JP2006287258A - Narrowband gas laser apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a narrowband gas laser apparatus such as a fluorine laser apparatus capable of controlling the discharge of ASE for the purpose of keeping the spectral line width to be no more than 0.2 pm and spectral purity no more than 0.5 pm. <P>SOLUTION: This narrowband gas laser apparatus has a discharge electrode 2 in a laser chamber 1, where laser gas containing F<SB>2</SB>is enclosed, and emits narrowband laser beams using a narrowband module 6 including a wavelength selection element 8 installed in a laser cavity. Time duration from the point of emitting light by electric discharge to the generation of laser beams is set so that laser beams emitted from the laser cavity may have spectral line width (FWHM) of no more than 0.2 pm and/or spectral purity of no more than 0.5 pm. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

The present invention relates to a narrow-band gas laser device, and in particular, a spectral line width of 0.2 pm or less and a spectral purity such that an ASE (Amplified Spontaneous Emission) is not included in a laser beam from a gas laser device such as an F ₂ laser device. The present invention relates to a gas laser device narrowed to 0.5 pm or less.

In addition, ASE was not included in the laser beam from a gas laser device using a laser gas containing F ₂ such as a KrF laser device or an ArF laser device, and the spectral line width and spectral purity were further narrowed. The present invention relates to a gas laser device.

  With the miniaturization and high integration of semiconductor integrated circuits, there is a demand for improvement in resolving power in an exposure apparatus for manufacturing the semiconductor integrated circuit. For this reason, the wavelength of the exposure light emitted from the exposure light source has been shortened. As a light source for semiconductor exposure, a KrF laser device having a wavelength of 248 nm from a conventional mercury lamp and a further short wavelength light source have a wavelength of 193 nm. ArF laser devices are beginning to be used.

Further, in an exposure technique for realizing a semiconductor integrated circuit having a line width of 70 nm or less on a semiconductor, an exposure light source having a wavelength of 160 nm or less is required. At present, an F ₂ (fluorine molecule) laser device that emits ultraviolet rays having a wavelength of about 157 nm is regarded as a promising light source, and research and development for mounting an exposure device is being carried out at a rapid pace. The present invention, KrF laser device, the performance improvement of ArF laser device, and, in order to mount the F ₂ laser device to a semiconductor exposure apparatus, those relating to the performance improvement of the F ₂ laser system.

  There are the following two types of optical technology.

1) Dioptric system
2) Catadioptric system
As a refractive system, there is a projection optical system generally used in a conventional exposure apparatus. In the exposure technique, the chromatic aberration correction method in the optical system is one big problem. In the refraction system, chromatic aberration correction has been realized by combining optical elements such as lenses having different refractive indexes. However, there is a limit to the types of optical materials that can be used in the wavelength region near 157 nm, and currently only CaF ₂ (fluorite) can be used.

  When the catadioptric system is used for the exposure technique, since the reflective optical element has no chromatic dispersion, the occurrence of chromatic aberration can be suppressed by combining the reflective optical element and the refractive optical element. Therefore, an exposure apparatus using a catadioptric system is promising in the current wavelength range near 157 nm. However, the catadioptric system tends to be avoided because it is difficult to adjust the optical axis of the exposure machine as compared to the conventional refractive system.

At present, there is one effective means for adapting a refraction system, which has been common as a prior art, to the wavelength region near 157 nm. That is to use an F ₂ laser device in which the emitted laser beam is narrowed as the light source of the exposure apparatus.

Although it depends on the operating conditions such as the total discharge gas pressure, the full width at half maximum (FWHM) of the spectrum of the F ₂ laser beam is generally about 1.5-1. It is around 2pm. In a refraction system, it is required to narrow the full width at half maximum of this spectrum to 0.2 pm or less. In addition, the KrF laser device and the ArF laser device are also required to have a narrow band because the full width at half maximum (FWHM) of the laser beam spectrum during free running operation is as wide as several hundred nm. The present invention relates to the narrow band technology.

FIG. 1 shows a configuration example of a narrow-band F ₂ laser device using one or more magnifying prisms and a grating (diffraction grating). The narrow-band KrF laser device and ArF laser device have the same configuration.

The laser chamber 1 is filled with F ₂ laser excitation medium gas (hereinafter referred to as laser gas). A high voltage pulse is applied from a high voltage pulse generator 3 to a pair of electrodes 2 arranged opposite to each other in the laser chamber 1 by a predetermined distance, and a discharge is generated between the electrodes 2, and the laser gas in this discharge part is excited. Is done. Seed light that becomes a laser beam is generated from the excited laser gas. A fan 4 and a radiator (not shown) are further installed in the laser chamber 1. The laser gas is circulated in the laser chamber 1 by the fan 4, and the laser gas heated to high temperature by the discharge is heat-exchanged with the radiator and cooled. As shown in FIG. 1, the laser chamber 1 is provided with a window member in a window shape 5 in the shape of a letter C at a Brewster angle or a parallel Brewster angle. The electrode 2 includes an anode electrode and a cathode electrode that are separated from each other by a predetermined distance in the direction perpendicular to the paper surface.

  The laser resonator is composed of a diffraction grating 8 and an output mirror 9 mounted on a band-narrowing module 6 described later.

  The seed light that becomes the laser beam described above reciprocates between the diffraction grating (grating) 8 and the magnifying prism 7, the discharge unit, and the output mirror 9 that are installed in the band narrowing module 6. Is extracted as a laser beam.

  A part of the laser beam emitted from the output mirror 9 is introduced into the wavelength monitor 11 by the beam splitter 10, and the output, center wavelength, etc. are measured by the wavelength monitor 11.

  The band narrowing is performed by arranging an optical band narrowing module 6 having a spectral function in the laser resonator. For example, the band narrowing module 6 includes a casing and a diffraction grating (grading) 8 and a magnifying prism 7 installed in the casing, and the spectrum narrowing is realized by wavelength selection by the diffraction grating 8.

  The oscillation center wavelength can be changed by rotating any one of the diffraction grating 8 and the magnifying prism 7.

  In addition, a high reflection mirror is installed at an arbitrary position between the laser chamber 1 and the diffraction grating 8, and the oscillation center wavelength is changed by rotating the high reflection mirror to change the incident angle of light to the diffraction grating 8. Is also possible.

Based on the center wavelength signal from the wavelength monitor 11, the controller 12 is arranged at the diffraction grating 8 in the narrowband module 6, the magnifying prism 7, or an arbitrary position between the laser chamber 1 and the diffraction grating 8 (not shown). The wavelength control is performed by rotating any one of the high reflection mirrors installed.
Proc. SPIE Vol. 3679. (1999) 1030-1037

Even if an optical narrow band module having a spectroscopic function is used in the laser resonator as the narrow band means of such a narrow band F ₂ laser apparatus, the spectral line required in the refractive system of the exposure apparatus is used. It was difficult to obtain a narrow band performance with a width (FWHM) of 0.2 pm or less.

If the spectral line width is Δλ, the light beam incident on the diffraction grating is W, and the blaze angle (= Littrow angle) of the diffraction grating is θ,
Δλ∝cosθ / W
There is a relationship. That is, the spectral width Δλ becomes narrower as the blaze angle θ of the diffraction grating and the width of the light ray W incident on the diffraction grating become larger.

  In order to increase the light beam width W, it is necessary to increase the magnification ratio of the magnifying prism, increase the number of magnifying prisms, and increase the width of the diffraction grating. However, when the light source for exposure is used, the size of the narrowband module is also limited due to restrictions on the apparatus size at the exposure site. Further, since the light transmittance of the magnifying prism is not 100% with respect to light having a wavelength of 157 nm, the oscillation efficiency decreases as the number increases. That is, there is a limit to increasing the number of magnifying prisms or increasing the width of the diffraction grating.

  In addition, there are limits that come from the manufacturing technology of optical components. For example, the blaze angle cannot be increased beyond a predetermined value due to the manufacturing limit of the diffraction grating.

  Due to the above circumstances, there is a limit to optical narrowing.

  Here, Non-Patent Document 1 describes that as the laser pulse width increases, the spectral line width of the laser light decreases accordingly. Proven.

  That is, in order to realize further narrowing beyond the limit of the optical narrowing described above, it is necessary to make the laser beam long pulse (pulse stretch).

  However, it is difficult to obtain a narrow-band performance with a spectral line width (FWHM) of 0.2 pm or less even when a long pulse is applied.

