TW564585B - Line selected F2 two chamber laser system - Google Patents

Abstract

An injection seeded modular gas discharge laser system capable of producing high quality pulsed laser beams at pulse rates of about 4,000 Hz or greater and at pulse energies of about 5 mJ or greater. Two separate discharge chambers are provided, one of which is a part of a master oscillator producing a very narrow band seed beam which is amplified in the second discharge chamber. The chambers can be controlled separately permitting separate optimization of wavelength parameters in the master oscillator and optimization of pulse energy parameters in the amplifying chamber. A preferred embodiment in a F2 laser system configured as a MOPA and specifically designed for use as a light source for integrated circuit lithography. In the preferred MOPA embodiment, each chamber comprises a single tangential fan providing sufficient gas flow to permit operation at pulse rates of 4000 Hz or greater by clearing debris from the discharge region in less time than the approximately 0.25 milliseconds between pulses. The master oscillator is equipped with a line selection package for selecting the strongest F2 spectral line.

Description

564585 A7 _B7___ V. Description of the Invention (1) The present invention claims priority to the following patent applications: Patent Application No. 10 / 056,619 (filed on January 23, 2002), Patent Application No. 10 / 036,676 (12.2001) Filed on May 21), patent application number 10 / 036,727 (filed on December 21, 2001), patent application number 10 / 012,002 (filed on November 30, 2001), patent application number 10 / 006,913 (2001 Application on November 29), Patent Application No. 10 / 000,991 (filed on November 14, 2001), Patent Application No. 09 / 970,503 (filed on October 3, 2001), Patent Application No. 09 / 943,343 ( Filed on August 29, 2001), Patent Application No. 09 / 879,311 (filed on June 6, 2001), Patent Application No. 09 / 855,310 (filed on May 14, 2001), Patent Application No. 09 / 854,097 (Application dated May 11, 2001), patent application number 09 / 848,043 (filed on May 3, 2001), patent application number 09 / 829,475 (filed on April 9, 2001), patent application number 09 / 771,789 (filed on January 29, 2001), patent application number 09 / 768,753 (filed on January 23, 2001), patent application number 09/684, 629 (filed on October 6, 2000), patent application number 09 / 473,795 (filed on December 28, 1999), patent application number 09 / 473,852 (filed on December 27, 1999), patent application number 09 / 459,165 (filed on December 10, 1999), patent application number 09 / 309,478 (filed on May 10, 1999). The present invention relates to gas discharge lasers and, in particular, high-repetition-rate fluorine lasers. Background of the Invention Gas Discharge Laser 4 (Please read the notes on the back before filling out this page) This paper size applies to Chinese National Standard (CNS) A4 (210X297 mm). It has been a long time since the laser was described in the 196G towel. Discharge lasers are well known and can be used. The high $ discharge between the two electrodes excites the laser gas to produce a gas gain medium. The resonant cavity containing this gain medium allows the stimulated amplification of light to be enhanced and can be taken out from the resonant cavity in the form of a laser beam. Many such gas discharge lasers operate in pulsed mode. Excimer laser + Excimer laser is a special form of gas discharge laser and has been well known since the mid-1970s. The description of the excimer laser (which is useful for lithography of integrated circuits) is described in US Patent Case No. usp 5,023,884 “Compact Stimulated Excimer Knife Field Shot” published on June 1991 This patent is issued to the employer of the applicant of this case, and this patent is incorporated herein by reference. The excimer laser described in patent case '884 is a high-repetition type pulsed laser. These fork excimer lasers, when used in integrated circuit lithography, are typically operated in a production line 24 hours a day (“ar〇und_the_clOck”) to produce at every 犄Thousands of valuable integrated circuits. Therefore, downtime (d0Wn-time) is very expensive. For this reason, most components are organized into modules that can be replaced in minutes. Used in Lithography excimer lasers typically have to reduce the bandwidth of their output beam to a fraction of a micron (10-12m). A gas discharge laser in the form described in patent '884 uses a Electrical pulsed power system to generate between two electrodes Electrical discharge. In this conventional technology system, the DC power supply will charge the f container bank called "charging capacitor" or "c0" to the set and controlled 564585 A7 B7 V. Description of the invention (3) It is called the voltage of the charging voltage for each pulse. The magnitude of the charging voltage in these conventional technology units is in the range of about 500 volts of 1000 volts. After C0 is charged to a preset voltage, the solid-state switch Turning off allows the electrical energy stored on CO to flow very quickly through a series of magnetic compression circuits and transformers to generate a high voltage potential in the range of approximately 16,000 V (or greater) across the electrodes, which lasts approximately 20 to 50 nanoseconds (Ns) discharge. The main advances in lithography light sources, as described in Patent Case No. '884, became the main light source for integrated circuit lithography during the period 1989-2001. .More than 1,000 such lasers are being used in the most modern integrated circuit manufacturing plants. Almost all of these lasers have the basic design features described in Patent Case No. 184, These are: (1) A single-pulse power system is used to provide electrical pulses across, and its pulse rate is about 100 to 2500 pulses / second. (2) A single resonant cavity is composed of: a partial mirror output coupler; And a line narrowing unit, which consists of a chirped light diffuser, an adjustment mirror, and a grating. (3) A single discharge chamber contains a laser gas (KrF or ArF), two extension electrodes, and a tangential fan for The laser gas flows between the two electrodes fast enough between the pulses to clear the discharge area, and (4) the light monitor is used to monitor the pulse energy, wavelength and frequency bandwidth of the output pulse, and the feedback control system is used to Pulse-to-pulse based control of pulse energy, energy dose, and wavelength. During 1989 ~ 2001, the output power of these lasers gradually increased. This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 6 (Please read the precautions on the back before filling this page).可 | 564585 A7 _B7_ V. Description of the Invention (4) And the light quality specifications for pulse energy stability, wavelength stability, and frequency bandwidth have become increasingly strict. The operating parameters of this general lithography laser mode widely used in integrated circuit manufacturing include: pulse energy of 8mJ, pulse rate of 2500 pulses / second (providing average light power up to approximately 20 watts), and a frequency band width of approximately 0.5pm (FWHM) and pulse energy stability were ± 0.35%. F2 laser F2 laser is well known. These lasers are similar to KrF and ArF lasers. The basic difference is a gas mixture, which is a small portion of F2 in an F2 laser and has helium and / or neon as a buffer gas. The natural output spectrum of this F2 laser is focused on two narrow-bandwidth spectral lines: a relatively strong spectral line is concentrated at about 157.63 nanometers (nm), and a relatively weak spectral line is concentrated at about 157.52 nanometers. F2 laser radio frequency band width A typical KrF laser has a natural frequency band width of about 300pm (10-12m), which is a full width half maximum (FWHM) of about 248 nanometers and is used for lithography. It typically reduces the line width to less than 0.6 pm (in this specification, unless otherwise stated, the bandwidth value will be referred to as the FWHM bandwidth). ArF lasers have a frequency band width of about 500pm concentrated on about 193nm and typically narrow the line width to less than 0.5pm. These lasers can be adjusted relatively easily in a large part of their natural frequency band width using the grating-based optical narrowing module mentioned above. As explained above, F2 lasers typically produce laser light. Most of their energy is concentrated in two broad spectrums of approximately 157.63nm and 157.52nm. (Please read the precautions on the back before filling this page) This paper The scale applies to the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 ____B7 V. Description of the invention (5) In general (herein commonly referred to as "spectral lines,"). Often, the weaker of the two spectral lines (That is, the 157.52nm line) is suppressed and forced to operate in the 157.63nm line. The natural band width of the 157 · 63nm line depends on the pressure and gas content, and changes from about 6:00 pm to l_29pm (FWHM). Lasers with a band width in this range can be used with lithography devices. This device has a reflective-refractive lens design that uses refractive and reflective optics, but for total refractive lenses this laser light is designed. The frequency bandwidth may have to be reduced to about 0.1 pm to produce the desired result. Seed injection (light) A well-known technique for reducing gas discharge laser systems ( Including the excimer laser system), which involves injecting a narrow-band "seed light" into the gain medium. In this system, the laser that generates this seed light is called the "main oscillator", which is designed at The first gain medium provides a very narrow band of light, and this light is used in the second gain medium. This system is called the main oscillator power amplifier system (Μόρα). If the second gain medium itself has a resonant cavity ( Laser resonance is generated in it), this system is called injection seed resonator (ISO) system or main oscillator power oscillator (ΜΟΟ) system, in which case this seed laser is called main oscillator and this downstream The system is called a power oscillator. The laser system composed of these two systems tends to be more expensive, larger and more complicated than the comparable single-chamber laser system. Therefore, this two-chamber laser system Commercial applications are limited. Therefore, what is needed is a better laser paper size for pulsed gas discharge F2 lasers, which applies the Chinese National Standard (CNS) Α4 specification (210X297 mm) (Please read the back of the precautions to fill out this page)

564585 A7 __B7__ V. Description of the invention (6) (Please read the precautions on the back before filling out this page) The radio design is used for repetition rate operation in the range of about 4000 pulses / second or more, allowing accurate Controls all light quality parameters including wavelength and pulse energy. SUMMARY OF THE INVENTION The present invention provides a seed-injected modular gas discharge laser system capable of generating high-quality pulsed laser light under the following conditions: a pulse rate of about 4000 Hz or greater, and a pulse energy of about 5 to 10 mJ (millijoules) Or greater is used to produce an integrated output of approximately 20 to 40W (watts) or greater. There are two separate discharge cells, one of which is a part of the main oscillator that produces a very narrow band of seed light which is amplified in the second discharge cell. These chambers can be controlled separately to allow optimization of the wavelength parameters in the main oscillator and optimization of the pulse energy parameters in the amplification chamber. A preferred embodiment is that the F2 laser system is designed as MOPA, and specifically designed to be used as a light source for integrated circuit lithography. In this preferred embodiment, these chambers and laser optics are mounted on a vertical optical table in a laser housing. In the preferred MOPA embodiment, each chamber includes a single tangential fan that allows operation at a pulse rate of 4000 Hz or greater by removing debris from the discharge zone in less than about 0.25 milliseconds between pulses. This main oscillator is equipped with a line selection device for selecting the strongest F2 spectral line. This preferred embodiment also includes a pulse amplification module that divides each pulse from the power amplifier into two or four pulses to substantially reduce the damage rate of the lithography optical device. The preferred embodiment of the present invention uses a "three-wavelength platform." This includes an optical table with a housing and general equipment layout. Each of the three types of discharge laser systems (ie, KrF, ArF, and F2 lasers) is the same. These three systems are expected to be at 21 paper sizes. Applicable to China National Standard (CNS) A4 specification (210X297 mm) -9-564585 A7 B7 V. Description of invention (9) Repeat rate pulse energy pulse period 4 kHz 7 mJ 24 ns 4 kHz 7 mJ 40 ns 4 kHz 10 mJ 24 ns 4 kHz 12 mJ 12 ns The main components of this preferred laser system 2 are shown in Figure 1 and include: (1) Laser system frame 4 which is designed to accommodate other than AC / DC power modules All the modules of the laser, (2) AC / DC high voltage power supply module 6, (3) Resonant charging module 7 is used to charge the two charging capacitor banks to about 1000 volts at a rate of 4,000 charges per second (4) Each of the two converter modules 8A and 8B includes one of the above-mentioned charging capacitor banks and each includes a conversion circuit for forming a very short high-voltage electrical pulse from the energy stored in the charging capacitor banks. 16,000 volts with a period of approximately 1 microsecond (/ zs), (5) two A discharge chamber module is installed in the top and bottom structure in the frame 4. This module includes a main oscillator module 10 and a power amplifier module 120. Each module includes discharge chambers 10A and 12A and a compression head 10B installed on the top of the chamber With 12B. This compression head (by way of appointment) compresses the electrical pulse from the converter module from about 1 microsecond to about 50 nanoseconds with a corresponding increase in current. (6) The main oscillator optics includes a rear mirror 100 and line selection device LSP, 10C, which includes an output coupler and a 5mm line selector, 12 (Please read the precautions on the back before filling this page) This paper size applies to China National Standard (CNS) A4 specification (210X297 mm) 564585 A7 _B7_ V. Description of the invention (10) (7) The wavefront engineering box 14 includes optical devices and instruments for shaping and introducing seed light into a power amplifier and monitoring the output power of this MO (8) Light stabilizer module 16 includes wavelength, frequency bandwidth, and energy monitors, (9) Switch module 18, (10) Auxiliary room without gas control module 20, cooling water distribution module 22, and ventilation Modules 24, (11) Customer interface modules 26, (12) Laser control modules 28, and (13) Light-state lamps 30 For many applications, this laser system preferably includes a pulse extension unit (not shown) ) Extend the duration of the pulse beyond About 12 nanoseconds. U-shaped optical table In the embodiment of Fig. 1, as shown in Figs. 1A and 1B, the optical devices of MO and PA are mounted on the U-shaped optical table. This U-shaped optical table is mobilely mounted to the bottom of the laser in the manner described in U.S. Patent No. 5,863,017 and incorporated herein by reference. The rooms of MO and PA are not installed on the platform but each is supported on the track by three wheels (two on one side and one on the other side) (of which the bottom frame of chamber 2 is supported) (the wheel and the axle are better configured) As described in US Patent No. 6,109,574, which is incorporated herein by reference). This configuration enables the vibration isolation IT0 caused by the optical device to the room. The second general layout design is shown in Figure 1C. The second general layout design is similar to the above-mentioned first paper standard applicable to China National Standard (CNS) A4 (210X297) %) -13-(Please read the notes on the back before filling this page)

， 可 I ___ £ 7_ 564585 A7 V. Description of the invention (11) 2. General layout design, but including the following features: (1) The two chambers and laser optics are installed on the vertical optical table 151. The installation (as explained in the next section) is in the laser box 4. These chambers are supported on rigid cantilevers and bolted to the optical table. The main oscillator 10 is mounted on the power amplifier 12 in this design. (2) The back-up power supply 6B is included in the laser box 4. This two-chamber p2 4000Hz requires two 1200V power supplies. However, the laser box has room for other local voltage sources, which will be required for 6000Hz F2 laser systems. (3) Each of the two laser chambers and the pulse power source used in these chambers are very similar to those used in a 4000 Hz single-chamber ArF laser system and its pulse power source, which is in U.S. Patent Application No. 09/854 No. 097, which is incorporated herein by reference. (4) The pulse amplifier module 13 provided behind the optical table 11 is included in this embodiment to extend the period of the pulses emitted by the power amplifier. (5) The main oscillator light output light device in the on-line selection device iOC directs the output light from the MO to the power amplifier input optical device 14c and is used to pass through the power amplifier 12 at one time. This short pulse (approximately 12 nanoseconds) output from the power amplifier 12 is extended in a pulse extension unit 13 located behind the optical table u. The entire light path through the laser system including the pulse extender is contained in a vacuum housing (not shown), and this housing is designed with radon or helium clearance-~~~-Part of the third general layout design Copies are shown in Figure 1D. This layout design uses the laser chamber in the embodiment of the present invention, in which the length of the discharge area and the paper size are applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) 14 (Please read the precautions on the back before filling in this Page), ^ τ— 564585 A7 _B7_ 5. Description of the invention (12) is about half of the length of the discharge region of the first two embodiments. That is, compared to a typical length of about 53 cm, the length of this discharge region is about 26.5 cm. In this case, the resonant cavity of the main oscillator 10 (1) is defined by the light passing through the discharge region twice. This discharge region is between the maximum mirror 10E and the output coupler (which is 5 edges from the LSP 10C). Mirror selectors are set together). In this layout design, light passes through the power amplifier 12⑴ four times. The first pass is reflected from the mirror 15A and the electrode is aligned with the angle of the lower half of the discharge area (for example, the angle of the lower half of the right discharge area is about 10 millimeters from left to right). The second pass is After the light is reflected by the mirror 15B, the angle is about 4 degrees from right to left through the upper half. The third pass is reflected by two mirrors and the alignment electrode passes through the upper half of the discharge area, and the last pass is reflected by the mirror 15D and the alignment electrode passes through the lower half of the discharge area. This last pass sets up the power amplifier to output light. It bypasses the mirror 15C and is guided by the mirror (not shown) to the pulse amplifier unit (also not shown). Provisions are preferably provided in each of the above two layout designs to allow this output light 'line to be emitted from the left side of the laser housing or the right side of the laser housing so as to suit the customer's preference without major design changes. . In each of the above layout designs, performance improvement can be achieved by combining the converter and the shrink head into a single module. The applicant in this case has opposed such a combination in the past, because any component failure requires the entire module to be replaced. However, the applicant's experience in this case is that these units are very reliable :: that such a combination module is currently feasible. In fact, one of the causes of several failures in the pulsed power unit is the failure of the electrical power connected to the two modules.

