Differential interference inspection device and differential interference inspection method

Abstract

To make it possible to satisfactorily visualize unevenness of an object being inspected through an inspection method in which a differential interference method is used. Specifically, a differential interference inspection device 1 comprises: a prism 10 that separates and branches two polarized lights La, Lb that are orthogonal to each other by a shear amount H; a camera 20 that captures an image G obtained by the polarized lights La, Lb hitting an object Wa being inspected through the prism 10; and a processing unit 30 that inspects the object Wa being inspected on the basis of the image G. The prism 10 can be moved so that a prism position P that corresponds to the shear amount H changes. The processing unit 30 sets a reference value R0 of the prism position P on the basis of the relationship between the absolute value Q of the prism position P and the brightness E of the image G. The prism 10 moves on the basis of a relative value R of the prism position P starting from the reference value R0 of the prism position P.

Description

Differential interference inspection apparatus and differential interference inspection method

[0008] ,

[0007] ,

[0001] The present disclosure relates to a differential interference inspection apparatus and a differential interference inspection method.

[0002] For example, as shown in Patent Document 1, a technique for inspecting a test object using a differential interference microscope is known.

[0003] Japanese Patent Laid-Open No. 5-256795

[0004] A differential interference microscope is provided with a prism and a camera. When one polarized light passes through the prism, it is converted into two polarized lights orthogonal to each other. Further, the prism branches the two polarized lights orthogonal to each other by a shear amount. The two branched polarized lights are incident on different two points in the test object. The two incident polarized lights are reflected or transmitted by the test object, and when they pass through the prism again, they become one polarized light and merge into the same optical path to interfere. The image obtained at the time of interference is captured by the camera. <​​​​​​​​​Traditionally, the absolute value of the prism position was set according to the specifications of the object being inspected. Ideally, there should be a correspondence between the absolute value of the prism position and how the surface irregularities appear in the image captured by the camera. In other words, if the absolute value of the prism position is set appropriately, the irregularities should be clearly visible.

[0009] However, even after setting the appropriate absolute value for the prism position (which should make surface irregularities clearly visible), individual differences such as manufacturing tolerances in differential interference microscopes sometimes made it difficult to see surface irregularities in the image captured by the camera.

[0010] The purpose of this disclosure is to make the irregularities of the object being inspected more visible in inspection methods using differential interferometry.

[0011] The differential interference contrast inspection apparatus according to this disclosure comprises a prism that splits two mutually orthogonal polarized rays by a shear amount, a camera that captures an image obtained when the polarized rays strike an object to be inspected through the prism, and a processing unit that inspects the object to be inspected based on the image, wherein the prism is movable such that the prism position corresponding to the shear amount changes, the processing unit sets a reference value for the prism position based on the relationship between the absolute value of the prism position and the brightness of the image, and the prism moves based on the relative value of the prism position with respect to the reference value of the prism position.

[0012] The prism is moved based on the relative value of its position, rather than its absolute value. This allows for better visualization of surface irregularities in the object being inspected, regardless of individual differences in the differential interferometry inspection device.

[0013] In one embodiment, the processing unit sets the absolute value of the prism position at which the brightness of the image is at its minimum as the reference value of the prism position.

[0014] The reference value for the prism position can be easily set.

[0015] In one embodiment, when a plurality of absolute values ​​for which the brightness is the minimum measured value are measured, the processing unit sets the absolute value closest to the lower limit for which the brightness is the minimum measured value as the lower limit minimum absolute value, sets the absolute value closest to the upper limit for which the brightness is the minimum measured value as the upper limit minimum absolute value, and calculates the absolute value corresponding to the minimum value based on at least the lower limit minimum absolute value and the upper limit minimum absolute value.

[0016] Even if the absolute value of the prism position corresponding to the true minimum brightness has not been measured, the absolute value of the prism position corresponding to the true minimum brightness can be calculated.

