JP4095095B2 - Gas supply method for excimer laser device - Google Patents

Description

  The present invention relates to an excimer laser device used for a light source for reduced projection exposure, fine processing of a material, surface modification of a material, and the like, and more particularly to an improvement for stably oscillating a laser for a long period of time.

  Conventionally, when an excimer laser device is operated using a halogen gas, the halogen gas is consumed by evaporation of the electrode material and chemical reaction with the laser chamber constituent material according to the operation. Therefore, conventionally, the following control is performed in order to compensate for a decrease in laser output due to consumption of the halogen gas.

  That is, the laser output is obtained by charging the electric energy stored in the capacitor to excite the laser into the discharge space and discharging it in the laser medium gas. Will increase. Therefore, conventionally, the laser output is detected, and the laser output is stabilized by controlling the charging voltage value according to this detection. This control is called power lock control, and this charging voltage is hereinafter called power lock voltage.

  However, even if this control is continued for a long time, the oscillation efficiency is lowered due to the exhaustion of the halogen gas, and the predetermined output cannot be maintained unless the charging voltage (power lock voltage) is gradually increased.

  Therefore, conventionally, when a charging voltage rises above a predetermined voltage, a certain amount of halogen gas is replenished to cope with the consumption of the halogen gas.

  The halogen gas reinforcing method in this prior art will be described with reference to FIGS.

  That is, FIG. 29 shows a portion related to gas replenishment in a general fluorine-based excimer laser device. In this case, a halogen gas (F2) diluted with a buffer gas (Ne, He, etc.) is used as a gas injection cylinder. , Hcl, etc.), a cylinder 42 filled with a rare gas such as Kr, and a cylinder 43 filled with a buffer gas such as Ne or He. Gas is supplied to the laser chamber 47 by controlling the opening / closing of the valves 44, 45, 46, and gas is supplied via the on / off valves 48, 49 and the sub tank 50 when replenishing the halogen gas after operation.

  That is, when a new gas is injected into the laser chamber 47 before activation, first, the old gas in the laser chamber 47 is discharged by the on-off valve 51 and the vacuum pump 52. Next, Kr gas is introduced from the cylinder 42 into the 40 torr laser chamber 47 through the opening / closing valve 45, and then 80 torr of F2 gas diluted with Ne gas is introduced from the cylinder 41 through the opening / closing valve 44, and finally opened and closed. Ne gas is introduced from the cylinder 43 through the valve 46 so that the total pressure in the laser chamber 47 becomes 2500 torr.

  By such gas injection control, the gas composition in the laser chamber 47 in this apparatus is F2: Kr: Ne = 4: 40: 2456 (torr), that is, F2: Kr: Ne = 0.16: 1.60: 98.24 (%) concentration. It becomes a ratio.

  When new gas is filled in the laser chamber 47 in this way, gas replenishment control is performed according to the procedure shown in the flowchart of FIG. 30 during the subsequent laser operation.

  First, prior to operation of the excimer laser device, a target laser output Ec, an optimal control charging voltage range Vm (Vmin to Vmax), an increase / decrease charging voltage ΔV in one control, and a supplementary gas amount ΔG for one control are preset. (Step 500).

  Thereafter, when the operation is started, the laser output E detected by the laser output monitor 53 and the charging voltage V detected by the charging voltage detector 54 are taken into the controller 55 (step 510). The controller 55 compares the detected laser output E with the target laser output Ec (step 520). If E <Ec, the detected charging voltage V is increased by the minute voltage ΔV to the command charging voltage Va (step 530). If E = Ec, the detected charging voltage V is directly used as the command charging voltage Va (step 540), and if E> Ec, the detected charging voltage V is decreased by the minute voltage ΔV to be the command charging voltage Va (step). 550).

  Further, the controller 55 compares the command charging voltage Va with the maximum value Vmax of the optimum control charging voltage range Vm (step 560). If Va ≦ Vmax, the controller 55 returns to step 510 again to control the command voltage Va. . However, when Va> Vmax, F2 gas containing Ne gas is supplied from the cylinder 41 to the laser chamber 47 by a predetermined amount ΔG (step 570), and the total pressure in the laser chamber 47 is maintained at the predetermined pressure. The gas is partially exhausted (step 580).

  FIGS. 31A to 31D show time charts for the laser output E, the command charging voltage Va, the halogen gas (F2) concentration, and the rare gas (Kr) concentration by the above control. Takes the number of laser shots. In FIG. 5B, each time point t1 to t6 at which the charging voltage command Va to the capacitor suddenly falls corresponds to the replenishment time of the halogen gas.

  However, according to the conventional method, the gas cylinder 41 for replenishing the halogen gas does not contain a rare gas component. Moreover, according to this conventional method, every time gas is replenished, control is performed such that a part of the gas in the laser chamber 47 is exhausted to make the total pressure constant (step 580 in FIG. 30R> 0). As the gas is replenished many times, the rare gas (Kr) gradually decreases as shown in FIG. 31 (d), and 20% Kr gas disappears after 10 gas replenishments even by simple calculation. In other words, as shown in FIG. 31 (c), it means that the halogen gas (F2) is gradually oversupplied as the gas is replenished. Further, even if the single replenishment amount ΔG is set to be small so as not to be oversupplied, the replenishment interval is gradually narrowed as shown in FIG. That is, according to the above conventional method, every time the halogen gas is replenished, the optimum composition balance of the mixed gas in the laser chamber 47 is lost, and the laser output is kept constant no matter how much the charging voltage to the capacitor is controlled. It becomes impossible to do.

  In addition, according to the prior art, the halogen gas is replenished by a certain amount based on the charge voltage comparison unrelated to the optimum composition balance of the gas. Therefore, it is difficult to maintain the optimum composition balance, and the gas balance is caused by disturbance. There was a problem that it was easy to break.

