Investigations using 532nm Nanosecond Lasers Lujia Xu, Klaus Weber, - PowerPoint PPT Presentation

Understanding Laser Doping: Investigations using 532nm Nanosecond Lasers Lujia Xu, Klaus Weber, Andreas Fell, Evan Franklin, Ziv Hameiri

Outline • Motivation • Dopant measurement by Scanning Electron Microscopy • Characterisation of dielectric window edge regions • Impact of laser parameters / dielectrics on recombination in laser doped regions 2

Motivation All laser doped IBC cell concept High quality n-type wafer surface passivation Localized base contact: Localized pn-junction: n ++ laser doped / opened p ++ laser doped / opened 3

Motivation Simulated efficiency potential (Quokka) 4

First all-laser processed cell batch 5

Laser processing (532nm DPSS) Scanning mirror Optical ( mirrors, beam expander, collimation lens, and etc. ) Laser head F-theta lens Sample x-y-movable and rotatable stage 6

Laser doping • Small, inhomogeneously doped areas • Rough surfaces • How to characterise? • EBIC / SIMS? • SEM can be used to image dopant density under certain conditions n diffusion / p substrate p diffusion / n substrate 7

SEM Doping Contrast – Laser doped regions p doping / n substrate n doping / p substrate L. Xu, et al., "Secondary electron microscopy dopant contrast imaging (SEMDCI) for laser doping" IEEE Journal of Photovoltaics , vol. 3, pp. 762-768, 2013. 8

Comparison – SEM DCI / EBIC L. Xu, et al, "Comparison between SEMDCI and EBIC for Laser Doping of Crystalline Silicon" SiliconPV 2014 , 's-Hertogenbosch, The Netherlands, 2014. 9

Comparison – SEM DCI / EBIC 10

Can SEM DCI give doping concentration? • Establish empirical relationship (diffused samples) • Apply to laser (excimer) doped samples • Compare with ECV profiles 11

Laser doping – (active) doping density maps 1E21 0 2000 4000 1E20 0 0.5 1 1.5 2 2.5 3 nm 4 x 10 (a) 40kHz 15uJ 125nm 0 1E19 2000 4000 0 0.5 1 1.5 2 2.5 3 nm 4 x 10 (b) 40kHz 17.5uJ 500nm 1E18 0 1000 1E17 nm Concentration 0 1000 2000 3000 4000 color bar (cm -3 ) (c) Boron diffusion 12

SEM DCI – Surface profile characterisation Optical microscope image SEM DCI 13

Application of SEM DCI Single step doping process – what happens at the dielectric edges? How to check the edge region? 14

Application of SEM DCI 1. Use dielectric as etch mask, then use SEM 15

Application of SEM DCI 2. Direct imaging with dielectric film still present 0 1000 2000 nm 0 2000 4000 6000 8000 10000 12000 14000 40kHz 17.5uJ 1000nm Dielectric film Laser doping region 0 500 1000 1500 nm 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 40kHz 15uJ 125nm 16

Experiment: laser doping process SOD Dielectric Silicon 17

Application of SEM DCI 15000 15 10 10000 Gap distance (µm) Si 3 N 4 5000 5 60/170nm SiO2 19nm 0 0 Si -5000 -5 Thin Thick -10 -10000 125 250 500 125 250 500 1000 250 500 1000 2000 12.5 15 17.5 Pulse distance (nm) Pulse energy (µJ) L. Xu at el., "The impact of SiO 2 /SiN x stack thickness on laser doping of silicon solar cell" IEEE Journal of Photovoltaics , vol. 4, pp. 594-600, 2014. 18

Assessing damage from laser doping J0 depends on What really matters in the end? • 3-D doping profile  Dopant source  Laser properties  Dielectric  Surface roughness  Etc… • Defects  Dopant source  Laser properties  Dielectric  Surface roughness  Etc… 19

Assessing damage from laser doping Simple experiment to focus on laser-dielectric interactions Dielectric B Diffusion Silicon 20

PL Measurement test structure Alignment mark 4mm 4mm Unprocessed, B diffused, unpassivated Laser processed Pitch distance region 30µm Processed, B diffused, unpassivated 21

PL Measurements - calibration 100 2 Lowest background J 0e 1407.3 fA/cm 2 Second lowest background J 0e 1783.5 fA/cm 2 Average background J 0e 2236.6 fA/cm 80 2 Highest background J 0e 2544.4 fA/cm 10 9 Normalized J 0e 60 8 7 6 40 5 4 3 2 20 1 0.7 0.8 0.9 1.0 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized PL 22

Results – no dielectric 1600 1000 0.25  m J oen 0.25  m R sh 20ns 0.25  m R sh 400ns 0.25  m J oen 1  m J oen 1400 1  m R sh 1  m R sh 1  m J oen 800 2.5  m J oen 100 Sheet resistance ( ฀ ) 100 Sheet resistance ( ฀ ) 2.5  m R sh 2.5  m R sh 2.5  m J oen Normalized J 0e Normalized J 0e 1200 20  m J oen 20  m J oen 1000 600 800 10 10 400 600 400 200 200 1 1 0 0 4 5 6 7 8 9 10 11 12 13 14 15 16 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 Pulse energy (  J) Pulse energy (  J) L. Xu et al., "The influence of thermal effects and dielectric films on the electronic quality of p + -doped silicon processed by nanosecond laser" IEEE Journal of Photovoltaics , vol. PP, pp. 1-8, 2014. 23

Results – evaporation threshold 18 Evaporation threshold 0.25  m parameter window 16 1  m parameter window 2.5  m parameter window 14 Pulse energy (  J) 12 10 8 6 4 2 0 20 400 100 Pulse duration (ns) 24

Results – with dielectric films 1000 1000 20ns 400ns 100 100 Normalized J 0e Normalized J 0e 10 10 1 1 0.25  m Bare 1  m Bare 2.5  m Bare 20  m Bare 0.25  m Bare 1  m Bare 2.5  m Bare 20  m Bare 0.25  m SiO 2 1  m SiO 2 2.5  m SiO 2 20  m SiO 2 0.25  m SiO 2 1  m SiO 2 2.5  m SiO 2 20  m SiO 2 0.25  m SiN x A 1  m SiN x A 2.5  m SiN x A 20  m SiN x A 0.25  m SiN x A 1  m SiN x A 2.5  m SiN x A 20  m SiN x A 0.1 0.1 4 5 6 7 8 9 10 11 12 13 14 15 16 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 Pulse energy (  J) Pulse energy (  J) 25

Conclusions • Single step laser doping processes for IBC are very challenging • SEM DCI is a convenient tool to assess laser doping near dielectric edges • Under the right conditions, good doping is achieved under dielectric • Laser induced damage silicon damage can be kept to acceptable levels • More work is needed to understand dielectric edge regions / dopant precursors 26

Acknowledgements ANU Laser team Colleagues at ISE and ISFH (S. Kluska, B. Fleischmann, S. Hopmann, K. Bothe, B. Lim) Australian Renewable Energy Agency (3-GER002 “High quality laser doping for solar cells through improved characterisation techniques”) 27