JP4124609B2 - Method and apparatus for measuring film thickness of pattern part in semiconductor manufacturing process - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method and apparatus for measuring a film thickness of a pattern part in a semiconductor manufacturing process.
[0002]
[Prior art]
In order to form semiconductor elements and wirings in semiconductor manufacturing, the steps are repeated in the following order. First, a thin film is formed on the entire wafer surface by a film forming process. Next, a resist film is applied to the entire surface of the wafer, and exposure according to the pattern shape is performed by a lithography process to remove only the exposed resist portion or only the unexposed resist portion. Next, the thin film other than the part protected by the resist film remaining without being removed, that is, the thin film part having the exposed surface is removed by etching or the like. Finally, the resist film remaining as a protective film at the time of etching is removed. As a result, the exposure pattern is transferred to the initially formed thin film. A part of the thin film remaining without being removed is used as a semiconductor element or wiring. Since the surface of the pattern manufactured by such a process is protected by a resist, the film thickness does not change by etching. Therefore, when it is desired to know the film thickness of the pattern portion, the film thickness of the thin film before pattern formation may be measured. For this purpose, methods such as 4-terminal resistance measurement, eddy current measurement, optical interference measurement, ellipsometer measurement, and fluorescent X-ray measurement are used.
[0003]
[Problems to be solved by the invention]
Recently, a method of embedding a metal in a groove portion removed by etching and using it as a wiring is also employed. In this case, a metal thin film is formed on the entire wafer surface after the resist is removed, and the metal thin film other than the metal filling the groove is removed by an etch back process or a CMP process. CMP is an abbreviation for Chemical Mechanical Polishing. The CMP process is a process in which a surface is chemically and mechanically polished, and the surface is made uneven by planarizing the surface. In these steps, it is fundamental to remove all the metal in a portion higher than the surface of the thin film constituting the groove, and leave a metal wiring having the same thickness as the thickness of the thin film, that is, the depth of the groove. However, in practice, a part of the surface of the thin film itself constituting the groove is removed, and similarly, the thickness of the metal wiring becomes thinner than the depth of the original groove. In addition, the metal wiring portion may be removed more than the surface of the thin film, and the wiring film thickness may be further reduced. Thus, a situation occurs in which the film thickness of the metal wiring is different from the film thickness immediately after the thin film forming the groove is formed.
[0004]
When trying to measure the film thickness of the wiring pattern created in the above-described process, the conventional method of measuring the film thickness of the thin film before pattern formation cannot be used because it has the following problems. First, the thickness of the wiring pattern is different from the thickness of the thin film before pattern formation. Therefore, measurement after pattern formation is required. In 4-terminal resistance measurement and eddy current measurement, it is necessary to measure a wide range of thin films with no structure in the plane direction, so it cannot be used after pattern formation. Secondly, since the optical interference measurement and the ellipsometer measurement can only measure a thin film that transmits light, a metal thin film cannot be measured. Thirdly, in the fluorescent X-ray measurement, the film thickness of each wiring can be obtained by using the excitation X-ray having a beam diameter smaller than that of each wiring pattern. Not right.
[0005]
Similarly, when the upper layer is selectively grown in accordance with the pattern structure, there is no wide-area thin film having no structure in the plane direction during the process, so that the same problem occurs when measuring the film thickness.
[0006]
In order to solve the above problems, an object of the present invention is to propose a method and apparatus for simply measuring the film thickness of a repetitive thin film pattern that has been miniaturized using fluorescent X-ray measurement.
[0007]
[Means for Solving the Problems]
In order to solve the above problems, in the present invention, the film thickness of the pattern structure portion is measured using fluorescent X-rays. However, a configuration is adopted in which a large number of repetitive thin film patterns are included in the measurement region without gaps. Moreover, when converting the detected fluorescent X-ray dose into a film thickness, the in-plane dimension of the thin film pattern is used.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Although the method of using fluorescent X-rays for film thickness measurement is well known, the physical measurement object is the number of atoms in the sample. The thin film to be measured is assumed to be uniform within the excitation region irradiated with excitation X-rays, and the atomic number density in the film is assumed, and the detected fluorescent X-ray dose is converted into the film thickness. Considering such restrictions, in order to measure the film thickness of a thin film having a pattern structure, the beam diameter of excitation X-rays must be made smaller than the pattern dimension. Since the actual pattern size to be measured is as small as 1 micrometer or less, it is not realistic to prepare such an excitation X-ray in a small analytical measurement apparatus.
