Xinyong
Diao
ab,
Ying
Xiong
a,
Yawen
Shi
a,
Longlong
Ma
c,
Chenglong
Dong
d,
Shengbo
Zhang
a and
Na
Ji
*a
aSchool of Environmental Science and Engineering, Tianjin Key Laboratory of Biomass/Wastes Utilization, Tianjin University, Tianjin 300350, China. E-mail: jina@tju.edu.cn
bThe Key Laboratory of Functional Molecular Solids, Ministry of Education, Anhui Normal University, Wuhu 241000, China
cKey Laboratory of Energy Thermal Conversion and Control of Ministry of Education, School of Energy and Environment, Southeast University, Nanjing 210096, PR China
dSinopec Research Institute of Petroleum Processing, Beijing 100083, China
First published on 19th October 2024
Cycloalkanes with the carbon numbers C9–C16 are ideal jet-fuel components and are mainly synthesized by the hydrogenation of petroleum-derived benzenes and the cyclization reactions of linear alkanes. The catalytic conversion of lignin and its derivatives, intrinsically embodying carbocyclic structures, to jet-fuel-range cycloalkanes has been demonstrated as a potential green and economical route, which can improve the sustainability of sustainable aviation fuels (SAFs) as well as reduce the overall greenhouse gas emissions. Direct hydrodeoxygenation (HDO) as well as C–C coupling relay hydrodeoxygenation are the two main routes for the production of cycloalkanes from lignin and its derivatives. In this review, first, the HDO of lignin C–O derivatives to monocycloalkanes over metal–acid catalysts was considered a model reaction to provide an understanding of the catalytic structure–activity relationship of the indispensable HDO process. Then, the production of lignin jet fuel via the simultaneous depolymerization and HDO of real lignin was discussed, followed by the C–C coupling relay hydrodeoxygenation route for polycycloalkanes production, including the alkylation relay hydrodeoxygenation route, aldol condensation relay hydrodeoxygenation route and one-pot conversion route. Furthermore, this paper attempts to highlight the remaining challenges and provide some perspectives for the future design of structure-specific cycloalkanes, aiming to provide insights into the viable utilization of lignin to obtain C9–C16 ideal jet-fuel-range cycloalkanes.
Cycloalkanes are an essential component of fossil fuels, accounting for 30–40 wt% of diesel,2,3 20 wt% of jet fuel (Fig. 1a),4 and 10 wt% of gasoline.5 Generally, cycloalkanes with different total carbon numbers, C6 ring numbers and side-chain substitutions present significant difference in fuel density, flash points, boiling point, freezing point, etc. Therefore, the structural requirement of cycloalkanes for diesel, jet fuel, and gasoline are quite different. In the past decades, cycloalkanes have been mainly synthesized by the hydrogenation of petroleum-derived benzenes and the cyclization reactions of linear alkanes (Fig. 1b). Thus, it is crucial to develop more green and sustainable synthetic routes. Recently, the catalytic conversion of lignin and its derivatives, intrinsically embodying carbocyclic structures, has been demonstrated as a potential green and economical route for the production of cycloalkanes.
![]() | ||
Fig. 1 (a) Composition of jet fuel;4 (b) industrial production of cycloalkanes; (c) cycloalkanes in typical jet fuel components;9–11 (d) hydrocarbon-type analysis of lignin-based jet fuel (LJF) and the average conventional jet fuel carbon distribution (green shaded region); (e) tier α predictions on full distillate LJF (blue triangle and lines) and virtual distilled LJF (black circle and lines), with the conventional fuel experience range (green shaded region), ASTM D7566 specification limit (red bar), and violation of the specification (red shaded region) shown for reference. Reproduced from ref. 7 with permission. Copyright 2024, Elsevier. |
Considering the space and working environment of aviation aircraft, jet fuel is usually required to meet high standards, including high fuel density (>0.8 g mL−1), low freezing point, flash point >38 °C, and boiling point <300 °C. These standards define the structure and composition of jet-fuel-range cycloalkanes. The typical cycloalkane compounds of the conventional RP-3 and JP-8 jet fuels are shown in Fig. 1c, involving C7–C10 alkylated monocyclohexanes and C10–C12 polycycloalkanes. Table 1 shows the carbon number of the typical cycloalkane fraction in Jet A-1, demonstrating a carbon distribution of cycloalkanes as C7–C13.6 It also has been reported that the average conventional jet fuel carbon distribution is C7–C17 (Fig. 1d), and the average carbon number is 11.4.7 However, it should be noted that an elevated concentration of light hydrocarbons (≤C8) would result in the violation of the flash point (Fig. 1e).7 In detail, the flash points of methylcyclohexane (C7) and ethylcyclohexane (C8) are −3 °C and 18.9 °C, respectively, which are well below the standard requirement of 38 °C. The flash point of propylcyclohexane is 35 °C, which is near the range and thus can be present in jet fuel to some extent. Therefore, C9+ monocyclohexanes are considered as good components of jet fuel. Also, C7–C8 cycloalkanes if present in small portions only will not greatly affect the fuel properties of the jet fuel. Other standards also limit the structural of polycycloalkanes. For example, when the carbon ring number exceeds 3, the density of the fuel exceeds 0.90 g mL−1, but the freezing point and viscosity will also both significantly increase, which is not conducive to use at low temperatures for aviation aircraft. Moreover, when the carbon number exceeds 20, the boiling point of cycloalkanes may exceed 350 °C.8 In fact, the final boiling point of the fuel needs to be limited to a maximum temperature of 300 °C to exclude the heavy compound fraction. A very high concentration of polycycloalkanes along with a relatively low paraffin and monocycloalkanes content will contribute to a high boiling point. Hence, the ideal jet-fuel-range cycloalkanes carbon distribution should be C9–C16 when considering multi-factors.
Carbon atom number | Cycloalkanes | Aromatic hydrocarbons | Iso-alkanes | n-Alkanes | Overall content |
---|---|---|---|---|---|
C n H 2n | C n H 2n−6 | C n H 2n+2 | C n H 2n+2 | ||
6 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
7 | 0.2 | 0.2 | 0.0 | 0.0 | 0.4 |
8 | 1.5 | 4.5 | 0.3 | 1.1 | 7.3 |
9 | 7.8 | 10.9 | 1.4 | 6.3 | 26.4 |
10 | 8.7 | 4.3 | 13.0 | 8.9 | 35.0 |
11 | 2.1 | 2.5 | 3.7 | 5.9 | 14.1 |
12 | 0.4 | 0.7 | 1.3 | 2.8 | 5.2 |
13 | 0.1 | 0.2 | 0.9 | 1.4 | 2.6 |
14 | 0.0 | 0.0 | 0.4 | 0.7 | 1.1 |
15 | 0.0 | 0.0 | 0.1 | 0.3 | 0.4 |
16 | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 |
17 | 0.0 | 0.0 | 0.0 | 0.1 | 0.1 |
18 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
wt% sum identified | 21.2 | 23.2 | 21.2 | 27.6 | 92.6 |
wt% sum corrected | 24.9 | 23.2 | 24.9 | 27.6 | 100.0 |
Lignin is the second-most-abundant component of lignocellulose and also is the only renewable energy with an aromatic ring structure in nature (Fig. 2a). Lignin possesses a complex amorphous polymeric structure of phenylpropane subunits randomly linked by C–C and C–O bonds. It should be noted that the complex structure and numerous chemical bonds of real lignin make it difficult to achieve efficient conversion and the mechanism of real lignin conversion is rather complicated. In this regard, studies performed on lignin derivatives as a feedstock have been widely reported, including the development of catalysts and mechanism investigations.
