JPH081373B2 - Testing methods for monitoring the properties of reflective patterns - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention is directed to horizontal dimensions (e.g. line width) and at least one other characteristic (e.g. intensity, shade, texture, reflectance) of any of the features of metal forming patterns on substrates. , Or saturation) at the same time.

[0002]

BACKGROUND OF THE INVENTION Much effort is currently being put into the development of multichip electronic modules. Such modules, which are sometimes referred to as hybrid integrated circuits (HICs), generally consist of a substrate, usually ceramic, with one or more layers, each layer usually having a metal such as gold on it. Has a forming pattern. Large Scale Integrated Circuit (VLSI)
Chips to one or both of the major (outer) surfaces of the substrate, with each chip interconnected to the other by a metal forming pattern on the substrate. Selected from the pattern elements of the metal forming pattern of each layer of the substrate are connected to the pattern elements on adjacent layers by through-plated metal conductors.

In order not only to reduce the overall size of such a multichip module, but also to enable very high density circuits, the horizontal dimensions (eg line width) of the pattern elements of the pattern are Make it very small. The linewidth of the pattern elements is in the range of 2 to 10 mils (1 mil is 0.001 inch) depending on the type of process used to deposit the metal on the substrate. Maintaining precise control of the dimensions of every pattern element in each pattern is extremely important. If the dimensions of a particular pattern element in a pattern exceed the maximum allowable under valid design rules, the signal is much more likely to leak between such pattern element and its adjacent pattern elements. And the possibility of “crosstalk” arises. Conversely, if the size of a particular pattern element in the pattern is too small, the impedance of the signal path provided by that pattern element will exceed acceptable limits, adversely affecting the operation of the multichip module.
Breaking of the elements in the pattern is obviously also undesirable. In addition, the quality of pattern elements such as through conduits is also important.

Not only the horizontal dimension of the pattern elements in each pattern on the substrate, but also the overall quality of the forming metal in the pattern is important. In the process of depositing the pattern on each layer of the substrate, the metal formed in the pattern may be contaminated. Such contamination not only adversely affects the impedance of the circuit established by the pattern, but also adversely affects the ability to securely bond the integrated circuit chip to the forming metal on the exposed surface of the substrate. Another important parameter is the height uniformity of the pattern elements in the pattern. A significant change in the height of the pattern element is
It can adversely affect the quality of the electrical connections made to the components located on the substrate.

Inspection of metallization patterns on a substrate to detect unwanted changes in horizontal dimensions has hitherto been made by comparing an image of the pattern (captured by a television camera) with each of a pair of master patterns or models. Has been done by doing. The generation of such a model pair is such that the pattern elements of one model are packed (reduced) in the horizontal dimension to represent the maximum and minimum allowable values respectively, and the pattern elements of the other model are packed in the horizontal dimension. Except for some expansions, it is done so that each model contains an exact replica of the desired pattern of pattern elements. By comparing an image of the actual pattern of the forming metal with each of the two models and logically combining the results of the comparison, pattern elements that are oversized or undersized can be detected.

The above-described technique has been devised to perform an automatic pattern inspection for detecting a change in the horizontal dimension of the pattern element. However, in addition to the horizontal dimension, the height, color tone, strength and reflectance of the pattern element are also considered. No comparable method has been devised for automatically inspecting one or more important attributes of pattern elements, such as, texture, saturation, or combinations thereof. If there is a test for any of these attributes, the test is done manually. For example,
Inspecting the strength of pattern elements to determine the tone and shade of the pattern is usually done manually and is at best subjective. In addition, the results may be affected by the fatigue of the person in charge.

[0007]

The problem sought to be solved by the invention is therefore the automatic inspection of a metallization pattern on a substrate so as to detect not only changes in horizontal dimensions but also changes in at least one property of the pattern element. Is to give a way to do.

[0008]

Briefly, in addition to the horizontal dimension, one or more of the importance of the pattern element's reflectance, intensity, height, shade, texture, saturation, or combinations thereof, is important. A method of inspecting a pattern on a substrate, such as a metallization pattern on a multichip module, to monitor various attributes is also disclosed by the present invention. This method is typically performed by first capturing a gray scale image of the pattern with a television camera. Other important characteristics may require further processing of the image to develop information about that characteristic. For example, if height is another important attribute (besides the x and y horizontal dimensions of the pattern element), it is necessary to process the image of the pattern to develop a depth map using known techniques.

