Alleviating the intracrystalline diffusion limitation but not the alteration of acid centers was argued to be the major factor in hydroisomerization activity enhancement.

Zeolite Coating Catalysts
Traditional heterogeneous zeolite catalysts are generally in the form of microspheres and stacked into reactors, which is difficult to achieve on aircraft. This difficulty is due to their filling patterns, which lead to higher mass and heat transfer resistance [61] during the thermal/catalytic cracking of fuels. To reduce the impact, catalysts are supported or coated on the inner surfaces of the pipe, forming a catalyst coating with good catalytic activity, but wall-coated catalysts should be applied in very thin layers (2-3 μm) that may make the design and construction of the reactor very complex. Commonly used coating catalysts include the HZSM-5 zeolite coating/membrane [54,62,63], HY zeolite coating [17], and SAPO-34 zeolite coating. For supported noble metal zeolite catalysts, most noble metal particles are hidden inside the coating such that the catalytic activity of a noble metal will be weakened. This problem is solved by electroless plating technology to incorporate metal into the zeolite coating. In addition, the catalytic activity of wall-coated catalysts will be affected by the coating thickness [64] and Si/Al ratio [63]; among them, the catalytic activity and stability of HZSM-5 coating increase with the Si/Al ratio. However, catalyst coatings also have many disadvantages, such as a complex preparation process in practical applications, quick deactivation, a reduction in heat transfer efficiency caused by carbon deposition, and the need for strong adhesion [65] required between catalysts and the surface.
Zhao et al. [66] prepared a Pd/HZSM-5 coating catalyst on the inner wall of a tubular reactor and simultaneously selected HZSM-5 powder for application to the supercritical cracking of n-dodecane. The results showed that the cracking rate of n-dodecane with the Pd/HZSM-5 coating at 640 °C and 4 MPa was 55.5%, and the hydrogen production rate was 3.1%, which compared with the thermal cracking or catalytic cracking with HZSM-5 powder increased by 17.3%, 78.1%, or 8.5%, 58.7%, respectively. The increase in hydrogen production was due to the dehydrogenation of Pd, and the dehydrogenation reaction caused by Pd was very useful for the endothermic reaction. Zhao et al. [61] discussed the catalytic performance of ZSM-5 membranes and conducted cracking experiments under supercritical conditions (450-500 °C, 4 MPa) in a fixed bed reactor using n-dodecane as the model fuel. The experiments showed that the conversion of catalytic cracking with ZSM-5 membranes was much higher than that of thermal cracking. Wu and Li [67] synthesized ZSM-5 crystals on the surface of a Ni-based alloy and applied them in the catalytic cracking of n-heptane. The catalyst was analyzed to be catalytically active for cracking at 700 °C, and the conversion of n-heptane increased by 4.6% with ZSM-5 crystals grown on the wall of the tubular reactor, compared with the bare tubular reactor.
However, the thermal conductivity of nonmetallic nanoparticles (nanozeolites in Sect. 2.2.2) is low, so the thermal conductivity of fuels cannot be substantially improved. Therefore, new types of nanofluid additives with high thermal conductivity and catalytic cracking performance must be sought [68].
Finally, the advantages and disadvantages of zeolite, nanozeolite, and zeolite coating catalysts are summarized in Table 2.

High Thermal Conductivity Nanofluids
Because of its high thermal conductivity and large specific surface area, graphene can be used to stabilize metal nanoparticles. Therefore, Liu et al. [68] developed the Pt@ FGS/JP-10 nanofluids (0.1 wt% for Pt) by loading Pt nanoparticles on functionalized graphene sheets (FGS) and grafted octadecylamine (ODA) to improve the dispersion of FGS in hydrocarbon fuels (Fig. 6). The catalytic cracking performance of Pt@FGS/JP-10 nanofluids was evaluated at 360-420 °C: The thermal conductivity of JP-10 with 0.2 wt% Pt@FGS increased by 5.2% compared with that of pure JP-10, and the conversion of JP-10 (400 °C, 5 h) increased by 150% for 0.5 wt% Pt@FGS.
At present, much work remains necessary to develop nonmetallic nanoparticles and metal nanoparticles.

