R&D R&D sul sul fotovoltaicofotovoltaico in STMin …leos.unipv.it/slides/lecture/Foti-stm.pdf•...
Transcript of R&D R&D sul sul fotovoltaicofotovoltaico in STMin …leos.unipv.it/slides/lecture/Foti-stm.pdf•...
R&D R&D sulsul fotovoltaicofotovoltaico in STMin STMMarina FotiIMS R&DSTMicroelctronicsSTMicroelctronics
Convegno su
Tecnologie, tecniche impiantistiche e mercato del fotovoltaico
15 Ottobre 2012
Mondello (PA)
Outline
• Thin film module technology
• Amorphous silicon (a-Si:H) and microcrystalline (µc-Si)
• Tandem and multiple junction solar cells
• Enhancement of light absorption in thin film Si
• Development of TCO front and back electrodes
• Next steps on light trapping
• Thin film silicon outlook
• TF PV flexible application for smart systems
A New PV Joint Venture: 3SUNA New PV Joint Venture: 3SUN
ENEL GREEN POWER, ENEL Group Company,
dedicated to the development and
management of activities related to energy
production from renewable sources at an
international level, which operates in Europe
and the American Continent. It is a leading
Company in this sector at global level.
SHARP CORPORATION, a Japanese
Company, which operates at global level in
the manufacturing and distribution of
consumer products (LCD TV, LED TV, ecc). A
leading company at global level in the
photovoltaic sector (Solar Cells, and
Electronic Devices).
STMICROELECTRONICS, is one of the
largest manufacturers of semiconductors
in the world with customers in all
electronics segments. The Corporate
headquarter is in Geneva, advanced
research and development centers in 10
countries, 14 main manufacturing sites
and sales offices all around the world.
3SUN 3SUN –– Thin Film MultiThin Film Multi--Junction Modules Junction Modules FabFab
Numbers:
• 240 000 m2 surface area
• 60 000 m2 Fab area
• 300 employees
• 160 MW/y (2011)
•240MW/y, …possible extension
The biggest PV Italian fab destined to
compete with the most important players
of the sector
Thin film multi-junctions modules are
manufactured in the innovative plant M6
built in Catania
Large area modules: 1m × 1.4 m
Why TF Solar Cells?Solar cell Si raw material Efficiency Peak power Peak power
c-Si 1200-1300 g/m2 16% 160W/m2 0.13W/g
TF-Si 5 g/m2 10% 100W/m2 20W/g
Large area
multi-junction / glass
Amorphous or tandem / flex
Amorphous / glass
Technology options: thin film Technology options: thin film vsvs Si wafersSi wafers
BULK Si SOLAR CELLS
Series connection of individual solar
cells
Mature technology but needs
a lot of Si
THIN FILMS
Monolithic integration
(series connection by lasering)
CVD on very large areas
Potential for ultra low costs
Processing of wafers
Large area modules on glassLarge area modules on glass
Altomonte (CS - Italy): 8,2MW. 11 Millions of kWh. It can satisfy the needs of 4.000 families
Other than Ground PV Plants…Other than Ground PV Plants…
Parking area Roof
Easy installation and no particular
maintenance or cleaning required. No
specific accurate angle to the sun. Perfect
integration with the environment.
Residential, Commercial and Industrial
Roof
Various Applications Innovative Future Solutions
Roof
Installation following the roof profile and
good performance at any slope of the
roof. Nice appearance integrated to the
building design.
Deserts and hot climate Countries
Supplying high performance even at
50~60°C thanks to the low temperature
coefficient (-0,24%/°C). Good performance
even when the panels would be partially
covered by dust and sand thanks to the
feature to produce energy with diffuse
light.
Integration on Building Design
Building front designed with Glass/Glass
Frameless Thin-Film PV Modules
Car, Truck and Trailer PV
Roof
Stand-alone Applications
powered by PV panels
E.g. Water sweetening kit
Thin film PV on flexible substrate
Substrate and superstrate configurations
BC TCOBack electrode (Ag, Al, white pvb)
Opaque sealing
Front TCO
a-Si:H, uc-Si:H or multiJ
Transparent sealing
Thin film deposition at low temperatures on large area substrates
glassFront TCO
a-Si:H, uc-Si:H or multiJ
BC TCO
Metal, plastic..
