JP2718396B2 - Line width measurement method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a line width measuring method,
In particular, the present invention relates to a line width measuring method effective for line width measurement of an automatic pattern in a semiconductor manufacturing process.

[0002]

2. Description of the Related Art In a semiconductor manufacturing process, in order to improve the reliability and yield of a product, a pattern line width of a photoresist or a formed film after etching is usually measured to check the quality. Recently, semiconductor patterns have been miniaturized, and for the purpose of this measurement, the resolution has been insufficient with a conventionally used optical microscope, and an electron microscope having a high resolution has come to be used. Further, development of automatic devices for line width measurement using this electron microscope has been actively carried out. A typical method for measuring the dimensions of these devices is a peak detection method using a peak of a one-dimensional waveform obtained from an observation pattern as a pattern edge (`` High-precision line width measurement device '', Semiconductor World 1984.
12, pp87-93) or a straight line approximation method that directly approximates the slope line and the baseline of a one-dimensional waveform and uses the intersection as an edge ("Ultra LSI Micro Dimension Measurement System MEA-3000", Semic
onductor World 1985.1, pp120-128).

[0003]

The conventional method described above is based on the premise that a pattern dedicated to length measurement is used for length measurement. That is, in the peak detection method, one peak exists for the one-dimensional waveform corresponding to one edge of the pattern, and in the direct approximation method, one peak exists for the one-dimensional waveform corresponding to one edge of the pattern. It is assumed that one slope line and one baseline can be approximated. However, such an assumption can be made only for a length measurement pattern having a simple shape. On the other hand, recently, there has been a strong demand for measuring the length of an actual circuit pattern having a complicated shape for the purpose of failure analysis. In this case, since a plurality of circuit patterns overlap, the one-dimensional waveform near the edge often becomes a somewhat complicated waveform having a plurality of peaks and valleys. Therefore, for example, in the peak detection method, it is not known which peak corresponds to the pattern edge to be measured, and it is often difficult to find the baseline by the linear approximation method. SUMMARY OF THE INVENTION An object of the present invention is to provide a line width measuring method that solves the above-described problems in the conventional technology and enables the line width of a circuit pattern having a complicated edge waveform to be measured.

[0004]

In order to achieve the above object, a line width measuring method according to the present invention comprises: a first step of scanning a sample with an electron beam to detect a signal of a first pattern; A second step of storing the obtained signal of the first pattern in a memory; a third step of detecting a signal of a second pattern different from the first pattern; A fourth step of changing the standard deviation value of the Gaussian filter to obtain respective approximate waveforms, and a fifth step of obtaining coordinates of an inflection point of each approximate waveform in order to match each signal of the second pattern; A pair of inflection points corresponding to each of the approximate waveforms of the first and second patterns are detected by collation, a corresponding point of both patterns is obtained, and a line width to be measured set in the first pattern is measured. Measured corresponding to the start and end points of Detecting the start and end points of the second pattern is a turn, it is characterized by having a sixth step of the following I line width difference.

[0005]

In the line width measuring method according to the present invention, global matching and detailed matching of a complicated one-dimensional waveform are dynamically performed, so that an edge position is determined by an absolute value of a waveform or a fixed magnitude relationship. , Edge detection corresponding to waveforms of various shapes becomes possible. Also, for a one-dimensional waveform having noise, by appropriately setting the lower limit values Σ _g and σ _g of the filter size, peaks and valleys smaller than Σ _g and σ _g are not detected. There is an effect that it becomes possible. Further, when the one-dimensional waveform is deformed in the variable direction by interpolating the correspondence between the feature positions by polynomial approximation, for example, when the magnification of the electron microscope changes or the edge cross-sectional shape of the pattern to be measured changes. Etc.
There is also an effect that the position can be detected. In addition, the inflection point is obtained for each approximate waveform at the time of collation. However, since the inflection point has a very short staying time of the noise, there is an advantage that the influence of the noise is negligible.