Initiators for Cracking Endothermic Hydrocarbon Fuels
Several problems of heterogeneous and pseudo-homogeneous catalysts are still faced, such as insufficient utilization, quick deactivation by accumulating coke, and difficult regeneration. If fuel-soluble additives can act as catalysts, these issues may be solved. Thus, studies on initiating cracking have been conducted. Most of the initiators are stable [69] and green, so the fuels have excellent storage and transportation performance, but the physical properties of the fuels, like density and viscosity, may be altered because of a high added amount (more than 2 wt%) of homogeneous initiators, which may affect the use of fuel. Rapid deactivation, easy scratching Fig. 6 Schematic of the preparation of Pt@FGS/JP-10 nanofluid. Reproduced with permission from Ref. [68]. Copyright © 2019 Elsevier

Nitrogenous Compounds: Amine and Nitro Compounds
Amine initiators include triethylamine [70,71] (TEA), tributylamine, 1-nitropropane (NP), nitroethane, and nitromethane [72,73], which have C-N bonds. The bond dissociation energy of C-N bonds is lower than that of C-C bonds. Thus, homolytic fission will occur at C-N bonds, thereby releasing radicals and initiating cracking reactions. The molecular structures of triethylamine and 1-nitropropane are represented in Fig. 7. Guan et al. [72] illuminated the mechanism of initiating cracking of n-heptane with 1-nitropropane at T = 698 K. On the basis of the analysis, NP has two reaction paths at 698 K: the elimination of HONO and the rupture of the C-N bond to produce radicals. The radicals can capture an H atom from n-heptane to start the chain reaction and then accelerate the cracking of n-heptane. Wang et al. [71] discovered that triethylamine and tributylamine as initiators could effectively promote the cracking of heptane at high temperatures (550-650 °C). The results showed that the conversion of heptane with initiators was almost twice that of thermal cracking, and yields of ethylene and propylene were also improved at 550 °C. At 650 °C, the catalytic role of triethylamine was weakened. The conversion gradually increased as the mass fraction of TEA increased from 2 to 10%, but it decreased as this mass fraction continued to increase. Zhao et al. [74] investigated the effect of nitroethane as the initiator for pyrolysis of n-heptane using a continuous flow microreactor. The optimum addition of nitroethane was 2.0 wt%. Compared with thermal cracking, the conversion and yield of gas products increased. In addition, the overall yield of gas products and the yield of ethylene only increased by 5.67% and 2.41%, respectively. Liu et al. [75] chose 1-nitropropane, triethylamine, and 3,6,9-triethyl-3,6,9-trimethyl-1,4,7-triperoxonane (TEMPO) as initiative additives for the supercritical thermal cracking of a jet fuel model compound, n-dodecane (Dod). The results showed that initiators can substantially increase the conversion of Dod by 20-150%; when the added amount of initiators was within 2-4 wt%, the conversion of Dod reached its peak. Moreover, initiators had a slight effect on the selectivity of the gas products but promoted the yield of C 6 -C 11 n-alkanes in liquid products.

