silizium-photovoltaik – status und neue entwicklungen · large-area record cells on n-type...

Stefan Glunz, Fraunhofer ISE 2018 1

© Fraunhofer ISE

SILIZIUM-PHOTOVOLTAIK –STATUS UND NEUE ENTWICKLUNGEN

Stefan Glunz

Fraunhofer Institute for Solar Energy Systems ISE

Seminarreihe Erneuerbare Energien

27. Juni 2018,

Hochschule Karlsruhe

© Fraunhofer ISE, S. Glunz 2018

2

PV Module Production Development by TechnologyIt is still silicon …

Production 2016 (GWp)

Thin film 4.9

Multi-Si 57.5

Mono-Si 20.2

Data: from 2000 to 2010: Navigant; from 2011: IHS (Mono-/Multi- proportion from cell production). Graph: PSE 2017

© Fraunhofer ISE



3

PAST


4

PASTThe Bell labs 1954

Gerald Pearson, Darryl Chapin, and Calvin Fuller testing their silicon solar cell.



5

1940 1950 1960 1970 1980 1990 2000 2010 20200

5

10

15

20

25

30

mono-Si (p-type) mono-Si (n-type)

Effi

cie

ncy

[%]

PAST The Beginning

Strong increase of efficiency in the 1950s

n-type silicon dominates as base material

29.4%

G.L- Pearson, American J. Phys. 25, p.591 (1957)


6

PASTThe First Application Space

1957 Sputnik (USSR)

1958 Explorer 1 (USA)

1958 VanguardFirst solar-powered satellite

Sputnik 1

Vanguard 1



7

1940 1950 1960 1970 1980 1990 2000 2010 20200

5

10

15

20

25

30


Effi

cie

ncy

[%]

PASTFrom Space to Earth

Switch to p-type silicon due to higher radiation stability for space applications

Reduction of recombination losses

Model for current industrial cell generation

29.4%

Small area

Blakers et al., Appl. Phys. Lett. 55, 1363 (1989)


8

PRESENTAverage Price for PV Rooftop Systems in Germany

Data: BSW-Solar. Graph: PSE 2018

© Fraunhofer ISE

BOS incl. Inverter

Modules

Percentage of the Total Cost

System costs strongly area-related

Further increase of cell efficiency



9

PRESENTLarge-area Record Cells on n-type Silicon

Panasonic/Sanyo1 Masuko et al., IEEE PVSC (2014)

SunPower: 2 Smith et al., IEEE PVSC (2014)

3 Yoshikawa et al., Nature Energy (2017)

Interdigitated back contact back junction solar cells

Excellent contact passivation (a-Si/c-Si heterojunction, passivated contacts)

Sanyo1 (da=143.7 cm2)25.6% (Voc = 740 mV)

SunPower2 (ta=153.5 cm2)25.2%3 (Voc = 737 mV)

2016: Kaneka (da=180.4 cm2)26.6%3 (Voc = 744 mV)26.7%


10

1940 1950 1960 1970 1980 1990 2000 2010 20200

5

10

15

20

25

30


Effi

cie

ncy

[%]

PRESENTRecord Cells on Crystalline Silicon

Large-area (Sunpower, Sanyo/Panasonic, Kaneka)

Sophisticated cell architecture

Excellent material quality needed

29.4%

Large area

Small area



11

THE Working HorseScreen-printed Al-BSF solar cell on p-type silicon

Saw Damage Removal

Texture + Cleaning Diffusion

PSG RemovalandIsolation ARC

Screen-printing Firing

Process

n-Emitter

ARC

p-Base

p-Contact

n-Contact

pn-Junction

Al-BSF

Still the main technology of the PV technology (70 to 80% of the market)

Efficiencies up to 19% (multi-Si) or 20% (mono-Si)

Incremental improvements (better pastes, surface passivation)

It is hard to beat the main stream!


12

The Next Industrial Cell GenerationPartial Rear Contact Cells (PRC or PERC)

The successor of Al-BSF cells:Partial Rear Contact (PRC) cells also known as PERC (25 years after its invention!)

Improved electrical (= less recombination)

and optical (=higher internal reflection)

-

- +

+Passivation layer

Reducedcontact area



13

The Next Industrial Cell GenerationSwitching from Al-BSF to PRC (PERC)

Blakers et al., Appl. Phys. Lett. 55, 1363 (1989)

Mandelkorn and Lamneck,J. Appl. Phys. 44, 4785 (1973)

Al-alloyed back surface field (Al-BSF) solar cells have dominated the industry for decades

Partial rear contact cells (PRC) like PERC/PERL are introduced in production

= 19% - 20% = 20% - 21.5%


14

The Next Industrial Cell GenerationIndustrial process for PERC cells

Two additional process steps

Dielectric passivation

Local contact opening (LCO)

SDE/Texture

POCl diffusion

Edge Isolation

PSG etching

SiN ARC

SP Ag FS

Drying & Firing

SP Al/Ag RS

Al2O3/ SiN RS

Laser Opening



15

The Next Industrial Cell GenerationPartial Rear Contact Cells (PRC or PERC)

The successor of Al-BSF cells:Partial Rear Contact (PRC) cells also known as PERC (25 years after its invention!)