In the early studies, the hydrodeoxygenation (HDO) route, which only considered the cleavage of C–O bonds, was developed for the production of monocycloalkanes. For example, in the HDO of lignin monomers, such as phenol, guaiacol, or lignin dimers, such as diphenyl ether, benzyl phenyl ether, the unit C–O linkages and the side-chain C–O bonds were cleaved and the CC bonds on the aromatic ring were saturated to generate C6–C9 monocycloalkanes as the final products (Fig. 2a). Similarly, for the HDO of real lignin with remaining C–C bonds, usually C6–C18 monocycloalkanes and polycycloalkanes mixtures were obtained (Fig. 2b). Recently, researchers have focused on the simultaneous C–C and C–O bond cleavages of the lignin structure, which can convert lignin C–C linkage dimers and real lignin into C6–C9 monocycloalkanes. However, as mentioned above, only C9+ monocycloalkanes are considered as good components of jet fuel. This suggests that the C–C cleavage should be restrained in the HDO of C9+ lignin derivatives and real lignin to obtain more ideal jet-fuel-range cycloalkanes.
Beside the direct HDO route, researchers have found that lignin-derived compounds contain active groups such as hydroxyl, aldehyde, and carbonyl groups, which can be used to construct target carbon chain molecules through C–C coupling reactions (hydroxyalkylation/alkylation, aldol condensation, Michael addition, etc.) under acid catalysis, followed by HDO to obtain C9–C16 polycycloalkanes as the final products (Fig. 2b). The as-prepared polycycloalkanes presented excellent jet-fuel properties with comparable density, viscosity, lower heating value, freeze point, and flash point to conventional jet fuel.
To date, we can conclude that work on the catalytic HDO and C–C coupling of lignin and its derivatives into renewable jet-fuel-range cycloalkanes has reached an important stage that requires an in-depth review and constructive outlook. Heterogeneous catalysts, especially metal–acid catalysts, have proven to be efficient catalysts both in HDO and C–C coupling processes. It is necessary to summarize the synthesis pathways and provide a greater in-depth understanding of the corresponding catalytic structure–activity relationship.
To the best of our knowledge, while there are several excellent reviews focused on the HDO of lignin and its derivatives into monocycloalkanes,13–15 few reviews have focused on the C–C coupling of lignin derivatives into polycycloalkanes. Recently, Li et al.16 summarized the production of cycloalkanes based on two strategies: the utilization of biomass platform compounds intrinsically embodying a carbocyclic structure (HDO route) and the creation of a carbocyclic structure from a biomass platform compound via chemical reactions (C–C coupling). However, their comprehensive and profound review did not focus on the transformation of lignin and interpret the catalytic structure–activity relationship. Therefore, this review focuses on the catalytic HDO and C–C coupling of lignin and its derivatives into renewable jet-fuel-range cycloalkanes over heterogeneous catalysts, with an attempt to systematically summarize the research progress during the past two decades. For the direct HDO route, the catalytic structure–activity relationship is briefly summarized based on the studies on the HDO of lignin derivatives, and then the focus is on the production of jet-fuel-range cycloalkanes by the one-pot conversion of real lignin. Afterwards, the C–C coupling relay HDO route for polycycloalkanes production is emphatically discussed, including the alkylation relay HDO route, aldol condensation relay HDO route, and one-pot conversion route. Finally, based on the systematic summary and comparison of the direct HDO and C–C coupling relay HDO routes, the remaining challenges are highlighted and some perspectives are provided for the future design of structure-specific cycloalkanes. This review may provide insights into the viable utilization of lignin renewable jet-fuel-range cycloalkanes.
Noble metals, such as Pt, Pd, and Ru, are widely used as metal sites for bifunctional catalysts in industry because of their excellent hydrogenation function and stability. Compared to non-noble metals, noble metals can catalyze oxygen-containing compounds effectively in a hydrogen atmosphere under less severe conditions. In the preparation of metal catalysts, improving the metal dispersion on the support is an effective way to promote catalytic efficiency. Compared with metal clusters, isolated metal cations coordinate strongly with the support but weakly to adsorbates, while not being active in the dissociation of H2.19 Therefore, the efficiency of the HDO with a single isolated metal may be low. However, dispersing a small amount of isolated metal atoms into another metal to form a bimetallic alloy catalyst can improve the catalytic performance. In addition, the metal particle size is also an important factor affecting the catalytic activity. In general, smaller metal nanoparticles are more favorable for benzene ring hydrogenation.20 The deactivation rate is also dependent on the particle size, with the smallest Pt size of 1.5 nm showing a high anti-deactivation activity, while larger Pt sizes deactivated faster.21
Among many active metals, noble metals can activate H2 under mild conditions and catalyze the hydrogenation of unsaturated functional groups, thus obtaining higher catalytic performance. However, the scarcity and high price of noble metals hinder their industrial application. As alternatives to noble metals, cheap and readily available transition metals, such as Ni, Co, Fe, and Cu, also exhibit good hydrogenation activity.
Compared with noble metal catalysts, highly dispersed nickel nanoparticles (NNPs) have wider d electron bands, higher energy density, and a stronger activation activity for H2 and hydrogenated aromatic rings under mild conditions.22–25 In addition, Ni is less oxygenophilic than other non-noble metals and cannot easily overcome the energy barrier for direct C–O bond cleavage.26 Therefore, introducing a second active metal can provide new oxygenophilic active sites (e.g., Co, Mo, Fe, W), thus facilitating C–O bond breaking.27
Above all, the type of active metal sites significantly affect the hydrogenation performance of the catalysts. In most cases, the hydrogenation ability of noble metals is better than that of non-noble metals. However, to improve the deoxygenation performance, noble metals need to be doped with oxygenophilic metals or combined with acidic sites. This modulation method is also common in the design of non-metal catalysts. This is because the introduction of a second metal increases the oxygen vacancies and improves the dispersion of the metal to some extent. In addition to this, the particle size of the metal and the physical properties of the catalyst are also important factors. For example, in transition-metal phosphides, the synthesis of catalysts with larger specific surface areas and smaller and highly dispersed metal sites facilitates the exposure of more active sites as well as the adsorption and activation of H2. As for transition-metal sulfides, the reaction performance can be improved by increasing the number of MoS2 edges.
