JP5810926B2 - Sample evaluation method and sample evaluation program - Google Patents

Description

  The present invention relates to a sample evaluation method and a sample evaluation program.

  When stress is applied to the crystal material, lattice distortion occurs, and the band structure of the crystal material changes. In particular, since the element characteristics of an electronic device such as a transistor are greatly influenced by lattice distortion, grasping the stress of the strain source and its position is useful in the analysis of the electronic device.

  There are various methods for evaluating the strain in the sub-micron region, but none of them is sufficient for locating the strain source.

  For example, in the convergent beam electron diffraction (CBED) using a scanning transmission electron microscope (STEM), the stress of the strain source and the position thereof cannot be grasped at all.

  It has also been proposed to obtain stress using the fact that a split HOLZ (High Order Laue Zone) line or ZOLZ (Zero-th Order Laue Zone) image, which is a type of CBED image, changes according to crystal distortion. ing.

  However, this method uses the STEM image of the electronic device, so if the user refuses to provide the STEM image for reasons such as preventing the structure of the electronic device from leaking, the position of the strain source is specified. Can not do it.

JP 2007-093344 A JP 2009-2748 A

  An object of the sample evaluation method and the sample evaluation program is to evaluate a sample without knowing the structure of the sample.

  According to one aspect of the following disclosure, a step of dividing a sample into a plurality of sections, a step of acquiring an electron beam diffraction image of the sample in at least one or more sections, and using the electron beam diffraction image Calculating the strain of the sample in the section from which the electron beam diffraction image was acquired, and comparing the geometric patterns of each of the plurality of electron beam diffraction images to thereby determine the position of the interface of the strain source in the sample. Identifying a plurality of virtual strain sources along the interface, calculating a distance between the imaging point of the electron beam diffraction image and the interface, and using the strain and the distance Thus, there is provided a sample evaluation method including a step of obtaining, for each virtual strain source, an index for a physical property value to be possessed by a plurality of the virtual strain sources in order to realize the strain.

  According to the following disclosure, an electron beam diffraction image of a sample is used to obtain an index for a physical property value that a plurality of virtual strain sources should have, and a microscopic image of the sample is not used. Therefore, the user can ask other people to evaluate the sample with peace of mind without fearing that the structure of the sample will flow out to others.

FIG. 1A is a configuration diagram illustrating an example of a hardware configuration of a sample evaluation apparatus according to the present embodiment, and FIG. 1B is a diagram illustrating another hardware configuration of the sample evaluation apparatus according to the present embodiment. It is a block diagram which shows an example. FIG. 2 is a cross-sectional view showing an example of a sample to be evaluated. FIG. 3 is a flowchart (part 1) of the sample evaluation method according to the present embodiment. FIG. 4 is a flowchart (part 2) of the sample evaluation method according to the present embodiment. FIG. 5 is a schematic diagram showing the processing content of step S1 of the sample evaluation method according to the present embodiment. 6A to 6D are diagrams illustrating examples of CBED images acquired by the sample evaluation method according to the present embodiment. FIG. 7 is a diagram showing a display example in step S3 of the sample evaluation method according to the present embodiment. FIG. 8A is a CBED image of silicon, and FIG. 8B is a CBED image of cobalt silicide. FIG. 9A is a diagram schematically showing the position of the interface specified in step S4 of the sample evaluation method according to this embodiment, and FIG. 9B is a diagram of the sample evaluation method according to this embodiment. FIG. 6 is a diagram schematically showing the processing content of step S5. FIG. 10 is a diagram schematically showing the processing content of step S6 of the sample evaluation method according to the present embodiment. FIG. 11 is a schematic diagram for explaining a distortion evaluation index. FIG. 12A is a graph showing the relationship between the distance from the surface of silicon and the strain, and FIG. 12B is an image visualized by simulation of how the silicon substrate is pulled by the stress of the strain source. . FIG. 13 is a diagram illustrating an example of a fitting curve in step S7 of the sample evaluation method according to the present embodiment. FIG. 14 is a diagram schematically showing the processing content of step S8 of the sample evaluation method according to the present embodiment. FIG. 15 is a diagram schematically showing the processing content of step S11 of the sample evaluation method according to the present embodiment. FIG. 16 is a diagram schematically showing the processing content of step S12 of the sample evaluation method according to the present embodiment.

