Neuromorphic heat transport effects in a molecular junction

TL;DR

The paper addresses energy transport in nanoscale molecular junctions subjected to time-periodic temperature gradients $T_L(t)$ and $T_R(t)$, focusing on heat transport hysteresis as a memory phenomenon. Using nonequilibrium molecular dynamics with Langevin baths and stochastic thermodynamics, it reveals nonlinear hysteresis curves, including multi-pinched loops, whose geometry depends on chain length $N$, couplings $\gamma_L$, $\gamma_R$, and driving frequencies $\omega_L$, $\omega_R$. A calibrated reduced-order model—a 1D harmonic chain with terminal pinning—is shown to reproduce MD results in certain regimes by fitting the force constant $k$ via RMSE against the MD $J_{sys}$, enabling rapid exploration of parameter space. The results suggest molecular junctions could function as thermal memory or logic elements for neuromorphic thermal computing and guide future exploration of nonlinear and quantum effects.

Abstract

Understanding energy transport at the nanoscale is an open and fundamental challenge in the molecular sciences with direct implications for the design of new electronics, computing devices, and materials. While nanoscale energy transport under steady-state conditions has been studied extensively, there is much less known about energy transport under time-dependent driving forces, particularly in the far-from-equilibrium regime. In this work, we use nonequilibrium molecular dynamics simulations and stochastic thermodynamics to investigate energy transport in a well-studied nanoscale system, a molecular junction, subjected to a time-periodic temperature gradient. The primary observation is that molecular junctions can exhibit heat transport hysteresis, a phenomenon in which the heat flux through a system depends not only on the instantaneous value of a time-dependent temperature bias but also on the temporal history of that bias. The presented findings illustrate that molecular junctions can exhibit the specific memory effect, heat transport hysteresis, that is essential for the design of thermal neuromorphic computers. This work elucidates a potential pathway toward the realization of such devices.

Figures (12)

• Figure 1: Schematic diagram of an alkane molecular chain connecting two heat baths with time-dependent oscillating temperatures. The temperatures of the left and right baths, $T_\text{L}(t)$ and $T_\text{R}(t)$, are oscillatory as signified by the circular patterns representing each bath. The representative hysteresis curves in the bottom row illustrate the concept of hysteresis under thermal driving. The molecular simulation results presented in this paper often deviate from this simple hysteresis curve shape.
• Figure 2: System energy flux as a function of temperature difference for an ethane molecular junction. The parameters in the model are $N =2$ (ethane), $\gamma_\text{L} \approx 0.05$, $\gamma_\text{R} \approx 0.05$, $T^{(0)}_\text{L} = 300$, $T^{(0)}_\text{R} = 300$, $\Delta T_\text{L} = 0$, $\Delta T_\text{R} = 30$, $\omega_\text{L} =0$, $\omega_\text{R} = 5$. All parameters throughout the article are given in AKMA units.
• Figure 3: System energy flux as a function of temperature difference for an ethane molecular junction. The parameters in the model are $N =2$ (ethane), $\gamma_\text{L} =1$, $\gamma_\text{R} =1$, $T^{(0)}_\text{L} = 300$, $T^{(0)}_\text{R} = 600$, $\Delta T_\text{L} = 0$, $\Delta T_\text{R} = 200$, $\omega_\text{L} =0$, $\omega_\text{R} = 20$. The results are shown over two oscillation periods.
• Figure 4: System energy flux as a function of temperature difference for an ethane molecular junction. The parameters in the model are $N =2$ (ethane), $\gamma_\text{L} \approx 0.05$, $\gamma_\text{R} \approx 0.05$, $T^{(0)}_\text{L} = 200$, $T^{(0)}_\text{R} = 300$, $\Delta T_\text{L} = 0$, $\Delta T_\text{R} = 50$, $\omega_\text{L} =0$, $\omega_\text{R} = 5$.
• Figure 5: System energy flux as a function of temperature difference for (a) a septane molecular junction and (b) a methane molecular junction. The parameters in (a) are $N =7$, $\gamma_\text{L} \approx 0.05$, $\gamma_\text{R} \approx 0.05$, $T^{(0)}_\text{L} = 300$, $T^{(0)}_\text{R} = 300$, $\Delta T_\text{L} = 0$, $\Delta T_\text{R} = 30$, $\omega_\text{L} =0$, $\omega_\text{R} = 5$. The data in (a) was vertically shifted by a factor $\Delta J_\text{sys}/\gamma k_\text{B} T = 0.014$. The parameters in (b) are $N =1$, $\gamma_\text{L} =0.1$, $\gamma_\text{R} =0.1$, $T^{(0)}_\text{L} = 300$, $T^{(0)}_\text{R} = 200$, $\Delta T_\text{L} = 0$, $\Delta T_\text{R} = 100$, $\omega_\text{L} =0$, $\omega_\text{R} = 10$.