LPP light source pulse width adjustable seed light source device for extreme ultraviolet lithography machine

By using an electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 laser and a high-voltage pulse signal tunable device, the pulse width of the light source device is continuously adjustable, which solves the problem of low light source conversion efficiency in lithography machines and improves the performance of extreme ultraviolet lithography machines.

CN118426267BActive Publication Date: 2026-06-12HARBIN INST OF TECH

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HARBIN INST OF TECH
Filing Date
2024-05-14
Publication Date
2026-06-12

Smart Images

  • Figure CN118426267B_ABST
    Figure CN118426267B_ABST
Patent Text Reader

Abstract

The application relates to an LPP light source pulse width tunable seed light source device for an extreme ultraviolet lithography machine, and relates to an LPP light source pulse width tunable seed light source device.The application comprises a group of electro-optic Q-switching radio frequency waveguide cavity emptying CO2 laser devices and a high-voltage pulse signal tunable device.The high-voltage pulse signal tunable device at least comprises a high-speed high-voltage switch circuit, the high-speed high-voltage switch circuit comprises a first main MOSFET M1 and a second main MOSFET M2;the drain electrode of the first main MOSFET M1 is connected with a positive power supply +HV, the source electrode of the second main MOSFET M2 is connected with the ground GND, and the source electrode of the first main MOSFET M1 is connected with the drain electrode of the second main MOSFET M2;the high-speed high-voltage switch circuit is used for generating a high-speed high-voltage pulse signal with a falling edge slope program that can be adjusted, and the high-speed high-voltage pulse signal is used for controlling a modulation crystal in the emptying cavity structure of the electro-optic Q-switching radio frequency waveguide cavity emptying CO2 laser device.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of seed light source technology for lithography machine light sources, and specifically relates to a tunable pulse width seed light source device for LPP light sources in lithography machines. Background Technology

[0002] EUV (Extreme Ultraviolet) lithography machines, indispensable for high-end semiconductor manufacturing, are a hot topic of international research. One of the core technologies of EUV lithography machines is the EUV lithography light source technology. This technology is monopolized by the Cymer team at ASML in the Netherlands. Currently, the main pulse laser pulse width parameters of the EUV lithography machine light source sold are on the order of tens of nanoseconds. The main pulse laser pulse width has a significant impact on the generation efficiency of the EUV lithography machine light source. Different sizes of Sn targets require different main pulse laser pulse widths to obtain the 13.5nm EUV light with optimal conversion efficiency. The main pulse laser pulse width is determined by the seed light.

[0003] Existing technologies can achieve controllable pulse widths in seed light source devices, but not continuously adjustable ones. Research has revealed that the conversion efficiency of CO2 lasers, the light source for lithography machines, is related to their pulse width. Traditional methods show that the inability to continuously tunable the pulse width affects the conversion efficiency of CO2 lasers. While this approach doesn't pose a significant problem when used as a light source for other devices, it fails to achieve the required conversion efficiency when used as an LPP light source in extreme ultraviolet lithography machines, severely hindering lithography machine development. Summary of the Invention

[0004] This invention addresses the problem that existing light source devices cannot achieve continuously adjustable pulse widths, resulting in a failure to meet the conversion efficiency requirements of LPP light sources in extreme ultraviolet lithography machines.

[0005] A pulse-width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine includes a set of electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 lasers. The device also includes a high-voltage pulse signal tunable device, which includes at least one high-speed high-voltage switching circuit. This high-speed high-voltage switching circuit includes a first main MOSFET M1 and a second main MOSFET M2. The drain of the first main MOSFET M1 is connected to the positive power supply +HV, and the source of the second main MOSFET M2 is connected to ground GND. The source of the first main MOSFET M1 is connected to the drain of the second main MOSFET M2. The high-speed high-voltage switching circuit generates a high-speed high-voltage pulse signal with a programmable falling edge slope. This high-speed high-voltage pulse signal is used to control the modulation crystal in the cavity-emptied structure of the electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 laser.

[0006] Furthermore, the electro-optic Q-switched radio frequency waveguide cavity-emptied CO2 laser adopts a dual-channel structure. One channel adopts the electrode structure and resonant cavity structure of a Z-folded cavity-emptied laser, and the other channel adopts a straight tube waveguide.

[0007] The electrode structure and resonant cavity structure of the Z-fold cavity empty laser are as follows:

[0008] The upper and lower aluminum electrodes press the ceramic sheet together to form a Z-fold structure. An electro-optic Q-switching device is installed in the resonant cavity of the Z-fold channel. Multiple resonant inductors are connected in parallel on the electrodes. The electro-optic Q-switching device includes a modulation crystal, a Brewster window, and a λ / 4 waveplate. The optical axis of the λ / 4 waveplate forms a 45° angle with the polarization direction determined by the Brewster window. The Brewster window is coated with a polarizing dielectric film. A total reflection mirror is installed at one end of the fold channel and at the fold. A fixed concave mirror with a focal length of f and a piezoelectric ceramic are installed at a distance f from the waveguide opening. The modulation crystal and the λ / 4 waveplate are placed between the Brewster window and the concave mirror.

[0009] The straight waveguide has a total reflection mirror at one end and a half reflection mirror at the other end.

[0010] Furthermore, the modulation crystal in the electro-optic Q-switching device is a CdTe modulation crystal.

[0011] Furthermore, the λ / 4 waveplate in the electro-optic Q-switching device is a CdS quarter-waveplate.

[0012] Furthermore, the total reflection mirror installed at one end of the folded channel and at the fold is a gold-plated copper reflection mirror.

[0013] Furthermore, the ceramic sheet in the Z-fold structure formed by pressing the upper and lower aluminum electrodes together with the ceramic sheet is an Al2O3 ceramic sheet.

[0014] Furthermore, each segment of the Z-folded channel has a discharge length of 430 mm and a folding angle of 4.5°.

[0015] Furthermore, the high-voltage pulse signal tunable device also includes a programmable multi-pulse drive high-speed high-voltage pulse generator; the programmable multi-pulse drive high-speed high-voltage pulse generator includes a logic control module and a MOSFET isolation drive module;

[0016] The logic control module is used to generate a sequence of wide and narrow driving pulses that meet the timing requirements, and to make the time interval between the narrow driving pulses adjustable.

