Principal strategies for designing graphdiyne-based catalyst toward green hydrogen production from water electrolysis

Water electrolysis has attracted significant attention for large-scale production of green hydrogen as next-generation clean fuels. Recently, the development of graphdiyne (GDY), a new member of carbon allotropes, has been promisingly offering novel alternatives for acquisition of inexpensive and efficient catalysts in the water electrolyzer. The unique atomic arrangement in GDY architecture leads to coexistence of sp– and sp2–C, correspondingly brings numerous intriguing features such as heterogeneous electron distribution, wide tailorable natural bandgap, rapid electron/mass transport and rich chemical bonds. These unique intrinsic natures of GDY provide brilliant inspirations for scientists to design new-concept electrocatalyst toward cathodic hydrogen evolution reaction, anodic oxygen evolution reaction and the overall water-splitting. Based on the immense progress, in this short perspective, current principal design strategies of GDY-based catalysts are systematically summarized, including interface engineering, individual atom fixation, induced constrained growth and bottom-up fabrication. With abundant implementation examples for achieving highly efficient water electrolysis, in particular we focus on clarifying the decisive role of GDY on these design strategies with comprehensive theoretical and experimental evidences. The future direction in developing GDY-based electrocatalysts in hydrogen energy field is also depicted with the urgent anticipation of deeper understanding of structure-performance relationship and catalytic mechanism, especially those in real industry water electrolyzers.


Introduction
With the increasing crisis of climate challenges, carbon neutralization is becoming a heated topic of the current era. As a renewable resource with high energy density, high calorific value, easy accessibility and zero carbon emission, hydrogen has been widely considered to be one of the most ideal alternative energy forms [1,2]. While electrochemical water splitting using renewable electricity is believed as the most promising technical route for the large-scale production of green hydrogen in the coming decades. Currently, precious metals (Pt, Pd, Ru, Ir, etc) dominate the market as commercial catalyst for water electrolysis equipment due to their high intrinsic activity. While the natural scarcity, extortionate cost and poor durability bring an insurmountable obstacle for their universal application. To accelerate the formation of economic advantages for green hydrogen production, highly-efficient, long-lasting, cost-effective electrocatalyst is still in increasing urgent demand.
Along with numerous recent efforts in precious-metal-free alternatives as catalysts for electrochemical hydrogen production, carbon materials have exhibited great potential in designing state-of-art catalyst with comprehensive advantages in efficiency, stability and cost [3,4]. Depending on the sp2 hybridized carbon atom and their arrangement, various low-dimensional carbon allotropies have been precisely identified, including fullerenes [5], carbon nanotubes [6] and graphene [7]. Those sp2-hybrized carbon materials feature π-conjugated chemical architecture, large surface area and strong conductivity, which endow them remarkable benefits as effective support for catalysis application. Nevertheless, the even electron distribution of theses sp2-hybrized carbon architecture heavily limits their intrinsic catalytic activity. Compared with sp2-C, sp-hybridized carbon possesses stronger electronegativity, and its linear orbital leads to long-range conjugation with no cis-trans isomerism. Herein, a novel type of carbon material with both rich sp2-C and sp-C may share combined advantages and exhibit intriguing physiochemical properties. The contradiction between the urgent eager for this co-hybridized carbon material and the awkward situation of their absence was finally solved by Prof. Yuliang Li's pioneer work in 2010 [8]. In their innovative achievement, the first sp-/sp2-cohybridized carbon allotrope, graphdiyne (GDY), was artificially synthesized via cross-coupling reaction using hexaethynylbenzene (HEB) as precursor. The emergence of GDY paves a new path for the development of diverse carbon materials in mild conditions through controllable precise chemical synthesis. Since this leaping-over breakthrough, GDY has attracted increasing interests of researchers and scientists, and caught focusing sights in various fields including catalysis, energy storage and conversion, biomedicine, optical devices, artificial intelligence and environmental governance [9][10][11][12][13][14][15].
A brief introduction on structural characteristics and physiochemical properties of GDY may help to understand how this novel one-atom-thick carbon allotrope brings its superiority into full play for electrochemical green hydrogen production, an important branch of catalysis field. In GDY's architecture, interconnected hexagonal benzene rings and linear -C≡C-C≡C-linkages weave into a two-dimensional (2D) planar network structure with long-range π conjugation. Especially, the alternate arrangement of sp-C and sp2-C units leads to uneven distribution of electrons along GDY skeleton. The naturally generated electron-rich centers are more energetically inclined to overlap with outer electron orbital of metal species, and this intimate interaction contribute to both structural stabilization and catalytic activation for diverse GDY-based electrocatalyst [16,17]. Moreover, those above-mentioned special electronic enrichment sites may function as intrinsically active centers, making GDY an all-carbon metal-free electrocatalyst directly [18]. This unique electron distribution also endows GDY with Dirac cones with direct band gaps whose value is calculated to be 0.44 eV and 1.10 eV according to different method, respectively [19][20][21]. Thus, GDY processes semiconductor nature with both silicon-like conductivity (2.56 × 10 −1 S m −1 ) and high charge carrier mobility (2 × 10 5 cm 2 V −1 s −1 ) [22]. Apart from the carbon skeleton, the natural triangular cavity is an important part of GDY's structural characteristics, which allows rapid mass transport and gas evolution during green hydrogen production process [23]. Furthermore, the artificially chemical synthetic method of GDY on arbitrary substrates [24] brings technological advantages for the fabrication of GDY-based electrocatalyst and corresponding electrode assembly.
