Vous êtes sur la page 1sur 57

Thin Film Solar Cells (A Status Review)

Prof K L Chopra Former Director , IIT Kharagpur Founder, Thin Film Laboratory, IIT Delhi & Microscience Laboratory, IIT Kharagpur

OUTLINE
Requirements for an ideal solar cell Thin film materials for viable solar cells Strengths and Weaknesses of various thin film cells Comparative production status of various cells New concepts to enhance cell conversion efficiency Concluding Remarks

SOLAR Cell:PHOTOVOLTAICS

Direct Conversion of light into electrical energy is called PHOTOVOLTAICS (PV) Photovoltaic devices which convert solar energy into electricity are called SOLAR CELLS Two electronically dissimilar materials (with different free electron densities) brought together to form a junction with a barrier form a PV device. Typical examples are : metal1-oxide-metal2 metal-semiconductor (Schottky) p-type semiconductor-n-type semiconductor (Homojunction) n+-n semiconductor p-type semiconductor(1)-n-type semiconductor(2) (Heterojunction) p- (Insulator)-n (p-i-n)1-(p-i-n)2- p-i-n)3 . (Multijunction) Jct 1/Jct 2 /Jct 3 (Tandem)

SOLAR CELL
Solar Cell operations depend on : Absorption of light to create electron-hole pairs (carriers) Diffusion of carriers Separation of electrons and holes Collection of carriers A Solar cell is a light driven battery with an open current voltage (Voc), short circuit current (Isc), maximum power point current and voltage (In, Vm), and a series and a parallel resistance (Rs, Rsh). Solar Cell Efficiency output = Im Vm = I siVIL FT input nhv nhv depends on quantum efficiency of creation of carriers, effectiveness of separation of carriers before recombination and collection of the separated carriers. Highest Theoretical Efficiency of known Jct Materials Homojunction ~ 30% Heterojunction ~ 42% 36 Tandem Multigap Jctns 76% o o o o

What is required for an ideal Solar Cell ?

1.Cheap,Simple and Abundant Material 2.Integrated Large Scale Manufacturabilty 3.Cost (< 1$/watt)and Long Life
HIGH ABSORPTION COEFFICIENT > 105 cm-1 with direct band gap ~1.5 eV JUNCTION FORMATION ABILITY HIGH QUANTUM EFFICIENCY LONG DIFFUSION LENGTH LOW RECOMBINATION VELOCITY ABUNDANT,CHEAP & ECO-FRIENDLY MATERIAL CONVENIENCE OF SHAPES AND SIZES SIMPLE AND INEXPENSIVE INTEGRATED PROCESSING/MANUFACTURABILITY MINIMUM MATERIAL / WATT MINIMUM ENERGY INPUT/ WATT ENERGY PAY BACK PERIOD < 2 YEARS HIGH STABILTY and LONG LIFE (> 20 Years) COST (< 1$/Watt)

POSSIBLE Solar Cell Materials


Single Elements:

Si ( epi, mc, nc, mixed)


Carbon (nanotubes, DLC) Binary alloys / Compounds: Cu2S, Cu2O Cu-C, CdTe, CdSe,

GaP, GaAs, InP,ZnP , a-Si : H, Dye coated TiO2


Ternary (+) Alloys / Compounds: Cu-In-S, Cu-In-Se,Cu-Zn-S, CdZnSe , CdMnTe, Bi-Sb-S, Cu-Bi-S, Cu-Al-Te, Cu-Ga-Se, Ag-In-S, Pb-Ca-S,

Ag-Ga-S, Ga-In-P, Ga-In-Sb ,and so on.


Organic Materials: Semiconducting Organics / Polymers and Dyes

Solar Cell Technologies


Crystalline Silicon solar cells - Single, Multi, Ribbon Thin Film solar cells - Silicon, Cu2 S , a-Si, m-Si,n-Si, CdTe, CIGS,CNTS Concentrating solar cells - Si, GaAs Dye, Organic ,Hybrid & other emerging solar cells New Ideas

Spectral response of solar cells

Source: unknown

Laboratory for Thin Films and Photovoltaics: 9 and Research Swiss Federal Laboratories for Material Testing Courtesy :Ayodhya Tiwari