This is due to the following reason. The F ₂ (molecular) laser device has a large gain, similar to the excimer laser (KrF, ArF, XeCl, etc.). A large gain in the laser device means that a large amount of light (ASE: Amplified Spontaneous Emission) is emitted from the output mirror of the laser device without resonating using the narrowband module. Since ASE does not reciprocate the laser resonator and is emitted from the output mirror, it is considered that the light does not pass through the narrowband module or is passed once even though it passes through. Since the laser beam emitted from the conventional narrow-band F ₂ laser device contains this ASE component, it has been difficult to narrow the spectral line width (FWHM) of the laser beam to 0.2 pm or less. Hereinafter, the generation of ASE will be described in detail.

FIG. 2 shows a side light during laser oscillation (hereinafter referred to as a side light) waveform and laser in the case of an F ₂ laser device that cannot obtain a narrow bandwidth performance of a conventional spectral line width (FWHM) of 0.2 pm or less. It is a figure which shows the time transition of a pulse waveform and a spectrum line width. Note that the length of the laser resonator formed by the diffraction grating and the output mirror at the time of waveform acquisition was 1500 mm.

  Here, the light generated by the laser gas excited by the discharge generated between the pair of electrodes is referred to as “side light”. The sidelight is observed from a position not on the laser resonator (for example, an electrode side position in a direction substantially perpendicular to the longitudinal direction of the electrode).

  The sidelight waveform shows the temporal gain distribution of the laser. That is, this sidelight shows a gain distribution during laser oscillation.

  The laser pulse rises rapidly beyond the threshold at the sidelight peak. That is, the main laser oscillation (oscillation that is not ASE) suddenly rises starting from the first peak of the sidelight.

  In the laser pulse waveform, one peak is observed at a position exceeding 20 ns from the start of discharge excitation (0 ns). The spectral line width at time A at this peak position was substantially the same as the spectral line width during free running (Note that the vertical axis indicating the spectral line width in FIG. 2 is not linear. (The spectral linewidth appears to be about 0.4 pm, but it is actually much larger.)

As described above, the F ₂ laser device has a large gain, similarly to the excimer laser (KrF, ArF, XeCl, etc.). In a laser device having a large gain, when a predetermined value after the gain rises (that is, after a predetermined time delay from the rise of the gain), the light is oscillated by passing through one pass without reciprocating the resonator (ASE). ) Occurs.

  The peak at time A of the laser pulse waveform shown in FIG. 2 was observed even when the optical axis of the laser resonator was shifted and misaligned. Therefore, the light at the first peak portion of the laser pulse waveform is considered as ASE. As described above, ASE is light emitted from the output mirror without reciprocating the laser resonator, so it is considered that the light does not pass through the narrowing module or is passed once. Rarely done.

  Light that has not been extracted as ASE reciprocates through the laser resonator, is narrowed, and is eventually extracted as a laser beam.

  As schematically shown in FIG. 3A, the spectral line width of one laser pulse is a temporal integration value of each spectral line width at each point in the laser pulse, so that ASE is emitted at the initial stage of the laser pulse. As a result, it was difficult to set the spectral line width to 0.2 pm or less. That is, as schematically shown in FIG. 3B, ASE having a spectral line width of 0.6 pm or more is overlaid on the spectral characteristics of the laser pulse, and the integral spectrum of the entire laser pulse is the main laser oscillation. Even if the spectral line width is 0.2 pm or less, the spectral line width is 0.2 pm or more as a whole.

  On the other hand, with the above-described laser pulse waveform with ASE, it is difficult to satisfy the specifications as an exposure light source in terms of spectral purity.

  Here, “spectral purity” is one index related to the degree of concentration of spectral energy, and as shown in FIG. 4, refers to a line width including “a certain area ratio” of the spectrum waveform. For example, “95% purity”, which is commonly used, refers to a line width including an area of 95% from the center side in the total area of the spectrum waveform. Usually, the spectral purity required for a refractive exposure light source is 0.5 pm.

  As described above, ASE is light that is hardly narrowed. That is, no matter how long the pulse is made (pulse stretch), as long as the ASE component exists, the waveform around the bottom of the integrated waveform of the spectrum shown in FIG. 3 does not change. Therefore, depending on the requirements on the exposure apparatus side, the specifications for spectral purity cannot be satisfied. The required values of spectral line width and spectral purity vary depending on the wavelength, but the above-mentioned problems also apply to KrF laser devices and ArF laser devices.

The present invention has been made in view of such problems of the prior art, and its object is narrowed in the gas laser apparatus using a laser gas containing a KrF laser system, ArF laser device, the F _2, such as F ₂ laser system To provide a narrow-band gas laser device that suppresses ASE emission in order to achieve improved performance, particularly in an F ₂ laser device, to achieve a spectral line width of 0.2 pm or less and a spectral purity of 0.5 pm or less. .

The narrow-band gas laser apparatus of the present invention that achieves the above object has a narrow-band module having a discharge electrode in a laser chamber in which a laser gas containing F ₂ is sealed, and including a wavelength selection element installed in the laser resonator. In a narrow-band gas laser device that emits a laser beam narrowed by
At the time of generation of the laser beam emitted from the laser resonator, the laser beam is emitted from the time when light is emitted by discharge so that the spectral line width (FWHM) is 0.2 pm or less and / or the spectral purity is 0.5 pm or less. This is characterized in that a time until the occurrence of the error is set.

  Further, it is a two-stage narrow-band gas laser apparatus composed of an oscillation stage laser and an amplification stage laser, and the oscillation stage laser includes the narrow-band gas laser apparatus that is the above-mentioned narrow-band gas laser apparatus.

In this two-stage narrow-band gas laser device, the amplification stage laser is connected to a discharge light emission measuring instrument (side light detector) that measures a temporal pulse waveform of light emission by discharge, and to the discharge electrode of the amplification stage laser. A controller capable of controlling at least one of an applied voltage, an F ₂ concentration of the laser gas, and a laser gas pressure in the laser chamber, wherein the controller is timing data at which a laser pulse of the oscillation stage laser rises from the controller of the oscillation stage laser. It is desirable to start the discharge of the amplification stage laser after receiving T _LPS and starting the laser pulse of the oscillation stage laser based on this data.

  According to the present invention described above, in the narrow-band gas laser device such as a fluorine laser device, the rising of the sidelight is made gentle so that the starting point of the laser pulse exists after the time of the first peak of the sidelight, ASE emission can be suppressed, and an ultra-narrow band gas laser device having a spectral line width of 0.2 pm or less and a spectral purity of 0.5 pm or less can be realized. In addition, it has become possible to realize a narrow-band gas laser device with improved narrow-band performance in the KrF laser device and ArF laser device. In addition, in order to provide a high-power, ultra-narrow band fluorine laser device, the injection lock method and the MOPA method are adopted, and in the oscillation stage laser having the narrow band means, the rise of the sidelight is thus moderated. Thus, a laser beam having a spectral line width of 0.2 pm or less and a pulse energy of 5 mJ or more can be obtained by making the starting point of the laser pulse exist after the time of the first peak of the sidelight. It became so.

  Similarly, in order to provide a high-power / ultra-narrow band KrF laser apparatus and ArF laser apparatus, an oscillation stage laser that employs an injection lock system or a MOPA system and has a band narrowing means in this way The start of the laser pulse is present after the first peak of the sidelight so that the spectral line width is less than the required line width from the exposure apparatus, and the pulse energy is 5 mJ. The above laser beam can be obtained.

From the description of the problem to be solved by the invention, in order to realize the spectral line width of 0.2 pm or less and the spectral purity of 0.5 pm or less, the ASE component is obtained from the light emitted from the narrow-band F ₂ laser device. It is understood that it is necessary to suppress and eliminate. Further, it is understood that in order to improve the band narrowing performance, it is necessary to suppress and remove the ASE component from the light emitted from the band narrowing KrF laser device and ArF laser device.

As a result of diligent research, the present inventors have made the rise of the laser gain slow and suppress the laser pulse in order to suppress / remove the ASE component from the light emitted from the narrow-band F ₂ laser device as shown in FIG. We found that it is effective to delay the rise start time. The same applies to the narrow-band KrF laser device and ArF laser device. In the following, description will be given taking a narrow band F ₂ laser device as an example.

  FIG. 5 shows a temporal transition (a) of a laser pulse waveform, a sidelight waveform, and a spectral line width in the present invention (details will be described later) in comparison with a conventional temporal transition (b). FIG. 5B is the same as FIG. In the case of FIG. 5A as well, the length of the laser resonator formed by the diffraction grating and the output mirror at the time of waveform acquisition was 1500 mm as in FIG. The configuration of the narrowband module is also the same.

