1/f noise from levels in a linear or planar array

Publisher Website
Google Scholar
Abstract
This contribution expands on the analysis of an earlier study with the same title. There it was shown that charge transfer across a double layer to and from a monoenergetic set of surface states would yield 1/f noise if the fluctuations are strong enough for nonlinearity. It is proposed that the location of uncharged energy levels or sites fluctuates and can induce such nonlinearity. The uncharged species can randomly move far‐from or close‐to a charged species. If it is close to a charged species, the rate of charge transfer to it is low. Fluctuations in position thus result in significant fluctuations of charge transfer rates. With simplifications, the 1/f noise associated with ionosorption/desorption of O2 on a semiconductor is calculated as an example of this double‐layer noise. The computed noise spectrum of carrier density fluctuations follows the 1/f law over the frequency span predicted in the earlier study. A distribution of surface state energies leads to a substantial increase in the frequency span. In the example, noise from four discrete trapping energies (time constants) when summed yields 1/f noise over 6 decades. The example describes carrier density fluctuations. It is suggested that mobility fluctuations may dominate in many cases, especially metals, because there will be strong carrier scattering by bulk (and possibly surface) double layers. The temperature dependence of double layer 1/f noise is briefly discussed. The results for grain boundary noise by Madenach and Werner are shown to be consistent with the double‐layer model, as is the temperature variation as discussed by Dutta and Horn [Rev. Mod. Phys. 5 3, 497 (1971)].