Article

Optimized Adhesion of Plated Silicon Solar Cell Contacts by F 2 -Based Dry Atmospheric Pressure Nano-Roughening

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

The adhesion of a plated layer on a substrate is increased by an appropriate roughness at the interface. The paper reports on plated Ni‐Cu‐Ag contacts for silicon solar cells deposited into passivation layer openings that are created using a nano‐second pulsed laser. The adhesion of plated contacts on ns‐pulsed laser‐structured silicon ab initio is not sufficient. This is attributed to an improper surface topography. Atmospheric pressure dry etching in F2 is introduced as plating pre‐treatment to generate a beneficial nano‐roughness on the silicon substrate. Until now, this method is used in silicon photovoltaics only for the formation of black silicon. Process parameters are adjusted to allow the formation of cavities in the range of 5–30 nm. Thereby, the contact adhesion increases. In peel force tests on busbars, the average peel‐force raised from 0.3 to 2 N mm−1. In sheer‐test on finger contacts an increase of maximum sheer force and a decreasing length of the finger displacement are observed. Due to the high etch selectivity between silicon and silicon nitride, no additional etch mask is required to protect the passivation layer. As precursors with a non‐optimized emitter design are used, the roughening procedure affects the solar cell efficiency. Atmospheric pressure dry etching in F2 is introduced as plating pre‐treatment to generate a beneficial nano‐roughness on the silicon substrate. The method is shown to improve the adhesion of solar cell contacts that are structured by local ablation of the passivation layer using ns‐pulsed lasers.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

Thesis
Die vorliegende Arbeit befasst sich mit lasergestützten Diffusionsprozessen zur Herstellung hocheffizienter „passivated emitter and rear cells“ (PERC-Solarzellen) mit laserdotiertem selektiven Emitter (LDSE). Das zentrale Thema der Arbeit stellt die „Laserdiffusion aus dem Phosphorsilikatglas (PSG)“ dar. Dabei handelt es sich um einen weit verbreiteten Ansatz zur LDSE-Implementierung. Das nach der Rohrofendiffusion vorliegende PSG wird dabei als Dotierstoffquelle genutzt, um den Emitter in den designierten Bereichen der Vorderseitenmetallisierung nachzudotieren. Dies geschieht typischerweise mittels gepulster Laserstrahlung, durch die PSG und Silizium lokal und für weniger als eine Mikrosekunde pro Laserpuls aufgeschmolzen werden. Die Arbeit vereint drei Themenkomplexe: Solarzellen aus monokristallinem Silizium, die Lasertechnologie und speziell die Laserdiffusion aus dem PSG. Jedem dieser Themenkomplexe ist ein eigenes Kapitel gewidmet, in dem die für diese Arbeit relevanten theoretischen Grundlagen erläutert werden. Im Anschluss an diese ersten, grundlegenden Kapitel werden vielseitige simulative und experimentelle Untersuchungen zur LDSE-Implementierung vorgestellt (inklusive der zum Einsatz gekommenen Charakterisierungsmethoden). Dabei wird unter anderem den Fragen nachgegangen, welche Rolle die zwischen PSG und Silizium befindliche Siliziumdioxidschicht („Zwischenoxid“) bei der Laserdiffusion spielt und ob es im Zuge der Laserdiffusion zum Verdampfen des PSG-Zwischenoxid-Stapels („PSG-Schichtstapel“) kommt. Des Weiteren wird erforscht, inwiefern die Mikrostruktur an der Grenzfläche zwischen LDSE und Metallisierung den spezifischen Kontaktwiderstand und die Emittersättigungsstromdichte beeinflusst. Neben der Beantwortung dieser und weiterer wissenschaftlichen Fragestellungen widmet sich die vorliegende Arbeit drei, im Hinblick auf die zukünftige LDSE-Generation möglicherweise wegweisenden, technologischen Ansätzen: einem Verfahren zur hochpräzisen Alignierung zwischen LDSE und Vorderseitenmetallisierung, dem Einsatz eines Polygon-Scanners zur Beschleunigung der LDSE-Prozessierung und einem innovativen Ansatz zur LDSE-Implementierung, der die Veränderung gleich mehrerer Schritte in der Prozesskette zur Herstellung von PERC-Solarzellen vorsieht. Was das Prozessverständnis angeht, so liefert diese Arbeit wichtige Erkenntnisse bezüglich der Laserdiffusion. Hervorzuheben ist hierbei die Erkenntnis, dass im Fall einer „random pyramids“ Textur ein charakteristischer Schwellwert der Laserintensität existiert, durch den sich zwei Prozessregime ergeben. Oberhalb dieses Schwellwertes führt die Laserdiffusion zu einem Abplatzen des PSG-Schichtstapels sowie zur Planarisierung der Pyramidenstruktur. Je stärker die Planarisierung der Siliziumoberfläche, desto weniger Silberkristallite bilden