Method for controlling a scanning charged-particle microscope, and associated device and system

Abstract

The invention relates to a method for controlling a scanning charged-particle microscope, the method comprising: - controlling a position and / or emission or interruption of a charged particle beam 8 on a sample 10 so as to sequentially position the charged particle beam according to a plurality of positions on the sample, the method being characterized in that it comprises associating and / or programming, for each position of the beam, at least one time datum. The invention also relates to an associated device 1 and system 2.

Description

[0001] DESCRIPTION

[0002] TITLE: Control method for a charged particle scanning microscope, device and associated system.

[0003] technical field

[0004] The present invention relates to a method for controlling a charged particle scanning microscope, as well as a device or system for implementing this method.

[0005] The field of the invention is more particularly, but not limited to, that of electron, ion, imaging or lithography microscopes.

[0006] Prior art

[0007] In electron microscopy (EM), focused ion beam microscopy (FIB), or scanning probe microscopy, such as in a scanning tunneling microscope (STM) or an atomic force microscope (AFM), the probe position is controlled by scanning units. In general, these instruments have multi-dimensional multi-signal acquisition capabilities at each point of the scan. The signals acquired as a function of the probe position can be scalar signals, as in bright-field and dark-field ring imaging (BF and ADF, respectively); a one-dimensional signal (spectrum), as in electron energy loss spectroscopy (EELS) or scanning tunneling spectroscopy (STS); or a multi-dimensional signal, such as that obtained in the case of four-dimensional scanning electron microscopy (4D-STEM) techniques.The probe's position and the acquisition of the multidimensional signal are synchronized in time, which is now achieved using digital logic, for example, by using a field-programmable gate array (FPGA) or an application-specific integrated circuit (ASIC). The interface between the probe microscopes and the digital logic is via a set of converters, which can be analog-to-digital (ADC) or digital-to-analog (DAC).

[0008] The standard acquisition strategy is sequential rasterization, where the probe typically moves in lines from top to bottom and left to right within a line. In this scanning mode, signal acquisition is often interrupted while the probe returns to the beginning of the line. This time is called the return time. It is also possible to sample the return of the line to avoid the return time (although other distortions appear), which is commonly called a serpentine scan, as shown in Figure 1. The hardware design of this system is trivial, as, for example, a pair of n-bit counters controlled by a new pixel or a new line edge is sufficient for this application. We then have a typical generative scan using two counters, following the traditional scanning method on the left of Figure 1 and the serpentine scan in the center of Figure 1.The edges of pixels and lines are generated by a common clock. The entire scanning engine uses 2n bits for two n-bit DACS, as illustrated on the right of Figure 1.

[0009] In parallel, electron detection with pixel-based detectors has been revolutionized over the last decade by the advent of more sensitive and faster detectors. Furthermore, the development of event-based electron detectors, such as the Timepix, represents a new paradigm in data acquisition. In these detectors, electron strikes are recorded individually as they arrive, with a temporal resolution of a few nanoseconds. Consequently, the scanning speed is no longer limited by the detector's acquisition speed, but by the fundamental limitations of the scanning system.To fully exploit the capabilities of event-driven detectors in the context of scanning devices, the traditional sequential frame is unsuitable because these patterns are highly asymmetrical in the X and Y directions and also exhibit inappropriate voltage ramps, thus hindering their ultimate scanning speed potential. Other scanning strategies, such as Lissajous scanning, can reach the physical limits of the scanning system, but they suffer from their own problems, such as inhomogeneous dose distribution. Currently, no scanning unit exists that can reach the analog limits of the probe positioning system by implementing new scanning strategies and that can interface with event-driven detectors.The aim of the present invention is to solve at least one of the problems posed by the state of the art, namely to propose a method, device or control system for a charged particle scanning microscope:.

[0010] - Enabling faster scanning speed, and / or

[0011] - Enabling better homogeneity or better control of the dose of charged particles, and / or

[0012] - Enabling the analog limits of the probe positioning system controlling the charged particles to be reached, and / or

[0013] - Enabling better quality of collected measurement data or better image quality obtained, particularly without measurement artifacts, and / or

[0014] - capable of interfacing with event-based detectors.

[0015] Description of the invention

[0016] This objective is achieved with a method for controlling a charged-particle scanning microscope, comprising:

[0017] - A control of a position and / or an emission or interruption of a beam of charged particles on a sample, so as to sequentially position the beam of charged particles according to several positions of the beam on the sample the process being characterized in that it includes an association and / or programming, for each position of the beam, of at least one time data.

[0018] The method according to the invention may include:

[0019] - the reception of measurement data from at least one detector capturing this measurement data from the sample, and / or

[0020] - an assembly of measurement data corresponding to the different positions of the beam, preferably in a measurement image, at least one time data being preferably used during command, reception and / or assembly.

[0021] At least one time data point may include at least one time offset from a command instant on a considered beam position.

[0022] The at least one offset time may include a time offset, called measurement offset, between a time of receiving measurement data from the at least one detector and a time of command on the position of the beam considered, this measurement offset may be common or specific to each detector considered.

[0023] The control of the position and / or emission or interruption of the beam of charged particles on the sample may include a beam control along a first axis and a beam control along a second axis distinct from the first axis, and the at least one time delay may include a time delay, called a control delay ('delay, delay'), between a start time of the beam control along the first axis and a start time of the beam control along the second axis when controlling the beam position in question.

[0024] At least one time lag:

[0025] - can be specific to each beam position and programmable or individually associated with each beam position according to a value specific to that beam position, potentially different from the values ​​programmed or associated with other beam positions, or

[0026] - can have a value common to all positions of the beam.

[0027] At least one time datum may include an acquisition time (tacq) during which the charged particle beam reaches or irradiates the sample at the command on the beam position considered.

[0028] At least one time data point may include a pause time (tpause) during which the charged particle beam does not reach or irradiate the sample when the beam position is commanded.

[0029] The exposure time may be longer on a peripheral area of ​​the image than on a central area of ​​the image.

[0030] At least one time data point can be specific to each beam position and programmable or individually associatable to each beam position according to a value specific to that beam position potentially different from the values ​​programmed or associated with other beam positions.

