Method for non-destructively characterizing a sample using an instantaneous signal difference and metrology device thereof

Abstract

A measuring device detects an embedded structure, such as a void or other subsurface structure, in a sample based on a transient signal difference determined from a single signal. The single signal acquisition is generated using a series of primary pump pulses and a series of secondary pump pulses, the pulses being intensity modulated and out of phase. The primary and secondary pump pulses are combined to form a pump beam incident on the sample, thereby inducing a transient disturbance in the material within the sample. Detection pulses are also incident on the sample, each detection pulse modulated by the combined transient disturbance caused by a previous primary and secondary pump pulse. A series of reflected detection pulses are detected and demodulated to determine a transient signal difference generated in response to the combined primary and secondary pump pulses, from which the embedded structure is detected.

Description

[Technical Field]

[0001] Cross-reference to related applications

[0002] This application claims priority to U.S. Provisional Patent Application No. 63 / 636,334, filed April 19, 2024, entitled “Metrology Based on Time Resolved Non-Acoustic Signals”, 35 USC §119, the entire contents of which are incorporated herein by reference.

[0003] The subject matter described herein is generally related to microscopy, and more specifically to the use of time-resolved reflectance and deflection measurements. [Previous Technology]

[0004] Inspecting and measuring materials or products to help ensure the quality of said products during manufacturing is a useful step. This is also true for semiconductor wafers or similar products that include microscopic components that are difficult to measure. Metrology systems have previously been used to measure such wafers. Improvements to said metrology systems, including in terms of power consumption, throughput, and performance, are beneficial.

[0005] The forms disclosed herein are in relation to these and other general considerations. Furthermore, although relatively specific problems may be discussed, it should be understood that the examples are not limited to solving specific problems identified in the prior art or elsewhere in this disclosure. [Summary of the Invention]

[0006] Typically, photoacoustic metrology uses a pump bundle and a probe bundle, each containing light pulses with different time delays. The light pulses in the pump bundle generate an acoustic response within the sample under test that propagates to the sample's surface, which is detected by the probe bundle. For example, the acoustic response affects the reflectivity of the material in the sample or the deflection of the probe bundle. The different time delays between the light pulses are used to produce a time-resolved reflectance measurement of the sample. Photoacoustic metrology is useful because it allows for non-destructive detection and measurement of underlying structures within a sample that may be difficult to detect by other means. However, photoacoustic metrology is a relatively slow procedure because the time delay between the light pulses in the pump bundle and the probe bundle must vary with the range of time delays at each measurement point to obtain a time-resolved reflectance measurement. However, as discussed herein, photoacoustic metrology devices use multiple pump bundles and an instantaneous signal difference, where the pulses in each pump bundle have a fixed delay, and this instantaneous signal difference is based on the difference between the signals generated by the two pump bundles. For example, two pump bundles with two discrete delay times can be generated by splitting one pump bundle into a main pump bundle and a secondary pump bundle, which have different path lengths and are recombined before being incident on the sample. Using multiple pump bundles and instantaneous signal differences can increase measurement speed and throughput while maintaining the desired measurement accuracy.

[0007] Embedded structures (such as voids, inclusions, or other underlying structures) are detected using a time-resolved reflectivity meter based on the instantaneous signal difference determined from a single time-resolved transient signal. The time-resolved transient signal is generated using a series of main pump pulses and a series of secondary pump pulses, which are intensity-modulated and phase-opposite, and have a fixed time delay between them. Each of the main pump pulses and the series of secondary pump pulses generates a transient disturbance in the material of the sample. A corresponding series of probe pulses are also incident on and reflected from the sample, wherein each reflected probe pulse is modulated by a transient disturbance caused by both a previous main pump pulse and a previous secondary pump pulse. Each probe pulse has a time delay relative to the previous main pump pulse and a different time delay relative to the previous secondary pump pulse, and therefore each reflected probe pulse is simultaneously modulated by transient disturbances from different depths in the sample. A series of reflected detection pulses are used to detect and demodulate the intensity modulation of the main pump pulse and the secondary pump pulse to determine a transient signal difference generated in response to the combined main pump pulse and secondary pump pulse. This transient signal difference is used to determine a characteristic of the sample, such as the presence or absence of an embedded structure.

[0008] In one embodiment, a method for non-destructively characterizing a sample using a transient signal difference, the transient signal difference being a difference between signals generated from different delay times of multiple pump beams acquired with a single probe beam, the method comprising: irradiating the sample with a pump beam comprising a series of main pump pulses and a series of secondary pump pulses. The series of main pump pulses and the series of secondary pump pulses are intensity modulated and out of phase, and the secondary pump pulses have a fixed delay relative to the main pump pulses. Each main pump pulse and each secondary pulse induces a transient perturbation in the material of the sample. The method further comprises: irradiating the sample with a probe beam comprising a series of probe pulses. Each probe pulse has a first delay relative to a previous main pump pulse and a second delay relative to a previous secondary pump pulse, the second delay being different from the first delay. The series of probe pulses is reflected from the sample as a series of reflected probe pulses, wherein each reflected probe pulse is modulated by the transient perturbation in the material of the sample induced by both the previous main pump pulse and the previous secondary pump pulse. The method further includes: detecting an instantaneous signal difference in one of the series of reflected detection pulses generated in response to the series of main pump pulses and the series of secondary pump pulses; and determining a characteristic of the sample based on the instantaneous signal difference.

[0009] In one embodiment, a metrology device for non-destructively characterizing a sample using a transient signal difference includes a pump arm that irradiates the sample using a pump beam comprising a series of main pump pulses and a series of secondary pump pulses, the transient signal difference being a difference between signals generated from different delay times of multiple pump beams acquired with a single probe beam. The series of main pump pulses and the series of secondary pump pulses are intensity modulated and out of phase, and the secondary pump pulses have a fixed delay relative to the main pump pulses. Each main pump pulse and each secondary pulse induces a transient disturbance in the material of the sample. The metrology device further includes a probe arm that irradiates the sample using a probe beam comprising a series of probe pulses. Each probe pulse has a first delay relative to a previous main pump pulse and a second delay relative to a previous secondary pump pulse, the second delay being different from the first delay. The series of probe pulses is reflected from the sample as a series of reflected probe pulses, wherein each reflected probe pulse is modulated by the transient disturbance in the material of the sample caused by both the previous main pump pulse and the previous secondary pump pulse. The metering device further includes: a detector that detects the series of reflected probe pulses to determine an instantaneous signal difference among the series of reflected probe pulses generated in response to the series of main pump pulses and the series of secondary pump pulses; and at least one processor coupled to the detector and configured to determine a characteristic of the sample based on the instantaneous signal difference.

Implementation Method

[0011] Non-destructive measurement and inspection techniques can be used to ensure proper handling of semiconductors or other similar devices. For example, during processing, a series of manufacturing steps can be performed in which layers such as insulating layers, polysilicon layers, and metal layers are deposited and patterned. In another example, during processing, advanced packaging processes can be used to interconnect two or more devices during packaging. During processing (e.g., manufacturing and packaging), desired or undesired embedded structures can be created in a sample, such as structures under one or more layers. The sample can be a wafer, panel, or any type of substrate. Using non-destructive metrology techniques to detect or measure such structures can be necessary or desirable to ensure proper handling, enable the resulting devices to operate correctly, and improve yield.

[0012] For example, during advanced packaging processes, several physical and chemical processes can be used to bond two or more wafers or substrates together. During the bonding process, structures such as solid structures, voids, or inclusions may be intentionally or unintentionally formed between the bonding layers. The presence of such structures can be detected or measured during processing. For example, structures can be used to ensure proper wafer alignment. Similar to manufacturing processes, structures can be formed through a series of processing steps, such as the deposition and patterning of material layers such as insulating layers, polysilicon layers, and metal layers. Successful formation, patterning, and alignment of successive layers during manufacturing and packaging are sometimes critical for the proper operation of such devices. Furthermore, unintentionally formed structures can affect the final performance of the device and thus may adversely affect overall yield. If unwanted characteristics, such as incorrect alignment or unwanted structures, are detected, the bonded wafers may need to be reprocessed before additional processing, such as polishing, can be performed.

[0013] Various conventional optical techniques exist that can be used for nondestructive metrology or inspection of devices during processing (e.g., during manufacturing or packaging). Nondestructive techniques for metrology or inspection of devices during processing (e.g., during manufacturing or packaging) often rely on the use of light. For example, conventional techniques may use light of specific wavelengths, such as ultraviolet (UV), visible light, or infrared (IR), to image or otherwise detect one or more structures produced during processing.

[0014] Generally, to conventionally detect or image embedded structures (e.g., structures beneath one or more layers), light with a wavelength suitable for penetrating the overlying layer is used. However, conventional optical techniques are sometimes unsuitable for measuring or detecting embedded structures when the structure is located beneath an optically opaque layer or structure, or when it is optically transparent to the light of the specific wavelength used. In some cases, for example, the overlying layer or the layer in which the structure is formed may be formed of an opaque material, such as when the material is metallic. In such cases, it may not be conventionally possible to image the embedded structure. As another example, voids or inclusions may be embedded between layers beneath a solid layer of silicon (Si) (e.g., 750 µm). Such structures may be difficult to detect or image because, for example, the structure of voids or inclusions is optically transparent to light. In some cases, infrared imaging may be possible to image such structures unless the structure is covered by an opaque layer (e.g., a metallic layer). Unfortunately, the limited resolution of infrared imaging technology makes it generally unsuitable even for gaps not covered by a metal layer.

[0015] One type of non-destructive metrology technique that can be used to detect voids is confocal scanning acoustic microscopy (C-SAM), which uses acoustic signals. Unfortunately, in order to properly transmit acoustic signals using C-SAM, the sample is immersed in water, which is generally undesirable for many samples (such as semiconductors or other similar devices). Furthermore, C-SAM cannot image relatively small voids (e.g., smaller than 10 µm), thus limiting its applications.

[0016] Photoacoustic metrology, such as picosecond acoustic microscopy (PAM), is commonly used to detect and measure embedded structures, including voids, inclusions, or solid structures. For example, in hybrid bonding processes, interfacial voids generated during wafer-to-wafer, wafer-to-wafer, and wafer-to-wafer processes can be detected and imaged using photoacoustic metrology. Typically, photoacoustic metrology uses pump and probe beams with different time delays between the light pulses in the pump and probe beams to produce time-resolved reflectance measurements. Metrological apparatuses performing time-resolved reflectance measurements can implement variable time delays between pump and probe pulses, for example, by using a mechanically shifted delay line, which, for example, changes the length of the beam path of the pump or probe beam; or an asynchronous optical sampling (ASOPS) configuration in which two synchronized light sources (e.g., lasers) with slightly different repetition rates produce variable time delays without using a mechanically shifted delay line. A light pulse in a pump beam propagating to the surface of the sample under test generates an acoustic response within the sample, which is detected by a probe beam after a delay. For example, the acoustic response affects the reflectivity of the material in the sample or the deflection of the probe beam. Different time delays between light pulses are used to produce a time-resolved reflectance measurement of the sample. Photoacoustic metrology is useful because it allows for non-destructive detection and measurement of underlying structures within the sample that may be difficult to detect by other methods. However, photoacoustic metrology is a relatively slow procedure because the time delay between light pulses must vary with the range of time delays at each measurement point in order to obtain a time-resolved reflectance measurement before moving to the next measurement point and repeating the procedure.

