JP2002116088A - Wavelength detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a wavelength detector for excimer lasers which can obtain a stable reference light and can highly accurately measure an absolute wavelength of the reference light and a light to be detected. SOLUTION: A waveform of light interference fringes of the reference light emitted from a low-pressure mercury lamp is obtained according to a waveform of light interference fringes of each mercury isotope sealed in the low-pressure mercury lamp and a mixture ratio. The absolute wavelength of the reference light is obtained of the basis of two points corresponding to a specific intensity of the wavelength of one specific light interference fringe among the waveforms of the light interference fringes of the reference light.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wavelength detector for a narrow band excimer laser used as a light source for a semiconductor exposure apparatus.

[0002]

2. Description of the Related Art Conventionally, a monitor etalon has been used to detect the wavelength line width and wavelength of a narrow band excimer laser. The monitor etalon is configured by using an air gap etalon in which partial reflection mirrors are arranged facing each other at a predetermined interval, and the transmission wavelength λ of this air gap etalon is expressed as follows.

Mλ = 2nd · cos θ (1) m: integer d: interval between etalon and partial reflection mirror n: refractive index of partial reflection mirror θ: angle formed by normal of etalon and incident light, n, Assuming that d and m are constant, when the wavelength λ changes, θ changes. The monitor etalon utilizes this property to detect the wavelength of the light to be detected.

In the above-described monitor etalon, if the pressure in the air gap and the ambient temperature change, the angle θ changes even if the wavelength is constant.
Therefore, there is a case where the wavelength λ cannot be accurately detected based on the angle θ. Therefore, conventionally, when a monitor etalon is used, wavelength detection is performed by controlling the pressure in the air gap, the ambient temperature, and the like to be constant.

However, it is difficult to control the pressure and the ambient temperature in the air gap with high accuracy, and it has not been possible to detect the absolute wavelength with satisfactory accuracy.

Therefore, together with the light to be detected (output laser light), reference light having a known wavelength (for example, the emission line of a mercury lamp) is incident on the monitor etalon, and the relative wavelength of the light to be detected with respect to this reference light is calculated. An apparatus has been proposed in which the absolute wavelength of the light to be detected is detected with high accuracy by detection (for example, JP-A-1-101683 and JP-A-4-35697).

In such an apparatus, the transmitted light of the monitor etalon is directly incident on a photodetector such as a CCD image sensor to form an interference fringe on a detection surface of the photodetector, and based on the position of the interference fringe. The absolute wavelength is detected.

[0008]

However, even in these prior arts, if the spectrum waveform of the reference light source is distorted or changes due to a change in time or the temperature of the reference light source (see 253. Mercury lamp). The 7 nm line undergoes self-absorption and the spectrum waveform changes significantly), and the light intensity distribution of interference fringes generated by the monitor etalon is distorted. Therefore, in such a case, the wavelength of the reference light changes, and the absolute wavelength of the light to be detected cannot be specified with high accuracy.

Therefore, there has been a demand for a method for obtaining a reference light having a stable wavelength change and a method for measuring the absolute wavelength of the reference light with high accuracy.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and a wavelength detection of an excimer laser capable of obtaining a stable reference light and measuring the absolute wavelengths of the reference light and the light to be detected with high accuracy. It is intended to provide a device.

[0011]

According to the present invention, a low-pressure mercury lamp is used as a reference light source, and reference light and detected light generated from the reference light source are incident on a wavelength detector, and the wavelength detector is used as a reference light source. In a wavelength detection device that detects an absolute wavelength of light to be detected based on a detection output, a storage unit that stores a waveform of an optical interference fringe of each mercury isotope enclosed in the low-pressure mercury lamp, and a storage unit that stores each of the stored mercury. A waveform synthesizing unit for synthesizing the waveform of the optical interference fringe of the isotope according to the mixing ratio of each mercury isotope to obtain the waveform of the optical interference fringe of the reference light emitted from the low-pressure mercury lamp; The absolute wavelength of the reference light is obtained based on the positions of two points corresponding to the specific intensity of the waveform of the specific one optical interference fringe of the fringe waveform, and the detected light is determined based on the obtained absolute wavelength of the reference light. The absolute wavelength of So that comprise the calculated wavelength calculation means.

