Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

InAs/GaAs1-xSbx quantum dots for applications in intermediate band solar cells

Yang Cheng, A. J. Meleco, M. Fukuda, V. R. Whiteside, M. C. Debnath, P. J.

Vallely, A. J. Roeth, T. D. Mishima, M. B. Santos, and I. R. Sellers

Homer L. Dodge Department of Physics & Astronomy, University of Oklahoma, Norman, OK, USA

K. Hossain Amethyst Research Inc., Ardmore, OK, United States

S. Hatch and H-Y. LiuDepartment of Electrical & Electronic Engineering, University College London, London, United Kingdom

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Intermediate Band Solar Cells

2

• Allows harvesting photons with energy below band gap

• Retains Voc while increasing Jsc

• QDs proposed as a candidate system: discrete DOS/artificial atoms

• InAs/GaAs QDs well developed system: QD lasers, LEDs, QDSCs

A. Luque and A. Martí, Advanced Materials, 22: 160–174. (2010) H. Y. Liu, I. R. Sellers et al., Appl. Phys. Lett. 85, 704 (2004)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

InAs/GaAsSb QD IBSCs

3

M. Y. Levy, and C. Honsberg, IEEE Transactions on Electron

Devices 55 706-711 (2008)

• Simple IB formation, improved hole

extraction

• Increased carrier lifetime

• For the proposed system increased

QD density

• Less well-developed than InAs QDs

• Materials issues/still in development

Chiu (2002), Chang (2008), Tomic (2013), Nishikawa (2012), Hatch

(2014), Tang(2015), Yang (2016)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Power Dependence of Peak PL

Growth and Analysis

4

• QD density (3.5×1011cm-2) has been

achieved, one order of magnitude higher

than InAs QDs on GaAs

• Band alignment transition from type-I to

type-II is observed by varying Sb

composition in the GaAsSb matrix.

Diagram of

InAs/GaAsSb Structure

1 μm2 AFM of InAs QDs

(3 MLs) on GaAsSb

Y. Cheng et al., Solar Energy Materials and Solar Cells 147 94 (2016)

Band Diagram of InAs/GaAsSb StructureNormalized PL of Varying

Sb Compositions

M. C. Debnath et al., Journal of Applied Physics 119, 114301 (2016)

10%12% 16% 18%

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

PL, EQE, J-V measurements

Solar cell Structure

Y. Cheng et al., Solar Energy Materials and Solar Cells 147 94 (2016)

-7.5

-5.0

-2.5

0.0

-0.3 0.0 0.3 0.6 0.9 1.2

-7.5

-5.0

-2.5

0.0

77K EQE

Rela

tiv

e E

QE

Cu

rre

nt

De

nsit

y (

mA

/cm

2)

77K Dark

77K Light

400 600 800 1000 1200

230K EQE

Rela

tiv

e E

QE

Wavelength (nm)

Cu

rren

t D

en

sit

y (

mA

/cm

2)

230K Dark

230K Light

Voltage (V)

101

102

103

104

75K PL

PL

In

ten

sit

y (

arb

. u

)

101

102

103

230K PL

PL

In

ten

sit

y (

arb

. u

)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Temperature dependent PL and EQE measurements

10-1

100

101

102

QDs (IB-VB)

GaAsSb

300K90K

Re

lati

ve

EQ

E (

%)

GaAs

800 900 1000 1100 1200

101

102

103

104

210K

75K

Inte

ns

ity

(a

rb. u

)

• GaAs (red arrow) EQE peak shifts

horizontally.

• GaA0.86Sb0.14 (green arrow) EQE peak shifts

to the longer wavelength but also becomes

shallower.

• EQE related to the InAs QDs also has a red

shift, but with an additional enhancement.

• At low temperature, the EQE related to QDs

seems to be dominated by radiative

recombination.

• Reduction of the radiative recombination (PL)

results in the EQE enhancement.