Tremendous potential of GDY for green hydrogen production has been proved according to several recent laboratory explorations. By precise morphology control, Hui et al has reported an ultrathin GDY nanosheet catalyst possessing Pt-like hydrogen evolution reaction (HER) activities under universal pH [25]. And latest, similar catalytic performance of GDY was also realized using a chemical-bond targeted physical clipping strategy by Zhang et al [26]. These inspiring works establish fundamental of the recent rapid development of diverse multi-scale GDY-based electrocatalysts for ideally performed HER, oxygen evolution reaction (OER) and overall water-splitting (OWS), which will be detailedly introduced in the next section. At current stage, several principal strategies can be summarized from recent advances for rational design and controllable synthesis of GDY-based water splitting electrocatalysts, including interface engineering, individual atom fixation, induced constrained growth of catalysts and bottom-up modification, which are mainly guided by the two core demands of enhancing GDY's intrinsic activity and modulating GDY-metal interaction (figure 1). A summative discussion on these design strategies for GDY-based catalysts can be helpful to further understand the structure-activity relationship and deeply reveal the mechanism of water electrolysis, bringing great practical significance for accelerating the development of new generation commercial catalysts. Based on above considerations, in this short perspective, we firstly give a brief introduction on the principal strategies for GDY-based electrocatalyst and corresponding recent achievements on water electrolysis following its development context, where the internal logical relationship between intrinsic natural of GDY, catalyst design strategies and catalytic performance are clarified. Later, corresponding possible future directions are prospected, in the expectation of providing referential guidance for designing next generation GDY-based catalysts, so as to realize the large-scale production of low-cost green hydrogen.

Construction of GDY-engineered active interface
Transition metal (TM) materials have attracted focused attention for their moderate catalytic performance and controllable cost for electrochemical water splitting. Currently, nickel mesh or stainless mesh is universally used as electrode material in commercial alkaline water electrolyzers. Driven by the urgent anxious for the novel efficient TM material as alternative of precious metal, the combined heterostructure consisting of TMs and low dimensional carbon materials has attracted increasing research interests. Benefiting from the unique chemical/electronic structure and intrinsic physiochemical properties of GDY, along with its outstanding affinity with arbitrary substrate, constructing GDY-engineered heterostructure can no doubt be a promising solution. Due to the natural inhomogeneity of charge distribution and large π-conjugated network in GDY skeleton, numerous electron-rich centers contribute to the electronic perturbations from the interaction between heterostructure interfaces. And the seamless combination between the both components further facilitate an electron transfer stabilization mechanism. Such resultant effort leads to an extremely active interface within GDY-based heterostructure catalyst, which contributes to both higher intrinsic activities and abundant active sites. In 2016, Li's group made an earliest attempt using a cobalt/GDY heterostructure (CoNC/GD) as HER catalyst in universal PH [27]. Since then, continuous efforts on GDY-based heterostructure catalyst for efficient water splitting (HER, OER and OWS) has been reported with developing strategies and methodologies in interface construction or modulation, including GDY-engineered TM sulfides [28][29][30], oxides [31,32], nitrides [33], phosphides [34] and hydroxides [35][36][37][38][39]. And these works may help understand the critical role GDY acts in active interface engineering.
Due to its efficient electron-transport capability, GDY may function as an active part more than barely support in catalyst fabrication, which is favorable for efficient cathodic HER. In 2018, Yu et al [28] thus constructed a GDY/MoS 2 heterostructure (eGDY/MDS) for HER in both alkaline and acid conditions, successfully overcoming the obstacle of catalytically inert basal plane of MoS 2 . It is revealed that the combination of both semiconducting GDY and MoS 2 leads to an active interface with metallic conductive property. Electron transport from MoS 2 to GDY leads to a favored adsorption-free energy of hydrogen (△ GH ), contributing an even better catalytic activity than 20 wt.% Pt/C in alkaline electrolyte. Based on that, Hui et al [40] designed a GDY-nanosheet encapsulated MoS 2 catalyst (GDY-MoS 2 NS/CF) using a hydrothermal and in situ coupling method (figures 2(a) and (b)). Significantly increased interface surface area was constructed benefiting from the nanosheet morphology of both GDY and MoS 2 . And corresponding stronger electron hybridization leads to a phase transformation from semiconductive 2H-MoS 2 to conductive 1T-MoS 2 (figures 2(c) and (d)), which is decisive to the low overpotential of 90 mV at 10 mA cm -2 for alkaline HER electrocatalysis. Similarly, GDY successfully activated planar unsaturated sulfurs of CoS 2 in a recent work of a CoS 2 nanowire/GDY heterostructure catalyst [41], and lower △ GH was achieved owing to the interfacial electrons transfer between different active sites.
