According to the classification proposed by Vassilev et al., biomass is a non-fossil, complex and biogenic heterogeneous mixture of organic and inorganic matter that is primarily generated by photosynthesis, such as natural constituents from lands and water-based vegetation; animal or human food digestion; anthropogenic processing of wood, plants and other organic matter, such as manure or household waste (Vassilev et al., 2010). Biomass is typically utilized for solid, liquid and gaseous products from non-edible residues formed by floras and faunas, such as corn cobs, wheat stalk, cornstalk, and manure (Supplementary Table S1). Biomass is carbon-rich, sustainable, and renewable. This has made it a sought-after precursor for generating green CNs (solid component) as well as being used for the production of value-added chemicals and biofuels (predominantly liquid and gaseous component) (Bhaskar et al., 2011; Titirici et al., 2015; Anca-Couce, 2016; Kumar et al., 2016).

Biomass can be divided into two major groups with the same characteristics of origin and chemical structure, i.e., lignocellulosic and non-lignocellulosic biomass (Martin, 2010; Kosinkova et al., 2015).

Lignocellulosic biomass is considered one of the most promising renewable sources for producing carbon materials and fuels while remaining environmentally friendly (Cai et al., 2015; Kabir and Hameed, 2017; Li et al., 2017). Lignocellulosic biomass is mainly a waste of wooden and crop industries; primarily rich in hemicellulose (15%–40%), cellulose (25–50 wt%), and lignin (10%–40%) (Chen, 2014; Kumar et al., 2016; Maliutina et al., 2017; Li Y. et al., 2020).

Cellulose is a linear polysaccharide that unsoluble in water and consists of D-glucose monomers bonded together via β-(1–4) glycosidic bonds (Abraham et al., 2011). The inter and intramolecular hydrogen bonds that give cellulose its crystalline form are formed between the three hydroxyl groups on each glucose monomer (Harmsen et al., 2010). Cellulose microfibrils are formed due to the coalescence of the polymer chains to form fibers, which in turn gives cellulose its crystalline structure (Shafizadeh, 1982). Hemicellulose comprises of various monomers such as glucose, mannose, arabinose, and xylose, which vary in quantity according to the type of biomass (Li, 2014). Hemicellulose take place in association with cellulose in the cell wall, but unlike cellulose, hemicelluloses are soluble in dilute alkali. Hence, xylan, which is the most abundant monomeric hemicellulose unit up to about 10% and 30% of the dry weight of the species (Yaman, 2004). Unlike cellulose, hemicellulose does not exhibit a crystalline structure due to the high number of branched chains and acetyl groups linked to a polymer chain. Lignin is an aromatic polymer in nature that consists of p-coumaryl, coniferyl, and sinapyl alcohol monomers and hydroxyl, carbonyl, and methoxy functionalities to form a firm composition (Chen, 2014). During pyrolysis or carbonization processes, hemicellulose decomposition occurs first at 200°C–260°C. Cellulose fragments and lignin structure break up at 240°C–350°C (Wang et al., 2015) and 280°C–500°C (Ramanayaka et al., 2020), respectively. The ratio of every single compound of hemicellulose, cellulose, or lignin is not regular and constant, which is related to the species, age, soil, place of origin (Liu et al., 2019).

Non-lignocellulosic biomass is predominantly comprised of carbohydrates, lipids, proteins and inorganic minerals (Li and Jiang, 2017). Apart from algal biomass, non-lignocellulosic type includes animal hair, feathers, bones, fish and seafood waste, sewage sludge, and manure (Li and Jiang, 2017). Sewage sludge encompasses different constituents like organics, inorganics, and microbes (Yu et al., 2021). Algae is considered a significant component of the Earth’s biomass, whose growth rate and CO₂ fixation effectiveness are higher than any earthly plant (Zhao et al., 2015). Algae consist mainly of lipids, proteins, ash, and carbohydrates. The main elements in algae are C and O, which account for 25%–55% and 15%–50%, while H and N are accounts for 5%–10% of algae, respectively (Li et al., 2017).

The conversion of biomass to advanced nanostructures is influenced by source, physical and chemical composition, that is highly variable in terms of origin and growth factors (e.g., environment) which can either enrich or deplete different compounds and elements (Vassilev et al., 2015). On average, the elemental composition of biomass consists of 99.9% of C, Ca, Cl, H, K, Mg, N, Na, O, P, and S, with ≤0.1% of trace elements (Vassilev et al., 2013). In summary, biomass has proven to be the an ideal source for CNs, given a series of advantages, including its relatively low cost and worldwide abundance; the absence of competition with the food industry; the natural viability of N, S, P, and O, allowing for doping and formation of chemical functionalities, without the need of extra chemicals.

