EUV Resists Considered as Materials for Optics Tom Wallow GLOBALFOUNDRIES Strategic Lithography Technology IEUVI Resist TWG 2/27/2011
EUV Photoresists as Optical Materials This analysis emerges from fundamental optical physics merged with models of EUV photon energy cascades in resist materials Ultimately, there are fundamental limits here (RLS triangle) However, the RLS response surface is currently responsive to materials and process refinement IEUVI Resist TWG 2/27/2011
EUV Resists as Materials for Optics Reaction-Diffusion Confinement effects on Reaction-Diffusion There are fundamental limits here too. However, these limits are softer: - Materials are too complex for a priori theoretical descriptions - Impact and severity may be more gradual Confinement effects on physical properties Interfacial gelation Component segregation Swelling Modulus, etc. IEUVI Resist TWG
How does this Recipe work at all? 1) In a metal-ion free container, combine: 1) In a metal-ion free container, combine: -1 or more complex copolymers, R g ~ 3nm -1 or more complex copolymers, R g ~ 3nm -n additives -n additives -PAG(s) -PAG(s) -Quencher(s) -Quencher(s) -Dissolution modifiers -Dissolution modifiers -Leveling agents -Leveling agents -etc. -etc. -Solvent(s) -Solvent(s) Mix until dissolved, then filter really well. Mix until dissolved, then filter really well. 2) Coat to 20 R g thickness with ~1 R g uniformity over ~10 16 R g 2 wafer area. 2) Coat to 20 R g thickness with ~1 R g uniformity over ~10 16 R g 2 wafer area. 3) Lightly dust with modulated ionizing radiation. Hold flat. Don’t tilt or wiggle. 3) Lightly dust with modulated ionizing radiation. Hold flat. Don’t tilt or wiggle. 4) Bake immediately. Use same temp. and time everywhere on the wafer, every time. 4) Bake immediately. Use same temp. and time everywhere on the wafer, every time. 5) Soak in caustic, then rinse and spin dry. Very carefully. 5) Soak in caustic, then rinse and spin dry. Very carefully. Yields: ~10 12 patterns with ~7 R g width, height of ~ 15 R g , <~ 1 R g LWR 3 s . Yields: ~10 12 patterns with ~7 R g width, height of ~ 15 R g , <~ 1 R g LWR 3 s . Serve (in a particle-free container) to etch chamber for destruction. Serve (in a particle-free container) to etch chamber for destruction. -The Way to Cook , ITRS, 2014-2015 -The Way to Cook , ITRS, 2014-2015 IEUVI Resist TWG 2/27/2011
Interfacial Gradients from Resist Components 193 nm Resist, 1999 EUV Resist, 2009 Normalized F intensity Triflate PAG Normalized F intensity Nonaflate PAG -Narayan et al., J. Photopolym. Sci. Technol. 1999 -Sundaramoorthi et al., Proc. SPIE 2009 IEUVI Resist TWG 2/27/2011
Confinement Effects at the Substrate Interface p(2VP)/SiO 2 PMMA/SiO 2 ‘Protected’ Resist ‘Deprotected’ Resist -Soles et al., JVST B 2001 -Soles et al., Proc. SPIE 2006 -Pham, J. Q., in Materials Science and Engineering -Soles et al., Proc. SPIE 2006 - Rittigstein et al., Nat. Mater. 2007 Univ. of Texas Press, 2004 At 50 nm FT and below, interfacial properties become dominant IEUVI Resist TWG 2/27/2011
Confinement Effects at the Free Interface Relatively uncomplicated for simple copolymers- -Tg depression, local mechanical property changes For resists, PAG segregation effects need to be considered IEUVI Resist TWG 2/27/2011
Blur and the Deprotection Gradient -Hinsberg et al., 2002-2004 Photoacid reaction-diffusion is enhanced in high acid concentration areas This behavior can be viewed as the origin of local interfaces IEUVI Resist TWG 2/27/2011
The Deprotection Gradient (continued) -Rao et al., Langmuir 2006 -Prabhu et al., Macromolecules 2007 -Lavery et al., Proc. SPIE 2006 Deprotection confinement produced by reaction-diffusion results in discrete domains IEUVI Resist TWG 2/27/2011
Interfacial Gelation and Swelling Deprotection Deprotection -Wallow et al., Proc. SPIE 2002 Modern quartz crystal microbalances can distinguish dissolution and swelling Interfacial gelation is always (?) present at some level Case II diffusion is typical for resists- interfacial gel mediates transport IEUVI Resist TWG 2/27/2011
