Benjamin K. Tsai

Bidirectional reflectance distribution function of rough silicon wafers

The trend towards miniaturization of patterning features in integrated circuits (IC) has made traditional batch furnaces inadequate for many processes. Rapid thermal processing (RTP) of silicon wafers has become more popular in recent years for IC manufacturing. Light-pipe radiation thermometry is the method of choice for real-time temperature monitoring in RTP. However, the radiation environment can greatly affect the signal reaching the radiometer. The bidirectional reflectance distribution function (BRDF) of rough silicon wafers is needed for the prediction of the reflected radiation that reaches the radiometer and for reflective RTP furnace design. This paper presents the BRDF measurement results for several processing wafers in the wavelength range from 400 to 1100 nm with the spectral tri-function automated reference reflectometer (STARR) at the National Institute of Standards and Technology (NIST). The rms roughness of these samples ranges from 1 nm to 1 μm, as measured with an optical interferometric microscope. Correlations between the BRDF and surface parameters are obtained using different models by comparing theoretical predictions with experiments.

Modeling Radiative Properties of Silicon with Coatings and Comparison with Reflectance Measurements

Achieving high-accuracy temperature measurements in rapid thermal processing using radiation thermometry requires knowledge of the optical properties of silicon and related materials, such as silicon dioxide, silicon nitride, and polysilicon. However, available optical property models lack consistency and are not fully validated by experiments at the wavelength and temperature ranges critical to radiation thermometry. A critical survey is given of the existing optical models, with emphasis on the need for extrapolation and validation. Also described is an algorithm for calculating the radiative properties of lightly doped silicon with coatings. The effect of coatings covering one or both sides of a smooth silicon wafer is theoretically studied at room temperature, as well as at elevated temperatures. A spectrophotometer was used to measure the reflectance for selected samples in the wavelength region from (1.5 to 1 μm at room temperature. The measurements agree well with the predicted reflectance for bare silicon, a silicon wafer with a nitride coating, and wafers with an oxide coating of different thicknesses, whereas a larger deviation of as much as twice the measurement uncertainty is observed for a silicon wafer coated with polysilicon and oxide films.

A Monte Carlo model for predicting the effective emissivity of the silicon wafer in rapid thermal processing furnaces

Abstract Advances in microelectronics led to the development of rapid thermal processing (RTP). Accurate in situ temperature measurement and control are crucial for RTP furnaces to be largely accepted in the fabrication of semiconductor chips. This paper describes an effective emissivity model based on the Monte Carlo method to facilitate radiometric temperature measurements. The results showed that for non-diffuse wafers the “true” effective emissivity (defined in this paper) should be used, instead of the hemispherical effective emissivity, to correct thermometer readings. The geometric parameters and surface radiative properties can significantly influence the effective emissivity. The numerical aperture of the lightpipe radiation thermometer and the wafer-to-shield distance may be optimized to improve the measurement accuracy.

Radiative Calibration of Heat-Flux Sensors at NIST: Facilities and Techniques

We present an overview of the National Institute of Standards and Technology high temperature blackbodies, both in operation and in development, suitable for heat-flux sensor calibration. Typical results of calibrations using the transfer technique in the 25 mm Variable-Temperature Blackbody are presented to demonstrate the long-term repeatability of the calibration technique. A comparative study of the absolute and transfer calibrations of a Gardon gage in a spherical blackbody with a cooled enclosure surrounding the gage housing was conducted. Results of this study demonstrated the influence of convection associated with absolute calibration of sensors in a cooled enclosure. Plans for further development of the transfer technique to higher heat-flux levels and the associated technical issues are discussed.

High-heat-flux sensor calibration using black-body radiation

This paper deals with the radiative calibration aspects of high-heat-flux sensors using black-body radiation. In the last two years, several heat-flux sensors were calibrated up to 50 kW/m2 using a 25 mm diameter aperture variable-temperature black body and a reference room-temperature electrical-substitution radiometer. Tests on a typical Schmidt-Boelter heat-flux sensor showed long-term repeatability of calibration is within 0.6%. Plans for extending the present calibration capability to 100 kW/m2 are discussed.