At present stage, the success implements of interface-engineering strategy can be explained from multiple structural characterizations, with the detection of interfacial electron interactions and chemical bond formations. As a typical example, Shi et al synthesized a GDY/CuS nanosheet HER catalyst in 2019 [42]. From the detailed high-resolution x-ray photoelectron spectroscopy (XPS) results, it is clearly revealed how GDY-engineered interface facilitates HER behavior. Compared with pristine CuS, obvious red shifts were discovered in both XPS Cu 2p 1/2 and S 2p 1/2 spectra of GDY/CuS, with 0.4 eV and 0.22 eV lower binding energy, respectively. The XPS results suggested GDY function as electron-donor to CuS and construct activated heterostructure interface. This intimate interface interaction was also persuasively evidenced by the XPS C 1s spectra. The emergence of an extra peak at 290.0 eV typically indicates the occurrence of π-π * transition, which can be ascribed to the restoration of the delocalized p conjugation and electronic interaction between GDY and CuS. From another work reported by Yu et al, an ultrathin GDY wrapped NiO nano-cube electrocatalyst (NiO-GDY NC) [31] was developed (figure 2(e)). From corresponding XPS Ni 2p 1/2 and 2p 3/2 results of the catalyst, 0.3 eV and 0.4 eV blue shifts were observed, respectively, indicating a different mechanism occurred on this kind of GDY-engineered interface, where GDY attract electrons from NiO component ( figure 2(g)). This electron interaction was further supported according to the theoretic investigation (figure 2(f)), contributing to superior HER activity with only 58 mV@10 mA cm -2 , and 28 000 ultra-stable working cycles in alkaline media (figure 2(h)). In a recent work of GDY-supported MoO 3 HER catalyst (GDY/MoO 3 ) reported by Yao et al [43] (figure 2(i)), x-ray absorption near edge structure (XANES) spectroscopy was performed to bring a deeper sight into the interfacial derivation when constructing GDY-based catalyst. In the C K-edge XANES of GDY/MoO 3 (figure 2(j)), the characteristic peak indicating aromatic of GDY skeleton shifted to lower energy, which suggested the electron cloud migrate to the triple bonds that attached to the carbon ring of GDY. This hinted that sp-hybridized carbon atoms take part in the formation of interfacial C-O-Mo bonds. Density functional theory (DFT) calculations on H 2 O adsorption energy further demonstrated that enhanced C-O-Mo interactions facilitates energetically favored water electrolysis (figure 2(k)).
Compared with HER, the sluggish OER process with much more complicated four-electron mechanism is believed decisive for the OWS kinetic. Heterostructure catalysts with GDY-engineered interfaces have also displayed thrilling potential in facilitating OER kinetic and boosting OWS efficiency. Currently, TM-based layered double hydroxides (LDHs) have triggered increasing interest for OER and OWS application. Hui et al [37] adopted a two-step synthetic method for integrating ultrathin GDY-wrapped Fe-LDH nanosheets on nickel foam (FeCH@GDY/NF). Benefiting from interface electron transfer evidenced by XPS results, both OER and OWS activities were optimized using this FeCH@GDY/NF as bifunctional catalyst. Later, a GDY-supported NiFe-LDH with nano-wall morphology was constructed by Shi et al, displaying both enhanced OER activity and stability in alkaline electrolyzer as well [38]. By applying an air-plasma method, Li et al fabricated a superhydrophilic GDY substrate possessing abundant hydrophilic oxygenic groups (-O-, -OH, -COOH, etc) with larger electron density, thus enhancing the interface interaction with positively charged CoAl-bimetallic LDH [39]. Accordingly, the OER overpotential required to reach 10 mA cm -2 in 0.1 M KOH was reduced by approximate 100 mV of the CoAl-LDH/GDY catalyst compared with hydrophobic GDY based ones. During the construction of in situ GDY-decorated FeCo-bimetallic LDH (e-LDH@GDY/NF, figure 3(a)), Hui et al [35] found that GDY sheets were intercalated into LDH structure, replacing the pristine carbonate anions within interlayer regions of LDH stacking, and leading to the exfoliated ultrathin LDH nano-sheets (figures 3(b) and (c)). It has been proved to significantly reduce transition state barrier for water electrolysis accordingly. For OER in 1 M KOH, only 216 mV overpotential was needed at 10 mA cm -2 current density (figure 3(d)) with 37 000 cycles' ultra-stable durability, while for OWS, a low voltage of 1.47 V was applied to drive 100 mA cm -2 current density. This intriguing discovery suggested a brand-new path for the GDY-engineer interface construction strategy.
Similar experience can be rationally extended for designing those other GDY-decorated TM systems. Qi et al developed a 1D nano-wire MnCo 2 O 4 catalyst (NW-MnCo 2 O 4 /GDY) from corresponding LDH precursor [32]. Owing to the seamless coating of GDY layer, significant synergistic effect occurs on the interface between GDY and both Mn and Co spices. And the resulting exposure of more active sites leads to the improved OER efficiency. Likewise, a GDY-wrapped cobalt nitride catalyst [33] featuring porous nanosheet morphology (CoN x @GDY NS/NF) has also been prepared following a three-step syntesis route of hydrothermal-nitridation-GDY coating (figure 3(e)). With CoN x @GDY NS/NF as HER/OER bifunctional electrode in an alkaline water electrolyzer, a very small cell voltage of 1.48 V was required to reach the current density of 10 mA cm −2 (figure 3(f)). Subsequently, a 24 h chronoamperometric test further verified the reliable OWS durability (figure 3(g)). In 2021, Li's group synthesized well-ordered and large area nickel based ultrathin charge-transfer complexes (CT) nanosheets on GDY network (CTNS/GDY) [44]. It was found that Ni-O coordination adjustment and partial charge transfer from CT to GDY contribute to a fast reversible redox-switching mechanism, which dominates accelerating OER kinetic with ultra-high activity (155 mV @ 10 mA cm -2 ). Very recently, Gao et al [34] employed a hydrothermal-in situ coupling-phosphating method, and fabricated a GDY reinforced Cu/Ni bimetallic phosphide heterostructure (CuNiPx-GDY, figure 3(h)). The electrocatalyst with columnar-array-like nanosheet assembly displayed superb OER activity (178 mV @ 10 mA cm -2 ). According to the detailed theoretical investigation, five types of possible active sites were constructed on the GDY-engineered interfaces (figure 3(i)).
