JP2005331321A - Optical measuring apparatus and periodic array pattern measuring method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately measure a periodic array pattern without being affected by noise caused by surface roughness of a substrate.
An optical measuring device is constituted by a sensor head including a light projecting unit and a light receiving unit and a controller including a CPU and a memory. The light projecting unit 11 irradiates the measurement target substrate 1 with a long band-like beam in the electrode arrangement direction. The light receiving unit 12 includes a CCD, and is adjusted so that the reflected wave of the diffracted light from the substrate 1 can be received by being separated for each order. The CPU 21 of the controller 20 takes in the received light amount data from the light receiving unit 12 and recognizes the intensity of each order of diffracted light, and then removes 1st to 3rd order diffracted light, which is greatly affected by the surface roughness of the substrate 1. A process of comparing the intensity of the diffracted light of the order with the theoretical intensity is executed.
[Selection] Figure 4

Description

The present invention relates to a technique for measuring a periodic arrangement pattern of a structure on a substrate on which structures of a predetermined size are arranged with periodicity, such as a substrate for a liquid crystal display (LCD).
In this specification, an electrode is described as an example of the structure. However, the structure is not limited thereto, and a color filter, a black matrix, or the like may be used as the structure. For any structure, the size, shape, height, width, etc. can be measured.

  There is a technique for measuring the groove width and groove depth of a groove-like pattern by utilizing a phenomenon in which light is diffracted when light is applied to an uneven pattern (see Patent Documents 1 to 4).

JP-A-57-187604 JP-A-1-28580 Japanese Patent Laid-Open No. 9-5049 Japanese Patent Laid-Open No. 10-122835

  Among the above, Patent Document 1 discloses an approximate expression (equation (1) in the same document) relating to the intensity distribution of diffracted light obtained when light parallel to the groove pattern is irradiated. Further, Patent Document 1 discloses a calculation formula (formulas (5), (6), and (7)) for measuring the groove pattern from the diffraction angle and the intensity of diffracted light based on the formula (1). Yes. In the invention of Patent Document 1, diffracted light from an actual groove-shaped pattern is received and the diffraction angle of the first-order diffracted light, the intensity ratio between the 0th-order light and the first-order light, and the intensity between the first-order light and the second-order light. By measuring the ratio and applying these measured values to the above formulas (5), (6), and (7), the groove period p, groove width a, and groove depth h are obtained.

  Other Patent Documents 2 to 4 also describe methods for measuring a groove pattern using the intensity ratio of diffracted light. In Patent Documents 2 and 3, a theoretical curve indicating the relationship between the groove width and groove depth and the intensity ratio of diffracted light is obtained by diffraction calculation using the vector diffraction theory, and the intensity of each actually measured diffracted light is calculated. Measurement is performed by applying the ratio to the theoretical curve. Patent Document 4 describes that a periodic array pattern by a trapezoidal structure is measured using an intensity ratio for each combination of four sets of diffracted light.

  In the inventions described in Patent Documents 1 to 4 above, a groove pattern formed on an optical disk, a semiconductor substrate, or the like is an object to be measured, but a glass substrate for LCD is also measured by the same method as these inventions. It is considered possible.

  Furthermore, although not specified in the above-mentioned Patent Documents 1 to 4, a plurality of combinations of groove width and groove depth values are created, and a database in which the theoretical intensity of each diffracted light is associated with each combination is created. You can also In this case, the intensity distribution closest to the intensity distribution of each diffracted light obtained from the groove pattern to be measured can be extracted from the database, and the shape of the groove pattern corresponding to the extracted intensity distribution can be applied to the object. .

  However, since the glass substrate is difficult to polish due to its structure, the surface roughness is increased. For example, when the surface roughness has a concavo-convex change having a magnitude of about 1/100 or more of the wavelength of the irradiated light, the intensity distribution of the diffracted light may fluctuate due to a phase shift. In such a case, if the glass substrate is measured using the conventional technique as it is, a measurement error may occur, and it becomes difficult to perform stable measurement.

  The present invention has been made paying attention to the above problem, and when a substrate having a large surface roughness such as a glass substrate is a measurement target, the period on the substrate is not affected by a noise component due to the surface roughness. An object is to enable measurement of an array pattern with high accuracy.

  The present invention relates to a method and apparatus for determining the shape of a structure using diffracted light generated by irradiating light onto a substrate on which structures having periodicity are arranged, particularly for the wavelength of irradiation light. This method is applied to a case where a measurement target is a substrate having irregularities that cannot be ignored. Although a specific measurement object is a glass substrate, the present invention is not limited to this, and the present invention is applied to a substrate having a surface roughness of about 1/10 or more with respect to the wavelength of irradiation light. desirable.

  The magnitude (intensity) of the diffracted light generated by the periodic array pattern can be approximately obtained by scalar theory. Hereinafter, the basis for deriving the present invention will be described by deriving an equation showing the amplitude distribution of the diffracted light applied to the glass substrate based on the scalar theory. In the following formula, j is an imaginary component indicating a phase.

As a specific example, a case where the shape (height or width) of an electrode of a glass substrate (hereinafter simply referred to as “substrate”) on which a wiring pattern using aluminum electrodes is formed will be described. First, as shown in FIG. 1, it is assumed that the surface of the substrate 1 is completely flat, the arrangement period of the electrodes 2 is p, the width between the electrodes 2 is a, and the height of the electrode 2 is h. Further, the arrangement direction of the electrodes 2 is an x-axis, and on this x-axis, the x coordinate corresponding to the center point of the predetermined electrode 2 is 0, and the aluminum surface constituting the electrode surface (hereinafter referred to as “aluminum surface”). The range of the x coordinate corresponding to is A _n , and the range of the x coordinate corresponding to the glass surface located between the electrodes is B _n . Note that n is an integer indicating the order of the electrodes, and n = 0 at a position corresponding to x = 0.

The intensity U ₁ (x) of the reflected light from an arbitrary position x on the substrate is expressed by a different expression depending on whether the position is included in the range A _n or B _n . Specifically, the following equation (1) is obtained.

In Equation (1), r is the ratio of the amplitude reflectance r _{G on the} glass surface to the amplitude reflectance r _A (complex amplitude) on the aluminum surface (see Equation (2)). Φ is the phase difference between the reflected light from the glass surface and the reflected light from the aluminum surface. This phase difference φ corresponds to a difference in optical path length between the glass surface and the aluminum surface (corresponding to a value (2h / cos θ) obtained by doubling w in FIG. 1, where θ is an incident angle of light). Is equivalent to a value obtained by adding the phase difference φ _u due to the above and the phase difference φ _t due to the difference in reflectance of each surface. φ _u and φ _t are obtained by the equations (3) and (4), respectively. In the equation (4), Im (r _A ) is an imaginary part of the reflectance r _A of aluminum, and Re (r _A ) is a real part of the reflectance r _A.