[0006]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 3 shows the overall configuration of a measuring device employing the waveform matching method according to the present invention. In the drawing, reference numeral 1 denotes an A / D converter which samples an analog image signal 1s of a pattern to be measured or a reference pattern from an electron microscope (SEM) 9 at an appropriate timing and converts it into a digital image signal 1d. An address generation circuit for generating the address 2a of the image memory 3 using the scan synchronization signal 2s of the above, 4 is a memory for storing data relating to the reference pattern, and 5 is a memory for storing data relating to the pattern to be measured. A memory for storing a calculation procedure according to the present invention; and a computer for processing data.
(CPU), 8 denotes a display device for displaying the measurement result as numerical data or the like, 7a denotes an address bus, and 7d denotes a data bus. First, an outline of the measurement procedure of the present embodiment will be described. A reference pattern prepared in advance is scanned by an electron microscope, its analog image signal is digitized, and stored in the image memory 3. Next, according to the calculation procedure of the memory 6,
The image is processed by the computer 7 and the edge position of the reference pattern and necessary data are stored in the reference pattern data memory 4. At the time of measurement, the required data of the measured pattern is stored in the measured pattern data memory 5 in substantially the same procedure. Next, pattern matching is performed using these data, the edge position of the test pattern corresponding to the edge position of the reference pattern is detected, and the result is displayed on the display device 8.

FIG. 1A shows the procedure for processing a reference pattern. In the first step, 1-dimensional waveform position is predetermined reference pattern from the image memory 3, line profile: taking out the _{F (X) (X s ≦} X ≦ X a). A in FIG. 4 (a)
Shows an example of this waveform. In the second step, an approximate waveform G (X: Σ) obtained by applying a Gaussian filter having a standard value difference Σ represented by the following equation is obtained.

(Equation 1) In the third step, position coordinates satisfying d ² G / dX ² = 0 are obtained. These coordinates represent the changing points of the peaks and valleys of the original waveform, that is, the positions of the inflection points, and are herein referred to as reference characteristic positions. Required standard width difference (Σ _s Σ Σ Σ _a )
Only the second, repeat third step, the reference feature positions for each standard deviation _{(X Σ1, ..., X Σi} , ..., X Σn) Request.

Naturally, the number n of the reference feature positions at each standard deviation
Is different. The waveform of the thin line drawn on A of FIG. 4A is obtained when Σ is changed, and indicates a characteristic position. In the figure, the valley positions are also indicated by small circles for reference. From the figure, it can be seen that the overall / detail of the waveform can be described by the magnitude of the standard deviation. Note that the signal obtained from the low-acceleration electron microscope used for this type of measurement is considerably S / S
Although N is bad, it is possible to obtain a noise-free waveform by applying a Gaussian filter having an appropriate standard deviation Σ. Therefore, the minimum value sigma ₁ of consideration standard deviation to the value. In the fourth step, a human inputs an edge position (Xd) on the original waveform. In the fifth step, data on the reference pattern collected so far, that is,
Reference feature positions for each standard deviation _{(X Σ1, ..., X Σi} , ...,
_XΣn ), the number n thereof, and the edge position (Xd) are stored in the memory 4.
To be stored.

Next, the procedure of data processing when a pattern to be measured is input will be described. Also in this case, the processing is executed in substantially the same procedure as in the case of the reference pattern. That is, an analog image signal of the pattern to be measured from the electron microscope is digitized and stored in the image memory 3. Then, the image data is processed by the computer 7 in accordance with the calculation procedure of the memory 6. The calculation procedure is shown in FIG. In the first step, a one-dimensional waveform (line profile) f (x) ( _xs ≦ x
≦ x _a ). A 'in FIG. 4 (b) shows an example of this input waveform. Normally, even if the waveform sampling position is the same as the reference pattern due to slight differences in the charge state and magnification of the pattern, Deformation exists. In the second step,
Approximate waveform g with Gaussian filter with standard deviation σ
(x: σ) is obtained. In the third step, d ² g / dx ² =
The position coordinates satisfying 0 are obtained. Here, these coordinates will be referred to as feature positions. And the required standard deviation
Repeat the second and third steps for (σ _s ≦ σ ≦ σ _n )
The feature position ( _xσ1 ,..., _Xσi ,..., _Xσm ) for each standard deviation is obtained. Naturally, the number m of characteristic positions differs depending on each standard deviation. The waveform of the thin line drawn on A 'in FIG. 4 (b) is obtained when σ is changed. In the figure, the valley positions are also indicated by small white circles for reference. The minimum value σ _B of the standard deviation to be considered is such that the noise of this waveform is removed. In the fifth step,
The data on the pattern to be measured obtained so far, that is, the characteristic positions (x _σ1 ,..., X _σi,.
x _σm ) and the number m thereof are stored in the memory 5.

