View
248
Download
0
Category
Preview:
Citation preview
7/27/2019 Photonics for Photovoltaics.pdf
1/43
Photonics for Photovoltaics
Prof. D.M. Bagnall*,
M. Banakar, S.A. Boden, T.L. Temple, D.N.R. Payne, R.S.A. Sesuraj,
A. Asadollahbaik and M.A. Rind
7/27/2019 Photonics for Photovoltaics.pdf
2/43
Dark clouds gather over China's once-booming solar
industryChina's push into solar energy was supposed to be a proud example of how thecountry was advancing into hi-tech manufacturing. But now the whole sector ison the brink of bankruptcy.
Solar power, along with biotechnology and aerospace, was declared a "strategic emerging industry" and was given grants and lo w-cost
loans.Photo: Getty Images
7/27/2019 Photonics for Photovoltaics.pdf
3/43
There is a large diversity of technologies...!
C-Si mC-Si CIGS CdTe/CdS a-Si p-Si Multi-junction Si Dye-sensitised Polymer Nano-rod
Antireflection/ Light- trapping can benefit all device types
7/27/2019 Photonics for Photovoltaics.pdf
4/43
Overview: ten commandments (observations)
1) Real devices must have an A/R scheme and light-trapping is good for all devices
2) Most A/R schemes are textures (and play a part in light-trapping)
3) The Yablonovich limit is quite impressive
4) All layer thicknesses should be optimised for minimum reflection
5) We dont yet have the computational power we need for Finite element
modelling of real devices.6) Plasmonics might (or might not) allow us to make better solar cells
7) Plasmonic properties depend on metal, dielectric environment, size, shape
8) It is difficult to avoid unwanted absorption in metal nanoparticles
9) Scattering, absorption in metal nanoparticles is modified by the photonic
environment10) There are alternatives to plasmonics - wavelength-scale features
(11) Wondrous things are possible but are not always practical )
7/27/2019 Photonics for Photovoltaics.pdf
5/43
5
(1) Real devices must have an A/R scheme& light-trapping is good for all devices
7/27/2019 Photonics for Photovoltaics.pdf
6/43
Improved light-trapping provides opportunity to:
Increase efficiency by increasing ISC
Reduce absorber thickness
may increase carrier collection
decrease requirement on diffusion length
decrease material usage
increase fabrication throughput
7/27/2019 Photonics for Photovoltaics.pdf
7/43
Light-trapping has two key themes
taking more near the bandgap (C-Si, a-Si) narrow band
ultra-thin devices (a-Si, n-Si) - broadband
PVCDROM
7/27/2019 Photonics for Photovoltaics.pdf
8/43
(2) Most A/R schemes are textured (andplay a part in light-trapping)
7/27/2019 Photonics for Photovoltaics.pdf
9/43
Micron-scale texturing inverted pyramids
Sub-wavelength texturing
m0th-eyes
textured TCOs
mC-surfaces
Mie scatterers
Plasmonics (?)
A/R techniques are nearlyalways light-trapping
(and vice versa)
7/27/2019 Photonics for Photovoltaics.pdf
10/43
Boden & Bagnall, APL 93 (2008)
Moth-eye texturing can providebroadband (wavelength and angle)
low reflectance
Exact shape of features is hugelyimportant
Polycrystalline texturing often hassame effect
Modelling random textured surfacesis a big challenge (effective mediumtheory isnt good enough)
7/27/2019 Photonics for Photovoltaics.pdf
11/43
Broadband AR effects can be obtained (with scattering thrown in) but
Small particles: Strong, narrow resonance. Mainly absorbing
Large particles: Weak, broad resonance. Mainly scattering
Hard to avoid short-wavelength losses
What about plasmonics for AR?
In principle, we could exploit forward scattering and should be broadband
[Temple and Bagnall, Progress in Photovoltaics, 2012]
7/27/2019 Photonics for Photovoltaics.pdf
12/43
Enhancement and loss mechanisms
Scatteringtowards
semiconductor
Near-fieldenhancement
Carrierinjection
Scattering
away fromsemiconductor
Parasitic
absorption
e-
7/27/2019 Photonics for Photovoltaics.pdf
13/43
13
7/27/2019 Photonics for Photovoltaics.pdf
14/43
14
(3) All layer thicknesses should beoptimised for minimum reflection
7/27/2019 Photonics for Photovoltaics.pdf
15/43
15
Transfer-matrix methodology can provide analyticalsolutions for simple multi-layered systems
But interfaces are seldom sharp
Modelling for (ellipsometry, reflection spectrometry) useeffective medium approximations
We really need to use finite-element techniques (FDTD)
7/27/2019 Photonics for Photovoltaics.pdf
16/43
(4)The Yablonovich limit is quiteimpressive
7/27/2019 Photonics for Photovoltaics.pdf
17/43
17
A perfect Lambertian reflector allows a path-length
enhancement of4n2d (around 50x enhancement for silicon)
This is theYablonovith or Ergodic limit
[Yablonovitch J. Opt. Soc. Am. 1982)
7/27/2019 Photonics for Photovoltaics.pdf
18/43David Payne (PVSAT6), Ahktar Rind (PVSAT8)
Detector Angle ()
Wavelength(nm)
0 20 40 60 80
500
550
600
650
700
750
800
RelativeIntensity(Lo
g10
Scale)
-4
-3.5
-3
-2.5
-2
Textured TCOs arent perfect Lambertians
Practical limit on roughness as increasingroughness often degrades deviceperformance
x20 enhancements are realistic
Absorption coefficients are dropping byorders of magnitude as you approach the
band edge
7/27/2019 Photonics for Photovoltaics.pdf
19/43
19
(5)We dont yet have the computational
power we need for Finite elementmodelling of real devices
7/27/2019 Photonics for Photovoltaics.pdf
20/43
20
In principle large area, high-resolution AFM images of each
interface could allow reasonableapproximation
But nano-scale features areimportant and avoiding
diffraction requires large areas
7/27/2019 Photonics for Photovoltaics.pdf
21/43
21
(6) Plasmonics might (or might not)allow us to make better solar cells
(more scattering, flatter?)
