Yi Su

Design of a scanning ellipsometer by synchronous rotation of the polarizer and analyzer

We have designed and constructed a new type of spectroscopic ellipsometer to study the optical properties of materials in the 3500-8000-Å wavelength range. In the system, the analyzer and polarizer are driven 10(4) steps/revolution by two stepping motors that have hollow shafts and rotate synchronously with a speed ratio of 2:1, i.e., A = 2P. Both the polarizer and analyzer are mounted directly on the shafts to avoid mechanical transmission and vibration problems entirely and make the system simple and reliable. An additional source polarizer was placed in the optical path to reduce the slight polarization effects of the light source. The light intensity finally received by the detector contained five components, one dc and four ac, with frequencies of ω(0), 2ω(0), 3ω(0), and 4ω(0), respectively. One can independently obtain the ellipsometric parameters of ψ and Δ as well as the optical constants by calculating any one of the two sets of ac signals, with a raw data self-consistency of better than 0.5%. The incident angle, aligned precisely by a laser beam, was continuously variable through a mechanical system with a computercontrolled resolution of 0.001° or a visual resolution of 0.005°. The system operations, including data acquisition and reduction, high-voltage control of the photomultiplier, incident angle, as well as wavelength setting and scanning, were fully and automatically controlled by a 386-based microcomputer. We self-calibrated the system by adjusting and setting precisely the initial azimuthal angles of the prisms. The results from the measured spectra of the complex refractive index for a gold-film sample are presented, and we show that the data obtained at three different incident angles of 65°, 70°, and 75° are remarkably consistent with one another. A comparison of the two results from the ellipsometry and reflectance measurements is given. The experimental skills and system error reduction are discussed in detail.

Improved rotating analyser-polarizer type of scanning ellipsometer

Abstract A new type of spectroscopic ellipsometer in which the analyser and polarizer rotate synchronously has been designed and constructed. For the system, the analyser and polarizer were driven 10 4 steps per revolution by two stepping motors with a speed ratio of 2:1. A second polarizer was placed in the optical path to eliminate the source polarization effect. The light intensity thus contains four a.c. components, having frequencies of ω 0 , 2ω 0 , 3ω 0 and 4ω 0 . The ellipsometric parameters ψ and Δ, as well as the optical constants, can be independently obtained by calculating any one of the two sets of a.c. signals. The system was self-calibrated, and was fully and automatically controlled by a 386-based microcomputer. The results of measured spectra of the complex refractive index for a 3000 A thick Au film sample are given. An equation for calculating directly the principal angle to obtain the highest precision of the optical constants is presented.

Ferromagnetic resonance study of Co/Pt multilayers

Abstract In this paper the results of ferromagnetic resonance study of sputtered Co/Pt multilayers, with a fixed Co layer thickness of 15 A, are presented. For these multilayers with small Pt layer thickness, in addition to a uniform resonance mode, a spin wave resonance mode is found to exist, which confirms the existence of an interlayer coupling between the neighboring Co layers. Simultaneously, the effective magnetization 4πMeff and the resonance linewidth ΔH are found to change correlatively with varying Pt layer thickness. This phenomenon can be explained as a result of an interplay between the interlayer coupling and the low-dimensional effect.

An anomalous magneto-optical Kerr effect in Pd-based multilayers

Abstract In this paper we present the calculated magneto-optical Kerr rotation spectra of Fe/Pd and Co/Pd multilayers based on two different assumptions. Firstly the spin-polarization of the Pd layers and its additional magneto-optical Kerr effect in Fe/Pd and Co/Pd multilayers are assumed negligibly small. Secondly the Pd layer is assumed to be spin-polarized. The calculated spectra with the latter assumption agree with the measured spectra by modifying the parameters. The Pd layers are suggested to induce an additional magneto-optical Kerr effect and thus enhance magneto-optical Kerr effect in these multilayers.

