DE102011087460B3 - Method for checking defects on crystalline body i.e. sapphire crystal, involves irradiating sapphire crystal with light, and determining path of refracted light for reconstruction of distribution of defects on sapphire crystal - Google Patents

Abstract

The method involves irradiating a crystalline body i.e. sapphire crystal (4) with a light (L) from an LED source (14), and receiving the light reflected from the body by a camera (12), where the light reflected from the body are received from different positions for obtaining images. The distribution of defects (6) in the sapphire crystal is reconstructed from the images. The light is refracted on a complex surface (8) of the sapphire crystal, and the path of the refracted light (22) is determined for reconstruction of the distribution of defects on the sapphire crystal. An independent claim is also included for a device for checking defects on a crystalline body.

Description

The invention relates to a method for checking an optically transparent body for defects with the features of the preamble of claim 1, in particular for testing a body of a crystalline material, which is preferably used in the semiconductor industry, for example a sapphire crystal Apparatus for carrying out such a method.

Such a method and such a device are known from DE 10 2009 043 001 A1 refer to. In the known method, an optically transparent body is checked for volume defects such as gas inclusions with the aid of a scattered light measuring device. In this case, use is made of the fact that the intensity of a scattered radiation generated by the defect, which is detected in an azumite angle range, changes periodically. The intensity change is used to deduce the size and shape of the defect.

In many technical areas, efforts are made to be able to detect defects in a body early in a nondestructive testing. In particular, for products at the beginning of a production chain, from which after a large number of further manufacturing steps an end product is manufactured, early detection of defects is of particular importance.

For the production of semiconductor devices, the actual semiconductor material, such as silicon, as well as carrier substrates, such as sapphire, is obtained from crystalline blocks. Early detection of micro-defects in particular, such as micro-cracks, inclusions of air or vacuum or other crystal defects is sought, since such micro-defects can lead to failure of the finished mechanical components.

There are already known optical investigation methods for detecting defects in silicon blocks but also in individual silicon wafers. In which the respective silicon object is irradiated by means of an infrared light source. Optionally, in the transmitted light method, the light transmitted through the object or the light scattered at a defect is detected by a camera in the manner of a dark field image.

In the semiconductor or optical industry, sapphire crystals, in particular sapphire single crystals, are produced, from which in turn individual sapphire wafers, sapphire lenses or similar objects are produced. Sapphire is also used as an insulating carrier substrate in semiconductor technology, to which then the semiconductor material, such as silicon (silicon-on-sapphire) is applied. Such sapphire single crystals form approximately a cylindrical block, which usually has several 10 cm in diameter and several 10 cm in height. Such a sapphire single crystal typically has a microscopically smooth, but otherwise complex, irregular surface with undulations and bends.

An optical inspection of such a sapphire single crystal, generally an optically transparent body with an undefined surface geometry, on any defects in the volume is difficult due to the irregular surface topography. The reason for this is to be seen in the refractive index difference between the air and the body to be examined, so that the light undergoes a refraction when entering and exiting the body. In an optical inspection, in particular in an optical tomography, in which an object is scanned from different directions and from the resulting images, a two-dimensional or three-dimensional image is reconstructed, however, is the knowledge of the course of the light beam or light beam used for a respective recording of essential for the reconstruction.

From the essay "Tomographic Reconstruction of Transparent Objects", Borislav Trifonov et al., Eurographics Symposium on Rendering (2006), a tomography method for transparent objects is described. To address the problem of an unknown, undefined surface topography of the body to be examined is proposed herein, to put the body in a glass cylinder with a defined surface geometry, wherein the glass cylinder is filled with an oil having an at least substantially identical refractive index as the has examining body, so that no refraction occurs in the transition from oil to the body. A refraction occurring in the transition from air to glass cylinder or oil is considered due to the known surface geometry of the glass cylinder.

Proceeding from this, the object of the invention is to provide a simplified method for the optical inspection of optically transparent bodies.

