JP2005302800A - Resolution determination technique of lithography - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a resolution determination technique of lithography capable of accurately estimating the resolution without preparing an estimating photomask requiring minute processing burdened with a load. <P>SOLUTION: The resolution determination technique of lithography adds background light by performing all transmission exposure with the exposure amount of such a degree that an estimation pattern is resolved to intentionally deteriorate an exposure state followed by an exposure processing of the estimation pattern. Resolving power of the estimation pattern is determined by changing the background light. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a resolving power determination method for lithography, and relates to a resolving power determination method for lithography capable of accurately evaluating the resolution of a fine pattern.

In recent years, along with the miniaturization of patterns accompanying the increase in the density of semiconductor integrated circuits, improvement in resolving power has become a major issue in lithography. As means for improving the resolution, development of a high resolution resist and search for optimum exposure conditions can be mentioned. The main exposure conditions are an optical element and a photomask.
When discussing resolving power by simulation, an index called contrast or NILS (Normalized Image Log-Slope) is usually used. The contrast C is defined as follows.

C = (Imax−Imin) / (Imax + Imin) (1)

Here, Imax and Imin are the light intensities of the bright part and the dark part, respectively. NILS is defined as follows.

NILS = (dI / dx) / (W × Ith) (2)

Here, W is a desired dimension, Ith is a threshold value of light intensity giving W, and (dI / dx) is a gradient of the aerial image.

FIG. 10 is a chart showing the relationship of each parameter in the optical image.
From FIG. 10, when the contrast C is large, the amplitude of the optical image increases and the resolution of the resist pattern is improved. Further, if the NILS value is large, the optical image becomes steep and the dimensional controllability of the resist pattern is improved.
In general, in order to obtain sufficient resolution, it is desirable that the contrast C is 0.5 or more and the NILS is 2 or more. However, with the miniaturization of semiconductor design rules, it has become difficult to obtain sufficient contrast C and NILS. For this reason, for example, there is a demand for a resist process that resolves even with a contrast C of less than 0.4 and a NILS of about 1.5.

As a method for evaluating the resolving power, for example, there is a method of evaluating by resolving or unresolving a repetitive pattern (line and space: hereinafter referred to as L & S) in which the pitch is changed stepwise.
FIGS. 11 and 12 are schematic diagrams showing exposure using a typical L & S evaluation pattern. (A) is a photomask pattern, (b) is an optical image on the wafer, and (c) is a resist pattern.

In the photomask pattern of FIG. 11A, the pattern is an L & S pattern with a half pitch 1101, and a light shielding portion 1102 and a transmissive portion 1103 are continuously formed.
In the optical image of FIG. 11B, the portion of the exposure light shielded by the light shielding portion 1102 in FIG. 11A corresponds to the dark portion 1104 and the bright portion 1105 that has passed through the transmission portion 1103.
Since the resist pattern in FIG. 11C uses a positive resist, a pattern portion 1106 is formed for the dark portion 1104 and a space portion 1107 is formed for the bright portion 1105.

FIG. 12 shows a finer L & S evaluation pattern than FIG.
In the photomask pattern of FIG. 12A, the pattern is an L & S pattern having a finer half pitch 201 than that of FIG.
In the optical image of FIG. 12B, it can be seen that the difference in light intensity between the bright portion 1203 and the dark portion 1202 cannot be secured as compared with the optical image of FIG. That is, since the contrast C described above is insufficient, the resist pattern is difficult to resolve as shown in FIG. 12C, and a problem that the resist remains in the space 1204 occurs.

  Until now, the level of the resolution evaluation pattern has been provided in the photomask by changing it at a half pitch and with a step size of 5 to 10 nanometers on the wafer. However, with the rapid miniaturization of semiconductor design rules in recent years, the dimension of the evaluation pattern has begun to cut below 100 nanometers. If the half pitch is in increments of 5 to 10 nanometers with respect to a pattern width of 100 nanometers, it will be 5 to 10% of the pattern width. For this reason, at the level of the conventional evaluation pattern, the evaluation of resolving power has been hindered.

FIG. 13 is a chart showing the relationship between the half pitch of the L & S pattern and the NILS predicted from the simulation. The calculation conditions are a halftone mask having a wavelength of 157.6 nanometers (F2 laser), a numerical aperture of 0.90, annular illumination of 0.8 / 0.6, and a transmittance of 6%.
FIG. 13 shows that NILS decreases as the half pitch becomes finer. In particular, NILS is significantly reduced from 1.95 to 1.49 from 60 nanometers to 55 nanometers.

