RU2253194C2 - Radiation source built around plasma focus with improved switching-mode supply system - Google Patents

Abstract

FIELD: high-energy ultraviolet and roentgen radiation sources.

SUBSTANCE: high-energy photon source has electrode pair 8 for producing plasma pinch disposed in vacuum chamber of plasma pinch generating device 2. Chamber holds working gas incorporating noble cushion gas and active gas chosen to afford generation of desired spectral line. Switching-mode power supply 10 generates electric pulses of sufficiently high voltage to build up electrical discharge between electrodes. This discharge produces high-temperature plasma pinches of high density in working gas which generate radiation in spectral line of power supply. Electrodes are of coaxial configuration with anode on axis. Active gas is introduced through hollow anode.

EFFECT: ability of optimizing spectral line of source and optimizing the latter separately with respect to cushion gas.

22 cl, 17 dwg

Description

This application is a partial continuation of US Pat. No. 09/590,962, filed June 9, 2000, US Pat. No. 09/442,582, filed November 18, 1999, and US Patent Application per. No. 09/324,526, filed June 2, 1999, which was a partial continuation of US Pat. No. 09/268,243, filed March 15, 1999, which is now US Pat. No. 6,064,072, and US Pat. No. 09/093,416, filed June 8, 1998, which is now US Pat. No. 6,051,841, and this application was a partial continuation of US Pat. No. 08 / 854,507, which is now US Pat. No. 5,763,930. This invention relates to sources of high-energy photons, and in particular to very reliable sources of X-rays and high-energy ultraviolet radiation.

BACKGROUND OF THE INVENTION

In the semiconductor industry, the development of lithographic technologies continues, which allows lithography to be performed with sizes even smaller than integrated circuits. These systems must have high reliability, high performance and reasonable latitude or process adaptability. Recently, the industry for the manufacture of integrated circuits has switched from radiation sources with the G-line (436 nm) and the I-line (365 nm) of mercury to sources in the form of excimer lasers with radiation wavelengths of 248 nm and 193 nm. This transition was accelerated by the need for higher lithographic resolution with minimal loss in depth of focus.

The needs of the integrated circuit industry will soon exceed the resolving power of radiation sources with a wavelength of 193 nm, therefore, it becomes necessary to have a reliable radiation source with a wavelength much shorter than 193 nm. There is an excimer line at 157 nm, however, it is difficult to obtain optical materials with sufficient transmission at this wavelength and a sufficiently high optical quality. Therefore, full reflection imaging systems may be required. For full reflection optical systems, a smaller numerical aperture is required than for transmission systems. Losses in resolution due to a smaller numerical aperture (NA) can only be compensated by reducing the wavelength by a factor having a larger value. Thus, if the resolution of optical lithography should be improved by a larger amount than that which can be achieved using a source with a wavelength of 193 nm or 157 nm, then a light source with radiation in the region of 10 nm is required.

Currently, in the field of technology associated with sources of high-energy ultraviolet and x-ray radiation, plasma is used, obtained by bombarding targets from various materials with laser beams, electrons or other particles. Solid targets are used, but contamination resulting from ablation (erosion) of the solid target adversely affects various elements of the system designed to produce spectral line radiation. The proposed solution to overcome the pollution problem is to use a target from a frozen liquid or frozen gas, so that the contaminants will not be deposited on optical equipment. However, in practice, no such system has been demonstrated to produce spectral line radiation.

It has been well known for many years that x-rays and high-energy ultraviolet radiation can be obtained by creating a plasma pinch. In a plasma pinch, an electric current passes through the plasma in one of several possible configurations so that the magnetic field created by the flowing electric current accelerates the electrons and ions in the plasma, concentrating them in a small volume, where their energy becomes sufficient to cause substantial pulling out "external electrons from ions and the subsequent generation of x-rays and high-voltage ultraviolet radiation. The following patents describe various known methods for generating high-energy radiation by a plasma focus or pinch plasma:

J.M. Dawson, "X-Ray Generator". U.S. Patent No. 3,961,197, June 1, 1976.

T. G. Roberts, et al., "Intense, Energetic Electron Beam Assisted X-Ray Generator" (X-ray Generator Generated by Intense High-Energy Electron Beam). U.S. Patent No. 3,969,628, July 13, 1976.

J.H. Lee, "Hypocycloidal Pinch Device", US Patent No. 4,042,848, August 16, 1977

L. Cartz, et al., "Laser Beam Plasma Pinch X-Ray System" (System for producing X-ray radiation from a plasma pinch generated by a laser beam). U.S. Patent No. 4,504,964, March 12, 1985.

A. Weiss, et al., "Plasma Pinch X-Ray Apparatus" (Device for receiving x-ray radiation of a plasma pinch). U.S. Patent No. 4,536,884, August 20, 1985

S. Iwamatsu, "X-Ray Source". U.S. Patent No. 4,538,291, August 27, 1985

G. Herziger, W. Neff, "Apparatus for Generating a Source of Plasma with High Radiation Intensity in the X-ray Region" (Device for creating a plasma source with intense radiation in the x-ray region of the spectrum). U.S. Patent 4,596,030, June 17, 1986

A. Weiss, et al., "X-Ray Lithography System" (X-ray Lithography System). U.S. Patent No. 4,618,971, October 21, 1986.

A. Weiss, et al., "Plasma Pinch X-ray Method" (Method for producing X-ray radiation from a plasma pinch). U.S. Patent No. 4,633,492, December 30, 1986.

I.Okada, Y. Saitoh, "X-Ray Source and X-Ray Lithography Method" (X-ray source and X-ray lithography method). U.S. Patent No. 4,635,282, June 6, 1987.

R.P. Gupta, et al., "Multiple Vacuum Arc Derived Plasma Pinch X-Ray Source" (X-ray source from a plasma pinch obtained from multiple arc discharges in a vacuum). U.S. Patent No. 4,751,723, June 14, 1988.

R.P. Gupta, et al., "Gas Discharge Derived Annular Plasma Pinch X-Ray Source" (X-ray source from a ring plasma pinch derived from a gas discharge). U.S. Patent No. 4,752,946, June 21, 1988.

J.C. Riordan, J.S. Peariman, "Filter Apparatus for use with an X-ray Source". U.S. Patent No. 4,837,794, June 6, 1989.

W. Neff, et al., "Device for Generating X-radiation with a Plasma Source". U.S. Patent No. 5,023,897, June 11, 1991.

D.A. Hammer, D.H. Kalantar, "Method and Apparatus for Microlithography Using X-Pinch X-Ray Source" (Method and apparatus for microlithography using an X-pin x-ray source). U.S. Patent No. 5,102,776, April 7, 1992.

M.W. McGeoch, "Plasma X-Ray Source". U.S. Patent No. 5,504,795, April 2, 1996.

G. Schriever, et al., "Laser-produced Lithium Plasma as a Narrow-band Extended Ultraviolet Radiation Source for Photoelectron Spectroscopy" (Laser-produced lithium plasma as a narrow-band extended ultraviolet radiation source for photoelectron spectroscopy). Applied Optics, Vol. 37, No. 7, pp. 1243-1248, March 1998.

R.Lebert, et al., "A Gas Discharged Based Radiation Source for EUV Lithography" (VUV lithography). Int.Conf. On Micro and Nano Engineering (International Conference on Micro- and Nanotechnology), September, 1998.

W.T. Silfast, et al., "High-power Plasma Discharge Source at 13.5 nm and 11.4 nm for EUV Lithography" (Powerful source of gas-discharge plasma with 13.5 nm and 11.4 nm radiation for VUV lithography). SPIE Proc. On Emerging Lithographic Technologies III, Vol. 3676, pp. 272-275, March 1999.

F.Wu, et al., “The Vacuum Spark and Spherical Pinch X-ray / EUV Point Sources” (SPIE and Spherical Pinch), SPIE Proc. On Emerging Lithographic Technologies III, Vol. 3676, pp. 410-420, March 1999.

I. Fomenkov, W. Partlo, D. Birx. "Chracterization of a 13.5 nm for EUV Lithography based on a Dense Plasma Focus and Lithium Emission" (Definition of 13.5 nm spectral line parameters for VUV lithography based on dense plasma focus and lithium radiation), Sematech International Workshop on EUV Lithography, October, 1999.

Plasma focus devices known in the art can generate intense radiation suitable for lithography with irradiation near the x-ray spectrum, but they have a limited repetition rate due to the high electrical energy required per pulse and due to short-lived internal components. Requirements for the amount of stored electrical energy for these systems range from 1 to 100 kJ. Repetition rates usually do not exceed a few pulses per second.

At the same time, a simple and reliable system for obtaining a spectral line is required to generate high-energy ultraviolet and X-ray radiation, and this system should work at high repetition rates and not have the problems inherent in analogues associated with the formation of contaminants.

