ICMCTF1999 Session C4: Ellipsometry: Measurement and Modeling

Monday, April 12, 1999 1:30 PM in Sunset Room

Monday Afternoon

Time Period MoA Sessions | Abstract Timeline | Topic C Sessions | Time Periods | Topics | ICMCTF1999 Schedule

Start Invited? Item
1:30 PM C4-1 Some Recent Developments in Optical Coatings
A. Macleod (Thin Film Center, Inc.)
Since the immediate purpose of optical coatings is to modify the reflectance and/or transmittance of an optical surface, they are indispensable features of virtually all optical systems. However, for many years optical coatings tended to be used only when absolutely necessary. The discouraging fact that the coating operation is often the final one in the manufacture of a component enhanced the impact of uncertainties in coating yield. Yields have improved enormously in recent years enabling applications well outside the traditional instrumental ones. There is even a large market for multilayer optical coatings in applications that are purely aesthetic. Current uses for optical coatings range from color shifting pigments for inks and paints, through wavelength division multiplexing filters, through anti-glare filters for displays of all kinds, through chirped reflectors for ultrafast pulses, to detectors for medical and biological purposes. This talk will examine some recent developments in the design, production and application of optical coatings.
2:10 PM C4-3 Two-Modulator Generalized Ellipsometry: Applications to Anisotropic Materials
G.E. Jellison (Oak Ridge National Laboratory)
Generally, ellipsometry measures the ratio of the complex reflection coefficients for light reflected from a sample surface. Because the phase of this ratio is also measured, ellipsometry is ideal for measuring thicknesses of very thin films. However, the interpretation of most ellipsometry measurements assumes that the sample is isotropic, that there are no cross-polarization effects, and that the sample does not de-polarize the incident light beam. For many of the applications for which ellipsometry has been used to date, these assumptions are valid, but recently there has been increased interest in using optical probes to examine anisotropic samples where these assumptions are not valid. Anisotropic samples may include crystalline anisotropic bulk materials or thin films grown using heteroepitaxy, diffraction gratings, or semiconductor integrated circuits at various stages of the fabrication process. Two-Modulator Generalized Ellipsometry (2-MGE) is a technique that capable of measuring all of the complex reflection coefficient ratios with a single measurement. Using spectroscopic 2-MGE, we have been able to obtain very accurate values of the optical functions of various anisotropic materials, including rutile (TiO2), zinc oxide (ZnO), bismuth tri-iodide (BiI3) and many tetragonal rare-earth phosphates from 850 to 240 nm (1.46 to 5.16 eV). Cross-polarization effects can be easily examined from one-dimensional diffraction structures (such as diffraction gratings) and two-dimensional structures such as some integrated circuits. The 2-MGE has also been useful for thin film measurements, where depolarization and interfacial cross-polarization effects are easily seen.
2:50 PM C4-5 A Web Accessible Ellipsometry Solving Engine Using the Variably Damped Least Squares Algorithm Assisted by Measured Data Point and Derivative Preselection for Improvement of Solutions
F.K. Urban, III, D. Barton, J. Pontillo (Florida International University)
A web accessible ellipsometry solving engine has been developed which uses a modification of the popular variably damped least squares (VDLS or Levenberg-Marquardt) solution algorithm. Although this algorithm is sensitive to local minima in the solution surface, especially in the case of solving for more than a few unknowns, it is widely used. One of its features is that it can readily use many measured data which provides for a kind of "signal averaging" that generally results in the algorithm achieving a solution (though normally with significant fiddling with initial estimates and user-imposed bounds on values of parameters to be solved). An underappreciated drawback to this approach is that no distinction is made between best and worst values in the dataset beyond considering the variability in obtaining each measured value. While this does reduce the impact of hard-to-measure points, it does not take into account the sensitivity of each point to desired parameters and the condition of the resulting system of equations used in the solver. The new adaptation of the VDLS does both these and allows selection of psi or delta as an objective function for each parameter to be computed as well. Results comparing the improvements obtained using these adaptations will be presented.
3:30 PM C4-7 Influence of Matrix Material on the Optical Properties of Plasma-Deposited Metal/Dielectric Nanocomposite Films for Nonlinear Optical Applications.
D. Dalacu, A.P. Brown, J.E. Klemberg-Sapieha, M.R. Wertheimer, L. Martinu, S.I. Najafi (Ecole Polytechnique, Canada); M.P. Andrews (McGill University, Canada)
