Understanding the boron-oxygen defect: properties, kinetics and deactivation mechanisms

Transcription

1 Faculty of Engineering \ School of Photovoltaic and Renewable Energy Engineering Advanced Hydrogenation Group Understanding the boron-oxygen defect: properties, kinetics and deactivation mechanisms SPREE Seminar University of New South Wales Sydney, Australia School of Photovoltaic & Renewable Energy Engineering 9 November 2017 Nitin Nampalli Supervisors: Malcolm Abbott, Stuart Wenham, Matthew Edwards

2 Motivation p-type silicon dominant for forseeable future PERC cells seeing an increasing market share Surface passivation quality is improving Reliable kwh is increasingly important Strong imperative to: (a) Improve bulk quality (b) Minimise cell degradation International Technology Roadmap for Photovoltaics, 8 th Edition, 2017.

3 Motivation p-type silicon dominant for forseeable future PERC cells seeing an increasing market share Surface passivation quality is improving Reliable kwh is increasingly important Strong imperative to: (a) Improve bulk quality (b) Minimise cell degradation International Technology Roadmap for Photovoltaics, 8 th Edition, 2017.

4 Motivation p-type silicon dominant for forseeable future PERC cells seeing an increasing market share Surface passivation quality is improving Reliable kwh is increasingly important Strong imperative to: (a) Improve bulk quality (b) Minimise cell degradation International Technology Roadmap for Photovoltaics, 8 th Edition, 2017.

5 Motivation p-type silicon dominant for forseeable future PERC cells seeing an increasing market share Surface passivation quality is improving Reliable kwh is increasingly important Strong imperative to: (a) Improve bulk quality (b) Minimise cell degradation International Technology Roadmap for Photovoltaics, 8 th Edition, 2017.

6 Motivation Boron-oxygen defects are the most important source of LID in commercial Cz solar cells Cell efficiency loss: PERC: 1-12% rel Al-BSF: 1-6% rel Mitigation of BO defects is of vital importance For a cell manufacturer producing 1,000 MW p /year: MW p /year in lost production USD $12-24 million/year in lost savings

7 Open Circuit Voltage The boron-oxygen defect B O Light Soaking (1 sun, 25 C) causes Light Induced Degradation Illuminated Annealing Dark Regeneration Annealing (~1 sun, (200 >85 C) C) but can be Light Soaking (1 sun, 25 C) Temporarily Deactivated Permanently Deactivated Stable Deactivation (Passivation) Deactivation is unstable! Time Time Time

8 The boron-oxygen defect Behaviour described by 3-state model Focus of this work: 1. Recombination properties 2. Reaction kinetics 3. Deactivation mechanisms State A State C 2 Annihilation Direct stabilisation Direct destabilisation Destabilisation Degradation State B 1 3 Regeneration

9 Recombination properties of the boron-oxygen defect

10 Recombination properties of BO 1 = 1 + τ eff,da τ surf 1 τ bulk,non BO DA Si wafer BO? 1 = τ eff,ls τ surf τ bulk,non BO 1 τ SRH,BO LS BO p 1 = n i,eff e E i E trap k B T 1 τ SRH = n τ p0 n + n 1 np n i 2 + τ n0 p + p 1 1 τ n0 = N trap σ n v th,e 1 τ p0 = N trap σ p v th,h k = σ n σ p

11 Recombination properties of BO N trap E trap k TDLS Is the value of E trap,bo correct? What is the real value of k BO? Statistical uncertainty? IDLS Is k BO temperature-dependent? k BO = 14 k BO = 9.3 S. Rein and S. Glunz, Applied Physics Letters 82(7), pp , X. Wang et al., Energy Procedia 55, pp , 2014.

