High throughput multi-beam charged particle inspection system with dynamic control

Abstract

To provide a charged particle system and an operation method of a charged particle system that allow high throughput inspection of integrated semiconductor features with the resolution of at least the critical dimension during the development or during manufacturing or for reverse engineering of semiconductor devices.SOLUTION: There are provided a multi-beam charged particle inspection system and a method of operating a multi-beam charged particle inspection system for wafer inspection with high throughput and with high resolution and high reliability. The method and the multi-beam charged particle beam inspection system are configured to extract, from a plurality pieces of sensor data, a set of control signals to control the multi-beam charged particle beam inspection system and thereby maintain the imaging specifications including a movement of a wafer stage during a wafer inspection task. In particular, the system is operated such that a time interval Tr overlaps with a time interval Ts1 and / or Ts2 for image acquisition.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a multi-beam charged particle inspection system and a method for operating the same, and more particularly to a multi-beam charged particle inspection system, related methods, and computer program products for high-throughput, high-resolution, and reliable wafer inspection, the method and the multi-beam charged particle inspection system being configured to derive a set of control signals from a plurality of sensor data to control the multi-beam charged particle inspection system. [Background technology]

[0002] The continued miniaturization and high functionality of microstructures, such as semiconductor devices, necessitates further development and optimization of planar processing techniques and inspection systems for processing and inspecting critical dimensions of microstructures. Semiconductor device development and processing require, for example, design verification of test wafers, and planar processing techniques involve process optimization for reliable, high-throughput processing. In recent years, analysis of semiconductor wafers has become necessary for reverse engineering and individual, customized configuration of semiconductor devices. Therefore, there is a need for high-throughput inspection tools that can accurately examine microstructures on wafers.

[0003] Typical silicon wafers used in semiconductor device manufacturing are up to 12 inches (300 mm) in diameter. Each wafer is divided into 30–60 repeating areas ("dies"), each measuring up to approximately 800 square mm. Semiconductors contain multiple semiconductor structures fabricated in layers on the surface of the wafer using planar integration techniques. Semiconductor wafers typically have flat surfaces due to the fabrication processes involved. Feature sizes of integrated semiconductor structures range from a few microns to 5 nm critical dimensions (CDs), with feature sizes expected to shrink further in the near future to feature sizes or critical dimensions of less than 3 nm (e.g., 2 nm) or even 1 nm. These small feature sizes necessitate identifying defects the size of the critical dimensions over very large areas in a short time.

[0004] It is therefore an object of the present invention to provide a charged particle system and a method of operating a charged particle system that allows high-throughput inspection of integrated semiconductor features with at least critical dimension resolution during development or manufacturing of semiconductor devices or in reverse engineering, and also allows for the acquisition of high-resolution images of a set of specific locations on a wafer (e.g., so-called process control monitors PCMs or only critical areas).

[0005] A recent development in the field of charged particle microscopes (CPM) is the multibeam scanning electron microscope (MSEM). Multibeam charged particle microscopes are disclosed, for example, in U.S. Pat. No. 7,244,949, U.S. Patent Application Publication No. 20190355545, or U.S. Patent Application Publication No. US20190355544. In a multibeam charged particle microscope, such as a multibeam electron microscope or MSEM, a sample is irradiated with an array of electron beamlets, including, for example, 4 to 10,000 electron beams as primary radiation, with each electron beam separated from adjacent electron beams by a distance of 1 to 200 micrometers. For example, in an MSEM, approximately 100 separated electron beams, or beamlets, are arranged in a hexagonal array and separated by a distance of approximately 10 micrometers. Multiple primary charged particle beamlets are focused by a common objective lens onto the surface of the sample under investigation (e.g., a semiconductor wafer secured to a wafer chuck mounted on a movable stage). During irradiation of the wafer surface by the primary charged particle beamlets, interaction products (e.g., secondary electrons) are generated from multiple intersection points formed by the focal points of the primary charged particle beamlets, while the amount and energy of the interaction products depend on the material composition and topography of the wafer surface. The interaction products form multiple secondary charged particle beamlets, which are collected by a common objective lens and guided onto a detector arranged at a detector plane by a projection imaging system of the multi-beam inspection system. The detector includes multiple detection areas, each with multiple detection pixels, to detect the intensity distribution of each of the multiple secondary charged particle beamlets, resulting in an image patch of, for example, 100 μm × 100 μm.

[0006] Prior art multi-beam charged particle microscopes include a series of electrostatic and magnetic elements. By adjusting at least some of the electrostatic and magnetic elements, the focal position and stigma of multiple secondary charged particle beams can be adjusted. For example, U.S. Pat. No. 10,535,494 proposes readjusting a charged particle microscope when the detected intensity distribution of the focal points of secondary charged particle beamlets deviates from a predetermined intensity distribution. Adjustment is achieved when the detected intensity distribution conforms to the predetermined intensity distribution. Global displacement or deformation of the intensity distribution of secondary charged particle beamlets can lead to conclusions about topography effects, sample shape or tilt, or sample charging effects. U.S. Pat. No. 9,336,982 discloses a secondary charged particle detector with a scintillator plate that converts secondary charged particles into light. To reduce the decrease in the conversion efficiency of the scintillator plate, the relative lateral positions of the focal spots of multiple secondary charged particle beamlets and the scintillator plate can be varied, for example, by a charged particle beam deflector or an actuator for lateral displacement of the scintillator plate.

[0007] A prior art multi-beam charged particle microscope includes at least one crossover plane of primary or secondary charged particle beamlets. A prior art multi-beam charged particle microscope includes a detection system and method that facilitates alignment.

[0008] It is generally desirable to change the imaging settings of a charged particle microscope. A method for changing the image acquisition settings of a multi-beam charged particle microscope from a first imaging setting to a second, different imaging setting is described in U.S. Pat. No. 9,799,485.

[0009] However, in charged particle microscopes for wafer inspection, it is desirable to maintain stable imaging conditions so that reliable and reproducible imaging can be performed. Throughput is determined by several parameters (e.g., the speed of the stage and repositioning at a new measurement site) as well as the measurement area per acquisition time itself. The latter is determined by the dwell time, resolution, and number of beamlets. The wafer is moved laterally to the next point of interest by the wafer stage between the acquisition of two image patches. The movement and precise alignment of the stage to the next position for image acquisition are among the factors limiting the throughput of a multi-beam inspection system. During high-throughput image acquisition, unnecessary stage movement or drift can degrade image resolution. During high-throughput image acquisition, drift and deviation of the predetermined paths of the primary and secondary charged particle beams can adversely affect the image quality and reliability of the measurement results. For example, raster configuration within a planar area segment can degrade the number of primary charged particle beamlets and change the resolution of a multi-beam charged particle inspection system.

[0010] Single-beam electron microscopes typically use a so-called beam error function (BEF) to improve the positioning accuracy of the electron beam and stage movement. For this purpose, the BEF provides feedback from a stage holding the sample in the beam deflection system. A recent example is described in International Publication WO 2020 / 136094. However, while multi-beam charged particle microscopes are more complex, the simple approach of single-beam electron microscopes is insufficient. For example, conventional techniques cannot compensate for the rotation of the multiple focal points of the multiple primary charged particle beamlets relative to the wafer stage. Furthermore, multi-beam charged particle microscopes have a projection imaging system for imaging the multiple secondary electron beamlets onto the detector, so accurate imaging of the multiple secondary electrons must be maintained. Aberrations in the secondary beam paths must also be considered separately.

[0011] U.S. Patent No. 9,530,613 describes a method for focus control of a multi-beam charged particle microscope. A subset of multiple charged particle beamlets is astigmatized and used to detect shifts in focus position. From the elliptical shape of each of the astigmatized beamlets, an error signal is generated, and the vertical position of the sample stage is adjusted or the current through one or more lenses of the charged particle microscope is changed. This optimizes the focal spots of the multiple charged particle beamlets. This method is performed in parallel with the normal operation of the scanning electron microscope. However, this method only provides a feedback loop for focus control; it does not provide predictive control or take into account sensor signals from a stage position sensor.

[0012] U.S. Patent Application Publication No. 20190355544 and U.S. Patent Application Publication No. 20190355545 disclose a multi-beam charged particle microscope with an adjustable projection system that compensates for sample charging during scanning. Therefore, the projection system is configured with high-speed electrostatic elements that maintain proper imaging of secondary charged particle beamlets from the sample to the detector. Both documents use an image detector that analyzes the imaging quality of the secondary beamlets and compensates for degradation due to sample charging in the secondary electron beam path. Both documents describe methods and apparatus for controlling the secondary electron beam path when the secondary electron beamlets originate at the sample surface. However, as a subject of the present invention, error sources also exist in the primary beam path, causing degradation of the spot position and spot shape of multiple primary charged particle beamlets on the substrate surface. Furthermore, additional error sources can result in positioning errors or movements of the substrate table, resulting in aberrations in the acquired digital image of the object without degradation of the primary or secondary beam paths. These additional aberrations and errors can be variable on different time scales (e.g., slow-varying drifts such as thermal drift). Another example is fast-varying dynamic aberrations, e.g., due to acoustic vibrations. These errors cannot be compensated for by means in the secondary beam path alone. It is an object of the present invention to provide a multi-beam charged particle inspection system with means for enabling high-throughput and reliable high-precision and high-resolution image acquisition. It is an object of the present invention to provide a multi-beam charged particle inspection system with a high-speed stage including means for maintaining the lateral positions and focal points of multiple primary charged particle beamlets in a predetermined raster configuration with a predetermined positional accuracy, even if the time for precise stage alignment is reduced. It is an object of the present invention to provide a multi-beam charged particle inspection system with means for maintaining high resolution and high image contrast during high-throughput and reliable image acquisition of a series of image patches. It is an object of the present invention to provide a high-throughput and reliable multi-beam charged particle inspection system with a stage for moving a wafer from a first inspection site to a second inspection site.It is an object of the present invention to provide a multi-beam charged particle inspection system that provides means for compensating for drift in predetermined primary and secondary charged particle beam paths as well as stage motion (eg, parasitic stage motion).

[0013] Stage movement, including stage acceleration, deceleration, and ringdown, is one of the factors limiting the throughput of multi-beam inspection systems. Acceleration and deceleration of the stage over short periods of time requires complex and expensive stages. It is an object of the present invention to provide a multi-beam charged particle inspection system that provides a means for high-throughput, reliable, and highly accurate high-resolution image acquisition using stages with reduced technical complexity and cost.

[0014] Generally, it is an object of the present invention to provide a multi-beam charged particle inspection system for wafer inspection that provides reliable, high-throughput, high-accuracy, high-resolution image acquisition. Summary of the Invention

[0015] An embodiment of the present invention achieves the object of the present invention by a multi-beam charged particle microscope having a set of compensators that compensate for changes in error amplitudes during image acquisition of an image patch. The multi-beam charged particle microscope includes a plurality of detectors or sensors that provide a plurality of sensor data and extract a set of actual error amplitudes for a predetermined set of normalized error vectors from the plurality of sensor data. The derivation of the normalized error vector allows for separation of contributions from different error sources, including error sources in a primary charged particle beam path, a secondary electron beam path, and a stage position. The multi-beam charged particle microscope includes a control unit that derives drive signals to drive a set of compensators that keep the actual error amplitudes below a predetermined threshold during image acquisition of a digital image of the image patch by compensating for a set of error amplitudes corresponding to a set of imaging aberrations. Drive signals for the set of compensators, including at least one of a first compensator in the primary charged particle beam path and a second compensator in the secondary electron beam path, are derived from the normalized error vector representing the contributions from the different error sources. Alternative compensators include computational image post-processing of the acquired digital image or a compensator in the wafer stage.

[0016] In one example, the multi-beam charged particle microscope is configured to predict a change in at least one error amplitude of the set of error amplitudes and, in response, provide a corresponding drive signal to the set of compensators. In one example, the plurality of sensor data includes data from a stage position sensor or a stage acceleration sensor. In one example, the set of compensators includes first and second deflection systems or deflection scanners of the multi-beam charged particle microscope. In another example, the set of compensators includes a third deflection system in a detection unit of the multi-beam charged particle microscope. In one example, the set of compensators further includes at least a high-speed electrostatic compensator or a multi-aperture active array element.

[0017] According to one embodiment of the present invention, a multi-beam charged particle inspection system is provided that includes means for enabling high-throughput, high-reliability, and accurate, high-resolution image acquisition. A wafer stage and means for controlling the position of the wafer stage are provided. The wafer stage is configured to hold a sample, such as a wafer, and is movable in at least one of the x-, y-, or z-directions. The stage typically includes a stage motion controller with multiple independently actuable or controllable motors or actuators. The motors or actuators may include at least one of a piezoelectric motor, a piezoelectric actuator, or an ultrasonic piezoelectric motor. The system also includes a position sensing system configured to determine lateral and vertical displacement or rotation of the stage. The position sensing system may use a laser interferometer, a capacitance sensor, a confocal sensor array, a grating interferometer, or a combination thereof.

[0018] The multi-beam charged particle inspection system includes means for maintaining the lateral position of the focal points of multiple primary charged particle beamlets on the wafer surface and means for maintaining the lateral position of the focal points of multiple secondary electron beamlets, each in a predetermined raster configuration and with a predetermined positional accuracy below a set of thresholds. This, in one example, reduces the time required for precise stage alignment. In another example, the time intervals required for image acquisition and wafer stage movement are overlapped, further improving throughput. The additional means include a first deflection unit for scanning the deflection of the multiple primary charged particle beamlets and at least a second deflection unit for scanning the deflection of the multiple secondary electron beamlets.

[0019] According to one embodiment of the present invention, a multi-beam charged particle inspection system is provided with means for maintaining high resolution and high image contrast during high-throughput and reliable image acquisition of a series of image patches. During first and second image acquisitions, a plurality of sensor data is generated, including sensor data from an image sensor and a stage position sensor. The multi-beam charged particle inspection system includes a control unit configured to generate a set of control signals from the plurality of sensor data. The set of control signals is provided to a control module that controls a set of compensators. According to one embodiment of the present invention, a multi-beam charged particle inspection system is provided with means for compensating for predetermined primary and secondary charged particle beam path drift as well as stage movement.

[0020] According to one example, a multi-beam charged particle beam system includes a controller or control unit configured to at least partially compensate for lateral displacement of a stage by applying a first signal to deflect multiple primary charged particle beamlets incident on a sample and to at least partially compensate for displacement of multiple secondary electron beamlets resulting from positions of the deflected primary charged particle beamlets on the sample by applying a second signal to deflect multiple secondary electron beamlets. The first signal includes an electrical signal that affects the deflection of the multiple primary charged particle beamlets in at least one of an X-axis or a Y-axis. The controller is further configured to dynamically adjust at least one of the first signal or the second signal during scanning of the multiple primary charged particle beamlets over the sample. The controller is connected to a stage motion controller, and multiple motors are each independently controlled to adjust the tilt of the stage so that the stage is substantially perpendicular to the optical axis of the primary charged particle beam. According to one embodiment of the present invention, the multi-beam charged particle microscope system comprises a charged particle source configured to generate a first charged particle beam in use and a multi-beam generator configured to generate a plurality of primary charged particle beamlets from the incident first charged particle beam in use, wherein each individual beamlet of the plurality of primary charged particle beamlets is spatially separated from all other beamlets of the plurality of charged particle beamlets. The multi-beam charged particle microscope system further comprises an object illumination unit including an objective lens configured to focus the primary charged particle beamlets incident on an object plane where the wafer surface is located such that a first image subfield in which a first individual primary beamlet of the plurality of charged particle beamlets impinges on the object plane is spatially separated from a second image subfield in which a second individual primary beamlet of the plurality of primary charged particle beamlets impinges on the object plane. The multi-beam charged particle microscope system further comprises a detection unit including a projection system and an image sensor with a plurality of individual detectors.The projection system is configured to image secondary electrons emitted from the wafer in a first image subfield in the object plane by collision of primary charged particles against a first detector or a first group of detectors among the plurality of individual detectors, and to image secondary electrons emitted from the wafer in a second image subfield in the object plane by collision of primary charged particles against a second detector or a second group of detectors among the plurality of individual detectors.

[0021] In one embodiment, the multi-beam charged particle microscope system includes a subset of high-speed compensators that provide high-speed compensation for dynamic changes in error amplitude. The subset of high-speed compensators includes at least one of an electrostatic lens, an electrostatic deflector, an electrostatic astigmatism corrector, an electrostatic microlens array, an electrostatic astigmatism corrector array, or an electrostatic deflector array. Electrostatic elements such as electrostatic deflectors and / or electrostatic astigmatism correctors have the advantage of being eddy current-free and inductive, providing adjustment times in the range of less than 10 μs for compensation for dynamic changes in error amplitude.

[0022] The subcomponent providing high-speed dynamic compensation can provide an adjustment frequency comparable to the scan frequency at which the primary charged particle beamlets are scanned. That is, the high-speed dynamic compensation can be performed multiple times, i.e., two or more times, during image acquisition of an image patch on the wafer surface by multiple primary charged particle beamlets. Typical line scan frequencies are on the order of 1 kHz to 5 kHz, and the frequency bandwidth of the electrical drive signal for the dynamic compensation element can range from 0.1 kHz to 10 kHz, thereby providing compensation, for example, every 50 scan lines or every 10 scan lines.

[0023] In one embodiment, the multi-beam charged particle microscope system includes a subset of slow-acting compensators that provide compensation for slow changes or drift in error amplitude, the subset of slow-acting compensators including at least one of a magnetic lens, a magnetic deflector, a magnetic astigmatism corrector, or a magnetic beam splitter.

[0024] In one embodiment, a multi-beam charged particle microscope for wafer inspection is provided. The multi-beam charged particle microscope for wafer inspection includes a charged particle multi-beamlet generator that generates a plurality of primary charged particle beamlets and an object illumination unit with a first deflection system that generates a plurality of secondary electron beamlets emitted from the wafer surface by scanning an area of ​​the wafer surface arranged in an object plane with the plurality of primary charged particle beamlets. The multi-beam charged particle microscope for wafer inspection further includes a projection system that images the plurality of secondary electron beamlets and, in use, acquires a digital image of a first image patch of the wafer surface, a second deflection system, and a detection unit with an image sensor. The multi-beam charged particle microscope for wafer inspection further includes a sample stage with a stage position sensor that positions and holds the wafer surface in the object plane during acquisition of the digital image of the first image patch. The first deflection system scans the plurality of primary charged particle beamlets along a predetermined scan path over the wafer surface while the wafer is held on the wafer stage, and the second deflection unit scans the plurality of secondary electron beamlets along the predetermined scan path so that image points of the plurality of secondary electron beamlets are fixed and constant on the image sensor of the detection unit. The multi-beam charged particle microscope for wafer inspection further comprises a control unit and a plurality of detectors including a stage position sensor and an image sensor configured, in use, to generate a plurality of sensor data including position and orientation data of the sample stage. The multi-beam charged particle microscope for wafer inspection further comprises a set of compensators with at least the first and second deflection systems. The control unit generates a set of P control signals C from the plurality of sensor data during acquisition of the digital image of the first image patch. p and configured to generate a set of K amplitudes A of the K error vectors to control the set of compensators. The set of compensators may further include at least one of a compensator of the charged particle multi-beamlet generator and a compensator of the detection unit. In one example, the control unit, when in use, analyzes the plurality of sensor data and generates a set of K amplitudes A of the K error vectors. kIn one embodiment, the control unit further comprises an image data acquisition unit configured, in use, to reduce image sensor data from the image sensor to an image sensor data fragment representing less than 10%, preferably less than 2%, of the image sensor data, and to provide the image sensor data fragment to the sensor data analysis system. In one example, the image sensor data fragment comprises digital image data of a plurality of secondary electron beamlets at a low sampling rate. In one example, the image sensor data fragment comprises digital image data of a reduced set of secondary electron beamlets (9).

[0025] In one example, the sensor data analysis system calculates a set of amplitudes A of the error vectors. k At least one of the amplitudes A n The method is further configured to derive or predict the time course of the

[0026] In one example, the control unit may calculate a set of amplitudes A of the error vector. k A set of control signals C p In one example, the deriving of at least one of the plurality or set of control signals is further based on a predictive model of the actuation output of the stage.

[0027] In one example, the sensor data analysis system is configured to derive a sensor data vector DV of length L (L≧K) from a plurality of sensor data.

[0028] In one example, the control unit generates a first set of control signals C p and providing at least one of the control signals to the first and second deflection units to compensate for changes in the position or orientation of the sample stage, the changes in the position or orientation of the sample stage being given by lateral displacement of the stage and corresponding to a difference between a current position and rotation of the stage in at least one of the X and Y axes and a target position and rotation of the stage.

[0029] The control unit is configured to derive a drive signal for a first compensator in the object illumination unit from the plurality of sensor data to realize an additional displacement of the scanning spot positions of the plurality of primary charged particle beamlets in synchronization with the lateral displacement of the wafer surface. In one example, the additional displacement includes a rotation of the raster configuration of the plurality of primary charged particle beamlets. The control unit is further configured to compensate for the additional displacement of the spot positions on the displaced wafer surface by a second compensator in the projection system, the second compensator in the projection system being configured to operate synchronously with the first compensator in the object illumination unit to keep the spot positions of the plurality of secondary electron beamlets on the image detector constant. In one example, the first compensator in the object illumination unit is a first deflection system, and the control unit is configured to calculate a control signal for the additional displacement or rotation of the scanning spot positions of the plurality of primary charged particle beamlets and provide it to the first deflection system to compensate for the displacement or rotation of the sample stage. In one example, the second compensator in the projection system is a second deflection system, and the control unit is configured to calculate and provide a control signal to the second deflection system to compensate for additional displacements or rotations of the scanning spot positions of the multiple primary charged particle beamlets on the displaced wafer surface, thereby keeping the spot positions of the secondary electron beamlets constant at the image sensor despite modifications of the scan path in response to the displacement or movement of the wafer stage.

[0030] In one embodiment, the charged particle multi-beamlet generator of the multi-beam charged particle microscope further comprises a fast compensator, and the control unit generates a first set of control signals C pand providing at least one of the control signals to the fast compensator to induce a rotation of the plurality of primary charged particle beamlets, thereby compensating for the rotation of the sample stage. In one embodiment, the control unit of the multi-beam charged particle microscope is further configured to generate a third control signal to move the wafer surface by the wafer stage to a second center position of the second image patch in the object plane for image acquisition of a digital image of the second image patch. In one embodiment, the control unit calculates a second set of P control signals C from the plurality of sensor data during a time interval Tr of the movement of the wafer stage to the second center position of the second image patch. p In one embodiment, the control unit is further configured to calculate a start time of image acquisition of the second image patch during a time interval Tr, and to start image acquisition of the second image patch during a deceleration time interval Td of the wafer stage, and is further configured to provide at least an offset signal of a predicted offset position of the wafer stage during the time interval Td to the first and second deflection systems.

[0031] In one embodiment, a method for wafer inspection with a multi-beam charged particle microscope is provided, the multi-beam charged particle microscope comprising a plurality of detectors including an image sensor and a stage position sensor, and a set of compensators including at least first and second deflection systems. The method includes: a. positioning and aligning a wafer surface of a wafer relative to a location in a local wafer coordinate system that includes a line of sight of a multi-beam charged particle microscope; b. performing image acquisition to acquire a digital image of a first image patch of the wafer surface; c. collecting multiple sensor data from multiple detectors in the image acquisition step; d. From multiple sensor data, a set of K error amplitudes A k and deriving e. A set of error amplitudes A k From the set of P control signals Cp and deriving f. In step b of image acquisition, a set of control signals C p to a set of compensators; Includes.

[0032] In one embodiment, the method of wafer inspection further includes the step (g) of deriving a sensor data vector DV of length L (L≧K) from the plurality of sensor data. In one embodiment, the method of wafer inspection further includes deriving a set of amplitudes A of the error vector. k At least one of the amplitudes A n In one embodiment, the method for wafer inspection further includes a step (h) of deriving a time course of the control signal C p to the first and second deflection units. In one embodiment, the method of wafer inspection further comprises the step of (i) compensating for changes in the position or orientation of the sample stage by providing a set of error amplitudes A k to a second set of control signals C p and providing a second set of control signals in step a) of positioning and aligning the wafer surface of the wafer.

[0033] In one embodiment of the present invention, a high-throughput, high-resolution charged particle microscope and a method for operating a charged particle microscope are provided that meet imaging specification requirements for a wafer inspection task. The method includes a first image acquisition of a first image patch during a first time interval Ts1 and a second image acquisition of a second image patch during a second time interval Ts2, and further includes a third time interval Tr for moving a sample stage from a first center position of the first image patch to a second center position of the second image patch, wherein the series of image patches are imaged in a series of image acquisition steps, with at least one of the first and second time intervals Ts1 and Ts2 overlapping with the third time interval Tr. The total time interval from the start of the first time interval Ts1 to the end of the second time interval Ts2 is shorter than the sum of the three time intervals Ts1, Tr, and Ts2, thereby achieving high-throughput, high-speed wafer inspection. In one example, the second image acquisition of the second image patch begins before the end of the third time interval Tr when the sample stage has completely stopped. In one example, the third time interval Tr of sample movement begins before the end of time interval Ts1 when image acquisition of the first image patch ends. In one example of this method, the calculation of the start time of the third time interval Tr of wafer movement is performed during the first time interval Ts1 of image acquisition of the first image patch so that a displacement of the first center position of the first image patch from the line of sight of the multi-beam charged particle microscope or a sample stage movement speed falls below a predetermined threshold. In one example of this method, the calculation of the start time of the second time interval Ts2 of second image acquisition is performed during the time interval Tr of sample stage movement so that a displacement of the second center position of the second image patch from the line of sight of the multi-beam charged particle microscope or a sample stage movement speed falls below a predetermined threshold.

[0034] In one example method of operating a multi-beam charged particle microscope, the method includes: predicting a series of sample stage positions over a time interval Tr of wafer stage movement; calculating at least first and second control signals from the predicted sample stage position; providing a first control signal to a first deflection system in a primary beam path of the multi-beam charged particle microscope and a second control signal to a second deflection system in a secondary beam path of the multi-beam charged particle microscope; Further includes:

[0035] In one example, the charged particle microscope comprises a control unit configured to calculate a start time of a sample stage movement from the first image patch to a second image patch during a first image acquisition of the first image patch. In one example of the present invention, the charged particle microscope comprises a control unit configured to calculate a start time of a second image acquisition of the second image patch during a sample stage movement from the first image patch to the second image patch.