[0017] As discussed herein, photoacoustic metrology systems (such as PAM) use multiple pump bundles, each with a fixed delay between pulses, and acquire instantaneous signal differences (e.g., signal differences from two discrete delay times) using a single probe beam. By using two or more fixed time delays from multiple pump bundles, rather than different time delays and instantaneous signal differences from a single pump bundle, measurement speed and throughput can be increased while maintaining the desired measurement accuracy. In some embodiments, the probe beam pulse and the pulses in the multiple pump bundles may have different time delays. In some embodiments, the probe beam pulse and the pulses in the multiple pump bundles may have fixed time delays, such that there are multiple fixed time delays between the pump pulse and the probe pulse, rather than different time delays. Signals received with two or more fixed time delays can have sufficient information and sensitivity to distinguish the presence or absence of embedded structures (e.g., voids, inclusions, or solid structures) in the sample.

[0018] By using a fixed time delay between pulses in the pump beam, an instantaneous signal difference, such as the signal difference from two discrete delay times acquired with a single probe beam, can be obtained. For example, the signal difference can be obtained by collecting conventionally rejected modulated pump laser pulses and reusing these pump laser pulses as a secondary pump pulse train in combination with the primary pump pulse train. For example, the pump pulses can be intensity modulated using an optical modulator, such as an electro-optic modulator (EOM), or other suitable modulator, followed by a linear polarizing element. The pulses in the primary and secondary pump trains (e.g., using the difference due to the separate beam paths) are delayed relative to each other to produce two time delays different from those in the probe pulse train. The pump-probe time delays of the primary and secondary pump pulse trains can be fixed during measurement but can be varied, for example, by changing the length of one or more beam paths between measurements, for example, to increase sensitivity.

[0019] The modulator and polarization element apply intensity modulation to the main pump pulse train and the secondary pump pulse train before they are combined and directed to the sample. Both the main pump pulse train and the secondary pump pulse train are intensity modulated but out of phase, i.e., 180 degrees out of phase with each other. At the measurement position, the probe pulse train, also pointing to the sample, interacts with the photoexcited sample and picks up the pump modulation frequency through acoustic generation and detection, as well as other transient procedures driven by the pump excitation, which is generated by the combined main pump pulse train and the secondary pump pulse train. The signal generated by the probe pulse pump can be demodulated to determine the instantaneous signal difference resulting from the different time delays of the main pump pulse train and the secondary pump pulse train.

[0020] Using instantaneous signal differential is advantageous compared to sequential differential procedures. For example, a sequential differential procedure acquires signals at different times with a first time delay and a second time delay, and then calculates the difference between the signals. Using instantaneous signal differential, the same signal is generated in response to both the first and second time delays in the main probe pulse pump and the secondary probe pulse. Therefore, by using simultaneous acquisition of instantaneous signal differential, results can be obtained in approximately half the time of a sequential procedure. Furthermore, simultaneous acquisition ensures that the precise positions of the main probe pulse pump and the secondary probe pulse are consistent at each measurement position of the scan (lateral position on the sample), which is particularly important for achieving the maximum benefit of using the differential signal to suppress unwanted signal artifacts.

[0021] The signal differential procedure discussed herein is compatible with both: a null-difference configuration using a single light source that generates light split into pump and probe arms and uses a mechanical delay line to generate a pump-probe delay; and a heterodyne configuration using two light sources (e.g., lasers) that have slightly different repetition rates and are electronically synchronized to generate a pump-probe delay without a mechanical delay line, such as asynchronous optical sampling (ASOPS). The two systems can be configured to set different fixed pump-probe delays for the primary and secondary pump pulses, while in other embodiments, the pump-probe delays of the primary and secondary pump pulses can be varied during measurement to obtain multiple instantaneous signal differences under different pump-probe delays. Either system can be amplified to capture and reuse discarded pump pulses from the intensity modulation unit, thereby achieving instantaneous signal differences, as described herein, provided the configuration includes pump intensity modulation.

[0022] Figure 1 illustrates a schematic diagram of an example time-resolved measurement device 100 configured to perform time-resolved reflectivity measurements using instantaneous signal differential, as discussed herein. For example, signal differential can be used to detect and image embedded structures, such as voids, inclusions, and solid structures.

[0023] As illustrated, device 100 includes a light source 102 that generates a beam comprising a series of light pulses. For example, light source 102 may be a laser, such as a 520 nm, 200 fs, 60 MHz laser, but other types of light sources or other characteristics may also be used. For example, device 100 may use a light source that operates in the infrared wavelength range for example, for imaging embedded structures. The pulses in the beam may be generated in various ways, such as by a pulsed laser, or in some embodiments by a chopper located outside the laser but which may be considered part of light source 102. The light generated by light source 102 may be guided through intensity control 103 and through beam expander 104, the intensity control including a half-wave plate HWP1 and a polarizer P1. The beam may be guided by one or more optical elements, such as a mirror M1, to beam splitter 106, which splits the light into a pump beam in pump arm 120 and a probe beam in probe arm 130.

[0024] In the pump arm 120, the pump bundle is guided to the pump bundle optical modulator 122 by the mirror M2. The pump bundle optical modulator 122 may be, for example, an electro-optic modulator (EOM) or other suitable modulator, thereby modulating the pump bundle intensity at a desired frequency, which may be in the range of several megahertz (MHz), such as about 5 or 5.5 MHz, but other frequencies may also be used.

[0025] The modulated pump beam is received by the pump beam splitter 124. The pump beam splitter 124 splits the modulated pump beam into a main pump beam traveling along the main pump beam path 125 and a secondary pump beam traveling along the secondary pump beam path 126. The main and secondary pump beams are intensity modulated by the pump beam optical modulator 122, but the intensity modulation of the secondary pump beam is due to the pump beam splitter 124 being 180 degrees out of phase with respect to the main pump beam.

[0026] Conventionally, secondary pump beams are rejected by the system, for example, by being received by a beam gather. However, as illustrated in FIG1, a secondary pump beam is used in metrology device 100 and recombined with the primary pump beam using beam splitter 129. As illustrated, for example, the secondary pump beam travels along a secondary pump beam path 126, where it is guided by various guide mirrors, such as M3 and M4, to an optical delay 128, which creates a delay between pulses in the primary and secondary pump beams. For example, optical delay 128 is illustrated by mirrors M5 and M6. Mirror M6 may be a regressor. In some embodiments, mirror M6 may be coupled to an actuator or voice coil that can be controlled to vary the delay generated by optical delay 128. As discussed herein, the delay generated by optical delay 128 may be fixed during sample measurement, such that the delay between pulses in the primary and secondary pump beams is constant, but the delay may be varied between measurements, for example, to improve sensitivity.

[0027] The main pump beam will have a first delay between pump beam splitter 124 and beam splitter 129, while the secondary pump beam will have a different second delay between pump beam splitter 124 and beam splitter 129. The optical delay 128 in the secondary pump beam path 126 can be controlled to set a delay difference relative to the main pump beam path 125, for example, to improve signal sensitivity. Due to the additional reflective elements in the secondary pump beam path 126, such as mirrors M3, M4, M5, and M6, the total path length of the secondary pump beam is greater than the total path length of the main pump beam. Therefore, the pulse in the secondary pump beam will arrive at the focusing optics L1 and eventually the sample 101 after the corresponding pulse in the main pump beam.

[0028] Furthermore, the series of reflections in the secondary pump beam path 126 (e.g., via mirrors M3, M4, M5, and M6) can be configured to produce a 90-degree rotation in the polarization orientation of the secondary pump beam, such that the primary and secondary pump beams have the same polarization orientation when combined by beam splitter 129. In some embodiments, an additional mirror can be positioned in the primary pump beam path 125 between pump beam splitter 124 and beam splitter 129 to assist in controlling the relative polarization orientation of the primary and secondary pump beams and controlling the first delay in the primary pump beam path 125. Additionally, a polarizing element P2, such as a polarizer or waveplate, can be positioned in the secondary pump beam path 126 prior to beam splitter 129 to ensure that the primary and secondary pump beams have the same polarization orientation. After recombination by beam splitter 129, the primary and secondary pump beams are collinear, have the same polarization, are intensity modulated at the same frequency but with opposite phases, and the pulses in the secondary pump beam are delayed relative to the pulses in the primary pump beam.

[0029] The pump bundle, i.e., the combined main pump bundle and secondary pump bundle, is guided to the focusing unit 140 by beam deflection mirrors (e.g., mirrors M7, M8, and M9). At least one of mirrors M7, M8, and M9 may be attached to a piezoelectric motor to adjust the direction of the pump bundle. As illustrated in FIG1, the focusing unit 140 may include a beam splitter 142 that guides the pump bundle to be incident normally onto the sample 101 through lens L1. Lens L1 focuses the pump bundle over the region of the sample 101 that includes the structure to be imaged. In some embodiments, the pump bundle may be guided to be incident obliquely onto the sample 101 (e.g., along the same beam path as the probe beam focused by lens L2).

[0030] In the probe arm 130, after the beam splitter 106, the probe beam can pass through the half-wave plate HWP2, which can be powered to rotate. The probe beam can be guided to an optical delay 132, which includes mirrors M10, M11, M12, and M13. For example, mirror M12 may be a regressor and may be coupled to an actuator or voice coil that can be controlled to vary the delay of the probe beam relative to the pump beam. The probe delay 132 can be used to control the delay between pulses in the pump beam and pulses in the probe beam, i.e., the delay between pulses in the main pump beam and pulses in the secondary pump beam. In some embodiments, the delay 132 may be located in the pump arm 120, for example, before the optical modulator 122 rather than in the probe arm 130. In some embodiments, separate delays may be located in both the pump arm 120 and the probe arm 130. The detection delay 132 can remain stationary during the measurement of a fixed pump-detection delay, such that the pulses in the detector beam have a first fixed delay relative to the pulses in the main pump beam and a second different fixed delay relative to the pulses in the secondary pump beam. In some embodiments, the detection delay 132 can move during the measurement of different pump-detection delays, such that the pulses in the detector beam have a variable delay relative to the pulses in the main pump beam and a variable delay relative to the pulses in the secondary pump beam, wherein there is a fixed delay between the pulses in the main pump beam and the secondary pump beam. The detector beam is guided to the focusing unit 140 by beam deflection mirrors (e.g., mirrors M14, M8, M15). At least one of mirrors M14, M8, and M15 can be attached to a piezoelectric motor to adjust the direction of the detector beam. As illustrated in FIG1, the focusing unit 140 can guide the detector beam through lens L2 to obliquely incident on the sample 101. Lens L2 focuses the detector beam over the region of the sample 101 including the structure to be imaged and can coincide with the incident region of the pump beam. In some implementations, similar to a pump beam, the probe beam may pass through a modulator (e.g., an EOM) and then through a polarizer to modulate the probe beam in intensity (e.g., using a frequency comb different from the pump beam). Additional optical components, such as those for polarization control, may also be included if needed.

