SOLA2020/SOLA9006
Photovoltaic Technology
and Manufacturing
Lecture 10 – Thin Film Technologies
Alison Lennon
Housekeeping
• Minor
Mi assignment
i t
– Discussion of answers in Wk 12 (also discussion of exam
material in 2nd hr)
• Tutorial sessions
– Attendance at remaining tutorials is optional
– Limited time for revision in lectures so please use your
tutorial time to ask revision questions.
• 2nd hour today: Rhett Evans from CSG Solar
Major Assignment
• Mustt use Version
M V i 2.2 2 2 off VPL.
VPL
• Do not change wafer parameters.
• Use a SiNx ARC not TiO2.
• Must submit final VPL batch file (version 2.2).
• Objective is to maximise average efficiency – why?
• Marking:
– Presentation 10 marks
– Written expression 10 marks
– Final average efficiency 30 marks
– Explanation 50 marks
• Should
Sho ld get at least 14%!
Thin FilmTechnologies
• Thin
Thi film
fil approachh
– Sequentially depositing thin layers onto substrate.
• Cost advantage
g through:
g
– Greatly reduced use of materials
– Inexpensive deposition & processing methods
– Low cost substrate material
• Challenge for thin film technologies (Lecture 6):
– Low efficiencies?
– D bili → maybe
Durability b new niche
i h applications?
li i ?
– Abundance of materials
– Stability
– Toxicity
– Limited history
Thin Film Market Share
• A
Average thin
hi film
fil panell cost expectedd to fall
f ll to US
$1.40/Wp in 2010 (cf US $2.00/Wp for c-Si
technologies).
• Rising availability of turn-key TF production line
(Applied Materials, Oerlikon, and Centrotherm)
accelerating TF market growth.
iSuppli, USA, Sept. 2009
Thin FilmTechnologies
• CdTe – First Solar
• CIGS – Global Solar & Nanosolar
• a-Si – Unisolar
• Crystalline Silicon on Glass – CSG Solar (Rhett)
• Dye Sensitised Solar Cells – Dyesol,
Dyesol Sony
• Organic PV - Konarka
CdTe Thin Film Technology
• Ideal bandgap for terrestrial applications
– Bandgap is 1.44 eV 300 K (temp dependent EG)
• Almost fully absorbs visible light within 1 μm.
μm
• Large percentage of carriers generated close to
depletion layer.
• Cell structure: n-CdS and p-CdTe heterojunction
deposited on a glass substrate coated with a TCO.
• Record lab efficiency for a CdTe cell is 16% for a
1 cm2 cell.
– First Solar average module efficiencies 11%.
First Solar: CdTe Thin Films
First Solar: CdTe Thin Films
• Currently manufacturing cost is US$0
US$0.81/W
81/W .
– Expects to be between US$0.52/W and US$0.63/W by 2014.
• First thin film company to reach 1 GW in installations.
• Proprietary high-speed deposition process that deposits
uniform semiconductor layers in a matter of seconds,
with little material waste.
iSuppli, USA, Sept. 2009
First Solar: CdTe Thin Films
• Rapid growth , founded in 1999,
1999 launched production 2002.
2002
• First plants were in Kedah, Malaysia.
• Smart replication of factories → manufacturing lesson.
• The manufacture process from glass → module takes only 22.55 hrs
hrs.
First Solar: Recycling
• Fi
First Solar
S l takek into
i account entire
i lifecycle
lif l off operation
i
from raw material sourcing to recycling of modules.
• See http://www.firstsolar.com/en/recycle
p://www. s so .co /e / ecyc e_pprogram.php
og .p p
• 1st pre-funded recycling program.
• 90% of module by weight recycled.
• First Solar sets aside funds to cover cost of collection
and recycling.
– Funds will be available regardless
g of First Solar’s financial
situation (in the hands of a trustee).
