EUV lithography
08/01/1997
FOURTH IN A SERIES
EUV lithography
Andrew M. Hawryluk, Natale M. Ceglio, David A. Markle, Ultratech Stepper, San Jose, California
During the past seven years, extreme ultraviolet lithography (EUVL) has evolved from a simple concept into a possible candidate for mass production of future integrated circuits. Part 1 of this article (in Solid State Technology, July 1997, p.151) reviewed the history of EUV lithography and discussed critical challenges in EUV optics. This part looks at the other pieces of the EUV puzzle - reticle, resist, and source technologies.
Reticle technologies
The EUV optical system projects a demagnified image of the reticle pattern onto the resist. The minimum feature size on the reticle, at 4? demagnification, is 400 nm. EUV reticles, like other EUV optics, rely on resonant reflective multilayer coatings, so they can be fabricated on opaque substrates. Silicon wafers may be the substrate of choice, offering a smooth surface, good thermal conductivity, compatibility with existing manufacturing equipment, and low cost.
To fabricate EUV reticles, a molybdenum-silicon multilayer reflective coating deposited on the substrate (silicon wafer) is followed by a thin protective buffer layer and a thin metal absorber layer [1, 2]. Next, resist is spun on and exposed by an e-beam pattern writer [3, 4]. While various defect repair procedures for the metal pattern have been tested [2, 5], no available technology can repair defects in the multilayer coating [6]. A multilayer coating deposition process with a defect density less than 10-3 defects/cm2 is required to produce reticles with greater than 90% yield.
As recently as 1994, magnetron sputter deposition, which has been used to fabricate high-reflectivity EUV multilayer mirrors, produced reticles with more than 104 defects/cm2. An extensive effort to reduce the defect density in EUV multilayer mirrors achieved <100 defects/cm2, but further reductions also degraded reflectivity [7]. A different approach was required.
Ion beam sputter deposition, which has successfully fabricated low-scatter optical multilayer coatings, is an attractive alternative. The reticle is decoupled from the plasma, a significant source of defects in magnetron sputtering, and the energetics of the deposited ions are well controlled. An ion beam tool [8] designed for low-defect density EUV multilayer coatings (see figure) achieved approximately 2 ? 10-2 defects/cm2, about 1000? lower than the best multilayer mirrors grown in a magnetron sputter system.
 |
A collaborative effort between Veeco and LLNL has designed, built, and demonstrated a low-defect coating tool for fabricating EUV reticle blanks using conventional particulate control practices. This tool has a demonstrated defect density of ~2 ? 10-2/cm2.
Source technologies
Many sources of EUV radiation were considered during the early development of EUV lithography, including synchrotrons (conventional bending magnets, undulators, and wigglers) [9], free electron lasers [10], and laser-produced plasmas. The source must be compatible with the optical system design (matching the etendu of both the condenser and imaging systems) and emit sufficient power into the condenser system (within the bandwidth of the multilayer mirrors) for a commercially relevant throughput. Because of these constraints, the laser-produced plasma source was preferred.
A laser-produced plasma source for EUV lithography uses a pulsed laser focused onto a target. Under the appropriate conditions, the laser`s interaction with the target produces a plasma that emits EUV radiation. The development of the plasma source can be divided into three tasks:
achieving efficient conversion from laser to EUV radiation,
developing a high average-power laser, and
developing a "debris-free" source.
The laser and material conditions needed to generate EUV radiation are very different from those required for generation of proximity print x-ray radiation. EUV is generated most efficiently [11, 12] with a laser pulse energy of 0.5-1 joules, a duration of 5-10 nsec, and an intensity on target of ~3 ? 1011 watts/cm2. Laser to EUV radiation conversion efficiencies as high as 2% (within the bandwidth of the imaging system) have been reported [12] using a tin target. Unfortunately, this target also generates considerable debris. Conversion efficiencies of approximately half the tin value have been reported for other high-atomic-weight metals. These materials generate less debris, but still exceed the "debris-free" requirement for EUVL.
Commercially relevant throughputs require a high average power pulsed laser system [13]. Diode pumped lasers have produced up to 1 kW at 2.5 kHz [14]. A laser system suitable for an EUV source driver is likely to be commercially available before the end of the decade [15].