[0031] Lenses L1 and L2 may be configured, for example, to irradiate sample 101 using a pump beam and a probe beam. The pump beam and the probe beam may coincide at the same measurement location on sample 101. In some embodiments, the measurement location may be at least the size of the structure under test on sample 101, so that scanning is not required to detect the desired structure (such as alignment or overlay patterns). For example, in some embodiments, lenses L1 and L2 may have a focal area greater than 10 µm. In some embodiments, the measurement location may be laterally scanned above the surface of sample 101 by, for example, using a stage 105 holding sample 101 to create relative movement between sample 101 and the optical system, so that various positions on sample 101 can be measured.

[0032] The reflected detection beam (and optionally the reflected pump beam, if obliquely incident) is received by collecting optics 150, which includes, for example, lens L3 and mirrors M16 and M17. The reflected beam is guided to detector 160 via lens 164. Detector 160 is a photodetector or a multi-pixel array of photodetectors. An image of sample 101 may be generated, for example, using the multi-pixel array in detector 160 (if present) or by scanning sample 101 (and / or optics) to a plurality of locations and performing measurements at each separate location.

[0033] Detector 160 may be coupled to demodulator 162, such as a lock-in amplifier configured for phase locking during signal acquisition. In some embodiments, detector 160 may be locked to a camera including a multi-pixel array and independent phase locking for each pixel in the multi-pixel array. Phase locking is used to demodulate the received probe beam based on the intensity modulation frequency of the pump beam, or, if the probe beam is also modulated, to demodulate the received probe beam based on a combination of frequencies in the pump beam and the probe beam. In embodiments where both the pump pulse in the pump beam and the probe pulse in the probe beam are modulated at two different frequency combs (e.g., by modulators in both pump arm 120 and probe arm 130), phase locking can be used to demodulate frequency combinations (e.g., sum or difference). Furthermore, phase locking can generate in-phase and quadrature measurements from the image. Detector 160 may record changes in the reflectivity of sample 101 based on instantaneous signal differences caused by a fixed time delay between the primary pump pulse and the secondary pump pulse. By measuring reflectance, the reflectance or deflection of sample 101 can be determined as an instantaneous signal difference. For example, the instantaneous signal difference can be a change in differential reflectance or reflectance measurement (ΔR / R), which is attributed to the presence of strain in the material of sample 101 and the associated change in optical constants; or a surface or interface deflection measurement, which is attributed to the physical deflection of the beam caused by strain at the surface or interface of sample 101. It should be understood that, for ease of reference, the reflectance or deflection of the sample may sometimes be collectively referred to as reflectance in this document.

[0034] Furthermore, the time resolution metering device 100 may be coupled to the imaging device 144, which may be configured to image the top structure of the sample 101 via the beam splitter 142 and the lens L1. For example, the imaging device 144 may be a flight path camera.

[0035] Sample 101 is held on a stage 105, which includes or is coupled to one or more actuators configured to move sample 101 relative to the optical system of the time-resolved reflectance metrology apparatus 100, thereby enabling measurement of various positions on sample 101. In the depicted embodiments, the apparatus may include additional components and subsystems, such as beam management and adjustment components (e.g., beam expanders, collimators, polarizers, half-wave plates, etc.), and beam power detectors and height detectors. Those skilled in the art will understand that variations of the apparatus depicted in FIG1 will still be suitable for implementing the time-resolved reflectance metrology techniques described herein.

[0036] Detector 160 and other components of the time-resolved reflectance metering device 100 may be coupled to processing system 170 (such as a workstation, personal computer, central processing unit, or other suitable computer system, or multiple systems). It should be understood that a single processor, multiple separate processors, or multiple interconnected processors may be used, all of which are interchangeably referred to herein as processing system 170. Processing system 170 is preferably included in, connected to, or otherwise associated with the time-resolved reflectance metering device 100. For example, processing system 170 may control the positioning of sample 101, for example, by controlling the movement of pedestal 105 on which sample 101 is held. For example, pedestal 105 may be able to move horizontally in either Cartesian (i.e., X and Y) coordinates or polar (i.e., R and θ) coordinates or a combination of both. Pedestal 105 may also be able to make vertical movements along the Z coordinate. The processing system 170 can further control the operation of the clamps on the pedestal 105 used to hold or release the sample 101.

[0037] Processing system 170 can collect and analyze data obtained from detector 160 and demodulator 162. In some embodiments, processing system 170 can act as demodulator 162. Processing system 170 can analyze time-analytical measurement data to detect and image embedded structures in sample 101, such as voids, inclusions, and solid structures. For example, in some embodiments, the underlying structure can be detected and imaged based on analysis of transient signal differences to distinguish various properties or characteristics of transient signals from different locations. The properties or characteristics of transient signals from different locations can be used to determine whether the measurement was performed at a location with voids or without voids. For example, transient signals can be acoustic transient signals or non-acoustic transient signals, i.e., signals in which the contribution of any acoustic signal is less than the contribution of other physical phenomena such as one or more of the following: heat dissipation, electron-hole recombination, plasma diffusion, symmetry breaking of top and bottom oxide surfaces, etalon effect, etc. The underlying structure can be detected based on a comparison of signal differences generated from transient signals from multiple different locations. The processing system 170 may alternatively or additionally process time-resolved reflectance measurement data for edge detection or triangulation (e.g., using a classification library or neural network generated on a reference sample by the time-resolved reflectance measurement device 100 (or another device)).

[0038] Processing system 170 may include at least one processor 172 and memory 174, and a user interface 176 may include a display and input devices such as a keyboard and mouse, all of which may be interconnected via bus 171. A non-transitory computer-usable storage medium 178 having embodied computer-readable code may be used by processing system 170 to enable processing system 170 to control time analysis metering device 100 and perform functions including the analysis described herein. Data structures, software code, etc., for automatically implementing one or more of the actions described in this embodiment may be implemented by those skilled in the art in accordance with this disclosure and stored, for example, on computer-usable storage medium 178, which may be any device or medium capable of storing code and / or data for use by a computer system (such as at least one processor 172). Computer-usable storage medium 178 may be, but is not limited to, flash drives, magnetic and optical storage devices, such as disk drives, magnetic tapes, optical discs, and DVDs (Digital Multifunction Discs or Digital Video Discs). Communication port 179 can also be used to receive instructions for programmable processing system 170 to perform any or more of the functions described herein, and can represent any type of communication connection such as to the Internet or any other computer network. Communication port 179 can further output signals (e.g., measurement or detection results and / or instructions) in a feedforward or feedback procedure to another system (e.g., an external processing tool) to adjust program parameters associated with the program steps of the sample and to provide network instructions. Additionally, the functions described herein can be embodied, wholly or partially, in a circuit system of an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), and these functions can be represented in a computer-understandable descriptor language that can be used to build an ASIC or PLD operating as described herein. Results from data analysis can, for example, be stored in memory 174 associated with the sample and / or provided to the user (e.g., via UI 176, an alarm, or other output device).

[0039] Figure 2 illustrates a block diagram of another example of a time-resolved measurement device 200, which is configured to perform time-resolved reflectivity measurements using instantaneous signal differential, as discussed herein.

[0040] The device 200 includes: a pump arm 210 including a pump laser 211 (also referred to herein as an excitation laser); a detector arm 220 including a detector laser 221 (also referred to herein as a detection laser); and optical elements such as a steering mirror 222 and a beam splitter 224. The device 200 further includes lenses 236 and 238, filters, polarizers, and the like (not shown), each of which directs light from the pump laser 211 and the detector laser 221 to a sample 201 including an embedded structure 202 to be detected or imaged. The pump laser 211 and the detector laser 221 can be synchronized at different repetition rates to produce different pump-detection delays in an asynchronous optical sampling (ASOPS) configuration. In some embodiments, the pump laser 211 and the detector laser 221 can be synchronized at the same repetition rate to produce a fixed pump-detection delay during measurement.

[0041] Similar to the metering device 100 described in Figure 1, the pump arm 210 includes an optical modulator 212, which may be an electro-optic modulator (EOM) or other suitable modulator, to intensity modulate the pump bundle at a desired frequency, which may be in the range of several megahertz (MHz), such as about 5 or 5.5 MHz, but other frequencies may also be used. In some embodiments, a second modulator may be present in the probe arm 220 to modulate the probe pulse using a different modulation frequency, for example, different frequency combs may be used. The pump arm 210 further includes a beam splitter 213 that receives the modulated pump bundle from the modulator 212 and splits the modulated pump bundle into a main pump bundle traveling along a main pump bundle path 214 and a secondary pump bundle traveling along a secondary pump bundle path 215. The main and secondary pump bundles are intensity modulated at the same frequency by the modulator 212, but are out of phase by the pump bundle splitter 213.

[0042] The secondary pump beam is recombined with the primary pump beam via beam splitter 218. For example, as illustrated, the secondary pump beam travels along a secondary pump beam path 215, illustrated by mirrors 216 and 217, which includes an optical retarder 219 that generates a delay between pulses in the primary and secondary pump beams. If desired, the optical retarder 219 may include a regressor that can be coupled to an actuator or voice coil to controllably vary the delay generated by the optical retarder 219. As discussed herein, the delay generated by the optical retarder 219 may be fixed during the measurement of a sample, such that the delay between pulses in the primary and secondary pump beams is constant, but the delay may be varied between measurements, for example, to improve sensitivity.

[0043] Therefore, the optical retarder 219 in the secondary pump beam path 215 can be controlled to set a delay difference relative to the primary pump beam path 214. Due to the additional reflective elements illustrated by mirrors 216 and 217 in the secondary pump beam path 215, the total path length of the secondary pump beam is greater than the total path length of the primary pump beam. Therefore, the pulse in the secondary pump beam will arrive at the focusing optics 236 and eventually the sample 201 after the corresponding pulse in the primary pump beam.