[0017] In one embodiment, the processing unit sets the absolute value at which the brightness is maximum closer to the lower limit than the minimum absolute value on the lower limit as the maximum absolute value on the lower limit, sets the brightness corresponding to the maximum absolute value on the lower limit as the first maximum measured value, sets the absolute value at which the brightness is maximum closer to the upper limit than the minimum absolute value on the upper limit as the maximum absolute value on the upper limit, sets the brightness corresponding to the maximum absolute value on the upper limit as the second maximum measured value, and sets the maximum value of the brightness in the first approximation curve generated based on at least the maximum absolute value on the lower limit and the first maximum measured value as the first maximum theoretical value. The following steps are performed: the absolute value corresponding to the first maximum theoretical value is set as the lower limit theoretical absolute value; the maximum value of brightness in the second approximation curve generated based on at least the upper limit maximum absolute value and the second maximum measured value is set as the second maximum theoretical value; the absolute value corresponding to the second maximum theoretical value is set as the upper limit theoretical absolute value; a third approximation curve is generated based on at least the first maximum theoretical value and the lower limit theoretical absolute value, and the second maximum theoretical value and the upper limit theoretical absolute value; and the absolute value corresponding to the minimum value of the third approximation curve is calculated as the absolute value corresponding to the minimum value.

[0018] The absolute value of the prism position corresponding to the true minimum brightness can be calculated with high accuracy.

[0019] The differential interference contrast inspection method according to this disclosure is a differential interference contrast inspection method that inspects an object to be inspected based on an image obtained when two mutually orthogonal polarized lights strike the object to be inspected through a prism that separates them by a shear amount, wherein the prism is movable such that the prism position corresponding to the shear amount changes, a reference value for the prism position is set based on the relationship between the absolute value of the prism position and the brightness of the image, and the prism is moved based on the relative value of the prism position with respect to the reference value of the prism position.

[0020] According to this disclosure, in an inspection method using differential interferometry, it is possible to make the irregularities of the object being inspected more visible.

[0021] Figure 1 schematically shows a differential interference contrast inspection apparatus. Figure 2 shows the first shear amount and first height difference when two polarized rays are incident on the substrate surface when the absolute value of the prism position is the first value. Figure 3 shows the second shear amount and second height difference when two polarized rays are incident on the substrate surface when the absolute value of the prism position is the second value. Figure 4 illustrates the concept of the relative value of the prism position. Figure 5 shows the method for calculating the absolute value corresponding to the minimum brightness value.

[0022] Embodiments of the present disclosure will be described in detail below with reference to the drawings. The following description of preferred embodiments is illustrative in nature and is not intended to limit the present disclosure, its applications, or its uses in any way.

[0023] (Differential Interferometry Inspection Apparatus) Figure 1 schematically shows the differential interferometry inspection apparatus 1. The differential interferometry inspection apparatus 1 consists of a differential interference microscope. The differential interferometry inspection apparatus 1 inspects the surface Wa of a substrate W, which is the object to be inspected. The substrate W is, for example, a semiconductor substrate. Defects N are formed on the surface Wa of the substrate W. The differential interferometry inspection apparatus 1 may detect the defects N.

[0024] Examples of defects N include foreign matter, cracks (scratches, fissures, tears, fissures), poor edge rinsing, abnormal through-holes, uneven coating, peeling, scratches, abnormal bonding pads, uneven color, abnormal patterns, stains, discoloration, deformation, etc. In this example, cracks are used as an example of defect N.

[0025] The differential interference contrast inspection apparatus 1 comprises a prism 10, a camera 20, a computer 30 as a processing unit, and an actuator 40 as a prism movement mechanism.

[0026] Light emitted from a light source (not shown) is converted into a single polarization L as it passes through a polarizer (not shown). This single polarization L is then converted into two mutually orthogonal polarizations La and Lb as it passes through a prism 10. Furthermore, the prism 10 splits the two mutually orthogonal polarizations La and Lb, separating them by a shear amount H.

[0027] The prism 10 is made by bonding together two birefringent crystals with their crystal axes offset from each other. The Shear amount H is the amount of displacement in the transverse direction perpendicular to the optical path direction (direction of propagation) of the two polarized light La and Lb.

