JP2003511865A - Energy control of excimer or molecular fluorine laser - Google Patents

Abstract

  (57) [Summary] Gas replenishment methods and algorithms for excimer and molecular fluorine lasers are based on parameters such as the input energy applied to the electric discharge, and preferably time, over which the aging of the gas depends more closely than the pulse count. . Burst control methods and algorithms include measuring the energy of the initial pulse of the first burst that occurs after a long burst pause and the initial pulse that will bring the output energy of the individual laser pulses or pulses to approximately the same value. Calculating the value of the input voltage for the pulse and applying the calculated voltage during a subsequent first burst after a long burst pause to achieve approximately the same predetermined output energy value for the pulse or group of pulses. To do. A similar operation is performed for one or more subsequent bursts following the first burst. The values for the first burst are maintained in a first table of input voltage values and are read by a processor that signals the power supply circuit to apply a voltage according to the voltage values in the table. Values for subsequent bursts are maintained in a second table, a third table, and so on. A final table, such as the third table, is used for all subsequent bursts until another long burst pause again occurs, after which the first table is used again for the first burst following the long burst pause. You.

Description

Detailed Description of the Invention

(Priority) This application is based on US provisional patent application No. 60 / 159,5 filed on October 15, 1999.
No. 25 priority benefit, and this application claims benefit of US Provisional Patent Application No. 60 / 124,785 filed Mar. 17, 1999 Nov. 2, 1999
This is a continuation-in-part application of U.S. patent application Ser. No. 09 / 447,882 filed on March 3, and the present application is filed on Mar. 12, 1999.
U.S. patent application Ser. No. 09/41, filed October 14, 1999, claiming benefit of No. 8
No. 8,052, which is a continuation-in-part application, and this application was filed on January 18, 2000, claiming the benefit of US Provisional Patent Application No. 60 / 127,062, filed March 31,1999. It is a continuation-in-part application of US patent application Ser. No. 09 / 484,818, all of which are incorporated herein by reference. (Background of the Invention) 1. FIELD OF THE INVENTION The present invention relates to excimers and molecular fluorine gas lasers.
The energy control of ine gas lasers is particularly concerned with constant laser output radiation pulse energy and / or application process.
A) control and feedback software algorithms to keep the energy dose constant and gas replenishment.

2. 2. Discussion of Related Art The energy of the output radiation pulse of an excimer or molecular fluorine laser constantly decreases during the operation of the laser unless some input parameters or conditions are controlled. This is because the halogen is consumed by the reaction of the halogen gas in the gas container and the halogen is burned by the gas discharge. In addition, the output power is reduced due to the accumulation of gas pollution.

Excimer lasers can be operated at constant energy levels for a period of time if they continuously increase the charging voltage to compensate for these factors causing energy loss. For demanding applications such as lithography or TFT annealing, it is desirable to maintain control of the charging voltage along with other beam parameters in addition to pulse energy or laser output power. Therefore, more sophisticated treatments have been developed. When the maximum applied charging voltage is reached, a gas replenishment operation is performed, further extending the time that the laser operates at constant energy. This gas replenishment operation is performed to compensate for the consumption of halogen and reduce pollution. Halogen depletion is usually compensated by halogen injection (HI). Reduction of pollution is achieved by partial gas replacement (PGR).
t) is achieved.

Gas replenishment was introduced for excimer lasers around 1986 (see US Pat. No. 4,997,573, incorporated by reference into the description of this application). The gas refill operation is initiated when the charging voltage exceeds a preset level. Gas refueling operations have traditionally been characterized by a large drop in charging voltage. However, such large changes in charging voltage are disadvantageous as large changes in charging voltage during long constant energy operation can affect various beam parameters other than beam energy or power. In other words, large changes in the charging voltage to stabilize the output energy will act to destabilize other important beam parameters.

[0005] In order to maintain both the charging voltage and the output energy or power of the laser at a nearly constant level, a large amount of gas replenishment operation involves micro halogen injection (HI) and micro partial gas replacement. Or replaced by a small amount of gas manipulation such as mini gas replacement (GR or mGR) (US patent application Ser. No. 09/447, each assigned to the same assignee and incorporated by reference into the description of the present application. , 882 and 60 / 171,7
(See No. 17). Such micro- or mini-gas replenishment operation preferably results in little or no charge voltage variation that can be detected with sufficient accuracy under industrial operating conditions. Therefore, it is desirable to use other parameters besides the change in charging voltage to initiate the micro or mini gas refill operation. One parameter that can be used as a suitable trigger for micro or mini gas refueling operations is the number of laser pulses, or pulse count. This is disclosed in US Pat. No. 5,097,291 and subsequently US Pat. No. 5,337,215, each incorporated by reference into the description of the present application. For example, gas replenishment operation is about 1
It is performed periodically every 100,000 pulses.

For microlithography scanner systems, it is desirable to maintain a constant energy dose as the die sites on the wafer are scanned.
Scanning speed, exposure slit width, and laser repetition rate (laser
The repetition rate determines the number of pulses that are superimposed on each site on the wafer. The number of overlaid pulses depends on the application process. For example, the number of pulses that are overlapped at one die site is about 40, and the normal length of one burst is 100 to 500 pulses.

The constant energy dose for each site on the wafer is the transfer energy average (
moving energy average). Accurate dose control (dose cont
In the case of rol), the fluctuation of the transfer energy average can be observed as being low. The output energy of the laser is the high voltage (HV) used for the discharge inside the laser tube.
Can be controlled by changing. The output energy can be, and usually is, measured for each pulse, and the HV can be varied for each individual pulse.

Excimer and molecular fluorine lasers usually operate in burst mode. This is because the laser is 100-500 at a constant repetition rate as mentioned above.
A burst of pulses, such as a pulse, followed by a burst break of a few milliseconds to a few seconds during a stepper / scanner performing some type of wafer positioning. Means to occur. Burst pauses can be short burst pauses, such as those that occur when the beam spot moves to different positions on the same wafer, or long bursts, such as those that occur when a stepper / scanner swaps wafers. Sometimes it is a break.

When the excimer or molecular fluorine laser is operating in burst mode, the first few pulses of each burst have a higher pulse energy than the subsequent pulses if left uncompensated. Therefore, the moving average at the beginning of the burst will be higher than at any later time during the burst. It is desirable to compensate for this overshoot in order to achieve a constant energy dose. Overshoot compensation is achieved by reducing the charging voltage for the first few pulses (
See US Pat. Nos. 5,463,650, 5,710,787 and 6,084,897, each incorporated by reference into the description of the present application). This keeps the energy dose constant at the beginning of the burst sequence. The charging voltage is adjusted for each laser pulse and is below the value applied to the later pulses in the burst at the beginning of the burst.

The problem that the first few pulses after the burst pause (at the beginning of the burst) have a higher energy-to-HV ratio than the pulses at the middle or end of the burst is, as illustrated in the sketch of FIG. It is understood by observing what happens when the medium HV is kept constant. In the sketch of Figure 1, the first 5-10
The pulse has a high energy, after which the energy is rapid, then 20 to 1
After 00 pulses the energy decays more slowly until it reaches a certain level. This phenomenon is called overshoot or spiking.

To keep the pulse energy or energy dose constant (as desired during laser operation), a low HV is used for the first pulse of the burst, after which the HV is increased to a constant level during the burst. This is done in response to pulse energy overshoots that would otherwise occur as described immediately above.

The exact behavior of the energy is influenced by various parameters in an unpredictable way. It would be desirable to have a technique that predicts the HV for the next pulse so that the energy of the next pulse or the energy dose in the application process matches the target energy or target energy dose. RECOGNITION IN THE PRESENT INVENTION There are short-term and long-term effects that affect the energy-related behavior of pulses during and from burst to burst. Short-term effects last less than a few seconds. Long-term effects include gas aging (several days), tube aging (several months) and possibly optical effects (several years). These effects are taken into account by changing the controller parameters. Advantageously, the parameter adaptation is done automatically.

The energy behavior changes depending on the burst pause length, laser repetition rate, energy of the most recent pulse and other influences. Controlling the energy of the first few pulses of a burst is more difficult than keeping the energy or energy dose of the pulse at the middle or end of the burst constant, but this is because the gas state This is because it does not change rapidly with time over the duration. Therefore, it is desirable to have a pulse energy or energy dose control algorithm that results in high pulse energy or energy dose stability at the beginning of a burst and throughout the burst.

Gas aging depends on time and input energy to the electrical discharge. Typical time constants for time-based gas replenishment operations are several hours, for example 8 hours. Depending on the laser wavelength, the individual laser structure, operating mode and other parameters, this time constant can be as short as one hour or less, or even more than one day. State-of-the-art gassing is based on pulse counting (eg, mentioned above
See the '097 and' 215 patents). Gas replenishment algorithms based on pulse counting work very well when the energy input to the laser discharge for each laser pulse or pulse group is constant. However, it is recognized in the present invention that the input energy to the laser discharge is not constant for each laser pulse or pulse group during the period of laser operation. Especially in industrial lithographic processes, the input energy to the discharge is not constant, especially for many laser pulses or pulse groups over thousands of pulses. Therefore, since gas aging more closely depends on time and energy input to the electrical discharge rather than pulse counting, it is desirable to have a gas replenishment algorithm that bases gas replenishment operations on these parameters. SUMMARY OF THE INVENTION Accordingly, one object of the present invention is to provide a gas replenishment algorithm based on parameters such as input energy to an electrical discharge, and preferably time, over which gas aging more closely depends on pulse counting. Is to provide.

A further object of the invention is a pulse energy or energy dose control algorithm that results in high pulse energy stability or energy dose stability for the initial pulse of the burst and also preferably throughout the burst. Is to provide. In a first aspect of the present invention for the above purposes, a discharge chamber having a number of electrodes therein, containing a gas mixture containing one or more components to be consumed,
A power circuit coupled to the electrodes for energizing the gas mixture, a laser
A gas discharge laser system is provided that includes a beam producing resonator. The processor monitors the stored energy applied to the discharge of the laser as a measure of the gas mixture state, and the gas control unit serves to replenish the gas mixture based on the monitored stored energy applied to the discharge. The processor also preferably monitors time as an additional measure of the state of the gas mixture, and the gas control unit replenishes the gas mixture based on the monitored time in addition to the stored energy applied to the discharge. The charging voltage and changes in charging voltage are also monitored with the stored energy and / or time applied to the discharge, and the gas control unit may monitor the charging voltage and / or charging voltage in addition to the applied energy and / or time to the discharge. Replenish the gas mixture based on the change in

An initial composition containing one or more constituent gases to be consumed.
Also provided is a method of stabilizing a gas mixture initially provided in a discharge chamber of a gas discharge laser having a ition) during operation. The method comprises the steps of monitoring the stored energy applied to the discharge of the laser, determining the state of the gas mixture and / or adjusting the gas mixture based on the monitored stored energy applied to the discharge. Steps and are included. Time is also preferably monitored with the stored energy applied to the discharge. The charging voltage and changes in charging voltage are also monitored with the stored energy applied to the discharge and / or with time.

Therefore, according to the first aspect of the invention, it is advantageous to initiate a gas refill operation based on the stored energy applied to the discharge. The new method is more flexible and thus more powerful than, for example, the method of initiating a gas refill operation based on pulse counting. As the pulse energy changes, the new algorithm advantageously lengthens or shortens the gas refill interval accordingly, thereby improving the performance of the laser.

In a second aspect of the invention according to the above objective, there is provided a method for controlling the output energy of a continuous pulse from a gas discharge laser operating in burst mode, each of which is followed by an application process specification. A method is provided, characterized in that it emits a burst of several pulses, which is followed by one of a long burst pause and a short burst pause. The method comprises measuring the energy of at least a predetermined number of initial pulses of a first burst occurring after a long burst pause.
Calculating the value of the input voltage for each initial pulse that will bring the corresponding output energy of the initial pulse to approximately the same predetermined value for the next first burst following a similar long burst pause. Applying an input voltage corresponding to the calculated value in the next first burst after a similar long burst pause so that the pulses have approximately the same predetermined output energy value.

The method further comprises measuring the energy of at least a predetermined number of initial pulses of at least one second burst that results in a short burst pause after the first burst following the long burst pause, and a similar long burst. A second burst that will bring the corresponding output energy of the initial pulse of the second burst to about the same predetermined value for the next second burst following a similar short burst pause after the first burst following the pause. Calculating the value of the input voltage for each initial pulse of the first burst of the first burst after the similar long burst pause so that the generated pulses have approximately the same predetermined output energy value. Mark the input voltage corresponding to the value calculated for the initial value in the next second burst following a similar short burst pause as described above. Adding and may be included.

The method further comprises measuring the energy of at least a predetermined number of initial pulses of at least one third or subsequent burst occurring in at least two short burst pauses after a long burst pause. A third burst that will bring the corresponding output energy of the initial pulse of the third burst to about the same predetermined value for one or more subsequent bursts occurring after at least two short burst pauses following the burst pause. Calculating the value of the input voltage for each of the initial pulses of the, and one or more after two short burst pauses followed by a long burst pause so that the generated pulses each have approximately the same predetermined output energy value. Applying an input voltage corresponding to the calculated value for the next burst above, and May be.

The method further comprises for the next burst after two short burst pauses followed by a long burst pause to generate additional bursts each having an initial pulse having approximately the same predetermined value. Repetitive application of an input voltage corresponding to the calculated value may be included. The method also further comprises measuring the output laser energy corresponding to the initial pulse plus subsequent pulses in a burst, and the output energy dose of the laser corresponding to the sum of the pulse energy of successive pulses. And calculating the value of the input voltage corresponding to each pulse after these that will each lead to approximately the same predetermined energy dose value, and the energy doses associated with the pulses produced are each approximately Applying an input voltage corresponding to the calculated value to bring the output energy doses to approximately the same predetermined value so as to have the same predetermined energy dose value.

Further in accordance with a second aspect of the invention, an algorithm for controlling the output energy of successive pulses during a burst of pulses from a gas discharge laser operating in burst mode, after each An energy control software algorithm is provided, characterized in that it emits a burst of several pulses followed by one of a long burst pause and a short burst pause, depending on the specifications of the application process. This algorithm sends a signal to the power supply circuit to apply a voltage according to the voltage value in the first table, whereby the next first burst of output laser pulses after a long burst pause, each having approximately the same energy value. A first table of input voltage values to be read by a processor that generates an initial pulse therein is provided. The input voltage values in the first table are calculated from the measured data of the initial pulse in the previous first burst after a long burst pause. The input voltage values are used to generate the initial pulse, each with approximately the same energy value.

