No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
P-Type Versus n-Type Silicon Wafers: Prospects for High-Efficiency Commercial Silicon Solar Cells
Abstract
Chemical and crystallographic defects are a reality of solar-grade silicon wafers and industrial production processes. Long overlooked, phosphorus as a bulk dopant in silicon wafers is an excellent way to mitigate recombination associated with these defects. This paper details the connection between defect recombination and solar cell terminal characteristics for the specific case of unequal electron and hole lifetimes. It then looks at a detailed case study of the impact of diffusion-induced dislocations on the recombination statistics in n-type and p-type silicon wafers and the terminal characteristics of high-efficiency double-sided buried contact silicon solar cells made on both types of wafers. Several additional short case studies examine the recombination associated with other industrially relevant situations-process-induced dislocations, surface passivation, and unwanted contamination. For the defects studied here, n-type silicon wafers are more tolerant to chemical and crystallographic defects, and as such, they have exceptional potential as a wafer for high-efficiency commercial silicon solar cells
... Although p-type wafers displayed poorer electrical properties, they exhibited better tolerance to radiation experienced in space, extra-terrestrial applications as such were the primary applications during the 1960s. 114,115 There are several technical factors that also make p-type advantageous for manufacture of industrial cells. First, the segregation coefficient of boron at the silicon liquid-solid interface is much higher than phosphorous. ...
... Second, minority carriers in p-type are electrons, which have a higher mobility than holes by approximately a factor of 3 (depending on the doping level). 115 This can relax the requirements for the bulk diffusion length compared to n-type. Additionally, forming an n-type emitter in p-type silicon is less technically challenging than the formation of a boron emitter. ...
This review details the design considerations for the bottom silicon cell in perovskite/silicon tandems. The review highlights the shift in mindset required when transitioning to the mass production of tandem solar cells.
... Although in N-type silicon the bipolar diffusion coefficient is three times smaller than in P-type, the lifetime of N-type silicon carriers is in the order of milliseconds (i.e., an order of magnitude higher than in boron-doped P-type Si), and longer diffusion lengths can be achieved [33]. For this reason, N-type monocrystalline wafers are suitable as a starting material for high-efficiency solar cells. ...
... N-type-based crystalline silicon technologies-interdigitated back contact (IBC), heterojunction technology (HJT), and tunnel oxide passivated contact (TOPCon)-have higher efficiencies (23-25%) and a lower temperature coefficient in comparison with boron-doped P-type-based technologies [33,34]. However, their market share remains relatively low (5% in 2015, 5% in 2019, and 5% in 2020), probably due to the higher costs of fabrication [35] (especially in the case of IBC and HJT) and consequently, they come with a higher price. ...
The key components of photovoltaic (PV) systems are PV modules representing basic devices, which are able to operate in outdoor conditions for a long time. PV modules can be manufactured from different materials using different production technologies. The main criterion supporting or limiting the successful placement of specific technologies on the market is the price of electricity produced by PV systems. The levelized cost of energy (LCOE) method considers investment costs, operating costs, and the total energy produced during a PV system’s service life. The influence of price, efficiency, and service life of PV modules on the LCOE (together with the availability of materials) sets limits for applicable technologies. Increasing the efficiency of the modules from 21% to 23% could lead to a reduction of the area-dependent part of the PV system costs by 8.7%. Extending the service life from 25 to 30 years could reduce the LCOE by about 10%. As shown in the work, wafer-based crystalline silicon technologies best meet these criteria due to their high efficiency, low costs, long service life, and the availability of materials at present. Technological innovations make it possible to increase the efficiency of the modules closer to the physical limits and to extend the service life of the modules.
... High-efficiency silicon solar cell is one of the promising candidates to solve the energy crisis all over the world [1,2]. So far, one of the major bottlenecks limiting the efficiency of solar cells is optical losses stemmed from serious reflection, insufficient transparency and spectral mismatch, in addition to thermalization losses from high energy charge carriers with the crystal lattice of the absorber in the range of short-wavelength [3][4][5][6][7]. ...
... Antireflection coating in front of active layer, typically composed of a single or multiple dielectric thin films following the optical interference modeled by the Fresnel equation [8,9], is generally employed to minimize the irradiation reflection and enhance light transmittance for silicon solar cells. According to Fresnel equation [8], the reflectivity can be estimated by R = ( n 2 a − n0ns n 2 a +n0ns ) 2 , which indicates that the reflectivity can be suppressed by decreasing the index of refraction of antireflection layer (n a ) close to the square root of the product of the indices of the surrounding medium (n 0 ) and the beneath substrate (n s ), n 2 a →n 0 n s . Generally, optical elements based on glass or polymers will lead to over 4% reflection loss from air-substrate interface because of their relatively high indices of refraction around 1.5-1.7. ...
Photon management is an efficient route to ameliorate optical issues for improving the efficiency of solar cells. On account of reflection loss and spectral mismatch for silicon solar cells, we herein firstly demonstrate a photon management of combining antireflection and luminescence down-shifting coatings to simultaneously suppress the reflection of incident light with nanoporous structure and improve the spectra response of silicon solar cells in short wavelength with in-situ fabricated perovskite quantum dots composite film. Antireflection film can be obtained by sacrificing Cd x Zn 1-x Se y S 1-y /ZnS quantum dots template in PDMS composite film, which can bring 9.7% improvement in photocurrent for amorphous silicon solar cells. By further combining perovskite quantum dots down-shifting luminescence coating, an absolute improvement of 1.05% in power conversion efficiency was achieved. The photon management combining the antireflection and luminescence down-shifting coatings could provide a low-cost, simple but effective way to address the optical issues and improve the device performance of optoelectronic devices.
... Besides the absence of LID, the efficiency of n-type devices are less affected by metallic contaminants common in silicon solar cell production 9,10 . For instance, the interstitial Fe is more effective in capturing electrons than holes due to the positive charge state. ...
... Figure 6 illustrates the distribution of the minority carrier diffusion length obtained by the LBIC technique of the thin n + np + device with P diffused at 855 °C and T Firing = 840 °C. The average value of the diffusion length (L DAverage ) was of around 1200 μm, that is, twelve times the wafer thickness and corresponds to the high minority carrier lifetime of around 1.1 ms 9,14 . ...
The silicon solar cells achieved relatively low prices in the last years and to introduce a new structure in the PV industry, the amount of silicon per watt has to be reduced, requiring a cost-effective manufacturing process. The use of n-type solar grade silicon has the advantages of presenting higher minority carrier lifetime than p-type one and the absence of the boron-oxygen defects. The aim of this paper is to present the development of 100 μm thick n+np+ silicon solar cells with a selective p+ rear emitter formed by boron deposited by spin-on and an Al/Ag grid deposited by screen-printing. The firing temperature of Ag/Al (rear face) e Ag (front face) was optimized and the temperature of 840 °C produced the devices with higher efficiency. The solar cells presented efficiencies of 16%, achieving a low silicon consumption of 1.6 g/W, 40% lower than thick p-type devices produced by the same process.
... Besides, about 95% of the production is based on p-type c-Si wafers. 1 This situation results from technological simplicity, since the superior lifetime and stability of n-type c-Si has been demonstrated. 5 However, a simple full-area solution as the Al-BSF contact has no equivalent on n-type c-Si. Thus, developing more cost-effective solar cell structures for n-type wafers is particularly appealing. ...
... Of course, the magnitude of the dipoles and their packing density differ between different materials. Theoretically, the electrostatic potential due to partially oriented dipolar molecules can be calculated by using the Helmholtz equation: 5) where N is the packing density of the PFN molecules and p z their net dipolar moment normal to the substrate. The expected V PFN value for spin-coated films is at least 0.4 V according to studies on this polymeric interlayer. ...
This work studies a novel electron-selective contact for n-type silicon solar cells based on the modification of the cathode with a polymeric interlayer. Specifically, a thin layer of the conjugated polyelectrolyte PFN is intercalated before the Al contact. During solution-processing, the amine groups in the PFN polymer form intense dipoles by the aggregation of acetate radicals. These dipoles are partially oriented in spin-coated layers because of the different interaction with the substrate and the solvent. As a result, an internal electrostatic potential appears that causes an apparent reduction of the cathode work-function. In this way, electron-selectivity at the rear contact is enhanced by inducing a favourable band-bending near the surface. A significant improvement is obtained in conventional electron-selective contacts with n-doped a-Si:H layers. Furthermore, good-performing solar cells can be also obtained even without intentionally doped layers. Compared to other solutions like the evaporation of metal-fluoride salts, the PFN cathode-modification can be readily done in ambient. Besides, the soft solution-processing kindly preserves the properties of the underlying passivating layers.
... 1) PERC cells, with a current average power conversion efficiency (PCE) of 23.5% in mass production, are steadily progressing toward a practical efficiency ceiling of 24% [2]. 2) The trap-assisted (SRH) recombination lifetime of minority carriers in n-type substrates is less impacted by donor-like impurities, like iron (Fe), Tungsten (W), or the Boron-Oxygen complex, which are the dominant contaminants in silicon, compared to p-type substrates [3]. ...
Silver paste accounts for a substantial portion of the nonsilicon cost of tunnel oxide polysilicon contact solar cells. Silver consumption is as well a major concern for material sustainability of global PV manufacturing. It is necessary to propose innovative grid line designs to reduce the amount of silver paste. Partially interrupting the metal fingers (also known as “Finger Break”) between the bus bars in areas where the fingers carry a relatively small current is a good method for reducing silver consumption in industrial production. Under the assumption of uniform generated current density (ISO-current), this article discusses the impact of the width of finger interruption on the cell efficiency under two different scenarios. Through simulations, experiments, and outdoor testing, it was found that one of the considered novel grid designs can save silver paste to a great extent, but suffers a rapid decline in power conversion efficiency, and also poses potential stability risks, while the other grid design not only saves silver, it does not severely affect the efficiency, and can be used outdoors while maintaining stability. Based on the different light intensities, the finger interruption width can be as high as 5 mm, resulting in a saving of 9.32 mg (∼1.61 mg/W) of silver paste on the front surface of 158.75 cm×158.75 cm solar cell, for a reduction of 12.97% of the original front silver consumption.
... Wafer fabrication and etching: At CEA and in the SHJ cell industry, n-type wafers are usually chosen as their minority carriers bulk recombination activity at a given resistivity is significantly lower than that of p-type wafer [18] because the absence of high temperature steps, does not allow to cure boron-oxygen defects [19] [20]. However, Lim et al. have recently proven that boron-oxygen defect in p-type wafer can be cured without high T annealing [21] which is a promising first step towards the production of SHJ with p-type wafers know to be less expensive than n-type wafers. ...
