Solarzellen aus dünnen Siliziumschichten –Stand der Technik und Herausforderungen für die Zukunft
Bernd RechHelmholtz-Zentrum Berlin (HZB) and Technische Universität Berlin
Many thanks to my colleagues from HZB and FZ-Jülich (Uwe Rau et al.), Michael Powalla from ZSW – Stuttgart and industry partners
WIAS 2008
2Bernd Rech, WIAS 2008
Outline
� Motivation and Background� Thin Film Solar Cell Technologies and Applications� Amorphous and Microcrystalline Based Silicon and Tandem Cells� Poly-Crystalline Si Thin-Films� R&D Challenges and Conclusions
3Bernd Rech, WIAS 2008
Thermodynamic limits – Generation of electricity in a Carnot process:
Carnot efficieny = Tsun – Tearth
Tsun
= 95 %
This is an absolute upper limit, however, unavoidable lossesof entropy reduce the thermodynamic limit towards 85 %(see e.g. Würfel, Physik der Solarzellen)
Note: due to the Tsun of 5800 K solar radiation is of high energetic value
“Theoretical Efficiency”
4Bernd Rech, WIAS 2008
++++
++++
----
----
x
E
pppp
nnnn
----
++++
Thermalisation
Recombination
Reflection
LBLBLBLB
VBVBVBVB
EEEEGapGapGapGap
---- --------
++++++++
++++
Working Principle and Losses (p/n-junction solar cell)
voltagevoltagevoltagevoltage
ηηηη====PPPPLichtLichtLichtLicht
PPPPmaxmaxmaxmax
Working point PPPPmaxmaxmaxmax
currentcurrentcurrentcurrentdenisitydenisitydenisitydenisity
jjjjSCSCSCSC
VVVVOCOCOCOC
FF=PPPPmaxmaxmaxmax
VVVVOCOCOCOC * * * * jjjjSCSCSCSC
losses
5Bernd Rech, WIAS 2008
(η = 12 – 17 %)
c-Si solar cell
c-Si wafer technology a-Si thin-filmtechnology
(η = 5 – 7 %)Si-thickness200-300 µm
Si-thickness0.5 µm
“Some Commercial Efficiencies”
6Bernd Rech, WIAS 2008
Source Sontor
Source Solar Integrated TechnologiesSource Sulfurcell
Thin film advantages
� Material usage/cost (1-5 vs 200 µm)
� High productivity (large area)
� Monolithic series connection
� Short energy pay back time
� New products (e.g.. flexible)
7Bernd Rech, WIAS 2008
Sol
ar C
ells Compound
Semiconductor
Silicon
Organic
Dye cells
New Concepts
CdTe
ChalkopyrideCIGSSe
GaAs (conc)
Thin Film
Crystallinewafers Mono
Poly
kristallin
amorphmikrokristallin
CdTe
ChalcopyriteCIGSSe
aSi/µcSi
Crystalline
Based on M. Powalla,,ZSW
Overview of photovoltaic material classes
This talk
8Bernd Rech, WIAS 2008 Source: Brabec, MRS Bulletin, Jan. 2005
Evolution of Record Solar Cells
a-Si,µc-Si∼14%
CdTe∼17%∼20%
CIS
9Bernd Rech, WIAS 2008
Thin film PV technologies
The primary idea is a tiny amount of expensive material (1 micron or so) and lots of cheap glass and wire and metal and plastic
Ken Zweibel, NREL, 2004
Frontkontakt
Rückkontakt
encapsulation
Substrat
Superstrate technology(transparent substrate)
Absorber / p/n
front contact - TCO
back contact
Substrat
absorber - p/n-structure
light
Rückkontakt
Frontkontakt
Substrat
encapsulation
Substrate technology
Absorber / p/nback contact
front contact -TCO
substrate
glass, foil + polymer
glass, foil + polymer
absorber / p/n-structure
light
thin SiCISCdTe
10Bernd Rech, WIAS 2008
ZnO
Light
Glass (1-4mm)
TCO
n
Ag
p
i
n
pi
µµcc--Si:HSi:H
aa--Si:HSi:H∼ 3µm
a-Si technologyExample:a-Si/µc-Si tandem cell(„Micromorph“)
Mo (0.5 µm)
ZnO:Al (1 µm)
CdS (0.05 µm)
Glass (3 mm)
CIGS (2 µm)
i-ZnO (0.05 µm)
p
n
∼ 4µm
CIGS-solar cells: CuIn1-xGaxSe1-ySy
Light
CdTe-solar cells: CdTe
Light
Glass (1-4mm)
TCO
CdS (0.1 µm)
CdTe (3-8 µm)
Metal
Thin film PV technologies
ZnO (n-type)
Cu(In,Ga)(S,Se)2 (p-type)