SUMMARY OF THE INVENTION

The present invention provides a source of high-energy photons, comprising: A) a vacuum chamber; C) at least two electrodes mounted coaxially in said vacuum chamber defining an electric discharge region and configured to create high-frequency plasma pinches at the pinch site during electric discharge; C) a working gas containing an active gas and a buffer gas, wherein the buffer gas is a noble gas and the active gas is selected so as to provide emission of at least one spectral line; D) an active gas supply system for supplying active gas to said discharge region; and E) a pulsed power system providing electric pulses with an energy of at least 12 J per pulse at a pulse repetition rate of at least 2000 Hz, said system comprising: 1) a battery of charging capacitors; 2) a high-speed charger for charging said battery of charging capacitors for periods of less than about 0.5 microseconds; 3) means for controlling the voltage of the charging capacitors to control the charge on the battery of charging capacitors; 4) an electromagnetic compression circuit comprising at least one capacitor bank and at least one saturable core inductor; 5) a charging capacitor battery trigger for discharging said charging capacitor bank into said electromagnetic compression circuit; and 6) a pulse transformer for increasing pulse voltages by at least a factor of 4, comprising at least three single-turn primary windings and a single single-turn secondary winding. A pair of electrodes is located in the vacuum chamber to create a plasma pinch. The chamber is filled with a working gas, which includes a noble buffer gas and an active gas selected so as to obtain the desired spectral line. The switching power supply provides the supply of electrical voltage pulses, the magnitude of which is sufficient to create an electric discharge between the electrodes in order to obtain plasma pinches in the working gas with a very high temperature and high density, which provide radiation in the spectral line of the source or active gas. Preferably, the electrodes have a coaxial configuration, with the anode located on the axis. The anode is preferably hollow, and active gas is introduced through the anode. This allows optimization of the spectral line of the source and, separately, optimization of the buffer gas. The preferred options for the presented optimization of the capacitance, length and shape of the anode are described, as well as preferred active gas supply systems. Preferred options also include a pulse power system comprising a charging capacitor and an electromagnetic compression circuit comprising a pulse transformer. A cooling system with a heat pipe for cooling the central anode is described. In preferred embodiments, an external reflective radiation collector, which is also a directing device, collects the radiation generated in the plasma pinch and directs this radiation in the desired direction. Options for forming a focused beam and a parallel beam are described. In addition, in preferred embodiments, the active gas is lithium vapor, and the buffer gas is helium, and the radiation collector is made of material or coated with a material having a high reflectivity during sliding incidence. Suitable reflective materials are molybdenum, palladium, ruthenium, rhodium, gold or tungsten.

In other preferred embodiments, the buffer gas is argon, and lithium gas is produced by evaporating solid or liquid lithium located in the hole along the axis of the central electrode of the coaxial electrode configuration. In preferred embodiments, the pollutant particles are collected on a conical nested collector of pollutant particles having surfaces aligned with the light rays exiting the pinch point and directed toward the radiation collector. The conical embedded collector of polluting particles and the collector - radiation guide device are maintained at a temperature in the range of about 400 ° C, which is higher than the melting point of lithium and significantly lower than the melting point of tungsten. Tungsten and lithium vapors are collected on the pollutant collector, but lithium will evaporate from the pollutant collector and collector-guiding device, while tungsten will constantly remain on the pollutant collector and therefore it will not collect on the collector-guiding radiation device and impair its reflectivity. Reflective collector - a radiation directing device and a conical embedded collector of polluting particles can be made together as a single structural element or they can be separate structural elements centered (coaxial) with each other and with the pinch.

If required, the camera can be equipped with a special window, which is designed to transmit light in the VUV range and reflect light (radiation) with lower energy, including visible light. This window is preferably a small diameter window consisting of an extremely thin material such as silicon, zirconium or beryllium.

Applicants describe here a prototype third-generation dense plasma focus (PFD) device, created by the applicants and their staff as a source for lithography using high-energy ultraviolet irradiation (VUV), using a fully solid-state device for switching pump power supply (excitation). Using the results obtained using a vacuum diffraction spectrometer combined with the results of measurements with a silicon photodiode, the applicants found that using a 13.5 nm emission line of a doubly charged lithium ion, a significant amount of radiation can be generated within the reflection band of the Mo / Si mirrors. In this prototype PPF device, 25 J of stored electrical energy per pulse is converted to approximately 0.76 J of radiation in the 13.5 nm band (line) emitted at a solid angle of 4π steradians. The characteristic of this device was investigated by the parameter of pulse repetition rate up to the maximum frequency of a DC power supply (DC) of 200 Hz. Up to this repetition rate (frequency), no significant decrease in VUV output per pulse was detected. At 200 Hz, the measured energy stability from pulse to pulse was σ = 6% and no drop in pulse energy was observed. The electrical circuitry and operation of this prototype PPF device are presented in the description of several preferred modifications, which are supposed to provide improved stability, efficiency and operating parameters.

In addition, a fourth-generation PPF device is described that is capable of operating at a frequency of 2000 Hz, which is capable of producing about 20 mJ per pulse of useful VUV radiation in an angle of 2π steradian.

The present invention provides a practical implementation of VUV lithography in a reliable, high-brightness VUV radiation source, the emissive characteristics of which are in good agreement with the reflection band of Mo / Si or Mo / Be mirror systems.

BRIEF DESCRIPTION OF THE DRAWINGS

1 and 1A are diagrams of high-energy photon sources representing preferred embodiments of the present invention;

figure 2 - diagram of a device for forming a three-dimensional plasma pinch with electrodes in the form of a disk;

3 is a diagram of a fourth preferred embodiment of the present invention;

4 is a preferred circuit diagram for a preferred embodiment of the present invention;

figa is a view of a prototype device created by the applicants and their employees;

5B is a sectional view showing the electrodes of a prototype with preionization means in the form of spark pins;

figv1-5v6 - the stages of formation of a plasma pinch;

figs - section of the electrode region with a protective screen;

5C1-5C6 - the steps of forming a plasma pinch with a protective shield installed;

6 is a pulse shape generated by the prototype device;

Fig.7 - part of the VUV beam formed by a hyperbolic collector;

figa is a perspective image of a hyperbolic collector;

figv - part of the VUV beam formed by an ellipsoidal collector;

Fig - peak lithium at 13.5 nm in comparison with the reflection coefficient of coatings MoSi;

Fig.9 - nested conical collector of polluting particles;

figure 10 - a window of thin Be to reflect visible light and transmitting VUV radiation;

11 is a diagram representing the reflection coefficient of various materials for ultraviolet radiation of 13.5 nm;

12 is a diagram showing a method of introducing a source gas and a working gas;

Fig. 13 is a timing chart representing the voltage at the anode and the intensity of the VUV;

figa, 14B, 14C and 14D - the effect of various designs of the anode on the plasma pinch;

15 is a diagram illustrating a method of using RF energy to operate a gas source, which is a pair of lithium;

Fig.16, 16A, 16B, 16C, 16D and 16E are the elements and technical characteristics of the prototype of the fourth-generation plasma pinch device;

17 is a heat pipe cooling method for cooling an anode in a preferred plasma pinch device.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Basic construction

Simplified drawings representing the basic structure of a source with high-energy ultraviolet radiation are shown in FIGS. 1 and 1A. The main elements are a device for creating a plasma pinch, a collector 4 of high-energy photons and a hollow fiber 6. The source of the plasma pinch contains a coaxial electrode 8, fed from a low-inductance switching power supply 10. The pulsed power supply in this preferred embodiment is a high-voltage circuit with high energy efficiency, capable of providing a very fast (approximately 50 ns) increase in the pulse amplitude in the range from 1.4 to 2.5 kV on coaxial electrode 8 with a pulse repetition rate of at least 1000 Hz.

Figure 1 shows the case where a small amount of a working gas, such as a mixture of helium and lithium vapor, is near the base of the electrode 8, as shown in figure 1. For each high-voltage pulse between the internal and external electrodes of the coaxial electrode 8, an avalanche breakdown occurs either due to preionization or self-breakdown. The avalanche process that occurs in the buffer gas ionizes the gas and creates a conductive plasma between the electrodes at the base of the electrodes. When a conductive plasma exists, current flows between the inner and outer electrodes. In this preferred embodiment, the inner electrode is at a high positive potential, and the outer electrode is under the ground potential. Current will flow from the inner electrode to the outer electrode and, therefore, the electrons will move toward the center, and positive ions will move from the center. This current generates a magnetic field that acts on the moving charge carriers, accelerating them in the direction from the base of the coaxial electrode 8.

