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

Carbon-Based Electrocatalyst Derived from Porous Organic Polymer in Oxygen Reduction Reaction for Fuel Cells

  • Sun Hanxue , * ,
  • Wang Juanjuan ,
  • Zhu Zhaoqi ,
  • Li An , *
Expand
  • Lanzhou University of Technology,Lanzhou 730050, China
*Corresponding author e-mail: (Hanxue Sun),

Received date: 2023-03-31

  Revised date: 2023-06-19

  Online published: 2023-09-11

Supported by

National Natural Science Foundation of China(22006061)

National Natural Science Foundation of China(21975113)

Gansu Provincial Science Fund for Distinguished Young Scholars(23JRRA808)

Postgraduate “Innovation Star” program of Gansu Province(2023CXZX423)

Abstract

Fuel cell, a kind of energy conversion device that can directly convert chemical energy into electric energy, is a new and an important energy technology during China’s 14th Five-Year Plan. In recent years, the fuel cell technology has undergone iterative upgrading, which effectively promotes the transition of hydrogen energy industry from mode-exploration to multiple demonstration, and helps the high-quality development of the new energy. Cathodic oxygen reduction reaction (ORR) is one of the basic and core reactions of fuel cells, but its slow kinetic process restricts the large-scale application of fuel cells. Although metal Pt-based catalysts have high catalytic activity and can improve the reaction rate of ORR, they are not conducive to wide commercial use because of their scarcity, high cost and poor durability. The development of non-Pt-based ORR catalysts is of great practical significance to promote the development of fuel cells. Porous Organic Polymers (POPs) are an important branch of porous materials. Due to their controllable composition and diverse structure, heteroatoms and metal species can be incorporated into the skeleton to enhance the overall catalytic activity of materials. As ideal candidate materials for electrocatalysts with high-efficiency, POPs have attracted wide attention in promoting the slow kinetics of ORR. In this paper, the research progress in the synthesis strategy, composition, morphology, structure regulation and electrocatalytic performance of POPs-derived carbon-based ORR electrocatalysts in recent years are emphatically introduced. The challenges faced by POPs-derived carbon-based ORR electrocatalysts at present are discussed, and their future development directions are summarized.

Contents

1 Introduction

2 Oxygen reduction reaction(ORR)mechanism

3 Design and performance of porous organic polymer derived carbon-based ORR electrocatalysts

3.1 Conjugated microporous polymers(CMPs)derived carbon-based ORR catalysts

3.2 Covalent organic frameworks(COFs)derived carbon-based ORR catalysts

3.3 Hyper-cross-linked polymers(HCPs)derived carbon-based ORR catalysts

3.4 Covalent triazine frameworks(CTFs)derived carbon-based ORR catalysts

3.5 Polymers of intrinsic microporosity(PIMs)derived carbon-based ORR catalysts

3.6 Porous aromatic frameworks(PAFs)derived carbon-based ORR catalysts

4 Conclusion and outlook

Cite this article

Sun Hanxue , Wang Juanjuan , Zhu Zhaoqi , Li An . Carbon-Based Electrocatalyst Derived from Porous Organic Polymer in Oxygen Reduction Reaction for Fuel Cells[J]. Progress in Chemistry, 2023 , 35(11) : 1638 -1654 . DOI: 10.7536/PC230332

1 Introduction

Since the 21st century, coal, oil and other non-renewable energy sources are facing depletion, and the problems of global climate change and living environment deterioration caused by massive combustion and uncontrolled use of fossil fuels have become increasingly prominent[1]. Under the common goal of "carbon peak" and "carbon neutrality", the transformation of energy structure (from traditional fossil energy to renewable energy) and green and low-carbon energy have become the main theme of world energy science and technology development in the 21st century and the main direction of energy development layout in various countries. Hydrogen energy is a clean energy (the only product is water) with abundant reserves and high energy density. It is an ideal medium to promote the decarbonization transformation of traditional fossil energy and support the large-scale utilization of renewable energy. It is known as the most promising alternative energy in the 21st century[2]. At present, fuel cell is one of the safest energy conversion methods, which can directly convert chemical energy stored in hydrogen (fuel) and oxygen (oxidant) into electrical energy. Compared with other batteries (such as lithium-ion batteries), fuel cells work more like "internal combustion engines", with high energy conversion efficiency, environmental friendliness and high reliability[3]. In 1839, William Grove, the "father of fuel cells", proposed the world's first fuel cell model. In the 1960s, fuel cells were first used as auxiliary power on the Apollo lunar landing spacecraft, marking the transition of fuel cell research and development from the experimental stage to the military stage. In 2014, Toyota launched the world's first commercial fuel cell vehicle Mirai, which opened the exploration of the commercial mode of fuel cell vehicles.
The core structure of fuel cell includes anode material, electrolyte and cathode material. Hydrogen Oxidation Reaction (HOR) takes place at the anode, while Oxygen Reduction Reaction (ORR) takes place at the cathode[4]. Compared with anodic HOR, the reaction rate of ORR is very slow kinetically due to the high bond energy (~494 kJ·mol-1) of the O — O bond of oxygen molecules, which makes the oxygen-containing species generated in the reaction process difficult to desorb and occupy the active sites of the catalyst. In addition, the cathodic overpotential is large and the reaction process is complex, which makes ORR a major factor restricting the development of fuel cells[5]. At present, the most widely used cathode catalyst is platinum (Pt) and its compounds[6,7]. However, Pt is rare in nature and expensive, which is one of the main factors restricting the price of fuel cell vehicles. In addition, although Pt has high activity, it is easy to be poisoned by CO and sulfide, which will affect the working life of fuel cells[8]. Therefore, the development of non-platinum-based ORR catalysts with high catalytic activity, good durability and low cost is the key to promote large-scale commercial application of fuel cells.
Carbon-based porous materials are ideal candidates for replacing Pt-based electrocatalysts because of their high conductivity, wide sources, and diverse structures, as well as their excellent methanol resistance and long-term cycle stability in both acidic and alkaline media. However, the catalytic activity of pure carbon materials is not ideal. In order to overcome its inherent defects, effective strategies such as optimizing the porous structure or doping heteroatoms are usually adopted[9,10]. However, due to the difference of carbon sources and the complexity of carbonization process, the controllability of carbon-based catalyst products is poor, especially the precise regulation of the location and number of active sites in the structure is difficult to achieve. Therefore, the research and development of efficient, stable and controllable carbon-based ORR electrocatalysts is still a hot topic in the field of fuel cells[11~13].
Porous Organic Polymers (POPs) is an important member of the family of Porous materials, which has many advantages, such as high specific surface area, diverse synthesis methods, adjustable pore structure, strong functional design and so on[14,15]. Through "bottom-up" or post-modification synthesis methods, the active material can be incorporated into its skeleton structure and directly used as an effective carbon precursor to synthesize carbon-based ORR catalysts. In this paper, the research status of POPs derived carbon-based ORR electrocatalysts suitable for fuel cells is discussed, starting from the ORR reaction mechanism of fuel cells, combined with the design concept of POPs materials.The application of important members of POPs family in ORR and the research progress in the past five years were summarized, and the advantages, existing problems and future development directions of POPs-based ORR electrocatalysts were summarized.

2 ORR mechanism

ORR is a fundamental reaction in electrochemistry, and understanding its mechanism will be helpful for the design and performance optimization of POPs-based electrochemical catalysts. ORR is a complex multi-electron reaction process, in which a variety of oxygen-containing intermediates are produced[18]. At present, there is no test method to observe the oxygen reduction process in situ, so there is no unified conclusion about the ORR mechanism, and we can only rely on the detection of the intermediate products to speculate the possible reaction pathway.
(Ⅰ) Under acidic conditions, the multi-electron reaction model of metal surface proposed by Wroblowa et al. In 1976 is the generally accepted ORR reaction mechanism[16]. As shown in Fig. 1A, the O2 molecules dissolved in the solution are first adsorbed on the surface of the catalyst by diffusion to form adsorbed O2 molecules, and the reduction reaction occurs under the electrochemical action, followed by two possible reaction pathways.
图1 (a) 酸性条件下金属表面多电子ORR反应模型[16];(b)酸性条件下N掺杂碳材料ORR反应机理[17]

Fig.1 (a) Model of multielectron ORR reaction on metal surface under acid condition[16]. Copyright 1976, Elsevier. (b) ORR reaction mechanism of N-doped carbon materials under acidic conditions[17]. Copyright 2016, American Association for the Advancement of Science

(A)4e- reduction reaction mechanism: The adsorbed O2 molecule (O 2 *) directly obtains 4e- from the solution and generates H2O according to the reaction equation (1).
O 2 + 4 H + + 4 e - 2 H 2 O E 0 = 1.229 V
(B) Mechanism of continuous 2e- reduction reaction: O 2 * (adsorbed O2 molecule) gets 2e- from the solution and generates the intermediate product H2O2 according to the reaction equation (2). This stage may undergo three reactions: :①H2O2 regenerating O 2 *;The ②H2O2 obtains the 2e- to generate the H2O(3), and the process is a (2+2)e- process; The ③H2O2 is dissolved into the solution by diffusion.
O 2 + 2 H + + 2 e - H 2 O 2 E 0 = 0.670 V
H 2 O 2 + 2 H + + 2 e - 2 H 2 O E 0 = 1.763 V
Guo et al. Used highly oriented pyrolytic graphite as a model, combined with in situ X-ray photoelectron spectroscopy (XPS) and CO2 adsorption means to reveal the ORR active sites and reaction mechanism of N-doped carbon materials for the first time, which is similar to the model proposed by Wroblowa et al[17]. As shown in fig. 1b, due to the presence of pyridine N, the adjacent C atom has Lewis basicity, so that the O2 molecule is first adsorbed on the C atom adjacent to pyridine N. Under acidic conditions, the adsorbed O2 molecule is protonated to form -OOH (4) by e- from the solution.
O 2 * + H + + e - O O H
Two reaction pathways may occur next.
The reaction mechanism of (A)4e- reduction is that :H+ attacks the oxygen atom, resulting in the cleavage of the O — OH bond to form -OH and H2O(5),-OH to obtain e- to produce H2O(6).
O O H + 2 H + + 2 e - O H + H 2 O
O H + H + + e - H 2 O
(B)2e- reduction reaction mechanism: -OOH obtains e- and combines with H+ to generate H2O2,It can be directly desorbed from the catalyst and dissolved in the solution (7), or 2e- can be obtained to form water H2O(3), which can be regarded as the (2+2)e- reaction process.
O O H + H + + e - H 2 O 2
(Ⅱ) Under alkaline conditions, the H2O molecule acts as a proton donor, and the reaction mechanism can also be summarized as a complete 4e- reduction process and a 2e- reduction process[19~23].
(A)4e- direct reduction reaction process: According to the reaction formula, (8),O2 molecules obtain 4e-, which is completely reduced to OH-.
O 2 + 2 H 2 O + 4 e - 4 O H - E 0 = 0.401 V
The (B)2e- reduction reaction process reduces the :O2 molecule to obtain the 2e- part to the intermediate H O 2 - (9),H O 2 - is further reduced (10) or directly decomposed to OH-(11).
O 2 + H 2 O + 2 e - H O 2 - + O H - E 0 = 0.065 V
H O 2 - + H 2 O + 2 e - 3 O H - E 0 = 0.867 V
2 H O 2 - 2 O H - + O 2
From the point of view of reduction potential, whether in acidic or alkaline conditions, the 4e- reaction path can better reduce O2 molecules to H2O completely, showing a higher output voltage. Therefore, the 4e- reaction process is an ideal reaction pathway for the current carbon-based ORR catalysts.