The algorithm preferably further signals the power supply circuit in a similar manner to apply a voltage according to the voltage value in the second table, whereby a short burst following a long burst pause followed by a short burst. A second table of input voltage values to be read by the processor to generate an initial pulse in the next second burst of output laser pulses occurring after a burst pause is provided. This algorithm also signals the power supply circuit in a similar manner to apply a voltage according to the voltage value in the third table, whereby a long burst pause followed by a short burst pause following the first and second bursts. It provides a third table of input voltage values to be read by the processor to generate an initial pulse in a subsequent third or subsequent burst that occurs later.

Further in accordance with a second aspect of the invention there is provided an algorithm for controlling the output energy of successive pulses during a burst of pulses from a gas discharge laser operating in burst mode, after each An energy control software algorithm is provided, characterized in that it emits a burst of several pulses followed by one of a long burst pause and a short burst pause, depending on the specifications of the application process. This algorithm sends a signal to the power supply circuit to apply a voltage according to the voltage values in the first table, whereby the next first burst of output laser pulses after a long burst pause, each having approximately the same energy value. 3 provides a table of input voltage values to be read by a processor to generate an initial pulse therein. The input voltage values in the first table are calculated from the measured data of the initial pulse in at least one previous first burst following a long burst pause. The input voltage values are used to generate the initial pulse, each with approximately the same energy value.

This table measures the energy of the initial pulse of the first burst following a long burst pause and the laser output energy corresponding to each initial pulse to the next first burst following a similar long burst pause. Calculating a value of the input voltage for the initial pulse based on the measured initial pulse energy that will lead to approximately the same predetermined value, and being generated according to the input voltage value stored in the first table. So that the pulses have approximately the same predetermined output energy value,
Storing the calculated values of the input voltage for the initial pulse as a table.

The algorithm preferably further comprises the further step that the table further measures the energy of the initial pulse of the next first burst following the next long burst pause, and the output energy of the laser corresponding to each initial pulse. The input voltage for the initial pulse based on the energy of the measured initial pulse of the next first burst that will lead to approximately the same predetermined value for another next first burst following another next long burst pause. And the measured initial pulse energy of the next first burst so that the pulses generated according to the input voltage values stored in the first table have approximately the same predetermined output energy values. Input voltage for the initial pulse in the first table using the calculated input voltage for the initial pulse based on And updating the values in the table of

Preferably, in the second aspect of the invention, the pulse group after the initial pulse or all energy doses of the pulse group are kept constant. This means that the energy of n pulses, eg the sum of the energies of n = 40 pulses, is kept constant for each package of n subsequent pulses. This sum divided by the number of pulses in it is called the moving average. INCORPORATION BY REFERENCE The following is in addition to the references cited in the Priority section above as a disclosure of alternative embodiments of the preferred embodiment elements or features not otherwise detailed below. And each is a citation list of references, each incorporated by reference into the detailed description of the preferred embodiments of the present application below. One or a combination of two or more of these references may be considered in order to obtain variations of the preferred embodiment described in the detailed description below. Additional patents, patent applications, and non-patent references are cited in the written description and are incorporated by reference in the description of the preferred embodiments with the same effect as described with respect to the following references.

US Pat. Nos. 4,997,573, 5,337,215 and 5,097,
No. 291, No. 5,140,600, No. 5,450,436, No. 4,674.
099, 5,463,650, 5,710,787, 6,084,
897, U.S. Patent Application Nos. 60 / 123,928, 60 / 124,785, 60
/ 137,907, 09 / 379,034, 09 / 447,882,
09 / 379,034, 09 / 484,818, 09 / 418,05
2, No. 09 / 513,025, No. 60 / 171,717, which are each assigned to the same assignee as the present application, Japanese Patent Application Publication No. 63-86593. Detailed Description of the Preferred Embodiments Gas Discharge Laser System Referring to FIG. 2, for example, ArF or KrF excimers for deep ultraviolet (DUV) or vacuum ultraviolet (VUV) lithography, for example. A schematic of a gas discharge laser system, such as a laser, or molecular fluorine (F ₂ ) laser system, preferably a DUV or VUV laser system, is shown. Alternative configurations of laser systems used in other industrial applications, such as TFT annealing and / or micromachining, are similar to and / or modified from the system shown in FIG. 2 to meet the needs of that application. It will be understood by those skilled in the art as being done. Alternative DUV or VUV laser system and component configurations for this purpose are described in US patent application Ser. Nos. 09 / 317,695, 09 / 317,526, 09/1.
No. 30,277, No. 09 / 244,554, No. 09 / 452,353, No. 0
9 / 317,527, 09 / 343,333, 60 / 122,145, 60 / 140,531, 60 / 162,735, 60 / 166,9
52, 60 / 171,172, 60 / 141,678, 60/17
3,993, 60 / 166,967, 60 / 147,219, 60
/ 170,342, 60 / 162,735, 60 / 178,445,
60 / 166,277, 60 / 167,835, 60 / 171,91
No. 9, 60 / 202,564, 60 / 204,095, 60/172
, 674, 09 / 574,921 and 60 / 181,156, and US Pat. Nos. 6,005,880, 6,061,382, 6,020,72.
3, No. 5,946,337, No. 6,014,206, No. 5,559,81
No. 6, No. 4,611,270, No. 5,761,236,
Each of which is assigned to the same assignee as the present patent application and incorporated herein by reference.

The system shown in FIG. 2 generally includes a solid pulser module.
e) Includes a laser chamber 2 having a pair of main discharge electrodes 3 connected to a 4 and a gas handling module 6. Solid pulser module 4
Is powered by a high voltage power supply 8. The laser chamber 2 has a resonator (re
It is surrounded by an optical module 10 and an optical module 12 forming a sonator). The optics modules 10 and 12 are controlled by the optics control module 14, or alternatively are directly controlled by the computer 16.

The laser control computer 16 receives various inputs and controls various operating parameters of the system. The diagnostic module 18 is preferably via optics, such as a beam splitter module 22 as shown, for deflecting a small portion of the beam in the direction of the module 18. One or more parameters of the split off portion of the main beam 20 are received and measured. Beam 20 is preferably an imaging system.
stem) (not shown) and finally a workpiece (also not shown)
Laser output to and can be output directly to the application process. Laser control computer 16 communicates with stepper / scanner computer 26 and other control units 28 through interface 24. Laser Chamber The laser chamber 2 contains a laser gas mixture and includes a pair of main discharge electrodes 3 plus one or more preionization electrodes (not shown). A suitable main electrode 3 is US patent application Ser. Nos. 09 / 453,670, 60/60/60/60 each of which is assigned to the same assignee as the present application and incorporated herein by reference.
184,705 and 60 / 128,227. Other electrode configurations are US Pat. Nos. 5,729,565 and 4,4, each assigned to the same assignee.
No. 4,691,322, 5,535,233 and 5,557,629, all of which are incorporated herein by reference. It is described in. A preferred preionization unit is US Patent Application No. 60 / 162,845, each of which is assigned to the same assignee as the present patent application.
60 / 160,182, 60 / 127,237, 09 / 535,27
6 and 09 / 247,887, and alternative embodiments are U.S. Pat. Nos. 5,337,330, 5,818,865 and 5,991,32.
No. 4, but all of the above patents and patent applications are incorporated herein by reference. (Power Supply Circuit and Pulser Module) The solid-state pulser module 14 and the high-voltage power supply 8 are compressed electric pulses (comp).
The electrical energy in the reseded electrical pulse is supplied to the preionization and main electrode 3 in the laser chamber 2 to impart energy to the gas mixture. Suitable pulser modules and high voltage power supply components are each assigned to the same assignee as the present application and are incorporated by reference into US patent application Ser. No. 60 / 149,3.
92, 60 / 198,058, 60 / 204,095, 09/43
No. 2,348 and No. 09 / 390,146, and No. 60 / 204,095,
And US Pat. Nos. 6,005,880 and 6,020,723. Other alternative pulser modules are disclosed in US Pat. Nos. 5,982,800, 5,982,795, 5,94, each of which is incorporated herein by reference.
0,421, 5,914,974, 5,949,806, 5,93
6,988, 6,028,872 and 5,729,562. Conventional pulser modules can generate electrical pulses with powers in excess of 3 Joules (see the '988 patent referenced above). (Laser Cavity) The laser cavity surrounding the laser chamber 2 containing the laser gas mixture includes:
Line narrowing optics for narrow linewidth excimer or molecular fluorine lasers (line-narrowi
optics module 10 including ng optics, which may be undesired narrowing, or where narrowing is done in the front optics module 12, or a spectrum outside the resonator. • Filters are used, or the line narrowing optics are high reflect mirrors to reduce the line width of the output beam.
a high reflectivity mirror in the laser system (if placed in front of the
HR mirror) and so on. The laser chamber 2 is enclosed by a window transparent to the wavelength of the emitted laser radiation 14. This window may be a Brewster window and may be aligned at another angle to the optical path of the resonant beam, eg 5 °. One of the windows can also serve to output couple the beams. Features Other Than the Resonator After a portion of the output beam 20 has passed through the outcoupler of the optical module 12, the output portion either deflects a portion of the beam towards the diagnostic module 18 or separates it. Beam splitter module 2 including optics that allows a small portion of the beam that has been outcoupled in the manner described above to reach the diagnostic module 18.
2 while the main beam portion 20 is made sustainable as the output beam 20 of the laser system. Suitable optics include beamsplitters or other methods of partially reflecting surface optics. The optical device may also include a mirror or beam splitter as the second reflective optical device.
More than one beam splitter and / or HR mirror, and / or dichroic mirror may be used to direct a portion of the beam to components of diagnostic module 18. Holographic beam sampler
am sampler), transmission diffraction grating, partially transparent diffraction grating,
Grism, prism or other refraction type, dispersive and / or
Alternatively, a transmissive optical device alone or multiple devices may be used to separate a small beam portion from the main beam 20 for detection at the diagnostic module 18, while a majority of the main beam 20 is being used. Directly or via the imaging system
Or otherwise allowed to reach the application process.

The output beam 20 can be transmitted through the beam splitter module 21, while the reflected beam portion is directed to the diagnostic module 18, or the main beam 2
The 0 can be reflected while a small portion is transmitted towards the diagnostic module 18. The part of the outcoupled beam that persists past the beam splitter module 21 is the output beam 20 of the laser, which is an industrial or experimental application such as an imaging system and workpiece in the case of photolithography applications. Propagate to. In particular, variants of the beam splitter module for molecular fluorine laser systems are each assigned to the same assignee as the present application and are incorporated by reference in US patent application Ser. No. 09 / 598,552 and 60/1
40,530. (Beam Path Enclosure) In particular, in the case of a molecular fluorine laser system and an ArF laser system, an enclosure (not shown) is used in order to keep the beam path from being affected by photoabsorbing species. Can seal the beam path of the beam 20. A small enclosure can seal the beam path between the chamber 2 and the optical modules 10 and 12, and between the beam splitter 22 and the diagnostic module. Suitable enclosures are each of 09 / 343,3, each of which is assigned to the same assignee as the present application and incorporated herein by reference.
No. 33, No. 09 / 598,552, No. 09 / 594,892, No. 09/13
Nos. 1,580 and 60 / 140,530 and U.S. Pat. Nos. 5,559,584, 5,221,8, which are incorporated herein by reference in their entirety.
No. 23, No. 5,763,855, No. 5,811,753 and No. 4,616.
No. 908 for further details. Diagnostic Module The diagnostic module 18 preferably includes at least one energy detector. This detector measures the total energy of the beam portion which corresponds directly to the energy of the output beam 20 (US Pat. No. 4,611, each assigned to the same assignee of the present application and incorporated herein by reference). 270 and US Patent Application No. 09 /
379, 034). An optical arrangement, such as an optical attenuator such as a plate or coating, or other optical device is formed on or near the detector or beam splitter module 21 to impinge on the detector. Control the intensity, spectral distribution and / or other parameters of the radiation (US patent application Ser. Nos. 09 / 172,805, 60, each assigned to the same assignee of the present application and incorporated herein by reference. / 172,749, No. 6
0 / 166,952 and 60 / 178,620).

One of the other components of diagnostic module 18 is preferably a monitor etalon.
etalon) or wavelength and / or bandwidth detection components, such as a grating spectrometer (US patent application Ser. Nos. 09 / 416,344, 60 / each assigned to the same assignee as this application). No. 186,003, 60 / 158,80
No. 8, No. 60 / 186,096, No. 60 / 186,096 and No. 60/18
6,096 and 60 / 202,564 and U.S. Pat. No. 4,905,243.
No. 5,978,391, 5,450,207, 4,926,428
Nos. 5,748,346, 5,025,445, and 5,978,3.
94, all of the above wavelength and / or bandwidth detection and monitoring components are incorporated herein by reference).

Other components of the diagnostic module include US patent application Ser. Nos. 09 / 484,818 and 09/418, each of which is assigned to the same assignee as the present application and incorporated herein by reference. A pulse shape detector or an ASE detector, such as those respectively described in No. 052, may be included to provide gas control and / or output beam energy stabilization, etc. , And as described in more detail below, the amplified spontaneous emission in the beam (am
Monitor the amount of plunged spontaneous emission (ASE) to ensure that ASE stays below a certain level. For example, there is a beam alignment monitor as described in US Pat. No. 6,014,206, which is assigned to the same assignee as the present application and incorporated herein by reference. You can also Control Processor The processor or control computer 16, along with other input or output parameters of the laser system and output beam, pulse shape, energy, ASE, energy stability, energy overshoot in burst mode operation, wavelength. , And receives and processes some values in spectral purity and / or bandwidth. Processor 16 also includes a line narrowing module (line n
arrowing module) to tune wavelength and / or bandwidth or spectral purity, and power supply and pulser modules 4 and 8 to preferably control moving average pulse power or energy, and The energy irradiation amount at each point of the work piece is stabilized near a desired value. In addition, the computer 16 controls the gas treatment module 6 including gas supply valves connected to various gas sources. Further functions of processor 16 are described in further detail below. Gas Mixture The laser gas mixture is first filled into the laser chamber 2 during a new fill. The gas composition for a highly stable excimer or molecular fluorine laser according to the preferred embodiment uses helium or neon or a mixture of helium and neon as a buffer gas, depending on the particular laser used. Suitable gas compositions are disclosed in U.S. Pat. Nos. 4,393,405 and 4,977,573, each of which is assigned to the same assignee as the present application and incorporated herein by reference. 317, 526, 09/513, 025, 60/124
, 785, 09 / 418,052, 60 / 159,525 and 60
/ 160,126. The concentration of fluorine in the gas mixture is 0.00
It is in the range of 3% to 1.00%, preferably about 0.1%. Additional gas additives, such as noble gases, increase energy stability and / or attenuator (at
No. 09/5, incorporated herein by reference above, as a tenuator)
It may be added as described in the 13,025 application. That is, F2
For lasers, the addition of xenon and / or argon may be used.
The xenon or argon concentration in the mixture ranges from 0.0001% to 0.1%. For ArF lasers, the addition of xenon or krypton, which also has a concentration of 0.0001% to 0.1%, may be used. In case of KrF laser,
Addition of xenon or argon, which also has a concentration of 0.0001% to 0.1%, may be used. Halogen and noble gas injection, total pressure regulation and gas replenishment procedures are preferably vacuum pumps, valve networks and one or more gas compartme.
nt). The gas treatment module 6 receives gas through a gas line connected to a gas container, tank, canister and / or bottle. Some suitable alternative gas treatments and / or make-up treatments other than those specifically described herein (see below) are each assigned US Pat. No. 4,977,573 to the same assignee as the present application. No. 5,396,514 and US Patent Application Nos. 60 / 124,785, 09 / 418,052, 09.
/ 379,034, 60 / 171,717, and 60 / 159,525
And US Pat. Nos. 5,978,406, 6,014,398 and 6,
No. 028,880, all of which are incorporated herein by reference. The '025 application referred to above may include a xenon gas source either inside or outside the laser system. Narrowing A general description of the line-narrowing features of some embodiments of the present invention is provided here, followed by a high spectral purity or bandwidth (eg, 1 pm).
The patents and patent applications incorporated herein by reference are cited as illustrations of variations and features used within the scope of the invention to provide an output beam having For narrowband lasers such as those used with refractive or catadioptric optical lithographic imaging systems, the exemplary line narrowing optics included in optical module 10 include a beam expander ( beam e
xpander), an optional etalon, and a relatively high dispersion (degr
ee of dispersion) is included. As noted above, the front optics module may also include line narrowing optics (60th / each assigned to the same assignee as this application and incorporated herein by reference). 16
6,277, 60 / 173,993 and 60 / 166,967 applications).

