JP4886169B2 - Mask, design method thereof, exposure method, and device manufacturing method - Google Patents

Description

  The present invention generally relates to exposure, and in particular, a semiconductor chip such as an IC or LSI, a display element such as a liquid crystal panel, a detection element such as a magnetic head, an imaging element such as a CCD, a mask used in micromechanics, and its mask The present invention relates to a manufacturing method, an exposure apparatus and method, a device manufacturing method, and a device manufactured from an object to be exposed. Here, micromechanics refers to a micron-scale mechanical system with advanced functions and a technology for making it by applying semiconductor integrated circuit manufacturing technology to the fabrication of fine structures.

  2. Description of the Related Art When a device is manufactured using a photolithography technique, a projection exposure apparatus that projects a pattern drawn on a mask (reticle) onto a wafer by a projection optical system and transfers the pattern has been conventionally used. The projection optical system causes the diffracted light from the pattern to interfere and form an image on the wafer, and in normal exposure, the 0th and ± 1st order diffracted lights (that is, three light beams) from the pattern are interfered.

  Mask patterns include adjacent periodic line and space (L & S) patterns, adjacent and periodic contact hole rows (that is, arranged at the same level as the hole diameter), isolated contact holes isolated without being adjacent to each other, etc. In order to transfer a pattern with high resolution, it is necessary to select an optimal exposure condition (such as illumination condition and exposure amount) according to the type of pattern.

  The resolution R of the projection exposure apparatus is given by the following Rayleigh equation using the wavelength λ of the light source and the numerical aperture (NA) of the projection optical system.

Here, _{k 1} is a constant determined by such developing process, in the case of normal exposure is _{k 1} is about 0.5 to 0.7.

In response to the recent high integration of devices, there is an increasing demand for miniaturization of transferred patterns, that is, higher resolution. In order to obtain high resolution, it is effective to increase the numerical aperture NA from the above equation and to reduce the wavelength λ. However, these improvements have reached the limit at this stage, and in the case of normal exposure. It is difficult to form a pattern of 0.15 μm or less on the wafer. Therefore, a phase shift mask technique that interferes and forms an image of two light beams out of the diffracted light that has passed through the pattern has been proposed (for example, see Patent Document 1). In the phase shift mask, the phase of adjacent light transmitting portions of the mask is inverted by 180 ° to cancel the 0th-order diffracted light, and the two ± 1st-order diffracted lights are interfered to form an image. According to this technique, it is possible to a k ₁ of the above equation substantially 0.25, it is possible to form the following pattern 0.15μm on the wafer to improve the resolution R.
US Patent Application Publication No. 2002/177048

  However, in the case of a contact hole close to the resolution limit, if the adjacent phase is changed by 180 degrees, the diffracted light is diffracted at a large angle from the optical axis in the direction of 45 degrees on the pupil plane. Jumps out and cannot pass through the pupil of the projection lens and is not resolved. The resolution can be up to a fine pattern of √2 times the limit line width of L & S.

  In the recent semiconductor industry, production is shifting to higher-value-added system chips that include a wide variety of patterns, and it is necessary to mix multiple types of contact patterns in the mask, which is equivalent to the resolution of the L & S pattern. The contact hole of the resolution is also needed. However, only the conventional phase shift mask technology cannot simultaneously expose a contact hole pattern in which contact hole arrays and isolated contacts are mixed with high resolution. On the other hand, it is conceivable to use double exposure (or multiple exposure) in which different types of patterns are separately exposed using two masks, but conventional double exposure uses two or more masks. Therefore, the cost is increased, the throughput is lowered due to the exposure twice, and the high overlay accuracy of the exposure twice in the mask exchange is required. Therefore, there are many problems to be solved in practice.

  Therefore, a mask having a fine hole diameter (for example, 0.15 μm or less) and capable of exposing a pattern having one or a plurality of holes at high resolution without exchanging the mask, a manufacturing method thereof, an exposure method, and It is an exemplary object of the present invention to provide an apparatus.

A mask according to an aspect of the present invention is a mask having a plurality of contact hole patterns and a plurality of auxiliary patterns having dimensions smaller than the plurality of contact hole patterns, and the mask is a binary mask or a halftone mask. The phases of the plurality of contact hole patterns and the auxiliary patterns are the same, and the plurality of auxiliary patterns include a first auxiliary pattern and a plurality of second auxiliary patterns, center and the center of the first auxiliary pattern is arranged at equal intervals on a straight line, the center of the plurality of second auxiliary patterns is equal from the center of two adjacent patterns of the plurality of contact holes pattern situated, the plurality of second auxiliary patterns, the center of the straight line and the second auxiliary pattern Equal distance of the cycle of the plurality of contact holes, characterized in that it is arranged in said plurality of contact holes the bilateral periodic with the same period of the linear and the pattern line and on a straight line parallel.

  Further objects and other features of the present invention will become apparent from the preferred embodiments described below with reference to the accompanying drawings.

  According to the present invention, there is provided a mask capable of exposing a pattern having a fine hole diameter and having one or a plurality of contact holes at a high resolution without exchanging the mask, a manufacturing method thereof, and an exposure method and apparatus. Can be provided.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. Hereinafter, unless otherwise specified, the exposure apparatus is assumed to have a light source of KrF excimer laser (exposure wavelength λ = 248 nm), a numerical aperture of its projection optical system of 0.73, and a reduction ratio of 4: 1. A reduction exposure apparatus is used. The projection exposure apparatus is generally reduced projection exposure. In the case of reduced projection exposure, the pattern size to be created and the mask pattern differ by a magnification that depends on the exposure apparatus. Since the magnification of the exposure apparatus varies with each model, in the following, the pattern size on the mask is converted into the dimension on the wafer. For example, when the projection exposure apparatus has a magnification of 0.25, when it is desired to create a 120 nm pattern, a pattern of 480 nm must actually be created on the mask. In the case of 20 times, a 600 nm pattern must be created on the mask. However, in the following, in order to eliminate the distinction between these situations, the size of the mask pattern is converted into a dimension on the wafer and referred to as a 120 nm pattern. Each pattern is composed of one or a plurality of contact holes. In the present application, the term “pattern” may mean a part of a pattern or one contact hole.

  The present inventors have arranged a small contact hole around the desired pattern of the binary mask by arranging an auxiliary pattern that is smaller than the desired pattern and large enough not to be resolved, and performing special oblique incidence illumination. The pattern has already been successfully exposed on an object to be exposed such as a wafer. As an example, in the desired pattern as shown in FIG. 3C, it is assumed that the hole diameter is 100 nm and the minimum hole interval is 100 nm.

  When the above-described projection exposure apparatus is used, normally, as schematically shown in FIG. 3A, a circular shape is used by using a mask in which a light transmitting portion 31 and a light shielding portion 33 are arranged according to a desired pattern. Although the wafer is exposed by illuminating the mask with illumination having an effective light source with an illuminance distribution, the pattern is not resolved by this method. On the other hand, as schematically shown in FIG. 3B, a mask in which the minute light-transmitting portion 32 is arranged around the light-transmitting portion 31 with respect to the desired pattern in the light-shielding portion 33, and FIG. As described above, when the wafer was exposed by an illumination system having an effective light source distribution having a cross-shaped light shielding portion 42A, the pattern was resolved as shown in FIG.

  The mask shown in FIG. 3B is a desired pattern 31 and auxiliary patterns that are periodically arranged in the vertical and horizontal directions of the desired pattern 31 and have dimensions smaller than the desired pattern 31 and that do not resolve. 32 is a binary mask. In FIG. 4A, a white portion 41 represents a light irradiation portion, and a portion 42 filled with black represents a light shielding portion. The wafer was a silicon base, and the film thickness was 350 nm using TOK-DP746HC as a resist.

  Such an exposure method is to form a mask in which a desired contact hole pattern and a dummy contact hole pattern having a hole diameter smaller than the hole diameter of the pattern are arranged, and to resolve only the desired contact hole pattern portion. Yes (this application may be referred to as exposure method I in this application). Further, the present inventors can obtain the same result even when the illumination as shown in FIGS. 4B, 4C, and 4D is used as the effective light source shape used in the exposure method I. It was confirmed.

  The basic idea of the exposure method I is to arrange auxiliary patterns periodically. For example, when it is desired to expose a contact hole pattern as shown in FIG. 5A, it is conceivable to insert auxiliary patterns vertically and horizontally in a desired pattern as shown in FIG. 5B. The exposure method has a greater depth of focus than a normal exposure method, and requires a smaller amount of exposure for pattern formation, and therefore has a high so-called throughput.

