JP2005338063A - Apparatus for measuring physical characteristics of sample - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an apparatus for measuring physical characteristics of samples by using short optical pulses for generating and detecting short-wavelength sound waves. <P>SOLUTION: For an initial stage of measurement, short initial pump pulses 4 of optical radiation are condensed by a lens 5 to a local region 13 on a sample 6, having a free surface at perpendicular or almost perpendicular angle of incidence. Light is absorbed, at a region close to the surface to generate an initial local sound wave source. Especially, in the example of an elastically isotropic solid, short-wavelength pulses of longitudinal sound waves are projected to the depth direction of the sample 6. It is advantageous for facilitating data analysis, to select a horizontal space profile of an initial pump beam 10 in the sample 6, in such a way as to be axially symmetric as in the case of a Gaussian spot. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an apparatus for measuring a physical property of a sample, and more particularly to an apparatus for measuring a physical property of a substance by an acoustic method.

  Nondestructive measurement of physical properties of materials using acoustic waves has been proposed in various fields because it is possible to probe subsurface properties. Of particular interest among applications relating to the probe of complex materials, including many acoustic scatterers, such as voids, defects, cracks or other inhomogeneous regions, is the method of acoustic time reversal. For example, there are various technologies in this field, as clearly shown in Patent Documents 1 to 4 below. These methods are particularly superior, regardless of the crystal state of the sample, such as isotropic or anisotropy, and even if acoustic scatterers are distributed, repeated time reversal methods or arrays of acoustic sources. It is a point that identification and analysis of the individual acoustic scatterers in the sample can be performed by the matrix method using the.

  The basic principle is to generate a short acoustic pulse in a local region of the sample and constitute an initial local acoustic wave source. It is advantageous that the acoustic disturbance in the sample includes a wider range of acoustic wave vectors. The acoustic response is recorded as a function of time, typically by an ultrasonic transducer that can also be used for sound generation, in the same or another region of the sample, referred to as the initial probe region of the sample.

  The time required for this recording is often so long that the detected acoustic signal, that is, the coder wave completely disappears. This signal, referred to as the initial signal group, is stored and electronically time-reversed. The same piezoelectric transducer is then played back at the same initial probe area of the sample. If the acoustic propagation in the sample is not very strongly influenced by the dissipation process or controlled by drift or creep, the acoustic propagation is time-reversed according to an elastic wave equation that includes only an even order of time t in a reasonable approximation. Part or all of the playback acoustic disturbance is again focused on the region of the initial local acoustic source, following the same path in the sample, but in the reverse direction. The focused sound wave can then be detected, for example, by the same ultrasonic transducer used for the initial impulse local sound source. This is true even if there is a scattering medium between the initial local acoustic wave source and the initial probe region of the sample.

  The playback acoustic disturbance may not be selected as a complete copy of the electronically time-reversed signal. For example, it may be digitized so that only one bit of information remains. Alternatively, for example, time gate processing may be performed so that only a part of the time-reversed signal is played back. This time-gating method has been found to be effective in identifying acoustic scatterers embedded in the material by repeating the time reversal process described above. If the signal representing the playback acoustic disturbance detected in the initial local acoustic source region is electronically time-reversed and played back on the same part of the sample, the entire process consisting of the initial and playback stages of the measurement Can be executed as many times as desired. By such an iterative method, the acoustic disturbance changes and converges, ignoring the weak scatterers, and generating a large amplitude at the location of one or more acoustic scatterers that predominate in the sample. This method has proven useful for identifying scatterers in materials.

  A generalized method of the above time gate processing method has also been proposed and is known as a time reversal operator decomposition method. This method uses an array of local acoustic sources, such as ultrasonic transducers, to identify individual scatterers. If one transducer records an acoustic signal generated by another transducer and executes this with all combinations of transducers, a complete acoustic signal of the sample to be measured is produced as a matrix. This matrix operation makes it possible to identify the main acoustic scatterers in the medium in an iterative process involving sound generation and time reversal playback.

  Another important area of application of this method relates to the realization of high acoustic intensity using many playback acoustic disturbances that are spatially distributed but all converge to the same point. In order to produce a particularly high acoustic intensity, it can be carried out, for example, in an acoustic waveguide.

  A field of application related to acoustic focusing is the use of disordered acoustic cavities that include an ultrasonic transducer acting as an initial local acoustic source, as described, for example, in Non-Patent Document 1 below. With the knowledge of the time reversal signal required for playback in this ultrasonic transducer, it is possible to focus on a preselected point inside or outside the disordered acoustic cavity. This required time reversal signal can be measured in advance by placing a local sound source at a preselected point and detecting the acoustic signal with an ultrasonic transducer in a disordered acoustic cavity. Necessary signals can be obtained by time-reversing the acoustic signals.

  It has also been shown that in the presence of significant acoustic non-linearity or dissipation processes, the acoustic time reversal method can be used to obtain useful physical information about the sample.

  These promising techniques based on acoustic time reversal rely primarily on conventional ultrasonic piezoelectric transducers. Such transducers are generally limited to relatively low MHz frequency bands such as the corresponding wavelength range of 0.1-1 mm. Although it is possible to use piezoelectric transducers at frequencies up to higher GHz, it is very inefficient, very difficult to implement and expensive. Furthermore, such conventional acoustic time reversal methods require contact with a sample or a binding medium such as water. Thus, these methods are not applicable, for example, to the evaluation of microstructures or nanostructures that require short acoustic wavelengths up to the sub-nanometer (nm) range, or high frequencies up to 1 THz or higher.

  Some progress has been made to non-contact sound generation combined with acoustic time reversal method. For example, Non-Patent Document 2 below reports the use of a laser pulse that generates an initial local acoustic wave source and an optical interferometer for time-reversed playback acoustic disturbance detection based on continuous wave radiation. However, this conventional laser acoustic method cannot be used for very high acoustic frequencies exceeding 1 GHz.

  An acoustic pulse echo technique for measuring the thickness of a micron or submicron thin film has been proposed by the following Patent Document 5. In this method, a short-wavelength longitudinal acoustic pulse is excited in a sample by an ultra-short duration pump light pulse, and a strain-induced change in the sample optical constant using a delayed ultra-short duration probe light pulse. In particular by measuring the intensity change of the probe beam reflected from the sample. However, since only the acoustic propagation time is measured in such an acoustic pulse echo apparatus, this apparatus cannot be used for determining the film thickness unless the sound velocity of the film depending on the elastic constant and density is known. This device is therefore disadvantageous in that it requires sound speed calibration. Furthermore, this device is useless when there is a significant number of acoustic scatterers or inhomogeneous regions in the sample and when there are too many layers, it is difficult to analyze the complex and relatively long echo waveforms detected.

Similar devices that can be used to circumvent the limitation that the film thickness and sound velocity cannot be determined simultaneously, and generate and detect short wavelength sound waves in a sample with ultrashort wavelength light pulses are disclosed in US Pat. Proposed to. Detection is performed with an interferometer using a delayed ultra-short duration probe light pulse, and in particular, monitoring of the surface displacement of the sample is possible, and a two-dimensional spatial image of surface acoustic wave propagation can be obtained as a function of time. This method is suitable for measuring surface acoustic waves in addition to longitudinal acoustic waves, for example, it is possible to derive thin film thickness and sound velocity. However, since this method relies on a series of simple excitation and detection processes, it is still not suitable for measuring samples containing multiple acoustic scatterers or inhomogeneous regions.
US Pat. No. 6,161,434 US Pat. No. 5,431,053 US Pat. No. 5,428,999 US Pat. No. 5,513,532 U.S. Pat. No. 4,710,030 JP 2001-228121 A (US Pat. No. 6,549,285) JP 2001-228122 A (US Pat. No. 6,552,800) JP 2001-228123 A (US Pat. No. 6,552,799) Japanese Patent Laid-Open No. 05-172739 M.M. Fink et al. The Journal of Nonlinearity, Volume 15, R1-R18, 2002 C. Draeger et al. , Applied Physics Letters, 72, 1567-1569, 1998. O. Matsuda. et al. , Journal of the Optical Society of America B, 19, 3028-3 041, 2002 O. B. Wright. et al. , Physical Review Letters, 69, pp. 1668-1671, 1992

  In general, short optical pump pulses with durations from 0.02 ps to 20 ns are used to generate sound waves that can typically include acoustic wavelengths on the order of 20 μm from sub-nanometers, and then similar durations as pump pulses By detecting this with a delayed short optical probe pulse having a physical property of a material such as a microstructure or nanostructure, typically on a relatively short length scale from 0.2 nm to 20 μm Non-contact measurement can be performed. This generation method generally produces short-wavelength acoustic disturbances that propagate in various directions of the sample. However, for samples containing a significant number of acoustic scatterers or inhomogeneous regions, or multiple layers, complex, relatively long coder waves are detected and very difficult to analyze.

  The present invention relates to an apparatus for measuring the physical properties of a sample based on the principle of acoustic time reversal and the generation of ultrashort light pulses and the detection of short wavelength sound waves. The present invention allows inspection of samples with characteristic structures that can be very small, but which may include a significant number of scatterers or inhomogeneous regions, or multiple layers. The present invention can be realized in a simple manner. In particular, by incorporating acoustic time reversal, there is no need to perform a predetermined playback of acoustic disturbances in a single stage, including a modulated sound generation method, and by using a series of playback local acoustic sources, the initial probe region It is clarified that a processed playback signal group that is effectively generated based on the acoustic propagation generated by the time reversal of the initial acoustic disturbance is obtained. For example, for each such local acoustic source, a single short pulse of playback optical pump radiation may be used, thereby necessarily requiring extra equipment for light modulation on a short time scale. Without reversing, the acoustic time reversal is realized. Furthermore, by using the synthesized playback signal group, the acoustic time reversal is realized without necessarily directly measuring the playback acoustic disturbance.

  The present invention can use acoustic time reversal of short wavelength acoustic waves on a short time scale, so that non-destructive of microstructures and nanostructures with a significant number of scatterers or inhomogeneous regions, or multiple layers. Can be evaluated. Moreover, since the present invention is an optical method, it does not require mechanical contact with the sample nor a binding medium in contact with the sample. The present invention also provides more valuable information at any stage of the measurement process, if necessary, such as spatial imaging of the propagation of acoustic disturbances in the sample being analyzed as a function of time. it can.

  It is an object of the present invention to provide an improved apparatus for measuring physical properties of a sample by generating and detecting short wavelength sound waves using short light pulses.