  In this embodiment, the position of the strain source of the sample and the magnitude of the stress at the strain source are obtained as follows.

  FIG. 1A is a configuration diagram illustrating an example of a hardware configuration of the sample evaluation apparatus according to the present embodiment.

  The sample evaluation apparatus 1 is an electronic computer such as a personal computer, and includes a monitor 2, an input unit 3 such as a keyboard, and a control unit 4.

  Among these, the control unit 4 includes a calculation unit 4a such as a CPU and a storage unit 4b such as a RAM. In actual use, the sample evaluation program 6 recorded in the storage medium 5 is stored in the storage unit 4b. The sample evaluation program 6 is developed by the calculation unit 4a.

  A user who wishes to evaluate a sample may perform the evaluation by himself using the sample evaluation apparatus 1 as described later, or may request an evaluation from another person who owns the sample evaluation apparatus 1.

  Note that the hardware configuration is not limited to this.

  FIG. 1B is a configuration diagram illustrating another example of the hardware configuration of the sample evaluation apparatus according to the present embodiment.

  In this example, the control unit 4 is connected via a network system 7 to another person's server 9 that provides a sample evaluation service. In this case, the sample evaluation program 6 is developed on the server 9, and the server 9 executes the sample evaluation program 6. Then, the user obtains the execution result of the sample evaluation program 6 from another person via the network system 7.

  In this example, the control unit 4 is connected via a network system 7 to another person's server 9 that provides a sample evaluation service. In this case, the sample evaluation program 6 is developed on the server 9, and the server 9 executes the sample evaluation program 6. Then, the user obtains the execution result of the sample evaluation program 6 from another person via the network system 7.

  FIG. 2 is a cross-sectional view showing an example of a sample to be evaluated.

  This sample P is a part of a MOS transistor formed on the silicon substrate 10. In the silicon substrate 10, a silicon oxide film is embedded as an element isolation insulating film 11, and a cobalt silicide film 12 is formed on the surface of the silicon substrate 10. The cobalt silicide film 12 is connected to the aluminum wiring 14 through the tungsten plug 13.

  As a strain source for distorting the silicon substrate 10 in such a sample P, there are a cobalt silicide film 12 and a tungsten plug 13.

  However, some users who ask others to evaluate the sample P do not want to clarify the positions of strain sources such as the cobalt silicide film 12 and the tungsten plug 13 in order to prevent the structure of the sample P from flowing out to others. There is.

  In the present embodiment, even when the strain source is not clear as described above, the position of the strain source and the magnitude of the stress of the strain source are obtained as follows.

  3 and 4 are flowcharts of the sample evaluation method according to the present embodiment.

  Each step of the flowchart is performed by the control unit 4 of FIG. 1A based on the above-described sample evaluation program 6 except for those described as being performed by the user below. When the network system 7 is used as shown in FIG. 1B, the server 9 executes the sample evaluation program 6.

  In the first step S1, as shown in FIG. 5, the user sets a rectangular evaluation region 20 on the cross section of the sample P. The evaluation area 20 is an area where the user wants to evaluate distortion in the silicon substrate 10.

  Further, it is preferable that an xy orthogonal coordinate system is set on the cross section of the silicon substrate 10 and the evaluation region 20 is set so that each side is parallel to the coordinate axis.

  In this step, the user further divides the evaluation area 20 into a plurality of square sections Q. The size of each section Q is arbitrary.

  Further, since this step is performed by the user himself and not by another person who performs the evaluation service, the sample P does not flow out to another person.

  Next, it moves to step S2 and the control part 4 acquires the CBED image of the silicon substrate 10. FIG. The CBED image is an example of an electron diffraction image.

As shown in FIG. 5, when the photographing point for photographing the CBED image is represented by T _k (k = 1, 2,...), Only one photographing point T _k needs to be set in one section Q. Further, it is not necessary to set the shooting point T _k for all the sections Q, and it is sufficient to set the shooting point T _k for at least one section Q.

6A to 6D are diagrams showing examples of the CBED images acquired in this step, and are CBED images at the photographing points T _{1 to} T ₄ , respectively. In this example, a CBED image of silicon in the <110> direction is illustrated.

