Optical coatings for automotive applications: a case study in translating fundamental materials science into commercial reality

Technology-oriented organizations and businesses generally adopt one of two strategies when it comes to the product they wish to sell in the market place: technology push or market pull [1]. Technology-based companies, especially those dealing in the business-to-business arena, are traditionally more likely to adopt a technology push approach, whereas those dealing in the business-to-consumer sector will focus more on the end-user and hence be influenced by market pull factors [2]. Regardless of the strategy utilized, both approaches require a level of scientific/engineering endeavour to translate an idea or perceived need into a final product or service; be it incremental developmental or a new design paradigm. Most small to medium enterprises (SMEs) and even some multi-nationals do not have the critical mass, commercial wealth and technical infrastructure to undertake research that may be prolonged, expensive, and importantly has no guarantee of success attached at the end. Such a situation opens the door for externally funded research bodies to partner with industry in an effort to mitigate the risk side of the ledger in the risk versus reward equation. At first glance, universities appear to have the prerequisites needed as they are able to provide scientific knowledge, technical expertise and skilled personnel. Such assets are the basic ingredients needed to generate new intellectual property (IP [3]). Rightly or wrongly, however, many traditional academic institutions have adhered to a hierarchal mindset where pure or fundamental research is valued above applied research, which is valued above engineering/product-oriented research. An institutional value-system such as this introduces significant road blocks into effectively engaging with industry needs. It produces a situation that is at odds with the vast majority of industrial research, where the focus is on applied/engineering solutions in which the problem solving process is carried out with a view to achieving a specific outcome. That outcome may not necessarily involve understanding why the phenomenon behaves the way that it does, all that is important is that the solution works. This targeted research process is in stark contrast to fundamental research that is akin to adopting a blue-sky [4] or curiosity-driven approach in which all the nuances of the process need to be fully investigated. The research value-system mismatch has in the past resulted in a rather poor translation of fundamental academic work making its way into commercial success, and is a contributing factor as to why industry has tended to shy away from such interactions. If one takes a different philosophical approach to research in which no such hierarchical value-system exists, be it at the institutional level or just as importantly at the individual researcher level, a more productive engagement process between industry and academia may ensue. In short both parties need to fully appreciate the operating constraints that the other partner works in and to allow for this from the outset.

Even if there is broad acceptance at the institutional level that industry engagement and the ensuing flow on effects of technology transfer are strategies worth adopting, academics may be loath to pursue such activities when being judged against performance metrics that favour traditional fundamental research. Furthermore, Thursby et al [5, 6] have put forward three additional reasons as to why such an engagement may not reach fruition when acted upon by research groups (or individual researchers). First, individuals involved in fundamental research may have an unwillingness to spend the often considerable time needed on the applied and engineering tasks being requested by the industry partner. Second, the researcher believes that research associated with a commercial activity is incongruent with fundamental research. Third, researchers may not disclose results to industry partners because they are unwilling to risk publication delays, which leads to the so-called academic 'publish or perish' dilemma [7]. All three reasons will dilute any meaningful collaboration between academia and industry. If, however, the group's research leader has an innate scientific entrepreneurial outlook and imprints this onto the research team, participation in the knowledge transfer process is likely to be high [8]. Therefore, rather than having a mindset in which applied/engineering research is largely viewed as being of 'lesser value' and devoid of fundamental questioning, it should be seen as having the potential to raise a manifold of basic questions. Unfortunately, studies reveal that the profile of academic/industry interactions is highly skewed, with relatively few academic researchers being responsible for the vast majority of these interactions [9–11]. A means of improving this current state of play is to construct a scientific environment in which fundamental, applied and engineering research can run concurrently as each has the potential to feed off the other. Adopting a concurrent fundamental/applied/engineering construct to academic research has the potential of bridging the so-called 'valley of death [12, 13]' between the end of a fundamental academic investigation and the beginning of the technology acceptance and scale-up processes, which is invariably required to bring a product to market. In short, one of the prerequisites for translating fundamental research into a commercial outcome is the need for a scientific entrepreneurial culture to exist at the institutional level, but even more importantly this culture must exist at the group leader/researcher level [14].