The paper size is suitable for the financial family (⑽) M size ⑽X29 ^ D 15 (Please read the precautions on the back before filling in this page) Order-! 564585 A7 ____B7 V. Description of the invention (l3) It is not necessary in the combined module cable. (Please read the precautions on the back before filling this page) The design and operation of the above-mentioned preferred laser systems and modules will be explained in more detail below. Main Oscillator The main oscillator 10 shown in Figures 1 and 1C is similar in many respects to ArF lasers of known technology, such as described in U.S. Patent Nos. 5, 〇23,884 and 6,128'323 and In the US patent, the ArF laser stated in other states is requested, with the exception that its chamber and cavity optics annihilation juice is used for & laser operation in the spectral range of 157.63 nm. And its output pulse energy is about 0.11 ^ to 1.01 ^ instead of about 5mJ. However, major improvements are provided for '323 lasers to allow operation at 4,000 Hz or greater. This main reduction can be optimized for spectral performance including line width control. As shown in Figures 1, 2 and 3, the main oscillator includes a discharge chamber 10A, without a pair of extension electrodes 1082 and 1083, each about 50 cm long and about 12 mm apart. . The anode 1084 is mounted on the anode support rod 10A6 that forms the current. There are four separate water-cooled heat exchange units 10A8. The tangential fan 10A10 is driven by two motors (not shown), and provides a laser airflow at a speed of about 80 m / s between the electrodes. This room includes a window unit (not shown) with a CaF2 window set to approximately 47 from the laser light. . The electrostatic filter unit has an inlet in the center of the chamber. A small portion of the filtered airflow is shown as 11 in Figure 2. The purified gas is introduced into the window unit (in the same manner as in US Patent No. 5,359,620. And is incorporated herein by reference) to remove discharge debris from the window. The gain region of this main oscillator is caused by the discharge of laser gas between the electrodes. In this embodiment, the paper size of this paper applies the Chinese National Standard (CNS) A4 specification (210X297 mm) -16-564585 A7 B7 V. Description of the Invention (M) Rhenium is composed of about 0.5% F2 and the rest is a combination of neon, helium, or neon helium. This airflow removes debris from each discharge from the discharge area before the next pulse. The resonant cavity is generated on its output side by a wheel-out coupler located in the LSP 10C. This output coupler consists of an uncoated CaF2 reflective optical device, which is mounted perpendicular to the direction of the light so that it reflects about 5% of the laser light at about 15711111 and passes about 95% of this 157nm light. As shown in Figure 1, the opposite boundary of this cavity is the largest mirror 100. This line selection device 10C is described in more detail below with reference to FIG. Variations In a preferred embodiment, the main charging capacitor bank for the main oscillator and the power amplifier is charged in parallel to reduce the problem of instability. This is expected because the time for pulse compression in the pulse compression circuit of these two pulse power systems depends on the charge level of the charging capacitor. The pulsed energy output from the power amplifier is preferably controlled on a pulse-to-pulse basis by adjusting the charging voltage. This somewhat limits the use of voltage to control the main vibration light parameters. However, the 'laser gas pressure and dagger concentration can be easily controlled to achieve the desired light parameters over a wide range of pulse energies. The bandwidth decreases with decreasing F2 concentration and laser gas pressure. For the main oscillator, the time between discharge and light emission is a function of the F2 concentration (Ins / kPa), so the F2 concentration can be changed to change the timing, but this may be undesirable because It complicates other aspects of thunder light control. Power Amplifier The power amplifier in each of these three embodiments consists of a laser chamber, which is very similar to the corresponding main oscillator discharge chamber. With these two 17 (please read the notes on the back before filling this page) This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 B7 V. Description of the invention (l5) Allowed by each room The pulse energy and the integrated energy (called the dose) in a series of pulses are controlled to a considerable extent by the wavelength and the bandwidth respectively. This allows better dose stabilization. All components in this chamber are identical and interchangeable during the manufacturing process. In operation, however, the gas pressure in M0 is preferably substantially lower than that of PA. The compression head 12B of the power amplifier shown in FIG. 1 is also substantially the same as the compression head 10B in this embodiment, and the components of the compression head can also be exchanged with each other during manufacturing. The difference is that compared to PA, this compression head capacitor bank capacitor is more widely set for M0 to produce a substantially higher inductance. The electrode spacing in PA is preferably about 8 mm (compared to about 12 mm for M0). The chamber of this pulsed power system is similar to the electrical components, helping to ensure that the timing characteristics of this pulsed circuit are the same or substantially the same, so as to minimize vibration problems. As explained above, this power amplifier is designed to be used for a single pass of the laser light in the discharge area of the power amplifier discharge chamber in the embodiments of Figs. 1 and 1C, and for four passes in the embodiment of Fig. 1D. Discharge area. In the embodiment of FIG. 1, the line selection occurs in LSP 10C and the five-line selector shown in FIG. 16 is used. The selection of the line is preferably downstream of the gain medium of MO, because the gain in these F2 lasers is very high compared to KrF and ArF lasers. This line selects the seed rays which are reflected upward by the mirror 14A and horizontally for a one-pass power amplifier as described above. This charging voltage is preferably selected on a pulse-to-pulse basis to maintain the desired pulse and dose energy. The F2 concentration and laser gas pressure can be adjusted in the power amplifier to provide the desired operating range of the charging voltage. The desired range can be selected to produce the desired dE / dV value, as this paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 18 (Please read the notes on the back before filling this page) — 564585 A7 --------- B7 _ 5. Description of the invention (16) '— (Please read the precautions on the back before filling this page) The change in energy to voltage is Fa concentration and laser gas pressure Of functions. The timing of this light injection is preferably based on the charging voltage. The frequency of this light injection = good question and keep the laser chamber condition quite constant and can be continuous or almost continuous. Some users of these embodiments may prefer a longer period (e.g., two hours) between & injections. Pulse Power Circuit In the preferred embodiment shown in Figures 1, 1C, and 1D, this basic pulse power circuit is a pulse power circuit similar to the conventional excimer laser light source used in lithography. However, a separate pulse power circuit is provided downstream of the charging capacitor for each discharge cell. Preferably, a single resonant charger charges two parallel capacitor banks to ensure that the two capacitor banks are charged to exactly the same electrical base. There are also important improvements to adjust the temperature of 70 pieces of pulse power circuit. In the preferred embodiment, the temperature of the magnetic core of these saturated inductors is monitored, and a temperature signal is used in the feedback circuit to adjust the relative time of discharge in the two chambers. Figures 5-8 and 5; 6 show important components of a better basic pulsed power circuit for M0. This same basic circuit is also used in PA. Resonant Charger Figure 5B shows the preferred resonant charging system. The main circuit components are: II: Three-phase power supply 300 with constant DC current output. C-1: Power supply capacitor 302, which is equal to or larger than the conventional c0 capacitor 42.

Ql, Q2, Q3: switches to control the current flow for charging, and this paper size applies the Chinese National Standard (CNS) A4 specification (210X297 public love) · 19 564585 A7 £ 7_ V. Description of the invention (17) — — The regulated voltage is maintained at C0.

Dl, D2, D3: Provide unidirectional current flow. R1 and R2: provide voltage feedback to the control circuit. R3: Allows the voltage on q to be quickly discharged when a small amount of overcharging occurs. L1: Resonant inductor between C-1 capacitor 302 and Co capacitor bank 42 to limit current flow and set charge transfer time. Control board 304: Command Qi, q2, q3 to open and close according to the circuit feedback parameters. This circuit includes a switch Q2 and a diode D3, which together are known as a De-Qing switch. This switch improves the regulation of the circuit by allowing the control unit to turn off the inductor during the resonant charging process. This "De-Qing" switch prevents the extra energy stored in the current of the charging inductor L1 from being transferred to the capacitor C0. Before the laser pulse is required, the voltage on capacitor C-1 is charged to 600-800 volts and switches Q1-Q3 are turned on (0pen). Qi will be closed under the laser command. At this time, the current will flow from C-1 to c0 via the charging inductance gL1. As explained in the previous section, the computer evaluates the voltage and current on CG and the current on L1 with respect to the command voltage set point from the laser on the control board. Q1 will turn on when the voltage on the CO capacitor bank plus the equivalent energy stored in capacitor B equals the desired command voltage. This calculation is Vf = [Vc〇s2 + ((Li * ILis2) / C〇)] 0 · 5 where:

Vf = voltage on C0 after Q1 is turned on and current in L1 becomes Wen. Vc0s = voltage on Co when Q1 is on. 20 (Please read the notes on the back before filling this page).

This paper size applies to China National Standard (CNS) A4 specifications (210X297 mm) 564585 A7 — ~ ----— B7____ 5. Description of the invention (18) lus = current flowing through l 1 when Q1 is on. After Q1 is turned on, the energy stored in L1 is transferred to the C0 capacitor bank through D2 until the voltage on the C0 capacitor bank is approximately equal to the command voltage. At this time Q2 is turned off and the current stops flowing to co and is guided by D3. In addition to this de qing circuit, Q3 and R3 form a bleed-down circuit to allow additional fine adjustment of the voltage on C0. When the current flowing through the inductor L1 stops, the control board will order the switch Q3 of the down-regulation circuit 216 to close and the voltage on c0 will be reduced to the desired control voltage; then the switch Q3 will be turned on. The time constant of capacitor C0 and capacitor R3 should be fast enough to reduce capacitor C0 to the command voltage without occupying a considerable amount of time during the entire charging cycle. Therefore, this resonance charger can be designed with three-level down-regulation control. The slightly coarse adjustment is provided by the energy calculator and the switch Q1 is turned on during the charging cycle. When the voltage on the capacitor bank of Co is close to the target value, turn off the de-qing switch. When the voltage on Co capacitor is at or slightly exceed the target value, stop this resonance charging. In the preferred embodiment, switches Q1 and de-qing switches are used to provide an adjustment accuracy better than ± 0.0%. If additional adjustment is required, a third control with voltage adjustment can be used. This is a step-down circuit for switch Q3 and resistor R3 (shown as 216 in Figure 5B) to reduce the CO discharge to an accurate target value. CO Downstream Improvements As shown above, the MO and PA pulse power systems of the present invention each use the same basic design as used in conventional technology systems (Figure 5A). However, some major improvements in the basic design are required for this paper. The size of the paper applies the Chinese National Standard A4 (21 × 297 mm) 21 (Please read the precautions on the back before filling this page) Order —

564585 A7 ____B7_ V. Description of the invention (19) Due to the greatly increased repeat rate, the approximate factor is to increase the thermal load. This improvement is discussed below. Detailed description of the converter and compression head In this section, the manufacturing details of the converter and compression head are explained. The solid-state switch 46 is a p / N CM 800 HA-34H IGBT switch provided by powerex. Its office is in Young Wood, Pennsylvania. In the preferred embodiment, two such switches are connected in parallel. use. Inductors The inductors 48, 54 and 64 are similar to saturated inductors used in conventional technology systems' as described in U.S. Patent Nos. 5,448,580 and 5,315,611. FIG. 6 shows a preferred design of the L0 inductor 48. In this inductor, four wires from two IGBT switches 46B are formed through 16 iron rings 49 to form a part 48A. The hollow cylinder is 8 inches long and has a very high permeability material. Inside diameter and outside diameter of about 1.5 inches. Each of these four wires is wound twice around the core of the insulating donut shape to form part 48B. This four wire is then connected to the board, which is connected to the high voltage side of the C1 capacitor bank 52. A preferred diagram of the saturation inductor 54 is shown in FIG. 8. In this case, the inductor is a single-turn geometry, and the top and bottom covers 541 and 542 and the central mandrel 543 of this assembly are all at a high voltage to form a single-wound coil through the inductor's magnetic core. The outer case 545 is at a ground potential. This magnetic core is 0.0005-inch thick tape wound with 50-50% nickel-iron alloy and used by Magnetics of Butler, Pennsylvania or National Arnold of. The paper size is applicable to China National Standard (CNS) A4 (210X297 mm) -22-( (Please read the notes on the back before filling out this page)