[0036] In one embodiment, a method of operating a multi-beam charged particle microscope configured for wafer inspection includes the steps of: a preliminary step of defining a set of predetermined normalized error vectors representing a set of image qualities and deviations from the set of image qualities; a preliminary step of determining a set of thresholds for the amplitudes of a set of normalized error vectors; a preliminary step of selecting a set of compensators for the multi-beam charged particle microscope; a preliminary step of determining a sensitivity matrix according to a linear and / or nonlinear perturbation model by variations in at least the drive signal of each of the set of compensators; a preliminary step of deriving a set of normalized drive signals that compensate for each of the set of normalized error vectors; a preliminary step of storing the normalized drive signal and the set of thresholds in a memory of a control unit of the multi-beam charged particle microscope; A method is described that includes:

[0037] In one example, the set of compensators comprises a first deflection unit of a multi-beam charged particle microscope for scanning and deflecting a plurality of primary charged particles, and a second deflection unit for scanning and deflecting a plurality of secondary electrons generated when using the multi-beam charged particle microscope.

[0038] The sensitivity matrix is ​​analyzed, for example, by singular value decomposition or a similar algorithm. In one example, the sensitivity matrix is ​​decomposed by dividing it into two, three, or more kernels or independent subsets of image qualities, thereby reducing computational complexity and suppressing nonlinear or higher-order effects.

[0039] In use (e.g., during wafer inspection), the method of operation includes using a normalized error vector, a normalized drive signal, and a set of thresholds stored in a memory of a control unit of the multi-beam charged particle microscope. In use, the method comprises receiving a plurality of sensor data from a plurality of sensors of a multi-beam charged particle microscope, the sensor data vector comprising: expanding the sensor data vector into a set of normalized error vectors stored in a memory of the control unit, and determining from the sensor data vector the actual amplitudes of the set of normalized error vectors; comparing the set of actual amplitudes to a set of thresholds stored in a memory of the control unit; deriving a set of control signals from the set of actual amplitudes based on the results of the comparison; deriving a set of actual drive signals from a set of normalized drive signals stored in a memory of the control unit using a set of control signals; providing a set of actual drive signals to a set of compensators of the multi-beam charged particle microscope such that, during operation of the multi-beam charged particle microscope, the set of actual amplitudes of the set of normalized error vectors is less than a set of thresholds; Includes.

[0040] In one example, the plurality of sensor data includes at least one of position and velocity information relating to the actual position and actual velocity of a wafer stage that holds the wafer during inspection by the multi-beam charged particle microscope. A set of actual amplitudes of a normalized error vector from the sensor data vector represents a set of actual image quality conditions of the multi-beam charged particle microscope. A set of control signals is derived by comparison with predetermined stored thresholds. A set of actual drive signals is calculated from the control signals, for example, by multiplying the control signals by a predetermined set of normalized drive signals. During image scanning or image acquisition of at least one image patch, the set of actual drive signals is provided to a set of compensators so that, during operation of the multi-beam charged particle microscope, a subset of the actual amplitudes is below a subset of the predetermined thresholds stored in the memory of the control unit. The steps of this method are repeated at least twice, at least ten times, preferably for each scan line, during acquisition of each image patch.

[0041] In one example, the method further includes predicting a subset of transition amplitudes of at least a subset of the set of actual amplitudes according to a predicted transition of the multi-beam charged particle microscope over a prediction time interval during wafer inspection. The method may further include recording at least a subset of the set of actual amplitudes of the multi-beam charged particle microscope during use to generate a history of the subset of the set of actual amplitudes. The method of operating the multi-beam charged particle microscope may further include deriving a set of predicted control signals from the set of transition amplitudes and deriving a set of predicted drive signals from the set of predicted control signals during wafer inspection, and providing the set of predicted drive signals to a set of compensators in a time sequence during wafer inspection to cause the subset of actual amplitudes to be below a set of thresholds during operation of the multi-beam charged particle microscope over the prediction time interval.

[0042] The predicted progression of the multi-beam charged particle microscope over the prediction time interval is determined according to one of a predictive model function or a linear, quadratic, or higher-order extrapolation of the set of historical actual amplitudes. In one example, the method further includes, during use, recording at least a subset of the set of actual amplitudes of the multi-beam charged particle microscope to generate a historical set of subsets of actual amplitudes. The method further includes, during wafer inspection, deriving a set of predicted control signals from the set of progression amplitudes and deriving a set of predicted drive signals from the set of predicted control signals, and, during wafer inspection, supplying the set of predicted drive signals to a set of compensators in a time sequence to cause the subset of actual amplitudes to be less than a subset of threshold values ​​during operation of the multi-beam charged particle microscope over the prediction time interval. This embodiment includes a multi-beam charged particle microscope configured, during use, to apply the above-described method steps.

[0043] In an embodiment, the error amplitude derived from the sensor data represents image performance specifications for the wafer inspection task, such as at least one of the relative position and orientation of the wafer stage to the line of sight of the multi-beam charged particle microscope or the image coordinate system of the multi-beam charged particle microscope, the magnification or pitch of the multi-beam charged particle microscope, the decentering state, contrast state, absolute position accuracy of the multiple charged particle beamlets, and higher order aberrations such as distortion aberrations, astigmatism, and chromatic aberrations of the multiple charged particle beamlets.

[0044] In one embodiment, a multi-beam charged particle microscope and software code is disclosed, the multi-beam charged particle microscope comprising a set of compensators including deflectors, a control unit, and installed software code, configured to apply any of the methods according to any of the method steps described above.

[0045] In one embodiment, one or more processors of a multi-beam charged particle device including a charged particle source generating a plurality of primary charged particle beamlets, determining a lateral displacement of a movable stage in at least one of the X and Y axes; instructing the controller to at least partially compensate for the lateral displacement by application of a first signal that deflects a plurality of primary charged particle beamlets incident on the sample; Disclosed is a non-transitory computer-readable medium comprising a set of instructions executable to cause an apparatus to perform a method including: instructing a controller to at least partially compensate for lateral displacement of the sample stage by application of a second signal that deflects a plurality of secondary electron beamlets emitted from the sample.

[0046] Further details are disclosed below with reference to the accompanying drawings. [Brief explanation of the drawings]

[0047] [Figure 1] FIG. 1 illustrates a multi-beam charged particle microscope system according to one embodiment. [Figure 2] FIG. 2 shows a first test site and a second test site including first and second image patches. [Figure 3a] FIG. 1 is a schematic diagram of an image coordinate system displaced and rotated relative to the local wafer coordinate system. [Figure 3b] FIG. 10 is a schematic diagram of an image patch rotated relative to the local wafer coordinate system. [Figure 4] 1A and 1B show the slowly varying drift component of the error amplitude (a) before and (b) after compensation according to the present invention. [Figure 5] 1A and 1B show fast-varying components or dynamic changes in error amplitude (a) before and (b) after compensation according to the present invention. [Figure 6] FIG. 8 is a block diagram of a multi-beam charged particle microscope system showing control unit 800 in greater detail, according to one embodiment of the present invention. [Figure 7] FIG. 1 is a block diagram of a method of operation of a multi-beam charged particle microscope system for wafer inspection according to an embodiment of the present invention. [Figure 8] FIG. 1 illustrates an active multi-aperture plate. DETAILED DESCRIPTION OF THE INVENTION

[0048] In the exemplary embodiments described below, functionally and structurally similar components are denoted by similar or identical reference numerals whenever possible.

[0049] The schematic representation of FIG. 1 illustrates basic features and functionality of a multi-beam charged particle microscope system 1 according to an embodiment of the present invention. Note that the symbols used in the figure are chosen to symbolically represent the respective functions of the components shown, rather than to represent the physical configuration of the components. The system as shown is that of a scanning electron microscope (SEM) that uses multiple primary electron beamlets 3 to generate multiple primary charged particle beam spots 5 on the surface of an object 7, such as a wafer, positioned in the object plane 101 of an objective lens 102. For simplicity, only five primary charged particle beamlets 3 and five primary charged particle beam spots 5 are shown. The features and functionality of the multi-beamlet charged particle microscope system 1 can be implemented using other types of primary charged particles, such as electrons or ions (particularly helium ions).

[0050] The microscope system 1 includes an object illumination unit 100, a detection unit 200, and a beam splitter unit 400 that separates a secondary charged particle beam path 11 from a primary charged particle beam path 13. The object illumination unit 100 includes a charged particle multi-beamlet generator 300 that generates a plurality of primary charged particle beamlets 3, and is configured to focus the primary charged particle beamlets 3 onto an object plane 101 where a surface 25 of a wafer 7 is located by a sample stage 500. The sample stage 500 includes a stage motion controller that includes a plurality of motors that are configured to be independently controlled by control signals. The stage motion controller is connected to a control unit 800.

[0051] The primary beamlet generator 300 generates a plurality of primary charged particle beamlet spots 311 at an intermediate image plane 321, which is typically a spherically curved surface to compensate for the field curvature of the object-illumination unit 100. The primary beamlet generator 300 comprises a primary charged particle (e.g., electron) source 301. The primary charged particle source 301 emits a diverging primary charged particle beam 309, which is collimated, for example, by collimator lenses 303.1 and 303.2 to form a parallel beam. The collimator lenses 303.1 and 303.2 typically consist of one or more electrostatic or magnetic lenses, or a combination of electrostatic and magnetic lenses. The collimated primary charged particle beam enters a primary multi-beamlet configuration unit 305. The multi-beamlet configuration unit 305 essentially comprises a first multi-aperture plate 306.1, which is illuminated by the primary charged particle beam 309. The first multi-aperture plate 306.1 comprises a plurality of apertures in a raster arrangement for generating a plurality of primary charged particle beamlets 3, which are generated by transmission of the collimated primary charged particle beam 309. The multi-beamlet forming unit 305 comprises at least another multi-aperture plate 306.2 arranged downstream of the first multi-aperture plate 306.1 with respect to the electron movement direction of the beam 309. For example, the second multi-aperture plate 306.2 preferably has the function of a microlens array and is set to a defined potential so that the focal positions of the plurality of primary beamlets 3 at the intermediate image plane 321 are adjusted. The third active multi-aperture plate arrangement 306.3 (not shown) comprises an individual electrostatic element for each of the plurality of apertures, thereby individually influencing each of the plurality of beamlets. The active multi-aperture plate arrangement 306.3 consists of one or more multi-aperture plates with electrostatic elements such as circular electrodes, multi-pole electrodes or a series of multi-pole electrodes for the micro-lenses that make up a deflector array, a micro-lens array or an astigmatism corrector array.The multi-beamlet forming unit 305 includes adjacent first electrostatic field lenses 307, which, together with a second field lens 308 and a second multi-aperture plate 306.2, focus multiple primary charged particle beamlets 3 onto or near an intermediate image plane 321.

[0052] Located at or near the intermediate image plane 321 is a beam steering multi-aperture plate 390 including a plurality of apertures with electrostatic elements (e.g., deflectors) that individually steer each of the plurality of charged particle beamlets 3. The apertures of the beam steering multi-aperture plate 390 have larger diameters that allow the passage of the plurality of primary charged particle beamlets 3 even if their focal spots are displaced from their respective design positions.

[0053] The multiple focal points of the primary charged particle beamlets 3 passing through the intermediate image plane 321 are imaged by the field lens groups 103.1 and 103.2 and the objective lens 102 at an image plane 101, where the investigation surface of the wafer 7 is positioned by an object mount on the sample stage 500. The object illumination system 100 further comprises a deflection system 110, close to the first beam crossover 108, capable of deflecting the multiple charged particle beamlets 3 in a direction perpendicular to the beam propagation direction (here, the z-direction). The deflection system 110 is connected to the control unit 800. The objective lens 102 and the deflection system 110 are centered on an optical axis 105 of the multi-beamlet charged particle microscope system 1, which is perpendicular to the wafer surface 25. The wafer surface 25 arranged in the image plane 101 is then raster-scanned by the deflection system 110. This causes the multiple primary charged particle beamlets 3 constituting the multiple beam spots 5 arranged in a raster configuration to be synchronously scanned over the wafer surface 101. In one example, the raster configuration of the focal spots 5 of the multiple primary charged particles 3 is a hexagonal raster of approximately 100 or more primary charged particle beamlets 3. The primary beam spots 5 are spaced apart from one another by approximately 6 μm to 15 μm and have diameters less than 5 nm (e.g., 3 nm, 2 nm, or less). In one example, the beam spot size is approximately 1.5 nm, and the distance between two adjacent beam spots is 8 μm. At each scanning position of each of the multiple primary beam spots 5, multiple secondary electrons are generated, forming multiple secondary electron beamlets 9 in the same raster configuration as the primary beam spots 5. The intensity of the multiple secondary charged particles generated at each beam spot 5 depends on the intensity of the impinging primary charged particle beamlet illuminating the corresponding spot and the material composition and topography of the object under the beam spot. The secondary charged particle beamlets 9 are accelerated by an electrostatic field generated by the sample charging unit 503, collected by the objective lens 102, and guided to the detection unit 200 by the beam splitter 400. The detection unit 200 images the secondary electron beamlets 9 onto an image sensor 207 to form a plurality of secondary charged particle image spots 15. The detector includes a plurality of detector pixels or individual detectors.The intensity of each of the multiple secondary charged particle beam spots 15 is detected separately, allowing for high-resolution detection of the material composition of the wafer surface, resulting in large image patches with high throughput. For example, a raster of 10 × 10 beamlets with an 8 μm pitch generates an image patch of approximately 88 μm × 88 μm in one image scan by the deflection system 110, resulting in an image resolution of, for example, 2 nm. The image patch is sampled at half the 2 nm beam spot size, for example, resulting in 8,000 pixels per image line for each beamlet, and an image patch generated by 100 beamlets contains 6.4 billion pixels. Image data is collected by the control unit 800. Details of image data collection and processing, for example, using parallel processing, are described in German Patent Application No. 102019000470.1 (incorporated herein) and the aforementioned U.S. Patent No. 9,536,702.

[0054] The multiple secondary electron beamlets 9 pass through the first deflection system 110, are scan-deflected by the first deflection system 110, and are guided by the beam splitter unit 400 to follow the secondary beam path 11 of the detection unit 200. The multiple secondary electron beamlets 9 travel in the opposite direction to the primary charged particle beamlets 3, and the beam splitter unit 400 is configured to separate the secondary beam path 11 from the primary beam path 13, typically by a magnetic field or a combination of a magnetic field and an electrostatic field. Optionally, an additional magnetic correction element 420 is present in the secondary beam path as well as in the primary beam path. The projection system 205 further comprises at least a second deflection system 222 connected to a projection system control unit 820. The control unit 800 is configured to compensate for residual errors in the positions of the multiple focal points 15 of the multiple secondary electron beamlets 9, so that the positions of the multiple secondary electron focal spots 15 remain constant on the image sensor 207.

[0055] The projection system 205 of the detection unit 200 includes at least a second crossover 212 of the multiple secondary electron beamlets 9, where an aperture 214 is arranged. In one example, the aperture 214 further comprises a detector (not shown) connected to a projection system control unit 820. The projection system control unit 820 is connected to at least one electrostatic lens 206 of the projection system 205, which further comprises electrostatic or magnetic lenses 208, 209, 210, and is further connected to a third deflection unit 218. The projection system 205 further comprises at least a first multi-aperture corrector 220 with an aperture and an electrode that individually influences each of the multiple secondary electron beamlets 9, and optionally further active elements 216, each connected to the control unit 800.

[0056] The image sensor 207 is configured with an array of detection areas whose pattern matches the raster arrangement of the secondary electron beamlets 9 focused onto the image sensor 207 by the projection lens 205. This allows individual secondary electron beamlets 9 to be detected independently of the other secondary electron beamlets 9 incident on the image sensor 207. A plurality of electrical signals are generated, converted into digital image data, and processed in the control unit 800. During an image scan, the control unit 800 is configured to trigger the image sensor 207 to detect a plurality of time-resolved intensity signals from the plurality of secondary electron beamlets 9 over a predetermined time interval, and digital images of image patches from all scanning positions of the plurality of primary charged particle beamlets 3 are accumulated and stitched together.

[0057] The image sensor 207 shown in FIG. 1 can be an electron-sensitive detector array, such as a CMOS or CCD sensor. Such an electron-sensitive detector array can include an electron-to-photon conversion unit, such as a scintillator element or an array of scintillator elements. In another embodiment, the image sensor 207 can be configured as an electron-to-photon conversion unit or scintillator plate positioned at the focal plane of the multiple secondary electron particle image spots 15. In this embodiment, the image sensor 207 can further include relay optics that image and guide photons generated by the electron-to-photon conversion unit at the secondary charged particle image spots 15 onto dedicated photon detection elements, such as multiple photomultiplier tubes or avalanche photodiodes (not shown). Such an image sensor is disclosed in U.S. Pat. No. 9,536,702, which is incorporated herein by reference. In one example, the relay optics further includes a beam splitter that splits and guides the light to a first slow photodetector and a second fast photodetector. The second high-speed photodetector is formed by a photodiode array, e.g., avalanche photodiodes, and is fast enough to resolve the image signals of the multiple secondary electron beamlets according to the scan rate of the multiple primary charged particle beamlets. The first low-speed photodetector is preferably a CMOS or CCD sensor that monitors the focal spots 15 of the multiple secondary electron beamlets 9 and provides high-resolution sensor data signals for controlling the operation of the multi-beam charged particle microscope as described in more detail below.

[0058] In the illustrated example, the primary charged particle source is implemented in the form of an electron source 301 featuring an emitter tip and an extraction electrode. If primary charged particles other than electrons are used, such as helium ions, the configuration of the primary charged particle source 301 can differ from that shown. The primary charged particle source 301, the active multi-aperture plate configurations 306.1...306.3, and the beam steering multi-aperture plate 390 are controlled by a primary beamlet control module 830 connected to the control unit 800.

[0059] Preferably, the stage 500 does not move during the acquisition of image patches by scanning the multiple primary charged particle beamlets 3, but moves after the acquisition of an image patch to the next image patch to be acquired. Stage movement and stage position are monitored and controlled by sensors known in the art, such as laser interferometers, grating interferometers, confocal microlens arrays, etc. For example, a position sensing system may use a laser interferometer, a capacitance sensor, a confocal sensor array, a grating interferometer, or a combination thereof to determine the lateral and vertical displacement and rotation of the stage. In one embodiment of the present invention, as described below, the movement of the stage 500 from one image patch to the next is overlapped with the acquisition of the image patches, thereby improving throughput.

[0060] One embodiment of a method for wafer inspection by image patch acquisition is described in more detail in FIG. 2. The wafer is positioned with its wafer surface 25 in the focal plane of the multiple primary charged particle beamlets 3 at the center 21.1 of the first image patch 17.1. Predetermined positions of the image patches 17.1···k correspond to wafer inspection sites for semiconductor feature inspection. Predetermined positions of the first inspection site 33 and the second inspection site 35 are loaded from an inspection file in a standard file format. The predetermined first inspection site 33 is divided into multiple image patches (e.g., first image patch 17.1 and second image patch 17.2), and the first center position 21.1 of the first image patch 17.1 is aligned under the optical axis of the multi-beam charged particle microscope system for the first image acquisition step of the inspection task. The first center 21.1 of the first image patch is selected as the origin of a first local wafer coordinate system for the acquisition of the first image patch 17.1. Methods for aligning a wafer so that the wafer surface 25 is positioned and a coordinate system of wafer coordinates is created are well known in the art.

[0061] In each image patch, the primary beamlets are distributed in a regular raster configuration 41, and a digital image of the image patch is generated by scanning with a scanning mechanism. In this example, the primary charged particle beamlets 3 are arranged in a rectangular raster configuration 41, which has n primary beam spots 5.11, 5.12 to 5.1N in a first line having N beam spots, and M lines having beam spots 5.11 to 5.MN. For simplicity, M = 5 × N = 5 beam spots are shown, but the number of beam spots M × N can be larger, and the beam spots 5.11 to 5.MN can have various raster configurations 41, such as a hexagonal or circular shape.

[0062] Each primary charged particle beamlet is scanned over the wafer surface 25. As shown, in the example of a primary charged particle beamlet including beam spots 5.11-5.MN, there are multiple scan paths 27.11-27.MN of the primary charged particle beamlets. The scanning of each of the multiple primary charged particle beamlets is performed by, for example, moving back and forth along scan paths 27.11-27.MN, while each focal point 5.11-5.MN of each primary charged particle beamlet is moved by scanning deflector 110 in the x-direction from the start of the image line, which in this example is the leftmost image point of image subfield 31.MN. After each focal point is scanned by scanning the primary charged particle beamlet to the right position, the scanning deflector 110 moves the multiple charged particle beamlets parallel to the line start position of the next line in each subfield 31.11-31.MN. The movement back to the line start position of the subsequent scan line is called a flyback. The multiple primary charged particle beamlets follow the scan paths 27.11-27.MN in parallel, thereby obtaining multiple scan images of each subfield 31.11-31.MN in parallel. Regarding image acquisition, as described above, multiple secondary electrons are emitted at the focal points 5.11-5.MN, generating multiple secondary electron beamlets 9. The multiple secondary electron beamlets 9 are collected by the objective lens 102, pass through the first deflection system 110, and are guided to the detection unit 200 and detected by the image sensor 207. A continuous stream of data from each of the multiple secondary electron beamlets 9 is converted into multiple two-dimensional data sets in synchronization with the scan paths 27.11-27.MN, constituting digital image data for each subfield. According to a preselected scan program, the multiple primary charged particle beamlets follow predetermined scan paths 27.11-27.MN. The digital images of the subfields are finally stitched together by an image stitching unit to form the digital image of the first image patch 17.1. Each image subfield has a small overlap area with the adjacent image subfield, as shown by the overlap area 39 of subfield 31.mn and subfield 31.m(n+1).The pitch between the multiple primary charged particle beam spots 5.11-5.MN in the prior art typically varies due to drift, lens distortion, and other aberrations. Therefore, the overlap area 39 in the prior art is typically set large enough to cover the entire image patch in a single image scan, despite variations in beam spot position.

[0063] In one embodiment of the method for wafer inspection, the throughput of a multi-beam charged particle microscope system for wafer inspection is improved by reducing the size of the overlap area 39. This increases the throughput by increasing the size of each image patch. In one example, the beam pitch of the focal points 5 of the multiple primary charged particle beamlets is 10 μm. For example, a 25% reduction in the width of each overlap area 39 by 200 nm increases the size of the image patch by approximately 1%, improving the throughput by approximately 1%. Furthermore, a 65% reduction in the width of the overlap area improves the throughput by 2.5%. The reduction in the overlap area 39 is achieved by controlling the pitch of the multiple primary charged particle beamlets 3. A compensator (multi-beam multipole deflection device), such as the active multi-aperture plate 306.3 in FIG. 1, precisely controls the position of the beam spot 5 formed by the multiple primary charged particle beamlets 3. For control calculations, a detector (e.g., the image sensor 207 of the detection unit 200 detecting the multiple secondary electron beamlets 9) is configured to provide a sensor signal representing the positions of the multiple beam spots 5. Then, the beam position deviations of the multiple primary beam spots 5 are corrected, thereby reducing the overlap area. Accurate control of each primary charged particle beam spot 5 of the multiple primary charged particle beamlets 3 at corresponding raster positions with an accuracy of less than 70 nm realizes a 2% improvement in throughput. Even more accurate control of the primary charged particle beam spot positions to less than 30 nm can improve throughput by more than 3.5%. In a next step, after acquiring the digital image of the first image patch, the wafer stage moves the wafer under sensor control to an adjacent predetermined center position 21.2, and a new local wafer coordinate system centered on the predetermined center position 21.2 is defined. Then, a second image patch 17.2 is determined, such that two adjacent image patches 17.1 and 17.2 are obtained with an overlap area 19. Again, similar to the reduction in overlap area 39 discussed above, the reduction in size of overlap area 19 improves throughput. The two image patches 17.1 and 17.2 are then stitched together to form an image of a given wafer area.After acquiring a digital image of the first inspection site 33, the wafer stage moves the wafer to a predetermined center position 21.k for inspecting a process control monitor (PCM) in a predetermined wafer area, for example, by acquiring an image of the second inspection site 35. A scanning operation (not shown) is then performed to obtain image patch 17.k. As shown in this simplified example, multiple inspection sites on a wafer are inspected sequentially in this manner.

[0064] Next, we will explain the requirements or specifications for the wafer inspection task. High-throughput wafer inspection requires high speed image acquisition of 17.1···k image patches as well as fast stage movement between the 17.1···k image patches. Meanwhile, strict image quality specifications, such as image resolution, image accuracy, and repeatability, must be maintained. For example, the image resolution requirement is typically 2 nm or less with high repeatability. Image accuracy, also known as image fidelity, determines the edge position of a feature, or more generally, the absolute position accuracy of a feature, with high absolute accuracy. For example, the absolute lateral position accuracy of each of the multiple primary charged particle beamlets must be less than 10 nm, and the absolute lateral position of each of the multiple primary charged particle beamlets must be known with an accuracy of less than 1 nm. Typically, the position accuracy requirement is approximately 50% or less of the resolution requirement. Next, high image uniformity must be achieved. The image uniformity error is defined as dU = (Imax - Imin) / (Imax + Imin), where Imax and Imin are the maximum and minimum image intensities of the uniform object under image acquisition. Typically, an image uniformity error dU of less than 5% is required. The image contrast and dynamic range must be sufficient to accurately represent the semiconductor features and material composition of the semiconductor wafer under inspection. Typically, the dynamic range should be greater than 6 or 8 bits and the image contrast should be greater than 80%.

[0065] High image repeatability means that repeated image acquisition of the same area produces first and second repeated digital images, with the difference between the first and second repeated digital images being below a predetermined threshold. For example, the difference in image distortion between the first and second repeated digital images should be less than 1 nm, preferably less than 0.5 nm, and the difference in image contrast should be less than 10%. In this way, repeated imaging operations produce similar image results. This is important, for example, for acquiring and comparing images of similar semiconductor structures on different wafer dies, or for comparing acquired images to representative or reference images obtained from CAD-based image simulations or databases.