  As is clear from the comparison between the two, the rise of the sidelight (that is, the rise of the laser gain) of the present invention (FIG. 5A) is slower than that of the conventional example (FIG. 5B), and the ASE component is also increased. It is not present in the laser pulse waveform. Further, when the start point of the sidelight rises (0 ns), the start point of the rise of the laser pulse waveform is higher in the present invention (FIG. 5A) than in the conventional example (FIG. 5B). slow.

  The reason why the ASE component is suppressed in this way when the sidelight rise (laser gain rise) is slow is considered to be as follows. As described above, in the case of the conventional example (FIG. 2), the rise of the sidelight (rise of the laser gain) is fast, the light generated by the discharge is rapidly amplified, and the laser beam exceeds the predetermined threshold as a laser beam. Before being extracted from the resonator, the amplified light is extracted as ASE without reciprocating the laser resonator.

  On the other hand, in the case of the present invention (FIG. 5A), the sidelight rises (laser gain rises) slowly, so that the light generated by the discharge is gradually amplified. Then, the light is not rapidly amplified as it is extracted from the laser resonator as ASE, but reciprocates in the laser resonator, and is eventually extracted as a laser beam from the laser resonator exceeding the threshold value. That is, there is no ASE in the light emitted from the laser device.

  Conventionally, as a method for realizing a narrow band, a long pulse (pulse stretch) of a laser pulse waveform has been mentioned. In this method, depending on conditions, an ASE component is included in the laser pulse waveform as shown in FIG. In some cases, it was difficult to realize a spectral line width of 0.2 pm or less and a spectral purity of 0.5 pm or less.

On the other hand, the inventors have been able to suppress the occurrence of ASE as described above by controlling the rise of the sidelight (rise of the laser gain), which has not been considered in the past. Therefore, the spectral line width and spectral purity required for a light source for refractive exposure systems are satisfied by setting the performance of the narrowband module, appropriately setting the laser resonator length, and increasing the laser pulse width. The narrow-band F ₂ laser device can be realized.

  By the way, the temporal transition of the spectral line width and the laser pulse waveform in FIGS. 5A and 5B are summarized as shown in FIG.

  As described above, in the case of the present invention (FIG. 5A), the rise of the sidelight (rise of the laser gain) is slow, so the amplification speed of the light generated by the discharge is slow. Therefore, when the side light rises as a starting point, as clearly shown in FIG. 6, the present invention (FIG. 5 (a)) is the conventional example (FIG. 5 (b)) at the rising start point of the laser pulse waveform. Slow compared to).

  Here, the temporal transitions of the spectral line widths of the conventional example (FIG. 2: FIG. 5B) and the present invention (FIG. 5A) are on substantially the same curve, and the laser pulse waveform of the present invention. At the time of rising, the spectral line is substantially the same as the spectral line width in the conventional example at the same point.

  That is, the light that travels back and forth in the laser resonator until the light generated after the start of discharge (after the start of sidelight emission) is taken out from the laser resonator as a laser beam (until the laser pulse rises) Since it passes several times, it is considered that the band is narrowed to some extent.

  Therefore, by controlling the rise of the sidelight (rise of the laser gain) so that the laser pulse rises after the light in the laser resonator is narrowed to a predetermined spectral linewidth, the spectral linewidth ( It is possible to narrow the integral value) to a predetermined value and improve the spectral purity.

To summarize the above, the inventors have controlled the rise of the sidelight (rising of the laser gain), which has not been considered in the past, thereby suppressing the generation of ASE and narrowing the band to some extent from the time of laser pulse generation. As a result, it is possible to provide a narrow-band F ₂ laser device that satisfies the spectral line width and spectral purity required for a light source for a refractive exposure apparatus.

  FIG. 7 shows the relationship between the time from the start of sidelight emission until the first peak of sidelight occurs (black circle), the start point (square) of the laser pulse waveform including ASE, and the spectral line width (FWHM) of the laser pulse. Show. Here, the spectral line width (FWHM) in FIG. 7 is a temporal integration value of each spectral line width at each time point in the laser pulse as shown in FIG. 3 as described above. The length of the laser resonator formed by the diffraction grating and the output mirror was 1500 mm.

  In addition, since the start point of the side light waveform and the laser pulse waveform is difficult to accurately determine from the waveform, here, the time point before the first peak occurrence of each waveform and 5% of the intensity of the first peak is obtained. The starting point of the sidelight waveform and laser pulse waveform was used. However, since the difference between the time when the intensity is less than 5%, for example, the time when the intensity is 1% and the time when the intensity is 5% is very small, before the first peak occurrence of each waveform, and An arbitrary time point that is less than 5% of the intensity of the first peak may be set as the starting point of the sidelight waveform and the laser pulse waveform.

  In each measurement shown in FIG. 7, when the laser pulse waveform was observed, the ASE component was included in the laser pulse waveform when the first pulse of the sidelight occurred about 35 ns before the start point of the sidelight defined above. . At this time, as shown in FIG. 2, the starting point of the laser pulse waveform occurred before the time of the first peak of the sidelight. On the other hand, when the first pulse of the sidelight is generated after about 35 ns from the starting point of the sidelight, the ASE component is not included in the laser pulse waveform. At this time, as shown in FIG. 5A, the starting point of the laser pulse waveform occurred near or after the first peak of the sidelight.

  Further, as apparent from FIG. 7, when the first peak of the sidelight occurs after 40 ns from the starting point of the sidelight, the spectral line width is less than 0.2 pm. Further, although not shown in FIG. 7, the spectral purity was 0.5 pm or less.

  To summarize the above, it was found that when the start of the sidelight is slow and the start point of the laser pulse exists after the time of the first peak of the sidelight, the ASE component is not included in the laser pulse waveform.

  Next, side light rising control (ASE suppression control) according to the present invention will be described.

  As described above, the present invention controls the rise of the sidelight to suppress the generation of ASE, and the rise of the laser pulse until the light reciprocating through the laser resonator is narrowed to a predetermined band or less. It is characterized by delaying.

  It has been found that control parameters for controlling the rise of the sidelight (that is, controlling the start point of the rise of the laser pulse) are mainly as follows.

(1) Applied voltage FIG. 8 shows that the applied voltage applied to the electrode from the high-voltage pulse generator is changed under the same narrowband module and the same gas conditions (gas mixing ratio and gas pressure in the laser chamber). It is a figure which shows the delay time Td from the starting point of the rise of the spectral line width and side light to the starting point of the rise of the laser pulse.

  When the applied voltage applied to the electrode from the high voltage pulse generator is lowered, the delay time Td is delayed as shown in FIG. That is, the laser gain decreases due to the decrease in applied voltage, and the starting point of the rise of the laser pulse is delayed.

  In FIG. 8, when the applied voltage C has a large voltage value, ASE occurs because the gain is high, the delay time is as short as T3, and the spectral line width is more than 0.2 pm.

  At the applied voltage B, the voltage value is smaller than that at the applied voltage C, and no ASE occurs. The delay time at that time is T2, which is longer than T3, but the number of times that the light reciprocates through the laser resonator is not sufficient, and thus the spectral line width exceeds 0.2 pm.

  On the other hand, when the applied voltage is A, the voltage value is smaller than when the applied voltage is B, and ASE is not generated. The delay time at that time is T1, which is longer than T2, the number of times the light reciprocates through the laser resonator is sufficient, and the spectral line width is less than 0.2 pm.

  In this way, under the same narrowband module and the same gas conditions, by controlling the applied voltage applied to the electrode from the high voltage pulse generator, the rise time of the sidelight is controlled and the generation of ASE is suppressed. It is also possible to control the delay time Td from the start point of the sidelight rise to the start point of the laser pulse so that the laser pulse rises after the band is sufficiently narrowed.

(2) F ₂ concentration FIG. 9 shows the spectral line width and the starting point of the sidelight when the F ₂ concentration contained in the laser gas is changed under the same narrowband module, the same applied voltage, and the same gas pressure. It is a figure which shows the delay time Td to the starting point of the rise of a laser pulse.

When the F ₂ concentration in the laser gas is decreased, the delay time Td is delayed as shown in FIG. That is, the laser gain decreases due to the decrease in the F ₂ concentration, and the starting point of the rise of the laser pulse is delayed.