  The inventors of the present invention have further studied the above-described exposure method in order to further improve the resolving power, increase the depth of focus, and improve the throughput. As a result, the mask in which the auxiliary pattern is arranged as shown in FIG. Has been found to be preferred. The mask shown in FIG. 5C has several features as follows. First, assuming an orthogonal virtual lattice having an intersection on the center of a desired contact hole pattern, the center of the auxiliary pattern exists at a position different from the orthogonal virtual lattice intersection. Next, the closest contact hole among the auxiliary patterns arranged obliquely as viewed from the desired contact hole on the orthogonal virtual lattice is between 0 ° and 45 ° (greater than 0 °, 45 ° as will be described later). Smaller). Further, auxiliary patterns having dimensions smaller than the pattern are arranged shifted from the intersection of the virtual lattices in every other row or every other column in the virtual lattice that is faithful to the cycle of the desired contact hole pattern.

  It is quite natural to insert auxiliary patterns above, below, left and right of a desired pattern. Although the auxiliary pattern itself is not resolved, the resolution is improved because the auxiliary pattern affects the desired pattern. In view of this, it is most natural to arrange the auxiliary pattern closest to the desired pattern, that is, vertically and horizontally. However, the present inventors have found that it is possible to increase the depth of focus by arranging the auxiliary patterns alternately rather than in the vertical and horizontal directions. That is, for the pattern as shown in FIG. 5 (a), the depth of focus is greater when the auxiliary pattern is arranged as shown in FIG. 5 (c) than when the auxiliary pattern is inserted as shown in FIG. 5 (b). I found it big.

  Hereinafter, this principle will be described. First, the terms “normal arrangement” and “alternate arrangement” are defined. “Normal arrangement” means that, as schematically shown in FIG. 6A, the auxiliary pattern 62 is arranged at each grid point in consideration of the virtual grid 63 faithfully to the cycle of the desired pattern 61. On the other hand, the “alternate arrangement” means that every other line in the virtual lattice 63 faithful to the period of the desired pattern 61 is shifted to the right or left as shown schematically in FIG. An auxiliary pattern is arranged. Similarly, it is an alternate arrangement to arrange auxiliary patterns at positions shifted up or down every other column in a virtual lattice that is faithful to the cycle of the desired pattern.

  Consider a case in which the pattern of the mask 90A shown in FIG. 9A and the mask 90B shown in FIG. In the mask 90A, holes of 120 nm are arranged in 5 rows and 5 columns with a space of 120 nm. In the mask 90B, the second and fourth hole columns of the mask 90A are laterally shifted by 120 nm.

The light intensity distribution on the pupil plane of the projection optical system when the masks 90A and 90B are illuminated coherently are as shown in FIGS. 10 (a) and 10 (b), respectively. Coordinate values shown in FIG. 10 (a) and FIG. 10 (b) is normalized _{(k 1} equivalent) in lambda / NA. In FIGS. 10A and 10B, a circle having a radius of 1 represents the pupil, and the state in which the diffracted light is shifted and taken into the pupil by appropriate oblique illumination from the vertical and horizontal directions is also shown. It represents at the same time.

  Since the appearance of the diffracted light also changes depending on the mask pattern, the way the diffracted light enters the pupil also changes. When the mask 90A is used, when oblique incidence illumination is performed, the diffracted light 102 enters the pupil 101 only as shown in FIG. 10 (c) or FIG. 10 (d). On the other hand, when the mask 90B is used, in the case of oblique incidence illumination, the three diffracted lights 102 enter the pupil 101 without being aligned on a straight line as shown in FIG. 10 (e) or FIG. 10 (f). It is understood. Such a difference in the distribution of diffracted light within the pupil causes a decisive difference in imaging performance.

  First, the difference in resolution in the vertical direction will be described. In the mask 90A, there are mainly two types of interference that determine the resolution in the vertical direction. That is, the interference between the diffracted light 101a having the intensity 1.00 and the diffracted light 102a having the intensity 0.41 and the interference between the diffracted light 101a having the intensity 1.00 and the diffracted light 103a having the intensity 0.41. On the other hand, in the mask 90B, there are mainly four types of interference that determine the resolution in the vertical direction. That is, the interference between the diffracted light 101b having an intensity of 1.00 and the diffracted light 102b having an intensity of 0.41, the interference between the diffracted light 101b having an intensity of 1.00 and the diffracted light 103b having an intensity of 0.41, and the diffracted light 104b having an intensity of 0.33. Interference between the diffracted light 105b having an intensity of 0.33, and interference between the diffracted light 106b having an intensity of 0.33 and the diffracted light 107b having an intensity of 0.33. As described above, when the vertical resolution of the masks 90A and 90B is examined, it is understood that the mask 90B has a larger amount of light contributing to the resolution.

  Next, the difference in resolution in the horizontal direction will be described. In the mask 90A, there are mainly two types of interference that determine the resolution in the horizontal direction. That is, the interference between the diffracted light 101a having the intensity 1.00 and the diffracted light 104a having the intensity 0.41 and the interference between the diffracted light 101a having the intensity 1.00 and the diffracted light 105a having the intensity 0.41. On the other hand, in the mask 90B, there are mainly four types of interference that determine the resolution in the vertical direction. That is, the interference between the diffracted light 101b having the intensity 1.00 and the diffracted light 104b having the intensity 0.33, the interference between the diffracted light 101b having the intensity 1.00 and the diffracted light 105b having the intensity 0.33, and the diffracted light 101b having the intensity 1.00. Interference between the diffracted light 106b having an intensity of 0.33, and interference between the diffracted light 101b having an intensity of 1.00 and the diffracted light 107b having an intensity of 0.33. From the above, even if the lateral resolution of the masks 90A and 90B is examined, it can be seen that the mask 90B contributes more light to the resolution.

  Next, pattern formation based on mask differences will be described. First, the pattern formation by the mask 90A is shown in FIG. The interference generated from the diffracted light 101a and the diffracted light 102a, or the interference generated from the diffracted light 101a and the diffracted light 103a forms the light intensity distribution 103g and contributes to the pattern formation in the vertical direction. The interference generated from the diffracted light 101a and the diffracted light 104a, or the interference generated from the diffracted light 101a and the diffracted light 105a forms the light intensity distribution 102g and contributes to the formation of the pattern in the horizontal direction. Since the light intensity distribution 102g and the light intensity distribution 103g have sinusoidal intensity distributions, it is understood that the pattern 101g is roughly diamond-shaped when they overlap each other.

  Next, the pattern formation by the mask 90B is shown in FIG. Interference generated from diffracted light 101b and diffracted light 102b, interference generated from diffracted light 101b and diffracted light 103b, interference generated from diffracted light 104b and diffracted light 105b, or interference generated from diffracted light 106b and diffracted light 107b Forms a light intensity distribution 102h and contributes to vertical pattern formation. The interference generated from the diffracted light 101b and the diffracted light 105b, or the interference generated from the diffracted light 101b and the diffracted light 106b shows the light intensity distribution 103h, the interference generated from the diffracted light 101b and the diffracted light 104b, or the diffracted light 101b and the diffracted light. The interference generated from 107b forms a light intensity distribution 104h, and thus contributes to the resolution in the oblique direction. Since the light intensity distribution 102h, the light intensity distribution 103h, and the light intensity distribution 104h have sinusoidal intensity distributions, the patterns 101h are roughly hexagonal when they overlap each other.

  Originally, it is desirable that the contact hole has a square shape faithful to the mask pattern. However, due to the nature of exposure, the corners are rounded to form a circular hole. Thus, the phenomenon that the actual hole shape deviates from the square is unavoidable to some extent. However, as shown in FIG. 10C, when a diamond-shaped hole is formed, the area of the hole is considerably smaller than a square or circular hole originally expected. On the other hand, as shown in FIG. 10 (d), it is understood that the hexagonal hole has a larger area than the rhombus, and is more preferable for approaching the originally required hole area.

  From the above results, it can be seen that it is better to insert the auxiliary patterns alternately instead of inserting the auxiliary patterns only in the vertical and horizontal directions of the desired pattern.

When the auxiliary pattern is arranged in this way, it is effective to use illumination with a dark center when the mask is a binary mask or a halftone mask. Illumination with a dark central portion means annular illumination shown in FIG. 4 (e), quadrupole illumination shown in FIG. 4 (f), and illumination distribution as shown in FIGS. 4 (a) to (d). It is the illumination which forms the effective light source which has. In particular, when the illumination described in FIGS. 4A to 4D is used, there are bright portions in the 0 degree direction and the 90 degree direction. That is, when the shape of the effective light source distribution is converted to σ, a portion of about 1 / (4 × α) is bright in the 0 degree direction and the 90 degree direction. Here, α is a value obtained by converting the half period of the virtual lattice into k ₁ (= half period × NA / λ), and “about” mentioned here is a range of ± 0.05 in terms of σ. It is. For a phase shift mask in which the phase difference between adjacent holes is 180 degrees, so-called small σ illumination may be used.