In a context where commercially available monocrystalline silicon substrates allow charge carrier lifetimes of several milliseconds, the quality of the passivation of the silicon surfaces in silicon heterojunction solar cells comes to the forefront for improving the performance of these industrially produced cells. First, the passivation defects of the front and back surfaces of the device have been studied by combining simulations (Silvaco) and experiments. This allowed the understanding and quantification of the physical phenomena involved in the losses associated with passivation defects. The influence of the variation of the defect properties was then revealed under standard test conditions as well as under different illuminations and operating temperatures in order to reproduce the real operating conditions of the cells. Furthermore, it is known that the passivation quality of the cell edges is lower compared to the front and back surfaces. This is even more important with the development of photovoltaic modules with cut cells. Since these cells are smaller, the current flowing through them, and therefore the associated resistive losses, are reduced. In these cells, edges are generated by the cuts and the ratio perimeter/surface being high, the edge effects are exacerbated. In a second step, a novel simulation code was developed to study these effects on the performance of the cells. This allowed to understand and quantify the phenomena involved in the losses related to the edges of whole and cut cells. In particular, the performance of a cell was related to the recombination rate on its edges whatever the size and geometry of the cell. For each of these studies, practical recommendations for the reduction of these losses on industrially produced cells have been formulated.
Translated with www.DeepL.com/Translator (free version)
... Amongst the various c-Si solar cell technology pathways displayed in Table 1, low-temperature (< 200 • C) deposition and processing of dopant-free CSCs are far less complex compared to that for doped passivation layers. In fact, n-Si wafers are highly favored for industrial manufacturing processes because of the greater tolerance to common metal contaminants and induced/introduced defects [49,50]. Hence, compared to p-Si substrates, n-type float-zone (Fz)-Si or even Czochralski-grown (Cz)-Si posseses superior minority carrier diffusion lengths so that bulk recombination losses are smaller. ...
Carrier-selective crystalline silicon heterojunction (SHJ) solar cells have already reached superior lab-scale efficiencies. Besides judicious wafer thickness design, the optimal choice of passivation schemes and carrier-selective materials is essential for industry adoption. Appropriate reduction of process complexity and performance benefits through minimal recombination losses are key. Thus, along with well-designed back contacts, the development of low-temperature processable transparent passivating stacks that act as carrier-selective contacts (CSCs) is highlighted for their potential in circumventing the limited open-circuit photovoltage and contact-related losses in mainstream solar cells. In this review, effective passivation schemes deploying materials ranging from undoped metal oxides (MOs) to doped silicon are evaluated, with a focus on their significance for industrially viable passivating contact development. Passivation stack architectures with SiOx/heavily doped polycrystalline silicon (n+-/p+-poly-Si) realize the most attractive polysilicon-on-oxide (POLO) junctions and related schemes, e.g., combined with tunnel oxide passivated contact (TOPCon) and interdigitated back contact (IBC) solar cells. It is envisioned that the industrial trend is to eventually shift from the p-Si passivating emitter rear contact (PERC) and passivated emitter and rear polysilicon (PERPoly), towards TOPCon architectures, due to high manufacturing yields and compatibility with large-area metal screen printing and alternative bifacial designs.
... Despite the advantages, p-type silicon is the one that dominates the PV market, though for homojunction solar cells. The reason behind this is low degradation to irradiation compared to n-type silicon wafers; thus, p-type silicon solar cells are preferred for space application over n-type silicon [4]. Electrons, the minority carrier in a p-type silicon wafer, possess longer diffusion length and higher mobility than holes in n-type crystalline silicon. ...
Performance of highly efficient one-sided short PIN diode heterojunction solar cell model is studied using AFORS-HET simulation software. Here, p-type crystalline silicon is used as an absorber layer, n-type amorphous silicon as emitter layer, and a thin intrinsic amorphous silicon (i-a-Si:H) layer is sandwiched in between, which acts as a passivation layer. The diode model concept has been introduced mainly to lower the reverse saturation current by appropriate selection of the cell parameters. Short diode reduces recombination phenomenon in the short quasi-neutral region (n-a-Si:H in this article), and one-sided junction allows the maximum amount of light absorption in the absorber layer (p-c-Si in this article). The one-sided junction nature has been confirmed by capacitance–voltage (C–V) analysis. Further thickness variation of layers and role of the carrier concentration of n-a-Si layer on various parameters, such as short-circuit current density (J_sc), open-circuit voltage (V_oc), and fill factor (FF) have also been studied. The structure has exhibited an open circuit voltage of 0.714 V, fill factor 0.80, and maximum efficiency of 24.2%. Further, experimental validation is also established by comparing the theoretical results with the fabricated NIP structure.
... A solar cell is a semiconductor diode with p-type and n-type doped regions that generate electricity from the absorption of light. [41], [42] Depending on the type of semiconductor used, the properties of light reflection and light absorption vary, and they depend on many factors summarized in [41, eq. (3) and eq. ...
One of the common failures in photovoltaic modules is the degradation of the ethylene-vinyl acetate (EVA) encapsulant due to prolonged ultraviolet exposure and other environmental stress factors, such as temperature and humidity. Experimental studies have shown that significant reduction in the optical transmission due to EVA degradation leads to loss in the available power by more than 50%. In this article, a novel approach to predict the early degradation of EVA encapsulant is proposed by correlating EVA degradation with short-circuit current (I
$_\rm{SC}$
). An electrical circuit simulator, simulation program with integrated circuit emphasis (SPICE), is used to evaluate the short-circuit current obtained under varying optical transmission caused by EVA discoloration. The simulation follows three steps: simulation of the transmitted solar spectrum; simulation of the spectral short-circuit current density; and simulation of the current-voltage (I–V) curve to obtain short-circuit current (I
$_\rm{SC}$
), maximum power output (P
$_\rm{max}$
), open-circuit voltage (V
$_\rm{OC}$
) and fill factor. Results show that the reduction in short-circuit current due to EVA degradation differs from the reductions expected due to a spectrally-uniform reduction of intensity of the solar irradiance. Both types of variation are linear, however, the slope due to EVA degradation is larger than the slope obtained for normal intensity variations in the solar irradiance. This model, when applied in conjunction with solar irradiance measurements, can predict early onset of EVA encapsulant failure, thereby enabling preventative measures to be taken.
... This is largely because n-type Si is exempt from boron-oxygen defects and the associated light-induced degradation, is more tolerant of most metal impurities, and thus usually presents a higher minority carrier lifetime than p-type. 133,134 In cell structures that rely on ll Joule 5, 1-18, April 21, 2021 lateral transport of majority carriers, n-type silicon also offers approximately three times better conductivity than p-type c-Si at equal doping. Not surprisingly, the first application of poly-Si junctions has been to n-type Si cells ( Figure 6B), where it has proved its worth by lifting the efficiency of champion cells made in R&D production lines up to 24.79%. ...
Despite the maturity of crystalline silicon photovoltaics (c-Si PV), the last 6 years have seen a string of efficiency improvements, most of which are centered around reducing the losses related to the directly metallized, heavily doped regions found in conventional c-Si solar cells. Among these advancements, polysilicon (poly-Si) passivated junctions, formed by embedding a thin silicon oxide (SiO2) layer between the c-Si wafer and a highly doped poly-Si layer, are emerging as one of the most promising alternatives, and efficiencies above 26% have already been demonstrated. The excellent performance of this junction architecture has been found to be remarkably independent of the deposition and/or doping technique used—even extending to techniques already prevalent in industry. This greatly reduces the capital and retraining expenditure needed to integrate the new technology into mainstream production lines, allowing it to be an evolutionary, rather than disruptive advancement. This has led to the rapid demonstration of large-area cells featuring poly-Si contacts by multiple PV manufacturing companies, with efficiencies above 24.5%. Although a bright future for poly-Si junctions is anticipated, as supported by the predictions of the International Technology Roadmap of Photovoltaics, several issues remain to be resolved, including those associated with the cost of and damage to the poly-Si layers due to the cell’s metallization process. This paper provides a perspective of the remaining challenges and potential of poly-Si junctions to transform the PV industry.
... Crystalline silicon (c-Si) solar cells are the most prevalent in the current photovoltaic market owing to their high conversion efficiency, simple fabrication process with industrial compatibility, and long-term stability [1,2]. The cost reduction schemes of c-Si solar cells often include the optimization of wafer thickness and quality, but a simplified structure with a low-temperature fabrication approach is highly desirable. ...
Silicon heterojunction (SHJ) solar cells are gaining prevalence owing to their high conversion efficiency and simple, low-cost, and industrially compatible fabrication process. However, the high parasitic absorption exhibited by the doped amorphous silicon a:Si(p/n) layer substantially limits their performance. This can be alleviated by substituting traditional amorphous doped p/n layers with alternative materials such as transition metal oxides. MoOx is one of the most promising candidates for replacing the a:Si(p) layer, and it has yielded impressive results in recent studies; however, the thermal instability of MoOx-based devices is a major limitation. In this work, the MoOx work function, thermal stability of the ITO/MoOx/a-Si(i) interface, and the effect of MoOx thickness variation on the performance of a MoOx SHJ solar cell were thoroughly studied. The MoOx composition and work function were analyzed using X-ray photoelectron spectroscopy, electron energy loss spectroscopy, and ultraviolet photoelectron spectroscopy. The O2-treated MoOx film showed minimal oxygen deficiencies accompanied by a higher work function of 6.22 eV compared with those of H2-, N2-, and Ar-treated MoOx films. The ITO/MoOx/a-Si(i) stacks have a clear contact interface up to 150 °C and in-out diffusion between the layers is initiated upon further increase of the annealing temperature. Thus, 150 °C was proposed as the optimal temperature and used for Ag paste sintering without degrading the device performance. Additionally, a thin sub-stoichiometric SiOx layer was formed on a-Si(i) during MoOx deposition, irrespective of the annealing environment, temperature, and thickness of MoOx. The fabricated SHJ solar cell with an optimized thickness of MoOx (7 nm) exhibited an efficiency of 19.86% with a Voc of 716 mV, Jsc of 37.50 mA/cm², and FF of 74.01%. The achieved Jsc is 1.40 mA/cm² higher and the external quantum efficiency in the range of 300–600 nm is 13%–15% higher than those for the reference SHJ solar cell. Therefore, our results show that MoOx is an efficient and convenient alternative emitter material for Si solar cells, and the proposed optimized parameters for MoOx processing pave the way for more efficient SHJ solar cells.
