Modeling The Deposit Thickness Distribution in Copper Electroplating of Semiconductor Wafer Interconnects Eugene Malyshev 1, Uziel Landau 2, and Sergey.

• Slide 1

Modeling The Deposit Thickness Distribution in Copper Electroplating of Semiconductor Wafer Interconnects Eugene Malyshev 1, Uziel Landau 2, and Sergey Chivilikhin 1 1 L-Chem, Inc Beachwood, OH 44122 and 2 Department of Chemical Engineering, Case Western Reserve University, Cleveland OH 44106 AIChE Annual Meeting,San Fransisco, CA. Slide 2 Analyze the effects of the different process parameters Provide a convenient (for non-expert users) & comprehensive tool for: Cell Design Scale-up Process Optimization Objectives Slide 3 Deposit- Deposit thickness uniformity (+/- ~3% across the wafer) Minimal edge exclusion ( Slide 4 Parameters Analyzed Cell Configuration (Dimensions, Edge gap, Shields) Flow (Rotation and Convective Flow) Seed Layer Thickness Electrolyte Composition Acid Concentration (Conductivity) Reactant Concentration (Mass-Transport) Additives (Kinetics/Polarization Curve) Operating Parameters: Current/Voltage Slide 5 Cell Generic configuration DISTRIBUTED FLOW = 4 gpm 10 mm 100 mm ANODE WAFER HOLDER 150 mm 60 rpm GAP Applied Voltage WAFERHOLDER Seed thickness 10 mm GAP Base Case: r = 100 mm, gap =10 mm i = 20 mA/cm2, K= 0.55 S/cm, seed thickness = 1000A rotation = 60 rpm impinging flow = 4 gpm Slide 6 Flow Map: Modified Design Flow Map: Base Case Rotating Disk vs. Combined Flow Flow effects r/R Delta, cm Base case Modified Levich eqn. Slide 7 Numerical comparison with analytical model i lim, A/cm 2 Delta, cm Cell-Design Levich eqn. Cell-Design Model system: rotating disk, r = 7 mm, C b = 0.28 mole/L, D = 6.7*10 -6 cm 2 /s Slide 8 i lim, A/cm 2 Effect of edge-gap Wafer r = 100 mm Simulated gaps: 5 mm, 10 mm, and 15 mm; C b = 0.28 mole/L, D = 6.7*10 -6 ; Impinging flow = 4 gpm 5 mm 50 mm 100 mm 150 mm Slide 9 DISTRIBUTED FLOW = 4 gpm 10 mm 100 mm ANODE WAFER HOLDER 150 mm GAP Applied Voltage WAFERHOLDER Seed thickness 10 mm GAP Resistive substrate effect i average = 10 mA/cm 2 i average = 40 mA/cm 2 Current, A/cm 2 no seed resistance 500 1000 2000 500 1000 2000 Seed thicknesses = 500, 1000 and 2000 . i average = 10 and 40 mA/cm 2. Wafer r = 100 mm. Rotation = 60 rpm. Impinging flow = 4 gpm. C b = 0.28 mol/L, = 0.55 S/cm, D = 6.7*10 6 cm 2 /s. Slide 10 Effect of edge-gap DISTRIBUTED FLOW = 4 gpm ANODE WAFER HOLDER 60 rpm Gap i = 20 mA/cm 2 gap = variable seed = 1000 A Deposit, [micron] Deposit, [micron] Deposit, [micron] 150 sec 180 sec 1-3 time steps = 20 sec, 4-7 time steps = 30 sec Gap = 0 mmGap = 10 mmGap = 50 mm Slide 11 Shield design 60 rpm ANODE WAFER HOLDER 60 rpm DISTRIBUTED FLOW = 4gpm HOLDER DISTRIBUTED FLOW = 4 gpm WAFER ANODE 60 rpm DISTRIBUTED FLOW = 4 gpm HOLDER WAFER DISTRIBUTED FLOW = 4 gpm i, A/cm 2 10% variation Slide 12 200 mm wafer vs. 300 mm wafer DISTRIBUTED FLOW = 4 gpm 10 mm 100 mm WAFER HOLDER 150 mm 60 rpm GAP DISTRIBUTED FLOW = 9 gpm 10 mm 150 mm Deposit, micron 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec deposit(r/R=1) / deposit(r/R=0) = 1.646 deposit(r/R=1) / deposit (r/R=0) = 1.847 200 mm wafer 300 mm wafer Seed thickness = 1000 . C b = 0.28 mol/L, k = 0.55 S/cm, D = 6.7*10 6 cm 2 /s. r/R Slide 13 Electrolyte conductivity (pH) r/R Deposit, [micron] Deposit, [micron] Deposit, [micron] Deposit, [micron] 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec 150 sec 120 sec 90 sec 60 sec 40 sec 20 sec 180 sec = 0.55 S/cm = 0.055 S/cm 300 mm wafer 200 mm wafer i average = 20 mA/cm 2 seed th = 1000 A i average = 20 mA/cm 2 seed th = 1000 A i average = 20 mA/cm 2 seed th = 1000 A i average = 20 mA/cm 2 seed th = 1000 A High (normal) acidity Low acidity Slide 14 Additives effect Current density, [A/cm 2 ] r/R Pure copper sulfate (0.5 M, pH = 2, no additives ) With additives * * - Plating from copper sulfate in the presence of 70 ppm Cl -, 50 ppm SPS and 200 ppm Polyethylene glycol [PEG] - (molecular weight = 4000 ) Slide 15 Conclusions The effects of the various process parameters have been simulated The simulated results are in general agreement with observations. Some Specifics: A proper shield design at the wafer edge significantly enhances uniformity Electrode rotation has a larger effect than the convective flow (in the practiced range) Wafer plating (macroscopic scale) does not typically operate under mass transport control The edge-gap has a major effect on the flow and the current density near the wafer edge The resistive seed effect is noticed mostly at higher current densities (~40 mA/cm 2 ) Scaling to 300 mm enhances the non-uniformity effects, unless compensating measures are taken,.