Optical transmission of graphite and potassium graphite intercalation compounds

  • PDF / 637,697 Bytes
  • 6 Pages / 595.44 x 841.68 pts Page_size
  • 5 Downloads / 293 Views

DOWNLOAD

REPORT


I. INTRODUCTION The first optical study of graphite intercalation compounds (GIC's) 1 ' 2 was reported by Hennig in 1965.3 In this work transmission spectra through the c face of thin pristine and intercalated single-crystal graphite flakes were obtained. Spectra of the GIC's showed clearly the signature of a metallic free carrier response. The screened free carrier plasma frequency for each GlC was identified with a strong transmission maximum that was observed to fall in the range 1-3 eV. Since this initial paper, numerous optical reflectance studies have been carried out in graphite intercalation compounds and have endeavored to put Henriig's work on a more solid quantitative footing.4'5 However, due to experimental limitations, these reflectance studies have been carried out in a somewhat narrow energy range and Kramers-Kronig (K-K) analyses of the data have been performed using various high-energy data extensions in order to obtain the dielectric function e(0>«) =e,(0,6;) +/e 2 (0,«). To the best of our knowledge, there has been no effort, until the present work, to check via transmission data whether or not the previous K-K analyses have resulted in reliable values for the dielectric function, or to see if recent energy band models for GIC's are in agreement with transmission data. In this article we present transmission results for pristine highly oriented pyrolytic graphite (HOPG) and the prototype binary donor type1'2 GIC's KC 8 (stage 1) and KC 24 (stage 2). The stage index refers to the number of carbon layers stacked between periodically inserted intercalate layers.1'2 We will discuss the various forms of transmission formulas applicable to weak and strong absorption, and in the presence (or absence) of interference effects due to multiple internal reflections. The KC 24 data are analyzed in terms of a twodimensional (2-D) rigid band model Of Blinowski and Rigaux et al.6 However, this simple 2-D model is not applicable to KC 8 due to hybridization near EF .4 We 858

J. Mater. Res. 2 (6), Nov/Dec 1987

http://journals.cambridge.org

also show below how the main features in a general GIC transmission spectrum depend on the Fermi energy in the carbon w bands. II. EXPERIMENTAL DETAILS The samples used in our experiments were thin flakes cleaved from a thick slab (1 c m X l cm X0.5 mm) of HOPG using ordinary transparent tape (Scotch 3M). This brand was chosen because the tape readily dissolved in a suitable solvent (methylene chroride, 99 + % A.C.S. reagent). The HOPG flakes had a thickness of ~ 1000 A that rendered them gray in transmission to room light. Irregularly shaped semitransparent flakes were trimmed (still stuck to the tape) to rectangular shapes (—4x5 mm 2 ) using an abrasive wire saw. The tape was then dissolved, leaving an unsupported HOPG flake that was cleaned carefully in spectralgrade acetone and then mounted on a stainless steel sample holder. The thin flake was placed over a small hole (diam ~ 2 mm) on a thin (~0.1 mm) rectangular stainless steel backing sheet as shown in Fig. 1. A second