Higher concentrations of acid sites have been reported to favor the formation of carbonium ions, which contribute to dehydroxylation and demethoxylation.29,30 Catalysts with mesoporous structures and higher concentrations of Brønsted acid sites have better HDO activity for lignin phenolics.31 Strong acidic sites facilitate dehydration, while weak acid sites adsorb the oxygen atoms in hydroxyl or methoxy groups and improve the cleavage ability of C–O bonds.32 However, acidity is not the main reason for the catalytic activity. Too large a pore size and strong acidity can lead to coupling reactions and reduce the selectivity for monocycloalkanes. Therefore, the pore-size structure and acidic sites of zeolites need to be regulated rationally.
In short, weakly acidic metal oxides can be mixed with other acid-strength metal oxides to form mixed supports with a large specific surface area and pore size. Metal oxides with a particular acidic strength can be prepared by changing the preparation method to optimize the morphological characteristics of the supports, thus improving the catalytic activity and stability.
Chen et al.45 used an organic ligand with hydrothermal stability, i.e., 4-trifluoromethyl salicylic acid (TFMSA), to modify Ru/γ-Al2O3 for the catalytic conversion of lignin phenolic compounds to jet fuel. It was shown that adding TFMSA to Ru/γ-Al2O3 helped improve the dispersion of metal nanoparticles, change the ratio of Brønsted/Lewis acid, and significantly increase the amount of Brønsted acid.
In the HDO conversion of phenol, the cyclohexanol hydrogenation rate is much higher than the dehydration rate, indicating that acid-catalyzed cyclohexanol dehydration is the rate-limiting step. Therefore, adjusting the metal and acidactive sites at the bifunctional catalyst interface is crucial to improve the dehydration efficiency. On bifunctional catalysts, the cyclohexanol generated at the metal site can be immediately dehydrated at the nearby acidic site. In contrast, cyclohexanol must repeatedly diffuse between the separated metal and acidic sites on conventional bifunctional catalysts, resulting in a relatively low catalytic activity.
The Brønsted acid site contributes to the dehydration reaction during catalysis, and in some cases, Lewis acids can exhibit better catalysis than Brønsted acids. Lewis acids can selectively bind to and activate specific functional groups (e.g., ether bonds and hydroxyl groups) during the conversion process.46 For example, metal triflates with strong Lewis acids could effectively catalyze the cleavage of lignin β-O-4 ether bonds.47
In summary, acid sites play an important role in removing oxygen from oxygen-containing compounds and facilitating hydrocarbon production. The type and number of acid functions vary slightly from one acid support to another. A reasonable Brønsted/Lewis acid ratio can be obtained by modulation. Brønsted acids offer the more prominent cleavage of C–O bonds than Lewis acids. However, the degree of increase in the strength and the number of acid sites needs to be moderate.
Ju et al.56 obtained a highly stable bifunctional catalyst based on Pt-A/Z by the selective deposition of Pt on alumina in an Al2O3-ZSM-5 nanocomposite. Shifting the metal–acid distance from the millimeter to nanometer scale improved the catalytic performance for the HDO of eugenol. However, further shortening the distance between the acidic and metal sites did not improve the catalytic performance. Therefore, the generally accepted “closer is better” criterion for metal/acid bifunctional catalysts has some distance limitation in the HDO of bio-oxygenated compounds.
In general, it is more beneficial to improve the catalytic performance when the distance between the metal and the acid site is on the nanometer scale.57,58 The zeolite encapsulation approach can shorten the active site spacing and thus improve the catalytic performance because of the synergistic effect between the active sites.59
In summary, in the metal–acid-catalyzed HDO system, it is the metal nanostructure (including the active metal type, particle size, and metal–metal interactions), acid property (including the amount and strength of the acid), and synergistic effects (including the ratio balance and spatial proximity) that determine the catalytic activity for the production of monocycloalkanes. Table 2 summarizes the state-of-the-art metal–acid catalysts in the HDO of lignin derivatives to monocycloalkanes. It can be seen that zeolite-supported noble metal catalysts are the most promising options, offering high HDO activity under mild conditions. The efficient HDO progress plays a crucial role not only in the direct HDO of lignin derivatives, but also has great significance in the cascaded HDO of C–C coupling precursors. Therefore, understanding the catalytic structure–activity relationship of the HDO is essential for both the HDO route and C–C coupling route.
Entry | Substrate | Conv. (%) | Catalyst | T (°C) | P (MPa) | t (h) | Solvent | Products | Molar yield (%) | Ref. |
---|---|---|---|---|---|---|---|---|---|---|
a The molar yield of monocycloalkane was calculated based on the C6 ring molar of initial substrate and product. | ||||||||||
1 | Diphenyl ether | 100 | Ru5%/HZSM-5 | 210 | 1 | 2 | n-Hexane | Cycloalkane | 100.0 | 32 |
2 | Diphenyl ether | 100 | Ru/Ga-HZSM-5 | 180 | 1 | 2 | n-Hexane | Cycloalkane | 100.0 | 52 |
4 | Guaiacol | 100 | 5Ni–5Co/NbOx | 300 | 3 | 2 | Dodecane | Cycloalkane | 98.9 | 27 |
5 | Guaiacol | 100 | Fe(24)Ni(6)–ZrO2 | 300 | 4 | 8 | Octane | Cycloalkane | 89.4 | 53 |
9 | Guaiacol | 100 | Ru2%/BEA-12.5 | 250 | 4 | 2 | — | Cycloalkane | 71.9 | 31 |
10 | Guaiacol | 100 | Co-Al2O3@USY | 180 | 3 | 4 | n-Hexane | Cycloalkane | 93.6 | 48 |
11 | Guaiacol | >99 | Ru–Cu/HY | 250 | 4 | 2 | H2O | Cycloalkane | 44.8 | 54 |
13 | Diphenyl ether | 100 | Ru/SHZSM-5–100 | 150 | 1 | 2 | n-Hexane | Cycloalkane | 100.0 | 18 |
14 | Guaiacol | 100 | Ni/ZrO2–SiO2 | 300 | 5 | — | Dodecane | Cycloalkane | 96.8 | 41 |
15 | Guaiacol | 100 | Ru/TiO2(TNP) | 250 | 1 | 4 | Octane | Cycloalkane | 100.0 | 55 |
16 | Phenol | 100 | Ru-20TFMSA/γ-Al2O3 | 240 | 2 | 1 | Dodecane | Cycloalkane | 84.0 | 45 |
17 | Guaiacol | 99.9 | Ru/C-HPW | 200 | 1 | 4 | Octane | Cycloalkane | 98.6 | 50 |
