JP2001074419A - Method and device for measuring patternized structure - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method and device for measuring parameters of a patternized structure. SOLUTION: A structure represents a grid having one cycle 12 comprising two adjacent elements 12A, 12B having different optical properties related to incident radiation rays. An optical model based on a few characteristics of the structure can determine theoretical data Dt, representing spectral intensity of an optical element of different wavelengths fully reflected from the structure, and at least one desirable parameter of the structure is computable. A substantially larger measuring region than a surface region of the structure provided by a grid cycle is irradiated by the incident radiation rays in the prescribed wavelength range. The optical element substantially reflected fully from the measuring region is detected, thereby obtaining measuring data. The measuring data and theoretical data are analyzed to optimize the optical model, until they satisfy the conditions set previously by the theoretical data. If the conditions set previously are satisfied, at least one parameter of the structure is calculated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention belongs to the field of measurement technology and relates to a method and an apparatus for measuring parameters of a patterned structure.

[0002]

2. Description of the Related Art Techniques for measuring the thickness of patterned structures have been developed. The term "patterned structure" as used herein refers to a structure formed of regions having different optical properties with respect to incident light (radiation). In particular, the patterned structure represents a grid in which each cycle has one or more cycles of at least two adjacent stacks. Each stack consists of multiple layers with different optical properties.

[0003] The fabrication of integrated circuits on semiconductor wafers requires maintaining tight control over the various dimensions of small structures. With certain measurement techniques,
Although the local dimensions of the wafer can be measured with relatively high resolution, the use of intermittent use of the wafer during manufacture must be sacrificed. For example, inspection using a scanning electron microscope allows measurement of the parameters of the patterned structure, which may be broken and must be excluded from continued processing. For mass production of such patterned structures such as wafers, non-destructive methods of controlling thin film parameters in a manner that allows local measurements are required.

One prior art technique for measuring thin film thickness is described in US Pat. No. 4,999,014. This technology is
It is based on using a small spot size for measurement and a large aperture value on a small area. Unfortunately, for small structures, this technique has, on the one hand, the general disadvantages associated with the use of small spot sizes and, on the other hand, the accumulation of higher order diffraction due to the large aperture value. It has associated disadvantages. The term "small spot size" means a spot diameter dimensionally close to a line or space of the measurement structure, ie a single grid cycle. by this,
A variety of difficult problems arise. In practice, not all stacks are at the focal point of the optical system used to collect the reflected light, so the optical system is large and complex. The detection signal is affected by a minute change in the grid shape and a deviation of the spot arrangement. Diffraction effects are highly dependent on grid geometry and topography and must be accounted for.

Another example of the above prior art is disclosed in US Pat.
No. 1,137, relates to a method and apparatus for measuring sub-micron line widths of patterned structures. The measurement is performed on a so-called "test pattern" of the diffraction grating type, which is placed in the test area of the wafer. Here, as in most conventional systems, black and white incident light is used to form and analyze the diffraction pattern. But,
A large number of test areas are used, and information on a plurality of parameters cannot be obtained.

Some prior art, such as US Pat.
According to 5,087,121, the part with and without the trench is illuminated with broadband light, the reflection spectrum is measured and the corresponding results, represented by the height or depth of the structure, are compared. . However, the structure under inspection is
Often the different parts cannot be displayed separately. this is,
Due to unavoidable limitations associated with the diameter of the incident beam impinging on the structure.

The above technique utilizes a frequency filtering technique that allows the separation of interfering signals from different layers. this is,
Due to the limited number of reflected oscillations, this cannot be achieved for layers with small thicknesses or layers with small thickness variations.

[0008] Yet another example of the prior art for implementing depth measurement is disclosed in US Patent No. 5,702,956. This method utilizes a test section that represents a patterned structure similar to a wafer (circuit section). The test section is in the form of a plurality of test areas, each located in a space between two adjacent circuit areas. This test area is designed to be large enough to have a trench depth measured by an in-line measurement tool. Assuming that the process is not dependent on feature size, the measurement is performed by comparing the parameters of different test areas. For many processes in the art, such as etching and photoresist development, this assumption is incorrect and therefore this method is not available.

[0009]

SUMMARY OF THE INVENTION It is a primary object of the present invention to overcome the above and other disadvantages of the prior art and to provide a method and apparatus for non-destructive non-contact measurement of parameters of a patterned structure. That is.

It is another object of the present invention to provide a method and system for obtaining relatively little information representing the condition of a structure and for successfully performing measurements on extremely complex structures. is there.