[0031] The method according to the invention may include programming or associating the tpause / tacq ratio, or a variable proportional to or dependent on tpause / tacq, individually with each beam position according to a value specific to that beam position, potentially different from the values ​​programmed or associated with other beam positions. The method according to the invention may further include associating and / or programming, for each beam position, the activation or deactivation of at least one trigger data point used during command, reception, and / or assembly, from among:

[0032] - a time-based trigger input (Trigin), preferably from at least one detector, and preferably to trigger the position command or the start of the pause time or acquisition time, and / or

[0033] - a time-based trigger output (Trig ou (t), preferably for the attention of at least one detector, and preferably for a trigger triggered at the end of the position command or pause time or acquisition time.

[0034] The control of the position and / or emission or interruption of the beam of charged particles on the sample may include control according to several positions of the beam successively distributed along a trajectory parameterized by at least two sinusoids or substantially sinusoidal functions.

[0035] The sample can be a reference sample and the method according to the invention can include, for a given trajectory of successive beam positions, an optimization of at least one time data point of each beam position and then a storage in a memory, and in association with this given trajectory, of at least one time data point of each beam position.

[0036] The association and / or programming, for each beam position, of at least one time data point, can be carried out at least by programmable digital means, and the method according to the invention may include a conversion by an analog-to-digital converter between:

[0037] - programmable digital means and at least one detector and / or

[0038] - between programmable digital means and physical means for deflecting and / or emitting or interrupting the beam of charged particles. According to yet another aspect of the invention, a control device for a charged-particle scanning microscope is proposed, comprising:

[0039] - means for storing or receiving beam positions

[0040] - at least one control output arranged and / or programmed to send, to the attention of physical means for controlling the charged particle beam, a command for a position and / or an emission or interruption of the charged particle beam,

[0041] - Electronic and / or computer processing means, the processing means preferably being arranged and / or programmed to associate and / or program, for each position of the beam, at least one time data point.

[0042] Furthermore :

[0043] - the device according to the invention may include a receiver arranged for receiving measurement data from at least one detector capturing this measurement data, and / or the processing means may be arranged and / or programmed for assembling the measurement data corresponding to the different positions of the beam, preferably into a measurement image, the processing means being preferably arranged and / or programmed to exploit at least one time data during the command, reception and / or assembly.

[0044] At least one time data point may include at least one time offset from a command instant on a considered beam position.

[0045] At least one offset time may include a time offset, called a measurement offset between a moment of receiving measurement data from at least one detector and a moment of command on the position of the beam considered, this measurement offset may be common or specific to each detector considered.

[0046] The control of the position and / or emission or interruption of the beam of charged particles on the sample may include a beam control along a first axis and a beam control along a second axis distinct from the first axis, and the at least one time delay may include a time delay, called a control delay ('delay, delay'), between a start time of the beam control along the first axis and a start time of the beam control along the second axis when controlling the beam position in question.

[0047] At least one time difference:

[0048] - can be specific to each beam position and programmable or individually associated with each beam position according to a value specific to that beam position, potentially different from the values ​​programmed or associated with other beam positions, or

[0049] - can have a value common to all positions of the beam.

[0050] At least one time datum may include an acquisition time (tacq) during which the charged particle beam reaches or irradiates the sample at the command on the beam position considered.

[0051] At least one time data point may include a pause time (tpause) during which the charged particle beam does not reach or irradiate the sample when the beam position is commanded.

[0052] The exposure time may be longer on a peripheral area of ​​the image than on a central area of ​​the image.

[0053] At least one time data point can be specific to each beam position and programmable or individually associatable to each beam position according to a value specific to that beam position potentially different from the values ​​programmed or associated with other beam positions.

[0054] The processing means can be arranged and / or programmed for programming or association of the tpause / tacq ratio or a variable proportional to tpause / tacq or which depends on tpause / tacq, individually at each beam position according to a value specific to that beam position potentially different from the values ​​programmed or associated with other beam positions.

[0055] The processing means can be arranged and / or programmed for association and / or programming, for each beam position, of the activation or deactivation of at least one trigger data used during command, reception and / or assembly, from among:

[0056] - an input trigger time data (Trigin), preferably from at least one detector, and preferably to trigger the position command or the start of the pause time or acquisition time, and / or - an output trigger time data (Trig ou (t), preferably for the attention of at least one detector, and preferably for a trigger triggered at the end of the position command or pause time or acquisition time.

[0057] The control of the position and / or emission or interruption of the beam of charged particles on the sample may include a control according to several positions of the beam successively distributed along a trajectory parameterized by at least two sinusoids or substantially sinusoidal functions.

[0058] The sample can be a reference sample and the device according to the invention can include, for a given trajectory of successive beam positions, an optimization of at least one time data point of each beam position and then a storage in a memory, and in association with this given trajectory, of at least one time data point of each beam position.

[0059] The processing means may include programmable digital means arranged and / or programmed for the association and / or programming, for each position of the beam, of at least one time data point, and a converter arranged and / or programmed for analog / digital conversion between the programmable digital means and at least one detector and / or between the programmable digital means and magnetic and / or physical means for deflecting and / or emitting or interrupting the beam of charged particles.

[0060] According to yet another aspect of the invention, a control system is proposed, comprising:

[0061] - a control device according to the invention, and

[0062] At least one of the following: o Means for generating the charged particle beam, o Physical means for controlling the charged particle beam, these means being connected to at least one control output of the control device, o At least one detector arranged to capture measurement data from a sample, Description of figures and embodiments

[0063] Other advantages and features of the invention will become apparent upon reading the detailed description of implementations and embodiments, which are by no means limiting, and the following attached drawings:

[0064] [Fig. 1] Figure 1 illustrates scans according to the state of the art, traditional on the left and serpentine in the center, with the structure of the n-bit and DAC on the right, [Fig. 2] Figure 2 is a schematic view of a first embodiment of device 1 and system 2 according to the invention, which is the preferred embodiment of the invention,

[0065] [Fig. 3] Figure 3 illustrates the storage or reception means 11, the processing means 13, the programmable digital means 131 (without the conversion means 132, not illustrated in this figure) of the first embodiment of device 1 and system 2 according to the invention,

[0066] [Fig. 4] Figure 4 illustrates different delays in system 2, T deiay being obtained by delaying the reception of data from detector 9, while T' deiay and T”deiay are managed by delaying the analog signals sent to the scanning deflectors 4, 5,

[0067] [Fig. 5] Figure 5 is an illustration of images obtained by a Lissajous scan in the first embodiment of system 2,

[0068] [Fig. 6] Figure 6 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous scan (sinusoidal scan), the deflectors of means 4, 5 operating at approximately 120 kHz, and with an optimal setting of T deiay , T' delay , and T” deiay ,

