Photovoltaic properties of PSi impregnated with eumelanin
© Mula et al.; licensee Springer. 2012
Received: 23 April 2012
Accepted: 9 July 2012
Published: 9 July 2012
Skip to main content
© Mula et al.; licensee Springer. 2012
Received: 23 April 2012
Accepted: 9 July 2012
Published: 9 July 2012
A bulk heterojunction of porous silicon and eumelanin, where the columnar pores of porous silicon are filled with eumelanin, is proposed as a new organic-inorganic hybrid material for photovoltaic applications. The addition of eumelanin, whose absorption in the near infrared region is significantly higher than porous silicon, should greatly enhance the light absorption capabilities of the empty porous silicon matrix, which are very low in the low energy side of the visible spectral range (from about 600 nm downwards). The experimental results show that indeed the photocarrier collection efficiency at longer wavelengths in eumelanin-impregnated samples is clearly higher with respect to empty porous silicon matrices.
The relevance of solar power in the renewable energy field is constantly increasing due to its ready availability and to the fact that the available amount exceeds by several orders of magnitude the needs of the human race. The search for new materials with better performances than standard Si-based solar cells is also constantly increasing. Organic materials  emerged as a very attractive solution for this scope, their lower efficiencies with respect to inorganic materials being compensated by lower fabrication costs and higher flexibility. Hybrid materials have also been investigated as a way to combine the low production costs of organic materials with the high efficiency of inorganic materials [2–8].
Among the adopted strategies for new materials, interface geometry often plays a major role in the collection of photogenerated carriers, and bulk heterojunctions [3, 9–12] - intimately mixing the two junction materials while keeping them separate in a ‘fractal like’ high-surface interface - are a very promising design for solar cells. This concept was introduced in the mid 1990s for organic solar cells [5, 13–15] to shorten the exciton travel distance from the photon absorption site towards the charge-separating interface and then both to reduce the spontaneous recombination and to increase the collection efficiency. The very large interfacial area available for charge separation processes also increases the carrier collection efficiency.
We investigate here a new hybrid material for photovoltaic applications composed by n-type porous silicon (PSi) and eumelanin, a natural pigment featuring relatively high electrical conductivity  and believed to rely mainly on proton-based conduction [16–18]. Porous Si is a large specific area material, whose properties depend on Si substrate doping and on fabrication parameters . Its application span ranges from biosensor  to drug delivery  and optoelectronics . In the photovoltaic field, PSi has been considered up to now mainly as an antireflection coating for crystalline Si , and there are very few studies about its photovoltaic properties [23, 24]. Porous Si-organic hybrids have recently been considered , but literature reports are mainly on amorphous  or crystalline [27, 28] Si. An exception are the interesting results by Nahor et al.  reporting a study of a PSi-organic hybrid material realized using conjugate polymers for solar cells, which highlight the potential of PSi-based bulk heterojunctions.
Melanins are a class of natural pigments responsible for the colorations of human skin and hair  in the range from light, yellow-reddish (pheomelanins) and dark, brownish black (eumelanins). Their unique status among natural pigments is due to their socioeconomic and biomedical relevance, encompassing racial pigmentation, skin photoprotection, sun tanning, and pigmentary disorders. Moreover, they display a quite unusual set of physicochemical properties such as broadband monotonic absorption in the ultraviolet-visible .
These features have suggested the possible use of synthetic (artificial) eumelanins in the development of a new generation of bioinspired electrically active devices [30–32]. More recently, eumelanin biopolymers have also been proposed for optoelectronic and photovoltaic applications [33, 34]. Due to their wide absorbance covering the whole visible light spectrum, eumelanins behave as very efficient photoreceptors .
As a part of a large project aimed to assess the scope of melanins and melanin-like materials for the development of novel hybrid functional architectures [35, 36], we investigate here ‘5,6-dihydroxyindole-melanin immobilized PSi’ as a prototypal device for the generation of white light-induced photocurrent.