As we have seen, benefiting from the first mover advantage, multiple catalysts with GDY-engineered active interface have already shown great potentials in water electrolysis. Current achievements are mainly concentrated on electrochemical water splitting in alkaline environment. While for the proton exchange member (PEM) water electrolyzers, corrosive acid electrolyte brings great obstacle for the chemical stability of TM structures and the maintenance of functional GDY-TM interfaces as well. Till now, limited breakthrough has been reported on this interface engineering strategy for developing long-term stable HER/OER bifunctional catalyst toward PEM systems, which may be a critical technical concern in the forthcoming stage. After summarizing the previous efforts, by taking composite mode and morphology regulation into consideration, we may carefully draw that the controllable in-situ coating of ultrathin GDY layers is still the optimal choice. And the dual role of GDY for both mechanical/chemical protection and catalytic activation will be increasingly highlighted for future construction of GDY-engineered catalysts.

Fixation of individual atomic active site
Though interface engineering has been proved a widely applicable strategy for designing GDY-based catalyst, the unexposed internal structure of metal species is actually catalytic inert. Single atom catalysts (SACs) emerge as a heated topic for maximizing metal atom efficiency, allowing each metal atom capable functioning as catalytically active site and fulfilling maximum surface free energy, which endow them with super-high activity and selectivity. Driven by the bright blueprint, increasing efforts have been continuously contributed to developing various SACs systems since 2011 [45]. Nevertheless, several critical challenges still remain intractable: (i) the ineluctable agglomeration tendency destabilizing individual distribution of metal atoms; (ii) undetermined configuration and valence state of SACs leading to an ambiguous catalytic mechanism at atomic level; (iii) complicated synthetic procedure and harsh conditions restricting the easy acquisition of SACs. The unique structure and property of GDY make it an ideal platform for the stable fixation of individual catalytically-active metal atom. Mostly, the coexisting of sp-C (p x -p y π/π * state) and sp2-C (p z π/π * state) allows the rotation of π/π * orbitals toward arbitrary direction perpendicular to the leaner diyne linkage, which affords an intimate p-d orbital overlapping for stabilizing adjacent metal atoms. And the triangular cavity within GDY structure serves optimized sites for the fixation of the individual atom. Since 2018, a series of vanguard works by Prof. Yuliang Li's group forcefully substantiated that it is an effective strategy to construct GDY-based atomic catalyst, not only for efficient hydrogen production via water electrolysis [46][47][48][49][50], but various applications including electrochemical ammonia synthesis [49,51], CO 2 reduction [52] and methanol oxidation [53]. And combined theoretical and experimental results continuously unveiled new effects and properties of those individual active sites designed by GDY fixation.
At the moment, for accelerating water electrolysis kinetic, efforts on GDY-based atomic catalyst mainly focus on cathodic HER. In 2018, Li's group [46] firstly anchored individual Fe and Ni atoms on GDY (Fe/GD and Ni/GD) through a simple in-situ electrochemical reduction method (figure 4(a)). High-angle annular dark-field (HAADF) images clearly proved the uniform distribution of both singly fixed Ni and Fe atoms. Extended x-ray absorption fine structure (EXAFS) spectra further confirmed metal species exists predominantly as isolated atoms, with the dominant characteristic peak indicating metal-C interaction. A unique unprecedented discovery is that for both Fe and Ni atoms anchored on GDY, the first derivative of their XANES curves in pre-edge zone are coincident with those of corresponding bulk metal references, which suggests the identical zero-valence with corresponding metallic simple substance (figures 4(b) and (c)). DFT analysis revealed that both Ni and Fe atoms are energetically inclined to be anchored on natural triangular cavity corners rather than hexagonal benzene rings in GDY architecture, and stabilized by the adjacent acetylene units. Theoretical studies also indicated that the metal-C interaction in this system performs as orbital charge overlaps rather than conventional covalent/ionic bonds, which brings about a specific incomplete CT effect to stabilize zero-valent metal atoms ( figure 4(d)). The fixation of individual zero-valent atomic active site on GDY leads to significantly boosted HER kinetic with higher mass activities than most of reported traditional bulk catalyst, nano-heterostructures and even SACs.
For commercial PEM water electrolyzers, GDY-supported precious-metal catalysts offer an advisable solution for controllable equipment cost. Following a similar electrochemical reduction method, Yu et al fabricated an atomic Pd catalyst on ultrathin GDY nanosheet (Pd 0 /GDY) [47]. Both theoretical studies and characterization results (HAADF, EXAFS, XANES) firmly proved the successful fixation of uniformly dispersed zero-valent Pd atoms on the same alkyne ring site as Fe/GD and Ni/GD systems. And it is worth noting that XPS Pd 3d spectra firstly gives a direct proof for the major existence of Pd 0 in this work (figure 4(e)). Benefiting from the charge compensation mechanism originated from strong p-d overlapping, 55 mV overpotential was required to deliver 10 mA cm -2 current density for Pd 0 /GDY with only 0.2 wt.% Pd loading in an acidic HER electrolyzer (figure 4(f)). Likely, Yin et al [48] prepared a GDY-supported Pt atomic catalyst (Pt-GDY2) for HER, fulfilling a 26.9 time-higher mass activity than commercial Pt/C. Moreover, Li's group employed a solvothermal method to synthesize zero-valent atomic Mo catalyst (Mo 0 /GDY) with high loading amount of 7.5 wt.% [49]. For the first time, the theoretically predicted alkyne-ring-corner anchoring site of GDY-supported atomic catalyst was directly observed from multiple HAADF images ( figure 4(g)).
With determined configurations and metal valence, Mo 0 /GDY performed superior HER activity to that of Pt/C (48 mV vs. 59 mV@10 mA cm -2 ). Later, Li's group similarly developed Cu 0 /GDY catalyst via in-situ reduction treatment and contributed to facilitated HER performance [54].
