Patent Application
20090296075
The present invention relates to overlay metrology, and in particular to diffraction based overlay metrology.
Semiconductor processing for forming integrated circuits requires a series of processing steps. These processing steps include the deposition and patterning of material layers such as insulating layers, polysilicon layers, and metal layers. The material layers are typically patterned using a photoresist layer that is patterned over the material layer using a photomask or reticle. Typically the photomask has alignment targets or keys that are aligned to fiduciary marks formed in the previous layer on the substrate. However, as the integrated circuit feature sizes continue to decrease to provide increasing circuit density, it becomes increasingly difficult to measure the alignment accuracy of one patterning step to the previous patterning step. This overlay metrology problem is becoming particularly difficult as overlay alignment tolerances required to provide reliable semiconductor devices are getting increasingly tighter.
Conventional overlay metrology uses imaging of non-diffraction based targets, such as Box-in-Box or Bar-in-Bar targets. These imaging targets typically include a large structure on one layer and a smaller structure on a different layer. The centers of the larger and smaller structures should coincide when the layers are properly aligned. This conventional overlay metrology technique, however, requires high magnification imaging and suffers from disadvantages such as optical distortions and sensitivity to vibration. Moreover, conventional imaging devices suffer from a trade-off between depth-of-focus and optical resolution. Additionally, edge-detection algorithms used to analyze these images for the purpose of extracting overlay error are inaccurate when the imaged target is inherently low-contrast or when the target suffers from asymmetries due to wafer processing.
Diffraction based overlay metrology utilizes overlying gratings that diffract incident light. Data is acquired, e.g., in the form of spectra, from multiple individual pads in the overlay target. The resulting spectra from each pad can then be compared and used to determine the overlay error. Conventionally, diffraction based overlay targets must be large enough that the measurement spot can only be incident on one individual pad at a time because each pad must be measured individually. Accordingly, the diffraction based targets have an undesirably large footprint on the sample. Additionally, due to the time associated with the acquisition of multiple pads, conventional diffraction based overlay measurements have a relatively low throughput.
Thus, there is a need in the semiconductor industry for improved overlay metrology techniques.
An overlay error is determined using a diffraction based overlay target by forming multiple images of the overlay target using different narrow ranges of wavelengths. The images can be used to construct spectra for the overlay target and the spectra is used to determine overlay error.
In one embodiment, a plurality of sample beams is generated, each beam having a different range of wavelengths. A diffraction based overlay target is imaged for each sample beam resulting in a plurality of images of the diffraction based overlay target. An intensity value for each of the plurality of images is determined and used to construct a spectrum. The constructed spectrum is then used to determine the overlay error, which is then recorded.
In another embodiment, an apparatus for measuring overlay error includes a light source that produces a plurality of light beams having different ranges of wavelengths, an optical system configured to illuminate a diffraction based overlay target and an image detector positioned to receive images of the diffraction based overlay target illuminated by each of the plurality of light beams having different ranges of wavelengths. A processor is coupled to the image detector and receives the images of the diffraction based overlay target illuminated by each of the plurality of light beams. The processor includes a computer-readable storage medium storing a computer program executable by the processor and the computer program includes instructions for determining an intensity value for each of the plurality of images; combining the determined intensity values for each of the plurality of images to produce a constructed spectrum; using the constructed spectrum to determine an overlay error; and recording the overlay error.
In another embodiment, the overlay error is determined using a diffraction based overlay target that includes a plurality of measurement pads. In this embodiment, narrow band illumination beams each having a different range of wavelengths are repeatedly generated. Each narrow band illumination beam simultaneously illuminates the plurality of measurement pads in the diffraction based overlay target. The plurality of measurement pads in the diffraction based overlay target are repeatedly imaged to produce an image of the plurality of measurement pads for each range of wavelengths. An intensity value for each measurement pad in each image is determined for each range of wavelengths and used to construct spectra for each measurement pad. The constructed spectra for each measurement pad is then used to determine the overlay error, and the overlay error is recorded.
In another embodiment, after determining the optimal range of wavelength for a specific application, a band pass filter or equivalent is inserted in the illumination light path to allow this range of wavelength to illuminate the target. An intensity value for each measurement pad is determined for this range of wavelengths. Using the intensity values for the measurement pads, the overlay error is determined and recorded.
In accordance with an embodiment of the present invention, overlay errors between patterning steps are measured by imaging the diffraction based overlay target.
Metrology device 100 uses a broadband light source 102 that generates a broadband light beam 104. By way of example, broadband light source 102 may be a Xenon lamp or other appropriate light supply or supplies that produce a desired broad range of wavelengths. A narrow band pass filter, such as monochromator 106 receives the beam 104 and generates a narrow band illumination beam 108, which has a narrow band of wavelengths. As illustrated in
The narrow band beam 108 is partially reflected by a beam splitter 114 towards the sample 200 having a diffraction based overlay target 202 with a plurality of measurement pads 204. The sample 200 may be, e.g., a semiconductor wafer or flat panel display or any other substrate, and is supported by a stage 116, which may be a polar coordinate, i.e., R-θ, stage or an x-y translation stage. The beam 108 is focused at normal incidence by an optical system, such as lens 118 (or series of lenses), to form an illumination spot 119 on the sample 200. In one embodiment, the spot 119 is large enough to cover a plurality of or all of the measurement pads 204 in the overlay target 202. The lens 118 may have a high NA, such as 0.5 to 0.7. If desired, polarized light may be used, e.g., using polarizer 115. Alternatively two polarizers may be used, one between the light source 102 and the beam splitter 114 and another between the beam splitter 114 and detector 120.
