"The Origin of 1/f Noise, Maybe"

Dr. Paul Norton, U.S. Army Night Vision Laboratory

May 6th (Friday), 3:00pm
Engineering Science Building (ESB), Rm 2001

The origin of 1/f noise (sometimes referred to as flicker or pink noise) in semiconductor devices has not been convincingly explained since it was first reported by Schokly in 1918. This paper presents a possible explanation namely that the noise arises due to turbulence in the flow of the electron (or hole) fluid/gas in the sample. Turbulence in fluid or gas flow arises naturally from the convective forces imposed by a fluid¹s viscosity. Historically, transport in semiconductor materials has been modelled using linear equations of drift and diffusion that do not exhibit turbulence. Transport of gases and fluids in the discipline of hydrodynamics also uses the equations of drift and diffusion, but in addition includes the Navier-Stokes equation to describe flows that are turbulent. The Navier-Stokes equation naturally produces turbulence as the Reynold¹s number exceeds a certain range. For a uniform flow the critical Reynolds number is large several thousand. However, for narrow jets impinging upon a reservoir, the Reynolds number may be more than two orders of magnitude smaller. Scattering of electron flow by phonons, impurities, and other carriers may precipitate turbulent flow to arise at low values of the Reynolds number in semiconductors. In this paper we report on Navier-Stokes simulations of the spectral properties of turbulent flows showing that they are consistent with a 1/f spectrum. We also report on a series of measurements of photodiodes as a function of illumination intensity to observe how the 1/f properties vary with current so that in the future we can compare that variation to our simulations.

About Dr. Paul Norton:

Dr. Norton has worked in the field of infrared detectors for 49 years, receiving his Ph.D from Professor Henry Levinstein in 1970 at Syracuse University.

His work experience includes:
• Bell Telephone Laboratories at Allentown, PA and Murray Hill, NJ
• Honeywell Research Center in Minnesota
• Santa Barbara Research Center (Hughes Aircraft), now Raytheon Vision Systems
• Army Night Vision Laboratory at Fort Belvoir, Virginia

He has worked on PbS, PbSe, InSb, Ge:Hg, Si:As, PbSnTe, and HgCdTe detectors in his career. One of his device designs, a HgCdTe phototransistor was used for production of first-generation Army FLIRs as well as a number of strategic and civilian space missions. His current interest is in future Army night vision technologies and in particular defect behaviors that impact their performance. For several years now he has been focused on 1/f noise.

Dr. Norton has authored a number of review papers on infrared focal plane arrays and the chapter on photodetectors in the Optics Encyclopedia for the Optical Society of America.

He is a co-chair of SPIE’s conference on Infrared Technology and Applications.

Hosted by: Professor John Bowers