According to the invention, the waveform of the optical interference fringe of each mercury isotope enclosed in the low-pressure mercury lamp and the waveform of the optical interference fringe of the reference light emitted from the low-pressure mercury lamp are determined according to the mixing ratio thereof. The absolute wavelength of the reference light is determined based on the positions of two points corresponding to the specific intensity of the waveform of one specific optical interference fringe among the waveforms of the optical interference fringes of the reference light.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the embodiments shown in the accompanying drawings.

FIG. 1 shows an example of a wavelength control device for an excimer laser.

A part of the laser light La of the narrow band oscillation excimer laser 1 whose band has been narrowed is reflected by the beam splitter 2 and is incident on the shutter 3 as sampling light. The light is then incident on the beam splitter 6 of the wavelength detector 5 via the optical disc.

On the other hand, the low-pressure mercury lamp 7 for generating the reference light Lb is disposed at a position substantially coincident with the focal position of the condenser lens 8, thereby increasing the amount of interference fringes due to the reference light. Reference light Lb generated by low-pressure mercury lamp 7
Is the band pass filter 10 when the shutter 9 is open.
, And a wavelength of 253.7 n
Only the m-line is selected. The selected reference light passes through the condenser lens 8 and enters the beam splitter 6.

The wavelength detector 5 comprises a beam splitter 6, a monitor etalon 11, a condenser lens 12, and an optical position detector 13.

The monitor etalon 11 has an inner surface formed of two transparent plates 11a and 11b of a partial reflection mirror,
It acts to vary the transmission wavelength according to the angle of the incident light with respect to the etalon.

The light transmitted through the etalon 11 is collected by the condenser lens 1.
2 is incident. The condenser lens 12 is, for example, an achromatic lens that has been subjected to chromatic aberration correction for both the wavelength of the light to be detected and the reference light, and both interference fringes coincide with the focal position of the achromatic range. Occur. Note that a concave mirror may be used as the condenser lens 12.

The light position detector 13 is, for example, a one-dimensional or two-dimensional image sensor disposed on the focal plane of the condenser lens 12, and the light passing through the condenser lens 12 is connected to the light position detector 13. The first interference fringes 13a corresponding to the wavelength of the detected light La and the second interference fringes 13b corresponding to the wavelength of the reference light Lb are formed on the detection surface of the optical position detector 13. Signals corresponding to the positions of the first and second interference fringes are output from the optical position detector 13 to the wavelength detection controller 20.
Is output to

The wavelength detection controller 20 calculates the relative wavelength of the detected light La with respect to the reference light Lb from the interference fringe position and the light intensity signal. Then, the absolute wavelength of the detected light is calculated based on the relative wavelength and the absolute wavelength of the reference light. Further, the wavelength detection controller 20 calculates a deviation between the set wavelength and the detected absolute wavelength,
The band narrowing element driver 25 is controlled so as to change the selected wavelength of the wavelength selecting element by the deviation.

In this case, a temperature sensor 14 and a cooling fan 15 are attached to the low-pressure mercury lamp 7, and the temperature of the low-pressure mercury lamp 7 is controlled by a temperature controller 30 so as to maintain a predetermined temperature. I have.

In such a configuration, first, the isotopic purity of the low-pressure mercury lamp 7 used as a reference light source will be considered. In the low-pressure mercury lamp 7, a 253.7 nm line is used as reference light.

FIG. 2 shows 253. corresponding to the isotope of mercury.
The absolute wavelength of the 7 nm line and its relative light intensity are shown. The abundance ratio of each isotope of natural mercury Hgnat is 10% (2
01Hg), 13% (201Hg), 30% (202Hg), 7
% (204 Hg). Therefore, the 253.7 nm line of the natural mercury Hgnat has five emission lines overlapping each other. Note that 201Hg indicates mercury having a mass number of 201.

FIG. 3 shows the case where only 202 Hg is sealed in the low-pressure mercury lamp 7 (a) and the case where natural mercury Hgnat is sealed.
(b) shows the interference fringes. In this case, the monitor etalon 11 having a free spectral range (FSR) of 20 pm is used.

Comparing these interference fringes, it is difficult to specify the absolute wavelength of the reference light in the case of natural mercury Hgnat shown in FIG. 3B because the width of the interference fringes is wide and about 6 pm. On the other hand, in the case of the isotope of 202Hg, the shape of the interference fringe is very beautiful, the measured line width is very narrow, about 1.5 pm, and the absolute wavelength of the reference light is 253.652 from FIG.
It can be specified to be 77 nm (253.6063 + 0.00214).