3 Design and Performance of POPs Derived Carbon-based ORR Electrocatalyst

POPs are a kind of porous materials with highly crosslinked network framework, which are connected by organic structural units through covalent bonds, and usually contain C, H, B, N, O and a small amount of metal elements[24]. Compared with conventional porous materials, POPs have unique advantages: on the one hand, most POPs have interconnected pore structure space, and the derived carbon-based catalyst can largely inherit the porosity of the original POPs, thus providing a fast and smooth electron and material transmission channel. On the other hand, the structure and composition of POPs materials can be designed to facilitate the specific binding of heteroatoms to improve the ORR catalytic activity. At present, POPs family mainly consists of Conjugated Microporous Polymers (CMPs), Covalent Organic Frameworks (COFs), Hyper-Cross-linked Polymers (HCPs), Covalent Triazine Frameworks (CTFs), Porous Aromatic Frameworks (PAFs) and Polymers of Intrinsic Microporosity (PIMs). In the past decade, with the advantages of porous structure and functional designability, the development and research of POPs in the field of ORR catalysis have shown an obvious growth trend (Figure 3).
图2 POPs的典型合成过程和分子结构单元示例

Fig.2 Examples of typical synthesis process and molecular structural unit of POPs

图3 在Web of Science上搜索主题如“conjugated microporous polymers”和“oxygen reduction”获得的文献量数据(2014~2022)

Fig.3 The statistics of the paper by topic of “conjugated microporous polymers” as an example and “oxygen reduction” indexed in Web of Science from 2014 to 2022

POPs-derived carbon-based ORR electrocatalysts are mainly prepared by pyrolysis, and the introduction of heteroatoms (N, P, S, Fe, Co, etc.) into the catalyst system can be summarized as pre-doping and post-doping (Fig. 4). Pre-doping is mainly based on the "bottom-up" design strategy, in which heteroatom-containing POPs (i-POPs, I = N, P, S, Fe, Co, etc.) Are first synthesized from heteroatom-containing building block units, and then the i-POPs are directly carbonized to obtain heteroatom-doped carbon materials. Post-doping is mainly through physical blending and chemical bonding to obtain carbon-based catalysts containing heteroatoms by one-step carbonization of pre-synthesized POPs with dopants (such as metal salts, phytic acid, etc.).
图4 POPs衍生碳基ORR电催化剂的设计流程

Fig.4 Schematic of POPs derived carbon-based ORR electrocatalysts

3.1 CMPs-derived carbon-based ORR catalyst

In 2007, Copper et al. Of the University of Liverpool reported the first network of poly (arylenevinylene) CMPs[25]. CMPs are a class of polymer materials with permanent porous structure constructed by π-conjugated structural units[26~28]. These materials have various synthetic routes, high specific surface area, controllable pore structure and modifiable surface chemical properties. As ORR catalysts, CMPs have the following characteristics: (1) The network of CMPs is mainly composed of rigid aromatic structural units, which on the one hand endows CMP with excellent thermal stability (thermal decomposition temperature is higher than 300 ℃), and on the other hand, its pore structure is not easy to collapse during pyrolysis, thus maintaining the inherent micro-and nano-pore structure to provide more active sites for ORR. (2) The modularity of CMPs facilitates the modification of the structural skeleton by functional units, thereby constructing heteroatom-supported ORR catalysts. Compared with other POPs materials, CMPs have abundant synthetic routes, but CMPs themselves have poor conductivity, which is not conducive to electron transport when they are used as ORR catalysts alone. The main research direction in this field is to use CMPs as precursors to produce carbon-based ORR catalysts doped with non-Pt atoms by pyrolysis. At present, non-metallic elements such as N, S and P, and non-noble metal elements such as Fe, Co and Zn are introduced into CMPs-derived carbon-based ORR catalyst structures by monomer polymerization or metal salt blending to enhance the oxygen reduction capacity of materials. Table 1 summarizes the ORR performance data of CMPs and their derived carbons reported in recent years.
表1 近年来CMPs及其衍生碳基ORR催化剂的电化学性能汇总

Table 1 Summary of electrochemical properties of CMPs and their derived carbon-based ORR catalysts

Catalysts Heteroatom Method for CMPs Eonset vs.RHE (V) E1/2 vs.RHE (V) Jd
(mA·cm-2)
Pt/C Jd
(mA·cm-2)
Electrolyte ref
1DPC-L3 B, N, S Sonogashira-Hagihara coupling reaction / 0.75 4.6 5.3 0.1 mol/L KOH 29
N-HsGDY-900 ℃ N Sonogashira-Hagihara coupling reaction 0.86 0.64 4.7 / 0.1 mol/L HClO4 18
1.02 0.85 6.5 / 0.1 mol/L KOH
TPA-BP-1 N Sonogashira-Hagihara coupling reaction 0.80 / / / 0.1 mol/L KOH 27
TPA-TPE-2 0.82 / /
ZnPcFePor-CMP Fe, Zn, N Sonogashira-Hagihara coupling reaction 0.902 0.724 -5.31 / 0.1 mol/L KOH 26
FePcZnPor-CMP 0.936 0.866 -5.59
CPP-P1 N Sonogashira-Hagihara coupling reaction 0.87 0.73 4.71 4.88 0.1 mol/L KOH 30
BP-800 B, N, Co, Fe Sonogashira-Hagihara coupling reaction 0.85 0.66 5.97 / 0.1 mol/L HClO4 31
0.93 0.80 5.95 5.57 0.1 mol/L KOH
0.85 0.66 / / 0.1 mol/L PBS
BPCMP-Fe-800 Fe, N Sonogashira-Hagihara coupling reaction 0.97 0.85 4.98 / 0.1 mol/L KOH 32
BBCMP-Fe-800 0.81 0.71 /
C-CMPs-NP N, S Sonogashira-Hagihara coupling reaction 0.98 0.82 4.2 4.3 0.1 mol/L KOH 33
NHCNT-1 N Sonogashira-Hagihara coupling reaction 0.87 0.76 3.8 / 0.1 mol/L KOH 34
1.15 0.45 4.4 / 0.1 mol/L HClO4
Fe/N-CMP-1000 Fe, N Sonogashira-Hagihara coupling reaction 0.95 0.85 5.10 4.10 0.1 mol/L KOH 35
CMP-NP-800 N Sonogashira-Hagihara coupling reaction 0.903 0.815 / 4.25 0.1 mol/L KOH 36
CMP-NP-900 0.930 0.857 4.45
CMP-NP-1000 0.872 0.766 /
N-Fc-800 Fe, N Schiff base reaction 0.96 0.82 5.3 4.6 0.1 mol/L KOH 37
CoNCs800 Co, N Schiff base reaction 0.905 0.807 -4.72 / 0.1 mol/L KOH 38
0.80 0.70 -4.40 3.57 0.5 mol/L H2SO4
CoPP-FePc-CMPs Co, Fe, N Schiff base reaction 0.837 0.426 1.537 5.85 0.1 mol/L KOH 39
CoFeNC 0.904 0.775 3.68
CoFeNG 0.957 0.777 4.00
C-POP-2-900 N, P Schiff base reaction -0.11 -0.19 / / 0.1 mol/L KOH 40
Fe/Co-CMP-800 N, Fe, Co Suzuki coupling reaction 0.88 0.78 4.5 / 0.5 mol/L H2SO4 41
TT-TPB S Suzuki coupling reaction 0.9 0.89 / / 0.1 mol/L KOH 28
TPP-CMP-900 N Suzuki coupling reaction 0.95 0.83 4.05 4.1 0.1 mol/L KOH 42
XWB-CMP-1000 N, S one-pot catalyst-free
procedure
-0.11 -0.19 -5.2 / 0.1 mol/L KOH 43
CoO/ZnO@N-PC Co, Zn, N Molten salt-templated
approach
0.91 0.85 / / 0.1 mol/L KOH 44
N, P-CMP-1000 N, P Acid-catalyzed con-
densation
0.94 0.84 / / 0.1 mol/L KOH 45
0.75 0.57 / / 0.1 mol/L HClO4
/ 0.48 / / 0.01 mol/L PBS