[0069] [Fig. 7] Figure 7 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous scan (sinusoidal scan), the deflectors of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of de(ay which has an effect on the X and Y axes,

[0070] [Fig. 8] Figure 8 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous scan (sinusoidal scan), the deflectors of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of T" deiay , which only affects the Y-axis, [Fig. 9] Figure 9 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous scan (sinusoidal scan), the deflectors of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of T' deiay which only has an effect on the X-axis,

[0071] [Fig. 10] Figure 10 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a standard scan from top left to bottom right as illustrated on the left side of Figure 1, and with a customization of the pause time tpause depending on the position of the beam 8, in particular with a longer pause time tpause on the left edge of the image,

[0072] [Fig. 11] Figure 11 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a standard scan from top left to bottom right as illustrated on the left side of Figure 1, and, compared to Figure 10, a uniform pause time tpause for all beam positions 8 and therefore over the entire image, and [Fig. 12] Figure 12:

[0073] - in its parts A and B, illustrates a comparison of Lissajous scanning with frame detectors (A) and event detectors (B). Lissajous scanning can be associated with frame detectors, but the minimum exposure time of these detectors limits the maximum achievable frequency of the pattern. When a nanosecond resolution electron detector is used, higher frequencies become possible. In the image-based case, the pixel residence time is 30 times longer than in the event-based case, as also indicated by the radius of the scattering sphere. This shows how the use of detectors 9 whose clock frequency is equal to or greater than the clock frequency of the converters 132 of the scanning unit 1 thus allows arbitrary scanning schemes.

[0074] Parts C, D, and E illustrate an example of this system applied to event-driven EELS spectroscopy. This involves hyperspectral imaging with a Lissajous scan and a Timepix3 event detector. Part C: ADF image of the region of interest. The Lissajous scan is spatially inhomogeneous, and the deposited dose is higher in the corners than in the center of the image. Part D: Spectrum associated with the region of interest, marked 2 in Part E. Part E: Hyperspectral image integrated over the energy window between 185 and 190 eV, including the n* signature of h-BN. The viability of spectroscopy with an event detector is thus demonstrated using the sample illustrated in (C). Since the scan is synchronized with the detector, the reconstruction of a hyperspectral image becomes possible, as shown by the results in (D) and (E).

[0075] [Fig. 13] Figure 13 compares Lissajous scanning (top line of this figure) with normal or conventional raster scanning (bottom line of this figure). Lissajous scanning enables fast, multi-resolution imaging. As the scan progresses, spatial resolution improves through the gradual acquisition of new sampling points. The imaging process thus transitions from sparse to dense sampling, allowing the application of conventional reconstruction techniques from incomplete data, such as interpolation or inpainting, from the earliest acquisition phases. In both scanning modes, the total dose deposited is identical, but due to the flyback time, conventional raster scanning is approximately 30% slower. Lissajous scanning operates at a frequency of approximately 1 kHz. The scale bars represent 100 nm.This illustrates how it is possible to have images of the sample before the scanning system finishes an entire cycle ("compressed sensing").

[0076] These embodiments are not exhaustive; in particular, variants of the invention may be considered that comprise only a selection of features described or illustrated hereafter, isolated from the other described or illustrated features (even if this selection is isolated within a sentence including these other features), provided that this selection of features is sufficient to confer a technical advantage or to differentiate the invention from the prior art. This selection includes at least one preferably functional feature without structural details, and / or with only a portion of the structural details if this portion alone is sufficient to confer a technical advantage or to differentiate the invention from the prior art.We will first describe, with reference to figures 2 to 11, a first embodiment of device 1 (also called scanning unit) and system 2 according to the invention of control of a scanning microscope with charged particles (typically electron or ion), and a first embodiment of process implemented by this device 1 or system 2.

[0077] Control system 2 includes:

[0078] - the control device 1,

[0079] - means 3 for generating the beam 8 (also called probe) of charged particles, such as, for example, means for generating an electron beam, for example, a generator of the thermionic emission type, Schottky emission, or cold field emission,

[0080] - means (preferably magnetic and / or electrostatic) 4, 5, 6 for controlling the charged particle beam 8, these means being connected to at least one control output 7 of the control device 1 (more precisely, means 13); these physical control means typically include: o means 4 for controlling the position of the charged particle beam 8 along a first X-axis, these means 4 typically comprising at least one deflector (for example, a pair of solenoids generating a magnetic field in Figure 2), the voltage and / or current supplying these means 4 being controlled by at least one control output 7, the deflectors of the means 4 typically comprising a pair of solenoids from a commercial electron microscope and driven by the means 13 with current and / or voltage over a range corresponding, according to the data of the electron microscope manufacturer,to a deflection of the beam 8 of a few angstroms to a few tens of micrometers about the X-axis, o means 5 for controlling the position of the charged particle beam 8 along a second axis Y, distinct from (preferably perpendicular to) the first axis X, these means 5 typically comprising at least one deflector (for example, a pair of solenoids generating a magnetic field in Figure 2), the voltage and / or current supplying these means 5 being controlled by at least one control output 7, the deflectors of the means 5 typically comprising a pair of solenoids from a commercial electron microscope and driven by the means 13 with current and / or voltage over a range corresponding, according to the data of the electron microscope manufacturer, to a deflection of the beam 8 of a few angstroms to a few tens of micrometers about the Y-axis, o A beam blocker or shutter 6, arranged to interrupt the beam 8 on command from the device 1,this blocker 6 being controlled by at least one control output 7; one can, for example, use an electrostatic or magnetic type shutter 6; the blocker 6 is preferably an electrostatic blocker called a "blanker" in English (typically of a commercial electron microscope) in the form of a capacitor subjecting the beam 8 to an electric field gradient,

[0081] - at least one detector 9 arranged to capture measurement data from a sample 10 subjected to the beam 8, for example of the type a photomultiplier tube, a pixelated detector, a photodiode, and / or a CCD camera.

[0082] Beam 8 is, for example, an accelerated electron beam, for example at 100 kV.