Oxygenated Compounds: Ethers, Alcohols, Esters, and Peroxides
The ether initiators mainly include diethyl ether (DEE) and methyl tert-butyl ether (MTBE), the alcohol initiators mainly include n-propanol and isopropanol (IPA), the ester initiators mainly include isopropyl nitrate (IPN) and isooctyl nitrate (ION), and the peroxide initiators mainly include tert-butyl hydroperoxide, di-tert-butyl peroxide [76], diacetonediperodixe (DADP) [77], and cumene peroxide. The diethyl ether has a weak C-O bond, and the alkyl group is directly linked with oxygen: these bonds are easy to break and produce alkyl radicals. The structure of MTBE contains a tert-butyl group branching three methyl groups, and the tert-butyl peroxides contain -O-O-bonds. The IPN and isooctyl nitrate contain -O-bonds and side chains that can generate alkyl radicals. The molecular structures of MTBE, IPN, IPA, and di-tert-butyl peroxide are pictured in Fig. 8.
Chakraborty and Kunzru [76] estimated the effectiveness of di-tert-butyl peroxide (DTBP), diisopropylamine (DIPA), and triethylamine as initiators for pyrolysis of n-heptane. They concluded that DTBP, DIPA, and TEA are effective initiators because of the initial release of radicals after weak C-N or O-O bonds are broken, and the order of initiation activity is TEA ≈ DIPA > DTBP. Wang et al. [77] investigated the influence of diethyl ether, MTBE, 1-nitropropane, 3,6,9-triethyl-3,6,9-trimethyl-1,2,4,5,7,8-hexaoxonane, triethylamine, and diacetonediperodixe for thermal cracking of n-decane by reactive molecular dynamics (MD) simulations. The results showed that DEE, MTBE, NP, TEMPO, Combining detection methods, such as infrared and isotope labeling, can better track the reaction of the initiator and reveal the cracking mechanism. As shown in Fig. 9, Mi et al. [78] monitored the effect of the polyester compound macromolecular initiator (MI) on the cracking process of JP-10 and confirmed that the polyester initiator mainly generates free radicals by alkyl-oxygen homolysis, which finally appeared to promote the H-abstraction reaction. Guo et al. [79] also studied the improvement of olefin selectivity in the cracking of naphtha using initiators. According to speculation, the initiator captures hydrogen atoms after generating free radicals and is more inclined to capture secondary hydrogen atoms than primary hydrogens to generate secondary carbon radicals, which eventually generate more olefins.

Hyperbranched Polymers Initiators
Fuel compatibility is an important factor restricting the practical application of various additives. Especially strong polar heteroatom compounds, such as amines and esters, are very easily polymerized in hydrocarbon fuels and even reduce fuel stability in severe cases. In addition, small-molecule initiators need to be added in larger amounts to ensure a detectable initiation effect. On the basis of the above considerations, while ensuring the hyperbranched structure and the presence of functional groups, such as amine and carboxyl groups, the hyperbranched compounds also introduce longchain alkyl groups such as n-dodecyl to ensure the solubility of additives in fuels (Fig. 10).
Metal nanoparticles or nanozeolite catalysts in quasihomogeneous catalytic cracking are prone to aggregate at high temperatures. Additionally, the physical properties of the fuel, like density and viscosity, may be altered by the high addition of common initiators, which may affect the use of fuel. Therefore, a macroinitiator [78,80,81] offers a promising method for obtaining high catalytic cracking activity of hydrocarbon fuels. He et al. [80] prepared a fuelsoluble macroinitiator, palmitoyl-hyperbranched polyglycerol (PHPG). Under supercritical conditions (3.5 MPa, 600-720 °C), PHPG could promote the cracking of n-tridecane and substantially improve the conversion and heat sink of n-tridecane. For example, at 690 °C, the conversion and heat sink of n-tridecane increased by 17.6% and 0.5 MJ/kg, respectively, compared with thermal cracking.
The functional groups of the hyperbranched initiators, such as amine and carboxyl groups, are still unit structures that easily generate free radicals. Hyperbranched compounds are characterized by their abundant terminal functional groups and oil solubility. Guo et al. [79] used the automatic reaction mechanism generator to simulate the function mechanism of the hyperbranched polyamide-amine initiators, inferring that the hyperbranched compounds can continuously release free radicals in low-temperature pyrolysis to induce the cracking reaction. Simultaneously, they can release many free radicals in a short time, which can locally increase the radical concentration sharply to substantially improve the cracking conversion rate, compared with smallmolecule initiators.
Typically, thermal cracking of hydrocarbon fuels is carried out through a chain reaction of free radicals from C-C bond cleavage, which is involved in the speed-control step of generating free radicals, and initiators enhance cracking reactions by generating free radicals more easily. A common initiator design strategy is to introduce chemical bonds (such as C-N, C-O, C-P, and O-O bonds) that are weaker than C-C or C-H bonds to rapidly generate many active free radicals through homolytic fission for attack, thus accelerating the initiation stage of the cracking reaction and promoting the cracking of hydrocarbon fuels. The BDE of these weak bonds is lower (less than 300 kJ/mol) than the BDE of C-C bonds in fuel molecules (greater than 350 kJ/mol) [76].
Finally, the advantages and disadvantages of heterogeneous and pseudo-homogeneous catalysts and homogeneous initiators are summarized in Table 3.

