JP4992930B2 - Method for generating backscattering intensity based on underlying structure in charged particle beam exposure and method for manufacturing semiconductor device using the method - Google Patents

Description

  The present invention relates to a method for generating backscattering intensity based on a lower layer structure in charged particle beam exposure and a method for manufacturing a semiconductor device using the method, and in particular, considers the influence of charged particles on the surrounding in a lower layer structure having a plurality of layers. Thus, a method for obtaining a more accurate backscattering intensity is provided, and a method for manufacturing a semiconductor device that enables more accurate proximity effect correction is provided.

  In recent years, as the degree of integration of semiconductor devices has been improved, the required pattern size has been miniaturized, and the resolution by conventional exposure methods using light is insufficient, making it difficult to form fine patterns. Therefore, an exposure method using a charged particle beam, particularly an electron beam (electron beam) has been used. The electron beam exposure method includes a point beam exposure method with high resolution but low throughput, a variable shaping exposure method that improves the throughput by exposing the pattern in small rectangular units, and a pattern that repeatedly appears in the chip. There are a partial batch exposure method (block exposure method) in which batch transfer is performed using a stencil mask, a projection type exposure method in which masks of all patterns are created, and a large area is collectively transferred, as in light exposure. In the partial batch exposure method and the projection type exposure method, the number of electron beam shots can be reduced, so that the throughput can be improved, but the exposure amount cannot be changed according to the exposure pattern, and the exposure accuracy can be improved. It tends to cause a decline.

  One problem common to such charged particle beam exposure methods (hereinafter simply referred to as electron beam exposure methods) is that the resist dimensions vary due to proximity effects. Electrons that have passed through the resist are scattered by the substance constituting the substrate and returned to the resist, and the resist is re-exposed. Since the amount of electrons returning to the resist is proportional to the density of the pattern, the variation width of the resist dimension due to the proximity effect varies depending on the pattern layout.

  As a method for correcting this proximity effect, the influence of electrons returning from the substrate in each region is estimated based on the energy distribution (EID: Exposure Intensity Distribution) function given to the resist by electrons incident from one point, and accordingly, It has been proposed to optimize the exposure amount in the region or change the pattern dimension. For example, it is described in Patent Documents 1 and 2 described later.

  It is generally known that this EID function is represented by the sum of two Gaussian distributions as shown in the following equation when the substrate is made of one kind of substance.

  Here, βf: forward scattering length, βb: backscattering length, η: forward / backscattering ratio, 1 item is energy given to resist by incident electrons, 2 item is energy given to resist by electrons reflected from substrate These are called forward scattering terms and backscattering terms, respectively.

In the case of performing electron beam exposure in the process of manufacturing a semiconductor integrated circuit (LSI), structures such as wirings and contact plugs are already formed under the resist, and the substrate is composed of a plurality of substances. Since the parameter of the backscattering term differs depending on the material constituting the lower layer, the influence on the exposure intensity distribution cannot be estimated with a simple Gaussian distribution function as shown in the above formula (1). In order to solve this problem, a technique for correcting the proximity effect in consideration of the structure one layer below the resist has been proposed (for example, Non-Patent Document 1 described later). In this method, for example, in the case where the lower layer is composed of a tungsten (W) contact plug and a silicon oxide film (SiO ₂ ) filling the space between the W plugs, exposure is performed when the backscattering intensity is obtained by the area density method. The influence of backscattered electrons returning from the region where the pattern density is α and the density of the lower layer W is α _w is obtained by the following equation.

η _w and η _SiO2 are forward / backward scattering ratios obtained when the entire surface is W or SiO ₂ .

However, since the material composing the wiring or contact hole is usually as thin as 1 μm or less (μ is expressed by u for convenience), the electrons further reach the lower layer. When electrons that have passed through one layer return to the resist, the electrons are affected by wiring and contact holes in the process, and therefore the backscattering intensity cannot be obtained simply by the above equation (2). For example, if the lower layer of the Al layer to be patterned after exposure is composed of a W contact hole and SiO ₂ filling the space between them, W easily scatters electrons, so there are few electrons that escape to the lower layer. , The electrons incident on SiO ₂ enter relatively deeply. However, since electrons that have entered deeply into SiO ₂ are scattered in the W contact hole in the process of returning to the resist, the number of electrons that are back-scattered and returned is reduced compared to the case where the lower layer is entirely SiO ₂ . Further, since the wiring structure of the LSI overlaps rather than two layers, the electrons that have passed through the lower layer are also affected by the lower layer and the lower layer.

  Thus, it has been proposed to obtain the backscattering intensity in consideration of the multilayer structure below the resist layer (for example, Non-Patent Document 2 described later). In the method described in this prior document, the region is classified according to the lower layer structure in order to consider the lower layer multiple layer structure. For example, a region where W is not present in a lower layer, a region where W is present only under one layer, a region where W is present only under two layers, a region where W is present under both one layer and two layers, and the like. Thereafter, the backscattering calculation by the area density method is performed using the forward / backscattering ratio and the backscattering length which are different for each region.

  Furthermore, it has also been proposed to add subexposure so that the backscattering intensity is uniform in consideration of a plurality of wiring layer structures existing below the resist layer (for example, Patent Document 3 described later). In this prior document, the subexposure is generated according to the pattern density of each of the plurality of lower wiring layers so that the backscattering intensity is uniform without depending on the lower wiring layer structure.

U.S. Pat. No. 6,610,990 B1 (issued Aug. 26, 2003) corresponding to JP 2001-52999 A (published on Feb. 23, 2001) JP 2002-313693 A (published on October 25, 2002) Corresponding US Publication Number 2002-0177056 (published on November 28, 2002) Patent No. 2950280 (issued on September 20, 1999) Corresponding US Patent No. 6243487 B1 (issued on June 5, 2001)

J. Vac. Sci. Technol. B, Vol. 10, No. 6, page3072-3076 (1992) Emerging Lithographic Technologies VII, Roxann L, Engelstd, Editor, Proceedings of SPIE Vol. 5037 (2003)

  However, although the above-mentioned prior art documents deal with the problem of backscattering in consideration of the multi-layer structure of the lower layer, all of them are only simple methods. In particular, the influence on adjacent regions due to electron scattering in a plurality of layers under the resist layer is not taken into consideration. Therefore, it is not possible to obtain an accurate backscattering intensity considering the influence of the multi-layer structure.

  Accordingly, the present invention provides a method capable of accurately obtaining the backscattering intensity in consideration of the influence of a complex multilayer structure under the resist layer, and an exposure method and a semiconductor device that perform proximity effect correction using the method. It is in providing the manufacturing method of.

  Furthermore, the present invention provides a method capable of accurately obtaining a backscattering intensity capable of reducing the calculation time in consideration of the influence of a complex multilayer structure under the resist layer, and a proximity effect correction using the method. It is an object to provide an exposure method and a method for manufacturing a semiconductor device.

In order to achieve the above object, a first aspect of the present invention provides a method in which a charged particle beam is applied to a resist layer formed on a plurality of layers each including a single material or a pattern of a plurality of materials. In the method for generating the backscattering intensity of the charged particles to the resist layer,
A reflection coefficient rn corresponding to the number of particles reflected by the n-th layer from the n-1th layer with respect to the n-th layer from the resist layer among the plurality of layers, and the n-th layer The nth layer includes a transmission coefficient tn corresponding to the number of particles through which the charged particles that have passed through the nth layer pass, and a diffusion distribution in which the charged particles scatter within the nth layer. Given for each substance
The generation method is:
The method includes a first step of generating the backscattering intensity using the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution.

  According to a more preferred embodiment of the above aspect of the invention, in the first step, the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution are used for each of the plurality of layers. A charged particle scattering intensity that causes the backscattering intensity is generated.

  Furthermore, according to a more preferred embodiment of the above aspect of the invention, in the first step, charged particles from a surrounding region including the region of interest with respect to a predetermined region of interest of the nth layer. The scattering intensity is divided into areas according to the area density αn of the substance in the n-th layer and the distance coefficient an corresponding to the distance between the surrounding region and the region of interest and obtained from the scattering distribution. Features.

  Furthermore, according to a more preferred embodiment of the above aspect of the present invention, the charged particle scattering intensity in the first step is (1) the first charged particle intensity transmitted through the n−1th layer is Lower transmitted charged particle intensity obtained by multiplying the transmission coefficient tn, (2) reflected charged particle intensity obtained by multiplying the first charged particle intensity by the reflection coefficient rn, and returning from the (n + 1) th layer. And a third charged particle intensity obtained by adding an upper transmitted charged particle intensity obtained by multiplying the second charged particle intensity by the transmission coefficient tn.

  Furthermore, according to a more preferred embodiment of the above aspect of the invention, the first step for the nth layer is recursively performed from the first layer to the lowest layer of the plurality of layers, The third charged particle intensity obtained in the lower first layer is the backscattering intensity.

  In order to achieve the above object, a second aspect of the present invention is a program for causing a computer to execute a procedure for generating the backscattering intensity of the charged particle.

  Further, in order to achieve the above object, in the third aspect of the present invention, the exposure is performed by the exposure data generated by the proximity effect correction step including the step of generating the backscattering intensity of the charged particle. A method of manufacturing a semiconductor device having a process.

  According to the first aspect, since the backscattering intensity to the resist layer is obtained in consideration of the reflection coefficient, transmission coefficient, and scattering of the charged particles in the multilayer structure below the resist layer, the charged particles in each layer are obtained. The backscattering intensity can be obtained with higher accuracy by taking into consideration the influence on the surroundings due to the scattering of. In addition, it is only necessary to obtain the distribution of the lower transmitted charged particle intensity, the reflected transmitted charged particle intensity, and the upper transmitted charged particle intensity in consideration of the charged particle transmission number and reflection number in each layer. The process can be simplified and the calculation time can be shortened.

