KR20100020665A - Regular lithography method - Google Patents

Abstract

PURPOSE: A regular lithography method is provided to supply a pitch set condition for forming a uniform projection structure on equal sides by analyzing the projecting structure which change according to the pitch. CONSTITUTION: In a regular lithography method, a projection angel is determined less than 3 pie/4 radian(S110). B of vertical cell number and A of transverse cell number are relatively prime and they are over 1. A time resolution S, a reflection length ratio Z, an image interval C, the transverse cell number A, and the vertical cell number B are decided so that the scan resolution D and unit step number L are integral number. A mirror angle is formed based on the reflection length ratio Z, the transverse cell number A and vertical cell number B(S450).

Description

Regular Lithography Method

The present invention relates to a digital lithography method using a spatial light modulator, in particular digital lithography based on a mathematical definition of a projection structure in which beams projected onto a substrate are formed on the substrate in accordance with a correlation between the movement of the substrate and the light modulation step. The present invention relates to a regular lithography method in which the regularity, accuracy, economy, efficiency and speed of the system are guaranteed at the same time.

In a digital lithography system using a spatial light modulator or a digital mirror device, the micromirrors constituting the micromirror array of the spatial light modulator are placed on a substrate surface moving with time. The pattern is exposed by light beams that selectively reflect. Whether the light beam is reflected by the micromirror array corresponding to each substrate movement serves as a partial mask, and the total set of light beam reflections by the micromirror arrays corresponding to the entire substrate movement is the role of one complete digital mask. Do this. Therefore, digital lithography using a spatial light modulator generates a pattern and mask data suitable for each substrate movement, and transmits them to the micromirror controller according to the substrate movement, respectively. The patterning precision of the exposure pattern is determined.

On the other hand, when the number of micromirrors participating in the exposure of the spatial light modulator is N horizontal M vertically, the mask data is N * M bit digital data represented by a series of 1's and 0's. For example, in the case of the XGA-class Digital Micromirror Device (DMD) of Texas Instrument, a representative spatial light modulator, since N is 1024 and M is 768, the mask data is 786,432 bits of digital data. For example, if 10000 light modulation steps are performed per second to match the resolution and processing time required in an actual process, the corresponding 10000 mask data are 7,864,320,000 bits of digital data. 7.86432 Gbps (Giga bit per second), which must generate and transmit large amounts of data at very high speeds.

At present, at home and abroad, maskless lithography to digital lithography methods using various spatial light modulators have been patented or filed, and in Korea, Korean Patent Publication Nos. 2007-0020410, 2006-0047613, 2006-0043024, and 2006-0043024 2006-0045355, 2006-0109724 and Patent No. 0660045, 655165, etc., in the US Patent No. 6,870,604, No. 6,473,237, U.S. Patent No. 6,379,867 and the like. Of these, the patents referring to the resolution of the scan ratios directly related to the present invention are described in "High Resolution point array" U.S. Patent No. 6,870,604, which refers to a projection structure directly related to the present invention in terms of resolution, is a "super resolution digital lithography" of 10-2008-0041846 filed by the applicants, and is a digital mask directly related to the present invention. There is no patent referring to the structural compression of the projection structure.

Background art of the present invention is based on T. Kanatake of Ball Semiconductor Inc., USA, which is directly related to the present invention. Patent No. 6,870,604 is mentioned as an example of the resolution related to the scan ratio, and 10-2008-0041846 "Super Resolution Digital Lithography", which is patented by the applicants, is mentioned as an example the resolution related to the projection structure. The technology belongs to the technical field to which the invention of the "occupation area based pattern generation method for maskless lithography" of Patent No. 655165, which is filed and filed by the applicants, and the prior art of the field and the patent application by the applicants 10-2007-0046450, "Super Resolution Digital Lithography," of 10-2008-0041846, to which the invention of the "Inline Virtual Masking Method for Maskless Lithography" belongs, and the prior art in that field and patented by the applicants It is the same as that described in detail in the technical field to which the invention belongs and the prior art of the field and the like And the patents incorporated by reference in the specification of the present application.

In a maskless lithography system using a spatial light modulator, the micromirror array of the spatial light modulator is rotated at an angle with respect to the direction of movement of the substrate, and the light beams selectively reflect on the surface of the substrate where the micromirrors move over time. Since the pattern is exposed by means of exposure, the degree of the pattern resulting from the exposure is improved as the rotation angle of the micromirror array decreases and the pitch decreases, with the arrangement and illuminance distribution of the light beam reflected from the micromirror determined. Known. However, depending on the setting of the pitch, in some cases, even if the pitch decreases, the degree of pattern resulting from exposure does not improve, but rather decreases, and in some cases, the degree of actually exposed pattern is targeted or predicted. It may get bigger. Therefore, a careful analysis of the effect of pitch on the degree of pattern is necessary. On the other hand, if the pitch decreases to match the resolution required in the actual mass production process, the number of mask data to be generated or transmitted within the unit time increases in order to match the process time required in the actual mass production process, thereby generating the mask data or the like. Speed up transmission. Therefore, there is an urgent need for a method capable of speeding up generation or transmission of mask data.

U.S. Patent No. The lithographic method based on the rotation angle of the micromirror array proposed in 6,870,604 focuses on the rotation angle of the micromirror array, so it does not mention any projection structure that depends on the pitch. There is a problem in that the degree of the exposed pattern becomes larger than the resolution determined as the target.

The lithography method based on the pitch proposed by the applicants of 10-2008-0041846 is derived under the premise that the horizontal and vertical spacing of the beams projected onto the substrate is the same, so that a perfect equilateral triangle or perfect isosceles triangle shape is obtained. Instead, it proposes a projection structure close to this, and it is derived only for the case where the number of projection images of the longitudinal mirror, which is the basis of the projection structure, is 1. Furthermore, it does not mention how to increase the projection area without increasing the total number of micromirrors involved in exposure, or how to quantitatively and losslessly compress a large amount of pattern data or digital mask.

Therefore, in view of the problems of the prior art as described above, the improvement method of the prior art as described above is a projection image of the longitudinal mirror which is the basis of the projection structure when the horizontal and vertical intervals of the beams projected on the substrate are different from each other. A wide range of methods that can be applied to form a projection structure of equilateral triangles, right-sided isosceles triangles, and isosceles triangles, which are set as targets, are sought by considering all the projection structures that vary depending on the pitch when the number of times is 1 or more. Required, and furthermore, it is possible to increase the projection area while maintaining the resolution without increasing the total number of micromirrors involved in the exposure, which is essential for accelerating the application of the actual substrate to digital lithography using spatial light modulators. How to and Quantify Massive Digital Mask Data There is an urgent need for a method of generating or transmitting at high speed by compressing the yarn.

One representative indicator for determining patterning precision in a lithography process is the critical dimension (CD) of the exposed pattern. Line width to line space and the like are the critical dimensions when the pattern is generated using the "occupation area based pattern generation method for maskless lithography" of Patent No. 655165, filed by the applicants. It is also possible to freely adjust the line width by about 25% without modifying the drawing by using the reflection confirmation occupancy area ratio. However, according to Patent 655165, stepwise jumps of line width appear when discrete rotation angles or discrete pitches of a micromirror array are applied, and jumps of line width are amplified when they are applied simultaneously. The stepped jump of the line width is represented by the line edge roughness (LER) of the exposed pattern, which greatly reduces the uniformity of the critical dimension of the exposed pattern, which is another indicator of patterning accuracy. It becomes a factor.

FIG. 1 shows a light beam having a Gaussian illuminance distribution with a beam radius of 0.5 microns on a substrate for a moving substrate when two discrete pitches of 0.7035 microns and 0.3519 microns are applied under the same 2.4 degree discrete angle of rotation. Discrete method that irradiates light beams for a certain moment while the substrate travels one pitch, and blocks the rest for the rest of the time, so that the line produces 5 micron 90, 60, 45 and 0 degree straight lines After performing simulations of an analog exposure process in which a light beam is continuously irradiated for one pitch while the substrate is moved, the results of the simulations are all scaled by fixing the maximum illuminance to 1 for accurate comparison. 1 (a) and 1 (b) are simulation results of the discrete exposure process, and FIGS. 1 (c) and 1 (d) are simulation results of the analog exposure process, and FIGS. 1 (a) and 1 (c). Has a pitch of 0.7035 microns, and FIGS. 1 (b) and 1 (d) have a pitch of 0.3519 microns. Thus, comparing all the other conditions except for the pitches of Fig. 1 (a) and Fig. 1 (b) with each other and Fig. 1 (c) and 1 (d) with each other, Fig. 1 (b) and 1 (d) Although the pitch of) is reduced than the pitches of FIGS. 1 (a) and 1 (c), the result shows that the line edge roughness which is not shown in FIGS. 1 (a) and 1 (c) is not shown in FIGS. Appearing on the 90 degree, 60 degree, and 45 degree straight lines excluding the horizontal line, it can be seen that the patterning precision is reduced in both the discrete method and the analog method. The results of FIG. 1 suggest that patterning precision and line edge roughness vary depending on how the discrete pitch is set even under the same discrete rotation angle. Of course, even in the case of Figs. 1 (b) and 1 (d), if the pitch is made very small, the line edge roughness will be reduced, but if the pitch is decreased, the exposure time required for exposing the same area increases, which makes the exposure process inefficient and inefficient. It becomes impractical, making it impossible to apply digital lithography using a spatial light modulator to the actual production process of the substrate. Therefore, in order to apply digital lithography using a spatial light modulator to the actual production process of the substrate, it is necessary to maintain the pitch at a realistic level and to reduce the line edge roughness of the exposed pattern to ensure uniformity of the critical dimension.

The first object of the present invention is to analyze the pitch-dependent projection structure in digital lithography using a spatial light modulator, to form a regular projection structure of equilateral equilateral triangles of equilateral triangles, right isosceles triangles and isosceles triangles. It is to provide a digital lithography method that guarantees regularity and accuracy by presenting setting conditions.

A second object of the present invention is to analyze the projection structure that depends on the pitch and the distance between the horizontal and vertical beams in digital lithography using a spatial light modulator, to be involved in exposure while preserving the regular projection structure of equilateral sides of the same resolution. The present invention aims to provide a digital lithography method that guarantees accuracy and economy simultaneously by suggesting a method of increasing the projection length without increasing the total number of micromirrors.

The third object of the present invention is to closely analyze the correlation between the scan ratio, the pitch and the resolution in digital lithography using a spatial light modulator. The present invention aims to provide a digital lithography method in which accuracy and efficiency are guaranteed at the same time by suggesting an increase condition of the pitch such that is preserved.

A fourth object of the present invention is to analyze a projection structure that depends on the scan ratio and resolution in digital lithography using a spatial light modulator, thereby structurally storing a large amount of digital mask data while preserving a regular projection structure of equilateral sides of the same resolution. It is to provide a digital lithography method that ensures accuracy and speed at the same time by quantitative lossless compression and ultra high speed generation or transmission.

The normal lithographic method of the present invention for achieving the above object,

Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians;

Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And

The scanning resolution D, the projection length ratio Z, the image spacing C, the image spacing C, and the horizontal cell number A, with the projection form angle Ψ in between, including calculating the mirror angle θ based on the projection length ratio Z, the horizontal cell number A, and the vertical cell number B. The scan direction resolution calculated as the vertical cell number B is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the super resolution S calculated as the horizontal cell number A and the vertical cell number B, so that the angle between the super resolution isosceles It is characterized by maintaining an isosceles triangle projection structure with a projection shape angle Ψ.

Another normal lithography method of the invention is

Determining a projection shape angle Ψ that defines a projection structure as π / 3 radians or 2π / 3 radians;

Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And

The scanning resolution D, the projection length ratio Z, the image spacing C, the image spacing C, and the horizontal cell number A, with the projection angle Ψ between The resolution of the scan direction, which is calculated as the number of vertical cells B, is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the number of horizontal cells A, and the super resolution S calculated as the number of vertical cells B. It is characterized by maintaining the delta projection structure of ().

Another regular lithography method of the present invention is

Determining a projection form angle Ψ that defines the projection structure as π / 2 radians;

Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And

The scanning resolution D, the projection length ratio Z, the image spacing C, and the horizontal cell number A with the projection form angle Ψ between, including calculating the mirror angle θ based on the projection length ratio Z, the horizontal cell number A, and the vertical cell number B. And the scanning direction resolution calculated as the vertical cell number B is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the super resolution S calculated as the horizontal cell number A and the vertical cell number B, between the super resolution isosceles. A stripe projection structure of right-angled isosceles triangles having right angles is maintained.

Preferably, the super resolution S is calculated as cscΨ * Z * C / (A ² + B ² * Z ² ) ^1/2 or C * (A ² + B ² * Z ² ) ^1/2 / D will be.

Preferably, the scan resolution D is calculated as C * (A ² + B ² * Z ² ) ^1/2 / S or sinΨ * (A ² + B ² * Z ² ) / Z, and the unit step number L Is calculated as sinΨ * A / Z-cosΨ * B.

Preferably, the micromirror array of the spatial light modulator is rotated with respect to the substrate at the mirror angle θ, and the scan pitch P, which is the unit travel distance of the substrate corresponding to the unit light modulation time of the micromirror array, is set to super resolution S. The step is further provided so that the position of the image center of row J column I of any K th light modulation step is structurally identical to the position of the image center of row J + B row I + A of the KD th light modulation step.

Preferably, the micromirror array of the spatial light modulator is rotated with respect to the substrate at the mirror angle θ, and the scan ratio H is determined by an integer greater than 1 and mutually independent of the scan resolution D to correspond to the unit light modulation time of the micromirror array. And setting the scan pitch P, which is the unit moving distance of the substrate, to H * S, which is the product of the scan ratio H and the super resolution S, while increasing the scan ratio while maintaining the super resolution. The position of the image center of the row I column is structurally coincident with the position of the image center of the row I + H * A of the row J + H * B of the KD th light modulation step.

Another regular lithography method of the present invention is

Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians;

Based on the projection form angle Ψ, the fixed resolution G, the projection length ratio Z, the image spacing C, the horizontal cell number A and the vertical cell number B are integers of 1 or more, where the vertical cell number B is the horizontal cell number A and the other (coprime) Determining to make;

Calculating a scan resolution D based on the projection shape angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B; And

The scan resolution S is calculated based on the projection length ratio Z, the image interval C, the horizontal cell number A, the vertical cell number B, and the scan resolution D, and the mirror angle θ is calculated based on the projection length ratio Z, the horizontal cell number A and the vertical cell number B. And maintaining a triangular projection structure in which the angle between the fixed resolution side and the scan resolution side is the projection shape angle Ψ.

Preferably, the fixed resolution of G is cscΨ * Z * C / (A 2 + B 2 * Z 2) 1/2 calculated, and the scan resolution, D is the first place under the R a ROUND {R, 0} point SinΨ * (A ² + B ² * Z ² ) / Z rounded up at (ROUND {sinΨ * A / Z-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² * Z ² ) / A, ROUND {sinΨ * (A ² + B ² * Z ² ) / Z, 0} and ROUND {(ROUND {sinΨ * A / Z-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² * Z ² ) / A, 0} is selected and calculated, and the scan resolution S is calculated as C * (A ² + B ² * Z ² ) ^1/2 / D.

Preferably, the projection length ratio Z is determined to be cscΨ * V / W or cscΨ using two integers V and W which are mutually different.

Preferably, the method further comprises the step of setting the horizontal interval between the image centers to the image interval C, and setting the vertical interval between the image centers to Z * C, which is the product of the projection length ratio Z and the image interval C. .

Preferably, the mirror angle θ is calculated in tan ⁻¹ (B * Z / A) radians.

Another regular lithography method of the present invention is

Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians;

Determining the fixed resolution G, the image interval C, the horizontal cell number A, and the vertical cell number B such that the vertical cell number B is an integer equal to or greater than each other (coprime) based on the projection form angle Ψ;

Calculating a scan resolution D based on the projection shape angle Ψ, the horizontal cell number A and the vertical cell number B; And

Calculating the scan resolution S based on the image interval C, the horizontal cell A, the vertical cell B, and the scan resolution D, and calculating the mirror angle θ based on the horizontal cell A and the vertical cell B; And maintaining a triangular projection structure in which the angle between resolution sides is the projection shape angle Ψ.

Preferably, in the fixed resolution of G is ^{C * cscΨ / (A 2 +} B 2) 1 / a ² calculated, and the scan resolution D are rounded to the first place after the a R a ROUND {R, 0}-point value SinΨ * (A ² + B ² ), (A ² + B ² ) / (1-B / (A * tanΨ)), ROUND {(A ² + B ² ) / (1-B / (A * tanΨ) )), 0}, (ROUND {sinΨ * A-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² ) / A, ROUND {sinΨ * (A ² + B ² ), 0} and ROUND {(ROUND {sinΨ * A-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² ) / A, 0} is selected and calculated, and the scan resolution S is C * ( A ² + B ² ) to calculate ^1/2 / D.

Preferably, the mirror angle θ is to be calculated in tan ⁻¹ (B / A) radians.

Preferably, the micromirror array of the spatial light modulator is rotated with respect to the substrate at the mirror angle θ, and the scan pitch P which is the unit travel distance of the substrate corresponding to the unit light modulation time of the micromirror array is set to the scan resolution S. The step is further provided so that the position of the image center of row J column I of any K th light modulation step is structurally identical to the position of the image center of row J + B row I + A of the KD th light modulation step.

Preferably, the micromirror array of the spatial light modulator is rotated with respect to the substrate at the mirror angle θ, and the scan ratio H is determined by an integer greater than 1 and mutually independent of the scan resolution D to correspond to the unit light modulation time of the micromirror array. And setting the scan pitch P, which is the unit moving distance of the substrate, to H * S, which is the product of the scan ratio H and the scan resolution S, while increasing the scan ratio while maintaining the scan resolution. The position of the image center of the row I column is structurally coincident with the position of the image center of the row I + H * A of the row J + H * B of the KD th light modulation step.

Preferably, row J of the Kth light modulation step of the image center array consisting of N rows and M columns in the total Q + 1 light modulation steps by the array of N vertical and horizontal M micromirrors involved in the exposure. Based on the regularity in which the position of the image center of the column is structurally consistent with the position of the image center of the row J + B of the row K + B of the KD light modulation step,

For each Kth light modulation step where K is equal to or greater than 0 and equal to or equal to D-1, the mirror unit mask corresponding to each row J of column I for columns J of column I satisfying that J is 1 or more and N or less and I is 1 or more and M or less. Generating a K th digital mask;

In each K-th optical modulation step where K is greater than or equal to D and less than or equal to Q, row J columns I satisfying that J is greater than or equal to N-B + 1 and less than N, and that I is greater than or equal to 1 and less than or equal to MA, and I is greater than or equal to 1 and less than or equal to N, Quantitatively compressing the K-th digital mask by generating a mirror unit mask corresponding to each J row I column for the J row I columns satisfying A + 1 or more and M or less; And

For each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, for each row J column I satisfying that J is greater than or equal to 1 and less than or equal to NB and I is greater than or equal to 1 and less than or equal to MA, row J + B is greater than or equal to Shifting the + A mirror unit mask by -B rows and -A columns to the row J mask of the Kth light modulation step to update the column I masks in the Kth optical modulation step, further including lossless restoring the compressed portion of the Kth digital mask. To structurally quantitative lossless compression.

Preferably, prior to creation, compression and decompression of the digital mask,

In a total Q + 1 optical modulation step with N vertical and horizontal M micromirror arrays involved in exposure, vertical N is an integer multiple of B, horizontal M is an integer multiple of A, and Q is an integer multiple of D. The step of determining N, M, and Q to be a subtracted integer is to be further provided.