Researchers have also made struggling attempt to employ this individual-atom-fixation method for efficient OER. In 2020, Li's group anchored ruthenium atoms on GDY (Ru-GDY) through a simple electrochemical reduction treatment [50]. While obviously distinguished from previously reported zero-valent atomic atoms, Ru presented in tetravalence in this system. The DFT simulation indicates that the strong coupling between Ru atoms and adjacent acetylenic -C atoms makes atomic Ru a unique electron-mediating-vehicle for fast reversible redox-switching. For OER in 0.5 M H 2 SO 4 , the atomic catalyst showed exceedingly higher mass activity and turnover frequency (TOF) than corresponding Ru nanoparticle reference and commercial RuO 2 . However, both the required overpotential for OER (531 mV@10 mA cm -2 ) and the applied bias for OWS (1.81 V@10 mA cm -2 ) suggest an unsatisfied improvement far from being adequate.
Summarizing these implemented examples of individual atom fixation strategy, we can judiciously conclude that GDY-based zero-valent atomic catalysts have undoubtedly brought key breakthrough in acidic HER. Nevertheless, corresponding expanded investigation for HER in alkaline electrolyte still remains blank, which is in the contrary situation to those GDY-engineered heterostructure catalysts. For the extensive screening of valid candidate, Sun et al have theoretically investigated 26 TMs from IIIB to IB of 3d to 5d electronic configurations in the periodic table, to provide guidance for designing GDY-fixed atomic TM electrocatalysts with zero-valence [55]. And according Huang's another work in 2020 [56] using combined technique of DFT and machine learning, for those TMs and lanthanide elements, the fixation of individual atom on GDY leads to kinetically favored HER performance in perspectives of the adsorption energies, adsorption trend, electronic structures, reaction pathway, and active sites. On the other hand, atomic catalysts bifunctionally efficient for both HER and OER are badly desired yet. The only reported work (Ru/GDY) [50] to date may provide some hints for tentative exploration, where tetravalent Ru functions as advantageous electron acceptor for OER. Given that the determined structure of GDY-based atomic catalysts has been omnidirectionally evidenced, the coordination environment of metal atoms can be optimized by introducing some auxiliary groups [48], which may contribute to favored OER kinetic. Also, we believe the fixation of bimetallic atoms will bring another optional solution. By precisely controllable anchoring individual atoms of different metal elements within one acetylene ring, the complex ternary electron interactions between adjacent metal atoms and GDY may create fully-new opportunity for OER in diverse electrolytes. And such intriguing attempt has never been experimentally reported yet. Anyway, we are looking forward the victorious news of electrocatalysts realizing highly efficient OER and alkaline HER using this GDY-dominated individual atom fixation strategy.

Induced constrained growth
Encouraged by the exhilaratingly successful strategies of GDY-engineered interface and individual atomic site fixation, researchers bend their efforts for the combination of advantages of the stable heterostructure and active atomic catalysts for efficient water electrolysis recently. In several latest works, an emerging strategy was reported for designing a new category of GDY-based catalysts. By rationally regulating catalyst synthetic conditions, the abundant natural cavity of GDY and the incomplete charge transfer mechanism contribute to both spatially and energetically favored anchoring and subsequent aggregation steps of metal atoms. This GDY-induced nucleation and grain-growth kinetic may lead to assembling of constrained-sized metal catalysts at quantum dot level. With such crystal size, electrons in the material are confined in all three dimensions to a length scale of the order of the Fermi wavelength. Owing to the size confinement, quantum effects will dominate over the bulk properties, and atom-like discrete energy spectrum can be observed. These unique phenomena were found advantageous for creating numerous catalytic centers and facilitating intrinsic activities toward electrochemical water splitting.
As a very recent example of this GDY-induced constrained growth strategy, Zhang et al [57] developed a simple self-reduction wet chemical method to prepared uniformly distributed 3.84 nm sized Pd quantum dots on GDY (GDY-Pd1). In this work, a special triple-anchored Pd crystal growth mechanism was illustrated. The GDY-induced arrangement sequence of Pd dominated the construction of single facet (111) structure, leading to a nearly thermal neutral H * adsorption free energy (∆GH * ≈ −0.08 eV). And thus, notable catalytic activity was realized with only 261 mV overpotential to deliver a large current density of 1000 mA cm -2 for HER in acidic media. In 2021, Gao et al fixed bimetallic vanadium-ruthenium oxide clusters on porous GDY with constrained size of 2.75 ± 0.04 nm (VRuO x /GDY, figure 5(a)) [58]. The success modulation of bimetallic synergism was verified according to EXAFS measurements with corresponding chemical bonds evolution ( figure 5(b)). As the result, VRuO x /GDY exhibited the best-performed durable HER activity with approximate 10 mV overpotential to afford 10 mA cm −2 in both alkaline and neutral environments (figures 5(c) and (d)). Later, Gao et al applied this GDY-induced constrained growth strategy to assembling of 2.86 ± 0.02 nm sized bimetallic niobium-rhodium oxide on ultrathin GDY nanosheets (Nb y RhO x /GDY, figures 5(e) and (f)) [59]. By coordination state modulation and electron distribution tailoring, Nb y RhO x /GDY overwhelmingly outperformed commercial Pt/C catalysts in HER activity together with 20 000 cycles' stability. As for OWS investigation, Wang et al [60] reported an IrO x quantum dot catalyst via GDY-induced in situ growth (IrO x /GDY). With average size of 1.48 ± 0.04 nm and modulated Ir 4+ /Ir 3+ ratio, IrO x /GDY features larger electrochemical active surface area (ECSA), higher conductivity, faster CT behavior, which are favorable for both accelerated HER and OER kinetic in acidic electrolyte.