[0044] Furthermore, the series of reflections in the secondary pump beam path 215 (e.g., illustrated by mirrors 216 and 217) can be configured to produce a 90-degree rotation in the polarization orientation of the secondary pump beam, such that the primary and secondary pump beams have the same polarization orientation when combined by beam splitter 218. In some embodiments, mirrors can be positioned in the primary pump beam path 214 to assist in controlling the relative polarization orientation of the primary and secondary pump beams and controlling the delay in the primary pump beam path 214. Additionally, polarization elements, such as polarizers or waveplates, can be positioned in the secondary pump beam path 215 prior to beam splitter 218 to ensure that the primary and secondary pump beams have the same polarization orientation. After recombination by beam splitter 218, the primary and secondary pump beams are collinear, have the same polarization, are intensity modulated at the same frequency but opposite in phase, and the pulses in the secondary pump beam are delayed relative to the pulses in the primary pump beam.

[0045] The device 200 may include optical components (such as a beam splitter 225 and a steering mirror 227) and may include a beam stack 226 for capturing radiation from the pump laser returning from the sample 201. The device 200 includes a detector 228 that detects changes in reflectivity from the reflected probe beam, such as changes attributable to changes in reflectivity of the sample 201 or surface deformation.

[0046] In some embodiments, the detector 228 may be coupled to the demodulator 229, such as a lock-in amplifier configured for phase locking during signal acquisition. In some embodiments, the detector 228 may be a lock-in camera, which includes a multi-pixel array and independent phase locking for each pixel in the multi-pixel array.

[0047] Phase locking is used to demodulate the received probe beam based on the intensity modulation frequency of the pump beam, or, if the probe beam is also modulated, to demodulate the received probe beam based on a combination of frequencies in the pump beam and the probe beam. In an embodiment where both the pump pulse in the pump beam and the probe pulse in the probe beam are modulated at two different frequency combs (e.g., by modulators in both pump arm 210 and probe arm 220), phase locking can be used to demodulate the frequency combination (e.g., sum or difference). Furthermore, phase locking can generate in-phase and quadrature measurements from the image. Detector 228 records the reflectivity of sample 201 based on the instantaneous signal difference caused by a fixed time delay between the main pump pulse and the secondary pump pulse.

[0048] The apparatus 200 further includes an electromechanical support pedestal 205 for a sample 201 having an embedded structure 202, the pedestal 205 being adapted to move the sample 201 relative to the pump laser 211 and the probe laser 221 to obtain measurements at multiple locations, for example, by sequentially scanning a grid.

[0049] Detector 228 and other components of the time-resolved reflectivity metering device 200 may be coupled to processing system 230, which may be similar to processing system 170 discussed with reference to FIG. 1. It should be understood that processing system 230 may be a self-contained or distributed computing device capable of performing calculations, receiving and sending instructions or commands, and receiving, storing, and sending information related to the metering function of the device.

[0050] In the depicted embodiments, the pump laser 211 and the probe laser 221 in the embodiment of the time-analysis metering device 200 shown in FIG. 2 may share at least a portion of the optical path leading to and from the sample 201. For example, the lasers may have several different relative arrangements, including configurations in which the paths are identical, partially overlapping, adjacent, or coaxial. In some embodiments, the pump beam and the probe beam may be derived from the same pulsed laser. In some embodiments, as illustrated in FIG. 2, separate lasers may be used for the pump beam and the probe beam; for example, separate synchronized lasers with slightly different repetition rates may be used in an ASOPS configuration. In other embodiments, the pump laser 211 and the probe laser 221, as well as the beam stack 226 and the detector 228, do not share optical paths. For example, the pump beam from the pump laser 211 may be incident normally onto the sample 201, while the probe beam from the probe laser 221 may be incident obliquely onto the sample 201. Pump laser 211 and detector laser 221 can be directly controlled to obtain the time interval between light pulses guided to structure 202.

[0051] It should be understood that many optical configurations are feasible. In some configurations, the pump may be a pulsed laser with a pulse width in the range of hundreds of femtoseconds to hundreds of nanoseconds, and the probe beam is coupled to a beam deflection system. For example, in some embodiments, pump arm 210 and / or probe arm 220 may include a mechanical delay stage (not shown) to increase or decrease the length of the optical path difference between the pump beam and the probe beam. Where a delay stage is provided, the delay stage will be controlled by processing system 230 to obtain and control the time delay between the pump pulse and the probe pulse incident on the object. Many other alternative configurations are also feasible. In other embodiments (such as those with ASOPS configurations), the device may not include a delay stage. It should be understood that the schematic diagram of FIG2 is not intended to be limiting, but rather to depict one of several example configurations.

[0052] In operation, the time-analyzing meter 200 directs a series of primary pump pulses and a series of secondary pump pulses from the pump laser 211 to the structure 202. These light pulses are incident on the sample 201 at any angle, for example, between zero and 90 degrees (including, for example, 45 degrees and 90 degrees). If the sample 201 includes at least a partially absorbing transducer layer (e.g., a metallic layer) above the structure 202, the primary and secondary pump pulses from the pump laser 211 are at least partially absorbed, resulting in transient expansion (i.e., an acoustic signal) in the material of the transducer layer. The expansion is short enough that it induces ultrasonic waves that propagate substantially vertically through the structure 202 and are reflected at the respective underlying interfaces and return to the top surface. The light reflected from the pump laser 211 from the structure 202 is transmitted to the beam stack 226, which attenuates or absorbs the pump radiation.

[0053] If sample 201 does not include strongly absorbing materials (such as a metallic layer), but only includes materials that are optically transparent to the wavelength used by the pump laser 211, then there may be no (or only a tiny) transient spread, i.e., an acoustic signal. However, the non-acoustic transient signal in sample 201 is generated in response to the main pump pulse and the secondary pump pulse from one or more different physical phenomena such as: heat dissipation, electron-hole recombination (e.g., possibly generated in Si), plasma diffusion, symmetry breaking of the top and bottom oxide surfaces, etalon effect within the voids, etc. In the absence of a strongly absorbing material to generate an acoustic signal, the non-acoustic contribution to the returned signal becomes more prominent and is sensitive to the presence of structures (e.g., voids in the oxide layer).

[0054] In addition to guiding the operation of the pump laser 211, the processing system 230 guides the operation of the probe laser 221. The probe laser 221 guides radiation incident on the substrate 201 in the form of a series of probe pulses, which are reflected from the sample 201 and affected by the resulting transient signals, such as reflected acoustic signals if the sample 201 includes strong absorbing materials to generate acoustic signals, or non-acoustic transient signals if strong absorbing materials are not present in the sample 201.

[0055] The device 200 includes optics (such as lens 236) that can be configured to adjust the dot size of the pump and probe beams. The dot sizes of the individual beams may be similar or dissimilar. For example, the optics (such as lens 236) may be configured, for example, to adjust the focal area of ​​the pump and probe pulses on the sample 201 to include the size of a plurality of locations to be measured, using a lock-on camera including a multi-pixel array as detector 228, or to the size corresponding to a single location if detector 228 is a single-pixel detector and scanning is used to measure a plurality of locations.

[0056] The light reflected from the surface of sample 201 is guided to detector 228, for example, by beam splitter 225. The reflectivity of the reflected probe beam is modulated by a change in reflectivity or surface deformation, which is attributed to the reflected sound wave or non-acoustic transient signals responding to the main pump pulse and secondary pump pulse. Detector 228 may be configured to receive and demodulate the reflected probe pulse, for example, using demodulator 229.

[0057] In embodiments where the detector 228 includes a multi-pixel array, optics such as lens 238 can adjust the magnification of the probe beam on the multi-pixel array for efficiency. The detector 228 may include a demodulator 229 configured for phase locking to acquire transient signals. If the detector 228 includes a multi-pixel array, the demodulator 229 may be configured to perform independent phase locking on each pixel in the multi-pixel array to acquire transient signals in parallel. In some embodiments, the demodulator 229 may be independent of the detector 228 (e.g., in a separate processor or field-programmable gate array (FPGA) or in processing system 230). Phase locking can be used to demodulate the frequency of the pump pulse in the received probe beam. If both the pump pulse and the probe pulse are modulated, the combination (e.g., sum or difference) of frequencies in the received probe beam can be demodulated. The detector 228 can respond to the different time delays between the main pump pulse train and the secondary pump pulse train, for example, to record the change in reflectivity of the sample 201 for each illuminated pixel as a transient signal difference.

[0058] Furthermore, the time-resolved reflectance meter 200 may be coupled to an imaging device 240 configured to image the top surface of the sample 201 (e.g., for alignment or overlay purposes). For example, the imaging device 240 may be a flight path camera. The imaging device 240 may perform optical imaging of the sample 201.

[0059] The metering devices 100 and 200 shown in Figures 1 and 2 are configured to generate a fixed time delay between pulses in the main pump bundle and pulses in the secondary pump bundle. This can be used to obtain the instantaneous signal difference (the signal difference from two discrete delay times) using a single acquisition. Traditionally, a single pump bundle and sequential signal acquisition are used to determine the signal difference, for example, two separate signals are acquired at different times, and the difference between the signals is calculated. For example, in sequential acquisition, a signal is acquired, for example, during a scan over the region of interest on the sample, with a pump-probe delay set to a first time delay A, followed by adjusting the pump-probe delay to a second time delay B and acquiring a second set of signals, for example, during a second scan over the region of interest. The image of the region of interest can be determined based on the signal difference (Diff = AB = A + -1 * B) at each measurement position in the region of interest, where A represents the signal acquired with pump-probe delay A, B represents the signal acquired with pump-probe delay B, and Diff represents the difference.

[0060] The metering devices 100 and 200 shown in Figures 1 and 2 utilize a secondary pump bundle with a fixed delay relative to the main pump bundle. Therefore, the pulses in the main pump bundle and the pulses in the secondary pump bundle have different time delays relative to the same pulses in the probe bundle. In this example, A may represent the signal obtained in response to the pump-probe delay A of the main pump bundle, and B may represent the signal obtained in response to the pump-probe delay B of the secondary pump bundle. The secondary pump bundle is intensity modulated at the same frequency as the main pump bundle, but with a phase difference of 180 degrees. The signal generated by the probe bundle is simultaneously responded to by the combined main and secondary pump bundles. Therefore, the signal obtained in response to the pump-probe delay B of the secondary pump bundle is inverted relative to the signal obtained in response to the pump-probe delay A, and can be represented as -1 * B. By combining the main and secondary pump bundles modulated at the same frequency but with opposite phases, the demodulated signal obtained from the detected probe bundle will be AB, thus providing an instantaneous signal difference.

[0061] Figure 3 is a graph 300 showing the intensity modulation curves of the main pump bundle and the secondary pump bundle in response to the combination of optical modulator 122 and pump bundle splitter 124 shown in Figure 1 and the combination of optical modulator 212 and pump bundle splitter 213 shown in Figure 2. In Figure 3, the X-axis represents time, and the Y-axis represents the intensity between 0 and 1 using arbitrary units (AU).

