National Center for Photovoltaics

Photovoltaics:
Energy for the New Millennium

Thomas Surek

National Renewable Energy Laboratory

Golden, Colorado, U.S.A.

tom_surek@nrel.gov
http://www.nrel.gov/ncpv
02679666
 National Center for Photovoltaics

Photovoltaics (PV)
• Direct conversion of
sunlight to electricity

Advantages
• Modular (mW to many MW)
• No (or few) moving parts
• Noise and pollution free
• Reliable; low operating costs
• Abundant, indigenous resource
(30,000 km2 PV for 700 GW) 02679658
 National Center for Photovoltaics

Photovoltaic Technology—Progress and Status

20
201.3
200 World PV Module Shipments CuInSe2
Consumer and Commercial (MW) 16 CdTe Univ. of So. Florida NREL NREL

160 154.9 Amorphous silicon

Rest of the World BP Solar
EuroCIS
Europe 125.8 (stabilized) Boeing Boeing

Efficiency (%)
ARCO
Japan 12 Kodak Univ. of So. FL
120 Photon Energy
Boeing AMETEK
U.S. 88.6 United Solar
Monosolar
77.6 Kodak
Boeing
80 69.4 8
55.4 57.9 60.0
Matsushita
Boeing
40.2 46.5 Univ. of Maine
The Best One-of-a-Kind
40 26.0 29.2 33.6 4 Laboratory Cell Efficiencies
for Thin Films (Standard
0 Conditions)
1986 1988 1990 1992 1994 1996 1998 0
Source: PV News, February 2000 1975 1980 1985 1990 1995 2000

• World market (1999): 201.3 MW, ~$1.7 billion • Record laboratory efficiencies, new materials, and
• Average growth rate (1994–99): >20% innovations
• Remote markets (telecommunications, developing • Commercial product efficiencies and reliability
countries); PV/buildings increasing
• Module and system prices decreasing • Multiple technology paths
• New products and applications – Crystalline silicon (ingot- and non-ingot-based)
– Thin films (amorphous Si, CdTe, copper indium
• Manufacturing capacities increasing: crystalline Si diselenide)
dominates, but thin film scale-up underway – Concentrators and high efficiency 02679601
 National Center for Photovoltaics

220
201.3
200 World PV Module Shipments
180 Consumer and Commercial (MW)
160 Rest of the World 154.9
140 Europe
125.8
120 Japan
U.S.
100 88.6
77.6
80 69.4
60 55.4 57.9 60.0
46.5
40.2
40 33.6
26.0 29.2
20
0
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
02679619
Source: PV News, February 2000
 National Center for Photovoltaics

Progress in Laboratory Efficiencies

40

02669404
Concentrator cells
Conversion efficiency (%)

30
Flat-plate crystalline silicon cells

20 Flat-plate thin films

10
Commercial module efficiencies lag 10–15
years behind laboratory efficiencies
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
02679662
 National Center for Photovoltaics

20
CuInSe2

16 CdTe NREL NREL

Univ. of So. Florida

Amorphous silicon BP Solar

EuroCIS
(stabilized) Boeing
Efficiency (%)

Boeing
ARCO Univ. of So. Florida
12 Kodak
Photon Energy
Boeing AMETEK
United Solar
Monosolar Boeing
Kodak
8 Matsushita
Boeing
Univ. of Maine
The Best One-of-a-Kind
4 Laboratory Cell Efficiencies for
Thin Films (Standard Conditions)
0
1975 1980 1985 1990 1995 2000
02679620
 National Center for Photovoltaics

