�������� �� �������� ��� ��������� �������������� �� � ��� ����

����������________________________________________________________________________________

�������� �������� �� ������ ������������� ���

���� ���� ����� ������

�� ������������ ����� �������� ������

�������

Current Progress in FutureOpportunities for Thin Film Solar Cells

Satyen K. DebDirector, Basic Sciences Center

National Renewable Energy LaboratoryGolden, CO 80401

Presented at the Workshop on Physics for Energy SourcesInternational Center for Theoretical Physics (ICTP)

Trieste, ItalyOctober 17–29, 2005

034016345

Photovoltaics is Solar Electricity

Good for oureconomy and

energyindependence

Good for our environment

Clean and abundantenergy for the21st Century

DOE PV Program Goal:U.S. leadership intechnology, industry,and markets

High-technologymanufacturing

jobs

02803203

Solar can supply all electricity for the U.S.using this area (100 x 100 mi.) in the SW

OR

Distributed applicationsthroughout the U.S. (vacantland, building-integrated, etc.)

World PV Cell/Module Production (MW)

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 20010

200

400

600

800

Rest of world

Europe

Japan

U.S.

2002

1000

1200

761.1

2003

40.2 46.5 55.4 60.1 69.4 77.6 88.6125.8

154.9201.3

287.7

390.5

561.8

2004

57.9

1195.4

Source: Paul Maycock, PV News, March 2005

034016413

PV Module Production Experience(or Learning) Curve

0.1

1.0

10.0

100.0

0 1 10 100 1,000 10,000 100,000

Cumulative Production (MWp)

PV

Mo

du

leP

rice

(2003$/W

p)

1976

2003

90%

80%

70%

75 GW2020 @25% growth

New, unofficial,

thin film learning

curve starting at

lower price and

volume, but same

slopeBINGO!

80% Learning Curve: Moduleprice decreases by 20% for everydoubling of cumulative production

Why Thin Films ?

• Substantial Cost Advantage

• Lower Consumption of Materials

• Ease of Manufacturing Large Area Devices

• Fewer Processing Steps

• Wider Selection of Materials

• Easier integration of Monolithic Devices

• Greater Tolerance on Materials Quality

03514202

Thin-Film Solar Cells, Y. Hamakawa, Springer

Absorption Coefficient of Chalcopyrite CompoundsTogether with Other Semiconductors Applied in PV

1.0 1.5 2.0 2.5

105

104

103

102

101

Abso

rptio

n c

onst

ant (c

m-1

)

Photon energy (eV)

CuInSe2

InP

GaAs

c-Si

Cu2S

CuInS2a-Si:H 10-5

10-4

10-3

10-2

10-1

Abso

rptio

n le

ngth

(cm)

034016356

Thin Film Solar Cells - Present & Future

(1) First Generation Thin Film Solar Cells

• Amorphous Silicon Alloys

• CdS/CdTe Thin Films

• CIGS/CdS Thin Films

(2) Next Generation Thin Film Solar Cells

• Dye-sensitized TiO2 Thin Film

• Crystalline Si Thin Films

• Microcrystalline Si

• GaAs Thin Film

• Organic Solar Cells

• Novel Ternary and Multinary Compound

(3) Novel Concepts for High Efficiency Devices

03250202

Amorphous Si:H Solar Cell

h�

Triple-Junction a-Si:H Solar Cell: (a) Substrateand (b) Superstrate Configuration

Glass

SnO2:(In, F, …) ITO

ZnO:(Al, Ga, …)p �c-Si:Hi a-SiC:Hn Si:Hp �c-Si:H

i a-SiGe:H

n Si:H

i a-SiGe:H

n Si:HZnO:(Al, Ga, …)

Al, Ag Stainlesssteel

p �c-Si:H

h�

Pro

cess

ing s

equence

Pro

cess

ing s

equence

(a)

(b)

