JP3321220B2 - Electron beam writing method and writing apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electron beam lithography technique, and more particularly to an electron beam lithography method and a lithography apparatus for reducing a proximity effect.

[0002]

2. Description of the Related Art In recent years, an electron beam drawing apparatus has been used to draw a fine pattern on a sample such as a semiconductor wafer or a mask substrate. In this apparatus, the pattern becomes thick and thin due to backscattered electrons. The effect of the so-called proximity effect becomes a problem. Therefore, recently, a proximity effect correction method called a ghost method has attracted attention.

The ghost method (for example, JP-A-59-921, G. Owen and P. Rissman, J. Appl. Phys. 54 (1983) 357)
In 3), first, a pattern to be drawn with the incident electron charge density Qp is drawn by using a normally focused electron beam (hereinafter, this is referred to as pattern drawing). Next, the beam is defocused to a diameter dc, and a portion having no pattern with the incident electron charge density Qc is irradiated with the beam (hereinafter, this is referred to as correction irradiation). Here, the beam diameter dc and the charge density Qc at the time of defocusing are as follows: dc = 2σb / (1 + η _E ) ^1/4 (1) Qc = Qp × η _E / (1 + η _E ) (2) Have been. However, .sigma.b radius the intensity of backscattered electrons becomes 1 / e, eta _E is a back scattering energy factor underlying the acceleration voltage of 40 kV .sigma.b = 7
μm, η _E = 0.7.

However, in the above ghost method,
When correcting and irradiating an area without a pattern, an area without a pattern generally has a larger area and a larger number of figures than an area with a pattern. It takes more time for correction irradiation than for drawing. Further, there is a problem that it takes a long time for data conversion to create an inverted pattern.

As another method for reducing the proximity effect, there is a dose correction method in which the dose is adjusted depending on the location based on the size and density of the pattern. Conventionally, when determining the dose by this dose correction method, a method using a matrix (M. Parikh, J. App.Phys. 19, p4371, P4378, p4383 (1979)
) Is used. The matrix method is a method in which the relationship between the irradiation amount and the light exposure amount at each position is expressed by a matrix, and the optimum irradiation amount at each position is obtained by obtaining an inverse matrix of the matrix.

However, the above-described irradiation dose correction method has a problem that the calculation time for determining the optimum irradiation dose continues to increase as the pattern becomes finer and more highly integrated. In particular, when the matrix method is used, the calculation time increases by the cube of the pattern density, and it is practically impossible to determine the optimal dose with the miniaturization of the pattern.

As a means for overcoming such a situation, for example, a representative figure method (JJAP vol 30, 3B, p521-531 (199)
1) A proximity effect correction method called T. Abe, S. Yamasaki, R. Yoshikawa, and T. Takigawa) has attracted attention. This correction method divides the entire correction area into small areas smaller than the spread of backscattering of the electron beam and larger than the smallest figure that can be drawn, and as a unit figure for irradiating the small area, a desired pattern in the small area. By setting a smaller number of representative unit figures than the number of irradiations when drawing a pattern in which is inverted from black and white, the number of shots of the correction irradiation is reduced, and the time required for the correction irradiation is shortened.

However, the following problems also exist in the representative figure method in which the number of correction irradiation shots can be significantly reduced as compared with the ghost method, and the time required for correction irradiation can be shortened. In other words, the correction irradiation sets the baseline of the energy application of the entire correction region to the backscattered electron amount at the place where the influence of the proximity effect is the largest, in other words, the place where the amount of backscattered electrons from the surrounding pattern is the largest. By matching, the proximity effect correction is performed so that the pattern size after development does not depend on the pattern density. For this reason, the developing process must be performed under the condition that the energy application contrast in the developing process is the strictest, and it has been difficult to resolve a fine pattern.

In the representative figure method, the number of reference figures is reduced by using the representative figure at the time of calculating the optimum dose in the dose correction, and the time required for determining the dose does not depend on the degree of integration of the pattern. It can be greatly reduced.
However, when a representative figure is set, the representative figure is set regardless of the total area of the figures included in the small area. Therefore, a small area, that is, a representative figure having little influence of the proximity effect on the peripheral pattern is also used for reference. Calculated as a figure,
A pattern including many small-area representative figures has a problem in that the time required to determine the optimum dose is long.

[0010]

As described above, conventionally, in the correction irradiation using the representative figure method, a sufficient energy storage amount contrast cannot be obtained, and it has been difficult to resolve a fine pattern. Further, in the dose correction using the representative figure method, there is a problem that the time required for determining the optimum dose is long in a pattern including many small-area representative figures.

The present invention has been made in view of the above circumstances, and has as its object to improve the energy storage contrast, which is a drawback of the correction irradiation in the ghost method and the representative figure method, and to improve the pattern at the time of development. An object of the present invention is to provide an electron beam lithography method and a lithography apparatus capable of improving the resolution and further reducing the time required for the correction irradiation to improve the throughput.

It is another object of the present invention to correct the irradiation dose to prevent a decrease in pattern dimensional accuracy due to the proximity effect, and to reduce the time required to determine the irradiation dose in the degree of pattern integration. Independent electron beam writing method
And a drawing device .

[0013]

The gist of the present invention is to divide an entire drawing area into small areas Δ at the time of correction irradiation, set a representative unit figure for each small area, and further represent a representative of all the small areas. By subtracting the reference area from the figure, the pattern density is high, and the place where the energy applied contrast is poor is not subjected to correction irradiation or the amount thereof is reduced, thereby improving the pattern resolution and improving the correction effect. An object of the present invention is to reduce the number of shots of the correction irradiation which is relatively small, thereby further reducing the time of the correction irradiation.

That is, according to the present invention (claim 1), a step of irradiating a sample with an electron beam to draw a desired pattern and, before or after this step, reducing a proximity effect due to backscattered electrons accompanying the pattern drawing. Therefore, in the electron beam writing method including the step of correcting and irradiating the sample with an electron beam, the correction irradiation step is performed as described below.

As a pre-process using a computer independent of the drawing system, or as a process inside the drawing system, the entire correction region is made smaller than the spread of the backscattered electron beam (or the distribution of the spread of the backscattered electron). (Same as or less than) and divided into small areas larger than the smallest drawable figure.