FIG. 4 shows the case where only 202 Hg is sealed in the low-pressure mercury lamp 7 (a) and the case where natural mercury Hgnat is sealed.
(b) shows the interference fringes. In this case, the monitor etalon 11 having a free spectral range (FSR) of 5 pm is used.

Comparing these interference fringes, it is difficult to specify the absolute wavelength of the reference light in the case of natural mercury shown in FIG. 4B because the contrast of the interference fringes is very small. On the other hand, in the case of only the isotope of 202Hg shown in FIG. 4 (a), the shape of the interference fringe is very beautiful, the measured line width is very narrow, about 1.0 pm, and the absolute value of the reference light 202Hg The wavelength can be specified at 253.665277 nm from FIG.

Here, the resolution R of the etalon is expressed by the following equation.

R = FSR / F (2) F: Finesse of the etalon FSR: Free spectral range F is determined by the surface accuracy, the reflectance, and the parallelism of the etalon. Usually, the maximum is around the wavelength of 248 nm. It is about 30.

Therefore, in order to improve the resolution of the monitor etalon, it is necessary to reduce the FSR.
At this time, as shown in FIG. 4, it is better to enclose one type of isotope in the low-pressure mercury lamp 7.

FIGS. 5 (a) to 5 (d) show that the low-pressure mercury lamp 7 has a purity of 99%, 70%, 61% and 30% of 202Hg of isotope mercury.
Are shown. As the purity P increases, the contrast of interference fringes improves. P = 61
%, The contrast is considerably reduced.
Since a sharp peak due to g can be detected, the absolute wavelength of the reference light can be specified.

FIG. 6 shows the relationship between the purity P of the isotope and the difference between the absolute wavelength λHg ′ of the detected reference light (253.7 nm line) and the ideal absolute wavelength λHg. According to FIG. 6, P =
Above 49%, the difference is almost zero. Also P
At ≧ 49%, a sharp isotope peak can be detected as shown in FIG. Therefore, in the low-pressure mercury lamp 7, if the purity P of the specific isotope is set to 49% or more, the theoretical value can be used as it is without the detection wavelength of the reference light deviating from the theoretical value.

As described above, by encapsulating a specific (-type) isotope of mercury in a lamp at a specific purity or higher, a reference light with a beautiful spectral waveform and a narrow line width can be obtained. It can be used as a reference light even for a small high-resolution monitor etalon. For this reason,
The absolute wavelength of the light to be detected can be detected with extremely high accuracy.

Next, the temperature of the low-pressure mercury lamp will be considered.

FIG. 7 shows interference fringes when the temperature T on the glass surface of the low-pressure mercury lamp 7 is 20 ° C. and 60 ° C. (time has elapsed after lighting) at 202 Hg. Temperature T = 25
When ゜ C, the waveform is very beautiful, but T = 60
In the case of ゜ C, self-absorption occurs, and the center of the interference fringes is depressed. The waveform of such interference fringes depends on the temperature T with good reproducibility, and has a shape corresponding to each temperature T. In other words, if the temperature is constant, the shape of the interference fringe is constant.

FIG. 8 shows the surface temperature T of the low-pressure mercury lamp 7 and the absolute wavelength λHg of the detected reference light (253.7 nm line).
And the difference ΔλHg between the ideal absolute wavelength λHg. According to this figure, when self-absorption does not occur, that is, when the temperature of the low-pressure mercury lamp is 40 ° C. or less, the difference ΔλHg is almost zero. In the low-pressure mercury lamp 7, if the surface temperature is controlled to be 40 ° C. or less,
The theoretical value can be used as it is without causing the detection wavelength of the reference light to deviate from the theoretical value, whereby various correction calculations for the reference light wavelength can be omitted.

Next, the type of the low-pressure mercury lamp will be considered.

FIG. 9 shows the structure of the low-pressure mercury lamp 7.

FIG. 9A shows the case of a hot cathode type, in which a glass (Vycor glass, quartz glass, etc.) tube 31 that transmits a reference light of 253.7 nm is filled with a carrier gas such as argon. It is filled with high purity isotope mercury. In addition, in this glass tube 31,
A filament 32 coated with an electron-emitting oxide (Ba0 or SrO or the like) for emitting thermoelectrons is provided therein. When a current flows through the filament 32, the filament 32 is heated to emit thermoelectrons, and isotopes Mercury is excited to emit light. Ballast 33 is filament 3
This is for making the current flowing through 2 constant.

The hot cathode type isotope lamp has the following advantages.