Biochar is a carbon-rich material that is produced by heating organic matter, such as wood chips, agricultural waste, or manure, in an oxygen-depleted atmosphere (Li Z. et al., 2019). The chemical nature of biochar is complex and depends on the type of biomass used, the thermal treatment conditions, and the post-treatment processes. Biochar is predominantly composed of carbon, with content ranging from 50% to 95% depending on the production process (Li Y. et al., 2020). The carbon in biochar is present in various forms, including elemental carbon, amorphous carbon, and graphite-like structures. These different forms of carbon are distributed throughout the biochar, creating a porous structure that provides high surface area and high adsorption capacity (Omoriyekomwan et al., 2021a). Apart from C, biochar also contains other elements such as H, O, N, and S, which are present in smaller quantities, and that can be included in C-based structures as dopants (Maliutina et al., 2018). The elemental composition of biochar significantly depends on the type of biomass used and the production conditions (Maliutina et al., 2017).

Biochar also contains a range of functional groups, such as carboxylic acids, phenols, ketones, and aldehydes. These functional groups are in large extent responsible for biochar chemical reactivity. The possibility of rationally designing these types of moieties on biochar surfaces presents a powerful strategy for tuning their potential catalytic properties, as well as influencing the adsorption of ions and molecules in the environment, by adjusting the overall acidity and basicity of the biochar surfaces. The surface area and porosity of biochar are further crucial features influencing chemical and catalytic properties. The porous structure of biochar is ordinarily emerging from the thermal preparation process, when the volatile constituents in the biomass are released, leaving behind a charred residue (Sch et al., 2011). Once again, the possibility of rationally designing high surface area and porosity, would provide a large number of active sites for adsorption and catalysis. Besides its potential applications as a catalyst or catalyst support, biochar is being investigated and employed in advanced sorbents, fertilizers, etc. (Huang et al., 2016; Uzun et al., 2016).

The chemical nature of biochar can also be modified through post-treatment processes, such as acid washing or chemical activation. These processes can alter the surface chemistry and functional groups of biochar, resulting in enhanced adsorption capacity and improved performance in environmental applications. Besides more or less complex C-based structures, thermal treatments of biomass for biochar production, often leaves K–Ca–Mg–Na carbonates, Ca–K–Mn silicates, K–Na–Ca chlorides, Ca–Al–Mn oxides, sulphates, and phosphates (Vassilev et al., 2014; Omoriyekomwan et al., 2021a; Vassilev et al., 2012), which content in the final biochar product can be adjusted via post-treatment processes.

The phenomena of mass transfer on the surface of biochar during heterogeneous catalysis can be divided into the following stages (Lee et al., 2017): a) diffusion of the reagent from bulk gas or liquid film to the external surface of the catalyst; b) diffusion of the reagent to the inner surface of the catalyst through the pores of the catalyst; c) adsorption of the reagent on the surface of the catalyst; d) reactions occurring on the catalytic active centres on the catalyst surface; e) desorption of product from the catalyst surface; f) diffusion of product to the internal surface of the catalyst through catalyst pores; g) diffusion of product from the external catalyst surface to bulk gas or liquid film (Lin and Huber, 2009). Thus, catalytic activity is highly contingent on accessibility to catalytic active sites dispersed throughout internal pores. Although biochar is a porous material, the morphology and porosity of biochar without activation demonstrates poor catalytic properties. Various pretreatments can be applied to modify biochar morphology and porosity, as highlighted in Section 2.3. Surface area, pore size, pore volume, pore type [i.e., micropores, mesopores and macropores (Bikbulatova et al., 2017)] are vital properties, that impact overall catalytic ability.

Despite discussed benefits, the effective conversion of biomass into high-performing nanomaterials on a large scale for electrocatalytic purposes remains challenging (Alston and Arnold, 2011; Sharma et al., 2011). Amongst the major technical barriers presenting significant challenges in terms of materials comparison or standardisation of production processes, we can mention the high water soluble fraction, the (already discussed) huge variability of organic compositions of different biomasses, and to some extent, the large variability of inorganic components too (Alston and Arnold, 2011; Sharma et al., 2011).

