TW202307578A - Forming multiple aerial images in a single lithography exposure pass - Google Patents

Abstract

A method for controlling an energy of a pulsed light beam is described. The method includes: producing a plurality of intermingled sets of pulses of the light beam from an optical source, each set of light beam pulses associated with a distinct primary wavelength and a distinct target energy; receiving a measurement of an energy of a prior pulse of the light beam; determining an energy error including: comparing the measured energy of the prior light beam pulse to a particular target energy associated with a particular set of light beam pulses if the prior light beam pulse is in the particular set of light beam pulses; and adjusting at least one component of the optical source to thereby adjust, based on the determined energy error, the energy of a subsequent pulse in the particular set of light beam pulses.

Description

Form multiple spatial images in a single lithography exposure pass

The present invention relates to the formation of multiple aerial images in a single lithographic exposure pass. For example, the techniques discussed below may be used to form three-dimensional semiconductor devices.

Photolithography is the process of patterning semiconductor circuits on substrates such as silicon wafers. Photolithography optical sources provide deep ultraviolet (DUV) light for exposing photoresists on the wafer. The DUV light used for photolithography is generated by an excimer optical source. Typically, the optical source is a laser source and the pulsed beam is a pulsed laser beam. The beam passes through a beam delivery unit, a reticle or mask, and is then projected onto the prepared silicon wafer. In this way, the wafer design is patterned onto the photoresist, which is then etched and cleaned, and the process then repeated.

In some general aspects, a method for controlling the energy of a pulsed beam of light is described. The method includes: generating a plurality of blended sets of pulses of the beam from an optical source, each set of beam pulses associated with a distinct dominant wavelength and a distinct target energy; receiving one of a previous pulse of the beam A measurement of energy; determining an energy error comprising comparing the measured energy of the previous beam pulse with a specific energy associated with the particular set of beam pulses if the previous beam pulse is in a particular set of beam pulses. a target energy; and adjusting at least one component of the optical source to thereby adjust the energy of a subsequent pulse in the particular set of beam pulses based on the determined energy error.

Implementations can include one or more of the following features. For example, the method may include receiving each distinct target energy associated with each set of beam pulses. The method may include classifying whether the previous beam pulse is in the particular set of beam pulses. The method may include determining an amount to adjust the at least one component of the optical source. The method may include correcting the adjustment of the at least one component of the optical source based on whether the previous beam pulse was in the particular set of beam pulses.

At least one component of the optical source can be adjusted by varying a voltage supplied to electrodes associated with an optical oscillator of the optical source.

In other general aspects, a system includes an optical source device and an energy control device in communication with the optical source device. The optical source device includes: an optical oscillator configured to generate an optical pulse in response to an excitation signal, the optical pulse having a spectral property; and a spectral adjustment device configured to control the optical pulse of the spectral properties. The energy control device is configured to: determine a target energy associated with the spectral property of the generated light pulse; and determine an adjustment to the excitation signal based at least on the determined target energy, the adjustment causing the optical The oscillator generates one or more subsequent light pulses to account for a change in a configuration of the spectral adjustment device.

Implementations can include one or more of the following features. For example, the adjustment of the excitation signal can cause an adjustment of the energy of one or more subsequently generated light pulses.

The target energy associated with the spectral property of the generated light pulse can be previously defined as being associated with the spectral property of the generated light pulse.

The optical oscillator can be associated with a plurality of transfer functions, each transfer function being associated with a specific configuration of the spectral tuning device and a specific value of the spectral property. Also, the energy control device may be configured to determine the adjustment of the excitation signal based on the transfer function associated with the particular configuration of the spectral adjustment device used to generate the one or more subsequent light pulses.

The spectral adjustment device may comprise at least one plenum and a diffractive element arranged in optical communication with each other, and each transfer function is associated with a different state of the at least one plenum.

The spectral property of an optical pulse may be a central wavelength of that optical pulse, and each configuration of the spectral adjustment device may correspond to a specific value of this wavelength.

The system can further include a measurement device configured to measure an energy of the light pulse. The energy control device may be configured to determine an energy error by comparing the target energy to the measured energy, and the determination of the adjustment to the excitation signal may also be based on the energy error.

The energy control device may be configured to determine the The adjustment of the excitation signal causes the optical oscillator to produce the one or more subsequent light pulses.

The energy control apparatus may be configured to determine the target energy associated with the spectral property of the generated light pulse by receiving a communication from a lithographic exposure apparatus configured to receive the light pulse, the communication A set of target energies is provided, each target energy in the set being associated with a spectral attribute.

In other general aspects, an energy control device includes a control module. The control module is configured to receive an energy value of a previous light pulse emitted from an optical source. The control module is configured to perform a comparison comprising comparing the received energy value to a first set of beam pulses only if the previous light pulse was in a first set of beam pulses associated with a first dominant wavelength. target energy; or comparing the received energy value with a value different from the first dominant wavelength only if the previous light pulse was in a second set of beam pulses associated with a second dominant wavelength different from the first dominant wavelength One of the target energies is the second target energy. The control module is configured to adjust at least one component of the optical source based on the comparison to thereby adjust the energy of a subsequent light pulse having the dominant wavelength associated with the previous light pulse.

Implementations can include one or more of the following features. For example, the control module may include a classification module configured to classify the previous light pulses as being in the first set of beam pulses or in the second set of beam pulses. The control module may include a comparator configured to determine whether the previous light pulse is in the first set of beam pulses or the second set of beam pulses, and based on the determination to provide the first target energy or The second target energy. The control module may include a signal module configured to determine an amount of adjustment to be made to the at least one component of the optical source.

The control module may include a calibration module configured to correct the at least one optical source to be based on whether the previous light pulse was in the first set of beam pulses or the second set of beam pulses The adjustment made to the serving size. The correction module can be configured to correct the adjustment by applying a filter to the adjustment. The filter may include a notch filter that transmits information having a frequency in a first frequency band and substantially blocks information having a frequency outside the first frequency band. The filter may include a Kalman filter. The calibration module can be configured to correct the adjustment by applying a forward correction to the adjustment.

The control module configured to adjust at least one component of the optical source based on the comparison to thereby adjust the energy of the subsequent optical pulse having the dominant wavelength associated with the previous optical pulse may include sending a signal to the optical source to thereby vary a voltage supplied to electrodes associated with an optical oscillator of the optical source. The control module configured to receive the energy value of the previous light pulse may include the control module configured to receive the energy value of a plurality of previous light pulses emitted from the optical source. The control module can be configured to adjust the at least one component of the optical source based on the comparison to thereby adjust the energy of subsequent light pulses having the dominant wavelength associated with the previous light pulse. The control module can be configured to maintain the energy of a subsequent light pulse without the dominant wavelength associated with the previous light pulse based on the comparison.

Discussed herein are techniques for forming more than one aerial image, each in a different plane, in a single lithography pass, and using the aerial images to form three-dimensional semiconductor devices.

Referring to FIG. 1A , photolithography system 100 includes an optical (or light) source 105 that provides a light beam 160 to a lithography exposure apparatus 169 that processes light received by a wafer holder or stage 171. Wafer 170. The light beam 160 is a pulsed light beam comprising light pulses separated in time from each other. The lithographic exposure equipment 169 includes a projection optical system 175 and a metrology system 172 , and the light beam 160 passes through the projection optical system 175 before reaching the wafer 170 . Metrology system 172 may include, for example, a camera or other device capable of capturing images of wafer 170 and/or beam 160 at wafer 170, or capable of capturing characteristics describing beam 160, such as the beam at wafer 170 in the x-y plane. 160 intensity) data optical detector. The lithographic exposure apparatus 169 can be a liquid immersion system or a dry system. The photolithography system 100 may also include a control system 150 to control the light source 105 and/or the lithography exposure apparatus 169 .

Microelectronic features are formed on wafer 170 by exposing a layer of radiation-sensitive photoresist material onto wafer 170 , for example, using beam 160 . Referring also to FIG. 1B , the projection optics system 175 includes a slit 176 , a mask 174 , and a projection objective including a lens 177 . The light beams 160 enter the optical system 175 and impinge on the slits 176 , and at least some of the light beams 160 pass through the slits 176 . In the example of FIGS. 1A and 1B , slit 176 is rectangular and shapes beam 160 into an elongated rectangular shaped beam. A pattern is formed on the mask 174 and the pattern determines which portions of the shaped beam are transmitted and which are blocked by the mask 174 . The design of the pattern is dictated by the specific microelectronic circuit design to be formed on wafer 170 .

The shaped beam interacts with the mask 174 . A portion of the shaped beam transmitted by mask 174 passes through (and may be focused by) projection lens 177 and exposes wafer 170 . The portion of the shaped beam transmitted by the mask 174 forms an aerial image in the x-y plane in the wafer 170 . The aerial image is the intensity pattern formed by light that reaches wafer 170 after interacting with mask 174 . The aerial image is at wafer 170 and generally extends in the x-y plane.

System 100 is capable of forming a plurality of aerial images during a single exposure pass, where each of the aerial images is at a spatially different location in wafer 170 along the z-axis. Referring also to FIG. 1C , which shows a cross-sectional view of wafer 170 in the y-z plane, projection optics 175 form two aerial images 173a, 173b at different planes along the z-axis in a single exposure pass. As discussed in more detail below, each of the aerial images 173a, 173b is formed from light having different dominant wavelengths.

The position of the aerial image along the z-axis depends on the characteristics of optical system 175 (including projection lens 177 and mask 174 ), and the wavelength of light beam 160 . The focusing position of the lens 177 depends on the wavelength of the light incident on the lens 177 . Thus, varying or controlling the wavelength of light beam 160 allows control of the positioning of the aerial image. By providing pulses of light with different dominant wavelengths during a single exposure pass, the optical system 175 (or any component of the optical system 175) and the wafer 170 can be shifted relative to each other along the z-axis. A plurality (two or more) of spatial images each at a different position along the z-axis are formed in a single exposure pass.

In the example of FIG. 1A , light passing through mask 174 is focused to the focal plane by projection lens 177 . The focal plane of projection lens 177 is between projection lens 177 and wafer stage 171 , where the positioning of the focal plane along the z-axis depends on the properties of optical system 175 and the wavelength of light beam 160 . The aerial images 173 a , 173 b are formed by light having different wavelengths, and thus, the aerial images 173 a , 173 b are at different positions in the wafer 170 . The aerial images 173a, 173b are separated from each other by a separation distance 179 along the z-axis. The separation distance 179 depends on the difference between the wavelength of the light forming the aerial image 173a and the wavelength of the light forming the aerial image 173b.

The wafer stage 171 and mask 174 (or other parts of the optical system 175) generally move relative to each other in the x, y, and z directions during scanning for routine performance calibration and manipulation, e.g., motion may be used to achieve Compensation for basic leveling, lens distortion and stage positioning errors. This relative motion is called para-operational motion. However, in the system of FIG. 1A , relative motion of wafer stage 171 and optical system 175 is not relied upon to create separation distance 179 . Alternatively, separation distance 179 results from the ability to control the dominant wavelength in the pulses passing through mask 174 during an exposure pass. Thus, unlike some prior systems, separation distance 179 is not created solely by moving optical system 175 and wafer 170 relative to each other along the z-direction. Additionally, aerial images 173a and 173b are both present at wafer 170 during the same exposure pass. In other words, the system 100 does not need to form the aerial image 173a in a first exposure pass and form the aerial image 173b in a second subsequent exposure pass.

The light in the first aerial image 173a interacts with the wafer at portion 178a, and the light in the second aerial image 173b interacts with the wafer at portion 178b. These interactions may form electronic features or other physical features, such as openings or holes, on wafer 170 . Since the aerial images 173 a , 173 b are in different planes along the z-axis, the aerial images 173 a , 173 b can be used to form three-dimensional features on the wafer 170 . For example, aerial image 173a can be used to form a perimeter region, and aerial image 173b can be used to form channels, grooves or recesses at different positions along the z-axis. Accordingly, the techniques discussed herein can be used to form three-dimensional semiconductor devices, such as three-dimensional NAND flash memory devices.

Example implementations of light source 105 and photolithography system 100 are discussed with respect to FIGS. 2A-2C , 3A-3C , and 4 before discussing additional details related to forming multiple aerial images in a single exposure pass.

Referring to FIG. 2A , a block diagram of a photolithography system 200 is shown. System 200 is an example of an implementation for system 100 (FIG. 1A). For example, in photolithography system 200, optical source 205 is used as optical source 105 (FIG. 1A). Optical source 205 generates pulsed light beam 260 that is provided to lithography exposure apparatus 169 . The optical source 205 may be, for example, an excimer optical source that outputs a pulsed beam 260 (which may be a laser beam). As pulsed beam 260 enters lithography exposure apparatus 169 , it is directed through projection optics 175 and projected onto wafer 170 . In this way, one or more microelectronic features are patterned onto the photoresist on wafer 170, and the wafer is developed and cleaned prior to subsequent processing steps, and the process is repeated. Photolithography system 200 also includes control system 250 , which in the example of FIG. 2A is connected to components of optical source 205 and to lithography exposure apparatus 169 to control various operations of system 200 . Control system 250 is an example of an implementation for control system 250 of FIG. 1A .

In the example shown in FIG. 2A , the optical source 205 is a two-stage laser system that includes a master oscillator (MO) 212 that provides a seed beam 224 to a power amplifier (PA) 230 . MO 212 and PA 230 may be considered subsystems of optical source 205 , or systems that are part of optical source 205 . The power amplifier 230 receives the seed beam 224 from the master oscillator 212 and amplifies the seed beam 224 to generate a beam 260 for use in the lithography exposure apparatus 169 . For example, the master oscillator 212 can emit a pulsed seed beam with a seed pulse energy of approximately 1 milliJoule (mJ) per pulse, and these seed pulses can be amplified by the power amplifier 230 to about 10 to 15 mJ.