12 Recombination properties of BO N trap N trap = 1 τ n0 σ n v th,e E trap At low injection (Δn = 0.1 N A ): k N trap 1 τ SRH,BO 1 τ LS 1 τ DA = NDD (Normalised Defect Density) NDD(t) NDD max

13 Recombination properties of BO º N trap E trap k Firing What is the impact of firing on NDD? Could k also be affected by firing? K. Bothe et al., Solid State Phenomena 95-96, p (2004) D. Walter et al., Solar Energy Materials and Solar Cells 131, p (2014)

14 Recombination properties of BO SRH properties of the BO defect ( k, E trap ) Impact of firing on k Impact of firing on defect density

15 Recombination properties of BO SRH properties of the BO defect ( k, E trap ) Impact of firing on k Impact of firing on defect density

16 SRH Recombination Properties Methods for determining recombination properties: IDLS: Injection-Dependent Lifetime Spectroscopy A common characterization method Good sensitivity to k (if E trap is known) Low sensitivity to E trap (for mid-gap defects) TDLS: Temperature-Dependent Lifetime Spectroscopy Good sensitivity to E trap, k Analysis at single injection level (Δn) TIDLS: Temperature- and Injection-Dependent Lifetime Spectroscopy Best sensitivity to E trap, k Analysis over full range of Δn

17 Inverse Lifetime (ms -1 ) SRH Recombination Properties IDLS analysis 1 = τ DA τ surf τ bulk,fixed 1 = 1 + τ LS τ surf 1 τ bulk,fixed + 1 τ SRH,non BO 1 τ SRH,non BO + 1 τ SRH,BO Method B (Constrained) Method D (Free J 0e ) 15 k 1 = k 1 = J 0e,DA = J 0e,LS = 84.7 fa/cm 2 GOF Q DA = 1.60 Q LS = 2.93 Q global = 2.22 J 0e,DA = 84.7 fa/cm 2 J 0e,LS = 80.8 fa/cm 2 GOF Q DA = 1.60 Q LS = 1.02 Q global = Excess Carrier Density, n (cm -3 ) N. Nampalli et al., Frontiers in Energy 11(1), 2017.

18 SRH Recombination Properties IDLS analysis 1 = τ DA τ surf τ bulk,fixed 1 = 1 + τ LS τ surf 1 τ bulk,fixed + 1 τ SRH,non BO τ SRH,non BO τ SRH,BO Median MAD A B C D E F G H I J Not all fitting methods are accurate Best Fit Capture Cross Section Ratio, k k BO > 9.3 k k BO > 9.3 BO = 11.9 ± 0.45 A B C D E F G H I J k lit = 9.3 Constrained Partially unconstrained Free N. Nampalli et al., Frontiers in Energy 11(1), 2017.

19 Inverse lifetime at elevated T, DA (T) -1 (s -1 ) Inverse lifetime at elevated T, LS (T) -1 (s -1 ) SRH Recombination Properties TIDLS analysis Elevated T (K) Increasing T Elevated T (K) Increasing T Excess Carrier Density, n (cm -3 ) Excess Carrier Density, n (cm -3 )

20 SRH Recombination Properties TIDLS analysis Determine E trap, k E C E trap = 0.41 ± 0.10 k BO = 11.5 ± 1.00

21 SRH Recombination Properties TIDLS analysis Determine E trap, k E C E trap = 0.41 ± 0.10 k BO = 11.5 ± 1.00 τ n0 T = N trap σ n T v th,e T 1 T-dependence of τ SRH,BO α BO = 2.3 σ n T = σ n 300K Assumption: α n = α p = α BO T 300 α n σ n/p T T 2.3

22 Recombination properties of BO SRH properties of the BO defect ( k, E trap ) Impact of firing on k Impact of firing on defect density

23 Impact of firing Experiment details NDD Meas. 1 PECVD SiN x :H Measure Normalised Defect Density (NDD) Dark Anneal: 200 o C, 15 min Light Soak: 35 o C 0.77 suns, 48 hr

24 Temperature, ( C) Impact of firing Experiment details NDD Meas. 1 NDD Meas. 2 PECVD SiN x :H Relative change due to firing Varying T peak Fired N fire = 1, 2, 3 Not Fired T peak = 827 C T peak = 790 C T peak = 735 C T peak = 686 C T peak = 632 C T peak = 580 C T peak = 518 C T peak = 463 C Time, t (s)

25 Impact of firing on k BO Best Fit Capture Cross Section Ratio, k After Firing Before Firing New, non-bo CID defect k BO 4 N fire T peak ( C)

26 Normalised Defect Density, NDD after (ms -1 ) Impact of firing on NDD BO 100 New, non-bo CID defect After Firing Before Firing 10 N fire T peak ( C)

27 Summary Properties of the BO defect SRH properties of BO: Determined k BO = 11.9 (> 9.3) Confirmed that E trap = E C 0.41 ev Determined that σ n/p (T) T 2.3 Impact of firing on BO properties: Confirmed that k BO is not affected by firing Demonstrated that firing reduces NDD BO Firing can induce other (non-bo) CID defects in Cz silicon

