We investigate hole-selective passivating contacts that consist of an interfacial layer of silicon oxide (SiOx) and a layer of boron-doped SiCx(p). The fabrication process of these contacts involves an annealing step at temperatures above 750 °C which crystallizes the initially amorphous layer and diffuses dopants across the interfacial oxide into the wafer to facilitate charge transport, but it can also disrupt the SiOx layer necessary for wafer-surface passivation. To investigate the transport mechanism of the charge carriers through the selective contact and its changes during the annealing process, we utilize various characterization methods, such as transmission electron microscopy, micro Raman spectroscopy, and conductive atomic force microscopy. Combining the latter with a sequential removal of material, we assemble a tomographic reconstruction of the crystallized layer that reveals the presence of preferential vertical transport channels.

We present a p-type passivating rear contact that complies with integration into standard solar cell manufacturing with phosphorus-diffused front side. Our contact structure consists of a thin SiOx tunneling layer grown by wet chemistry and a stack of layers deposited in one single run by plasma-enhanced chemical vapor deposition. The layers of the stack were tailored to protect the interfacial oxide layer, to act as a source for boron diffusion into the wafer and to connect to the external metallisation with low contact resistivity. We found that this stack tolerated annealing at 900 °C over a wide range of dwell times: for 15 min anneals we obtained dark saturation current densities (J o ) as low as 10 fA·cm -2 (after hydrogenation) and after 12-fold increase of the annealing time to 180 min, J 0 was only increased to 12 fA·cm -2 . These values corresponded to implied open circuit voltages (iV oc ) of 718 and 715 mV, respectively. To test passivating rear contacts under realistic operation conditions, we combined them with an n-type heterojunction into hybrid solar cells. With conversion efficiencies abovementioned 22% and V oc > 705 mV, these devices demonstrated high level of rear surface passivation. Finally, we demonstrated the integration of the hole selective rear contact with a POCl 3 diffusion process. To this end, we added a phosphorus diffusion barrier to our layer stack by depositing one additional layer of amorphous SiO x on top of the stack. For symmetric samples with this layer structure on both sides, we observed iVoc values of 714 and 712 mV on n- and p-type silicon wafers after hydrogenation, respectively. Co-diffused cells with POCl 3 front diffused emitter and rear passivating contact resulted so far in efficiencies of 20.4% and 20.1% for n- and p-type wafers, respectively.

This work presents a systematic analysis of the transport mechanism and surface passivation of tunneling oxide (SiO2)/p-type poly-silicon (poly-Si(p)) junctions applied to p-type crystalline silicon (c-Si) solar cells by means of TCAD numerical simulations. We report on the impact of the buried doped region (BDR) in the c-Si wafer on the transport and passivation of SiO2/poly-Si(p) junctions. We show that a BDR is not necessary for carrier selective contacts (CSCs) with a tunnel oxide thinner than 1.2 nm and for surface recombination velocity at SiO2/c-Si interface below 1.10(3) cm/s. Then, we explore alternative semiconductors to poly-Si for tunnel oxide passivating contacts. We rind that 3C-SiC(p) is a promising candidate thanks to its valence band offset with respect to silicon, driving the wafer surface into a condition of strong accumulation. We show that excellent SiO2/3C-SiC(p) junctions are obtained for doping density of the 3C-SiC(p) larger than 5.10(19) cm(-3) and for SiO2 thinner than < 1.2 nm. Finally, with the aim of deriving guidelines for material selection, we present an investigation on the influence of the electron affinity and bandgap of the semiconductor layer forming the passivating contact, demonstrating that conversion efficiency is maximized for built-in voltages between 0.4 and 2.6 eV.

