Fabrication and processing of polymer solar cells: A review of printing and coating techniques

Abstract

Polymer solar cells are reviewed in the context of the processing techniques leading to complete devices. A distinction is made between the film-forming techniques that are used currently such as spincoating, doctor blading and casting and the, from a processing point of view, more desirable film-forming techniques such as slot-die coating, gravure coating, knife-over-edge coating, off-set coating, spray coating and printing techniques such as ink jet printing, pad printing and screen printing. The former are used almost exclusively and are not suited for high-volume production whereas the latter are highly suited, but little explored in the context of polymer solar cells. A further distinction is made between printing and coating when a film is formed. The entire process leading to polymer solar cells is broken down into the individual steps and the available techniques and materials for each step are described with focus on the particular advantages and disadvantages associated with each case.

Introduction

Why polymer solar cells? As a technology polymer solar cells are unrivalled in terms of processing cost, processing speed, processing simplicity and thermal budget. It is the only photovoltaic technology that potentially offers a convincing solution to the problem of a high cost commonly encountered for photovoltaic technologies. There are, however, unsolved problems of low power conversion efficiency, poor operational stability, materials cost and environmental impact.

The field of polymer and organic solar cells has been the subject of reviews [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], special issues [21], [22], [23], [24], [25], [26], [27] and books [28], [29], [30], [31], [32] on several occasions during the past 5 years and the definitions of polymer solar cells are quite broad spanning all-polymer solar cells, polymer fullerene solar cells, small molecule and hybrid solar cells. By far the most successful of them are the polymer–fullerene solar cells that comprise a mixture of the polymer, which exclusively is the donor material and typically a soluble fullerene derivative as the acceptor material. While research on polymer and organic solar cells date back to the 1980s [2] the first example of a polymer solar cell with a convincing underlying understanding of the physics and chemistry is the bilayer heterojunction between the soluble polymer 2-methoxy-5-(2-ethylhexyloxy)-polyphenylenevinylene (MEHPPV) and the Buckminsterfullerene C₆₀ [33], [34] where a power conversion efficiency of 0.04% was obtained using monochromatic light. It is reasonable to define this report as the prototypical example of a polymer solar cell from which all modern polymer solar cells stem. The next convincing step was the application of a dispersed bulk heterojunction of MEHPPV and C₆₀ [35] and later soluble derivatives of C₆₀ [36], [37], [38], which increased the power conversion efficiency to 2.5%. The third step that represents the state of the art today is the exemplification of the interplay between the morphology reached by processing of the active layer and the function of the device which allowed efficiencies of up to around 5% to be reached for mixtures of poly-3-hexylthiophene (P3HT) and phenyl-C61-butyric acid methyl ester ([60]PCBM) [39], [40]. A lot of original research has detailed methods for improving the function through physics and chemistry. The most important physical means for improving performance are the use of a thin layer of an insulator between the active layer and the low work function metal electrode [41] and more recently also in inverted devices [42], the use of optical spacers [43], [44], [45], the understanding of how the open circuit voltage is obtained [46] and by these means deriving an efficient method for predicting performance of materials combinations based on measurable materials properties [47] and finally the use of various mixing and annealing methods to control the morphology of the active layer [48]. It should be mentioned that the latter methods are not generic and generally should be viewed as materials specific. Thermal annealing is a good example that will work well for some material combinations, but not others. Sometimes even the same material gives different results as in the case of P3HT where regioregularity and molecular weight plays a crucial role [17]. In terms of chemistry work has focussed on improving the properties of both donor and acceptor material. For the polymers (the donor materials), which in many of the reported devices are responsible for the light harvesting there has been special focus on obtaining a low band gap such that as much of the available energy from the Sun is harvested [5], [11], [18]. Many materials with low band gap have been prepared and while band gaps as low as 0.5 eV has been prepared it is remarkably few of the low band gap polymers that give devices that rival those obtained from materials with a larger band gap. The most successful fullerene (the acceptor material) is the derivative known as [60]PCBM [36]. In spite of intensive research for better fullerene derivatives (which would be a review topic of its own) there has been no derivatives being as performing as [60]PCBM except perhaps for the more difficultly accessible [70]PCBM [49] and the recent Bis[60]PCBM [50].