Chemical pretreatment of biomass is often regarded as a strategy to impart a particular porosity, microstructure and/or morphology to the final CNs. Omoriyekomwan et al. (Omoriyekomwan et al., 2017) compared microwave-assisted and conventional (fixed-bed) of raw palm kernel shell (PKS) pyrolysis approaches. Hollow carbon nanofibers (HCNFs) were successfully produced on the surface of bio-chars during microwave pyrolysis of PKS at 500 °C and 600°C. The observed HCNFs onto the PKS biochar was attributed to the microwave heating condition only and not detected under conventional heating. Authors explained this phenomenon by the difference in heat transfer mechanism in microwave and conventional heating. The yield of observed hollow carbon nanofibers (HCNFs) increased with temperature and reached 9.88 wt% at 600°C (Omoriyekomwan et al., 2017). A self-extrusion mechanism of volatiles from inside of the biomass particle through pores on biomass surface and resolidification of volatiles on particle surface bio-component was attributed for CNTs formation under microwave irradiation at 600°C (Figure 2). Additionally, the presence of Fe, K, and Ca in raw PKS biomass and in the HCNFs structure was considered to have an important role during their formation and growth (Omoriyekomwan et al., 2017).

In another study of Omoriyekomwan et al. (Omoriyekomwan et al., 2019), microwave pyrolysis for 1 h at 600°C was carried out on the raw PKS (Figures 2B, C) and isolated cellulose without addition of external catalyst (Figures 2D, E). Two different pretreatment methods, i.e., alkali-acid and formic acid/acetic acid were systematically compared. The higher yield of formed CNTs was correlated with the two-step alkali-acid (NaOH for lignin isolation/H₂SO₄ for cellulose isolation) route of the raw PKS biomass (Omoriyekomwan et al., 2019). The chemical alkali-acid pretreatment displayed effective synthesis of CNTs that greatly enhanced their length (600–1800 nm in this study vs 200–1,600 nm of commercial CNTs) (Figure 2E) (Omoriyekomwan et al., 2019). The observed Raman I_D/I_G ratio for CNTs synthesised in this study was lower (0.86) compared to that of commercial CNTs (0.96) was attributed to lower amount of defects and higher order of the nanostructure, resulting in higher quality (Omoriyekomwan et al., 2019).

Recently, Maliutina et al. (Maliutina et al., 2021b) reported two-step chemical pretreatment and pyrolysis methods that were employed to pomelo peel biomass to obtain mesoporous carbon nanosheets (Maliutina et al., 2021b). The formation of carbon nanosheets with a thickness of about 107 nm (Figure 3A) was obtained from irregularly shaped pomelo peel as a precursor using a chemical pretreatment process, involving alkaline etching with KOH and acid hydrolysis with H₂SO₄ (Figure 3B). The final cell-PP-NS-700-Ta-900 nanosheet sample also exhibited a moderate graphitization degree (ID₁/IG ratio of 1.11) and high specific surface area (SSA) of 336.82 m² g^-1. This chemical pretreatment method was regarded as a promising route for fast and tunable conversion of the complex pristine biomass microstructure into a nano framework with desirable properties.

Cyclic oxidation pretreatment of biomass for CNs synthesis has been reported in other works (Kang et al., 2005; Goodell et al., 2008). The cyclic oxidation process allows for the conversion of biomass into tubular carbon nanostructure via a repetitive (i.e., cyclic) oxidation process. Goodell et al. (Goodell et al., 2008) used six different feedstock, including wood fibers, avicel cellulose, bamboo, alpha-cellulose, organosolv lignin, and wood fibers. All samples were preheated before cyclic oxidation at 240°C and 400°C, respectively. They reported hard tubular CNTs formation in cellulose samples, organosolv lignin, and filter paper. It was concluded that the original biomass structure played an essential role in the formation of CNs via the cyclic oxidation method.

The Authors proposed a nanoscale channel formed at low-temperature carbonization treatment that acted as the templates for the CNTs’ growth (Goodell et al., 2008). The formation of nano-channels at low-temperature carbonization promoted cellulose ablation, which contributed to CNTs synthesis. Nevertheless, when pure biomass components (such as lignin and cellulose) were subjected to cyclic oxidation, tubular CNTs were not formed.

