The current standard practice for forecasting the energy yield of a c-Si system includes the assumption that spectral effects are negligible. In this work, we compute the error that this assumption introduces. We conclude that the error in annual yield forecasts is up to +/- 3% but within +/- 1.5% for most locations. We also conclude that the variability due to daily and seasonal effects is +/- 2% to +/- 3% when the sun's zenith angle is less than 85 degrees. With simulation, we demonstrate how a spectral correction factor f(lambda) can be determined for any system-including complicated installations such as bifacial single-axis trackers-and we describe how f(lambda) depends on latitude, zenith angle, albedo and several atmospheric conditions.

Silicon heterojunction (HJT) solar cells have world-leading efficiencies due to outstanding surface passivation. Yet, maintaining their performance during the lifetime of a photovoltaic module requires excellent quality and stability of the surface regions. It is well known that HJT solar cells can show an increase or reduction in performance under illumination, and this instability has been related to changes in the surface regions. This work investigates the stability of surface passivation in HJT solar cells by modelling the injection-dependent minority carrier lifetime of a range of symmetrically a-Si passivated silicon wafers. Fixed charges and defects at the interface are varied in the model to find the best fit to the injection-dependent lifetime before and after a high-intensity illumination treatment. The results indicate that the laser process induces an increase in field effect passivation at the surface, which is then reduced upon storage in the dark. The results show that lifetime spectroscopy is a useful tool to investigate the nature of a-Si passivation degradation.

Silicon heterojunction technology (HJT) is expected to gain a significant market share in the near future. For HJT to deliver a low levelized cost of electricity (LCOE), it needs to have a high initial efficiency and degrade less than 0.5% relative per year. This work investigates damp heat-induced failure modes in silicon HJT glass-backsheet modules. Four unique failure modes are identified after damp heat (DH) testing: point failure (Type-1); failure around the interconnected regions of the busbars and ribbon wires (Type-2); failure between the busbars (Type -3); and failure at/on the interconnected regions of busbars and ribbon wires (Type-4). The Type-1 failure mode is likely caused by a chemical reaction between surface contaminants (introduced to the cells during handling or characterization before encapsulation) and moisture that increase charge carrier recombination and lead to a loss in maximum power (Pmax) of up to 40%rel in this study. Type-2 and Type-3 failure modes cause Pmax losses of -5%rel and 50%rel, respectively, in this study and can appear due to exposure to soldering flux used for con-necting the ribbon wires and busbars. Finally, the Type-4 failure mode causes a Pmax loss of-16%rel in this study after the DH test. The evidence suggests that this failure mode is likely due to the interaction of acetic acid, generated from a reaction between the encapsulation material and moisture, ribbon wires, and silver paste (busbars), resulting in recombination loss. We believe these failure modes must be well understood and mitigated at preferably the solar cell level to ensure that HJT can meet its LCOE potential.

Phosphorous dopant diffusion profiles feature in many silicon semiconductor devices, including the vast majority of silicon solar cells. Accurate spatially resolved dopant profiling is crucial for understanding the performance of these diffused regions, however, it is very challenging to obtain such profiles in non-planar samples. Scanning electron microscopy for dopant contrast imaging (SEMDCI), where the secondary electron (SE) image contrast is used to determine the dopant level of a semiconductor sample, is an ideal candidate for Si dopant profiling, especially for silicon samples with surface nanotexturing or black silicon (BSi) technology. However, in previous SEMDCI studies, the dopant concentration of heavily doped n-type layers in silicon samples have shown a poor correlation with the SE signal contrast. In this work, 1) good contrast for n-type diffused silicon without contrast-enhancing techniques; 2) a new contrast definition to account for imaging non-uniformities; 3) clear correlations between SE contrast and sample work function for phosphorus-diffused planar silicon specimens across a wide range of emitter profiles; 4) implementation of an empirical baseline correction to normalize scanning electron microscopy image condition variations, are presented. This SEMDCI method is subsequently used for the first time to obtain 2D electron concentration maps for both planar and BSi samples.