  Thus, it is clear that the resolution cannot be effectively evaluated unless the half pitch of the evaluation pattern is changed more finely than in the conventional 5-nanometer increments. However, finely changing the half pitch places a large burden on the mask creation. Specifically, the grid size at the time of photomask drawing must be made fine, and mask data and drawing time become enormous. Furthermore, mask manufacturing errors such as the half pitch itself and the L & S duty ratio cannot be ignored.

Although there is no direct relationship with the resolution evaluation method, Patent Document 1 discloses a double exposure method in which different reticles are used in each exposure step as a method of exposing the irradiated object twice. Yes.
JP-A-11-111601

  As described above, in the lithography evaluation method for fine patterns, a considerable load is applied to the production of the evaluation pattern, and it is difficult to evaluate the resolution in detail.

  The present invention has been made based on recognition of such a problem, and an object of the present invention is to provide a lithography resolving power determination method capable of accurately evaluating resolution without producing a heavy photomask. There is.

  In the lithography resolving power judgment method of the present invention, subsequent to the exposure processing of the evaluation pattern, by performing total transmission exposure with an exposure amount sufficient to resolve the evaluation pattern, background light is added and the exposure state is intentionally changed. Deterioration. Then, the resolution of the evaluation pattern is determined by changing the background light.

  That is, according to the present invention, a method for determining the resolving power of lithography using a photomask provided with an evaluation pattern, and using the first evaluation pattern provided with the pattern, the different positions on the wafer. In addition, exposure is performed by changing the exposure amount to each of the different positions on the wafer using the first exposure process in which exposure is performed by changing the exposure amount and a total transmission pattern in which no pattern is provided. A determination process for determining the resolving power by performing a development operation after observing the formed composite evaluation pattern after performing the second exposure process and the first and second exposure processes in an overlapping manner, and A method for determining the resolution of lithography is provided.

  Here, the first exposure process can be performed by changing the exposure amount in the vicinity of the exposure amount that the first evaluation pattern resolves.

  In addition, the second exposure process can be performed by changing the exposure amount within a range of 0 to 30% of the exposure amount resolved by the first evaluation pattern.

Further, the first exposure process may be performed after the second exposure process.
Further, in the first exposure process, the exposure amount is changed m times with n steps, and in the second exposure process, the exposure amount is changed n times with m steps, and m × n composite evaluation patterns. Can be created.

In the determination process, it can be determined that the larger the number of composite evaluation patterns having a desired dimension is, the higher the resolution is.
In the determination process, a resolution limit pattern condition is defined as a resolution limit pattern condition including a ratio of an exposure amount in the second exposure process to an exposure amount in the first exposure process among resolution limit patterns. It can also be determined that the larger the condition, the higher the resolution.
The first evaluation pattern and the total transmission pattern may be provided on the same photomask.

  According to the present invention, subsequent to the exposure processing of the evaluation pattern, by performing total transmission exposure with an exposure amount sufficient to resolve the evaluation pattern, background light is added and the exposure state is intentionally deteriorated. . Then, the resolution of the evaluation pattern is determined by changing the background light. By using such a method, it becomes possible to observe the resolving power in detail without producing a fine evaluation pattern. That is, the same resolution evaluation can be performed as when the pitch of the evaluation pattern is finely chopped.

  FIG. 1 is a flowchart showing a procedure of a resolution determination method for lithography according to the present invention. Hereinafter, embodiments of the present invention will be described with reference to this flowchart.

  First, a first exposure process is performed using the first evaluation pattern. At this time, the exposure amount is changed stepwise in the vicinity of a value at which the evaluation pattern is resolved well. For example, the exposure amount is changed to n stages and is repeated m times to obtain a total of n × m shots (S101).

2A and 2B are schematic views showing the first exposure process, where FIG. 2A is a photomask in the first exposure process, and FIG. 2B is a shot map thereof.
In the first exposure process, exposure is performed by changing the exposure amount only in the horizontal direction of the wafer 202 as shown in FIG. 2B using the photomask 201 provided with the evaluation pattern shown in FIG. To do. The maximum value of the exposure amount is the optimum exposure amount (Eop) of the evaluation pattern, and the step size of the exposure amount is about several percent of Eop. Here, Eop is 24 mJ / cm ² , and the step size is changed in 7 steps from ² mJ / cm ² . Exposure processing is sequentially performed with the left end set to 12 mJ / cm ² so that Eop comes to the right end of the wafer 202. As shown in FIG. 2B, a total of 7 × 7 = 49 shots are obtained by performing shots with the same exposure amount seven times in the vertical direction of the wafer 202. At this time, the focus is set to the best focus state obtained in advance.

