Time-resolved cathodoluminescence sample probing

Abstract

A method for investigating a sample by time-series emission of light in a cathodoluminescence (CL) microscope having an electron beam and a light sensor, comprising the steps of: setting parameters of the electron beam to irradiate at a first dose rate; flushing a buffer of the light sensor; scanning the electron beam over a region of interest of the sample while collecting CL emission with the light sensor while preventing data from being read from the buffer until the entire scan is complete; blanking the electron beam once the entire scan is complete, interrogating the buffer to identify a first CL image; and then interrogating the buffer to retrieve all remaining CL images and tagging all retrieved CL images according to time series starting with the first CL image.

Description

[Technical Field]

[0001] (Related Applications) This application claims the benefit of priority from U.S. Provisional Application No. 63 / 053969, filed July 20, 2020, the disclosure of which is incorporated herein by reference in its entirety. This application also relates to subject matter disclosed in PCT Patent Application No. PCT / EP2020 / 063093, the disclosure of which is incorporated herein by reference in its entirety.

[0002] (Technical field) The present invention relates to the use of a scanning electron microscope to investigate samples by cathodoluminescence spectroscopy, particularly but not exclusively to the investigation of samples by observing changes in spectra and / or images obtained with a time-resolved scanning cathodoluminescence microscope. [Background technology]

[0003] Traditionally, most commercially available cathodoluminescence (CL) measurement solutions are third-party add-ons to existing scanning electron microscopes (SEMs). They consist of at least a light collection means and a photodetector installed in the scanning electron microscope. The applicant disclosed a state-of-the-art scanning cathodoluminescence microscope in the above-mentioned PCT application. CL systems have been used with either an airborne detector (SP mode) that collects the spectral distribution of emitted CL over a scanned area of ​​the sample, or a point detector (PC mode) that generates a panchromatic CL image. In either of these modes, the acquisition resolution is related to the size (diameter) and focus of the SEM's electron beam and the focus of the CL collection optics.

[0004] Irradiating a sample with an electron beam can induce (at least local) changes in certain sample properties. For example, Gelhausen et al. employed electron beam irradiation during cathodoluminescence to determine the chemical origin of the dominant photoemission in moderately and heavily Mg-doped GaN (see “Dissociation of H-Related Defect Complexes in Mg-Doped GaN,” Physical Review B69, 125210, 2004). Similarly, electron-beam-induced current (EBIC) and CL were used to study the effects of electron injection on the increase in the minority carrier diffusion length in p-type GaN (see Czerniak et al., “CL Study of Effects Induced by Electron Injection in GaN,” Applied Physics Letters, Vol. 82, No. 21, 2003). These studies employed SP mode to investigate or confirm suspected phenomena in electron-implanted GaN samples. Specifically, they observed changes in the intensity and frequency of specific CL emissions and attributed them to the phenomenon they investigated. Summary of the Invention

[0005] The following summary of the disclosure is included to provide a basic understanding of some aspects and features of the present invention. This summary is not an extensive overview of the invention, and as such, it is not intended to particularly identify key or critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concepts of the invention in a simplified form as a prelude to the more detailed description that is presented below.

[0006] The present invention has been developed by examining the following questions: What properties and changes in a sample can be observed using time-resolved CL studies? What conditions are necessary to enable the observation of these changes? We investigated the time-dependent changes in CL spectra with increasing radiation dose and attempted to quantitatively analyze them. Furthermore, we found that electron beam irradiation can also cause changes in CL panchromatic and color filter images (PC mode) with increasing electron beam radiation. We found that certain features in the images disappear with increasing radiation and, in some cases, reappear with further increases in radiation. Therefore, the disclosed embodiments relate to both SP mode and PC mode studies. However, to efficiently implement some of the disclosed embodiments, it may be necessary to modify the operation of a standard CL microscope, as described in more detail below.

[0007] The disclosed embodiments provide a method for investigating various properties and / or changes in a sample caused by an electron beam using time-resolved CL spectra and / or images. The properties and changes are investigated by observing changes in the optoelectronic properties of the sample caused by electron injection. These changes are discovered by observing changes in the spectrum (SP mode) or image (PC mode) that correlate with the dose irradiated to the sample. In this disclosure, a phenomenon appearing in either the CL image or CL spectrum due to electron beam irradiation is referred to as a change event.

[0008] In the disclosed embodiments, various parameters of the electron beam can be adjusted to precisely control the dose delivered to the sample and / or layers of interest within the sample. For new, untested samples, the beam is adjusted to deliver a low dose with each pass, allowing for continuous CL collection. The optical collection system is also set to a very fast collection speed, i.e., a very short acquisition time per pixel. The time-series spectra or images are then processed to determine which doses reveal changes in the sample, i.e., change events. Subsequent investigations are performed by adjusting the electron beam parameters to deliver the doses observed to produce changes in the spectra or images, thus accelerating the observation of those events.