Mechanical activation treatment was proposed by Chen et al. (Chen and Chadderton, 2004) when they reported the synthesis of aligned CNTs from iron phthalocyanine pyrolysis after ball milling treatment. Milling time was observed to be the most important controlling parameter in this process (Chen and Yu, 2005; Onishchenko et al., 2012; Onishchenko et al., 2013a). Onishchenko et al. (Onishchenko et al., 2013b) reported the formation of CNTs from sphagnum moss and agricultural wastes via a mechanical activation process. The CNTs content after mechanical activation was 11.52 wt%, 24.45 wt%, 34.06 wt%, and 42.75 wt% for mechanical pretreatment duration of 7, 10, 16, and 27 h, respectively (Onishchenko et al., 2013b). It was observed that after 10 h of mechanical treatment, most of the materials were enriched with CNTs. In other study of Reva et al. (Reva et al., 2016a), abundant CNTs were presented in all bulk material derived from brown sphagnum moss biomass after 36 h treatment. Electrostatic interaction of the CNs was reported to be responsible for the felt-like aggregates forming with a size of 20–100 μm, consisting of CNTs and amorphous carbon (Reva et al., 2016a; Reva et al., 2016b). It is worth noting that there was no catalyst involved in synthesizing CNs from biomass using mechanical treatment, while amorphous carbons acting as the precursor of CNs before mechanical activation were readily obtained from biomass. Therefore, mechanical pretreatment is a favorable technique for the synthesis of CNs as a result of the low cost of technology and feedstock.

H₃PO₄, and ZnCl₂ are also used alongside KOH as pretreatment agents prior to pyrolysis (Chaparro-Garnica et al., 2021). This method has showed to enhance porosity and ensurehigh specific surface area (SSA) of the final CNs (Chen et al., 2016). Li et al. pretreated willow catkin in KOH aqueous solution based on a mass ratio of 1:1 and obtained an interconnected porous carbon nanosheet after pyrolysis (Li et al., 2016). Meanwhile, large SSA of 2,385 m²g^-1, a large pore volume, and easily accessible open surfaces of graphene nanosheets were achieved from spruce bark after HTC at 180°C and KOH pretreatment (Figures 4A–H) (Sun et al., 2018). Chemical pretreatments of biomass were also proven to be beneficial in the case of biochar synthesis from graphitisation processes (Sun et al., 2013; Mahmoudian et al., 2016; Su et al., 2017). Interesting observations were also reported in the case of chemical pretreatment prior pyrolytic biochar synthesis via microwave heating rather than conventional heating. Zhang et al. (Zhang et al., 2021) synthesized biochar-CNTs composite from char from pine nutshell char via microwave heating (Figures 5A–D).

The Authors impregnated the source biomass with Ni catalyst and applied microwave heating in an inert environment of CH₄/N₂ gas (60/40 vol%) for 20 min reaching 600°C. Enhanced CNTs growth was reported and ascribed to the conversion of Ni nanoparticles into highly-crystallized NiO nanoparticles (Zhang et al., 2021). Supplementary Table S2 summarises biomass-based CNs obtained from different pretreatment and fabrication methods that were examined by this review.

It can be seen in Supplementary Table S2, that the preferable temperature range to produce CNTs enriched biochar is 700°C–800°C, while production of carbon nanosheets correlates with the range of 700°C–900°C. It is worth to note, that the microwave pyrolysis method assists to produce CNTs at the lower temperature range of 600°C–700°C. It can be explained by the difference in heat transfer mechanism in microwave and conventional heating. Unlike conventional heating, during microwave irradiation the particle temperature is higher than the surrounding area due to the selective heating under microwave. During microwave pyrolysis, the released volatiles are extruded through pores on biomass particle surface. Upon release, the high molecular weight tar precursors solidify due to lower temperatures at particle surface compared to particle core, initiating CNTs/CNFs formation and growth via CVD mechanism. Graphene nanosheets can be obtained at temperatures above 850°C during the graphitisation process, but requires a catalyst, such as Fe or Zn based.

As an alternative to precious metal-based catalysts, various CNs, such as CNTs/CNFs (Wu Z.-Y. et al., 2018), and graphene-based nanosheets (Gong et al., 2019; Tang and Qiao, 2019) have been reported as ORR/OER electrocatalysts. CNs are generally functionalized via doping, allowing to generate functionalised electrocatalysts with a superior electrocatalytic activity (Zhang et al., 2016b). Dopants alter the amount of crystal and surface defects acting as potential catalytic active sites and modify the electronic structure of the CNs (Zheng et al., 2014; Wu et al., 2017), with a direct influence on the thermodynamics and charge carrier distribution in the carbon-based electrocatalysts (Sharifi et al., 2012; Lu et al., 2017; Mamtani et al., 2018; Jiang et al., 2019b). Typical types of doping include mono/dual/multiple heteroatoms doping; defective/vacancy doping; and charge‐transfer doping.