[0009]
In the present invention, based on a new idea, a fluorescent X-ray measurement is proposed for measuring a region wider than the individual pattern dimensions to be measured and determining the film thickness of the thin film pattern. The new idea is that the measurement object should be a wide area where a large number of patterns are repeated, eliminating the conventional limitation that the measurement object must be a uniform thin film within the measurement area. The information obtained by the measurement is not the film thickness of each pattern but the average film thickness of the patterns included in the measurement region. Therefore, it is not necessary to make the measurement area smaller than the dimension of the microstructure to be measured. Conversely, the larger the number of samples contained in the measurement area, the better the accuracy of the average value. Further, as the measurement area is smaller, it is necessary to suppress the vibration to be low so that the measurement position does not shift. In the present invention, such a requirement can be suppressed. Therefore, the point that it is easy to cope with further miniaturization of the pattern dimension to be measured is an excellent point of the present invention.
[0010]
The details of the present invention will be described below using mathematical expressions. First, the fluorescent X-ray dose detected by fluorescent X-ray measurement is expressed as follows.
[0011]
B = C · ∫∫∫ (a (x, y) · G1 (x, y, z)) · D (x, y, z) ·
(E (x, y) · G2 (x, y, z)) dx · dy · dz
(Integration range: entire measurement area x thickness)
Here, B is the detected fluorescent X-ray dose, D (x, y, z) is the volume atom number density of the element to be measured in the film at position (x, y, z), and (a (x, y) G1 (x, y, z)) is the irradiation density of the excitation X-rays at the position (x, y, z), and (e (x, y) G2 (x, y, z)) is at the position (x, y, z). The detection efficiency for fluorescent X-rays emitted at y, z) is shown. G1 (x, y, z) indicates the transmittance when the excited X-ray enters the position (x, y, z). Therefore, G1 (x, y, 0) = 1. Similarly, G2 (x, y, z) indicates the transmittance until the fluorescent X-ray emitted at the position (x, y, z) reaches the sample surface. Therefore, G2 (x, y, 0) = 1. C represents a constant determined by the energy of excited X-rays, the type of element to be measured, the fluorescent X-ray energy to be detected, the measurement time, the unit system in the formula, and the like. x and y represent coordinates on two axes orthogonal to each other along the sample surface, and z represents a depth from the sample surface. “Thickness” in the integration range indicates the thickness of the target thin film pattern. Only the part where both a (x, y) and e (x, y) are positive in the integration contributes to the fluorescent X-ray dose B. The contribution area is the measurement area.
[0012]
When the pattern is repeated two-dimensionally along the sample surface, the following equation can be obtained by appropriately approximating the above equation. Here, S1 represents the area of the unit structure.
[0013]
∫∫∫G1 (x, y, z) · D (x, y, z) ·
G2 (x, y, z) dx · dy · dz (integral range: in unit structure)
= S1 · B / (C · (∫∫a (x, y) · e (x, y) dx · dy))
(Integration range: entire measurement area)
When the repeating structure along the sample surface exists only in one direction, the repeating width on the axis without the repeating structure is defined as any appropriate width. By doing so, the above basic formula can be used without changing S1 in the above formula and the definition of “in unit structure” of the integration range.
[0014]
Since (∫∫a (x, y) · e (x, y) dx · dy) in the equation is determined by the characteristics and arrangement of the optical element, a value calibrated by some reference measurement may be used as a constant. it can.
[0015]
Next, the equation for calculating the film thickness is derived by rewriting the integral on the left side of the above equation. First, it is assumed that the attenuation of X-rays in the sample is negligibly small so that the subsequent calculations can be performed clearly, and G1 (x, y, z) = G2 (x, y, z) = 1. And As a result, it can be rewritten as follows.