[0017] The MOSFET isolation drive module is used to isolate the transmission drive pulse signal and provide a physical circuit for the charging and discharging of the input capacitors of the first main MOSFET M1 and the second main MOSFET M2.

[0018] Furthermore, the logic control module includes a microcontroller unit, n delay line chips, and a drive pulse generation circuit;

[0019] When n=3, the n delay line chips include the first delay line D1, the second delay line D2 and the third delay line D3;

[0020] The logic control module receives an externally input PWM signal, which is used as an external trigger signal to drive the pulse generation circuit. At the same time, the PWM signal is input to the drive pulse generation circuit through a series of first delay lines D1, second delay lines D2 and third delay lines D3 in sequence.

[0021] The PWM signal, the output signal of the first delay line D1, the output signal of the second delay line D2, and the output signal of the third delay line D3 are all used as trigger signals input to the drive pulse generation circuit. The drive pulse generation circuit detects the rising or falling edge of the trigger signal and sequentially generates drive wide pulse signals P1, P2, P3 and drive narrow pulse signals P4, P5, P6 with timing sequence.

[0022] Furthermore, the MOSFET isolation drive module includes a first transformer T1, a second transformer T2, a third transformer T3, a fourth transformer T4, a fifth transformer T5, a sixth transformer T6, a first MOSFET S1, a second MOSFET S2, a third MOSFET S3, a fourth MOSFET S4, a fifth MOSFET S5, and a sixth MOSFET S6;

[0023] The driving pulse P1 is isolated and transmitted through the first transformer T1 to control the first MOSFET S1 to turn on, providing a discharge circuit for the input capacitor of the second main MOSFET M2;

[0024] The driving pulse P2 is isolated and transmitted through the second transformer T2 to control the second MOSFET S2 to turn on, and provides a charging circuit for the input capacitor of the first main MOSFET M1;

[0025] The driving pulse P3 is transmitted through the isolation of the third transformer T3, which turns on the third MOSFET S3 and provides a discharge circuit for the input capacitor of the first main MOSFET M1.

[0026] The driving narrow pulse signals P4, P5, and P6 are transmitted in isolation through the fourth transformer T4, the fifth transformer T5, and the sixth transformer T6, respectively, to control the fourth MOSFET S4, the fifth MOSFET S5, and the sixth MOSFET S6 to turn on, thereby charging the input capacitor of the second main MOSFET M2 one by one and turning it on slowly.

[0027] The time interval between the driving narrow pulses P4, P5, and P6 is controlled by the microcontroller unit.

[0028] Beneficial effects:

[0029] This invention employs a dual-channel seed light source device. One channel utilizes the electrode and resonant cavity structures of a Z-folded cavity laser, while the other channel is a straight-tube waveguide output used for real-time monitoring of the laser cavity's operating status. It also provides a high-voltage pulse signal tunable device. This device includes a high-speed high-voltage switching circuit consisting of a first main MOSFET M1 and a second main MOSFET M2 to control the high-voltage signal from the CdTe modulation crystal. A programmable multi-pulse driven high-speed high-voltage pulse generator is applied to the series-connected MOSFETs. While the MOSFETs cannot achieve a complete turn-on transient in a single narrow driving pulse, they gradually turn on over multiple consecutive pulses, resulting in a slower descent rate. This achieves smooth and continuously programmable control of the descent rate of the Q-switched high-voltage pulse loaded on the CdTe crystal. Ultimately, the pulse width of the radio frequency excited waveguide carbon dioxide cavity emptied laser is continuously tunable, thus enabling it to be adapted to the expansion diameter D of different Sn droplets. Even without knowing the expansion diameter D, the laser duration can be continuously adjusted to achieve the adaptation to the expansion diameter D, which can effectively improve the conversion efficiency and achieve the conversion efficiency required by the LPP light source of the extreme ultraviolet lithography machine. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the LPP light source structure of an EUV lithography machine.

[0031] Figure 2 This is a structural diagram of a radio frequency waveguide cavity emptied carbon dioxide laser.

[0032] Figure 3 The timing relationship between Q-switched laser and cavity-emptied laser is shown in the diagram.

[0033] Figure 4 This is a graph showing the variation of the cavity emptying pulse width under the falling edge width of the high-voltage pulse.

[0034] Figure 5 This is a schematic diagram of a linearly adjustable signal on the falling edge.

[0035] Figure 6 This is a schematic diagram of the high-speed high-voltage pulse signal tunable device with adjustable falling edge slope in the embodiment.

[0036] Figure 7 This is a diagram showing the current flow of a high-speed, high-voltage pulse signal tunable device.

[0037] Figure 8 This is a diagram showing the current flow of a high-speed, high-voltage pulse signal tunable device.

[0038] Figure 9 This is a signal timing diagram for a high-speed high-voltage pulse signal tunable device with adjustable falling edge slope.

[0039] Figure 10 The graph shows the results of the Q-switched high-voltage signal (yellow) and the laser pulse width (blue). Detailed Implementation

[0040] This invention proposes a pulse-width tunable master pulse seed light source device, which will improve the power of EUV light generation with optimal conversion efficiency, enabling EUV lithography machines to meet commercial standards and providing important technical support for the development of EUV lithography machines. The invention will be described below with reference to specific embodiments. Specific implementation method one:

[0042] This embodiment describes a pulse-width-tunable seed light source device for an LPP light source in an extreme ultraviolet (EUV) lithography machine. Before proceeding with the specific details, the key technologies of extreme ultraviolet (EUV) lithography machines will first be explained. Currently, the light source used in EUV lithography machines is an LPP (Light Produced Plasma) light source, such as... Figure 1 The structure of the LPP lithography light source for EUV is shown in the figure. The structure of the light source includes the following components:

[0043] Main pulse seed light: A waveguide CO2 laser is used as the laser source. The continuous output is modulated into a high repetition frequency pulse output with a wavelength of 10.6μm by a modulation crystal CdTe. This part can be called seed light.

[0044] The main pulse amplifier typically consists of four or more stages. High-repetition-frequency (not less than 100kHz) CO2 pulses with a certain pulse width from the seed light output are injected into the CO2 amplifier to achieve power amplification.

[0045] Pre-pulse laser: A narrow pulse laser is used to ionize Sn droplets into low-valence vapor plasma through the shock wave effect.

[0046] Sn droplet generator: It liquefies metallic Sn and forms uniform, high-speed Sn droplets, which then enter the vacuum target chamber.