a-Si:H, uc-Si:H or multiJ
Back electrodeBC TCO
Superstrate Substrate
11Glass with
TCO
Edge seaming
Cleaning
Laser scribe P1
TCO
Laser edge
deletion
Bus bars and
wires
connection
Lamination
In line solar
simulator (IV)
Laser scribe
Isolation P4
•The modules are fabricatedmonolithically on a glasssubstrate during front end process
•The back end is dedicated toadd electrical connection,
Thin Film Module process flow Thin Film Module process flow
PECVD
deposition
Laser Scribe cell
P2
PVD deposition
Back contact
Laser scribe
back contact P3
Lamination
with PVB and
back glass
J-box
connection
2nd in line solar
simulator (IV)
Packaging
add electrical connection, protection layers, frame and junction box
•Typical process flow tandem modules
•1 x 1.4 m2
ThinThin film on film on glassglass: FEOL : FEOL processprocess Glass with TCO Layer
Cleaning
Laser Scribe P1
PECVD Depositiona-Si:H -pin
SOIR
µc-Si:H - pin
Laser ScribeP2
TCO Deposition
Laser ScribeBack Contact
P3
Cleaning
Scheme of thin film moduleScheme of thin film module
load
- +
TCO
glass
cell
back contact
PECVD
High
automation
Glass Size Matters for Thin Film
PECVD
deposition
TCO
depositionlarge area and high
throughput is needed to
achieve low cost/Wp
TF Silicon Costs breakdown
3%
5%6%
15%6%14%
5%3%
46%
TCO Gas/Chem Target Back glass Encapsulant Terminal Box Silver Paste/ Bus Bar/ Packing/Other Lead Wire / MultiFrame J
Amorphous siliconAmorphous Si: a-Si:H layers were first deposited by R. Chittick
(1969) experimenting with SiH4 in a plasma reactor.
First systematic study by Spear et al Phil Magaz, 33, 935 (1976)
Tetrahedrally
bonded
c-Si structure
Amorphous Si:
absence of
Long range order
Distribution of density of allowed energystates for electrons
17
due to the disorder direct
optical transitions are not
forbidden in amorphous Si
Eg = ~1.8 eV
better light absorption
than c-Si
Amorphous Si for thin film PV• Deposited by plasma-enhanced CVD of SiH4 at 150-300°C. Low gas utilization (10-30%). Heavilyhydrogenated 1-10 at.% H.
• PN (PIN) junctions formed through boron or phosphorous containing gases.
• Total thicknesses in some cases below 1 µm (100 timesthinner than c-Si).
• Multiple junction devices with two or three junctionsgrown one upon the other and current matched.
The BIG three challenges
• Improve efficiency from 6-8% up to 12-15%;
• Minimize or eliminate the self-limited degradation
• Increase deposition rate
StaeblerStaebler--WronskiWronski EffectEffect
L
• Exposure to light induced degradation, which stabilizes with time
• Typically after 1000h of continuous light soaking at 1 Sun AM1.5G
• New dangling bonds (from 1e15 to 1E17 cm-3) are created under light exposure
• Degradation is recovered after annealing at T<150C
Typically 10-13 % of degradation
For a-Si:H of 150-300nm
Limitation on the thickness
annealing at T<150C
Amorphous a-Si:H: p-i-n
Drift charge transport – p-i-n junction
ip n
20
• Carriers are photogenerated in the
intrinsic region and collected by drift
Amorphous and Microcrystalline silicon
Two materials with
the same process
21
the same process
PECVD
a-Si:H Eg=1.8eV
µµµµc-Si:H Eg=1.1 eVColumnar microstructure
Amorphous
Eg=1.8eV
«High» absorption
in the green-blue
Microcrystalline
Eg=1.1eV
«High» absorption in
the red-near I.R.
Enhanced absorption: double junction/tandemEnhanced absorption: double junction/tandemLi
ght i
nten
sity
(kW
/m2 µ
m) “spectrum splitting.”