  Also, according to the prior art, there are some which control the injection of rare gas, but since only the control of injecting a constant amount of rare gas intermittently has been performed, the stability of the laser output is lacking. There was a problem.

  In addition, according to the prior art, the gas replacement time and maintenance time are not automatically notified, so the operator determines this based on window dirt, gas deterioration, laser chamber life, etc. It was impossible for an inexperienced operator.

  By the way, when the laser oscillation is stopped, the gas in the laser chamber naturally deteriorates with the elapsed time. However, in the prior art, there has been a problem that the laser output is not stable when laser oscillation is started because no countermeasure is taken against the laser stop. This is particularly noticeable when the laser is stopped for a long time.

  The present invention has been made in view of such circumstances, and even if gas replenishment is repeated many times, the optimal composition balance of gas is not easily lost, and the gas replenishment time point and gas replenishment amount can be determined accurately, and laser An object of the present invention is to provide a gas replenishing method for an excimer laser device that can obtain a stable laser output even when the operation is stopped for a long period of time, and can automatically and properly notify the operator of the gas replacement time and maintenance time. And

  In the first invention, in an excimer laser device that performs laser oscillation by injecting a halogen gas, a rare gas, or a buffer gas into the laser chamber, the oscillation stop time is calculated before the laser oscillation, and the calculated oscillation stop time is a predetermined time. If it exceeds, the injection amount of the mixed gas of the rare gas and the buffer gas is calculated based on the calculated oscillation stop time, and the mixed gas is injected before the laser oscillation by the calculated injection amount. (Power correction subroutine 1).

  According to this invention, before the laser oscillation, the rare gas and the buffer gas are replenished by an amount corresponding to the oscillation stop time, so that the output that naturally deteriorates and decreases during the laser stop is returned to the original proper state. After that, laser oscillation is performed.

  In the second invention, the mixed gas of the rare gas and the buffer gas is not injected before the laser oscillation, and immediately after the laser oscillation, when the laser output during the laser oscillation exceeds the rated laser output, the charging voltage is detected, A mixed gas of a rare gas and a buffer gas is injected in an amount calculated according to the detected charging voltage (power correction subroutine 3).

  According to the second aspect of the invention, the rare gas and the buffer gas are not injected before the laser oscillation. Then, immediately after the laser oscillation, if the laser output during laser oscillation exceeds the rated laser output, the rare gas and the buffer gas are injected by the amount calculated according to the detected charging voltage. That is, when the laser output exceeds the rated laser output, the normal charging voltage is a considerably large value. In such a case, a large amount of noble gas and buffer corresponding to the large charging voltage value are used. By injecting gas at once, the charging voltage value is lowered at once.

  In the third invention, when the gas exchange request signal is output before the laser oscillation, the first power lock voltage in the current filling gas is detected, the detected value is compared with a predetermined threshold value, and the detected value is less than the threshold value. If it is smaller, gas exchange is performed with the same gas composition as before, and if the detected value is larger than the threshold, the injection amount of the mixed gas of rare gas and buffer gas is changed according to the deviation between the detected value and the threshold, Gas exchange is performed with the gas composition of the changed injection amount (power correction subroutine 1, power correction subroutine 2).

  According to the invention, when changing the gas, it is determined whether to change the gas composition according to the power lock voltage, and when changing the gas composition, the rare gas and the buffer gas are changed according to the power lock voltage. The amount of injection is determined.

  Hereinafter, the present invention will be described in detail according to embodiments shown in the accompanying drawings.

  When the operation of the excimer laser device is continued, the laser output decreases due to the consumption of halogen gas and the generation of impurities as the operation progresses. The laser output is obtained by charging the electric energy charged in the capacitor to excite the laser gas into the discharge space and discharging it in the laser medium gas.

Here, there are the following three methods as general means for compensating for the decrease in laser output. (1) Increase the charging voltage V to the capacitor.
(2) Inject Kr / Ne mixed gas to increase the total pressure.
(3) The F2 / Ne mixture gas is injected to compensate for the decrease in F2 gas.

First Embodiment A first embodiment of the present invention will be described with reference to FIGS.

  FIG. 1 shows the configuration of a gas injection system of a KrF excimer laser. Gas injection into the laser chamber 1 is performed by a cylinder 2 filled with halogen gas (fluorine in this case) F2 diluted with Ne as a buffer gas. A cylinder 3 filled with a buffer gas Ne and a rare gas (krypton in this case) Kr at an arbitrary mixing ratio, and a gas injection opening / closing disposed on each gas supply path from each cylinder to the laser chamber 1 This is done by a configuration consisting of valves 4 and 5.

  Further, gas is exhausted from the laser chamber 1 by a configuration including a gas exhaust opening / closing valve 6 and a vacuum pump 7.

  The cylinder 2 is filled with F2 and Ne at a concentration ratio of F2: Ne = 5: 95 (%), and the cylinder 3 contains Kr and Ne with Kr: Ne = 1.5: 98.5 (%). It shall be filled in the concentration ratio.

  The oscillator 8 is an internal trigger oscillator for performing pulse oscillation. An oscillation instruction signal as a pulse signal is output from the oscillator 8 to the discharge power supply 9 and is synchronized with the oscillation instruction pulse signal. A signal S is output to the controller 10.

  The discharge power source 9 discharges between two electrodes in the laser chamber 1 at a period corresponding to the oscillation instruction pulse signal input from the oscillator 8, and a command charge voltage V input from the controller 10. The charging circuit temporarily charges at a voltage according to the voltage, and discharge is performed by the operation of a switching element such as a thyratron. When discharge is performed in the laser chamber 1 and the laser gas is excited, laser oscillation is performed by an optical resonator (not shown), and laser light is output. Discharge is performed as pulse oscillation.

  The output monitor 11 detects the energy E (hereinafter referred to as laser output) of the output laser light and also detects that the laser light has been output, and the detection value E is input to the controller 10.