034016382

Triple-Junction Cell Structure

03250204

Deposition Recipes for MW-PCVD and RF-PCVD

Recipe for the

low pressure MW-PCVD

MW power: 100~1000 W

Pressure: 0.1~30 mTorr

TS: 200~400°C

Deposition rate: >4.0 nm/s

Source gases: SiH4, GeH4, and H2

Thin-Film Solar Cells, Y. Hamakawa, Springer

034016358

Recipe for the RF-PCVD

RF power: 0.1~10 Torr

Pressure: 0.1~30 mTorr

TS: 100~400°C

Deposition rate: >0.1 nm/s

Source gases: SiH4, H2, PH3, and BF3

Stabilized Efficiency of a Few Representativea-Si:H-Based Solar Cells and Modules

034016383

Efficiency status: Cell 13.0(stabilized) Submodule 10.4

Module 7–8Commercial 5–7

• Engineered “solution” for degradation: thinabsorber layers and multijunctions

• Extensive fundamental research, leveragedby many other applications

Thin-Film Amorphous Silicon PV—Progress and Status

Key companies: BP Solar, United Solar/ ECD, EPV, IowaThin Films; Sanyo, Kaneka; Phototronics, DunaSolar

• Glass, stainless steel, plastic substrates

• Multi-MW/year in consumer products

• 5 and 10 MW factories for power products operational;many tens of MW in near term

• Unique products for building integration (e.g., roofing,cladding, semi-transparent canopies)

United Solar Systems Corp.

BP Solar

United Solar

03250203

� Improved fundamental understanding:– Metastability (e.g., hydrogen collision

modeland kinetics)

– Role of hydrogen– Alloys with Ge, C, …– Characterization techniques

� Improved cell/module efficiencies; newdevice structures

� Long-term field performance

� Manufacturing throughput and yield — impacton equipment cost

� Novel growth techniques– e.g., hot-wire deposition, VHF plasma– Gas-phase chemistry and control– Nucleation and growth– High-rate deposition (10–100 vs. 1–3 Å/s)– Amorphous to microcrystalline

structures– Low-bandgap (~1 eV) materials

Thin-Film Amorphous Silicon PV—Research Issues and Directions

Advantages of HW-CVD

• Extremely high deposition rates

• High gas utilization

• Better control of [H] in films• More stable films• Lower H2:SiH4 to get �c-Si inclusions• Wide parameter window for quality films

03250205

Key Issues for Efficiency Improvement of a-Si Solar Cells

Physical Process

Efficient guidance of optical energy

Efficiently guided photon confinement

Carrier confinement

Reduction of photogenerated carrier

recombination

Reduction of voltage factor losses

Reduction of series resistance losses

Technical Solution

• Antireflection coating (ARC)

• Multi-energy-gap stacked juntion

• Textured surface treatment

• Use of back-surface-reflection (BSR) effect

• Refractive index arrangement

• Minority carrier mirror effect by heterojunction• Increase of ��-product in the PV active layer

• Film quality improvement by controlling the deposition condition such as

RH, Ts, RF-frequency

• Drift-type effect with p-i-n junction

• Graded-gap PV active layer (bandgap profiling)

• Graded impurity-doping involving back surface field (BSF) effect

• Band profile control of the PV active layer

• Insertion of proper buffer layer in the interface of the p-i and i-n junction

• Optimum design of electrode pattern

• Decrease of transparent conductive oxide (TCO) resistance

• Use of superlattice tunneling junction

Thin-Film Solar Cells, Y. Hamakawa, Springer034016357

Hydrogen Collision Model

Problem: No experimental evidence forlong-range hydrogen motion.

Solution

Stainless Steel Substrate

a-WO3

200 nm

Laser

a-Si:H 1.3 m�

Difference Raman Spectra

200 300 400 500

-35

0

35

70

(c)

(b)

(a)

Diff

ere

nce

Ram

an S

ignal

Raman Shift (cm-1)

(c)

a-WO3 only

(b)

a-Si:H only

(a)

a-WO3/

a-Si:H

Evidence for Hydrogen insertion from a-Si:H to a-WO3

Deb et al. Appl. Phys. Lett. 77, 2686 (2000)

Comparative Stability ofa-Si:H and �c-Si:H Solar Cells

03250206

0.0 0.2 0.4

25

20

15

10

5Curr

entdensi

ty(m

A/c

m2)