Here, when a light receiving unit such as a one-dimensional CCD is disposed at a distance l from the substrate, the intensity distribution of the reflected light extracted by the light receiving unit is expressed by U ₂ (X) in the following equation (5). The Incidentally, X of the equation (5), the distance l, represented by the wavelength lambda, and the using the coordinates x ₂ on the light receiving means (6) of the illumination light. N is the number of periods p included in the light irradiation range.

  The above equation (5) is almost the same as the equation (1) in Patent Document 1. In other words, if the surface of the substrate is flat, the conventional technique can be applied as it is for measurement. However, since surface roughness is generated on the actual surface of the substrate 1, it is necessary to derive the intensity of reflected light in consideration of this surface shape.

In FIG. 2, it is assumed that the surface roughness of the substrate 1 changes in a sine wave shape, and the change period is q and the maximum amplitude is g. Here, when the magnitude of the amplitude g has a magnitude that cannot be ignored with respect to the wavelength λ of the irradiation light, the difference in optical path length due to the amplitude causes a phase shift, and the intensity of the reflected light is expressed by the formula (5). May be different from the value of U ₁ (X).

Next, a method for obtaining the intensity of each diffracted light in consideration of the surface roughness of the substrate 1 will be described step by step. As shown in FIG. 2, when the change in the surface roughness of the substrate 1 is approximated to a sine wave, the intensity U _1A (x) of the reflected light from the arbitrary position x on the substrate 1 is expressed by equation (7).

U ₁ (x−np) in the latter half F of the equation (7) is based on the equation (1) ((x−np) corresponds to x in the equation (1)). That is, the portion F represents a phase reflecting the periodic arrangement pattern of the glass surface and the aluminum surface.

On the other hand, the first half G shows a phase delay caused by the amplitude g and the period q of the surface roughness. Here, the change in the phase lag, believes cosine wave maximum amplitude and phi _g. Note that φ _g is a phase difference generated when light reciprocates in a portion corresponding to the maximum amplitude g, and is obtained by Equation (8) based on the amplitude g, wavelength λ, and incident angle θ.

Since the first half G of the equation (7) is an exp function, and U ₁ (x−np) of the second half F is also expressed by the exp function, the phase corresponding to G is represented by the equation (7). It can be considered that an operation of subtracting the phase delay indicated by F from is executed.

Here, by applying the above equation (7) to the Franhofer diffraction equation, the amplitude distribution of the reflected light obtained at a position separated from the substrate 1 by the distance l can be obtained by the following equation (9). . Note that U _2A (X) in the expression (9) also represents the intensity of reflected light extracted by a light receiving means such as a one-dimensional CCD, similarly to the intensity U ₂ (X) in the expression (5). . X is also represented by the above formula (6).

Furthermore, when φ _g << 1, U _2A (X) can be expressed by an approximate expression such as the expression (10).

  In the above equation (10), terms that are not recognized in equation (5) are enclosed by dotted line frames A1, A2, A3, and A4. These A1, A2, A3 and A4 portions all include a sinc function. The sinc function is a function expressed as sinc (x) = sinc (πx) / πx, and has a characteristic that the value is close to 0 when x ≠ 0 and the value becomes remarkably large when x = 0. . Therefore, it can be considered that the portions A1 and A3 become larger when X = −1 / q, and the portions A2 and A4 become larger when X = + 1 / q.

  Further, according to the fraction at the end in the equation (10), it can be seen that the amplitude distribution according to the equation (10) takes a local maximum value with a period of sinπpX. Since these maximum values can be considered to correspond to m-order (m is an arbitrary integer) diffracted light, πpX = mπ.

Therefore, from X = m / p = ± 1 / q, the values of A1, A2, A3, and A4 in the amplitude distribution U _2A (X) shown by the above equation (10) increase ± p / q order It can be considered that it is the diffracted light.

  According to the above, when the possible range of the electrode period p is known, the order of the diffracted light in which the noise corresponding to the A1 to A4 overlaps is specified based on the range and the roughness period q. can do. For example, in a large LCD substrate, the surface roughness period is 100 μm or more, and the electrode period is about 300 to 500 μm. Therefore, if the period is 300 to 400 μm, there is a high possibility that the noise overlaps at least ± 1 to 3 order diffracted light, and if the period is 400 to 500 μm, at least ± 1 to 4 order diffracted light. It can be considered that the possibility that the noise overlaps increases.

  In this invention, based on the principle described above, the reflected wave of the diffracted light generated by irradiating the substrate on which the periodic structures are arranged is separated and received for each order, and the intensity of each diffracted light is received. When the shape of the structure is obtained by comparing each with the theoretical intensity, the order affected by the noise component due to the surface roughness of the substrate among the diffracted light of each order It is removed from the target.

  The order affected by the noise component can be obtained from a numerical range that can be taken by each of the period q approximating the surface roughness and the period p of the structure based on the theory described above. In addition, the measurement of the diffracted light is executed a plurality of times using the model of the substrate to be measured, and the order in which the intensity variation exceeds a predetermined value among the diffracted lights of each order is extracted as the order overlapping the noise component. You can also

  When obtaining a degree overlapping with a noise component caused by the surface roughness based on a numerical range that can be taken by the period q approximating the surface roughness and the period p of the structure on the substrate, for example, taking ± p / q A range to be obtained can be obtained, and each order included in this range can be set as an order overlapping with the noise component. However, the present invention is not limited to this, and it is also possible to extract a range slightly wider than the range that can be taken by ± p / q, and to set each order that falls within that range as an order that overlaps the noise component.

  An optical measurement apparatus according to the present invention separates, for each order, a light projecting unit that irradiates light on a substrate on which periodic structures are arranged, and a reflected wave of diffracted light generated on the substrate by the irradiated light. A light receiving means for receiving light and a measuring means for determining the shape of the structure by comparing the intensity of each diffracted light obtained by the light receiving means with the theoretical intensity. The light projecting means includes a light source and a lens group, and is preferably configured to irradiate light extending along the arrangement direction of the structures. The light receiving means includes a light receiving element such as a CCD, and is adjusted so that a plurality of orders of diffracted light from the substrate can be incident on different positions on the light receiving surface. The measuring means is preferably a computer that operates by a program for comparison processing. Furthermore, it is desirable that a database including the theoretical strength is set in the memory of the computer.