  As can be seen from FIGS. 6A to 6D, the geometric pattern of the CBED image varies depending on the location. This is because the amount of strain of silicon changes depending on the distance from the strain source and the like, and the phase and intensity of each of the transmitted wave and diffracted wave of incident electrons change.

  Since the user takes the CBED image itself, the execution of this step does not cause the sample P to flow out to others.

Then, the flow proceeds to step S3, the control unit 4 calculates the local distortion C in the x direction of the silicon in the compartment Q shooting point T _k of CBED image of said it belongs.

  As described above, the difference in distortion appears in the difference in the geometric pattern of the CBED image. A method for calculating the strain from the geometric pattern of the CBED image has already been established, and in this embodiment, the strain C is calculated using simulation software using dynamic diffraction theory.

  In addition, after calculating the distortion C, the control unit 4 may display the section Q on the monitor 2 in different colors according to the magnitude of the distortion C. FIG. 7 shows a display example. In FIG. 7, the distortion C of the section Q for which the CBED image is not acquired is also shown for reference.

  Subsequently, the process proceeds to step S4, and the position of the interface of the strain source in the sample P is specified as follows using the CBED image acquired in step S2.

  FIG. 8A is a <110> direction CBED image of strained silicon, and FIG. 8B is a <110> direction CBED image of cobalt silicide.

  Cobalt silicide is a strain source for silicon, but the geometric pattern of the CBED image of cobalt silicide is significantly different from that of silicon as shown in FIGS. 8 (a) and 8 (b).

  Therefore, the geometric patterns of the two CBED images acquired in step S2 are compared with each other, and if they are not similar, it can be determined that the interface of the strain source exists between the sections Q in which these CBED images are captured. .

  The controller 4 determines whether or not the geometric patterns of the CBED image are similar based on the Zernike moment of the CBED image.

Here, the Zernike moment _{An, m} of the CBED image whose intensity is expressed as f (x, y) in the xy orthogonal coordinate system is defined by the following equation (1).

なお、V_n,m(x,y)は次の式（２）で定義される。

V _{n, m} (x, y) is defined by the following equation (2).

但し、n、mは｜ｍ｜≦nを満たす整数であって、n−｜ｍ｜は偶数である。そして、ρ＝√（x²+y²）、θ＝tan^-1(y/x)である。

However, n and m are integers satisfying | m | ≦ n, and n− | m | is an even number. Ρ = √ (x ² + y ² ) and θ = tan ⁻¹ (y / x).

The Zernike moment A _{n, m} is included only in exp (imθ) of the angle dependent term V _{n, m} (x, y), so its intensity | A _{n, m} | It does not depend on it.

In addition, _since the order of the Zernike moment _{An, m} is defined by _n, the number of components of the _n- th order Zernike moment _{An, m} is n or more.

In the present embodiment, it is determined whether or not the geometric patterns of the two CBED images are similar by using a complex vector formed by each component of the Zernike moments A _{n, m} of an arbitrary order n.

  The determination criterion is not particularly limited. For example, it is preferable to calculate the cosine similarity of each of the complex vectors of the two CBED images. The cosine similarity is defined by the inner product of each complex vector normalized to a unit vector. When the inner product is 1, both complex vectors are parallel, and when the inner product is 0, both complex vectors are orthogonal. That is, the cosine similarity indicates a value between 0 and 1, and the closer to 1, the more similar the both complex vectors are.

  In step S4, when the cosine similarity exceeds the threshold set by the user, it is determined that the geometric patterns of the CBED image are similar to each other. Otherwise, the geometric patterns are determined to be dissimilar.

  FIG. 9A is a diagram schematically illustrating the position of the interface F specified in this step.

  Since only the CBED image is used for specifying the interface F as described above, the structure of the sample P is not specified by others alone.

  Next, the process proceeds to step S5.

  FIG. 9B schematically shows the processing content of step S5. In this step, as shown in FIG. 9B, the control unit 4 arranges a plurality of virtual strain sources 25 along the interface F described above.