In the case of a semi-narrow band laser, such as used in an all-reflective imaging system, the diffraction grating is replaced by a high-reflectivity mirror and a dispersive prism. prism) results in low dispersity. Semi-narrowband lasers typically have an output beam linewidth in excess of 1 pm and can be as high as 100 pm depending on the laser system's broadband bandwidth characteristics.

The beam expander of the above-illustrated line narrowing optics of optical module 10 preferably includes one or more prisms. The beam expander may include other beam expanding optics such as a lens assembly or a focusing / diverging lens pair.
The grating or high reflectance mirror is preferably rotatable so that the wavelength reflected within the acceptance angle of the resonator can be selected or tuned. Alternatively, the diffraction grating, or other optical device or devices, or the entire line narrowing module are each assigned to the same assignee as the present application and are incorporated herein by reference. , 445 and No. 09 / 317,527 applications, and may be pressure tuned. Gratings can be used both to disperse the beam to achieve a narrow bandwidth and, preferably, to retroreflect the beam in the direction of the laser tube. Alternatively, a high-reflectance mirror is placed after the diffraction grating to receive the reflection from the diffraction grating and reflect the beam back in the direction of the diffraction grating in the Littman configuration, or the diffraction grating causes the diffraction grating to be transmitted. Lattice (transmissi
on grating). One or more dispersive prisms may be used and more than one etalon may be used.

Many alternative optics configurations can be used, depending on the type and degree of line narrowing and / or the desired selection and tuning, and the particular laser in which the line narrowing optics should be installed. For this purpose, they are each assigned to the same assignee as the present application, U.S. Pat. Nos. 4,399,540, 4,905,243, 5,226.
No. 050, No. 5,559,816, No. 5,659,419, No. 5,663.
973, 5,761,236, and 5,946,337, and US Patent Applications Nos. 09 / 317,695, 09 / 130,277, 09/244.
, 554, 09 / 317,527, 09 / 073,070, 60 /
124,241, 60 / 140,532, 60 / 147,219 and 60 / 140,531, 60 / 147,219, 60 / 170,34
No. 2, No. 60 / 172,749, No. 60 / 178,620, No. 60/173
, 993, 60 / 166,277, 60 / 166,967, 60 /
167,835, 60 / 170,919, 60 / 186,096, and also US Pat. Nos. 5,095,492, 5,684,822, 5
, 835,520, 5,852,627, 5,856,991, 5
, 898,725, 5,901,163, 5,917,849, 5
, 970,082, 5,404,366, 4,975,919, 5th
, 142,543, 5,596,596, 5,802,094, 4th
, 856, 018, 5,970, 082, 5,978, 409, 5
, 999,318, 5,150,370 and 4,829,536, and German Patent DE 298 22 090.3 are each incorporated by reference into the description of the present application.

The optical module 12 is preferably a partially reflective resonator.
Means for outcoupling the beam 20 such as a tor reflector) is included. Beam 2
0 may also be outcoupled in other ways, such as by an intra-resonator beam splitter or a partially reflecting surface of another optical element, in which case the optical module 12 may be coupled to a high reflectance mirror. Is included. The optical control module 14 preferably receives and interprets signals from the processor 16 and controls the optical modules 10 and 12, such as by initiating a realignment or reconfiguration procedure (as mentioned above). No. 241, '695
No., '277,' 554, and '527 applications). Gas Replenishment Based on Stored Energy Applied at Least Partly to the Discharge of the Laser, General Here, a preferred embodiment relating to the first aspect of the invention will now be described. This first aspect of the present invention discloses a gas replenishment algorithm that is also dependent on the stored input energy to the laser discharge and, preferably, time. The preferred method is more flexible than algorithms based on simple pulse counting. For example, if the irradiation process requires low pulse energy and the charging voltage should be adjusted accordingly, the new gas replenishment algorithm extends the refill interval. At high pulse energies the spacing is reduced. By this method, the make-up is automatically adjusted to the actual halogen combustion by gas discharge.

In this connection, the gas replenishment scheme, preferably carried out by the above-referenced 09 / 447,882 or 60 / 171,717 applications, provides a storage energy applied to the discharge. It can be set so that gas operation is performed at a certain value. That is, if the pulse energy is reduced at the same repetition rate, the gas operation will have a longer time interval that corresponds proportionally to the stored energy applied to the discharge. However, another factor of total time may be introduced into the calculation due to gas aging that may occur based on something separate from the total time from laser gas operation. For example, the gas mixture is known to degrade even when the laser is not operating, although at a much slower rate than when an electrical discharge is occurring in the laser chamber 2.

The preferred algorithm also considers a cross factor that depends on both the stored energy applied to the discharge and the time. A more preferred algorithm considers the high voltage and also considers the high voltage and one of the stored energy given to time and discharge.
Mutual factors, including one or both, may be considered. Other parameters may also be considered in this algorithm and therefore this algorithm should not be construed as limited to considering only one or another of the parameters mentioned above or any combination thereof. Absent. The input energy E ₁ to the laser discharge depends on the charging voltage U ₁ . The capacity of the main storage capacitor of the power supply 3 and the pulser module 4 is C. The input energy of one laser pulse is typically in the range of 2-10J for a lithographic laser system. Therefore, the input energy can be expressed by the following equation by the well-known capacitor energy formula.

The E = 1/2 * C * U ² gas replenishment operation is preferably initiated when the accumulated input energy of multiple laser pulses reaches a predetermined level. At the counter, the input energy of the laser pulse is added until a predetermined level E _repl is reached. This level is lithography
For laser systems, the range is preferably 0.1-10 MJ.

E _repl = ΣE _i = 1/2 * C * ΣU _i ² , where Σ is the standard symbol for the sum, and E _i and U _i are the energy corresponding to the i-th specific pulse. Voltage. (Exemplary Simplification Calculation Method) This algorithm can be suitably simplified when the expected change in the charging voltage is small. This simplifies the calculation of activation level.

If U _i = U + ΔU _i , then U _i ² = (U + ΔU _i ) ² ≈U ² + 2 * U * ΔU _i , where ΔU _i ² has been omitted by the above approximation. Therefore, E _{repl ≉1} / 2 * C * Σ (U ² + 2 * U * ΔU _i ) This is because the trigger level E _repl of gas supply is accumulated (accumulate).
d variation).

E _repl = [1/2 * C * n * U ² ] + [C * U * ΣΔU _i ]. When performing various gas replenishment operations such as halogen injection or gas replenishment of different amounts or compositions of gas, there may be many trigger levels such as E _repl ¹ , E _repl ² , etc. As mentioned above, the preferred gas replenishment algorithm considers more combinations than simply stored energy applied to the discharge, such as including time or charging voltage or other parameters that may indicate halogen depletion in the laser tube 2. May be required. Gas Control FIG. 3 illustrates a schematic of the gas processing module or gas control box 6 of FIG. 2 combined with a laser chamber 2 according to the preferred embodiment. The gas control box 6 is connected to the laser tube 2 and supplies gas based on the control signals received from the processor 16 as shown in FIG. The processor 16 regulates the delivery of the gas or mixture of gases to the laser tube 2 via the valve assembly 46 or system of valves. The valve assembly 46 preferably comprises a reservoir or compartment 47 having a known volume and having a pressure gauge P mounted to measure the pressure within the compartment 47. The compartments 47, like the laser tube 2, preferably each have means such as a thermocouple device for measuring the temperature of the gas in the compartment and the tube. The compartment 47 may have a volume size of, for example, 20 cm ³ (in contrast, the laser tube 2 has a volume of, for example, 42,000 cm ³ ). Gas compartment 4 housed in an external gas container
Four valves 8a-8d are shown in FIG. 3 as controlling the flow to 7. Of course, more or less than four such valves may be provided. Halogen filter H. F. Through a vacuum pump V.V. to compartment 47 shown connected through P. Another valve 32 is shown controlling the access of the.
Another valve 34 is shown controlling the flow of gas between the compartment 47 and the laser tube 2. Further valve (s) (not shown) are provided along the line 35 from the valve 34 to the line 2, for example using a pump to exhaust the line 35 to the line 3
The atmosphere in 5 can be controlled.

The small amount of gas or gas mixture is preferably HI or enhanced HI (enhanced HI).
) Or during PGR or MGR operation from the compartment 47 into the discharge chamber 2. As an example, the gas supply connected to the valve assembly 46 via the gas line 36a is a premix A containing, for example, 1% F ₂ to 99% Ne, through the gas line 36b. Is 1% Kr vs. 99% for a KrF laser
Can be a premixed gas B containing Ne. For ArF lasers, the premixed gas B will have Ar instead of Kr, and for the F ₂ laser no premixed gas B is used. The F ₂ laser can have a combination of inert gases such as Ne and He in its premixed gas A and / or in the gas mixture in the tube 2 like the ArF laser, or both lasers can be used as buffer gas. It is also possible to have only He (see for example 09 / 317,526 and 4,393,5).
No. 05 patent). Therefore, by injecting the premixed gas A and the premixed gas B into the tube 2 through the valve assembly, the concentrations of fluorine and krypton in the laser tube 2 can be replenished in the case of the KrF laser, respectively. Gas lines 36c and 36d may be used for different additional gas mixtures such as those with gas additives such as Xe as mentioned above (see 09 / 513,025 application). ). Although not shown, tube 2 preferably has additional means for venting gas, and alternatively gas is vented through the valve assembly, such as through valves 34 and 32.

New fill, partial and mini gas replacement and gas injection processes, such as enhanced and regular micro-halogen injection, and any and all other gas replacement operations can be performed in various feedback loops. Based on the input information, valve assembly 46 and pump V.V. P. Is initiated and controlled by the processor 16 which controls With each of the gas injections, according to the preferred embodiment and the approximation used in the above calculations to determine the interval between gas replenishment operations based on the stored energy input to the discharge,
An exemplary method according to the invention for accurately and precisely replenishing the fluorine concentration in the laser tube 2 by a small amount such that the important output beam parameters are not disturbed too much, i.e. very small, will now be described. To be done. The processor 16 monitors a parameter indicating the fluorine concentration in the laser tube 2 and performs micro halogen injection (
HI), that is, the amount E _repl of stored energy in the discharge is reached, for example. As will be appreciated, a replenishment operation with a known correlation between the exhaustion of halogen and / or active noble gas in the tube 2 or the accumulation of pollutants in the tube 2 is desirable here.

The processor 16 then sends a signal to open the valve 8a and fill the compartment 47 with the premixed gas A to a predetermined pressure, for example 5 bar. Next, since the valve 8a is closed and the valve 34 is opened, at least a part of the premixed gas A with which the compartment 47 is filled is emitted to the laser tube 2. If the pressure in the tube before injection is eg 3 bar, the tube has 42,000 cm ³ and the injection is such that the pressure in the accumulator has dropped to 3 bar after injection, then 2 × 20 / 40,000 bar or 1 mbar
Would be an approximate pressure rise in tube 2 as a result of injection. Premixed gas A is 1
With% F ₂ to 99% Ne, the increase in the partial pressure of F _{2 in} the laser tube as a result of the injection would be about 0.01 mbar.

The above calculation may be performed by the processor 16 to more accurately determine how much F ₂ premix gas was injected, or it may be received by the processor 16 prior to injection, eg Pressure in compartment 47 calculated by processor 16 regarding status information of monitored parameters such as accumulated energy input to discharge, time, etc., or how much F ₂ should be injected based on pre-programmed criteria. Can be set. Correction for the difference in temperature between the gas in the compartment 47 and the gas in the tube 2 may also be performed by the processor 16 for more accuracy, and the temperature of the gas in the compartment 47 may be, for example, in the laser tube 2. The temperature may be preset.

Suitably, a quantity of premixed gas corresponding to a total gas pressure of 10 mbar or a pressure increase in the tube 2 of less than 0.1 mbar F ₂ partial pressure is injected from the compartment 47. More preferably, a pressure increase in the laser tube 2 of less than 5 mbar, or even less than or less than ₂ mbar, or a total gas pressure (F ₂ partial pressure of 0.05 or 0.02 mbar) as a result of gas injection. Occurs.

The compartment 47 may simply be the valve assembly 46 itself, or may be an additional accumulator (as described in detail below). Compartment 47 is also made of KrF
, ArF or F ₂ lasers, a small amount of continuous gas in order to compensate for the degradation of the halogen gas and / or another gas (s) in the discharge chamber 2 of an excimer or molecular laser. Configured to be injected at short intervals.

As explained above, there may be more than one compartment, such as compartment 47, having different properties, such as each having a different volumetric space. . For example, there may be two compartments, one for HI and the other for reinforced HI. There are more than two compartments because of the still further versatility of the amount of halogen to be injected during gas operation and to adjust the drive voltage range in response to different gas operation algorithms. May be. Different premixed gases may be injected from different compartments. Also, while the exemplary method has been described using a premixed gas of a particular gas composition, many different gas compositions can be used in accordance with the present invention. For example, gas compositions having a higher percentage concentration of fluorine (or hydrogen chloride) such as 5% or 2% instead of 1% can be used. Also, for example, a 100% buffer gas container of Ne and / or He (buffer)
There may be additional valves connected to the gas container).