  The fluorine concentration monitor 12 detects the fluorine concentration F in the laser chamber 1, and the detection value F is input to the controller 10.

  The pressure monitor 13 detects the total pressure P of the gas in the laser chamber 1, and the detected value P is input to the controller 10.

  The controller 10 calculates a command charge voltage V based on the output E of the output monitor 11 as will be described later, and outputs the command charge voltage V to the discharge power source 9 to control the discharge voltage. In addition, the controller 10 controls the opening and closing of the valves 4 to 6 so that the gas injection amount in the laser chamber 1 becomes a predetermined amount before and during the laser operation and during gas exchange and gas injection. Further, the controller 10 calculates the fluorine partial pressure Fp from the fluorine concentration F and the gas total pressure P. The controller 10 monitors a power lock charging voltage (charging voltage when the laser output E reaches a specified value) Vp and the like, and has a storage table for storing these monitoring values.

  The operation of the first embodiment mainly performed by the controller 10 will be described below according to the flowcharts of FIGS.

  First, various parameter values listed below are set to predetermined values (step 100a).

・ Tt: Output correction lower limit time ... Lower limit time that can be corrected by stopping oscillation
Since the laser output decreases when the oscillation stop time is long, if the oscillation stop time exceeds the Tt time, a Kr / Ne mixed gas is injected to compensate for the laser output decrease.
-Pa: Total pressure at the end of injection ... Upper limit total pressure during gas injection
When the total pressure exceeds Pa, the injection of the Kr / Ne mixed gas is stopped.
Vc: charging voltage for injecting Kr / Ne mixed gas: threshold voltage for charging voltage for injecting Kr / Ne mixed gas
After the previous injection of Kr / Ne mixed gas, when the laser oscillation time exceeds Tg from when the charging voltage becomes Vc or higher, a predetermined amount of Kr / Ne mixed gas is injected.
・ Tg: Kr / Ne mixed gas injection interval time
After the previous injection of Kr / Ne mixed gas, when the laser oscillation time exceeds Tg from when the charging voltage becomes Vc or higher, a predetermined amount of Kr / Ne mixed gas is injected.
・ Fc: Target fluorine partial pressure
・ ΔFc: Allowable width of the target fluorine partial pressure (± ΔFc)
・ Ec: Rated laser output
ΔVc: Minimum increase / decrease amount of charging voltage
・ Va: Charging voltage upper limit
・ Sc: Number of target oscillation shots of the charged gas
・ Dc: Target filling time of 1 filling gas
・ Gn: Kr / Ne mixed gas injection amount

Each variable listed below is not initially set in step 100a, but is used in the processing described later. The definition of each variable is shown here.

・ ΔVm: Allowable width of charging voltage (± ΔVm)
・ St: Number of shots of injection interval of Kr / Ne mixed gas
・ Sn: Number of F2 injection interval shots
・ Λc: Target spectral line width
・ Δλc: Allowable width of target spectral line width (± Δλc)
・ Dt: Elapsed time since last gas exchange
・ St: Number of oscillation shots of the current filling gas
・ Vp: Power lock charging voltage
・ Sn: Number of shots since the previous Kr / Ne mixed gas injection
・ Fp: F2 partial pressure
・ ΔFp ： Fp−Fc
・ Pb: Threshold for total gas pressure for maintenance request
When the setting of the various parameter values is completed, the procedure proceeds to the power correction subroutine 1 shown in FIG. 3 (step 101a).

  This power correction subroutine 1 is a routine for compensating for a decrease in laser output due to gas exchange and long-time stoppage.

  The controller 10 first checks the presence or absence of a gas replacement request signal (step 120). If there is a replacement request, the controller 10 reads the first power lock voltage Vp of the current filling gas from the storage table (step 129), and this power lock. The charging voltage Vp is compared with the Kr / Ne mixed gas injection charging voltage Vc (step 130). As a result of this comparison, if Vp ≦ Vc, gas exchange is performed with the same gas composition as in the previous gas exchange (step 133). However, when Vp> Vc, gas exchange is performed with a gas composition in which the filling pressure of the Kr / Ne mixed gas is changed according to ΔV (= Vp−Vc) (steps 132 and 133). That is, as the value of ΔV increases, the gas exchange is performed by increasing the filling pressure of the Kr / Ne mixed gas.

  That is, according to the above control, when Vp at the time of the previous laser oscillation is higher than Vc, Kr / Ne mixed gas is added by an amount corresponding to ΔV to perform gas exchange. In, a predetermined laser output can be obtained with a low power lock charging voltage.

  Further, when there is no request for gas replacement in the determination of step 120, the controller 10 first calculates the oscillation stop time T. Then, this oscillation stop time T is compared with the output correction lower limit time Tt (step 123). If T ≦ Tt, the process returns to the original main routine. If T> Tt, Kr · Ne is determined according to the oscillation stop time T. A mixed gas injection amount Gn is calculated (step 124).

  Next, the controller 10 reads the total pressure Pp before the oscillation is stopped from the storage table (step 125), and the total pressure Pt after the injection is calculated from the calculated Kr / Ne mixed gas injection amount Gn and the total pressure Pp before the oscillation is stopped. (Step 126) and the calculated value Pt is compared with the total pressure Pa at the end of injection. If the calculated value Pt exceeds the injection end total pressure Pa, the process returns to the main routine without doing anything. If the calculated value Pt does not exceed the injection end total pressure Pa, only KnNe gas is injected. (Steps 127 and 128).

  That is, according to the above control, an amount of the Kr / Ne mixed gas corresponding to the oscillation stop time T is injected before the laser oscillation, so that the laser output is not lowered due to the natural deterioration of the gas during the oscillation stop. Can be prevented.