  Although the arrangement of the virtual strain sources 25 is not particularly limited, in this embodiment, one layer Q on the interface F is set as the virtual strain source 25. When it is considered that the virtual strain source 25 is outside the evaluation region 20, the virtual strain source 25 may be disposed outside the evaluation region 20.

  In the following, in order to identify each virtual distortion source 25, the label i (= 1, 2, 3,...) In the x direction and the label j (= 1, 2, 3,...) In the y direction are used.

  Since only the position of the interface F is used in this step, the structure of the sample P does not flow out to others by executing this step.

  Then, it moves to step S6.

  FIG. 10 is a diagram schematically showing the processing content of step S6.

In this step, the distances d _ks ⁽¹⁾ and d _ks ⁽²⁾ from the virtual distortion source 25 are _obtained for each of the shooting points T _k at which the CBED image is taken. At the distance d _ks , the subscript s represents the above i or j that specifies the virtual distortion source 25, and the subscript k represents the photographing point T _k . The subscript (1) represents the distance from the virtual distortion source 25 with the label i (= 1, 2, 3,...) In the x direction as the subscript s, and the subscript (2) is y. This represents the distance from the virtual strain source 25 with the direction label j (= 1, 2, 3,...).

Further, since only shooting point T _k and surfactant F is to utilize in this step, structure of the sample P on others by executing this step does not flow out.

  Next, the process proceeds to step S7, and the individual stress of each virtual strain source 25 is evaluated. In calculating the stress, the following strain evaluation index ξ is used.

  FIG. 11 is a schematic diagram for explaining the strain evaluation index ξ.

  As shown in FIG. 11, a point M that is separated from the virtual strain source 25 by a distance d in the depth direction inside the silicon substrate 10 is assumed. In this case, the inventor of the present application has found that the strain C generated at the point M due to the virtual strain source 25 can be empirically expressed by the following equation (3).

式（３）において、Aは定数であり、C₀は界面Fにおけるシリコンの歪みである。

In equation (3), A is a constant and C ₀ is the strain of silicon at the interface F.

  Ξ is expressed by the following equation (4).

式（４）において、σ_jは応力、E_jはヤング率、ρ_jはポアソン比を示す。なお、式（４）では材料の相違を添え字j（＝０、１）で表しており、０が付された記号はシリコンの物性値を示し、１が付された記号は仮想歪み源２５の物性値を示す。

In Equation (4), σ _j is stress, E _j is Young's modulus, and ρ _j is Poisson's ratio. In the equation (4), the difference in material is represented by the subscript j (= 0, 1). The symbol with 0 indicates the physical property value of silicon, and the symbol with 1 indicates the virtual strain source 25. The physical property values of are shown.

As in equation (4), the strain evaluation index ξ is a function of physical property values of the virtual strain source 25 such as the stress σ ₁ , Young's modulus E ₁ , and Poisson's ratio ρ ₁ . The function form is a linear function of each of the stress σ ₁ and Young's modulus E ₁ of the virtual strain source 25 and depends on the square of the Poisson's ratio ρ ₁ . This is a reflection of the empirical rule that the strain C increases as the stress σ ₁ , Young's modulus E ₁ , and Poisson's ratio ρ ₁ increase in the strain evaluation index ξ.

  Further, according to Equation (3), even if the distance is the same, the strain C increases as the strain evaluation index ξ increases. From this, the strain evaluation index ξ can also be said to be an index representing the ease with which the strain C is caused by the virtual strain source 25.

  FIG. 12A is a graph representing the above-described formula (3). In this example, a graph in the case where each of cobalt silicide (CoSi), nickel silicide (NiSi), silicon oxide (SiO), and tungsten (W) is used as the material of the virtual strain source 25 is shown.

  As shown in FIG. 12A, the shape of the graph is different if the material of the virtual strain source 25 is different. This is because each physical property value in ξ defined by the equation (4) is different for each material.

FIG. 12B is an image obtained by visualizing the state in which the silicon substrate 10 is pulled by the stress σ ₁ of the strain source 25 by simulation.

Here, for simplification, it is assumed that the material of the virtual strain source 25 is silicon. In this case, since the silicon substrate 10 and the virtual strain source 25 are formed of the same type of material, the Young's moduli E ₀ and E _{1 of both} are the same. For the same reason, Poisson's ratios ρ ₀ and ρ ₁ are also the same.