Crystalline Silicon :Present Scenario

Efficiency of single crystal Si cells (Laboratory) has been rising steadily to ~ 25% as a result of better understanding of the junction properties and innovations in cell design and fabrication technologies. Efficiency gap between best laboratory cells, submodules/modules, and mass produced modules varies with the maturity of technology and can be at least 10% lower at every step so that the manufactured cell may be as low as 50% of the efficiency of the best laboratory cell. The world PV production of ~ 7900 MW in FY 2009 is primarily (~ 93%) based on single, crystal and polycrystalline silicon. With increasing production of Si-PV from 200 kW in 1976 to 6900 MW in 2008, the cost of solar cells has decreased from $100 to about $3/Wp With the existing technology and the material cost, the cost of Si cells can not be decreased significantly unless major innovations in the production of appropriate quality silicon I thin sheets take place. Present day technology uses 8or larger pseudo square of ~ 200 m thickness, with an efficiency of ~ 15-16%. The energy (16-5 kWH/Wp) pay back period of such cells is ~3-4 years.The module life is about 25 years Specially designed silicon solar cells with efficiency ~ 18-20% are being manufactured on a limited scale for special applications (e.g for concentration). Polycrystalline silicon solar cells with efficiency ~ 12-14% are being produced on large scale. Specially designed thin(~ 20 m) films silicon solar cells with efficiency ~ 12% have been fabricated on a lab scale . Production of hybrid thin film Si cells on MW scale is being pursued

Solar module production for different technologies


CIGS is emerging with about 1% share CdTe is leading with over 6% share a-Si:H: About 5% share
C- Si dominates with ~ 90% share

Source: Paula Mint, Navigant Consulting

EPIA expects thin film shares will grow: 20% in 2010 with about 4 GW 25% in 2013 with about 9 GW

Laboratory for Thin Films and Photovoltaics

Swiss Federal Laboratories for Material Science and Technology 11

------

WHY THIN FILM SOLAR CELLS ?


SMALL THICKNESS REQUIRED DUE TO HIGH ABSORPTION, SMALL DIFFUSION LENGTH & HIGH RECOMBINATION VELOCITY MATERIALS ECONOMY, VERY LOW WEIGHT GHT PER UNIT POWER VARIOUS SIMPLE & SOPHISTICATED DEPOSITION TECHNIQUES A VARIETY OF STRUCTURES AVAILABLE : AMORPHOUS, PLOYCRYSTALLINE, EPITAXIAL

TOPOGRAPHY RANGING FROM VERY ROUGH TO ATOMICALLY SMOOTH DIFFERENT TYPES OF JUNCTIONS POSSIBLE HOMO, HETERO, SCHOTTKY, PEC
TANDEM AND MULTI JUNCTION CELLS POSSIBLE IN-SITU CELL INTEGRATION TO FORM MODULES COMPATIBILITY WITH SOLAR THERMAL DEVICES

TAILORABILITY OF VARIOUS OPTO-ELECTRONIC PROPERTIES ( e.g; Energy Gap ,Electron Affinity ,Work function ,Graded Gap ,etc)

One of the simplest solar cell to produce with simple chemical conversion technique Highest efficiency obtained ~10 % Large scale production of modules with ~5% efficiency demonstrated during 70s Stability of cells due to cuprous-cupric conversion remained an issue Due to the emergence of higher efficiency Si cells, this cell lost the battle of survival Revival of this cell with suitable modifications is a possibility

Thin Film Cu2S CdS Cell

Crystalline states of Si: Long range or short range order of atoms

C-Si & Poly-Si a-Si amorphous Si

Uncoordinated atoms and broken bonds (called dangling bonds are characteristics of a-Si

a-Si:H amorphous hydrogenated Si

Hydrogen passivates the dangling bonds in a-Si:H. Almost uc-Si:H microcrystalline Si (hydrogenated) any impurity can be added to this open structure to obtain asuitable semiconducting behaviour

Absorption coefficient of Si can change with the crystalline state


Different Eg Different optical properties

105 104
Absorption coefficient (cm-1)

103
102 101

c-Si:H

c-Si

a-Si:H

100
10-1 0.5 1
Energy (eV)

