Heterogeneous Catalysts

There is direct formation of noble metal nanoparticles (<5 nm size) on CNT by high‐energy ball milling of a mixture of CNT and acetate precursor [48]. In fact, transition metal CUS on graphene can be achieved, for example, through the mechanochemical ball milling of graphene and phthalocyanine precursors [49]. Ball milling supplies the activation energy necessary to form bonds between the nitrogen atoms of phthalocyanine with dangling carbon atoms at the defects of graphene. This method is aggressive to the support, forming defects on the materials or decreasing the particle size. Accordingly, it is only applicable to carbon supports in powder form, which could be subsequently pelletized or structured macroscopically, if necessary.

Electrically conductive supports such as carbon are amenable to the deposition of metal nanoparticles by electrodeposition. This method allows the use of bulk metal, which is electrochemically stripped and subsequently electrochemically or electroless deposited on the carbon. It is a waste‐free method, comparing advantageously to other methods using metal precursors based on salts such as nitrates. A ligand is used, but this is recycled upon metal deposition. Generally, smaller particles are prepared by decreasing the concentration of the metal solution, the deposition time, or pH. The method must be optimized to prepare small particles of typically 7–20 nm [50]. Since the effects of loading and particle size are coupled, there is a trade‐off between achieving higher metal loadings and having smaller particle sizes.

This method is mainly used for the deposition of metals on semiconductor materials (see for examples Chapters 11, 31, and 37on the principles of photocatalysis). Carbon nitride is a carbon‐based semiconductor material. Pd nanoparticles have been deposited on carbon nitride, by photoassisted reduction of Pd(NO 3) 2using ultraviolet (UV) illumination [51]. This method led to uniform distribution of Pd nanoparticles with the size of 6 ± 0.7 nm, and exhibited higher stability in ORR than the commercial Pt on carbon catalyst. A variant of photodeposition is the photo‐Fenton process, which capitalizes on the homogeneous photoreaction between dissolved Fe 2+and H 2O 2. On CNTs, Co and Fe nanoparticles (2–5 nm) have been deposited under UV illumination through a photo‐Fenton process [52]. In this process, UV‐induced hydroxyl radicals ( •OH) oxidize the CNT surface in the presence of Fe or Co ions that are reduced and subsequently precipitated on the CNT surface in the form of nanoparticles.

The optimization of catalyst prepared on activated carbon materials by conventional wet chemistry methods has usually relied on a trial and error approach due to its complex surface chemistry. Hence, the degree of control on particle size distribution remains very modest. The recent advancements in new carbon materials (CNTs, graphene, nanodiamonds, carbon nitride) along with the advances in the nanotechnology synthesis and characterization techniques have enabled the “catalyst by design” approach as complemented by theoretical and computational techniques.

High level of control of catalyst size and distribution has been achieved on carbon materials, from single‐atom catalysts and clusters of few atoms to nanometric particles. This has been possible in part because the nanostructure and surface chemistry of new carbon materials are well defined and tunable. The structure can be chosen between 0D (nanodiamonds), 1D (CNT), 2D (graphene, carbon nitride), and 3D carbon materials (aerogels). There is a variety of carbons from pure sp 2(graphene) to pure sp 3(nanodiamond) hybridization, carbons doped with different types and amounts of heteroatoms (doped graphene, carbon nitride).

Although some of the new carbon materials have extremely high surface area, they lack microporosity, rendering the entire surface accessible for the metal precursor and catalytic reactions. The porosity can be also tuned for synthetic laminated carbon materials (carbon nitride, CTF, graphene derivatives). It is recognized that the support material is not innocent in catalytic reaction, modulating the activity of metal catalyst, absorbing reactants or intermediates of reaction. Thus, metals can be supported on carbon supports affording different degrees of metal–support interactions. The electronic charge transfer between the metal and carbon material can also be tuned because the carbon supports can vary from highly conductive (graphene) to semiconductors (carbon nitride). All these tunable parameters can affect the subsequent catalytic activity and stability. Therefore, tailored catalysts can be prepared to meet the requirements of each specific reaction.

The range of carbon materials to choose as catalyst support is also open to carbon materials synthesized at the laboratory, starting from molecular building blocks. This group includes graphene derivatives (graphyne and graphdiyne), hydrothermal carbonization of carbohydrate precursors, or pyrolysis of MOFs. The preparation of catalyst on these carbon materials is still in its infancy and brings a wide field to be explored. Heteroatoms, especially nitrogen, doped on carbon materials can act as a surrogate of ligands in homogeneous catalysts stabilizing (SACs and tuning their activity. Therefore, SACs stabilized on carbon materials benefit from the high turnover and selectivity of homogeneous and enzymatic catalysts. On the other side, they also benefit from high stability and robustness of heterogeneous catalysis. Traditional wet chemistry synthesis techniques are applicable with an enhanced level of control on the new carbon materials. Emerging synthesis techniques (e.g. colloidal nanotechnology, ALD) have been successfully applied for the preparation of a myriad of new carbon‐based catalysts. Therefore, the array of carbon materials and the advances in preparation techniques have enriched the “carbon toolbox” for the preparation of tailored catalysts.

1 1 Lam, E. and Luong, J.H.T. (2014). ACS Catal. 4: 3393–3410.

2 2 Cao, S., Monnier, J.R., Williams, C.T. et al. (2015). J. Catal. 326: 69–81.

3 3 Dreyer, D.R. and Bielawski, C.W. (2011). Chem. Sci. 2: 1233–1240.

4 4 Su, D.S., Wen, G., Wu, S. et al. (2017). Angew. Chem. Int. Ed. 56: 936–964.

5 5 Mistry, H., Behafarid, F., Reske, R. et al. (2016). ACS Catal. 6: 1075–1080.

6 6 Dreyer, D.R., Park, S., Bielawski, C.W., and Ruoff, R.S. (2010). Chem. Soc. Rev. 39: 228–240.

7 7 Tang, X.‐Z., Li, X., Cao, Z. et al. (2013). Carbon 59: 93–99.

8 8 Jeon, E.K., Seo, E., Lee, E. et al. (2013). Chem. Commun. 49: 3392–3394.

9 9 Su, C., Tandiana, R., Balapanuru, J. et al. (2015). J. Am. Chem. Soc. 137: 685–690.

10 10 Yin, H., Tang, H., Wang, D. et al. (2012). ACS Nano 6: 8288–8297.

11 11 Ferdinand, H., Alessandro, B., Giovanni, V. et al. (2017). Chem. Eur. J. 23: 15283–15288.

12 12 Chen, Y., Zhu, Q.‐L., Tsumori, N., and Xu, Q. (2015). J. Am. Chem. Soc. 137: 106–109.

13 13 Gao, Y., Chen, X., Zhang, J. et al. (2015). Adv. Mater. 27: 4688–4694.

14 14 García‐Bordejé, E., Víctor‐Román, S., Sanahuja‐Parejo, O. et al. (2018). Nanoscale 10: 3526–3539.

15 15 Bai, H., Li, C., Wang, X., and Shi, G. (2011). J. Phys. Chem. C 115: 5545–5551.

16 16 Cong, H.‐P., Ren, X.‐C., Wang, P., and Yu, S.‐H. (2012). ACS Nano 6: 2693–2703.

17 17 Wufeng, C., Sirong, L., Chunhua, C., and Lifeng, Y. (2011). Adv. Mater. 23: 5679–5683.

18 18 Fan, L., Li, B., Rooney, D.W. et al. (2015). Chem. Commun. 51: 1597–1600.