Wavelength (nm)
Ligh
t int
ensi
ty (
kW/m
Micromorph cell efficiency 11-14%
Micromorph module efficiency 8.5-10.8%
Tandem configuration: Top Tandem configuration: Top aa--Si:HSi:H, Bottom , Bottom µµµµµµµµcc--Si:HSi:H
TCO
0.40
0.50
0.60
0.70
0.80
0.90
1.00
EX
TE
RN
AL Q
UA
NT
UM
EF
FIC
IEN
CY
a-Si:H
23
Multiple junction devices with two junctions grown one upon the other
and current matched spectrum splitting enables higher absorption
and higher efficiency
0.00
0.10
0.20
0.30
250 350 450 550 650 750 850 950 1050 1150
EX
TE
RN
AL Q
UA
NT
UM
EF
FIC
IEN
CY
Wavelength (nm)
a-Si:H
µµµµc-Si:H
FromFrom Single Single toto Multiple Multiple junctionsjunctions
glass
• Single Junction–aSi:H cell with enhanced light trapping – TCO and Texturing
• Double Junction / Tandem cell–highest efficiency: combination of absorber materials having band gap 1.8 eV and 1.1 eV for the top and bottom cell..
• Triple junction–aSiGe:H middle absorber more than 12% on largeareas
• Best stabilized efficiencies above 12%
glass
textured TCO
a-Si:H top absorber
a-SiGe:Hmiddle absorber
µµµµc-Si:H bottom absorber
ZnOAg
• Higher efficiencies (from 12 to 20%) are possible with additional junctions
• But so far :
•Reduced throughput:~ 30% lower for triple Junction
•Costs ~ 20% higher than tandem
•Despite the lower efficiency of tandem technology higher throughput in MW/years
Issues limiting a-Si:H and µc-Si:Hefficiencies
• a-Si:H : Voc too low 0.9V instead of 1.4V (bandtails contacts)
25
• µc-Si:H: Low Jsc. Improve absorption, light trapping
Light trapping increases the absorption because increases the optical thickness
Light can be captured in the desired parts of a solar cell (absorber layers)
and can be confined in it.
26
The cell current can be enhanced by increasing the effective
optical path in the absorber layer (a-Si:H or µµµµcSi:H)
300nm a-Si:H
M. Zeman, J ELECTRICAL ENG, VOL. 61, NO. 5, 2010, 271–276
Light trapping
p-i-n a-Si:H
p-i-n uc-Si:H
TCO
glass
light~700nm
~250nm
~1.6µm
Asahi VU (SnO2:F) Asahi W
ZnO:B -MOCVD W text ZnOp-i-n uc-Si:H
TCOBack reflector
~1.6µm
~50nm
• Improvement: about 50 % reduction of the deposition time
• (today limiting process step).
Figures of merit for TCO 28
• A good TCO must have a high figure of merit (conductivity/absorbtion coefficient)
• Rs is the sheet resistivity
• R is the reflectance
R.G. Gordon MRS Bulletin 2000
BAND GAP BAND GAP Engineering: Engineering: Impact of work function on solar cell conductivityImpact of work function on solar cell conductivity
TCO / aSi:H (n doped)
barrier
TCO / aSi:H (p doped)
barrier
electrons rich area
barrier
TCO objective: WF < 4,3eV
Hole rich area
barrier
TCO objective: WF > 5eV
Haze impact
• Apart from the transmittance and the low sheet resistance (~7-10 Ω/) a TCO must have:
• reduced reflection due to refractive index grading (1<n<3); this effect applies to
100
Index grading at the
TCO/p interface
(whole spectral range)
Light trapping and index
Grading at the back reflector
(red spectral range)
λλλλ/n
H=Tdiffused/Ttotal
(1<n<3); this effect applies to the whole wavelength range of the spectral response
• light scattering and subsequent trapping in the silicon absorber; this applies more to the weakly absorbed light that penetrates up to the back contact400 500 600 700 800
0
20
40
60
80
Cel
l ref
lect
ivity
(%
)
Wavelength (nm)
low haze
high haze
Impact of texturing and TCO material
low haze TCO high haze TCO
Haze: ratio between diffusely scattered and total
intensity
0
10
20
30
40
50
60
70
80
90
100
200 300 400 500 600 700 800 900 1000 1100
Tran
smitt
ance
(%)
Wavelength (nm)
UV-typeANX10OE_B_TDOE_C_TDANX10
Total T
Diffused T
SnO2:F
ZnO:B
TCO
a-Si:H
µµµµc-Si:H
BR
• Texturing (increases optical path) can improve the currents generated in the top and bottom cells
intensity
• θ1 direction incident beam
• θ2 direction scattering beam