We present an electron selective passivating contact based on a tunneling SiOx capped with a phosphorous doped silicon carbide and prepared with a high-temperature thermal anneal. We investigate in detail the effects of the preparation conditions of the SiCx (n) (i.e., gas flow precursor and annealing temperature) on the interface recombination rate, dopant in-diffusion, and optical properties using test structures and solar cells. On test structures, our investigation reveals that the samples annealed at temperatures of 800-850 degrees C exhibit an increased surface passivation toward higher gas flow ratio (r = CH4/(SiH4 + CH4)). On textured and planar samples, we obtained best implied open-circuit voltages (i-V-OC) of 737 and 746 mV, respectively, with corresponding dark saturation current densities (J(0)) of similar to 8 and similar to 4 fA/cm(2). The SiCx (n) layers with different r values were applied on the textured front side of p-type c-Si solar cells in combination with a boron-doped SiCx(p) as rear hole selective passivating contact. Our cell results show a tradeoff between V-OC and short-circuit current density (J(SC)) dictated by the C-content in the front-side SiCx (n). On p-type wafers, best V-OC = 706 mV, FF = 80.2%, and J(SC) = 38.0 mA/cm(2) with a final conversion efficiency of 21.5% are demonstrated for 2 x 2 cm(2) screen-printed cells, with a simple and patterning-free process based on plasma depositions and one annealing step 800 degrees C < T < 850 degrees C for the formation of both passivating contacts.

Passivating contacts are indispensable for achieving high conversion efficiency in crystalline-silicon solar cells. Their realization and integration into a convenient process flow have become crucial research objectives. Here, we report an alternative passivating contact that is formed in a single post-deposition annealing step called ‘firing’, an essential step for current solar cell manufacturing. As firing is a fast (<10 s) and high-temperature (>750 °C) anneal, the required microstructural and electrical properties of the passivating contact are stringent. We demonstrate that tuning the carbon content of boron-doped silicon-based thin films inhibits firing-induced layer delamination without preventing a partial crystallization. The latter promotes charge-carrier selectivity, even in the absence of a diffused doped region beyond the oxide, by inducing hole accumulation near the wafer surface. We fabricated proof-of-concept solar cells employing the developed technology, demonstrating an open circuit voltage of 698 mV and an efficiency of 21.9%, and show how it could be a drop-in replacement for today’s rear contacts based on locally opened dielectric passivation stacks. To minimize recombination losses and therefore increase the conversion efficiency of crystalline silicon solar cells, researchers have relied on passivating contacts. Here, the authors demonstrate a hole-selective passivating contact that exploits the firing step currently employed in industrial manufacturing.

We analyze the recombination properties of passivating electron selective contacts based on nanostructured silicon oxide. Our contact design is based on an interfacial buffer oxide capped with a bilayer structure of phosphorus-doped silicon oxide and silicon which is annealed at 900 degrees C. We investigate in detail the effects of the initial dopant concentration in the bilayer and of the anneal dwell time on dopant in-diffusion, contact formation, and interface recombination. Our investigation addresses also the hydrogenation of interface defects and the effect of indium-tinoxide (ITO) sputtering, allowing us to separate the interplay between enhanced field-effect passivation, Auger recombination, and interface recombination. After thermal annealing, the passivating electron selective contact presented here attains a saturation current density (J(0)) of 12.4 fA cm(-2) for medium doping, which improves further upon hydrogenation to J(0) = 8.1 fA cm(-2). For specific contact resistances <500 m Omega cm(2), however, higher doping concentrations are required. For those doping concentrations, the saturation current density is 13.9 fA cm(-2) and increases by 10% upon sputter-deposition of an ITO layer on top of the electron selective stack.

In devices with intrinsic amorphous silicon layer on a crystalline silicon substrate, the light absorbed in the amorphous layer can be weakly electronically coupled into the silicon base. Such carrier injection has previously been reported from measurements on finished devices containing stacks of intrinsic and doped amorphous silicon layers. Here, we use spectral response of photoluminescence, a contactless approach, to investigate this carrier injection on significantly simpler structures. In such devices, the effect of absorption in the front layer can be measured by the internal quantum efficiency. A highly absorbing front layer is expected to cause a drop in the quantum efficiency at short wavelengths. However, if electron-hole pairs that are generated in the front layer are subsequently injected into the base, the optical losses will be reduced, resulting in a partial recovery of the quantum efficiency at short wavelengths. Here, we quantify the efficiency of carrier injection from the intrinsic amorphous silicon front layer to the crystalline silicon base, by measuring the spectral response of photoluminescence heterojunction test structures. For devices with just an intrinsic amorphous silicon layer, the carrier injection from the layer was found to be close to unity.