[0066] One of the requirements or specifications for a wafer inspection task is throughput. It depends on several parameters (e.g., the speed of the sample stage, the time required to accelerate and decelerate the stage, and the number of iterations required to align the stage at each new measurement site), as well as the measurement area per acquisition time itself. An example of improving throughput by increasing the image patch size by reducing the overlap area has been described above. The measurement area per acquisition time depends on the dwell time, resolution, and number of beamlets. Typical examples of dwell times are 20 ns to 80 ns. Therefore, the pixel rate of the high-speed image sensor 207 is in the range of 12 MHz to 50 MHz, and it is possible to acquire approximately 20 image patches or frames per minute. However, between the acquisition of two image patches, the wafer is moved laterally by the wafer stage to the next point of interest. In one example, the time interval Tr for the wafer to move from the first image patch to the second image patch is approximately 1 second, reducing the frame rate to approximately 15 frames per minute. The typical time interval Tr for moving a wafer from a first image patch to a second image patch with a standard stage exceeds 1 s, including the time interval for precise adjustment of the second image patch, and can be on the order of 3 s or more (e.g., 5 s). For 100 beamlets, a typical example throughput in high-resolution mode with a pixel size of 0.5 nm is approximately 0.045 mm² / min (millimeters per minute), while for increased beamlets and lower resolution (e.g., 10,000 beamlets and a dwell time of 25 ns), throughputs of over 7 mm² / min are possible. Stage movement, including stage acceleration and deceleration, is one of the factors limiting the throughput of multi-beam inspection systems. Faster stage acceleration and deceleration in a short period of time requires complex and expensive stages or induces dynamic vibrations in multi-beam charged particle systems. According to embodiments of the present invention, high throughput of wafer inspection tasks can be achieved while maintaining image performance specifications well within the requirements, e.g., as described above.

[0067] Typically, uncontrolled high-speed, high-throughput image acquisition suffers from dynamic effects such as drift and residual and unwanted stage movement. Deviations from ideal image acquisition conditions are generally described by an error function. An example of an error function for the rotation and displacement of multiple image spots 5 relative to the wafer 7 is shown in FIG. 3a for a circular arrangement of multiple beam spots 5. An image coordinate system 51, including image coordinates xi and yi, is defined by a virtual coordinate system at the center of an image patch obtained by scanning a set of primary charged particle beamlets with the beam spots 5 (three shown). The center line of the set of primary charged particle beamlets at a predetermined central scanning position is called the line of sight 53, and the line of sight 53 and the z-axis of the image coordinate system are identical. In an ideal situation, after proper calibration of the multi-beamlet charged particle microscope system 1, the line of sight 53 and the optical axis 105 of the multi-beamlet charged particle microscope system 1 are identical. In a real imaging situation, the line of sight 53 deviates from the optical axis 105 of the multi-beamlet charged particle microscope system 1. This deviation is due to, for example, drift of the object illumination unit 100, aberrations in the first scanning deflector 110 or other electrostatic and magnetic elements in the primary charged particle beam path 13, such as the multi-beamlet generator active element 330 or the beam splitter 400. In a real imaging situation, the deviation of the line of sight 53 varies with time, such as the image acquisition time of one image scan of each image patch 17.1···k.

[0068] A local wafer coordinate system 551 is defined at the wafer inspection site by local wafer coordinates xl and yl. In a real imaging situation, the local wafer coordinate system 551 is displaced from the image coordinate system 51 by the line of sight 53. The displacement vector 55 is due to, for example, wafer stage misalignment, wafer stage drift, or image coordinate system 51 drift, or both. In a real imaging situation, the deviation of the local wafer coordinate system 551 varies with time, such as the image acquisition time of a single image scan. The displacement vector 55 is generally described as a time-dependent vector D(t) = [Dx, Dy, Dz](t). In a real imaging situation, the displacement vector 55 includes the difference between the line of sight 53 deviation and the wafer stage 500 drift, both of which vary independently with time, such as the image acquisition time of a single image scan of each image patch 17.1···k.

[0069] The image coordinate system 51 can be rotated about the z-axis or line of sight 53 relative to the local wafer coordinate system 551 by a rotation angle Rz indicated by arrow 57, and the image of the image patch 17 from the wafer surface 25 is obtained in a rotated image coordinate system 59 containing coordinates (xi', yi'). The rotation angle can occur about either axis and, depending on the time, constitutes a rotation angle vector R(t) = [Rx, Ry, Rz](t). Rotation about the z-axis rotates all image spots 5 to image spots 5' (as shown), as indicated by the displacement vector 61 between the unrotated image spots 5 and the rotated image spots 5'. The image rotational shift can result from rotation of the image spots 5 about the vertical or z-axis, stage rotation, or both.

[0070] FIG. 3b illustrates the image rotation situation for the example image patch 17.1 of FIG. 2. The same reference numerals as in FIG. 2 are used, but the image coordinate system 51 is rotated relative to the wafer coordinate system 551. The multiple focal spots 5 arranged in a raster configuration are rotated, the image patch 31 is rotated, and the scan path 27 is rotated. In one embodiment of the present invention, described in more detail below, the image rotation is compensated for by rotating the raster configuration of the multiple focal spots 5. This differs from single-beam charged particle microscopes, which effectively achieve rotation of a single scan path by dynamic scan rotation, i.e., compensating for image rotation by changing a single scan path. As an example of the effect of scan rotation on multiple primary charged particle beams in the example of two primary charged particle beamlets, the effect of scan rotation is shown in subfields 37.1 and 37.2. While the scan path 27 can be rotated by the scan beam deflector of a multi-beam charged particle microscope, the raster configuration of the multiple spots 5 cannot be rotated by the scan deflector. To compensate for rotation, including dynamic changes in the rotation of the raster arrangement of the plurality of spots 5, additional means are required, as provided in some embodiments of the present invention.

[0071] The multi-beamlet charged particle microscope system according to an embodiment of the present invention includes multiple sensors that provide sensor signals during image acquisition. The sensors may be, for example, a stage position sensor of the stage 500, a sensor disposed at an aperture such as the aperture 214, or the image sensor 207. The control unit 800 is configured to extract error functions, such as an image displacement vector D(t) or an image rotation R(t), from the sensor signals, which include changes in the focal position or the image plane tilt. In general, the control unit 800 is configured to analyze the sensor signals and decompose them into a set of individual model error functions by methods known in the art, such as fitting a set of predetermined model error functions with error amplitudes to the sensor data. Such fitting operations can be, for example, a least-squares fitting operation or singular value decomposition, and multiple error amplitudes are calculated for each of the set of model error functions. The calculation of the error amplitudes significantly reduces the amount of data for controlling the multiple primary and secondary charged particle beamlets 3 and 9 and the stage 500, for example to six error amplitudes. However, in one embodiment of the present invention, many more error amplitudes are considered as well, such as magnification error, various higher-order distortions, and individual field-dependent image aberration patterns. The normalized error amplitudes may represent, for example, line-of-sight displacement in both lateral directions, lateral and axial wafer stage displacement, wafer stage rotation, line-of-sight rotation, magnification error, focus error, astigmatism error, or distortion error. The decomposition of the sensor signal into a limited set of error amplitudes significantly improves the speed of computation and control of the correction signal.

[0072] In one embodiment, the control unit 800 is configured to analyze the evolution of the error amplitude over time. A history of the change in the error amplitude over time is recorded, and the control unit is configured to expand the change in the error amplitude into a time-dependent model function. The control unit 800 is configured to predict the change in at least a subset of the error amplitude over a short period of time, e.g., a fraction of an image scan over a scan time interval Ts. The scan time interval Ts of an image patch ranges from 1 to 5 seconds depending on the dwell time. In a typical example, the scan time interval Ts of an image patch is approximately 3 seconds. In one example, slow fluctuations in the predicted change in the error amplitude, commonly referred to as drift, are separated from fast dynamic fluctuations in the predicted evolution of the error amplitude, commonly referred to as dynamic changes. In one example, the control unit 800 is configured to predict the change in at least a subset of the error amplitude over a time interval Tr during which the stage moves from a first image patch to a second image patch. The time interval Tr during which the stage moves from the first image patch to the second image patch ranges from 0.5 to 5 seconds. In one example, the control unit 800 is configured to predict a change in at least a subset of the error amplitudes during a time interval Td when the stage decelerates from high speed movement to a rest position. Generally, the control unit 800 is configured to derive at least one of the plurality of control signals based on a predictive model of the stage actuation output.

[0073] In one embodiment, the slowly varying part (drift) and the dynamic part (i.e., the dynamic change course) of the error amplitude are extrapolated separately. The drift part typically exhibits, for example, linear or asymptotic behavior. For example, thermal effects typically result in a slow drift with asymptotic behavior. Using prior knowledge of the time course of the error amplitude, the drift course is derived based on a model function of a predetermined asymptotic behavior, and the control unit 800 is configured to generate a control signal that anticipates the predicted error amplitude. The slowly varying course or drift of the error amplitude is separated from the fast course, and the stage 500 is controlled, for example, by direct transfer of the error amplitude drift. Figures 4 and 5 illustrate the evolution of a typical error amplitude over time. Figure 4a shows an example of the drift or slowly varying error amplitude Sn(t) including the predetermined asymptotic behavior of the error amplitude model function 907 with respect to time t. Such behavior is typically for thermal drift or the drift of electrostatic or electromagnetic elements, but other effects also have similar time courses. Other sources of drift may include variable electrostrictive forces or drift induced by charging of the conductive parts or wafer during image scanning. In operation, the control unit 800 is configured to continuously derive a drift error amplitude Sn(t) from the sensor data. The operating time includes a first time interval Ts1 of the first image scan of the first image patch 17.1, a time interval Tr of the wafer stage movement from the first center position 21.1 of the first image patch 17.1 to the second center position 21.2 of the second image patch 17.2, and a second time interval Ts2 of the second image scan of the second image patch 17.1 (see FIG. 2 for reference numerals). For example, at an actual time Ta in the first time interval Ts1, a temporal gradient 903 of the error amplitude Sn(t) is determined, or a model function 907 is fitted to the measured error amplitude Sn(t). The error amplitude model function 907 or gradient vector 903 predicts the evolution of the error amplitude Sn(t) and predicts that at a future time tc in the second time interval Ts2, the drift portion of the error amplitude Sn(t) will reach a predetermined threshold Sn_max as shown by line 901. This threshold is predetermined, for example, from specifications of the image quality parameters associated with the error vector Sn(t).During the time interval Tr between two subsequent image scans of the two image patches 17.1 and 17.2, the control unit 800 is configured to change the control values ​​of the compensators accordingly, so that the drift component of the error amplitude Sn(t) is reduced by adjusting the active elements of the multi-beam charged particle microscope 1. The active elements may include slow-acting compensators (e.g., magnetic elements or stages). In one embodiment of the present invention, the lateral drift of the line of sight 53 or the image coordinate system 51 is compensated for, for example, by adding an offset to the lateral position of the wafer stage 500, and the drift of the focus position is compensated for, for example, by adding an offset to the z-position of the wafer stage 500. In one embodiment of the present invention, the drift of the imaging magnification of the multiple primary charged particle beamlets 3, which results in a change in the pitch of the multiple primary charged particle beamlets 3, is compensated for, for example, by adding an offset current to a dedicated magnetic lens element of the objective lens 102. In one embodiment of the present invention, the rotational drift of the multiple primary charged particle beamlets 3 as described in FIG. 3 is compensated by a corresponding rotation of the stage 500 about the z-axis or by correcting the rotation of the multiple primary charged particle beamlets 3, for example, by adding an offset current to the second dedicated magnetic lens element of the objective lens 102. The result is shown in FIG. 4b. This adjustment sufficiently controls the corrected slowly varying drift error amplitude Sn(t) above the error amplitude threshold Sn_max. Due to the slow variation in time of the drift portion, the error amplitude Sn(t) can be at least partially adjusted and compensated for during the time Tr between subsequent image scans. This provides a method for wafer inspection with a multi-beam charged particle microscope, comprising: a first image acquisition step of a first image patch in a first time interval Ts1; moving the wafer stage from the position of the first image patch to the second image patch during a time interval Tr; a second image acquisition step of a second image patch in a second time interval Ts2; Therefore, calculating at least a first error amplitude from a plurality of sensor signals during a first time interval Ts1; predicting a progression of the first error amplitude over at least the first time interval Ts1, the travel time interval Tr, and the second time interval Ts2 in a first time interval Ts1; - during at least the transfer time interval Tr, supplying a control signal to the control unit of the multi-beam charged particle microscope to keep the predicted error amplitude progression below a predetermined threshold during a second time interval Ts2; A method is provided, comprising:

[0074] In one example, the prediction of the evolution of the first error amplitude is generated by a predictive model or by extrapolation.

[0075] In one example, a control signal is provided to a control unit of a multi-beam charged particle microscope to keep the predicted error amplitude progression below a predetermined threshold during an image scan of time interval Ts1 or Ts2. For example, if a slow drift of the image coordinate system is predicted, the drift can be compensated for by a slow compensating movement of stage 500 such that Sn(t) is controlled sufficiently below a threshold Sn_max during image acquisition.

[0076] 5 shows the fast dynamic change of the image misalignment, represented by the dynamic change of the error amplitude Nn(t). Such dynamic change of the image misalignment can be introduced by internal noise sources, such as a vacuum pump or other internal noise sources, such as vibrations induced by fast acceleration and deceleration of the wafer stage. Other noise sources can be external sources.

[0077] The dynamic change Nn(t) exhibits a simple periodic behavior, with a half-period shorter than the scanning time interval Ts1 or Ts2 of one image patch. In one embodiment of the present invention, the control unit 800 is configured to derive the dynamic change of the error amplitude Nn(t) and to rapidly determine control signals for high-speed active elements of the multi-beam charged particle microscope 1. Such active elements are, for example, electrostatic beam deflection scanners or electrostatic correctors, and are rapidly adjustable. During the image scanning of the first image patch in the scanning time Ts, the uncontrolled error amplitude Nn(t) exceeds a predetermined error amplitude window DNn, indicated by reference numeral 905, at least twice, at times tc1 and tc2. The error amplitude window 905, which has upper and lower thresholds for the error amplitude Nn(t), represents the specification requirement for the image quality parameter represented by the error amplitude Nn(t). The control unit 800 is further configured to provide dynamic control signals to the control units of the high-speed active elements so that the corrected dynamic deviation or error amplitude Nn(t) shown in FIG. 5b is controlled between upper and lower thresholds of a predetermined error amplitude window 905. The control operator 800 is configured with a high-speed control loop (e.g., an open control loop) that allows adjustment and control over a bandwidth at least 50 times, preferably at least 100 times, and more preferably 1000 times the image scanning frequency or frame rate 1 / Ts of approximately 0.3 Hz. In one example, calculation of the error vector and extraction of the control signal for compensation in the case of imaging aberrations are performed at least once per line scan with a control frequency of approximately 2.5 kHz or greater. Thus, the electrical control signal includes a signal having a bandwidth in the range of 0.1 kHz to 10 kHz or greater.

[0078] Note that the frequency of the corrected error magnitude Nn(t) shown in FIG. 5b may differ from the frequency of the uncorrected error magnitude Nn(t) shown in FIG. 5a depending on the frequency response of the control loop of the control operator 800.

[0079] In one example, the control unit 800 is configured to predict dynamic changes in the error amplitude Nn(t). For example, the control unit 800 is configured to derive a control signal for a high-speed active element for dynamic control during image scanning in a time interval Ts1 by deriving a local gradient 909 at time Ta of the error amplitude Nn(t).

[0080] The driving error source of the wafer stage misalignment or drift is in the time interval Tr given to move the stage from the first image patch 17.1 to the second image patch 17.2. In particular, the wafer stage misalignment or drift depends on the number of adjustment iterations and the time Td required to decelerate the stage from the moving speed to a stop position near the second image patch 17.2. In one embodiment of the present invention, the time difference between the two image acquisition steps of the first image patch 17.1 and the second image patch 17.2 is significantly reduced, thereby improving throughput. According to one embodiment of the present invention, a charged particle microscope and a method for operating a charged particle microscope are provided, which include a first image acquisition of a first image patch 17.1 during a first time interval Ts1 and a second image acquisition of a second image patch 17.2 during a second time interval Ts2, and further include a third time interval Tr for moving a wafer stage 500 from a first center position 21.1 of the first image patch 17.1 to a second center position 21.2 of the second image patch 17.2, such that at least one of the first and second time intervals Ts1 and Ts2 overlaps with the third time interval Tr. The total time interval from the start of the first time interval Ts1 to the end of the second time interval Ts2 is shorter than the sum of the three time intervals Ts1, Tr, and Ts2, thereby improving throughput and enabling a high-speed inspection mode. Figures 5a and 5c illustrate this embodiment in a high-throughput, high-speed inspection mode. In the first example of FIG. 5a, image acquisition of the second image patch 17.2 begins before the wafer stage 500 comes to a complete stop. Image acquisition begins during the time period Td during which the wafer stage is decelerated to its end position, and the image acquisition time period Ts2 overlaps with the time period Td for deceleration of the stage 500. The deceleration time period Td includes repeated stage adjustments and the time required for the stage to come to a complete stop. After high-speed movement, the stage may drift, wobble, or vibrate, and the time period Td for stage deceleration includes the time required for the stage to slow down until its position is aligned with the line of sight of the multi-beam charged particle microscope with an accuracy below a first predetermined threshold and a dynamic position stability below a second predetermined threshold.The control unit 800 is configured to monitor or predict the expected lateral positions Xl(t), Yl(t) and the movement velocity of the wafer stage 500 at time Td. The control unit 800 derives a control signal for the scan deflection unit of the charged particle microscope to compensate for residual movement of the wafer stage during the deceleration time Td by variable offsets Dx(t), Dy(t) of the line of sight 53. The control unit 800 is configured to calculate the start time of image acquisition of the second image patch 17.2 from the predicted movement velocity of the wafer stage. For example, the start time of the time interval Ts2 of image acquisition of the second image patch 17.2 is determined as the time when the predicted velocity of the wafer stage falls below a predetermined threshold so that residual movement of the wafer stage during the deceleration time interval Td can be compensated for. The control unit 800 is configured to start image acquisition by scanning the second image patch 17.2 and to compensate for residual movement of the wafer stage by applying a time function of the offset coordinate to the deflection unit during at least a portion of a deceleration time interval Td that overlaps with the time interval Ts2 of the second image acquisition, resulting in a shortened time interval Tr' between the first image scan in the first time interval Ts1 and the second image scan in the second time interval Ts2.

[0081] 5c shows a second example of this embodiment in more detail. In this example, the control unit 800 is configured to derive a start time r1 of wafer stage acceleration during the time interval Ts1 of the first image acquisition of the first image patch so that compensation for wafer movement due to the deflector scanning is within the maximum range of the scanning deflector of the charged particle microscope. During image acquisition and at least a portion of the time interval Tu for accelerating the wafer stage, the control unit 800 is configured to provide a control signal to the deflection unit so that the error amplitude Nn(t) (in this example, the lateral position offset in the coordinate system as described above) is sufficiently within the specified threshold range 905.1 for the lateral position offset of the first image patch, and image acquisition continues after the start time r1 of wafer movement until the end time t1 of the time interval Ts1 of the first image acquisition. During wafer movement in the time interval Tr, the control unit 800 is configured to derive a start time t0' of the second time interval Ts2 of the second image acquisition of the second image patch so that compensation of the wafer movement by the scan deflector is within the maximum range of the scan deflector of the charged particle microscope and the lateral position offset of the coordinate system is sufficiently within a specified threshold range 905.2 for the lateral position offset of the second image patch. The second image acquisition begins at start time t0' during wafer movement and ends at time r2 when the wafer stage reaches the vicinity of its target position. As a result, the time interval Tr' between the first image scan in the first time interval Ts1 and the second image scan in the second time interval Ts2 is short. The overlap time interval between the start t0' of the second image acquisition and the deceleration of the wafer stage until the end time rs of the wafer stage movement time interval Tr is typically longer than the overlap time interval between the end t1 of the first image acquisition and the acceleration of the wafer stage, which begins at time ri of the wafer stage movement time interval Tr. In one example, the wafer stage deceleration time interval Td includes at least one iteration of precise alignment of the wafer stage during image acquisition of the image patch at each inspection site, so that wafer movement is controlled by the control unit in synchronization with the deflection unit, and the predicted and monitored positions of the wafer stage during wafer stage movement are compensated by a set of offset coordinates or functions applied to the deflection unit corresponding to the position of the wafer stage.The iterative precision alignment of the wafer stage involves repeated readjustment of the wafer stage position from a first position that is significantly off-target to a second position where the deviation from the target position is below a predetermined threshold. In one example, the threshold is determined by a reduction in the overlap area between two adjacent image patches, e.g., less than 100 nm, less than 50 nm, or less than 30 nm. This improves throughput by shortening the time interval between subsequent image acquisitions and by reducing the overlap area between adjacent image patches. In one example, the time interval between two subsequent image acquisitions of the first and second image patches is shortened by a factor of two, improving the throughput or frame rate of the multi-beam charged particle microscope by approximately 10 to approximately 14 frames per minute. In another example, the time interval between two subsequent image acquisitions of the first and second image patches is shortened by a factor of three, and the throughput or frame rate of the multi-beam charged particle microscope is increased by more than approximately 10-15 frames / minute, and the method for controlling image quality during wafer movement according to the present invention increases throughput by more than 50%. In general, the provided method enables image acquisition of at least two separate image patches 17.1 and 17.2 within a shorter time interval T G compared to the time intervals T S1 and T S2 required for acquiring the two separate image patches 17.1 and 17.2, respectively, and the time T R required to move the sample from the first inspection site to the second inspection site (T G ). <Ts1+Ts2+Tr)。

[0082] The separation of the error amplitude profile into drift and dynamic change is achieved, for example, by fast Fourier analysis or moving average calculation. Other methods known in the art are also possible. In one example, a predetermined threshold value for the maximum slope of the error amplitude profile is applied to obtain a decomposition into a linear drift at the maximum slope and a residual dynamic change where the error amplitude portion exceeds the maximum slope. The linear portion of the error amplitude below the maximum slope is subtracted, and the dynamic change is obtained by the error amplitude profile from which the linear drift has been subtracted. The maximum slope of the error amplitude is determined by the maximum speed of the compensator that compensates for the linear drift. Such a slow-acting compensator can be, for example, a magnetic element in a multi-beam charged particle microscope. In another example, a predetermined threshold value for the maximum frequency of the error amplitude profile is applied, and the drift portion is determined by low-pass filtering of the error amplitude profile. In one example, the separation into drift and dynamic portions takes into account the dwell time, line scan rate, and frame rate. For example, with a dwell time of 50 ns, the line scan rate is approximately 2.5 kHz. Changes or deviations in imaging performance in the frequency range of approximately 10 kHz or greater can be compensated for by the control unit 800 and the high-speed compensator of the multi-beam charged particle microscope. Thus, during multiple line scans with multiple primary charged particle beamlets, rapid and dynamic changes or deviations from imaging performance can be controlled. Thus, dynamic changes are compensated for multiple times during image acquisition over a time interval Ts of approximately 3 s (e.g., every flyback with a control frequency of approximately 2.5 kHz) or during each line scan with a control frequency greater than 2.5 kHz (e.g., 5 kHz or 10 kHz or greater). Slow drifts over a time interval of seconds are compensated for by the low-speed compensator within a time interval Tr' of, for example, less than approximately 0.5 s, for example, during the time interval Tr' between two consecutive image scans. To synchronize compensators with different response times, a delay line can be included in the control unit, for example.

[0083] The prediction of the error amplitude progression is calculated according to an approximation using polynomial expansion and extrapolation (e.g., linear extrapolation), although other higher-order extrapolations, such as second-order or higher, are also possible. An example of higher-order polynomial extrapolation is the Runge-Kutta method. In the example of a slow-fluctuation compensator, such as a moving wafer stage, the prediction of the error amplitude progression (e.g., stage position) is achieved by controlling and monitoring the calibration performance of the slow-fluctuation compensator, such as the wafer stage. The prediction of the error amplitude progression can also be based on a model; a so-called model-based predictor generates a predicted error amplitude according to a model function of the expected progression of the error amplitude. Such a predetermined model function is generated by simulation or representative test operations of the multi-beam charged particle microscope and stored in the memory of the control unit 800. In one example, such a predetermined model function is individual for each individual multi-beam charged particle microscope. In many examples, the estimation of error behavior according to a predictive model includes frequency analysis, low-pass filtering, and polynomial approximation.

[0084] The above-mentioned transition and extrapolation methods, such as applying a model function with separation or prior knowledge to drift and dynamic changes of error amplitude, can be selected differently for different deviations of the image performance parameters represented by the time evolution of the error amplitude. In one example, the control unit 800 performs the following during image acquisition of a series of image patches: A) expanding a data stream comprising a plurality of sensor data into a set of error amplitudes; E) extracting a set of drift control signals and a set of dynamic control signals; F) providing a set of drift control signals to a slow acting compensator; G) providing a set of dynamic control signals to a fast acting compensator; The device is configured to perform a series of operational steps including:

[0085] In one embodiment, the control unit 800 is further configured to include a step B of approximating a time course of at least one of the error amplitudes. In one embodiment, the control unit 800 is further configured to include a step C of predicting a slow-varying drift of at least one of the error amplitudes. In one embodiment, the control unit 800 is further configured to include a step D of predicting a fast-varying dynamic change of at least one of the error amplitudes.

[0086] In one example, configuring the control unit 800 includes performing step G of providing a set of dynamic control signals to a fast acting compensator during a time interval Ts1 of an image scan of a first image patch in the series of image patches.

[0087] In one example, the setting of the control unit 800 includes executing step F of providing a set of drift control signals to the slow motion compensator during a time interval Tr between a first image scan of a first image patch and a subsequent second image scan of a second image patch in the series of image patches. The time interval Tr is defined by the time interval required for the wafer stage 500 to move from a first center position of a first image patch to a second center position of a subsequent second image patch obtained by scanning imaging with the multi-beam charged particle microscope 1. In one example, the setting of the control unit 800 includes executing step F of providing a set of drift control parameters to the slow motion compensator during a time interval Ts of one image scan of one image patch.

[0088] In one example, configuring the control unit 800 includes performing step G during a time interval Ts1 or Ts2 of image scanning of the image patch during at least one overlapping time interval that overlaps with a time interval Tr of stage movement. In one example, the at least one overlapping time interval is at least a portion of a time interval Tu for accelerating the wafer stage, at least a portion of a time interval Td for decelerating the wafer stage, or both time intervals.

[0089] One embodiment of the present invention is a method of operation of the multi-beamlet charged particle microscope system 1 to perform a wafer inspection task and a software product for such wafer inspection task. The method for performing the wafer inspection task includes software code for performing steps A-G described above. This method is described in more detail in FIG. 7 below.

[0090] Therefore, in one embodiment of the present invention, a multi-beamlet charged particle microscope system 1 for wafer inspection has multiple measures to compensate for drift, dynamic effects, and residual and unwanted stage movement. An example is shown in FIG. 6. The same reference numerals as in the previous drawings are used, so reference is made to the previous drawings. The multi-beam charged particle microscope (1) for wafer inspection comprises a charged particle multi-beamlet generator (300) that generates multiple primary charged particle beamlets (3), and an object illumination unit (100) with a first deflection system (110) that scans an area of ​​a wafer surface (25) arranged in an object plane (101) with the multiple primary charged particle beamlets (3) to generate multiple secondary electron beamlets (9) emitted from the wafer surface (25). The plurality of secondary electron beamlets (9) are imaged by a detection unit (200) comprising a projection system (205) and a second deflection system (222) for imaging the plurality of secondary electron beamlets (9) onto an image sensor (207) and for, in use, acquiring a digital image of a first image patch (17.1) of the wafer surface (25). The multi-beam charged particle microscope (1) further comprises a sample stage (500) comprising a stage position sensor (520) for positioning and holding the wafer surface (25) in the object plane (101) during acquisition of the digital image of the first image patch (17.1).

