Home Journals Progress in Chemistry
Progress in Chemistry

Abbreviation (ISO4): Prog Chem      Editor in chief: Jincai ZHAO

About  /  Aim & scope  /  Editorial board  /  Indexed  /  Contact  / 
Review

Efficient Catalysts for the Selective Hydrogenation of Unsaturated Aldehydes

  • Xingyue Yang ,
  • Shijie Zhou ,
  • Yusen Yang , * ,
  • Min Wei
Expand
  • State Key Laboratory of Chemical Resource Engineering, College of Chemistry, Beijing University of Chemical Technology, Beijing 100029, China

Received date: 2023-07-31

  Revised date: 2023-09-27

  Online published: 2024-02-26

Supported by

National Key Research and Development Program(2021YFC2103500)

National Natural Science Foundation of China(22172006)

National Natural Science Foundation of China(22102006)

National Natural Science Foundation of China(22288102)

Abstract

The selective hydrogenation of unsaturated aldehydes is an important process of fine chemical processing that is widely used in the fields of flavor, medicine and food production, agricultural product processing, and so on. However, the hydrogenation reactivity of current catalysts still needs to be improved and further modulation of catalyst structures is needed. Three design strategies for the selective hydrogenation catalysts are summarized in this paper, modifying the electronic properties of metal active sites, enhancing the synergistic effect between the metal active sites and the electrophilic sites, and utilizing the structural effect to change the adsorption strength and hydrogenation activity of C=O bond or C=C bond. The influences of hydrogen source types, reaction solvents, temperatures and hydrogen pressures on catalytic performance are also summarized. The density functional theory (DFT) calculation, the reaction kinetic model, and the structure-activity relationship of catalysts related to the selective hydrogenation of unsaturated aldehydes are summarized. In the final section, problems, and challenges in the selective hydrogenation of unsaturated aldehydes are discussed, and some feasible solutions are further proposed.

Contents

1 Introduction

2 Design strategy of catalysts

2.1 Modifying electronic properties of metal active sites

2.2 Enhancing the synergistic effect between the metal active sites and the electrophilic sites

2.3 Utilizing the structural effect

3 the influence of reaction conditions on The catalytic performance

3.1 Hydrogen source types

3.2 Reaction solvents

3.3 Reaction temperatures

3.4 Hydrogen pressures

4 The density functional theory calculation

5 Kinetic study of the hydrogenation of unsaturated Aldehydes

6 The hydrogenation mechanism of unsaturated aldehydes

7 Conclusion and outlook

Cite this article

Xingyue Yang , Shijie Zhou , Yusen Yang , Min Wei . Efficient Catalysts for the Selective Hydrogenation of Unsaturated Aldehydes[J]. Progress in Chemistry, 2024 , 36(3) : 297 -318 . DOI: 10.7536/PC230728

1 Introduction

Unsaturated aldehydes (UAL) have special flavor and unique conjugated structure, which can be used as platform molecules for the production of various value-added chemicals, and are important raw materials and intermediates in the fields of perfume, medicine and chemical industry. Its selective hydrogenation is an important processing and conversion process of fine chemicals, which is widely used in the fields of flavor and fragrance, pharmaceutical food production, agricultural products processing and so on[1]. For example, Cinnamyl alcohol (COL) produced by the hydrogenation of the C = O bond of Cinnamaldehyde (CAL) is commonly used to prepare the vasoconstrictor antagonist cinnarizine, and Allyl alcohol (COL) produced by the hydrogenation of the C = O bond of acrolein.Crotyl alcohol (CROL) produced by hydrogenation of the C = O bond of crotonaldehyde is widely used as a plasticizer and herbicide, and nerol/geraniol produced by selective hydrogenation of citral has important applications in food and flavors[2][3][4][5]. Selective hydrogenation of UAL to unsaturated alcohols (UOL) is often used as a probe reaction to verify the catalytic performance of catalysts and to explore the effects of hydrogen spillover on the surface of catalysts[6]. The selective hydrogenation of UAL to saturated aldehyde (SAL) is mainly studied for CAL, and its C = C bond hydrogenation product hydrocinnamaldehyde (HCAL) also has high economic value and is an intermediate for the synthesis of antiviral drugs (especially HIV protease reagents). Other SALs are usually synthesized by other methods such as olefin carbonyl synthesis[7]. The hydrogenation reaction of UAL involves two parallel and successive reductions of different functional groups, that is, the selective hydrogenation of C = O bond and C = C bond in UAL molecule, and the corresponding products are generated (Fig. 1). The hydrogenation products of UAL are more, the boiling point and molecular weight are closer, the separation is more difficult, and the purification cost is increased[8]. From the thermodynamic point of view, C = O bond (715 kJ·mol-1) is more stable than C = C bond (615 kJ·mol-1), and it is difficult to carry out selective hydrogenation of C = O bond efficiently. Therefore, it is of great academic significance to study and prepare economical and efficient selective hydrogenation catalysts to improve the selectivity of UAL through kinetic pathway. At present, the catalysts used in industry still can not meet the requirements of high selectivity to the target product, and the catalyst structure needs to be regulated to improve the selectivity and production rate of the target product.
图1 不饱和醛选择性加氢的反应路径

Fig. 1 Reaction routes of selective hydrogenation of unsaturated aldehyde

In this paper, three design ideas of catalysts are summarized: (1) adjusting the electronic effect of metal sites; (2) regulating the cooperative effect of metal sites; (3) Regulatory structure effect. The external factors affecting the performance of the catalyst, such as solvent, reaction temperature, hydrogen source and hydrogen partial pressure, were discussed. The structure-activity relationship in the reaction is also discussed in detail, which provides a theoretical basis for the design and synthesis of new efficient catalysts for the selective hydrogenation of UAL. Finally, the density functional theory calculations and kinetic models related to the selective hydrogenation of UAL were summarized, the reaction mechanism was revealed, the opportunities and challenges of the selective hydrogenation of UAL in the future were discussed, and some feasible strategies to solve the current problems were proposed. It provides some ideas for better catalyst design and deeper understanding of the reaction mechanism of UAL selective hydrogenation.

2 Catalyst design strategy

The catalytic process includes the adsorption of reactants, chemical reaction and desorption of products. The adsorption of reactant molecules on the surface of catalyst is one of the key processes of catalytic reaction, and the molecular structure, bond energy and adsorption configuration are essential for the selective hydrogenation of UAL. The study of the adsorption mode of molecules on the catalyst surface is helpful to understand the reaction mechanism of the catalyst. Generally speaking, there are four main adsorption modes of UAL on the metal surface, such as :η1 mode (adsorption through the top carbonyl oxygen), η2 mode (adsorption through the C = C bond or C = O bond), η3 mode (absorption through the C = C bond and carbonyl oxygen) and η4 mode (simultaneous adsorption through both C = C and C = O bonds) as shown in Fig. 2[9]. The adsorption mode has a significant impact on the selective hydrogenation of UAL. The adsorption of C = O bond is beneficial to the formation of unsaturated alcohol, while the adsorption of C = C bond is beneficial to the formation of SAL, and the planar adsorption mode is easy to form mixed products.
图2 丙烯醛的吸附模式:(a) 通过顶部羰基O的η1-模式。(b) 通过C=C或C=O键的η2-模式。(c) 通过C=C键和羰基O以及末端羰基氧的η3-模式。(d) 涉及所有主链原子的η4-模式

Fig. 2 Adsorption modes of acrolein: (a) η1-mode (atop) via the carbonyl O. (b) η2-modes via either the C=C or the C=O bond. (c) η3-mode via the C=C bond and the carbonyl O, as well as a metallocycle via the terminal atoms. (d) η4-modes involving all backbone atoms

At present, the difficulty of realizing the high selectivity hydrogenation of UAL lies in the research and preparation of high performance catalysts. Generally, there are three methods to design, adjust and optimize the catalyst structure to improve the activity and selectivity of the catalyst.

2.1 Changing the electronic properties of the metal active site

The electronic properties of the metal active site are changed, and then the adsorption selectivity of C = C and C = O bonds is changed. In general, the electron-rich active sites prefer to adsorb C = O bonds during hydrogenation, while the electron-deficient active sites prefer to adsorb C = C bonds. Therefore, the electron density and charge of the catalyst can be adjusted by the selective addition of electron auxiliaries. Carbon supports, second-class metals, NaCl, sulfur, and organic ligands, etc., are commonly used electron auxiliaries.
Carbon materials have high specific surface area, the ability to transfer electrons and the number of active metal sites due to their abundant mesoporous characteristics and a small amount of oxygen-containing functional groups, which in turn improve the catalytic activity and C = O bond selectivity of the catalyst[10]. Three-dimensional porous carbon materials simplify the transport and diffusion paths of reaction molecules and product molecules, and increase the local concentration of UAL at the active site, thereby improving the catalytic efficiency of the catalyst[11]. Carbon nanosheets (SNS) with different degrees of graphitization, three-dimensional hierarchical porous carbon framework (3D-HPC), monodispersed nitrogen-doped hollow carbon spheres (N-HCS), and three-dimensional N-doped honeycomb porous carbon (3D-NHPC) supported Pt were used to obtain the catalysts (Fig. 3) with conversions of 87.5%, 92.7%, 99.9%, and 96.7%, respectively, and selectivities of COL of 90.5%, 91.1%, 99.9%, and 95.6%, respectively[10][11][12][13]. Tian et al. Prepared a stable Ni-C catalyst by simple pyrolysis of Ni-BTC, and found that Ni nanoparticles were uniformly dispersed in the carbon layer by characterization, which could directly carry out catalytic hydrogenation without pre-reduction, and the reaction conversion was 96. 2% and the selectivity to HCAL was 92. 5% at 120 ℃ and 2 MPa H2 for 6 H[14]. Steffan et al. Found that the catalyst supported on graphite had higher selectivity for the hydrogenation of C = C bond of citral and CAL than that supported on activated carbon[15]. Vilella et al. Compared the catalytic performance of Pt and Pt-Sn clusters supported on activated carbon powder and activated carbon felt for UAL, and found that the catalyst supported on activated carbon powder had higher selectivity for COL due to the different acidity of the support surface[16]. Li et al. Prepared interconnected porous three-dimensional graphene aerogel supported highly dispersed Ir nanoparticle composite (Ir/GA), which improved the number of active Ir sites, electron transfer ability and hydrogenation adsorption of C = O in the catalyst due to its abundant porosity, a small amount of oxygen-containing functional groups and a certain amount of Irδ- species[17]. Ir/GA achieved a COL selectivity of 83.2% at a conversion of 85.8% and a Turnover Frequency (TOF) value of 17.5 h−1.
图3 (a) 不同以碳材料作为载体的催化剂:(a) 不同石墨化程度的碳纳米片[10],(b) 三维分级多孔碳骨架[11], (c) 单分散的氮掺杂空心碳球[12], (d) 三维N掺杂蜂窝状多孔碳[13]负载Pt所制备的催化剂

Fig. 3 Different catalysts using carbon materials as support: (a) carbon nanosheets with different degrees of graphitization[10] (Copyright 2016, Royal Society of Chemistry) (b) Three dimensional hierarchical porous carbon skeleton[11] (Copyright 2018, Wiley-VCH Verlag Gmbh), (c) monodisperse nitrogen doped hollow carbon spheres[12] (Copyright 2016, Elsevier), (d) three-dimensional N-doped honeycomb porous carbon[13] (Copyright 2022, Elsevier BV) supported Pt catalysts

Nitrogen-doped mesoporous carbon materials play a key role in improving the performance of catalysts due to their abundant negative electron defect sites. Nagpure et al. Found that the nitrogen-doped mesoporous carbon (NMC) supported gold catalyst (Au-NMC-SI) with uniformly dispersed gold nanoparticles (1. 6 nm) synthesized by sol immobilization method showed high catalytic activity, and the selectivity for C = O bond hydrogenation could reach 78% and the conversion rate was 94. 2%[18]. Shaikh et al. Prepared efficient, reusable and magnetically separable nanocomposite catalyst Co3O4/N-Gr/Fe3O4-800-1[19]. Encapsulation of Co3O4 nanoparticles with porous nitrogen-rich graphitic carbon can prevent poisoning by quinoline and its hydrogenation products and improve the catalytic selectivity and durability of the catalyst. HCAL selectivity of 99% and CAL conversion of 99% were achieved in non-polar solution toluene.
When Pd is used as the active center of the catalyst, the selectivity for C = C bond hydrogenation of the catalyst supported by Pd nanoparticles on carbon materials such as carbon nanofibers, multi-walled carbon nanotubes, amorphous carbon and few-layer graphene is 98%, 91.3%, 87% and 92%, respectively, and the conversion is nearly 100%. The carbon support improves the dispersion of Pd by introducing functional groups, promotes the hydrogenation adsorption of C = C bond, and improves the selectivity of the catalyst to HCAL. By adjusting the surface chemistry of carbon nanotubes and the size of metal particles, Radhika et al. Found that the ubiquitous oxygen-containing functional groups on carbon materials significantly changed the work function of these supports[6]. Combining ultraviolet photoelectron spectroscopy and in situ near-ambient pressure X-ray photoelectron spectroscopy, the researchers unambiguously correlated the oxygen concentration at the carbon surface with the work function of the support and the interfacial charge transfer at the Pd − C interface. Combining experiments and density functional theory (DFT) calculations, it is demonstrated that carbon support-induced metal-support electronic interaction (EMSI) changes the binding energy of the relevant species and their reaction paths, generating more active Pdδ+ sites at the interface and enhancing the selectivity of the catalyst. The introduction of N into carbon-based materials can improve the hydrogenation selectivity of the catalyst. Nie et al. Synthesized N-doped reduced graphene oxide (NRGO) by simple hydrothermal treatment of graphene oxide with urea, and found that the N in NRGO could improve the dispersion of Pd NPs, and the selectivity of the obtained catalyst for C = C bond reached 95.9%[20]. Nagpure et al. Synthesized palladium nanoparticles deposited on nitrogen-doped mesoporous carbon (NMC) by a simple ultrasound-assisted method. Due to the mesoporous structure of NMC and the high N content of 11.6 wt%, the size of Pd nanoparticles is small, and the selectivity of the catalyst for C = C bond hydrogenation is 93%[21]. The Pd/OGF catalyst with layered structure was obtained by oxygen functionalization of graphite felt. The carbon support was treated by acid to form a porous structure with abundant oxygen-containing groups, which provided a channel for charge transfer between Pd and the support[22]. Electron-deficient Pdδ+ species and abundant defects on the support are beneficial to the cycling stability of the catalyst. Pd nanoparticles were supported on organosilicon-modified graphene oxide to prepare Pickering emulsion as an efficient catalyst for selective hydrogenation of CAL. Organosilicon-modified GO enhanced the lipophilicity of Pd/GO-Si surface.Therefore, the adsorption of CAL on the interface of Pickering emulsion prepared by Pd/GO-Si was promoted, and the Pickering emulsion was stabilized. With the increase of silane grafting amount, the droplets in Pickering emulsion became smaller and denser, which enhanced the bifunctional synergy of Pickering emulsion interface and showed more excellent activity[23]. Poly (N-isopropylacrylamide) (PNIPAM) was grafted on the surface of carbon nanotubes to construct a wettability-controlled surface on carbon nanotubes, which was loaded with Pd nanoparticles for the selective hydrogenation of CAL.It was found that the hydrophilic surface of the modified catalyst at 25 ℃ improved the dispersion of Pd nanoparticles on the support surface, and the hydrophilic surface changed into hydrophobic surface at 80 ℃, which was beneficial to the adsorption of reactant molecules and enhanced the catalytic activity of the catalyst[24]. In industrial applications, the interaction between noble metals (such as Pd and Pt) and sulfur-containing compounds (such as thiourea) is strong, and the active sites are occupied, which greatly reduces the activity of the catalyst. To solve the problem of catalyst deactivation by poisoning, Chen et al. Reported that Pd NPs encapsulated by mesoporous carbon matrix showed strong resistance to poisoning by thiourea and maintained high catalytic activity under prolonged exposure to thiourea[25].
In addition to the carbon support, the addition of electronic promoters such as NaCl and sulfur also has a significant effect on the catalytic performance of the catalyst[26][27]. Wang et al. Prepared a series of NaCl-modified Pt/ZrO2 catalysts, and ammonia temperature-programmed desorption and pyridine adsorption-infrared spectroscopy showed that the strength of Lewis acid on the surface of the catalysts decreased with the increase of NaCl content.The results of diffuse reflectance infrared spectroscopy and X-ray photoelectron spectroscopy of CO adsorption showed that NaCl modification increased the electron density of Pt active sites and enhanced the selectivity of the catalyst to COL[26]. However, the addition of too much NaCl will lead to the agglomeration of Pt particles, reduce the Pt sites at the interface, and reduce the performance of the catalyst. Through high-resolution X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption fine structure spectroscopy (NEXAFS), Chiu et al. Found that the incorporation of sulfur activated the Cu (111) surface, changed the adsorption configuration of crotonaldehyde, inhibited the adsorption of C = C, and converted crotonaldehyde to CROL by 100%[27,28]. Abid et al. Prepared Pt/CeO2 catalyst by using chlorine-free and chlorine-containing precursors respectively, and found that the catalytic activity of chlorine-free catalyst was lower than that of chlorine-containing catalyst, but the selectivity for CROL was higher (83.0%), which was due to the fact that after the chlorine-containing precursor catalyst was reduced at high temperature, there were still residual chloride ions in the catalyst, which formed CeOCl with the carrier and was not conducive to the production of CROL[29].
The addition of the second type of metal during the preparation of the catalyst is also an effective means of regulating the activity of the catalyst. When metals such as Fe, Co, Sn and Ga are added (Fig. 4), electron transfer occurs between the two metals, which increases the charge density of the active site, thus promoting the selective adsorption of C = O and weakening the adsorption of C = C bond of UAL, thus improving the selectivity of the catalyst to unsaturated alcohols[30,31][32][33][34]. Wang et al. Loaded Pt3Co nanoparticles on Co(OH)2 nanosheets by a facile one-pot solvothermal method, which enhanced the interaction between metal supports, resulting in electron transfer and an increase in the electron density of Pt sites; the conversion of CAL was 99.6%, and the selectivity to COL was 91.3%, with good stability[35]. Gu et al. Developed an amino acid-mediated conversion of NH2-MIL-125(Ti) to synthesize titanium oxide (PCT) porous cages. Due to the synergistic effect between Pt and Co and the influence of the specific structure of PCT, the Pt-Co/PCT catalyst has high catalytic activity. Under the conditions of 0.2 MPa H2 and 80 ℃ for 3 H, a COL selectivity of 96% at 100% CAL conversion was obtained[36]. Li et al. Successfully prepared a series of catalysts such as PtMx/SBA-15 by simple ultrasonic impregnation, and found that the improvement effect of Co on the selective hydrogenation of crotonaldehyde to CROL was much better than that of Cu, Ni and Zn[37]. When the atomic ratio of Pt/Co increases, its conversion and selectivity to CROL decrease.
图4 添加第二类金属制备催化剂:(a) 添加Fe[30], (b) Co[32], (c) Sn[33], (d) Ga[34]制备Pt基高效催化剂