[0028] The two branched polarities La and Lb are incident on (strike) two different points on the surface Wa of the substrate W. The two incident (strike) polarities La and Lb are reflected by the surface Wa of the substrate W. When the two reflected polarities La and Lb pass through the prism 10 again, they become a single polarity L and merge into the same optical path and interfere.

[0029] The image G (differential interference pattern) obtained during interference is captured by camera 20. Camera 20 captures the image G obtained when two mutually orthogonal polarizations La and Lb strike the surface Wa of the substrate W through the prism 10. Camera 20 is, for example, a known digital camera. An analyzer may be placed between the prism 10 and camera 20.

[0030] The computer 30 inspects the surface Wa of the substrate W based on the image G captured by the camera 20. The computer 30 detects defects N on the surface Wa of the substrate W based on the image G captured by the camera 20. The computer 30 includes, for example, a processor and a memory that stores software for operating the processor. The computer 30 is connected to the camera 20 and the actuator 40 by wire or wireless. The computer 30 controls the actuator 40 based on input from the camera 20.

[0031] When there is a difference in elevation δ (gradient) between two points Pa and Pb where two polarized light signals La and Lb are incident (strike), the phases of the two polarized light signals La and Lb after reflection are shifted relative to each other. The interference intensity of the two polarized light signals La and Lb differs depending on the magnitude of the difference in elevation δ between the two points Pa and Pb where they are incident (strike). Therefore, if there is a distribution of elevation differences δ on the surface Wa of the substrate W, a contrast of light and dark is generated, and the unevenness is emphasized. In this way, the differential interference contrast analyzer 1 visualizes minute unevenness on the surface Wa of the substrate W.

[0032] The prism 10 is movable so that its prism position P changes. The prism position P is the position of the prism 10. The direction of movement of the prism 10 is transverse (perpendicular to the optical path directions of the two polarized light La and Lb). By changing its prism position P, the prism 10 changes the shear amount H between the two polarized light La and Lb. The prism position P corresponds to the shear amount H.

[0033] When the prism position P is changed, the shear amount H between the two polarized rays La and Lb changes, the height difference δ between the two points Pa and Pb on the surface Wa of the substrate W where the two polarized rays La and Lb are incident (strike) changes, the phase difference between the two polarized rays La and Lb after they are reflected from the surface Wa of the substrate W changes, and ultimately the appearance of the surface irregularities in the image G captured by the camera 20 changes.

[0034] The prism position P corresponds to the shear amount H. The prism position P corresponds to the height difference δ between two points Pa and Pb where two polarized light La and Lb are incident (strike). The prism position P corresponds to the phase difference between the two polarized light La and Lb. The prism position P corresponds to how the surface irregularities appear in the image G captured by the camera 20.

[0035] The actuator 40 moves the prism 10 laterally, changing the prism position P. The actuator 40 is, for example, an electromagnetic solenoid. In the actuator 40 (electromagnetic solenoid), the plunger 41 (movable iron core) moves laterally. The tip of the plunger 41 is connected to one lateral end of the prism 10. By moving the plunger 41 laterally, the actuator 40 moves the prism 10 laterally, changing the prism position P.

[0036] The (lateral) coordinate of the prism 10 in a fixed absolute coordinate system is called the absolute value Q of the prism position P. In Figure 1, the absolute value Q of the prism position P is shown as the distance from the origin O of the absolute coordinate system to one end of the prism 10 (laterally). The absolute value Q takes a positive value.

[0037] When the plunger 41 of the actuator 40 retracts (when the prism 10 moves toward the origin O), the absolute value Q of the prism position P becomes small. When the plunger 41 of the actuator 40 is fully retracted (when the prism 10 moves furthest toward the origin O), the absolute value Q of the prism position P becomes the lower limit Qa. When the plunger 41 of the actuator 40 protrudes (when the prism 10 moves toward the opposite side of the origin O), the absolute value Q of the prism position P becomes large. When the plunger 41 of the actuator 40 protrudes furthest (when the prism 10 moves furthest toward the opposite side of the origin O), the absolute value Q of the prism position P becomes the upper limit Qb.