... A passivated emitter and rear cell (PERC) based on P-type Czochralski-grown silicon (Cz-Si) substrate has been introduced into production [1]. However, with absence light induced degradation (LID), higher lifetimes of minority charge carriers, and lower sensitivity towards metal impurities, N-type solar cells could offer greater efficiency possibilities and advantages over P-type solar cells [2][3][4][5][6][7]. ...
A kind of low recombination firing-through screen-printing aluminum (Al) paste is proposed in this work to be used for a boron-diffused N-type solar cell front side metallization. A front side fire-through contact (FTC) approach has been carried out for the formation of local contacts for a front surface passivated solar cell. With a low contact resistivity (ρc) of 1.0 mΩ·cm2, good ohmic contact between the boron-doped front surface of the silicon sample and the Al paste was realized. To obtain a good energy conversion efficiency, a balance can be achieved between the open circuit voltage (Voc) and contact resistivity (ρc) of the cell by combining suitable Al powders and appropriate additives. The detailed micro-contact difference in Si/metallization between the firing-through Al paste and silver-aluminum (Ag-Al) paste was analyzed. The dark saturation current density beneath the metal contact (J0, metal) of the Si/metallization region using our firing-through Al paste was discussed, which was proven to be 61% lower than using Ag-Al paste. The pseudo energy conversion efficiency of the cell using Al paste measured by Suns-VOC was also higher than using Ag-Al paste. The role of Al paste in low surface metal recombination is discussed. The utilization of this new kind of Al paste was much cheaper and more convenient, compared to the traditional process using Ag or Ag-Al paste.
... Sentaurus technology computer-aided design (TCAD) software developed by Synopsys, Inc. 27) is used for numerical analysis. Our previous work evaluated TOPCon solar cell structural parameters for p-type Si base of resistivity 1 Ω cm, 28) and in this study, we focused on both ptype (current popular material for c-Si in the market 29) ) and ntype Si (higher bulk lifetime 30) ). The effects of bulk doping property including bulk doping type, bulk resistivity, and bulk quality and tunneling oxide thickness were evaluated, showing that n-type Si bulk is suitable for TOPCon solar cells, in addition to its higher bulk lifetime. ...
... P-type monocrystalline silicon solar cells receive benefits of hydrogen passivation of metastable defects such as the carrier- 38 This has been a major driver for research into the use of alternative materials such as n-type Cz silicon. 39 Even in the early development of the PERC and passivated emitter and rear, locally diffused (PERL) cell technologies in the 1980s and 1990s at UNSW, the incredible importance of hydrogen passivation in silicon solar cells had been noted. The successful development of F I G U R E 1 Process flow for PERC solar cells these technologies capitalised on work from investigations of hydrogen passivation in the semiconductor industry during the 1960s. ...
The understanding and development of advanced hydrogenation processes for silicon solar cells are presented. Hydrogen passivation is incorporated into virtually all silicon solar cells, yet the properties of hydrogen in silicon are still poorly understood. This is largely due to the complex behaviour of hydrogen in silicon and its ability to exist in many different forms in the lattice. For commercial solar cells, hydrogen is introduced into the device through the deposition of hydrogen‐containing dielectric layers and the subsequent metallisation firing process. This process can readily passivate structural defects such as grain boundaries but is ineffective at passivating numerous defects in silicon solar cells such as the boron‐oxygen complex, responsible for light‐induced degradation in p‐type Czochralski silicon. This difficulty is due to the need to first form the boron‐oxygen defect and also due to atomic hydrogen naturally occupying low‐mobility and low‐reactivity charge states. However, these challenges can be overcome using advanced hydrogenation processes incorporating excess carrier generation from illumination or current injection that increase the concentration of the highly mobile and reactive neutral charge state. As a result, after fast firing, additional low‐temperature advanced hydrogenation processes incorporating illumination can be implemented to enable the passivation of difficult defects like the boron‐oxygen complex. With the implementation of such processes for industrial silicon solar cells, efficiency improvements of 1.1% absolute can be obtained.
... Effective surface passivation of crystalline silicon has been performed using amorphous silicon [4], amorphous silicon carbide [5] and aluminum oxide [6]. Surface passivation is also achieved by amorphous hydrogenated silicon nitride (abbreviated SiN) [7][8][9][10][11]. Another passivating material is thermally grown silicon dioxide (SiO 2 ). ...
The SiN/SiO 2 stack is widely used to passivate the surface of n-type monocrystalline silicon solar cells. In this work, we have undertaken a study to compare the stack layer obtained with SiO 2 grown by both rapid thermal and chemical ways to passivate n-type monocrystalline silicon surface. By varying the plateau time and the plateau temperature of the rapid thermal oxidation, we determined the parameters to grow 10 nm thick oxide. Two-step nitric acid oxidation was used to grow 2 nm thick silicon oxide. Silicon nitride films with three refractive indices were used to produce the SiN/SiO 2 stack. Regarding this parameter, the minority carrier lifetime measured by means of QSSPC revealed that the refractive index of 1.9 ensured the best passivation quality of silicon wafer surface. We also found that stacks with nitric acid oxidation showed definitely the best passivation quality. In addition to produce the most efficient passivation, this technique has the lowest thermal budget.
... The most common dopants for PV cell are Boron and Phosphorous [46]. When doped with Phosphorous, the silicon solar cell becomes a negative-type (n-type) while when it is doped with Boron, the cell become a positive-type (p-type) [47,48]. ...
This thesis work presents the design and simulation of a 100kVA hybrid solar power system to be developed for Gollis University’s administrative block. Prior to the system design, a preliminary field work on the site was performed to essentially measure the power/energy consumption of Gollis University’s administrative block. The results from the site survey were then used to select the appropriate equipment and instrument required for the design. This was achieved by calculating the energy consumption and then sizing the solar panel, battery, inverter and charge controller. The inverter system was modeled and simulated using the MATLAB /Simulink software package. The simulation was used to study the reliability of the size of inverter chosen for the design, since the failure of most photovoltaic systems is ascribed to inverter failures. The implications of the results are then discussed before presenting the recommendations for future works.
... Indeed, most effort in improving cells for thin wafers is related to light trapping and increased absorption due to the low IR absorption of silicon. Further, since the LID has been shown to account for a 2-4% power loss in p-type wafers, which is absent in n-type wafers the prospects for higher efficiency are far greater for n-type wafers [101]. Notably, most high efficiency manufacturers of technologically advanced solar cell architectures such as interdigitated back contact (IBC), heterojunction intrinsic thin films (HIT) utilize n-type wafers. ...
... N-type Si wafers have advantages over p-type Si wafers for the fabrication of high-efficiency Si solar cells. N-type Si wafers have a high tolerance to metal impurities, resulting in a longer bulk minority carrier lifetime [3], [4]. Moreover, they suffer less from light-induced degradation, which is caused by complexes of the p-type dopants B and O acting as recombination centers [5]- [9]. ...
We investigate the effect of current injection during contact formation of an Ag-based screen-printed electrode to boron-doped emitters, which differ by their sheet resistances. The average contact resistivities between the metal electrode and silicon of all the boron-doped emitter samples are ~3 mΩ cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
, regardless of the sheet resistance (75-145 Ω/sq), and the lowest values are below 1 mΩ cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
using the injection of a current density of 5 A/cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
during the metallization process. Additionally, the injection of current to a phosphorus-doped emitter in the opposite direction suppressed the formation of the Ag precipitates and crystallites and increased the contact resistivity of over 300 mΩ cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
, which is comparable to that obtained when Ag paste is applied to a boron-doped emitter with no current injection. This finding indicates that electrons are essential for the reduction of Ag ions during high-temperature metallization process using the screenprinting technique and that the injection of current can control the contact formation and enhance the efficiency of solar cells. Finally, we suggest a suitable process for reducing the contact resistivity in manufacturing n-type Si solar cells.
... N-type Si solar cells are of great interest for widespread photovoltaic applications. They exhibit many advantages compared to p-type Si substrates, such as higher minority carrier diffusion lengths, higher tolerance to metallic impurities and immunity from boron-oxygen related light-induced degradation [1][2][3][4][5][6][7] . In general, these superior properties allow n-type Si solar cells to attain higher photovoltaic conversion efficiencies. ...
In boron-doped p⁺-n crystalline silicon (Si) solar cells, p-type boron doping control and surface passivation play a vital role in the realization of high-efficiency and low cost pursuit. In this study, boron-doped p⁺-emitters are formed by boron diffusion in an open-tube furnace using borontribromide (BBr3) as precursor. The formed emitters are characterized in detail in terms of shape of the doping profile, surface doping concentration, junction depth, sheet resistance and removal of the boron-rich layer (BRL). In the aspect of BRL removal, three different methods were adopted to investigate their influence on device performance. The results demonstrate that our proposed chemical etch treatment (CET) with the proper etching time could be an effective way to remove the BRL. After removal of the BRL, Al2O3/SiNx stacks are deposited by atomic layer deposition (ALD) and plasma-enhanced chemical vapor deposition (PECVD) to passivate the cell surface. It was found that a reasonably-high implied V oc of 680 mV has been achieved for the fabricated n-type Si solar cells.
Energetics at the carrier selective layer/absorber interface plays a subtle yet crucial role in influencing the performance of any photovoltaic technology. This review focuses on the interface energetics and resulting aspects of device physics associated with n‐type molybdenum oxide, which is adopted as a hole selective layer over diverse class of photovoltaic (PV) technologies ranging from silicon heterojunction (SHJ), organic and perovskite solar cells (PSCs). As existing literatures provide a very inconsistent picture of the electronic structure of transition metal oxides, specifically in the case of molybdenum oxide (MoO x ), this review makes a first‐of‐its‐kind attempt to connect the scattered reports and thus highlight the ways in which device performance varies for diverse class of photon absorbers used in conjunction with n‐type MoO x hole selective layer. In spite of the interfacial energy requirements and in turn fundamental working mechanism of varied photovoltaic technologies being the same, an attempt to compare and collate the contributions of MoO x as a charge selective layer for each of these technologies is yet to be done. In this review, we focus primarily on the role played by MoO x /absorber interface engineering on the performance of varied photovoltaic technologies, which remains a fundamental and yet unaddressed theme till date.