Next, an outline of a procedure for comparing the data of the pattern to be measured with the reference pattern obtained by the above procedure will be described with reference to FIGS. 4 (a) and 4 (b). As described above, when the standard deviation value of the Gaussian filter is large, it represents the general shape of the waveform (upper part of the figure), and when it is small, it represents a detailed shape close to the sampled waveform (lower part of the figure). Therefore, the matching of the reference characteristic position of the reference waveform with the characteristic position of the input waveform is performed starting from the case of the large standard deviation value. A specific method of detecting the matching feature position will be described later.
The double circles in FIGS. 4A and 4B are the first matching feature positions. Next, the matching feature positions are collated again in detail. That is, the waveform starting point ~ a and the waveform starting point ~ a ', a ~ b
The matching is performed while reducing the standard deviation value for each part between matching feature positions of the a-b ', b-c, b-c', c-waveform end points and c'-waveform end points. The start point and the end point of each partial waveform when the standard deviation value is changed are the coordinates of the characteristic position corresponding to the first coincidence characteristic position (a, b, c / a ', b', c '). Therefore, as shown by the arrows in FIGS. The result of the second verification is indicated by ●. By repeating such matching recursively up to the lower limit values 標準_R and σ _R of the standard deviation values, global matching to detailed matching can be unified.

[0011] The following is a program in which the collation procedure is performed. The language used is the C language, and the declaration of variables and data transfer descriptions irrelevant to the essence of the present invention are omitted or simplified, and do not always adhere to the grammar. _{MATCH (Σ a, Σ i,} X a, X i, σ a, σ i, x a, x i) / * Σ a, Σ i, σ a, σ i: Min Max standard standard deviation of the reference and input waveforms Value * / / * X _a , X _i / x _a , x _i : coordinates of search for minimum and maximum inflection point of reference and input waveforms * / {/ * stop condition-step 1-* / if (Σ _i = lower limit standard deviation Value) {goto end;} if (σ _i = lower limit standard deviation) {goto end;} if (no feature position between Σ _a , Σ _i , X _a , X _i ) {goto end:} if (σ _a, between sigma _i, x _a, no characteristic position between _{x i) {goto end;}} / * matching - step 2 - * / for (σ = σ n, err0 = 9999; σ ≦ σ r; σ ++ ) {for (Σ = Σ _n ; Σ ≦ Σ _F ; Σ ++) (/ * in the case of standard deviation width σ, 間, between X _n , X, and x ₀ , x ₁ of the reference and input waveforms matching feature position X [i], x [i ] and obtains the error err at that time. * / lesq (σ, Σ , X n, X 1, x 0, x i, X [i], x [i ], err); / * Store each data (l ≦ i ≦ N) of the minimum error * / if (err <err0) (σ ′ = σ; Σ ′ = Σ; X [i] X ′ [i], x [i] → x ′ [i]; err0 = err;})} if (err0 ≧ const) (/ * large error * / goto end;} / * recursive matching—step 3 * / for (i = 1; i ≦ N−1; i ++) (MATCH (Σ _n , Σ ′, X ′ [i], X ′ [i + 1], σ _n , σ ′, x ′ [ i], x '[i + 1]);} end:}