[0009] The disclosed embodiments are advantageous in that they can discover previously unknown phenomena, and then provide a faster way to examine samples for these phenomena. The disclosed embodiments also include a method for performing CL acquisition without reading data from the sensor's buffer until the full acquisition is complete. This process allows for the collection of finer data as the illumination progresses, since no data is lost due to querying the sensor's buffer.

[0010] Disclosed aspects include a method for investigating a sample by generating a time-series emission of light in a cathodoluminescence (CL) microscope having an electron beam and a light sensor, comprising the steps of: setting parameters of the electron beam to irradiate the sample at a first dose rate; flushing a buffer of the light sensor; scanning the electron beam over a region of interest of the sample while collecting CL emission with the light sensor while preventing data from being read from the buffer until the entire scan is complete; blanking the electron beam once the entire scan is complete, interrogating the buffer to identify a first CL image; interrogating the buffer to retrieve all remaining CL images and tagging all retrieved CL images according to time sequence starting with the first CL image. The CL images may be panchromatic (i.e., integrating a large swath of the emitted CL spectrum) cathodoluminescence images or color-filtered (i.e., selecting a precise portion of the spectrum) cathodoluminescence spectral images.

[0011] The step of collecting CL luminescence with an optical sensor may begin before the start of electron beam scanning. The method may include identifying change events, where changes are visible in the series of CL images, and determining a change dose corresponding to the dose delivered to cause each change event. The method may also include setting electron beam parameters to irradiate the sample at a second dose rate higher than the first dose rate, performing a second scan to irradiate the second sample at the change dose, and collecting CL luminescence with an optical sensor to confirm the occurrence of a change event. The electron beam may also be defocused to reduce the irradiation rate or dose rate, or CL images may be acquired at a rate of 50 ns to 2 μs per pixel. Conversely, the dose may be controlled by changing the electron beam imaging field of view, pixel density, or scan speed. A height map of the region of interest may be created in advance, and the distance from the collection optics to the sample may be adjusted to be constant during the electron beam scan according to the height map, or the distance between the sample and the collection optics may be directly adjusted according to a height sensor that probes the distance in real time.

[0012] Another disclosed aspect includes a method for investigating a sample by generating a time series of luminescence emissions from a cathodoluminescence (CL) microscope having an electron beam and a photosensor, comprising: performing a discovery scan of the sample, in which the electron beam delivers a first dose rate (exposure rate) to the sample while collecting CL emissions with the photosensor; generating a time series of CL images collected during the discovery scan; identifying a change event in the time series where a change between successive images is observable; determining a change dose corresponding to the total dose delivered to the sample up to the occurrence of the change event; performing a test scan of a second sample using the change dose obtained from the discovery scan while collecting CL emissions with the photosensor; and generating a second time series of CL images collected during the test scan. The CL images may include panchromatic cathodoluminescence images or cathodoluminescence spectral images, and preventing data from being read from the photosensor's memory buffer during the discovery scan and, optionally, during the test scan, until the discovery scan is complete. The photosensor's memory buffer is flushed before starting the discovery scan and before starting the test scan. In addition, in each of the discovery scan and the inspection scan, acquisition of a CL image by the optical sensor is started before scanning of the electron beam is started. During the discovery scan, CL emission is collected by the optical sensor by acquiring a CL image in a first field of view, and during the inspection scan, CL images are acquired in a second field of view that is the same as or smaller than the first field of view.

[0013] The method may include generating a height map of the sample before performing the discovery scan and using the height map to maintain a constant distance from the sample to the focusing objective during the discovery scan. The height map may be generated using a focusing fiducial of the focusing objective or may be generated using a dedicated sensor calibrated to the latter fiducial. The method may also include defocusing the electron beam. [Brief explanation of the drawings]

[0014] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain and illustrate the principles of the invention. The drawings are intended to diagrammatically illustrate major features of exemplary embodiments. The drawings are not intended to depict every feature of an actual embodiment or the relative dimensions of the depicted elements, and are not drawn to scale.

[0015] Other characteristics and advantages of the invention will become apparent from the following description of non-limiting exemplary embodiments with reference to the accompanying drawings, in which: [Figure 1] FIG. 1 is a cross-sectional view of an integrated quantitative cathodoluminescence scanning electron microscope modified for performing time-resolved CL studies according to disclosed embodiments. [Figure 2] FIG. 2 is a schematic cross-sectional view of a portion of the sample, showing the energy distribution from the electron beam inside the sample. [Figure 3] FIG. 3 is a flow chart of sample testing according to one embodiment. [Figure 4] FIG. 4 is a flow chart of a sample inspection according to another embodiment. [Figure 5] 5A and 5B are CL spectra of a sample showing the transformation events induced by electron beam irradiation, according to one embodiment. [Figure 6] 6A and 6B are CL images of a sample showing the transformation events induced by electron beam irradiation, while FIG. 6C shows the CL spectrum corresponding to FIGS. 6A and 6B, according to one embodiment. [Figure 7] FIG. 7 is a flowchart of a process for performing an early discovery process and using the results to perform testing on a sample, according to one embodiment. [Figure 8] FIG. 8 is a chart showing the time evolution of some of the parameters in FIGS. 5A and 5B, further illustrating the discovery process in spectrum mode. DETAILED DESCRIPTION OF THE INVENTION

[0016] Some embodiments of the invention will now be described in more detail with reference to the accompanying drawings, in which identical functional and structural elements appearing in different drawings are assigned the same reference numerals.