1. Light is emitted at a small voltage (about 30 V), no special power is turned on. 2. Lighting is instantaneously performed when current is applied, and light emission is instantly stabilized. 3. Spectral waveform due to low light emission temperature Is clean and the line width (1pm or less) is narrow. 4. It is inexpensive. 5. There is no deterioration due to turning on / off.

FIG. 9 (b) shows the case of an electrodeless discharge tube, in which a glass (Vycor glass, quartz glass, etc.) tube 34 that transmits a reference light of 253.7 nm is filled with argon or the like as described above. It contains a carrier gas and one type of high purity isotope, mercury. The excitation of the lamp is performed by a high frequency coil 35 and a high frequency power supply 36.
In the case of this lamp, a large energy is first input for lighting, and after the light is turned on, the input energy is reduced to obtain a 253.7 nm line reference light with a narrow spectral line width. It should be noted that this electrodeless isotope lamp has the disadvantage that it takes time to light up and stabilize the discharge state, and requires a large power supply voltage for excitation.

FIG. 9C shows a case of a cold cathode type, in which a glass (Vycor glass, quartz glass, etc.) tube 37 transmitting a reference light of 253.7 nm is provided with a carrier such as argon as described above. Gas and high purity isotope mercury are enclosed. Two electrodes are sealed in the glass tube 37, and a high voltage (several KV) is applied by an AC power supply 39 and a step-up transformer 38 to apply a high voltage between the two electrodes to discharge. In this case, as in the case of electrodeless discharge, a certain time is required for lighting and stabilization. Also, in the case of this cold cathode type, by narrowing the input energy after lighting, it is possible to obtain reference light of a 253.7 nm line having a narrow spectral line width to some extent.

FIG. 10 shows an example of lighting control suitable for the electrodeless discharge or cold cathode type low-pressure mercury lamp 7 shown in FIGS. 9B and 9C. The power is turned on (step 101), and then the power is supplied with a predetermined input energy A (the power supply 36 in the case of FIG. 9B, the power supply 36 in FIG. 9C).
In this case, the power supply 39) is operated (step 102).
In the next step 103, it is confirmed whether or not the lamp 7 emits light. After the confirmation, the power-on energy is reduced to B (<A) (step 104). Finally, it is confirmed whether or not the reference light is stable (step 105), and the process proceeds to the next procedure. In step 105, it may be possible to wait for a predetermined time to elapse. This control may be applied to a hot cathode type.

That is, according to the reference light lighting control shown in FIG. 10, a large energy is first supplied for lighting, and after the light is turned on, the input energy is reduced to thereby reduce the 253.7 nm line having a narrow spectral line width. The reference light is obtained.

By the way, when a low-pressure mercury isotope lamp is used as a reference lamp and is a hot cathode type lamp, as described above, if a current is applied, lighting is performed instantaneously, and light emission is instantaneously stabilized. Further, as shown in FIG. 8, if the temperature of the lamp is 40 ° C. or less, the absolute wavelength of the reference light is constant. Therefore, by cooling the mercury lamp or detecting the reference light immediately after lighting, the absolute wavelength of the reference light can be obtained with high accuracy without detecting the temperature of the reference light lamp.

Therefore, in the configuration shown in FIG. 1, the temperature sensor 14, the cooling fan 15, the temperature controller 30
Is omitted, the reference light is turned on at a constant cycle, and immediately (0.1
(About 2 seconds) The interference fringes of the reference light may be detected and the reference light may be turned off. In this case, it is not necessary to detect the temperature of the lamp.

Further, only the cooling fan 15 is mounted, the temperature sensor 14 and the temperature controller 30 are omitted, and the process of turning on the fan, turning on the reference light, and detecting the interference fringe of the reference light is executed at a constant period. Good. In this control, only the fan 15 may be always on.

Further, the cooling fan 15 and the reference light 7 may be always turned on, and the interference fringes of the reference light may be periodically detected.

Next, a method for obtaining the absolute wavelength of the reference light and the absolute wavelength of the light to be detected will be considered.

FIG. 11 shows the principle of a Fabry-Perot interferometer (monitor etalon 11). As shown in FIG. 11A, light L is incident on the etalon 11 at a mirror distance d at an incident angle θ. When the light passes through the etalon 11 and the condenser lens 12, an interference fringe 13 c of the light L is formed on the detection surface of the photodetector 13 which is separated from the condenser lens 12 by a focal length f.