Record cell results:

p-type mono-Si:22.1% (Trina)22.0% (Solar World)> 23% (Jolywood)

p-type multi-Si21.25% (Trina)22% (Jinko Solar)

P. Verlinden, WCPEC, Kyoto, 2014


16

PERC ModulesBifacial p-type PERL Solar Cells

Starting point: PERC-process

Fingers on front and back side

Bifacial power conversion

Dullweber et al. PiP 2015 SolarWorld BiSun technology



17

Module TechnologyNo More Busbars

Cell has only fingers and no busbars

Interconnection with solder-coatedwires

Lower reflection losses

Improved aesthetics

LG Neon moduleMeyer&Burger Smart Wire


18

Bridging the GapState-of-the-Art Silicon Solar Cell: Cost Analysis

p-t

yp

e m

c A

l-B

SF

n-type CzWorld Record

????

Allowed costs for same levelized costs of electricity (LCOE) as a function of cell efficiency

M. Hermle et. al., 29th EUPVSC 2014.

18 20 22 24 26

100

150

200

250

norm

alis

ed c

ost

of

cell

pro

duc

tion

[%

]

cell efficiency [%]



19

Bridging the Gap State-of-the-Art Silicon Solar Cell: Cost Analysis

p-t

yp

e m

c A

l-B

SF

n-type CzWorld Record

Allowed costs for same levelized costs of electricity (LCOE) as a function of cell efficiency

Bridging the gap:- Higher efficiency- Reasonable complexity

18 20 22 24 26

100

150

200

250

no

rma

lise

d co

st o

f ce

ll pr

od

uctio

n [

%]

cell efficiency [%]


20

PRESENTPERC – Restrictions

n-base

p+

n++

J0,met > J0,pass

PERC cells have high efficiency potential

But: Strongly two-dimensional cell structure (contact distance = 5 cell thickness)

Trade-off in cell design: Voc vs. FF

Influence of base resistivity

Additional patterning step required

local contact



21

Bridging the Gap n-Type PRC – Restrictions

J0,met = J0,pass

n-base

p+

passivating contact

n-base

p+

n++

J0,met > J0,pass

local contact


22

Bridging the GapHybrid TOPCon Cell

J0,met = J0,pass

n-base

p+

passivating contact

n-type hybrid cell with boron emitter at the front and a passivated rear side offers

1. transparent front side

2. less influence of base resistivity

3. no patterning of the rear side



23

TOPCon (Tunnel Oxide Passivated Contact)Combining a-Si Hetero and poly-Si/Tunnel Oxide Approach

Higher band gap Very good selectivity

Better thermal stability

Tunnel oxide

n-Si BaseAmorphous/Polycryst.Si (n)-Layer

ECEF

EV

Tunnel oxide

ECEF

EV

ECEF

EV

n-Si Base PolycrystallineSi(n)-Layer

n-Si Base Amorphous

Si(n)-Layer

F. Feldmann et al., SOLMAT 120 (2014)


24

a-/nc-Si

TOPCon (Tunnel Oxide Passivated Contact)Combining a-Si Hetero and poly-Si/Tunnel Oxide Approach



25

Si Thin Film

TOPCon (Tunnel Oxide Passivating Contact)A New Rear Contact

Tunnel oxide

n-base

Boron emitter

Highly dopedsilicon thin film

Metal

F. Feldmann et al. ,SOLMAT (2014)


26

High-Efficiency Solar CellsCells with Top/Rear Contacts

Material Area Voc Jsc FF η

[cm2] [mV] [mA/cm2] [%] [%]

UNSW/ PERL1 p-type 400 μm 4 (da) 706 42.7 82.8 25.0

Kaneka/ HJT2 n-type 200 μm 152 (ap) 737 40.8 83.5 25.1

2Adachi et al., Appl. Phys. Lett. (2015)1Zhao et al., Progr. Photovolt. (1999)



27



[cm2] [mV] [mA/cm2] [%] [%]

UNSW/ PERL p-type mono-Si 4 (da) 706 42.7 82.8 25.0

Kaneka/ HJT n-type mono-Si 152(ap) 737 40.8 83.5 25.1

ISE / TOPCon1 n-type mono-Si 4 (da) 719 41.5 83.4

TOPCon: J0,rear º 3-5 fA/cm²

n-basep++

*ISE CalLab Measurement


28



[cm2] [mV] [mA/cm2] [%] [%]



ISE / TOPCon1 n-type mono-Si 4 (da) 719 41.5 83.4 24.9*


n-basep++


Resistive Losses; 0.05

Bulk Recomb. 0.12

Front Recomb. 0.70

Rear Recomb.