In FIG. 9, when the F ₂ concentration is high, the gain is high, so ASE occurs, the delay time is as short as T3, and the spectral line width is more than 0.2 pm.

At the density E, the F ₂ density is smaller than at the density F, and no ASE occurs. The delay time at that time is T2, which is longer than T3, but the number of times that the light reciprocates through the laser resonator is not sufficient, and thus the spectral line width exceeds 0.2 pm.

On the other hand, at the concentration D, the F ₂ concentration is smaller than that at the concentration B, and no ASE is generated. The delay time at that time is T1, which is longer than T2, the number of times the light reciprocates through the laser resonator is sufficient, and the spectral line width is less than 0.2 pm.

Thus, under the same narrowband module, the same applied voltage, and the same gas pressure, by controlling the F ₂ concentration contained in the laser gas, the rise of the sidelight can be controlled to suppress the generation of ASE. In addition, the delay time Td from the start point of the sidelight rise to the start point of the laser pulse can be controlled so that the laser pulse rises after the band is sufficiently narrowed.

(3) Laser gas pressure FIG. 10 shows the spectral line width when the laser gas pressure (hereinafter referred to as total pressure) in the laser chamber is changed under the same narrowband module, the same applied voltage, and the same F ₂ concentration. FIG. 6 is a diagram showing a delay time Td from the start point of the sidelight rise to the start point of the rise of the laser pulse.

  When the total pressure is lowered, the delay time Td is delayed as shown in FIG. That is, the laser gain decreases due to the decrease in the total pressure, and the starting point of the rise of the laser pulse is delayed.

  In FIG. 10, when the total pressure I has a large pressure value, ASE occurs because the gain is high, the delay time is as short as T3, and the spectral line width exceeds 0.2 pm.

  When the total pressure is H, the total pressure is lower than when the total pressure is I, and no ASE is generated. The delay time at that time is T2, which is longer than T3, but the number of times that the light reciprocates through the laser resonator is not sufficient, and thus the spectral line width exceeds 0.2 pm.

  On the other hand, when the total pressure is G, the total pressure is lower than when the total pressure is H, and ASE is not generated. The delay time at that time is T1, which is longer than T2, the number of times the light reciprocates through the laser resonator is sufficient, and the spectral line width is less than 0.2 pm.

Thus, under the same narrowband module, the same applied voltage, and the same F ₂ concentration, by controlling the total pressure, the rise of the sidelight can be controlled, and the occurrence of ASE can be suppressed sufficiently. It is also possible to control the delay time Td from the start point of the sidelight rise to the start point of the laser pulse so that the laser pulse rises after being narrowed.

  Next, a specific example for controlling the rise of such a sidelight will be described.

FIG. 11 shows a configuration example of the narrow-band F ₂ laser apparatus of the present invention. Here, portions common to those in FIG. 1 have equivalent functions, and thus description thereof is omitted.

Connected to the laser chamber 1 are a gas supply line 26 for supplying buffer gas such as F ₂ gas and rare gas into the laser chamber 1 and an exhaust line 27 for exhausting gas in the laser chamber. .

The gas supply line 26 includes a line for supplying F ₂ gas and a line for supplying buffer gas such as rare gas. A line for supplying a buffer gas such as a rare gas is connected to a buffer gas supply source (not shown) via a valve V1, and a line for supplying F ₂ gas is an F ₂ gas (not shown) via a valve V2. Connected to the supply source. Since F ₂ gas has a very high reactivity, diluted F ₂ gas diluted with F ₂ gas with rare gas or the like is supplied from the normal F ₂ gas supply source.

  The two lines are integrated downstream of the valves V1 and V2 and connected to the laser chamber 1.

  The exhaust line 27 is connected to an exhaust means (not shown) via a valve V3.

  The controller 22 controls the opening and closing of the valves V1, V2, and V3 to supply and exhaust gas into the laser chamber 1.

  On the side surface of the laser chamber 1 (electrode side position substantially perpendicular to the longitudinal direction of the electrode 2), an observation window 23 for observing the sidelight is provided. The temporal pulse waveform of the sidelight is measured by the sidelight detector 24, and the measurement data is sent to the controller 22.

  On the other hand, a part of the laser beam emitted from the output mirror 9 is measured by the laser pulse detector 21 by the beam splitter 10, and the measurement data is sent to the controller 22.

  The controller 22 also controls the high voltage pulse generator 3 to control the value of the applied voltage applied to the electrode 2.

  The laser chamber 1 is provided with a pressure gauge 25 that measures the pressure of the laser gas in the laser chamber 1, and pressure data is sent to the controller 22.

By the way, as described above, the present inventors have obtained the following knowledge by analyzing the experimental results and the like in detail. That is, as shown in FIGS. 12 and 13, when the first peak time T _{SL1 of the sidelight} and the first peak time T _LP1 of the laser pulse are compared, ASE occurs when T _SL1 > T _LP1 (FIG. 12). If T _SL1 ≦ T _LP1, no ASE occurs (FIG. 13). In FIGS. 12 and 13, T _LPS is the starting point of the laser pulse. Based on such knowledge, generation of ASE is suppressed by controlling as follows.
(1) Voltage control FIG. 14 shows a flowchart for suppressing the generation of ASE by voltage control.

  First, the controller 22 outputs a command to the high voltage pulse generator (hereinafter referred to as a power source) 3 to generate laser discharge, and starts laser oscillation (step S101).

  Next, the sidelight detector 24 and the laser pulse detector 21 detect the temporal pulse waveform of the sidelight and the temporal pulse waveform of the laser pulse (step S102), and send the detected data to the controller 22 (step S103). ).

Next, the controller 22 calculates the first peak time T _SL1 of the side light and the first peak time T _LP1 of the laser pulse from the received waveform data with the origin of the side light as the origin (step S104), and T _SL1. And _TLP1 are compared in magnitude (step S105).

When T _SL1 ≦ T _LP1, it is determined that no ASE has occurred. On the other hand, when T _SL1 > T _LP1 , as shown in FIG. 8, ASE does not occur when the voltage applied to the electrode 2 is less than a predetermined value, so that the controller 22 outputs a command to the power source 3 to the electrode 2. The applied voltage is reduced by a predetermined value (step S106).

Thereafter, steps S101 to S106 are repeated until no ASE occurs.
(2) F ₂ concentration and total pressure control In the voltage control of (1) above, the voltage applied to the electrode 2 was controlled to suppress the generation of ASE, but as apparent from FIGS. 9 and 10 described above. The generation of ASE may be suppressed by controlling the F ₂ concentration in the laser gas and the pressure of the laser gas.

When controlling the F ₂ concentration, Steps S101 to S105 in FIG. 15 (same as Steps S101 to S105 in FIG. 14) are performed in the same manner as in the voltage control, and one T _SL1 > T _LP1 in Step S105. The controller 22 opens the valve V3 while observing the pressure data from the pressure gauge 25 to exhaust the laser gas in the laser chamber 1, and closes the valve V3 when the laser gas pressure reaches a predetermined pressure (step S107).

  Thereafter, the controller 22 opens the valve V1 while observing the pressure data from the pressure gauge 25 (at this time, the valve V2 is closed), fills the laser chamber 1 with buffer gas, and the laser gas pressure reaches a predetermined pressure. Then, the valve V1 is closed (step S108).

That is, the laser gas is exhausted by a predetermined amount and filled with the same amount of buffer gas, thereby reducing the F ₂ concentration in the laser gas.

  Thereafter, steps S101 to S108 are repeated until no ASE occurs.

  Note that when the generation of ASE is suppressed by controlling the pressure of the laser gas, step 108 may be omitted.

As described above with reference to FIG. 7, when the start of the side light is made gentle and the start point of the laser pulse exists after the time of the first peak of the side light, an ASE component is included in the laser pulse waveform. It has been found that Therefore, the determination of the occurrence of ASE may be performed by comparing the first peak time T _{SL1 of the sidelight} and the starting point T _{LPS of the} laser pulse (FIGS. 12 and 13). The specific control is achieved by using the starting point T _{LPS of the} laser pulse instead of the first peak time T _LP1 of the laser pulse in the above control.

Here, as described above, the starting point T _LPS of the laser pulse waveform is a time point before the occurrence of the first peak of the laser pulse waveform and 5% of the intensity of the first peak. An arbitrary time point before the first peak of the pulse waveform and less than 5% of the intensity of the first peak may be set as the starting point T _LPS of the laser pulse waveform.