It is a flowchart figure of the exposure development process of the manufacturing method of the semiconductor device in this Embodiment. It is explanatory drawing of the method of calculating | requiring the backscattering intensity | strength in this Embodiment. It is a figure explaining the area density method. FIG. 4 is a diagram showing the types of backscattered electrons. It is sectional drawing which shows the multiple types of layer structure in this Embodiment. FIG. 6 is a development view of each layer in FIG. 5. 6 is a chart showing parameters of each layer in FIG. 5. It is a figure which shows the outline of scattering of the incident electron with respect to the multilayer structure of FIG. It is a figure for demonstrating the area density method in this Embodiment. It is a figure explaining the production | generation method of the generalized backscattering intensity in this Embodiment. It is a flowchart figure of the production | generation process of backscattering intensity | strength. It is a flowchart figure of a backscattering intensity production | generation subroutine process. It is a flowchart figure of proximity effect amendment to which the backscattering intensity generation method in this embodiment is applied. It is a figure which shows the electron intensity distribution explaining the proximity effect correction | amendment which applied the backscattering intensity production | generation method in this Embodiment. It is a figure which shows the effect of this Embodiment. It is a figure explaining the procedure which produces | generates the exposure data of each layer in 1st Embodiment. It is a principle figure of 2nd Embodiment. It is a flowchart figure of the procedure which produces | generates the exposure data in 2nd Embodiment. It is a figure for demonstrating the principle of the charged particle intensity coefficient map Mn. It is a figure which shows the structure of a charged particle intensity coefficient map. It is a figure which shows the structure of a charged particle intensity coefficient map. It is a flowchart figure which shows the procedure for producing _| generating the charged particle intensity coefficient map _Mn .

  Hereinafter, the best mode for carrying out the present invention will be described. However, the technical scope of the present invention is not limited to such embodiments, but covers the matters described in the claims and equivalents thereof. The present invention is applied to charged particle beam exposure. Hereinafter, an example of electron beam exposure will be described. Therefore, an electron is illustrated as a charged particle, and an electron beam is illustrated as a charged particle beam.

[Whole process of this embodiment]
FIG. 1 is a flowchart of the exposure and development process of the semiconductor device manufacturing method according to the present embodiment. The design pattern data D1 generated by the integrated circuit design tool includes a substance or material to be processed and its pattern data, and is converted into exposure data D2 having an exposure pattern and an exposure amount (S10). That is, the exposure data includes the exposure amount of each shot in electron beam exposure and the pattern data of each shot. This electron beam exposure may be any of the above-described point beam exposure method, variable shaping exposure method, partial batch exposure method, and projection type exposure method. However, when a projection type exposure method is used, the exposure amount is constant regardless of the pattern.

  When a pattern according to a design value is exposed with an exposure amount according to a design value, as described above, the development pattern varies due to the proximity effect. Therefore, the proximity effect corrected exposure data D3 is generated by the proximity effect correction step of correcting one or both of the exposure pattern and the exposure amount in consideration of the proximity effect (S12). This proximity effect correction step S12 includes a step of generating the backscattering intensity with high accuracy. Then, in accordance with the corrected exposure data D3, the exposure apparatus performs electron beam exposure on the semiconductor wafer or mask (S14), the exposed resist is developed, and the layer below the resist layer is patterned (S16). . The patterning target film includes, for example, a wiring layer such as Al and Cu, and a via hole layer such as W and Cu.

  As shown in FIG. 1, in the proximity effect correction step S12, a step of obtaining electron backscattering intensity is performed. A desired semiconductor device is manufactured using the exposure, development, and patterning steps shown in FIG.

[Principle of method for obtaining backscattering intensity in this embodiment]
FIG. 2 is an explanatory diagram of a method for obtaining the backscattering intensity in the present embodiment. In this example, a W / SiO ₂ layer layer 12 having a W contact hole with a pattern density α in a SiO ₂ film is formed on a silicon substrate 10, and an unpatterned Al layer 13 formed thereon is formed. A resist layer 14 is formed thereon. That is, the model of FIG. 2 is a model in which a multilayer structure of the Al layer 13, the W / SiO ₂ layer 12, and the silicon substrate 10 exists under the resist layer 14 to be exposed and developed. A method for obtaining the backscattering intensity when the resist layer 14 is irradiated with an electron beam in such a structure will be described. For simplicity, the electrons incident on the resist layer 14 are scattered and spread on the uniform resist layer and the Al layer for wiring, reach the W / SiO ₂ layer, and the electrons are scattered on the resist layer and the Al layer. However, suppose that most of the electrons fall down just by changing the direction.

As described above, when the Al layer 13 is formed on the W / SiO ₂ layer 12 in which the contact hole is formed and the resist layer 14 on the Al layer 13 is exposed for wiring, it may be backscattering of electrons. (1) Reflected back to the Al layer 13 (2) Transmitted through the Al layer 13 and reflected back to the W / SiO ₂ layer 12 (3) Al layers 13 and W / SiO ₂ layers 12 , And reflected by the Si substrate 10 thereunder to return to the resist. Therefore, a ratio r returning from the film when one electron is incident on a certain layer and a ratio t reaching the back surface through the film are defined. That is, the reflection ratio r _w of Al layer 13, the transmittance t _al, W / reflection ratio r _w of W of the SiO ₂ layer 12, the transmittance t _w, reflection of SiO ₂ ratio r _sio2, the transmittance t _sio2, silicon substrate A reflection ratio r _{si of} 10 is defined respectively. The pattern density of the W contact holes in the W / SiO ₂ layer 12 is α _w and the pattern density of SiO ₂ is (1−α _w ). Further, for the sake of simplicity, it is assumed that the pattern density of the W / SiO ₂ layer 12 is uniform over the entire surface as in the above assumption. In that case, the number of electrons in the above three cases is as follows.

In case (1), the number of electrons (electron intensity) is simply given by r _al . In the case (2), the light passes through the Al layer 13, is scattered by the W / SiO ₂ layer 12, and passes through the Al layer 13 again. Therefore, the number of electrons finally returning to the resist is t _al × [r _w α _w + r _SiO2 (1-α _w )] × t _al . In the case (3), after passing through the Al layer 13, it passes through W or SiO ₂ and is reflected by the Si substrate 10, and again passes through the W or SiO ₂ and Al layer 13 to return to the resist layer. Therefore, the number is t _al × [t _w α _w + t _SiO2 (1 −α _w )] × r _si × [t _w α _w + t _{SiO 2} (1 −α _w )] × t _al . Therefore, since Al has a low specific gravity and a thin film thickness, assuming that r _al = 0 and t _al = 1, the final backscattering intensity is the sum of the number of electrons in the above three cases.

It is obtained like this.

  Next, in the above, it is assumed that the area density of the exposure pattern and the contact hole is uniform. However, in the actual model, the area density varies depending on the region, and accordingly, the area density due to the scattering of electrons is reduced. The impact is also different. Therefore, in order to obtain the backscattering intensity in consideration of the area density that varies depending on these regions, it is necessary to obtain the distribution of the number of electrons using the area density method described in Non-Patent Document 1 described above.

FIG. 3 is a diagram for explaining the area density method. In the area density method, each layer is divided into a mesh shape into a plurality of areas, and the surrounding area (i + l, j) in the region 20 defined by the electron scattering length with respect to the area of interest (i, j). Accumulate the number of electrons scattered from (+ m) to obtain the number of electrons in the area of interest (i, j). In this case, the distribution that spreads due to the scattering of electrons decreases as the scattering distance becomes longer, such as a Gaussian distribution. Therefore, the scattering distribution depending on the distance is defined as the distance coefficient a _{l, m} . In that case, the number of electrons P ′ _{i, j} in the area (i, j) is affected by scattering by the number of electrons P _{i + l, j + m} reaching the surrounding area (i + l, j + m). As shown in FIG.

It is obtained by adding. Thereby, an area is substantially performed.

Therefore, in the present embodiment, the electron scattering length in each layer is defined as σ _Al , σ _W , and σ _{SiO 2} . The distance weighting coefficients corresponding to these scattering lengths are defined as a _Al , a _W , and a _{SiO 2} . First, the electrons incident on the Al layer 13 reach the lower surface of the Al layer while being scattered in the Al layer. Therefore, the electron distribution on the lower surface of the Al layer is the Gaussian distribution whose length range is σ _Al . The distribution is divided by the area. Therefore, using the above-described area density method, the electron distribution P _{0i, j} on the lower surface of the Al layer is obtained for each area.

As described with reference to FIG. 2, of the electron distribution P _{0i, j} on the lower surface of the Al layer, r _W α _{Wi, j} + r _SiO2 (1 −α _{Wi, j} ) is reflected by the W / SiO ₂ layer 12 and is Then, t _W α _{Wi, j} + t _SiO2 (1 − α _{Wi, j} ) reaches the lower surface while being scattered by W and SiO ₂ . Accordingly, the electron distribution P _{1i, j} reaching the lower surface of the W / SiO ₂ layer 12 is similarly determined by the area density method.

Is required. Here since scattering length of electrons in each material is different, the coefficient of the distance _{a Wl, m, a SiO2l,} m are different for each material.

Further, the electrons reaching the lower surface of the W / SiO ₂ layer are scattered and spread on the bulk Si substrate 10 and reach the lower surface of the W / SiO ₂ film again. The electron distribution P _{1i, j} ' at this time is given by the following equation.

Furthermore, electronic P _1i which reaches this W / SiO ₂ layer lower _{surface, j} ' of the electronic P _2i returning the interface between the Al layer and the W / SiO _{2 layer, j} ' is, W / SiO ₂ layer and electrons medium passed through while spreading in scattering length sigma _W and sigma _SiO2 and so the combined electrons plus the electrons reflected scattered within said W / SiO ₂ film, the number of electrons, P _{2i, j} '

It becomes.

The distribution of electrons reaching the upper surface of the Al layer is obtained by spreading the number of electrons P _{2i, j} ' in the equation (7) by the scattering length σ _Al , so that the scattering in the Al layer is taken in by the area density method. The electron intensity distribution Intensity _{i, j}

It is obtained. Again, in the above example, it is assumed that all electrons are transmitted and reflected in the Al layer as a precondition.

The distribution of the intensity of electrons returning to the resist layer 14 based on the above principle can be verified from another viewpoint. FIG. 4 is a diagram showing the types of backscattered electrons. In the same multilayer structure as in FIG. 2, the trajectories of electrons returning to the resist among the electrons incident on the W / SiO ₂ mixed layer 12 can be classified into the following six shown in FIG.

(1) An electron E1 that is reflected by W and comes out. (2) An electron E2 that passes through W and comes out of W. (3) Electrons E3 that pass through W and emerge from SiO ₂ . (4) An electron E4 incident on SiO ₂ and coming out of W. (5) electrons incident on the SiO ₂ coming out of the SiO ₂ E5. (6) Electron E6 that is reflected by SiO ₂ and emerges.

  The number of electrons from E1 to E6 is determined by the following concept.