[0011]

SUMMARY OF THE INVENTION In one aspect of the invention, a grid having at least one cycle (12) of at least two adjacent elements (12a, 12b) having different optical properties with respect to incident radiation. (10, 100, 110, 21)
0, 310, 410, 510), wherein the method has at least one desired parameter of a plurality of features defined in a manufacturing process, the method comprising: a) at least some of the features of the structure; Preparing an optical model capable of determining theoretical data (Dt) representing the spectral intensity of light components of different wavelengths specularly reflected from the structure, based on the following: b) the surface area of the structure determined by the grid cycle Irradiating a substantially larger measurement area with incident radiation in a preset substantially wide wavelength band; c) detecting substantially zero-order reflected light from the irradiation unit and detecting each wavelength within the wavelength band. Obtaining measured data representing the spectral intensity of the optical data; d) analyzing the measured theoretical data, until the data satisfies predetermined conditions; The step of optimizing the Le, and e) when detecting that the predetermined condition is satisfied,
Calculating the at least one parameter of the structure.

Therefore, the main technical idea of the present invention is as follows. The patterned structure whose parameters are to be measured is formed by several successive steps of a certain technical process that has been completed before the measurement. The features of the actual design rules can often be found together in the structure (eg, memory read lines). The term "design rule features" refers to the dimensions of a set of allowed patterns used throughout the wafer. Thus, information about the desired parameters can be obtained using a supermicron tool, such as a large spot focused on a set of lines.

The present invention utilizes a spectrophotometer that receives substantially zero-order reflected light, unlike conventional methods. Zero-order signals are unaffected by small changes in the grid contour of the structure, such as rounded edges or slopes. This eliminates the need to consider effects associated with diffracted light, and therefore
The optical model can be simplified with the optical system. Also, a large spot size allows for a large depth of focus, including the entire depth of the structure to be measured. When the spot includes several grid cycles, the measurement is not affected by local defects, precise spot placement or focusing.

In the case of a wafer, each element of the grid cycle consists of a stack of different layers. The features of such structures (wafers) that are determined by the manufacturing process and to be considered by the optical model represent the following known effects: Specular reflection from separate cycles in the grid cycle, interference of reflected light from layers in each stack, vanishing in the transparent stack due to the cavity shape formed in the grid-like structure, specular reflection due to the width of the stack relative to wavelength,
Polarization (if any) due to the interaction of the conductive grid-like structure with the incident beam, the effect of limited coherence of illumination Reflected from each stack in the grid cycle, considering the above effects The interference between the light beams and the degree of contribution of each of the above effects to the theoretical data are calculated according to known physical laws.

An optical model based on some features actually requires that certain factors of the optical model be considered in order to perform an accurate calculation of the desired parameters. If information on all features is not available and the model is not optimized before the measurement, this is done by a so-called early "learning" process. That is, on the one hand,
Depending on the variability of all features, on the other hand, there are several optical model factors that determine the contribution of each optical effect present to the detected signal. The values of these optical model factors are
During the learning process, it is adjusted together with unknown desired parameters to satisfy a predetermined condition. The latter (parameter)
Is usually in the form of a merit function that specifies the so-called “fitness” between the measured data and the theoretical data. The adjusted optical model factors are used in conjunction with known features to enable accurate calculation of desired parameters of the structure.

Preferably, the measurement area is part of a measurement structure. Alternatively, the measurement area represents the actual structure to be measured, ie is located on a test pattern having the same design rules and layer stack. The necessity of such a test pattern arises for the following two reasons. 1) If the measurement area is not much smaller than the available surface area specified by the actual structure being measured, the test location is implemented to include an enlarged structure. 2) If the structure is very complex or consists of an ambiguous underlying structure, the test locations are of the same shape as the actual structure being measured, but are implemented with a simplified underlying design, A simplified measurement of the upper layer is possible.

According to another aspect of the invention, there is provided a grid having at least one grid cycle (12) of at least two adjacent elements (12a, 12b) having different optical properties with respect to incident radiation. Patterned structures (10, 100, 110, 21)
0, 310, 410, 510) for measuring at least one desired parameter having a plurality of characteristics determined by a certain manufacturing process, a) the surface of the structure determined by a grid cycle A measurement area that is substantially larger than the area is irradiated with incident radiation in a predetermined wide wavelength band, and a specular reflection element of reflected light from the measurement area is detected to indicate a spectral intensity of the measurement light within the wavelength band. A spectrophotometer (14) for providing data; b) a processing unit (2) coupled to said spectrophotometer.
0), comprising pattern recognition software and conversion means, responding to the measurement data, positioning the measurement values,
Utilizing an optical model based on at least some of the features of the structure to obtain theoretical data representing a spectral intensity of light specularly reflected from the structure within the wavelength band, the at least one desired parameter; Is calculated, and the measured data is compared with theoretical data to determine whether the theoretical data satisfies a predetermined condition.