When the plasma reaches the end of the central electrode, the electric and magnetic forces acting on the plasma compress the plasma into a “focus” near point 11 along the central line and at a small distance from the end of the central electrode, and the plasma pressure and temperature increase sharply, reaching extremely high temperatures, in some cases much higher than the temperature on the surface of the sun! The dimensions of the electrodes and the total electrical energy in the circuit are preferably optimized to obtain the desired black body temperature in the plasma. To generate radiation in the 13 nm region, the temperature of the black body must be above 20-100 eV. In general, for a particular coaxial configuration, the temperature will increase with increasing voltage in the electric pulse. The shape of the spot or point of radiation is somewhat irregular in the axial direction, and in the radial direction in a rough approximation corresponds to the Gaussian shape. The typical radial size of the source is 300 microns and its length is approximately 4 mm.

In most known plasma pinch devices described in the technical literature, the radiation spot emits radiation in all directions and its spectrum is approximately close to the spectrum of a black body. The purpose of lithium in the working gas is to narrow the spectrum of radiation from this radiation spot.

A second basic structure is shown in FIG. 1A. In this case, lithium vapor is an active gas and is introduced through the center of the anode. The buffer gas is helium, and it is introduced into another (separated) place. A suction pump is used to create the required vacuum in the chamber. The lithium feed lines are heated to maintain lithium in a vapor state.

Lithium vapor

Double-ionized lithium has an electronic transition at 13.5 nm and in the buffer helium gas it acts as an atom, which is the source of radiation. Twice-ionized lithium is an excellent choice for two reasons. Firstly, it has a low melting point of lithium and a high vapor pressure. Lithium ejected from the radiation spot can be prevented from being deposited on the walls of the chamber and collecting optical elements simply by heating these surfaces above 180 ° C. Then, lithium in the vapor phase can be pumped out of the chamber along with a buffer gas, helium, using the standard technology of turbomolecular pumping. And besides, lithium can be easily separated from helium simply by cooling these two gases.

To provide good reflection at 13.5 nm, coating materials are available. On Fig shows the peak emission of lithium in comparison with published data on the reflection coefficients of MoSi.

A third advantage of using lithium as a radiation source atom is that non-ionized lithium has a small absorption cross section for 13.5 nm radiation. Moreover, any ionized lithium emitted from the radiation spot can easily be removed using a weak electric field. The remaining non-ionized lithium is substantially transparent to 13.5 nm radiation. Currently, laser ablation (destruction) of a frozen xenon jet is used in the most popular radiation source proposed at 13 nm. In such a system, almost all injected xenon should be captured until the next pulse, since xenon has a large absorption cross section at 13 nm.

Xenon wide bandwidth

Another preferred radiation source atom is xenon, which has a broad emission line in the 13.5 nm region. Applicants describe in the subsequent section of this description how to solve the problem of strong xenon absorption.

Radiation collector

The radiation that occurs in the radiation spot is radiated uniformly into the full angle of 4π steradian. To capture this radiation and direct its nalitographic instrument, some elements of collecting optics are required. In previously proposed sources with radiation at 13 nm, it was proposed to use collecting optics based on mirrors with a multilayer dielectric coating. Mirrors with a multilayer dielectric coating are used to achieve high efficiency of collecting radiation in a wide angular range. Any source of radiation that creates contamination would cover these dielectric mirrors with impairment, worsen their reflection coefficient, and, consequently, the collected output radiation from the source would decrease. This preferred system would be susceptible to erosion of the electrodes and, therefore, over time, it would degrade the quality of any dielectric mirror located near the radiation spot.

Some materials are available with a high reflectance at small moving incidence angles for UV radiation with a wavelength of 13.5 nm. Graphs for some of them are shown in Fig. 11. Good choices are molybdenum, rhodium and tungsten. The collector may be made of these materials, but they are preferably used as a coating on a structural base material such as nickel. This conical section can be made by electrodeposition of nickel on a removable template.

To obtain a collector capable of receiving radiation from a large conical angle, several conical sections can be embedded inside each other. Each conical section can perform more than one reflection of radiation in order to redirect its sector of the radiation cone in the desired direction. Creating a collector for work near a sliding fall will allow you to get a collector that will be most resistant to the deposition of material of the destroyed electrode. The reflection coefficient of mirrors during a sliding incidence is such that it strongly depends on the surface roughness of the mirror. The dependence on surface roughness decreases as the angle of incidence approaches the sliding incidence. Evaluation shows that we can collect and direct radiation at 13 nm emitted into a solid angle of at least 25 degrees. Preferred collectors for directing radiation into optical fibers are shown in FIGS. 1, 2, and 3.

Tungsten electrodes - tungsten collector coatings

A preferred method of selecting material for an external reflective collector is that the covering material on the collector is the same as the electrode material. Tungsten is a promising candidate because it has demonstrated its function as an electrode, and the real part of its 13 nm reflection coefficient is 0.945. When using the same material for the electrode and coating the mirror, the deterioration of the reflectivity of the mirror is minimized when the material of the destroyed electrode is deposited on collecting mirrors.

Silver electrodes and coatings

Silver is also an excellent choice for electrodes and coatings, because it also has a low refractive index of 13 nm and has a high thermal conductivity, allowing for work with a higher pulse repetition rate.

Conical nested particle collector

In another preferred embodiment, the collector-guiding device is protected from surface contamination by the material of the evaporating electrode using a contaminant collector that collects all tungsten vapors before they can reach the collector-guiding device 4. Fig. 9 shows a conical embedded collector 5 of contaminant particles to collect contaminants flying out of a plasma pinch. The collector 5 of polluting particles consists of nested conical sections, the surfaces of which are combined with light rays passing from the center of the pinch spot (spot) and directed to the collector - guide device 4.

The collected contaminants are tungsten vaporized from tungsten electrodes and vaporized lithium. The collector of polluting particles is attached to the collector - the radiation directing device 4 or represents part of it. Both collectors consist of a base coated with nickel. Part of the collector — the radiation directing device 4 — is coated with molybdenum or rhodium to obtain a very high reflection coefficient. Preferably, both collectors are heated to about 400 ° C., which is well above the melting point of lithium and well below the melting point of tungsten. Lithium and tungsten vapors will collect on the surfaces of the collector 5 of contaminants, but lithium will evaporate; in addition, lithium is collected on a collector - a guiding device 4 - from where it will soon also evaporate. Tungsten, which is collected on a collector of 5 polluting particles, will remain on it constantly.

7 shows the optical characteristics of the collector created by the applicants. The collector consists of five nested parabolic sliding-fall reflectors, but only three out of five reflectors are shown in the drawing. Two internal reflectors are not shown. In this design, the angle from which the radiation is collected is about 0.4 steradian. As discussed below, the surface of the collector is coated and heated to prevent lithium deposition. This design allows you to get a parallel beam. Other preferred designs, such as those shown in FIGS. 1, 3, and 10, focus the beam. The collector should be coated with a material with a high reflection coefficient with a sliding incidence in the region of the wavelength of 13.5 nm. Two of these materials are palladium and ruthenium.

Another collector — a guiding device — designed to focus the beam, is shown in FIG. This collector — a guiding device — uses an ellipsoidal mirror 30 to focus the VUV source. Mirrors of this type can be obtained commercially from suppliers, such as, for example, Reflex SVO, which has production facilities in the Czech Republic, in the United States, Bede Scientific Instrument Ltd., whose offices are located in the United Kingdom and Inglewood, distributes such products. , PCS. Colorado. The reader should pay attention to the fact that such a mirror collects rays located only in the angles shown at 32 in FIG. However, additional mirror elements may be included inside the mirror 30 and outside the mirror 30 to collect and focus additional rays. The reader should also note that along the propagation of the rays after the mirror 30, other mirror elements may be located to collect the rays from a narrow angle, or mirror elements may be located before the mirror 30 to collect the rays from a wider angle.

Light guide

It is important to avoid deposition of the above materials on the lighting optics of a lithographic instrument. Therefore, to further ensure their separation, it is preferable to use a light guide 6. The light guide 6 is a hollow light guide, in which substantially total external reflection is also carried out on its internal surfaces. The main collecting optics can be designed to reduce the conical angle of the collected radiation so that it matches the receiving angle of the hollow fiber. This principle is shown in figure 1.

The dielectric mirrors of the lithographic instrument would then be very well protected from any particles flying off the electrodes, since tungsten, silver or lithium atoms would be scattered in the upper part of the fiber in the event of a collision with a stream of buffer gas introduced below into the hollow fiber, as shown in FIG. .

Switching power supply

A preferred switching power supply 10 is a solid state, high frequency, high voltage switching power supply that uses a solid state trigger and a magnetic key circuit, such as a switching power supply, as described in US Pat. No. 5,936,988. These sources are very reliable and can operate continuously without significant maintenance costs for months or billions of pulses. The description of US patent No. 5,936,988 is incorporated into this description by reference.

Figure 4 shows a simplified electrical circuit providing pulsed power. A preferred embodiment includes a direct current power supply 40, which is a controlled resonance charging source of the same type as used in excimer lasers. C ₀ is a battery of standard (non-special) capacitors with a total capacity of 65 μF, capacitor C ₁ is for the formation of a sharp maximum, which is also a battery of finished capacitors with a total capacity of 65 μF. A saturable core inductor 42 has an inductance at saturation corresponding to a trigger of about 1.5 nH. Trigger 44 - IGBT. The diode 46 and inductor 48 create an energy recovery circuit similar to that described in US Pat. No. 5,729,562, which allows the reflected electrical energy to be stored in C ₀ from one pulse to the next pulse.