Voltage (V)0.6 0.8 1.0

0

AM1, 100 mW/cm2

Voc = 0.905 V

Jsc = 18.8 mA/cm2

Fill factor = 73.6%

Efficiency = 12.5%

Incident light

h�

Glass

i a-SiC

i a-Si

n �c-Si

TCOp �c-Si (C)

p a-SiC

ITOAg

Cell PerformanceCharacteristics

Structure andPerformance of a-Si

Double Heterojunction

Thin-Film Solar Cells, Y. Hamakawa, Springer

034016351

UniSolar Roll-to-roll triple-junction a-Sideposition plant of 30 MW annual capacity

Cadmium Telluride

SnO2

Glass substrate

Back contact

CdTe

CdS

Back contact

CdTe

ZnxCd1-xS/CdS

Zn2SnO4 (ZTO)

Cd2SnO4 (CTO)

Glass substrate

Typical CdTe Device Structure: (a) Conventionaland (b) Modified Version

(a) (b)

Frontcontact

Frontcontact

034016384

CdTe Thin Film Deposition Technologies

• Sublimation-condensation

• Close-spaced sublimation (CSS)

• Chemical spraying

• Electrodeposition

• Screen printing

• Chemical vapor deposition

• Atomic layer epitaxy

• Sputtering

034016372

Apparatus Used for the Deposition of CdTe bythe CSS Technique

034016366

� Efficiency status: Cell 16.5Module 11.0Commercial 7–9

� Many deposition approaches for >10%efficiency

� Early fundamental scientific and engineeringbase for materials and devices

� ES&H issues studied and under control (e.g.,recycling)—Cd perception issue?

Thin-Film Cadmium Telluride PV—Progress and Status

� Key companies: BP Solar, First Solar;Matsushita; Antec Solar

� ~1 MW/year in consumer products for years

� Successful first-time manufacturing underway:– High-rate vapor transport (vacuum)– Electrodeposition (non-vacuum)– Few tens of MW in near term

� Field testing of large power modules (50–90 W)shows promise

BP Solar

First Solar

03250207

� Film deposition development:– Nucleation and growth– Gas-phase and surface chemistries– Annealing and heat treatment (CdCl2)– Grain growth, native defects, dopants– CdS/CdTe interdiffusion– Alternate transparent conductors and

impacts film growth

Thin-Film Cadmium Telluride PV—Research Issues and Directions

� Front and back contacts– Alternate transparent conductors– Low resistance, stable back contacts– Role of Cu; Cu-free contact strategies?

� Close efficiency gap (cell to module)� Compatibility of manufacturing process steps� Low-cost module packaging for long-term

reliability(>20 years); edge sealants and moisture ingress

� Accelerated module test procedures

As-grown film

Annealed in CdCl2at 450ºC, ~20 mins

AFM images of 0.8 �m film —— 1 �m

10 kWBP Solar

03250208

CdTe Research Issues

• Improved contacts to p-CdTe

• Effective p-type doping of CdTe to improve Voc

• CdTe-alloys that allow device design gradients

• Investigation of materials and device properties allowingultra-thin CdTe layers (to 0.25 micron) while maintaininghigh efficiencies

• Reducing tellurium usage by replacement of telluriumwith other elements while maintaining performance

• Materials Availability, Safety,and Environmental Issues

• Closing Gap Between Small and Large Area Devices

Module Structure and Processing SequenceUsed by Solar Cells, Inc.

034016367

03540713

Thin Film CdTeSolar Cell Back Contacts

Metals

– Cu

– Au

– Cu/Au

– Ni

– Ni/Al

– Sb/Al

– Sb/Au

Others

– Graphite – Sb2Te3/Metal

– Graphite (Cu, HgTe, Ni2P)– ZnTe: Cu/Metal

– As2Te3/Metal – ZnTe: N/Metal

– Cu2Te/Metal

– Ni2P/Metal

– NiTe2/Metal

– Te/As2Te3/Metal

First Solar CdTe 25 MWp VaporTransport Manufacturing Line

CIGS Solar Cell

Thin Film CISSolar Cell Structure

I-V Curve for a 19.3%Thin Film CIGS Solar Cell

MgF2/ZnO/CdS/CIGS/Mo/glass

16

12

8

4

0

-4-0.6 -0.2 0.2 0.6

Voltage

Curr

ent(m

A) Jsc = 34.6 mA/cm2

Joc = 0.70 V

FF = 0.796

� = 19.3%

MgF2

ITO/ZnO

CdS

CuInGaSe2

Mo

Glass/SS/polymer/foil

Ni/Al

0.1 �m

0.5–1.5 �m

0.03–0.05 �m

1.5–2.0 �m

0.5–1.5 �m

0.05/3 �m

� = 21.1% (14x) � = 19.3%

034016373

18.8%-Efficient CIGS/CdS/ZnO Solar Cell: (a)Device Structure and (b) Elemental Fluxes and