  If a virtual lattice is obtained in the present embodiment, auxiliary patterns can be inserted in a staggered arrangement. Hereinafter, a method for determining a virtual lattice will be described with reference to FIG. Here, FIG. 23 is a flowchart for explaining a method of determining a virtual lattice.

  In many cases, the resolution of the projection exposure apparatus changes with the numbers 0.25 and √2. Therefore, in FIG. 23, g1 is 0.25 or more and 0.25 × √2 or less, g2 is 0.25 × √2 or more and 0.5 or less, g3 is 1.0 or more and √2 or less, and g4 is 0.5 ×. √2 to 1.0 and g5 may be 0.25 to 0.25 × √2 or less. In addition, the experimental result of FIG. 8 is a thing when the value of g1, g2, g3, g4, and g5 is 0.29, 0.40, 1.20, 0.80, and 0.25, respectively. In order to set a virtual grid for an isolated pattern, the isolated pattern may be placed on an appropriate virtual grid and arranged in a staggered manner. Basically, the half period of the virtual lattice should be the same as that of the isolated pattern, but if there is a periodic pattern at a little distance, it may be adjusted to that period.

First, the hole diameter D of a desired pattern k ₁ obtained by converting (i.e., D × NA / lambda) to determine whether it is less than the first threshold value g1 (step 1202). If step 1202 determines that this is the case, the process ends abnormally (step 1204). On the other hand, if step 1202 determines that this is not the case, it is determined whether the hole diameter of the desired pattern is not less than the first threshold value g1 and not more than the second threshold value g2 (step 1206). Here, the period of the periodic pattern is P1.

Step 1206, if it is determined that the first threshold value g1 or more and a second threshold value g2 below, after standing and P2 = P1, the process proceeds to "PartI", the third threshold value when further the P2 _{k 1} converted It is determined whether or not g3 or more (step 1210).

  When step 1210 determines that it is not greater than or equal to the third threshold g3, the period of the virtual lattice is determined as P2 (step 1212). If it is determined that step 1210 is greater than or equal to the third threshold value g3, i = i + 1 is set (step 1214), P1 is divided into i (P12) (step 1216), and the process returns to step 1210. Finally, the period of the virtual lattice is determined as P2 (step 1212).

If the step 1206 determines that not less than the first threshold value g1 equal to or higher than the second threshold g2, after standing and P2 = P1, the process proceeds to "PartII", further the fourth threshold value g4 or more when the P2 _{k 1} converted Is determined (step 1220). Step 1220 If it is determined that the fourth threshold value g4 or greater, it is determined whether or not those k ₁ in terms of the value obtained by subtracting the hole diameter D of P2 is the fifth threshold g5 or less (Step 1222 Otherwise, treat as an isolated pattern or go to “Part I” (step 1226). On the other hand, when it is determined that step 1120 is not equal to or greater than the fourth threshold value g4 or when step 1222 is determined to be equal to or less than the fifth threshold value g5, the period of the virtual lattice is determined to be P2 (step 1224).

  In the virtual lattice thus obtained, auxiliary patterns are arranged at positions shifted to the right or left in every other row, or similarly, every other row in the virtual lattice that is faithful to the cycle of the desired pattern. If an auxiliary pattern is arranged at a position, an alternate arrangement can be created. The shift amount at this time may be not less than 1/6 and not more than half of the period of the virtual lattice. This point will be introduced later in Examples. Although it has been confirmed that the auxiliary pattern size should be within 50% to 90% of the desired pattern, it is usually about 75%.

  Thus, a very effective auxiliary pattern insertion method has been clarified. Hereinafter, an auxiliary pattern insertion method and a lighting condition determination method suitable for the method will be described. Here, FIG. 24 is a flowchart for explaining a method of setting a mask pattern and illumination conditions.

  Referring to FIG. 24, first, the transmittance where there is no desired pattern is set to 0 and the transmittance where there is a desired pattern is 1 according to the pattern to be formed after exposure, and the corresponding desired pattern data (Dpd). ) Is created (step 1402). Here, the dimension and arrangement of a desired pattern to be exposed on the wafer are determined.

  Next, a mask type (for example, binary mask, halftone mask, phase shift mask, etc.) used after setting a desired pattern is determined (step 1404).

  Here, as schematically shown in the sectional view of FIG. 2C, the halftone mask attenuates the light intensity of the portion corresponding to the light-shielding portion of the binary mask, and the phase difference from the light-transmitting portion. The mask is changed to a substance 24 that maintains 180 degrees. Halftone masks are more difficult to create than binary masks, but are still used in exposure apparatuses. A phase shift mask is a mask designed to have a predetermined phase difference when light incident on the mask passes through adjacent light transmitting parts. Design to the degree. FIG. 2A schematically shows a cross-sectional structure of the phase shift mask, and reference numeral 23 denotes a phase shifter introduced in order to set the phase difference between adjacent light transmitting portions to 180 degrees. Various types of phase shift masks have been proposed. Depending on the type of phase shift mask, a line width approximately half that of a binary mask can be realized. This is because the light adjacent to each other cancels out the amplitude of the central part of the light adjacent to each other by maintaining the phase difference of 180 degrees.

  Next, an auxiliary pattern (dummy contact hole pattern) to be inserted and an illumination method are determined for the desired pattern data (step 1406).

  The size of the auxiliary pattern is the first step of extracting necessary dummy contact hole pattern data (Dum) based on Dpd and creating mask data (Fpd) suitable for the exposure method I, based on logical operation. A second step of creating data and creating Fpd, and a third step of determining whether Fpd satisfies the mask pattern design rules are included. Usually, the steps are repeated as necessary in the order of the first step or the second step or both steps and the third step.

  The present invention can be applied regardless of whether the desired pattern is a periodic pattern or an isolated pattern. Here, the “periodic pattern” is a pattern having at least two contact holes aligned in any one of two orthogonal directions. An “isolated pattern” is a pattern that includes only one contact hole and does not have other contact holes that are aligned in any of two orthogonal directions.

  In determining the lighting method, calculation is performed, the lighting conditions are checked, and if the lighting conditions are created within the predetermined design rule, the process is terminated. If the lighting conditions are not created within the predetermined design rule, the calculation step is performed. Returning to is repeated a predetermined number of times. If the creation of the lighting condition is not determined to be acceptable within the predetermined number of times, the process ends as an abnormality. The alternative lighting method is determined by pulling out the database (table data), checking the lighting conditions, and if the lighting conditions are created within the predetermined design rules, the lighting conditions are created within the predetermined design rules. If not, returning to the calculation step is repeated a predetermined number of times. If the creation of the lighting condition is not determined to be acceptable within the predetermined number of times, the process ends as an abnormality.

  Step 1406 is also used when correcting the auxiliary pattern insertion method and / or illumination method when the processing is fed back from step 1422 described below.

  Next, the desired pattern is checked (step 1408). Here, based on the mask pattern data created by inserting the auxiliary pattern into the desired pattern and the illumination condition data, it is determined whether the desired pattern is accurately formed. That is, step 1408 determines whether only the desired pattern is resolved with high accuracy without resolving the auxiliary pattern. The degree of accuracy is determined according to a certain standard, but may be determined by the user. If there are a plurality of auxiliary patterns or illumination condition candidates for resolving only a desired pattern, it is preferable to select the one having the larger contrast and the smaller error variation in the line width (CD: critical dimension).

  If step 1408 determines that the desired pattern is not resolved, correction is made (step 1412). In the correction, a desired pattern, auxiliary pattern, etc. are corrected. The auxiliary pattern and other corrections are mainly performed in step 1406 fed back in step 1422 described later, but may be performed in step 1412 if fine adjustment is performed.

  The correction of a desired pattern will be described in optical proximity correction and Example 1 described later. Optical proximity correction (hereinafter referred to as OPC: Optical Proximity Correction) is a technique for enabling transfer of a desired pattern with high accuracy.

  For example, consider OPC when a pattern to be transferred is created smaller than a desired size. In FIGS. 25 (a) and 25 (b), the dotted line has a desired hole diameter, but a pattern smaller than the desired size can be formed due to the exposure amount with other patterns, so FIGS. 25 (a) and 25 (b). The mask pattern may be changed as indicated by the solid line portion in FIG. The OPC technique shown in FIG. 25A is effective when the pattern is sparse, and the OPC technique shown in FIG. 25B is effective when the pattern is dense and it is difficult to enlarge the mask pattern. However, there is a drawback that the mask data becomes large.