  It is a further object of the present invention to provide such an apparatus that measures the physical properties of a sample by making the acoustic time reversal method applicable to short wavelength sound waves.

  Providing an apparatus that allows inspection of samples with a characteristic structure that may be very small, but that may include a significant number of scatterers or inhomogeneous regions, or multiple layers It is a further object of the present invention.

  It is a further object of the present invention to provide such an apparatus that does not require mechanical contact with the sample and can be non-destructive.

  It is a further object of the present invention to provide such an apparatus that can measure the physical properties of a sample on a short length scale.

  It is a further object of the present invention to provide such an apparatus that can produce an image of acoustic disturbances propagating in the region of the sample with a small spatial resolution on the order of 1 micron or less.

  It is a further object of the present invention to provide an apparatus that can measure the physical properties of a sample and at the same time generate an aerial image of acoustic disturbances propagating in the sample as a function of time.

  Physical properties of samples including microstructures, nanostructures, thin films, multilayers, composites, polycrystalline samples, and other inhomogeneous regions characterized by voids, defects, cracks, dislocations or short length scales It is a further object of the present invention to provide such an apparatus that can be measured.

  It is a further object of the present invention to provide such an apparatus that can measure the physical properties of a sample composed of an opaque, transparent or translucent part or a combination thereof.

  It is a further object of the present invention to provide such an apparatus that can measure samples that are isotropic or anisotropic.

  All the operations necessary for a complete acoustic time reversal process, including the initial and playback phases of the measurement, can be achieved, and also by using a series of short light pulses that are not modulated in amplitude, phase or polarization state It is a further object of the present invention to provide such an apparatus that can be operated.

  A device that can operate all operations necessary for complete acoustic time reversal processing by using the playback signal group synthesized from the measurement results of the initial stage, without performing the actual playback stage measurement. It is a further object of the present invention to provide.

  An object of the present invention is to provide an apparatus for measuring the physical characteristics of a sample as described above.

  A feature of the present invention is that an optical pump probe method and an acoustic time reversal method are combined. While the upper limit of the acoustic frequency that can be handled by the existing acoustic time reversal method is about MHz, in the present invention, it is possible to handle an acoustic frequency in a much higher THz region. The measuring apparatus according to the present invention can not only acquire one-dimensional information mainly along the stacking direction of thin films, multilayer films, and nanostructures, but also can perform two-dimensional measurements along the surfaces of these samples. . It is also possible to refocus the sound wave at an arbitrary three-dimensional position. This method eliminates the need for ultrafast time modulation of the pump light pulse, so that the established standard laser picosecond acoustic technique can be used. According to the present invention, defects in the thin film, multilayer film, and nanostructure and the film thickness of each layer can be evaluated.

  The present invention allows a truly effective measurement instrument to monitor the maximum amount of information about the acoustic response of a sample on a relatively short length scale when illuminated by short light pulses, while at the same time a significant number of acoustics in the sample. It is based on the recognition that scatterers or inhomogeneous regions or multiple layers should also be acceptable.

  The present invention measures the physical properties of a sample by using a short pump light pulse to generate a local acoustic source that generates an acoustic disturbance in the sample, and then detecting the acoustic disturbance with a short probe light pulse. Features the device. A coherent or partially coherent radiation source providing a pulsed pump beam consisting of a series of periodic short light pulses, typically having a duration of 0.002 picoseconds (ps) to 20 nanoseconds (ns) There is. A typical example of such a radiation source is a high repetition rate mode-locked solid state laser. There are also radiation sources that provide a pulsed probe beam, which consists of a periodic series of short probe light pulses, usually from the same radiation source as the measurement beam, typically having a duration of 0.002 ps to 20 ns. Pump and probe radiation can be configured in the wavelength range of 10 nanometers (nm) to 100 microns and need not have the same central wavelength or optical spectral components. These pump and probe beams can be focused on the sample with a generally axisymmetric cross-section, such as a Gaussian cross-section.

  A single initial pump beam is directed onto the sample from a selected direction that may be collinear with the initial probe boom when incident on the sample. The role of the initial pump light pulse is to provide an initial local acoustic source that includes a wide range of acoustic wave number vectors and frequency components. The role of the single or multiple initial probe beams is to probe initial acoustic disturbances that occur on or inside the sample. This is made possible by modulation of the amplitude, phase or polarization state of the initial probe light emission due to the initial acoustic disturbance induced by the arrival of the initial pump light pulse at the sample. Alternatively, it can also be realized by modulating the difference in amplitude, phase, or polarization state of two initial probe light pulses incident on the sample after the initial pump light pulse arrives at the sample.

  The acoustic disturbance may be, for example, a longitudinal wave, a shear wave, a quasi-longitudinal wave, a quasi-shear wave, a surface acoustic wave, or an interface wave. Specific examples of waves that propagate near the surface include Rayleigh waves, Lamb waves, Love waves, Stoneley waves, general surface acoustic waves, pseudo surface acoustic waves, or surface skimming bulk waves . They create mechanical motion of the sample surface or interface within the sample, and various acoustic distortion fields in the sample. It is these disturbances that modulate the amplitude, phase or polarization state of the light pulse containing the initial probe beam.

  The interval of arrival time, called the initial delay time, is changed between the initial pump pulse and the initial probe pulse, causing an acoustically induced change as a function of this initial delay time during the initial probe light emission. A set of initial signals proportional to these acoustically induced changes as a function of the initial delay time during initial probe light emission is stored, for example, in a computer memory. The initial signal group is then processed to calculate the playback function. The playback function is calculated because it is implemented in a device and used to perform acoustic time reversal. This playback function is used for the convolution operation. Its purpose is to perform this convolution operation either when controlling the generation of the playback local sound source in the sample by a series of short light emission playback pump pulses or when processing the playback signal group later. Is to apply. The latter method has the advantage of not requiring any light modulation device.

  A single playback pump beam is directed to the sample from the chosen direction. When incident on the sample, it may be collinear with the playback probe beam. The role of the playback pump light pulse is to provide a playback local acoustic wave source. The role of the playback probe beam is to probe playback acoustic disturbances that occur on or inside the sample. This can be done by modulating the amplitude, phase, or polarization state of the playback probe light emission by the playback acoustic disturbances induced by the arrival of the playback pump light pulse at the sample, as in the initial stage of measurement. It becomes. Alternatively, this can also be achieved by modulating the difference in amplitude, phase or polarization state of two playback probe light pulses that are continuously incident on the sample after the playback pump light pulse arrives at the sample. A series of short playback pump pulses of light emission are directed to or near the initial probe region, where a playback local acoustic wave source is generated in the sample. This region may or may not be the same as the initial local acoustic wave source. By changing the interval called playback delay time between the arrival times of the playback pump pulse and playback probe pulse, acoustically induced changes as a function of this playback delay time during playback probe light emission. cause. A set of playback signals proportional to these acoustically induced changes as a function of playback delay time during playback probe light emission is stored, for example, in a computer memory.

  It is also possible to achieve acoustic time reversal without taking measurements in the actual playback phase. In this case, the acoustic time reversal is realized using a playback signal group synthesized from the initial signal group obtained in the initial stage measurement.

  In one embodiment of the invention, there are means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time. This is a method that depends on the playback function. Alternatively, in another embodiment, the playback signals are processed without the need for means for modulating the amplitude, phase or polarization state of the playback light pump radiation. In yet another embodiment, a playback pump pulse of light radiation may be selected by adjusting the amplitude, phase, or polarization state to facilitate generation of a playback local acoustic wave source. it can. The purpose of these operations is to effectively superimpose based on the acoustic propagation generated by the generation of the time-reversed initial acoustic disturbance or the time-gated portion of the initial acoustic disturbance in the initial probe region. Using the principle, the processed playback signal group is reproduced with a predetermined approximation.

  Traditional acoustic time reversal often uses an array of ultrasonic transducers, but one array can serve as both the initial and playback stages of the measurement, or two or more arrays in different regions of the sample In some cases, an initial stage and a playback stage are executed. In the present invention, the same effect as one or more ultrasonic transducer arrays can be obtained by scanning the position of the light beam interacting with the sample. The present invention is suitable for assembling necessary signals by measuring points one by one and summing them in space and time.

  For this purpose, means can be introduced for scanning the position or relative position of the initial local acoustic source or the initial probe region in order to generate a group of data representing the probed initial acoustic disturbances. A typical example of such a measurement is to use a single initial local source for the purpose of simulating a linear array of ultrasonic transducers for this initial stage of measurement, and to a linear array of initial probe regions. This is a method of creating a corresponding data group.

  The same is true when scanning the position or relative position of the playback local acoustic wave source and the playback probe region in order to create a data group representing the probed playback acoustic disturbance. A typical example of such a measurement uses a single playback local source for the purpose of simulating a linear array of ultrasound transducers for this playback stage of the measurement, and also for the playback probe region. This is a method for creating a data group corresponding to a linear array.

  In conventional acoustic time reversal, it is sometimes necessary to repeat the measurement, and it is necessary to utilize the time-reversed playback of the processed playback signal group in the playback probe region. This initiates a stage called the second initial stage of measurement. By repeating the initial stage and the playback stage, the acoustic disturbances in the sample tend to focus on the strongest acoustic scatterers present in the region exhibiting significant acoustic disturbances. With the present invention, exactly the same measurement is possible. The present invention is also suitable for time reversal operator decomposition methods that still require the creation of an array of one or more acoustic sources.

  Just as with conventional acoustic time reversal, it is possible to use a random acoustic cavity to focus the sound on the required area of the sample, and this method can be used to study a specific area of the sample. it can.

  In an aspect of the invention involving means for processing the playback signal group, this process includes convolution of the playback signal group with part or all of the time domain of the processing function a (t). Here, t is a delay time. By calculating the playback function, the time reversal of the initial acoustic disturbance in the initial probe region is effectively performed. This processing function a (t) is an impulse response function d (t) for the light detection of the initial acoustic disturbance in the initial probe region, and an acoustically induced change as a function of the initial delay time during the initial probe light emission. Calculated using knowledge of the proportional initial signal group g (t) and the impulse response function h (t) for light generation of the playback local acoustic source. The processing function a (t) can be filtered before use if necessary, one example being to digitally process the processing function a (t) with a constant bit number accuracy. .

  In one embodiment of the invention where there is a means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time, this modulation is a function of the playback delay time. Includes multiplication of the playback light pump radiation intensity by the playback function b (t). Here, the playback function b (t) is calculated using knowledge of d (t), g (t) and h (t). The playback function b (t) can be filtered before use if necessary, one example being digitally processing the playback function b (t) with a constant number of bits precision. .