This article is protected by copyright. All rights reserved.
Células solares de silício base n com emissor posterior de alumínio ou boro têm potencial de alcançar alta eficiência, apresentam baixa degradação à radiação solar e podem ser fabricadas com a tecnologia atual da indústria. Nesta estrutura de célula solar, o campo retrodifusor frontal (FSF – front surface field) é produzido pela difusão de fósforo, que em dispositivos base p, forma o emissor. O objetivo deste artigo é avaliar a influência da concentração de POCl3 nas características do campo retrodifusor frontal de fósforo em células solares base n, por meio da análise da resistência de folha e do perfil de fósforo. A difusão do fósforo foi realizada na temperatura de 845 °C e variou-se a concentração de POCl3 (CPOCl3) na câmara de processamento de 0,026 % a 0,064 %. As lâminas de Si foram colocadas com a face posterior juntas para reduzir a difusão de fósforo na face posterior onde será formado o emissor. Constatou-se que a resistência de folha no campo retrodifusor frontal de fósforo variou de (48 ± 3) Ω/□ a (72 ± 3) Ω/□, com a redução da concentração de POCl3 de 0,064 % e 0,026 %. A concentração em superfície (CS) de fósforo variou de 7,7x1020 a 1,0x1020 átomos/cm3 e observou-se a tendência de aumento de CS com o aumento da concentração de POCl3. Em relação ao perfil de fósforo verificou-se que, em geral, a concentração de fósforo em função da profundidade é maior para os perfis com maior CPOCl3. A concentração de POCl3 afeta principalmente a concentração de fósforo próximo à superfície, e este parâmetro aumenta até a profundidade de aproximadamente 0,025 µm, formando a “camada morta”. Na face posterior, constatou-se que a profundidade da região altamente dopada e a concentração em superfície são menores que os valores obtidos no campo retrodifusor frontal e não se observou a “camada morta”.
A homo-hetero junction solar cell made up of crystalline and amorphous silicon is investigated using AFORS-HET simulator. The ability of a homo-hetero junction cell to improve photoelectric behavior is well established. The study is carried out on a p-type crystalline silicon substrate which forms a shallow junction with n-type crystalline silicon. The addition of highly doped amorphous silicon at the top and bottom surfaces creates extra field, thus reducing the interface recombination and series resistances. The study includes the thickness variation of n-type amorphous silicon, n-type crystalline silicon, and p-type substrate, to find out the optimum result. Also, the effect on cell performances due to the variation of n-type amorphous silicon doping is performed. In comparison with a heterojunction with thin intrinsic layer (HIT) cell, homo-hetero junction is different in a way that it includes the additional crystalline silicon layer which actually enhances the field-effect passivation, thereby further reducing the series resistance and enhancing the fill factor to a greater extent. The excess amount of field also makes it less sensitive toward the interfacial defect states, and to prove it, we increase the defect states up to 102 times greater than that of a HIT cell. The calculated total recombination factor, Jo, is found to be as low as 3.45 fA/cm2. Finally, we achieved an efficiency of 28.33% with enhanced fill factor (FF) of 0.88.KeywordsAFORS-HETAmorphous siliconCrystalline siliconHIT
Silicon solar cells are well known as a major technology used in solar photovoltaics due to their high lifetime and highly efficient performance. The conventional process of silicon solar cell wells involves a highly toxic process that is harmful to the environment and human health because of the application of liquid-based phosphorous oxychloride (POCl3) and boron tribromide (BBr3) for emitter formation. This paper discusses the potential of water-based phosphoric acid (H3PO4) as emitter formation on n-type Si wafer. An alternative method for fabrication of Si solar cells by replacing POCl3 or BBr3 by phosphoric acid (H3PO4) is proposed. Phosphoric acid is less harmful toward the environment and health because of in-house synthesis. The emitter formation uses heavily doped/ highly concentrated phosphoric acid by dip coating at various temperatures and times. Temperature was varied within the high range of 875 °C to 975 °C and low range of 700 °C to 850 °C. Deeper junction depth was observed at high temperatures with sheet resistance value of 7 Ω/sq, whereas a heavily doped junction was determined at 700 °C for 5 min with sheet resistance (Rsheet) value of 81 Ω/sq. The variation in sheet resistance (Rsheet) ranging from 5 Ω/sq to 100 Ω/sq depended on time and temperature, where deeper junction depth occurred at high temperatures while shallow junction occurred at low temperatures during diffusion. Resistance measured was consistent from the front and back surface of n-Si wafer. Junction depth (xj) was determined through mathematical calculation depending on time, temperature, peak doping and base doping of n-Si wafers. The junction depth before and after PSG removal on n-Si wafers during low-temperature diffusion of H3PO4 ranged from 0.08801 μm to 4 μm. The value of junction depth, (xj) at high temperature was approximately 4 μm to 28.844 μm. These data were supported by the activation of phosphorous by 20% H3PO4 on n-Si wafer by using dark I–V, with series and shunt resistance of 7.763 Ω and 18.315 Ω, respectively. PC1D simulation was used to estimate the junction depth, which was found to be correlated with the measured sheet resistance from the experiment. The overall efficiency of the n-type emitter on n-Si solar cell was determined as 8.6% by using the average peak doping concentration. The FESEM and EDX profiles indicated the presence of 20% H3PO4 on n-Si wafer. Thus, fully coated 20% H3PO4 can be applied to a toxic-less approach for commercial fabrication of silicon solar cells.
Both polarity
poly
-Si/SiO
<sub>x</sub>
passivating contacts in a front/back device configuration may represent the next solar cell architecture after tunnel oxide passivating contacts (TOPCon) cells, but high parasitic absorption in the front
poly
-Si layer often limits its performance. This work explores a wet etching technique to remove the front
poly
-Si in the nonmetallized regions using self-aligned metal grids as an etch mask. We systematically examine various dielectric layers (SiN
<sub>x</sub>
, Al
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>
O
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub>
, and stacks thereof) to study the repassivation of the etched
n
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup>
surface, and find that an SiN
<sub>x</sub>
/Al
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>
O
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub>
passivation stack can effectively repassivate the etched surface, which we attribute to field-effect passivation from the positive fixed charge from the SiN
<sub>x</sub>
layer, and excellent chemical passivation property from Al
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>
O
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub>
in the form of atomic H. We demonstrate a front/back
poly
-Si/SiO
<sub>x</sub>
passivating contact device, with an open-circuit voltage (
V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">oc</sub>
) of 690 mV, short-circuit current density (
J
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sc</sub>
) of 39.8 mA/cm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
, fill factor of 78%, and power conversion efficiency of 21.4%. Furthermore, simulations using SunSolve and Quokka 3 show good agreement with both the optical and electrical properties of the experimental device. The power loss analysis reveals improvements in the optical loss from the back Ti adhesion and front SiN
<sub>x</sub>
layers would lead to a 23.5% device. Lastly, a techno-economic model compares the production cost of this improved cell with the current TOPCon cells. Our results highlight that further cost reductions in single-sided doped
poly
-Si layers are needed to compete with mainstream passivated emitter and rear cell, and TOPCon technologies.
Recently, the parameterisation of Auger recombination in c-Si has been revised by two separate studies, both of which reached very similar conclusions. A key change is that the ratio of the Auger coefficients for n- and p-type Si (Cn/Cp) has been found to be significantly lower than previously accepted. In this work, we explore the implications of these findings for c-Si solar cells. In particular, we seek to answer the question of whether any intrinsic advantage is expected for n-type vs p-type doping for c-Si solar cells in general, or for particular device architectures. We focus on simple analytical parameters and models in order to elucidate the relevant physical mechanisms, making use of more complex numerical modelling to complement these and validate our conclusions. A key conclusion is that the new models predict improved intrinsic performance potential for n-type devices, with p-type Si retaining an intrinsic advantage only when surface recombination and transport losses are very close to zero. n-type quickly becomes more efficient when non-zero surface recombination is introduced, and widens its advantage with increasing surface saturation current density J0. In particular, n-type back-junction devices show higher efficiency potential and less sensitivity to bulk resistivity variation than p-type equivalents, thanks to lower lateral transport losses for majority carriers in the Si bulk. Meanwhile, an analysis of the intrinsic performance potential of highly doped n- and p-type Si reveals no significant intrinsic advantage for either type as transport layers.
The first reports of both boron‐oxygen (BO) related light‐induced degradation (BO‐LID) and amorphous/crystalline silicon heterojunction (SHJ) solar cell fabrication date back to the early 1970s. However, the complete development of the ‘modern’ SHJ structure took place well before BO defect stabilisation processes were developed. Due to the susceptibility of p‐type Czochralski‐grown silicon to BO‐LID, such wafers were deemed unsuitable for SHJ solar cells. In addition to stability issues, lower charge carrier lifetimes due to contamination and challenges with surface passivation posed barriers to the adoption of p‐type wafers in SHJ applications. In this perspective, these three key challenges are discussed in detail. Kinetic modelling and experimental results reveal the severe impact of BO‐LID in p‐type SHJ solar cells and provide possible explanations as to why earlier attempts using p‐type wafers might have failed. The role of gettering and advanced hydrogenation in stabilising BO defects in SHJ solar cells is demonstrated experimentally. Finally, a summary of the effective surface recombination velocities reported in literature for hydrogenated intrinsic amorphous silicon passivation of p‐ and n‐type crystalline silicon wafers is presented. Based on these findings, the potential of p‐type wafers to enable a next‐generation of high‐efficiency solar cells featuring carrier‐selective contacts is discussed. This article is protected by copyright. All rights reserved.
Tandem solar cells (TSCs) represent a promising route to break the Shockley–Queisser limit of single‐junction solar cells. Organic–inorganic hybrid perovskite is an ideal candidate for fabricating TSCs with the superiority of excellent photoelectric properties, bandgap tunability, and low‐temperature solution processability. In this chapter, we summarize the recent progress of perovskite‐based TSCs, including perovskite/crystalline silicon (PVSK/c‐Si), perovskite/organic (PVSK/OPV), and perovskite/perovskite (PVSK/PVSK) solar cells. Then, we introduce the emergence of multi‐junction perovskite‐based tandems. Subsequently, the challenges and potential strategies toward high‐efficiency perovskite‐based TSCs are discussed. Lastly, the conclusions and perspectives on the promising perovskite‐based TSCs are provided.