Potential-induced degradation of the shunting type (PID-s) is a well-known problem for photovoltaic modules. However, standard PID testing at the cell or module level takes 96 h due to the time it takes for the Na ions to diffuse from the glass to the solar cell through the ethylene-vinyl acetate (EVA) encapsulant. This is too slow to monitor production quality in the fast-paced solar cell industry. Moreover, it results in the encapsulation of solar cells, significantly restricting characterization options after PID testing. In this work, we replace the EVA in the PID testing setup with a salt-enriched hybrid polymer, which provides the solar cell with direct access to sodium (Na) ions while at the same time avoiding encapsulation during the testing process. Solar cells that underwent PID-s testing with the salt-enriched hybrid polymer showed a maximum degradation after only 12 min, over two orders of magnitude faster than standard PID-s testing using EVA, which took 4,300 min (similar to 72 h) to reach maximum degradation. Hence, PID-s testing using the salt-enriched hybrid is significantly faster than conventional techniques and avoids encapsulation, thus allowing for more detailed post-PID cell analysis. This new test shows that some solar cells can recover from PID-s, which was corroborated by extending the standard test to 300 h. This important novel insight shines a new light on the PID-s mechanism and the impact of this failure mode in field operation.

Nanostructured black silicon (b-Si) surfaces with an extremely low reflectance are a promising light-trapping solution for silicon solar cells. However, it is challenging to develop a high-efficiency front-junction b-Si solar cell due to the inferior electrical performance of b-Si emitters, which outweighs any optical gain. This article uses three-dimensional numerical nanoscale simulations, which are corroborated with experiment results, to investigate the effect of the surface nanofeature sizes on the b-Si emitter performance in terms of the sheet resistance (${\rm{R_{sheet}}}$) and the saturation current density (${\rm{J_{0e}}}$). We show that the specific surface area (SSA) is an effective parameter to evaluate the nanofeature size. A shallow surface nanofeature with a large SSA will contribute to a better electrical performance. We will show that b-Si emitter ${\rm{R_{sheet}}}$ measured by a four-point probe is not a measure of the doping level in the nanofeature, but is ruled by the doping level in the underlying substrate region. We also show that a small nanofeature with SSA > 100 mu m-1 and height < 100 nm can lead to a relatively low ${\rm{J_{0e}}}$ (33 fA/cm2 lower than the best b-Si results reported in the literature) by suppressing surface minority carrier density and minimizing the total Auger recombination loss.

Photovoltaic devices operate under a wide range of temperature and illumination conditions. While the temperature coefficients (TC) of crystalline silicon solar cells have been well-studied, there have been only a few investigations regarding the effect of illumination on TCs. In this study, the TCs of the main electrical parameters of various silicon solar cell technologies are first determined. The illumination spectrum dependence of the TC of the short-circuit current and the illumination intensity dependence of the TC of the open-circuit voltage are then investigated. For the latter investigation, a custom-designed temperature-dependent Suns-VOC system is used. It is found that: (1) the TC of the short-circuit current measured using an A-rated spectrum may differ from the TC measured using the AM1.5G spectrum by up to 30%, (2) the TC of the open-circuit voltage of all technologies at 0.001 suns approximately doubles compared to at one-sun, and (3) silicon heterojunction cells seem to have the overall best TC performance at medium to high intensities.

Light-ray tracing (RT) and the transfer matrix method (TMM) allow detailed optical simulation of single-junction silicon and perovskite solar cells, which critically aids device design towards record performance. However, their accuracy is compromised when simulating monolithic perovskite/silicon tandem devices built from textured silicon substrates, due to the resulting complex top-cell morphologies. The associated front surfaces of such tandem devices are typically either roughly conformal to the textured silicon underneath (for thermally evaporated or hybrid-deposited perovskites) or flattened (for solution-processed perovskites). Here, we develop accurate optical models for each configuration. For the conformal-like morphology, we apply a texture-on-texture model to accommodate for local imperfect conformalities. Contrastingly, for the flattened morphology, we develop a multi-subcell model to solve the limitation that films must be conformal in one-step RT + TMM simulations. We verify our models with experimental light absorption and quantum efficiency data for photovoltaic cells and modules in both morphological configurations and identify primary sources of parasitic light absorption. Finally, we extend our simulations to module-scale optics and find that textured module glass can effectively suppress photon escape for both tandem-device morphologies, improving their module performance.

