Chapter 15 Summary and Future Trends

Transcription

1 Chapter 15 Summary and Future Trends Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 1

2 The 1960s First IC product Bipolar dominant PMOS Diffusion for doping Metal gate Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 2

3 PMOS in the 1960s Gate Oxide CVD Cap Oxide p + N-Silicon p + Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 3

4 The 1970s Bipolar dominant NMOS Ion implantation for doping after mid-1970s Self-aligned source/drain Polysilicon gate Main driving force: electronic watches and calculators Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 4

5 NMOS in the 1970s Al Si SiN PSG Poly n + p-si n + Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 5

6 The 1980s MOSFET surpassed bipolar CMOS Multi-level interconnections Main driving force: personal computer Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 6

7 CMOS in the 1980s PD2 Nitride PD1 Oxide Metal 2, Al Cu Si p + n + IMD PMD Poly Si Gate n + Al Cu Si BPSG LOCOS SiO 2 p + P-type substrate USG dep/etch/dep p + p + N-well Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 7

8 The 1990s MOSFET dominant CMOS Multi-level interconnections Tungsten Silicide CMP Main driving force: PC, network, internet Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 8

9 CMOS Passivation 2 Silicon Nitride in the Al Cu Alloy Al Cu Passivation 1 USG 1990s Ti/TiN IMD 3 USG Metal 4 TiN ARC Metal 3 Al Cu Alloy Ti Ti/TiN IMD 2 USG W M2 Al Cu IMD 1 USG W TiSi 2 M1 Al Cu Alloy Sidewall Spacer, USG PMD W BPSG Poly-Si STI n + n + USG p + p + P-Well N-Well Hong Xiao, Ph. D. P-Epi www2.austin.cc.tx.us/hongxiao/book.htm P-Wafer 9 PMD Barrier Nitride

10 The 2000s MOSFET dominant CMOS SOI substrate Copper and low-κ interconnection Main driving force: telecommunication, network, internet, PC Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 10

11 CMOS Cr, Cu, and Au liners Copper 5 Lead-tin alloy SOD Nitride PSG SiN seal layer in the SOD SOD Copper s SiC etch stop layers SOD Ta/TaN barrier layer Cu 3 SOD Cu 3 CoSi 2 Copper 2 Cu 1 Cu 1 SOD Cu 1 Cu 1 W Tungsten PSG n + P-well n + p + N-well p + Buried SiO 2 P-wafer Poly Si gate SiN barrier Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm STI USG USG layer 11 SOD SOD SOD SOD SiC seal layer PE-TEOS cap SiC seal layer

12 New Materials: Copper Metal interconnection: Cu replace Al and W Lower resistivity Improve device speed Higher electromigration resistance Higher current Reduce metal layers that reduce processing steps Lower the production cost Improves overall yield Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 12

13 New Materials: Low-κ Low-κ dielectrics replace silicate glass for the interconnection applications CVD: CSG and α-fc SOD: HSQ and porous silica. Combination of copper and low-κ to improves IC chip speed Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 13

14 New Materials: High-κ Capacitance of MOS gate capacitor has to be large enough to hold enough charges Feature size reducing, gate capacitance reduces High-κ, thicker gate dielectric to prevent the leakage and breakdown Candidates: TiO 2 (κ ~ 60), Ta 2 O 5 (κ ~ 25), and possibly HfO 2 Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 14

15 New Materials: High-κ BST (Ba ½ Sr ½ TiO 3, κ up to 600 DRAM capacitor dielectric Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 15

16 Next Generation Lithography Photolithography limit: ~ nm, after 2010 Next generation lithography (NGL) technology EUV lithography Projection electron beam lithography Still in R&D Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 16

17 Developing Industry After more than 50 years, semiconductor industry is still a developing industry, not a matured industry as automobile industry New technology is introduced almost daily Technology less than ten years could become obsolete Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 17

18 300 mm and Beyond Larger wafer size, more chip can be made Currently 300 mm (12 inch) transition Will become mainstream Cost more than 2 billion dollar to build 400 mm wafer fabs may start by 2010 Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 18

19 World Chip Demands Lower chip price, cheaper consumer electronics TVs, VCRs, telephones, and PCs. Steady economy development of developing country, especially China and India Dramatically increase the demands Need more chips! Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 19

20 Auto-chip Global location system and voice activated internet access may become standard feature in the future automobiles In the near future, General Motors (GM) will consume more IC chips than International Business Machines (better known as IBM) every year Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 20

21 Bio-chip Miniaturize test probes and analysis circuits Medical IC chips for DNA test Fast, accurate diagnosis of DNA related diseases. Lab-on-chip Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 21

22 Telecommunication, Internet The worldwide development of the telecommunication and internet will still be the main driving force of the continuing rapid development of the IC industry in the near future Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 22

23 Future is Bright Boom-bust cycle of the IC industry Demands for IC chips will steadily grow So will the demands for skillful and knowledgeable workers Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 23

24 Life After the Final Limit Physical limit may be reached by ~2030 Feature size 10 to 5 nm IC feature size can no longer shrink IC industry will finally become matured Less frequent technology change Still need large number of workers, just like auto industry Hong Xiao, Ph. D. www2.austin.cc.tx.us/hongxiao/book.htm 24