The silicon heterojunction (SHJ) solar cell has long been considered as one of the most promising candidates for the next-generation PV market. Transition metal oxides (TMOs) shows good carrier selectivity when combined with c-Si solar cells. This has led to the rapid demonstration of the remarkable potential of TMOs (especially MoO x ) with high work function to replace the p-type a-Si:H emitting layer. MoO x can induce a strong inversion layer on the interface of n-type c-Si, which is beneficial to the extraction and conduction of holes. In this paper, the radio-frequency magnetron sputtering was used to deposit MoO x films. The optical, electrical and structural properties of MoO x films are measured and analyzed, with focus on the inherent compositions and work function. Then the MoO x films are applied into SHJ solar cells. When the MoO x works as a buffer layer between ITO/p-a-Si:H interface in the reference SHJ solar cell, a conversion efficiency of 19.1% can be obtained. When the MoO x is used as a hole transport layer (HTL), the device indicates a desirable conversion efficiency of 17.5%. To the best of our knowledge, this current efficiency is the highest one for the MoO x film as HTL by RF sputtering.
Inverted pyramid texture is used to improve the performance of single crystalline silicon (sc-Si) solar cell due to its excellent light-trapping properties. In this paper, inverted pyramid structures are fabricated on large area sc-Si wafers by metal assisted chemical etching method using CuSO4 instead of Cu(NO3)2. Three groups of solar cells were fabricated with textured structures prepared by CuSO4, Cu(NO3)2 and NaOH respectively. The surface microstructure, average reflectivity, surface recombination and auger recombination, internal quantum efficiency and external quantum efficiency of the sc-Si solar cells were studied in detail. The results show that CuSO4 can prepare uniform inverted pyramid structures on silicon substrate by reducing the generation of bubbles and promoting the deposition copper nano-particles on the surface of sc-Si wafer, and thus improve the properties of solar cells. More importantly, the chemical reaction mechanism for etching silicon wafer with CuSO4 and Cu(NO3)2 during the texturing process was compared and analyzed. Finally, the average efficiency of large-area sc-Si PERC solar cells fabricated by CuSO4 was 22.63%, which was 0.60% absolutely higher than that of Cu(NO3)2, and 1.56% higher than normal pyramid group.
An a-Si:H (p) window layer is used in silicon heterojunction (SHJ) solar cells; however, it is limited by short-circuit current density (JSC). In general, an emitter with a high doping concentration is appropriate for contact with a transparent conducting oxide (TCO); however, it is influenced by side effects such as a reduction of JSC through optical absorption. The conductivity of the emitter is lowered as its doping concentration is reduced, resulting in a decrease in VOC and FF. We investigated p-type emitters such as those made of a-Si:H, a-SiC:H, and µc-SiO:H through film analysis and AFORS-HET simulation to improve the conversion efficiency of the device. Prior to conducting a simulation, a fabricated SHJ solar cell was used to theoretically calculate the precise parameter values. The obtained efficiency was 22.03% when VOC =730 mV, JSC =39.63 mA/cm², and FF = 76.13%. Based on the fitted structure, we conducted experiments to test the emitter materials within a wide band gap and performed a simulation. In the case of µc-SiO:H (p), the achieved efficiency was 24.23% when VOC =736.6 mV, JSC =40.15 mA/cm², and FF = 81.93%.
We describe a newly developed multiscale computational method, combining quantum mechanics with classical electrodynamics for simulations of photovoltaic devices. In this quantum mechanics/electromagnetics (QM/EM) method, the regions of the system where charge excitation and migration processes take place are treated quantum mechanically, while the surroundings are described by Maxwell’s equations coupled with a semiclassical drift-diffusion model. The QM model and the EM model are solved, respectively, in different regions of the system in a self-consistent manner. Potential distributions and current densities at the interface between QM and EM regions are employed as the boundary conditions for the quantum mechanical and electromagnetic simulations, respectively. In this chapter, we first demonstrate the method by studying the plasmonic scattering and light trapping effects in silicon nanowire array solar cells. Our results show that there exists an optimal nanowire number density in terms of optical confinement. The method is then applied to study a tandem solar cell where the subcells are treated quantum mechanically. The QM/EM simulation results demonstrate that a significant enhancement of open-circuit voltage is achieved by using the tandem architecture.
Silicon heterojunction solar cells can achieve high conversion efficiency with a simple structure. In this study, we investigate the passivation characteristics of VOx thin films as a hole-selective contact layer using ALD (atomic layer deposition). Passivation characteristics improve with iVoc (implied open-circuit voltage) of 662 mV and minority carrier lifetime of 73.9 µs after post-deposition annealing (PDA) at 100 °C. The improved values are mainly attributed to a decrease in carbon during the VOx thin film process after PDA. However, once it is annealed at temperatures above 250 °C the properties are rapidly degraded. X-ray photoelectron spectroscopy is used to analyze the chemical states of the VOx thin film. As the annealing temperature increases, it shows more formation of SiOx at the interface increases. The ratio of V⁵⁺ to V⁴⁺, which is the oxidation states of vanadium oxide thin films, are 6:4 for both as-deposition and annealing at 100 °C, and 5:5 for annealing at 300 °C. The lower the carbon content of the ALD VOx film and the higher the V⁵⁺ ratio, the better the passivation characteristics.
Organic‐inorganic hybrid heterojunction solar cells have gained extensive attention due to simple device structure and low‐cost technological processes, when the previous researches were mainly use n‐type silicon as inorganic absorber. In this work, a hybrid heterojunction solar cells based on p‐type silicon were successfully prepared through a low‐temperature solution method. The preliminary power conversion efficiency over 4% was observed, via optimizing the organic thin film deposition and organic/inorganic interface which enhance the built‐in field and promote the separation of photo‐generated carriers in the heterojunction. Therefore, a new idea and possibility have been provided for developing low‐cost, high‐performance solar cells.
This article is protected by copyright. All rights reserved.
This PhD thesis aims at evaluating low temperature silicon epitaxy (< 200°C) by RF-PECVD as an alternative to diffusion for the formation of the emitter layer of crystalline silicon (c-Si) solar cells. The epitaxial growth of intrinsic, n-type and p-type films grown on (100) c-Si substrates has been confirmed by ellipsometry. Surprisingly, the as-grown doped epitaxial silicon (epi-Si) films present a larger out-of-plane lattice parameter than the substrate, and are fully strained. Boron and phosphorus concentration of 3 and 3.5x1019 at/cm3, respectively, were achieved as deduced from SIMS measurements. However, in the as-grown state B-H complexes are formed and annealing is required to activate boron atoms. After annealing, we reached a doping efficiency of approximately 60%. As expected, the activation of boron leads to a drop of the mobility to ~20 cm2/Vs. Moreover, we put forward a correlation between the evolution upon annealing of the structure, electrical properties and the hydrogen content. Finally, XRD measurements have shown that by providing energy, annealing at 350°C induces the relaxation of the epi-Si structure, thus the apparition of defects, confirmed by TEM and low temperature PL.
This doctoral work aimed to assess the potential of low-temperature (200-300°C) epitaxy by plasma-enhanced chemical vapor deposition (PECVD) using SiF4/H2/Ar gas mixtures for the emitter formation in nPERT solar cells. The first part of this PhD thesis concerned the identification and the optimization of the process conditions to perform lowly strained intrinsic epi-layers with a smooth epi/wafer interface. We also investigated the causes of epitaxy breakdown and found out that a twinning-induced mechanism was responsible. Subsequently we focused on the growth mechanisms by studying the initial stages of growth and a Volmer-Weber growth mode has been highlighted. Finally, the process conditions for intrinsic epitaxy were transferred from a researchPECVD reactor to a 6 inch semi-industrial one. Inhomogeneity and growth rate issues have been tackled by fluid dynamics simulations resulting in the design of a new shower head. Boron-doped epi-layers grown at 300°C with an as-deposited hole concentration of 4.1019 cm-3 and a doping efficiency up to 70 % have been achieved keeping a low mosaicity and a low variation of the lattice parameter. The growth rate in these conditions reached 1.1 Å/s, i.e 15 times higher than what obtained at the beginning of this PhD for boron-doped epi-layers. Finally, the passivation of epitaxial layers has been investigated and lifetimes up to 160 μs for a 200 nm thick intrinsic layer passivated with
Carbon nanotube electrode‐laminated perovskite solar cells in combination with n‐type tunnel oxide passivated contact silicon solar cells demonstrate a high power conversion efficiency of 24.42% when stacked in tandem. This is compared with conventional indium tin oxide/MoOx‐deposited perovskite solar cells which give an efficiency of 22.35% when stacked in the same four‐terminal tandem system. Despite higher transmittance of the carbon nanotube electrode than that of the indium tin oxide/MoOx in the infrared range, the carbon nanotube electrode‐laminated devices show lower transmittance in the same region due to the total internal reflection and scattering as evidenced by optical simulation. Yet, the exceptionally high power conversion efficiency of the carbon nanotube electrode‐laminated semi‐transparent devices far exceeding that of the indium tin oxide/MoOx‐deposited semi‐transparent top cell outweighs the effect of the optical transparency. Four types of silicon solar cells are compared as the bottom subcells and the n‐type tunnel oxide passivated contact silicon solar cells is the best choice owing mainly to their high absorption in the long‐wavelength region. The obtained 24.42% efficiency is one of the high PCEs among the reported four‐terminal perovskite‐silicon tandem solar cells and this work is the first demonstration of the carbon nanotube electrode application in tandem solar cells.
This article is protected by copyright. All rights reserved.
High-purity silicon is a readily available material of utility in realizing a variety of long-wavelength optical and guided wave components. The transmittance of uncompensated for silicon is measured in the far- and mid-infrared regimes at room and cryogenic temperatures. The experimental and analysis techniques used to extract the refractive index from 100-1000cm-1 (100-10 µm) are presented, and the results are compared to the literature. An average refractive index below 300cm-1, n^(300K)=3.417+i8.9×10-5, which transitions in cooling to n^(10K)=3.389+i4.9×10-6, is observed.
Silicon heterojunction (SHJ) solar cells formed using n-type Cz silicon wafers are attracting increasing industrial interest. Cheaper p-type Cz silicon wafers can also be used to form SHJ cells; however, they achieve lower efficiencies. In this work, a Monte Carlo simulation approach is used to provide a comprehensive commercial comparison between n-type and p-type wafers, considering a wide range of uncertainty in the cost of production, the cost of the wafers, and cell performance. The most critical factors influencing the commercial comparison between wafer types were identified as the difference in cell efficiency, the difference in cost between n-type and p-type wafers, and the SHJ processing costs. The analysis provides a target for p-type SHJ solar cells of being within 0.4% absolute of that obtained with n-type wafers. This work motivates and sets research targets for the development of SHJ solar cells fabricated on p-type wafers.