Preferably, row J of the Kth light modulation step of the image center array consisting of N rows and M columns in the total Q + 1 light modulation steps by the array of N vertical and horizontal M micromirrors involved in the exposure. Based on the regularity in which the position of the image center of the column is structurally consistent with the position of the I + H * A column of the image center of row J + H * B of the KD light modulation step,

For each Kth light modulation step where K is equal to or greater than 0 and equal to or equal to D-1, the mirror unit mask corresponding to each row J of column I for columns J of column I satisfying that J is 1 or more and N or less and I is 1 or more and M or less. Generating a K th digital mask;

In each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, row J columns I and J are greater than or equal to 1 and less than or equal to N while at the same time J is NH * B + 1 or more and N or less and I is greater than or equal to 1 or more MH * A Quantitatively compressing the K-th digital mask by generating a mirror unit mask corresponding to each J-row I column for the J-row I columns satisfying MH * A + 1 or more and M or less; And

In each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, for each row J of column I satisfying that J is greater than or equal to 1 and less than NH * B and I is greater than or equal to 1 and less than MH * A, J of the KDth light modulation step Compresses the K-th digital mask by shifting the + H * B-row I + H * A-column mirror unit mask by -H * B-row -H * A column to the J-row I-mirror mask of the K-th optical modulation step The method further includes lossless restoring a portion to structurally quantitatively lossless compress the digital mask.

Preferably, prior to creation, compression and decompression of the digital mask,

In a total Q + 1 optical modulation step with N vertical and horizontal M micromirror arrays involved in exposure, vertical N is an integer multiple of H * B, horizontal M is an integer multiple of H * A, and Q is Steps to determine N, M and Q to be an integer less than 1 from the integer multiples to be provided.

The present invention has the following effects.

Firstly, by implementing the structured regular projection structure unique to the present invention by defining the unique projection shape angle of the present invention, the regular digital lithography having the structure of the regular structure in the delta form of the equilateral triangle, the stripe form of the right isosceles triangle, and the isosceles triangle form This is possible.

Second, by determining the horizontal and vertical spacings between the image centers according to the projection projection ratio, the equilateral isosceles triangle and the uniform super resolution isosceles triangle are uniform. It is possible to realize accurate digital lithography having a precise projection structure in the form of an isosceles triangle with an angle between one super resolution isosceles.

Third, it is economical by increasing the projection length while maintaining the regular and accurate projection structure of super resolution without increasing the number of micromirrors of the optical modulator by setting the vertical gap larger than the horizontal gap between image centers with the unique projection length ratio of the present invention. Digital lithography is possible.

Fourth, even if a scan pitch larger than the super resolution is used by using the resolution and the scan ratio in relation to each other unique to the present invention, it is possible to implement digital lithography efficiently by maintaining the normal projection structure of the super resolution.

Fifth, it is possible to implement digital lithography quickly by structurally quantitatively compressing, losslessly restoring, and transmitting a digital mask based on a regular projection structure unique to the present invention.

Sixth, it is possible to apply the actual substrate mass production process of digital lithography by the regular lithography method, which ensures regularity, accuracy, economy, efficiency and speed unique to the present invention.

1. Projection Structure

In digital lithography using a spatial light modulator, a projection in which light beams reflected by micromirrors of a light modulator in a fixed position are formed on a substrate, which is an exposed object that moves from a predetermined initial position to a predetermined final position. The degree of the pattern is determined by the structure. FIG. 2 illustrates a substrate on an exposed object which is reflected by ten horizontal (column) three micromirrors in a fixed position and moves from a predetermined initial position to a predetermined final position. The projection structure formed in Fig. 1 is shown for each light modulation step to light modulation switch unit. 3 illustrates only the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 18400 in the projection structure of FIG. 2. The boundary between the blue squares 10001 of FIGS. 2 and 3 is regarded as the maximum boundary of the projection image formed by one light beam, and the blue square 10001 is defined as the image cell. do. The red dots to circles 10002 of FIGS. 2 and 3 are regarded as the center of the projection image formed by one light beam, and the red dots to circles 10002 are defined as the image center. 2 and 3, there is only one shortest path through which the initial light modulation step unit projection structure 10000 moves to the position of the final light modulation step unit projection structure 18400, but the final position at the initial position. There is a range of choices as to how many steps of light modulation are performed during the movement, and FIG. 2 shows an example where 84 light modulation steps were performed. FIG. 2 shows the position of the upper left vertex 19002 of the one-row, one-column image cell after 84 optical modulation steps, and the position of the lower-right vertex 19001 of the three-row, ten-column image cell at the initial position. As shown, 10 vertical 3 image cells in a row such that the initial row J column I image center goes to the position of the initial row (J + 3) row (I + 10) image center after 84 light modulation steps. The movement of the light modulation step unit projection structure is performed based on the reference numerals 10001 to 10002. Referring to the image centers 10002 shown in FIG. 3 as an example, an image center of an initial 1 row, 1 column, 1 row, 2 columns, 2 rows, 5 columns, 2 rows 9 columns, 3 rows 8 columns, and 3 rows 9 columns ( Each of the 10011, 10012, 10025, 10029, 10038, and 10039 moves to the image centers (18411, 18412, 18425, 18429, 18438, 18439) after 84 light modulation steps, respectively. 18429, 18438, 18439) Each position is an initial 4 rows 11 columns, 4 rows 12 columns, 5 rows 15 columns, 5 rows 19 columns, 6 rows 18 columns, 6 rows 19 columns when the number of image cells increases. Each of them will match the location of the center. Therefore, the direction 10007 from the image center 10029 shown in FIG. 3 to the image center 18429 becomes the opposite direction to the substrate moving direction. In the present invention, the direction 10007 opposite to the substrate movement direction is defined as the horizontal direction or the scan direction, and the direction 10008 perpendicular to the horizontal direction 10007 is defined as the vertical direction. In addition, the direction 10005 from the image center 10025 to the image center 10029 shown in FIG. 3 rotates the horizontal direction 10007 counter-clock-wise (CCW) by an angle θ 10009. Direction 10005. In the present invention, this direction 10005 is defined as the horizontal direction, the direction 10006 perpendicular to the horizontal direction 10005 is defined as the vertical direction, and angle θ 10009 is defined as the mirror angle hereinafter. In the present invention, the image cell (10001) to the image center in the horizontal direction to determine the position of the initial light modulation step unit projection structure 10000 and the position of the final light modulation step unit projection structure (18400) as a reference of the projection structure The number of (10002) is expressed as A, and the number of horizontal cells is defined, and the number of image cells (10001) and image centers (10002) in the vertical direction is expressed as B, and the number of vertical cells is defined, and the initial light modulation step unit projection structure ( Scanned resoluble degree, denoted by the number of optical modulation steps based on the horizontal cell number A and the vertical cell number B, while 10000 moves the shortest path to the position of the final optical modulation step unit projection structure 18400. Defined as In the example shown in Figs. 2 and 3, the number of horizontal cells A is 10, the number of vertical cells B is 3, and the scan resolution D is 84. Image spacing, denoted by the horizontal interval (10003) to the horizontal length (19003) of the image cell between the light beam image centers projected onto the substrate in a stationary state by the micromirrors of the spatial light modulator shown in FIG. It is defined as The image spacing C, which is the horizontal spacing between the image centers, is the reference length, and the value obtained by dividing the vertical spacing (10004) by the horizontal spacing (10003) between the image centers is expressed as Z. For example, the vertical interval 10004 between the image centers shown in FIG. 3 and the vertical length 19004 of the image cells are expressed as C * Z, which is a value obtained by multiplying the image interval C by the projection length ratio Z, and as the vertical image interval. define. When the unit movement distance of the substrate, which is the unit optical modulation time of the micromirrors of the spatial light modulator, or the movement distance of the optical modulation step unit image center is expressed as P, and defined as scan pitch, scan pitch P Is generalized to SQRT (A * A * C * C + B * B * C * C * Z * Z) / D, and mirror angle θ (10009) is generalized to arctan (B * Z / A) radians ) As shown in Figs. 2 and 3, the image spacing C (10003, 19003) is 10 * SQRT (3) / 2 micron and the projection length ratio Z is 2 / SQRT (3) and thus the vertical image spacing C For example, if Z (10004,19004) is 10 microns, the scan pitch P is about 1.091 microns, and the mirror angle θ 10009 is about 19.1 degrees. Accordingly, when the projection structure of FIG. 2 is rotated about 19.1 degrees counterclockwise, FIG. 2 is a projection structure formed on the substrate moving in the negative horizontal direction by light beams reflected by the micromirror array inclined about 19.1 degrees. Can be regarded as. ArctanN3 is an inverse tangent value of N3. In the present invention, an inverse tangent value of any number N3 is represented by tan ⁻¹ N3 to arctanN3. The SQRT (N4) is a square root of N4. Hereinafter, in the present invention, the square root of any number N4 is represented by (N4) ^1/2 to SQRT (N4). In the present invention, for any number N5, the tangent of N5 is tanN5, the sine of N5 is sinN5, the cosine of N5 is cosN5, and the sine of N5. The cosecant value of N5, which is the inverse (ie 1 / sinN5), is denoted cscN5 or cosecN5.

4 illustrates only the image centers 10002 in the projection structure of FIG. 2. As shown in FIG. 4, each of the image centers 10011, 10012, 10025, and 10039 of the first row, the first row, the first row, the second row, the second row, the fifth row, and the third row of the nine columns has a horizontal direction 10007 for each light modulation step. After the 84 light modulation steps, the image centers 10038 are moved to the image centers 18184, 18412, 18425, and 18439, respectively. It moves to the image center 10638 after six light modulation steps, to the image center 10838 after nine light modulation steps, and to the image center 18842 after 84 light modulation steps. The distribution of the image centers 10002 uniformly shown in the center of FIG. 4 is a projection structure formed by regular movement of the light modulation step unit projection structure, and has a triangular structure by three image centers. Initially, two parallel straight lines (10038-18438 and 10039-18439) formed by the horizontal direction 10007 movement according to the optical modulation step of the eighth-row image center 10038 and the nine-row image center 10039 located in three rows. ), Two straight lines 10012-formed by the movement of the horizontal direction 10007 according to the optical modulation step of the two-row image center 10012 located in the first row and the five-row image center 10025 located in the first row. 18412 and 10025-18425 are placed in parallel, dividing the two parallel lines 10038-18438 and 10039-18439 into three regions, which are again identically arranged in eight rows of image centers 10038 and 9 The triangular projection structure was formed by dividing the large triangles 10039, 10938, and 10638 formed by the movement of the image centers 10039 into nine smaller triangles having a length of 1/3. This is because the number of vertical cells B of FIG. 4 is 3, and if the number of vertical cells B is 1, the initial rows are paralleled between parallel lines formed by the horizontal movement according to the optical modulation step of two adjacent image centers located in the same row. The straight line formed by the horizontal shift according to the optical modulation step of the located image center cannot be formed and the two parallel lines are not divided. If the number of vertical cells B is 2, the optical modulation of two adjacent image centers positioned in the same row is initially performed. Between the parallel lines formed by the horizontal movement according to the step, the straight lines formed by the horizontal movement according to the optical modulation step of the image center located in one initial different row are located in parallel, and the two parallel lines are divided into two regions. It is self-evident.

FIG. 5 is an enlarged view of only a portion of the green rectangle 10010 seen in the center of FIG. 4. The distribution of the image centers 10002 uniformly shown in the center of FIG. 4 is a projection structure formed by the regular movement of the light modulation step unit projection structure. It is determined by the repetition of the inverted triangular structure by the triangular structure to the image centers (15812, 15912, 13325). As discussed above, since the vertical cell number B is 3, the two cells are initially located in two different rows between the parallel lines formed by the horizontal direction 10007 movement according to the optical modulation step of two adjacent image centers located in the same row. The two straight lines formed by the horizontal movement according to the optical modulation step of the image center are positioned in parallel so that the two parallel lines are divided into three regions, so that two adjacent image centers 10038 and 10039 are positioned in the same row. The triangular structure by the image centers 10039, 10938, 10638 formed by the horizontal movement according to the optical modulation step of the first, the nine triangular structures (10039, 15912, 15812) by the image centers located in different rows ), (15812, 13325, 13225), (15812, 15912, 13325), (15912, 13425, 13325), (13225, 10738, 10638), (13225, 13325, 10738), (13325, 10838, 10738), (13325, 13425, 10838), (13425, 10938, 10838), the triangular structure by the image centers 10039, 15912, and 15812 is formed into a three-fold enlarged structure.

FIG. 6 is a view illustrating an enlarged portion of the purple rectangle 10020 shown in the center of FIG. 5 and then the image centers are reduced to small dots or circles, according to the light modulation step of two adjacent image centers positioned in the same row as described above. Image centers (10039, 15912, 15812) formed by the horizontal movement according to the optical modulation step of the image center located at different initial rows and the triangular structure by the image centers (10039, 10938, 10638) formed by the horizontal movement. The triangular structure by) is shown. As shown in FIG. 6, the arrowed straight line 198698 moves in the horizontal direction 10007 according to the image center 10038 located in the initial three rows and eighth columns and the optical modulation steps 5, 6, and 9 of the image center. The image centers 1038, 10638, and 10938 are formed to be connected to each other, and a straight line 1995 is a negative vertical direction so as to be orthogonal to the arrowed straight line 19998 in the image center 10039 located in the first three rows and nine columns. A line drawn at 10008, and a point 19999 is an intersection point of the arrowed straight line 1986 and straight line 19997. Therefore, the angle between the straight line 19996 and the straight line 19997 drawn in the negative longitudinal direction 10006 in the image center 10039 is the mirror angle θ 10009.

FIG. 7 is a view showing the image centers being reduced to small dots or circles after only expanding the triangular structure by the three image centers 10039, 15912, and 15812 that are the basis of the projection structure shown in FIG. 6. The distance (19995) from the image center (10538) to the image center (10638) of FIG. 6 to the image center (15912) represented by the red straight line in FIG. The image center 10039 is defined as a h-offset and is represented by a red straight line in FIG. 7 through a distance (19994) to the image center 10039 from the image center 15812 of FIG. 6 to the image center 10039. The distance (19994) to η is defined as v-offset, defined as a v-offset, a straight line connecting the image center 10638 and the image center 10938 of FIG. 6, the image center 10638 and the image center. The angle between the angle (19993) between the straight lines connecting (10039) to (η) (19995) and (η) (19994) represented by the red arrowed curve in FIG. To be defined). Therefore, in the digital lithography proposed in the present invention, the triangles that are the basis of the projection structure are determined by the h-offset ξ (19995), the v-offset η (19994), and the projection shape angle Ψ (19993).

In the current display industry, when manufacturing display substrates such as TFT LCDs, pixels are arranged in a stripe configuration and a delta configuration in order to improve sharpness and obtain a clear image. In addition, there are numerous types of exposure patterns used in industrial fields such as displays, semiconductors, and printed circuit boards, and unfortunately, most of them are not simple patterns consisting of a combination of vertical lines and horizontal lines. Are complex patterns consisting of: Therefore, in order to further improve the quality of the pattern resulting from the exposure, it is necessary to adjust the projection structure formed on the substrate by the beams projected on the substrate according to the arrangement of the pixels used in manufacturing the display substrate or the shape characteristics of the exposure pattern. It can be necessary.

On the other hand, when looking at the distribution of the triangular form to the inverted triangle form formed by the image centers (10002) shown in Figures 4 to 7, when the projection shape angle Ψ (19993) is π / 4 radians or more and 3 π / 4 radians or less, When the triangle that is the basis of the projection structure becomes an isosceles triangle, and the h-offset ξ (19995) and v-offset η (19994) become the same, the pattern that appears as a result of exposure It can be seen that the degree of improvement will be improved. In the case illustrated in FIG. 7, an ideal case conforming to the delta shape, the dotted purple line 10030 is part of a circle whose image center 15812 is the origin and whose radius is the same as the h-offset ξ 199595. Since 10039 is located on the purple dotted line 10030, the v-offset η (19994) is equal to the h-offset ξ (19995), and the projection shape angle Ψ (19993) is π / 3 radians. As shown in FIG. 7, the image center 10039 moves along the purple dotted line 10030 while the image center 15158 and the image center 15912 are fixed and the h-offset ξ 199595 is fixed. When located in the image center 10039A, the v-offset η (19994A) is equal to the h-offset ξ (19995), and the projection form angle Ψ (19993A) becomes π / 2 radians and is located in the image center 10039B. The v-offset η (19994B) is equal to the h-offset ξ (19995), and the projection shape angle Ψ (19993B) is 2π / 3 radians. The image center 10039 to be considered in the present invention is located in the image center 10039S when the projection shape angle Ψ is π / 4 radians, and is located in the image center 10039E when 3π / 4 radians and the values therebetween. It is located on the purple dotted line 10030 between the centers 10039S and 10039E. Therefore, when the projection shape angle Ψ (19993) is π / 4 radians or more and 3π / 4 radians or less, the triangle that is the basis of the projection structure becomes an isosceles triangle, so that the h-offset ξ (19995) and the v-offset η ( The horizontal cell number A, the vertical cell number B, the image interval C (10003), the projection length ratio Z to the vertical image interval C * Z (10004), and the scan resolution D to the same. Correlate scan pitch P.

When the number of horizontal cells as the reference of the projection structure is A, the number of vertical cells is B, the image interval is C, and the projection length ratio is Z, the mirror angle θ is expressed by Equation 1, and the sine of the mirror angle θ is represented by Equation 2 and The cosine of the mirror angle θ is expressed as shown in Equation 3 below.

In FIG. 6, the horizontal distance 10070 from the image center 10038 to the image center 10039 is the image interval C 10003, and the vertical distance 19997 from the point 19999 to the image center 10039 is shown in FIG. 6. C * sinθ, and the distance from the image center 10638 to the image center 10039 is C * sinθ / sinψ. Since the number of vertical cells B described above is 3, a large triangle formed by the horizontal movement according to the light modulation step of two adjacent image centers located in the same row is initially applied to the light modulation step of the image centers located in two different rows. By recalling that the triangular projection structure is formed by dividing into nine small triangles whose length is one third by the horizontal movement, the v-offset η (19994) is obtained from the image center 10638 to the image center 10039. It can be seen that it is 1/3 of the distance, and when this is generalized, Cx * sinθ / (B * sinψ). Substituting Equation 2 above, the generalized v-offset η is expressed as Equation 4.

3 to 4, the distance from the image center 10038 to the image center 18184 and the distance from the image center 10012 to the image center 18412 are SQRT (A * A * C * C + B * B * C). * C * Z * Z). Therefore, the generalized h-offset ξ is expressed as Equation 5 as a function of scan resolution D.

V-offset eta of Equation 4 and h-offset of Equation 5 for obtaining an isosceles triangle having a projection shape angle Ψ (19993), which is the basis of the projection structure of the present invention, not less than π / 4 radians and not more than 3π / 4 radians. Scan resolution D having the same ξ is calculated as a function of the number of horizontal cells A, the number of vertical cells B, the projection length ratio Z and the projection shape angle Ψ as shown in Equation 6 below.

In the present invention, the scan resolution D is an integer number of light modulation steps performed while the initial light modulation step unit projection structure based on the horizontal cell number A and the vertical cell number B travels the shortest path to the position of the final light modulation step unit projection structure. must be an integer. Accordingly, under the given or predetermined projection angle π / 4 radians or more and 3π / 4 radians or less, the number of horizontal cells A, vertical cells B, and projection length such that the scan resolution D according to Equation 6 becomes an integer Determination of Z is essential. In the example shown in Figs. 2 to 7, the projection form angle Ψ is π / 3 radians, the horizontal cell number A is 10, the vertical cell number B is 3, and the projection length ratio Z is 2 / SQRT (3). . The scan resolution D obtained by substituting these into equation (6) is 84, which is an integer. However, the image spacing C of the example shown in Figs. 2 to 7 in which the sine value sin (Ψ) of the projection form angle Ψ is SQRT (3) / 2 is 10 * SQRT (3) / 2 microns, which is accurately realized by the optical system. It is a difficult value. If the size of the image cell of the example shown in Figs. 2 to 7 is multiplied by 2 / SQRT (3), the image spacing C becomes 10 microns so that it can be accurately realized by the optical system, but the projection length ratio Z is 2 / SQRT (3). Because of this, the vertical image interval of C * Z becomes 10 * 2 / SQRT (3), which is also difficult to be accurately realized by the optical system. The error in realizing such an image interval C to a vertical image interval C * Z by the optical system occurs as an error in obtaining the scan resolution D of Equation 6 as an integer. Therefore, it is necessary to allow an error in which scan resolution D can be regarded as an integer. However, since the error in determining the scan resolution D will eventually affect the degree of the pattern, the tolerance in considering the scan resolution D as an integer should be determined by the user according to the purpose. . Therefore, in the present invention, when the error when the number near the integer is integer is within the tolerance range determined by the user for the purpose, the number near the integer is regarded as an integer and is called an integer. The h-offset ξ obtained by substituting 84 into the scan resolution D of Equation 5 is about 1.091 microns, which is the same as the v-offset η according to Equation 4, and the projection structure shown in FIGS. An isosceles triangle with π / 3 radians was formed. Accordingly, one condition for obtaining an isosceles triangle-like projection structure having the same v-offset η and h-offset ξ under a given π / 4 radian or greater than 3π / 4 radian of the present invention is a scan of Equation 5 The resolution D is an integer.