Source debris must be eliminated before EUV can be a viable lithography candidate for advanced integrated circuits. The first condenser optic looks directly at the source, so any debris could degrade the optic and reduce the system throughput. Several different source architectures (solid targets [16], low mass targets [17], water, ice [18], and cryogenic targets [19]) and a number of debris intervention techniques (high-speed rotating shutters, electric and magnetic fields, and flowing gas) have been tested. All of these techniques helped reduce the debris reaching the first condenser optic. The latest low-debris source uses a supersonic gas jet [20].
Resist technologies
Resists for EUV lithography will probably use some form of surface-imaging because the absorption depth of EUV radiation in conventional resist is very shallow. Surface-imaging processes under evaluation include a tri-layer resist process (a metal layer sandwiched in between an imaging layer and a pattern transfer layer), an organometallic process (a pattern transfer layer, activated by the exposure, permits plating of a metal pattern), and a silylation process.
Silylation for IC fabrication has been under development for many years. A typical EUV silylation process [21] employs a thick conventional resist (~500-700 nm) as a pattern transfer/planarization layer, followed by a thin (~100-200 nm) silylation resist layer. After the EUV exposure, the wafer undergoes a post-exposure bake, silylation, and a pattern transfer etch.
EUV resists need high sensitivity (typically 10? more sensitive than proximity print x-ray resists) because of the short absorption depth at EUV wavelengths. A resist sensitivity of <5 mJ/cm2 is required to achieve adequate throughput [13]. High-quality, 100- to 130-nm-wide lines (with sidewalls >858) have been fabricated with a ~15 mJ/cm2 resist [21]. Resists with suitable sensitivity, resolution, etch resistance, and sidewalls appear to be possible, but much more development is required.
Conclusion
The encouraging progress in the US EUV lithography program over the past half-dozen years is largely attributable to stable and continuous funding by the Department of Energy, a strong commitment and technology base at the National Laboratories, and active participation by US industry, which provided technical guidance and focus for the program. The time is at hand to demonstrate engineering viability with an EUVL alpha machine, and then production-worthiness with an EUVL beta machine.
After a cumulative investment of 21 man-years in the development of EUVL technology, we are encouraged by the progress that has been made, yet we retain a healthy respect for the hard challenges ahead. EUV lithography is sometimes described as the natural extension of high-resolution lithography to shorter wavelengths, and as such, it is argued, requires only minimal investment to be successful. On the contrary, EUV is unlike all previous lithographies. It employs a series of resonant reflectors in its condenser and imaging optics, placing important constraints on the EUVL optical system. Use of such an optical system for exposure of large fields with negligible wavefront distortion and high throughput remains to be validated. Nevertheless, the emergence of additional lithography options should be viewed as a hopeful sign. A viable production technology at EUV wavelengths will be challenging and costly, but will sustain the evolution of ICs for at least three more device generations.
References
1. A.M. Hawryluk, et al., "EUV Reticle Pattern Repair Experiments Using 10Kev Neon Ions," OSA Proceedings on Extreme Ultraviolet Lithography, Vol. 23, 1994.
2. D. Tennent, et al., "Mask technologies for Soft X-ray Projection Lithography," Applied Optics, Vol. 32, No. 34, p. 7007, 1993.
3. A. M. Hawryluk, G. Shelden, P. Troccolo, "EUVL Reticle Factory Model and Reticle Cost Analysis," Trends in Optics and Photronics, OSA TOPS on Extreme Ultraviolet Lithography, Vol. IV, G. Kubiak, D. Kania, eds, 1996.
4. A.M. Hawryluk, et al., "Reflection Mask for Soft X-ray Projection Lithography," SPIE Proceedings, Vol. 1547, 1991.
5. A. Hawryluk, D. Stewart, "Reflection mask defect repair" Applied Optics, Vol. 32, No. 34, p. 7012, 1993.
6. A. M. Hawryluk, "Characterization and possible repair of defects in SXPL Masks," OSA Topical meeting on EUV Lithography, Monterey, CA, A.M. Hawryluk, R. Stulen, eds., 1993.
7. K. Nguyen, et al., "Defects in coatings deposited by planar magnetron sputtering; measurements with a Tencor Surfscan 6200," OSA Proc. on EUVL, Vol. 23, Optical Soc of America, F. Zernike, D. Attwood, eds., 1994.