Here, the basic formula of the etalon is as described above: mλ = 2nd · cos θ (1) m: integer d: interval between the etalon and the partial reflection mirror n: refractive index of the partial reflection mirror θ: etalon method Where m is m0 and wavelength is λ0 when the angle θ = 0 in the above equation.
Then, 2nd = m0 · λ0 (3) Subtracting equation (1) from equation (3) above and applying the half-angle formula (cos θ = 1−2 sin ^ 2)) gives 2 sin ^ 2 (θ / 2) = (λ / 2nd) (m0−m ) (4) is obtained. Note that ^ 2 indicates a square.

When the angle θ is a relatively small angle, sin
(θ / 2) = θ / 2, which can be approximated by substituting this into the above equation (4) to obtain θ ^ 2 = (λ / nd) (m0−m) (5)

If the distance from the center of the interference fringes 13c is r, as shown in FIG.
Since the focal length of f is f, r = fθ = f (λ / nd) ^ 1/2 (m0−m) ^ 1/2 (6) Note that ^ 1/2 indicates a 1/2 power, that is, a route.

Here, when both sides of the equation (6) are squared with c = f ＝ 2 · (λ / nd), the following equation is obtained: r ^ 2 = c (m0−m) (7) (FIG. 11 (b) )reference).

Here, considering the p-th and p + 1-th peaks, c = c (mp + 1-mp) = rp ^ 2-rp + 1 ^ 2 from equation (7). In equation (3), when the integer m is differentiated with respect to the wavelength λ, −Δλ = (λ ^ 2 / 2nd) · Δm = FSR · Δm (9)

Here, when the equation (7) is substituted into the equation (9), the following equation is obtained: -Δλ = FSR ・ r 2 / c (10) Here, if the wavelength at m = m0 is λ0,
The required wavelength λ can be obtained from the equation (10) as λ = λ0−FSR · r ^ 2 · c (11)

Here, equation (11) indicates that the wavelength λ is the interference fringe 1
3c is proportional to the square of the radius r of 3c.
Therefore, if the square of the radius of the interference fringes can be determined correctly, the wavelength of the light to be detected can be determined accurately.

That is, in the basic formula (1) of the etalon, mHg is the order corresponding to the reference light, mex is the order corresponding to the light to be detected, λHg is the wavelength of the reference light, λex is the wavelength of the light to be detected, and nHg Is the refractive index of the reference light in the air gap of the etalon 11 and nex is the refractive index of the detected light in the air gap of the etalon 11. mHg · λHg = 2nHg · d · cos θ (12) mex · λex = 2nex D · cos θ (13)

Here, the wavelength λe when the interference fringe of the reference light and the interference fringe of the detected light coincide with each other is the above-mentioned (12).
Eliminating d · cos θ from the equation (13) and rearranging it, λe = (nex / nHg) · (mHg / mex) · λHg (14) Therefore, the radius of the interference fringe of the detected light is rex,
Assuming that the radius of the interference fringe of the reference light is rHg, the wavelength λex of the light to be detected is given by the following equation from the above equation (10): λex−λe = FSR · (rHg ^ 2-rex ^ 2) / c (15) You can ask. Here, since the reference light and the light to be detected are transmitted through the same etalon, even if the temperature of the etalon changes, the error due to the change is canceled out, and the absolute wavelength can be detected accurately. it can.

Here, the calculation of the square rm ^ 2 of the radius of the interference fringe is executed as shown in FIG.

That is, in the case of FIG. 12 (a), the maximum value Imax of the light intensity is detected, the half value Imax / 2 is calculated, and the two positions A and B corresponding to the half value Imax / 2 are respectively inside. The diameter 2r1 of the circle and the diameter 2r2 of the outer circle are determined. Then, the squares r1 ^ 2 and r2 ^ 2 of both radii are obtained from the respective diameters, and the squares rm ^ 2 of the radii of the interference fringes are obtained by obtaining the average values thereof as in the following equation (16).

Rm ^ 2 = (r1 ^ 2 + r2 ^ 2) / 2 (16) In the case of FIG. 12B, the radii r1 and r2 corresponding to 1 / a of the maximum value Imax are obtained. The square rm ^ 2 of the radius of the interference fringe is determined from the radius.

In the case of FIG. 12C, the maximum value Imax and the minimum value Imin of the light intensity are detected, and the average value Iav
(= (Imax + Imin) / 2), and the square rm ^ 2 of the radius of the interference fringe is determined from the radii r1 and r2 corresponding to the value.