3.1.1 Direct introduction of heteroatoms by monomer polymerization

CMPs are synthesized by various methods. Monomers containing N, S, Fe, Co, Cu and other atoms are introduced into the network structure of CMPs after polymerization (such as Sonogashira-Hagihara coupling, Buchwald-Hartwig coupling, Suzuki-Miyaura reaction). After carbonization, these heteroatoms are retained in the porous carbon[46][47][48]. In N-doped CMPs, Li et al. Constructed N-doped CMPs network by Sonogashira-Hagihara coupling reaction with 2-amino-3,5-dibromopyridine as N-containing monomer. After pyrolysis, the maximum content of N element is 9.91 wt%, and the onset potential (Eo) reaches 1.15 V in 0.1 mol/L HClO4 solution[34]. In 2022, they prepared porphyrin-based CMPs by providing N and C sources from m-tetra (4-chlorophenyl) porphyrin monomer, and the limiting current density (Jd) of the sample pyrolyzed at 900 ° C reached 4.45 mA·cm-2, which was higher than that of Pt/C catalyst under the same conditions[36]. Interestingly, in this work, the catalytic performance of N-doped derived carbon was optimized only by controlling the morphology of CMPs precursor (from tubular to hollow spherical), which provided a design idea for the design of highly active ORR catalysts.
Lv et al. Pyrolyzed 1,3,5-triethynylbenzene self-polymerized CMPs in ammonia to prepare N-doped CMPs based derived carbon (N-HsGDY)[18]. By X-ray absorption near edge structure (XANES) analysis of N K-edge combined with density functional theory (DFT) calculation, only pyridine-type N,Jd is present in the N-HsGDY-900 ℃ sample with a value of 6.2 mA·cm-2, which is the highest value reported so far. To better understand the catalytic activity, DFT was used to compare the calculated ORR free energy diagrams of N-HsGDY and pyridine N-graphene at zero electrode potential (U = 0 V vs. NHE) and equilibrium potential (U = 0.455 V vs. NHE). It was found that the C atom of the ethynyl group adjacent to the benzene ring in the N-HsGDY structure is the active site for OOH*, O*, and OH* with a higher positive charge density (0.694 vs. 0.392). In alkaline medium, different charge distributions lead to different rate-determining steps. The adsorption of OOH* is the controlling step when N-HsGDY is used as a catalyst, while the desorption of OH* is the rate-determining step in pyridinic N-graphene. Comparing the free energy change of the control step of the two catalysts, the ORR diffusion resistance of N-HsGDY is significantly smaller than that of N-graphene at 0.455 V (0.87 eV vs. 0.94 eV) and 0 V (0.42 eV vs. 0.49 eV), which further verifies the excellent ORR activity of N-HsGDY.
N and S co-doped CMPs based derived carbon also showed better ORR performance, especially the N atom has higher electronegativity (3.04 vs. 2.55), which often acts as an electron donor in the molecular structure and endows the N-doped CMPs materials with redox activity. Bhosale et al. Introduced 2,4,6-tris (5-bromothiophen-2-yl) -1,3,5-triazine as N and S element donor monomer into the structure of CMPs through Sonogashira-Hagihara coupling reaction, and produced N and S double-doped ORR catalyst after pyrolysis[49]. Ren et al. Constructed new bis-heterocyclic CMPs as catalyst precursors through condensation cyclization between aldehyde group and thiocarboxamide group[43]. In these works, although the electronegativity of S is similar to that of C, the synergistic effect between S and N atoms and the difference of electron spin density of S atoms lead to the improvement of the overall ORR activity of carbon materials.
Porphyrin is a macrocyclic conjugated compound with 18π electrons. Porphyrin has a hole in the center of the ring, which can combine with a variety of metal ions (such as Fe, Co, Zn, Cu, etc.) To form metal complexes[50,51]. On the one hand, the network constructed by these conjugated macrocycles and aromatic groups has a porous and π-conjugated structure, which can ensure the exposure of active sites and electron transfer during electrocatalysis; On the other hand, the N4 coordination site in the porphyrin group structure can form the M-N4 active site. The introduction of metalloporphyrin building blocks into the network structure of CMPs by chemical bonding provides an effective way to prepare CMPs-based ORR catalysts doped with transition metals and N atoms. In addition, the advantage of using metalloporphyrins is that after carbonization, metal nanoparticles can usually be anchored on the nitrogen-doped carbon skeleton, which effectively avoids the agglomeration of metal nanoparticles. Li et al. Synthesized different metal (Co, Cu, Fe) porphyrin-based CMPs by Schiff reaction, and obtained monometal-doped ORR catalysts by pyrolysis at 700 ~ 900 ℃[38]. In the Co-doped porous carbon, the high content of Co (1. 1%) and N (4. 26%) synergistically improved the overall catalytic activity of the material, and the Eo and E1/2 were 0. 80 V and 0. 7 V, respectively, in the 0.5 mol/L H2SO4 test solution, which were close to those of the Pt/C catalyst. M Müllen et al. Prepared Fe, Co bimetallic doped carbon-based ORR catalyst (Fe/Co-CMP-800) by carbonized metalloporphyrin-based CMPs[41]. In this work, Fe and Co atoms were derived from two metalloporphyrin monomers, respectively, and simultaneously entered the structure of CMPs through Suzuki coupling reaction, and two active sites of FeN4 and CoN4 were formed after carbonization treatment (Fig. 5). The presence of CoN4 can inhibit the formation of low activity nanoparticles and increase the proportion of active FeN4 sites. In addition, the vicinal position of the CoN4 and FeN4 sites compensates for the poor catalytic selectivity of CoN4, promoting higher electron density for FeN4. Therefore, Fe/Co-CMP-800 combines the advantages of the physical characteristics of cobalt-based catalysts (discontinuous, ribbon structure) and the electrochemical characteristics of iron-based catalysts (high Eo value, low H2O2 evolution). Fe/Co-CMP-800 exhibited ORR performance close to that of Pt/C in acidic medium (half-wave potential (E1/2) was only 0.08 V lower) compared with the single transition metal doped ORR catalyst. Dou et al. Used CMPs doped with B, N, Co and Fe obtained by Sonogashira-Hagihara coupling reaction as precursors to prepare ORR catalysts. Through KSCN poisoning experiment and acid leaching experiment combined with XPS characterization analysis, it was considered that Co3Fe7 and Fe/Co-N were the active sites of the system[31].
图5 (a) Fe/Co-CMP, (b) Co-CMP和(c) Fe-CMP的合成和结构示意图及其SEM和TEM成像的纳米尺寸形貌[41]

Fig.5 Schematic representation of the synthesis and structure of (a) Fe/Co-CMP, (b) Co-CMP and (c) Fe-CMP and their nanosized morphology imaged by SEM and TEM[41]. Copyright 2015, Royal Society of Chemistry

3.1.2 Introduction of heteroatoms by blending of metal salts

In the post-doping design strategy based on CMPs, CMPs can act as either a carbon source or a heteroatom carrier. Among them, FeCl3 is a commonly used Fe source in the design of ORR catalyst. Sun et al. Prepared porphyrin-based CMPs, and physically ground and blended FeCl3 to form Fe-N-C structure after carbonization, which effectively improved the ORR activity of the material[35]. Dai et al. Used 1,2,4,5-benzenetetramine as N source to react with hexanone cyclohexane under acid catalysis to synthesize N-CMPs, which were then blended with phytic acid (N source) and carbonized at 900 ℃ and 1000 ℃ to prepare N, P-doped CMPs derived carbon (N, P-CMP-900 and N, P-CMP-1000)[45]. Affected by the monomer, this CMPs system contains a large number of O atoms, and the presence of these O atoms makes the carbonization process beneficial to the construction of micropores in the porous network of the catalyst, which acts like a C atom "etchant", and the BET surface area increases from 27 m2·g-1 to 171 m2·g-1. It showed good ORR activity in 0.1 mol/L KOH, 0.1 mol/L HClO4 and PBS solutions. In addition, this work is the first to use N and P co-doped electrocatalyst as a biosensor to detect dissolved oxygen in tears. Jiao et al. Used ionic liquids containing N and S as dopants and activators. S plays two roles in the material system: ① Polymer fragments containing S will decompose into small molecular substances under high temperature, which will escape from the system and form micropores, similar to the physical activation process, thus improving the specific surface area of the material[33]. (2) S in the liquid will interact with the carbon skeleton to form thiophene-S, which can change the electron distribution on the surface of porous carbon and cooperate with N to improve the catalytic activity of the material. It should be noted that the specific surface area and the amount of microporosity are not the only factors affecting the performance of ORR. In this system, C-CMPs-1 NP has the highest Eo and E1/2 values, but its Jd value is the lowest, which is due to the high sulfur content and partial mesopores in C-CMPs-1 NP, which is beneficial to current transport. Li et al. Prepared Fe, Co and N ternary doped graphene composite catalyst by introducing three heteroatoms of Fe, Co and N into CMPs through monomers containing iron porphyrin and cobalt porphyrin, and then blending and pyrolyzing with graphene oxide[39]. In this system, CMPs act as both carbon source and heteroatom dopant.
In metal-doped carbon-based catalysts, the agglomeration of metal particles is an important reason for the decrease of ORR performance of materials[52]. To overcome this problem, Hu et al. Mixed triazine-based CMPs with CoO and ZnCl2 by ball milling blending method, and prepared CoO, ZnO, N ternary doped CMPs-based porous carbon (CoO/ZnO @ N-PC) by assisted molten salt activation method (Fig. 6)[44]. In the reaction system, the triazine-based CMPs provide a N source and a C source, and the ZnCl2 can be used as a template agent and an activator to prevent pore collapse and promote the generation of a porous structure at high temperature; On the other hand, the generated ZnO can be used as a "dispersant" to effectively inhibit the agglomeration of CoO particles. Compared with the triazine-based CMPs direct carbonization materials, CoO/ZnO @ N-PC exhibited higher Eo(0.91 V) and E1/2 values (0.85 V). Wang et al. Prepared Fe-N-C carbon-based catalyst (BPCMP-Fe-800) based on the coordination of bipyridine N atom and Fe3+ in the unit structure of CMPs, and the doping amount of Fe element was more than 10 wt%[32]. Compared with the control sample (BBCMP-Fe-800) prepared by conventional blending method, the V vs. Of BPCMP-Fe-800 was improved to nearly 20% (0. 97 V vs. 0. 81 V).
图6 CoO/ZnO@N-PC的合成示意图[44]

Fig.6 Schematic illustration of synthesis of CoO/ZnO@N-PC[44]. Copyright 2020, Elsevier

3.2 COFs-derived carbon-based ORR catalyst

COFs are a kind of porous polymers with two-dimensional or three-dimensional structure formed by light elements such as C, N, B, O through covalent bonds, which are the only crystalline materials with highly ordered and regular structure in POPs family[32,53]. The organic monomer of COFs has rich designability. By changing the type, connection mode and spatial structure of the organic block structural unit, it can produce extended and highly crosslinked topology, so that two-dimensional or three-dimensional COFs often have high specific surface area and rich pores. However, COFs synthesized by conventional methods have poor conductivity, and usually need to be mixed with conductive materials (such as graphene, carbon black) or grown in situ on conductive substrates (such as carbon nanotubes, carbon paper, ITO glass) to improve the conductivity[54]. Direct pyrolysis of COFs materials provides an effective way to prepare efficient ORR catalysts. As mentioned above, the rich pore structure of COFs can provide more anchoring sites for metal nanoparticles or oxides after pyrolysis, and at the same time, it is helpful for the uniform dispersion of active substances to improve the utilization of active sites. In addition, the heteroatoms such as N, B and O naturally contained in the structure of COFs are retained after pyrolysis, which avoids the tedious step of doping modification by additional non-metallic heteroatoms. These characteristics make it possible for COFs to derive carbon-based ORR catalysts with catalytic properties.
Multi-atom doping modification (N, P, Fe, Co, etc.) is still a hot direction to improve the performance of COFs-based derived carbon ORR catalysts. Yang et al. Constructed four kinds of two-dimensional COFs materials through [4 + 4] connection, carbonized them at 1000 ℃, and then synthesized N and P double-doped COFs-based derived carbon by introducing P element into the Na2HPO2[55]. Among the four kinds of COFs, the carbon sample with 1,3,6,8-tetra- (p-aminophenyl) -pyrene and 1,2,4,5-tetrakis (4-formylphenyl) benzene as building monomer molecules has the highest P doping amount and graphite N content, accelerates the increase of charge transfer and active sites, and shows the optimal E1/2(0.81 V).

