AVS1996 Session EM-TuA: In-situ Characterization of Materials and Processes
Tuesday, October 15, 1996 2:00 PM in Room 204A
Tuesday Afternoon
Time Period TuA Sessions | Abstract Timeline | Topic EM Sessions | Time Periods | Topics | AVS1996 Schedule
Start | Invited? | Item |
---|---|---|
2:00 PM | Invited |
EM-TuA-1 In Situ Characterization of Materials and Processes
N. Dietz (North Carolina State University) =20Understanding surface chemistry under steady-state growth involving organo-metallic chemical precursor molecules is essential for optimizing growth processes. Surface-sensitive optical real-time sensor techniques are very well suited for this task since their application is not limited to a high vacuum environment. In addition to the optical signatures, information about surface coverage of reactions species and their fragments are crucial input parameter in the identification and modeling of chemical reaction kinetics and their pathways. In this contribution we describe the combined application of p-polarized reflectance spectroscopy with in-situ mass spectroscopy (MS) for the real-time monitoring of deposition processes. We demonstrate low temperature growth of epitaxial GaP/Ga\sub x\In\sub 1-x\P/GaP heterostructures on Si(100) as well as GaN heterostructures embedded in GaP epilayers by pulse chemical beam epitaxy. Under the conditions discussed here, single wavelength PR follows the growth process with sub-monolayer resolution during the sequential precursor exposure of the surface that causes periodic alterations in composition and thickness of a surface reaction layer (SRL), the effect of which is monitored by PR as a periodic fine structure. This fine structure is superimposed on interference oscillations, resulting from back reflection at the substrate-layer interface with increasing layer thickness. The fine structure undergoes an amplitude modulation that has the same period as the interference oscillation but are shifted if the SRL is optically absorbing. The extrema in the fine structure modulation are characterized by turning points at which the response to the first precursor exposure changes direction. The relative positions of these turning point with respect to the extrema in the interference oscillations are linked to the absorbivity in the ultra-thin SRL. The kinetics of the heteroepitaxial growth process is assessed on the basis of PR and laser light scattering (LLS) and link to mass spectroscopic results to allow an assessment of surface reactions path ways during steady-state growth conditions. |
2:40 PM |
EM-TuA-3 Extended Virtual-Interface Approaches for Determining Compositions from Real-Time Reflectance and Ellipsometric Spectra
D. Aspnes (North Carolina State University); W. Gilmore III, C. Lee (North Carolina A&T State University) For real-time compositional control one must determine compositional fluctuations of the most recently deposited material, which requires that the dielectric function \epsilon\\sub o\ of an arbitrarily thin layer just beneath the surface be determined from kinetic reflectometric or ellipsometric data. Fresnel analysis cannot be used because its feed-forward nature makes it intrinsically unstable. The virtual-interface (V-I) approach eliminates these instabilities, but the information contained in the value and thickness derivative of the reflectometric or ellipsometric data is not adequate to obtain \epsilon\\sub o\ without making some assumptions or approximations. To overcome this limitation without sacrificing stability we investigate hybrid extensions of the V-I approach, where enough parameters are fed forward to yield an exact solution. We develop a formalism for determining the sensitivity of the calculated value of \epsilon\\sub o\ to the other parameters, allowing the least critical to be identified and thereby keeping the risk of instability low. Numerical calculations for AlGaAs show that ellipsometric data are particularly insensitive to the s-polarized virtual reflectance, which explains the accuracy of the virtual-substrate approximation (VSA) in semiconductor heteroepitaxy. The results will be useful for the analysis of reflectometric data and for ellipsometric measurements on systems where substantially different dielectric responses are involved. The procedure is general and can be applied to any combination of materials. |
|
3:00 PM |
EM-TuA-4 In-situ Optical Flux Monitor to Probe Chemistry of Molecular Beam Epitaxy