The spectrophotometer preferably has an aperture stop arranged on the optical path of the specularly reflected light component. The diameter of the aperture stop is automatically set according to the measured grid cycle of the structure.

Preferably, the incident radiation and the reflected light received by the detector are directed along a substantially specular axis.

Since the present invention particularly relates to the measurement of the height, depth and width of a semiconductor wafer, this application will be described below.

[0021]

1 (a) and 1 (b) are a partial sectional view and a top view, respectively, of a grid-like wafer structure 10 whose parameters are to be measured. The structure comprises a plurality of cells 12, each constituting a grid cycle. In this embodiment, only three adjacent cells 12 are shown. For simplicity of illustration, it is assumed that each cell contains only two stacks (or elements). That is, cell 12
Are two stacks 12a and 12 of separate layers
b. Specifically, the stack 12a includes six layers L1
-L6, of which L1 and L2, L6 respectively constitute the two layers L1 and L2,6 of the stack 12b. As is known in conventional semiconductor devices, semiconductor structures, such as sources, drains, gate electrodes, capacitors, are typically formed on a semiconductor substrate (layer L1) of silicon material, which is a metal conductor ( (Aluminum). The substrate is covered with an insulating silicon oxide compound (layer L2). A first level metal layer L4 (a single layer in this embodiment) is formed;
It is inserted between the upper barrier layer L3 and the lower barrier layer L5 made of titanium nitride (TiN). The uppermost insulating silicon oxide layer L6 is covered, and it is thinned by chemical mechanical polishing (CMP) to complete the structure. The structure of such a structure and its manufacturing method itself are well known and need not be described in detail here.

In this embodiment, the parameters to be measured are the widths W1 and W2 of the stacks 12a and 12b,
Depth d1 of uppermost silicon oxide layers L6 and L2,6
And d2. It should be noted that other parameters, such as material, optical properties, etc. of the patterned structure can also be measured. Reference is now made to FIG. 2, which illustrates a system 14 suitable for performing measurements. System 14 represents one of the workstations on a production line (not shown), with wafers 10 moving between upstream and downstream stations on the production line. The system 14 includes an indicator frame 16 for holding the structure 10 during the inspection phase, the spectrophotometer 1
8 and a processing unit 20 connected thereto.
The spectrophotometer 18 generally includes a light source 22, which emits a light beam 24 in a predetermined wavelength band, a light directing optical system 26, and a detection unit 28. Light directing optical system 26
Is usually in the form of a beam deflector comprising an objective lens 30, a beam splitter 32 and a mirror. The detection unit 28 includes an imaging lens 36, a motor 40
Aperture stop 3 coupled with and operated by
8 and a spectrophotometric detector 42. The structure and operation of the spectrophotometer 18 may be, for example, those known in U.S. Pat. No. 5,517,312, assigned to the assignee of the present application. Therefore, the spectrophotometer 18 will not need to be described in detail except as follows.

The light beam 24 passes through the light directing optical system 26, and the structure 1 at a predetermined position that defines the measurement area S1.
Incident at 0. The light component 44 specularly reflected from the reflection area in the area S1 is directed to the detector unit 28.

In general, the irradiated area of the structure is larger than the measurement area S1, and in that case, a suitable optical system for taking in the light reflected only from the area S1 in a conventional manner is provided in the irradiated area. Is provided. That is, the measurement area is
When the light beam 24 is incident on the structure, it is contained within the spot size provided thereby. For ease of understanding, it is assumed that the irradiation area specified by the diameter of the incident beam is the measurement area S1.

Light directing optical system 26 and detector unit 28
Is designed so that only the zero-order light component of the reflected light from the structure 10 is detected by the spectrophotometric detector 41. The incident light beam and the detection light beam are substantially parallel to each other and substantially perpendicular to the surface of the structure 10. The diameter of the aperture stop 38 is variable and is automatically set according to the measured lattice cycle of the structure. Generally speaking, the diameter of the aperture stop is optimized to collect the maximum reflection intensity except for those of multiple diffraction orders.

The incident beam 2 that defines the measurement area S1
4 has a surface area S0 defined by the cell 12.
Much bigger. That is,

(Equation 7)

According to this embodiment, the patterned structure 10
Is a so-called "one-dimensional" structure. As shown in FIG. 1B, the stacks 12a and 12b
Aligned along the axis, their stacks continue along the Y-axis to infinity (uniform structure) with respect to the measurement area S1. In other words, the measurement area S1 has one or more grid cycles extending along the X axis and Y
Including structures that are uniform along the axis.

The total surface area S of the structure to be inspected is the measurement area S defined by the diameter of the incident beam.
Must be much larger than one.