  The gas exchange performed in step 133 and the Kr / Ne mixed gas injection performed in step 128 are performed by controlling the time during which the valve 4 or 5 is opened to inject a desired amount of gas. At the time of this gas injection, the total pressure P in the laser chamber 1 is monitored by the pressure monitor 13, and the opening / closing of the valve 4 or 5 is controlled based on the detected value P so that the pressure in the laser chamber 1 becomes a predetermined value. Is done. Further, gas exhaust at the time of gas exchange is executed by starting the vacuum pump 7 and opening the valve 6.

  When the power correction subroutine 1 is completed as described above, laser oscillation is started in the main routine (step 102 in FIG. 2). During this laser oscillation, the laser output E, the charging voltage V, the total pressure P, and the fluorine concentration F detected by the output monitor 11, the discharge power supply 9, the pressure monitor 13, and the fluorine concentration monitor 12 are input to the controller 10 ( Step 103).

  Next, the procedure proceeds to the Kr / Ne mixed gas injection subroutine 1 shown in FIG. 4 (step 104a). This Kr / Ne mixed gas injection subroutine 1 is a routine for compensating for an output that decreases as the laser oscillation time elapses.

  First, the charging voltage V is compared with the Kr / Ne mixed gas injection charging voltage Vc (step 140), and if V> Vc, the time when the laser oscillates with the charging voltage Vc or higher after the previous Kr / Ne mixed gas injection. Tv is calculated (step 141). Then, the time Tv is compared with the Kr / Ne mixed gas injection interval time Tg (step 142). If Tv> Tg, the total pressure P and the injection end total pressure Pa are then compared (step 143). As a result of the comparison, if P <Pa, a fixed amount of Kr / Ne mixed gas is injected (step 144), and the process returns to the main routine. If V ≦ Vc in the comparison in step 140, Tv ≦ Tg in the comparison in step 142, or P ≧ Pa in the comparison in step 143, the process returns to the main routine without doing anything.

  In step 144, a certain amount of Kr / Ne mixed gas is injected by opening the valve 5 for a certain period of time.

  As described above, in the Kr / Ne mixed gas injection subroutine, when the charging voltage V exceeds the predetermined threshold value Vc, the charging voltage is equal to or higher than the threshold value for a predetermined time interval (Tg). Since a predetermined amount of the Kr / Ne mixed gas is injected, accurate and high-precision laser output constant control by rare gas and buffer gas injection can be achieved even in a constant amount of gas supply control.

  When the process of the Kr / Ne mixed gas injection subroutine 1 is completed as described above, the fluorine partial pressure Fp is calculated from the F2 concentration F and the total pressure P (step 105 in FIG. 1). The difference ΔFp from the F2 partial pressure Fc is calculated (step 106), and this difference ΔFp is compared with the allowable width ΔFc of the target F2 partial pressure (step 107). If | ΔFp | ≧ ΔFc, the F2 injection shown in FIG. The process proceeds to subroutine 1 (step 108a). If | ΔFp | <ΔFc, the process proceeds to the charge voltage command subroutine shown in FIG. 6R> 6 (step 109).

  In the F2 injection subroutine 1 of FIG. 5, the total pressure P is compared with the injection end total pressure Pa (step 150). If P ≧ Pa, the process returns to the main routine, and if P <Pa, the previously calculated ΔFp is set. Accordingly, the injection amount Gt is calculated (step 151), the F2 gas is injected by the calculated injection amount Gt, and the process returns to the main routine (step 152).

  Thus, in the F2 gas injection subroutine 1, the F2 gas partial pressure Fp is detected, and the replenishment amount of the halogen gas is determined in accordance with the detected partial pressure value Fp. Halogen gas can be supplied accurately by the reduced amount.

  In this case, the F2 gas is injected by an amount corresponding to ΔFp, but a predetermined amount of F2 gas may always be injected regardless of ΔFp. In this embodiment, the F2 partial pressure is controlled to a constant value, but the F2 concentration may be controlled to a constant value.

  As described above, when the F2 injection subroutine 1 is completed, the procedure is shifted to the charging voltage command subroutine.

  The charging voltage command subroutine shown in FIG. 6 is a routine for setting the charging voltage V corresponding to the laser output E. In this charging voltage command subroutine, the detected laser output E is compared with the rated laser output Ec (step 160), and if E <Ec, an increase / decrease amount ΔVc is set to the charging voltage V to increase the laser output to the rated output. In addition (step 161), the addition result (V + ΔVc) is output to the discharge power source 3 as the command charge voltage V (step 163). If the comparison result in step 160 is E = Ec, the detected charging voltage V is output to the discharge power source 3 as a command charging voltage (step 163). Further, if the comparison result of the step 160 is E> Ec, the increase / decrease amount ΔVc is subtracted from the charging voltage V to reduce the laser output to the rated output (step 161), and the subtraction result (V−ΔVc) is commanded. The charge voltage V is output to the discharge power supply 3 (step 163).

  When the charging voltage command subroutine is completed as described above, the charging voltage V is compared with the charging voltage upper limit value Va (step 110). If this result V ≦ Va, the procedure is again transferred to step 103 and the same as described above. If the comparison result is V> Va, the laser oscillation is stopped and the routine proceeds to a gas exchange request signal output subroutine shown in FIG.

  This gas exchange request signal output subroutine is a subroutine for outputting a gas exchange request signal, a window exchange request signal of the laser chamber 1, and a maintenance request signal.

  In the gas exchange request signal output subroutine shown in FIG. 7, a gas exchange request signal is first output from the controller 10 (step 170), thereby warning the operator that the command charge voltage V is outside the allowable range. At the same time, a display and a warning sound for prompting gas exchange are generated.

