SPIE Startup Challenge 2015 Founding Partner - JENOPTIK Get updates from SPIE Newsroom

Micro/Nano Lithography

Getting up to speed with roadmap requirements for extreme-UV lithography

A major research effort under way at the SEMATECH consortium focuses on new tools and techniques for better understanding what causes mask blank defects and how to prevent them.
19 April 2011, SPIE Newsroom. DOI: 10.1117/2.1201104.003542

The availability of defect-free masks is the most critical technology gap hindering the commercialization of extreme-UV (EUV) lithography.1–3 Multilayer defects are generated during the manufacture of EUV mask blanks, the basis for the photomasks used to pattern computer chips. The insufficient rate of progress toward reducing the defects in mask blanks4 across the EUV community over the past two years has prompted us at the SEMATECH (Semiconductor Manufacturing Technology) consortium to establish a Mask Blank Defect Reduction Program to provide immediate and long-term solutions to the industry. This Program has been tasked with identifying the source of defects within the multilayer deposition process, implementing mitigation techniques, and demonstrating an improved process that yields a low density of defects.

SEMATECH's Mask Blank Defect Reduction Program is supported by the Mask Blank Development Center, a facility for developing defect-free EUV blanks built at the College of Nanoscale Science and Engineering, State University of New York (SUNY) at Albany (see Figure 1). Two Nexus low-defect-density tools by Veeco Instruments are being used for process development of the deposition of molybdenum/silicon with an integrated ruthenium capping layer, as well as targeted experiments to minimize specific defect types.5 The low-defect-density tools each consist of two loadlocks to hold the substrate in place , a standard mechanical interface (SMIF) unit, an aligner, a transfer module, and the deposition chamber. Each of the deposition tools (see Figure 2) can accommodate 6-inch-square mask substrates. The substrates are loaded from the SMIF pod into the loadlock with the SMIF indexer. The loadlock is exposed to low- and high-speed pumping (i.e., it is soft-roughed and then fast-roughed) before going into the high-vacuum-deposition module. The deposition module consists of an ion source with silicon, molybdenum, and ruthenium water-cooled targets and an electrostatic ‘chuck’ to hold the mask substrates. After the substrates are deposited with the multilayer and capping films, the mask blanks are characterized by x-ray analysis to determine the thickness of the bilayers, and by an EUV reflectometer to determine the overall reflectivity, uniformity, and centroid wavelength (the arithmetic mean value of the two wavelengths defining the full-width half-maximum points for the reflectivity curve). Defect densities are obtained using inspection tools at a wavelength of either 433 or 266nm. Defect composition is then characterized by energy-dispersive spectrometry (EDS) on a focused ion-beam/scanning electron microscope (FIB/SEM), transmission electron microscope (TEM) imaging, or through Auger analysis. The shapes and volumes of the defects can also be measured with an atomic force microscope (AFM) or TEM.


Figure 1. An overview of SEMATECH's Mask Blank Development Center in Albany, NY.

Figure 2. Schematic illustration of the Veeco Nexus ion-beam sputter-deposition chamber in the two systems (process modules 1 and 2) installed in the Mask Blank Development Center.

Ongoing research at our Mask Blank Development Center has provided opportunities for blank suppliers to learn about the tools at an early stage and to develop and evaluate new processes for multilayer deposition and metrology of EUV mask blanks. The performance of current deposition tools is one of the factors limiting the availability of defect-free blanks. Over the past year, we have been concentrating significant resources on modeling and developing a fundamental understanding of the deposition process, which is then verified by experimental programs to characterize the deposition tool. Success in this effort requires characterizing the defects incorporated in multilayer films, modulating defect types (i.e., changing parameters to determine the effect on defects), and developing novel solutions for mitigating defects. The fundamental understanding thus generated has been shown to improve the performance of the existing ion-beam deposition (IBD) tool and to help demonstrate the design and feasibility of next-generation multilayer deposition tools to meet mask-blank-pilot-line defect and reflectivity requirements.

We now have the facilities and expertise at SEMATECH to support the metrology, characterization, and evaluation of EUV mask blank defects as small as tens of nanometers. The widely used metrology technique EDS (implemented in the SEM) provides elemental information about defects, and also extracts further information on their shape and composition. However, due to its penetration depth, the technique is restricted to large defects (hundreds of nanometers). In addition, EDS cannot provide quantifiable composition analysis of the defect or information about its chemical state and crystallinity. These limitations impede the identification of defect sources. However, we have devised a comprehensive metrology strategy to identify any defect with a core size larger than 10nm.