sich während des Feuerprozesses pro Flächenelement und desto größer ist daher der spezifische Kontaktwiderstand. So lautet das Fazit vielschichtigen Untersuchungen zur Mikrostruktur der Grenzfläche zwischen LDSE und Metallisierung. Diese umfassen eine Rasterelektronenmikroskop Studie, „scanning spreading resistance microscopy“ Messungen (die erstmalig die Dotierstoffverteilung in einem LDSE auf Mikrometer-Skala auflösen) und die Modellierung des Kontaktwiderstands an den Silberkristalliten. Für Laserintensitäten unterhalb des charakteristischen Schwellwertes wird in dieser Arbeit ein neuartiges Modell entwickelt, das den Ablauf der Laserdiffusion beschreibt. Diesem Modell zufolge bleibt im Prozessregime „geringer Laserintensitäten“ die Pyramidenstruktur der Siliziumoberfläche größtenteils erhalten, ebenso der PSG-Schichtstapel. Insbesondere gilt dies für die Pyramidentäler, in denen die Temperatur geringer ist als an den Pyramidenspitzen. Angenommen werden kann, dass im Bereich der Pyramidentäler die Schmelztemperatur des Zwischenoxids nicht überschritten wird. Das Zwischenoxid bleibt fest und verhindert somit die Eindiffusion von Phosphoratomen aus dem PSG ins Silizium. Anders verhält es sich im Bereich der Pyramidenspitzen. Hier staut sich die laserinduzierte Wärme, sodass Silizium, PSG und Zwischenoxid geschmolzen werden. Entsprechend stark werden die Pyramidenspitzen dotiert bzw. verformt. Die obigen Ausführungen lassen erahnen, wie komplex die Entwicklung laserdotierter selektiver Emitter ist. Dieser Umstand wird in der vorliegenden Arbeit eingehend diskutiert, zum einen anhand von Simulationen, die mittels der Software „Quokka3“ durchgeführt wurden, zum anderen im Kontext zweier Experimente, in denen durch die LDSE-Implementierung eine Steigerung von offener Klemmspannung und Füllfaktor im Bereich von 7,3 mV bzw. 1,6% erzielt werden konnte. Die Diskussion über die Herausforderungen bei der Herstellung von Solarzellen mit LDSE mündet in der Evaluierung eines innovativen Ansatzes zur LDSE-Implementierung. Dieser geht auf die Promotion von Marius Meßmer am Fraunhofer ISE zurück und wurde oben bereits angesprochen. Bei diesem Ansatz wird ein Großteil des Temperaturbudgets, das im Zuge der Rohrofendiffusion zur Emitterausbildung aufgewandt wird, in die thermische Oxidation „verlagert“. Letztere erfolgt im Anschluss an die Laserdiffusion, die PSG-Ätze, die Kantenisolation und eine Reinigung der Wafer. Die vorliegende Arbeit zeigt, dass bei diesem Ansatz der spezifische Kontaktwiderstand weitestgehend unabhängig von den zur LDSE-Prozessierung gewählten Prozessparametern ist, und mit ≤0,6 mΩcm2 gleichzeitig extrem niedrig. Nachgewiesen wird außerdem, dass es zur Ausheilung laserinduzierter Defekte kommen kann. Demonstriert wird dies für eine neuartige Form der LDSE-Prozessierung, die oben bereits anklang, die Kombination eines Polygon-Scanners mit einer infraroten, „continuous wave“ Laserquelle. Dieser innovative Ansatz zur LDSE-Prozessierung ermöglicht eine Prozessdauer von 0,4 s pro Solarzellenwafer und setzt in dieser Hinsicht neue Maßstäbe. Neue Maßstäbe setzt auch der dritte technologische Ansatz, der zuvor erwähnt wurde, das Verfahren zur hochpräzisen Alignierung zwischen LDSE und Vorderseitenmetallisierung. Demonstriert wird in dieser Arbeit die passgenaue Alignierung zwischen einem äußerst schmalen LDSE mit 26 µm Breite und 7 µm breiten „laser contact openings“, die zur Kontaktierung des LDSE mittels Plating dienen. Die vorliegende Arbeit leistet somit einen relevanten Beitrag zur Forschung im Bereich laserdotierter selektiver Emitter, nicht nur durch die Erweiterung des Prozessverständnisses, sondern auch in technologischer Hinsicht.
Article
Full-text available
The metallization of silicon heterojunction (SHJ) solar cells by electroplating of highly conductive copper onto a multifunctional patterned metal layer stack is demonstrated. The approach features several advantages: low temperature processing, high metal conductivity of plated copper, no organic making, and low material costs (almost Ag‐free). A PVD layer stack of copper and aluminum is deposited onto the cell subsequently to TCO deposition. The aluminum layer is patterned with a printed etchant and its native oxide on the remaining areas inhibits plating. The full area aluminum layer while electroplating supports plating current distribution and allows homogeneous plating height distributions over the cell. The NOBLE (native oxide barrier layer for selective electroplating) approach allows reaching a first encouraging SHJ solar cell efficiency of 20.2% with low contact resistivity.