Crystalline Silicon (Ingot-Based) PV—Progress and Status

Key companies: BP Solarex, Siemens Solar, Kyocera, Efficiency Status Cells Modules
AstroPower, Solec/Sanyo, Sharp, Photowatt, Shell Float-zone 24.7 22.7
• ~85% of today's market Czochralski 22.0 13–15
• 150 MW capacity (to double in near-term) Cast poly 19.8 10–13
• Proven products, 20-year warranties • Batch/continuous processing
• Large ingots: 100 kg CZ, 200 kg casting • High-efficiency devices in production
• Multiple ingots w/ melt replenishment • Well-developed technology base—new understanding
of defects/impurities
• Wire saw: < 300 µm wafers, < 200 µm kerf
• Continuous electromagnetic casting in production
02679603
 National Center for Photovoltaics

Crystalline Silicon (Ingot-Based) PV—

Research Issues and Directions

B Concentration
(atoms/cm3)
vs. Depth (µm)

µm

• Impurity/defect gettering and passivation • Czochralski “tri-crystals” to strengthen thin wafers

• High-throughput, low-cost, high-efficiency processes— • Thin-wafer (down to 100 µm) handling
rapid thermal processing? • Wire-saw slurry recycling
• Environmentally benign processing; waste-stream • Dedicated silicon feedstock supply—new processes
reduction for “solar grade” silicon?
• Manufacturing automation and module packaging for • Novel purification for mg-silicon (e.g., porous silicon
30-year life etching/gettering)
02679604
 National Center for Photovoltaics

Crystalline Silicon (Non-Ingot-Based) PV—

Progress and Status

Key companies: ASE Americas, Evergreen Solar, Efficiency Status Cells Modules
AstroPower, Ebara Solar, Bayer, Pacific Solar, Kaneka EFG 14–15 10–12
• Status varies from prototype modules to pilot String ribbon 14–15 10–12
production to commercial products Thick Si/substrate 16.6 >10
• Proven products (~ 3% of market) Dendritic web 16–17 14
Thin Si/substrate up to 11 n/a
• Capacity increases underway—few tens of MW in
near term • Improved performance from defect/ impurity and
passivation studies
• New interest in thin silicon growth 02679605
 National Center for Photovoltaics

Crystalline Silicon (Non-Ingot-Based) PV—

Research Issues and Directions

• Manufacturing yield and throughput • Thin silicon on low-cost substrate

• Novel processing for continuous sheets
• Impurity/defect gettering and passivation
• New technology directions
– ASE Americas — 1 m diameter cylinders
– Evergreen — < 100 µm thick, wide ribbon
– AstroPower — < 50 µm, monolithic interconnects
on substrate
– Fast deposition rate (> 1 µm/min)
– Grain size comparable to thickness
– Diffusion length greater than thickness
– Insulating substrate for interconnects
• Many novel approaches: solid-state recrystallization,
nucleation and growth, LPE, CVD, CVT, laser ablation, …
• Orientation-independent processes for light-trapping
and passivation in thin Si 02679606
National Center for Photovoltaics

EFG Silicon Cylinders — ASE Americas

Current: 0.5 m diameter

150 µm thick
Future: 1.0 m diameter
<100 µm thick
02679657
 National Center for Photovoltaics

Thin-Film Amorphous Silicon PV—Progress and Status

Key companies: BP Solarex, United Solar/ ECD, EPV, Efficiency status: Cell 12.7
Iowa Thin Films, Sanyo, Canon, Phototronics, DunaSolar (stabilized) Submodule 10.4
• Multi-MW/year in consumer products Module 7–8
• 5 and 10 MW plants operational; few tens of MW in Commercial 5–7
near term • Engineered “solution” for degradation: thin absorber
• Unique products for building integration (e.g., roofing, layers and multijunctions
cladding) • Extensive fundamental research, leveraged by
many other applications 02679607
 National Center for Photovoltaics

Thin-Film Amorphous Silicon PV—

Research Issues and Directions

• Manufacturing throughput and yield • Improved fundamental understanding:

• 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; medium-
range ordering
– Metastability (e.g., hydrogen collision model and
kinetics)
– Molecular dynamics models
– Alloys with Ge, C, …
– Role of hydrogen
– Characterization techniques
• Improved cell/module efficiencies; new device structures
• Long-term field performance 02679608
 National Center for Photovoltaics

