The present application is based on, and claims priority from, Taiwan Application Serial Number 94105298, filed Feb. 22, 2005, the disclosure of which is hereby incorporated by reference herein in its entirety.
1. Field of Invention
The invention relates to a highly sensitive defect detection method and, in particular, to a method for detecting bottom defects of a device pattern in semiconductor wafers.
2. Related Art
Semiconductor processes employ complex steps such as deposition, etching, chemical mechanical polishing (CMP), to form devices (e.g. transistors, metal wires, and capacitors) with various kinds of shapes and patterns. However, there may be minute particles or residues left on the surface or bottom of the device pattern in these processes. Such residues or particles are considered as defects in the device pattern and may affect the performance and quality of semiconductor devices.
Therefore, the defect detection of semiconductor devices is an important step in the fabrication. A good defect detection method can find out small defects on the device pattern to ensure the semiconductor product quality and to provide process engineers with faithful data for adjustments.
The conventional defect detection method uses a defect detecting system to project an incident beam at a particular angle on the semiconductor wafer and scan through the whole wafer. When the beam hits a defect on the device pattern, the incident light is scattered. Several detectors are disposed around the device to receive the light reflected by the surface of the device pattern and scattered by the defect. Such a signal is compared with that from the neighboring regions of the same wafer, thereby determining the position and number of defects. For example,
However, the defect 108 on the surface of the device pattern 102 can be directly hit by the incident light, whereas the bottom defect 110 receives less light due to the height and pitch of the pattern. As a result, the intensity of the light scattered from the bottom defect 110 is much weaker than that from the surface defect 108. Therefore, it is harder to detect the bottom defect 108 of the device pattern 102. If one increases the intensity of the incident light 104 for enhancing the intensity of the light scattered from the bottom defect 110 of the device pattern 102, the intensity of the light scattered from the surface defect 108 is also increased. Moreover, the noise intensity is enhanced also. Thus, the signal-to-noise (S/N) ratio is reduced. In the end, it is still very difficult to detect the bottom defects with high sensitivity. The detection result may be even worse than before.
On the other hand, there is a relation between the device resistance and its dimension:
R=p×1/(h×w).
As the size of the semiconductor device pattern is further reduced while the coefficient of resistance p and the length of the device pattern 1 remain the same, one has to increase the height h if the width w is to be reduced so that the resistance R is not increased. We take a metal wire as an example in
Therefore, we need a defect detection method that is not only able to detect defects on the surface of a device pattern, but also sensitive to defects at the bottom of the device pattern.
It is an object of the invention to provide a highly sensitive defect detection method, particularly a method for detecting bottom defects on a device pattern of a sample.
In accord with the above object, a preferred embodiment of the invention forms a medium on a sample with several device patterns. A beam of light is projected by a defect detecting system on the medium-covered sample. The light penetrates through the medium and scans through the sample surface. At least one detector is used to receive the scattering light and reflected light from the sample to obtain the intensity of the reflected and scattering light of each point on the sample. Afterwards, the intensity of the reflected and scattering light from a few consecutive device patterns is compared with one another to determine the defects in them.
The sample is a substrate with semiconductor devices disposed on its surface. In particular, it is a semiconductor wafer having device patterns with a high aspect ratio. The medium coated on the sample can be a material with a refractive index greater than 1, transparent for incident light, and not affecting the sample. Such materials can be, for example, the dielectric material (e.g. silicon oxide) in a semiconductor process, the photo resist, or any liquid that does not erode, oxidize, or dissolve the sample (e.g. deionized water and mineral oils). The defect detecting system can be any existing optical defect detecting system.
An advantage of the invention is to increase the ability of a conventional defect detecting system to detect defects at the bottom of device patterns on a sample.
Another advantage of the invention is to solve the problem in the conventional defect detecting method which, when the device patterns get smaller, is unable to detect bottom defects of device patterns with high sensitivity because either the signals of bottom defects in a device pattern are weak or the signals of bottom defects are interfering with the signals of the surface defects or noises.
A further advantage of the invention is that the invention can be conveniently used in various kinds of conventional defect detecting systems without the need to make complicated modifications.
These and other features, aspects and advantages of the invention will become apparent by reference to the following description and accompanying drawings which are given by way of illustration only, and thus are not limitative of the invention, and wherein:
The present invention will be apparent from the following detailed description, which proceeds with reference to the accompanying drawings, wherein the same references relate to the same elements.
The disclosed method forms a medium on a sample. A defect detecting system then detects defects on the sample. The defect detecting system projects a beam of light on the sample and scans through each point of the sample after the light penetrates through the medium. At least one detector of the defect detecting system is used to receive light scattered and reflected from each point on the sample. The intensity of the scattered and reflected light of all points is compared to find out at least one defect or a defect distribution on the sample.
