Intermediate band materials for high efficiency solar cells: - - PowerPoint PPT Presentation

Intermediate band materials for high efficiency solar cells:

• verview and future directions

Jacob J Krich Department of Physics & School of Electrical Engineering and Computer Science University of Ottawa

University of Ottawa Jacob J. Krich

Collaborators

uOttawa Eduard Dumitrescu Akshay Krishna Kyle Robertson Luc Robichaud Peter Rose Anna Trojnar Josh Wheeler Matt Wilkins Daixi Xia Emily Zhang Karin Hinzer Ross Cheriton Alex Walker US Army Research Jeff Warrender Quentin Hudspeth Philippe Chow McGill/Michigan Zetian Mi Ashfiqua Connie Hieu Nguyen Stanford Aaron Lindenberg Middlebury Renee Sher Australian National University Jim Williams Shao Qi Lim Wenjie Yang Harvard Bertrand Halperin Michael Aziz Dan Recht Eric Mazur Toronto Alán Aspuru-Guzik MIT Tonio Buonassisi Austin Akey Christie Simmons Joe Sullivan Mark Winkler University of Dayton Jay Mathews Yining Liu

University of Ottawa Jacob J. Krich

Present topics in the group

Intermediate band materials Nanowire PV Quantum biology/ nonlinear spectroscopy IB device modeling Monochromatic PV

Delay Time T

IB IR photodetectors

10-4 10-2 100

α w

Si:S (Sher 2014)

0.25 0.5 0.75

η

10-6 10-5 10-4 10-3 10-2 10-1 100 101 102

νD

1 2 3

H

University of Ottawa Jacob J. Krich

Outline

• Intermediate band solar cells

– 3 material classes

• Figure of merit

– Measurements – Predictions

• New developments

– InGaN quantum dots in nanowires – Device model

University of Ottawa Jacob J. Krich

Photon flux Photon energy (eV) Photon flux Photon energy (eV)

Intermediate band PV

All è carriers Halfè carriers

n p IB Eg

Max efficiency Standard PV IBPV Concentrated 41% 63% Unconcentrated 33% 47%

n p

Conduction Band Valence Band

Eg

All photons ècarriers

Standard PV

Ee Ee Eg

Detailed balance efficiency limits

F F

Ei

Krishna, Krich, J Optics (2016)

Full concentration Shockley-Queisser limit

University of Ottawa Jacob J. Krich

Ei

1 1

Krishna, Krich, J Optics (2016)

Full concentration 1 sun

Detailed balance efficiency limits

A s s u m p t i

• n

: e a c h p h

• t
• n

e n e r g y

• n

l y a b s

• r

b e d i n

• n

e t r a n s i t i

• n

. è c u r r e n t m i s m a t c h

F F

Photon flux Photon energy (eV)

3 1

Eg

Eg − Ei

Ei

2

3 1 2

University of Ottawa Jacob J. Krich

Ei

1 1

Krishna, Krich, J Optics (2016)

Full concentration 1 sun

Detailed balance efficiency limits

A s s u m p t i

• n

: e a c h p h

• t
• n

e n e r g y

• n

l y a b s

• r

b e d i n

• n

e t r a n s i t i

• n

. è c u r r e n t m i s m a t c h

F

Photon flux Photon energy (eV)

Eg

Eg − Ei

Ei

F

3 1 2

3 1 2 Photon flux Photon energy (eV)

3 1

Eg

Eg − Ei

Ei

2

University of Ottawa Jacob J. Krich

Full concentration 1 sun Overlapping absorptions (1 sun)

Ei

Krishna, Krich, J Optics (2016)

• Detailed balance efficiency limits

1 1 F F

University of Ottawa Jacob J. Krich

IBPV variants

E1 E2

g

E′ Eg DOS E n(↑) n(↓) EF

Olsson, Domain, Guillemoles, PRL 2009

Magnetic IB Strong subgap absorption, Weak nonradiative recombination

VB IB CB RB EF,CB EF,IB ECB EIB EVB ECB EIB EVB ERB ∆E (b) EF,CB GVC RVC GVI RVI GIC RIC

Ratchet band Turn off recombination pathways Better voltage matching

Yoshida, Ekins-Daukes, Farrell, Phillips, APL 2012

University of Ottawa Jacob J. Krich

3 Candidate Materials Classes

Quantum dots Highly-mismatched alloys ZnTeO

• δ-doping

InAs QD GaAs Contact Contact p+ n Substrate (n+ GaAs) Wetting layer

Luque et al., Nat Photon 2012 Wang et al., APL 2009

1 2

Hyperdoped silicon

3

• Pan et al, APL (2011)