CdTe Thin Films
• Advantages:
– Direct bandgap 3 um 1.5eV – more efficient
– Better at high T and low light
– Si l device
Simple d i structure
t t
– Automated, high throughput processes ~ 2.5hrs
– Environmental – recycling program
– Convert Cd & Te into stable and inert semiconductor
– Uses by-products from Cu, Zn mining
– 11% efficient
– Energy payback ~ 1 yr1 cf 2.7 yrs for c-Si cells
• Disadvantages:
– Cd toxic
– Te rare
1 http://www1.eere.energy.gov/solar/pv_basics.html
Copper Indium Gallium DiSelenide (CIGS)
• 20.1% efficient in lab (Apr 2010; Stuttgart).
• Direct band-gap material and most sunlight can be
absorbed in 1-2 μm of material.
• Bandgap varies from 1.0 (CIS) to 1.65 eV depending on
the Ga/(In+Ga) ratio
• Deposition onto flexible substrates is possible.
possible
– Active layer is formed by co-evaporating (or co-sputtering)
copper, gallium, and indium, then annealing the resulting film
with a selenide vapour to form the final CIGS layer.
layer
• Cells in production use evaporation
– Sputtering is faster but difficult to get sputter targets.
• Potentially
i ll low
l material
i l andd fabrication
f b i i costs.
CIGS
• CIGs is p-doped from intrinsic defects.
defects
• A homojunction is formed (in the CIGS) by the
deposition of a thin CdS layer.
• Can also form a junction without CdS by depositing
intrinsic ZnO (TCO)
• Buffer layer: prevention of shunts.
shunts
High efficiency CdTe and CIGS thin film solar cells [electronic resource] : highlights of the
technologies challenges / Rommel Noufi. 2006.
CIGS
• Leading companies:
– Ascent Solar
– Global Solar
– Sh ll Solar
Shell S l
– Showa Shell
– MiaSole
– Wuerth Solar (commercial partner of ZSW)
– Nanosolar – 2002
• Deposition
p
– Vacuum, sputtering – best quality
– Low cost ink-printing methods – lowest cost
Global Solar: CIGS
• CIGS on flexible stainless steel substrate.
• Efficiencies of 13% using standard manufacturing
equipment and processes (best in lab 1515.4%).
4%)
• Focussing on BIPV.
CIGS Solar Field (Global Solar)
• 750 kW solar power PV system in Tucson
Tucson, Arizona
– 6,600 solar panels and covers over 5 acres of land.
– 1,200,000 kWh annually, y providing
p g about 30% of Global Solar
Energy's annual energy needs.
– First commercial-scale installation of Global Solar Energy's thin
film CIGS technology.
technology
– Will offset 30,000 tons of CO2 over its 25 year lifetime.
http://www.globalsolar.com/en/technology/cigs-solar-field.html
NanoSolar: CIGS
• CIGS semiconductor printed on low-cost conductive
metal foil (rear electrode) and back-contacted through a
Metal-Wrap-Through (MWT) architecture.
• Cells
C ll are light
li ht weight
i ht bendable
b d bl andd can be
b maded to
t be
b
any size.
• Proprietary
p y CIGS nanoparticle
p ink
• Printed/coated onto a metal foil.
– Printing press technology
• Cells of any size can cut from the sheet.
sheet
• Max efficiency 16.4%, median ~ 11%.
• Produces 640 MW/yr; 1 panel every 10 s; 24/7 capacity.
http://www.betasights.net/wordpress/?p=863
CIGS
• Advantages:
– Stable over time
– 99% light absorbed in 1st micron
– 20% efficient in lab
– 13 % efficiency in production (Global Solar 2009)
– Amenable to large area automated production
– R ll to
Roll t roll
ll processing
i
– Good at low light intensity
– Available on flexible substrates
– Energy payback < 1 mth!
– Demonstrated in production
• Disadvantages:
g
– $/W more than First Solar
– Indium - rare
Amorphous Si (a-Si) Cells
• One of first thin film technologies to be developed.
developed
• Layers of a-Si a few um thick deposited on an
inexpensive backing such as glass, flexible plastic, or
stainless
t i l steel.
t l
• Amorphous - small, disorderly variation in the angles
g
between the bonds eliminates regular lattice structure.