Metal/dielectric nanocomposite films are currently receiving considerable attention as a potential material for use in optical devices due to their high optical nonlinearity present in the vicinity of the surface plasmon resonance (SPR). The optical properties, both linear and nonlinear, are dependent on the microstructural characteristics of the embedded metal nanoparticles (size, shape and concentration) and the composition of the surrounding matrix material. In the present work we extend our previous study of gold/flouropolymer nanocomposites to include SiOxNy and TiO2 matrices deposited by plasma-enhanced chemical vapour deposition (PECVD). This allows us to perform a systematic investigation of the influence of the matrix index of refraction on the SPR for materials ranging from plasma polymerized fluorocarbon (PPFC) (n =1.4), to SiOxNy (continuous change from 1.45 for SiO2 to 1.90 for Si3N4), and to TiO2 (n ~2.3). The optical constants of the films were determined using variable angle spectroscopic ellipsometry (VASE), and modeled using the Maxwell-Garnett (M-G) effective medium theory (EMT). The use of spectroellipsometric measurements was identified as a useful non-invasive technique capable of complete microstructural characterization of the nanocomposite films. By comparing the particle size extracted from the spectroellipsometric measurements with that determined by TEM observations and with results from XPS analysis, we were able to elucidate the influence of the gold/matrix interfacial region on the SPR band. The nonlinear optical (NLO) response is discussed in relation to the nanocomposite film microstructure.
3:50 PM C4-8 In-situ Ellipsometry of Sputtered ZnO Films
N.K. Zayer (University of Kent, United Kingdom); R. Greef (Southampton University, United Kingdom); C.N. Pannell (University of Kent, United Kingdom)
Thin orientated polycrystalline films of zinc oxide grown by RF reactive sputtering on various substrates are being investigated for use as high frequency acoustic transducers for (e.g.) fibre-optic modulators. Although the films need not be monocrystalline for this application, it is important that the films are reproducible and of high quality, having a small degree of surface roughness and as dense as possible, with crystallites well oriented with the c-axis perpendicular to the substrate. The deposition process we have used consisted of RF sputtering using a pure zinc target and mixtures of argon and oxygen. Process variables known to strongly influence the morphology of the films include the partial pressures of the gases, the RF power density on the target, target-substrate spacing and substrate temperature. We demonstrate the use of in-situ ellipsometry to monitor the formation of the films, using simple optical models to chart the rate of growth and their optical properties. Our early experiments resulted in films that were duplex in nature and of poor piezoelectric activity, having a dense layer in contact with the substrate, and a less dense upper layer. The two layers grew at measurably different rates. Further experiments enabled us to identify substrate temperature as an important variable in controlling the nature of the films. By changing it from 300 °C to 200 °C we obtained reproducibly high quality smooth films having high piezoelectric activity, with a uniform refractive index of 1.94. This compares with an index of 1.99 for bulk ZnO crystals. Despite the fact that the index remained constant during growth, the rate of growth was not quite linear, increasing slowly with time according to a weakly quadratic law. This can be interpreted as a sticking coefficient for ZnO which increases with film thickness.
4:10 PM C4-9 Modelisation in Spectroscopic Ellipsometry
J.-L. Stehle (SOPRA, France)
Spectroscopic Ellipsometry (SE) is a non destructive Technique which measures the complex reflectance of polarized light at glanzing incidence on the surface of samples, versus wavelength. Using modeling and regression algorithm, the user can determine the main physical parameters of the sample: layers thicknesses, refractive indices dispersion laws, interfaces, roughness, gradients, anisotropy, The choice of the model becomes as critical as the measurement quality. There are many models which can be applied to solid state physics, during this presentation, only the refractive indices and the sample parameters modeling will be addressed. Examples on real samples will demonstrate the critical choice of the modeling and its advantages and limitations. Golden Rules of regression applied in SE will be presented. Finally the need for accurate information about the sample structure coming from other techniques as XTEM or SIMS will be emphasized and the new combination of SE with grazing X Ray Reflectance, which gives complementary measurement on thin films on flat surfaces, will be assessed. In conclusion: a SE is a very powerful technique when the modeling is correctly used. Other characterization techniques may be needed also to determine the best model for SE.
Time Period MoA Sessions | Abstract Timeline | Topic C Sessions | Time Periods | Topics | ICMCTF1999 Schedule