Thin-Film Cadmium Telluride PV—Progress and Status

8 130
Array efficiency

Array efficiency (%)

6 90

Temperature (oC)
SCI CdTe 1 kW System:
Efficiency vs. Time
4 50

2 10
Air temperature
0 -30
6/1/95 5/31/97 5/31/99

Key companies: First Solar, BP Solarex, Matsushita, Efficiency status: Cell 15.8
Antec Module 10.8
• ~1 MW/year in consumer products Commercial 7–9
• Manufacturing expansions underway: • Many approaches for >10% efficiency
– High-rate vapor transport (vacuum) • Lacking fundamental scientific and engineering base
– Electrodeposition (non-vacuum) for materials/devices
– Few tens of MW in near term • ES&H issues studied and under control (e.g.,
• Field testing of modules shows promise recycling)—Cd perception issue?
02679609
 National Center for Photovoltaics
Thin-Film Cadmium Telluride PV—
Research Issues and Directions
GRAIN BOUNDARY

Thick
ZnTe:Cu
SIMS Profile
of Cu Thin
Diffusion ZnTe:Cu

• Film deposition development • Front and back contacts

– Nucleation and growth
– Gas-phase or surface chemistries
– Annealing and heat treatment (CdCl2)
– Grain growth, native defects, dopants
– CdS/CdTe interdiffusion
– Alternate transparent conductors; impact on film
growth
• Successful first-time manufacturing
– Alternate transparent conductors
– Low resistance, stable back contacts
– Role of Cu; Cu-free contact strategies?
• Close efficiency gap (cell –> module)
• Compatibility of manufacturing process steps
(e.g., tie to glass float-line?)
• Low-cost module packaging for long-term reliability
(>20 years)
• Accelerated module test procedures 02679610
 National Center for Photovoltaics
Thin-Film Copper Indium Diselenide (CIS) PV—
Progress and Status
MgF2 (0.1 µm)—anti-reflection coating
ZnO (0.4 µm)—transparent conducting
oxide
CdS (0.05 µm)—n-type window
CuInGaSe2 (3.0 µm)—p-type absorber
Mo (1.0 µm)—back contact

Glass (1–3 mm), plastic, or stainless steel

foil—structural support

Key companies: Siemens Solar, Global Solar/ITN, Efficiency status: Cell 18.8
ISET, EPV, Wurth Solar, Showa/Shell Submodule 14.7
Module 12.1
• Prototype production started in 1998:
Commercial >10
– First commercial products (5–10 W)
Others: Stainless steel substrate 17.5
– Efficient, large modules (>12%)
Electrodeposition 15.4
– Expansion to multi-MW in near term
With ZnO (no buffer) 15.0
• Field testing of modules shows promise; • Alloying with Ga and S; role of Na
>10 years outdoors, no degradation • Progress mostly empirical; little understanding of
materials/devices/processes 02679611
 National Center for Photovoltaics
Thin-Film Copper Indium Diselenide (CIS) PV—
Research Issues and Directions
Electrodeposited CIGS Absorber CIGS from Electro-
Precursor Film deposited Precursor Film

• Scalability of current processes • Device research and development

– 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
– 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,
defect physics, doping 02679612
 National Center for Photovoltaics

High-Efficiency and Concentrator PV—Progress and Status

Key companies: PV International, Entech, Amonix, Efficiencies: Si (up to 400X) 27

Sunpower, Tecstar, Spectrolab, Honda, BP Solar GaAs (up to 1000X) 28
• Manufacturability demonstrated GaInP2/GaAs (1X) 30.3
– Low-concentration, line focus GaInP2/GaAs (180X) 30.2
– High-concentration, point focus GaInP2/GaAs/Ge (40–560X) 32.3
– High efficiency cells (Si, GaAs, multijunctions) • Module efficiencies: 15-17% (Si); best prototypes:
in production >20% (Si), >24% (GaAs), 28% (GaInP2/GaAs/Ge at 10X
• Limited applications in today's markets • Large space markets drive GaInP2/GaAs commercial
cell production 02679613
 National Center for Photovoltaics