[0017] Next, embodiments of time-resolved cathodoluminescence inspection of samples according to the present invention will be described with reference to the drawings. Different embodiments or combinations thereof can be used for different applications or to achieve different advantages. Depending on the results to be achieved, different features disclosed herein can be partially or fully utilized alone or in combination with other features, while balancing the advantages with requirements and constraints. Therefore, certain advantages will be emphasized with reference to various embodiments, but are not limited to the disclosed embodiments. That is, the features disclosed herein are not limited to the described embodiments, but can be "mixed and matched" with other features and incorporated into other embodiments, even if such is not explicitly stated herein.

[0018] FIG. 1 shows the relevant portions of a CL microscope in which the CL collection optics are integrated with the electron beam column. However, it will be understood that the embodiments disclosed herein can be implemented in situations where the CL device is included as an add-on tool in a scanning electron microscope. To maximize the benefits of the disclosed embodiments, good control over electron beam parameters (e.g., focus, current, beam width, etc.), good control over the CL collection optics (focus, sensitivity, sensor speed, etc.), and access to modify signal collection and image formation software and / or hardware are desirable. A more detailed description of CL microscopes is provided in the cited International Application No. PCT / EP2020 / 063093 (WO 2020 / 225453), which is incorporated herein by reference; however, for convenience, some relevant explanations are provided herein with necessary modifications for the disclosed embodiments.

[0019] (Embodiment of the Invention) As shown in FIG. 1 , a microscope typically includes an electron column 41 housed in a vacuum enclosure 10 and an optical (CL) imaging element 42 in an atmospheric environment. The integrated microscope shown in FIG. 1 can generate electron beam images, CL images (PC mode) using a point detector such as a photomultiplier tube (PMT) or an InGaAs diode array 46, and CL spectroscopic images (SP mode) using an array detector such as a CCD camera 45 or an InGaAs diode array 46. Incidentally, in FIG. 1 , the light source 26 may be replaced by a point detector (e.g., a PMT) and used to generate a CL image instead of the detector 46. Therefore, element 26 in FIG. 1 can represent either a light source or a point detector. Note that the term point detector does not necessarily imply that the detector has a single light-sensitive element, but also implies that the detector's output signal represents the integral of multiple emitted light energies from the sample.

[0020] It should be noted that both CCD cameras and point detectors have been shown to be usable for capturing CL spectra. Point detectors are typically used to generate CL spectra by spatially orienting a point detector with a slit to capture desired wavelengths of the wavelength-dispersed emission. In the example shown in Figure 1, a CCD camera can be used as a line detector by summing all the pixels in each column, simultaneously capturing multiple wavelengths of the dispersed emission, and spatially orienting the CCD so that each summed column represents a single or defined bandwidth of the full spectrum.

[0021] The imaged CL emission can be correlated with the structure and quality of the sample's material at the nanoscale, revealing material stress, impurities, crystallographic, and subsurface defects that are invisible using other imaging modes. Importantly, while CL imaging is generally considered a non-destructive method of examining a sample, the embodiments disclosed herein investigate changes in the sample's properties caused by irradiation with an electron beam. Thus, while the sample remains fully operational after CL imaging, the injection of the electron beam by the SEM alters the sample's optoelectronic properties.

[0022] The electron column includes an electron source 1, such as a thermionic or field emission source, that emits electrons. The emitted electrons are shaped into an electron beam 9 by various particle-optical elements, such as an electromagnetic condenser lens 5', an electromagnetic objective lens 5, and aperture disks (sometimes called stops) 6. Note that any of the aperture disks 6 can function as electrostatic lenses by applying an electric potential thereto. In FIG. 1, pulse 2 indicates that electron emission from electron source 1 may be pulsed in this particular example, but this is not necessarily the case. For example, the emission may be continuous or intermittent using a shutter. Regardless of the electron source's operating method, it is desirable to provide a good method for blocking the beam (e.g., beam blocker, beam deflector, etc.) to improve results.