[0016]
∫∫∫G1 (x, y, z) · D (x, y, z) ·
G2 (x, y, z) dx · dy · dz (integral range: in unit structure)
= ∫∫∫D (x, y, z) dx · dy · dz (integral range: in unit structure)
Here, when the thin film sample to be measured is viewed from above, the region containing the element to be measured and the other region are clearly separated in the xy plane, and the atom number density D (x, y, z in each region). ) Is D and 0. Hereinafter, the range including the measurement target element in the unit structure is referred to as a unit measurement target range, and the area in the xy plane is defined as a unit measurement target area S2. The film thickness in the unit measurement target range is T (x, y). Using these definitions, the formula is further transformed. Subsequently, the case where the film thickness T (x, y) within the unit measurement target range is set to a constant value T will also be examined.
[0017]
∫∫∫D (x, y, z) dx · dy · dz (integral range: in unit structure)
= ∫∫∫Ddx / dy / dz
(Integration range: Unit measurement target range x thickness)
= D · ∫∫T (x, y) dx · dy (Integration range: Unit measurement target range)
= DT ・ ∫∫dx ・ dy (Integration range: Unit measurement target range)
= DT ・ S2
Therefore, the following two formulas can be obtained.
[0018]
T = (S1 / S2) · B /
(C · D · (∫∫a (x, y) · e (x, y) dx · dy))
(Integration range: entire measurement area)
∫∫T (x, y) dx · dy (Integration range: Unit measurement target range)
= S1 · B / (C · D · (∫∫a (x, y) · e (x, y) dx · dy))
(Integration range: entire measurement area)
From the former equation, it can be seen that T can be obtained based on the fluorescent X-ray dose B and the pattern area ratio (S1 / S2).
[0019]
For example, in the case of a repetitive pattern wiring structure (see FIG. 2), if two dimensions of the wiring width W and the repetitive pitch width P are known, the film thickness T is obtained because (S1 / S2) = P / W. be able to. FIG. 2 is a cross-sectional view of the thin film, and shows a state in which a large number of wirings are arranged in a direction perpendicular to the paper surface.
[0020]
For example, if the plug structure has a repetitive pattern (see FIGS. 3A and 3B), the film thickness T can be obtained if the two values of the area S1 of the unit structure and the wiring cross-sectional area S2 of the plug are known. FIG. 3A is a view of the sample from the top, and shows a repetitive structure of the plug in the xy plane and a range corresponding to S1 and S2. The plug is a short wiring that vertically connects between a lower wiring and an upper wiring parallel to the wafer surface. The wiring cross-sectional area of the plug refers to a cross-sectional area when the plug wiring is cut along a plane (xy plane) parallel to the wafer surface. FIG. 3B is a perspective view showing the plug structure in three dimensions.
[0021]
When obtaining the pattern area ratio (S1 / S2), there is a method of using the design value of the pattern part as it is, but if there is a concern about deviation from the design value, the in-plane dimension of the pattern part is used as another measuring means. It is necessary to ask. Therefore, by incorporating a surface length measuring function such as an optical microscope, electron microscope, or SPM (Scanning Probe Microscope) into a fluorescent X-ray film thickness measuring device incorporating the above method, The film thickness measurement in the depth direction can be performed collectively. SPM is a generic name for the surface, such as STM (Scanning Tunneling Microscope), AFM (Atomic Force Microscope), SNOAM (Scanning Near-Field Optical), It is a microscope that detects the optical and optical characteristics and measures the surface shape and characteristics of the sample with atomic and molecular sizes.
[0022]
Regarding the latter formula, the following rewriting is performed using F (x, y) indicating the surface shape. Here, the relationship of F (x, y) −F0 = T (x, y) is used. When the surface shape F (x, y) is measured by measurement using SPM or the like, the information inside the sample is not obtained by itself, so the offset F0 must be taken into consideration with the film thickness T (x, y). .
[0023]
∫∫T (x, y) dx · dy
= ∫∫ (F (x, y) −F0) dx · dy
= ∫∫F (x, y) dx · dy-F0 · S2
= ∫∫F (x, y) dx · dy− (F (x, y) −T (x, y)) · S2
(Integration range: Unit measurement target range)
Finally, the following equation can be obtained.
[0024]
T (x, y)
= F (x, y)-(1 / S2) · ∫∫F (x, y) dx · dy
(Integration range: Unit measurement target range)
+ (S1 / S2) ・ B /
(C · D · (∫∫a (x, y) · e (x, y) dx · dy))
(Integration range: entire measurement area)
Accordingly, it is understood that T (x, y) can be obtained based on the fluorescent X-ray doses B, S1, S2, and the surface shape F (x, y). T (x, y) is obtained by adding film thickness information to the surface shape F (x, y) and represents a cross-sectional shape.
