Woods Energy Seminar, 28 May 2008

Third Generation Photovoltaics

Photovoltaics Centre of Excellence supported by the Australian Research Council, the Global Climate and Energy Project and Toyota CRDL

Gavin Conibeer

Deputy DirectorARC Photovoltaics Centre of Excellence

University of New South Wales

Transforming the global energy mix: The exemplary path until 2050/ 2100appointed for a term of four years by the federal cabinet (Bundeskabinett)

Meeting the IPCC target of 60% reduction in GHG emission by 2050

The business case for early action60% reduction in GHG emission by 2050

Outline

• The importance of Photovoltaics• Three generations of Photovoltaics• The main losses in photovoltaic cells• Third Generation approaches

• Silicon nanostructure tandem cells• Band gap engineering – quantum confinement• Fabrication of materials / devices

• Hot Carrier cells• Contacts – energy filtering • Hot Carrier cooling – energy loss to phonons

• Summary

Booming Photovoltaics

Global PV market US$6.5 billion in 2006 → $16.4 billion in 2012

Market growth at 35%/yr for last 10 years, 60%+ in 2007Approx 1 million jobs in PV by 2020Approx 1 million jobs in RE by 2010

Driven by rebates/tariffs: Japan, Germany

Now other Euro. Countries and S Australia

USA: Power purchase agreements

Japan: market is stablewith reducing rebates

1988

1991

1994

1997

2000

2003

2006

0

1000

2000

3000

4000

5000

MW

p

USAEuropeJapanRest of WorldTotal

Annual capacity increase

2000

2001

2002

2003

2004

2005

2006

Solar Heating

Wind

Nuclear

Photovoltaics0

5

10

15

New

Cap

acity

, GW

.

Sources: Photon International, WNA, Sources: Photon International, WNA, WWEA, IEAWWEA, IEA

Google’s Mountain View campus

Learning curves

bulk-Si(~10%)

2003

US

$ /kW (~20%)

20000

10000

5000

2000

1000

500

2000.01 0.1 1.0 10.0 100.0

Gas turbines (USA)

1980

1963

1982

19871993

2001

Wind turbines

2002

Cumulative GW installed

1981

Photovoltaics

. more potential for learning

. lower cost at smaller volumes

20000

10000

5000

2000

1000

500

2000.01 0.1 1.0 10.0 100.0

2002

Cumulative GW installed

1981

Photovoltaics

bulk-Si(~10%)

2003

US

$/kW (~20%)

Thin-film PV

2nd Generation

2002

3rd Generation

1st Generation

100

80

60

20

0 100 200 300 400 500

US$0.50/W

US$1.00/W

US$3.50/W

Cost, US$/m2

US$0.10/W US$0.20/W

Present limit

Thermodynamiclimit

40

Effic

ienc

y,%

Photovoltaics: Three Generations

III

III

mc-Si

III-V tandem

c-Si

thin film

a-Si tandem

concentration

Efficiency Loss Mechanisms

Two major losses – 50%

Limiting efficiencies 1 sun Max concn.Single p-n junction: 31% 40.8%Multiple threshold: 68.2% 86.8%

qV

2. Lattice thermalisation

2

2

1. Sub bandgap losses

1

Energy

3

Also: 3. Junction loss

4

44. Contact loss

5

5

5. Recombination

Third generation options

100%

74%68%

54%49%44%39%31%

0%

58%

circulators

tandem (n )hot carrier

impurity PV & band, up-converterstandem (n = 3)thermal, thermoPV, thermionics

impact ionisationtandem (n = 2)down-converterssingle cell

tandem (n = 6)65%

Eg

Eh

Elintermediate level

Ef

Jl

Jh

JVC

CB

VB

Erela

x

h+

e-

e-e-

h+

E0,e

E2,h

E2,e

One photonMultiple electrons-

E0,h

Intrinsic Intrinsic radiativeradiative and Auger losses includedand Auger losses included

Free choiceor Si cell

Decreasing band gap

Sunlight

42.5%47.5%

50.5%

Number of cells1 2 3

10

20

30

40

00

Si bottomcellFreechoice

29%

33%

45%

AM1.

5GEf

ficie

ncy

Silicon based Tandem Cell

Silicon based Tandem Cell

SiO2

SRO

Substrate

Thin film Si cellEg = 1.1eV

2nm QD, Eg =1.7eV

Si QDs

defect or tunnel

junction

SiO2barriers

Engineer a wider band gap – Si QDsSo

lar

Cel

l 1

Sola

r C

ell 2

Sola

r C

ell 3

Decreasing band gapSo

lar

Cel

l 1

Sola

r C

ell 2

Sola

r C

ell 3

Decreasing band gap

SixSiOxSiOx ⋅⎟⎠⎞

⎜⎝⎛ −+→

21

2 2

SubstrateSubstrate

Anneal 1100°C – Si precipitation

Si nanostructure tandem cell

1.0

1.5

2.0

2.5

3.0

3.5

0 1 2 3 4 5 6 7Dot diameter [nm]

PL e

nerg

y [e

V]

Y. Kanemitsu et alH. Takagi et alS. Takeoka et alT. Y. Kim et alT. W. Kim et al Oxide (UNSW) Nitride (UNSW)

Si QDs in oxide/nitride

Si QDs in SiC

400 800 1200 1600 2000

a-Si

Nanocrystalline SiC

508 cm-1Si nanocrystal

As deposited

800 oC

1000 oC

1100 oC

Inte

nsity

(a.u

.)