0.39External Rs

and Rp

0.49

FELA[mW/cm2]



29



[cm2] [mV] [mA/cm2] [%] [%]



ISE / TOPCon1 n-type mono-Si 4 (da) 725 42.5 83.3


n-basep++



30



[cm2] [mV] [mA/cm2] [%] [%]



ISE / TOPCon1 n-type mono-Si 4 (da) 725 42.5 83.3 25.7*


n-basep++


1Richter et al., SOLMAT (2017)



31



[cm2] [mV] [mA/cm2] [%] [%]



ISE / TOPCon

ISE/ TOPCon1

n-type mono-Si

n-type multi-Si

4 (da)

4

725

674

42.5

41.1

83.3

80.5

25.7*

22.3*


1Benick et al., IEEE JPV (2017)


32

Record Efficiencies

October 2017



33

Record Efficiencies

Mono-Si heterojunctionMono-Si homojunction

Perovskite (not stabilized)CIGSMulti-SiCdTe





October 2017


34

What is the limit?Detailed Balance

Shockley und Queisser, 1961

Detailed balance between sun and solar cell

Assumption: Solar cell emits photons via radiative recombination

Radiative recombination in a direct semiconductor

E

k

W. Shockley & H. Queisser



35

What is the limit?Maximum Efficiency as a Function of Bandgap

Max. Efficiency ~ 33%

High thermalisation for low bandgaps

High transmission for high bandgaps

Silicon and GaAs are close to the optimum

But: Record values of GaAsare closer to the limit.

Is III/V-R&D better than silicon-R&D ?

Thermalisation Transmission

0.5 1.0 1.5 2.0 2.50.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Ge

GaSb

mc-Si

c-Si

CIGSInP

GaAs

CdTe

a-Si AlGaAs

Shockley-Queisser

Record Efficiencies

EG [eV]

Eff

icie

ncy


36

What is the Limit?Other Recombination Channels

Assumption:Ideal solar cell: only radiative recombination (Shockley and Queisser, J. Appl. Phys. 1961)

But silicon is an indirect semiconductor radiative recombination has a low probability

Radiative recombination in an indirect semiconductor

E

k

Phonon



37

What is the limit?Auger-Recombination

In silicon solar cells Auger-recombination is the limiting intrinsic loss mechanism

Auger recombination in an indirect semiconductor

Pierre Auger 1899 - 1993

E

k

BZB


38

What is the limit?Taking Auger Recombination into Account

Shockley, Queisser (1961)= 33% (AM1.5)

Theoretical efficiency limit for silicon (taking actual Auger model 1 into account)= 29.4%2

1Richter, Glunz et al., Phys. Rev. B 86 (2013)

2Richter, Hermle, Glunz, IEEE J. Photovolt. (2013)



39

0.5 1.0 1.5 2.0 2.50.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Ge

GaSb

mc-Si

c-Si

CIGSInP

GaAs

CdTe

a-Si AlGaAs

Shockley-Queisser

Record Efficiencies

EG [eV]

Eff

icie

ncy

What is the Limit?Taking Auger Recombination into Account

Shockley, Queisser (1961)= 33% (AM1.5)

Theoretical efficiency limit for silicon (incl. Auger)= 29.4%1

Best silicon solar cells= 25.6%2

Corresponds to 87% of theoretical efficiency limit

(GaAs = 87% )

1Richter, Hermle, Glunz, IEEE J. Photovolt. (2013)

2Masuko et al., IEEE-PVSC (2014)


40

FUTUREBeyond the Shockley-Queisser-Limit

Shockley, Queisser (1961):Efficiency limit of solar cell made of one material = 33%

Theoretical efficiency for silicon (Auger limit) = 29.4%1

Light management

Up-conversion

Beam splitting

Tandem cells with silicon as bottom cell

Perovskite top cell

Silicon quantum dots

III/V top cell

!

