and ethylene glycol are generally used as reductants. For transition‐metal nanoparticles (Fe, Co, Mn), NaOH or aqueous ammonia is often used to hydrolyze metal salts (to metal hydroxide precipitates). Sometimes, reduction is carried out under high pressures using supercritical water or CO2 where the graphene support would act as a reductant (carbothermal reduction). Although small noble metal clusters (< 2 nm) could be formed directly on rGO without adding any reductant or capping agent [10], the alternative is to presynthesize the fine nanoparticles using the organometallic approach, and followed by the deposition on rGO and removal of capping agents.
TrGO tends to have higher concentration of defective (heteronuclear) sites, which turns to be beneficial as the adsorption sites for metal cations. With its mildly reducing behavior, the rGO sheets subsequently act as electron donors for metal clusters to grow on its surface. The formation of small clusters only occurs when using rGO with low density of heteronuclear sites, where nucleation sites are distanced adequately. If the GO substrate is not sufficiently reduced, the rGO contains excessive nucleation sites and results in large nanoparticles. In a recent work [11], the reducing potential of graphite intercalation compounds (GICs) as precursors for graphenide solutions is used to deposit transition‐metal nanoparticles and metal oxides on graphene under mild conditions and without the use of other reductants. Small Fe nanoparticles (2–5 nm) were prepared by this method.
In general, the main problem encountered when using graphene as a catalyst support is that the basal planes lack the anchoring points for the metal and the metal sinters in the subsequent treatments of removal of capping agents or reduction. A nonpolar solvent or a mildly polar solvent is preferred for impregnation of the metal precursor. The advantage is that graphene is intrinsically reductive (albeit mildly) to circumvent the use of reductants. Sometimes, sacrificial stabilizing agents such as metal oxides are used to prevent the sintering of nanoparticles during thermal treatment [12]. Another option is to use GO to anchor a metal precursor and reduce both GO and the metal precursor simultaneously, i.e. GO reduced to rGO and metal precursor to metal nanoparticles. This approach is explained in Section 4.2.2.2.
4.2.2.2 Graphene Oxide as Precursor of Graphene‐Supported Catalyst
This method is derived from the preparation described in Section 4.2.1, but the reduction is stronger and leads to the simultaneous reduction of both the metal precursor and GO. This is the most commonly used method to prepare metal/graphene catalysts due to the strong interactions of GO with the metal precursor and the possibility of carrying out impregnation in aqueous phase. A composite based on metal ions and GO is prepared first, followed by a harsh reduction treatment using a reducing agent either in liquid or gas phase. The final product consists of metal nanoparticles on rGO. When the reduction is carried out in liquid phase, nucleation and reduction of the metal take place simultaneously. This is the case of solvothermal reduction in liquid phase at moderate temperatures. Generally, the mixture of GO, metal precursor, and reducing agent (e.g., hydrazine, ethylene glycol, NaBH4) is heated (100–150 °C) by conventional heating or microwave, leading to the simultaneous reduction of GO that becomes rGO, resulting in a metal supported on graphene material. The simultaneous reduction of GO and metal in solution usually leads to a broad metal distribution or to relatively large nanoparticles (50–100 nm).
The reduction can also be carried out in gas phase (e.g., H2 gas) after drying the pre‐impregnated metal precursor on GO. The method involves several steps of impregnation, drying, and reduction, leading to nanometric particles, i.e. Pd and Pt nanoparticles of around 2 nm, whereas the size of Ni and Mn nanoparticles is between 4 and 6 nm [13]. The first step is the intercalation of a chosen metal salt into GO via impregnation, forming the metal salt anchored on GO oxygenated groups. In the second step, this composite undergoes an explosive reaction or “popping” at around 200 °C, leading to the formation of large‐surface‐area graphene with well‐dispersed, partially decomposed metal precursors strongly anchored on the graphene sheets.
A special case is the preparation of catalyst nanoparticles on 3D macrostructures consisting of 2D graphene nanosheets with a foam‐like structure, namely, graphene aerogels. When GO is reduced under solvothermal mild conditions, the graphene nanosheets self‐assemble into 3D hydrogels, which can be dried subsequently by freeze‐drying or supercritical CO2 to produce graphene aerogels [14]. These 3D structures are favorable as structured catalytic reactors due to the ease in handling and separation. It is very interesting that the formation of both metal nanoparticles and the graphene aerogel support can take place in a one‐pot solvothermal synthesis using hot water as a mild reducing agent. The addition of divalent and trivalent ions (e.g. Ca2+, Mg2+, Cu2+, Pb2+, Cr3+, Fe3+) in GO dispersion promotes the formation of GO hydrogels [15]. Fe2+ ions are anchored onto the oxygenated functional groups of GO and hydrolyzed at acid pH to give Fe oxide nanorods (60 nm) or precipitated as oxide nanoparticles (30 nm) at basic pH [16]. By adding a reducing agent, 10 nm Fe oxide nanoparticles wrapped by graphene aerogels could be prepared [17]. By using also the one‐pot hydrothermal reduction method but adding some mild reductants such as sodium citrate and sodium acetate, the size of Fe3O4 clusters was reduced to 5 nm on graphene aerogel [18]. Adding ascorbic acid as reductant and 100 °C solvothermal treatment, 5–13 nm noble metal nanoparticles have been prepared on graphene aerogels[19]. Instead of using aqueous dispersion of GO, a GO dispersion in ethylene glycol and subsequent hydrothermal method was used to prepare 4–6 nm Ru nanoparticles on graphene aerogel [20]. It is possible to decouple the formation of the rGO hydrogel and the deposition of the metal using a two‐step approach. By capitalizing on the advantage of high water content within the rGO hydrogel, an aqueous solution of the metal is infiltrated into the hydrogel. This approach has been used to introduce MoS2 in the hydrogel, which is subsequently freeze‐dried [21].
4.2.2.3 Graphene Derivatives: Doped Graphene and Synthetic Derivatives
Graphyne and graphdiyne are synthetic derivatives of graphene. They are synthetic flat single atomic layers consisting of carbon hexagons connected by linear carbon chains instead of only carbon hexagons as in graphene. In graphyne structure, the hexagons are bonded by linear acetylenic chains, whereas in graphdiyne the hexagons are bonded by two acetylenic chains (Figure 4.2A–C). Graphdiyne is a new man‐made carbon allotrope prepared from a molecular precursor (hexaethynylbenzene) that possesses uniform 18 C‐hexagonal pores formed by three butadiyne linkages (‐C between the benzene rings), which can provide ideal anchoring sites for SACs with high stability as demonstrated by the results of theoretical calculations and in experiments in hydrogen evolution reaction (HER) [22]. The research about these graphene derivatives as catalyst support is almost unexplored due to the novelty of the material. It is foreseen that these derivatives are more amenable to functionalization with SACs than their parent material graphene. Therefore, a research field remains open for new researchers.
Figure 4.2 Structure of some 2D structures related to graphene: (A) graphene; (B, C) synthetic graphene derivatives (graphyne and graphdiyne); (D) graphitic carbon nitride, based on heptazine unit (a) and triazine unit (b); (E) covalent triazine frameworks (CTFs) with different structures: (a) CTF‐1 with