However, in the present stage, if the horizontal cell number A, vertical cell number B and the projection length ratio Z are correctly determined and only one of the above conditions is satisfied, whether the actually formed projection structure becomes an isosceles triangular projection structure whose projection shape angle is Ψ. You should check FIG. 8 illustrates a projection structure formed on a substrate, which is an object to be exposed, which is reflected by three horizontal and ten micromirrors and moved from a predetermined initial position to a final position determined by 57 light modulation steps. 9 shows only the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 15700 in the projection structure of FIG. 8, and FIG. 10 shows the image centers in the projection structure of FIG. 8. Only (10002) is selected and shown. The image center 10011 in the first row and the first column moves to the image center 15711 after 57 light modulation steps, where the position of the image center 15711 is an initial four-row and nine-column image when the number of image cells increases. Will match the location of the center. FIG. 11 illustrates a projection structure formed on a substrate, which is an object to be exposed, which is reflected by three horizontal and ten micromirrors and moved from a predetermined initial position to a predetermined final position by 36 light modulation steps. 12 shows only the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 13600 in the projection structure of FIG. 11, and FIG. 13 shows the image centers ( Only 10002) is shown. The image center 10011 in the first row and the first column moves to the image center 13611 after 36 light modulation steps, and the position of the image center 13611 is the initial four-row and seven-column image when the number of image cells increases. Will match the location of the center. The image center 10023 of the first 2 rows and 3 columns is the image center 10035. 8 to 10, the projection shape angle Ψ is π / 3 radians, the horizontal cell number A is 8, the vertical cell number B is 3, the projection length ratio Z is 2 / SQRT (3), and the image interval C is 10 * SQRT (3) / 2 micron, and the scan resolution D obtained by substituting these in equation (5) is 57, which is also an integer. 11 to 13, the projection form angle Ψ is π / 3 radians, the horizontal cell number A is 6, the vertical cell number B is 3, the projection length ratio Z is 2 / SQRT (3), and the image interval C is 10 * SQRT (3) / 2 micron, and the scan resolution D obtained by substituting them into Equation 5 is 36, which is also an integer. However, the projection structure shown in the green rectangle 10010 enlarged in the center of FIGS. 10 to 10 appears as a projection structure different from the prediction, in which the projection shape angle Ψ is neither π / 3 radians nor isosceles triangles. The projection structure shown in the green rectangle 10010 in which the central portion of FIGS. 13 to 13 is enlarged is a triangle structure in which the projection shape angle Ψ is not π / 3 radians and the difference between v-offset η and h-offset ξ is large. It is difficult to see, but it is different from projections. Therefore, even when the projection shape angle Ψ is π / 4 radian or more and 3π / 4 radian or less, even if the horizontal cell number A, vertical cell number B, and projection length ratio Z are correctly determined so that one of the above conditions that the scan resolution D becomes an integer is satisfied. In fact, the projection structure actually obtained may not be an isosceles triangle-shaped projection structure with a projection angle of Ψ, thus leading to the need for additional conditions. Therefore, a more in-depth analysis of the correlation between the projection shape angle Ψ, scan resolution D, horizontal cell number A, vertical cell number B, projection length ratio Z and the projection structure is necessary.

The large triangular structure of the three image centers 10039, 10938, and 10638 described above takes into account the angle 19993 formed by the arrowed straight line 1986 and the straight line connecting the image center 10638 and the image center 10039. For the case where the projection shape angle Ψ is π / 4 radians or more and π / 2 radians or less, generalizing it based on the J row I column image center of any K th light modulation step, in other words, the J row I of the K th light modulation step Triangle formed by the image center of the column and the image center of row J (I-1) of the (K + L) -th optical modulation step and the image center of the row J (I-1) of the (K + L) -th optical modulation step. Is equal to, and the projection shape angle Ψ is greater than π / 2 radians and less than 3π / 4 radians in consideration of the angle formed by the arrowed straight line 1986 and the straight line connecting the image center 10109 and the image center 10039 This is based on the J row I column image center of any K th light modulation step. In other words, the image center of row J of column I of the Kth light modulation step and the image center of row (I-1) of row J of the (K + L) th light modulation step and the J of the (K + LB) th light modulation step. It is identical to the triangle formed by the image center of row (I-1). In this case, since B is the number of vertical cells and L is the number of light modulation steps, it should be an integer according to the discussion of scan resolution D of Equation 6 above. In the present invention, L is defined as the unit step number in the sense of the unit light modulation step number which is the reference of the large triangular structure formed by the horizontal movement according to the optical modulation step of the image center, and the unit step number L is defined as the projection shape angle Ψ Is determined by the number of horizontal cells A, the number of vertical cells B, the projection length ratio Z and the scan resolution D.

As discussed in the description of FIG. 6, the horizontal distance 10070 from the image center 10038 to the image center 10039 is the image interval C 10003, and from the point 19999 to the image center 10039. The vertical distance (19997) is C * sinθ, and the angle between the straight line (19996) and the straight line (19997) drawn in the negative longitudinal direction (0606) in the image center (10039) is the mirror angle θ (10009). 19999 to the horizontal distance 10060 from the image center 10039 is C * sinθ * sinθ. The distance (10050) from the image center (10638) to the image center (10039) is generalized as three times the distance (19994) η from the image center (15812) to the image center (10039), and becomes B * η, and v- When the offset η and h-offset ξ are equal, the lateral distance 10050 from image center 10638 to point 19999 is B * ξ * cosΨ * cosθ. 6, the horizontal distance 10040 from the image center 10038 to the image center 10638 is a distance obtained by subtracting the sum of the horizontal distances 10050 + 10060 from the horizontal distance 10070. 10070-10050-10060. In addition, the horizontal distance 10040 from the image center 10038 to the image center 10638 is a product of the unit number of steps L and the distance 10080, which is the distance from the image center 10538 to the image center 10638. Since (19995) is the h-offset ξ and the horizontal distance 10080 from the image center 10538 to the image center 10638 is ξ * cosθ, it is L * ξ * cosθ. Since the horizontal distance 10040 is L * ξ * cosθ, which is 10070-10050-10060, this relationship is expressed as Equation 7.

Recalling that Equation 7 has the same v-offset η and h-offset ξ, and substituting Equation 2, Equation 3 and Equation 4 to sum up the unit step number L, It calculates as a function of the horizontal cell number A, the vertical cell number B, the projection length ratio Z, and the projection shape angle Ψ as in Equation 8.

 According to the discussion that the unit step number L is also an optical modulation step number like scan resolution D, it should be an integer, and v-offsets η and h− under a projection form angle Ψ of greater than or equal to π / 4 radians of less than 3π / 4 radians of the present invention. Another condition for obtaining an isosceles triangle-like projection structure with the same offset ξ is that the unit step number L in Equation 8 becomes an integer.

In Equation 8, the projection shape angle Ψ is π / 3 radians, the horizontal cell number A is 10, the vertical cell number B is 3, the projection length ratio Z is 2 / SQRT (3), and the same as in Figs. The number of unit steps L obtained by substituting l is 6 as an integer. Therefore, in the case of FIGS. 2 to 7, two isotropic triangular projection structures having the same v-offset η and h-offset ξ under the given π / 4 radians or more and 3π / 4 radians or less of the present invention are obtained. The conditions are all satisfied. Thus, the results shown in FIGS. 2 to 7 form an isosceles triangle projection structure in which the projection shape angle Ψ is π / 3 radians, and the large triangle structure described above has a projection shape angle Ψ of π / 4 radians or more π / 2 or less radians and the number of unit steps L is 6, so that the image center 10039 of the 3rd row and 9th column of the 0th light modulation step and the image center of the 3rd row (9-1) of the (0 + 6 + 3) th light modulation step (10938) and (0 + 6) th light modulation step are formed into a triangle formed by the image center 10638 of the third row (9-1) columns. In Equation 8, the projection form angle Ψ is π / 3 radians, the horizontal cell number A is 8, the vertical cell number B is 3, the projection length ratio Z is Z / 2 / SQRT (3), and FIGS. 8 to 10 The unit step number L obtained by substitution is the same as 4.50000 ....., which is not an integer. Therefore, in the case of Figs. 8 to 10, two isotropic triangular projection structures having the same v-offset η and h-offset ξ under the given π / 4 radians or more and 3π / 4 radians or less of the present invention are obtained. Only one of the conditions is met. Therefore, the results shown in FIGS. 8 to 10 formed a projection structure different from the prediction, in which the projection angle Ψ was neither π / 3 radians nor isosceles triangles.

In Equation 8, the projection form angle Ψ is π / 3 radians, the horizontal cell number A is 6, the vertical cell number B is 3, the projection length ratio Z is Z is 2 / SQRT (3), and FIGS. 11 to 13 are shown. The unit step number L obtained by substitution is the same as 3, which is an integer. Therefore, in the case of Figs. 11 to 13, two isotropic triangular projection structures having the same v-offset η and h-offset ξ under the given π / 4 radians or more and 3π / 4 radians or less of the present invention are obtained. The conditions are all satisfied. However, the results shown in FIGS. 11 to 13 show that the projection shape angle Ψ is not π / 3 radians and that the difference between the v-offset η and the h-offset ξ is large and difficult to be considered as a triangular structure. . Therefore, when the projection shape angle Ψ is π / 4 radian or more and 3π / 4 radian or less, the horizontal cell number A, vertical cell number B, Even if the projection length ratio Z is correctly determined, the actual projection structure obtained may not be an isosceles triangle-shaped projection structure with the projection shape angle Ψ, thus leading to the need for another additional condition. Referring to FIG. 13, an initial two rows and three columns of image centers are formed on a straight line formed by the horizontal direction 10007 movement of the image center 10011 of the first one row and one column to the image center 13601 by 36 optical modulation steps. 10023) and the image center 10035 of the first three rows and five columns are located, and it can be seen that the projection structure having a different shape from that of the prediction whose v-offset η is larger than the h-offset ξ is formed. This result is different from the case of FIGS. 4 and 10, since the horizontal cell number A of FIG. 11 to FIG. 13 is 6, the vertical cell number B is 3, and the horizontal cell number A divided by the vertical cell number B becomes an integer. This is because a straight line formed by the horizontal movement according to the optical modulation step of the image center located in another row is not formed between the parallel lines formed by the optical modulation step of the image centers of two adjacent columns located in the row. . Therefore, an isosceles with the same v-offset η and h-offset ξ under the projection form angle Ψ of given π / 4 radians or more and 3π / 4 radians or less so that the value of the horizontal cell number A divided by the vertical cell number B is not an integer. Another condition to be added to the above two conditions for obtaining a triangular projection structure is that a relationship between each other (coprime) is established between the horizontal cell number A and the vertical cell number B. Up to this stage, we have only discussed the case where the projection angle is π / 3 radians. Therefore, even when projection angles are angles with Ψ, if the horizontal cell number A, vertical cell number B, and projection length ratio Z are correctly determined so that the three conditions are satisfied, the actually formed projection structure has an isosceles triangle shape with the projection shape angle Ψ. You need to make sure that it is a projection structure.

In the case of FIGS. 2 to 7, when only the projection shape angle is set to 2π / 3 radians while all other conditions are maintained, the scan resolution D obtained by Equation 6 is 84, and the projection shape angle is π / 3 radians. It is an integer similar to the case, and the unit step number L obtained by Formula (8) is 9, which is a different integer from the case where the projection form angle is π / 3 radians. In this case, as described above, the angle formed by the straight line 19998 and the straight line connecting the image center 10109 and the image center 10039 to the large triangular structure by the three image centers 10039, 10938 and 10638 is formed. Considering that the projection form angle Ψ is larger than π / 2 radians and less than 3π / 4 radians, the large triangular structure according to the above description has a unit step L of 9, so that the image centers of 3 rows and 9 columns of the 0th optical modulation step Image center (10938) in the third row (9-1) of the (10039) and (0 + 9) th optical modulation steps and image center in the third row (9-1) of the (0 + 9-3) th optical modulation step It is obvious that the 1010638 is formed by a triangle to form the same projection structure as that of FIGS. 2 to 7.

FIG. 14 illustrates a projection structure formed on a substrate, which is an object to be exposed, which is reflected by three horizontal and vertical micromirrors and moves from a predetermined initial position to a predetermined final position by 68 light modulation steps. 15 shows only the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 16800 in the projection structure of FIG. 14, and FIG. 16 shows image centers (FIG. Only 10002) is shown. The image center 10011 of the first one-row column moves to the image center 16811 after 68 light modulation steps, and the position of the image center 16811 is the initial four-row, eleven-column image when the number of image cells increases. Will match the location of the center. 14 to 16, the projection shape angle Ψ is π / 2 radians, the horizontal cell number A is 10, the vertical cell number B is 3, and a mutual relationship is established between the horizontal cell number A and the vertical cell number B, and the projection length ratio Z is 2, the image interval C is 10 * SQRT (3) / 2 micron, and the scan resolution D obtained by substituting them into Equation 5 is an integer 68, and the number of unit steps L obtained by substituting them into Equation 8 is an integer 5. Thus, the schnitch P is about 1.485 microns and the mirror angle θ 10009 is about 30.96 degrees. 14-16 is an isosceles triangle with the same v-offset η and h-offset ξ under the projected perception Ψ of a given π / 4 radian or more and 3π / 4 radian or less, as in the case of FIG. 4. All three conditions, such as the number of horizontal cells A, the number of vertical cells B, and the projection length ratio Z, for which the projection structure is formed are satisfied. Therefore, the h-offset ξ obtained by substituting 68 for the scan resolution D of Equation 5 is 1.485 microns, which is the same as the v-offset η according to Equation 4 and the scan pitch P, and the central part of FIGS. 16 to 16. The projection structure shown in the green square (10010), which has been enlarged, has a right angled isosceles triangle shape in which the projection shape angle Ψ is π / 2 radians.

In summary, the number of horizontal cells A and the number of vertical cells for forming an isosceles triangular projection structure having the same v-offset η and h-offset ξ under a given π / 4 radian or more and 3π / 4 radian or less according to the present invention. The three conditions of B and the projection length ratio Z determination are that condition 1 is a mutual relationship between the horizontal cell number A and the vertical cell number B, condition 2 is the scan resolution D of Equation 5 becomes an integer, condition 3 The unit step number L in this expression (8) becomes an integer. In the present invention, the h-offset ξ of Equation 5 is newly denoted by S to define scan resolution or scan direction resolution. 4 and 16 satisfying all three conditions of the present invention, the distance h-offset ξ to scan resolution S, which is the moving distance of the optical modulation step unit image center, is also the v-offset η, It is also the scan pitch P which is the unit moving distance of the board | substrate which is a corresponding to-be-exposed body. However, in the cases described below, the scan pitch P and the scan resolution S may not be the same. At this time, even if the scan pitch P increases above the scan resolution S, the efficiency of lithography may be considered to be improved if the h-offset ξ and the v-offset η remain the same at the scan resolution S. In the present invention, the ratio of the scan pitch P to the scan resolution S, i.e., P / S is defined as a scan ratio, denoted by H, and the efficiency of lithography is determined by the scan ratio. Therefore, as in the case of FIGS. 4 and 16, when the scan pitch P is equal to the scan resolution S, the scan ratio is 1.

A horizontal cell number A, a vertical cell number B, and a projection to form an isosceles triangle with the same v-offset η and h-offset ξ under a given π / 4 radian or greater than 3π / 4 radian of the present invention. In the case of FIGS. 2 to 7, which satisfy all three conditions of the length ratio Z crystal, an equiangular projection angle Ψ between isosceles is π / 3 radian or 2π / 3 radian, so that all the intervals between the image centers are the same equilateral triangle. 14 to 16, and in the case of FIGS. 14 to 16, a right-angled isosceles triangle, that is, a square structure, since the projection angle Ψ between the isosceles is π / 2 radians. 1/2 of. Thus, the projection structure angle Ψ is π / 3 radians or 2π / 3 radians and the projection structure satisfying all three conditions of the present invention is the delta-type projection structure, and the projection shape angle Ψ is π / 2 radians and the bone The projection structure that satisfies all three conditions of the invention is the stripe type projection structure. If the angle between the isosceles projection angle Ψ is greater than π / 4 radians and less than π / 3 radians, greater than π / 3 radians and less than π / 2 radians, greater than π / 2 radians and less than 2π / 3 radians And larger than 2π / 3 radians and smaller than 3π / 4 radians, the projection structure that satisfies all three conditions of the present invention may be an isosceles triangle type projection shape or a rhombus type projection structure. The delta-shaped projection structure and the stripe-shaped projection structure are also an isosceles triangle type projection structure or a rhombus type projection structure. However, in the case of isosceles triangles with projection angles Ψ between isosceles other than π / 3 radians, 2π / 3 radians and π / 2 radians, the display produced by the exposure process may be improved or improved. In addition, the delta type projection structure or the stripe type projection structure is not advantageous for improving the sharpness of the semiconductor and the printed circuit board and obtaining a clear image. Therefore, the description mainly focuses on the delta projection structure (hereinafter, referred to as a delta projection structure) and the stripe projection structure (hereinafter, referred to as a stripe projection structure), and the projection form including a delta projection structure and a stripe projection structure. For an isosceles triangular projection structure where angle Ψ is π / 4 radians or more and 3π / 4 radians or less (hereafter isosceles triangle projection structure), the delta projection structure and the stripe projection structure without specifying the value of the projection form angle Ψ As discussed in the discussion.

2. Delta Projection Structure

The three conditions for determining the horizontal cell number A, vertical cell number B, and projection length ratio Z forming the delta projection structure of the present invention are such that condition 1 is mutually established between horizontal cell number A and vertical cell number B, and condition 2 The scan resolution D calculated by Equation 9 below by substituting π / 3 radians or 2π / 3 radians into the projection form angle Ψ in Equation 6 becomes an integer.

  In condition 3, when the projection form angle Ψ is π / 3, the unit step number L calculated by Equation 10 below by substituting π / 3 radians into the projection form angle Ψ in Equation 8 becomes an integer, and the projection In the case where the shape angle Ψ is 2π / 3, the unit step number L having SQRT (3) * A / (2 * Z) + B / 2, in which 2π / 3 radians is substituted for the projection shape angle Ψ, becomes an integer.

The scan resolution S of the delta projection structure of the present invention is obtained by substituting π / 3 radians or 2π / 3 radians into the projection form angle Ψ of Equation 4 or by substituting Equation 9 into the scan resolution D of Equation 5. It calculates by 11.

3. Striped Projection Structure

Three conditions for determining the horizontal cell number A, the vertical cell number B, and the projection length ratio Z forming the stripe projection structure of the present invention are such that condition 1 is mutually established between horizontal cell number A and vertical cell number B, and condition 2 The scan resolution D calculated by the following Equation 12, in which π / 2 radians is substituted into the projection form angle Ψ of Equation 6, becomes an integer.

  The condition 3 is the unit step number L calculated by Equation 13 below in which π / 2 radians is substituted for the projection form angle Ψ in Equation 8, which is an integer.

The scan resolution S of the stripe projection structure of the present invention is calculated by Equation 14 obtained by substituting π / 2 radians into the projection form angle Ψ of Equation 4 or by substituting Equation 12 into the scan resolution D of Equation 5. .

