A comparison of light-coupling into high and low index nanostructured photovoltaic thin films

Periodically structured electrodes are typically introduced to thin-film photovoltaics for the purpose of light management. Highly effective light-trapping and optimal in-coupling of light is crucial to enhance the overall device performance in such thin-film systems. Here, wavelength-scale structures are transferred via direct laser interference patterning to electron-selective TiO2 electrodes. Two representative thinfilm solar cell architectures are deposited on top: an organic solar cell featuring blended P3HT:PCBM as active material, and a hybrid solar cell with Sb2S3 as inorganic active material. A direct correlation in the asymmetry in total absorption enhancement and in structure-induced light in-coupling is spectroscopically observed for the two systems. The structuring is shown to be beneficial for the total absorption enhancement if a high n active material is deposited on TiO2, but detrimental for a low n material. The refractive indices of the employed materials are determined via spectroscopic ellipsometry. The study outlines that the macroscopic Fresnel equations can be used to investigate the spectroscopically observed asymmetry in light in-coupling at the nanostructured TiO2 active material interfaces by visualizing the difference in reflectivity caused by the asymmetry in refractive indices. C 2015 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 Unported License. [http://dx.doi.org/10.1063/1.4921955]

8][19] Ideally, such structuring is realized using upscalable, low-cost, and fast synthesis strategies compatible with mass production.For instance, Lasagni et al. discussed the scalability of periodic structures fabricated via direct laser interference patterning (DLIP) at a rate of 0.1 m 2 /min with sub-µm resolution. 20Here, we apply DLIP on anatase TiO 2 (Figure 1) to obtain highly ordered periodic structures.In order to assess their potential, these structures are implemented into common thin-film solar cell architectures that either use a blend of P3HT:PCBM (poly(3-hexylthiophene): [6,6]-phenyl-C 61 -butyric acid methyl ester) or Sb 2 S 3 as photoactive materials.Based on our findings, we are able to spectroscopically show that wavelength-scale patterns support the in-coupling of light (E ≥ E bandgap ) into an active material with higher refractive index n, but simultaneously suppress the in-coupling into a low n material like an organic semiconductor.
2][23][24] To the best of our knowledge, this is the first time this method is applied to TiO 2 .The experimental realization of a two-beam interference setup is schematically shown in Figure 1(a).A Q-switched Nd:YAG laser produces 12 ns pulses at a repetition rate of 10 Hz and was scanned over the surface.Due to the strong absorption of TiO 2 at 266 nm, the fourth harmonic of the fundamental beam (1064 nm) is chosen to transfer the interference pattern into the TiO 2 surface.The periodicity of the interference pattern is determined by the angle of incidence θ (Figures 1(a) and 1(b)).Details on laser patterning of TiO 2 and structure formation are provided through the supplementary material. 25he inset of Figure 1(a) shows a photography, which illustrates the occurrence of far-field diffraction patterns due to the reflectance of visible light at the periodically patterned substrates.Corresponding height profiles of the employed patterns are summarized in Figure 1(c).The shown profiles are statistically averaged over 750 line scans, which are extracted from atomic force microscopy (AFM) topographic height profiles such as exemplarily shown for the 470 nm period structure (Figure 1(b)).Two different photovoltaic thin films are conformally coated on top of a flat reference and two patterned substrates featuring a 470 nm and 1600 nm pitch, respectively.Both material systems are finalized by thermal evaporation of a 150 nm thick Ag back reflector (Figure 1(d)).The first architecture consists of approximately 125 nm P3HT:PCBM with 5-10 nm PEDOT:PSS (poly(3,4-ethylenedioxythiophene):polystyrene sulfonate) on top and serves as a model system for an inverted organic solar cell.The second architecture features approximately 50 nm Sb 2 S 3 with 30 nm P3HT on top, a common combination in (extremely) thin absorber photovoltaics.Note that the photocurrent contribution of P3HT is negligible in the Sb 2 S 3 /P3HT system. 9The total absorption A tot of the respective solar cell architectures is determined by measuring the total reflectance R tot at the backport of a 150 mm diameter integrating sphere.Accordingly, A tot = 1 − R tot since the transmission through the Ag back-reflector can be neglected as shown by Snaith and Ducati. 26The results of the spectroscopic examination are summarized in Figure 2. The wavelength-scale 470 nm structure shows superior total absorption enhancement in comparison to the 1600 nm structure for both organic and inorganic active materials.Therefore, the 470 nm structure is the exemplary object of discussion.