[0047] Vacuum Target Chamber: A vacuum environment is created within the target chamber. Sn droplets are irradiated with a pre-pulse to form a low-valence plasma, which is then irradiated with a 10.6 μm laser with a high repetition rate and high peak power output from a seed laser amplifier. Due to the thermal effect of the main pulse laser, the low-valence Sn plasma is further ionized into +8 to +14 valence Sn ions. These +8 to +14 valence Sn ions then transition to lower energy levels, generating 13.5 nm extreme ultraviolet light.

[0048] Optical system: The 13.5nm extreme ultraviolet light is collected and extracted by the optical system and then used for photolithography on the wafer, thus completing the current EUV LPP light source extreme ultraviolet photolithography process.

[0049] The LPP light source structure described above can generate 13.5nm extreme ultraviolet light. The principle behind the generation of 13.5nm EUV light is as follows:

[0050] In the physical process of amplifying a Sn droplet by a main pulse waveguide carbon dioxide laser, the Sn droplet is first bombarded into a vaporized low-valence state plasma by a narrow pulse laser pre-pulse. Then, the low-valence Sn plasma is ionized into high-valence Sn ions of 8-14 valence through thermal effect by CO2 main pulse irradiation, and the energy level transition radiates 13.5nm extreme ultraviolet light.

[0051] In the process of vaporizing low-valence Sn plasma irradiated by a high-power CO2 laser main pulse, the plasma density (expansion diameter D) varies depending on the timing of vaporization and expansion. When the laser aperture and power are constant, the laser emission timing (determining the expansion diameter D) and the laser duration (pulse width τ) significantly affect the conversion efficiency for generating 13.5 nm extreme ultraviolet light. The optimal conditions for obtaining the conversion efficiency are:

[0052]

[0053] Where c is the speed of light.

[0054] When the pulse width is greater than this value, laser energy is wasted, resulting in a decrease in conversion efficiency.

[0055] When the pulse width is less than this value, the low-valence Sn plasma does not fully participate in ionization, which will also cause a decrease in conversion efficiency.

[0056] Because the processes used in low-cost Sn plasma generation equipment from different manufacturers vary, the expansion diameter D of the generated Sn droplets also differs. Current technology cannot achieve continuously adjustable laser duration, thus failing to adequately match the expansion diameter D. Furthermore, since the expansion diameter D of Sn droplets generated by some manufacturers' low-cost Sn plasma generation equipment is unknown, it is even more difficult to achieve a good match between laser duration and expansion diameter D. Therefore, the pulse width of the seed light has a significant impact on the performance of the EUV lithography machine's light source, and the aforementioned situations all contribute to the inability to meet the requirements for extreme ultraviolet light sources. This invention achieves continuously adjustable laser duration by continuously adjusting the falling edge of the control pulse loaded onto the CdTe crystal. This not only allows for adaptation to the expansion diameter D of different Sn droplets, thereby improving conversion efficiency, but also allows for optimal matching of the expansion diameter D even when the expansion diameter D is unknown, achieving the highest conversion efficiency by continuously adjusting the laser pulse width (duration).

[0057] The pulse-width-tunable seed light device developed in this invention can serve as the basis for the light source of an EUV LPP lithography machine. The radio frequency waveguide CO2 laser used in this invention has a continuous output of 10.6 μm. Its continuous output is modulated into a high repetition rate pulse output using a CdTe modulation crystal. This embodiment provides a high-voltage pulse signal tunable device based on an electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 laser, thereby providing a pulse-width-tunable seed light source device.

[0058] The principle of the electro-optically Q-switched radio frequency waveguide cavity emptied CO2 laser is as follows:

[0059] A dual-channel design is employed, with one channel featuring the electrode and resonant cavity structure of a Z-folded cavity laser, and the other channel being a straight waveguide. Figure 2 As shown, one end of the straight waveguide is a total reflection mirror, and the other end is a partial reflection mirror. The electrode and resonant cavity structures of the Z-fold cavity-emptied laser adopt a cavity-emptied operating mode, with the Z-fold section serving as the master oscillator. The laser's electrode structure is as follows: upper and lower aluminum electrodes are pressed together with Al2O3 ceramic sheets to form a Z-fold structure. An electro-optic Q-switching device is inserted into the resonant cavity of the Z-fold channel. The waveguide dimensions are 2.25 × 2.25 mm. 2Each segment of the Z-folded channel has a discharge length of 430 mm and a folding angle of 4.5°. Multiple resonant inductors are connected in parallel on the electrodes, along with a matching circuit to ensure resonance at the RF source frequency and maintain uniform voltage distribution. A type-II waveguide resonant cavity structure is employed. The electro-optic Q-switching device includes a modulation crystal, a Brewster window, and a λ / 4 waveplate. The optical axis of the λ / 4 waveplate forms a 45° angle with the polarization direction determined by the Brewster window. The Brewster window is coated with a polarizing dielectric film, which provides high transmittance for lasers polarized in the p-direction determined by the incident plane and high reflectivity for lasers polarized in the s-direction. Total internal reflection mirrors are installed at one end of the folded channel and at the fold; in this embodiment, gold-plated copper mirrors with a reflectivity of 99% are used.

[0060] According to waveguide resonant cavity theory, a concave reflector with focal length f and a piezoelectric ceramic are selected and fixedly installed at a distance f from the waveguide opening, where f satisfies:

[0061]

[0062] Where ω0 = 0.7032a, a is the half-width of the square waveguide, and the coupling loss is minimized at this point.

[0063] In some embodiments, a ZnSe concave mirror with a radius of curvature of 370 mm (the concave mirror is coated with ZnSe) is selected, and the modulation crystal and λ / 4 wave plate are CdTe crystal and CdS quarter-wave plate. The CdTe crystal and CdS quarter-wave plate of the electro-optic Q-switching device are placed between the Brewster window and the concave mirror.

[0064] Analysis shows that when no quarter-wave voltage is applied to the crystal, the laser electrodes are constantly discharging, resulting in a much larger number of particles in the upper energy level than in the lower energy level. This leads to a low Q-value, severe losses, and the inability to generate laser oscillation. However, when a quarter-wave voltage is applied to the crystal, the laser's Q-value suddenly increases, rapidly generating laser oscillation within the laser cavity. This creates a Q-switched laser pulse within the cavity, and when the voltage is removed, cavity-emptied laser output can be obtained from the Brewster window side. The dual-channel laser polarization state is as follows: Figure 2 As shown.