• Plasma-deposited a-Si contains a lot of H.
– a-Si is generally known as “hydrogenated amorphous silicon”,
or aa-Si:H
Si:H.
• a-Si absorbs light more efficiently than c-Si but less
efficient energy conversion.
Amorphous Si (a-Si) Cells
• Initial efficiencies of ~ 13% for small cells in lab but
commercial efficiencies much lower (~ 8%).
• Require p-i-n device configuration because of high
d f t densities.
defect d iti Higher
Hi h electric
l t i field
fi ld helps
h l tot speedd up
of the transport of charge carriers.
• Good low lightg pperformance,,
• But typically experience a 15-35% drop in efficiency
when first exposed to sun-light.
– Staebler-Wronski
Staebler Wronski effect.
effect
• Commercial producers of a-Si cells
– Unisolar
– Solarex
– Intersolar
a-Si (Unisolar)
• Triple junction cells
– a-Si layers of different bandgap are stacked together
interconnected by heavily doped layers that form tunnel
junctions
Layer Bandgap Material
(eV)
Blue 18
1.8 a Si
a-Si
Green 1.6 a-SiGe (10-15% Ge)
Orange
g 1.4 A-SiGe ((40-50% Ge))
a-Si (Unisolar)
• Triple junction cells
– a-Si layers of different bandgap are stacked together
interconnected by heavily doped layers that form tunnel
junctions
Amorphous Si (a-Si)
• Advantages:
– Much less Si required
– Deposited on cheap substrate – steel
– Roll-to-roll vacuum process
– Good in low light, high T
• Disadvantages:
Disad antages:
– Degrade on exposure to light– Staebler-Wronski effect
– Slow deposition rate → high capital costs
– Low efficiency ~ 8%
CSG Solar
• UNSW invention 1995
• Next lecture: Rhett Evans (CSG Solar)
CSG Solar
• Advantages:
– < 1% Si - Non-toxic, Abundant
– Direct encapsulation onto glass
– E ll t metallisation
Excellent t lli ti based
b d on ink
i k jet
j t printing
i ti
– Stable
– Good light trapping
– Potential for high efficiency
• Disadvantages:
– Only 8% efficient
– Custom made equipment
– Difficulty in production
– Ink-jet
– Defects Æ yield losses
Dye Sensitised Solar Cells (DSSC)
• Dye Sensitised Solar Cells
– Invented 1988 Michael Graetzel
• IInvolves
l a semiconductor
i d t formed
f d between
b t a photo-
h t
sensitized anode and an electrolyte.
• I.e.,, it is a photo-electrochemical
p system.
y
• Low-cost material.
• Inexpensive to manufacture.
• Lab
L b efficiencies
ffi i i off ~ 11 %. %
DSSC Operation
• Dye on the surface of TiO2
particles absorbs light →
excited state → inject
electrons into CB of TiO2.
• Electrons move to anode.
• Dye is reduced (to its initial
state) by a reducing agent (tri-
(tri
iodide) in the electrolyte.
• Oxidised iodide → cathode
where it is reduced.
reduced
DyeSol (Based in Queanbeyan)
• Supplies
pp materials,, technology
gy and equipment
q p for
the manufacture of DSSC.
• Sells small demonstrators.
• Close to making commercial cells.
Sony’s “Concerto Effect”
• Mix 2 dyes with different absorption spectrums to
increase the light absorption wavelength range.
• Increased the adsorption of the dye molecule to the
TiO2 surface.
• Prototype cell efficiencies of ~ 10%.
• Targeting consumer-related applications.
G24 Innovations: “Solar on the Go”
• Announced at an electronics show
in 2009.
• Flexible DSSC panels
incorporated into bags enable
recharging of electronic
components.
components
• Produces 0.5 W of power under
direct sunlight.
• Can be patched onto clothing,
tents, etc.