The master oscillator 212 includes a discharge chamber 214 containing two elongated electrodes 217 , a gain medium 219 as a mixed gas, and a fan for circulating the gas between the electrodes 217 . A resonator is formed between line narrowing module 216 on one side of discharge chamber 214 and output coupler 218 on a second side of discharge chamber 214 . Line narrowing module 216 may include diffractive optics, such as a grating, to finely tune the spectral output of discharge chamber 214 . 2B and 2C provide additional details regarding line narrowing module 216 .

FIG. 2B is a block diagram of an example of an implementation of spectral feature selection module 258 including one or more instances of line narrowing module 216 . The spectral feature selection module 258 is coupled to the light propagating in the optical source 205 . In some embodiments, such as shown in FIG. 2B , spectral feature selection module 258 receives light from chamber 214 of master oscillator 212 to allow for spectral features such as wavelength and bandwidth to be analyzed within master oscillator 212 . Perform fine tuning.

Spectral signature selection module 258 may include a control module, such as spectral signature control module 254 in the form of an electronic device including any combination of firmware and software. The control module 254 is connected to one or more actuation systems, such as spectral signature actuation systems 255_1 to 255_n. Each of the actuation systems 255_1 to 255 — n may include one or more actuators connected to a respective optical feature 256_1 to 256 — n of the optical system 257 . The optical features 256_1 to 256 — n are configured to adjust certain characteristics of the generated light beam 260 to thereby adjust the spectral characteristics of the light beam 260 . The control module 254 receives control signals from the control system 250 , and the control signals include specific commands to operate or control one or more of the actuation systems 255_1 to 255_n. The actuation systems 255_1 to 255_n can be selected and designed to work together, ie in series, or the actuation systems 255_1 to 255_n can be configured to work individually. Furthermore, each actuation system 255_1 to 255_n can be optimized to respond to a particular type of disturbance.

Each optical feature 256_1 to 256 — n is optically coupled to the light beam 260 generated by the optical source 105 . Optical system 257 may be implemented as line narrowing module 216C as shown in Figure 2C. The line narrowing module includes dispersive optical elements (such as reflective grating 291 ) and refractive optical elements (such as gratings 292 , 293 , 294 , 295 ) as optical features 256_1 to 256_n. One or more of the bells 292, 293, 294, 295 may be rotatable. An example of this line narrowing module can be found in the application titled "SYSTEM METHOD AN APPARATUS FOR SELECTING AND CONTROLLING LIGHT SOURCE BANDWIDTH" (the '306 application) on October 23, 2009 and as US Patent 8,144,739 on March 27, 2012 In issued US application Ser. No. 12/605,306, the contents of that application are incorporated herein by reference as if set forth in its entirety. In the '306 application, a line narrowing module is described that includes a beam expander (including one or more beams 292 , 293 , 294 , 295 ) and a dispersive element, such as a grating 291 . The respective actuation systems for one or more of the actuatable optical features such as the grating 291 and the gates 292, 293, 294, 295 are not shown in FIG. 2C.

Each of the actuators of the actuation systems 255_1 to 255 — n is a mechanical device for moving or controlling a respective optical feature 256_1 to 256 — n of the optical system 257 . The actuator receives energy from the module 254 and converts that energy into a certain motion imparted to the optical features 256_1 to 256_n of the optical system 257 . For example, in the '306 application, actuation systems such as force devices (to apply force to grating regions) and rotating stages for expanding one or more of the beam expander's beams are described. The actuation systems 255_1 to 255_n may include, for example, motors such as stepper motors, valves, pressure-controlled devices, piezoelectric devices, linear motors, hydraulic actuators, and/or voice coils.

Returning to FIG. 2A , the master oscillator 212 also includes a line center analysis module 220 that receives the output beam from the output coupler 218 , and a beam coupling optical system 222 that modifies the size or shape of the output beam as needed to form a seed beam 224 . The line center analysis module 220 is a measurement system that can be used to measure or monitor the wavelength of the seed beam 224 . The line center analysis module 220 may be placed elsewhere in the optical source 205 , or it may be placed at the output of the optical source 205 .

The gas mixture used in the discharge chamber 214 can be any gas suitable for generating a light beam at the wavelength and bandwidth required by the application. For the excimer source, the gas mixture may contain inert gases (noble gases) such as argon or krypton, halogens such as fluorine or chlorine, and traces of xenon, in addition to helium and/or neon as buffer gases. Specific examples of gas mixtures include argon fluoride (ArF) which emits light at a wavelength of about 193 nm, krypton fluoride (KrF) which emits light at a wavelength of about 248 nm, or krypton fluoride (KrF) which emits light at a wavelength of about 351 nm xenon chloride (XeCl). By applying a voltage to the elongate electrode 217, the excimer gain medium (gas mixture) is pumped with short (eg, nanosecond) current pulses in a high voltage discharge.

Power amplifier 230 includes beam coupling optics 232 that receive seed beam 224 from master oscillator 212 and direct the beam through discharge chamber 240 and to beam steering optics 248, which beam steering optics Element 248 modifies or changes the direction of seed beam 224 so that it is sent back to discharge chamber 240 . The discharge chamber 240 includes a pair of elongated electrodes 241 , a gain medium 219 as a mixed gas, and a fan for circulating the mixed gas between the electrodes 241 .

The output beam 260 is directed through a bandwidth analysis module 262, where various parameters of the beam 260 (such as bandwidth or wavelength) can be measured. The output beam 260 may also be directed through a beam preparation system 263 . Beam preparation system 263 may include, for example, a pulse stretcher in which each of the pulses of output beam 260 is stretched in time (eg, in an optical delay unit) to adjust the performance properties of the beam striking lithographic exposure apparatus 169 . Beam preparation system 263 may also include other components capable of acting on beam 260, such as reflective and/or refractive optical elements (such as lenses and mirrors), filters, and optical apertures (including automatic shutters).

The photolithography system 200 also includes a control system 250 . In the embodiment shown in FIG. 2A , a control system 250 is connected to the various components of the optical source 205 . For example, control system 250 may control when optical source 205 emits a pulse of light or a burst of pulses of light comprising one or more pulses of light by sending one or more signals to light source 205 . The control system 250 is also connected to the lithography exposure apparatus 169 . Therefore, the control system 250 can also control various aspects of the lithography exposure apparatus 169 . For example, control system 250 may control the exposure of wafer 170 and thus may be used to control how electronic features are printed on wafer 170 . In some embodiments, the control system 250 can control the scanning of the wafer 170 by controlling the movement of the slit 176 in the x-y plane (FIG. 1B). Additionally, control system 250 may exchange data with metrology system 172 and/or optical system 175 .

Lithographic exposure apparatus 169 may also include, for example, temperature control devices, such as air conditioning devices and/or heating devices, and/or power supplies for various electrical components. Control system 250 may also control these components. In some embodiments, the control system 250 is implemented to include more than one sub-control system, and at least one sub-control system (lithography controller) is dedicated to controlling aspects of the lithography exposure apparatus 169 . In such embodiments, the control system 250 may be used to control aspects of the lithography exposure apparatus 169 as an alternative or in addition to using a lithography controller.

Control system 250 includes electronic processor 251 , electronic storage 252 and I/O interface 253 . Electronic processor 251 includes one or more processors suitable for executing computer programs, such as general or special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, electronic processors receive instructions and data from read-only memory, random-access memory, or both. Electronic processor 251 may be any type of electronic processor.

Electronic storage 252 may be volatile memory, such as RAM, or non-volatile memory. In some implementations, and electronic storage 252 includes non-volatile and volatile portions or components. Electronic storage 252 may store data and information for the operation of control system 250 , components of control system 250 , and/or systems controlled by control system 250 . The information may be stored, for example, in a lookup table or a database. For example, electronic storage 252 may store data indicative of values of different properties of light beam 260 under different operating conditions and performance scenarios.

Additionally, electronic storage 252 may store various recipes or processing programs 259 indicative of parameters of light beam 260 during use. For example, electronic storage 252 may store a recipe indicating the wavelength of each pulse in beam 260 in a particular exposure pass. Recipes may dictate different wavelengths for different exposure passes. The wavelength control techniques discussed below can be applied on a pulse-by-pulse basis. In other words, the wavelength content can be controlled for each individual pulse in an exposure pass to facilitate the formation of an aerial image at a desired location along the z-axis.

Electronic storage 252 may also store instructions (possibly as a computer program) that, when executed, cause processor 251 to communicate with components in control system 250 , optical system 205 , and/or lithographic exposure apparatus 169 .

I/O interface 253 is a function that allows control of system 250 from an operator, optical system 205, lithographic exposure apparatus 169, any component or system within optical system 205 and/or lithographic exposure apparatus 169, and/or in another electronic device Electronic interfaces of any kind that receive data and signals from automated programs running on them and/or provide data and signals to such entities. For example, I/O interface 253 may include one or more of a visual display, a keyboard, and a communication interface.

Beam 260 (and beam 160) is a pulsed beam, and may include one or more bursts of pulses that are separated in time from each other. Each burst may include one or more light pulses. In some embodiments, a burst includes hundreds of pulses, such as 100 to 400 pulses. 3A-3C provide an overview of the generation of pulses and bursts in the optical source 205 . FIG. 3A shows the amplitude of the wafer exposure signal 300 over time, FIG. 3B shows the amplitude of the gate signal 315 over time, and FIG. 3C shows the amplitude of the trigger signal over time.

Control system 250 may be configured to send wafer exposure signal 300 to optical source 205 to control optical source 205 to generate light beam 260 . In the example shown in FIG. 3A , wafer exposure signal 300 has a high value 305 (eg, 1) during time period 307 during which optical source 205 produces a burst of light pulses. Wafer exposure signal 300 additionally has a low bit value 310 (eg, 0) when wafer 170 is not exposed.

Referring to FIG. 3B , beam 260 is a pulsed beam, and beam 260 includes bursts of pulses. The control system 250 also controls the duration and frequency of the burst of pulses by sending a gate signal 315 to the optical source 205 . The gate signal 315 has a high value 320 (eg, 1) during bursts of pulses and a low value 325 (eg, 0) during the time between consecutive bursts. In the example shown, the duration when gate signal 315 has a high value is also the duration of burst 316 . The bursts are separated in time by inter-burst intervals. During the inter-burst intervals, lithography exposure apparatus 169 may position the next die on wafer 170 for exposure.

Referring to FIG. 3C, the control system 250 also uses the trigger signal 330 to control the repetition rate of pulses within each burst. Trigger signal 330 includes trigger 340, one of which is provided to optical source 205 to cause optical source 205 to generate a pulse of light. The control system 250 may send a trigger 340 to the source 205 each time a pulse is generated. Thus, the repetition rate (time between two consecutive pulses) of the pulses generated by the optical source 205 can be set by the trigger signal 330 .

As discussed above, when the gain medium 219 is pumped by applying a voltage to the electrodes 217, the gain medium 219 emits light. When a voltage is applied to electrode 217 in pulses, the light emitted from medium 219 is also pulsed. Thus, the repetition rate of the pulsed beam 260 is determined by the rate at which a voltage is applied to the electrode 217, wherein each application of the voltage produces a pulse of light. The light pulse propagates through gain medium 219 and exits chamber 214 through output coupler 218 . Thus, a pulse train is generated by periodically repeating the application of a voltage to the electrode 217 . Trigger signal 330 may be used, for example, to control the application of voltage to electrodes 217 and the repetition rate of pulses, which for most applications may range between about 500 and 6,000 Hz. In some embodiments, the repetition rate may be greater than 6,000 Hz, and may be, for example, 12,000 Hz or greater.

Signals from the control system 250 may also be used to control the electrodes 217, 241 within the master oscillator 212 and the power amplifier 230, respectively, for controlling the respective pulse energies of the master oscillator 212 and the power amplifier 230, and thus control the beam 260 energy. There may be a delay between the signal provided to electrode 217 and the signal provided to electrode 241 . The amount of delay can affect properties of beam 260 , such as the amount of coherence in pulsed beam 260 .

Pulsed beam 260 may have an average output power in the range of tens of watts, eg, about 50 W to about 130 W. The irradiance (ie, the average power per unit area) of the beam 260 at the output can range from 60 W/cm ² to 90 W/cm ² .

Referring also to FIG. 4 , wafer 170 is irradiated by beam 260 . The lithographic exposure apparatus 169 includes an optical system 175 (FIGS. 1A and 1B). In the example of FIG. 4 , optical system 175 (not shown) includes illuminator system 429 that includes objective lens arrangement 432 . Objective arrangement 432 includes projection lens 177 ( FIG. 1B ) and enables image transfer from mask 174 to photoresist on wafer 170 . The illuminator system 429 adjusts the range of angles at which the light beam 260 impinges on the mask 174 . The illuminator system 429 can also homogenize (make uniform) the intensity distribution of the light beam 260 in the x-y plane across the mask 174 .

In some embodiments, an immersion medium may be supplied to cover the wafer 170 . The immersion medium may be a liquid such as water used in liquid immersion lithography. In other embodiments where the lithography is a dry system, the immersion medium can be a gas, such as dry nitrogen, dry air, or clean air. In other embodiments, wafer 170 may be exposed in a pressure-controlled environment, such as a vacuum or partial vacuum.

During an exposure pass, a plurality N of pulses of beam 260 illuminate the same area of wafer 170 . N can be any integer greater than one. The number N of pulses of the beam 110 illuminating the same area may be referred to as the exposure window or exposure pass 400 . The size of the window 400 can be controlled by the slit 176 . For example, the slot 176 may include a plurality of movable vanes such that the vanes form an aperture in one configuration and close the aperture in another configuration. The size of the window 400 can also be controlled by configuring the vanes of the slit 176 to form an orifice of a specific size.

