Image of phonon spectrum in the 1/f noise of semiconductors

Abstract

features noise peaks which develop at some specific temperatures. We compared this noise structure with either phonon density of states or Raman spectrum of each topological insulator (TI), respectively. In (BiSb)2Te3, the comparison revealed that the noise peaks track the van Hove singularities in the phonon density of states. It resulted that bulk atomic oscillators are responsible for the noise peaks. The most intense noise peak observed in (BiSb)2Te3 at 50 K is attributed to the thermal motion of the Bi atoms. Other less intense noise peaks are assigned to either a single phonon mode or multi-phonon combinations. We found that thermal motions of Bi and Te 2 atoms in different symmetry directions are involved in most of the phonon combinations, which stand for the signature of the lattice anharmonicity. The noise increase observed in both (Bi,Sb)2Te3 and BiSbTeSe1.6 above a specific temperature threshold is attributed to the anharmonicity-induced strengthening of the carrier-phonon coupling. In the case of BiSbTeSe1.6, we show that all noise singularities are mirrored in the Raman spectrum of a structurally close TI (BiSbTeSe2) in the whole temperature range. This indicates that although transport can be at the surface or in the bulk or both of them, the carrier-phonon interaction is the only source of 1/f fluctuations in TIs. Inherently, these results imply that microscopic origin of 1/f noise in solid is in the perpetual thermal motion of the atoms.For almost a century[1], 1/f noise is puzzling the scientific community with its omnipresence in solid and solidstate devices[2-7]. In the last decade, it was observed [8][9][10][11][12][13][14][15][16][17][18][19] in topological insulators (TI), which are quantum materials with metallic conductivity at the surface, while their bulk is insulating [20][21][22][23][24][25][26][27]. These properties are due to the presence of surface electronic states with a Dirac cone energy dispersion located in the bulk band gap [20][21][22][23][24][25][26][27]. These surface states are protected by inversion symmetry to backscattering, a property which opens fascinating possibilities for TIs applications in spintronics and quantum topological computing. However, the presence of the 1/f noise in the surface conduction is a factor which can lead to certain technology roadblocks in applications such as quantum computing, where it acts as source of decoherence [28,29]. Therefore, finding its origin is of great practical interest. Hossain et al. [9] suggested that the noise dominated by the surface states in TIs could be very low because the time-reversal symmetry topologically protects the surface states from backscattering. However, in Bi 2 Se 3 samples wherein the carrier transport was dominated by the surface, they found a high 1/f noise parameter (α ≈ 0.2), therefore the noise would be due to a "mixed volume-surface transport regime" [9]. Consequently, to reduce the noise level, it would be necessary to protect the surface conduction from the bulk influences. Nevertheless, in str...

An alternative form of Hooge's relation for 1/f noise in semiconductor materials

“…Hence, the intermittent g-r process generating 1/f noise is also phonon-induced. Indeed, there is ample empirical evidence that electronphonon scattering is the dominate mechanism generating 1/f noise in semiconductors [13][14][15]. As was already mentioned in a previous paper [19], this suggests that the intermissions (= off-states) originate from phonons due to an unknown mechanism.…”

“…Based on empirical findings, Hooge, Kleinpenning and Vandamme [13] concluded that 1/f noise is a bulk phenomenon; they favor phonon scattering as an origin of 1/f noise. This perspective is also supported by Mihaila [14] who claims that 1/f noise comes from the perpetual equilibrium atomic motion. Musha, Gabor and Minoru [15] investigated the scattering of laser light in quartz and found 1/f fluctuations in the number of phonons.…”