(Equation 8)

In some cases, the above conditions may not be obtained in the structure 10. For example, the structure may have a single lattice cycle. For this purpose, a measurement area S1 composed of a plurality of cells 12 should be arranged on a test location (not shown).

For example, if the system 14 has a numerical aperture of 0.2 and 15 μm
When the spot diameter (measurement area S1) is about the same, the minimum surface area S of the test location is 20 μm. No released by Nova Measuring Instruments, Israel
The vaScan 210 spectrophotometer can be used in the system 14.

The spectrophotometer 18 measures the spectral intensities of different wavelengths contained in the detected zero-order light component of the reflected beam 44. This is shown in FIG. 3 as a dotted curve Dm serving as measurement data. The processing unit 20 includes pattern recognition software and conversion means and responds to the measurement data to determine the measurement. It is pre-programmed with an optical model based on at least some features of the structure to theoretically calculate the spectral intensity of light having different wavelengths reflected from the patterned structure. This is shown as a solid curve Dt serving as theoretical data in FIG.

In order to design an optical model that can predict all possible optical effects determined by the features of the structure that affect the measured production data, the following should be considered: .

Normally, the total regular reflection R from the lattice structure is composed of a coherent portion Rcoh and a non-coherent portion Rincoh. When broadband radiation is used, the coherence effect plays a very important role in the measurement. The coherence length L of light in the optical system is determined by the radiation source and the optical system (spectrophotometer) itself. Reflected light amplitudes from features of the structure that are shorter than the coherence distance interact coherently, thereby producing an interference effect on light reflected by separate stacks of cells. For large features, a non-negligible portion of the light reflected by the different stacks undergoes non-coherent interactions without causing interference. The coherence length L limits the mutual coherence ν of light originating from points separated by a half cycle of the grating structure, and ultimately limits the degree of coherence γ. That is,

(Equation 9) Becomes Where D is a variable parameter experimentally determined for actual optical systems and stack structures based on measured reflectance spectra (measured data) for gratings having various cycle dimensions, and J1 is a known Bessel function It is. Approximate initial inputs to the determination of parameter D can be given by nominal optical system characteristics. Therefore, total specular reflection R is expressed by the following equation.

(Equation 10)

To calculate the optical effects that affect each term of the total reflection signal in the above equation, the following main factors should be considered. Patterned structure 10 (FIGS. 1A and 1B)
1) Filling factors a1 and a2 are represented by the following equations.

[Equation 11]

These factors represent the zero-order contributions based on the area ratio of the stacks 12a and 12b, respectively, in the calculation of reflection. Zero-order signals are insensitive to grid-like details of the structure, such as rounded edges and local slopes. Therefore, effects related to diffracted light do not need to be considered.

2) Size coupling fact (size coupling fact)
or) c1 and c2 When the width of the stack is close to the wavelength of the radiation, the fill factors a1, a2 should be modified to reduce the coupling of the incident radiation into each stack. Therefore, the so-called “binding factors” c1 and c2 are replaced by the filling factors a1 and a1, respectively.
2 should be introduced. The coupling factor has negligible effect when the stack width is relatively larger than the wavelength, but when the stack width is much smaller than the wavelength,
Completely disable the interaction. Using a newly discovered exponential function that gives such a dependency, the binding factor is expressed by the following equation.

(Equation 12) Where λ is the wavelength of each light component and A is a variable that depends on the size and material of the structure, and is determined experimentally on the actual stack structure, as described below. You.

3) Dissipation in the hollow structure
ion) b2 Often, one of the stacks is inherently dissipative due to the geometric effects of reduced reflection that normally occur in hollow structures. Some of these geometric effects include high aspect ratios, subsurface waveguiding of metal-like structures, and the like. High aspect ratios are characterized by dissipative effects that reduce the amount of light that is completely reflected back by phase effects.
For example, multiple reflections in deep trenches in metal reduce the amount of complete return reflection and destroy the phase relationship. These effects above are strong for deep features and weak (with respect to wavelength) for shallow features. Using a newly discovered exponential function that gives such a dependency, the dissipation factor b2 is expressed as follows.

(Equation 13) Here, B represents a variable size parameter experimentally determined for the actual stack structure, and d2 represents the depth of the cavity of the stack. In this embodiment, the stack 12b is dissipative.

To create a modified filling factor model, it is assumed that light radiation that is not reflected from a given cell stack due to coupling is originally reflected by other cell stacks. Dissipation factor b2 takes into account the reduced effective filling factor of the geometrically dissipative area. Therefore, the modified filling factor is expressed by the following equation.