18 | Eugenol | 77.3 | Pt-A/Z | 300 | 4 | 4 | n-Hexane | Propylcycloalkane | 38.0 | 56 |
19 | Phenol | 100 | Ru@H-ZSM-5 | 150 | 5 | 4 | Decalin | Cycloalkane | 90.0 | 57 |
20 | Guaiacol | 100 | Ru@HMCM-22-IN | 160 | 3 | 5 | Dodecane | Cycloalkane | >95.0 | 58 |
Linkage | Bond dissociation energy/kJ mol−1 | Number/100 ppu | |||
---|---|---|---|---|---|
Softwood | Hardwood | Grasses | |||
C–O bonds | β-O-4 | 226–301 | 43–50 | 50–65 | 74–84 |
α-O-4 | 209–234 | 6–8 | 4–8 | 5–11 | |
4-O-5 | 327–348 | 4 | 6–7 | n.d. | |
5–5 | 481–494 | 10–25 | 4–10 | n.d. | |
C–C bonds | β-5 | 226–264 | 9–12 | 4–6 | 5–11 |
β–β | 281–339 | 2–4 | 3–7 | 1–7 | |
β-1 | 272–289 | 3–7 | 5–7 | n.d. |
Most studies only demonstrated the production of cycloalkanes as bio-fuels or bulk chemicals, but did not investigate their properties. As for lignin-based jet fuel, it should meet the current ASTM D7566 standard specifications.68Fig. 4a illustrates the proposed processing pathway for converting lignin to jet-fuel-range C9+ mono- and polycycloalkanes.69 However, the formation of light hydrocarbons (≤C8) cannot be avoided and sometimes they exist as main products, as shown in Table 4. Recently, researchers have paid more attention to the chemical compositions and properties of lignin-based jet-fuel-range hydrocarbons. Prof. Yang et al.8 adopted comprehensive two-dimensional gas chromatography (GC × GC) with both mass spectrometry and flame ionization (FID) detection to identify and quantify the species in lignin-based jet fuel (US patent 9518076 B2). As shown in Fig. 4b, the main compositions were n-paraffins, iso-paraffins, and mono-, di-, and tri-cycloalkanes, of which the majority contained carbon numbers in the range of 7–20. Among them, cycloalkanes were the most abundant components with the concentrations of mono-, di-, tri-, and tetra-cycloalkanes being 14.42, 31.95, 21.12, and 10.43 wt%, respectively (Fig. 4c). Fig. 4d presents a comparison of the simulated distillation (SimDis) (ASTM D2887) data of lignin-based jet fuel, a fractionated lignin-based jet fuel, a fractionated and 50/50 wt% blend with conventional jet fuel, and a range of jet fuels with extreme operability properties. The result showed that the neat lignin-based jet fuel could hardly meet the standard of jet fuel. It is therefore usually blended with conventional jet fuel, with a strict blend limit (<10 wt%). Further, Prof. Yang et al.70 removed 10% of C6 monocycloalkanes and C17+ molecules from lignin-based jet fuel in an attempt to achieve a higher blending ratio with conventional jet fuel. As mentioned above, removing the C17+ molecules should improve the low-temperature viscosity and freezing point of the neat LJF sample, while removing the C6 cyclohexane will retain the flash point above the specification limit. The results also proved that the blend limit could be improved to 20% after the removal. Their latest research also claimed that an elevated concentration of C7 and C8 will lead to a violation of the flash point, and these components should be removed by distillation.7
![]() | ||
Fig. 4 (a) Proposed processing pathway for converting lignin into jet fuel. Reproduced from ref. 69 with permission. Copyright 2018, the American Institute of Aeronautics and Astronautics, Inc. (b) Bubble plot representation of a ‘normal’ GC × GC-FID contour plot for the sample; (c) summary of the GC × GC-FID hydrocarbon composition (wt%); (d) simulated distillation curves of lignin jet fuel compared to some reference jet fuels. Reproduced from ref. 8 with permission. Copyright 2019, Elsevier. |
Entry | Substrate | Catalyst | T (°C) | P (MPa) | t (h) | Products | Yield (wt%) | Ref. |
---|---|---|---|---|---|---|---|---|
a Calculation of the product mass yield was based on the masses of the initial lignin reactant and final products. | ||||||||
1 | Birch lignin | Ru/C | 250 | 0.7 | 20 | Monocycloalkanes | 29.3 | 64 |
2 | Pyrolysis bio-oil from pinewood | RANEY® Ni + H-BEA-35 | 160 | 2-PrOH | 2 | Cycloalkanes | 7.2 | 71 |
3 | Birch lignin | Ru/NbOPO4 | 310 | 5 | 40 | C6–C9 cycloalkanes | 12.0 | 42 |
4 | Poplar lignin | Mo0.06–Co9S8/Al2O3 | 265 | 3 | 20 | Monocycloalkanes | 5.6 | 72 |
5 | Pyrolysis bio-oil from poplar wood | Ni/Nb2O5 | 300 | 7 | 16 | Cycloalkanes | 25.4 | 66 |
6 | Kraft lignin | S–NiMo/MgO–La2O3 | 350 | 10 | 4 | Cycloalkanes | 1.1 | 67 |
7 | Organosolv lignin | Ir–ReOx/SiO2 | 260 | 4 | 10 | Monocycloalkanes | 19.5 | 67 |
Polycycloalkanes | 1.6 | |||||||
8 | Birch lignin | Pt/NbOPO4 | 190 | 5 | 20 | Alkylcyclohexanes | 4.8 | 65 |
9 | Enzymatic lignin | Ni/ASA-1 | 300 | 6 | 160 min | Monocycloalkanes | 35.3 | 73 |
Polycycloalkanes | 6.1 | |||||||
10 | Lignin | Ni/S-1 | 300 | 6 | 2 | C6–C9 cycloalkanes | 24.0 | 74 |
C10–C17 cycloalkanes | 9.0 | |||||||
11 | Organosolv lignin | Pt/HAP + Ni/ASA | 300 | 6 | 4.5 | Monocycloalkanes | 26.5 | 75 |
Polycycloalkanes | 15.5 | |||||||
12 | Organosolv lignin | Ni/HBEA | 250 | 2 | 6 | Monocycloalkanes | 29.1 | 76 |
Polycycloalkanes | 6.0 | |||||||
13 | Poplar lignin | Ni2P–Al2O3(H)71 | 250 | 5 | 15 | Monocycloalkanes | 6.0 | 77 |
Polycycloalkanes | 1.0 | |||||||
14 | Pine wood lignin | RuNi/HY | 250 | 4 | 4 | Hydrocarbon products | 32.0 | 54 |
15 | Dimers of pine sawdust | Pd/C | 300 | 3 | 16 | C16–C18 cycloalkanes | 28.0 | 78 |
16 | Alkali lignin | HY + Ru/Al2O3 | 300 | 4 | 4 | C7–C11 cycloalkanes | 3.4 | 63 |
C12–C18 cycloalkanes | 18.4 | |||||||
17 | Pyrolysis bio-oil from eucalyptus wood | Pd/m-MoO3–P2O5/SiO2 | 250 | 1 | 15 | C6–C9 cycloalkanes | 13.4 | 79 |
18 | Birch lignin | Pd/Nb2O5 | 250 | 0.7 | 20 | C7–C9 cycloalkanes | 24.4 | 80 |
19 | Oligomer of hardwood | Pt/H-MOR | 300 | 4 | 24 | C6–C9 cycloalkanes | 36.8 | 62 |
To sum up, real lignin can be converted into mono- and polycycloalkanes mixtures via a simultaneous depolymerization and HDO, and so-called lignin-based jet fuel with carbon numbers in the range of C6–C20 could be obtained. This lignin-based jet fuel still has some distance to meet the current jet fuel standards due to the inevitably generated light cycloalkanes (≤C8) and C17+ polycycloalkanes. The high concentration of light cycloalkanes and C17+ polycycloalkanes will contribute to violations of the flash point and boiling point limits, respectively. More studies should be performed on the after treatment of lignin-based jet fuel to meet the ASTM standard.