High-Efficiency and Concentrator PV—

Research Issues and Directions
• Want 1 eV material Calculated efficiencies (ideal)
lattice-matched to GaAs 500X AM1.5D: 36% 47% 52%
⇒ Try GaInNAs one sun AM0: 31% 38% 41%

GaInP GaInP GaInP

1.8 eV 1.8 eV 1.8 eV
GaAs GaAs GaAs
1.4 eV 1.4 eV 1.4 eV
New New
1.0 eV 1.0 eV
Ge
0.7 eV
4 5 6 7 8 9 2 3 4
In
1 Future generation
Energy (eV) production

• High-efficiency (>40%) multijunctions • Novel concentrating concepts:

– GaInAsN and Ge for 3rd and 4th cells – Reflecting troughs
– Lattice match to GaAs – Concentrating dish
– Short diffusion lengths to date – PV/thermal hybrids
• Monolithically integrated modules (e.g., for dish – Non-imaging optics
concentrators) – Low-profile concentrators
• New products for today's markets • Standards for qualification and performance
testing of concentrators
02679614
 National Center for Photovoltaics
Novel PV Concepts and New Materials—
Research Issues and Directions

Nanoparticle-derived precursors for PV Dye-sensitized TiO2 photochemical cells

• Potential for very low cost
• Low process temperatures, non-vacuum
• Potential for smooth, dense films
• Absorbers (CdTe, CIS, ...)
• Transparent conductors (SnO2, CdS, ...)
• Contacts (Ag, Au, Pt, ...)
• Nanocrystals, nanotubes, nanorods
• Nanocharacterization
• Potential for very low cost
• Nanocrystalline TiO2, with monolayer dye sensitizer,
in liquid electrolyte
• 11%-efficient cell; scale-up for consumer products
underway
• Dye stability issue
• Gel or solid-state electrolytes
• Photoelectrochromic window (with WO3) 02679615
 National Center for Photovoltaics
Novel PV Concepts and New Materials—
Research Issues and Directions
100
Room heat Room heat

80 SnO2
Absorbance (%)

60 Exhaust
Quartz shield
Cd2SnO4 Window

40 PV array Matched
emitter
Cooling fins
20
Propane
0 Combustion
Cooling Cooling
0.4 0.8 1.2 1.6 2.0 air Fan
air
air
Wavelength (µm)

Transparent conducting oxides Thermophotovoltaics (TPV)

• Critical component in thin-film devices • Conversion of IR radiation (1300–1800° K) by
• Conventional TCOs: In, Sn, Zn-based low-bandgap (0.3–0.7 eV) PV cells
• Many fuel choices and radiators; system design
• New/improved materials: high mobility key includes heat recycling/recovery
• Non-oxides, p-type TCOs, single crystals • High power densities (1–5 W/cm2)
• Theory for defects, doping, role of grain boundaries • Single and multijunction III-V cells: GaSb, InGaAs,
• For PV: thin, high rate/low T, non-toxic, good electro- GaInAsSb; thin-film PV cells
optical properties, easy to etch • Many applications and system concepts; prototypes
available on market
02679616
 National Center for Photovoltaics

PV Cell and Module Efficiency Measurements

Eff = Eff =
10.4% 18.8%

• Spectral response
• Efficiency measurements with respect to standard measurements:
reporting conditions - Extended wavelength range
• Multiple test-beds for accurate and reliable results - QE of each cell in a module
• Any PV cell or module technology - Measure non-uniformities
• Spectrally adjustable light sources for multijunction devices in photocurrent
• Concentrator cells (to 2000 suns)
• Reference cell calibrations 02679665
 National Center for Photovoltaics

