The objective of the 1st year of our project was to demonstrate the feasibility of synthesis of conjugated polymers by surface-confined polymerization. We have focused on model building blocks which are structurally relevant to those that we intend on using for 2D polymerization, yet polymerize in 1D fashion, which simplifies the control of the reaction and the analysis of products.
      Three independent approaches were considered:
a) topological 1,4-addition polymerization
b) electrochemical polymerization
c) catalytic polymerization
      Bidentate monomers 1-4 have been used as models to study the surface-confined polymerization (Scheme 1). Our goal was to adopt the polymerization technique in our lab and to understand the intrinsic limitations of the methods in order to apply them for 2D polymerization.
1,4-Addition Polymerization (a). Before us, solid-confined polymerization of diacetylenes (1 and similar derivatives) have been studied by Aono et al. and De Feyter et al.[1] Although the general feasibility of the concept has been demonstrated, the detailed study reveal two main challenges, which limit the “topological” polymerization and lead to structural imperfections:  (i) poor self-assembly of the monomer and (ii) quenching of the polymerization transient by the surface (graphite). In addressing challenge (i) we have developed a method for “stimulated” self-assembly which leads to ordering of the monomer molecules along the STM fast scan direction (Fig. 1). The challenge (ii) is still a limitation of polymerization, as demonstrated by finite polymer length, and other surfaces, with shifted density of states are being currently explored.
Electropolymerization (b). The only precedent to ordered growth of conjugated polymers (in a monolayer) via electropolymeriation is the work of Sakaguchi.[2] Our first objective along this direction was simply to adopt the technique in order to be able using it for 2D polymers. However, in the course of studies we also have realized that the exact Sakaguchi's approach might not be feasible for 2D polymers. It seems that in his approach, the monomers (substituted with long alkyl chains) first form rather long oligomers in solution near the electrode, and only then get adsorbed on the electrode. This is confirmed by relatively high number of “cis” defects in the polymer structure of polythiophenes (eg, 1, R = H). For highly planar 2D polymers (e.g, from 5), even small oligomers would be totally insoluble and would crash out on the electrode in, possible, a  disordered state. Although further studies are clearly needed to achieve better control over the “surface confined” polymerization (as oppose to solution polymerization/adsorption), our initial studies showed that change of the solvent can result in formation of highly linear 1D chains of polythiophenes.
Catalytic Ullmann polymerization (c). The only relevant studies of Ullmann polymerization in on flat metal surfaces (Cu) was performed by Weiss et al for diiodobenzene 3.[3] Their work failed to show formation of linear poly(p-phenylene) (PPP), instead an unidentified “pre-polymer” structure was suggested. We believe that the reason for this negative result was very low temperature (77K) used in this study as well as unsuitable symmetry of the surface (trigonal Cu(111)). Although the initial breaking of C-I bonds occurs well below room temperature, an increased temperature is required to achieve high degree of polymerization and ordering of the polymer lines on the surface. We found that performing the reaction at elevated temperature (200°C), under Ultra High Vacuum (UHV) conditions transforms di-iodinated aromatic molecules into lines of conjugated polymers (Fig. 2). The most successful experimental runs were performed on Cu(110), exploiting the anisotropy of this surface, while Cu(001) and Cu(111) were much less effective. The method appears to be rather general and the polymer lines can obtained from different monomers (3b,c, 4). 
      We also established that the reaction byproduct (copper iodide) at high coverage creates separate phases on the surface with the characteristic 2×2 reconstruction characteristic of I on Cu(110) (not shown). This phase segregation effectively “stops” the growths of the polymer and complete coverage of the surface cannot be achieved under these conditions. In future work, we intend to overcome this obstacle by using Ni surface, which would yield volatile (under UHV) NiI2 compound, that can be removed by heating.
      Further characterization by Infrared Spectroscopy, X-Ray Photoelectron Spectroscopy and Ultraviolet Photoelectron Spectroscopy are under way. In the same time in this coming year we will start exploring multidentate monomers (such as 5) to tackle the challenges related to 2D polymerization.

[1] (a) Okawa, Y.; Aono, M. Nature 2001, 409, 683; (b) Grim, P. C. M.; De Feyter, S.; et al. Angew. Chem. Int. Ed. 1997, 36, 2601. [2] Sakaguchi, H.; et al. Science 2005, 310, 1002. [3] McCarty, G.S.; Weiss, P.S. J. Amer. Chem. Soc. 2004, 126, 16772