4. Isosceles triangle projection structure

When the projection shape angle Ψ is π / 4 radians or more and 3 π / 4 radians or less, three conditions for determining the horizontal cell number A, the vertical cell number B, and the projection length ratio Z forming the isosceles triangular projection structure of the present invention are horizontal. A mutual relationship is established between the number of cells A and the number of vertical cells B, and the scan resolution D calculated by substituting a value of π / 4 radian or more and 3π / 4 radian or less into the projection form angle Ψ of equation (6) is an integer. In the condition 3, the unit step number L calculated by substituting π / 4 radians or more and 3π / 4 radians or less into the projection form angle Ψ of Equation 8 is an integer, and the isosceles triangular projection structure of the present invention. The scan resolution S of is obtained by substituting a value of π / 4 radians or more and 3π / 4 radians or less into the projection form angle Ψ of Equation 4 by v-offset η or the scan resolution D of Equation 5 Scan resolution D of an isosceles triangular projection structure It is the same as h-offset ξ obtained by substitution.

5. Triangle Projection Structure

Looking at a digital lithography apparatus using a spatial light modulator that is currently developed or under development, the projection length ratio Z is all 1. That is, the image interval C between the image centers and the vertical interval C * Z between the image centers are the same. Considering the delta projection structure in which the projection angle Ψ is π / 3 radians or 2π / 3 radians, the scan resolution D in Equation 9 and the unit step number L in Equation 10 are integers when the projection length ratio Z is 1. It can be seen that there are no two integer horizontal cells A and B horizontal cells in mutual relation. That is, when the projection length ratio Z is 1, there is no delta projection structure in which the projection form angle Ψ is π / 3 radians or 2π / 3 radians. Therefore, in the present invention, based on the results of the above discussion, by including the unit step number L in the calculation and calculation of the scan resolution D, a digital lithography apparatus using a spatial light modulator with a projection length ratio Z of 1 or being developed is being developed. A method of forming a projection structure similar to an isosceles triangle including an equilateral triangle and an equilateral isosceles triangle, and including a unit step number L in calculating and calculating the scan resolution D, the projection length ratio Z is 1 or 1 In other cases, a method of forming a similar projection structure (hereinafter, referred to as a triangular projection structure) to an isosceles triangle including an equilateral triangle and an isosceles triangle applicable to both will be described as follows.

When the projection length ratio Z is 1, the unit step number L of the stripe projection structure of Equation 13 is equal to the horizontal cell number A, and the unit step number L of the isosceles triangle projection structure of Equation 8 is sinΨ * A-cosΨ * B. On the other hand, when the projection length ratio Z is 1, the scan resolution D of the isosceles triangle projection structure of Equation 6 is sinΨ * (A * A + B * B), and it is assumed that the unit step number L is equal to the horizontal cell number A. The numerator is multiplied by A, the denominator by L, Ψ sinΨ * A-cosΨ * B, and ROUND {(B1 / B2), 0} divided by B1 divided by B2, rounded to the nearest decimal place to be an integer. In this case, the scan resolution D which is an integer of the triangular projection structure in which the projection length ratio Z is 1, the unit step number L is equal to the horizontal cell number A, and the projection form angle is Ψ is calculated by the following equation (15).

The denominator in equation (15) is 1 when the projection form angle Ψ is π / 2 radians, and less than 1 when the projection form angle Ψ is less than or equal to π / 2 radians, and is greater than 1 when the projection form angle Ψ is greater than π / 2 radians. Big. Therefore, when the projection length ratio Z is 1, of course, it may be somewhat different depending on the number of horizontal cells A and the number of vertical cells B, but is calculated by substituting a value of π / 4 radians or more and π / 2 radians or less into the projection form angle Ψ in equation (15). The h-offset ξ calculated by substituting one scan resolution D into Equation 5 is equal to or less than the v-offset η calculated by Equation 4, but is larger than π / 2 radians in the projection angle Ψ in Equation 15. The h-offset ξ calculated by substituting the scan resolution D calculated by substituting a value less than 3π / 4 radians into Equation 5 is larger than the v-offset η calculated by Equation 4. Further, as the number of vertical cells B increases and the projection shape angle Ψ moves away from π / 2 radians, the difference between h-offset ξ and v-offset η increases. Therefore, in order to use the scan resolution D of Equation 15, it is recommended to determine the projection form angle Ψ as π / 2 radians or less and the number of vertical cells B to 1, and in another method in consideration of the disadvantage of Equation 15 The scan resolution D of the triangular projection structure is obtained by integrating the scan resolution D with the number of steps L as follows.

When the projection length ratio Z is 1, the scan resolution D of the isosceles triangle projection structure of Equation 6 is sinΨ * (A * A + B * B), and the number of unit steps L of the isosceles triangle projection structure of Equation 8 is sinΨ * A-cosΨ * B. If Equation 8 is summed and sinΨ is written as a function of L, it becomes (L + cosΨ * B) / A, and ROUND {sinΨ * A-cosΨ * B, 0}, where sinΨ * A-cosΨ * B is integerized to L. Substituting for, sinΨ is (ROUND {sinΨ * A-cosΨ * B, 0} + cosΨ * B) / A. By substituting this again into sinΨ in Equation 6, another scan resolution D of a triangular projection structure having a projection shape angle of Ψ is calculated by Equation 16 below.

If the projection length ratio Z is 1 or not 1, the scan resolution D of the isosceles triangle projection structure of Equation 6 is sinΨ * (A * A / Z + B * B * Z), and the isosceles triangle projection structure of Equation 8 The unit step number L of is sinΨ * A / Z-cosΨ * B. If Equation 8 is summed and sinΨ is written as a function of L, it becomes (L + cosΨ * B) * Z / A, and ROUND {sinΨ * A / Z- is an integer number of sinΨ * A / Z-cosΨ * B in L. If you substitute cosΨ * B, 0}, sinΨ is (ROUND {sinΨ * A / Z-cosΨ * B, 0} + cosΨ * B) * Z / A. By substituting this again into sinΨ in Equation 6, another scan resolution D of the triangular projection structure having the projection angle of Ψ, which is applicable even when the projection length ratio Z is not 1, is calculated by Equation 17 below. .

In the triangular projection structure of the present invention, if the projection shape angle Ψ is π / 2 radians, the scan resolution D of equations (15), (16), and (17) is equal to the scan resolution D of the stripe projection structure of equation (12). If v-offset η of 4 and h-offset ξ to Equation 5 to scan resolution S are the same, but the projection angle Ψ is not π / 2 radians, then v-offset η of Equation 4 and h-offset of Equation 5 ξ to scan resolution S may be different. Therefore, in the present invention, the equation is divided into h-offset ξ to scan resolution S calculated by substituting the scan resolution D of Equation 15 or Equation 16 or Equation 17 into Equation 5 under the triangular projection structure. V-offset η is newly denoted by G and defined as fixed resolution.

In order to verify the delta projection structure, the stripe projection structure, the isosceles triangle projection structure and the triangular projection structure described above, and to determine the importance of condition 2 and condition 3 of the present invention and the effect on the projection structure, the horizontal 600 A micromirror array of optical modulators consisting of 180 vertical and vertical mirrors emits a light beam having a Gaussian illuminance distribution with a Gaussian beam radius of 0.5 microns to a moving substrate at 90 degrees and 60 degrees in the horizontal direction 10007. Discrete exposure process simulations were performed to create a pattern consisting of straight lines with a line width of 5 microns tilted at 45 degrees and 0 degrees. All exposure simulation results are scaled by fixing the maximum illuminance to 1 for accurate comparison, and the illuminance becomes closer to 1 as the color becomes red.

17 are the results of the delta projection structure simulation. 17 (a) and 17 (b), the image interval C is 10 * SQRT (3) / 2 micron, the horizontal cell number A is 10, the vertical cell number B is 3, and the projection length ratio is the same as that of FIGS. Since Z is determined to be 2 / SQRT (3), all three conditions are satisfied, and the scan resolution D is calculated as 84 by Equation 6 or 9, and the unit step number by Equation 8 or 10. This is the case where L is calculated as 6. Fig. 17 (a) shows the distribution of light beams irradiated onto the substrate in a stationary state, and Fig. 17 (b) shows that scan pitch P is equal to scan resolution S and fixed resolution G because scan resolution D is 84 and scan ratio is 1; As a result of simulation with the same 1.091 micron, the delta projection structure of the equilateral triangle is shown as expected. Fig. 17 (c) shows that the condition 2 of the present invention is not satisfied, wherein the image interval C is 10 * SQRT (3) / 2 microns, the horizontal cell number A is 12, the vertical cell number B is 1, and the projection length ratio Z is By determining 12 / (5 * SQRT (3)), the scan resolution D is calculated as 91.2 by equation (6) or (9), and the unit step number L is calculated by 7 by equation (8) or equation (10). Since the scan ratio P is 1, the scan pitch P is simulated using a scan pitch P of 1.147 microns which is equal to the scan resolution S calculated from π / 3 radians to π / 3 radians. Despite setting the projection angle Ψ to π / 3 radians, the scan resolution D is not an integer so that the h-offset ξ or scan resolution S does not remain at v-offset η or fixed resolution G and the delta projection structure collapses as a whole. The result was a loss. Fig. 17 (d) shows that when condition 3 of the present invention is not satisfied, the image interval C is 10 * SQRT (3) / 2 micron, the number of horizontal cells A is 12, the number of vertical cells B is 1, and the projection length ratio Z is shown. Is determined by 2 / SQRT (3), the scan resolution D is calculated to be 109 by equation (6) or (9), the unit step number L is calculated by 8.5 by equation (8) or equation (10), and the scan ratio is The result is a simulation of the scan pitch P as 0.958 microns, which is equal to the scan resolution S calculated from π / 3 radians to the fixed resolution G, because it is 1. Although the projection shape angle Ψ was set to π / 3 radians, similar results were obtained with the stripe projection structure rather than the delta projection structure. This is because the unit step number L becomes 8.5 instead of an integer, so that the delta projection structure is virtually formed by the virtual image center. In other words, the delta projection structure of the triangle with B equal to 1 includes the image center of row J of column I of the Kth light modulation step and the image center of row (I-1) of row (I +) of the (K + 8.5) th light modulation step. 7.5) The image center of row J (I-1) of the light modulation step is virtually formed, and the actual projection structure is the image center of row J (I-1) of the (K + 9) light modulation step and (K). It was formed as an image center of row J (I-1) of the +8) th light modulation step. Therefore, in the actual projection structure, since the scan resolution D is an integer, the h-offset ξ or the scan resolution S is maintained at 0.958 microns as it is, but since the decimal point portion of the unit step number L is 0.5, the projection form angle Ψ is π / 2 radians. This is similar to the stripe projection structure, but the actual v-offset η or the fixed resolution G is 0.830 microns, forming a similar projection structure rather than an exact stripe projection structure.

18 are results of a triangular projection structure simulation. In the triangular projection structure simulation of Fig. 18A, the projection length ratio Z is 1 and the number of vertical cells B is 1 while the other conditions are maintained in the simulation of Fig. 17B, and the scan is performed according to the equation (15). This is the case where the resolution D is calculated to be 107. Fig. 18 (a) shows a simulation result of scanning pitch P of 0.812 microns which is equal to scan resolution S because scan resolution D is 107 and scan ratio is 1, and as expected, v-offset? Or fixed resolution G is 0.995 micron and h. Show a triangular projection structure with an offset ξ or scan resolution S of 0.812 microns. In order to determine the influence due to the error of the projection shape angle, the triangular projection structure simulations of FIGS. 18 (b) and (c) have the projection length ratios while maintaining all other conditions in the simulation of FIG. 17 (b). Z was set to 1, π / 3 radians (i.e., 60 degrees), and the projection angle Ψ was 58 degrees, and the number of vertical cells B was changed. The number of vertical cells B of FIG. 18 (b) is 1, which is the same as that of FIG. 18 (a), and since the scan resolution D of Equation 16 is 86 and the scan ratio is 1, the scan pitch P is 1.012 microns equal to the scan resolution S. As predicted, the triangular projection structure is similar to the equilateral triangle with v-offset η or fixed resolution G of 1.016 micron and h-offset ξ or scan resolution S of 1.012 micron. The number of vertical cells B of FIG. 18 (c) is 3, which is the same as that of FIG. 17 (b), and the scan resolution D according to Equation 16 is calculated as 94 and the scan ratio is 1. As predicted, the triangular projection structure is similar to the equilateral triangle with v-offset? Or fixed resolution G of 0.978 microns and h-offset? Or scan resolution S of 0.962 microns. In the case of Fig. 18 (b), even if the projection shape angle Ψ is π / 3 radians, the scan resolution D according to Equation 16 is calculated to be 86. Furthermore, if the projection shape angle Ψ is a value between 56 and 62 degrees, The scan resolutions D according to (16) are all 86. Therefore, even if the projection shape angle Ψ is set to π / 3 radians or a value between 56 and 62 degrees, it is obvious that the simulation result is not the same as that of Fig. 18 (b). This is because the number of horizontal cells A is small and, of course, may vary somewhat depending on the number of horizontal cells A and B, but this indicates that there is a need to allow an error in the projection form angle Ψ values of the delta projection structure and the stripe projection structure. . Therefore, in the present invention, the tolerance is ± π / 36 radians, and the projection angle Ψ is from π / 3-π / 36 radians to π / 3 + π / 36 radians and 2π / 3-π / 36 radians to 2π / A case of up to 3 + π / 36 radians is regarded as a delta projection, and π / 3-π / 36 radians to π / 3 + π / 36 radians are expressed in π / 3 radians and 2π / 3-π / 36 radians To 2π / 3 + π / 36 radians in 2π / 3 radians, where the projection angle Ψ is from π / 2-π / 36 radians to π / 2 + π / 36 radians is considered a striped projection structure. And denote π / 2 radians from π / 2-π / 36 radians to π / 2 + π / 36 radians. The triangular projection structure simulation of FIG. 18 (d) is expressed by Equation 17 with the projection shape angle Ψ as 5π / 12 radians (ie, 75 degrees) while all other conditions remain the same in the simulation of FIG. 17 (b). The scan resolution D is calculated as 98. 18 (d) shows that the scan resolution P is 98 and the scan ratio is 1, and the scan pitch P is simulated with 0.935 microns equal to the scan resolution S. As expected, the v-offset η or the fixed resolution G is 0.978 microns and h. It shows a triangular projection structure with a 75 degree projection angle with an offset ξ or scan resolution S of 0.935 microns. In the case of Fig. 18 (d), when the projection form angle Ψ is pi / 3 radians, the scan resolution D according to equation (17) becomes 84. This is the same as the scan resolution 84 calculated by Equation 6 or Equation 9 in the simulation of FIG. This is because all of the above conditions 1, 2 and 3 are satisfied. Therefore, in the case of FIG. 18 (d), when the projection shape angle Ψ is π / 3 radians, the simulation result is shown even when the scan resolution D according to Equation 17 is used. The same as 17 (b) is obvious even if not shown. Therefore, if all of the above conditions 1, 2 and 3 are satisfied, the scan resolution D according to Equation 17 becomes the scan resolution D according to Equation 6 when the projection shape angle is Ψ, and the projection shape angle is π / 3 radians. In this case, the scan resolution D according to Equation 17 becomes the scan resolution D according to Equation 9, and the scan resolution D according to Equation 17 becomes the scan resolution D according to Equation 12 when the projection form angle is π / 2 radians.

19 are the results of the stripe projection structure simulation. 19 (a) and 19 (b), the image interval C is 10 * SQRT (3) / 2 micron, the horizontal cell number A is 10, the vertical cell number B is 3, and the projection length ratio is the same as that of FIGS. When Z is determined to be 2, all of the above conditions 1, 2 and 3 are satisfied, and the scan resolution D is calculated as 68 according to Equation 6, Equation 12 or Equation 17, and Equation 8 or Equation 13 This is the case where the unit step number L is calculated to be 5 by. Fig. 19 (a) shows the distribution of the light beams irradiated onto the substrate in the stationary state, and Fig. 19 (b) shows that the scan pitch P is 1.485 microns which is equal to the scan resolution S since the scan resolution D is 68 and the scan ratio is 1; The simulation results show that, as expected, a stripe projection structure in the form of a right isosceles triangle is shown. 19 (c) is the case where condition 3 of the present invention is not satisfied, the image interval C is determined to be 9 microns, the number of horizontal cells A is 10, the number of vertical cells B is 3, and the projection length ratio Z is 4/3. Equation 6 or 12 calculates the scan resolution D as 87, unit 8 L calculates the number of unit steps L as 7.5, and the scan ratio P is 1. This is a simulation result of 1.114 microns which is equal to the scan resolution S to the fixed resolution G calculated in π / 2 radians. Although the projection shape angle Ψ is set to π / 2 radians, the results are similar to the delta projection structure, not the stripe projection structure. This is because the unit projection number L becomes 7.5 instead of an integer, so that the stripe projection structure is virtually formed by the virtual image center. That is, a stripe projection structure with a large triangle having a B of 3 has an image center of row J of column I of the K th light modulation step and an image center of row J (I-1) of the (K + 7.5) th light modulation step and (K +4.5) is virtually formed as the image center of row J (I-1) of the light modulation step, and the actual projection structure is the image center of row J (I-1) of the (K + 7) light modulation step and ( It was formed as an image center of row J (I-1) of the K + 4) th light modulation step. Therefore, in the actual projection structure, since the scan resolution D is an integer, the h-offset ξ or the scan resolution S is kept at 1.114 microns as it is, but since the decimal point portion of the unit step number L is 0.5, the projection form angle Ψ is arctan (1 / 0.5) The radian is 63.435 degrees, similar to the delta projection structure, but the actual v-offset η or fixed resolution G is 1.246 microns, forming a similar projection structure rather than an exact delta projection structure. 19 (d) is a case where condition 3 of the present invention is not satisfied, and the image interval C is determined to be 10 microns, the number of horizontal cells A is 12, the number of vertical cells B is 5, and the projection length ratio Z is 9/5. Equation 6 or 12 calculates the scan resolution D to 125, the unit step number L is calculated to be 6.6666667 by Equation 8 or 13, and the scan pitch P is 1, so that the scan pitch P It is the simulation result set to 1.2 micron which is equal to the scan resolution S calculated from (pi) / 2 radians, and the fixed resolution G. Although the projection shape angle Ψ is set to π / 2 radians, the results are similar to the triangular projection structure, not the stripe projection structure. This is because the unit step number L becomes 6.6666667 instead of an integer, so that the stripe projection structure is virtually formed by the virtual image center. That is, a stripe projection structure with a large triangle having a B of 5 includes image centers in row J of column I of the K-th optical modulation step and image centers in row J (I-1) of (K + 6.6666667) -th optical modulation step and (K It is virtually formed as the image center of row J (I-1) of the (1) th light modulation step, and the actual projection structure is the image center of row J (I-1) of the (K + 6) th light modulation step and ( It was formed as an image center of row J (I-1) of the K + 1) th light modulation step. Therefore, in the actual projection structure, since the scan resolution D is an integer, the h-offset ξ or the scan resolution S is maintained at 1.2 microns as it is, but since the decimal point portion of the unit step number L is 0.6666667, one angle is arctan (1 / 0.6666667). It formed a triangular projection structure of 56.310 degrees radians, 1.442 microns on one side, and 71.565 degrees radians on the other, arctan (1 / (1-0.6666667)), and 1.265 microns on the other. According to the discussion of FIG. 17 to FIG. 19, if the resolution 2 of the present invention is not satisfied because the scan resolution D is not an integer, the v-offset η or the fixed resolution G is maintained but the h-offset by the scan resolution D. If ξ or scan resolution S is not maintained, the projection structure is totally collapsed, and if the condition 3 of the present invention is not satisfied because the unit step number L is not an integer, the scan resolution D is an integer, so the h-offset ξ or scan resolution S is It is maintained, but the projection angle Ψ and v-offset η or fixed resolution G are not maintained, resulting in the projection structure changing to another unwanted structure and into a triangular structure rather than an isosceles. Therefore, the importance of the conditions of the present invention has been verified once more, and the triangular projection structure close to the isosceles triangle, the isosceles triangle projection structure, as well as the delta projection structure of the equilateral triangle shape in which the uniform scan resolution in all directions is maintained and the horizontal It has been confirmed that digital lithography using a precise spatial light modulator is possible by forming a stripe projection structure of a right-angled isosceles triangle shape which maintains uniform scan resolution in the vertical direction.