[0038] The absolute value Q of the prism position P is measured by a sensor (for example, an axial position sensor built into the actuator 40).

[0039] (Relationship between the absolute value of the prism position and the appearance of unevenness) The relationship between the absolute value Q of the prism position P and the appearance of unevenness in the image G captured by the camera 20 will be described.

[0040] Fig. 2 shows the first shear amount H1 and the first height difference δ1 when two polarized lights La and Lb are incident on the surface Wa of the substrate W when the absolute value Q of the prism position P is the first value Q1. Fig. 3 shows the second shear amount H2 and the second height difference δ2 when two polarized lights La and Lb are incident on the surface Wa of the substrate W when the absolute value Q of the prism position P is the second value Q2.

[0041] The first value Q1 is smaller than the second value Q2. The second value Q2 is larger than the first value Q1. The first value Q1 is closer to the lower limit Qa. The second value Q2 is closer to the upper limit Qb.

[0042] When the absolute value Q of the prism position P is the first value Q1, the two polarized lights La and Lb passing through the prism 10 change from being oblique to straight (parallel) at an earlier stage (upstream side) compared to the case where the absolute value Q of the prism position P is the second value Q2. For this reason, the first shear amount H1 is smaller than the second shear amount H2 (the second shear amount H2 is larger than the first shear amount H1). Assuming that the gradient on the surface Wa of the substrate W is constant, the first height difference δ1 is smaller than the second height difference δ2 (the second height difference δ2 is larger than the first height difference δ1).

[0043] The prism position P is adjusted so that the unevenness can be clearly seen in the image G captured by the camera 20. Here, depending on the specifications of the surface Wa of the substrate W, the prism position P (for clear visibility of unevenness) is determined in advance. When the prism position P is determined according to the specifications of the surface Wa of the substrate W, the unevenness can be clearly seen in the image G captured by the camera 20.

[0044] Conventionally, the absolute value Q of the prism position P has been set according to the specifications of the surface Wa of the substrate W. Originally, the absolute value Q of the prism position P and the appearance of unevenness in the image G captured by the camera 20 should be in a corresponding relationship. That is, if the absolute value Q of the prism position P is set appropriately, the unevenness should be clearly visible.

[0045] For example, in this example, when the absolute value Q of the prism position P is the first value Q1 (FIG. 2), the unevenness in the image G (the first image G1) captured by the camera 20 is blurred and difficult to see (appears flat), while when the absolute value Q of the prism position P is the second value Q2 (FIG. 3), the unevenness in the image G (the second image G2) captured by the camera 20 is clearly visible (appears sharp). Note that there may also be a pattern opposite to the above.

[0046] However, even though an appropriate absolute value Q of the prism position P (where the unevenness should be clearly visible) is set, due to individual differences such as manufacturing errors of the differential interference inspection device 1, the unevenness in the image G captured by the camera 20 may become difficult to see.

[0047] In the present embodiment, by making the following improvements, in the differential interference inspection device 1 using the differential interference method, regardless of the individual differences of the differential interference inspection device 1, the unevenness on the surface Wa of the substrate W as the inspection object is made clearly visible.

[0048] (Relative value of prism position) FIG. 4 shows the concept of the relative value R of the prism position P. FIG. 4 is a graph, where the absolute value Q of the prism position P is shown on the horizontal axis, and the luminance E as the brightness of the image G captured by the camera 20 is shown on the vertical axis. On the horizontal axis, the absolute value Q becomes smaller as going to the left, and at the leftmost side, the absolute value Q becomes the lower limit Qa. On the horizontal axis, the absolute value Q becomes larger as going to the right, and at the rightmost side, the absolute value Q becomes the upper limit Qb.

[0049] The surface Wa of the calibration substrate W is imaged by the camera 20. The surface Wa of the calibration substrate W has no unevenness but has a pattern Wb. The image G captured by the camera 20 is divided into a plurality of pixels. Each pixel is given a numerical value of luminance E in the range of 0 to 255. The larger the numerical value of the luminance E, the brighter, and the smaller the numerical value of the luminance E, the darker. The vertical axis indicates the luminance E at a specific pixel in the image G captured by the camera 20. The absolute value Q and the luminance E correspond to each other.

