Direct measurement of electrical conductance through a self-assembled molecular layer F. Song 1,2 , J. W. Wells 1,3 , K. Handrup 1 , Z. S. Li 4 , S. N. Bao 2 , K. Schulte 5† , M. Ahola-Tuomi 6 , L. C. Mayor 5 , J. C. Swarbrick 5† , E. W. Perkins 1,5 , L. Gammelgaard 7 and Ph. Hofmann 1† * The self-assembly of organic molecules on surfaces is a promising approach for the development of nanoelectronic devices 1,2 . Although a variety of strategies have been used to establish stable links between molecules 2–11 , little is known about the elec- trical conductance of these links. Extended electronic states, a prerequisite for good conductance, have been observed for mol- ecules adsorbed on metal surfaces 12–16 . However, direct conduc- tance measurements through a single layer of molecules are only possible if the molecules are adsorbed on a poorly conduct- ing substrate. Here we use a nanoscale four-point probe 17 to measure the conductivity of a self-assembled layer of cobalt phthalocyanine on a silver-terminated silicon surface as a func- tion of thickness. For low thicknesses, the cobalt phthalocyanine molecules lie flat on the substrate, and their main effect is to reduce the conductivity of the substrate. At higher thicknesses, the cobalt phthalocyanine molecules stand up to form stacks and begin to conduct. These results connect the electronic struc- ture and orientation of molecular monolayer and few-layer systems to their transport properties, and should aid in the rational design of future devices. Using a good metal as the substrate when studying the electric con- ductivity of molecular layers on surfaces is problematic because charge will flow through the metal rather than the molecules. A promising sub- strate for conductance measurements is the silver-terminated silicon surface (Ag/Si(111)–( p 3  p 3); hereafter referred to as Ag/Si), on which the ordered growth of molecular layers can be achieved 5,18–20 and where the substrate consists of a semimetallic surface on a semicon- ducting bulk. This approach has been chosen in the present paper. Figure 1a shows the measured conductance of the self-assembled layer of cobalt phthalocyanine (CoPc) thickness on the Ag/Si surface. To achieve surface sensitivity, the conductance was measured with a TiW-coated four-point probe with a contact spacing of only 500 nm (refs. 17, 21). Also presented are a model of the geometric structure of the film (Fig. 1c) and an image of the type of probe used in the conductance measurements (Fig. 1a, inset). Two regions of different behaviour can be distinguished. Initially the conductance drops below that of the clean surface (shown in pink). Once the thickness is of the order of one monolayer, the con- ductance increases again and is eventually even higher than for the clean surface. We interpret this behaviour as the sum of the conduc- tance through two channels: one through the Ag/Si layer and the other as direct conductance through the molecules. The filled blue circle indicates a situation in which the surface is covered by one layer of flat-lying molecules, defining one monolayer, prepared by evaporating the CoPc molecules on an Ag/Si substrate held at elevated temperature (see preparation details). To facilitate a comparison with the main dataset it is necess- ary to assign a film thickness to this measurement; here we estimate that one monolayer is equal to 0.3 nm. The conductance for one monolayer prepared in this way agrees well with the main data series for which CoPc was deposited at room temperature, and cor- responds to the minimum in the conductance. There are two factors contributing to the lower conductance when CoPc is adsorbed on Ag/Si. The first is the introduction of disorder, which leads to a higher scattering rate for the carriers. The second, which we propose to be more significant, is related to an electron transfer from the surface to the CoPc molecules. Figure 2a shows X-ray photoelectron spectroscopy data from the cobalt 2p core level, for increasing thicknesses of CoPc on Ag/Si. For very small thicknesses, a low binding energy component dominates, whereas at higher thickness, the bulk component dominates. As a compari- son, the cobalt 2p core level of lithium-doped (that is electron- doped) bulk CoPc is shown in Fig. 2b. Again, two contributions to each peak can be readily identified. For highly doped CoPc, the lower binding energy component dominates whereas for undoped and weakly doped samples, the higher (bulk) component dominates. Thus, by comparing these measurements with those of CoPc on Ag/Si, we propose that the first monolayer of CoPc on Ag/Si is strongly electron-doped. Also from Fig. 2a, it is clear that the lower binding energy com- ponent does not vanish for thicknesses greater than one monolayer. Rather, the higher binding energy component appears in addition. This indicates that the interfacial monolayer of CoPc does not undergo any re-organization. Thus, for thicker films (in which the molecules form in the bulk a geometry), we understand that we still have a strongly electron-doped and flat-lying first monolayer, as shown schematically in Fig. 1c. We later show that the initially confined electron-doping of the first layer is then able to propagate freely through the entire film, leading to increased conductivity. The photoemission measurements in Fig. 2c show the evolution of the valence band with increasing CoPc deposition. For very thin films (,1 monolayer), there is an increase of intensity close to the Fermi level, but no signature of a peak derived from the highest occupied molecular orbital (HOMO). This, combined with the observed lower conductivity in this thickness range, points towards a strong and loca- lized bonding of the first-layer molecules to the substrate. For thicker films (.1 monolayer), the HOMO begins to appear in the spectra, and from its position and work function measurements (not shown), the 1 Institute for Storage Ring Facilities (ISA) and Interdisciplinary Nanoscience Center (iNANO), University of Aarhus, 8000 Aarhus C, Denmark, 2 Department of Physics, Zhejiang University, Hangzhou, 310027 China, 3 Norwegian University of Science and Technology (NTNU), Trondheim, Norway, 4 Institute for Storage Ring Facilities (ISA), University of Aarhus, 8000 Aarhus C, Denmark, 5 School of Physics and Astronomy, University of Nottingham, Nottingham NG7 2RD, UK, 6 Department of Physics and Astronomy, University of Turku, 20014 Turku, Finland, 7 Capres A/S, 2800 Kgs. Lyngby, Denmark; † Present address: MAX-lab, Lund University, P.O. Box 118, 22100 Lund, Sweden (K.S.), European Synchrotron Radiation Facility, B.P. 220, 38043 Grenoble cedex 9, France (J.C.S.), Surface Science Research Centre, University of Liverpool, Liverpool L69 3BX, UK (Ph.H.). *e-mail: philip@phys.au.dk LETTERS PUBLISHED ONLINE: 19 APRIL 2009 | DOI: 10.1038/NNANO.2009.82 NATURE NANOTECHNOLOGY | VOL 4 | JUNE 2009 | www.nature.com/naturenanotechnology 373 © 2009 Macmillan Publishers Limited. All rights reserved.