Hui Tian Charles Reece Jefferson Lab TTC April 2010 Tutorial in spirit, see published literature for more precise use of language. TTC cer 4/20/2010 1
We continue to deepen our understanding of what “EP” does to niobium surfaces and apply that knowledge to optimize the process. We want to understand the scale-specific details of surface leveling. We pursue a reliable, cost-minimized process for JLab, ILC and other applications. TTC cer 4/20/2010 2
Diffusion-limited access of F - to the surface produces “best” polishing Summary Diffusion Layer (~ um) • Anodization of Nb in H 2 SO 4 forces growth of Nb 2 O 5 . Nb F - dissolves Nb 2 O 5 . • Bulk Electrolyte • These competing processes result in current flow and material removal. F - % • Above a certain anodization potential, the reaction rate plateaus, limited by how fast fresh F - can arrive at the surface. ( diffusion-limited) Distance • In this steady-state case, this Nb 2 O 5 layer is a “compact salt film” with specific resistivity. • The thickness of the salt film increases with Compact Salt Film applied potential, although the steady-state (~ nm) F - % current does not change ( plateau ). • In the true diffusion-limited circumstance, material removal is blind to crystallography (avoids crystallographic etching ). Distance Local temperature, flow and • The diffusion coefficient sets a scale for the most electrolyte composition affect the effective leveling local F - gradient TTC cer 4/20/2010 3
Diffusion-limited access of F - to the surface produces “best” polishing Summary Diffusion Layer (~ um) • Anodization of Nb in H 2 SO 4 forces growth of Nb 2 O 5 . Nb F - dissolves Nb 2 O 5 . • Bulk Electrolyte • These competing processes result in current flow and material removal. F - % • Above a certain anodization potential, the reaction rate plateaus, limited by how fast fresh F - can arrive at the surface. ( diffusion-limited) Distance • In this steady-state case, this Nb 2 O 5 layer is a “compact salt film” with specific resistivity. • The thickness of the salt film increases with Compact Salt Film applied potential, although the steady-state (~ nm) F - % current does not change ( plateau ). • In the true diffusion-limited circumstance, material removal is blind to crystallography (avoids crystallographic etching ). Distance Local temperature, flow and • The diffusion coefficient sets a scale for the most electrolyte composition affect the effective leveling local F - gradient So we want to understand this diffusion coefficient TTC cer 4/20/2010 4
We have successfully characterized the • temperature-dependent viscosity of the EP fluid • diffusion constant of F - in the fluid This allows us to calculate the scale of most effective leveling. We have also clearly identified that a parallel etching process is present at higher temperatures – this works against obtaining the smoothest surfaces, yielding a reaction rate that is spatially varying with local chemical potential – grain orientations and lattice stresses. TTC cer 4/20/2010 5
I-V curves of Nb electropolishing at different temperatures with RDE 10 °C 0 – 4000 rpm 19 °C 0.015 0.040 HF(49% ):H 2 SO 4 (96% )=1:10 0.014 o C T=10 0.013 0.035 4000 RPM 0.012 3600 RPM HF(49% ):H 2 SO 4 (96% )=1:10 0.030 0.011 3200 RPM o C T=19 2800 RPM 0.010 2400 RPM 0.025 Current (mA) Current (mA) 0.009 2000 RPM 0.008 4000 RPM 1600 RPM 0.020 3600 RPM 0.007 3200 RPM 1200 RPM 2800 RPM 0.006 2400 RPM 0.015 2000 RPM 800 RPM 1600 RPM 0.005 Note only ~20% 1200 RPM 0.010 400 RPM 0.004 800 RPM current rise 400 RPM 0.003 0 RPM 0.005 0 RPM with 400 rpm, 0.002 0.001 0.000 at constant 0.000 0 2 4 6 8 10 12 14 16 18 20 temperature 0 2 4 6 8 10 12 14 16 18 20 22 Voltage( vs.MSE) Voltage (vs. MSE) 41 °C 30 °C 400 250 HF(49%):H 2 SO 4 (96%)=1:10 350 T=41 o C 225 300 200 2 ) HF(49% ):H 2 SO 4 (96% )=1:10 Current Density (mA/cm 2 ) o C T=30 4000 RPM Current Density (mA/cm 175 250 3600 RPM 3200 RPM 2400 RPM 150 4000 RPM 2000 RPM 3600 RPM 200 1600 RPM 3200 RPM 125 2800 RPM 1200 RPM 2400 RPM 2000 RPM 150 800 RPM 400 RPM 100 1600 RPM 1200 RPM 800 RPM 100 75 0 RPM 400 RPM 50 50 0 RPM 25 0 0 2 4 6 8 10 12 14 16 18 20 0 0 2 4 6 8 10 12 14 16 18 20 Voltage(V) (vs. MSE) TTC cer 4/20/2010 Voltage (V) (vs. MSE) 6