[0065] For CO2 lasers with a pulse width of approximately 100 ns, the high repetition rate pulse output pulse width can be reduced to 15-20 ns after employing cavity emptying technology. It should be noted that these parameters are the basic parameters before using the high-voltage pulse signal tunable device of this invention.

[0066] The relationship between the Q-switched laser pulse settling time and the cavity-emptied CO2 laser pulse is as follows:

[0067] Theoretical analysis and calculation of the dynamics of the Q-switched CO2 laser yielded the rate equations for equation (1):

[0068]

[0069] In the formula, n represents the number of photons within one wavelength mode of the laser, where t represents time; J’ 00 0 The number of particles in a rotating energy level with rotational quantum number J' in energy level 1; n J” 10 0 The number of particles in the rotational energy level with rotational quantum number J” = J'+1 in energy level 0; n v’ For the upper energy level of laser vibration (00) 0 1) Number of particles; n v” For laser vibrational energy level (10) 0 0,02 0 0) Number of particles; represented by k J Let k represent the relaxation rate of the rotational energy level. J' Let k be the relaxation rate of the rotational energy level with rotational quantum number J'. J” k” represents the relaxation rate of the rotational energy level with rotational quantum number J” = J' + 1; k” represents the vibrational energy level (10 0 0,02 0 The relaxation rate of a particle with a velocity of 0 to other vibrational states.

[0070] In the equation above, all particle numbers and photon numbers are inverted above the energy level threshold. As a unit of measurement, all rates are expressed in cavity lifetime t. c It is a unit of measurement.

[0071] Using a combination of experimental and theoretical methods, the timing relationship between electro-optic Q-switched laser output and cavity-emptied laser output was analyzed and measured. Theoretically, the timing relationship between quarter-wave voltage, electro-optic Q-switched laser, and cavity-emptied laser can be obtained as follows: Figure 3 As shown, when a quarter-wave voltage is applied to the CdTe modulation crystal, the Q-value of the laser resonator rapidly increases from low to high, quickly establishing laser oscillation and forming a Q-switched laser pulse. Since the transmittance of the ZnSe concave mirror (coated with ZnSe) is only 1%, the light intensity inside the cavity remains high and continues to oscillate. As the quarter-wave voltage decreases on the crystal, the p-polarized light inside the resonator passes through the crystal twice (once in the forward direction and once after reflection) and becomes s-beam. The s-beam exits the cavity through the Brewster window, resulting in the highest cavity-emptied laser pulse power. The width of the Q-switching voltage can be adjusted by changing the width of the trigger signal.

[0072] The process for adjusting the pulse width of a cavity-emptying CO2 laser is as follows:

[0073] Cavity emptying utilizes the electro-optic effect of the electro-optic modulator crystal to change the polarization direction of light, and the aforementioned properties of the Brewster window to achieve pulsed laser output, typically achieving narrow pulse widths of tens or even hundreds of nanoseconds. Cavity emptying based on voltage-applied electro-optic Q-switching is called de-voltage pulsed cavity emptying, and cavity emptying based on de-voltaged electro-optic Q-switching is called voltage-applied pulsed cavity emptying. The former occurs during the time when the voltage abruptly changes from high voltage to 0, while the latter occurs within the time range of voltage rising from 0 to the required high voltage. During the high voltage change, the electro-optic effect of the electro-optic crystal changes, corresponding to a change in the polarization direction of light within the resonant cavity. The intensity of light reflected and transmitted through the Brewster window changes, which can be viewed as a change in the reflectivity of the Brewster window, and the corresponding function can be equivalently represented by the following equation:

[0074]

[0075] In the formula, τ d This represents the time of Vt's descent, with the zero point of time t taken at the starting point of Vt's descent.

[0076] The optical length of the resonant cavity is

[0077] L=l+(n a -1)d a +(n c -1)d c +(n b -1)d b (3)

[0078] Where l is the distance from the total reflection mirror to the output mirror, and n a ,n c ,n b d represents the refractive index of the Brewster window, the crystal, and the waveplate, respectively; a ,d c ,d b These represent the length (thickness) of the beam passing through the Brewster window, the crystal, and the waveplate, respectively.

[0079] Based on the formula here and the purpose of continuous adjustment that this invention ultimately aims to achieve, the final designed Z-fold channel has a discharge length of 430mm for each segment.

[0080] The time τ required for light to travel one round trip within the cavity L =2L / C, assuming that during the cavity emptying process, the gain and loss of light traveling one round trip within the cavity are equal, then the time characteristic of the output laser power P can be expressed as:

[0081]

[0082] Where P0 represents the power achieved when a quarter-wave voltage is applied to the crystal, at which point the total loss in the cavity is minimized and laser oscillation is formed.

[0083] The first five equations in formula (4) each represent the laser power of p-direction polarized light passing through the Bucher window, circulating once within the cavity, partially becoming s-direction polarized light, and then reflected back from the Bucher window. The sixth equation represents the laser power during the fifth cycle, at τ... d -4τ L The laser power that passes through the Brewster window and circulates once inside the cavity within a certain time, and is then completely reflected from the Brewster window.

[0084] The laser waveform after cavity emptying can be calculated from the above formula, and it can be seen that the de-voltage process determines the pulse width of the cavity emptying laser. The slope of the falling edge (i.e., the speed of the falling edge of the Q-switched pulse signal) has a significant impact on the pulse width of the seed light of the main pulse.

[0085] like Figure 4 As shown, the laser pulse width increases accordingly with the increase of the high-voltage falling edge width. The traditional method involves connecting a resistor in series with the CdTe crystal, forming an RC circuit with the high-voltage power supply load. After passing through the RC circuit, the falling edge of the high-voltage pulse is extended. Subsequently, the high-voltage falling edge width is fixed, and the laser pulse width is also fixed. However, because this method involves a relatively large resistor value and poor linearity, it cannot achieve continuously tunable pulse width.