The N pulses also determine the illumination dose for the exposure pass. Illumination dose is the amount of optical energy delivered to the wafer during an exposure pass. Thus, properties of the N pulses, such as the optical energy in each pulse, determine the illumination dose. Furthermore, and as discussed in more detail below, N pulses may also be used to determine the amount of light in each of the aerial images 173a, 173b. In particular, the recipe may specify that of the N pulses, a certain number of pulses have a first dominant wavelength forming aerial image 173a and a certain number of pulses have a second dominant wavelength forming aerial image 173b.

Additionally, slit 176 and/or mask 174 may be moved in the scan direction in the x-y plane such that only a portion of wafer 170 is exposed at a given time or during a particular exposure scan (or exposure pass). The size of the area on wafer 170 exposed by beam 160 is determined by the distance between the blades in the non-scanning direction and the length (distance) of the scan in the scanning direction. In some embodiments, the value of N is several tens, such as from 10 to 100 pulses. In other embodiments, the value of N is greater than 100 pulses, eg, from 100 to 500 pulses. Exposure field 479 of wafer 170 is the physical area of wafer 170 that is exposed in one scan of an exposure slit or window within lithography exposure apparatus 169 .

The wafer stage 171 , mask 174 and objective lens arrangement 432 are secured to associated actuation systems, thereby forming a scanning arrangement. In the scanning configuration, one or more of mask 174, objective lens arrangement 432, and wafer 170 (via stage 171) are movable relative to each other in the x-y plane. However, apart from incidental relative operational motion between wafer stage 171, mask 174, and objective lens arrangement 432, these elements do not move relative to each other along the z-axis during or during an exposure pass.

Referring to FIG. 5 , a flow diagram of a process 500 is shown. Process 500 is an example of a process for forming a three-dimensional semiconductor component or a portion of such a component. Process 500 may be performed using photolithography system 100 or 200 . Process 500 is discussed with respect to system 200 shown in FIG. 2A. Process 500 is also discussed with respect to FIGS. 6A-10B .

Light beam 260 is directed toward mask 174 (510). Beam 260 is a pulsed beam comprising a plurality of pulses, each of which is separated in time from each other, such as shown in Figure 3C. 6A and 6B show an example of the spectrum of a single pulse as part of beam 260 . Other pulses in beam 260 may have different spectra.

Referring to FIG. 6A , a spectrum 601A of a light pulse 600A is shown. Light pulse 600A has a non-zero intensity within a wavelength band. The wavelength band may also be referred to as the bandwidth of the pulse 600A.

The information shown in FIG. 6A is the instantaneous spectrum 601A (or emission spectrum) of pulse 600A. Spectrum 601A contains information about how the optical energy or power of the pulses of beam 260 is distributed over different wavelengths (or frequencies). Spectrum 601A is depicted in graphical form, where spectral intensity is plotted against wavelength or optical frequency (not necessarily with absolute calibration). Spectrum 601A may be referred to as the spectral shape or intensity spectrum of the pulses of beam 260 . Pulse 600A has a dominant wavelength 602A, which in the example of FIG. 6A is a peak intensity. Although the discussion of the pulses of the beam 260 and the spatial image formed by the pulses of the beam 260 refer to the dominant wavelength of the pulse, the pulses include wavelengths other than the dominant wavelength, and the pulses have a finite bandwidth that can be characterized by a metric. For example, the full width of the spectrum at the fraction of maximum peak intensity (X) of the shape of spectrum 601A (referred to as FWXM) can be used to characterize the beam bandwidth. As another example, the width of the spectrum containing the fraction (Y) of the integrated spectral intensity (referred to as EY) can be used to characterize the beam bandwidth.

Pulse 600A is shown as an example of a pulse that may be in beam 260 . When pulse 600A is used to expose a portion of wafer 120, the light in the pulse forms an aerial image. The position of the aerial image in the z direction (FIGS. 1C and 4) is determined by the value of the dominant wavelength 602A. Different pulses in beam 260 may have different dominant wavelengths. For example, to produce two spatial images during a single exposure pass, some of the pulses of the beam 260 have one dominant wavelength (the first dominant wavelength) and other pulses of the beam 260 have the other dominant wavelength (the second dominant wavelength). wavelength). The first and second dominant wavelengths are different wavelengths. The wavelength difference between the first and second dominant wavelengths may be referred to as spectral separation. For example, the spectral separation may be 200 femtometer (fm) to 50 picometer (pm). Although the wavelengths of the various pulses in beam 260 may be different, the spectral shape of the pulses may be the same.

The light source 205 may dither or switch the dominant wavelength between the first and second dominant wavelengths on a pulse-to-pulse basis such that each pulse has a different dominant wavelength than the pulse immediately preceding or following the pulse in time. In these embodiments, distributing the first and second dominant wavelengths in this way produces two spatial images with the same intensity at different positions in the z-direction, assuming that all pulses in beam 260 have the same intensity.

In some implementations, some portion of the pulses (eg, 33%) has the first dominant wavelength and the remainder (67% in this example) has the second dominant wavelength. In these embodiments, assuming that all pulses in beam 260 have the same intensity, the two aerial images are formed with different intensities. The aerial image formed by the pulse having the first dominant wavelength has approximately half the intensity of the aerial image formed by the pulse having the second dominant wavelength. In this way, the dose provided to a particular location in wafer 170 along the z-axis can be controlled by controlling the fraction of N pulses having each of the first and second dominant wavelengths.

Pulse portions having a particular dominant wavelength for an exposure pass may be specified in a recipe file 259 stored on electronic storage 252 . Recipe 259 specifies ratios of the various dominant wavelengths for exposure passes. Recipe 259 can also specify ratios for other exposure passes so that different ratios can be used for other exposure passes and the spatial image can be adjusted or controlled on a field-by-field basis.

Referring to Figure 6B, a spectrum 601B of pulse 600B is shown. Pulse 600B is another example of a pulse for beam 260 . Spectrum 601B of pulse 600B has a different shape than spectrum 601A. In particular, spectrum 601B has two peaks corresponding to the two dominant wavelengths 602B_1 and 602B_2 of pulse 600B. Pulse 600B is a portion of beam 260 . When pulse 600B is used to expose a portion of wafer 120, the light in the pulse forms two aerial images at different locations on the wafer along the z-axis. The position of the spatial image is determined by the wavelengths of the dominant wavelengths 602B_1 and 602B_2.

The pulses shown in Figures 6A and 6B can be formed by any hardware capable of forming such pulses. For example, a line narrowing module similar to line narrowing module 216C of FIG. 2C may be used to form a pulse train of pulses such as pulse 600A. The wavelength of the light diffracted by the grating 291 depends on the angle of the light incident on the grating. A mechanism to vary the angle of incidence of light interacting with the grating 291 can be used with this line narrowing module to generate a pulse train of N pulses for an exposure pass, wherein at least one of the N pulses One has a dominant wavelength different from that of another of the N pulses. For example, one of the beams 292, 293, 294, 295 may be rotated to change the angle of light incident on the grating 291 on a pulse-by-pulse basis. In some embodiments, the line narrowing module includes a mirror that is in the path of the light beam 260 and is movable to change the angle of light incident on the grating 291 . An example of such an implementation is discussed, for example, in US Patent No. 6,192,064, issued February 20, 2001, entitled "NARROW BAND LASER WITH FINE WAVELENGTH CONTROL."

A pulse, such as pulse 600B (FIG. 6B), may be formed using a line narrowing module similar to line narrowing module 216C of FIG. 2C. For example, a stimulated optical element such as an acousto-optic modulator may be placed in the path of beam 260 in line narrowing module 216C. An acousto-optic modulator deflects incident light at an angle that depends on the frequency of the acoustic wave used to excite the modulator. An acoustic modulator includes a material, such as glass or quartz, that allows sound waves to propagate, and a transducer coupled to the material. The transducer vibrates in response to the excitation signal, and the vibration generates sound waves in the material. The sound waves create moving planes of expansion and compression that change the refractive index of the material. Thus, the acoustic waves act as a diffraction grating, causing incident light to diffract and exit the material at several different angles simultaneously. Light from two or more orders may be allowed to reach the grating 291 with light in each of the various diffracted orders having different angles of incidence on the grating 291 . In this way, a single pulse comprising two or more dominant wavelengths can be formed. Examples of line narrowing modules including acousto-optic modulators are discussed, for example, in US Patent No. 7,154,928, issued December 26, 2006, entitled "LASER OUTPUT BEAM WAVEFRONT SPLITTER FOR BANDWIDTH SPECTRUM CONTROL."

During a single exposure pass, a set of light pulses is directed toward wafer 170 through mask 174 (520). As discussed above, N light pulses may be provided to wafer 170 during an exposure pass. The N light pulses may be consecutive light pulses in beam 260 . The exposed portion of wafer 170 sees the average of the spectra of each of the N pulses in the exposure pass. Thus, if a portion of the N pulses has a first dominant wavelength and the remaining N pulses have a second dominant wavelength, the average spectrum at wafer 170 will include a peak at the first dominant wavelength and a peak at the second dominant wavelength spectrum. Similarly, if all or some individual pulses of the N pulses have more than one dominant wavelength, those dominant wavelengths may form a peak in the averaged spectrum. FIG. 7 shows an example of an averaged spectrum 701 at wafer 170 . The average spectrum 701 includes a first dominant wavelength 702_1 and a second dominant wavelength 702_2. In the example of FIG. 7 , the first dominant wavelength 702_1 and the second dominant wavelength 702_2 are separated by a spectral separation 703 of about 500 fm, however, other combinations are also contemplated. The spectral separation 703 is such that the first dominant wavelength 702_1 and the second dominant wavelength 702_2 are distinct, and the averaged spectrum 701 includes a spectral region 704 with little to no intensity between the wavelengths 702_1 and 702_2.

For example, based on the averaged spectrum, two or more spatial images are formed at wafer 170, a first image based on the first dominant wavelength and a second image based on the second dominant wavelength (530). Continuing with the example of averaging spectrum 701 , and referring also to FIG. 8A , two aerial images 873a and 873b are formed in a single exposure pass based on N pulses. The N pulses include a first set of pulses with a first dominant wavelength 702_1 and a second set of pulses with a second dominant wavelength 702_2. For example, such pulses are single peak pulses such as shown in Figure 6A. Pulses with the first dominant wavelength 702_1 form a first aerial image 873a, and pulses with a second dominant wavelength 702_2 form a second aerial image 873b. Aerial image 873a is formed at first plane 878a, and aerial image 873b is formed at second plane 878b. Planes 878a and 878b are perpendicular to the direction of propagation of beam 260 at wafer 170 . Planes 878a and 878b are separated by a separation distance 879 along the z-direction.

For an averaged spectrum with a single dominant wavelength, the separation distance 879 is greater than the depth of focus of the lithography apparatus 169 . Depth of focus can be defined as the range of focus along the z-direction for the dose value (the amount of optical energy delivered to the wafer) at which to provide 170 within an acceptable range of feature sizes for the process applied to wafer 170. feature size. Process 500 can increase the depth of focus of lithographic exposure apparatus 169 by providing more than one distinct spatial image at wafer 170 during a single exposure pass. This is because the plurality of aerial images can all expose the wafer at different positions in the z-direction, with features within acceptable ranges of feature sizes. In other words, process 500 is able to provide lithography exposure apparatus 169 with a greater range of depth of focus during a single exposure pass. As discussed above, the operator of the lithography exposure apparatus 169 can control various parameters of the exposure process via the recipe file 259 . In some embodiments, the operator of the lithography exposure apparatus 169 may receive information from a simulation program such as Tachyon Source-Mask Optimization (SMO) available from Brion (ASML Corporation) , and this information can be used to program or otherwise specify the parameters of the recipe file 259 . For example, an operator of lithography exposure apparatus 169 may know that an upcoming batch does not require as much depth of focus as a previous exposure batch. In this example, the operator can specify the depth of focus and dose variation for the simulation program, and the simulation program returns values for spectral separation 703 to achieve the desired parameters. The operator can then specify values for the spectral separation 703 for the upcoming batch by programming the recipe file 259 via the I/O interface 253 . In some embodiments, an operator may use simulations to determine whether a greater depth of focus is required for a particular exposure pass (such as perhaps by exposing wafer 170 with multiple aerial images at different planes). In cases where a greater depth of focus is not required to form a particular portion of a semiconductor device, the recipe file 259 may be structured such that, for example, the exposure passes used to form that particular portion of the semiconductor device have an average spectrum that includes a single dominant wavelength.

In addition, the operator and/or simulator may receive information about the formed three-dimensional component as measured by the metrology system 172 or by another sensor. For example, metrology system 172 may provide data related to sidewall angles of formed 3D semiconductor devices, and the data may be used to program parameters in recipe file 259 for subsequent exposure passes.

Figure 8B shows aerial image 873a in the x-y plane at plane 878a (see page in Figure 8A). Aerial images 873a and 873b are generally two-dimensional intensity patterns formed in the x-y plane. The nature of the intensity pattern depends on the characteristics of the mask 174 . The first plane 878 a and the second plane 878 b are part of the wafer 170 . As illustrated in FIG. 8B , the first plane 878a may be only a small portion of the entire wafer 170 .