As a unit figure for irradiating each small area, 1
Or a smaller number of representative unit figures than the number of figures included in the small area.

The reference area (or the minimum area of the representative unit graphic set for the small area) is uniformly subtracted from the representative unit graphic set for the small area in the entire area.

An irradiation amount is set for the representative unit figure for each small area.

The beam size is blurred to the extent of the backscattering, and a representative unit figure of each small area is drawn with the irradiation amount set in.

Further, according to the present invention (claim 2), prior to irradiating a sample with an electron beam to draw a desired pattern, an optimum dose is obtained for each position in the pattern to be drawn. In the electron beam writing method for writing each pattern, the following method is used.

The drawing area is divided into small areas smaller than the spread of backscattering of the electron beam and larger than the smallest figure that can be drawn.

A representative figure representing a pattern to be drawn in each small area is set.

A reference pattern is created by uniformly subtracting the reference area (or the minimum area of the representative pattern set as a small area) from each of the representative patterns obtained above.

Assuming that the above-mentioned reference figure is to be drawn, the optimum dose for the representative figure obtained above is determined.

The optimum dose determined for the representative figure of each small area is determined as the optimum dose of the pattern to be drawn included in the small area.

[0026]

According to the present invention (claim 1), at the time of correction irradiation, the entire drawing area is divided into small areas Δ, and a representative unit figure is set for each small area, so that one or a small number of representative figures are set. The number of shots at the time of correction irradiation can be reduced by representing a large number of figures included in the small area Δ in the unit figure.
In addition, in the present invention, when the representative unit graphic is set, the representative unit graphic set for the small area in the entire area is uniformly determined from the reference area (or the minimum of the representative unit graphic set for the small area). Area) is subtracted, and the correction irradiation is performed, thereby improving the energy storage contrast in a place where the pattern density is large as compared with the correction irradiation as in the conventional ghost method, and improving the pattern resolution during development. It can be improved. Further, the number of shots at the time of the correction irradiation can be further reduced by the above subtraction, the time required for the correction irradiation can be shortened, and the throughput can be improved.

First, the reason why it is possible to improve the resolution of a pattern at the time of development by performing the correction irradiation as described above is as follows: isolated space, line and space (1: 1), drawing of isolated line. This will be described with an example of a pattern.

The correction of the proximity effect is to make the amount of energy stored in the resist accompanying the irradiation of the electron beam as uniform as possible for all the drawing patterns. In the ghost method, an inverted pattern of a desired pattern to be drawn as a pattern for correction irradiation is used. In the representative figure method, the whole drawing area is divided into small areas at the time of correction irradiation, and a rectangular figure having a total area of the entire pattern to be drawn existing in the small area is used.

FIG. 9 shows a relationship between a drawing pattern or a corrected irradiation pattern and an energy absorption distribution when the pattern is drawn. (A) is a drawing pattern, (b) is an energy absorption distribution when drawing (a), (c) is a corrected irradiation pattern by the conventional method, (d) is an energy absorption distribution when drawing (c), (E) shows the corrected irradiation pattern according to the present invention, and (f) shows the energy absorption distribution when drawing (e).

As shown in FIG. 9D, the contrast of the energy storage amount in the conventional correction irradiation is equal to that of an isolated space where the influence of backscattered electrons from the surroundings is greatest. Here, the contrast of the energy storage amount indicates B / B ′ in FIG. 9D.

In the case of the conventional method, in the case of an isolated space, the development proceeds from the beginning in a direction in which the contrast of the amount of energy storage required during development is further deteriorated.
It becomes difficult to form a fine pattern.

On the other hand, as shown in FIG. 9 (f), the contrast of the energy storage amount in the correction irradiation according to the present invention is the contrast (C / C) in the isolated space portion.
C ′) is the contrast obtained in the step of irradiating the electron beam onto the sample to draw a desired pattern (A in FIG. 9B).
/ A '). Further, the line and space (1: 1) and the energy storage amount contrast of the isolated line are also different from the contrast (C
/ C ').

The contrast of energy storage (A /
A ', B / B', C / C ') can be calculated as follows. The energy storage amount (A, B, C in FIG. 9) of the pattern to be drawn is expressed as the sum of the energy absorption amount obtained in the step of drawing the desired pattern with the electron beam and the step of performing the correction irradiation.

E = E ₁ + E ₂ (3) where E is the amount of stored energy, E ₁ is the amount of energy absorbed in the first step of drawing a desired pattern, and E ₂ is the second step of correcting irradiation. Is the amount of energy absorbed by The energy storage amount E 'at the baseline in the resist after the correction irradiation is expressed as follows.

E ′ = E ₁ ′ + E ₂ (4) where E ′ is the energy storage amount of the baseline,
E ₁ ′ is the amount of energy absorbed by the application of backscattered electrons in the step of drawing a desired pattern.

The energy storage contrast is E /
It can be represented as E '. The contrast of each pattern after the correction irradiation is equal to the place where the backscattered electrons make the largest contribution, in this case, the isolated space portion.

First, the energy storage amount contrast when the pattern density ρ in the small area of the isolated space portion is set to 9/10 will be calculated using the conventional method. Here, the ratio η of the energy application of the backscattered electrons to the forward scattered electrons is set to 0.7. This is the value of η when an electron beam is irradiated on the Si substrate at an acceleration voltage of 40 kV.

First, when the first step of drawing a desired pattern is drawn at a dose of ₁ , E1 can be expressed as follows.

E ₁ = 1 + ρ · η = 1 + (9/10) · 0.7 = 1.63 In the step of performing the second correction irradiation, the correction irradiation amount is (1-1).
ρ) η / (1 + η). Since the amount of energy absorbed in the correction irradiation also contributes to the backscattering due to the correction irradiation,
It looks like this:

E ₂ = (1 + η) · (1-ρ) η / (1 + η) = (1-ρ) η = 0.07 Here, the contrast ratio B / B ′ of the amount of energy stored by the conventional correction irradiation is calculated. .

B = E ₁ + E ₂ = 1.63 + 0.07 = 1.7 B ′ = E ₁ ′ + E ₂ = 0.63 + 0.07 = 0.7 B / B ′ = 2.4 Next, the contrast of the energy storage amount of the correction irradiation in the present invention. Calculate the ratio C / C '. In this case, the correction irradiation is not performed in the isolated space by setting the reference area corresponding to the amount of the correction irradiation to the isolated space.