[0086] This invention has revealed that the conversion efficiency of a CO2 laser is related to its pulse width. Based on the aforementioned traditional methods, it is known that the inability to continuously tunably adjust the pulse width affects the conversion efficiency of the CO2 laser. While this approach is not a major problem when used as a light source for other devices, it fails to achieve the required conversion efficiency when used as an LPP light source in extreme ultraviolet lithography machines, severely hindering the development of these machines. To address the issue of continuously tunable pulse width and thus improve conversion efficiency, this invention proposes the following... Figure 5 The example shown is a linearly adjustable falling edge device. Based on the aforementioned electro-optically Q-switched RF waveguide cavity-emptied CO2 laser, this embodiment provides a high-voltage pulse signal tunable device, which together constitutes a pulse-width tunable seed light source device, specifically a pulse-width tunable seed light source device for an extreme ultraviolet lithography LPP light source. The high-voltage pulse signal tunable device is as follows... Figure 5As shown, the high-voltage pulse signal tunable device is used to control the high-voltage signal of the modulated crystal CdTe. Improvements are made to the control of the high-voltage signal to the crystal: a smooth, continuously controllable falling edge of the crystal high-voltage signal is obtained using a multi-pulse signal. For the programmable control of the multi-pulse generation signal, the smooth falling edge of the crystal high-voltage signal is continuously adjustable through a MOSFET, thereby achieving continuous controllability of the pulse width. Therefore, the method applied in this invention can make the high-voltage falling edge continuously controllable by program modulation, and the linearity is greatly improved. Finally, cavity-empty CO2 laser output with continuously tunable pulse width can be obtained through program control.

[0087] The high-voltage pulse signal tunable device includes at least a high-speed high-voltage switching circuit, and may also include a programmable multi-pulse drive high-speed high-voltage pulse generator.

[0088] The high-speed high-voltage switching circuit includes a first main MOSFET M1 and a second main MOSFET M2; the drain of the first main MOSFET M1 is connected to the positive power supply +HV, the source of the second main MOSFET M2 is connected to ground GND, and the source of the first main MOSFET M1 is connected to the drain of the second main MOSFET M2; the high-speed high-voltage switching circuit is used to generate a high-speed high-voltage pulse signal with an adjustable falling edge slope.

[0089] The programmable multi-pulse drive high-speed high-voltage pulse generator includes a logic control module and a MOSFET isolation drive module;

[0090] The logic control module is used to generate a sequence of wide and narrow driving pulses that meet the timing requirements, and to make the time interval between the narrow driving pulses adjustable.

[0091] The MOSFET isolation drive circuit is used to isolate the transmission drive pulse signal and provide a physical circuit for the charging and discharging of the input capacitors of the first main MOSFET M1 and the second main MOSFET M2.

[0092] The high-speed high-voltage pulse generation process with adjustable falling edge slope in the above device is as follows: two MOSFETs, the first main MOSFET M1 and the second main MOSFET M2, connected in series, are used as switching devices to generate high-speed high-voltage pulses. n series-connected delay line chips generate n driving narrow pulses to charge the input capacitor of the second main MOSFET M2 successively, where n≥3, so that the gate voltage of the second main MOSFET M2 rises slowly. By adjusting the time interval between any two adjacent delay line chips, the time interval between the driving narrow pulses is programmably adjustable, thereby controlling the rising speed of the gate voltage of the second main MOSFET M2, and thus controlling the falling edge slope of the drain-source voltage curve of the second main MOSFET M2, so as to realize the adjustable falling edge slope of the high-speed high-voltage pulse signal output by the switching device.

[0093] The core technology of this invention lies in using multiple driving narrow pulses to charge the input capacitor of the second main MOSFET M2 successively, so that the gate voltage of the second main MOSFET M2 rises slowly, thereby widening the width of the falling edge of the high-speed high-voltage pulse signal. Furthermore, the time interval of the driving narrow pulses can be controlled by a program. By adjusting the time interval, the speed at which the gate voltage of the second main MOSFET M2 rises can be controlled, which in turn controls the slope of the falling edge of the high-speed high-voltage pulse signal.

[0094] Example

[0095] by Figures 6 to 9 The high-voltage pulse signal tunable device shown is used as an example for illustration. The high-voltage pulse signal tunable device includes a logic control module, a MOSFET isolation drive module, and a high-speed high-voltage switching circuit;

[0096] The logic control module is used to generate a sequence of wide and narrow driving pulses that meet the timing requirements, and to make the time interval between the narrow driving pulses adjustable.

[0097] The MOSFET isolation drive circuit is used to isolate the transmission of drive pulse signals and to provide a physical circuit for the charging and discharging of the input capacitors of the first main MOSFET M1 and the second main MOSFET M2.

[0098] The high-speed high-voltage switching circuit is used to generate a high-speed high-voltage pulse signal with an adjustable falling edge slope.

[0099] The logic control module includes a microcontroller unit, n delay line chips, and a drive pulse generation circuit.

[0100] In this embodiment, n=3 is taken as an example. The n delay line chips include the first delay line D1, the second delay line D2 and the third delay line D3. When the number of driving narrow pulses increases, the corresponding pulse width and pulse amplitude need to be reduced.

[0101] The logic control module receives an externally input PWM signal, which is used as an external trigger signal to drive the pulse generation circuit. At the same time, the PWM signal is input to the drive pulse generation circuit through a series of first delay lines D1, second delay lines D2 and third delay lines D3 in sequence.

[0102] The PWM signal, the output signal of the first delay line D1, the output signal of the second delay line D2, and the output signal of the third delay line D3 are all used as trigger signals input to the drive pulse generation circuit. The drive pulse generation circuit detects the rising or falling edge of the trigger signal and sequentially generates drive wide pulse signals P1, P2, P3 and drive narrow pulse signals P4, P5, P6 with timing sequence.

[0103] The high-speed high-voltage switching circuit includes a first main MOSFET M1 and a second main MOSFET M2; the drain of the first main MOSFET M1 is connected to the positive power supply +HV, the source of the second main MOSFET M2 is connected to ground GND, the source of the first main MOSFET M1 is connected to the drain of the second main MOSFET M2, and the connection point outputs a high-speed high-voltage pulse signal with an adjustable falling edge slope.

[0104] The MOSFET isolation drive module includes a first transformer T1, a second transformer T2, a third transformer T3, a fourth transformer T4, a fifth transformer T5, a sixth transformer T6, a first MOSFET S1, a second MOSFET S2, a third MOSFET S3, a fourth MOSFET S4, a fifth MOSFET S5, and a sixth MOSFET S6;

[0105] The driving pulse P1 is isolated and transmitted through the first transformer T1 to control the first MOSFET S1 to turn on, providing a discharge circuit for the input capacitor of the second main MOSFET M2;

[0106] The driving pulse P2 is isolated and transmitted through the second transformer T2 to control the second MOSFET S2 to turn on, and provides a charging circuit for the input capacitor of the first main MOSFET M1;

[0107] The driving pulse P3 is transmitted through the isolation of the third transformer T3, which turns on the third MOSFET S3 and provides a discharge circuit for the input capacitor of the first main MOSFET M1.