Magnetic and magneto‐optical properties of multilayers Fe‐Si/Pd

The magnetic and magneto‐optical properties of Fe‐Si/Pd multilayers prepared by rf sputtering were studied. For Fe‐Si/Pd multilayers with a fixed Fe‐Si layer thickness dm of 15 A and the Pd layer thickness dp increasing from 10.8 to 18 A, the saturation magnetization Ms and the Kerr rotation θk decrease sharply and reach a minimum. With dp further increasing, the multilayers increase. Ms is almost constant and θk decreases slightly when dp≳36 A. For Fe‐Si/Pd (54 A) multilayers the room temperature Ms decreases and θk increases, respectively, with increasing dm. This is caused by the spin‐polarization effect of Pd atoms near the interfaces.

New design to measure absolute spectral reflectivity

A new type of normal incidence scanning reflectometer is designed and constructed. In the system design, instead of using a reference reflective sample or the optical path adjustment methods as in traditional design, we use a fused quartz M-type prism to split the beam of incidence light into two beams in a total internal reflection configuration. The spot sizes, intensities, and spectral responses of these two light beams, therefore, are identical. One reference beam goes directly to the detector. Another sampling beam goes to the sample and is measured by the same detector after beam reflection. A metal disk having three holes is driven precisely by a stepping motor and is used to control the reference and reflection beams as well as the background signals, which are measured in sequence. In terms of the three signals, the absolute reflectivity of the sample at a certain wavelength can be determined immediately using the computer. The system is controlled automatically by the computer with a working wavelength range from 400 to 800 nm. The incidence angle is fixed and equals about 5 deg. The details of the system design, optical configuration, and error reduction are given and discussed. The measured spectral results of the reflectivity for a few testing samples are also given, and are shown to be in good agreement with those measured by other optical methods. The designed system, however, is simpler and can be used in many optical studies.

Study of optical and magneto-optical anisotropy of the Cd1−xMnxTe crystal

Abstract By using the magneto-optical and transmission-ellipsometric methods, giant anisotropy in both optical and magneto-optical properties was found for the single crystal Cd1−xMnxTe (x = 0.45). Two orthogonal axes E⊥ and E|, were in the (1 1 3) plane, and along the [ 4 7 1 ] and [2 1 1 ] directions, respectively. The crystal has two refractive indexes of n⊥ and n| and can be made a zero-field retarder. Not only the micro twin crystalline structure but also possibly the order replacement of Cd2+ by Mn2+ with induced lattice distortion through the whole body of the crystal along the [1 1 1] orientation can cause the optical and magneto-optical anisotropy. Both transmission-ellipsometric and magneto-optic Faraday rotation spectra located peaks at the same ∼2 eV EMn position, that was in the gap between the conduction and valence bands. In discussion, EMn was attributed to the optical transition of Te p → Mn d↓. A better understanding of Cd1−xMnxTe anisotropy will reveal further the behavior of Mn2+ that can account for the properties of the crystal.

New design to measure the absolute spectral reflectivity

A new type of the normal incidence scanning reflectometer has been designed and constructed. In the system design, instead of using a reference reflective sample or adjusting the optical path as did in the traditional design, we use a fixed M-type fused quartz prism to split the incidence light source beam into two ones by a total internal reflection configuration. Therefore, these two light beam have the same spot sizes, intensities and spectral response. One beam goes directly to the detector and is used as the incidence reference beam. The second sampling beam goes to the sample first, and then is reflected by the sample with the beam intensity measured finally by the same detector. A metal disk with three holes is driven precisely by a stepping motor and is used to control the reference beam, reflection beam and background signals to be measured in sequence. Afterwards, the absolute reflectivity of the sample at the certain wavelength can be determined immediately by the computer through calculating those three signals. The system is controlled automatically by the computer and worked in a wavelength range from 400 to 800 nm under a 5-degree nearly normal incidence angle condition. In the paper, the details of the system design, optical element configuration and error reduction are given and discussed. The measured reflective spectral results for testing samples are also given and shown in good agreement with those measured by other optical method. The system designed in this work, however, is simpler and more reliable to be used in many optical measurements of the sample.