1.5

2.5
Courtesy : Vikram Dalal

(small areaeff ~15%

Triple junction a-Si:H/SiGe:H/nc-Si:H solar cell


Area: 0.25 cm2

Initial efficiency: 15.1%; Stable efficiency: 13.3%

Laboratory for Thin Films and Photovoltaics


Swiss Federal Laboratories for Material Testing and19 Research

Amorphous Silicon (a-Si-H) : A Review


The glow discharge technology is well established production process. The highest efficiency obtained in the lab cells is ~ 15%. Single junction cells degrade down to ~ 5-7% efficiency over a period dependent on how these are used. Numerous innovations such as cell integration, graded gap, multi-junctions, light trapping have contributed to the improvements in the cell performance. Stability has been improved with double and triple layer cells. Large MW plants forsingle and multiple junction cells have been set up . The best stabilized (claimed !) module efficiency is ~ 8%. The present day cost/watt of a-Si:H cells and modules is comparable (about $3) to that of single crystal silicon. Because of the lower throughput, complex and expensive deposition technology for triple junction cells, and material cost, the cost can be brought down only with much larger (>100MW_ scale production, or with breakthroughs which help stabilize simpler single junction cells. Major applications of a-Si-H cells are for small scale, small power,flexible power packs ,value -added electronics. a-Si-H PV technology has lost ground from ~ 39% world PV share in 1988 down to ~ 10% in 1997 and less than 4% in Y 2010. .

Thin Film Si Cells


Thick Films (Etched, EFG, melt spun /drawn) : 20% efficiency for 50 micron films demonstrated Micro / Nano Crystalline and Mesoporous Thin Films ( Vacuum Evaporated, CVD ) : 10% efficiency for 2 micron films demonstrated Hybrid and tandem amorphous and microcystalline films/ junctions : 12 % efficiency demonstrated Large Scale ( upto 50 MW) production established PROBLEMS : Thin Film deposition throughput limited to 2-3 microns / min which is not cost effective Higher throughput with good quality opto-electronic properties required Photon trapping structures , Passivation and Cheap Substrate required for lowering the cost

Thin Film CIGS, CdTe, a-Si Solar Cells


Metal back contact

TCO contact CdS or ZnS window CuInGaSe2 absorber

Metal contact CdTe absorber

TCO n I absorber p n I absorber p TCO contact Substrate

mc-Si:H

CdS window

Metal contact
Substrate

TCO contact Substrate

a-Si:H

Highest: 20.3%
Cell area: ~0.5 cm2

Highest: 16.5%
Cell area: ~1 cm2

Highest: 13.3%
Cell area: ~0.25 cm2

Typical range: Cells: 12% - 20%

Typical range: Cell: 10% - 16.5%

Typical range: Cell: 8% - 13.3%

Lower efficiency of large area solar modules


Module: 8% - 13.5% Highest: 15% - 16%
Laboratory for Thin Films and Photovoltaics

Module: 9% - 11% Highest: 11.5%

Module: 4% - 9% Highest: 10.3%

Swiss Federal Laboratories for Material Science and Technology

NANOSOLAR

Thin Film CdTe/CdS Cell


Theoretical Efficiency : ~ 30% Deposition Techniques : CdTe by Evaporation/sublimation/Chemical Solution/Screen Printing CdS by Evaporation/Sublimation/Chemical Solution Lab Cell Efficiency Achieved : ~ 16% Module Efficiency : ~ 10% Nature of Junction : Controversial Formation of Good Junction : Empirical requiring Suitable Heat, Chemical and CdCl2 Treatment required Estimated Production Cost ~ 1$/Wp for 100 MW plant Pay back Time : 1.6 months for 10MW plant Stability : Good Problems : Cd Toxicity and Te Availability Production Technology : Empirical & Temperamental

Thin Film Cu-In (Ga)-Se(S) Based Cell


Theoretical Efficiency : ~ 28-30% Deposition Techniques : Co-evaporation and homogenization Layered vacuum deposition followed by selenization with Se or H2Se Sputter deposition followed by selenization Spray deposition Screen printing followed by selenization Electroplating Lab Cell Efficiency Achieved : ~ 20.3% Module Efficiency ~ up to 15.7% on flexible substrate Estimated Cost : ~ 1$/Wp at > 50 MW Production Pay Back Time : ~ 4 months for 100 MW plant Stability : Good Problems : Multiple Binary Phases; Polymorphism;Structural and Electronic Disorder Availability of In and Ga Sensitive Structure ;Role of Na ? Sophisticated Controls required Upscaling Problematic

PROBLEMS with CIGS Technology


1. 2. 3. Incompatibility of deposition processes for CIS and CdS (Evp / Sputt/ED-followed by selenization for CIS and Evp/CS for CDS). Complex deposition processes and controls. CIS synthesis : Narrow stoichiometry range, polymorphism Multiple Binaries Numerous Structural Defects Nonunifority Electronic Disorder Non Stoichiometry / defect dominated conductivity typedepend on deposition parameters. CdS Microstructure and Morphology very sensitive to deposition process. Mo/CIS Adhesion & Interficial strain.