28 Transition kinetics of the BO system

29 Defect transition rate constant, ij ij (s -1 ) ) Transition kinetics of the BO system κ BA κ AB κ CB κ BC Dependence of transition rates on T and illumination (Δn) are of vital importance! Temperature, T ( C) AB BA BC CB

30 Transition kinetics of the BO system Degradation: κ AB appears to be independent of illumination intensity (>0.1 suns) Other studies show κ deg p 0 2 Which is true? 2 κ AB p 0 κ AB p p 0 κ AB p 2 J. Schmidt and K. Bothe, Physical Review B 69, p , K. Bothe and J. Schmidt, Journal of Applied Physics 99, p , 2006.

31 Transition kinetics of the BO system Annealing: Known to occur in dark (no carrier dependence assumed) But one study showed κ BA 1 p 0 for compensated Si Does κ BA have a carrier dependence? B. Lim et al., Applied Physics Letters 98, p , 2011.

32 Issue with reaction rate studies Process T Measurement T! p(t) = p 0 (T) + Δp(T) p(300k) = p 0 (300K) + Δp(300K) G T = n T τ eff T, n G 300K = n 300K τ eff 300K, n Need a method to determine n T from τ eff 300K and G 300K

33 Reaction kinetics of the BO system Model to obtain τ eff (T) from τ eff (300 K) Temporary deactivation (annealing) kinetics Degradation kinetics

34 Generation rate correction factor, f corr (T) Temperature dependence of lifetime n T G T = τ eff T, n 1.06 Data Fit Measurement temperature, T (K)

35 Temperature dependence of lifetime n T G T = τ eff T, n 1 τ DA T, Δn = 1 τ surf T, Δn + 1 τ bulk,fixed T + 1 τ SRH,non BO T 1 τ LS T, Δn = 1 τ surf T, Δn + 1 τ bulk,fixed T + 1 τ SRH,non BO T + 1 τ SRH,BO T, Δn τ DA T, Δn τ LS T, Δn

36 Temperature dependence of lifetime n T G T = τ eff T, n 1 τ DA T, Δn = 1 τ surf T, Δn + 1 τ bulk,fixed T + 1 τ SRH,non BO T 1 τ LS T, Δn = 1 τ surf T, Δn + 1 τ bulk,fixed T + 1 τ SRH,non BO T + 1 τ SRH,BO T, Δn τ n0,bo T = τ n0,bo 300 K T 300 α n0,bo τ n0,non BO T = τ n0,non BO 300 K T 300 α n0,non BO τ bulk,fixed T = τ bulk,fixed 300 K T 300 α bulk,fixed Param T = Param 300K T 300 α param 1 τ surf T, Δn = 2 p 0 + Δn q W 1 τ surf T, Δn = 2 S eff(300 K W J 0e 300K T 300 T 300 α J0e α Seff 1 + Δn p 0

37 Parameter value relative to 300 K J 0e relative to 300 K Temperature dependence of lifetime 10 Temperature, T ( C) J 0e bulk,fixed n0,bo n0,non-bo S eff T / 300 K Parameter Relevant lifetime component Exponent of temperature dependence Value for α param τ bulk,fixed * τ bulk,fixed (T) α b ± S eff τ surf (T) α Seff ± J 0e τ surf (T) α J0e ± τ n0,bo τ SRH, BO (T) α n0,bo ± τ n0,non-bo * τ SRH,non-BO (T) α n0,non-bo ± * may be specific to wafers used in this study

38 Reaction kinetics of the BO system Model to obtain τ eff (T) from τ eff (300 K) Temporary deactivation (annealing) kinetics Degradation kinetics

39 Inverse lifetime, T (t,t=300 K) (ms -1 ) Annealing kinetics Defect dissociation T = 478 K (205 C) T = 448 K (175 C) T = 438 K (165 C) T = 433 K (160 C) T = 428 K (155 C) (B A) κ BA 1 p(t) Anneal time, t (s) LS BO DA

40 ln( BA [s -1 ] ) Annealing kinetics κ BA κ BA (p/n V ) Dark anneal temperature ( C) E a,ba = ev BA ' = s This work Rein et al. (2001) Schmidt et al. (2004) Schmidt et al. (2002) Lim et al. (2011) Fit line to data in this work /T (K -1 ) E a,ba κ BA = ν BA e k B T κ BA = ν 0,BA N C (T p(t) Ea,BA e k B T

