Schematic representation of the molecular junction under consideration. When an electron hops from the bridging site to the LUMO, a hole is left behind. Due to the interface Coulomb interaction, the electron on the LUMO and the hole on the bridging site are attracted, forming a bound exciton. This process renormalizes and enhances the tunneling matrix element between the bridging site and the LUMO, thereby affecting the current through the junction.
I-V curves for the molecular junction for different values of the interface Coulomb interaction, U 1 = 0, 0.4, …, 2.8 eV (see text for other numerical parameters). As U 1 increases, the current develops a peak near the LUMO resonance and NDR. (Inset) Blow-up of current at low bias, demonstrating an increase in the conductance with increasing U 1.
Current (blue) and effective coupling Γ (red) as a function of the bias voltage for the interface Coulomb interaction U 1 = 2.6 eV. The coupling constant is maximal at the LUMO resonance, followed by a peek in the current. In the inset the experimental data of Ref. 5 is shown, presenting the current (blue) and force constant (red). The dashed lines show the position of the maximum coupling, demonstrating the agreement between theory and experiment.
(a) The parts f L (ω) + f R (ω) (blue line) and ℜG r ω (purple line) composing the integrand I τ(ω) (Eq. (6) ), for a non-interacting example (see text). ℜG r ω changes sign as the molecular level is crossed. (b) I τ(ω) for different bias voltages V = 0, 0.1, 0.2, 0.3. As long as V < ε0, I τ(ω) is mostly positive, and reaches maximal area when V = ε0 (dashed line), after which it develops a negative part, resulting in a decrease in τ.
Peak-to-valley ratio (PVR) as a function of the molecule-electrode distance r, which affects both the interface interaction U 1 and the bare tunneling elements t (see text). The strong dependence of the PVR on r can be tested experimentally, and is in accord with the experimental findings. (Inset) I-V curves for different temperatures T = 60, 120, …, 300 K, demonstrating a very weak dependence of the current on the temperature. This weak dependence can also serve as a test for the origin of the NDR.
Power output vs. bias voltage of the molecular junction in the presence of a constant temperature gradient ΔT = 50 K, for U 1 = 2.6 eV (blue solid line) and U 1 = 0 eV (dashed red line).
Article metrics loading...
Full text loading...