[0083] The control device 1 for a charged particle scanning microscope (typically electron or ion) comprises:

[0084] - electronic and / or computer means for storing or receiving beam positions 8,

[0085] - at least one control output 7 arranged and / or programmed to send, to the attention of the physical control means 4, 5, 6 of the charged particle beam 8, a command for an emission or interruption / pause (using the blocker 6) of the charged particle beam 8 on the sample 10 and / or a position of the charged particle beam 8 on the sample 10 (for example by controlling the voltage or current individually for each of the means 4 and / or 5 or each of the solenoids of the means 4 and / or 5) so as to sequentially position the charged particle beam 8 according to the several positions of the beam 8 on the sample 10,

[0086] - electronic and / or computer processing means 13. The processing means 13 are preferably arranged and / or programmed to associate and / or program, for each position of the beam 8, at least one time data point.

[0087] Device 1 includes a receiver 12 arranged for receiving measurement data from at least one detector 9 capturing this measurement data from the sample 10. The receiver 12 is, for example, of the type of any analog or digital input port of the means 13 and / or the analog / digital converter 132.

[0088] The processing means 13 are arranged and / or programmed to assemble the measurement data corresponding to the different positions of the beam 8, preferably into a measurement image. This image is not necessarily a visualized or displayed image, but can simply be in the form of a grouping of measurement data, preferably gathered into a single computer file.

[0089] The processing means 13 are arranged and / or programmed to exploit at least one temporal data point during ordering, receiving and / or assembly.

[0090] In this description, the "position of beam 8" may also be called a "pixel," but not in the sense of a matrix of one pixel. Indeed, the positions of beam 8 are not necessarily distributed in a matrix fashion. Two positions of beam 8, whether temporally successive or not, may overlap spatially.

[0091] Each of the means of device 1 and system 2 are technical means.

[0092] The storage or receiving means 11 comprise at least one computer, a central processing unit or computing unit, an analog electronic circuit (preferably dedicated), a digital electronic circuit (preferably dedicated), a memory, and / or a microprocessor (preferably dedicated), and / or software means. Preferably, the means 11 consist solely of a memory.

[0093] The processing means 13 include at least one computer, central processing unit or computing unit, analog electronic circuit (preferably dedicated), digital electronic circuit (preferably dedicated), and / or microprocessor (preferably dedicated), and / or software means.

[0094] The processing means 13 include programmable digital means 131 arranged and / or programmed for the association and / or programming, for each position of the beam 8, of at least one time data point. These programmable digital means 131 typically include a "field-programmable gate array" (FPGA), for example, the Artix series manufactured by Xilinx Inc., Artix-7 series.

[0095] With reference to Figure 3, memory 11 is external to the FPGA, typically one or more DDRs external to the FPGA, but in a variant memory 11 is or includes internal memories on the FPGA chip.

[0096] The processing means 13 include a converter 132 arranged and / or programmed for analog-to-digital conversion (analog to digital and digital to analog) between the programmable digital means 131 and at least one detector 9 and / or between the programmable digital means 131 and magnetic and / or physical means

[0097] 3, 4, 5, 6 of deflection and / or emission or interruption of the beam 8 of charged particles. This converter 132 includes, for example, at least one chip manufactured by Texas Instruments or Analog Devices (AD), for example Analog Devices AD9764 (14-bit 100 MHz DAC) and / or AD9225 (12-bit 25 MHz ADC).

[0098] The means 13 control the analog voltage of the coils of the means

[0099] 4, 5.

[0100] Means 3, 4, 5, 6 and 9 typically constitute the microscope of system 2.

[0101] Converter 132 (analog to digital or digital to analog) connects microscope 3, 4, 5, 6 and 9 to digital electronics 11, 13, 131.

[0102] The first embodiment of the method according to the invention for controlling a charged-particle scanning microscope, implemented by device 1 or system 2, comprises as follows:

[0103] - A command, by means 13 (sending commands to means 3, 4, 5 and / or 6), for a position and / or an emission or interruption of the beam 8 of charged particles onto the sample 10, so as to sequentially position the beam 8 of charged particles according to several positions of the beam 8 on the sample 10

[0104] - an association and / or programming, by means 13, 131, of at least one time data for each position of the beam 8.

[0105] The first embodiment of the method according to the invention comprises: - receiving, by means 13 via receiver 12, measurement data from at least one detector 9 capturing this measurement data from the sample 10 subjected to the beam 8,

[0106] - an assembly, by means 13, 131 of the measurement data corresponding to the different positions of the beam 8, preferably in a measurement image, at least one time data being exploited during the command, reception and / or assembly by means 13.

[0107] Each position of beam 8 is decomposed into a component along the X direction and a component along the Y direction.

[0108] The at least one time data includes at least one offset time preferably relative to an instant of the command on a position of the beam 8 considered.

[0109] At least one time offset includes a time offset, said between a moment of receiving measurement data from at least one detector and a moment of command on the position of the beam 8 considered, this measurement offset may be common or specific to each detector considered.

[0110] The first embodiment of the method according to the invention comprises associating, by means 13 and in memory 11 (which are arranged and / or programmed for this purpose), the trajectory of the chosen beam 8 (in particular the position of the beam considered and the times tacq and tpause) and the measurement data obtained T deiayAfter the command is sent to a given beam position 8, preferably for the purpose of assembling the measurement data into the measurement image, deiay takes into account that the precise moment of a command to a given beam position 8 does not necessarily correspond to the same moment at which the data corresponding to that beam position 8 will be acquired by at least one detector 9; in particular, there may be a latency and a phase difference that depends on the beam positions 8 and the chosen tacq and tpause times. Thus, to increase speed, it is necessary to take this latency and time difference into account without slowing down the exploration rate of the different beam positions 8, while improving the quality of the image or the set of measurement data obtained.

[0111] TThe delay is typically greater than 1 ns and / or less than 1 ms. The control of the position and / or emission or interruption of the beam 8 of charged particles onto the sample 10 comprises a control of the beam 8 along a first X-axis and a control of the beam 8 along a second Y-axis distinct from the first X-axis and preferably perpendicular to the first X-axis, and the at least one delay time includes a time delay, called the control delay ('deiay-T' delay, between a start time of the control of the beam 8 along the first X-axis for the position of the beam 8 considered and a start time of the control of the beam 8 along the second Y-axis for this same position of the beam 8 considered. The control of the position of the beam 8 considered is directly programmable via the means 131. is the time lag between the start of the reading, by means 13, of memory 11 for a given position of beam 8 and the sending by output 7 of the analog command of means 4 (along the first X axis) corresponding to this given position of beam 8. is directly programmable via means 131.

[0112] ^"deiay is the time lag between the start of the reading, by means 13, of memory 11 for a given position of beam 8 and the sending by output 7 of the analog control of means 5 (along the second Y axis) corresponding to this given position of beam 8.

