AVS2010 Session EN+TF-TuA: Thin Films for Photovoltaics

The transition from amorphous (a-Si:H) to microcrystalline (µc-Si:H) silicon film growth has been ascribed to the interaction of atomic hydrogen with the (sub)surface of the growing film [1]. To gain more insight into the formation of microcrystalline silicon and to the role of atomic hydrogen, several studies have been dedicated to hydrogen treatment of a-Si:H films [2,3,4,5]. However, the observed film crystallization is often wrongly ascribed to the impinging hydrogen atoms. What tended to be overlooked, is that the counter electrode is covered with an a-Si:H film during the deposition step. Interaction of the H₂ plasma with the coated electrode will result in formation of molecules and radicals originating from silane [4]. Hence, the resulting plasma conditions are similar to those used for µc-Si:H, which explains the observed crystal formation.

Here, we show that this is not only true for direct plasmas, but that also in remote plasmas the reactor wall has big influences on the growth process. By using in situ spectroscopic ellipsometry and ex situ Raman spectroscopy, we show that H₂ plasma treatment of a-Si:H only results in the formation of crystals when there is a source of silicon-based molecules and radicals. Together with a plasma study, comprising mass spectrometry and ion probe measurements, we address the origin and kinetics of the crystal formation and discuss the implications for µc-Si:H growth.

1. A. Matsuda, Journal of Non-Crystalline Solids 338, 1-12 (2004).

2. A. F. I. Morral, J. Bertomeu, and P. R. I. Cabarrocas, Materials Science and Engineering B-Solid State Materials for Advanced Technology 69, 559-563 (2000).

3. A. F. I. Morral and P. R. I. Cabarrocas, Journal of Non-Crystalline Solids 299, 196-200 (2002).

4. K. Saitoh, M. Kondo, M. Fukawa, T. Nishimiya, A. Matsuda, W. Futako, and I. Shimizu, Applied Physics Letters 71, 3403-3405 (1997).

Amorphous hydrogenated silicon (a-Si:H) is widely used in photovoltaic applications. The high absorption renders a-Si:H technically relevant especially for thin film solar cells. Despite lower efficiency, an amorphous silicon solar panel possesses the advantage of higher absorption rate and easier processing at lower production cost.

Here the focus lies on highly (p and n) doped amorphous silicon films. The samples are prepared using d.c.-pulsed magnetron sputtering of crystalline silicon targets. A controlled hydrogen flow is added to the sputtering plasma. Hydrogen in amorphous silicon is known to saturate dangling bonds and improves the short range atomic order [1]. To probe the influence of hydrogen in the sputtering process various spectroscopic techniques were applied for sample characterisation.

Raman spectroscopy is a technique sensitive to the morphological aspects of the film. The relaxation of quasi-momentum conservation in amorphous films results in drastically different spectra of amorphous and crystalline silicon. A broad band at ~485 cm^-1 appears instead of the sharp crystalline phonon feature at 520 cm^-1. Its shape and asymmetry unveils further information on the short range order like the average dispersion angle from tetrahedral conformation. With the help of Fourier transformed transmission infrared spectroscopy the concentration of hydrogen in the sample is studied. Vibrational hydrogen-silicon stretching modes in the region around 2000 cm^-1 are therefore assessed by a modified [2] Brodsky-Cardona-Cuomo approach [3]. Access to the optical constants n and k and therefore the complex dielectric function ε of the sputtered material is granted by variable angle spectroscopic ellipsometry. Thereby important parameters like the Tauc-Lorentz band gap which is mainly determined by interband gap defects are revealed. The combination of these spectroscopic techniques provides a detailed picture of morphological, electrical, and optical parameters of the system. An in depth discussion of the degree of structural improvement, the decrease of interband gap defects, the saturation of hydrogen content, and evolution of optical properties in correlation with the hydrogen flow will be presented.

[1] R. A. Street, “Hydrogenated Amorphous Silicon”, chapter 2.3, Cambridge University Press.

[2] A. A. Langford, M. L. Fleet, B. P. Nelson, W. A. Lanford, N. Maley: Phys. Rev. B 45 (1992) 13367.

[3] M. H. Brodsky, M. Cardona, J. J. Cuomo: Phys. Rev. B 16 (1977) 3556.