[0091] The multi-beam charged particle microscope 1 includes a set of compensators including at least first and second deflection systems (110, 222) and a set of slow-acting compensators, such as magnetic elements or mechanical actuators. In one example, the slow-acting compensators include a wafer stage 500. The set of compensators further includes a set of fast-acting compensators (132, 232, 332), such as electrostatic elements or low-mass mechanical actuators. The multi-beamlet charged particle microscope system 1 includes a stage position sensor (520) and an image sensor (207), and includes a plurality of detectors configured to generate a plurality of sensor data, when in use. The plurality of sensor data includes position and orientation data of the sample stage (500) provided by the stage position sensor (520).

[0092] The multi-beam charged particle microscope 1 further includes a control unit (800), which generates a first set of P control signals C from the plurality of sensor data. p In use, operational control is achieved by generating a set of compensators configured to control a set of compensators during acquisition of a digital image of a first image patch (17.1) such that the aforementioned specifications are maintained during image acquisition of a series of image patches.

[0093] During stage movement of the stage 500, the stage movement is monitored by a stage position sensor 520. The stage position sensor 520 is known in the art and may include a laser interferometer, a grating sensor, or a confocal lens array sensor. During the time interval Ts of an image scan of one image patch in the series of image patches, the relative position of the wafer stage 500 is preferably controlled with high stability (e.g., less than 1 nm, preferably less than 0.5 nm). As described above, between the first and second image scans of the first and subsequent second image patches, the stage 500 is triggered by the control unit 800 to move from the first inspection site to the second inspection site. At the second inspection site, a new local wafer coordinate system is defined and controlled by the stage control module 880 to position the stage 500 at its predicted position, with its position relative to the line of sight controlled with high stability. The stage position sensor 520 measures the stage position and movement in six degrees of freedom with an accuracy of less than 1 nm, preferably less than 0.5 nm. In one example (not shown), the stage position sensor 520, through a direct connection to the stage control module 880, forms a direct feedback loop for the control of stage position and movement. However, such a direct feedback loop and control of a high-mass wafer stage are typically slow and do not provide sufficient precision during image scanning. The feedback loop can also introduce unwanted stage jitter or lag. Therefore, according to an embodiment of the present invention, the stage position sensor 520 is connected to the sensor data analysis system 818 of the control unit 800.

[0094] According to one example of this embodiment, the control unit (800) includes an image data acquisition unit (810) configured, in use, to reduce image sensor data from the image sensor (207) to an image sensor data fragment representing, for example, less than 10% of the image sensor data and provide the image sensor data fragment to a sensor data analysis system (818). In use, the electron-sensitive image sensor 207 receives a large image data stream of image sensor data of multiple secondary electron intensity values ​​and provides the image data to the image data acquisition unit 810 of the control unit 800. This large amount of image data is not directly used for monitoring image operations of the multi-beamlet charged particle microscope system 1. A small fraction of the image data stream is then branched off, and the image sensor data fragment is guided to the sensor data analysis system 818. For example, the image data acquisition unit 810 is configured to branch off a subset of secondary charged particle signals generated at a predetermined scan position of the multiple charged particle beamlets. Alternatively, signals generated during the flyback of the scanning charged particle beam 3 are extracted and forwarded to the sensor data analysis system 818. The predetermined scan position can be, for example, the line start position of a subset of scan lines (e.g., every fifth scan line or the center position of each). In one example, image data from a subset of primary charged particle beamlets (e.g., only one beamlet at spot position 5.11 (see FIG. 2)) is used to generate the image sensor data fragments. U.S. Pat. No. 9,530,613 shows an example of a dedicated subset of peripherally arranged primary charged particle beamlets to provide a sensor signal for controlling a multi-beam charged particle microscope, and is incorporated herein by reference. U.S. Pat. No. 9,536,702 shows an example of generating a live view image by branching a dedicated subset of image data from each of multiple subfields, and is incorporated herein by reference. At least a portion of the live view image data can be applied as the image sensor data fragments.By branching off signals from a predetermined subset of the charged particle beamlets or using signals at predetermined scan positions of the charged particle beamlets, the image sensor data fraction transferred to the sensor data analysis system 818 is significantly reduced to a small fraction of the image data stream, approximately less than 2%, less than 1%, preferably less than 0.5%, more preferably less than 0.1%, or even less than 0.01%. In one embodiment, the image sensor 207 includes a first low-speed, high-resolution image sensor and a second high-speed image sensor, as described above in conjunction with FIG. 1. In this embodiment, the image sensor data fraction is formed by sensor data provided by the first low-speed image sensor, and the image data acquisition unit 810 is configured to provide the sensor data provided by the first low-speed image sensor to the sensor data analysis system 818 and to provide the sensor signal of the second high-speed image sensor to the image stitching unit 812.

[0095] The image sensor data fragments and the stage position data from the stage sensor 520 are combined in a sensor data analysis system 818. The sensor data analysis system 818 analyzes the image sensor data fragments from the image sensor 207 and the position information from the stage sensor 520, as described in the example of Figure 3, and extracts wafer stage position information relative to the actual image coordinate system of the multiple primary charged particle beamlets 3.

[0096] The control unit (800) of the multi-beam charged particle microscope (1) includes a sensor data analysis system (818) configured to derive a sensor data vector DV of length L from a plurality of sensor data and to analyze the sensor data vector DV to extract error functions such as image displacement, image rotation, and changes in focus position and image plane tilt from the sensor data vector DV. The sensor data analysis system 818 calculates a set of K amplitudes A of K error vectors (K≦L). kGenerally, the sensor data analysis system 818 is configured to analyze the plurality of sensor signals and decompose the plurality of sensor signals into a set of normalized error functions by methods known in the art (e.g., fitting a set of normalized error functions to the plurality of sensor signals).

[0097] The control unit (800) calculates a set of amplitudes A of the error vector. k to a first set of control signals C p In the example of dynamic changes in the lateral displacement of the image point 5, the control arithmetic processor 840 is configured to derive corrections or control signals for dynamic changes in the error amplitude. The control unit (800) generates a first set of control signals C pand supplying at least one of the control signals to the first and second deflection units (110, 222) to compensate for changes in the position or orientation of the sample stage (500). The sensor data from the image sensor 207 is synchronized and combined with information from the stage position sensor 520. A relative lateral displacement vector 55 between the local wafer coordinate system at the inspection site and the image coordinate system defined by the line of sight is then derived by the sensor data analysis system 818. The control and calculation processor 840 is configured to supply correction or control signals to a deflection control module 860, which controls the operation of the first scanning deflector 110 of the multi-beamlet charged particle microscope system 1. As a result, the first electrostatic scanning deflector 110 controls the scanning movement of the primary charged particle beamlet 3 in synchronization with unwanted dynamic changes of the wafer stage 500 in the lateral directions (here, the x and y directions). In parallel, the deflection control module 860 controls the operation of the second scanning deflector 222 such that the positions of the plurality of secondary electron beamlets 9 on the image sensor 207 remain constant. The control unit 800 is thereby configured to compensate for dynamic changes in the position of the lateral stage 500 by correcting the scanning movement of the primary and secondary charged particle beamlets by the first and second deflectors 110 and 222, such that image acquisition with high image fidelity and high image contrast is sufficiently maintained within the requirements or specifications of the wafer inspection task. The control unit 800 is therefore configured to calculate and apply at least additional voltage signals to the beam deflector 110 in the primary charged particle beam path 13 to generate, in use, an additional displacement or rotation of the plurality of primary charged particle beamlets 3 to at least partially compensate for the lateral displacement or rotation of the stage relative to the line of sight. Therefore, the control unit 800 is configured to at least partially compensate for the additional displacement or rotation of the multiple secondary electron beamlets resulting from the beam spot 5 of the multiple primary charged particle beamlets 3 during the adjustment scan by calculating and applying at least a second additional voltage signal to the beam deflector 222 in the secondary electron beam path 11.

[0098] The following shows an example of an error function for drift of the wafer stage 500 in the vertical, i.e., z-direction. The set of compensators in the multi-beam charged particle microscope 1 includes at least one of the compensator (332) of the charged particle multi-beamlet generator (300), the fast compensator (132) of the object illumination unit, and the compensators (230, 232) of the detection unit (200). Again, the image sensor data fragments, along with stage position data from the stage position sensor 520, are analyzed by the sensor data analysis system 818. The sensor data analysis system 818 analyzes the image sensor data fragments from the image sensor 207 and the position information from the stage sensor 520, and extracts position information of the wafer stage relative to the actual scanning positions and line of sight of the multiple primary charged particle beamlets 3. The control arithmetic processor 840 extracts control signals for focus control of the multiple primary and secondary charged particle beamlets 3 and 9. Therefore, the control arithmetic processor 840 of the control unit 800 is connected via the primary beam path control module 830 to at least one fast compensator 332 (e.g., an electrostatic focusing lens in the primary charged particle beam path 13, such as the electrostatic field lens 308 (see FIG. 1)) of the multi-beamlet generator 300 or a fast compensator 132 of the object illumination unit that controls the focus of the multiple primary charged particle beamlets 3. The control arithmetic processor 840 is also connected to a projection system control module 820 that controls at least one fast compensator 232 (e.g., the electrostatic focusing lens 206 (see FIG. 1)) of the detection unit 200 so that the focus positions of the multiple secondary electron beamlets 9 on the image sensor 207 are kept constant. In this way, the primary beam path control module 830 and the projection system control module 820 compensate for the stage drift of the stage 500 in the vertical direction, i.e., the z-direction, so that high-contrast and high-resolution image acquisition is maintained sufficiently within the requirements or specifications of the wafer inspection task.

[0099] In one example, the sensor data analysis system (818) of the multi-beam charged particle microscope (1) calculates a set of amplitudes A of the error vectors. k At least one of the amplitudes A nThe control unit (800) is further configured to generate a third signal for positioning the wafer surface (25) by the wafer stage (500) in the object plane (101) for image acquisition of the digital image of the second image patch (17.2), and to control a set of compensators in positioning the wafer stage (500) relative to the position of the second image patch (17.2) by providing a second set of drift control signals from the plurality of sensor data. In one example, the control arithmetic processor 840 is connected to the primary beam path control module 830. The primary beam path control module 830 is connected to at least one slow compensator 130 of the object illumination unit 100 or the magnetic element 430 of the beam splitter 400 (see FIG. 1 ) to correct the rotational drift, i.e., slow fluctuation, of the set of primary charged particle beamlets 3. In one example, static image rotation is further compensated for by a predetermined rotation of the image sensor 207. The control arithmetic processor 840 is further connected to the projection system control module 820, which controls the slow compensator 230 (e.g., a magnetic lens) of the secondary electron beam path. However, the magnetic element can only compensate for a limited speed of rotational drift.

[0100] In one example, the charged particle multi-beamlet generator (300) further comprises a fast compensator (332), and the control unit (800) generates a first set of control signals C pand supplying at least one of the control signals to a fast compensator (332) to induce rotation of the plurality of primary charged particle beamlets, thereby compensating for rotation of the sample stage (500). For example, dynamic changes in the rotation of the wafer stage cause fast changes and deviations of the plurality of primary charged particle beamlets from a predetermined orientation relative to the wafer stage. In this example, the dynamic changes in rotation are fast compensated for. The control arithmetic processor 840 is connected to the primary beam path control module 830. In this example, the primary beam path control module 830 is further connected to the fast compensator 332 (e.g., active multi-aperture plate 306.3 (see FIG. 1)) of the multi-beamlet generator 300, which in this example comprises an electrostatic deflector array that compensates for unwanted dynamic changes in the rotation of the set of primary beamlets 3 relative to the local wafer coordinate system by fast deflection of each primary charged particle beamlet individually. The projection system control module 820 is connected to a high speed compensator 232 of the detection unit 200 with a second multi-aperture plate, which may include, for example, an array of electrostatic deflectors that compensate for unwanted dynamic changes in the rotation of the plurality of secondary electron beamlets 9. This compensates for image rotation during image scanning of an image patch in the series of image patches, and maintains high image fidelity and image contrast well within the specifications of the wafer inspection task.

[0101] In one example, the stage position sensor 520 comprises a position and rotation detection sensor, such as a dual interferometer for each of the x and y axes.

[0102] In one example, compensation for wafer stage rotation is performed during the time interval Tr of wafer stage movement from the first image patch to the second image patch, as described above in conjunction with Figures 5a and 5c, thereby improving throughput.

[0103] In one embodiment, the control and arithmetic processor 840 is further connected to the image stitching unit 812. The image stitching unit 812 receives the large image data stream from the image data acquisition unit 810 and converts the image data stream into a 2D image by time-series deconvolution of the data stream and image stitching of image subfields 27 to obtain a single image patch 17 (see FIG. 2). A 2D image representation of an area on the wafer surface 25 is obtained by stitching multiple image patches (e.g., first and second image patches 17.1 and 17.2) together. To compensate for high-speed image rotation due to, for example, stage jitter and high-speed rotation of the wafer 7 relative to the image coordinate system, the control and arithmetic processor 840 is configured to extract residual rotation of multiple image spots 5 during scanning and provide the residual rotation of the spots 5 to the image stitching unit 812. The image stitching unit 812 is configured to obtain a 2D image with high image fidelity from the data stream of a single image patch by compensating for the residual rotation of the spots 5 using known digital image processing methods. The final image is finally compressed and stored in image data memory 814.

[0104] In one example, the control arithmetic processor 840 of the control unit 800 is configured to compensate for image rotation by parallel drift and dynamic compensation. Since the range of image rotation compensation by a multi-aperture plate configured as a deflector array is limited, the range of fast-varying dynamic compensation can be reduced by continuously changing the slow-varying drift offset using a drift compensator (130, 230, 330) including a magnetic lens, and can be realized by a multi-aperture plate configured as a deflector array in the primary charged particle beam path 13 and the secondary electron beam path 11.

[0105] Next, an example of an error function in which multiple image spots 5 are formed in an image plane tilted with respect to the wafer surface is shown. In this example, the control arithmetic processor 840 is configured to derive an image tilt correction signal that is forwarded to the primary beam path control module 830. The primary beam path control module 830 is configured to control the fast compensator 332 (e.g., the active multi-aperture plate 306 (see FIG. 1 )) of the multi-beamlet generator 300 configured to change the focal position of each primary charged particle beamlet 3, thereby effectively realizing a tilted focal plane of the multiple focal spots 5. This allows each primary charged particle beamlet 3 to be focused at the wafer surface 25 even when the wafer stage 500 is tilted or its tilt angle is changed. The control arithmetic processor 840 is further connected to the projection system control module 820, which controls the fast compensator 232 (e.g., including the multi-aperture corrector 220) of the detection unit 200. The fast compensator 232 of the detection unit 200 corrects the focal position of each secondary electron beamlet 9 so that the beam spot 15 remains constant at the focal position of the image sensor 207. The control and arithmetic processor 840, the primary beam path control module 830, and the projection system control module 820 are thereby configured to compensate for image tilt, maintaining high contrast and high resolution image acquisition across the image patch 17.

[0106] In one example of the present invention, similar to the above example, a direct feedback loop is provided in the control unit 800 between the stage position sensor 520 and the first deflection system 110, where the control unit 800 is configured to receive a stage position signal from the stage position sensor 520 and provide at least a first offset signal to the first deflection system 110 to control the first deflection system 110 to compensate for movement of the wafer stage 500 and deviation of the wafer stage 500 from a target position. The control unit 800 is further configured to provide at least a corresponding second offset signal to the second deflection system 222. This allows for fast compensation of wafer stage position errors or movements, improving throughput while maintaining the required specifications for the wafer inspection task.

[0107] Of course, the above examples can be used in parallel as well as independently. The above-described apparatus and error correction method are not limited to the above examples. The control and calculation processor 840 is configured to derive control signals in parallel by direct feedback, predictive correction, or model-based correction as described above for a set of error amplitudes of a set of imaging misalignments. In one example, the projection system control module 820 is connected to the sample voltage source 503 to control the extraction field for extracting secondary charged particles, thereby controlling the secondary electron collection efficiency and thus the intensity of the secondary electron beamlet 9, as well as the kinetic energy of the secondary electrons. The kinetic energy is responsible for several other characteristics, such as image contrast. In one example, the projection system control module 820 is connected to other active elements 230 and 232 (such as the third deflection system 218) of the detection unit 200 or correctors such as the multipole lens 216 (see FIG. 1). In one example, a sensor 238 (such as a sensor on an aperture element) in the secondary electron beam path supplies an additional sensor signal to the sensor data analysis system 818. In one example, a multipole sensor is arranged around the periphery of the aperture element 214 located at the crossover 212 of the secondary charged particle beam path 11. The signal provided by the multipole sensor measures the decentering state of the secondary charged particle beam path 11. In another example, the charged particle microscope 1 includes an active high-speed element, such as a beam steering multi-aperture plate 390 (see FIG. 1 ), for decentering correction of multiple primary charged particle beamlets 3. The beam steering multi-aperture plate 390 is connected to a primary beam path control module 830, which receives control signals from a control arithmetic processor 840. In one example, a sensor 138 (such as a sensor near an aperture element or on a multi-aperture plate) included in the object illumination unit 100 provides additional sensor signals to the sensor data analysis system 818. In one example, an array of coils is included in different orientations to measure electromagnetic noise. In one example, the primary beam path control module 830 is coupled to the particle source 301 and configured to control the particle source power or charged particle dose provided by the particle source 301, thereby maintaining a constant charged particle dose throughout a series of image scans of a set of image patches.In one example, a vibration sensor, such as an accelerometer or gyroscope, is attached to an element of the charged particle microscope (e.g., wafer stage 500). The vibration sensor measures vibrations and provides a signal to the sensor data analysis system 818. A temperature sensor (e.g., a temperature sensor in a magnetic lens or in the return path of a cooling fluid) provides an indication of the status of the system's elements and the expected drift behavior of some image quality. All sensor signals can be calibrated, for example, during a simulated inspection task on a test sample, to provide representative sensor data for a wafer inspection task. The representative sensor data can be used to establish a sensor data vector and to extract the amplitude of the normalized error vector from the sensor data vector of a real wafer inspection task.

[0108] Generally, the control and arithmetic processor 840 of the control unit 800 is configured to derive a correction signal from the error amplitude to compensate for slow-varying transitions of the error function, such as slow drift of the stage 500. From dynamic changes in the error amplitude, the control and arithmetic processor 840 derives a correction strategy for fast compensation of the dynamic changes and distributes control signals to the primary beamlet control module 830, the projection system control module 820, and the deflection control module 860 to compensate for fast or dynamic changes in the error amplitude, such as fast vibrations of the stage 500. The drift and dynamic changes in the error amplitude are calculated by the sensor data analysis system 818 of the control unit 800, but they can also be derived directly based on extrapolation or model-based control. The correction strategy can follow a look-up table. Alternatively, a linear decomposition is used to decompose the error amplitude into predetermined correction functions provided by different active elements of the charged particle microscope 1. Therefore, the control and arithmetic processor 840 also monitors the actual status and status changes of the active elements of the charged particle microscope 1. In one example, the control and calculation processor 840 is configured to predict the actual status of the active elements of the charged particle microscope 1 by accumulating a history of control signals provided to the active elements, such as the secondary electron path active elements 230, 232, the primary beam path active elements 330 and 332, and the deflector units 110 or 222.

[0109] One aspect of the present invention is the derivation of error vectors and drive signals to drive compensators to optimize image quality parameters when using a multi-beam charged particle microscope, as shown in Figure 1 in conjunction with Figure 6. While this aspect is illustrated in the primary beam path, similar considerations apply to elements of the detection unit 200 as well. Figures 1 and 6 show a typical subset of elements in the primary beam path of a charged particle microscope, including a charged particle source 301, first and second collimator lenses 303.1 and 303.2, first and second active multi-aperture plate arrangements 306.1 and 306.2 (only one shown), first field lens 308, second field lens 307, third field lens 103.1 and fourth field lens 103.2, beam steering multi-aperture plate 390, first and second objective lenses 102.1 and 102.2 (only one shown), as well as a sample voltage source 503 and a stage 500. The control unit 800 is configured, in use, to provide at least one control signal (e.g., a voltage, a current, or both) to all of the above elements. The multi-aperture configuration is provided with multiple voltages (e.g., individual voltages for at least a plurality of primary charged particle beamlets). For a multi-beamlet charged particle microscope system including 100 primary charged particle beamlets, approximately 50 different drive signals are applied to the global elements and approximately 200-800 different voltages are applied to each multi-aperture configuration in use, with the number of individual voltages or currents exceeding approximately 10 times the number of primary charged particle beamlets. Prior to operation of the multi-beamlet charged particle microscope system according to an embodiment of the present invention, a set of image qualities is defined according to the wafer inspection task specifications. Some of the specifications are as described above. The set of image qualities constitutes an image quality vector, and the deviation of the image quality corresponds to the amplitude of the error vector. For convenience, a set of normalized error vectors is defined by normalizing the set of error vectors. The sensitivity, ie, the amount of change in the set of image qualities for a change in the drive signal applied to each of the set of elements in the primary beam path, is determined, for example, by simulation or calibration measurements.For example, in a calibration measurement, a representative set of sensor data is measured by a set of sensors or detectors, generating a sensor data vector for each sensitivity. A sensitivity matrix of the sensitivities of the elements in the primary beam path is then formed. The sensitivity matrix constitutes a linear perturbation model of the multi-beam charged particle microscope for a set of image qualities suitable for the wafer inspection task and is typically not an orthogonal matrix. The sensitivity matrix is ​​analyzed, for example, by singular value decomposition or a similar algorithm, and for each image quality, at least one set of elementary control signals is selected as control signals for a compensator that compensates for the deviations or aberrations of the image quality, thereby reducing the amplitude of the corresponding error vector. In one example, the sensitivity matrix is ​​decomposed by dividing it into two, three, or more kernels or subsets of independent sensitivity kernels corresponding to specific subsets of the set of image qualities. This reduces computational complexity and suppresses nonlinear or higher-order effects.

[0110] In one example, at least the kernel of the sensitivity matrix is ​​determined by the temperature of the multi-beam charged particle microscope. For example, temperature changes in the column or column elements of the multi-beam charged particle microscope cause focus drift, magnification drift, and stigma drift. The detector installed in the multi-beam charged particle microscope includes a temperature sensor (e.g., a temperature sensor in the cooling water or a temperature sensor attached inside a mechanical component, a multi-aperture plate, or a magnetic element). This allows each kernel of the sensitivity matrix to be orthogonalized with multiple representative temperatures, and the temperature-corrected sensitivity matrix can be used to calculate corresponding drive signals for the compensator in response to the temperature signal. Taking the actual temperature into account and applying the temperature-corrected sensitivity matrix and corresponding drive signals is particularly suitable for the iterative calibration step of the multi-beam charged particle microscope system, as described below. In a simplified example, multiple temperature sensors are eliminated, and the expected temperature is predicted from the operating history of the multi-beam charged particle microscope system.

[0111] In one example, a first set of basic drive signals is selected for the fast compensators (e.g., including electrostatic compensators and deflectors such as fast compensator 332 of the multi-beamlet generator, deflection system 110, and fast compensator 132 of the object illumination unit 100), and a second set of basic drive signals is selected for the slow-acting compensators (e.g., including magnetic elements such as the slow compensators of the object illumination unit 130 of FIG. 6). In one example, the number of basic drive signals in each set is minimized to a minimum number of drive signals so as to reduce the number of control operators for individual elements, shorten computation time, and enable control of the set of image quality within the required specifications of the wafer inspection task.

[0112] Each set of fundamental drive signals is stored in a memory of the control unit 800 (e.g., a memory of the primary beam path control module 830). The control arithmetic processor 840 derives a set of control signals from a set of amplitudes of the error vector. The primary beam path control module 830 derives a set of drive signals from the set of fundamental drive signals, for example, by multiplication with the set of control signals calculated by the control arithmetic processor 840. The secondary beam path control module 820 derives a set of drive signals from the set of fundamental drive signals, for example, by multiplication with the set of control signals calculated by the control arithmetic processor 840.

[0113] Thus, a method for preparing a multi-beam charged particle microscope configured for wafer inspection includes defining a set of image qualities and a set of normalized error vectors representing deviations from the set of image qualities in conjunction with a sensor data vector. A set of thresholds for the amplitudes of the set of normalized error vectors is determined according to the imaging specifications of the wafer inspection task as described above, and a pre-selection of a set of compensators for the multi-beam charged particle microscope is performed. The set of compensators includes a first deflection unit of the multi-beam charged particle microscope for scanning and deflecting a plurality of primary charged particles and a second deflection unit for scanning and deflecting a plurality of secondary electrons generated when using the multi-beam charged particle microscope. The method for preparing a multi-beam charged particle microscope for wafer inspection further includes determining a sensitivity matrix according to a linear and / or nonlinear perturbation model by varying at least the drive signals of each of the set of compensators. The sensitivity matrix is ​​analyzed, for example, by singular value decomposition or a similar algorithm. In one example, the sensitivity matrix is ​​decomposed by dividing it into two, three, or more independent subsets of kernels or image qualities, thereby reducing computational complexity and reducing nonlinear or higher-order effects. The method for preparing the operation of a multi-beam charged particle microscope for wafer inspection further includes deriving a set of normalized drive signals for compensating for each of the set of normalized error vectors. The normalized error vectors, the normalized drive signals, and the set of thresholds are stored in a memory of a control unit of the multi-beam charged particle microscope to constitute predetermined error vectors and predetermined drive signals.