[0113] The setting of T' deiay and T" deiay therefore allows T to be adjusted" deiay - T' delaywhich is the time lag, called the control lag, between the start time of the control of beam 8 along the first X axis and the start time of the control of beam 8 along the second Y axis when the control is applied to the position of beam 8 considered.

[0114] ^deiay and T” de iay take into account that means 4 and 5, combined with means 13, will not necessarily have the same frequency responses, which consequently results in different time responses (due, for example, to differences in the inductances and / or resistances of means 4 and 5, or to differences between the analog circuits driving means 4 and 5). There may be a difference or latency between the two. Thus, to speed things up, allow this difference to be taken into account without slowing down the exploration rate of the different positions of beam 8, while improving the quality of the image or the set of measurement data obtained. Delay is typically greater than 1 ns and / or less than 1 ms. i:" Delay typically greater than 1 ns and / or less than 1 ms.

[0115] The difference r" de(ay - T' de iay is typically greater than 0 ns and / or less than 500 ps.

[0116] At least one time difference:

[0117] ■ ^deiay, and / or T” deiay preferably has a value common to all positions of beam 8. Indeed, the values ​​of r de(ay , v'deiay and T” deiaytake into account the differences (particularly in reaction time) or analog defects of the different coils of means 4, 5 and / or at least one detector 9. A priori, these differences or defects or latencies depend intrinsically on means 4, 5, 9 but not or only slightly on the positions of the beam 8, or

[0118] ■ T deiay, and / or T” deiay In some cases, it can be specific to each position of the beam 8 and programmable or individually associatable by means 13 to each position of the beam 8 according to a value specific to that position of the beam 8 potentially different from the values ​​programmed or associated with other positions of the beam 8. This can be particularly useful for forming asymmetrical scan curves (X axis and Y axis very different) or curves with multiple frequency components.

[0119] Figure 6 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous sweep (sinusoidal sweep), the coils of means 4, 5 operating at approximately 120 kHz, and with an optimal setting of T deiay , T' delay, and T" deiay In Figure 6 and the following Figures 7 to 11, the imaged sample 10 is always the same gold cross, 30 nm thick and approximately 400 nm on each side, on a SisIXU substrate, with its axes substantially oblique to the horizontal in these figures. We have T- d elay 2800 150 islands.

[0120] For each of images 5 to 11, the system includes:

[0121] - As means of generation 3: cold field electron source;

[0122] - As means 4: deflectors comprising a pair of solenoids as previously described;

[0123] - As means 5: deflectors comprising a pair of solenoids as previously described; - As blocker 6: electrostatic deflector as previously described;

[0124] - As detector 9: ADF detector (or annular dark field for "Annular Dark Field");

[0125] - As sample 10: Gold cross with SisIX substrate;

[0126] - Like FPGA 131: Xilinx Inc. Artix-7 range;

[0127] - As a 132 converter: Analog Devices AD9764 (14-bit 100 MHz DAC) and AD9225 (12-bit 25 MHz ADC);

[0128] Figure 7 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous sweep (sinusoidal sweep), the coils of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of de(ayThis has a detrimental effect on the X and Y axes. In this example, the difference between the optimal and non-optimal settings is 200 ns.

[0129] Figure 8 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous sweep (sinusoidal sweep), the coils of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of T". deiay , which only has a detrimental effect on the Y axis. The difference between the optimal and non-optimal settings is 600 ns in this example.

[0130] Figure 9 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a Lissajous sweep (sinusoidal sweep), the coils of means 4, 5 operating at approximately 120 kHz, and with, compared to Figure 6, a non-optimal setting of T' deiayThis only has a detrimental effect on the X-axis. The difference between the optimal and non-optimal settings is 600 ns in this example.

[0131] The at least one time data includes an acquisition time tacq during which the beam 8 of charged particles reaches or irradiates the sample 10 at the command on the position of the beam 8 considered.

[0132] At least one time-domain parameter includes a pause time tpause (which can be zero everywhere) during which the beam 8 of charged particles does not reach or irradiate the sample 10 when the beam position 8 is commanded, typically by blocking the beam 8 (for example, by the blocker 6, which, depending on its position, can either block or allow the beam 8 to pass). tacq is specific to each beam position 8 and is programmable or individually associated with each beam position 8 according to a value specific to that beam position 8, potentially different from the values ​​programmed or associated with other beam positions 8. tpause is specific to each beam position 8 and is programmable or individually associated with each beam position 8 according to a value specific to that beam position 8, potentially different from the values ​​programmed or associated with other beam positions 8.

[0133] The sum of tacq programmed or associated with a position of the beam 8 considered and tpause programmed or associated with this same position of the beam 8 considered corresponds to the total time spent by the system 2 on this position of the beam 8.

[0134] Thus, the first embodiment of the method according to the invention comprises programming or associating, by means 13, the ratio tpause / tacq or a variable proportional to tpause / tacq or dependent on tpause / tacq, individually to each position of the beam 8 according to a value specific to that position of the beam 8 potentially different from the values ​​programmed or associated with other positions of the beam 8. tacq is typically greater than 0 ns and / or less than 512 ps. tpause is typically greater than 0 ns and / or less than 512 ps. tpause / tacq is typically greater than 0 and / or less than 100.

[0135] Figure 10 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a standard scan from top left to bottom right as illustrated on the left part of Figure 1, and with a customization of the pause time tpause according to the position of the beam 8, in particular with a pause time tpause of 30 ps (microseconds) on the left edge of the image while tpause is zero on the rest of the image, tacq remains constant during the acquisition, which therefore has a tpause / tacq ratio between 0 and 30.

[0136] Figure 11 illustrates an image obtained by the first embodiment of system 2 according to the invention, with a standard scan from top left to bottom right as illustrated on the left side of Figure 1, and, compared to Figure 10, a uniform pause time tpause, equal to zero, for all positions of the beam 8 and therefore over the entire image. Figure 11 shows the detrimental effect of this uniformity of tpause by the appearance of a white band on the left edge of the image, which has no actual presence in the sample 10 but is merely an imaging artifact.

