The sustainable production of hydrogen is key to the delivery of clean energy in a hydrogen economy; however, lower-cost alternatives to platinum electrocatalysts are needed. Now, isolated, earth-abundant cobalt atoms dispersed over nitrogen-doped graphene are shown to efficiently electrolyse water to generate hydrogen.

The success of the proposed hydrogen economy — a system based on hydrogen as an energy carrier — will in part depend on the ability to generate molecular hydrogen (H2) in a clean, sustainable and cheap way. One established method for hydrogen generation is electrochemical water splitting, in which water is converted into molecular hydrogen by means of an electric current. Hydrogen production can be sustainable if the current is generated by renewable sources, such as solar energy. In contrast to such sources, which are intermittent, once electricity is converted into chemical energy in the form of hydrogen it can be stored until it is required. However, water electrolysis is energy intensive and requires expensive catalysts, typically containing platinum nanoparticles. Now, writing in Nature Communications, Dongliang Chen, James Tour and colleagues in China and the US report a highly active and stable catalyst consisting of isolated cobalt atoms on a nitrogen-doped graphene substrate1. This material is cheap and is produced by a scalable synthesis method, making it a promising candidate to replace platinum catalysts.

The hydrogen evolution reaction (HER), which occurs at the cathode, is kinetically sluggish in neutral or acidic electrolytes. Thus, the role of catalysts in water electrolysis is to lower the kinetic barrier for the reaction and in doing so decrease the overpotential required to drive the reaction. Platinum is the best HER catalyst, but its scarcity and high cost raise serious concerns about the long-term viability of platinum-based catalyst systems. Various cheaper, earth-abundant metal catalysts, such as transition metal oxides2, sulfides3, phosphides4 and carbides5 have been explored as potential replacements for platinum, but they are often inefficient because only a small proportion of the atoms located on the surface of the particles are catalytically active. To achieve the same total activity of platinum, a higher metal loading would be needed. However, this can cause aggregation of the metal atoms, which leads to a significant loss in the fraction of active metal atoms.

Chen, Tour and colleagues show that metal aggregation can be circumvented by combining transition metal atoms and nitrogen-doped graphene. Individual cobalt atoms disperse on the graphene substrate, instead of forming discrete nanoparticles (Fig. 1a), through coordination to the nitrogen atoms in the graphene. This synthesis method is both cheap and scalable, and the material can be produced as either a powder or a free-standing film, which enables easy integration of the catalytic layer into an electrochemical device. The electrocatalytic performance of the cobalt-based material is excellent. The intrinsic activity (called turnover frequency) per cobalt atom is similar or higher than that of most non-noble metal catalysts, the overpotentials are very low (Fig. 1b) and the catalyst is durable in both acidic and basic media. Moreover, owing to the atomic and uniform dispersion of cobalt throughout the catalyst, the data suggest that the utilization efficiency of cobalt may approach the ideal value of 100%, and the density of active sites per unit area is high. This result is indeed impressive considering that the cobalt content is low (0.57 at%), and it demonstrates that atomic dispersion is a powerful strategy for constructing highly active and stable catalysts.

Figure 1: Atomically dispersing cobalt atoms on nitrogen-doped graphene creates an effective catalyst for the electrolytic production of hydrogen from water.

Figure 1

a, High-angle annular dark-field scanning transmission electron micrograph showing isolated cobalt atoms (bright spots) on the nitrogen-doped graphene substrate. Scale bar, 0.5 nm. b, Linear sweep voltammetry (in 0.5 M H2SO4 at a scan rate of 2 mV s−1) demonstrates the low overpotential required to drive the hydrogen evolution reaction on the reported catalyst (Co-NG). The equivalent data for the metal-free nitrogen-doped graphene substrate (NG), the graphene-supported cobalt (Co-G) and a commercial carbon-supported platinum catalyst (Pt/C) are shown for comparison. The inset displays a close-up of the main plot. j, current density; E, potential; RHE, reversible hydrogen electrode. Figure adapted from ref. 1, NPG.

Single-atom catalysts (SACs)6, such as those demonstrated in this work, could also provide excellent model systems for the identification of catalytically active sites in other reactions. For example, nitrogen-doped carbon materials containing metals such as cobalt or iron are catalysts for the oxygen reduction reaction (ORR), which is important in metal–air batteries and fuel cells. At present, the typical structure of these catalysts is highly heterogeneous, comprising a mixture of large nanoparticles, small clusters and single atoms, which makes identifying the catalytically active site challenging. Therefore, the study of catalysts containing just one type of metal coordination environment simplifies the task greatly. Although in this work the researchers were not able to give a clear image of how the single cobalt atoms coordinate to the nitrogen atoms, a recent study7 made strides by unequivocally determining the exact structure of the active site (FeN4C12) in an iron-based SAC for ORR. Determining whether the single cobalt atoms on nitrogen-doped graphene adopt a similar structure would be valuable to understanding the catalytic properties of the material, and could aid the design of catalysts with better performance for multiple reactions.

Although this work provides a good approach for the production of efficient single-atom catalysts from earth-abundant materials there is still a long way to go to compete with noble metals. Future studies may focus on how to increase the metal content (to above 10 wt%) while maintaining the single-atom dispersion, or on doping single atoms of a second metal to generate synergy and modify the electrocatalytic properties.


  1. 1.

    et al. Nature Commun. 6, 8668 (2015).

  2. 2.

    et al. Science 334, 1256–1260 (2011).

  3. 3.

    et al. Science 335, 698–702 (2012).

  4. 4.

    , , & J. Am. Chem. Soc. 136, 7587–7590 (2014).

  5. 5.

    & Angew. Chem. Int. Ed. 51, 12703–12706 (2012).

  6. 6.

    et al. Acc. Chem. Res. 46, 1740–1748 (2013).

  7. 7.

    et al. Nature Mater. 14, 937–942 (2015).