41 Reaction kinetics of the BO system Model to obtain τ eff (T) from τ eff (300 K) Temporary deactivation (annealing) kinetics Degradation kinetics

42 Degradation kinetics κ AB p T 2 DA LS BO J. Schmidt and K. Bothe, Physical Review B 69, p , 2004.

43 Degradation kinetics κ AB κ AB (p/n C ) 2 (This work) E a,ab κ AB = ν AB e k B T κ AB = ν 0,AB p(t) N C (T) 2 E a,ab e k B T K. Bothe and J. Schmidt, Journal of Applied Physics 99, p , 2006.

44 Implications of degradation kinetics Fast initial degradation ( FRC ) May be partially related to p 2 dependence κ AB is not constant (due to p 2 dependence) Fast initial degradation occurs in same time scale as change in κ AB K. Bothe and J. Schmidt, Journal of Applied Physics 99, p , 2006.

45 Implications of degradation kinetics Regeneration rates Regen (B C) occurs only after degradation (A B) κ BC will be limited by κ AB A: ~100% B:?% B: (<100%) κ C: ~100% BC must be determined starting from State B (LS), not State A (DA) Starting State: B (fast recovery) Starting State: A (slow recovery) A. Herguth et al., Progress in Photovoltaics: Research and applications 16, pp , 2008 P. Hamer et al., Solar Energy Materials and Solar Cells 145, pp , 2016

46 Parameter value relative to 300 K J 0e relative to 300 K Summary Reaction Kinetics Developed parameterization to obtain τ eff (T) from τ eff (300 K) Temperature, T ( C) J 0e bulk,fixed n0,bo n0,non-bo S eff T / 300 K Carrier dependence confirmed for annealing (A B), degradation (B A) Carrier dependence explains other observed kinetics phenomena A, B: ~0 % C: ~1 00 %

47 Mechanisms for permanent deactivation of BO defects

48 Permanent deactivation State A Firing Direct stabilisation Direct destabilisation State C Annihilation Destabilisation Degradation Why does permanent deactivation occur? Regeneration 1 State B Defect precursors Thermal formation BO defect 2 Thermal deactivation D. Walter et al., Solar Energy Materials and Solar Cells 131, p (2014) K. Bothe et al., Solid State Phenomena 95-96, p (2004) V. Voronkov and R. Falster, Physica Status Solidi C 13(10-12), p (2016)

49 Deactivation mechanisms Why does regeneration occur? Why does thermal deactivation occur? Are they related?

50 Experimental Details NDD Meas. 1 PECVD SiN x :H Thermal SiO 2 Measure Normalised Defect Density (NDD) Dark Anneal: 200 o C, 15 min Light Soak: 35 o C 0.77 suns, 48 hr Bare

51 Regeneration NDD Meas. 1 NDD Meas. 2 NDD Meas. 3 PECVD SiN x :H (hydrogen) Relative change due to firing Relative change due to regeneration Thermal SiO 2 (no hydrogen) Bare (no hydrogen) Fired Not Fired Regeneration Illuminated Anneal: 172 o C, 0.66 suns, 2 hr

52 Regeneration Rel. FDD change after in regen. NDD due (rel. to NDD regen. ) after Some improvement in hydrogen-lean wafers Hydrogen-related reduction in FDD due to regeneration Regeneration observed in hydrogen-lean samples (very long time scales) Fired Non-fired Fired Non-fired Fired Non-fired No dielectric SiN x :H SiO 2 D. Walter and J. Schmidt, Solar Energy Materials and Solar Cells 158, p (2016)

53 Deactivation mechanisms Why does regeneration occur? Why does thermal deactivation occur? Are they related?

54 Temperature, ( C) Thermal deactivation NDD Meas. 1 NDD Meas. 2 NDD Meas. 3 PECVD SiN x :H (hydrogen) Thermal SiO 2 (no hydrogen) Bare (no hydrogen) Relative change due to firing Varying T peak Fired N fire = 1, 2, 3 Not Fired 800 T peak = 827 C Regeneration Illuminated Anneal: 172 o C, 0.66 suns, 2 hr T peak = 790 C T peak = 735 C T peak = 686 C T peak = 632 C T peak = 580 C T peak = 518 C T peak = 463 C Time, t (s) 350