  Next, the controller 10 calculates the number of oscillation shots St and the gas filling time (days) Dt of the current filling gas (step 171). Then, the number of oscillation shots St is compared with the lower limit Sc of the number of oscillation shots of the filling gas (step 172). If St ≧ Sc, the process returns to the main routine, and if St <Sc, the previously obtained gas filling is performed. The time Dt is compared with the lower limit value of the filling time of the filling gas (step 173). If Dt ≧ Dc as a result of the comparison, the process returns to the main routine. If Dt <Dc, the total gas pressure P is compared with the threshold value Pb (step 174). If P <Pb, the process returns to the main routine. If P ≧ Pb, whether or not the current gas exchange is the first gas exchange after the window exchange (in other words, the current gas is changed after the window exchange). It is confirmed whether or not the gas is initially filled (step 175). If it is not the first gas exchange, a window exchange request signal is output and a notification operation to that effect is performed to cause window exchange and return to the main routine (step 177). If it is the first gas exchange, a maintenance request signal is output, the operator is informed of this, and a maintenance check is performed before returning to the main routine (step 176).

  As described above, in the first embodiment, before the laser oscillation, the Kr / Ne mixed gas is injected according to the oscillation stop time T to compensate for the laser output decrease due to the long-time stop. After the laser oscillation, the Kr / Ne mixed gas is injected based on the charging voltage, the F2 gas is compensated based on the partial pressure value of the F2 gas, and the laser is further increased until the charging voltage rises to a predetermined upper limit value. Laser output compensation control is performed in which the charging voltage is increased or decreased according to the comparison between the output and the rated laser output.

-2nd Example Next, 2nd Example which is a modification of 1st Example is described with reference to the flowchart of FIGS. 8-10.

  In the second embodiment, the power correction subroutine 1 of the main routine of FIG. 2 of the first embodiment is changed to a power correction subroutine 2 (step 101b in FIG. 8), and the power is corrected as a power correction after laser oscillation. A correction subroutine 3 (step 101c) is added, and the others are the same as in the first embodiment. A duplicate description is omitted.

  The power correction subroutine 2 shown in FIG. 9 is obtained by omitting steps 121 to 128 of the power correction subroutine 1 of the first embodiment.

  That is, the controller 10 first checks the presence / absence of a gas replacement request signal (step 120 in FIG. 9), and if there is a replacement request, reads the first power lock voltage Vp of the current filling gas from the storage table (step 129). The power lock charging voltage Vp is compared with the Kr / Ne mixed gas injection charging voltage Vc (step 130). As a result of this comparison, if Vp ≦ Vc, gas exchange is performed with the same gas composition as in the previous gas exchange (step 133). However, if Vp> Vc, the gas exchange is performed by changing the filling pressure of the Kr / Ne mixed gas in accordance with ΔV (= Vp−Vc) (steps 132 and 133).

  Further, when there is no gas exchange request in step 120, the controller 10 returns to the main routine.

  The power correction subroutine 3 shown in FIG. 10 is a procedure performed after laser oscillation, and is intended to compensate for a decrease in laser output due to long-term laser stoppage. The first embodiment omitted in the power correction subroutine 2 described above. This is a substitute procedure of steps 121 to 128 of the power correction subroutine 1 of the example.

  In this power correction subroutine 3, the charging voltage V when the laser output E reaches the rated laser output Ec is detected (steps 180 to 182), and when V> Vc, Kr / Ne mixing is performed according to this charging voltage. A gas injection amount Gn is calculated (step 183), Kr / Ne mixed gas is injected by the calculated amount Gn, and then the process returns to the main routine.

  That is, in the second embodiment, even when the laser is stopped for a long time, laser oscillation is started without injecting Kr · Ne, and Kr · Ne mixture gas is injected at a stretch. Therefore, even if the charging voltage V is initially high due to the laser being stopped for a long time, a large amount of noble gas and buffer gas corresponding to the large charging voltage value can be injected all at once. Compared with the conventional method of injecting, the charging voltage value can be lowered at a stretch.

  In the second embodiment, the Kr / Ne mixed gas injection amount Gn may be calculated from the oscillation stop time T as in the first embodiment. In addition, if a Kr / Ne mixed gas of several Torr is injected at a time during laser oscillation, the fluctuation of the laser output becomes large. Therefore, when the Kr / Ne mixed gas injection amount Gn is 3 Torr or more, the injection is divided into several times. It is better to do so.

Third Embodiment A third embodiment of the present invention will be described with reference to FIGS.

  In the third embodiment, the Kr / Ne mixed gas injection subroutine 1 (step 104a in FIG. 2) of the first embodiment is replaced with the Kr / Ne mixed gas injection subroutine 2 (step 104b in FIG. 11) shown in FIG. To replace it. Further, the Kr / Ne mixed gas injection amount Gn is added to the parameter initially set in step 100b. The rest is exactly the same as in the first embodiment.

  That is, in the Kr / Ne mixed gas injection subroutine 2 of FIG. 12, the number of laser shots Sn after the Kr / Ne mixed gas injection is calculated (step 190), and the shot number Sn and the Kr / Ne mixed gas injection interval shot are calculated. The number St (fixed value) is compared (step 191). If Sn ≦ St, the process returns to the main routine. If Sn> St, the Kr / Ne mixture is injected by a predetermined amount Gn that is initially set. Returning to the main routine (step 192).

  That is, in this Kr / Ne mixed gas injection subroutine 2, the number of laser shots from the previous injection of the Kr / Ne mixed gas is calculated, and the mixed gas is injected when the number of shots exceeds a predetermined threshold. Therefore, the rare gas is injected based on the number of laser shots directly related to the laser oscillation speed, and the laser output constant control can be performed accurately and accurately.

Fourth Embodiment A fourth embodiment of the present invention will be described with reference to FIGS.

  In the fourth embodiment shown in FIG. 13, a mass flow controller (mass flow rate controller) 14 and an on-off valve 15 for supplying a Kr / Ne mixed gas are added to the configuration shown in FIG.