等式(1)， 其中ΔD為以奈米(nm)為單位之分離距離879，C為色像差(定義為焦平面在傳播方向上針對波長改變而移動之距離，該波長改變為投影透鏡177之已知屬性)，且Δλ為以皮米為單位之光譜分離873。對於具有C = 500 nm/pm之值的透鏡177，為實現5000 nm (5 µm)之聚焦分離距離875，光譜分離873可為約10 fm。 The value of separation distance 879 depends on spectral separation 703 and on the properties of optical system 275 . For example, the value of separation distance 879 may depend on the focal length, aberrations, and other properties of the lenses and other optical elements in optical system 275 . For a scanner lens with chromatic aberration C, the separation distance 879 can be determined from Equation 1:

Equation (1), where ΔD is the separation distance 879 in nanometers (nm), and C is the chromatic aberration (defined as the distance the focal plane moves in the direction of propagation for a change in the wavelength that the projection lens 177), and Δλ is the spectral separation 873 in picometers. For a lens 177 with a value of C = 500 nm/pm, to achieve a focus separation distance 875 of 5000 nm (5 μm), the spectral separation 873 may be about 10 fm.

Furthermore, for a particular instance of a certain type of exposure apparatus 169, a different dominant wavelength may be required to achieve the desired separation distance 879 due to variations in the manufacturing and installation process and/or modifications made by the end user. As discussed above, recipe or program control program 259 may be stored on electronic storage 252 of control system 250 . Recipe 259 can be modified or programmed to be customized for a particular exposure equipment or type of exposure equipment. Recipe 259 may be programmed when lithography system 200 is manufactured, and/or may be programmed by an end user or other operator familiar with the performance of system 200 via, for example, I/O interface 253 .

Recipe 259 may also specify different separation distances 879 for different exposure passes used to expose different regions of wafer 170 . Additionally or alternatively, recipe 259 may specify separation distance 879 on a per-lot or per-layer basis, or on a per-wafer basis. A lot or layer is a group of wafers processed by the same exposure equipment under the same nominal conditions. The recipe 259 also allows specifying other parameters related to the aerial images 873a, 873b, such as the dose provided by each image. For example, the recipe 259 may specify a ratio of the number of pulses out of N pulses having the first dominant wavelength 702_1 to the number of pulses having the second dominant wavelength 702_2. These other parameters may also be specified on a per field, per lot (or per layer) and/or per wafer basis.

Furthermore, the recipe 259 may specify that some layers are not exposed by the first dominant wavelength 702_1 and the second dominant wavelength 702_2, and are instead exposed by pulses having a spectrum comprising a single dominant wavelength. This spectrum can be used, for example, when planar rather than three-dimensional semiconductor components should be formed. I/O interface 253 allows end users and/or manufacturers to program or generate recipes to specify the number of dominant wavelengths, including, for example, scenarios where a single dominant wavelength is used for a particular layer or lot.

Additionally, while the examples above discuss averaged spectrum 701 having two dominant wavelengths, in other examples averaged spectrum 701 may have more than two dominant wavelengths (e.g., three, four, or five dominant wavelengths), of which Each is separated from the nearest other dominant wavelength by spectral separation and regions such as region 704 . I/O interface 253 allows end users and/or manufacturers to program or generate recipes to specify these parameters.

A three-dimensional (3D) semiconductor component is formed (540). FIG. 9A shows a cross-sectional view of an example of a 3D semiconductor component 995 . Figure 9B shows wafer 170 and component 995 in the x-y plane at first plane 878a. 3D semiconductor component 995 may be a complete component or part of a larger component. 3D semiconductor component 995 may be any type of semiconductor component that has features that are not all formed at one z-position in wafer 170 . For example, a 3D semiconductor component may be a device comprising a recess or opening extending along the z-axis. 3D semiconductor components can be used in any type of electronic application. For example, the 3D semiconductor device can be all or part of a 3D NAND flash memory device. 3D NAND flash memory is memory in which memory cells are stacked in layers along the z-axis.

In the example of FIG. 9A , a 3D semiconductor component 995 includes a recess 996 formed in a perimeter 999 . Recess 996 includes a bottom surface 997 and a sidewall 998 extending generally along the z-axis between perimeter 999 and bottom surface 997 . Bottom surface 997 is formed by exposing the photoresist at plane 878b with light in second aerial image 873b (FIG. 8A). Features on perimeter 999 are formed using light in first aerial image 873a (FIG. 8A).

It is also possible to use process 500 to produce side wall angles 992 equal to or closer to 90° than other processes may produce. Side wall angle 992 is the angle between bottom surface 997 and side wall 998 . If the sidewalls 998 extend in the x-z plane and the bottom surface extends in the x-y plane, the sidewall angle 992 is 90° and can be considered vertical in this example. Sidewall angles that are closer to vertical are desirable because, for example, they may allow for more well-defined features in 3D semiconductor devices. Process 500 achieves sidewall angle 992 equal to or close to 90° because the locations of first aerial image 873a and second aerial image 873b (first plane 878a and second plane 878b, respectively) are on different parts of wafer 170 Individual images at . Forming separate spatial images in a single exposure pass allows improving the quality of each of the images, resulting in more defined features that are more vertically oriented than features formed by lower quality single spatial images.

10A and 10B are examples of simulation data related to process 500 . Figure 1OA shows three plots 1001, 1002, 1003 of spatial image intensity versus mask positioning (Figure 9A) along the y-axis. Each of the plots 1001, 1002, 1003 represents the intensity and mask positioning for an aerial image. In FIG. 10A , plot 1001 represents a simulation of the average spectrum forming two aerial images during a single exposure pass, such as discussed above with respect to FIG. 5 . Plot 1002 represents a simulation of wafer stage tilting according to ASML's EFESE technique, a procedure for increasing depth of focus to facilitate printing of three-dimensional features such as vias and holes on a wafer. In the EFESE technique, the wafer stage is tilted at an angle to scan the aerial image through the focal point when exposing the wafer. The EFESE technique generally produces a greater depth of focus. In FIG. 1OA, only plot 1002 represents data simulated using the EFESE technique. The rest of the data shown in Figure 10A does not use the EFESE technique. Plot 1003 represents data from a dose-based best focus simulation.

The spatial image intensity as a function of mask positioning shown in Figure 10A illustrates that forming two or more aerial images in a single exposure pass can produce contrast similar to tilting the wafer stage. Greater contrast indicates that three-dimensional features at different locations along the z-axis (FIG. 8A) are more likely to be properly formed.

Figure 10B shows three plots 1004, 1005, 1006 of critical dimension as a function of focus position for three different aerial images, each averaged over exposure passes. In Figure 10B, plot 10004 represents data from a simulation that did not apply the EFESE technique and formed a single spatial image. Plot 1005 represents data from a simulation applying the EFESE technique. As shown, the EFESE technique increases the depth of focus compared to the no-EFESE simulation because the critical dimension value remains constant at greater distances from zero focus. Plot 1005 represents data from a simulation that produced two aerial images in a single exposure pass and did not employ the EFESE technique. Depth of focus for EFESE-free simulations using multiple aerial images is equal to or better than EFESE techniques. Thus, process 500 can be used to achieve greater depth of focus in a single exposure pass without relying on techniques such as EFESE.

Referring to FIG. 11A , an embodiment 1150 of a control system 250 is shown as part of a photolithography system 1100 . Control system 1150 includes processor 251, electronic storage 252, and I/O interface 253, which together are configured to interface with spectral feature selection module 258 within optical source 1105 to thereby enable Adjustment of the spectral characteristics of the pulsed beam 1160 output by 1105. Additionally, control system 1150 includes an energy control module 1161E configured to provide an excitation signal 1168E to optical source 1105 for controlling a master oscillator of optical source 1105 (such as the master oscillator of FIG. 2A ). 212) within the electrodes. The energy control module 1161E may also be configured to provide an excitation signal to one or more other oscillators within the optical source 1105 . The control system 1150 can be used with any type of optical source 1105 . Control system 1150 may be used with optical source 1105 comprising a single optical oscillator. Control system 1150 may be used with multi-stage optical source 1105, such as optical source 205 of FIG. 2A, including one or more optical oscillators and one or more power amplifiers.

Optical source 1105 provides pulsed beam 1160 to lithography exposure apparatus 1169 . Energy control device 1160E is formed by energy control module 1161E and optical detection system 1145E. Optical detection system 1145E is configured to sense light, such as pulsed beam 1160, and generate energy property signal 1146E. Optical detection system 1145E is any type of optical sensor or detector capable of measuring optical energy in pulsed beam 1160 and generating an energy property signal 1146E based on this measurement. Energy attribute signal 1146E includes information about the energy in one or more pulses of beam 1160 . The energy attribute may be, for example, the optical energy of the optical pulses in the pulsed beam 1160 or an energy error associated with the optical pulses in the pulsed beam 1160 .

Energy control module 1161E generates excitation signal 1168E or causes excitation signal 1168E to be generated by a separate device, such as source supply 1197E, configured to amplify the signal and apply a voltage to one or more electrodes 217, as shown in FIG. 2A shown in . When an excitation signal 1168E is applied to one or more optical oscillators in optical source 1105, that optical oscillator generates a pulse of light. The excitation signal 1168E and the pulses in the light beam 1160 are time-varying signals. In the following discussion, individual instances of excitation signal 1168E, pulse and energy property signal 1146E may be indexed by k, where k is an integer number. For example, the kth instance of excitation signal 1168E (excitation signal 1168E(k)) produces pulse k of beam 1160 . Energy control module 1161E receives an instance of energy attribute signal 1146E and generates an instance of excitation signal 1168E for each pulse in beam 1160 .

The amount of optical energy (ie, the energy in the pulses of light beam 1160) produced in response to application of excitation signal 1168E depends on the characteristics of excitation signal 1168E. For example, the excitation signal 1168E may be a train of voltage pulses, and characteristics of the excitation signal 1168E may include the amplitude and/or temporal duration of the voltage pulses. The energy control module 1161E determines the excitation signal 1168E or a characteristic of the excitation signal 1168E. In the following discussion, energy control module 1161E and its various implementations are described as generating or determining excitation signal 1168E. However, in some implementations, the energy control module 1161E (or any of its various implementations) generates the characteristic of the signal 1168E that is provided to the source provider 1197E that generates the signal 1168E based on the characteristic. For example, the excitation signal 1168E may be a high voltage signal generated by the source supply 1197E.

The energy control module 1161E is implemented to achieve spectral feature dependent (eg, wavelength dependent) dose or energy control at the lithography exposure apparatus 1169 . In particular, energy control module 1161E enables the dose and/or energy of a current pulse in beam 1160 to be varied relative to previous and adjacent pulses in beam 1160 . This change may be performed for each pulse of the beam 1160 such that the energy changes with each pulse of the beam 1160 . The energy control module 1161E is configured to provide pulse-to-pulse control of the dose and/or energy of the pulses in the beam 1160 by varying and correcting the excitation signal 1168E provided to the optical source 1105 .

It may be desirable to generate different energies for different pulses in a manner dependent on the wavelength selected for that pulse. In this way, it may be desirable that the value of the dose and/or energy of the pulses depend on the wavelength of the pulses (or other spectral characteristics or simply the number or timing of pulses). For example, referring to FIG. 7 , it may be necessary to generate a first pulse set with a first dominant wavelength 702_1 at a first target energy Etarget1, and generate a set of pulses with a second dominant wavelength 702_1 at a second target energy Etarget2 different from the first target energy Etarget1. The second pulse set of dominant wavelength 702_2. In this way, the dose and/or energy of the pulses of the light beam 1160 can be optimized at each aerial image 873a, 873b within the field.

As discussed above with reference to FIGS. 2A-2C , optical source 1105 includes spectral feature selection module 258 coupled to light propagating in optical source 1105 to enable fine tuning of spectral features, such as wavelength within master oscillator 212 and bandwidth. In multi-focus imaging, the spectral feature selection module 258 can change its configuration by every pulse or by every nth pulse, where n is an integer greater than 1. Optical oscillator 212 is associated with a plurality of transfer functions, each transfer function is associated with a particular configuration of spectral signature selection module 258, and each transfer function is related to an efficiency characteristic of that configuration. A particular transfer function relates the characteristics of the excitation signal 1168E to the amount of optical output (within the pulsed beam 224 or 260) produced by the optical oscillator 212 when in that particular configuration. As a specific example with reference to FIG. 2A , the transfer function for a particular configuration of the optical oscillator 212 (and a particular configuration of the spectral feature selection module 258) relates the magnitude of the voltage applied to the electrode 217 in the chamber 214 to the voltage generated by the chamber 214. The optical energy generated by the gain medium in 214 is related.

Referring to FIG. 11B , the transfer function TF (the optical energy generated by the single optical oscillator 212 that varies with the supplied excitation energy) varies with the wavelength of the emitted pulsed beam. 11B includes: transfer function TF(1), which is the efficiency of the optical oscillator 212 when the center or dominant wavelength of the pulse is the first wavelength (λp1); and transfer function TF(2), which is the center or dominant wavelength of the pulse Efficiency of the optical oscillator 212 at the second wavelength (λp2). The transfer functions TF(1) and TF(2) relate the voltage V applied to the excitation mechanism of the optical oscillator 212 to the optical energy of the pulses of the light beam 1160 generated by the optical oscillator 212 . Both transfer functions TF(1) and TF(2) are locally close to linear, but have different slopes and different y-intercepts. The pulse energy (Epulse) depends on the transfer function TF, as follows: Epulse = TF × [HVSetPoint - OffsetV] + OffsetE + noise, where HVSetPoint is the discharge voltage set point, and OffsetV is the voltage bias applied to the excitation mechanism of the optical oscillator 212 shift, and OffsetE is the energy offset.

In one example, the optical oscillator 212 alternates between generating light pulses at a first dominant wavelength (λp1) and light pulses at a second dominant wavelength (λp2) to generate light with a spectral peak at the first dominant wavelength and Pulsed beam 1160 at the spectral peak of the second dominant wavelength. In this way, light pulses at the first dominant wavelength (λp1) are generally blended (and in some embodiments, interleaved) with light pulses at the second dominant wavelength (λp2).

