1. David E. Carlson Chief Scientist, BP Solar July 9, 2009
Basic Science Issues in the Development of Photovoltaics
David E. Carlson
Chief Scientist, BP Solar
July 9, 2009

2. Projected PV Installations and Revenues
New investment in clean energy has collapsed to just US$13.3bn in the first quarter of 2009, down by 44% on the fourth quarter of last year and 53% below the level achieved in the first quarter of (Renewable Energy, 13Apr09)
The PV industry has grown at CAGR of ~ 45% over the last decade.
However, 2009 looks to be tough year – iSuppli predicts a 32% decrease in PV installations and a 12% decrease in prices in 2009.
Lux Research believes average selling prices may fall > 25% in 2009.

3. Forecast for PV Electricity Production
Sharp forecasts that PV will supply 10% of the world’s electricity by 2032
Assuming a CAGR of 35% (average over the last few decades), the cumulative PV production would be ~ 3.5 TWp by 2026.
3 TWp of solar electricity will reduce carbon emissions by about 1 Gton per year (7 Gtons of carbon were emitted as CO2 in 2000)

4. The Major Players Crystalline Si a-Si/µc-Si CIGS CdTe
Sharp
Kyocera
BP Solar
Q-Cells
Mitsubishi
SolarWorld
Sanyo
Schott Solar
Isofoton
Motech
Suntech
Evergreen Solar
GE Energy
United Solar
Kaneka
Fuji Electric
Sharp
Mitsubisihi
Schott Solar
AMAT licensees
SunTech
PowerFilm
OptiSolar
EPV
Avancis
Showa Shell
Wurth Solar
DayStar
Nanosolar
First Solar
Antec Solar
AVA
PrimeStar Solar
There are currently about 325 companies developing or producing solar cells.
Total PV production is forecasted to be ~10 GWp in 2009, but demand may only be ~ 3.5 GWp (which can be met by top 10 companies).

5. The Typical Silicon Solar Cell
This device structure is used by most manufacturers today.
The front contact is usually formed by POCl3 diffusion
The rear contact is formed by firing screen-printed Al to form a back-surface field
The cell efficiencies for screen-printed multicrystalline silicon cells are typically in the range of 14 – 17%.

6. Operation of a Solar Cell
The theoretical limit for a crystalline silicon solar cell is ~ 29%.

7. Research Needs for Photovoltaics
Lower Costs
Efficiency improvements will help to lower costs of PV electricity
Low-cost storage required for significant penetration of the grid (> 10%)
In the case of silicon solar cells, there is a need for a high-quality, thin silicon wafer or sheet (10 – 50 m) that can be produced at low cost
Automated high throughput process with intelligent process control
We’ll need to replace silver contacts with lower cost materials such as copper or carbon-based materials (nanotubes?)
At an annual production of ~ 154 GWp (~ 2016), the PV industry would be using the entire annual production of Ag (~ 19,000 tons/yr)
Low-cost (i.e. abundant) materials must be incorporated into reliable, high performance PV modules and systems (annual total U.S. electricity demand is ~ 3 x 1012 kWh which can be met by PV arrays with an area of 100 miles x 100 miles in the U.S. Southwest).

8. PV Experience Curve
PV module prices have followed an experience curve with a slope of ~ 80% (a 20% decrease in price with every doubling of cumulative production).

9. Research Needs for Photovoltaics
Efficiency
Efficiency is a major driver in reducing costs and meeting energy requirements in limited areas
Further efficiency improvements are possible with all PV technologies

10. Conversion Efficiencies vs. Time (NREL)
There has been steady progress in the improvement of conversion efficiencies for a number of PV technologies over the last few decades.

11. PV Module Conversion Efficiencies
Modules Cells (Lab)
Dye-sensitized solar cells 3 – 5% 8.2%
Amorphous silicon (multijunction) % %
Cadmium Telluride (CdTe) thin film % 16.5%
Copper-Indium-Gallium-Selenium (CIGS) % 19.9%
Multicrystalline or polycrystalline silicon % 20.3%
Monocrystalline silicon % 23.4%
High performance monocrystalline silicon % 24.7%
Triple-junction (GaInP/GaAs/Ge) cell (~ 250 suns) %
Triple-junction (GaInP/GaInAs/Ge ) (454 suns) %
For most PV technologies there is a large gap between the best laboratory efficiencies and those achieved in production PV modules

12. Paths to Ultra-High Conversion Efficiencies
Multijunction solar cells (currently used for some thin-film cells and for the highest efficiency cells)
Multiple absorption path solar cells (impact ionization, multiple exciton generation )
Multiple energy level solar cells (localized levels or intermediate bands)
Multiple spectrum solar cells (up and down conversion of photons)
Multiple temperature solar cells (utilization of hot carriers)
All these approaches have theoretical efficiency limits > 60%.
The theoretical efficiency limit is > 80% for multijunction cells utilizing other high efficiency approaches.

