In the semiconductor industry, one of the greatest recurring challenges is the continuous research and the subsequent production of integrated circuits with smaller and smaller critical dimensions (CD). This term not only defines the semiconductor’s smallest possible feature size, it is also interconnected with the design and implementation of a feasible, advanced, simple, and accurate apparatus for measuring different parameters.
The characterization of these parameters, such as the line width roughness (LWR), the sidewall roughness (SWR), and the line edge roughness (LER), that determine the overall roughness and shape of device patterns for manufacturers is extremely important because they directly influence device performance.
Optical lithography was initially used to create patterns in the semiconductor manufacturing process, but it is severely limited in terms of resolution. Scanning electron microscopy (SEM) with its image analysis software was primarily used to measure these parameters prior to 3D atomic force microscopy (AFM).
This technique can provide substantial advantages such as automation and compatibility with traditional critical dimension SEM tools, but cannot offer the user with high resolution LER data as the SEM resolution reaches its limits, therefore 3D AFM provides a highly desirable alternative solution.
Park 3D AFM can measure LER, SWR, and resist profiles in a highly accurate, cost-effective, and non-destructive manner that has been implemented by leading manufacturers. The accurate and complete characterization of these features is extremely necessary during the pattern transfer process because it helps to image all surfaces of the pattern.
Figure 1. LER, LWR and SWR are the limiting factors of resolution in optical lithography
Non–Contact 3D AFM
The fundamental principle of non-contact 3D-AFM is that the cantilever oscillates rapidly just above the surface of the imaging sample. This offers several advantages compared to traditional contact and intermittent modes. One of its essential advantages is the lack of physical contact between the tip and the sample surface.
As shown in Figure 2, the Z scanner that moves the tip is decoupled from the XY scanner that solely moves the sample, providing brilliant flat scanning and an improved Z scan bandwidth. In addition, the Z scanner can be tilted to gain access to the sidewall of the nanostructures and measure the critical dimensions of top, middle, and bottom photoresist lines whilst simultaneously carrying out roughness measurements along the sidewalls (see Figure 2).
Figure 2. The independent tilted Z-scanner used in 3D-AFM from Park Systems
A conical tip performs data acquisition at predefined tilted angles, typically 0º, a, and -aº. Consequently, 3D patterns can be reconstructed by combining these three scans (Figure 3). This process is called image stitching.
Figure 3. Combination of the three acquired images for 3D AFM pattern reconstruction.
This offers a very precise and excellent method that utilizes the interference pattern of standing waves to measure features, such as the top, middle, and bottom width as well as the total height.
The 3D AFM system can provide advanced three-dimensional imaging of isolated and dense line profiles. AFM is less expensive compared to the alternative methods, such as focused ion beam (FIB) and CD-SEM, for imaging and measuring the line-profile parameters as the preparation of the sample is much simpler.
Noise Levels in 3D-AFM
When handling metrology tools, there is a critical requirement associated with limiting the noise level in the manufacturing environment. Park has thoroughly studied and provided evidence on the correlation of noise levels with productivity.
A noise level study conducted on a 300 mm wafer demonstrated that Park 3D-AFM is a powerful automated nanocharacterization system with high resolution and has the ability to limit the overall system noise levels to less than 0.05 nm (0.5 angstrom) (see Figure 4).
Figure 4.
Roughness Measurements
The quality of the patterns can be described and determined using roughness measurements, as it is possible to transfer roughness into the final etched profile. The unique tilted Z scanner in combination with the low noise levels during an AFM scan process helps deliver accurate sidewall roughness measurements.
Figure 5 shows the 3D AFM image of a photoresist semi-dense line pattern, revealing the rough structure of its sidewalls. The excellent repeatability (0.08 nm 1 sigma for 5 sites wafer mean) for the sidewall roughness of about 6.0 nm validates the precision with which the SWR was measured.
Figure 5. Park 3D AFM image of a photoresist semi-dense line pattern imaged with Z-scanner tilt. The bottom figure clearly depicts the grainy structure of the sidewall.
Roughness relies on the aerial image contrast (AIC) or the physics of exposure, among others. AIC is defined as the quotient between the subtraction and the addition of the maximum and minimum image intensities.
Many consequent series of images with adjustable exposure show that LER can increase considerably when the AIC is reduced. This fact highlights that AIC can be a controlling factor for LER.
The reduced levels of AIC generated profile images of the resist that were more blunt with smaller sidewall angles (SWA) (see Figure 6). Park 3D AFM is capable of imaging all surfaces of the pattern (Figure 7) compared to the standard AFM or the SEM that cannot provide a complete characterization of the surface data and obtain other information, such as top, bottom and sidewall roughness from sidewall characterization.
Figure 6. Park 3D AFM line profiles at different AIC levels reveal the proportionate relationship between SWA and AIC.
Using this system, a 300 nm photoresist line pattern was imaged and the respective line profiles were obtained. From the profiles, a substantial variation in terms of SWR can be observed between 97% and 40% AIC.
Specifically, the lower the value of AIC the greater the measured roughness. This intense decrease in roughness clearly reveals the correlation between LER and the measured sidewall roughness.
Finally, the role of non-contact 3D AFM needs to be highlighted in terms of preserving the sharpness of the cantilever tip. In an independent study, scientists used the same tip to perform 150 consecutive measurements, demonstrating minimal tip wear.
This is a prominent feature of AFM that eliminates continuous expensive replacement of the tip and the damage caused to the sample by the AFM cantilever. As the tip sharpness is preserved, high-resolution roughness data can be continuously measured.
Figure 7. a) A Park 3D AFM image of a 300 nm photoresist line pattern yields full information regarding the morphology of the sidewall. b) Side-wall roughness is different at different AIC levels, a fact that indicates the connection between LER and SWR.
Conclusion
This article has discussed the potential of the innovative, nondestructive imaging methods of 3D AFM over existing SEM system. Clear examples of the features of 3D AFM include the incorporation of an independent and tilted Z-scanner that has been proven to overcome the drawbacks of alternative metrology tools. It does this by measuring parameters such as roughness and detailed sidewall morphology, and providing sidewall angle characterization that results in an easier and detailed optimization and evaluation process.
Park Systems’ 3D AFM has been proven to meet the requirements and challenges of continuously shrinking semiconductor device critical dimensions by providing a solution that includes high precision and accuracy, easy sample preparation, and excellent resolution. 3D AFM sets new benchmarks in the nanotechnology performance and measurement in an industry that is rapidly changing and where new technologies and advancements lead to future opportunities.
This information has been sourced, reviewed and adapted from materials provided by Park Systems Inc.
For more information on this source, please visit Park Systems Inc.