Advantageously, the processor 16 and the gas supply unit 6 are arranged to allow a very small amount of one or more gases or gas mixtures to be supplied or injected into the discharge chamber 2. The injection of a small amount of gas or gas mixture results in an increase in gas pressure in the discharge chamber 2 of less than 10 mbar, preferably 0.1 to 2 mbar. Since each gas in the gas mixture in the discharge chamber 2 can be adjusted separately, the gas composition in the discharge chamber 2 can be precisely controlled. For example, Kr, Ar
Or a similar injection of Xe can be done to replenish these gases in the laser tube 2.

The amount of gas injected during the gas injection or replacement procedure is advantageously small so that the laser output beam parameters do not change significantly with each injection. This injection is preferably carried out regularly at predetermined intervals corresponding to a known exhaustion of gas. For example, under the current operating conditions, the halogen partial pressure in the gas mixture of the F ₂ laser is about 3 bar after fresh filling, every X minutes or every Y shot or the stored energy E _input (any If it is known to consume only 0.1 mbar for every Z amount of energy units), halogen injection with a premixed gas of, for example, 1 mbar (pressure increase in tube 2) containing 1% F ₂ is essential. According to the invention it is carried out every X / 10 minutes or every Y / 10 shots or every Z / E _input to maintain the halogen concentration, or a halogen injection of 2 mbar premixed gas is carried out every X / 5 minutes etc. It is possible. Also, a micro halogen injection (HI) of 1 mbar premixed gas A containing 1% F2 and 99% Ne buffer was performed every X / 5 minutes for 100 minutes and no injection was performed for 100 minutes. , Or 100E _input for each Z / 5E _input and no injection is performed for 100E _input , and so on. Many variations are possible within the spirit of the invention, including irregular gas operation determined by the processor 16.

In contrast to the present invention, for example, injection of a premixed gas A of 50 mbar (pressure increase in tube 2) (again a partial pressure increase of F ₂ in tube 2 of 0.5 mbar results in If 1% and having a F2) of the 5X minute or per 5Y each shot or 5E _{input the} basis, or carried out at any time, such as corresponding to about 17% increase in F ₂ concentration, the laser by a large injection volume The output beam parameters of the beam have a noticeably undesired variation in response. For example, if a large implant is made, the pulse energy or drive voltage can vary by 10% or more. Even with large implants, if the laser is not stopped or the industrial process is not interrupted, meaningful output beam parameter disturbances will result in inaccurate industrial processes.

The halogen injection algorithm of the present invention allows total halogen injection to be performed for longer periods of time.
It is also considered to extend to the pulse count or amount, the stored energy input to the discharge, the charging voltage or a change thereof. The stability of the pulse energy, moving average energy and energy dose, among other things, is because the high voltage and the F ₂ concentration do not change significantly over the duration of some halogen implants, so there are other meaningful output beam parameters. Large changes in are eliminated. Again, some of these other output beam parameters are listed above, but each of them is quite stable using the method of the invention.

FIG. 4 schematically shows another configuration of a gas line for injecting halogen into the discharge chamber 2 of the laser of FIG. 2 using the accumulator 46 a. The accumulator 46a is connected to the laser tube 2 through a laser head valve LH. The accumulator 46a is also connected to the gas line 12a through a halogen valve H connected to a gas bottle 13 containing a halogen or halogen premixed gas. For example, the gas bottle 13 contains a gas mixture containing an F ₂ mixture (among other possibilities, for example 5% F 2/95% Ne or 5% HCl / 1% in a neon mixture. H2 or 1% F2 to 99% Ne premixed gas). Pumps connected to the accumulator 46a and the laser tube 2 respectively through vacuum valves V are shown. Tube 2 contains buffer gas through valve B, noble gas or noble gas premixed gas through valve R (eg, used with KrF, ArF, XeCl and XeF excimer lasers) and inert gas through valve I. It is shown to be valved to additional gas lines and valves including. As a valve connecting to the source of Xe to be used as an additive in the gas mixture in the tube, an inert gas valve I or another valve not shown may be used. One or more additional accumulators may be added to the system.

[0056] The accumulator 46a has a special advantage that a small amount of a gas containing F ₂ in F ₂ premixed gas to be injected by each silver implantation according to the invention can be accurately controlled. The accumulator 46a is easily pumped to a low pressure. The exact amount of F ₂ gas or F ₂ premixed gas is discharged into the accumulator 46a, F ₂
The amount of the total gas pressure in the accumulator 46a, a known volume of the accumulators 46a and the laser tube 2, and is determined by the known percent concentrations of concentration i.e. F ₂ of F ₂ in the premixed gas. Partial pressure increase in the F ₂ laser tube 2 after injection is determined based on the amount of F ₂ that are known to exist before (and even after injectable) in the accumulator 46a infusion.

Interval and / or before this determination and / or before (for example as measured by stored energy or charging voltage or changes in charging voltage up to time or pulse counting or discharge) and / or before, Based on other factors such as the value of the drive voltage at the time of the current and / or next gas operation, the interval between the current and next gas operation and / or
Alternatively, the amount of halogen-containing gas or the total gas to be injected in the next gas operation is determined and the exact amount of each gas, in particular the halogen-containing gas, is injected in the next gas operation. Also, the type of gas operation to be performed may be determined based on these or other factors.

Various gas operations and procedures are now described. These treatments are potentially applicable to all gas discharge lasers, although excimer lasers (eg KrF, ArF, XeCl and XeF) and F ₂ lasers would greatly benefit from the present invention. In the following, a KrF laser will be used as a particular example. The process begins with a new fill that occurs prior to operation of the laser system. For a new filling, the laser tube 2 is evacuated and then filled with a new gas mixture. As a result of the new filling of the KrF laser, usually around F2: Kr: Ne = 0.1%:
A gas mixture with a distribution of gas of 1.0%: 98.9%. If the gas mixture in the KrF laser discharge chamber 2 has a normal total pressure of about p = 3000 mbar, the partial pressures of F ₂ and Kr will usually be about 3 mbar and about 30 mbar, respectively. In the case of an F ₂ laser, the new fill is F2: Ne and / or He = 0.1%: 99.
It produces a normal gas distribution of 9%. As suggested, for the F ₂ laser,
Instead of only Ne, He or a mixture of He and Ne is used as a buffer (see the above '526 application).

The new filling procedure can be performed using a separate gas line supplying pure or premixed gas. The usual premixed gases that are always used in semiconductor manufacturing are premixed gases A and B, where premixed gas A has a 1% F ₂ /1% Kr in Ne and a premixed gas B. Has a Kr of 1% in Ne. After a new fill is made, the halogen gas begins to react with the components of the laser tube 2 with which it comes into contact, whether or not the laser is operating. “Gas replenishment” refers to gas replacement (PGR and MGR, respectively, which is affected by changing the amount and composition of the injected and released gases) and gas injection (again, performed to return the state of the gas mixture to a state close to the new fill). It is a general term including HI and enhanced HI, each of which is affected by varying the amount and composition of the injected gas.

An optional gas replenishment procedure is that each gas in the gas mixture is now depleted at a different depletion rate due to the halogen depletion described above and the gas replenishment procedure performed in response. It is done with that in mind. For example, in the case of a narrow band KrF laser, the consumption of F ₂ occurs at a rate of about 0.1% to 0.3% (and sometimes close to 1%) for every one million shots. Consumption is about 10-5
It occurs 0 times slower. The Ne buffer in this case is less important, but may also be considered as part of the overall gas refueling operation, for example to maintain the desired pressure in the tube 2.

Separate gas operations are preferably performed to replenish each constituent gas of the gas mixture.
For example, in the case of a KrF laser, F ₂ is replenished by injection of halogen or halogen / rare gas or premixed gas A, and Kr is replenished by injection of rare gas or premixed gas B. Other gas additives such as Xe may be supplemented with Xe gas or further premixed gases C, D, etc. Individual wear rates also refer to whether the laser is operating in wideband or narrowband modes, operating energy level, laser turn off or continuous, standby or other burst pattern operation. It also depends on the operating state of the laser, such as whether the laser is operating at the same time and the operating repetition rate. Processor 16 is programmed to account for all these variations in laser operation.

The state of the gas mixture has a deviation of the content of fluorine and krypton of 5
% Or less, preferably 3% or less is considered to be sufficiently stable in the present invention. Without any gas replenishment operation, the partial pressures of F ₂ and Kr after 100 million shots deteriorate by between 30% and 100% and between 0.5% and 5%, respectively. In order to compensate for various gas depletion rates in the discharge chamber, the laser of the preferred embodiment
The system performs a variety of separate and cross-linked gas replenishment procedures, which refer to a comprehensive database of various laser operating conditions to produce a variety of individual degradation rates. rate) is taken into consideration. Suitable techniques are disclosed in the 09 / 379,034 application already mentioned above. The behavior of individual lasers in operation and the experience associated with gas degradation under various operating conditions are stored in their database and used by a processor-controlled "expert system" to determine the current state of the laser. Manage gas replenishment or refurbishment operation. The history of gas operations performed during the current operation of the laser is also used by the present invention.

As mentioned above, a series of small gas injections (called enhanced normal micro-gas or halogen injections, or HI) are used to either excimer or molecular laser composition gases, In particular, the very active halogen is returned to its optimum concentration in the discharge chamber without changing important output beam parameters. However, the gas mixture also deteriorates over time as contamination accumulates in the discharge chamber 2. Therefore, the preferred method is to use mini gas replacement.
replacement: MGR and partial gas replacement: PGR
) Is also done. Gas replacement generally involves the release of some gas from the discharge chamber, including the discharge of some pollution. MGR involves periodic small gas replacements at longer intervals where small HIs are performed. PGR involves greater gas displacement, which is generally done at longer regular intervals to "clean" the gas mixture. Exact intervals in each case are determined by consulting current laser operating conditions (consulting
) And an expert system and a comprehensive database. This interval is a change in the parameter that changes in a known relationship with the deterioration of the gas mixture. Thus, the intervals are: time, pulse count, stored energy input to discharge, charge or drive voltage or charge voltage change, pulse shape, pulse duration, pulse stability, beam profile, coherence, One or a combination of discharge widths or bandwidths. In addition, the accumulated pulse energy dose may be used as this interval. Each of HI, MGR, and PGR may be performed during operation of the laser system, and thus the uptime of the laser (
uptime) does not have a harmful effect (compromise).

Three exemplary gas replenishment methods to stabilize the optimal gas mixture are described below.
Many other methods are possible, including combinations of those described below. The methods and parameters used can also be changed during laser operation based on a database and expert system depending on the operating conditions of the laser. The processor 2 and the gas supply unit 6 are arranged to perform a number of methods based on a comprehensive database of laser operating conditions and gas mixture states.

Each method comprises a distinct, very small gas operation with a small continuous gas injection, preferably by injecting a premixed gas of less than 10 mbar and more preferably 0.1-2 mbar. The gas preferably contains less than 5% halogen containing spe in order not to fluctuate the operation of the laser and the parameters of the output beam.
cies) is included. Whatever the composition of the premixed gas, the most important is the amount of halogen in the premixed gas. That is, a suitable amount of halogen-containing substance injected in a small gas operation is preferably less than 0.1 or 0.2 mbar in the laser tube 1, and more preferably a partial pressure of 0.001-0.02 mbar. Respond to the rise. (At least in part time-based gas replenishment) The first parameter referred to above upon which gas replenishment was based was the stored energy applied to the discharge of the laser. A second exemplary gas stabilization method involves performing gas injection based at least in part on operating time. Preferably, the method used takes into account both of these parameters. This method determines whether the laser is operating, i.e. the laser system is up and doing industrial processing,
Consider whether it is in standby mode or just stopped. That is, the method is useful for maintaining a gas composition that is either active or passive. Note that the following description of various time-based gas replenishment procedures is used in conjunction with the stored energy-based gas replenishment procedures applied to the discharge. Therefore, a separate discussion in which time is replaced by the stored energy applied to the discharge is omitted, but in general these and other parameters such as charging voltage and changes in charging voltage are interchangeable in the following discussion. , Gas replenishment algorithms within the scope of the present invention are combinable.

HI, MGR, and PGR that are time-correlated are performed at selectable time intervals based on operating conditions. For example, HI is performed after time interval t1, MGR is performed after time interval t2, and PGR is performed after time interval t3. According to a preferred embodiment, the time intervals t1, t2 and t3 are adjusted in real time as well as the amount and / or composition of gas injected during gas operation. Suitably, the time intervals and the amount and composition of gas are adjusted between gas operations. further,
The drive voltage range in which a particular gas operation is carried out is also preferably adjusted at least for each new filling based on the aging of the optical components of the tube and the laser cavity. This range is adjusted during operation, even during new filling, for example on the basis of the beam-induced effect exerted on the optical components of the narrowing module. (See US patent application Ser. No. 09 / 454,803, assigned to the same assignee as the present application and incorporated by reference herein by reference).
.

In the following, detailed graphs will be described for a laser system operating according to the preferred embodiment. Lasers are usually non-pulsing or have low repetition rates (<1
When in standby mode with a (00 Hz) pulse, gas operation occurs after a few hours. If the laser is completely off (power-off mode), the battery-powered internal clock is still running and the expert system has a reasonable number of controlled times during the warm-up phase after laser restart. The injection can be released. The number and amount of injections is also related to certain drive voltage start conditions that initiate a preferred sequence of gas operations to restore optimum gas quality.

5 and 6 show a fully operating system according to the present invention.
3 is a graph of drive voltage vs. time also illustrating the periodic HI and periodic HI and MGR intervals, respectively, for. FIG. 5 includes a plot of drive voltage versus time (A), in which a vertical line on the graph (part one) is shown for a narrowband laser operating in 2000 Hz burst mode with 10 mJ output beam energy. Part is indicated by "B" for reference), the HI is about 1
Every 2 minutes. The vertical axis of the figure corresponds to graph A only. As shown by graph A, this small HI does not cause a noticeable discontinuity in the drive voltage. The top horizontal axis shows the increase in stored energy in the discharge.

FIG. 6 is a graph of drive voltage versus time (labeled “A”), where HI is for a narrowband laser operating in 2000 Hz burst mode with 10 mJ output beam energy. , Every 12 minutes, as shown by the short vertical line on the graph (also partly shown by a "B" for reference, but nowhere on the vertical axis indicates a halogen injection). Done and MGR
Is the longer vertical line on the graph (some of which are indicated by a "C" for reference,
Again the vertical axis has no meaning with respect to the illustrated MGR), about every 90 minutes. Again, the driving voltage is almost constant around 1.8 KV, and no large change exceeding 5%, for example, is observed. Similar to FIG. 5, the top horizontal axis shows the increase in stored energy in the discharge.