Hybrid solar cells with active and transport layers based on conjugated polymers and/or inorganic semiconductor nanoparticles are an alternative to all-organic or all-inorganic solar cells. In hybrid cells, inorganic nanoparticles complement the absorption of the organic phase and provide better charge transport properties due to higher carrier mobility, while still maintaining the ability to solution-process. These properties will be illustrated first in hybrid solar cells with a mixed active layer based on poly(3-hexyl thiophene) (P3HT) and colloidal CdSe nanospheres, and with a ZnO nanoparticle buffer layer. The CdSe and ZnO nanoparticles were synthesized using a micelle and a sol-gel method, respectively. Both the active and buffer layers were spin-coated from solution onto a poly(3,4-ethylene dioxythiophene) doped with polystyrenesulfonic acid (PEDOT:PSS) layer on an ITO/glass substrate, and finished by deposition of the Al cathode. Compared to control devices without the ZnO layer, devices with the layer showed only slight changes in open-circuit voltage and fill factor, but showed 40-70% higher short-circuit current density, depending on the size of the CdSe nanospheres. ZnO-containing devices showed a maximum power conversion efficiency of 2.5-2.8%, compared to approximately 1.6-1.9% for the best P3HT/CdSe nanosphere devices without the ZnO layer. Using a ZnO layer and a low-gap poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]-dithiophene)-alt-4,7-(2,1,3-benzothiadiazole)] (PCPDTBT) to better harvest near-infrared photons, we have achieved a maximum power conversion efficiency of 3.3-3.5%. In addition to the efficiency enhancement, the ZnO layer also drastically improved the air stability of both types of hybrid solar cells. While devices without the ZnO layer degraded completely after one to three days of air exposure, devices with the ZnO layer exhibited only a modest 35% efficiency decrease after >70 days of storage in laboratory air. The mechanisms leading to higher efficiencies and reduced degradation will be discussed.

It is important to investigate new materials for thin film solar cells for 2^nd generation devices. Materials extraction cost and annual electricity production considerations highlighted several potential new materials including cuprous oxide (Cu₂O). In this work, we report on the growth of Cu₂O and highlighted the importance of the oxygen partial pressure during growth. Namely, the partial pressure of oxygen determines the transition of Cu₂O to CuO with increasing partial pressure. This is accomplished at a fixed total pressure as this may influence Cu₂O formation. We then discuss about the interface energy alignments, first between between Cu₂O/ZnO and then ZnO/Si. The former is of importance as inorganic thin film p-n junction that is suitable for 2^nd generation solar cell devices. For the latter case, we fabricated device structure on differently doped Si, to investigate influence on doping on the transport characteristics of the hetero- pn junction. It is found that forward bias characteristics for a heterojunction, is not critically dependant on the band offsets, but rather the build-in-field at the heterojunction. If the physics is considered from the point of view of quais-fermi level separation during light illumination, this build-in-field will also determine the V_oc. In this sense, band offset measurements can only give an indication of the maximum limit of the V_oc for differently doped semiconductors (non-degenerate) heterojunction solar cells. In addition, we show that under illumination, the current conduction for the ZnO/Si at zero-bias is a “forward bias” current, unlike all homojunction devices. This can be understood with a detail examination of the energy band diagrams. This work shows the importance of using measured band offsets to aid in understanding the relative Fermi-level alignment instead of using bulk electron affinity values. The work also demonstrates a whole array of playground possible for thin film heterojunction of different materials to engineer an ideal junction for solar cell devices.

II-VI direct band gap semiconductors are attractive for thin film solar cell (TFSC) applications owing to their potential flexibility in tunable opto-electronic properties and possible application in tandem cells for being band gap materials (E_G > 2 eV). For the n-ZnSe/p-ZnTe heterojunction solar cell, the defect states and electronic band alignment at the ZnSe/ZnTe interface are crucial for device performance. We have employed Al-Kα X-ray photoelectron spectroscopy as well as synchrotron source ultra-violet photoelectron spectroscopy to study the surface chemical composition and electronic structures at heterojunction interface. Scanning electron microscopy (SEM) was used to study observe the film microstructure morphology of the interface.

The wide band gap tunability of In_xGa_1-xN thin films (0.7 eV to 3.4 eV, 1>x>0) makes them ideal for efficient photovoltaic (PV) devices. However, growing high-quality In-rich In_xGa_1-xN films with strong photoluminescence in the green-to-red portions of the visible spectrum has faced considerable challenges due to indium phase segregation and other materials issues. These challenges have precluded the growth of both In-rich InGaN and compositionally graded InGaN materials, and make it difficult to grow higher bandgap Ga-rich materials on top of lower bandgap In-rich materials. Overcoming these difficulties using conventional epitaxial techniques is challenging due to the low decomposition temperatures of In-rich materials (e.g. InN~550°C) and the relatively high growth temperatures for Ga-rich materials (e.g. GaN >800°C).

Energetic neutral atom beam lithography & epitaxy (ENABLE) is a low-temperature thin film growth technology recently developed at LANL that utilizes a collimated beam of energetic neutral N atoms (kinetic energies 0.5 to 5.0 eV) to react with evaporated Ga and In metals to grow InGaN. ENABLE is similar to MBE, but provides a much larger N atom flux and correspondingly high film growth rate. The high kinetic energy of the reactive N atoms substantially reduces the need for high substrate temperatures, making isothermal growth over the entire InGaN alloy composition range possible at rates of >3 microns/hr with no toxic precursors or waste products.

Data on film photoluminescence, crystallinity, electrical properties, doping, and electro-luminescence of In_xGa_1-xN, graded In_xGa_1-xN, and GaN films grown using ENABLE over the full composition range will be presented. ENABLE-grown In_xGa_1-xN films show strong photo- and electro-luminescence spanning the entire visible region of the spectrum, with reasonable carrier mobilities background carrier concentrations typically in the low 10¹⁷ range. Evidence for p-type doping of In-rich InGaN films and characterization of p/n junctions will be discussed along with the prospects for using ENABLE to fabricate efficient PV devices.