A. Ott (National Institute of Standards & Technology and University of Colorado, Boulder); S. Casey (University of Colorado, Boulder); S. Leone (National Institute of Standards & Technology) Single-photon ionization time-of-flight mass spectrometry (SPI-TOFMS) is applied to molecular beam epitaxy (MBE) to study the chemistry of semiconductor growth in ultra-high vacuum (UHV). In this technique, a pulsed beam of photons ionizes gaseous molecular species in front of an epitaxially growing substrate. The ions are measured in real time by time-of-flight mass spectrometry. The experimental configuration allows for simultaneous monitoring of mass signals and reflection high-energy electron diffraction (RHEED). The laser position can be adjusted so that the photons intersect only scattering molecules from the surface for desorption kinetics and surface chemical reactions or the photons can intersect only incident fluxes to track pressure bursts and contaminants from the source ovens. SPI-TOFMS is sensitive to small changes in fluxes and is therefore an excellent monitor of slight changes in sticking coefficients and incorporation kinetics.The method is applied to the homoepitaxy of GaAs(100) with simultaneous RHEED to analyze the mechanisms of growth. During growth using incident fluxes of Ga and As\sub 4\, the incorporation of As\sub 2\ and As\sub 4\ is measured. When the incident Ga flux is varied, the incorporation of arsenic indicates whether the growth rate is limited by the incident flux of arsenic or gallium. When the substrate temperature is increased, the measured arsenic incorporation decreases. Over this temperature range, the probe is sensitive to scattering Ga atoms from the surface that reflect less than 0.1 ML changes in the Ga sticking probability and account for the observed decrease in arsenic incorporation. |
|
3:20 PM |
EM-TuA-5 Optimization of an ECR Etch Process using Full Wafer CCD Interferometry
S. Pendharkar, D. Resnick, W. Dauksher (Motorola); I. Tepermeister, W. Conner (Low Entropy Systems) Full wafer CCD interferometry is a relatively new technique that can be used for in-situ monitoring of plasma etching. A CCD camera is used to monitor the light intensity reflected from distinct regions on a wafer. Plasma emission is used as the incident light, and changes in the optical emission intensity can also be monitored from an off-wafer region. For dielectric films, endpoint is observed by the loss of an interference signal. For metal films, an endpoint is indicated by a change in the reflected light intensity from a particular region on the wafer. A threshold algorithm is then used to calculate endpoint time, and Fast Fourier Transforms (FFT) are used to calculate etch rates. Since data from the entire wafer is recorded, both selectivity and etch uniformity can also be readily determined. In this paper, we describe the use of a Low Entropy Systems' 1000-IS full wafer interferometer to optimize an ECR etch process.A Plasma-Therm 770 ECR system was used to develop an etch process for tantalum silicon nitride (TaSiN), an absorber material for X-ray masks. Good selectivity to both a hard mask and an etch stop layer is necessary for defining sub-250 nm features. In addition, good etch uniformity is critical for avoiding long overetch times, which may cause pattern distortion on the mask. The microwave power, collimating magnet current and gas flow were optimized using the LES 1000-IS. Etch rate uniformity under the optimized conditions was 8% (3 \sigma\) over a 4 inch substrate, and structures as small as 100 nm have been fabricated with this process. |
|
3:40 PM |
EM-TuA-6 In Situ Analysis of Si(100) Surface Damage Induced by Low-energy Rare-gas Ion Bombardment using X-ray Photoelectron Spectroscopy
M. Ishii, Y. Hirose, T. Sato, T. Ohwaki, Y. Taga (Toyota Central Research and Development Laboratories, Inc., Japan) The damage layer of a Si(100) surface induced by low-energy (300-600 eV) rare-gas (Ne, Ar, or Xe) ion bombardment was in situ analyzed by x-ray photoelectron spectroscopy (XPS). The thickness of an amorphous Si (a-Si) layer was estimated quantitatively by a curve-fitting analysis of the XPS Si2p spectrum from the bombarded Si surface.\super 1\ We also observed the bombarded Si with cross-sectional transmission electron microscopy (TEM) to confirm the quantitativity for the thickness measured with the XPS analysis. The XPS analysis showed that, irrespective of the ionic species, an a-Si layer with a thickness of 1-3 nm was formed and saturated by the dose of 10\super 15\ cm\super -2\. It also showed that the thickness increased linearly with the ion energy and Ne ions formed the thickest a-Si layer. The saturated thickness measured with the XPS analysis has been confirmed to be valid by TEM observation. Formation process of the a-Si layer can be explained in terms of the Poisson process. As a result, the size of an a-Si cluster formed by single ion bombardment was estimated. The cluster, which is supposed to be a cylinder, has a radius of 0.2-0.4 nm and a height of 1-3 nm for each ionic species and energy. In situ XPS analysis for the Si surface bombarded with ion beams revealed the growth process of the a-Si layer in addition to the dependence of the thickness on the ionic species, energy, and dose. 1) Z. H. Lu et al., Appl. Phys. Lett. 65 (1994) 552. |