Most improvements have been found through materials such as P3HT and the fullerenes [60]PCBM and [70]PCBM that represent the state–of-the-art well from a materials point of view. In terms of device structure efforts have been relatively limited for various reasons. Firstly, there is a considerable drive towards achieving high power conversion efficiency and this implies small devices and a high conductivity of the electrodes. Since the back electrode for “hero” cells is typically an evaporated metal electrode the limiting electrode in terms of conductivity is the necessarily transparent front electrode. The best performing transparent electrode material that combines high transparency over a broad range of wavelengths and high electrical conductivity is indium tin oxide (ITO). From the point of view of processing polymer solar cells on a large scale there are several impossibilities connected to the use of ITO but for the purpose of breaking power conversion efficiency records there are few. The reason that nearly all reported devices with any significant power conversion efficiency are prepared on ITO covered glass substrates is understandable. This has severely limited the evolution of both device geometry and processing techniques.

The most significant step is undoubtedly the stacking of both the high and the low band gap materials into a tandem solar cell with the highest reported power conversion efficiency of 6.5% [51]. In Fig. 1 an example of a tandem solar cell structure is shown along with examples of high and low band gap materials [52]. The tandem solar cell is a pleasing concept that itself has been subject to some evolution due to the complexity of its realisation. Since it is a multilayer structure where ideally all layers are processed without affecting underlying layers during processing this has taken some time and should by no means be considered completely evolved. Initially, the problems associated with multilayer solution processing was solved by processing small molecules entirely from vacuum [53], [54] or by the use of solution processing for the first device in the stack followed by vacuum processing of the second device in the stack [55]. More recently, a careful choice of solvents allowed for solution processing of the entire stack when employing soluble polymer materials [56], [57]. The most recent developments are the use of a folded reflective geometry, which offers simplicity in device processing but complexity in module assembly [58] and finally the use of thermocleavable materials that efficiently solves the problems associated with solubility of the underlying layers during processing. [52] While both of the latter examples have given rise to inferior performance they do prove the point that technical solutions are possible and that they should be sought. Another reason that has limited the evolution of device geometries is in part a result of the use of ITO in conjunction with the hole transporting layer poly-(ethylenedioxythiophene):polystyrenesulphonic acid (PEDOT:PSS), which implies the use of a back electrode with a low work function such as aluminium or calcium. The use of these metals necessitates thermal evaporation (physical vapour deposition). It is thus reasonable to assume that this highly limited choice in device geometry have had some influence on both the choices made when engineering new materials and in the observed state-of-the-art. In essence all the new materials that have been prepared have been prepared for electrode combinations such as ITO-PEDOT:PSS and aluminium (or calcium) and it is interesting to speculate what the state-of-the-art would have looked like if research into the use of different electrode materials had been more popular.

When the selective pressure on research is on one particular aspect such as the power conversion efficiency, other equally important areas of research may be neglected while they share the same importance when it comes to realizing the objective of stable, efficient, low-cost polymer solar cells prepared by very fast processing techniques. This can be summarized in a Venn diagram as the unification challenge, which has been discussed earlier [16], [31] (Fig. 2). A material that gives rise to highly efficient devices is thus of little consequence if its operation is not stable or if the process leading to the final device is difficult. There has been a recent interest in the operational stability of devices [16], [59], [60], [61], [62], [63], [64] and more importantly on the understanding of why devices and materials break down [16], [65], [66], [67], [68], [69], [70], [71], [72], [73], [74]. There is hope that new materials can be derived that combines both efficiency and stability. The interest in processing of polymer solar cells has been limited to methodologies for improving power conversion efficiency and it is remarkable that there has been no interest in actually demonstrating the alleged significance of large-scale processing of polymer solar cells. After all, if large-scale processing proves to be problematic or impossible the concept of polymer solar cells would become a scientific curiosity with little consequence to mankind. The view held in this review is thus that the problems associated with industrial and large-scale processing must be addressed and taken into account alongside considerations of operational stability and power conversion efficiency before polymer solar cells can be attributed any importance (i.e. that they generate more power than is consumed during their preparation).