If the incident electrons are sufficiently scattered and spread by the resist film 14 and the Al layer 13, the electron distribution on the upper surface of the W / SiO ₂ layer 12 becomes almost uniform, and the electrons are incident on W or on the SiO ₂ . It is considered that the incidence depends on the area ratio of W and SiO ₂ . Some of the electrons incident on W or SiO ₂ are reflected and partially transmitted according to their respective transmittances. If the transmittances of the W film and SiO ₂ film are T _w and T _SiO2 , respectively, and the area ratios are α _w and 1−α _w , for example, the number of electrons incident upon W and reflected and returned is α _w (1-T _w ). On the other hand, when electrons return, similarly, the number of electrons can be obtained from the area ratio and transmittance of W and SiO ₂ . As a result, the number of electrons returning to the upper surface of W and SiO ₂ is expressed by the following equation, assuming that the number of incident electrons is 1.
(1) Probability of being reflected by W: α _w (1-T _w )
(2) Probability of passing through W and coming out of W: α _w T _w × α _w T _w
(3) Probability of coming out of SiO ₂ through W: α _w T _w × (1-α _w ) T _SiO2
(4) Probability of entering SiO ₂ and coming out of W: (1-α _w ) T _SiO2 × α _w T _W
(5) the probability is incident on the SiO ₂ coming out of the _{SiO 2: (1 - α w} ) T SiO2 × (1 - α w) T SiO2
(6) Probability of being reflected by SiO ₂ : (1-α _w ) (1-T _SiO2 )
It is considered that the energy given to the resist by the electrons E1 to E6 taking the above six processes varies depending on the scattering process. This is because the average energy of electrons changes due to scattering, and the effective resist sensitivity (resist exposure sensitivity with respect to electron energy) may change. Therefore, the energy given to the resist by the electrons through each process is defined as e ₁ to e ₆ and the sum of the back scattering received by the resist layer when exposed to form an Al wiring of density α is It is as follows.

In order to simplify the above equation (9), (1 − T _W ) e ₁ = η ₁ , T _W ² e _{2 =} η ₂ , T _W T _SiO2 e ₃ = η ₃ , T _SiO2 T _{w w} = If η ₄ , T _SiO2 ² e ₅ = η ₅ , (1−T _siO2 ) e ₆ = η ₆ , the equation (9) is as follows.

These η ₁ to η ₆ are obtained by multiplying the decrease in the number of electrons in the W film and the SiO ₂ film by the resist sensitivity for each path, and when one electron passes through the path. It represents the energy given to the resist.

Density The above formula (9) (10), the density alpha and W contact holes density alpha _W of Al wiring is based on the assumption that a uniform, the area density method similar to the model of Figure 2 The energy in the region of interest can be obtained in consideration of the influence of scattering of electrons from the surrounding region. However, when Formula (10) is applied to a structure of two or more layers, it becomes somewhat complicated and inferior in practicality.

Comparing the equations (3) and (10), the area ratios α _w and 1-α _w are exactly the same. Since both should naturally be equal, the parameters according to each have the following relationship:

  In the model of FIG. 4, the parameter is expressed by the product of the number of electrons and the energy that the electron gives to the resist, whereas in the model of FIG. 2, the number r of reflected electrons with respect to the incident electrons is r. It can be seen that this is expressed only by a parameter relating to the number of electrons, that is, the number of transmitted electrons t. This means that in the model of FIG. 4, the effective resist sensitivity changes depending on the path of scattered electrons, and in the model of FIG. 2, it is considered that the number of electrons has changed. Note that the reflectivity r and the transmittance t do not represent a simple number of electrons, so r + t is not always 1.

  As described above, it was verified that the model in FIG. 2 is substantially the same as the model in FIG.

[Example applied to multiple types of layer structures]
Next, an example in which the present embodiment is applied to a plurality of types of layer structures obtained by further extending the model of FIG. 2 will be described. FIG. 5 is a cross-sectional view showing a plurality of types of layer structures in the present embodiment. FIG. 6 is a development view of each layer of FIG. Further, FIG. 7 is a chart showing parameters of each layer.

The layer structure of FIG. 5 is a resist layer 14, a wiring Al layer 13, a W / SiO ₂ layer 16, an Al / SiO ₂ wiring layer 15, a W / SiO ₂ layer 12, and a Si substrate 10 from the top. As described above, the reflection coefficient r, the transmission coefficient t, and the distance coefficient a with respect to the scattering length σ are given to the substances in each layer, and a part of electrons incident on each layer are reflected while being scattered, and a part is scattered. However, the distribution of electrons that finally return to the resist layer 14 is required because of the idea of transmission.

FIG. 8 is a diagram showing an outline of scattering of incident electrons with respect to the multi-layer structure of FIG. Some of the electrons incident on the resist layer 14 are scattered forward and reach the interface between the resist layer and the Al layer. This number of electrons P0 is partially reflected and partially transmitted within the Al layer 13, and the number of electrons P1 reaches the interface between the Al layer 13 and the W / SiO ₂ layer 16. This number of electrons P1 is also partially reflected and partially transmitted within the W / SiO ₂ layer 16, and the number of electrons P2 reaches the interface between the W / SiO ₂ layer 16 and the Al / SiO ₂ layer 15. This number of electrons P2 is partially reflected and partially transmitted within the Al / SiO ₂ layer 15, and the number of electrons P3 reaches the interface of the W / SiO ₂ layer 12. This number of electrons P3 is also partially reflected and partially transmitted within the W / SiO ₂ layer 12, and the number of electrons P4 reaches the silicon bulk 10. This number of electrons P4 is partially reflected in the silicon bulk 10, and the number of electrons P′4 reaches the lower surface of the W / SiO ₂ layer 12. The number of electrons P′4 partially transmits through the W / SiO ₂ layer 12. As a result, the total P′3 of the electrons reflected and transmitted through the layer 15 reaches the lower surface of the Al / SiO ₂ layer 15. Similarly, the lower surface of the W / SiO ₂ layer 16 reaches a total P′2 of electrons reflected and transmitted through the layer 16, and the lower surface of the Al layer 13 includes electrons reflected from the layer 13. A total P′1 of transmitted electrons reaches, and a total P′0 of electrons reflected and transmitted through the layer 14 reaches the lower surface of the resist layer 14. As described above, electrons that have reached each layer from the top or bottom are partially reflected and partially transmitted, and electrons that have reached the silicon bulk 10 are also partially reflected to rise up each layer, and finally the resist layer 14 is reached. This number of electrons P′0 is the backscattering intensity to the resist layer 14.

  Furthermore, since the above electrons are scattered in each layer, they spread to the surrounding area within the range of the scattering length. Therefore, in order to consider the influence of such scattering, it is necessary to obtain the amount of electrons in the area of interest by applying the area density method described in FIG. In order to apply this area density method, the pattern of each layer is divided into correction areas 18 having a size of A × A, as indicated by the wavy lines in the development view of FIG. Then, the amount of electrons in the target correction area is obtained by dividing the amount of electron scattering from a plurality of areas in the surrounding area with respect to the target correction area. Hereinafter, the electron distribution of each layer is specifically obtained.

First, the amount of electrons P _{0i, j} irradiated to the correction area (i, j) is obtained. The amount of electrons P _{0i, j} irradiated to the correction area (i, j) is determined based on the density occupied by the Al pattern of the Al layer 13 in the correction area. In this case, assuming that the optimum exposure amount for a large area pattern with the highest backscattering intensity is 1, the exposure amount correction for giving an exposure amount that is d times as large as the surrounding pattern with a low pattern density is considered. The exposure amount correction is described in the above-mentioned Patent Document 1, and will be described in an application example described later. Considering this exposure amount correction, the exposure amount is multiplied by d for each exposure pattern. As a result, when there are a plurality of patterns in a certain correction area (i, j), the kth pattern has an area density α _k and the correction exposure amount is d _k , effective electrons irradiated to the area The number (electron intensity) is

It is required as follows. However, the number of electrons is d _k = 1 in the case of the projection exposure method in which the exposure amount cannot be corrected for each pattern.

Next, electrons reaching the interface of each layer are obtained in order by the area density method. FIG. 9 is a diagram for explaining the area density method in the present embodiment. 9, electrons that have reached the upper surface of the second layer of W / SiO ₂ layer 16, a case of obtaining the electronic distribution reaching the interface between the W / SiO ₂ layer 16 and the Al / SiO ₂ layer 15 as an example Show. That, (i + l, j + m) th reaches the correction area electron _{P 1 i + l, j +} m is, W / SiO ₂ layer 16 W and SiO ₂ in the area ratio alpha _W, (1- An amount of electrons corresponding to α _W ) is incident on each of the electrons and scattered in W and SiO ₂ and spreads to the (i, j) th correction area of interest. Further, the spreading method differs between electrons incident on W and electrons incident on SiO ₂ . Therefore, the electron distribution P _{1 i} at the interface between the W / SiO ₂ layer 16 and the Al / SiO ₂ layer 15 takes into account the extent of spread due to scattering of electrons incident on W and electrons incident on SiO _2. Multiply the effects by multiplying _{+ l, j + m} by different distance coefficients a _AL and a _{SiO 2} . That is, the electrons P _{1 i + l, j + m that} have reached the (i + l, j + m) -th correction area have an area ratio α _{W i + l, j + m} , (1−α _{Wi + l, j + m} ) is incident on W and SiO ₂ and expands to the (i, j) -th correction area of interest according to the distance coefficients a _{Al l, m} and a _{SiO2 l, m} .

Based on the above idea, the electron distribution reaching each interface is obtained as follows.
Electron distribution P _{1 i, j} reaching the interface between the Al layer 13 and the W / SiO ₂ layer 16

Electron distribution P _{2 i, j} reaching the interface between the W / SiO ₂ layer 16 and the Al / SiO ₂ layer 15

Electron distribution P _{3 i, j} reaching the interface between the Al / SiO ₂ layer 15 and the W / SiO ₂ layer 12

Electron distribution P _{4 i, j} reaching the interface between the W / SiO ₂ layer 12 and the Si substrate 10

An electron distribution P _{4 i, j} ' which enters the Si substrate 10 and returns to the interface between the Si substrate 10 and the W / SiO ₂ layer 12.

Electron distribution P _{3 i, j} ' returning to the interface between the W / SiO ₂ layer 12 and the Al / SiO ₂ layer 15

Electron distribution P _{2 i, j} ' returning to the interface between the Al / SiO ₂ layer 15 and the W / SiO ₂ layer 16

Electron distribution P _{1 i, j} ' returning to the interface between the W / SiO ₂ layer 16 and the Al layer 13

Electron distribution P _{0 i, j} ' returning to the interface between the Al layer 13 and the resist layer 14

That is, the backscattering intensity finally returning to the resist layer 14 is P _{0 i, j} ' in the equation (25).