Doping of graphene frameworks with non-metal elements such as N or O has been reported as a convenient strategy to tune the electrochemical as well as optical properties of graphene nanostructures (Inagaki et al., 2018). Zhang L. et al. (Zhang and Xia, 2011) investigated the role of nitrogen heteroatom towards ORR based on density functional theory (DFT) calculations applied to a N-doped nanographene model (C₄₅NH₂₀) in acidic medium. Nitrogen doping introduces an unpaired electron (the latter causing localized distribution of the molecule orbitals) with the overall result of enhancing the chemical reactivity of the graphene itself. Their simulations demonstrated that ORR on N-doped graphene is a direct four electron pathway, and consistent with experimental observations. Interestingly, the Authors identified the active catalytic sites on single nitrogen doped graphene, having either high positive spin density or high positive atomic charge density. N doping introduces asymmetry spin density and atomic charge density, allowing for N-graphene to exhibit high electroncatalytic activities for the ORR (Zhang and Xia, 2011). The electrocatalytic performance of the heteroatom-doped CNs partially depends on the electronic structure of the heteroatoms in comparison to the positively charged C atoms. In this case, the selectivity of a specific reduced product strongly depends on the affinity of active sites towards the corresponding intermediate motifs. This is quite different from the conventional pristine CNs.

Experimentally, different N functionalities in the carbonaceous skeleton can be easily probed and interrogated via X-Ray photoelectron spectroscopy (XPS) given the distinguishable N 1s binding energy of pyridinic-N (ca. 398 eV), pyrrolic-N (399–400 eV), graphitic-N or quaternary-N (ca. 401 eV), and pyridinic N oxide (402–406 eV) (Maliutina et al., 2018; Jiang et al., 2019a; Maliutina et al., 2021b). Pyridinic-N exists on the edge of the graphene layer and bonds to two carbon atoms, containing a lone electron pair, which can boost the electron-donor capability of the carbon material, resulting in the improvement of the adsorption capacity of O₂ and onset potential for ORR (Lai et al., 2012). Whilst incorporated into the graphene layer and bonded to surrounding three carbon atoms, graphitic-N donates one extra electron to the carbonaceous skeleton that weaken the O-O bonds and decrease the energy barrier for the first electron transfer (rate-determining step). Overall, graphitic-N functionalities facilitate the limiting current density and electron transfer for ORR. Wang and co-workers (Wang D.-W. et al., 2012) postulated that pyridinic-N and pyrrolic-N improve the storage of hydrogen atoms and accelerate protonation. At the same time, the oxidized-N functionalities exhibit a high affinity to electrons, unlike less positively charged pyridinic-N. The edge plane graphitic-N (valley) acts the same as pyridinic-N centers via a four-electron pathway, while graphitic-N (center) acts via a two-electron pathway, as elucidated by Sharifi and co-authors (Sharifi et al., 2012). Xing and co-authors postulated that the pyridinic-N form and the neighboring carbon atom play a crucial role in ORR activity (Xing et al., 2014). The ORR and OER activity was observed to increase with an increase in pyridinic-N site density. Mamtani and co-workers postulated the correlation between higher content of the pyridinic-N active site and electrocatalytic activity towards both ORR and OER (Mamtani et al., 2018). Notwithstanding, the research community still debates what form of nitrogen, pyridinic-N, graphitic-N, or their particular ratio make a more significant contribution to electrochemical performance based on both computational and empiric features (Matter et al., 2006; Guo et al., 2016; Singh et al., 2019).

Incorporation of the elements from the second row of the Periodic Table with a larger radius than carbon leads to significant distortion of planar structures creating defects that act as active centers towards ORR/OER as well as HER. S-doping can effectively alter spin/charge density distribution and distorts the carbon lattice by creating large electroactive sites. Sulfur doping was observed to alter spin density distribution via formation of several different sulfur-containing functionalities, including thiophene (aromatic sulfur), thiol, thioesters, sulfoxide, sulfone, and sulfonic acid (Zhao et al., 2012).