[0023] The magnetic field generated by the magnetic lens creates a converging electron beam 9, which can be focused and scanned onto the surface of the sample 7. In this example, the electron beam 9 generated by the electron emitter 1 propagates from top to bottom in the figure. The electron beam's divergence can be altered to diverge, collimate, or converge by a condenser arrangement, such as lens 5'. The condenser arrangement can be located below the electron emitter. Because the resolution of the electron image is directly related to the beam spot size on the sample, conventional efforts have been made to properly focus the electron beam onto the sample. The position of the focal plane, more specifically its height above the sample, can be adjusted by changing the strength of the magnetic field flowing through the electromagnetic objective lens 5.

[0024] As can be seen in FIG. 1 , the reflective objective lens within the electromagnetic objective lens 5 includes a first mirror M1, also referred to as a primary mirror in this example, which is spherical and concave, and a second mirror M2, also referred to as a secondary mirror in this example, which is spherical and convex. The diameter of the first mirror M1 is larger than the diameter of the second mirror M2. The first mirror M1 is positioned above the second mirror M2 and is arranged to reflect light coming from the sample 7 as a result of the electron beam 9 impinging on the surface of the sample 7 and direct the light to the second mirror M2, which is located between the sample and the first mirror M1. The second mirror M2 is arranged to redirect the light along the optical axis of the electromagnetic objective lens, and a third mirror M3, which is planar in this example, is arranged to redirect the light beam 4 toward the output. In this example, the third mirror M3 is at a 45° angle with respect to the axis of the electron beam 9 and is used to redirect the light out of the vacuum enclosure 10. All three mirrors M1, M2, M3 have openings along the electron beam path so as not to obstruct the electron beam.

[0025] The microscope shown in FIG. 1 also includes a first deflector 17 and a second deflector 15. Each deflector can include, for example, four longitudinal electrical conductors (quadrupoles), also called electrodes. It can also include four magnetic coils (magnetic deflectors). However, the number of electrodes or magnetic poles can be different from four. For example, there can be eight electrodes or magnetic poles (octopoles) instead of four, so that the astigmatism of the electron beam can also be corrected. In this case, the four electrodes or magnetic poles used to correct the astigmatism are called stigmators.

[0026] Electron beam characteristics can be affected by the current applied to the electron source, the accelerating voltage, electromagnetic and electrostatic elements along its path, and the working distance to the sample. In the embodiments disclosed herein, it is important to have good control over the dose, particularly the rate (dose per time), delivered to the sample by the electron beam, including the beam's electron energy, beam current, and spot size. Note that deflectors are used to deflect the electron beam's trajectory so that it can scan across the surface of the sample, but do not control the characteristics of the beam itself. However, scanning parameters, such as pixel density and field of view when rastering the beam over the region of interest, can be used to control the dose delivered to the sample surface, provided the dwell time per pixel is sufficiently low.

[0027] To accurately calculate the delivered dose, it is useful to consider the energy distribution as pear-shaped, as shown in Figure 2. Figure 2 shows a cross-section of a portion of a sample with, for example, three different layers, each made of a different material. Depending on the control signals applied to the electron column, the beam can be focused to different depths within the sample with different spot sizes on the sample. Specifically, the acceleration voltage of the primary beam must be high enough that a non-negligible fraction of electrons actually reaches the region or layer of interest, for example, if the region or layer is buried, as shown by layers B and C. However, regardless of the acceleration voltage, the actual energy distribution is approximated by the pear shape shown at 9'. Therefore, depending on the layer of interest, the actual delivered dose can be approximated by considering "slices" of the pear deposited within that layer. Note also that the spatial width of the energy distribution is larger than the beam spot size at the surface of the sample.

[0028] Typically, to achieve high resolution, the electron beam is focused as tightly as possible, down to sub-nanometer sizes. However, in certain embodiments, as described in more detail below, to reduce the dose delivered per scan, the electron beam is actually defocused, e.g., to a diameter or spot size of 1-10 micrometers, distributing the energy over a larger area and thus reducing the dose per pixel. Conversely, in other embodiments, the electron beam is tightly focused but quickly rastered across the field of view (FOV). For example, by increasing the scan speed, i.e., shortening the dwell time per pixel (e.g., from 20 μs to 100 ns), and reducing the number of pixels in the FOV (e.g., from 2048 to 128), it is possible to obtain high-quality images for each frame captured for image sequence mode or to have a statistically relevant number of frames for spectral acquisition. Ideally, for spectral acquisition, it is desirable to capture several frames of the FOV for each spectrum acquired.

[0029] Returning to Figure 1, light reflected by mirror M3 is focused by lens 22 onto imaging monochromator 43. This example provides two imagers: a CCD camera 45 or InGaAs diode array 46-InGaAs for SP mode, and a point detector such as PMT 46-PMT26 for PC mode. In other embodiments, CCD camera 45 can also be used in PC mode by summing all pixels in a column into a single pixel, thereby forming a linear detector. If mirror 24 is a half mirror, both imagers can be operated simultaneously. Conversely, mirror 24 can be a flip mirror, allowing one imager to be operated at a time. With this arrangement, detector 46-PMT can be used to detect the light intensity of a specific wavelength, and CCD camera 45 or InGaAs diode array 46-InGaAs can be used to simultaneously detect the light intensity of several wavelengths, especially when operated in linear mode.