3.2.1 Metal chelating effect

A variety of metal and N-doped COFs-based derived carbons can be synthesized by using the strategy that pyridine and its derivatives can form stable complexes with transition metal ions such as Fe and Co. Park et al. Synthesized pyridine-linked triazine COFs (PTCOF) by nucleophilic substitution reaction using cyanuric chloride and diaminopyridine as monomers[56]. Compared with conventional COFs, the crystallinity of the synthesized PTCOF is lower, while the amorphous or low-crystalline carbon structure formed after pyrolysis is more conducive to ion migration and mass transport at the solid-liquid interface[57]. In order to adjust the electronic structure of PTCOF, Co-doped COFs-based porous carbon (CoNP-PTCOF) was prepared by blending PTCOF with Co(NO3)2 and carbonizing at 800 ℃. Due to the coordination effect, the Co2+ is fixed at the pyridine and triazine groups, so the Co nanoparticles are uniformly dispersed on the surface of the carbon skeleton after carbonization. In addition, due to the partial charge transfer from Co to the C framework after Co doping, the p-band center of the C active site in CoNP-PTCOF is downshifted compared with that in pristine PTCOF, which is more favorable for the oxygen intermediate to adsorb and desorb on the pyridine carbon active site, and the measured E1/2 is 0.87 V, which is close to that of Pt/C (0.85 V).
Among COFs-based M-N-C catalysts, single-atom catalysts reported in recent years have become the research frontier. Due to their special structure, M-N-C single-atom catalysts tend to exhibit higher atom utilization and significant catalytic activity[58]. At present, the mainstream synthesis method is to use wet chemistry to load metal ions onto COFs carriers for pyrolysis treatment[59,60]. Yang et al. Synthesized COFs precursors using 1,3,6,8-tetrakis (4-formylphenyl) pyrene and 2,2 '- (1,4-phenylene) diacetonitrile as monomers, and then Fe2+ ions (derived from iron porphyrin) were adsorbed on the surface and pores of COFs to form Fe-COF, which was finally pyrolyzed at high temperature to synthesize nanorod-derived carbon materials[60]. Fe in the material was found to have a Fe-N4 and Fe-Fe structure by XANES spectroscopy, and the Fe-Fe distance (2. 5252 Å) was much lower than the graphite interlayer spacing (~ 3.33 Å).Combined with high-angle annular dark-field scanning transmission electron microscopy, it was inferred that two Fe atoms coordinated two adjacent N atoms to form an approximately monoatomic structure. Using the synthetic strategy of "COF adsorption-pyrolysis", Zhang et al. First prepared COFs nanospheres by solution method using 1,3,5-tris (4-aminophenyl) benzene and 1,3,5-benzaldehyde as monomers. Secondly, COFs were used to adsorb Co, Fe and Pt ions in aqueous solution. Finally, the Pt, Fe, Co/N-C ternary monatomic catalyst is obtained through high-temperature pyrolysis[59]. On the one hand, the introduction of Fe and Co elements avoids the clustering or aggregation of metal atoms through the chelation with N element, thereby improving the stability of the catalyst, and the current loss is only 6%; On the other hand, it can optimize the central local coordination environment and electron distribution of Pt atom, weaken the adsorption of oxygen intermediate, and improve the ORR catalytic activity, with a E1/2 value of 0. 845 V in 0. 1 mol/L potassium hydroxide solution.

3.2.2 Template effect

Direct pyrolysis of COFs often leads to unexpected structural collapse and pore structure changes, which can be effectively avoided by using templates or compositing with substrate materials. Zhao et al. Pyrolyzed Fe, N-doped COFs-based mesoporous carbon (mC-TpBpy-Fe) using p-toluenesulfonic acid-assisted mechanochemical synthesis with trialdehydotriphenol and 2,2 '-bipyridine-5,5' -diamine as raw materials and SiO2 beads as hard templates[61]. The SiO2 sphere template plays an important role in promoting the formation of mesoporous structure, thereby improving the mass transfer and the utilization rate of active sites. At the same time, the bipyridine in the structure of COFs is coordinated with Fe3+, so that the Fe atoms after pyrolysis can be uniformly dispersed in the carbon skeleton, forming Fe-N active sites to improve the ORR performance. Chen et al. Used a similar synthetic strategy to introduce Co2+ into TpBpy-COFs[62]. The difference is that the surface of the SiO2 bead template is modified by amino groups in advance to increase the compatibility between the surface and the monomer molecules. After pyrolysis, the Co2+ was converted into Co subnanoclusters rather than Co-Nx bonds. Due to the strong synergistic effect between Co subnanoclusters and N/C, they proposed that the decomposition and adsorption of oxygen occurred simultaneously on the surface of the catalyst in this system, and OH* was directly generated without the intermediate step of OOH*.

3.2.3 Morphology control

Zhang et al. Reported a COFs-based derived carbon with core-shell structure (GC @ COF-NCT) for ORR[63]. ZIF-67 was first carbonized, and the organic linker in the ZIF-67 crystal was carbonized to form a graphitic carbon (GC) core during pyrolysis, while the Co2+ was converted to Co/CoO nanoparticles, which were transferred from within the ZIF- 67 structure to the surface of the carbon skeleton. Then a fibrous COFs layer (DAAQ-TFP-COF) was formed on the GC surface by Schiff reaction before further carbonization. The strong interfacial interaction between DAAQ-TFP-COF and ZIF-derived graphitic carbon provides a significant confinement effect, allowing the inner core GC layer to abut against the inner surface of COFs during the following pyrolysis process, resulting in a uniform hollow microporous or mesoporous structure. The ORR performance was affected by the pyrolysis temperature and the monomer concentration of DAAQ-TFP-COF, and the sample with the monomer concentration of 0. 04 mmol/L and carbonized at 800 ℃ had the best ORR performance with Eo~0.923 V,E1/2~0.841 V, which was close to that of Pt/C. Liu et al. Also used a similar synthetic strategy to prepare COF layer (TP-BPY-COF) -coated ZIF-derived carbon (Fig. 7)[64]. Different from the previous work, in this work, the ZIF was not carbonized in advance, and the COF layer was directly grown on the surface of ZIF-67. Compared with the ZIF-67 direct carbonization product (ZIF800), the core-shell COF @ ZIF800 has increased pyridine N and Co-N contents, and the Eo and E1/2 are higher than those of ZIF800 and Pt/C. They believed that the pyrolysis product of TP-BPY-COF had a stronger structural framework than that of ZIF-67, and the thin COFs shell not only inhibited the structural collapse of the overall structure of COF @ ZIF due to heat and promoted the formation of mesopores, but also effectively avoided the agglomeration of Co particles, which provided a reference for the design of composite COFs-based derived carbon ORR catalysts. Liu et al. Prepared COFs-derived carbon materials with flower-like morphology, and the carbon nanosheets stacked by a large number of graphite layers provided abundant edge sites for the catalytic reaction, which could expose more N atoms, thus showing excellent ORR catalytic activity. The ORR catalytic activity of COF800 in alkaline environment was comparable to that of commercial Pt/C catalyst, and the Eo, E1/2 and Jd reached 0.86 V, 0.79 V and 5.0 mA·cm-2, respectively[65].
图7 (a)在ZIF-67表面配合生长TP-BPY-COF合成COF@ZIF800催化剂的工艺示意图。(b) TP-BPY-COF的合成与结构。ZIF-67 (c)、COF@ZIF (d)、COF@ZIF800 (e)的SEM图像[64]

Fig.7 (a) Schematic of the synthesis procedure of COF@ZIF800 catalyst by in-situ growing TP-BPY-COF on the surface of ZIF-67. (b) The synthesis and structure of TP-BPY-COF. The SEM images of ZIF-67 (c), COF@ZIF (d), and COF@ZIF800 (e)[64]. Copyright 2022, Royal Society of Chemistry

3.3 HCPs derived carbon-based ORR catalyst

HCPs are a class of amorphous microporous polymer materials that have been developed rapidly in recent years and have attracted more and more attention[66]. Compared with other POPs materials, HCPs have the advantages of more extensive sources of building monomers (mainly aromatic monomers and vinyl monomers), cheap catalysts, mild synthesis conditions and high yields. In addition, the structure of most HCPs is composed of interconnected aromatic rings, and the higher specific surface area and porosity after pyrolysis contribute to the exposure of active sites and material transport[67,68]. At present, there are three main synthesis methods of HCPs: post-crosslinking polymer precursor method, one-step polycondensation method of functional monomer and external crosslinking agent braiding method[68]. External crosslinking agent braiding method is a new synthesis strategy proposed by Tan Bien et al. In 2011. This method mainly uses anhydrous FeCl3 as catalyst and formaldehyde dimethyl acetal (FDA) as crosslinking agent to construct polymer networks with microporosity and high specific surface area through Friedel-Crafts reaction[69]. This strategy has more friendly synthesis conditions and does not require high monomer functional groups, which is the mainstream preparation method of HCP-based ORR catalyst precursors.
At present, the HCP-based derived carbon used for ORR mainly includes N-doped, S-doped, N/S Co-doped, Fe (Co), N co-doped and other types. Aromatic compounds containing N and S atoms, such as aniline, pyrrole, carbazole, porphyrin and thiophene, are mainly used as monomers in the preparation of HCPs[70][71,72][73][74][75]. Transition metal atoms such as Fe and Co are usually introduced into the derived carbon structure by direct pyrolysis or Co-pyrolysis. Table 2 summarizes the electrochemical performance of HCPs-derived carbon ORR catalysts reported in the literature in the past five years.
表2 基于外部交联剂编织法的HCPs衍生碳基ORR催化剂的电化学性能汇总

Table 2 Summary of electrochemical properties of HCPs derived carbon-based ORR catalysts

Catalysts Heteroatom Monomer Eonset vs.
RHE (V)
E1/2 vs.
RHE (V)
Jd
(mA·cm-2)
Pt/C Jd
(mA·cm-2)
Electrolyte ref
N-PHCP-900 N Pitch / 0.883 / / 0.1 mol/L KOH 76
Fe/HCPs Fe Zeolitic imidazolate
frameworks
0.960 0.850 5.59 5.73 0.1 mol/L KOH 77
Fe/HCPs-etching 0.900 0.799 4.96
TPP-HCP-1 Fe, N 5,10,15,20-tetrakis (4-chlorophenyl) porphyrin 0.96 0.85 4.00 3.9 0.1 mol/L KOH 78
MPH-Fe/C Fe, N metalloporphyrin / 0.816 / / 0.1 mol/L KOH 74
PCF-HCP-900 Fe, N Pyrrole 0.95 0.84 4.8 5.2 0.1 mol/L KOH 72
NCP-An-900 Fe, N Aniline, pyrrole,
methylbenzene
0.96 0.85 -5.31 -5.20 0.1 mol/L KOH 70
Co-TSP-HCP-900 Co, N Carbazole 0.90 0.80 4.75 4.3 0.1 mol/L KOH 73
Fe-TSP-HCP-900 Fe, N 0.89 0.76 4.50
PPFeC-800 Fe, N Pyrrole, thiophene 0.977 0.833 5.15 / 0.1 mol/L KOH 71
PTFeC-800 Fe,S 0.942 0.825 5.17
MixFeC-800 Fe, N, S 0.983 0.844 5.07
HCP-NSZn-900 Fe, N, S Triazine derivative 0.98 0.86 4.72 4.59 0.1 mol/L KOH 75
0.85 0.68 / / 0.1 mol/L HClO4
FeCoP/NPC Fe, Co, N Poly(bis(N-carbazolyl)-1,2,4,5-tetrazine) 0.948 0.855 5.23 / 0.1 mol/L KOH 79
HCP-NT-NH3-800 O, S Hexakis(benzylthoi)
benzene, thiophene
1.01 0.85 4.99 / 0.1 mol/L KOH 80