[0114] The method for operating a multi-beam charged particle microscope during use (e.g., during wafer inspection) further includes receiving a plurality of sensor data constituting a sensor data vector from a plurality of sensors of the multi-beam charged particle microscope. In one example, the plurality of sensor data includes at least one of position and velocity information related to an actual position and an actual velocity of a wafer stage that holds the wafer during inspection by the multi-beam charged particle microscope. The set of sensors generating the plurality of sensor data are prepared and configured to enable unambiguous derivation of a predetermined error vector, and during use, a set of actual amplitudes of normalized error vectors are derived from the sensor data vector to represent a set of actual image quality conditions of the multi-beam charged particle microscope. A set of control signals is derived from the set of actual amplitudes, e.g., by multiplication with control signals, and a set of actual drive signals is derived from the predetermined normalized drive signals. The control unit controls compensators of the multi-beam charged particle microscope and provides the set of actual drive signals to the set of compensators so that the set of actual amplitudes is kept below a set of thresholds to maintain operation of the wafer inspection task sufficiently within imaging specifications. A method of operation according to one embodiment of the present invention is explained in more detail in Figure 7. For illustrative purposes, the same reference numerals as in Figures 1-6 are used. For wafer inspection, the multi-beam charged particle microscope (1) comprises a plurality of detectors including an image sensor (207) and a stage position sensor (520), and a set of compensators including at least first and second deflection systems (110, 222). The memory of the control unit 800 of the multi-beam charged particle microscope (1) stores the amplitude of the error vector and at least one set of threshold values ​​for the normalized drive signal.

[0115] In the first step SR, a wafer inspection task is registered, for example, by an operator or a command is given by an external operating system. The loaded wafer is aligned and positioned in a predetermined global wafer coordinate system of the multi-beam charged particle microscope system 1. The wafer inspection task includes a series of inspection sites (e.g., 33 and 35 in FIG. 2). From the series of inspection sites, a series of image acquisition tasks for multiple wafer areas in at least first and second inspection sites 33 and 35 are generated. At least one inspection site may include at least first and second image patches 17.1 and 17.2. In the multi-beam charged particle microscope 1, multiple primary charged particle beamlets 3 arranged in a raster array are used to image each image patch having a lateral dimension PX, and each of the multiple primary charged particle beamlets 3 is scanned over each of the subfields 31 having a lateral dimension SX. The image patch 17 is formed by stitching together the multiple subfields scanned by the multiple primary charged particle beamlets 3. The lateral dimension SX of a subfield is typically 10 μm or less, and the image dimension PX of one image patch 17 is typically approximately 100 μm or more. The number of primary beamlets 3 is typically 10×10 beamlets or more (such as 300 or 1000 beamlets). Suitable raster configurations are, for example, a hexagonal raster, a rectangular raster, or a circular raster with at least a circular arrangement of the beamlets, although other raster configurations are possible as well.

[0116] First and second patch center locations 21.1 and 21.2 of image patches 17.1 and 17.2 are calculated from an inspection task list that includes the location of the inspection site 33 on the wafer surface and the area of ​​the inspection task. If the lateral dimension of the inspection site area exceeds the image patch, the inspection site area is divided into at least two image patches 17.1 and 17.2 having at least first and second patch center locations 21.1 and 21.2. The first and second patch center locations 21.1 and 21.2 are transformed into wafer coordinates to define first and second local wafer coordinate systems relative to the global wafer coordinate system. This generates a list of multiple local wafer coordinate systems for the acquisition of corresponding image patches 17.1 and 17.2.

[0117] Whenever a plurality of elements is described in terms of first and second elements, it is understood that the plurality of elements may include more than two elements. For example, it is understood that an inspection task may include a plurality of 50, 100, or more inspection sites, and each inspection site may include a plurality of 2, 4, or more image patches.

[0118] In step S1, the status of the multi-beamlet charged particle microscope system 1 is determined, for example, by the operation history or initialization of the multi-beamlet charged particle microscope system 1. The initialization of the multi-beamlet charged particle microscope system 1 may include calibrating the system when a corresponding trigger signal is provided. A selected beamlet from the plurality of charged particle beamlets 3 can be used for the system calibration. At least one reference sample mounted on a dedicated holder on the wafer stage 500 or the second metrology stage can be used for the system calibration and for determining the line of sight 53 of the charged particle microscope and the position of the wafer stage. Two or more reference samples at different positions can be used for the calibration of different image performance functions, such as magnification, distortion, or astigmatism.

[0119] Step S0 involves positioning and aligning the wafer surface (25) of the wafer relative to a location in a local wafer coordinate system (551) that includes the line of sight of the multi-beam charged particle microscope (1). The wafer is positioned and aligned with the next local wafer coordinate system 551 directly under the optical axis or line of sight 53 of the multi-beamlet charged particle microscope 1. The next local wafer coordinate system 551 can be the first or any subsequent local wafer coordinate system 551 from the list of local wafer coordinate systems generated in step SR.

[0120] The wafer stage 500 is triggered by wafer movement to align the local wafer coordinate system 551 with the line of sight 53 of the charged particle microscope. Alignment of each local wafer coordinate system by wafer stage movement is optionally performed using a pattern formed or visualized on the wafer surface. Adjustment stops when the difference vector between the image coordinate system 51, which includes the line of sight 53 as the z-axis of the multi-beam charged particle microscope 1, and the local wafer coordinate system 551 falls below a threshold. The difference vector 55 is a vector containing six degrees of freedom for stage movement, such as displacement and rotation or tilt. For fine adjustment, the difference vector can be 50 nm or less in the lateral direction and less than 100 nm in the line of sight or focus direction. The threshold for image rotation in the z-axis is typically 0.5 mrad, and the threshold for tilt relative to the xy-plane image coordinate system is typically 1 mrad. Fine adjustment can include multiple iterations of at least the imaging step of the inspection site and stage movement.

[0121] In one embodiment of the present invention, a high-speed operating mode for the wafer inspection task is selected, and the requirement for precision adjustment is relaxed by increasing the threshold by a factor of 2, 10, or more, and residual difference vectors exceeding the threshold are compensated by a set of compensators of the multi-beam charged particle microscope 1. To this end, a plurality of offset error vector amplitudes are generated by the control unit 800 and provided to step S2 below.

[0122] In step S1, an image acquisition is performed to acquire a digital image of a first image patch (17.1) of the wafer surface (25), collect multiple sensor data from multiple detectors, and perform a series of inspection tasks for the inspection site. Step S1 includes at least: In step S1-1, the process of image acquisition by scanning imaging of a first image patch 17.1 begins. Each image patch 17 is preferably imaged with the stage 500 aligned with the line of sight 53 of the multi-beam charged particle microscope 1 according to step S0, or with the stage moving at a slow speed as described above. In step S1-2, in parallel with step S1-1, multiple sensor data are generated by multiple detectors. The multiple detectors include at least stage position sensor 520 and image sensor 207. In one example, the multiple detectors also include other detectors of the multi-beamlet charged particle microscope, such as sensors 238 and 138, that generate sensor data during image acquisition. The multiple sensor data may also include currents or voltages applied to electrostatic and magnetic elements. Another example of a sensor that provides sensor data is a temperature sensor (e.g., a temperature sensor that monitors the temperature in a cooling fluid or a magnetic element).

[0123] The imaging of step S1-1 involves collecting a plurality of secondary electrons generated at locations where the plurality of primary charged particle beamlets 3 intersect the wafer surface 25. A plurality of secondary electron beamlets 9 are formed from the secondary electrons. Each of the plurality of secondary charged particle beamlets 9 is separately detected to obtain a digital image of the image patch 17 in each local wafer coordinate system 551.

[0124] In step S2, the plurality of sensor data are evaluated in the sensor data analysis system 818. Step S2 includes at least the following. In S2-1, real time T a Given a sensor data vector DV(i) of length L in kFor the expanded sensor data vector DV(i) in (i), a plurality of sensor data from different sensors are combined. As described above, a set of predetermined error vectors E k (i) represents the deviation of a set of image quality parameters for image acquisition. A set of predetermined error vectors E k The set of K error amplitudes A k of (i) are calculated as follows. DV(i)=Σ k A k ·E k (i)+ε The residual error vector ε is below a predetermined threshold. In the calculation of the error amplitude A k , the plurality of offset error vector amplitudes or the predicted temporal behavior of the offset error vector amplitudes generated in step S0 are considered. In one example, the size of K is 6, representing the six degrees of freedom difference between the local wafer coordinate system and the line of sight. Generally, K is larger than 6. For example, a set of error vectors may include K = 14 error vectors including the six degrees of freedom difference between the wafer stage position, the local wafer coordinate system and the line of sight, magnification change, anamorphic distortion change, spherical aberration, field curvature, third-order distortion, and chromatic aberration. Generally, K is smaller than L (i.e., K < L). Although the number L of sensor data constituting the sensor data vector DV(i) can be 10 or more, it is preferable to use a small value to increase the calculation speed (e.g., K < L < 4K). For example, when K = 14, L is preferably less than 50. By reducing the plurality of sensor data and using the error vector of K error amplitudes with the reduced length L, the data amount is suppressed and the calculation speed is improved. In S2-2, according to the predicted transition of the multi-beam charged particle microscope in the prediction time interval, a subset of the transition amplitudes of at least a subset of the set of actual amplitudes is predicted. The temporal transition of at least a subset of the n error amplitudes A n is derived, and the subset of error amplitudes is considered as a time-dependent function A n (t). An example of the derivation of the temporal transition of the error amplitude A a in the prediction time interval after the real time T n (t > T a ) is the error amplitude An (t <T a ) by either linear or higher order extrapolation as described above. Thus, in use, at least a subset of the set of actual amplitudes of the multi-beam charged particle microscope is recorded to generate a history of the subset of the set of actual amplitudes. n (t>T a As another example of the derivation of the time course of ), given a set of model functions M n Error amplitude A for (t) n There is an approximation of the history of (t). In S2-3, the error amplitude A n The time evolution of at least a subset of (t) has amplitude A n The drift part S of the time transition of (t) n (t) and dynamic change N n (t) is separated into (A n (t)=S n (t)+N n (t)).

[0125] In one example, the predetermined error vector includes image aberrations introduced into the primary charged particle beam path, such as magnification error, anamorphic distortions such as keystone, or distortions such as third order or higher distortions. Other aberrations are, for example, field curvature, astigmatism, or chromatic aberrations.

[0126] In one example, the predetermined error vector includes image aberrations introduced in the secondary electron beam path 11. Aberrations in the secondary electron beam path 11, for example, reduce the collection efficiency of secondary electrons, for example, reducing image contrast and increasing noise.

[0127] In one example, combining multiple sensor data into a sensor data vector DV(i) includes calculating the difference between the multiple sensor data (e.g., the difference between the line-of-sight position coordinates of a multi-beam charged particle microscope and the position and orientation data of a stage position sensor).

[0128] In one example, the particular signature of the selected error amplitude is the error amplitude A nThe error amplitude A is subtracted from the time course of (t). n A filter is applied to at least one of the time series of (t), which results in an error amplitude A that does not affect the image quality. n A specific signature of the time course of (t) is subtracted, reducing the amount of control calculations.

[0129] In one example, the set of error vectors is derived from the capabilities of a set of compensators available in the multi-beam charged particle microscope, such that the set of error vectors can be compensated for by controlling the operation of the set of compensators. In one example, a set of possible error vectors is derived from an imaging experiment, and the multi-beamlet charged particle microscope is provided with a set of compensators capable of compensating for the set of error vectors.

[0130] In one example, the amplitude or amplitude trends of a set of error vectors are compared against a predetermined set of thresholds stored in the memory of the control unit 800 .

[0131] In a wafer inspection task, a set of predicted control signals is determined from the amplitude progression, a set of predicted drive signals is determined from the set of predicted control signals, and the set of predicted drive signals is supplied to a set of compensators in a time series manner so that a subset of actual amplitudes falls below respective thresholds during a prediction time interval.

[0132] In step S3, a set of error amplitudes A k From the set of P control signals C p The error function E n’ The error amplitude A n The error amplitude A including the time transition of k The deviations and amplitudes of are analyzed by the control processor 840 using a predetermined mapping function MF to generate a set of P control signals C p is mapped to MF:A k →C p

[0133] The mapping of the set of K error amplitudes to the set of P control signals by a predetermined mapping function MF is achieved by a numerical fitting operation, such as a look-up table, a matrix inversion, or a singular value decomposition.

[0134] In one example, different groups of error vectors are processed in parallel in different error vector categories. For example, coordinate system drift in two coordinate systems defined by the line of sight and the local wafer coordinate system is processed separately in one coordinate error category. Higher-order imaging aberrations or decentering aberrations are processed in their respective error vector categories. This allows a set of control signals for a set of error vectors to be calculated in parallel at high speed.

[0135] In step S3-1, a set of control signals C p A set of deflection control signals is derived from

[0136] In step S3-2, a set of control signals C p A set of primary control signals is derived from the primary beam path, which are selected to compensate for imaging aberrations such as defocus, field tilt, field curvature, magnification, astigmatism, chromatic aberration, decentration, or other higher order aberrations by controlling compensators in the primary beam path.

[0137] In step S3-3, a set of control signals C p A set of secondary control signals is derived from

[0138] In step S3-4, a set of control signals C p The set of image processing control signals is derived from the image stitching components IS p Contains a subset of .

[0139] In optional step S3-5 (not shown), a set of control signals C p A set of stage control signals is derived from

[0140] In step S4, a set of control signals C is sent to at least one of a set of control modules including a projection system control module 820, a primary beam path control module 830, a deflection control module 860, a stage control module 880, and an image stitching unit 812. p The control modules each provide a set of control signals C that are provided to at least one compensator in the set of compensators to generate a set of actuation values ​​or drive signals (e.g., a series of voltages or currents) that compensate for the imaging aberration represented by an error vector in the set of error vectors. p In step S1 of image acquisition, a first set of control signals is provided to a set of compensators, and a second subset of control signals is stored in memory and provided to step S0 for application in subsequent positioning and alignment of the local wafer coordinate system.

[0141] In step S4-1, a set of deflection control signals is provided to the deflection control module 860. To compensate for a set of error vectors representing image aberrations, the lateral positions of the focal points 5 of the plurality of primary charged particle beamlets 3 are first corrected so that the focal points are formed on the wafer surface 25 at predetermined lateral positions defined by the local wafer coordinate system 551 and a predetermined raster configuration with a lateral position accuracy of 10 nm or better. The lateral alignment of the focal points 5 at the predetermined positions is controlled by the first deflection unit 110, which deflects the plurality of primary charged particle beamlets 3. This allows, for example, the control signal C p to the first and second deflection units (110, 222) to compensate for changes in the position or orientation of the sample stage (500).

[0142] Therefore, the control signal C of the set of deflection control signals pAn example of this is a primary offset signal supplied to a deflection control module 860. The deflection control module 860 derives a first offset signal for a first deflection unit 110 equipped with an electrostatic deflection scanner, and the multiple primary charged particle beamlets 3 are scanned on the wafer surface 25 by a scan path 27 at an offset position. This compensates for a lateral displacement vector 55 between the local wafer coordinate system 551 and the line of sight 53 of the multi-beam charged particle microscope 1, thereby realizing correction of the line of sight 53 so that it deviates from the local wafer coordinate system 551 by less than a predetermined threshold (e.g., less than 10 nm, 5 nm, 2 nm, or 1 nm).

[0143] Furthermore, by providing a second offset signal to the second deflection unit 222, the focal points 15 of the secondary electron beamlets 9 are kept at a constant position on the image detector 207, thereby achieving high image contrast and image fidelity. To keep the positions of the focal points 15 of the multiple secondary electron beamlets 9 at a constant position on the image sensor 207, the multiple secondary electron beamlets 9 pass through the first deflection unit 110 and the second deflection unit 222. After changing the scan paths 27 of the multiple primary charged particle beamlets 5 on the wafer surface 25 by the offset position, a second offset signal of the set of deflection control signals is provided to the second independent deflection unit 222, and the offset position of the focal points 5 on the wafer surface 5 is compensated by the second deflection unit 222 so that the focal points 15 of the multiple secondary electron beamlets 9 are kept constant on the image sensor 207.

[0144] The offset position may vary over time, and the offset control signal may be altered during an image scan of the image patch 17, for example to compensate for lateral drift or jitter of the sample stage 500.

[0145] In step S4-2, a set of primary control signals is provided to the primary beam path control module 830. To compensate for a set of error vectors representing image aberrations, the longitudinal positions of the focal points 5 of the multiple primary charged particle beamlets 3 are corrected so that the focal points are formed on the wafer surface 25 with an accuracy less than the depth of field of the multi-beam charged particle microscope. Multi-beam scanning electron microscopes typically have a depth of field of approximately 10 nm to 100 nm, and a specification for the maximum focal spot deviation from the image plane of less than 10 nm, preferably less than 5 nm. The image aberrations of the focal points 5 of the multiple primary charged particle beamlets 3 include defocus, field tilt, and field curvature.

[0146] For example, when a primary control signal for correcting the image plane tilt is supplied to the primary beam path control module 830, the primary beam path control module 830 derives a set of focus correction voltages for the active multi-aperture plate arrangement 306 (e.g., a multi-aperture lens array), which changes the focus positions of the individual primary charged particle beamlets 3 individually, thereby achieving correction of the image plane tilt, for example to compensate for the tilt of the sample stage 500 relative to the image coordinate system 51.

[0147] In another example, when a primary control signal for correcting the defocus is provided to the primary beam path control module 830, the primary beam path control module 830 derives a voltage change on the field lens 306 that changes the image plane position in the z direction as a whole, thereby compensating for a movement of the sample stage 500 in the z direction, which is the propagation direction of the primary beamlets 3, for example, by changing the focal positions of the multiple primary charged particle beamlets 3.

[0148] In another example, when a primary control signal correcting for the rotation between the image coordinate system 51 and the local wafer coordinate system 551 is supplied to the primary beam path control module 830, the primary beam path control module 830 derives a set of deflection voltages for the active multi-aperture plate arrangement 306 (e.g., a multi-aperture deflector array) to individually deflect individual primary charged particle beamlets 3, thereby achieving a rotation of the image coordinate system 51, for example to compensate for a rotation of the sample stage 500 relative to the image coordinate system 51.

[0149] Other control signals are also provided as appropriate to correct other image aberrations in the primary beam path, including magnification change, astigmatism, chromatic aberration, etc. The set of primary control signals includes control signals for controlling compensators in the primary beam path, including the compensators of the charged particle multi-beamlet generator 300 and the object illumination unit 100.

[0150] In step S4-3, a set of secondary control signals is provided to the projection system control module 820. The imaging aberrations of the secondary beam path or detection unit are corrected to compensate for the set of error vectors representing the image aberrations.

[0151] For example, when a secondary control signal correcting the image plane tilt is provided to the projection system control module 820, the projection system control module 820 derives a set of focus correction voltages for the multi-aperture corrector 220 (e.g., a multi-aperture lens array) to individually change the focus positions of the individual secondary electron beamlets 9, thereby achieving correction of the image plane tilt, e.g., to compensate for tilt of the sample stage 500 relative to the image coordinate system 51, while maintaining imaging of the secondary electron beamlets 9 from the tilted wafer surface 25 onto the image detector 207.

[0152] In another example, when a secondary control signal that corrects the defocus is provided to the projection system control module 820, the projection system control module 820 derives a voltage change on the electrostatic lens 206 that changes the image plane position as a whole, thereby changing the focal position of the multiple secondary electron beamlets 9, for example to compensate for movement of the sample stage 500 in the z-direction, which is the direction of propagation of the primary beamlets, while maintaining imaging of the secondary electron beamlets 9 from the defocused wafer surface 25 onto the image detector 207.

[0153] In another example, when a secondary control signal that corrects for the rotation between the image coordinate system 51 and the local wafer coordinate system 551 is supplied to the projection system control module 820, the projection system control module 820 derives a set of deflection voltages for the multi-aperture corrector 220 (e.g., a multi-aperture deflector array) that individually deflects each secondary electron beamlet 9 to compensate for the rotation of the image coordinate system 51 for imaging the multiple secondary electron beamlets 9 at predetermined fixed positions on the image detector.

[0154] The above example shows the compensation of image aberrations in both the primary beam path 13 and the secondary electron beam 13. Some primary control signals for correcting image aberrations in the primary beam path, such as astigmatism or field curvature, are supplied to the primary beam path control module 830, and image aberrations are compensated only in the primary beam path 13. Some secondary control signals for correcting image aberrations in the secondary beam path, such as astigmatism or field curvature, are supplied to the projection system control module 820 as appropriate, and image aberrations are compensated only in the secondary beam path 11.

[0155] In step S4-4, a set of image processing control signals is provided to image stitching unit 812. The set of image processing control signals IS p are applied directly or stored along with the image data stream for application in image processing and stitching operations performed by image stitching unit 812.

[0156] In optional step S4-5 (not shown), a set of stage control signals is provided to the stage control module 880. In one example, in step S2, slow drift of the sample stage 500 is detected and compensated for by the stage control signals.

[0157] In one example, a set of control signals C p At least a subset of the drift part S n (t) and dynamic change N n Error amplitude A at (t) n According to the separation of the time course of at least a subset of (t), the drift control component CS p A subset of the dynamic control component CN p The drift control component CS p A subset of CS is fed to a set of compensators or active elements of a multi-beam charged particle microscope system. In one example, a drift control component CS p is supplied to a subset of slowly varying active elements, including magnetic elements, which change their state upon activation. In one example, a drift control component CS is applied to a fast varying active element, such as an electrostatic deflector, an electrostatic multipole corrector, or an electrostatic multi-aperture element. p In one example, both subsets of compensators are provided with a drift control component CS p Dynamic control component CN p The subset of the charged particle beamlets is supplied to a fast-varying active component of the multi-beamlet charged particle microscope system, and the fast-varying active component is driven to vary its effect on the charged particle beamlets. The fast-varying active component is an electrostatic element such as an electrostatic deflector, an electrostatic multipole corrector, an electrostatic lens, or an electrostatic multi-aperture element.

[0158] In general, the number of control signals P is proportional to the error amplitude A k The control signal C can exceed the number K (P≧K). p Each of these can vary over time. pis varied at least in part during the image scan of the image patch 17, thereby compensating for lateral drift of the sample stage 500, for example, during image acquisition of an image patch. In one example, at least one control signal is a time-dependent function representing the predicted evolution of the error amplitude over subsequent time intervals, thereby realizing a continuous control operation for compensation of the predicted imaging deviation.

[0159] In step S5, the drift control component CS p and the dynamic control component CN p A set of control signals C containing a subset of p By monitoring and accumulating the data, the history of changes to the multi-beam charged particle microscope is recorded.

[0160] In step S6, the actual system status of the multi-beam charged particle microscope is estimated based on the history of changes.

[0161] In optional step S7, a time-varying model function M n (t) is adapted to the history of changes and the actual system status of the multi-beam charged particle microscope is given in step S2.

[0162] In step S8, the actual system status of the multi-beam charged particle microscope is analyzed, and a transition of the system status is predicted during a subsequent image scan. If the prediction of the system status indicates that the transition of the error vector will reach an uncompensable value, for example, because the range of the actuators for compensation may be reached during the subsequent image scan, recalibration and resetting of the actuators of the multi-beam charged particle microscope is triggered before the subsequent image scan. In this case, a trigger signal is provided to step S1. If the prediction of the system status indicates that the next imaging task is possible, the method continues with steps S0 to S7 for image acquisition of a subsequent image patch in the next local wafer coordinate system from the list of inspection tasks.

[0163] In one example, in step S8, a drift component of the control signal is calculated, for example, by predicting a transition of the error amplitude, and provided to step S0. In step S0, the drift component is compensated for by operating a compensator while the stage moves from a first image patch to a subsequent second image patch or a subsequent inspection site. Accordingly, in step S0, the control unit 800 provides a control signal to the stage control module 880 to move the stage 500 from the first local wafer coordinate system to the subsequent local wafer coordinate system, and further provides the drift component of the control signal to at least one of the control modules including the primary beam path control module 830, the projection system control module 820, or the deflection control module 860.

[0164] As is apparent from the above description, steps S1-S7 of the method of operation run in parallel, are performed in real time during image acquisition of an image patch, and interact with each other. Those skilled in the art will recognize that variations and modifications of the above-described method are possible.

[0165] In one embodiment of the present invention, a change in the object plane 101 or focal position of the multiple primary charged particle beamlets is enabled while maintaining the specification requirements of the wafer inspection task. The reason for the change in the object plane 101 may be a predetermined change in the imaging settings for image acquisition (e.g., a change in magnification or numerical aperture, a change in desired resolution, or drift of elements arranged in the primary beam path 13 or the secondary beam path 11), as monitored in step S7 or step S8. The change in the focal plane by the magnetic objective lens 102 of the multi-beam charged particle microscope 1 has a rotational effect on the multiple primary charged particle beamlets 3. The change in the focal plane or object plane 101 causes a rotation of the raster arrangement of the multiple primary charged particle beamlets relative to the optical axis 105 of the multi-beam charged particle microscope 1, as shown in FIG. 3, resulting in a rotation of the image coordinate system 51 relative to the local wafer coordinate system 551. The semiconductor structures arranged on the wafer surface 25 are typically arranged orthogonally to each other. Rotation of the image coordinate system or scan path 27 of the primary charged particle beamlet 3 relative to the semiconductor structure arrangement as shown in FIG. 3b makes at least some of the specification requirements for high-throughput wafer inspection tasks impossible to achieve. Other image performance parameters, such as the decentering of the multiple primary charged particle beamlets or the magnification or pitch of the multiple primary charged particle beamlets, are also changed. The change in the image performance parameters increases when multiple secondary electrons emitted at changed positions of the multiple primary charged particle beam spots are collected by the adjustment objective lens 102. In this embodiment, unwanted changes in the image performance parameters induced by the change in the image plane or focal plane are compensated for by the control unit 800. The control unit 800 is configured to predict a control signal that compensates for the error amplitude induced by the change in the image plane position or focal position from the first image plane position to the second image plane position. The control unit 800 is configured to provide control signals to compensators and a wafer stage in the primary beam path 13 and the secondary beam path 11. The compensator of the primary beam path may include, for example, a second objective lens (not shown in FIG. 1) arranged near the intermediate image plane 321, a field lens 103.1 or 103.2, a multi-aperture deflector array 306.3, or a multi-aperture deflector array 390.The compensator of the secondary beam path may include, for example, a magnetic lens, an astigmatism corrector, or a multi-aperture array element. After a change in the image plane or focal position from the first position to the second position is triggered, the control unit 800 controls a combination of multiple elements including the compensators or wafer stages of the primary and secondary beam paths.

[0166] In one example, the control unit 800 or the multi-beam charged particle microscope 1 is configured to derive, from the plurality of sensor data, an error vector representing a deviation of the orientation of the semiconductor structure from the image coordinate system 51 or the scan path 27 or the direction of the plurality of primary charged particle beamlets, and is further configured to derive and provide a set of control signals to the control module. The control module is configured to cause at least one of a rotation of the plurality of primary charged particle beamlets, a rotation of the plurality of secondary electron beamlets, and a rotation of the sample stage 500. For example, the control unit 800 is configured to cause a rotation by a slow-acting compensator including the wafer stage 500 or the objective lens 102 by providing the control signal in step 0 of the above-described method. For example, the control unit 800 is further configured to rotate a fast-acting compensator, such as an electrostatic deflector array, disposed in the primary charged particle beam path, the secondary electron beam path, or both, by providing the control signal.