Analytical Microscopy — Microcharacterization Labs

AFM ASEM (a) M CIGS ITO
o

Device
(c) junction
EPMA STEM
Twin boundary

(b)

5 nm

(a) AFM studies of recrystallization of polycrystalline

CdTe versus treatment
• Structural, topographical, chemical, defect, and (b) ASEM cross-section of CdS-less CIGS with EBIC
electrical properties of PV materials and devices linescan showing junction location
• Down to atomic-scale resolution (c) Atomic structure of defects in CdTe by TEM 02679664
 National Center for Photovoltaics

S ubm icron U ltrafast Studies on PV M at eria ls

1000

Photoluminescence (arb. units)

100

10

1
0 1 2 3 4
Ultrafast Near-Field Scanning Optical Time (nsec)

Microscopy (NSOM): Time-resolved photoluminescence (TRPL) across a grain

• Spatial resolution = 0.1 micron boundary:
• Time resolution = 1 psec • CdTe bicrystal — lifetime at grain boundary(lower curve)
• Directly probe effects of passivation decreases by factor of 2 compared to few microns from
on recombination at defects boundary (upper curve) 02679663
 National Center for Photovoltaics

PV Markets and Applications — International

Rural electrification:
• Water pumping
• Desalination/disinfection

Brazil India

India China
Brazil South Africa

Home and security

lighting
02679653
National Center for Photovoltaics

PV Markets and Applications — International

Rural electrification:
Schools and clinics

Zambia India

South Africa
Brazil Mexico

Village power

02679654
National Center for Photovoltaics

PV Markets and Applications — Domestic

Remote, off-grid power:
• Telecommunications
• Signaling

• Disaster mitigation
• Emergency power
02679656
National Center for Photovoltaics

PV Markets and Applications — Domestic

Remote, off-grid power:
• Water pumping
• National Parks

• Stand-alone power
• EV charging
02679652
National Center for Photovoltaics

PV Markets and Applications — Buildings

Grid-connected:
Commercial and residential

02679655
 National Center for Photovoltaics

Some FAQs Concerning Photovoltaics

• Balance-of-Systems (BOS) components issues?

• Materials availability: limits to large-scale PV?
• Is photovoltaics safe?
• What is energy payback period for PV systems?
• How much land area for large-scale PV?

02679667
 National Center for Photovoltaics
Balance-of-Systems (BOS) Components
• Key companies: Trace, S&C (Omnion), Applied
Power (Ascension), UPG, AES, Mastervolt, NKF,
Nippon Denchi, Toshiba
• BOS related to area (e.g., installation), power (e.g.,
inverter, storage), and applications (e.g., pumps, lights)
• 30–70% of today's system cost; source of nearly all
reliability issues
• Proven inverters (2–4 kW and 20 kW); experimental
for 100s of kW
• Hybrid systems; new ac-modules
• Reliability addressed by HALT and quality programs
in manufacturing:
- Improved MTBF (30,000 to 114,000 h)
- Reduced failure rates (14% to 2%)
• Research issues and directions:
- Interconnection and islanding
- Surge resistance (lightning, transients)
- Advanced inverters (e.g., soft switching)
- Battery management
- Reliability testing; standards
- Lower costs and improved reliability 02679671
 National Center for Photovoltaics

Materials Availability: Limits to Large-Scale PV?

Percent of Today’s Production for 1 GW PV
40
Drinkard Metalox

30

20

10

0
Indium Tellurium Gallium*
*<1% if all Ga is recovered.

• Si availability: limited only by quality and cost.
“Solar-grade” silicon?
• “Rare” metals are by-products of primary metal
refinement: Te (from Cu), In (from Zn), Ga (from Al)
• Other pure materials (e.g., CdS, TCOs)
• Today's production and costs are functions of demand;
competing uses
• 2 to 3-fold price increases will not impact PV costs;
• Thinner layers; higher cell efficiencies
• Better materials utilization in deposition
• Substitutions: Ga for In, S (or Zn) for Te, ZnO for CdS, .…
• Recycling in manufacturing and disposal
• Improved recovery techniques and new sources
for production of rare metals
02679670

>8-fold increase will

 National Center for Photovoltaics

Photovoltaics IS Safe!