We present electron- and hole-selective passivating contacts based on wet-chemically grown interfacial SiOx and overlying in-situ doped silicon carbide (SiCx) deposited by plasma-enhanced chemical vapor deposition. After annealing at 850 degrees C, excellent surface passivation on the p-type planar crystalline silicon wafer is obtained for both electron- and hole-selective contacts. Their potential is demonstrated at the device level by employing a simple process flow, in which the junction formation of the two polarities is achieved with a single coannealing step. Both-side-contacted patterning-free planar p-type cells with an area of 4 cm(2) and screen-printed metallization reach a fill factor of 83.4% and a open-circuit voltage of 726 mV. Zirconium-doped indium oxide with excellent optoelectrical properties is used as a front electrode. The decrease in the parasitic absorption in the front electrode results in higher photogenerated current. By realizing front-side-textured and rear-side-planar p-type cells, an efficiency of up to 22.6% is achieved.

Intrinsic amorphous silicon provides excellent surface passivation on crystalline silicon. It has previously been shown, that carriers that are photo generated in the amorphous silicon can be efficiently electronically injected into the crystalline silicon. A method to quantify the efficiency of such carrier injection using the spectral response of photoluminescence has recently been demonstrated. This is a contactless method and it can be applied to incomplete device structures. Here we use this technique to measure partially processed heterojunction devices with different capping layers to quantify their impact on the carrier injection efficiency. Silicon nitride capping on amorphous silicon is shown to have minimum impact on the high carrier injection efficiency of the amorphous layer but doped amorphous capping layer on the other hand were seen to have a strong effect on the carrier injection efficiency. A model was developed to understand the material properties of the amorphous layer. The reduction in carrier injection efficiency with doped amorphous silicon capping layers were attributed to the large defects in the doped layer.

Transition metal oxides based high efficiency silicon heterojunction (SHJ) solar cells have emerged as promising candidate due to their low manufacturing cost and avoidance of poisonous dopant gases. Temperature dependent J V curves in such solar cells could reveal the physical insight about the carrier collection mechanism. In this work, the temperature dependence of Molybdenum Oxide (MoOx) based SHJ is modelled and compared with the experimental J-V curves (under both dark and 1-sun illumination). This study validates the hole collection mechanism using MoOx and explains the disappearance of the kink in J-V curves at higher temperatures. In addition, we have extracted the temperature coefficient for MoOx-SHJ solar cells and found that they are almost identical to that of standard SHJ cells which supports MoOx-SHJ cells as a promising candidate for future low cost high efficiency solar cells.

Silicon heterojunction solar cells enable high conversion efficiencies, thanks to their passivating contacts which consist of layered stacks of intrinsic and doped amorphous silicon. However, such contacts may reduce the photo current, when present on the illuminated side of the cell. This motivates the search for wider bandgap contacting materials, such as metal oxides. In this paper, we elucidate the precise impact of the material parameters of MoOx on device characteristics, based on numerical simulations. The simulation results allow us to propose design principles for hole-collecting induced junctions. We find that if MoOx has a sufficiently high electron affinity (>= 5.7 eV), direct band-to-band tunneling is the dominant transport mechanism; whereas if it has a lower electron affinity (< 5.7 eV), trap-assisted tunneling dominates, which might introduce additional series resistance. At even lower electron affinity, S-shaped J-V curves may appear for these solar cells, which are found to be due to an insufficient trap state density in the MoOx film in contrast to the expectation of better performance at low trap density. These traps may assist carrier transport when present near the conduction band edge of the MoOx film. Our simulations predict that performance optimization for the MoOx film has to target either 1) a high electron affinity and a moderate doping density film or, 2) if the electron affinity is lower than the optimum value, a high defect density not exceeding the doping density inside the film.