By the way, it has been found that the determination of the occurrence of ASE is possible by the comparison result between the first peak time T _{SL1 of the sidelight} and the first peak time T _LP1 of the laser pulse as described above. It has been clarified by a test by the present inventors that the determination of the occurrence of ASE can also be made by this discrimination method.

That is, it has been clarified by the inventors' tests that the occurrence of ASE occurs when the maximum gradient ΔT _s of the sidelight rising exceeds a predetermined value ΔT ₁ . As shown in FIG. 16, the sidelight waveform and the laser pulse waveform, the laser pulse 1 obtained in the case of the sidelight 1 includes ASE. At this time, the maximum gradient ΔT _s at the time of rising of the sidelight is ΔT _s1 , and the relationship between the predetermined values ΔT ₁ and ΔT _s1 is ΔT ₁ <ΔT _s1 .

On the other hand, the laser pulse 2 obtained in the case of the sidelight 2 does not contain ASE. At this time, the maximum gradient ΔT _s at the time of rising of the sidelight is ΔT _s2 , and the relationship between the predetermined values ΔT ₁ and ΔT _s2 is ΔT ₁ ≧ ΔT _s1 .

As described above, the maximum gradient ΔT _s of the sidelight rising is measured, and whether or not ASE has occurred can be determined based on the gradual increase / decrease of the maximum gradient ΔT _s with respect to the predetermined value ΔT ₁ .

The maximum gradient ΔT _{s of the} sidelight rising also changes depending on the gain of the laser excitation medium, similarly to the sidelight rising time. When the gain is low, ΔT _s becomes moderate, and when the gain is high, it becomes steep. The control parameters for the maximum gradient ΔT _s of the sidelight rising are the aforementioned applied voltage, F ₂ concentration, and total pressure.

In this case, based on the above knowledge, the following control is performed to suppress the generation of ASE.
(1) Voltage Control FIG. 17 shows a flowchart for suppressing the generation of ASE by voltage control in this case.

The predetermined value ΔT ₁ depends on the performance of the narrowband module 6 (for example, the enlargement ratio of the enlargement prism 7 and the blaze angle of the diffraction grating 8) and the laser resonator length. The controller 22 stores in advance a predetermined value ΔT ₁ obtained based on the dependence condition of the laser device.

  Then, the controller 22 outputs a command to the power supply 3 to generate laser discharge, and starts laser oscillation (step S201).

  Next, the sidelight detector 24 detects the temporal pulse waveform of the sidelight (step S202), and sends the detection data to the controller 22 (step S203).

Next, the controller 22, from the received waveform data to calculate the maximum gradient [Delta] T _s range sidelight rise of from the origin of the sidelight to the first peak (step S204), the magnitude of [Delta] T _s and [Delta] T ₁ Compare (step S205).

When ΔT _s ≦ ΔT _1, it is determined that no ASE has occurred. On the other hand, when ΔT _s > ΔT ₁ , as shown in FIG. 8, ASE does not occur when the voltage applied to the electrode 2 is equal to or lower than a predetermined value, so the controller 22 outputs a command to the power source 3 to the electrode 2. The applied voltage is reduced by a predetermined value (step S206).

Thereafter, steps S201 to S206 are repeated until no ASE occurs.
(2) F ₂ concentration and total pressure control In the voltage control of (1) above, the voltage applied to the electrode 2 is controlled to suppress the generation of ASE, but as described above, the F ₂ concentration in the laser gas and the laser gas are controlled. The generation of ASE may be suppressed by controlling the pressure.

When controlling the F ₂ concentration, similarly to the voltage control, steps S201 to S205 in FIG. 18 (same as steps S201 to S205 in FIG. 17) are performed, and one ΔT _s > ΔT ₁ in step S205. The controller 22 opens the valve V3 while observing the pressure data from the pressure gauge 25 to exhaust the laser gas in the laser chamber 1, and closes the valve V3 when the laser gas pressure reaches a predetermined pressure (step S207).

  Thereafter, the controller 22 opens the valve V1 while observing the pressure data from the pressure gauge 25 (the valve V2 is closed at this time), fills the laser chamber 1 with the buffer gas, and when the laser gas pressure reaches a predetermined pressure, The valve V1 is closed (step S208).

That is, the laser gas is exhausted by a predetermined amount and filled with the same amount of buffer gas, thereby reducing the F ₂ concentration in the laser gas.

  Thereafter, steps S201 to S208 are repeated until no ASE occurs.

  Note that step 208 may be omitted when the generation of ASE is suppressed by controlling the pressure of the laser gas.

In the ASE control, in the above-described example, any one of the voltage applied to the electrode 2, the F ₂ gas concentration in the laser gas in the laser chamber 1, and the laser gas pressure in the laser chamber 1 is used as the control parameter. However, the present invention is not limited to this, and it is also possible to perform control by combining a plurality of the above parameters.

  By the ASE suppression control as described above, the ASE component disappears in the laser pulse, and by performing the narrow band module 6 performance, appropriately setting the laser resonator length, increasing the laser pulse width, etc. It was possible to satisfy the spectral line width and spectral purity required for a light source for a refractive exposure apparatus.

By the way, depending on the performance of the narrowband module 6, for example, even if generation of ASE is suppressed, a predetermined spectral line width (for example, 0.2 pm or less) may not be obtained. For example, at the applied voltage B in FIG. 8, at the F ₂ concentration E in FIG. 9, or at the laser gas total pressure H in FIG. 10, no ASE has occurred, but the spectral line width exceeds 0.2 pm. ing.

  This is because, in the performance of the band narrowing module 6 when the characteristics of FIGS. 8, 9, and 10 are observed, the delay time Td (rising edge of the sidelight) under the above conditions (applied voltage B, concentration E, total pressure H) The delay time from the starting point of the laser pulse to the starting point of the rise of the laser pulse) is that the number of times the light reciprocates through the laser resonator is not sufficient, and the laser pulse rises before the light is sufficiently narrowed. Conceivable.

  A control method for controlling the spectral line width and the spectral purity below predetermined values will be described below.

  As described above, from the result shown in FIG. 6, the light generated after the start of discharge (after the start of sidelight emission) is extracted from the laser resonator as a laser beam (until the laser pulse rises). Since the light traveling back and forth passes through the band narrowing module 6 several times, it is considered that the band is narrowed to some extent.

Therefore, by controlling at least one of the voltage applied to the electrode 2, the F ₂ concentration in the laser gas, and the laser gas pressure, the delay time from the start point of the sidelight to the start point of the laser pulse is changed. A plurality of laser pulses having different values are obtained. Then, the temporal transition of the spectral line width at that time is measured.

  As apparent from FIG. 6, under the same narrowband module performance and the same laser resonator length conditions, the temporal transitions of the spectral line widths of the plurality of laser pulses (the starting point of the sidelight rising is the origin). ) Are on the same curve.

  That is, the temporal transition characteristic of the spectral line width with the origin of the sidelight rise as shown in FIG. 19 is obtained, and the spectral line width of the light reciprocating in the laser resonator before the laser pulse rises. Transition characteristics are estimated. The temporal transition characteristics of the spectral line width with the starting point of the sidelight rise obtained in this way as the origin are the performance of the narrowband module of the laser device used at that time, the laser resonator length, and the long pulse. This is considered to be a characteristic determined from the laser pulse width.

  It is determined whether or not the value of the spectral line width is equal to or smaller than a predetermined value from the temporal transition characteristic of the spectral line width with the origin of the sidelight rising as shown in FIG. 19 and the sidelight waveform and laser pulse waveform.

In the temporal transition characteristic of the spectral line width with the origin as the starting point of the sidelight rising as previously shown in FIG. 19, the time T _bw when the spectral line width becomes 0.2 pm is obtained (FIG. 20). . In general, when ASE does not occur, the starting point of the laser pulse waveform occurs near the time of the first peak of the sidelight. Therefore, by comparing the time T _bw and the first peak generation time T _{SL1 of the sidelight} , the light reciprocating through the laser resonator before generation of the laser pulse is narrowed to a predetermined value (for example, 0.2 pm or less). Determine whether it has been.

That is, when the spectral line width is compared with the first peak time T _SL1 of time T _bw and side lights becomes 0.2pm, T _SL1 <T _bw if light before the laser pulse generating reciprocating laser resonator zero. The band is not narrowed to 2 pm or less, and when T _SL1 ≧ T _bw , the light reciprocating through the laser resonator before the generation of the laser pulse is narrowed to 0.2 pm or less.