Step 1 of this program is a determination of a collation stop condition. When this program is called recursively and reaches the lower limit of the standard deviation value, the execution of this program ends. In step 2, for each combination within the range of the standard deviation value, in the subroutine lesq, the reference characteristic position and the characteristic position data previously obtained are retrieved from the memories 4 and 5, and collation is performed. The reference feature position X [i], the feature position x [i], and the error err at that time are output. Next, among the data obtained in this way, only those having the smallest error are stored with their standard deviation values Σ ′, σ ′ and coincident feature positions X ′ [i], x ′ [i]. If the minimum error itself at this time is larger than a certain value (const), the deformation between the two waveforms is too large, so that the obtained data is invalidated and the comparison at this stage is finished. In step 3, the matching feature position X '[i], X' of the result obtained in the previous stage
Using [i], the collation is performed again on the partial regions of both waveforms. At this time, the upper limits of the standard deviation values are Σ ′ and σ ′, respectively, and more detailed waveform matching is performed. The initial value of this program is Σ ₀ = Σ _s , Σ _i = Σ
_n , X _a = X _s , X ₁ = X _n , σ _n = σ _s , σ ₁ = σ _s , x ₀ = x _s ,
Let x ₁ = x ₀ .

Next, the subroutine le in the previous program
sq corresponding to, n-number of reference feature positions of the reference pattern of the standard deviation _{Σ (X Σ1, ..., X} Σi, ..., X Σn) ( provided that, X _n ≦ X
_{Σ1 <... <X Σi <...} <X Σn ≦ with X _i), m pieces of feature position of a measurement pattern of the standard deviation _{σ (x ns, ..., x} ni, ...,
x _σm ) (where x _σ ≦ x _σ1 <... <x _σi <... <x _σm ≦
The method of collating x ₁₎ will be described. Here, n pieces of reference feature positions of the reference pattern _{(X Σi, ... X Σ1,} ..., X Σn) ( Fig. 3
(a)) is subjected to coordinate transformations for n ′ points, and these are subjected to m characteristic positions (x _σ1 ,..., x _σi,.
x _σm ) (FIG. 3B) (where 3 ≦ n ′ ≦ m
in [n, m]). Assuming that the coordinate transformation equation is a linear equation of the input variable X, the polynomial T (X) for minimizing the square error δ as follows is the optimal coordinate transformation equation.

(Equation 2) A ₀ and A ₁ are constants. Note that A ₀ represents fluctuation due to shift, and A ₁ represents fluctuation due to magnification or distortion. Then, for each combination taking n ′ from the reference feature position of the reference pattern and the feature position of the pattern to be measured, this least squares method is applied to obtain δ at that time, and α · δ /
(n ') ² finds the minimum combination (α: constant). However,
n 'itself is also changed from 3 to min [n, m]. simply,
The reason why the weight coefficient α / (n ′) ² is added instead of finding the minimum of δ is that the error δ becomes smaller as n ′ becomes smaller, so the error evaluation is performed appropriately. That's why. Therefore, the above-mentioned coefficient may be another one as long as this object is realized.

FIG. 2 shows the procedure described above. In the first step, the standard deviation value is larger than X _{0 in} Σ,
X ₁ is taken out of smaller feature location. In the second step,
3 in the initial combination element variables n ', leave substituting a value larger than the error possible in variable err, n pieces of reference feature positions of the reference pattern in the third step _{(X Σ1, ..., X Σi} , ..., X
_Σn ) and n ′ feature positions (x
_σ1 ,..., _xσi ,..., _xσm ). In the fourth step, the least square errors δ and δ ′ = α · δ in this combination
/ (n ′) ² , and from the fifth step to the sixth step,
After setting err = δ 'only when δ'<err, the combination at that time is stored as output arguments X [i], x [i]. In the seventh step, it is checked whether all the combinations taking n 'have been completed, and if not completed, the process returns to the fourth step and the process is repeated in the same manner. When finished,
In the eighth step, n 'is incremented by 1 and the same operation as described above is repeated until n' = min [n, m].

