Iron in crystalline silicon solar cells: fundamental properties, - PowerPoint PPT Presentation

Iron in crystalline silicon solar cells: fundamental properties, detection techniques, and gettering Daniel Macdonald, AnYao Liu, and Sieu Pheng Phang Research School of Engineering The Australian National University, Canberra

Outline • Origins of Fe in multicrystalline Si ingots • Chemical states and recombination activity of Fe in silicon • Measuring [Fe i ] by QSSPC and PL imaging • Gettering of Fe during ingot growth and cell fabrication: – Internal gettering – at GBs, dislocations and within grains – External gettering by P, Al and B diffusions 2

Origins of Fe in multicrystalline Si ingots • One of the most common metal contaminants • Total Fe concentration, measured by NAA before and after P gettering • Comes from the crucible, not the feedstock • Typically between 10 12 -10 15 cm -3 Macdonald et al. 29 th IEEE PVSC New Orleans (2002) 3

Origins of Fe in multicrystalline Si ingots 10 17 • Concentration increases B towards top - segregation 10 16 k eff = 0.65 Impurity concentration (cm -3 ) • Also increases at bottom – 10 15 total Fe solid-state diffusion from crucible 10 14 k eff < 0.05 • Only a small fraction is interstitial - around 1% 10 13 • Remainder is precipitated 10 12 (or substitutional) interstitial Fe k eff < 0.05 • The dissolved fraction has a 10 11 much larger impact on 0.1 1 lifetime Fraction from top of ingot Macdonald et al. J. Appl. Phys. 97 033523 (2005) 4

Recombination activity of interstitial Fe in silicon • Interstitial Fe (Fe i ) introduces a deep donor level • Positively charged in p-type Si - mobile at RT - forms pairs with ionised acceptors. • Two FeB levels – acceptor and donor • Pairs break under illumination – only Fe i present in working cells Courtesy of J. Schmidt, ISFH conduction band FeB 0/- (E C -0.26 eV) 0/+ (E V +0.38 eV) Fe i FeB 0/+ (E V +0.1 eV) valence band 5

Recombination activity of Fe in p-type silicon • Different energy levels and capture cross sections - different lifetime curves • SRH modelling on left, QSSPC data on right (trapping restricts range) 1 Ω .cm p -type Si 100 Carrier lifetime ( µ s) 10 FeB lifetime crossover 1 Fe i lifetime point 12 13 14 15 16 17 10 10 10 10 10 10 Excess carrier density ∆ n (cm -3 ) Macdonald et al. J. Appl. Phys. 95 1021 (2004) 6

Detecting Fe i using FeB pairing • Manipulating the SRH equations shows that: 16 cm -3 , [Fe 12 cm -3 p -Si, N A =10 i ]=10 1000   1 1   Fe i lifetime = × − [ Fe ] C Carrier lifetime ( µ s)   τ τ i FeB lifetime   Fe FeB Auger lifetime i 100 • Zoth and Bergholz developed a famous method based on SPV SPV • Extended to other methods (uW- µ W-PCD PCD and QSSPC) QSSPC 10 crossover point • Very sensitive (similar to DLTS) 11 12 13 14 15 16 17 • Only works in p-Si 10 10 10 10 10 10 10 -3 ) Excess carrier concentration (cm • Must avoid the crossover point 7

Detecting Fe i using FeB pairing   1 1   = × − -13 6x10 [ Fe ] C   τ τ i   16 cm -3 N A = 3x10 Fe FeB -13 5x10 i 16 1x10 • Pre-factor C is a function of -13 15 4x10 3x10 15 1x10 -3 ) doping and excess carrier -13 3x10 1/C ( µ s cm 14 3x10 density. 14 1x10 -13 2x10 • In true low injection, C becomes -13 constant – ideal measurement 1x10 region. 0 • Not accessible to QSSPC or -13 -1x10 uW-PCD (trapping effects). 9 11 13 15 17 10 10 10 10 10 Excess carrier density ∆ n (cm -3 ) Macdonald et al. J. Appl. Phys. 95 1021 (2004) 8

Iron imaging with photoluminescence (PL) • Band-to-band PL imaging - rapid and highly-resolved method for low-injection lifetime imaging – no trapping effects. • Allows low-injection Fe imaging - similar to original SPV technique • Two PL images required, before and after breaking FeB pairs. Auger 100% Fe i : 0% FeB 1000 lifetime 75% : 25% 50% : 50% Carrier lifetime ( µ s) 25% : 75% 0% : 100% 100 PL imaging SPV µ W-PCD 10 QSSPC crossover point p -Si, N A =10 16 cm -3 , [Fe i ]+[FeB]=10 12 cm -3 1 10 11 10 12 10 13 10 14 10 15 10 16 10 17 Excess carrier concentration ∆ n (cm -3 ) Macdonald et al. J. Appl. Phys. 103 , 073710 (2008) 9

