JP3284514B2 - Fine size standard structure and fine size standardization 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 standard for dimensions on the order of nanometers.

[0002]

2. Description of the Related Art For high integration and high performance of a semiconductor integrated circuit, it is very important that the characteristics of individual devices, which are constituent elements of the circuit, be well matched. For example, regarding the MOSFET gate length, 3
Only a length fluctuation of about 10% of the gate length in σ (standard deviation) is allowed. As the miniaturization progresses, the gate length decreases, and the allowable value of the fluctuation also gradually decreases. For example, when the gate length becomes 100 nm, the fluctuation needs to be 10 nm (3σ) or less. In order to achieve such a goal, it is very important to inspect the finished condition with high accuracy in the manufacturing process.
This inspection is generally called length measurement, and it is necessary to measure the length of the component of the device with high accuracy. For that measurement, of course, the measurement method is also important,
Exact length criteria are also required.

Hitherto, as a reference for the length, a linear pattern formed at regular intervals formed by machining (ruling engine) or interference exposure has been used (FIG. 2). The minimum value of the period of these patterns is about 1 μm because the former is manufactured by machining, and the maximum pitch of the latter is about 100 nm because it is necessary to secure a sufficient contrast by interference exposure. Also, in the current linear pattern, the average pitch can be determined accurately, but the width of each one varies, so that the length cannot be determined by one pitch. ~
The length is calibrated in a field of view of about several tens of μm. In the case of observing a structure in which the object to be observed becomes fine and has a size of 100 nm or less, it is necessary to observe at a magnification higher than the calibrated magnification. In such a case, the dimensions at that magnification must be calibrated by electrical extrapolation. For this reason, a shift occurs due to aberration or the like, and accurate measurement cannot be performed. In particular, when piezo is used for scanning as in a scanning probe microscope, the distortion is large and the problem is large.

Further, in the current length standard pattern, only the pitch is the standard of length, and there is no structure whose structure itself can be the standard of length. In most cases, the length actually measured is not the pitch of the pattern but the length of the structure itself. In this sense, the structure itself needs to be a length reference. For this purpose, for example, a method of forming a pattern whose line width is well controlled by an electron beam lithography technique, which is a high-precision exposure technique, can be considered. However, in actual pattern formation, there is a so-called pattern conversion difference (the pattern width changes every time the transfer process is performed), and the size is as large as about 10 to 100 nm. It is usual that the degree is different. Therefore, it is difficult to obtain a pattern whose width (length) is accurately determined by an ordinary method.

[0005]

The problem to be solved by the present invention can be applied to an area shorter than the conventional length reference, and the pattern itself can be used as the length reference instead of the pattern pitch. It is to fulfill the length criterion. If the difference between the design value of the pattern dimension (the value on the exposure pattern data) and the actual finished dimension, that is, the so-called conversion difference, can be accurately estimated, if the pattern is about 10 to 100 nm, the electron beam lithography technique will be used. Since the pattern can be easily formed, the pattern itself can be used as a dimensional reference. It is possible in principle to estimate the conversion difference by an inspection means such as an SEM. However, this conversion difference generally has a considerable dimensional fluctuation in a wafer surface or between wafers, and it is practically impossible to use it as a dimensional reference. Therefore, it is only necessary to prevent the dimensional fluctuation between the wafers and between the wafers from appearing in the dimensional reference structure. SUMMARY OF THE INVENTION An object of the present invention is to provide a solution to the problem of how to produce a dimensional standard that does not reflect line width fluctuation in a wafer surface or between wafers.

[0006]

In order to achieve the above object, the present invention provides: (1) a line and space structure comprising a first group, a second group and a third group in the longitudinal direction of a line; Are sandwiched between the first group and the second group, the widths of the lines of the first group and the second group are equal, the repetition period is different, and the width of the line of the third group is adjacent to the line having a width of 0. A fine dimension standard structure characterized by sequentially increasing from a line by a difference between a repetition period of the first group and a repetition period of the second group. (2) When forming the fine dimension standard structure on a substrate,
In the vicinity of the fine dimension standard structure, a plurality of
A line structure is formed.
The pattern width conversion difference is measured from the
Is determined, and the plurality of line widths are determined based on the pattern width conversion difference.
Correct the line width of the line structure of
A fine dimension standardization method characterized by using a structure as a dimension standard . Are the features of the invention.

