Patterning materials at the nanoscale 6.12J/3.155J Microelectronic processing Outline 1. Introduction- How small would you like to go?... can you go 2. Optical and other lithography 3. Subtractive and additive patterning 4. More exotic patterning methods: Interference (Imprint,) Block copolymers elf asse (Building on last years notes of Caroline ross) Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Patterning materials at the nanoscale Outline 1. Introduction - How small would you like to go?… can you go? 2. Optical and other lithography 3. Subtractive and additive patterning 4. More exotic patterning methods: Interference, (Imprint,) Block copolymers Self assembly (Building on last year’s notes of Caroline Ross)
How small would you like to go?...and can you go? with semiconductors?. with lithography 6.12J/3.155J Microelectronic processing Semiconductor scaling drivers: Speed of light in global interconnects, vac/vK Increased information density ◆ Increased logic speed ◆ Economies of scale Semiconductor scaling challenges diffusion length control, statistics, especially for channe Thermal management, V2/R Power density: Pentium: Pentium Il Pentium Il: Nuclear Reactor Dielectric breakdown field, vd ◆ Interconnect cross-talk, delay,C↓,T=LC Information stability: kBT< CV2 Screening lengths(range of band bending, depletion regions Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing How small would you like to go? …and can you go? with semiconductors?…with lithography? Semiconductor scaling drivers: w Speed of light in global interconnects, v µ c/√k w Increased information density w Increased logic speed w Economies of scale Semiconductor scaling challenges: w diffusion length control, statistics, especially for channel w Thermal management, V 2/R Power density: Pentium: Pentium III :: Pentium III: Nuclear Reactor w Dielectric breakdown field, V/d w Interconnect cross-talk, delay, CØ , t = L/C w Information stability: kBT < CV 2 w Screening lengths (range of band bending, depletion regions)
How small can MOSFETs go? 6.12J/3.155J Microelectronic processing 10 Four problems as size shrinks 1. Drai doping concentration Gate le ng th limited by thermodynamics 导 2. Physical tunnel through gate oxide (lot< lon) 01 3. Statistical fluctuations Junction depth of doping: Nd s 50/ (10 nm)3 ate 0.0 Oxide thic kness source 1970 1980 1990 2000 201[ Year Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Four problems as size shrinks: 1. Drain, source doping concentration limited by thermodynamics 2. Physical tunneling through gate oxide (Ioff < Ion) 3. Statistical fluctuations of doping: Nd ≈ 50/(10 nm)3 4. Economic cost of Fab How small can MOSFETs go?
International Technology roadmap for Semiconductors: 2002 6.12J/3. 155J Microelectronic processing YEAR OF PRODUCTION200120021203204205120012007 DRAM Pitch (m)130 115 100 0 80 70 65 MPU为 Pitch(nn 1301079080 MPUPrinted Gate Length(n) 75655345 4 35 MPU Physical Gate Length (en) 534537322825 YEAR OFPRODUCTION 201020132016 DRAM Pitch (nm) 453222 MPU Pitch (* y 4532|22 MPU Printed Gate Length(nm) 251813 MPU Physical Gate Length(/w) 18139 N≈50n10×10×10m3 Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing International Technology Roadmap for Semiconductors; 2002 Nd ≈ 50 in 10x10x10 nm3
How small can you go with semiconductors?. with lithography? 6.12J/3.155J Microelectronic processing Lithography limitations fresnel and fraunhoffer limitations other light sources or particles? When the optical resolution exceeds physical resolution e-beam, secondary generation blurs lines Particle dama Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing How small can you go with semiconductors?…with lithography? Lithography limitations: w Fresnel and Fraunhoffer limitations w Other light sources or particles? w When the optical resolution exceeds physical resolution w e-beam, secondary generation blurs lines w Particle damage
Can Lithography get you there? Proximity lithography 6.12J/3.155J Microelectronic processing OBJECT PLACEHOLDER This object has been removed because it is not by MIT. It may be reinstated future Fresnel (near field diffraction): n3 microns Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Fresnel (near field diffraction): l 3 microns OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future
Can Projection lithography get you there? 6.12J/3.155J Microelectronic processing OBJECT PLACEHOLDER OBJECT PLACEHOLDER This object has been removed because it is not This object has been owned by MIT. It may be removed because it is not reinstated or replaced in owned by MT. It may be reinstated or replaced in Fraunhoffer(far field diffraction) imum resolved feature nfld(no limit on lens diameter, d For f=d, n=436 nm=>0.4 microns Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Fraunhoffer (far field diffraction): Minimum resolved feature = lf/d (no limit on lens diameter, d) Can Projection lithography get you there? For f ≈ d, l = 436 nm => 0.4 microns OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future. OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future
Beyond Optical a X-Ray Masks Wavelength Energy ■ Sources Light U\ 400 nm E-Beam a Speed E Deep uv 250 nm 4.96eV Electron Scattering X-Ray 0.5nm 2480el a Nano-imprint Soft-Litho h1.23 mv ve(e Fall 2003-MA Schmidt 3.] 6. 152]-Lecture 11-Slide 18
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing l = hc E l = h mv = 1.23 E(eV) (Å)
Recall semiconductor physical limits 6.12J/3.155J Microelectronic processing Semiconductor scaling challenges diffusion length control, statistics, especially for channel Thermal management, v2/R Power den: pentium: pentium lll a pentium l: nuclear reactor Dielectric breakdown field, v/d ◆ Interconnect cross-talk, delay,c↓,T=LC Information stability: KBT> CV2 ◆ Screening lengths (range of band bending, depletion regions) Electron concentration If not semiconductors, then what other information technologies? Dec.10,2003
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Semiconductor scaling challenges: w diffusion length control, statistics, especially for channel w Thermal management, V 2/R Power den.: Pentium: Pentium III :: Pentium III: Nuclear Reactor w Dielectric breakdown field, V/d w Interconnect cross-talk, delay, CØ , t = L/C w Information stability: kBT > CV 2 w Screening lengths (range of band bending, depletion regions) Recall semiconductor physical limits Electron concentration If not semiconductors, then what other information technologies? …
Size limitations on other information technologies 6.12J/3.155J Microelectronic processing Magnetic disk storage media OBJECT PLACEHOLDER owned by MIT. It may be reinstated or replaced in Schematic of longitudinal recording medium OBJECT PLACEHOLDER OBJECT PLACEHOLDER This object has beer his object has been removed because it is not removed because it is not owned by MIT. It may be owned by MIT. It may be reinstated or replaced in reinstated or replaced in Each bit(currently about 0.5 X 0.08 X0.03 um -=10 Gb/in) Consists of 1000 grains; noise a N-l/2 Dec
D e c. 10 , 2 0 0 3 6.12J / 3.155J Microelectronic processing Size limitations on other information technologies Schematic of longitudinal recording medium Magnetic disk storage media Each bit (currently about 0.5 x 0.08 x 0.03 µm - ≈ 10 Gb/in2 ) Consists of 1000 grains; noise µ N-1/2 OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future. OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future. OBJECT PLACEHOLDER This object has been removed because it is not owned by MIT. It may be reinstated or replaced in future