System 1160 attempts to maintain a first target energy Etarget1 for light pulses at a first dominant wavelength (λp1) and a second target energy Etarget2 for light pulses at a second dominant wavelength (λp2). For example, the kth pulse has energy E1 and a dominant wavelength of λp2. After the kth pulse is generated, the optical elements within the spectral feature selection module 258 are activated such that the dominant wavelength of the k+1th pulse will be λp2. Thus, the system 1160 determines the voltage suitable for the optical oscillator 212 to generate the k+1th pulse based on an estimate of the transfer function TF(2) that is configured to generate the k+1th pulse where the dominant wavelength is the second An accurate representation of the efficiency of the optical oscillator 212 configuration for pulses at the dominant wavelength (λp2).

The energy control module 1161E can be configured to determine the excitation signal 1168E based on a transfer function associated with a particular configuration of the spectral feature selection module 258 used to generate subsequent light pulses. For example, the spectral feature selection module 258 includes at least one sputum, and each transfer function can be associated with a different location of the at least one scallop.

Energy disturbances and their associated imperfect estimates for each state of the spectral feature selection module 258 of the optical source 1105 arise due to differences in transfer functions whenever the energy of the current pulse changes relative to previous and neighboring pulses. Furthermore, undesirable oscillations in the energy of the pulses of beam 1160 may occur due to coupling between the energy of the pulses of beam 1160 and the wavelength. Without any sort of correction mechanism to quickly account for such energy disturbances, the dose and/or energy of the pulses of beam 1160 may be wrong or non-optimal, and this further causes errors at wafer 170 . The energy control module 1161E corrects or adjusts the excitation signal 1168E using a calibration module and also a modeling module that estimates the transfer function for each state of the spectral feature selection module 258 such that the energy control module 1161E can move Eliminate or reduce energy interference. In addition, the energy control module 1161E performs this control on a pulse-to-pulse basis to account for errors that occur on a per-pulse basis.

Referring to Figure 12, an implementation 1261E of an energy control module 1161E is shown for use with an optical oscillator 1212E. Energy control module 1261E is configured to be implemented as part of control system 1150 or 250 . Optical oscillator 1212E may be one of two or more optical oscillators in a multi-stage optical source, such as optical source 205 of FIG. 2A . The output of optical oscillator 1212E is a pulsed beam, such as seed beam 224 or output beam 260 (FIG. 2A). In some embodiments, it is possible to have separate energy control modules 1261E configured for each optical oscillator in a multi-level optical source. For example, the first energy control module 1261E can be configured for the master oscillator 212, and the second energy control module 1261E can be configured for the power amplifier 230 (see FIG. 2A). In other embodiments, a single energy control module 1261E may be configured for both the master oscillator 212 and the power amplifier 230 (see FIG. 2A ).

The energy control module 1261E includes a comparator 1263E and an energy controller 1262E. The energy control module 1261E also includes a target energy generator 1270E. Comparator 1263E receives energy attribute signal 1246E from optical detection system 1145E and also receives target energy Etarget 1271E from target energy generator 1270E. Comparator 1263E implements a comparison function, such as subtraction to determine error signal 1266E. Energy controller 1262E includes one or more modules configured to determine excitation signal 1268E, which corresponds to excitation signal 1168E referenced in FIG. 11A. The excitation signal 1268E takes into account the error signal 1266E, and also takes into account changes in the transfer function of the optical oscillator 1212E, as discussed below.

Referring to Figure 13, a more detailed view of the master oscillator 212 is shown. The two elongated electrodes 217 include a cathode 217 - a and an anode 217 - b contained in the discharge chamber 214 . The potential difference between cathode 217 - a and anode 217 - b creates an electric field in gaseous gain medium 219 . The potential difference is created by controlling the source supplier 1197E to apply a voltage to the cathode 217-a and/or the anode 217-b. In this example, source provider 1197E is controlled by excitation signal 1168E. The excitation signal 1168E includes information sufficient to cause the source provider 1197E to generate a voltage signal 1168Ev and apply the voltage signal 1168Ev to the master oscillator 212 in accordance with the trigger signal 330 (FIG. 3C). Voltage signal 1168Ev has an amplitude specified by excitation signal 1168E. Source supplier 1197E applies voltage signal 1168Ev to thereby apply a voltage of a particular amplitude to cathode 217-a and/or anode 217-b such that the electric field will be sufficient to cause population inversion and enable generation of light beam 224 by means of stimulated emission The energy of the pulse is provided to the gain medium 219. Repeated generation of this potential difference forms pulse trains that are emitted as light beam 224 and thus as light beam 260 (FIG. 2A).

Referring again to 12, comparator 1263E implements a comparison function such as subtraction. Comparator 1263E receives energy attribute signal 1246E from optical detection system 1145E, and receives the value of target energy Etarget 1271E from target energy generator 1270E. Energy attribute signal 1246E includes an indication of the amount of optical energy in pulse k-1, which is the pulse immediately preceding pulse k.

Target energy Etarget 1271E is the target or desired optical energy value for a subset of optical pulses in beam 1160 . The target energy Etarget 1271E is a predefined optical energy associated with acceptable or optimal performance of the photolithography system 1100 . The value of Etarget 1271E may be stored in electronic storage 252 or another location within optical source 1105 and may be ready for use by comparator 1263E when needed. In some embodiments, the value of Etarget 1271E may be indicated by lithography exposure apparatus 1169 (as shown by arrow 1165). As discussed above, the energy control module 1161E is implemented to achieve spectral feature dependent dose or energy control at the lithography exposure apparatus 1169 . To implement wavelength-dependent dose or energy control, a target energy generator 1270E provides or determines a target energy Etarget 1271E, which correlates to a spectral property, such as the dominant wavelength λp, of the pulses of the light beam 1160 generated by the optical oscillator 1212E.

For example, FIG. 14 shows a table in which each target energy Etarget 1271E_i is related to each possible dominant wavelength λp 1402_i in the pulse set of beam 1160, where i is an integer greater than 1 and has a maximum value of M. This table may be stored within optical source 1105 or lithographic exposure apparatus 1169 and accessed by target energy generator 1270E after generating a pulse of light beam 1160 .

As another example, Figure 15A shows a graph of target energy Etarget 1571E versus four dominant wavelengths 1502, each of which is associated with a set of beam pulses. Thus, dominant wavelength 1502a is associated with target energy Etarget 1571Ea; dominant wavelength 1502b is associated with target energy Etarget 1571Eb; dominant wavelength 1502c is associated with target energy Etarget 1571Ec; and dominant wavelength 1502d is associated with target energy Etarget 1571Ed. In this example, as shown in FIG. 15B, four distinct aerial images 1573a, 1573b, 1573c, 1573d are formed at wafer 170 during the same exposure pass, each aerial image 1573a, 1573b, 1573c, 1573d along The z-axes are formed at separate and distinct planes 1578a, 1578b, 1578c, 1578d. The location of the plane depends on the dominant wavelength 1502 . Thus, for example, aerial image 1573a is formed at plane 1578a, the position along the z-axis being dependent on dominant wavelength 1502a. Thus, each aerial image 1573a, 1573b, 1573c, 1573d is associated with a respective distinct energy 1571Ea, 1571Eb, 1571Ec, 1571Ed. In Figure 15B, each distinct energy 1571Ea, 1571Eb, 1571Ec, 1571Ed is represented by a different light and dark level within a respective spatial image 1573a, 1573b, 1573c, 1573d.

Referring again to FIG. 12 , to determine the target energy Etarget 1271E associated with the dominant wavelength of pulse k−1 of beam 1160, target energy generator 1270E may access information from optical source 1105 about the dominant wavelength of pulse k−1 or material. For example, if energy attribute signal 1246E is associated with pulse k-1, target energy generator 1270E may output target energy Etarget 1271E associated with the dominant wavelength of pulse k-1. As another example, the target energy generator 1270E may output the target energy Etarget 1271E based on the pulse number or index. For example, referring to FIGS. 15A and 15B , if pulse k−1 has a dominant wavelength 1502c (corresponding to the set of pulses specified by c), target energy generator 1270E determines that target energy Etarget 1271E for pulse k−1 is 1571Ec. On the other hand, if pulse k−1 has dominant wavelength 1502a (corresponding to the set of pulses specified by a), target energy generator 1270E determines target energy Etarget 1271E for pulse k−1 to be 1571Ea.

Additionally, the value of Etarget 1271E and/or an indication of the amount of optical energy in energy attribute signal 1246E may be processed prior to being received by comparator 1246E. For example, if the value of Etarget 1271E is in units of energy (joules) and the indication of the amount of optical energy in energy attribute signal 1246E is in units of power (watts), the indication may be converted to A unit of energy (joule). Comparator 1263E determines energy error 1266E associated with pulse k-1 of beam 1160, energy error 1266E corresponding to the difference between the amount of energy in pulse k-1 and Etarget 1271E.

Energy error 1266E is provided to energy controller 1272E, which evaluates excitation signal 1268E. The characteristic of excitation signal 1268E is based on energy error 1266E (which in turn is based on an indication of the amount of energy in energy property signal 1246E). In addition, energy controller 1272E corrects excitation signal 1268E to account for changes in the transfer function of optical oscillator 1212E. The transfer function varies due to the fact that the spectral properties (wavelengths) of the pulses in beam 1160 are intentionally not all the same. For example, the center or dominant wavelength of each pulse can be changed on a pulse-by-pulse basis before the pulse is generated, thereby changing the configuration of the spectral feature selection module 258 . The dominant wavelengths can be alternated between a plurality of values to form a pulsed beam 1160 having spectral peaks at each dominant wavelength, wherein any two peaks are separated from each other by a spectral distance which is the distance between the dominant wavelengths of the two peaks. Difference. At wavelengths between two adjacent dominant wavelengths, little or no light is present in the pulsed beam.

Corrected excitation signal 1268E is applied to optical oscillator 1212E to correct for variations in the efficiency of optical oscillator 1212E. By correcting the excitation signal 1268E, the energy control module 1261E causes the pulse energy of a particular dominant wavelength in the pulsed beam 1160 to be at or within an acceptable range of a target energy 1271E associated with that particular dominant wavelength.

Referring to FIG. 16, an implementation 1661E of an energy control module 1161E is shown for use with an optical oscillator 1212E. In this embodiment, the energy control module 1661E includes a plurality of energy controllers 1672E, one for each dominant wavelength λp. In this embodiment, two energy directors 1672E_1 and 1672E_2 are shown, one energy director for each of the two dominant wavelengths, such that the optical oscillator 1212E produces a first dominant wavelength λp1 and A pulse at a second dominant wavelength λp2 at a second target energy 1671E_2. In other embodiments, the energy control module 1661E may include more than two energy directors 1672E, and as many energy directors 1672E as there are dominant wavelengths in the light beam 1160 . Each of energy directors 1672E may have any suitable design or operation. Furthermore, any one energy director 1672E within the energy control module 1661E may have a different design or operation than the other energy directors 1672E within the energy control module 1661E.

Energy control module 1661E includes a set of comparators 1663E, one comparator for each energy controller 1672E. In the embodiment shown, the first comparator 1663E_1 is associated with the first energy controller 1672E_1 and the second comparator 1663E_2 is associated with the second energy controller 1672E_2. The energy control module 1661E also includes a target energy generator 1670E for generating a target energy Etarget for each dominant wavelength λp. Therefore, in this embodiment, the target energy generator 1670E generates the first target energy Etarget 1671E_1 for the first dominant wavelength λp1, the first target energy Etarget 1671E_1 is provided to the first comparator 1663E_1, and is generated for the second dominant wavelength λp2 The second target energy Etarget 1671E_2, the second target energy Etarget 1671E_2 is provided to the second comparator 1663E_1.

The energy control module 1661E includes a switch 1646Es configured to determine where to send the energy attribute signal 1646E. Specifically, if the current pulse has the first dominant wavelength λp1, the switch 1646Es provides the energy attribute signal 1646E to the first comparator 1663E_1, and if the current pulse has the second dominant wavelength λp2, the switch 1646Es provides the energy attribute signal 1646E 1646E is provided to the second comparator 1663E_2. In other embodiments, instead of a switch at the optical detection system 1145E, respective switches may be implemented at the comparators 1663E_1, 1663E_2.

An energy director, such as energy director 1272E configured to operate on all dominant wavelengths λp, or energy directors 1672E_1, 1672E_2 each configured to operate on a single dominant wavelength λp, may be of any suitable design or operation. Next, several embodiments of suitable energy controllers are discussed with reference to FIGS. 17-20 . Any of these energy directors may be implemented as any of energy directors 1272E, 1672E_1 or 1672E_2. Furthermore, it is possible to combine the operations of multiple energy directors into a single energy director in the energy directors 1272E, 1672E_1, 1672E_2.

Referring to Figure 17, an implementation of the energy controller 1772E uses a notch filter. The energy controller 1772E includes a delay module 1767E, an excitation determination module 1762E, and a correction module 1764E. Delay module 1767E receives energy error 1766E from comparator 1763E, which can be any one of comparators 1263E, 1663E_1, 1663E_2. Delay module 1767E introduces a time delay into energy error 1766E to ensure proper causality; so that actions taken by energy controller 1772E are not applied to received measured pulses (from energy attribute signal 1246E). While delay module 1767E is shown as a separate block, its functionality may be implemented within optical detection system 1145E or stimulus decision module 1762E.

Energy error 1766E is provided to excitation decision module 1762E, which determines excitation signal 1768Ep. The characteristic of the excitation signal 1768Ep is based on the energy error 1766E, which in turn is based on an indication of the amount of energy in the energy property signal 1246E, 1646E. Thus, for example, the excitation determination module 1762E may determine how much to adjust the voltage of the electrodes of the oscillator 1212E to offset the error in the energy of the beam output from the oscillator 1212E.