Conference Paper
Full-text available
One of the key topics for industrial implementation of fully plated NiCu contacts on Si solar cells has so far been metal adhesion to silicon. The evaluation of adhesion is not entirely consistent throughout PV research, which is discussed at first. The adhesion issue of plated NiCu contacts has then been addressed by following two different routes for adhesion promotion. Firstly an "Interim Anneal" route is focussing on controlled silicide formation as a bonding agent and reliably excellent adhesion greater than wafer breakage force is shown independent of surface roughness (above 2 N/mm for measurements on clamped cells and above 4 N/mm for fortified wafers). Secondly a "Post Anneal" route is focussing on creation of high Si surface roughness by ps laser ablation of the ARC and annealing after full metallization. Excellent adhesion above wafer breakage force is demonstrated. Comparisons to different methods of ARC patterning and resulting surface roughness are made. The presented metallization technology is feasible for standard BSF cells and next generation solar cells for both p-type and n-type emitters. Solar cell efficiencies of up to 19.4% abs for p-type BSF cells, 20.5% abs for n-type BSF cells and 22.1% abs for n-type PassDop cells are presented.
Article
Full-text available
This paper presents first results of combining two promising future technologies: i) metal catalyzed textured diamond-wire sawn (DWS) mc-Si wafers and ii) nickel-copper-silver (Ni/CuAg) plated front contact processing. Results of first optimizations of laser patterning, annealing and module string soldering are presented. A 60-cell module with a power output of 266 W was successfully manufactured. The pseudo fill-factor (pFF) has been identified as the parameter with the highest potential for further improvements. A very promising approach is moving from an emitter profile that is mainly adapted to silver paste properties to an optimized emitter profile for Ni/Cu/Ag-plated contacts. The idea is to reduce the surface doping concentration (reduces Auger recombination) and increase the junction depth (increased pFF and FF).
Article
Full-text available
In this work, we highlight the benefits of alternative plating routes compared to the standard plating route primarily discussed in literature and currently introduced in pilot production [1, 2]. The common plating route starts with the laser ablation of the front side grid after rear side Al-printing and firing. Then, an HF dip removes the native oxide for the subsequent light induced nickel-, copper plating (LIP) with a silver capping as a finishing. Afterwards, an annealing step improves the adhesion of the plated grid and its contact resistance. Here, we want to show the advantage of three alternative process routes. In the first alternative route, the laser opening of the front side grid is performed before the firing step. During firing, the laser damage is partly cured and can lead to an increase in open circuit voltage (Voc) of 7 mV on state of the art industrial PERC solar cells. For the second process route, the removal of the native oxide is eliminated. To achieve this, special laser and plating conditions are needed. Finally, for the third alternative process route, the annealing of the plated stack can be combined with a stabilization process suppressing the light induced degradation (LID). The presented alternative routes give new degrees of freedom for process optimizations regarding precursor-induced features such as high laser damage on shallow emitters, parasitic plating (PP) for passivation layers with high pinhole density or LID.
Article
Full-text available
A narrow size distribution of quantum dots (QDs) is needed for their application in photovoltaics but collection of such information is difficult. This paper demonstrates the application of Raman spectroscopy as a characterisation tool to extract the size distribution and crystalline fraction of Si QD samples fabricated through the sputter-anneal method. Measured Raman spectra of Si QD materials are de-convoluted into four components according to their origins and Raman scattering by Si QD cores is described by a modified one phonon confinement model, while other components are reproduced with Gaussian functions. Through fitting of Raman spectra, Si QD size distributions and Si crystalline fractions are obtained. The results are compared with the values extracted from PL modelling on a series of B doped Si QD samples. The good consistency between the values extracted by these two methods confirms the validity of the Raman model. The result confirms that Si crystallization has been suppressed by B doping as the average Si QD size and Si crystalline fraction are reduced with increased B doping level.
Article
Full-text available
The 6th Metallization Workshop took place in Constance, Germany on 2 and 3 May 2016. At the workshop the latest progress in the understanding and application of metallization and interconnection was presented. Screen printed metallization continues to dominate. Material and application technologies are constantly further improved, with sub-40μm fingers with high cell performance and low Ag consumption demonstrated. Cu plating technology is further perfected in anticipation of large scale industrial implementation, with improvements on adhesion and long term reliability. In interconnection, alternatives to the traditional ribbon soldering technology are proposed. Among them, the multi-wires interconnection schemes are shown to have a dramatic impact on metallization design and technology.