  Returning again to FIG. 1, the description will be continued. Subsequent to the exposure process of the first exposure process S101, the second exposure process is performed using the second evaluation pattern so as to overlap the shot obtained by the first exposure process. At this time, the exposure amount is changed stepwise within a range of 0 to 30% of a value at which the first evaluation pattern is well resolved. For example, the exposure amount is changed to m steps and repeated n times, so that a total of m × n shots are obtained (S102).

FIG. 3 is a schematic diagram showing the second exposure process, (a) is a photomask in the second exposure process, and (b) is a shot map thereof.
In the second exposure process, the exposure amount is changed only in the vertical direction of the wafer 202 as shown in FIG. 3B using the all-transmissive photomask 301 not provided with the pattern shown in FIG. To expose. The minimum value of the exposure amount is 0 mJ / cm ² , and the step size of the exposure amount may be about several percent of Eop in the first exposure process. Here, the step width is 0.5 mJ / cm ² and is changed in seven steps. As the lower end the upper end of the wafer 202 is at 0 mJ / cm ² is 3.0 mJ / cm ^2, subjected to sequential exposure process. As shown in FIG. 3B, a total of 7 × 7 = 49 shots are obtained by performing shots with the same exposure amount seven times in the horizontal direction of the wafer 202. At this time, the focus is set to the best focus state obtained in advance, but it is not always necessary to be the best focus.

  Returning again to FIG. 1, the description will be continued. After the first exposure process (S101) and the second exposure process (S102), the wafer 202 is developed. Since no development process is performed between the first and second exposure processes, the second exposure process acts as a background component of the first exposure process. That is, a pattern with a degraded resolution is obtained compared to the case where only the first exposure process is performed (S103).

  After the development process (S103), the resolution of the lithography is evaluated using the obtained m × n patterns (S104).

  FIG. 4 is a shot map showing the tendency of the resist pattern resolution after the first and second exposure processes are completed. When measuring the L & S pattern with a positive photoresist, the dimension of the line pattern becomes thinner as the sum of the exposure amounts of the first and second exposure processes increases. Therefore, the size of the line pattern tends to be thick at the upper left of the shot map and thin at the lower right of the shot map. As shown in FIG. 4, a shot having a desired pattern dimension appears in a hatched area. On the other hand, the resolution of the L & S becomes better as it goes to the upper right of the shot map, and the resolution of the L & S deteriorates as the lower left. That is, as E2 / E1, which is the ratio of the exposure amount E1 in the first exposure process and the exposure amount E2 in the second exposure process, increases, the resolution of the L & S deteriorates. Therefore, the resolution of the L & S deteriorates in the order of shots A, B,..., And is not resolved in the blackened area.

In FIG. 4, shot X corresponds to the resolution limit of the L & S pattern. The exposure amount ratio E2 / E1 in shot X is

2.0 / 16.0 = 12.5%

Is calculated.

In the embodiment of the present invention, the resolution of lithography is determined using the exposure amount ratio E2 / E1 at the resolution limit.
For example, when discussing the superiority or inferiority of the two types of resists A and B, a resolution limit shot may be obtained for each of the resists A and B. When the exposure limit ratio E2 / E1 of the resolution limit shot is 10% for the resist A and 15% for the resist B, it can be determined that the resist B has better resolution. This means that the resist B resolves even at a lower contrast.

  Also, even if the specifications are the same for some exposure machines, the resolution may vary depending on resolution impediment factors such as lens aberrations. Even in this case, the superiority or inferiority of the two exposure machines In order to discuss, the exposure amount ratio E2 / E1 of the resolution limit shot can be used. For each of the devices X and Y, the exposure limit ratio E2 / E1 of the resolution limit shot is obtained and compared. When the exposure limit ratio E2 / E1 of the resolution limit shot is 10% with the apparatus X and 15% with the apparatus Y, it can be determined that the apparatus Y has better resolution. Device Y can resolve lower contrast patterns.

Up to this point, in the description of the embodiment of the present invention, it has been assumed that separate photomasks are used during the first and second exposure processes. However, some people are concerned about the deterioration of throughput caused by exchanging the photomask during double exposure (Japanese Patent Laid-Open No. 11-11601). For this reason, in the present invention, the first and second evaluation patterns may be provided on the same photomask, and the exposure process may be performed while shifting the position. From the obtained evaluation pattern, exactly the same effect can be obtained as when separate photomasks are used.
In addition, there are no restrictions on the photomask, the photoresist, and the evaluation pattern used for the first exposure process, and a phase shift mask or the like as the photomask, a negative type as the photoresist, an isolated pattern or a hole pattern as the evaluation pattern, etc. It doesn't matter.