  As described above, in the multilayer structure model shown in FIGS. 5 and 6, if the parameters of each layer (see FIG. 7) obtained in advance by experiments are given, the number of electrons reached corresponds to the area ratio and scattering length of each layer. By dividing the area according to the distance coefficient, the backscattering intensity returning to the resist layer can be obtained for each correction area, and a backscattering intensity distribution can be obtained. According to the above method, the backscattering intensity can be determined in the same manner even if the number of layers of the multilayer structure and the substances and materials constituting each layer are different.

[Method for obtaining generalized backscattering intensity]
FIG. 10 is a diagram for explaining a generalized method of generating backscattering intensity in the present embodiment. FIG. 10 is a model in which a multilayer structure composed of N layers is formed on a substrate such as a silicon bulk of the (N + 1) th layer, and a resist is formed thereon. For the sake of generalization, the calculation for the n-th layer will be described. Among the electron intensities (or the number of electrons) Pn−1 transmitted through the n−1th layer and reaching the interface with the nth layer, the reflectance Electron intensity rn * Pn-1 corresponding to rn is reflected, and electron intensity tn * Pn-1 corresponding to transmittance tn is transmitted to reach the (n + 1) th layer. Therefore, the electron intensity Pn reaching the n + 1 layer is Pn = tn * Pn-1. Also, among the electron intensities P'n reaching the nth layer from the lower layer side, the electron intensity tn * P'n corresponding to the transmittance tn is transmitted and reflected by the nth layer rn * Pn-1 Is the electron intensity P′n−1 = tn * P′n + rn * Pn−1 reaching the n−1th layer of the upper layer. As described above, the area ratio α of the substance constituting the n-th layer and the distance coefficient a corresponding to the electron scattering length are used as parameters for the area of the area density method.

  However, the electron intensity P′n reaching the nth layer from the lower layer side cannot be obtained unless the calculation of the (n + 1) th layer is performed. Accordingly, the electron intensity P′n−1 that returns to the interface between the nth layer and the (n−1) th layer by performing the above calculation of the nth layer recursively or recursively from the first layer to the Nth layer. Can be requested. This recursive processing is a processing method that can be easily handled in a program executed by a computer.

  FIG. 11 is a flowchart of the backscattering intensity generation process. FIG. 12 is a flowchart of the backscattering intensity generation subroutine process. In this flowchart, the notation of integration calculation by the area density method is omitted.

  In the present embodiment, the backscattering intensity can be obtained by calculation using a program installed in a general-purpose computer. According to this program, first, the initial value of the model to be obtained is input (S20). Specifically, the initial values are the layer structure of the model, parameters such as reflectance, transmittance, and distance coefficient of each layer (see FIG. 7), the exposure intensity for each electron beam shot irradiated to the resist, and the initial number of layers. The value n = 0 or the like. Then, the backscattering intensity generation subroutine step S22 is executed.

  Moving on to the subroutine of FIG. 12, the number n of layers is incremented (S30). From the electron intensity Pn-1 reaching the nth layer from the upper layer, the transmittance tn, and the reflectance rn, the nth layer and n + The transmitted electron intensity Pn = tn * Pn-1 reaching the interface with the first layer and the reflected electron intensity rn * Pn-1 are obtained (S32). Since it is not the lowest layer at first, step S34 is NO. Furthermore, since the electron intensity Pn ′ reaching from the lower layer has not been determined, the electron intensity Pn−1 ′ reaching the interface between the nth layer and the n−1th layer cannot be determined. Accordingly, step S36 is NO, and the backscattering intensity generation subroutine S22 is further executed. This is a recursive process.

Each time each subroutine step S22 is entered, the layer number n is incremented (S30). Then, when it finally reaches the lowest layer (n = N + 1), the electron intensity P ′ _N = r _{N + 1} * P _N corresponding to the reflectance r _{N + 1} at the silicon bulk of the (N + 1) th layer is obtained (S40). The subroutine is returned, and the process returns to step S42 of the subroutine that called the subroutine. In step S42, the electron intensity P _N-1 & p = ' _{N N} * P _N ' + r _N * P _N-1 that returns to the interface between the Nth layer and the N-1th layer is obtained, and the subroutine returns. . Therefore, in the nth layer subroutine, the electron intensity Pn-1 ′ = tn * Pn ′ + rn * Pn−1 that returns to the interface between the nth layer and the n−1th layer is obtained. When all the subroutines are returned, finally, the backscattering intensity P ₀ ' = r1 * P ₀ + t1 * P ₁ ' returning to the resist layer is obtained.

[Application example of the backscattering intensity generation method of the present embodiment]
FIG. 13 is a flowchart of proximity effect correction to which the backscattering intensity generation method according to the present embodiment is applied. This proximity effect correction is disclosed in Patent Document 1 described above, and is included in this specification by reference. The proximity effect correction disclosed in Patent Document 1 is applied to the variable shaping exposure method, and changes the exposure pattern size based on the forward scattering distribution, and each beam shot based on the forward and back scattering distributions. And a step of adding auxiliary exposure.

  In this application example, first, an electron intensity distribution by forward scattering is obtained (S50). This is obtained from the forward scattering term of the aforementioned equation (1). According to the forward scattering distribution, electrons spread in the lateral direction. For example, the half width of the forward scattering distribution (the width of the distribution at 50% of the maximum intensity) is wider than the exposure pattern width. Therefore, for example, a graphic change is performed to narrow the exposure pattern width so that the half-value width of the forward scattering distribution becomes the width as designed (S52). Specifically, correction is performed to reduce the exposure beam size according to the forward scattering distribution.

  Next, for the corrected beam size, a distribution due to forward scattering and back scattering is obtained (S54). This calculation is performed based on the forward scattering term of the above-described formula (1) for the forward scattering intensity, and the backward scattering intensity is calculated using the method of the present invention. FIG. 14A shows an example of distribution of exposure intensity (electron intensity) before correction. In this example, the intensity gradually decreases in the shot areas i, i + 1, i + 2. This is mainly due to the influence of backscattered electron intensity from the periphery depending on the pattern density around the shot region. Accordingly, correction values di + 1 and di + 2 are added in the shot areas i + 1 and i + 2 so that the exposure amount of each shot area becomes the appropriate value ep. That is, the exposure amount for each beam shot is corrected. The exposure amount to which the correction value is added is the corrected exposure amount (S56). As a result, as shown in FIG. 14B, the exposure amount of each shot area is substantially the same.

  However, even within each correction shot, there is a difference in exposure intensity depending on the surrounding pattern density, so in addition to the electron beam shot corresponding to the exposure pattern, an electron beam shot with a small exposure amount is used as auxiliary exposure. Add (S58). As a result, as shown in FIG. 14C, the exposure intensity is corrected to be constant even in the correction area.

  It is necessary to review the exposure correction and the auxiliary exposure again in consideration of the change in backscattering due to the addition of the exposure correction and the auxiliary exposure. Therefore, steps S54, S56, and S58 are repeated until it is confirmed that the correction is appropriately performed. By repeating these steps several times, the exposure intensity is optimized.

  In the above proximity effect correction process, it is necessary to obtain the electron intensity (exposure intensity) due to backscattering in step S54. The production | generation process of the backscattered electron intensity in this Embodiment is performed in this process S54.

FIG. 15 is a diagram illustrating the effect of the present embodiment. In this experiment, an example in which only one W / SiO ₂ contact hole is formed under an Al layer to be exposed, an example in which two layers are formed, and an example in which three layers are formed are formed by a conventional method. A corrected exposure amount 100 obtained by taking in only one contact hole layer under the Al layer and a corrected exposure amount 1002 obtained in consideration of all contact holes under the Al layer according to the present embodiment are shown. ing. The horizontal axis represents the number of contact hole layers, and the vertical axis represents the ratio of the exposure amount given by the correction calculation when the optimum exposure amount obtained by experiment is 1. In addition, this experiment is a result when a line-and-space pattern having a line width of 240 nm and a pitch of 540 nm is exposed on one, two, and three layers of W / SiO ₂ layers, respectively. The area density of all three layers is 16.2%, and the thickness is 0.75um.

  The exposure amount 100 obtained by the conventional method is deviated from the optimum exposure amount 1 as the number of contact hole layers is increased. On the other hand, the exposure amount 102 obtained by the present embodiment substantially matches the optimum exposure amount 1 even when the number of contact hole layers is increased.

The method of generating the backscattered electron intensity in the present embodiment can also be applied in the proximity effect correction of Patent Document 2 described above. The proximity effect correction of Patent Document 2 relates to a projection type exposure method that batch-transfers a large area, and corrects the mask pattern width for batch transfer in consideration of the distribution by forward scattering and the distribution by backscattering. The feature is that the exposure intensity at the design dimension position is made equal in all patterns. Specifically, if the design width (target width) is W ₀ and the exposure intensity of the reference pattern is ε ₀ , the corrected pattern width W can be obtained by solving the following equation.

  Two items of this formula are the backscattered electron intensity, and this backscattered electron intensity is obtained by the method of the present embodiment. This formula is applied to an actual pattern to create corrected data, and then the mask substrate is exposed by an electron beam exposure apparatus for mask drawing. Thereafter, development and patterning are performed to generate a mask having a corrected pattern width. Then, batch transfer is performed by irradiating the semiconductor wafer with an electron beam through this mask.

[Modification]
In the above embodiment, the area density method is applied using correction areas having the same size in all layers of the multilayer structure. However, as the electrons repeat scattering, the distribution becomes gentle and the change in the distribution becomes gentle. Therefore, in this modified example, the area density method is applied using a correction area having a size that increases toward the lower layer. Thereby, the calculation time of a computer can be shortened. That is, the size of the correction area in the lower layer farther from the resist layer is smaller than the size of the correction area in the upper layer close to the resist layer. Thereby, calculation time can be shortened, keeping calculation accuracy high.

Furthermore, in the above embodiment, since the first and third Al wiring layers and the second and fourth contact hole layers have the same film thickness, the same parameters r, t, and σ are used. ing. However, when any film thickness is thick (thin), the reflectance r is large (small), the transmittance t is small (large), and the diffusion length σ is long (short). Therefore, for example, if the parameters when the film thickness is T are r, t, and σ, the parameters of the film thickness 2T are r ′ = r (1 + t ² ), t ′ = t ² ,

  And When the ideal value and the actual value are slightly different from each other, the correct backscattering intensity can be obtained by using experimentally obtained parameter values.