• Made using a low-cost
low cost, roll-to-roll
roll to roll
process to make the flexible
panels.
DSSC Future Directions
• Liquid electrolyte is a problem.
– It can freeze in low temperatures and expand at higher
temperatures resulting in sealing problems
problems.
• Electrolyte can also contain solvents which permeate
plastics – this has limited flexible applications.
• Therefore
Th f replacing
l i the h liquid
li id electrolyte
l l with i h a solid
lid
has been a major ongoing field of research.
• Recent
ece eexperiments
pe e s using
us g solidified
so d ed melted
e ed sasaltss have
ave
shown some promise, but currently suffer from higher
degradation during continued operation, and are not
flexible – also typically lower efficiencies.
• Replace electrolytes with hole conducting polymers.
DSSC
• Advantages:
– Very llow cost
V
– Low embodied energy
– Simple to manufacture
– Low cost equipment
– Pilot production demonstration
– Good in low light
g conditions and high
g temperatures
p
– Good application for tinted windows
– Work well indoor because diffuse and fluorescent light is
absorbed
• Disadvantages:
– Problems with electrolyte (can freeze/expand resulting in
sealing problems)
– Uses solvents which are volatile and in some cases, toxic
– Low efficiencies
– Unstable – dye material
– Low life expectancy
Organic PV (OPV)
• Use conductive organic polymers or small organic
molecules for light absorption and charge transport.
• Photo excitation is fundamentally different in OPV.
OPV
• Absorption in OSCs results in the formation of excitons
(electron-hole pairs that are bound together by Coulomb
attraction).
i )
• Need to dissociate the excitons to generate current.
• Can dissociate excitons at an interface between materials
with different energy levels.
– Electron transfer from donor → acceptor material
• Excitons typically travel ~ 10 nm before decaying
– Need to keep active layers thin.
OPV Lifetimes & Manufacturing
• For “low cost” devices a lifetime of 3-5 years
(operational lifetime of 3000-5000 hours) is regarded as
market entry point.
point
• This lifetime is derived from the usability lifetime of
consumer electronics.
• So needd to consider
id lifetime
lif i together
h with
i h cost.
– Break with the 25 yr guarantee power generation.
• See Konarka videos on http://www.solarpv.tv/
Charge Transport in Organic Semiconductors
• Charge transport occurs via overlap between molecular
orbitals. This is called carrier hopping.
• Transport increases with increasing temperature.
Bandgap is 1-4 eV
Mobilities < 1 cm2V-1s-1 (low
compared to inorganic SCs)
Organic Semiconductors (OSCs)
OPV Cell Types
• Bilayer or planar donor-acceptor
donor acceptor heterojunction:
– The layer with higher electron affinity and ionization potential is
the electron acceptor, and the other layer is the electron donor.
• Dispersed
Di d heterojunction:
h j i
– Electron acceptor and donor are mixed together in a polymer
blend.
Konarka’s “Power Plastic”
• Tunable cell chemistry.
chemistry
• P3HT:PCBM bulk heterojunction technology.
• Highest
g efficiency
y 6.4%.
• Focus is on applications requiring flexibility.
• Uses recyclable materials.
OPV Lifetimes & Manufacturing
• For “low cost” devices a lifetime of 3-5 years
(operational lifetime of 3000-5000 hours) is
regarded as market entry point.
point
• This lifetime is derived from the usability lifetime of
consumer electronics.
• So need to consider lifetime together with cost.
– Break with the 25 yr guarantee power generation.
• See Konarka videos on http://www.solarpv.tv/
Organic PV
• Advantages:
– Low cost
– Flexible
– Well-suited to consumer applications
– Production of consumer applications demonstrated
– G d iin llow light
Good li ht conditions
diti
– Good application for tinted windows
• Disadvantages:
g
– Low efficiencies
– Questionable chemical stability
– Low life expectancy – is this a problem if sufficiently cheap?
Next Week
• BP Solar’s manufacturing experience - David Jordan
• Cost Modelling