|
4:00 PM |
EM-TuA-7 Ion-Scattering Spectroscopy during InGaAs Molecular Beam Epitaxy: Reduction of Sputtering using Glancing-Angle Ar Ions
J. Labanda, S. Barnett (Northwestern University) Ion scattering spectroscopy (ISS) has been proposed as a method for in situ monitoring of surface composition during molecular beam epitaxy (MBE). In this talk, we show how alteration of GaAs and InAs growth surfaces due to ion sputtering during ISS can be essentially eliminated. Ar ions, with energy E = 1-3 keV with impingement angles \phi\ = 3-15\super o\ from the plane, were used for ISS measurement of InAs(001) and GaAs(001) surfaces. Ar was used because of the better mass resolution than He. Sputtering during MBE growth was observed by changes in reflection high-energy electron diffraction (RHEED) oscillation periods. MBE growth rates and the ion sputtering rates were found to subtract algebraically during simultaneous growth and ion bombardment. At \phi\ = 15\super o\, sputtering rates from 0.122-0.007 ML/s were observed when E was reduced from 3 to 1 keV. At E = 3 keV, the sputtering rate was reduced from 0.122 ML/s to 0.013 ML/s when \phi\ was reduced from 15\super o\ to 3\super o\. No change in RHEED oscillation period was observed at E = 1 keV and \phi\ \<= \ 9\super o\. As a further test, ISS measurements of 1-ML-thick InAs films on GaAs(001) were made using 1 keV Ar ions at \phi\ = 3\super o\ and a 35\super o\ scattering angle. No change in the In peak was observed after >20 minutes of continuous ISS scans corresponding to a dose of ~1.3 x 10\super 16\ ions-cm\super -2\ (estimated yield < 0.005 atoms/ion ). In coverages <0.1 ML were readily detected. Indium segregation during subsequent GaAs on InAs monolayers was observed. In addition to minimizing sputtering, the glancing ion incidence and small scattering angle is compatible with typical MBE systems. |
|
4:20 PM |
EM-TuA-8 In Situ Determinations of the Role and Concentrations of the Chemical Species Present in a Diamond Deposition Plasma
R. Blumenthal (Auburn University) Pulsed supersonic, plasma sampling mass spectrometry has been previously demonstrated to have a uniform high sensitivity to the broad range of chemical species present in a diamond deposition plasma. In this work, this technique has been used to determine the concentrations of each gas-phase chemical species present during diamond film growth as the deposition parameters (substrate temperature, feed-gas composition, and microwave power) are systematically varied. Correlations are determined between the measured concentrations of individual species and the quality of the resulting film, as determined through post-situ film analysis using a combination of Raman spectroscopy, AFM and electron microscopy. Correlations between the concentrations of individual chemical species and specific film properties are used to unravel the individual chemical role of each species within the overall diamond deposition process. |
|
4:40 PM |
EM-TuA-9 Design of a Scanning Tunneling Microscope for In Situ Topographic and Spectroscopic Measurements within a Commercial Molecular Beam Epitaxy Machine
C. Ventrice, Jr., V. LaBella, L. Schowalter (Rensselaer Polytechnic Institute) A scanning tunneling microscope (STM) that performs STM, scanning tunneling spectroscopy (STS), and ballistic electron emission microscopy (BEEM) measurements on 2 inch wafers has been designed and constructed. The STM is incorporated into the Si prep chamber of our cryo-pumped Fisons V90H Si/III-V molecular beam epitaxy (MBE) machine. The STM design uses two commercial Burleigh inchworms: one for performing STM measurements and a second for making a front contact to allow BEEM measurements. The substrate holder for the V90H system is designed to handle wafers up to 6 inches in diameter. Therefore, a custom 6 inch diameter holder has been constructed which supports two 2 inch holders: one for performing reflection high energy electron diffraction (RHEED) measurements and a second which allows transfer of the wafer to the STM. Although the chamber vibrations generated by the cryopump and its compressor are quite severe, atomic-resolution images have been obtained. The enhanced stability of our design is attributed to our unique support system of the inchworm and wafer which are both manufactured from machinable ceramic, the use of 16 inch long Be-Cu springs for vibration isolation, and the use of silicone O-rings for vibration damping. |