  Furthermore, in the above embodiment, the lowermost layer is a silicon substrate, but in a multilayer structure, the transmission and reflection scattering of electrons are considered up to a predetermined number of layers, and the lowermost layer having a predetermined number of layers is reflected. Only scattering may be applied.

In the above embodiment, the same parameter is used for the different W / SiO ₂ layers 12 and 16, but even if these layers have the same thickness, different parameters are used. May be. Further, different transmission coefficients tn may be used for the upper transmission electrons and the lower transmission electrons.

  In the above embodiment, the case where the resist layer 14 is formed on the Al wiring layer 13 and the resist layer 14 is exposed to the electron beam has been described. However, the present invention is applied to the following cases regardless of the above embodiment. Is done.

For example, a SiO ₂ film is formed on the Al / SiO ₂ layer 15, a resist layer is formed on the SiO ₂ film, the resist layer is exposed by electron beam exposure, and a resist pattern formed by development or the like is used. When the SiO ₂ layer is dry-etched to form a contact hole for embedding W, then the W layer is formed on the entire surface, W plugs are formed by the damascene method, and the W / SiO ₂ layer 15 is formed It is.

That is, when a contact hole or a wiring groove is formed in an insulating film such as SiO ₂ film and a conductive film such as W or copper is embedded therein, the resist layer formed on the insulating film is exposed when exposed. Needless to say, the present invention can also be applied to the case where the backscattering intensity affected by the wiring layer under the film is obtained.

[Second Embodiment]
According to the calculation procedure of the influence of backscattering in the first embodiment described above, in each layer, the reflection coefficient, the transmission coefficient, and the diffusion distribution are defined for each configured material, and the area density (occupation) where each material exists The flow of the number of electrons in the layer is calculated. That is, as shown in FIG. 10, the flow of the number of electrons P is combined for each layer, and recursively calculated from the top layer to the bottom layer. The number of electrons P0 ′ finally returned to the resist is the energy absorbed by the resist by backscattering. According to this method, since the reflection, transmission, shielding effect, etc. of electrons by each material of each layer are expressed, the backscattering intensity can be accurately calculated.

  FIG. 16 is a diagram for explaining a procedure for generating exposure data of each layer in the first embodiment. In this figure, the layer numbers are 1, 2,... . . n. In order to perform the proximity effect correction for the first layer immediately above the silicon substrate, the area density α of the substance in the area is obtained based on the design data IN1 which is the pattern data at the design stage of the first layer, and the reflection coefficient is calculated. The backscattering intensity is obtained from r, the transmission coefficient t, the diffusion distribution coefficient a, etc., and the design data is corrected in consideration of the proximity effect due to the influence. As a result of the proximity effect correction, exposure data OUT1 of the first layer is obtained.

  Next, in order to obtain the exposure data OUT2 of the second layer, the area density of the substance in the area of each layer is obtained based on the design data IN2 and IN1 of the second and first layers, and the reflection coefficient and transmission of each layer. The backscattering intensity to the second layer is obtained by a recursive method based on the coefficient, diffusion distribution, etc., and proximity effect correction is performed.

  Similarly, in order to obtain the exposure data OUTn of the nth layer, the backscattering intensity to the nth layer is obtained by a recursive method using the design data IN1 to INn of the first layer to the nth layer. Effect correction is performed to obtain exposure data OUTn.

  As described above, according to the first embodiment, in order to obtain the backscattering intensity in the nth layer, the flow of electrons (charged particles) must be sequentially calculated for all affected lower layers. In other words, there is a problem that the proximity effect correction processing takes a longer time as it goes to the upper layer of the multilayer structure. Furthermore, with the recursive process of starting from the top layer, going to the bottom layer, and returning to the top layer again, all the data in the lower layer must remain, and the amount of data retained in the recursive process as it goes to the upper layer There is a problem that increases.

  Therefore, in the second embodiment, it is possible to perform the proximity effect correction by obtaining the backscattering intensity in order from the lower layer without requiring the recursive processing as in the first embodiment.

FIG. 17 is a principle diagram of the second embodiment. FIG. 18 is a flowchart of a procedure for generating exposure data in the second embodiment. As shown in FIG. 17, the silicon substrate is defined as the 0th layer, and the layers above it are defined as 1, 2 to N-1, and the Nth layer in order. Then, in order from the first layer to the upper layer, charged particle intensity coefficient maps M _{0 to} M _n-1 in each layer are obtained, and in the nth layer, the charged particle intensity coefficient map M _n in the n−1th layer. _-1 and the nth layer design data INn are used to determine the backscattering intensity, and proximity effect correction is performed based on the backscattering intensity. In this charged particle intensity coefficient map M _n , the charged particles incident on an arbitrary position (i1, j1) of the nth layer are reflected in the nth layer, and after passing through the nth layer, n−1. This is coefficient data indicating the ratio of charged particles that are reflected by the layers below the first layer and return to the surrounding positions (i2, j2) including the position (i1, j1). That is, the charged particle intensity coefficient map M _n corresponds to the distance coefficient a shown in FIG. 3 and the like in the first embodiment. However, the charged particle intensity coefficient map M _n is a ratio depending on the incident position and the backscattering position because the charged particle incident at one position (i1, j1) returns to another position (i2, j2). On the other hand, the distance coefficient is a coefficient depending on the distance without depending on the position, and the two coefficients are different in that respect.

In the first embodiment, the flow of charged particles in each layer is calculated based on the charged particle intensity (or the number of charged particles) corresponding to the amount of exposure with which the uppermost layer is irradiated. For this reason, as a result of the distribution of the charged particle intensity incident on each layer due to scattering, how the charged particle intensity is distributed is obtained while dividing the area. On the other hand, in the second embodiment, in each layer, the ratio of how much the charged particles incident on an arbitrary position (i1, j1) are backscattered to the surrounding position (i2, j2) is shown. The coefficient is obtained as a charged particle intensity coefficient map. This charged particle intensity coefficient map M _n is independent of the exposure dose, and the pattern data (design data INn) of the material of the nth layer and the charged particle intensity coefficient map M _n-1 of the lower layer immediately below the pattern data Can be obtained from Then, if the charged particle intensity coefficient map M _n−1 of the lower layer immediately below is obtained, the backscattering intensity in the n th layer is the charged particle intensity P incident on the n th layer and the n th layer. And the charged particle intensity coefficient map M _{n−1 of the} (n−1) th layer can be obtained.

Therefore, as shown in FIG. 18, the exposure data OUT is obtained in the proximity effect correction steps S72, S76, and S80 in order from the lower layer of the integrated circuit, and the charged particle intensity coefficient maps M _{1 to} M _{n of} each layer are also obtained at the same time. (S70, S74, S78). That is, when determined by the proximity effect correction exposure data OUT of the n-th layer, the n-th charged particle intensity map M _n-1 of the (n-1) th layer immediately below the design data IN _n layer, n The backscattering intensity to the second layer can be obtained, and there is no need to perform recursive processing as in the first embodiment.

FIG. 19 is a diagram for explaining the principle of the charged particle intensity coefficient map M _n . Here, the charged particle intensity coefficient that affects the surrounding position (i2, j2) including the position (i1, j1) of the charged particle incident at an arbitrary position (i1, j1) is obtained. First, as shown in FIG. 19A, charged particles incident on an arbitrary position (i1, j1) of the n-th layer are reflected in the n-th layer and are moved to the position (i2, j2). To reach. Since it scatters in the nth layer, it has a distribution E1 at position (i2, j2). This distribution is, for example, a Gaussian distribution.

Next, as shown in FIG. 19B, the charged particles incident on the position (i1, j1) are transmitted through the n-th layer while being scattered, and the interface with the (n−1) -th layer is transmitted. The position (i3, j3) is reached. Also in this case, the distribution spreads at the position (i3, j3) with the distribution E2. This distribution E2 can be obtained from the transmission coefficient T, the area density α of the material of the nth layer, and the distance coefficient A indicating the distribution. Further, the charged particles that have reached the position (i3, j3) are reflected while being scattered in the lower layer below the (n−1) th layer, and arrive at the interface position (i4, j4) with the distribution E3. This reaching ratio is shown in the charged particle intensity coefficient map M _n−1 of the (n−1) th layer. Therefore, the distribution E3 can be obtained by E3 = M _n-1 * E2. However, since E2 has a distribution, it is necessary to divide the area of the distribution E2 in order to obtain the distribution E3. The charged particles that have reached the position (i4, j4) are transmitted through the n-th layer while being scattered, and reach the position (i2, j2) with the spread of the distribution E4. This distribution E4 can be obtained from the transmission coefficient T, the area density α of the material of the nth layer, and the distance coefficient A indicating the distribution. In addition, the distribution E3 needs to be divided into areas.

Eventually, the charged particle intensity map M _n indicating the influence of the charged particles that have reached the position (i1, j1) of the nth layer on the position (i2, j2) is the sum of the distributions E1 and E4. The charged particle intensity coefficient map M _n has charged particle intensity coefficients corresponding to all positions (i1, j1) of the nth layer for all positions (i2, j2) around it.

  20 and 21 are diagrams showing the configuration of a charged particle intensity coefficient map. The charged particle intensity coefficient map is configured using a plurality of divided areas. That is, the charged particle intensity coefficient map is a data structure having as an element a charged particle intensity coefficient that enters an arbitrary area (position) (i1, j1) and reaches an area (i2, j2) in the vicinity thereof. As shown in FIG. 20, the data structure includes a table 30 composed of a two-dimensional array of M × N areas, and a table 40 of an L × L two-dimensional array provided corresponding to each area of the table 30. Consists of. For example, charged particle intensity coefficients that enter an arbitrary area (i1, j1) and reach an L × L area (i2-i1, j2-j1) in the vicinity thereof are each element of the table 40. Here, the area (i2-i1, j2-j1) of the table 40 corresponds to the surrounding area centered at the area (i1, j1) (0, 0), and each of the local coordinates (i2-i1, j2-). It is specified in j1). The table 40 corresponds to each area of the table 30, and an element of the table 30 is a reference number (index) to the table 40.

  FIG. 21 shows a more specific data structure. Reference numbers A00 to A22 are stored in each element of the table 30 composed of a 3 × 3 two-dimensional array. The charged particle intensity coefficients M-1-1 to M00 to M11 are stored in each element of the table 40 referred to by the reference number A00. Since the table 30 consists of nine areas, the table 40 is also composed of nine each. However, in the case of the same pattern structure, the charged particle intensity coefficient of the corresponding table 40 may be the same. In that case, the elements of the table 30 can refer to the same table 40, and the amount of data can be reduced.

