Role of Surface Chemistry in Photovoltaics · -4 vac-3-2-1 0 2 ΔEvac=0.24eV ZnTe 2.24eV E VB 2.68eV As-deposited Surface ZnSe Interface 1 (eV) Acceptor Donor Max V OC Fermi =1.44eV

Department of Materials Science & Engineering 1 University of Delaware

Role of Surface Chemistry in Photovoltaics

R. L. Opila1,2,3,4, F. Fang1, L. L. Costello1, B. E. McCandless2, N. Kotulak4, D. Yang1, A. Teplyakov3, F. Tian3

1) Department of Materials Science and Engineering, University of Delaware 2) Institute of Energy Conversion, University of Delaware, Newark, DE 3) Department of Chemistry and Biochemistry, University of Delaware

4) Department of Electrical and Computer Engineering, University of Delaware Outline

I.  Moore’s Law for Photovoltaics

II.  Energy Bands in Heterojunctions

III. Surface Recombination in Silicon Solar Cells


Sustainability of Electricity l  Increasing sustainability requires significant PhotoVoltaic

generation capacity within a decade l  While PV has realized compound growth rates > 40% for a

decade, it presently accounts for <1% of electrical generating capacity

l  Can PV reach TeraWatt production capacity in a decade – is there a Moore’s Law analog to PV?

Need to redo fig

�  All new US electricity needs met by PV – within 5 years

�  New world electricity needs met by PV – within about 10 years

�  US electricity needs met by PV – within 15 years

� World electricity needs met by PV – within 20 years

Example: 40% growth rates


Moore’s Law Analog for PV •  Integrated circuits have had a transforma4onal impact by increasing performance, reducing costs and integra4ng electronic circuitry into a wide range of systems, allowing rapid expansion in their use.

•  Moore’s Law expressed the sustained, rapid growth of transistor. •  Staying on Moore’s Law requires focused, integrated projects coupled with scien4fic and technological innova4on driven by a roadmap– e.g., voltage scaling, photolithography, high k-‐dielectrics, etc.


QESST Approach l  Improve efficiency, manufacturability and $/kWh for silicon, thin

film, and tandem solar cells l  Efficiency in a manufacturable process is key metric

l  Develop advanced approaches which increase efficiency and are compatible with existing production

l  Integrate with other components to increase functionality, performance and enable new applications.


II-VI ZnTe/ZnSe Thin Film Solar Cell Structures & Band

Alignment


Heterojunction --- Band alignment at interface

Energy conversion efficiency is the percentage of incident energy of sunlight or heat that actually ends up as electric power

Carrier transport behavior at interface severely affects the device performance

Suitable band offset at interface is critical for carrier transport

P

Band Bending

Egn

Egp

ϕpχp

VBi

ΔEC

Ef

EVn

ECn

ϕnχn

Vacuum Level

Ef EVp

ECp

N

Department of Materials Science & Engineering University of Delaware 7

ZnTe-based Solar Cell

A promising II-VI material for TFSC:

Ø  Wide direct band gap (EG>2 eV)

Ø  Tunable bandgap via alloying

Ø  Nontoxic materials Ø  Various growth techniques

High diffusion voltage at ZnSe/ZnTe heterojunction, VOC > 1V is feasible

F. Buch, et al, J. Appl. Phys., 48 (4), 1977


ZnTe-based Solar Cell

A promising II-VI material for TFSC:

Ø  Wide direct band gap (EG 2.24 eV)

Ø  Tunable bandgap via alloying

Ø  Nontoxic materials Ø  Various growth techniques

High diffusion voltage at ZnSe/ZnTe heterojunction, VOC > 1V is feasible

ZnSe/ZnTe thin film device structure

Terminals

Transparent Conducting Oxide (ITO)

7059Borosilicate Glass

ZnSe Window Layer

ZnTe Absorber Layer

F r o n t Contact

Incident Light

n(-)

p(+) Metal/Graphite back contact


Film growth --- Close Space Sublimation/ Vapor transport

Thin Film

Source T (°C)

Sub T (°C)

Growth Rate

(nm/s)

ZnSe 730 575 0.1

ZnTe 670 575 2.5

CSS growth

Low cost --- No HV required Reliability & Reproducibility

Scalable (compatible to roll-roll processing)