Extracting the parameters, energy level and electron-to-hole capture cross-section ratio, of efficiency-limiting bulk defects in silicon solar cells is a critical step in identifying those defects and potentially eliminating their impact. Typically, this is achieved on specially prepared test structures. However, in some cases, this is not possible, especially in mass production lines when only completed solar cells are available. In this study, a method that is based on temperature-dependent Suns-Voc measurements is introduced to extract the defect parameters in metalized solar cells. The method is validated by comparing the parameters of the boron-oxygenrelated defect extracted from cells and those extracted from wafers using the commonly used temperatureand injection-dependent lifetime spectroscopy. It is shown that this method has the benefit of a more accurate lifetime at low injection levels compared with photoconductance-based lifetime measurement since it is not impacted by minority carrier traps. The proposed technique is then applied to determine the parameters of the defect causing light-induced degradation in gallium-doped silicon solar cells. We determined an energy level, with respect to the intrinsic level, of -0.26 +/- 0.04 eV and a capture cross-section ratio of 34 +/- 2 for this defect. Finally, a sensitivity analysis is performed by considering the system's limited measurement temperature range. The findings demonstrate the potential of the temperature-dependent Suns-Voc method as a fast and easy-toapply method for defect characterization in metalized cells.

The wide variety of silicon materials used by various groups to investigate LeTID make it difficult to directly compare the defect concentrations (N t) using the typical normalised defect density (NDD) metric. Here, we propose a new formulation for a relative defect concentration (β) as a correction for NDD that allows flexibility to perform lifetime analysis at arbitrary injection levels (Δn), away from the required ratio between Δn and the background doping density (N dop) for NDD of Δn/N dop = 0.1. As such, β allows for a meaningful comparison of the maximum degradation extent between different samples in different studies and also gives a more accurate representative value to estimate the defect concentration. It also allows an extraction at the cross-over point in the undesirable presence of iron or flexibility to reduce the impact of modulation in surface passivation. Although the accurate determination of β at a given Δn requires knowledge of the capture cross-section ratio (k), the injection-independent property of the β formulation allows a self-consistent determination of k. Experimental verification is also demonstrated for boron-oxygen related defects and LeTID defects, yielding k-values of 10.6 ± 3.2 and 30.7 ± 4.0, respectively, which are within the ranges reported in the literature. With this, when extracting the defect density at different Δn ranging between 1014 cm−3 to 1015 cm−3 with N dop = 9.1 × 1015 cm−3, the error is less than 12% using β, allowing for a greatly improved understanding of the defect concentration in a material.

Illuminated solar cells are susceptible to various degradation mechanisms that can act to reduce the total energy yield when deployed. One potentially severe form is an increase in carrier recombination in the surface regions. This effect has been reported at both the undoped rear surface and phosphorous diffused emitter of PERC solar cells. This work investigates the influence of a range of surface conditions on the surface-related degradation (SRD) behaviour in PERC solar cells. It is shown that SRD is strongly affected by the doping profile of phosphorous emitters, the use of thin thermal oxides with SiNx:H dielectric passivation layers, the substrate material, and the configuration of the rear surface passivation. It finds that more lightly doped emitters result in more front side SRD, with its extent increasing with the introduction of the SiO2/SiNx:H surface passivation layers. Czochralski silicon (Cz-Si) wafers were observed to be significantly more susceptible to surface degradation than multi-crystalline silicon (mc-Si) wafers, which we attribute to less trapping of hydrogen in the bulk of those substrates. On the rear side of PERC cells, surface degradation was only observed in structures that incorporated the combination of SiO2/SiNx:H rear layers. No SRD was observed in the existing Al2O3/SiNx:H technology used in the industrial PERC cells studied. However, the results presented have implications for future commercial solar cell technologies, which are transitioning towards lightly doped emitters and commonly incorporate thermal oxides for surface passivation.

The optimal tilt β opt of the modules in a single-axis tracker is often determined by assuming all sunlight is direct and ground reflectance is zero. Prior works, however, have demonstrated that β opt is smaller when diffuse light is significant. In this article, we determine how β opt decreases as the conditions change from clear sky to overcast, accounting for many complicated effects like row-to-row shading of diffuse light, ground reflection, structural shading, cell-to-cell mismatch, and angular and spectral dependencies. We find that when compared to monofacial systems, bifacial systems tend to have a higher β opt when it is cloudier and a lower β opt when it is sunnier. We also quantify the increase in annual yield that arises when accounting for indirect light at three example sites with different climates, finding it to be ∼30% lower for a bifacial system than a monofacial system; the gain was 0.8%–1.5% for bifacial systems and 1.1%–2.1% for monofacial systems, where the highest gain was attained in the cloudiest climate.