The above methods (b) and (c) are effective in the case of a 253.7 nm line where the purity P of the isotope mercury is low.

FIG. 13 schematically shows an approximate calculation method when the interference fringes are detected by a one-dimensional or two-dimensional photodiode array sensor (such as a CCD) as the photodetector 13.

In the case of a photodiode array sensor,
Each light amount is calculated at the position of each channel. The position resolution is about 13 microns, which is a high resolution.
Also, since the dispersion per channel is 0.1 pm / ch, it is necessary to perform an approximate calculation to detect a change of 0.01 pm.

FIGS. 13 (a) and 13 (b) show an approximate curve superimposed on the signal waveform of each channel detected by the photodiode array sensor. The approximate curve connects the output between each channel with a straight line, Created using the least squares method or multi-order equation.

Then, the approximate curve and, for example, the half value Imax
By calculating the intersection with / 2, the inner and outer radii of the interference fringes are obtained, and the square of the radius of the interference fringes is calculated in the same manner as described above. By performing such an approximate calculation, 0.0
The absolute wavelength of the light to be detected can be detected with a resolution of 1 pm.

FIG. 14 shows an example of a method of calculating the square of the interference fringe radius by the above-described approximate calculation. In this case, first, an approximate straight line is obtained by connecting the light intensity values of the respective channels with straight lines.
Next, the maximum light intensity Imax is obtained from the light intensity of each channel, and the half value Imax / 2 is calculated. Next, the intersection between the above approximate straight line and the half value Imax / 2 is determined, and the outer radius r2 and the inner radius r1 of the interference fringe are calculated from the determined intersection.
Then, by calculating the average of these values, the square of the radius of the interference fringe is calculated as rm ^ 2 according to the above equation (16).

FIG. 15 shows another method of the above calculation method. In this case, an approximate curve is obtained by using a least square method or a multi-order equation instead of a linear approximation. The procedure after obtaining the approximate curve is the same as the method shown in FIG.

FIG. 16 shows that the maximum value Imax and the minimum value Imin of the light intensity are detected after linear approximation between the channels, the average value Iav (= (Imax + Imin) / 2) is obtained, and the average value Iav is obtained. And an intersection of the approximate straight line and the calculated intersection, calculate an outer radius r2 and an inner radius r1 of the interference fringe from the obtained intersection, and obtain an average of these values to obtain the square of the radius of the interference fringe as rm. ^ 2 is calculated according to the above equation (16).

In FIG. 17, the method of FIG. 16 is performed not by linear approximation but by using an approximate curve using a least squares method or a multi-order equation.

In each of the above methods, an approximate curve is calculated from the light intensity values of the respective channels, and the intersection between the approximate curve and a predetermined value such as a half value is determined. In a region where the relationship monotonically increases, the intersection is obtained by performing linear interpolation from the light intensities of both the channel that has become larger than the predetermined value and the channel immediately before the channel, thereby reducing the calculation time. You may.

Next, a method of obtaining the absolute wavelength of the reference light from the mixing ratio of the isotopes of mercury will be considered with reference to the flowchart of FIG.

First, the spectrum waveform in the case of each isotope alone is converted into data or a function (198 Hg (λ, T), 199 Hg
(Λ, T), 200Hg (λ, T), 201Hg (λ, T),
They are stored as 202Hg (λ, T) and 204Hg (λ, T)) (step 200). Next, the mixing ratio of each isotope (198K, 199K, 200K, 201K, 202K, 204K) and the lamp temperature T are input (step 210).

Then, a composite spectrum Hg (λ, T) is calculated using the spectrum waveform data and the mixing ratio according to the following equation (step 220).

Hg (λ, T) = 198K · 198Hg (λ, T) + 199K · 199Hg (λ, T) + 200K · 200Hg (λ, T) + 201K · 201Hg (λ, T) + 202K · 202Hg (λ, T) + 204K · 204Hg (λ, T) (17) FIG. 19 shows the spectrum waveforms of mercury lamps having various isotope mixing ratios, (a) showing a purity of 99% of 202Hg, and (b) showing 198.
Hg purity is 99%, (c) is 202Hg51% and 198Hg49.
% And mixed Hg.

In these spectral waveforms, (a) and (b)
When the spectrum waveforms of the above were superimposed, the superimposed waveform almost coincided with the waveform of the mixed Hg in (c). This indicates that the composite spectrum waveform of the mixed Hg can be obtained by performing the calculation of the equation (17).