6. Projection Length Ratio

In a digital lithography using a spatial light modulator composed of micromirrors involved in exposures of length N horizontal M with image spacing C and projection length ratio Z, when the mirror angle θ is less than 45 degrees, one optical modulation step is performed. The vertical exposure length mainly depends on N * C * Z * cosθ and the horizontal exposure length mainly depends on M * C * cosθ. However, instead of one optical modulation step, Q + 1 (hereinafter, referred to as the total number of steps) from the initial 0th optical modulation step to the final Q (hereinafter referred to as the final step number) optical modulation step. Taking into account the vertical exposure length and the horizontal exposure length, the vertical exposure length is directly dependent on N * C * Z * cosθ, but the horizontal exposure length is the horizontal exposure length because the light is accumulated in the direction of movement of the substrate. Is more affected by the product of the scan pitch and the number of final steps than M * C * cosθ, and M * C * cosθ affects the exposure energy amount rather than the horizontal exposure length. Therefore, the number of longitudinal mirrors (hereinafter, referred to as the total vertical mirror number) of the spatial light modulator involved in the exposure is N, the number of transverse mirrors (hereinafter referred to as the total mirror length) is M, the image interval is C, and the projection length ratio is At Z, when the scan speed, the scan pitch, the number of final steps and the mirror angle are all fixed, the exposure area is mainly influenced by N * C * Z and the amount of exposure energy is mainly influenced by M. With the image interval, projection length ratio, scan speed, scan pitch, the number of final steps and the mirror angle all fixed, the problem of exposure energy due to the increase and decrease of the total mirror number M is solved with a photoresist or a light source. Although this is possible, the problem of the exposure area according to the increase and decrease of the total vertical mirror number N can be solved only by increasing or decreasing the number of spatial light modulators required for the exposure process. Therefore, especially in the design of a digital exposure machine using a mass production-compatible spatial light modulator for exposing large-area display substrates such as currently produced TFT LCDs, most of them adopt a method of increasing the exposure area by installing a plurality of spatial light modulators. It is true.

Comparing Fig. 17 (a) with Fig. 19 (a) with focus on the exposure area, since the projection length ratio Z in Fig. 19 (a) is larger than the projection length ratio Z in Fig. 17 (a), the same area is exposed. It can be found that the number of longitudinal mirrors used to do this is small. 17 (b) and 19 (b), of course, not all conditions are the same, so the scan resolution S of FIG. 19 (b) is increased than the scan resolution S of FIG. 17 (b). It can be found that this is not as much as the increase in the projection length ratio of Fig. 19 (a) with respect to a). This is important. Because if the number of mirrors used to expose the same area can be reduced while maintaining uniform scan resolution, the result is that the number of spatial light modulators required for the exposure of a large area substrate can be reduced, resulting in economical ripple effects. Is significant. However, comparison of Figs. 17 (a), 17 (b), 19 (a), and 19 (b) alone cannot be prompt, and confirmation of this fact must be made under the same conditions. Therefore, in the simulation of FIG. 17 (b), the projection length ratio Z is determined to be 4 / SQRT (3) twice that of FIG. 17 (b), leaving all other conditions intact, and the scan resolution S of FIG. 17 (b). Exposure simulation is performed when all three conditions are satisfied by determining the horizontal cell number A to 20 so that the number of horizontal cells A is kept at 1.091 microns as it is, and the scan resolution D is calculated as 168 and the unit step number L is calculated as 6 as it is. It was.

FIG. 20 (a) shows the distribution of the light beams irradiated onto the substrate in a stationary state, and FIG. 20 (b) shows the scan pitch P of FIG. And 1.091 microns in the same manner as the scan resolution S. As expected, FIG. 20 (b) shows the same delta projection structure as the equilateral triangle of 1.091 micron scan resolution as shown in FIG. 17 (b). FIG. 21 (a) shows the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 18400 similar to FIG. 3, which is the basis of the projection structure of FIG. 17 (b), and FIG. ) Shows an initial light modulation step unit projection structure 20000 and a final light modulation step unit projection structure 36800 that are the basis of the projection structure of FIG. 20 (b).

Comparing Fig. 17 (a) with Fig. 20 (a), the longitudinal mirror Z used in exposing the same area is twice the projection length ratio Z of Fig. 20 (a). It can be found that the number of is reduced to 1/2. In FIG. 21 (a), the portion of the projection structure 10000 and the upper portion of the straight line 40000 of the projection structure 18400 is virtually moved to the upper portion of the straight line 40001, and the projection structure 20000 and the projection are shown in FIG. 21 (b). 21A and 21 assuming large-area substrate exposure in which the portion located above the straight line 40000 of the structure 36800 is virtually moved above the straight line 40002, so that a plurality of spatial light modulators are required. Comparing (b), the projection length ratio Z in Fig. 21 (b) is twice the projection length ratio Z in Fig. 21 (a), so that the number of longitudinal mirrors used to expose the same vertical length 40011 is obtained. It can be seen that Fig. 21 (a) is six and Fig. 21 (b) is three, and the number of spatial light modulators to be required to expose the same vertical length 40012 is shown in Fig. 21 (a) and Fig. 21 (b). It can be seen that it should be twice. On the other hand, as the projection length ratio Z of FIG. 21 (b) increases to 2 times of FIG. 21 (a), the horizontal cell number A of FIG. 21 (b) is equal to 2 of FIG. 21 (a) in order to maintain the same scan resolution S. By doubling, it can be expected that the amount of exposure energy will be reduced by half when the exposure speed is kept the same. However, increasing the amount of exposure energy may be desirable, as discussed above, rather than increasing the number of spatial light modulators required from an economical point of view.

In the case of Fig. 17 (a), 2 / SQRT (3), which is the projection length ratio Z, is the cscΨ * 1/1 associated with cscΨ or sinΨ, In the case of sinΨ * 4/3, the horizontal cell number A is 10, and in FIG. 20, 4 / SQRT (3) having a projection length ratio Z is cscΨ * 2/1 or sinΨ * 2/3 and the horizontal cell number A is 2 * 10. 17 (b) and 20, the vertical exposure length is increased by 7 times while maintaining the delta projection structure in which the projection shape angle Ψ of the same scan resolution is maintained at π / 3 radians using the same exposure speed as in FIGS. It is obvious that the reduction of the number of spatial light modulators to 1/7 is achieved by setting the projection length ratio Z to 7 * cscΨ and the horizontal cell number A to 7 * 10 and increasing the exposure energy amount by 7 times. . On the other hand, in the case of Fig. 19 (a), in which the projection form angle Ψ is a stripe projection structure in which an isosceles triangle is correctly maintained in π / 2 radians, 2, the projection length ratio Z, is also related to cscΨ or sinΨ cscΨ * 2/1 or sinΨ * 2/1 and the horizontal cell number A is 10. Considering the equations (6) and (8) representing the condition that the isosceles triangular projection structure is formed under the projection shape angle Ψ of π / 4 radian or more and 3π / 4 radian or less, the projection length ratio Z is a function of sinΨ or cscΨ. It can be seen that the scan resolution D and the number of unit steps L become integers at the same time, which increases the likelihood that condition 2 and condition 3 will be satisfied at the same time. Examining when the projection shape angle Ψ is π / 2 radians and the projection length ratio Z is 3.5, of course, the scan resolution S will be different from that in Fig. 19 (b), but the scan resolution D calculated by Equation 12 and Equation 13 The number of unit steps L calculated by is simultaneously an integer, and the horizontal cell number A and vertical cell number B which are mutually related are A, 7 is B, 2 is A, 14 is B, 2 is 2, ... 3.5, the ratio Z, can be written as cscΨ * 7/2 or sinΨ * 7/2. In addition, when the projection angle Ψ is π / 3 radians, sinΨ is cscΨ * 3/4 is sinΨ and cscΨ is sinΨ * 4/3, and when the projection angle Ψ is π / 2 radians, sinΨ is cscΨ. Therefore, in the present invention, based on the above discussion, the projection length ratio Z is calculated based on the projection form angle Ψ, and the projection length ratio Z is a value obtained by multiplying the constant V by the cosecant value csc or cscΨ of the projection form angle Ψ. Is generalized as a function of the projection form angle Ψ as cscΨ * V / W or V / (W * sinΨ), which is divided by V and an integer W that is disjoint.

17 (b) and FIG. 20 with reference to FIG. 21 and by the above discussion, in the present invention, cscΨ or cscΨ, which is the cosecant value of the projection form angle Ψ in which the projection length ratio Z is not only 1 but also a number greater than 1 and less than 1 By setting the product of multiplying the integer V to cscΨ * V / W divided by V and the dissimilar integer W, the projection angle Ψ is π / 3 radians or 2π / as well as the case that the projection angle Ψ is π / 2 radians. Even in the case of 3 radians, it was found that it is possible to accurately maintain the projection structure of the equilateral triangle structured as the projection shape angle Ψ. Furthermore, by increasing the projection length ratio Z to 1, the projection length ratio in the vertical direction is maintained while maintaining the scan resolution. It has been found that it can be increased by to reduce the number of spatial light modulators required. Therefore, according to the present invention, a digital apparatus using an economical spatial light modulator by reducing the number of mirrors used to expose the same area and the number of spatial light modulators required for exposure of a large area substrate while maintaining the exact projection structure of the same scan resolution is achieved. It has been found that lithography is possible.

7. Scanning Ratio

In digital lithography using a spatial light modulator, the degree of pattern resulting from the exposure is generally known to improve the scan resolution S as the scan pitch P decreases. However, as scan pitch P decreases, the exposure time required to expose the same area increases. Considering the case where only the scan pitch P is changed while all other conditions are fixed, the h-offset ξ and the v-offset η are always scanned even if the scan ratio H is increased by more than 1 so that the scan pitch P is increased by the scan resolution S. If the same delta projection structure, stripe projection structure, or isosceles triangle projection structure as the resolution S can be maintained, it is possible to realize an economical and efficient exposure process with short exposure time. Accordingly, in the present invention, under the delta projection structure, the stripe projection structure, and the isosceles triangle projection structure, scan ratio H is used as an index for determining the efficiency of lithography, and scan resolution S is used as an index for determining the accuracy of lithography. If the scan resolution at the larger value remains the same as the scan resolution at the scan ratio of 1, it is considered that the lithography efficiency is improved while maintaining the lithography accuracy.

FIG. 22 illustrates a projection structure formed on a substrate, which is an object to be exposed, which is reflected by twelve horizontally three micromirrors and moved from a predetermined initial position to a predetermined final position by eight light modulation steps. FIG. 23 shows only the initial light modulation step unit projection structure 10000 and the final light modulation step unit projection structure 10800 in the projection structure of FIG. 22, and FIG. 24 shows the image centers (FIG. Only 10002) will be shown respectively. The image center 10011 of the first one-row column moves to the image center 10811 after eight light modulation steps, and the position of the image center 10811 coincides with the position of the initial two-row five-column image center. 22 to 24, the total length mirror number N is 3 and the total length mirror number M is 12, the projection angle Ψ is π / 3 radians, the horizontal cell number A is 4, and the vertical cell number B is 1 A mutual relationship is established between A and the vertical cell number B, the projection length ratio Z is 4 / SQRT (3), and the image spacing C (10003) is 10 * SQRT (3) / 2 micron, and these are expressed in equation (5). The scan resolution D obtained by substitution is an integer of 8, and the unit step number L obtained by substituting these into equation (8) is an integer of 1. Accordingly, the vertical image interval C * Z 10004 is 20 microns, the scan pitch P is about 5.0 microns, and the mirror angle θ 10009 is about 30.0 degrees. In the case of Figs. 22 to 24, all three conditions of the horizontal cell number A, the vertical cell number B, and the projection length ratio Z determination for forming the delta projection structure of the present invention are satisfied. Therefore, the h-offset ξ obtained by substituting 8 into the scan resolution D of Equation 5 is 5.0 microns, so the v-offset η according to Equation 4 and the scan pitch P are the same, and the projection structure shown in the center of FIG. The delta projection structure in the form of equilateral triangle was formed. In the case of FIGS. 22 to 24, the h-offset ξ, the v-offset η, the scan resolution S and the scan pitch P are all the same 5.0 microns. Accordingly, the scan ratio H of FIGS. 22 to 24 is one.

In order to determine the possibility of improving the efficiency of the lithography described above, in Figs. 22 to 24, the scan ratio P is doubled to 2, and the scan pitch P is tripled, while all other conditions are fixed. The projection structure formed on the substrate is shown by increasing the scan ratio H to three. FIG. 25 illustrates a projection structure formed on a substrate in units of light modulation steps when the scan ratio H is 2, and FIG. 26 illustrates an initial light modulation step unit projection structure 10000 and a final light modulation step unit in the projection structure of FIG. 25. Only the projection structure 10800 is selected and shown, and FIG. 27 is a diagram illustrating only the image centers 10002 in the projection structure of FIG. 25. The image center 10011 of the first one-row column moves to the image center 10811 after eight light modulation steps, where the position of the image center 10811 coincides with the position of the initial three-row, nine-column image center. FIG. 28 illustrates a case where the scan ratio H is 3, and shows the projection structure formed on the substrate for each light modulation step unit, and FIG. 29 shows the initial light modulation step unit projection structure 10000 and the final light modulation step unit in the projection structure of FIG. Only the projection structure 10800 is selected and shown, and FIG. 30 illustrates only the image centers 10002 in the projection structure of FIG. 28. The image center 10011 in the first row and the first column moves to the image center 10811 after eight light modulation steps, where the position of the image center 10811 is the initial four-row and thirteen-column image when the number of image cells increases. Will match the location of the center.

In the case of FIGS. 25 to 30, only the scan pitch P is different from the case of FIGS. 22 to 24, and all other conditions are the same. The total length mirror number N is 3 and the total length mirror M is 12, and the projection shape angle Ψ Is π / 3 radian, the horizontal cell number A is 4, the vertical cell number B is 1, and a mutual relationship is established between the horizontal cell number A and the vertical cell number B, and the projection length ratio Z is 4 / SQRT (3). The interval C (10003) is 10 * SQRT (3) / 2 microns, the scan resolution D obtained by substituting them into Equation 5 is an integer 8, and the unit step number L obtained by substituting these into Equation 8 is 1, The vertical image spacing C * Z (10004) is 20.0 microns. Referring to FIGS. 25 to 30 closely compared to FIGS. 22 to 24, in the case of the mirror angle θ, the result of considering the scan ratio H is the same as that of Equation 1, and Equation 1 may be used as it is. However, the scan pitch P defined as equal to the h-offset ξ to the scan resolution S of Equation 5 is newly calculated by Equation 18 considering the scan ratio H as the scan ratio H is not one.

Accordingly, the mirror angle θ 10009 is 30.0 degrees in the case of FIGS. 25 to 30 as in FIGS. 22 to 24, and the scan pitch P is 10.0 microns according to Equation 18 in FIGS. 25 to 27. 28 to 30 is 15.0 microns according to Equation 18.

The projection structure shown in the center of FIG. 27 in which the scan pitch P has increased twice the scan pitch P of FIG. 24 is different from FIG. 24, in which the horizontal cell number A, the vertical cell number B, and the projection length allow the delta projection structure of the present invention to be formed. Although all three conditions of the ratio Z crystal were satisfied, the h-offset ξ and v-offset η did not remain at the scan resolution of FIG. 24 but appeared different from the projection structure. However, in the projection structure shown in the center of FIG. 30 in which the scan pitch P increased three times the scan pitch P of FIG. 24, a delta projection structure in the form of an equilateral triangle having a scan resolution S of 5.0 microns was formed in the same manner as in FIG. 24. Of course, in the case of FIG. 30, the absolute amount of light irradiated onto the substrate may be different from that of FIG. 24 depending on the number of mirrors according to the moving speed of the substrate, but in terms of distribution of light quantity and line edge roughness related to patterning precision, Suggests that it will result in nearly similar exposure results. Thus, in the case of Fig. 30, it can be said that the efficiency of lithography is improved while maintaining the lithography accuracy.

22 to 30 are cases where only the scan ratio H and thus the scan pitch P are different and all other conditions are the same. However, the results show that the scan resolution S and the projection structure may vary depending on the scan ratio H. The cause of such a result can be easily found by recalling that condition 1 of the three conditions for forming the projection structure of the present invention was a condition that a mutual relationship must be established between the horizontal cell number A and the vertical cell number B. The scan resolutions D of FIGS. 22 to 30 are all eight. The scan ratio H of FIGS. 22 to 24 is 1, the scan ratio H of FIGS. 25 to 27 is 2, and the scan ratio H of FIGS. 28 to 30 is 3. FIG. When the scan ratio H is 1 or 3 mutually different from the scan resolution D, the projection structure of scan resolution S is maintained, but the scan ratio H is not 2 mutually different from the scan resolution D. Therefore, when the scan ratio H is not 1, which is mutually different from the scan resolution D, in order to form the delta projection structure, the stripe projection structure, the isosceles triangle projection structure, or the triangular projection structure, the scan resolution D and the scan ratio H are formed. It is obvious that the relationship between cows should be established. When the mutual relationship between the scan resolution D and the scan ratio H is established, even if the scan ratio H is increased by more than 1 to increase the scan pitch P, the delta projection structure is used to confirm that the uniform projection structure is maintained. In the simulations of FIG. 17 (b) and FIG. 19 (b), which are the stratified projection structure, exposure simulation was performed by changing only the scan ratio H and the scan pitch P accordingly, leaving all other conditions unchanged.

In the case of Fig. 17 (b) which is a delta projection structure, the scan resolution D is 84 which can be expressed as the product of prime numbers of (2 * 2 * 3 * 7). FIG. 31 illustrates a case in which only the scan ratio H and the corresponding scan pitch P are different in the simulation of FIG. 17B, that is, the scan ratios H of FIGS. 31A, 31B, and 3D are each 2; , 3, 5, and 6, and thus the scan pitch P is 2.182 microns, 3.273 microns, 5.455 microns, and 6.547 microns, respectively, according to Equation 18. As expected, only in the case of FIG. 31 (c) where a mutual relationship is established between the scan resolution D and the scan ratio H, an delta projection structure having an equilateral triangle shape having a scan resolution S of 1.091 microns is formed in the same manner as in FIG. 17 (b). In (a), (b), and (d) of FIG. 31 in which no mutual relation is established between the remaining scan resolutions D and the scan ratio H, the h-offset ξ and the v-offset η are scanned in FIG. 17 (b). It is not maintained in resolution but appears as a projection structure different from the prediction. In the case of Fig. 19 (b), which is a striking projection structure, the scan resolution D is 68, which can be expressed as a product of a prime number of (2 * 2 * 17). 32 is a case where only scan ratio H and corresponding scan pitch P are different in the simulation of FIG. 19 (b), that is, scan ratios H of FIGS. 32A and 32B are 2 and 3, respectively. P is 2.970 microns and 4.456, respectively. As expected, in FIG. 32 (b) where a mutual relationship is established between the scan resolution D and the scan ratio H, as in FIG. 19 (b), a right-angled isosceles triangle shaped projection structure having a scan resolution of S is 1.485 microns. In the case of FIG. 32 (a) where the relationship between the scan resolution D and the scan ratio H is not established, the h-offset ξ and v-offset η are not maintained at the scan resolution of FIG. 19 (b). It appears to be a different type of projection structure. Therefore, if conditions 1, 2, and 3 peculiar to the present invention are satisfied, and the scan ratio H is mutually different from the scan resolution D, even if the scan ratio H is increased by 1 and the scan pitch P is increased, the projection structure of uniform scan resolution S is increased. Was found to remain the same. Therefore, it has been found by the present invention that digital lithography using an efficient spatial light modulator is possible by increasing the scan ratio H.