  Thus, the charged particle intensity coefficient map in the present embodiment is composed of the following two types of tables.

  (1) First, L × having a charged particle intensity coefficient of charged particles that reach an arbitrary relative area (i2-i1, j2-j1) around an area (i1, j1) as data. It is a table 40 of a two-dimensional array of L. Here, L is the number of areas that can be arranged in a region where charged particles reach with the area (i1, j1) as the center.

  (2) Second, there is an M × N two-dimensional array table 30 having reference data to the corresponding two-dimensional array 40. M and N are the numbers of areas in the horizontal and vertical directions of the integrated circuit chip area, respectively.

Therefore, when the distribution E3 of the nth layer is obtained using the charged particle intensity coefficient map M _n−1 , charged particles of electrons that enter the area (i1, j1) and reach the area (i2, j2) The intensity coefficient is obtained by referring to the element corresponding to the area (i1, j1) of the table 30 and extracting the element corresponding to the relative area (i2-i1, j2-j1) of the table 40 referenced therefrom. It is done. If the surrounding structures are the same in a plurality of areas, the tables 40 for those areas are also the same, so the same table 40 can be referred to, and in this case, the amount of data can be reduced.

FIG. 22 is a flowchart showing a procedure for generating the n-th layer charged particle intensity coefficient map M _n from the n-th layer design data and the n−1th charged particle intensity coefficient map M _n−1. FIG. A method of generating a charged particle intensity coefficient map will be specifically described with reference to FIG.

  First, as a pre-process, the area where the pattern exists is divided into meshes for the design data INn of the nth layer, the pattern area density of each divided area is calculated, and a pattern area density map α is generated. (S90). At that time, even if the hole pattern of the contact hole layer is rectangular on the design data, the hole pattern is actually rounded and rounded. Therefore, it is desirable to resize the design data pattern or perform a process of multiplying the pattern area density by a ratio to bring it close to the actually measured pattern area density.

Furthermore, for each material contained in the nth layer, the ratio _Tk (transmission coefficient) of electrons incident on the nth layer through the layer, the ratio _Rk (reflection coefficient) of reflection in the layer, Define the diffusion distribution (scattering length: σ _k ). Here, k is the name of each material, and the reflection coefficient R _k and the transmission coefficient T _k are the same as the reflection coefficient r and the transmission coefficient t in the first embodiment. For example, when the n-th layer is formed by W and SiO _2, the transmission coefficient of each material, the reflection coefficient, the scattering length is _{T W, R W, σ W} , T SiO2, R SiO2, the sigma _SiO2 Become. The diffusion distribution is assumed to be Gaussian. In the following, in order to simplify the explanation, the case where the substrate is formed of Si and the 1st to nth layers are formed of W and SiO ₂ will be treated.

  Next, the area (i1, j1) is fixed, and the charged particle intensity coefficient exerted on the adjacent area by charged particles (for example, electrons) incident on the area is obtained. As described in FIG. 19, the charged particle intensity coefficient reaching the adjacent area (i2, j2) is the sum of the charged particle intensity coefficients calculated by the following two paths.

(1) E1: Path through which charged particles incident on the area (i1, j1) are reflected in the nth layer and reach the area (i2, j2) (FIG. 19A)
(2) E2, E3, E4: Charged particles incident on the area (i1, j1) are transmitted through the nth layer, reflected by the n−1 and lower layers, and further transmitted through the nth layer. , Route to area (i2, j2) (Fig. 19B)
The path of (1) is obtained by the following equation weighted by the area density α of each material using the reflection coefficient R of each material (S93).

Here, A _{W, i2-i1, j2-j1} and A _{SiO2, i2-i1, j2-j1} are diffusion intensity coefficients from area (i1, j1) to area (i2, j2) in W and SiO ₂ respectively. It is the same as the distance coefficient a in the first embodiment. The diffusion intensity coefficient A has an area suffix (i2-i1, j2-j1), but is actually a distance coefficient corresponding to only the distance. As shown in the above equation (26), the charged particle intensity coefficient E1 in the path of (1) is the area density α _{W, i1, j1} , (1-α _{W, i1} ) of each material in the area (i1, j1). _{, j1} ), reflection coefficient R _W , R _SiO2 and diffusion intensity coefficient A _{W, i2-i1, j2-j1} , A _{SiO2, i2-i1, j2-j1} .

  The route (2) is represented by a connection of three routes (E2, E3, E4). First, the charged particle intensity coefficient E2 of electrons reaching the area (i3, j3) through the nth layer from the area (i1, j1) and reaching the area (i3, j3) is as follows, similarly to the equation (26). (S94).

Here, A _{W, i3-i1, j3-j1} and A _{SiO2, i3-i1, j3-j1} are diffusion intensity coefficients from area (i1, j1) to area (i3, j3) in W and SiO ₂ respectively. (Distance coefficient).

Next, the charged particle intensity coefficient E3 of electrons that reach the area (i4, j4) after being reflected and diffused by the (n−1) th and lower layers from the area (i3, j3) is E2 _i3 obtained by the equation (27). _{, j3} and the element coefficient E _n-1 (i3, j3; i4, j4) of the n−1th charged particle intensity coefficient map M _n−1 is expressed as follows.

As described above, E _n-1 (i3, j3; i4, j4) enters the area (i3, j3) in the ( _n−1 ) th charged particle intensity coefficient map M _n-1 and enters the area (i4, j4). This is the charged particle intensity coefficient of the electrons returned to. Since E2 has a distribution, the charged particle intensity coefficient E3 of the charged particles that enter the area (i1, j1) and reach each area (i4, j4) can be expressed by the above equation (28) for the distribution E2 ( It is obtained by accumulating with i3, j3). That is, the charged particle intensity coefficient E3 is a cumulative value of the equation (28) for a plurality of areas (i3, j3) in the distribution E2, and is expressed as follows (S95).

Finally, the charged particle intensity coefficient of electrons that reach the area (i2, j2) from the bottom through the nth layer from the area (i4, j4) uses E3 _{i4, j4} obtained by the equation (29) As in the equation (27), the following is given.

Here, A _{W, i2-i4, j2-j4} and A _{SiO2, i2-i4, j2-j4} are diffusion intensity coefficients from area (i4, j4) to area (i2, j2) in W and SiO ₂ respectively. (Distance coefficient). Therefore, the charged particle intensity coefficient E4 of electrons reaching the area (i2, j2) from the area (i1, j1) is a distribution in which E3 includes a plurality of areas (i4, j4). This is the cumulative value of equation (30) for the surrounding area (i4, j4) and is expressed as follows (S96).

From the above, the charged particle intensity coefficient En of the electrons that enter the area (i1, j1) and reach the area (i2, j2) is _expressed by the sum E1 + E4 of the equations (26) and (31) as follows: Given (S97).

Then, the charged particle intensity coefficient map of n th layer M _n is, for all combinations of all the areas (i1, j1) and its neighboring area (i2, j2), wherein the charged particle intensity coefficient E _n It is generated by obtaining in (32). That is, the charged particle intensity coefficient E _n obtained by the equation (32) as the elements (i2-i1, j2-j1) of the L × L table 40 for all the elements (i1, j1) of the M × N table 30. Is given to become a charged particle intensity coefficient map Mn (S91, S92). The data structure of the charged particle intensity coefficient map M _n is as shown in FIGS.

Next, the proximity effect correction steps S72, S76, S80 in FIG. 18 will be described. In this proximity correction step, it is necessary to obtain the backscattering intensity, and the nth from the above-described particle intensity coefficient map _Mn-1 of the ( _n-1 ) th layer and the design data INn of the _nth layer. A method for obtaining the backscattering intensity Fb in each layer will be described.

Since the charged particle intensity map M _n-1 represents the ratio (charged particle intensity coefficient) of electrons incident on one area returning to another area, the exposure dose irradiated to the area (i + l, j + m) The backscattering intensity that Q gives to area (i, j) is given by the following equation.

Where α _{i + l, j + m} is the pattern area density in area (i + l, j + m), and Q _{i + l, j + m} is the exposure irradiated to area (i + l, j + m). Amount (charged particle amount, charged particle intensity). Therefore, the backscattering intensity Fb _{i, j} given from the surrounding area to the area (i, j) is as follows by dividing the equation (33) by the area density method of FIG.

When this equation (34) is compared with the equation (4) in FIG. 3 of the first embodiment, the number of electrons P in the equation (4) becomes the exposure amount Q and the pattern area density α in the equation (34). The distance coefficient a in the equation (4) corresponds to the charged particle intensity coefficient En _-1 in the equation (34).

  In the proximity effect correction steps S72, S76, and S80, exposure amount correction, figure change, or figure change + exposure amount correction is performed using the backscattering intensity distribution given by Expression (34). The proximity effect correction method disclosed in Patent Document 1 is as described in detail with reference to FIG. 13, and is calculated by the formula (34) when performing graphic change, exposure amount correction, and addition of auxiliary exposure. Scattering intensity is used. Further, in the proximity effect correction method disclosed in Patent Document 2, since the transfer is batch transfer, correction is performed only by changing the figure without adding exposure amount correction or auxiliary exposure. Also in this graphic change, the backscattering intensity calculated | required by Formula (34) is utilized.

  Finally, returning to FIG. 18, the exposure data generation process of the integrated circuit device having the multilayer wiring structure using the second embodiment is as follows. From the first layer formed on the silicon substrate to the upper layer, for each layer, the exposure data OUT is generated by proximity effect correction from the design data and the lower layer charged particle intensity coefficient map, and the design data and the lower layer are generated. A charged particle intensity coefficient map in the layer is generated from the charged particle intensity coefficient map.

First, the backscattering intensity is obtained from the design data IN1 of the first layer and the charged particle intensity coefficient M0 of the silicon substrate by the equation (34), and the exposure data OUT1 is obtained by proximity effect correction. At the same time, the charged particle intensity coefficient map M1 of the first layer is obtained from the design data IN1 and the charged particle intensity coefficient map M0 by Expression (32). At this time, an area density map α for each area and for each material is generated from the design data IN1, and parameters given in advance (see FIG. 7) are used. Note that the charged particle intensity coefficient map M0 of the silicon substrate is the same as the distance coefficient a independent of position because the silicon substrate is made of only a silicon material and the table 40 for all areas is the same. That is, the charged particles that have passed through the first layer are reflected while being scattered according to the distance coefficient a (diffusion distribution parameter) in the silicon substrate and returned to the first layer. The charged particle intensity coefficient E3 ₁ (see FIG. 17) can be obtained from the equation (29).