  In addition, changing the dimension and shape of the auxiliary pattern also leads to correction of a desired pattern. For example, when the dimension of the desired pattern is smaller than the desired value, the dimension of the auxiliary pattern arranged around the desired pattern is increased, or the arrangement period (pitch) is reduced. On the other hand, when the dimension of the desired pattern is larger than a predetermined value, there are methods such as reducing the dimension of the auxiliary pattern arranged around the desired pattern or increasing the arrangement period (pitch).

  Even if the number of holes in the auxiliary pattern arranged around the desired pattern is changed, the desired pattern can be corrected. For example, the amount of light of the desired pattern can be reduced by reducing the number of holes in the auxiliary pattern arranged around the desired pattern, and the amount of light of the desired pattern can be increased by increasing the number of holes in the auxiliary pattern.

  Even if the illumination conditions are changed, the desired pattern can be corrected. For example, in the binary mask, when illumination that forms an effective light source having the illuminance distribution shown in FIGS. 4B to 4D is used, the depth of focus can be made deeper than that in FIG. For other corrections, the minimum pitch, the type of mask, the threshold at which the photoresist applied to the exposure object is exposed, and the coherence factor σ of the effective light source may be changed. For example, since the phase shift mask has the effect of extending the depth of focus or reducing the line width error variation somewhat, it is also effective to change the phase shift mask when the binary mask has insufficient depth of focus.

  After correction, the desired pattern is checked again (step 1414). Similar to the check in step 1408, it is determined whether only the desired pattern is resolved with high accuracy without resolving the auxiliary pattern. If the check in step 1414 still does not pass, the process of returning to step 1412 is repeated a predetermined number of times (kmax) (steps 1416 and 1418). When the predetermined number of times kmax is exceeded, the process of returning to step 1406 is repeated for the predetermined number of times jmax (steps 1420 and 1422).

  If the check in step 1408 or 1414 does not pass, the process ends as abnormal (step 1424). If the check in step 1408 or 1414 is finally passed, mask pattern data and illumination conditions are determined (step 1410).

  Since most of the methods shown in FIG. 24 can be executed by a computer, the creator only creates and inputs a pattern to be finally formed on the resist, and the subsequent generation of mask pattern data and illumination conditions is as described above. Since the process can be automatically performed by a computer, an optimum mask pattern and illumination conditions can be efficiently created even in designing a large-scale semiconductor integrated circuit. Even if a large amount of mask data is not processed at once, the mask pattern data can be divided and processed, and the method of combining them at the end is also convenient, which is convenient for a computer.

  These principles can be applied to various patterns. Examples thereof will be clarified in the following examples.

  A projection optical image for an ArF excimer laser (exposure wavelength 193 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.70 and a reduction ratio of 4: 1 according to the present embodiment will be described. Consider exposing a pattern having a hole diameter of 100 nm and a hole interval of 100 nm arranged in a line using a binary mask.

  In the mask 110A schematically shown in FIG. 11A, the distance between the desired pattern center and the auxiliary pattern center is 200 nm in both the vertical and horizontal directions, and the distance between the auxiliary pattern center and the auxiliary pattern center is 200 nm in both the vertical and horizontal directions. Suppose that The mask 110B in FIG. 11B shifts the auxiliary pattern in the horizontal row of the first row and the auxiliary pattern in the horizontal row of the third row by 100 nm to the right of the mask 110A compared to the mask 110A. The auxiliary pattern on the right side is deleted, and the auxiliary patterns are arranged alternately.

  Since it is a binary mask, the results when the illumination having a schematic shape as shown in FIG. 4A is used for the masks 110A and 110B are FIG. 11C and FIG. 11D, respectively. From a comparison between FIG. 11C and FIG. 11D, it can be seen that when the auxiliary pattern is inserted alternately, the amount of light is stronger on the exposure surface and the depth of focus is larger. In addition, it is understood that the external shape of the desired contact hole 113A becomes a diamond shape when auxiliary patterns are inserted in the vertical and horizontal directions. On the other hand, if auxiliary patterns are inserted alternately, it can be seen that the outer shape of the desired contact hole 113B is close to a circle. This is due to the difference in the distribution of diffracted light that contributes to pattern formation.

  If it is desired to further increase the amount of light, auxiliary patterns for two turns may be inserted around the desired pattern as shown in FIGS. 11 (e) and 11 (f). As shown in FIG. 11 (e), the auxiliary pattern is inserted alternately in the vertical and horizontal directions of the desired pattern because the amount of light is larger and the depth of focus is larger. An auxiliary pattern may be inserted as shown in FIG.

  As described above, a good transfer pattern can be obtained by alternately inserting auxiliary assist patterns. In this embodiment, a binary mask is used as a mask. However, the above effect can be similarly obtained even when a halftone mask is used as a mask.

  In this example, a rectangular contact hole pattern was exposed with a binary mask using a KrF excimer laser (exposure wavelength 248 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.73 and a reduction ratio of 4: 1. Thus, it is shown that the method of arranging the auxiliary patterns alternately can be applied not only to the square contact hole pattern but also to the rectangular contact hole pattern (line pattern).

  The illumination condition is a condition for forming an effective light source having the illuminance distribution shown in FIG. 4A, and the target pattern is a line pattern having a width of 120 nm and a length of 120 × 13 nm. 12A shows a mask 120A in which auxiliary patterns 122 are arranged vertically and horizontally, and FIGS. 12B and 12C show masks 120B and 120C in which auxiliary patterns are alternately arranged. 12A to 12C, virtual lines with a spacing of 240 nm are drawn in the horizontal direction, and the desired pattern 121 and the auxiliary pattern 122 are arranged on the lines. In FIGS. 12B and 12C, the auxiliary pattern 122 is provided. Are staggered.

  As a result of simulation, when the mask 120A was used, the width was 120.0 nm and the length was 1439.8 nm at the best focus. On the other hand, when the mask 120B was used, the width was 120.0 nm and the length was 1443.7 nm at the best focus. That is, it is understood that the pattern reproducibility is better when the auxiliary patterns are alternately arranged at the best focus. Further, when the defocus is 0.3 μm, the mask 120A has a width of 75.9 nm and a length of 1330.8 nm. In contrast, the mask 120B has a width of 78.3 nm and a length of 1342.3 nm. In other words, it can be seen that alternately arranging the auxiliary patterns has an advantage over the normal auxiliary pattern arrangement even at the time of defocusing.

  Furthermore, it should be noted that in the normal arrangement, 36 auxiliary patterns are arranged, whereas in the alternate arrangement, only 32 auxiliary patterns are arranged. When comparing the size of the mask pattern area including the auxiliary pattern, the mask pattern area 123a is 1080 nm × 2040 nm in the normal arrangement, whereas the mask pattern area 123b is as small as 1080 nm × 1920 nm in the staggered arrangement.

  By comparing the pattern reproducibility, the pattern area, and the number of auxiliary patterns, it can be seen that alternate auxiliary pattern arrangements are also effective for line patterns. The masks 120B and 120C have the same effect.

  The same effect applies to short line patterns. For example, when the mask is illuminated with illumination that forms an effective light source distribution as shown in FIG. 4A for a short line pattern, an auxiliary pattern may be inserted as shown in FIG. 13A or 13B. Although it is conceivable, alternate auxiliary pattern arrangement can also be applied in such a case. For example, auxiliary patterns may be arranged as shown in FIG. 13 (c) or (d). The auxiliary pattern interval d1 in FIG. 13C or 13D is determined by the desired pattern size and the auxiliary pattern size. In most cases, it is better to set the value to a quarter or more of the value obtained by dividing the wavelength by the numerical aperture. The present inventors have confirmed that there is no problem even if it is about 1/8 of the value obtained by dividing the wavelength by the numerical aperture if the mask can be produced as designed.

  In the present embodiment, only an example of the auxiliary pattern insertion method into the line pattern is given, and a line pattern longer or shorter than the line pattern of the present embodiment, and only a binary mask as in the present embodiment. In addition, the present invention can be applied to a halftone mask.

  As described with reference to FIG. 9, the imaging performance is improved by making the pattern arrangement staggered due to the difference in the appearance of the diffracted light. In this embodiment, a binary mask is taken as an example to examine how much the pattern arrangement changes to change the appearance of diffracted light.