Further, in order to simplify the equation (4), the relative stress σ is defined as σ ₁ / σ ₀ . As is clear from this definition, the relative stress σ is assumed that the actual stress σ ₁ of the virtual strain source 25 and all the sections Q including the virtual strain source 25 are made of the same material such as silicon. It is a ratio with the stress σ ₀ of the virtual strain source 25 in this case, and is a relative value based on the stress σ ₀ .

  In this case, since ξ = σ from Equation (4), Equation (3) becomes

となる。

It becomes.

Here, the distance d has been calculated in step S6, and the distortion C has been calculated in step S3. Furthermore, if the CBED image of the section Q having the interface F as one side is acquired in step S2, the distortion C ₀ based on the CBED image has also been calculated in step S3.

Therefore, while using these d, C, and C ₀ as known parameters, is calculated by successive approximations the combination of constant A and the relative stress σ satisfying formula (5). As the successive approximation algorithm, for example, a single-objective optimization algorithm such as the Simplex method can be used.

  Table 1 below is an example of the distance d and distortion C used for the calculation.

表１の数値を利用して実際に計算を行うと、A=１０４．４９、σ＝２が式（４）を満足する最適な解であることが分かった。図１３は、これらの解で得られたフィッティング曲線である。

When the calculation was actually performed using the numerical values in Table 1, it was found that A = 104.49 and σ = 2 were the optimum solutions satisfying Expression (4). FIG. 13 shows fitting curves obtained from these solutions.

When the control unit 4 performs such calculation for all the virtual strain sources 25, in this step S9, the individual relative stress σ and the constant A of these virtual strain sources 25 are obtained. Hereinafter, the relative stress σ of the s-th (s = i or j) virtual strain source 25 is represented by σ _s .

Thus, the relative stress σ _s that each virtual strain source 25 should have in order to realize the strain C obtained in step S3 is obtained for each virtual strain source 25.

Next, the control unit 4 calculates the value of the strain evaluation index ξ by substituting the relative stress σ _s into the equation (4). Thus, the strain evaluation factor ξ that each virtual strain source 25 should have in order to realize the strain C determined in step S3 is determined for each virtual strain source 25.

In FIG. 11 and Formula 1, only the distance d in the y direction is considered for simplification. However, in step S6, the distances d _ks ⁽¹⁾ , all the virtual distortion sources 25 with respect to one photographing point T _k . d _ks ⁽²⁾ is being sought. Therefore, a combination of the constant A and the relative stress σ satisfying the equation (4) may be calculated using each distance d _ks and the above C and C ₀ .

  Further, the calculation method is not limited to the successive approximation method such as the Simplex method, and may be calculated by other numerical calculation such as a finite element method (FEM) method.

  It should be noted that the structure of the sample P does not flow out to others just by performing numerical calculations in this way.

  Next, the process proceeds to step S8.

  FIG. 14 is a schematic diagram for explaining the processing content of step S8.

  In this step, the control unit 4 color-codes each virtual strain source 25 according to the value of the strain evaluation index ξ and displays it on the monitor 2.

  As described above, the strain evaluation index ξ is an index representing the ease with which the strain C is caused by the virtual strain source 25. Therefore, by color-coding each virtual distortion source 25 as in this step, the user can visually recognize the position of the virtual distortion source 25 that generates a strong distortion.

  Further, in order to emphasize a strong strain source, the virtual strain source 25 having a value of ξ of, for example, 30% or more than that of the adjacent virtual strain source 25 may be highlighted.

  Further, since it is difficult for another person to grasp the specific structure of the sample P only with the strain evaluation index ξ, the sample P does not flow out to the other person in this step.

  Next, the process moves to step S9, and the user determines whether or not the virtual strain source 25 is a single layer in an actual device.

  If it is determined that the layer is a single layer (YES), the process proceeds to step S10.

  In step S9, it is assumed that the material of the virtual strain source 25 is silicon. However, in an actual device, the virtual strain source 25 may be formed of a material different from silicon, such as cobalt silicide.

Therefore, in step S10, the control unit 4 uses the physical property value of the material actually used as the virtual strain source 25 as follows, and the control unit 4 calculates the relative stress of each virtual strain source 25 from the following equation (6). to calculate the σ ^_'s.