[0137] The first embodiment of the method according to the invention further comprises an association and / or programming, by means 13, and for each position of the beam 8, of the activation or deactivation of at least one trigger data used during command, reception and / or assembly, from among:

[0138] - a Trigin input trigger time data, preferably from at least one detector (not necessarily detector referenced as 9), and preferably to trigger the position command or the start of the pause time tpause or the acquisition time tacq, and / or

[0139] - a trigger time data output Trig ou t, preferably: o for the attention of at least one detector (not necessarily detector referenced 9), and preferably for a trigger triggered at the end of the position command or the pause time tpause or the acquisition time tacq t OR o directed to another part of system 2, to trigger another action (laser pickup, synchronization with other detectors, etc.)

[0140] The control of the position and / or emission or interruption of the beam 8 of charged particles on the sample 10 typically includes, but is not limited to, control according to several positions of the beam 8 successively distributed along a trajectory parameterized by at least two sinusoids or substantially sinusoidal functions, for example of the Lissajous type.

[0141] Depending on the variant considered, the pause time tpause may be longer in a peripheral area of ​​the image than in a central area. This is particularly relevant in the case of a Lissajous-type command.

[0142] Figure 4 illustrates a Lissajous sweep model using an LR circuit (containing a resistor and an inductor) as a model, with an L / R ratio of 5.10' 6s. The voltage supplied by the DAC (digital-to-analog converter), the current, and the voltage of each coil of means 4 or 5 are indicated. For a 10 kHz scan under these conditions, a delay appears between the current and the supplied voltage. The increase in coil impedance explains this characteristic, which is corrected by the first embodiment of the method according to the invention. This also explains why the image field of view must decrease during a rapid scan.

[0143] As illustrated in Figure 4, the scan frequency and the L / R ratio are highlighted on each of its parts:

[0144] - on its part A: For low frequency scans, the delay is mainly caused by the desynchronization between the reception of the data and the current / voltage value put on the deflectors 4, 5.

[0145] - on its part B: For high-frequency sweeps, the current and voltage become desynchronized, requiring an adjustment of T deiay accordingly.

[0146] - on its part C: The scanning trajectories may differ due to variations in frequency, impedance, and other factors. To correct these imperfections, it is necessary to adjust T' deiay and T" deiay .

[0147] Figure 5 is an illustration of images obtained with a Lissajous scan by system 2:

[0148] - Part A of Figure 5: The Lissajous scanning frequency is progressively increased from 3 kHz to 175 kHz. The adjustable delay changes by approximately 1 ps between the two. Also noted is the reduction in the field of view of microscopes 3, 4, 5, 6, and 9 due to the higher impedance of each coil in means 4 or 5 during high-frequency scans. The circular mask is due to the complete suppression of beam positions 8 near the corners of the image.

[0149] - Part B of Figure 5: a similar principle is applied, but for a list following a Lissajous scheme using triangular waves. Although this scan is more uniform in terms of dose, the delay correction is not as effective, which creates significant distortions in the image.

[0150] This first embodiment of the method according to the invention is implemented on samples 10 on which measurements are to be performed. However, the first embodiment of the method according to the invention is preferably implemented beforehand with a target or a reference sample 10, and this prior implementation of the first embodiment of the method according to the invention comprises, for a given trajectory of successive beam positions 8, an optimization (manual, or by means known to those skilled in the art such as automatic software optimization varying the time value or each of at least one temporal data point by optimizing the quality of the image obtained, or such as solutions based on artificial intelligence) by means 13 of at least one tpause, and / OR tpause / tacq, of each position of the beam 8 then a storage in memory 11 (and in association with this given trajectory) of at least one time data of each position of the beam 8.

[0151] This reference sample is, for example, a Magnification Reference Standard (MRS).

[0152] The association and / or programming, for each position of the beam 8, of at least one time data, is carried out at least by programmable digital means 131.

[0153] The 132 converter implements analog-to-digital and digital-to-analog conversion between:

[0154] - programmable digital means 131 and at least one detector

[0155] 9 often analog and / or

[0156] - between the programmable digital means 131 and magnetic and / or physical means 3, 4, 5, 6 for deflection and / or emission or interruption of the beam 8 of charged particles, for example for the control of the analog voltage of the coils of the means 4,5.

[0157] Thus, the first embodiment described above proposes a new implementation of the scanning unit logic, which can significantly increase the maximum achievable scanning speeds by allowing fully customizable arbitrary scanning strategies. It is based on the principle of list scanning: a set of information is defined, by beam position 8, in the memory 11 of device 1 by the control unit 13 of system 2 (typically comprising a user-programmed computer), and this information is read sequentially as soon as the new beam position flag 8 is ready. This has the advantage of being adaptable and not requiring a complete replacement of existing units.The only requirement is the presence of sufficient available memory in the digital system, for example in the form of readily available double data rate memory (DDR) or block random access memory (BRAM), capable of storing the number of beam positions.

[0158] 8 or desired probe points.

[0159] The proposed scanning unit 1 architecture is based on lists for microscope setup and data acquisition. The scanning list is stored in memory 11 and provides instructions for the beam position 8 behavior, its acquisition time tacq, and its pause time tpause. Similar logic exists for image measurement data acquisition.

[0160] For each position of beam 8, a desired number of bits of information is read from memory 11, comprising at least:

[0161] - the component of the position of beam 8 along the X axis, the component of the position of beam 8 along the Y axis, the acquisition time tacq, the suppression / pause time tpause,

[0162] - optionally the Trigin logic trigger input, the Trig trigger output ou t and

[0163] - the logical end of the EOF frame (which signals a circular buffer to read again from the first position of beam 8), as shown in Figure 3. For each new value read, a unique counter runs until the sum of the acquisition time tacq and the pause time tpause has elapsed, which triggers the reading of the information of the next beam 8 position in memory 11.