FIG. 7 shows the absolute pulse energy (abso) for the system according to the invention.
value of standard devia as a percentage of lute pulse energy
tion: SDEV) (indicated by "A") and moving average stability value (moving ave)
rage stability: ± MAV) is included, which is a graph of pulse energy stability versus pulse time of a laser pulse. The graphs labeled "B" and "C" show the moving averages of groups of 40 pulses each. During this operation, a micro-halogen injection was performed resulting in a very stable continuous laser operation without any detectable deviation caused by the gas refill operation. Gas Replenishment Based on Stored Energy (and / or Pulse Count) Applied at least Partly to the Discharge A second exemplary gas stabilization method is the stored energy applied to the discharge using a stored energy counter, and Alternatively, it involves performing gas injection based on shot or pulse counting using a shot or pulse counter. Depending on the amount of stored energy, or also the operating mode of the laser, eg N (μHI),
After a certain number of laser pulses such as N (MGR) and N (PGR), μHI
, MGR and PGR can be performed respectively. Usually, μHI is KrF, A
rF, fluorine premixed gas in the case of XeF or F ₂ laser (e.g., 1-5% F ₂ versus 95-99% of Ne) (Ne is replaceable in a mixture of He or He and Ne), also XeCl or For a KrCl laser, HCl premixed gas (eg N 2
0.5% of 1-5% HCl in He or He: 1%). ． ． It has a pressure of 20 mbar and is emitted after hundreds of thousands, or even millions of laser shots. Each μH
I compensates for the halogen depletion since the last gas operation, typically less than 0.1 mbar of halogen containing s in the laser tube 1 per million shots, for example.
pecies) and more preferably a partial pressure increase of 0.001-0.02 mbar. The actual amount and shot spacing will vary depending on the type of laser, the composition of the discharge chamber, the composition of the original gas mixture, and the operating mode used, eg, energy or repetition rate.

As alluded to above, additional methods not specifically mentioned, according to the preferred embodiment, include many of the details described above, including the stored energy applied to the discharge, or the time, charging voltage and / Or use a combination with changes in charging voltage. The total input electrical energy to the discharge is maintained in the counter for that purpose, and gassing takes place at certain intervals or after a certain amount of this input electrical energy has been applied.

Also, according to the preferred embodiment, any intervals in the exemplary method are dynamically adjusted from injection to injection, as well as the amount of halogen injected by each gas operation. The interval between the current and next implants is set based on any one or combination of parameters such as the drive voltage or any output beam parameters described above. Additionally, the amount of halogen injected in the current implant and / or the interval between the previous and current implants may be considered.

According to the present invention, the amount of halogen injected with any HI or enhanced HI can be determined by the present invention at the time of injection and / or immediately before and / or immediately after the injection.
4 to 5) and the pressure in the laser tube 2 is measured.
Moreover, the temperature of the gas in the accumulators 46, 46a and the tube 2 may also be measured. The internal volumes of the tube 2 and the accumulators 46, 46a are known in advance. The well known formula PV = N for calculating the amount of halogen injected into a tube during any injection.
kBT is used.

For example, if the accumulators 46, 46a have a measured halogen partial pressure Pa and temperature Ta, and a volume Va, the accumulators 46, 46a contain Na fluorine molecules. During the injection, all the molecules of Na were injected into the laser tube and the tube was heated to the temperature TT
And a volume VT, the change in fluorine partial pressure in the tube resulting from the injection will be P (F ₂ ) T = PaVaTT / VTTa. Since it is desirable to maintain the total number of fluorine molecules in the tube, the change in the number of fluorine molecules in the tube, that is, N (F ₂ )
T = P (F2) TVT / kBTT is calculated, and the amount is constantly monitored (beep track
) Is more appropriate. Then, the amount of halogen and / or the interval to the next injection can be calculated by calculating the calculated amount of halogen injected in the previous injection, the partial pressure of halogen in the tube after the previous injection, and / or in the tube after the next injection. It is determined based on the amount of halogen that it is desired to have.

The overall calculation also depends on the amount of exhaustion the halogen gas suffers (or will suffer) between injections. This depletion is known primarily as a function of many factors, including, for example, time and energy stored in the discharge (and possibly any of the parameters listed above or elsewhere). For example, the halogen partial pressure in the laser tube 2 (and alternatively the number of halogen molecules) in the interval between injections.
Can be calculated by kt × t and kp × E _input , where kt and kE _input are constants depending on the rate of halogen consumption with time and the stored energy input to the discharge, respectively, in the interval under consideration. And t and E _input are respectively
It is the amount of time and the stored energy input to the discharge. The amount of stored energy in the discharge itself depends on the repetition rate and pulse energy, and for lasers operating in burst mode, the number of pulses in a burst and the dwell interval between bursts are also considered. Again, other parameters may have an effect and be an additional term included in this calculation.

Now, from one interval to the next, the calculation could be done as follows. The increase in fluorine partial pressure in the laser tube (or the decrease represented by the sum becoming negative) over that interval is calculated as:

[0077]

[Equation 1]

Where ΔE _i is the total energy applied to the discharge since the last halogen injection, and Δt is the time since the final injection. Again, additional terms are associated with charging voltage, changes in charging voltage and / or pulse counting, and the k _i terms for any of these parameters are repetition rate, pulse energy, laser output power, It depends on many factors such as the conditions of the burst, the age of the laser, the gas mixture and / or the optics or electrodes of the laser.

Since it is the total number of fluorine molecules that is desired to be kept constant, the change in the number of molecules is calculated by the following equation.

[0080]

[Equation 2]

Here, the constants k _t and k _Ei differ from the case of partial pressure calculation due to unit conversion. The whole algorithm will try to keep the total number of halogen molecules (or halogen partial pressure) constant. That is, the change in particle number (or partial pressure) will be continuously summed over many intervals, and preferably all intervals since the last new fill. This grand total will be kept as close to zero as possible by the present invention.

As already discussed, a counter of stored energy to the discharge can also be used in combination with a time-related gas make-up, and a stored energy counter or a time-related gas can also be used. Either supply
As mentioned above, it can be used in combination with pulse counting, charging voltage and / or changes in charging voltage, along with many other parameters such as ASE, pulse shape over time, etc. Laser pulse operating modes with different stored energy counters or total applied energy, eg burst pattern, or continuous pulsing mode with different pulse repetition rates
, But a number of individual times or counters N _i (μHI) such as shots or input energies are used. All of these various counters can be stored in the expert system's database. Which of these various counters N _i (HI) should be used at any given time is determined by the software of the expert system.

FIG. 8 qualitatively illustrates the free of discontinuity drive voltage when a small partial pressure rise is performed in the laser discharge chamber by the HI according to the present invention. Driving voltage has been shown to be almost constant in the vicinity of 1.7KV over 150E _i, where E _i is in connection with the gas supply for a given operating laser conditions, for example, 150 million pulses It is the magnitude of the corresponding input energy to the discharge, eg in Joules, while the HI is done approximately once every 12E _i , corresponding to eg 12 million pulses. The pulse energy is also maintained at a constant level.

[0084]   Expert systems have other types for any parameter that is counted
Different gas operations (ie N_iDifferent types of counsel (different from HI)
That is, N_i(MGR) and / or N_i(PGR) can be used
It Gas displacement operations, such as MGR and PGR, can be applied to various gas mixtures in the laser tube.
A predetermined amount of gas is replaced or substituted. As already mentioned, MGR and
PGR involves gas injection with the release of gas from the laser tube, whereas HI
Does not release gas. Outgassing is commonly used to expel pollution from gas mixtures.
Alternatively, it may be done simply to reduce the pressure in the laser tube. Individual in the gas mixture
Uneven degradation of gas components is well compensated using MGR and PGR
Again, according to the expert system's decision, various operating modes and conditions
In contrast, various numbers of N_i(MGR) and N_i(PGR) is used. All of these
Setting, ie N for injection for each gas_i(HI), N_i(MGR), N _i (PGR) and the individually selectable part are
Considering the conditions of gas consumption and replenishment that change with the progress of laser, the change over time of the laser tube
And / or adapted to the aging of the resonator optics. The amount of compensation is set manually
By default or in a computer controlled expert system database
Preselected based on the settings of In case of MGR, injected gas as in HI
Is the total pressure rise in the laser tube of a few mbar (or only a few percent).
MGR is suitable in combination with a small pressure release of a few mbar to 10 mbar of tube pressure
The pressure in the tube is returned to a pressure close to the pressure in the tube immediately after the last new filling.

More than one gas may be injected or replaced in the same gas operation. For example, a quantity of halogen and a quantity of active noble gas and / or gas additive for an excimer or molecular fluorine laser may be injected together in the laser tube. This infusion may be accompanied by a small pressure release similar to MGR or PGR. Also, this mixture of halogen and noble or additive gas may be injected simply to raise the partial pressure of each gas in the discharge chamber without any gas evolution. Gas Replenishment Based at Least Partly on Charging Voltage or Change in Charging Voltage A third exemplary gas stabilization method involves performing gas injection based on the operating drive voltage value of the laser. This method may in particular be combined with any of the first, second and third exemplary methods. That is, the time-related t ₁ (HI), t ₂ (MGR, discussed above.
), T ₃ (PGR) and Ni associated with the counter of input electrical energy to the discharge
The (HI), Ni (MGR), Ni (PGR) gas operation is generally adjusted during operation according to the operating drive voltage or charge voltage, and preferably the operation band of the drive voltage.

Referring to FIG. 9, it is possible to define a number of drive voltage levels (HVi) at which the levels at which individual gas operations should be performed are predetermined. The processor 16 monitors the drive voltage, and the gas supply unit determines that the value of the drive voltage, ie Depending on which preset range the current operating drive voltage is in (y-axis in FIG. 9), the degree of gas injection and the degree of change and mini gas replacement are performed.

An example according to the invention will now be described with reference to FIG. Laser system may operate at a driving voltage between the HV _{mi n} ~HV _max. Actual minimum and maximum driving voltage is
It is set to be in a much smaller range between HV ₁ and HV ₆ as indicated by the broken coordinate axis. The advantage of this embodiment is that the range HV _{1 to} HV ₆ itself can be reduced to a very small window so that the operating voltage does not change significantly during operation of the laser. Since this operating range itself lies between HV _{min and} HV _max , the actual voltage range (in Volts) corresponding to this range can be adjusted to, for example, output energy attenuating gas additive.
) Can be adjusted to improve the lifetime of the optical components of the resonator and the laser tube (see the '025 application mentioned above).

The coordinate axes of FIG. 9 show gas manipulations performed based on one or more storage parameters when the drive voltage is within each interval. The general sequence of performing gas operations is left to right as the gas mixture ages. However, as each gas operation is performed, the drive voltage is checked and the next gas operation is in the same drive voltage range,
Or it may correspond to different drive voltage ranges shown to the left or right of that range. For example, after PGR is performed (when it is determined that the drive voltage is above HV ₅ ), the drive voltage is reduced to HV ₂ to HV ₃ and thus the system is restored to the original HI.
And MGR1 gas control operation may be returned.

Within the operating range between HV _{1 to} HV ₆ , several other ranges are defined. For example, the drive voltage HV is between HV ₁ ~HV ₂ (i.e., HV ₁ <HV <HV ₂₎ if, gas operation due to the presence of halogen in an amount sufficient gas mixture is not performed.
Driving voltage is between HV ₂ ~HV _{3 (ie, HV 2 <HV <HV 3} ) time,
MGR1 and regular HI are performed periodically based on accumulated parameters (ie, input electrical energy to discharge, time, and / or pulse count, etc.). This is the normal operating range of the system according to the preferred embodiment.

[0090] driving voltage is between HV ₃ ~HV _{4 (ie, HV 3>HV> HV 4} ) when the injection amount of one or both of the HI and MGR increases with corresponding gas discharge. In this example, only HI increases. That is, the range between HV ₃ and HV ₄ in FIG. 9 is the range in which the enhanced HI is performed, and the same MGR amount as the previous range between HV ₂ and HV ₃ is maintained. Enhanced HI differs from normal HI in one or both of two ways. First, the volume per injection can be increased. Second, the spacing between consecutive HIs may increase.

[0091] range between HV ₄ ~HV _{5 (ie, HV 4 <HV <HV5 5} ) is, HI and M
One or both GR doses represent a new range that increases with corresponding outgassing. In this example, only MGR is increased. That is, the amount the drive voltage when it is in the range between HV ₄ ~HV ₅ is that increased than the normal amount mgr1 (enhanced AMOUNT
) Halogen gas is injected into each MGR2 (along with the corresponding gas release). HI is also performed regularly in this range to restore the gas mixture to its original state. It should be noted that several ranges in which either or both of the amounts injected into HI and MGR are adjusted are each defined corresponding to one defined drive voltage range. is there. Also, adjusting the amount injected for each injection, as mentioned above in connection with monitoring the pressure (and optionally temperature) in the accumulator (and optionally the laser tube). You can

When the driving voltage is higher than HV ₅ (that is, HV ₅ <HV <HV ₆ ), a larger gas displacement PGR is performed. PGR can be used to replace up to 10 percent or more gas mixtures. Here, some kind of protective measure (sa
feguard) may be used. One is to allow some time (such as a few minutes) after the HV ₅ level has been exceeded before allowing gas manipulation to take place, thus ensuring that the drive voltage actually rises due to deterioration of the gas mixture. It is something to do. When the drive voltage exceeds HV ₆ , it is the time for new filling of the laser tube. Note that the size of the drive voltage range shown in FIG. 9 is not necessarily drawn to scale.

FIG. 10 is a flow chart for performing normal and enhanced HI, MGR and PGR according to the preferred embodiment and the example described in FIG. The procedure starts with a new filling, where the discharge chamber 2 is filled with the optimum gas mixture. The laser then operates for industrial applications under different conditions, such as different modes at different repetition rates and emission at certain pulse energies or moving averages, either in standby mode or completely shut down. can do. A drive voltage test (HV test) is then performed to determine which gas operation is a candidate for the next gas operation. In one embodiment, gas manipulation may be based solely on drive voltage. In the preferred embodiment, the following steps are performed first.

The measured drive voltage (HV) is compared with predetermined values for HV _{1 to} HV ₆ . The processor is between HV is HV ₁ ～HV ₂ (i.e., HV ₁ <HV <HV ₂₎ or to determine and thus follows the path (1), gas operation is not performed the procedure returns to the previous step. Although not shown, when HV is lower than HV ₁ , the halogen concentration in the laser tube is lowered by discharging a part of the laser gas from the laser tube and / or injecting some buffer gas into the laser tube. The steps to be taken are taken.