C = E ₁ = 1.63 C ′ = E ₁ ′ = 0.63 C / C ′ = 1.63 / 0.63 = 2.6 According to the present invention, the contrast of the energy storage amount is improved by about 8% as compared with the conventional method. It becomes possible.

Further, in the present invention, when patterns having extremely different pattern densities such as spaces and isolated lines do not coexist on the same layer, the effect of improving the energy storage amount contrast becomes greater. The description is given below. It is assumed that a pattern to be drawn on the same layer has a pattern density of only about line and space (1: 1) or isolated line as shown in FIG. Note that (a) to (f) of FIG. 10 correspond to (a) to (f) of FIG.

According to the conventional method, in the correction irradiation, the irradiation of the inverted pattern or the same pattern density as that of the inverted pattern is performed, so that the energy accumulation distribution is as shown in FIG.

On the other hand, when the present invention is used, in the line-and-space portion, the amount of energy accumulated by the correction irradiation can be completely eliminated. However, in this case, the reference area to be subtracted has an area equivalent to the pattern for correcting and irradiating the line and space portion. The energy storage amount at this time is as shown in FIG. When the present invention is applied, the contrast (C / C '= 3.9) is improved by about 60% as compared with the contrast (B / B' = 2.4) of the conventional method.

Next, the reason why the time required for the correction irradiation can be reduced and the throughput can be improved when the present invention is applied will be described. A pattern shown in FIG. 11A is considered as a drawing pattern used for correction irradiation. Here, a pattern to be drawn is included in a place where the central representative figure area is small. Further, it is assumed that the pattern density in the small area of the drawing pattern is equal.

FIG. 11B shows a corrected irradiation pattern obtained by using the present invention. If the density of the pattern to be drawn is uniform and the area of the figure used for correction irradiation is the smallest in the correction region, there is no need to perform correction irradiation on the drawing pattern as shown in this example. Therefore, the time required for the correction irradiation can be reduced, and the throughput can be improved.

The correction accuracy when the present invention is applied will be described. FIG. 12A shows the relationship between the small area size and the correction error when a line and space with a pitch of 0.25 μm is used as the evaluation pattern. Also here
As a figure to be used for the correction irradiation, a representative figure having a pattern density ρ in a small area of １ ／ is used, and a result having an area of 25%, 50%, 75%, and 100% of the representative figure as a reference area is shown. Show.

It can be seen that the correction error when the size of the small area is 2Δ = σb and the reference area is 100% of the representative figure area is within 1.5%. here,
σb represents the distance from the beam irradiation position at which the amount of backscattered electrons becomes 1 / e.

FIG. 12B shows the result of the correction accuracy when the density of the drawing pattern changes. A line and space of 0.25 μm was used as an evaluation pattern, and the ratio of line to space was 1: 7, 1: 2, and 1:
1, 2: 1, 7: 1. The reference area was equal to the representative figure used for correction irradiation. Thus, even when the pattern density is high, that is, even when the correction irradiation area is small, the correction error can be kept within 1.5% by setting the small region size to about 2Δ = σb.

As described above, even when the correction irradiation according to the present invention is performed, the proximity effect correction can be performed with the accuracy within the correction error range, the energy storage amount contrast is improved, and the pattern resolution is improved. Performance can be improved. Further, the number of correction irradiations can be reduced, the correction irradiation time can be shortened, and the throughput can be improved.

According to the present invention (claim 2), the irradiation amount is calculated for the representative figure set for each small area, and the result is set in the original LSI pattern.
The time required for calculating the optimum irradiation amount can be reduced. In addition, when calculating the dose, the representative dose equal to or smaller than the reference area is calculated by subtracting the reference area uniformly from each unit figure set for the small area in the entire area. Sometimes, it can be deleted from the number of figures to be referred to, which can also reduce the calculation time required for setting the dose.

First, the reason why the above-described subtraction processing can reduce the calculation time required for setting the optimum dose will be described below. Here, a method using an approximation formula of the corrected dose which has been recently proposed as an algorithm used for dose correction (JMParkovich, J. Vacsci.Techn.
d B4 p.195 (1986), T.Abe, K.Koyama, R.Yoshikawa and T.T
akigawa, Proceeding of 1st Micro Process Congerence
(1988) p40) and the representative figure method (Japanese Patent Application No. 3-732).

The algorithm of the correction calculation is as shown in FIG. First, in step i, data for storing the result of the correction calculation is created using the pattern to be drawn. At this time, the figure is divided by the size of the small area, and if two or more figures exist in one small area, one representative figure is set in this small area instead of these.

Next, in step ii, reference rectangle data used for correction calculation is created. At this time, the figure is divided into rectangles as large as possible irrespective of the size of σ, and if there are two or more figures in a small area, a representative figure having a total area of these figures is set to 1 Set the number.

Next, in step iii, the correction dose for the small figure and the representative figure in the small area is determined.
At this time, the optimum dose is determined as follows.

From the reference rectangles in the data, FIG.
As shown in the figure, a rectangle existing within 3σ around the small figure i is extracted. Here, the hatched portions (2 to 7) in the figure are figures used when correcting the small figure i, and the other figures (1) are not used.

The center position of the reference rectangle selected by (xi, yi) in the small figure i or the representative figure in the small area is (xj, xj), and the size is (2Aj, 2B).
As j), the _{Di Di = D 0 / (1} + ηVi) ... (5) Vi = Σj {erf (Xj-xi + Aj / σ) -erf (Xj-xi-Aj / σ) × {erf (Yj-yi + Bj / σ) −erf (Yj−yi−Bj / σ) (6), which is defined as the irradiation amount of the small figure i in the small area. Here, Pubkovich's equation is used as a correction formula, η is a parameter for correction, Si substrate, acceleration voltage 40 k
In the case of V, it is set to 0.7. D ₀ is a reference irradiation amount.

Then, the irradiation time is set to the original pattern in step iv using the output result of step iii.

When the above-described correction calculation algorithm is used, the calculation time required for setting the dose depends on the number of reference figures. In the above example, there can be a maximum of the number of reference graphics equal to the number of small areas in the drawing pattern. However, the contribution of the proximity effect from the peripheral pattern in the correction target graphic largely depends on the distance between the correction target graphic and the reference graphic and the size of the reference graphic.