[0030] In the embodiment of FIG. 1, an electron detector 19 is provided to detect secondary electrons emitted from the sample or backscattered electrons reflected by the sample. The detector signal can be used to generate a scanning electron microscope (SEM) image. Also in the embodiment of FIG. 1, the sample holder 47 is in the form of a cryostage, which keeps the sample cool enough that CL emission is more efficient because many non-radiative pathways become impossible below a certain threshold. This also improves spectral resolution by reducing "thermal fluctuations" in each emitted photon.

[0031] Typically, in SP mode, the CCD detector 45 or InGaAs diode array 46-InGaAs has a built-in buffer, and as each image is formed, it is stored in the internal buffer. The controller 52 retrieves each image from the buffer and displays the most recent image on the monitor. This configuration does not guarantee accurate retrieval of subsequent images, since retrieving several images consecutively can result in the time overhead incurred by retrieving the images exceeding the acquisition time of the next few images. Images that could have been acquired during this overhead time are skipped and unavailable. However, according to the disclosed embodiment, it is desirable to generate a new image for each small change in the sample caused by the electron beam irradiation. Therefore, in the disclosed embodiment, the operation of the controller is modified so that images are not retrieved from the CCD detector 45 until all images have been generated and the scan is complete. Conversely, this prevents any querying of the buffer during the electron beam scan of the sample, thereby ensuring that all acquired images remain intact in the buffer and that the controller 52 can link them to the acquisition order. When using PC mode, e.g., PMT46-PMT26, the point detector is processed using a different channel than the array detector, so controller 52 itself includes a buffer to store each PC image. In this case, the same processing is applied to the buffers included in controller 52. In a further embodiment, the two approaches can be combined to achieve the same result using an array detector in PC mode.

[0032] Figure 3 is a flowchart showing a general process for acquiring CL spectra and images to identify change events. This general process can be employed for both samples where the correlation of events to the applied dose is unknown (a discovery process) and samples where a certain amount of dose is expected to result in observable change events (an inspection process). A key difference between the operation of a discovery process and an inspection process is that in a discovery process, capturing each minute increase in dose is important, and dose delivery must be minimized during each pixel acquisition. In contrast, in an inspection process, the dose required to produce an observable event is known in advance, so the setup is positioned to directly capture a known amount of dose. As a result, the inspection process is much faster than a discovery process, but the time evolution of the CL images obtained from a discovery process can provide equally important inspection results.

[0033] The process shown in Figure 3 begins with setting appropriate acquisition parameters. For the discovery process, in step 300, electron beam parameters such as acceleration voltage, beam current, and dwell time are set to minimize the dose delivered by the electron beam to each pixel at each dwell time, ensuring that all small changes are detected. This is done by considering the energy distribution and the location of the layer being investigated, as described with reference to Figure 2. If necessary, the delivered dose can be further minimized by defocusing the electron beam to generate a spot with a diameter of 0.1 to 10 microns. Similarly, in step 305, the parameters of the CL acquisition system (including the parameters of the camera 45 and PMT 46) are set. In this step, the field of view can be reduced, if necessary. Note that there is an inverse relationship between beam focus / spot size and field of view: defocusing the beam reduces the delivered dose, while reducing the field of view increases the delivered dose.

[0034] As previously mentioned, conventionally, the buffers of the camera 45 and PMT 46 contain data from previous acquisitions (regardless of whether such data has been previously retrieved by the controller). Typically, data remains in the buffer until it is overwritten by an acquisition when the buffer is full. Conversely, in this embodiment, in step 310, the buffer is flushed of all data in preparation for each acquisition. To record all CL emissions, in 315 the controller is programmed not to retrieve any data from the sensor (camera 45 and / or PMT 46) until the entire acquisition run is complete. The purpose is to keep communication with the sensor to a minimum so that the time series can be reconstructed in the proper chronological order, since multiple calls to the buffer would result in losing track of the timing of individual images or spectra. For example, using a straightforward method based on typical SDK commands, communication with the controller, reading and transferring data would discard several spectra recorded by the sensor with each acquisition, whereas in this embodiment, no spectra are discarded at all. Thus, while conventional methods have the advantage that the image or spectrum is immediately available during acquisition, the disclosed method forgoes immediate viewing of the image in order to avoid losing CL emission data and thus maintain full control over the image / dose relationship, allowing the image to be displayed only once the entire acquisition is complete.

