Optics: measuring and testing – By light interference – Spectroscopy
Reexamination Certificate
2000-02-25
2001-11-20
Font, Frank G. (Department: 2877)
Optics: measuring and testing
By light interference
Spectroscopy
C372S032000
Reexamination Certificate
active
06320663
ABSTRACT:
BACKGROUND OF THE INVENTION
Specrometers are well known devices for measuring the intensity of light in a laser beam as a function of wavelengths. Spectrometers used to measure the spectrum of the excimer laser can be divided into two main categories diffraction grating based spectrometers and Fabri-Perot etalon based spectromters.
FIG. 1
shows the features of a prior art etalon spectrometer used for measurement of wavelength and bandwidth of a laser beam
16
. The beam is diffused by diffuser
2
so that rays propagating in a very large number of angles illuminate etalon
4
.
FIG. 1
shows a single ray
20
being reflected many times within the etalon gap between surfaces
8
A and
8
B which are coated to reflect about 90% of the light. Spectral components which are transmitted through the etalon are focused by lens
14
onto photo diode array
12
. Photo diode array
12
registers a fringe pattern
15
which can be read using electronic data acquisition board
18
. The transmission or reflection of light incident on an etalon such as that depicted is well understood and depends on the design of the etalon, particularly the reflectance of the two reflecting surfaces.
A particularly important use of etalon spectrometers is to measure the bandwidth of line narrowed excimer lasers such as the line narrowed KrF or ArF excimer lasers. These lasers are used, for example, as light sources for deep-UV microlithography.
A description of a KrF laser is provided in U.S. Pat. No. 5,991,324 which is incorporated herein by reference.
There are two spectral characteristics of these lasers which are very important for microlithography applications. These are the spectral bandwidth of the laser measured at 50 percent of the peak intensity, called its full width at half maximum bandwidth (abbreviated &Dgr;&lgr;
FWHM
), and the spectral bandwidth, which contains 95% of laser energy called the 95% integral bandwidth (abbreviated &Dgr;&lgr;
95
%). It is very important that the laser is always operating within specifications during microlithography chip manufacturing because spectral broadening would cause blurring of the integrated circuit features being printed on silicon wafers which will result in yield problems. Therefore, it is very important to provide continuous monitoring capabilities for the laser spectrum.
The prior art etalon spectrometer is capable of accurately measuring &Dgr;&lgr;
FWHM
values, and is currently used for this purpose in production microlithography lasers, such as manufactured by CYMER, Inc. (San Diego, Calif.). However, prior art etalon spectrometers are not very suitable for accurately measuring &Dgr;&lgr;
95
% values. Typical production quality KrF excimer lasers should have a &Dgr;&lgr;
FWHM
of about 0.6 pm and &Dgr;&lgr;
95
% of about 2 pm, if operating properly.
FIG. 2
shows the calculated so called “slit function” spectrum of a typical prior art etalon having a free spectral range (FSR) of 5 pm and a coefficient of finesse (finesse) of 38. (The terms FSR and finesse are defined and explained in a variety of optic texts such as OPTICS by Eugene Hecht/Alfred Zajae published by Addison-Wesley, Reading, Mass.) The slit function spectrum of
FIG. 2
can be derived from one of the peaks of fringe pattern
15
. The calculation graphed in
FIG. 2
assumes that the light illuminating the etalon is monochomatic (i.e., an infinitely narrow bandwidth). If such an etalon is used to measure the bandwidth of a laser beam, the slit function bandwidth of the etalon is a source of error and contributes to uncertainty or error in the measurement. The calculated FWHM bandwidth for this prior art etalon is 0.13 pm and the 95% integral bandwidth for the etalon is about 1.5 pm.
For the etalon to accurately measure spectrum of a real laser, the slit function bandwidth of the etalon itself should be substantially smaller than the laser bandwidth. While this condition is satisfied for &Dgr;&lgr;
FWHM
measurements, where etalon slit function FWHM of 0.13 pm is substantially smaller than typical laser &Dgr;&lgr;
FWHM
of about 0.6 pm, the same is not true for &Dgr;&lgr;
95
% measurements, where etalon slit function bandwidth of about 1.5 pm is a substantial fraction of the expected laser bandwidth of about 2 pm.
Therefore, if the prior art etalon spectrometer with the
FIG. 2
slit function is used to measure &Dgr;&lgr;
95
%, a complicated numerical analysis is needed to deconvolve the real &Dgr;&lgr;
95
% value. Such analysis is prone to errors and ambiguous results, so no reliable &Dgr;&lgr;
95
% information is available during the microlithography process. As a result, a laser can go out of specification unnoticed. This can lead to very expensive yield problems and should be avoided.
Another way of accurately measuring laser spectrum is to use a high resolution grating spectrometers. These instruments can provide accurate spectral measurement including accurate &Dgr;&lgr;
95
% measurements, but are very bulky and expensive. These instruments are successfully used in the laboratory but are not well suited for production line microlithography use.
Another problem which needs to be solved is connected with the use of photodiode array to measure the light intensity distribution at the exit of spectrometer. Although photodiode arrays (PDA's) allow a “snap shot” of complete spectrum to be made in a single laser pulse, they do have a problem which is especially important for &Dgr;&lgr;
95
% measurements. A PDA has significant noise which is usually referred to as a “dark current” noise. This noise adds to the signal being measured, thus reducing the accuracy of measurements. The noise amplitude might be up to a few percent of maximum of the measured signal, so for the FWHM measurement it is usually not a big problem. However, it becomes a very significant problem in &Dgr;&lgr;
95
% measurements, which are very sensitive to small signal levels at the tails of the spectrum.
Currently, two methods are used to reduce the effect of dark current noise. According to the first method, few pixels on a PDA are shielded from the light coming from spectrometer. The signal from these pixels is read at the same time as the signal from the rest of the PDA and is used as a dark current reference. The problem with this method is that the dark current is assumed to be the same for all pixels, while in practice it is not. Experiments done by Applicant show that this dark current can vary by as much as two times on some PDA's. Accordingly to the second method, a PDA scan is done when the light is blocked from illuminating the PDA. The signal from every pixel is read and saved to be used as a dark current reference later when the spectral measurements are done. This technique allows individual dark current correction for every pixel. The problem, however, is that dark current noise can change over time. In fact, this noise can change significantly in just a few minutes. This requires frequent dark current calibration measurements. Even though this can be done in the laboratory, it is very difficult to do in a microlithography production environment.
What is needed is a compact spectrometer, capable of accurate measurement of both &Dgr;&lgr;
FWHM
and &Dgr;&lgr;
95
% which can be built as a part of internal electric discharge laser diagnostic set, so that it can be used in the field during the microlithography process.
SUMMARY OF THE INVENTION
The present invention provides a fast response etalon based spectrometer for spectral measurement of a pulse laser beam. A portion of the beam is directed through a double pass etalon device which provides angular separation of spectral components of the beam. The spectral components are measured by a fast photodiode array. In a preferred embodiment periodic photodiode data is also collected between pulses to record background dark current values for each photodiode in the array. In a preferred embodiment at least one set of dark current data is collected for each set of laser spectral data embodiments and the dark current data are subtracted from the laser s
Cymer Inc.
Font Frank G.
Natividad Phil
Ross, Esq. John R.
LandOfFree
Method and device for spectral measurements of laser beam 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 spectral measurements of laser beam, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for spectral measurements of laser beam will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2602263