Substrate Temperature vs Deposition Time

(a) (b)

034016386

03540717

SEM-Thin Film CIGS Solar Cell

Thin-Film Copper Indium Diselenide (CIS)PV—Progress and Status

Key companies: Siemens/Shell Solar, GlobalSolar/ITN, ISET, EPV; Wurth Solar; Showa/Shell

• Prototype production started in 1998:– First commercial products (5–40 W)– Efficient, large modules (>12%)– Expansion to multi-MW in near term

• Field testing of modules shows promise;>10 years outdoors, no degradation

Efficiency status: Cell 19.3Module 12.1Commercial >10

Others: Stainless steel substrate 17.5Electrodeposition 15.4With ZnO (no CdS buffer) 15.7Concentrator (14X) 21.5

� Understanding of film growth, microstructures,defects, and device physics

� Reproducible high-efficiency processes

40.8 kWSiemens Solar

Global Solar

03250209

• Scalability of current processes– Predictive models of materials growth,

devices, and processes– Real-time process controls– Yield and throughput

• New techniques and materials– Non-vacuum approaches– Low-temperature depositions

• Device research and development– Heterojunction vs. homojunction– Role of window materials;

improvements in blue response– Alternate front and back contacts– Higher bandgaps and multijunctions– Device models and characterization

• Theory: Band structures,optoelectronic properties, defectphysics, doping

Electrodeposited CIGSPrecursor Film

Absorber CIGS from Electro-deposited Precursor Film

Thin-Film Copper Indium Diselenide (CIS)PV—Research Issues and Directions

03250210

CuInSe2-alloys Research Issues

• Understanding materials science of complexcompositions, alloys and gradients

• Understanding the complex properties andinteractions of key interfaces

• Investigation of materials and device propertiesallowing ultra-thin CIS layers (to 0.25 micron) whilemaintaining high efficiencies

• Reducing indium usage by replacement of indium withother elements while maintaining performance

• Investigating low-cost processes, and the science ofsuch processes to establish the control and flexibilityneeded to reach high performance and high yield

18

16

14

12

10

Eff

icie

ncy (

%)

1.61.51.41.31.21.11.0

Absorber band gap (eV)

Efficiency vs. CIGS Bandgap

Efficiencies of Cd-Free Buffer Layers inCIGS Solar Cells

034016370

Stability of Thin Film CIS-Based ModulesFabricated by Siemens Solar, Inc.

034016371

03540721

Polycrystalline Thin FilmPhotovoltaic Modules

Organization

BP Solar

Wurth Solar

First Solar

Shell Solar - GmbH

Matsushita Battery

Global Solar

Antec Solar

Shell Solar

Showa Shell

* NREL Confirmed; All aperture-area efficiency

Material

CdTe

CIGS

CdTe

CIGSS

CdTe

CIGS

CdTe

CIGSS

CIGS

Area (cm2)

8390

6507

6612

4938

5413

7714

6633

3626

3600

Eff (%)

11.0*

12.2

10.1*

13.1

11.0

7.3*

7.0

12.8*

12.8

Power (W)

92.5*

79.2

67.1*

64.8

59.0

56.8*

46.7

46.5*

44.15

Date

09/01

05/02

12/01

05/03

05/00

03/02

11/01

03/03

05/03

Dye-Sensitized TiO2 Solar Cell

Dye Sensitized TiO2 Solar Cell

03250216

Dye-sensitized Nano-structureTiO2 Solar Cells

Advantages• Relatively simple and inexpensive fabrication processes with

production cost potential of ~50¢/Wp

• Demonstrated cell efficiency (h�10%) comparable toconventional amorphous silicon solar cell