[0050] The graph in Figure 4 is roughly a sixth-degree polynomial (with a positive coefficient for the highest order). As the absolute value Q moves from the lower limit Qa to the upper limit Qb, the luminance E takes on a maximum value near the lower limit Qa, a maximum value near the upper limit Qb, and a minimum value midway between these two maximum values. At the intermediate minimum, the luminance E reaches its minimum value Es. The numerical value of the minimum luminance value Es is basically zero. When the luminance E is at the minimum luminance value Es, the pattern Wb on the surface Wa of the calibration substrate W is not visible at all in the image G(Gs). When the luminance E is greater than the minimum luminance value Es, the pattern Wb on the surface Wa of the calibration substrate W is visible in the image G(Gt).

[0051] The computer 30 sets a reference value R0 for the prism position P based on the relationship between the absolute value Q of the prism position P and the brightness E of the image G (captured by the camera 20). The prism 10 moves based on the relative value R of the prism position P, with the reference value R0 as the starting point. Specifically, the computer 30 moves the prism 10 by advancing and retracting the plunger 41 of the actuator 40 based on the relative value R of the prism position P. The relative value R is obtained by subtracting the reference value R0 from the absolute value Q (R = Q - R0). When the absolute value Q is smaller than the reference value R0, the relative value R is a negative value. When the absolute value Q is larger than the reference value R0, the relative value R is a positive value.

[0052] The computer 30 sets the absolute value Q of the prism position P at which the brightness E of the image G (captured by the camera 20) becomes the minimum brightness value Es as the reference value R0 of the prism position P.

[0053] (Calculation of absolute value corresponding to minimum luminance value) Figure 5 shows the method for calculating the absolute value Q corresponding to the minimum luminance value Es. The absolute value Q and the corresponding luminance E are measured, for example, by taking measurements of the absolute value Q at intervals of 100 μm. Therefore, the absolute value Q (corresponding to the true minimum luminance value Es) when the luminance E is the true minimum luminance value Es is not necessarily measured. If the absolute value Q corresponding to the true minimum luminance value Es is not determined, the reference value R0 cannot be set.

[0054] Therefore, we utilize the fact that the graph consisting of the absolute value Q and the luminance E is an even function of degree four or higher (with a positive coefficient for the highest degree).

[0055] When the computer 30 measures multiple (at least two) absolute values ​​Q such that the luminance E is the minimum luminance measurement value Esi, it sets the absolute value Q closest to the lower limit Qa where the luminance E is the minimum luminance measurement value Esi as the lower limit minimum absolute value Qsa, and sets the absolute value Q closest to the upper limit Qb where the luminance E is the minimum luminance measurement value Esi as the upper limit minimum absolute value Qsb. Based on at least the lower limit minimum absolute value Qsa and the upper limit minimum absolute value Qsb, the computer 30 calculates the absolute value Q corresponding to the true minimum luminance value Es. The computer 30 sets the calculated absolute value Q corresponding to the true minimum luminance value Es as the reference value R0.

[0056] The minimum luminance measurement value Esi for luminance E may be the same as the true minimum luminance value Es for luminance E, or it may be greater than the true minimum luminance value Es for luminance E. The numerical value of the minimum luminance measurement value Esi may be zero, or it may be greater than zero.

[0057] The minimum luminance measurement value Esi is the smallest among the luminance E values ​​that have actually been measured (data obtained). In this example, there are four absolute values ​​Q for which luminance E results in the minimum luminance measurement value Esi (corresponding to the minimum luminance measurement value Esi).

[0058] For example, the computer 30 calculates the median between the minimum absolute value Qsa on the lower limit and the minimum absolute value Qsb on the upper limit as the absolute value Q corresponding to the true minimum luminance value Es (set as the reference value R0). Alternatively, the computer 30 calculates either the minimum absolute value Qsa on the lower limit or the minimum absolute value Qsb on the upper limit as the absolute value Q corresponding to the true minimum luminance value Es (set as the reference value R0).