The reactive combination of electrocatalysis and photocatalysis furnishes a more efficient and direct path for the transformation of renewable solar energy to green hydrogen. Metal quantum dot catalysts from GDY-induced constrained growth feature enlarged band gaps to energetically drive charge transfer and facilitate photocatalytic reaction. And strong quantum-confined effect promotes kinetically favored photoelectroncatalytic (PEC) water splitting. As early as 2016, with assistance of introduced 4-mercaptopyridine functional group, CdSe quantum dots were firstly constrained on GDY by Wu et al [61]. The as-synthesized CdSe QDs/GDY shows exciting PEC performance with a HER rate of 27 000 µmol h -1 g -1 cm -2 . In this work, GDY played multiple roles as both exciton recombination inhibiter and charge transfer promoter. In 2021, Du et al directly fabricated GDY-induced OsO x quantum dots with approximate 2.6 nm in size (OsO x QDs/GDY) [62]. The incorporation between GDY and OsO x leads to high coordination environment of O 4+ , which can be a promoter for HER performance (42.5 mV@100 mA cm -2 ) under irradiation. In a latest report on GDY-supported molybdenum oxides (GDY@MoO x ) [63] by Chen Apart from the efficient utilization of solar energy, direct seawater electrolysis provides a promising choice for the conversion from offshore wind energy to green hydrogen. Very recently, Gao et al took the advantage of GDY-induced in situ nucleation and growth mechanism (figure 5(g)), successfully prepared Rh nanocrystals on GDY (Rh/GDY) with size of 5.16 ± 1.20 nm by formic acid-assisted wet chemical reduction [64]. Benefiting from the complementary reductive effect of GDY and the formation of sufficient sp -C∼Rh bonds (figure 5(h)), controllable regulation and stabilization of high-density atomic steps on the faces of Rh nanocrystals were realized according to detailed HAADF observation ( figure 5(i)). The abundant unsaturated coordination sites on those stepped Rh edges (figure 5(j)) function as much more active centers for HER in saline environment, which leads to remarkably accelerated catalytic activity (65 mV@1000 mA cm -2 ) as well as long-lasting stability (8000 working cycles, figure 5(k)). Later, the similar constrained growth strategy was employed for designing a GDY-RhO x crystal-GDY bilayer catalyst (GDY/RhO x /GDY) [65]. The double layer GDY networks facilitate the generation of increasing sp-C∼O-Rh active centers, resulting in superb seawater electrolysis performance with a small applied bias of 1.52 V to reach 500 mA cm -2 .
Above recent achievements have already accessed to several highly efficient GDY-based catalysts for HER, OER and OWS in electrolyte systems with wide pH range. Though corresponding researches are still in infancy stage, we can reasonably predict that this GDY-induced constrained growth strategy will dominate the popularity in the forthcoming period. As we are seeing, GDY-induced size effect leads to more exposed metal atoms as catalytically active sites. In contrast with heterostructure catalysts obtained from traditional interface engineering, these metal quantum dot catalysts thus provide feasibility for characteristically accurate investigation of surface coordination environment, defects as well as metal-support bonding. Moreover, the resulting unique quantum effect further delivers opportunity for those novel catalysts in PEC systems. While compared with anchored individual metallic atom center, the constrained nanoscale crystal allows more flexibly in omnidirectionally modulating catalysts' morphology, composition, valence state and other structural features. As another future tendency, extensive implements of GDY-induced constrained growth strategy may guide heated research focus on controllable fabrication of electrocatalyst at sub-nano cluster level [66].

Bottom-up fabrication for novel GDY architecture
Paralleling with the development of those strategies for incorporating GDY with metal materials, accelerating efforts are currently paid for constructing active sites within GDY skeleton directly to realize efficient water electrolysis. Benefiting from the unique cross-coupling synthetic approach of GDY using HEB as precursor, bottom-up fabrication is a reliable option for constructing novel GDY architectures based on molecular engineering of monomers. By using this method, determined heteroatom doping, functional group modification and periodically structural arrangement can be conveniently achieved. Those GDY derivatives from bottom-up fabrication feature tailored electronic structures, modulated band gap, enhanced conductivity and charge mobility, providing the possibility to optimize their intrinsic catalytic activity. Moreover, novel GDY architecture with determined modification sites and content may provide an ideal platform to accurately probe the catalytic mechanism during electrochemical hydrogen production process.