Mo
glass substrate
Flexible solar cell on titanium foil
Example CIGS-Solar Cell
12Bernd Rech, WIAS 2008
HZB: Spin-off Company
13Bernd Rech, WIAS 2008
0,1
1,0
10,0
100,0
10 100 1.000 10.000 100.000
akumulierte Gesamtproduktion (MWp)
PV
Mod
ulpr
eis
(€/W
p)
O. Hartley, J. Malmström, A. Milner,21st EUPVSC, Dresden 2006
Cost reduction has different options
80%
status 2005
source: T. Surek NREL
long term target
Required:� Mass production� Technology development� Fundamental R&D
accumulated production (MWp)
PV
mod
ule
pric
e(€
/Wp)
Cost reduction strategies
14Bernd Rech, WIAS 2008
a view
years ago
Capacity (thin film Si)> 10 MWp< 10 MWpCapacity (CIS/CdTe)> 10 MWp< 10 MWp
Thin Film PV Fabs:under constructionor in operation
Example Germany
15Bernd Rech, WIAS 2008
a view
years ago
Capacity (thin film Si)> 10 MWp< 10 MWpCapacity (CIS/CdTe)> 10 MWp< 10 MWp
Thin Film PV Fabs:under constructionor in operation
Example Germany
16Bernd Rech, WIAS 2008
Source: A. Jäger-Waldau, PV-Status Report 2007
Announced capacity by thin film type - world
2010/2011: 6 GWp!
130 TF companiesof which 21 wereactive in 2006
Note: CdTe growth even higher(First Solar)
17Bernd Rech, WIAS 2008
Quelle: A. Jäger-Waldau, PV status report 2007
Thin Film vs. Wafer Based Crystalline Si
18Bernd Rech, WIAS 2008
Solarpark Buttenwiesen – amorphous siliconQuelle: Phönix SonnenStrom AG
Thin-Film PV applications
19Bernd Rech, WIAS 2008
Gescher-EsternEntsorgungs-Gesellschaft Westmünsterland (EGW) ,put in operation August 2006.One of the biggest roof-top installations(1.4 MWp, CdTe, First Solar) 23 430 thin-film modules on an area of ca. 17 000 m² and an investment of € 5.6 Mio.
Source: Reinecke + Pohl Sun Energy AG
20Bernd Rech, WIAS 2008
BIPV (3S AG Megaslate System)
Roof integration – family homes
Würth-Solar CIGS modules
21Bernd Rech, WIAS 2008
Optic Center Wales, CIS-Module, 85 kWP, 2004source: AVANCIS (www.avancis.de)
Solar fassade
22Bernd Rech, WIAS 2008
Semitransparent Modules
Schott Solar – a-Si
Würth Solar
23Bernd Rech, WIAS 2008
Source: Unisolar
Flexible thin-film PV
24Bernd Rech, WIAS 2008
Flexible thin-film PV
25Bernd Rech, WIAS 2008
Outline
� Motivation and Background� Thin Film Solar Cell Technologies and Applications� Amorphous Silicon and Microcrystalline BasedSilicon and Tandem Cells� Poly-Crystalline Si Thin-Films� R&D Challenges and Conclusions
26Bernd Rech, WIAS 2008
1 µm
a-Si:H
Ag
TCO
a-Si:H solar cell cross section
BIPV: Stillwell Avenue Terminal, New Yorkca. 210 kWp, installed 2004Source: SCHOTT Solar
Amorphous Silicon Based Solar Cells
27Bernd Rech, WIAS 2008
Si Si
SiSi
Si Si
SiSi
Si Si
SiSi
Si Si
SiSi
single-crystalline Sic-Si
amorphous Sia-Si:H
mikro-crystalline Si
(µc-Si:H)
5-15 % Hydrogen
Structure of Silicon
28Bernd Rech, WIAS 2008
EF
n+
p
n
pi
mono- and poly c-Si
2222----300 300 300 300 µµµµmmmm 0.3 0.3 0.3 0.3 –––– 2 2 2 2 µµµµmmmm
a-Si:H, µc-Si:H
“diffusion controlled”“interface limited”
“drift controlled”“bulk limited”
Device Structure of Si Solar Cells
29Bernd Rech, WIAS 2008
0
0,2
0,4
0,6
0,8
1
300 400 500 600 700 800 900 1000wavelength (nm)
quan
tum
effi
cien
cy
top cella-Si:H
bottom cellµc-Si:H
ZnO
Licht
glass (1-4mm)
TCO
n
Ag
p
i
n
pi
Advantages and challenges :� „red/IR-response“ µc-Si:H� no/small SWE ⇒ high stability� preparation with PECVD� indirect semiconductor: light trapping!� high growth rate and process control!