  That is, when a Kr / Ne mixed gas is injected several Torr at a time, the laser output varies. This variation becomes smaller as the amount of one injection is smaller. The injection by the on-off valve 4 is the minimum of about 1 Torr. However, if the mass flow controller 14 is used, a smaller amount of injection can be performed and fluctuations in laser output can be suppressed. This is the reason for using the mass flow controller 14.

  FIG. 14 shows the main routine of the fourth embodiment. In the fourth embodiment, the Kr / Ne mixed gas injection subroutine 1 of the first embodiment is replaced with a Kr / Ne mixed gas injection subroutine 3 (step 104c in FIG. 14) shown in FIG. Further, the allowable range ΔVm of the charging voltage is added to the parameter initially set in step 100c. Others are the same as the first embodiment.

  In the Kr / Ne mixed gas injection subroutine 3 of FIG. 15, a difference ΔV (= V−Vc) between the charging voltage V and the Kr / Ne mixed gas injection charging voltage Vc is obtained (step 200), and the difference ΔV is determined as the charging voltage. (Step 201). As a result of this comparison, if | ΔV |> ΔVm, the mass flow controller flow rate of the Kr / Ne mixed gas is changed according to ΔV (step 202), and the process returns to the main routine, and | ΔV | ≦ ΔVm. Return to the main routine.

Fifth Embodiment A fifth embodiment of the present invention will be described with reference to FIGS.

  In the fifth embodiment shown in FIG. 16, a mass flow controller 16 for supplying F2 gas and an on-off valve 17 are added to the configuration shown in FIG. 1 to prevent fluctuations in laser output due to fluorine gas injection. ing.

  That is, in the fifth embodiment, the amount of gas passing through the gas supply path is controlled so that the mass flow rate of the halogen gas supplied by the mass flow controller 16 becomes a desired constant value.

  FIG. 17 shows a main routine according to the fifth embodiment. The F2 gas injection subroutine 1 of the first embodiment is replaced with the F2 gas injection subroutine 2 (step 108b in FIG. 17) shown in FIG. I have to. The rest is exactly the same as in the first embodiment.

  In the F2 gas injection subroutine 2 shown in FIG. 18, the flow rate of fluorine gas supplied by the mass flow controller 16 is changed according to ΔFp (= Fp−Fc).

Sixth Embodiment In the sixth embodiment, neither the fluorine concentration monitor 12 in FIG. 1 nor the spectrum width monitor 20 employed in the later embodiments is provided. That is, in this sixth embodiment, the shortage of F2 is estimated by the number of laser shots after the fluorine gas is injected, and when it is determined that F2 is insufficient, fluorine is supplied by a predetermined amount. Yes.

  FIG. 20 shows the main routine in the sixth embodiment. In this embodiment, the fluorine injection processing procedure (step 105 to step 108a) in the first embodiment is replaced with the processing in steps 220 to 222. I have to. Further, in the initial setting of the parameter in step 100d, the injection amount Gt of F2 gas is set. The rest is exactly the same as in the first embodiment.

  That is, when the Kr / Ne mixed gas injection subroutine 1 is completed (step 104a), the number of shots St since the previous fluorine gas was injected is calculated (step 220), and the calculated value St and the F2 injection interval shot number Sh are calculated. Compare (step 221). If St ≧ Sh, the F2 gas is injected by a predetermined amount Gt set in advance (step 222), and if St <Sh, the next step is executed.

Seventh Example A seventh example will be described with reference to FIGS.

  In the seventh embodiment, a spectral line width monitor 20 is installed instead of the fluorine concentration monitor, and such an embodiment is applied to a narrow-band laser.

  That is, in the narrow-band laser, the spectral line width needs to be equal to or less than a predetermined value, but the laser oscillation efficiency decreases as the spectral line width is reduced. For this reason, it is effective to perform laser oscillation by controlling the spectral line width to be close to the predetermined value.

  Here, the present inventors pay attention to the fact that the spectral line width λ has a large correlation with the fluorine partial pressure Fp, and determines the halogen gas injection amount according to the spectral line width λ that is proportional to the number of halogen gas molecules. I am doing so. That is, when the spectral line width is controlled to a predetermined target value, the spectral line width becomes narrower as the halogen gas is consumed. In this case, the halogen gas supply amount is increased so that the spectral line width becomes thicker. Let Further, when the spectral line width is increased, the halogen gas supply amount is decreased so as to narrow the spectral line width.

  FIG. 22 shows the experimental results showing the relationship between the F2 concentration and the spectral line width, and FIG. 23 shows the experimental results showing the relationship between the F2 partial pressure and the spectral line width. From these graphs, there is no correlation between the F2 concentration and the spectral line width shown in FIG. 22, but there is a clear proportional relationship between the F2 partial pressure and the spectral line width shown in FIG. Appears.

  Therefore, the present inventors have found from these experimental results that “the line width is proportional to the number of F2 molecules (proportional to the F2 partial pressure). However, the% concentration in the mixed gas also changes depending on the increase or decrease of other gases. It does not accurately reflect the molecular or spectral linewidth ".

  This can be proved from the gas equation of state.

  Now, the partial pressure of each gas in the mixed gas is Pi (i = ˜n), the number of molecules of each gas in the mixed gas is Ni (i = ˜n), the volume in the laser chamber is v, the gas constant is R, If the gas temperature is t, the following equation is established.

.SIGMA.Pi.v = .SIGMA.Ni.R.t (1)
Here, assuming that a certain gas is added to increase the number of molecules to Nj + δN, the following equation (2) can be obtained from the above equation (1).

(P1 + P2 + ... + Pj ... + Pi + δP) × v
= (N1 + N2 + ... + Nj ... + Ni + .delta.N) .times.R.times.t (2)
The following formula (3) can be obtained by subtracting the formula (1) from the formula (2).

