Etching a substrate: processes – Forming or treating mask used for its nonetching function
Reexamination Certificate
2000-10-24
2004-04-06
Alanko, Anita (Department: 1765)
Etching a substrate: processes
Forming or treating mask used for its nonetching function
C216S060000, C216S085000, C438S008000, C438S009000, C438S016000
Reexamination Certificate
active
06716362
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to endpoint detection, and more particularly to thin film laser reflectance correlation for quartz etch endpoint.
2. Description of the Related Art
Several conventional endpoint detection techniques exist in the industry for detecting the endpoint of substrates/materials such as quartz, etc. Examples include laser reflectance, optical emission spectroscopy, and mass spectroscopy. These techniques have been proven to work quite well for several plasma applications, etc. In each case, a change in either the film or the plasma, signals the endpoint of the film being etched.
However, these methods prove to be useless in the fabrication of alternating phase-shift masks because this particular process requires that portions of the photomask substrate (e.g., quartz) be etched to a specific depth. Since this involves etching into a bulk quartz substrate, where there is no etch stop, no change in either the film or plasma characteristics can be observed. This is a problem.
Previously, etching to the correct depth in quartz substrate was almost an art form, which also required some luck. The process, referred to as an “iterative etch process”, would be typically performed as follows.
First, the operator assumes a quartz etch rate (e.g., based on the last run completed) and determines the time needed to etch about 80% of the desired depth. Next, the part is etched. Then, the part is unloaded and taken to a profilometer, where the approximate etch depth is measured. Etch rate is then re-calculated and then the part is loaded back into the etch tool to complete the remaining etch.
However, not only does this procedure require a lot of work, it also requires an extremely stable process.
Moreover, if the process etch rate fluctuates, the chances of “hitting” the correct depth drops dramatically. The procedure also demands a skilled operator, and several tool transfers, both of which are costly and not manufacturing-friendly.
Thus, these methods are very inaccurate and result in tedious, time-consuming operations requiring highly-skill, highly-trained operators. Otherwise, yield is low and waste of materials is high.
FIGS. 1A-3B
illustrate the conventional etching process and the laser reflectance signal for the conventional dry etching of a chrome on glass etch.
In
FIG. 1A
, a substrate
1
(e.g., quartz) is shown having a chrome material
2
deposited thereon and a chromium oxide
3
is formed over the chrome
2
. A laser
4
is shown for etching through the chromium oxide
3
(CrOx) to remove the chrome
2
on the substrate. A resist is shown for use as a patterning mask.
FIG. 1B
shows the voltage signal as a function of time at the start of etching to represent the reflectance of the laser.
FIGS. 2A and 2B
respectively illustrate the etching of the CrOx and the associated waveform of the reflectance signal.
FIGS. 3A and 3B
respectively illustrate the etching and removal of the Cr from the quartz substrate and the associated waveform of the reflectance signal until the endpoint is detected. As is evident, the reflectance signal changes as the etching continues.
Thus,
FIGS. 1A-3B
illustrate a typical laser reflectance trace used to etch chrome on the quartz substrate. This also represents typical methods used for endpointing on several types of dry etching processes, by simply changing the films involved, or the curve. The voltage signal could also be a wavelength signal emitted from the plasma (used in optical emission spectroscopy). In this particular case, in the conventional method, the laser is focused on the primary film (chrome). Since the chrome is reflective, and the substrate (quartz) below it is not reflective, by observing the laser reflectance voltage signal, which represents the amount of reflectivity, it is easy to judge when the chrome film has been completely etched. As the chrome oxide (CrOx), or anti-reflective layer, is quickly etched away the signal rises. At the peak of the signal, the chrome is exposed and begins to be etched away. When the chrome becomes thin, some of the signal is lost to transmission. Finally, when no chrome is remaining the signal flattens out due to the presence of the secondary film, quartz, which is transmissive.
Typical endpoint methods use software algorithms to detect a change in the output curve to trigger endpoint, as was done in the above example. However, as described above, the problem with etching quartz lies in the fact that there will be no change in the quartz film, or plasma to trigger an endpoint.
Typically, in alternating photomask production, a standard binary mask is used (e.g., chrome on a quartz substrate). As mentioned above, to etch the quartz substrate to a desired target depth, the iterative etch process is used. The idea behind this process is to etch to a pre-selected time. The etch depth is then measured to determine the etch rate. That calculated etch rate is then used to determine the etch time for the second etch pass. As mentioned above, this process has several problems including several tool in/outs (tool transfers leading to additional defects such as reduced yield), potential for operator error, increased defects, and it would be very susceptible to process variations. The use of an endpoint method would show tremendous benefits.
SUMMARY OF THE INVENTION
In view of the foregoing and other problems, drawbacks, and disadvantages of the conventional methods, an object of the present invention is to provide a method for thin film laser reflectance correlation.
In a first aspect, a method of etching a substrate according to the present invention includes measuring a reflectance signal from a reflective material deposited on the substrate as the substrate is being etched, correlating the substrate etch rate to the reflectance signal from the reflective material, and using the etch relation between the substrate and the reflective material to determine the etch target.
In a second aspect, a method of etching a material includes measuring a reflectance signal from a correlation material that is removed from the path of a second material that is to be etched as the second material is etched, correlating the second material etch rate to the reflectance signal from the correlation material, and using the etch ratio between the correlation material and the second material to determine the etch target.
In a third aspect, a method of etching a semiconductor substrate includes measuring a reflectance signal from an opaque material deposited on the semiconductor substrate as the semiconductor substrate is being etched, correlating the semiconductor substrate etch rate to the reflectance signal from the opaque material, and using the etch relation between the semiconductor substrate and the opaque material to determine the etch target.
Generally, the invention takes advantage of a metal film (e.g., a chrome film) which is already on a photomask used with the etching process. For purposes whereinbelow, chrome will be assumed to be the metal film, but of course, as would be known by one ordinarily skilled in the art after taking the present specification, any metal (or other opaque material) providing a predetermined reflectance signal could be used. The surface of the chrome film contains an anti-reflective chrome oxide which isolates the chrome from the etching process. This film is etched during the quartz etch process. By correlating the quartz etch to the rate of the chrome oxide etch, the reflectance signal from the chrome can be used to determine an endpoint for the quartz etch process.
With the unique and unobvious features of the invention, even if the process etch rate luctuates, the correct depth will still be etched. Further, a skilled operator is not required nor are the several tool transfers which are required by the conventional methods. Thus, a low-cost, efficient method is provided, thereby resulting in greater yield and less waste of materials.
REFERENCES:
patent: 4367044 (1983-01-01), Booth et al.
patent: 4
Alanko Anita
Kotulak, Esq. Richard M.
McGinn & Gibb PLLC
LandOfFree
Method for thin film laser reflectance correlation for... 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 for thin film laser reflectance correlation for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for thin film laser reflectance correlation for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3193844