[0108] The driving narrow pulse signals P4, P5, and P6 are transmitted in isolation through the fourth transformer T4, the fifth transformer T5, and the sixth transformer T6, respectively, to control the fourth MOSFET S4, the fifth MOSFET S5, and the sixth MOSFET S6 to turn on, thereby charging the input capacitor of the second main MOSFET M2 one by one and turning it on slowly.

[0109] The time interval between the driving narrow pulses P4, P5, and P6 is controlled by the microcontroller unit.

[0110] Regarding the types of wide and narrow driving pulses: Wide driving pulse signals P1 and P3 are positive pulses, used to turn off the second main MOSFET M2 and the first main MOSFET M1, respectively; wide driving pulse P2 is a negative pulse, used to turn on the first main MOSFET M1; narrow driving pulse signals P4, P5, and P6 are all negative pulses, used to slowly turn on the second main MOSFET M2, realizing the adjustable falling edge slope of the high-speed high-voltage pulse signal.

[0111] The timing sequence of the driving wide pulse signals P1, P2, P3 and the driving narrow pulse signals P4, P5, P6, and the generation process of the high-speed high-voltage pulse signal are as follows:

[0112] When the drive pulse generation circuit detects the rising edge of the PWM signal, it generates drive wide pulse signals P1 and P2 before and after it. When it detects the falling edge of the PWM signal, it generates drive wide pulse signal P3, and then generates drive narrow pulse signals P4, P5, and P6 sequentially based on the delay time of the three delay line chips.

[0113] After the PWM signal is input as an external trigger signal, it is divided into two paths. One path is input to the first delay line D1, and the other path is directly input to the drive pulse generation circuit. A drive pulse P1 is generated at the rising edge of the PWM signal and input to the MOSFET isolation drive module. After being isolated and transmitted through the first transformer T1, the first MOSFET S1 is quickly turned on, causing the second main MOSFET M2 to be turned off.

[0114] Following this, the drive pulse generation circuit generates another drive wide pulse P2, which is input to the MOSFET isolated drive module. After being isolated and transmitted through the second transformer T2, the second MOSFET S2 is quickly turned on, which turns on the first main MOSFET M1. The output high-speed high-voltage pulse signal with an adjustable falling edge slope rises to +HV.

[0115] The falling edge of the PWM signal triggers the drive pulse generation circuit to generate a drive pulse P3, which is input to the MOSFET isolation drive module and then transmitted through the isolation of the third transformer T3. The third MOSFET S3 is quickly turned on, causing the first main MOSFET M1 to turn off.

[0116] Meanwhile, the falling edge of the output signals from the first delay line D1, the second delay line D2, and the third delay line D3 all trigger the drive pulse generation circuit, generating three drive narrow pulses P4, P5, and P6. The time interval between the three drive narrow pulses is controlled by the microcontroller and is input to the MOSFET isolation drive module. After being isolated and transmitted through the fourth transformer T4, the fifth transformer T5, and the sixth transformer T6, the fourth MOSFET S4, the fifth MOSFET S5, and the sixth MOSFET S6 are quickly turned on, causing the second main MOSFET M2 to slowly turn on, thus realizing the programmable adjustment of the falling edge slope of the output high-speed high-voltage pulse signal.

[0117] The first main MOSFET M1 and the second main MOSFET M2 are each composed of at least one MOSFET. If they are composed of multiple MOSFETs, they are connected in series to increase the maximum amplitude of the output signal.

[0118] The voltage difference between the positive power supply +HV and the ground GND is any value within the range of the sum of the drain-source voltages of the first main MOSFET M1 and the second main MOSFET M2.

[0119] The pulse width of the high-speed high-voltage pulse signal with adjustable falling edge slope is equal to the difference between the arrival times of the wide driving pulse at the first transformer T1 and the narrow driving pulse at the fourth transformer T4.

[0120] For more specific details on the circuit connections, please refer to [link / reference]. Figure 6 It includes a logic control module, a MOSFET isolation drive circuit, and a high-speed high-voltage switching circuit;

[0121] The logic control module includes a microcontroller unit, a first delay line D1, a second delay line D2, a third delay line D3, and a drive pulse generation circuit;

[0122] The logic control module receives an externally input PWM signal. The PWM signal is simultaneously input to the first delay line D1 as an external trigger signal. The output terminal of the first delay line D1 is connected to the input terminal of the second delay line D2, and the output terminal of the second delay line D2 is connected to the input terminal of the third delay line D3. The PWM signal, the output signal of the first delay line D1, the output signal of the second delay line D2, and the output signal of the third delay line D3 are all input to the drive pulse generation circuit as trigger signals. The drive pulse generation circuit detects the rising edge or falling edge of the trigger signal and generates a timing-sequential drive wide pulse signal and a drive narrow pulse signal.

[0123] The MOSFET isolation drive module includes a first transformer T1, a second transformer T2, a third transformer T3, a fourth transformer T4, a fifth transformer T5, a sixth transformer T6, a first MOSFET S1, a second MOSFET S2, a third MOSFET S3, a fourth MOSFET S4, a fifth MOSFET S5, and a sixth MOSFET S6;

[0124] The driving wide pulse signals are respectively input to the primary terminals of the first transformer T1, the second transformer T2, and the third transformer T3. The secondary terminal of the first transformer T1 is connected to the gate of the first MOSFET S1, and the secondary terminal of the first transformer T1 is connected to the source of the first MOSFET S1. The secondary terminal of the second transformer T2 is connected to the gate of the second MOSFET S2, and the secondary terminal of the second transformer T2 is connected to the source of the second MOSFET S2. The secondary terminal of the third transformer T3 is connected to the gate of the third MOSFET S3, and the secondary terminal of the third transformer T3 is connected to the source of the third MOSFET S3. The gate of the first MOSFET S1 is connected to the source of the second main MOSFET M2, and the source of the first MOSFET S1 is connected to the gate of the first main MOSFET M1. The gate of the second MOSFET S2 is connected to the source of the second main MOSFET M1, and the source of the second MOSFET S2 is connected to the gate of the first main MOSFET M1. The gate of the third MOSFET S3 is connected to the source of the first main MOSFET M1. The source of S3 is connected to the gate of the first main MOSFET M1;