After processing the image as needed, the first includes a pattern image that is respectively scaled and scaled to the horizontal dimension of the pattern element and scaled to values of other important attributes (eg, intensity), respectively. And comparing the second model with the captured image. The results from such comparisons are logically combined and the results of this combination are used to determine if the dimensions and other important attributes (eg, strength) of the pattern elements are within acceptable limits.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 1 shows a substrate 12 containing a forming metal pattern 14.
1 is a schematic block diagram of a system 10 for inspecting (For purposes of explanation, the substrate 12 and the metal forming pattern 14 thereon are
The thickness is exaggerated. The system 10 comprises an xy table which not only supports the substrate 10 but also serves to move it precisely along the x and y axes. The movement of the xy table is controlled by the xy controller.

The television camera 20 is optically coupled by a microscope 21 aiming at an xy table so as to capture an image of the substrate 12 and the metal forming pattern 14 thereon. Generally, due to the optical conditions of the camera 20 and the microscope 21, only a small portion of the pattern 14 is within the field of view of the camera. By moving the xy table in stages, different parts of the pattern 14 can be brought into the field of view of the camera 20.
A ring lamp 22 driven by a light source 24 is provided to illuminate the surface of the substrate 12. Another lamp (not shown) may be provided to illuminate below the surface of the substrate.

The output signal of the camera 20 is the image system 2
6 is passed through the first television monitor 25 before being fed. In the exemplary embodiment, the imaging system is an IRI model P25.
6 or P512, both available from IRI, Carlsbad, CA.
The imaging system 26 serves to process the images of the camera 20 as well as to control the movement of the xy table by providing control signals to the xy controller 18. A second television monitor 28, similar to monitor 25, is coupled to the output of image system 26 for displaying the images produced by image system 26. Image system 28
Are generally programmed through terminal 30.

As will be described in further detail below, the imaging system 26 includes a horizontal dimension and at least one other characteristic of the pattern elements of the metal forming pattern 14 on the substrate 12 (eg, strength, height, reflectivity, The images captured by the camera are processed to determine if the texture, hue, saturation, and combinations thereof) are acceptable, and the characteristics to be inspected are quantified. Except for the height of the pattern elements of pattern 14, the actual values of the other properties listed above can be obtained directly from system 10. To obtain height information, see US Patent Application No. 440,
No. 948 (Inventor: I. Amir) Applicant:
A system as disclosed in AT & T) may be used.

In the exemplary embodiment, system 10 is designed to monitor the horizontal dimension of pattern 14 as well as its strength. FIG. 2 shows a partial image of the metal forming pattern 14 captured by the camera of FIG. In the portion of the metal forming pattern 14 shown in FIG. 2, there is an extremely bright region 34 near the upper part of the figure, and an extremely dark region 35 near the lower part of the figure.
Included is a broken stripe 34 having. In the pattern 14 shown in FIG. 2, in addition to the cut stripes 34, an extremely bright irrelevant metal forming region 36 is formed on the left side of the cut stripes 32.
Also included to the right of 2 is an extremely dark extraneous metal forming region 38. Extremely bright areas 34 and extremely dark areas 3
5 (as well as regions 36 and 38) represent regions within the pattern portion of FIG. 1 whose strength is out of tolerance, for example due to contamination of the forming metal. If there is no defect in the portion of the metal forming pattern 14 shown in FIG.
Should be of consistent strength and should be free of extraneous regions 36 and 38 of forming metal.

When the output signal of the camera of FIG. 1 is supplied,
The imaging system of FIG. 1 serves to determine the relative intensities of each of the plurality of small pixels (pixels) that make up the captured image shown in FIG. FIG. 3 is a map of the intensities of the pixels in the image of FIG. 2 measured on a 0 to 255 scale. As can be seen, the metal-free regions within the image of FIG. 2 are generally in the range of 48 to 51 intensities in FIG. In FIG. 2, the metallized region (eg, the region consisting of stripes 32) is shown in FIG.
Is in the range of 12 to 40, with lower values representing extremely dark areas and higher values representing extremely bright areas.