University of Ottawa Jacob J. Krich

Best bulk IBPV devices so far

GaAs/GaNAs <1% efficient

López et al., PRL 2011

ZnTe:O 1.4% efficient

Tanaka et al., JPV 2014

• Devices not
• ptimized, but

clear lifetime problem. ZnTe:Cr 5.9% efficient

Lee et al., Sol. Energ. Mat. Sol. Cells 2017

University of Ottawa Jacob J. Krich

GaNAs and ZnTe:Cr

López et al., PRL 2011 Lee et al., Sol. Energ. Mat. Sol. Cells 2017

University of Ottawa Jacob J. Krich

3 Candidate Materials Classes

Quantum dots Highly-mismatched alloys ZnTeO

• δ-doping

InAs QD GaAs Contact Contact p+ n Substrate (n+ GaAs) Wetting layer

Luque et al., Nat Photon 2012 Wang et al., APL 2009

1 2

Hyperdoped silicon

3

• Pan et al, APL (2011)

University of Ottawa Jacob J. Krich

Ion implantation

Damaged Silicon M+ M+ M+ M+ M+ M+

Slide from Dan Recht

ANU

University of Ottawa Jacob J. Krich

Nanosecond laser melting Ion implantation

Damaged Silicon M+ M+ M+ M+ M+ M+ Molten Silicon

Slide from Dan Recht

ANU Benét Labs New York

University of Ottawa Jacob J. Krich

Nanosecond laser melting Rapid epitaxial solidification Ion implantation

Damaged Silicon M+ M+ M+ M+ M+ M+ Hyperdoped Silicon Molten Silicon

Slide from Dan Recht

ANU Benét Labs New York

University of Ottawa Jacob J. Krich

Hyperdoping in Silicon

Which materials are promising? What doping to use?

https://commons.wikimedia.org/wiki/File:Modern_Periodic_Table.jpg

Figures of merit

Devices

• Efficiency
• Voc
• Jsc
• FF

Absorber materials

• Band gaps
• Absorptivity α
• Lifetime τ
• Mobility μ

Want:

• Absorber only
• High predictive value

Indicate when to work on devices

n p IB

University of Ottawa Jacob J. Krich

αw ≈ 1

w

n-type p-type IB

A good device has

Krich, Halperin, and Aspuru-Guzik, J App Phys 2012

and

τ t 1

subgap absorptivity

Motivating a figure of merit

Sullivan et al., JAP 2013

Si:S

α

carrier lifetime in IB region transit time through IB region

University of Ottawa Jacob J. Krich

Drift Device

τ t ≈

w

VB CB n-type p-type IB

Vbi ≈ Eg q

E = Vbi w ≈ Eg qw t = w vdrift = w µE ≈ w2q µEg ≈ q µEgα2

Krich, Halperin, and Aspuru-Guzik, J App Phys 2012

νdrift = Eg q µα2τ

University of Ottawa Jacob J. Krich

Comparison of figures of merit

Measurable parameters of IB material alone. Good devices: large ν for both electrons and holes.

Krich, Halperin, and Aspuru-Guzik, J App Phys 2012

νdrift = Eg q µα2τ νdiff = kbT q µα2τ

VB CB CB VB

University of Ottawa Jacob J. Krich

üüüüüüüüüüüüüüüüüüüüüüüü üüüüüüüüüüüüüüüüüüüüüüüü

0.001 0.01 0.1 1 10 100 1000 10 20 30 40 50 60 n Efficiency H%L

„„„„„„„„„„„„„„„„„„„„„„„„ „„„„„„„„„„„„„„„„„„„„„„„„

1 2 3 4 5 6 7 OD

Eg = 1.9 eV ΔE = 0.7 eV full concentration Best without IB IB improves efficiency

∆E Eg

CB VB

ν = kbTα2µτ/q

CB VB

Krich et al., SPIE 2014

= ! w

University of Ottawa Jacob J. Krich

Light trapping

Effectively increase α by 4n2≈47 Increases ν by (4n2)2≈2000

Figure from Yu et al, PNAS 2010

University of Ottawa Jacob J. Krich

Outline

• Intermediate band solar cells

– 3 material classes

• Figure of merit

– Measurements – Predictions

• New developments

– InGaN quantum dots in nanowires – Device model

University of Ottawa Jacob J. Krich

THz measurements of carrier lifetime

Pump

• 50 fs, 400 nm
• Absorbed mostly in

hyperdoped region Probe

• ~ps, ~2 THz
• Transmission measures
• conductivity. Proxy for n

CB 1.1 eV Pump 0.1-0.3 eV Probe substrate hyperdoped layer

University of Ottawa Jacob J. Krich

Sher, Simmons, Krich, Akey, Winkler, Recht, Buonassisi, Aziz, and Lindenberg, APL 2014