  Hereinafter, the effects of the embodiment of the present invention will be described based on the simulation results.

FIG. 5 is a chart showing an aerial image of L & S in the first exposure process.
The horizontal axis represents the position, and the vertical axis represents the normalized light intensity when the intensity of the entire transmission portion in the first exposure process is normalized to 1. Ith is a standardized light intensity threshold value, and usually takes a value of 0.25 to 0.3, depending on the pattern.

FIG. 6 is a chart showing an aerial image of L & S in the first and second exposure processes.
In order to align the position with FIG. 5 on the horizontal axis, the vertical axis indicates the normalized light intensity when the light intensity of the total transmission part in the first and second exposure processes is normalized to 1. Take. In the figure, the position indicated by the arrow on the vertical axis is a value corresponding to 1 in FIG. 5, and the hatched area is the background component, which is represented by the first and second exposure dose ratio E2 / E1. Is done. In order to study the contrast C and NILS with the pattern dimension as a desired value, the light intensity threshold Vth in FIG. 6 is set to the same value as the light intensity threshold Ith in FIG.
From FIG. 6, the aerial image of the L & S pattern in the exposure processing of the present invention corresponds to the aerial image shown in FIG. 5 contracted in the vertical axis direction with the light intensity threshold Ith as the center. Therefore, the maximum value of the aerial image of the L & S pattern decreases as the background component increases.

Contrast _{C B} and NILS _B when the background component is present, is represented by the following equation.

C _B = C × {1− (E2 / E1) / Ith} (3)

NILS _B = NILS × {1- (E2 / E1) / Ith} (4)

Here, C and NILS are the contrast C and NILS when there is no background component, and are expressed by the equations (1) and (2) described above. The background component E2 / E1 should not be larger than the light intensity threshold value Ith. This is because the pattern disappears when E2 becomes too large.

FIG. 7 is a chart showing the relationship between the half pitch of the L & S pattern by simulation and NILS. The horizontal axis is half pitch, the vertical axis is NILS, and the background component is taken into account from 0% to 10%. As in FIG. 13, the calculation conditions are a halftone mask having a wavelength of 157.6 nanometers (F2 laser), a numerical aperture of 0.90, annular illumination of 0.8 / 0.6, and a transmittance of 6%.
From FIG. 7, for example, the NILS is the same for an L & S pattern with a background light of 4% and a half pitch of 65 nanometers and an L & S pattern with a background light of 0% and a half pitch of 60 nanometers. Similarly, the NILS is the same for an L & S pattern with a background light of 8% and a half pitch of 60 nanometers and an L & S pattern with a background light of 0% and a half pitch of 55 nanometers.
Therefore, if the background light is changed within this range, the resolution of the intermediate half pitch can be evaluated. That is, by using an L & S pattern with a half pitch of 65 nanometers during the first exposure process, and adjusting the exposure amount so that the background component E2 / E1 is 0 to 4% during the second exposure process, the half pitch Can be evaluated in detail in the range of 60 to 65 nanometers. Similarly, by using an L & S pattern with a half pitch of 60 nanometers during the first exposure process, the exposure amount is adjusted so that the background component E2 / E1 is 0 to 8% during the second exposure process. The resolution in the range of 55 to 60 nanometers can be evaluated in detail.

  Thus, the level fluctuation of the background light only needs to change the total transmission exposure amount in the second exposure process, and can theoretically change infinitely. This means that just by preparing a normal evaluation pattern, the same evaluation as when a plurality of evaluation patterns are used in small increments can be performed.

The effect of evaluating the superiority or inferiority of the two types of resists using the embodiment of the present invention will be described below.
FIG. 8 is an enlarged view of a part of the chart of FIG.
For example, it is assumed that NILS necessary for resolving two types of resists A and B are 1.90 and 1.55, respectively. According to FIG. 8, when the background light is 0%, the NILS of the 60 nanometer L & S pattern and the 55 nanometer L & S pattern are 1.95 and 1.49, respectively. If an evaluation mask provided with a nanometer L & S is used, both the resists A and B are resolved with the 60 nanometer L & S, and neither the resists A and B are resolved with the 55 nanometer L & S. .
On the other hand, when NILS with background light added is used, NILS when 6% background is added to 60 nanometer L & S is 1.59. Under this condition, only resist B is resolved, and resist A will not be resolved. Therefore, it can be determined that the resist B has better resolution.

