Important issues facing model-based approaches to tunneling transport in molecular junctions

We systematically examine popular models based on tunneling barriers or tight-binding pictures and demonstrate that, for a quantitative description at biases of interest, cubic expansions do not suffice

Ioan Bâldea


Scholarcy highlights

  • Introduction and backgroundThe roughly parabolic shape of the conductance G(V) qI/qV, or the related cubic dependence of the current on bias, was considered a prominent characteristic of transport via tunneling.This conclusion emerged from extensive studies on a variety of macroscopic thin film junctions of oxides, insulators, superconductors up to relatively large biases. A quick glance at I–V measurements in a variety of molecular junctions may convey the impression that this cubic dependence is satisfactory for such systems. As an illustration, we have chosen in Fig. 1a a measured I–V curve, for which fitting with a cubic polynomial looks accurate
  • The view based on such third-order Taylor expansions) was able to qualitatively describe a series of interesting aspects related to charge transport in molecular junctions
  • In order to demonstrate that the cubic polynomial approximation is insufficiently accurate to describe Vt, we have presented in Fig. 1a a raw I–V trace measured on a conducting probe atomic force microscope junction
  • With the manifest aim of providing experimental colleagues a comprehensive working framework enabling them to process and interpret measurements of transport by tunneling in molecular junctions, this paper have presented a detailed collection of analytical formulae, emphasizing on the fact that these formulae only hold if specific conditions of applicability are satisfied, which often impose severe restrictions on the model parameters
  • At biases of experimental interest, within all the models examined here, the description based on a third-order expansion I = I(V) of transport measurements in molecular junctions is insufficient for quantitative purposes
  • We have shown this by systematically analyzing the fifth-order expansions, which turned out to be reasonably accurate at least for biases up to the transition voltages

Need more features? Save interactive summary cards to your Scholarcy Library.