564585 A7 B7 V. Description of the invention (20)

Courtesy of Adelanto, California. The heat sink 546 on the inductor case facilitates the transfer of the internally distributed heat to achieve mandatory air cooling. In addition, a ceramic plate (not shown) is installed under the bottom cover of the reactor to help transfer heat from the central part of the assembly to the bottom plate of the module. Figure 8 also shows the high-voltage wire connected to one of the capacitors connected to the C1 capacitor bank 52 and to one of the sensing units of the 1:25 boost pulse transformer 56. This case 545 is the ground wire to the unit 56 . A top view and a cross-sectional view of the saturation inductor 64 are shown in Figs. 9A and 9B, respectively. In the inductor of this embodiment, the flux exclusion metal pieces 301, 302, 303, and 304 are added as shown in FIG. 9B in order to reduce the leakage flux in the inductor. These flux exclusions substantially reduce the area where the magnetic flux is permeable and therefore help minimize the inductor's saturation inductor. The current passes around the magnetic core 307 five times via a vertical conductor rod in the inductor assembly. This current entering at 305 flows down through the large-diameter conductor marked "1" in the center shown in Figure 9A and up through the six smaller conductors also marked "1" around it. This current then flows down through the two conductors labeled 2 inside, then flows up through the six conductors labeled 2 outside, and then flows down through the flux that excludes the metal, and then flows upwards through the outside labeled 3 Six conductors then flow down through the two conductors marked 3 inside, then flow up through the six conductors marked 4 outside, and then down through the conductor marked 4 inside. This flux exclusion metal element is maintained across the conductor at half the full pulse voltage, allowing the safety separation gap between the flux exclusion metal portion and other wound metal rods to be reduced. This magnetic core 307 is formed by winding a 0.005-inch-thick 80-20% nickel-iron alloy strip. 23 (Please read the precautions on the back before filling this page.) The paper size applies to the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 _B7_ V. Description of the invention (21) (Please read the precautions on the back before filling out this page) Coils 307A, B and C, this band is made of Magnetics, Inc. of Butler , Available from Pennsylvania or National Arnold of Adelanto, California. The reader should note that nano-crystalline materials such as VITROPERMTM are available from VACUUM SCHITELZE GmbH Germany and FINEMETTM & Hitachi Metals Japan, both of which can be used for inductors 54 and 64. In conventional pulsed power systems, leakage of lubricant from electronic components is a potential problem. In this preferred embodiment, the lubricant isolation element is limited to a saturated inductor. In addition, as shown in FIG. 9B, the saturated inductor 64 is provided in a pot-shaped housing containing lubricating oil, and all closed connections are located at the level of the lubricating oil to substantially eliminate the possibility of oil leakage. . For example, the lowest confinement of inductor 64 is shown in Fig. 8B, 308. Because the normal lubricating oil level is under the top cover of the housing 306, it is almost impossible for the oil to leak out of the assembly as long as the housing is kept upright. The capacitor banks 42, 52, 62, and 82 shown in FIG. 5 (i.e., C., C !, <: 1 > _1 and (^) are all formed by a bank of existing capacitors connected in parallel. Capacitor 42 And 52 are film capacitors and are available from suppliers: for example, Vishay Roederstein (its office is in Statesville, North Carolina) or Wima, Germany. The preferred method for connecting capacitors and inductors is to solder them to The positive and negative terminals on a special printed circuit board coated with thick nickel copper wires are similar to those described in U.S. Patent No. 5,448,580. Capacitor banks 62 and 64 are typically composed of a parallel array of high voltage ceramic capacitors And the Japanese supplier Murata or this paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) -24-564585 A7 B7 V. Description of the invention (22) TDK supply. Compared with the use of ArF laser In the preferred embodiment, the capacitor bank 82 (ie, Cp) is composed of 33 banks of 0.3nF capacitors, and its total capacitance is 9.9nF; Cp- 丨 is composed of 24 banks of 0.40nF capacitors, and its total The capacitance is 9.6nF; (^ is 5. 7 //? Capacitor bank and C. Capacitor bank of 5.3 / zF. Pulse transformer Pulse transformer is also similar to that described in US Patent Nos. 5,448,580 and 5,3 13,481. However, the pulse transformer of this embodiment is in the secondary There are only one number of turns in the winding and 24 induction units are equal to 1/24 of the number of single primary turns for an equal 1:24 boost ratio. The diagram of the pulse transformer 56 is shown in Figure 10. Each of the 24 induction units includes The short-pitch aluminum tube 56A has two flanges (each has a flat edge and has threaded bolt holes), and it is bolted to the positive and negative terminals on the printed circuit board 56B (this negative terminal It is the high-voltage terminal of the 24 primary winding.) The isolator 56C separates the positive terminal of each short tube from the negative terminal of the adjacent short tube. It is 1 1/16 inch long between the flanges of the short tube. A chamber cylinder with an outer diameter of 0.875 and a wall thickness of about 1/32 inch. This short pitch tube is one inch wide by 0.7mil (one thousandth of an inch) thick Metglas ™ 2605 S3A and O.lmil thick Polyester film wrapped and wound up to this insulation 1 ^^ (8135 ^ outside the package diameter Up to 2.24 inches. Figure 10A shows a perspective view of the primary winding formed by a single wrapped short tube. The secondary winding of the transformer is a single 0D stainless steel rod installed in a firmly installed PTFE (Teflon) insulation tube. The winding shown in Figure 10 has four parts. This is the low voltage 25 of the stainless steel second winding shown at 56D in Figure 10. (Please read the precautions on the back before filling this page) This paper size applies to China Standard (CNS) A4 specification (210X297 mm) 564585 A7 B7 V. Description of the invention (23 The terminal is the main high-voltage line connected to 56E on the printed circuit board 56B. This high-voltage terminal is shown at 56F. Therefore, this transformer adopts an automatic transformer structure, and the boost ratio becomes 1:25 instead of 1:24. Therefore, a pulse of about -1400v between the ten and one terminals of the induction unit will generate a pulse of about -35,000 volts at the terminal 56F on the secondary winding side. This single-turn second winding design provides very low leakage inductance while allowing very fast output rise times. Details of the electrical components of the laser chamber

The Cp capacitor 82 includes a library of 0 3 nF capacitors mounted on the top of the chamber pressure vessel. C is typically an ArF laser operated in a laser gas, of which 3.5% argon, 0.1% fluorine, and the rest are composed of neon). These electrodes are about 28 inches long with a spacing of about 0.5 to 1-0 inches, preferably about 5/8 inches. The preferred electrodes are discussed below. In this embodiment, the top electrode is referred to as the cathode and the bottom electrode is referred to as the anode and is grounded as shown in FIG. Discharge time In ArF, KrF, and F2 discharge lasers, this discharge lasts only about 50 nanoseconds (ie, 50x10-9 seconds). This discharge causes an overall reversal for continuous operation, but this reversal only exists during the discharge. Therefore, when the whole is inverted in the laser gas so that the amplification of the seed light can be generated. An important requirement for injecting seeded ArF, KrF or F2 lasers is to ensure that the seed light from the main oscillator passes through the discharge zone of the power amplifier within approximately 50 nanoseconds. An important obstacle to accurate time control of this discharge is the switch 42 (As shown in Figure 5) there is a time delay of about 5 microseconds between being triggered and turning off and the beginning of discharge (which only lasts about 40-50 nanoseconds). This pulse uses approximately 5 microseconds to flow through the circuit between Co and the electrode. This interval time is 26% of the charging voltage (please read the precautions on the back before filling this page) The paper size applies to the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 ___B7_ V. Description of the invention (24) Size The temperature of the inductor in the AND circuit substantially changes. However, in the preferred embodiment of the invention described herein, the applicant of this case has developed an electrical pulsed power circuit that provides the discharge time control of two discharge cells to less than about 2 nanoseconds (ie, 2 × 10 · 9 Seconds) relative accuracy. A block diagram of these two circuits is shown in Figure 4. When the inventors of the present case conducted tests, they showed that the timing control changed with the charging voltage at about 5-10 nanoseconds / volt. This is a strict requirement for the accuracy and repeatability of the high-voltage power supply for charging the rechargeable capacitor. For example, if a timing control of 5 nanoseconds is desired and it has a movement sensitivity of 10 nanoseconds per volt ', its discrimination accuracy will be 0.5 volts. For a rated charging voltage of 1,000 volts, this will require a charging accuracy of 0.05%, which is very difficult to achieve, especially when these capacitors must be charged to this special value 4,000 times per second. The applicant's preferred solution to this problem is as shown in Figures 丨 and 4 and as described above, a single resonant charger 7 pairs 馗 〇 and? The eighth parallel charging capacitor is charged. It is therefore important to design two pulse compression / amplification circuits for the two systems so that the time delay versus charge voltage curve matches the one shown in Figure 4A. The same can be used in each circuit as much as possible The easiest to achieve. Therefore, in order to minimize the time variation (referred to as fluctuation) in the preferred embodiment, the applicant of the present application designed a hybrid power element for two discharge cells with similar elements, and confirmed that this time delay is proportional to (vs The curves of the voltages actually follow each other as shown in Figure 4A. The applicant in this case noted that in the normal operating range of the charging voltage, as the electric castle will be substantially & this paper size applies the Chinese national standard (Ο®) A4 specification (210X297 public love) (Please read the precautions on the back first (Fill in this page again)

27 564585 A7 ___B7 V. Description of the invention (25) The change of the delay, but this time delay with the change of voltage is the same for the two circuits. Therefore, for these two charging capacitors charged in parallel, the charging voltage can be changed in a large operating range without changing the relative time of discharge. The temperature control of electrical components in pulsed power circuits is also important because temperature changes can affect pulse compression time (especially temperature changes in saturated inductors). Therefore, its design purpose is to minimize the temperature change and the second method is to monitor the temperature of the temperature sensitive element and use the four-feed control to adjust the trigger time to reduce the compensation. It can also provide control by a processor, which is privatized with a learning algorithm that adjusts based on historical data about past time changes (with known operating history). This historical data is then applied to predict changes in time based on the current operation of the laser system. Control of Triggering The discharge triggering of each of the two chambers is accomplished individually using trigger circuits for each circuit (as described in U.S. Patent No. 6,016,325). These circuits add a time delay to correct the timing change caused by the change in the charging voltage and temperature in the electrical component of the pulse power, so that the time (delay) between triggering and discharging is kept practically constant. As shown above, since the two circuits are substantially the same, the changes after this correction are almost the same (i.e., within about 2 nanoseconds of each other). As shown in Figures 6C, D and E, the performance in this preferred embodiment will be greatly enhanced if the discharge in the power amplifier occurs in the main oscillator approximately 40 to 50 nanoseconds after discharge. This is because it takes several nanoseconds for the laser pulse to develop in the main oscillator and another several nanoseconds for the laser-1 "-_ The paper size is suitable for financial standards (CNS) A4 specifications ( 210X297 public love) "% ·--

、? Τ— (Please read the notes on the back before filling this page) 564585 A7 B7 V. Description of the invention (26) The front part of the light reaches the amplifier from the oscillator, and because of the frequency band behind the laser pulse from the main oscillator The width is much narrower than the width of the front band. For this reason, a separate trigger signal is provided to trigger the switch 46 for each room. This actual delay selected to achieve the desired light quality is based on the actual performance curves as shown in Figures 6C, D and E. The reader should note that, for example, narrower bandwidths and longer pulses can be obtained at the cost of increasing the pulse energy delay between the M0 trigger and the PA trigger. Other techniques for controlling the discharge time As shown in Figures 6C, D, and E, the relative time of discharge can have a significant effect on the quality of light. It is reasonable to have other steps to control the discharge time. For example, some modes of laser operation can cause large changes in charging voltage or large changes in inductor temperature. These large changes can complicate discharge time control. Monitoring time The discharge time can be monitored with pulse-to-pulse basis, and the time difference can be used in the feedback control system to adjust the time of the trigger signal of the switch 42. This PA discharge is preferably monitored using a photovoltaic cell to observe the discharge fluorescent light (called ASE) rather than using laser pulses, because if laser light is not generated in the PA, it will cause a very bad time. For MO you can use ASE or seed laser pulses. The bias adjustment is as shown in Figure 5. The pulse time can be adjusted by adjusting the bias current flowing through the inductors LB1, 1 ^ 2, and 1 ^ 3 (which provides the bias for inductors 48, 54 and 64) The current can be increased or decreased. Other technologies can be used to increase the required paper size. This paper applies the Chinese National Standard (CNS) A4 specification (210X297 mm) -29-(Please read the precautions on the back before filling this page)