  Further, in the above optical measuring apparatus, the measuring means is configured to exclude a predetermined order of diffracted light selected in advance from the comparison target. The order can be selected, for example, by inputting a numerical value (1, 2, 3, etc.) indicating the order overlapping the noise component. Alternatively, for the period q approximating the surface roughness and the period p of the structure, input of numerical ranges that can be taken by the period may be received, and the order overlapping the noise component may be calculated based on the above theory. . Further, when only specific types of substrates are targeted, and the periods p and q are substantially fixed, a predetermined order based on these periods may be set to be always selected in advance. .

  Further, in one aspect of the optical measurement apparatus described above, the order selected in advance for each of the plurality of shape data of the structure (referred to a combination of width and height of the structure; the same applies to the following). A memory in which the theoretical intensity of each diffracted light is registered is included. In this case, the measuring means executes the comparison processing for each diffracted light separated by the light receiving means, for which the theoretical intensity is registered.

  However, theoretical intensities may be registered for diffracted light of multiple orders including the selected order. The measuring means in this case can exclude the diffracted light of the selected order from the diffracted lights separated by the light receiving means, and execute the comparison process using the remaining orders of diffracted light.

  Next, another optical measurement apparatus according to the present invention includes a light projection unit, a light reception unit, and a measurement unit similar to those described above, and includes noise components caused by the surface roughness of the substrate in the diffracted light of each order. Selection means for selecting the affected order; In addition, the measuring unit is configured to exclude the diffracted light of the order selected by the selecting unit from the diffracted lights separated by the light receiving unit from the comparison target.

  In the above, the selection means can be an arithmetic circuit that derives the order overlapping the noise component from the period q or the period p of the structure based on the principle described above. In this case, it is desirable to provide an input means for inputting a numerical range that can be taken by the period q and the period p of the structure for the substrate to be measured. The selection means can be configured to select, for example, an order corresponding to a possible range of p / q (either one of the positive and negative orders or both of the positive and negative orders).

In addition, the optical measuring apparatus may include an input unit that inputs sample data of the intensity distribution of the diffracted light obtained by using the light projecting unit and the light receiving unit for the measurement target substrate. In this case, the selection means can be an arithmetic circuit that obtains the variation in the intensity of the diffracted light among a plurality of sample data for each order and selects the order in which the variation of a predetermined value or more occurs. The above input means may be a means for inputting sample data acquired in advance from the outside, but is configured so that the received light amount data from the light receiving means can be directly input when the model substrate is being measured. May be.
In addition, when the selection unit takes any form, the selection unit is preferably configured by a computer, but is not limited thereto, and may be configured by an ASIC (specific application IC).

  In the optical measurement apparatus, the theoretical intensity of the plurality of diffracted lights that can be received by the light receiving means for each of the plurality of shape data of the structure, regardless of which aspect the selecting means takes. It is desirable to register.

  In the optical measurement apparatus according to the present invention, a substrate for a liquid crystal television can be a main measurement object. In a popular type 18-30 inch W-XGA type liquid crystal substrate, the pixel pitch is 300-400 μm. As an apparatus for measuring such a liquid crystal substrate, in the present invention, a light projecting means for irradiating the substrate with light, and a light receiving means for separating and receiving diffracted light from the substrate with respect to the irradiated light for each order. And measuring means for determining the shape of the structure by comparing the intensity of each diffracted light separated by the light receiving means with the theoretical intensity, and the measuring means comprises at least first to third order diffracted lights. Is configured to be excluded from the subject of the comparison process.

  According to the measurement by the inventors, the surface roughness period q of this type of liquid crystal substrate is considered to be 100 μm or more. Therefore, according to the above theory, when the period of the structure (electrode) is 300 to 400 μm, the noise component due to the surface roughness is highly likely to affect the ± 1 to 3rd order diffracted light. . In the optical measuring device, depending on the range of diffracted light that can be received by the light receiving means, +1 to 3rd order, or -1 to 3rd order, or ± 1 to 3rd order diffracted light is excluded from the comparison target, The intensities of orders of diffracted light that are not easily affected by other noises can be compared with the theoretical intensities, respectively. When the period of the structure is close to 400 μm, fourth-order diffracted light can be excluded from the comparison target.

  Furthermore, in the present invention, as an optical device having a liquid crystal substrate in which structures are arranged with a period of 400 to 500 μm larger than the above as an object to be measured, a light projecting means for irradiating the substrate with light, A light receiving means for separating and receiving the diffracted light from the substrate for each order; and a measuring means for determining the shape of the structure by comparing the intensity of each diffracted light separated by the light receiving means with the theoretical intensity. And the measurement means provides an apparatus configured to exclude at least 1-4th order diffracted light from the object of the comparison process. This is also based on the fact that the surface roughness period q of the liquid crystal substrate is 100 μm or more, as described above. When the period of the structure is close to 500 μm, the fifth-order diffracted light can be excluded from the comparison target.

  According to the present invention, when a measurement target is a substrate having a surface roughness that cannot be ignored with respect to the wavelength of the irradiation light, the order affected by the noise component due to the surface roughness can be measured. Since it is not used, the shape of the structure can be stably measured without being affected by noise components.

FIG. 3 shows the appearance of an optical measuring apparatus according to one embodiment of the present invention together with a usage example.
This optical measuring device is for measuring the size and arrangement period of electrodes formed on a glass substrate 1 for LCD, and comprises a sensor head 10 and a controller 20 connected by a cable 4. . The controller 20 is a personal computer in which an interface board and a program for the sensor head 1 are incorporated, and includes peripheral devices such as a keyboard 201 and a monitor 202.

  The sensor head 10 includes a light projecting unit using a laser diode as a light source, a light receiving unit having a one-dimensional CCD (hereinafter simply referred to as “CCD”), and the like. The controller 20 controls the operation of the sensor head 10 and takes in the output from the CCD and executes a measurement process described later.

  FIG. 4 shows the electrical configuration of the optical measurement apparatus. In addition to the light projecting unit 11 and the light receiving unit 12, the sensor head 10 includes a light projecting circuit 13, a light receiving circuit 14, a timing control circuit 15, and the like. The configurations of the light projecting unit 11 and the light receiving unit 12 will be described later with reference to FIG. 6. The light projecting unit 11 includes a laser diode 111 (shown in FIG. 6), and the light receiving unit 12 includes a CCD 122. (Shown in FIG. 6) are respectively deployed.