  The function g (t) can be expressed for a desired variable such as strain, dislocation or surface particle velocity, as well as the function h (t). Similarly, for example, the function g (t) may be expressed in terms of sample reflectance change, relative reflectance change, or optical phase change.

  In another embodiment of the present invention, as a playback pump pulse of optical radiation, a pulse having a desired shape is selected by adjusting an amplitude, a phase, or a polarization state so as to facilitate generation of a playback local acoustic wave source. be able to. One example in this case is that the processing function a (t) or the adjustment function proportional to the time gated portion of the processing function a (t) is the shape of the intensity of the playback pump pulse as a function of the real time t. When it is chosen as. Here, a is the same function as the processing function. The shape adjustment can be realized by a spatial light modulator, for example.

  The processed playback signal group or the playback signal group corresponding to the result of the repetition is used to extract the acoustic properties in the sample in order to extract physical properties such as the size, elastic constant or density of the sample or part of the sample. It can be compared with a suitable theoretical model of the propagation and scattering of disturbances and the process of acoustic light generation and detection in the sample. This can be done, for example, by changing the trial value for the size, elastic constant or density of the sample or part of the sample and minimizing the difference between the predicted value of the appropriate theoretical model and the processed playback signal group. realizable.

  Other physical properties of the sample can also be derived, such as film thickness in the case of thin film samples. Relations such as internal stress, adhesion, ion implantation, moisture exposure, size and location of acoustic scatterers, particle size distribution, porosity, defect distribution, dislocation, adsorbent thickness on the surface, or microstructure It is also possible to measure quantities via their influence on the propagation of acoustic disturbances.

  In general, the sample may consist of parts that are transparent, translucent or opaque to optical probe or pump light radiation. In order to excite the sample, the sample should not be completely transparent to the pump light radiation. Samples are not limited to solid materials. Samples with solid, liquid or gaseous parts or other intermediate layers of matter can be studied.

  In order to increase the spatial resolution of the measurement above the optical diffraction limit, it is also possible to focus the light beam on the sample using near-field optical techniques such as tapered fibers. This method can increase the horizontal resolution to a fraction of 1 micron.

  It is also possible to obtain an aerial image representing the initial acoustic disturbance or the playback acoustic disturbance as a function of the initial or playback delay time. This is very useful for obtaining more information about acoustic disturbances.

  The number of probe pulses corresponding to each pump pulse can be arbitrarily selected and used. It is also possible to use a number of initial probe pulses or playback probe pulses and collect them together to form a continuous probe beam.

  The present invention is applicable to on-line monitoring of semiconductor wafers or integrated circuits. In particular, the present invention can be used to evaluate the presence and location of very shallow subsurface defects in a sample. In particular, the present invention can be used to detect the presence and location of subsurface defects in the presence of other inhomogeneous distributions. The present invention provides density, elastic constants or quantities that affect these properties, such as ion implantation, internal stress, surface roughness, interface roughness, adsorbed material on the surface, adhesion, particle size distribution, porosity, microstructure Alternatively, it can be used for evaluating the influence of moisture and the like.

  With the present invention, sample characteristics can be determined non-destructively without contact with the sample and can be used on-line conditions, eg, for semiconductor wafer products coated with a single or multiple layers of thin films. The invention can also be used for complex sample shapes.

  The feature of the present invention is that the optical pump probe method and the acoustic time reversal method are combined. While the upper limit of the acoustic frequency that can be handled by the existing acoustic time reversal method is about MHz, in the present invention, it is possible to handle an acoustic frequency in a much higher THz region. The measurement apparatus according to the present invention can not only acquire one-dimensional information of thin films, multilayer films, and nanostructures mainly along the stacking direction, but also can perform two-dimensional measurements along the surface of these samples. is there. It is also possible to refocus the sound wave at an arbitrary three-dimensional position. This method eliminates the need for ultrafast time modulation of the pump light pulse, so that the established standard laser picosecond acoustic technique can be used. According to the present invention, defects in the thin film, multilayer film, and nanostructure and the thickness of each layer can be evaluated.

  Embodiments of the present invention will be described below with reference to the drawings.

  1 and 2 show a first embodiment of the invention incorporating a time domain interferometer 3, which is particularly suitable for samples that are sufficiently opaque to the light irradiation used. This embodiment is particularly suitable for acoustic measurements at short wavelengths. An example of the time domain interferometer 3 used in this embodiment is described in detail in the above-mentioned Patent Documents 6 to 8 by the present inventor. This interferometer 3 can be used to take the difference in reflectivity change or phase change of two short probe pulses of optical radiation. The simplest form of measurement process consists of two stages: an initial stage and a playback stage. FIG. 1 represents the initial stage, and FIG. 2 represents the playback stage. Here, in FIG. 1, the initial pump beam and the initial probe beam are focused on the sample through the same lens. Also in FIG. 2, the playback pump beam and the playback probe beam are focused on the sample through the same lens.

  In the initial stage of the measurement, a short initial pump pulse 4 of light emission is focused by a lens 5 onto a local region 13 on a sample 6 having a free surface, with a normal or nearly normal incidence angle. Light absorption occurs in a region close to the surface of the sample 6 to create an initial local acoustic wave source. In particular, in the case of an elastically isotropic solid, a short wavelength pulse of a longitudinal sound wave is generated in the depth direction of the sample 6. The horizontal spatial profile of the initial pump beam 10 in the sample 6 is advantageously chosen to be axially symmetric as in the case of a Gaussian spot, for example, to facilitate data analysis.

In the first embodiment, two short initial probe pulses of light emission, called initial probe pulse 1 and initial probe pulse 2, separated by a selected time interval T _P are used. These initial probe pulses 1 and 2 form an initial probe beam 9 that is collinear with the initial pump beam 10, is focused by the lens 5, and is incident on the initial probe region 11 of the sample 6 at an incident angle that is perpendicular or almost perpendicular. . The initial probe region 11 is selected so as to optimally overlap with the local region 13 related to the initial pump beam 10 incidence. The horizontal spatial profile in the sample 6 of the initial probe beam 9 is advantageously chosen to be axisymmetric as in the case of a Gaussian spot, for example, to facilitate data analysis. If the initial pump beam 10 and the initial probe beam 9 do not have the same center wavelength, a dichroic beam splitter 12 can be selected. As time passes, the initial acoustic disturbance spreads in and out of the sample 6. As is well known to those skilled in the art, it is advantageous to use a periodic sequence of initial pump pulses and initial probe pulses and average the resulting signal group over time.

  In general, a coupling may occur between the initial probe light emission and the initial acoustic disturbance due to acoustic motion at or below the sample 6 or acoustic distortion within the sample 6. The out-of-plane component of the acoustic displacement of the surface or interface of the sample 6, the in-plane component of the acoustic displacement of the surface or interface of the sample 6, the out-of-plane component of the acoustic particle velocity at the surface or interface of the sample 6, the surface or interface of the sample 6 If one of the in-plane component of the surface particle velocity and the six components of the acoustic distortion tensor is coupled to the initial probe light emission, in principle, the initial acoustic disturbance can be detected. The acoustic distortion within the light penetration depth range of the initial probe beam 9 into the sample 6 generally affects the modulation of the initial probe light emission due to the photoelastic effect.

  After interacting with the initial acoustic disturbance in the sample 6, the initial probe light emission is generally monitored by a photodetector incorporated in the time domain interferometer 3 of FIG. In this embodiment, the initial probe irradiation is detected after reflection from the sample 6. As is well known to those skilled in the art, for example, optical chopping of the initial pump beam 10 with modulator 8 and lock-in detection can be used to improve the signal-to-noise ratio. To guide the initial pump beam 10, mirrors 14 and 15 are used.

  Changes that are acoustically induced during the initial probe light emission and that characterize the initial acoustic disturbance are the arrival time of the initial pump pulse 4 to the sample 6 and the arrival times of the two initial probe pulses 1 and 2 to the sample 6. Is detected with a predetermined initial time delay between This initial time delay can be introduced using an optical delay line or an electronically controlled delay between the initial pump pulse 4 and the initial probe pulse. These initial pump pulses 4 and initial probe pulses are usually chosen to be periodic pulse trains with the same repetition rate. Next, corresponding to the average arrival time of the two initial probe pulses 1 and 2 to the sample 6, the delay time between the arrival time of the initial pump pulse 4 to the sample 6 and the probe time of the initial acoustic disturbance is calculated. By varying, an initial signal group can be obtained that is proportional to the acoustically induced change in the initial probe light emission as a function of the initial delay time. The initial signal group is stored in a memory of a data processor 7 which is a computer, for example. The delay time scan can be performed in a continuous or sequential manner, which applies equally to the initial stage and the playback stage.

In the apparatus of FIG. 1, the time interval T _P of arrival of the two initial probe pulses 1 and 2 to the sample 6 is fixed, and the average arrival time of the two initial probe pulses 1 and 2 and the initial pump pulse 4 The initial delay time between is scanned. The same applies to the playback stage of FIG.

In one measurement example using the first embodiment of the present invention, the initial probe pulse 1 is set to enter the sample 6 earlier than the initial pump pulse 4. With such a configuration, it is possible to measure the reflectance change or phase change of the initial probe pulse 2 due to the interaction with the sample 6 by the interferometer 3. In another measurement example according to the first embodiment, both the initial probe pulses 1 and 2 are incident on the sample 6 after the initial pump pulse 4. If the time interval T _P is chosen to be sufficiently small, the interferometer 3 can measure the difference between the reflectivity change or phase change at this time interval that is approximately proportional to the time derivative of the reflectivity change or phase change. it can. Equivalent considerations apply to the playback stage of FIG.

  After recording the initial signal group, the measurement playback phase can begin as shown in FIG. In the first embodiment, exactly the same characteristics and incident light emission and the same device are used for the initial stage and the playback stage. Adopting this method to avoid duplication of equipment is generally but not essential. During the playback phase of the measurement, a playback pump beam 25 consisting of a short playback pump pulse 24 of a series of light emission is focused by the lens 5 and is incident at an angle of incidence perpendicular or nearly perpendicular to the local region 31 on the sample 6. To produce a playback local acoustic wave source. The local region 31 is optimally selected so as to overlap with the initial probe region 11 of the sample, but may be in the vicinity of the initial probe region 11. As time passes, the playback acoustic disturbance spreads in and out of the sample 6.