One of the largest manufacturing sectors that has benefitted from actively embracing plastic as its preferred cosmetic and engineering material in lieu of traditional (metal) materials is the automotive industry. Plastic now takes the place of metal, which, for decades, was the material-of-choice in many automotive components such as body panels, facia, brackets, door linings, lighting assemblies, etc. The amount of plastic in a modern large car can now exceed 200 kg, a 20 fold increase since the 1960s [15]. As a result of careful plastic material selection and adopting an integrated design and manufacture philosophy, many technical and cost advantages have been realized. Some of the more obvious ones are the elimination of corrosion, weight reduction and parts count, increased design flexibility and better impact resistance. Two applications which until recently had defied this materials transition, however, were general automotive large-area glazing and rear-view reflectors. On the upside, glass is a highly scratch resistant substrate, relatively cheap (in flat form), highly transparent, and for reflectors the glass substrate is easy to metallize. The downside is that glass is relatively heavy, complex shapes are possible but difficult and usually cost prohibitive and, glass is prone to shattering during an impact (i.e. vehicle collisions). By way of comparison, a plastic substrate, for example polycarbonate, is virtually shatter proof, can be injection moulded quickly with a high degree of accuracy and offers design freedoms not available with glass. Additionally, the injection moulding process can be used to integrate parts more effectively, which leads to a reduction in the parts count (i.e. the design of parts that have an integrated clipping system). These advantages make plastic an appealing alternative in the automotive industry, and this is fuelling the continued development and incorporation of plastic materials.

Where high transparency and durability are required, glass is still the dominant material of choice within the automotive sector (e.g. windscreens, side windows, rear-view reflectors, etc). That said, the automotive industry has successfully converted from using glass/metal in headlight and tail-light assemblies to now accepting the (almost exclusive) use of plastic [16]. This transition has freed up the design constraints imposed by glass and allowed the seamless integration of headlights and tail-lights into the overall styling of modern cars (see figure 1). The adoption of this alternate material has been driven by the need for reduced cost, increased safety, enhanced functionality, reduced weight, and increased fuel efficiency [16, 17]. A critical technological hurdle in the use of plastic components for automotive optical applications was overcome with the development of abrasion-resistant hardcoats that produced components with 'glass-like' properties [18]. These coatings were introduced into the ophthalmic industry [19] in the 1970s, and by the 1980s the same technology had slowly started to establish itself in the automotive industry [20]. The two critical issues addressed by using a suitable hardcoat were that of reducing UV associated degradation and increasing the wear characteristics of the optical parts.

Zoom In Zoom Out Reset image size

This investigation outlines a scientific case study that is now a mature product in the market place but one that undertook a fundamental investigation, applied development, and rigorous engineering and product validation. It is the development and production of the world's first original equipment manufacturer (OEM) plastic automotive reflector. With respect to the OEM plastic reflector, a nano-engineered first surface reflector deposited onto a polycarbonate substrate with performance that matches that of first surface glass reflectors is reported. Herein, the contractual issues associated with the ownership of intellectual property (IP) and scientific publications, as well as the fundamental science and product development of the plastic reflector, are outlined. Additionally, the process involved in translating that fundamental research from lab-based results into a final product is also discussed. Whilst one cannot say that the contractual and problem solving approach taken herein was unique, the project itself was atypical when compared to the modus operandi of many academic investigations in which scientific progress beyond the fundamental stages is never formally incorporated into the project scope. An industry-focused target-driven problem solving path was employed in which dedicated academic material scientists (with prior industrial experience) and industrial process engineers were assigned to the project from inception. By employing such an approach the so-called 'valley of death' scenario between the end of fundamental science and the commencement of an applied investigation and product engineering scale-up was largely overcome with a greatly reduced time and dollar outlay.

Project commencement was the result of a three way agreement between the Automotive Australia 2020 Cooperative Research Centre (AA2020CRC), the University of South Australia—Thin Film Coatings Group—and an Australian automotive component manufacturer. In fact, referencing the third party as a generic Australian component manufacturer (which was at the request of the manufacturer) provides some insight into the level of on-going negotiations that invariably followed post-contractual signing. Even after contractual arrangements had been finalized and the basic project framework implemented, maintaining a high level of mutual confidence (and respect) between all parties was an essential prerequisite for the successful outcomes that followed.

Ownership of IP generated within this project was equally owned between the University of South Australia and the Australian automotive component manufacturer, with the CRC acting as the administrative arm for the Commonwealth of Australia. Project terms and conditions were established early in the piece and well-finalized prior to any contract signing. The cost associated with any IP arising from the project that proceeded to patent filing was borne by the party wishing to pursue it. A field of use clause was incorporated into the contract in which royalty-free arrangements existed. Any IP ultimately used within the automotive sector would be royalty-free for the automotive component partner, with the university able to pursue royalty-free use outside the automotive sector. With respect to public disclosure by means of journal publications etc, this needed the approval of all three parties prior to submission. At times, this resulted in publications being delayed due to commercially sensitive trigger points not having been reached, or journal content having to be re-profiled to keep all the parties satisfied. Regardless of the existing contract set in place at the commencement of the project, the take home message is that a good working relationship needs to exist from the outset and that this must be maintained throughout the life of the project. This involved continual on-going negotiations at general group meetings, technical meetings etc, so that each party could present their concerns or requests. This format helped to alleviate any misunderstandings or unrealistic demands that could have created ill-will and quickly eroded confidence regardless of any contract that was set in place.