Passivated emitter and rear cell (PERC) with laser‐doped selective emitter (SE) has become mainstream in the PV industry. In this work, we report a solid strategy to realize heteroface monocrystalline silicon (mono‐Si) wafers for PERC‐SE solar cells by employing alkaline polishing to form a polished surface for the rear side and well‐established metal‐catalyzed chemical etching to form a honeycomb texture for the front side in one wet process successively. The key to success lies in the fact that the two back‐to‐back wafers inserted into one slot in the cassette are tightly attached together in MCCE etching so that only the exposed surfaces are etched to form textures, while the rear polished surfaces are still retained to avoid wrap‐around etching. With the strategy, the mono‐Si PERC‐SE solar cells achieve an average efficiency of over 22.0%, no poorer than that of the reference system (traditional alkaline texturing and rear acidic polishing), and have good light trapping capability for oblique incident light. Moreover, the total Si removal in the novel process is only ~0.4 g, which is far less than that in the traditional process. More importantly, the strategy can also double the throughput of existing texturing processes and significantly reduce the amount of etching waste. Therefore, the work is expected to provide a promising way to mass produce efficient mono‐Si PERC‐SE solar cells with a superior rear surface, achieved without increasing the number of processing steps, and lower cost.

Black silicon (BSi) is a branch of silicon material whose surface is specially processed to a micro/nanoscale structure, which can achieve ultra‐low reflectance or ultra‐high electrochemical reactivity. The diversity and complex surface structures of BSi make it challenging to commercialize BSi devices. Modeling and simulation are commonly used in the semiconductor industry to help in better understanding the material properties, predict the device performance, and provide guidelines for fabrication parameters’ optimization. The biggest challenge for BSi device modeling and simulation is obtaining accurate input surface morphological data. In this work, the 3D models of challenging BSi textures are compared as obtained by atomic force microscopy (AFM) and plasma focused ion beam (PFIB) tomography techniques. In previous work, the PFIB tomography workflow toward the application of surface topography is optimized. In this work, the 3D models obtained from both AFM and PFIB are comprehensively compared, by using the surface models as inputs for finite‐difference time‐domain‐based optical simulation. The results provide strong evidence that PFIB tomography is a better choice for characterizing highly roughened surface such as BSi and provides surface 3D models with better reliability and consistency.

We measure the module temperature $T_{m}$ in 1P and 2P single-axis trackers, analyzing how $T_{m}$ depends on wind speed and direction, as well as on irradiance, ambient temperature and module tilt ß. On a clear day, we find that the typical temperature variation within a tracker is 1-4°C for 1P and 2–6 °C for 2P trackers, where the coolest region tends to be nearest the torque tube. Whether for 1P or 2P, we find that when ß < ~25°, the wind cools the windward side of the tracker by 1-1.5 °C more than the leeward side; but when ß > ~25°, the upper side is 0.5-1.5 °C cooler than the lower side, irrespective of wind direction (for wind speeds < 5 m/s), We also find that the commonly used NOCT and Faiman temperature models overpredict $T_{m}$ by, on average, 7.4 °C and 3.3 °C. Even after calibrating these models to our trackers, they only predict $T_{m}$ at any given time to ±6.6 °C with 95% confidence. Without adding any free variables, the modelling accuracy is improved to ±3.8 °C by accounting for radiative loss to the sky and transient effects; the accuracy is improved further to ±2.8 °C by accounting for module tilt, wind direction and ground temperature. This study expands upon the PV industry's understanding of how single-axis trackers are influenced by wind speed, wind direction, and tilt, and it refines our ability to accurately predict $T_{m}$ of FTC's Voyager 2P tracker.

At present, the commercially dominant and rapidly expanding PV-device technology is based on the passivated emitter and rear cell (PERC) design developed at UNSW. However, this technology has been found to suffer from a carrier-induced degradation commonly referred to as 'light- and elevated temperature-induced degradation' (LeTID) and can result in up to 16% relative performance losses. LeTID was recently shown to occur in almost every type of silicon wafer, independent of the doping material. Even though the degradation mechanism is known to recover under normal operation conditions, it is a lengthy process that drastically affects the energy yield, stability and, ultimately, the levelized cost of electricity (LCOE) of installed systems. Despite the joint effort of many research groups, the root cause of the degradation is still unknown. Here, we provide an overview of the existing literature and describe key LeTID characteristics and how these have led to the development of various theories of the underlying mechanism. Further, given the continuously appearing and strong evidence of hydrogen involvement in LeTID, many mitigation methods concerning hydrogenation have been suggested. We discuss such reported methods, bearing in mind crucial consumer necessities in terms of sustained cell performance and minimised LCOE.

