Slide 1: Unique Technology for Exponential Growth
Unique Technology for Exponential Growth Dr. Shay Wolfling, CTO September 2022
Slide 2: Speaker biography: Dr. Shay Wolfling, CTO
Joined Nova as CTO in 2011. R&D manager at KLA-Tencor-Belgium, leading multidisciplinary metrology & inspection projects. Founder and VP R&D of Nano-Or-Technologies, a start-up company with a proprietary 3D optical technology, acquired in 2005. PhD in physics from the Hebrew University. 59 © Nova pr
Slide 3: Outline
Technology Trends Outline Key Challenges Unique Solutions 60 © Nova proprietary information
Slide 4: Technology Trends
Technology Trends 61 © Nova proprietary information
Slide 5: Training Compute and Advanced Technology Portfolio
I.E+25 Training Computer (FLOPs) I.E+22 I.E+19 I.E+16 I.E+13 Higher performance IoT Edge Computing I.E+10 with less power and space I.E+07 Compute needs for AI expanding – I.E+04 decision intelligence, generative AI & constant optimization I.E+01 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 200
Slide 6: Linear scaling is only part of the way forward
HKMGGate last SiGe, Cu Linear scaling is only part of the way forward Nanosheets, back side delivery Normalized Cost per Transistor ($) CoContact over gate LaO in the Gate EUV FinFET More than Moore is essential: SiGe, in Fin • New integrations • Novel materials Scaling & Integration • True 3D – dev
Slide 7: Solutions & Technology Inflections
Solutions & Technology Inflections Inflections Key value Patterning High NA EUV, Additive patterning Density New architectures – Logic transistor Performance, Density, Power & Cost Nanosheet, Forksheet, Complementary FET Buried Power Rail, new materials, Interconnect Performance, Power selective pro
Slide 8: Dimensional & Material Implications
Dimensional Advanced Packaging & Material Implications New Materials Dimensional Vertical Materials Chemical Integration Expanding Material Metrology Device to Chemical Scaling Metrology 65 © Nova proprietary information
Slide 9: Measure Chemical Solutions
- Electrochemical + Plating Measure Chemical Wafer Main components: Solutions Acids, bases, organics Rather than on Wafer Plating Solution Plating Bath Secondary components: Clean & Etch Composition Contaminants, breakdown products, Metal Source leftovers Optimizing wet processes Front End Damascene
Slide 10: Key Challenges
Key Challenges 67 © Nova proprietary information
Slide 11: From Single Deck to Multi Deck
Overlay 5 Slit Dimensional 3 TCD Slit 7,8,9 MCD 4 1. Full profile – per deck and in total 1 2. Logic under Memory cell 6 3. Bottom parameters Slit 4. Tiers and liner thickness BCD 5. Tilt and Overlay between decks 3 3 2 Material 6. Dielectric composition & thickness control 7. Channel Poly Si crysta
Slide 12: Dimensional Material: DRAM challenges
1,3 1,6 Capacitor Bit Line Dimensional Word 1. High-Aspect Ratio Line 4 capacitor holes & supporters 2. Bottom gate parameters 2 3. Tight contact holes 4. Lateral selective etches for thinner capacitors 5,6, 7 Material 5. Dielectric composition & thickness control 6. Higher dielectric constants 7. M
Slide 13: Dimensional Material: Nanosheet challenges
Nanosheet Forksheet Complementary 2D Materials FET <10nm 4 Dimensional 1. Multiple nanosheets – shapes, spacing & thickness 1,2,5, 6,7 2. Thin deposition surrounding nanosheets 3. Buried structures 4. Local topography variations 3 Adding a dielectric wall True 3D 2D Atomic between NFet and PFet NFet
Slide 14: Dimensional Material: Interconnect challenges
Dimensional 1. HAR Cu nano-TSV profile control 7,8 2. Topography for Wafer-to-Wafer 1,3,5 bonding 3. Voids & Delamination 4. Surface roughness Material 5. Via Pitch and CD too narrow for plated copper 2,4 6. Cobalt etch - Single damascene 7. Cu fill - Buried Power Rails 8. Cu, Ru & Co interconnects
Slide 15: Dimensional Material: Fan Out and 3D-IC
Fan Out 3D-IC Dimensional • The move from 2.5D to 3D - Topography & Profile • High aspect ratio structures • Buried structures • Local topography variations Material • Optimal plating • Highly saturated copper bath management • By-Products metrology • Contaminants and photoresist leach control • Die