In accordance with the present invention, changes in horizontal dimensions and / or changes in strength of the metal forming pattern 14 on the substrate 12 of FIG. 1 are first shown in FIGS. 4 and 5 if the specified tolerances are exceeded. It can be detected by generating two such models 40 and 42 respectively. Model 40 represents an acceptable, and preferably perfect, metallization pattern 14 that is typically 10% compacted (reduced) in x and y dimensions, but with its reflection intensity expanded (enlarged) by 10%. There is. Note that the reduction in the y direction is not shown in FIG. Conversely, model 42 represents an acceptable and complete metal forming pattern 14 (ie, unbroken stripes 32) with a 10% expanded horizontal dimension. Its reflection intensity is reduced by 10%. Model 40 of FIGS. 4 and 5
The 10% change in line width and reflected intensity at and 42
Represents the maximum allowable horizontal dimension change and strength, respectively. Increasing the allowed tolerances requires some adjustment of the horizontal dimension variations and intensity values in models 40 and 42. It should be noted that the allowance for the horizontal dimension in the x and y directions and the allowance for the strength need not be the same.

FIGS. 6 and 7 are pixel intensity maps corresponding to the models 40 and 42 of FIGS. 4 and 5, respectively. Areas that appear dark and areas that appear bright in models 40 and 42 of FIGS.
Respectively corresponding to the small and large intensity values at. The pixel intensity values associated with the actual metallization area (stripe 32) in FIGS. 6 and 7 differ by 10% from the corresponding intensity values for the stripes in FIG.
The models 40 and 42 of FIGS. 4 and 5 represent acceptable and reduced patterns of the metal forming pattern 14 of FIG. 2 respectively, so that the metal forming regions 36 and 38 irrelevant to either model (see FIG. 2). Does not exist.

At the beginning of the test, the model 40 of FIGS.
And 42 are each the mechanical imaging system 26 of FIG.
Is input to Models 40 and 42 generally have acceptable (preferably defect-free) metal forming patterns 14.
Are imaged and the resulting pixel map is stored in the imaging system 26. This map is then manipulated to give the maps shown in FIGS. 6 and 7. The values of this map are adjusted to effectively scale and scale the horizontal dimension, respectively, and strength and scale, respectively.

Once the intensity maps of FIGS. 6 and 7 (representing models 40 and 42, respectively) have been established, inspection of substrate 12 is actually performed by capturing at least a portion of metal forming pattern 14 with the camera of FIG. To be done.
Then, an intensity map of the pixels of the captured image (see FIG. 3) is established. After the intensity map of the pixels of the imaged part of the pattern 14 of FIG. 1 has been established, the map is
It is compared with each of the pixel intensity maps (see FIGS. 6 and 7) associated with corresponding portions of models 40 and 42, respectively.

The comparison above compares the pixel intensity map of FIG. 3 (corresponding to the actual or "run-time" image of FIG. 2) and the map of FIG. 6 (corresponding to model 40 of FIG. 4) (from the former). This is done by logically combining the former and the latter by subtracting the map of FIG. 3 from the map of FIG. 7 (corresponding to the model 42 of FIG. 4), as well as logically combining the latter. The following results of these separate operations are logically combined (by addition) to give a set of intensity values corresponding to those of the image represented in FIG. Only the extremely bright regions 34 and the extremely dark regions 35 in FIG. 2 and the extraneous metal forming regions 36 and 38 collectively representing defects in the portion of the metal forming pattern 14 shown in FIG. 2 appear in FIG. is there. Pattern 14 of FIG.
The image shown in FIG. 8 should appear perfectly bright (completely blank) if there is no defect in the area.

The way in which the intensity map of FIG. 3 is compared with the intensity maps of FIGS. 6 and 7 depends on the contrast between the metal forming pattern 14 and the surrounding background. If the pattern is dark and the background looks light, the results of the comparisons are combined as described. On the contrary, when the pattern looks bright and the background looks dark, the intensity map of FIG. 7 is subtracted from the map of FIG. 3, and further the map of FIG. 3 is subtracted from the map of FIG.

The above description relates to one embodiment of the present invention, and various modifications of the present invention are conceivable to those skilled in the art, but all of them are within the technical scope of the present invention. Included in.

