Opto-electrical modelling of CIGS solar cells

One of the key approaches to slow down and eventually prevent dramatic climate change is direct electricity generation from sunlight. Thin-film copper indium gallium (di)selenide (CIGS) is an excellent candidate for highly efficient and stable solar cells. A tuneable and direct bandgap as well as a high absorption coefficient allow for CIGS solar cells to be nearly 100 times thinner than their crystalline silicon (c-Si) counterparts; a feature suitable for flexible photovoltaic (PV) applications. In this thesis, light management for sub-micron CIGS solar cells is studied with the help of opto-electrical simulations. In Chapter 2, the theoretical optical limits for CIGS solar cells as well as the various available opto-electrical modelling platforms are briefly discussed. We study the Green absorption benchmark as a function of thickness and bandgap. Our modelling tools of choice, namely Ansys HFSS for the optical simulations, and Sentaurus TCAD for the electrical simulations are introduced in more details. The interface between CIGS and molybdenum (Mo) back contact is subject to a considerable amount of optical and electrical loss. This issue is investigated in Chapter 3, where we firstly discuss the plasmonic nature of the optical losses. Later, we introduce a double-layer dielectric spacer consisting of MgF2 and Al2O3 with periodic point contacts to quench the Mo-associated losses. We optimize the spacer thickness and the point contact area coverage for maximal photo-current density (Jph) in a CIGS solar cell with 750-nm thick absorber. The front reflection losses, contributing to roughly 10% of optical losses, are addressed in Chapter 4. We show that an MgF2-based double-layer porous-on-compact anti-reflection coating (ARC) allows for gradual refractive index change from air to CIGS and, therefore, according to the Rayleigh effect leads to a wideband antireflection effect. This is done by means of Bruggemann’s effective medium approximation and sequential nonlinear programming (SNLP) for the optimization process. Our models suggest that the proposed ARC surpasses the conventional single-layer ARC in resiliency against angle of incidence. A hybrid light management, employing both the suggested ARC at the front side and MgF2 / Al2O3 dielectric spacer at the rear side, proves to increase Jph of a 750-nm thick CIGS solar cell beyond that of a 1600-nm thick absorber (without light managment). In the rest of the thesis, we take an approach beyond the state-of-the-are architecture of CIGS solar cells and, for the first time, introduce the inter-digitated back-contacted (IBC) structure for CIGS technology. This structure, which no longer suffers from parasitic absorption (associated with the buffer and window layers), is optically studied in Chapter 5. We compare the results with a reference front- and back-contacted (FBC) solar cell with the same absorber volume, and take the Green limit as the benchmark. Two ARC schemes are studied; (i) high-aspect ratio features at the front side of the absorber and, (ii) the as-grown CIGS morphology with optimized MgF2 / Al2O3 layers. Once the optical potential of the IBC CIGS solar cells is realized, we continue our studies with an opto-electrical analysis in TCAD Sentaurus environment (Chapter 6). We not only optimize the geometry of electron- and hole-contacts, the gap between them and the contacts’ period, but also, study the CIGS bandgap grading and its defect density. The electric field map around the gap region is used to highlight the importance of electrical passivation in achieving a high performance. Our models (calibrated with real FBC solar cells fabricated at Solliance at the High-tech campus in Eindhoven) show the high potential of IBC CIGS solar cells for high efficiency PV applications