The excitation signal 1768Ep is provided to the correction module 1764E. Correction module 1764E determines corrected excitation signal 1768E based on excitation signal 1768Ep. In particular, correction module 1764E corrects excitation signal 1768Ep to account for variations in the transfer function of optical oscillator 1212E. The transfer function varies due to the fact that the spectral properties of the pulses in beam 1160 are intentionally not all identical. For example, the center or dominant wavelength λp of each pulse may be changed on a pulse-by-pulse basis before the pulse is generated, thereby changing the configuration of the spectral feature selection module 258 . The dominant wavelengths can be alternated between a plurality of values to form a pulsed beam 1160 having spectral peaks at each dominant wavelength, wherein any two peaks are separated from each other by a spectral distance which is the distance between the dominant wavelengths of the two peaks. Difference. At wavelengths between two adjacent dominant wavelengths, little or no light is present in the pulsed beam.

等式(2)， 其中k為將脈衝數目索引化之整數數目，V _sp為經校正激勵信號1768E，且特定言之，V _sp(k+1)為用於k+1脈衝之經校正激勵信號1768E，G _N為K _H/K _E，其中K _H為與劑量誤差有關之增益的調諧參數，K _E為與能量誤差有關之調諧參數或增益，且Vservo為根據如下等式(3)計算之電壓命令：

等式(3)， 其中e(k)為第k個脈衝之能量誤差1766E，D(k)為第k個脈衝之累積能量誤差或劑量誤差，且dEdV(k)為在光學振盪器1212E產生第k個脈衝時的轉移函數。 Correction module 1764E implements a filter, such as a notch filter, that determines corrected excitation signal 1768E based on at least the transfer function TF(k) at which the optical oscillator 1212E generates the kth pulse, the kth Energy error 1766E for pulses, cumulative energy error for the kth pulse, one or more values of the previous excitation signal for a pulse having the same dominant wavelength as the kth pulse, and one related to energy and/or dose error or multiple tuning parameters or gains. In general, a notch filter rejects signals with frequencies that are in-band and passes signals with frequencies that are out-of-band. The notch filter is configured to reject energy disturbances that may arise due to using optical pulses of different configurations (different transfer functions) from the optical oscillator 1212E. The notch filter can be expressed by the following equation:

Equation (2), where k is an integer number indexing the number of pulses, _Vsp is the corrected excitation signal 1768E, and in particular, _Vsp (k+1) is the corrected excitation for k+1 pulses Signal 1768E, G _N is K _H /K _E , wherein K _H is the tuning parameter of the gain related to the dose error, K _E is the tuning parameter or gain related to the energy error, and Vservo is calculated according to the following equation (3) The voltage command:

Equation (3), where e(k) is the energy error 1766E of the kth pulse, D(k) is the cumulative energy error or dose error of the kth pulse, and dEdV(k) is the energy error generated in the optical oscillator 1212E The transfer function at the kth pulse.

Corrected excitation signal 1768E is applied to optical oscillator 1212E to correct for variations in the efficiency of optical oscillator 1212E. By correcting the excitation signal 1268E, the energy control module 1261E makes the pulse energy of the specific dominant wavelength in the pulsed beam 1160 within the acceptable range of the target energy (such as Etarget 1271E, or Etarget 1671E_1, Etarget 1671E_2) associated with the specific dominant wavelength or within that acceptable range.

Referring to Figure 18, an energy controller implementation 1872E uses a Kalman filter that uses linear quadratic estimation. The energy controller 1872E includes a delay module 1867E that receives an error signal from a comparator 1863E, which can be any one of the comparators 1263E, 1663E_1, 1663E_2. As discussed above, the delay module 1767E introduces a time delay into the energy error 1866E in order to ensure proper causality; so that actions taken by the energy controller 1872E are not applied to the received measured pulse (from the energy attribute signal 1246E) . While delay module 1867E is shown as a separate block, its functionality may be implemented within another component of optical detection system 1145E or energy controller 1872E. The energy controller 1872E includes an excitation determination module 1862E, a correction module 1864E and a second comparator 1869E.

等式(4)， 其中k為表示光束1160之脈衝的脈衝數目的大於或等於1之整數，Ch _k為在光束1160中產生第k個脈衝之光學振盪器1212E的狀態，且dedv(Ch _k)為模型化產生第k個脈衝之光學振盪器1212E之轉移函數的模型M(TF)。V ^*及E ^*作為模型化之部分而判定。V(k+1)為針對k+1脈衝而判定之激勵信號1868Ep。 Similar to the stimulus decision module 1762E, the stimulus decision module 1862E decides the stimulus signal 1868Ep based on the energy error 1866E output from the delay module 1867E. In particular, excitation decision module 1862E includes a set of transfer function models, where each transfer function model is associated with a respective one of the states of optical oscillator 1212E. In particular, each transfer function TF of optical oscillator 1212E is associated with a particular configuration of spectral feature selection module 258 that produces a distinct dominant wavelength λp, and each transfer function TF is related to the efficiency characteristics of that configuration. The excitation decision module 1862E selects the model M(TF) associated with the transfer function TF of the optical oscillator 1212E generating the kth pulse for computing the excitation signal 1868Ep. In equation form, this can be represented by:

Equation (4), where k is an integer greater than or equal to 1 representing the number of pulses of the pulses of beam 1160, Ch _k is the state of optical oscillator 1212E generating the kth pulse in beam 1160, and dedv(Ch _k ) is the model M(TF) modeling the transfer function of the optical oscillator 1212E generating the kth pulse. V ^* and E ^* are determined as part of the modeling. V(k+1) is the excitation signal 1868Ep asserted for the k+1 pulse.

Correction module 1864E is implemented as a Kalman filter that efficiently rejects interpulse energy interference with a known period. Kalman filter 1864E uses energy error 1866E from comparator 1863E and excitation signal 1868Ep from excitation decision module 1862E to determine output signal 1864Eo. The output signal 1864Eo is provided to a second comparator 1869E. The second comparator 1869E determines the corrected excitation signal 1868E based on the output signal 1864Eo and the excitation signal 1868Ep.

The output signal 1864Eo of the Karman filter 1864E is based on a factor directly related to the energy error 1866E of the kth pulse, a model M(TF) associated with the period in which the configuration of the spectral feature selection module 258 changes, and applying An excitation signal 1868E for the kth pulse is generated. The output signal 1864Eo of the Kalman filter 1864E also takes into account the gain and tuning parameters of the Kalman filter 1864E.

等式(5)， 其中A = -1，且KXpost(k)為針對第k個脈衝之卡門濾波器1864E之輸出的估計且藉由以下給出： KXpost(k)=KXpred(k)+K_K(k)*Ke(k)          等式(6)， 其中K_K為卡門濾波器1864E之增益，藉由以下給出：

等式(7)， 且Ke(k)藉由以下給出：

等式(8)， 其中誤差(k)為第k個脈衝之能量誤差1866E，dedv(M(TF))為與用於在光束1160中產生第k個脈衝的光學振盪器1212E相關聯之模型，且HVcommand(k)為施加以產生第k個脈衝之激勵信號。 The output signal 1864Eo of the Karman filter can be expressed according to the following equation:

Equation (5), where A = -1, and KXpost(k) is the estimate of the output of the Kalman filter 1864E for the kth pulse and is given by: KXpost(k) = KXpred(k) + K_K (k)*Ke(k) Equation (6), where K_K is the gain of the Kalman filter 1864E, given by:

Equation (7), and Ke(k) is given by:

Equation (8), where error(k) is the energy error 1866E for the kth pulse and dedv(M(TF)) is the model associated with the optical oscillator 1212E used to generate the kth pulse in the beam 1160 , and HVcommand(k) is the excitation signal applied to generate the kth pulse.

等式(10)， 其中C為卡門濾波器1864E之調諧參數且在此實施方式中可等於1，且R為調諧參數。 K_S(k) is given by K_S(k) = KPpred(k) + R, where R is a tuning parameter. And, KPpred(k) is the covariance of KXpred(k). KPpred(k) can also be regarded as the confidence level for determining the gain K_K of the Kalman filter 1864E. Therefore, if KPpred(k) is 0, then K_K = 0, and this means that we are very confident in the model prediction and will not need the output from the optical detection system 1145E. On the other hand, if KPpred(k) is very large compared to the noise R in the optical detection system 1145E, then K_K = 1, and this means that one can only trust the optical detection system 1145E. KPpred(k+1) is given by: KPpred(k+1) = A * KPpost(k) *A' + Q Equation (9), where A = -1 and Q is the tuning of the Karman filter 1864E parameters, and KPpost(k) is given by:

Equation (10), where C is the tuning parameter of the Kalman filter 1864E and may be equal to 1 in this embodiment, and R is the tuning parameter.

The second comparator 1869E determines the corrected excitation signal 1868E as follows: HVSP(k) = HVCommand(k) + HVDefault - KXpred(k), Equation (11), Wherein HVSP(k) is the corrected excitation signal 1868E, HVCommand(k) is the uncorrected excitation signal 1868Ep determined by the excitation determination module 1862E for the kth pulse, and HVDefault is the estimated optical oscillator for the kth pulse Parameters of the nominal excitation signal of the transfer function TF of 1212E, and KXpred(k) is the output signal 1864Eo of the Kalman filter 1864E for the kth pulse. The value of HVDefault may be stored in electronic storage and retrieved by energy controller 1872E as needed. The value of HVDefault may be a magnitude of voltage, and may be a value greater than 100 volts, for example.

Referring to FIG. 19A , an embodiment 1961E of the energy control module 1161E uses a feedforward approach to reject or reduce interpulse energy disturbances or energy variations due to intentional changes associated with the optical oscillator 1212E. This occurs in conjunction with the configuration of the spectral feature selection module 258 to alter the spectral properties of the light beam 1160 produced by the optical oscillator 1212E. Energy control module 1961E relies on a set of estimates EvsV(λp), each estimate having an input (excitation signal or V) to optical oscillator 1212E and an output from optical oscillator 1212E (beam 1160 The relationship between the energy E). The energy control module 1961E includes a target energy generator 1970E (which operates similarly to the target energy generator 1270E), a comparator 1963E, and an energy controller 1972E.

Referring to Figures 19B and 19C, details of the energy controller 1972E are shown. The energy controller 1972E includes a delay module 1967E and an excitation determination module 1962E. The output of delay module 1967E is energy error 1966E output from comparator 1963E, which corresponds to a measure of the difference between energy attribute signal 1946E (which is the energy in the previous pulse, see FIG. 19A ) and energy target 1971E. The excitation determination module 1962E determines the corrected excitation signal 1968E and provides the corrected excitation signal 1968E to the optical oscillator 1212E.

Figure 19C is a block diagram of the stimulus determination module 1962E. The incentive determination module 1962E may include a feedback controller FC. In some embodiments, feedback controller FC is a proportional-integral-derivative (PID) controller that receives error signal 1966E and produces an output that is applied to one of downstream selected transfer functions (as discussed next ). For example, a PID controller includes a proportional term, an integral term, and a derivative term. Although a PID controller is discussed, any regenerative controller can be used as the regenerative controller FC.

The stimulus decision module 1962E includes a transfer function selector 1974E, which selects a transfer function TF(1), TF(2), . . . TF(N). Each of the transfer functions TF(1), TF(2)...TF(N) is an estimated transfer function of the optical oscillator 1212E for a particular dominant wavelength λp, and the transfer functions TF(1), TF( 2) Each of...TF(N) is associated with a particular configuration of spectral feature selection module 258 (FIG. 13). Spectral feature selection module 258 has N different configurations, each of which is associated with a different spectral parameter of output beam 1160 (eg, central or dominant wavelength or bandwidth). N is an integer number greater than one that indexes all possible configurations of the spectral feature selection module 258 that are relevant to a particular application. Each of the N configurations of spectral feature selection module 258 is associated with a respective transfer function TF(1), TF(2)...TF(N) of optical oscillator 1212E. For example, the index value of N associated with a particular one of the transfer functions TF(1), TF(2)...TF(N) can be obtained by defining the data of the transfer function TF and by the spectral feature selection model The central or principal wavelength λp resulting from this configuration of group 258 is stored in a lookup table or database. The transfer functions TF(1), TF(2), . . . TF(N) may be stored on electronic storage 252 and may be accessed by energy control module 1961E. The transfer functions TF(1), TF(2)...TF(N) may be associated with the N configurations by the manufacturer, or may be provided by the operator of the system 1100, or may use an input (discharge voltage) And the history of output (measured energy) is estimated and updated online.

The transfer function selector 1974E determines which of the transfer functions TF(1), TF(2)....TF(N) is associated with the configuration of the kth pulse generated from the output beam 1160 emitted by the optical oscillator 1212E couplet. Transfer function selector 1974E can select from among transfer functions TF(1), TF(2).....TF(N) by implementing a remainder function that returns the remainder of a division operation that divides k by M, where M is an integer representing the number of N configurations of spectral feature selection module 258 that alternate or cycle between generating optical pulses to generate optical pulses and k indexes the number of pulses. Thus, M is two, N, or any number greater than 2 and less than or equal to N. If transfer function selector 1974E is implemented as a remainder function and M=2, then transfer function selector 1974E returns 0 for pulses with even k-index numbers and 1 for odd k-index numbers. In these embodiments, transfer function TF(1) is selected when transfer function selector 1974E returns 0, and transfer function TF(2) is selected when transfer function selector 1974E returns 1.

In another example, the center or dominant wavelength λp of the optical pulses generated by the optical oscillator 1212E varies pulse-by-pulse according to a predetermined recipe. For example, optical oscillator 1212E and spectral feature selection module 258 may be controlled such that the dominant wavelength λp cycles in a sequential fashion among four predetermined dominant wavelengths such as those shown in FIGS. 15A and 15B . Thus, transfer function selector 1974E selects transfer function TF(2) for the second and sixth pulses, and selects transfer function TF(3) for the third and seventh pulses, and so on.