Fig. 4 Adding the second kind of metal to prepare the catalyst: (a) adding Fe[30] (Copyright 2018, Elsevier Science), (b) adding Co[32] (Copyright 2018, Elsevier Science), (c) adding Sn[33] (Copyright 2020, Elsevier), (d) adding Ga[34] (Copyright 2020, Elsevier Science) to prepare the Pt-based high efficiency catalyst

Dai et al. Synthesized a series of carbon nanotube (CNT) supported bimetallic PtFe nanoparticles, and the participation of water and the synergistic effect between bimetals improved the activity and selectivity of the catalyst, with a TOF of 1200 h-1 and a selectivity of > 97%[30]. Wang et al. Obtained HPZSM-5 with hierarchical porous structure by simple desilication of conventional ZSM-5 in NaOH aqueous solution, and prepared PtFe/HPZSM-5 catalyst with Fe/Pt molar ratio of 0.25 by impregnation method[38]. The selectivity of the catalyst to COL was 87.6%, the conversion was 97.9%, and the TOF was 3.41 s-1. Compared with the traditional ZSM-5, the interaction of metal support of the catalyst after partial desilication becomes stronger, electron transfer occurs between Pt and Fe, the surface geometry of Pt nanoparticles changes, and the Pt (111) and Pt (100) crystal planes are exposed, which is beneficial to improving the catalytic activity and selectivity of the catalyst. By doping Fe into the hollow CeO2 supported Pt catalyst, Wang et al. Found that the prepared hollow CeO2 had different spatial positions and microenvironments, and the metal-support interaction was significant. After doping Fe, the electronic structure and crystal structure of the catalyst changed, and the catalyst selectivity to C = O bond and catalytic activity were improved, with a selectivity to COL of 88.9% and a conversion of 97.2%[39]. Changing the proportion of doped Fe will also affect the selectivity of the catalyst. Zhang et al. Found that when mesoporous TS-1 (MTS-1-MS) was loaded with Pt-Fe bimetallic catalysts with different Fe/Pt molar ratios, the catalytic performance of the catalysts was different[40]. When the molar ratio of Fe/Pt is 0.25, the catalyst has the best COL selectivity (89.2%) and conversion (95.7%).
The bimetallic catalyst was prepared by loading metal nanoparticles on functionalized carbon materials, which improved the catalytic activity of the catalyst. Tian et al. Prepared Pt-Co bimetallic catalyst by impregnation-reduction-deposition method after functionalizing carbon nanotubes with ammonia and hydrogen peroxide[41]. The results show that the functionalization increases the density of N and O elements in carbon nanotubes, and increases the distribution of positively charged Pt (Ⅱ) and Pt (Ⅳ) species, which is beneficial to the adsorption and activation of C = O bonds. Su et al. Synthesized a series of PtxCoy bimetallic catalysts with different Pt to Co ratios supported on oxidized carbon nanotubes (OCNTs) by a facile oleylamine reduction method[42]. The introduction of Co affects the electronic structure of Pt NPs and obviously improves the selectivity of the catalyst to COL. Among all the PtxCoy-OCNTs catalysts, PtCo3-OCNTs showed the best catalytic performance with 76% selectivity to COL at 99% conversion. Wang et al. Deposited highly dispersed Pt-Co bimetallic NPs on MWCNTs by atomic deposition method, and the prepared catalyst achieved 90.5% selectivity to COL at 94.5% conversion[32]. A series of experiments and DFT calculations show that the interaction between Pt and Co promotes the highly selective hydrogenation adsorption of C = O bond. The synergistic effect between Pt-Co and MWCNTs greatly enhanced the activity of Pt-Co bimetallic catalyst.
In addition to Fe and Co, Sn can also act as a promoter for Pt-based catalysts to improve their catalytic activity. Zhu et al. Preparation of Pt-Sn nanowires (Pt-Sn NWs) for selective hydrogenation of UAL to the desired unsaturated alcohol[43]. The optimized Pt1.5Sn nanowires exhibited high conversion efficiency (98.1%) and excellent selectivity (90.6%) for the hydrogenation of CAL, which were superior to those of Pt1.5Sn nanoparticles as well as Pt nanoparticles. Dai et al. Synthesized thin (about 3.8 nm) Pt-Sn cross-linked nanowire free-standing composites with Pt: Sn molar ratio of 3 – 4:1 by one-pot hydrothermal method[44]. The growth mechanism of CNs involves spontaneous hydrolysis and reduction of metal ions in aqueous solution followed by oriented attachment of M(OH)x(M=Pt,Sn) particulate matter reduced by polyvinylpyrrolidone during hydrothermal treatment. The PtSn/SnO2 catalyst exhibited 87% selectivity to COL due to the formation of Pt-Sn alloy in the CNs, and the free-standing thin Pt-Sn CNs combined the advantages of bimetallic alloy catalyst and nanowire structure with a high proportion of step/edge atoms with high selectivity and good durability. Wang et al. Used hierarchical porous ZSM-5 (HPZSM) loaded with 3 wt% Pt to prepare the catalyst and changed the weight ratio of Pt/Sn, and found that the yield and C = C bond hydrogenation selectivity of 3Pt 0.05Sn/HPZSM-5 catalyst were significantly improved. The characterization results showed that the surface acidity of ZSM-5 was changed, which enhanced the interaction between Pt and the support[45]. The electron transfer from Sn to Pt adjusts the surface electronic properties of Pt nanoparticles and changes the surface geometry of Pt nanoparticles, so that Pt (111) and Pt (100) crystal planes are exposed on the surface.
Hui et al. Modified the surface of PtGa alloy particles by covering an amorphous carbon thin layer. Because the amorphous carbon thin layer has the characteristics of geometric modification and permeability, electron-rich Pt sites are formed at the C @ PtGa interface, and the reactants can easily reach the Pt sites. The selectivity and conversion of the catalyst are improved, and the reusability is enhanced. After five cycles, the conversion is 88.3%, and the selectivity to COL is 91.9%[34]. Zhang et al. found that the catalytic activity and selectivity of the catalyst were related to the Fe content through a unique catalyst prepared by porous Pt-Ni NWs encapsulated by MOFs in situ (Pt-Ni NWs @ Ni/Fe-MOFs), and the optimal catalyst porous PtNi2.20NWs@Ni/Fe4-MOF obtained after optimization had high activity and selectivity, with a conversion rate of 99.5% and a COL selectivity of 83.3%[46]. In general, when a Pt-based catalyst is used to catalyze the selective hydrogenation of CAL, the hydrogenation product is generally COL. However, on the Pt3Ni@Ni32Cu(OH)2-2NWs catalyst, the selectivity to HCAL reached 87.9%, which was attributed to the limitation of the coated Ni32Cu(OH)2 film, poisoning effect, and the interaction between Pt3Ni NW and the film[47]. Specifically, the coated Ni32Cu(OH)2 film can hinder the diffusion of CAL and HCAL, thereby controlling the reaction rate and avoiding over-hydrogenation. In addition, the doped Cu2+ can poison Pt3Ni@Ni32Cu(OH)2-2NWs and inhibit the hydrogenation reaction, thereby improving the yield of HCAL.
When Pd is used as the active center of the catalyst, because Pd nanoparticles are easy to agglomerate and sinter, the introduction of second metals such as Ag and Au leads to electron transfer, which improves the dispersion of Pd, promotes the hydrogenation adsorption of C = C bonds, and improves the selectivity of the catalyst to HCAL[48][49]. The introduction of Ag into the Pd catalyst increased the selectivity of the catalyst to HCAL to 90%, and the catalyst prepared by MCM-41 supported Pd-Ag alloy achieved 98.6% HCAL selectivity and almost 100% CAL conversion due to the ultra-small particle size of Pd and the geometric and electronic effects between Pd[48]. In addition to Ag, tungsten nitride (WN), Ni, and Sn were also used as promoters for Pd catalysts. However, when Pd-Sn alloy is used as active metal, the adsorption energy of C = O on the surface of Pd3Sn(111) is lower according to DFT calculation, so the main product is COL, not HCAL. Li et al. Doped Ag into Pd-based catalysts to improve their catalytic activity and HCAL selectivity[50]. Pd-0.3Ag/MCM-41 exhibited 98.6% HCAL selectivity, > 99.9% CALD conversion. In comparison, the HCAL selectivity of Pd/MCM-41 was 84. 5%, and Pd/AC was 77. 3%, which were not as good as the catalytic performance of Pd-0. 3Ag/MCM-41. Ma et al. Synthesized a calcium-promoted palladium/carbon nanotube hybrid (Ca-Pd @ CNT) with a monolithic structure by a one-pot alginate gel process[51]. The characterization results show that the electron transfer between the added Ca promoter and Pd NPs can change the electronic structure of Pd species and form partially positively charged Pdδ+ species on the surface of Pd NPs in Ca-Pd @ CNTHCl-2 H. The as-prepared Ca-Pd @ CNTHCl-2h exhibited excellent hydrogenation activity with 99.9% CAL conversion and 86.4% HCAL selectivity. Cattaneo et al. Prepared a series of Au-Pd nanoparticles with different Au: Pd molar ratios by sol immobilization technique to catalyze the selective hydrogenation of CAL under mild reaction conditions[49]. Au50Pd50/TiO2(Au:Pd molar ratio of 1:1) has the best CAL hydrogenation activity, but the monometallic Pd/TiO2 has the strongest selectivity for HCAL. The structure of the catalyst was characterized, and Fourier transform infrared spectroscopy (FTIR) analysis showed that the presence of adsorbed carbonyl surface species in the used catalyst material was related to Pd leaching, which was the main cause of catalyst deactivation. The effect of calcination on the Au-Pd/TiO2 of the best performing catalyst was studied in the range of 110 – 400 ° C, and a direct correlation between the increase in calcination temperature and catalyst stability and selectivity was observed. Li et al. Used Pd, Ir, and bimetallic Pd-Ir nanoparticles to uniformly disperse and deposit on the SiC surface by a simple impregnation method, respectively[52]. The prepared Pd/SiC catalyst effectively hydrogenates CAL to HCAL at room temperature and atmospheric pressure, and it is found that the activity of Pd/SiC can be further improved by introducing Ir component. The activity of Pd-Ir catalyst is related to the molar ratio of Pd/Ir, which further confirms the synergistic effect between Ir and Pd. Because of the electron transfer between Pd and Ir, the electronic structure of the catalyst is adjusted and the catalytic performance of the catalyst is improved. Dai et al. Synthesized CeO2 supported ternary RuSnB catalyst by incipient wetness impregnation method and applied it to CAL hydrogenation with pure water as solvent[53]. The catalyst gave 96.0% conversion and 93.9% selectivity to COL. The structural characterization results show that the Ru site is incorporated into the SnOx matrix, and the electronic interaction is enhanced by Ru and B addition. The optimal Ru2Sn1B1/CeO2 catalyst can provide the best H2O activation ability, thus accelerating the reaction rate of H2O participating in hydrogenation.
In addition to the addition of the second type of metal to the noble metal, the bimetallic catalyst formed by doping transition metal and transition metal also has good catalytic activity for the selective hydrogenation of UAL. The loading of Cu, Ni bimetallic alloy particles on the redox graphene and functionalized graphene oxide nanocomposite showed good catalytic activity in the selective hydrogenation of CAL, with 100% selectivity and 82% conversion for HCAL for Ni-Cu @ RGO and 59.8% selectivity and 85% conversion for COL for Cu (0) -Ni (0) -AAPTMS-GO[54]. The high activity of Ni-Cu @ RGO is due to the synergistic effect between Ni and Cu, and the electron transfer between Ni and Cu adjusts the electronic structure and geometry of the active metal. The main product of Cu (0) -Ni (0) -AAPTMS-GO is COL instead of HCAL, which may be due to the synergistic effect between Co and Ni and the uniform distribution of Cu-Ni bimetallic alloy particles. Yang et al. Prepared two intermetallic compounds (IMCs:CoIn3 and CoGa3) by structural topological transformation of layered hydroxide precursor, and the hydrogenation selectivity of COL over CoGa3 catalyst was 96%, which was significantly higher than the selectivity of 80% over CoIn3 catalyst and 42% over monometallic Co catalyst[55]. Electron transfer from Ga (or In) to Co was found by characterization, leading to the formation of Co-Ga (or Co-In) coordination. From FTIR measurements and DFT calculations, it is shown that Ga (or In) In IMCs acts as an active site, promoting the selective adsorption of C = O. Mohire et al. Synthesized Ni-Cu @ RGO catalyst using a simple solvothermal method[56]. After adjusting the ratio of Ni: Cu, the catalyst prepared by 5 wt% Ni-Cu (1:1) supported on reduced graphene oxide showed 100% selectivity to HCAL. Prakash et al. Prepared a NiCu bimetallic catalyst supported on TiO2-P-25 by chemical reduction using glucose as a reducing agent[57]. Temperature-programmed reduction and diffuse reflectance spectroscopy studies indicate the formation of Ni-Cu alloy, and the changes in XPS binding energy values for the Ni 2p1/2 and Ni 2p3/2 levels indicate that Cu tends to increase the electron density around Ni, which hinders the C = C bond adsorption of CAL. Bimetallic catalysts show higher catalytic activity at lower temperatures than the corresponding monometallic catalysts.