[0035] In step 315, continuous acquisition occurs without accessing the sensor data. Rather, the data is collected in the sensor's buffer. Once the entire acquisition is complete, i.e., once the region of interest has been scanned a predetermined number of times, the buffer is queried to retrieve the data in step 320, and the data is tagged according to the timing of its acquisition. Because the data is tagged, any image or spectrum obtained at any exposure time can be displayed on the monitor. Furthermore, the entire time series can be continuously displayed or examined to identify change events.

[0036] Observable changes in the time-series images and / or spectra can include those caused by the activation of the material by the electron beam itself (e.g., changes in its properties caused by irradiation with the electron beam, either through charge accumulation in the material, ionization of dopants in the material, or physical effects such as kicking interstitial atoms back into place or disrupting chemical complexes such as Mg:H complexes in gallium nitride). Other changes can be associated with the formation of a dynamic equilibrium between charge carriers in the sample and carriers injected by the electron beam. Examples of this effect are carrier-induced band bending or the relative saturation of a particular emissive state, resulting in a relative decrease in emission from that transition compared to other transitions. The embodiments disclosed herein provide a method for observing static or dynamic changes in the optoelectronic properties of a material and their temporal evolution. That is, changes in the optoelectronic properties associated with accumulated charged particle irradiation cause certain features to appear and disappear over time in the SP and / or PC modes.

[0037] According to the disclosed embodiments, since it is desirable to deliver small increments of dose when investigating transformation events, it is also beneficial to control the timing of when the beam actually hits the sample. Also, as explained further below, in the disclosed embodiments, it is also beneficial to start image acquisition before starting the electron beam hits the sample. To support these objectives, it is desirable to perform good beam blanking. Blanking can be achieved using electrostatic or electromagnetic deflectors already present in the electron microscope, or by using a dedicated beam blanker added to the electron microscope column.

[0038] Additionally, the distance between the collection optics and the sample must be maintained with high precision during the measurement to ensure there are no artifacts associated with out-of-focus acquisition of the collection optics. This can be done in several ways, for example, by creating a height map of the area to be scanned prior to the measurement using the optical focus as a reference or using a capacitive or other distance measurement sensor. In this way, moving the collection optics or the sample during image acquisition of the sample can keep the distance constant, especially if the sample has a curvature or uneven surface.

[0039] In view of the above considerations, a method for acquiring time-dependent CL images or spectra is provided, as shown in the flowchart of FIG. 4. As a first step, a height map is optionally generated, as indicated by dashed line step 400. Next, similar to the embodiment of FIG. 3, in step 401, the electron beam and optical collection parameters are set to provide the required dose increment for each pixel in each scan. In step 402, the electron beam is blanked, for example, by deflection into a Faraday cup or other sufficiently fast beam-blanking method. In step 404, the sample is moved to a position such that the area selected for measurement is exposed to the charged particle beam. As mentioned above, the distance between the sample and the collection optics is a critical parameter for maintaining focus of the CL collection optics, and in step 406, the sample height is adjusted to achieve proper focus. This height can be changed during the scan as needed for uneven samples according to the height map generated in step 400.

[0040] Before acquisition begins, the sensor buffer is flushed at 410. This step can potentially be performed any time before step 412, but must ensure that the buffer is completely empty before CL acquisition begins. At step 412, sensor acquisition begins while blocking any queries requesting data from the buffer. Then, at step 414, the beam is unblanked and scanning of the region of interest begins. As shown in FIG. 4, it is preferable to start CL acquisition before starting the electron beam scan. This ensures that the CL acquisition captures the start of the electron beam scan. At step 416, the entire scanning and CL acquisition process is performed without retrieving data from the sensor buffer. That is, all acquired CL data is maintained only in the buffer.

[0041] Once all tests are complete, the electron beam is blanked again in step 418. At this point, data can be retrieved from the sensor. In step 420, it is determined which data in the buffer corresponds to the first captured CL image. This data is used to tag the remaining data so that a time series can be generated from the data in the buffer in step 425.

[0042] Generally, cathodoluminescence emission is quite weak. Therefore, it is conventional in the art to use a high electron beam current to enhance CL emission in order to obtain good CL detection. However, contrary to expectations, the inventors have found that reducing the electron beam current so that the dose rate irradiating the sample is very low results in previously unobserved changes appearing in the acquired CL signal. Therefore, at least for the initial survey scan, also known as a discovery scan, the disclosed embodiments aim to minimize the dose delivered per time, for example, by setting a low electron beam current or by defining a large field of view so that a low dose is delivered per unit area at a given pixel density. We have also unexpectedly found that fast acquisition speeds of 50 ns to 2 μs per pixel, particularly 100 ns to 200 ns, allow previously unobserved changes to be seen. Furthermore, to achieve such acquisition speeds and generate continuous images with small incremental doses, it is beneficial to not perform any readout to the sensor buffer until the entire scan is completed.