The object is achieved according to the invention by a method having the features of claim 1 and with a device having the features of claim 12.

Thereafter, it is provided that the body to be examined is irradiated by means of a light source. The body is in particular one crystalline material, preferably for use in the semiconductor industry and especially a sapphire single crystal or other semiconductor crystals. The light emerging from the body at a decoupling point is picked up by a camera. In terms of tomography, a large number of such recordings is recorded at different recording positions. By this is generally understood that the body to be examined is successively scanned by the optical arrangement consisting of camera and light source in the manner of a scanning operation. Thus, there is a relative movement between the optical components camera, light source and body to be examined. In this case, a fixed relative position between the camera and the light source can be provided, but preferably a relative movement between the camera and the light source is provided, at least in some areas.

From the large number of such images, a particular three-dimensional image is subsequently reconstructed in the context of a tomography method. If there are defects in the body to be examined, they can be recognized in the reconstruction with regard to their position, orientation and size.

According to the method, it is provided that the light detected by the camera or the light irradiated into the body by the light source is refracted at the complex, that is geometrically unknown, surface topography of the body. This means that the body just is not immersed in an oil bath, but comes directly into contact with the surrounding medium, in particular air. In order to enable the reconstruction, it is further provided that the course of the light beam in the body is determined with the aid of an additional measuring arrangement and used as the basis for the reconstruction. In this case, the light path in the body is preferably determined by a possible defect to the outcoupling point in the direction of the camera.

The key advantage here is that immersion in a liquid is not required. As a result, on the one hand a simplified measurement setup is possible. In addition, the use of an oil often provided with highly toxic substances and the associated with the immersion contamination of the body to be examined is avoided.

Preferably, the respective course of the corresponding light beam, which is recorded by the camera, is determined for each of the recordings. With only slight changes in the recording position, ie the relative position between the optical components and the body, it can be provided that the measurement is not made for each individual shot, but the course of the light beam calculatory, for example by approximation, interpolation or simply by Acquisition of the determined data from an immediately adjacent recording position is determined.

For the quality of the reconstruction, it is advantageous if individual volumes of the body to be examined are covered several times from different acquisition positions. The higher this coverage density, the better the reconstruction quality. In order to achieve a sufficient quality of reconstruction, several thousands of images are provided in each case for the examination of an object. In a typical size of a sapphire single crystal described above with a diameter of about 20 cm and a height of about 30 cm, for example, several 10,000, in particular more than 100,000 shots are made.

As a camera in this case, for example, a line and preferably a matrix camera is used, which thus has individual arranged in line or matrix shape measuring points (pixels). The surface area covered by the camera on the body is preferably in the range of a few mm ² to a cm ² .

The light source selected is a light source with a wavelength of light suitable for the illumination of the object, so that the body is transparent for the selected wavelength. A sapphire crystal is transparent to light in the visible range. As a light source, in particular an LED light source, preferably a surface radiator of a plurality of LEDs arranged in a matrix-like manner is used. Generally, monochromatic light is preferably used in the visible region, for example in the red wavelength region. For example, when examining silicon bodies, an IR light source is used because silicon is transparent to IR light.

By optically transparent body is generally understood a body which is transparent to light of a suitable wavelength. There is no restriction on the visible light. In particular, optically transparent is understood to mean that the body is transparent to light in the visible range as well as in the IR range or also in the UV range. In principle, this principle can also be applied to other beams, such as X-rays. However, there is no problem here of refraction on an unknown surface since, for example in medical examination by means of X-ray tomography, there are no or negligible refractive index differences for the X-ray radiation.

According to a first preferred embodiment for determining the course of the light beam, a surface normal of the body is determined at the point at which the light is refracted, that is to say in particular at the outcoupling point by the measuring device. This embodiment is based on the consideration that under the assumption of a constant refractive index within the body, the propagation of light within the body is rectilinear and insofar can be assumed at the interface where the light is refracted, in principle geometric optics on the basis of the law of refraction. For this purpose, the determination of the orientation of the surface, ie in particular their normal is required in order to determine the angle at which the light beam impinges on the respective surface element.