2.2 Synergism of the enhanced metal site

The ability of the metal surface to dissociate hydrogen and the ability of the adjacent sites to adsorb substrate are regulated by enhancing the synergy between the metal active sites and electrophilic sites, which enhances the hydrogenation activity of the catalyst for C = O and thus improves the selectivity for unsaturated alcohols. In general, metal active sites can be generated by a variety of ways, such as local positive charges caused by electron transfer in metal catalysts, oxygen vacancies in reducible metal oxides, Lewis acid sites in supports, metal nodes in MOF materials, and metal boride amorphous alloys.Electrophilic sites can also be generated by adding additives such as alkali to improve the catalytic activity of the catalyst[58][59][60][61][62]. When the catalyst is doped with Re, Ge, Zn, Sn, etc., the doped second metal often forms oxides such as ReOx, ZnOx, etc., which form electrophilic sites in the vicinity of Pt sites and facilitate the C = O bond adsorption of UAL. Pt50Re50/rGO has the best catalytic activity compared to monometallic Pt/rGO, Pt50Sn50/rGO, Pt50Er50/rGO, Pt50La50/rGO, Pt5OY50/rGO and monometallic Re/rGO[63]. Pt-Re/rGO catalysts with different metal ratios were prepared, and it was found that Pt50Re50/rGO showed the highest catalytic performance under the optimal reaction conditions of 120 ℃ and 2.0 MPa H2 for 4 H, with a conversion of 94% and a selectivity for C = O bond of 89%. The good catalytic performance may be due to the successful synthesis of Pt-Re alloy, which improves the metal particle size distribution and dispersion of Pt. The ReOx can selectively adsorb C = O to the active sites near the metal. The mesoporous, macroporous and aromatic graphite structure of rGO is conducive to the adsorption of CAL, which is conducive to improving the hydrogenation rate of the catalyst and the selectivity for COL. Cao et al. Reported the use of a metal alloy consisting of a small amount of Pt single atoms in Cu nanoparticles to selectively promote the hydrogenation of C = O bonds in UAL[64]. The surface of the dispersed CuPtx catalyst is mainly composed of a thin copper oxide film. The surface of Cu-based catalyst tends to bond UAL through the terminal oxygen atom, which is beneficial to the selective adsorption of C = O bond, while Pt sites promote the activation and cleavage of molecular hydrogen, which improves the catalytic activity of the catalyst through synergistic effect. Weng et al. Used the atomic deposition method to grow a silicon oxide film on the surface of an alumina-supported Pt catalyst to modify it, and found that the silicon oxide film could greatly improve the stability of Pt nanoparticles, provide more Brønsted and Lewis acid sites, improve the synergistic effect with the metal surface, and increase the selectivity to C = O bonds to 85%[59]. The components of the CeO2-ZrO2 composite have a great influence on the surface structure of the catalyst. Wei et al. Prepared a CeO2-ZrO2 composite support with a high surface area micro/mesoporous structure by a low-temperature liquid phase method, and found that when the molar ratio of Ce/Zr is 1:The selectivity of the as-prepared Pt/CeO2-ZrO2 catalyst is the best, up to 89%, because of the synergistic effect between the electron-rich metal Pt species and the Ce3+ species and oxygen vacancies on the support surface, which promotes the selective adsorption of C = O bonds and improves the reactivity of the catalyst[58]. Liu et al. Reported the preparation of Pt/(WO3-300H2) from Pt supported on pre-reduced WOx[65]. The prepared catalyst showed excellent performance with 93.4% selectivity and 91.1% conversion for COL at 100 ° C. The pre-reduction of WO3 makes more defect sites on the support, which is beneficial to the reduction and dispersion of Pt on the support. The defect-rich WOx supported Pt catalyst has more surface oxygen vacancies and better H spillover ability, which is beneficial to the adsorption hydrogenation of C = O bond. The introduction of FeOx and ReOx into the Ir catalyst as electrophilic sites to preferentially adsorb C = O bonds improves the catalytic performance of the catalyst. Tamura et al. Prepared a Ir-ReOx/SiO2 catalyst with 99% conversion and 96% COL selectivity at 0.8 MPa H2 pressure and low temperature of 303 K.Their study showed that the introduction of ReOx promoted the adsorption of C = O and the heterolytic cleavage of H2 into H+ and H species, which improved the catalytic activity and selectivity of the catalyst[66]. Martinez et al. Prepared a Ir-FeOx/SiO2 catalyst, which increased the selectivity of the catalyst to COL from 57% to 83% compared with the Ir/SiO2[67]. Zhao et al. Found that the hydrogenation rate of Au-Ir/TiO2 was 5 times higher than that of Au/TiO2, which may be due to the strong interaction between Au and Ir and the electron transfer of Ir, which promoted the selective adsorption hydrogenation of C = O bond[68]. Lin et al. Prepared Ni-Ir/TiO2 catalyst by continuous deposition method, and its catalytic activity was 93.4 h-1, which was about four times higher than that of Ni/TiO2, with a conversion of 97.8% and a C = O bond hydrogenation selectivity of 95.4%[69]. Through the introduction of Ir, the size of Ni nanoparticles is reduced, the electronic structure of Ni surface is changed, the interaction between Ni-Ir is enhanced, the hydrogen overflow effect is enhanced, and the catalyst performance is improved. Zhao et al. Used SBA-15 as a support and FeOx as a modifier to construct a 0.73%Au-4.13%FeOx@SBA-15 catalyst through a synergistic effect[70]. It showed better catalytic activity than 0.71%Au/SiO2, 0.69% Au @ SBA-15 and 0.71%Au/bulk-FeOx, and higher selectivity under the same conditions. The interface between the introduced FeOx and Au forms H2 dissociation, which makes 0.73%Au-4.13%FeOx@SBA-15 have higher catalytic activity.
Zahid et al. Synthesized a series of Pt-Co intermetallic compound nanoparticles supported on MIL-101 (Cr) MOF (3% Pty% Cr/MIL-101 (Cr)) by using a polyol reduction method[60]. It has higher selectivity than monometallic Pt/catalyst under mild conditions. 3% Pt3% Co/MIL-101 (Cr) can achieve 95% conversion and 91% COL selectivity. The Pt-Co intermetallic compound adjusts the electronic structure of Pt active sites and enhances the Lewis acidity of the catalyst surface, and Lewis acidic sites are formed at the MOF-metal interface, which is beneficial to the adsorption hydrogenation of C = O bonds.
Lin et Al. Used a layered hydrotalcite-like material, Mg3Al1−xFex, to prepare supported Ir catalysts with different amounts of Fe and Al species as supports[31,71]. It was found that during the variation of X from 0 to 1, the CAL hydrogenation rate reached a maximum at about X = 0.25, while the selectivity to COL increased monotonically from 44.9% to 80.3%. Through a series of instrumental characterizations, it was found that the improvement of catalyst performance was due to the electron transfer from Fe2+ in the Mg3Al1-xFex support to Ir particles and the modification of surface geometry. Gao et al. Used CoAl-HTs to support Pt to obtain Pt/CoAl-HTs with a COL selectivity of 89%[72]. The water-mediated hydrogen exchange reaction enhances the catalytic activity of the catalyst. Gao et al. Prepared a Pt/CoAl-LDHs catalyst and achieved 93.6% selectivity at 95.4% conversion[73]. The Co2+-O2--Al3+ interface formed in CoAl-LDHs is helpful to strengthen the metal-support interaction and form electron-rich Ptδ sites. Electron-rich Ptδ- provides better activation ability for H2, while it also exposes abundant unsaturated sites, which optimizes the adsorption mode of CAL and is beneficial to C = O hydrogenation. Mateen et al. Prepared Ir1@CeO2 three-dimensional spherical catalyst based on emulsion-based molecule-nanoparticle self-assembly[74]. The strong metal-support interaction makes the charge transfer from Ir to CeO2, so that Ce4+ is partially reduced to Ce3+, and a new Irδ+-O2--Ce3+ interface is formed. The electrophilic Ir species at the interface site preferentially undergoes adsorption hydrogenation of the C = O bond, resulting in a catalyst with a selectivity of about 100%. He et al. Prepared Ir/H-MoOx for selective hydrogenation of CAL to COL, and the selectivity reached 93% under the condition of conversion > 99%, which showed great advantages over Ir/CNT 68% selectivity and Ir/TiO264% selectivity[75]. The C = O bond hydrogenation adsorption is promoted by the active Irδ species on H-MoOx due to the strong electronic metal-support interaction. At the same time, He et al. Found that the prepared Ir/H-MoOx could be used as a general catalyst for the selective reduction of substrates with reducible groups, and could achieve more than 90% of the C = O bond hydrogenation selectivity in the selective hydrogenation of citral, furfural, 1-bromo-4-nitrobenzene and 4-nitrobenzoic acid. Zhang et al. Prepared Co @ BN/BN model catalyst by ball milling under NH3 atmosphere, and the Co particles encapsulated by BN layer were dispersed on the defective BN support[76]. The thickness and crystallinity of the BN shell were adjusted by controlling the temperature (600 ~ 900 ℃). The H2 molecules are dissociated into hydrogen atoms by the encapsulated Co particles through the defect sites, and then selectively hydrogenated by the overflow of hydrogen atoms on the BN surface. Li et al. Reported a low-temperature synthesis strategy to create atomically dispersed palladium atoms anchored on defective hexagonal boron nitride (h-BN) nanosheets[77]. Nitrogen-containing B vacancies can provide stable anchor sites for palladium atoms. The catalyst showed exceptional efficiency in CAL selective hydrogenation, as well as excellent recyclability, sinter resistance, and scalability. Zhang et al. Adjusted the reduction temperature to obtain Co/p-BN-T (T = 300, 500, 800 ° C) catalysts with different surface B − O and N − H species contents[78]. The content of surface B − O species decreased with the increase of reduction temperature, and the metal-support interaction weakened. Co/p-BN-500 had the best catalytic performance, and showed good conversion and selectivity for CAL hydrogenation to COL.
The amorphous alloy catalyst has a unique isotropic structure and coordination unsaturated sites, which has a good promotion effect on the selective hydrogenation of UAL. Mo et al. Prepared one-dimensional amorphous alloy Co-B NWs using lamellar liquid crystal as a template[61]. The results of low temperature plasma treatment show that the plasma treatment with appropriate power and time has little effect on the morphology and amorphous structure of Co-B NWs, but it can increase the relative content of Co0 on the surface, optimize the type of surface active sites, and improve the specific surface area, which is beneficial to the improvement of catalytic hydrogenation performance. Hidalgo-Carrillo et al. Found that the selectivity and reaction rate of COL were significantly improved when NaOH or KOH was added to the solution, which was due to the fact that OH inhibited the adsorption of C = C bond and Na+ and K+ as Lewis acid sites promoted the adsorption of C = O bond[62].

2.3 Use Structure Effect

In addition to the electronic effect and synergistic effect, the structural effect also has a great influence on the selectivity and activity of the catalyst. The hydrogenation performance of the catalyst is controlled by means of interaction with the surface environment of the metal site to hinder the adsorption of C = C, controlling the exposure of different crystal faces and the size of crystal grains, synthesizing porous materials and core-shell structures, and adjusting the type and coverage of ligands. The selectivity and hydrogenation activity of Pt catalysts for unsaturated alcohols are generally in the order of Pt nanocuboctahedra > Pt nanocubes > Pt nanospheres, and the order of crystal planes is Pt (111) > Pt (100) > circular (stepped) surface[79]. The most stable adsorption configuration of CAL molecules on the catalyst surface depends on the size of the nanocluster (Fig. 5),It usually adsorbs on Pt6 and Pt10 in a η2(CC) configuration,The adsorption configurations on Pt13 and Pt14 are η2-π(CO) and η2-diσ(CO)[9]. Pt NPs with smaller particle size expose more low coordination sites, which is more beneficial to improve the catalytic activity, but the exposure of more low coordination sites is beneficial to C = C bond adsorption, which reduces the selectivity to COL. Hu et al. Modified Pt nanoparticles with FeOx by atomic layer deposition (ALD), and the selectivity to COL increased from 45% for the single Pt-based catalyst to 84% for the Pt-based catalyst after 30 ALD cycles[80]. FeOx precisely covers the low coordination site of Pt, creating a Pt-FeOx interface. The interaction between Pt nanoparticles and iron oxide improves the selectivity of the catalyst to COL. Cui et al. Used a two-solvent strategy to construct ultrafine platinum nanoparticles (Pt NPs) supported on the inner and outer surfaces of halloysite nanotubes (HNTs), and the best catalyst with an average Pt particle size of 2.98 nm showed excellent catalytic activity for CAL hydrogenation to COL, with a conversion of 94.1% and a COL selectivity of 95.1%[81]. The selectivity of the catalyst to COL and HCAL can be switched by changing the molar ratio of Ni/ (Fe + Ni) in the NixFe1-xAl2O4+δ spinel composite metal oxide. Xin et al. Found a selectivity of 92.2% for COL over Pt/FeAl2O4+δ[82]. The selectivity for HCAL over Pt/NiAl2O4+δ was 80.7%. The selectivity to COL is attributed to the modification of FeOx species on Pt/FeAl2O4+δ and the exposure of more Pt (111) facets. While the selectivity to HCAL is attributed to the exposure of more low-coordinated active sites on the catalyst surface. Peres et al. Grown Pt concave nanocubes with Pt (110) crystal face on FLG support, and found that compared with Pt (111) crystal face, the exposure of Pt (110) crystal face resulted in higher COL selectivity and activity of the catalyst, with a selectivity of 88% at 95% conversion[83].
图5 在Pt13/CeO2(111)的金属-载体界面处,巴豆醛中C=O键优先吸附[9]

Fig. 5 Preferential adsorption of C=O bond in crotonaldehyde at metal-carrier interface of Pt13/CeO2(111)[9]. (Copyright 202020, American Chemical Society)