FIG. 5 shows the correspondence between the final feature positions obtained by this method. If this correspondence is known, the edge position (xd) of the input waveform corresponding to the edge position (Xd) in the reference waveform can be determined by appropriate interpolation or the coordinate conversion formula T (X) in the partial waveform including the same. Easy to calculate. Then, the result is displayed on the display device 8. In the above-described embodiment, an example of collation for a one-dimensional waveform has been described. However, the present invention is not limited to this, and can be easily applied to collation of a two-dimensional image. Hereinafter, this will be described.

FIGS. 6A and 6B show a binary reference line image P.
An example of collation between (X, Y) and an input line image p (x, y) is shown. Here, the line image refers to an image drawn with a line having a width of one pixel. Such a line image is represented by a function using a coordinate L on a line from a starting point on a certain line as a parameter, X = S (L) and Y
= T (L). Therefore, each standard deviation value (Σ _s Σ Σ
≤ Σ _a ) A function obtained by applying _a Gaussian filter to the functions S (L) and T (L), and S '(L; Σ) and T'(L; Σ) are obtained. The shape of the line image represented by S ′ (L; Σ) and T ′ (L; Σ) depends on the magnitude of the standard deviation value.
It will be big or detailed. Therefore, the curvature Θ = 0 or dΘ / dL = 0 which can be calculated using the functions S ′ (L; Σ) and T ′ (L; に お け る) at each standard deviation value.
Coordinates _{L Σ1, ..., L Σi,} ..., L reference feature coordinates _.SIGMA.n, similarly determined l _.sigma.1 input line _{image, ..., l σi, ...,} l σm: if said coordinates (l line coordinates) Can be reduced to the above one-dimensional waveform collation method. Naturally, the line image is decomposed into a partial line image using the result of the first collation, and the collation in the same process is performed again on this partial line image. If the line image is a closed loop, it is necessary to allow the correspondence beyond the starting point when associating the reference feature coordinates with the feature coordinates. At that time, the perimeter of the line drawing is added to the coordinates that have been exceeded. In this way, the global shape and the detailed shape of the line image can be compared on each line, and the line image having the deformed shape can be compared.

FIG. 7 shows the correspondence between the final feature positions obtained by this method. In each of the above embodiments, a Gaussian filter is used as a filter. However, another simple averaging filter may be used as long as it can describe a general or detailed waveform.

As described above, according to the present invention, the criterion 1
In the method for detecting a specific position of the input one-dimensional signal waveform f (x) by comparing the one-dimensional waveform F (X) with the input one-dimensional signal waveform f (x), the reference one-dimensional waveform F (X) A first step of performing filtering of a filter size Σ at each position to obtain an approximate waveform G (X);
In (X)

(Equation 3) (X ₁ ,..., X _i ,..., X _n ) as reference feature coordinates, and at each position of the input one-dimensional signal waveform f (x), A third step of performing filtering to obtain an approximate waveform g (x);
In the approximate waveform g (x)

(Equation 4) A fourth step of obtaining x coordinates (x ₁ ,..., X _i ,..., X _n ) satisfying the following as feature coordinates; A fifth step, and T (X) is defined as a coordinate transformation expression represented by a polynomial of X, and the reference feature coordinates (X ₁ ,..., X _{i ,.}
And some of the features coordinates _{(x 1, ..., x i} , ..., x m) combinations of several _{(X 1, ..., X i} , ..., X h) and _{(x 1, ..., x i} , …, X
_h ), the square error

(Equation 5) When the error δ at this time is equal to or less than a certain value, the corresponding position is determined as a collation result. When the error δ exceeds the certain value, the collation result is determined as having no corresponding position. The steps are as follows.

[0019]

As described above, according to the present invention,
Even when measuring the line width of a circuit pattern having a complicated edge waveform, the start point and end point of the circuit pattern to be measured can be automatically detected according to the set start point and end point of the prepared circuit pattern. Therefore, there is an effect that the line width can be accurately measured based on these differences.

[Brief description of the drawings]

FIG. 1A is a flowchart showing a processing procedure of a reference pattern, and FIG. 1B is a flowchart showing a processing procedure of a pattern to be measured in an embodiment of the present invention.