For example, in FIG. 20, in the case of _{sidelight 1} where T _SL1 <T _bw , the light reciprocating through the laser resonator before laser pulse generation is not narrowed to 0.2 pm or less, and T _{SL1 ′} ≧ T _{In the case of the sidelight} 2 having _bw , the light traveling back and forth through the laser resonator is narrowed to 0.2 pm or less before the generation of the laser pulse.

According to the laser device having a narrowband module, laser resonator length, and laser pulse width designed and set to achieve a narrowband of 0.2 pm or less used in the experiments by the present inventors, the side When the laser pulse was generated after the time from the light starting point to the time point T _bw , ASE did not occur.

Based on the above knowledge, the light that reciprocates the laser resonator before the generation of the laser pulse is narrowed to 0.2 pm or less before the laser pulse is generated. As a result, a laser beam having a spectral line width of 0.2 pm or less is obtained. Obtained.
(1) Voltage control FIG. 21 shows a flowchart for controlling the spectral line width to 0.2 pm or less by voltage control in this case.

First, the temporal transition of the spectral line width with the origin of the sidelight as the origin is obtained by the above-described procedure in advance, and the time T _{bw at} which the spectral line width becomes a predetermined value (for example, 0.2 pm) is obtained. The controller 22 stores the time T _bw in advance. Next, the controller 22 outputs a command to the power supply 3 to generate laser discharge, and starts laser oscillation (step S301).

  Next, the sidelight detector 24 detects the temporal pulse waveform of the sidelight (step S302), and sends the detection data to the controller 22 (step S303).

The controller 22 calculates the first peak time T _SL1 of the sidelight from the received waveform data with the start point of the side light as the origin (step S304), and compares the magnitudes of T _SL1 and T _bw (step S305). .

When T _SL1 ≧ T _bw , it is determined that the light traveling back and forth in the laser resonator is narrowed to 0.2 pm or less before the laser pulse is generated. On the other hand, when T _SL1 <T _bw , the controller 22 outputs a command to the power source 3 to lower the voltage applied to the electrode 2 by a predetermined value (step S306).

Hereinafter, until T _SL1 ≧ T _bw, repeats step S301~ step S306.
(2) F ₂ concentration and total pressure control In the voltage control of (1) above, the voltage applied to the electrode 2 is controlled so that the light reciprocating through the laser resonator before the laser pulse rises is a predetermined value (0.2 pm). However, the F ₂ concentration in the laser gas and the pressure of the laser gas may be controlled.

When controlling the F ₂ concentration, steps S301 to S305 in FIG. 22 (same as steps S301 to S305 in FIG. 21) are performed as in the voltage control. When T _SL1 <T _bw in step S305, the controller 22 observes the pressure data from the pressure gauge 25, opens the valve V3, exhausts the laser gas in the laser chamber 1, and closes the valve V3 when the laser gas pressure reaches a predetermined pressure (step S307).

  Thereafter, the controller 22 opens the valve V1 while observing the pressure data from the pressure gauge 25 (the valve V2 is closed at this time), fills the laser chamber 1 with the buffer gas, and when the laser gas pressure reaches a predetermined pressure, The valve V1 is closed (step S308).

That is, the laser gas is exhausted by a predetermined amount and filled with the same amount of buffer gas, thereby reducing the F ₂ concentration in the laser gas.

Hereinafter, until T _SL1 ≧ T _bw, repeats step S301~ step S308.

  If the pressure of the laser gas is controlled so that the light traveling back and forth through the laser resonator has a predetermined value (0.2 pm or less) before the laser pulse rises, step 308 may be omitted.

Whether or not the light traveling back and forth through the laser resonator has fallen below the predetermined value (0.2 pm or less) before the laser pulse rises is determined in advance by the first peak time T _SL of the sidelight as described above. It was found that the temporal transition characteristic of the spectral line width with the starting point of the rising sidelight as the origin is possible by the comparison result with the time T _bw when the spectral line width becomes 0.2 pm. As shown in FIG. 4, the determination may be made based on the comparison result between the starting point T _LPS of the rising edge of the laser pulse and the time point T _bw .

That is, when the time point T _bw when the spectral line width becomes 0.2 pm and the start point T _LPS of the rise of the laser pulse are compared, if T _LPS ≦ T _bw, the light reciprocating through the laser resonator is 0. The band is not narrowed to 2 pm or less, and when T _LPS > T _bw , the light reciprocating through the laser resonator before the generation of the laser pulse is narrowed to 0.2 pm or less.

In FIG. 23, when the laser pulse 1 satisfies T _LPS1 <T _bw , the light reciprocating through the laser resonator before the generation of the laser pulse is not narrowed to 0.2 pm or less, and T _LPS2 > T _bw In the case of the laser pulse 2, the light traveling back and forth through the laser resonator is narrowed to 0.2 pm or less before the laser pulse is generated.

Here, as described above, the starting point T _LPS of the laser pulse waveform is a time point before the occurrence of the first peak of the laser pulse waveform and 5% of the intensity of the first peak. An arbitrary time point before the first peak of the pulse waveform and less than 5% of the intensity of the first peak may be set as the starting point T _LPS of the laser pulse waveform.

According to the laser device having a narrowband module, laser resonator length, and laser pulse width designed and set to achieve a narrowband of 0.2 pm or less used in the experiments by the present inventors, the side When the laser pulse was generated after the time from the light starting point to the time point T _bw , ASE did not occur.

Based on the above knowledge, the light that travels back and forth through the laser resonator before the generation of the laser pulse is narrowed to 0.2 pm or less before the laser pulse is generated. As a result, a laser beam having a spectral line width of 0.2 pm or less is obtained. It was.
(1) Voltage control FIG. 24 shows a flowchart for controlling the spectral line width to 0.2 pm or less by voltage control in this case.

First, the temporal transition of the spectral line width with the origin of the sidelight as the origin is obtained by the above-described procedure in advance, and the time T _{bw at} which the spectral line width becomes a predetermined value (for example, 0.2 pm) is obtained. The controller 22 stores the time T _bw in advance. Next, the controller 22 outputs a command to the power supply 3 to generate laser discharge, and starts laser oscillation (step S401).

  Next, the laser pulse detector 21 detects a temporal pulse waveform of the laser pulse (step S402), and sends detection data to the controller 22 (step S403).

The controller 22 calculates the starting point T _LPS of the rising edge of the laser pulse from the received waveform data with the starting point of the laser pulse as the origin (step S404), and compares the magnitudes of T _LPS and T _bw (step S405).

When T _LPS ≧ T _bw , it is determined that the light traveling back and forth through the laser resonator is narrowed to 0.2 pm or less before the generation of the laser pulse. On the other hand, when T _LPS <T _bw , the controller 22 outputs a command to the power source 3 to lower the voltage applied to the electrode 2 by a predetermined value (step S406).

Thereafter, steps S401 to S406 are repeated until T _LPS ≧ T _bw .
(2) F ₂ concentration and total pressure control In the voltage control of (1) above, the voltage applied to the electrode 2 is controlled so that the light reciprocating through the laser resonator before the laser pulse rises is a predetermined value (0.2 pm). However, the F ₂ concentration in the laser gas and the pressure of the laser gas may be controlled.

When controlling the F ₂ concentration, steps S401 to S405 in FIG. 25 (same as steps S401 to S405 in FIG. 24) are performed as in the voltage control, and when T _LPS <T _bw in step S405, The controller 22 opens the valve V3 while observing the pressure data from the pressure gauge 25 to exhaust the laser gas in the laser chamber 1, and closes the valve V3 when the laser gas pressure reaches a predetermined pressure (step S407).

  Thereafter, the controller 22 opens the valve V1 while observing the pressure data from the pressure gauge 25 (the valve V2 is closed at this time), fills the laser chamber 1 with the buffer gas, and when the laser gas pressure reaches a predetermined pressure, The valve V1 is closed (step S408).

That is, the laser gas is exhausted by a predetermined amount and filled with the same amount of buffer gas, thereby reducing the F ₂ concentration in the laser gas.

Thereafter, steps S401 to S408 are repeated until T _LPS ≧ T _bw .

  Note that when the pressure of the laser gas is controlled so that the light traveling back and forth through the laser resonator has a predetermined value (0.2 pm or less) before the laser pulse rises, step 408 may be omitted.

As described above, in the laser apparatus to be used, the temporal transition of the spectral line width with the origin of the sidelight depending on the band narrowing module, the laser resonator length, etc. as the origin is obtained in advance and reciprocates in the laser resonator. The time T _{bw at} which the light has a predetermined spectral line width is obtained and is compared with the detected first sidelight peak time T _SL or laser pulse rise time T _LPS in the laser resonator before the laser pulse rises. It is possible to make the light traveling back and forth have a predetermined spectral line width.