[Equation 14]

4) Polarization factor representing the degree of contribution to the polarization effect that occurs in the case of a metal grid. When the width of the cell stack is close to the wavelength of the incident radiation, the polarization is transferred to each stack due to the boundary state at the edge of the metal line. Incident T
A correction factor should be introduced to reduce the coupling of E-radiation. The polarization factor has a negligible effect when the stack width is relatively large with respect to wavelength, but has the effect of completely nullifying reflections when the stack width is much smaller than the wavelength. Therefore, the polarization factor p
1, p2 It is expressed by the following equation.

(Equation 15) Here, C is a variable parameter experimentally determined for an actual stack structure. If there is no pattern formed by metal lines, the optical factor C is considered equal to zero.

Similarly, to use the modified fill factor, it is assumed that radiation that is not reflected from a given cell stack due to polarization is essentially reflected by other cell stacks. The corrected filling factor is represented by the following equation.

(Equation 16)

The intensity of the reflected signal r (λ) from each stack is calculated using layer thickness information and material optical parameters (known features). For this, a standard formula for reflection from a multilayer stack is used. This expression is
It is based on the Fresnel coefficient for reflection and transmission at the boundary as a function of the wavelength of the normally incident light. The thickness of each layer is either known (provided by the user) or calculated internally by the program. The material of the layers, and thus the optical coefficients, such as the refractive index and the absorptivity, are known or can be calculated.

The above point, and both the coherent and incoherent portions, have two polarizations (eg, Rcoh
= R ^(p) + R ^(s) ), the total reflection R, which is theoretical data obtained by the optical model, is given by the following equation.

[Equation 17] Here, r1 and r2 indicate the amplitudes of the reflected waves from the first and second stacks of cells, in this embodiment, the stacks 12a and 12b, respectively.

Other effects of the prior art (lateral reflection, roughness,
Has been found to have a negligible contribution under the given conditions, so it is only necessary to adjust the parameters A, B, C and D. Returning to FIG. 3, curves Dm and D
It clearly shows that t does not match, ie the theoretical data does not exactly match the measured data. An appropriate merit function is used to determine whether or not the obtained result is appropriate. An optical model is defined by setting the values of the optical model parameters A, B, C, and D. By adjusting the values of the desired parameters W1, W2, d1 and d2, the theoretical data is optimized until the fitness reaches a certain desired value (constituting the required state). In detecting that the optimized theoretical data satisfies the required state, the desired parameter W of the structure
1, W2, d1, and d2 are calculated from the various equations described above.

It should be noted that in the most general case, the optical model is still correct when the grid cycle includes two or more adjacent discrete elements (eg, stacks). The mutual coherence ν ′ is as follows.

(Equation 18) Here, i is the ith element (stack) in the grid cycle, n is the number of elements in the grid cycle, L
Is the coherence distance. The main factors on which the above optical model is based are the filling factors:

[Equation 19] Binding factor:

(Equation 20) Dissipation factor:

(Equation 21) Here, m is the number of dissipative elements in n stacks, and dm is
2 represents the depth of a cavity of a stack relative to an adjacent stack. For a non-dissipative stack, bn = 1 (n ≠ m). Polarization factor:

(Equation 22) Modified filling factor:

(Equation 23) From the above equation, total reflection is expressed as follows.

(Equation 24)

FIG. 4 shows a so-called two-dimensional structure 100,
That is, a periodic structure is shown in both the X and Y axes. This structure features multiple grid cycles aligned along both the X and Y axes. The cycle aligned along the X axis is a pair of elements W1 and W2
The cycle aligned along the Y-axis is made up of a pair of elements G1 and G2 (the length of the stack). For example, elements G1, G2
May be a metal layer stack and an interlayer dielectric (ILD) stack, respectively. The measurement area defined by the diameter of the incident beam is at least one cycle in the X-axis direction and Y
At least one cycle in the axial direction (a plurality of cycles in this embodiment) is included.

Generally speaking, a cycle in either the X-axis or Y-axis direction may be composed of several elements (eg, stacks). If the measurement area S1 is smaller than the surface area defined by the grid cycle along either the X-axis or the Y-axis, the total reflection (theoretical data)
(FIGS. 1 (a) and (b)). If the measurement area is larger than the surface area defined by the grid cycles along both the X and Y axes, the total reflection R _2D of such a two-dimensional structure is given by:

(Equation 25) Here, R _G1 and R _G2 are the intensities of the reflected signals from the two one-dimensional structures aligned along the Y axis having widths G ₁ and G ₂ , respectively. In that case, the Y-axis is merely a display and has no physical significance and can be interchanged with the X-axis. In the general case where k elements are aligned during a cycle along the Y axis, R _2D is:

(Equation 26) Here, R _Gi and Gi are the reflection intensity from the i-th element and the width of the i-th element, respectively.

In general, the axis arrangement for calculating the reflection intensity R _G1 is selected so as to satisfy the following equation.

[Equation 27] If the above conditions are not met, the two axes are exchanged.