  A short playback probe pulse of a series of light radiation composed of pairs of light pulses creates a playback probe beam 29 that is focused by the lens 5 and is at a normal or nearly normal incidence angle at the playback probe region 33 of the sample 6. Is incident on. In the first embodiment, in particular, the playback probe region 33 is actually optimally selected so as to overlap the local region 13 and the initial probe region 11 for incidence of the initial pump beam 10. Each pair of probe pulses with a short light emission in the playback phase is referred to as a playback probe pulse 21 and a playback probe pulse 22. As is well known to those skilled in the art, it is advantageous to use a periodic sequence of playback pump pulses 24 and playback probe pulses 21, 22 and average the resulting signal group over time.

  The playback probe light emission after interaction with the playback acoustic disturbance in the sample 6 is monitored by the same photodetector used in the time domain interferometer 3 in the initial stage. As is well known to those skilled in the art, for example, optical chopping and lock-in detection of the playback pump beam 25 can be used to improve the signal to noise ratio.

  Changes that are acoustically induced during the playback of the playback probe light and are characteristic of playback acoustic disturbances are the arrival time of the playback pump pulse 24 to the sample 6 and the samples 6 of the two playback probe pulses 21 and 22. Detected with a predetermined playback time delay between the average arrival time to. In the first embodiment, the playback time delay is introduced by using the same equipment used in the initial stage. Corresponding to the average arrival time of the two playback probe pulses 21 and 22 to the sample 6, changing the delay time between the arrival time of the playback pump pulse 24 to the sample 6 and the probe time of the playback acoustic disturbance Thus, a playback signal group proportional to the acoustically induced change during playback probe light emission can be obtained as a function of the playback delay time. This playback signal group is stored in the memory of the data processor 7.

  In the first embodiment, there is no means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time. A group of playback signals processed using the principle of superposition, effectively by acoustic propagation generated by the generation of time-reversed initial acoustic disturbances or time-gated parts of the initial acoustic disturbances in the initial probe region Is reproduced with a predetermined approximation.

The processing function required for the first embodiment can be derived by considering the generation and detection of acoustic disturbances in the sample. d (t) is the impulse response function for the light detection of the initial acoustic disturbance in the initial probe region, and g (t) is the initial proportional to the acoustically induced change as a function of the initial delay time during the initial probe light emission. Let it be a signal group. Here, t represents the initial delay time. Let h (t) be an impulse response function for generating light from the playback local acoustic wave source, and let z ₁ (t) be a playback signal group before processing. Here, the same symbol t represents the playback delay time.

  The response function is defined with respect to the quantities to be measured. A specific example is to define an impulse response function with respect to the outward surface displacement u (t) of the surface of the sample 6 having a free surface. This analysis is derived on the assumption that the acoustic propagation in the sample 6 is not dissipative and follows a linear differential equation to meet the requirements of acoustic time reversal. Further assume that sample 6 is not subject to significant drift or creep. Also assume that the acoustic light generation and detection process is a linear process. However, the present invention can still be applied to general samples in various approximations, just like conventional acoustic time reversal. The function u (t) is defined as follows.

  Also, the function h (t) is the outward surface displacement u (t) generated at the position of the initial local source by the initial pump pulse 4 having an intensity that varies with time according to the Dirac delta function δ (t). Defined against. In this case, the function g (t) is defined by the following equation (2).

したがって

Therefore

初期プローブ領域の位置で時間反転された変位を再構成するためには、そこで変位ｕ（−ｔ）を発生させる必要がある。

In order to reconstruct the time-reversed displacement at the position of the initial probe region, it is necessary to generate a displacement u (−t) there.

First consider the case where a means for modulating the playback pump pulses in the real-time domain can be used, rather than in the first embodiment. This can be done, for example, using a playback pump pulse 24 whose intensity is proportional to the playback function a ₀ (t) as a function of real time (where the symbol t represents time in the real time domain). . Then, in the real-time domain,

を満たす必要がある。フーリエ（Ｆｏｕｒｉｅｒ）変換をとり、上記式（２）および（４）を使うと、
Ａ₀ （ω）＝Ｕ（−ω）／Ｈ（ω）＝Ｇ（−ω）／Ｈ（ω）Ｄ（−ω）
（ここで大文字はそれぞれ小文字のフーリエ変換を表し、ωは音響角振動数）となる。

It is necessary to satisfy. Taking a Fourier transform and using the above equations (2) and (4),
A ₀ (ω) = U (−ω) / H (ω) = G (−ω) / H (ω) D (−ω)
Where uppercase letters represent lowercase Fourier transforms and ω is the acoustic angular frequency.

Therefore, the necessary playback function a ₀ (t) can be obtained from the knowledge already obtained for the functions h (t) and d (t) and the measured initial signal group g (t). If the acoustic propagation in the sample 6 is not dissipative, follows a linear differential equation and is not subject to much drift or creep, the acoustic disturbance produced by the known reciprocity of the transformation function m (t) [in this case play It is easy to verify the time variation of the displacement u ₁ (t) at the position of the back probe region. That is, in the real time domain:

これはさらに式（１）を用いて、

This further uses equation (1),

となる。ここで、

It becomes. here,

を仮定した。これは十分な散乱体を含む媒体においては良い近似である。予期されるように、プレイバックプローブ領域の位置で測定された試料６の変位は、同じ位置における初期局所音波源発生後の対応する変位に対して時間反転されている。

Was assumed. This is a good approximation for media with sufficient scatterers. As expected, the displacement of the sample 6 measured at the position of the playback probe region is time-reversed with the corresponding displacement after the initial local sound source generation at the same position.

In the first embodiment, there is no means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time. However, it is still possible to perform acoustic time reversal by utilizing the superposition principle that applies to the sample 6 having a linear response. In order to generalize, an impulse response function for light detection of the playback acoustic disturbance in the playback probe region is displayed as d ₁ (t). However, in the case of the first embodiment, this is the same as the impulse response function d (t) for the light detection of the initial acoustic disturbance in the initial probe region.

この方程式は、上記の式を用いて解くことができ、遅延時間ｔの関数として、

This equation can be solved using the above equation, and as a function of the delay time t:

または
Ａ（ω）＝Ｇ（−ω）／Ｈ（ω）Ｄ（−ω）＝Ｄ（ω）Ｈ（−ω）／Ｇ（ω）（１０）
が得られる。ここで考慮する関数は実関数であり、したがってＦ（−ω）＝Ｆ（ω）^* である（ここでＦは任意の関数であり、＊は複素共役を表す）。処理関数ａ（ｔ）は、実際には上記した異なる例で得られた関数ａ₀ （ｔ）と同じものである。処理関数ａ（ｔ）を用いる畳み込み演算の結果として、初期プローブ領域における時間反転された初期音響擾乱または初期音響擾乱の時間ゲート処理された部分によって生成される音響伝播によって効果的に、重ね合わせの原理を用いて、処理されたプレイバック信号群が所定の近似で再現される。処理関数ａ（ｔ）は、例えば、理論または実験または両方から関数ｈ（ｔ）、ｄ（ｔ）を推定することにより、および測定された初期信号群ｇ（ｔ）を用いて、得ることができる。

Or A (ω) = G (−ω) / H (ω) D (−ω) = D (ω) H (−ω) / G (ω) (10)
Is obtained. The function considered here is a real function, and therefore F (−ω) = F (ω) ^* (where F is an arbitrary function and * represents a complex conjugate). The processing function a (t) is actually the same as the function a ₀ (t) obtained in the different example described above. As a result of the convolution operation using the processing function a (t), the time-reversed initial acoustic disturbance in the initial probe region or the acoustic propagation generated by the time-gated portion of the initial acoustic disturbance is effectively Using the principle, the processed playback signal group is reproduced with a predetermined approximation. The processing function a (t) can be obtained, for example, by estimating the function h (t), d (t) from theory or experiment or both, and using the measured initial signal group g (t). it can.

The above measurement using a series of single short playback pump pulses of equal intensity requires only one or more scans of the playback delay time over a selected time interval. This selected time interval can include, if desired, all of the coder wave that essentially represents the playback acoustic disturbance or the time-gated portion of this coder wave. Since a ₀ (t) and a (t) correspond to each other, the time-gated playback in the same real time as the time-gated function a ₀ (t) It is clear that it can be obtained by convolution with a time gated processing function a (t) as well. This is important because such time-gated playback is often useful in acoustic time reversal.

  In order to theoretically verify these results with respect to the first embodiment, a simulation of longitudinal acoustic generation by a light pulse with a short duration of 100 fs was performed. For such short light pulse durations, this duration is ignored compared to the acoustic pulse duration. FIG. 3 shows a multilayer comprising a series of polycrystalline metal films: 400 nm thick molybdenum, 100 nm thick silver, 100 nm thick nickel, 100 nm thick gold, 100 nm thick nickel, 100 nm thick silver and finally a silica substrate. A selected sample consisting of: The wavelengths of the pump pulses 4 and 24 are 630 nm, and the wavelengths of the probe pulses 1, 2, 21, and 22 are also 630 nm. It is assumed that the acoustic pulse is generated only by the thermoelastic effect. As will be apparent to those skilled in the art, this means that the main acoustic frequency is ν / 2πL = 64 GHz and corresponds to a sound wave wavelength of 80 nm. Here, ν = 6440 m / s is the longitudinal sound velocity of the molybdenum top layer film, and L = 16 nm is the optical absorption depth of the pump pulses 4 and 24.

As known to those skilled in the art, the photoelastic response of the sample 6 can be ignored at a certain probe light wavelength for a certain substance. In that case, the optical phase change δφ of the initial probe pulse 2 or the playback probe pulse 22 due to the interaction with the sample 6 is proportional to the outward surface displacement u (t), and g (t) = δφ = −2 ku. (T). Here, k is the wave number of the initial probe light emission. This means that d (t) = − 2 kδ (t), where δ (t) is a Dirac delta function. This is considered to be a good approximation for the optical wavelength used here, as shown in Non-Patent Document 4 above. Consider this example below. The calculated surface displacement h (t) due to the initial short pump pulse 4 is shown in FIG. As is well known to those skilled in the art, this function h (t) is proportional to θ (t) (1−e ^{− (νt / L)} ). Here, θ (t) is a unit step function. The initial probe signal group g (t) is shown in FIG. 5 and shows a long coder wave due to multiple scattering in multiple layers. The playback signal group z ₁ (t) obtained using a single short playback pump pulse 24 for a series of equal intensity impulses without time gating shows that the position of the playback local acoustic source is at this position. Since it is the same as the position of the initial local acoustic wave source in one embodiment, in this case it is the same function as the initial probe signal group except for the sign and offset. FIG. 6 shows the calculated processing function a (t). The processed playback signal group z (t) obtained from the convolution with the processing function a (t) is shown in FIG. As expected, the form of this function is the same as u (t). Depending on the accuracy of the numerical calculations used, noise is superimposed on the resulting signal. By improving the calculation accuracy and increasing the duration of the calculated coder signal, this noise can be arbitrarily reduced. The function z (t) exists on both sides of the point t = 0, but u (t) exists only when t> 0. This is a characteristic of the acoustic time reversal method.