等式(12)， 其中k為大於或等於1且表示由光學振盪器1212E輸出之光束中的脈衝之脈衝數目的整數，λ _k為由光學振盪器1212E產生之第k個脈衝的波長，E為能量值，V為電壓值，且TF(λ _k)為在第k個脈衝中產生波長之光學振盪器1212E之轉移函數TF(1)、TF(2)…..TF(N)中之一者。V*及E*為經濾波版本，分別如原始電壓及能量值之移動平均值。 Error signal 1966E is provided to selected transfer function TF (via transfer function selector 1974E), and the output of selected transfer function TF is provided to gain 1984E and then to integrator 1985E. Forward correction signal 1967E is provided to integrator 1985E and is based on an EvsV curve selected based on the dominant wavelength of the next pulse of beam 1160 . The forward correction signal 1967E removes, reduces or rejects energy interference. Signal 1967E corrects for energy differences caused by changing the configuration of spectral feature selection module 258 and changing Etarget during operation of optical oscillator 1212E, and determines corrected excitation signal 1968E. Corrected excitation signal 1968E (V(k+1)) is determined based on the following equation:

Equation (12), wherein k is an integer greater than or equal to 1 and represents the pulse number of the pulse in the beam output by the optical oscillator 1212E, _λk is the wavelength of the k-th pulse generated by the optical oscillator 1212E, E is the energy value, V is the voltage value, and TF(λ _k ) is one of the transfer functions TF(1), TF(2)...TF(N) of the optical oscillator 1212E that generates the wavelength in the kth pulse one. V* and E* are filtered versions, like moving averages of raw voltage and energy values, respectively.

Referring to FIG. 20, an implementation 2061E of the energy control module 1161E uses a repetitive control method that relies on forwarding energy to invert or cancel any repetitive interference (shown as a pattern in the energy attribute signal 2046E), which Repeat interference occurs due to intentionally changing the configuration of spectral feature selection module 258 associated with optical oscillator 1212E in order to change the spectral properties of beam 1160 produced by optical oscillator 1212E. The energy control module 2061E includes a target energy generator 2070E (which operates similarly to the target energy generator 1270E), a comparator 2063E, and an energy controller 2072E.

To invert or cancel any repetitive disturbance, the energy controller 2072E requires deductive knowledge of the properties of the random repetitive disturbance to model the disturbance and make changes to the excitation signal 2068E in batches (i.e., for a set number of future pulses) . The energy controller 2072E may include a deductive module 2072ED configured to obtain this deductive knowledge by, for example, measuring the disturbance on each burst of pulses, and then use this information to generate a model of the disturbance. The energy controller 2072E may also include a correction module 2072EC configured to make changes to the excitation signal 2068E based on the disturbance model.

Energy Director 2072E relies heavily on direct observation of these disturbances; therefore, it is important to incorporate deductive knowledge about disturbances, including how disturbances vary or change. For example, if interference is known to vary significantly with the repetition rate of the pulses of beam 1160, it may be useful to ensure that the derivation module takes this dependency into account.

In some embodiments, the derivation module 2072ED within the energy controller 2072E may acquire the energy property signal 2046E over a certain period of time, such as for a set number of pulses in a burst of the light beam 1160 (referred to as a disturbance period). Deduction module 2072ED may then compare each of these energy attribute signals 2046E to target energy Etarget 2071E to thereby generate an energy error (which is part of energy signal 2066E) for each pulse in the disturbance cycle. The deduction module 2072ED can calculate how much of this energy error the other feedback controllers within the energy controller 2072E estimate have removed, and add it back to the measured energy error to obtain the total error. Other feedback controllers may include the controllers discussed herein. The remainder is the amount of energy error that the other feedback controllers will not remove or cannot remove and the energy derivation module 2072ED will remove. For each pulse during the glitch period, the deduction module 2072ED updates the magnitude and (alternatively) flag of the glitch, and the error is passed through an integrator (inside the deduction module 2072ED) with a gain of less than one to obtain the The interference shape. Each pulse in the interference period is processed independently, and no correlation between pulses is assumed. On the next disturbance cycle, the correction module 2072EC adds the latest disturbance shape to the excitation signal 2068E. This technique is repeated for each jamming cycle and results in training of the forward control.

Referring to FIG. 21 , the procedure 2100 is executed by the photolithography system 1100 . Routine 2100 determines a corrected input signal (excitation signal 1168E) for application to optical source 1105, and in particular optical oscillator 1212E. Process 2100 is implemented at least in part by control system 1150 including energy control module 1161E. Control system 1150 and/or portions of control system 1150, such as energy control module 1161E, may be implemented as part of optical source 1105, as part of lithographic exposure apparatus 1169, or in conjunction with optical source 1105 and/or lithographic exposure apparatus 1169 The two separate (but communicate with them).

A plurality of pulse sets of beam 1160 are generated (2105). In particular, optical source 1105 generates beam 1160 such that each pulse in beam 1160 is associated with a different dominant wavelength λp, and possibly also a different target energy Etarget. The distinct dominant wavelength λp is determined based on the configuration of the spectral feature selection module 258 .

Next, a measurement of the energy of the previous pulse of beam 1160 is received (2110). For example, control system 1150 (and in particular energy control module 1161E) receives energy attribute signal 1146E for the kth pulse (which may be considered a previous pulse) from detection system 1145E.

The energy error of the previous pulse is determined (2115). For example, referring to FIG. 12, the error signal 1266E is output from the comparator 1263E, or referring to FIG. 16, the error signal 1666E_1 is output from the comparator 1663E_1. If the kth pulse has a dominant wavelength λp2, a comparator (such as 1263E or 1663E_2) compares the measured energy of the previous beam pulse (determined from the energy attribute signal 1146E) with the second target energy Etarget2, which is due to the second target energy Etarget2 is associated with the second set of pulses. On the other hand, if the kth pulse has a dominant wavelength λp1, then a comparator (such as 1263E or 1663E_1) compares the measured energy of the previous beam pulse (determined from the energy attribute signal 1146E) with the first target energy Etarget1 because A first target energy Etarget1 is associated with a first set of pulses.

At least one component of optical source 1105 is adjusted to thereby adjust the energy of subsequent pulses of beam 1160 (2120) based on the determined error (2115). The adjusted subsequent pulse has the same dominant wavelength as that of the previous pulse. Thus, for example, if spectral feature selection module 258 is configured to generate pulses with a first dominant wavelength λp1 between generating pulses with a second dominant wavelength λp2 (such as shown in FIGS. 7 and 8A ) Alternately, and the previous pulse (which is the kth pulse) has the first dominant wavelength λp1, then the k+2x pulse (where x is a positive integer) is adjusted since the k+2x pulse also has the first dominant wavelength λp1.

In particular, corrected excitation signal 1168E is applied to optical oscillator 1212E. For example, the voltage of electrodes 217-a, 217-b is adjusted based on this corrected excitation signal 1168E.

Embodiments can be further described using the following terms: 1. A method for controlling the energy of a pulsed light beam, the method comprising: generating a plurality of blended sets of pulses from a beam of light from the optical source, each set of beam pulses associated with a different dominant wavelength and a different target energy; Measurement of the energy of previous pulses of the received beam; determining an energy error comprising, if the previous beam pulse is in the particular set of beam pulses, comparing the measured energy of the previous beam pulse to a particular target energy associated with the particular set of beam pulses; and At least one component of the optical source is adjusted to thereby adjust the energy of subsequent pulses in a particular set of beam pulses based on the determined energy error. 2. The method of clause 1, further comprising receiving each distinct target energy associated with each set of beam pulses. 3. The method of clause 1, further comprising classifying whether a previous beam pulse is in the particular set of beam pulses. 4. The method of clause 1, further comprising determining an adjustment amount to at least one component of the optical source. 5. The method of clause 1, further comprising correcting an adjustment to at least one component of the optical source based on whether a previous beam pulse was in a particular set of beam pulses. 6. The method of clause 1, wherein adjusting at least one component of the optical source comprises varying a voltage provided to an electrode associated with an optical oscillator of the optical source. 7. A system comprising: Optical source equipment comprising: an optical oscillator configured to generate an optical pulse in response to an excitation signal, the optical pulse having spectral properties; and a spectral tuning device configured to control the spectral properties of the light pulses; and An energy control device in communication with the optical source device, the energy control device configured to: Determining target energies associated with spectral properties of the generated light pulses; and An adjustment to the excitation signal is determined based at least on the determined target energy, the adjustment causing the optical oscillator to generate one or more subsequent light pulses to account for changes in the configuration of the spectral adjustment device. 8. The system of clause 7, wherein the adjustment of the excitation signal results in an adjustment of the energy of one or more subsequently generated light pulses. 9. The system of clause 7, wherein the target energy associated with the spectral property of the generated light pulse was previously defined as being associated with the spectral property of the generated light pulse. 10. The system of clause 7, wherein the optical oscillator is associated with a plurality of transfer functions, each transfer function associated with a specific configuration of the spectral tuning device and a specific value of a spectral property; and the energy control device is configured to Adjustments to the excitation signal are determined based on a transfer function associated with the particular configuration of the spectral adjustment device used to generate the one or more subsequent light pulses. 11. The system of clause 7, wherein the spectral adjustment device comprises at least one smelt and a diffractive element arranged in optical communication with each other, and each transfer function is associated with a different state of at least one smelt. 12. The system of clause 7, wherein the spectral property of the light pulse is the center wavelength of that light pulse, and each configuration of the spectral adjustment device corresponds to a specific value of the wavelength. 13. The system of clause 7, further comprising a measurement device configured to measure the energy of the light pulse. 14. The system of clause 13, wherein the energy control device is configured to determine the energy error by comparing the target energy to the measured energy, and the determination of the adjustment to the excitation signal is also based on the energy error. 15. The system of clause 7, wherein the energy control device is configured to determine by determining an adjustment to the excitation signal that causes the optical oscillator to produce one or more subsequent light pulses associated with the spectral properties of the generated light pulse. An adjustment to the excitation signal to cause the optical oscillator to generate one or more subsequent light pulses is determined. 16. The system of clause 7, wherein the energy control device is configured to determine a target energy associated with a spectral property of the generated light pulse by receiving a communication from a lithography exposure device configured to receive the light pulse, The communication provides a set of target energies, each target energy in the set being associated with a spectral property. 17. An energy control device comprising: A control module configured to: received energy values of previous light pulses emitted from the optical source; Perform a comparison, which includes: comparing the received energy value to the first target energy only if the previous light pulse was in the first set of beam pulses associated with the first dominant wavelength; or comparing the received energy value to a second target energy different from the first target energy only if the previous light pulse was in a second set of light beam pulses associated with a second dominant wavelength different from the first dominant wavelength; and At least one component of the optical source is adjusted based on the comparison to thereby adjust the energy of a subsequent light pulse having a dominant wavelength associated with the previous light pulse. 18. The energy control apparatus of clause 17, wherein the control module comprises a classification module configured to classify whether the previous light pulses are in the first set of beam pulses or in the second set of beam pulses. 19. The energy control device of clause 17, wherein the control module includes a comparator configured to determine whether the previous light pulse was in the first set of beam pulses or the second set of beam pulses, and based on the determination to provide The first target energy or the second target energy. 20. The energy control device of clause 17, wherein the control module includes a signal module configured to determine an amount of adjustment to be made to at least one component of the optical source. 21. The energy control apparatus of clause 17, wherein the control module comprises a correction module configured to correct the optical beam to be treated based on whether the previous light pulse was in the first set of beam pulses or the second set of beam pulses The amount of adjustment made to at least one component of the source. 22. The energy control apparatus of clause 21, wherein the correction module is configured to correct the adjustment amount by applying a filter to the adjustment amount. 23. The energy management device of clause 22, wherein the filter comprises a notch filter that transmits information having frequencies in the first frequency band and substantially blocks information having frequencies outside the first frequency band Information. 24. The energy control device of clause 22, wherein the filter comprises a Kalman filter. 25. The energy control apparatus of clause 21, wherein the correction module is configured to correct the adjustment amount by applying a feed-forward correction to the adjustment amount. 26. The energy control device of clause 17, wherein the control module is configured to adjust at least one component of the optical source based on the comparison to thereby adjust the energy of a subsequent light pulse having a dominant wavelength associated with a previous light pulse comprising A signal is sent to the optical source to thereby vary a voltage provided to an electrode associated with an optical oscillator of the optical source. 27. The energy control device of clause 17, wherein the control module is configured to receive energy values of previous light pulses comprises the control module configured to receive energy values of a plurality of previous light pulses emitted from the optical source. 28. The energy control device of clause 17, wherein the control module is configured to adjust at least one component of the optical source based on the comparison to thereby adjust a plurality of subsequent light pulses having a dominant wavelength associated with a previous light pulse of energy. 29. The energy control apparatus of clause 17, wherein the control module is configured to maintain the energy of a subsequent light pulse that does not have a dominant wavelength associated with a previous light pulse based on the comparison.

Other implementations are within the scope of the patent application.

