On the determination of the free carrier relaxation time by transverse magneto‐plasma‐reflectivity measurements on Sn‐doped n‐type GaAs

Abstract

F r o m many investigations of the optical semiconductor properties due to free was found to be quite different from the carriers the optical relaxation time z dc relaxation time zdc /1 to 5/. This i s true in particular for ir plasma reflectivity effects /2 to 4/. In the manner used mostly the plasma frequency w and 5 be found from the reflectivity minimum with and without magnetic field and f r o m the line shape over a more or less broad part of the spectrum. However, because of the questionable validity of the Drude theory with a constant z value for including scattering of carriers, this method often yields erroneous 5 values. Therefore, the determination of z range. This condition can be fulfilled, for example, by measurements of the frequency o of the zero field reflectivity minimum and, if possible, of the resonance frequency w itself. In this note we report on this possibility using the transverse magneto-plasma reflectivity /6/.opt can P OPt opt should be reasonable only for a very small frequency OPt min P We have measured the reflectivity on n-type GaAs epitaxial layers doped with Sn in the infrared region with A, = (2 to 15) pm up to magnetic fields B = 2 . 8 T a t room temperature using a sensitive difference method /7/.In Fig. 1 some experimental spectra of the difference AR/R = (R(B) -R(O))/R(O) measured on GaAs samples at different magnetic fields are shown. The strong field dependence of the effect is clearly seen. However, the high-frequency zero point w does not depend ot? the magnetic field within the experiniental e r r o r . 0For comparison we have calculated AR/R applying the Drude theory / 6 / . Some theoretical AR/R spectra for indicating the damping influence a r e given in Fig. 1, too. In particular, the theoretical analysis has shown the very narrow neighborhood of w (plasma frequency) and w . In Fig. 2