Photovoltaics F.-J. Haug Ecole Polytechnique Fédérale de Lausanne (EPFL), Institute of Microengineering (IMT), Photovoltaics and Thin Film Electronics.

Photovoltaics

F.-J. HaugEcole Polytechnique Fédérale de Lausanne (EPFL), Institute of Microengineering (IMT),

Photovoltaics and Thin Film Electronics Laboratory (PV-Lab),

2000 Neuchâtel, Switzerland.

F.-J. Haug – Photovoltaics 2

Outline

• Why Solar Cells?

• How do they work?

• Why thin film silicon solar cells?

• What do we do at PV-Lab?– thin film solar cells on glass and plastic substrates– thin films for wafer-based cells (hetero-junction cells)– transparent conducting oxides– (sensor applications)– (module design and reliability)


The power source: Our Sun

• A low to medium size star

• Heated by fusion(essentially , processes for , etc.)

• Power output: (convert to energy)

• Delivers to the earth

• 10’000 the global energy consumption


Solar irradiation atlas

SolarGIS

Central Europe: ca 1 kW/m2 under clear sky at noonYearly average: 1000 kWh/m2, ca 3 hours every day


The solar spectrum

See www.pvlighthouse.com.au

The solar spectrum resembles white-hot body glowing at T=5700 K• Usually measured with respect to wavelength (use a grating or a prism)• In PV: often necessary to convert to photon energy:


Filtering by atmosphere: air mass (AM)

AM1.5: central Europe

AM2: 30° above horizon

AM3: 20° above horizon

Air mass number: , latitudePV standard: => , central Europe, Canada/US border

AM0: space

AM1: equator

400 600 800 10000.0

0.5

1.0

1.5

2.0 AM0 (data) AM1 AM1.5 (data) AM2 AM3

Spect

ral i

rradia

ton d

ensi

ty (W

/m2nm

)

Wavelength (nm)

WaterOzone

1000 W/m2

integrated: 1366 W/m2


Is 1 kW/m2 a low energy density?

Example Germany: 30% of electricity supply by lignite (brown coal)exploitation in open strip mines

some data for Hambach (Germany): - lower ground water table by 400 m- dislocate 3 villages (2 more planned)- 85 km2 open mine and waste-deposit - 40 Mio tons of coal per year

~7GW electricity

Formation of lignite: 25 Mio y exploitation: 1984-2050

Perspective: solar illumination on 80 km2: 80 GWcover with solar cells of 10% efficiency => 8 GWp (alas, not continuously)


Other “conventional” energy sources

Less obvious, but still impact on large area: dropping ground over coal shafts, shale-gas fracking, deep-sea oil drilling, etc.

MTR (mountain top removal) for Appalachian coal (USA)

Strip mining of tar sands in Alberta (Canada)


Solar electricity (photovoltaics)

Roof top installation (Kaneka a-Si modules)Teplin Airstrip, Germany: 130 MW on 2 km2 installed in 4 months

c.f. Centrale de Gd Dixence: 2 GW, 8 km2 lakes + catchment area

Suitable roof top area in Switzerland: 138 km2 (residential, commercial, etc.)

(IEA)


What is inside a solar cell?

thin film silicon solar cell p-i-n junction

glass

reflecting back electroden-layer

semiconductorabsorber layer

p-layer

transparent front contact

eh

glass

encapsulation

glass

metallic back contact

p-type wafer

n-contact

encapsulation

eh

glass

encapsulation

silicon wafer solar cell p-n junction


What is inside a solar cell?

active part: semiconducting silicon (thin film)(also used: CdTe, CIGS etc.)

glass

glass

encapsulation

glass

n-contact

encapsulation

glass

encapsulation

active part: semiconducting silicon (crystalline)

2-3 μm200-300 μm


What is a semiconductor?

• It’s conducting, but not as good as metals

• Electric conductivity is normally associated with the flow of electrons in metals

• Conductivity in semiconductors is different

• The flow of water can serve as analogue


Conductivity in metals

Flow of electrons similar to flow of waterHorizontal surface, no (very small) potential difference

Analogue:a pond of water


Electronic conductivity in a semiconductor

Winter: the pond is covered with ice


Electronic conductivity in a semiconductor


Flow of electrons similar to flow of water (like metals)Difference: Slower movement, needs potential difference



Hole conductivity in a semiconductor

Imagine: blow air bubbles under the ice sheet

A new phenomenon, exclusive to semiconductors !Bubbles under the barrier move upwardsIn semiconductors: positive charge carriers called holes


Why semiconductors for solar cells?

• Metals normally reflect light; semiconductors can absorb it

• Absorbed light creates pairs of electrons and holes (water droplets above and bubbles below the ice sheet)

• An electric field between doped regions separates electron-hole pairs (imagine you inclined the ice sheet)

e-

h+


Absorption in semiconductors: bandgap

+

-

-

+

High bandgap:high el. potentiallow current

Low bandgap:low el. potentialhigh current

Photons with energy less than the band gap are not absorbedPhoton energy in excess of band gap is lost to thermalization

𝜂=𝑉 𝑜𝑐 ⋅ 𝑗𝑠𝑐 ⋅ 𝐹𝐹⏟¿ 0.7 ...0.8


Limiting efficiency

0.5 1.0 1.5 2.0 2.50

10

20

30

40

CdTe

CuGaSe2

a-Si

CuInSe2

GaAs

Ge

Si

Effi

cien

cy (

%)