• Device constituents (TiO2, dye,electrolyte) are abundant andenvironmentally benign

• Optional color and device transparency leads to multiplicity ofproducts and applications

Disadvantages• Use of liquid electrolyte is not an optimum solution

• Very long-term stability of dyes questionable

• Significantly higher efficiency difficult to achieve

03425923

03250215

The First U.S.Patents on Dye-Sentised TiO2

Solar CellsIssued to Deb

etal in 1978

TIC/TIO2

TIC/CdS/TIO2

TIC/CdSe

Relative energydistribution of

solar spectrum

40

30

20

10

0

Qu

an

tum

eff

icie

ncy

(%)

300 400 500

Wavelength (nm)

100

10

1

0.1Q

ua

ntu

me

ffic

ien

cy(%

)

300 380 460

Wavelength (nm)

Bare cell

Bare cellwith NMP+

SpectralSensitization ofTiO2 PEC Cell

03425925

Action Spectra of a Bare Cell andthe Same Cell with NMP+

03250218

Research Issues and Directions forDye-TiO2 Solar Cells

• Dynamics of electron transfer processes

• Surface and interface properties

• Charge transport in TiO2 film and electrolytes

• Role of crystal structure and film morphology

• Electrolyte properties and solid electrolytes

• New Dyes and novel approaches to sensitization

• Efficiency enhancement- multi-junction devices

• Degradation mechanisms

Quantum Dot Sensitized TiO2 Solar Cell

3.2 eV

TiO2

h�

e-

e-e-

e-

e-e-Vph

TransparentTCO

Electrode

electrolyte

PtCounterElectrode

h+

QD

I3-/I-

-Analogous to dye-sensitized TiO2 solar cells-10 to 20 µm film of NC TiO2 (10-30 nm)-Ru dyes � Efficiency ~ 11%

-Advantages of QD’s as sensitizers:-possibility of slowed hot e- cooling-possibility of impact ionization-tunable absorption

TCOelectrode

300 Å TiO2

InP QDs30-60 Å

0342

5905

0342

5906

Schematic diagrams of a dye-sensitizedelectrochromic smart window.B.A. Gregg, Endeavour Vol. 21(2) 1997.

Transmittance spectra of an experimentalsolid-state electrochromic cell in both thebleached and colored states.

Bleached Colored

03425937

Dye-sensitized Solar Cells (DSC)

attractive application

light weight

colorful

sharp cut in production cost

environmentally benign points

NIKKEI 2003.3

Thin Film Si Solar Cell

Calculated Efficiency of Solar Cells with Base Diffusion Length,Ln, and Base Thickness d, Having Very Good Emitters

(Cell B [thin, with back surface field BSF and optical confinement OC] isbetter than cell A [thick, no BSF, no OC], though its diffusion length is lower.)

100 200 300 400

20

18

16

12

�(%

)

d (�m)

14

500 �m

200 �m

100 �m

50 �m

Ln =

0

A

B

034016387

Calculated MACD for SiSolar Cells with DifferentTexture Shapes

Various Surface Structures(a) Random Pyramids (b)Textured Pyramids (c)Inverted Pyramids (d)Perpendicular Slats

(a) (b)

(c)

(d)

AR coatingMedium 1Medium 2Air

Si

034016405

Approaches to Thin-Film PolycrystallineSi Solar Cells on Different Substrates

Objective: To fabricate 10-20 � Si film of sufficient electronic quality withhigh throughput (>1�/min) on low-cost substrates at relatively low processingtemperature.

Approaches

(1) Single-crystal substrates (Cz or Fz growth)

• Epitaxial growth on porous silicon followed by separation by

chemical etching

• Hydrogen implantation in subsurface Si-wafer followed by separation

(demonstrated for 1� Si layer)

• “Epilift” process consisting of deposition of epilayer on

patterned single-crystal substrate

(2) Multicrystalline Si-substrate — metallurgical-grade Si-substrate

(3) Low-cost, non-silicon substrates — glass, ceramic, metals

03250219

STAR Structure

03250220

0.0 0.2 0.4

40

30

20

10

Curr

entdensi

ty(m

A/c

m2)

Voltage (V)

0.60

AM1, 100 mW/cm2

Cell area = 0.16 cm2

Voc = 0.578 V

Jsc = 37.2 mA/cm2

Fill factor = 80.0%

Efficiency = 17.2%

Al contact

p �c-SiC (7.5 nm)ITO (80 nm)

p a-SiC (7.5 nm)

n poly-Si (300 �m)

n �c-Si (20 nm)Al contact

p �c-SiC/n poly-Si heterojunction cell

�c-SiC/Poly-Si Heterojunction Solar Cell andIts Output Characteristics

(Presented by Osaka Univ.)