In the example described above, any one of the voltage applied to the electrode 2, the F ₂ gas concentration in the laser gas in the laser chamber 1 and the laser gas pressure in the laser chamber 1 is used as the control parameter. It is also possible to perform control by combining a plurality of the above parameters.

  In the above example, the temporal transition of the spectral line width with the origin of the sidelight as the origin is obtained in advance, and the light that reciprocates in the laser resonator before the laser pulse rises using the spectral line width is used as a predetermined spectral line. Although it was controlled so as to have a width, the temporal transition of the spectral purity with the origin of the sidelight as the origin was obtained in advance, and this was used to reciprocate within the laser resonator before the laser pulse rose in the same procedure as above. It is also possible to control the light to be a predetermined spectral purity.

  Furthermore, the temporal transition of the spectral line width and spectral purity with the origin of the sidelight as the origin is obtained in advance, and the light that reciprocates in the laser resonator before the laser pulse rises in the same procedure as described above has a predetermined spectrum. It is also possible to control the line width and spectral purity.

  In the above-described spectral line width and spectral purity control, when the narrow band narrowing module or the current narrow band narrowing module for realizing the ultra narrow band with the spectral line width of 0.2 pm or less designed by the present inventors is used. When the laser pulse was generated after the above time, ASE was not generated.

  Here, if the suppression of ASE generation, spectral line width, and spectral purity are reliably controlled, the ASE control (FIGS. 14, 15, 17, and 18), spectral line width, and spectral purity shown above. Control (FIGS. 21, 22, 24, and 25) may be combined.

A combination example is shown below. As an example, an example is shown in which the maximum slope ΔT _s of the sidelight rise is obtained to suppress ASE (FIG. 17), and the starting point T _LPS of the rise of the laser pulse is obtained to control the spectral line width (FIG. 24). FIG. 26 is a flowchart for performing the control.

As described above, it has been found that generation of ASE occurs when the maximum gradient ΔT _s of the sidelight rising exceeds a predetermined value ΔT ₁ . The predetermined value ΔT ₁ depends on the performance of the narrowband module 6 (for example, the enlargement ratio of the enlargement prism 7 and the blaze angle of the diffraction grating 8) and the laser resonator length. The controller 22 stores in advance a predetermined value ΔT ₁ obtained based on the dependence condition of the laser device.

Further, the temporal transition of the spectral line width with the origin of the sidelight as the origin is obtained in the above-described procedure, and the time T _{bw at} which the spectral line width becomes a predetermined value (for example, 0.2 pm) is obtained. The controller 22 stores the time T _bw in advance. Then, the controller 22 outputs a command to the power supply 3 to generate laser discharge, and starts laser oscillation (step S501).

  Next, the sidelight detector 24 detects the temporal pulse waveform of the sidelight (step S502), and sends the detection data to the controller 22 (step S503).

Next, the controller 22, from the received waveform data to calculate the maximum gradient [Delta] T _s range sidelight rise of from the origin of the sidelight to the first peak (step S504), the magnitude of [Delta] T _s and [Delta] T ₁ Compare (step S505).

When ΔT _s ≦ ΔT _1, it is determined that no ASE has occurred. On the other hand, when ΔT _s > ΔT ₁ , as shown in FIG. 8, ASE does not occur when the voltage applied to the electrode 2 is equal to or lower than a predetermined value, so the controller 22 outputs a command to the power source 3 to the electrode 2. The applied voltage is reduced by a predetermined value (step S506).

  Thereafter, steps S501 to S506 are repeated until no ASE occurs.

  If it is confirmed that ASE has not occurred, the laser pulse detector 21 detects the temporal pulse waveform of the laser pulse (step S507), and sends the detection data to the controller 22 (step S508).

The controller 22 calculates the starting point T _LPS of the rising edge of the laser pulse from the received waveform data with the starting point of the laser pulse as the origin (step S509), and compares the magnitudes of T _LPS and T _bw (step S510).

When T _LPS ≧ T _bw , it is determined that the light traveling back and forth through the laser resonator is narrowed to 0.2 pm or less before the generation of the laser pulse. On the other hand, when T _LPS <T _bw , the controller 22 outputs a command to the power supply 3 to lower the voltage applied to the electrode 2 by a predetermined value (step S511).

Thereafter, step S508 to step S511 are repeated until T _LPS ≧ T _bw is satisfied.

  In the example described above, the spectral line width is controlled after the ASE suppression control is performed first. However, the present invention is not limited to this, and the ASE suppression control is performed after the spectral line width control is performed first. May be performed.

In the ASE control portion, the ASE determination is performed using the maximum gradient of the sidelight rising. However, the present invention is not limited to this, and as described above, the first peak time _{TSL1 of the sidelight} and the laser pulse The determination of ASE may be performed using the first peak time T _LP1 (FIGS. 14 and 15) or the starting point T _{LPS of the} laser pulse.

Further, in the above-described example, the spectral line width is controlled by detecting the start point T _LPS of the rise time of the laser pulse. However, as described above, the first peak time T _{SL1 of the sidelight} (FIGS. 21 and 22). ) May be detected.

  In the above example, the ASE control and the spectral line width control are performed. However, the spectral purity control may be performed instead of the spectral line width control, and both the spectral line width and the spectral purity control are performed. You may make it perform.

As described above, the F ₂ laser device to which the present invention is applied has been described as an example. However, the present invention can also be applied to a KrF laser device and an ArF laser device. For example, the spectral line width may be controlled so as to be equal to or smaller than the spectral line width required from the exposure apparatus to the KrF laser apparatus and ArF laser apparatus.

  In the configuration of the laser apparatus of FIG. 11 (FIG. 1), the band narrowing module 6 includes a grating (diffraction grating) 8 and a magnifying prism 7 so that the spectrum is narrowed. As the band narrowing module 6, a configuration using an etalon other than such a grating-prism system can be used. When the etalon is used, the oscillation center wavelength can be changed by the rotation of the etalon or the change in gas pressure between the etalon gaps (change in the refractive index of the gas).

By the way, the average output required for the F ₂ laser device as the light source for semiconductor exposure is, for example, 20 W. That is, when the repetition frequency of the F ₂ laser device is 2 kHz, the pulse energy per pulse is 10 mJ, and when the repetition frequency is 4 kHz, the pulse energy per pulse is 5 mJ.

  If the energy injected into the laser gas is increased by the discharge between the electrodes, the rise of the laser pulse becomes faster and ASE also appears, so it is difficult to narrow the band with a pulse energy of 5 to 10 mJ. It is.

  In order to obtain a laser beam having a spectral line width of 0.2 pm or less and a pulse energy of 5 mJ or more under the circumstances as described above, for example, a two-stage laser system including an oscillation stage laser and an amplification stage laser is employed. do it. That is, a laser beam having a low output but a spectral line width of 0.2 pm or less is generated by an oscillation stage laser, the laser beam is amplified by an amplification stage laser, and the spectral line width is 0.2 pm or less. A laser beam having a pulse energy of 5 mJ or more can be obtained.

  Configuration examples of the two-stage laser system include injection locking and MOPA (Master Oscillator Power Amplifier). The former is a configuration in which the amplification stage laser is provided with a laser resonator, and the latter is a configuration in which the amplification stage laser is not provided with a laser resonator.

  Here, injection locking will be described with reference to FIG. 27 as a two-stage laser system. As described above, in the injection locking system, the oscillation stage laser 30 and the amplification stage laser 40 are used. The oscillation stage (oscillator) laser 30 has a function as a seed laser (seed laser beam) of the laser system. The amplification stage (amplifier) laser 40 has a function of amplifying the seed laser.

  That is, the overall spectral characteristics of the laser system are determined by the spectral characteristics of the oscillation stage laser 30. Then, the laser output (energy or power) from the laser system is determined by the amplification stage laser 40.

  Therefore, for example, as shown in FIG. 27, the oscillation stage laser 30 is equipped with a narrowband module 6 composed of a grating (diffraction grating) 8 and a magnifying prism 7, and a spectrum narrowing laser is provided. Output from the oscillation stage laser 30. In the case of FIG. 27, apertures 31 for limiting the laser beam of the oscillation stage laser 30 are provided in the resonators on both sides of the laser chamber 1 of the oscillation stage laser 30.