[0062] In Figure 3, solid curve 310 illustrates the intensity modulation of the main pump beam after it is modulated by optical modulator 122 and split by pump beam splitter 124 in Figure 1, for example, along the main pump beam path 125, or similarly, after it is modulated by optical modulator 212 and split by pump beam splitter 213 in Figure 2, along the main pump beam path 214. On the other hand, dashed curve 320 illustrates the intensity modulation of the secondary pump beam, for example, after it is modulated by optical modulator 122 in Figure 1 and split by pump beam splitter 124, along the secondary pump beam path 126, or similarly, after it is modulated by optical modulator 212 and split by pump beam splitter 213 in Figure 2, along the secondary pump beam path 215. As shown by curves 310 and 320, the intensity modulation of the main pump beam and the secondary pump beam is 180 degrees out of phase.

[0063] It should be understood that, regardless of the continuous curves 310 and 320 shown in Figure 3, the main pump bundle and the secondary pump bundle are not continuous bundles, but rather generate a plurality of optical pulses, sometimes referred to as pulse trains, in which the intensity is modulated as shown by curves 310 and 320. For example, there may be tens, hundreds or thousands of optical pulses within a single period of intensity modulation.

[0064] Due to the fixed delay between the pulses in the main pump bundle and the secondary pump bundle, the pulse in the probe bundle will have a first delay relative to the corresponding pulse in the main pump bundle, and a different second delay relative to the corresponding pulse in the secondary pump bundle. If the time delay between the probe bundle and the pump bundle remains constant during the measurement, the first delay and the second delay will be fixed. If the time delay between the probe bundle and the pump bundle varies during the measurement, the probe bundle will have the first delay and the second delay will vary, but will still maintain the offset to the fixed delay between the pulses in the main pump bundle and the secondary pump bundle.

[0065] Figure 4 illustrates, graphically by way of example, the main pumping beam 410, the secondary pumping beam 420, the probe beam 430, and the detected reflected probe beam 440. Figure 4 illustrates a fixed pump-detection time delay. As illustrated, the main pumping beam 410, the secondary pumping beam 420, and the probe beam 430 are generated by multiple optical pulses. The intensity of the pulses in the main pumping beam 410 is modulated, as shown by the solid curve 412; and the intensity of the pulses in the secondary pumping beam 420 is modulated, as shown by the dashed curve 422. The pulses in the probe beam 430 may not be modulated, but in some embodiments, the probe beam 430 may also be modulated at a different frequency than the main pumping beam 410 and the secondary pumping beam 420.

[0066] Due to the fixed pump-detector time delay shown in Figure 4, each pulse in the main pump bundle 410 has a fixed delay D1 relative to the corresponding pulse in the detector bundle 430. Additionally, due to the longer path length of the secondary pump bundle, each pulse in the secondary pump bundle 420 has a relatively small fixed delay d relative to the corresponding pulse in the main pump bundle 410, as shown in Figures 1 and 2. Furthermore, each pulse in the secondary pump bundle 420 has a fixed delay D2 (D2 = D1 - d) relative to the corresponding pulse in the detector bundle 430. The intensity modulation of the pulses in the secondary pump bundle 420 is 180 degrees out of phase with respect to the main pump bundle 410, but after shifting to a fixed delay d, this delay may be small relative to the intensity modulation period.

[0067] Figure 4 further illustrates the idealized contribution of the main pump bundle and the secondary pump bundle to the detection intensity of the reflected detection beam 440. Each pulse in the reflected detection beam 440 is generated in response to the main pump bundle 410 and the secondary pump bundle 420. The instantaneous signal difference from the contributions of the main pump bundle 410 and the secondary pump bundle 420 is recovered by demodulating the reflected detection beam 440 using a frequency based on the intensity modulation of the pump bundles.

[0068] Figures 5A and 5B illustrate the use of transient signal differential detection to detect embedded structures in samples having one or more metal layers. The transient signal differential is derived from time-resolved reflectivity measurements in response to acoustic transient signals. Figure 5A illustrates a sample 500 formed from two bonded wafers 520 and 530, wherein an embedded structure 502 is disposed between wafers 520 and 530. The embedded structure 502 may be, for example, a void or other type of inclusion, and is sometimes referred to herein as void 502, but it should be understood that the embedded structure 502 may be a solid structure. Wafer 520 may, for example, include a silicon substrate 522 (which may have a thickness of 750 µm, serving as the top layer of sample 500) and a metal layer 524 (which may be, for example, copper (Cu), tungsten (W), or titanium (Ti) and may have a thickness of 50 nm). Wafer 530 may similarly include a metal layer 534, which may be, for example, Cu, W, or Ti and may have a thickness of 50 nm. On a silicon substrate 532, the silicon substrate may have a thickness of 750 µm and serve as the bottom layer of sample 500. Layers 532 and 534 form an interface 533. Metal layers 524 and 534 are bonded together at interface 510, as shown by line 510, with voids 502 disposed between the metal layers.

[0069] Figure 5A further illustrates the measurement of acoustic transient signals at three different locations 540, 550, and 560 on the sample 500, where location 550 includes an embedded void 502. The acoustic transient signals are attributed to the presence of an optically opaque material, such as a metal layer 534. The measurement of the acoustic transient signals may occur sequentially, for example, during a lateral scan of the sample, or in parallel, for example, using a detector with a multi-pixel array.

[0070] At positions 540, 550, and 560, the irradiation from pulses in pump bundles 542, 552, and 562 is depicted as solid arrows of normal incidence. Pump bundles 542, 552, and 562 may be combinations of main and secondary pump bundles, as discussed above. For example, pump bundles 542, 552, and 562 may use infrared wavelengths, which penetrate the silicon substrate 522 without significant absorption, but when incident on the metal layer 524, they cause transient expansion of the metal layer 524 at the interface with the silicon substrate 522, thereby generating acoustic disturbances 544, 554, and 564, respectively, as depicted by the solid curves. Acoustic disturbances 544, 554, and 564 propagate through the metal layer 524 over time.

[0071] At positions 540 and 560, acoustic disturbances 544 and 564 may be reflected at interface 510 between metal layers 524 and 534, and after a delay d1, return to the surface of metal layer 524 as reflected acoustic disturbances 545, as illustrated by the dashed curve. The remaining portions of acoustic disturbances 544 and 564 continue to propagate through metal layer 534 until at least part of the remaining portions are reflected at interface 533 between metal layer 534 and silicon substrate 532, and after a delay d2, return to the surface of metal layer 524 as reflected acoustic disturbances 546 and 566, as illustrated by the dashed curve.

[0072] At location 550, including the gap 502, a portion of the acoustic disturbance 554 is reflected at the interface between the metal layer 524 and the gap 502, and returns to the surface of the metal layer 524 as a reflected acoustic disturbance 555 after a delay of d1, as illustrated by the dashed curve. The remaining portion of the acoustic disturbance 554 may continue to propagate through the metal layer 534 until it is at least partially reflected at the interface 533 between the metal layer 534 and the silicon substrate 532, and returns to the surface of the metal layer 524 as a reflected acoustic disturbance 556 after a delay of d2, as illustrated by the dashed curve.

[0073] The reflectivity at positions 540, 550, and 560 is measured by probe beams 543, 553, and 563, which are plotted as being incident on sample 500 at an illegal incident angle and reflected by the sample. However, it should be understood that probe beams 543, 553, and 563 may be collinear with pump beams 542, 552, and 562, or, if necessary, pump beams 542, 552, and 562 may be incident on sample 500 at an illegal incident angle, and probe beams 543, 553, and 563 may be incident on sample 500 at a normal incident angle and reflected by the sample. The reflectivity of samples (e.g., at the interface between metal layer 524 and silicon substrate 522) at locations 540, 550, and 560, as measured by probe beams 543, 553, and 563, is attributed to changes in reflectivity or surface deformation caused by reflected transient disturbances, such as combinations of reflected acoustic disturbances 545 / 546, 555 / 556, and 565 / 566, respectively.

[0074] As discussed above, pump bundles 542, 552, and 562 are combinations of main pump bundles and secondary pump bundles. The pulses in pump bundles 542, 552, and 562 have a fixed time delay between the main pump bundles and the secondary pump bundles. The pulses in probe bundles 543, 553, and 563 may have a fixed time delay relative to the pulses in pump bundles 542, 552, and 562. In this case, there is a first fixed time delay relative to the corresponding pulse in the main pump bundle and a second fixed time delay relative to the corresponding pulse in the secondary pump bundle. When pulses from probe beams 543, 553, and 563 are incident on the surface of metal layer 524, the reflected acoustic disturbances 545 / 546 at position 540, 555 / 556 at position 550, and 565 / 566 at position 560 are reflected back to the surface of metal layer 524, i.e., the interface between metal layer 524 and silicon substrate 522. Therefore, at position 540, the pulse in probe beam 543 is modulated due to the change in reflectivity caused by the reflected acoustic disturbance 545, which is a combination of the reflected acoustic disturbance 546 from interface 510 and the reflected acoustic disturbance 546 from interface 533. Similarly, at position 560, the pulse in probe beam 563 is modulated due to the change in reflectivity caused by the reflected acoustic disturbance 565, which is a combination of the reflected acoustic disturbance 566 from interface 510 and the reflected acoustic disturbance 566 from interface 533. However, at position 550, the probe beam 553 is modulated by the change in reflectivity caused by the combination of the reflected acoustic disturbance 555 from the gap 502 and the reflected acoustic disturbance 556 from the interface 533.

[0075] In some embodiments, measurements at each location 540, 550, and 560 may be performed using a variable pump-probe delay, such that the pulses in the probe bundles 543, 553, and 563 have a variable delay relative to the pulses in the pump bundles 542, 552, and 562, and vary a first time delay relative to the corresponding pulse in the main pump bundle and a second time delay relative to the corresponding pulse in the secondary pump bundle, but offset to a fixed delay between the main pump bundle and the secondary pulse bundle.

[0076] By using demodulation probes 543, 553 and 563, the instantaneous signal difference generated by the fixed time delay between the main pump pulse and the secondary pump pulse can be determined.

[0077] Figure 5B illustrates examples of acoustic transient signals 547, 557, and 567 received at positions 540, 550, and 560 in Figure 5A, respectively. In the graph of Figure 5B, the X-axis represents the delay time in picoseconds (ps) between the pump pulse and the probe pulse, and the Y-axis represents the differential reflectivity in arbitrary units (AU), i.e., the total reflectivity attributable to the pump's disturbance (at the sample) such as the effect of the pump modulation frequency being temporarily stored by the lock-in amplifier. Transient signals 547, 557, and 567 are responses to acoustic signals generated at positions 540, 550, and 560 by a single pump bundle, for example, a pump bundle that does not include a secondary pump bundle and has a pump-probe delay varying from 1000 ps to approximately 3000 ps in the sample. It should be understood that the original transient signals from samples including metal layers or other strongly absorbing materials may be a combination of acoustic signals and background signals (e.g., generated by thermal dissipation). The original transient signals are typically processed to remove any background signals (such as thermal dissipation, which is typically a DC component), thereby generating acoustic transient signals 547, 557, and 567.