In an actual exposure process, there exists a target value of resolution related to the critical dimension of the exposed pattern, the line edge roughness of the exposed pattern, and the like. The exposure simulations of FIGS. 17, 19, 20, 31, and 32 are designed to increase the scan resolution S or the fixed resolution G regardless of the target value or the size of the pattern for the purpose of clearly showing the projection structure. It was performed in a determined state. However, in the actual exposure process, it is only natural to set the scan resolution S or the fixed resolution G to be equal to or smaller than the target value of the resolution related to the critical dimension of the exposed pattern or the line edge roughness of the exposed pattern. Meanwhile, in the present invention, the scan resolution S and the fixed resolution G are both functions of the projection shape angle Ψ, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B. Therefore, in the present invention, after the projection structure is defined as the projection shape angle Ψ, if the resolution is more important than the projection area, the scan length S to the fixed resolution G which is equal to or smaller than the target value is first determined, and then the projection length corresponding thereto is determined. The ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B may be determined to satisfy the above three conditions. If the projection area is more important than the resolution, the projection length ratio Z and the image interval C are determined first according to the projection area, and then projected. The scan resolution S to the fixed resolution G may be determined based on the length ratio Z and the image interval C, and the horizontal cell number A and the vertical cell number B may be determined to satisfy the three conditions, or the projection area and the resolution may be simultaneously considered. Scanning resolution S to fixed resolution G, projection length ratio Z, image interval C, horizontal cell number A and vertical cell number B are It may decide that together. Since the order of such determinations should be different for the user, the order of such determinations is left to the user in the present invention, and in the present invention, the projection structure is determined as the projection shape angle Ψ, and then the projection area and resolution are appropriate for the three aspects of the present invention. In order to satisfy all of the conditions, the scan resolution S to the fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B are determined so that all the procedures of the determination are included. Accordingly, the three conditions of the present invention are three conditions of scan resolution S to fixed resolution G, projection length ratio Z, image interval C, horizontal cell number A and vertical cell number B crystals.

In order to expose the 5 micron pattern consisting of the 90 degree, 60 degree, 45 degree and 0 degree straight lines, it is required to maintain the resolution below 0.15 micron, the delta projection structure is required, and the projection length ratio Z is 4 / SQRT ( 3), the image interval C 10003 is 10 * SQRT (3) / 2 micron, and the number of vertical cells B is 3, as in FIG. 20, and there may be a predetermined exposure process. The minimum value A _MIN of the horizontal cell number A to form an equilateral triangle-shaped delta projection structure having a scan resolution S of 0.15 micron or less is obtained by substituting 0.15 micron to the scan resolution S of Equation 11 and arranging for A. The calculation is possible.

The minimum value A _{MIN according} to Equation 19 is 153.8041099. The minimum number of horizontal cells, A _MIN or more, is 154, 155, 157, ..., 164, 165, .... As the number of horizontal cells A increases, the scan resolution S decreases from 0.15 microns, thereby improving the degree of the pattern, but this is insignificant, and the scan pitch P decreases to increase the exposure time. Therefore, the horizontal cell number A is the minimum value A _MIN It is preferable to determine the above integers closest to the minimum value A _MIN . Considering the integers between 154 and 165 of these, 156, 159, 162 and 165 are excluded because they do not satisfy condition 1 of the present invention because they are not multiples of the vertical cell number B as multiples of 3. Of the remaining integers, 154, 155, 157, 158, 161, and 163 are excluded because the scan resolution D calculated by substituting them into Equation 9 is not an integer and does not satisfy the condition 2 of the present invention. 160 of the remaining integers is excluded because the unit step number L calculated by substituting this in Equation 10 is not an integer and does not satisfy condition 3 of the present invention. The scan resolution D is calculated as an integer 10104 by determining the remaining one integer 164, which satisfies all the conditions 1, 2, and 3 of the present invention, and which is the minimum number A _MIN of the horizontal cell number as the horizontal cell number A. According to Equation 10, the unit step number L is calculated as an integer 60, and the scan resolution S is calculated as 0.1407 microns. 1 (a) and 1 (c) are 1640 horizontal lines for the case in which the scan pitch P is calculated to be 0.7035 microns by determining the horizontal cell count A to 164 and the scan ratio H to be 5 to be different from the scan resolution D. The exposure simulation results of 180 vertical mirrors are shown. FIGS. 1 (b) and 1 (d) show the image interval C, the horizontal cell number A, the vertical cell number B, and the projection length ratio Z as shown in FIG. ), While maintaining the resolution calculated by Equation 11 in the same manner as in Fig. 1 (a), the scan resolution D is determined to be an arbitrary integer 4040 that does not satisfy the condition 2 of the present invention, and the scan ratio H It is the result of performing exposure simulation with the same number of mirrors as in FIG. 1 (a) for the case where the scan pitch P is calculated as 0.3519 microns by determining 1 as 1. The values actually set in the exposure process are the total length mirror number N, the total length mirror number M, the image interval C, the projection length ratio Z, the mirror angle θ and the scan pitch P. 1 (a), 1 (b), 1 (c), and 1 (d) have the same image interval C, horizontal cell number A, vertical cell number B, and projection length ratio Z, the mirror angle θ is also all 2.4 degrees. same. Only scan pitch P differs from FIGS. 1 (a) and 1 (c) with 0.7035 microns and FIGS. 1 (b) and 1 (d) with 0.3519 microns. 1 (a) and 1 (b) are simulation results of the discrete exposure process and FIGS. 1 (c) and 1 (d) are simulation results of the analog exposure process, except for the scan pitch P. When (a) and (b) are compared with each other and the values actually set in the exposure process are compared with each other, and (c) and 1 (d) are compared with each other, Although scan pitch P is reduced than scan pitch P of FIGS. 1 (a) and 1 (c), the result shows that the line edge roughness which is not shown in FIGS. 1 (a) and 1 (c) is shown in FIGS. ), It appears in the 90 degree, 60 degree, and 45 degree straight lines except a horizontal line, and it turns out that the patterning precision is reduced whether the exposure method is the discrete method or the analog method. 1 (b) and 1 (d) which do not satisfy the condition 2 of the present invention are because the resolution calculated by Equation 11 is not the scan resolution and the projection structure is not maintained. Thus, the results of FIG. 1 suggest that the patterning accuracy does not always improve when the scan pitch P decreases, and the efficiency of increasing the scan ratio H so as to deviate from the importance of the three conditions of the present invention and the scan resolution D. . Therefore, the exposure process is a discrete method or an analog method, or the conditions 1, 2, and 2 of the scan resolution S to the fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B are determined. It has been confirmed once again that digital lithography using an efficient and precise spatial light modulator is possible by determining the scan ratio H so as to satisfy both the condition 3 and the scan resolution D.

In summary, the three conditions of the scan resolution S to the fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B, which are proposed in the present invention, are defined by the condition 1 being between the horizontal cell number A and the vertical cell number B. The relation between each other is established, condition 2 is the scan resolution D of the equations 5 to 12 to 15, and condition 3 is the unit step number L of the equations 8 to 13 to 16, and the scan ratio H must be determined by an integer equal to or greater than 1 that is mutually equal to scan resolution D. All three conditions peculiar to the present invention are satisfied so that the scan resolution S to the fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A and the vertical cell number B are determined, and the scan ratio H is an integer greater than or equal to 1 with the scan resolution D. Once determined, 1) a delta projection structure of equilateral triangles to a stripe projection structure of equilateral isosceles triangles to an isosceles triangle projection structure is formed, and 2) an increase in the projection length ratio results in the preservation of the scan resolution and the projection structure. It is possible to increase the longitudinal length or area projected by the optical modulator, and 3) increase the scan pitch to shorten the exposure time while preserving the scan resolution and the projection structure. Therefore, the scan resolution according to the present invention is a resolution that is always maintained even if the scan pitch is increased, that is, super-resolution meaning superior resolution that is superior to the scan pitch. Therefore, the delta projection structure to the stripe projection structure to the projection where the projection form angle Ψ is π / 4 radians or more and 3 π / 4 radians or less and the v-offset? Scan resolution S according to Equations 11 to 14 or Equations 4 to 5 of an isosceles triangular projection structure having a shape angle Ψ is defined as super resolution hereinafter and denoted as S. Meanwhile, in the triangular projection structure, the v-offset η of Equation 4 and the h-offset ξ of Equation 5 may be different from each other. Therefore, in the triangular projection structure, the scan resolution S such as the h-offset ξ of Equation 5 and the fixed resolution G such as v-offset η of Equation 4 are used without defining the super resolution.

A first feature of the present invention is to define a structured regular projection structure as the unique projection shape angle Ψ of the present invention. A second feature of the present invention is to determine the ratio of the longitudinal spacing to the horizontal spacing between image centers suitable for a given projection structure as the projection length ratio Z, which is a function of the projection form angle Ψ of equilateral triangles, equilateral isosceles triangles and isosceles triangles. It is to maintain the normal projection structure of the projection shape angle Ψ and increase the projection length or the projection area while maintaining super resolution without increasing the number of micromirrors of the optical modulator. The third aspect of the present invention is based on scan resolution S to fixed resolution G, projection length ratio Z, image interval C, horizontal cell number A and vertical cell number B, which are determined to satisfy all three conditions peculiar to the present invention, and the scan resolution D is different from each other. By calculating and setting the mirror angle θ and the scan pitch P using scan ratio H which is an integer greater than or equal to one, the delta projection structure of the equilateral triangle of super resolution S to the stripe projection structure of the equilateral isosceles triangle or the isosceles triangle projection structure of super resolution S The normal projection structure in which the angle between the super resolution equilateral sides is structured as the projection shape angle Ψ is maintained. The fourth feature of the present invention is that the projection length ratio Z, the fixed resolution G, the image interval C, the horizontal cell number A and the vertical cell number B are determined in advance so that all three conditions peculiar to the present invention are difficult to satisfy. The scan resolution S is determined by taking into account the scan resolution S, and based on these, the mirror angle θ and the scan pitch P are calculated and set using the scan resolution H, which is an integer greater than or equal to 1, to exceed the scan pitch P. While maintaining the scan resolution S, the normal projection structure in which the angle between the fixed resolution side and the scan resolution side of the triangular projection structure similar to the equilateral triangle, the equilateral isosceles triangle, and the isosceles triangle is maintained at the projection form angle Ψ is maintained. Therefore, the digital lithography method of the present invention in which the normal projection structure of the projection form angle Ψ is maintained is called a regular lithography method (RLM).

An embodiment of the normal lithographic method of the present invention will be described with reference to the flowchart of FIG. 33 which is a representative view as follows. As described above, the normal lithography method of the present invention is largely divided into two methods.

The first method uses the super resolution S, the projection length ratio Z, the image spacing C, the horizontal cell number A, and the vertical cell number B to satisfy all three conditions peculiar to the present invention to ensure that a predetermined projection structure is formed accurately based on the projection form angle Ψ. And multiplying the scan resolution H and scan ratio H, which is an integer equal to or greater than 1, by the super resolution S and calculating and setting the scan pitch P, thereby setting the delta projection structure of the equilateral triangle of super resolution S to the stripe projection structure of the equilateral isosceles triangle to the isosceles triangle projection It is a method in which a regular projection structure structured with the equilateral sides of the given projection shape angle Ψ of the structure is accurately maintained. The second method is based on the projection form angle Ψ. Of the three conditions peculiar to the present invention, the condition 2 related to the scan resolution D and the condition 3 related to the number of unit steps L are not considered. After determining the length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B, or under a predetermined fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B, By determining the scan resolution D so as to form the closest projection structure, and calculating and setting the scan pitch P by multiplying the scan resolution H calculated using the scan resolution D by a scan ratio H that is an integer greater than or equal to 1 and a fixed resolution. Projection form of triangular projection structure similar to orthogonal to right isosceles to isosceles triangles with scan resolution S equal to or close to G The normal projection structure structured at an angle Ψ is maintained.

The embodiments described below and shown in the chart of FIG. 33 are examples of the first method of forming an isosceles triangle projection structure of projection form angle Ψ and of the second method of forming a triangle projection structure of projection form angle Ψ. The embodiment of the first method is represented by a solid line in FIG. 33, and the embodiment of the second method is indicated by a dashed line while other parts S320 to S340 are left as are, while the parts common to the first method are left as they are. The following examples are intended to illustrate the invention, not to limit the scope thereof, and the scope of the invention is defined by the claims.

<Example of the first method>

(1) Determine the projection shape angle Ψ that defines the projection structure. (Step S110)

(2) Determine the super resolution S, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B based on the projection form angle Ψ. (Step S220)

(2a) Examine whether the relationship between horizontal cell number A and vertical cell number B holds. If A and B are not mutually different, the flow advances to step S220 to determine S, Z, C, A, and B again. (Step S221)

(3) The unit step number L is calculated by substituting the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B into the equation (8) (step S240).

(3a) It is checked whether the unit step number L is an integer. If L is not an integer, the flow advances to step S220 to determine S, Z, C, A, and B again. (Step S241)

(4) The scan resolution D is calculated by substituting the projection shape angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B into Equation 6. (Step S230)

(4a) Check if the scan resolution D is an integer. If D is an integer, the flow advances to step S450. If D is not an integer, the flow advances to step S220 to determine S, Z, C, A, and B again. (Step S231)

(5) The mirror angle θ is calculated by substituting the projection length ratio Z, the number of horizontal cells A, and the number of vertical cells B into Equation 1 (step S450).

(6) The scan ratio H is determined as an integer of 1 or more. (Step S560)

(6a) Examine whether the scan ratio H is in relation to the scan resolution D. If H and D are mutually different, the flow advances to step S571. If H and D are not mutually different, the flow advances to step S560 to determine H again (step S561).

(7) Set the mirror angle θ to the calculated θ value, and set the scan pitch P to the value obtained by multiplying the super resolution S by the scan ratio H, or the projection length ratio Z, the image interval C, the horizontal cell number A, the vertical cell number B, and The scan pitch P is set to a value obtained by substituting the scan resolution D and the scan ratio H into Equation 18 (step S571).

(8) On the basis of the determined Z and C, the horizontal distance between the image centers is set to the image interval C, and the vertical distance between the image centers is set to C * Z (step S572).

(9) N, M, and Q are calculated so that the total length of the maze N becomes an integer multiple of H * B, the total number of mirrors, M, becomes an integer multiple of H * A, and the final step number Q is an integer less than 1 from the integer multiple of D. (Step S680)

(10) A digital mask is generated, compressed, and restored by a structural quantitative lossless compression method (step S790).

Steps S680 of (9) and Steps S790 of (10) are steps related to compression of the following digital mask, and detailed descriptions and specific implementation methods of steps S680 of (9) and (S10) of (9) are as follows. The structural quantitative lossless compression method of the digital mask is described in detail.

<Example of Second Method>

As shown in Fig. 33, (Step S110) of (1) and Step S450 of (5), Step S560 of (6), (Step S561) of (6a), and (7) (Step S571) of the second method. ), Step S572 of (8), step S680 of (9), and step S790 of (10) are the same as the first method, and the other parts (S320 to S340) of the second method shown by dotted lines in FIG. It demonstrates in (-4).

(2) Determine the fixed resolution G, the projection length ratio Z, the image interval C, the horizontal cell number A, and the vertical cell number B based on the projection form angle Ψ. (Step S320)

(2a) Examine whether the relationship between horizontal cell number A and vertical cell number B holds. If A and B are not mutually different, the flow advances to step S320 to determine G, Z, C, A, and B again. (Step S321)

(3) The scan resolution D is calculated by substituting the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B into Equation 17. (Step S330)

(4) The scan resolution S is calculated by substituting the projection length ratio Z, the image interval C, the horizontal cell number A, the vertical cell number B, and the scan resolution D into Equation 5. (Step S340)

8. Compression of Digital Masks

In digital lithography using a spatial light modulator, a pattern is formed by light beams that selectively reflect at each light modulation step on a substrate on which the micromirrors of the light modulator are moving. Whether to reflect light beams to each micromirror to the micromirror driving information may be obtained according to Patent No. 655165, which is patented and registered by the applicants, or 10-2007-0046450, which is patented by the applicants. Or by simply comparing the position of the image center 10002 with the input pattern drawing to reflect the light beam only when the image center 10002 is inside the pattern area on the input drawing. have. Regardless of the method of determining whether the light beam is reflected on each micromirror, the resultant light beam reflection (hereinafter referred to as a mirror unit mask) corresponding to one mirror is a binary digit represented by 1 to 0 or One bit of digital data. When the number of micromirrors involved in the exposure of the spatial light modulator is N horizontal M vertically, whether the light beam is reflected on all the micromirrors of the spatial light modulator corresponding to one light modulation step (hereinafter, referred to as a digital mask) N * M bit digital data represented by 1 and 0, and when the exposure process is performed by 10000 light modulation steps, whether the light beam is reflected on all the micromirrors of the spatial light modulator corresponding to 10000 light modulation steps (hereinafter, Pattern mask) is 10000 * N * M bit digital data. For example, in the case of a typical spatial light modulator, XGA-class Digital Micromirror Device (DMD) of Texas Instrument, since N is 1024 and M is 768, the digital mask is 786,432 bits of digital data. 980,432 micromirrors corresponding to 10000 light modulation steps each 10000 light modulation steps are performed per second to match the resolution and tact time required for the actual process. For example, the pattern mask is 7,864,320,000 bits of digital data, and the pattern mask is generated and transmitted at 7.86432 Gbps (Giga bit per second). The conclusion is that the mask must be generated and transmitted at very high speed. In order to apply digital lithography using a spatial light modulator to a mass production process of a large area high resolution substrate such as an actual large area LCD, as described above, a plurality of spatial light modulators of 15 or more are required, and 10000 light modulation steps per second described above. It is known that the pattern mask generation and transmission speed should be improved by more than twice as long as this is performed. Therefore, there is an urgent need for a method of generating and transmitting a large amount of pattern mask at a very high speed. Therefore, in the present invention, a fixed amount of lossless compression for the generation or transmission of the digital mask based on the proposed projection structure unique to the present invention is performed as follows.

FIG. 34 shows an initial 10000 and a thirty-fourth formed on a substrate which is reflected by 40 horizontal and 15 micromirrors and moved by a light modulating step 370 from a predetermined initial position to a predetermined final position. 13400), 84th (18400), 118th (21800), 168th (26800), 202th (30200), 252th (35200), 286th (38600), 336th (43600), 370th (47000) FIG. 35 illustrates only the initial 10000, 84th (18400), 286th (38600), and 370th (47000) light modulation step unit projection structures. 34 to 35 show only the number of mirrors in FIGS. 2 to 7, and the rest of the conditions as shown in FIGS. 2 to 7 to maintain the same delta projection structure, showing the projection structure formed on the substrate. to be. That is, the image interval C of FIGS. 34 to 35 is 10 * SQRT (3) / 2 micron, the horizontal cell number A is 10, the vertical cell number B is 3, and the projection length ratio Z is 2 / SQRT (3), which is calculated accordingly. The integer scan resolution D of the delta projection structure is 84, the scan ratio H is 1 and the scan pitch P is 1.091 microns equal to the scan resolution S, all of which are the same as in FIGS. Accordingly, after all intermediate light modulation step unit projection structures are shown in FIG. 34, the image cell 10001 is not displayed, only the image centers 10002 are selected, and the center portion is enlarged. Being clear is not necessary. Looking closely at FIG. 35, as described with reference to FIGS. 2 to 7, the one-row, one-column image center 18411 of the 84th light modulation step 18400 is the four-row, eleven-column image center 10088 of the initial 10000. The second row three-column image center 18184 of the 84th optical modulation step 18400 is located at the fifth row 13-column image center 10099 of the initial 10000, and the 370th optical modulation step 47000. 12-row 30-column image center (47088) is located in 15-row 40-column image center (38688) of 286 th light modulation step 38600, and 11-row 28-column image center of 370 th light modulation step (47000). It can be seen that 47099 is located in the 14-row 38-column image center 38699 of the 286 th light modulation step 38600.