The main principle of the method according to the present invention will be described with reference to FIGS. The structure of the required measurement area is inspected (step 46), and it is determined whether or not the measurement area condition is satisfied within the existing pattern (step 48). If this condition is not satisfied, a test point structure that satisfies this condition is designed on a mesh line (step 50). In this case, the test location is usually located in a so-called “margin area”.

Next, an initial learning mode of operation is executed in step 52. The learning mode aims, on the one hand, for obtaining measurement data and, on the other hand, for optimizing the optical mode. During the learning mode, the system 14 operates in the manner described above to detect near zero order reflected light from the illuminated area and obtain measurement data in the form of spectral intensities at each wavelength within the wavelength band of incident radiation. (Step 5
4). At the same time, processor 20 applies the optical model to obtain theoretical data (step 56) and compares it with measured data (step 58). The optical model is based on some known features of the structure and the nominal values of unknown features provided by the user (ie, the desired parameters to be measured). At this stage, the relationship between the theoretical data and the measured data is compared to certain conditions (step 62). If this condition is satisfied, the parameters A, B, C, and D are calculated (step 64), and an optical model is obtained (step 66). If the condition is not satisfied, the optical model factors A, B, C, D and unknown features are adjusted until the condition is satisfied (step 60).
Although not specifically shown, it should be noted that at this early stage, the desired parameters can be calculated.

Thereafter, the operation in the measurement mode is executed (step 68). Therefore, measurement data and theoretical data are generated simultaneously (steps 70 and 7 respectively).
2). The theoretical data obtained here are converted to known parameters of the structure, ie, the pre-calculated correct values of the optical model factors A, B, C, D, and the nominal values of the desired parameters to be measured. Note that it is based. Similarly, the measured theoretical data is compared with the optimized theoretical data, and it is determined whether the theoretical data satisfies a predetermined condition, for example, whether the fitness has a desired value (step 76). If so, the desired parameters are calculated (step 76). If not, adjustment is made until the theoretical data substantially matches the measured data (step 78). If desired, the measurement mode (step 68) is repeated (step 80) to inspect the next location on the structure 10.

FIGS. 7 and 8 show two examples 110 and 210, respectively, of a patterned structure that can be inspected in the manner described above by the system 14. Each structure 110, 210 is a cell 112, 21 respectively.
Each cell contains two stacks of different layers. The parameters to be measured in these structures are the width of the photoresist layer on top of aluminum,
And the depth of the etch region (atmosphere) of the silicon oxide layer.

FIGS. 9 and 10 show two more examples 310 and 410 of patterned structures whose parameters can be measured in accordance with the present invention. Here, the parameters to be measured are the width, thickness, and chemical mechanical polishing (CM) of the aluminum layer on the silicon oxide.
The remaining thickness of the metal layer on the silicon oxide layer after P).

It should be noted that in structures 310, 410, there is a weak polarization effect in structures 110, 210 due to the presence of the patterned metal.

FIG. 11 shows a patterned structure 510 using copper under and between any of the silicon dioxide based layers known as interlayer dielectric (ILD) insulating layers. When such a copper-based structure 510 is subjected to CMP, a copper deficient portion Pi usually occurs due to a so-called “dishing effect”. This effect is related to the properties of copper (eg, being softer than other metals) and the chemistry of copper-based CMP processes. The parameters to be measured are the depth d1 of the uppermost ILD insulating layer and the depth d2 of the dishing-related portion Pi, respectively. In some cases, depending on the layer stack, the thickness of the metal can be determined by d1-d2.

Those skilled in the art will be able to make various changes and modifications to the invention described above without departing from the scope of the claims. For example, any other number of cells may comprise the patterned structure, with each cell comprising any other number of stacks.

[Brief description of the drawings]

1A is a schematic cross-sectional view of a kind of patterned structure to be measured, and FIG. 1B is a cross-sectional view thereof.

FIG. 2 is a schematic diagram of the main elements of an apparatus for measuring parameters of a patterned structure according to the present invention.

FIG. 3 is a graph showing a main principle of the present invention, and is a diagram showing a relationship between measured data and theoretical data obtained by the apparatus shown in FIG. 2;

FIG. 4 illustrates yet another example of a patterned structure measured by the apparatus of FIG. 2;

FIG. 5 is a flow chart of the main steps of the method according to the present invention.

FIG. 6 is a flowchart of the main steps of the method according to the present invention, following FIG. 5;

FIG. 7 is a cross-sectional view of yet another example of a patterned structure suitable for inspection by the apparatus of FIG.

8 is a cross-sectional view of yet another example of a patterned structure suitable for being inspected by the apparatus of FIG.

9 is a cross-sectional view of yet another example of a patterned structure suitable for being inspected by the apparatus of FIG.