[0125] The driving narrow pulse signals are respectively input to the primary terminals of the fourth transformer T4, the fifth transformer T5, and the sixth transformer T6. The secondary terminal of the fourth transformer T4 is connected to the gate of the fourth MOSFET S4, and the secondary terminal of the fourth transformer T4 is connected to the source of the fourth MOSFET S4. The secondary terminal of the fifth transformer T5 is connected to the gate of the fifth MOSFET S5, and the secondary terminal of the fifth transformer T5 is connected to the source of the fifth MOSFET S5. The secondary terminal of the sixth transformer T6 is connected to the gate of the sixth MOSFET S6, and the secondary terminal of the sixth transformer T6 is connected to the source of the sixth MOSFET S6. The gate of the fourth MOSFET S4 is connected to the source of the second main MOSFET M2, and the source of the fourth MOSFET S4 is connected to the gate of the second main MOSFET M2. The gate of the fifth MOSFET S5 is connected to the source of the second main MOSFET M2, and the source of the fifth MOSFET S5 is connected to the gate of the second main MOSFET M2. The gate of the sixth MOSFET S6 is connected to the source of the second main MOSFET M2. The source of S6 is connected to the gate of the second main MOSFET M2;

[0126] The high-speed high-voltage switching circuit includes a first main MOSFET M1 and a second main MOSFET M2;

[0127] The drain of the first main MOSFET M1 is connected to the positive power supply +HV, the source of the second main MOSFET M2 is connected to ground GND, and the source of the first main MOSFET M1 is connected to the drain of the second main MOSFET M2.

[0128] Work process:

[0129] At the rising edge of the input PWM signal, refer to Figure 2 The system generates two wide-range driving pulses, P1 and P2, sequentially. P1 is transmitted through isolation via the first transformer T1, turning on the first MOSFET S1 and providing a discharge path for the input capacitor of the second main MOSFET M2, causing M2 to quickly turn off. P2 is transmitted through isolation via the second transformer T2, turning on the second MOSFET S2 and charging the input capacitor of the first main MOSFET M1, causing it to quickly turn on as well. At this point, the output signal amplitude rises to +HV.

[0130] At the falling edge of the input PWM signal, refer to Figure 3 The system generates a wide driving pulse P3 and narrow driving pulses P4, P5, and P6 sequentially. The wide driving pulse P3 is transmitted through isolation via the third transformer T3, turning on the third MOSFET S3 and providing a discharge circuit for the input capacitor of the first main MOSFET M1, causing M1 to quickly turn off. The narrow driving pulses P4, P5, and P6 are transmitted through isolation via the fourth transformer T4, fifth transformer T5, and sixth transformer T6, respectively, turning on the fourth MOSFET S4, fifth MOSFET S5, and sixth MOSFET S6, successively charging the input capacitor of the second main MOSFET M2 and causing it to slowly turn on. The time interval between the narrow driving pulses P4, P5, and P6 is program-controlled by the microcontroller unit. A longer time interval results in a slower turn-on speed for the second main MOSFET M2 and a slower decrease in the output signal amplitude, achieving a high-speed, high-voltage pulse signal output with a programmable falling edge slope.

[0131] Reference Figure 4 The operating timing of this invention within a single cycle is as follows:

[0132] The PWM signal is used as an external trigger signal input. At its rising edge, it triggers the generation of drive pulses P1 and P2. Drive pulse P1 is transmitted through the isolation of the first transformer T1, which turns on the first MOSFET S1 and quickly turns off the second main MOSFET M2.

[0133] The drive pulse is transmitted through isolation via the second transformer T2, turning on the second MOSFET S2 and quickly turning on the first main MOSFET M1. At this time, the amplitude of the output signal rises to its maximum value.

[0134] At the falling edge of the PWM signal, drive pulses P3, P4, P5, and P6 are triggered. Drive pulse P3 is transmitted through isolation via the third transformer T3, turning on the third MOSFET S3 and quickly turning off the first main MOSFET M1.

[0135] The time interval between driving narrow pulses P4, P5, and P6 is programmable via a microcontroller, first delay line D1, second delay line D2, and third delay line D3. It is then transmitted in isolation via fourth transformer T4, fifth transformer T5, and sixth transformer T6, causing fourth MOSFET S4, fifth MOSFET S5, and sixth MOSFET S6 to conduct, and second main MOSFET M2 to conduct slowly. At this time, the amplitude of the output signal gradually decreases according to the conduction speed of the second main MOSFET M2.

[0136] In this embodiment, the falling edge slope of the output high-speed high-voltage pulse signal is determined by the conduction speed of the second main MOSFET M2. The time interval between the driving narrow pulses P4, P5, and P6 is controlled by the microcontroller unit, the first delay line D1, the second delay line D2, and the third delay line D3. The larger the time interval, the slower the conduction speed of the second main MOSFET M2, and the slower the falling edge of the output high-speed high-voltage pulse signal.

[0137] Experimental Test

[0138] The results of the Q-switched high-voltage signal (yellow) and laser pulse width (blue) of this invention are shown in the figure below. Figure 10 As shown, (A) is a waveform with a pulse width of 32ns, (B) is a waveform with a pulse width of 40ns, (C) is a waveform with a pulse width of 50ns, (D) is a waveform with a pulse width of 55ns, (E) is a waveform with a pulse width of 60ns, (F) is a waveform with a pulse width of 80ns, and (G) is a waveform with a pulse width of 110ns.

[0139] This invention proposes a seed laser device for EUV lithography machine light sources with continuously tunable pulse width. This device utilizes a programmable multi-narrow pulse driving method to smoothly and continuously programmably control the falling slope of the Q-switched high-voltage pulse, thereby achieving continuously tunable seed laser pulse width. To achieve linearly adjustable falling edge of the high-voltage Q-switched signal, this invention employs a programmable multi-pulse driven high-speed high-voltage pulse generator, applied to a series-connected MOSFET. The MOSFET cannot achieve full turn-on transients in a single narrow driving pulse, but gradually turns on over multiple consecutive pulses, resulting in a slower falling slope. The time interval between each adjacent narrow pulse is programmable. Therefore, smooth and continuously programmable control of the falling slope of the Q-switched high-voltage pulse loaded on the CdTe crystal is achieved. Ultimately, a continuously tunable pulse width output is achieved for the radio frequency excited waveguide CO2 cavity emptied laser, allowing adaptation to different Sn droplet expansion diameters D. Even without knowing the expansion diameter D, the optimal adaptation to D can be achieved by continuously adjusting the laser duration, effectively improving conversion efficiency to meet the conversion efficiency requirements of the LPP light source in EUV lithography machines.

