Graphene nanoribbons-narrow and straight-edged stripes of graphene, or single-layer graphite-are predicted to exhibit electronic properties that make them attractive for the fabrication of nanoscale electronic devices. In particular, although the two-dimensional parent material graphene exhibits semimetallic behaviour, quantum confinement and edge effects should render all graphene nanoribbons with widths smaller than 10 nm semiconducting. But exploring the potential of graphene nanoribbons is hampered by their limited availability: although they have been made using chemical, sonochemical and lithographic methods as well as through the unzipping of carbon nanotubes, the reliable production of graphene nanoribbons smaller than 10 nm with chemical precision remains a significant challenge. Here we report a simple method for the production of atomically precise graphene nanoribbons of different topologies and widths, which uses surface-assisted coupling of molecular precursors into linear polyphenylenes and their subsequent cyclodehydrogenation. The topology, width and edge periphery of the graphene nanoribbon products are defined by the structure of the precursor monomers, which can be designed to give access to a wide range of different graphene nanoribbons. We expect that our bottom-up approach to the atomically precise fabrication of graphene nanoribbons will finally enable detailed experimental investigations of the properties of this exciting class of materials. It should even provide a route to graphene nanoribbon structures with engineered chemical and electronic properties, including the theoretically predicted intraribbon quantum dots, superlattice structures and magnetic devices based on specific graphene nanoribbon edge states.

successfully applied to the growth of AGNRs 11-13 and related structures [14][15][16] . Here, we describe the successful bottom-up synthesis of ZGNRs, which are fabricated by the surface-assisted colligation and cyclodehydrogenation of specifically designed precursor monomers including carbon groups that yield atomically precise zigzag edges. Using scanning tunnelling spectroscopy we prove the existence of edge-localized states with large energy splittings. We expect that the availability of ZGNRs will finally allow the characterization of their predicted spin-related properties such as spin confinement 17 and filtering 18,19 , and ultimately add the spin degree of freedom to graphene-based circuitry.To explore the fundamental electronic and magnetic properties related to zigzag edges and to realize specific carbon nanostructures for the controlled manipulation of their spin states,ZGNRs with atomically precise edges are required. For GNRs with armchair edges, it was demonstrated that atomic precision can indeed be achieved by a bottom-up approach based on the surface-assisted polymerization and subsequent cyclodehydrogenation of specifically designed oligophenylene precursor monomers 11 . The on-surface synthesis has been applied by many groups to fabricate a number of different AGNR structures [11][12][13] , N-doped AGNRs 14,15 as well as AGNR heterostructures 15,16 . It is, however, not directly suited forZGNRs since polymerization of monomers via aryl-aryl coupling does not take place along the zigzag but along the armchair direction (Fig. 1a). In addition, dehydrogenative cyclization of phenyl subgroups is not sufficient to form pure zigzag edges, thus calling for a totally new chemical design. Thereby, additional carbon functions must be placed at the edges of the monomers to complete the tiling toolbox needed for the bottom-up fabrication of arbitrary GNR structures.Here, we report a bottom-up fabrication approach to ZGNRs. In our unique protocol, surfaceassisted polymerization and subsequent cyclization of suitably designed molecular precursors carrying the full structural information of the final ZGNR afford atomic precision with respect to ribbon width and edge morphology. The groundbreaking idea depends upon the choice of a unique U-shaped monomer as 1 shown in Fig. 1b. With two halogen functions for thermally induced aryl-aryl-coupling at the R 1 positions, it allows the polymerization toward a snake-like polymer. It is the beauty of this design that additional phenyl groups at the R 2 position fill the holes in the interior of the undulating polymer. The crucial precursor is monomer 1a which carries two additional methyl groups. In such a case, apart from the 3 polymerization and planarization, an oxidative ring closure including the methyl groups is expected which would then establish two new six-membered rings together with the zigzag edge structure. To our delight, this concept could indeed be synthetically realized under reaction monitoring and structure proof by scanning tunneling microscopy (S...

No abstract