55 Normalised Defect Density, NDD after (ms -1 ) Thermal deactivation 100 Wafers with SiNx:H dielectric New, non-bo CID defect After Firing Before Firing 10 N fire T peak ( C)

56 Relative NDD after firing ln [ -ln ( NDD %,BO ) ] Thermal deactivation Simple 450 Arrhenius 400 model Reaction: BO X N fire = 1 N fire = 2 N fire = 1 E a,f = 0.77 Peak Firing Temperature, T peak ( C) 0.1 ev 4.15 ± 0.6 ( f t step ) = 10 Data Model N fire = 1 All N fire N fire = 2 E a,f = ev N fire = ± 0.5 ( f t step ) = 10 N fire = 3 Linear fit (N fire = 1 only) Linear fit (all N fire ) Increasing N fire Peak firing temperature, T peak ( C) 1000/T (K -1 ) V. Voronkov and R. Falster, Physica Status Solidi C 13(10-12), p (2016)

57 Relative NDD after firing Thermal deactivation Data Model - Freeze-in Model - Freeze-in (adj.) Model - Arrhenius Peak firing temperature, T peak ( C) V. Voronkov and R. Falster, Physica Status Solidi C 13(10-12), p (2016)

58 Deactivation mechanisms Why does regeneration occur? Why does thermal deactivation occur? Are they related?

59 Normalised Defect Density (ms -1 ) Rel. change in NDD due to firing Thermal deactivation FDD after firing (rel. to NDD before ) Thermal reduction in N FDD due to firing SiN x :H, NDD before SiN x :H, NDD after SiO 2, NDD after Thermal reduction in NDD due to firing Fired Non-fired Fired Non-fired SiN x :H SiO Peak firing temperature, T peak ( C)

60 Regeneration Rel. FDD change after in regen. NDD due (rel. to NDD regen. ) after Hydrogen-related reduction in FDD due to regeneration Fired Non-fired Fired Non-fired Fired Non-fired No dielectric SiN x :H SiO 2

61 Revision to 3 state model?

62 4 state model State D High temperature Thermal effects Hydrogenation + other minor effects

63 Summary Permanent deactivation Effective regeneration requires sufficient quantities of hydrogen. Regeneration could involve other mechanisms (slower). Thermal deactivation occurs independent of regeneration. Likely not hydrogen-related. FDD after firing (rel. to NDD before ) Thermal reduction in FDD due to firing Fired Non-fired Fired Non-fired SiN x :H SiO 2 Thermal deactivation is likely to be defect dissociation 4-state model proposed to account for thermal deactivation

64 Implications of this work

65 SiN x :H SiO 2 Implications 1. Improved understanding of the BO defect Recombination properties Mechanism of permanent deactivation FDD after firing (rel. to NDD before ) Thermal reduction in FDD due to firing Fired Non-fired Fired Non-fired 4-state model Impact: - Easier identification of the BO defect - Multiple, tailored solutions to mitigate BO defect

66 Parameter value relative to 300 K J 0e relative to 300 K Implications 2. Accurate reaction kinetics modelling for BO 4-state model Temperature, T ( C) Conversion from τ eff (300K) to τ eff (T) 1 J 0e bulk,fixed n0,bo n0,non-bo 0.1 S eff T / 300 K Carrier dependence of reactions Impact: - Better estimate of regeneration time-scales for industrial wafers / solar cells

67 Acknowledgements Funding sources Australian Renewable Energy Agency (ARENA) Australian Center for Advanced Photovoltaics (ACAP) UK Institution of Engineering and Technology (IET) / A.F. Harvey Engineering Prize. Commercial partners (ARENA 1-A060)

68 Acknowledgements Supervisors: Malcolm Abbott, Stuart Wenham, Matt Edwards Hydrogenation group Laboratory: MAiA processing team LDOT team Others Admin staff: TETB SPREE Friends and family

69 Thank you for your attention

70 Motivation p-type silicon dominant for forseeable future PERC cells seeing an increasing market share Surface passivation quality is improving Reliable kwh is increasingly important Strong imperative to: (a) Improve bulk quality (b) Minimise cell degradation International Technology Roadmap for Photovoltaics, 8 th Edition, 2017.