XPS showed that a series of four major peaks (S 2p peaks appear as doublets of 2p3/2 and S 2p1/2 due to spin-orbit splitting) can be expected in the C lattice; i.e., at binding energies of 163.80 eV and 165.20 eV corresponding to the S 2p3/2 and S 2p1/2 peaks of the C–S–C bonds, and at binding energies of 167.7 eV and 169.3 eV corresponding to S 2p3/2 and S 2p1/2 indicative of C–SOx–C species, where x is 2, 3, and 4, respectively. (Wang Z. et al., 2014; Zhang J. et al., 2020; Maliutina et al., 2021b). Furthermore, the doping of carbon framework with phosphorous introduces reduced unstable P–O groups as phosphate, phosphine oxide, phosphonic acid, and substituted C-P-C phosphine, thereby improving the electrochemical properties. The XPS P 2p spectra of carbonaceous metal-free materials exhibit a predominant P–C characteristic peak and P–O bond at 131.0 ±0.7 eV and 133.7 ± 0.5 eV (Huang et al., 2018). In particular, it has been reported that the co-doping of CNs with two [such as N/S (Wang X. et al., 2014; Ito et al., 2015), N/P (Borghei et al., 2017; Cheng et al., 2021), or N/F (Akula et al., 2021)] or even multi-doping [such as N, S, P and F (Liu et al., 2016; Huang et al., 2018; Li Y. et al., 2019)] with several heteroatoms is an effective way to boost their electrochemical activity synergistically. Recently, Li et al. reported no synergetic effect between N and S dual-doping towards ORR, while the sequence of doped heteroatoms played a critical role (Li J.-C. et al., 2019). Due to this phenomenon, doping of S followed by doping of N promotes the formation of pyridinic-N functionalities. This is attributed as the most active functional group in N, S dual-doped electrocatalyst towards the ORR, while on the counter doping of N followed by doping of S reduces the content of pyridinic-N species, which leads to a nadir of ORR performance (Li J.-C. et al., 2019).

For oxygen adsorption, only the sample with pyridinic-N functionalities possesses binding energy of −1.4 eV, which is superior to that of pyridinic-N-S (−1.0 eV), graphitic N (−1.1 eV), and graphitic-N-S (−1.0 eV) (Li J.-C. et al., 2019). Additionally, the Gibbs free energy diagrams for different catalyst models in an alkaline environment were presented by authors (Li J.-C. et al., 2019) observed all the reaction intermediates (O₂*, OOH*, O* and OH*) formed during ORR. Pyridinic N (green line) plentifully contributes the ORR via the OOH* formation with a small energy barrier (0.3 eV), as well as the fact that only the pyridinic N can chemically adsorb O₂ that overall contribute to the ORR. Pyridinic-N functionality was also mentioned as the major responsible active site in other valuable studies (Rao et al., 2010; Zhang and Xia, 2011; Vikkisk et al., 2014; Xing et al., 2014). Notwithstanding, it is still difficult to distinguish the role of secondary or especially tertiary doped heteroatoms not only by computational DTF simulations but especially for such a complex biomass-derived carbonaceous matrix both experientially and theoretically.

The synergy between various techniques can avoid, to a great extent, the limitations of using one sole technique. For example, the advances in using a combination of operando XRD and Raman spectroscopy are desired, which enables the complementary characterisation of nanocatalysts in both surface and bulk levels (Zhang H. et al., 2023). Under operando conditions, the adsorbate and intermediate species, as well as the structural parameters of the catalyst can be simultaneously traced for an extensive understanding of the reaction pathways and elucidate the role of active centres. Structural defects, such as vacancies, substitutional or interstitial impurities, are regions of absorptive binding for gases and ions are successfully examined by scanning tunnelling microscopy (STM) (Feng et al., 2018; Xiao et al., 2021). X-ray absorption spectroscopy (XAS) has become a universal and in-depth characterisation technique that has been widely used in many fields due to the rapid development of synchrotron radiation sources. (Gao et al., 2017). Because of the difference in the relative absorption threshold of energy, the XAS can be divided into two parts: XANES (X-ray absorption near-edge structure, approximately 40 eV and below) and EXAFS (extended X-ray absorption fine structure, beyond the XANES region) (Xiao et al., 2017). The XAS analysis provides a myriad of information about the structure CNs and their dopants, such as the structure symmetries, element valence, bond distance and adjacent coordination environment of the atom (Liu et al., 2017). Although tremendous amount of works has described in-situ electrocatalytic characterisation, in-situ and operando characterization techniques and devices with a high resolution are necessary to elucidate the mechanism of defects due to the complexity of electrochemical reactions and the variety of imperfections. Consequently, the synergetic studies based on models of biomass compounds and comprehensive studies on in situ tunable engineering and activity of heteroatom-based active sites are highly desired for carbon-based electrocatalysts and a sustainable future.