System

Thus, as shown in FIG. 1, the working gas mixture of helium and lithium vapor is supplied to the coaxial electrode 8. Electrical pulses from the pulsed power supply 10 create a dense plasma focus at the location indicated at 11 with sufficiently high temperatures and pressures to there was a double ionization of lithium atoms in the working gas, causing the generation of ultraviolet radiation with a wavelength of about 13.5 nm.

This light is collected in the collector 4 with total external reflection and sent to the hollow fiber 6, where the light is then sent to a lithographic tool (not shown). A vacuum of about 4 Torr is maintained in the discharge chamber 1 using a turbo-suction pump 12. A certain amount of helium in the working gas is separated in the helium separator 14 and is used to purge the fiber, as shown in figure 1, at 16. The pressure of the helium in the fiber preferably meets the requirements of pressure for a lithographic tool, which is usually maintained at low pressure or in a vacuum. The temperature of the working gas is maintained at the required temperature by means of a heat exchanger 20, and the gas is purified using an electrostatic filter 22. The working gas is fed into the volume of the coaxial electrode, as shown in FIG.

Prototype device

A drawing of a prototype plasma pinch device created and tested by the applicant and his staff is shown in FIG. 5A. The main elements are a set of capacitors C ₁ , a set of capacitors C ₀ , IGBT switches, a saturated core inductor 42, a vacuum chamber 3 and a coaxial electrode 8.

Test results

FIG. 6 shows a typical pulse shape measured by the applicant when using the device shown in FIG. 5A. Applicants recorded voltage at C ₁ , current at C _1, and radiation intensity at 13.5 nm over a time period of 8 microseconds. The total energy in this typical pulse is about 0.8 J. The pulse width (at half maximum) is about 280 ns. The voltage at C ₁ before the breakdown is less than 1 kV.

This pilot version can operate with a pulse repetition rate of up to 200 Hz. The measured averaged radiation in the frequency band of 13.5 nm at 200 Hz is 152 W at an angle of 4π steradian. Energy stability at 1 sigma is about 6%. Applicants have estimated that 3.2% of the energy can be directed to a useful 13.5 nm radiation beam using collector 4 shown in FIG.

The second plasma pinch device

A second plasma pinch device is shown in FIG. This device is similar to the plasma pinch device described in US Pat. No. 4,042,848. This device contains two external electrodes 30 and 32 of a disk shape and an internal electrode 36 of a disk shape. The pinch is created from three directions, as described in patent No. 4.042.848 and as indicated in figure 2. The pinch begins near the peripheral circumference of the electrodes and extends toward the center, and the radiation spot develops along the axis of symmetry and in the center of the inner electrode, as shown in FIG. 2 at 34. The radiation can be collected and directed, as described in relation to the embodiment shown in FIG. 1. However, it is possible to capture radiation in two directions emerging from both sides of the device, as shown in FIG. In addition, due to the location of the dielectric mirror at position 38, a significant percentage of the radiation initially reflected to the left could be reflected back through the radiation spot. This would increase the radiation in the direction to the right.

Third plasma pinch device

The third option can be described referring to Fig.3. This embodiment is similar to the first preferred embodiment. However, in this embodiment, the buffer gas is argon. Helium possesses the required relative transparency property for 13 nm radiation, but it also has the undesirable property that it has a small atomic mass. Low atomic mass forces us to work at the main pressure in the 2-4 Torr system. An additional disadvantage of the small atomic mass is not that the length of the electrode is required so that the distance at which the acceleration occurs is consistent with the time characteristic of the electrical excitation circuit. Since Not light, the electrode should be longer than desired, so that It does not reach the end of the electrode simultaneously with the peak of the current flowing through the excitation circuit. A heavier atom, such as Ar, at a given pressure will produce lower transmittance than He, but due to its higher mass, it can create a stable pinch at a lower pressure. The reduced operating pressure Ar in excess compensates for the properties of higher absorption of Ar radiation. In addition, the required electrode length is reduced due to the higher atomic mass. A shorter electrode has an advantage for two reasons. First, the resulting reduction in circuit inductance when using a shorter electrode. A lower inductance makes the excitation circuit more efficient and therefore reduces the required electric pump energy. The second advantage of a shorter electrode is the reduction in the length of the heat-conducting path from the top of the electrode to the base. The main part of the thermal energy is transferred to the electrode at the apex, and the electrode is cooled due to thermal conductivity mainly at the base (although radiation cooling also takes place). A shorter electrode results in a lower temperature drop along the length of the electrode from the hot peak to the cold base. A lower pump energy per pulse and an improved cooling path allow the system to operate at a higher repetition rate. An increase in the repetition rate directly increases the average optical output of the system. A proportional increase in the output power due to an increase in the repetition rate, in contrast to an increase in energy per pulse, is the most desirable way to increase the average output power of lithographic light sources.

In this preferred embodiment, lithium is not introduced into the chamber in a gaseous form, as in the first and second embodiments. Instead, solid lithium is placed in a hole in the center of the central electrode, as shown in FIG. The heat from the electrode then leads to lithium heating up to the evaporation temperature. By adjusting the height (distance) of lithium relative to the hot top of the electrode, it is possible to control the partial pressure of lithium near the top of the electrode. One preferred way of doing this is shown in FIG. A mechanism is proposed for adjusting the tip of a solid lithium rod relative to the tip of the electrode. Preferably, the system is positioned vertically, so that the open side of the coaxial electrodes 8 is the apex, so that molten lithium simply forms a “puddle” near the apex of the central electrode. The beam will exit straight up in the vertical direction, as indicated in FIG. 5A. (An alternative approach is to heat the electrode to a temperature above the melting point of lithium so that lithium is added as a liquid.) To pump liquid at the speed required for any particular pulse repetition rate, extremely low feed pumps are available. A tungsten wick can be used for capillary flow of liquid lithium into the region of the apex of the central electrode.

A hole down the center of the electrode provides another important advantage of this option. Since a plasma pinch is formed near the center of the apex of the central electrode, a lot of energy is dissipated (i.e. sprayed) in this region. The electrode material near this point will evaporate and ultimately fall onto other surfaces inside the low pressure chamber. The use of an electrode with a central hole significantly reduces the erosion of the material. In addition, the applicant’s experiments showed that the existence of lithium vapors in this region further reduces the erosion rate of the electrode material. Bellows or other suitable sealing methods should be used to maintain good tightness at the point where the electrode equipment is introduced into the chamber. Replacement fully filled solid lithium electrodes can be made easily and cheaply and can easily be replaced in a chamber.

Small window in the vacuum chamber

The pinch generates a very large amount of visible light, which must be separated from the VUV radiation. In addition, a window is desirable in order to provide an additional guarantee that lithographic optics are not contaminated with lithium or tungsten. The high energy ultraviolet beam generated in the device according to the present invention is well absorbed in the solid. Therefore, the implementation of the window for the beam is a very complex technological problem. The applicants' preferred window solution is to use an extremely thin foil that will transmit VUV and reflect visible light. The window proposed by the applicants is a beryllium foil (about 0.2 to 0.5 microns) inclined at an incidence angle of about 10 ° relative to the axis of the incoming beam. With this configuration, almost all visible light is reflected and about 50-80 percent of the VUV is transmitted. Of course, such a thin window is not very durable. Therefore, applicants use a window of very small diameter and the beam is focused through this small window. Preferably, the diameter of the thin beryllium window is about 10 mm. The heating of a small window must be taken into account, and at high repetition frequencies, special cooling of this window will be required.

In some designs, this element can be made simply as a beam splitter, which will simplify the design, since the pressure difference will not affect the thin optical element.

Figure 10 shows a preferred embodiment in which the radiation collector 4 is elongated by an elongated collector end 4A to focus the beam 9 through a beryllium window 7 with a diameter of 1 mm, a thickness of 0.5 microns.

Preionization

Applicants' experiments have shown that good results can be obtained without preionization, but with preionization, work improves. The prototype device shown in Fig. 5A contains preionizers in the form of spark gap excited by direct current to preionize the gas between the electrodes. With the use of advanced means of preionization, applicants are able to significantly improve energy stability and improve other operating parameters. Preionization is a well-developed method used by applicants and other professionals to improve excimer laser performance. Preferred ionization methods include:

1) spark gap with direct current excitation

2) RF spark arrester

3) surface discharge with rf excitation

4) corona discharge

5) the chain of formation of an acute maximum in combination with preionization.

These methods are well described in the scientific and technical literature related to excimer lasers, and are well known.