In our experiments, DHI  and synthetic eumelanin  were prepared according to the literature: a solution of DHI (50 mmol) in phosphate buffer, 0.1 M, pH 7.4, (10 ml) was treated with tyrosinase (400 units) under a stream of oxygen for 4 h at 25 °C, then acidified to pH 4.5 and washed first with distilled water and then in methanol/water 7/3.
Eumelanin immobilization on porous Si was obtained by treating the substrates with the appropriate solution/suspension of DHI or synthetic melanin: methanol, methanol/water, or methanol/phosphate buffer pH 7.0 were used as liquid phase. The substrates were processed with three cycles, 5 min each, in ultrasound bath, followed by 4 h oxygen exposition. When suitable, tyrosinase was also added to the mixture to promote oxidation process.
Electrical contacts were realized by deposition of gold spots on top of the empty and impregnated porous layers by sputtering using an Emitech K450 sputter coater (Quorum Technologies Ltd, East Grinstead, West Sussex, UK). On each sample, four contacts were realized to test the reproducibility of the procedure.
The samples' photoconductivity was tested using a PM8 Analytical prober and a Keithley multimeter (Keithley Instruments Inc., Cleveland, OH, USA). The light source was a tungsten-halogen lamp, whose spectral range at the output of the optical system was in the 400 to 850 nm interval. The active external surface of the samples involved in the photocurrent generation is estimated of the order of a squared millimeter.
Different approaches for the immobilization of eumelanin in the porous matrix have been explored, including treatment of the porous Si with preformed synthetic eumelanin and in situ-induced oxidative polymerization of DHI.
To gain insight into the mode of coupling (Figure 3) of DHI during polymer buildup in this procedure, which is important to match the basic structural features of the immobilized DHI-derived polymer [30, 32] and to validate the procedure, mother liquors were collected and the reaction was stopped in the early stages by the addition of sodium dithionite to reduce the oxidized species. The crude oligomer-containing mixture was acetylated according to an established protocol  and examined by thin layer chromatography (TLC). This treatment of the mixture allowed the isolation of two main eumelanin oligomer intermediates, which were identified as the acetylated 2,7′- and 2,4′-biindolyls.
The oligomer identification has been performed using the protocols reported in  and . In detail, after 2 min of the substrate treatment, the liquors were removed and the oxidation was halted by the addition of 5 % w/w aqueous solution of sodium dithionite and acidified to pH 4 with 3 M HCl. The reaction mixture was extracted repeatedly with ethyl acetate (3 × 250 mL), and the combined organic layers were dried over sodium sulfate and taken to dryness.
The residue was acetylated with acetic anhydride-pyridine 95:5 (v/v) and fractionated by preparative TLC (CHCl3/MeOH 98:2) to identify the constituents by comparison with synthetic standards with known structure.
The data shown in Figure 6 clearly evidenced that the optical absorption coefficient of both thin and thick eumelanin films is significantly greater than that of PSi and even bulk silicon for photon energies lower than 2.5 eV (500 nm). For wavelengths longer than 600 to 650 nm, the eumelanin absorption coefficient is still significant (well above than 104 cm−1) and more than a factor of ten larger than that of PSi whose absolute value becomes less than 103 cm−1. This is where the effect of eumelanin may be expected to generate a more significant difference in the photoconductive behavior of empty and impregnated porous layers. It is important to note that from , the eumelanin absorption coefficient at 1,400 nm is still more than 6,000 cm−1 for thin films and 8,000 cm−1 for thick films.
The photoconductive properties of porous Si samples were studied with and without eumelanin for unoxidized layers. A few samples were oxidized at several oxidation levels. This test has been done because, although the partial oxidation process may reduce the PSi conductivity, the oxidation-induced modification of the eumelanin adhesion to the pore walls could, in principle, more than counterbalance the effect. In all cases, however, all the partially oxidized samples showed no measurable photosensitivity. Whatever the impinging light intensity value (up to a maximum of about 200 W/m2) is, with and without eumelanin, our results conclusively show that even a thin oxide layer on the pore's wall is sufficient to severely limit the layers' photosensitivity.