Up to date, there are tons of works that report N-doped electrocatalysts towards ORR and a major part is belongs to alkaline electrolyte catalysis. Predominantly, their catalytic performance is still far from that of the Pt/C benchmark, although the characterization of nanomaterials very often describes similar nitrogenized functionalities and analogous distributions (Supplementary Table S3). Rationally incorporating one specific type of heteroatom for designing catalytic active sites remains one of the major challenges when thermochemically valorising biomass for production of efficient electrocatalysts for water splitting.

Zheng and co-authors derived a nitrogen self-doped carbon hollow cubes (NCHCs) was derived from L-lysine precursor at 1,000°C (Zheng et al., 2017). When used as electrocatalysts for the ORR, this composite exhibited an onset potential of 0.92 V vs RHE. This corresponds to a cathodic shift of only 61 mV of the half-wave potential of NCHCs when compared to that of commercial Pt/C (20 wt%), i.e., a common and widely used benchmark for ORR. Interestingly, the NCHCs composite also showed activity for the OER statistically comparable to another widely used commercially available OER standard, i.e., RuO₂/C (20 wt%). The E_onset potential close of that of Pt/C was attributed to the role of pyridinic-N and graphitic-N active structures. Wang and others used chitin biomass to fabricate N-doped 3D graphene-like nanosheets (N-DC/G) by doping with pretreating biomass with urea and proceeding with carbonization at 800°C (Wang et al., 2016).

The HRTEM image of the as-prepared N-DC/G sample displayed thickness of the layer less than 50 nm coupled with uniformly distributed porous structure. The corresponding K-L plots showed first-order kinetics towards ORR at different potentials and electron transfer number (n) of about 4, suggesting a direct four-electron pathway attributed to the presence of pyridinic-N and graphitic-N active sites. In another study, He and co-authors (He et al., 2019) used Taro stem plant biomass, impregnation with melamine, followed by KOH activation and carbonization at 800°C. The obtained biochar (3DNPC-800) displayed meso-/microporous 3D architecture with large SSA of 1,012 m² g^-1 and enriched nitrogen content, measured equal to 4.8 at%. The as-prepared biomass-derived 3DNPC-800 electrocatalyst exhibited ORR long-term stability (96.5% retention of the current density after 5.5 h) and resistance to methanol poisoning, which was attributed to the promoted formation of graphitic-N functionalities and developed 3D porous structure. Zhang and co-workers (Zhang et al., 2019) synthesised N-doped CNTs from paper mill sludge (PMS) by treatment with melamine (PMS/CNTs-b) and carbonization at 900°C. Results are illustrated in Figures 6A–C.

The as-fabricated biochar displayed enriched formation of CNTs with predominant graphitic-N functionalities. The exceptional ORR activity of PMS/CNTs-b catalyst in contrast to that of Pt/C benchmark results in the most positive E_onset potential of 0.99 V vs RHE and current density of −6.96 mA cm² shown in Figure 6D. The calculated electron transfer number of the PMS/CNTs-b catalyst is a direct four-electron pathway, as can be seen in Figure 6E. The formation of CNTs was attributed to the addition of melamine, which was at first just considered as a nitrogen source to develop N-doped PMS-based catalysis in this work, during the pyrolysis of PMS. The remarkable ORR performance of PMS-CNT/C-b was correlated with the formed nanotube/nanoporous structure and the synergistic effect of abundant N groups, iron nitrides and thiophene-S. It can be seen in Figures 6F,G that the PMS/CNTs-b catalyst was employed in microbial fuel cells and possessed a power density of 675.72 W/m² that is improved when compared to Pt/C cathode (580.46 W/m²).

A collection of biomass-based nanostructures with single heteroatom doping are reported in Supplementary Table S3 alongside physical chemical properties, and claimed active sites. From a comparison of the data reported in Supplementary Table S3 (as well as for most of the considered bifunctional or trifunctional electrocatalysts), graphitic-N was mentioned as the predominant active site.

