Thermal measuring and testing – Temperature measurement – In spaced noncontact relationship to specimen
Reexamination Certificate
2001-04-19
2003-05-13
Fulton, Christopher W. (Department: 2859)
Thermal measuring and testing
Temperature measurement
In spaced noncontact relationship to specimen
C374S132000, C374S001000, C374S129000
Reexamination Certificate
active
06561694
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a method and apparatus for calibrating temperature measurements that are taken with at least one first radiation detector for measuring heat radiation emitted by at least one substrate.
Methods and apparatus of this type are known, for example, in connection with the production of semiconductor substrates in a reaction chamber. Preferred in such chambers are methods that are independent of emissivity and in which heat radiation coming from a first substrate is compared to the actual temperature of the substrate largely independent of the current emissivity. Such a method compensates differences in emissivity between different substrates. Used for achieving a measurement method that is independent of emissivity are, for example, the so-called “ripple technique”, which is described, for example in U.S. Pat. No. 5,490,728 and in DE 197 54 386 A (not previously published) from the same applicant, and the cavity principle, in which, for example, a mirrored chamber is closed on one side by an article to be measured, whereby an approximation of a cavity radiator with emissivity
1
is achieved. Another method that uses the cavity principle is described, for example, in DE 197 37 802 (not previously published) from the same applicant, which is incorporated into this application in order to avoid repetition. In the known methods wafers are used and a thermoelement (TC—ThermoCouple) is glued to the tops or bottoms thereof. Experience has demonstrated that the measurement deviation from TC to TC is very small as long as the TCs were produced from pairs of wires in the same series. The deviation then ranges from about 1 to 2° K. Based on this slight deviation, the temperature measurements can be adjusted to be independent of emissivity since in this case the issue in particular is slight deviation between the TCs. The uncertainty in the measurement of the absolute temperature is substantially greater, however, whereby the measurement uncertainty is at best in the region of 2-3 K and is more than 10 K-20 K in less favorable cases.
These measurement uncertainties are the result of different factors. Among these factors is that the thermoelectromotive force of the thermopair depends not only on the temperature but also on the alloy, but it is subject to certain deviation determined by production factors. Furthermore, provided between the TC and its amplifier are a plurality of electrical connections that also constitute the thermopair, so that unsymmetrical transitions produce additional thermoelectromotive force. In addition, the adhesive sites for the TCs have a different degree of absorption than the wafer surface surrounding the adhesive sites. The equilibrium temperature of the TCs is thus not determined solely by means of heat conducted between wafer and TC, which would be ideal, but rather the temperature of the TC is also influenced by the heat radiated from the lamp, which is why the temperature of the TCs frequently does not precisely equal the temperature of the substrate.
The measurement uncertainties thus arise primarily from erroneous sources that affect all TCs in the same manner as measurement errors. TCs are therefore measurement recorders that have great absolute measurement uncertainty, although they have a low deviation.
For achieving improved measurement accuracy, the TCs were calibrated in a TC calibrator, i.e., in an oven with very homogenous temperature distribution in the interior in terms of absolute temperature measurement. A plurality of uncalibrated TCs were placed in the oven together with a reference TC, whereby the reference TC itself was calibrated by a separate calibration service using a transfer pyrometer with respect to a primary reference. These multiple calibrations require different apparatus, they are very expensive, and, due to the many steps involved, there are numerous opportunities for errors to be introduced into the calibration process, which could again result in measurement uncertainties at the end of the process.
Furthermore, known from U.S. Pat. No. 5,265,957 is an apparatus and a method for calibrating a temperature sensor in which a wafer is provided with a plurality of calibration islands made of a reference material with a melting point in the range of 150° C. to 550° C. While such a wafer is being heated, the effective reflectivity of the wafer is measured by the temperature sensor and a first step change in an output signal of the temperature sensor is equated to a wafer temperature that equals the melting point of the reference material. Then temperature sensor calibration parameters are calculated. This principle is illustrated in FIG.
6
. The signal I of a temperature sensor (pyrometer) is recorded as a function of time t while the wafer is heated. For the reflectivity of the wafer to change when the reference material reaches phase transition, the reference material must be arranged close to the surface so that it is in the range of the penetration depth of the measurement wavelength. The aforementioned step change in the pyrometer signal occurs at the phase transition of the reference material, as shown in FIG.
6
. The method shown in U.S. Pat. No. 5,265,957 has substantial disadvantages. For instance, no unique pyrometer value can be allocated to melting point T
m1
due to the step change in the pyrometer signal, which results in a systematic measurement error &Dgr;I for the calibration method.
The object of the invention is therefore to provide a method and an apparatus of the type cited in the foregoing in which temperature measurements can be calibrated with great accuracy in a simple and cost-effective manner.
SUMMARY OF THE INVENTION
This object is inventively achieved using a method of the type described in the foregoing that has the following process steps: heating a reference substrate carrying at least one reference material with a known melting point to the melting point or over the melting point; measuring the thermal radiation of the reference substrate during the heating and/or during a cooling period following the heating; comparing a measurement plateau during the measurement process to the known melting point.
Heating the reference material located on the reference substrate raises the temperature of the reference substrate and the reference material until it reaches the melting point of the reference material. Once the melting point has been achieved, the temperature does not increase further until all of the reference material has converted to the liquid phase, that is, until the latent heat has been conveyed to the reference material. During cooling, this process reverses in a known manner. Since the melting point of the reference material is known precisely, it is possible to compare a measurement plateau measured during the heating and/or during a cooling period following the heating to the known melting point, thus achieving simple calibration of an absolute temperature measurement.
Advantageously the measurement plateau is determined during the heating and/or cooling of the reference substrate. However, the measurement plateau is preferably determined during the cooling period since the reference material in its melted state prior to complete solidification has particularly good contact to the reference substrate in terms of heat conductivity.
The method in accordance with the invention has substantial advantages compared to the method in U.S. Pat. No. 5,265,957 described in the foregoing. Since the method does not depend on changes in the emissivity of the reference substrate, it is possible to surround the reference material with a thick protective coating or to arrange the reference material in the interior of the reference substrate. It is not necessary to arrange it near the surface. The advantage of this is that the reference material cannot contaminate the processing chamber. This is the basic prerequisite for broad use of the reference substrate in semiconductor technology.
An additional advantage results from the occurrence of a measur
Hauf Markus
Lerch Wilfried
Becker R W
Fulton Christopher W.
R W Becker & Associates
STEAG RTP Systems GmbH
Verbitsky Gail
LandOfFree
Method and device for calibrating measurements of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and device for calibrating measurements of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for calibrating measurements of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3032542