Slide 16: More steps and higher sampling
Shorter More steps Tighter specs as In-die & New R&D cycles & higher sampling design rule complex structures materials shrinks Good Fill Void Seam Chemical control to enable Topography & Profile 3D Local variation is critical: Monitoring more parameters filling of HAR structures Packaging local dens
Slide 17: Nova Solutions
Nova Solutions 74 © Nova proprietary information
Slide 18: Critical Dimensions
Stand Alone performance in IM An additional source of Extending Nova IM leadership information to SR and SE Critical Dimensions 75 © Nova proprietary information
Slide 19: Material Properties
In-Line composition & thickness In-Line stress, strain and crystallinity In line full Composition profile Market leading technology Measuring material properties optically Bringing SIMS from Lab to Fab High volume manufacturing Material Properties 76 © Nova proprietary information
Slide 20: Chemical Analysis & Replenishment
Automated separation Quantitative analysis using optics technique whereby all to directly measure component components are separated concentration before being measured Chemical Analysis & Replenishment 77 © Nova proprietary information
Slide 21: Across all Product Lines
Combining ML with physical modeling Inherently accurate sample description AI-enhanced Machine Learning Central management, Expanding beyond OCD Dramatically reducing time to solution performance monitoring and big data analytics Across all Product Lines 78 © Nova proprietary information
Slide 22: Solutions for Nanosheets
Solutions for Nanosheets Unique Technologies to Answer Key Challenges Fin Patterning • XPS - Multiple nanosheets thickness • SIMS - Multiple nanosheets full profile • OCD - Multiple CDs along gate and fin HAR profile; Buried structures • Raman - SiGe stress post Etch of nanosheets Replacing Metal Ga
Slide 23: METRION Inline SIMS: Nanosheet Use Case
METRION – Inline SIMS Nanosheet Use Case: SiGe Deposition Wafer 01 Wafer 02 In-Line RESULT Uniformity Germanium Deposition Variations in nanosheet SiGe Uniformity growth affect etch selectivity Across Wafer and impact device performance Growth Difference SiGe Growth Uniformity In-line SIMS monitors
Slide 24: Solutions for Memory - 3D NAND
Solutions for Memory - 3D NAND Unique Technologies to Answer Key Challenges Pillar Sidewall Liner Deposition Channel Hole Etch at Top Decks • XPS - Liner thickness • XPS - ONON thickness & composition • SIMS – Contamination detection & Individual • Raman- Channel poly crystallinity & Grain size laye
Slide 25: Zoom in - Solutions for 3D NAND
Zoom in- Solutions for 3D NAND In-Die Measurement Top Oxide CMP Word Line Contamination Detection Channel Etch Profiling Polysilicon on Sidewall 82 © Nova proprietary information
Slide 26: Solutions for Packaging
Solutions for Packaging Unique Technologies to Answer Key Challenges Dimensions Materials • Al bond pad monitoring • 3D Interconnect: TSV, hybrid bonding – Cu2Cu • High K composition & thickness • Cobalt cap & liner monitoring • RDL CD and thickness • Multi-Layer dielectric thickness • Organic addit
Slide 27: ancolyzer: High Aspect Ratio Plating Use Case
ancolyzer chemical metrology Packaging Use Case: High Aspect Ratio Plating Excellent Control, Accuracy and Reproducibility Concentration Electroplated Copper micron- size vias plating quality is strongly affected by the changes in the plating solution RESULT 0 5 10 15 20 Measurement Automated inline
Slide 28: Nova Future Directions
Nova Future Directions Lasers Different energy Additional material Chemical metrology Physical and ML sources and extended technologies from enhanced with Nova combined solutions wavelength range Lab to Fab HW and algorithms enhancing materials core competency metrology 85 © Nova proprietary informa
Slide 29: Unique technology solutions: Challenges, Trends, Solutions
Unique technology solutions Exponential growth in data Multiple technology inflections across all product lines generation Logic- architecture evolution with Enhanced by physical modeling More than a single increased complexity and machine learning track forward Challenges Solutions 3D NAND - multip
Slide 30: Thank You
Thank You