V. Description of the invention (1) (27) Saturates these inductors. For example, the core material of an 'inductor' can be mechanically ㈣ instead a ρζτ element with a very fast response, which can be feedback-controlled based on a feedback signal from a pulse timing monitor. Adjustable parasitic load A positive parasitic load can be added to one or both of the pulsed power circuits downstream of the CD. Other feedback control In addition to the pulse time monitoring signal, the charging voltage and inductor temperature 仏 can break the feedback control, in addition to adjusting the trigger time described above, you can also adjust the bias or core mechanical separation described above . X burst type operation When the laser is operated on a continuous basis, the feedback control of time is quite easy and effective. However, under normal circumstances lithography lasers are, for example, the following operating in burst mode to process 20 areas on each of many wafers: Shut down (stop emitting lasers) for one minute and place the wafer into position to 4000 Hz repeat rate irradiates the area 1 〇2 seconds shut down 0.3 seconds to move to area 2 0.2 seconds to illuminate the area 2 0.2 seconds turn off 0.3 seconds to move to area 3 irradiates the area 3 with 4000 Hz repeat ratio Illuminate the area at a repetition rate of 4000 Hz in seconds, 0.2 seconds, shut down in 0.3 seconds, and move to the area in 200. 30 This paper size applies the Chinese national standard (⑽) A4 specification (210X297 public love) 564585 A7 B7 V. Description of the invention (28) at 4000Hz The repetition rate irradiation area 200 was turned off for 0.2 minutes in 0.2 seconds to replace the wafer. On the next wafer, the area was irradiated at the repetition rate of 4000 Hz for 1 0.2 seconds. This process can be repeated for many hours, but will often be interrupted over a period of 1 minute. The length of this down time will affect the relative time between the pulse power system of the MO and the PA, and may need to be adjusted in the trigger control to ensure that: when the seed light from the MO is at the desired position, it is generated in the PA Discharge. By monitoring the discharge and time of light emission from each chamber, the laser operator can adjust the trigger time (accurate to within about 2 nanoseconds) to achieve the best performance. The laser control processor is preferably programmed to monitor time and light quality and automatically adjust the time for optimal performance. In the preferred embodiment of the present invention, a timing algorithm is used to develop stored value groups that can be applied to various different operation mode groups. These algorithms in the preferred embodiment are designed to switch to feedback control during continuous operation. The timing value for the current pulse is based on the feedback data collected for one or more previous pulses (such as the previous pulse). While setting. No Output Discharge The timing algorithms discussed above work well for continuous or regular repeated operations. However, the accuracy of this timing may be poor in unusual situations, such as the first pulse after an unusual period of laser shutdown, such as 5 minutes. In some cases this is used for the inaccuracy of the first or second pulse 31 (please read the precautions on the back before filling this page) This paper size applies to China National Standard (CNS) A4 (210X297 mm) ) 564585 A7 B7 5. Description of the invention (29) Timing may not cause problems. A better technique is to pre-design the laser so that the discharge of MO and PA is intentionally out of order for one or two pulses, so that it is impossible to amplify the seed light from M0. For example, a laser can be programmed to discharge PA trigger for 80 nanoseconds before triggering M0. In this case there will be no significant output from the laser, but the laser measurement sensor can determine the timing parameters so that the time parameters used for the first output pulse are accurate. Alternatively, the triggering of the MO may be earlier than that of the PA, so that the light of the MO passes through the PA before the PA is discharged. The 4D, 4D1, 4E, and 4E1 are flowcharts showing possible control methods using these technologies. . In order to accommodate the larger thermal load, in addition to the normal forced air cooling provided by the cooling fan in the laser box, the water cooling of the pulse power element is also provided to support 4KHz or larger pulses. Operating rate. The disadvantage of water cooling is that it traditionally has the potential to leak water near electrical components or high-voltage wiring. The particular embodiment of this case substantially avoids this potential problem by using a single sturdy cooling tube that passes through the module to cool these components, which normally dissipates most of the heat stored in the module . Because there are no joints or connections in the module case, and the cooling pipe is a continuous piece of solid metal (such as copper, stainless steel, etc.), this greatly reduces the chance of water leakage in the module. The connection of the module to the cooling water is therefore implemented outside this assembled sheet metal housing, where the cooling water pipe is mated with the quick-release connector. Saturated inductor 32 (Please read the precautions on the back before filling this page) This paper size applies to Chinese National Standard (CNS) A4 (210X297 mm) 564585 A7 _________B7 V. Description of the invention (3〇) In the converter module In the case, for example, the water-cooled saturated inductor II54A is provided in the u figure, which is similar to the inductor 54 shown in FIG. 8 except that the heat sink of the inductor 54 is shown in the figure. Water cooling jacket 54A1 replaced. The cooling line 54A2 is wound around the sleeve 54A1 through the module and through the aluminum base plate. A 1 × ^ switch and a series triode are installed on this board. These three components constitute most of the power consumption in the module. Other components that also dissipate heat (buffering diodes, resistors, capacitors, etc.) are cooled by forced air provided by two fans behind the module. Because the sleeve 54A1 is kept at ground potential, it will cool down There is no problem of voltage isolation when the tube is directly attached to the reactor casing. This is achieved by pressing the cooling tube into the dovetail slot, which is cut out of the casing as shown in 54-8 and used The heat-conducting mixture helps to create a good thermal contact between the cooling tube and the housing. Although the IGBT switches are "floating" at high voltages, they are mounted on the bottom of the aluminum, and with a 1 / 16-inch thick aluminum plate It is electrically isolated from the switch. This aluminum base plate is used as a radiator and operates at ground potential and is easy to cool, because high voltage isolation is not required in the cooling circuit. Figure 7 shows the figure of the water-cooled aluminum base plate. In this case, the cooling piping is pressed into a groove in the aluminum bottom, and the IGBT is mounted on the aluminum bottom. Like the inductor 54a 'uses a thermally conductive mixture to improve the overall joint between the piping and the base plate. These series diodes are in the right place It is also “floating at high pressure” during operation. In this case, the diode housing typically used in this design does not provide the paper size applicable to the Chinese National Standard (CNS) A4 specification (21〇 > < 297 public love)

(Please read the notes on the back before filling this page) Order 丨

564585 A7 --- -B7_— V. Description of the invention (31) ~ ^ ~ ~ Back pressure isolation. In order to provide the required isolation, the diode "hockey pUck", the device is clamped in a heat sink assembly, and then It is mounted on top of a ceramic base, which is then mounted on top of a water-cooled aluminum base plate. This ceramic base is just thick enough to provide the required electrical isolation, but not too thick to cause more than required heat Impedance. For this special design, the ceramic is 1/16 inch thick aluminum, although other more special materials such as beryllium can also be used to reduce the thermal resistance between the diode junction and the cooling water. Water-cooled conversion The second embodiment of the device is assembled using a single cooling plate, which is connected to the chassis bottom plate for the IGBT and the diode. This cooling plate can be welded to the two aluminum "tops," and "bottoms" by welding a single piece of nickel pipe As explained above, these IGBTs and diodes are designed to transfer heat to the cooling plate by using the previously mentioned ceramic discs under assembly. In the comparison of the present invention In the preferred embodiment, a cold plate cooling method is also used to cool the The IGBT and diode in the charger can also use a heat transfer rod or heat pipe to transfer heat from the outer casing to the chassis plate. Detailed description of the compression head Compared with water-cooled compression head, its electrical design is similar to air cooling of conventional technology (The same form of ceramic capacitors and similar materials are used in the reactor design.) The main difference in this case is that the module must operate at higher repetition rates and therefore higher average power. During compression In the case of the head module, most of the heat is dissipated in the modified saturated inductor 64A. Cooling the subassembly is not a simple matter, so the entire housing is operated with very high voltage short pulses. The solution to this problem As shown in Figures 12, 12A, and 12B, the case and the ground potential are isolated by inductance. This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm)

Yes, (please read the precautions on the back before filling this page) 564585 A7 B7 V. Description of the Invention (32) The inductor is provided by winding a cooling pipe around two cylindrical shapes (including the ferromagnetic core). The input and output cooling lines are wound around the cylindrical part of the iron core. It consists of two cylindrical parts and two iron blocks as shown in sections 12, 12A and 12B. This iron piece is made of CN-20 material and is made by Ceramic Magnetics, Inc., of Fairfield, New Jersey. The one-piece copper piping (0.187 inch diameter) is press-fitted to surround the case 64A1 of the inductor 64A and the shape surrounding the second winding is wound into a winding shape. Allow sufficient length at the end to extend through the equipment in the metal head cover of the compression head so that no cooling pipe joints are present in the chassis. Inductor 64A includes a dovetail slot shown at 64A2, which is similar to that used in the first stage reactor housing water-cooled converter. This shell is very similar to the previous air-cooled shell except for the dovetail slot. This copper cold water pipe is pressed into the groove to form a good thermal connection between the shell and the cold water distribution. A thermally conductive mixture is also added to minimize thermal resistance. The electrical design of the inductor 64A is slightly changed from that of the inductor 64 shown in Figs. 9A and 9B. The inductor 64A only provides two turns (rather than five turns) around the magnetic core 64A3, which is composed of four turns instead of three turns. Because the guidance route of this water-cooled piping is from the output voltage to the ground potential, the bias current circuit is now slightly different. As before, the bias current is provided by a dc-dc converter in the converter via a cable to the compression head. This current passes through the "positive" bias inductor LB2 and is connected to the Cp-1 voltage node. This current then separates a part of it through the high-voltage cable (via the secondary winding of the transformer to ground and back to the dc-dc converter) back to converter 35 (please read the precautions on the back before filling this page) Applicable to China National Standard (CNS) A4 specification (210X297 mm) 564585 A7 B7 V. Invention description (33) (commutator). The other part passes the compression head reactor Lp-1 (biasing the magnetic switch) and then "negatively" biases the inductor LB3 through the cold water pipe and returns to ground and the dc-dc converter. By balancing the resistance in each branch, the designer can ensure that sufficient bias current can be used in the compression head reactor and the converter transformer. This "positive" bias inductor LB2 is made very similar to the "negative" bias inductor LB3. In this case, the same iron rod and iron block are used as the magnetic core. However, the use of two 0.125-inch-thick plastic spacers creates an air gap in the magnetic circuit so that the core is not saturated with DC current. Instead of surrounding the inductor with a cold water pipe, surround the shape with an 18AWG Teflon wire. Quick Connect In this preferred embodiment, three of the pulsed power electrical film sets use blindly paired electrical connections, so that all electrical connections to the laser system section are made only by sliding the module into its place. This can be achieved in the position of the laser box. These are A C distribution core group, power handle group and resonance charging module. In each case, the male or female plugs on the module mate with the female or male plugs installed behind the box. In each case, two approximately 3-inch terminal pins on the module guide the module into its correct position so that the electrical plugs are properly mated. Such blind mating connectors are, for example, AMP model number No. 194242-1 and are commercially available from AMP, Inc. (whose corporate office is in Harrisburg, Pennsylvania). In this embodiment, the connector is used for various power circuits such as: 208 volt AC, 400 volt AC, 1000 volt DC (power out and resonance charging in) and for several signal voltages. This paper size applies to China National Standard (CNS) A4 specifications (210X297 public love) -36-(Please read the precautions on the back before filling this page)

564585 A7 B7 V. Description of the invention (34) The blind pairing connection allows the ground waiting module to be used for maintenance and replacement within seconds or minutes. In this embodiment, the blind paired connection is used side by side for the converter module. The output voltage of the module is in the range of 20 to 30,000 volts. And use typical high-voltage connectors. Discharge Element Figures 13 and 13A (1) show a modified discharge structure used in the preferred embodiment of the present invention. This structure includes an electrode structure referred to by the applicant as a "blade dielectric electrode. In this design, the 'anode 5 40' includes a pure blade-shaped electrode 542 with a dielectric spacer 544 mounted on both sides of the anode as shown, to Improve the gas flow in the discharge zone. This spacer is installed outside the discharge zone with screws on the anode support rod 546. This screw allows thermal expansion and sliding between the spacer and the rod. The anode is 26.4 inches and 0.439 inches to height. It is 0.284 inches wide at the bottom and 0.141 inches wide at the top. It is connected by screws through a sleeve to form a flowing anode support rod 546, which allows the electrode to undergo differential thermal expansion from its central position. The anode It is composed of a copper-based alloy, preferably C36000, C95400, or C19400. The cathode 541 has a cross-sectional shape as shown in Figure 13A. The cathode material is preferably C36000. Other details of the blade dielectric structure Is provided in U.S. Patent Application No. 09 / 768J53 and is incorporated herein by reference. The current reentrant 548 in this structure is composed of a whale-shaped portion with 27 ribs along the electrode 542 The length is equidistantly spaced, and its cross section is shown in Figure 13A (1). As mentioned above, this current retracement is made of metal sheet, and these whale ribs (each with a cross section size of about 0.15 inches x 0. 09 inches) is bent so that the length of each rib is in the direction of current flow. This paper size applies to China National Standard (CNS) A4 (210X297 mm) 37 (Please read the precautions on the back before filling in this page)