Next, the absolute wavelength λHg ′ of the low-pressure mercury lamp is calculated from the composite spectrum waveform thus obtained (step 230).

This calculation method is as shown in FIGS.
There are various methods shown in (c).

In the case of the method shown in FIG. 12A, the absolute wavelengths λHg1 and λHg2 at two points in the half value of the combined spectrum are calculated, and the average value is set as the absolute wavelength λHg ′ (λHg ′ =
(λHg1 + λHg2) / 2).

In the case of FIG. 12B, 1 / a (1 / a)
The average value of the two wavelengths is defined as the absolute wavelength λHg ′.

In the case of FIG. 12C, the synthesized spectrum Hg
The interference fringes when (λ, T) is generated by the monitor etalon 11 are theoretically calculated, and the maximum value Ima of the interference fringes is calculated.
x and the minimum value Imin are determined, and the Iav is determined from these values. Further, the Iav and the synthesized spectrum waveform Hg
The wavelengths at two points, which are the intersections with (λ, T), are determined, and the average value of them is defined as the absolute wavelength λHg ′.

The obtained absolute wavelength λHg ′ and (1)
The absolute wavelength of the light to be detected is determined using the expressions 4) and (15) (step 240).

Next, the difference Δλ between the detected absolute wavelength λHg ′ of the reference light and the theoretical absolute wavelength λHg of the reference light shown in FIG.
A method of calculating the detected absolute wavelength λHg ′ of the reference light from the relationship between Hg and the temperature T will be considered with reference to FIG.

That is, in the relationship shown in FIG. 8, since the theoretical absolute wavelength λHg is known (253.65277 nm),
The graph of FIG. 8 shows the lamp temperature T and the detected absolute wavelength λ of the reference light.
Hg ′ (= λHg + ΔλHg). Therefore, the relationship between the lamp temperature T and the detected absolute wavelength λHg ′ of the reference light is stored in a predetermined memory in advance, for example, in a table format.

When detecting the wavelength of the reference light,
First, the glass surface temperature T of the low-pressure mercury lamp 7 is detected by the temperature sensor 14 (Step 300).

Next, λHg ′ corresponding to the detected temperature T is read from the above-mentioned correspondence table, and this is set as the absolute wavelength of the reference light. (Step 310).

Next, the absolute wavelength λHg ′ of the reference light and the respective radii of the interference fringes of the reference light and the detected light measured by the photodetector 13 are substituted into the above-mentioned expressions (14) and (15). To obtain the absolute wavelength λex of the light to be detected (step 32).
0).

As described above, since the accurate absolute wavelength of the reference light can be obtained from the temperature of the reference light, it is not necessary to control the temperature of the reference light source, and the temperature of the lamp 7 need only be detected. .

[0093]

As described above, according to the present invention,
253. A low-pressure mercury lamp was used as a reference light source, and at least 49% of a specific isotope mercury was enclosed in the low-pressure mercury lamp.
Since a 7 nm line is used, a reference light having a beautiful spectral waveform and a narrow line width can be obtained.

Further, in the present invention, the surface temperature of the low-pressure mercury lamp is controlled to a predetermined temperature (for example, 40 ° C.) or less, so that a clean and easy-to-measure spectrum waveform is obtained and the absolute value of the reference light is obtained. The wavelength can always be made to coincide with the theoretical value, whereby the calculation for correcting the reference light wavelength can be omitted.

Further, in the present invention, since the low-pressure mercury lamp is of a hot cathode type, there are the following advantages.

1. Light is emitted at a small voltage (about 30 V), no special power supply is turned on. 2. Lighting is instantaneously performed when current is applied, and light emission is instantaneously stabilized. 3. Spectral waveform due to low light emission temperature Is clean and the line width (1pm or less) is narrow. 4. It is inexpensive. 5. There is no deterioration due to turning on / off.

Further, according to the present invention, the absolute wavelength of the reference light is obtained from the relationship between the absolute wavelength of the reference light stored in advance and the temperature of the reference light source. An accurate absolute wavelength of the reference light can be obtained.

Further, according to the present invention, the spectral waveform of the reference light emitted from the low-pressure mercury lamp is determined according to the spectral waveform of each mercury isotope enclosed in the low-pressure mercury lamp and the mixture ratio thereof, and the reference light Since the absolute wavelength is determined, a configuration for measuring the absolute wavelength of the reference light is not required.

