Figure shows the SEM images of the silicon nanostalagmite array on the glass substrate. Length of SiNS can be controlled and varies linearly with the duration of catalyst etching process. Figure shows high magnification cross-sectional SEM image of μc-SiNSs fabricated with different catalyst chemical etching times of 30 and 90 sec, producing nanostalagmites with a length of around 300 nm and 1 μm respectively. The SiNSs are randomly distributed due to PS nanosphere template which is aperiodic as shown in Figure . Although the SiNSs were not strictly vertical, this structure is still similar to that of the anti-reflection feathers on the eyes of moths [26
SEM image of microcrystalline silicon nanostalagmite arrays. High magnification cross-sectional SEM images of sample after etching for (a) 30 sec (b) 90 sec.
Nanowires provide not only the advantage of more efficient charge transport over planar material but also present the potential for improved optical absorption characteristics [27
]. Nanostalagmites are similar in structure with nanowires. The SiNSs were expected to have an efficient light-trapping effect. The incident light will have multiple internal reflections to cause an optical path for light absorption. Optical measurements were performed on the microcrystalline samples before and after μc-SiNS fabrication. The prepared μc-SiNSs samples were black in appearance and highly non-reflective to the naked eye. Transmittance and reflectance spectra of thin μc-Si film before (light-grey) and after etching to form nanostalagmites (black) were shown in Figure . The transmittance for μc-SiNS over the entire spectral range from 300 to 800 nm is around 0.3%. In comparison, the planar control shows increased transmittance and strong interference patterns after 600 nm. The reflectance of μc-SiNSs and μc-Si film were also depicted in Figure . The ultra-low reflectance of the μc-SiNSs array, which is around 0.3% over 300 to 800 nm, was also observed. In comparison, the planar μc-Si film exhibits around 20% to 30% reflectance over the measured spectrum. The ultra-low reflectance of SiNSs is caused by a strong absorption due to the strong light-trapping property of the dense SiNSs. The nanostalagmite simply acts as a very effective anti-reflection layer. This might be due to the 'moth-eye-like' biomimetic effect.
Optical measurement on thin microcrystalline Silicon film with and without nanostalagmites arrays: (a) transmittance and reflectance and (b) absorption.
The acquisition of reflectance [R
] and transmittance [T
] spectra from μc-Si nanostalagmites allows further obtaining of their absorbance [A
] spectra, expressed as A
(%) = 100 - R
(%) - T
(%) as shown in Figure . Photographs of bare μc-Si and a chemically etched substrate are shown in inset Figure . The absorption over 98% was obtained for SiNSs over the measured spectrum from 300 to 800 nm. This indicates that the SiNSs could extend the absorption to the infrared regime, harvesting more light to increase photocurrent. Our nanostructure SiNS has low reflectance and high absorption; similarly, silicon nanocone [29
] and nanodome [30
] have higher absorption also. However, their process using reactive ion etching which means a high vacuum was used. Here, we demonstrated an alternative way by using simple and chemical process to get silicon nanostalagmite structure. This remarkable property suggests that SiNS arrays are appropriate candidates for antireflective surfaces and absorption materials used in photovoltaic cells. The effect of light trapping could be understood by the absorption coefficient. By taking the R
and transmittance [T0
] spectrum into account, the absorption coefficient can be calculated by the following equation:
in which T is the transmission, α is the absorption coefficient and d is the thickness of the μc-Si layer. Figure depicts the absorption coefficient of SiNSs layer. The highest absorption coefficient for the SiNSs layer is also approximately 7 × 104/cm at 620 nm. The absorption coefficient at wavelength shorter than 620 nm cannot be deduced because of the measured 'zero' transmittance, which is limited by the instrument. The absorption coefficient for the planar μc-Si is around 6 × 104/cm at 550 nm. For the planar μc-Si, the absorption decreases with increase of the wavelength. However, after 600 nm, since planar μc-Si shows clear interference in the transmittance and reflectance spectra, the Equation 1 is not applied well. The absorption coefficient for μc-Si after 600 nm can only be used for estimation of the advantage of SiNSs layer. By this estimation, the average absorption coefficient for μc-Si at 750 nm is around 2 × 103/cm. In comparison, the α value for SiNSs still retained around 6 × 104/cm. There is about 27 times difference. This indicates that the light in the SiNSs at this wavelength is multiple-reflected 27 times by rough estimation. A good antireflective coating should show low reflectance over a wide angle of incidence [AOI], which is important for applications in sunrise-to-sunset solar cells. Figure shows a wide range of AOI in the wide range wavelength. For all possible angles of incidence, we have measured in the range wavelength from 300 to 800 nm. One of the unique features of this catalytic chemical etching SiNS is that efficient light trapping occurs irrespective of the angle of incidence. Angle-dependent reflective of SiNS array was shown in Figure . The performance of SiNSs arrays showed a reduced dependence on the angle of incidence and significantly higher absorption at any angle. At angle of incidence up to 60°, the total reflectance was maintained almost approximately 0.3%.
The absorption coefficient for thin microcrystalline silicon film with and without nanostalagmites arrays.
Optical measurements of AOI and SiNS. (a) Wide range of AOI in wide range wavelength (b) Angle-dependent reflectives of SiNS.