564585 A7

5. Description of the invention (35) (Please read the notes on the back before filling out this page) Figure 13 A2 shows the design of an alternative dielectric spacer for the anode to improve the flow. In this case the spacers are perfectly matched to form a flowing anode support rod to provide a better gas flow path. The applicant in this case called it a “quick-return” blade dielectric anode design. Alternative Pulsed Power Circuits The second preferred pulsed power circuit is shown in Figures 5C1, 5C2 and 5C3. This circuit is similar to the one described above, but uses a higher voltage power supply to charge co to a higher value. As in the embodiment described above, the high-voltage pulsed power supply unit is operated at 230 or 460 volts AC by power from a power plant. As mentioned above, it is a power source for fast charging resonance chargers and is designed to operate at frequency 4000 to 6000 Hz charges two 2.17uF capacitors to a voltage in the range of approximately 1100V to 2250 volts. The electrical components in these converters and compression heads for the main oscillator are as far as possible equivalent to the corresponding components in the power amplifier. This is done to keep the time response as feasible as possible in both circuits. Switch 46 is a bank of two IGBT switches each at 33,000 volts and configured in parallel. The Co container library 42 is configured by 128 0.68 μP 1600V capacitors into 64 parallel pairs to provide a 2.17 pF Co library. (^ Capacitor bank 52 is composed of 136 0.068pF 1600V capacitors and configured into 68 parallel pairs to provide a 2.33pF bank capacitance. (: ... and the core capacitor bank are the same as those described above with reference to Figure 5). Saturated inductor 54 is a single-turn inductor that provides approximately 3.3nH of saturated inductance. It has 5 cores consisting of 0.5 inches to 50% -50% thickness * ^ 6, with a diameter of 4.9 inches and 3.8 inches. Internal diameter. Saturated inductor 64 is a two-turn inductor that provides approximately 38Nh of saturated inductance. Each of the five cores is made of 0.5-inch thick 80% -20%, Ni-Fe. (CNS) A4 specification (210X297 mm) 38 564585 A7 —___ B7_ V. Description of the invention (36) Outside diameter and 2.28 inches inside diameter. A trigger circuit is used to set the time accuracy with two nanoseconds. IGBT 46 is off. This main oscillator is typically triggered 40ns before the IGBT trigger for the power amplifier. However, the exact time is completely determined by the feedback signal from the sensor, which measures the main oscillation Output time of the amplifier and discharge time of the power amplifier. The length of the output pulses measured by the applicant in this case for these F2 lasers in the test is in the range of about 12 nanoseconds, and to some extent is a function of the relative time of the two discharges. Longer pulses The length (same in other respects) can increase the service life of the optical components of the lithography equipment. The applicant in this case identified several techniques for increasing the pulse length. As shown above, the relative time between discharges can be optimized For pulse length, both the MO and PA pulse power circuits can be optimized for longer pulses. The technology used is described in US Patent Application No. 09 / 451,995, which is incorporated herein by reference. An optical pulse amplification system (such as one described in US Patent No. 6,067,311, which is incorporated herein by reference) can be added downstream of the PA to reduce the intensity of individual pulses. The better pulses will be explained in the next section Amplifier unit (also known as pulse extender). This pulse amplifier can be made as part of the light path to the lens component of the lithography tool. This chamber can Made longer and the electrodes can be designed to produce traveling wave discharges for longer pulse lengths. The pulse amplifier unit is shown in Figure 22A as a better pulse amplifier unit. The light 20 from the laser 50 hits the beam splitter 22 The spectroscope has a reflectivity of about 40%. 39 (Please read the precautions on the back before filling this page) This paper size applies to China National Standard (CNS) A4 (210X297 mm) 564585 A7 B7 V. Description of the invention ( 37) Approximately 40% of the light is reflected as the first part of the output light 30. The remaining part of the incident light is transmitted to the light 24 through the beam splitter 22. This light is reflected back at a small angle by the mirror 26, which is a spherical mirror with a focal length equal to the distance from the beam splitter 22 to the mirror. Therefore, this light is focused to the point 27 near the beam splitter 22 but deviates slightly. This light continues and is reflected by mirror 28, which is also a spherical mirror and has a focal length equal to the distance from this mirror to point 27. The mirror 28 reflects the light back at a small angle and aligns the reflected light. This reflected light 32 is transmitted to the right and reflected by the mirror 29 to the beam splitter 22, where about 60% of the light is transmitted through the beam splitter 22 and merges into the second part of the output light 30. A portion (approximately 40%) of the ray 34 is reflected by the beam splitter 22 in the direction of the ray 24 and repeats the distance traveled by the ray 32. Therefore, the short input pulse is divided into several parts, so the entire period of light transmission increases and its peak intensity decreases. The mirrors 26 and 28 constitute a transmission system which reflects one part of the output light onto each other. Because of this illumination (reflection), the parts of the output light are substantially the same (if the mirrors 26 and 28 are flat mirrors, the light dispersion will diffuse each subsequent repeated reflection, so the size of each repeated reflection light will be different). The length of the entire ray path from beam splitter 22 to mirror 26 to mirror 28 to mirror 27 and finally to beam splitter 22 determines the time delay between repetitions. Figure 22B1 shows the pulse profile of a typical pulse generated by an ArF excimer laser (these results can be applied to an F2 laser, the difference is that the unamplified pulse for an F2 laser is approximately 12 Nanoseconds, not about 18 nanoseconds for ArF). Figure 22B2 shows the simulated output pulse profile of an ArF-like laser pulse after diffusion in a pulse extender made according to Figure 6. In this example, the Tis of the pulse is from 40 (please read the precautions on the back before filling this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 B7 V. Description of the invention (38) 18.16 ns increased to 45.78ns. (Tis is a measurement used to describe the pulse period of a laser pulse. It is about the complete square pulse period.) Figure 22C shows a layout design similar to the layout in Figure 24A but with an additional delay path. In this case, the first beam splitter 22A is designed to reflect 25% and the second beam splitter 22B is designed to reflect 40% of light. Figure 22D shows the shape of the light generated by the computer simulation. The Tis used to extend the pulse is about 73.2ns. In the embodiment of FIG. 22C, the part of the light transmitted through the beam splitter 22B is inverted and added to the output light 30 when it returns. This phenomenon greatly reduces the spatial cohesion of this light. Figures 22E and F show the design of the beam splitter using uncoated optics. Figure 22E shows the beamsplitter design which uses failed internal reflections, and Figure 22F is a slanted, transparent uncoated plate that reflects Fresnel from both sides of the plate to achieve the desired reflection / transmission ratio. This pulse extension unit can be installed in the back of the vertical optical table 11 as suggested above, or it can be installed on top of the table or even inside the table. Pulse and dose energy control The pulse energy and dose energy are preferably controlled by the feedback control system and method described above. This pulse energy monitor can be closer to the wafer in laser due to the lithography tool. Use this technique to select the charging voltage to produce the desired pulse energy. In the above preferred embodiment, both MO and PA are provided with the same charging voltage, because the Co capacitor is charged in parallel. As discussed above, the applicant in this case determined that the technology worked well and minimized the problem of time instability. However, this technology does ask this question 41 (please read the notes on the back before filling this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 B7 V. Description of the invention (39) Reduction to this level enables the laser operator to control M0 independently of PA. However, some operating parameters of MO and PA can be controlled separately to optimize the performance of each unit. These other parameters include: laser gas pressure, F2 concentration, and laser gas temperature. These parameters are preferably controlled independently in each of the two chambers and adjusted in a processor-controlled feedback configuration. Other improvements in optical quality The present invention provides a laser system that can achieve greater pulse energy and output power than conventional single-cell high-repetition gas discharge lasers. In this system, the oscillator determines the wavelength and frequency bandwidth to a large extent, and the power amplifier mainly controls the pulse energy. As shown in Figure 6B, the pulse energy required for the effective seed light of the power amplifier can be as low as a fraction of a millijoule (mJ). Because this main oscillating laser can easily generate a 5mJ pulse, it has energy remaining. This additional pulse energy provides an opportunity to use certain techniques for improving light quality, which are not currently particularly energy efficient. These techniques include: • Pulse trimming, as described in U.S. Patent No. 5,852,621 and incorporated herein by reference. The pulse energy is monitored, the pulse is delayed, and a portion of this delayed pulse is trimmed using a very fast optical switch, such as a Pockels cell. • Use a line narrowing module, which has very high light expansion and a small aperture, which will be explained later in this case. • The wavefront correction function between the wavefront engineering cavities can be added to the main oscillator or main 42 (please read the precautions on the back before filling this page) This paper size applies to China National Standard (CNS) A4 specifications (210X297 mm) ) 564585 A7 _______B7_ V. Description of the invention (40) Downstream of the oscillator. This may include the use of a grating having one or more bends as described in U.S. Patent Application No. 09 / 703,317 and incorporated herein by reference. The deformable mirror wavefront correction can also be a static correction ', such as designing a non-planar mirror and a non-planar mirror surface to correct known wavefront distortions. Light-passing light filters, such as spatial filters, as described in US Patent Application No. 09 / 309,478, incorporated herein by reference, and can be added to reduce the bandwidth as shown at 11 in FIG. 23 . The light filter can be placed in the MO cavity or between M0 and PA, or it can be added downstream of the PA. A better spatial filter, which does not require light to propagate through the focus, is a fully internal spatial filter described in the next section. • Cohesiveness control The cohesiveness of laser light can be a problem for integrated circuit manufacturers. Gas discharge lasers typically produce laser light with low cohesiveness. However, when the bandwidth is made very narrow, the result is a large cohesiveness of the output light. For this reason, some introduction of space non-depolymerization may be expected. The optical element for reducing cohesion is preferably added to the MO cavity or between M0 and PA. Several types of optical elements are known for reducing the cohesiveness of light, such as mobile phase skins or acoustic optical devices. • Set the aperture. The light quality of the seed light can also be narrower. The paper size applies the Chinese National Standard (CNS) A4 (210X297 mm) -43 (Please read the precautions on the back before filling this page) ， 可 | 564585 A7 B7 Invention description (41 improvements. Total internal reflection spatial filter spatial filtering can effectively reduce 95% of the integrated frequency bandwidth. However, all the previously proposed direct spatial filtering techniques need to focus at least the light, and In most cases, it is necessary to actually focus the light. In addition, all previous designs relied on multiple optical elements. If spatial filtering is desired, 'a simple and compact spatial wave filter that does not need to focus light can be easier Suitable for installation inside a laser resonator. A preferred filter is a single chirp that is approximately two inches long. The entrance and exit sides of the chirp are parallel to each other and perpendicular to the incident light. The other two faces are parallel to each other , But its direction is equal to the critical angle (cdtical angle) with respect to the entrance and exit planes. In CaF2, the critical angle at a wavelength of about 157 nm It is 39 89 degrees. The only coating required is the normal incidence anti-reflection coating on the entrance and exit surfaces of the radon. This spatial filter works in the following way: light will enter the radon into the entrance of the radon The surface. The light then travels to the corner of the puppet surface. If the light is adjusted, all the light will be incident on the first surface at a critical angle. However, if the light is divergent or concentrated, some of the light will be greater or less than the critical angle. The angle hits this surface. All light rays that are greater than the critical angle in this surface will be reflected 100%. Light rays that hit this angle at angles less than the critical angle will be reflected less than 100% and will attenuate. All reflected rays will be The same angle is incident on the opposite side of 稜鏡 and the light will be attenuated by the same number. In the proposed design, there will be a total of six reflections for each pass of the light. This is less than the critical angle / radius (mrad ) Of the angle of the __ polar polar strikes the face of the total (please read the precautions on the back before filling out this page)-Order 丨 · 564585 V. Description of the invention (42) The angle of reflection A7 B7 is about 71%. Therefore, all rays whose angle of incidence differs from the critical angle / mrad or more will transmit less than 13% of their original intensity at their exit surface. However, the rays will pass through this filter once. All The incident light whose angle of incidence is greater than the critical angle will reflect 100/0. Once it leaves the spatial filter, the wave prism 'this light will be incident on the mirror. In laser resonance writing, this mirror can be an output coupler. After reflecting from the mirror, the light will re-enter the spatial filter, but there are significant differences. All rays leaving the spatial filter at a greater than critical angle will be inverted after reflecting from the mirror. These rays will now The angle re-enters the beam and will attenuate. The light passes through the beam a second time and changes the transmission function of the beam from a side filter to a true band-pass filter. Figure 23B shows the design of the spatial filter. The input and output sides of this frame are 1/2 inches. The critical angle is about 2 inches. The input light width is 2.6 mm and represents the width of the light in the short axis. This frame will have an inch height in the plane of the figure. This figure shows three groups of light. This first group of rays is aligned and hits the surface at a critical angle. These are green lights. The second group of rays enters this surface at an angle less than the critical angle and ends at the first reflection. These are blue light. These rays are more visible in the enlarged part. They represent light that decays during one pass of k. The last-group of rays is incident at a greater angle than the critical angle. These rays travel through the entire first pass, but end at the first reflection of the second pass. They represent the attenuation of light on the second pass. Telescopes between rooms This paper scale is in accordance with Chinese national standard A4 (21〇 > < 297 public love) (Please read the precautions on the back before filling this page)

-45 564585 A7 ______B7 V. Description of the Invention (43) In a preferred embodiment, a cylindrical refracting telescope is provided between the output of the main oscillator and the input of the power amplifier. This controls the level of light entering the power amplifier. This telescope has previously been designed using well-known techniques to control the horizontal spread of light. Metrology In the preferred embodiment of the present invention, the pulse energy is monitored and controlled on a pulse-to-pulse basis with feedback from a fast photodiode mass monitor. Pulse-to-pulse monitoring of wavelength and bandwidth is not provided in many applications because the natural centerline wavelength and bandwidth of the main F2 line are quite constant. However, if desired, both wavelength and frequency can be monitored in the same way as excimer lasers are conventionally known, but at a wavelength of 157 nm. Power monitors (p-cells) should preferably be set at the output of the main oscillator, after power amplification and after pulse amplification. It should preferably also be provided for monitoring any retro-reflection into the main oscillator. This retro-reflection amplifies and damages the main oscillator optics in the oscillator. If danger is exceeded, this signal from the retro-reflective monitor is used to turn the laser off. This system should also be designed to avoid back reflections in the light path which could cause any significant retroreflections. The following describes the measurement and control of light parameters for this laser. The wavemeter used in this embodiment is similar to that described in U.S. Patent No. 5,978,394, and some of the following descriptions are extracted from this application. In the range of 157 nanometer wavelengths, the wavelength and bandwidth measuring elements will be damaged by radiation. Therefore, the applicant of this case recommends that these measurement periods apply the Chinese National Standard (CNS) A4 specification (210X297). 46 (Please read the notes on the back before filling this page), \ t— 564585 A7 ----- ^ ___ B7__ V. Description of the invention (44) It is not implemented on a pulse-to-pulse basis. For example, the wavelength and frequency bandwidth are monitored with 30 pulses at a period of 10 clocks each. At this rate, the measurement element used for lasers should have a service life comparable to at least £] ^ and eight 1 ^ lasers. To this end, a switch for a wavelength meter should be provided to prevent light from entering the wavelength and bandwidth measuring elements. The optical equipment in these units measures pulse energy, wavelength, and frequency band search. These measurements are used with feedback circuits to maintain the pulse energy and wavelength within the desired limits. A small portion of the laser light is reflected to the energy detector, which includes a very fast light diode whose energy is measured at the energy of individual pulses at a pulse rate of 4,000 per second. This pulse energy is about 10mJ, and the output of the detector 69 is fed to the computer controller, which uses a special algorithm to adjust the laser charging voltage 'and accurately control the pulse energy of future pulses based on the stored pulse energy data. In order to limit the change of individual pulse energy and the overall energy of the burst pulse. According to the measurement of the pulse energy of each pulse described above, the pulse energy of the pulses thereafter can be controlled to maintain the desired pulse energy and the desired integrated dose of a specific number of pulses, all of which are in the US patent Case No. 6,005 [879 "Pulse Energy Control of Excimer Laser", which is incorporated herein by reference. Line Selection Line Selector with 稜鏡 In the preferred embodiment, the strongest natural F2 resonance line is to use a five 稜鏡 line selector to select one as shown in Figures 16A and B. These five prism paper sizes are in accordance with China National Standard (CNS) A4 specifications (210X297 mm) 47 (Please read the precautions on the back before filling this page), ^ τ— 564585 A7 B7 V. Description of the invention (45) Mirror The 112 AE is precisely mounted on a single board located in the LSP 10C (not shown) as shown in Figure 1. This LSP is located downstream of the main oscillator and a short distance downstream of the main oscillator output coupler. Each of these five angles is 65 degrees (apex angle), and the horizontal configuration is as shown in Figure 16A. The angle of incidence for this particular embodiment is as follows, for each of U2A-E: 79.6. , 61.4. , 47.7. , 71.7. And 42.1. (Many other tritium structures can provide similar results). As shown in FIG. 1, the mirror 114B is positioned to reflect light upward to the mirror 114C, and it is positioned to reflect this light into the discharge area of the power amplifier through the aperture (not shown) as in FIG. 12A. As shown. This penta-line selector produces an angular spread of 10.56 milli-radian between the 157.63 nm and 157.52 nm F2 lines, which produces an approximate aperture in front of the power amplifier about 0.5 m downstream 5 5 亳 m (mni) space separation. This diffusion is easy enough to separate 157 52 nm wires. Circle Line Selector Another line selection unit is shown in Fig. 16E, which is a circle structure. It can be inserted into the light path without disturbing the direction of the light. In this embodiment, the ring consists of four 450-term corners and four 65s. Horned owl.