FIG. 2 is a flowchart showing a procedure for associating a reference pattern with a pattern to be measured in one embodiment of the present invention.

FIG. 3 is a configuration diagram of a measuring device showing one embodiment of the present invention.

FIG. 4 is a diagram showing a waveform of a reference pattern and a pattern to be measured and an approximate waveform thereof according to the present invention.

FIG. 5 is a diagram showing a final correspondence between characteristic positions.

FIG. 6 shows two reference line images P (X, Y) and an input line image p
It is a figure showing an example of collation of (x, y).

FIG. 7 is a diagram showing correspondence between feature positions in FIG. 6;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... A / D converter, 2 ... Address generation circuit, 3 ... Image memory, 4 ... Reference pattern data memory, 5 ... Measured pattern data memory, 6 ... Calculation procedure data memory, 7 ... CPU, 8 ... Display device, 9 ... Electron microscope.

 ──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Hidenori Inuchi 1-280 Higashi Koikebo, Kokubunji-shi, Tokyo Inside the Central Research Laboratory of Hitachi, Ltd. (56) References JP-A-62-237307 (JP, A) JP-A-62 -73105 (JP, A)

Claims (4)

(57) [Claims] A first step of scanning a sample with an electron beam to detect a signal of a first pattern; a second step of storing the detected signal of the first pattern in a memory; A third step of detecting a signal of a second pattern different from the first pattern, and changing a standard deviation value of a Gaussian filter to match each signal of the first pattern and the second pattern. A fourth step of obtaining respective approximate waveforms, a fifth step of obtaining coordinates of inflection points of the respective approximate waveforms, and a step of obtaining inflection points corresponding to the respective approximate waveforms of the first and second patterns. A pair is detected by collation, a corresponding point of both patterns is obtained, and a start point and an end point of a second pattern which is a pattern to be measured corresponding to a start point and an end point of a line width to be measured set in the first pattern. Is detected, and the difference is Line width measuring method characterized by having a sixth step of the line width Te. 2. A one-dimensional waveform obtained by scanning a prepared reference pattern with an electron beam and a one-dimensional signal waveform obtained by scanning a pattern to be measured with an electron beam are compared. In a method of detecting two specific positions of a signal waveform and measuring a line width from a difference between the specific positions, a first filter for performing filtering of a certain filter size at each position of the reference one-dimensional waveform to obtain an approximate waveform A second step of obtaining an X coordinate of an inflection point in the approximate waveform as a reference feature coordinate; and a step of performing filtering of a certain filter size at each position of the one-dimensional signal waveform to be measured to obtain an approximate waveform. Step 3, a fourth step of obtaining the x-coordinate of the inflection point in the approximate waveform as characteristic coordinates, and a step between the reference one-dimensional waveform and each approximate waveform of the measured one-dimensional signal waveform. In order to eliminate the influence of the difference in magnification and the effect of distortion, a predetermined coefficient of a coordinate conversion equation expressed by a polynomial of X is assumed, and a coefficient that minimizes the error of the coordinate conversion equation including each coefficient is specified. And a sixth step of obtaining a corresponding set of reference feature coordinates and feature coordinates in each approximate waveform including the coefficient, and a corresponding set of both signal waveforms being obtained from the corresponding set of the approximate waveforms. Seventh step of detecting the start point and end point of the one-dimensional waveform to be measured corresponding to the start point and end point of the line width to be measured set in the one-dimensional waveform, and the difference between the end point and the start point of the one-dimensional waveform to be measured An eighth step of setting the line width of the pattern to be measured. 3. When finding a corresponding set of feature coordinates in the sixth step, the filter size is reduced for each corresponding set of partial waveforms, and the partial area of both waveforms is again reduced.
3. The line width measuring method according to claim 2, wherein collation is performed to perform more detailed waveform collation. 4. The step of obtaining the coordinates of the inflection point,
3. The line width measuring method according to claim 1, wherein an X coordinate at which a differential value of the approximated waveform becomes zero, or an X coordinate at which a second-order derivative or higher derivative becomes zero is used.