As apparent from direct comparison of Figures 2(a) and 2(b), there is an asymmetry in total absorption enhancement and therefore in light in-coupling into the respective active material that is caused by the patterning.In case of the organic active material, the structure induces less absorption (14% for the 470 nm and 26% for the 1600 nm structure) in the main absorption region of P3HT:PCBM above the optical bandgap (i.e., 450 nm to 650 nm).The P3HT:PCBM peak absorption (Figure 2(a)) at around 480 nm (flat) which is caused by resonant Fabry-Perot modes (Figure S3, 25 simulated following Burkhard et al.) is suppressed due to the structuring. 27In contrast, the total absorption of the inorganic active material greatly benefits from the structuring.There is absorption enhancement observable for all states above the optical bandgap of Sb 2 S 3 (46% for the 470 nm structure and 12% for the 1600 nm structure in the wavelength range 400 nm to 720 nm).Even the Sb 2 S 3 peak absorption (Figure S2 25 ) of the coherent standing wave at around 450 nm (flat) FIG. 2. Total absorption measurements of structured (according to Figure 1) thin-film solar cell systems using either P3HT:PCBM (a) or Sb 2 S 3 (b) as active material.The patterning is beneficial for light absorption in the Sb 2 S 3 system, but detrimental for absorption in the P3HT:PCBM system in the main absorption region.The total absorption is determined by measuring the total reflectance at the backport of an integrating sphere.
is outperformed by the structure (Figure 2(b)). 25We note the occurrence of resonant absorption features below the optical bandgap of P3HT:PCBM at around 700 nm (Figure 2(a)), which we discuss in detail in another publication. 28n optics, the macroscopic properties of a material are described by its dispersive refractive index ñ = n + ik, where n is the real part of the refractive index and k is the corresponding imaginary part, the so called extinction.Both are directly related to the intrinsic dielectric material properties. 29The attenuation of a propagating electromagnetic wave is related to the extinction k, whereas the phase velocity within a medium and furthermore the refraction at an interface of two media are determined by the real refractive indices n.Reflection and transmission of light at an interface between media of differing n can be described by the Fresnel equations.Therefore, one single bounce at a flat interface is considered and the angle-and polarization-dependent reflection is predicted based on a geometrical optics approach.Despite the original formulation to describe light-coupling at flat interfaces, we use Fresnel equations to investigate the experimentally observed asymmetry in absorption enhancement, e.g., light in-coupling at the TiO 2 /active material interfaces although the difference in total absorption originates from the implementation of wavelength-scale patterns in thin-film solar cell architectures.Therefore, we use spectroscopic ellipsometry to determine ñ of pure Sb 2 S 3 , TiO 2 , and P3HT:PCBM, respectively (Figure 3(a), Figure S3 25 ).Details on the employed ellipsometry system and the corresponding data processing are provided through the supplementary material. 25he inorganic Sb 2 S 3 absorber shows n values around 3.5, whereas the organic P3HT:PCBM absorber shows values around 1.8 for all investigated wavelength ranging from 400 nm to 880 nm (Figure 3(a)).Most important, the n values of the employed anatase TiO 2 are located between the values of Sb 2 S 3 and P3HT:PCBM at around 2.5.Thus, light is coupled into the active material of the solar cell architecture (Figure 1(d)) through the TiO 2 , either into the high n inorganic Sb 2 S 3 or into the low n organic P3HT:PCBM.In order to visualize the difference in light-coupling, we plot (following equations S1) the dispersive Fresnel equations for unpolarized light, passing either through a flat TiO 2 /P3HT:PCBM interface (Figure 3(b)) or through a flat TiO 2 /Sb 2 S 3 interface (Figure 3(c)) for different angle of incidence. 