100: Photolithography system 105: Optical source 150: Control system 160: Beam 169:Lithographic exposure equipment 170: Wafer 171: wafer stage 172: Weights and measures system 173a: Spatial imagery 173b: Spatial imagery 174: mask 175:Projection optical system 176: Slit 177: lens 178a: part 178b: part 179: separation distance 200: Photolithography system 205: Optical source 212: master oscillator 214: discharge chamber 216: Line narrowing module 216C: Line narrowing module 217: electrode 217-a: Cathode 217-b: anode 218: output coupler 219: Gain medium 220: Line center analysis module 222: Beam coupling optical system 224: Seed Beam 230: power amplifier 232: Beam coupling optical system 240: discharge chamber 241: electrode 248: Beam steering optics 250: Control system 251: electronic processor 252: Electronic storage 253: I/O interface 254: Spectral feature control module 255_1~255_n: spectral feature actuation system 256_1~256_n: optical features 257: Optical system 258: Spectral feature selection module 259: Recipe file 260: Beam 262:Bandwidth analysis module 263: Beam preparation system 291: Reflective grating 292: 稜鏡 293: 稜鏡 294: 稜鏡 295: 稜鏡 300: wafer exposure signal 305: high value 307: time period 310: low value 315: gate signal 316:burst 320: high value 325: low value 330: trigger signal 340:Trigger 400: window 429: Illuminator system 432: Objective Lens Configuration 479: exposure field 500: process 510: Operation 520: Operation 530: Operation 540: Operation 600A: Pulse 600B: Pulse 601A: Spectrum 601B: Spectrum 602A: Dominant wavelength 602B_1: Dominant wavelength 602B_2: Dominant wavelength 701: average spectrum 702_1: The first dominant wavelength 702_2: Second dominant wavelength 703:Spectrum separation 704: spectral region 873: Spectral Separation 873a: Spatial imagery 873b: Spatial imagery 875: Focus separation distance 878a: plane 878b: plane 879: separation distance 992: side wall angle 995: 3D semiconductor components 996: Concave 997: Bottom 998: side wall 999: Surroundings 1001: Plotting 1002: Plotting 1003: Plotting 1004: Plotting 1005: Plotting 1006::plot 1100: Photolithography system 1105: optical source 1145E: Optical Detection System 1146E: Energy attribute signal 1150: Control system 1160: pulse beam 1160E: Energy Control Equipment 1161E: Energy Control Module 1165: Arrow 1168E: Excitation signal 1168Ev: voltage signal 1169: Lithography Exposure Equipment 1197E: Source Provider 1212E: Optical oscillator 1246E: Energy attribute signal 1261E: Energy Control Module 1263E: Comparator 1266E: error signal 1268E: Excitation signal 1270E: Target Energy Generator 1271E: target energy 1271E_i: target energy 1272E: Energy Controller 1402_i: dominant wavelength 1502: Dominant wavelength 1502a: dominant wavelength 1502b: Dominant wavelength 1502c: dominant wavelength 1502d: dominant wavelength 1571E: target energy 1571Ea: target energy 1571Eb: target energy 1571Ec: target energy 1571Ed: Target energy 1573a: Spatial imagery 1573b: Spatial imagery 1573c: Spatial imagery 1573d: Spatial imagery 1578a: plane 1578b: plane 1578c: plane 1578d: plane 1646E: Energy attribute signal 1646Es: Switcher 1661E: Energy Control Module 1663E: Comparator 1663E_1: The first comparator 1663E_2: second comparator 1666E_1: error signal 1670E: Target Energy Generator 1671E_1: The first target energy 1671E_2: Second target energy 1672E: Energy Controller 1672E_1: Energy Controller 1672E_2: Energy Controller 1762E: Incentive Judgment Module 1763E: Comparator 1764E: Calibration module 1766E: Energy error 1767E: delay module 1768E: Corrected excitation signal 1768Ep: Excitation signal 1772E: Energy Controller 1862E: Incentive Judgment Module 1863E: Comparator 1864E: Calibration module 1864Eo: output signal 1866E: Energy error 1867E: Delay Module 1868E: Corrected excitation signal 1868Ep: Excitation signal 1869E: second comparator 1872E: Energy Controller 1946E: Energy attribute signal 1961E: Energy Control Module 1962E: Incentive Judgment Module 1963E: Comparator 1966E: Energy error 1967E: Delay Module / Forward Correction Signal 1968E: Corrected excitation signal 1970E: Target Energy Generator 1971E: Energy Target 1972E: Energy Controller 1974E: Transfer function selector 1984E: Gain 1985E: Integrator 2046E: Energy attribute signal 2061E: Energy Control Module 2063E: Comparator 2066E: Energy Signal 2068E: Excitation signal 2070E: Target Energy Generator 2071E: target energy 2072E: Energy Controller 2072EC: Calibration module 2072ED: Deduction Module 2100: Procedure 2105: Operation 2110: Operation 2115: Operation 2120: Operation Epulse: pulse energy Etarget: target energy Etarget1: the first target energy Etarget2: Second target energy EvsV(λp): estimate FC: Feedback Controller TF(1): transfer function TF(2): transfer function TF(N): transfer function λp: dominant wavelength λp1: the first dominant wavelength λp2: second dominant wavelength

1A is a block diagram of an example of an implementation of a photolithography system.

Figure IB is a block diagram of an example of an implementation of an optical system for the photolithography system of Figure IA.

1C is a cross-sectional view of an example of a wafer exposed by the photolithography system of FIG. 1A.

2A is a block diagram of another example of an implementation of a photolithography system.

2B is a block diagram of an example of an implementation of a spectral feature selection module that may be used in a photolithography system.

2C is a block diagram of an example of an implementation of a line narrowing module.

3A, 3B and 3C are plots of data related to pulse generation and/or pulse bursts in an optical source.

4 is a block diagram of another example of an implementation of a photolithography system.

5 is a flowchart of an example of a process for forming a three-dimensional semiconductor device.

6A and 6B each show an example of the spectrum of a single light pulse.

Figure 7 shows an example of an averaged spectrum for a single exposure pass.

8A and 8B show side and top cross-sectional views, respectively, of an example of a wafer.

9A and 9B show side and top cross-sectional views, respectively, of an example of a three-dimensional semiconductor device.

Figures 10A and 10B show examples of simulated data.

11A is a block diagram of a photolithography system in which the control system includes an energy control module configured to provide an excitation signal to an optical source for controlling electrodes within an optical oscillator of the optical source.

11B is a diagram showing an example of a transfer function TF (optical energy generated by a single optical oscillator that varies with supplied excitation energy) of an optical oscillator showing how optical energy varies with the wavelength of the emitted pulsed beam .

12 is a block diagram of an embodiment of the energy control module of FIG. 11A for use with an optical oscillator.

FIG. 13 is a block diagram of an embodiment of a master oscillator that may constitute an optical oscillator.

14 is a table showing the correlation between each target energy and each possible dominant wavelength of a light beam output from an optical source comprising an optical oscillator.

15A is a graph of target energy sets for each of the four dominant wavelengths of a light beam output from an optical source comprising an optical oscillator.

15B is a cross-sectional view of an example of a wafer exposed in the photolithography system of FIG. 11 in which light beams output from an optical source including an optical oscillator are generated at the four dominant wavelengths provided in FIG. 15A.

FIG. 16 is a block diagram of an embodiment of the energy control module of FIG. 11A for use with an optical oscillator and including a plurality of energy directors, each associated with a master of a light beam output from an optical source including an optical oscillator. related to the wavelength.

FIG. 17 is a block diagram showing an embodiment of an energy controller that may be used in the energy control module of any one or more of FIGS. 11 , 12 and 16 .

FIG. 18 is a block diagram showing an embodiment of an energy controller that may be used in the energy control module of any one or more of FIGS. 11 , 12 and 16 .

19A is a block diagram of an embodiment of the energy control module of FIG. 11A for use with an optical oscillator and including a feed-forward energy controller.

FIG. 19B is a block diagram of an embodiment of an energy controller prior to FIG. 19A.

FIG. 19C is a block diagram of an implementation of the excitation determination module of the feedforward energy controller of FIG. 19B .

20 is a block diagram of an embodiment of the energy control module of FIG. 11A for use with an optical oscillator and including a repetitive control energy controller.

FIG. 21 is a flowchart of a process executed by the photolithography system of FIG. 11A.

1502: Dominant wavelength

1502a: dominant wavelength

1502b: Dominant wavelength

1502c: dominant wavelength

1502d: dominant wavelength

1571E: target energy

1571Ea: target energy

1571Eb: target energy

1571Ec: target energy

1571Ed: Target Energy

Claims (29)

A method for controlling the energy of one of a pulsed light beam, the method comprising: generating a plurality of blended sets of pulses of the beam from an optical source, each set of beam pulses associated with a distinct dominant wavelength and a distinct target energy; receiving a measure of the energy of a previous pulse of the beam; determining an energy error comprising, if the previous beam pulse is in a particular set of beam pulses, comparing the measured energy of the previous beam pulse to a particular target energy associated with the particular set of beam pulses; and At least one component of the optical source is adjusted to thereby adjust the energy of a subsequent pulse in the particular set of beam pulses based on the determined energy error. The method of claim 1, further comprising receiving each distinct target energy associated with each set of beam pulses. The method of claim 1, further comprising classifying whether the previous beam pulse is in the particular set of beam pulses. The method of claim 1, further comprising determining an adjustment amount of the at least one component of the optical source. The method of claim 1, further comprising correcting the adjustment amount of the at least one component of the optical source based on whether the previous beam pulse is in the particular set of beam pulses. The method of claim 1, wherein adjusting at least one component of the optical source comprises varying a voltage supplied to an electrode associated with an optical oscillator of the optical source. A system comprising: An optical source device comprising: an optical oscillator configured to generate an optical pulse in response to an excitation signal, the optical pulse having a spectral property; and a spectral tuning device configured to control the spectral property of the light pulse; and an energy control device in communication with the optical source device, the energy control device configured to: determining a target energy associated with the spectral property of the generated light pulse; and An adjustment to the excitation signal is determined based at least on the determined target energy, the adjustment causing the optical oscillator to generate one or more subsequent light pulses to account for a change in a configuration of the spectral adjustment device. The system of claim 7, wherein the adjustment of the excitation signal causes an adjustment of the energy of one or more subsequently generated light pulses. The system of claim 7, wherein the target energy associated with the spectral property of the generated light pulse was previously defined as being associated with the spectral property of the generated light pulse. The system of claim 7, wherein the optical oscillator is associated with a plurality of transfer functions, each transfer function associated with a specific configuration of the spectral tuning device and a specific value of the spectral property; and the energy control device configured to determine the adjustment of the excitation signal based on the transfer function associated with the particular configuration of the spectral adjustment device used to generate the one or more subsequent light pulses. The system according to claim 7, wherein the spectral adjustment device comprises at least one indium and a diffractive element arranged in optical communication with each other, and each transfer function is associated with a different state of the at least one indium. The system according to claim 7, wherein the spectral property of an optical pulse is a central wavelength of the optical pulse, and each configuration of the spectral adjustment device corresponds to a specific value of the wavelength. The system of claim 7, further comprising a measurement device configured to measure an energy of the light pulse. The system of claim 13, wherein the energy control device is configured to determine an energy error by comparing the target energy and the measured energy, and the determination of the adjustment to the excitation signal is also based on the energy error . The system of claim 7, wherein the energy control device is configured to cause the optical oscillator to generate one or more subsequent light pulses associated with the spectral property of the generated light pulse by determining the excitation signal The adjustment of the excitation signal to cause the optical oscillator to generate the one or more subsequent light pulses is determined. The system of claim 7, wherein the energy control device is configured to determine the spectral property associated with the generated light pulse by receiving a communication from a lithography exposure device configured to receive the light pulse The communication provides a set of target energies for which each target energy in the set is associated with a spectral property. An energy control device comprising: a control module configured to: receiving an energy value of a previous light pulse emitted from an optical source; Perform a comparison that includes: comparing the received energy value to a first target energy only if the previous light pulse was in a first set of beam pulses associated with a first dominant wavelength; or comparing the received energy value to one different from the first target energy only if the previous light pulse was in a second set of beam pulses associated with a second dominant wavelength different from the first dominant wavelength the second target energy; and At least one component of the optical source is adjusted based on the comparison to thereby adjust the energy of a subsequent light pulse having the dominant wavelength associated with the previous light pulse. The energy control device of claim 17, wherein the control module includes a class module configured to classify whether the previous light pulse is in the first set of beam pulses or in the second set of beam pulses . The energy control device of claim 17, wherein the control module includes a comparator configured to determine whether the previous light pulse is in the first set of beam pulses or the second set of beam pulses, and based on the Determine to provide the first target energy or the second target energy. The energy control device of claim 17, wherein the control module includes a signal module configured to determine an amount of adjustment to be made to the at least one component of the optical source. The energy control device of claim 17, wherein the control module includes a calibration module configured to correct based on whether the previous light pulse was in the first set of beam pulses or the second set of beam pulses The amount of adjustment to be made to the at least one component of the optical source. The energy control device according to claim 21, wherein the correction module is configured to correct the adjustment amount by applying a filter to the adjustment amount. The energy control device of claim 22, wherein the filter comprises a notch filter that transmits information having a frequency in a first frequency band and substantially blocks information having a frequency outside the first frequency band One frequency information. The energy control device of claim 22, wherein the filter comprises a Karman filter. The energy control device of claim 21, wherein the correction module is configured to correct the adjustment amount by applying a forward correction to the adjustment amount. The energy control device of claim 17, wherein the control module is configured to adjust at least one component of the optical source based on the comparison to thereby adjust the subsequent light having the dominant wavelength associated with the preceding light pulse The energy of the pulse includes sending a signal to the optical source to thereby vary a voltage supplied to electrodes associated with an optical oscillator of the optical source. The energy control device of claim 17, wherein the control module is configured to receive the energy value of the previous light pulse comprising the control module configured to receive the energy value of a plurality of previous light pulses emitted from the optical source Energy value. The energy control device of claim 17, wherein the control module is configured to adjust the at least one component of the optical source based on the comparison to thereby adjust a complex number having the dominant wavelength associated with the previous optical pulse The energy of subsequent light pulses. The energy control device of claim 17, wherein the control module is configured to maintain the energy of a subsequent light pulse that does not have the dominant wavelength associated with the previous light pulse based on the comparison.

Images