Device diagram of a CSS system

Thermo couple

Sub CdS CdTe CdCl2 Back Heat Contact

Completed Device Glass Substrate

Vacuum Boundary

Continuous Conveyor

Air to Vacuum to Air Seal

schematic of process

Abound Solar (formerly known as AVA Solar) Production Prototype (National CdTe Team Meeting April 5 and 6, 2005 )


SEM cross-section image

4 um ZnTe on ZnSe/ ITO/BSG glass

ZnTe ZnSe ITO

Energy-dispersive X-ray spectroscopy (EDS) confirmed the chemical composition of distinctive films.


Photoemission Spectroscopy EB = hv – EK – Φ

/UV-light

/UV-light

http://en.wikipedia.org/wiki/File:ARPESgeneral.png

Surface sensitive technique: Quantitative analysis of small chemical shifts depending on the chemical environment of the atom which is ionized, allowing chemical structure and chemical identification to be determined.

590 585 580 575 570

(b)

Sb2Te3 film as-deposited

Sb2Te3

Te oxideSb2Te3

Te oxide

Te 3d5/2Te 3d3/2

Te 4d

Binding Energy /eV


XPS Measurements on ZnSe Surface X-Ray photoelectron spectroscopy using Al-Kα source measured core level for ZnSe film surface, confirmed chemical compositions of the film surfaces. High-resolution surface science system PHI 5600

Probed oxide on ZnSe surface Binding energy of Se 3d core level for Se2- in ZnSe is 54.1eV as for Se4+ in SeO2 at 59eV

1040 1035 1030 1025 1020 101570 65 60 55 50

ZnSe Core level ZnSe: As-deposited

Binding Energy (eV)

Inte

nsity

(arb

. uni

ts) ZnSe

Se 3d

SeO2

ZnSeZn 2p3/2


Surface oxide removed by sputtering or chemical etch

Prior Treatment of XPS measurement •  Ar+ ion Sputtering •  Wet chemical etching

•  40 % H2SO4 at 50 ºC •  1 % HCl RT

could successfully remove the oxide.

70 65 60 55 50

Binding Energy (eV)

Inte

nsity

(arb

. uni

ts)

As-deposited film

After 10 min Ar+ion sputtering

40%H2SO4 at 50 degree C

ZnSe XPS Spectra Se 3dNormalized to As-deposited


Elemental Se residual observed after etching--- Post annealing

56 54 52 50 48 46 44

Annealed at 150 degree C Etched by sulfuric acid Decomposed to elemental

Se and ZnSe peak

Se 3d 3/2ZnSe

Se 3d

Inte

nsity

/ au

Binding Energy eV

Elemental Se 3d

Elemental Se is observed after etching; Same mild anneal in-situ: inside the UHV photoemission chamber desorb it.


Fermi Level Alignment l  UPS measurement of Valence Band Maximum (VBM) at Au film

surface → Fermi Level

l  Aligned with all the other VBM measurements l  Assume all Fermi levels are aligned

-5 -10

0

100

200

300

400

500

Inte

nsity

/ au

B.E eV

Au Fermi @140V beamSSRL 8-1 01/07/2009

[-6.45+(-6.66)]/2= - 6.55 eVBias -4.96 eVAu Fermi level= -1.59 eV


10 8 6 4 2 0

ZnSe VB

Ar+ Sputtered

As-deposited

Binding Energy (eV)

Inte

nsity

(arb

. uni

ts)

UPS Measurements

Synchrotron radiation at Brookhaven National Laboratory and SSRL

Valence band (VB) structures for ZnSe and ZnTe films Surface oxide altered valence band of ZnSe

Extrapolation of the rising edges give the valence-band maximum:

10 8 6 4 2 0

Ar+ Sputtered

ZnTe VB

ZnSe VB

As-deposited

Binding Energy (eV)

Inte

nsity

(arb

. uni

ts)

