Carrier-Lifetime Testing to Monitor Material Quality through Silicon Solar Cell Manufacturing

Transcription

1 Carrier-Lifetime Testing to Monitor Material Quality through Silicon Solar Cell Manufacturing James Swirhun Sinton Instruments, Inc., Boulder, Colorado USA PVMC c-si Metrology Workshop 12 July 2012

2 Overview Lifetime Theory & Background Lifetime Measurements through the process Brick and Ingot As Cut Wafer In-Process Wafer SEMI Standard & Round Robin 2

3 Lifet ime Theory Carrier lifetime (τ) is: Average time for photogenerated excess carriers to recombine Key parameter for solar cell operation longer lifetime enables more carriers to be collected at higher cell voltage e e e e e e e e e N A 3

4 Cell Efficiency (%) Solar Cell Operat ion Cell Efficiency (%) Higher efficiency cell design leads to higher sensitivity to bulk lifetime Physical Measurements allow accurate simulation PC1D Simulation Standard Production Cell PC1D Simulation of Highefficiency n-type IBC Cell Bulk Lifetime (ms) Bulk Lifetime (ms) 4

5 Why Measure Lifet ime in Bricks and Ingot s? After crystal growth: Process monitoring of feedstock contamination, contamination from crucible, bad growth conditions, furnace quality. Quickly qualify material representing many hundreds of wafers. Sensitivity to measure lifetime in the ,000 us range. After wafering: As-cut wafers have very low measured lifetimes due to surface recombination low sensitivity to defects or contaminants. Process Optimization: Dopant Diffusion Passivation layer 5

6 Measurement Theory Illumination Illumination Photoconductance Photoconductance Two methods to measure τ eff using eddy-current instrument Quasi-Steady-State (QSSPC): Transient PCD: QSS n G Both methods applied to 2-4mm slug samples in this study flash Si trans n d n dt QSSPC E E E E-02 Time Transient PCD E E E E-03 Time

7 Bulk Measurement s on Silicon Ingot s and Slabs IR light source Eddy current sensor

8 True Bulk Lifetime (ms) QSSPC Technique on Bricks or Ingot s Bulk Lifetime Measurement (Broadband light, real instrument parameters) IR-pass 1000 pass Perfect 100 QSSPC 10 Transient (after surface modes decay) 2.5% of peak carrier density 1 1 Ohm-cm p-type silicon Measured Lifetime (ms) PC1D simulation: Swirhun, PIP Res. Appl (2010) pg. 1029

9 Lifetime ( ms) Example P-t ype Mult i Brick, 1 ohm-cm 1 ohm-cm MultiX: Lifetime scan from bottom to top RG1000 RG850 RG1000 as-measured Cm from bottom of brick

10 Measured Lifetime (sec) N-t ype FZ 3 ohm-cm C: Minority-Carrier Lifetime vs. Carrier Density 1.4E-02 Tau = µs at 5.0E+14 cm-3 1.2E E E E E-03 Sensitivity much greater than critical threshold for 22% cells 2.0E E E E E E+16 Minority- Carrier Density (cm -3 ) = 1 ms

11 As-Cut Multicrystalline Wafers Some cell manufacturers do not have access to bricks or ingots and start with wafers 11

12 Bulk lifet ime can be predict ed from as-cut lifet imes As-cut QSSPC measured lifetime vs. nitride passivated lifetime. Model permits bulk lifetime to be measured on as-cut wafers. Good accuracy for L<W Bothe et al., PIP

13 Bulk Lifetime (ms) Trap Density (cm -3 ) Different QSSPC Dat a Set : Product ion dat a from Phot ovolt ech 100 As-cut bulk lifetime Bulk lifetime after phosphorus 10 1E14 Traps (right axis) 1 1E Wafer Number from Brick Bottom T. Mankad, R. Sinton, A. Ristow, N. Akil, E. Picard, NREL Silicon Workshop, Breckenridge, CO

14 Process Opt imizat ion and Cont rol Passivation Process Control Lifetime allows Surface Recombination Velocity measurements Used to design and optimize new passivation technologies such as: a-si, Al 2 O 3, SiN x 1 = 1 + 2S τ eff τ bulk W J 0 measurement used to optimize dopant diffusion Bulk lifetime can also be extrapolated at this stage to measure impurity gettering or bulk contamination 1 = 1 + J 0(N A + Δn) τ eff τ bulk qn 2 i W 14

15 SEMI St andards Recently Published SEMI documents: AUX-017 technical overview of lifetime measurement and analysis techniques SEMI PV13 Eddy current lifetime measurement standard These documents enable physics-based measurement and analysis in R&D and production Recently completed Round Robin shows precision of measurement method 15

16 SEMI PV13 Round Robin 23 labs participated spanning Research and Industry in US, Europe, Asia 5 samples with lifetimes from 2μs to 3.5ms were measured. 3 wafer samples, 2 Bulk samples Results: QSS Lifetime ~ 7-10% RSD, 4-7% RSD after Recalibration Transient Lifetime ~ 3-6% RSD 16

17 Short Term Met rology Challenges Implement physics-based measurement and analysis in production. (See SEMI AUX-017, PV13) Shift to high efficiency improved process means higher sensitivity to bulk material. This requires work on incoming material quality Achieve tight distribution on >23% efficient cells Improve CZ quality minimize/understand common defects in both p- and n-type 17

18 Long Term Met rology Challenges Develop models to predict final efficiency based on parameters measured through all process steps. 18

19 Industry-leading Technology for Silicon PV Process Control Founded in 1992 Boulder, CO USA 19