• Brookhaven National Laboratory: Photovoltaic EH&S • Codes and regulations:

Assistance Center
• Work with industry and labs
• EH&S from R&D to manufacturing to deployment to
disposal/recycling
• Toxic, explosive, and carcinogenic materials in manufacturing
(e.g., SiH4, H2Se, Cd, Pb, As, solvents, effluents)
• Small amounts in products; well-sealed and in inert form
• Proactive approach: hazard identification; monitoring and
control strategies
• Semiconductor industry experience
- Manufacturing (OSHA, EPA, States)
- Deployment (UL, NEC)
- Disposal (EPA, States)
• Reduced solvents and waste streams
• Pb-free solder
• Recycling: <$0.05/W
• 1 kW PV system in typical location:
- Avoids about 2 MT of CO2 per year
- Reduces other pollutants (NOx, SOx)
• PV offsets projected C increase (~2025)
02679669
 National Center for Photovoltaics

Energy Payback for PV Systems

4-3-2-1 Energy Payback for PV Systems Cumulative Net Clean Energy Payoff
4.0
For PV System (415 kWh/month) Meeting
System components 120 Half of Average Household Use
Balance of system
3.0

MWh clean energy

Frame
Module 80
Years

2.0
Multicrystalline
present

40 Return
Multicrystalline

Investment
Thin-film

future

1.0
present

0
future
Thin-
film

0.0 0 5 10 15 20 25 30
PV technology Years

• Conventional technology is energy intensive • BOS components and support structures add less
(Si purification and growth, cell processing, than 1 year to energy payback
module packaging) - Will decrease as efficiency increases
- 4 years for present technology • Energy credit for PV as building component (e.g.,
- 2 years for future technology rooftop, facade)
• Thin films use less materials, lower-energy processing; • With 30-year projected life, PV system provides
substrate and module packaging are key 15 to >30-fold energy return
- 3 years for present technology • Avoidance of emissions (CO2, NOx, SOx, particulates)
- < 1 year for future technology during PV system life 02679668
 National Center for Photovoltaics

Photovoltaics is Solar Electricity

Good for our DOE PV Program Goal:
economy and U.S. leadership in technology,
energy
independence industry, and markets

Good for our environment

Solar can supply all electricity for the U.S.

using this area (100 x 100 mi.) in the SW
OR
Distributed applications
throughout the U.S. (vacant land, Clean and abundant
building-integrated, etc.) energy for the
21st Century

High-
High-technology
manufacturing jobs

02803203
 National Center for Photovoltaics
Photovoltaic Technology—
Prospects and Future Directions
6 20
Average Direct Manufacturing Cost for
Module cost ($/Wp)
PVMaT Partners
4
2 Projected
0
0 200 400 600 800
Total manufacturing capacity (MW)
• PV works, reliable, competitive in many rapidly-
growing markets
• Continuing technical advances: higher efficiencies
and lower costs
• Significant manufacturing expansions underway:
few 100s of MW in near term
• Crystalline silicon dominates markets
• Many challenges for new technologies: technical,
market, and financial risks
50
U.S. and Worldwide PV Module Shipments
U.S. cumulative shipments (GWp)
2000–2020 (25% per Year Growth)
16 40
Total world shipments (GWp/yr)
U.S. industry shipments worldwide (GWp/yr)
GWp per year
12 30
Cumulative U.S. shipments worldwide (GWp)
8 20
4 10
0 0
2000 2010 2020
• PV industry projection of 25% growth rate
• Significant contributions to energy and environment
starting in 2020–2030
• Market incentives support manufacturing expansion,
but impact on cost reduction not apparent
• Sustainable markets require lower cost technology
• Key: continued technology development
02679602