Lyot filter As shown in Fig. 16A, the other method of the 稜鏡 -based line selector is a Lyot filter. This filter uses, for example, birefringent diffusion of MgF2i anisotropic crystalline material, and rotates polarization-dependent polarization of light. By appropriately selecting the thickness of the crystals, two νυν 48 (please read the precautions on the back before filling in this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 The entire polarization rotation angle of 46 wavelengths is substantially different. The difference between these rotation waves can be achieved by optical components that depend on polarization (such as the Brewster window of a laser discharge cell). (Unit price of the material) ⑽The transmission intensity ratio between P-polarized and s-polarized waves is displayed in a window: 〇17. The loss of s-polarized waves is due to reflection on the surface. Because of this The chamber has two windows. This value must be taken at its fourth intensity to produce a correct complete round-trip ratio of 1: 〇24. If the entire polarization rotation angle on the pass through the crystal pair is exactly 90 degrees, Then one of the most appropriate differences of these lines is reached. This will be achieved by adjusting the thickness of the crystal to the thickness of the characteristic quarter wave plate. However, the wavelengths of other lines should not undergo such rotation. In fact, at this wavelength The polarization rotation should be an integral multiple of 180 degrees (half-wave plate), so that the discriminating element does not affect the transmission of this wave. Therefore, this diffusion affects the refractive crystal, the polarization element (Brewster window), and the rear The combination of a mirror (for the second time the light returns through the crystal) suppresses one of the wavelengths, while the other wavelength remains unaffected. The advantage of this delta is that it is firm, easy to adjust, and reduces optical components And the possibility of using only anti-reflection coatings on the crystal for small (near-zero) incident angles. Figure 16C1 outlines a line-selective h laser system with an intra-cavity Lyot filter. This resonator is made up of The following structure, high reflection mirror 6 A, birefringent diffusion crystal 116B, (chamber) Btewster windows 116C and 116D, and output coupling mirror 116E (which is partially reflective). Its optical gain is generated in the chamber 丨 丨 矸Figure 16C2 shows another design, which uses one or several other paper sizes that are applicable to China National Standard (CNS) A4 (210X297 mm) (Please read the note on the back first And then fill out the entry page) - Order ----- 564585 A7 ______B7_ V. Description (47 invention)

Brewster element 116G to increase the difference between p- and s-polarization. Figure 16C3 shows another embodiment for a high-reflection mirror and crystal. In fact, the two elements can be combined by applying a dielectric reflective coating 116H directly on the back of the crystal 116B, which will reduce the number of required optical elements. The line selector behind M0 is shown in Figure 1 The illustrated embodiment provides a line selector 'under the main oscillator and no line selector in the resonant cavity of the main oscillator. In another embodiment, the line selector may be included in the resonance cavity, for example, the rear of the shooting chamber 108. This line selector can be added to or replace the 16A line selector. When the chirped line selector is used in a resonant cavity, it is more important to minimize the optical loss in the selected polarization. The preferred design of this line selector consists of five CaF or MgF 稜鏡, the direction of which is to enter and leave the light at a Brewster angle (about 57.3 °). This allows the use of radon without anti-reflective coating. The diagram shown in Figure 6 shows how this line is selected. This configuration includes five 稜鏡 118A which have a vertex angle equal to 2X (90 ° -ΘΒ) 'and ㊀BSBrewster's angle. This reflective optical device 118B is a half-thickness having a maximum reflective coating 118C behind it. The new feature here is to avoid the need for anti-reflection on the incident surface by using Brewster's angle (providing zero reflection coefficient for appropriately polarized lasers) and placing the reflective coating directly on the rear surface of this optical device coating.稜鏡 Wheel out coupler This is designed as the output surface of the gas discharge laser of the exhaustor. It is typically a partial mirror, which is usually a wedge-shaped optical element with a surface. A4 specifications (210X297 mm) 50!… ::: ",'# ------ · :: .......----- (Please read the precautions on the back before filling in this Page), τ · 564585 A7 -----_— B7__ 5. Description of the invention (48) The light path is loosened and is coated to reflect the desired part of the light and transmit the rest. This other surface is often coated with an anti-reflection coating and angles out of the lateral direction for the optical path diameter so that any reflection from this surface does not return to the gain region. When used in high-intensity UV applications, this coating surface often causes problems with usage. A solution is shown in Figure 17. In this case, the wheel-out coupler 120 is in the shape of 稜鏡, whose front surface (closest to the gain region) 疋 is oriented at the lowest loss angle (for P-polarized polarization), and the second surface is a thunder perpendicular to the refraction. Rays of light are provided to provide reflected light with magnification. This design eliminates the need for AR coatings and provides additional spectral separation due to light diffusion. For this F2 application, 稜鏡 is composed of CaF2 with an apex angle of 32 7 degrees and an incident angle of 57.2 degrees. In this preferred embodiment, there is no coating on the second surface and a Fresnel reflection of about 4.7% provides sufficient reflection for the main oscillator. Although the optical element's light steering strives to maintain a constant condition in the light path, as explained above, many laser operations such as burst mode operations can cause transient conditions, which in some cases cause a momentary moment when the output laser light is significant. Turn. This momentary turning of light can be corrected by an active light direction control system, which includes a light direction monitor and a light direction control mechanism. In a preferred embodiment, the light direction monitor is a partial detector and is known as a bi-cell detector or a segment detector. This type of detector has two individual optically sensitive elements separated by a small gap. The output ratio of these two components is measured in the direction of the light. This light direction control mechanism can be a pivot lens, preferably 51 (Please read the precautions on the back before filling out this page) This paper size is applicable to China National Standard (CNS) A4 (210X297 mm) 564585 5. Description of the invention ( 49 is in the line selection device 1GC in Fig. 1. Alternatively, one of the 稜鏡 in the line selector unit shown in Fig. 16A can be rotated by dragging the axis. It is mounted on a cymbal plate, so the plate itself can rotate with a polar axis. The driver that generates the pivot is preferably a piezo drive port or can be a voice coil or a stepper motor drive unit or any other type of cymbal unit. The control for the directional control mechanism should preferably include processing benefits, which are programmed with appropriate feedback algorithms and other electronic controls and swivel to allow the operator to adjust the direction of the light. The light steering with cleaning pressure is compensated as above It is shown that due to, for example, operations commonly used in laser integrated circuit lithography burst mode operations, a small amount of light in the light path can be diverted. In the light direction Even very small changes are extremely undesirable. As shown above, techniques can be used to eliminate the cause of light turning. Also, pivoting optical elements such as chirps or mirrors can correct undesired changes in light direction Another method is to correct the undesired light direction change by controlling the pressure of the cleaning gas in the part of the light path. In a preferred embodiment, the cleaning pressure in the online selection device is controlled to compensate for the change in light direction. This case The applicant has decided to use the five-line selector shown in Figure 16A. The output light direction for the five-line structure is about the pressure of the clean gas in the range of about 1 atm by the following coefficient: Δφ = 15 millimeters Diameter / atm. This light direction is preferably monitored by the branch detector as discussed above, and the feedback signal adjusts the pressure in the LSP by controlling the clean air flow valve. Another method is to use a temperature sensor to provide the feedback signal . -52 The size of the paper used in this edition applies to the Chinese National Standard (CNS) A4 (210X297 mm) 564585 A7

564585 A7 B7 V. Description of the invention (51) C is provided with a sphere and a hole slot to restrict rotation around the A-B axis. Laser Chamber 4 Hz Operation The preferred embodiment is designed to operate at a pulse repetition rate of 4000 pulses per second. Clearing this discharge-affected gas discharge zone between pulses requires an airflow between electrodes 18A and 20A to reach approximately 67 m / s. To achieve this rate, the diameter of the tangent fan is set to 5 inches (the blade structure is 26 inches long), and the rotation rate is increased to about 3500 revolutions per minute. In order to achieve this performance, this embodiment uses two motors which together generate a driving power up to about 4kw to this fan blade structure. At a pulse rate of 4000Hz, this discharge will add about 12kw (仟 W) of thermal energy to the laser gas. In order to remove the heat generated by the discharge and the heat applied by the fan, four individual water-cooled heat sink heat exchange units 58A are provided. The motor and heat exchanger will be described in detail below. The preferred embodiment of the present invention uses four heat sink water-cooled heat exchangers 58A shown generally in FIG. Each of these heat exchangers is somewhat similar to the single heat exchanger shown in Fig. 58 but has substantial improvements. Heat exchanger element A cross-sectional view of the heat exchanger is shown in Figure 21. The middle section of this heat exchanger is cut away but shows both ends. Figure 21A shows an enlarged view of a heat exchanger terminal that adapts to thermal expansion and contraction. The elements of this heat exchanger include a heat sink structure 302, which is machine made of solid copper (CU 11000) and includes 12 heat sinks per inch. 3030 Water flows through an axial channel with a hole diameter of 0.33 inches. The plastic spoiler 306 in the channel prevents water from forming a layer in the channel, and prevents the paper size in the channel from applying the Chinese National Standard (CNS) A4 specification (210X297 mm) 54 ---------- mushroom ... : (Please read the precautions on the back before filling out this page) Order ·… line-564585 A7 B7 V. Description of the invention (52) A thermal boundary layer is formed on the surface. The flexible flange unit 304 is a welded unit and is composed of an inner flange 304A, a blower 304B, and an outer flange 304C. The heat exchanger unit includes three c-closures 308 to seal the water flow in the heat exchanger from the laser gas. The blower 304B allows the heat exchanger to expand and contract the chamber. The Sunwell nut 400 connects the heat exchanger channel to a standard 5/16 inch positioning elbow device, which itself is connected to a water source. The 0-ring 402 provides a seal between the nut 400 and the heat sink structure 302. In this embodiment, the direction of the cold flow in the two units is opposite to the direction of the cold flow in the other two units to minimize the axial temperature gradient. Spoiler In a preferred embodiment, the spoiler is composed of four off-the-shelf, long-term on-line mixing elements, which are typically used to mix epoxy components and can be used by 3M Corporation (static mixer, (Part No .: 06-01229-00). This online mixer is shown outside 306 in Figures 21 and 21A. The on-line mixer forces water along a generally spiral path that reverses it clockwise about every distance (which is 0.3 inches). The spoiler substantially improves the performance of the heat exchanger. Tests performed by the applicant in this case have shown that adding this spoiler can reduce the required water flow by a factor of about 5 to maintain comparable gas temperature conditions. Airflow path and sound effect In this embodiment, the inflow and outflow of gas into the discharge zone is greatly improved for the laser chamber of the conventional technology. This discharge is formed with the upstream region near the exit of the AC fan to form a stable transition from a large cross section to a small cross section of the discharge. Before this gas is forced to turn 90 ° and enter the heat exchanger, 55 ----------_: · (Please read the precautions on the back before filling this page) This paper size applies to Chinese national standards (CNS) A4 specification (210X297 mm) 564585 A7 ------------ B7 __ V. Description of the invention (53) '' "--The cross section of the direct downstream area of this discharge is small The discharge value increases smoothly to a larger value. This configuration minimizes the pressure drop and associated perturbations caused by high-speed airflow in quick steps. Providing a smooth and gradually expanding airflow path in the direction away from the laser also reduces the undesirable sound effect, which is caused by the sound wave reflected by the pulse back to the discharge area at the next pulse. Techniques for reducing this effect are described in U.S. Patent Nos. 6,212,221 and 6,317,447, both of which are incorporated herein by reference. The time required for this sound wave to return to the discharge zone depends to a considerable extent. The result of this reflection from a particular surface is only problematic if the combination of repetition rate and gas temperature is special. If this reflective surface cannot be easily removed, the other approach is to avoid operation at a combination of problematic temperature repeat rates. One solution is to program this laser controller to automatically change the required gas temperature to avoid operating in problematic combinations. Blower motor and large blower The first preferred embodiment of the present invention provides a large tangential fan driven by a dual motor. The preferred configuration shown in Figure 24 provides an air flow of 67 m / sec between the electrodes, which is sufficient to clear a space of approximately 1.7 cm in the discharge area between 4,000 Hz pulses. 64A in Fig. 4 shows a cross section of this fan blade structure. The perspective view is shown in Figure 18A. Its blade structure has a diameter of 5 inches and is machine-made from a solid lead alloy 6061-T6 rod material. Individual blades in each section are slightly offset from adjacent sections as shown in Figures 18 and 8. This offset is better made uneven to avoid any pressure wave front. As an alternative, these 4 individual pieces can slightly angle the blade axis (again to avoid this paper size applying the Chinese National Standard (CNS) A4 specification (210X297 mm) 56

(Please read the precautions on the back before filling this page) Order — 564585 A7 _____B7_ V. Description of the invention (54) Generation of pressure wavefront). These blades have sharp front edges in front of them to reduce sound reflections from the blade edges facing the discharge zone. The embodiment shown in FIG. 18 uses two three-phase brushless dc motors each with a magnetic rotor contained in a metal pressure cup. This rotor separates the wound of the motor from the laser gas environment as in the United States Illustrated in Patent No. 4,950,840. In this embodiment, the pressure cup is a thin-walled nickel alloy 400, 0.016 inches thick, which acts as a laser gas barrier. The two motors 530 and 532 drive the same shaft and are stylized to rotate in opposite directions. These two motors are sensorless motors (ie, they operate without a position sensor). The 'right motor controller 534 controls the right motor 530, and its role is to control the slave motor control via analog and digital signals through the master The device 536 performs start / stop, current command, current feedback, and the like. Communication with the laser controller 24A 进入 Enter the main controller 534 via the RS-232 serial port. Cleaning system A first embodiment of the present invention includes an extremely pure N2 removal system that provides significantly improved performance and substantially increases the useful life of the components. Fig. 19 is a block diagram showing important features of the first preferred embodiment of the present invention. In this embodiment of the system, the five excimer field shots cleaned by nitrogen L · LNP 2P, the high-voltage element 4P installed on the laser chamber 6p, connecting the high-voltage element 4p and the upstream pulsed power element 10p. The high-voltage cable 8p outputs the converter 12p and the wavelength meter 14p. Each component, outside, kiss and 14P are each contained in a sealed container or chamber, and each has only two ports: one port for input and one port for output. One ~ source tells them that the core of the integrated circuit is typically a large plutonium bucket (typically kept at the temperature of liquid nitrogen Λ4 ^ -Π7-- (Please read the precautions on the back before filling this page)