Within the scope of the project, moving to a plastic instead of glass-based material would help overcome certain engineering limitations in the design of the reflector and associated parts, as well as providing the industry partner with a technical point of differentiation in a competitive market place. The durability of the reflector (also applicable to large-area glazing), however, needed to be significantly improved to meet the strict performance criteria that applies in the automotive sector. With respect to the replacement of reflector substrates from glass to plastic, recent attempts have been made to develop a coated plastic automotive reflector by Venture-source, VTEC Mirror Systems LLC, Ficocipa SAS, Specchidea s.r.l. and Saudi Basic Industries Corporation (SABIC [21]). However, none have been taken up by OEMs with durability, and cosmetic issues are cited as the reason preventing their widespread use [21]. Currently, manufacturers offer both first and second surface glass reflectors. A first surface reflector has the reflective surface on the front of the glass substrate, while a second surface reflector has the reflective surface on the back with the first optical surface being the glass substrate itself. From an optical performance viewpoint, the first surface option is superior in performance, as the second surface reflector can produce double-imaging, which emanates from the first optical surface (glass substrate) and the second reflector surface (metallic back-coat). At night this double-imaging can be especially distracting to the driver [22], leading to possible delays in information processing similar to other well documented cognitive distractions such as multi-tasking operations and mobile phone use [23, 24]. However, when considering the lifetime performance of these reflectors in durability tests, the first surface reflector has traditionally been inferior to that of the second surface reflector. Locating the metal layer on the outermost surface leaves it exposed to the surrounding environment and makes it susceptible to atmospheric degradation, and most notably makes it susceptible to fine scratches (due to washing/wiping/cleaning). The performance of both first and second surface glass-based reflectors has become accepted as the benchmark by which alternate materials are now judged against. To realize the best optical properties and capitalize on the design of parts with an integrated clipping system, a first surface plastic reflector was decided upon as the most logical design target.

Conventional coating approaches have metal layers deposited directly onto the substrate and then, if needed, over-coated with a transparent protective layer. This over-coat provides protection against UV weathering, corrosive agents, abrasion damage etc. For polycarbonate substrates many hard-coats are formulated so that they (partially) solvate the substrate, and this action helps produce a high adhesion interface [25]. When metal layers are, however, deposited directly onto polycarbonate substrates, a previous study has found that the hard-coat adhesion mechanism can be disrupted [21]. Any disruption to this adhesion mechanism may result in delamination. Additionally, thin optically transparent protective layers positioned on top of reflective surfaces are prone to producing interference fringes if there is a refractive index mismatch between the two layers. From a consumer view point such a product is cosmetically unacceptable. To avoid such issues the protective over-coat can be made thick (10–20 microns) so as to shift the optical interference away from the visible range [20]. Doing so, however, renders the coating susceptible to crazing and can impart a low level of scratch resistance.

To overcome these issues the order of the reflector stack was reassessed and, based on conventional ophthalmic industry design, the hard-coat was placed directly onto the polycarbonate substrate and an additional SiO₂ layer incorporated [26]. While placing the protective hard-coat directly onto the polycarbonate substrate appears to be a dramatic departure from accepted wisdom, car headlight covers have successfully utilized this simple approach [27]. By adopting this method the delamination issues were resolved and the additional SiO₂ layer formed a 'glass-like' coating with enhanced abrasion resistance. This redesigned stack enabled the protective coating to pass all of the relevant automotive durability and performance tests. However, this meant that the reflective metallic layer was deposited on top of the hard-coat, effectively making it the 'first surface' in the stack. As such this metallic layer had to be re-designed so that it was inherently scratch and corrosion resistant. Figure 2 shows the evolution of the stack from that first proposed (i.e. polycarbonate/chrome/hardcoat; PC/Cr/HC), to the last iteration which was implemented during the industrial equipment commissioning phase (i.e. polycarbonate/hardcoat/SiO₂/chrome/twin-capping layers [28]).

