Determination of optical constants of thin films and multilayer stacks by use of concurrent reflectance, transmittance, and ellipsometric measurements.

Using measurements of reflectance, transmittance, and the ellipsometric parameter D, we have determined the thickness, refractive index, and the absorption coefficient of various thin films and thin-film stacks. (D, the relative phase between the p- and s-polarized components, is measured for both reflected and transmitted light.) These optical measurements are performed with a specially designed system at the fixed wavelength of lambda = 633 nm over the 10 degrees -75 degrees range of angles of incidence. The examined samples, prepared by means of sputtering on fused-silica substrates, consist of monolayers and trilayers of various materials of differing thickness and optical constants. These samples, which are representative of the media of rewritable phase-change optical disks, include a dielectric mixture of ZnS and SiO(2), an amorphous film of the Ge(2)Sb(2.3)Te(5) alloy, and an aluminum chromium alloy film. To avoid complications arising from reflection and transmission losses at the air-substrate interface, the samples are immersed in an index-matching fluid that eliminates the contributions of the substrate to reflected and transmitted light. A computer program estimates the unknown parameters of the film(s) by matching the experimental data to theoretically calculated values. Although our system can be used for measurements over a broad range of wavelengths, we describe only the results obtained at lambda = 633 nm.

Crystallization studies on phase-change optical recording media by use of a two-dimensional periodic mark array.

We present the results of crystallization studies in thin-film samples of amorphous and crystalline Ge(x)Sb(y)Te(z). The experiments, conducted at moderately elevated temperatures, are based on measurements of the first-order diffraction efficiency from a two-dimensional periodic array of recorded marks. When the samples are slowly heated above room temperature, changes in the efficiencies of various diffracted orders give information about the on-going crystallization process within the sample. Two different compositions of the GeSbTe alloy are used in these experiments. Measurements on Ge(2)Sb(2.3)Te(5) films show crystallization dominated by nucleation. For the Sb-rich eutectic composition Ge-(SbTe), crystallization is found to be dominated by growth from crystalline boundaries. We also show that crystalline marks written by relatively high-power laser pulses are different in their optical properties from the regions crystallized by slow heating of the sample to moderate temperatures.

High-performance readout and recording by a combination aperture.

A solid immersion lens combined with a conical dielectric tip exhibits good resolution and efficiency in reading and recording data marks on optical storage media. We demonstrate a combination aperture that produces ~200-nm full-width 1/e(2) spot size and achieves 50% optical efficiency in an edge-scan experiment. A comparison of recording with the combination aperture, with an unmodified solid immersion lens, and with a far-field system is made.

Variation on Zernike's phase-contrast microscope.

We describe the design, construction, and testing of a variant of Zernike's phase-contrast microscope. The sample is illuminated with a white-light source through an annular aperture, which is projected onto the entrance pupil of the objective lens. In the return path the light diffracted by the sample and appearing in the interior of the objective's aperture (i.e., the test beam) is separated from the light returning in the annular region near the rim of the objective (i.e., the reference beam). The separated beams are relatively phase shifted and then combined to create an interferogram of the sample's surface on a CCD camera. It is fairly straightforward to use this system as a conventional bright-field or dark-field microscope, but its most interesting application is as a Zernike phase-contrast microscope with adjustable amplitude ratio and phase shift between test and reference beams. The ability to continuously adjust the phase of the reference beam also enables quantitative measurement of the phase imparted by the sample to the incident beam.

Measurement of the relative optical phase between amorphous and crystalline regions of the phase-change media of optical recording.

We describe a method of measuring the relative optical phase on reflection between amorphous and crystalline regions of the phase-change media of optical data storage. With a red He-Ne laser (wavelength, 632.8 nm) the relative phases on two quadrilayer optical disk stacks were measured and found to be ~40 degrees . The results are in good agreement with the calculated values based on the known layer thicknesses and refractive indices of the stacks. For calibration purposes the height of a known step on an otherwise flat silicon substrate was measured with the same apparatus. The proposed method is fairly simple to set up, can measure both front-surface and through-substrate types of optical disk, and can be used with any laser that has long coherence length.

Static tester for characterization of phase-change, dye-polymer, and magneto-optical media for optical data storage.

We have designed and built a static tester around a commercially available polarized light microscope. This device employs two semiconductor laser diodes (at 643- and 680-nm wavelengths) for the purpose of recording small marks on various media for optical data storage and for the simultaneous monitoring of the recording process. We use one of the lasers in the single-pulse mode to write a mark on the sample and operate the other laser in the cw mode to monitor the recording process. The two laser beams are brought to coincident focus on the sample through the objective lens of the microscope. The reflected beams are sent through a polarizing beam splitter and thus divided into two branches, depending on whether they are p or s polarized. In each branch the beam is further divided into two according to the wavelength. The four beams thus produced are sent to four high-speed photodetectors, and the resulting signals are used to monitor the reflectance as well as the polarization state of the beam on reflection from the sample. We provide a comprehensive description of the tester's design and operating principles. We also report preliminary results of measurements of phase-change, dye-polymer, and magneto-optical samples, which are currently of interest in the areas of writable and rewritable optical data storage.

Leaky polarizing beam splitter with adjustable leak ratio for operation in the wavelength range of 440-690 nm.

We discuss the optomechanical design and fabrication of a novel wideband (440-690-nm), leaky polarizing beam splitter with an adjustable leak ratio. This beam splitter is an important component of a multiwavelength dynamic testbed that we have constructed for testing optical disks. The multilayer thin-film structure of the beam splitter is essentially a stacked pair of narrow-band dielectric reflectors that have been fine tuned for optimal performance. The characteristics of the fabricated device are in good agreement with our theoretical calculations.