In other words, in order to solve the above-mentioned problem, as shown in FIG. 1, the area formed between two types of linear patterns having different pitches and having the same line width is formed. A structure in which the line width is expressed by (the line width described above—the pitch shift between the upper and lower patterns) may be used. In this structure, for example, in the example of FIG. 1, the right pattern edge is defined by the upper pattern with the shorter pitch, and the left pattern edge is defined by the lower pattern with the longer pitch. If an appropriate number of patterns are formed as shown in the figure and the pattern is formed up to a region where (actual line width)-(pattern position shift amount due to pitch shift) is 0 or minus, an absolute pattern is formed. The zero reference of the width is obtained.
The position where the width is 0 in the pattern, that is, the position where the pattern of the joining portion disappears is the position of the width 0, and a structure in which the pitch shift amount is increased based on the position is obtained. That is, in this structure, even if the absolute value of the line width of the original line is unknown, an absolute position of 0 can be determined, and a structure in which the width increases in units of the pitch shift amount based on the position can be obtained. Therefore, it can be used as a reference for the absolute width.

In such a structure (for example, as shown in FIG. 1, the right pattern edge is defined by a pattern having a shorter upper pitch, and the left pattern edge is defined by a pattern having a longer lower pitch. The structure specified in)
If an anisotropic etching technique is used, it can be realized on a single crystal substrate. For example, as shown in FIG.
(1-10) A line pattern in the <11-2> direction is shifted on the substrate and drawn.
Large in crystal anisotropy (in this case, (111)
When etching is performed using an etching solution (the etching rate of the surface is much smaller than the etching rate of the other surface), a structure as shown in FIG. 3B is formed.
This is because a structure having a (111) plane on the side wall is formed at a normal line segment, but etching proceeds because the side wall is shifted from the (111) plane at the junction of the two patterns, and the pattern edge is formed. They recede in different directions, and the structure shown in the figure is formed.

In this structure, even if the fluctuation in the wafer surface is generally on the order of cm and mm, and the fluctuation is smaller in a region smaller than that, there is almost no fluctuation. If the structure itself does not fluctuate), even if the conversion difference fluctuates between wafers, the above pattern only shifts the position of 0, and if the position can be confirmed, the width is specified based on that position can do. In this pattern,
In principle, the accuracy of the width is determined by the pitch shift amount. The finer the pitch shift amount, the higher the accuracy of the dimensional reference.

In the structure resulting from the above-described pitch shift, the width is determined by the pitch accuracy of the original pattern. In order to ensure the accuracy in a region on the order of cm, such as the entire surface of the wafer, a technique such as an interference exposure method is advantageous. However, in the present invention, the accuracy may be ensured in a region on the order of μm. It is possible to realize a standard deviation of about 1 nm as the pitch accuracy using the current electron beam writing apparatus, and there is a possibility that it will be further improved in the future. Therefore, the structure of the present invention can be sufficiently used as a dimension standard on the order of nm.

Further, the above-mentioned structure is essentially constituted by a pitch structure, is compatible with a conventionally used dimensional standard at this level, and is suitable as a dimensional standard. Further, the width of the structure itself is used as the length reference instead of the pitch as in the conventional length reference, and the consistency with the actual length measurement procedure for measuring the structure width is good. (The length measurement performed in a general LSI manufacturing process generally measures the width of a structure such as a line, a groove, and a hole.) Further, a pattern without a pitch shift (for example, a pattern at the left end in FIG. 1). By inserting the pattern, the pattern width of the upper and lower L & S regions can be known, and this can be used as a dimensional standard. That is, if it is known which pattern (NO) corresponds to the pattern having no pitch deviation from the pattern having the width of 0, the width L (L = (P2−P1) × (NO-1) + α, [α: error] ) Can be known. This means that the pattern conversion difference δ
L, that is, the deviation of the actual structure length Leff from the design length Ld is also known. (ΔL = Ld−Leff) A region where the pattern conversion difference does not change (generally, the conversion difference changes between wafers (on the order of mm or more) and between wafers), that is, the dimensional standard structure (pitch shift) If a line-shaped structure with a different design length is formed in the vicinity of the line-shaped structure, the effective width of the pattern can be known by observing the dimensional standard structure. Can also be used as a dimensional standard.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention provides a line and space structure including a first group, a second group, and a third group in the longitudinal direction of a line, wherein the third group is sandwiched between the first group and the second group. The widths of the lines of the first group and the second group are equal and the repetition period is different, and the widths of the lines of the third group are the first group and the second group starting from the line next to the line having the width of 0. It is characterized by a fine size standard structure and a fine size standardization method that are sequentially increased by a difference from the repetition period of the two groups.