Band gap energy (eV)

certified records: 28.8%: crystalline GaAs (Alta Devices)25.6%: crystalline silicon (Panasonic)20.5%: thin film Cu(In.Ga)Se2, (Solibro)

Theoretical limit of ~33%, optimum bandgap 1.44 eV


Better use of incident light: tandem cells

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

Top cell bandgap (eV)

Bot

tom

cel

l ban

dgap

(eV

)

0510152025303540

4 Terminal (no constraint on current matching)

Theoretical limit of ~43%, optimum bandgaps 1.1 and 1.7 eV

certified records (triples): 37.9: crystalline (In,Ga)P/GaAs/(In,Ga)As (Sharp)13.4%: thin film a-Si/nc-Si/nc-Si (LG)

high gap

low gap

glue


Wafers or thin films?Crystalline cells: size determined by wafer

e.g. c-Si:single-crystalline: max 30cm diametermulti-crystalline: typically cut to 15x15 cm2

Thin film modules: size determined by glass and machinery

e.g. thin film silicon tandem modules5.3 m2 (Applied Materials machinery)1.4 m2 (TEL Solar (ex Oerlikon) machinery)

high efficiency – high cost moderate efficiency – low cost


Wafers or thin films?Crystalline material: perfect propertiesexcellent lifetime and collection

Polycrystalline or amorphous materials: defects and grain boundariesrecombination and short lifetime

Electric field: Electric field:

e-

h+

e- e-

h+ h+

e-

h+


Module interconnection

Wafer based: solder cells into strings(e.g. SunPower: 22.4%)

Thin film: 3 x laser scribing during process(e.g. TEL Solar: 12.2%)

SunPower

SaFlex/TEL Solar

Rowell, En. Env. Sci


Thin film advantage

Semi-transparent a-Si modules (Schott)

Colorful TF-Si modules(PV-Lab prototypes)


Thin film advantage

Roof tile laminate(Bressler)

Flexible a-Si modules(Flexcell)

(UniSolar)


Thin film dilemma

PV modules are sold by power, currently:

~0.8 $/W crystalline

~0.6 $/W thin film

3.5$/W in 3.5 y

PV system cost (currently ~5 $/W)

includes items that scale with area

(mount, wiring, labour, etc.)

For given power kWp, a larger area is needed, higher cost takes longer for amortisation

3.5$/W in 3.5 y

Balance of System (BOS)“cost baseline”


PV landscape in Switzerland

PV-Centre

Neuchâtel

dye sensitizedsolar cells

flexible CIGS

3S Laminators

MB wafer sawing

Pasan sun simulators

Roth and Rau (CH)

PV-Labsystens

module testing

PV-Labc-Si, TF-Si

dye cells

inverters

Equipment(end 2014)


What we do at PV-Lab

TF-Sichallenge: weak absorberption especially for long wavelengths (red light)

Solution: add back reflector => 2x absorptionadd texture for scattering => up to 50x

TCOp

i

n

TCO

p-i-n (usually on rigid glass)

Superstrate

(glass)

back reflector

grow

th


Textured ZnO:B electrodes

Requirements:- conductivity: add B-dopant (diborane)- surface texture for light scatteringExamples:

ZnO:B (PV-Lab std.)SnO2:F (commercial, Asahi)

W. Wenas, Jap. J. Appl. Phys (1991)S. Faÿ, EU PV Conf. (2000)


Take care with texturing!

Collision of growth fronts on facets yields defective material !

H. Sakai, Jap. J. Appl. Phys. (1990)Y. Nasuno, Jap. J. Appl. Phys. (2001) M. Python, J. Non-Cryst. Sol., (2008)

Here: μc-Si, but also observed in a-Si


Cells on flexible plastic substrate

Nano-imprinting used for texturing of plastic (here: periodic grating structure)

T. Söderström, Appl. Phys. Lett, (2009)K. Söderström, Prog. in PV, (2010)

J. Escarré, J. Optics, (2012)R. Biron, Sol. En. Mat. (2013)


Heterojunction solar cells

c-Si is a perfect crystal; cell-performance is not bulk- but surface-limited

Intrinsic (undoped) a-Si:H provides excellent passivation of c-Si surface

Charge is extracted through passivating a-Si:H bi-layers

High efficiency cells >22% (certified)

Surfacepassivation

A. Descoedre, EU-PV Conf., (2012)


Research on hetero-junction cells

• Thin a-Si layers: max 5 nm

• Pure a-Si layers: avoid epitaxy on underlying c-Si

• Chemical annealing: alternate SiH4 and H2 plasma

• Other:– Texture etch (avoid sharp valleys)– Highly transparent TCOs (very high transmission at 1100 nm)– Replace screen printing of Ag contacts by galvanic Cu

34F.-J. Haug – Photovoltaics

• Acknowledgement for funding– Swiss Federal Office for Energy

EU FP6 and FP7, CTI, FNS, CCEM-CH, Swiss Electric Research, Axpo Naturstromfonds, Velux-Stiftung, IBM…

– Bosch, Oerlikon Solar, Pasan, Flexcell, Solvay, Dupont, Metalor,Meyer-Burger (3S Moduletec, Roth and Rau), Indeotec SA, …

• Thanks also to the PV-Lab members

Thanks for your attention


Photovoltaics F.-J. Haug Ecole Polytechnique Fédérale de Lausanne (EPFL), Institute of Microengineering (IMT), Photovoltaics and Thin Film Electronics.

Documents

haug photovoltaics2

haug photovoltaics3

haug photovoltaics12

film solar cells

haug photovoltaics8

haug photovoltaics5

haug photovoltaics10

haug photovoltaics11