FIG. 10 is a cross-sectional view of yet another example of a patterned structure suitable for inspection by the apparatus of FIG.

11 is a cross-sectional view of yet another example of a patterned structure suitable for being inspected by the apparatus of FIG.

[Explanation of symbols]

10, 100, 110, 210, 310, 410, 51
0 Patterned structure 12 Grid cycle 12a, 12b Element S1 Measurement area 14 Spectrophotometer 20 Processing unit

 ──────────────────────────────────────────────────の Continuing on the front page (72) Inventor David Shiner, Israel, Ganay Jeuda 56905, Iris Street 7 (72) Inventor, Moshe Finarov, Israel, Rehoboat 76209, Sh Kornik Street 4/25 F-term (reference) 2F065 AA24 AA25 AA30 AA54 BB02 CC19 CC31 FF41 FF49 FF51 FF61 GG22 HH04 JJ15 LL12 LL30 LL46 LL67 QQ33 4M106 AA01 BA04 CA19 CA21 CA48 DH03 DH12 DH31 DJ20

Claims (35)

[Claims] At least two adjacent elements (12a, 12a) having different optical properties with respect to incident radiation.
b) a patterned structure (10, 10) representing a grid having at least one cycle (12)
0, 110, 210, 310, 410, 510), wherein at least one desired parameter of a structure having a plurality of characteristics determined by a certain manufacturing process is measured. Based on at least some of the features, theoretical data (Dt) representing spectral intensities of light components of different wavelengths specularly reflected from the structure can be determined and the at least one desired parameter of the structure B) providing an optical model capable of calculating the following: b) Incident radiation in a substantially wide wavelength band by presetting a measurement area (S1) substantially larger than the surface area of the structure determined by the grid cycle. C) detecting a light component substantially specularly reflected from the measurement area (S1), and detecting each wave within the wavelength band. D) obtaining measurement data (Dn) representing the spectral intensity of the above; d) analyzing the measurement data and the theoretical data to optimize the optical model until the theoretical data satisfies a preset condition; and e) Calculating the at least one parameter of the structure when detecting that the preset condition is satisfied. 2. The method according to claim 1, wherein at least some features of the structure on which the optical model is based are obtained before measurement. 3. The method according to claim 1, wherein at least some features of the structure on which the optical model is based include a nominal value of the desired parameter to be measured. 4. The method of claim 1, wherein at least some features of the structure underlying the optical model include a material forming each of the at least two elements. 5. The method according to claim 1, wherein the step of preparing the optical model comprises calculating, in response to the incident radiation, a known optical effect occurring in the structure and a contribution of the optical effect to the detected light component. The method of claim 1, comprising a step. 6. The method of claim 1, wherein said at least two elements are stacks of layers having different optical properties. 7. The at least two elements are a stack of layers having different optical properties, respectively.
The calculating step relates to a) the width of the stack with respect to the wavelength of the incident radiation,
And calculating the specular reflection within the grid cycle associated with the disappearance of the stack having a shape that reduces reflection due to the hollow structure; b) reflected within the transparent layer of each stack and from each stack within the grid cycle Calculate the interference between the light beams, c) Calculate the polarization due to the interaction of the incident radiation with the patterned conductive layer of the grid-like structure, and d) Calculate the effects related to the coherence length of the irradiating radiation. The method of claim 5, comprising: 8. The method according to claim 1, wherein the analyzing step includes comparing the theoretical data and the measured data to obtain data indicating a relationship between the theoretical data and the measured data. 9. The optimizing step includes adjusting a certain variable factor of the optical model until the theoretical data satisfies a preset condition, and obtaining a correct value of the optical model factor. The method of claim 1. 10. The method of claim 9 wherein said certain variable of said optical model determines the contribution of a known optical effect to a detected light component. 11. The method according to claim 1, wherein the preset condition represents a merit function for determining a degree of goodness of fit between measured data and theoretical data. 12. The method of claim 1, wherein said optimizing step comprises adjusting said at least one desired parameter until theoretical data satisfies said preset condition. 13. The method according to claim 1, wherein the measurement area is a part of a structure to be measured. 14. The measurement area is arranged on a test area representing a test pattern similar to a pattern of a structure,
The method of claim 1, wherein the test patterns have identical design rules and layer stacks. 15. The theoretical data is represented by the following equation: Where r1 and r2 are the amplitudes of the reflected signals from the two elements, respectively; A _1p , A _1s and A _2p , A _2s are the size coupling and dissipation effects of the hollow structure and the s of the two elements 2. The method according to claim 1, wherein the filling factor modified according to the effects of -polarization and p-polarization; γ is determined by the degree of coherence. 16. The method according to claim 1, wherein the theoretical data is represented by the following formula: Where ri is i in the at least one cycle.