a-Si/µc-Si tandem cell
µµcc--Si:HSi:H
aa--Si:HSi:H∼ 3µm
pioneered by University of Neuchatel 1994see: Uwe Rau next talk, tomorrowfirst solar modules by Kaneka, J (2001)
Microcrystalline Silicon (µc-Si:H)
30Bernd Rech, WIAS 2008
a-Si:H/µc-Si:H development
0.1 1 10 100 10006
7
8
9
10
11
12
effic
ienc
y (%
)
degradation time (h)
a-Si:H/µc-Si:H module µc-Si:H module
10x10 cm2 substrate(aperture area 64 cm2)
NREL confirmed: 10.1 %
NREL confirmed: 8.1 %
BR et al. TSF 2006
source: FZ-Jülich
31Bernd Rech, WIAS 2008
SiH4+H2
HF
HF-Electrode
Substrate
Silicon growth by PECVD
(plasma enhanced chemical vapour deposition)
process parameters: amorphous ⇔ µ-crystalline growth
� SiH4 / H2-gas flow ratio� Pressure, � RF power, � Excitation Frequency� Substrate temperature� ...
source: FZ-Jülich
32Bernd Rech, WIAS 2008
L. Houben, Dissertation, FZJ (IFF/IPV), Uni Düsseldorf O. Vetterl et al., Sol. Energ. Mat. Sol. Cells 62 (2000) 97-108
microcrystalline amorphous
Structure of µc-Si:H
33Bernd Rech, WIAS 2008
very narrow process window for optimised µc-SI:H
4
6
8
4 5 6 7 8 9 10400
500
600
700
4 5 6 7 8 9 1010
15
20
0,6 0,8 1,0
50
60
70
0,6 0,8 1,0
a-Si
a-Si
a-Si
a-Si
µc-Si
η (%
)
µc-Si
µc-Si
µc-Si
VO
C (
mV
)
deposition pressure (Torr) deposition pressure (Torr)
JS
C (mA
/cm2)
FF
(%)
[SiH4]/[SiH4+H2] (%) [SiH4]/[SiH4+H2] (%)
µc-Si:H Solar Cell Process: 13.56 MHz Regime
BR et al, TSF 2003
34Bernd Rech, WIAS 2008
Rotational temperature
SiH4 dissociation
Optical Emission Spectroscopy (OES)
35Bernd Rech, WIAS 2008
µc-Si:H – control of film growth
-10 0 10 20 30 40 50 60
140
160
180
200
Sub
stra
te te
mpe
ratu
re (
°C)
time (min)
200 W, 11 Å/s 60 W, 5 Å/s
Plasma-induced substrate heating
-10 0 10 20 30 40 50 600,5
1,0
1,5
Em
issi
on in
tens
ity (
a.u.
)time (s)
SiH* emission (414 nm) H* emission (656 nm)
Amorphous growth conditions @plasma start
plasmaoff
plasmaon
M.N. van den Donker et al., TSF 2006
36Bernd Rech, WIAS 2008
HF
x (cm) 1
Pot
enzi
alU
(V
)
Si-Schicht
• high power⇒ high plasma density (ne)
⇒ high growth rate
• high pressure⇒ low plasma potential⇒ low ion bombardment
“soft“ deposition
very narrow process windowand drift of the plasma properties
during growth*!⇒⇒⇒⇒
up-scaling extremely challenging !