  Assume that a KrF excimer laser (exposure wavelength λ = 248 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.73 and a reduction ratio of 4: 1. In the mask pattern as shown in FIG. 14A, holes having a diameter of 120 nm are arranged in 5 rows and 5 columns with a space of 120 nm. On the other hand, in FIG. 14B, the holes in the second and fourth hole columns in FIG. 14A are relatively displaced laterally by d2 compared to the first and third hole columns. Suppose that

  FIG. 15A shows the intensity distribution of illumination light on the pupil plane of the projection system when d2 illuminates the mask with 0 nm, and FIG. 15B shows the projection system when d2 illuminates the mask with 40 nm. Illumination light intensity distribution on the pupil plane, FIG. 15C shows the illumination light intensity distribution on the pupil plane of the projection system when d2 illuminates the mask of 80 nm, and FIG. 15D shows d2 Is the intensity distribution of the illumination light on the pupil plane of the projection system when a 120 nm mask is illuminated. Note that the values of the pupil plane coordinates in FIG. 15 are normalized by λ / NA, and the black circle in the figure represents the diffracted light, and the size of the black circle represents the intensity of the diffracted light.

  When the distribution and intensity of the four types of diffracted light are examined, when d2 is 40 nm or more, the change in the distribution of diffracted light becomes more significant than when d2 is 0 nm, and the amount of light is larger than when d2 is 0 nm. I found out.

  That is, the auxiliary pattern is staggered by shifting the position in a certain direction so that the shift amount is not less than 1/6 and not more than 5/6 of the period p1 with respect to the period of the virtual lattice that can be formed from the desired pattern. By arranging, the imaging performance can be improved as compared with the normal auxiliary pattern arrangement. It is more preferable that the shift amount is not less than one third and not more than two thirds of the period p1, and the shift amount is approximately one half of the period p1 so that the opening of the auxiliary pattern has a grid intersection and an intersection point. It is even better if it exists at the midpoint. A further auxiliary pattern insertion method becomes clear based on the idea of the virtual lattice. That is, when the virtual lattice is taken as an axis, the line connecting the auxiliary pattern center and the desired pattern center is within an angle within a certain range from the axis. This is shown in FIG. In FIG. 26, the hatched portion represents a desired pattern, and the dotted line represents a virtual lattice. As shown in FIG. 26, it can be considered that the angle formed by the line connecting the center of the auxiliary pattern and the center of the desired pattern with the axis of the virtual lattice should be θ1 or more and θ2 or less. However, θ1 is 9 degrees and θ2 is 40 degrees.

  This is independent of the hole size and the exposure apparatus, and the light flux generated by the auxiliary pattern is (-1, 1/2) diffracted light, (-1, -1/2) diffracted light, (1, 1/2) diffracted light, (1, -1/2) diffracted light, (1 / 2,1) diffracted light, (1/2, -1) diffracted light, (-1 / 2,1) diffracted light, It is understood that it is preferable to include at least one of (−1 / 2, −1) diffracted light.

  Here, the definition of the name of the diffracted light is as follows.

  When there is a pattern only in the normal arrangement, that is, in the vertical and horizontal directions, diffracted light appears in a grid pattern according to the pattern period as shown in FIG. The diffracted light that appears at the center of the pupil is called (0,0) -order diffracted light, and is named as the (n, m) -order diffracted light as nth in the x direction and mth in the y direction (where n and m are integers) Turn on.

  In the alternate arrangement, diffracted light appears at a position different from the normal arrangement as shown in FIG. In particular, diffracted light is generated between the (n, m) order diffracted light and the (n, m ± 1) order diffracted light. These diffracted lights are referred to as (n, m ± 1/2) order diffracted lights in this specification. Similarly, the diffracted light generated between the (n, m) -order diffracted light and the (n ± 1, m) -order diffracted light is called (n ± 1/2, m) -order diffracted light.

  As described above, in this embodiment, a binary mask is used as a mask. However, a halftone mask may be used, and the same effect as described above can be obtained.

  Consider the case of a mask pattern having a plurality of patterns in which holes are arranged in a line. For example, as shown in FIG. 16A, there is a pattern 162 in which holes 161 are aligned in the horizontal direction with a hole diameter s and a hole interval s. Further, holes 161 are vertically formed with a hole diameter s and a hole interval s. Suppose that there is a lined pattern 163. The interval between the pattern 162 and the pattern 163 in the x direction is 3 s, and the virtual lattice point formed by the pattern 162 and the virtual lattice point formed by the pattern 163 are shifted by half the lattice point interval in the y direction. Assume that the illumination conditions are such that an effective light source having an illuminance distribution as shown in FIG. 4A is formed using a binary mask or a halftone mask.

In such a case, in the normal auxiliary pattern arrangement, contradiction occurs in the auxiliary pattern insertion (hereinafter referred to as auxiliary pattern conflict). Therefore, in such a pattern arrangement, auxiliary patterns may be inserted alternately into either one of the patterns. An example is shown in FIG. In FIG. 16B, auxiliary patterns are alternately inserted into the pattern 163. Since the light quantity is stronger when the auxiliary patterns are alternately inserted, the light quantity does not match the pattern with the normal auxiliary pattern arrangement. There is. In that case, the area of the peripheral auxiliary pattern having the normal auxiliary pattern arrangement may be slightly increased. This is shown in FIG. In FIG. 16C, there are two types of auxiliary patterns, and the area of the auxiliary pattern 164 is smaller than that of the auxiliary pattern 165. Specifically, for a projection optical image for an ArF excimer laser (exposure wavelength 193 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.70 and a reduction ratio of 4: 1, the hole diameter is 100 nm and the hole interval is 100 nm. If the pattern is arranged in one row, for example, the auxiliary pattern 165 is set to be about 5 to 10 nm larger than the auxiliary pattern 164. What is necessary is just to change the magnitude | size of an auxiliary pattern by the magnitude | size equivalent to 5 to 10% of a desired pattern, and what is necessary is just to enlarge about 10%-30% in terms of area.

  Here, by changing the size of the auxiliary pattern, the variation in size when the desired pattern was formed on the wafer was corrected.

  However, the size variation of the desired pattern on the wafer can be reduced by changing the size of the desired pattern on the mask to various sizes instead of changing the size of the auxiliary pattern. It may be reduced.

  In this case, the size of the desired pattern on the mask on which the normal auxiliary pattern is arranged may be increased by about 1 nm to 7 nm under the exposure conditions. On the contrary, the size of the auxiliary pattern arranged alternately among the desired patterns on the mask may be reduced by about −1 nm to −7 nm. Needless to say, the above-described method of changing the size of the auxiliary pattern may be used in combination.

  Consider the case where a mask pattern is formed by combining patterns in which holes are arranged in a line. Taking FIG. 17A as an example, there are patterns 172, 173, and 174 in which holes 171 are arranged in a line with a hole diameter t and a hole interval t, and the interval between the patterns 173 and 172 in the x direction is 0 and y The interval in the direction is 3t, the interval between the patterns 173 and 174 in the x direction is 0, and the interval in the y direction is 3t. Assume that the illumination conditions are such that an effective light source having an illuminance distribution as shown in FIG. 4A is formed using a binary mask or a halftone mask.

  In such a mask pattern arrangement, an auxiliary pattern conflict occurs when the normal auxiliary pattern arrangement is performed. Therefore, the normal auxiliary pattern arrangement and the alternate auxiliary pattern arrangement may be combined. This is shown in FIG. In FIG. 17B, auxiliary patterns are alternately inserted into the pattern 173, and auxiliary patterns are inserted into the patterns 172 and 174 as usual.

  It should be noted that since the amount of light is stronger when the auxiliary patterns are inserted alternately, the amount of light may not match the pattern with the normal auxiliary pattern arrangement. In that case, the size of the peripheral auxiliary pattern in which the normal auxiliary pattern is arranged may be slightly increased. This is shown in FIG. In FIG. 17C, there are two types of auxiliary patterns, and the area of the auxiliary pattern 175 is smaller than that of the auxiliary pattern 176. Specifically, for a projection optical image for an ArF excimer laser (exposure wavelength 193 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.70 and a reduction ratio of 4: 1, the hole diameter is 100 nm and the hole interval is 100 nm. If it is a desired pattern arranged in one row, for example, the auxiliary pattern 176 is set to be about 5 to 10 nm larger than the auxiliary pattern 175. That is, the size of the auxiliary pattern may be changed by a size corresponding to 5 to 10% of the desired pattern, and the size may be increased by about 10% to 30% in terms of area.