Conveniently, it is provided that the surface normal is measured at each recording position together with the recording at this particular recording position, so that the course of the light beam is known for each shot.

In order to measure the surface normals, fundamentally different methods are available, such as (laser) triangulation or so-called deflectometry, etc.

In the sense of the simplest possible measuring setup, it is provided in an expedient embodiment that a measuring light beam, in particular a laser light beam, is directed onto the body coaxially to a camera viewing direction which is defined by its optical axis. The measuring light beam is therefore directed directly to the coupling-out point, ie to the point at which the light detected by the camera for a respective recording emerges. Furthermore, a further camera is provided by means of which a light component of the coaxially irradiated light reflected on the surface of the body is detected. Therefore, in addition to the camera viewing direction, the direction of reflection is also determined, and the surface normal is determined from these two known measured variables taking into account the reflection law, according to which the angle of incidence equals the angle of reflection.

To determine the reflected light, in particular a measuring arrangement with a ground glass is provided, on which the reflected beam impinges, wherein the point of impact is detected by means of the further camera.

In a preferred embodiment, moreover, it is provided that the ground glass is provided with individual holes or is transparent to a third camera, which is directed to the decoupling point. The two other cameras are preferably designed as matrix cameras. From the positions of the laser points on the one hand at the decoupling point and on the other hand at the point of impact on the screen, the spatial coordinates for these two points are determined and then determined from the normal vector (surface normal) at the entry point of the laser beam. Thus, the local surface topography is known.

Alternatively or additionally, it is provided that, in addition to the point of entry of the coaxially coupled measuring light beam, the exit point of this measuring light beam is also detected and a simple straight line between these then measured points is determined to determine the course of the light beam.

Both methods are suitable both for isotropic and anisotropic bodies, the latter having different refractive indices and thus birefringence for different beam directions or different polarizations. The only requirement is that the refractive index is constant in the defined beam direction. This is regularly fulfilled with crystals. In principle, it is also possible to measure amorphous materials, such as glasses.

In a preferred development, the data obtained on the basis of this surface measurement are further processed in such a way that the surface topography of the body is also shown in the reconstructed image. At the same time, this makes it possible to measure the entire body. With the help of appropriate image processing programs can therefore be immediately recognized which areas are error-free, how large they are, etc. For further use of the body can therefore be determined exactly those places in the body, which will continue to be used. As a result, the body as a whole can be optimally utilized for the further production process.

From the determined reconstruction of the body, error-free regions are identified in a preferred development by means of an automated evaluation and an optimized decomposition of the body, in particular by drilling out of cylinders with a defined diameter, depending on the geometry of secondary products, such as wafer wafers, etc., is calculated. The automatic evaluation preferably takes place on the basis of the reconstructed 3D image by means of an image evaluation. By this measure, an optimized material utilization is achieved.

With regard to the positioning of the optical arrangement, that is, the light source and the camera, the light source is preferably laterally and in particular perpendicular to the optical axis of the camera in Arranged sense of a dark field illumination. The light scattered at the defects is detected by the camera in this case.

Alternatively or in addition to this, the light source is arranged opposite the camera in the sense of transmitted light illumination. In this case, the defect in the camera does not appear as lighter but as a dark spot due to the light scattered on it, in contrast to the dark field illumination.

In this case, the body is preferably irradiated with a wide-spread light bundle, in particular over the entire surface, and thus preferably homogeneously illuminated over its entire cross-sectional area. For this purpose, a suitably designed light source is optionally used, using appropriate apertures, lenses, etc.