Developer Impact on Interfacial Gelation -A. Sawano, T. Kumagai, TOK IEUVI Resist TWG 2/27/2011
Deformation of Resist Beams φ δ max Δ P θ s Y = yield stress W E = Young’s modulus a = geometric parameter - Yoshimoto et al. J. Appl. Phys. 2004 Multiple scaling issues for pattern collapse: -Young’s modulus -Yield stress -Interfacial gelation IEUVI Resist TWG 2/27/2011
Materials Properties Scaling -Van Workum and de Pablo, Phys. Rev. Lett. 2003 Weakening arises from local modulus fluctuations Scaling may be compounded by other effects such as swelling, etc. Recent Henderson group experimental match is excellent IEUVI Resist TWG 2/27/2011
Critical Aspect Ratio for Collapse - Yoshimoto et al., SPIE 2011 Resist Session 5, Tuesday 11:20 AM IEUVI Resist TWG 2/27/2011
Interfaces and LER Yes, but… What about the -Patsis, Microelect. Eng. 2004 interfaces? IEUVI Resist TWG 2/27/2011
Film Thickness Effects at EUV ca. 2006 FT = 80 nm FT = 60 nm FT = 40 nm 40L80P, 11.3mJ/cm 2 40L80P, 10.8mJ/cm 2 40L80P, 10.8mJ/cm 2 LER=3.7±0.7 LER=7.1±1.1 LER=4.2±0.9 -Wallow et al., EUVL Symposium 2006 Top-loss and roughness become worse as thickness decreases Confinement effects are the likely root cause, but which ones??? What does ‘LER’ actually mean at these film thicknesses? What solutions should we pursue? IEUVI Resist TWG 2/27/2011
Sidewall LER Studies and Observations -Foucher et al. SPIE 2005 -Goldfarb et al. JVSTB 2004 Isotropic sidewall roughness is observed for large resist structures Sidewall roughness in ultrathin resists is much more complex IEUVI Resist TWG 2/27/2011
Anisotropic Sidewall LER in EUV Resists Std. Develop: LER 5.7 nm Surf. Rinse: LER 4.4 nm -George et al., SPIE 2010 IEUVI Resist TWG 2/27/2011
Underlayer Effects on LER -George et al., SPIE 2010 -Koh et al., SPIE 2010 POR processes for many current EUV integration studies Linkage between underlayer smoothing, LER anisotropy, and interfacial confinement can be inferred, but definitive studies are needed IEUVI Resist TWG 2/27/2011
Process Mitigation of Collapse and LER -Petrillo et al., EUVL Symposium 2010 IEUVI Resist TWG 2/27/2011
Summary • Ultrathin polymeric resists (<100 nm thickness) are better viewed as a collection of dissimilar interfaces than as a bulk material • This materials model underlies at least part of the currently still responsive RLS surface for EUV resists • Numerous process enhancements can mitigate aspects of materials performance limits. Initial implementations have largely moved from research to development • Simulation and experiment indicate that maintaining patterning performance will continue to become more challenging as resist films become thinner. This is probably a manifestation of the onset of ‘soft’ materials limits. • As an overly broad statement of historical inevitability, there will be a limit for polymeric chemically amplified resists. • “Why?,” “when?,” and “what then?” are very interesting questions. IEUVI Resist TWG 2/27/2011
Extras
Resist Copolymers … Functional Monomers … … … Typical MW ~ 5000-10000 (~25-75 monomers/chain) Typical radius of gyration ~ 3 nm Polymeric systems are inevitably statistical mixtures -chain microarchitecture -molecular weight distribution IEUVI Resist TWG 2/27/2011
Film Heterogeneity from Chain-Chain Interactions -Chan and Dunstan, J. Phys. Chem. B 2010 Interchain behavior in copolymers is highly cooperative Cooperative behavior is observed at multi-R g length scales The behavior of even simple statistical random copolymers is very challenging to describe IEUVI Resist TWG 2/27/2011
Free-standing Film Moduli- Surface Softening at Small Dimensions Molecular Dynamics simulations Glassy core; softened exterior s (MD segment size; s ~1.5 nm) -Yoshimoto et al. J. Chem. Phys. 2005 Softened exterior grows both in absolute and relative thickness below ~40 nm CD XRR: similar observations IEUVI Resist TWG 2/27/2011
Recommend
More recommend