  An application example of alternate auxiliary pattern arrangement will be described with reference to FIGS. It is assumed that the patterns 181 in FIG. 18A are desired patterns and are arranged at the vertices of the virtual quadrangle in the virtual lattice. The length of the virtual rectangle in the x direction is pa, and the length in the y direction is pb. It is assumed that pa and pb are too small to insert an auxiliary pattern, but when the auxiliary pattern is not inserted, the length of light is insufficient compared to other dense patterns. In other words, it is preferable to place auxiliary patterns on or near the intersection of diagonal lines of a virtual quadrangle and to have periodicity in the direction of the diagonal lines. More specifically, when the exposure wavelength is λ and the numerical aperture of the projection optical system is NA, this is applicable when pa and pb are 0.5 × λ / NA or more and 1.0 × λ / NA or less (of course. , Pa and pb are not limited to this range, which is an example).

  In such a case, the auxiliary pattern 182 may be inserted at the midpoint of the oblique direction period of the desired pattern, and the auxiliary pattern may be inserted periodically in that direction. This is an application example of staggered arrangement. When a binary mask or a halftone mask is used, the illumination condition for forming an effective light source having an illuminance distribution as shown in FIG.

  Furthermore, a normal auxiliary pattern arrangement can be added to the pattern. This is shown in FIG. The auxiliary pattern 183 is a pattern in which auxiliary patterns are normally arranged in the vertical and horizontal directions based on a virtual quadrangle pattern, and the auxiliary pattern 182 is provided in a staggered arrangement as described above.

  By arranging the auxiliary pattern as described above, the desired pattern can be rounded and the amount of light can be increased.

  Consider the case where the halls are lined up. FIG. 19A shows how to insert the auxiliary pattern 192 at the right end when many holes form one row. Basically, it is only necessary to insert the auxiliary patterns 192 around the desired pattern 191 alternately. However, the light quantity at the end portion inevitably falls. Therefore, it is better to add more auxiliary patterns only at the pattern end portions than at other portions. The auxiliary patterns to be added should also be arranged in a staggered manner. FIG. 19A shows the mask pattern thus obtained.

  FIG. 19B also shows how to insert the auxiliary pattern 192 at the right end when many holes form one row. Basically, it is only necessary to insert the auxiliary patterns 192 around the desired pattern 191 alternately. However, the light quantity at the end portion inevitably falls. Therefore, for example, in a projection optical image for an ArF excimer laser (exposure wavelength 193 nm) reduction projection exposure apparatus having a numerical aperture NA = 0.70 and a reduction ratio of 4: 1, the hole diameter is 100 nm, the hole interval is 100 nm, and 1 If the pattern is arranged in a row, the auxiliary pattern 193 to be inserted at the pattern end is set to be about 5 to 10 nm larger than the other auxiliary patterns. That is, the size of the auxiliary pattern may be changed by a size corresponding to 5 to 10% of the desired pattern, and the size may be increased by about 10% to 30% in terms of area.

  If necessary, it is possible to accurately form a desired pattern by changing the number of auxiliary patterns inserted depending on the location, changing the size of the auxiliary pattern, or a combination of both. Again, at this time, when a binary mask or a halftone mask is used, oblique incidence illumination having an effective light source distribution with a dark center as shown in FIGS. 4A to 4E is most effective.

  In the exposure method I, it has been found that the illumination conditions should be devised to increase the depth of focus, so a specific shape of the effective light source distribution will be introduced.

  The principle of the present invention will be described with reference to FIG. FIG. 20 shows a state of so-called two-beam interference in which a certain plane wave 202 is incident on the optical axis 201 at an angle φ1, and another plane wave 203 is incident at an angle φ2, and they interfere at z = 0. . However, assuming that there are certain angles φ and δ, φ1 = φ−δ and φ2 = φ + δ. At this time, the light intensity distribution I (x) at z = 0 can be expressed by the following equation.

  If φ1 and φ2 are equal, δ is 0, so that I (x, z) does not depend on z. That is, the depth of focus is theoretically infinite. However, since δ has a value when φ1 and φ2 are different, I (x, z) depends on z. The z dependence of I (x, z) is repeated in the following equation.

As a result, the smaller the δ is, that is, the greater the equality of φ1 and φ2, the greater the depth of focus.

  Based on the above principle and the effect of oblique incidence illumination, when a binary mask or a halftone mask is used, the effective light source distribution having a shape as shown in FIG. It can be seen that the effective light source distribution having the shapes as shown in FIGS. The point of improvement is to provide a narrow illuminance distribution in the vertical and horizontal directions of illumination. Considering the experience of the present inventors, the influence of the eyelet lens and the illuminance, the vertical and horizontal illuminance distribution diagrams, that is, b3, c3 and dd3 in FIG. A width of about 0.15 should be used.

In particular, when the mask pattern has a strong periodicity, the shape can be easily obtained as follows. When the pattern half pitch is p, the numerical aperture of the projection optical system is NA, and the wavelength is λ, the k ₁ factor is p × NA / λ. In this case, when the shape of the effective light source distribution is obtained by the conversion sigma, about 1 / a (4 × _{k 1} factor), b3, c3, dd3 may be 0.1.

  The effective light source having the general shape as shown in FIGS. 4B, 4C, and 4D will be described in more detail. In FIG. 4B, the effect can be exhibited if 0.3 ≦ b1 ≦ b2 <1.0. In FIG. 4C, the effect can be exhibited if 0.1 ≦ c1 ≦ c2 <0.9. In FIG. 4D, the effect can be exhibited if 0.3 <dd1 ≦ dd2 <1.0. In actual exposure apparatuses, there are few apparatuses that guarantee σ to 1, but in practice, it can be considered that σ corresponds to 1 or more by narrowing the NA.

  Consider an auxiliary pattern insertion method for isolated holes. Conventionally, as shown in FIG. 21A, the auxiliary pattern 212 is inserted around the desired pattern 211 in the vertical and horizontal directions. However, the staggered arrangement according to the present invention is adopted, and the auxiliary pattern 212 is inserted around the desired pattern 211 as shown in FIG. When illumination was performed under illumination conditions according to the mask, it was found that FIG. 21 (b) had better performance in terms of depth of focus, exposure amount, and the like than FIG. 21 (a). That is, the present invention can also be applied to isolated patterns. In order to apply the present invention to an isolated pattern, it is only necessary to place the isolated pattern on a suitable virtual lattice and to arrange them alternately. Basically, the half period of the virtual lattice should be the same as the hole diameter of the isolated pattern, but when there is a periodic pattern that is slightly apart, it may be adjusted to the period.

  The present invention can also be applied when two holes are lined up. Conventionally, as shown in FIG. 22A, auxiliary patterns 222 are arranged around the desired pattern 221 in the vertical and horizontal directions. However, according to the present invention, a virtual lattice composed of two holes is assumed, and the auxiliary patterns 22 are arranged alternately. When illuminated with appropriate oblique incidence illumination, it was found that FIG. 22 (b) had better imaging performance in terms of depth of focus, exposure amount, etc. than FIG. 22 (a). That is, the present invention can also be applied when two holes are arranged.

  Needless to say, the present invention can be similarly applied when three or more holes are arranged.

  In this embodiment, it is assumed that the exposure apparatus is an apparatus having KrF (wavelength 248 nm) and a numerical aperture of 0.73. Consider a case in which 13 contact holes of 120 nm are aligned with a space of 120 nm under such conditions. This situation is shown in FIG. FIG. 29 is a diagram for illustrating the arrangement of contact hole patterns (desired patterns).

  Further, as a mask, a halftone mask having a pattern portion as a light transmissive portion and a member other than the pattern portion having a transmittance of about 6% and maintaining a phase difference of 180 degrees with light passing through the light transmissive portion. (Attenuated phase shifting mask).

  Since the pattern period in the horizontal direction is 240 nm, a virtual grating having a period of 240 nm in the vertical direction, that is, a virtual grating of 240 nm in both the vertical and horizontal directions is considered, and an auxiliary pattern having a size of 90 nm is arranged on the lattice point. .

  When exposure was performed in an illumination mode in which the shape of the effective light source was as shown in FIG. 4A, the result shown in FIG. 30 was obtained. FIG. 30 is a contour diagram showing the light intensity distribution of the aerial image on the wafer surface.

  As can be seen from the results of FIG. 30, it can be seen that the light intensity of the desired pattern portion becomes comparable to the light intensity of the background, and the desired pattern is not resolved. Even if the auxiliary patterns are arranged alternately, the result is the same and the desired pattern is not resolved. Therefore, without changing the period of the virtual lattice, the size of the light transmitting part corresponding to the desired pattern is 140 nm (that is, the space is 100 nm), the size of the auxiliary pattern is 110 nm, and the normal arrangement is obtained. The corresponding light intensity became stronger than the background light intensity, and the desired pattern was resolved. The result is shown in FIG. FIG. 31 is a contour diagram showing the light intensity distribution of the aerial image on the wafer surface.