SEMATECH is currently equipped to perform state-of-the-art Auger and TEM analysis in-house to support the analysis of small core defects for pilot-line EUV masks. Moreover, these capabilities are continuously being developed and the throughput increased. Our newly established Auger tool can analyze a standard 6-inch mask blank and is already providing important information about sub-100nm defects on EUV blanks (see Figure 3). Operation on a complete 6-inch mask allows sample analysis without glass cutting, which contaminates the blank and hinders sample navigation and determination of defect position. Use of TEM complements Auger analysis, providing ultimate resolution in the imaging of subnanometer structures. Crystalline and phase information generated by the TEM also indicates the sources of defects. In addition, the EDS and electron-energy-loss spectroscopy equipment at SEMATECH provides higher analytical power than similar techniques in traditional SEMs (see Figure 4).


Figure 3. A carbon defect as identified by Auger electron microscopy. The image on the right shows an elemental map of carbon as compared with the scanning electron microscope image on the left.

Figure 4. Scanning transmission electron microscopy image of a defect on the mask substrate disrupting the multilayer structure. The bottom two images show elemental scans of molybdenum (Mo) and silicon (Si) on the substrate.

The current industry requirement specifies that EUV blanks should contain zero defects over more than 150nm, considered to be ‘killer’ defects, and a density of no more than 0.13 defects between 50 and 150nm in size per square centimeter (see Figure 5). At SEMATECH we are producing EUV blanks with densities of 0.1/cm2 for defects greater than 70nm at a yield of 70. We consider the target values to be achievable in 2011 with the current multilayer deposition tool and Lasertec 7360 inspection capability. However, high-volume manufacture requirements for future EUV blanks are much more stringent, with a required defect density of no more than 0.13 defects greater than 35nm/cm2 and no defects greater than 80nm. Achieving these targets will be very challenging and will require a new multilayer deposition tool in 2013 and an inspection capability that can detect defects up to 25–30nm. Future research at SEMATECH is committed to progress in EUV lithography technology and bridging the critical gap in progress toward producing defect-free EUV mask blanks. This will require developing a metrology infrastructure for learning about defects, as well as a state-of-the-art tool set and advanced analytics to support the program.


Figure 5. SEMATECH's roadmap for the EUV Mask Blank Defect Reduction Program. HVM: High-volume manufacture. Q: The quarter of the year. YE 12: Year ending 2012. M7360: Commercial tool name for inspection by Lasertec. ML Dep.: Multilayer deposition process. SMT: SEMATECH. LDD1: Low-defect-density tool by Veeco.

Vibhu Jindal, Francis Goodwin, Stefan Wurm
SEMATECH
Albany, NY 

Vibhu Jindal is currently a project manager at SEMATECH working on the EUV Mask Blank Defect Reduction Program. He received his BS and MS in materials science from the Indian Institute of Technology in Bombay, and a PhD in nanoscale science and Engineering from SUNY-Albany.

Francis Goodwin is the program manager for SEMATECH's EUV Mask Blank Defect Reduction Program and Mask Blank Development Center. He has more than 20 years' experience in lithography, holding management and engineering positions throughout his career, and has supported application-specific integrated circuit logic and advanced dynamic random access memory operations.

Stefan Wurm is SEMATECH's associate director for lithography research on assignment from GlobalFoundries. He holds a physics diploma and a PhD in physics from the Technische Universität München, Germany.


References:
1. S. Wurm, EUV lithography development and research challenges for the 22nm half-pitch, J. Photopolym. Sci. Technol. 22, no. 1, pp. 31-42, 2009.
2. S. Huh, A. Rastegar, S. Wurm, K. Goldberg, I. Mochi, T. Nakajima, M. Kishimoto, M. Komakine, Study of real defects on EUV blanks and a strategy for EUV mask inspection, Proc. SPIE 7545, pp. 7545N, 2010. doi:10.1117/12.863559
3. H. Yun, F. Goodwin, S. Huh, K. Orvek, B. Cha, A. Rastegar, P. Kearney, SEMATECH EUVL mask program status, Proc. SPIE 7379, pp. 73790G, 2009. doi:10.1117/12.824257
4. http://www.sematech.org/meetings/archives/litho/8653/pres/O_M1-05_Liang_Intel.pdf Defect-free mask fabrication to enable EUVL. Accessed 22 March 2011.
5. P. Kearney, C. C. Lin, T. Sugiyama, C.-U. Jeon, R. Randive, I. Reiss, R. Rajan, P. Mirkarimi, E. Spiller, Ion beam deposition for defect-free EUVL mask blanks, Proc. SPIE 6921, pp. 69211X, 2009. doi:10.1117/12.774505