Article
Full-text available
Copper plating can reduce the consumption of silver for silicon photovoltaic manufacturing, whilst also offering the potential to increase cell efficiency by way of reduced shading due to very narrow fingers and contacting silicon surfaces with low phosphorus concentrations. However, it can be challenging to plate busbars and fingers that have sufficient adhesion to the silicon. To date adhesion has typically been assessed by busbar pull tests however we propose that this measurement does not consider the properties of copper-plated fingers and may not be a good indicator of whether fingers may dislodge or peel with subsequent processing or during module fabrication. In this paper we investigate the factors of silicon interface roughness and plated copper properties using a combination of busbar pull tests and stylus-based adhesion measurements. We show that average 180° pull test forces of 2.1 N/mm can be achieved when a UV ps laser is used to ablate the silicon nitride, however ensuring strong finger adhesion is a far more complex problem with no accepted standard to determine what is “sufficient” adhesion. Although use of a ps laser to ablate the silicon nitride can result in plated metal adhering so strongly to the silicon that fragments of silicon are broken off with the finger when it is dislodged by the stylus, use of fast plating rates can result in reduced finger dislodgement forces and excessive finger peeling whereas busbar pull test forces are largely unchanged because the increased plating current is directly mostly through the fingers of the plated metal grid during light-induced plating. The plating of busbars and fingers on a cell presents challenges for uniform silicon-metal adhesion and this paper highlights the importance of finger adhesion measurement for process quality control for nickel/copper plating of p-type silicon cells in a manufacturing environment.
Conference Paper
Full-text available
Replacing Ag paste contacts in silicon solar cells by plated Ni/Cu contacts seems a logical next step in the evolution of industrial Si solar cell manufacturing. Ag paste contacts cause a significant share of the solar cell cost today and limit the efficiency of advanced Si solar cells. However, replacing a proven technology by another requires reliability of this technology. Cost and efficiency advantages alone are a high motivation for adopting a new metallization technology in mass production. Reliability of the contacts is a must. In this contribution we show that laser ablation followed by light induced plating of Ni and Cu and plating of a thin capping layer results in good and reliable contacts on industrial solar cell precursors. After plating of the complete metal stack a thermal annealing step is used to increase mechanical adhesion and to reduce resistive losses of the plated contacts (contact resistance and grid resistance). Excellent solar cell efficiencies can be combined with reliable contacts. Adhesion data and data after 60 cell module testing (as part of IEC61215) are reported next to the most influencing factors.
Patent
Full-text available
The present disclosure describes methods of fabricating a semiconductor device. An exemplary method includes forming a metal pattern on a substrate and etching the metal pattern using an etchant including at least an alkaline solution and an oxidant to form a metal electrode, where at least a portion of the surface of the metal electrode is uneven.
Article
Full-text available
Ni/Cu two-step plating is a promising metallization technique because low contact resistance and improved contact adhesion can be achieved after the Ni annealing process. Also, narrow fingers, which are required for high-efficiency solar cells, can be formed by plating. However, the reliability of contact adhesion is still considered one obstacle to industrializing solar cells with plated metal contacts. In this experiment, the influence of ARC opening methods on plated contact adhesion was investigated because the roughnesses of the Si surfaces produced by using pico-second laser ablation and photolithography may be different. Also, the annealing process was conducted before and after plating Cu/Ag metal stacks. The sequence of the annealing can be significant for efficient production because plating is a wet process while annealing is a dry process. The contact adhesion was measured by using a peel-off test. The test was conducted on a 1.5-mm-wide by a 60 ~ 70- mm-long bus bar area. A 3.2-N/mm adhesion force was recorded as a highest average value along the bus bar.
Article
Full-text available
Plated Ni/Cu/Ag contacts are an industrially feasible metallization approach for high efficiency c-Si solar cells with low surface doping concentrations (1018 cm-3 < ND < 1020 cm-3). The 2d-simulations of this work define the minimum requirements on the contact resistivity of metal contacts in a high efficiency solar cell design. The following experimental study of the contact resistivity of plated Ni/Cu/Ag contacts on lowly doped phosphorus emitter demonstrates low contact resistivities in the mΩcm2 regime, which enable solar cells with high fill factors. Furthermore, the paper analyzes the influence of the thermal silicidation process on pseudo-fill factor losses and on the mechanical contact adhesion. The contact adhesion is also studied with respect to the laser contact opening process. The results of this work demonstrate that the right choice of back-end processes enable plated Ni/Cu/Ag contacts with low contact resistivities in combination with high contact adhesions above 1 N/mm.
Article
Full-text available
This work deals with requirements regarding the solar cell process that allow or facilitate the introduction of fabrication processes for front side metallization. By taking experience with plating on solar cells both from the literature and from practical lab work, design rules for the solar cell and the plating process have been derived. Regarding the surface texture, small features (e.g., random pyramids < 5 μm) have been found beneficial for plating processes both on printed seed layers and directly on silicon. A dense, pinhole-free anti-reflective coating suppresses so-called ghost plating as well as careful handling in production. Shunting, which is possible for laser edge isolation can be prevented by applying wet chemical edge isolation. An adjusted soldering process which brings in high energy within a short time is recommended. Infrared or inductive soldering are ideal for the connection of plated surfaces with ribbon. For plating on paste it is recommended that using plating electrolytes with pH>2 result in a good adhesion of the seed layer on the silicon surface. Further the glass content in the silver paste and the size of the glass particles have a big influence on the adhesion. A high glass content and small particles are beneficial for a following plating process. For direct plating on silicon an ablation of the ARC with a 355 nm ps-laser is the best option for adhesion.