To study the effect of Pd particle size on catalyst performance, Jiang et al. Prepared a series of catalysts using activated carbon, SiO2, TiO2, γ-Al2O3, SiC and graphene oxide as supports, and Pd particles with different sizes could be directly obtained by using different Pd precursors[84]. On reducible TiO2 support, Pd particle size can also be indirectly controlled by the influence of strong metal-support interactions. The incorporation of Ag metal into the γ-Al2O3 supported palladium catalyst indirectly obtained Pd particles of different sizes by Ag deposition on large Pd NPs. It was found by DFT calculation that the smaller Pd particle (Pd4) is favorable for C = C bond adsorption of CAL, while C = O is more easily adsorbed on the exposed Pd (111) crystal face of the larger particle surface. When Yin et al. Used Pd@MIL-101(Cr)-NH2 in the selective hydrogenation of furfural, furfural was completely converted to tetrahydrofurfuryl alcohol after 6 H of reaction under mild conditions of 40 ℃ and 2 MPa H2, and its selectivity was close to 100%, showing more excellent catalytic performance than that reported in the literature[85]. This is due to the highly uniform dispersion of Pd nanoparticles, as well as the coordination and π-π interaction between the catalyst support and Pd nanoparticles. Wang et al. Prepared PtFe/HPZSM-5 catalyst with Fe/Pt molar ratio of 0.25 using the impregnation method[38]. The selectivity of the catalyst to COL was 87.6%, the conversion was 97.9%, and the TOF was 3.41 s-1. Compared with the conventional ZSM-5, the interaction of the metal support of the partially desilicified catalyst becomes stronger, electron transfer occurs between Pt and Fe, and the surface geometry of Pt nanoparticles changes, exposing more Pt (111) and Pt (100) crystal planes. Wang et al. Used hierarchical porous ZSM-5 (HPZSM) loaded with 3 wt% Pt to prepare the catalyst and changed the weight ratio of Pt/Sn, and found that the yield and C = C bond hydrogenation selectivity of 3Pt 0.05Sn/HPZSM-5 catalyst were significantly improved. The characterization results showed that the surface acidity of ZSM-5 was changed, which enhanced the interaction between Pt and the support[45]. The electron transfer from Sn to Pt adjusts the surface electronic properties of Pt nanoparticles and changes the surface geometry of Pt nanoparticles, so that Pt (111) and Pt (100) crystal planes are exposed on the surface. The catalyst was prepared by loading Pt on manganese oxide octahedral molecular sieve (OMS-2), and DFT calculations showed that the H2 dissociated on the Mn site at the (001) crystal plane of OMS-2, and water participated in the dissociation of the H2. The reason for the high C = O bond selectivity of the catalyst is that the C = O bond is preferentially adsorbed at the edge of the Pt (211) step[86].
O 'Brien et al. Studied the effect of Pd nanoparticle size on the selectivity of acrolein hydrogenation, and found that nearly 100% selectivity to the target product AA was observed on Pd (111)[87]. No AA formation was observed on 4 – 7 nm sized nanoparticles, while a large amount of AA was formed on more than 12 nm Pd nanoparticles in the temperature range around 250 K. The introduction of Ag into the Pd catalyst increased the selectivity of the catalyst to HCAL to 90%, and the catalyst prepared by MCM-41 supported Pd-Ag alloy achieved 98.6% HCAL selectivity and almost 100% CAL conversion due to the ultra-small particle size of Pd and the geometric and electronic effects between Pd[48]. In addition to Ag, tungsten nitride (WN), Ni, and Sn were also used as promoters for Pd catalysts. However, when Pd-Sn alloy is used as active metal, the adsorption energy of C = O on the surface of Pd3Sn(111) is lower according to DFT calculation, so the main product is COL, not HCAL. Li et al. Doped Ag into Pd-based catalysts to improve their catalytic activity and HCAL selectivity[50]. Pd-0.3Ag/MCM-41 exhibited 98.6% HCAL selectivity, > 99.9% CALD conversion. In comparison, the HCAL selectivity of Pd/MCM-41 was 84. 5%, and Pd/AC was 77. 3%, which were not as good as the catalytic performance of Pd-0. 3Ag/MCM-41. Wang et al. Synthesized nanorods, nanoparticles and triangular nanosheets of La2O2CO3 by a simple hydrothermal method, and supported Pd to prepare the catalyst. The selectivity of the Pd/La2O2CO3-NPs to HCAL was as high as 96%, but the conversion was very poor, only 17%[88,89]. Pd/La2O2CO3-NRs showed high catalytic activity with 86% HCAL selectivity at 3 H of 100% CAL conversion. The characterization shows that there are more basic sites (La3+-O2-) on the surface of the Pd/La2O2CO3-NRs, which improves the hydrogenation activity of the catalyst. Compared with Pd/La2O2CO3-NR, Pd/La2O2CO3-TNS exhibited significantly superior catalytic activity in CAL hydrogenation with a TOF of 41 238 h−1, whereas the TOF of Pd/La2O2CO3-NR was 1251 h-1. Pd/La2O2CO3-TNS achieved a high conversion of 90% and COL selectivity of 92% after 0.5 H of reaction time. This enhanced activity of Pd/La2O2CO3-TNS is due to the strong metal-support interaction. The oxygen-rich La2O2CO3-TNS exposed (001) facets on the surface facilitate the charge transfer between Pd nanoparticles and triangular La2O2CO3 nanosheets, increasing the electron density of Pd. In addition, the electronic state of the Pd/La2O2CO3-TNS catalyst can enhance the hydrogen adsorption and activation, thereby improving the hydrogenation activity. The catalytic performance of Ir catalyst is greatly affected by the particle size, morphology and support, and the performance is better when the diameter of Ir cluster is in the range of 0. 6 ~ 1.6 nm[90]. Li et al. Prepared a composite of highly dispersed Ir nanoparticles (Ir/GA). Due to the diameter of Ir clusters, the number of active Ir sites in the catalyst was increased, which promoted the hydrogenation adsorption of C = O and reached 83.2% COL selectivity at 85.8% conversion[17]. Prakash et al. Used nickel acetate as precursor, hydrazine hydrate (HH) and glucose (GL) as reductants for reduction, and prepared 15 wt%Ni/TiO2-P-25 catalysts by four different methods including direct impregnation (IM), urea deposition-precipitation (DP) and chemical methods[57]. The results showed that Ni/TiO2-HH had the highest activity and selectivity for COL, and Ni/TiO2-DP had the highest selectivity for HCAL. The report shows that the preparation method can affect the Ni crystal size and the nature of the metal-support interaction, thereby affecting the performance of the catalyst.
Materials such as MOF, zeolite porous materials and core-shell structures are prepared to limit the particle size of metal nanoparticles and improve the selectivity of catalysts by steric hindrance. Encapsulation of Pt nanoparticles into UiO-66's layered defect metal-organic framework, zeolite framework-1, and yolk-shell metal-organic framework with controllable spatial positioning (Fig. 6) adjusts the appropriate pore structure, which is conducive to the exposure of Lewis acidic sites and the enhancement of metal-support interaction.The adsorption of reactants and the desorption of products were optimized to improve the selectivity of the catalyst to COL. The selectivity of the obtained catalyst was 85%, 98.7% and 98.2%, and the conversion was 76%, 99.8% and 97%, respectively[91][92][93]. The dispersion of Pt nanoparticles in the porous structure also affects the selectivity of the catalyst. Geng et al. Prepared three series of UiO-66 by using sol-gel method, wet impregnation and in situ encapsulation. The position and dispersion degree of Pt species on the support affect the interaction between the metal and the support. Pt/U-720 with more MLD has higher catalytic activity and COL selectivity[95]. Yuan et al. Used conjugated microporous and mesoporous polymers with iron porphyrin to prepare MIL-101 @ Pt @ FeP-CMP[96]. The as-prepared MIL-101 @ Pt @ FeP-CMP is hydrophobic and porous, which can enrich the reactants, and has iron sites on the surface to activate the C = O bond.MIL-101 @ Pt @ FeP-CMP can achieve high turnover frequency (1516.1 h−1) with 97.3% selectivity to COL at 97.6% conversion. Liu et al. Prepared iridium nanoclusters (Ir NCs) confined in hollow MIL-101 (Fe), which exhibited 93.9% conversion and 96.2% selectivity in the catalytic selective hydrogenation of CAL to CAL[94]. The electron transfer from Ir to MIL-101 (Fe) and the Lewis acid sites in MIL-101 (Fe) are beneficial to the selective hydrogenation adsorption of C = O bond and improve the catalytic performance of the catalyst. Liu et al. Synthesized Rh nanoclusters encapsulated in MIL-101 (Cr) by a two-solvent method to catalyze the selective hydrogenation of CAL to HCAL, achieving 98% selectivity and 98% conversion under mild conditions[97]. Zhang et al. Designed a hollow flower-like Zr-MOF double-shelled nanosphere to confine Pt nanoparticles, decorated with functional groups such as -H, -NH2, -Br, and -NO2 on the shell, and used it as a catalytic microreactor to catalyze the selective hydrogenation of CAL C = O bonds[98]. The double shell of the outer decoration can not only be used as a stabilizer to ensure the high dispersion of Pt nanoparticles, but also the medium-sized channel formed on the shell can be used to control the direction of CAL entering the reactor. Lan et al. Developed a core-shell structure of Pt@SnOx/SiO2 for the selective hydrogenation of acrolein[99]. The structure significantly promoted the synergistic hydrogenation effect of Pt and SnOx, showing high activity and selectivity for allylic alcohols. Patil et al. Synthesized a 30 wt% Ni @ OCNT catalyst, in which Ni nanoparticles were precisely dispersed on the external and internal surfaces of OCNT[100]. The steric hindrance effect of Ni NPs loaded in the OCNT channel is obvious, and the electronic configuration and chemical properties of Ni NPs are changed, which leads to the hydrogenation of CAL to HCAL. In addition, the highly dispersed Ni-NPs outside the OCNT surface also contribute to the selective formation of HCAL. The Ni @ OCNT catalyst had 89% selectivity at 96% CAL conversion. Cui et al. Successfully prepared ZIF-67 and ZIF-67@SiO2-CPTEOS and applied them in the environmentally friendly transfer hydrogenation of CAL to COL[101]. However, the CAL transfer hydrogenation over ZIF-67 is actually a homogeneous reaction in a heterogeneous system, which has a serious recovery problem, so hydrophobic modified core-shell ZIF-67@SiO2-CPTEOS were prepared, and the active sites were restricted in the nanoreactor. Therefore, ZIF-67@SiO2-CPTEOS has better catalytic performance and stability. Conversion of CAL was > 99% and selectivity to COL was 93.25% at 453 K and 1 MPa N2. According to DFT calculations, the activation energy barrier of Co-N in ZIF-67 is lower than that of metallic Co dissociated by Co(NO3)2·6H2O, which illustrates the higher catalytic activity of CoN than metallic Co.
图6 将金属纳米粒子封装到(a) UiO-66的分层缺陷金属有机骨架[91], (b) 沸石骨架-1[92], (c) 可控空间定位的蛋黄壳金属有机框架[93], (d) MIL-101(Fe)[94]内制备催化剂

Fig. 6 The catalyst is prepared by packing metal nanoparticles into (a) UiO-66[91] (Copyright 2022, American Chemical Society), (b) silicalite-1 framework[92] (Copyright 2022, Elsevier), (c) in Yolk-Shell MOFs[93] (Copyright 2020, Wiley-VCH Verlag), (d) MIL-101(Fe)[94] (Copyright 2022, Elsevier)

Tuning the anions bound at the Pt nanoparticles can reconstruct the surface state of the catalyst, thereby tuning the selectivity of C = C and C = O bonds. Tao et al. Additionally introduced sodium hydroxide and sodium formate into Pt-coated titania (Pt @ P25) to remodel the surface by competitive exchange with carbonate and bicarbonate anions weakly bound to the surface, resulting in 94.7% and 97.6% hydrogenation selectivity of the catalyst for C = O and C = C bonds, respectively[102]. Zahid et al. Functionalized MOF with amine and uniformly anchored Pt nanoparticles by polyol reduction method[103]. The as-prepared 3-Pt/MOF-NH2(x) catalyst retained the intrinsic characteristics of the MOF-NH2(x) support, such as crystallinity, surface area, pore structure, and surface acidity. The prepared catalyst achieved 78.9% C = O bond hydrogenation selectivity at 72.3% conversion. The improvement of the catalytic performance of the catalyst is due to the fact that the doping of N atoms helps the small Pt nanoparticles to be highly dispersed and stabilized, and the electron density is adjusted by adjusting the cooperative effect of the empty d-orbital of the active Pt particles, resulting in more interfacial electrophilic sites and nucleophilic sites. The steric hindrance effect induced by the mesopores of MOFs is also beneficial to the hydrogenation adsorption of C = O bonds. Vu et al. Stabilized PtFe NPs by carboxylate surface ligands, and the selectivity of C10F-Fe0.33Pt0.67 to COL was as high as 94%. They found that the nanoparticles stabilized by longer chain fluorocarboxylate ligands promoted the adsorption of C = O bonds and improved the catalytic ability of the catalyst, but if hydrocarbon ligands were used as stabilizers, they would occupy the active sites, resulting in a significant decrease in catalyst activity[104]. Sole et al. Prepared a novel surfactant triazolyl sulfide as a ligand, and used a two-step strategy of copper azide-alkyne cycloaddition (CuAAC) and thiol-ene click reaction with [Ir(COD)Cl]2 as a metal precursor for the in situ preparation of Ir NPs as a catalyst for the selective hydrogenation of CAL, which achieved 92% conversion and 91% COL selectivity[105]. Tan et al. Synthesized ZnAl-hydrotalcite supported cysteine-terminated Au25 nanoclusters for CAL selective hydrogenation[106]. The reaction was carried out at 130 ° C and 15 atm H2 with 98.3% conversion and 95.4% COL selectivity for the Au25/ZnAl-300 catalyst. Ye et al. Kinetic test results showed that crotonaldehyde had a higher adsorption equilibrium constant and a higher intrinsic reaction rate constant on the Ir/BN-C3 catalyst compared with Ir/BN-C5, indicating that the presence of cetyltrimethylammonium bromide (CTAB) effectively enhanced the adsorption of crotonaldehyde on the catalyst surface.On the other hand, the Ir-CTAB interfacial active sites have higher intrinsic activity than Ir-Ir active sites, thus verifying that the Ir-CTAB interface plays a key role in enhancing reactant adsorption and promoting the reaction rate[107]. Peres et al. Grown Pt nanocubes with (110) crystal plane on FLG support in situ by wet chemical method, and found that in the presence of octadecylamine as a stabilizer,The nanocubes were directly grown on a few layers of graphene support, and the organic ligand was beneficial to stabilize the morphology of the nanoparticles. When the ligand content was low, the selectivity of the catalyst to COL was improved, but when the ligand content was high, the performance of the catalyst was affected[83]. Peres et al. Also found that the exposure of the (110) crystal face resulted in higher selectivity and activity of the catalyst compared to the (111) crystal face, with a COL selectivity of 88% at 95% conversion. Schr Schröder et al. Measured the isothermal reaction kinetics by multiple molecular beam techniques, and simultaneously monitored the evolution of the reaction intermediates, the chemical composition of the ligand layer, and their dynamic changes by operational infrared reflection-absorption spectroscopy (IRAS)[108]. Three types of ligand precursors were employed to functionalize the Pd (111) surface: allyl cyanide, 2-methyl-2-pentenal, and acetophenone. It is shown that the interaction of acrolein with the ligand-modified surface can be effectively tuned by changing the functional group and chain length of the ligand to produce the desired reaction intermediate (propenoxy species) and finally the target product AA. Ligand modification of the catalyst surface, Farrag et al. Proposed a new synthetic method to prepare monodisperse palladium clusters protected by N-acetyl-l-cysteine[109]. This approach allowed the generation of very closely sized palladium clusters (~ 1 nm) anchored to the ligand via thiol groups, and the metal ligand ratio (M/L) and average chemical formula of the prepared clusters were calculated by thermogravimetric analysis and elemental analysis. The synthesized palladium clusters were supported on α-Fe2O3 at a percentage of 1 wt%. The synthesized palladium clusters have special properties and do not block the mesopores of the iron oxide support after loading. The as-prepared catalysts (1 wt%Pdn/α-Fe2O3, 1 wt%Pdn(NALC)m/α-Fe2O3, and Pdn(NALC)m) exhibited high catalytic activity for the reduction of CAL at room temperature with 100% HCAL selectivity. Murata et al. Studied the role of carbon monoxide molecules as a structural capping agent for metal nanoparticles[110]. The shape and surface structure of the Pd particles were controlled by reducing the supported Pd precursor with CO. It was found that the reduction of Pd nanoparticles supported on SiO2 by CO promoted the exposure of step sites and the formation of spherical and concave tetrahedral Pd particles on carbon. The conventional H2 reduced Pd particles exhibit a flat shape. With the increase of the proportion of step sites on Pd nanoparticles, the C = C bond of CAL is preferentially adsorbed, and the hydrogenation activity and selectivity of HCAL increase. The liquid-liquid interface microenvironment also has a regulatory effect on the selectivity of the reaction. Zhang et al. Controlled the assembly of tubular catalyst particles at the narrow internal interface of Pickering emulsion droplets, and found that the selectivity of the catalyst particles at the internal interface of water droplets to C = O bonds reached 92.0% ~ 98.0%[111].