FIG. 36 is a view illustrating a two-step (K-D) 50000 and a K-th (60000) light modulation step unit projection structures in FIG. 34. The two step light modulation step unit projection structures of FIG. 36 are the two steps of the initial (10000) and the 84th (18400) of FIG. 34, the two steps of the 34th (13400) and the 118th (21800), and the 84th. 2 steps of (18400) and 168th (26800), 2 steps of 118th (21800) and 202th (30200), 2 steps of 168th (26800) and 252th (35200), ... It may be two steps, 286 th (38600) and 370 th (47000). Based on the description in FIG. 35, for the image centers in the region where the (KD) th light modulation step 50000 and the K th light modulation step 60000 shown in FIG. Generalized based on the J row I column image center of any K-th optical modulation step 60000, all of condition 1, condition 2, and condition 3 inherent in the present invention are satisfied, and the scan ratio H is mutually independent of the scan resolution D. Since it is an integer greater than or equal to 1, the image center of the (J + H * B) row (I + H * A) rows of the (KD) th optical modulation step 50000 is from the (K-D + 1) th to the (K-1) th It is not located the same as any of the image centers of the light modulation step up to, and is located in the K-th light modulation step 60000, the same as the image center of the row J column I.

Referring to Fig. 34 and Fig. 35, two steps of the initial (10000) and the 84th (18400) are described. The (J + H * B) row (I + H * A) column image centers of the initial 10000 are described. Is not located the same as any of the image centers during the first to the 83rd light modulation steps, and is located in the 84th light modulation step (18400) in the same manner as the image centers in row J columns. Although not shown in Fig. 34, the row J column I image center of the 84th optical modulation step 18400 is the same as the row column image center I + H * A of the initial (10000). The image center of row J of column 85 of the optical modulation step is located in the (J + H * B) row of image center of the first optical modulation step, and the image center of column 86 of the 86th optical modulation step Row J image center is the same as the (J + H * B) row (I + H * A) column image center of the second light modulation step, ...., row J column I of the 370th light modulation step The image center is located in the (J + H * B) row (I + H * A) column of the 286th optical modulation step.

FIG. 37 shows three regions of 50007, 50008 to 60008, and 60009 of four zones of the waste zone 50007, the regeneration zone 50008, the new zone 60008, and the new zone 60009. Figures separated by). Since the image centers of the (KD) th optical modulation step 50000 of FIG. 36 and the K th light modulation step 60000 are identically positioned, the overlapping regions (50008 to 60008 of FIG. 37) are the (KD) th light modulation step. A regeneration zone 50008 of 50000 and a renew zone 60008 of the K-th optical modulation step 60000. The discarded area 50007 is an area obtained by subtracting the reproduction area 50008 from the (KD) th light modulation step 50000, and the new area 60009 is an area obtained by subtracting the renewal area 60008 from the Kth light modulation step 60000. to be. Accordingly, the (KD) th light modulation step 50000 is composed of a waste zone 50007 and a regeneration zone 50008, and the Kth light modulation step 60000 is composed of a renovation zone 60008 and a new zone 60009. do.

Each of the image centers in row J column I of the renewal zone 60008 of the K-th optical modulation step 60000 of FIGS. 36 to 37 corresponds to (J +) of the reproduction zone 50008 of the (KD) -th optical modulation step 50000. Each of the image centers in row H * B) (I + H * A) will be identical. The number of horizontal mirrors to image cells used in FIGS. 34 to 37 is 40, the number of vertical mirrors to image cells is 15, the scan resolution D is 84, the number of horizontal cells A is 10, the number of vertical cells B is 3, The scan ratio H is one. For example, if K is 118, the (KD) th light modulation step 50000 of FIGS. 36 to 37 is the 34 th light modulation step 13400 of FIG. 34, and the K th light modulation step 60000 is shown in FIG. 34 is the 118th light modulation step 21800. Looking at the rows and columns to which the image centers or image cells constituting the reproduction zone 50008 belong, the image center 50001 for the (KD) th light modulation step 50000, that is, the 34 th light modulation step 13400 of FIG. 34. These four rows are 11 columns, 50002 is 4 rows and 40 columns, 50003 is 15 rows and 40 columns, and 50004 is 15 rows and 11 columns. Looking at the rows and columns to which the image centers or image cells belonging to the new zone 60008 belong, in the case of the K th light modulation step 60000, that is, the 118 th light modulation step 21800 of FIG. 34, the image center 50001 is 1. Row 1 column, 50002 is 1 row 30 columns, 50003 is 12 rows 30 columns, 50004 is 12 rows 1 column. That is, each of the image centers in rows 1 through 12 of the 118th light modulation step, columns 1 through 30, respectively, has each of the image centers in columns 4 through 15 of the 34th light modulation step, 11 through 40 columns, respectively. Each is located the same. In addition, the waste area 50007 includes image centers 4 to 4 and 1 to 3 rows 1 to 3 of the (KD) th light modulation step 50000, that is, the 34 th light modulation step 13400 of FIG. Is composed of image centers of columns 1 to 10 in rows 15 to 15, and the new zone 60009 is located in rows 13 to 15 of the K th light modulation step 60000, that is, the 118 th light modulation step 21800 of FIG. It consists of image centers in rows 1 through 30 and rows of image centers in rows 1 through 15 and columns 31 through 40.

In general, the total number of vertical mirrors is N, the number of total mirrors is M, the number of horizontal cells is A, the number of vertical cells is B, the scan ratio is H, the scan resolution is D, and the image of column J row I of the K-th optical modulation step. Referring to the projection structure whose center coincides with the position of the (J + H * B) row (I + H * A) column image center of the (KD) th light modulation step, the (KD) th light modulation step 50000 For playback area 50008, Image Center 50001 is (H * B + 1), (H * A + 1), 50002 is (H * B + 1), M, 50003, N, M, 50004 Is N rows (H * A + 1) columns, and in the update zone 60008 of the Kth light modulation step 60000, image center 50001 is 1 row, 1 row, 50002 is 1 row (MH * A) columns, 50003 is (NH * B) rows (MH * A) columns and 50004 is (NH * B) rows 1 columns. That is, each of the image centers from column 1 to column (MH * A) of row 1 to (NH * B) in the update zone 60008 of the K-th light modulation step 60000 is the (KD) th light modulation step. In the reproduction zone 50008 of (50000), each of the image centers (H * B + 1) to N rows (H * A + 1) to M columns is identical to each other. In addition, the waste area 50007 includes image centers in columns 1 to M and rows N in (H * B + 1) from row 1 to (H * B) in the (KD) th light modulation step 50000. Rows consist of image centers from column 1 to column H * A, and new area 60009 is column 1 from row (NH * B + 1) to row N of the Kth light modulation step 60000. It consists of image centers from (MH * A) to columns and image centers from (MH * A + 1) to M to rows 1 to N.

34 to 37 show a case where the scan ratio H is 1. In order to examine the positional change of the image centers according to the scan ratio H, the total vertical mirror number N is 6 and the total horizontal mirror number M is 6 for the case of FIG. It is increased to 24, and the rest is a view showing the projection structure formed on the substrate while keeping all the conditions to have the same delta projection structure as in FIGS. FIG. 38 illustrates a case where the scan ratio H is 1 as shown in FIG. 22 and FIG. 41 illustrates a case where the scan ratio H is 3 as shown in FIG. 28. The pino moves by eight light modulation steps from a predetermined initial position to a predetermined final position. Initial 0th (10000), 1st (10100), 2nd (10200), 3rd (10300), 4th (10400), 5th (10500), 6th (10600) The tenth (10700), eighth (10800) light modulation step unit projection structures, FIG. 39 is the initial zeroth (50000) and eighth (60000) light modulation step unit projection structures in FIG. 38 and FIG. 40 is shown in FIG. 40 and FIG. 43 shows FIG. 42 in the four regions of the waste zone 50007, the regeneration zone 50008, the regeneration zone 60008 and the new zone 60009. And divided into three regions of 60008 and 60009. 40, in the generalized description, M is 24, N is 6, A is 4, B is 1, H is 1, K is 8, and D is 8. It consists of image centers of 5 to 24 columns from 2 to 6 rows of modulation step 50000, and update zone 60008 is 1 column to 5 rows of 8th light modulation step 60000. To 20 columns, and the waste area 50007 contains image centers from columns 1 to 1 and rows 1 to 24 and rows 2 to 6 of the 0th light modulation step 50000. The new area 60009 is comprised of 6 to 6 rows of image centers and 1 to 20 columns of the 8th light modulation step 60000. It is described that it is composed of image centers of 21 to 24 columns up to 6 rows, which is the same as the result directly confirmed in FIGS. 39 and 40. In the case of FIG. 43, when M is 24, N is 6, A is 4, B is 1, H is 3, K is 8, and D is 8, the regeneration area 50008 is 0th light. It consists of image centers from 13 to 24 columns from 4 to 6 rows of modulation steps, and the new zone 60008 has images from 1 to 12 columns from 1 to 3 rows of the 8th optical modulation step. The waste area 50007 consists of the centers, and image centers 1 to 3 rows 1 to 24 columns and 4 to 6 rows 1 to 12 columns of the 0th light modulation step 50000. The new area 60009 is composed of image centers in rows 1 through 12 and columns 4 through 6 of the 8th light modulation step 60000, and in 13 columns in rows 1 through 6. It is described that it consists of up to 24 image centers, which is the same as the result directly confirmed in FIGS. 42 and 43. Therefore, the generalized description is valid even when the scan ratio H is not one. In addition, even if the scan resolution D is an arbitrary number of optical modulation steps, not an optical modulation step number based on the horizontal cell number A, the vertical cell number B, and the scan ratio H forming the unique normal projection structure of the present invention, the initial one row 1 It is obvious that the generalized description above is not valid if the image center of the column is a projection structure coinciding with the position of the initial (H * B + 1) row (H * A + 1) column image center after the D optical modulation step. Accordingly, the generalized description of the normal projection structure is performed so that the image center of row J column I of the K th light modulation step is (J + H * B) row (I + H *) of the (KD) th light modulation step. A) Expand for a general projection structure that matches the position of the thermal image center.

The same location of the image centers is the same as that of the boundary of the rectangular image cell centered on the image center. Accordingly, the K-th optical modulation step 60000 may be obtained by selecting any one of the above three methods, whether the light beams are reflected on the micromirrors corresponding to each image cell or the image center, the micromirror driving information, or the mirror unit mask. Each of the mirror unit masks in row J, column I, located inside the renewal zone 60008 of the () row is (J + H * B) row (I +) located inside the playback zone 50008 of the (KD) th light modulation step 50000. Each of the mirror unit masks of the H * A) column is the same. Therefore, in the present invention, the mirror unit masks in row J column I located in the renewal zone 60008 of the K-th optical modulation step are located within the regeneration zone 50008 of the (KD) th light modulation step without obtaining a new one (J + By using the mirror unit masks in the H * B) row (I + H * A) column as they are by shifting, the digital mask data is quantitatively losslessly compressed, thereby speeding up the generation or transmission of the digital mask.

Fig. 44 shows the total vertical mirror number N and the total horizontal mirror number composed of the masks in the J-row I columns of the K-th optical modulation step based on the horizontal cell number A, vertical cell number B, scan ratio H, and scan resolution D characteristic of the present invention. As an example of the case where the scan resolution D is 84 and the final step Q Q is 671, the method of structural quantitative compression and structural lossless recovery of the Kth digital mask corresponding to M is shown. The first light modulation step is shown in two rows and one column, the second light modulation step is shown in three rows and one column, and the 83rd light modulation step is shown in five rows and one column. The columns are shown in three columns from the 168th optical modulation step to the 251nd optical modulation step, and the five columns from the 588th optical modulation step to the 671th optical modulation step, and the order of the rows is shown in the same order as the rows described in the first column. .

As shown in column 1 of FIG. 44, from the initial 0 th light modulation step to the 83 th light modulation step, every N vertical M horizontal M J row I column mirror unit masks are generated for each K th light modulation step to generate the K th digital. Construct a mask. That is, in the 0th light modulation step, all the vertical N horizontal M mirror unit masks are generated to form the 0th digital mask 10000, and in the first light modulation step, all the vertical N horizontal M mirror unit masks are generated. To configure the first digital mask 10100, the second digital mask 10200 is generated by generating all N vertical and horizontal M mirror unit masks in the second optical modulation step, and the 83rd optical modulation. The 83rd digital mask 18300 is configured by generating all N vertical M mirror unit masks in a step.

As shown in column 2 of FIG. 44, from the 84th light modulation step to the 167th light modulation step, for each Kth light modulation step, (J + H *) of the reproduction zone 50008 of the (KD) th light modulation step. B) Row (I + H * A) mirror units Mirror rows (I-H * B) by column (-H * A) to mirror row J column I in mirror area 60008 of the k-th optical modulation step. The masks are updated, and the J-row I column mirror unit masks of the new zone 60009 of the K-th optical modulation step are generated to form the K-th digital mask. That is, in the 84th light modulation step, the mirror unit masks of the reproduction zone 50008 of the 0th light modulation step are shifted to the mirror unit masks of the update zone 60008 (18400A), and the mirror unit masks of the new zone are generated. (18400B) to configure the 84th digital mask 18400, and shift the mirror unit masks of the reproduction area 50008 of the first light modulation step from the 85th light modulation step to the mirror unit masks of the update area 60008. Update (18500A) and generate mirror unit masks of the new zone (18500B) to form the 84th digital mask 18500, and mirror unit masks of the reproduction zone 50008 of the second light modulation step in the 86th light modulation step. Shift to update with mirror unit masks in the new zone 60008 (18600A) and create mirror unit masks in the new zone (18600B) to form the 86th digital mask 18600, and the 167th optical modulation mask. Shifts the mirror unit masks of the reproduction zone 50008 of the 83rd light modulation step to update the mirror unit masks of the update zone 60008 (26700A), and generates the mirror unit masks of the new zone (26700B) and the 167th digital A mask 26700 is configured.

As shown in the third column of Fig. 44, the regeneration zone 50008 of the (KD) th light modulation step is performed for each K th light modulation step in the same manner as in the second column from the 168 th light modulation step to the 251 th light modulation step. (J + H * B) rows of (I + H * A) rows of mirror unit masks are shifted by (-H * B) rows (-H * A) columns in the area of the update zone 60008 of the kth optical modulation step. The J row I column mirror unit masks are updated, and the J row I column mirror unit masks of the new region 60009 of the K th light modulation step are generated to form the K th digital mask. That is, the mirror unit masks of the reproduction zone 50008 of the 84th light modulation step are shifted to the mirror unit masks of the update zone 60008 at the 168th light modulation step (26800A), and the mirror unit masks of the new zone are generated. (26800B) to configure the 168th digital mask 26800, and shift the mirror unit masks of the reproduction area 50008 of the 167th light modulation step from the 251th light modulation step to mirror the renewal area 60008. The unit masks are updated to unit masks (35100A) and mirror unit masks of the new zone (35100B) to form a 251 th digital mask 35100.

As shown in column 5 of FIG. 44, the reproduction zone of the (KD) th light modulation step is performed for each Kth light modulation step from the 588th light modulation step to the 671th light modulation step in the same manner as the second to third columns. (J + H * B) (I + H * A) column mirror unit masks of (50008) by (-H * B) row (-H * A) columns to change the area of the k-th optical modulation step ( 6000 row), and then, row J masks are generated, and the row J column mirror masks of the new region 60009 of the K th light modulation step are generated to form the K th digital mask. That is, in the 588 th light modulation step, the mirror unit masks of the reproduction area 50008 of the 504 th light modulation step are shifted to be updated to the mirror unit masks of the update area 60008 (68800A), and the mirror unit masks of the new area are generated. (68800B) to construct the 588th digital mask 68800, and shift the mirror unit masks of the reproduction area 50008 of the 587th light modulation step from the 671th light modulation step to mirror the renewal area 60008. The unit masks are updated to unit masks (77100A) and mirror unit masks of the new zone (77100B) to form the 671th digital mask 77100. As shown in Fig. 44, (N * M) mirror unit masks should be generated when the light modulation step K is 0 or more and D-1 or less, but when the light modulation step K is D or more and Q or less, the ((NH) By updating * B) * (MH * A)) mirror unit masks with shift only, only generation of mirror unit masks of [(N * M)-(NH * B) * (MH * A)] new zones need.

Therefore, the compression method of the digital mask of the present invention is a structural compression method, and the structural compression is N horizontal in each K-th optical modulation step from the initial 0 th light modulation step to the D-1 th light modulation step. Create all M J row I column mirror unit masks and, from the D th light modulation step to the Q th light modulation step, the J row I column mirrors of the new zone 60009 of the K th light modulation step for each K th light modulation step. By creating only unit masks. The amount of data compressed by the present invention is always a fixed quantity, and the compressed pattern mask by the entire light modulation steps is 10000, 10100, 10200, ...., 18300 and 18400B, 18500B, 18600B, ...., 26700B and 26800B, 26900B, 27000B, ...., 35100B, ...., and 68800B, 68900B, 69000B, ...., 777100B. The decompression method of the digital mask of the present invention is a lossless restoration method, wherein the lossless restoration is the reproduction of the (KD) th optical modulation step for each Kth optical modulation step from the Dth optical modulation step to the Qth optical modulation step. (J + H * B) row (I + H * A) column mirror unit masks in area 50008 by (-H * B) row (-H * A) columns to update the K-th optical modulation step This is accomplished by updating to the J row I column mirror unit masks of (60008). Therefore, according to the present invention, when the light modulation step K is (D≤K≤Q), the amount of data that is quantitatively compressed or losslessly restored for each Kth light modulation step is ((NH * B) * ( The amount of MH * A)) mirror unit mask data. Therefore, the compression (including restoration) method of the digital mask according to the present invention is a quantitative lossless compression method, and the compression rate when the optical modulation step K is (D≤K≤Q) is {[(N * M)-(NH *). B) * (MH * A)] * 100 / (N * M)}%. For example, when N is 1024, M is 768, A is 16, B is 1, and H is 1, the compression ratio is about 2.179%. The simulation of the discrete type exposure process of FIG. 1 (a), which satisfies all three conditions of the present invention and the scan ratio H is different from the scan resolution D, and the analog type exposure process simulation of 1 (c) Digital masks were generated by compressing and restoring by shift by a quantitative lossless compression method. N in Figs. 1 (a) and 1 (c) is 180, M is 1640, A is 164, B is 3, and H is 5, so the compression ratio of Figs. 1 (a) and 1 (c) is about 11.500%. . Therefore, by using the structural quantitative lossless compression method of the present invention, digital pattern lithography using a spatial light modulator that can be applied to the mass production process of a large-area high-resolution substrate is possible by generating and transmitting a large amount of pattern mask structurally and quantitatively and losslessly. This was confirmed.

The structural quantitative lossless compression method of the present invention is represented by the step S790 in the flowchart of FIG. 33, which is a representative view, and as the step (10) of the description of the embodiment of the normal lithography method. Quantitative lossless compression methods. In addition, in the present invention, in order to ensure a uniform distribution of the amount of exposure energy formed on the substrate, in determining the total vertical maze number, the total horizontal mirror number M, and the final step number Q in digital lithography using a spatial light modulator. It is recommended that N, M, and Q be determined so that the vertical maze number N is an integer multiple of H * B, the total vertical mirror number M is an integer multiple of H * A, and the final step number Q is an integer minus one from the integer multiple of D. 33, which is shown in the flowchart of FIG. 33 as step S680 and included in step (9) of the description section of the embodiment of the normal lithography method, which is included in the normal lithography method of the present invention. 34 to 44 summarize and summarize the above discussion, the total number of mirror mirrors performed by the Q + 1 optical modulation steps from the initial 0th optical modulation step to the final Qth optical modulation step is M, In digital lithography using a spatial light modulator having a mirror number of N, a unique characteristic of the present invention for the generation or transmission of a K-th digital mask composed of N vertical M horizontal M J-row I-column mirror masks of a K-th optical modulation step Specific examples of the structural quantitative lossless compression method (step S790) based on the horizontal cell number A, vertical cell number B, scan ratio H, and scan resolution D will be described as follows. The following examples are intended to illustrate the invention, not to limit the scope thereof, and the scope of the invention is defined by the claims.

(10) Structural Quantitative Lossless Compression Method (Step S790)

(10a) Initial generation of the digital mask at (0≤K≤D-1), where (1≤J≤N) && (1≤I≤M), the J-row I column mirror units of the K-th optical modulation step Create a mask.

In (10b) (D≤K≤Q), the quantitative compression and lossless recovery of the digital mask is composed of the following two steps.