  Next, the backscattering intensity is obtained from the design data IN2 of the second layer and the charged particle intensity coefficient map M1 of the first layer, and the exposure data OUT2 is obtained by proximity effect correction (S76). At the same time, a charged particle intensity coefficient map M2 of the second layer is generated from the design data IN2 and the charged particle intensity coefficient map M1.

By repeating the proximity effect correction and the generation of the charged particle intensity coefficient map, the exposure data OUT and the charged particle intensity coefficient map M for all layers can be generated. In order to obtain the backscattering intensity in the nth layer, it suffices if there is a charged particle intensity coefficient map _Mn-1 of the ( _n-1 ) th layer and design data of the nth layer. There is no need to refer to state or data. Similarly, it is not necessary to calculate recursively as in the first embodiment.

  The above embodiment is summarized as follows.

(Appendix 1)
In a method for generating a backscattering intensity of the charged particles to the resist layer when each layer is irradiated with a charged particle beam on a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances ,
A reflection coefficient rn corresponding to the number of particles reflected by the n-th layer from the n-1th layer with respect to the n-th layer from the resist layer among the plurality of layers, and the n-th layer The nth layer includes a transmission coefficient tn corresponding to the number of particles through which the charged particles that have passed through the nth layer pass, and a diffusion distribution in which the charged particles scatter within the nth layer. Given for each substance
The generation method is:
A method for generating a backscattering intensity of charged particles, comprising: a first step of generating the backscattering intensity using the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution.

(Appendix 2)
In Appendix 1,
In the first step, a charged particle scattering intensity that causes the backscattering intensity is generated for each of the plurality of layers by using the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution. A method for generating backscattering intensity, comprising:

(Appendix 3)
In Appendix 2,
In the first step,
With respect to a predetermined region of interest in the nth layer, the charged particle scattering intensity from the surrounding region including the region of interest is expressed by the area density αn of the substance in the nth layer, the surrounding region, and A method for generating a backscattering intensity of charged particles, characterized in that the area is divided according to a distance coefficient an obtained from a scattering distribution corresponding to a distance between regions of interest.

(Appendix 4)
In Appendix 3,
The charged particle scattering intensity in the first step is (1) a downward transmitted charged particle intensity obtained by multiplying the first charged particle intensity transmitted through the n−1th layer by the transmission coefficient tn, (2) The reflected charged particle intensity obtained by multiplying the first charged particle intensity by the reflection coefficient rn and the second charged particle intensity returned from the (n + 1) th layer are multiplied by the transmission coefficient tn. And a third charged particle intensity obtained by adding the upward transmitted charged particle intensity. A method for generating the backscattered intensity of charged particles.

(Appendix 5)
In Appendix 4,
The method for generating a backscattering intensity of charged particles, wherein the first step for the nth layer is recursively performed from the first layer to the lowest layer of the plurality of layers.

(Appendix 6)
In Appendix 5,
A method for generating a backscattering intensity of charged particles, wherein the third charged particle intensity obtained in the first layer below the resist layer is set as the backscattering intensity.

(Appendix 7)
In any of Supplementary Notes 4 to 6,
When performing the first step on the first layer of the plurality of layers, the first charged particle intensity is obtained according to an exposure pattern density from the charged particle intensity irradiated on the resist layer. A method for generating a characteristic backscattering intensity of charged particles.

(Appendix 8)
In any of Supplementary Notes 4 to 6,
In the area of the first step, the nth layer is divided into a plurality of subdivided areas, and in each area, the first charged particle intensity, the lower transmitted charged particle intensity, the reflected charged particle intensity, the upper A method for generating a backscattering intensity of a charged particle, characterized in that a transmitted charged particle intensity and a third charged particle intensity are obtained.

(Appendix 9)
In Appendix 8,
The method for generating a backscattering intensity of charged particles, wherein the size of the area is larger in a lower layer farther from the resist layer than in an upper layer near the resist layer.

(Appendix 10)
In any one of appendices 1 to 9,
The substance contained in each layer includes at least one or both of a conductive material, an insulating material, and a semiconductor material.

(Appendix 11)
A step of generating a backscattering intensity of the charged particles on the resist layer when a charged particle beam is irradiated on a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances. In a backscatter intensity generation program to be executed by a computer,
A reflection coefficient rn corresponding to the number of particles reflected by the n-th layer from the n-1th layer with respect to the n-th layer from the resist layer among the plurality of layers, and the n-th layer The nth layer includes a transmission coefficient tn corresponding to the number of particles through which the charged particles have reached the nth layer and a scattering distribution in which the charged particles scatter within the nth layer. Given for each substance
The generation procedure is as follows:
A charged particle backscattering intensity generation program, comprising: a step of generating the backscattering intensity using the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution.

(Appendix 12)
In a method for manufacturing a semiconductor device, each layer having a plurality of layers each including a pattern of one substance or a plurality of substances,
Generating a backscattering intensity of the charged particles to the resist layer when the resist layer formed on at least one of the plurality of layers is irradiated with a charged particle beam;
Exposure data having an exposure pattern, performing proximity effect correction according to the backscattering intensity to generate corrected exposure data;
A step of performing exposure, development, and patterning with the corrected exposure data,
In the step of generating the backscattering intensity, the charged particles that have passed through the (n-1) th layer correspond to the number of particles reflected by the nth layer with respect to the nth layer from the resist layer among the plurality of layers. A reflection coefficient rn, a transmission coefficient tn corresponding to the number of particles through which the charged particles that have reached the n-th layer pass through the n-th layer, and a scattering distribution in which charged particles scatter within the n-th layer. Is provided for each of the substances contained in the nth layer,
A method for manufacturing a semiconductor device, comprising: a first step of obtaining the backscattering intensity using the reflection coefficient rn, the transmission coefficient tn, and the diffusion distribution.

(Appendix 13)
In Appendix 12,
In the first step,
With respect to a predetermined region of interest in the nth layer, the charged particle scattering intensity from the surrounding region including the region of interest is expressed by the area density αn of the substance in the nth layer, the surrounding region, and A method of manufacturing a semiconductor device, comprising: a first step of dividing an area according to a distance coefficient an obtained from the scattering distribution corresponding to a distance between regions of interest.

(Appendix 14)
In Appendix 13,
The charged particle scattering intensity in the first step is (1) a downward transmitted charged particle intensity obtained by multiplying the first charged particle intensity transmitted through the n−1th layer by the transmission coefficient tn, (2) The reflected charged particle intensity obtained by multiplying the first charged particle intensity by the reflection coefficient rn and the second charged particle intensity returned from the (n + 1) th layer are multiplied by the transmission coefficient tn. And a third charged particle intensity obtained by adding the upward transmitted charged particle intensity. A method for manufacturing a semiconductor device, comprising:

(Appendix 15)
In Appendix 14,
The method for manufacturing a semiconductor device, wherein the first step for the nth layer is recursively performed from the first layer to the lowest layer of the plurality of layers.

(Appendix 16)
In Appendix 15,
A method of manufacturing a semiconductor device, wherein the intensity of the third charged particle obtained in the first layer below the resist layer is set as the backscattering intensity.

(Appendix 17)
In any one of appendixes 14 to 16,
When the first step is performed on the first layer of the plurality of layers, the first charged particle intensity is obtained in accordance with an exposure pattern density. A method of manufacturing a semiconductor device.

(Appendix 18)
In any one of appendixes 14 to 16,
In the area of the first step, the nth layer is divided into a plurality of subdivided areas, and in each area, the first charged particle intensity, the lower transmitted charged particle intensity, the reflected charged particle intensity, the upper A method for manufacturing a semiconductor device, comprising: obtaining a transmitted charged particle intensity and a third charged particle intensity.

(Appendix 19)
In Appendix 18,
A method of manufacturing a semiconductor device, wherein the size of the area is larger in a lower layer farther from the resist layer than in an upper layer near the resist layer.

(Appendix 20)
In any one of appendices 12 to 19,
A method for manufacturing a semiconductor device, wherein the lowermost layer of the plurality of layers is a semiconductor substrate.

(Appendix 21)
In any one of appendixes 12 to 20,
The substance contained in each layer includes at least one or both of a conductive material, an insulating material, and a semiconductor material.

(Appendix 22)
A method for generating a backscattering intensity of the charged particles to the resist layer when each layer is irradiated with a charged particle beam on a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances. In the method for generating the backscattering intensity in the nth resist layer from the substrate toward the upper layer,
Of the area divided into a plurality of minute regions, the ratio of the charged particles incident on the first area being reflected by the layers below the n−1th layer below the nth layer and returning to the second area A charged particle intensity map (M _n-1 ) obtained for the first and second areas, and design data (IN _n ) including the nth pattern data The method for generating the backscattering intensity of the charged particles, further comprising the step of obtaining the backscattering intensity.

(Appendix 23)
In Appendix 22,
Each material of the nth layer has a transmission coefficient ( _Tn ) corresponding to a ratio of charged particles incident on the nth layer through the nth layer and a ratio of reflecting the nth layer. Is given a reflection coefficient (R _n ) corresponding to, and a diffusion distribution coefficient (A _n ) in the n th layer,
The method of generating the backscattering intensity is:
Furthermore, the method includes a step of obtaining the charged particle intensity coefficient map (M _n ) in a layer below the nth layer,
The step of obtaining the charged particle intensity coefficient map (M _n )
Using the pattern area density (α _n ) of each material included in the n-th layer obtained from the design data (IN _n ) of the n-th layer, the reflection coefficient, and the diffusion distribution coefficient, Obtaining a first reflected charged particle intensity coefficient (E1) indicative of a proportion of charged particles reflecting in the nth layer;
Using the pattern area density, the transmission coefficient, the diffusion distribution coefficient, and the charged particle intensity coefficient map (M _n-1 ) in the layers below the n-1 th layer, the n-1 th layer Obtaining a second reflected charged particle intensity coefficient (E4) indicating the proportion of charged particles reflected from the following layers;
And a step of obtaining the sum (E1 + E4) of the first and second reflected electrified particle intensity coefficients.

(Appendix 24)
In Appendix 23,
The step of obtaining the charged particle intensity coefficient map includes:
A method of generating a backscattered intensity of charged particles, wherein the sum (E1 + E4) of the reflected charged particle intensity coefficients is obtained for each second area group corresponding to all of the first areas.