3.3.1 Carbazolyl HCPs

Zhu et al. Prepared carbazole-based HCPs with carbazole as monomer and FDA as external crosslinking agent[73]. By optimizing the monomer/catalyst/crosslinker ratio and pyrolysis temperature, the best ORR performance (Eo~0.89 V,E1/2~0.78 V) of TSP-HCP-900 was obtained, although there is still a certain distance between this value and Pt/C, which provides an idea for the preparation of N-containing monomers required for N-doped HCPs. In addition, doping transition metal elements such as Fe and Co into the structure of carbazole-based HCPs is also an effective strategy to improve the electrocatalytic performance. Under the same test conditions, the Eo and E1/2 values of Co-TSP-HCP-900 were increased by 10 mV and 20 mV, respectively. Chen et al. Used a similar design idea to synthesize carbazole-based HCPs, using bis- (N-carbazolyl) -1,2,4,5-tetrazine as a monomer and AlCl3 as a catalyst to first synthesize carbazole-based HCPs (HCTCz @ PA), adsorbing phytic acid (P source), then introducing Fe3+ and Co2+ through coordination, and pyrolyzing to obtain Fe, Co, P, N-doped porous carbon (FeCoP/NPC)[79]. They systematically compared the effects of Fe, Co, and phytic acid on the catalytic activity of the materials, and found that in the FeCoP/NPC structure, Fe and Co mainly exist in the form of FeP and CoP, rather than the form of M-Nx reported in the general literature. The Eo and E1/2 of FeCoP/NPC are 0.948 V and 0.855 V, respectively. Although they are slightly lower than those of Pt/C, their lower Tafel slope (58 mV·dec-1vs.77 mV·dec-1) indicates that FeCoP/NPC has a faster reaction kinetic rate during ORR.

3.3.2 Asphalt based HCPs

Pitch is a by-product of petroleum and coal chemical industry. Yang et al. Synthesized pitch-based HCPs by Friedel-Crafts reaction with FDA as crosslinking agent, and prepared N-doped HCP-based ORR catalyst (N-PHCP-x) by high temperature NH3 activation process[76]. During pyrolysis at high temperature, ammonia acts as an etchant, and the surface area of N-PHCP-900 is up to 992 m2·g-1, which increases the exposure of active sites. In addition, it can be seen from the pore size distribution that there are a large number of micropores and continuous mesopores in the sample, which is beneficial to the mass transfer and electrolyte diffusion in the ORR, so that the E1/2 value of N-PHCP-900 shows a positive shift of 10 mV compared with that of the commercial Pt/C catalyst, reaching 0. 883 V.

3.3.3 Porphyrin-based HCPs

Compared with metalloporphyrin-based MOFs or CMPs, the construction process of porphyrin-based HCPs is simpler. Tan et al. Prepared metal (Fe, Co) porphyrin-based HCPs by external crosslinker braiding method, and prepared carbon-based porous materials (MPH-Fe/C and MPH-Co/C) with high specific surface area after KOH activation; the specific surface area of MPH-Fe/C was as high as 2768 m2·g-1,Fe, and the Co content was as high as 9.97 wt% and 10.35 wt%, respectively, showing ORR performance comparable to that of Pt/C[74]. Zhu et al. Prepared porphyrin-based HCPs (TPP-HCPs) using 5,10,15,20-4 '-chlorophenyl porphyrin as a monomer, and obtained TPP-HCP-based derived carbon after one-step pyrolysis[78]. They found that when the FeCl3 and FDA dosage were increased, the catalyst could obtain higher specific surface area at the same pyrolysis temperature, thus optimizing the ORR performance.

3.3.4 Morphology control

Most of the HCPs-based derived carbons are amorphous or granular, while there are relatively few reports on special morphologies. Qian et al. Prepared hollow tubular HCPs (HCP-NT) with hexakis (benzylthio) benzene and thiophene as monomers, chloroform as crosslinking agent, and FeCl3 as catalyst by adjusting the amount of crosslinking agent[80]. The hollow tubular structure remains intact after carbonization. In the preparation of HCP-based derived carbon, the researchers used three methods, as shown in Figure 8: ① Direct inert gas pyrolysis to obtain S-doped material (HCP-NT-Ar-800). Pyrolysis under ②NH3 atmosphere, S atoms were almost replaced by N atoms, and the specific surface area of HCP-NT-NH3-800 was greatly increased due to the etching effect of NH3, forming C-N active sites. Due to the difference of electron density caused by doping atoms and the different porous structure of the samples, the Eo and E1/2 of the HCP-NT-NH3-800 are higher than those of HCP-NT-Ar-800, reaching 1. 01 V and 0. 85 V, respectively. (3) The surface of HCP-NT-Ar-800 was modified by Ar Plasma to obtain S and O co-doped HCP-NT-Plasma, which not only increased the oxygen content in the structure, but also introduced additional defect sites. Unlike the previous two samples, the 2e-ORR selectivity of HCP-NT-Plasma is more than 80%, which is attributed to the increase of C = O and COOH species in the structure after Plasma treatment.
图8 超交联聚合物纳米管(HCP-NT)和衍生多孔碳纳米管的示意图[80]

Fig.8 Schematic illustration of the hyper-cross-linked polymer nanotubes (HCP-NT) and derived porous carbon nanotubes[80]. Copyright 2023, Elsevier

3.3.5 Catalyst action

Compared with other POPs materials, another advantage of HCPs-based derived carbon ORR catalyst is that in this system, the FeCl3 is not only the catalyst for building HCPs network, but also acts as metal dopant and pore-forming agent. Zhang et al. Systematically compared the performance of Fe, N and S heteroatoms on HCPs-derived carbon-based ORR catalysts[71]. Supercrosslinked polypyrrole and polythiophene were synthesized by Friedel-Crafts reaction using similar pyrrole and thiophene as monomer molecules, and Fe-N-based, Fe-S-based and Fe-N-S-based catalysts were obtained after pyrolysis at different temperatures. It was found that Fe in the PPFeC-800 structure prepared with pyrrole as a monomer mainly exists in the form of Fe3O4, while Fe1-xS nanocrystals mainly exist in the PTFeC-800 structure prepared with thiophene as a monomer. They suggested that Fe element is favorable for N doping, and the stronger electronegativity of N atom changes the electronic properties of the neighboring C atom or transition metal atom, which contributes to O2 adsorption, so N/Fe3O4 is the catalytic active site of PPFeC-800. In the PTFeC-800 structure, S/Fe1-xS reduces the catalytic activity of the material for ORR, but S doping can change the spin density distribution near the Fe center and reduce the activation barrier on Fe/N/C, so the Fe-N-S based catalyst (MixFeC-800) shows the highest Eo value (0.983 V) and E1/2 value (0.844 V) among the catalysts with three compositions.

3.4 CTFs-based ORR catalyst

In 2008, Thomas et al., Technical University of Berlin, Germany, proposed the concept of CTFs, which are a class of POPs materials composed of rigid triazine rings connected by covalent bonds, belonging to the branch of COFs materials, with excellent physical and chemical stability and thermal stability (comparable to CMPs)[81][82]. Compared with other POPs materials that need to introduce additional N element to improve the catalytic activity of materials, the framework structure of CTFs is naturally rich in N element, and the purpose of adjusting the content and type of N can be achieved by selecting appropriate aromatic nitriles as monomers combined with trimerization reaction, and predictable two-dimensional porous structure and three-dimensional porous network can be obtained at the same time. In 2014, CTFs were first applied in the field of electrocatalysis as ORR catalysts[83]. However, the conductivity of CTFs itself is poor. In order to improve its conductivity, carbonized CTFs is an ideal choice. The N-rich CTFs-derived carbon can be directly used as an ORR catalyst without doping additional metal elements[84]. In addition, CTFs are usually compounded with carbon black, graphene, etc. To further improve the overall conductivity of the material[84,85]. Cao et al. Mixed part of carbon black (Super P) in the synthesis of CTFs[86]. The addition of carbon black effectively avoids the agglomeration of CTFs during pyrolysis, thus increasing the specific surface area and porosity of the material, thereby affecting the ORR performance of the carbonized catalyst. Jiao et al. Pyrolyzed CTF/rGO composite at 950 ° C to prepare NHC/rGO composite with "sandwich" structure[87]. Compared with commercial Pt/C, NHC/rGO-950 has a higher Jd value in alkaline solution.
Blending with metal salts and pyrolysis to dope metal atoms into the carbon skeleton is also an effective means to improve the catalytic performance of CTFs-based derived carbon for ORR. Zhang et al. Prepared CTFs by ZnCl2 induced ionothermal method using isophthalonitrile as monomer, and then pyrolyzed by blending with FeCl3·6H2O[88]. The mixing ratio of FeCl3 and CTFs, pyrolysis temperature have a direct effect on the ORR of the catalyst. When the FeCl3·6H2O/CTF is 1 ∶ 0.5 and the pyrolysis temperature is 900 ℃, the open circuit potential is the highest. They suggested that FeCl3 is the main active site of the catalyst, which accelerates the cleavage of the O — O bond in *OOH, follows the 4e- reaction path, and favors the ORR process. Table 3 summarizes the electrochemical performance of CTFs-derived carbon ORR catalysts.
表3 CTFs衍生碳基ORR催化剂的电化学性能汇总

Table 3 Summary of electrochemical properties of CTFs derived carbon-based ORR catalysts