Article
Full-text available
Laser microstructuring of thin dielectric layers on sensitive electronic devices, such as crystalline silicon solar cells, requires a careful design of the laser ablation process. For instance, degradation of the substrate’s crystallinity can vastly decrease minority carrier lifetime and consequently impair the efficiency of such devices. Short-pulse laser ablation seems well suited for clean and spatially confined structuring because of the small heat-affected zone in the remaining substrate material [Dube and Gonsiorawski in Conference record of the twenty first IEEE photovoltaic specialists conference, 624–628 1990]. The short-time regimes, however, generate steep temperature gradients that can lead to amorphization of the remaining silicon surface. By ‘heating’ the substrate via a non-ablative laser pulse in the nanosecond regime before the actual ablation pulse occurs we are able to prevent amorphization of the surface of the silicon solar cell substrate, while lowering the ablation thresholds of a SiNx layer on crystalline silicon wafers.
Conference Paper
Full-text available
Strong contact adhesion is an important requirement for durable, manufacturable solar cells. Advanced contacting technologies require new methods to measure adhesion. We describe a scratch test for measuring contact adhesion that involves scanning a weighted stylus across the cell while measuring the horizontal force FD required to dislodge the contacts. FD is characteristic of the adhesive bond but independent of the contact height, stylus weight and scan speed. We observe that contact peeling depends also on the tensile strength of the metal finger. The tests provide a valuable way to assess and optimize the adhesion of metal contacts.
Article
Full-text available
This work discusses the impact of laser annealing on a picosecond laser ablation process of anti-reflection layers on damage etched and random pyramid textured silicon wafers. The laser ablation is realized using picosecond pulsed laser radiation which facilitates a continuously ablated passivation layer but induces a significant reduction in charge carrier lifetime. It is demonstrated that the application of a nanosecond pulsed laser annealing step can improve the electrical properties of the picosecond laser treated area.
Data
Full-text available
This paper describes a galvanic deposition method for the formation of uniformly thin nickel seed layers for silicon solar cells. Unlike the previously-reported electroless and light-induced plating methods, where the thickness of nickel seed layers can be vary significantly due to non-uniform nucleation of metal which can be exacerbated by variations in surface dopant concentrations and roughness, this method saturates resulting in Ni seed layers of uniform thicknesses in the range of 250-400 nm depending on the plating time and patterning process. The method was successfully used in the fabrication of nickel/copper plated homogeneous 100 Ω/□ emitter Si solar cells and laser-doped selective emitter (LDSE) cells, both with screen-printed, Al-alloyed back surface fields. Average cell efficiencies of 18.4% and 18.9% were achieved for the homogeneous emitter cells and LDSE cells, respectively, with the best LDSE cell having an efficiency of 19.2%.
Article
Confocal micro-Raman spectroscopy allows for spatially resolved measurements of the phonon energy in silicon, which is correlated to mechanical stress. Mechanical stress is a tensorial quantity. For the confocal measurement geometry and certain crystal orientations approximations have been derived in the past which correlate the shift of the Raman frequency to a scalar stress value. For optimization of mono-crystalline solar cell manufacturing steps the determination of induced mechanical stress from the top view perspective is desirable. However, this method is so far restricted to planar wafers or cross sections. we find that the anti-reflection surface texture strongly affects the measurement result. To enable quantitative stress determination of alkaline textured silicon a suited measurement procedure is investigated using 3-point bended solar cell segments. Preceding elasticity test revealed Young’s module and allowed a prediction of maximum stress values for any bending radius. We propose a measurement procedure which yields an observable shift of the Raman frequency proportional to the induced stress. We state the requirements for determination of a calibration factor to quantify stress on any alkaline textured solar cell. Adapting the conversion factors allows calibrated measurements for textures with different average pyramid heights. By varying the numerical aperture and the focus setting even changes of the induced stress within the texture pyramids are resolved.
Article
Both buried contact solar cells (BCSC) and laser doped selective emitter (LDSE) solar cells have achieved considerable success in large-scale manufacturing. Both technologies are based on plated contacts. High metal aspect ratios achieved by BCSC allow low shading loss while the buried metal contacts in the grooves provide good contact adhesion strength. In comparison, although the LDSE cell achieves significantly higher efficiencies and is a much simpler approach for forming the selective emitter region and self-aligned metal plating, the metal adhesion strength falls well short of that achieved by the BCSC. Recent studies show that plated contacts based on the latter can be more durable than screen-printed contacts. This work introduces a new concept of laser doping with grooving to form narrow grooves with heavily doped walls in a simultaneous step, with the self-aligned metal contact subsequently formed by plating. This process capitalizes on the benefits of both BCSC and LDSE cells. The laser-doped grooves are only 3–5 µm wide and 10–15 µm deep; the very steep walls of these grooves remain exposed even after the subsequent deposition of the antireflection coating (ARC). This unique feature significantly reduces the formation of laser-induced defects since the stress due to the thermal expansion mismatch between the ARC and silicon is avoided. Furthermore, the exposed walls allow nucleation of the subsequent metal plating. This novel structure also benefits from greatly enhanced adhesion of the plated contact due to it being buried underneath the silicon surface in the same way as the BCSC. Cell efficiencies over 19% are achieved by using this technology on p-type Czochralski (Cz) wafers with a full area aluminum (Al) back surface field (BSF) rear contact. It is expected that much higher voltages and consequently higher efficiencies could be achieved if this technology is combined with a passivated rear approach.