Doping biomass derived carbonaceous nanostructured with Earth abundant, or not rare, metals can exploit the electrocatalytic properties of transition metals at relatively low costs, compositional flexibility, and provide opportunities for large scale production and deployment eliminating major barriers associated with the economy, origin, abundancy, production and supply chain typical of rare and precious metals (Han et al., 2019). Rationally designing earth-abundant hybrid electrocatalysts for ORR/HER/OER approaches as rare and precious metal substitutes is of extraordinarily importance in the quest to reach net zero carbon emissions. In order to enhance the electrocatalytic activity of carbonaceous catalysts and further extend their feasibility for electrochemical energy generation and storage devices, introducing certain TMs (Fe, Mn, Co, Ni, Cu, Mo, Ti, Ta, etc.) and their derivatives in form of oxides/suboxide, sulfides, selenides, nitrides/oxynitrides, carbonitrides is a sought after strategy. Previous studies indicate that adding iron (Wang K. et al., 2020) or cobalt (Zhang et al., 2018) salts into a porous carbon catalyst can significantly improve the ORR catalytic performance of the resulting samples. Furthermore, nickel-based (Ding et al., 2019) carbon materials have been extensively considered as a highly efficient HER/OER electrocatalyst. In synthesis of bifunctional/trifunctional electrocatalysts, doping of metal or metal oxides onto a biomass-derived CNs may be an alternative manner. This section represents recent progress towards the next-generation of multifunctional hybrid metal-loaded green catalysts, as summarized in Table 2 and Supplementary Table S5. It can be seen in Table 2 and Supplementary Table S5, that the most of widespread metal-based interractions are presented by Fe-N_x (or Fe-N-C) and Co-N_x NPs (or Co-N-C) heterojunctions, i.e., a central TM cation coordinated to pyridinic nitrogen on edges of the graphitic surface. Despite that, pyridinic-N can be found as the most common active sive for ORR/OER of biomass-derived carbonaceous nanosupport.

Recently, Zhang et al. (2018) employed chitosan biomass for in situ synthesis of Co,N-codoped CNTs, amongst other N and Co doped nanosheets. It can be seen that the synthesized CNT electrocatalysts possess tubular morphology with uniform distribution of N and Co atoms. A sample of Co(16%) and N codoped CNT (Co16%-NCNT-T800) synthesized at 800°C has been found to display the highest ORR current density (6.92 mA cm^-2) among other carbon based catalysts and Pt (40 wt%)/C (5.06 mA cm^-2) and a half-wave potential closest to Pt (40 wt%)/C. In line with recent research showing that M–N–C active sites formed via coordination of the metal with pyridinic nitrogen plays a crucial role in enhancing ORR activities, Zhang et al. (2018) showed that key for high ORR activity is an effective formation of the Co–N–C active sites (via pyridinic nitrogen). The Authors also found that the Co16%-NCNT-T800 sample exhibited the highest specific capacitance of all electrocatalysts. The higher the specific capacitance (the latter resulting directly from electrochemical specific surface area), the higher the amount of active sites for current density enhancement. Furthemore, the as-prepared Co16%-NCNT-T800 electrocatalyst showed exceptional resistance to the methanol poisoning with no obvious drop in current (in contrast to that of Pt/C benchmark) and long-time stability.

In a recent study, of Lv and others (Lv et al., 2019) fabricated a hierarchical porous Co,N-codoped bio-carbon bifunctional electrocatalyst (CoTBrPP@bio-C) for an all-pH ORR and HER by using a simple one-pot co-pyrolysis over a commercially available mushroom template with adsorbed Br-substituted porphyrinato cobalt (CoTBrPP). The as-prepared nanosheets displayed a porous stratified volcanic rock-like structure with Co-N_x functionalities. XPS was executed to determine the impact of Br on the formation and distribution of nitrogenized functionalities. Autors claimed that the addition of Br can promote the formation of Co-N_x and graphitic-N species, attributed to active centers for ORR and HER. Further, ORR and HER electrocatalytic activity were evaluated in 0.1 M KOH and 1 M KOH environments, respectively. The obtained biomass-derived CoTBrPP@bio-C hybrid electrocatalyst possessed outstanding activity towards ORR (E_onset=0.93 V vs. RHE and E_1/2=0.85 V vs. RHE) as well as HER activity. The excellent Zn-air battery performance of the CoTBrPP@bio-C hybrid with the open-circuit voltage remained at 1.51 V, while the discharging polarization and the corresponding power density of 100 mW cm^-2 at 155 mA cm^-2, were superior to that of Pt/C (78 mW cm^-2 at 119 mA cm^-2).