Anti-reflection coatings are used on 92% of today’s module glass to reduce the front-surface reflection and increase the power output of the module. Currently, most anti-reflection coatings are designed to maximize transmittance at normal incidence to improve the power output at standard test conditions. However, in the field, where the meteorological conditions vary widely in time and space, it is not clear whether a maximum power gain at standard test conditions leads to maximum energy yield. Here, we use SunSolve Yield, a ray-tracing software, to investigate the optimum anti-reflection coating design by simulating the energy yield from a PV system under real-world conditions. We found that manufacturers could immediately provide a 0.3% increase in energy yield relative to their products today by simply increasing the ARC thickness by 20–40 nm thicker.

Black silicon (b-Si) surfaces typically have a high density of extreme nanofeatures and a significantly large surface area. This makes high-quality surface passivation even more critical for devices such as solar cells with b-Si surfaces. It has been hypothesized that conformal dielectrics with a high fixed charge density (Q(f)) are preferred as the nanoscale features of b-Si result in a significant enhancement of field-effect passivation. This article uses 1-D, 2-D, and 3-D numerical simulations to study surface passivation of b-Si, where we particularly focus on the charge carrier control by vertical bar Q(f)vertical bar up to 1 x 10(13) cm(-2) under accumulation conditions. We will show that there is a significant space charge region compression in b-Si nanofeatures, which affects the charge carrier population control formoderate vertical bar Q(f)vertical bar up to approximate to 1x10(12) cm(-2). The average surfaceminority charge carrier density can be reduced by 70% in some cases, resulting in an equivalent reduction in area-normalized surface recombination losses if the effective surface recombination velocity (S-eff) is limited by minority carriers. This provides a possible solution for the empirical S-eff proportional to 1/Q(f)(4) reported previously. We will also show that the situation is more complicated for surface passivation films where the ratio between the electron and hole capture cross section (sigma(n)/sigma(p)) is higher than 10 for p-type surfaces. For commonly used surface passivation films with a vertical bar Q(f)vertical bar larger than approximate to 1 x 10(12) cm(-2), there is little space charge compression for b-Si. Consequently, S-eff simply scales with the surface area, i.e., there is no enhanced reduction of surface recombination by field-effect passivation on b-Si.

Light and elevated-temperature-induced degradation (LeTID) is a well-known phenomenon that reduces the bulk lifetime in silicon wafers. The cause of this degradation mechanism is still under investigation. However, a wide range of empirical trends that correlate LeTID with multiple physical and processing parameters have been reported, including the observation that wafers thinner than 120 μm do not show significant LeTID. In this work, we extend that study by varying the thickness of the wafers, the temperature of the firing step, and testing LeTID at the accelerated stability testing conditions. We demonstrate that the extent of degradation reduces with the thickness of the wafer, in agreement with the earlier work. However, silicon wafers with a thickness below 120 μm still suffer from LeTID when fired at sufficiently high temperatures, demonstrating that thinner wafers are not inherently immune to LeTID. By performing accelerated testing using a high-intensity laser and fitting the degradation and regeneration data, we observe that thinner wafers do not necessarily exhibit a faster recovery, as suggested earlier. However, their reduced degradation extent could be a consequence of relatively higher out-diffusion of hydrogen per unit volume in thinner wafers during firing. We further report that the method used for thinning the wafers results in a variation in the surface morphology of the samples, and that may partly be responsible for the observed correlation between the thickness of the wafers and LeTID extent. Finally, we discuss how these new findings can be explained by the involvement of hydrogen and other impurities in LeTID.

The integration of nanotextured black silicon (B-Si) into solar cells is often complicated by its enhanced phosphorus doping effect, which is typically attributed to increased surface area. In this article, we show that B-Si's surface-to-volume ratio, or specific surface area (SSA), which is directly related to surface reactivity, is a better indicator of reduced sheet resistance. We investigate six B-Si conditions with varying dimensions based on two morphology types prepared using metal-catalyzed chemical etching and reactive-ion etching. We demonstrate that for a POCl3 diffusion, B-Si sheet resistance decreases with increasing SSA, regardless of surface area. 2-D dopant contrast imaging of different textures with similar surface areas also indicates that the extent of doping is enhanced with increasing SSA. 3-D diffusion simulations of nanocones show that both the extent of radial doping within a texture feature and the metallurgical junction depth in the underlying substrate increase with increasing SSA. We suggest SSA should be considered more readily when studying B-Si and its integration into solar cells.