Intrinsic amorphous silicon provides excellent surface passivation on crystalline silicon. It has previously been shown, that carriers that are photo generated in the amorphous silicon can be efficiently electronically injected into the crystalline silicon. A method to quantify the efficiency of such carrier injection using the spectral response of photoluminescence has recently been demonstrated. As this is a contactless method, it can be applied to incomplete device structures. Here we, use this technique to measure partially processed heterojunction devices with different capping layers to quantify their impact on the carrier injection efficiency. Silicon nitride capping on amorphous silicon is shown to have minimum impact on the high carrier injection efficiency of the amorphous layer whereas phosphorus doped amorphous capping layers on the other hand were seen to have a strong effect on the carrier injection efficiency.

The charge carrier transport mechanism of passivating contacts which feature an ultra-thin oxide layer is investigated by studying temperature-dependent current-voltage characteristics. 4-Terminal dark J-V measurements at low temperatures reveal non-linear J-V characteristics of passivating contacts with a homogeneously grown silicon oxide, which result in an exponential increase in contact resistance towards lower temperature. The attempt to describe the R(T) characteristic solely by thermionic emission of charge carriers across an energy barrier leads to a significant underestimation of the resistance by several orders of magnitude. However, the data can be described properly with the metal-insulator-semiconductor (MIS) theory if tunneling of charge carriers through the silicon oxide layer is taken into account. Furthermore, temperature-dependent light J-V characteristics of solar cells featuring passivating contacts at the rear revealed a FF drop at T < 205 K, which is near the onset temperature of the exponential increase in contact resistivity.

Cesium–formamidinium-based mixed-halide perovskite materials with optical band gaps ranging from 1.5 to 1.8 eV are investigated by variable-angle spectroscopic ellipsometry. The determined complex refractive indices are shown to depend on the fabrication procedure and environmental conditions during processing. This data is complemented by additional optical and structural characterization, as well as the demonstration of efficient perovskite solar cells. Finally, the data is used in optical simulations to provide guidelines for the optimization of perovskite/silicon tandem solar cells.

Molybdenum oxide (MoOX) combines a high work function with broadband optical transparency. Sandwiched between a hydrogenated intrinsic amorphous silicon passivation layer and a transparent conductive oxide, this material allows a highly efficient hole‐selective front contact stack for crystalline silicon solar cells. However, hole extraction from the Si wafer and transport through this stack degrades upon annealing at 190 °C, which is needed to cure the screen‐printed Ag metallization applied to typical Si solar cells. Here, we show that effusion of hydrogen from the adjacent layers is a likely cause for this degradation, highlighting the need for hydrogen‐lean passivation layers when using such metal‐oxide‐based carrier‐selective contacts. Pre‐MoOX‐deposition annealing of the passivating a‐Si:H layer is shown to be a straightforward approach to manufacturing MoOX‐based devices with high fill factors using screen‐printed metallization cured at 190 °C.

We present a detailed optimization of a hole selective rear contact for p–type crystalline silicon solar cells which relies on full-area processes and provides full-area passivation. The passivating hole-contact is based on a layer stack comprising a chemically grown thin silicon oxide, an intrinsic silicon interlayer, and an in-situ boron-doped non-stoichiometric silicon-rich silicon-carbide layer on top. After deposition, the structure is annealed at 775–900°C to diffuse dopant impurities to the c-Si wafer and a hydrogenation step is carried out. It is shown that hydrogenation is essential to obtain high quality surface passivation. In particular, we compare the effect of annealing in forming gas and annealing with a silicon-nitride overlayer as hydrogen source. We present a systematic optimization of the hole-selective contact, for which we varied the doping concentration, annealing parameters and report the implied open circuit voltage (iVoc) and combined specific contact resistivity (ρc). It is observed that for highly doped layers the optimum annealing temperature for high quality surface passivation is 800°C while for lowly doped layers the optimum annealing condition shifts to 850°C. Excellent surface passivation and efficient current transport is evidenced by an iVoc value of 718mV which corresponds to a saturation current density (J0) of 11.5fA/cm2 and a ρc of 17mΩcm2 on p−type wafers. Moreover, the evolution of the boron diffusion profiles with different annealing conditions is investigated. Finally, we demonstrate proof-of-concept p−type hybrid solar cells employing the full-area hole-selective rear contact presented here and standard heterojunction front electron contact. The excellent efficiency potential of our passivating rear contact is highlighted by conversion efficiencies up to of 21.9%, enabling Voc of 708mV, FF of 79.9% and Jsc of 38.7mA/cm2.

