Optical Indices of Tin-Doped Indium Oxide and Tungsten Oxide Electrochromic Coatings

  • PDF / 986,429 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 35 Downloads / 195 Views

DOWNLOAD

REPORT


homogeneous structural model found that a frequency dependent relaxation time in the Drude model was required for good agreement with their radiometric data. The optical properties of a commercial grade of ITO are carefully measured by both radiometry and ellipsometry. Several plausible models of the underlying optical response and the layer structure are constructed. The models are examined by fitting the optical data to each. Finally, the data for the conducting layer is used to construct a model in which the properties of a W0 3 top layer are unknown. The properties of the W0 3 are then determined in turn. METHODS The optical properties are measured using both ellipsometric and radiometric techniques. A variable-angle spectroscopic ellipsometer by J.A Woollam Co. is used over the extended visible range (300-1000 nm), Transmittance and reflectance measurements are made over the solar spectrum from 3002500 nm. All measurements were made at normal incidence using either a Perkin-Elmer Lambda 9 or Lambda 19 spectrophotometer. For these highly specular samples, the detectors could be exposed directly, rather than together with an integrating sphere. This detector arrangement provides maximum accuracy and also permits the backsurface to be roughened, which would otherwise contribute to the integrating sphere signal. Reflectance measurements were made with relative reflectance attachment with an angle of incidence of 8 degrees. A front-surface Al mirror calibrated by NIST was used as a reference. As a further check, measurements were made at 670 nm using a feedback-stabilized laser diode with a short coherence length and a large-area silicon detector. The error for this system is typically 0.1%. Some of the surface features of the structural model were verified by atomic force microscopy (AFM) using a Park Scientific Instruments Autoprobe. Additional information on compositional variations was obtained at Donnelly Corporation by etching the ITO film. Film thickness was measured in some cases using a Veeco Dektak II surface profilometer. Ellipsometric and radiometric data were fitted together weighting both data types according to their standard deviations. The standard deviations of the reflectance and transmittance measurements were not directly measured, but had to be assigned by comparison measurements and were taken as 0.1% over the whole spectral range from 300-2500nm Backside reflections were generally accounted for and included as fit parameters. In all cases we considered the material to be represented by a superposition of a single Lorentz oscillator and a Drude free-electron component: _

2

0j2

)__o_+7

W

(-02

2-

2

2

-0a)

r r2a)) 2

PP damping factors, 2e_______= where y and 13are

W

=2

2

Ne e 2 3comc

552

=p n2 -2

with the carrier concentration N,, the elementary charge e, the vacuum permittivity &oand mn the effective electron mass in the conduction band. Lattice vibrations in the far infrared have been observed, but they are too distant to contribute significantly [4]. RESULTS Ellipsometric