  As is known from conventional acoustic time reversal, it is possible to probe various time regions of multiple scattering in the sample 6 by choosing a time-gated function for the processing function a (t). Is possible. It is also possible to repeat the measurement process. The signal group z (t) that effectively represents the sum of all playback acoustic disturbances detected in the initial local acoustic source region is time-reversed and effectively played back again in the same region of the sample 6. The entire process consisting of the initial and playback phases of the measurement can be executed by folding the processing function a (t) appropriate for each time as many times as desired. Similar to the conventional acoustic time reversal, such repetition causes the acoustic disturbance to change and converge to a high acoustic wave amplitude in the single or multiple major acoustic scatterers in the sample 6 in exchange for weak scatterers. Is generated. The processing function a (t) can be filtered before use if necessary, one example is by processing the processing function a (t) digitally with a constant bit number accuracy. is there.

  In the first embodiment, the impulse response function may also be defined in terms of longitudinal distortion in the sample. In the homogeneous region of the sample, a longitudinal distortion pulse having a general shape returning to the free surface of the sample 6 after interaction with the acoustic scatterer in the sample can be expressed by the following equation.

η (z, t) = η ₀ (t + z / ν) −η ₀ (tz / ν) (11)
Actually, η ₀ = − (1 / 2ν) du (t) / dt. As is generally done, the measured initial signal group g (t) can be used to represent the reflectivity change R or relative reflectivity change δR / R of the initial probe pulse 2 or playback probe pulse 22. The complex relative reflectance change δr / r with respect to the distortion pulse of Expression (11) is well known to those skilled in the art as described in Non-Patent Document 3 above. The impulse response function related to the longitudinal strain η ₀ (t) is simply derived, and instead of equation (2), the following equation is used:

ｇ（ｔ）が複素相対反射率変化の実部であるとすると、

If g (t) is the real part of the complex relative reflectance change,

が得られる。ここで、Ｒｅは実部、ｋは初期プローブ光放射の波数、Ｎは初期プローブ放射に対する試料６の複素屈折率、ｄＮ／ｄηは初期プローブ放射に対する適当な光弾性定数、νは試料６の縦方向音速、θ（ｔ）は単位ステップ関数、ｉ＝√（−１）である。式（１３）は、実験ではよく起こりうる、相対反射率変化が比較的小さい場合に適用される。その場合、Ｒｅ（δｒ／ｒ）＝（１／２）δＲ／Ｒである。

Is obtained. Here, Re is the real part, k is the wave number of the initial probe radiation, N is the complex refractive index of the sample 6 with respect to the initial probe radiation, dN / dη is an appropriate photoelastic constant for the initial probe radiation, and ν is the longitudinal length of the sample 6. Directional sound velocity, θ (t) is a unit step function, i = √ (−1). Equation (13) is applied when the relative reflectance change, which can occur frequently in experiments, is relatively small. In that case, Re (δr / r) = (1/2) δR / R.

  Alternatively, if g (t) is the imaginary part of the complex relative reflectance change

が得られる。式（１４）も相対反射率変化が比較的小さい場合に適用される。その場合Ｉｍ（δｒ／ｒ）＝δφである。ここでδφは初期プローブパルス２またはプレイバックプローブパルス２２の光学位相変化である。

Is obtained. Equation (14) is also applied when the relative reflectance change is relatively small. In that case, Im (δr / r) = δφ. Here, δφ is an optical phase change of the initial probe pulse 2 or the playback probe pulse 22.

  Equations (13) and (14) are useful in estimating the processing function a (t) required in the first embodiment. Both of these equations (13) and (14) take into account the effects of the photoelastic response and surface displacement of the sample 6. These definitions are useful when the probe pulse 1 or 21 is set to be incident on the sample 6 before the pump pulse 4 or 24, respectively.

  As known to those skilled in the art, the photoelastic response of the sample 6 can be ignored at a certain probe light wavelength for a certain substance. Such an example has been discussed above. In that case, the optical phase change δφ of the initial probe pulse 2 is proportional to the outward surface displacement u (t), whereby g (t) = δφ = −2 ku (t) can be defined. Here, k is the wave number of the initial probe light emission. This means d (t) = − 2 kδ (t), where δ (t) is a Dirac delta function. These definitions are useful when the probe pulse 1 or 21 is set to be incident on the sample 6 before the pump pulse 4 or 24, respectively.

In another type of measurement associated with the first embodiment, initial probe pulses 1 and 2 or playback probe pulses 21 and 22 are applied to sample 6 after initial pump pulse 4 or playback pump pulse 24, respectively. Incident. If the time interval T _P is selected to be sufficiently short, the interferometer 3 can measure the difference in reflectance change or phase change in this time interval. These differences are approximately proportional to the time derivative of reflectivity change or phase change.

For the sample 6 in which the photoelastic response can be ignored, the change in the optical phase difference Δφ between the initial probe pulse 2 and the initial probe pulse 1 due to the interaction with the sample is the time of the outward surface displacement u (t). Proportional to differentiation. g (t) = Δφ / T _P = −2 kdu (t) / dt. This again means d (t) = − 2 kδ (t).

  In the general case where the photoelastic effect and surface or interface displacement of the sample 6 affects the acoustically induced changes in probe light emission, the function d (t) is more complex as discussed above. Take a shape. The form of d (t) can be derived from theory or experiment or both. For example, as reported in Non-Patent Document 3 above, a general multilayer theoretical form can be used.

The impulse response function h (t) can be defined in terms of convenient quantities such as u (t) or η ₀ (t). This function h (t) can also be determined from theory or experiment or both. As is known to those skilled in the art, when estimating h (t) for a sample 6 such as a metal or semiconductor, the effects of thermal diffusion and carrier diffusion may have to be taken into account. The type of sound wave generated in the sample 6 depends on, but is not limited to, the geometry of the sample and the generation of carriers and temperature changes in the sample 6. Sound waves are generated, for example, by thermoelastic effects, by deformation potential effects, by electrostriction, or by ablation.

  The use of a single lens 5 for both the pump beam and the probe beam in the first embodiment is not the only possible design. The use of two lenses, one for the pump beam and one for the probe beam, is also an option. In general, any desired focusing system can be used, including the use of mirrors. It is also possible to select various light incident angles. It is also possible to select different light components at the beginning of the measurement and at the playback stage. Furthermore, the pump beam and the probe beam may be incident from the opposite side of the sample 6 or from different sides.

In yet another design similar to the first embodiment, the playback signal group z _l (t) need not be obtained by direct measurement. For example, if the sample can be identified in the vicinity of the initial pump region and the initial probe region, a playback signal group synthesized under the assumption of d _l (t) = d (t)

が得られる。この場合、式（３）よりｚ_l （ｔ）はｇ（ｔ）に等しい。ｄ_l （ｔ）≠ｄ（ｔ）の場合でもｍ（ｔ）を測定し、理論に基づいてｄ_l （ｔ）を計算すれば、ｚ_l （ｔ）を合成できる。ｍ（ｔ）は式（３）のフーリエ変換と既知のＤ（ω）、Ｈ（ω）、およびＧ（ω）より、Ｍ（ω）＝Ｇ（ω）／〔Ｈ (ω）Ｄ (ω）〕として求められる。このようにしてプレイバック段階の測定を省略することにより、測定時間の短縮が図られる。

Is obtained. In this case, z _l (t) is equal to g (t) from equation (3). m (t) is determined even if d _l (t) ≠ d in (t), by calculating the d _l (t) based on the theory, can be synthesized z _l (t). m (t) is M (ω) = G (ω) / [H (ω) D (ω from the Fourier transform of equation (3) and known D (ω), H (ω), and G (ω). )]. By omitting the measurement at the playback stage in this way, the measurement time can be shortened.

  In a second embodiment, also corresponding to FIGS. 1 and 2, there are means for modulating the amplitude, phase or polarization state of the playback optical pump radiation as a function of the playback delay time. In this embodiment, a multiplication of the intensity of the playback optical pump radiation that is a function of the playback delay time t and the playback function b (t) is included. This can be done with a modulator 8, which can be, for example, an electro-optic modulator or an acousto-optic modulator. This embodiment is also particularly suitable for short wavelength acoustic measurements and for samples 6 that are sufficiently opaque to the light radiation used.

  Taking again the example of defining the impulse response function with respect to the outward surface displacement u (t) of the surface of the sample 6 having a free surface, the required form of the playback function b (t) is the equation:

から得ることができる。ここでｔは遅延時間である。これは、次の式

Can be obtained from Here, t is a delay time. This is the formula

に変換することができる。したがって、プレイバック関数ｂ（ｔ）は、例えば、理論または実験、またはその両方から関数ｈ（ｔ）およびｄ（ｔ）を推定することにより、および測定した初期信号群ｇ（ｔ）を用いて得ることができる。必要であれば、プレイバック関数ｂ（ｔ）には、使用前にフィルターをかけてもよく、一つの例は一定のビット数精度でのｂ（ｔ）のディジタル処理である。この第２の実施態様においては、プレイバック信号群を処理する必要はない。プレイバック関数はまた、原則として、それによりプレイバック音響擾乱中に対応する変化が生じる場合には、プレイバック光ポンプ放射の位相または偏光を変調するために適用することができる。

Can be converted to Thus, the playback function b (t) is, for example, by estimating the functions h (t) and d (t) from theory or experiment, or both, and using the measured initial signal group g (t). Can be obtained. If necessary, the playback function b (t) may be filtered before use, one example being digital processing of b (t) with constant bit number accuracy. In the second embodiment, it is not necessary to process the playback signal group. The playback function can also be applied in principle to modulate the phase or polarization of the playback light pump radiation if it causes a corresponding change during the playback acoustic disturbance.