[0164] Thus, the first embodiment of the process according to the invention possesses unique characteristics, allowing full exploitation of all the capabilities of system 2 and in particular of the microscope 3, 4, 5, 6, 9: a) the addition of both the acquisition time tacq and the pause time tpause allows for non-uniform and dose-controllable residence times of the beam positions 8; b) the addition of the Trig output out and the Trigin beam position trigger input 8, allows synchronization with other hardware; c) other parameters, associated and customized in memory 11 for each beam position 8, can be added (illustrated in Figure 3 by the "Extra" parameter) at the cost of increased use of memory 11 of the digital system; d) the same list scanning principle is applied to the signals of at least one detector 9. For each beam position 8, multiple analog-to-digital conversions take place by the converter 132. Traditional scanning units sum (or average) all these values ​​in the same beam position 8, which minimizes Poisson noise and increases the signal-to-noise ratio by Vn, where n is the number of samples, and stores the net value in the memory channels.This moving average filter, however, performs poorly in the frequency domain, following a sine function (the Fourier transform of a finite square wave). In the case of the first embodiment of the method according to the invention, for each image sampling, a configurable weight is added to memory 11. By shaping the intensity in the time domain, it is possible to obtain a desired response in the frequency domain. This image list principle is also used to access the temporal response of a given set of microscopes by sampling a single conversion at different elapsed times. With this configurable weight, one can have a temporal resolution given by the sampling rate of the ADC (analog-to-digital converter), for example 25 MHz or 40 ns; e) a configurable delay between the image signal and data storage, and between the position and the control output also performed (r. de(ay, delay, et i:" d(eiay in Figure 3). This is important to account for the delay in the electronics, but also to correct the phase shifts induced by each coil of the means 4, 5. More generally, it is important to account for any physical delay that may occur during acquisition (variations in the velocity of the primary and secondary particles between settings, for example). During a fast scan, the impedance of each scanning coil increases, which creates a time lag between the applied voltage and the current, but this lag can be compensated for by the controllable delay, as shown in Figure 4; f) after reading a new position of beam 8 into memory 11, it is possible to overwrite its value on the fly, which means that the next scan image is transformed relative to the previous one.This has potential applications in dose-sensitive applications where the scanning pattern is not only arbitrary, but where each new scan is different from the previous one without software intervention. In this case, the invention can exploit a Lissajous scan with irrational frequency ratios. The Lissajous previously described used rational frequency ratios, for example, 175 kHz and 174.9 kHz on X and Y axes. If irrational ratios are used, such as 100(2) kHz and 140 kHz, the Lissajous never closes, meaning that the beam does not pass over the same point twice, which can lead to applications for dose control on samples.

[0165] The first embodiment of the method, device, or system according to the invention described above can be used for multiple applications, including, but not limited to: a) traditional frame-based spectroscopy, in which the scan awaits an external trigger (such as detector acquisition 9, a laser trigger, etc.) to move the electronic probe 8 upon receiving a logic signal at the Trigger input value. The first embodiment of the method, device, or system according to the invention allows for further exploration of such measurements by defining a custom number of beam positions 8 with the Trigger input flag, the others being driven by the internal clock generator of the means 131.For charge-coupled devices (CCDs), for example, which are very popular for two-dimensional acquisitions, typical readout speeds are ~1 ms, enabling fast scanning images (with beam position times 8 on the order of zs). This can find important applications in correcting drift in live images; b) spectroscopy or event imaging (as illustrated in Figure 12), when using detectors 9 whose clock frequency is equal to or greater than the clock frequency of the converters 132 of the scanning unit 1. This allows arbitrary scanning patterns. In this mode, the trigger indicator is periodically activated and sent to detector 9 for synchronization.This allows for event-based electron energy loss spectroscopy (EELS), 4D scanning transmission electron microscopy, fast tomographic reconstructions, etc.; c) the input and output (Trigin and Trig. out) Triggering signals can be used to synchronize the scanning unit 1 with multichannel time-to-digital converters (TDCs). With such systems, the trigger output (Trigout) is used to identify, in a timely manner, a desired position of the beam or probe 8. With multichannel systems, this can be extended to any digital signal, such as photon-counting photomultiplier tubes, for the identification of a single electron or photon, and can be used for electron energy-loss spectroscopy, cathodoluminescence measurements, dark-field ring imaging, etc.; d) Much faster scan rates are possible. In particular, one can consider using scan patterns based on the Lissajous method, where the X and Y frequencies are slightly different and where a phase shift can be defined.For comparison, a typical sequential scan for 2048x2048 pixels and a stopping time of 80 ns per pixel performs one back-and-forth scan of 2048 pixels in the X direction and a single back-and-forth scan in the Y direction. With the same scan parameters, but loading a Lissajous-based scan list, we demonstrate over 50,000 back-and-forth movements in the X and Y directions with less than 5% redundancy in beam position 8 (with a scan frequency of approximately 200 kHz). By adjusting the delay T. delay , T' delay and T" deiayIt is possible to obtain images such as those in Figure 5. As the impedance of each coil of means 4 or 5 increases with the scanning frequency, the image field of view is progressively reduced, which can be calibrated and taken into account in the DAC value. Figure 5 also compares the Lissajous and sawtooth Lissajous scanning schemes, the latter using a sawtooth pattern and the former a sinusoidal pattern. The sinusoidal pattern has fewer frequency components. The phase correction therefore gives better results. It is important to note that Figure 5 demonstrates the reaching of the analog limits of the scanning coils of means 4 and 5, and that it is therefore impossible to be faster; e) as illustrated in Figure 13, the Lissajous model also finds important applications in dose-controlled applications, which are very important in materials science for sensitive samples.These high-frequency scanning patterns mean that an image can already be partially reconstructed by interpolation and / or inpainting algorithms very early in the frame; thus, the beam 8 can be directed (in position and / or emission or interruption) to a desired dose deposition to avoid further damaging the sample 10 while still being able to reconstruct the image (also called "compressed sensing" or "sparse sampling"). These advantages are unattainable without the high-frequency scan list and the configurable delay. Furthermore, Lissajous patterns are not uniform, concentrating most of the rasterization time at the edges and much less at the center of the image. Playing with the acquisition time and the pause time while keeping the total time per position of the beam 8 constant significantly reduces this problem, as shown by the circular mask in Figure 5A.The applications of this space-spread dose deposition and the controllable duty cycle of the electron beam can find numerous applications in the fields of materials science, biology, earth sciences, fragile samples, 2D materials, battery research, pharmaceuticals, etc.; f) frequency-based scanning can be useful for effectively suppressing 50 Hz line noise. By setting the scanning frequency to a submultiple of this value, the 50 Hz line frequency and the beam scan 8 are synchronized; g) it is also possible to obtain digital processing of the detector input signal 9 based on the frequency response. By choosing the coefficients, a customized frequency response can be obtained. This can have considerable applications in microscopy, given that controlling electron noise is one of its most important aspects.Since the general average filter produces a sinusoidal frequency response (s / nc), the output low-pass filter is also non-uniform. This list-based approach allows for the construction of a more uniform low-pass filter; h) the input weighting of detector 9 can also be used to obtain time-resolved acquisitions. By setting the weight to one for a single time acquisition, while the other remains at zero, clock-time sampling is obtained; i) a fairly general application in electron microscopy, which is multidisciplinary, notably, without loss of generality, in scanning electron microscopy (SEM), transmission electron microscopy (TEM), STEM, and focused ion beam (FIB).The flexibility of the scanning device 1 is also fully backward compatible, as it can operate with traditional synchronization schemes, such as frame-based hyperspectroscopy. There are no drawbacks compared to previous scanning unit technologies.