[0140] The above examples of the present invention are merely illustrative of the computational model and process of the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is impossible to exhaustively list all possible implementations here. Any obvious variations or modifications derived from the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A pulse-width tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine, comprising a set of electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 lasers, characterized in that, The device also includes a high-voltage pulse signal tunable device, which includes at least one high-speed high-voltage switching circuit, including a first main MOSFET. M 1. Second main MOSFET M 2; The first main MOSFET M The drain of MOSFET 1 is connected to the positive power supply +HV, and the second main MOSFET... M The source of 2 is connected to ground (GND), and the first main MOSFET M The source of 1 and the second main MOSFET M 2. Drain connection; The high-speed high-voltage switching circuit is used to generate a high-speed high-voltage pulse signal with adjustable falling edge slope. The high-voltage pulse signal tunable device includes a programmable multi-pulse driven high-speed high-voltage pulse generator. The programmable multi-pulse drive high-speed high-voltage pulse generator includes a logic control module and a MOSFET isolated drive module; the logic control module is used to generate drive wide pulse and drive narrow pulse sequences that meet the timing requirements, and to realize the adjustable time interval between drive narrow pulses; The MOSFET isolation drive module is used to isolate the transmission of drive pulse signals and to provide power to the first main MOSFET. M 1 and second main MOSFET M 2. The charging and discharging of the input capacitors provide a physical circuit; High-speed, high-voltage pulse signals are used to control the modulation crystal in the cavity structure of an electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 laser.

2. The pulse width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 1, characterized in that, The electro-optically Q-switched radio frequency waveguide cavity-emptied CO2 laser adopts a dual-channel structure. One channel uses the electrode structure and resonant cavity structure of a Z-folded cavity-emptied laser, and the other channel uses a straight tube waveguide. The electrode structure and resonant cavity structure of the Z-fold cavity empty laser are as follows: The upper and lower aluminum electrodes press against the ceramic sheet to form a Z-fold structure. An electro-optic Q-switching device is installed in the Z-fold channel resonant cavity. Multiple resonant inductors are connected in parallel on the electrodes. The electro-optic Q-switching device includes a modulation crystal, a Brewster window, and a λ / 4 waveplate. The optical axis of the λ / 4 waveplate is at a 45° angle to the polarization direction determined by the Brewster window. o Angle; Brewster window coated with polarizing dielectric film; total reflection mirrors installed at one end of the folded channel and at the fold; a fixed integrated focal length is installed at a distance f from the waveguide opening. f The concave mirror and piezoelectric ceramic; the modulation crystal and λ / 4 waveplate are placed between the Brewster window and the concave mirror; The straight waveguide has a total reflection mirror at one end and a half reflection mirror at the other end.

3. The pulse width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 2, characterized in that, The modulation crystal in the electro-optic Q-switching device is a CdTe modulation crystal.

4. The pulse width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 2, characterized in that, The λ / 4 waveplate in the electro-optic Q-switching device is a CdS quarter-wave plate.

5. A pulse-width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 2, characterized in that, The total reflection mirror installed at one end of the folded channel and at the fold is a gold-plated copper reflector.

6. A pulse-width tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 2, characterized in that, The ceramic sheet in the Z-fold structure formed by pressing the upper and lower aluminum electrodes together is an Al2O3 ceramic sheet.

7. A pulse-width tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 2, characterized in that, Each segment of the Z-folded channel has a discharge length of 430 mm and a folding angle of 4.5°. o .

8. A pulse-width tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 1, characterized in that, The logic control module includes a microcontroller unit, n delay line chips, and a drive pulse generation circuit. When n=3, the n delay line chips include the first delay line. D 1. Second delay line D 2 and the third delay line D 3; The logic control module receives an externally input PWM signal, which serves as an external trigger signal to drive the pulse generation circuit. Simultaneously, the PWM signal passes sequentially through a first, series-connected delay line. D 1. Second delay line D 2 and the third delay line D 3. Input drive pulse generation circuit; The PWM signal and the first delay line D 1's output signal, second delay line D 2's output signal, third delay line D The output signals of circuit 3 are all used as trigger signals to drive the pulse generation circuit. The drive pulse generation circuit detects the rising or falling edge of the trigger signal and sequentially generates timing-sequential wide drive pulse signals. P 1. P 2. P 3 and drive narrow pulse signal P 4. P 5. P 6.

9. A pulse-width-tunable seed light source device for an LPP light source in an extreme ultraviolet lithography machine according to claim 8, characterized in that, The MOSFET isolation drive module includes a first transformer. T 1. Second transformer T 2. Third Transformer T 3. Fourth Transformer T 4. Fifth Transformer T 5. The Sixth Transformer T 6. First MOSFET S 1. Second MOSFET S 2. Third MOSFET S 3. Fourth MOSFET S 4. Fifth MOSFET S 5 and the sixth MOSFET S 6; Drive wide pulse P 1. Through the first transformer T 1. Isolation transmission control first MOSFET S 1 is turned on, serving as the second main MOSFET. M 2. Input capacitors provide a discharge circuit; Drive wide pulse P 2 via the second transformer T 2. Isolation transmission control second MOSFET S 2 is turned on, and is the first main MOSFET. M 1. The input capacitor provides the charging circuit; Drive wide pulse P 3 via the third transformer T 3. Isolation transmission enables the third MOSFET S 3 is turned on, serving as the first main MOSFET. M The input capacitor 1 provides a discharge circuit; Drive narrow pulse signal P 4. P 5. P 6 are respectively connected to the fourth transformer T 4. Fifth Transformer T 5. The Sixth Transformer T 6. Isolation transmission, controlling the fourth MOSFET S 4. Fifth MOSFET S 5. Sixth MOSFET S 6 is turned on, serving as the second main MOSFET. M The input capacitor of 2 is charged successively, causing it to conduct slowly; Drive narrow pulse P 4. P 5. P The time interval between 6 is controlled by a microcontroller program.