3 Effect of Reaction Conditions on Catalytic Performance

3.1 Hydrogen source species

Hydrogen is usually used as the hydrogen source in the hydrogenation reaction, while when proton hydrogen sources such as alcohols or hydrogen storage materials such as ammonia borane are used as hydrogen donors, the reaction is mostly carried out at atmospheric pressure, and the requirements for equipment are low, which reduces the risk of the reaction. In addition, it was found that some reactions using other hydrogen sources also had higher C = O bond hydrogenation selectivity and hydrogenation rate. The diversity of hydrogen sources provides a new way to improve the selectivity of the reaction (Fig. 7). Yamada et al. Used hydrogen storage alloy Mg2Ni as a hydrogen source and found that its selectivity depended on the release rate of stored hydrogen, and that 100% selectivity to unsaturated alcohols was achieved at 393 K when ethanol was added to the reactants and Co was deposited on the Mg2Ni[112]. Siddqui et al. Used cyclohexanol as hydrogen donor and Cu-MgO as catalyst to simultaneously catalyze the dehydrogenation of cyclohexanol to cyclohexanone and the selective hydrogenation of UAL to unsaturated alcohol, and the catalyst had high catalytic activity[113]. Luo et al. Obtained only unsaturated alcohol products with HCOOH as hydrogen donor at low pH, and preferentially formed saturated alcohol products with HCOONa as hydrogen donor at high pH[114]. Cai et al. Used isopropanol as a hydrogen source to obtain a high-performance catalyst, TiOx/γ-Al2O3-NT, for the reduction of CAL to COL by dispersing titania on an alumina support[115]. The selectivity of COL was higher than 99% and the conversion of CAL was higher than 95%. Zhou et al. Combined the two hydrogenation mechanisms of transfer and catalytic hydrogenation, through the transfer hydrogenation of ammonia borane, unsaturated alcohols can be selectively generated, and saturated alcohols can be generated through the synergistic effect of ammonia borane and Pt-loaded organometallic layer[116]. Zhou et al. Precisely regulated the selectivity of products by using ammonia borane as both transfer hydrogenation agent and H2 source with high catalytic activity in catalytic hydrogenation[117]. Wang et al. Catalyzed the selective hydrogenation of CAL with ethanol as solvent and hydrogen donor, and obtained excellent conversion (97.8%) and selectivity (96.8%)[118]. Farrar-Tobar et al. Used iPrOH and EtOH as hydrogen sources and commercially available Ru-MACHOTM-BH as a catalyst to catalyze the transfer hydrogenation of crotonaldehyde to unsaturated alcohols[119].
图7 使用(a) HCOOH[114], (b) 异丙醇[115], (c) 氨硼烷[116],(d) 氨硼烷[117], (e) 乙醇[118], (f) iPrOH和EtOH[119]作为氢源进行加氢反应

Fig. 7 Hydrogenation was performed using (a) HCOOH[114] (Copyright 2019, American Chemical Society), (b) isopropyl alcohol[115] (Copyright 2017, Science Press), (c) aminoborane[116] (Copyright 2020, Wiley-VCH Verlag), (d) aminoborane[117] (Copyright 2022, Elsevier), (e) ethanol[118] (Copyright 2020, American Chemical Society), (f) iPrOH and EtOH[119] (Copyright 2018, Wiley-VCH), as hydrogen sources

3.2 Solvent type

The solvent plays an important role in the catalytic reaction, and the selection of the appropriate solvent can accelerate the reaction rate, improve the reaction yield, and improve the selectivity of the reaction product (Fig. 8). Li et al. Used a conventional 5 wt% Pd/C catalyst and 12 organic solvents, including three nonpolar solvents, two protic solvents, and seven aprotic solvents[120]. The overall rate of CAL hydrogenation was largely dependent on the solvent used, and the relationship between the observed CAL hydrogenation rate and some different solvent parameters was examined. The major product was HCAL with 80% or greater selectivity in most solvents except pyridine and 4-methylpyridine. In both solvents, pyridine and 4-methylpyridine, CAL hydrogenation was slow and the major product changed to COL with a selectivity of about 60%. The addition of small amounts of pyridine to other solvents, such as 2-propanol and tetrahydrofuran, can alter the hydrogenation rate and switch the major product from HCAL to COL.
图8 在(a) 甲醇[121], (b) 2-丙醇[123], (c) 不同分子大小的醇溶剂[124], (d) 水[72]中的选择性加氢反应

Fig. 8 Selective hydrogenation of (a) methanol[121] (Copyright 2018, Elsevier Masson), (b) 2-propanol[123] (Copyright 2021, Elsevier Science), (c) alcohols with different molecular sizes[124] (Copyright 2018, American Chemical Society), and (d) water[72] (Copyright 2020, American Chemical Society)

In general, because protic solvents such as 2-propanol and ethanol have the ability of high solubility H2, they can be selected as solvents for hydrogenation reaction to accelerate the reaction. Leng et al. Prepared a catalyst with ruthenium nanoparticles deposited on ruthenium fullerenate nanospheres by a simple one-pot method, and its catalytic activity was the highest when methanol was used as the solvent, with 77% conversion, 76% selectivity to unsaturated alcohol, and TOF=128 h−1[121]. Daimon et al. Used supercritical 2-propanol to selectively reduce UAL to the corresponding unsaturated alcohol at 573 K without further addition of catalyst[122]. The Ni single-atom catalyst reported by Chen et al. Plays a key role in activating H in the dehydrogenation of 2-propanol (2-POL), with basic sites inducing α-CH cleavage in 2-POL and Lewis acidic sites adsorbing activated C = O bonds[123]. The synergistic effect between Ni SAs and acid-base sites makes the catalyst have excellent catalytic activity, and the conversion and COL selectivity reach 100%. Wang et al. Catalyzed the selective hydrogenation of CAL to COL over Al2O3 with ethanol as solvent and hydrogen donor[118]. Under the optimized reaction conditions (120 ° C, 12 H), 97.8% CAL conversion and 96.8% COL selectivity were exhibited. It has been shown by various characterizations and DFT calculations that both Lewis acidic and basic sites of Al2O3 are active sites for the catalytic reaction. It also has high catalytic activity in the selective hydrogenation of 4-fluorocinnamaldehyde, 4-chlorocinnamaldehyde L, 4-methylcinnamaldehyde, etc. (Conversion > 94%, UOL selectivity > 95%). Deng et al. Reported that molybdenum carbide (α-MoC), β-Mo2C), and molybdenum nitride (γ-Mo2N) selectively catalyzed the liquid-phase hydrogenation of furfural to furfuryl alcohol and 2-methylfuran at low temperature, and the study showed that they were the first choice for the selective hydrogenation of furfural because the hydrogen donating ability of alcoholic solution promoted the hydrogenation reaction rate[124]. By varying the size of the alcohol solvent molecule, the selectivity of the product can be controlled. Wang et al. Studied the hydrogenation of furfural over Pt (111) in different solvents such as water vapor, methanol and toluene by DFT calculation and microkinetic analysis, and found that methanol was an excellent solvent for furfural hydrogenation, while toluene was a poor solvent for furfural hydrogenation[125]. The stable adsorption of furfural in methanol increases the coverage of furfural on the surface of the catalyst and improves the catalytic activity.
Water as a reaction solvent has the advantage of being environmentally friendly. Joubert et al. Studied the mechanism of competitive C = C and C = O hydrogenation of UAL by [RuH2(PH3)3] based on DFT calculations, and proposed a new water-assisted mechanism to improve the selectivity of C = O hydrogenation[126]. Dai et al. Found that Pt3Fe/CNT showed higher activity in water than in organic solution. Through experimental and theoretical studies, it was shown that water participated in the hydrogen exchange reaction pathway and accelerated the hydrogenation reaction with a lower energy barrier[30]. The TOF was 1200 h−1 and the COL selectivity was up to 97%. Gao et al. Showed that the interlayer water species of HTs support provided a faster hydrogenation reaction path and accelerated the reaction process through isotope labeling experiments and theoretical DFT calculations[72]. Lv et al. Experimentally showed that H2O enhanced the polarity of the catalyst surface (in the form of H3O+ or dissociated OH/H+), forming a hydrogen bond network and promoting the adsorption of polar C = O groups[127]. Dai et al. Synthesized CeO2 supported ternary RuSnB catalyst by incipient wetness impregnation method and applied it to CAL hydrogenation with pure water as solvent[53]. The conversion of the catalyst is 96.0%, and the C = O bond hydrogenation selectivity is 93.9%. The kinetic isotope experiment of the substitution of D2O for H2O shows that the hydrogen exchange reaction path involving water coexists with the direct hydrogenation path of H2 dissociation.

3.3 Reaction temperature

Hydrogenation reaction is an exothermic reaction. From the thermodynamic point of view, increasing the temperature is unfavorable to the exothermic reaction, but from the kinetic point of view, increasing the temperature can accelerate the reaction rate. Therefore, the reaction rate of the catalyst can be increased by adjusting the reaction temperature. Wei et al. Found that when the reaction of selective hydrogenation of CAL was catalyzed over Pt/CeO2-ZrO2, the CAL conversion increased to 100% as the temperature increased from 50 ° C to 70 ° C, but the selectivity decreased from 97% to 79% due to further hydrogenation of COL to form hydrogenated COL[58]. Li et al. Synthesized noble metal (Rh, Pt, Ru, Ir, Pd, and Au) nanoparticles stabilized by temperature-regulating ligands Ph2P(CH2CH2O)nCH3(n=16 or 22), and realized their temperature-regulating phase transfer function in a water/alcohol biphasic system. When the reaction temperature increased to 70 ℃, the conversion of CAL reached a maximum (> 99%), while the selectivity to C = O groups began to decrease when the reaction temperature exceeded 70 ℃[128]. Cui et al. Found that the CAL conversion and COL selectivity increased from 32.7% to 94.1% and 81.4% to 95.1%, respectively, as the temperature increased from 50 ° C to 80 ° C, indicating that C = O hydrogenation over Pt/HNTs at low temperature was more favorable than C = C[81]. However, at temperatures over 80 ° C, the CAL conversion and COL selectivity decreased, and the HCOL selectivity increased significantly, which was caused by excessive hydrogenation at high temperature. Zhou et al. Found that the CAL conversion increased from 8.1% to 100% as the temperature increased from 25 ° C to 80 ° C[117]. The HCAL selectivity was 92% at 80 ° C. Guo et al. Found that when the reaction temperature was increased from 40 to 70 ° C, the conversion, COL selectivity, and yield all increased significantly, indicating that high temperature was beneficial to the target reaction[129]. When the reaction temperature rises to 80 ℃, the conversion increases because the reaction temperature reaches the boiling temperature of the reactant maleimide (MAL) at 80 ℃, but too high temperature will lead to side reactions and polymerization of the product MAA, thus reducing the selectivity and yield of the reaction.

3.4 Hydrogen partial pressure

For the hydrogenation reaction, the effect of pressure on the hydrogenation reaction is realized by the hydrogen partial pressure, and increasing the pressure increases the reaction rate of the hydrogenation reaction. When the H2 pressure was increased from 5 to 10 bar at 60 ° C, the conversion of CAL also increased from 80 to 95%. This is due to the increase of H2 pressure and the improvement of H2 solubility, which makes the H2 molecule easily accessible to the substrate and the active site. However, with further increase in H2 pressure, the conversion of CAL did not change appreciably because the solubility of H2 was close to saturation. Cui et al. Studied the dependence of CAL hydrogenation on H2 pressure at 80 ° C for 2 H[81]. The conversion of CAL increases with the increase of H2 pressure in the range of 0.5 MPa to 2.5 MPa. This is because increasing the pressure can improve the solubility of H2 and make more H2 molecules close to the catalyst in the reaction solution, thus increasing the CAL conversion, while the COL selectivity reaches the highest value of 95.1% at 2 MPa. Xin et al. Found that the CAL conversion increased with the increase of hydrogen pressure in the range of 0.5 ~ 2 MPa, and then remained stable[130]. When the hydrogen pressure is in the range of 0.5 – 4.0 MPa, the selectivity to hydrogenated COL increases slightly, indicating that too high hydrogen pressure promotes the full hydrogenation of CAL. Liu et al. Carried out the reaction at different H2 pressures[97]. With the increase of gas pressure, the reaction conversion increased rapidly, while the selectivity of HCAL was always above 98%. This means that Rh @ MIL-101 can selectively obtain hydrogenation products and avoid over-hydrogenation of products.