With regard to the scanning or scanning process, it is expediently provided that the body is arranged on a slide and is rotated about an axis of rotation. It is further provided that the camera for scanning the body in the body longitudinal direction parallel to the axis of rotation, hereinafter referred to as Y-direction, is moved. Preferably, it is additionally provided that the camera is moved perpendicular to the axis of rotation, so that it is ensured that not only radial, passing through the body center rays are detected, which would then only have a low information content.

An embodiment of the invention will be explained in more detail below with reference to FIGS. These show in a schematic representation:

1 a view of a device for checking an optically transparent body on defects,

2 a side view of such a device similar to in 1 illustrated device, but without the in 1 illustrated measuring arrangement for measuring the course of a respective light beam in the body to be examined,

3 an exemplary reconstructed image of a sapphire crystal examined with the device.

In the figures, like-acting parts are provided with the same reference numerals.

The in the 1 and 2 The device shown comprises a slide 2 , in the embodiment, a sapphire crystal 4 is arranged, on flaws 6 to be examined in its volume. The sapphire crystal 4 roughly has an approximately cylindrical configuration, which is due to the crystal growth in the production of the single crystal. The sapphire crystal 4 has a complex, irregular surface 8th on. The slide 2 is about a rotation axis 10 rotatably mounted, allowing the sapphire crystal 4 is rotatably arranged.

The device is designed to perform an optical tomography, that is, to reconstruct a preferably three-dimensional image from a plurality of individual images. The individual images are each taken with the aid of an optical recording device, which is a recording camera 12 as well as a light source 14 includes. In the exemplary embodiment, an arrangement is shown with a dark field illumination, on the basis of which the method is also explained further below. Accordingly, the light source 14 laterally to an optical axis 16 arranged the recording camera and illuminates the crystal 4 from above.

The light source generates a planar irradiation field, thus illuminating the crystal 4 from above over the entire surface. For this purpose, suitable lenses and diaphragms are provided in a manner not shown here. The crystal 4 is therefore from the light source 14 with light L preferably full surface and homogeneously illuminated.

The recording camera 12 is from a camera carrier 20 kept movable. One is the recording camera 12 parallel to the axis of rotation 10 movable in a Y-direction, so that successively individual height levels of the crystal 4 can be sampled. In addition, the recording camera 12 perpendicular to the axis of rotation 10 movable in an X-direction.

At the optical inspection of the crystal 4 on possible defects 6 in its volume, a large number of exposures are taken in different recording situations at constant dark-field illumination from above. This is understood to mean that the recording camera 12 , in particular its optical axis 16 , for each shot on a different surface area of the crystal 4 is directed.

Furthermore, a control and evaluation unit 18 provided in which all information is processed, in particular the image data captured by the recording camera, and further performs the actual reconstruction. The reconstructed image can be output or stored on a display (not shown) or in any other way. About this unit 18 The control of the individual components is also carried out.

While capturing the multitude of shots, the crystal becomes 4 either continuously or gradually around the axis of rotation 10 turned. The recording camera 12 initially remains at a predetermined height position on the Y axis. Within this height position is the recording camera 12 shifted in the X direction, so that preferably for each changed camera position along the X direction at least one full turn of the crystal 4 takes place, with a large number of recordings per full revolution. This is then successively performed layer by layer for all height positions along the Y-axis, ie the recording camera is moved to a further height positions.

Depending on the desired quality of the reconstructed image, the number of recording positions can be changed. The goodness can additionally by a further freedom of movement of the recording camera 12 be improved, for example, a tilting movement within the XZ plane.

From this large number of taken pictures, for example 150,000, a three-dimensional picture is subsequently reconstructed with the aid of a suitable algorithm. The basic procedure for this and the manner of the required principles in the design of such reconstruction algorithms is known, for example, from X-ray computed tomography. In order to obtain meaningful images with such a reconstruction, however, it is necessary that the course of a scattered and the recording camera 12 detected light beam 22 within the crystal 4 is known to make a statement about the location of the defect 6 to meet. The light beam 22 occurs at a decoupling point 23 from the crystal 4 out.