Further, in the present invention, an electrodeless discharge type or a cold cathode type is used as the low-pressure mercury lamp, and the low-pressure mercury lamp is first turned on by operating a power source with high energy. When the lighting is confirmed, the input energy of the power supply is reduced, so that reference light with a narrow spectral line width can be obtained.

Further, according to the present invention, a spectrum waveform approximation calculating means for calculating an approximate waveform of the spectrum waveform of the reference light and the laser oscillation light by applying a predetermined approximation process to the detection output of each photodetector of the photodetector array. With
Since the radius of the interference fringes of the reference light and the laser oscillation light is calculated from the approximate spectrum waveform, and the absolute wavelength of the laser oscillation light is calculated from the calculated radius of each interference fringe.
A high-resolution spectral waveform can be obtained, and the calculation at that time can be speeded up.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a wavelength distribution diagram of 253.7 nm line of various Hg isotopes.

FIG. 3 is a diagram showing spectrum waveforms of 202Hg and natural mercury.

FIG. 4 is a diagram showing spectrum waveforms of 202Hg and natural mercury.

FIG. 5 is a view showing a spectrum waveform corresponding to various mixing ratios of 202Hg.

FIG. 6 is a diagram showing the relationship between the wavelength of reference light and the purity of specific mercury.

FIG. 7 is a diagram showing a spectrum waveform of 202Hg according to temperature.

FIG. 8 is a diagram showing the relationship between the wavelength of the reference light and the temperature of the reference light source.

FIG. 9 is a diagram showing various types of low-pressure mercury lamps.

FIG. 10 is a flowchart showing an example of lighting control of a low-pressure mercury lamp.

FIG. 11 is a diagram showing the principle of a monitor etalon.

FIG. 12 is a diagram showing various methods for obtaining the square of the radius of an interference fringe.

FIG. 13 is a diagram showing an approximate calculation when obtaining the square of the radius of an interference fringe.

FIG. 14 is a flowchart illustrating an example of a procedure for obtaining the square of the radius of an interference fringe.

FIG. 15 is a flowchart showing an example of a procedure for obtaining the square of the radius of an interference fringe.

FIG. 16 is a flowchart showing an example of a procedure for obtaining the square of the radius of an interference fringe.

FIG. 17 is a flowchart showing an example of a procedure for obtaining the square of the radius of an interference fringe.

FIG. 18 is a flowchart showing a procedure for obtaining λe from a mercury isotope mixing ratio.

FIG. 19 shows spectral waveforms of mercury at various mixing ratios.

FIG. 20 is a flowchart showing a procedure for obtaining the absolute wavelength of light to be detected from the temperature of reference light.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Narrow band oscillation excimer laser 2 ... Beam splitter 3 ... Shutter 4 ... Ground glass 5 ... Wavelength detector 6 ... Beam splitter 7 ... Low pressure mercury lamp 8 ... Condensing lens 9 ... Shutter 10 ... Bandpass filter 11 ... Monitor etalon 12 ... Condensing lens 13 Optical position detector 14 Temperature sensor 15 Cooling fan 20 Wavelength detection controller 25 Bandwidth narrowing element driver 30 Temperature controller

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H01S 3/00 H01L 21/30 527 F term (Reference) 2G020 AA05 CB06 CB23 CB34 CB43 CB51 CC23 CC26 CC48 CD04 CD16 CD22 CD36 5F046 CA04 CA10 CB22 5F072 AA06 HH02 JJ08 KK08 KK15 RR05 TT27

Claims (1)

[Claims] 1. A low-pressure mercury lamp is used as a reference light source,
In a wavelength detection device, a reference light generated from the reference light source and a light to be detected are incident on a wavelength detector, and an absolute wavelength of the light to be detected is detected based on a detection output of the wavelength detector. Storage means for respectively storing the waveforms of the optical interference fringes of the respective mercury isotopes, and the low-pressure mercury by synthesizing the stored waveforms of the optical interference fringes of the respective mercury isotopes in accordance with the mixing ratio of the respective mercury isotopes. Waveform synthesizing means for determining the waveform of the optical interference fringe of the reference light emitted from the lamp; and two points corresponding to the specific intensity of the waveform of one specific optical interference fringe among the waveforms of the optical interference fringes of the reference light. A wavelength calculating unit for obtaining an absolute wavelength of the reference light based on the position, and calculating an absolute wavelength of the light to be detected based on the obtained absolute wavelength of the reference light. .