COMPARISON OF EUV SINGLE EXPOSURE VS. 193i MULTIPLE PATTERING FOR N10 BEOL CHRISTOPHER J. WILSON - CONTRIBUTORS FORM ADVANCED LITHO AND CU-LOW-K IIAP PROGRAMS - ASML VELDHOVEN DEMO LAB FOR EXPOSURES
EUV 193i COMPARISON VEHICLE FOR 24NM HALF PITCH 2 level BEOL vehicle M0/V0/M1/V1/M2 with Product Like layout 193i and EUV versions for Pattering Assessment M1 LE 3 EUV SP V0/V1 LE 2 EUV SP M2 SADP+Keep EUV SP 9-193i Masks Fields V0 LE 2 M1 LE 3 V1 LE 2 M2 Core M2 Keep 4 - EUV Masks Fields All single exposure 2
ADVANCED LITHOGRAPHY CLUSTERS IMMERSION EUV ASML NXT:1950 Sokudo DUO (since Mar 12 and continuing ASML NXE:3100 TEL Lithius Pro (from Oct 11 till Feb 15 ASML NXT:1970 TEL Lithius Pro Zi (since April 15 ASML NXE:3300 TEL Lithius Pro Z EUV (Since July 2015 3
4 COMPARING PROCESS COMPLEXITY
COMPARING PROCESS COMPLEXITY 193i LE 3 PRINTS 42NM LINES ON 144NM PITCH LEA LEB LEC Litho Shrink in SOG SOC etch Oxide etch Strip TiN etch 5
COMPARING PROCESS COMPLEXITY 193i LE 3 PRINTS 42NM LINES ON 144NM PITCH LEA 27 LEB Proces s Steps LEC Litho Shrink in SOG SOC etch Oxide etch Strip TiN etch 6
COMPARING PROCESS COMPLEXITY 193i SADP+KEEP USES 48NM LINES ON 96NM PITCH Core Litho Core etch/trim Core Etch, Trim and And SiOC SiOCstrip strp Spacer dep Dep Spacer Etch etch and Core pull Keep Litho Shrink SOG etch in SOG and PR strip SOC etch and and SOG strip strip TiN TiN Open Etch and and SOC Strip strip 7
COMPARING PROCESS COMPLEXITY 193i SADP+KEEP USES 48NM LINES ON 96NM PITCH 18 Core Litho Core etch/trim Core Etch, Trim and And SiOC SiOCstrip strp Proces Spacer dep Dep s Steps Spacer Etch etch and Core pull KeepLitho Shrink SOG etch in SOG and PR strip SOC etch and and SOG strip strip TiN TiN Open Etch and and SOC Strip strip 8
COMPARING PROCESS COMPLEXITY EUV SP PRINTS 24NM LINES ON 48NM PITCH Litho UL etch Oxide etch PR Strip TiN etch 9
COMPARING PROCESS COMPLEXITY EUV SP PRINTS 24NM LINES ON 48NM PITCH 8 Proces s Steps Litho UL etch Oxide etch PR Strip TiN etch 10
Cost (a.u. COMPARING PROCESS COMPLEXITY COST MODEL 450 400 350 300 250 200 150 100 50 0 Lithography Depostion Metallization DryEtch WetClean CMP Metrology 4% 5% 4% 6% 9% 5% 9% 6% 18% 5% 18% 8% 21% 23% 23% 4% 5% 7% 38% 48% 34% MxLE^3_VxLE^2 MxSADP+SP-Blk_Vx-LE^2 MxEUV-SP_VxEUV-SP High Deposition, Etch and Clean cost High Litho Cost 11
12 REAL WAFER DATA
WAFER DATA 193i LE 3 VS EUV SP BCABCABCA A (nm B (nm C (nm 193i LE 3 13nm LE bias EUV SP 0nm LE bias Litho Etch Litho Etch 39.3(1.12 39.6(1.34 39.4(1.19 24nm HP 27.7(2.53 26.8(2.41 27.4(2.68 24.6(1.84 24.6(1.77 24.6(1.66 25.1(2.25 25.3(1.89 25.2(1.96 SWR 5.85 3.50 5.18 4.90 M1C M1A 193i LE 3 13nm LE bias M1B EUV SP 0nm LE bias Litho Etch Litho Etch 51.9 (3.53 - - 27.2(4.31 27.0(3.00 24.3(3.60 26.2 (3.61 25.7(2.88 25.2(3.67 25.5(3.85 24.9(3.02 25.6(3.97 11 8.6 - - 13 LOGIC Better CDU for EUV SP but slightly worse LWR/SWR In each case CDU and LWR can lead to open or bridging 193i LE 3 route results in corner rounding EUV SP offers clearly better patterning fidelity than 193i LE 3