Protective screen

On figv shows the location of two of the eight spark pins 138, providing preionization in the preferred embodiment. 5B also shows a cathode 111 and anode 123, which are an external stainless steel element and an internal tungsten element. The insulator in a circle covers the lower part of the anode 123, and the film insulator 125 with a thickness of 5 mils (127 μm) completes the isolation of the anode from the cathode. FIGS. 5B1-6 show the development of a typical pulse leading to a pinch that is fully formed in FIG. 5B5 approximately 1.2 μs after the start of the discharge. During the discharge, the plasma accelerates toward the top of the anode under the action of the Lorentz forces acting on ions and electrons created by the current flowing through the plasma. Upon reaching the top of the electrode, shown at 121 in FIG. 5B, radially directed component forces compress and heat the plasma to high temperatures.

When the plasma is compressed, the existing axially directed forces acting on the plasma tend to lengthen the plasma column, as shown especially in FIG. 5B6. It is this lengthening that leads to instabilities. As soon as the plasma column “sprouts” along the axis beyond a certain point, the voltage drop across the region of the compressed plasma becomes too large to be supported by the low-pressure gas in this region around and near the top of the anode. An arc discharge occurs and most of the current or all of the current flows through a shorter gas region near the top of the low density anode, as shown by the dashed line in FIG. 5B6. This arc discharge is harmful because it creates instabilities in the pulse and causes relatively fast erosion of the electrodes.

The solution to this problem is to provide a physical barrier to the axial movement of the plasma column. Such a barrier is shown as an element indicated by 143 in FIG. 5C and is called a protective shield because it acts as a shield to prevent the release of plasma in the PPF device. The protective screen should be made of electrical insulating material with good mechanical and thermal strength. In addition, the chemical conformity of the material of the protective shield must be taken into account when high reactivity elements such as lithium are used during operation of the device. Lithium is proposed as a radiating element for this VUV source because of its intense radiation at 13.5 nm. An excellent candidate (for a protective shield) is monocrystalline alumina, sapphire or amorphous sapphire, such as, for example, a material manufactured by General Electric under the brand name Lucalux.

It was found that the optimal shape of the protective screen is a dome centered on the anode, the radius of the dome is equal to the diameter of the anode, as shown in figs. This form is in good agreement with the streamlines in the plasma, which arise naturally when the plasma is in the most compressed state. If the protective screen is located farther from the top of the anode, then the plasma column will be too elongated for sufficient plasma heating to occur, and there is a danger of an arc discharge. If the protective shield is too close to the top of the anode, then the current flowing from the central axis and lower to the cathode is limited, which also leads to insufficient plasma heating.

The hole at the top of the shield 143, indicated by 144, is required so that the VUV radiation can exit and this radiation can be collected for use. This hole should be made as small as possible due to the fact that the plasma tends to flow out through this hole and form a long narrow pillar above the protective screen. An inclined section in this hole, as shown at 144, allows for the collection of a larger amount of radiation remote from the VUV axis generated by the plasma pinch device.

FIGS. 5C1-6 show how a protective shield restrains a plasma pinch and prevents the formation of an arc discharge.

Combining types and densities of gas

Applicants have found that a single (one) gas does not satisfy the requirements for a source gas and the requirements for an optimized buffer gas. The source gas should be similar to lithium with radiation at 13.5 nm in a narrow frequency range or like xenon with radiation in a wide frequency range of about 13.5 nm. Parameters such as density, breakdown characteristics, and absorption of lithium and xenon are not optimal for use as a buffer gas. For example, xenon has too much intrinsic absorption, and lithium has insufficient density to be used as a buffer gas.

To solve this problem of conflicting requirements, the applicants, as shown in Fig. 12, separate the working gas into a source gas and a buffer gas and supply 42 a source gas, such as a mixture of 5% Xe and 95% He, up the center of the anode 40 Applicants then provide an optimized buffer gas, such as helium or a mixture of helium and argon, in the main region of the chamber, maintained at constant pressure. The source gas inside the anode will then be under the pressure of the buffer gas, and the flow rate of the source gas 42 determines the partial pressure of the source gas mixed with the buffer gas in the main part of the chamber. It is optimal to have a low flow rate of the source gas in order to minimize the partial pressure of the source gas in the main region of the chamber. (The working gas pressure in the main region of the chamber is controlled by a pressure control system not shown.) Buffer gas circulates between the anode 40 and the cathode 44, as shown at 46.

Capacity optimization

Applicants have discovered that the highest plasma temperature exists when a plasma pinch occurs simultaneously with the maximum current provided by the field capacitor bank. For a given configuration of the anode and the density of the buffer gas, the plasma front will propagate down the length of the anode in a given time for a given value of the charging voltage. The maximum radiation efficiency is obtained by adjusting the value of the capacitance and the charging voltage so that the maximum value of the capacitor current occurs during a plasma pinch.

If a higher level of input energy is required and, therefore, a higher charging voltage, then the excitation capacitance should be reduced so that the duration of the waveform of the excitation corresponds to the propagation time of the plasma down along the length of the anode. Since the energy stored in the capacitor is proportional to the square of the voltage and linearly proportional to the capacitance, the stored energy will linearly increase with increasing voltage, since the capacitance decreases in proportion to the increase in voltage.

13 is a graph showing the measured voltage across the drive capacitance, the measured voltage at the anode, and the VUV intensity versus time for a preferred embodiment in which the capacitance is best selected to obtain the maximum capacitor current during the pinch. In this case, for a 2 cm long anode, the pressure of the buffer gas He was 2.5 Torr and the capacitance C ₁ - 3 μF.

Optimal anode shape

Applicants have discovered that in hollow-anode configurations, the plasma pinch grows rapidly along the axis as soon as the pinch has formed and will propagate down into the hole in the hollow anode. Since this pinch grows in length, ultimately, a too large voltage drop occurs along its length and an arc discharge arises on the surface of the anode. One solution to prevent such an arc discharge is to provide a protective shield that provides a physical barrier to the growth of the length of the pinch extending outward from the anode, as described above. Another solution to reduce the growth rate of the pinch length down into the hollow anode is to increase the open diameter inside a narrow region of the anode. This will slow down the growth of the length of the pinch and prevent the occurrence of an arc.

All previous literature shows a hollow cathode with a constant diameter of the hollow part. On figa, 14B, 14C and 14D shows examples of pinch shapes for various forms of a hollow anode. The configuration shown in FIG. 14D represents the shape with the shortest pinch.

The length of the open part of the anode

Since the time of plasma propagation downward determines the place (moment) where a pinch appears on the wave-like curve of the excitation voltage, the applicants were able to adjust the duration of the pinch section of the plasma focus device by changing the magnitude of the open part of the anode and, therefore, the duration of plasma propagation downward.

The density of the buffer gas is determined by the required diameter of the plasma pinch, and the excitation capacity in practice is limited to a certain range. These two parameters, combined with the excitation voltage, determine the required downstream propagation time of the plasma. Then the propagation time of the plasma down can be adjusted by increasing or decreasing the length of the open part of the anode. Preferably, the downward propagation time of the plasma is chosen so that the moment of the plasma pinch occurs during the maximum value of the wave-like curve of the excitation current. If a longer plasma pinch is required, then the open length of the anode can be reduced, therefore, the plasma propagation time is shortened, and this leads to the fact that the plasma pinch appears earlier on the wave-like excitation curve.

Lithium feed method

The lithium supply circuits described above depend on increasing the temperature of the anode to a sufficiently high value at which lithium vapor pressure reaches the required level. This temperature is in the range of 1000-1300 ° C.

An alternative method is to fabricate an RF antenna from a material such as porous tungsten impregnated with lithium. This porous tungsten antenna 50 filled with lithium is located below the inside of the anode, as shown in FIG. A source of RF energy 52 creates a plasma layer on and around the antenna, it will excite atoms that move upward under the action of a gas stream 54 through the center of the hollow anode, and lithium atoms will be transferred to the end of the anode.

The rate of lithium ion formation is easily controlled by adjusting the power level of the RF source. In addition, the porous tungsten anode can be maintained by this RF excitation at a temperature sufficient for capillary supply of liquid lithium upward from the reservoir 56 located at the bottom of the anode.

Anode cooling

In preferred embodiments of the present invention, the central anode has an outer diameter in the range of about 0.5 cm to 1.25 cm. It is assumed that the anode absorbs a significant amount of energy due to the plasma moving down during discharge and due to absorption of radiation from the plasma pinch. Cooling in the range of about 15 kW may be required. Since the gas pressure is very low, strong cooling cannot occur by convection through the buffer gas. Radiative cooling could only be effective at very high anode temperatures. It would take a very large drop in temperature to remove heat down the length of the anode.