In the above example, a small reference figure, that is, a figure with a small contribution of the proximity effect, is to be corrected regardless of the size of the reference figure. However, if the reference area is subtracted from the area of the reference figure, Reference figures smaller than the reference area can be excluded from correction calculation targets.

FIG. 15A shows an example of a reference figure when the conventional representative figure method is used, and FIG. 15B shows an example of a reference figure when the subtraction processing according to the present invention is applied. In this example, in the representative figure method of (a), the number of reference figures is nine, whereas in (b) after the subtraction processing according to the present invention, the number of reference figures is three. The time required for the calculation can be reduced to about 1/3. In this way, the time required for the optimal dose calculation can be shortened by using a value obtained by uniformly subtracting the reference area from each of the representative unit figures set for the small areas in the entire area as a reference figure used in the correction calculation. It becomes possible.

In the present invention, the ratio of the energy absorption amount of backscattered electrons from the underlying substrate to the forward scattered electrons in the resist in the resist is η, the maximum pattern density among the patterns to be drawn is ρ, and the subtraction processing is performed. When the pattern density of the reference area used for ρ is ρ ′, the parameter η ′ is set as (1 + ρη) / (1 + ρ′η) = 1 + (ρ−ρ ′) η ′ (7) and correction calculation is performed. Is used instead of η. In this case, the reason why the proximity effect can be corrected substantially in the same manner as in the representative graphic method will be described below.

In the dose correction, the ratio of the optimum dose to each pattern having different densities is represented by the reciprocal of the amount of energy absorption in the resist of each pattern. The energy absorption E ₁ of a certain pattern can be easily set to 1
In this case, E ₁ = 1 + ρη (8) Here, ρ is the pattern density, and η is the ratio of the energy absorption of the backscattered electrons to the forward scattered electrons. The optimum irradiation amount at this time is represented by the reciprocal 1 / E of the energy absorption amount.

On the other hand, according to the present invention, when the reference area is ρ ′, the influence of the proximity effect corresponding to the reference area is removed by subtraction processing. Therefore, the apparent energy absorption amount E 'of each pattern in the present invention.
Is represented by the following equation.

E ′ = 1 + (ρ−ρ ′) η (9) In this case, the optimal dose is set to 1 / E ′.

As an example, let us consider a drawing pattern having different densities as shown in FIG. FIG. 16B shows the amount of energy absorption when drawing is performed with the same irradiation amount. Here, when the maximum pattern density is ρa and the minimum pattern density is ρc, the proximity effect correction by the representative figure method (FIG. 16)
In (c)), the optimum doses Da and Dc are respectively expressed as follows.

Da = 1 / Ea = 1 / (1 + ρa η) (10) Dc = 1 / Ec = 1 / (1 + ρc η) (11) On the other hand, the present invention was applied to the pattern shown in FIG. In the case, the pattern density ρ ′ of the reference area is
If it is equal to c, the optimal doses Da 'and Dc' are as follows.

Ρ ′ = ρc (12) Da ′ = 1 / Ea ′ = 1 / {1+ (ρa−ρc) η ′} (13) Dc ′ = 1 / Ec ′ = 1 / ｛1+ (ρc− ρc) η ′｝ (14) In the above equation, η ′ indicates a parameter used for correction calculation when the present invention is applied. At this time, the ratio (Da '/ Dc') of the optimum dose due to the pattern density is different from that (Da / Dc) in the conventional dose correction method because the influence of the proximity effect for the reference area is not taken into account. have. However, by setting the correction parameter η ′ so as to satisfy the relationship of Da ′ / Dc ′ = Da / Dc, it is possible to set the optimum irradiation amount due to the pattern density within the error range in the same manner as in the representative graphic method. Can be. Η and η 'in this case
Is shown below.

Dc / Da = (1 / Ec) / (1 / Ea) = {1 / (1 + ρc η)} / {1 / (1 + ρa η)} = (1 + ρa η) / (1 + ρc η) (15) Dc ′ / Da ′ = (1 / Ec ′) / (1 / Ea ′) = [1 / {1+ (ρc−ρc) η ′}] / [1 / {1+ (ρa−ρc) η ′}} = 1 + (ρa−ρc) η ′ (16) Here, η, η ′ when Da / Dc = Da ′ / Dc ′
Is as follows.

(1 + ρa η) / (1 + ρc η) = 1 + (ρa−ρc) η ′ (17) In the above equation, ρa and ρc are values obtained from the drawing pattern, and η is the acceleration voltage and the type of the base substrate. Is a value uniquely obtained by That is, η 'can be easily obtained by substituting these values into the above equation. By using the parameter η ′ obtained as described above instead of the correction parameter η in the representative graphic method in the correction calculation in the present invention, it is possible to set the optimum irradiation amount due to the density of the pattern in the same manner as the conventional method. It becomes possible.

As an example, ρc = 1/5, ρa = 9/1
0, the case where the pattern density ρ '= ρc of the reference area is considered. In this case, the maximum ratio due to the density of the pattern according to the conventional method is as follows from Expression (15).

Dc / Da = {1 / (1 + η / 5)} / {1 / (1 + 9η / 10)} = 1.43 Here, η used a value of 0.7. This is the ratio of forward scatter to back scatter when an electron beam is incident on a Si substrate at an acceleration voltage of 40 kV.

Η 'in the case of applying the above example and the present invention is obtained as 0.614 from the above equation (17). In the present invention, the ratio of the maximum and minimum irradiation doses due to the density of the pattern is as follows from the above equation (16).

Da ′ / Dc ′ = [1 / {1+ (1 / 5-1 / 5) η ′}] / [1 / {1+ (9 / 10-1 / 5) η ′}] = 1 .43 Further, the conventional method and the present invention will be compared with respect to a pattern density having an intermediate value between ρa and ρc. Where ρb = 1/2
When the irradiation amount of the pattern having the pattern density ρb in the conventional method is set to 1 when the irradiation amount of the pattern having the pattern density of ρa is 1, Db / Dc = (1 + η / 2) as in the above equation (15). (1 + 9η / 10) = 1.18 In the case of the present invention, Db ′ / Dc ′ = 1 + (1/2/5) η ′ = 1.18 in the same manner as in the above equation (15). '= 0.614 was used.