The amplitude of the reflected signal from the i th element; A _ip , A _is the filling factor modified according to the disappearance of the hollow structure and the effects of the s-polarization and p-polarization of the i-th element in the at least one cycle. The method of claim 1, wherein γ is determined by a degree of coherence. 17. The structure has at least one additional cycle comprised of at least two different adjacent elements arranged along an axis perpendicular to the axis of arrangement of the at least one cycle element. The method of claim 1, wherein the method comprises: 18. The theoretical data is represented by the following equation: 18. The method according to claim 17, wherein RGi, Gi are determined by the reflection intensity from the i-th element and the width of the i-th element; k is the total number of elements. 19. The method of claim 1, wherein said at least one desired parameter to be measured is a width of at least one of at least two neighboring elements during a grid cycle. 20. The method of claim 6, wherein the at least one desired parameter to be measured is a width of at least one of at least two neighboring elements during a grid cycle. 21. The method of claim 6, wherein the at least one desired parameter to be measured is a depth of at least one layer of the at least one stack. 22. The method of claim 6, wherein the at least one desired parameter to be measured is a depth of a metal loss resulting from chemical mechanical polishing performed on the structure. Method. 23. At least two adjacent elements (12a, 12a) having different optical properties with respect to the incident radiation.
b) at least one grid cycle (1
2) a patterned structure (10, 100, 110, 210, 310, 410, 5) representing a grid having
10) An apparatus for measuring at least one desired parameter having a plurality of features defined by a manufacturing process, wherein: i) a measurement substantially greater than a surface area of the structure determined by a grid cycle; A region (S1) is irradiated with incident radiation of a preset substantially wide wavelength band, a specular reflection component of light reflected from the measurement region is detected, and a measurement representing a spectral intensity of the measurement light within the wavelength range is performed. A spectrophotometer (14) for providing data (Dm); b) a processing unit (2) coupled to said spectrophotometer.
0) comprising pattern recognition software and conversion means responsive to the measurement data, locating the measurement, utilizing an optical model based on at least some of the features of the structure, Obtaining theoretical data (Dt) representing the spectral intensity of light that is specularly reflected from the structure inside, calculating the at least one desired parameter, comparing the measured data with the theoretical data, and determining the theoretical data. The parameter measuring device, comprising: the processing unit that determines whether or not the above condition is satisfied. 24. A spectrophotometer, comprising: a spectrophotometer (42).
And a variable aperture stop (3) placed in the optical path reaching the detector.
24. Apparatus according to claim 23, comprising 8), wherein the diameter of the aperture stop varies according to the measured grid cycle of the structure. 25. The apparatus according to claim 23, wherein the measurement area is located within a patterned area of the structure. 26. The apparatus of claim 23, wherein the measurement area is located in a test location located outside a patterned area of the structure. 27. The theoretical data has the following formula: Where r1 and r2 are the amplitudes of the reflected signals from the two elements, respectively; A _1p , A _1s and A _2p , A _2s are the size coupling and dissipation effects of the hollow structure and the s of the two elements 24. Apparatus according to claim 23, wherein the filling factor modified according to the effects of polarization and p-polarization; γ is determined by the degree of coherence. (28) The theoretical data is represented by the following equation: Where ri is i in the at least one cycle.
Th amplitude of the reflected signals from the elements; A _ip, A _is the impact of the size bond and dissipation of the cavity-like structure, and said i-th element in at least one cycle s- polarization and p
24. The device according to claim 23, wherein the filling factor modified according to the effect of polarization; γ is determined by the degree of coherence. 29. The theoretical data is represented by the following equation: 24. The method according to claim 23, wherein RGi, Gi are determined by the reflection intensity from the i-th element and the width of the i-th element; k is the total number of elements. 30. The structure has at least one additional cycle composed of at least two different adjacent elements arranged along an axis perpendicular to the axis of arrangement of the at least one cycle element. 24. The device according to claim 23, characterized in that: 31. The apparatus of claim 23, wherein each of the at least two elements is a stack including layers having different optical properties. 32. The method according to claim 23, wherein the at least one desired parameter to be measured is at least one width (W1, W2) of at least two neighboring elements during a grid cycle. apparatus. 33. The method according to claim 31, wherein the at least one desired parameter to be measured is at least one width (W1, W2) of at least two neighboring elements during a grid cycle. apparatus. 34. The apparatus of claim 31, wherein the at least one desired parameter to be measured is a depth (d1, d2) of at least one layer of the at least one stack. 35. The at least one desired parameter to be measured is a depth (d2) of a metal loss portion (Pi) resulting from chemical mechanical polishing performed on the structure. The method of claim 6.