δP × v = δN × R × t
Here, since v, R, and t are all constants, when attention is paid to each gas in the mixed gas, it can be seen that the increment of the number of molecules of the gas is proportional to the increment of the partial pressure. The amount of increase in line width can be predicted quantitatively.

  FIG. 24 shows the main routine in the seventh embodiment. In this embodiment, the fluorine injection processing procedure (step 105 to step 108a) of the first embodiment is performed in steps 230, 231 and 108c. Has been replaced. Further, in the initial setting of the parameters in step 100e, the target spectral line width λc and the allowable spectral line width Δλc are set. Others are exactly the same.

  That is, when laser oscillation is started (step 102), during this laser oscillation, the laser output E, the charging voltage V, the total voltage detected by the output monitor 11, the discharge power supply 9, the pressure monitor 13, and the spectral line width monitor 20, respectively. The pressure P and the spectral line width λ are input to the controller 10 (step 103 ′).

  Thereafter, when the Kr / Ne mixed gas injection subroutine 1 is completed (step 104), a difference Δλ between the detected spectral line width λ and the target spectral line width λc is calculated (step 230). It is compared with the allowable width Δλc (step 231). In this comparison, if | Δλ | ≧ Δλc, the process proceeds to the F2 injection subroutine 3 shown in FIG. 25 (step 108c). If | Δλ | <Δλc, the process proceeds to the next step and the charge voltage command subroutine is executed. .

  In the F2 injection subroutine 3 of FIG. 25, the total pressure P is compared with the injection end total pressure Pa (step 240), and if P <Pa, the F2 gas injection amount is calculated based on the previously calculated Δλ (step 241). ) F2 gas is injected by the calculated amount (step 242). That is, Δλ and the injection amount Gt are substantially in inverse proportion, and when the line width λ is smaller than the target value, the supply amount of the halogen gas is increased so that the spectral line width becomes thick, and the spectral line width becomes thicker than the target value. In this case, the supply amount of halogen gas is decreased so as to narrow the spectral line width.

  If P ≧ Pa in the comparison in the previous step 240, the process returns to the main routine.

  The following processing is the same as in the first embodiment.

Eighth Example An eighth example will be described with reference to FIGS.

  In the eighth embodiment, as shown in FIG. 26, the spectral line width monitor 20 is installed and the mass flow controller 16 is used for supplying F2 gas, as in the previous seventh embodiment. .

  FIG. 27 shows a main routine in the eighth embodiment. In the eighth embodiment, after laser oscillation starts (step 102), F2 gas injection control is performed by comparing line widths (steps 229 to 231). 108c) After the Kr / Ne mixed gas injection subroutine 1 (step 104a), the injection control of the F2 gas by the line width comparison is performed again. Others are the same as the seventh embodiment.

  That is, laser oscillation is started (step 102), the spectral line width λ is detected (step 229), and the difference Δλ between the detected spectral line width λ and the target spectral line width λc is calculated (step 230). The difference Δλ is compared with the line width allowable width Δλc (step 231). In this comparison, if | Δλ | ≧ Δλc, the process proceeds to the F2 injection subroutine 3 shown in FIG. 25 (step 108c), and if | Δλ | <Δλc, the process proceeds to the next step 103 ′.

  In the F2 injection subroutine 3 of FIG. 25, the total pressure P is compared with the injection end total pressure Pa (step 240), and if P <Pa, the F2 gas injection amount is calculated based on the previously calculated Δλ (step 241). ) F2 gas is injected by the calculated amount (step 242).

  When the Kr / Ne mixed gas injection subroutine 1 is completed (step 104a), the difference Δλ between the detected spectral line width λ and the target spectral line width λc is calculated again (step 250). The width is compared with an allowable width Δλc (step 251). In this comparison, if | Δλ | ≧ Δλc, the process proceeds to the F2 injection subroutine 4 shown in FIG. 28R> 8 (step 108d). If | Δλ | <Δλc, the process proceeds to the next step and the charge voltage command subroutine is executed. Execute (step 109).

  In the F2 injection subroutine 4 of FIG. 28, the flow rate of F2 gas passing through the mass flow controller 16 is changed according to Δλ (step 260).

  In the present invention, He or a mixed gas of Ne and He may be used as the buffer gas, HCl may be used as the halogen gas, and Xe or Ar may be used as the rare gas.

  In the embodiment, the Kr / Ne mixed gas is filled in one cylinder. However, these are filled in different cylinders and controlled so as to have a predetermined mixing ratio inside the laser chamber. Also good.

  As described above, according to the present invention, the rare gas and the buffer gas are replenished by an amount corresponding to the oscillation stop time before the laser oscillation. Thus, laser oscillation can be performed after returning to the proper state, whereby a stable laser output can be obtained from the initial stage of laser oscillation.

  Further, according to the present invention, the charging voltage is controlled so that the desired laser output is obtained immediately after the laser oscillation, the charging voltage when the desired laser output is obtained is detected, and only the amount calculated according to the detected charging voltage is obtained. Since the rare gas or the buffer gas is injected, a large amount of the rare gas or the buffer gas corresponding to the large charge voltage value can be injected all at once even in a state where the charge voltage is high from the beginning. Compared with the conventional method of gas injection, the charging voltage value can be lowered at once.

  Further, according to the present invention, the halogen gas partial pressure in the laser chamber is detected, and the replenishment amount of the halogen gas is determined according to the detected partial pressure value. Halogen gas can be supplied.

  According to the invention, the oscillation spectral line width proportional to the halogen gas partial pressure is detected, and the halogen gas injection amount is determined according to the difference between the detected line width value and the target spectral line width. Therefore, an accurate amount of halogen gas can be supplied while maintaining the target spectral line width.

  Further, according to the present invention, when the gas exchange request signal is output before the laser oscillation, the gas composition relating to the rare gas and the buffer gas is changed according to the first power lock voltage in the current filling gas. Therefore, for example, when the initial power lock voltage is high, gas exchange is performed with a gas composition in which a rare gas and a buffer gas are added to the previous gas composition. Laser output can be obtained.