Article
Nickel galvanic displacement (NiGD) plating to silicon surfaces allows for the deposition of self-limiting sub-micron nickel layers. This paper reports the use of NiGD plating as an adhesion-promoting seed layer for light-induced plated nickel and copper (LIP-NiCu) contacts on chemically-etched silicon surfaces. The improved adhesion, which is attributed to surface roughening caused by the oxidation and subsequent etching of silicon in the fluoride-containing electrolyte, is quantified by stylus-based scratch measurements and shown to increase with the duration of sintering at 350 °C. The width-normalised cut-off force for NiGD layers, sintered for 10 min, was approximated by a normal distribution with a mean of 114±32 N/mm. Unsintered NiGD and sintered LIP-only control samples, plated onto similar chemically-etched silicon surfaces, were insufficiently adherent to be measured. The poor adhesion of the unsintered NiGD contacts was attributed to the formation of voids and an oxide-rich interface layer during the galvanic displacement process. Although sintering at 350 °C appeared to reduce the thickness of the interfacial oxide and eliminate the voids, the oxide was not totally removed and contributed to the measured contact resistance of sintered NiGD-treated contacts of ~ 15 mΩ cm² being more than an order of magnitude higher than laser-ablated LIP-NiCu contacts on a similar n-type emitter surface. Adhesion priming layers formed using NiGD were used to metallise selective-emitter cells patterned using aerosol jet etching and having planarised contact regions. Although the LIP-NiCu contact grid was strongly adherent, the average area-normalised series resistance (Rs) was 0.96±0.18 Ω cm², the majority of which is attributed to high contact resistance arising from a residual interfacial oxide.
Conference Paper
Laser-induced periodic surface structures (LIPSS, ripples) are a universal phenomenon that can be observed on almost any material after the irradiation by linearly polarized laser beams, particularly when using ultrashort laser pulses with durations in the picosecond to femtosecond range. During the past few years significantly increasing research activities have been reported in the field of LIPSS, since their generation in a single-step process provides a simple way of nanostructuring and surface functionalization towards the control of optical, mechanical or chemical properties. In this contribution current applications of LIPSS are reviewed, including the colorization of technical surfaces, the control of surface wetting, the tailoring of surface colonization by bacterial biofilms, and the improvement of the tribological performance of nanostructured metal surfaces.
Article
Historically, busbar pull tests have been used as a measure of metal-silicon adhesion for silicon solar cells; however, such measurements cannot be easily applied to evaluate finger adhesion and the propensity of metal fingers to peel. Finger adhesion will be increasingly important as the width of fingers decrease and busbars are effectively removed from the cell metallization. In this paper, we correlate metal-plated finger dislodgement measurements, which have been obtained using a stylus-based metallization testing tool, and busbar pull test forces with nanoindentation measurements of the Young's modulus in order to determine key determinants of strong finger adhesion. It is proposed that metal fingers with a higher Young's modulus dislodge at lower stylus impact forces because the energy associated with the impact is less easily dissipated along the fingers and consequently remains more focused on the impact location, causing not only finger dislodgement but more extensive finger peeling as well. It is shown how plating rate, chemistry, grid geometry, and postplating annealing can all contribute to plated metal finger adhesion, therefore necessitating an understanding of these factors for reliable plated metallization.
Article
An approach for n-type passivated emitter and rear locally diffused (PERL) solar cells is presented that combines the PassDop technology with Ni- and Cu-plated front contacts. As a step toward an eased industrial implementation, a new PassDop layer-based on a-SiNx:P is presented and compared with the original layer based on a-SiCx:P We show that the PassDop concept can be used as a rear-side approach for these solar cells that reach energy-conversion efficiencies up to 23.5% on a small-area (2 × 2 cm2) when using the layer based on a-SiCx:P, and 22.8% when using the layer based on a-SiNx:P. By applying the same technology on a larger scale, we achieved efficiencies that exceed 22% on 143.2 cm2. We show that Ni-plating on boron emitters allows for excellent contact properties. With laser contact opening in conjunction with Ni- and Cu-plating, similar results for photolithographic opening (TiPdAg seed layer and Ag-plating) can be reached. Combining these technologies, cells are presented using PassDop as the rear-side approach and Ni-plating to contact the boron emitter with first results to achieve up to 21% cell efficiency on 148.6 cm2.
