Anti-Reflecting Nanostructures - Biomimetic Architectures by Plasma Processing - page 50

Boden and Bagnall [38] compared different biomimetic profiles
(shown in Fig. 2.7c-e) have demonstrated that the superiority of
that shown in Fig. 2.7c, where the reflectance of less than 1% was
obtained for height h > 400 nm and period d > 300 nm, calculated

at exists with

sub-1% reflectance. This window was extremely narrow for the
other two profiles (Fig. 2.7d,e), indicating the practical limitation in
achieving sub-1% reflectivity.

l

= 1000 nm. For h > 1200 nm, a wider window of d

2.3.4 Textured AR Surfaces

Textured surfaces have been implemented on a variety of materials,
such as semiconductors [39,40], metals [41], insulators [42] and
textured metal-dielectric composite films [43] for broadband AR
application, ensuring that the spacing of the arrays is smaller than
the

l l

. For
a columnar textured surface [26], if the texturing dimension is larger
than solar wavelengths, the radiation is trapped in the crevices by
multiple reflections. In contrast, if the texturing is on a smaller scale,
multiple reflections will not happen.
Properly selected semiconductor material can be used as
spectrally selective absorber, by appropriate choice of depth of
texturing, for solar thermal conversation and photovoltaic cells.
Generally, semiconductors, such as Si and GaAs, are potentially
efficient material for solar cell application, which usually suffer from
high reflectance (

concerned but the depth is at least a significant fraction of

∼

40%) and limitations in solar absorption due
to high RI. Upon surface texturing, a single crystal Si can provide a
reduction of reflectance by ~10% [44], while textured amorphous
Si layers can reduce reflection down to

∼

1% [39,40]. These kind of
textured surfaces consist of pillar arrays having SW diameters and
spacings, while heights are equal or larger than

l

.

2.4 Anti-Relecting Nanostructures

With the progress in optical science and technology, it has been
realised that nature not only provides a self-optimised solution, it
also supplements the self-optimised way to get there [11]. Here,
for AR applications, it is interesting to observe how both the
transfer of the nature's solution into man-made systems and also