Comparative Calibration of Heat Flux Sensors in Two Blackbody Facilities

This paper presents the results of heat flux sensor calibrations in two blackbody facilities: the 25 mm variable temperature blackbody (VTBB) primary facility and a recently developed 51 mm aperture spherical blackbody (SPBB) facility. Three Schmidt-Boelter gages and a Gardon gage were calibrated with reference to an electrical substitution radiometer in the VTBB. One of the Schmidt-Boelter gages thus calibrated was used as a reference standard to calibrate other gages in the SPBB. Comparison of the Schmidt-Boelter gages calibrations in the SPBB and the VTBB agreed within the measurement uncertainties. For the Gardon gage, the measured responsivity in the SPBB showed a gradual decrease with increasing distance from the aperture. When the gage was located close to the aperture, a distance less than the aperture radius, the responsivity in the SPBB agreed with VTBB measurements. At a distance of about three times the aperture radius, the responsivity showed a decrease of about 4 %. This is probably due to higher convection loss from the Gardon gage surface compared to the Schmidt-Boelter sensor.

Transfer Calibration Validation Tests on a Heat Flux Sensor in the 51 mm High-Temperature Blackbody

Facilities and techniques to characterize heat flux sensors are under development at the National Institute of Standards and Technology. As a part of this effort, a large aperture high-temperature blackbody was commissioned recently. The graphite tube blackbody, heated electrically, has a cavity diameter of 51 mm and can operate up to a maximum temperature of 2773 K. A closed-loop cooling system using a water-to-water heat exchanger cools electrodes and the outer reflecting shield. This paper describes the newly developed blackbody facility and the validation tests conducted using a reference standard Schmidt-Boelter heat flux sensor. The transfer calibration results obtained on the Schmidt-Boelter sensor agreed with the previous data within the experimental uncertainty limits.

Modeling of solvent evaporation effects for hot plate baking of photoresist

A heat transfer model for hotplate baking is combined with a mass transfer model for solvent diffusion to predict the major effects of photoresist prebaking for photolithography. Solvent diffusivity as a function of solvent concentration and temperature is determined experimentally. The results of the model are a complete time-temperature history of the wafer, final solvent distribution within the resist film, and final resist thickness.

The extension of the NIST BRDF scale from 1100 nm to 2500 nm

Measurements of bi-directional reflectance factor for diffuse reflectance from 1100 nm to 2500 nm using extended-range indium gallium arsenide (exInGaAs) detectors in the NIST Spectral Tri-function Automated Reference Reflectometer (STARR) facility are described. The determination of bi-directional reflectance factor with low uncertainties requires the exInGaAs radiometer to be characterized for low-noise performance, linearity and spatial uniformity. The instrument characterizations will be used to establish a total uncertainty budget for the reflectance factor. To independently check the bi-directional reflectance factors, measurements also were made in a separate facility in which the reflectance factor is determined using calibrated spectral irradiance and radiance standards. The total combined uncertainties for the diffuse reflectances range from < 1 % at 1100 nm to 2.5 % at 2500 nm. At NIST, these measurement capabilities will evolve into a calibration service for diffuse spectral reflectance in this wavelength region.

Effects of wafer emissivity on rapid thermal processing temperature measurement

Lightpipe radiation thermometers (LPRTs) are widely used to measure wafer temperatures in rapid thermal processing (RTP) tools. Using blackbody-calibrated LPRTs to infer the wafer temperature, it is necessary to build a model to predict the effective emissivity accounting for the wafer and chamber radiative properties as well as geometrical features of the chamber. The uncertainty associated with model-corrected temperatures can be investigated using test wafers instrumented with thin-film thermocouples (TFTCs) on which the LPRT target spot has been coated with films of different emissivity. A model of the wafer-chamber arrangement was used to investigate the effects of Pt (/spl epsiv//sub s/=0.25) and Au (/spl epsiv//sub s/=0.05) spots on the temperature distribution of the test wafers with the emissivity of 0.65 and 0.84. The effects of the shield reflectivity and the cool lightpipe (LP) tip on the wafer temperature were evaluated. A radiance analysis method was developed and a comparison of model-based predictions with experimental observations was made on a 200 mm wafer in the NIST RTP test bed. The temperature rises caused by the low-emissivity spot were predicted and the cooling effect of the LP tip was determined. The results of the study are important for developing the model-corrected temperature measurement and uncertainty estimates using LPRT in semiconductor thermal processes.