4. 5.

6.
7. 8. 9.

TCO/CdS Interface (?)


Role of sodium ? Cell-to-Cell mismatch. Encapsulation

CZTS (Se) Cell


Band Gap :1.4-1.6 eV Direct Deposition Techniques : PVD;Sputtering;Spray Pyrolysis; Electrodeposition; Screen Printing Theoretical Efficiency : ~30% Efficiency Obtained : up to about 9.6 % Abundant, cheap and green materials Problems :
Multiphasic ;Mixed Phases (monoclinic,orthorhombic,cubic,tetragonal,stannite) Multistructural;Structural and Electronic Inhomogeneities Difficult to control complicated synthesis process Time and temperature stability questionable

Thin Film GaAs Cell


Deposition techniques include MBE, MOCVD, CVD and LPE Homo, Hetero, Stacked ,Multijunction, Tandem Junction and PEC possible Efficiencies : Homo (23.3%), AlGaAs/Si (26.9%), AlGaAs/GaSb Tandem (32.6%), GaAs/InGaP (30%), Stacked InGaAs and InGap (33%) Junction Formation Straightforward Various types of junctions possible Suitable for stacked cell application Stable Cell. Good for high temperature applications Expensive materials and processing Limited Laboratory batch size production for specialized applications

New & Emerging Excitonic Cells


Photoeletrochemical (PEC) Cell : Efficiency up to 12% Dyed TiO2/ Electrolyte/TCO (Gretzel Cell) & Variations with Polymeric Solid , Gel, and Hybrid Electrolytes Organic ( Plastic ) Cells:Polymer / Polymer , Polymer/ Inorganic Semicon Jct : Efficencies up to 9 % Carbon Nanotube Cells- Hybrid , and Hetero Jct ( concept stage ) PROBLEMS
_

Stability , Empirical Processing and understanding Low Efficiency, Excitonic Transport and Charge Transfer Processes not well understood Encapsulation problems

Organic Exitonic Solar Cell


Principles of photovoltaic energy generation process :

Transparent anode

1 Photon absorption

3 Exciton diffusion

4 Exciton dissociation by charge transfert at interface Acceptor

5 Holes transport

D-A interface

Cathode

Donor 2 - Generation of excitons

5 Electron transport

Gretzel- Dye Sensitive Solar Cell

Organic Solar Cell

Cathode
.

Al LiF

Active layer PEDOT Anode ITO Glass ITO

+ -

h?

Organic Exitonic Solar Cells


Development of tailored conducting and semiconducting polymers and co-polymers has made possible photonic junction devices with these materials. Solar cells of about 8% efficiency have been achieved. Impressively rapid progress is being made to understand the physics of the cells to improve the efficiency on large area cells Fabrication techniques are simple and manufacturable on a large scale Polymers used so far are rather expensive and thus cost of cells remains a question mark Poor stabilty of the cells is a major concern Sophisticated encalpsulation techniques need to be developed

Modeled losses from an ideal solar cell

Other losses 18%

Incident solar radiation 100%

Useful energy 29% Thermalization 32%

Sub-bandgap losses 21%


The most noticeable loss mechanism in solar energy conversion relates to the fact that the basic electronic excitation process in Photovoltaics and also in photochemical processes & photobiological such as photosynthesis

Efficiency Enhancement by Fundamental Processes

Multiple Junction and Tandem cells (feasible and useful) Graded bandgap cells (feasible but complicated) Quantum Well & Q-Dot structured cells ( feasible on
small area cells)

Hot electron cells ( questionable) Multi-carrier generation cells (possible by using inverse
Auger Effect,impact ionization ,field emission if e-ph interaction can be controlled which is the main limitation today )

Up and Down wavelength conversion cells ( not much


to gain from poor efficiency)