In 2019, inspired by the synthesis method of HEB, Xing et al [67] prepared a novel 1,3,5-triethynyl-2,4,6-trifluoro-benzene monomer with coupling of 1,3,5-trifluoro-2,4,6-triiodobenzene and ethynyltrimethylsilane, and subsequently followed a similar cross-coupling process to successfully prepare 3D porous fluoro-GDY nanostructures on carbon cloth substrate (p-FGDY/CC, figure 6(a)). Owing to this bottom-up synthetic strategy, massive strong C-F bonds was cleverly created, contributing to significant charge-polarization and enhanced electron-localization near adjacent C sites. And as expected, p-FGDY/CC exhibited excellent HER, OER and OWS performance in electrolyte with wide pH range (0-14). Especially, low overpotentials of 82 mV and 92 mV were needed to reach 10 mA cm -2 for HER in alkaline and acidic environment, respectively (figures 6(b) and (c)). Later, He et al developed triamino-GDY (TAGDY) electrocatalyst with ultrathin nanosheet morphology following similar bottom-up synthetic steps [68]. In TAGDY architecture, the acetylenic linkages and amino groups are alternately distributed on hexagonal benzene rings. And the resulting charge redistribution makes C2 sites electron-rich centers for favored H+ adsorption. Meanwhile, an intriguing alkyne-alkene transition phenomenon is revealed from C-C bonding length change by theoretical investigation (figures 6(d)-(f)), which may function as another energetic booster for the facilitated HER performance (82 mV@10 mA cm -2 ). And according to another work of Sakamoto et al a pyrazine-incorporated GDY analogue (PR-GDY) featuring pyrazinic core was polymerized using liquid/liquid interfacial synthetic method based on tetraethynylpyrazine monomer [69]. The precise introduction of pyridinic-N is believed to be responsible for the enhancement in HER activities of PR-GDY in various electrolytes. As for PEC hydrogen production, the bottom-up fabrication strategy has also proved its extensive possibilities. Li et al [70] selected 1,3,6,8-tetraethynylpyrene (TEP) and 1,3,4,6-tetraethynylbenzene (TEB) as precursors for the acquisition of pyrenyl GDY (Pyr-GDY) nanofibers and phenyl GDY (Phe-GDY) nanosheets, respectively (figure 6(g)). And optimized HER performance was realized owing to the special electronic structures and modulated band gaps of the both architectures (figures 6(h) and (i)).
In fact, advanced studies on bottom-up synthetic methods have already prompted a series of GDY derivatives. While only a moderated part of corresponding investigations was involved in catalysis field, not to mention that for water electrolysis. To our knowledge, the design strategy of fluoro-GDY can be similarly applied for preparing hydrogen-substituted GDY (HsGDY) [71] and chlorine-substituted GDY (Cl-GDY) [72] featuring the identical sites. Pyridinic and triazinic nitrogen substitution within GDY architecture has also been realized [73]. And a more radical attempt is to substitute a single boron center for the entire benzene ring to fabricate GDY analogues consisting of only sp-C and B atoms [74]. While moving our sights to the diyne linkages, Zhao et al has proved the precise sp-N cation substitution brought about an impressive GDY-based electrocatalyst for oxygen reduction reaction [75]. Thus, based on current situation, we suggested that corresponding supplementary researches will gradually unleash potentials of these bottom-up designed GDY derivatives as efficient water electrolysis catalysts. Of course, continuous development of this strategy is hopefully proceeding as well with emerging novel precursor or GDY network structures. Besides heteroatom doping/substitution or functionalized modification, bottom-up fabrication strategies may also be implemented for modulation spatial configuration of novel GDY architectures. In 2021, Wang et al [76] artificially synthesized two kinds of 3D GDY analogues (tetraphenylmethane-GDY and tetraphenylmethane-GDY, denoted as TPM-GDY and TPN-GDY, respectively) with numerous exposed 3D spatially extended acetylenic bonds. And both of TPM-GDY and TPN-GDY performed promising HER activities with overpotentials of 105 mV and 211 mV at 10 mA cm -2 , respectively. Anyway, benefiting from synchronous development tendency of bottom-up designed GDY derivatives for various applications, future blueprint of GDY-based metal-free catalysts toward water electrolysis is quite clear.

Future perspectives
From the examples presented above, we reviewed the present principal strategies for designing and fabricating GDY-based catalysts toward efficient water electrolysis. Although encouraging progresses have firmly proved the availability and success of the above-mentioned design strategies, additional investigations and further improvements are still required for the commercial application of those various GDY-based catalysts. And in our perspective, the current challenges and future opportunities lie in the following aspects, which should be specifically discussed.
Firstly, an important point for developing new-generation electrocatalyst is the further exploration of structure-performance relationship, which relies on various surface science characterization techniques. Typically, XPS, x-ray diffraction, Fourier transform infrared spectroscopy, Raman spectroscopy, XANES spectroscopy, extended x-ray absorption fine structure spectroscopy (EXAFS) and other spectroscopic techniques are usually performed to for structural characterizations of materials. For catalyst systems involving GDY and metal components, the construction of massive sp-C∼ metal sites is undoubtedly the main origin of enhanced catalytic activity, which serve as the guiding ideology of current strategies of interface engineering, individual atom fixation and catalyst constrained growth. From earlier XPS studies to present XANES and EXAFS determinations, the sp-C∼ metal interaction can be directly identified rather than ambiguous observation of general electron transfer. We hope the formation mechanism of this sp-C∼ metal site near reaction interfaces between catalysts and electrolyte can be revealed with developing characterization technics, which benefits instructively designing novel GDY-metal catalyst systems at the next stage. And for GDY derivatives obtained by bottom-up strategy, diverse spectra characterizations will be continuously expected to verify the precise design of doped or modified precursors, and corresponding successful cross-coupling. Combined with the accurate evaluation of electronic properties, efforts on intrinsic activity optimization will be the long-term goal for developing bottom-up synthesis. While extensive morphological characterizations including scanning electron microscopy (SEM), transmission electron microscopy (TEM), HAADF scanning transmission electron microscopy (HAADF-STEM) images, have been providing substantial reference for the optimization of catalysts' ECSA. At present, employing of HAADF-STEM makes it possible for direct observation of catalysts at atomic level, and has already contributed to the remarkable success of individual atom fixation strategy. We can foresee that in recent future, increasing attention will be paid to the direct investigation of the atomic arrangement at the GDY/metal interface, and high-resolution images of dopant or functional group sites within GDY derivative structures if possible.