  The controller 20 has a normal configuration in a personal computer such as a CPU 21, a memory 21, a hard disk 23, and an input / output unit 24, and has the dedicated interface board 200 described above. An interface unit 25, an A / D conversion circuit 26, a sensor control circuit 27, a trigger input unit 28, a power supply circuit 29, and the like are mounted on the interface board 200, and are connected to the CPU bus 201 via the interface unit 25. Yes.

  In the above, the hard disk 23 of the controller 20 stores a program for executing measurement processing, a determination table, and the like. The memory 22 is used to temporarily store received light amount data used for measurement. The input / output unit 24 includes a keyboard 201, a monitor 202, an output terminal to an external device (not shown), and the like.

  The sensor control circuit 27 of the interface board 200 is for giving a trigger signal that informs the timing control circuit 15 of the sensor head 1 of the timing of the measurement process. The trigger signal can be generated based on an external signal from the trigger input unit 28 or can be generated based on a command from the CPU 21. Which trigger signal is used can be set by data input in advance. The trigger input unit 28 is connected to a substrate detection sensor or the like.

  The timing control circuit 15 generates a timing signal having a predetermined length according to the trigger signal, and outputs the timing signal to the light projecting circuit 13 and the light receiving circuit 14. The light projecting circuit 13 drives the laser diode 111 of the light projecting unit 11 based on this timing signal. The light receiving circuit 14 drives the CCD 122 of the light receiving unit 12 based on the timing signal. Thereby, the reflected light from the glass substrate 1 can be received at the timing synchronized with the light emission of the laser diode 111.

  The received light amount signal generated by the CCD 122 is input to the A / D conversion circuit 26 of the interface board 200 and digitally converted. The CPU 21 takes in the digital data of the received light amount signal (hereinafter referred to as “received light amount data”) via the interface unit 25 and stores it in the memory 22. Then, the CPU 21 uses the received light amount data in the memory 22 to store the data. Execute the measured process.

  The power supply circuit 29 is for supplying power to the sensor head 1. Although not shown in FIG. 4, the power supply line from the power supply circuit 29, the transmission line for the trigger signal from the sensor control circuit 27, the transmission line for the received light amount signal from the CCD 122, and the like are accommodated in the cable 4. It will be.

  FIG. 5 shows an enlarged side and top surface of the glass substrate 1. This glass substrate 1 (hereinafter simply referred to as “substrate 1”) has a predetermined thickness t, and rectangular electrodes 2 each having a predetermined width are provided on the upper surface 1a along vertical and horizontal directions at regular intervals. Deployed every other time. In this embodiment, attention is paid to the periodic arrangement of the electrodes 2 in either one of the two directions (of course, it is also possible to pay attention to the periodic arrangement in each direction in order), and a predetermined length along the arrangement direction. Irradiate the band-like light 5. Thereby, a plurality of diffracted lights reflecting the uneven state of the electrode 2 are generated and reflected on the front surface and the back surface of the substrate 1. The surface of the substrate 1 corresponds to the upper surface 1a on which the electrodes 2 are arranged. Further, the back surface of the substrate 1 can be considered as the inner surface 1b of the bottom surface. The bottom surface of the substrate 1 is supported on a surface plate (not shown).

  In FIG. 5, h represents the height of the electrode 2, d represents the width of the electrode 2, a represents the spacing between the electrodes 2, and p represents the period of the arrangement of the electrodes 2. In this embodiment, the thickness t and the period p of the substrate 1 are constant, and the values of h and d are measured. For this measurement, in this embodiment, the theoretical intensity of the surface reflected light of each diffracted light is obtained in advance for a plurality of periodic array patterns having different combinations of the values of h and d, and stored in the hard disk 23. Registered. Specifically, a determination table in which data including the theoretical intensity of each diffracted light (hereinafter referred to as “theoretical value data”) is registered for each shape of the periodic array pattern. CPU21 measures the intensity | strength of the surface reflected light concerning each order using the received light amount data obtained from the board | substrate 1 of a process target, and collates the measured value with the said determination table, h, d in the said board | substrate 1 Specify the value of. Details of this processing will be described later.

  FIG. 6 shows the main components of the light projecting unit 11 and the light receiving unit 12 in the sensor head 1 together with their functions. In the figure, x and y are coordinate axes based on the arrangement of the electrodes to be measured, and x corresponds to the arrangement direction of the electrodes and y corresponds to the direction orthogonal to the arrangement direction. In addition, a balloon (A) in the figure indicates a light irradiation state on the substrate 1, and a balloon (B) indicates a condensing state of reflected light on the CCD 122.

  The light projecting unit 11 includes a collimating lens 112, a cylindrical lens 113, and a condensing lens 114 arranged in this order in front of a laser diode 111 serving as a light source. On the other hand, the light receiving unit 12 has a configuration in which a cylindrical lens 121 is provided in front of the CCD 122. The laser diode 111 and the lenses 112, 113, and 114 on the light projecting unit 11 side are supported by dedicated holders 115, 116, 117, and 118, respectively. Similarly, the lens 121 and the CCD 122 on the light receiving unit 12 side are also supported by dedicated holders 123 and 124. As the laser diode 111, a laser diode that emits laser light having a wavelength of approximately 650 to 850 nm is used.

  In the above, the light emitted from the laser diode 111 is collimated by the collimator lens 112 and then passes through the cylindrical lens 113 and the condenser lens 114 in order. The cylindrical lens 113 narrows down the parallel light that has passed through the collimating lens 112 in the y direction, and generates band-shaped light that is long in the x direction. The parallel light constituting the belt-like light is converted into focused light by the condenser lens 114. However, since a condenser lens having a focal length sufficiently longer than a standard distance from the lens 114 to the substrate 1 is used as the condenser lens 114, as shown in the balloon (A), The belt-like light 5 having a length that crosses the predetermined number of electrodes 2 can be irradiated.

  The CCD 122 on the light receiving unit 12 side is disposed at a position corresponding to the focal length of the condensing lens 114 and capable of separating and receiving the reflected light of the diffracted light with respect to the band light 5 (details will be described later). .) Note that, due to the action of the cylindrical lens 113 on the light projecting unit 11 side, each reflected light becomes light spreading in a direction opposite to the direction of narrowing on the light projecting unit 12 side, but the cylindrical lens 121 on the light receiving unit 12 side is This spread is narrowed down, and the reflected light functions to form a strip-shaped light 6 having a predetermined length.

In this embodiment, the curvature of the cylindrical lens 121 is adjusted so that the strip light 6 is longer than the width of the pixel 122 a of the CCD 122. The CCD 122 is arranged with its pixel arrangement corresponding to the direction in which the strip light 6 is arranged. In the balloon (B), the pixel arrangement direction of the CCD 122 is shown as x ₂ , and the direction orthogonal to this is shown as y ₂ .