Plasmonic Effects for enhancing optical absorption(promising if reproducibleQ-dots can be printed)

Surface Plasmons

Scattering Increase in EM field near particle (Near Field Effect) Direct electron emission from metal nanoparticles Increase in Photonic Mode Density near the particles
Scattering: The light hitting the solar cell excites a surface plasmon on the metal nanoparticle, which then re-radiates most of its energy into the silicon in such a way that the light is trapped inside the cell. Increase in EM field near particle (Near Field Effect): The strong interaction between light and metal nanoparticles also leads to increases in the electromagnetic field around the particles. The particles effectively concentrate the light into small regions. If a semiconductor is close to or surrounding the metal particles, this will increase the light absorbed by the semiconductor in that region. Courtesy: Dr Vamsi

Enhancement of Optical Absorption


(antireflection,scattering , path length increase,plasmonics)

Plasmonic Nanostructure as AR coating(size and shape dependent) Surface Plasma Polaritons Localised Surface Plasmons Nano-imprinted Back Reflector Textured Back Electrode Nano-dome ,Nano-moth eye graded index AR structures Integrated Diffraction &Light Coupled Grating (Limited feasibility for small area applications)

Optical Absorption of Thin Discontinous Silver Films (Source : Thin Film Phenomena)

SPR position depends on material and size and shape of Islands or Q-dots

(Possible Choice of Materials : Si ,Fe, Cu,Al ,C,Ca, Pb,Ba,Zn,S)

Concluding Remarks
1. Hybridized micro- and nano-crystalline and aa-Si:H silicon thin films technologies with efficiencies ~ 10-12% have started competing with mulicrystalline silicon wafer technology . 2. a-Si:H PV technology will continue at a limited level and will cater to portable small/medium power and other photo-electronic application.
3. Both vapour deposited and screen printed , thin film solar cells on flexible and hard substrates, based on CIGS and CdTe films have reached MW scale production with claimed module efficiencies ~ 12-15% at a production cost of about $1/watt . 4.CdTe and CIGS based solar cells have only short range prospects. Only cells based on abundant,cheap and green materials such as Cu ans Fe will have a brighter future. Research on binary or at most tertiary Cu based cells hold the future key.Stabilized CuxS and CuxO thin films need a serious re-visit

5. Small area hybridised/ hybrized organic - inorganic thin film with efficiencies up to 9% and Dye sensitized solid state electrochemical cells with efficiencies upto 12% are opening new vistas for Thin Film Solar Cells. 6.Economic viability and sustainabilitywill ultimately determine the successful thin film technologies. High efficiency at high cost , or low efficiency at low cost are two competing options depending on applications

NEW CONCEPTS :coOLING of HOT CARRIERS

NEW CONCEPTS: Multiphoton Generation

Multiple electron-hole pair generation


Schaller et al., Nano Letters 6, 424 (2006)

The challenge for photovoltaic application (a) Separating electron-hole pairs (b) Collecting them efficiently

Roll-to-roll CIGS solar module production concept

P1 scribe Mo sputter deposition for back contact pressure reduction CIGS absorber layer pressure adjustment Chemical bath deposition for buffer layer

Challenge: Transfer of static deposition processes to dynamic deposition on moving foils


P2 scribe pressure reduction ZnO/ZnO:Al sputter deposition for front contact & anti reflection layer P3 scribe pressure adjustment electrical contacts lamination & protection

Thermal mismatch induced stress

Laboratory for Thin Films and Photovoltaics (Courtesy : Ayodhya Tiwari)

Swiss Federal Laboratories for Material Science and Technology

Consequences of Nucleation & Growth of Films


Grain Structure : Nano to Micro Size; Dense; Porous ; Columnar ; Granular Morphology : Particles ; Quantum Dots; Nano-wires,- rods, -tubes,-sponges ;Films ;Multilayers (Superlattices, QWells) Microstructure :Amorphous ; Nano to Micro-Crystalline ; Oriented ; Epitaxial Topography :Atomically smooth to micron scale rough Crystal Structure :Normal ; Polymorphic ; Metastable Chemical Structure : Normal ;Variable and Extended Solubility ; Non-equilibrium structures Opto-electronic Properties of Micro & Nano-structured Films depend very strongly on nucleation and growth processes and hence on numerous deposition paramaters