ScaleUp Activities in Atomic Layer Deposition at Argonne

Scale-Up Activities in Atomic Layer Deposition at Argonne Jeffrey Elam, Anil Mane, Joe Libera December 9, 2011 Large Area Picosecond Photodetector Collaboration Meeting December 9 -10, 2011, Argonne National Laboratory

Outline § § Introduction to ALD (1 slide) ALD on 33 mm plates Challenges with scale-up to larger substrates Conclusions 2

Binary Reaction Sequence for Al 2 O 3 ALD Trimethyl Aluminum (TMA) A) OH OH OH CH 3 Al CH 3 CH 4 CH 3 Al 3 CHAl Al CH CH 3 3 OH OH OH Al(CH 3)3 H 2 O B) CH CH 33 CH 3 Al Al CH 3 3 OH OH OH H 2 O CH 4 H 2 O OH CH 3 OH OH CH Al 3 CH 3 Al OH OH OH n 1 ALD Cycle of TMA/H 2 O Deposits 1 Al 2 O 3 “Monolayer” 3

Multiple 33 mm MCPs in Tubular ALD Reactor Chem-2 Resistive Coating • Thickness uniformity on monitor Si(100) <2% • The resistive layer thickness ~800 A • Similar thickness trend observed on second batch of 5 MCPs Excellent batch-to-batch reproducibility 4

Resistance Comparison for 9 MCPs (air vs. vacuum) Chem-2 Resistive Coating • Average resistance (2 batches of 5 plates) 115 ± 12 M ~10% resistance variation 5

Challenges to Coating Larger Areas § § Need a bigger reactor Non-ideal ALD surface reactions High aspect ratios High surface areas 6

Large Substrate Reactor 12 x 18” box 2” tube § From 1 x 1 -in plates to 12 -in x 18 -in plates 7

Coating 8” MCPs in Beneq 300 mm chamber 8

Non-Ideal ALD Surface Reactions In 2 O 3 ALD in Large Substrate Reactor In. Cp + O 3 →In 2 O 3 n In 2 O 3 films were thinner downstream in reactor n O 3 In 2 O 3 O* + O 2 n Indium oxide catalyzes ozone destruction 9

Scale-up of ALD In 2 O 3 to 12”x 18” Substrates In. Cp + O 3 § Thickness deviation: 45% In. Cp + H 2 O + O 2 § Thickness deviation: 2. 5% § Resistivity deviation: 6% § Non-self limiting ALD reactions can lead to non-uniform films in larger ALD systems 10

Coating High Aspect Ratios: ALD W in Anodic Alumina 40 nm pores, 70 microns long 15 μm § Conformal coating on all exposed surfaces § Aspect ratio ~ 1000 11

Large Substrate Surface Area: Silica Gel Powder § 100 micron particles, 30 nm pores (aspect ratio ~ 2000) § Surface area = 100 m 2/g § Powder bed fixture for ~1 g support 12

Self-Limiting Al 2 O 3 ALD Planar Surface Silica Gel n Self-limiting growth on planar and porous surfaces n Exposures increased by x 100 13

Surface Areas of Glass Capillary Arrays: MCP Type d (µm) γ = l/d 33 mm diam 40 40 20 60 8 in square 40 40 20 60 α 0. 65 0. 83 0. 65 ATotal (cm 2) 881 1, 120 1, 317 43, 244 64, 712 (thanks Jason M. ) Empty tube reactor: ~2000 cm 2 Empty LSR: ~4000 cm 2 Empty Beneq 300 mm: ~1600 cm 2 n MCP aspect ratio = 60 (we’ve done 10^5) n 8” MCP surface area = 6. 5 m 2 (we’ve done 10^3) n Piece of cake… 14

Work Plan 1) Qualify Beneq for Chem 2 coatings using coupons and 33 mm MCPs in new 300 mm chamber 200 300 mm mmchamber 2) Coat single 8”x 8” MCPs in 300 mm chamber 3) Coat multiple (1 -4) 8”x 8” MCPs in 300 mm chamber X) Coat multiple (20) 8”x 8” MCPs in 3 D chamber 15

Al 2 O 3 ALD in 300 mm Reactor <1% non-uniformity in 100 nm coating Works great! 16

Chem-2 ALD in 300 mm Reactor Test Metrics: • Optimized chemistry -2 baseline process on 300 mm chamber: • Deposit 300 mm wafer at same condition as MCP deposition. • Evaluated the resistivity uniformity across the large area Precursor inlet direction 5 11 17 1 6 12 18 23 2 7 13 19 24 3 8 14 20 25 4 9 15 21 26 10 16 22 300 mm 17