[0166] The first embodiment of the method, device, or system according to the invention allows, for the first time, complete control of the scanning unit 1 in a probe microscope, and it is especially suited to applications using traditional and modern multidimensional detectors 9, particularly event-driven detectors such as Timepix3 or Timepix4, because it properly implements the tools for running fast scanning patterns. It resolves the persistent problems of sequential dithering in the standard scanning unit, which is severely asymmetrical because one direction scans much faster than the other, and the voltage ramps are not suitable for fast scanning. The advantage stems from the need for sufficient memory connected to the FPGA pinout, which is currently the standard for commercially available chips.It complements the arrival of event-driven detectors by providing a much more suitable scanning engine, capable of synchronizing and customizing the new timescales provided by these detectors. This first embodiment is also particularly well-suited to lithography applications, such as electron or ion beam lithography. Indeed, in these applications, two parameters are crucial. The first is the precise control of the exposure time at every point of the sample to be irradiated. The second is the ability to access the fastest possible scanning modes, thus significantly reducing the production time of lithographed samples.

[0167] Of course, the invention is not limited to the examples just described and many modifications can be made to these examples without departing from the scope of the invention.

[0168] Of course, the various features, forms, variants, and embodiments of the invention can be combined in various ways, provided they are not incompatible or mutually exclusive. In particular, all the variants and embodiments described above are combinable.

Claims

DEMANDS 1. A method for controlling a charged-particle scanning microscope, comprising: - A control of a position and / or an emission or interruption of a beam of charged particles on a sample, so as to sequentially position the beam of charged particles according to several positions of the beam on the sample the process being characterized in that it includes an association and / or programming, for each position of the beam, of at least one time data.

2. A method according to claim 1, characterized in that it comprises: - the reception of measurement data from at least one detector capturing this measurement data from the sample, - an assembly of measurement data corresponding to the different positions of the beam, preferably in one image, with at least one time data being used during command, reception and / or assembly.

3. Method according to claim 1 or 2, characterized in that at least one time datum includes at least one time offset relative to an instant of the command on a position of the beam considered.

4. A method according to claim 3, considered dependent on claim 2, characterized in that at least one offset time comprises a time offset, referred to as a measurement offset. between a moment of receiving measurement data from at least one detector and a moment of command on the position of the beam considered, this measurement offset may be common or specific to each detector considered.

5. A method according to any one of claims 3 or 4, characterized in that the control of the position and / or the emission or interruption of the beam of charged particles onto the sample comprises controlling the beam along a first axis and a beam control along a second axis distinct from the first axis, and in that at least one shift time includes a time shift, called control shift between a start time of beam control along the first axis and a start time of beam control along the second axis during the control on the beam position considered.

6. A method according to any one of claims 3 to 5, characterized in that at least one offset time: - is specific to each beam position and programmable or individually associatable to each beam position according to a value specific to that beam position, potentially different from the values ​​programmed or associated with other beam positions, or - has a value common to all positions of the beam.

7. A method according to any one of the preceding claims, characterized in that at least one time datum includes an acquisition time (tacq) during which the charged particle beam irradiates the sample when the beam position is controlled.

8. A method according to any one of the preceding claims, characterized in that at least one time datum includes a pause time (tpause) during which the beam of charged particles does not irradiate the sample when the beam position is being controlled.

9. Method according to the preceding claim, characterized in that the pause time is longer on a peripheral area of ​​the image than on a central area of ​​the image.

10. A method according to any one of the preceding claims, characterized in that at least one time datum is specific to each beam position and individually programmable or associatable to each beam position according to an eigenvalue at this beam position potentially different from the programmed values ​​or those associated with other beam positions.

11. Method according to claim 10 considered as dependent on claim 8 or 9 itself considered as dependent on claim 7, characterized in that it comprises a programming or association of the ratio tpause / tacq or of a variable proportional to tpause / tacq or which depends on tpause / tacq, individually to each position of the beam according to a value specific to that position of the beam potentially different from the values ​​programmed or associated with other positions of the beam.

12. A method according to any one of the preceding claims, characterized in that it further comprises an association and / or programming, for each beam position, of the activation or deactivation of at least one trigger data used during command, reception and / or assembly, from among: - a time-based trigger input (Trigin), preferably from at least one detector according to claim 2, and preferably for triggering the position command or the start of the pause time according to claim 8 or 9 or the acquisition time according to claim 7, and / or - a time-based trigger output (Trig ou(t), preferably directed at at least one detector according to claim 2, and preferably for a trigger triggered at the end of the position command or pause time according to claim 8 or 9 or the acquisition time according to claim 7.

13. A method according to any one of the preceding claims, characterized in that the control of the position and / or emission or interruption of the beam of charged particles on the sample comprises a control according to several positions of the beam successively distributed along a trajectory parameterized by at least two sinusoids or substantially sinusoidal functions.

14. A method according to any one of the preceding claims, characterized in that the sample is a reference sample and in what it includes, for a given trajectory of successive beam positions, an optimization of at least one time data point for each beam position and then a storage in a memory, and in association with this given trajectory, of at least one time data point for each beam position.

15. A method according to any one of the preceding claims, characterized in that the association and / or programming, for each beam position, of at least one time data point, is carried out at least by programmable digital means, and in that it comprises a conversion by an analog-to-digital converter between: - programmable digital means and at least one detector and / or - between programmable digital means and physical means of deflection and / or emission or interruption of the beam of charged particles.

16. Control device for a charged-particle scanning microscope, comprising: - means for storing or receiving beam positions - at least one control output arranged and / or programmed to send, to the attention of physical means for controlling the charged particle beam, a command for a position and / or an emission or interruption of the charged particle beam, - Electronic and / or computer processing means, the processing means preferably being arranged and / or programmed to associate and / or program, for each position of the beam, at least one time data point.

17. Control system, comprising: - a control device according to claim 16, and At least one of the following: o Means for generating the charged particle beam, o Physical means for controlling the charged particle beam, these means being connected to at least one control output of the control device, o At least one detector arranged to capture measurement data from a sample.

Images