13. Research Needs for Photovoltaics
New Device Structures
A number of new devices structures have appeared over the years resulting in higher performance
Some devices utilize new materials in novel structures
e.g. dye-sensitized solar cells, bulk heterojunction organic solar cells, luminescent quantum dot concentrators, etc.

14. SunPower Back Contact Solar Cell
The SunPower cell has all its electrical contacts on the rear surface of the cell.
Production cells ~ 22.4% efficiency; new prototypes at 23.4%.
Diffusion lengths > 3 x cell thickness (using 145 m thick CZ-Si at end of 2008).

15. Sanyo HIT Solar Cell
The HIT cell utilizes amorphous Si intrinsic layers (~ 5 nm) as passivation layers. The cell is symmetric except for the a-Si p+ emitter layer (~ 10 nm) on the front and the a-Si n+ contact layer (~ 15 nm) on the rear.
Best lab efficiency = 22.3% (open-circuit voltages as high as 739 mV).

16. Thin Film Silicon Solar Cells
Delft Univ. of Technology
Companies such as Sharp and Mitsubishi are developing variants of the micromorph solar cell.
Applied Materials and Oerlikon have each sold several manufacturing lines that can produce single-junction amorphous silicon and micromorph solar cells.

17. Cadmium Telluride Solar Cells
The CdS/CdTe heterojunction solar cell is typically formed by using a chemical bath technique to deposit the CdS and close space vacuum sublimation to deposit the CdTe.
Toxicity of Cd is perceived by some to be an issue.
Best lab efficiency = 16.5%.
First Solar reports a manufacturing cost ~ $0.94/Wp.
They’re selling modules at ~ $2/Wp.
They plan to have 570 MWp of capacity by the end of 2009.

18. Copper-Indium-Gallium-Diselenide Cell
NREL has demonstrated an efficiency of 19.9% for the CIGS solar cell.
Typically requires relatively high temperature processing (> 500C).

19. Spectrolab’s Triple-Junction Solar Cell
Spectrolab has reported a conversion efficiency of 40.7% with this solar cell structure operating at ~ 250 suns.
More recently Fraunhofer ISE has obtained an efficiency of 41.1% with a triple-junction cell operating at ~ 454 suns.

20. Research Needs for Photovoltaics
Environmental Issues
Some photovoltaic materials are highly toxic and must be processed with appropriate safeguards and obsolete/damaged product should be recycled
Research on new materials and devices should define all potential hazards before commercialization
Sustainability
Need to use materials that are plentiful - some PV materials may be constrained by availability when production reaches tens of GWp/yr; e.g. In (byproduct of Zn refining), Se and Te (byproducts of Cu refining)

21. Research Needs for Photovoltaics
Reliability
Further reliabilty improvements are required in PV systems (inverters, batteries, etc.)
Encapsulation is necessary to protect contacts, interconnects and some PV materials (crystalline silicon PV modules are warranted for 25 years)
Possible failure mechanisms must be identified for new PV materials and devices
Diagnostics
There is a need for rapid characterization of critical parameters such as minority carrier lifetime, efficiency of light trapping, junction quality, etc. for quality control of production lines

22. Research Needs for Photovoltaics - Diagnostics
Trupke et al., 4th World Conf. on PV (2006)
PL imaging can provide a high resolution map of the minority carrier lifetime of a silicon wafer in ~ 1 second

23. Research Needs for Photovoltaics
Conclusions
Many of us believe that photovoltaics could become the major energy source for the world in the later part of this century, but continued research and development are required in several areas:
PV system costs must be reduced significantly
Improved conversion efficiencies
Long-term reliability must be ensured for new PV materials & devices
Future PV systems should be environmentally benign