  In this optical measurement apparatus, in order to stabilize the measurement accuracy, a process of removing two types of noise from the intensity distribution of the reflected light extracted by the CCD 122 is executed. The first noise is caused by reflected light from the back surface 1 b of the substrate 1, and the second noise is caused by the surface roughness of the substrate 1.

The reason why the reflected light from the back surface 1b is considered as noise is as follows.
As described above, since this embodiment uses a glass substrate as a measurement target, a part of the diffracted light generated on the substrate surface is transmitted through the substrate 1 and reflected on the back surface 1b. Hereinafter, the diffracted light reflected by the back surface 1b is referred to as “back surface reflected light”. On the other hand, the diffracted light reflected by the surface 1a of the substrate 1 is referred to as “surface reflected light”.

  When the thickness t of the substrate 1 is sufficiently larger than the coherent length of the irradiation light, the front surface reflected light and the back surface reflected light do not interfere with each other, and theoretically, the front surface reflected light and the back surface reflected light are at the same position on the CCD. The optical system is adjusted so as to be incident on the light beam, and the intensity of the reflected light for each order may be compared with the theoretical intensity. However, in the actual measurement process, since the substrate 1 is fixed on the surface plate, the back surface is affected by the influence of multiple reflections generated between the air layers generated between the surface plate and the substrate and the variation in the reflectivity of the surface of the surface plate. It becomes difficult to obtain the intensity of the reflected light correctly. For this reason, it is necessary to remove the intensity resulting from the back surface reflected light as noise.

  Next, since the surface roughness of the glass substrate has an amplitude that cannot be ignored with respect to the wavelength of irradiation light (650 to 850 nm), the optical path length (from the light projecting portion 11 to the substrate 1 and from the substrate 1). Variation of the length of the optical path to the light receiving unit 12). Due to the phase shift due to the variation in the optical path length, the reflected light may not exhibit the theoretical intensity. As described in detail in the previous section “Means for Solving the Problems”, the noise component due to the surface roughness becomes remarkable at a specific order, but does not become so large at other orders. Therefore, if the order in which the noise component becomes significant is excluded from the comparison processing targets, it is possible to prevent measurement errors from occurring.

In this embodiment, for the first noise, the optical system is adjusted so that the front surface reflected light and the back surface reflected light on the CCD 122 can be received separately, and only the intensity of the front surface reflected light is extracted. Yes.
Further, the second noise is subjected to a measurement process excluding the surface reflection light of the order in which the noise component becomes significant.

First, the process for removing the back surface reflected light will be described.
FIG. 7 schematically shows the progress of light by the optical system of FIG. In FIG. 6, irradiation light and front surface reflected light to the substrate 1 are indicated by solid lines, and transmitted light and back surface reflected light to the substrate 1 are indicated by alternate long and short dash lines. In FIG. 7, only the optical path for the 0th-order diffracted light is shown in order to clarify the relationship between the front surface reflected light and the back surface reflected light, but the same relationship can be obtained for other diffracted light. .

The parallel light generated by the light projecting unit 11 is converted into focused light by passing through the condenser lens 114 and then irradiated onto the substrate 1. Since the substrate 1 is a mirror surface, the focused state is maintained even in the surface reflected light, and the substrate 1 is focused at a predetermined position. The converged state is similarly maintained even with the back surface reflected light, but this back surface reflected light is reflected farther than the surface reflected light with respect to the light projecting unit 11, and thus is condensed at a position different from the surface reflected light. It becomes like this.
When attention is paid to the front surface reflected light and the back surface reflected light corresponding to one optical path of the irradiation light, these reflected light becomes light traveling in parallel above the substrate 1.

  In this embodiment, the height d and width a of the electrode 2 described above are predetermined values, and the collection of surface reflected light when the condenser lens 114 is placed at the standard distance from the substrate 1. The CCD 122 is arranged according to the light position. Further, when the arrangement pattern of the electrodes 2 and the height position of the sensor head 10 are changed, the condensing position of the surface reflected light with respect to the CCD 112 is also considered to change. Further, since the back surface reflected light has a longer distance from the condensing lens 114 to the reflection position than the front surface reflected light, if the CCD 122 is arranged in accordance with the condensing position of the front surface reflected light as described above, Condensed to the front. However, if the focal depth of the condensing lens 114 is increased, the deviation of these condensing positions becomes slight, and any reflected light can be incident on the CCD 112 in a state that can be regarded as a condensing state. Therefore, the front surface reflected light and the back surface reflected light are separately incident on the CCD 122, and a clear image of each reflected light can be obtained.

  FIG. 8 shows a desirable state of the intensity distribution of the received light amount obtained by each pixel of the CCD 122 when the optical system is adjusted as shown in FIG. FIG. 8A is a distribution curve when it is assumed that only the surface reflected light is collected on the CCD 122, and a mountain-shaped image corresponding to the surface reflected light of each order appears at a predetermined interval. Yes. Hereinafter, this mountain-shaped image is referred to as a “surface reflected light image”. FIG. 8 (2) is a distribution curve when it is assumed that only the back surface reflected light is collected. Similarly, mountain-shaped images corresponding to the back surface reflected light of each order appear at a predetermined interval. Yes. Hereinafter, this mountain-shaped image is referred to as a “back surface reflected light image”. FIG. 8 (3) is a curve obtained by adding the intensity indicated by the distribution curve (2) to the intensity indicated by the distribution curve (1). When the thickness t of the substrate 1 is larger than the coherent length of the light source and the front surface reflected light and the back surface reflected light do not interfere with each other, a received light amount distribution curve as shown in (3) is obtained.

Incidentally, any of the above curves, the horizontal axis represents the coordinates of the received light amount data, corresponding to the x ₂ direction balloon (B) of FIG. 6. When the light projecting unit 11 and the light receiving unit 12 are arranged with the relationship shown in FIG. 6, in FIG. 8, the coordinates increase as the distance from the light projecting unit 11 increases.

When the reflected light of each diffracted light is collected on the CCD 122 based on the principle shown in FIG. 7, the surface reflected light image P _m and the back surface of the m-th order diffracted light as in the example of FIG. a reflected light image p _m comes to appear at different positions. Further, in the example of FIG. 8, m order of the back reflected light image p _m is positioned between the (m + 1) The following surface reflection image P _{m + 1} of Totsugi m following surface reflected light image P _m . In other words, the front surface reflected light image and the back surface reflected light image are arranged alternately and in order.