As described above, by using the present invention, not only the time required for the correction calculation can be shortened, but also the correction effect can be obtained by using the parameter η 'as in the conventional method.

[0077]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, an embodiment of the present invention (claim 1) will be described below.

FIG. 1 is a schematic diagram showing an electron beam writing apparatus used in the method of one embodiment of the present invention. In the figure, 10 is a sample chamber, 11 is a target (sample), 12 is a sample stage, 2
0 denotes an electron optical column, 21 denotes an electron gun, 22a to 22e denote various lens systems, 23 to 26 denote various deflection systems, 27a denotes a blanking plate, and 27b and 27c denote beam forming aperture masks. Reference numeral 31 denotes a sample stage drive circuit, 3
2 is a laser measurement system, 33 is a deflection control circuit, 34 is a variable shaping beam size control circuit, 35 is a blanking control circuit, 36 is a buffer memory and control circuit, 37 is a control computer, and 38 is data conversion. The computer 39 indicates a CAD system.

The electron beam emitted from the electron gun 21 is turned on and off by the blanking deflector 23. By adjusting the irradiation time at this time, the present apparatus can change the irradiation amount according to the irradiation position.
The beam that has passed through the blanking plate 27a is supplied to the beam forming deflector 24 and the beam forming aperture mask 27.
The beam is formed into a rectangular beam by b and 27c, and the size of the rectangle is changed. The shaped beam is deflected and scanned on the target 11 by the scanning deflectors 25 and 26, and the target 11 is drawn in a desired pattern by the beam scanning. The standard acceleration voltage of the electron beam in this apparatus is 40 KV, and the maximum size of the variable shaped beam that can be generated is 2 μm in height,
It is a rectangle having a width of 2 μm.

Next, an electron beam writing method using the above-described apparatus, particularly a method for correcting the proximity effect, will be described with reference to FIGS.

First, the acceleration voltage of the apparatus was set to a standard value of 40 KV. At this time, as described above, the small area may be 2 μm × 2 μm. Since the size of the small region is equal to the maximum beam size that can be set by the system, here, the irradiation amount Q at the time of correction is the same for all representative unit figures. Q = Qc = Qp × η _E / (1 + η _E ) (18) And control only the beam size.
Here, Qc is the standard correction dose for the original figure, and Qp is the dose in the pattern drawing process.

FIG. 2 shows an image of a small area occupied in a wafer (or a mask). FIG. 2A shows a state in which a plurality of chip areas 41 are arranged on a wafer 40, and FIG.
Indicates a state in which a plurality of small areas 42 are arranged in the chip area 41. FIG. 2C shows the four small areas in an enlarged manner, and the hatched portions in this figure are the desired patterns 43 to be drawn.

In this case, if the ghost method is used, the pattern to be corrected and radiated is as shown by the hatched portion in FIG. 3A, and the number of patterns to be corrected and radiated is 8 in the upper left small area. Further, when the conventional representative figure method is used, if a representative unit figure is set for each small area, the pattern to be corrected and irradiated becomes one figure in each small area as shown in FIG. The unit figure representing the small area at the upper left of FIG.
It is a square of × L μm, and its size is determined as L ² = (１ ／ × 2) × (S _{1 to} S ₈ ) (19). Here, S _{1 to} S ₈ are the total areas of the figures in the small region at the upper left of FIG.

FIG. 3C shows a pattern to be subjected to correction irradiation when the method of this embodiment is used. here,
The figure area at the top of FIG. 3B was used as a reference area. In this case, it is sufficient to draw only the hatched portion in FIG. 3C. For example, in the ghost method, it is necessary to irradiate eight times in the small region at the upper left and once even when the representative figure is used. The example method eliminates the need for irradiation at all.

FIG. 3 (d) shows the dose distribution on the wafer for one of the representative unit figures formed as shown in FIG. 3 (b) or (c). The aperture mask 27
The beam is formed as shown in FIG. 3C using b and 27c. However, since the beam is blurred on the wafer surface, FIG.
become that way.

Next, the method of the above embodiment will be described in more detail.

<Embodiment 1> First, two types of EB data are created offline. One is EB data (DATA-1) corresponding to a desired pattern, which corresponds to FIG. 2C, and the other is EB data (DATA-
2) and corresponds to FIG. These are stored in the disk of the EB device, and first, the pattern of DATA-1 is drawn while the beam is focused on the wafer surface. Here, the irradiation amount was 50 μc / cm ² .

Next, correction irradiation for proximity effect correction is performed, and various controls in this correction irradiation are performed in a portion surrounded by a broken line in FIG. FIG. 4 shows a specific configuration of this part. Beam diameter dc [12.2μ on wafer surface
m]. Then, send the DATA-1 from the control computer in the buffer memory M _1, the length of 2.0μm of the small area of one side and stored in the register R _1, R _2, to start the circuit A _1, A _2.

In this state, the irradiation amount is changed to the standard correction irradiation amount Qc.
Is drawn at 16.6 μc / cm ² . At this time, the circuits A ₁ and A ₂ perform the following operations.

The drawing area is divided into rectangles of 2 μm × 2 μm. (Circuit A ₁ ) All figures are allocated to each small area (Circuit A ₁ ). At this time, if a certain figure extends over two or more areas, this figure is divided and the divided figure is allocated to small areas.

A representative form is set for each small area as follows (circuit A ₂ ).

(A) The areas of figures included in each small area are summed up.

(A) The reference area S 'is subtracted from the value of (A).

(C) Take the square root of the value of (a).

(D) The value of (c) is changed to the value R (2.
Divide by 0).

(E) The representative unit figure is a square having the value of (d).

When the drawing and the correction irradiation were performed as described above, the correction accuracy of the proximity effect substantially coincided with the conventional method within the measurement error.

As described above, in the method of this embodiment, one representative unit figure is set in each small area, and the size S of the representative unit figure is set.
Is the area of the original unit figure i contained in the small area as S
i, the reference area is set to S ', and S = {Si-S' (20) is set, and the irradiation amount Q of the representative unit figure is set to be the same as the standard correction irradiation amount Qc to perform correction irradiation. The proximity effect can be corrected similarly to the ghost method.