  In a third embodiment, also corresponding to FIGS. 1 and 2, there is no means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time, but instead the light emission A modulator 8 that modulates the playback pump pulse 24 in real time is used. This embodiment is also particularly suitable for samples 6 that are sufficiently opaque to short wavelength acoustic measurements and the light radiation used. These playback pump pulses can be selected to be pulses with the desired shape of amplitude, phase or polarization state to facilitate generation of the playback local acoustic wave source. One example in this case is that the function proportional to a (t) or the time-gated part of a (t) is the shape of the intensity of the playback pump pulse as a function of real time t. This is the case. Here, a is the same function as the processing function. The shape adjustment can be realized by a spatial light modulator, for example. As known to those skilled in the art, it is difficult to generate light pulses of arbitrary duration and shape on a picosecond time scale. Thus, combining the techniques used in the first three embodiments may be advantageous over using the third embodiment alone.

  In the fourth embodiment, the same apparatus as in FIGS. 1 and 2 is used, but the probe pulses 1 and 21 are not used. This embodiment is also particularly suitable for short wavelength acoustic measurements and for samples 6 that are sufficiently opaque to the light radiation used. In this case, the interferometer 3 can be, for example, a Michelson interferometer, as disclosed in the above-mentioned Patent Document 9. In this fourth embodiment, the interferometer 3 can also be replaced by a simpler system for reflectance measurements only. In this embodiment, the function is similar to the first three embodiments, except that no difference in reflectance or phase change between the two probe pulses is measured. Instead, the reflectivity or phase change will be measured for a single probe pulse that interacts with the sample 6.

  Conventional acoustic time reversal often uses an array of ultrasonic transducers, but a single array that serves as both the initial and playback phases of the measurement, and the initial and playback in different regions of the sample 6 In some cases, two or more sequences are used to perform the steps. By scanning the position where the light beam interacts with the sample, the present invention can achieve the same effect as an array of single or multiple ultrasonic transducers.

  A fifth embodiment is shown in FIGS. FIGS. 8 and 9 represent the same apparatus as FIGS. 1 and 2, but the location or relative position of the initial local acoustic source or initial probe region is used to generate a data set representing the probed initial acoustic disturbances. Means for scanning, and means for scanning the position or relative position of one or more local acoustic sources in the playback probe region probed to create a data group representative of the playback acoustic disturbances probed. Each has been introduced. In the fifth embodiment, these scanning means correspond to means for scanning the positions of the initial pump beam 10 and the playback pump beam 25 by adjusting the angles of the mirrors 14 and 15. Further, the position of the sample 6 can be scanned. This embodiment can be used, for example, for measurement of acoustic disturbance in the sample 6 that propagates in the in-plane direction such as surface acoustic waves.

  In this embodiment, in general, the initial probe region 11 is not the same as the local region 13 for initial pump beam incidence. However, as in the case of the first embodiment, the local region 31 for incidence of the playback pump beam 25 is optimally selected so as to overlap the initial probe region 11 of the sample 6.

  In the simplest implementation of the fifth embodiment, the playback probe region 33 is chosen to optimally overlap the local region 13 for the initial pump beam 10 incidence. To achieve this, the playback pump beam 25 and the sample 6 are scanned after the initial stage of measurement. By this means, based on the acoustic propagation generated by the time-reversed initial acoustic disturbance or the time-gated part of the initial acoustic disturbance in the initial probe region 11, the local region is effectively The processed playback signal group detected at 33 can be reproduced with a predetermined approximation.

  In this fifth embodiment, the same reasons as described in the first embodiment can be used to achieve the playback stage. As before, it is sufficient to determine the impulse response function and the required processing function or playback function from experiments and / or theory to perform one of the methods of performing the playback stage.

  In a more complex implementation of the fifth embodiment, a single initial local source is used to simulate a linear array of ultrasound transducers for this initial stage of measurement, and a linear array of initial probe regions 11 is used. A series of initial signal groups corresponding to is created. The position of the local region 31 relative to the incidence of the playback pump beam 25 is then scanned one after another in each of these initial probe regions 11 and a playback phase is performed for each of these positions. Furthermore, the position of the playback probe region 33 corresponding to the predetermined initial probe region 11 can be scanned for each of the playback stages. In this way, a linear array of ultrasonic transducers for the measurement playback phase can be simulated.

  The fifth embodiment is very suitable for performing measurement independently point by point while forming a necessary signal group by a summation method in space and time. This embodiment is also suitable for time reversal operator decomposition methods that require an array of one or more acoustic sources to be created.

  In connection with the fifth embodiment, just as in the case of conventional acoustic time reversal, a random acoustic cavity can also be used to allow acoustic focusing to a desired area of the sample, In this way, specific areas of the sample can be studied. An example of such a cavity is a D-shaped region on the surface of the sample 6, on which a metal film having an acoustic impedance different from that of the neighboring portion of the sample 6 is deposited.

  The scanning area in the fifth embodiment need not be limited to two dimensions. A three-dimensional space scan is also possible. This may be necessary, for example, when the sample 6 has a curved surface or is partially transparent. Using such scanning means, it is also possible to obtain an aerial image representing an initial acoustic disturbance or a playback acoustic disturbance as a function of the initial or playback delay time. This is very useful for obtaining more information about acoustic disturbances.

  In all of these embodiments, the processed playback signal group or the signal group corresponding to the result of the repetition extracts physical properties such as the size, elastic constant or density of the sample 6 or part of the sample 6. Therefore, propagation and scattering of acoustic disturbances in the sample 6 and comparison with an appropriate theoretical model of the acoustic light generation and detection process in the sample can be made. This can be done, for example, by changing the trial value of sample 6, or part of sample 6, the elastic constant or density to minimize the difference between the predicted value of the appropriate theoretical model and the processed playback signal group. Can be executed.

  Other physical characteristics of the sample 6 can also be derived, for example, the film thickness in a thin film sample. Also, for example, internal stress, ion implantation, adhesion, moisture exposure, size and location of acoustic scattering objects, particle size distribution, porosity, defect distribution, dislocation, thickness of adsorbed material on the surface, or micro It is also possible to measure related quantities such as structure through the influence of these properties on the propagation of acoustic disturbances.

  In general, the sample 6 may consist of parts that are transparent, translucent or opaque to the optical probe or pump light radiation. In order to generate a sound source in the sample 6, the sample 6 should not be completely transparent to the pump light radiation. Sample 6 is not limited to a solid material. Samples 6 containing solids, liquids or gases or other intermediate phases can be studied. For example, as in the case of the sample 6 having a transparent coating, the positions of the local regions 11, 13, 31 and 33 do not necessarily correspond to the region near the surface of the sample 6. Similarly, the medium in front of the sample 6 may be a solid, liquid or gas. The sample 6 is not limited to one having two parallel surfaces. For example, it is possible to have a local acoustic wave source on one face of the crystal and a probe region on another face inclined at an angle.

  In order to increase the spatial resolution of the measurement beyond the diffraction limit of the light, it is also possible to focus the light beam on the sample 6 using a near-field optical technique such as a tapered fiber. In this way, the horizontal resolution can be reduced to an order of a fraction of a micron.

  The experimental apparatus that can be used in connection with the present invention is not limited to the specific apparatus described above. It is also possible to use an apparatus that includes modulation of the polarization state of the probe beam or modulation of the probe beam transmitted through the sample. Any number of probe pulses can be selected and used for each pump pulse. It is also possible to use a number of initial or playback probe pulses and integrate these pulses to form a continuous probe beam. In this case, a real-time data acquisition system including a streak camera or a high-frequency photodetector can be used in the apparatus instead of the optical delay line. It is also important to recognize that the sample 6 may be transparent to the optical probe radiation, since the phase change of the optical probe radiation can be detected. When scanning the position or relative position of the local acoustic wave source on the sample 6, the position of the probe beam relative to the apparatus and the position of the sample 6 and the pump beam can also be scanned.

  The incident angle of the pump beam or the probe beam may be oblique, but if necessary, a circularly symmetric point may be created on the sample 6 using a cylindrical lens to correct the beam profile. Tilting the optical probe beam in the sample 6 allows the amplitude of the acoustically induced change in the initial or playback probe light emission through changes in coupling with various components of the acoustic distortion within the sample 6. It may be advantageous to improve. In addition, it is not necessary to use a single center wavelength or a single polarization of the pump beam or probe beam. The use of multiple wavelengths or polarizations of the pump beam or probe beam may be advantageous, for example in the case of a multilayer sample, to increase the efficiency of sound generation or detection at various parts in the sample 6.

  As a horizontal spatial distribution of the initial or playback probe pulse of light emission in the sample 6, a smooth axisymmetric profile is usually selected, but other profiles can be selected. Similarly, as the time profile of the initial light emission or playback probe pulse in the sample 6, a smooth time profile is usually selected, but other profiles can also be selected.

  As another method for obtaining a necessary signal group, there is a parallel acquisition technique (Parallel Acquisition Technique). In this technique, a series of points, lines or substantial areas of the sample 6 are irradiated with a pump beam or a probe beam in order to probe multiple portions of the sample 6 simultaneously. Multiple light sources or photodetectors can be used for this purpose. As a result, a necessary signal group can be acquired more quickly.

  Since this invention was comprised as mentioned above, there can exist the following effects.

  The apparatus described in the present invention is very effective in a wide range of applications relating to the measurement of material properties on a short length scale. The apparatus of the present invention is versatile and is isotropic or anisotropic independent of various samples, or samples composed of a combination of opaque, transparent or translucent parts, and crystalline state, It can be applied to a sample in which an acoustic scatterer is distributed.

  The device also allows non-contact and non-destructive measurement of physical properties of the sample such as size, density or elastic constant. In addition, for example, internal stress, ion implantation, adhesion, moisture exposure, acoustic scatterer size and location, particle size distribution, porosity, defect distribution, dislocation, adsorbent thickness on the surface, or microstructure And related physical properties can be probed. It is also possible to probe the physical properties of different layers or parts of the sample. With this device, before deriving the physical properties, imaging of the acoustic disturbances on or inside the sample and whether the sample has the required quality, for example sufficient horizontal elastic or mechanical homogeneity Can be detected simultaneously.

  This approach allows a wider range of applications for measuring physical properties on short length scales than other non-contact acoustic devices that do not include acoustic time reversal. In particular, this approach allows non-destructive evaluation of microstructures and nanostructures with a significant number of scatterers or inhomogeneous regions, or multiple layers.