8. S. Vernon, et al., "Reticle Blanks for EUV Lithography: Ion beam sputter deposition of low defect density Mo/Si Multilayers," OSA TOPS on Extreme Ultraviolet Lithography, Vol. 4, 1996.
9. J. Murphy, et al., "Synchrotron radiation sources and condensers for projection x-ray lithography," Applied Optics, Vol. 32, No. 34, p. 6920, 1993.
10. B.E. Newman, "XUV Free Electron Laser Based Projection Lithography," Proc. SPIE, X-ray/EUV Optics, Vol. 1343, San Diego, R. Hoover, ed., 1990.
11. C. Cerjan, "Spectral Characterization of Lithographic Sources," OSA Topical meeting on EUVL, Monterey, CA, A.M. Hawryluk, R. Stulen, eds., 1993.
12. R. Spitzer, et al., "X-ray production from Laser Produced Plasmas for SXPL Applications," OSA Topical meeting on EUV Lithography, Monterey, CA, A.M. Hawryluk, R. Stulen, eds., 1993.
13. A.M. Hawryluk, N.M. Ceglio, "EUV Lithography Cost of Ownership Analysis," OSA Proceedings on Extreme Ultraviolet Lithography, Vol. 23, Optical Society of America, F. Zernike, D. Attwood, eds., 1994.
14. B. Comansky, et al., "A one kilowatt average power diode pumped Nd:Yag folded zigzag slab laser," SPIE, Vol. 1865, 1993.
15. B. Krupke, LLNL, private communication.
16. G. Kubiak, et al., "Laser Plasma Sources for SXPL: Production and Mitigation of Debris," OSA Topical meeting on EUV Lithography, Monterey, CA, A.M. Hawryluk, R. Stulen, eds., 1993.
17. D. Torres, et al., "Characterization of Mass-limited, Ice Droplet Laser Plasmas," OSA TOPS on Extreme Ultraviolet Lithography, Vol. 4., 1996.
18. F. Jin, M. Richardson, "Conversion Efficiency and Debris Studies of Ice Targets for EUV Projection Lithography," OSA Proc. on EUVL, Vol. 23, Optical Society of America, F. Zernike, D. Attwood, eds., 1994.
19. M. Richardson, et al., "Cryogenic Targets for Laser Plasma X-ray Lithography," OSA Topical meeting on EUV Lithography, Monterey, CA, A.M. Hawryluk, R. Stulen, eds., 1993.
20. G. Kubiak, et al., "Debris-free EUVL Sources Based On Gas Jets," OSA TOPS on Extreme Ultraviolet Lithography, Vol. 4, 1996.
21. D. Wheeler, et al., "Basic Issues Associated with Four Potential EUV Resist Schemes," OSA TOPS on Extreme Ultraviolet Lithography, Vol. 4, 1996.
ANDY HAWRYLUK received his BS, MS, and PhD degrees in electrical engineering from the Massachusetts Institute of Technology. He is Director of Critical Projects at Ultratech Stepper, where his responsibilities include project planning and technical analysis of strategic programs, including optics, stage technologies, and new product development. Hawryluk holds 11 patents, primarily in applications of lithography, and has received numerous awards, including the DOE Award of Excellence in 1985. Ultratech Stepper, 3050 Zanker Rd., San Jose, CA 95134; ph 408/321-8835, fax 408/325-6444.
NAT CEGLIO received his BA and BS degrees in physics from Columbia University in 1966 and 1967, respectively, and his MS and PhD degrees from MIT in 1969 and 1976, respectively. He is VP of Engineering and Product Development at Ultratech Stepper. Prior to joining Ultratech, Ceglio worked at Lawrence Livermore National Laboratory from 1976 through 1996, leading the Advanced Microtechnology Program. He directed the Laboratory`s program in giant magnetoresistive head development and EUV lithography.
DAVID A. MARKLE is vice president of Advanced Technology at Ultratech Stepper. During his 19-year tenure with Perkin-Elmer Corp., he led the invention and development of the Micralign, a scanning projection aligner, and the Micrascan step-and-scan, deep ultraviolet lithography system. In addition to holding more than 20 patents, Markle developed technology used in leading research projects such as Skylab. In 1994, he was honored by the Optical Society of America with the David Richardson Medal, in recognition of his contributions to the field of applied optics.