Y. Cheng et al., Solar Energy Materials and Solar Cells 147 94 (2016)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Temperature dependent J-V measurements

7

• Large Voc drop observed coupled with rising dark saturation current

• Temperature dependent Jdark shows a barrier to transport that decreases with increasing

temperature

• “S-shaped” Jsc behavior from non-radiative recombination due to defects

Y. Cheng et al., Solar Energy Materials and Solar Cells 147 94 (2016)

100 150 200 250 3005.1

5.4

5.7

6.0

6.3

Total Jsc from EQE

Jsc from GaAs

Cu

rre

nt

De

ns

ity

(m

A/c

m2)

Temperature (K)

0.2

0.3

0.4

0.5 Jsc from GaAsSb+ QDs

Cu

rre

nt

De

ns

ity

(m

A/c

m2)

0.0 0.3 0.6 0.9 1.2 1.5

10-3

10-1

101

103

10-37

10-31

10-25

10-19

10-13

10-7

50 100 150 200 250 300 350

Temperature (K)

J0

J0 (

mA

/cm

2)

Jd

ark

(m

A/c

m2)

Voltage (V)50 100 150 200 250 300 350

0.4

0.6

0.8

1.0

Temperature (K)V

oc (

V)

Voc

6.8

6.9

7.0

7.1

Jsc

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Comparison of control and 3 Layer sample at 300K

Dark, Light J-VEQE

400 600 800 1000 120010

-2

10-1

100

101

102

Control

3 Layers

EQ

E (

%)

Wavelength (nm)

Control

3 Layer

-0.2 0.0 0.2 0.4 0.6-10

-5

0

5

Control

3 Layer

J (

mA

/cm

2)

Voltage (V)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Comparison of multilayer QDSCs and control

• All samples show a large temperature dependence in open circuit voltage

• Above 130 K open circuit voltage of 3 and 5 QD samples larger than control

• Voc larger with decreasing QD layers

• Non-conventional JSC again evident in all samples

A. Meleco, Y. Cheng et al., in preparation (2016)

100 200 300 400

6.8

7.0

7.2

7.4

7.6

7.8

Jsc (

mA

/cm

2)

Temperature (K)

Control

3 Layer

5 Layer

7 Layer

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Equivalent Circuit of Model

Single diode model cannot fit dark J-V

1. J01 represents the main junction diode.

2. J02 represents the second diode to achieve a better fitting for the forward bias regime.

3. The third term represents the current due to ohmic shunt and series resistance.

4. the last term represents a non-ohmic (SCLC) leakage current.

Equivalent Circuit

B. L. Williams et al., Progress in Photovoltaics: Res. Appl. 23, 1516 (2015)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Dark Current Analysis/Fits

10-9

10-7

10-5

10-3

10-1

-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.51E-9

1E-7

1E-5

1E-3

0.1

-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0

80K

240K

J (

mA

/cm

2)

Data

Fit

140K

J (

mA

/cm

2)

Voltage (V)

300K

Voltage (V)

• Two-diode model produces good fits

although they are less so a lower

temperatures

• Non-ideality of system

(inhomogeneities) also require large

contribution from non-ohmic-space-

charge type effects in reverse bias.

• At higher temperatures a single diode

fit becomes dominant

A. Meleco, Y. Cheng et al., in preparation (2016)

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Dark Saturation Coefficient

10-46

10-36

10-26

10-16

10-6

10-13

10-12

10-11

10-10

50 100 150 200 250 300 35010

-13

10-12

10-11

10-10

J01

(a)

(b)

J (

mA

/cm

2) J02

(c)

Temperature (K)

J01+J02

10-44

10-34

10-24

10-14

10-12

10-11

10-10

50 100 150 200 250 300 350

10-12

10-11

10-10

(f)

(e)

(d)

J01

J02

Temperature (K)

J01+J02

Control 5 Layer

Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html Photovoltaics Materials & Device Group, University of Oklahoma: http://www.nhn.ou.edu/~sellers/group/index.html

Conclusions/Future Work

• Growth conditions for InAs QDs on GaAsSb are optimized. Enhancement of QD density is achieved.

• A transition from type-I to type-II band alignment is observed through power dependent PL measurements.

• Lattice mismatch between GaAs and the matrix contributes to defect formation

• Those defects in the intrinsic region facilitate carrier escape process and dramatically decrease the Voc.

• The escape of electrons compromises the formation of an isolated intermediate band even at low temperature in the present samples

• Phenomenological diode analysis now underway to further elucidate the unusual PV characteristics

• Cross-sectional TEM is important to validate our hypothesis for a defect mediated tunneling model.