Resistivity of Chem 2 coating at different locations on 300 mm wafer Works great for Chem-2! 18

Chem-2 thickness NU on 300 mm wafer with 8”x 8” MCP on top: 2 D thickness map 300 mm 1 -sigma thickness: 10% Items Run-3 Average (nm) 42. 905 STDV 4. 484 1 sigma 0. 105 %sigma 10. 450 Max (nm) 57. 821 Min(nm) 36. 843 % diff 48. 893 August 22, 2011 • We may need square chamber? 300 mm 19

8” 40µm MCP Pair Gain Map Y gain slice (thanks Ossy, Jason, SSL) Y gain slice X gain slice FIRST Gain map – Looks awful - multifiber visible, lots of gain non-uniformity This is from the thickness non-uniformity 20

Troubleshooting in Beneq reactor wall 300 mm wafer 8” MCP Side View assumed flow distribution Top View n ALD conductive coating with MCP substrate installed over Si(100) 300 mm wafer n Measure film thickness on Si wafer using 4 -point probe conductivity (easy, quantitative) 21

Process testing on 8” MCP on Beneq • Baseline process: 50 cycles conductive coating (No MCP) Precursor inlet Items Thickness data from Resistance Average (nm) 75. 169 STDV 3. 412 1 sigma 0. 045 %sigma 4. 539 Max(nm) 78. 186 Min(nm) 63. 107 % diff 20. 060 1 -sigma = 4% 22

Thickness values form resistance and 2 D map: Items Baseline Run-1 Run-2 Run-3 # of cycles 50 cycles 30 cycles 25 cycles Expected thickness (nm) 75 75 45 45 37 Dose time (s) 1 10 20 40 60 Average (nm) 75. 169 56. 289 25. 263 42. 905 35. 491 G (nm/cycle) 1. 50 1. 13 0. 842 1. 43 1. 42 STDV 3. 412 32. 299 13. 352 4. 484 8. 448 1 sigma 0. 045 0. 574 0. 529 0. 105 0. 238 %sigma 4. 539 57. 381 52. 851 10. 450 23. 804 Max (nm) 78. 186 110. 438 46. 746 57. 821 47. 500 Min(nm) 63. 107 0. 236 2. 196 36. 843 14. 884 % diff 20. 060 195. 776 176. 348 48. 893 91. 900 57% 52% 10% 24% 1 -sigma = 4% Tested many ideas, hardware, software, reactor breaks, etc. (3 months) 23

Troubleshooting in Beneq: Al 2 O 3 Side View n ALD Al 2 O 3 coating with MCP substrate installed over Si(100) 300 mm wafer n Visually assess coating uniformity n Measure film thickness on Si using ellipsometry 24

Troubleshooting in Beneq: Al 2 O 3 Precursor inlet 50 nm 100 nm m 150 n Anil: “We have a flow problem… we need a square chamber” 25

Troubleshooting in Tubular Reactor 2 cm x 30 cm Si(100) substrate, rest 2 x 20 cm MCP 2 mm above Si 500 cycles TMA/H 2 O for Al 2 O 3 2 inches MCP Si • Narrow gap under MCP • High surface area 26 26

Troubleshooting in Tubular Reactor 2 -5 -2 -5, 0. 05 Torr doses 2 -5 -2 -5, 0. 2 Torr doses 2 -5 -2 -5, 1 Torr doses n Some non-uniformity, not as bad as Beneq 27

Troubleshooting in Tubular Reactor Si MCP Si • Narrow gap under MCP • Narrow gap over MCP • High surface area 28

Troubleshooting in Tubular Reactor 2 -5 -2 -5, 0. 05 Torr doses n Same bad non-uniformity as in the Beneq 29

Troubleshooting in Tubular Reactor Si • Narrow gap under MCP” • Narrow gap over MCP • No high surface area Si 2 -5 -2 -5, 0. 05 Torr doses n Nearly perfect uniformity 30

What is going wrong in Beneq? reactor wall • Narrow gap under MCP • Narrow gap over MCP • High surface area 300 mm wafer 8” MCP Side View actual flow distribution Top View • No flow in the gap • Slow outgassing/diffusion from MCP • Precursors mix, CVD 31

What is going wrong in Beneq? Precursor inlet 50 nm 100 nm m 150 n n The problem: flow is bypassing the MCP n Solution: we need to confine the flow so that it is forced to pass in the gap between the plate(s) and the reactor wall. n We have a flow problem… we need a square reactor. 32

Potential Solution, Conclusions reactor wall 300 mm wafer 8” MCP Side View Top View n Equal gap n Convert circle into square n CFD modeling would be helpful n Listen to Anil 33