7/27/2019 Photonics for Photovoltaics.pdf
22/43
7/27/2019 Photonics for Photovoltaics.pdf
23/43
23
Our focus (at Southampton)
7/27/2019 Photonics for Photovoltaics.pdf
24/43
24
(7) Plasmonic resonances are dependent onthe metal, dielectric environment, size, shape
7/27/2019 Photonics for Photovoltaics.pdf
25/43
Surrounding medium
The higher the refractive index the better?
Tune the peak position by increasing the refractive index of the surrounding
medium
7/27/2019 Photonics for Photovoltaics.pdf
26/43
Surrounding medium
Increasing the refractive index means that the maximum totalscattering is achieved at smaller radii.
However, increasing the refractive index reduces the maximum totalscattering and increases absorption.
7/27/2019 Photonics for Photovoltaics.pdf
27/43
Shape: anisotropy
Tuning by shape results instrong, narrow resonances.
Polarization dependency is offsetby extremely high extinction
efficiency
Selectively target wavelengthrange to scatter, or combinemultiple particle types forbroadband scattering.
Reduced effect of higher-ordermodes and interband absorption.
More complex design andfabrication than spheres.
7/27/2019 Photonics for Photovoltaics.pdf
28/43
28
(8) It is difficult to avoid unwantedabsorption in metal nanoparticles
7/27/2019 Photonics for Photovoltaics.pdf
29/43
Radiative efficiency overview
Ag sphere
N=1.550 nm radius
7/27/2019 Photonics for Photovoltaics.pdf
30/43
30[Temple and Bagnall, Progress in Photovoltaics, 2012]
7/27/2019 Photonics for Photovoltaics.pdf
31/43
31[Temple and Bagnall, Progress in Photovoltaics, 2012]
Ag: any position, any materialAu: rear of any device
Al: a-Si onlyCu: rear thick c-Si
7/27/2019 Photonics for Photovoltaics.pdf
32/43
(9) Plasmonic properties are modifiedby the photonic environment
7/27/2019 Photonics for Photovoltaics.pdf
33/43
7/27/2019 Photonics for Photovoltaics.pdf
34/43
Plasmonic mirror
plasmonic mirror should reflect all light
maximise diffuse reflection, minimise absorption
at the moment our experiments dont include silicon (adding silicon will
make a big difference)
The plasmonic mirror
7/27/2019 Photonics for Photovoltaics.pdf
35/43
35
FIG. 1. (a) Cross- section SEM image of the fabricated plasmonic mirror. Image taken at atilt of 36o to show the nanodisc array (b) Measured total and diffuse reflectance spectrafor random arrays of 100nm Ag nanodiscs at 6%surface coverage, on varying thicknessesof SiO2 on a Ag mirror.
Sesuraj and Bagnall, rejected APL (2012)(!)
7/27/2019 Photonics for Photovoltaics.pdf
36/43
Simulated spectra of the scattered power for Ag nanodiscs of 100nm, 150nm and 200nm diameteron 110nm SiO2- on- Ag mirror (b) Optical characterisation results and (c- e) SEM images for
plasmonic mirror samples with 100nm, 150nm and 200nm nanodiscs at 6%surface coverage, on110nm SiO2- on- mirror substrate
7/27/2019 Photonics for Photovoltaics.pdf
37/43
37
(10) There are alternatives to plasmonics? -Wavelength-scale features
7/27/2019 Photonics for Photovoltaics.pdf
38/43
38
7/27/2019 Photonics for Photovoltaics.pdf
39/43
39
7/27/2019 Photonics for Photovoltaics.pdf
40/43
40Banakar, PVSAT8
Nanowires are:
Anti-reflective
Light-trapping
Scattering
Photonic bandgap
Bulk-junctions
7/27/2019 Photonics for Photovoltaics.pdf
41/43
41Naughton et al, Phys. Status Solidi (2010)
7/27/2019 Photonics for Photovoltaics.pdf
42/43
42
(11) Wondrous things are possible(but are they practical?)
7/27/2019 Photonics for Photovoltaics.pdf
43/43
Fraunhofer: Ralf B. Wehrspohn, Johannes pping,Thomas Beckers and Reinhard Carius (SPIE 2011)
Recommended