• May be I 564585 A7 B7 V. Description of the invention (55 degrees), but it can be quite small N2. ^ The gas is discharged from N2 # 16p, passes through the N2 cleaning module 17p, and passes through the N2 filter 18p to the distribution panel 20p. It contains an airflow control valve for controlling the N2 flow to the element to be cleaned. With regard to each component, this clean flow is led back to the module 17p to the airflow monitoring unit 22p, where the airflow returned by each cleaning unit is monitored, and if the monitored airflow is less than a preset value, it is activated (not (Illustrated) Alarm bell. FIG. 19 is a line drawing showing the special elements of this preferred embodiment, which includes some other N2 features that are not particularly relevant to the cleaning features of the present invention. N2 filter An important feature of the present invention is the inclusion of an N2 filter 18. In the past, manufacturers of excimer lasers for integrated circuit lithography believed that filters for N2 clean gas were not necessary, because the Q2 N2 gas specification available at high fields was almost always sufficient So good that the gas that meets the specifications is clean enough. However, the applicant in this case found that sometimes the source gas may not meet specifications or the N2 pipeline leading to this cleaning system may contain contamination. This line can become contaminated during maintenance or operating procedures. The applicant of this case has determined that the cost of this filter is a very good insurance against damage caused by even a low probability of contamination. A preferred N2 filter is a 500K inert gas purifier available from Aeronex Inc. (its office is in San Diego, California). This filter removes N20,02, C0, co2, H2 and non-methane hydrocarbons to a level below 1 part per billion. It removes all particles of 0.003 microns or larger that are 99.9999999. 58 (Please read the precautions on the back before filling this page) This paper size applies to Chinese National Standard (CNS) A4 (210X297 mm) 564585 A7 B7 V. Description of the invention (56) Airflow monitor for each of these five The element is provided with an airflow monitor in the unit 22. This is a unit available in shopping malls with alarm features for low airflow. Piping All piping is preferably made of stainless steel (316SST) and has an electronically polished interior. Plastic pipes made of PFA 400 or ultra-high-purity Teflon can also be used. Recycling and cleaning As shown in Figure 19B, some or all of the cleaning gas can be reused. In this case, add a blower and water cooling or heat exchanger to the cleaning module. For example, the cleaning stream from the optical optics may be recycled and the cleaning stream from the electrical components may be depleted, or a portion of this combined stream may be depleted. Therefore, an ozone cleaning element can be added to remove ozone by the light path enclosed thereby. This may include a filter made of one of several materials that react with 03. Helium cleaning of LSP In a preferred embodiment, the line selection device is cleaned with helium, and the rest of the optical path diameter is cleaned with nitrogen. Helium has a much lower refractive index than hydrogen, so the thermal effect in LNP is minimized with the use of helium. However, helium is about 1,000 times more expensive than nitrogen. And, with helium, it will be more difficult to control the turning of light by generating cleaning pressure. The better technology of the improved seal for accommodating the light path is described in US Patent Application No. 10 / 000,991, filed on November 14, 2001, and marked 59 (please read the precautions on the back before filling (This page) This paper size is in accordance with China National Standard (CNS) A4 (210X297 mm) 564585 A7 -------B7 _ 5. Description of the invention (57) '一 " entitled "With & good light The path of the gas discharge laser is incorporated here as a reference. Figures 1, 3, 4, and 5 show the easy-to-seal blower seal ", ', which is used to provide the Sealed, but allows quick and easy removal of the module to allow quick replacement of the module. Easy-to-seal blower seal The applicant of this case has developed an easy-to-seal blower seal that allows the light path to be quickly sealed to meet the requirements of the vacuum when the laser module is reinstalled in the light path. Readers of this manual should pay attention Although the vacuum quality seal of each sealing part of the light path provided by this seal is not operated in a vacuum, it is typically operated in a pressure slightly exceeding the atmosphere. It is very important to be quickly sealed ’because there is a great need for these modules to be replaced within minutes. The basic design of the easy-to-seal blower seal is shown in Figures 8A-E. This easy-to-seal blower seal is a four-part seal. These four parts are ... (1) the blower part 93A shown in Figure 8A, (2) the flange part 93B shown in Figures 8A and 8B, and (3) shown in Figure 8A The metal c seal ring 93C shown in the figure, and (4) the first compression ring clip 93D shown in FIG. 8C. Another second compression ring clamp is shown in Figure 8E. The assembled easy-to-seal blower is shown in Figure 8D. These two additional metal c seals can be used to seal the seal flange portion 93B to the first laser portion 93E and the seal blower portion 93A to the second laser portion 93F. These additional seals are provided in the grooves 102 and 104. The flange portion 93B is sealed to the first laser portion by a bolt through the anti-sinking hole 106, and the bolt is tightened by a square-hole screw head wrench through the hole 108. This paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) 60;: .............. J-! (Please read the precautions on the back before filling this page) Order 丨564585 A7 B7 V. Description of the Invention (58) The flange portion 93B includes an inclined flange 120. As shown in Figure 8A, this flange has a slope of 20 °. The flange 114 also has a 20 ° slope. The compression clip 39D is then opened by unscrewing the finger bolt 118 and placed around the inclined flanges 120 and 114. The compression clip 93D has a hinge portion 122 and a bolt portion 124. It has the form of mating flanges 114 and 120 around the angled slotted interior. The diameter of the hole where the bolt 118 is completely inserted is slightly smaller than the inclined surfaces of the matching flanges 114 and 120, so that when the bolt 118 is bolted, the two flanges are forced together to compress the seal 93C between them. Create a vacuum-tight seal. The applicant in this case determined that a compression force of 400 pounds is preferred to ensure the required vacuum seal. This requires applying a torque of about 40 inches per pound to the handle of the bolt 118 of the first compression ring clamp. In this preferred embodiment, the handle is only 1 inch long, so most technicians need a quick wrench (or similar tool) to provide the 40 inch one pound torque. If a two-inch handle is provided, this seal can be achieved by hand. When the bent lever arm 119 is pushed into a position around the ring, the second compression ring clamp shown in Fig. 8E forces the two oblique flanges together. The clip is opened by rotating the arm around the ring, and then the two halves of the ring clip can be separated. The applicant in this case estimated that this 40 pound force applied to the end of the lever arm 119 caused a compression force of about 400 pounds on the c-seal 93C. The design of this clip is known as a “pull-out toggle clip” based on what is available in the mall. The important advantages of this sealing system are: (1) The time it takes to produce this seal is not important (about 1 to 2 minutes). (2) Produce excellent vacuum seal; (3) Avoid substantial vibration coupling between the chamber and optical components; and this paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) 61 ----- \ ----- · --.....----- (Please read the notes on the back before filling out this page) ·· τ · 564585 V. Description of the invention (59 Compared with most In terms of other vacuum sealing technologies, this seal is not as shown in Figure 8A. This seal is made between the flange part gauge and 93A. It has a metal e seal. 9 is sandwiched between these two parts. The metal ring seal 93D is used as shown in Fig. 8D. The metal seal is installed in the groove 10. These shells are made to be slightly rounded to fit into the circular groove. 110. The longer diameter of this e-seal ring is 1.946 inches and the shorter diameter is 1.84 inches. The spring force in the elliptical c-seal ring bears against the edge of the groove 110, which prevents the seal ring. Detach during assembly. The blower section 93A includes a round ridge 112 and its protective sealing ring 93C prevents it from being scratched by the component 93B when it and the component 93B slide against each other during assembly. Lateral clean airflow The applicant of this case found through experiments with an F2 laser (at a high repetition rate of 4000 Hz, several milH Joules per pulse of light) that due to the interaction between light and clean gas The light will be diffused and deflected instantaneously. The applicant of this case has also determined that these effects can be minimized by generating a clean air flow across the path of the light. The applicant of this case identified several techniques to achieve this. In 19C1 Figures 19C2, 19C3, and 19C4 show four such techniques. Figures 19C1 and 19C2 show adjustment devices in the optical path that encourage this traversing clean flow. Figure 19C3 shows a fan in the cleaning circuit used to apply The cleaning gas is re-circulated and the cleaning stream is directed across the light path by the cleaning nozzle in Fig. 19C4. Aligning the laser provides a restricted vacuum cleaning path so that the laser optics applies the Chinese national standard for this paper size ( CNS) A4 specification (210X297 mm) 564585 5. The description of the standard of the invention (60) becomes complicated. In the system of the conventional technology, this clean road must be Is cut to insert a small visible light alignment laser. In the preferred embodiment, a small visible light laser can be included as a permanent part of the light path, which is very useful during maintenance operations. This alignment laser is more A helium-neon laser or a small diode laser installed on the rear side of the high-reflection mirror 10D is preferred; for this design, the mirror should be composed of a CaF plate, which has a dielectric reflective cold layer and is designed for Reflects a very large part of ultraviolet light at 157 nm, but transmits a very large part of visible light. This alignment laser can then be used to align the entire optical line through mo, pa and light extender, and This cleaning path was not cut off. (The reader of this case should understand that any line selection optics will change the direction of the laser. You can remove this line selection optics to align the rest of the system, or you can adjust this considering the line selection optics. Align with the direction of the laser light.) ° Advantages of the vacuum-quality light path The vacuum-quality cleaning system described here represents a major improvement in the long-term performance of excimer lasers, especially F2 lasers. The problem of contamination is basically eliminated, which results in a substantial increase in the component life and light quality. In addition, since the leakage except through the output port has been eliminated, the airflow can be controlled to a desired value, which has the effect of reducing the amount of demand by about 50%. Closed Switching Unit with Power Meter As shown in Figures 20, 20A and 20B, this first embodiment includes a closed switching unit 500 which has a built-in power meter. With this important improvement, this switch port has two functions: No.-as a switch to block the laser light, and the paper size is oriented toward the towel (4) A4 size (0 × 297). '--- 564585 A7' ^ ----—- _ B7 V. Description of the invention (61) Second 'When measurement is needed, it is used as a complete light power meter to monitor light power. Figure 20 is a top view showing the important components of the switch unit. These are the switch 502, the light dump 504 and the power meter 506. Figure 51 in Figure 20 shows the path of the laser output light with the switch in the off position. The path with light on is shown at 512. The active surface of the switch of the light terminating element 5 丨 6 is 45 with the direction in which the light exits the chamber. When the switch is turned off, the daylight 'light is absorbed in the surface of the switch and reflected to the light dump 504. Both the active surface of the light dump and the active surface of the switch are plated with chrome to highly absorb laser light. In this embodiment, the light terminating element 516 is mounted on the flexible elastic yellow steel arm 518. As shown in Figure 20B, the switch is opened by applying a current to the coil 514, which pulls the flexible arm 5 and 8 and the light terminating element 516 to the coil, and outputs the light terminating element 516 from the laser. The path to go away is removed. This switch is closed by stopping the current flowing through the coil 514, which allows the permanent magnet 520 to pull the light terminating element 516 and the flexible arm 518 back into the closed position. In the preferred embodiment, this current is carefully designed to cause easy movement of the components and arms between the open and closed positions. As shown in Figures 20 and 20A, the power meter 506 operates in a similar manner and places a pyroelectric photodetector in the path of the laser output light. In this case, the coil 520 and the magnet 522 pull the detection unit 524 and its flexible arm 526 in and out from the light path for output power measurement. This power meter can be operated with on and off switches. The current and switch to the coil are controlled so that the unit 524 can be easily sent in and out from the light path. This paper size applies to China National Standard (CNS) A4 specification (210X297 mm) 64 (Please read the precautions on the back before filling this page), OK — 564585

V. Description of the Invention (62) The modified polarization technique with room window angles greater than the Brewster angle is often placed at the Brewster's angle, which causes about 58% of the polarization in the p-polarization direction 100% of the direction of transmission. In other conventional technical designs, this window is set at about 45 degrees. In this case, the transmission of the s-polarization is slightly smaller and the p-polarization is slightly larger than the above value. For F2 lasers, there is substantial competition in the gain region between 8 and 1) polarization. This is because, compared with the conventional technologies KrF and ArF lasers, there is typically a very high gain in the discharge area of the dagger laser. This competition is undesired because s-polarized light is useless in most applications and is typically lost as unwanted heat. Therefore, it is necessary to minimize the amount of S-polarization generated in the laser. A better technique that can be achieved fairly easily is to increase the angle of incidence of the chamber window to be substantially larger than the Brewster's angle. For example, in Brewster, the s-angle is approximately looG / ^ ip-polarized light transmission for 157 nm F2 rays and approximately 83% of s-polarized light transmission. If you increase the incident angle to the material. , Then the & polarized transmission is reduced to about 99%, but only 76% of the polarized light transmission. Because the output light from the gain region in the main oscillator passes through these two windows approximately two times (for 4 passes through the window and two surfaces for each window). The ratio of the two polarizations in the output light (assuming a window angle of 64.) is 89% Ό.76Υ 099 The polarization disappearance ratio = -1 + 1 + 47 for the applicant in this case. Window angle test, about 72〇 / 〇 light is p-biased

564585 A7 __ΒΊ_ 5. Description of the invention (63) polarization and 28% is s-polarized. By changing the angle of the window to 64% and adding another window at the 64 angle before the high mirror 10D in Figure 1, the proportion of s-polarization in the output light is reduced to about 4%, and the rest 96% of the light is p-polarized. Various modifications can be made to the invention without changing its scope, and those skilled in the art will appreciate many other possible variations. For example, as shown in Fig. 5, the pulse power circuit up to the output of the pulse transformer 56 may be a common circuit. This method provides a further reduction in instability as described in U.S. Patent Application No. 09 / 848,043, which is incorporated herein by reference. Fig. 3B of the present invention shows the input and output to the pulse transformer. It is included as Fig. 13 for the convenience of the reader. Other heat exchange designs should be a significant modification of the structure shown here. For example, all four units can be combined into a single unit. The use of larger fans on heat exchangers can have significant advantages to reduce the effects of rapid changes in gas temperature due to sudden laser operation. Readers of this case should understand that feedback control on pulse energy at very high pulse rates is not necessarily fast enough to control the pulse energy of a specific pulse using the pulse just passed. For example, control technology can be provided and the measured pulse energy for a particular pulse can be used to control the following second or third pulse. Many layout designs other than those shown in Figure 1 can be used. For example, these chambers can be mounted side-to-side or with PA on the bottom. Moreover, by including an output coupler (such as a partial mirror), this second laser unit can be designed as a driven oscillator, and other changes are possible. Fans other than tangential fans can be used. This may require much greater repetition rates than 4KHz. These fans and this paper size are applicable to China National Standard (CNS) A4 specification (210X297 mm) -66-(Please read the precautions on the back before filling this page)

564585 A7 B7 5. Description of the Invention (64) The heat exchanger may be located outside the discharge chamber. The pulse timing technique described in U.S. Patent Application 09 / 837,035 (herein incorporated by reference) may also be used. You can use line selection techniques other than those described above. For example, 3, 4, or 6 cymbals can be used and the design techniques discussed above are used to select the strongest line. You may want to measure frequency bandwidth accurately. This can be achieved by using a standard with a smaller free frequency range than the one described above. Other well-known techniques can be adjusted to accurately measure the bandwidth. Therefore, the above disclosure is not intended to be limiting, and the scope of the present invention should be determined by the scope of the attached patent application and its legal equivalent. (Please read the precautions on the back before filling this page) •, ^ Ti This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A7 _B7 V. Description of the invention (65) Component label comparison 2 ... 12B ... Compression head 4 ... Laser box 13 ... Pulse amplification module 6 ... Power module 14 ... Wavefront engineering box 6B ... High voltage power supply 15A ... Mirror 8A ... Converter module 15C ... Mirror 10 ... Main oscillator 15D ... mirror 10A ... discharge chamber 18 ... switch module 10B ... compression head 20 ... gas control module 10C ... line selection device 22 ... water distribution module 11 ... vertical optical table 28 ... laser control module Group 12 ... Power amplifier 30 ... Status light 12A ... Discharge chamber ...: Ί:; ---------- Binding: ... (Please read the precautions on the back before filling this page) This paper size applies to China Standard (CNS) A4 Specification (210X297 Public Love) 68