Although the density of bicyclohexane is high, the freezing point value (2.6 °C) needs to be reduced but it can be used as a fuel additive (Table 5). In contrast, polycycloalkanes containing a five-membered ring structure, such as cyclopentyl, cyclohexane, and (cyclopentyl methyl)cyclohexane, have good low-temperature properties due to their asymmetric structure. Such substances can be obtained either by alkylation reactions of cyclopentanol and lignin phenolic compounds85 or by promoting isomerization reactions during the HDO of lignin phenolic compounds.86 In addition, decahydronaphthalene has been widely researched as an aviation fuel additive with a high thermal stability and energy density. Decahydronaphthalene is mainly obtained from the hydrogenation of naphthalene in petroleum. In the long term, there is still a need to develop renewable synthetic routes for decahydronaphthalene. Structurally, decahydronaphthalene consists of two cyclohexanes, so lignin derivatives with a six-membered ring structure can be used as raw materials for the preparation of decahydronaphthalene. Generally, besides the strategy involving the direct HDO of lignin C–C dimers or multimers intrinsically embodying a multi-carbocyclic structure without C–C cleavage, another strategy involves the creation of a multi-carbocyclic structure via the C–C coupling of lignin C–O derivatives followed by HDO. The catalysis system, including reaction route, catalysts, mechanism, presents significant differences based on the different detailed reactions. In this chapter, we start with the design thoughts on the reaction route, and then emphatically discuss the corresponding catalytic structure–activity relationship.
Unlike lignin C–C dimers, lignin C–O derivatives first require C–C coupling to form dimer precursors with a multi-carbocyclic structure, followed by the HDO reaction to produce polycycloalkanes. According to the different structures of various lignin C–O derivatives, mainly three C–C coupling routes have been designed to synthesize multi-carbocyclic precursors, including the alkylation relay HDO route, aldol condensation relay HDO route, and one-pot conversion (Fig. 5).
![]() | ||
Fig. 5 Reaction routes for converting lignin derivatives into polycycloalkane. (a) Alkylation relay HDO route, (b) aldol condensation relay HDO route, (c) one-pot conversion route. |
Nie et al.91 investigated the alkylation reaction of phenol with cyclopentanol catalyzed by hydrophobic acid resin (R-x) under mild conditions (140 °C for 4 h). Among the catalytic results for R-x (x = 0.25–1.5), R-0.25 (79.8% phenol conversion) was much better than Hβ (51.6% phenol conversion), mainly due to the more acidic and better hydrophobic nature of R-x than Hβ. In bifunctional catalysts, the acid site also has the function of dehydration, which is a competitive reaction with alkylation. Therefore, it is not desirable to excessively pursue the acid site strength, and it is more important to consider how to balance the relationship between excessive dehydration and alkylation. At the same time, extreme acid sites may lead to products with larger molecular weights, which is not favorable for forming the target alkylation products.92 Therefore, the strength and number of acid sites need to be appropriate.
In addition to the acidity, the solid acid morphology characteristics are important factors that can affecting the product distribution. Large-pore-sized zeolites facilitate the diffusion of macromolecular intermediates, while small-pore-sized supports have spatial hindrances and are not conductive to forming polycycloalkanes.93 However, the macroporous zeolite produced after excessive alkali treatment causes a partial collapse of the zeolite framework structure, reducing its specific surface area and total acid content.94 In addition, solid acids with a high specific surface area at appropriate acid amounts and acidity make it easy for the molecules to come into contact with the acid sites.95
Zhao et al.96 found that the alkylation of phenol and cyclohexanol was high enough for effective alkylation reactions when the molar ratio of phenol to metal was relatively high. At 473 K, the optimized phenol/Pd ratio was 2254–4508 mol mol−1 with a 67%–85% yield of the alkylation products. When the phenol/Pd ratio was reduced to 564 mol mol−1, the dominant product was cyclohexane (98% yield), while the alkylation product accounted for only 2.1%. In addition, Pd/HBEA showed higher selectivity for alkylation than the physically mixed catalyst, indicating that the shorter distance between the metal and acid sites significantly increased the probability of the C–C coupling reaction occurring between the phenol and intermediate. Moreover, strong Brønsted solid acids catalyze dehydration but not C–C bond formation. Only in combination with the pore size of the zeolite pore channel can phenol undergo C–C coupling at the Brønsted acid site and then be hydrodeoxygenated to polycycloalkanes.
Zou's group95,97 reported the alkylation of benzyl ether (or benzyl alcohols) with phenols, including phenol, anisole, and guaiacol. After the HDO of the alkylation product over the Pd/C+ HZSM-5 catalyst, polycycloalkanes with a bicyclic structure were obtained. They also synthesized substituted diphenyl methane by the acid-catalytic alkylation of lignin-derived phenols (phenol, anisole, guaiacol) with benzyl ether or benzyl alcohols (Fig. 6).95 The result showed that among the different solid acid catalysts, MMT-K10 exhibited superior activity than HPW, Amberlyst-15, and Al-MCM41, with 80.6% selectivity for the substituted diphenylmethane precursor, ascribed to its modest acid property and open lamellar structure (Fig. 6a). Besides, the effects of the reactant ratio and type were evaluated to obtain an optimized activity (Fig. 6b and c). They also synthesized ethyl-substituted bicyclic cycloalkanes from the lignin-derived 4-ethylphenol and phenylmethanol.97 In this alkylation reaction, HPW was found to work as a better catalyst with 71% selectivity for the monoalkylated products (2-benzyl-4-ethylphenol and 3-benzyl-4-ethylphenol) (Fig. 6d and e). The resulting fuel had a density and viscosity of 0.873 g cm−3 and 10.7 mm2 s−1, respectively, at 20 °C. Moreover, the freezing point of the fuel was −42 °C, much lower than that of dicyclohexylmethane and dicyclohexane.