In the following, we describe the disclosed method with reference to the appropriate drawings. As shown in
In accord with the invention, an appropriate medium is selected to cover the sample. The medium fills the gap between device patterns 302 or the concave portion, as shown in the right plot. The incident light 304 projected by the defect detecting system enters the medium in the gap from the surface of the device pattern 302. When the refractive index of the used medium is greater than that of the original medium (usually air) on the surface of the device pattern 302, the incident beam 304 with an incident angle θ changes its direction of propagation toward the normal 305, as indicated by the arrow. The incident light 304 then proceeds in the direction with an outgoing angle θ1. The change in its propagation direction enables the incident beam 304 to reach the defect 311 at the bottom of the device pattern 302 in fewer reflections. The intensity of the beam at the bottom defect is less attenuated. Therefore, the incident light projected at the defect 311 and the light scattered from it are stronger in intensity than from the defect 310. The sensitivity of the defect detecting system on the bottom defects of device patterns is thus enhanced. Please refer to
90°>θ1>[90−tan−1(h/d)]°.
However, the angle between the incident beam and the horizontal line in a usual defect detecting system is normally greater than 30°. In other words, the incident angle is usually less than 60°. Therefore, the range of the incident angle θ1 is between 60° and [90−tan−1(h/d)] degrees; i.e., 60°>θ1>[90−tan−1(h/d)]°.
If the incident light cannot directly reach the bottom of the device pattern 402, then the number of reflections T experienced by the incident light can be expressed as:
T=h/(d×cotθ2) (1)
Deflection angle θ2=cot−1[h/(T×d)] (2)
Table 1 shows the relation between the different generations of manufacturing processes and the number of reflections. It can be seen that when the incident angle of the detecting beam is 60° (i.e. when the angle between the incident beam and the horizontal line is 30°), the number of reflections of an incident beam increases as it proceeds in the device pattern. Thus, this inevitably increases the difficulty in detecting the bottom defects.
To solve the problem of increasing number of reflections for smaller device patterns, one can use the following formulas to determine a medium material with an appropriate refractive index to reduce the number of reflections during detection. As shown in
n1×sin θ1=n2×sin θ2. (3)
With the help of Eq. (2), one can compute the refractive index of the medium on the sample as follows:
n2=n1×sin θ1/sin {cot−1[h/(T×d)]}. (4)
The refractive index of air is about 1. Therefore, the refractive index of the medium is:
n2=sin θ1/sin {cot−1[h/(T×d)]}. (5)
Since the number of reflections T≧1, the refractive index of the medium on the sample can be expressed in another way as:
1<n2<sin θ1/sin {cot−1[h/(d)]}. (6)
Besides, the fewer reflections the incident light has, the less attenuation the incident light has. Take a device size of 0.12 μm as an example. If, with respect to the normal, the incident angle θ1 is 60° and the aspect ratio of the device h/d=4.13, the reflection occurs approximately 8 times; To enhance the sensitivity in detecting the bottom defects of the device pattern, the number of reflections has to be less than 8. For example, if the number of reflections of the incident light is expected to be between 3 and 4, then one should use a medium with a refractive index between 1.24 and 1.47.
Therefore, one can pick an appropriate medium according to the device aspect ratio (h/d) of the sample, the number of reflections an incident beam has (e.g., from 0 to 7), and the incident angle of the detecting beam from the selected defect detecting system. The sample can be a substrate with semiconductor device patterns formed on its surface. In particular, it is a semiconductor wafer having device patterns with a high aspect ratio. The medium covering the sample can be any material with a refractive index of n1×sin θ1/sin {cot−1[h /(T×d)]} greater than 1, transparent to the incident beam, and not affecting the sample. Such materials include the dielectric materials (e.g. silicon oxide), photo resist, or a liquid that does not erode, oxidize, or dissolve the sample (e.g., deionized water and mineral oils) commonly used in semiconductor processes. The defect detecting system can be any conventional optical defect detecting system. The method of forming the medium on the sample can be coating, depositing, or any other approach.
In the following, we use a preferred embodiment to explain how the disclosed method is employed to detect defects on a semiconductor wafer. As shown in
It is seen from the preferred embodiment that the invention has the following advantages. First, the invention can avoid or reduce the problem of attenuating energy due to repeated reflections of the incident light in the gap of device patterns.
Another advantage of the invention is that the ability of detecting the bottom defects in device patterns can be enhanced without affecting the signal from the surface defects of the device pattern.
A further advantage of the invention is to avoid complicated modifications to the existing defect detecting system.
Yet another advantage of the invention is to solve the problem that it is difficult for a conventional defect detecting system to detect bottom defects in device patterns as the size of the device patterns gets smaller.
While the invention has been described by way of example and in terms of the preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments. To the contrary, it is intended to cover various modifications and similar arrangements as would be apparent to those skilled in the art. Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.
Number | Date | Country | Kind |
---|---|---|---|
94105298 | Feb 2005 | TW | national |