Quantitative compression of the digital mask at (10b-1) (D≤K≤Q), where [(NH * B + 1≤J≤N) && (1≤I≤MH * A)] || [(1≤J ≦ N) && (MH * A + 1 ≦ I ≦ M)], a J-row I column mirror unit mask of the K-th optical modulation step is generated.

(10b-2) Lossless recovery of the digital mask at (D≤K≤Q), where (1≤J≤NH * B) && (1≤I≤MH * A), where the (KD) th optical modulation step (J + H * B) row (I + H * A) column mirror unit The mask is shifted by (-H * B) row (-H * A) columns, and the J row I column mirror units of the Kth optical modulation step Renew with a mask.

The symbol ≤ used represents N6 ≤ N7, for example, that N6 is equal to or less than N7, and the symbol && used represents N6 and N7 at the same time, for example, in the case of N6 && N7. Satisfaction is expressed, and the symbol || used indicates that only one of N6 or N7 is satisfied in the sense of or of N6 and N7, for example. 1 to 44 are intended to illustrate the invention, not to limit its scope, and the scope of the invention is defined by the claims.

1 is a view showing that the patterning precision is different according to the pitch, Figure 1 (a) and 1 (b) is a discrete method Figure 1 (c) and 1 (d) is an analog method Figure 1 (a) and 1 (c) is 0.7035 micron pitch and FIGS. 1 (b) and 1 (d) are exposure simulation results of 0.3519 micron pitch.

FIG. 2 is a view showing a projection structure formed of 10 horizontal cells and 3 vertical cells by 84 optical modulation steps on a substrate that is reflected by 10 horizontally and vertically 3 micromirrors. FIG.

3 is a view showing only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG.

4 is a view showing a projection structure formed by only the image centers in FIG.

5 is an enlarged view of a portion 10010 of FIG. 4;

FIG. 6 is a view illustrating reduced image centers after enlarging portion 10020 of FIG. 5; FIG.

FIG. 7 illustrates an isosceles triangle structure by three image centers as the basis of the projection structure in FIG. 6; FIG.

FIG. 8 is a view showing a projection structure formed of eight horizontal and three vertical cells by 57 light modulation steps on a substrate reflected and moved by three horizontal and ten micromirrors. FIG.

FIG. 9 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 8; FIG.

FIG. 10 is a view showing a projection structure formed only of image centers in FIG. 8; FIG.

FIG. 11 is a view showing a projection structure formed of 6 horizontal cells and 3 vertical cells by 36 light modulation steps on a substrate reflected and moved by 10 horizontal 3 vertical micromirrors. FIG.

FIG. 12 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 11; FIG.

FIG. 13 illustrates a projection structure formed only of image centers in FIG. 11; FIG.

FIG. 14 is a view showing a projection structure formed by the number of horizontal cells 10 and the number of vertical cells 3 by 68 light modulation steps on a substrate reflected and moved by the horizontal 10 vertical 3 micromirrors. FIG.

FIG. 15 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 14; FIG.

FIG. 16 is a view showing a projection structure formed only of image centers in FIG. 14; FIG.

FIG. 17 is a delta projection structure simulation result. FIG. 17 (a) shows a distribution of light beams irradiated onto a substrate in a stationary state, and FIG. 17 (b) shows a projection angle of 60 degrees with a projection length ratio of 2 / SQRT. (3) is a diagram of an exposure simulation result in which the horizontal cell count is 10 and the vertical cell count is 3, and the resolution is 84. FIG. 17 (c) shows that the scan resolution is 91.2 and the unit step number is 7, and the condition 2 is not satisfied. Fig. 17 (d) is a diagram of exposure simulation results when the scan resolution is 109 and the unit step number is 8.5 and condition 3 is not satisfied.

Fig. 18 is a simulation result of a triangular projection structure. 18 (b) is a diagram of an exposure simulation result obtained by scanning a projection angle of 58 degrees, a projection length ratio of 1, a horizontal cell number of 10, and a vertical cell number of 1 with a resolution of 86; and FIG. 18 (c) shows a projection shape. The angle is 58 degrees, the projection length ratio is 1, the horizontal cell count is 10, the vertical cell count is 3, and the resolution is 94. The drawing shows the result of the exposure simulation. A diagram of the exposure simulation result obtained by scanning the horizontal cell number 10 and the vertical cell number 3 with / SQRT (3).

19 (a) shows a distribution of light beams irradiated onto a substrate in a stationary state as a result of the simulation of a stripe projection structure. FIG. 19 (b) shows a projection length ratio of 90 degrees at a projection shape angle of 2. FIG. Fig. 19 (c) is a diagram of exposure simulation results in which the cell count is 10 and the vertical cell count is 3 and the resolution is 68. FIG. 19 (c) shows the exposure simulation result when the scan resolution is 87 and the unit step count is 7.5 and condition 3 is not satisfied. 19 (d) is a diagram of exposure simulation results when the scan resolution is 125 and the number of unit steps is 6.6666667 and condition 3 is not satisfied.

20 (a) shows the distribution of the light beams irradiated onto the substrate in a stationary state as a result of the delta projection structure simulation. FIG. 20 (b) shows the projection length ratio of 60 degrees at 4 / SQRT. Fig. 3 shows the results of exposure simulations in which the horizontal cell count is 20 and the vertical cell count is 3, and the resolution is 168.

FIG. 21 shows that the projection length can be increased while increasing the projection length ratio while keeping the projection structure the same. FIG. 21 (a) shows an initial light modulation step which is the basis of the projection structure of FIG. A unit projection structure and a final light modulation step unit projection structure are shown, and FIG. 21 (b) shows an initial light modulation step unit projection structure and a final light modulation step unit projection structure which are the basis of the projection structure of FIG. 20 (b). Figure.

FIG. 22 is a view showing a projection structure in which a scan ratio is set to 4 in a horizontal cell number and 1 in a vertical cell number by 8 light modulation steps on a substrate reflected and moved by 12 horizontal and 3 fine micromirrors. FIG.

FIG. 23 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 22;

FIG. 24 is a view showing a projection structure formed only of image centers in FIG. 22; FIG.

FIG. 25 is a view showing a projection structure in which a scan ratio is set to 4 in a horizontal cell number and 1 in a vertical cell number by 8 optical modulation steps on a substrate reflected and moved by 12 horizontal and 3 fine micromirrors. FIG.

FIG. 26 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 25; FIG.

FIG. 27 is a view showing a projection structure formed only of image centers in FIG. 25; FIG.

FIG. 28 is a diagram showing a projection structure in which the scan ratio is 3 and the vertical cell number 3 is 8 by 8 light modulation steps on a substrate reflected and moved by 12 horizontal and 3 fine mirrors.

FIG. 29 is a view illustrating only the initial light modulation step unit projection structure and the final light modulation step unit projection structure in FIG. 28;

30 is a view showing a projection structure formed only of image centers in FIG. 28;

FIG. 31 shows exposure simulation results of increasing the scan ratio in the delta projection structure of FIG. 17. FIG. 31A shows the scan ratio 2, FIG. 31B shows the scan ratio 3, and FIG. 31C shows the scan ratio 5, FIG. (d) is a diagram of exposure simulation results performed at scan ratio 6.

32 is a diagram illustrating exposure simulation results of increasing scan ratio in the stripe projection structure of FIG. 19. FIG. 32 (a) shows scan ratio 2 and FIG. 32 (b) shows scan simulation result.

Figure 33 is a chart showing a normal lithography method of the present invention.

FIG. 34 shows the initial, 34th, 84th, 118th, 168th, among the light modulation step unit projection structures formed by the 370 light modulation steps on the substrate reflected and moved by the 40 vertical and 15 fine mirrors; A drawing showing only the 202th, 252th, 286th, 336th, and 370th optical modulation step unit projection structures.

FIG. 35 is a view illustrating only the initial, 84th, 286th and 370th light modulation step unit projection structures in FIG. 34; FIG.

FIG. 36 is a view illustrating (K-D) th and Kth light modulation step unit projection structures in FIG. 34; FIG.

FIG. 37 shows FIG. 36 divided into three regions of four zones.

FIG. 38 is a view showing a projection structure formed by eight light modulation steps by increasing the total length mirror number to 6 and the total length mirror number to 24 in FIG. 22; FIG.

FIG. 39 is a view showing only the initial and eighth light modulation step unit projection structures in FIG. 38;

FIG. 40 shows FIG. 39 divided into three regions of four zones.

FIG. 41 is a view showing a projection structure formed by eight light modulation steps by increasing the total length mirror number to 6 and the total length mirror number to 24 in FIG. 28; FIG.

FIG. 42 is a view showing only the initial and eighth optical modulation step unit projection structures in FIG. 41;

FIG. 43 shows FIG. 42 divided into three regions of four zones.

FIG. 44 illustrates an embodiment of a method for structural quantitative lossless compression of a digital mask of the normal lithography method of the present invention. FIG.

Claims (21)

Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians; Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And The scanning resolution D, the projection length ratio Z, the image spacing C, the image spacing C, and the horizontal cell number A, with the projection angle Ψ between The scan direction resolution calculated as the vertical cell number B is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the super resolution S calculated as the horizontal cell number A and the vertical cell number B, so that the angle between the super resolution isosceles A regular lithographic method, characterized by maintaining an isosceles triangle projection structure with a projection shape angle Ψ. Determining a projection shape angle Ψ that defines a projection structure as π / 3 radians or 2π / 3 radians; Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And The scanning resolution D, the projection length ratio Z, the image spacing C, the image spacing C, and the horizontal cell number A, with the projection form angle Ψ between, including calculating the mirror angle θ based on the projection length ratio Z, the horizontal cell number A, and the vertical cell number B The resolution of the scan direction, which is calculated as the number of vertical cells B, is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the number of horizontal cells A, and the super resolution S calculated as the number of vertical cells B. And a delta projection structure of Determining a projection form angle Ψ that defines the projection structure as π / 2 radians; Based on the projection form angle Ψ, super resolution S, projection length ratio Z, image interval C, horizontal cell number A, and vertical cell number B are integers of 1 or more, where vertical cell number B is horizontal cell number A and mutually (coprime). Determining that the scan resolution D and the unit step number L, which are calculated as the projection form angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B, are both integers; And The scanning resolution D, the projection length ratio Z, the image spacing C, the image spacing C, and the horizontal cell number A, with the projection angle Ψ between The scan direction resolution calculated as the vertical cell number B is equal to the projection form angle Ψ, the projection length ratio Z, the image interval C, and the super resolution S calculated as the horizontal cell number A and the vertical cell number B, so that the angle between the super resolution isosceles And a stripe projection structure of this right-angled isosceles triangle. The method according to any one of claims 1, 2 and 3, The super resolution S is calculated as cscΨ * Z * C / (A ² + B ² * Z ² ) ^1/2 or C * (A ² + B ² * Z ² ) ^1/2 / D Lithography method. The method according to any one of claims 1, 2 and 3, The scan resolution D is calculated as C * (A ² + B ² * Z ² ) ^1/2 / S or sinΨ * (A ² + B ² * Z ² ) / Z, and the unit step number L is sinΨ * A A normal lithographic method, characterized by / Z-cosΨ * B. The method according to any one of claims 1, 2 and 3, Rotating the micromirror array of the spatial light modulator with respect to the substrate at the mirror angle θ, and setting the scan pitch P, which is the unit travel distance of the substrate corresponding to the unit light modulation time of the micromirror array, to super resolution S; And so that the position of the image center in row J column I of the arbitrary K th light modulation step is structurally identical to the position of the image center in row J + B row I + A column of the KD th light modulation step. The method according to any one of claims 1, 2 and 3, The unit of the substrate corresponding to the optical mirror modulation time of the micromirror array by rotating the micromirror array of the spatial light modulator at the mirror angle θ with respect to the substrate, and determining the scan ratio H by an integer greater than 1 and dissimilar to the scan resolution D. And further setting the scan pitch P, which is the moving distance, to H * S, which is the product of the scan ratio H and the super resolution S, while increasing the scan ratio while maintaining super resolution, while at the same time, image of column J, column I of any K th light modulation step. And the position of the center is structurally coincident with the position of the image center of row I + H * B in row J + H * B of the KD th light modulation step. Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians; Based on the projection form angle Ψ, the fixed resolution G, the projection length ratio Z, the image spacing C, the horizontal cell number A and the vertical cell number B are integers of 1 or more, where the vertical cell number B is the horizontal cell number A and the other (coprime) Determining to make; Calculating a scan resolution D based on the projection shape angle Ψ, the projection length ratio Z, the horizontal cell number A, and the vertical cell number B; And The scan resolution S is calculated based on the projection length ratio Z, the image interval C, the horizontal cell number A, the vertical cell number B, and the scan resolution D, and the mirror angle θ is calculated based on the projection length ratio Z, the horizontal cell number A and the vertical cell number B. And maintaining a triangular projection structure in which the angle between the fixed resolution side and the scan resolution side is the projection shape angle Ψ. The method of claim 8, The fixed resolution G is cscΨ * Z * C / (A 2 + B 2 * Z 2) 1/2 calculated, and the scan resolution D is a value rounded off to the first place below the R a ROUND {R, 0} point SinΨ * (A ² + B ² * Z ² ) / Z, (ROUND {sinΨ * A / Z-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² * Z ² ) / A , ROUND {sinΨ * (A ² + B ² * Z ² ) / Z, 0} and ROUND {(ROUND {sinΨ * A / Z-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² And calculating any one of * Z ² ) / A, 0}, and the scan resolution S is calculated by C * (A ² + B ² * Z ² ) ^1/2 / D. . The method according to any one of claims 1, 2, 3 and 8, And the projection length ratio Z is determined by cscΨ * V / W or cscΨ using two integers, V and W, which are mutually different. The method according to any one of claims 1, 2, 3 and 8, And setting the horizontal interval between image centers to the image interval C, and setting the vertical interval between image centers to Z * C, which is the product of the projection length ratio Z and the image interval C. Way. The method according to any one of claims 1, 2, 3 and 8, And the mirror angle θ is calculated in tan ⁻¹ (B * Z / A) radians. Determining a projection shape angle Ψ defining a projection structure to not more than π / 4 radians but not more than 3 pi / 4 radians; Determining the fixed resolution G, the image interval C, the horizontal cell number A, and the vertical cell number B such that the vertical cell number B is an integer equal to or greater than each other (coprime) based on the projection form angle Ψ; Calculating a scan resolution D based on the projection shape angle Ψ, the horizontal cell number A and the vertical cell number B; And Calculating the scan resolution S based on the image interval C, the horizontal cell number A, the vertical cell number B, and the scan resolution D, and calculating the mirror angle θ based on the horizontal cell number A and the vertical cell number B. A regular lithographic method, characterized by maintaining a triangular projection structure in which the angle between resolution sides is the projection form angle Ψ. The method of claim 13, The fixed resolution of G is ^{C * cscΨ / (A 2 +} B 2) 1/2 calculated, and the scan resolution D by a value rounded off to the first place below the R a ROUND {R, 0} point sinΨ * ( A ² + B ² ), (A ² + B ² ) / (1-B / (A * tanΨ)), ROUND {(A ² + B ² ) / (1-B / (A * tanΨ)), 0 }, (ROUND {sinΨ * A-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² ) / A, ROUND {sinΨ * (A ² + B ² ), 0} and ROUND {(ROUND {sinΨ * A-cosΨ * B, 0} + cosΨ * B) * (A ² + B ² ) / A, 0} is selected and calculated, and the scan resolution S is C * (A ² + B ² ) A regular lithographic method, characterized by calculating ^1/2 / D. 14. The lithographic method of claim 13, wherein the mirror angle θ is calculated in tan ⁻¹ (B / A) radians. The method according to claim 8 or 13, Rotating the micromirror array of the spatial light modulator with respect to the substrate at the mirror angle θ, and setting the scan pitch P, which is the unit travel distance of the substrate corresponding to the unit light modulation time of the micromirror array, to the scan resolution S; And so that the position of the image center in row J column I of the arbitrary K th light modulation step is structurally identical to the position of the image center in row J + B row I + A column of the KD th light modulation step. The method according to claim 8 or 13, The unit of the substrate corresponding to the optical modulation time of the micromirror array by rotating the micromirror array of the spatial light modulator at the mirror angle θ with respect to the substrate, and determining the scan ratio H by an integer greater than 1 and dissimilar to the scan resolution D. Further comprising setting the scan pitch P, which is the moving distance, to H * S, which is the product of the scan ratio H and the scan resolution S, while increasing the scan ratio while maintaining the scan resolution, at the same time, image of column J, column I of any K th light modulation step. And the position of the center is structurally coincident with the position of the image center of row I + H * B in row J + H * B of the KD th light modulation step. 17. The K-th light of the image center array of N rows and M columns according to claim 6 or 16, in a total Q + 1 light modulation step by N vertical M horizontal mirror mirror arrays involved in exposure. Based on the regularity in which the position of the image center of row J of column I of the modulation step is structurally identical to the position of the image center of row J + B of column I + B of the KDth optical modulation step, For each Kth light modulation step where K is equal to or greater than 0 and equal to or equal to D-1, the mirror unit mask corresponding to each row J of column I for columns J of column I satisfying that J is 1 or more and N or less and I is 1 or more and M or less. Generating a K th digital mask; In each K-th optical modulation step where K is greater than or equal to D and less than or equal to Q, row J columns I satisfying that J is greater than or equal to N-B + 1 and less than N, and that I is greater than or equal to 1 and less than or equal to MA, and I is greater than or equal to 1 and less than or equal to N, Quantitatively compressing the K-th digital mask by generating a mirror unit mask corresponding to each J row I column for the J row I columns satisfying A + 1 or more and M or less; And For each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, for each row J column I satisfying that J is greater than or equal to 1 and less than or equal to NB and I is greater than or equal to 1 and less than or equal to MA, row J + B is greater than or equal to Shifting the + A mirror unit mask by -B rows and -A columns to the row J mask of the Kth light modulation step to update the column I masks in the Kth optical modulation step, further including lossless restoring the compressed portion of the Kth digital mask. A lithographic method, characterized in that structurally quantitative lossless compression. 19. The method of claim 18, wherein prior to creation, compression and decompression of the digital mask: In a total Q + 1 optical modulation step with N vertical and horizontal M micromirror arrays involved in exposure, vertical N is an integer multiple of B, horizontal M is an integer multiple of A, and Q is an integer multiple of D. And a step of determining N, M, and Q to be a subtracted integer. 18. The K-th light of an image center array of N rows and M columns in a total Q + 1 optical modulation step by N vertical and horizontal M micromirror arrays involved in exposure. Based on the regularity in which the position of the image center in row J column I of the modulation step is structurally identical to the position of the image center in row J + H * B row I + H * A of the KD th light modulation step, For each Kth light modulation step where K is equal to or greater than 0 and equal to or equal to D-1, the mirror unit mask corresponding to each row J of column I for columns J of column I satisfying that J is 1 or more and N or less and I is 1 or more and M or less. Generating a K th digital mask; In each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, row J columns I and J are greater than or equal to 1 and less than or equal to N while at the same time J is NH * B + 1 or more and N or less and I is greater than or equal to 1 or more MH * A Quantitatively compressing the K-th digital mask by generating a mirror unit mask corresponding to each J-row I column for the J-row I columns satisfying MH * A + 1 or more and M or less; And In each Kth light modulation step where K is greater than or equal to D and less than or equal to Q, for each row J of column I satisfying that J is greater than or equal to 1 and less than NH * B and I is greater than or equal to 1 and less than MH * A, J of the KDth light modulation step Compresses the K-th digital mask by shifting the + H * B-row I + H * A-column mirror unit mask by -H * B-row -H * A column to the J-row I-mirror mask of the K-th optical modulation step And losslessly restoring the portion to structurally quantitatively and losslessly compress the digital mask. The method of claim 20, wherein before the digital mask is generated, compressed and decompressed, In a total Q + 1 optical modulation step with N vertical and horizontal M micromirror arrays involved in exposure, vertical N is an integer multiple of H * B, horizontal M is an integer multiple of H * A, and Q is D And determining N, M, and Q to be an integer less than 1 from an integer multiple.

Images