We present the development of passivating contacts for high-efficiency silicon solar cells using silicon oxide (SiO x ) and silicon carbide (SiCx)-based layers. We discuss a comprehensive optimization of a SiCx-based passivating hole contact reaching implied open circuit voltages >715 mV. In addition, we introduce a passivating hole contact based on nanocrystalline SiO x (nc-SiO x ) targeting compatibility with higher process temperatures as well as increased optical transparency for front side application. First planar test devices employing nc-SiO x -based passivating contacts for both charge carrier types are presented, yielding short-circuit current densities >34 mA/cm 2 and fill factors >78%, showing that efficient current extraction is possible despite the added SiO x phase and also indicating potential optical advantages of the concept.

Highly-transparent carrier-selective front contacts open a pathway towards entirely dopant free Si solar cells. Holeselective a-Si:H/MoO x /ITO front contact stacks were already successfully applied in such novel devices. However, for optimum device performance, further improvements are required: We evaluate the use of the high-work-function material WO x as a replacement for MoO x in an attempt to reduce optical absorption losses. In addition, we investigate the use of thin hydrogenated SiO X instead of a-Si:H, and the impact of the residual pressure for MoO x evaporation.

Optical absorptance spectroscopy of polycrystalline CH3NH3PbI3 films usually indicates the presence of a PbI2 phase, either as a preparation residue or due to film degradation, but gives no insight on how this may affect electrical properties. Here, we apply photocurrent spectroscopy to both perovskite solar cells and coplanar-contacted layers at various stages of degradation. In both cases, we find that the presence of a PbI2 phase restricts charge-carrier transport, suggesting that PbI2 encapsulates CH3NH3PbI3 grains. We also find that PbI2 injects holes into the CH3NH3PbI3 grains, increasing the apparent photosensitivity of PbI2. This phenomenon, known as modulation doping, is absent in the photocurrent spectra of solar cells, where holes and electrons have to be collected in pairs. This interpretation provides insights into the photogeneration and carrier transport in dual-phase perovskites.

The use of passivating contacts compatible with typical homojunction thermal processes is one of the most promising approaches to realizing high-efficiency silicon solar cells. In this work, we investigate an alternative rear-passivating contact targeting facile implementation to industrial p-type solar cells. The contact structure consists of a chemically grown thin silicon oxide layer, which is capped with a boron-doped silicon-rich silicon carbide [SiCx(p)] layer and then annealed at 800-900 °C. Transmission electron microscopy reveals that the thin chemical oxide layer disappears upon thermal annealing up to 900 °C, leading to degraded surface passivation. We interpret this in terms of a chemical reaction between carbon atoms in the SiCx(p) layer and the adjacent chemical oxide layer. To prevent this reaction, an intrinsic silicon interlayer was introduced between the chemical oxide and the SiCx(p) layer. We show that this intrinsic silicon interlayer is beneficial for surface passivation. Optimized passivation is obtained with a 10-nm-thick intrinsic silicon interlayer, yielding an emitter saturation current density of 17 fA cm-2 on p-type wafers, which translates into an implied open-circuit voltage of 708 mV. The potential of the developed contact at the rear side is further investigated by realizing a proof-of-concept hybrid solar cell, featuring a heterojunction front-side contact made of intrinsic amorphous silicon and phosphorus-doped amorphous silicon. Even though the presented cells are limited by front-side reflection and front-side parasitic absorption, the obtained cell with a Voc of 694.7 mV, a FF of 79.1%, and an efficiency of 20.44% demonstrates the potential of the p+/p-wafer full-side-passivated rear-side scheme shown here.