なお、式（６）において、E₀はシリコンのヤング率（１３０．２）であり、ρ₀はシリコンのポアソン比（０．２８）である。そして、E_sは、仮想歪み源２５として実際に使用される材料のヤング率であり、ρ_sは当該材料のポアソン比である。これらの物性値E₀、ρ₀、E_s、ρ_sは、入力部３（図１（ａ）、（ｂ）参照）を介してユーザが制御部４に入力する。

In Equation (6), E ₀ is the Young's modulus (130.2) of silicon, and ρ ₀ is the Poisson's ratio (0.28) of silicon. E _s is the Young's modulus of the material actually used as the virtual strain source 25, and ρ _s is the Poisson's ratio of the material. These physical property values E ₀ , ρ ₀ , E _s , and ρ _s are input to the control unit 4 by the user via the input unit 3 (see FIGS. 1A and 1B).

Equation (6) shows that the actual relative stress σ ^′ _s of the virtual strain source 25 is (E ₀ / E _s ) {(1−ρ _s ² ) / of the relative stress σ _s when silicon is used as the material. (1-ρ ₀ ² )} times. This is because the strain C is proportional to σ _s (E ₁ / E ₀ ) {(1-ρ ₀ ² ) / (1-ρ ₁ ² )} in the equations (3) and (4).

In this step, by substituting the values E _s and ρ _s into the formula (6), the control unit 4 uses the material in the actual device as the material of the virtual strain source 25 to determine the relative value of each virtual strain source 25. Calculate the stress σ ^' _s .

Similarly to step S8, since it is difficult for another person to grasp the specific structure of the sample P only with the relative stress σ ^′ _s , the sample P does not flow out to the other person in this step.

  On the other hand, if it is determined in step S9 that the layer is not a single layer (NO), in the actual device, the virtual strain source 25 has a multilayer structure.

  Therefore, in this case, the user operates the input unit 3 to input the number of layers of the virtual strain source 25 to the control unit 4.

  Next, the process proceeds to step S11.

  FIG. 15 is a schematic diagram for explaining the processing content of step S11. In this example, a case where one virtual strain source 25 has a three-layer structure of first to third layers 25a to 25c will be described.

  Note that the dotted line in FIG. 15 is a line that schematically represents a state in which strain propagates from the upper layer toward the lower layer.

In this step, the relative stresses σ _{a to} σ _c of the first to third layers 25a to 25c are obtained as follows.

First, for simplification, it is assumed that the material of each of the first to third layers 25a to 25c is silicon. In this case, from the analogy with the equation (5) assuming that the strain source is silicon, the relative stress σ _{a to} σ _c propagates to the silicon substrate 10 and the silicon substrate at the point M as shown in the following equation (7). Assume that 10 causes distortion C.

なお、h_a、h_bはそれぞれ第１の層２５ａと第２の層２５ｂの厚さを示す。また、C^' _a〜C^' _cは、それぞれ第１〜第３の層２５ａ〜２５ｃの下面の歪みである。そして、dは、シリコン基板１０の表面と点Mとの距離であり、Aは、式（１）と同一の定数である。

_Here , h _a and h _b indicate the thicknesses of the first layer 25a and the second layer 25b, respectively. C ^′ _{a to} C ^′ _c are distortions of the lower surfaces of the first to third layers 25a to 25c, respectively. D is the distance between the surface of the silicon substrate 10 and the point M, and A is the same constant as in equation (1).

  In equation (7), the distance d has been calculated in step S6, and the distortion C has been calculated in step S3. The constant A has been calculated in step S9.

Accordingly, the unknown parameters in the equation (7) are the relative stresses σ _{a to} σ _c , the strains C ^′ _{a to} C ^′ _c , and the constants A _a and A _b . In this step, these unknown parameters σ _{a to} σ _c , C ^′ _{a to} C ^′ _c , and constants A _a and A _b are calculated by the Simplex method or the finite element method.

Note that it is difficult to specify the structure of the sample P only with these unknown parameters σ _{a to} σ _c , C ^' _{a to} C ^' _c , and constants A _a and A _b , The structure of the sample P does not flow out.

  Next, the process proceeds to step S12.