Due to the undefined, complex surface 8th and the different refractive indices between crystal 4 and the surrounding air is the course of the light beam 22 within the crystal 4 but not known because the light beam 22 on the surface 8th due to the complex surface structure is broken in a way not defined for the viewer. Therefore, initially alone from the received light beam 22 without knowledge of the course of the light beam 22 within the crystal 4 no statement about the position of the defect 6 to be hit.

To determine the course of the light beam 22 in the exemplary embodiment is a measuring arrangement 24 provided in 1 is shown in addition. This measuring arrangement 24 includes at least one, in the embodiment, two more cameras 26a , b, one in the embodiment with holes 28 provided ground glass 30 , in the embodiment as a point laser 32 formed further light source for generating a measuring light beam 34 and an optical deflection element, which in the exemplary embodiment as a beam splitter 36 is trained. The measuring light beam 34 becomes coaxial with the optical axis 16 the recording camera 12 coupled and meets the decoupling point assigned to a respective recording 23 on, out of which the light beam 22 exit.

The measuring light beam 34 occurs in the embodiment, first through the ground glass 30 through, hits the surface 8th at the decoupling point 23 and is reflected there at least partially according to the law of reflection and hits the screen 30 in a meeting point 38 on. The position of the decoupling point 23 is with the help of the other camera 26a captured by a transparent area (hole 28 ) on the crystal 4 is directed. For the determination of the position of the decoupling point 23 in addition, the recording camera in the room 12 be used. The position of the point of impact 38 is through the camera 26b detected. In order to determine their position in space, the knowledge of the position of the ground glass is additionally added 30 used. Under knowledge of the impact point 38 and the decoupling point 23 , which are recognizable in respective pictures, which with the further cameras 26a , b, can be - with still known position of the two cameras 26a , b a reflection direction of the reflected laser beam 34 determine. Taking into account the known irradiation direction in the direction of the optical axis 16 becomes a surface normal on the basis of the normal refraction law 40 in the area of the decoupling point 23 certainly. Furthermore, based on the Snellius law of refraction at the refractive index known for the crystal, the course of the light beam 22 in the crystal 4 determined. This determination of the course of the light beam 22 is preferably carried out for each recording in the different recording positions.

Sapphire is an anisotropic crystal in which the refractive indices differ only slightly in the different propagation directions. In a simplified method, an (averaged) refractive index is therefore assumed for such anisotropic bodies.

In 1 is still a second alternative way to determine the course of the light beam 22 shown. This can also be used, for example, when the crystal 4 has a strong anisotropy. And it is provided here that - identical to the first case - the measuring light beam 34 at the decoupling point 23 incident. Since only a part is reflected, this also defines an entry point for a transmitted portion of the measuring light beam 34 which is at an opposite exit point 42 exits again. The position of this exit point 42 on the surface 8th of the crystal 4 is done with the help of other cameras 26c , d recorded. The course of the light beam 22 then results from a simple straight connection between the exit point 42 and the decoupling point 23 like this in the 1 is shown.

The result of a tomographic reconstruction performed with the help of such a device is in 3 as a reconstruction image 44 shown. The structure shown on the one hand gives the surface 8th of the crystal 4 as a surface topography 8th' again. The individual approximately circular lines give individual contour lines of the surface topography 8th' again.

The one in the crystal 4 possibly contained defects 6 are recognizable inside. In the exemplary embodiment form the flaws 6 an approximately hose-like structure extending in the Y-direction.

From this reconstructed 3D representation, therefore, defect-free areas in the sapphire crystal can be achieved with high accuracy 4 in terms of their location (size and orientation) in the volume of the crystal 4 , With the help of such a 3D image, therefore, areas of the crystal 4 defined and determined from which elements for further processing can be worked out. In particular, the optimum cutting or drilling positions are calculated by automated evaluation in order to optimize the crystal with regard to secondary products (wafer products) such as wafer wafers, optical elements etc., ie to disassemble them with as little waste material as possible. The calculation of the optimal cutting positions is carried out depending on the geometry of the secondary products.