[0025]
Next, the average film thickness T within the unit measurement target range is obtained using the above equation. The result is as follows, and the same result as the film thickness T when a sample having a constant film thickness is measured is obtained.
[0026]
T = (S1 / S2) · B /
(C · D · (∫∫a (x, y) · e (x, y) dx · dy))
(Integration range: entire measurement area)
When determining the surface shape F (x, y), S1, and S2, the surface shape is measured by SPM or the like. By incorporating the surface shape measuring function into the fluorescent X-ray film thickness measuring device of the present invention, The cross-sectional shape can be measured in the apparatus.
[0027]
Up to this point, the calculation has been made assuming that G1 (x, y, z) = G2 (x, y, z) = 1. However, if this assumption is not appropriate, the film thickness T is calculated as in the above calculation. Or T (x, y) cannot be expressed by a simple expression. In such a case, first, the film thickness T or T (x, y) is appropriately assumed, and G1 (x, y, z) and G2 (x, y) are based on the specific structure of the thin film pattern. , Z) is obtained by calculation. Next, using the G1 (x, y, z) and G2 (x, y, z), the film thickness is obtained from the detected fluorescent X-ray dose. If the film thickness obtained here agrees with the initially assumed film thickness, it can be understood that the initially assumed film thickness was correct. If there is a deviation, the same process is repeated using the corrected assumed film thickness. When the assumed film thickness coincides with the film thickness of the analysis result, the correct film thickness is obtained. Further, depending on the actual structure of the thin film pattern, an equation for obtaining the film thickness T or T (x, y) may be calculated in the same manner as described above by incorporating a specific structure into the calculation. In that case, there is no need to repeat the calculation.
[0028]
【Example】
Embodiments will be described with reference to the drawings. In this embodiment, the measurement example for the wiring structure sample is shown, but the repeated wiring structure is not the only sample. For example, a repeated pattern structure of a plug is also included in the subject of the present invention.
[0029]
FIG. 1 is a diagram including a measurement sample and a fluorescent X-ray measurement device. A TEG 5 is provided on the surface of the wafer 4 and the excitation X-ray 1 is irradiated there. A TEG (Test Element Group) is a measurement-dedicated area created in accordance with measurement conditions, and is used to monitor the manufacturing process of an actual device that cannot be measured directly. Only the portion irradiated by the excitation X-ray 1 becomes the measurement region 15, and the fluorescent X-ray 14 emitted therefrom is measured using the fluorescent X-ray detector 3. Further, in order to narrow down the irradiation range of the excitation X-ray 1, the excitation X-ray condensing mirror 2 is used. The TEG used in this method is created simultaneously with the management target pattern so that the pattern shape is the same as the management target pattern. In addition, the lower structure formed by the previous process is limited so that the lower structure does not affect the measurement. The beam diameter of the excitation X-ray and the size of the TEG are set in accordance with the pattern dimensions so that a large number of patterns are included in the irradiation range of the excitation X-ray. An excitation X-ray beam having a required beam diameter may be prepared using a collimator instead of the condenser mirror. In an actual apparatus, it is necessary to specify the position of the TEG 5 in the wafer 4 and to align the measurement area with the position, and an optical microscope or the like built in for that purpose is used.
[0030]
FIG. 4 shows a wiring pattern after CMP. The wiring part 8 left after the CMP has a rough shape, not a horizontal surface. The wiring pitch 6 and the wiring width 7 of the wiring pattern are obtained by SEM measurement, and the film thickness of the wiring part is obtained based on the pattern area ratio (S1 / S2) obtained therefrom. In the film thickness measuring method according to the present invention, an average film thickness 9 as indicated by a broken line is obtained.
