Low frequency noise in AC biased metallic tunnel junctions

Nicolas Fontaine,Alexandre Dumont,Bertrand Reulet
DOI: https://doi.org/10.1109/ICNF57520.2023.10472762
2023-12-05
Abstract:We study the effect of an AC bias on the low frequency noise, notably $1/f^\gamma$ with $\gamma<2$, of metal-insulator-metal tunnel junctions at room temperature. The measurement is performed in the 6Hz-100kHz frequency range with an AC excitation above 1MHz. We observe that $1/f^\gamma$ noise is dominant across our measurements though the shape of the spectra varies. The effect of the DC excitation seems to be very different on the noise generated by the junction than that of the AC excitation, thus questioning the fact that the observed noise is due to resistance fluctuations that the bias only reveals.
Mesoscale and Nanoscale Physics
What problem does this paper attempt to address?
This paper aims to study the low - frequency noise characteristics of metal - insulator - metal (MIM) tunnel junctions under alternating - current bias (AC bias), especially the 1/fγ noise (where γ < 2), and measure them at room temperature. Specifically, the author hopes to experimentally verify the following issues: 1. **Sources of low - frequency noise**: The traditional view is that low - frequency noise (such as 1/f noise) is caused by resistance fluctuations, and these fluctuations are independent of the bias voltage. However, through experiments, the author found that direct - current bias (DC bias) and alternating - current bias have different effects on noise, which has raised doubts about this traditional view. 2. **The influence of AC bias on noise**: According to theoretical expectations, applying an AC bias should generate sideband noise on both sides of the fundamental frequency, and the 1/f noise in the low - frequency region should be suppressed. However, the experimental results show that at lower AC bias voltages, the noise spectrum hardly changes; while at higher AC bias voltages, the noise spectrum changes significantly. These results are inconsistent with the expectations based on the resistance fluctuation model. 3. **Differences in noise characteristics under DC and AC biases**: The experimental results show that as the bias voltage increases, the noise spectrum under DC bias gradually changes from a complex shape to a nearly pure power - law distribution (slightly less than 1/f). The noise under AC bias shows different frequency dependencies, and at higher bias voltages, the noise generated by AC bias increases significantly. ### Formula summary 1. **Johnson - Nyquist noise formula**: \[ S_V = 4 k_B T R \] where \( S_V \) is the power spectral density of voltage fluctuations, \( k_B \) is the Boltzmann constant, \( T \) is the temperature, and \( R \) is the resistance of the sample. 2. **Voltage fluctuations under DC bias**: \[ S_V(f) = S_R(f) I_{DC}^2 \] where \( S_R(f) \) is the power spectral density of resistance fluctuations, and \( I_{DC} \) is the direct - current. 3. **Voltage fluctuations under AC bias**: \[ S_V(f) = \frac{1}{2} \left[ S_R(|f - f_0|) + S_R(f + f_0) \right] I_{AC}^2 \] where \( f_0 \) is the frequency of the alternating - current, and \( I_{AC} \) is the amplitude of the alternating - current. 4. **Noise spectral density described by the Hooge parameter**: \[ S_V(f) = \frac{\alpha}{A f} V^2 \] where \( \alpha \) is the Hooge parameter, \( A \) is the area of the tunnel junction, and \( V \) is the bias voltage. ### Conclusion The experimental results show that the behavior of low - frequency noise is inconsistent with the traditional resistance fluctuation model. In particular, at higher bias voltages, the influence of AC bias on noise is more significant. These findings help to reveal the underlying mechanisms of 1/f noise in metal tunnel junctions and provide a new perspective for understanding flicker noise in other conductive materials.