![]() | ||
Fig. 6 (a) Performance of catalysts in the alkylation of anisole with benzyl ether over different catalysts. Reaction conditions: 23.3 mmol anisole, 23.3 mmol benzyl ether, 0.1 g catalyst, 110 °C, 2 h. (b) Effect of anisole to benzyl ether ratio in the alkylation reaction over MMT-K10. (c) Alkylation of different phenols with benzyl ether over MMT-K10. Reaction conditions: 23.3 mmol phenols, 5.83 mmol benzyl ether (2.92 mmol for guaiacol), 0.1 g MMT-K10, 110 °C (90 °C for guaiacol), 2 h (9 h for guaiacol). Reproduced from ref. 95 with permission. Copyright 2017, Elsevier. (d) Schematic alkylation reaction of 4-ethylphenol with phenylmethanol. (e) Distribution of substances during the process of the alkylation reaction. Reaction conditions: 9.77 g (80 mmol) 4-ethylphenol, 4.32 g (40 mmol) phenylmethanol, 0.85 wt% HPW, 110 °C. Reproduced from ref. 97 with permission. Copyright 2018, Elsevier. (f) Time-on-course process of the production over Au/CdS and CdS. Reaction conditions: 1 mL of 4-ethyl-1-methoxybenzene, 50 mg of Au/CdS, 10 mL of CH3CN, 455 nm LED (60 W) irradiation, Ar atmosphere. (g) Proposed reaction mechanism for the photocatalytic dehydrocoupling of 4-ethyl-1-methoxybenzene. Reproduced from ref. 78 with permission, Copyright 2021. Wiley. |
Beside thermocatalytic alkylation over an acid catalyst, photocatalytic C–C coupling was also developed for some thermodynamically unfavorable reactions. Dou et al.78 performed the photocatalytic coupling of lignin-derived 4-ethyl-1-methoxybenzene using an Au/CdS photocatalyst under visible-light (455 nm LEDs) irradiation in MeCN solvent at room temperature. The yield of the dimer reached 83 wt% in 24 h accompanied with 11 mmol gcatal−1 H2 production (Fig. 6f). The catalyst was also active for the self-coupling of various lignin-derived phenolics to form dimers with 57–83 wt% yields. The proposed reaction mechanism revealed the combination of Au and CdS could significantly improve the separation of photogenerated electrons and holes, which finally enhances the coupling activity of Au/CdS (Fig. 6g).
Methyl benzaldehyde and vanillin, typical downstream products obtained via the catalytic oxidation of lignin, are usually employed as aldehyde reagents to participate in the cross condensation, while cyclohexanone, cyclopentanone, methyl isobutyl ketone, etc. are regarded as ketone reagents. Xu et al.98 reported the synthesis of C14 oxygenates (i.e., 2-(2-methylbenzylidene)cyclohexanone or 2-(4-methylbenzylidene)cyclohexapnone) from the aldol condensation of 2-methyl benzaldehyde (or 4-methyl benzaldehyde) and cyclohexanone (Fig. 7a). Among the investigated catalysts, they found that the EAOAc ionic liquid presented superior activity and good stability, with carbon yields 85% under 353 K for 6 h. The mechanism study showed that the excellent performance of the EAOAc ionic liquid could be ascribed to its special chemical structure and/or the synergetic effect of ethanolamine and acetic acid (Fig. 7b).
![]() | ||
Fig. 7 (a) Conversions of 2-methyl benzaldehyde and the carbon yields of 1A or 1B over ionic liquid catalysts. Reaction conditions: 10 mmol 2-methyl benzaldehyde, 10 mmol cyclohexanone, 1 mmol ionic liquid; 333 K, 4 h. (b) Reaction mechanism for the generation of 1A from the aldol condensation of 2-methyl benzaldehyde and cyclohexanone. Reproduced from ref. 98. with permission, Copyright 2018. Royal Society of Chemistry. (c) 2-Methyl benzaldehyde conversions and 1A yields over solid bases. Reaction conditions: 10 mmol MIBK, 10 mmol 2-methyl benzaldehyde, 0.2 g catalyst; 403 K, 6 h. (d) Results for the aldol condensation reactions of 2-methyl benzaldehyde with different ketones over K2CO3/Al2O3. Reaction conditions: 20 mmol ketone, 10 mmol 2-methyl benzaldehyde, 0.2 g catalyst; 353 K, 6 h. Reproduced from ref. 99 with permission, Copyright 2019. American Chemical Society. |
Beside acid catalysts, base catalysts can also be used in the aldol condensation. Zhang et al.99 reported the aldol condensation of methyl benzaldehyde and methyl isobutyl ketone to synthesize C–C coupling precursors over solid base catalysts. Interestingly, the as-synthesized precursors could be converted into polycycloalkanes via intramolecular dehydration/alkylation/hydrogenation reactions (Fig. 7c). Among the studied solid bases, K2CO3/Al2O3 showed the highest activity in the aldol condensation reaction, which might be rationalized by its larger base site concentration and higher specific surface area. Moreover, K2CO3/Al2O3 was effective for the aldol condensation of other ketones, and the as-synthesized precursors could be transformed into octahydro-indenes with different structures (Fig. 7d). The cycloalkane mixtures obtained in this work had a density of 0.857–0.944 g mL−1, higher than the currently used jet fuel (∼0.8 g mL−1). In addition, they had freezing points of 227.7–240 K and could be used to improve the volumetric heat values and/or thermal stability of jet fuels.
It is well known that cyclohexanone can be obtained from the hydrogenation reaction of lignin-derived phenol, and it can undergo self-condensation to construct interunit C–C bonds between C6 rings. Liu et al.100 demonstrated that phenol could be selectively hydrogenated to cyclohexanone over a dual supported Pd-Lewis acid catalyst, with both the conversion and selectivity exceeding 99.9%. However, it should be noted that phenol hydrogenation to cyclohexanol was much easier than cyclohexanone. In most cases, the self-condensation of cyclohexanone happened during the one-pot conversion of phenol, which is discussed in the next section. Until now, few studies have been reported on the direct self-condensation of cyclohexanone (employed as an initial reactant) to produce bicyclohexane. Sun et al.101 performed the self-condensation of cyclohexanone over the base-Mg–Zr–O catalyst, with the formation of a bicyclic oxygenate precursor. However, the self-coupling depth was insufficient, and limited the production of larger molecules.
Generally, the condensation between cyclic ketones presents a greater priority than that between cyclic ketones and linear aldehydes or ketones, leading to the production of bicyclic oxygenate precursors of jet fuel with a higher C number. Both acid and base catalysts have been proven to be effective in cross- and self-condensation reactions. In addition, similar to the alkylation relay HDO route, two different catalysts for the cascade reaction are essential in the aldol condensation relay HDO route to achieve polycycloalkanes as the final products (Table 6).