25The obtained contour plot shows the reflectivity R on a logarithmic scale.Dark red indicates high reflectivity, whereas dark blue indicates low reflectivity.The asymmetry in n observed for the two active materials with respect to TiO 2 has tremendous consequences on light-coupling into the photoactive layers, namely, higher reflection losses and a wavelength-dependent line of total reflection (Figure 3(b)) for a flat TiO 2 /P3HT:PCBM interface.All the incident light that is reflected at the TiO 2 /active material interface does not contribute to any photocurrent generation in the active materials.In general, light-trapping due to textured in-scattering layers is aiming at an enhancement of the optical path length in the photoactive material.Ideally, the incident solar flux is guided along the thin film, and the limited absorption in thin film devices is circumvented by absorbing along the lateral direction.Diffraction at the structured TiO 2 /active material interface causes the incident light to be laterally scattered, but the angle dependent reflection (Figures 3(b) and 3(c)) simultaneously plays an important role.There is a trade-off between lateral scattering and enhanced reflection both caused by the structuring of the TiO 2 /active material interface.The more efficient light-coupling into high n Sb 2 S 3 (i.e., less reflection losses) causes the structuring to be beneficial for the total absorption enhancement (Figure 2(b)).In contrast, high reflection losses at the low n P3HT:PCBM are responsible for the overall reduced total absorption (Figure 2(a)).We therefore conclude that the observed asymmetry in absorption enhancement (Figures 2(a In conclusion, we investigated the influence of wavelength-scale patterns on light in-coupling at TiO 2 active material interfaces.DLIP is employed to implement highly ordered periodic structures into two different representative thin-film solar cell architectures: an organic solar cell featuring blended P3HT:PCBM as active material and a hybrid solar cell with Sb 2 S 3 as inorganic active material.These serve as model systems for low n and high n thin film absorbers, respectively.A quantitative spectroscopic examination shows an asymmetry in absorption enhancement for the two solar cell architectures.The implementation of a wavelength-scale pattern is beneficial for the total absorption enhancement if an active material of higher n is deposited on the structure, but detrimental for the absorption if a low n material is deposited on top.We determined the refractive indices of the pure materials with spectroscopic ellipsometry and used the Fresnel formalism to understand and to visualize the asymmetry in refractive index and correspondingly the asymmetry in light-coupling.Our findings outline that considerations based on the macroscopic Fresnel equations enable a qualitative understanding of light coupling at nanostructured electrode-active layer interfaces, which feature a wavelength-scale pitch.This provides an easy way to assess the potential of a functional nanostructure for light absorption enhancement in thin film solar cells.

066101- 2 Pfadler
FIG. 1.(a) Schematic of a two-beam direct laser interference patterning (DLIP) setup.The periodicity of the interference pattern is determined by the angle of incidence θ.The inset of (a) shows a photography, which illustrates the occurrence of far-field diffraction patterns due to the wavelength-scale topography pattern of TiO 2 .(b) AFM topographic height profile of laser structured TiO 2 featuring a periodicity of 470 nm.(c) Summary of the investigated structures.The shown profiles are statistically averaged over 750 AFM line scans.(d) Schematic drawing of a structured solar cell architecture.The gap is either filled with an organic (P3HT:PCBM) or an inorganic (Sb 2 S 3 ) active material system.
) vs. 2(b)), which is induced by the wavelength-scale pattern, is correlated to the angle-dependent efficiency in light in-coupling (Figures3(b) and 3(c)) that is driven by the asymmetry in n (Figure3(a)).

FIG. 3 .
FIG. 3. (a) Spectroscopic ellipsometry is employed to determine the refractive indices of pure Sb 2 S 3 , TiO 2 , and P3HT:PCBM, respectively.The inorganic Sb 2 S 3 shows high n values, whereas the organic P3HT:PCBM shows low n values in comparison to TiO 2 .The contour plots show the reflectivity R on a logarithmic scale for light-coupling (b) at a flat TiO 2 /P3HT:PCBM interface and (c) at a flat TiO 2 /Sb 2 S 3 interface.The reflectivity is visualized based on Fresnel equations (equations S1), for unpolarized incident light and refractive indices determined in (a).Dark red indicates high reflectivity, whereas dark blue indicates low reflectivity.Total reflectance is exclusively observed in (b).