(Experimental result)
To experimentally test the apparatus of the present invention, the apparatus of FIGS. 1 and 2 is employed. The experimental method is based on an optical pump and probe technique that periodically excites sound waves using a sub-picosecond initial pump pulse 4 of light emission in the visible light range. Sample 6 is composed of multiple layers, and its cross section is shown in FIG. A polished silica substrate was coated with 100 nm polycrystalline nickel, and an upper layer of 400 nm polycrystalline gold was formed thereon. Sample 6 diameter 6 μm (peak half-width intensity) at normal incidence of double frequency initial pump pulse 4 from a mode-locked Ti: sapphire pulse laser with a center wavelength of 415 nm, duration of the order of 1 ps (picosecond) and repetition rate of 80 MHz ) Focus to a Gaussian spot on the flat gold surface of sample 6 using lens 5 to the spot of order. The energy of the incident pump pulse is typically on the order of 0.3 nJ. Thereby, in the case of a gold thin film, a transient temperature rise of the sample occurs on the order of 10K. These initial pump pulses 4 or playback pump pulses 24 generate a complex three-dimensional acoustic field in the sample 6. In particular, longitudinal acoustic pulses are generated in the depth direction. These longitudinal acoustic wave wavelengths are mainly dominated by the thermoelastic effect combined with electron diffusion into the sample to a subsurface depth on the order of 140 nm, and are generally in the 600 nm region in this experiment. This corresponds to a broadband frequency range on the order of 5 GHz.

The apparatus includes a very stable common path Sagnac interferometer 3 using an optical probe pulse with a center wavelength of 830 nm and duration on the order of 200 fs, focusing at normal incidence through a lens 5 to a Gaussian spot of approximately 3 μm in diameter. This interferometer is installed to measure the optical phase change. In the initial stage, two initial probe pulses 1 and 2 extracted from the same laser in synchronism with the optical pump pulse are used. The initial probe pulse 1 is set to enter the sample 6 before the initial pump pulse 4 so that the change of the initial probe pulse 2 due to the interaction with the sample 6 is measured by the interferometer 3. The arrival time interval T _P of the two probe pulses to the sample is chosen to be 3200 ps, which is long enough to allow recording of signals over a long delay time interval. For the gold thin film used in this experiment, the change proportional to the optical phase change of the initial probe pulse 2 or the playback probe pulse 22 that is acoustically induced during the initial probe or playback probe light emission is mainly the sample. Expected to result from 6 outward surface displacements. However, at probe wavelengths other than those used here, or at thin films of other materials such as, for example, aluminum, the photoelastic effect can primarily affect acoustically induced changes during probe light emission. . By using a variable time delay between the pump and probe pulses, the initial or playback signal group can be recorded. To improve the signal-to-noise ratio, optical chopping and lock-in detection at 1 MHz of the pump beam is used.

  FIG. 11 shows the initial signal group g (t) recorded for the multilayer sample. A series of acoustic echoes is visible as a result of multiple acoustic scattering of sound waves inside the multilayer sample 6. The dissipative acoustic attenuation effect on the ultrasonic frequencies included here is small.

Fig. 4 shows the usage of the first embodiment when the position of the playback local acoustic source is the same as the position of the initial local acoustic source. Since the playback stage is performed using the same light pulse, the playback signal group z ₁ (t) in the first embodiment that does not use optical pump modulation is therefore the initial signal group g (t) of FIG. It is considered the same and no additional experimentation is required to obtain it. Therefore, it can be assumed that z ₁ (t) = g (t).

In this embodiment, it is necessary to perform convolution with the playback processing function a (t). FIG. 12 shows the calculated function a (t).

  The processed playback signal group z (t) obtained using equations (9) and (10) is shown in FIG. As expected, the shape of z (t) is a good approximation to the time reversal of the initial signal group g (t), except for the sign and offset. In this case, the time shapes of g (t) and z (t) are approximately symmetrical in the vicinity of the respective local maximum values.

  In the analysis, the slow background change produced by thermal diffusion into the sample was subtracted from the data. This is generally done on an as-needed basis, as the thermal variation changes slowly in time compared to the signal representing the acoustic disturbance.

  This experiment demonstrates the feasibility of the present invention in a simple but effective manner.

  In addition, this invention is not limited to the said Example, Based on the meaning of this invention, a various deformation | transformation is possible and these are not excluded from the scope of the present invention.

  The present invention is applicable to on-line monitoring of semiconductor wafers or integrated circuits. In particular, the present invention can be used to infer the presence and location of very shallow subsurface defects in a sample. In particular, the presence and location of subsurface defects can be detected using the present invention even if other inhomogeneous regions are distributed. The present invention also includes density, elastic constants or quantities that affect these properties, such as ion implantation, internal stress, surface roughness, interface roughness, adsorbed material on the surface, adhesion, particle size, porosity, micro It can be used to evaluate the effect of structure or moisture.

  According to the present invention, the characteristics of a sample can be determined in a non-destructive and non-contact manner, and can be used, for example, under on-line conditions in the production of a semiconductor wafer coated with a single layer or a multilayer thin film. It can also be used when the sample geometry is complex.

  There are many other industrially applicable fields for the present invention. According to the present invention, when a thin film or layer is formed on a sample, when subsurface structures are present in the sample, or when it is necessary to measure the properties of a sample on a short length scale, the performance of the final product is achieved. A wide range of information regarding the physical properties of the sample can be provided at any time, such as mechanical or elastic properties that are critical to the environment.

It is a schematic diagram which shows the apparatus of the 1st embodiment of this invention, and shows the initial stage of a measurement. FIG. 2 is a schematic diagram showing the apparatus of the first embodiment of the present invention, showing the playback stage of measurement. It is sectional drawing of the multilayer sample selected for the simulation of the acoustic generation of the vertical direction by the short light pulse in the 1st embodiment of this invention. FIG. 4 shows the calculated surface displacement h (t) due to a short initial pump pulse for a multilayer sample in the first embodiment of the invention. It is a figure which shows the initial stage probe signal group g (t) in the 1st embodiment of this invention. It is a figure which shows the calculated process function a (t) in the 1st embodiment of this invention. It is a figure which shows the processed playback signal group z (t) obtained from the convolution by the processing function a (t) in the 1st embodiment of this invention. It is a schematic diagram which shows the apparatus containing the means of position scanning which shows the 5th embodiment of this invention, and shows the initial stage of a measurement. FIG. 7 is a schematic diagram showing an apparatus including means for position scanning showing a fifth embodiment of the present invention, and showing a measurement playback stage; It is sectional drawing of the multilayer sample selected for experiment of this invention. It is a figure which shows the initial stage signal group g (t) recorded with respect to the multilayer sample of FIG. FIG. 5 shows a calculated function a (t) corresponding to a multilayer sample selected for the experiment of the present invention. FIG. 4 shows a processed playback signal set z (t) corresponding to a multilayer sample selected for the experiment of the present invention.

Explanation of symbols

1, 2 Initial probe pulse 3 Time domain interferometer 4 Initial pump pulse 5 Lens 6 Sample 7 Data processor 8 Modulator 9 Initial probe beam 10 Initial pump beam 11 Initial probe region 12 Beam splitter 13, 31 Local region 14, 15 Mirror 21 , 22 Playback probe pulse 24 Playback pump pulse 25 Playback pump beam 29 Playback probe beam 33 Playback probe region

Claims (43)

