Search

International audience; Although solar cells operate under illumination, most electrical characterization methods are carried out in darkness, which implies some bias. In this work, we study the influence of light on the contact resistivity of the electron and hole contacts of a silicon heterojunction (SHJ) cell using the transfer length method (TLM) method in order to determine them in conditions representative of an operating solar cell at maximum power point. A specific fabrication process has been developed to preserve the passivation level. Therefore, we first focus on the patterning approach used for processing, and show that we can fabricate TLM samples with good passivation properties. Using simulations, we also discuss on the influence of inhomogeneities in the excess minority carrier spatial concentration on the TLM analysis. These inhomogeneities, due to shading, local damages in the passivation or to carrier drift under bias voltage make the measurement complicated to carry out without significant error. Our results suggest that the measurement is more precise in darkness and at very high injection levels, under rear illumination and a low voltage bias.

International audience; Choice of topic: This work focuses on the modeling and characterization of silicon heterojunction solar cells. Scientific innovation and relevance: Series resistance (R S) in SHJ cells is related to various carrier transport mechanisms in both bulk materials and at the interfaces, in combined transverse and lateral directions. We propose to review the influence of wafer dark resistivity on R S of SHJ cells with the rear emitter configuration and propose a new approach for improved modeling of such devices. Aim and approach used: Our work aims at improving understanding and modeling of R S various contributions in SHJ cells with rear emitter configuration. For this purpose, we prepared rear emitter n-type SHJ cells varying the substrate dark resistivity from 0.49 to 14.1 Ω. cm, as well as special samples to allow the measurement of electron contact resistance (ρ C,e −). We examined variations of effective lifetime, efficiency and series resistance with c-Si dark resistivity. Results: Efficiency varies only slightly with wafer dark resistivity in our experiment. Injection level is found to be very important to examine performance, both for passivation and series resistance matters. We propose a series resistance model that takes into account the lateral transport in parallel between the c-Si wafer and the ITO layer under varying injection levels.

International audience; Rods of highly magnetostrictive polycrystalline TbxDy1-xFey materials are elaborated by casting in a cylindrical mould with cooling imposed from the bottom which leads to a fast vertical unidirectional growth. A [1 1 0] orientation is obtained along the axis of the rods parallel to the growth direction. The magnetostrictive properties are remarkable along the rods axis comparable to those of [1 1 1] rods. The saturated magnetostriction measured in a parallel magnetic field is lambda(parallel to) = 1124 x 10(-6) under 0 MPa and increases to 1760 x 10(-6) under a compressive stress of 25 MPa. The pseudolinear variation of the magnetostriction with the internal magnetic field Delta lambda(parallel to)/mu(0)Delta H-int reaches 640 x 10(-6)/0.04 T under 9.9 Mfa. The texture and the good magnetostrictive properties are reproducible.

The first part of this paper is devoted to the study of contaminated surface of the uranium materials U, UFe2 and U6Fe a few hours after the preparation. These oxidation conditions are selected with the aim to reproduce the same contamination surfaces that occurred in microprobe analytical conditions. The surface characterization techniques are SIMS and Auger spectroscopy. The contaminated surfaces are described. They are made of successive layers: a carbon layer, one oxidized iron layer followed by an iron depletion layer (only in UFe2 and U6Fe), and a ternary oxide layer (U-Fe-O for UFe2 et U6Fe and UO2+x for the uranium). The second part of this paper is devoted to estimating errors caused by surface contamination during quantitative analysis. EPMA-WDS is used to make the estimation. The analyses were carried out with a CAMECA SX100 and the simulation with the X-Film software.