  A laser beam (seed laser beam) from the oscillation stage laser 30 is guided and injected into the amplification stage laser 40 by a beam propagation system including a reflection mirror 32 and the like. In the injection lock method, an unstable resonator having a concave mirror 42 and a convex mirror 43 and having a magnification of 3 times or more, for example, is employed for the amplification stage laser 40 so that it can be amplified even with a small input.

  The concave mirror 42 of the unstable resonator of the amplification stage laser 40 has a hole, and the seed laser beam introduced through this hole is reflected and expanded by the convex mirror 43 as indicated by an arrow, and the laser chamber 41 discharges. The laser beam power is increased by effectively passing through the section. Then, a laser is emitted from the convex mirror 43. A spatial hole is provided at the center of the concave mirror 42, and a high reflectivity mirror coat is provided around the center. A high reflectivity mirror coat is applied to the central portion of the convex mirror 43, and an antireflection coat is applied to the surrounding laser emission portion. The hole of the concave mirror 42 is not spatially open, and a mirror substrate on which only an antireflection coating is applied may be used. Moreover, you may use the unstable resonator which does not give a transmission part to a mirror.

  When the concave mirror 42 and the convex mirror 43 are not used in the amplification stage laser 40 of FIG. 27, the present system is MOPA. That is, since the amplification stage laser 40 does not have a resonance mirror (concave mirror 42 and convex mirror 43), the amplification stage laser 40 functions as a one-pass amplifier of the oscillation stage laser 30.

  In a two-stage laser system such as an injection locking system or MOPA, the characteristics of the laser beam emitted from the amplification stage laser are affected by the oscillation stage laser light injected at the rise of the amplification stage laser light. That is, in the laser pulse from the oscillation stage laser injected into the laser gas of the amplification stage laser, the spectrum of the oscillation stage laser pulse instantaneously at the rise of the amplification stage laser beam directly affects the spectral linewidth characteristics of the two stage laser. Effect.

Based on the above-described present invention, the narrowed F ₂ laser apparatus is applied to the two-stage laser apparatus of FIG. 27 as follows, for example. That is, the oscillation stage laser 30 of FIG. 27 is used as the laser apparatus of the present invention, and the starting point T _LPS of the rise of the laser pulse of the oscillation stage laser 30 is detected. The side light of the amplification stage laser 40 rises after the rise start point (that is, discharge starts). By doing this, the spectrum of the oscillation stage laser pulse instant at the rise of the amplification stage laser light becomes 0.2 pm or less, so that the spectral linewidth characteristics of the two-stage laser are also affected, resulting in a high frequency. An output, ultra-narrow band laser device is realized.

  Note that the above-described two-stage laser system to which the present invention is applied can also be applied to a KrF laser device and an ArF laser device. The spectral line width is equal to or less than the required line width from the exposure apparatus, and the pulse energy is 5 mJ or more. The laser beam can be obtained.

  As described above, the narrow-band gas laser device of the present invention has been described based on the principle and description of the embodiments. However, the present invention is not limited to these embodiments, and various modifications are possible.

It is a figure which shows the structural example of a narrow band fluorine laser apparatus. It is a figure which shows the time transition of the sidelight of the conventional narrow band fluorine laser apparatus, the waveform of a laser pulse, and a spectral line width. It is a figure which shows the time change and component of a spectral line width within one laser pulse. It is a figure for demonstrating 95% purity. It is a figure which compares and shows the time transition (a) of the laser pulse waveform in this invention, a sidelight waveform, and a spectrum line width, and the conventional time transition (b). It is the figure which put together the time transition and laser pulse waveform of the spectral line width of Fig.5 (a) and (b). It is a figure which shows the relationship between the time from the start of sidelight emission to the occurrence of the first sidelight peak, the starting point of the laser pulse waveform including ASE, and the spectral line width of the laser pulse. It is a figure which shows the delay time from the starting point of the rise of the spectral line width and the sidelight when the applied voltage is changed to the starting point of the rise of the laser pulse. It is a figure which shows the delay time from the starting point of the rise of the spectral line width and the sidelight when the F ₂ concentration is changed to the starting point of the rise of the laser pulse. It is a figure which shows the delay time from the starting point of the rise of a spectral line width and a sidelight when the laser gas pressure is changed to the starting point of the rise of a laser pulse. It is a diagram illustrating a configuration example of a narrowed F ₂ laser system of the present invention. Laser pulse waveform when the first peak time T _LP1 of the first peak time T _SL1 and laser pulses sidelight ASE occurs at T _SL1> T _LP1, a diagram illustrating a sidelight waveform. Laser pulse waveform when the first peak time of the sidelight T _SL1 and laser pulses first peak time T _LP1 is ASE does not occur in the T _SL1 ≦ T _LP1, a diagram illustrating a sidelight waveform. It is a flowchart of one Example which suppresses generation | occurrence | production of ASE by voltage control. F ₂ concentration, a flow chart of one embodiment suppresses the generation of the ASE by the total pressure control. Is a diagram showing a maximum gradient [Delta] T _s and the laser pulse waveform rising sidelight. It is a flowchart of another Example which suppresses generation | occurrence | production of ASE by voltage control. F ₂ concentration, a flow chart of another example of suppressing the occurrence of ASE by the total pressure control. It is a figure which shows the time transition characteristic of the spectral line width which makes the origin the starting point of a sidelight. It is a figure which shows the time of a spectral line width becoming 0.2 pm in the time transition characteristic of the spectral line width which makes the origin the starting point of a sidelight. It is a figure which shows the flowchart of one Example controlled to the spectral line width of 0.2 pm or less by voltage control. F ₂ concentration is a diagram showing a flowchart of one embodiment for controlling the following spectral linewidth 0.2pm by total pressure control. Origin T _LPS and spectral line width of the rise of the laser pulse is a diagram showing the relationship of time T _bw and laser pulse waveform becomes 0.2 pm. It is a figure which shows the flowchart of another Example controlled to the spectrum line width of 0.2 pm or less by voltage control. F ₂ concentration is a diagram showing a flow chart of another embodiment for controlling the following spectral linewidth 0.2pm by total pressure control. Suppresses ASE seeking maximum gradient [Delta] T _s of rising sidelight is a diagram showing a flowchart of an embodiment for controlling the spectral linewidth seek starting point T _LPS of the rise of the laser pulse. A configuration example of a narrowed F ₂ laser system of the injection locking method of applying the control method according to the present invention. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Laser chamber 2 ... Electrode 3 ... High voltage pulse generator (power supply)
4 ... Fan 5 ... Window 6 ... Narrow band module 7 ... Magnifying prism 8 ... Diffraction grating (grating)
DESCRIPTION OF SYMBOLS 9 ... Output mirror 10 ... Beam splitter 11 ... Wavelength monitor 12 ... Controller 21 ... Laser pulse detector 22 ... Controller 23 ... Observation window part 24 ... Sidelight detector 25 ... Pressure gauge 26 ... Gas supply line 27 ... Exhaust line 30 ... Oscillation stage laser 31 ... Aperture 32 ... Reflection mirror 40 ... Amplification stage laser 41 ... Laser chamber 42 ... Concave mirror 43 ... Convex mirror V1, V2, V3 ... Valve

Claims (3)

A narrow-band gas laser having a discharge electrode in a laser chamber in which a laser gas containing F ₂ is sealed and emitting a laser beam narrowed by a narrow-band module including a wavelength selection element installed in the laser resonator In the device
At the time of generation of the laser beam emitted from the laser resonator, the laser beam is emitted from the time when light is emitted by discharge so that the spectral line width (FWHM) is 0.2 pm or less and / or the spectral purity is 0.5 pm or less. A narrow-band gas laser device characterized in that a time until occurrence of is set. A narrow-band gas laser apparatus according to claim 1, wherein the narrow-band gas laser apparatus is a two-stage narrow-band gas laser apparatus comprising an oscillation stage laser and an amplification stage laser. A discharge light emission measuring device (side light detector) that measures a temporal pulse waveform of light emission due to discharge by the amplification stage laser, applied voltage to the discharge electrode of the amplification stage laser, F ₂ concentration of laser gas, and in the laser chamber And a controller capable of controlling at least one of the laser gas pressures of the laser. The controller receives timing data T _{LPS at} which a laser pulse of the oscillation stage laser rises from the controller of the oscillation stage laser, and based on this data, the controller 3. The narrow-band gas laser device according to claim 2, wherein discharge of the amplification stage laser is started after the laser pulse of the oscillation stage laser rises.

Images