Catalysts Heteroatom Method Eonset vs.
RHE (V)
E 1 / 2 vs.
RHE (V)
Jd
(mA·cm-2)
Electrolyte ref
NHC/rGO-950 N Situ “bottom-up” trimerization 0.95 0.83 -5.64 0.1 mol/L KOH 87
N-HCNFs-2-1000 N Step-wise polymerization 1.01 0.84 5.56 0.1 mol/L KOH 89
CTF-CSU1 N Bottom-up technology 0.79 0.57 5.6 0.1 mol/L KOH 90
CTF-Super P-10 N Ionothermal synthesis 0.981 0.883 5.31 0.1 mol/L KOH 86
0.840 0.717 5.40 0.1 mol/L HClO4
BINOL-CTF-10-500 N Ionothermal synthesis 0.793 0.737 / 0.1 mol/L KOH 91
BINOL-CTF-5-400 0.758 0.684 /
BINOL-CTF-5-500 0.760 0.688 /
BINOL-CTF-10-400 0.737 0.659 /
DCBP-750 N Ionothermal synthesis 0.90 0.79 -5.1 0.1 mol/L KOH 92
PDCB-750 0.88 0.75 -6.2
PDCB-600 0.84 0.68 -6.6
PDCB-400 0.65 / -2.0
PDCB 0.84 0.68 -6.6
DCBP 0.85 0.71 -6.3
FB7 Fe, N Ionothermal synthesis 0.871 0.757 / 0.1 mol/L HClO4 93
Fe-C3N3-750 Step-wise polymerization subse, quent pyrolysis, NH3 activation 0.956 0.794 4.75 0.1 mol/L KOH 85
Fe-C3N3-750-NH3 1.03 0.840 4.82
TEBCB-Fe-N/S/C Fe, N, S Post-polymerization 0.899 0.79 / 0.1 mol/L HClO4 94
Cu-CTF/CP Cu, N Hybridization 0.81 / / Phosphate buffer
solution
95
Co-CTF/KB Co, N Ketjen Black hybridization / 0.83 6.14 0.1 mol/L KOH 96
NPF-CNS-2 N, P, F Self-templated carbonization strategy 0.90 0.81 5.42 0.1 mol/L KOH 97
0.82 0.69 5.01 0.1 mol/L HClO4
0.83 0.70 5.03 0.5 mol/L H2SO4
0.70 0.58 4.23 0.1 mol/L PBS

3.4.1 Construction of M-N-C catalyst

The CTF framework structure has high nitrogen content, which can provide abundant nitrogen coordination sites to anchor the filter metal at a fixed position while changing the electron distribution of the carbon material, so as to construct the M-N-C type ORR catalyst, realize the uniform dispersion of transition metal atoms and improve the utilization rate. Zhu et al. Designed a CTFs with abundant N4structural units based on o-phenanthroline monomer to anchor Fe3+,After pyrolysis treatment, the Fe element in the sample is uniformly distributed and maintains the FeN4 configuration (there is C-FeN4(—O) or C-FeN4(—OH)), but with the increase of carbonization temperature, the non-planarity of FeN4/C increases significantly, which makes the 2e- selectivity decrease gradually[98]. In the free energy diagram (Fig. 9), the intensity of *OOH adsorption increases obviously with the introduction of Fe. At the same time, with the increase of non-planarity of Fe/N/C structure, the adsorption of (2-C-FeN44-C-FeN46-C-FeN4),*OOH on C site becomes stronger, and the reaction pathway switches from 2e- to 4e-. This work shows that the electronic structure of the catalyst can be adjusted by controlling the pyrolysis temperature to optimize the reaction path, and also explains the reason for the improvement of the ORR performance of the catalyst by Fe doping from the structural change.
图9 C-N4、C-FeN4(O)和C-FeN4(OH)的结构模型(左)及不同压缩应变下C位点上的2e-反应途径自由能变化(右)[98]

Fig.9 Structural models of C-N4, C-FeN4(O) and C-FeN4(OH) (left), and the corresponding free energy diagram of the 2e- pathway at C sites of varying compressive strain[98]. Copyright 2021, American Chemical Society

Zhu et al. Developed a series of S-doped Fe-N-C catalysts with six different aromatic nitriles as building blocks[94]. Different from the traditional ZnCl2 molten salt method, the researchers used the molten mixture of FeCl3 and S to prepare CTFs. Molten FeCl3 can not only be used as a strong Lewis acid to catalyze the polymerization of CTFs, but also can be mixed with CTFs at the molecular level to become the Fe source of Fe-N/C catalyst, and FeCl3 can be used as a pore-forming agent in the pyrolysis process to optimize the pore structure of carbon materials. In the reaction process, S element can alleviate the decomposition of Fe-Nx active sites at high temperature and improve the ORR performance of Fe-N-C catalyst. They found that the number ratio of nitrile group to benzene ring (Nnitrile/Nbenzene) in the structural module had a great influence on the ORR activity of the catalyst. DCBP-Fe-N/S/C has the highest measured N/C ratio, but its catalytic performance is the worst. However, DCP-Fe-N/S/C has the lowest measured N/C ratio, but shows moderate ORR performance. The smaller the Nnitrile/Nbenzene, the fewer the crosslinking nodes (triazines) and the more benzene rings in CTFs, resulting in an increase in the degree of graphitization of the final catalyst during pyrolysis. Therefore, the aromatic ring is a necessary condition for Fe-N-C catalyst to obtain high ORR activity in thermal treatment.
Dun et al. Constructed a CTF-derived Fe/Co bimetallic single-atom electrocatalyst (Fe/Co-N-C, Fig. 10) using isophthalonitrile as monomer and CoCl2 and FeCl3 as metal sources by sequential chemical doping and complexation adsorption strategy[99]. The porous structure and abundant N atoms of the CTFs precursor can provide suitable sites for the anchoring of Fe and Co. Compared with the mono-metal catalyst, the atomically dispersed binary metal catalyst can enhance its ORR activity by optimizing and modifying the local geometric and electronic structure, and the V vs. And E1/2 reach 1. 02 V and 0. 878 V, respectively, showing a stable open circuit voltage (1. 52 V vs. 1. 47 V) superior to that of the commercial Pt/C catalyst in Zn-air battery.
图10 基于CTFs的Fe/Co双金属单原子催化剂的制备流程[99]

Fig.10 Preparation process of Fe and Co bimetallic monatomic catalyst based on CTFs[99]. Copyright 2023, Royal Society of Chemistry

3.4.2 Morphology control

Zheng et al. Prepared hollow nanoflower CTFs-based carbon materials with high N-doping content through a simple synthesis strategy[89]. The key of this synthesis strategy is to attach the CTFs monomer (cyanuric chloride) to the melamine-cyanuric acid (MCA) substrate through covalent bonds, which serves as an anchor point for the subsequent growth of CTFs on the MCA surface to avoid random growth in the liquid reaction medium. During the pyrolysis process, due to the structural stability of CTFs, the hollow nanoflower morphology is retained after the MCA crystal is decomposed into organic vapor. In the same year, they rapidly synthesized N, P, F tri-doped carbon nanospheres (NPF-CNSs) with spherical morphologies by ultrasound-assisted precipitation polycondensation based on the self-templated carbonization strategy[97]. According to XPS analysis, the content of heteroatoms in the material is also higher than that in other traditional synthesis methods (F ~ 0.57 at%, P ~ 1.21 at%, N ~ 3.89 at%). At the same time, this work is also the first time that F element is introduced into the carbon structure derived from CTFs, on the one hand, the surface hydrophilicity of the bulk material is improved, on the other hand, the fluorination treatment of the carbon matrix induces the formation of ionic C — F bonds (684. 9 eV) and semi-ionic C — F bonds (687. 8 eV).The increased adsorption of oxygen molecules, which in turn stimulates the cleavage of O — O bonds, enhances the electrocatalytic performance of NPF-CNS, with Eo and E1/2 values of 0.93 V and 0.81 V, respectively.

3.5 PIMs derived carbon-based ORR catalyst

PIMs is a POPs material reported by McKeown and Budd of the University of Manchester, UK. PIMs are usually formed by the polymerization of a monomer containing a spiro structure or a monomer containing a non-planar structure with another rigid monomer. Due to the various rigid macromolecular chains and distorted structures in the molecule, the macromolecules can not effectively occupy the internal space, resulting in a large number of micropores in the concave part (most of the pores are below 2 nm).PIMs have unique advantages in the field of electrocatalysts because of their high specific surface area, but compared with CMPs, COFs and other materials, it is still difficult to precisely design and control the pore size of PIMs[100]. At present, there are not many reports of ORR based on PIMs-derived carbon. In the work reported in recent years, doping PIMs-based ORR catalysts with N, Co and other elements is the main way to improve the catalytic performance. Metal ion adsorption is the conventional method of metal doping, especially the ingenious selection of appropriate ligands, the introduction of coordination sites in the molecule to coordinate with the metal (such as Co) has gradually become the main direction. Quernheim et al. Selected phenanthroline-indole macrocycle (PIM) with chelating property as ligand and cobalt acetate as metal source to form cobalt complex Co-3c by acetate method[100]. Co-3c shows very high ORR performance in alkaline solution, while the ligand 3C/C itself has very low activity for ORR, which illustrates that the introduction of Co can effectively increase the active sites and promote the direct 4e- reduction of oxygen. In addition, Patil et al. First used a combination of electrospinning and pyrolysis to spin amide, amine, and amidoxime-functionalized PIMs into ultrafine fibers and pyrolyzed them to obtain nitrogen-doped carbon fibers, which were used as ORR electrocatalysts[101]. This work focuses on the comparison of the effects of different functional groups on the properties of PIM-1 based nitrogen-doped carbon fibers. Among them, the nitrogen-doped carbon fiber in hydrolyzed PIM-1 has the characteristics of high specific surface area (430 m2·g-1), small pore size, high pyridine nitrogen and graphite nitrogen (34. 29%, 54.57%), and the highest E1/2 value (0. 985 V). Taking advantage of the rigid structure of PIMs, He et al. Used them as a protective agent for other highly active catalysts (Pt/C catalyst containing 40% Pt on Vulcan-72), which made the surface transfer rate of Pt/C catalyst high, solved the common agglomeration phenomenon of Pt nanoparticles, significantly reduced the loss of Pt, and made up for the shortcomings of Pt/C in durability and stability[102].

3.6 PAFs derived carbon-based ORR catalyst

PAFs are porous polymers constructed from rigid phenyl units with extremely high surface area[103]. Compared with traditional porous polymer materials, the interconnection of C-C covalent bonds makes PAFs extremely stable and less prone to structural changes during carbonization[104,105]. However, at present, the synthesis cost of PAFs is relatively expensive, which often leads to the decrease of porosity when functionalizing PAFs with high specific surface area, and the residual noble metal catalyst in the synthesis process will also block the pores, so the application of PAFs in energy storage and ORR is still in its infancy[106][104]. In 2015, Xiang et al. Used tetrakis (4-bromophenyl) methane as a monomer to synthesize PAFs containing only C and H elements, and pyrolyzed them under NH3 atmosphere to prepare N-doped PAFs-derived Carbon-based ORR catalyst (N-PAF-carbon)[107]. Compared with the sample without NH3 treatment, the E1/2 of N-PAF-Carbon was enhanced by 70 mV, the limiting current was increased by 80%, and the electrocatalytic activity was significantly enhanced, which was comparable to that of commercial Pt/C.

