progress of ibc si solar cells - apviapvi.org.au/solar-research-conference/wp-content/... · anu...

Progress of IBC Si solar cells(ACAP PP1.2a)

ANU: Kean Chern Fong, Teng Choon Kho(PhD), WenSheng Liang, Pheng

Phang, Marco Ernst, Daniel Walter, Matthew Stocks, and Andrew Blakers.

PVLighthouse: Keith McIntosh

Contact: kean.fong@anu.edu.au

ANU 25% IBC Solar Cell

In 2018, ANU received certification

for a 24.95 ± 0.61% single-junction

Si solar cell.

– VOC 717 mV

– JSC 42.9 mA cm-2,

– FF 81.1

Key features of the cell:

– Oxide-Nitride-Oxide passivation

– High bulk lifetime >50ms

– Excellent light trapping

2

ANU 25% IBC Solar Cell

3

Front surface:• Textured

• Phosphorus diffusion

• Oxide-Nitride-Oxide

TLM

Finger resistance

test structures

Dielectric

test structures

IBC Cells

-

+

Rear surface:• Non-textured

• Phosphorus diffusion

• Oxide-Nitride-Oxide

ANU 25% IBC Solar Cell

4

N-type BSF

+ metal finger- metal finger

Light Phosphorus

- metal finger

P-type emitter

Cross section

Bottom-up

View

Light Phosphorus

(ONO)

(ONO)

(100 Ω cm)

5

Key design improvements

Franklin E., Fong K.C., et al, Progress in Photovoltaics: Research and Applications, 2014

Need high lifetimeTR, Gettering, Cz → FZ,

1.5 →100 Ω-cm

Need better optics,

Thinner wafer or edge

treatment

Front Surface: ONO

Reduce Boron regions

Thinner wafer

Loss analysis of (2014) 24.4% IBC cell:

Oxide-Nitride-Oxide

6

Si wafer

Thermal oxide

PECVD SiNx

PECVD SiOx

Si wafer

Origin:

The idea of ONO passivation is derived from SONOS

memory cells, which has properties of being chargeable

with good charge retention capacity.

It turns out, that an ONO performs excellently as a

passivation layer, and surface charges can be tuned from

+ve to –ve.

Oxide-Nitride-Oxide

7

O-N-O on undiffused Si

T.C. Kho, K.C. Fong, et al., “Exceptional silicon surface passivation by

an ONO dielectric stack”, Solar Energy Materials and Solar Cells, 2018.

Si waferThermal oxide

PECVD SiNx

PECVD SiOx

Planar:

• < 0 fAcm-2, exceeding (Richter)

Auger lifetime limits.

Random Textured:

• ~ 1-5 fAcm-2

Phos diffused ONO

8

Si wafer

O-N-O

O-N-O

Phos diffusion

Phos diffusionSi wafer

O-N-O

O-N-O

Phos diffusion

Phos diffusion

T.C Kho, PhD thesis , in preparation. (est. 2019)

Boron diffused- ONO

• Uncharged 25 fAcm-2

• Negatively charged J0 at 9 fAcm-2

9

Si wafer

O-N-O

O-N-O

Boron diffusion

Boron diffusion

T.C Kho, PhD thesis , in preparation. (est. 2019)

Optics of ONO

10

43.6543.20

43.60

43.30

43.55

43.50

43.40

43.300 5 10 15 20 25 30

36

40

44

48

52

56

60

64

68(a) ONO OPAL2 Simulation

Thic

kness o

f S

iNx (

nm

)

Thickness of thermal SiO2 (nm)

42.70

42.90

43.10

43.30

43.50

43.70

JGen

(mA/cm2)

0.0

0.5

1.0

1.5

2.0

Spectr

al in

t. (

W/m

2/n

m)

400 600 800 1000

0

5

10

15

20

25

30

35

40(b) Reflectance Measurement

Simulated SiNx

Simulated NO

Simulated ONO

Measured ONO Cell

Wavelength (nm)

Reflecta

nce (

%)

25% cell

T.C Kho, PhD thesis , in preparation. (est. 2019)

Optimal Optics is achieved when SiO2 is thinnest.

• Trade-off between optics & passivation

Application of ONO to IBC cells

11

7 fAcm-2

(~500 Ω∕)

2 fAcm-2

(~500 Ω∕)

16 fAcm-2

(~170 Ω∕)210 fAcm-2

(~15 Ω∕)

Area Weighted J0 indicates the surfaces and contacts

contribute approximately 16 fAcm-2 to the entire cell.

Improvement to Texturing

• ANU adopted the use of monoTEX additive for the formation

of random pyramid structures.

– Physical : Uniform, repeatable, smaller pyramids (~3um), consumes less silicon

(~10um)

– Electrical : Slightly higher bulk lifetimes achieved in controlled test versus TMAH-

IPA texturing.

12

W.S.Liang, K.C. Fong, et al, “TMAH etchant and monoTEX agent induced highly reproducible and non-metal contamination

c-Si texturing“, ASPRC 2018

TMAH + MonoTEX

(presented improvement)TMAH + IPA

Quantum Efficiency

13

+0.5 mA cm-2-0.04 mA cm-2

300-400 nm

+0.4 mA cm-2

400-950 nm 950-1200 nm

IV measurements

14

• High VOC of 717 mV..

• Bulk lifetime >50ms

before metallisation.

• High JSC of

42.9mAcm-2.

• But weak knee, low

FF of 81.1.

Future Improvements

• Low FF, identifying

source of non-ideality

• Application of

passivated contacts in

combination with ONO

15

T.C Kho, in preparation for presentation at SiliconPV 2019.

ANU IBC Cell progression

16

+ metal finger

(shaded)- metal finger

(shaded)

Light phosphorus

Metal contact

(black dots)

- metal finger

(shaded)

Heavy Boron

Boron diffusionundiffused

Heavy Phos

diffusion

2013: 24.4% (independently confirmed)

• 703mV, 42.0 mAcm-2, FF 82.7%

• 1.5 Ohm-cm n-type

• PECVD SiN/SiO front, LPCVD SiO/SiN rear.

2016: 24.7% (in-house)

• 714mV, 42.0 mAcm-2, FF 82.3%

• High resistivity (100 Ohm-cm)

• ONO passivation

• Phosphorus gettering

2018: 25.0% (independently confirmed)

• 717mV, 42.9 mAcm-2, FF 81.1%

• Phos + BRL gettering

• Improved texturing (MonoTex)

undiffused

Conclusion

• 25% efficient IBC cells fabricated at ANU.

• Key features: ONO, TMA-MonoTEX, Bulk lifetime (TR,

BRL & Phos gettering.)

• Publications:– T. K. Choon, K.C. Fong, et al., SOLMAT, 2018.

– W. S. Liang, K.C. Fong, et al., ASPVC, 2018.

– E. Franklin, K.C. Fong, et al., PIPV, 2016.

– K.C Fong, et,al, JPV, 2016.

– K.C Fong, et,al, JPV, 2015.

– A.Fell, K.C. Fong, JPV, 2014

– K.R. McIntosh, PVSC 2014, ... many more...

• Open to collaboration.

17