4 DFT study on selective hydrogenation of UAL

It is known from previous work that the electronic structure determines the catalytic performance. The correlation of thermodynamic properties with catalytic activity and selectivity is a general descriptor method that can be used instead of calculating the complete reaction pathway, and the simplest and most easily obtained one is the chemisorption energy, which is often closely related to the activation energy[9,131]. As further descriptors to elucidate the underlying electronic structure, studies typically focus on charge distributions (Bader topological atoms, ionic radius volume partitioning, bond valence and density isosurfaces) and atomically resolved spectral densities of States.
AA has been extensively studied as the smallest UAL and industrially important molecule. It has the advantages of small size, simple catalytic reaction and less hydrogenation by-products. Selective acrolein hydrogenation over platinum nanoparticles is a relevant model system to explore structure-sensitive and chemical effects. The adsorption thermodynamics of acrolein determines the selectivity of the reaction. The adsorption strength of AA produced by hydrogenation is the same as that of acrolein, but greater than that of propionaldehyde (also applicable to Ni (111) and Pd (111) crystal faces), so further hydrogenation produces n-propanol[9]. For the addition of the second type of metal, such as Sn and Fe, the electron is transferred to Pt, which weakens the adsorption of C = C bond. However, the addition of electrophilic sites can coordinate and activate C = O, which improves the hydrogenation selectivity of C = O bond. Engelhardt et al. simulated the adsorption and hydrogenation of acrolein on BaxSr1−xTiO3 supported platinum nanoparticle (Ptn),Catalysis by Ptn generated by atomic layer deposition on SrTiO3, Ba0.5Sr0.5TiO3, and BaTiO3 nanocubes,The allylic alcohol selectivity was found to increase with increasing Sr concentration in the support, which is related to the modification of the stable Winterbarton shape and lattice strain of Pt nanoparticles by the interfacial energy between Pt and BaxSr1−xTiO3 support[131].
The theoretical approach of Loffreda et al., based on DFT calculations and kinetic models, showed that the partial hydrogenation of C = O or C = C bond in acrolein over Pt (111) catalyst is determined by the equilibrium between the surface hydrogenation step and the desorption of the partial hydrogenation products[132].
Tuokko et al. Used DFT calculations to study the selective hydrogenation of acrolein on Pt (111) and Pd (111) surfaces, and found that the most stable adsorption configuration of acrolein was the di-δ1 * configuration with C = C bond adsorption.When the coverage increases, the steric repulsion between acrolein molecules increases and the adsorption energy of di-δ configuration decreases, that is, the formation of allylic alcohol is more favorable at low coverage, indicating that the steric hindrance effect is an important factor in determining the selectivity[133]. Murillo et al. Used temperature-programmed desorption (TPD), high-resolution electron energy loss spectroscopy (HREELS) and density functional theory modeling to study the trend of hydrogenation activity and selectivity to the corresponding unsaturated alcohols of 3D/Pt (111) bimetallic catalyst.The catalyst performance was found to decrease from Pt − Ni − Pt (111) to Pt − Co − Pt (111) to Pt − Cu − Pt (11), and DFT calculations showed that the different hydrogenation activity between the subsurface and surface structures seems to be related to the difference in the binding energy of acrolein on the corresponding bimetallic surface[134].
Liu et al. Discussed the adsorption configuration of CAL on Ir and Ir-Cd alloy by DFT calculation[135]. Selecting the (111) crystal plane for calculation, the adsorption energy (Ead) obtained from the perpendicular configuration calculation is − 0.57 and − 0.36 eV for Ir (111) and Ir-Cd (111), respectively, while the adsorption energy (Ead) obtained from the parallel configuration calculation is − 1.46 and − 0.13 eV, respectively, which indicates that CAL is more easily adsorbed in the perpendicular configuration on the (111 ′) plane of the alloy, while it is more easily adsorbed in the parallel configuration on Ir (111 ′).
When CAL adsorbed on the top site of Au (111) surface with C = O, the adsorption configuration is the most stable, and the adsorption energy is −93.7 kJ·mol-1. The p orbital of CAL interacts with the d orbital of the metal, and the electron is transferred from CAL to Au (111) surface[136]. For the selective hydrogenation of CAL on Au (111) surface, the possibility of COL formation is the greatest, and the conversion of CAL hydrogenation reaction can be improved by reducing the temperature. The specific reaction path is as follows: the H atom first combines with the C2 atom of CAL, and then the H atom further combines with the O1 atom to form COL. Among them, COL is adsorbed perpendicularly on the Au (111) surface, which makes the C3 and C4 atoms far away from the H atoms, thus reducing the possibility of further hydrogenation and improving the selectivity of COL. In addition, the first step of CAL hydrogenation requires the highest activation energy, which is the rate-limiting step of the reaction.
Cao et al. Found that the adsorption energy was the largest when CAL was adsorbed on the Pt atom of the Au19Pt-V cluster in parallel with C = C double bond by DFT calculation, the number of electrons transferred from CAL to the cluster was the largest, and the adsorption model was the most stable[137]. Based on the most stable adsorption model, six possible mechanisms of three kinds of reactions (1,2-addition, 3,4-addition and 1,4-addition) in the selective hydrogenation of CAL were explored, and the reaction heat, reaction energy barrier and configuration change were obtained by the transition state search of elementary reactions.The CAL molecule on the Au19Pt-V cluster is most likely to proceed via a 3,4-addition mechanism, in which an active H atom preferentially bonds with a C3 atom to form an intermediate, and another H atom adds to the intermediate to form a C4−H bond, followed by a transition state to form the final product HCAL.
Cao et al. Used density functional theory to systematically study the selective hydrogenation reaction mechanism (C = O, C = C as well as 1,4-conjugate addition mechanism) of CAL on regular icosahedral M13(M=Au,Pt) clusters[138]. On the basis of the stable adsorption configuration, the transition States of each elementary reaction in the selective hydrogenation of CAL were explored by the Complete LST/QST method, and the activation energy and reaction heat of each reaction were obtained. The calculated results show that the enol is most likely to be obtained by the 1,4-conjugate addition of CAL to the Au13 cluster, but the enol is very unstable and can be isomerized to phenylpropanal. CAL most likely undergoes C = O addition at the Pt13 cluster to give COL, which is further hydrogenated to give the saturated alcohol.
According to DFT calculations, Li et al. Concluded that the selective hydrogenation of CAL catalyzed by Ptn(n=6, 10,14,18) cluster could produce 3-phenylpropanal, 3-phenylallyl alcohol and COL[139]. Each product can be generated by two different pathways. Different atom numbers of the cluster only change the catalytic activity but not the reaction mechanism. The Ptn cluster is beneficial to catalyze the hydrogenation reaction on the C = O bond, and the catalytic effect of different Ptn clusters is different. When the catalyst is the Pt14 cluster, the activation energy is obviously reduced, and its catalytic effect is the best. From the calculation of the frontier orbital gap in the Ptn cluster, it was concluded that the catalytic activation is stronger when the gap is smaller. Li et al. Also studied the reaction mechanism of selective hydrogenation of CAL catalyzed by Co-doped Pt clusters[140]. The conclusion is similar to that of the Ptn cluster, when mixed with Co, the activation energy of the Co−Pt5 cluster for catalyzing the CAL C = O bond is lower, which is more conducive to the selective hydrogenation of the C = O bond.
Yang et al. Used M19(M=Ag, Au and Pt) clusters and M (111) surface to simulate bulk metal catalysts, and carried out density functional calculation. It was found that crotonaldehyde was adsorbed on the (Ag19) of Ag clusters in a η2-diσ(CO) mode.On the Pt (111) surface, it adsorbs in the η4 and η2-diσ(CC) modes, and on the Au cluster (Au19) on the low coordination, it adsorbs in the η2-π(CC) mode[141].
Jiang et al. Explored the effect of Pd nanoparticle size on CAL hydrogenation selectivity, and constructed a Pd4 cluster and a Pd (111) surface for DFT calculation study, and found that on the Pd4 cluster, CAL was most stable by C = C bond adsorption, with an adsorption energy of 1.24 eV[84]. On the Pd (111) surface, the C = O bond is the most stable, and the adsorption energy is 0. 75 eV, which proves that the smaller Pd particles are beneficial to the adsorption of C = C, and the larger Pd particles are beneficial to the adsorption of C = O.
Delbecq et al. Used DFT method to study the adsorption configuration and reaction performance of UAL on two alloy surfaces, p(2×2)Pt3Sn(111) and (3×3)R30°Pt2Sn(111)[142]. Where the charge is transferred from Sn to Pt, the work function decreases and the d-band center moves away from the Fermi level. The adsorption configurations of different UALs on the two alloys were compared, and the results were compared with those on Pt (111), which showed that the adsorption energy was reduced and the catalytic efficiency of the catalyst for C = O was improved.
Jiang et al. Combined DFT calculations, TPD, and high-resolution electron energy loss spectroscopy (HREELS) to investigate the hydrodeoxygenation (HDO) activity of furfural on bimetallic surfaces prepared by modifying Pt (111) with 3D transition metals (Cu, Ni, Fe, and Co)[143]. The HDO activity of furfural for 2-MF production on Co/Pt (111) surface is higher than that on Co/Pt (111) surface. For the 3D terminated bimetallic surface, the yield of 2-MF is in the order of CoPtPt (11l) > FePtPt (11l) > NiPtPt (lll) > CuPtPt (lll). It is also revealed that the presence of surface oxygen can lead to a decrease in 2-MF yield.
Joubert et al. Proposed a water-assisted mechanism based on DFT calculations, showing that although the η2-π(CC) adsorption mode is thermodynamically dominant, the Eyring theoretical kinetic model analysis constructed by the DFT value of the activation barrier found that the hydrogenation selectivity of the C = O bond is higher[126].
Cao et al. Used DFT calculations to explore the selectivity behavior of crotonaldehyde during Cu surface hydrogenation[64]. The addition of the first hydrogen atom to the adsorbed crotonaldehyde is the rate-determining step of the reaction. From the kinetic point of view, the Cu surface is conducive to the production of CTO. The relative energy of the adsorbed HCTA(E(HCTAads)=−60 kJ·mol-1) and the adsorbed CTO(E(CTOads)=−19 kJ·mol-1) is lower than that of CTA, and both products show larger adsorption energy, indicating that the surface adsorbed products will hinder the further catalytic reaction.
In order to reveal the microscopic origin of hydrogenation chemoselectivity, Liu et al. Carried out a DFT theoretical study on the reactivity of acrolein on seven metal surfaces, including Pd, Pt, Rh, Ir, Cu, Ag and Au, and found that the broadening of the frontier molecular orbital on the C = C bond was greater than that of the C = O orbital[145]. The degree of internal molecular orbital broadening serves as a guiding principle for predicting the chemoselectivity of a large class of catalytic reactions on metal surfaces.

5 Kinetic Study on Selective Hydrogenation of UAL

In order to study the product distribution of selective hydrogenation of UAL, Tan et al. Carried out dynamic product analysis of CAL to COL at Au25/ZnAl-300, and found that COL initially increased linearly with the conversion of CAL, and after about 5 H, CAL was almost completely converted, and the selectivity of COL reached 95.4%[106]. At this time, HCOL appears only in very small amounts as a by-product, and HCAL is barely detectable. In the complete conversion of CAL, the concentration of COL can be kept at a high level. Even if the reaction time was increased from 5 H to 15 H, the yield of by-products remained below 5% and the selectivity of COL remained above 95%. Wang et al. Found that when Co3O4/MC was used as a catalyst, furfural was completely converted to furfuryl alcohol (selectivity > 97%) after 8 H at 120 ℃, and the selectivity to furfuryl alcohol did not change when the hydrogenation time was extended to 24 H[8]. Komal et al. carried out dynamic product analysis for Ni-C-600, and found that the conversion rate of CAL was only 25.0% when the reaction time was 1 H, gradually increased to 96.2% when the reaction time was 6 H, and increased to 99.8% when the reactions time was extended to 8 H[144]. During the whole process, the selectivity to HCAL slowly decreased from 93.9% (1 H) to 85.2%, because part of HCAL continued to hydrogenate to HCOL, which led to a slow decrease in selectivity. This is because the hydrogenation of UAL to UOL is a thermodynamically controlled reaction. With the extension of reaction time, some selective hydrogenation products are further hydrogenated, resulting in the decrease of catalyst selectivity. Therefore, for the preparation of HCAL by selective hydrogenation, a reasonable reaction time should be selected to improve the selectivity of the catalyst.
In most studies on selective hydrogenation of UAL, the Langmuir-Hinshelwood adsorption kinetic model is used to fit the experimental data and optimize the factors added to improve the selectivity of the catalyst in specific experiments. For example, the effect of the introduction of the second metal on the number of catalytic sites, the effect of grain size on the chemisorption configuration of UAL, and the effect of phase interface. For the intermittent reactor system, it is found that the UAL selective hydrogenation reaction is a zero-order reaction with respect to the UAL concentration and a first-order reaction with respect to the hydrogen partial pressure, and the activation energy of the hydrogenation reaction is determined by Arrhenius formula.It was found that the activation energy of the reaction could be reduced and the catalytic activity of the catalyst could be enhanced by exposing specific crystal planes, introducing second type metals and enhancing the interaction between metal supports.
CAL hydrogenation was carried out on Pt/C catalyst with aqueous potassium hydroxide as promoter and toluene as solvent. The Langmuir-Hinshelwood adsorption kinetic model was used to fit the experimental data with the parameters of catalyst loading, hydrogen pressure and initial substrate concentration, and the effects of various parameters on the initial hydrogenation rate, conversion and selectivity were studied.It can be found that the initial hydrogenation rate of all substances increases in proportion to the initial substrate concentration, catalyst loading and hydrogen partial pressure, with excellent conversion and COL selectivity in the range of 1.5 – 3.1 M KOH concentration, from which it can be concluded that the C = O bond undergoes adsorption hydrogenation at the Pt-K+ site affected by KOH, and the C = C bond only undergoes adsorption hydrogenation at the Pt site[146].
Wang et al. Used DFT calculations to study the acrolein partial hydrogenation selectivity[147]. The partial hydrogenation of acrolein on Ag (111), Ag (100), Ag (211) and Ag (111) single crystal surfaces has been investigated using two hydrogenation mechanisms, namely Horiuti-Polanyi and non-Horiuti-Polanyi mechanisms. It was found that the hydrogenation of acrolein at the terminal carbon and oxygen atoms follows a non-Horiuti-Polanyi mechanism. The selectivity of the partial hydrogenation products (i.e., AA, propanal, and enol) on silver surfaces with different morphologies was compared by calculating the energy difference between the rate-determining transition States. It is found that the selectivity of AA increases with the coordination number of silver atoms on the surface.
Ning et al. Examined a series of FeOx promoted Ir/BN catalysts for the gas-phase selective hydrogenation of crotonaldehyde[148]. The addition of FeOx greatly improved the performance of the catalyst, and the yield of CROL (6.9%) was five times higher than that of the non-added Ir/BN catalyst (7.2%). The characterization results show that the Ir-FeOx interaction leads to the generation of new active centers at the interface. Kinetic experiments were carried out on representative 3Ir/BN and 3Ir-0.05 Fe/BN catalysts. For the 3Ir/BN catalyst, the formation rate increased with the increase of crotonaldehyde and hydrogen partial pressure. A similar trend was observed on the 3Ir-0.05 Fe/BN catalyst. The reaction orders of crotonaldehyde are 0.4 and 0.1, respectively, while the reaction orders of hydrogen to CROL over Ir/SiO2 and Ir-NbOx/SiO2 catalysts are 0.6 and 0.8, respectively. The apparent activation energies for the formation of CROL on the 3Ir/BN and 3Ir-0.05Fe/BN catalysts are (37.6±5.2)kJ-1 and (24.3±1.6)kJ-1, respectively. Because the Ea(24.3±1.6)kJ-1 in 3Ir-0.05 Fe/BN is lower than the Ea(37.6±5.2)kJ-1 in 3Ir/BN. The addition of FeOx greatly increased the surface coverage of crotonaldehyde and improved the activity of the catalyst.
Guo et al. Studied the kinetics of hydrogen transfer reaction between UAL and allyl alcohol. Because the homogeneous catalyst was used in the reaction, the effect of internal and external diffusion on the reaction was not considered[129]. The effect of IPA concentration on the reaction rate was investigated, and the results showed that the reaction order of IPA in the hydrogen transfer reaction was 0. There is a good linear relationship between the initial reaction rate and the amount of catalyst. The reaction order for the catalyst is therefore 1. The Arrhenius formula was used to determine the activation energy of the hydrogen transfer reaction as 49.02 kJ·mol-1. The activation energy Ea of the tubular reactor was determined to be 51.18 kJ·mol-1 using the Arrhenius formula.
M Mäki-Arvela et al. Studied the kinetics of three-phase citral hydrogenation over Ni/Al2O3 catalyst. The main parameters were hydrogen pressure (5 ~ 21 bar) and temperature (50 ~ 90 ℃). The kinetic data when the bulk density of the catalyst increased were compared with the kinetic data when the bulk density of the liquid phase remained unchanged[149]. The results show that only part of the kinetic data is not affected by the change of catalyst bulk density. For example, only at lower temperatures is the C = O bond hydrogenation selectivity independent of the packing density of the catalyst. A mechanistic kinetic model for the continuous hydrogenation of citral was derived, and since the product selectivity varied with the catalyst packing density, the kinetics were modeled for a limited data set, resulting in a good description of the experiment.
Sykes et al. Used a single crystal model and surface characterization techniques to test the PdCu bimetallic system under ultra-high vacuum (UHV) conditions, and showed that a single Pd atom on the Cu (111) single crystal surface greatly reduced the energy barrier for hydrogen absorption and subsequent desorption on the Cu metal surface.Able to dissociate H2 from the gas phase and produce a single adsorbed H atom, overflowing onto the Cu substrate for hydrogenation of the unsaturated organic reactant adsorbed on the Cu (111) surface[150]. Cao et al. Studied the kinetics of UAL catalytic hydrogenation promoted by CuPtx/SBA-15 monatomic alloy catalyst, and found that the catalyst had the highest catalytic activity when the ratio of K ′ CMO/K ′ HCMA was X = 0.005[151].
Luo et al. Used spin-polarized periodic density functional theory to calculate the hydrogenation of acrolein to propanol via allyl alcohol, propionaldehyde or enol on Ni (111) surface, and found that acrolein has an adsorption energy close to that of molecular hydrogen on Ni (111) surface[152]. This suggests that there may be co-adsorption of them and the reaction should follow the Langmuir-Hinshelwood mechanism proposed on the surface. The hydrogenation of propanal and allyl alcohol should follow the Eley-Rideal mechanism.
The activation energy (Ea)Pt for different catalysts and Ea for Pd-based catalysts were calculated by Arrhenius plot to be approximately between 17~30 kJ·mol-1,For example, the Ea of Pt/G is 18.8 kJ·mol-1,Pt/Fe3O4 and the Ea is 24.7 kJ·mol−1[153][154]. The Ea on Ir/C catalyst was 37 kJ·mol-1, and the Ea on Ni-Cu/RGO catalyst was 52 kJ·mol-1[155][56]. Pt and Pd-based catalysts are more active in the selective hydrogenation of UAL due to their lower apparent activation energy. Wang et al. Found that the Ea of Pd-WN/SBA-15 was lower than that of Pd/SBA-15 due to the synergistic effect of Pd and WN, which reduced the activation energy barrier of the reaction[156]. Yang et al. Prepared CoGa3 intermetallic compound and concluded that Ea on CoGa3 is 18.6 kJ·mol-1, which is lower than Ea(25.1 kJ·mol-1) on Co[55]. Electrons are transferred between Co and Ga to form positively charged Ga active centers, which is beneficial to the adsorption and activation of C = O and improves the selectivity of COL.
Zhang et al. Studied the diffusion behavior of citronellal molecules in UiO-66, ZIF-8, and ZIF-67 channels by using molecular dynamics simulations[157]. A 2 × 2 × 2 unit cell was chosen to represent the crystal structure of ZIF and UiO-66. First, 245 citronellal molecules were used as a sample to study their diffusion trend in ZIF. The tendency of the primary C = O group to diffuse into the ZIF-8 framework (97.75%) and ZIF-67 framework (100%) is much higher than that of the secondary C = C group (2.25% and 0%). When the citronellal molecules enter the channel of the ZIF-8 crystal, they can only move in a straight line and cannot rotate freely, so only the C = O bond can carry out the adsorption hydrogenation reaction on the surface of the encapsulated Pt NPs. Therefore, highly value-added citronellol was produced with high selectivity under the catalysis of Pt @ ZIF-8 and Pt @ ZIF-67.