In silicon heterojunction solar cells made with high-lifetime wafers, resistive losses in the contacts dominate the total electrical power loss. Moreover, it is widely believed that the hole contact stack-a-Si:H(i)/a-Si:H(p)/ITO/Ag-is responsible for more of this power loss than the electron contact stack. In this article, we vary the a-Si:H(i) layer thickness, the a-Si:H(p) layer thickness and doping, and the indium tin oxide (ITO) doping, and determine the effect of each variation on the contact resistivity of the hole contact stack. In addition, we make complete solar cells with the same variations and correlate their series resistivity to the hole contact resistivity. We find that the contact resistivity is most sensitive to the thickness of the a-Si:H(i) layer and the oxygen partial pressure during ITO sputtering. Increasing the former from 4 to 16 nm results in a fourfold increase in contact resistivity, whereas increasing the latter from 0.14 to 0.85 mTorr raises the contact resistivity almost 30-fold. Optimized conditions produce a contact resistivity of 0.10 Ωcm
<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>
, while maintaining an implied open-circuit voltage of 720 mV measured on cell precursors, which is the lowest contact resistivity value reported in the literature for an a-Si:H hole contact.
Molybdenum trioxide (MoOX, X < 3), with a large work function, can induce upward band bending in crystalline silicon (c‐Si) when constructing a heterojunction, which makes it an attractive candidate for hole‐selective contact in c‐Si solar cells. In this work, the passivation property and hole selectivity of MoOX thin films are investigated on p‐type c‐Si wafers using MoOX/aluminum (Al) as rear contacts. To elevate the performance from the aspect of light management, silver (Ag) and copper (Cu) are further used as back electrodes instead of Al. Solar cells with Ag electrodes deliver the best performance with a power conversion efficiency of 18.74%, followed by Cu (17.61%) and Al (16.36%) electrodes, attributing to the better reflectivity of Ag and Cu. It is also noted that solar cells with MoOX/Ag and MoOX/Cu contacts show significant degradation under room temperature storage. The interfacial evolutions are then carefully studied as a function of elevated temperature that would accelerate the thermodynamic process. The degradation mechanism involves redox reaction and metal diffusion at the MoOX/metal interfaces. This work points out the importance of selecting the adjacent layers of MoOX and regulating the interfaces to stabilize the MoOX‐based c‐Si solar cells. This article is protected by copyright. All rights reserved.
Perovskite/silicon hybrid tandem solar cells are very close to commercialization owing to their low cost and relatively high efficiency compared to tandem cells based on III-V compound semiconductors. However, most hybrid tandem cell research is based on n-type heterojunction Si cells, which occupy only a small fraction of the total solar market. Here, we propose a new method for optimizing the design of low-cost and high-efficiency monolithic tandem cells based on p-type homojunction Si cells by realizing lossless current matching by simultaneously controlling the band gap energy and thickness of the perovskite film. In addition, systematic studies have been conducted to determine the optimal hole transport layer applicable to the tandem cell from the viewpoint of band alignment and process compatibility, in order to reduce the open-circuit voltage loss. Optimized tandem cells, which were fabricated with a 310 nm thick perovskite layer of (FAPbI 3 ) 0.8 (MAPbBr 3 ) 0.2 and a hole transport layer of poly(triaryl amine), had a significantly increased efficiency of 21.19% compared to semi-transparent stand-alone perovskite (13.4%) and Si cells (12.8%). Our tandem cell represented the highest efficiency increment among all monolithic perovskite/Si tandem cells as well as the highest efficiency among monolithic perovskite/Si tandem cells based on p-type homojunction Si cells with Al back-surface fields. The design rules suggested in this study could also be applicable to different types of perovskite/Si tandem cells.
Recently bifacial n-type passivated emitter and rear totally diffused (PERT) monocrystalline silicon solar cells have become one hot spot in photovoltaic (PV) industries, due to good bifaciality, high and stabilised conversion efficiency. Unlike passivated emitter and rear contacts (PERC) structure, n-PERT solar cells require the use of a thin and uniform back surface field (BSF) layer, usually achieved by phosphorus doping. In this study, we optimised the phosphorus diffusion process in terms of surface concentration, junction depth and carrier lifetime. The effects of different phosphorus BSF were investigated by fabricating n-type front and back contact (nFAB) PERT solar cells using industrial feasible approaches and M2 size Czochralski (Cz) monocrystalline Si wafers (6 in., 244.32 cm²). Good phosphorus profiles were developed, which gives low parasitic absorption, low contact resistance and low J0 value when passivated with silicon nitride (SiNx) layer. The optimised champion cell shows a high Voc of 666.5 mV, Jsc of 40.2 mA/cm², fill factor of 79.9%, efficiency of 21.43% from front side-illumination, together with a good bifaciality factor of 93.0%.
This paper describes an augmented reality (AR) aided smart sensing technique, for in-line condition monitoring of IGBT wafers. A series of signal processing algorithms are applied for enabling sensor intelligence. Based on electromagnetic infrared-visible-fusion (IVF), a supplementary palpable 3-D thermography layer is integrated with an IGBT wafer in real world environment. Before the IVF, independent component analysis (ICA) is implemented to identify defects in the wafer. The proposed AR aided smart sensing technique enhances user's perception and interaction between the industrial systems and the surrounding world. In contrast to conventional sensor techniques, it provides a non-destructive testing and evaluation (NDT&E) based high-throughput in-line condition monitoring method. The advantages of non-contact and time efficient of this smart sensing technique potentially bring huge benefit to yield management and production efficiency. AR aided smart sensing can improve the productivity, quality and reliability of power electronic materials and devices, as well as in other industrial applications.
This paper reports recent progress by SunPower Corporation to commercialize silicon solar cells with efficiency greater than 20%. Large-area (149cm 2) cells with efficiency as high as 21.5% (confirmed by NREL) have been made on a 1 MW/yr pilot line, and a production line with 25 MW/yr capacity has been constructed. Using a back-contact cell design and novel manufacturing techniques, cells with efficiency over 21% were produced with techniques suitable for high-volume manufacturing using Photovoltaic-grade Float Zone (PVFZ) silicon. Modules have been built. All test sequences for IEEE 1262 qualification have been passed and a 5kW demonstration array has been installed. Advantages of the cell design include a grid-less front surface and n-type starting material that does not suffer the initial light-induced degradation of commonly-used p-type wafers. Additional technical information about cell design and experiment results is also provided.
Interstitial iron in crystalline silicon has a much larger capture cross section for electrons than holes. According to the Shockley–Read–Hall model, the low-injection carrier lifetime in p-type silicon should therefore be much lower that in n-type silicon, while in high injection they should be equal. In this work we confirm this modeling using purposely iron-contaminated samples. A survey of other transition metal impurities in silicon reveals that those which tend to occupy interstitial sites at room temperature also have significantly larger capture cross sections for electrons. Since these are also the most probable metal point defects to occur during high temperature processing, using n-type wafers for devices such as solar cells may offer greater immunity to the effects of metal contaminants.
We have investigated the effect of the light-induced deep-level recombination centre specific to boron-doped, oxygen-contaminated Czochralski (Cz) silicon on the current–voltage characteristic of Cz silicon solar cells by means of numerical simulation and experiment. The device simulation predicts the occurrence of a shoulder in the current–voltage curve after activating the characteristic recombination centre. The physical reason for the non-ideal diode behaviour, characterised by a local ideality factor greater unity, is the strongly injection-level-dependent bulk lifetime produced by the deep-level centre. The increased ideality factor causes a degradation in fill factor with the magnitude of degradation depending on the doping concentration of the Cz silicon base. In order to verify the theoretical predictions experimentally, we have performed measurements on high-efficiency Cz silicon solar cells. Current–voltage curves recorded before and after light degradation clearly show the theoretically predicted change in shape and the reduction in fill factor. An excellent quantitative agreement between calculation and experiment is obtained for the subtracted current–voltage curves measured after and before illumination. Copyright © 2001 John Wiley & Sons, Ltd.
The high lifetimes recently measured for n-type multicrystalline and CZ silicon make innovative solar cells possible. This paper presents back junction solar cells made on n-type silicon, including a variety of simple designs based on the use of aluminum to form the pn junction, either with uniform or localized p/sup +/ regions. The feasibility of n/sup +/np/sup +/ continuous junction cast multicrystalline Si cells is demonstrated with a 15% efficiency back junction device (J/sub sc/=33.3 mAcm/sup -2/,V/sub oc/=580 mV, FF=0.776, 4 cm/sup 2/, 180 /spl mu/m thickness). Voltages up to 594 mV have been measured on some mc-Si devices. Using n-type CZ silicon, a 15.8% efficiency has been obtained (J/sub sc/=33 mAcm/sup -2/, V/sub oc/=600 mV, FF=0.8). Localized rear Al junction devices on n-type FZ silicon have reached V/sub oc/=645 mV and J/sub sc/=27 mAcm/sup -2/ (no AR coating), showing the way for improved performance.
Different approaches to reduce the light-induced degradation of Czochralski silicon (Cz-Si) solar cells are investigated. In the first part of this paper the very promising possibility of using overcompensated oxygen-rich n-type silicon with residual boron as solar cell substrate material is demonstrated. Stable bulk carrier lifetimes in the millisecond range are achievable in this material. The second part of this work deals with new technological approaches to reduce the concentration of the metastable defect responsible for the light-induced carrier lifetime degradation in boron-doped Czochralski silicon. A permanent reduction of the defect concentration by a factor of up to 3.5 is achieved with an optimized emitter diffusion process at 850°C in a conventional quartz tube furnace using fast ramping conditions. Long-term annealing treatments performed at low temperature (450°C) are shown to reduce the metastable defect concentration by a factor of up to 3.3 and, at the same time, lead to the formation of thermal donors (TDs). These results are in good agreement with our new defect model which assumes a fast-diffusing oxygen dimer as part of the Cz-specific defect.