10 8 6 4 2 0

EtchedZnSe VB

As-deposited ZnSe VB


Energy Band Diagram & Device VOC and JSC

Evac

ECB p

n E

VB

ECB

Evac-4

-3

-2

-1

0

2

ΔEvac=0.24eV

ZnTe

2.24eV

EVB

2.68eV

As-deposited Surface

ZnSe Interface

1

(eV)Acceptor Donor

Max VOC

=1.44eVFermiLevel

As-deposited ZnSe

Diagram based on VB, Ef (UPS) and EG (optical absorption)

3

pn

EVB

ECB

Evac-4

-3

-2

-1

0

2

ΔEvac=1.57eV

ZnTe

2.24eV

EVB

ECB

Evac

2.68eV

Ar+ ionSputtered Bulk

ΔEVB=1 eV

ZnSe Interface

1

(eV)

Acceptor Donor

Max VOC

=1.68eV

Sputtered ZnSe


Energy Band Diagram & Device VOC and JSC

ZnSe Preparation Voc (mV) JSC(mA/cm2)

None 450 <5 Rapid transfer

(<1 min) 600 ~5 Chemical

Etched 750 >5

Solar cell parameters AM1.5 Illumination at 25°C

3

pn

EVB

ECB

Evac-4

-3

-2

-1

0

2

ΔEvac=1.57eV

ZnTe

2.24eV

EVB

ECB

Evac

2.68eV

Ar+ ionSputtered Bulk

ΔEVB=1 eV

ZnSe Interface

1

(eV)

Acceptor Donor

Max VOC

=1.68eV


Direct measurement of CBM --- Inverse photoemission Inverse photoemission is the time-reversed process of photoemission, complementary technique to measure the conduction band edge directly.

PE EB = hv – EK IPE E (above EF) = hv – EK

http://www.physics.rutgers.edu/%7Ebart/grouphome/PE_IPE_RAB.htm


IPES Schematic Diagram

Low energy electron gun (5-30 eV)

Position sensitive MCP detector (Micro-channel plate detector )

Concave spherical diffraction grating

Main blocks: UHV system / <10-7 Torr) • Source: electron gun • Diffraction grating • Position sensitive detector • Analysis programming


Thin Film Solar Cell Summary Ø  CSS & evaporation of sequential ZnSe/ZnTe growth demonstrated:

Ø  Significant chemical oxidation and sensitivity to storage of ZnSe

l  Oxidation state observed by PES l  UPS using synchrotron radiation detected band structure offsets

changed by oxidation l  Robust device baseline → minimize exposure of films to oxygen

Ø  Film structure characterization: Optimize CSS grown ZnSe film → diminish lateral facets Refine the CSS equipment and enable better control of the ZnTe growth Increase grain size of ZnSe and ZnTe in evaporation growth; Recrystallization →post annealling


Roadmap for silicon devices


0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.51.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

Bottom Cell Bandgap (eV)

Top

Cel

l Ban

dgap

(eV)

35

35

35

35

35

35

35

35

38

38

3838

38

38

38

40

40

40

40

Dual-Junction “Limit” Results l  III-V/Si Device

l  Unconstrained l  1.72eV/1.12eV l  42% Efficiency

l  III-V/SiGe Device l  Lattice-matched

l  1.58eV/0.84eV GaAsP/SiGe

l  40% Efficiency

GaInP/SiGe Lattice-Match

GaAsP/SiGe Lattice-Match


GaAsP/SiGe Tandem Device Schmeider, Diaz, Barnett, Veeco, Amber Wave l  Improved TJs, Window layers, Integration (Martin, Brianna, ….)

l  Voc as high as 1.32 V l  Efficiency as high as 15.2% (AR corrected) l  JSC-VOC FF = 78% l  Top cell ideality factor = 1.84 (Assumes n=1 in SiGe)

l  Bottom cell not optimized (Xin, Dun, Anastasia, ….) l  Low bottom cell current—Ge:Si not optimized

Si Substrate

SiGe Graded Buffer

SiGe Solar Cell

III-V Nucleation

GaAsP TJs

GaInP BSF

GaAsP Solar Cell GaInP Window

GaAsP Contact

Generation 3

Solar Cell

VOC = 1.30 V

JSC = 12mA/cm2

FF = 70.5%

Efficiency = 11%

1cm2 Device


Roadmap for silicon devices


Recombination Processes ð  Process where an electron & a hole annihilate

ð  Both Carriers disappear - no collection!!