1Richter, Hermle, Glunz et al., IEEE J. Photovolt. 2013

Thermalisation Transmission

0.5 1.0 1.5 2.0 2.50.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Ge

GaSb

mc-Si

c-Si

CIGSInP

GaAs

CdTe

a-Si AlGaAs

Shockley-Queisser

Record Efficiencies

EG [eV]

Eff

icie

ncy



41

Si (1.12 eV)

Beyond the LimitMagic Antireflection Coatings (AAA-Coating®)

Si (Eg = 1.12 eV)

Silicon Solar Cells with AAA-Coating®

(amazingly active antireflection coating)

Si (1.12 eV)

GaAs (1.42 eV)

GaInP (1.88 eV)

AAA-Coating®


42

Beyond the LimitSilicon Solar Cell

Theoretical limit for Eg,Si: 33% (29.4% consideringAuger recombination)

Word record for siliconsolar cells: 26.7%

400 600 800 1000 1200 1400 1600 1800 20000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Transmission loss

Bandgap

Usable power

Thermalization loss

Inte

nsi

ty [

W m

-2n

m-1]

Wavelength [nm]



43

Beyond the LimitMultijunction Solar Cells

Tandem cells with Si bottom cell

III/V on Si

Silicon quantum dots

Perovskites

“Silicon Solar Cell 2.0”

400 600 800 1000 1200 1400 1600 1800 20000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

AM1.5 1. Cell 2. Cell 3. Cell (Si)

Inte

nsi

ty [

W m

-2n

m-1]

Wavelength [nm]


44

Beyond the LimitIII-V on Silicon Tandem Solar Cells



45

Beyond the LimitSilicon-based Multijunction Cells

Si (1.12 eV)

GaAs (1.42 eV)

GaInP (1.88 eV)

Top cells with high bandgap to utilize blue and visible light

c-Si bottom cells for IR light

Deposition by direct epitaxial growth or wafer bonding


46


GaInP pn-junction

GaAs pn-junction

Si bottom cell

III-V Substrate

Bonding to newsubstrate

III-V substrate lift-off and recycling



47


GaInP pn-junction

GaAs pn-junction

Si bottom cell

Processing of solar cell contacts andARC

5-10 μm


48

10 nmSilicon- cell or wafer

GaAs- cell

High Resolution TEM Image, Bright Field, Zone Axis Si, Universität Kiel, Group Prof. Dr. Jäger, 2011

HRTEM-image of Si/GaAs interface 4 nm thick amorphous layer

Beyond the Limit2-terminal GaInP/AlGaAs//Si

GaInP- cell

1 µm

GaAs

Silicon



49

Beyond the Limit2-terminal GaInP/AlGaAs//Si

Very good current matching:Each cell around 12 mA/cm2

Efficient utilization of spectrum

400 600 800 1000 1200 14000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

AM1.5g GaInP GaAs Si Total

Inte

nsity

[W

m-2nm

-1]

Wavelength [nm]

Si (1.12 eV)

GaAs (1.42 eV)

GaInP (1.88 eV)


50

Beyond the Limit2-terminal GaInP/AlGaAs//Si >30% @1-Sun AM1.5g

R. Cariou et al., Fraunhofer ISEIEEE Journal of Photovoltaics 2017


Very good current matching

Efficiency = 30.2 > 29.4%

R. Cariou et al., Fraunhofer ISENature Energy 2018



51

Wafer-Bonded 2-Terminal GaInP/AlGaAs//Si TandemTOPCon with Nanostructured Diffraction Grating

Voc

[V]Jsc

[mA/cm2]FF[%]

η[%]

Gen 3

3.127 12.7 83.8 33.3

Ag

p+ poly-Si

Resist

1 µm

R. Cariou, Nature Energy (2018) doi:10.1038/s41560-018-0165-5


52

Beyond the LimitIII-V on Silicon Tandem Solar Cells

1.9 μm of III-V material on Si sufficient to realize = 33.3% solar cell by wafer bonding

= 19.7% efficiency reached forGaInP/GaAs/Si triple-junction bydirect growth

Both values included in PIP Efficiency tables 2018



53

Beyond the Limit2-terminal GaInP/AlGaAs//Si > 29.4% @ 1-Sun AM1.5g


Very good current matching

Efficiency = 33.3 > 29.4%

Beyond the Auger limit for crystalline silicon solar cells

R. Cariou et al., Fraunhofer ISENature Energy 2018


54

Conclusion

Photovoltaics is a significant player in the energy market.

Prices are already very low. Conversion efficiency is the key to further bring down the levelized costs of electricity and to survive competition.



55

Conclusion



New cell structures with high industrial potential.


56

Conclusion




New fascinating concepts for an old technology:Crystalline silicon solar cells 2.0

Si (1.12 eV)

GaAs (1.42 eV)

GaInP (1.88 eV)



57

Conclusion




New fascinating concepts for an old technology:Crystalline silicon solar cells 2.0

Si (1.12 eV)

GaAs (1.42 eV)

GaInP (1.88 eV)

silizium-photovoltaik – status und neue entwicklungen · large-area record cells on n-type...

Documents