In this case, only one representative figure for correction irradiation is required for each small area. Further, when the total figure area included in the small area is equal to or smaller than the reference area S ', no correction irradiation is performed, so that the time required for correction irradiation is further shortened, and the throughput can be improved.
Further, as described above, the contrast of the energy storage amount in the resist is improved, and the resolution of the fine pattern is improved.

In the above-described embodiment, the correction irradiation is performed by adjusting only the area of each representative unit figure. However, the following embodiments 2 to 8 may be performed. is there.

[0102] <Example 2> Example 1 of the function of the circuit A, and the following functions of (a) to (c), the register R ₂ '[area of Qc / small area] = 16.6 / 4 = 4.1 is set.

(A) ′ The areas of figures included in each small area are summed.

(B) The value of the reference area S 'is subtracted from the value of (a)'.

(C) Multiply the value of (a) 'by the value of the register R ₂ '.

(D) The representative unit figure is a square having the value (2) of the register R as shown in FIG. 5, and the irradiation amount is (c) '.

As described above, by making the size of the representative unit figure the same as the size of the small area, and setting the irradiation amount according to the total area of the drawing figure included in each small area,
The correction of the proximity effect can be performed as in the first embodiment.

<Embodiment 3> In determining the representative unit figure performed by the circuit A in the first and second embodiments, the minimum area Smin of the representative figure in the entire correction area may be used instead of the reference area S '. Good. In this case, the functions (a) and (b) of the circuit A performed in the first embodiment and the functions of the circuit A performed in the second embodiment
(A) '(a)' is as follows.

(A) The areas of figures included in each small area are summed up.

(A) If the value of (a) is smaller than the value of the register R ₃ already input, the value of (a) is substituted for the register R ₃ .

(C) The value of (a) is input to the memory M.
Here, for each small area of the total correction region, when the above process is completed, the value of R ₃ is the smallest representative figure area Smin
Is entered. Further, the total figure area ΣSi in each small area is input to the memory M.

(D) The total area ΣSi in each small region input to the memory M is sequentially read, and the value of Smin input to the register R ₃ is reduced.

Of course, in the first and second embodiments, the minimum representative figure area may be input in advance as the reference area.

<Embodiment 4> The determination of the representative unit figure performed by the circuit A in the first, second, and third embodiments is performed by a control computer or a computer independent of the system, and each representative unit figure is a unit figure to be drawn ( EB data may be created.

In determining the representative unit figure, the input does not necessarily have to be subdivided as shown in FIG. 3A, and the pattern is inverted by a CAD system as shown in FIG. The figure before the subdivision described in the case may be used.

Embodiment 5 The electron optical system was adjusted by the apparatus shown in FIG. 1 and the acceleration voltage was set to 70 kV. At this time, σb ~ 1
0 μm, ηE〜0.5. The maximum beam size Smax obtained is 1.4 μm, and the maximum irradiation dose Dmax of the apparatus is 70 μc / cm ² . Then, the standard correction dose Qc for setting the patterning dose to 50 μc / cm ²
Is 16.6 μc / cm ² .

On the other hand, the backscattering spread of 70 kV is σ
b = 10 μm, 3σb = 30 μm, and the size of the small region may be up to 3 μm. Although the small area size of 1.4 μm may be used and the methods of Embodiments 1 to 4 may be used as they are, the correction irradiation is performed at a higher speed by performing the following.
Each small area has a maximum size of 1.4 μm.
The representative unit figures up to are selected as follows.

(A) The areas Si of the figure i included in each small area are summed (ΣSi).

(A) The reference area S '(or the minimum total figure area Smin) is subtracted from the value of (A) to obtain S.

(C) The value y of the following equation is obtained.

Y = [value of (a) × (16.6) μc / cm ² ] (d) The number N of representative unit figures is determined by the following equation.

N = [y / ｛(70 μc / cm ² ) × (1.4 μm
m ² )} The smallest natural number not less than (e) A representative unit figure is determined as follows.

The dose Q is a standard dose Qc.

The representative unit figure is a square, and the length of one side is (y / N) ^1/2 .

The center of the representative unit figure coincides with the center of the small area.

As described above, ΣSi-S '(or ΣS
(i-Smin), the number of representative figures in each small area can be reduced.

As described above, in this embodiment, when the maximum beam size that can be set by the drawing system is Smax and the maximum irradiation amount is Dmax, the number N of representative unit figures set in the small area is represented by (S × Qc) / (Smax × Dmax) or {({Si−S ′) × Qc} / (Smax × Dmax)... Since the minimum natural number not less than the value of (21) is set, the time required for correction irradiation is minimized. , And higher-speed correction irradiation can be performed.

<Embodiment 6> In Embodiment 5, the EB data is processed off-line. However, a dedicated circuit may be used as in Embodiments 1, 2, and 3. Further, in the fifth embodiment, the center of the representative unit graphic coincides with the center of the small area. However, the figure may be moved in a range where the center of gravity of the representative unit graphic is included in the small area.

When the maximum shot size ≧ the total area of the figures in the small area, one representative figure may be set in the small area, as shown in FIG. 7A. When the maximum shot size <the total area of the figures in the small area, two or more pieces are set.
At this time, two or more figures may overlap as shown in FIG. 7B, or may not overlap as shown in FIG. 7C. Instead of setting a plurality of figures, the time may be lengthened by one shot.

Next, a specific embodiment of the present invention (claim 2) will be described.

<Embodiment 7> As an algorithm of the correction calculation, a method using a recently proposed approximate formula of the correction dose (JMPavkovnh, J. Vac, Sci Technl B4 p195 (198)
6), T.Abe, K.Koyama, R.Yoshikawa and T.Takigawa Prece
edys of 1st Micro Process (Conference (1988) p40)
Can be used as an algorithm for correction calculation, and this can be applied to the present invention. The algorithm of the correction calculation is as described in the section of (action).

When this method is applied in the present invention, the method may be as shown in FIG. That is, step i in FIG.
In step a, data for storing the result of the correction calculation is created. At this time, the small area is divided by a size smaller than σb.

Next, in step b instead of step ii, a reference rectangle used for correction calculation is created. At this time, the created data has an area obtained by subtracting the reference area S 'from each figure in the reference data created in step ii.