[0078] As illustrated in Figure 5B, the presence or absence of a gap can be easily identified from acoustic transient signals 547, 557, and 567. It can be seen that acoustic transient signals 547, 557, and 567 are generally similar except for the presence of structures, such as gap 502. The presence or absence of a gap is determined based on the difference in signal profiles at a specific time delay corresponding to the time it takes for the acoustic echo to return. For example, as illustrated in Figure 5B, at approximately 2210 ps, ​​acoustic transient signals 547 and 567 experience negative peaks, while acoustic transient signal 557 experiences a positive peak. Based on the difference between acoustic transient signals 547, 557, and 567 at a specific time delay (e.g., 2210 ps) corresponding to the depth of the expected underlying structure, the presence of gap 502 at location 550 can be detected.

[0079] By rationally selecting the fixed time delay between the pulses in the main pump bundle and the pulses in the secondary pump bundle, the sensitivity of the resulting instantaneous signal difference can be optimized. For example, the fixed time delay D between the pulses in the main pump bundle and the pulses in the secondary pump bundle is illustrated as the distance between the following two: a first delay between the main pump pulse and the probe pulse, such as delay d1 illustrated in FIG. 5B; and a second delay between the secondary pump pulse and the probe pulse, such as delay d2 illustrated in FIG. 5B. Due to the presence of gap 502, the transient signal difference generated by acoustic transient signal 557 is significantly different from the transient signal difference generated by acoustic transient signals 547 and 567, and the presence of gap 502 can be determined based on the instantaneous signal difference.

[0080] If the pump-detection delay varies during the measurement, multiple transient signal differences generated from a fixed time delay D will be produced under different pump-detection delays, for example, the fixed time delay D is scanned across the X-axis in Figure 5B. Figure 5C illustrates a graph of transient signal differences under different pump-detection delays with the aid of examples, where the X-axis represents different pump-detection delay times in picoseconds (ps), and the Y-axis represents the differential reflectivity of the transient signal difference in arbitrary units (AU). Transient signal differences 548, 558, and 568 are generated at positions 540, 550, and 560 respectively in response to the fixed time delay between the main pump bundle and the secondary pump bundle. As can be seen, the resulting transient signal difference 558 generated at position 550 is different from the resulting transient signal differences 548 and 568 generated at positions 540 and 560, thereby enabling the detection and characterization of samples.

[0081] If the pump-probe delay is fixed during measurement, a single instantaneous signal difference generated from a fixed time delay D under a fixed pump-probe delay will be generated, for example, at a position of the fixed time delay D on the X-axis plotted in Figure 5B. Figure 5D illustrates a graph of the instantaneous signal difference under a fixed pump-probe delay by way of example, where the X-axis represents the position on the sample and the Y-axis represents the differential reflectivity of the instantaneous signal difference in arbitrary units (AU). For example, the instantaneous signal difference 570 is generated by scanning the measurement points above positions 540, 550, and 560 while maintaining a fixed pump-probe time delay and a fixed time delay between the main pump bundle and the secondary pump bundle. As can be seen, the gap 502, for example, generates a distinct peak 572 in the resulting instantaneous signal difference 570, thereby enabling the detection and characterization of the sample.

[0082] In some embodiments, the fixed time delay D between the pulses in the main pump bundle and the pulses in the secondary pump bundle can be changed, for example, by changing one or both of the probe bundle paths to adjust one or both of the first delay d1 and the second delay d2, thereby increasing the sensitivity of the instantaneous signal difference. Therefore, embedded structures, such as layer interfaces and voids 502, can be detected using the fixed time delay between the main and secondary pump bundles using instantaneous signal differences. By way of example, based on the instantaneous signal difference from the reflected acoustic disturbances from interfaces 510 and 533 corresponding to the thickness of the metal layer 534 at location 540, changes in the instantaneous signal difference can be used to indicate changes in the thickness of the metal layer 534. Furthermore, an image of the sample 500, including deviations from the nominal thickness of the metal layer 534 or the presence or absence of embedded structures (such as voids), can be generated, for example, by laterally scanning the measurement location above the sample based on the instantaneous signal differences of acoustic transient signals from multiple locations.

[0083] Figures 6A and 6B illustrate the use of time-resolved reflectance measurements in response to non-acoustic transient signals to detect embedded structures in samples that do not have a metal layer (or other strongly absorbing materials capable of generating acoustic signals). For example, reflectance measurements in response to non-acoustic transient signals can be attributed to changes in reflectance, although it is possible that the non-acoustic transient signals may also, or alternatively, cause some changes in surface deformation that may be detected in the reflectance measurements. Figure 6A illustrates a sample 600 similar to the sample 600 shown in Figure 6A, except that sample 600 includes silicon oxide (SiO2) instead of a metal layer. As illustrated, sample 600 is formed from two bonded wafers 620 and 630, with an embedded structure in the form of a void 602 between the two wafers. Wafer 620 includes, for example, a silicon substrate 622 (which may have a thickness of 750 µm, serving as the top layer of sample 600) and a SiO2 layer 624 (which may have a thickness of 50 nm). Wafer 630 similarly includes a SiO2 layer 634, which may have a thickness of 50 nm, on the silicon substrate 632, which may have a thickness of 750 µm and serve as the bottom layer of sample 600. SiO2 layers 624 and 634 are bonded together as shown in line 610, with a void 602 disposed between SiO2 layers 624 and 634.

[0084] Figure 6A further illustrates the measurement of non-acoustic transient signals at three different locations 640, 650, and 660 on the sample 600, where location 650 includes an embedded void 602. The sample 600 comprises only an optically transparent layer, i.e., there is no strongly absorbing material that generates an acoustic signal in response to pump irradiation, and therefore non-acoustic transient signals are generated and measured at locations 640, 650, and 660. The measurement of non-acoustic transient signals at locations 640, 650, and 660 may occur sequentially, for example, during a lateral scan of the sample, or in parallel, for example, using a detector with a multi-pixel array.

[0085] At positions 640, 650, and 660, the irradiation from pulses in pump bundles 642, 652, and 662 is depicted as solid arrows with normal incidence. Pump bundles 642, 652, and 662 may be combinations of main and secondary pump bundles, as discussed above. Pump bundles 642, 652, and 662 may use infrared wavelengths that penetrate the silicon substrate 622 without significant absorption. SiO2 layers 624 and 634 are not strongly absorbing materials and do not undergo transient expansion in response to pulses in pump bundles 642, 652, and 662, and therefore no (or very small) acoustic signals are generated in SiO2 layers 624 and 634. However, pump bundles 642, 652, and 662 generate non-acoustic transient perturbations, for example, due to absorption of pulses from the pump bundles via multiphoton ionization or due to distortions in the properties of the sample material at the interface. Non-acoustic transient disturbances 644, 654, and 664 are generated in response to the main pump bundles at positions 640, 650, and 660, respectively, as indicated by the solid arrows radiating outwards. Non-acoustic transient disturbances 645, 655, and 665 are generated in response to the secondary pump bundles at positions 640, 650, and 660, respectively, as indicated by the dashed arrows radiating outwards. Non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 are generated in response to the main pump pulses and secondary pump pulses in pump bundles 642, 652, and 662, respectively, and are caused by non-acoustic physical phenomena, such as heat dissipation, electro-electric recombination, plasma diffusion, symmetric damage at the top and bottom oxide surfaces, and the etalon effect within the void 602, or one or more of these phenomena. Unlike the acoustic transient disturbances illustrated in Figure 5A, the non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 are not reflected back by the structure, but instead are generated in response to absorption, for example, from pulses from the pump bundle, and decay over time. The presence of structures such as void 602 affects the decay rate of the non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665.

[0086] The reflectivity at positions 640, 650, and 660 is measured by pulses in probe beams 643, 653, and 663, which are plotted as being incident on sample 600 at an illegal angle of incidence and reflected by the sample. However, it should be understood that probe beams 643, 653, and 663 may be collinear with pump beams 642, 652, and 662, or, if necessary, pump beams 642, 652, and 662 may be incident on sample 600 at an illegal angle of incidence, and probe beams 643, 653, and 663 may be incident on sample 600 at a normal angle of incidence and reflected by the sample. The reflectance of samples measured by probe beams 643, 653, and 663 at locations 640, 650, and 660, for example at the interface between SiO2 layer 624 and silicon substrate 622, is modulated by changes in the reflectance of SiO2 layer 624 caused by non-acoustic transient disturbances 644, 654, and 664. In some cases, the reflectance of the sample may also be due to changes in surface deformation or alternatively, attributable to changes in surface deformation. The non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 decay over time, and therefore the measured reflectance in response to the non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 will also change over time. Generally, for the measurement of acoustic transient signals, as illustrated in Figure 5A, non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 are considered background signals and removed from the original transient signal measurement. However, in this embodiment, there is little or no acoustic information, and therefore the original transient signals measured and analyzed for positions 640, 650, and 660 are non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665.

[0087] As discussed above, pump bundles 642, 652, and 662 are combinations of main pump bundles and secondary pump bundles. The pulses in pump bundles 642, 652, and 662 have a fixed time delay between the main pump bundles and the secondary pump bundles. The pulses in probe bundles 643, 653, and 663 may have a fixed time delay relative to the pulses in pump bundles 642, 652, and 662. In this case, there is a first fixed time delay relative to the corresponding pulse in the main pump bundle and a second fixed time delay relative to the corresponding pulse in the secondary pump bundle. When the pulses from probe bundles 643, 653, and 663 are incident on the surface of metal layer 624, the non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 generated in response to the main pump pulses and secondary pump pulses will return to the surface of metal layer 624, i.e., the interface between metal layer 624 and silicon substrate 622. Non-acoustic transient disturbances 644 / 645, 654 / 655, and 664 / 665 decay over time, and therefore, when measured by pulses from probe beams 643, 653, and 663, the non-acoustic transient disturbances 644, 654, and 664 generated in response to the main pump pulse decay faster than the non-acoustic transient disturbances 645, 655, and 665 generated by the secondary pump pulse. Therefore, at position 640, the pulse in probe beam 643 will be modulated due to the reflectivity change caused by the combination of non-acoustic transient disturbances 644 and 645. Similarly, at position 660, the pulse in probe beam 663 will be modulated due to the reflectivity change caused by non-acoustic transient disturbances 664 and 665. At position 650, the probe beam 653 will be modulated due to the change in reflectivity caused by non-acoustic transient disturbances 654 and 655, which are attributed to the presence of gap 602, unlike the non-acoustic transient disturbances 644 / 645 and 664 / 665.

[0088] In some embodiments, measurements at positions 640, 650, and 660 may be performed using a variable pump-probe delay, such that the pulses in the probe bundles 643, 653, and 663 have a variable delay relative to the pulses in the pump bundles 642, 652, and 662, and vary with a first delay relative to the corresponding pulse in the main pump bundle and a second delay relative to the corresponding pulse in the secondary pump bundle, but offset to a fixed delay between the main pump bundle and the secondary pulse bundle.