  FIG. 16 is a schematic diagram for explaining the processing content of step S12.

In this step, the control unit 4 obtains the value of the evaluation index ξ for each of the first to third layers 25a to 25c by substituting the relative stresses σ _{a to} σ _c into the equation (4), and the evaluation index Each of the first to third layers 25a to 25c is color-coded according to the value of ξ and displayed on the monitor 2.

  By color-coding in this way, the user can visually recognize which one of the first to third layers 25a to 25c actually has a strong strain source.

However, the stresses σ _{a to} σ _c calculated in the above step S11 are values when the material of each of the first to third layers 25a to 25c is silicon as described above. Therefore, when the first to third layers 25a to 25c are formed of a material other than silicon, the stresses σ _{a to} σ _c are different from the values calculated in step S11 .

Therefore , using the physical property values of the materials actually used for each of the first to third layers 25a to 25c, the control unit 4 uses the analogy with the expression (6) to obtain the following expression (8). Based on this, the relative stresses σ ^′ _{a to} σ ^′ _c of the first to third layers 25a to 25c are calculated.

但し、m=a,b,c。

However, m = a, b, c.

In Equation (8), E ₀ is the Young's modulus (130.2) of silicon, and ρ ₀ is the Poisson's ratio (0.28) of silicon.

E _m (m = a, b, c) is the Young's modulus of the material actually used for each of the first to third layers 25a to 25c, and ρ _m is the Poisson's ratio of the material. . In this step, by substituting these values E _m and ρ _m into the equation (8), the relative stress σ ^′ _m (m = m = _m ) of the first to third layers 25a to 25c when an actual material is used. a, b, c) are calculated.

Thereby, the user can know the relative stress σ ^′ _m of each of the first to third layers 25 a to 25 c and can quantitatively evaluate the virtual strain source 25.

Even if the stress σ ^′ _m is calculated in this way, it is difficult to specify the structure of the sample P only by this, and therefore, the structure of the sample P does not flow out to others by executing this step.

  The basic steps of the sample evaluation method according to the present embodiment are thus completed.

  In the above-described embodiment, since the strain evaluation index ξ of the virtual strain source 25 can be known without using a microscopic image such as an STEM image of the sample P, the structure of the sample P flows out to others. The user who is concerned about the sample can evaluate the sample with peace of mind.

  Further, since the strain evaluation index ξ is an index representing the ease with which the strain C is caused by the virtual strain source 25, it is qualitatively evaluated at what position of the sample P and how strong the strain source is. You can also.

  The following additional notes are disclosed for each embodiment described above.

(Supplementary note 1) dividing the sample into a plurality of sections;
Obtaining an electron diffraction image of the sample in at least one of the compartments;
Using the electron diffraction image to calculate strain of the sample in the section from which the electron diffraction image was acquired;
Identifying the position of the interface of the strain source in the sample by comparing each geometric pattern of the plurality of electron diffraction images; and
Arranging a plurality of virtual strain sources along the interface;
Calculating a distance between the imaging point of the electron diffraction image and the interface;
Using each of the strain and the distance to obtain an index for a physical property value that the plurality of virtual strain sources should have in order to realize the strain for each virtual strain source;
A sample evaluation method characterized by comprising:

  (Additional remark 2) The said parameter | index is a function of the said physical-property value, Comprising: This physical-property value is either a stress, a Young's modulus, and Poisson's ratio, The sample evaluation method of Additional remark 1 characterized by the above-mentioned.

  (Supplementary note 3) The sample evaluation method according to supplementary note 1 or supplementary note 2, further comprising a step of displaying each of the plurality of virtual strain sources in different colors according to the value of the index.

(Supplementary note 4) dividing the virtual strain source into a plurality of layers;
The sample evaluation method according to any one of appendix 1 to appendix 3, further comprising: obtaining the index for each of the plurality of layers.

  (Supplementary Note 5) The step of specifying the position of the interface is by determining that the interface exists between the sections where the two geometric patterns are photographed when the two geometric patterns are not similar to each other. The sample evaluation method according to any one of appendix 1 to appendix 4, wherein the sample assessment method is performed.