[0043] 5A and 5B show the spectrograms of the sample, where FIG. 5A shows three peaks at three different energies taken at time t0 before the sample was sufficiently irradiated to indicate a transformation event, and FIG. 5B shows three peaks at three different energies taken at time t0 after the sample had received a sufficient dose of radiation to indicate a transformation event. N In this respect, t0 does not indicate the time when the electron beam scan started, but the time t N t0 and t N An example of the determination is shown in Figure 8. N is the transition period t Tt0 is chosen as the point before the start of the transient period, when the values ​​of Peak 1 and Peak 2 reach a plateau after t0. As can be seen from Figures 5A and 5B, the intensity of Peak 1 has increased, but the full width at half maximum (dotted FWHM) remains relatively the same. The amplitude of Peak 2 has decreased significantly, while the curve of Peak 3 has not changed at all. Thus, the disclosed embodiments provide a method for detecting the change in the amplitude over time, particularly during the transient period t T The amplitude change during the transient, e.g., the rate of change within the transient, the peak wavelength (energy) shift over time and within the transient, the change in the peak half-width over time, or the rate of change of these parameters over time (i.e., t in Figure 8). N This involves quantitative analysis of various peaks in the time series spectrum by examining the rate of change of a parameter with respect to cumulative exposure (as at the previous time).

[0044] The dopant concentration of a sample can be investigated by examining the events and rates of change that cause wavelength or energy shifts, amplitude changes, and / or FWHM changes in the CL spectrum. Furthermore, charge accumulation at the interface between two layers of different materials can cause peaks to appear and then disappear over time, making dose control even more important for determining the measurement window.

[0045] As mentioned above, we have also found that CL images (PC mode) can change over time due to the accumulation of electron beam radiation. For example, when performing feature counting in InGaN / GaN multiple quantum wells, certain features may appear early in the irradiation process, then disappear, and eventually return once the irradiation dose exceeds a certain threshold. Figures 6A and 6B show CL images selected from two different times of time-series PC mode CL imaging, displaying two types of defects in the sample. Such features can appear as either dark or bright spots on a panchromatic cathodoluminescence image. Type 2 defects have disappeared from the image in Figure 6B, while Type 1 defects are still visible. Figure 6C shows the spectra at these two times (t0 is shown as a solid line, t N The graph shows the time course of the spectrum (dashed line), indicating that there is no significant change. This explains why time-dependent changes in the spectrum do not necessarily indicate corresponding changes in the CL image, and changes in the CL image do not necessarily indicate changes in the spectrum.

[0046] By acquiring time-series CL images at low doses, quantitative analysis can be obtained by correlating the appearance and / or disappearance of specific features in the images with the recorded applied dose up to the observed change event. Once these parameters are known, the information can be used to inspect different locations on the sample or different samples. For example, if you observe that Type 2 defects have disappeared from the CL image at the recorded dose, and then in a second inspection where you want to count the number of Type 1 defects, you can immediately apply the recorded dose to obtain an image of only Type 1 defects, making it easier to count them. Alternatively, you can compare the image at time t0 with the image at time t N By subtracting the image of type 1 from the image of type 2, the density and location of type 2 defects can be determined. In addition, the type of defect can be identified by the change in the appearance of the defect points in the image, and can be distinguished from defects whose appearance does not change over time.

[0047] FIG. 7 is a flowchart illustrating a process that may be performed using the processes of FIGS. 3 and / or 4. At 700, the electron beam and CL acquisition system are configured to perform time-resolved CL imaging at a first exposure dose rate, which is a slow exposure dose rate. As explained above, this can be done, for example, by defocusing the electron beam, performing a fast scan with a large field of view that produces a short dwell time per pixel, or simply reducing the probe current. It is also possible to configure the system to prevent data from being read from the sensor until the entire test is complete. At 705, a scan is performed for the discovery process, generating a CL time series using SP mode, PC mode, or both. Because the dose rate is set low, each image in the CL time series corresponds to a small increment of dose delivered to the sample at the corresponding pixel. At step 710, the resulting series is analyzed by any of the methods disclosed herein to identify change events and to determine the change events and specific changes and rates in the CL images. Additionally, the specific dose required to reach any change event, e.g., the dose delivered to reach time t and the dose delivered to reach time t, are also determined. N The dose required to reach the alteration event is recorded as the alteration event dose. In 715, the alteration event dose obtained in step 710 is used to modify the parameters of the electron beam and CL system. For example, the electron beam is refocused and / or the field of view is reduced so that the electron beam must raster over a smaller area. A second test is then performed at 720 on a different site on the sample or on a different sample. This time, the system can much more quickly deliver the alteration event dose required to observe the alteration event. As a result, subsequent scans using knowledge of when the alteration event occurs and the required dose will be much faster than the discovery scan. This is particularly true for situations where a review scan is performed with a reduced field of view.