  Similarly, even when the auxiliary patterns are arranged alternately, the pattern is resolved when the desired pattern size is 140 nm and the auxiliary pattern size is 110 nm.

  From the above results, when a halftone mask is used, the desired pattern size on the mask is multiplied by the pattern size corresponding to the pattern size to be transferred onto the wafer (multiplying the projection optical system magnification of the exposure apparatus). It was found that it must be made somewhat larger than (size).

  However, if it is too large, the pattern becomes dense and it becomes difficult to insert the auxiliary pattern, or the shape is distorted due to the influence of the pattern in which the desired pattern itself is close. For example, FIG. 32 shows the results when the desired pattern is 180 nm and the auxiliary pattern is 150 m. This is still okay, but you can see that the hole is a little longer (this shows that it corresponds to a decrease in contrast).

  After all, in a halftone mask having a desired pattern and an auxiliary pattern having a smaller dimension, the size of the desired pattern is 120% of the size corresponding to the size to be formed on the substrate (wafer). It can be seen that it is preferably 150% or less.

  Next, a binary mask was used as a mask. The exposure method proposed by the present invention is characterized by virtually creating a periodic pattern. In this case, the intensity ratio between the diffracted beams that interfere with each other affects the contrast of the aerial image. Of course, the contrast increases as the intensity ratio of the diffracted light that interferes is closer to 1. In the binary mask, the intensity of the zero-order light is stronger than the intensity of the primary light. In order to bring the intensity of the primary light closer to the intensity of the 0th order light, the desired pattern size on the mask may be made smaller than the pattern size to be transferred. As a result, the contrast of the aerial image increases. As an example, consider exposing a contact hole pattern as shown in FIG.

  First, the lateral contrast of the central hole estimated from an aerial image with a mask having a desired pattern size of 120 nm and an auxiliary pattern size of 90 nm was about 36%. On the other hand, the lateral contrast of the central hole estimated from an aerial image with a mask having a desired pattern of 140 nm and an auxiliary pattern of 110 nm was about 31%.

  In other words, in the case of a binary mask, the pattern size corresponding to the pattern size to be transferred onto the wafer (projection optical system of the exposure apparatus) is reversed from the halftone mask described above. It was found that the contrast of the image is higher when the size is set smaller than (the size multiplied by the magnification of). However, if the pattern size is too small, the throughput decreases and it becomes difficult to create a mask. Practically, the lower limit is preferably about 80%.

  After all, in a binary mask having a desired pattern and an auxiliary pattern having a smaller dimension, the size of the desired pattern is 80% or more of the size corresponding to the size desired to be formed on the substrate (wafer). It turns out that it is 95% or less.

  In this embodiment, a KrF (wavelength 248 nm), numerical aperture 0.73 apparatus is used as an exposure apparatus, and exposure is performed in an illumination mode in which the effective light source has a shape as shown in FIG. 4A. ing.

  Consider that the contact hole pattern shown in FIG. 33 is exposed under such conditions. In FIG. 33, eleven contact hole patterns with a period of 250 nm and a length of 120 nm are arranged in the vertical direction (in FIG. 33, only the three contact hole patterns at the center are shown for the sake of simplicity). Two rows are arranged at an interval of 520 nm.

  According to the flowchart shown in FIG. 23, it is only necessary to insert auxiliary patterns in the horizontal direction at a cycle of about 253 nm, ie, 760 nm ÷ 3. However, in this embodiment, the period of the desired pattern is also applied in the horizontal direction as shown in FIG. 34, that is, a virtual lattice is formed in the horizontal direction with a period of 250 nm. FIG. 34 is a diagram showing a virtual lattice applied to the desired pattern of FIG.

  Thus, even when the virtual lattice was configured and the auxiliary pattern was arranged in a normal arrangement, the desired pattern was resolved (FIG. 35). FIG. 35 is a view showing an aerial image of a desired pattern on the wafer.

  Next, by utilizing the fact that the imaging performance is good when the hole diameter and the space dimension are 1: 1, a virtual lattice is formed with a period of 240 nm in the lateral direction as shown in FIG. . Even when the auxiliary pattern was arranged in a normal arrangement on the virtual lattice formed in this way, the desired pattern was resolved (FIG. 37). FIG. 37 is a diagram showing an aerial image of a desired pattern on the wafer.

  In the present embodiment, only the result of the normal arrangement of the auxiliary pattern is shown, but the auxiliary pattern insertion method of the present embodiment is not limited to the normal arrangement. That is, after the virtual lattice is determined by the method described in the present embodiment, the auxiliary pattern may be arranged in an arrangement in which the lattice point and the auxiliary pattern center are alternate.

  While specific embodiments of the invention have been described above, it will be appreciated that the invention may be practiced otherwise than as described. This description is not intended to limit the invention. It should be clearly noted that the present invention can be applied to any mask pattern.

  According to this embodiment, it is possible to transfer patterns having different periods onto the wafer in a batch. Further, as a secondary effect, the resolving power is improved as compared with a mask in which only a desired pattern portion is arranged. As described above, according to this embodiment, it is possible to perform fine processing by a projection exposure apparatus using a conventional photomask.

  Hereinafter, an exposure apparatus 300 for exposing a mask of the present invention will be described with reference to FIG. Here, FIG. 1 is a schematic block sectional view of an exemplary exposure apparatus 300 of the present invention. As shown in FIG. 1, the exposure apparatus 300 includes an illumination apparatus 310 that illuminates the mask, a projection optical system 330 that projects diffracted light generated from the illuminated mask pattern onto the plate 340, and a stage that supports the plate 340. 345.

  The exposure apparatus 300 is a projection exposure apparatus that exposes a circuit pattern formed on the mask 320 to the plate 340 by, for example, a step-and-scan method or a step-and-repeat method. Such an exposure apparatus is suitable for a lithography process of sub-micron or quarter-micron or less, and in the present embodiment, a step-and-scan method (also referred to as “scanner”) will be described below as an example. Here, in the “step-and-scan method”, the wafer is continuously scanned (scanned) with respect to the mask to expose the mask pattern onto the wafer, and the wafer is stepped after the exposure of one shot is completed. An exposure method for moving to an exposure area. The “step and repeat method” is an exposure method in which the wafer is stepped and moved to the exposure area of the next shot for every batch exposure of the wafer.

  The illumination device 310 illuminates the mask 320 on which a transfer circuit pattern is formed, and includes a light source unit 312 and an illumination optical system 314.

  The light source unit 312 can use, for example, an ArF excimer laser with a wavelength of about 193 nm, a KrF excimer laser with a wavelength of about 248 nm, an F2 excimer laser with a wavelength of about 157 nm, and the type of light source is an excimer laser. The number of light sources is not limited. When a laser is used for the light source unit 312, a light beam shaping optical system that shapes a parallel light beam from the laser light source into a desired beam shape and an incoherent optical system that makes a coherent laser beam incoherent are used. Is preferred.

  The illumination optical system 314 is an optical system that illuminates the mask 320 and includes a lens, a mirror, a light integrator, a stop, and the like. For example, a condenser lens, a fly-eye lens, an aperture stop, a condenser lens, a slit, and an imaging optical system are arranged in this order. The illumination optical system 314 can be used regardless of on-axis light or off-axis light. The light integrator includes an integrator configured by stacking a fly-eye lens and two sets of cylindrical lens array (or lenticular lens) plates, but may be replaced with an optical rod or a diffractive element.

  The mask 320 is made of, for example, quartz, on which a circuit pattern (or image) to be transferred is formed, and is supported and driven by a mask stage (not shown). Diffracted light emitted from the mask 320 passes through the projection optical system 330 and is projected onto the plate 340. The mask 320 and the plate 340 are optically conjugate. Since the exposure apparatus 300 of this embodiment is a scanner, the pattern of the mask 320 is transferred onto the plate 340 by scanning the mask 320 and the plate 340 at a speed ratio of the reduction magnification ratio. In the case of a step-and-repeat type exposure apparatus (also referred to as a “stepper”), exposure is performed with the mask 320 and the plate 340 stationary. On the mask 320, the above-described desired pattern and auxiliary pattern are formed.

  The projection optical system 330 includes an optical system composed only of a plurality of lens elements, an optical system (catadioptric optical system) having a plurality of lens elements and at least one concave mirror, a plurality of lens elements, and at least one kinoform. An optical system having a diffractive optical element such as an all-mirror optical system can be used. When correction of chromatic aberration is required, a plurality of lens elements made of glass materials having different dispersion values (Abbe values) can be used, or the diffractive optical element can be configured to generate dispersion in the opposite direction to the lens element. To do.