If lithium vapors are used as the active gas and they are introduced through the center of the anode, then it will be necessary to maintain the temperature of the anode in the range of 1000 ° C-1300 ° C or higher. Such a high operating temperature, the requirement to remove a significant amount of heat, taking into account environmental conditions and high voltage limit the choice of cooling method. However, one method with a lithium (or other alkali metal) heat pipe represents the possibility of a relatively simple and reliable way to solve this problem. Lithium heat pipes begin to work effectively at temperatures around 1000 ° C. In the specific design of such devices, refractory metals — molybdenum and tungsten — are usually used for the casing and internal wick, and therefore they can operate at very high temperatures. Initial studies have revealed that there is confidence that such a heat pipe is able to satisfy the cooling requirements of the PPF device.

The simplest option could be in the form of a tubular or annular heat pipe, which is combined with the PPF anode for the best thermal connection (heat transfer). A possible option would be an annular heat pipe to provide liquid or vapor lithium to the PPF plasma. For example, a solid heat pipe with a diameter of 0.5 inches (1.27 cm), diverting 15 kW, would have a power density of 75 kW / inch ² (11.8 kW / cm ³ ). An annular heat pipe having an external diameter of 1 inch and an internal diameter of 0.5 inch, which removes 15 kW of heat, would have a power density of 25.4 kW / inch ² (3.9 kW / cm ² ). Both of these examples illustrate the possibility of this method, since power densities far exceeding 15 kW / cm ^{2 are} demonstrated with lithium heat pipes. During operation, heat pipes have only a very small temperature gradient along the length and for practical purposes can be considered as having a constant temperature along the length. Therefore, the "cold" (refrigerator) end of the heat pipe will also be at a temperature of about or above 1000 ° C. In order to remove heat from the end of the heat pipe cooler, radiation jacket cooling may preferably be used in a jacket with liquid refrigerant (such as water). Radiation heat transfer is proportional to the fourth degree of temperature, therefore, at the expected operating temperatures, high heat transfer rates will be possible. The heat pipe would be surrounded by an annular water heat exchanger capable of providing a stable working condition with a 15 kW tap. In other embodiments, the end-cooler of the heat pipe may be insulated with another material, such as stainless steel, and the outer surface of this material may be cooled with liquid refrigerant. Whatever technology is used, it is important that this heat pipe is not “damaged” by the refrigerant at the end of the refrigerator, i.e. so that more cooling does not occur on this end than on the evaporator end. This can seriously impair performance. In addition, if the temperature of the heat pipe falls below the freezing temperature of the working fluid at any point along the length (~ 180 ° C for lithium), then it will not work at all.

Limitations on the operating temperature of the cells near the base of the central electrode (anode) may require that the heat transferred to this area be minimized. This condition can be realized, for example, by coating the outside of the heat pipe with a material with low emissivity near the region of the lower temperature allowable limit. Then, between the heat pipe and the elements that must be at a reduced temperature, a vacuum gap can be made. Since vacuum has a very low thermal conductivity and the heat pipe is coated with a material with low emissivity, minimal heat transfer will occur between the heat pipe and cooler elements. Another condition that must be taken into account is maintaining an anode temperature at varying load power levels. This can be done by placing the cylinder between the heat pipe and the water-cooled outer jacket. This cylinder could be coated or finished with a material with high reflectivity in inner diameter and low emissivity in its outer diameter. If the cylinder is completely inserted between the radiating heat pipe and the water-cooled jacket, the radiation will be reflected back to the heat pipe, thereby reducing the power flow from the heat pipe to the jacket. When the “limiter” cylinder is removed, most of the heat pipe cooler can radiate directly into the heat exchanger of the water jacket. Therefore, by adjusting the position of the "limiter", the power flow is controlled to set the operating temperature of the steady state of the heat pipe and, ultimately, the anode.

A preferred embodiment in which a heat pipe is used for cooling is illustrated in FIG. 17, where anode 8A, cathode 8B and insulating element 9 are shown. In this case, lithium vapor is used as the active gas and it is supplied to the discharge chamber through the center of the anode 8A, as shown at 440. The anode 8A is cooled by a lithium heat pipe system 442 containing a lithium heat pipe 444. Lithium in the heat transfer region 446 of the heat pipe 444 is vaporized near the hot end of the electrode 8A and steam flows to the cold end of the heat pipe, where is transferred from the heat pipe by radiation cooling to a heat sink module 446 having a heat sink surface 448 cooled by a water coil 450. Cooling of lithium vapor causes a change in its phase state and transitions to a liquid state, and the liquid rises capillary back to the hot end due to capillary injection in accordance with a well-known method of operating a heat pipe. In this embodiment, the restriction cylinder 452 slides up and down, as shown at 454, inside the heat sink module 448 using an actuator that is part of a feedback temperature adjustment device not shown. The anode portion with the heat pipe also preferably includes an auxiliary heating system to maintain lithium at a temperature above the freezing point, when the device for creating a plasma pinch does not generate enough heat.

Fourth generation device

FIG. 16 shows a cross-sectional drawing of a prototype fourth-generation plasma pinch device 400 manufactured and tested by the applicants. Figa is an enlarged part of the device, illustrating in more detail the pinch region 401. Fig.16B is an electrical diagram representing important electrical elements of the excitation system with a high voltage power source for this embodiment. This device creates a plasma pinch at a pulse repetition rate up to about 2 kHz. The electrical energy released during the discharge between the electrodes is about 12 J per pulse. The applicants estimated that the useful light energy generated by each pinch in the VUV range of interest, radiated into an angle of 2π steradian, has a magnitude of about 20 mJ.

Essentially all of the elements shown in FIG. 16 are part of a solid-state pulsed power supply system 404 for feeding electrodes with discharge electric pulses. In this embodiment, a positive voltage pulse of about 4-5 kV is supplied to the central anode 8A. Cathode 8B has ground potential. Preionization is provided using 8 spark pins 138, which produce spark discharges of preionization in the lower part of the space between the cathode and anode. These spark pins operate at 20 kV using a sine wave generator (not shown) as a 30 kV 10 MHz power source.

Electric circuit

A description of the electrical circuit of this preferred switching power system is described below with reference to FIG. 16B and partially to FIGS. 16 and 16A.

Plasma Focus Power System

A conventional DC power supply of approximately 700 V is used to convert variable electrical power from a three-phase power supply of a 208 V public utility to approximately 700 V to 50 A DC. This power supply 400 provides power for the resonant charger 402. Power supply 400 charges a large 1500 uF C-1 filter capacitor bank. Based on a signal from an external trigger, the resonant charger initiates a charging cycle by closing the controlled charging switch S1. When the switch is closed, a resonant circuit is formed from the capacitor C-1, the charging induction coil L1 and the capacitor bank C0, which forms part of the solid-state switching power supply system (SSPPS) 404. Therefore, the current begins to “discharge” from C-1 through the inductor L1 in C0 charging this capacity. Since the capacitance C-1 is many, much larger than the capacitance C0, the voltage at C0 can reach a value approximately 2 times greater than the initial voltage at C-1 during this resonant charging process. The shape of the impulse of the charging current is assumed to be in the form of a half sine wave, and the voltage at C0 is similar to the “1-cosine” waveform.

In order to control the final voltage at C0, several actions can be performed. Firstly, the managed charging switch S1 can be opened at any time during a normal charging cycle. In this case, the current stops flowing from C-1, but the current that is already on the charging inductor continues to flow in C0 through the "free rotation" diode D3. This leads to the cessation of further energy transfer from C-1 to C0. Only the energy that remains in the charging inductor L1 (which can be substantial) continues to be transmitted to C0 and charge it to a higher voltage.

In addition, the S2 de-qing switch can be closed through the charging inductor, shorting the charging inductor and the de-qing resonant circuit. This essentially eliminates the inductor from the resonant circuit and prevents any further supply of current to the inductor from the C0 that continues to be charged. Then, the current in the inductor is shunted from the load and captured in the circuit formed by the charging inductor L1, switch S2 de-qing and diode D4 de-qing. The D4 diode is included in the circuit because the IGBT has a reverse antiparallel diode included in a device that would normally conduct reverse current. As a result, diode D4 blocks this reverse current, which could otherwise bypass the charging inductance during the charging cycle.

In conclusion, a bleed switch or a shunt switch and a series resistor (both not shown in this preferred embodiment) can be used to discharge energy from C0 as soon as the charging cycle is complete to achieve very fine voltage control at C0.

The DC power source is a power source with an adjustable output voltage: 800 V, 50 A DC; input 208 V, 90 A AC. Such a power source is provided by companies such as Universal Voltronics, Lambda / EMI, Kaiser Systems, Sorensen, etc. In the second embodiment, several power sources can be used, which are lower power sources connected in series and / or in parallel, in order to to provide full voltage, current, and average power that meet the requirements for this system.

Capacitor C-1 is two electrolytic capacitors, 450 V DC, 3100 uF, connected in series. The resulting capacitance is 1500 μF, rated at 900 V, providing a sufficient margin in the normal operating range of 700-800 V. These capacitors can be obtained from suppliers, such as Sprague, Mallory, Aerovox, etc.