[0059] While the methods described above can reduce computation time, they suffer from accuracy issues. Therefore, the following methods offer higher accuracy.

[0060] Computer 30 sets the absolute value Q at which the luminance E is maximum closer to the lower limit Qa than the lower limit minimum absolute value Qsa as the lower limit maximum absolute value Qla. Computer 30 sets the luminance E corresponding to the lower limit maximum absolute value Qla as the first maximum luminance measurement value Ela. Computer 30 sets the absolute value Q at which the luminance E is maximum closer to the upper limit Qb than the upper limit minimum absolute value Qsb as the upper limit maximum absolute value Qlb. Computer 30 sets the luminance E corresponding to the upper limit maximum absolute value Qlb as the second maximum luminance measurement value Elb.

[0061] The first maximum luminance measurement value, Ela, is the largest among the measured luminance E values ​​that were actually measured (data obtained) in the interval where the absolute value Q is closer to the lower limit Qa than the lower limit minimum absolute value Qsa. The second maximum luminance measurement value, Elb, is the largest among the measured luminance E values ​​that were actually measured (data obtained) in the interval where the absolute value Q is closer to the upper limit Qb than the upper limit minimum absolute value Qsb.

[0062] The computer 30 sets the maximum value of luminance E in the first approximation curve Ua, which is generated based on at least the lower limit maximum absolute value Qla and the first maximum luminance measurement value Ela, as the first maximum luminance theoretical value Elua. The computer 30 sets the absolute value Q corresponding to the first maximum luminance theoretical value Elua as the lower limit theoretical absolute value Qlua.

[0063] Computer 30 sets the maximum value of luminance E in the second approximation curve Ub, which is generated based on at least the upper limit maximum absolute value Qlb and the second maximum luminance measurement value Elb, as the second maximum luminance theoretical value Elub. Computer 30 sets the absolute value Q corresponding to the second maximum luminance theoretical value Elub as the upper limit theoretical absolute value Qlb. The first approximation curve Ua is generated by the least squares method based on the measurement data of absolute value Q and the measurement data of luminance E in the interval where the luminance E is a predetermined ratio (e.g., 100% to 60%) of the first maximum luminance measurement value Ela.

[0064] The second approximation curve Ub is generated by the least squares method based on the measurement data of absolute value Q and luminance E in the interval where luminance E is a predetermined ratio (e.g., 100% to 60%) of the second maximum luminance measurement value Elb. In this example, the first maximum luminance theoretical value Elua and the second maximum luminance theoretical value Elub are equal to each other. However, they may be different from each other.

[0065] The computer 30 generates a third approximation curve Uc based on at least a first maximum luminance theoretical value Elua and a lower limit theoretical absolute value Qlua, and a second maximum luminance theoretical value Elub and an upper limit theoretical absolute value Qlub.

[0066] The third approximation curve Uc is generated by the least fourth squares method based on the measured absolute value Q and luminance E data in the interval between the first point A, which is composed of the first maximum luminance theoretical value Elua and the lower limit theoretical absolute value Qlua, and the second point B, which is composed of the second maximum luminance theoretical value Elub and the upper limit theoretical absolute value Qlub. The least squares method may also be used.

[0067] Computer 30 calculates the absolute value Q corresponding to the local minimum of the third approximation curve Uc as the absolute value Q corresponding to the true minimum luminance value Es. Computer 30 sets the absolute value Q corresponding to the true minimum luminance value Es as the reference value R0.

[0068] (Effects) Conventionally, even when an appropriate absolute value Q was set for the prism position P (where surface irregularities should be clearly visible), surface irregularities were sometimes difficult to see in the image G captured by the camera 20 due to individual differences such as manufacturing errors in the differential interference contrast inspection device 1.

[0069] Therefore, the computer 30 sets a reference value R0 for the prism position P based on the relationship between the absolute value Q of the prism position P and the brightness E of the image G (captured by the camera 20). The prism 10 moves based on the relative value R of the prism position P, with the reference value R0 of the prism position P as the starting point.