The image 121 of the target 202 is resolved on a detector 120 by lens 118 through the beam splitter 114, as illustrated by beam 117. Additional lenses, e.g., between the beam splitter 114 and the detector 120 may be used to resolve the image 121 on the detector 120 if desired. The detector 120 is a two dimensional photodetector array, such as a high speed CCD array, CMOS array, or other appropriate device. The image of the target 202 is received by a processor 122 and stored in memory 122s. The processor 122 includes a computer-usable medium 122i having computer-readable program code embodied therein for causing the processor to control the device 100 and to perform a desired analysis, as described herein. The data structures and software code for automatically implementing one or more acts described in this detailed description can be implemented by one of ordinary skill in the art in light of the present disclosure and stored on a computer readable storage medium, which may be any device or medium that can store code and/or data for use by a computer system such as processor 122. The computer-usable medium 122i may be, but is not limited to, magnetic and optical storage devices such as disk drives, magnetic tape, compact discs, and DVDs (digital versatile discs or digital video discs). The processor 122 also includes storage 122s and a display 122d for storing and/or displaying the results of the analysis of the data.
The illuminated overlay target 202 is then imaged using the first range of wavelengths (block 154). The diffraction based overlay target 202 includes multiple measurement pads 204, which may be imaged simultaneously as illustrated in
The range of wavelengths is then changed so that a different range of wavelengths is produced (block 156). The diffraction based overlay target 202 is again imaged time using the different range of wavelengths (block 158). The range of wavelengths is changed and additional images are taken of the overlay target until images have been formed for all desired ranges of wavelengths (block 160).
An intensity value for each measurement pad 204 in each image is determined (block 162). If desired, the intensity value for each measurement pad 204 may be determined prior to or after imaging the overlay target at a different range of wavelengths. In one embodiment, the intensity value of each measurement pad 204 is determined by summing the intensities of each pixel in an image of the measurement pad 204. Of course, if desired other statistical techniques may be used to generate the intensity for each pad for each image, such as finding the median or mean of the intensities of the pixels, or other similar techniques. In one embodiment, the intensity value for each measurement pad 204 is determined using less than all the pixels in the image of the measurement pad 204. By way of example, the central 50% to 90% of the pixels in the image of a measurement pad 204 may be used.
Once an intensity value for each measurement pad 204 in each image is determined, the spectra for each measurement pad is constructed (block 164). The constructed spectrum for each measurement pad consists of the determined intensity values I for each wavelength range λ corresponding to each image.
Once the spectra for each measurement pad 204 is constructed, the overlay error can be determined using known methods (block 166) and the results are recorded (block 168), e.g., by storing in memory, such as storage 122s (
The method of determining overlay error depends on the type of metrology target used.
When the top layer 208 is perfectly aligned with the bottom layer 206, the top pattern 209 will be offset slightly with respect to the bottom layer 207. The offset of each measurement pad 204 is different in magnitude and/or direction. By way of example, measurement pad 204a has an offset that has a magnitude of D towards the right, referred to herein as +D, while measurement pad 204b has an offset of the same magnitude by towards the left, referred to herein as −D. Measurement pads 204c and 204d include the same magnitude offset, i.e., D, with a reference offset. Thus, measurement pad 204c has an offset towards the right with a magnitude of +D+d, and the measurement pad 204d has an offset that is the same magnitude, i.e., |D+d|, but in an opposite direction towards the left, and is referred to herein as −D−d. The magnitude of the reference offset d can be fairly small, e.g., approximately 1% to 15% and in particular 5% of the pitch of the patterns. Of course, the precise magnitude and direction of designed in offset D and reference offset d may be varied to suit the particular materials and dimensions of the overlay patterns, along with the wavelength or wavelengths of light used by the metrology equipment.
With the use of metrology target 202 show in
where Ra, Rb, Rc, and Rd are the intensities at selected wavelengths of the constructed spectra for measurement pads 204a, 204b, 204c, 204d and d is the absolute value of the reference offset.
If desired, different types of overlay targets may be used with the present invention. For example, U.S. Pat. No. 6,982,793, which is incorporated herein by reference, describes the overlay target 202 shown in
By way of example,
In general, to measure the alignment error e it is necessary to determine the change in the diffracted light with respect to the change in alignment error. This may be written as follows:
where R is the measured light at selected wavelengths and e is the alignment error. The factor φ for an overlay target maybe determined using, e.g., modeling techniques or using additional measurement locations as reference locations, as discussed above. Once the factor φ is determined, the value of the overlay error e can then be determined using the following equation.
Ra−Rb=2eφ eq. 4
where Ra and Rb are the intensities at selected wavelengths of the constructed spectra for measurement pads 302 and 304.
The illuminated overlay target 202 is then imaged using the optimal range of wavelengths (block 406). As discussed above, the diffraction based overlay target 202 includes multiple measurement pads 204, which may be imaged simultaneously as illustrated in
Although the present invention is illustrated in connection with specific embodiments for instructional purposes, the present invention is not limited thereto. Various adaptations and modifications may be made without departing from the scope of the invention. Therefore, the spirit and scope of the appended claims should not be limited to the foregoing description.