e-
e-
e-
e-e-
e-
e-
SiH4+H2
e-
e-
e-
+
Summary: µc-Si:H Process Conditions
* see PL0001 van de Sanden et al.BR et al, TSF 2006
37Bernd Rech, WIAS 2008
Starting point
10 x 10 cm2
30 x 30 cm2
>1 m2
Goal:cost-effective production technologyfor highly efficient solar cells
a-Si:H/µc-Si:H: From laboratory towards production
Development@IPV, FZ-Jülich
Joint development:Applied Materials, Sontor, FAP, FZJ
38Bernd Rech, WIAS 2008
Successful Scale-Up @ Sontor
Module size : 1.8 m2
Status: ~7.5 % (stab. tot. area) production average
39Bernd Rech, WIAS 2008
silver
TCO
superstrate
TCO
a-Si pin????c-Si pin
1-4 mm
700 nm
)1-
1-3 µm
100 nm
silver
TCO
superstrate
TCO
a-Si pin????c-Si pin
Light
silver
TCO
Glas
TCO
µc-Si pin
1-4 mm
700
)1
100 nm
200-300 nm
APCVD SnO2 - Asahi U
sputtered & etched ZnO
0.5 µm
„Light-Trapping“ by Surface-Textured ZnO
source: FZ-Jülich
40Bernd Rech, WIAS 2008
Properties of ZnO:Al
• highly transparent & conductive
• c-axis oriented
• resistant against H2-plasmas
• smooth surface
• surface-texture by etching(depends on initial film properties!)
wurtzite-structure
c-axis
oxygen
zinc
Sputter techniques:
• rf/dc ceramic targets• mf metallic targets, high rate
Ar+RF ZnO
target: ZnO:Al 2O3
glass substrate
Sputtered ZnO:Al Films as Front TCO
41Bernd Rech, WIAS 2008
1µm
smooth short dip optimised texture
• δrms up to 150 nm, high transparency and conductivity• crater size 100-2000nm
• cristallite size δ < 50 nm
• phenomenologic model(Kluth et al. TSF, 2003)
glass
ZnO
HCl 0.5 %
120°
c-axis
Schematic cross section
microscopicmodel?
Optimised ZnO for Si Thin-Film Solar Cells
42Bernd Rech, WIAS 2008
Tailor-made surface roughness and surface features ⇒⇒⇒⇒
• AR-effect by index-matching• efficient light trapping for long wavelength light • low free carrier absorption
Licht
ZnO
Glas
µc-Si:H:1 µm
ZnO
Ag
“Light Trapping” in µc-Si:H Solarzellen
B R et al. TSF, 2006
400 600 800 10000,0
0,2
0,4
0,6
0,8
1,0
wavelength (nm)
qua
ntum
effi
cien
cy
2 µm low Al content 1 µm low Al content 1 µm rf standard 1 µm rf smooth
15,623,0
24,3
26,8
mA/cm2
43Bernd Rech, WIAS 2008
Rsheet after etching:
Large Area Surface-Textured ZnO
Sontor / Applied Materials
44Bernd Rech, WIAS 2008
Outline
� Motivation and Background� Thin Film Solar Cell Technologies and Applications� Amorphous Silicon Based Solar Cells� Microcrystalline Silicon and Tandem Cells� Poly-Crystalline Si Thin-Films� R&D Challenges and Conclusions
45Bernd Rech, WIAS 2008
Towards High Efficiency Thin Film Si Solar Cells
today mid term very long term
efficiency
long term
20 %
15 %
Wide-gap Si(quantum size effects)
High EfficiencyTandem Cells
„nano“-scaledlocal contacts!
Materials Research +Materials Research +Materials Research +Materials Research +ProcessProcessProcessProcess DevelopmentDevelopmentDevelopmentDevelopment ++++DeviceDeviceDeviceDevice DevelopmentDevelopmentDevelopmentDevelopmentin parallel !in parallel !in parallel !in parallel !