  (Additional remark 6) It further has the step which calculates | requires the relative value of the stress which the said several virtual strain source should have in order to implement | achieve the said distortion for every said virtual strain source using the said strain and the said distance. The sample evaluation method according to any one of Appendix 1 to Appendix 5.

  (Supplementary Note 7) The relative value is a value based on the stress of the virtual strain source when it is assumed that all the sections and the virtual strain source are formed of the same material. The sample evaluation method according to appendix 6.

(Supplementary Note 8) In the step of acquiring the electron beam diffraction image, the electron beam diffraction image in the section having the interface as one side is acquired,
In the step of calculating the strain, the strain of the sample in the section having the interface as one side is calculated,
In the step of obtaining the index, the distortion of the sample in the section having the interface as one side, the distortion of the sample in the section from which the electron diffraction image is acquired, and the distance are set as known parameters, and sequentially The sample evaluation method according to any one of appendix 1 to appendix 7, wherein the index is obtained by approximation.

(Additional remark 9) The step which acquires the electron beam diffraction image of at least 1 or more of the above-mentioned division among the samples provided with the evaluation field divided into a plurality of divisions,
Using the electron diffraction image to calculate strain of the sample in the section from which the electron diffraction image was acquired;
Identifying the position of the interface of the strain source in the sample by comparing each geometric pattern of the plurality of electron diffraction images; and
Arranging a plurality of virtual strain sources along the interface;
Calculating a distance between the imaging point of the electron diffraction image and the interface;
Using each of the strain and the distance to obtain an index for a physical property value that the plurality of virtual strain sources should have in order to realize the strain for each virtual strain source;
Sample evaluation program that causes a computer to execute.

(Supplementary note 10) dividing the virtual strain source into a plurality of layers;
The sample evaluation program according to appendix 9, further comprising the step of obtaining the index for each of the plurality of layers.

DESCRIPTION OF SYMBOLS 1 ... Sample evaluation apparatus, 2 ... Monitor, 3 ... Input part, 4 ... Control part, 4a ... Calculation part, 4b ... Memory | storage part, 5 ... Storage medium, 6 ... Sample evaluation program, 7 ... Network system, 9 ... Server, DESCRIPTION OF SYMBOLS 10 ... Silicon substrate, 11 ... Element isolation insulating film, 12 ... Cobalt silicide film, 13 ... Tungsten plug, 14 ... Aluminum wiring, 20 ... Evaluation area | region, 25 ... Virtual strain source, 25a-25c ... 1st-3rd layer .

Claims (5)

Dividing the sample into a plurality of compartments;
Obtaining an electron diffraction image of the sample in at least one of the compartments;
Using the electron diffraction image to calculate strain of the sample in the section from which the electron diffraction image was acquired;
Identifying the position of the interface of the strain source in the sample by comparing each geometric pattern of the plurality of electron diffraction images; and
Arranging a plurality of virtual strain sources along the interface;
Calculating a distance between the imaging point of the electron diffraction image and the interface;
Using each of the strain and the distance to obtain an index for a physical property value that the plurality of virtual strain sources should have in order to realize the strain for each virtual strain source;
A sample evaluation method characterized by comprising:   The sample evaluation method according to claim 1, wherein the index is a function of the physical property value, and the physical property value is any one of stress, Young's modulus, and Poisson's ratio.   The sample evaluation method according to claim 1, further comprising a step of displaying each of the plurality of virtual strain sources in a color-coded manner according to the value of the index. Dividing the virtual strain source into a plurality of layers;
The sample evaluation method according to any one of claims 1 to 3, further comprising a step of obtaining the index for each of the plurality of layers. Obtaining an electron diffraction image of at least one or more of the sections of the sample having an evaluation region divided into a plurality of sections;
Using the electron diffraction image to calculate strain of the sample in the section from which the electron diffraction image was acquired;
Identifying the position of the interface of the strain source in the sample by comparing each geometric pattern of the plurality of electron diffraction images; and
Arranging a plurality of virtual strain sources along the interface;
Calculating a distance between the imaging point of the electron diffraction image and the interface;
Using each of the strain and the distance to obtain an index for a physical property value that the plurality of virtual strain sources should have in order to realize the strain for each virtual strain source;
Sample evaluation program that causes a computer to execute.

Images