All non-oxidized samples, for all eumelanin immobilization approaches considered here, showed a marked photosensitivity. When illuminating with the whole lamp spectrum, the variation in the photocurrent intensity from very low ambient light to the maximum impinging light intensity showed an increase of more than three orders of magnitude, fully reproducible for the same gold plot. However, while the wavelength dependence was highly reproducible, as discussed later in more detail, a variability was observed in the maximum measured photocurrent when using different gold contacts on the same sample and/or on different samples, even if nominally identical. At the same time, the maximum photogenerated voltage showed a much lower variation, being between 100 and 150 mV for impregnated samples and between 120 and 200 mV for empty samples, suggesting a contact-related and non material-related issue. Empty PSi layers showed more stable results with respect to the impregnated samples. The maximum absolute photocurrent values measured with full illumination for impregnated and empty PSi were within about a factor of two, while the photovoltage measurement were much closer (5.1 mA/200 mV for empty PSi and 2.2 mA/150 mV for impregnated samples). The observed variability did not appear to depend on the eumelanin immobilization procedure.
The striking feature of Figure 8 is the almost complete superposition of the two curves, indicating that the relative behavior of the two impregnated samples is the same, despite the factor of 20 within their maximum values. This clearly indicates, as stated above, that the observed photocurrent intensity fluctuation is fundamentally related to the contact fabrication process and does not point to a poor reproducibility of the hybrid material behavior.
From the results shown in Figures 7 and 8, there are several considerations that can be made. First, all curves show a monotonic photocurrent increase when going from longer to shorter wavelengths, indicating that the absorption takes place in the whole available light spectrum for both impregnated and empty PSi layers. The second significant feature is that impregnated samples show a steep photocurrent increase in the lower energy part of the spectrum. This is coherent with what can be expected by the comparison of the optical absorption coefficients of the materials, as discussed earlier in this work. If we compare empty and impregnated samples with similar maximum photocurrent, about 40 % of the total photocurrent is generated, for impregnated samples, when using wavelengths longer than 780 nm; while for empty PSi layers, the same wavelength range gives only about 15 % of the maximum photocurrent.
The latter effect is even more significant if we keep in mind the very low intensity of the impinging light in the 700 to 850 nm range: for the impregnated samples, 40 % of the total photocurrent is obtained with a very limited part of the total spectrum; while to obtain the same amount of photocurrent with the empty PSi samples, we need to increase the wavelength range up to 550 nm, including most of the available lamp spectral range.
These results show how the impregnation of the PSi matrix with eumelanin significantly increases the capability of the layer to efficiently photogenerate carriers from light especially when approaching the infrared region.
We have shown that the photovoltaic properties of PSi may be significantly improved by impregnation with eumelanin. In particular, we showed that introducing the pigment in the porous Si matrix leads to a significantly more efficient photocurrent generation in the lower energy part of the experimental wavelength range explored with respect to empty porous Si layers.
This result not only contributes to expand the scope of heterojunctions in developing a new hybrid material but also provides the first evidence of the eumelanin capability to efficiently collect the photon energy. In a given substrate, eumelanin seems to act as a kind of ‘antenna’, modifying the range of the useful wavelength range for photoconversion application.
Although further experimental and theoretical studies are needed to improve the electrical contacts reproducibility and to reach a deeper understanding of the observed behavior in view of future development and applications, the proof of principle device presented here opens a new window into the evolving panorama of eumelanin-based devices and contributes to the development of bioinspired and biocompatible optoelectronic devices.
One of the authors (AP) acknowledges partial financial support by Ministry of Education, Universities and Research (MIUR, Italy), PRIN 2008 project. Andrea Falqui from the Italian Institute of Technology, Genova, Italy is gratefully acknowledged for the SEM measurements.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.