Versatile polychromatic dynamic testbed for optical disks.

A dynamic testbed for the evaluation of optical disks has been designed and constructed. The system is achromatic within the wavelength range 440-690 nm, allowing any light source in this range to be utilized for read-write-erase experiments. In addition, the system accepts disks with substrate thicknesses ranging from 0 to 1.7 mm. The polarization handling capabilities of the testbed are such that, with the turn of a knob, one can generate either linearly polarized or circularly polarized light at the disk surface. This feature permits the testing of both magneto-optical and phase-change disks, in addition to compact disks and digital versatile disks, without any modifications to the system. A leaky polarizing beam splitter (LPBS) has been specially designed and built for this tester. The LPBS allows continuous adjustment of the ratio between p- and s-polarized components of the reflected beam that reach the detectors. This feature is especially useful for magneto-optical disks, where one can achieve an optimum signal-to-noise ratio by adjusting the relative amounts of the two components of polarization at the detection module. Focus-error detection is based on the astigmatic method, and the primary track-error detection scheme is the push-pull method, although other focusing and tracking schemes may also be implemented. The rf data signal and the focusing and tracking servo signals are all derived from the same detectors, thus allowing the optical power returning from the disk to be used in its entirety for these multiple purposes. The detection channel consists of two high-speed quad detectors mounted on the two arms of a differential detection module. By combining the various outputs of these detectors it is possible to generate the astigmatic focus-error signal, the push-pull track-error signal, the differential magneto-optical readout signal, the conventional sum signal for phase-change disk readout, and the differential edge-signal for mark-edge detection on various types of optical media.

Investigation of certain diffraction effects in an optical disk.

We report certain diffraction effects that are pertinent to the operation of double-layer optical recording media. For simulating cross-talk effects for double layers, the diffraction of light from the out-of-focus layer and the resulting distribution on the in-focus layer are studied by use of computer simulations. The findings are then verified qualitatively by direct measurements. We also describe a technique for analyzing (by computer simulation) the focus-error signal (FES), taking into account the cross talk between two layers, in systems that use the astigmatic method in conjunction with the double-layer disk. The results of our computer simulations of the FES give us a 10% cross-talk contribution to the original signal. The results of the FES evaluation are compared with those measured in an actual disk drive; good agreement between computation and measurement is obtained.

Retroreflecting ellipsometer for measuring the birefringence of optical disk substrates.

A retroreflecting ellipsometer has been constructed for measuring the birefringence of optical disk substrates. In contrast to conventional ellipsometers with two mechanical arms, this system has only one arm along which both the incident and reflected beams travel. This construction eliminates the mechanical limitations of conventional ellipsometers, thereby permitting normal incidence on the sample. In addition, the single arm is adjustable in two dimensions, with the polar incident angle, θ(inc), varying from 0° to 70°, and the azimuthal incident angle, Φ(inc), varying from 0° to 360°. The condition of normal incidence permits accurate measurement of in-plane birefringence. The adjustability of both θ(inc) and Φ(inc) is necessary for the measurement of possible tilts of the index ellipsoid, and also for the variation of birefringence through the substrate thickness. Measurement results showing the useful features of the equipment are presented. The optics of the hemispherical assembly used for retroreflection as well as for the elimination of undesirable refractions are also studied by use of the ZEMAX lens design program.

Measurement of birefringence for optical recording of disk substrates.

We experimentally investigate the birefringence of bare and coated substrates for magneto-optical recording, using ellipsometry at wavelengths of 632.8 and 780 nm. The polarization rotation and ellipticity of reflected or transmitted light are measured for different incident angles and for different orientations of the incident linear polarization. The measured data are then fitted by the MULTILAYER computer program, which solves the Maxwell equations for a plane wave propagating in a multilayer structure. This approach makes it possible to determine, with high accuracy, the orientations of the principal axes of the substrate and the corresponding refractive indices. The results show that one of the principal axes is always along the substrate's normal direction, but the orientations of the in-plane principal axes can be much different from the radial and track directions. A special feature of the ellipsometers that were used is that a glass hemisphere is placed in contact with the substrate to eliminate refraction of the incident beam. This enables a maximum propagation angle of 70° (with respect to the normal) in the substrate and hence increases the measurement sensitivity. Certain anomalies have been observed, which we believe are associated with the variation of birefringence properties along the thickness direction.

Ring-lens focusing and push-pull tracking scheme for optical disk systems.

An experimental comparison of the ring-lens and the astigmatic techniques of generating a focus-error signal (FES) in optical disk systems reveals that the ring lens generates a FES curve over two times steeper than that produced by the astigmat. Partly because of this large slope and partly because of its diffraction-limited behavior, the ring-lens scheme exhibits superior performance characteristics. In particular, the undesirable signal known as feedthrough (induced on the FES by track crossings during the seek operation) is substantially lower than that observed with the astigmatic method. The ring lens is easy to align and has reasonable tolerance for positioning errors.

Dielectric tensor characterization for magneto-optical recording media.

A new and comprehensive dielectric-tensor characterization method was used for the characterization of magneto-optical recording media. The effect of film thickness and bilayer thickness on the magneto-optic Kerr effect was studied for a series of Co-Pd samples. The composition dependencies of the dielectric tensor for both Co-Pt and Co-Pd superlattice samples were measured, and the enhancement of the magneto-optical Kerr effect for these samples was studied.

Measuring the wavelength dependence of magnetooptical Kerr (or Faraday) rotation and ellipticity: a technique.

A technique is proposed that allows accurate determination of the magnetooptic rotation and ellipticity angles. This method requires only a few standard optical elements and is expected to be effective over a relatively wide range of wavelengths.