[0089] By using demodulation probes 643, 653 and 663, the instantaneous signal difference generated by the fixed time delay between the main pump pulse and the secondary pump pulse can be determined.

[0090] Figure 6B illustrates example graphs of nonacoustic transient signals 646, 656, and 666 detected by probe beams 643, 653, and 663 at positions 640, 650, and 660, respectively, in Figure 6A. In the graphs of Figure 6B, the X-axis represents the delay time between the pump beam and the probe beam in picoseconds (ps), and the Y-axis represents the differential reflectivity in arbitrary units (AU). The measured nonacoustic transient signals 646, 656, and 666 are responses to nonacoustic disturbances generated by a single pump beam at positions 640, 650, and 660, which, for example, does not include a secondary pump beam and has a pump-probe delay time varying from 1000 ps to approximately 3000 ps. However, the presence of gap 602 at position 650 results in a difference between the non-acoustic disturbance 654 and the non-acoustic transient disturbances 644 and 664 at positions 640 and 660, and therefore, the resulting measured non-acoustic transient signals in response to these disturbances will also differ. As shown in Figure 6B, the non-acoustic transient signals 646 and 666 from positions 640 and 660, where gap 602 is absent, are substantially similar in properties or characteristics (e.g., the shape or slope of the transient signal). Conversely, the non-acoustic transient signal 656 from position 650, where gap 602 is present, is not similar to the non-acoustic transient signals 646 and 666 at positions 640 and 660 on sample 600 in terms of such properties or characteristics as shape or slope. Therefore, transient signal analysis can be performed to distinguish the various properties or characteristics of transient signals from different positions. The properties or characteristics of transient signals from different positions can be used to determine whether the measurement was performed at a position with a gap or without a gap.

[0091] If the measurement system uses a variable delay time between the pump pulse and the probe pulse, curves such as those plotted by 646, 656, and 666 within the delay time range can be generated, and said curves can be analyzed, for example, based on polynomial or other curve fitting (e.g., exponential) or other types of analysis (such as principal component analysis, PCA) of transient signals at different locations, or comparison of attributes or characteristics (e.g., the analysis rate of shape).

[0092] Although the presence or absence of a void is determined based on the difference in transient signals at a specific time delay (e.g., a single point on the time delay axis), if the transient signal is generated by acoustic disturbances, as illustrated in Figure 6B, when the transient signal is generated by one or more of non-acoustic disturbances such as heat dissipation, electron-hole recombination, plasma diffusion, symmetric breaking of the top and bottom oxide surfaces, etalon effect, etc., the presence or absence of a void is determined based on the difference in transient signals over the time delay range (e.g., multiple points on the time delay axis). For example, the number of time delays used to analyze the transient signal may depend on the sample being tested, but should be sufficient to identify one or more features of the non-acoustic transient signal, such as shape, slope, rate of change, etc., which characterize desired properties or characteristics, such as the attenuation rate of the non-acoustic transient disturbance, from which the presence of an embedded structure (such as a void) can be determined.

[0093] As discussed above, pump bundles 642, 652, and 662 are combinations of main pump bundles and secondary pump bundles. By reasonably selecting the fixed time delay between the pulses in the main pump bundle and the pulses in the secondary pump bundle, the sensitivity of the resulting instantaneous signal difference can be optimized. For example, the fixed time delay D between the pulses in the main pump bundle and the pulses in the secondary pump bundle is plotted as the distance between the following two: a first delay between the main pump pulse and the probe pulse, such as delay d1 shown in FIG. 6B; and a second delay between the secondary pump pulse and the probe pulse, such as delay d2 shown in FIG. 6B. The instantaneous signal difference generated by the non-acoustic transient signal 656 due to the presence of gap 602 will be significantly different from the instantaneous signal difference generated by the acoustic transient signals 646 and 666, and therefore the presence of gap 602 can be determined based on the instantaneous signal difference. As discussed above, the pump-probe delay can vary during measurement, and multiple instantaneous signal differences generated from a fixed time delay D will be produced under different pump-probe delays, for example, the fixed time delay D is scanned across the X-axis in Figure 6B. In another embodiment, the pump-probe delay can be fixed during measurement, for example, at a fixed position of the fixed time delay D on the X-axis shown in Figure 6B and at the measurement position scanned above the sample.

[0094] In some embodiments, the fixed time delay D between the pulses in the main pump bundle and the pulses in the secondary pump bundle can be changed, for example, by changing one or both of the probe bundle paths to adjust one or both of the first delay d1 and the second delay d2, thereby increasing the sensitivity of the transient signal difference. As can be seen in FIG6B, the resulting difference in the signals generated in response to the first time delay d1 and the second time delay d2 will be different for the non-acoustic transient signals 646 and 666 from positions 640 and 660 where there is no gap, and for the non-acoustic transient signal 656 from position 650 where there is a gap 602. For example, FIG6B illustrates that the signal difference in the non-acoustic transient signal 656 from position 650 at the first time delay d1 and the second time delay d2 is approximately 0, while the signal difference in the non-acoustic transient signals 646 and 666 from positions 640 and 660 at the first time delay d1 and the second time delay d2 will not be zero. Therefore, the gap 602 can be detected using instantaneous signal differential with a fixed time delay between the main pump bundle and the secondary pump bundle. Furthermore, an image of the sample 600, including the presence of the gap 602, can be generated, for example, by laterally scanning the measurement position above the sample based on instantaneous signal differential from non-acoustic transient signals from multiple locations.

[0095] Figure 7 is a flowchart 700 illustrating a method for non-destructively characterizing samples using instantaneous signal differences, which are the differences between signals generated from different delay times of multiple pump beams acquired with a single probe beam, as discussed herein. For example, the procedure can be performed using the time-resolved reflectivity metering device 100 or 200 shown in Figure 1 or Figure 2, respectively.

[0096] As illustrated, at block 710, the procedure includes: irradiating a sample using a pump bundle comprising a series of main pump pulses and a series of secondary pump pulses, wherein the series of main pump pulses and the series of secondary pump pulses are intensity modulated and out of phase, the secondary pump pulses having a fixed delay relative to the main pump pulses, and each main pump pulse and each secondary pump pulse causing transient disturbances in the material of the sample, such as those illustrated by pump arms 120 and 210 in metering devices 100 and 200 in Figures 1 and 2 respectively, and discussed with reference to Figures 4, 5A and 6A.

[0097] At block 720, a probe beam is used to irradiate the sample. The probe beam comprises a series of probe pulses, each probe pulse having a first delay relative to a previous main pump pulse and a second delay relative to a previous secondary pump pulse, the second delay being different from the first delay. The series of probe pulses is reflected from the sample as a series of reflected probe pulses, each of which is modulated by the transient disturbance in the material of the sample caused by both the previous main pump pulse and the previous secondary pump pulse, for example as illustrated by probe arms 130 and 220 in the metering devices 100 and 200 in Figures 1 and 2, respectively, and discussed with reference to Figures 4, 5A, and 6A. In some embodiments, the first and second delays are fixed delays. In some embodiments, the probe pulses have variable delays relative to the main pump pulse and the secondary pump pulse.

[0098] At block 730, the instantaneous signal difference in the series of reflected detection pulses generated in response to the series of main pump pulses and the series of secondary pump pulses is detected, for example as illustrated by detectors 160 and 228 in metering devices 100 and 200 in Figures 1 and 2 respectively, and discussed with reference to Figures 4, 5B and 6B.

[0099] At block 740, the characteristics of the sample are determined based on the instantaneous signal difference, for example, as illustrated by the processing systems 170 and 230 in the metering devices 100 and 200 in Figures 1 and 2, respectively, and discussed with reference to Figures 5A and 6A. In some embodiments, the characteristics of the sample are the presence or absence of an embedded structure in the sample, for example, as discussed with reference to Figures 5A and 5B. In some embodiments, the characteristics of the sample are the presence or absence of voids in the material of the sample that is transparent to the pump bundle, for example, as discussed with reference to Figures 6A and 6B.

[0100] In some embodiments, for example, irradiating a sample with a pump beam may include generating a series of pump pulses of light and modulating the intensity of such series of pump pulses, as illustrated, for example, by light sources 102 and 211 and optical modulators 122 and 212 in metrology devices 100 and 200 in Figures 1 and 2, respectively, and discussed in Figures 3 and 4. Each pump pulse in the series of pump pulses may be split into main pump pulses and secondary pump pulses to generate the series of main pump pulses and secondary pump pulses with modulated intensity and opposite phase, for example, by beam splitters 124 and 213 in metrology devices 100 and 200 in Figures 1 and 2, respectively, and discussed in Figures 3 and 4. Furthermore, the difference between the main pump path length of the series of main pump pulses and the secondary pump path length of the series of secondary pump pulses is generated to produce a fixed delay between the main pump pulses and the secondary pump pulses, for example, as illustrated by the main beam path 125 and secondary beam path 126 and the main beam path 214 and secondary beam path 215 in the metering devices 100 and 200 in Figures 1 and 2, respectively, and discussed in Figures 3 and 4.

[0101] In some embodiments, the procedure may further include using a light source to generate a series of light pulses and using a beam splitter to split the series of light pulses into a pump beam and a probe beam, for example, as illustrated by the light source 102 and beam splitter 106 in the metering device 100 in FIG1.

[0102] In some embodiments, the procedure may further include using a first light source to generate a pump beam and using a second light source to generate a probe beam, for example, as illustrated by light sources 211 and 221 in the metering device 200 in FIG2.

[0103] The sample can be irradiated at the measurement location using a probe beam, and the procedure can further include maintaining the first delay and the second delay as fixed delays while laterally scanning the measurement location above the sample, for example, as discussed with respect to the measuring devices 100 and 200 in Figures 1 and 2 and Figures 5A and 6A.

[0104] The procedure may further include: adjusting the fixed delay between the main pump pulse and the secondary pump pulse while using the pump beam to irradiate the sample and using the probe beam to irradiate the sample; and using the different fixed delays between the main pump pulse and the secondary pump pulse to detect the instantaneous signal difference to determine the characteristics of the sample, as discussed with respect to the metering devices 100 and 200 in Figures 1 and 2, and Figures 5A and 6A. For example, adjusting the fixed delay between the main pump pulse and the secondary pump pulse may include: varying the difference between the main pump path length of the main pump pulse series and the secondary pump path length of the secondary pump pulse series, for example, as discussed with respect to the main beam path 125 and the secondary beam path 126 and the main beam path 214 and the secondary beam path 215 in the metering devices 100 and 200 in Figures 1 and 2, respectively, and as discussed in Figures 3 and 4.