As highlighted above polymer solar cells typically comprise a multilayer structure where each layer in the stack may be formed by an individual film-forming technique. This should be viewed as a particular advantage of the technology for several reasons. Firstly, it is very versatile that one in principle can arrive at the same device geometry (i.e. layer sequence and materials combination) through many different routes. Secondly, it is of some importance from an intellectual property rights (IPR) point of view as this implies that it will be almost impossible to efficiently protect a polymer solar cell product unless it is materials specific. Since both the available device geometries and many materials are prior art there is little hope that one can protect ideas through a particular process sequence and even if possible it would be very easy to circumvent. A further complication lies in the fact that it is not always possible with a printed film at hand to establish how it was made. While some printing methods give rise to characteristics in the printed film that does allow for their identification it can be anticipated to be very difficult for multilayer films. This should be viewed as strengths of the technology as it increases competition and places focus of the competition on what matters, namely overall performance, and places the judgement in the hand of the consumer, where it should be, rather than Olympic records in scientific articles. Thirdly, this aspect makes it desirable to invent new device structures and electrodes as this would become valuable IPR in the event that a competitive device could be prepared.

It is noticeable that there are so many known, well developed and currently explored film-forming techniques, and yet so few of them have made their way into the world of polymer solar cells. There are at least three reasons that can account for this. Firstly, many of the techniques require large amounts of material, and secondly, reproducibility is sometimes difficult, and thirdly many techniques are unsuited for the small scale commonly employed in laboratory trials. The techniques that have been explored are generally well suited for individual processing of small substrates (i.e. spin coating, doctor blading and casting). There are, however, many more film-forming techniques available that have been developed for high volume processing of paper, plastic and textile materials where the substrate is in the form of a continuous roll of material. This is often called roll-to-roll coating or reel-to-reel coating (abbreviated R2R coating) and the processing equipment generally comprises: unwinding, coating and rewinding of the material. Many more processes may be involved such as cleaning, pre/post treatments of the fabric, heating, drying, etc. The importance of the coating techniques is that they are suited for high-speed coating or printing and they have been developed with the aim of achieving a very low process cost. While this has not been shown repeatedly in the context of polymer solar cells it is generally accepted that R2R processing is cheap and fast. There has been one report [75] that details large-scale processing of polymer solar cells which in part was based on R2R processing and that study confirmed that processing by coating/printing is of low cost which is not surprising. The study also showed that while polymer solar cells have the potential to fulfil the expectations of low-cost processing this will not come easily and a dedicated effort will be required. One of the challenges currently is thus to identify the ideal coating techniques for polymer solar cells. Some techniques are suited only for coating an even layer over the entire substrate surface thus giving 0-dimensional control (i.e. no ability to create a pattern), other techniques allow for 1-dimensional control of the pattern and in the context of a substrate passing a coating head this implies that a striped pattern can be created. Further, some methods allow for full 2-dimensional control of the printed pattern where any shape can be reproduced on the substrates. Finally, some coating methods allow for the formation of multilayer films in the same coating step. Combinations of these techniques can in principle enable pseudo 3-dimensional control during printing (i.e. both pattern and multilayer structure). A second challenge is that with each of these coating techniques there are often a window of operation in terms of processing speed and achievable wet layer thickness that all hinges on the properties of the ink and the interplay between the surface that is to be coated. Viscosity, surface tension, surface energy, volatility are a few of the important ink properties that have to be taken into account when procuring the ink. It should be noted that it is customary in the coating industry to employ additives and adjuvants that are added to the ink to adjust the properties such that printability/coatability of the ink is achieved. This is all very well for traditional coating and printing where the transfer of a pattern is all that is required. In the case of polymer solar cells the coated pattern has to be functional after the film is formed. It should thus be anticipated that there is limited freedom when choosing additives to aid printing/coating.

The ideal process should involve solution processing of all layers on flexible substrates by the combination of as few coating and printing steps as possible. The process should be free from costly indium, toxic solvents and chemicals and the final polymer solar cell product should have a low environmental impact and a high degree of recyclability.

The main purpose of this work is to guide the reader through some of the available coating techniques and allowing for a judgement as to the suitability of a given technique in the context of polymer solar cells. It is especially relevant for people involved in upscaling, preproduction and production of polymer solar cells and it is my hope that it will serve as a motivation for addressing the unification challenge and that it will bring attention to the hitherto little explored area of processing polymer solar cells on an industrial scale and enabling their production.