EXAMPLE 1 Two types of linear repeating patterns having the same design line width of 20 nm but different pitches of 100 nm and 102 nm, respectively, on a Si (1-10) substrate, each having 30 lines, FIG.
As shown in FIG. 3A, the pattern is formed by EB exposure using an existing negative resist so that the pattern positions on the left end are aligned and the drawing areas are in contact with each other. At this time, it is desirable that the direction of the line be accurately aligned with the <11-2> direction. The resist pattern formed in this way is generally thicker than the design line width, and even a pattern that is not connected on the design pattern is connected and formed as shown in FIG. Next, based on this pattern, KO
The processing is performed using etching with a large crystal anisotropy such as H (a phenomenon in which the etching rate of the Si (111) plane is much smaller than the etching rate of other crystal planes). As a result, a structure having a width of (actual line width) − (amount of pattern position shift due to pitch shift) is formed at the junction of the two types of patterns as shown in FIG. During this processing,
The following principle is used. In etching with large crystal anisotropy such as KOH, the etching is effectively stopped when the Si (111) plane is exposed. For this reason, FIG.
<11-2> on a Si (1-10) substrate as shown in FIG.
The structure formed based on the linear resist pattern formed in the direction has a Si (111) surface on the side wall and has a line width having a certain conversion difference from the original resist pattern. On the other hand, at the junction of two types of patterns having different pitches, a pattern which is locally shifted from the <11-2> direction is formed. For this reason, the etching progresses, and as shown in FIG. 3B, the joints behave in such a manner as to move in opposite directions on the left and right. As a result, the (actual line width) − ( (Amount of pattern position shift due to pitch shift) A structure having a width as shown in FIG. For example,
Assuming that the final conversion difference is +20 nm in the above process, the actual line width is 40 nm, and the width is 0 in the 21st pattern from the leftmost aligned pattern.
Becomes 2, 4, 6,
.., 36, 38, and 40 nm can be realized.

In this embodiment, an example using a negative resist has been described. However, a pattern formed using a positive resist can be formed by a positive / negative inversion method (for example, a metal lift-off method,
Alternatively, the ECR oxygen plasma oxidation inversion method (Japanese Patent Application No. 6-2)
It is needless to say that a pattern inverted by using 15522)) may be used. Further, in this embodiment, KOH is used as the anisotropic etching solution.
11) The etching rate of the surface is sufficiently lower than that of the other surfaces, and an etching solution capable of forming the target structure in the present invention is, for example, KOH to which an alcohol such as ethylenediamine, hydrazine, or isopropyl alcohol is added. It goes without saying that an aqueous solution or the like may be used. Further, in this embodiment, the speed difference between the Si (111) surface and the other Si surface is used, but there is no problem in other surfaces as long as there is an effective etching method. Needless to say, the substrate is not limited to Si but may be another single crystal material.