Article
In this work we present large area p-type PERL solar cells featuring local p+ Back-Surface-Field (BSF) obtained by pico-second (ps) laser processing of thin ALD Al2O3 layers capped either by PECVD SiNx or PECVD SiOx. In this specific approach, the laser processing is performed through the rear passivation stack, whereby the laser simultaneously patterns the dielectrics for the subsequent Al–Si contact formation, whilst incorporating the Al atoms from the Al2O3 film into the underlying Si bulk. With this technique, we register substantial dopant incorporation for 10 nm of ALD Al2O3, without using additional doping sources. When this process is coupled to a front metallization scheme based on Cu plating, p-PERL cells can be fabricated adopting only low temperature process steps, which avoids thermal stability issues related to the Al2O3 layer, such as crystallization and blistering, and improves the properties of the rear reflector by avoiding Al–Si alloying. Efficiencies up to 20.7% are reported on 6-in. Cz–Si solar cells featuring local p+ BSF formed by laser doping through a rear stack composed of 10 nm ALD Al2O3 and 120 nm PECVD SiNx.
Article
In this paper, a two-layer metallization for silicon based solar cells is presented. The metallization consists of thin nickel barrier and thick copper conductive layers, both obtained by electrodeposition technique suitable for phosphorus-doped 70–90 Ω/sq solar cell emitter formed on p-type silicon substrate. To ensure the adhesion between metal contact and emitter a very thin layer of mesoporous silicon is introduced on the emitter surface before metal deposition. This approach allows metal anchoring inside pores and improves silicon–nickel interface uniformity. Optimization of metal contact parameters is achieved varying the anodization and electrodeposition conditions. Characterization of contacts between metal and emitter is carried out by scanning electron microscopy, specific contact resistance and current–voltage measurements. Mechanical strength of nickel–copper contacts is evaluated by the peel test. Adhesion strength of more than 4.5 N/mm and contact resistance of 350 μΩ cm2 on 80 Ω/sq emitter are achieved.
Article
Continuous wave laser processes used to create solar cells with selective emitter and plated Ni-Cu front contacts are a widely discussed topic. Particularly low adhesion of the front metal contact has been identified as a serious challenge. In this work a detailed surface characterization of laser doped and patterned front sides of solar cells shows that formation of silicon oxynitride hinders nickel silicide formation and reduces contact adhesion of Ni-Cu plated contacts. In order to overcome the observed tradeoff between metal contact adhesion and penetration of the pn-junction, this paper presents a novel process sequence based on the formation of a deep selective emitter using a green continuous waver laser and subsequent patterning of the dielectric using a nanosecond-pulsed laser.
Article
A novel atmospheric pressure dry texture process is investigated in order to create nanostructures at the c-Si surface. The texture process uses diluted molecular fluorine (F2) as the process gas. F2 is partially dissociated at an elevated temperature before it is delivered to the c-Si wafer. Thermal activation of fluorine occurs on Si wafer surface in a dissociative chemisorption process leading to the removal of Si in the form of volatile SiFx species. The etching process can be controlled to form nanostructures with different aspect ratios and surface reflection values. In this work, we dry textured multicrystalline (mc) Si wafers to reach weighted surface reflection ∼12% in the wavelength range of 250–1200 nm. Nanotextured mc Si wafers were used to prepare p-type Al-BSF solar cells. The fabricated nanostructured cells show a gain in short circuit current (Jsc) of ∼0.5 mA/cm2 and reached a conversion efficiency of 17.3%.
Conference Paper
Good metal to silicon adhesion is essential in achieving high fill factor and low series resistance in high-efficiency solar cells, and electroless plating has provided a cost-effective method of metallising the grooves of these cells. Unfortunately the smoothness and shallowness of grooves created using inkjet printing can limit the mechanical adhesion of the deposited metal. In this paper we examine the importance of roughness in achieving a successful adhesion between electroless-deposited nickel and the silicon surface in inkjet printed grooves. We investigate several low HF or HF-free chemical solutions commonly used for etching the surface of inkjet-printed grooves with minimum impact on the sensitive dielectric layer. We conclude by presenting a roughening technique suitable for inkjet printed solar cell fabrication.
Article
The adhesion of Ni–Cu-plated contacts requires annealing, which can lead to an electrical degradation of the solar cell. A combination of optical imaging methods and electron microscopical cross-section analysis was used to investigate the annealing-induced shunts on monocrystalline solar cells. The results show that Ni spikes cause the lowering of the pseudo fill factor and reveal that these nonlinear shunts show the same behavior as recombination active breakdown sites on multicrystalline silicon solar cells, including radiation of visible light while breakdown. Using reverse biased electroluminescence set-up the shunts could be localized in top view with µm size lateral resolution. Subsequently, a cross-section was prepared on a radiative breakdown spot. Advanced electron microscopic investigations reveal defect structures featuring nickel precipitates on the position of the light source. (© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)
Article
Local mechanical stress is currently an important topic of concern in microelectronics processing. A technique that has become increasingly popular for local mechanical stress measurements is micro-Raman spectroscopy. In this paper, the theoretical background of Raman spectroscopy, with special attention to its sensitivity for mechanical stress, is discussed, and practical information is given for the application of this technique to stress measurements in silicon integrated circuits. An overview is given of some important applications of the technique, illustrated with examples from the literature: the first studies of the influence of external stress on the Si Raman modes are reviewed; the application of this technique to measure stress in silicon-on-insulator films is discussed; results of measurements of local stress in isolation structures and trenches are reviewed; and the use of micro-Raman spectroscopy to obtain more information on stress in metals, by measuring the stress in the surrounding Si substrate is explained.
