@article{38a92def100340f8b421e9d21a68ced1,
title = "1/f noise in metal films: The role of the substrate",
abstract = "We have examined the temperature dependence of voltage ({"}1/f{"}) noise in Cu and Ag films on quartz and sapphire substrates. Our data suggest that two types of voltage noise occur simultaneously in metal films. One type of noise depends on the substrate and is weakly temperature-dependent; the remaining noise is strongly temperature-dependent and independent of the substrate.",
author = "P. Dutta and Eberhard, {J. W.} and Horn, {P. M.}",
note = "Funding Information: EBERHARD AND HORN \[1\]have recently reported measurements of the temperature dependence of voltage noise in silver and copper films on sapphire substrates. They have found that the temperature dependence is in both cases strikingly different from the prediction of the thermal fluctuation model of Voss and Clarke \[21;in fact at low temperatures the voltage noise drops orders of magnitude below the prediction. However, thermal fluctuations certainly do exist and should cause voltage noise (although the shape of its frequency spectrum is far less certain). Thus if the properties reported in \[1\] are to be taken to mean that an as yet unidentified source is responsible for the observed noise, one must first explain how temperature fluctuations are so strongly suppressed near the experimental frequency of 20 Hz. If, on the other hand, the observed noise is stifi interpreted as temperature fluctuations, one is left with a similar problem: identifying the mechanism responsible for the strong distortion of the predicted temperature dependence. The primary feature that distinguishes our experimental system from the existing thermal fluctuation models is the thermal contact to a substrate. While the total power of the thermal fluctuations (integrated over all frequencies) is determined according to thermodynamics by the nature and size of the sample itself, the substrate may affect the spectrum of the fluctuations and thus change the noise magnitude at any experimentally observable frequency. To investigate this possibility * Work supported by N.S.F. Grant DMR75-14360. We have also benefited from support of the Materials Research Laboratory by the National Science Foun-dation.",
year = "1978",
month = sep,
doi = "10.1016/0038-1098(78)91578-8",
language = "English (US)",
volume = "27",
pages = "1389--1391",
journal = "Solid State Communications",
issn = "0038-1098",
publisher = "Elsevier Limited",
number = "12",
}