Further, instead of step iii, step c
In, the optimal dose for the small figure and the representative figure is determined. Then, in step d instead of step iv, the dose is determined for the original pattern from the output result of step c.

The test pattern used here is ρmin = 1
/ 5, ρmax = 1, and the pattern density occupies 1/5 of 80% of the entire writing area. Here, the irradiation area setting small area size is set to 2 μm × 2 μm, the reference figure small area size is set to 10 μm × 10 μm, and the reference area S ′ is set to １ ／ which is equal to ρmin. Then, when the drawing was performed, the correction accuracy of the proximity effect was almost the same as the conventional method within the measurement error.

In this case, the correction calculation was performed by expanding the entire pattern without using the array definition of the pattern.
By using the above algorithm, an area with a pattern density of 1/5, which occupies 80% of the entire drawing area, is deleted from the reference graphic, so that the time required for the correction calculation can be greatly reduced.

<Eighth Embodiment> In the above-described seventh embodiment, the entire pattern is developed without using the array definition of the pattern. However, the calculation time can be further reduced by using the array definition. In this case, the following is performed.

The correction processing according to the present invention is performed on the subfield group at the outer peripheral portion in the drawing area defined by the cell array.

With respect to the internal subfield surrounded by the outer peripheral portion, one subfield is selected and correction processing is performed.

For those in which no correction processing is performed in the internal subfields, the correction result of the graphic in the subfield obtained in is substituted.

Here, the sub-field indicates a unit of an area when processing is performed by a computer. If the cell array size is larger than the set subfield,
The above process is performed after all the patterns once defined in the array are developed.

In Embodiments 7 and 8, the drawing pattern is used as data to be processed as a reference graphic. However, an inverted pattern of the pattern to be drawn can be used. Further, in the seventh and eighth embodiments, the position in the small area of the representative figure for setting the dose is centered, but this position may be the center of gravity of the pattern included in the small area. In the seventh and eighth embodiments, the correction calculation is performed by a computer independent of the system. However, the correction calculation may be performed off-line using a dedicated circuit.

The present invention is not limited to the above embodiments. Although the embodiment uses the variable shaped beam type electron beam drawing apparatus, the present invention can be applied to other apparatuses such as a character projection type electron beam drawing apparatus. The present invention does not limit the purpose of use of the electron beam writing apparatus. For example, the present invention can be used not only for forming a resist pattern directly on a wafer but also for forming an X-ray mask, an optical stepper mask, a reticle, and the like. In addition, various modifications can be made without departing from the scope of the present invention.

[0143]

As described in detail above, the present invention (claim 1)
According to the method described above, the entire writing area is divided into small areas (areas smaller than the backscattering spread of the electron beam, or areas having a size equal to or smaller than the half width of the backscattering spread) during the correction irradiation. At the same time, a representative unit figure is set for each small area by subtracting the reference area S 'or the minimum total figure area Smin from the total figure area in each small area, and by irradiating the representative unit figure, a correction The number of shots can be reduced and the time required for the correction irradiation can be shortened, and the throughput can be improved. In addition to this,
Since the contrast of the energy storage amount in the resist can be improved, the pattern resolution can be improved.

Further, according to the present invention (claim 2), when performing the correction calculation for the representative figure set for each small area, the total figure area in each small area and the reference area S ′ Or the figure obtained by subtracting the minimum total figure Smin is used as a reference figure used in the correction calculation, so that the time required for the correction calculation can be greatly reduced. A correction effect can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an electron beam drawing apparatus used in a method according to an embodiment of the present invention.

FIG. 2 is a diagram showing an image of a small area occupying in a wafer.

FIG. 3 is a diagram showing a pattern to be corrected for irradiation and a beam irradiation distribution on a wafer.

FIG. 4 is a block diagram specifically showing a main part of FIG. 1;

FIG. 5 is a diagram showing a representative unit graphic according to the second embodiment.

FIG. 6 is a view for explaining a fourth embodiment, showing a figure before subdivision described in a CAD system.

FIG. 7 is a diagram showing a representative basic graphic according to a sixth embodiment.

FIG. 8 is a diagram illustrating a correction algorithm according to a seventh embodiment.

FIG. 9 is a diagram showing a relationship between a drawing pattern or a correction irradiation pattern and an energy absorption distribution when the pattern is drawn.

FIG. 10 is a diagram showing a contrast improvement effect when patterns having extremely different pattern densities are not mixed on the same layer.

FIG. 11 is a diagram illustrating a shot number reduction effect (correction irradiation time reduction effect) by a subtraction process.

FIG. 12 is a diagram illustrating a relationship between correction accuracy and a small area size when a subtraction process is performed.

FIG. 13 is a diagram showing an algorithm for correction calculation in the representative graphic method.

FIG. 14 is a diagram showing a small graphic to be corrected and a reference rectangle.

FIG. 15 is a diagram showing a decrease in the number of reference graphics by performing a subtraction process.

FIG. 16 is a diagram showing an optimum irradiation amount of a pattern and an absorption energy in a resist in the irradiation amount correction.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Sample chamber 11 ... Target (sample) 12 ... Sample stand 20 ... Electron optical column 21 ... Electron gun 22a-22e ... Lens system 23-26 ... Deflection system 27a-27c ... Aperture mask 36 ... Buffer memory and control circuit 37 ... Control computer 38 ... Data conversion computer 39 ... CAD system 40 ... Wafer 41 ... Chip area 42 ... Small area 43 ... Drawing pattern

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-5-160010 (JP, A) JP-A-4-212407 (JP, A) JP-A-2-130815 (JP, A) JP-A-6-130815 5502 (JP, A) JP-A-5-343305 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H01L 21/027 G03F 7/20

Claims (8)