The managed charging switch S1 and the output serial switch S3 in this embodiment are IGBT switches (1200 V, 300 A). The actual model number for these SM300NA-24H switches is from Powerex. The S2 de-qing switch is a 1700 V, 400 A IGBT switch, also from Powerex, model number SM400NA-34N.

Inductance L1 Charging Coil - a conventionally manufactured inductor made with 2 rows of parallel windings (each 20 turns) of a high-frequency stranded winding wire made on a toroidal core wrapped with 50-50% NiFe tape, with two 1/8 inch air gaps (3.175 mm) and a resulting inductance of approximately 140 μH.

This special core is made by National Arnold. In other embodiments, a variety of magnetic materials may be used for the core, including molypermal, Metglas ™, and others.

All diodes - serial, de-qing and "free-wheeling" - are 1400 V, 300 A diodes from Powerex, model number R6221430PS.

As mentioned above, the SSPPS scheme is similar to the one referred to in the description of the prior art. As soon as the resonant charger 402 charges C0, a trigger signal is generated using a control unit (not shown) in the resonant charger, which initiates the closure of the IGBT switches S4. Although the diagram shows only one such switch (for clarity of the circuit), there are eight parallel S4 IGBT switches that are used to discharge C0 into C1. Then the current from the capacitors C0 is discharged through the IGBT and into the first electromagnetic switch LS1. The design of this electromagnetic switch provides a sufficient “volt-second” value so that all 8 parallel IGBT switches are fully turned on (ie, closed) before the current rises significantly in the discharge circuit. After the closure, the main current pulse is formed, and it is used to transfer energy from C0 to C1. The transfer time from C0 to C1 is usually of the order of 5 μs, while the inductance LS1 at saturation is approximately 230 nH. As the voltage on C1 rises to the full required voltage value, the “volt-seconds” value on the second electromagnetic switch LS2 drops and this switch is in a saturated state, transferring energy from C1 to 1: 4 pulse transformer 406, which is described below in details. The transformer mainly consists of three single-turn primary "windings" connected in parallel, and one secondary "winding". The secondary conductor is connected to the high-voltage output of the primary, this leads to the fact that the ratio of the stepwise increase becomes 1: 4 instead of 1: 3, as in the configuration of the autotransformer. The secondary “winding” is then connected to a capacitor bank C2, which is then charged by transferring energy from C1 (via a pulse transformer). The transfer time from C1 to C2 is approximately 500 ns, while the inductance LS2 at saturation is approximately 2.3 nH. As the voltage rises by C2, the required volt-second product of the third electromagnetic switch LS3 is reached and it also saturates by transferring the voltage available on C2 to the anode 8a, as shown in FIGS. 14A and 14B. Inductance LS3 at saturation of approximately 1.5 nH.

The fourth electromagnetic switch is provided as a protective device in case the PPF does not function properly. If the preionization pulse is not supplied at the required time (just before the main pulse), then the voltage value of the main pulse is insufficient to breakdown the insulator between the anode and cathode. As a result, the pulse voltage under these open circuit conditions can essentially lead to an undesirable breakdown in the installation at some other place, and not between the PPF electrodes. In this case, most of the energy is then reflected back to the "front end" of the SSPPS. Such a large reverse voltage pulse can cause an avalanche breakdown of the serial diode in SSPPS, leading to possible damage or destruction of devices. This fourth electromagnetic switch is designed so that the product of "volt-seconds" is redundant if there is no breakdown between the main electrodes of the PPF. In this case, the electromagnetic switch is designed to short-circuit the load before voltage doubling occurs and significant damage to the device occurs. The saturation inductance of the fourth electromagnetic switch LS4 is approximately 22 nH and it ends with a parallel load RL with a resistance of ~ 1.5 Ohms and an inductance of ~ 75 μH.

The bias circuit shown in FIG. 16B at 408 is also used to properly bias the four electromagnetic switches. The current from the bias power supply V1 passes through the electromagnetic switches LS4 and LS3. Then it is separated and part of the current passes through the bias circuit inductance coil L5 and returns back to the bias power supply V1. The remaining part of the current passes through the secondary winding of the pulse transformer and then through the electromagnetic switches LS2 and LS1 and the magnetization inductance coil L3 to the magnetization power supply V1. The bias circuit inductance coil L2 provides a path back to the power source for the current that passes through the primary winding of the pulse transformer to earth. The bias circuit inductors L3, L5 also provide insulation voltage during the pulse in SSPPS, because the bias power supply V1 works near the ground potential (as opposed to the potentials generated in SSPPS, where the bias connections are made).

Capacities C0, C1 and C2 are created from a series of parallel, polypropylene film capacitors mounted on a printed circuit board with a thick (6-10 oz.) Copper coating. The printed circuit boards are wedge-shaped, so that the 4 boards form a cylindrical capacitor board, which is fed from the cylindrical bus for both high voltage and ground connections. In this embodiment, a low inductance compound is formed, which is important both for pulsed compression and for the stability of the plasma pinch in the PPF itself. The total capacitance of each of C0 and C1 is 21.6 microfarads, and the total capacitance of C2 is 1.33 microfarads. Capacitors C0 and C1 are capacitors of 0.1 uF, 1600 V, obtained from suppliers, such as, for example, Wima in Germany or Vishay Roederstein in North Carolina. Capacitance C2 consists of three sections of capacitors assembled in series to obtain the full rated voltage, since the voltage on the secondary winding of a pulse transformer is ~ 5 kV. Capacitors C2 are elements with parameters: 0.01 μF, 2000 V DC, obtained again from Wima or Vishay Roederstein.

SSPPS switches are IGBT 1400 V, 1000 A. switches. The actual model number is SM1000NA-28 from Powerex. As mentioned earlier, 8 parallel IGBT switches are used to discharge C0 in C1.

Serial SSPPS diodes are all 1400 V, 300 A diodes from Powerex, model number R6221430. Two diodes are used for each IGBT switch, resulting in a total of sixteen parallel devices.

The LS1 solenoid switch is a conventionally manufactured inductor with 16 rows of parallel windings (6 turns each) of a high-frequency stranded winding wire made on a toroidal ferrite core. This particular core is from Ceremic Magnetics of New Jersey and is made of CN-20 ferrite material. Top thickness is 0.5 inches (1.27 cm), inner diameter is 5.0 inches (7.54 cm) and outer diameter is 8.0 inches (20.32 cm).

The LS2 solenoid switch is a single-turn toroidal inductor. A magnetic core is a tape wound on a mandrel with an outer diameter of 8.875 inches (~ 22.54 cm) using a 2605-S3A Metglas ™ 2 inches wide (~ 5 cm), 0.7 mils (17.78 microns) thick by Honeyweil with Mylar wound with a thickness of 0.1 mil (2.54 μm) between the layers to an outer diameter of 10.94 inches (~ 27.8 cm).

The LS3 solenoid switch is also a single-turn toroidal inductor. A magnetic core is a tape wound around a mandrel with an outer diameter of 9.5 inches (24.13 cm) using a 2605-S3A Metglas ™ 1 inch wide (2.54 cm) and 0.7 mils (17.78 μm) thick Honeywell with mylar wound 0.1 mils (2.54 μm) thick between layers to an outer diameter of 10.94 inches (~ 27.8 cm).

The pulse transformer has a design similar to that described in US patent No. 5,936,988. Each of the three cores of the transformer is a tape wound around mandrel 422 with an outer diameter of 12.8 inches (~ 32.5 cm) using a 2605-S3A Metglas ™ 1 inch wide (2.54 cm) and 0.7 mil thick 17.78 μm) from Honeywell with mylar wound with a thickness of 0.1 mils (2.54 μm) between the layers to an outer diameter of 14.65 inches (~ 37.2 cm). Each of the three cores 418 has a ring shape with an inner diameter of 12.8 inches (~ 32.5 cm) and an outer diameter of about 14 inches (35.56 cm), a height of 1 inch (2.54 cm). On figs shows an axial section schematically showing the physical location of the three cores, as well as the primary and secondary "windings". Each primary winding is actually formed of two circular rings 420A and 420B bolted to the mandrel 422 and spacers 424 in the form of rods. The secondary "winding" consists of 48 bolts 426 mounted in a circle with gaps.

The transformer operates on a principle similar to that laid down in a linear accelerator. A high voltage current pulse in the three primary "windings" induces an increase in voltage in the secondary "winding" to approximately equal the primary voltage. As a result, the voltage generated on the secondary winding (i.e., rods 426) is equal to three times the primary voltage in the pulse. But, since the side of the secondary winding with a low voltage is connected to the primary windings, a four-fold conversion is performed.