Entry | Substrate | C–C coupling | HDO | Ref. | ||||
---|---|---|---|---|---|---|---|---|
Catalyst A | C–C coupling precursor | Molar yield (%) | Catalyst B | Polycycloalkanes | Molar yield (%) | |||
a The molar yield of polycycloalkane was calculated based on the C6 ring molar of initial substrate and product. b Total selectivity of 2-benzyl-4-ethylphenol and 3-benzyl-4-ethylphenol. | ||||||||
1 | Cyclopentanone | R-2 | Bicyclic compounds | ∼85.0 | Pd/Hβ | Bicyclic alkanes | 57.0 | 91 |
Cyclohexanone | Tricyclic compounds | Tricyclic alkanes | 41.8 | |||||
2 | Phenol | Hβ | Bicyclic compounds | 52.6 | Pd/C | Bicyclic alkanes | 74.3 | 95 |
Cyclopentanol | Tricyclic compounds | 14.3 | Tricyclic alkanes | 16.2 | ||||
3 | 4-Ethylphenol | HPW | 2-Benzyl-4-ethylphenol | 71b | Pd/C | Bicyclic alkanes | 79.2 | 97 |
Phenylmethanol | 3-Benzyl-4-ethylphenol | HZSM-5 | ||||||
4 | 4-Ethyl-1-methoxybenzene | Au/CdS | Dimers | 83.0 | Pd/C | C16–C18 cyclic alkanes | 58.0 | 78 |
5 | Methyl benzaldehyde | EAOAc | 2-(2-Methylbenzylidene)cyclohexanone | 85.0 | Pd/C | 1-Methyldodecahydro-1H-fluorene | 92.5 | 98 |
Cyclohexanone | ||||||||
6 | Methyl isobutyl ketone | K2CO3/Al2O3 | Branched octahydro-indenes | 86.2 | Pd/C | Branched octahydro-indene | 96.8 | 99 |
Methyl benzaldehyde |
However, as discussed in Chapter 2, metal–acid catalysts are efficient for the production of monocycloalkanes from lignin C–O derivatives. Scheme 2 displays the reaction pathway comparison for monocycloalkanes and polycycloalkanes production from phenol conversion over metal–acid catalysts. Recently, several studies have achieved important progress on the metal–acid-catalyzed one-pot conversion of phenol to polycycloalkane via regulating the microstructure of the catalysts and reaction conditions. There are mainly two different synthetic routes: (a) adding cyclohexylphenol or cyclopentanol as electrophilic reagents, which react with phenol via alkylation; or (b) utilizing phenol as a single substrate to in situ generate cyclohexylphenol or cyclohexanone, in which the former reacts with phenol via alkylation and the latter undergoes self-condensation.
![]() | ||
Scheme 2 One-pot reaction pathway comparison for monocycloalkanes and polycycloalkanes production from phenol conversion. |
For route (a), with the presence of extra-added electrophilic reagents, the alkylation of phenol and electrophilic reagents can be easily achieved over an acid catalysts, leading to a high yield of multi-cyclic oxygenate precursors. Nie et al.85 developed a one-pot synthesis of bi- and tricyclic cyclohexane from phenol with cyclopentanol through an enhanced alkylation followed by HDO over Pd/C+ Hβ dual catalysts, affording a high molar yield of 83.9% from the starting reactant (Table 7, entry 1). The mixture had a density of 0.89 g mL−1 at 20° C, net heat of combustion of 37.9 MJ L−1, freezing point lower than −75 °C, and viscosity of 22.2 mm2 s−1 at −40 °C. Its high-density characteristic makes it promising as an aviation jet fuel or additive.
Entry | Substrate | Conv. (%) | Catalyst | T (°C) | P (MPa) | t (h) | Solvent | Products | Sel./yield (%) | Ref. |
---|---|---|---|---|---|---|---|---|---|---|
a The molar yield and selectivity of polycycloalkane was calculated based on the C6 ring molar of initial substrate and product. | ||||||||||
1 | Phenol | 66.7 | Pd/C | 200 | 0.2 (N2) | 7 | — | Bicyclic alkanes | 74.3 (S) | 85 |
Cyclopentanol | 100.0 | HBEA | 6.0 (H2) | Tricyclic alkanes | 16.2 (S) | |||||
2 | Phenol | 83.0 | Pt/Hβ | 180 | 0.5 (N2) | 8 | Decalin | Bicyclic & tricyclic | 83.4 (Y) | 94 |
Cyclohexanol | 100.0 | 0.5 (H2) | 8 | |||||||
3 | Guaiacol | 100.0 | Pd2%/Hβ | 220 | 3 (H2) | 4 | n-Hexane | Cycloalkane | 39.8 (Y) | 20 |
Bicyclohexane | 21.8 (Y) | |||||||||
(Cyclopentylmethyl)cyclohexane | 16.0 (Y) | |||||||||
4 | Diphenyl ether | 100.0 | Ni10%/YBCZ | 160 | 5 (H2) | 18 | n-Pentane | Cycloalkane | 87.9 (Y) | 22 |
Bicyclohexane | 10.8 (Y) | |||||||||
5 | Cyclohexylphenol | 100.0 | Pd/C | 150 | 4 (H2) | 30 | H2O | Bicyclohexane | 62.0 (S) | 87 |
HZSM-5 | (Cyclopentylmethyl)cyclohexane | 32.0 (S) | ||||||||
6 | Diphenyl ether | 97.0 | Pd/HY | 200 | 3.4 (H2) | 1 | Decalin | Cycloalkane | 36.0 (Y) | 102 |
[1′-Bicyclohexyl]-2-one | 23.4 (Y) | |||||||||
7 | Guaiacol | 98.7 | Ni/Y | 300 | 4 (H2) | 200 min | — | Cycloalkane | 36.6 (S) | 93 |
Bicyclohexane | 20.5 (S) | |||||||||
8 | Phenol | 100.0 | Pd/C | 160 | 5 (H2) | 12 | H2O | Cycloalkane | 40.0 (S) | 96 |
HBEA | Bicyclohexane | 55.0 (S) | ||||||||
9 | Phenol | 90.0 | Co2P/Beta | 240 | 4 (H2) | 2 | Decalin | Cycloalkane | 54.0 (S) | 103 |
Cyclohexylphenol | 34.0 (S) | |||||||||
10 | Phenol | 60.0 | Pd/C | 200 | 4 (H2) | 4 | H2O | Hydroalkylation products | 25.0 (Y) | 104 |
La-BEA |
Similarly, Shen et al.94 developed a bifunctional Pt/Hβ catalyst for the one-pot conversion of phenol with cyclopentanol, which obtained an 83.4% yield of bicyclic and tricyclic cyclohexane in the final product (Table 7, entry 2). They found that the larger pore size of the alkali-treated Hβ would not limit the production of polycyclic compounds, and a good acid content in a particular mesoporous specific surface area could speed up the reaction rate. The bifunctional catalyst synthesized in this study makes it possible to synthesize fuels directly from lignin oil or other biomass molecules in a one-pot process.
For route (b), since dehydration reactions also occur at the acid active site, HDO and alkylation (or aldol condensation) are two competing pathways over metal–acid catalysts, even if the rate of alkylation (or aldol condensation) is much lower than the hydrogenation of aromatic rings. Scaldaferri et al.102 observed that diphenyl ether hydrogenation occurs significantly more than C–C coupling reactions over Pd/OTS-HY catalysts. This is because C–C coupling must occur after hydrogenation and C–O cleavage of the reactants. If the metal function is too active compared to the acid function, it will increase the rate of secondary hydrogenation of the reactants, and the yield of the C–C coupling product will decrease. Therefore, it is important to balance the competition rate between the hydrogenation over the metal sites and the alkylation or aldol condensation over the acid sites.
García-Minguillán et al.92 performed a one-pot synthesis of cyclohexylphenol via isopropanol-assisted phenol conversion on a tandem catalytic system formed by RANEY®Ni plus hierarchical Beta zeolites. Under this tandem catalytic system, the synthesis of cyclohexylphenol was achieved with remarkable selectivity (∼70%) and high conversion (64%) after 1 h of reaction at 150 °C. The hierarchical β-zeolite with more Brønsted acid sites and stronger Brønsted acid sites promoted the formation of cyclohexylphenol at a relatively low temperature compared to Al-SBA-15. However, the steric hindrance of micropore formation in ZSM-5 made it much less selective for cyclohexylphenol than H-Beta.