Claims (1)

’, Patent application scope 0 Yue Xi 10 15 20 types of two-chamber high-repetition 6-gas discharge laser systems with very narrow frequency bands' include: A) a first laser unit, including: D a first discharge chamber, including a) a first laser gas, b) a first pair of extended electrodes defining a first discharge zone, 2) a first fan, when the laser unit is operated at a repetition rate in the range of 4,000 pulses per second or higher, the Sufficient speed to generate the first laser gas in the first discharge zone, and substantially remove all ions generated by the discharge from the first discharge zone after each pulse before the next pulse, 3) the first heat exchange system Can remove at least 16kw of thermal energy from the first laser gas, B) a line selection unit for minimizing energy outside the selected first line spectrum, C) a second laser unit, including: 1) a Two discharge chambers, including: a) a second laser gas, b) a second pair of extended and separated electrodes defining a second discharge zone, 2) a second fan, when the laser unit emits 4,000 pulses per second When the repetition rate operation in the range of or higher is generated in the second discharge region, Sufficient speed of the second laser gas to substantially remove all ions generated by the discharge from the second discharge area after each pulse and before the next pulse, the paper size applies the Chinese National Standard (CNS) A4 specification (210X : 297 mm) gutter 69 VI. Patent application scope 3) The first-parent heat exchange system can remove at least 16kw of thermal energy from this second laser gas, D) Pulse power system, which is designed to provide electrical properties Pulse to the first pair of electrodes and the second pair of electrodes, enough to generate a laser pulse at a pulse rate of about 5 pulses per second of about 4000 pulses per second, which has an accurately controlled pulse energy of more than about 5 mJ, E) laser light A measurement and control system for measuring the pulse energy of the laser output pulse generated by the two-chamber laser system, and controlling the laser output pulse in a feedback control configuration, 10 is characterized by using the first- The output laser light of the laser unit is used as a seed light for seeding the second laser unit. 2. The laser system according to item 丨 of the patent application scope, wherein the first laser unit is designed as a main oscillator, and the second 15 laser unit is designed as a power amplifier. 3. The laser system according to item 2 of the patent application range, wherein the first laser gas includes fluorine and neon. 4. The laser system according to item 2 of the patent application scope, wherein the first laser gas includes fluorine and helium. 20.5. The laser system according to item 2 of the patent application range, wherein the first and second laser gases include fluorine and a buffer gas selected from the group consisting of neon, helium, or a mixture of neon and helium. 6. The laser system according to item 2 of the patent application scope, in which the power amplifier is designed for a single discharge of light through the second discharge. 6. Patent application area 8882? ABCD 〇 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 10 7. The laser system of item 2 in the scope of patent application, in which the power amplifier is designed to pass light through the second discharge multiple times. The laser system of claim, wherein the main oscillator includes an optical element which provides a resonance path which passes through the first discharge region twice. 9. The laser system according to item 2 of the patent application scope, wherein the main oscillator includes an optical element which provides a resonance path which passes through the first discharge region twice, and wherein the power amplifier includes an optical element to provide the light to pass through multiple times Second electric zone. 10. The laser system according to item 1 of the scope of patent application, further comprising: an optical table for supporting a resonant cavity optical device of the first laser unit regardless of the first discharge chamber. 11. The laser system according to item 10 of the patent application scope, wherein: the optical table 5 is generally a type 11 and defines a u-shaped cavity, and wherein the first discharge chamber is installed in the ϋ cavity. 12. The laser system according to item 1 of the patent application scope further includes: a vertically mounted optical table having a second discharge chamber mounted on the vertical optical table. 13. The laser system according to item 1 of the scope of the patent application, wherein: each of the first and second laser chambers defines an airflow path, which has a gradually increasing cross section downstream of the electrode to allow recovery in the discharge zone. The percentage of large static pressure drops that occur. Home Standard (CNS) A4 specification (210 X 297 mm) Please read the back first. Fill in the notes and write the binding page 2 A8 B8 C8 ---__ 21 、 Applicable patent scope M · If applying for patent scope The laser system of item 2, wherein the first and first chambers each include a wind vane structure downstream of the discharge zone for normalizing the gas velocity downstream of the discharge zone. 5. If the laser system of the first scope of the patent application, where. The Heidi fan and the second fan are each a tangential fan, and each includes a shaft driven by two brushless DC motors. 16. The laser system of claim 15 in which the motor is a water-cooled motor. 17. The laser system according to item 15 of the scope of patent application, wherein each α-hai motor includes a mule and each of the motors includes a magnetic rotor contained in a pressure cup, which separates the stator from the laser gas. 18. The laser system according to the scope of the patent application, wherein the first and second fans are each a tangential fan, each of which includes a blade structure made of the lead raw material by a machine. 19. The laser system of claim 15 in which the blade structure has a diameter other than about five inches. 20. The laser system of claim 19, wherein the blade structure includes a blade element having a sharp front edge. 21. The laser system according to item 15 of the patent application scope, wherein the motor is a sensorless motor and further includes: a master motor control state for controlling one of the motors, and a slave motor controller for controlling the other motor. 22. The laser system according to item 15 of the patent application, wherein each of the tangential fans includes a blade forming an angle with the shaft. Scope of patent application jlxj 23 · If the laser system of item i of the scope of patent application is applied, the heat exchange system of the heat sinks of the complex / A is water-cooled. 24. The laser system according to item 23 of the patent application, wherein each heat exchange system includes at least four separate water-cooled plutonium exchanges 10 ′, 25. The laser system according to item 23 of the patent application, wherein Each heat exchange system includes at least one heat exchanger having a tubular water flow channel, wherein at least one spoiler is located in the channel. 26. The laser system of claim 25, wherein each of the four heat exchangers includes a tubular water flow channel without a spoiler. 27. The laser system according to item 1 of the patent application range, wherein the pulse power system includes a water-cooled electrical component. 28. The laser system of claim 27, in which at least one of the water-cooled elements is operated at a high voltage exceeding 12,000 volts. 29 > The laser system of claim 28, wherein the high voltage is isolated from the ground using an inductor through which cold water flows. 30. The laser system according to item 1 of the patent application range, wherein the pulse power system includes: a first charging capacitor bank and a first pulse compression circuit for providing an electrical pulse to the first pair of electrodes; and a second charging capacitor The bank and the second pulse compression circuit are used to provide electrical pulses to the second pair of electrodes; and the resonance charging system is used to charge the first and * ·; * · · second charging capacitor banks in parallel to an accurately controlled voltage . 73 This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 564585 A8 B8 C8 D8 92 years / 0 only /;? R: scope of patent application 15 20 31. If the scope of patent application item 30 Laser system, wherein the resonance charging system includes a De-Qing circuit. 32. The laser system of claim 30, wherein the resonant charging system includes a bleed-down circuit. 33. The laser system of claim 30, wherein the resonant charging system includes a Q-removing circuit and a down-regulating circuit. 34. The laser system according to item 1 of the scope of patent application, wherein the pulse power system includes a charging system ′, which is composed of at least three power sources configured in parallel. 35. The laser system according to item 1 of the patent application range, wherein the line selection unit is located in the main oscillator. 36. The laser system of claim 35, wherein the line selection unit includes a plurality of thorium. 37. The laser system of claim 36 in the scope of patent application, wherein: the plurality of puppets are five puppets. 38. The laser system of claim 36 in the scope of patent application, wherein: the plurality of chirps are configured in a loop so that the laser light from the first f laser unit will turn 36 before entering the second laser unit. . 39. The laser system according to item 1 of the patent application scope further includes: visible light directed at the laser. 40. The laser system according to item 1 of the patent application scope, wherein: the line selection unit includes a Lyot filter. 41. If the laser system of the first scope of the patent application, where: the first discharge chamber and the second discharge chamber including the window, they are set in another standard (applicable to this page, please read the precautions on the back) ) Order ·) () 4585 ίο 15 20 bits ′ makes the incident angle of the laser light on this window larger than the Brewster angle. 42. The laser system according to item 1 of the patent application scope, further comprising: a light manipulating device for manipulating a laser light generated in the first laser unit. 43. The laser system according to item 42 of the patent application range, wherein the manipulation device includes a device for pivoting the optical element. 44. The laser system according to item 42 of the patent application, wherein the light control device includes a device for adjusting the pressure in the line selection unit. 45. If the laser system of the scope of patent application item 1, wherein the laser system includes a chirp output coupler which partially defines a resonant cavity for the -laser single it, the coupling device It includes two surfaces. The direction of the first surface is a low loss angle for p-polarization, and the second surface is located perpendicular to the laser light from the first laser unit. 46. If the laser system in the scope of the patent application, it further includes: A) a first temperature monitor for monitoring the temperature of the gas in the first discharge chamber, B) a-gas temperature control system, here The gas temperature control system includes a control method for adjusting the gas temperature to avoid adverse sound effects caused by reflected sound waves. 47. The laser system according to item 丨 of the patent application scope further includes: A) a second temperature monitor for monitoring the paper size application (CNS) Α4β (210X297 ^) in the second discharge chamber-> ------------ Raid ------------------, but ......... line (Please read the precautions on the back before filling this page) 564585 A8 B8 C8 D8 Patent application scope 10 15 20 Gas temperature, B) Second gas temperature control system, here this gas temperature control system includes control methods for adjusting Gas temperature to avoid undesirable sound effects caused by reflected sound waves. 48. The laser system according to item 1 of the patent application scope further includes a nitrogen filter. 49. If the laser system of the first patent application scope, further includes: Nitrogen cleaning system which includes a cleaning module, this module includes a flow monitor; the laser also includes a cleaning exhaust pipe for g to send from the laser Clean exhaust gas. 50. The laser system of item 1 of the scope of patent application, further comprising: a switch unit including an electrically operated switch and a power meter 'This unit can be positioned in the laser output optical path diameter by a command signal. 51. The laser system of claim 50, further comprising a light housing system comprising: A) at least one light seal including a metal blower, and B) a cleaning device for cleaning the air with a cleaning gas Light shell. 52. The laser system according to item 51 of the patent application scope, wherein: the light-containing device includes an airflow guiding device for generating a clean airflow 'which traverses the laser light generated in the second laser unit. 53. The laser system of claim 51, wherein: the at least one light seal is designed to allow easy replacement of the shooting chamber. 54_ If the laser system of the scope of patent application No. 51, in which: 'paper? Ft standard applies to the Chinese standard 芈 (CNS) M specifications (2) 0X297 round), ... [install ... ...........- ΙΤ ........ line (please read the note on the back before filling this page) Λ8 Λ8 The cleaning device is used to clean the isolation area with a cleaning gas. 59. The laser system according to item 1 of the scope of the patent application, wherein: The system is designed to be slightly modified to operate a KrF laser system, an ArF laser system or an f2 laser system. 60. The laser system according to item 1 of the scope of the patent application, wherein substantially all components are contained in the laser case, but the system includes an AC / DC module physically separated from the case. This paper size is applicable to China National Standard (CNS) A4 (2K) ×: 297 mm. 77 61. The laser system according to the scope of the patent application, wherein: the pulse power system includes a main oscillator charging capacitor bank, a power amplifier charging capacitor bank, and a resonance charger, which are designed to use the two charging capacitor banks Charging in parallel. 5 62. The laser system of claim 61, wherein the pulsed power system includes a power source, which is designed to provide at least 2000V to the resonant charger. 63. The laser system according to item 1 of the patent application scope further includes: a W gas control system for controlling the ρ2 concentration in the first laser gas to control the light parameters of the main oscillator. 64. The laser system according to item 1 of the patent application scope further includes: a gas control system for controlling the pressure of the laser gas in the first laser gas to control the light parameters of the main oscillator. 65. The laser system according to item 2 of the patent application scope, further comprising: 15. The discharge timing controller is used to trigger the discharge in the power amplifier 20 to 60 nanoseconds after discharging in the main oscillator. 66. The laser system according to item 2 of the scope of patent application, further including: The discharge controller is programmed to cause a discharge in some cases and such timing avoids any significant output pulse energy. 2 0 67 · The laser system of item 66 of the patent application scope, wherein: the controller is programmed in some cases and the discharge of the power amplifier is caused in the main oscillator in some cases Occurs at least 20 nanoseconds before. 68 · If the laser system in the first scope of the patent application, it also includes:-_ This paper size is applicable to the Chinese National Standard (CMS) A4 specification (210X297 mm) _-564585 ABCD 倏 See 7T, Pulse amplification unit for patent scope Used to increase the period of the output pulse from the laser. 69. The laser system according to item 68 of the patent application scope, wherein: the pulse amplifying unit is configured to receive the output pulse laser light, and the number of pulses per second is amplified by a factor of at least 2, so that a larger number of pulses is generated The single amplified output pulse light has a substantially reduced intensity value compared with the laser output pulse, and the pulse amplification unit includes: (1) a first beam splitter configured to separate a part of the output light, and 10 The separated portion defines a delay portion, and the output light is defined by the light size and angular spread of the first beam splitter; (2) the first delay path is started and ended by the first beam splitter, the The first delay path includes at least two focusing mirrors, which are configured to focus the delayed portion to a focal point in the first delay path, 15 and return the delayed portion to the first beam splitter, It has a light size and angular spread equal to or approximately equal to the light size and angular spread of the output light of the first beam splitter. 70. The laser system of claim 69, wherein the at least two focusing lenses are spherical mirrors. 207. The laser system according to item 69 of the scope of patent application, further comprising: The second delay circuit includes at least two spherical mirrors. 72. The laser system of claim 69, wherein: the first delay path includes four focusing mirrors. 73 · If the laser system in the scope of patent application No. 72, including: 79 9 > Year / Month The scope of patent application consists of the second and the second delay path located in the first delay path. 74 of the fourth series, such as the laser system of the 69th scope of the patent application, in which the first delay path includes a second beam splitter and further includes a second delay path, this path includes at least a focusing lens Is configured to focus the delayed portion on the focal point of the first delay path, and return the money delayed portion to the first beam splitter, which has a light size and an angular spread equal to or approximately equal to the first beam splitter The light from the wheel ^ ray size and angle spread. 75. The laser system of claim 69, wherein 5 uses the first beam splitter and uses the optical characteristics of unsuccessful internal reflection to direct the laser light in at least two directions. Order 76. If the laser system of the scope of patent application number 69, the first beam splitter is composed of two transparent optical elements, each element has a flat surface, and these optical elements are positioned so that these surfaces are on each other Parallel and separated less than 200 nanometers (nm). Line 77. The laser system of claim 69, wherein the first beam splitter is an uncoated optical element and forms an angle with the output laser light so that a desired reflection-transmission ratio is achieved. 564585 V / O //> Fixed 10D 6B