[0167] In one embodiment, a method for operating a multi-beam charged particle microscope configured for wafer inspection includes: (a) loading into memory a set of predetermined normalized error vectors representing deviations from a set of image qualities; (b) loading into memory a set of predetermined thresholds for amplitudes of the set of predetermined normalized error vectors; and (c) loading into memory a set of predetermined normalized drive signals for compensating for each of the set of normalized error vectors. While a wafer inspection task is being performed, the method for operating the multi-beam charged particle microscope also includes: (d) receiving a plurality of sensor data from a plurality of sensors of the multi-beam charged particle microscope, the plurality of sensor data comprising a sensor data vector. In one example, the plurality of sensor data includes at least one of position or velocity information regarding an actual position and an actual velocity of a wafer stage that holds the wafer during inspection by the multi-beam charged particle microscope. The method for operating a multi-beam charged particle microscope while a wafer inspection task is being performed further includes the steps of: e) determining from the sensor data vector a set of actual amplitudes of a predetermined normalized error vector representing a set of actual states of image quality of the multi-beam charged particle microscope; f) deriving a set of control signals from the set of actual amplitudes and a set of actual drive signals from the set of predetermined normalized drive signals during wafer inspection; and g) providing the set of actual drive signals to a set of compensators during wafer inspection to cause a subset of the actual amplitudes to be less than the subset of thresholds determined in step b) during operation of the multi-beam charged particle microscope. In one example, the method for operating a multi-beam charged particle microscope further includes the step of h) predicting a subset of transition amplitudes of at least a subset of the set of actual amplitudes during wafer inspection according to an expected transition of the multi-beam charged particle microscope over a prediction time interval. In one example, the predicted progression of the multi-beam charged particle microscope over a prediction time interval is determined according to one of a predictive model function or a linear, quadratic, or higher order extrapolation of a set of actual amplitude histories.The method for operating a multi-beam charged particle microscope further includes the steps of: i) deriving a set of predicted control signals from the set of transition amplitudes during wafer inspection, and deriving a set of predicted drive signals from the set of predicted control signals; j) supplying the set of predicted drive signals to a set of compensators in a time-series manner during wafer inspection, thereby causing a subset of actual amplitudes to be less than a subset of threshold values ​​during operation of the multi-beam charged particle microscope during a prediction time interval; and k) recording at least a subset of the set of actual amplitudes of the multi-beam charged particle microscope during wafer inspection to generate a history of the subset of actual amplitudes.

[0168] The method for operating a multi-beam charged particle microscope includes, prior to operation, selecting a set of compensators for the multi-beam charged particle microscope. In one example, the set of compensators includes a first deflection unit of the multi-beam charged particle microscope for scanning and deflecting a plurality of primary charged particles and a second deflection unit for scanning and deflecting a plurality of secondary electrons generated during use of the multi-beam charged particle microscope. The method for operating a multi-beam charged particle microscope further includes, prior to operation, determining a set of predetermined normalized error vectors representing deviations from a set of image qualities, determining a sensitivity matrix according to a linear perturbation model by varying at least the drive signals of each of the set of compensators, and determining, from the sensitivity matrix, a set of predetermined normalized drive signals that compensate for each of the set of predetermined normalized error vectors.

[0169] It will be appreciated that the components of the multi-beam charged particle microscope described in conjunction with Figure 6 and the method steps described in conjunction with Figure 7 are simplified examples for illustrating the construction and operation of a multi-beam charged particle microscope for wafer inspection according to the present invention. At least some of the method steps or components may be combined. For example, the control and arithmetic processor 840 and the sensor data analysis system 818 may be combined together, or the primary beam path control module 820 may be incorporated into the control and arithmetic processor 840.

[0170] At least one of the compensators used in the above-described embodiments is a multi-beam active array element. Each individual primary charged particle beamlet among the plurality of primary charged particle beamlets is individually affected by an electrostatic microlens array, an electrostatic astigmatism corrector array, or an electrostatic deflector array in the primary charged particle beam path. As an example, such a multi-aperture array 601 is shown in FIG. 8. The multi-aperture array 601 comprises a plurality of apertures arranged in a raster configuration (in this example, a hexagonal raster configuration) of the plurality of primary charged particle beamlets. Two of the apertures are indicated by reference numerals 685.1 and 685.2. A plurality of electrodes 681.1 to 681.8 are arranged around each of the plurality of apertures. In this example, there are eight electrodes, but other numbers are possible, such as one, two, four, or more. The electrodes are electrically isolated from each other and from the carrier of the multi-aperture array 601. Each of the plurality of electrodes is connected to a control module by one of the conductive lines 607. By applying separate, predetermined voltages to each of the electrodes 681, different effects can be achieved for each of the multiple primary charged particle beamlets passing through each aperture 685. Because only electrostatic effects are used, the charged particle beamlets passing through the aperture 685 can be individually adjusted or modified at high speed and frequency. For example, such effects can include deflection, focal plane modification, and astigmatism correction of the primary charged particle beamlets. In one example, multiple (e.g., two or three) such multi-aperture plates are arranged in series. Similarly, each individual secondary electron beamlet among the multiple secondary electron beamlets can be individually influenced in a similar manner by an electrostatic microlens array, electrostatic astigmatism corrector array, or electrostatic deflector array in the secondary electron beam path.

[0171] Another embodiment of the present invention will now be described in more detail. A multi-beam charged particle microscope configured for wafer inspection and a method of operating such a microscope have been described with reference to FIG. 1 . As can be seen from the above description, during acquisition of a digital image of, for example, a first image patch 17.1, a plurality of primary charged particle beamlets 3 and a plurality of secondary electrons 9 are scanned and deflected together by a first deflection system 110 in a common beam path, and a plurality of secondary electrons 9 are further scanned and deflected by a second deflection system 222 in a secondary beam path 11 in a detection unit 200. This keeps the focal spots 15 of the plurality of secondary electron beamlets 9 on an image sensor 207 at a constant position during image scanning. The detection unit 200 includes an aperture 214 through which the plurality of secondary electron beamlets 9 are filtered. The aperture filter 214 therefore controls the topographic contrast of the secondary electron beamlets provided to the image sensor 207. Misalignment of the detection unit 200 (e.g., a shift in the center of the crossover 212 of the multiple secondary electron beamlets 9) or the second deflection system 222 changes the image contrast. According to this embodiment, unwanted changes in topography contrast are detected and compensated for. Therefore, the detection unit 200 includes a third deflection system 218, and the combined action of the first, second, and third deflection units 110, 222, and 218 keeps the center of the crossover 212 of the multiple secondary electron beamlets 9 aligned with the aperture stop position of the aperture stop 214, and keeps the position of the secondary charged particle image spot 15 constant on the image sensor 207. This enables wafer inspection according to the specifications of the wafer inspection task. A constant image contrast is maintained for each of the multiple secondary electron beamlets along the scan paths 27.11-27.MN of the image patch 17.1 and within different subfields 31.11-31.MN. Figure 1 shows the positions of the second and third deflection systems 222 and 218 in the projection system 205 of the detection unit 200 as an example, but other positions of the second and third deflection systems 222 and 218 in the projection system 205 are also possible to achieve a constant image contrast on the image sensor 207 as well as a constant position of the focal point 15 of the multiple secondary electron beamlets 9.For example, both the second and third deflection systems 222 and 218 can be arranged before the aperture filter 214. The control unit 800 is configured to derive, from the sensor data vector, the amplitude of an error vector representing contrast variations across the plurality of secondary electron beamlets 9, and is further configured to derive and supply a first control signal to a deflection control module 860. The deflection control module 860 is configured to derive a deflection system change drive signal for the deflection system, including the second and third deflection systems 222 and 218, arranged in the secondary beam path 11 of the detection unit 200. In one example, the control module 800 is further configured to derive and supply a second control signal to a projection system control module 820. The projection system control module 820 is configured to derive a second drive signal for controlling another high-speed compensator 232 of the projection system 205 (e.g., an electrostatic lens or an astigmatism corrector of the multi-array active element 220). This maintains image contrast sufficiently within performance specifications for high-throughput wafer inspection tasks. In another example, the first deflection system 110, which scans the deflection of the multiple primary charged particle beamlets 3, is preferably positioned close to the first beam crossover 108 of the multiple primary charged particle beamlets 3. However, misalignment of the primary beam path 13 may cause the position of the first beam crossover 108 to deviate from its designed position, introducing an off-center error for the multiple primary charged particle beamlets 3. The control unit 800 is configured to derive from the sensor data an amplitude of an error vector representing a deviation from off-center illumination of the wafer surface 25 by the multiple primary charged particle beamlets 3, and to derive a control signal from the deviation, for example, to provide a drive signal to the multi-aperture deflector 390 near the intermediate image plane 321. This maintains off-center illumination of the wafer surface by the multiple primary charged particle beamlets 3. Off-center illumination refers to illumination in which each of the multiple primary charged particle beamlets 3 impinges on the wafer surface 25 parallel and nearly perpendicular (eg, at an angular deviation from the surface normal of less than 25 milliradians).In an embodiment, the actual error amplitude derived from the sensor data vector represents at least one of the image performance specifications for the wafer inspection task, such as the relative position and orientation of the wafer stage with respect to the line of sight of the multi-beam charged particle microscope and the image coordinate system of the multi-beam charged particle microscope, the decentering condition, the contrast condition, the absolute position accuracy of the multiple charged particle beamlets, the magnification or pitch of the multi-beam charged particle microscope, or the numerical aperture of the primary charged particle beamlet of the multi-beam charged particle microscope. During the image scan, other deviations from the image performance specifications for the wafer inspection task, such as distortion of the multiple charged particle beamlets, astigmatism, and higher-order aberrations such as chromatic aberration, can also be monitored and compensated for. For example, the amplitude of the error vector representing astigmatism can be derived from the data fragments of the image sensor and compensated for by an electrostatic compensator. For example, the amplitude of the error vector representing chromatic aberration of the primary charged particle beamlets can be compensated for by the additional magnetic lens 420 and voltage supply unit 503 of the beam splitter unit 400. The multi-beam charged particle microscope according to the above-described embodiments or examples allows for rapid scanning of wafer surfaces, providing high-throughput inspection of integrated semiconductor features with a critical dimension resolution of at least a few nanometers (e.g., less than 2 nm) during development or manufacturing of semiconductor devices or in reverse engineering.

[0172] Multiple primary charged particle beamlets are scanned in parallel over the surface of the wafer, generating secondary charged particles to form digital images of image patches, e.g., 100 μm to 1000 μm in diameter. After acquiring a first digital image of a first image patch, the substrate or wafer stage moves to a second image patch, and a second digital image of the second image patch is acquired by rescanning the multiple primary charged particle beamlets. During operation and each image acquisition, multiple detectors, including an image sensor and a stage position sensor, generate multiple sensor data and a set of control signals. The control signals are supplied to a control module that controls the operation of active elements, such as a deflection unit, electrostatic lens, magnetic lens, astigmatism corrector, multi-aperture active array, or other compensator, that scan the multiple primary and secondary charged particle beamlets. For example, during the stage movement from the first image patch to the second image patch between the acquisition of the first and second digital images, at least a subset of imaging aberrations is compensated for by a slow compensator, e.g., a magnetic element. During image acquisition of the digital image of the first or second image patch, a subset of control signals is provided to a control module including a deflection unit. This compensates for, for example, wafer stage position errors or drift relative to the line of sight of a multi-beam charged particle microscope during the image scan. Other aberrations or deviations from imaging performance specifications are determined and predicted from the sensor data, and respective control signals are generated in real time and provided to high-speed actuators. This allows for the stitching together of multiple image subfields or patches to form a high-resolution digital image with high image fidelity and precision, to within 5 nm, 2 nm, or even 1 nm. The stage is then rapidly moved between the first and second image patches or to the next location of interest (e.g., the next PCM or an adjacent image field), reducing the time required for, for example, repeated precision stage alignment.

[0173] As will be apparent from the above description, combinations and various modifications of the above examples and embodiments are possible and are equally applicable to the above embodiments or examples. The charged particles of the primary beam can be, for example, electrons, but can also be other charged particles, such as He ions. Secondary electrons include secondary electrons in the strict sense as well as any other secondary charged particles generated by the interaction of the primary charged particle beamlet with the sample, such as backscattered electrons or second secondary electrons generated by backscattered electrons. In another example, secondary ions can be collected instead of secondary electrons.

[0174] Some embodiments may be further described through the use of the following sets of clauses, although the invention is not intended to be limited to any of these sets of clauses.

[0175] First set of clauses Clause 1: A method of operating a high-throughput and high-resolution multi-beam charged particle microscope (1), comprising: a first image acquisition of a first image patch 17.1 in a first time interval Ts1; a second image acquisition of a second image patch 17.2 in a second time interval Ts2; a third time interval Tr during which the wafer stage (500) is moved from the first center position (21.1) of the first image patch (17.1) to the second center position (21.2) of the second image patch 17.2, wherein at least one of the first time interval Ts1 and the second time interval Ts2 overlap; A method for operating a multi-beam charged particle microscope (1), comprising:

[0176] Clause 2: A method of operating a multi-beam charged particle microscope (1) as described in clause 1, wherein the second image acquisition of the second image patch 17.2 begins before the end of the third time interval Tr when the wafer stage (500) has completely stopped.

[0177] Clause 3: A method of operating a multi-beam charged particle microscope (1) as described in clause 1 or 2, wherein a third time interval Tr of wafer movement begins before the end of time interval Ts1 when image acquisition of the first image patch 17.1 has ended.

[0178] Clause 4: A method for operating a multi-beam charged particle microscope (1) described in any one of clauses 1 to 3, further comprising calculating the start time of a third time interval Tr of wafer movement in a first time interval Ts1 of image acquisition of a first image patch 17.1 so that the positional deviation of a first center position of the first image patch 17.1 from the line of sight (53) of the multi-beam charged particle microscope (1) or the movement speed of the wafer stage (500) is below a predetermined threshold.

[0179] Clause 5: A method for operating a multi-beam charged particle microscope (1) described in any one of clauses 1 to 4, further comprising calculating the start time of a second time interval Ts2 of second image acquisition in the time interval Tr of wafer stage movement so that the positional deviation of the second center position 21.2 of the second image patch 17.2 from the line of sight (53) of the multi-beam charged particle microscope (1) or the movement speed of the wafer stage (500) is below a predetermined threshold.

[0180] Clause 6: predicting a series of wafer stage positions over a time interval Tr of wafer stage (500) movement; computing at least first and second control signals from the predicted wafer stage position; providing a first control signal to a first deflection system (110) in a primary beam path (13) of the multi-beam charged particle microscope (1) and a second control signal to a second deflection system (222) in a secondary beam path (11) of the multi-beam charged particle microscope (1); A method of operating a multi-beam charged particle microscope (1) according to any one of clauses 1 to 5, further comprising:

[0181] Clause 7: A high-throughput and high-resolution multi-beam charged particle system (1), a charged particle multi-beamlet generator (300) for generating a plurality of primary charged particle beamlets (3); an object irradiation unit (100) including a first deflection system (110) for scanning a wafer surface (25) with a plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) arranged in an object plane (101) at spot positions (5) of the plurality of primary charged particle beamlets (3); a detection unit (200) comprising a projection system (205), a second deflection system (222) and an image sensor (207) for imaging the plurality of secondary electron beamlets (9) onto the image sensor (207) and for, in use, acquiring digital images of a first image patch (17.1) and a second image patch (17.2) of the wafer surface (25); a sample stage (500) with a stage motion controller, the stage motion controller comprising a plurality of motors configured to be independently controlled, the stage configured to position and hold the wafer surface (25) at the object plane (101) during acquisition of the digital images of the first image patch (17.1) and the second image patch (17.2); a plurality of detectors including a stage position sensor (520) and an image sensor (207), configured, in use, to generate a plurality of sensor data including position data of the sample stage (500); a control unit (800) configured, in use, to perform a first image acquisition of a first image patch (17.1) in a first time interval Ts1 and a second image acquisition of a second image patch (17.2) in a second time interval Ts2, and configured to, upon triggering of the sample stage (500) in a third time interval Tr, move the sample stage (500) from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch (17.2) so that at least one of the first time interval Ts1 and the second time interval Ts2 overlaps with the third time interval Tr; A system equipped with

[0182] Clause 8: The system described in Clause 7, wherein the control unit is further configured to determine the start time of a third time interval Tr of wafer movement in the first time interval Ts1 of image acquisition of the first image patch 17.1 so that the positional deviation of the first center position of the first image patch 17.1 from the line of sight (53) of the multi-beam charged particle microscope (1) or the movement speed of the wafer stage (500) is below a predetermined threshold.

[0183] Clause 9: A system as described in clause 7 or 8, wherein the control unit is further configured to determine the start time of a second time interval Ts2 of second image acquisition in the time interval Tr of wafer stage movement so that the positional deviation of the second central position 21.2 of the second image patch 17.2 from the line of sight (53) of the multi-beam charged particle microscope (1) or the movement speed of the front wafer stage (500) is below a predetermined threshold.

[0184] Clause 10: A system described in any one of clauses 7 to 9, wherein the control unit is further configured to predict a series of wafer stage positions during a time interval Tr of movement of the wafer stage (500), calculate at least first and second control signals from the predicted wafer stage positions, supply the first control signal to a first deflection system (110) in a primary beam path (13) of the multi-beam charged particle microscope (1), and supply the second control signal to a second deflection system (222) in a secondary beam path (11) of the multi-beam charged particle microscope (1).

[0185] Clause 11: A method of operating a high-throughput and high-resolution multi-beam charged particle system (1), comprising: a first image acquisition of a first image patch 17.1, a second image acquisition of a second image patch 17.2, and moving a wafer stage (500) from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch 17.2 all within a time interval TG; the first image acquisition of the first image patch 17.1 is in the first time interval Ts1; a second image acquisition of a second image patch 17.2 in a second time interval Ts2; Moving the wafer stage (500) from the first center position (21.1) of the first image patch (17.1) to the second center position (21.2) of the second image patch 17.2 is in the third time interval Tr, A method of operating a multi-beam charged particle system (1) in which the time interval TG is shorter than the sum of Ts1, Ts2, and Tr, i.e., TG < Ts1 + Ts2 + Tr.

[0186] Clause 12: A multi-beam charged particle microscope (1) for wafer inspection, A charged particle multi-beamlet generator (300) that generates a plurality of primary charged particle beamlets (3), An object irradiation unit (100) including a first deflection system (110) that scans the wafer surface (25) with a plurality of primary charged particle beamlets (3) for generating a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) disposed on the object plane (101) at the scanning spot positions (5) of the plurality of primary charged particle beamlets (3), A detection unit (200) including a projection system (205), a second deflection system (222), and an image sensor (207) that forms an image of the plurality of secondary electron beamlets (9) on the image sensor (207) and acquires digital images of the first image patch (17.1) and the second image patch (17.2) of the wafer surface (25) during use, A sample stage (500) including a stage position sensor (520) that positions and holds the wafer surface (25) on the object plane (101) and moves the wafer surface from the first image patch (17.1) to the second image patch (17.2) when acquiring the digital image of the first image patch (17.1), A plurality of detectors including a stage position sensor (520) and an image sensor (207) configured to generate a plurality of sensor data including position data of the sample stage (500) during use, a first compensator in the object irradiation unit (100) configured to displace or rotate scanning spot positions (5) of the plurality of primary charged particle beamlets (3) on the wafer surface (25); a second compensator in the projection system (205) configured to compensate for displacement or rotation of the scanning spot positions (5) of the plurality of primary charged particle beamlets (3) and to keep the spot positions (15) of the plurality of secondary electron beamlets (9) constant on the image sensor (207); During acquisition of the digital image of the first image patch (17.1) or the second image patch (17.2), a first set of control signals C p a control unit (800) configured to synchronously control a first compensator in the object illumination unit (100) and a second compensator in the projection system (205) by generating A multi-beam charged particle microscope (1) equipped with

[0187] Clause 13: The control unit (800) generates a first set of control signals C p and providing the calculated value to the first compensator and the second compensator, thereby compensating for changes in the position or orientation of the sample stage.

[0188] Clause 14: The control unit (800) generates a first set of control signals C p and supplying the calculated value to the first compensator and the second compensator, thereby compensating for changes in the position of the line of sight (53) of the object illumination unit (100).

[0189] Clause 15: The control unit (800) generates a first set of control signals C p and supplying the calculated difference to a first compensator and a second compensator, thereby compensating for a difference between a change in position or orientation of the sample stage and a change in position of the line of sight of the object projection unit.

[0190] Clause 16: The control unit (800) generates a first set of control signals C p A multi-beam charged particle microscope (1) according to any one of clauses 12 to 15, configured to compensate for the movement speed of the sample stage (500) when acquiring a digital image of the first image patch (17.1) or the second image patch (17.2) by calculating and supplying the calculated value to the first compensator and the second compensator.

[0191] Clause 17: A method for wafer inspection by multi-beam charged particle microscopy, comprising: a first image acquisition step of a first image patch in a first time interval Ts1; moving the wafer stage from the position of the first image patch to the second image patch during a time interval Tr; a second image acquisition step of a second image patch in a second time interval Ts2; Therefore, calculating at least a first error amplitude from a plurality of sensor signals during a first time interval Ts1; predicting a transition of the first error amplitude over at least a travel time interval Tr and a second time interval Ts2 in a first time interval Ts1; - during at least the transfer time interval Tr, supplying a control signal to the control unit of the multi-beam charged particle microscope to keep the predicted error amplitude progression below a predetermined threshold during a second time interval Ts2; A method comprising:

[0192] Clause 18: The method of clause 17, wherein the prediction of the course of the first error amplitude is generated by a predictive model or by extrapolation.

[0193] Clause 19: The method of clause 17 or 18, wherein the first error amplitude represents at least one of line of sight displacement, wafer stage displacement, wafer stage rotation, line of sight rotation, magnification error, focus error, astigmatism error, or distortion error.

[0194] Clause 20: A method according to any one of clauses 17 to 19, controlling components including at least one of a wafer stage, a first deflection unit, a second deflection unit, a high-speed compensator of a multi-beamlet generating unit, or a high-speed compensator of a detection unit by supplying a control signal to a control unit of a multi-beam charged particle microscope.

[0195] Clause 21: A method of operation of a multi-beam charged particle microscope with a control unit, the method comprising, during image acquisition of a series of image patches including a first image patch and a second image patch: Decomposing a data stream comprising a plurality of sensor data into a set of error amplitudes; extracting a set of drift control signals and a set of dynamic control signals; providing a set of drift control signals to a slow acting compensator; providing a set of dynamic control signals to a fast acting compensator; A method comprising a series of operational steps including:

[0196] Clause 22: The method described in Clause 21, wherein the step of extracting a set of drift control signals and a set of dynamic control signals is performed during a time interval Ts1 of image acquisition of a first image patch, and the step of supplying the set of drift control signals to a slow motion compensator is performed during a time interval Tr of movement of the substrate by the substrate stage from the first image patch to the second image patch.

[0197] Clause 23: The method of clause 21 or 22, wherein the step of providing a set of dynamic control signals to the fast acting compensator is performed in a time interval Ts1.

[0198] Clause 24: The method of clause 22 or 23, wherein the step of providing a set of dynamic control signals to the fast-acting compensator is further performed during a time interval Ts2 of the image scanning of the second image patch.

[0199] Clause 25: The method according to any one of clauses 21 to 24, further comprising the step of predicting a time course of at least one of the error amplitudes.

[0200] Clause 26: The method of clause 25, comprising predicting a slow-varying drift of at least one of the error amplitudes; and predicting a fast-varying dynamic change of at least one of the error amplitudes.

[0201] Clause 27: by one or more processors of an apparatus including a charged particle source generating a plurality of primary charged particle beamlets, determining a lateral displacement of a movable stage in at least one of the X and Y axes; determining a lateral displacement of the line of sight of the object illumination unit; instructing the controller to at least partially compensate for the lateral displacement by application of a first signal that deflects a plurality of primary charged particle beamlets incident on the sample; A non-transitory computer-readable medium comprising a set of instructions executable to cause an apparatus to perform a method including:

[0202] Second set of clauses Clause 1: a movable stage configured to hold a sample; an object illumination unit configured to illuminate a surface of the sample with a plurality of focal spots of a plurality of primary charged particle beamlets; a charged particle beam generator configured to generate a plurality of primary charged particle beamlets from a charged particle source; a stage sensor configured to determine a lateral displacement or rotation of the stage; an image sensor configured to determine a lateral displacement of the line of sight of the object illumination unit; a control unit configured to generate and apply at least an additional voltage signal to a first beam deflector in the object illumination unit configured, in use, to generate an additional displacement or rotation of the plurality of primary charged particle beamlets to at least partially compensate for a difference between the lateral displacement of the line of sight and the lateral displacement or rotation of the stage; A multi-beam charged particle beam system equipped with

[0203] Clause 2: The system described in clause 1, wherein the control unit is further configured to calculate a lateral displacement or rotation of the stage corresponding to the difference between a current position of the stage and a target position of the stage during scanning of the multiple primary charged particle beamlets on the sample surface.

[0204] Clause 3: A system described in clause 1 or 2, wherein the control unit is further configured to calculate a lateral displacement of the line of sight corresponding to the difference between a current position of the line of sight and a target position of the line of sight during scanning of the multiple primary charged particle beamlets on the sample surface.

[0205] Clause 4: A system described in clause 2 or 3, wherein the control unit and the first beam deflector are further configured to dynamically adjust at least one drive voltage signal during scanning of the primary charged particle beamlet on the sample.

[0206] Clause 5: A system described in any one of clauses 1 to 4, further comprising a second beam deflector in the secondary electron beam path configured to at least partially compensate for additional displacement or rotation of multiple secondary electron beamlets resulting from the beam spot positions of multiple primary charged particle beamlets during scanning.

[0207] Clause 6: A system described in any one of clauses 1 to 5, wherein the control unit further comprises a stage motion controller, the stage motion controller comprising a plurality of motors configured to be independently controlled by control signals.

[0208] Clause 7: A system described in any one of clauses 1 to 6, wherein the control unit comprises a processor configured to derive a plurality of error vector amplitudes based on a plurality of sensor signals and to extract at least one of a plurality of control signals from the plurality of error vector amplitudes.