It should be noted that the reference numbers in the claims are not to be construed as limiting the technical scope of the invention for easy understanding of the invention.

[0024]

As described above, according to the present invention, each pattern of the pair of models 40 and 42 including the master copy of the metal forming pattern 14 on the substrate 12 is reduced and expanded, respectively. Both the horizontal dimension and the strength of the pattern can be monitored by comparing the horizontal dimension and the expanded and contracted strength respectively. still,
Although the system of FIG. 1 has been described as monitoring strength as one of the important pattern 14 attributes in addition to the horizontal dimensions of the pattern elements, similar techniques can be used to monitor other attributes. . In other words, to monitor other characteristics such as height, reflectance, texture, tint, saturation, or combinations thereof, the horizontal dimensions are scaled up and down and the value of the characteristic to be monitored is appropriate. Model 4 by enlarging and reducing by various tolerance factors
Generate two separate models similar to 0 and 42 respectively. Next, the actual images of pattern 14 are
The two models are compared and the results of the comparisons are combined as described to generate an image containing only the defects, if any.

[Brief description of drawings]

FIG. 1 is a schematic block diagram of a system for monitoring horizontal dimensions of pattern elements and at least one other important attribute of a metal forming pattern on a substrate.

2 is a schematic representation of a reflective metal forming pattern captured by the system of FIG.

FIG. 3 is a map containing intensity values of the image of FIG.

FIG. 4 represents an image of a first model used for comparison purposes with the image of FIG.

5 represents an image of a second model used for comparison purposes with the image of FIG.

FIG. 6 is a diagram showing an intensity map of the image of FIG.

7 is a diagram showing an intensity map of the image of FIG.

FIG. 8 is a representation of an exemplary image resulting from the logical combination of the intensity maps of FIGS. 3, 6 and 7.

[Explanation of symbols]

 10 System of the present invention 16 xy table 18 X-y controller 20 TV camera 26 Imaging system

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification number Internal reference number FI Technical indication location H01L 21/66 J 7514-4M (72) Inventor David Robert Lease United States 01835 Massachusetts Bradford, Fair Wood Drive 53 (56) Reference JP-A-1-121709 (JP, A) JP-A-61-139706 (JP, A)

Claims (5)

[Claims] 1. A step of: (a) capturing an actual image of the pattern 14 to monitor the horizontal dimensions of the pattern elements and at least one other important characteristic of the reflective pattern 14 on the substrate 12; ) The model and the capture representing an image of a comparison pattern with the value of the key attribute expanded to the maximum allowable value in the first model 40 and the horizontal dimension of the pattern element reduced to the minimum allowable value. (C) a comparison pattern having horizontal dimensions of the pattern element reduced to the minimum allowable value in the second model 42 and expanded to the maximum allowable value in the second model 42; Comparing the captured image with the model representing the image of (d) comparing the captured image with each of the first and second models. Logically combining the results obtained by: (e) determining according to the comparison result whether the horizontal dimension and the important characteristics of the pattern 14 are within defined tolerances. An inspection method for monitoring the characteristic of a reflective pattern, which comprises: 2. The inspection method according to claim 1, wherein the important characteristic is strength of a pattern element of the pattern. 3. Logically combining the comparison results by subtracting the captured image from the image of the first model and further subtracting the second model from the captured image. The inspection method according to claim 1, characterized in that 4. The step of the first model 40 capturing an image of a comparison pattern having pattern elements corresponding to the pattern elements in the pattern 14 on the substrate, the horizontal dimension of the pattern element in the image capturing the comparison pattern Established by the steps of reducing by an amount corresponding to a minimum acceptable value, and expanding the value of said important characteristic of the pattern element in the image capturing the comparison pattern by an amount corresponding to the maximum allowable value. The inspection method according to claim 1, wherein: 5. The step of the second model 42 capturing an image of a comparison pattern having a pattern element corresponding to the pattern element in the pattern on the substrate, allowing the horizontal dimension of the pattern element in the image capturing the comparison pattern. Established by expanding by an amount corresponding to the maximum possible value, and reducing the value of said important characteristic of the pattern element in the image capturing the comparison pattern by an amount corresponding to the minimum allowable value. The inspection method according to claim 1, wherein: