role of surface chemistry in photovoltaics · -4 vac-3-2-1 0 2 Δevac=0.24ev znte 2.24ev e vb...
TRANSCRIPT
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
Department of Materials Science & Engineering 2 University of Delaware
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
Department of Materials Science & Engineering 3 University of Delaware
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.
Department of Materials Science & Engineering 4 University of Delaware
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.
Department of Materials Science & Engineering 5 University of Delaware
II-VI ZnTe/ZnSe Thin Film Solar Cell Structures & Band
Alignment
Department of Materials Science & Engineering 6 University of Delaware
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
Department of Materials Science & Engineering 8 University of Delaware
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
Department of Materials Science & Engineering 9 University of Delaware
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 )
Department of Materials Science & Engineering 10 University of Delaware
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.
Department of Materials Science & Engineering 11 University of Delaware
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
Department of Materials Science & Engineering 12 University of Delaware
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
Department of Materials Science & Engineering 13 University of Delaware
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
Department of Materials Science & Engineering 14 University of Delaware
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.
Department of Materials Science & Engineering University of Delaware 15
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
Department of Materials Science & Engineering University of Delaware 16
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
Department of Materials Science & Engineering University of Delaware 17
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
Department of Materials Science & Engineering University of Delaware 18
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
Department of Materials Science & Engineering 19 University of Delaware
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
Department of Materials Science & Engineering 20 University of Delaware
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
Department of Materials Science & Engineering 21 University of Delaware
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
Department of Materials Science & Engineering University of Delaware 22
Roadmap for silicon devices
Department of Materials Science & Engineering University of Delaware 23
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
Department of Materials Science & Engineering University of Delaware 24
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
Department of Materials Science & Engineering University of Delaware 25
Roadmap for silicon devices
Department of Materials Science & Engineering University of Delaware 26
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
Department of Materials Science & Engineering University of Delaware 27
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
Department of Materials Science & Engineering University of Delaware 28
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
Department of Materials Science & Engineering University of Delaware 29
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
Lifetime (IN solution)
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)
Department of Materials Science & Engineering University of Delaware 30
• 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
Department of Materials Science & Engineering University of Delaware 31
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!!
Department of Materials Science & Engineering University of Delaware 32
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.
Department of Materials Science & Engineering University of Delaware 33
XPS Studies Si substrates after QYH/Me
Chhabra et al, Phys. Status Solidi A, DOI: 10.1002/pssa.201026101
Department of Materials Science & Engineering University of Delaware 34
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
Lifetime (IN solution)
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
Department of Materials Science & Engineering University of Delaware 36
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
Department of Materials Science & Engineering University of Delaware 37
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
Department of Materials Science & Engineering University of Delaware 38
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
Department of Materials Science & Engineering University of Delaware 39
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)
Department of Materials Science & Engineering University of Delaware 40
Department of Materials Science & Engineering University of Delaware 41
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)?
Department of Materials Science & Engineering University of Delaware 42
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
Department of Materials Science & Engineering University of Delaware 43
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!!!