[0209] Clause 8: A method for irradiating a sample disposed on a stage in a multi-beam charged particle beam system, comprising: generating a plurality of primary charged particle beamlets from a charged particle source; Determining a lateral displacement or rotation of the movable stage in the XY plane; determining a line of sight of a multi-beam charged particle system; determining a displacement vector from the lateral displacement or rotation of the stage and the line of sight position; applying at least an additional voltage signal to a beam deflector in the primary charged particle beam path to generate, in use, an additional displacement or rotation of the plurality of primary charged particle beamlets to at least partially compensate for a displacement vector corresponding to a lateral displacement or rotation of the stage relative to the line of sight position; A method comprising:

[0210] Clause 9: The method described in Clause 8, wherein the lateral displacement or rotation of the stage corresponds to the difference between the current position of the stage and the target position of the stage, and the rotational displacement varies during scanning of the multiple primary charged particle beamlets over the sample surface.

[0211] Clause 10: The method of clause 8 or 9, further comprising dynamically adjusting at least one of the voltage signals during scanning of the primary charged particle beamlets over the sample.

[0212] Clause 11: A method according to any one of clauses 8 to 10, further comprising applying at least a second additional voltage signal to a beam deflector in the secondary electron beam path to at least partially compensate for additional displacement or rotation of the multiple secondary electron beamlets resulting from the beam spot positions of the multiple primary charged particle beamlets during scanning.

[0213] Clause 12: The method of any one of clauses 8 to 11, further comprising applying the control signal to a stage motion controller, the stage motion controller comprising a plurality of motors configured to be independently controlled by the control signal.

[0214] Clause 13: A method according to any one of clauses 8 to 12, further comprising deriving a plurality of error vector amplitudes based on a plurality of sensor signals, and extracting at least one of a plurality of control signals from the plurality of error vector amplitudes.

[0215] Third set of clauses Clause 1: A multi-beam charged particle microscope (1) for wafer inspection, a charged particle multi-beamlet generator (300) configured to generate a plurality of primary charged particle beamlets (3) in a raster configuration (41); an object illumination unit (100) configured to illuminate a wafer surface (25) with a plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) arranged in an object plane (101) at scanning spot positions (5) of the plurality of primary charged particle beamlets (3); a detection unit (200) comprising a projection system (205) and an image sensor (207), configured to image the plurality of secondary electron beamlets (9) onto the image sensor (207) and to acquire a digital image of a first image patch (17.1) of the wafer surface (25); a sample stage (500) having a stage position sensor (520) configured, in use, to position and hold a wafer surface (25) in an object plane (101) of an object projection unit (100); a first compensator (132, 110) in the object illumination unit (100) configured, in use, to incrementally displace or rotate scanning spot positions (5) of a plurality of primary charged particle beamlets (3) on a wafer surface (25); a second compensator (232, 222) in the projection system (205) configured, in use, to keep the spot positions (15) of the plurality of secondary electron beamlets (9) constant on the image sensor (207) by compensating for additional displacements or rotations of the scanning spot positions (5) of the plurality of primary charged particle beamlets (3); a control unit (800) configured to compensate for displacement of the wafer surface (25) induced by movement of the sample stage (500) by at least a first compensator (132, 110) and a second compensator (232, 222); A multi-beam charged particle microscope (1) equipped with

[0216] Clause 2: A multi-beam charged particle microscope (1) as described in clause 1, wherein the first compensator (132, 110) comprises one of an electrostatic lens, an electrostatic deflector, an electrostatic astigmatism corrector, an electrostatic microlens array, an electrostatic astigmatism corrector array, or an electrostatic deflector array.

[0217] Clause 3: The control unit (800) generates a first set of control signals C during the acquisition of the digital image of the first image patch (17.1). p 3. A multi-beam charged particle microscope (1) according to claim 1 or 2, configured to synchronously control a first compensator (132, 110) in the object illumination unit (100) and a second compensator (232) in the projection system (205) by generating a

[0218] Clause 4: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 3, further comprising a third compensator (330, 332) in the charged particle multi-beamlet generator (300) configured to, when in use, additionally displace or rotate the scanning spot positions (5) of the multiple primary charged particle beamlets (3) on the wafer surface (25).

[0219] Clause 5: The control unit (800) generates a first set of control signals C during the acquisition of the digital image of the first image patch (17.1). p 5. A multi-beam charged particle microscope (1) according to clause 4, configured to synchronously control either a first compensator (132, 110) in an object illumination unit (100), a second compensator (232, 222) in a projection system (205), or a third compensator (330, 332) in a charged particle multi-beamlet generator (300) by generating a

[0220] Clause 6: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 5, further comprising a stage position sensor (520) and an image sensor (207), and a plurality of detectors configured to generate a plurality of sensor data when in use.

[0221] Clause 7: A multi-beam charged particle microscope (1) as described in Clause 6, wherein the control unit (800) is configured to derive a drive signal for a first compensator (132, 110) in the object irradiation unit (100) from multiple sensor data to achieve additional displacement of the scanning spot positions (5) of multiple primary charged particle beamlets (3) synchronized with the displacement of the wafer surface (25).

[0222] Clause 8: A multi-beam charged particle microscope (1) according to clause 7, wherein the additional displacement comprises a rotation of the raster arrangement (41) of the plurality of primary charged particle beamlets (3).

[0223] Clause 9: A multi-beam charged particle microscope (1) described in any one of clauses 6 to 8, wherein the control unit (800) is further configured to compensate for additional displacement of the spot position (5) on the displacement wafer surface (25) by a second compensator (232, 222) in the projection system (205), and the second compensator (232, 222) in the projection system (205) is configured to operate synchronously with the first compensator (132, 110) in the object irradiation unit (100) to keep the spot positions of the multiple secondary electron beamlets (9) on the image detector (207) constant.

[0224] Clause 10: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 9, wherein the first compensator in the object irradiation unit (100) is a first deflection system (110), and the control unit (800) is configured to calculate a control signal for additional displacement or rotation of the scanning spot position (5) of the multiple primary charged particle beamlets (3) and supply it to the first deflection system (110) to compensate for the displacement or rotation of the sample stage (500) relative to the line of sight (53) of the object irradiation unit (100).

[0225] Clause 11: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 10, wherein the second compensator in the projection system (205) is a second deflection system (222), and the control unit (800) is configured to calculate and supply a control signal to the second deflection system (110) to compensate for additional displacement or rotation of the scanning spot positions (5) of the multiple primary charged particle beamlets (3) on the displacement wafer surface (25).

[0226] Clause 12: A multi-beam charged particle microscope (1) according to any one of clauses 1 to 11, comprising at least one of a separate compensator in the charged particle multi-beamlet generator (300), a separate compensator in the detection unit (200), or a separate compensator in the object illumination unit (100).

[0227] Clause 13: The control unit (800), in use, analyzes the plurality of sensor data and, in use, calculates a set of K amplitudes A of K error vectors. k 13. The multi-beam charged particle microscope (1) according to any one of clauses 1 to 12, comprising a sensor data analysis system (818) configured to calculate:

[0228] Clause 14: A multi-beam charged particle microscope (1) as described in clause 13, wherein the control unit (800) is provided with an image data acquisition unit (810) configured to, in use, reduce image sensor data from the image sensor (207) to an image sensor data fragment representing less than 10%, preferably less than 2%, of the image sensor data, and provide this image sensor data fragment to a sensor data analysis system (818).

[0229] Clause 15: A multi-beam charged particle microscope (1) as described in clause 14, wherein the image data acquisition unit (810) is configured to, when in use, reduce image sensor data from the image sensor (207) to an image sensor data fragment comprising digital image data of multiple secondary electron beamlets at a low sampling rate.

[0230] Clause 16: A multi-beam charged particle microscope (1) as described in clause 14, wherein the image data acquisition unit (810) is configured to reduce image sensor data from the image sensor (207) to an image sensor data fragment comprising digital image data of a reduced set of secondary electron beamlets (9).

[0231] Clause 17: The sensor data analysis system (818) calculates a set of amplitudes A of the error vector. k At least one amplitude A n 17. The multi-beam charged particle microscope (1) according to any one of clauses 13 to 16, configured to predict the time course of

[0232] Clause 18: The control unit (800) calculates a set of amplitudes A of the error vector. k to a first set of control signals C p 18. The multi-beam charged particle microscope (1) according to any one of clauses 13 to 17, further comprising a control calculation processor (840) that calculates the above.

[0233] Clause 19: A multi-beam charged particle microscope (1) described in any one of clauses 13 to 18, wherein the sensor data analysis system (818) is configured to derive a sensor data vector DV of length L (L≧K) from multiple sensor data.

[0234] Clause 20: The control unit (800) generates a first set of control signals C p A multi-beam charged particle microscope (1) as described in any one of clauses 4 to 19, configured to calculate and supply at least one of the control signals to a third compensator (330, 332) to compensate for rotation of the sample stage (500) by inducing rotation of a raster arrangement (41) of a plurality of primary charged particle beamlets (3).

[0235] Clause 21: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 20, wherein the control unit (800) is further configured to generate a control signal for moving the wafer surface (25) by the wafer stage (500) to a second central position of the second image patch (17.2) in the object plane (101) for image acquisition of a digital image of the second image patch (17.2).

[0236] Clause 22: The control unit (800) generates a second set of P control signals C from the plurality of sensor data during a time interval Tr of the movement of the wafer stage (500) to a second center position of the second image patch (17.2). p 22. The multi-beam charged particle microscope (1) of clause 21, further configured to control any of the compensators by calculating:

[0237] Clause 23: A multi-beam charged particle microscope (1) as described in clause 21 or 22, wherein the control unit (800) is further configured to calculate the start time of image acquisition of the second image patch (17.2) in the time interval Tr and to start image acquisition of the second image patch (17.2) in the deceleration time interval Td of the wafer stage (500), and further configured to supply at least an offset signal of the predicted offset position of the wafer stage (500) in the time interval Td to the first and second compensators.

[0238] Fourth Set of Clauses Clause 1: a stage configured to hold a sample and movable in at least one of an X-axis, a Y-axis, and a Z-axis; a position sensing system configured to determine lateral and vertical displacement or rotation of the stage; a controller configured to at least partially compensate for lateral displacement of the stage by applying a first signal to deflect the plurality of primary charged particle beamlets incident on the sample, and to at least partially compensate for displacement of the plurality of secondary electron beamlets by applying a second signal to deflect the plurality of secondary electron beamlets resulting from positions of the deflected primary charged particle beamlets on the sample; A multi-beam charged particle beam system equipped with

[0239] Clause 2: The system described in clause 1, wherein the first signal includes an electrical signal that affects how the multiple primary charged particle beamlets are deflected in at least one of the X and Y axes.

[0240] Clause 3: The system of clause 2, wherein the electrical drive signal comprises a signal having a bandwidth in the range of 0.1 kHz to 10 kHz.

[0241] Clause 4: A system according to any one of clauses 1 to 3, wherein the lateral displacement corresponds to a difference between a current position of the stage and a target position of the stage in at least one of the X and Y axes.

[0242] Clause 5: A system described in any one of clauses 1 to 4, wherein the controller is further configured to dynamically adjust at least one of the first signal or the second signal during scanning of the multiple primary charged particle beamlets on the sample.

[0243] Clause 6: The system of any one of clauses 1 to 5, further comprising a stage motion controller having a plurality of motors configured to be independently controlled by a third signal.

[0244] Clause 7: The system described in clause 6, wherein each of the multiple motors is independently controlled to adjust the tilt of the stage so that the stage is substantially perpendicular to the optical axis of the primary charged particle beam.

[0245] Clause 8: The system of clause 6 or 7, wherein the plurality of motors includes at least one of a piezoelectric motor, a piezoelectric actuator, or an ultrasonic piezoelectric motor.

[0246] Clause 9: The system of any one of clauses 1 to 8, further comprising: a first component configured to form a plurality of error vector amplitudes based on a plurality of sensor signals; and a second component configured to extract at least one of a plurality of control signals from the plurality of error vector amplitudes.

[0247] Clause 10: The system described in clause 9, wherein the first component is configured to form a plurality of error vector amplitudes based on the lateral displacement of the stage and the actual position of the line of sight of the multi-beam charged particle beam system.

[0248] Clause 11: The system of clause 9 or 10, wherein the extraction of at least one of the plurality of control signals is based on a plurality of error vector magnitude predictive models.

[0249] Clause 12: A system described in any one of clauses 9 to 11, wherein the extraction of at least one of the plurality of control signals is further based on a predictive model of the actuation output of the stage.

[0250] Clause 13: A system described in any one of clauses 1 to 12, wherein the position sensing system uses a laser interferometer, a capacitance sensor, a confocal sensor array, a grating interferometer, or a combination thereof to determine the lateral and vertical displacement and rotation of the stage.

[0251] Fifth Set of Clauses Clause 1: A method for irradiating a sample disposed on a stage in a multi-beam charged particle beam system, comprising: generating a plurality of primary charged particle beamlets from a charged particle source; determining a lateral displacement and a rotation of the movable stage in at least one of an XY axis and a Z axis; at least partially compensating for lateral displacement or rotation of the stage by application of a first signal that deflects a plurality of primary charged particle beamlets incident on the sample; at least partially compensating for the displacement of the plurality of secondary electron beamlets by applying a second signal that deflects the plurality of secondary electron beamlets resulting from the position of the deflected primary charged particle beamlet on the sample; A method comprising:

[0252] Clause 2: The method of clause 1, wherein the first signal comprises an electrical signal that affects how the primary charged particle beam is deflected in at least one of the X and Y axes.

[0253] Clause 3: The method of clause 1 or 2, wherein the lateral displacement corresponds to a difference between a current position of the stage and a target position of the stage in at least one of the X and Y axes.

[0254] Clause 4: The method of any one of clauses 1 to 3, further comprising dynamically adjusting at least one of the first signal or the second signal during scanning of the multiple primary charged particle beams on the sample.

[0255] Clause 5: The method of any one of clauses 1 to 4, further comprising applying a third signal to a stage motion controller, the stage motion controller comprising a plurality of motors configured to be independently controlled by the third signal.

[0256] Clause 6: The method of any one of clauses 1 to 5, further comprising deriving a plurality of error vector amplitudes based on a plurality of sensor signals, and extracting at least one of a plurality of control signals from the plurality of error vector amplitudes.

[0257] Clause 7: The method of clause 6, further comprising predicting at least one of the control signals based on a predictive model of the temporal behavior of the plurality of error vector amplitudes.

[0258] Clause 8: The method of clause 6 or 7, further comprising predicting at least one of the plurality of control signals based on a predictive model of the actuation output of the stage.

[0259] Sixth Set of Clauses Clause 1: A method for wafer inspection using a multi-beam charged particle microscope (1) comprising a plurality of detectors including an image sensor (207) and a stage position sensor (520), and a set of compensators including at least first and second deflection systems (110, 222), comprising: Positioning and aligning a wafer surface (25) of the wafer relative to a position in a local wafer coordinate system (551) that includes the line of sight of the multi-beam charged particle microscope (1); performing image acquisition to acquire a digital image of a first image patch (17.1) of the wafer surface (25); collecting a plurality of sensor data from a plurality of detectors; From multiple sensor data, a set of K error amplitudes A k and deriving A set of error amplitudes A k from a first set of control signals C p and deriving providing a first set of control signals to a set of compensators in step b of image acquisition; A method for wafer inspection with a multi-beam charged particle microscope (1), comprising:

[0260] Clause 2: A method for wafer inspection using a multi-beam charged particle microscope (1) according to clause 2, further comprising the step of deriving a sensor data vector DV of length L (L≧K) from the plurality of sensor data.

[0261] Clause 3: A set of amplitudes A of the error vector k At least one of the amplitudes A n 3. A method for wafer inspection by a multi-beam charged particle microscope (1) according to clause 1 or 2, further comprising the step of deriving a time progression of

[0262] Article 4: Control Signal C p 4. A method for wafer inspection using a multi-beam charged particle microscope (1) according to any one of clauses 1 to 3, further comprising the step of compensating for changes in the position or orientation of the sample stage (500) by supplying to the first and second deflection units (110, 222).

[0263] Clause 5: A set of error amplitudes A k to a second set of control signals C p and providing a second set of control signals in step a) of positioning and aligning the wafer surface (25) of the wafer.

[0264] Seventh Set of Clauses Clause 1: A method of operating a multi-beam charged particle microscope (1) configured for wafer inspection, comprising: defining a set of predetermined normalized error vectors representing a set of image qualities and deviations from the set of image qualities; determining a set of thresholds for the amplitudes of the set of normalized error vectors; selecting a set of compensators for the multi-beam charged particle microscope; determining a sensitivity matrix according to a linear perturbation model by varying at least the drive signal of each of the set of compensators; deriving a set of normalized drive signals that compensate for each of the set of normalized error vectors; storing the normalized drive signal, the set of thresholds, and the normalized error vector in a memory of a control unit of the multi-beam charged particle microscope; A method for operating a multi-beam charged particle microscope (1), comprising:

[0265] Clause 2: A method of operating a multi-beam charged particle microscope (1) as described in clause 1, wherein a set of compensators comprises a first deflection unit (110) of the multi-beam charged particle microscope (1) for scanning and deflecting a plurality of primary charged particles (3) and a second deflection unit (222) for scanning and deflecting a plurality of secondary electron beamlets (9) generated when using the multi-beam charged particle microscope (1).

[0266] 8th Set of Clauses Clause 1: In use, receiving a plurality of sensor data from a plurality of sensors of a multi-beam charged particle microscope (1) and constructing a sensor data vector; expanding the sensor data vector into a set of normalized error vectors stored in a memory of the control unit, and determining from the sensor data vector the actual amplitudes of the set of normalized error vectors; comparing the set of actual amplitudes to a set of thresholds stored in a memory of the control unit; deriving a set of control signals from the set of actual amplitudes based on a comparison of the set of actual amplitudes to a stored set of threshold values; deriving a set of actual drive signals from a set of normalized drive signals stored in a memory of the control unit using a set of control signals; providing a set of actual drive signals to a set of compensators of the multi-beam charged particle microscope (1) so that, during operation of the multi-beam charged particle microscope (1), a set of actual amplitudes of a set of normalized error vectors is less than a set of thresholds; A method for operating a multi-beam charged particle microscope (1), comprising:

[0267] Clause 2: A method of operating a multi-beam charged particle microscope (1) as described in clause 1, wherein the plurality of sensor data includes at least one of position or velocity information regarding the actual position and actual velocity of a wafer stage (500) that holds or moves a wafer during inspection by the multi-beam charged particle microscope (1).

[0268] Clause 3: A method of operating a multi-beam charged particle microscope (1) described in clause 1 or 2, wherein the plurality of sensor data includes at least one of the actual positions of the line of sight (52) during wafer inspection by the multi-beam charged particle microscope (1).

[0269] Clause 4: A method of operating a multi-beam charged particle microscope (1) described in any one of clauses 1 to 3, wherein the above steps are repeated at least twice, at least ten times, preferably for each scan line when acquiring an image patch.

[0270] Clause 5: A method of operating a multi-beam charged particle microscope (1) according to any one of clauses 1 to 4, further comprising a step of predicting a subset of transition amplitudes of at least a subset of a set of actual amplitudes according to the predicted transition of the multi-beam charged particle microscope over a prediction time interval during wafer inspection.

[0271] Clause 6: A method of operating a multi-beam charged particle microscope (1) according to any one of clauses 1 to 5, further comprising the step of recording, in use, at least a subset of the set of actual amplitudes of the multi-beam charged particle microscope to generate a history of the subset of the set of actual amplitudes.

[0272] Clause 7: A method for operating a multi-beam charged particle microscope (1) according to any one of clauses 1 to 6, further comprising the steps of: deriving a set of predicted control signals from the set of transition amplitudes during wafer inspection, and deriving a set of predicted drive signals from the set of predicted control signals; and supplying the set of predicted drive signals to a set of compensators in a time series manner during wafer inspection, thereby causing a subset of actual error amplitudes to be below a set of thresholds during operation of the multi-beam charged particle microscope in a predicted time interval.

[0273] Clause 8: A multi-beam charged particle microscope comprising a control unit (800) and installed software code and configured to apply any of the methods described in any one of clauses 1 to 7.

[0274] 9th Set of Clauses Clause 1: A method for focusing a plurality of primary charged particle beamlets onto a sample, comprising: irradiating a sample disposed on a stage of a multi-beam charged particle beam system with a plurality of primary charged particle beamlets to form a plurality of focal spots on a surface of the sample; adjusting the position and rotation of the multiple focal spots of the multiple charged particle beamlets relative to the sample using at least a first component of the multi-beam charged particle system; scanning focal spots of the plurality of primary charged particle beamlets relative to the sample along a plurality of predetermined primary scanning beam paths using a second component of the multi-beam charged particle system; dynamically steering the predetermined scanning beam path relative to the sample using the first component, the second component, or the third component; A method comprising:

[0275] Clause 2: The method of clause 1, comprising scanning the focal spot of a plurality of primary charged particle beamlets using a second component by adding at least a first deflection voltage to a first scanning voltage.

[0276] Clause 3: generating a plurality of primary charged particle beamlets using a charged particle multi-beamlet generator; Using components of the charged particle multi-beamlet generator, adjusting or dynamically manipulating the position and rotation of the multiple focal spots of the multiple charged particle beamlets relative to the sample; 3. The method of clause 1 or 2, further comprising:

[0277] Clause 4: generating and collecting a plurality of secondary electron beamlets from a surface of the sample at a plurality of focal spots of a plurality of primary charged particle beamlets; scanning the plurality of secondary electron beamlets along a predetermined secondary electron beam path such that focal spots of the plurality of secondary electron beamlets are at constant positions on an image sensor by using a fourth component of a projection system of the multi-beam charged particle system; dynamically steering a predetermined secondary electron beam path relative to the image sensor by using a fourth or fifth component of a projection system of the multi-beam charged particle system; 4. The method of any one of clauses 1 to 3, further comprising:

[0278] Clause 5: The method of clause 4, comprising scanning the plurality of secondary charged particle beamlets using a fourth component by adding at least a second deflection voltage to the second scanning voltage.

[0279] Clause 6: Determining the current position of the stage; determining a lateral displacement or rotation of the stage from a difference between a current position of the stage and a target position of the stage; 6. The method of any one of clauses 1 to 5, further comprising:

[0280] Clause 7: determining a first deflection voltage to compensate for lateral displacement or rotation of the stage; supplying a first deflection voltage to at least the first, second, or third component to dynamically steer the predetermined first scanning beam path relative to the sample; 7. The method of clause 6, further comprising:

[0281] Clause 8: determining a second deflection voltage; providing a second deflection voltage to at least the fourth or fifth component to dynamically steer the predetermined secondary electron beam path relative to the image sensor; 8. The method of clause 6 or 7, further comprising:

[0282] Tenth Set of Clauses Clause 1: an object illumination unit configured to illuminate a surface of a sample with a plurality of focal spots of a plurality of primary charged particle beamlets; a first component of the object illumination unit configured to adjust the position and rotation of a plurality of focal spots of a plurality of charged particle beamlets relative to the sample; a second component of the object illumination unit configured to scan the focal spots of the plurality of primary charged particle beamlets relative to the sample along a plurality of predetermined primary scanning beam paths; a third component configured to dynamically steer the predetermined scanning beam path relative to the sample position; and A multi-beam charged particle beam system equipped with

[0283] Clause 2: A multi-beam charged particle beam system according to clause 2, wherein the third component is the first component.

[0284] Clause 3: A multi-beam charged particle beam system according to clause 2, wherein the third component is the second component.

[0285] Clause 4: A multi-beam charged particle beam system as described in Clause 3, further comprising a control unit configured to add at least a first deflection voltage configured to dynamically manipulate a predetermined scanning beam path to a first scanning voltage supplied to a second component configured to scan the focal spots of the plurality of primary charged particle beamlets.

[0286] Clause 5: A multi-beam charged particle beam system according to any one of clauses 1 to 4, further comprising a charged particle multi-beamlet generator configured to generate a plurality of primary charged particle beamlets.

[0287] Clause 6: A multi-beam charged particle beam system according to any one of clauses 1 to 5, further comprising a control unit configured to, in use, adjust the line of sight of the object irradiation unit by the first component.

[0288] Clause 7: A projection system configured to collect and image a plurality of secondary electron beamlets originating from a surface of the sample at a plurality of focal spots of a plurality of primary charged particle beamlets; an image sensor configured to detect a plurality of focal spots of the plurality of secondary electron beamlets; a fourth component of a projection system of the multi-beam charged particle system configured to scan the plurality of secondary electron beamlets along a predetermined secondary electron beam path such that focal spots of the plurality of secondary electron beamlets are at constant positions on the image sensor; a fifth component of a projection system of the multi-beam charged particle system configured to dynamically steer the predetermined secondary electron beam path relative to the image sensor; 7. The multi-beam charged particle beam system according to any one of clauses 1 to 6, further comprising:

[0289] Clause 8: A multi-beam charged particle beam system according to clause 7, wherein the fifth component is the fourth component.

[0290] Clause 9: A multi-beam charged particle beam system as described in Clause 8, wherein the control unit is further configured to add at least a second deflection voltage configured to dynamically manipulate a predetermined secondary electron beam path to a second scanning voltage supplied to a fourth component that scans the plurality of secondary charged particle beamlets.

[0291] Clause 10: A multi-beam charged particle beam system according to any one of clauses 1 to 9, further comprising a stage sensor configured to determine a lateral displacement or rotation of the stage.

[0292] Clause 11: A multi-beam charged particle beam system as described in clause 10, wherein the control unit is further configured to derive the first and second deflection voltages from a lateral displacement or rotation imparted by the stage sensor.

[0293] Clause 12: A system according to any one of clauses 1 to 11, wherein the first component is disposed upstream of the second component.

[0294] 11th Set of Clauses Clause 1: A method for performing wafer inspection with a multi-beam charged particle beam device, comprising: irradiating a sample disposed on a stage with a plurality of primary charged particle beamlets; performing static adjustment of focal points of a plurality of primary charged particle beamlets; performing dynamic manipulation of focal points of the plurality of primary charged particle beamlets; A method comprising:

[0295] Clause 2: Determining slowly varying fluctuations of a multi-beam charged particle beam device, including determining slowly varying fluctuations of an object irradiation unit and detecting drift of a stage configured to hold a sample; determining a first drift compensation signal that compensates for slowly varying fluctuations; applying a first drift compensation signal to at least a component of the object illumination unit to perform static adjustment of focal points of the plurality of primary charged particle beamlets; 2. The method of clause 1, further comprising:

[0296] Clause 3: The method of clause 2, wherein determining the slowly changing fluctuations of the object illumination unit includes determining the slowly changing fluctuations of the line of sight of the object illumination unit.