This paper reports the recent improvements in the energy conversion efficiencies of solar cells on magnetically-confined Czochralski grown (MCZ) and float zone (FZ) silicon substrates at the University of New South Wales. A PERT (passivated emitter, rear totally-diffused) cell structure has been used to reduce the cell series resistance from higher resistivity substrates. The total rear boron diffusion in this PERT structure appears to improve the surface passivation quality of MCZ and some FZ substrates. Hence, higher open-circuit voltages were observed for some PERT cells. One of these cells on MCZ substrates demonstrated 24·5% energy conversion efficiency at Sandia National Laboratories under the standard global spectrum (100 mW/cm²) at 25°C. This is the highest efficiency ever reported for a MCZ silicon solar cell. The cells made on MCZ substrates also showed stable cell performance rather than the usually reported unstable performance for boron-doped CZ substrates. Also reported is a PERL (passivated emitter, rear locally-diffused) cell on a FZ substrate of 24·7% efficiency, which is the highest efficiency ever reported for any silicon solar cell. Copyright © 1999 John Wiley & Sons, Ltd.
This paper reports the recent improvements in the energy conversion efficiencies of solar cells on magnetically-confined Czochralski grown (MCZ) and float zone (FZ) silicon substrates at the University of New South Wales. A PERT (passivated emitter, rear totally-diffused) cell structure has been used to reduce the cell series resistance from higher resistivity substrates. The total rear boron diffusion in this PERT structure appears to improve the sinface passivation quality of MCZ and some FZ substrates. Hence, higher open-circuit voltages were observed for some PERT cells. One of these cells on MCZ substrates demonstrated 24.5% energy conversion efficiency at Sandia National Laboratories under the standard global spectrum (100 mW/cm2) at 25 °C. This is the highest efficiency ever reported for a MCZ silicon solar cell. The cells made on MCZ substrates also showed stable cell performance rather than the usually reported unstable performance for boron-doped CZ substrates. Also reported is a PERL (passivated emitter, rear locally-diffused) cell on a FZ substrate of 24.7% efficiency, which is the highest efficiency ever reported for any silicon solar cell.
The misfit dislocations that are introduced into silicon (001), (111) and (110) wafers by boron diffusion are found to nucleated from the diffusion surface as half-loops with Burgers vectors inclined to the surface. The expansion of the dislocation half-loop on its glide plane leaves a 60° misfit dislocation in the diffusion front. Numerous dislocation reaction products which are the results of the interation among the half-loops during their expansion are characterized by cross-section and plan-view transmission electron microscopy. The dislocations with density of the order of 108 cm-2 are distnbuted mainly in the region that is in a certain depth below the surface, leaving the top region basically dislocation free. The misfit dislocations that are nucleated from surface are not enough to relax all the strain in the diffused layer.
An investigation of the CrO//3-H//2O-HF system shows that preferential etching of crystal defects on silicon surfaces is very sensitive to the concentration ratio of CrO//3 to HF. This leads to development of a new etch consisting of one part by volume of 1. 5 molal (M) CrO//3 (150 g/lH//2O) and one part of 49% HF. This etch can delineates a wide variety of crystal defects with sharp definition. The shape of dislocation etch pits is uniquely determined by the orientation of wafer surface and dislocation lines.
Area-averaged carrier lifetimes of 200 µs in n-type block-cast multicrystalline silicon (mc Si) and 40 µs in n-type edge-defined film-fed grown silicon (EFG Si) have been measured in the as-grown state. These life-time values are about one order of magnitude higher compared to lifetimes typically measured in as-grown p-type mc and EFG Si and they were found to be perfectly stable under illumination. Phosphorus gettering improved the life-times in both types of material. A subsequent bulk hydrogenation step was found to be particularly effective in n-type mc Si, resulting in exceptionally high area-averaged lifetimes of 700 µs, with local lifetimes exceeding 1 ms. Impor-tantly, no low-lifetime regions could be detected after hydrogenation. To our knowledge, a lifetime distribution of the observed outstanding homogeneity has not been observed before in any block-cast mc Si material. Finally, we dem-onstrate the feasibility of easy-to-fabricate n + np + solar cells featuring a screen-printed rear-junction Al-p + emitter. In our first cell batches we have achieved efficiencies up to 16.0 % on n-type monocrystalline silicon and efficiencies up to 11.8 % on n-type mc Si, demonstrating the potential of the developed low-cost cell process on n-type silicon mate-rials.
We investigate the steady-state photoconductance of an oxidized low-lifetime monocrystalline Si wafer with an inversion layer at its surfaces. Photogenerated electrons and holes reduce the band bending and decrease the width of the carrier depleted space-charge region. Mobile charge carriers are stored on both sides of the space-charge region and dominate the photoconductivity at a low illumination intensity. This charge storage effect disappears under accumulation. We present an analytic model for the experimental observations. It is necessary to account for the charge storage effect when deducing low (<10 μs) minority carrier lifetimes on surface-inverted solar Si wafers from one-sun steady-state photoconductance measurements. © 2003 American Institute of Physics.
Objectives in development of the solar plant for the Telstar spacecraft were to provide a power source which would withstand launching stresses and the expected space environment, with optimum end-of-life performance. Radiation damage to the silicon solar cells is the primary factor limiting their useful life; the effect of energetic protons or electrons is the generation of recombination centers in the silicon which reduce the minority-carrier diffusion length and therefore the long-wave response of the cell. The spacecraft solar cells use the n-on-p structure, in preference to conventional p-on-n structure, to obtain a factor of 3 to 10 increased life expectancy. Response to light in the 0.4 to 0.7 micron range is enhanced by using a thin n-layer (about 0.5 micron) and an antireflection coating with minimum reflectance at 0.55 micron wavelength. Early estimates of electron and proton fluxes in the satellite orbit showed that even the best cells would not give sufficient life without radiation shielding. Therefore the cells are protected against electrons of energy up to 1 Mev by 0.3 gm/cm2 sapphire cover plates. The cell mountings are designed to withstand peak vibration stresses of 200 g and repealed temperature cycles from + 65°C to −100°C. The 3600-cell solar power plant is composed of 300 twelve-cell groups of 1 em X 2 cm cells, yielding a nominal initial power of 14 watts at 28 volts for any spin-axis orientation relative to the sun. Telemetry information on performance of the solar plant indicates degradation of the shielded solar cells equal to that measured in the laboratory on unshielded cells with a 1-Mev normal incidence flux of 6 × 1012 electrons/(cm2 day). From this comparison it is estimated that the plant will degrade to 68 per cent of its initial output after two years in orbit.
A set of p-type Czochralski (Cz) silicon materials grown by Shin-Etsu Handotai was used for a comprehensive investigation, including carrier lifetime measurements and fabrication of high-efficiency solar cells at Fraunhofer ISE. The set of different materials consists of gallium and boron doped wafers grown with the Cz method and boron doped wafers grown with the magnetic Czochralski (MCz) method. A clear correlation of the Cz-specific lifetime degradation and the concentration of boron and interstitial oxygen was observed. Thus, gallium-doped wafers with a high concentration of interstitial oxygen of 13.7 ppm showed no degradation. Excellent stable lifetimes of 1098 mu s and 862 mu s were determined for boron-doped MCz wafers and for gallium-doped Cz wafers, respectively. This high lifetime level was maintained or even improved throughout the cell process optimized for Cz silicon and record efficiencies of 22.7% and 22.5% were achieved for boron-doped MCz silicon and gallium-doped Cz silicon, respectively.
In this paper, we report energy conversion efficiencies of 20.8% from CZ and 22.7% from FZ n-type silicon substrates, by a rear emitter PERT (passivated emitter, rear totally-diffused) cell structure. Record efficiency of 22.7% and excellent open-circuit voltage of 706 mV have been demonstrated by these rear emitter cells on n-type FZ substrates. The comparable high efficiency from the CZ substrates also confirmed that high cell efficiencies can be achieved from these lower quality CZ silicon substrates. Since the CZ cell results came from the first batch of such CZ cells, it shows the minimum performance boundary for these CZ cells. Higher performance is expected from future experiments, due to very little difference in the experimental minority carrier lifetimes between CZ and FZ substrates
Laser based processing has been incorporated into many successful solar cell technologies over the past 20 years; for example buried contact solar cells, laser-fired back contacts and laser texturing. However, the impact of crystal damage generated during laser processing on solar cell performance is still uncertain. This paper investigates laser-induced defects that result from a laser ablation process similar to that used to form grooves in the buried contact solar cell. The first part of this paper focuses on the formation of the defects and uses the Yang etching technique to investigate the impact of chemical etching and thermal cycles on their propagation. The second part of the paper takes a close look at the electrical properties of laser-induced defects, experimentally investigates the recombination of minority-carriers at laser-induced defects, and the potential of these defects to shunt pn-junctions.
A simplified laboratory process with one photolithographic step for front junction solar cells on n-type multicrystalline (mc) silicon has been developed. The emitter diffusion is done in an open tube furnace with BBr<sub>3</sub> and back-surface-field diffusion using POCl<sub>3</sub>, loading the wafers front-to-front and back-to-back respectively and thus avoiding additional etching steps. The front surface has been passivated by a 10 nm thermal oxide grown in a tube furnace. With this simple process, efficiencies of 11.0% on n-type mc-Si and 11.5% on n-type Cz-Si have been realized without antireflection coating and without surface texture. Applying a double layer antireflection coating (DARC) on these cells, efficiencies of 16.4% on Cz-Si and 14.7% on mc-Si have been achieved.
Boron and phosphorus diffusions are used in many high-efficiency mono-crystalline silicon solar cell designs to form collecting junctions, localized contact diffusions, and back surface fields. The diffusion of boron or phosphorus generates stress on the silicon lattice as a result of their atomic mismatch with silicon. If this stress exceeds the fracture stress of silicon, a misfit dislocation network is generated. This paper uses photo-conductance lifetime measurements and Yang defect etching to demonstrate that boron and phosphorus misfit dislocation networks result in bulk asymmetric Shockley-Read-Hall (SRH) recombination. Finally, the presence of diffusion-induced misfit dislocations in high-efficiency silicon solar cells is demonstrated to result in bulk asymmetric-SRH, and reduced fill factor.
N-type silicon wafers have been found to have higher bulk lifetimes compared to those of boron-doped p-type silicon wafers with the same resistivity, and proved to have no light-induced degradation associated with the boron-oxygen complex. In this paper, the laser-grooved buried contact (BC) technology was used to fabricate interdigitated backside contact (IBC) solar cells on the phosphorus-doped silicon wafers. Three obstacles that hindered the performance of n-type interdigitated backside buried contact (IBBC) solar cells - parasitic shunt resistance, metallization issues, and optimization of the diffused regions - are discussed. 19.2% efficiency has been achieved on the n-type IBBC solar cell, and all BC solar cells made on the phosphorus-doped n-type silicon wafers, regardless of the crystalline growth techniques, show no light-induced degradation.