WAFER DATA 193i SADP+KEEP VS EUV SP LOGIC 24nm HP 193i SADP+KEEP Gap Core EUV SP Core (nm 193i SADP+Keep 18nm LE bias EUV SP 0nm LE bias Litho Etch Litho Etch 42.5(0. 8 24.8(1.8 22.0(0.86 22.7(2.25 SWR (Core 5.1 2.1(1.0 4.61 3.9 Gap (nm - 24.2(1.6 - - SWR (Gap - 2.7(1.9 - - EUV single patterning offers clearly better patterning fidelity than 193i SADP + Keep, however, roughness is high 193i SADP + Keep shows better CDU and LER/LWR but has sharp line ends 14
WAFER DATA 193i SADP+KEEP Spacer Contour * Keep Contour * Final Contour * TiN Core Spacer Gap Oxide SiN *Contour used to guide the eye. Based on non-calibrated etch model 15
193i SADP+KEEP KEEP MULTIPLE PATTERNING KEEP CAN HELP LINE ENDS CORE REMOVAL Keep KEEP LITHO FINAL METAL TRENCH Spacer SP Target Contour Keep A Keep B DP Target Contour 16
WHAT IS IMPORTANT WHEN MAKING A COMPARISON?
ELECTRICAL INTEGRITY
ELECTRICAL INTEGRITY Resistance/Capacitance variability Open/Short due to Lines/Space narrowing CDU/LER/LWR Proximity effect Via landing Overlay Line end pull back Proximity M1 LE 3 Rounding Line end/corner shape Narrowing Pull back Comparison by human eye is easy but the challenge is to automate this M1 EUV 19
CONTOUR EXTRACTION FROM CD-SEM E.G. EUV CASE 100+ contours show the variability across the wafer Line-Ends Full wafer SEM collection Count 300 200 100 Mean = 1.8nm Standard Deviation = 2.0nm 0-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 EPE [nm] All ends show a wide range Count 40 35 30 25 20 15 10 5 Separate locations show pull in and push back Location 5 Location 6 Location 7 0-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 EPE [nm] 20
VARIABILITY BANDS EUV vs 193i LE 3 EUV Variability Band 193i Variability Band Extraction artefacts No doubt EUV cases look better qualitatively Immersion case shows pattern displacement errors (probably due to overlay errors, less sharp corner transitions (probably due to OPC, and larger variability bandwidths 21
LINE END PULLBACK IS DEPENDENT ON PROCESS EUV vs 193i LE 3 22
HOT SPOT BRIDGE EUV vs 193i LE 3 23
VIA AREA COVERAGE (NOMINAL VIA USED EUV vs 193i LE 3 24
SUMMARY CAN WE CONCLUDE WHICH PROCESS IS BETTER? DEPENDS! Masks (single level LE 3 SADP+Keep EUV 3 2 1 CDU (3s 2.5nn 1.8nm 2.0nm L/SWR 3.5nm 2.7nm 3.9nm Process steps 27 18 8 Full flow cost 396 a.u. 375 a.u. 343 a.u. Pattern Integrity Rounded corners Sharp line ends Good Variability in Line-end pull back (3s 16nm Close to 1nm * Close to 5nm Bridge 3-4 0 * 2-3 Via coverage 28% >75% * 25 75%... IT * Estimated based on visual data