(57) [Claims] 1. A step of irradiating a sample with an electron beam to draw a desired pattern, and before or after this step, the electron beam is irradiated on the sample to reduce a proximity effect due to backscattered electrons accompanying the pattern drawing. The correction irradiation step, wherein, as the correction irradiation step, the entire correction area is divided into small areas smaller than the spread of the backscattering of the electron beam and larger than the smallest figure that can be drawn, As a unit figure for irradiating each small area, the total area (ΣS
The area S obtained by uniformly subtracting the reference area S ' from i) is obtained, and the maximum beam size that can be set by the writing apparatus is Smax,
When the irradiation amount is Dmax and the correction irradiation amount is Qc,
In the area, N = (S × Qc) / (Smax × Dmax)
N representative unit figures having a minimum natural number equal to or greater than a certain value are set, the beam size is blurred to the extent of backscattering, and the representative unit figures of each small area are drawn at the set irradiation amount. An electron beam drawing method characterized by the above-mentioned. 2. A step of irradiating an electron beam on a sample to draw a desired pattern, and before or after this step, an electron beam is irradiated on the sample to reduce a proximity effect due to backscattered electrons accompanying the pattern writing. A step of performing a correction irradiation step, wherein the correction irradiation step includes setting the entire correction area to be smaller than the spread of the backscattering of the electron beam.
Much larger than the minimum figure drawable, and set in the drawing device
The area divided into small areas smaller than the maximum beam size Smax that can be radiated, and the reference area is uniformly subtracted from the total area of the pattern obtained by inverting the desired pattern in the small area between black and white as a unit figure for irradiating each small area. Is set, a uniform irradiation amount is set for the representative unit figure of each small area, the beam size is blurred to the extent of backscattering, and the representative unit figure of each small area is An electron beam drawing method characterized by drawing at a set irradiation dose. 3. A step of irradiating a sample with an electron beam to draw a desired pattern, and before or after this step, the electron beam is irradiated on the sample to reduce a proximity effect due to backscattered electrons accompanying the pattern drawing. A step of performing a correction irradiation step, wherein the correction irradiation step includes setting the entire correction area to be smaller than the spread of the backscattering of the electron beam.
Much larger than the minimum figure drawable, and set in the drawing device
Is divided into small areas smaller than the maximum possible beam size Smax , and the same large area as the small area is used as a unit figure for irradiating each small area.
Of the desired pattern in the small area is inverted
Surface obtained by uniformly subtracting reference area S 'from total area (面積 Si)
The product S is calculated, the amount of illumination of (S × Qc) / Smax is set for the representative unit figure of each small area, and the beam size is blurred to the extent of backscattering, and the An electron beam drawing method, wherein a representative unit figure is drawn at the set dose. 4. Prior to irradiating a sample with an electron beam to draw a desired pattern, an optimum dose is determined for each position in the pattern to be drawn, and an electron beam for writing each pattern at the optimum dose. In the beam writing method, the writing area is divided into small areas smaller than the spread of backscattering of the electron beam and larger than the minimum figure that can be drawn, and a representative figure representing a pattern to be drawn inside the small area. Is set, a rectangular reference pattern having an area obtained by uniformly subtracting the reference area from the area of the representative figure existing in the small area is set, and it is assumed that the reference figure of each small area is drawn. Determine the optimal dose for the figure, and determine the optimal dose for the representative figure in each small area,
An electron beam writing method characterized by determining an optimum dose of a pattern to be written included in the small area. 5. An electron beam drawing apparatus for irradiating an electron beam on a sample to draw a desired pattern and correcting and irradiating the sample with the electron beam in order to reduce a proximity effect due to backscattered electrons accompanying the pattern drawing. Means for dividing the entire correction area into smaller areas smaller than the spread of backscattering of the electron beam and larger than the smallest figure that can be drawn, in order to irradiate the corrected electron beam; As a figure, the total area of the pattern (の S
obtaining the area S obtained by subtracting the reference area S 'uniformly from i)
Means, the maximum beam size that can be set is Smax, and the maximum irradiation amount is Dm
ax, the correction dose is Qc, and N =
(S × Qc) / (Smax × Dmax)
Means for setting N representative unit figures which are small natural numbers; and means for blurring the beam size to the extent of backscattering and drawing the representative unit figures of each small area at the set dose. An electron beam writing apparatus characterized by comprising: 6. An electron beam writing apparatus for irradiating a sample with an electron beam to draw a desired pattern and correcting and irradiating the sample with the electron beam to reduce a proximity effect due to backscattered electrons accompanying the pattern writing. In order to correct and irradiate the electron beam, the entire correction area is smaller than the backscattering spread of the electron beam.
Much larger than the minimum figure drawable, and set in the drawing device
Means for dividing into small areas smaller than the maximum possible beam size Smax, and as a unit figure for irradiating each small area, a reference area is uniformly subtracted from a total area of a pattern obtained by inverting a desired pattern in the small area between black and white. Means for setting a rectangular representative unit figure having an area, means for setting a uniform irradiation amount for the representative unit figure of each small area, and blurring the beam size to the extent of backscattering, Means for drawing a representative unit figure at the set irradiation dose. 7. An electron beam drawing apparatus for irradiating a sample with an electron beam to draw a desired pattern and correcting and irradiating the sample with the electron beam to reduce a proximity effect due to backscattered electrons accompanying the pattern drawing. In order to correct and irradiate the electron beam, the entire correction area is smaller than the backscattering spread of the electron beam.
Much larger than the minimum figure drawable, and set in the drawing device
Means for dividing the maximum beam size Smax following small region may, as a unit figure to be irradiated to each small area, the same large as the small region
Kisano means for setting a representative unit figures, the desired pattern is a black and white reversal pattern in the small area if
Surface obtained by uniformly subtracting reference area S 'from total area (面積 Si)
Means for calculating the product S, means for setting an irradiation amount of (S × Qc) / Smax for the representative unit figure of each small area, and blurring the beam size to the extent of backscattering, Means for drawing a representative unit figure of each small area at the set irradiation dose. 8. An apparatus for irradiating a sample with an electron beam to draw a desired pattern, wherein an optimum dose is determined for each position in the pattern to be drawn, and each pattern is drawn with the optimum dose. Means for dividing a writing area into small areas smaller than the spread of backscattering of the electron beam and larger than the smallest figure that can be written, and a pattern to be drawn inside the small area. Means for setting a representative figure, means for setting a rectangular reference figure having an area obtained by uniformly subtracting a reference area from the area of a representative figure existing in a small area, and drawing a reference figure for each small area Means for determining the optimal dose for each representative figure, and the optimal dose determined for the representative figure in each small area,
Means for determining an optimal dose of a pattern to be written, which is included in the small area, the electron beam writing apparatus.