In addition, when light is applied to a concavo-convex pattern made of a rectangular structure such as the periodic pattern of the electrodes, the 0th-order diffracted light is predominantly dominant over other diffracted light in many cases. Surface reflection image P ₀ peak maximum in the example of FIG. 8 (1) is the back surface reflection image p ₀ peak maximum in the example of FIG. 8 (2), be thought of as respectively corresponding to the zero-order diffracted light it can. These reflected light images P ₀ and p ₀ can be considered to be adjacent to each other as shown in FIG. 8 (3) in the distribution curve of the actually received light amount. In the figure, the images P ₁ , p ₁ , P ₂ , p ₂ , P ₃ , p ₃ ... On the right side of the 0th order are the first, second, third, etc. surface reflected light images. And images P ₋₁ , p ₋₁ , P ₋₂ , p ₋₂ , P ₋₃ , and p ₋₃ on the left side of the 0th order are −1st order, −2nd order, −3 Next are a front surface reflected light image and a back surface reflected light image.

  In the optical measuring apparatus according to this embodiment, conditions are obtained in advance so that the front surface reflected light and the back surface reflected light of each diffracted light are separated and condensed on the CCD 122 in the state shown in FIG. Based on this condition, the focal length of the condenser lens 114 and the positional relationship between the condenser lens 114 and the CCD 122 are adjusted. Hereinafter, this condition will be described with reference to FIGS.

First, FIG. 9 shows conditions for arranging the front surface reflected light image and the back surface reflected light image according to each order alternately and in the order of the order as shown in FIG. 8 (3). In the figure, P _m and P _{m + 1} are the m-th order and (m + 1) -order surface reflected light images, respectively, and p _m and p _{m + 1} are the m-th order and (m + 1) -order back-surface reflected light images, respectively. FIG 8 (3), such to obtain the received light amount distribution curve as the distance bx surface reflection image P _m of the surface reflected light image P _m and the back reflected light image p _m according to the same _degree, p m + ₁ It needs to be smaller than the distance ax. That is, a relationship of ax> bx is required.

  FIG. 10 (1) shows the distance ax by the optical paths of the m-th order and (m + 1) -th order surface reflected light. L in the figure is the length of the optical path of the 0th-order surface reflected light when adjusted so that the 0th-order surface reflected light generated at the light irradiation position C on the substrate 1 is condensed on the CCD 122. Hereinafter, this L is referred to as a distance L. In this embodiment, a condition necessary for adjusting the optical system is indicated by a distance L.

In the figure, θ _m is the mth-order diffraction angle. Here, when the arrangement period of the electrodes 2 is p, the wavelength of irradiation light is λ, the thickness of the substrate is t, and the incident angle of light is θ, the diffraction angle θ _m can be obtained by the following equation (A). .

θ _{m + 1} is the (m + 1) -order diffraction angle, and can be obtained by replacing m in the above equation (1) with (m + 1). Here, when the diffraction angles θ _m and θ _{m + 1} are considered to be extremely small (θ _m , θ _{m + 1} << ₁ ), the distance ax approximates a value according to the following equation (B).

Next, FIG. 10B shows the distance bx by the optical path of the front surface reflected light and the back surface reflected light of the m-th order diffracted light. The angle θ _tr in the figure is an angle formed by the diffracted light transmitted through the substrate 1 with respect to the vertical direction. When the diffraction angle θ _m is extremely small, the distance bx approximates the following equation (C). In the formula (C), n ₂ is the refractive index of the back surface of the substrate.

  By applying the above expressions (B) and (C) to the above-described condition: ax> bx, the following expression (D) can be derived for the distance L.

In the optical measurement apparatus according to this embodiment, when the sensor head 1 is set apart from the substrate 1 by a predetermined distance, the collection is performed so that the distance between the substrate 1 and the CCD 122 satisfies the relationship of the above equation (D). The positional relationship between the optical lens 114 and the CCD 122 is adjusted. Further, the condensing lens 114 has a focal length such that, in the above positional relationship, the front surface reflected light is collected on the CCD 122 and the back surface reflected light can also be regarded as being collected. Things are adopted. The parameters p, λ, θ, t, and n ₂ necessary for specifying the distance L can be specified in advance based on the substrate to be measured, the pattern of the electrodes, the configuration of the light projecting unit 11, and the like.

  By performing the adjustment as described above, the CCD 122 can generate a distribution curve in which pairs of the front surface reflected light image and the back surface reflected light image of the same order are arranged in order. The CPU 21 of the controller 20 extracts data indicating the surface reflected light image of each order from the received light amount data indicating the distribution curve, and measures the periodic pattern of the electrodes using those intensities.

  FIG. 11 shows the procedure of the CPU 21 relating to the extraction of the intensity of the surface reflection light described above. Since this procedure corresponds to the detailed procedure of step 2 in FIG. 16 to be described later, each step is indicated by a number in the 20th generation. In FIGS. 10 and 16, each step is abbreviated as “ST”. In the following description, “ST” will be used accordingly.

  As described above, the front surface reflected light and the back surface reflected light applied to the 0th-order diffracted light are higher in intensity than the reflected light applied to other orders, and are positioned side by side in the received light amount distribution curve. . Therefore, in this procedure, in ST21, the one with the maximum peak and the one with the second largest peak are extracted from each reflected light image in the received light amount data. Then, after confirming that the distance between these peaks is a value close to the distance bx and that no other peak exists between them, two reflected light images corresponding to these peaks are converted into the 0th-order diffracted light. Specified as corresponding to

  In the next ST22, one of the pair of reflected light images corresponding to the 0th-order diffracted light is specified as the surface reflected light image. As shown in FIG. 7, the back surface reflected light is reflected at a position farther from the light projecting unit 11 than the front surface reflected light and travels in parallel with the surface reflected light. Also, the light is condensed at a position away from the light projecting unit 11. Therefore, in ST22, an image closer to the light projecting unit 11 among the pair of reflected light images (which is an image having a smaller coordinate according to the coordinate axis in FIG. 8) is specified as a surface reflected light image. .

  Further, in the distribution curve of the received light amount according to this embodiment, the surface reflected light images relating to the respective orders are arranged in the order of the orders with a distance close to ax in FIG. Therefore, in the next ST23, the position separated from the position of the 0th-order surface reflected light image by a value obtained by multiplying the distance ax by m (m <0, m> 0) is the position of the mth-order surface reflected light image. As specified. In addition, the specific value of ax can be calculated | required by above-described Formula (B).

In ST24, the intensity of each identified surface reflected light image is obtained. As the intensity, in this embodiment, as shown in FIG. 12, the integrated values of the data included in the ranges of the widths w _m , w _{m + 1} ... For each of the surface reflected light images P _m , P _{m + 1.} Asking for. Note that the widths w _m , w _{m + 1} ... In this case correspond to regions where an intensity of 1 / e ² or more of the peak of the reflected light image is obtained.