6 Hydrogenation mechanism

The hydrogenation mechanism is usually modeled on a monometallic system. As shown in Fig. 9, the stepwise addition of H leads to two possible pathways for the production of each partial hydrogenation product, and the hydrogenation reaction can be carried out through multiple pathways. In general, the first hydrogenation step is rate-limiting, and the terminal atoms (O1 and C4) are more prone to hydrogenation in a Markovnikov's rule manner. Although the reduction of C = O bond has a low energy barrier in most cases, it is still difficult to obtain unsaturated alcohols. On the one hand, the C = C bond is more adsorptive at the metal site. On the other hand, SAL is usually obtained by isomerization of enol. For metals with poor ability to dissociate hydrogen, weakly bound hydrogen molecules usually undergo hydrogenation reactions in a non-Horiuti-Polyani mechanism (Fig. 10), which can seriously affect the reaction rate due to the short residence time of hydrogen on the surface[9].
图9 (a) 肉桂醛和肉桂醇产率在Au25/ZnAl-300催化剂上的时间变化过程[106]; (b) 反应时间对于Ni-C-600上选择性和转化率的影响[144]

Fig. 9 (a) Time courses of the yield of cinnamaldehyde and cinnamyl alcohol over the Au25/ZnAl-300 catalyst[106] (Copyright 2021, Elsevier B.V); (b) the effect of reaction time on the conversion of CAL and selectivity to products on Ni-C-600[144] (Copyright 2022, Springer US)

图10 不饱和醛以Horiuti-Polyani机制进行依次加氢[9]

Fig. 10 Horiuti-Polyani mechanism for the hydrogenation of the unsaturated aldehyde, in which one H atom is added in each step[9]. (Copyright 2020, American Chemical Society)

Dong et al. Used a high-throughput molecular beam apparatus to describe the hydrogenation kinetics of crotonaldehyde on a Pt surface (Fig. 11), and found that the TOF had a progressive positive dependence on the crotonaldehyde pressure, which was due to the fact that the reaction rate was limited by the coverage of crotonaldehyde on the Pt surface and could be described by the Langmuir isotherm (θ = K · P/ [1 + K · P])[158]. Coadsorbed hydrogen has a certain change in the adsorption mode and intensity of unsaturated hydrocarbons, but has no obvious effect on their adsorption range. The hydrogen atoms are adsorbed on the Pt (111) plane, and the surface metal atoms are bonded to the hydrocarbon. It was found that UAL molecules were π-bonded to the surface of the catalyst, and the hydrogenation reaction followed the Horiuti-Polanyi mechanism by adding a single hydrogen atom to form four intermediate products. Crotonaldehyde molecules were more easily adsorbed on the surface of the catalyst at higher partial pressure of crotonaldehyde, and CROL was more easily formed.
图11 巴豆醛在铂表面加氢的机理以及随着反应物压力的增加选择性趋势的变化,即从饱和醇转向不饱和醛[158]

Fig. 11 The hydrogenation of crotonaldehyde on Pt surfaces to account for the trends seen in selectivity with increasing reactant pressure, namely, a shift from the saturated alcohol to the unsaturated aldehyde[158]. (Copyright 2018, American Chemical Society)

According to the kinetic results, the reaction pathway for the formation of octyl alcohol on the catalyst was inferred (Fig. 12)[148]. First, crotonaldehyde adsorbs on the surface Lewis acid sites, whereas the adsorption of H2 is on the metal active sites. For the 3Ir-0.05 Fe/BN catalyst, in addition to Irδ+, FeOx provides new sites for crotonaldehyde adsorption. Secondly, the heterolytic dissociation (H2→H++H-) of H2 at the metal active site occurs. The steps on Ir/BN catalyst include: adsorption of crotonaldehyde on oxidized Irδ+ sites; Adsorption and Heterolytic Dissociation of H2 on Metal Ir0 Sites; The H- attacks the C = O bond to form a RCH2O- intermediate; The attack of H+ on RCH2O- forms CROL; Finally, desorption of the product is carried out.
图12 在(a) 3Ir/BN和(b) 3Ir-0.05Fe/BN催化剂上生成辛基醇的反应路线[148]

Fig. 12 Reaction route for the formation of octylalcohol over (a) 3Ir/BN and (b) 3Ir-0.05Fe/BN catalysts[148]. (Copyright 2019, Elsevier)

7 Summary and Prospect

In this paper, the research results of selective hydrogenation of UAL with noble metal and non-noble metal catalysts reported so far are reviewed. C = O bond hydrogenation is thermodynamically unfavorable, so it is still difficult to achieve highly selective hydrogenation to produce a single product, and the structure of the catalyst needs to be designed to improve the selectivity and catalytic activity of the catalyst.
From the point of view of industrial application, the catalytic activity, stability and cost of catalysts are important factors affecting the commercialization of catalysts. However, there are still many problems and challenges in the development of cheap and long-life catalysts. (1) At present, most of the highly active catalysts are still based on precious metal catalysts. Ni, Co and other non-precious metal based catalysts have high selectivity, but there is still a gap in catalytic efficiency between them and precious metal based catalysts. Non-noble metal SACs catalysts are prone to further hydrogenation reactions, resulting in low selectivity of the catalyst to a single product. At present, the design and development of efficient non-noble metal catalysts is still a difficult problem to be solved. (2) Hydrogenation catalysts are often deactivated by carbon deposition, active metal leaching, or poisoning. Therefore, the development of efficient catalysts with long-term stability remains challenging. (3) Although UAL selective hydrogenation systems have been extensively explored, in most systems, the mechanism of H2 activation to form UOL or SAL is not fully understood, and further exploration is needed.
In recent years, molecular beam and other surface research techniques have been developed to better follow the hydrogenation kinetic process by means of high-throughput collimated beams and isotope labeling. For the above problems and challenges, the research can start from the following aspects. (1) To improve the catalytic activity and selectivity of non-precious metal catalysts by adjusting the catalyst structure by design, such as designing bifunctional catalysts that catalyze hydrogen activation and hydrogenation through different active sites, such as single-atom alloy catalysts[64]. (2) Catalysts for gas-phase hydrogenation can be kinetically measured by isotope labeling and the use of high-throughput collimated beams. The stability of catalyst in liquid phase hydrogenation is more suitable to be evaluated by trickle bed reactor, which eliminates the loss of catalyst caused by recycling treatment and plays an important role in industrial production[159]. (3) The combination of advanced characterization of high-surface-area catalysts, studies of single-crystal models at ultra-high vacuum, ambient pressure, and temperature, and theoretical modeling of multicomponent and power systems provides impetus for theoretical studies of catalysts in different metal systems[9].
[1]
Zhao M T, Yuan K, Wang Y, Li G D, Guo J, Gu L, Hu W P, Zhao H J, Tang Z Y. Nature, 2016, 539(7627): 76.

[2]
Ke J X, Xiao L Y, Yu G X, Wu H J, Shen G H, Zhang Z Q. Int. J. Biol. Macromol., 2019, 125: 642.

[3]
Sun D L, Yamada Y, Sato S, Ueda W. Green Chem., 2017, 19(14): 3186.

[4]
Hidalgo-Carrillo J, Sebti J, Marinas A, Marinas J M, Sebti S, Urbano F J. J. Colloid Interface Sci., 2012, 382(1): 67.

[5]
Golmakani M T, Tahsiri Z, Alavi N. Int. J. Food Prop., 2018, 21(1): 2660.

[6]
Radhika G R, Raoul B, Mark T G, Pei L, Thomas W Hansen, Kathleen S D, David D H, Jean P T. ACS Catal., 2022, 12(12): 7344.

[7]
Alfilfil L, Ran J S, Chen C L, Dong X L, Wang J J, Han Y. Microporous Mesoporous Mater., 2022, 330: 111566.

[8]
Wang G H, Deng X H, Gu D, Chen K, Tüysüz H, Spliethoff B, Bongard H J, Weidenthaler C, Schmidt W, Schüth F. Angew. Chem. Int. Ed., 2016, 55(37): 11101.

[9]
Luneau M, Lim J S, Patel D A, Sykes E C H, Friend C M, Sautet P. Chem. Rev., 2020, 120(23): 12834.

[10]
Han Q, Liu Y F, Wang D, Yuan F L, Niu X Y, Zhu Y J. RSC Adv., 2016, 6(100): 98356.

[11]
Hu D, Fan W Q, Liu Z, Li L. ChemCatChem, 2018, 10(4): 779.

[12]
Nongwe I, Ravat V, Meijboom R, Coville N J. Appl. Catal. A Gen., 2016, 517: 30.

[13]
Yu H T, Xu Y, Havener K, Zhang L, Wu W J, Liao X J, Huang K. Mol. Catal., 2022, 525: 112343.

[14]
Tian F P, Zhang M J, Zhang X C, Chen X, Wang J L, Zhang Y F, Meng C G, Liang C H. J. Mater. Sci., 2022, 57(5): 3168.

[15]
Steffan M, Klasovsky F, Arras J, Roth C, Radnik J, Hofmeister H, Claus P. Adv. Synth. Catal., 2008, 350(9): 1337.

[16]
Vilella I M J, de Miguel S R, de Lecea C S M, Linares-Solano Á, Scelza O A. Appl. Catal. A Gen., 2005, 281(1-2): 247.

[17]
Li L, Gao G, Zheng J, Shi X, Liu Z. Diam. Relat. Mater., 2019, 91: 272.

[18]
Nagpure A S, Gogoi P, Chilukuri S V. Chem., 2021, 16(18): 2702.

[19]
Shaikh M N, Abdelnaby M M, Hakeem A S, Nasser G A, Yamani Z H. ACS Appl. Nano Mater., 2021, 4(4): 3508.

[20]
Nie R F, Miao M, Du W C, Shi J J, Liu Y C, Hou Z Y. Appl. Catal. B Environ., 2016, 180: 607.

[21]
Nagpure A S, Gurrala L, Gogoi P, Chilukuri S V. RSC Adv., 2016, 6(50): 44333.

[22]
Xu Z X, Duong-Viet C, Liu Y F, Baaziz W, Li B, Nguyen-Dinh L, Ersen O, Pham-Huu C. Appl. Catal. B Environ., 2019, 244: 128.

[23]
Xue B, Xu T C, Li D S, Xu J, Li Y X, Wang F, Zhu J. Catal. Sci. Technol., 2020, 10(4): 1096.

[24]
Zhu J, Dou M D, Lu M H, Xiang X, Ding X J, Liu W X, Li M S. Appl. Catal. A Gen., 2019, 575: 11.

[25]
Chen S J, Meng L, Chen B X, Chen W Y, Duan X Z, Huang X, Zhang B S, Fu H B, Wan Y. ACS Catal., 2017, 7(3): 2074.

[26]
Wang X X, Zheng H Y, Liu X J, Xie G Q, Lu J Q, Jin L Y, Luo M F. Appl. Catal. A Gen., 2010, 388(1/2): 134.

[27]
Chiu M E, Watson D J, Kyriakou G, Tikhov M S, Lambert R M. Angew. Chem. Int. Ed., 2006, 45(45): 7530.

[28]
Chiu M E, Kyriakou G, Williams F J, Watson D J, Tikhov M S, Lambert R M. Chem. Commun., 2006, (12): 1283.

[29]
Abid M, Touroude R. Catal. Lett., 2000, 69: 139.

[30]
Dai Y H, Gao X, Chu X F, Jiang C Y, Yao Y, Guo Z, Zhou C M, Wang C, Wang H M, Yang Y H. J. Catal., 2018, 364: 192.

[31]
Lin W, Cheng H, Li X, Zhang C, Zhao F, Arai M. Chinese J. Catal., 2018, 395:988.

(林伟伟, 程海洋, 李小汝, 张弨, 赵凤玉, 荒井正彦. 催化学报, 2018, 39:988.)

[32]
Wang X F, He Y F, Liu Y Z, Park J, Liang X H. J. Catal., 2018, 366: 61.

[33]
Merlo A B, Machado B F, Vetere V, Faria J L, Casella M L. Appl. Catal. A Gen., 2010, 383(1-2): 43.

[34]
Hui T L, Miao C L, Feng J T, Liu Y N, Wang Q, Wang Y F, Li D Q. J. Catal., 2020, 389: 229.

[35]
Wang H P, Bai S X, Pi Y C, Shao Q, Tan Y M, Huang X Q. ACS Catal., 2019, 9(1): 154.

[36]
Gu Z Z, Chen L Y, Li X Z, Chen L, Zhang Y Y, Duan C Y. Chem. Sci., 2019, 10(7): 2111.

[37]
Li X J, Zhang S F, Zhu L H, Liu J, Zhang H, Zhao N, Chen B H. Mater. Chem. Phys., 2023, 294: 127003.

[38]
Wang G M, Xin H Y, Wang Q X, Wu P, Li X H. J. Catal., 2020, 382: 1.

[39]
Wang L W, Han R R, Ma Y F, Duyar M S, Liu W, Liu J. J. Phys. Chem. C, 2021, 125(41): 22603.

[40]
Zhang W B, Xin H Y, Zhang Y Q, Jin X, Wu P, Xie W H, Li X H. J. Catal., 2021, 395: 375.

[41]
Tian Z B, Liu C, Li Q Y, Hou J Y, Li Y, Ai S Y. Appl. Catal. A Gen., 2015, 506: 134.

[42]
Su J F, Shi W, Liu X C, Zhang L Y, Cheng S B, Zhang Y, Botton G A, Zhang B S. J. Catal., 2020, 388: 164.

[43]
Zhu M W, Huang B, Shao Q, Pi Y C, Qian Y, Huang X Q. ChemCatChem, 2018, 10(15): 3214.

[44]
Dai L X, Zhu W, Lin M, Zhang Z P, Gu J, Wang Y H, Zhang Y W. Inorg. Chem. Front., 2015, 2(10): 949.

[45]
Wang Q X, Wang G M, Xin H Y, Liu J X, Xiong G, Wu P, Li X H. Catal. Sci. Technol., 2019, 9(12): 3226.

[46]
Zhang N, Shao Q, Wang P T, Zhu X, Huang X Q. Small, 2018, 14(19): 1704318.

[47]
Wang P, Shao Q, Cui X, Zhu X, Huang X. Adv. Funct. Mater., 2018, 28: 1705918.

[48]
Wang X F, Liang X H, Geng P, Li Q B. ACS Catal., 2020, 10(4): 2395.

[48]
Wang X F, Liang X H, Geng P, Li Q B. ACS Catal., 2020, 10(4): 2395.

[49]
Cattaneo S, Freakley S J, Morgan D J, Sankar M, Dimitratos N, Hutchings G J. Catal. Sci. Technol., 2018, 8(6): 1677.

[49]
Cattaneo S, Freakley S J, Morgan D J, Sankar M, Dimitratos N, Hutchings G J. Catal. Sci. Technol., 2018, 8(6): 1677.

[50]
Li R R, Yao W B, Jin Y X, Jia W P, Chen X Y, Chen J J, Zheng J L, Hu Y, Han D M, Zhao J. Chem. Eng. J., 2018, 351: 995.

[51]
Ma Y, Feng L, Guo Z L, Deng J T, Pham-Huu C, Liu Y F. Front. Chem., 2019, 7: 751.

[52]
Li P H, Wang Y Y, Wang Y W, Jin G Q, Guo X Y, Tong X L. Chin. J. Chem., 2020, 38(4): 367.

[53]
Dai Y H, Chu X F, Gu J J, Gao X, Xu M, Lu D, Wan X Y, Qi W, Zhang B S, Yang Y H. Appl. Catal. A Gen., 2019, 582: 117098.

[54]
Rana S, Jonnalagadda S B. RSC Adv., 2017, 7(5): 2869.

[55]
Yang Y S, Rao D M, Chen Y D, Dong S Y, Wang B, Zhang X, Wei M. ACS Catal., 2018, 8(12): 11749.

[56]
Mohire S S, Yadav G D. Ind. Eng. Chem. Res., 2018, 57(28): 9083.

[57]
Prakash M G, Mahalakshmy R, Krishnamurthy K R, Viswanathan B. Catal. Today, 2016, 263: 105.

[58]
Wei S P, Zhao Y T, Fan G L, Yang L, Li F. Chem. Eng. J., 2017, 322: 234.

[59]
Weng Z H, Zaera F. ACS Catal., 2018, 8(9): 8513.

[60]
Zahid M, Li J, Ismail A, Zaera F, Zhu Y J. Catal. Sci. Technol., 2021, 11(7): 2433.