If the processor determines that the HV is between HV _{2 and} HV ₃ , then the system is within the normal operating drive voltage band. If the system is in the normal operating band, it is preferable to follow the path (2), whereby the expert system is pre-determined based on the operating conditions, preferably time, input electrical energy to the discharge and Normal HI and MGR ₁ may be performed based on the pulse count interval. Whether HI or MGR ₁ is performed is preferably reading of one or more counters such as time, stored energy applied to discharge, pulse count, ASE, pulse shape with respect to time, charge voltage change, etc. It will depend on (one or more indications). Again, each gas operation can be adjusted depending on the calculated partial pressure in the laser tube or the number of halogen molecules, as explained above.

HI and MGR ₁ performed when following pathway (2) are determined by any of the methods described in US patent application Ser. No. 09 / 379,034 already incorporated herein by reference. can do. If the HV is not within the normal operating band, then it is determined whether the HV is lower than HV ₂ (ie, HV <HV ₂ ). If HV is lower than HV ₂ , gas operation is not performed by following the path (1).

[0097] HV is between HV ₃ ～HV ₄ (i.e., HV ₃ <HV <HV ₄₎ case, follows a path (3), electrical energy again time, is applied to the pulse count and / or discharge Depending on the counter value (s) of the counter or other counters used. The exact amount and composition of gas injected and gas released will preferably be determined by an expert system and will depend on operating conditions.

[0098] HV is between HV ₄ ～HV ₅ (i.e., HV ₄ <HV <HV ₅₎ case follows the path (4), is reinforced HI and MGR _2, also based on the examination of the value of the counter It can be made depending on the decision. Again, the exact amount and composition of gas injected and gas released will preferably be determined by an expert system and will depend on operating conditions.

[0099] HV is between HV ₅ ~HV _{6 (ie, HV 5 <HV <HV 6} ) case, P
GR is performed. If HV is higher than HV ₆ (ie HV ₆ <HV), then a new fill is done. The method determines the mode of operation of the laser after any of the paths (2) to (5) has been subjected to a corresponding gas operation and preferably after a certain settling time. , The HV is measured again, and the process returns to the step of comparing with the predetermined HV levels HV _{1 to} HV ₆ .

All the settings of these different operating voltage levels and times, the electrical energy applied to the discharge, the change in drive voltage, the ASE, the pulse shape and / or the pulse counting schedule with respect to time, mentioned above among others, Reference can be made to a computer controlled database, which can be made individually or stored for different operating conditions. Operation of the laser at various HV levels under different operating conditions such as continuous pulse or burst mode can be considered.

The combination of all these various types of gas control and replenishment mechanisms requires the coordination of many factors and variables. Combined with an expert system and database, the processor-controlled laser system of the present invention extends the life of the gas before a new fill is needed. In principle, the bringing down of the laser system for a new filling is completely prevented. Therefore, the life of the laser system will depend on the scheduled maintenance intervals determined by other laser components, such as for replacement of laser tube windows or other optical components. Again, as mentioned above with reference to the '126 application, the life of the laser tube and resonator components can even be extended, and the spacing between the outage periods can be extended.

The gas replenishment procedure described above may be combined with cryogenic or electrostatic gas refining techniques, whereby rare gas fluorids are used.
e), ie pollutants such as AF _n molecules (where A = Kr, Ar or Xe, n = 2, 4 or 6) or other pollutants as mentioned above are removed from the gas mixture. . To this end, reference is made to US Pat. Nos. 4,534,034, 5,001,721, 5,111,473, 5,136,605 and 5,430,752. Incorporated into the description of this specification. Standard methods typically include a cold trap (cold t) before the gas is recirculated and returned to the discharge chamber.
rap) and freeze out the contaminants. Some of the contaminants frozen are the active noble gases of excimer lasers and molecular bonds of active gases such as halogen gases. Therefore, a significant amount of these important laser gases are removed from the gas mixture within the discharge chamber. The result is a sharp drop in rare gas and halogen gas concentrations, which has an undesirable effect on the output beam parameters.

In summary, the preferred embodiments provide methods and procedures for stabilizing the original or optimal gas composition of gas discharge lasers, especially excimer or molecular fluorine (F2) lasers. During long laser operation in operating or standby mode, laser gas depletion is due to high voltage, laser pulse shape, ASE, as well as accumulated electrical energy applied to the discharge, time and / or pulse count. , The elapsed time after a new fill, or in part, is continuously monitored by monitoring and controlling the additional laser parameters described above. With a known history and trend database of key operating parameters of various lasers operating under various conditions, processor control procedures are applied to replenish gas degradation. The stabilization process preferably relates to a parameter that changes in a known relationship to the specified time, the change in drive voltage, the input electrical energy to the discharge and / or the shot counting interval, a combination thereof, or the deterioration of the gas mixture. It involves the use of numerous small gas operations (micro-injections) that are performed based on some other interval of Careful combinations of μHI and MGR in varying amounts, and PGR are used to provide near complete stabilization of the laser gas mixture over a potentially unlimited unlimited duration.

Perhaps most importantly, the gas handling described herein is meaningful for lasers because it is smooth and controlled on the basis of an expert system that includes a wide variety of operating conditions of the laser system. It does not interfere with such output beam parameters or operation. Therefore, the laser can operate without interruption during the gas refilling operation and can perform industrial processing with high efficiency. Further details of suitable gas replenishment operations carried out in the wake of the first aspect of the present invention may be found in the 09 / 447,882 and 60 / 171,717 applications, and in the patent specifications mentioned above, for example US It is described elsewhere in Japanese Patent No. 4,997,573. Burst Control Algorithm A preferred embodiment relating to the second aspect of the invention will now be described below, in which the laser control computer or processor 16 preferably initiates the burst following each long burst pause. It is programmed by a self-learning algorithm to reduce the charging voltage for the pulse. In this way, the tendency of energy spikes or overshoots after a long burst pause, as illustrated in FIG. 1, is advantageously compensated for by lowering the high voltage for the initial pulse in the burst. In addition, the fast energy regulation loop achieves a nearly constant energy dose over the die site or wafer position. The output energy of a pulse or group of pulses in a burst is compared to a target energy or energy dose or moving average energy to change the moving energy average or energy dose over the burst following both short and long burst pauses. Minimize fluctuations, ie improve moving average energy stability or energy dose stability of the laser system.

In particular, according to this preferred embodiment preferably two energy control methods are used. The first algorithm, the overshoot algorithm, controls 1-10 pulses at the beginning of the burst, and the second algorithm, the control algorithm, controls the other pulses in the burst.

[0106]

[Table 1]

The energy control preferably comprises three tables as indicated above. Also, more than two or three tables may be used within the scope of this embodiment. Each table contains high voltage values for a number of initial pulses in the burst. In this example, the high voltage values for up to 10 pulses are held in the table, as shown in the table below. Table 1 is used for the start of the first burst after a long burst pause.

Table 2 is used for the start of the second burst after a long burst pause. -Table 3 is used for all other bursts. At laser start-up, all values in the table are preferably set to the intermediate HV. The value stored at the last laser stop is then preferably loaded, if available. When the laser is stopped, its actual value is stored on the disc.

In general, there are two normal burst pause periods: (1) long burst pauses and (2) short burst pauses. A short burst pause has a very short relative duration, eg about 20-50 ms. A long burst pause is usually much longer than the minimum time, eg 20 times the duration of a short burst pause.

The length of the short burst pause in units of ms can be set by software (typical value is 50 ms, but configurable to any selected amount (config
urable)). The length of a long burst pause is usually 2 times the length of a short burst pause.
It is set to 0 times. Thus, some exemplary burst pause definitions include, for example: That is, 0 to 50 ms for no pause, 50 ms to 20 × 50 ms = 1 s for short pause, and 1 s or more for long burst pause. The energy control device preferably has a timer that counts milliseconds. This counter is reset after each light pulse. When the counter reaches the number set as the length of the short burst pause, the short burst pause is recognized. When the counter reaches 20 times this number, a long burst pause is recognized.

During a long burst pause, the power supply charges to the value of HV given in the first position of Table 1, ie HV1, and prepares for the first pulse. When the first pulse is triggered and the value of HV1 from Table 1 is applied and the first pulse occurs, the resulting pulse energy is measured. If the energy is too low or too high, that is, if the measured pulse energy is different from the target energy, the HV value for the first pulse in Table 1 is the degree of variance from the target energy. And possibly some other factor, such as the result of a previous measurement of the first pulse, possibly after a previous long burst pause. That is, the power supply is charged to HV1 and the resulting measured energy of the first pulse is E1 _M. Then, the measured energy E1 _M is subtracted from the target energy E1 _T to obtain ± E1 = E1 _T −E1 _M. Next, ± HV1 = HV1 _new −HV1 is used, where ± HV1
Is calculated from ± E1 and ± HV1 is added to or subtracted from HV1 to replace HV1 in the first position of Table 1 by HV1 _new , etc. As a result, the first pulse of the first burst of the next burst sequence following a long burst pause approaches the target energy and the controller preferably performs continuous repetitions.
Continue learning the optimal HV setting by sive iteration).

This process preferably consists of a number of first pulses HV1 following the next long burst pause.
The same procedure of measuring the energy, comparing it with the target energy and updating Table 1 is carried out. The power supply then charges the high voltage HV2 applied to the second location of Table 1 in the same manner as just described with respect to the first pulse after the long burst pause, and enters the long burst pause. Prepare for the second pulse in the following first burst, and so on. That is, the same procedure is preferably performed for each initial pulse, for example the first 10 pulses as shown in the table, HV1
The value of ~ HV10 is calculated and updated. After a certain number of pulses configurable between 1 and more than 10 pulses exemplified in the table, the control mode is set to the second control algorithm (see below) for the remaining pulses of the first burst. Switch.

After the next short burst pause, which means for the second burst in the burst sequence following the long burst pause, a similar algorithm is applied except that Table 2 is used. It Again, the number of pulses in the second burst performed by the first algorithm is configurable, may be more or less than 10, and may be greater than the number of pulses used in Table 1 for the first burst. It can be high or low. Table 3 is preferably used after each next short burst pause, which means for all other bursts in the burst sequence. Again, the third algorithm in which the first algorithm is executed
The number of pulses in the burst is configurable and may be 10 or more or less than 10 and is the same as the number used in either Table 1 or Table 2,
Or it may be more or less. Preferably table 3 is used as this last table in the first control algorithm, but table 2 or the fourth or later table is followed by a long burst pause and the next long burst pause. It may also be used for this function which applies to all subsequent bursts before.

After the next long burst pause, everything is repeated starting from table 1. In this way, the energy controller will optimize the optimum H for the first pulse of every burst.
Learn V settings. However, note that after a long burst pause, the algorithm returns to Table 1 regardless of which table was used after the previous pause.

Preferably, an initial learning phase for the first 10 or so burst sequences is used, but a slower learning phase (sl) for subsequent burst sequences.
ower learning phase) is used. The fast learning phase is initiated when the energy setpoint is changed or the energy control mode is changed. The slow learning phase is preferably always active. That is, the amount of HV correction is preferably determined as follows.

For the first 10 or other number of burst sequences, fast learning is done because the HV is corrected in proportion to the energy deviation. □ For subsequent burst sequences, the HV is one or two or more least significant digits, eg two digits.
It is corrected by only or by adding or subtracting a small HV from the existing values in the table. That is, after the learning performed in the first few burst sequences, it is not necessary to make a large correction of the HV value in the table.

Preferably, the HV for the number of pulses n = 10 or less in the number m of bursts following a long burst pause is not strongly dependent on the energy of the immediately preceding pulse, but the same number in the previous burst sequence. Depends only on the pulse. Alternatively, the energy of both the preceding pulse and the same number of pulses in the previous burst sequence may be added to the calculation and determination of the HV for that pulse number in the subsequent burst number m.

The HV values in the table are determined such that the resulting pulse energy is approximately constant, eg, about 10 mJ for each initial pulse. However, the target energy of the laser may change, for example, from 9.5 mJ to 10.5 mJ. Therefore, the HV value is also changed to satisfy the actual target energy. This is preferably done with a gradient. That is, the HV value that produces a pulse with an energy of 9.5 mJ is determined, and the HV value that produces a pulse with an energy of 10.5 mJ is determined. The resulting 1
A slope of 00V / mJ is used to change the HV in the table. If the target energy is maintained at 10 mJ, the learning algorithm does not need to be modified in this way, nor is the burst that occurs after the target energy has been modified and adjusted. Description of Control Algorithm The HV value for the initial pulse in the burst is taken from the table and used to generate the initial pulse by the overshoot algorithm, then the control algorithm is used to determine the remaining values in each burst. The HV value for the pulse is set. The first embodiment of the control algorithm determines the energy deviation from the target energy for each pulse. The determined deviation is preferably multiplied by the amplification factor ( _kg ). This result is added to the HV used for the previous pulse and this value is used for the next pulse. This stabilizes the output pulse energy.

Alternatively, for each group of X pulses, the deviation from the target moving average energy is determined. This deviation is used to determine the HV for the next pulse or the HV value for some subsequent pulses. Alternatively, for each die site, multiple pulses may be incident during the scanning process. The die site and the total energy delivered to each die site, or energy dose, is determined and compared to the target energy dose.
The deviation is used to determine the HV for the next pulse or the HV value for some subsequent pulses. In this way, the moving average is associated with the energy dose. The sum of n pulse energies, eg 40 pulse energies, is n
The energy dose that is held constant for each package of individual subsequent pulses. This sum divided by the number of pulses in it is called the moving average.
This moves through the burst and can be illustrated as a time window that always contains the same number of pulses (except at the beginning of the burst). The initial pulse subject to the overshoot algorithm may or may not be used in this decision.

In the case of a stepper lithography system, a wafer stage
Note that) does not move during irradiation. Each pulse illuminates the entire area of the chip. In this case, the total energy dose, ie the sum of all pulse energies applied to the chip, is utilized, not the energy dose per die site. On the other hand, the scanner moves the wafer stage during irradiation. Each pulse illuminates a rectangular area that moves on the chip between pulses. The illuminated rectangles of subsequent pulses overlap. Each point on the chip is illuminated by several pulses, typically about 40 pulses. The different doses are preferably illuminated by different actual laser pulses, or in other words by pulses emitted at different times or divided portions of the same beam at the same time, but the dose is preferably on-chip. The same holds for each point or die site. That is,
The moving average is held constant.