• exponent: 2<β<3
−−−==β
ϑϑλπ
2211 coscos2
exp1 nnT
TH
total
diffused
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
200 300 400 500 600 700 800 900 1000 1100 1200
EX
TE
RN
AL
QU
AN
TU
M E
FF
ICIE
NC
Y
Wavelength (nm)
ZnO - H=20%
ZnO - H=20%
ZnO - H=20%
SnO2-H=10%
SnO2-H=10%
SnO2-H=10%
a-Si:H
µc-Si:H
SnO2:F
ZnO:B
Impact of TCO on the cell performances
• (1) ZnO:B 20% Haze higher Jsc
• (2) SnO2:F 10% Haze
• Φ < Φ
32
8
10
12
14
16
Cu
rre
nt D
en
sis
ty
(m
A/cm
2)
η=11.5%
η=12.5%(1)
(2)
• ΦZnO< ΦSnO2
• Difference of Workfunctions
differences in the Voc
0
2
4
6
8
0 0.5 1 1.5
Cu
rre
nt D
en
sis
ty
(m
A/cm
2)
Voltage (V)
Increase the efficiency: intermediate reflecting layer
• Currents (Jsc) matching (top cell –bottom cell
• IRL refractive index between 1 and 3
• Filters low energy photons
• Reflects high energy photons
• SOIR is obtained in the same PECVD chamber used for a-Si and µµµµc-Si
silicon oxide based
intermediate
reflector layer
(SOIR)
A. Feltrin et al, MRS 2009
Texturing of ZnO on front and on backside• Use of LPCVD or MOCVD ZnO
controlling texturing on front and backside
• Increased light path in a-Si and µc-Si
• Reduced absorber thickness of~ 50%
• Increased efficiency
Front TCO
Glass
Front TCO
Back TCO
a-Si:H
µc-Si:H
White pvb
EQ
E
Wavelength (nm)
Thin cell /Effect of white sheet reflection on microcrystalline
6.00E-01
7.00E-01
8.00E-01
9.00E-01
1.00E+00TOP (ZnO no white paper)
BOTTOM (ZnO no white paper)
SUM (ZnO no white paper)
TOP(ZnO with white paper)
BOTTOM(ZnO with white
White polymer can be used instead of expensive Ag for reflection
Textured thick TCO as back contact
Glass
White PVB contributes significantly to the reflection especially in the bottom
cell (µc-Si:H)
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
250 350 450 550 650 750 850 950 1050 1150
EQ
E (
%)
wavelength (nm)
BOTTOM(ZnO with white
paper)
SUM(ZnO with white paper)
a-Si:H
µc-Si:H
White pvb
Back TCO
3D Structures3D TCO
• 3D architectures by using TCO 3D patterning
– Higher efficiency 3D structures obtained by using TCO 3D
templates
– To increase light trapping and orthogonalize light
absorption and photocarrier collection
W. Soppe eta al 26th PVSEC 2011
• Planar waveguides with disordered pores to enhance the absorption of the light
(Anderson localization effects )
Riboli et al Optics Letter, 36, 127, 2011
Plasmonic enhancement effects by metal layers and nanoparticles
Plasmonics
Waveguide modesScattering Near-field enhancement
37
Waveguide modesScattering Near-field enhancement
V.E. Ferry, et.al., APL 95 183503 (2009)
• Developing cell architectures with silicon wires
in order to orthogonalize light absorption and photocarrier collection
Silicon wires and quantum dots
p-type
n-type
Traditional planar, single junction solar
cell
~L
ħω1/α
Idealized radial junctionwire solar cell
~L
1/α
• Quantum dot based heterojunction solar cells
38
cell
B.M. Kayes, et.al., J. APPL PHYS 97 114302 (2005).
Substrate chuck in atmospheric pressure
Liquidinjection head
Substrate spin chuck
host precursorc:Si synthesis
+
Colloidalnanocrystals
TCO precursor
Substrate chuck in atmospheric pressure
Liquidinjection head
Substrate spin chuck
host precursorc:Si synthesis
+
Colloidalnanocrystals
TCO precursor
Substrate chuck in atmospheric pressure
Liquidinjection head
Substrate chuck in atmospheric pressure
Liquidinjection head
Substrate spin chuckSubstrate spin chuck
host precursorc:Si synthesis
+
Colloidalnanocrystals
TCO precursor
host precursorc:Si synthesis
+
Colloidalnanocrystals
TCO precursor
Applied on Si Thin Film for efficiency > 20%
Si PV Thin film outlookThin film technology addresses low cost/Wp by using large area high throughput (e.g. PECVD with high dep, rate low T)e quipments with very high level of automation
To achieve the target material costs, especially th e front glass with TCO, need to be low .