[0070] The prism 10 is moved based on the relative value R of the prism position P, rather than the absolute value Q of the prism position P. This makes it possible to clearly see the surface irregularities Wa of the substrate W being inspected, regardless of individual differences in the differential interferometry inspection device 1.

[0071] The computer 30 sets the absolute value Q of the prism position P at which the brightness E of the image G (captured by the camera 20) becomes the minimum brightness value Es, as the reference value R0 of the prism position P. The reference value R0 of the prism position P can be easily set.

[0072] The computer 30 calculates the absolute value Q of the prism position P corresponding to the true minimum luminance value Es of luminance E, based on at least the lower limit minimum absolute value Qsa and the upper limit minimum absolute value Qsb. Even if the absolute value Q of the prism position P corresponding to the true minimum luminance value Es of luminance E has not been measured, the absolute value Q of the prism position P corresponding to the true minimum luminance value Es of luminance E can be calculated.

[0073] The computer 30 calculates the absolute value Q corresponding to the minimum value of the third approximation curve Uc as the absolute value Q corresponding to the true minimum luminance value Es. The absolute value Q of the prism position P corresponding to the true minimum luminance value Es of luminance E can be calculated with high accuracy.

[0074] (Other Embodiments) Although the present disclosure has been described above with reference to preferred embodiments, this description is not limiting, and of course, various modifications, substitutions, or combinations are possible.

[0075] Alternatively, the absolute value Q corresponding to the true minimum luminance value Es may be determined by finding the midpoint between the lower theoretical absolute value Qlua and the upper theoretical absolute value Qlub. Alternatively, the absolute value Q corresponding to the intersection of the line connecting the first point A and the upper minimum absolute value Qsb and the line connecting the second point B and the lower minimum absolute value Qsa may be determined as the absolute value Q corresponding to the true minimum luminance value Es.

[0076] The reference value R0 is not limited to the absolute value Q at which the luminance E becomes the minimum luminance value Es. The reference value R0 may be arbitrarily set based on the relationship between the absolute value Q and the luminance E.

[0077] The actuator 40, which serves as the prism moving mechanism, may also be a ball screw mechanism or a cylinder mechanism.

[0078] The indicator of brightness is not limited to luminance E; for example, it may also be hue, lightness, darkness, saturation, luminosity, illuminance, etc.

[0079] The object to be inspected is not limited to the surface Wa of the substrate W. The object to be inspected does not have to be a plate material. The object to be inspected is not limited to semiconductors; for example, it may be glass, metal, resin, or other materials.

[0080] The differential interference contrast inspection apparatus is not limited to a reflective type in which the polarized light is reflected from the object under inspection, as in the above embodiment, but may also be a transmissive type in which the polarized light is transmitted through the object under inspection (after hitting it).

[0081] The differential interference contrast inspection method according to this disclosure inspects an object to be inspected (e.g., the surface Wa of a substrate W) based on an image G obtained when two mutually orthogonal polarized beams La and Lb strike the object to be inspected (e.g., the surface Wa of a substrate W) through a prism 10 that splits two mutually orthogonal polarized beams La and Lb separated by a shear amount H. In the differential interference contrast inspection method, the prism 10 is movable so that the prism position P corresponding to the shear amount H changes. In the differential interference contrast inspection method, a reference value R0 of the prism position P is set based on the relationship between the absolute value Q of the prism position P and the brightness (e.g., luminance E) of the image G. In the differential interference contrast inspection method, the prism 10 is moved based on the relative value R of the prism position P, starting from the reference value R0 of the prism position P.

[0082] This disclosure is extremely useful and has high industrial applicability because it can be applied to differential interferometry apparatus and differential interferometry methods.