[61]
Mo M, Tang J S, Zou L J, Xun Y Y, Guan H R. RSC Adv., 2022, 12(51): 33099.

[62]
Hidalgo-Carrillo J, Aramendía M A, Marinas A, Marinas J M, Urbano F J. Appl. Catal. A Gen., 2010, 385(1-2): 190.

[63]
Wei Z J, Zhu X M, Liu X S, Xu H Q, Li X H, Hou Y X, Liu Y X. Chin. J. Chem. Eng., 2019, 27(2): 369.

[64]
Cao Y Q, Chen B, Guerrero-Sánchez J, Lee I, Zhou X G, Takeuchi N, Zaera F. ACS Catal., 2019, 9(10): 9150.

[65]
Liu Y C, Wang X C, Zhang C, Xu Q, Dang L L, Zhao X, Tan H Q, Li Y G, Zhao F Y. New J. Chem., 2022, 46(33): 15950.

[66]
Tamura M, Tokonami K, Nakagawa Y, Tomishige K. Chem. Commun., 2013, 49(63): 7034.

[67]
Martinez J, Rojas H, Castaneda C, Diaz G, Gomez-Cortes A, Arenas-Alatorre J. Curr. Org. Chem., 2012, 16(23): 2791.

[68]
Zhao J, Ni J, Xu J H, Xu J T, Cen J, Li X N. Catal. Commun., 2014, 54: 72.

[69]
Lin W W, Cheng H Y, He L M, Yu Y C, Zhao F Y. J. Catal., 2013, 303: 110.

[70]
Zhao J B, Yuan H F, Gui Y H, Li X M, Qin X M, Wei C Z, Liu Y F, Wang G Q, Zhou L M, Fang S M. J. Mater. Sci., 2021, 56(9): 5760.

[71]
Yu J, Yang Y S, Wei M. Acta Chim. Sinica, 2019, 77, 1129.

[72]
Gao X, Dai H, Peng L L, Lu D, Wan X Y, Zhou C M, Zheng J W, Dai Y H, Wang H M, Yang Y H. ACS Appl. Mater. Interfaces, 2020, 12(2): 2516.

[73]
Gao Z X, Cai L Y, Miao C L, Hui T L, Wang Q, Li D Q, Feng J T. ChemCatChem, 2022, 14(18): e202200634.

[74]
Mateen M, Akhtar M N, Gao L, Cheong W C M, Lv S S, Zhou Y, Chen Z. Nano Res., 2022, 15(8): 7107.

[75]
He S N, Xie L F, Che M W, Chan H C, Yang L C, Shi Z P, Tang Y, Gao Q S. J. Mol. Catal. A Chem., 2016, 425: 248.

[76]
Zhang R, Wang L C, Ren J Z, Hu C C, Lv B L. J. Colloid Interface Sci., 2023, 630: 549.

[77]
Li Z J, Wei W, Li H H, Li S H, Leng L P, Zhang M Y, Horton J H, Wang D S, Sun W W, Guo C M, Wu W, Wang J. ACS Nano, 2021, 15(6): 10175.

[78]
Zhang R, Wang L C, Yang X, Tao Z, Ren X B, Lv B L. Appl. Surf. Sci., 2019, 492: 736.

[79]
Kato S, Ohyama J, Machida M, Satsuma A. Catal. Sci. Technol., 2019, 9(9): 2097.

[80]
Hu Q M, Wang S, Gao Z, Li Y Q, Zhang Q, Xiang Q, Qin Y. Appl. Catal. B Environ., 2017, 218: 591.

[81]
Cui R Q, Zhou J Q, Wang D, Zhao Y F, Xiang X, Shang H S, Zhang B. Dalton Trans., 2023, 52(11): 3325.

[82]
Xin H Y, Zhang W B, Xiao X X, Chen L, Wu P, Li X H. J. Catal., 2021, 393: 126.

[83]
Peres L, Axet M R, Yi D L, Serp P, Soulantica K. Catal. Today, 2020, 357: 166.

[84]
Jiang F, Cai J, Liu B, Xu Y B, Liu X H. RSC Adv., 2016, 6(79): 75541.

[85]
Yin D, Ren H, Li C, Liu J, Liang C. Chinese J. Catal., 2018, 392, 319.

[86]
Manyar H G, Yang B, Daly H, Moor H, McMonagle S, Tao Y, Yadav G D, Goguet A, Hu P, Hardacre C. ChemCatChem, 2013, 5(2): 506.

[87]
O’Brien C P, Dostert K H, Schauermann S, Freund H J. Chem., 2016, 22(44): 15856.

[88]
Wang F, Bi Y S, Hu K, Wei X J. Chem., 2020, 26(21): 4874.

[89]
Wang F, Zhang Z N, Wei X J, Fang Q H, Jiang X M. Appl. Catal. A Gen., 2017, 543: 196.

[90]
Aydin C, Lu J, Browning N D, Gates B C. Angew. Chem. Int. Ed., 2012, 51(24): 5929.

[91]
Guo M Y, Du Y Z, Zhang M W, Wang L, Zhang X W, Li G Z. ACS Sustainable Chem. Eng., 2022, 10(23): 7485.

[92]
Liu C, Zhu P, Wang J S, Liu H O, Zhang X F. Chem. Eng. J., 2022, 446: 137064.

[93]
Zhou A W, Dou Y B, Zhou J, Li J R. ChemSusChem, 2020, 13(1): 205.

[94]
Liu Q L, Liu Q, Chen Y R, Li Y L, Su H, Liu Q H, Li G Q. Chin. Chemical Lett., 2022, 33(1): 374.

[95]
Geng R Y, Jia H H, Xie Y W, Pan D H, Yu F, Fan B B. Microporous Mesoporous Mater., 2022, 338: 111968.

[96]
Yuan K, Song T Q, Wang D W, Zhang X T, Gao X, Zou Y, Dong H L, Tang Z Y, Hu W P. Angew. Chem. Int. Ed., 2018, 57(20): 5708.

[97]
Liu Q L, Li Y L, Fan Y N, Su C Y, Li G Q. J. Mater. Chem. A, 2020, 8(22): 11442.

[98]
Zhang B Z, Li M Y, Lei H J, Chen J L, Wang S H, Chen C. Appl. Surf. Sci., 2022, 599: 153899.

[99]
Lan X C, Wang T F, Li X D, Huang N, Wang J F. Catal. Sci. Technol., 2016, 6(21): 7703.

[100]
Patil K N, Manikanta P, Srinivasappa P M, Jadhav A H, Nagaraja B M. J. Environ. Chem. Eng., 2022, 10(5): 108208.

[101]
Cui H S, Liu S H, Lv Y, Wu S T, Wang L P, Hao F, Liu P L, Xiong W, Luo H A. J. Catal., 2020, 381: 468.

[102]
Tao R, Shan B Q, Sun H D, Ding M, Xue Q S, Jiang J G, Wu P, Zhang K. J. Phys. Chem. C, 2021, 125(24): 13304.

[103]
Zahid M, Ismail A, Sohail M, Zhu Y J. J. Colloid Interface Sci., 2022, 628: 141.

[104]
Vu K B, Bukhryakov K V, Anjum D H, Rodionov V O. ACS Catal., 2015, 5(4): 2529.

[105]
Sole R, Buranello C, Bardella N, Di Michele A, Paganelli S, Beghetto V. Catalysts, 2021, 11(8): 914.

[106]
Tan Y, Liu X Y, Zhang L L, Liu F, Wang A Q, Zhang T. Chin. J. Catal., 2021, 42(3): 470.

[107]
Ye Y W, Hu Y M, Zheng W B, Jia A P, Wang Y, Lu J Q. Chin. J. Catal., 2023, 47: 265.

[108]
Schröder C, Baumann A K, Schmidt M C, Smyczek J, Haugg P A, Graap O C, Schauermann S. J. Phys. Chem. C, 2022, 126(10): 4907.

[109]
Farrag M. Microporous Mesoporous Mater., 2018, 257: 110.

[110]
Murata K, Ogura K, Ohyama J, Sawabe K, Yamamoto Y, Arai S, Satsuma A. ACS Appl. Mater. Interfaces, 2020, 12(23): 26002.

[111]
Zhang Y B, Ettelaie R, Binks B P, Yang H Q. ACS Catal., 2021, 11(3): 1485.

[112]
Yamada H, Fujimura M, Goto S. J. Alloys Compd., 2004, 375(1-2): 202.

[113]
Siddqui N, Sarkar B, Pendem C, Khatun R, Sivakumar Konthala L N, Sasaki T, Bordoloi A, Bal R. Catal. Sci. Technol., 2017, 7(13): 2828.

[114]
Luo N H, Liao J H, Ouyang L, Wen H L, Liu J T, Tang W P, Luo R S. Organometallics, 2019, 38(15): 3025.

[115]
Cai W M, Yang J, Sun H F, Wang Y B, Ling T, Guo X F, Peng L M, Ding W P. Chin. J. Catal., 2017, 38(8): 1330.

[116]
Zhou Y Y, Li Z H, Liu Y B, Huo J, Chen C, Li Q L, Niu S Y, Wang S Y. ChemSusChem, 2020, 13(7): 1746.

[117]
Zhou Y Y, Chen C, Li Q L, Liu Y B, Wei T, Liu Y Z, Zeng Z B, Bradshaw D, Zhang B, Huo J. Appl. Catal. B Environ., 2022, 311: 121348.

[118]
Wang H J, Liu B Y, Liu F, Wang Y N, Lan X C, Wang S Q, Ali B, Wang T F. ACS Sustainable Chem. Eng., 2020, 8(22): 8195.

[119]
Farrar-Tobar R A, Wei Z H, Jiao H J, Hinze S, De Vries J G. Chem., 2018, 24(11): 2725.

[120]
Li Y, Cheng H Y, Lin W W, Zhang C, Wu Q F, Zhao F Y, Arai M. Catal. Sci. Technol., 2018, 8(14): 3580.

[121]
Leng F Q, Gerber I C, Axet M R, Serp P. Comptes Rendus Chim., 2017, 21(3/4): 346.

[122]
Daimon A, Kamitanaka T, Kishida N, Matsuda T, Harada T. J. Supercrit. Fluids, 2006, 37(2): 215.

[123]
Chen Y Q, Liu W, Yin P, Ju M H, Wang J, Yang W Y, Yang Y S, Shen C. J. Catal., 2021, 393: 1.

[124]
Deng Y C, Gao R, Lin L L, Liu T, Wen X D, Wang S, Ma D. J. Am. Chem. Soc., 2018, 140(43): 14481.

[125]
Wang J, Lv C Q, Liu J H, Ren R R, Wang G C. Int. J. Hydrog. Energy, 2021, 46(2): 1592.

[126]
Joubert J, Delbecq F. Organometallics, 2006, 25(4): 854.

[127]
Lv Y, Han M M, Gong W B, Wang D D, Chen C, Wang G Z, Zhang H M, Zhao H J. Angew. Chem. Int. Ed., 2020, 59(52): 23521.

[128]
Li W J, Wang Y H, Chen P, Zeng M, Jiang J Y, Jin Z L. Catal. Sci. Technol., 2016, 6(20): 7386.

[129]
Guo H, Li Y C, Zuo C C, Zheng Y X, Guo X P, Hao J G, Huang H F, Wang M, Ge T T. AlChE. J., 2022, 68(7): e17683.

[130]
Xin H Y, Li M N, Chen L, Zhao C, Wu P, Li X H. Catal. Sci. Technol., 2023, 13(5): 1488.

[131]
Engelhardt C M, Kennedy R M, Enterkin J A, Poeppelmeier K R, Ellis D E, Marshall C L, Stair P C. ChemCatChem, 2018, 10(3): 632.

[132]
Loffreda D, Delbecq F, Vigné F, Sautet P. Angew. Chem. Int. Ed., 2005, 44(33): 5279.

[133]
Tuokko S, Pihko P M, Honkala K. Angew. Chem. Int. Ed., 2016, 55(5): 1670.

[134]
Murillo L E, Menning C A, Chen J G. J. Catal., 2009, 268(2): 335.

[135]
Liu Q L, Wu J Y, Kang J W, Liu Q, Liao P S, Li G Q. Nanoscale, 2022, 14(41): 15462.

[136]
Jiang J, Xia S, Ni Z, Zhang L. Chem. J. Chinese U., 2016, 374: 693.

(蒋军辉, 夏盛杰, 倪哲明, 张连阳. 高等学校化学学报, 2016, 374:693.)

[137]
Cao Y, Jiang J, Ni Z, Xia S, Qian M, Xue J. Chem. J. Chinese U., 2016, 377:1342.

(曹勇勇, 蒋军辉, 倪哲明, 夏盛杰, 钱梦丹, 薛继龙. 高等学校化学学报, 2016, 377: 1342.)

[138]
Chang S S, Meng S G, Fu X L, Zhang S J, Zheng X Z, Chen S F. Chemistry Select, 2019, 4(7): 2018.

[139]
Li L C, Wang W, Wang X L, Zhang L. J. Mol. Model., 2016, 22(8): 186.

[140]
Li L C, Wei W, Wang W, Wang X L, Zhang L, Tian A M. RSC Adv., 2016, 6(91): 88277.

[141]
Yang X F, Wang A Q, Wang X D, Zhang T, Han K L, Li J. J. Phys. Chem. C, 2009, 113(49): 20918.

[142]
Delbecq F, Sautet P. J. Catal., 2003, 220:115.

[143]
Jiang Z F, Wan W M, Lin Z X, Xie J M, Chen J G. ACS Catal., 2017, 7(9): 5758.

[144]
Komal N P, Manikanta P, Puneethkumar M S, Arvind H J, Bhari M N. J. Environ. Chem. Eng., 2023, 11(1): 109168.

[145]
Liu W, Jiang Y D, Dostert K H, O’Brien C P, Riedel W, Savara A, Schauermann S, Tkatchenko A. Sci. Adv., 2017, 3(7): e1700939.

[146]
Khan M, Joshi S, Ranade V. Chem. Eng. J., 2019, 377: 120512.

[147]
Wang K L, Yang B.. Catal. Sci. Technol., 2017, 7: 4024.

[148]
Ning X, Xu Y M, Wu A Q, Tang C, Jia A P, Luo M F, Lu J Q. Appl. Surf. Sci., 2019, 463: 463.

[149]
Mäki-Arvela P, Denecheau A, Alho K, Wärnå J, Eränen K R, Salmi T, Murzin D Y. Chem. Eng. J., 2007, 134(1-3): 268.

[150]
Kyriakou G, Boucher M B, Jewell A D, Lewis E A, Lawton T J, Baber A E, Tierney H L, Flytzani-Stephanopoulos M, Sykes E C H. Science, 2012, 335(6073): 1209.

[151]
Cao Y Q, Guerrero-Sańchez J, Lee I, Zhou X G, Takeuchi N, Zaera F. ACS Catal., 2020, 10(5): 3431.

[152]
Luo Q Q, Wang T, Beller M, Jiao H J. J. Phys. Chem. C, 2013, 117(24): 12715.

[153]
Ji X W, Niu X Y, Li B, Han Q, Yuan F L, Zaera F, Zhu Y J, Fu H G. ChemCatChem, 2014, 6(11): 3246.

[154]
Zhang Y, Chen C, Gong W B, Song J Y, Su Y P, Zhang H M, Wang G Z, Zhao H J. RSC Adv., 2017, 7(34): 21107.

[155]
Breen J P, Burch R, Gomez-Lopez J, Griffin K, Hayes M. Appl. Catal., A., 2004, 268: 267.

[156]
Wang D, Zhu Y J, Tian C G, Wang L, Zhou W, Dong Y L, Yan H J, Fu H G. ChemCatChem, 2016, 8(9): 1718.

[157]
Zhang W L, Shi W X, Ji W L, Wu H B, Gu Z D, Wang P, Li X H, Qin P S, Zhang J, Fan Y, Wu T Y, Fu Y, Zhang W N, Huo F W. ACS Catal., 2020, 10(10): 5805.

[158]
Dong Y, Zaera F. J. Phys. Chem. Lett., 2018, 9(6): 1301.

[159]
Lan X C, Wang T F. ACS Catal., 2020, 10(4): 2764.

Outlines

/