In n-type silicon the capture cross sections of metallic impurities, are neatly smaller than in p-type. So, lifetime and also diffusion length of minority carriers should be neatly higher. This is of a paramount interest for multicrystalline silicon wafers, in which the impurity-defects interaction governs the recombination strength of minority carriers. In 1.2 Ωcm wafer, lifetime is found around 200 μs and diffusion lengths around 220 μm, These values increase strongly after gettering treatments like phosphorus diffusion or AI-Si alloying. Scan maps reveal that extended defects are poorly active, even when the density of dislocation is higher than 10<sup>5</sup> cm<sup>-2</sup>. High quality abrupt p<sup>+</sup>n junctions are obtained by Al-Si alloying and annealing at 850 or 900 °C, which could be used for rear junction cells.
The silicon integrated‐circuits chip is built by contiguously embedding, butting, and overlaying structural elements of a large variety of materials of different elastic and thermal properties. Stress develops in the thermal cycling of the chip. Furthermore, many structural elements such as CVD (chemical vapor deposition) silicon nitride, silicon dioxide, polycrystalline silicon, etc., by virtue of their formation processes, exhibit intrinsic stresses. Large localized stresses are induced in the silicon substrate near the edges and corners of such structural elements. Oxidation of nonplanar silicon surfaces produces another kind of stress that can be very damaging, especially at low oxidation temperatures. Mismatch of atomic sizes between dopants and the silicon, and heteroepitaxy produce another class of strain that can lead to the formation of misfit dislocations. Here we review the achievements to date in understanding and modeling these diverse stress problems.
Thermal stress and subsequent plastic deformation can be caused in a thin slice of silicon by temperature gradients which arise during the transient periods of heat cycling. They arise either because of the radiation shielding effect of neighboring slices in a parallel row of slices, or the heat sink effect of the boat on which the slices stand, or because of the effect of both. Spacing between the parallel slices, type of boat used, heating and cooling rates, and furnace temperature affect the temperature gradients and the amount of plastic deformation. Study of slip patterns showed that the tangential component of thermal stress was responsible for slip, and that multiple slip was more common than single slip under the conditions investigated.
Depletion-region modulation (DRM) has recently been identified as a mechanism that influences photoconductance lifetime measurements. The effect is observed in semiconductor samples containing a depletion-region (i.e., p-n junction solar cells). Experimental measurements presented within demonstrate that the DRM effect dominates the conductance measurement at low excess carrier concentrations, resulting in an overestimation of the effective lifetime by several orders of magnitude. The influence of substrate thickness on the DRM effect is experimentally verified. The previously developed analytical equation for DRM is in agreement with our experimental data and can be used to correct DRM affected photoconductance lifetime measurements. Finally, the impact on the sensitivity of a photoconductance measurement is discussed for the DRM corrected case. © 2003 American Institute of Physics.
Silicon nitride (SiN) fabricated by plasma-enhanced chemical vapour deposition (PECVD) is increasingly used within the crystalline silicon (c-Si) photovoltaic industry as it offers the possibility to fabricate a surface and bulk passivating antireflection coating at low temperature (⩽450°C). This article presents an overview on the present status of SiN for industrial as well as laboratory-type c-Si solar cells. Topics covered include the fundamentals of the PECVD technology, the present status of high-throughput PECVD machines for the deposition of SiN onto c-Si wafers, and a review of the fundamental properties of Si–SiN interfaces fabricated by PECVD.
An investigation of n-type multicrystalline silicon grown by directional solidification has produced several important findings: i) demonstration of effective phosphorus gettering; ii) achievement of minority carrier lifetimes above one millisecond; iii) verification of good stability under illumination. The lifetimes after gettering show a strong dependence on doping: 1.6 ms for 2.3 /spl Omega/cm, 500 /spl mu/s for 0.9 /spl Omega/cm and 100 /spl mu/s for 0.36 /spl Omega/cm, respectively. Lifetime mapping by infrared carrier density imaging has revealed a large surface variability of this parameter, which is detrimental for large area devices. A minor degradation of the lifetime after light exposure has been observed for the highest lifetime regions, while other wafers and regions remained essentially stable.
A world record total area conversion efficiency of 20.7% and high
open circuit voltage (VOC) of 719 mV were achieved on a solar cell with
HIT (heterojunction with intrinsic thin-layer) structures on both sides
(wafer size: 100.5 cm<sup>2</sup>, n-type solar-grade CZ-Si). This solar
cell was fabricated with the same process as that used in our
mass-production lines. The essence of this high performance is derived
from the excellent passivation ability of the HIT structure on c-Si.
This report discusses research for excess of 20% efficiency HIT cell
(~100 cm<sup>2</sup>), focusing on the a-Si passivation effect estimated
from the carrier lifetime, and describes product development for the
industrialization of HIT cells
In the literature it is well known that the low-injection bulk
carrier lifetime of boron-doped Cz-grown silicon is not a constant
material property but, depending on previous thermal treatments and
light exposure, varies between two states corresponding to a high and a
low lifetime value. The upper state is obtained by means of
low-temperature annealing, while illumination degrades the lifetime
towards the value of the lower state. In order to improve the
understanding of this phenomenon, we performed comprehensive carrier
lifetime measurements on solar- and electronic-grade boron, gallium, and
phosphorus doped Cz wafers obtained from different manufacturers. Based
on the experimental results, a new model is introduced which attributes
the disappointingly low stable lifetimes of illuminated boron-doped Cz
silicon with resistivity around 1 Ωcm to boron-oxygen pairs. From
this model, simple recipes are derived which might lead to an
improvement of the efficiency of commercial Cz silicon solar cells
The buried contact (BC) technology has demonstrated both an
efficiency and cost advantage over conventional screen printed solar
cells. New BC structures, in particular the double sided (DS) BC cell,
allow further improvements in cost and efficiency. Improvements in
efficiency arise through improved rear surface passivation. Experimental
results from DSBC cells using various passivation methods demonstrate
that a floating junction (FJ) passivates as well as passivation schemes
used with high efficiency cells. 2D analysis and experimental results
both show localised defects have prevented FJ passivation from achieving
its potential and that optimization of the rear doping or by bifacial
operation can improve performance
In this paper, we demonstrate for the first time the use of the laser-grooved solar cell technology, a proved commercial technology, for the implementation of the rear junction backside contact solar cells. Laser-grooved backside contact solar cells, designated as interdigitated backside buried contact (IBBC) solar cells, have been fabricated on planar, n-type, 1 Ω·cm wafers with a single layer SiO<sub>2</sub> anti-reflection coating, achieving 17% efficiency with open-circuit voltage (V<sub>oc</sub>) of more than 680 mV. Front and rear surface recombination velocities of 350 cm/s and 4800 cm/s, and more than 1 ms of post-processing bulk lifetime confirm that commercial laser-grooved solar cell fabrication process is capable of obtaining the efficiency advantages of the rear junction, backside contact design. Moreover, this paper presents a side-by-side comparison between the more conventional double-sided buried contact solar cell and the IBBC solar cell. The advantages of higher short-circuit current in the latter design due to no contact shading loss, and higher V<sub>oc</sub> due to inherently lower surface recombination velocity of the IBBC structure are demonstrated.
Significant interest has recently been shown in the use of dark
and illuminated current-voltage (I-V) measurements for the
characterization of high-efficiency silicon solar cells. Similar
nonideal behavior, in the form of “humps” in dark I-V
curves, has been observed by various research groups but apparently
different interpretations of this effect given. In this paper we present
detailed computer simulations of solar cells with defects (producing
recombination centers within the bandgap) at a number of specific
positions in the devices. It is found that a distinct shoulder (or
“hump”) occurs in the I-V characteristics when the
recombination centers exhibit unequal electron and hole capture rates.
Furthermore, it is shown that these shoulders are a result of the
saturation of (Shockley-Read-Hall) recombination via the defect levels,
which dominates behavior at low forward bias. As the bias voltage is
increased, recombination in the defected region increases again, beyond
the saturation level. The simulations show conclusively that the
shoulders in the measured dark I-V curves of high-efficiency silicon
solar cells produced at the University of New South Wales arise from the
rear Si-SiO<sub>2</sub> interface
Solar cells on n-type silicon materials with screen-printed rear alumi-nium-p+ emitter
- A Schmiga
- M Froitzheim
- A Ghosh
- J Metz
- R Schmidt
- Brendel
Schmiga, A. Froitzheim, M. Ghosh, A. Metz, J. Schmidt, and R. Brendel, " Solar cells on n-type silicon materials with screen-printed rear alumi-nium-p+ emitter, " in Proc. 20th EUPVSEC, 2005, pp. 918–921.
where he worked on the devel-opment of thin-film crystalline-silicon solar cells Australia, as a Postdoctoral Fellow, working on crystalline-silicon solar cell research and devel-opment. He is currently a Senior Lecturer with the Centre of Excellence in Silicon Photonics and Photovoltaics
- A G J E Aberle
- B S E E Cotter
A. G. Aberle, Crystalline Silicon Solar Cells; Advanced Surface Passi-vation and Analysis. Sydney, Australia: Centre Photovoltiac Eng., Univ. New South Wales, 1999. J. E. Cotter received the B.S.E.E., M.S.E.E., and Ph.D. degrees from the University of Delaware, Newark, in 1987, 1989, and 1997, respectively. From 1989 to 1996, he was with AstroPower, Inc., Newark, DE, where he worked on the devel-opment of thin-film crystalline-silicon solar cells. In 1997, he joined the Photovoltaics Special Research Centre, University of New South Wales, Sydney, N.S.W., Australia, as a Postdoctoral Fellow, working on crystalline-silicon solar cell research and devel-opment. He is currently a Senior Lecturer with the Centre of Excellence in Silicon Photonics and Photovoltaics, University of New South Wales, where he leads the development of commercial silicon solar cells. J. H. Guo received the B.Sc. and M.Sc. degrees
The effect of SRH on the fill factor of double sided buried contact solar cells
- P J Cousins
- C B Honsberg
- A B Sproul
- J E Cotter
P. J. Cousins, C. B. Honsberg, A. B. Sproul, and J. E. Cotter, " The effect of SRH on the fill factor of double sided buried contact solar cells, " in Proc. Australian and New Zealand Solar Energy Soc. Conf., 2002, p. 3B4.