ð  Recombination types

v  Radiative v  Trap-Assisted v  Auger v  Surface v  Emitter

-

+ Photon

CB

VB

BULK

ð  How good are the surfaces?

v  Minority carrier lifetime (µs or ms)

v  Surface recombination velocity (S in cm/sec) WS

beff

211+=

ττ


Moderate-to-high Temp. Techs. v  a-Si (PECVD at 200°C)

v  SiN (PECVD at 200-350°C)

v  SiO2 (Diff. Furnaces at 800-900°C)

PECVD for a-Si & SiN SiO2 passivation

ROOM Temp Techs. v  Hydrogen Fluoride (HF)

v  Quinhydrone-Methanol (QHY/ME)

v  Iodine-Methanol (I2/ME)

Si Passivation Schemes

QHY/ME passivation

Surface recombination is controlled by growing a passivation


Why QHY-ME? ð  Easy to use

ð  Low cost

ð  Important characterization tool

ð  Room temperature operation

ð  Reversible

ð  Ideal passivation if stable

ð  QHY-ME = 0.01 mol/L

ð  Wafer cleaning: Piranha & HF

ð  Wafers in solution in acid resistant plastic bag

ð  Passivation time - 1 hour at room temp

ð  Measure lifetime/ Implied-Voc/ S

Procedure

Limitations Lifetime not stable if sample exposed to air

QuinHYdrone/MEthanol Passivation

O

O p-benzoquinone

OH

OH

hydroquinone

+

Quinhydrone Structure


Injection Level (cm-3)EffectiveLifetime(ms)

1014 1015 1016

0

1

2In solution (1 hr)Out of solution (20 mins)Out of solution (1 day)Out of solution (3 days)

p-type

3.3 ms at 1x1015

Lifetime Results (QHY/ME)

Lifetime (IN solution)

3.3 ms; 7 cm/s

Lifetime (20 mins OUT of solution)

2.7 ms; 8.7 cm/s

1.1 ms at 1x1015

p-type Si <100>, 3 ohm-cm, 170 µm


1.1 ms; 21 cm/s

Lifetime (20 mins OUT of solution)

0.436 ms; 53 cm/s

Chhabra et al, Appl. Phys. Lett., 96 (2010) 063502

n-type Si <100>, 100 ohm-cm, 460 µm

(20 mins)


•  hydrocarbons •  metallic impurities •  silicon oxide layer •  ionic contamination •  particles

Alternative wet cleaning method: Hot H2O2 –H2SO4 solution (Piranha clean)

Cleans heavily contaminated Si wafers. Dilute HF/water solution.

Premium RCA clean: SC-1: NH4OH, H2O2 , and H2O at typically 80 C for 10 min

Removal of hydrocarbons and may cause oxidation and metal contamination. Immersion in HF in H2O at 25 C

Removal of oxide and the ions dissolved in the oxide. SC-2: HCl, H2O2 , and H2O at 80 C.

Removes the remaining metallic contamination. Final HF rinse to remove oxide.

Hydrogen Passivation of Silicon


F. Tian, D. Yang, R. Opila and A. Teplyakov, Appl. Surf. Sci., 258, 3019-3026 (2012).

a)  RCA method; b)  HF-dip method with a float

zone crystal c)  HF-dip method with n-doped

crystal.

Infrared spectra of the Si-H stretching region

of H-terminated Si(111) surfaces

RCA clean has better surface chemistry and morphology (AFM)

Surface Preparation Charge-carrier Surface Recombination Lifetime, τ (µs) Velocity, S (cm s-1)

RCA 60.7+28.9 1002.7+577.3 HF-dip 169.9+14.0 295.3+24.4

But worse lifetimes!!


Chemical and electrical passiva:on of Si(111) surfaces

F. Tian, D. Yang, R. Opila and A. Teplyakov, Appl. Surf. Sci., 258, 3019-3026 (2012).

IR investigation of the C-H stretching spectral region of the 1-decene modified Si (111) surface produced by a) RCA method and b) HF-dip procedure.

IR investigation of the C-H stretching spectral region of the 1-octadecene modified Si (111) surface produced by a) RCA method and b) HF-dip procedure.