RDE measurements υ − ω = 0.5 0.67 0.166 ( . ) 0.62 slope J vs nFD c 350 Excellent linear fit provides o C T= 1 +/- 0.5 definitive evidence of a o C T= 9.0 +/- 0.5 300 diffusion-limited process. o C slope=10.14 T= 19.0 +/- 0.5 2 ) Knowing ν and c yields D . o C Current density(mA/cm T= 30.0 +/- 0.5 250 o C T= 41.0 +/- 0.5 o C T= 50.0 +/- 0.5 slope=9.82 200 150 slope=5.68 c F = 2.67×10 -3 mol/cm 3 slope=4.23 100 slope=2.87 50 slope=1.79 0 0 2 4 6 8 10 12 14 16 18 20 ω 1/2 (rad.s -1 ) 1/2 Strong evidence for temperature-dependent electrochemical etching in parallel with the diffusion-limited process. For analysis, we must separate these current contributions. TTC cer 4/20/2010 7
Determining Electrolyte Physical Properties Kinematic Viscosity of RDE measurements 1:10 HF/H 2 SO 4 Electrolyte + viscosity measurements 0.30 + concentration 0.25 determine the Diffusion coefficient 0.20 ν (cm2/s) 0.15 0.10 0.05 Diffusion Coefficient of 1:10 HF/H 2 SO 4 Electrolyte 0.00 0 10 20 30 40 50 3.5E-07 T ( ° C) 3.0E-07 H. Tian, JLab 2.5E-07 D (cm2/s) Measured using a 2.0E-07 Brookfield DV-II pro viscometer 1.5E-07 1.0E-07 5.0E-08 c F = 2.67×10 -3 mol/cm 3 0.0E+00 0 10 20 30 40 50 T ( ° C) TTC cer 4/20/2010 8
Estimation of diffusion layer thickness in 1:10 HF/H 2 SO 4 Electrolyte at different temperatures c = × × × c F = 2.67×10 -3 mol/cm 3 J n F D δ Bulk C − 34 F 32 Diffusion layer thickness ( µ m) 30 ≈ C (0, ) t 0 − 28 F 26 24 δ 22 δ 2 1 20 There exists a F - concentration 18 16 gradient within the 10-20 µm away 14 from the surface. 12 10 8 On this scale, peaks are 6 dissolved much faster 0 10 20 30 40 50 than valleys. o C) Temperature ( 2 µ m scale structure should vanish much faster that 40 µ m structure TTC cer 4/20/2010 9
Not all Nb “EPs” the same PSD of Fine CBP Nb Surface Before/ After EP AFM Measurement ( 50 µ m*50 µ m) 7 10 KEK fine CBP fine grain sample 2 With “standard”1:10 HF/H 2 SO 4 KEK fine CBP large grain sample 9 6 10 KEK fine CBP single crystal sample 13 Electrolyte at 30 ° C Nb KEK fine CBP large grain sample 9 after EP KEK fine CBP single crystal sample 13 after EP 5 crystallography affects the 10 KEK fine CBP fine grain sample 2 after EP polishing effectiveness. 4 10 With identical starting PSD(nm 2 ) 3 10 RMS~200nm topography from CBP, given identical 100 min “EP” at 30 ° C, 2 10 RMS~40nm single-crystal material was 1 10 significantly smoother. 0 10 Evidence for a significant etching RMS~7nm activity at 30 ° C, consistent with -1 10 RDE analysis and visual -2 10 experience. -2 -1 0 1 10 10 10 10 Spatial frequency ( µ m -1 ) TTC cer 4/20/2010 10
Avoid sulfur at the cathode • Most commercial electropolishing applications attempt to maximize the surface area of the cathode to avoid process complications (power cost and chemistry). • In contrast to this, typical horizontal cavity EP circumstances have a cathode:anode active area ratio of 1:10. (Worse if “masking” is applied.) • Result is high current density on the cathode and necessary high overpotential to drive the current. This risks driving other chemistry, such as S reduction . • ~5.5 V polarization potential @ cathode to drive 300 mA/cm 2 =~30 mA/cm 2 at anode. 2- + 8 H + + 6 e - → S + 4 H 2 O SO 4 may proceed if cathode overpotential is too high Power supply Power supply voltage -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 voltage 340 340 Area ratio of Nb/ Al 1.0 Area ratio of Nb/ Al 320 2 ) 2 ) 320 o C T = 20.5 +/- 1.3 Cathode Current Density ( mA/cm Anode Current Density ( mA/cm o C T = 20.5 +/- 1.3 300 300 2 ) 0.9 Al area kept unchanged ( 2.6 cm 2 ) Al area kept unchanged ( 2.6 cm 280 280 10 : 1 ; 8 : 1 ; 6 : 1 10 : 1 ; 8 : 1 ; 6 : 1 260 0.8 4 : 1 ; 2 : 1 ; 1 : 1 260 4 : 1 ; 2 : 1 ; 1 : 1 240 240 0.7 220 220 200 200 0.6 Current ( A ) 180 180 0.5 160 160 140 140 0.4 120 120 100 100 0.3 80 80 0.2 60 60 40 40 0.1 20 20 0 0 0.0 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 Voltage ( V ) Voltage ( V ) TTC cer 4/20/2010 11
If the objective is maximally smooth surfaces without sulfur particles : Implications: • We should expect the best micropolishing for topographic features smaller than ~ 10 µm, so start with surfaces that are consistently smooth to this scale: e.g. CBP. • This process we call “EP” also has a temperature-dependent etching process present. So, minimize the temperature as much as is practical (and minimize lattice strains). • Reduce or eliminate sulfur production at the cathode by minimizing cathode current density and improving the reaction kinetics for hydrolysis at the cathode. (1:10 HF/H 2 SO 4 Electrolyte with Nb) TTC cer 4/20/2010 12
Recommend
More recommend