  Next, the effect when evaluating the difference between the units of the exposure apparatus will be described.

  FIG. 9 is a chart showing the relationship between background light and NILS in the apparatus X and apparatus Y by simulation. The evaluation pattern uses 60 nanometer L & S, and the background light is changed from 0 to 8%.

The ideal NILS for the 60 nanometer L & S pattern is said to be 1.95, but NILS deteriorated to 1.60 and 1.85, respectively, due to factors such as lens aberration and the like in both devices X and Y. Yes. In order to evaluate the superiority or inferiority of these two devices, if resist B with a resolution NILS of 1.55 is used, both the devices X and Y resolve the 60 nanometer L & S pattern, and the resolution can be evaluated. Absent.
On the other hand, when NILS to which background light is added is used, NILS when the background light is 4% is 1.35 for device X and 1.60 for device Y. Under this condition, only device Y After the resolution, the device X will be resolved. Therefore, it can be determined that the resolution of the device Y is superior.

As can be seen from the above comparison with the simulation results, the lithography resolution evaluation method of the present invention can be used to reduce the resolution in a finer state without imposing a great burden on the production of the evaluation pattern photomask. Evaluation is possible, and the resolution accuracy of lithography is remarkably improved.
This greatly contributes to semiconductor microfabrication technology.

It is a flowchart showing the procedure of the resolution determination method of the lithography of this invention. It is a schematic diagram showing a 1st exposure process. It is a schematic diagram showing a 2nd exposure process. It is a shot map showing the tendency of resist pattern resolution after the end of the first and second exposure processes. It is a chart showing the aerial image of L & S in a 1st exposure process. It is a chart showing the aerial image of L & S in the 1st and 2nd exposure processing. It is a graph showing the relationship between the half pitch of the L & S pattern by simulation, and NILS. FIG. 8 is an enlarged view of a part of the chart of FIG. It is the graph showing the relationship between the background light and NILS of the apparatus X and the apparatus Y by simulation. It is a chart showing the relationship of each parameter in an optical image. It is a schematic diagram showing the mode of exposure using the evaluation pattern of L & S. It is a schematic diagram showing the mode of exposure using the evaluation pattern of L & S. It is a chart showing the relationship between the half pitch of the L & S pattern and NILS predicted from the simulation.

Explanation of symbols

1101, 201 Half pitch 1102 Light-shielding portion 1103 Transmission portion 1104, 1202 Dark portion 1105, 1203 Bright portion 1106 Pattern portion 1107, 1204 Space portion 201 Photomask 202 Wafer 301 All-transmissive photomask

Claims (8)

A method for determining the resolution of lithography using a photomask provided with an evaluation pattern,
Using the first evaluation pattern provided with the pattern, a first exposure process for performing exposure with different exposure amounts at different positions on the wafer; and
A second exposure process in which exposure is performed by changing an exposure amount to each of the different positions on the wafer using a total transmission pattern in which no pattern is provided; and
A determination process for determining the resolving power by performing the development work after overlapping the first and second exposure processes and observing the formed composite evaluation pattern;
A method for determining the resolving power of lithography, comprising:   2. The lithography resolving power determination method according to claim 1, wherein the first exposure process is performed by changing an exposure amount in the vicinity of an exposure amount that is resolved by the first evaluation pattern.   3. The lithography according to claim 1, wherein the second exposure process is performed by changing an exposure amount in a range of 0 to 30% of an exposure amount that is resolved by the first evaluation pattern. Resolution determination method.   The lithography resolving power determination method according to claim 1, wherein the first exposure process is performed after the second exposure process. In the first exposure process, the exposure amount is changed n times and is performed m times.
In the second exposure process, the exposure amount is changed n times and m times,
The lithographic resolving power determination method according to claim 1, wherein m × n synthesis evaluation patterns are created.   6. The lithographic resolving power determination method according to claim 1, wherein in the determination process, it is determined that the higher the number of composite evaluation patterns having a desired dimension is, the higher the resolving power is.   In the determination process, the ratio of the exposure amount in the second exposure process to the exposure amount in the first exposure process among the resolution limit patterns is set as a resolution limit pattern condition, and the resolution limit pattern condition is 6. The lithography resolving power determination method according to claim 1, wherein the larger the resolving power, the higher the resolving power is determined. 8. The lithography resolving power determination method according to claim 1, wherein the first evaluation pattern and the total transmission pattern are provided on the same photomask.

Images