  Further, according to the present invention, when the charging voltage exceeds a predetermined threshold after laser oscillation, the time when the charging voltage oscillates above the threshold after injecting the previous mixed gas of rare gas and buffer gas is predetermined. Since a predetermined amount of the mixed gas is injected when the time interval is exceeded, accurate and highly accurate laser output constant control by rare gas and buffer gas injection can be performed even in a constant amount of gas supply control.

  Further, according to the present invention, the mixed gas of the rare gas and the buffer gas is injected by the mass flow controller, the charging voltage for power-locking after the laser oscillation is detected, and the detected charging voltage and the threshold of the charging voltage are detected. Since the injection flow rate of the mixed gas is changed in accordance with the difference between them, accurate and highly accurate laser output constant control by rare gas and buffer gas injection can be achieved.

  According to the present invention, the number of laser shots since the previous injection of the mixed gas of the rare gas and the buffer gas is calculated, and the mixed gas is injected when the number of shots exceeds a predetermined threshold value. The rare gas is injected based on the number of laser shots directly related to the laser oscillation speed, and the laser output constant control can be performed accurately and accurately.

  Further, according to the present invention, the window replacement signal or the maintenance request signal is automatically generated according to the number of oscillation shots of the current filling gas, the filling time of the current filling gas, the gas pressure in the laser chamber, or the number of gas filling after the window replacement of the laser chamber. Since the output is performed automatically, the operator can recognize an appropriate time for maintenance and window replacement.

The block diagram used for the 1st, 2nd and 3rd Example of this invention. The main routine flowchart which shows the whole operation | movement of 1st Example. 5 is a flowchart showing a power correction subroutine 1; 7 is a flowchart showing a Kr / Ne mixed gas injection subroutine 1; The flowchart which shows F2 injection | pouring subroutine 1. FIG. The flowchart which shows a charging voltage command subroutine. The flowchart which shows a gas exchange request signal output subroutine. The main routine flowchart which shows 2nd Example of this invention. 7 is a flowchart showing a power correction subroutine 2; 7 is a flowchart showing a power correction subroutine 3; The main routine flowchart which shows 3rd Example of this invention. 9 is a flowchart showing a Kr / Ne mixed gas injection subroutine 2; The block diagram which shows 4th Example of this invention. The main routine flowchart which shows 4th Example of this invention. 10 is a flowchart showing a Kr / Ne mixed gas injection subroutine 3; The block diagram which shows 5th Example of this invention. The main routine flowchart which shows 5th Example of this invention. The flowchart which shows F2 injection | pouring subroutine 2. FIG. The block diagram which shows 6th Example of this invention. The main routine flowchart which shows 6th Example of this invention. The block diagram which shows 7th Example of this invention. The graph which shows the experimental result of the relationship between F2 density | concentration and a spectral line width. The graph which shows the experimental result of the relationship between F2 partial pressure and a spectrum line width. The main routine flowchart which shows 7th Example of this invention. 7 is a flowchart showing an F2 injection subroutine 3; The block diagram which shows 8th Example of this invention. The main routine flowchart which shows 8th Example of this invention. 10 is a flowchart showing an F2 injection subroutine 4; The figure which shows the control structure of the conventional gas supply. The flowchart which shows the conventional gas supply control. The graph which shows the various state quantities by the conventional gas supply control.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Laser chamber 2 ... F2 cylinder 3 ... Kr * Ne cylinder 4 ... On-off valve 5 ... On-off valve 6 ... On-off valve 7 ... Vacuum pump 8 ... Oscillator 9 ... Discharge power supply 10 ... Controller 11 ... Output monitor 12 ... Fluorine concentration Monitor 13 ... Pressure monitor 14 ... Mass flow controller 16 ... Mass flow controller 20 ... Spectral line width monitor

Claims (5)

In an excimer laser device that performs laser oscillation by injecting a halogen gas, a rare gas, or a buffer gas into a laser chamber,
Before the laser oscillation, the oscillation stop time is calculated, and when the calculated oscillation stop time exceeds a predetermined time, the injection amount of the mixed gas of the rare gas and the buffer gas is calculated based on the calculated oscillation stop time, A gas replenishing method for an excimer laser device, wherein the mixed gas is injected before the laser oscillation by the calculated injection amount.   The gas replenishment of the excimer laser apparatus according to claim 1, wherein the total pressure after the mixed gas injection is predicted and calculated, and the mixed gas is injected by the calculated injection amount only when the calculated total pressure is less than a predetermined value. Method. In an excimer laser device that performs laser oscillation by injecting a halogen gas, a rare gas, or a buffer gas into a laser chamber,
Before the laser oscillation, the mixed gas of the rare gas and the buffer gas is not injected. After the laser oscillation, when the laser output during the laser oscillation exceeds the rated laser output, the charging voltage is detected, and the detected charging voltage is set to the detected charging voltage. A gas replenishing method for an excimer laser device, wherein a mixed gas of a rare gas and a buffer gas is injected in an amount calculated accordingly. In an excimer laser device that performs laser oscillation by injecting a halogen gas, a rare gas, or a buffer gas into a laser chamber,
When the gas exchange request signal is output before the laser oscillation, the first power lock voltage in the current filling gas is detected, and the detected value is compared with a predetermined threshold value. Gas exchange is performed with the same gas composition, and when the detection value is larger than the threshold value, the injection amount of the mixed gas of the rare gas and the buffer gas is changed according to the deviation between the detection value and the threshold value. A gas replenishing method for an excimer laser device in which gas exchange is performed with a gas composition.   The excimer laser device gas supply method according to claim 4, wherein the gas exchange request signal is output when a charging voltage exceeds an upper limit value of the charging voltage.

Images