Tengjiao Hu, Ronald L. Jones, Eric K. Lin, Wen-li Wu
Polymers Division, National Institute of Science and Technology, Gaithersburg, MD 20899-8541
Nanometer scale control of lithographic pattern quality is required as the semiconductor industry moves to the mass production of sub-100nm semiconductor devices. Scanning electron microscopy (SEM) has been the mainstay metrology in lithography. Other methods, such as atomic force microscopy (AFM) and light scatterometry, have been proposed as alternate metrologies to measure the pattern quality. Unfortunately, each of these metrologies is expected to face challenges as device features continue to shrink. The detailed characterization of line edge roughness (LER) and sidewall angle of nanopatterns is listed as one of the five grand challenges in metrology in the new release of International Technology Roadmap for Semiconductors 2004.
Recently, we have been developing a new technique based on traditional X-ray crystallography and diffuse X-ray scattering, termed Critical Dimension Small Angle X-ray Scattering (CD-SAXS) to meet such a demand. CD-SAXS operates in transmission mode, in contrast to the reflective geometry of other metrologies, requiring only a simple data analysis. Earlier results show that CD-SAXS provides a fast, non-destructive measurement of pitch size, pattern width.
We will present our most recent work on CD-SAXS measurements on the line-edge roughness and sidewall angle of test gratings. We will first describe Bragg diffraction patterns of ideal gratings. We will then predict the characteristic scattering from more realistic gratings with LER using both analytical derivation and numerical simulations. Experimentally, we obtained the Fourier spectrum of LER with a combination of the intensity and the position of satellite peaks. By tilting the samples to different angles relative to the incident beam, we determined the average sidewall angle of the grating lines without complicated modeling. A precision better than 0.5 degree can be achieved in our current experimental setup. A 3D picture of the patterns will be discussed based on our current experimental results.
Finally, we will discuss the applications of this new metrology in the emerging nanoimprint field. Specifically, we will discuss the defect transfer from mold to imprinted samples across the wafer. The study of physical origins of the defects in nanopatterns such as thermal expansion coefficient and adhesion to the substrates is ongoing.
Polymers Division, Materials Science and Engineering Laboratory
224/B314, Stop 8541, NIST, Gaithersburg, MD
Sigma Xi Member: No