The magnetization circuit inductors L3, L4 are toroidal inductors wound on a magnetic core made of molypermaloy. Specific core sizes: height 0.8 inches (~ 2 cm), inner diameter 3.094 inches (~ 7.86 cm) and outer diameter 5.218 inches (~ 13.25 cm). Core model number: 430026-2 from Group Arnold. An L3 inductor has 90 turns wound with 12 AWG wire over a torus to produce an inductance of ~ 7.3 mH, and L4 has 140 turns of 12 AWG wound around it to produce an inductance of ~ 18 mH.

The bias inductor L6 is just 16 turns of 12 AWG wire wound around a diameter of 6 inches (15.24 cm). The bias circuit L4 inductance coil is 30 turns of 12 AWG wire on a diameter of 6 inches (15.24 cm). The bias circuit inductance coil L2 is 8 turns of 12 AWG wire on a diameter of 6 inches (15.24 cm).

Resistance R1 is a group of twenty parallel resistors, each of which is a composite resistor, 27 Ohms.

Energy recovery

To increase full efficiency, this fourth-generation plasma pinch device provides for the regeneration of energy from pulse to pulse based on the electrical energy of the pulse reflected from the discharge part of the circuit. The energy recovery method used here is similar to that described in US Pat. No. 5,729,562, which is incorporated herein by reference. With reference to FIG. 16B, an explanation is given of how energy recovery is achieved.

After discharge, C2 is negatively excited. When this happens, LS2 is already saturated so that current flows from C1 to C2. Therefore, instead of the energy looping in the device (which leads to erosion of the electrodes), the saturated state LS2 is the reason that the reverse charge is transferred back to C2 resonantly to C1. This transfer is due to the ongoing forward current flow through LS2. After the charge is transferred from C2 to C1, the latter has a negative potential relative to C0 (which at this moment is approximately at the grounding potential) and (as was the case with LS2), LS1 continues to pass forward current due to the fact that during the pulse, which only what happened, a large current flowed. As a result, the current flows from C0 to C1, while the potential C1 rises to approximately the ground potential and a negative potential forms on C0.

Readers should pay attention to the fact that this reverse energy transfer back to C0 is possible only if all saturable core coils (LS1, LS2 and LS3) support direct conductivity until all or almost all of the energy returns to C0. After the unused energy passes back to C0, C0 becomes negative with respect to its initial stored charge. At this point, the switch 54 is opened by the switching power control device. Inverting the circuit containing the inductor L1 and the solid-state diode D3 connected to ground causes a change in the sign of the polarity of C0 as a result of the resonant "free rotation" (i.e., half the cycle of the circuit call L1-C0, as blocked against changing the direction of the current to the opposite in the inductor L1 due to the diode D3, leads to the total result, namely, that the energy is regenerated (captured) due to the partial "overdischarge" of C0. Thus, the energy, which in another case would contribute to the erosion of the electric delivery, is regenerated, reducing the requirements for charging the circuit for the next pulse.

Test results

On fig.16D and 16E shows the test results of a prototype device of the fourth generation. On fig.16D shows the pulse shape on the capacitor C2 and the voltage between the electrodes. 16E shows a signal measured by a photodiode using xenon as an active gas.

It is understood that the foregoing description of the embodiments is an illustration of only a few of the many possible specific variations that may constitute an application of the basic principles of the present invention. For example, instead of recirculating the working gas, it is preferably possible to simply collect lithium and remove helium. The use of a different electrode with other combinations of coatings instead of tungsten and silver is also possible. For example, copper or platinum electrodes and coatings would be workable. Other methods of generating a plasma pinch can be used for the specific options described. Some of these methods are described in the patents referred to in the description section on the background of the invention, and all of these descriptions are incorporated by reference in the present description. Many methods for generating high-frequency high-voltage electrical pulses are available, and they can be used. An alternative would be to keep the light guide at room temperature and therefore freeze both lithium and tungsten as they tend to propagate down the length of the light guide. This concept with freezing would further reduce the amount of contaminants that would reach the optical elements used in the lithographic instrument, since atoms would constantly adhere to the walls of the fiber upon impact. The deposition of the electrode material on the optics of a lithographic instrument could be prevented by the creation of collecting optics to form a repeated image of the radiation spot through a small hole in the primary discharge chamber and the use of a differential pump circuit. Helium and argon can be supplied from the second discharge chamber through an opening in the first chamber. It is shown that this scheme is effective for preventing the deposition of material on the output windows of copper vapor lasers. Lithium hydride can be used instead of lithium. The device can also work as a stationary filled system without flowing working gas through the electrodes. Of course, a very wide range of pulse repetition rates is possible from single pulses in the amount of 5 pulses per second to several hundred or thousands of pulses per second. If required, to adjust the position of solid lithium, the adjustment mechanism could be modified so that the position of the top of the central electrode is also adjusted taking into account erosion of the top.

Many other electrode arrangements are possible instead of those described above. For example, the external electrode may be conical in shape, rather than cylindrical, as shown, with a large diameter in the direction of the pinch. In addition, the operating parameters in some embodiments could be improved if the inner electrode could protrude beyond the end of the outer electrode. This can be done with spark pins or other preionizers well known in the art. Another preferred alternative is to use for the outer electrode a group of rods arranged so as to form a generally cylindrical or conical shape. This approach will help maintain a symmetrical pinch centered along the axis of the electrode due to the ballast inductance resulting from it.

Thus, the reader is invited to determine the scope of the invention by the attached claims and their legal equivalents, and not by the examples that are given in the description.

Claims (22)

1. A source of high-energy photons containing: A) a vacuum chamber; B) at least two electrodes mounted coaxially in said vacuum chamber defining an electric discharge region and configured to create high frequency plasma pinches at the pinch site during electric discharge; C) a working gas containing an active gas and a buffer gas, wherein the buffer gas is a noble gas and the active gas is selected so as to provide emission of at least one spectral line; D) an active gas supply system for supplying active gas to said discharge region; and E) a pulse power system providing electrical pulses with an energy of at least 12 J per pulse at a pulse repetition rate of at least 2000 Hz, said system comprising: 1) a battery of charging capacitors; 2) a high-speed charger for charging said battery of charging capacitors for periods of less than about 0.5 microseconds; 3) means for controlling the voltage of the charging capacitors to control the charge on the battery of charging capacitors; 4) an electromagnetic compression circuit comprising at least one capacitor bank and at least one saturable core inductor; 5) a charging capacitor battery trigger for discharging said charging capacitor bank into said electromagnetic compression circuit; and 6) a pulse transformer to increase the voltage of the pulses by at least a factor of 4, containing at least three single-turn primary windings and a single single-turn secondary winding. 2. The source of claim 1, wherein one of said two electrodes is a hollow anode, and said active gas is introduced into said vacuum chamber through said hollow anode. 3. The source of claim 2, wherein said active gas contains lithium. 4. The source of claim 2, wherein said active gas contains xenon. 5. The source of claim 2, wherein said at least one capacitor bank comprises at least one capacitor bank optimized to provide maximum current from the capacitor simultaneously with said plasma pinch. 6. The source according to claim 2, in which said hollow anode determines the end of the pinch, the first inner diameter near the specified end of the pinch and the second inner diameter, which is further from the specified end of the pinch than the first inner diameter, while the second inner diameter is larger than the first inner diameter. 7. The source according to claim 6, in which the first inner diameter is separated from the said pinch end by a distance selected to prevent the formation of an arc discharge. 8. The source according to claim 2, in which said anode determines the length of the open part of the anode, and this length is selected so that the plasma pinch occurs approximately simultaneously with the maximum value of the excitation current. 9. The source according to claim 2, additionally containing a source of lithium, consisting of a porous material impregnated with lithium. 10. The source of claim 9, wherein said porous material is tungsten. 11. The source of claim 10, further comprising an RF source configured to create a plasma surrounding at least a portion of said porous material. 12. The source of claim 1, wherein said high-speed charging system is a resonant charging system for charging said charging capacitor. 13. The source according to claim 1, wherein said at least one saturable core inductor is at least two saturable core inductors and a bias circuit for magnetizing said at least two saturable core inductors. 14. The source of claim 1, further comprising an energy recovery circuit for regenerating, on said charging capacitor, energy reflected from said electrodes. 15. The source of claim 1, wherein said at least one capacitor bank is at least two capacitor banks. 16. The source of claim 1, further comprising a heat pipe for cooling at least one of said electrodes. 17. The source according to claim 2, further comprising a cooling system with a heat pipe for cooling said hollow anode. 18. The source of claim 17, wherein said heat pipe cooling system and said hollow anode comprise a hollow anode cooled by a heat pipe having a portion for introducing said active gas. 19. The source of claim 18, wherein said pulse transformer consists of a plurality of ring-shaped cores made of magnetic material and a primary winding that is in electromagnetic communication with each of said cores. 20. The source according to claim 19, in which said magnetic material consists of a film with high magnetic permeability wound on a mandrel. 21. The source according to claim 20, in which said pulse transformer determines a secondary winding consisting of a plurality of rods. 22. The source of claim 20, wherein said mandrel forms part of said primary winding for each primary winding.

Images