XPS Studies Si substrates after QYH/Me

Chhabra et al, Phys. Status Solidi A, DOI: 10.1002/pssa.201026101


H2O

Air

OH

Si

O

IN Solution OUT of Solution OH

O H

Si

OH

HF termination

Si

H H H H H H

Si

Oxide termination

OH O O OH OH

QHY/ME treated

Si

R H R H H H

First Proposed Surface Bonding from XPS


3.3 ms

Unstable

Lifetime (24 hrs OUT of solution)

1.9 ms

Department of Materials Science & Engineering University of Delaware 35 35

Recombination center

3.3ms 0

500

1000

1500

2000

2500

Life:me in solu:on at 1hr

QHY BQ HQ DDQ H:Si

DDQ


0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350

Life:m

e/us

Time/hr

Bare Si wafer in BQ/Me 0

40

80

120

160

0 10 20 30 40 Time/min


172.5

2438

2054

13.5

751

<10 <10 0

500

1000

1500

2000

2500

3000

H-‐Si QHY BQ HQ DDQ DMBQ Acetylacetone

Life4me/us

2,3-‐Dichloro-‐5,6-‐dicyano-‐1,4-‐benzoquinone (DDQ)

2,6-‐dimethoxy-‐(1,4)-‐benzoquinone(DMBQ)

Acetylacetone


XPS

Si

SiOx

Si0 Si0 Si0

38

C C-O

C-O

C-O

C-C C-C

C-C

Si

H H H

Si Si


Surface passivation of Si with hydroquinone l  Increases carrier lifetime dramatically. l  Slows oxidation rate compared to Hydrogen passivation l  How

l  What reaction? l  Density functional theory says BQ will not react with H-terminated

Si, but will react with Si(111)7x7. What defects are important? (D. Okeeva)

l  BQ is photoactive—is light from Sinton test important in observed passivity? (L. Costello, M. Chen)

l  Are protons in MeOH important in passivity (L. Costello)?

l  Is passivating site a charge center? l  Can we generalize it?

l  How do we make organic/Si induced junctions (N. Kotulak, ASU) l  Can organic passivation of Si nanowires be improved (Cal Tech)



Surface Photovoltage with Prof. Sefik Suzer, Bilkent University

n-‐Si

Si 2p

n-‐Si/SiOx

p-‐Si

Si 2pO 1s

Si4+Si0

Laser ON

p-‐Si/SiOx

Laser OFF

Time

106 104 102 100 538 536 534 532 530

-‐0.19 eV

0.18 eV

102 100 98

+0.11 eV

Can induce shifts in surface potential (binding energy) by up to 0.5 eV with laser

•  Can this induce changes in contact angle?

•  Different surface chemistry (redox chemistry)?


Lifetime tester, data Kotulak, Costello, Chen

0

500

1000

1500

2000

2500

0 5 10 15 20 25 30

Life

time

/ us

Passivation Time / min

p-type, methanol

control

p-benzoquinone

hydroquinone

0 500

1000 1500 2000 2500

0 5 10 15 20 25 30

Life

time,

us

Passivation Time / min

p-type, ether

0 500

1000 1500 2000 2500 3000 3500 4000

0 5 10 15 20 25 30

Life

time

/ us

Passivation time / min

n-type, methanol

0

1000

2000

3000

4000

0 5 10 15 20 25 30 Li

fetim

e, u

s

Passivation time /min

n-type, ether

Not simple redox

Protons are important


Christiana Honsberg, ASU Stuart Bowden, ASU Allen Barnett, UNSW Conan Weiland, NIST/NSLS Bhumika Chhabra, Micron Sefik Suzer, Bilkent University Darika Okeeva, Bilkent University Ulrike Salzner, Bilkent University Funding:

NSF-IGERT DOE/NSF-ERC University of Delaware DARPA, DARPA Fulbright Foundation

Acknowledgments

Surface chemistry is crucial in next generation solar cells!!!

Documents

Role of Surface Chemistry in Photovoltaics · -4 vac-3-2-1 0 2 ΔEvac=0.24eV ZnTe 2.24eV E VB 2.68eV As-deposited Surface ZnSe Interface 1 (eV) Acceptor Donor Max V OC Fermi =1.44eV