Improving light trapping is fundamental to increase the efficiency (12% on single junction) or reduce costs because lower Si a bsorber thickness is necessarynecessary
Multiple-junctions solar cells are necessary to inc rease the efficiency but to date they are still characterized by low throughp ut and higher costs
Silicon TF solar cells are expected to achieve a mu ch higher conversion efficiency (up to ~20%) than other TF technologies (CdTe, CIGS,..), which today are strong rival to Si, by exploiting new mat erials and by applying multi-junction structures.
Bring together the experience and know-how of resea rchers in applied physics to speed up the development of materials an d devices.
Electronic device integrated energy harvesting with flexible thin film PV
Harvesting in Smart Systems An example in Wireless Sensor Nodes for Automation
Harvesting Device
(PV, Piezo, etc)Low Power RF
Transceiver
Sensors
Ultra Low PowerEnergy Conversion
Integrating Harvesting in Smart
Systems
Energy
Autonomous Wireless Sensor Node
Ultra Low Power
Microcontroller
Energy Conversion
Battery Storage
Enabling wireless sensors for energy
autonomy
Harvesting system with flexible foils
Platform features• PV module collects energy from indoor light
(300lux minimum)
• Harvested energy stored in an micro-
battery (ST-TF)
• Managing of the system energy
• Sensing ambient temperature
• Powering an STM8L15 microprocessor
• Supplying an RF transceiver
• Processed data transmitted to a BST
Flexible PV Modules Modules of 30 cm2
Thin film solar cells are monolithically series
connected (13 cells) 250µW
Back contact
pin a-Si:H
TCO
43
Back contact
Polymeric substrate
4 patents of ST on the subject
13 cells module @ 300 lux F12
1.5E-04
2.0E-04
2.5E-04
3.0E-05
4.0E-05
5.0E-05
Po
we
r (W
)
Cu
rre
nt
(A)
Efficiency @ 300 Lux: 8% - 9%
0.00E+00
5.00E+02
1.00E+03
1.50E+03
2.00E+03
2.50E+03
3.00E+03
AM1.5G
F12
Sp
ect
ral
de
nsi
ty (
a.u
.)
0.0E+00
5.0E-05
1.0E-04
1.5E-04
0.0E+00
1.0E-05
2.0E-05
3.0E-05
0.0 2.0 4.0 6.0 8.0 10.0
Po
we
r (W
)
Cu
rre
nt
(A)
Voltage (V)
Jsc (µA) 43.60
Voc (V) 8.42
Pmax (µW) 238.52
eff (%) 7.98
Fluorescence lamp
spectrum
300 lux ~ 1W/m2
0.00E+00
300 800 1300 1800
wavelenght (nm)
Implementation: contact layer
Sequence SnO2:F/p-type a-Si:H/Mo : to study the interface between the contact layers and the p-type a-Si:H
Plays an important role on the PV cell performances
By C-V and I-V data coupled with modeling we find that the Mo provides a better Schottky contact
(a)
provides a better Schottky contact with p-type a-Si:H compared to SnO2:F.
(b)
M. Foti et al, ECS 2011
G. Cannella et al. JAP 2011
Strong synergy with CNR-IMM (S. Lombardo)
0
5
10
15
max power
Voc
Isc
var
iatio
n (%
)
Flex PV: Benchmark with competitors @ indoor light
Efficiency comparison at indoor
• ST Flex Module 30cm2,
• high robustness,
• less leakage
PV module is very robust
Mechanical stress test: Module is bent with very small
-15
-10
-5
0
0 200 400 600 800
Number of bending
var
iatio
n (%
)
46
3.9cm2.9cm1.9cm1.5cm1.2cm0.95cm
Mechanical stress test: Module is bent with very small
radius, r = 2cm
No significant changes in the electrical characteristics
(Voc, Isc, Max Power)
Roll to roll
polyethylene-naphtalate (PEN)
Roll to roll technology
Si TF development at low deposition T
from 150C to RT using new deposition techniques IC PECVD