Conclusion
The increase in the speed of hypersonic vehicles has caused severe thermal barriers. The active cooling system using endothermic hydrocarbon fuels as a coolant can substantially reduce the heat load of aircraft through thermal, catalytic, and initiated cracking reactions. Catalysts and initiators used for catalytic cracking can promote fuel cracking at low temperatures and increase the heat sink capacity by improving the selectivity of hydrogen and light hydrocarbons, compared with thermal cracking.
The synthesis and application of traditional heterogeneous catalysts are mature, while several problems are still faced in use, such as high fuel flow resistance, low fuel dispersion, quick deactivation, and difficult regeneration of the catalyst caused by coking. In comparison with traditional catalysts, coating catalysts have unique advantages in enhancing mass and heat transfer and reducing resistance, while the thickness and uniformity are difficult to control. By modifying the traditional catalysts, metal nanoparticles and nanozeolite catalysts are designed and widely applied in catalytic cracking. Moreover, soluble initiators can play a similar role as heterogeneous catalysts without affecting the homogeneity of the fuel.
In a word, the above-mentioned studies of traditional catalysts, wall-coated catalysts, nano-catalysts, and initiators 1 3 Fig. 9 a 3D temperature-resolved IR spectra, b the total ion chromatogram of MIP pyrolysis products, and c the main pyrolysis process of MIP. Reproduced with permission from Ref. [78]. Copyright © 2020 Elsevier will help to improve the further study of the catalytic cracking of endothermic hydrocarbon fuels.

Prospects
Based on full-text analysis, this paper indicates the following prospects for the development directions of catalysts and initiators.
First, zeolite catalysts with high cracking activity and good thermal stability are appropriate for cracking. To date, elements of F, P, transition metal ions, etc., have been used to improve the structure and electronic properties of zeolite catalysts. Therefore, in the future, researchers can try to modify zeolite catalysts with two or more elements/ compounds as above or develop other new elements or compounds to improve the acidity and pore size of zeolites.
Second, developing a new type of nanofluid with excellent dispersibility, high-temperature thermal stability, and catalytic cracking performance remains challenging. In the future, studies can continue to explore new lipotropic organic compounds to stabilize nano-metal particles or nanozeolites.
Third, initiators are considered effective in promoting hydrocarbon cracking, but small-molecule initiators require large additive amounts of approximately 2 wt%. To reduce additive amounts, metal nanoparticles capped by small-molecule initiators must be designed to effectively stabilize and disperse metal nanoparticles in fuels with catalysis and initiation. The development of fuel-soluble catalysts is promising.
With the development of modern equipment and analysis methods, predicting and designing novel catalysts and initiators can be anticipated with the assistance of intermediate analysis, DFT calculations, machine learning, and other methods, finally leading to a breakthrough in the cracking process. Fig. 10 Schematic of the preparation of the macroinitiator palmitoyl modified hyperbranched polyglycerol. Reproduced with permission from Ref. [80]. Copyright © 2020 Elsevier High added amount (more than 2 wt%) of small-molecule initiator, may affect physical properties of fuels

Declarations
Conflict of interest The authors declare that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.