Recombination activity of Fe in n-type silicon • Neutral charge state in n-type – less attractive for minority carriers compared to p-type. • Higher lifetime in n-type in low- to mid-injection. • Possible incentive for using n-type substrates… 10000 Recombination lifetime ( µ s) n-type FZ Si 10000 Effective lifetime τ Fe+intrinsic ( µ s) N D =2.4x10 15 cm -3 1000 [Fe i ]=3.8x10 12 cm -3 1000 n-type 100 100 10 p-type 10 p-type FZ Si 1 N A =2.8x10 15 cm -3 1 0.1 [Fe i ]=3.4x10 12 cm -3 16 cm -3 , 0.1 suns illumination N A/D = 10 0.1 0.01 10 12 10 13 10 14 10 15 10 16 10 17 10 11 12 13 14 10 10 10 10 10 Excess carrier concentration (cm -3 ) -3 ) Interstitial Fe concentration [Fe i ] (cm Macdonald and Geerligs, Appl. Phys. Lett. 85 4061 (2004) 10

Fe images on mc-Si • Wafer 20% from bottom of ingot High [Fe i ] (10 13 cm -3 ) • • Internal gettering of Fe during ingot cooling at GBs, dislocation clusters Liu et al. Progress in PV, 19 649 (2011) 11

Fe images on mc-Si • Wafer from near very bottom of ingot Moderate [Fe i ] (10 12 cm -3 ) • • Small grains • Fewer dislocation clusters • Lower [Fe i ] within grains – presence of precipitation sites Liu et al. Progress in PV, 19 649 (2011) 12

Fe images on mc-Si • Wafer from middle of ingot Low [Fe i ] (10 11 cm -3 ) • • Reduced gettering at GBs (due to precipitation starting at lower T) Liu et al. Progress in PV, 19 649 (2011) 13

Internal gettering of Fe at GBs during ingot cooling • Line-scans of PL images with resolution of 25 microns • 1D diffusion/capture model – 2 free parameters – diffusion length of Fe i L D (Fe i ) and precipitation velocity P of the GB • Have to take care of PL artifacts! – Image smearing in the CCD camera - use point-spread function de-convolution – Carrier spreading in the sample – use low-lifetime – i.e. high [Fe i ] samples) Liu and Macdonald, IEEE JPV 2 , 479, (2012) 14

Internal gettering of Fe at GBs during ingot cooling • Same GB on different wafers reveals diffusion length of Fe i L D (Fe i ) depends on initial [Fe i ] • Lower [Fe i ] means that precipitation begins later during cooling • Modelling ingot cooling time reveals data can only be explained if precipitation at GBs commences after super-saturation ratio of about 50 is reached! Liu and Macdonald, IEEE JPV 2 , 479, (2012) 15

Internal gettering of Fe at GBs during annealing • Annealing at low temps can drive further precipitation at GBs, and within grains. • Higher temperatures tend to homogenise the dissolved Fe 16

Internal gettering of Fe at GBs during annealing • At 600 ºC, [Fe i ] is far above solubility limit: • Strong super-saturation • Drives precipitation at GBs, and within grains • At 800 ºC, [Fe i ] is approx equal to the solubility limit: • No precipitation • At 900 ºC and above, [Fe i ] is below solubility limit: • No precipitation • Homogenization by diffusion • Dissolution of precipitates also possible 17

Internal gettering of Fe at GBs during annealing • 500 C annealing for various times • 1D diffusion/capture model: • Widening denuded zone • Reduced intra-grain [Fe i ] 18

Internal gettering of Fe at GBs during annealing • At fixed temp annealing, diffusion length of Fe can be calculated from literature values. • Very good agreement with fitted values (500 ºC) • A method to measure diffusivity? 19

Internal gettering of Fe within the grains • After homogenization on left, after 14 hour anneal at 500 °C on right. • Precipitation rate varies from grain to grain. 20

Internal gettering of Fe within the grains • Precipitation rate varies despite initial Fe concentrations being similar 21

Internal gettering of Fe within the grains Final [Fe i ] with respect to intra grain dislocation density, after annealing at 500 o C for 14.5 hours Microscope image of a defect etched mc-Si wafer • Less Fe precipitation in grains of low dislocation density • Average distance between dislocations is less than Fe diffusion length • Dislocations act as nucleation sites for Fe precipitation within grains 22

External gettering of Fe i by P, Al and B diffusions • Fe-implanted, annealed, mono FZ p-Si, detected by QSSPC. implant Fe anneal P, B or Al diffusion 23

Phosphorus gettering of Fe i • P gettering removes between 90-99% of Fe – better at lower temp • Adding a post-getter anneal improves gettering further – segregation ratio improves 100 Percentage of Fe remaining (%) 10 1 0.1 0.01 C C C C - - - - - - - - - - - - 0 0 0 0 5 0 5 5 8 8 7 6 , , , + P P P 0 5 8 , P Phang and Macdonald, JAP 109, 073521 2011 24

Phosphorus gettering of Fe i • Driving in P diffusion reduces gettering effectiveness Very heavily doped region (>10 20 cm -3 ) required for best gettering • Phang and Macdonald, IEEE JPV accepted 25