4 Conclusion and prospect

Porous organic polymers (POPs) are an attractive class of porous materials with rich porous structure and high designability, which provide unique conditions for the design and construction of non-Pt-based ORR electrocatalysts. In this paper, the application of six kinds of POPs materials (COFs, HCPs, CMPs, PIMs, CTFs and PAFs) as catalyst precursors in the oxygen reduction reaction of fuel cells was systematically summarized, and the research progress of preparation methods, active site construction and structural analysis, morphology control and catalytic mechanism of different pyrolysis POPs derived catalysts were emphatically discussed. From the perspective of material design, there is still much room for development in this research field, and it is likely to make breakthroughs in the following aspects in the future.
(1) The rational design and efficient synthesis of precursors are the key factors affecting the performance and application of pyrolytic oxygen reduction catalysts. However, most POPs materials have high monomer cost or undergo complex synthesis processes, so improving the existing synthesis methods or exploring new synthesis strategies is an urgent task to achieve the development of economical and low-cost POPs-based catalysts.
(2) Macro-scale morphology control and structure design of POPs are of great practical significance for the large-scale preparation of catalyst materials and the optimization of catalytic performance of actual battery modules in the future. Most of the POPs materials prepared by conventional synthesis methods are amorphous powders, although the POPs-based derived carbon materials with different dimensions and micro-morphologies have been prepared in small batches by means of regulating the monomer assembly polymerization process or post-auxiliary processing.However, the regulation of morphology assembly and the rational construction of multi-level structure are still important research topics in this field, and the relationship between morphology and electrocatalytic performance still needs to be further explored.
(3) The pyrolysis process will damage the structure of POPs precursors, which makes it difficult to understand the structure of active sites. Although DFT calculations provide more references for understanding the structure of the active site, the correlation with the experimental results is not clear, and new research methods are still needed to establish the structure-property relationship.
(4) The time/space distribution of reactive oxygen species (ROS) in the ORR process is the basis for establishing the reaction rate equation and studying the catalytic reaction mechanism. Recent studies have found that the temporal/spatial distribution of ROS in the diffusion layer of electrocatalysts is closely related to the durability of electrocatalysts[108]. However, the current testing methods are still challenging to quantitatively explore the intermediate species, and the development and design of new characterization and detection methods to obtain the ROS information of the catalytic reaction has become an important direction for the study of ORR catalytic mechanism and the design of POPs-based catalysts.
(5) At present, POPs-derived carbon-based catalysts show good ORR performance in alkaline medium, but the research in acidic medium is obviously insufficient. However, some fuel cells (such as proton exchange membrane fuel cells) are more suitable for operation in acidic medium. Therefore, the development of POPs-based ORR catalysts with excellent catalytic activity and stability in acidic medium is an important direction for future development.
(6) Under the actual working conditions of fuel cell, the catalyst coating, electrolyte, actual reaction temperature and cell structure will affect the oxygen reduction process, so higher requirements are put forward for the catalytic performance of POPs-based derived carbon-based catalysts under actual working conditions.
[1]
Wang Q, Han N, Bokhari A, Li X, Cao Y, Asif S, Shen Z F, Si W M, Wang F G, Klemeš J J, Zhao X L. Energy, 2022, 255: 124465.

[2]
Zhao G Q, Rui K, Dou S X, Sun W P. Adv. Funct. Mater., 2018, 28(43): 1803291.

[3]
Zaman S, Huang L, Douka A I, Yang H, You B, Xia B Y. Angewandte Chemie Int. Ed., 2021, 60(33): 17832.

[4]
Wei C, Rao R R, Peng J Y, Huang B T, Stephens I E L, Risch M, Xu Z J, Shao-Horn Y. Adv. Mater., 2019, 31(31): 1806296.

[5]
Wang H J, Ren H, Liu S L, Yin S L, Jiao S Q, Xu Y, Li X N, Wang Z Q, Wang L. Chem. Eng. J., 2022, 438: 135539.

[6]
Huang L, Wei M, Qi R J, Dong C L, Dang D, Yang C C, Xia C F, Chen C, Zaman S, Li F M, You B, Xia B Y. Nat. Commun., 2022, 13: 6703.

[7]
Li S W, Liu J J, Yin Z, Ren P J, Lin L L, Gong Y, Yang C, Zheng X S, Cao R C, Yao S Y, Deng Y C, Liu X, Gu L, Zhou W, Zhu J F, Wen X D, Xu B J, Ma D. ACS Catal., 2020, 10(1): 907.

[8]
Xia W, Mahmood A, Liang Z B, Zou R Q, Guo S J. Angewandte Chemie Int. Ed., 2016, 55(8): 2650.

[9]
Lin L X, Miao N H, Wallace G G, Chen J, Allwood D A. Adv. Energy Mater., 2021, 11(32): 2100695.

[10]
Wei W, Ge H T, Huang L S, Kuang M, Al-Enizi A M, Zhang L J, Zheng G F. J. Mater. Chem. A, 2017, 5(26): 13634.

[11]
He M, Song S J, Wang P, Fang Z, Wang W W, Yuan X L, Li C Y, Li H, Song W Y, Luo D, Li Z X. Carbon, 2022, 196: 483.

[12]
Feng L Y, Wang T T, Sun H, Jiang M, Chen Y G. ACS Appl. Mater. Interfaces, 2020, 12(51): 56954.

[13]
Xue W D, Zhou Q X, Cui X, Jia S R, Zhang J W, Lin Z Q. Nano Energy, 2021, 86: 106073.

[14]
Yang D H, Tao Y, Ding X S, Han B H. Chem. Soc. Rev., 2022, 51(2): 761.

[15]
Liu X, Liu C F, Xu S H, Cheng T, Wang S, Lai W Y, Huang W. Chem. Soc. Rev., 2022, 51(8): 3181.

[16]
Wroblowa H S, Pan Y C, Razumney G. J. Electroanal. Chem. Interfacial Electrochem., 1976, 69(2): 195.

[17]
Guo D H, Shibuya R, Akiba C, Saji S, Kondo T, Nakamura J. Science, 2016, 351(6271): 361.

[18]
Lv Q, Si W Y, He J J, Sun L, Zhang C F, Wang N, Yang Z, Li X D, Wang X, Deng W Q, Long Y Z, Huang C S, Li Y L. Nat. Commun., 2018, 9: 3376.

[19]
Yeager E. Electrochim. Acta, 1984, 29(11): 1527.

[20]
Liu M Y, Xiao X D, Li Q, Luo L Y, Ding M H, Zhang B, Li Y X, Zou J L, Jiang B J. J. Colloid Interface Sci., 2022, 607: 791.

[21]
Kim H W, Bukas V J, Park H, Park S, Diederichsen K M, Lim J, Cho Y H, Kim J, Kim W, Han T H, Voss J, Luntz A C, McCloskey B D. ACS Catal., 2020, 10(1): 852.

[22]
Zhang J T, Xia Z H, Dai L M. Sci. Adv., 2015, 1(7): e1500564.

[23]
Zhou R F, Zheng Y, Jaroniec M, Qiao S Z. ACS Catal., 2016, 6(7): 4720.

[24]
Li Z, Yang Y W. J. Mater. Chem. B, 2017, 5(47): 9278.

[25]
Jiang J X, Su F B, Trewin A, Wood C, Campbell N, Niu H J, Dickinson C, Ganin A, Rosseinsky M, Khimyak Y, Cooper A. Angew. Chem. Int. Ed., 2007, 46(45): 8574.

[26]
Liu W B, Wang K, Wang C M, Liu W P, Pan H H, Xiang Y J, Qi D D, Jiang J Z. J. Mater. Chem. A, 2018, 6(45): 22851.

[27]
Roy S, Bandyopadhyay A, Das M, Ray P P, Pati S K, Maji T K. J. Mater. Chem. A, 2018, 6(14): 5587.

[28]
Isci R, Balkan T, Tafazoli S, Sütay B, Eroglu M S, Ozturk T. ACS Appl. Energy Mater., 2022, 5(11): 13284.

[29]
He Y F, Gehrig D, Zhang F, Lu C B, Zhang C, Cai M, Wang Y Y, Laquai F, Zhuang X D, Feng X L. Adv. Funct. Mater., 2016, 26(45): 8255.

[30]
Bandyopadhyay S, Boukhvalov D W, Nayak A K, Ha S R, Shin H J, Kwon J, Song T, Choi H. Dyes Pigments, 2019, 170: 107557.

[31]
Dou J L, Luo H T, Zhang C L, Lu J J, Luan X J, Guo W X, Zhang T, Bian W W, Bai J K, Zhang X L, Zhou B L. New J. Chem., 2021, 45(45):21020.

[32]
Wang M Y, Cao Z L, Li L Y, Ren S J. Carbon, 2022, 200: 337.

[33]
Jiao R, Zhang W L, Sun H X, Zhu Z Q, Yang Z F, Liang W D, Li A. Mater. Today Energy, 2020, 16: 100382.

[34]
Zhang W L, Sun H X, Zhu Z Q, Jiao R, Mu P, Liang W D, Li A. Renew. Energy, 2020, 146: 2270.

[35]
Sun H X, Zhou P L, Tian Z Y, Ye X Y, Zhu Z Q, Ma C H, Liang W D, Li A. ChemPlusChem, 2022, 87(7): e202200168.

[36]
Sun H X, Zhou P L, Ye X Y, Wang J J, Tian Z Y, Zhu Z Q, Ma C H, Liang W D, Li A. J. Colloid Interface Sci., 2022, 617: 11.

[37]
Zhou B L, Liu L Z, Cai P W, Zeng G, Li X Q, Wen Z H, Chen L. J. Mater. Chem. A, 2017, 5(42): 22163.

[38]
Li Q, Shao Q, Wu Q, Duan Q, Li Y H, Wang H G. Catal. Sci. Technol., 2018, 8(14): 3572.

[39]
Li Y, Tao X S, Wei J C, Lv X L, Wang H G. Chem. Asian J., 2020, 15(13): 1970.

[40]
Yang J, Xu M, Wang J Y, Jin S B, Tan B E. Sci. Rep., 2018, 8: 4200.

[41]
Brüller S, Liang H W, Kramm U I, Krumpfer J W, Feng X L, Müllen K. J. Mater. Chem. A, 2015, 3(47): 23799.

[42]
Zhu Z Q, Yang Z F, Fan Y K, Liu C, Sun H X, Liang W D, Li A. ACS Appl. Energy Mater., 2020, 3(6):5260.

[43]
Ren S B, Chen X L, Li P X, Hu D Y, Liu H L, Chen W, Xie W B, Chen Y X, Yang X L, Han D M, Ning G H, Xia X H. J. Power Sources, 2020, 461: 228145.

[44]
Hu L L, Gu S, Yu W G, Zhang W J, Xie Q J, Pan C Y, Tang J T, Yu G P. J. Colloid Interface Sci., 2020, 562: 550.

[45]
Lin D Q, Hu C G, Chen H, Qu J, Dai L M. Chem. A Eur. J., 2018, 24(69): 18487.

[46]
Ma L, Liu Y L, Liu Y, Jiang S Y, Li P, Hao Y C, Shao P P, Yin A X, Feng X, Wang B. Angew. Chem. Int. Ed., 2019, 58(13): 4221.