[0048] In other words, the discovery scan is performed in a manner that allows for the collection of many data points representing each small increment of delivered dose per pixel. Conversely, during the inspection scan, the operator already knows the general shape of the curve per delivered dose and can perform the inspection scan in a manner that generates fewer data points and use only those data points to determine whether these points match the expected curve obtained from the discovery scan. Thus, according to disclosed embodiments, there is provided a method for investigating a sample by generating a time-series emission of a cathodoluminescence (CL) microscope having an electron beam and a light sensor, comprising the steps of: The method includes the steps of: performing a discovery scan of the sample, in which an electron beam scans the sample while collecting CL emission with an optical sensor to generate a first set of data points; generating a first time series of CL images using the first set of data points; identifying a transient period in the first time series during which a change in successive images is observable; performing a test scan of a second sample while collecting CL emission to generate a second set of data points, the second set having fewer data points than the first set; generating a second time series of CL images using the second set of data points; and determining parameters of the change observable during the transient period in the second time series of CL images. The parameters may include a shift in peak wavelength, a shift in peak amplitude, a rate of shift in peak wavelength and / or amplitude, etc. The second set of data points may be collected from a smaller area of ​​the sample than that used to obtain the first set of data points. The smaller area can be obtained by setting a smaller field of view for the test scan than that set for the discovery scan.

[0049] Although the present invention has been described in relation to particular embodiments thereof, many other variations and modifications and other uses will become apparent to those skilled in the art. It is therefore preferred that the invention be limited not by the specific disclosure herein, but only by the scope of the appended claims.

Claims

1. 1. A method for investigating a sample by generating a time series of emission light in a cathodoluminescence (CL) microscope having an electron beam and a light sensor, comprising: setting parameters of the electron beam to irradiate the sample at a first dose rate; flushing the photosensor buffer; performing a first scan by scanning the electron beam over a region of interest of the sample while collecting CL emissions with the photosensor while preventing reading of data from the buffer until an entire scan is completed; blanking the electron beam once the entire scan is complete and querying the buffer to identify a first CL image; querying the buffer to fetch all remaining CL images and tagging all fetched CL images in chronological order starting from the first CL image; identifying a change event, where a change is visible in the CL images of the series, and determining a change dose corresponding to the dose delivered to cause the change event; setting parameters of the electron beam to irradiate the sample at a second dose rate that is greater than the first dose rate; performing a second scan irradiating a second sample with the varying dose; and collecting CL emissions with said optical sensor to confirm the occurrence of a change event.

2. The method of claim 1 , wherein the CL image comprises a cathodoluminescence spectrum.

3. The method of claim 1 , wherein the CL image comprises a panchromatic or color-filtered cathodoluminescence image.

4. The method of claim 1 , wherein collecting CL emission with the photosensor begins before scanning of the electron beam begins.

5. The method of claim 2 , wherein setting the parameters of the electron beam comprises defocusing the electron beam, changing its probe current, or changing the electron image field of view.

6. 4. The method of claim 3, wherein collecting CL emissions with the optical sensor comprises acquiring CL images at a rate of 50 ns to 2 μs per pixel.

7. The method of claim 1 , wherein setting the electron beam parameters comprises forming a beam having a spot size diameter between 0.01 and 10 micrometers.

8. The method of claim 1 , wherein setting parameters of the electron beam to irradiate the sample at the second dose rate comprises reducing a field of view or changing a probe current.

9. The method of claim 1 , further comprising generating a height map of the region of interest and varying a distance from collection optics to the sample during the scan of the electron beam according to the height map.

10. The step of identifying the change event includes a change being observable in a plurality of successive images; 10. The method of claim 1, wherein the second dose rate corresponds to a total dose delivered to the sample up to the occurrence of the alteration event.

11. The method of claim 10 , wherein the CL image comprises a panchromatic or color filtered cathodoluminescence image, or a cathodoluminescence spectral image.

12. The method of claim 10 , further comprising flushing a buffer of the optical sensor before performing the second scan.

13. The method described in claim 12, further comprising a step of initiating CL image acquisition by the optical sensor before initiating scanning of the electron beam.

14. 11. The method of claim 10, further comprising determining a focal height of a CL optical system relative to the sample before performing the first scan, and varying the height of the sample using a height sensor while performing the second scan to keep the focal height constant.

15. The method of claim 1, wherein performing the first scan includes generating a first set of data points; the CL image is formed using the first set of data points; performing the second scan generates a second set of data points, the second set having fewer data points than the first set; 2. The method of claim 1, comprising generating a second time series of CL images using the second set of data points; and determining parameters of changes observable during transient periods in the second time series of CL images.

16. The method of claim 15 , wherein the parameters include at least one of a shift in peak wavelength, a shift in peak amplitude, a rate of shift in peak wavelength, and a rate of shift in peak amplitude.

17. 16. The method of claim 15, wherein the second set of data points is collected from an area of ​​the second sample that is smaller than the area used to obtain the first set of data points.

Images