(Appendix 25)
In Appendix 22,
The step of obtaining the backscattering intensity includes
The exposure dose (Q) incident on the nth layer, the pattern area density (α) obtained from the design data of the nth layer, and the charged particle intensity coefficient map of the n−1th layer A method of generating a backscattered intensity of charged particles, comprising a step of multiplying a charged particle intensity coefficient (E _n-1 ) included in (M _n-1 ) to obtain the backscattered intensity.

(Appendix 26)
A procedure for generating a backscattering intensity of the charged particles to the resist layer when a charged particle beam is irradiated onto a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances. A backscattering intensity generation program for causing a computer to execute a procedure for generating the backscattering intensity in the nth resist layer from the substrate toward the upper layer.
The generation procedure is as follows:
Of the area divided into a plurality of minute regions, the ratio of the charged particles incident on the first area being reflected by the layers below the n−1th layer below the nth layer and returning to the second area A charged particle intensity map (M _n-1 ) obtained for the first and second areas, and design data (IN _n ) including the nth pattern data A program for generating the backscattering intensity of the charged particles, characterized in that a program for obtaining the backscattering intensity is obtained.

(Appendix 27)
In a method for manufacturing a semiconductor device, each layer having a plurality of layers each including a pattern of one substance or a plurality of substances,
Generating a backscattering intensity of the charged particles to the resist layer when the resist layer formed on at least one of the plurality of layers is irradiated with a charged particle beam;
Exposure data having an exposure pattern, performing proximity effect correction according to the backscattering intensity to generate corrected exposure data;
A step of performing exposure, development, and patterning with the corrected exposure data,
The step of generating the backscattering intensity includes:
Of the area divided into a plurality of minute regions, the ratio of the charged particles incident on the first area being reflected by the layers below the n−1th layer below the nth layer and returning to the second area A charged particle intensity map (M _n-1 ) obtained for the first and second areas, and design data (IN _n ) including the nth pattern data And a step of obtaining the backscattering intensity.

(Appendix 28)
In Appendix 27,
Each material of the nth layer has a transmission coefficient ( _Tn ) corresponding to a ratio of charged particles incident on the nth layer through the nth layer and a ratio of reflecting the nth layer. Is given a reflection coefficient (R _n ) corresponding to, and a diffusion distribution coefficient (A _n ) in the n th layer,
The step of generating the backscattering intensity includes:
Furthermore, the method includes a step of obtaining the charged particle intensity coefficient map (M _n ) in a layer below the nth layer,
The step of obtaining the charged particle intensity coefficient map (M _n )
Using the pattern area density (α _n ) of each material included in the n-th layer obtained from the design data (IN _n ) of the n-th layer, the reflection coefficient, and the diffusion distribution coefficient, Obtaining a first reflected charged particle intensity coefficient (E1) indicative of a proportion of charged particles reflecting in the nth layer;
Using the pattern area density, the transmission coefficient, the diffusion distribution coefficient, and the charged particle intensity coefficient map (M _n-1 ) in the layers below the n-1 th layer, the n-1 th layer Obtaining a second reflected charged particle intensity coefficient (E4) indicating the proportion of charged particles reflected from the following layers;
And a step of obtaining a sum (E1 + E4) of the first and second reflected electrified particle strength coefficients.

(Appendix 29)
In Appendix 27,
The method of manufacturing a semiconductor device, wherein the step of generating the corrected exposure data and the step of obtaining the charged particle intensity coefficient map are performed for each layer from the substrate side toward the upper layer.

  According to the above invention, in the generation of exposure data used in the exposure method of the semiconductor device manufacturing method, the step of accurately obtaining the backscattering intensity of the charged particles in the resist layer in consideration of the plurality of layers under the resist layer Is available.

10: Silicon substrate, 12: W / SiO ₂ layer, 13: Al layer, 14: Resist layer
P0: Number of electrons, electron intensity

Claims (8)

A method for generating a backscattering intensity of the charged particles to the resist layer when each layer is irradiated with a charged particle beam on a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances. In the method for generating the backscattering intensity in the nth resist layer from the substrate toward the upper layer,
Among the areas divided into a plurality of minute regions, the charged particles incident on the first area are reflected by the layers below the i-1th layer below the i (i <n) th layer and reflected by the second The charged particle intensity coefficient map (M _i-1 ) obtained by calculating the charged particle intensity coefficient composed of the ratio of returning to the area corresponding to all the first and second areas, and the pattern data of the i-th layer The step of obtaining the charged particle intensity coefficient map (M _i ) of the i-th layer from the design data (IN _i ) included is repeated from the i = 1th layer to the i = n−1th layer, and n A step of obtaining a charged particle intensity coefficient map (M _n-1 ) in a layer below the first layer ,
A step of obtaining the backscattering intensity from a charged particle intensity coefficient map (M _n-1 ) in a layer below the n−1 th layer and design data (IN _n ) including the n th pattern data; A method for generating a backscattering intensity of a charged particle, comprising: In claim 1,
Each material of the i-th layer, the ratio of the charged particles incident on the i-th layer permeability coefficient corresponding to the ratio of transmitted through the i-th layer and (T _i), reflects the i-th layer And a reflection distribution coefficient ( A _i ) in the i- th layer, given by the reflection coefficient ( R _i ) corresponding to
The step of obtaining the charged particle intensity coefficient map ( M _i ) of the i-th layer includes:
Using the pattern area density of each material (alpha _i) included in the i th layer obtained from the design data of the i-th layer (IN _i), wherein the reflection coefficient, and the diffusion distribution coefficient, Obtaining a first reflected charged particle intensity coefficient (E1) indicative of a proportion of charged particles reflecting in the i- th layer;
And the pattern area density, and the transmission coefficient, the diffusion and distribution coefficient, i -1-th layer below charged in layer particle intensity coefficient map (M _{i -1)} and by using the i -1-th layer Obtaining a second reflected charged particle intensity coefficient (E4) indicating the proportion of charged particles reflected from the following layers;
And a step of obtaining the sum (E1 + E4) of the first and second reflected electrified particle intensity coefficients. In claim 2,
The step of obtaining the charged particle intensity coefficient map includes:
A method of generating a backscattered intensity of charged particles, wherein the sum (E1 + E4) of the reflected charged particle intensity coefficients is obtained for each second area group corresponding to all of the first areas. In claim 1,
The step of obtaining the backscattering intensity includes
The exposure dose (Q) incident on the nth layer, the pattern area density (α) obtained from the design data of the nth layer, and the charged particle intensity coefficient map of the n−1th layer A method of generating a backscattered intensity of charged particles, comprising a step of multiplying a charged particle intensity coefficient (E _n-1 ) included in (M _n-1 ) to obtain the backscattered intensity. A procedure for generating a backscattering intensity of the charged particles to the resist layer when a charged particle beam is irradiated onto a resist layer formed on a plurality of layers each including a pattern of one substance or a plurality of substances. A backscattering intensity generation program for causing a computer to execute a procedure for generating the backscattering intensity in the nth resist layer from the substrate toward the upper layer.
The generation procedure is as follows:
Among the areas divided into a plurality of minute regions, the charged particles incident on the first area are reflected by the layers below the i-1th layer below the i (i <n) th layer and reflected by the second The charged particle intensity coefficient map (M _i-1 ) obtained by calculating the charged particle intensity coefficient composed of the ratio of returning to the area corresponding to all the first and second areas, and the pattern data of the i-th layer The step of obtaining the charged particle intensity coefficient map (M _i ) of the i-th layer from the design data (IN _i ) included is repeated from the i = 1th layer to the i = n−1th layer, and n A step of obtaining a charged particle intensity coefficient map (M _n-1 ) in a layer below the first layer ,
A procedure for obtaining the backscattering intensity from a charged particle intensity coefficient map (M _n-1 ) in a layer below the n−1 th layer and design data (IN _n ) including the n th pattern data. A program for generating a backscattering intensity of a charged particle, comprising: In a method for manufacturing a semiconductor device, each layer having a plurality of layers each including a pattern of one substance or a plurality of substances,
Generating a backscattering intensity of the charged particles to the resist layer when the resist layer formed on at least one of the plurality of layers is irradiated with a charged particle beam;
Exposure data having an exposure pattern, performing proximity effect correction according to the backscattering intensity to generate corrected exposure data;
A step of performing exposure, development, and patterning with the corrected exposure data,
The step of generating the backscattering intensity includes:
Among the areas divided into a plurality of minute regions, the charged particles incident on the first area are reflected by the layers below the i-1th layer below the i (i <n) th layer and reflected by the second The charged particle intensity coefficient map (M _i-1 ) obtained by calculating the charged particle intensity coefficient composed of the ratio of returning to the area corresponding to all the first and second areas, and the pattern data of the i-th layer The step of obtaining the charged particle intensity coefficient map (M _i ) of the i-th layer from the design data (IN _i ) included is repeated from the i = 1th layer to the i = n−1th layer, and n A step of obtaining a charged particle intensity coefficient map (M _n-1 ) in a layer below the first layer ,
A step of obtaining the backscattering intensity from a charged particle intensity coefficient map (M _n-1 ) in a layer below the n−1 th layer and design data (IN _n ) including the n th pattern data; A method for manufacturing a semiconductor device, comprising: In claim 6,
Each material of the i-th layer, the ratio of the charged particles incident on the i-th layer permeability coefficient corresponding to the ratio of transmitted through the i-th layer and (T _i), reflects the i-th layer And a reflection distribution coefficient ( A _i ) in the i- th layer, given by the reflection coefficient ( R _i ) corresponding to
The step of obtaining the charged particle intensity coefficient map ( M _i ) of the i-th layer includes:
Using the pattern area density of each material (alpha _i) included in the i th layer obtained from the design data of the i-th layer (IN _i), wherein the reflection coefficient, and the diffusion distribution coefficient, Obtaining a first reflected charged particle intensity coefficient (E1) indicative of a proportion of charged particles reflecting in the i- th layer;
And the pattern area density, and the transmission coefficient, the diffusion and distribution coefficient, i -1-th layer below charged in layer particle intensity coefficient map (M _{i -1)} and by using the i -1-th layer Obtaining a second reflected charged particle intensity coefficient (E4) indicating the proportion of charged particles reflected from the following layers;
And a step of obtaining a sum (E1 + E4) of the first and second reflected electrified particle strength coefficients. In claim 6,
The method of manufacturing a semiconductor device, wherein the step of generating the corrected exposure data and the step of obtaining the charged particle intensity coefficient map are performed for each layer from the substrate side toward the upper layer.