Ab-initio model of Si:Ge by VASP: (a) 1D phonon DOS

for Si:Ge MQW. 3D DOS for QD arrays, Si dots of (b) 1.3 nm and (c) 2.1 nm. Insets show detail over 60-64 meV: indicating very small mini-gaps in 3D.

3D force constant modelling of hypothetical QD supe rlattice

Hot Carrier Solar Cells – structures for slowed carr ier coolingG. Conibeer, M.A. Green, D. König, S. Shrestha, S. Huang , P. Aliberti, L. Treiber, R. Patterson, B. Puthen Vee ttil, A. Hsieh

– ARC Photovoltaics Centre of Excellence, UNSW, Sydney

A.Luque, A. Marti, P.G. Linares, E. Cánovas, E. Ant olín, D. Fuertes Marrón, C. Tablero, E. Hernández- Instituto de Energía Solar-Universidad Politécnica de Madrid

J-F. Guillemoles, L. Huang, A. Lebris, S. Laribi, P . Olsson- IRDEP, joint EDF-CNRS Institute for Photovoltaic R&D , Paris

T.W. Schmidt, R.G.C.R. Clady, M.J.T. Tayebjee- School of Chemistry, University of Sydney

ConclusionsConclusionsConclusionsConclusions

Concept of the Hot Carrier cell [1,2]Concept of the Hot Carrier cell [1,2]Concept of the Hot Carrier cell [1,2]Concept of the Hot Carrier cell [1,2]

•Absorption of all solar photons with energies greater than the absorber threshold energy. •Collect carriers before they thermalise. Requires:Requires:Requires:Requires:- Selective energy contacts -- Slowing of carrier cooling -

Selective energy contacts

Hot carrier absorber HolecontactElectroncontact

Hot Optical phonon population“phonon bottleneck effect”

Slows further carrier cooling

Decay via O → LA + LA (only)

Electrons carry most energy

Cool predominantly via small wave vector optical phonon

emission - timescale of psinelastic – energy relaxation

LO

TO

LA & TA – quasi elastic

Hot Carrier coolingHot Carrier cell requirements

• InP has large phonon band gap - should block Klemens decay• GaAs has no phonon gap – but similar Eg

Eg

GaAs InP

Eg

E

k

TA

LA

LOTO

63meV

O →→→→ LA + LA (Klemens) – principal mode in semiconductors [3]

LO

TO

LALA

Large phonon gap in bulk materials → M >>m- Klemens blocked

LA

TO

TA

LA

XL Γ

LA

LA

TO

QD nanstructure - mini-gaps- folded phonon modesBlock Klemens?- need specific periodic

superstructureBlock Ridley?

•Peak intensities (black arrows) shift closer to the band gap (white lines) with timeInP stays further above the band gap for longer times than GaAs

• Slower cooling in InP due to suppression of Klemens’ decay

Low temperature QEadapted for the UV

Deposition ofDeposition ofDeposition ofDeposition of nanocrystalnanocrystalnanocrystalnanocrystal superstructuressuperstructuressuperstructuressuperstructures

Four characterisation techniques for HCSC: • Low temperature current-voltage • Low temperature quantum efficiency (QE) • Photoreflectance for band diagram analysis• Time resolved PL – measure carrier cooling

Low temperature I-Vhigh I and very low I

Time resolved PL spectra for bulk GaAs and InP 730 nm excitation – carrier density - 8.5x10-19 cm-3

• Carriers cool by emission of phonons - restricting Optical to Acoustic phonon decay can slow cooling • Binary semiconductors can have large band gaps between O and A modes,

e.g. InP: phonon gap large enough to block Klemens decay – from TRPL• Ridley mode allowed but this has lower energy loss • Cubic materials restrict Ridley loss through narrow optical dispersion

• Folding of Brillouin zone in QD nanostructures gives gaps in phonon DOS • These can prevent Klemens decay if tuned correctly • Modelling in 1D and 3D – Group IV and hypothetical superlattices • Complete gaps in reciprocal space can block Klemens decay• Langmuir Blodgett deposition of ordered arrays of NP superstructures

Ma

Tacoustic

12=ω

+=Mma

Tand

ma

Toptical

11212ω

Simple force constant model derived from the equationof motion gives phonon frequency, ω, as:

Acoustic max. and optical min. & max. phonon energies[‘M’ (heavy), ‘m’ (light) atomic masses; T/a = force constant]

Optical to acoustic phonon decay

BlockingBlockingBlockingBlockingOptical Phonon decayOptical Phonon decayOptical Phonon decayOptical Phonon decay

O → TO + LA – Ridley [4] smaller energy loss

Cubic << Ridley

Time resolved PL Time resolved PL Time resolved PL Time resolved PL ---- resultsresultsresultsresults

.

Maximum phonon energy, ωoptical

ωoptical / 2aligned with mini-gap

(a) 1D phonon DOS Si:Ge MQW

(b) 3D DOS

(c) 3D DOS

Very narrow mini-gaps – bigger for 1.3nm cf. 2.1nm QDs

(111)(110)diamond superlattice (100)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

BiB

BS

b

InN

SnO

BA

s

GaN

AlS

b

InP

BP

SiC

AlN

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Eoptical (M

ax -M

in)

Eacoustic

Eop

tical

−E

acou

stic

Eac

oust

ic

Phonon miniPhonon miniPhonon miniPhonon mini----gaps ingaps ingaps ingaps in nanostructuresnanostructuresnanostructuresnanostructures

3D phonon dispersions for a diamond QD superlattice, with 1nm diamond QDs. Mass ratio MMatrix:MQD = 1:7. There are complete gap in reciprocal space – sufficient to block Klemens’ decay, indicated by the arrow at ωoptical / 2

PhononicPhononicPhononicPhononic band gaps in bulk materialsband gaps in bulk materialsband gaps in bulk materialsband gaps in bulk materials

References: References: References: References: 1. P. Würfel, SOLMAT. 46 (1997) 43-47. 2. M.A. Green, Third Generation Photovoltaics (2003). 3. P. Klemens, Phys Rev 148 (1966) 845; 4. J.W. Pomeroy et al., AAPL. 86 (2005) 223501.

Characterisation for Hot Carrier Solar CellsCharacterisation for Hot Carrier Solar CellsCharacterisation for Hot Carrier Solar CellsCharacterisation for Hot Carrier Solar Cells

substrate

Langmuir-Blodgett deposition of amonolayer of encapsulated NPs

InP/CdS core-shell NPs: 1. Synthesis 2. Characterization3. Test of phonon engineering concepts (Planned)

Langmuir-Blodgett deposition to fabricate ordered arrays of nanoparticles (NP)• Initial work on Si & Au NPs of uniform size• Functionalise to give close packing• Langmuir-Blodgett deposition in layers• Characterise periodicity• Later work on NPs with MQD>>mmatrix

IES-UPM

Raman spectra of InP/CdS NP’s

(crossed polarization)Displaying LO & TO modes