In this respect, maintaining energy dose stability is a measure of pulse energy stability according to the preferred embodiment.
) Is different from maintaining. For example, over 40 pulses incident on the die site, the standard deviation varies more or less, but the energy dose remains the same. For example, at one point on the wafer, 20 alternating pulses of 9.9 mJ and 10.1 mJ are incident on the site, or 20 alternating pulses of 9 each.
． Suppose a pulse changing to 95 mJ and 10.05 mJ is incident on the site. In both cases a moving average of 10 mJ is observed, but the pulse energy stability is greater in the latter case than in the former. As the above examples show, as long as the total energy dose applied to each point on the wafer is about the same, the energy dose stability is high regardless of the pulse energy stability.

The moving average window can be set by the parameter w. That is, the controller 16 calculates the sum of the energy deviations of several pulses, eg i pulses, and multiplies this sum by the correction factor k _i to obtain the parameter w
Divided by, that is, the HV value is corrected by the coefficient s * ki / w. Since the parameter w is independent of the number of pulses i, the total energy deviation becomes more important as the number of pulses i increases. This sum is set to 0 each time a short or long burst pause is detected. This sum is preferably calculated as follows.

If 0 <i <(w−1), then s is the sum of the energy deviations of the first i pulses. Also, if i ≧ (w−1), s is the total energy deviation of the last (w−1) pulses. For example, w is set to 40 and the controller 16 calculates the sum of the deviations of the first 39 pulses in the burst, and thus by the first bullet above (ie,
Correct the 40th pulse (by the correction factor k _i ). The same is done for earlier pulses in the burst, i.e. for example the first one in the burst.
The sum of the deviations of the 5 pulses is calculated, thus correcting the HV of the 16th pulse. Then, for each pulse after the 40th pulse, only the previous 39 pulses are used due to the second bullet above. That is, the dose stability according to the control algorithm of the present invention is advantageously improved.

In this preferred control algorithm, the HV value for a pulse in a burst, and the corrections and updates made to it, are determined by the preceding pulse (pr) in that burst.
Note that it only depends on the eceding pulse) and not on the previous burst sequence. Moreover, overshoot control is simultaneously enhanced by the overshoot algorithm. That is, the overshoot algorithm is advantageously used to control the energy of the initial pulse in a burst, while the control algorithm uses the energy or moving average energy or energy dose corresponding to the pulse in the burst after the initial pulse. Is advantageously used to control That is, the control algorithm and the overshoot algorithm are advantageously combined for energy, energy stability, energy dose stability and energy overshoot control.

The laser system can be configured to operate in various modes. If the laser is set to operate in energy burst mode, the overshoot and control algorithms discussed above are preferably used. When the laser is set to the constant energy mode, it is assumed that the laser does not use bursts, and thus there is no overshoot such that HV compensation is used, so the control algorithm is as discussed above. , The overshoot algorithm is not used. Laser is H
It may be set to operate in the constant V mode, but in this case, energy control by changing the HV value does not exist. HV is set to a constant value for all pulses. Lasers operating at constant energy may use energy control algorithms that rely on changes in other parameters other than HV, which would otherwise result in poor application processi due to changes in energy.
ng) will be invited.

Some specific laser modes of operation include: External HV mode, where the internal energy control device 16 is not used and H
The V value is calculated directly by the external controller. ESF mode (exposure sensor feedback), where instead of using an internal laser energy sensor, the energy sensor in the scanner is used. The energy value is sent to the energy controller 16 for each pulse, for example, via a parallel interface. The rest of the control algorithm may be left unchanged.

ECI mode (energy control interface), where a serial interface is used and the controller parameters and energy settings are changed. This system is normally used only during burst pauses. The regulation loop for constant transfer energy average according to the preferred embodiment includes charge voltage regulation for each shot at the beginning of a burst. In the preferred laser system, the change in charging voltage that is made is sufficiently small that it does not significantly affect the other laser output beam parameters (09 / 447,882 and 60 for further discussion in this regard). / 171,717 application).

The average pulse energy or energy dose or moving average depends on the individual irradiation step. For example, different resists can accommodate different energy doses used. The target energy is achieved by a corresponding adjustment of the charging voltage. Normal charging voltage is in the range of 1.4-2.0 kV. Pulse energy conditioning can be accomplished by voltage regulation, preferably about 100-400V.

While the exemplary drawings and specific embodiments of the present invention have been illustrated and illustrated, it will be understood that the scope of the invention is not limited to the specific embodiments discussed herein. That is, the examples should be considered to be illustrative rather than limiting, and by one of ordinary skill in the art from the scope of the invention as claimed in the following claims and their equivalents. It should be understood that variations can be made in these embodiments without departing.

Further, in the following method claims, the steps are arranged in a selected typographical order. However, this order is chosen and arranged as such for the convenience of printing, except where the particular order of steps is explicitly stated or as required by one of ordinary skill in the art Is not meant to imply any particular order for performing.

[Brief description of drawings]

FIG. 1 illustrates energy versus pulse number for a pulse of a burst from a pulsed gas discharge laser with an input high voltage held constant during the burst.

FIG. 2 is a schematic diagram of an excimer or molecular fluorine laser system according to a preferred embodiment.

[Figure 3]   It is a figure which illustrates the outline of the gas control unit by a preferred embodiment.

4 is a diagram illustrating a schematic of a gas line for halogen injection into the discharge chamber of the laser of FIG. 2 using an accumulator.

FIG. 5 shows a graph of charge voltage versus time and stored energy applied to discharge, which also shows periodic halogen injection for a laser system according to a preferred embodiment.

FIG. 6 shows a graph of charge voltage versus time that also shows periodic halogen injection for a laser system according to the preferred embodiment.

FIG. 7 shows a graph of pulse energy stability (standard deviation, upper graph) versus time and moving average (maximum and minimum over 40 pulse intervals) for a laser system operating at 2 KHz, according to a preferred embodiment. Is.

FIG. 8: Periodic micro-halogen injection (HI) for the system according to the preferred embodiment.
FIG. 4 is a diagram showing a qualitative graph of drive voltage vs.

FIG. 9 is a qualitative graph of drive voltage versus pulse count also showing periodic halogen injection, mini gas replacement and partial gas replacement for the system according to the preferred embodiment.

FIG. 10 is a flow chart for performing halogen injection, mini gas replacement and partial gas replacement according to the preferred embodiment.

─────────────────────────────────────────────────── ─── Continued front page    (31) Priority claim number 09 / 447,882 (32) Priority date November 23, 1999 (November 23, 1999) (33) Priority claiming countries United States (US) (31) Priority claim number 60 / 171,717 (32) Priority date December 22, 1999 (December 22, 1999) (33) Priority claiming countries United States (US) (31) Priority claim number 09 / 484,818 (32) Priority date January 18, 2000 (January 18, 2000) (33) Priority claiming countries United States (US) (81) Designated countries DE, JP

Claims (81)