„„„„RoadmapRoadmapRoadmapRoadmap““““
increase in grain size
„perfect“ passivationlight trapping concept“
„high growth rates“
46Bernd Rech, WIAS 2008
a-Si:H (n, p)
a-Si:H (p, n)
1 cm²34.9 mA/cm²629 mV17.4 %Pyramidsa-Si(n)/c-Si(p)/a-Si(p)
1 cm²39.3 mA/cm²639 mV19.8 %Pyramidsa-Si(p)/c-Si(n)/a-Si(n)
0 100 200 300 400 500 600 700
5
10
15
20
25
30
35
40
a-Si:H(p)/c-Si(n)
a-Si:H(n)/c-Si(p)
Jsc(mA/cm²) 39.26 34.9 Voc(mV) 639.4 629FF(%) 78.9 79ηηηη(%) 19.8 17.4
|J| (
mA
/cm
2 )
VOC
(mV)
Independently confirmed at ISE Freiburg
Rear contact
TCO: 80nm
a-Si:H(n, p) ~ 5nm
c-Si(p, n)
FZ-Si
Front contact
a-Si:H(p, n) ~ 35nm
Note: Record efficiencies of these cell type > 22 % by Sanyo
M. Schmidt et al. TSF 2007
a-Si:H / c-Si Wafer Based Cells
47Bernd Rech, WIAS 2008
single crystalline
Thin Film Approaches
grain size g (µm)
10-2 10-1 100 101 102 103 104
open
circ
uit v
olta
ge V
oc
350
400
450
500
550
600
650
700
750
∞
Neuchatel
pn
(mV
)
Sanyo
Kaneka*
BPTonen
ETL
ASE/ISFH
Daido
Astropower
ISE
Mitsu-bishi
Sony
ISEMPI
UNSW
ipe
IMEC
IMEC
ISI
pin
PHASE
ZAE
Canon
CanonETL
10-2 10-1 100 101 102 103 104
10-1
100
101
102
N
RST
Q
U
P
O
KM
L J
I
H
GF
E
CD
B A
limit Leff,mono
=102 µm
103
105
107
SGB
= 101 cm/s
diffu
sion
leng
th L
eff,p
oly [
µm
]
grain size g [µm]
well passivated grain bounderies
Recombination @ grain boundaries
R. Bergmann, J.H. Werner, TSF 202-204 (2002) 162K. Taretto, U. Rau, J.H. Werner, J. Appl. Phys. 93 (2003)
IPV
48Bernd Rech, WIAS 2008
glasspoly-Si
Al
Epitaxial growthof absorber layer
by high rate deposition(e.g. E-Beam evaporation)
Growth rate >1.2µm/h
Glaspoly-Si
poly-Si
solar cellprocessing
Glasp+
p
TCOa-Si:H, n +
Glas Glaspoly-Si
Al
Seed layer conceptAluminium-induced layer exchange
(ALILE)
Ala-Si
Alternative Pathes: �solid phase crystallisation(first product by CSG Solar)�laser crystallisation�E-beam crystallisation
Poly Si thin film @ HZB
49Bernd Rech, WIAS 2008
Outline
� Motivation and Background� Thin Film Solar Cell Technologies and Applications� Amorphous Silicon and Microcrystalline BasedSilicon and Tandem Cells� Poly-Crystalline Si Thin-Films� R&D Challenges and Conclusions
50Bernd Rech, WIAS 2008
Si
SiH
Fundamental Materialproperties
Solar cellsand prototypes
Large area industrialprocessing
Technology Value Chain
In Thin-Film PV
Application in systems
Solar module development:From materials via devices towardscomplete systems
…and „vice versa“problems from application⇒ applied and basicR&D
51Bernd Rech, WIAS 2008
� Large area PECVD (high rate, process control)
� Alternative techniques forabsorber deposition
� Quantitative understanding of materials interfaces and device
� improved/new materials (e.g. μμμμc-SiGe,SiC,…)
Goals for applied and fundamental R&Dprerequisites to realise the production roadmap !
� New deposition reactor concepts (very high growth rates, full gas usage)
� Incorporate quantum or spectrum-converting effects
� Combine thin-film Si with other PV technology
�Understand fundamental limitations of thin-film Si
hot topics
source: Strategic Research Agenda for PV (PV Technology Platform)
Concept forηηηη > 15 %
Demonstrateη > 12%
Prototype/test modules
2013-20202008-2013
52Bernd Rech, WIAS 2008
� Thin-film solar modules will become major PV techno logies within the next decade. ⇒⇒⇒⇒ The proof of concept for a variety of technologies exists.
� The transfer of lab developments / prototypes into a cos teffective production is the challenge today.
� There is a strong need for broad R&D to improve existingconcepts and develop new thin-film technologies to openthe path for higher efficiencies and lower production cost s.
Conclusions
53Bernd Rech, WIAS 2008
Scenario for the world´s primary energy mix in 2100
54Bernd Rech, WIAS 2008
„ Key Product for the Bavarian Marketflexible thin-film solar cells for cold beer“