[0297] Clause 4: Determining dynamic fluctuations of the multi-beam charged particle beam device, including determining dynamic fluctuations of an object irradiation unit and detecting vibrations of a stage configured to hold a sample; determining a first dynamic compensation signal that compensates for dynamic variations; applying a first dynamic compensation signal to at least a component of the object illumination unit to perform dynamic manipulation of focal points of the plurality of primary charged particle beamlets; 4. The method of any one of clauses 1 to 3, further comprising:

[0298] Clause 5: The method of clause 4, wherein determining the dynamic variation of the object illumination unit includes determining a dynamic change in the line of sight of the object illumination unit.

[0299] Clause 6: determining a second drift compensation signal that compensates for slow-varying fluctuations; applying a second drift compensation signal to at least a component of the projection unit to compensate for static alignment of the plurality of secondary electron beamlets resulting from the aligned focal points of the plurality of primary charged particle beamlets; 6. The method of any one of clauses 2 to 5, further comprising:

[0300] Clause 7: determining a second dynamic compensation signal that compensates for dynamic fluctuations; applying a second dynamic compensation signal to at least a component of the projection unit to compensate for dynamic steering of the plurality of secondary electron beamlets resulting from the dynamic steering focal points of the plurality of primary charged particle beamlets; 7. The method of any one of clauses 4 to 6, further comprising:

[0301] Clause 8: The method of any one of clauses 1 to 7, wherein the determination of the first and second drift compensation signals is based on a predictive model of the temporal behavior of the multi-beam charged particle radiation device.

[0302] Clause 9: The method according to clause 8, wherein the determination of the first or second drift compensation signal and the first or second dynamic compensation signal is based on a frequency analysis of the temporal behavior of the multi-beam charged particle radiation device.

[0303] Clause 10: The method of any one of clauses 1 to 9, further comprising receiving a plurality of sensor signals including sensor signals from a stage position sensor and an image sensor.

[0304] Clause 11: The method of clause 10, further comprising determining, by using a control unit, a drift and dynamic compensation signal based on the received plurality of sensor signals.

[0305] Clause 12: Estimating a predictive model of the temporal behavior of the multi-beam charged particle radiation device by using a processor of the control unit; determining, using a control unit, drift and dynamic compensation signals based on a predictive model; 12. The method of any one of clauses 1 to 11, further comprising:

[0306] Clause 13: The method of clause 12, wherein estimating the predictive model includes frequency analysis, low-pass filtering, and polynomial approximation.

[0307] Clause 14: The method of clause 12 or 13, further comprising synchronizing the drift and dynamic compensation signal with the predictive model by using at least a delay line.

[0308] Clause 15: Generating a beam deflection signal based on the dynamic compensation signal; modifying the beam scanning signal with the beam deflection signal; providing the modified beam scanning signal to a scanning beam deflection unit; 15. The method of any one of clauses 1 to 14, further comprising:

[0309] 12th Set of Clauses Clause 1: A multi-beam charged particle microscope (1) for wafer inspection, a charged particle multi-beamlet generator (300) for generating a plurality of primary charged particle beamlets (3); an object irradiation unit (100) including a first deflection system (110) for scanning a wafer surface (25) with a plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) arranged in an object plane (101) at spot positions (5) of the plurality of primary charged particle beamlets (3); a detection unit (200) comprising a projection system (205), a second deflection system (222) and an image sensor (207) for imaging the plurality of secondary electron beamlets (9) onto the image sensor (207) and for, in use, acquiring a digital image of a first image patch (17.1) on the wafer surface (25); a sample stage (500) equipped with a stage position sensor (520) for positioning and holding the wafer surface (25) in the object plane (101) during acquisition of the digital image of the first image patch (17.1); a plurality of detectors including a stage position sensor (520) and an image sensor (207), configured, in use, to generate a plurality of sensor data including position data of the sample stage (500); a set of compensators including at least a compensator in the object illumination unit (100) and a compensator in the projection system (205); During acquisition of the digital image of the first image patch (17.1), a first set of control signals C is generated from the plurality of sensor data. p a control unit (800) configured to generate a set of compensators to control the set of compensators; Equipped with A multi-beam charged particle microscope (1), wherein the control unit (800) is configured to compensate for displacements of the wafer surface (25) induced by movements of the sample stage (500).

[0310] Clause 2: The control unit (800) generates a first set of control signals C p and supplying the calculated value to a first compensator and a second compensator, thereby compensating for changes in the position of the line of sight (53) of the object illumination unit (100).

[0311] Clause 3: The control unit (800) generates a first set of control signals C p and supplying the calculated difference between the change in position or orientation of the sample stage (500) and the change in position of the line of sight (53) of the object projection unit (100) to a first compensator and a second compensator.

[0312] 13th Set of Clauses Clause 1: A multi-beam charged particle microscope (1) for wafer inspection, a. a charged particle multi-beamlet generator (300) for generating a plurality of primary charged particle beamlets (3); b. an object irradiation unit (100) having a first deflection system (110) for scanning a wafer surface (25) arranged in an object plane (101) with a plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25); c. a detection unit (200) comprising a projection system (205), a second deflection system (222) and an image sensor (207) for imaging the plurality of secondary electron beamlets (9) onto the image sensor (207) and for, in use, acquiring a digital image of a first image patch (17.1) on the wafer surface (25); d. A sample stage (500) equipped with a stage position sensor (520) for positioning and holding the wafer surface (25) in the object plane (101) during acquisition of the digital image of the first image patch (17.1); e. a control unit (800); f. a plurality of detectors including a stage position sensor (520) and an image sensor (207), configured, in use, to generate a plurality of sensor data including position data of the sample stage (500); g. a set of compensators comprising at least first and second deflection systems (110, 222); Equipped with The control unit (800) generates a first set of control signals C from the plurality of sensor data during acquisition of the digital image of the first image patch (17.1). p and controlling a set of compensators.

[0313] Clause 2: A multi-beam charged particle microscope (1) as described in clause 1, wherein the set of compensators further includes at least one of a compensator (330, 332) of the charged particle multi-beamlet generator (300) and a compensator (230, 232) of the detection unit (200).

[0314] Clause 3: The control unit (800), in use, analyzes the plurality of sensor data and, in use, calculates a set of K amplitudes A of K error vectors. k 3. A multi-beam charged particle microscope (1) according to clause 1 or 2, comprising a sensor data analysis system (818) configured to calculate:

[0315] Clause 4: A multi-beam charged particle microscope (1) as described in clause 3, wherein the control unit (800) is provided with an image data acquisition unit (810) configured to, in use, reduce image sensor data from the image sensor (207) to an image sensor data fragment representing less than 10%, preferably less than 2%, of the image sensor data, and provide this image sensor data fragment to a sensor data analysis system (818).

[0316] Clause 5: The sensor data analysis system (818) calculates a set of amplitudes A of the error vector. k At least one amplitude A n 5. A multi-beam charged particle microscope (1) according to clause 3 or 4, configured to predict the time course of

[0317] Clause 6: The control unit (800) calculates a set of amplitudes A of the error vector. k to a first set of control signals C p 6. The multi-beam charged particle microscope (1) according to any one of clauses 3 to 5, further comprising a control calculation processor (840) that calculates the above.

[0318] Clause 7: A multi-beam charged particle microscope (1) described in any one of clauses 3 to 6, wherein the sensor data analysis system (818) is configured to derive a sensor data vector DV of length L (L > K) from multiple sensor data.

[0319] Clause 8: The control unit (800) generates a first set of control signals C p A multi-beam charged particle microscope (1) described in any one of clauses 1 to 7, configured to compensate for changes in position or orientation of the sample stage (500) by calculating and supplying at least one of the control signals to the first and second detection units (110, 222).

[0320] Clause 9: The control unit (800) generates a first set of control signals C p A multi-beam charged particle microscope (1) described in any one of clauses 1 to 8, configured to compensate for changes in the position of the line of sight (53) of the object illumination unit (100) by calculating and supplying at least one of the control signals to the first and second detection units (110, 222).

[0321] Clause 10: The control unit (800) generates a first set of control signals C p A multi-beam charged particle microscope (1) as described in any one of clauses 1 to 9, configured to compensate for differences between changes in position or orientation of the sample stage (500) and changes in position of the line of sight (53) of the object illumination unit (100) by calculating at least one of the control signals and supplying it to the first and second detection units (110, 222).

[0322] Clause 11: The charged particle multi-beamlet generator (300) further comprises a fast compensator (330), and the control unit (800) generates a first set of control signals C pA multi-beam charged particle microscope (1) according to any one of clauses 1 to 10, configured to calculate and supply at least one of the control signals to a fast compensator (330) to induce rotation of a plurality of primary charged particle beamlets, thereby compensating for rotation of the sample stage (500).

[0323] Clause 12: A multi-beam charged particle microscope (1) described in any one of clauses 1 to 11, wherein the control unit (800) is further configured to generate a third control signal to move the wafer surface (25) by the wafer stage (500) to a second central position of the second image patch 17.2 in the object plane (101) for image acquisition of a digital image of the second image patch (17.2).

[0324] Clause 13: The control unit (800) generates a second set of P control signals C from the plurality of sensor data during a time interval Tr of the movement of the wafer stage (500) to a second center position of the second image patch (17.2). p 13. The multi-beam charged particle microscope (1) of clause 12, further configured to control the set of compensators by computing:

[0325] Clause 14: A multi-beam charged particle microscope (1) as described in clause 12 or 13, wherein the control unit (800) is further configured to calculate the start time of image acquisition of the second image patch (17.2) in the time interval Tr and to start image acquisition of the second image patch (17.2) in the deceleration time interval Td of the wafer stage (500), and further configured to supply at least an offset signal of the predicted offset position of the wafer stage (500) in the time interval Td to the first and second detection units (110, 222).

[0326] 14th Set of Clauses Clause 1: A method for increasing throughput of a multi-beam charged particle microscope, comprising: generating a plurality of beam spots of a plurality of primary charged particle beamlets on a surface of the sample in a raster configuration with a beam pitch d1; collectively scanning a plurality of primary charged particles along a predetermined scan path; controlling a beam pitch d1 of the plurality of primary charged particle beamlets; Including, The method, wherein the controlling step includes reducing the overlap area by compensating for variations in the beam pitch d1 through the use of a compensator that manipulates the beam spot position.

[0327] Clause 2: The method described in clause 1, wherein the controlling step includes providing a control signal to a multi-beam multi-pole deflection device to dynamically control multiple beam spot positions on the sample surface with a high accuracy of less than 100 nm, less than 70 nm, or less than 30 nm.

[0328] Clause 3: The method of clause 1 or 2, further comprising using an image sensor to detect the focal spot position on the sample surface with a high accuracy of less than 100 nm, less than 70 nm, or less than 30 nm.

[0329] Clause 4: The method of any one of clauses 1 to 3, wherein the beam pitch d1 is approximately 10 μm.

[0330] A non-transitory computer-readable medium can be provided that stores instructions for a processor (e.g., in the control unit 800 or the sensor data analysis system 818) to perform operations for wafer inspection, wafer imaging, stage calibration, displacement error calibration, displacement error compensation, manipulation of an electromagnetic field associated with the sample, communication with the image data acquisition unit 810, operation of the acceleration sensor, execution of algorithms to estimate or predict performance of the multi-beam charged particle microscope system 1 including the sample stage 500, and control of the multi-beam charged particle system. Common forms of non-transitory medium include, for example, a hard disk, a solid-state drive, any optical data storage medium, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash EPROM or any other flash memory, non-volatile random access memory (NVRAM), cache, registers, or any other memory chip or cartridge.

[0331] The block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer hardware or software products according to various exemplary embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, including one or more executable instructions for implementing specified logical functions. It should be understood that in some alternative implementations, the functions shown in the blocks may occur out of the order depicted in the figures. For example, depending on the functionality involved, two blocks shown in succession may be executed or performed substantially simultaneously, or the two blocks may be executed in the reverse order. Also, some blocks may be omitted. It should be understood that each block and combination of blocks in the block diagrams may be implemented by a dedicated hardware-based system or a combination of dedicated hardware and computer instructions that performs the specified functions or operations.

[0332] It should be understood that the embodiments of the present disclosure are not limited to the exact configurations described above and illustrated in the accompanying drawings, and that various improvements and modifications can be made without departing from the scope thereof. It should be understood that the present invention is not limited to the above set of clauses, and that various improvements and modifications or combinations of clauses can be made without departing from the scope thereof. It should be understood that the present invention is not limited to methods or apparatus, and should cover any apparatus configured to operate according to any method, or any method utilizing the elements and configurations of any apparatus described above or set of clauses. [Explanation of symbols]

[0333] 1. Multi-beam charged particle microscope system 3 Primary charged particle beamlets (composed of multiple primary charged particle beamlets) 5 Primary charged particle beam focal spot 7 Object or wafer 9 Secondary electron beamlets (composed of multiple secondary electron beamlets) 11 Secondary electron beam path 13 Primary beam path 15 Secondary charged particle image spot or focal spot 17 Image patch (e.g., first or second image patch 17.1, 17.2) 19 Overlapping Area 21 Image patch center position 25 Wafer surface 27 Scan path of primary charged particle beamlet 29 Center of image subfield 31 Image Subfield 33 First Examination Site 35 Second Test Site 37 Image subfield after scan rotation 39 Subfield 31 overlap area 41 Raster Configuration 51 Image coordinate system 53 Line of sight of multi-beam charged particle microscope 55 Displacement Vector 59 Rotation Vector Components 61 Individual displacement of image spots 100 Object Irradiation Unit 101 Object plane 102 Objective Lens 103.1, 103.2 First and second field lenses 105 Optical axis of multi-beam charged particle microscope system 108 First beam crossover 110 First deflection system 130 Object illumination unit slow compensator 132 Dynamic or fast compensator for object illumination units 138 Object illumination unit sensor 200 detection units 205 Projection system 206 Electrostatic Lens 207 Image Sensor 208 Imaging Lens 209 Imaging Lens 212 Second Crossover 214 Aperture 216 Active Elements 218 Third deflection system 220 Multi-aperture corrector 222 Second deflection system 230 Slow Compensator for Secondary Electron Beam Path 232 Fast Compensator for Detection Unit 238 Secondary Electron Beam Path Sensor 300 Charged Particle Multi-Beamlet Generator 301 Charged Particle Source 303 Collimator Lens 305 Primary Multibeamlet Configuration Unit 306 Active multi-aperture plate configuration 307 First Field Lens 308 Second Field Lens 309 Diverging Electron Beam 311Focal spot of primary electron beamlet 321 Intermediate image plane 330 Slow Compensator for Multi-Beamlet Generator 332 Fast Compensator for Multi-Beamlet Generator 390 Beam Steering Array or Deflector Array 400 Beam Splitter Unit 420 Magnetic focusing lens 430 Beam Splitter Unit Slow Compensator 500 sample stage 503 Sample Voltage Source 520 Stage Position Sensor 551 Local Wafer Coordinate System 601 Multi-Aperture Active Array 607 Conductive Wire 681 Electrode 685 Aperture or Aperture Array 800 Control Unit 810 Image Data Acquisition Unit 812 Image Stitching Unit 814 Image data memory 818 Sensor Data Analysis System 820 Projection System Control Module 830 Primary Beam Path Control Module 840 Control Calculation Processor 860 Deflection Control Module 880 Stage Control Module 901 Error Amplitude Threshold 903 Error Amplitude Gradient 905 Error Amplitude Threshold Window 907 Error Magnitude Model Function 909 Error Amplitude Gradient

Claims

1. A method of operating a high-throughput, high-resolution multi-beam charged particle microscope (1), comprising: a first image acquisition of a first image patch 17.1 in a first time interval Ts1; a second image acquisition of a second image patch 17.2 in a second time interval Ts2; a third time interval Tr during which the wafer stage (500) is moved from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch 17.2, wherein at least one of the first time interval Ts1 and the second time interval Ts2 overlap; 1. A method for operating a multi-beam charged particle microscope (1), comprising:

2. 2. The method of operating a multi-beam charged particle microscope (1) according to claim 1, wherein the second image acquisition of the second image patch 17.2 starts before the end of the third time interval Tr when the wafer stage (500) has come to a complete stop.

3. 3. A method for operating a multi-beam charged particle microscope (1) according to claim 1 or 2, wherein the third time interval Tr of wafer movement starts before the end of the time interval Ts1 when image acquisition of the first image patch 17.1 has ended.

4. A method for operating a multi-beam charged particle microscope (1) according to any one of claims 1 to 3, further comprising calculating the start time of the third time interval Tr of wafer movement in the first time interval Ts1 of image acquisition of the first image patch 17.1 so that the positional deviation of the first center position of the first image patch 17.1 from the line of sight (53) of the multi-beam charged particle microscope (1) or the movement speed of the wafer stage (500) is below a predetermined threshold.

5. 5. A method for operating a multi-beam charged particle microscope (1) according to any one of claims 1 to 4, further comprising calculating a start time of the second time interval Ts2 of the second image acquisition in the time interval Tr of wafer stage movement so that a positional deviation of the second central position 21.2 of the second image patch 17.2 from the line of sight (53) of the multi-beam charged particle microscope (1) or a movement speed of the wafer stage (500) is below a predetermined threshold.

6. predicting a series of wafer stage positions during the time interval Tr of movement of the wafer stage (500); calculating at least first and second control signals from the predicted wafer stage position; supplying the first control signal to a first deflection system (110) in a primary beam path (13) of the multi-beam charged particle microscope (1) and supplying the second control signal to a second deflection system (222) in a secondary beam path (11) of the multi-beam charged particle microscope (1); A method for operating a multi-beam charged particle microscope (1) according to any one of claims 1 to 5, further comprising:

7. A high-throughput and high-resolution multi-beam charged particle system (1), comprising: a charged particle multi-beamlet generator (300) for generating a plurality of primary charged particle beamlets (3); an object irradiation unit (100) including a first deflection system (110) for scanning a wafer surface (25) with the plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) arranged in an object plane (101) at spot positions (5) of the plurality of primary charged particle beamlets (3); a detection unit (200) comprising a projection system (205), a second deflection system (222), and an image sensor (207), for imaging the plurality of secondary electron beamlets (9) onto the image sensor (207) and for, in use, acquiring digital images of a first image patch (17.1) and a second image patch (17.2) of the wafer surface (25); a sample stage (500) with a stage motion controller, the stage motion controller comprising a plurality of motors configured to be independently controlled, the stage configured to position and hold the wafer surface (25) at the object plane (101) during acquisition of the digital images of the first image patch (17.1) and the second image patch (17.2); a plurality of detectors comprising a stage position sensor (520) and the image sensor (207), configured, in use, to generate a plurality of sensor data including position data of the sample stage (500); a control unit (800) configured, in use, to perform a first image acquisition of the first image patch (17.1) in a first time interval Ts1 and a second image acquisition of the second image patch (17.2) in a second time interval Ts2, and configured to, upon triggering of the sample stage (500) in a third time interval Tr, move the sample stage (500) from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch (17.2) such that at least one of the first time interval Ts1 and the second time interval Ts2 overlaps with the third time interval Tr; A system equipped with

8. 8. The system of claim 7, wherein the control unit is further configured to determine a start time of the third time interval Tr of wafer movement in the first time interval Ts1 of image acquisition of the first image patch 17.1 so that a deviation of the first center position of the first image patch 17.1 from the line of sight (53) of the multi-beam charged particle microscope (1) or a movement speed of the wafer stage (500) is below a predetermined threshold.

9. 9. The system of claim 7, wherein the control unit is further configured to determine a start time of the second time interval Ts2 of the second image acquisition in the time interval Tr of wafer stage movement such that a deviation of the second center position 21.2 of the second image patch 17.2 from a line of sight (53) of a multi-beam charged particle microscope (1) or a movement speed of the wafer stage (500) is below a predetermined threshold.

10. 10. The system of claim 7, wherein the control unit is further configured to predict a series of wafer stage positions during the time interval Tr of movement of the wafer stage (500), calculate at least first and second control signals from the predicted wafer stage positions, provide the first control signal to a first deflection system (110) in a primary beam path (13) of the multi-beam charged particle microscope (1), and provide the second control signal to a second deflection system (222) in a secondary beam path (11) of the multi-beam charged particle microscope (1).

11. A method of operating a high-throughput, high-resolution multi-beam charged particle system (1), comprising: a first image acquisition of a first image patch 17.1, a second image acquisition of a second image patch 17.2, and moving a wafer stage (500) from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch 17.2 all within a time interval TG; the first image acquisition of a first image patch 17.1 is in a first time interval Ts1, the second image acquisition of a second image patch 17.2 is in a second time interval Ts2, moving the wafer stage (500) from a first center position (21.1) of the first image patch (17.1) to a second center position (21.2) of the second image patch (17.2) during a third time interval Tr; A method of operating a multi-beam charged particle system (1), wherein said time interval TG is less than the sum of Ts1, Ts2 and Tr, i.e. TG<Ts1+Ts2+Tr.

12. A multi-beam charged particle microscope (1) for wafer inspection, comprising: a charged particle multi-beamlet generator (300) for generating a plurality of primary charged particle beamlets (3); an object irradiation unit (100) including a first deflection system (110) for scanning a wafer surface (25) with the plurality of primary charged particle beamlets (3) to generate a plurality of secondary electron beamlets (9) emitted from the wafer surface (25) arranged in an object plane (101) at scanning spot positions (5) of the plurality of primary charged particle beamlets (3); a detection unit (200) comprising a projection system (205), a second deflection system (222) and an image sensor (207), for imaging the plurality of secondary electron beamlets (9) onto the image sensor (207) and for, in use, acquiring digital images of a first image patch (17.1) and a second image patch (17.2) of the wafer surface (25); a sample stage (500) with a stage position sensor (520) for positioning and holding the wafer surface (25) in the object plane (101) and for moving the wafer surface from the first image patch (17.1) to the second image patch (17.2) during acquisition of the digital image of the first image patch (17.1); a plurality of detectors comprising a stage position sensor (520) and the image sensor (207), configured, in use, to generate a plurality of sensor data including position data of the sample stage (500); a first compensator in the object illumination unit (100) configured to displace or rotate the scanning spot positions (5) of the plurality of primary charged particle beamlets (3) on the wafer surface (25); a second compensator in the projection system (205) configured to compensate for the displacement or rotation of the scanning spot positions (5) of the plurality of primary charged particle beamlets (3) and to keep the spot positions (15) of the plurality of secondary electron beamlets (9) constant on the image sensor (207); a control unit (800) configured to synchronously control the first compensator in the object illumination unit (100) and the second compensator in the projection system (205) by generating a first set of control signals Cp from the plurality of sensor data during acquisition of the digital image of the first image patch (17.1) or the second image patch (17.2); A multi-beam charged particle microscope (1) equipped with:

13. 13. The multi-beam charged particle microscope (1) of claim 12, wherein the control unit (800) is configured to calculate and supply the first set of control signals Cp to the first compensator and the second compensator to compensate for changes in position or orientation of the sample stage (500).

14. 14. The multi-beam charged particle microscope (1) of claim 12 or 13, wherein the control unit (800) is configured to calculate the first set of control signals Cp and supply them to the first compensator and the second compensator to compensate for changes in the position of the line of sight (53) of the object projection unit (100).

15. 15. The multi-beam charged particle microscope (1) of claim 12, wherein the control unit (800) is configured to calculate the first set of control signals Cp and supply them to the first compensator and the second compensator to compensate for a difference between a change in position or orientation of the sample stage (500) and a change in position of a line of sight (53) of the object projection unit (100).

16. 16. The multi-beam charged particle microscope (1) according to any one of claims 12 to 15, wherein the control unit (800) is configured to calculate the first set of control signals Cp and supply them to the first compensator and the second compensator, thereby compensating for a movement speed of the sample stage (500) when acquiring the digital image of the first image patch (17.1) or the second image patch (17.2).

17. 1. A method for wafer inspection with a multi-beam charged particle microscope, comprising: a first image acquisition step of a first image patch in a first time interval Ts1; moving a wafer stage from the position of the first image patch to a second image patch during a time interval Tr; a second image acquisition step of a second image patch in a second time interval Ts2; Therefore, calculating at least a first error amplitude from a plurality of sensor signals during the first time interval Ts1; predicting a transition of the first error amplitude over at least the moving time interval Tr and the second time interval Ts2 in the first time interval Ts1; - during at least the transfer time interval Tr, supplying a control signal to a control unit of the multi-beam charged particle microscope, thereby keeping the predicted error amplitude progression below a predetermined threshold during the second time interval Ts2; A method comprising:

18. 18. The method of claim 17, wherein the prediction of the evolution of the first error amplitude is generated by a predictive model or by extrapolation.

19. 19. The method of claim 17 or 18, wherein the first error amplitude represents at least one of line of sight displacement, wafer stage displacement, wafer stage rotation, line of sight rotation, magnification error, focus error, astigmatism error, or distortion error.

20. 20. The method of any one of claims 17 to 19, wherein the control signal is supplied to a control unit of the multi-beam charged particle microscope to control components including at least one of a wafer stage, a first deflection unit, a second deflection unit, a high-speed compensator of a multi-beamlet generating unit, or a high-speed compensator of a detection unit.

21. 1. A method of operation of a multi-beam charged particle microscope with a control unit, the method comprising: during image acquisition of a series of image patches including a first image patch and a second image patch, Decomposing a data stream comprising a plurality of sensor data into a set of error amplitudes; extracting a set of drift control signals and a set of dynamic control signals; providing the set of drift control signals to a slow acting compensator; providing said set of dynamic control signals to a fast acting compensator; A method comprising a series of operational steps including:

22. 22. The method of claim 21, wherein the step of extracting the set of drift control signals and the set of dynamic control signals is performed during a time interval Ts1 of image acquisition of a first image patch, and the step of supplying the set of drift control signals to a slow motion compensator is performed during a time interval Tr of movement of a substrate by a substrate stage from the first image patch to the second image patch.

23. 22. The method of claim 21, wherein the step of providing the set of dynamic control signals to a fast acting compensator is performed during the time interval Ts1.

24. 24. The method of claim 22 or 23, wherein the step of providing the set of dynamic control signals to a fast motion compensator is further performed during a time interval Ts2 of an image scan of the second image patch.

25. The method according to any one of claims 21 to 24, further comprising the step of predicting the time course of at least one of the error amplitudes.

26. 26. The method of claim 25, comprising predicting a slow-varying drift of at least one of the error amplitudes; and predicting a fast-varying dynamic change of at least one of the error amplitudes.

27. by one or more processors of an apparatus including a charged particle source generating a plurality of primary charged particle beamlets; Determining a lateral displacement of a stage movable in at least one of the X and Y axes; determining a lateral displacement of the line of sight of the object illumination unit; instructing a controller to at least partially compensate for the lateral displacement by applying a first signal that deflects the plurality of primary charged particle beamlets incident on a sample; A non-transitory computer-readable medium comprising a set of instructions executable to cause said device to perform a method including:

Images