Second Embodiment In the structure shown in the first embodiment, for example, in order to directly form a dimensional standard having a width of 100 nm and an accuracy of 1 nm, a linear structure having at least 100 pitches or more is required. This is inefficient. Therefore, in such a case, the following method is conceivable. Usually, the pattern conversion difference (the pattern width changes each time the transfer process is performed) has almost no pattern width dependence. Therefore, a narrow width of, for example, 10 nm
A pattern having a line width of about 100 nm and 101 nm is formed at a pitch of about 20 to 30 pitches, and the conversion difference is measured using this pattern. On top of that, the 100 nm
What is necessary is just to use a simple linear pattern of about a width as a reference. That is, for example, if the conversion difference measured from a pattern having a width of 10 nm is 10 nm, it can be specified that the one having a design length of 100 nm is 110 nm.

In the vicinity of the pattern shown in the first embodiment, for example, the design width is set to 10, 20, 30, 40, 50, 60, 7
Patterns of 0, 80, 90, and 100 nm are formed, and if the conversion difference is measured as +10 nm using the pattern shown in the first embodiment, for example, 20, 30, 40, and 40, respectively.
It can be used as a reference for line widths of 50, 60, 70, 80, 90, 100 and 110 nm. With such a structure, a dimensional standard of, for example, about 1 nm to about 100 nm can be realized.

[0017]

According to the present invention, there is an effect that a fine dimensional standard on the order of nm can be obtained.

[Brief description of the drawings]

FIG. 1 shows a conceptual diagram of a dimension reference pattern of the present invention.

FIG. 2 shows a conceptual diagram of a conventional dimension reference pattern.

3A and 3B are schematic diagrams showing a structure formation, wherein FIG. 3A shows a resist pattern, and FIG. 3B shows a structure after anisotropic etching.

FIG. 4 shows a schematic view of a process, wherein (a) is a design pattern,
(B) shows a resist pattern, and (c) shows a structure after anisotropic etching.

──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Takahiro Makino 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo Nippon Telegraph and Telephone Corporation (56) References JP-A-8-5313 (JP, A) JP-A-7-74222 (JP, A) JP-A-8-82632 (JP, A) JP-A-3-122514 (JP, A) JP-A-5-196451 (JP, A) JP-A-7-253456 ( JP, A) JP-A-6-218870 (JP, A) JP-A-6-58753 (JP, A) JP-A-4-289411 (JP, A) JP-A-8-94346 (JP, A) JP JP-A-7-176585 (JP, A) JP-A-9-129691 (JP, A) JP-A-8-35975 (JP, A) JP-A-6-66512 (JP, A) JP-B-3-71047 (JP) , B2) Yoshinori Nakayama, "The World's Smallest Ruler: Microscale", Clean Techno Over, Japan, July 1, 1994, Vol. 4 No. 7, p. 79-80 (58) Field surveyed (Int. Cl. ⁷ , DB name) G01N 13/10-13/24 G01N 1/00-1/44 G01B 21/00-21/32 JICST file (JOIS)

Claims (5)

(57) [Claims] A first group and a second group extending in a longitudinal direction of the line;
In the line & space structure composed of a group and a third group, the third group is sandwiched between the first group and the second group, the widths of the lines of the first group and the second group are equal, and the repetition period is different. The line width of the group starts from the line next to the line whose width is 0,
A fine dimension standard structure characterized by sequentially increasing by a difference between repetition periods of the first group and the second group. 2. The edge according to claim 1, wherein the edge defining the width of the line of the third group is one of the edge defining the width of the line of the first group and the edge defining the width of the line of the second group. A fine dimension standard structure, each of which corresponds to one. 3. The method according to claim 1, wherein a line having a width equal to the width of the line of the first group (or the second group) is a third line.
A fine dimension standard structure characterized by being included in a group. 4. The method according to claim 1, wherein:
A fine standard structure comprising a plurality of line structures having different line widths provided in the vicinity of the fine standard structure. Claim 5.The fine particle according to any one of claims 1 to 3.
A fine dimension standardization method using a fine dimension standard structure, Forming the fine dimension standard structure on a substrate,
Multiple line structures with different line widths near the standard structure
The pattern is formed from the pattern according to the fine dimension standard structure.
The tongue width conversion difference is measured, and the line width of the plurality of line structures is calculated.
And determining the plurality of lines based on the pattern width conversion difference.
Correct the line width of the structure and dimension the line structure with the corrected line width
As a reference A method for standardizing fine dimensions.