[0083] 1 Differential Interferometry Inspection System 10 Prism 20 Camera 30 Computer (Processing Unit) 40 Actuator (Prism Movement Mechanism) 41 Plunger W Substrate Wa Surface (Object under Inspection) Wb Pattern N Defect L Polarization La Polarization Lb Polarization Pa Point Pb Point H Shear Amount H1 First Shear Amount H2 Second Shear Amount δ Height Difference δ1 First Height Difference δ2 Second Height Difference G Image G1 First Image G2 Second Image Gs Image Gt Image O Origin P Prism Position Q Absolute Value Qa Lower Limit Qb Upper Limit Q1 First Value Q2 Second Value Qsa Minimum Absolute Value on Lower Limit Qsb Minimum Absolute Value on Upper Limit Qla Maximum Absolute Value on Lower Limit Qlb Maximum Absolute Value on Upper Limit Qlua Lower limit theoretical absolute value Qlub Upper limit theoretical absolute value R0 Reference value R R Relative value E Luminance (brightness) Es Minimum luminance value (minimum value) Esi Minimum luminance measurement value (minimum measurement value) Ela First maximum luminance measurement value (first maximum measurement value) Elb Second maximum luminance measurement value (second maximum measurement value) Elua First maximum luminance theoretical value (first maximum theoretical value) Elub Second maximum luminance theoretical value (second maximum theoretical value) Ua First approximation curve Ub Second approximation curve Uc Third approximation curve A First point B Second point

Claims

1. A differential interference contrast inspection device comprising: a prism that splits two mutually orthogonal polarized beams by a shear amount; a camera that captures an image obtained when the polarized beams strike an object to be inspected through the prism; and a processing unit that inspects the object to be inspected based on the image, wherein the prism is movable such that the prism position corresponding to the shear amount changes; the processing unit sets a reference value for the prism position based on the relationship between the absolute value of the prism position and the brightness of the image; and the prism moves based on the relative value of the prism position with respect to the reference value of the prism position.

2. The differential interference contrast analyzer according to claim 1, wherein the processing unit sets the absolute value of the prism position at which the brightness of the image is at its minimum as the reference value of the prism position.

3. The differential interference contrast inspection apparatus according to claim 2, wherein when a plurality of absolute values ​​for which the brightness is the minimum measured value are measured, the processing unit sets the absolute value for which the brightness is the minimum measured value closest to the lower limit as the lower limit minimum absolute value, sets the absolute value for which the brightness is the minimum measured value closest to the upper limit as the upper limit minimum absolute value, and calculates the absolute value corresponding to the minimum value based on at least the lower limit minimum absolute value and the upper limit minimum absolute value.

4. The processing unit sets the absolute value at which the brightness is maximum closer to the lower limit than the minimum absolute value on the lower limit as the maximum absolute value on the lower limit, sets the brightness corresponding to the maximum absolute value on the lower limit as the first maximum measured value, sets the absolute value at which the brightness is maximum closer to the upper limit than the minimum absolute value on the upper limit as the maximum absolute value on the upper limit, sets the brightness corresponding to the maximum absolute value on the upper limit as the second maximum measured value, sets the maximum value of the brightness in the first approximation curve generated based on at least the maximum absolute value on the lower limit and the first maximum measured value as the first maximum theoretical value, sets the absolute value corresponding to the first maximum theoretical value as the theoretical absolute value on the lower limit, sets the maximum value of the brightness in the second approximation curve generated based on at least the maximum absolute value on the upper limit and the second maximum measured value as the second maximum theoretical value, sets the absolute value corresponding to the second maximum theoretical value as the theoretical absolute value on the upper limit, A differential interference contrast analyzer according to claim 3, comprising generating a third approximation curve based on at least the first maximum theoretical value and the lower limit theoretical absolute value, and the second maximum theoretical value and the upper limit theoretical absolute value, and calculating the absolute value corresponding to the local minimum of the third approximation curve as the absolute value corresponding to the minimum value.

5. A differential interference contrast inspection method for inspecting an object under inspection based on an image obtained when two mutually orthogonal polarized lights strike the object under inspection through a prism that separates them by a Shear amount, wherein the prism is movable such that the prism position corresponding to the Shear amount changes, a reference value for the prism position is set based on the relationship between the absolute value of the prism position and the brightness of the image, and the prism is moved based on the relative value of the prism position with respect to the reference value of the prism position.

Images