[0031]
FIG. 5 shows a wiring pattern after CMP similar to FIG. In this example, the wiring pitch 6 and the wiring width 7 of the wiring pattern are obtained by AFM measurement, and the film thickness 9 of the wiring part is obtained based on the pattern area ratio (S1 / S2) obtained therefrom. In the AFM measurement, the surface shape 10 of each wiring can also be measured. Therefore, according to the present invention, not only the flat average film thickness 9 indicated by the broken line but also the average wiring cross-sectional shape 11 can be obtained. Therefore, the film thickness 12 near the wiring side wall, the film thickness 13 at the bottom of the chipped shape of the wiring, and the like can be obtained. The wiring cross-sectional area obtained from the flat film thickness 9 and the wiring cross-sectional area obtained from the film structure 11 are substantially equal.
[0032]
【The invention's effect】
The present invention is implemented in the form as described above, and has the following effects.
[0033]
By irradiating excitation X-rays so that a large number of repeated patterns are included in the irradiation range, and measuring the fluorescent X-ray dose from the irradiation range, the average film thickness of the pattern thin film can be obtained based on the pattern area ratio. it can. Further, the cross-sectional shape of the pattern thin film can be obtained based on the pattern surface shape.
[0034]
By incorporating surface length measurement functions such as an optical microscope, electron microscope, or SPM into the fluorescent X-ray film thickness measurement device, the length measurement in the surface direction and the thickness measurement in the depth direction are performed together in the same device. Can do. Further, by incorporating a surface shape measurement function such as SPM, the cross-sectional shape of the pattern portion can be measured in the same apparatus.
[0035]
Since the measurement area can be widened, it is not necessary to focus the beam as much as the pattern size, and it is not necessary to keep vibrations low, so that it is easy to cope with further miniaturization in future semiconductor processing.
[0036]
With respect to the wiring of the semiconductor integrated circuit, the film thickness of the wiring portion can be obtained based on the wiring width and pitch.
[Brief description of the drawings]
FIG. 1 is a diagram showing an embodiment of the present invention together with a TEG.
FIG. 2 is a cross-sectional view of a patterned thin film, showing a wiring structure of a repetitive pattern.
FIG. 3A is a view of a repetitive pattern of a plug structure as viewed from above a thin film. B is a perspective view three-dimensionally showing the repetitive pattern of the plug structure.
FIG. 4 is a diagram illustrating an example in which measurement of a wiring pattern is performed in combination with an SEM.
FIG. 5 is a diagram illustrating an example in which measurement of a wiring pattern is performed in combination with an AFM.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Excitation X-ray 2 Excitation X-ray condensing mirror 3 Fluorescence X-ray detector 4 Wafer 5 TEG
6 Wiring pitch 7 Wiring width 8 Wiring 9 Film thickness (measurement result)
10 Surface shape 11 Film structure (measurement result)
12 Film thickness in the vicinity of the wiring side wall 13 Film thickness at the bottom of the chipped shape of the wiring 14 Fluorescent X-ray 15 Measurement region

Claims (6)

Irradiating an excitation X-ray in the measurement area comprising a plurality of repetitive patterns,
Detecting fluorescent X-rays emitted from the measurement region;
Measure the in-plane dimensions of the measurement area with the surface length measurement function,
A pattern film thickness measurement method for obtaining a film thickness of the pattern based on the detected amount of the fluorescent X-ray and the measured in-plane dimension. Surface shape thin pattern film thickness measuring method according to claim 1, wherein determining said pattern cross-sectional shape based on the measured surface shape information by measuring how with the measurement function.   2. The thin film pattern film thickness measuring method according to claim 1, wherein the pattern is a semiconductor integrated circuit, and the measurement region is a wiring width and a pitch of the semiconductor integrated circuit wiring. Surface length measurement function to measure the in-plane dimension of the measurement area consisting of multiple repeating patterns,
An excitation X-ray source for irradiating the measurement region with excitation X-rays;
A fluorescent X-ray detector for detecting fluorescent X-rays emitted from the measurement region;
A plane dimension measured at the surface length measurement function, that said detected by X-ray fluorescence detector based on the detected amount of the fluorescent X-ray, with a, and analyzing means for determining the thickness of said pattern A pattern film thickness measuring device.   The thin film pattern film thickness measuring device according to claim 4, further comprising a surface shape measuring function capable of measuring the surface shape.   6. The thin film pattern film thickness measuring apparatus according to claim 5, wherein the pattern is a semiconductor integrated circuit wiring, and the in-plane dimension is a wiring width and a pitch of the semiconductor integrated circuit wiring.

Images