  Next, processing for noise caused by the surface roughness of the substrate 1 will be described. According to the theory described above, the order in which this noise overlaps can be obtained based on the period q approximating the surface roughness and the period p of the electrode 2.

  FIG. 13A is a graph showing the results of measuring the surface shape of the model substrate 1 with a stylus profiler. The horizontal axis of this graph indicates a change in position in a predetermined direction on the substrate, and the vertical axis indicates a change in height of the substrate surface. According to this graph, the height of the substrate surface varies in a range of approximately −100 to 350 angstroms, and the maximum amplitude is about 30 nm.

FIG. 13 (2) shows a spatial frequency distribution obtained by Fourier transforming the measurement result of FIG. 13 (1). According to this graph, it is recognized that high mountains are concentrated in the range of ⁰ to 0.01 μm ⁻¹ . When this frequency range is converted into a wavelength, it can be considered to be 100 μm or more.

  According to the measurement result of FIG. 13, it can be considered that the period q approximating the surface roughness of the substrate 1 is 100 μm or more. In addition, since the period p of the electrode 2 is 300 μm, according to the above-described theory, the influence of noise due to the surface roughness among the diffracted light of each order becomes remarkable. Can be considered.

  FIG. 14 shows the theoretical intensity of the diffracted light calculated for the case with surface roughness and the case without surface roughness, respectively, and contrasted. The graph (A) in the figure is the calculation result when the period q is 200 μm, and the graph (B) is the calculation result when the period q is 125 μm. The above-described equation (5) was used for the calculation when there was no surface roughness, and the above-mentioned equation (10) was used for the calculation when there was surface roughness. The amplitude g of the surface roughness was 10 nm, and the electrode period p was 300 μm. In FIG. 14 and the next FIG. 15, the intensities of the diffracted light of the 0th order and the positive order are respectively normalized by the intensities of the 0th order diffracted light.

  According to the theory described above, it is the order corresponding to ± p / q that is strongly influenced by noise due to surface roughness. In the graph (A), a large difference appears in the intensity of the first-order diffracted light. In the graph (B), a large difference appears in the intensity of the second-order diffracted light, and the above theory is almost supported. Can think.

  Next, the inventor has verified that measurement that is not affected by noise due to surface roughness can be performed by removing ± 1st to 3rd order diffracted light using an actual substrate. FIG. 15 shows the result. In this example, the substrate of the same type used for the measurement in FIG. 13 is irradiated four times while changing the light irradiation position, and the intensity of diffracted light with respect to hourly irradiation (the surface reflection light described above) Strength) was measured. FIG. 15 shows the measurement results, and the measurement results of each time are shown by dotted lines and chain lines. According to this measurement result, it can be seen that there are large variations in the measured values of the diffracted light in the first to third orders.

  Further, the inventor prepares theoretical value data including the theoretical intensity of the diffracted light for each of the plurality of types of periodic array patterns, and for each of the four measurement results, each of the theoretical value data is the most. A process for extracting the ones close to the measurement results was performed. In addition, each intensity | strength of theoretical value data was computed by the vector theory similar to employ | adopted by the said patent documents 2 and 3. FIG. Also, in the process of extracting the theoretical value data close to the measurement result, the least square method is used, and if the minimum error is within the predetermined value, the theoretical value data when the error is obtained corresponds to the measurement result. It was. As a result, the same theoretical value data could be extracted for each of the four measurement results. The solid line in FIG. 15 shows the theoretical value data. The measured values of the 1st to 3rd order diffracted light tend to show different values for this theoretical intensity.

  According to the measurement result of FIG. 15, when the surface roughness period q of the substrate 1 is 100 μm or more and the period p of the electrode 2 is about 300 μm, the 1st to 3rd order diffracted light is excluded. The theory that measurement accuracy can be stabilized can be supported.

  However, according to the graph of FIG. 15, the fourth-order diffracted light also has intensity variations comparable to the third-order diffracted light. Some of the diffracted lights of the fifth order and higher show some variation. For this reason, the intensity of a large number of 0th-order and 4th-order diffracted lights is extracted instead of the conventional process using 2-3 diffracted lights, and the least square error between these and the theoretical value data is extracted. It is desirable to perform measurement by the method of obtaining

FIG. 16 illustrates a series of procedures of measurement processing executed by the optical measurement device.
First, in the first ST1, a process for capturing received light amount data necessary for the measurement process is performed. In this processing, after the light projecting circuit 13 and the light receiving circuit 14 on the sensor head 1 side are driven via the sensor control circuit 27, the received light amount signal from the CCD 122 is A / D converted and stored in the memory 22. Become.

  In the next ST2, the intensity of the surface reflected light of each order is detected by executing the procedure of FIG. 12 for the received light amount data obtained in ST1. In the next ST3, the detected surface reflected light intensity is further narrowed down to the 0th and 4th to 15th order reflected light intensities. In this example, only 0th-order and positive-order diffracted light is used for measurement.

  Next, in ST4, theoretical value data relating to a predetermined arrangement pattern is selected from the above-described determination table, and processing for reading out the intensity values of the 0th order and 4th to 15th orders of diffracted light is executed.

  In ST5, an error σ between the received light amount data and the theoretical value data is calculated using each intensity narrowed down in ST3 and the theoretical intensity read in ST4.

  Hereinafter, for each array pattern registered in the determination table, the error σ is obtained in order by executing the processes of ST4 and ST5. When the error σ is obtained for all the registered arrangement patterns, ST6 becomes “YES”, and in the next ST7, the arrangement pattern when the error σ is minimized is specified as the structure to be measured (however, Confirm that the minimum error σ is within a predetermined value.)

  Thereafter, in ST8, the width d and height h of the electrode according to the specified periodic pattern are output to the outside as measurement results.

  The processing performed by the CPU 21 as an optical measuring device is as described above. Furthermore, the CPU 21 can also perform processing for determining whether the electrode arrangement pattern is good or bad from the measurement result.

  In the above measurement process, the first to third order diffracted lights are set to be excluded from the measurement process in advance. However, when the substrate 1 to be measured changes variously, it depends on the configuration of the substrate 1. Therefore, it is desirable that the order excluded from the measurement process can be appropriately selected. For example, prior to the measurement process, each possible numerical range is input for the period p of the substrate 1 and the period q of the surface roughness, and based on these input values, a numerical range where the value of p / q can be obtained is obtained. It is possible to set so that the orders included in the range are not used for the measurement process.