  The plate 340 is an object to be processed such as a wafer or a liquid crystal substrate, and is coated with a photoresist. The photoresist coating process includes a pretreatment, an adhesion improver coating process, a photoresist coating process, and a prebaking process. Pretreatment includes washing, drying and the like. The adhesion improver coating process is a surface modification process for improving the adhesion between the photoresist and the base (that is, a hydrophobic process by application of a surfactant), and an organic film such as HMDS (Hexmethyl-disilazane) is used. Coat or steam. Pre-baking is a baking (baking) step, but is softer than that after development, and removes the solvent.

  The stage 345 supports the plate 340. Since any structure known in the art can be applied to the stage 345, a detailed description of the structure and operation is omitted here. For example, the stage 345 can move the plate in the XY directions using a linear motor. The mask 320 and the plate 340 are synchronously scanned, for example, and the positions of the stage 345 and a mask stage (not shown) are monitored by a laser interferometer, for example, and both are driven at a constant speed ratio. The stage 345 is provided on a stage surface plate supported on a floor or the like via a damper, for example, and the mask stage and the projection optical system 330 are, for example, a damper or the like on a base frame placed on the floor or the like It is provided on a lens barrel surface plate (not shown) that is supported via the.

  In the exposure, the light beam emitted from the light source unit 312 illuminates the mask 320 by, for example, Koehler illumination by the illumination optical system 314. Light that passes through the mask 320 and reflects the mask pattern is imaged on the plate 340 by the projection optical system 330. The mask 320 used by the exposure apparatus 300 is a device (semiconductor) with higher quality than the conventional one because the auxiliary pattern enhances the imaging performance of the desired pattern (for example, close to a circle) and increases the depth of focus. Elements, LCD elements, imaging elements (CCD, etc.), thin film magnetic heads, etc.) can be provided.

  Next, with reference to FIGS. 7A and 7B, an embodiment of a device manufacturing method using the above-described exposure apparatus will be described. FIG. 7A is a flowchart for explaining how to fabricate devices (ie, semiconductor chips such as IC and LSI, LCDs, CCDs, and the like). Here, the manufacture of a semiconductor chip will be described as an example. In step 1 (circuit design), the device circuit is designed. In step 2 (mask production), a mask on which the designed circuit pattern is formed is produced. In step 3 (wafer manufacture), a wafer is manufactured using a material such as silicon. Step 4 (wafer process) is called a pre-process, and an actual circuit is formed on the wafer by the lithography technique of the present invention using the mask and the wafer. Step 5 (assembly) is called a post-process, and is a process for forming a semiconductor chip using the wafer created in step 4, and includes processes such as an assembly process (dicing and bonding) and a packaging process (chip encapsulation). . In step 6 (inspection), inspections such as an operation confirmation test and a durability test of the semiconductor device created in step 5 are performed. Through these steps, the semiconductor device is completed and shipped (step 7).

  FIG. 7B is a detailed flowchart of the wafer process in Step 4. In step 11 (oxidation), the wafer surface is oxidized. In step 12 (CVD), an insulating film is formed on the surface of the wafer. In step 13 (electrode formation), an electrode is formed on the wafer by vapor deposition or the like. In step 14 (ion implantation), ions are implanted into the wafer. In step 15 (resist process), a photosensitive agent is applied to the wafer. Step 16 (exposure) uses the exposure apparatus to expose a circuit pattern on the mask onto the wafer. In step 17 (development), the exposed wafer is developed. In step 18 (etching), portions other than the developed resist image are removed. In step 19 (resist stripping), the resist that has become unnecessary after the etching is removed. By repeatedly performing these steps, multiple circuit patterns are formed on the wafer. According to the device manufacturing method of the present embodiment, it is possible to manufacture a device of higher quality than before.

It is the simplified optical path figure of the exposure apparatus as one Embodiment of this invention. It is a schematic sectional drawing of a phase shift mask, a binary mask, and a halftone mask. It is a figure which shows the schematic plan view of a normal mask and the mask used for the exposure method as one side of this invention, and a desired contact hole pattern. It is a figure which shows the shape of the effective light source of the illumination system applicable to the exposure method of this invention. It is a schematic diagram of a desired contact hole pattern and a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. 5 is a flowchart for explaining a device manufacturing method using the exposure apparatus shown in FIG. 1 and a detailed flowchart of the wafer process in step 4; It is a figure which shows the exposure result by the exposure method of this invention. It is a schematic plan view of the mask applicable to the exposure method of this invention. It is a figure which shows typically distribution and image formation of the diffracted light on the pupil surface produced from the mask shown in FIG. It is a schematic diagram which shows the mask pattern applicable to the exposure method of this invention, and its image formation state. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic plan view of the mask applicable to the exposure method of this invention. It is a figure which shows distribution of the diffracted light on the pupil surface produced from the mask shown in FIG. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram for demonstrating 2 light beam interference. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a schematic diagram of a mask pattern applicable to the exposure method of the present invention. It is a flowchart for demonstrating the determination method of the virtual lattice of this invention. 3 is a flowchart for explaining a method of setting a mask pattern and illumination conditions according to the present invention. It is a top view for demonstrating OPC. It is a figure which shows typically the mask pattern production method applicable to the exposure method of this invention. It is a figure for demonstrating the definition of the name of a diffracted light. It is a figure for demonstrating the definition of the name of a diffracted light. It is a figure for showing the arrangement | sequence of a pattern. It is a contour figure showing the light intensity distribution of the aerial image on a wafer surface. It is a contour figure showing the light intensity distribution of the aerial image on a wafer surface. It is a figure which shows an exposure result. It is a figure for showing the arrangement of a pattern. It is a figure for showing the arrangement | sequence of a pattern. It is a figure which shows an exposure result. It is a figure for showing the arrangement | sequence of a pattern. It is a figure which shows an exposure result.

Explanation of symbols

21 mask base 22 light shielding member 23 phase shifter 24 halftone member 31, 34, 51, 52, 61, 111, 121, 131, 161, 171, 181, 191-193, 211, 221 desired pattern 32, 53, 62, 112, 122, 132, 164, 165, 175, 176, 182, 183, 212, 222 Auxiliary pattern 33, 92, 142 Light shielding part 41A-41F Illumination system light transmission part 42A-42F Illumination system light shading part 63 Virtual grating 91 , 141 Mask light transmitting part 101 Pupil 102, 101a-107b Diffracted light 101g Schematic shape of the formed pattern 101h-104h, 102g-103g Schematic shape of the formed pattern 113A, 113B Image cross-sectional view 123a, 123b Pattern region 162, 163, 172-174 The traveling direction of the pattern 201 optical axis 202 and 203 plane wave as

Claims (4)

A mask having a plurality of contact hole patterns and a plurality of auxiliary patterns having dimensions smaller than the plurality of contact hole patterns,
The mask is a binary mask or a halftone mask, and the phases of the plurality of contact hole patterns and the auxiliary pattern are the same,
The plurality of auxiliary patterns include a first auxiliary pattern and a plurality of second auxiliary patterns,
The centers of the plurality of contact hole patterns and the center of the first auxiliary pattern are arranged on a straight line at equal intervals,
The centers of the plurality of second auxiliary patterns are equidistant from the centers of two adjacent patterns of the plurality of contact hole patterns,
In the plurality of second auxiliary patterns, the distance between the straight line and the center of the second auxiliary pattern is equal to the period of the plurality of contact holes, and the distance between the lines is the same as the period of the plurality of contact hole patterns. A mask arranged on a straight line parallel to the straight line on both sides. The object according to claim 1, wherein the mask according to claim 1 is illuminated with oblique incidence illumination so that the plurality of contact hole patterns are resolved and the resolution of the plurality of auxiliary patterns is suppressed. An exposure method characterized by exposing . Multiple contact hole patterns;
A mask design method comprising a first auxiliary pattern and a plurality of second auxiliary patterns, the mask having a plurality of auxiliary patterns having dimensions smaller than the plurality of contact hole patterns,
The mask is a binary mask or a halftone mask, and the plurality of contact hole patterns and the auxiliary patterns have the same phase,
Disposing the first auxiliary pattern such that the center of the contact hole pattern and the center of the first auxiliary pattern are aligned on a straight line at equal intervals;
The centers of the plurality of second auxiliary patterns are equidistant from the centers of two adjacent patterns of the plurality of contact hole patterns, and the distance between the straight line and the center of the second auxiliary pattern is the plurality of the plurality of contact hole patterns. Disposing the plurality of second auxiliary patterns on a straight line parallel to the straight line on both sides of the straight line, and having the same period as that of the plurality of contact hole patterns. Mask design method. Exposing the object to be processed using the mask according to claim 1;
Developing the exposed object to be processed.