Clear illustration on electrochemical OWS mechanism is also naturally important for designing GDY-based catalysts. Universally, possible catalytic mechanisms can be inferred from comprehensive comparison of catalyst's structural or morphological information before and after water electrolysis. It has been widely reported that the reconstruction behavior of metal catalyst may occur after electrochemical process, including compositional changes, crystal phase transitions and coordination adjustments [77]. Especially for OER, the formation of metal oxides or (oxy)hydroxides contributes to surface activation of those actually regarded as pre-catalyst [78][79][80]. This reconstruction phenomenon has gradually been revealed according to recent explorations on GDY-metal systems. In 2021, Li's group illustrated the transformation of nickel species to corresponding oxy-compounds in CTNS/GDY catalyst during OER process. The gradual formation of surface Ni(OH) 2 and NiOOH phases was determined with N 2+ /Ni 3+ ratio increasing from 1.35 to 9.96 after 2500 electrochemical cycles, which was further supported by ex-situ measurements (figures 7(a)-(c)) [44]. And for IrO x nanocrystal from GDY-induced constrained growth, increased percentage of Ir 4+ indicated abundant newly formed metal-OH interfacial structures as OER proceeding in acidic media (figures 7(d)-(g)), leading to a self-modulation mechanism toward efficient water splitting [60]. Those latest findings guide a positive direction for developing strategies on GDY-based catalysts in the perspective of working mechanism. And here we suggest that future attentions can be complementarily paid on corresponding surface reconstructions on those GDY-involved HER catalysts, which is still hidden in the dark as to our knowledge. While from recent peer works, it is increasingly notable that the developing operando characterization technics under electrochemical conditions is leading the future tendency for catalytic mechanism probing [81][82][83]. Thus, we are looking forward that various emerging operando spectroscopy and imaging technics performed on GDY-based catalysts will facilitate detailed observation of real time structure evolution, accurate identification of critical intermediate, and direct real spatial information analysis of reaction kinetics at next stage. Particularly, we anticipate the gathering of more direct and convincing proofs for the substantial role GDY plays in dynamic electrochemical OWS processes. On this basis, we can truly judge the reliability of expected structural and morphological properties of the catalysts at atomic/molecular level, contributing to expanded strategies for cleverly designing GDY-based catalysts.
In the view of practical application demand, the establishment of more realistic water electrolysis systems will help provide more authentically referential guidance for developing novel GDY-based catalysts. For commercial alkaline and PEM water electrolyzers for industrial green hydrogen production, electrolyte temperature is typically maintained at approximate 50 • C-90 • C for effective running, while the working current density is universally set at hundreds or even thousands of mA cm -2 to meet different hydrogen generation rates. Researchers have already bent their efforts for evaluating electrochemical activities and corresponding long-term stabilities of GDY-based catalysts under industrially large current density (figure 7(h)) [35,57,64]. While far to our knowledge, investigations on catalytic temperature of GDY-based catalyst still remains scant, which should be emphasized in the future work. Till now, there are only few related works for estimating electrochemical activation energies involving the influence of different temperature, which mainly focus on catalytic activities (figures 7(i)-(k)). In fact, we suggest that for those reported GDY-based catalysts, corresponding HER, OER and OWS performance studies at industrial temperature should be supplemented to give undoubtable evidence for success implement of the strategies discussed above, both laboratorially and practically. In addition, it is found that for several water electrolysis catalysts, totally distinct structure-performance relationships and catalytic mechanisms can be detected under industrial temperature and ambient conditions, respectively. For instance, Liu et al [84] reported a thermally induced complete reconstruction phenomenon of nickel molybdate (NiMo 4 ) OER catalyst in alkaline electrolyzer at 51.9 • C, which involves the formation of (oxy)hydroxide, generation of grain boundaries and creation of oxygen vacancies. While this kinetically promotive complete reconstruction is forbidden at room temperature. Considering the fact that GDY and its derivatives are thermodynamically stable under operation temperatures of commercial water electrolyzers, there should be specific concerning on such temperature influence for designing GDY-metal catalyst systems, while their catalytic inertness at room temperature can be very deceptive. Anyway, for the present GDY-based water electrolysis catalysts, the absence of real condition simulation leads to not only incomplete performance database, but the ambiguous learning of realistic catalytic mechanism in practical hydrogen production device. Therefore, we can assert categorically that a series of works on GDY-based catalyst under more really simulated water electrolyzer environment will successively emerge, which can be of highly referential value for designing catalysts with practical potentials.

Conclusions
To sum up, we discussed four principal strategies for current design of GDY-based catalysts toward electrochemical water splitting, including construction of GDY-engineered active interface, fixation of individual metal atom on GDY network, GDY-induced constrained growth for metal quantum dots, and bottom-up fabrication for novel GDY architecture. From the adequate persuasive example above, we have seen the remarkable success of these strategies in creating more active sites and increasing the ECSA. And there are convincing reasons to believe that GDY is far from fully exerting its potential for electrochemical green hydrogen production. We may foresee the further understanding of structure-performance relationship and catalytic mechanism of GDY-based catalysts under industrial water electrolysis conditions. And large-scale synthesis of high quality GDY with large area, few layers and high crystallinity is gradually being possible. Therefore, many new concepts based on GDY will be more and more comprehensively verified, including the surface electron heterogeneity, the incomplete charge transfer phenomenon, the multicavity space limiting effect, the alkyne-alkene conversion mechanism of chemical bonds, the in situ induction of constrained growth, and etc. We are hopefully expecting the vigorous development of the design strategy of GDY-based catalyst in the forthcoming period, and it may be the optimization and combinations of current strategies, or the emergence of creative and revolutionary ones. We believe these fundamental studies on GDY-based catalysts design strategies will bring vitality and inspiration to the development of next-generation commercial water electrolyzers, and finally witness the prosperity of green hydrogen industry.