[47]
Chen J, Yan W, Townsend E J, Feng J T, Pan L, Del Angel Hernandez V, Faul C F J. Angew. Chem. Int. Ed., 2019, 58(34): 11715.

[48]
Wang Z J, Yang X Y, Yang T J, Zhao Y B, Wang F, Chen Y, Zeng J H, Yan C, Huang F, Jiang J X. ACS Catal., 2018, 8(9):8590.

[49]
Bhosale M E, Illathvalappil R, Kurungot S, Krishnamoorthy K. Chem. Commun., 2016, 52(2): 316.

[50]
Abdulaeva I A, Birin K P, Bessmertnykh-Lemeune A, Tsivadze A Y, Gorbunova Y G. Coord. Chem. Rev., 2020, 407: 213108.

[51]
Park J M, Hong K I, Lee H, Jang W D. Acc. Chem. Res., 2021, 54(9): 2249.

[52]
Xia W, Zou R Q, An L, Xia D G, Guo S J. Energy Environ. Sci., 2015, 8(2): 568.

[53]
Lohse M S, Bein T. Adv. Funct. Mater., 2018, 28(33): 1705553.

[54]
Bu R, Lu Y Y, Zhang B. Chem. Res. Chin. Univ., 2022, 38(5):1151.

[55]
Yang C, Tao S S, Huang N, Zhang X B, Duan J G, Makiura R, Maenosono S. ACS Appl. Nano Mater., 2020, 3(6):5481.

[56]
Park J H, Lee C H, Ju J M, Lee J H, Seol J, Lee S U, Kim J H. Adv. Funct. Mater., 2021, 31: 2101727.

[57]
Li Z, Liu Z W, Li Z Y, Wang T X, Zhao F L, Ding X S, Feng W, Han B H. Adv. Funct. Mater., 2020, 30: 1909267.

[58]
Wan X, Liu Q T, Liu J Y, Liu S Y, Liu X F, Zheng L R, Shang J X, Yu R H, Shui J L. Nat. Commun., 2022, 13: 2963.

[59]
Zhang R M, Wang K, Wang P, He Y, Liu Z M. Energies, 2023, 16(9): 3684.

[60]
Yang S, Li X W, Tan T Y, Mao J N, Xu Q, Liu M H, Miao Q Y, Mei B B, Qiao P Z, Gu S Q, Sun F F, Ma J Y, Zeng G F, Jiang Z. Appl. Catal. B Environ., 2022, 307: 121147.

[61]
Zhao X J, Pachfule P, Li S, Langenhahn T, Ye M Y, Tian G Y, Schmidt J, Thomas A. Chem. Mater., 2019, 31(9): 3274.

[62]
Chen H, Li Q H, Yan W S, Gu Z G, Zhang J. Chem. Eng. J., 2020, 401: 126149.

[63]
Zhang S H, Xia W, Yang Q, Valentino Kaneti Y, Xu X T, Alshehri S M, Ahamad T, Hossain M S A, Na J, Tang J, Yamauchi Y. Chem. Eng. J., 2020, 396: 125154.

[64]
Liu M H, Xu Q, Miao Q Y, Yang S, Wu P, Liu G J, He J, Yu C B, Zeng G F. J. Mater. Chem. A, 2022, 10(1): 228.

[65]
Liu M M, Zhu X H, Song Y J, Huang G L, Wei J M, Song X K, Xiao Q, Zhao T, Jiang W, Li X P, Luo W. Adv. Funct. Mater., 2023, 33(11): 2213395.

[66]
Tan L X, Tan B E. Chem. Soc. Rev., 2017, 46(11): 3322.

[67]
Lu Y, Liang J N, Deng S F, He Q M, Deng S Y, Hu Y Z, Wang D L. Nano Energy, 2019, 65: 103993.

[68]
Gu Y L, Son S U, Li T, Tan B E. Adv. Funct. Mater., 2021, 31(12): 2170082.

[69]
Li B Y, Gong R N, Luo Y L, Tan B E. Soft Matter, 2011, 7(22): 10910.

[70]
Xu J J, Shi L, Wang J N, Lu S Y, Wang Y K, Gao G X, Ding S J. Carbon, 2018, 138: 348.

[71]
Zhang J W, Xu D, Wang C C, Guo J N, Yan F. Adv. Mater. Interfaces, 2018, 5(7): 1701641.

[72]
Yang Z F, Han J X, Jiao R, Sun H X, Zhu Z Q, Liang W D, Li A. J. Colloid Interface Sci., 2019, 557: 664.

[73]
Zhu Z Q, Cui J, Cao X Y, Yang L J, Sun H X, Liang W D, Li J Y, Li A. Int. J. Hydrog. Energy, 2022, 47(16): 9504.

[74]
Zhai T L, Xuan C J, Xu J C, Ban L, Cheng Z M, Wang S L, Wang D L, Tan B E, Zhang C. Chem. Asian J., 2018, 13(18): 2671.

[75]
Barman J, Deka N, Maji P K, Dutta G K. ChemElectroChem, 2022, 9(15): e202200677.

[76]
Yang Y, He Z P, Liu Y, Wang S L, Wang H G, Zhu G S. Appl. Surf. Sci., 2021, 565: 150579.

[77]
Pei Y C, Qi Z Y, Li X L, Maligal-Ganesh R V, Goh T W, Xiao C X, Wang T Y, Huang W Y. J. Mater. Chem. A, 2017, 5(13): 6186.

[78]
Zhu Z Q, Cui J, Cao X Y, Yang L J, Sun H X, Liang W D, Li J Y, Li A. Energy Technol., 2022, 10(5): 2101097.

[79]
Chen H W, Liu Y J, Liu B, Yang M, Li H M, Chen H B. Nanoscale, 2022, 14(34): 12431.

[80]
Qian Q P, Hu H Y, Huang S D, Li Y, Lin L L, Duan F, Zhu H, Du M L, Lu S L. Carbon, 2023, 202: 81.

[81]
Kuhn P, Antonietti M, Thomas A. Angew. Chem. Int. Ed., 2008, 47(18): 3450.

[82]
Liu M Y, Guo L P, Jin S B, Tan B E. J. Mater. Chem. A, 2019, 7(10): 5153.

[83]
Kamiya K, Kamai R, Hashimoto K, Nakanishi S. Nat. Commun., 2014, 5: 5040.

[84]
Hao L, Zhang S S, Liu R J, Ning J, Zhang G J, Zhi L J. Adv. Mater., 2015, 27(20): 3190.

[85]
Chai D, Min X T, Harada T, Nakanishi S, Zhang X W. Colloids Surf. A Physicochem. Eng. Aspects, 2021, 628: 127240.

[86]
Cao Y Q, Zhu Y Z, Chen X F, Abraha B S, Peng W C, Li Y, Zhang G L, Zhang F B, Fan X B. Catal. Sci. Technol., 2019, 9(23): 6606.

[87]
Jiao L, Hu Y L, Ju H X, Wang C D, Gao M R, Yang Q, Zhu J F, Yu S H, Jiang H L. J. Mater. Chem. A, 2017, 5(44): 23170.

[88]
Zhang X L, Lin Z Y, Su W, Zhang M T, Wang X J, Li K X. Appl. Surf. Sci., 2022, 592: 153278.

[89]
Zheng Y, Chen S, Song H, Guo H L, Zhang K A I, Zhang C, Liu T X. Nanoscale, 2020, 12(27): 14441.

[90]
Yu W G, Gu S, Fu Y, Xiong S H, Pan C Y, Liu Y N, Yu G P. J. Catal., 2018, 362: 1.

[91]
Jena H S, Krishnaraj C, Parwaiz S, Lecoeuvre F, Schmidt J, Pradhan D, van der Voort P. ACS Appl. Mater. Interfaces, 2020, 12(40): 44689.

[92]
Sönmez T, Belthle K S, Iemhoff A, Uecker J, Artz J, Bisswanger T, Stampfer C, Hamzah H H, Alexandra Nicolae S, Titirici M M, Palkovits R. Catal. Sci. Technol., 2021, 11(18): 6191.

[93]
Tong L, Shao Z G, Qian Y S, Li W M. J. Mater. Chem. A, 2017, 5(8): 3832.

[94]
Zhu Y Z, Chen X F, Liu J, Zhang J F, Xu D Y, Peng W C, Li Y, Zhang G L, Zhang F B, Fan X B. ChemSusChem, 2018, 11(14): 2402.

[95]
Iwase K, Yoshioka T, Nakanishi S, Hashimoto K, Kamiya K. Angew. Chem. Int. Ed., 2015, 54(38): 11068.

[96]
Zhou S K, Xiao Z C, Yang Q, Huang X X, Niu Y, Ma Y J, Zhi L J. Sci. China Mater., 2021, 64(9): 2221.

[97]
Zheng Y, Song H, Chen S, Yu X H, Zhu J X, Xu J S, Zhang K, Zhang C, Liu T X. Small, 2020, 16(47): 2004342.

[98]
Zhu Y P, Li J J, Chen Y B, Zou J, Cheng Q Q, Chen C, Hu W B, Zou L L, Zou Z Q, Yang B, Yang H. ACS Catal., 2021, 11(21): 13020.

[99]
Dun R M, He X, Huang J, Wang W liu Y W, Li L H, Lu B W, Hua Z L, Shi J L. J. Mater. Chem. A, 2023, 11(11): 5902.

[100]
Quernheim M, Liang H W, Su Q, Baumgarten M, Koshino N, Higashimura H, Müllen K. Chem. A Eur. J., 2014, 20(44): 14178.

[101]
Patil B, Satilmis B, Uyar T. J. Power Sources, 2020, 451: 227799.

[102]
He D P, Rong Y Y, Kou Z K, Mu S C, Peng T, Malpass-Evans R, Carta M, McKeown N B, Marken F. Electrochem. Commun., 2015, 59: 72.

[103]
Tian Y Y, Zhu G S. Chem. Rev., 2020, 120(16): 8934.

[104]
Wu X, Shaibani M, Smith S J D, Konstas K, Hill M R, Wang H T, Zhang K S, Xie Z L. J. Mater. Chem. A, 2018, 6(24): 11327.

[105]
Chen J H, Jiang L C, Wang W T, Wang P Y, Li X, Ren H, Wang Y G. Micropor. Mesopor. Mater., 2021, 326: 111385.

[106]
Zhao S, Bian Z, Liu Z L, Wang Y H, Cui F C, Wang H G, Zhu G S. Adv. Funct. Mater., 2022, 32 (44):2204539.

[107]
Xiang Z H, Wang D, Xue Y H, Dai L M, Chen J F, Cao D P. Sci. Rep., 2015, 5: 8307.

[108]
Wu K Q, Chen R, Zhou Z X, Chen X H, Lv Y Q, Ma J, Shen Y F, Liu S Q, Zhang Y J. Angew. Chem. Int. Ed., 2023, 62(12): e202217078.

Outlines

/