(A) means for generating an initial acoustic wave source localized in the sample by a short initial pump pulse of light radiation;
(B) means for probing initial acoustic disturbances occurring in the initial probe region near or within the surface of the sample by a short initial probe pulse of light emission;
(C) To change the delay time, called the initial delay time, between the arrival time of the initial pump pulse of the predetermined light emission to the sample and the probe time of the initial acoustic disturbance due to the initial pulse of probe light emission. Means of
(D) after the interaction between the initial probe light radiation and the sample, a change in the characteristic of the initial acoustic disturbance in the initial probe region acoustically induced during the initial probe light radiation is represented by the initial delay; Means for detecting as a function of time;
(E) means for storing, as a function of the initial delay time, an initial signal group proportional to the change acoustically induced during the initial probe light emission;
(F) to calculate a playback function used to control the generation of a playback local acoustic wave source in the sample by a short playback pump pulse of a series of light emissions at or near the initial probe region; Means for processing the initial signal group;
(G) means for generating the playback local acoustic source by a short playback pump pulse of the series of light emissions at or near the initial probe region of the sample;
(H) means for probing playback acoustic disturbances occurring in the playback probe region near or within the surface of the sample with a short playback probe pulse of a series of light emissions;
(I) called the playback delay time between the arrival time of the playback pump pulse of the predetermined light emission to the sample and the probe time of the playback acoustic disturbance due to the playback pulse of the probe light emission Means for changing the delay time;
(J) means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time;
(K) After the interaction between the playback probe light radiation and the sample, a change in the characteristics of the playback sound disturbance in the playback probe region acoustically induced during the playback probe light radiation. Means for detecting;
(L) means for storing the playback signal group proportional to the change acoustically induced during the playback of the playback probe light as a function of the playback delay time;
(M) means for processing the playback signal group with a processing function to create a processed playback signal group;
(N) Scan the position of the initial local sound source or the initial probe region or the relative position of the initial local sound source and the initial probe region to create a data group representing the probed initial acoustic disturbance. Means for
(O) In order to create a data group representing the probed playback acoustic disturbance, the position of the playback local sound source or the playback probe region or the playback local sound source and the playback probe region Means for scanning the relative position;
(P) means for repeating the above steps if necessary;
(Q) Superimposing processed playback signal groups effectively resulting from acoustic propagation due to the occurrence of time-reversed initial acoustic disturbances or time-gated initial acoustic disturbances in the initial probe region Said means enabling reproduction with a predetermined approximation using the principle of:
(R) the processed playback signal group corresponding to the processed playback signal group or the result of repetition by the means (p), and propagation and scattering of the initial and playback acoustic disturbances in the sample; And means for comparison with a suitable theoretical model of the acoustic light generation and detection process in the sample,
As a result, an apparatus for measuring a physical property of a sample, wherein the physical property of the sample is extracted.   The apparatus for measuring physical characteristics of a sample according to claim 1, wherein the playback signal group is synthesized from the initial signal group without using the playback acoustic disturbance.   2. The method for measuring physical properties of a sample according to claim 1, wherein the initial pump pulse of light radiation or the playback pump pulse of a series of light radiations is focused into an axially symmetric beam. apparatus.   2. The physical properties of a sample according to claim 1, wherein the initial pump pulse of light radiation or the playback pump pulse of a series of light radiation is focused into a beam having a Gaussian cross section approximately. Device for measuring.   2. The method for measuring physical properties of a sample according to claim 1, wherein the initial probe pulse of light radiation or the playback probe pulse of a series of light radiation is focused into a beam having axial symmetry. apparatus.   2. The physical properties of a sample according to claim 1, wherein the initial probe pulse of light radiation or the playback probe pulse of a series of light radiation is focused into a beam having a Gaussian cross section approximately. Device for measuring.   The means for probing the initial or playback acoustic disturbance by the probe light radiation includes the use of modulation of the amplitude, phase or polarization state of the probe light radiation by interaction with the sample. An apparatus for measuring physical properties of a sample according to item 1.   Means for probing the initial or playback acoustic disturbance with the probe light radiation includes the incidence of light on the sample from the same side of the sample with respect to the initial or playback pump pulse of the light radiation. The apparatus for measuring physical properties of a sample according to claim 1.   Means for probing the initial or playback acoustic disturbance by the probe light radiation includes the incidence of light on the sample from a different side of the sample relative to the initial or playback pump pulse of the light radiation. The apparatus for measuring physical properties of a sample according to claim 1.   The apparatus for measuring physical properties of a sample according to claim 1, wherein the initial probe region is substantially the same as the region of the initial local acoustic wave source.   The apparatus for measuring physical properties of a sample according to claim 1, wherein the playback probe area is substantially the same as the area of the playback local acoustic source.   2. The sample according to claim 1, wherein the means for scanning the position of the initial local sound source or the initial probe region or the relative position between the initial local sound source and the initial probe region is not provided. A device for measuring physical properties.   2. A means for scanning the position of the playback local sound source or the playback probe region or the relative position of the playback local sound source and the playback probe region is not provided. A device for measuring the physical properties of samples.   The apparatus for measuring physical properties of a sample according to claim 1, wherein the acoustically induced change in the initial or playback probe light emission is proportional to an out-of-plane acoustic displacement.   The apparatus for measuring physical property properties of a sample according to claim 1, wherein the acoustically induced change in the initial or playback probe light emission is proportional to the out-of-plane acoustic particle velocity. .   2. A physical property of a sample according to claim 1, wherein the acoustically induced change in the initial or playback probe light emission is proportional to a component of an acoustic distortion tensor in the sample. Equipment for.   The acoustically induced change in the initial or playback probe light emission is an out-of-plane component of acoustic displacement of the sample surface or interface, an in-plane component of acoustic displacement of the sample surface or interface, the sample The in-plane component of the acoustic particle velocity of the surface or interface of the sample, the in-plane component of the acoustic particle velocity of the surface or interface of the sample, and the six components of the acoustic distortion tensor are proportional to any combination. An apparatus for measuring physical properties of a sample according to item 1.   The acoustically induced change in the initial or playback probe light emission is obtained from a difference between two measurements of the initial or playback acoustic disturbance measured at short time intervals. An apparatus for measuring physical properties of a sample according to item 1.   2. The physical property of a sample according to claim 1, wherein said means for modulating the amplitude, phase or polarization state of said playback light pump radiation as a function of said playback delay time is not present. Equipment for. The means for processing the playback signal group has a Fourier transform A (ω) of the form A (ω) = G (ω) ^* / H (ω) D (ω) ^*
(Where ω is the angular frequency of the sound wave, H (ω) is the Fourier transform of the impulse response function h (t) of light generation of the playback local sound source (where t is the delay time), and D (ω) is Fourier transform of light detection impulse response function d (t) of the initial acoustic disturbance in the initial probe region, G (ω) is the acoustically induced change as a function of the initial delay time during the initial probe light emission. Convolution of the playback signal group with part or all of the time domain of the processing function a (t) defined by a Fourier transform of the initial signal group g (t) proportional to 20. The apparatus for measuring physical properties of a sample according to claim 1 or 19, characterized by comprising:   21. Sample physical as claimed in claim 1, 19 or 20, characterized in that the processing function a (t) is filtered before being applied in the means for processing the playback signals. A device for measuring properties.   21. A digital processing with a certain number of bits precision before the processing function a (t) is applied in the means for processing the playback signal group. A device for measuring the physical properties of samples.   2. An apparatus for measuring physical properties of a sample according to claim 1, wherein said means for processing said playback signal group is not provided. The means for modulating the amplitude, phase or polarization state of the playback light pump radiation as a function of the playback delay time, the intensity of the playback light pump radiation as a function of the playback delay time, and formula

  Filter before the playback delay time function b (t) is applied in the means for modulating the amplitude, phase or polarization state of the playback optical pump radiation as a function of the playback delay time 25. An apparatus for measuring physical properties of a sample as claimed in claim 1, 23 or 24.   A constant bit before the function b (t) of the playback delay time is applied in the means for modulating the amplitude, phase or polarization state of the playback optical pump radiation as a function of the playback delay time 25. Apparatus for measuring physical properties of a sample according to claim 1, 23 or 24, characterized in that it is digitally processed with numerical accuracy. The impulse response function d (t) of the light detection of the initial acoustic disturbance in the initial probe region is for a sample having a free surface.

(Where Re is the real part, k is the wave number of the initial probe radiation, N is the complex refractive index of the sample with respect to the initial probe radiation, dN / dη is the appropriate photoelastic constant for the initial probe radiation, ν is the sound velocity in the longitudinal direction of the sample, θ (t) is a unit step function, i = √ (−1), and the impulse response function d (t) is propagated perpendicularly to the surface of the sample. , Wherein g (t) is a real part of the relative complex reflectivity change of the initial probe light radiation). The apparatus for measuring the physical characteristic of the sample as described in any one of 26. The impulse response function d (t) of the light detection of the initial acoustic disturbance in the initial probe region is for a sample having a free surface.

(Where Im is the imaginary part, k is the wave number of the initial probe radiation, N is the complex refractive index of the sample relative to the initial probe radiation, dN / dη is the appropriate photoelastic constant for the initial probe radiation, ν is the sound velocity in the longitudinal direction of the sample, θ (t) is a unit step function, i = √ (−1), and the impulse response function d (t) is propagated perpendicularly to the surface of the sample. 1, wherein g (t) is proportional to the phase change of the initial probe light radiation due to interaction with the sample). The apparatus for measuring the physical characteristic of the sample as described in any one of 26.   The impulse response function d (t) and its Fourier transform D (ω) of the light detection of the acoustic wave are d (t) = − 2 kδ (t) and D (ω) = − 2k ( Where k is the wave number of the initial probe light radiation, δ (t) is the Dirac delta function, and the impulse response function d (t) is defined in relation to the outward surface displacement of the sample surface, 27. The sample according to claim 1, wherein g (t) is approximated by an optical phase change of the initial probe light radiation due to an interaction with the sample. A device for measuring the physical properties of the.   The impulse response function d (t) of light detection of the initial acoustic disturbance in the initial probe region and its Fourier transform D (ω) are d (t) = − 2 kδ (t ) And D (ω) = − 2k (where k is the wave number of the initial probe light radiation, i = √ (−1), δ (t) is a Dirac delta function, and the impulse response function d (T) is defined in relation to the outward surface particle velocity of the sample, and g (t) is proportional to the optical phase difference between the two initial probe pulses of the light emission, where the initial probe pulse is Both are incident on the sample after the initial pump pulse and the time interval between the two initial probe pulses is selected to be sufficiently small). Any one of 26 Apparatus for measuring the physical properties of the sample.   2. The sample according to claim 1, wherein the playback pump pulse of the light emission is a pulse having a shape whose amplitude, phase or polarization state is adjusted to facilitate generation of the playback local acoustic wave source. A device for measuring physical properties.   The appropriate theoretical model of the propagation and scattering of the initial and playback acoustic disturbances in the sample and the acoustic light generation and detection process in the sample is known for the size, elastic constant and density distribution of a portion of the sample. Measure other unknown physical properties such as dimensions, elastic constants or density of other parts of the sample based on the amount of the The comparison changes the trial value of the size, elastic constant or density of the other part of the sample to minimize the difference between the predicted value of the appropriate theoretical model and the processed playback signal group. The apparatus for measuring physical properties of a sample according to claim 1, comprising:   Physical properties to be measured are (1) internal stress in the sample portion, (2) adhesion between the sample portions, (3) ion implantation amount into the sample portion, (4) Effects of sample exposure to moisture; (5) roughness of the sample surface or interface; (6) size and location of acoustic scattering sources in the sample; (7) particle size distribution of the sample; ) Porosity of the sample; (9) defect distribution in the sample; (10) dislocations in the sample; (11) thickness or distribution of adsorbed material on the sample; (12) microstructure of the sample; 32. The apparatus for measuring physical properties of a sample according to claim 31, wherein the apparatus is selected from:   The apparatus for measuring a physical property of a sample according to claim 1, wherein the sample is made of a material that is absorptive of light radiation of the initial pump pulse or the series of playback pump pulses.   The apparatus for measuring physical properties of a sample according to claim 1, wherein the sample is composed of a transparent part and an absorptive part for the light emission of the initial pump pulse or the series of playback pump pulses. .   The sample of claim 1, wherein the sample comprises one or more of a transparent portion, a translucent portion, and an absorptive portion with respect to the light emission of the initial pump pulse or the series of playback pump pulses. A device for measuring the physical properties of the.   The sample of claim 1, wherein the sample comprises one or more of a transparent portion, a translucent portion, and an absorptive portion with respect to the light emission of the initial probe pulse or the series of playback probe pulses. A device for measuring the physical properties of the.   The apparatus for measuring physical properties of a sample according to claim 1, wherein the sample is composed of one or more of a solid, liquid or gaseous portion.   The sample is composed of at least one region including a disordered acoustic cavity, and the cavity enables effective focusing of the initial acoustic disturbance or playback acoustic disturbance in a region of the sample. A device for measuring the physical properties of samples.   The apparatus for measuring physical properties of a sample according to claim 1, wherein an aerial image representing the initial acoustic disturbance or playback acoustic disturbance is obtained as a function of the initial or playback delay time for the region of the sample. .   The sample physics of claim 1, wherein one or more of the initial or playback pump pulses or probe pulses are focused on the sample using a near-field light technique such as a tapered fiber to increase the spatial resolution of the measurement. For measuring typical characteristics.   The apparatus for measuring a physical property of a sample according to claim 1, wherein one or more of the initial or playback probe pulses correspond to one of the initial or playback pump pulses, respectively.   The apparatus for measuring physical properties of a sample according to claim 1, wherein a number of said initial or playback probe pulses are collected and integrated to produce a continuous probe beam.

Images