Gutiérrez-Rubio et al.103 reported a one-pot hydroalkylation process to prepare cyclohexylphenol by loading Co2P onto acidic zeolites using an impregnation method with phenol as the only organic substrate. The results showed that the highest phenol conversions were obtained over Co2P/MCM-22 and Co2P/Beta catalysts, which had the smallest Co2P nanoparticles and the most considerable acid site accessibility due to their large mesoporous/external surface areas. The HDO/hydrogenation activity was weakened due to the formation of CoAlPO groups on the Co2P/Beta catalysts. As a result, the extension of the phenol HDO/hydrogenation to cyclohexane/methylcyclopentane pathway was reduced, allowing the alkylation reaction between unconverted phenol and cyclohexene molecules to drive the system to form more cyclohexylphenol.
Hu et al.20 prepared Pd-based zeolite catalysts for the HDO conversion of guaiacol using a modified deposition–precipitation (DP) method. The highest hydrocarbon production of guaiacol was achieved with Pd2%/Hβ (DP) catalysts, with a 39.79% cyclohexane yield and 21.84% bicyclohexane yield. For the same Si/Al ratio of zeolites, the catalytic activity of the Pd2%/Hβ catalyst for guaiacol HDO was higher than that of the Pd2%/M and Pd2%/HZSM-5 catalysts, which mainly originated from the synergistic effect between the metal nanoparticles and catalyst acid sites. Due to the presence of strong Lewis acid sites, the coupling reaction could significantly promote the rearrangement of cyclohexanone and cyclohexane to form bicyclohexyl-2-one. In addition, the coupling reaction also promoted the rearrangement of cyclohexane to polycycloalkanes, such as bicyclohexane.
Table 7 summarizes the state-of-the-art one-pot conversion of lignin C–O derivatives to polycycloalkanes. When adding cyclohexylphenol or cyclopentanol as electrophilic reagents to react with phenol, the C–C coupling was much easier to be carried out, leading to a high polycycloalkanes yield (>80%). When adopting phenol as a single substrate, the catalytic efficiency was relatively low due to the competition between the hydrogenation and C–C coupling, which generated monocycloalkanes as major products while the yield of polycycloalkanes was unsatisfactory (<50%). More studies should be performed on the catalyst design and on mechanism investigations to improve the production of polycycloalkanes.
For polycycloalkanes production from lignin derivatives, the two-step routes, including the alkylation relay HDO route and the aldol condensation relay HDO route, present relatively high yields of polycycloalkanes, but two different catalytic systems are needed. As for the one-pot conversion route, the catalytic efficiency is relatively low at present due to the competition between the HDO and C–C coupling reaction in the same catalytic system, which generate monocycloalkanes as the major products while the yield of polycycloalkanes is unsatisfactory. Strategies such as adding cyclohexylphenol or cyclopentanol as electrophilic reagents into the catalytic system could facilitate the C–C coupling to realize a higher polycycloalkanes yield.
For the catalytic structure–activity relationship of the polycycloalkanes production system, current research has mostly focused on the design of reaction routes rather than the regulation of catalyst structures. Yet it was found that by adjusting the properties of the metal–acid catalyst, the competition between the HDO and C–C coupling reaction could be balanced to afford polycycloalkanes as the final products. More studies with a focus on the catalyst design and reaction mechanism could further promote the efficient production of polycycloalkanes from lignin derivatives.
The pore size of the catalyst is also an important factor affecting the product. Smaller pore sizes with greater spatial obstruction allow only small molecular weight substances to pass through and are more suitable for the synthesis of monocycloalkanes. The synthesis of polycycloalkanes, on the other hand, requires larger pore sizes. The C–C coupling reactions are mostly carried out under milder conditions compared to the reaction conditions of HDO reactions. It is hypothesized that at high temperatures, the acidic sites would favor deoxygenation over C–C coupling.
(i) More attention on the combination property of lignin-derived cycloalkanes. Cycloalkanes with different total carbon numbers, C6 ring numbers, and side chain substitutions present significant difference in fuel density, flash point, boiling point, freezing point, etc. In particular, jet-fuel-range cycloalkanes should meet strict standards on fuel density, freezing point, flash point, boiling point, etc. Therefore, more studies should be performed on the relationship between the fuel properties and fuel components, which could aid designing proper fuel molecules with a suitable fuel density, freezing point, and other properties.
(ii) Further development of bifunctional catalysts with non-noble metals. In the preparation system of cycloalkanes, noble metals have good hydrogenation activity and can obtain satisfactory results under mild reaction conditions. Moreover, noble metals are widely used to prepare polycycloalkanes, while the reaction conditions and catalytic steps limit the application of non-metal metals. Therefore, developing more efficient and milder non-noble metal catalysts deserves great attention. The catalytic activity and selectivity can be improved by adjusting the metal particle size, dispersion, and metal–support interactions. In addition, oxygenophilic metals have an affinity for oxygen-containing groups in phenolic compounds, which can help the cleavage of C–O bonds by adding oxygenophilic metals.
(iii) Reasonable regulation of the acid active sites. In the catalytic system of polycycloalkanes, the acid site can catalyze the C–C coupling reaction. However, at the same time, acid sites can also participate in dehydration. Once the dehydration reaction dominates, the possibility of the C–C coupling reaction decreases. In addition, the strength and number of acid sites should be appropriate. Too strong or too many acid sites is detrimental to the formation of the target product. Therefore, further studies are needed to regulate the acid sites to improve the target products’ selectivity.
(iv) Focus on the synergistic effect between the metal and acid. The synergistic effect between the metal and acid is the focus of much research on bifunctional catalysts. The catalytic performance is usually better when the distance between the metal and acid sites is on the nanometer scale. This can be initiated with the preparation method of the catalyst, such as zeolite encapsulation. In addition, by adjusting the ratio of metal to acid sites, the yield of the target product can also be improved.
(v) Design of one-pot catalytic reactions. Most of the preparation pathways for monocycloalkanes are one-pot reactions. However, in the study of polycycloalkanes, cascade reactions are widely used because this reduces the catalytic difficulty. However, this process involves complex separation and purification of the target product at a high cost. From the principle of green economy, bifunctional catalysts are more attractive to catalyze the reaction sequentially in the same reactor.
(vi) In-depth investigation of the C–C coupling reaction mechanism. At present, alkylation and aldol condensation reactions are widely used in the study of the lignin preparation of polycycloalkanes. Reactions utilizing other C–C coupling mechanisms for catalytic synthesis are fewer. Therefore, catalytic pathways for the preparation of polycycloalkanes utilizing different C–C coupling mechanisms could be developed. In addition, the influence of the acidic site species in lignin C–C coupling needs to be clarified and requires further study.
This journal is © The Royal Society of Chemistry 2024 |