[0105] In some embodiments, the instantaneous signal difference can be detected by detecting the series of reflected probe pulses from the sample and demodulating the series of reflected probe pulses to determine the instantaneous signal difference generated in response to the series of main pump pulses and the series of secondary pump pulses, for example, as illustrated by detectors 160 and 228 and demodulators 162 and 229 in metering devices 100 and 200 in Figures 1 and 2, respectively, and as depicted in Figures 4, 5B, and 6B. Demodulating the series of reflected probe pulses to determine the instantaneous signal difference can be based on the frequency of intensity modulation of the series of main pump pulses and the series of secondary pump pulses, for example, as discussed with respect to detectors 160 and 228 and demodulators 162 and 229 in metering devices 100 and 200 in Figures 1 and 2, respectively.

[0106] In some embodiments, the transient disturbances in the material caused by each main pump pulse and each secondary pump pulse are acoustic transient disturbances, for example, as discussed with respect to Figures 5A and 5B. In some embodiments, the transient disturbances in the material caused by each main pump pulse and each secondary pump pulse are non-acoustic transient disturbances, for example, as discussed with respect to Figures 6A and 6B. For example, non-acoustic transient disturbances may be caused by one or more of the following: heat dissipation, electron-hole recombination (e.g., possibly generated in Si), plasma diffusion, symmetric fragmentation of the top and bottom oxide surfaces, and etalon effects.

[0107] The foregoing description is intended to be illustrative and not restrictive. For example, the above examples (or one or more of them) may be used in combination with each other. Other embodiments may be used, for instance, by those skilled in the art upon review of the foregoing description. Furthermore, in the above embodiments, various features may be combined together to simplify this disclosure. The subject matter of this invention may be less than all the features of the specifically disclosed embodiments. Therefore, the following embodiments are hereby incorporated as examples or implementations, wherein each embodiment is independently a separate implementation, and such implementations are contemplated to be combined with each other in various combinations or arrangements. Therefore, the spirit and scope of the appended claims should not be limited to the foregoing description. [Simplified Explanation of the Diagram]

[0010] [Figure 1] illustrates a schematic diagram of an example time-resolved metrology device configured to perform time-resolved reflectivity measurements using instantaneous signal differential, as discussed herein. [Figure 2] illustrates a block diagram of another example time-resolved metrology device configured to perform time-resolved reflectivity measurements using instantaneous signal differential, as discussed herein. [Figure 3] is a graph illustrating the intensity modulation of the main pump bundle and the secondary pump bundle. [Figure 4] is a graphical representation of the main pump bundle, the secondary pump bundle, the probe bundle, and the detected reflective probe bundle. [Figures 5A] through [Figure 5D] illustrate the use of instantaneous signal differential detection of embedded structures in samples with one or more metal layers, the instantaneous signal differential being derived from time-resolved reflectivity measurements generated in response to acoustic transient signals. [Figure 6A] and [Figure 6B] illustrate the use of time-resolved reflectivity measurements in response to non-acoustic transient signals to detect embedded structures in samples that do not have a metallic layer (or other strongly absorbing materials capable of generating acoustic signals). [Figure 7] is a flowchart illustrating a procedure for non-destructively characterizing a sample using transient signals, which employ instantaneous signal differential.

Claims

1. A method for characterizing a sample using a transient signal difference rather than destructively, the transient signal difference being the difference between signals generated from different delay times of multiple pump bundles acquired with a single probe beam, the method comprising: irradiating the sample with a pump bundle comprising a series of main pump pulses and a series of secondary pump pulses, wherein the series of main pump pulses and the series of secondary pump pulses are intensity modulated and out of phase, the secondary pump pulses having a fixed delay relative to the main pump pulses, and each main pump pulse and each secondary pump pulse causing a transient perturbation in the material of the sample; irradiating the sample with a probe beam comprising a series of probe pulses, wherein each probe pulse has a first delay relative to a previous main pump pulse and a second delay relative to a previous secondary pump pulse, the second delay being different from the first delay, and the series of probe pulses being reflected from the sample as a series of reflected probe pulses, wherein each reflected probe pulse is modulated by the transient perturbation in the material of the sample caused by both the previous main pump pulse and the previous secondary pump pulse; The instantaneous signal difference in the series of reflected detection pulses generated in response to the series of main pump pulses and the series of secondary pump pulses is detected; and the characteristics of the sample are determined based on the instantaneous signal difference.

2. The method of request item 1, wherein the first delay and the second delay are fixed delays.

3. The method of claim 1, wherein the probe pulse has a variable delay relative to the main pump pulse and the secondary pump pulse.

4. The method of claim 1, wherein irradiating the sample with the pump bundle comprising the series of main pump pulses and the series of secondary pump pulses comprises: generating a series of light pump pulses; modulating the intensity of the series of light pump pulses; splitting each pump pulse in the series of pump pulses into main pump pulses and secondary pump pulses to generate the series of main pump pulses and the series of secondary pump pulses that are intensity modulated and out of phase; and generating the difference between the main pump path length of the series of main pump pulses and the secondary pump path length of the series of secondary pump pulses to generate the fixed delay between the main pump pulses and the secondary pump pulses.

5. The method of claim 1, further comprising: generating a series of light pulses with a light source; and splitting the series of light pulses into the pump beam and the probe beam with a beam splitter.

6. The method of claim 1, further comprising: generating the pump beam with a first light source; and generating the probe beam with a second light source.

7. The method of claim 1, wherein the sample is irradiated with the probe beam at the measurement location, the method further comprising: while laterally scanning the measurement location above the sample, maintaining the first delay and the second delay as fixed delays.

8. The method of claim 1, further comprising: adjusting the fixed delay between the main pump pulse and the secondary pump pulse while irradiating the sample with the pump beam and the probe beam; and detecting the instantaneous signal difference with the different fixed delays between the main pump pulse and the secondary pump pulse to determine the characteristic of the sample.

9. The method of claim 8, wherein adjusting the fixed delay between the main pump pulse and the secondary pump pulse comprises: varying the difference between the main pump path length of the series of main pump pulses and the secondary pump path length of the series of secondary pump pulses.

10. The method of claim 1, wherein detecting the instantaneous signal difference in the series of reflected detection pulses generated in response to the series of main pump pulses and the series of secondary pump pulses comprises: detecting the series of reflected detection pulses from the sample; and demodulating the series of reflected detection pulses to determine the instantaneous signal difference generated in response to the series of main pump pulses and the series of secondary pump pulses, wherein demodulating the series of reflected detection pulses to determine the instantaneous signal difference is based on the intensity modulation frequency of the series of main pump pulses and the series of secondary pump pulses.

11. The method of claim 1, wherein the transient disturbances in the material caused by each main pump pulse and each secondary pump pulse are acoustic transient disturbances or non-acoustic transient disturbances generated by one or more of the following: heat dissipation, electron-hole recombination, plasma diffusion, symmetric fragmentation of the top and bottom oxide surfaces, and etalon effect.

12. The method of claim 1, wherein the characteristic of the sample is the presence or absence of voids in the embedded structure in the sample or in the material of the sample that is transparent to the pump bundle.

13. A metrology device for characterizing a sample using a transient signal difference rather than destructively, the transient signal difference being the difference between signals generated from different delay times of multiple pump bundles acquired with a single probe beam, the metrology device comprising: a pump arm that irradiates the sample with a pump bundle comprising a series of main pump pulses and a series of secondary pump pulses, wherein the series of main pump pulses and the series of secondary pump pulses are intensity modulated and out of phase, the secondary pump pulses having a fixed delay relative to the main pump pulses, and each main pump pulse and each secondary pump pulse causing a transient disturbance in the material of the sample; a probe arm that irradiates the sample with a probe beam comprising a series of probe pulses, wherein each probe pulse has a first delay relative to a previous main pump pulse and a second delay relative to a previous secondary pump pulse, the second delay being different from the first delay, and the series of probe pulses being reflected from the sample as a series of reflected probe pulses, wherein each reflected probe pulse is modulated by the transient disturbance in the material of the sample caused by both the previous main pump pulse and the previous secondary pump pulse; A detector that detects a series of reflected probe pulses to determine the instantaneous signal difference in the series of reflected probe pulses generated in response to the series of main pump pulses and the series of secondary pump pulses; and at least one processor coupled to the detector and configured to determine the characteristics of the sample based on the instantaneous signal difference.

14. The metering device as claimed in claim 13, wherein the first delay and the second delay are fixed delays.

15. The metering device of claim 13, wherein the detection pulse has a variable delay relative to the main pump pulse and the secondary pump pulse.

16. The metering device of claim 13, wherein the pump arm comprises: an optical modulator that modulates the intensity of a series of pump pulses; a beam splitter that splits each pump pulse in the series of pump pulses into a main pump pulse guided along a main pump path and a secondary pump pulse guided along a secondary pump path to generate the series of main pump pulses and the series of secondary pump pulses that are intensity modulated and out of phase, wherein the main pump path length and the secondary pump path length are different to generate the fixed delay between the main pump pulses and the secondary pump pulses; and a beam combiner that combines the series of main pump pulses and the series of secondary pump pulses into the pump bundle.

17. The metering device of claim 13, further comprising: a light source that generates a series of light pulses; and a beam splitter that splits the series of light pulses into the pump beam and the probe beam.

18. The metering device of claim 13, further comprising: a first light source that generates the pump beam; and a second light source that generates the probe beam.

19. The measuring device of claim 13, further comprising an actuator to generate relative lateral movement between the sample and the pump arm and the probe arm to laterally scan a measurement position above the sample, wherein the first delay and the second delay are maintained as fixed delays while the measurement position is laterally scanned above the sample.

20. The metering device of claim 13, further comprising a variable delay to adjust the fixed delay between the main pump pulse and the secondary pump pulse while irradiating the sample with the pump beam, and the detector detecting an instantaneous signal difference with a different fixed delay between the main pump pulse and the secondary pump pulse, and irradiating the sample with the probe beam to determine the characteristic of the sample.

21. The metering device of claim 20, wherein the variable delay causes the difference between the main pump path length of the series of main pump pulses and the secondary pump path length of the series of secondary pump pulses to vary.

22. The metering device of claim 13, further comprising a demodulator coupled to the detector to demodulate the series of reflected detection pulses to determine the instantaneous signal difference generated in response to the series of main pump pulses and the series of secondary pump pulses, and the demodulator demodulating the series of reflected detection pulses to determine the instantaneous signal difference by modulating the frequency based on the intensity of the series of main pump pulses and the series of secondary pump pulses.

23. The metering device of claim 13, wherein the transient disturbances in the material caused by each main pump pulse and each secondary pump pulse are acoustic transient disturbances or non-acoustic transient disturbances generated by one or more of the following: heat dissipation, electron-hole recombination, plasma diffusion, symmetric fragmentation of the top and bottom oxide surfaces, and etalon effect.

24. The metering device of claim 13, wherein the characteristic of the sample is the presence or absence of voids in the embedded structure of the sample or in the material of the sample that is transparent to the pump bundle.

Images