At the defects 6 These are, in particular, gas inclusions such as air or so-called voids, in which vacuum areas are thus included. Such gas inclusions can be seen very well in a dark field illumination.

Basically, this measuring principle can be determined by determining the course of the light beam 22 also use in transmitted-light lighting situations, ie where the light source 14 not laterally but opposite the camera 12 is arranged. By scattering effects on such defects 6 appear flaws 6 Unlike dark field illumination, as dark spots, while appearing as dark spots in dark field illumination.

Claims (12)

Method for checking an optically transparent body ( 4 ) on defects ( 6 ), in particular a crystalline body, in which - by means of a light source ( 14 ) Light (L) in the body ( 4 ), - one out of the body ( 4 ) at a decoupling point ( 23 ) exiting light (L) from a camera ( 12 ), - a plurality of such recordings is recorded at respectively different recording positions and - possibly in the body ( 4 ) located defects ( 6 ) are reconstructed from the plurality of recordings, characterized in that - the light (L) on a complex surface ( 8th ) of the body ( 4 ) is broken and - the course of the broken light beam ( 22 ) in the body and is used as the basis for the reconstruction. A method according to claim 1, characterized in that for determining the course of the refracted light beam ( 22 ) a surface normal ( 40 ) of the body ( 4 ) is measured at the point where the light is refracted. Method according to claim 2, characterized in that the surface normal ( 40 ) at each recording position and measured together with the recording at that recording position. Method according to claim 2 or 3, characterized in that for determining the respective surface normal ( 40 ) a measuring light beam ( 34 ) coaxial with an optical axis ( 16 ) the camera ( 12 ) on the body ( 4 ), one on the surface ( 8th ) reflected light component of the measuring light beam ( 34 ) and from this the surface normal ( 40 ) is determined. Method according to one of claims 2 to 4, characterized in that a measuring light beam ( 34 ) coaxial with an optical axis ( 16 ) the camera ( 12 ) at an entry point ( 23 ) in the body ( 4 ), an exit point ( 42 ) of the measurement light beam ( 34 ) and from the knowledge of the entry point ( 23 ) and the exit point ( 42 ) the course of the broken light beam ( 22 ) is derived. Method according to one of the preceding claims, characterized in that a three-dimensional image ( 44 ) of the detected defects ( 6 ) is reconstructed. Method according to one of the preceding claims, characterized in that the measured surface topography ( 8th' ) of the body ( 4 ) is pictured. Method according to one of the preceding claims, characterized in that from the determined reconstruction of the body ( 4 ) are identified by an automated evaluation error-free areas and an optimized decomposition of the body is calculated depending on the geometry of derived products. Method according to one of the preceding claims, characterized in that the body ( 4 ) laterally to the optical axis ( 16 ) the camera ( 12 ) is illuminated in the manner of a dark field illumination or opposite in the manner of a transmitted light illumination. Method according to one of the preceding claims, characterized in that body ( 4 ) on a microscope slide ( 2 ) is arranged and about a rotation axis ( 10 ) is rotated. A method according to claim 10, characterized in that the camera in an x-direction perpendicular to the axis of rotation ( 10 ) and in a y-direction parallel to the axis of rotation ( 10 ). Device in particular for carrying out the method according to one of the preceding claims, comprising a slide ( 2 ) for a body to be examined ( 4 ), a light source ( 14 ) for illuminating the body ( 4 ), a camera ( 12 ) for performing a plurality of recordings at different recording positions, an evaluation unit ( 18 ), in which an image is reconstructed from the plurality of recordings, characterized in that a measuring arrangement ( 24 ) is provided, via which the course of a respective recording associated light beam ( 22 ) in the body ( 4 ), which is used as the basis for the reconstruction.

Images