  • A Ciesla
  • R Chen
  • S Wang
  • J Ji
  • Z Shi
  • L Mai
  • C Chan
  • B Hallam
  • C Chong
  • S Wenham
  • M Green
A. Ciesla, R. Chen, S. Wang, J. Ji, Z. Shi, L. Mai, C. Chan, B. Hallam, C. Chong, S. Wenham, M. Green, Prog. Photovolt. Res. Appl. 2018, 1, 43.
  • A A Brand
  • F Meyer
  • J.-F Nekarda
  • R Preu
A. A. Brand, F. Meyer, J.-F. Nekarda, R. Preu, Appl. Phys. A 2014, 117, 237.
  • B Steinhauser
  • M Kamp
  • A A Brand
  • U Jager
  • J Bartsch
  • J Benick
  • M Hermle
B. Steinhauser, M. Kamp, A. A. Brand, U. Jager, J. Bartsch, J. Benick, M. Hermle, IEEE J. Photovolt. 2016, 6, 419.
  • E Cornagliotti
  • A Uruena
  • B Hallam
  • L Tous
  • R Russell
  • F Duerinckx
  • J Szlufcik
E. Cornagliotti, A. Uruena, B. Hallam, L. Tous, R. Russell, F. Duerinckx, J. Szlufcik, Sol. Energy Mater. Sol. Cells 2015, 138, 72.
  • C Geisler
  • W Hördt
  • S Kluska
  • A Mondon
  • S Hopman
  • M Glatthaar
C. Geisler, W. Hördt, S. Kluska, A. Mondon, S. Hopman, M. Glatthaar, Sol. Energy Mater. Sol. Cells 2015, 133, 48.
Verfahren zur Metallisierung von Substraten
  • D L Notarp
  • M Becker
D. L. Notarp, M. Becker, "Verfahren zur Metallisierung von Substraten," DE102007005161B4.
  • K Kholostov
  • L Serenelli
  • M Izzi
  • M Tucci
  • M Balucani
K. Kholostov, L. Serenelli, M. Izzi, M. Tucci, M. Balucani, Mater. Sci. Eng.: B 2015, 194, 78.
  • J Rodriguez
  • W Zhang
  • S Lim
  • A Lennon
J. Rodriguez, W. Zhang, S. Lim, A. Lennon, Sol. Energy Mater. Sol. Cells 2017, 165, 17.
  • A Büchler
  • S Kluska
  • M Kasemann
  • M Breitwieser
  • W Kwapil
  • A Hähnel
  • H Blumtritt
  • S Hopman
  • M Glatthaar
A. Büchler, S. Kluska, M. Kasemann, M. Breitwieser, W. Kwapil, A. Hähnel, H. Blumtritt, S. Hopman, M. Glatthaar, Phys. Status Solidi RRL 2014, 8, 385.
  • S Wang
  • L Mai
  • A Wenham
  • Z Hameiri
  • D Payne
  • C Chan
  • B Hallam
  • A Sugianto
  • C M Chong
  • J Ji
  • Z Shi
  • S Wenham
S. Wang, L. Mai, A. Wenham, Z. Hameiri, D. Payne, C. Chan, B. Hallam, A. Sugianto, C. M. Chong, J. Ji, Z. Shi, S. Wenham, Sol. Energy Mater. Sol. Cells 2017, 169, 151.
  • B Kafle
  • J Seiffe
  • M Hofmann
  • L Clochard
  • E Duffy
  • J Rentsch
B. Kafle, J. Seiffe, M. Hofmann, L. Clochard, E. Duffy, J. Rentsch, Phys. Status Solidi A 2015, 212, 307.
  • I. De Wolf
I. De Wolf, Semicond. Sci. Technol 1996, 11, 139.
  • X Wang
  • P.-C Hsiao
  • W Zhang
  • B Johnston
  • A Stokes
  • Q Wei
  • A Fell
  • S Surve
  • Y Shengzhao
  • P Verlinden
  • A Lennon
X. Wang, P.-C. Hsiao, W. Zhang, B. Johnston, A. Stokes, Q. Wei, A. Fell, S. Surve, Y. Shengzhao, P. Verlinden, A. Lennon, IEEE J. Photovolt. 2016, 6, 1167.