[Claims] 1. A gas discharge laser system comprising: a discharge chamber having a plurality of electrodes therein, containing a gas mixture containing one or more components to be consumed; and energizing the gas mixture. A power supply circuit coupled to the electrodes to provide a laser beam, a resonator for generating a laser beam, means for monitoring the stored energy applied to the discharge of the laser as a reference for the gas mixture state, Means for replenishing the gas mixture based at least in part on the monitored stored energy being stored. 2. The method of claim 1 further comprising means for monitoring time as an additional measure of said gas mixture status, wherein said replenishing means replenishes said gas mixture also based on said monitored time. The system described. 3. The method of claim 1, further comprising means for monitoring a charging voltage as an additional measure of the gas mixture state, wherein the replenishing means replenishes the gas mixture also based on the monitored charging voltage. Or the system according to 2. 4. Further comprising means for monitoring changes in charging voltage as an additional criterion for said gas mixture status, wherein said replenishing means replenishes said gas mixture also based on said monitored changes in charging voltage. The system according to claim 1 or 2, which comprises: 5. The method of claim 1, further comprising means for monitoring the stored energy dose, wherein the monitoring means controls the charging voltage based on the monitored stored energy dose. System. 6. The system according to claim 1 or 2, wherein the replenishing means comprises means for injecting into the discharge chamber components of the gas mixture to be consumed to replenish the components of the gas mixture. . 7. The system of claim 6, further comprising means for ejecting a portion of the gas mixture from the discharge chamber. 8. The system of claim 6, wherein the component is a halogen containing molecular material. 9. The system of claim 6, wherein the implanting occurs in response to a detected increment of stored energy applied to the discharge. 10. The replenishing means injects a dilute mixture containing components to be consumed of 0.0001 mbar to 0.5 mbar at predetermined intervals so that the partial pressure of the components is approximately equal to the initial pressure in the discharge chamber. 3. The system of claim 1 or 2 returning to partial pressure. 11. The system of claim 10, wherein the component is a halogen containing molecular material. 12. The system of claim 10, wherein the implanting occurs in response to a detected increment of stored energy applied to the discharge. 13. The monitoring means includes a processor and the replenishment means comprises:
A gas supply unit connected to the discharge chamber and in communication with a processor, wherein the processor may inject a diluted mixture containing 0.0001 mbar to 0.5 mbar halogen-containing molecular material into the discharge chamber at predetermined intervals. 3. The system according to claim 1 or 2, wherein the valve is controlled at predetermined intervals so as to do so. 14. The system of claim 13, wherein the spacing is based at least in part on increments of stored energy applied to the discharge. 15. A gas discharge laser system, comprising: a discharge chamber having a plurality of electrodes therein, containing a gas mixture containing one or more components to be consumed; and energizing the gas mixture. A power supply circuit coupled to the electrode to provide a laser beam, a resonator for generating a laser beam, a gas control unit for replenishing the gas mixture, and a gas mixture state applied to the discharge of the laser as a reference A processor in communication with the power supply circuit for monitoring stored energy, the processor responsive to the gas mixture based at least in part on the monitored stored energy applied to the discharge. A gas discharge laser system for controlling the gas control unit to replenish. 16. The processor of claim 15, wherein the processor further monitors time as an additional measure of the gas mixture status and the gas control unit replenishes the gas mixture also based on the monitored time. System. 17. The processor further monitors a charging voltage as an additional measure of the gas mixture status, wherein the gas control unit replenishes the gas mixture also based on the monitored charging voltage. Item 17. The system according to Item 15 or 16. 18. The processor further monitors a change in charging voltage as an additional measure of the gas mixture state, wherein the gas control unit also determines the gas mixture based on the monitored change in charging voltage. 17. A system according to claim 15 or 16 which replenishes. 19. The processor further monitors the stored energy dose,
17. The system of claim 15 or 16, wherein the processor controls a charging voltage based on the monitored stored energy dose. 20. The gas control unit further comprises a valve controlled to inject a predetermined amount of components of the gas mixture to be exhausted into the discharge chamber to replenish constituent gases. The system according to claim 15 or 16, which comprises: 21. The component is injected at regular intervals into a dilute mixture containing 0.0001 mbar to 0.5 mbar of the component to bring the partial pressure of the component back to about the initial partial pressure in the discharge chamber. 21. The system of claim 20, wherein 22. The system of claim 21, wherein the component is a halogen containing molecular material. 23. The system of claim 21, wherein the implanting occurs in response to a detected increment of stored energy applied to the discharge. 24. The system of claim 21, wherein the gas control unit further enables release of a portion of the gas mixture from the discharge chamber. 25. The system according to any one of claims 1-2 or 15-16, wherein said laser is an excimer laser. 26. The system according to any of claims 1-2 or 15-16, wherein said laser is an F2 laser. 27. A method of initially stabilizing, in operation, a gas mixture having an initial composition in a discharge chamber of a gas discharge laser, the gas mixture comprising one or more constituent gases to be consumed, said method comprising: Monitoring the stored energy applied to the discharge of the laser; and determining the state of the gas mixture based at least in part on the monitored stored energy applied to the discharge. How to stabilize during operation. 28. The method of claim 27, further comprising monitoring time and determining the condition of the gas mixture also based on the monitored time. 29. Further comprising the step of monitoring a charging voltage as an additional measure of the state of said gas mixture, wherein the step of replenishing said gas mixture also based on said monitored charging voltage. Item 29. The method according to Item 27 or 28. 30. A replenishment step for replenishing the gas mixture based on also the monitored change in charging voltage, further comprising the step of monitoring a change in charging voltage as an additional measure of the state of the gas mixture. 27 or 28 including
The method described in. 31. The method of claim 27 or 28, further comprising: monitoring a stored energy dose and controlling a charging voltage based at least in part on the monitored stored energy dose. The method described. 32. The method according to claim 27, further comprising injecting a diluted mixture containing a constituent gas to be consumed into the discharge chamber at predetermined intervals to replenish the constituent gas. 33. The step of injecting comprises selecting a halogen-containing molecular substance as the component gas, and 0.0001 mbar contained in the diluted mixture.
33. Choosing an amount of the component gas of ~ 0.5 mbar. 34. The method of claim 32, further comprising selecting the predetermined interval to be an increment of stored energy applied to the discharge. 35. The method of claim 32, further comprising discharging a portion of the gas mixture from the discharge chamber. 36. The method further comprising controlling the composition of gases of the gas mixture in the discharge chamber based at least in part on the monitored stored energy applied to the discharge. 27. The method according to 27 or 28. 37. The method according to claim 27 or 28, wherein the laser is an excimer laser. 38. The method of claim 27 or 28, wherein the laser is an F2 laser. 39. Further depletion of the gas mixture to replenish the constituent gases based at least in part on the determined state of the gas mixture as indicated by the stored energy applied to the discharge. 29. The method of claim 27 or 28, comprising injecting the component gas of interest into the discharge chamber. 40. The method of claim 39, wherein the injecting step is responsive to a detected increment of stored energy applied to the discharge. 41. The method of claim 39, further comprising discharging a portion of the gas mixture from the discharge chamber. 42. A method of initially stabilizing a gas mixture, having an initial composition, in a discharge chamber of a gas discharge laser, the gas mixture comprising a first constituent gas to be depleted during operation, said laser discharge Stabilizing the gas mixture during operation, comprising: monitoring a stored energy applied to the gas mixture; and adjusting the gas mixture based at least in part on the monitored stored energy applied to the discharge. How to make. 43. The method of claim 42, further comprising monitoring time and adjusting the gas mixture also based on the monitored time. 44. The method further comprises monitoring a charge voltage as an additional measure of the condition of the gas mixture, wherein the step of replenishing the gas mixture also based on the monitored charge voltage. Item 42. The method according to Item 42 or Item 43. 45. A replenishment step for replenishing the gas mixture based on the monitored change in charging voltage as well as the step of monitoring a change in charging voltage as an additional measure of the state of the gas mixture. 42 or 43, including
The method described in. 46. The method of claim 42 or 43, further comprising monitoring a stored energy dose and controlling a charging voltage based at least in part on the monitored stored energy dose. the method of. 47. The method according to claim 42, further comprising injecting a diluted mixture containing a constituent gas to be consumed into the discharge chamber at predetermined intervals to replenish the constituent gas. . 48. The step of injecting comprises selecting a halogen-containing molecular substance as the component gas, and 0.0001 mbar contained in the diluted mixture.
48. The method of claim 47, comprising selecting an amount of the component gas of ˜0.5 mbar. 49. The method of claim 47, further comprising the step of selecting the predetermined interval to be an increment of stored energy applied to the discharge. 50. The method of claim 47, further comprising releasing a portion of the gas mixture from the discharge chamber. 51. The method further comprising controlling the composition of the gas of the gas mixture in the discharge chamber based at least in part on the monitored stored energy applied to the discharge. Or the method according to 43. 52. The method of claim 42 or 43, wherein the laser is an excimer laser. 53. The method of claim 42 or 43, wherein the laser is an F2 laser. 54. Further, the component gas is replenished based on the determined state of the gas mixture, which is at least partially indicated by the stored energy applied to the discharge, and is therefore subject to exhaustion of the gas mixture. 44. The method of claim 42 or 43, comprising injecting the component gas into the discharge chamber. 55. The method of claim 54, wherein the step of injecting is performed in response to a detected increment of stored energy applied to the discharge. 56. The method of claim 54, further comprising discharging a portion of the gas mixture from the discharge chamber. 57. A method of controlling the output energy of successive pulses in a burst of pulses from a gas discharge laser operating in burst mode, wherein each burst is followed by a long burst pause, depending on the specifications of the application process. Measuring the energy of at least a predetermined number of initial pulses of the first burst occurring after a long burst pause, characterized by emitting said burst of several pulses followed by one of the short burst pauses; Based on the energy of the measured initial pulse that will bring the output energy of the laser corresponding to each of the initial pulses to about the same predetermined value for the first burst following the burst pause. Calculating the value of the input voltage for each of the initial pulses; So that the pulse delivered has approximately the same predetermined output energy value as described above,
Applying an input voltage corresponding to the calculated value during the subsequent first burst after the same long burst pause, and controlling the output energy. 58. The long burst pause is longer than 100 ms.
The method described in. 59. The method of claim 57, wherein the long burst pause is longer than 1 second. 60. Measuring the energy of at least a predetermined number of initial pulses of at least one second burst resulting in a short burst pause after the first burst following the long burst pause, and a similar long burst pause. For the subsequent second burst following a similar short burst pause after the first burst following, the output energy of the laser corresponding to each of the initial pulses of the second burst can be brought to approximately the same predetermined value. The second above
Calculating the value of the input voltage for each of the initial pulses of the burst, such that the generated pulses have approximately the same predetermined output energy values as above.
Applying an input voltage corresponding to the calculated value for the second burst in the subsequent second burst following the similar short burst pause after the first burst following the similar long burst pause. 58. The method of claim 57, comprising: 61. measuring the energy of at least two short burst pauses followed by at least two short burst pauses and at least a predetermined number of initial pulses of at least one third or subsequent burst of each burst. , The output energy of the laser corresponding to each of the initial pulses of the third or subsequent burst for one or more subsequent bursts occurring after at least two short burst pauses following a long burst pause. Calculating the value of the input voltage for each of the initial pulses of the third or subsequent bursts that will lead to the same predetermined value; and the pulses produced each have substantially the same predetermined output energy value. To have the at least two short bursts following the long burst pause. 61. Applying an input voltage corresponding to the value calculated for the one or more subsequent bursts that occur after a first pause. 62. Further, the subsequent bursts after the at least two short burst pauses after the long burst pauses to produce additional bursts each with an initial pulse having the substantially same predetermined value. 62. Repetitively applying the input voltage corresponding to the calculated value to a burst.
The method described in. 63. Furthermore, an output laser corresponding to the latter pulse in the burst in addition to the initial pulse.
Measuring energy, said energy after a predetermined number of initial pulses, which will bring the output energy dose of said laser corresponding to the sum of the pulse energy of successive pulses to approximately the same predetermined energy dose value. Calculating the value of the input voltage corresponding to each of the later pulses, and the output energy dose so that the energy doses associated with the generated pulses each have said substantially same predetermined energy dose value. 62. Applying an input voltage corresponding to said calculated value to bring an amount to said approximately the same predetermined value. 62. A method according to any one of claims 57, 60 or 61. 64. The method of claim 63, wherein the continuous pulse corresponds to substantially all laser output radiation pulses incident on a single location in an application process. 65. Further comprising measuring the energy of the initial pulse of the subsequent first burst, and the subsequent first burst for another subsequent first burst following another similar long burst pause. Of the input voltage for the initial pulse based on the measured initial pulse energy of the subsequent burst that will bring the output energy of the laser corresponding to each of the initial pulses to approximately the same predetermined value. Calculating different values, so that the generated pulses have approximately the same predetermined output energy values as above,
58. The method of claim 57, comprising applying an input voltage corresponding to the calculated different value during the other subsequent first burst after the other similar long burst pauses. 66. In an energy control software algorithm for controlling the output energy of successive pulses during a burst of pulses from a gas discharge laser operating in burst mode, according to the application process specifications after each burst. Emitting laser bursts of several pulses followed by one of a long burst pause and a short burst pause, each output laser pulse after a long burst pause having approximately the same energy value A first table of input voltage values to be read by a processor that signals a power supply circuit to apply a voltage according to the voltage values in the first table to generate an initial pulse in a subsequent first burst of The input voltage value in the first table is long. At least 1 after pause
From the measured data of the initial pulse in the previous two first bursts, a first table in which each said initial pulse was calculated to generate said initial pulse at said approximately the same energy value, and after a long burst pause The power supply circuit to apply a voltage according to the voltage value in the second table to produce an initial pulse in a subsequent second burst of output laser pulses following a short burst pause after the first burst of A second table of input voltage values to be read by the processor, wherein the initial pulses in the subsequent second burst each have approximately the same energy value, the input in the second table A measurement of the initial pulse in at least one previous second burst following a short burst pause after the first burst after a long voltage pause. A second table in which each said initial pulse is calculated from the determined data to produce said initial pulse at said approximately the same energy value, and an energy control software algorithm. 67. Further, in the third table to generate an initial pulse in a subsequent third or subsequent burst following a short burst pause after the first and second bursts after a long burst pause. The third table of input voltage values to be read by the processor to signal the power supply circuit to apply a voltage according to a voltage value, wherein the initial pulse in the subsequent third or subsequent burst is At least one previous third or each of which has said approximately the same energy value and whose input voltage value in said third table follows a short burst pause after a first burst and a second burst after a long burst pause; From the measured data of the initial pulse in the subsequent bursts, each initial pulse is calculated to produce the initial pulse at approximately the same energy value. 67. The energy control software algorithm of claim 66, comprising a third table that is customized. 68. The voltage value in the third table for generating the initial pulse in some bursts immediately following a short burst in addition to the third or subsequent bursts. 68. The method of claim 67, wherein the initial pulse in each of the plurality of bursts, each read by the processor for signaling the power supply circuit to apply a voltage according to, has the substantially same energy value. Energy control algorithm. 69. The third table further comprises: in addition to the third or subsequent bursts, other than the second burst following the short burst pause after the first burst after the long burst pause. Read by the processor to signal the power supply circuit to apply a voltage according to the voltage value in the third table to generate an initial pulse in all other bursts immediately following a short burst pause, 68. The energy control algorithm of claim 67, wherein the initial pulses in each of all the other bursts each have the substantially same energy value. 70. The first table measuring the energy of an initial pulse of a first burst following a long burst pause, and each said initial pulse for a subsequent first burst following a similar long burst pause. Calculating the value of the input voltage for the initial pulse based on the measured energy of the initial pulse, which will bring the output energy of the laser to approximately the same predetermined value. Storing the calculated value of the input voltage for the initial pulse as the first table so that the pulses generated by the input voltage value stored in have the same predetermined output energy value. 67. The algorithm of claim 66, generated by. 71. The first table further comprising: measuring the energy of an initial pulse of a subsequent first burst following a subsequent long burst pause; and another subsequent first burst following another subsequent long burst pause. For a burst, the initial based on the energy of the measured initial pulse of the subsequent first burst that will bring the output energy of the laser corresponding to each initial pulse to about the same predetermined value. Calculating the value of the input voltage of the pulse, the first burst following so that the pulse generated by the input voltage value stored in the first table has substantially the same predetermined output energy value. Using the calculated value of the input voltage for the initial pulse based on the energy of the measured initial pulse of Updating the value in the first table of the input voltage for the initial pulse in the first table;
71. The algorithm of claim 70, updated by. 72. An energy control software algorithm for controlling the output energy of a continuous pulse during a burst of pulses from a gas discharge laser operating in burst mode, according to the specifications of the application process after each burst. Characterized in that it emits the burst of several pulses followed by one of a long burst pause and a short burst pause, each output laser pulse after a long burst pause having approximately the same energy value. Said table of input voltage values to be read by a processor which signals the power supply circuit to apply a voltage according to the voltage values in the first table to generate an initial pulse in a subsequent first burst of pulses; , A burst pause in which the input voltage value in the first table is long From the measured data of the initial pulse in at least one previous first burst after, each said initial pulse comprises a table calculated to generate said initial pulse at said approximately the same energy value, wherein: , Said table measuring the energy of the initial pulse of the first burst following a long burst pause, and the output of said laser corresponding to each said initial pulse for a subsequent first burst following a similar long burst pause Calculating a value of the input voltage for the initial pulse based on the measured energy of the initial pulse that will bring the energy to approximately the same predetermined value; and storing the value of the input voltage for the initial pulse in the first table. The initial pulse is changed so that the pulse generated by the input voltage value has almost the same predetermined output energy value. Steps and energy control software algorithms generated by storing the calculated value of the input voltage as said table. 73. The table further comprising: measuring the energy of the initial pulse of a subsequent first burst following a subsequent long burst pause, and another subsequent first burst following another long burst pause. On the other hand, for the initial pulse based on the energy of the measured initial pulse of the subsequent first burst that will bring the output energy of the laser corresponding to each initial pulse to approximately the same predetermined value. Calculating a value of the input voltage, the pulse of the input voltage value stored in the first table having substantially the same predetermined output energy value, and the pulse of the subsequent first burst. Using the calculated value of the input voltage for the initial pulse based on the measured initial pulse energy of 73. The algorithm of claim 72, updated by updating the values in the table of the input voltage for the initial pulse in a table. 74. An energy control software algorithm for controlling the output energy of successive pulses during a burst of pulses from a gas discharge laser operating in burst mode, the application process specification after each burst. The output laser after a long burst pause in which each output laser pulse has approximately the same energy value, characterized in that the output laser emits a burst of several pulses followed by one of a long burst pause and a short burst pause. A second of said input voltage values to be read by a processor which signals the power supply circuit to apply a voltage corresponding to the voltage value in the first table to generate the initial pulse in the subsequent first burst of pulses. 1 table and a short bar after the first burst after a long burst pause To be read by the processor which signals the power supply circuit to apply a voltage according to the voltage value in the second table to generate an initial pulse in a subsequent second burst of output laser pulses following a power pause. A second table of input voltage values for the energy control software algorithm. 75. Further, in the third table to generate an initial pulse in a subsequent third or subsequent burst following a short burst pause after the first and second bursts after a long burst pause. 75. Energy control software according to claim 74, comprising said third table of said input voltage values to be read by said processor which signals said power supply circuit to apply a voltage according to a voltage value.
algorithm. 76. The third table further comprises: in the third table for generating an initial pulse in some bursts immediately following a short burst pause in addition to the third or subsequent bursts. 76. The energy control of claim 75, wherein the initial pulse in each of the some bursts read by the processor that signals the power supply circuit to apply a voltage according to a voltage value, each has the substantially same energy value. algorithm. 77. The third table further comprises: in addition to the third or subsequent bursts, other than the second burst following the short burst pause after the first burst after the long burst pause. Read by the processor to signal the power supply circuit to apply a voltage according to the voltage value in the third table to generate an initial pulse in all other bursts immediately following the short burst, 76. The energy control algorithm of claim 75, wherein the initial pulses in all other bursts each have the approximately same energy value. 78. The first table measuring the energy of an initial pulse of a first burst following a long burst pause, and each said initial pulse for a subsequent first burst following a similar long burst pause. Calculating the value of the input voltage for the initial pulse based on the measured energy of the initial pulse, which will bring the output energy of the laser to approximately the same predetermined value. Storing the calculated value of the input voltage for the initial pulse as the first table such that the pulse generated by the input voltage value stored therein has substantially the same predetermined output energy value. 75. The algorithm of claim 74, generated by. 79. The first table further comprises: measuring the energy of the initial pulse of a subsequent first burst following a subsequent long burst pause; and another subsequent first burst following another subsequent long burst pause. For a burst, the initial based on the energy of the measured initial pulse of the subsequent first burst that will bring the output energy of the laser corresponding to each initial pulse to about the same predetermined value. Calculating a value of the input voltage for a pulse, the first burst following so that the pulse generated by the input voltage value stored in the first table has substantially the same predetermined output energy value. Using the calculated value of the input voltage for the initial pulse based on the energy of the measured initial pulse of 79. The algorithm of claim 78, updated by updating the value in the first table of the input voltage for the initial pulse in the first table. 80. An energy control software algorithm for controlling the output energy of successive pulses during a burst of pulses from a gas discharge laser operating in burst mode, the application process specification after each burst. The output laser after a long burst pause in which each output laser pulse has approximately the same energy value. The table of input voltage values to be read by a processor which signals the power supply circuit to apply a voltage according to the voltage values in the first table to generate an initial pulse in a subsequent first burst of pulses. With the table following a long burst pause Measuring the energy of the initial pulse of the first burst and bringing the output energy of the laser corresponding to each of the initial pulses to approximately the same predetermined value for subsequent first bursts following a similar long burst pause. Calculating a value of the input voltage for the initial pulse based on the measured energy of the initial pulse, which will be substantially equal to the pulse generated by the input voltage value stored in the first table. Storing the calculated value of the input voltage for the initial pulse as the table so as to have the same predetermined output energy value, the energy control software algorithm generated by. 81. The table further comprising: measuring the energy of the initial pulse of a subsequent first burst following a subsequent long burst pause, and another subsequent first burst following another long burst pause. On the other hand, based on the energy of the measured initial pulse of the subsequent first burst, which will cause the output energy of the laser corresponding to each initial pulse to reach substantially the same predetermined value, Calculating the value of the input voltage and the measuring of the subsequent first burst such that the pulses generated by the input voltage values stored in the first table have substantially the same predetermined output value. The first table using the calculated value of the input voltage for the initial pulse based on the energy of the initial pulse generated Updating the values in the table of the input voltage for the initial pulse in
The algorithm according to claim 80.