Thin-Film Solar Cells, Y. Hamakawa, Springer

034016352

Current Status ofThin-Film Si Solar Cell Efficiency*

03250221

Summary of Various TF-Si Solar Cells

034016404

Summary of Various TF-Si Solar Cells

034016403

Future Developments ofThin-Film Si (polycrystalline) Solar Cells

• New approaches to improvements in materials quality (grain size, electron transport, grain-boundary

passivation, etc.) on low-cost substrates

• Breakthroughs in high-throughput growth rates

• Surface morphology and roughness control to achieve optimum light-trapping

• Novel approaches to converting indirect-bandgap Si to direct bandgap (co-doping, quantum confinement, Si:Ge superlattice structure, etc.)

03250222

GaAs Thin-Film Solar Cells

Current Status and Potential Advantages

• Material with optimum energy gap and light absorption characteristics

• High efficienty (~25%) achieved on epitaxially grown GaAs onGe-coated

GaAs single crystal substrate

• High efficiency GaAs thin solar cells fabricated on reusabel substrate

(CLEFT Process)

• Polycrystalline thin film GaAs solar cells (h = 11%) fabricated in early

1980s using low-cost w-coated graphite substrate

03250223

Current-Voltage Characteristics of p+/n/n+

Polycrystalline GaAs Thin FilmHomojunction Solar Cells*

03250224

Organic Thin Film Solar Cell

Progress in LED efficiencies.

Sheats et al., H.P. Lab, Science 273, 884 (1996).

03250225

Bulk HeterojunctionSolar Cell Connected To AnExternal Resistive Load &Screen Printing Technique

03250226

Absorption Coefficients of Films of Commonly UsedMaterials are Depicted in Comparison with the

Standard AM 1.5 Terrestrial Solar Spectrum(the overlap is generally small)

034016378

Single Layer Device witha Schottky Contact at the

Aluminum Contact

Bilayer HeterojunctionDevice

(The donor [D] contacts the higherand the acceptor [A] the lower workfunction metal, to achieve good hole

and electron, respectively.)

034016379

Bulk Heterojunction Device(The donor [D] is blended with the

acceptor [A] throughout the whole film.)

10

8

6

4

2

00 50 100 150

Active layer thickness (nm)

I sc

(iqe

=1)

[mA

/cm

2]

200 250 300

(PEDOT:PSS 150 nm. MDMO-PPV:PCBM 1:4)

Calculated Photocurrent ofa MDMO-PPV:PCBM-BasedSolar Cell under the IdealAssumption of an Internal

Quantum Efficiency of Unity

034016380

Al

Structure (left) and the Energetic Descriptionat Closed Circuit Conditions (right) of an

Organic p-i-n Solar Cell

N-doped C60

Blend ZnPc:C60

P-doped MeO-TPD

ITO

Glass substrate

ITO Metal

034016381

Reported Efficiencies of Some RecentOrganic Solar Cells at AM1.5

03250227

Global Views of Semiconductor Materialsfor PV Applications

Elemental, Binary, and Ternary Semiconductors

034016111

Science Topics Needed by All Thin Films

• Science Base

• Degradation and Metastable Mechanisms

• Device Characterization and Modeling

• In situ process diagnostics and controls

• Device protection from water vapor

• Innovative module design, including cellinterconnects, device protections, lower-costsubstrate, less-costly replacement packaging

Conclusions

• Future of thin-film solar cells looks very promising

• Major improvements in efficiency, stability, and reduction

in cost are being made continuously

• Multiple options are available

• Industries are gearing-up for large-scale production

• Performance gap between laboratory scale devices and

commercial modules needs to be narrowed

• Opportunities are enormous for new innovation in terms of

materials and device technologies

03250228