Optical properties of Cu2ZnSnSe4 thin films by

spectroscopic ellipsometry and photoluminescence

Özden Demircioğlu1, Jose Fabio Lopez Salas1, Germain Rey2, Thomas Weiss2,

Susanne Siebentritt2, Jürgen Parisi1, Levent Gütay1 1Laboratory for Chalcogenide Photovoltaics, Institute of Physics, University of Oldenburg, Oldenburg, Germany

2Laboratory for Photovoltaics, Physics and Materials Science Research Unit, University of Luxembourg, Luxembourg

EMRS – Spring Meeting 2016

Lille, France, 2 to 6 May 2016

Spectroscopic Ellipsometry (SE)

Data analysis and Model

Results

ZnSe and MoSe2 with different amounts at back side of all samples

ZnSe fraction on front side only in sample C

Confirmation of MoSe2 layer thickness by SEM (not shown)

Presence of ZnSe determined, quantities not confirmed

Extracted band gaps: SE measurement : Surface & bulk sensitive

Sample A: 0.95 ± 0.01 eV

Sample B: 0.92 ± 0.01 eV

Sample C: 0.93 ± 0.01 eV

R&T measurement : Bulk sensitive

Sample A: 0.97 ± 0.01 eV

Sample B: 0.94 ± 0.01 eV

Sample C: 0.95 ± 0.01 eV

Relative differences of extracted band gaps identical for SE and R&T

Differences in absolute band gaps from SE and R&T due to different

sensitivities for bulk and surface

Extraction of dielectric functions ε1 and ε2 by optical modeling of sample

Excitation wavelength at 532 nm and 457.9 nm

ZnSe at back side of all samples

Different amounts of MoSe2 on remaining substrates after absorber lift-off

ZnSe at the front side only in sample C confirmation of SE results

Conclusion

Application of SE as a non-destructive technique for detection of secondary

phases in layer and/or at the back sides

- Determination of presence and amount of MoSe2 at back side by SE

- Determination of presence of ZnSe for front and back side by SE

( confirmed by Raman spectroscopy and photoluminescence)

Reliable band gap determination by SE, band gap variations between

samples are identical for both methods

Selective probing of ZnSe defect peak PL by choice of excitation wavelength

Acknowledgements The work at the LCP research group (Oldenburg) is funded by EWE AG, Oldenburg, Germany, and the

BMBF (German ministry of Education and Science), funding Nr. 03SF0530A (project “Free-Inca”). The

work at the LPV group (Luxembourg) was supported by the FNR (National Research Fund) (Luxembourg).

Contact: oezden.demircioglu@uni-oldenburg.de

Raman spectroscopy

Photoluminescence (PL)

(ħω) ( Δ, Ψ ) surface roughness

Absorber

(CZTSe)

Intermix layer

Substrate (Mo)

MoSe2

Intermix layer

ZnSe

Intermix layer

EMA- (ZnSe:CZTSe)

Energy fit

range (eV)

0.75-3.0

Roughness

(nm)

EMA

(nm)

CZTSe

(nm)

Interface layer (nm)

MSE

(ZnSe:CZTSe)

mix. ZnSe mix. MoSe2 mix.

Sample A 14.0 0.0

(0.0:100)

1221 0.5 2.0 0.0 4.0 4.0 2

Sample B 0.5 0.0

(0.5:99.5)

1079 2.5 8.0 2.0 3.0 0.0 3

Sample C 12.0 7.0

(40:60)

1323 0.0 5.0 0.0 200 15.0 26

PL from CZTSe absorber at 0.83-0.86 eV

ZnSe defect transition at 1.2 eV at back side of each sample

Intensity dependence of ZnSe defect peak on excitation wavelength

No observation of ZnSe defect transition at front sides of samples

SE and Raman are more sensitive to detect ZnSe (compare sample C)

References [1] R. Djemour, M. Mousel, A. Redinger, L. Gütay, A. Crossay, D. Colombara, P. J. Dale, S.

Siebentritt, Appl. Phys. Lett. 102,222108 (2013)

[2] R.Kondrotas, R. Juskenas, A. Naujokaitis, G. Niaura, Z. Mockus, S. Kanapeckaite, B. Cechavicus,

K. Juskevicius, E. Saucedo, Y. Sánchez, Thin Solid Films 589 (2015) 165-172

α=𝐴(ℎ𝑣−𝐸𝑔)

ℎ𝑣

1/2

α:absorption coefficient

A:constant

hν:photon energy

α =−1

𝑡ln(

𝑇

(1−𝑅)2)

t:thickness

T:transmission

R:reflectance

*

*

*

*

* Surface and intermix layers modelled by

Effective Medium Approximation (EMA),

contain 50:50 mixture of adjacent layers.

Variable mixture of ZnSe:CZTSe layer.

Overview

Investigation of Cu2ZnSnSe4 (CZTSe) thin films by spectroscopic ellipsometry

(SE), Raman, photoluminescence (PL) and optical reflection & transmission

(R&T)

SE analysis for non-destructive detection of MoSe2 layer at the back and

ZnSe secondary phase at the front and back interface region of absorber

Complex optical modeling including multiple layers for evaluation of SE data

Confirmation of SE results by PL and resonant Raman spectroscopy,

Comparison of band gaps from SE and R&T

Samples:

-Samples A and B from a high temperature co-evaporation process

-Sample C from co-evaporation followed by an annealing step

-Back side of absorber accessed by lift-off from substrate

Black markers: CZTSe Red markers: ZnSe Orange markers : MoSe2