3.6E18 cm-3 1.4E19 cm-3 4.0E19 cm-3 1.2E20 cm-3 3.6E20 cm-3 S concentration

fs THz study of trapping times in Si:S and Si:Se

University of Ottawa Jacob J. Krich

Sher, Simmons, Krich, Akey, Winkler, Recht, Buonassisi, Aziz, and Lindenberg, APL 2014

fs THz study of trapping times in Si:S and Si:Se

No evidence of lifetime recovery

University of Ottawa Jacob J. Krich

Sher, Simmons, Krich, Akey, Winkler, Recht, Buonassisi, Aziz, and Lindenberg, APL 2014

ν peaks at lower concentration. Still hyperdoped. Hole figure of merit not measured.

University of Ottawa Jacob J. Krich

GaPyAs1-x-yNx

Heyman et al, PR Applied (2017) α ≈ 2 um-1

τ = 23 ps Eg = 2.1 eV

νdrift = 19

Estimate μ = 1000 cm2/Vs

University of Ottawa Jacob J. Krich

University of Ottawa Jacob J. Krich

Outline

• Intermediate band solar cells

– 3 material classes

• Figure of merit

– Measurements – Predictions

• New developments

– InGaN quantum dots in nanowires – Device model

University of Ottawa Jacob J. Krich

Predicting ν

• S. Li and W. Thurber, Solid State Electronics, 20, 609 (1977)

constant

1 µ = 1 µlatt + 1 µni µni ∝ 1 NI

νdrift = Eg q µτα2

Small : Large : NI ν ∝ NI ν → ν∗ NI

α ∝ NI

τ ∝ 1 NI

University of Ottawa Jacob J. Krich

200 400 600 800 1000 1120 E (meV)

In (A) Ag (D) Zn (A) Fe (D) Zn (dA) Ag (A) S (D) Pt (A) Pt (D) Au (A)

νdrift

* In (A) Ag (D) Pt (D) Zn (A) Fe (D) Au (A) Zn (dA) Ag (A) Ti (D) S (D) Pt (A) 1e−08 1e−06 0.0001 0.01 1 100 10000 1e+06 Electrons f=1/2 Electrons f=fcm Holes f=1/2 Holes f=fcm

With light trapping

Sullivan, Simmons, Buonassisi, Krich, JPV 2015

University of Ottawa Jacob J. Krich

3 Candidate Materials Classes

Quantum dots Highly-mismatched alloys ZnTeO

• δ-doping

InAs QD GaAs Contact Contact p+ n Substrate (n+ GaAs) Wetting layer

Luque et al., Nat Photon 2012 Wang et al., APL 2009

1 2

Hyperdoped silicon

3

• Pan et al, APL (2011)

University of Ottawa Jacob J. Krich

Nguyen et al., Nano Lett 2011 Ross Cheriton, PhD thesis, 2018

Eg = 3.4 eV Eg = 0.7-3.4 eV

University of Ottawa Jacob J. Krich

Ross Cheriton, PhD thesis, 2018

University of Ottawa Jacob J. Krich

Intermediate band device models

Essential to determine requirements for IB absorbers and optimize devices. Strandberg and Reenaas, PiP 2010

– Radiative recombination only. IB region only, depletion approximation.

Yoshida, Okada, Sano, JAP 2012

– No IB transport.

Martí, Cuadra, Luque IEEE TED 2002

– Diffusive only. Radiative recombination only.

Many detailed-balance based models

University of Ottawa Jacob J. Krich

Our device model

• 2D
• Finite element –

built using FEniCS

• Benchmark against

Synopsys Sentaurus

<10-3 deviation

University of Ottawa Jacob J. Krich

Our device model

Better reverse-bias convergence with 64 bits than Sentaurus at 128 bits

University of Ottawa Jacob J. Krich Eg = 2 eV Ei = 1.2 eV !I = 20 cm2/Vs

p n IB

dark

University of Ottawa Jacob J. Krich

• Model released within the year
• Enable better devices and

understanding of IB materials

Eg = 2 eV Ei = 1.2 eV !I = 20 cm2/Vs

p n IB

1-sun

14% 20% 31%

University of Ottawa Jacob J. Krich

Recap

Intermediate band solar cells

– Great potential – Need sufficient absorber materials

Figure of merit

– guide materials development – determine when to make a device

Device modeling

– Required to optimize device performance