  According to the procedure of FIG. 16 above, the component of the back surface reflected light related to the first noise is excluded, and the first to third order surface reflected light related to the second noise is excluded, and then the theoretical data and Since the comparison is performed, highly accurate measurement can be performed on the periodic pattern of the electrodes.

Note that the optical measurement apparatus described above can measure not only a rectangular structure but also other shapes such as a trapezoid.
FIG. 17 shows a specific example when the trapezoidal structure 2 </ b> A is a measurement target. In this example, the trapezoid is considered to approximate a shape in which a plurality of rectangles are stacked in a staircase pattern. In this way, the theoretical intensity of each diffracted light can be obtained by repeating the calculation based on the vector diffraction theory for each rectangle (see Non-Patent Document 1).

TK Gayload and MG Moharam, “analysis and applications of optical diffraction by gratings”, Proc. IEEE, 73, 894-937, 1985

  Therefore, also in this case, the theoretical data including the theoretical value of the intensity of the diffracted light for a plurality of orders is created for each of the plurality of periodic array patterns, and the structure 2A is approximated by executing the procedure of FIG. The shape can be determined.

It is a figure which shows the various parameters in the case of calculating the intensity | strength of diffracted light on the assumption that a board | substrate is flat. It is a figure which shows the various parameters in the case of calculating the intensity | strength of diffracted light in consideration of the surface roughness of a board | substrate. It is a figure which shows the external appearance and usage example of the optical measuring device concerning this invention. It is a block diagram which shows the electric constitution of an optical measuring device. It is a figure which shows the structure of a glass substrate with the parameter and measuring method of a measuring object. It is a figure which shows the main structure of a light projection part and a light-receiving part with an effect | action. It is a figure which shows the advancing state of irradiation light and reflected light. It is a figure which shows the relationship of the preferable distribution of front surface reflected light and back surface reflected light. It is a figure which shows the conditions required in order to obtain the distribution curve of FIG. It is the figure which showed ax and bx of FIG. 7 by the optical path of reflected light. It is a flowchart which shows the procedure of the intensity | strength detection process of surface reflected light. It is a figure which shows the measuring method of intensity | strength. It is a graph which shows the result of having measured the change of the surface shape of an actual board | substrate, and the result of having converted the measurement result into the spatial frequency. (5) It is a graph which shows the intensity | strength of each diffracted light calculated based on (10) Formula correspondingly. It is a graph which shows the relationship between the measurement result of the diffracted light from an actual board | substrate, and the theoretical value data corresponding to these measurement results. It is a flowchart which shows the procedure of a measurement process. It is a figure which shows the specific example in case a trapezoid structure is made into a measuring object.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Board | substrate 2 Electrode 10 Sensor head 11 Light projection part 12 Light reception part 20 Controller 21 CPU
22 Memory 122 CCD

Claims (10)

A light projecting means for irradiating light on a substrate on which a structure having periodicity is arranged; a light receiving means for separating and receiving a reflected wave of diffracted light generated on the substrate by the irradiated light for each order; and the light receiving means In the apparatus comprising the measuring means for determining the shape of the structure by comparing the intensity of each diffracted light obtained by the above with the theoretical intensity,
The optical measurement apparatus is configured so that the measurement unit excludes a predetermined-order diffracted light selected in advance from the comparison target. The apparatus of claim 1, wherein
A memory in which theoretical intensities of a plurality of orders of diffracted light excluding the preselected orders are registered for each shape data of the structure is provided, and the measuring means is separated by the light receiving means. An optical measurement apparatus that executes the comparison process on the diffracted light having the theoretical intensity registered. A light projecting means for irradiating light on a substrate on which a structure having periodicity is arranged; a light receiving means for separating and receiving a reflected wave of diffracted light from the substrate with respect to the irradiated light; and In the apparatus comprising a measuring means for determining the shape of the structure by comparing the intensity of each separated diffracted light with the theoretical intensity,
A selection means for selecting an order affected by a noise component caused by a period q approximating the surface roughness of the substrate among the diffracted light of each order;
The optical measuring apparatus configured to exclude the diffracted light of the order selected by the selecting means from the diffracted lights separated by the light receiving means from the comparison processing target. The apparatus of claim 3, wherein
An input means for inputting a range that can be taken by each of the period q approximating the surface roughness and the period p of the structure on the substrate to be measured is provided, and the selection means is input from the input means. An optical measuring device that selects the order affected by the noise component using the periods p and q.   The optical measuring apparatus according to claim 4, wherein the selection unit is set to select an order corresponding to a value of p / q. The apparatus of claim 3, wherein
An input means for inputting sample data of the intensity distribution of the diffracted light obtained using the light projecting means and the light receiving means for the substrate on which the structure having periodicity is arranged, and the selecting means comprises a plurality of An optical measurement device that obtains the variation in the intensity of diffracted light between the sample data for each order and selects the order in which a variation of a predetermined value or more occurs as the order affected by the noise component. Using a liquid crystal substrate in which structures are arranged with a period of 300 to 400 μm as a measurement target, a projection unit that irradiates light to the substrate and a reflected wave of diffracted light from the substrate with respect to the irradiation light are separated for each order. In an apparatus comprising: a light receiving means for receiving light; and a measuring means for determining the shape of the structure by comparing the intensity of each diffracted light separated by the light receiving means with the theoretical intensity.
The optical measurement apparatus is configured such that the measurement unit excludes at least the first to third order diffracted light from the object of the comparison process. Using a liquid crystal substrate in which structures are arranged with a period of 400 to 500 μm as a measurement object, a projection means for irradiating the substrate with light and a reflected wave of diffracted light from the substrate with respect to the irradiation light are separated for each order. In an apparatus comprising: a light receiving means for receiving light; and a measuring means for determining the shape of the structure by comparing the intensity of each diffracted light separated by the light receiving means with the theoretical intensity.
The optical measurement apparatus is configured so that the measurement unit excludes at least 1-4th order diffracted light from the object of the comparison process. The reflected wave of the diffracted light generated by irradiating the substrate on which the periodic structure is arranged is separated and received for each order, and the intensity of each diffracted light is compared with the theoretical intensity. In the method for obtaining the shape of the structure,
A method for measuring a periodic array pattern, wherein the orders affected by a noise component caused by the surface roughness of the substrate among the diffracted lights of each order are removed from the comparison target. The method according to claim 9, wherein
A method for measuring a periodic array pattern, in which an order affected by a noise component due to the surface roughness is obtained based on a numerical range that can be taken by the period q approximating the surface roughness and the period p of the structure on the substrate. .

Images