Raman shift (cm-1)

20 30 40 50 60 70 80

Annealed at 1100 oC

Si (311)

Si (220)

β-SiC (111)

Si (111)

Inte

nsity

(a.u

.)2 Theta (deg.)

c-Si

SiC

0.9 eV

1.1 eV

0.5 eV

Si3N4

c-Si

1.9 eV

1.1 eV

2.3 eV

c-Si

SiO2

3.2 eV

1.1 eV

4.7 eV

Alternate matrices

Gaussian modelling of Si QD in SiC

Alternate QDsTin QDs in SiO2

Ge QDs in SiO2

Various material combinationsQuantum Dot / Matrix combinations and current status of investigations

-POSPOSn

--SPGe

SPODSPOEDSPOEDSi

SiCSi3N4SiO2

S = Simulation (ab-initio modelling - DFT)P = Physical (electron microscopy, X-ray difraction)O = Optical (photoluminescence, absorptance)E = Electronic (conductivity, conductivity with Temp.)D = Devices (Diodes, Cells)

Increasing conductivity

Decreasing processing temperature

Si QD

SiO2

SiC substrate

n-SiC wafer

Light

quartz

Si QD

SiO2

SiC pseudo-substrate

sputtered SiC

Light

quartz

p-Si QDSiO2

Homojunction– front back contact

sputtered SiC

n-Si QD

Light

Si substrate - problem – Can’t be sure absorption is not in Si Hence – transparent substrate or pseudo-substrate

Voc = 93 mVSi substrate

N+ - Si NC:SiC (100nm, Sb doping)

SiN (70 nm)

3mm

P-Si NC:SiC (200nm, B doping)(600 nm B-doping)emitter

absorberBarrier

(200 nm Sb-doping)

Open circuit Voltage = 83mV

Hot Carrier cellExtract hot carriers before they can thermalise:

small Eg

TH

Ef

Hot carrier distribution

• Need to slow carrier coolingRoss & Nozik, 1982

Würfel, 1995Green, 2003Würfel, 2005

Takeda et al, SOLMAT, 08

Eδ

Ef(n)

TA

qV=ΔµA

Ese- energy selective contact

h+ energy selective contact

Es

Ef(p)

TA

• Collect carriers over narrow range of energies

Ross & Nozik, JAP, 53 (1982) 3813Würfel, SOLMAT, 46 (1997) 43 1995Green, 3rd Gen PV (S-Verlag) 2003Würfel, PIP, 13 (2005) 277

Si QD

Dielectricmatrix

ResonantTransport

Resonant Tunneling Transport

Energy

Energy Selective Contact

Filter

ECEf

Ef

I

V

0

0.01

0.02

0.03

0.04

0 0.5 1 1.5

Gate voltage (V)

Ig(A

) Two different sites on the wafer

NDR at 300K - Repeatable

Hot Carrier cooling

Hot Optical phonon population“phonon bottleneck effect”

Slows further carrier cooling

Decay of Optical phonons to Acoustic is critical

Electrons carry most energy

Cool predominantly via small wave vector optical phonon

emission - timescale of psinelastic – energy relaxation

Optical phonons emitted

Energy

Optical phonon decay

Optical phonon decay

O → LA + LA (Anharmonicity or Klemens mechanism)

Compound – e.g. InN

Element– e.g. Si

Phon

on e

nerg

ies

(den

sity

of s

tate

s) 60

meV

30

0

Allowed phonon energies

E

Nō

Optical phonons(standing waves)

Acoustic phonons(heat in the lattice)

Some evidence for slowed carrier cooling in InN: Chen & Cartwright, APL, 83 (2003) 4984

And for longer phonon lifetimes in GaN, AlSb, InP – all of which have large phonon gaps

Phononic gaps in nanostructures

Linear force constant model: l = 4a1 + 4a2– mass ratio = 2; force constant ratio = 5

Nanostructure

Phon

on e

nerg

ies

(den

sity

of s

tate

s)

40

meV

20

0

Phonon propagation in nanostructure

Acoustic phonon reflected from zone edges → standing wave

Towards a complete cell•Fabrication of slowed cooling absorber

•Energy Selective Contacts

•Transport and Renormalisation of carrier energies

Summary•Relevance and growth of Photovoltaics•Three PV Generations•Main energy losses•Third Generation approaches•Si nanostructure tandem cells

•Band gap eng.•Range of QD materials•Early devices

•Hot Carrier cells•Energy filter contacts•Phonon bottleneck•Nanostructures - QD based cell

•Third generation multi-energy level devices•tend to involve QD nanostructures •enable tailoring of material properties

Research Staff: Martin Green, Richard Corkish, Gavin Conibeer, Dirk König, Eun-ChelCho, Tom Puzzer, Yidan Huang, Shujuan Huang, Dengyuan Song, Santosh Shrestha, Ivan Perez-Wufl, Supriya Pillai

PhD students: Chris Flynn, Jeana Hao, Sangwook Park, Lara Treiber, Yong So, Pasquale Aliberti, Yong So, Andy Hsieh, Bo Zhang, Rob Patterson, Binesh Puthen Veettil,

Visiting researchers: Fei Gao, Dong-Ho Kim, Ke Ma, Veronique Gevaerts, Martin Kirkengen, Martina Schmid

Third Generation Strand (2008)