Zoom In Zoom Out Reset image size

An industry examination of current automotive first-surface glass reflectors was undertaken and revealed that chrome was almost universally the reflective material of choice. Intuitively this makes sense as chrome is a relatively hard metal that self-passivates by forming a thin CrO layer, making it inherently corrosion resistant [29]. Equally importantly the colour and level of reflectivity (industry requirement > 40%) offered by chrome are widely accepted within the automotive industry. On this basis chrome was selected as the reflective material and incorporated into the plastic reflector stack design from inception (see figure 2). This design initially 'ticked all the boxes' with respect to eradicating failures due to delamination and eliminating cosmetically unacceptable optical fringes due to thin film interference. Additionally, placing the chrome layer on top all but eliminated any UV light penetration into the lower layers and enhanced the UV resistance of the reflector. Thus by moving away from a traditional glass substrate to a plastic-based design, flow on advantages were realized with respect to increased functionality and design freedoms, reduced parts count, simplified actuator motor mounting, and allowed the direct application of a metal backing layer for defrost functionality.

While the automotive industry has many strict objective test procedures and protocols in place, the importance of meeting subjective customer-focused demands during the course of a project cannot be discounted. This, however, is very rarely discussed in the scientific literature. Not meeting any of the industry relevant tests will result in an unsuccessful conclusion to a project, but just as important, failure to meet any subjective appraisal by someone like a CEO or R&D manager can also result in the demise of a project. Unfortunately, consideration as to whether or not a product is 'cosmetically'acceptable rarely enters the thought processes of an academic charged with implementing a scientific investigation. First-hand knowledge gained in this project, however, dictated that the subjective decision-making process carried significant weighting and is as important as passing any of the strict automotive tests. As such it must be incorporated into the academics' thought processes as they deal with finding a scientific solution to the problem at hand. The two dominant opto-mechanical tests relevant to this project were the automotive industry corrosion and scratch resistant tests. Adding credence to the importance of needing to pass both subjective as well as objective tests, the reflector stack did not initially pass the scratch resistant testing. A scientific solution was found relatively early in the project life cycle but it rendered the cosmetics of the reflector with a yellow hue. Even though the yellowness index of the reflector was within the stipulated specification range, a simple statement by one of the key stakeholders to the effect: 'I don't like the look of the (yellow) reflector' meant that another technical solution needed to be found. The need to overcome this subjective 'fail' is discussed in more detail later.

Initial corrosion tests revealed that the stack was susceptible to salt spray, but interestingly it was not the upper chrome layer that failed but rather the sputtered SiO₂ layer which was delaminating from the hard-coat [30]. Measurements using a profilometer confirmed the failure was occurring at the hard-coat SiO₂ interface. In the presence of either acidic or alkaline conditions, SiO₂ is known to be susceptible to hydrolytic attack [31, 32]. This mechanism was deemed the most likely cause for the observed delamination as XPS measurements conducted on a number of plaques revealed no inconsistencies with the stoichiometric Si:O ratio during sputtering. As sputter deposition conditions are known to have an effect on the deposition rate [33] and structure of films [34, 35], a series of SiO₂ layers were deposited at low/high chamber pressures and low/high substrate temperatures. Direct evidence for changes in the surface structure by means of SEM imaging proved problematic as the insulating nature of such a surface is known to produce charging artifacts [36, 37]. Previous work [38] using AFM was able to correlate changes in the surface roughness of metallic depositions to changes in the structure of the sputter layers. Based on these findings, the same technique was utilized to examine the various SiO₂ layers generated in this study.

The sputter chamber conditions were altered using a simple high/low matrix for chamber pressure and substrate temperature and the horizontal red/green interface in figure 3 delineates a pass/fail result for the plaques during the salt spray corrosion test. To record a pass result the plaques needed to record a minimum of 288 h without any delamination being observed. Regardless of the substrate temperature samples sputtered under high pressure (i.e. 0.6 Pa), conditions were unable to meet the minimum 288 h requirement. The chamber was then run at low pressure (i.e. 0.2 Pa) and all samples achieved a pass result, albeit with the sample produced at a low temperature of 55 °C only just being able to meet the minimum industry requirement. The two samples sputtered at high temperatures (i.e. 110 and 130 °C) were able to record 1000 h of continuous salt spray testing without any degradation prior to the test being terminated. This result exceeded the minimum industry requirement by a factor of 3.5 (i.e. min. requirement is 288 h) and matched the performance of a first surface glass reflector which was used as the benchmark.

Zoom In Zoom Out Reset image size

To see if the pass/fail results could be correlated against surface topography and roughness, AFM imaging of the two most extreme samples was performed, namely the high pressure/low temperature (HP/LT) and low pressure/high temperature (LP/HT) samples (see table 1 and figure 4). Visual observation of the two samples at both 10 and 100 μm length scales revealed that the LP/HT sample contained smaller grains, which appeared to be closely packed. The smaller RMS roughness value and positive kurtosis (peaked/narrow RMS roughness distribution) for the LP/HT sample as opposed to the larger RMS roughness and negative kurtosis (flat/broad RMS roughness distribution) for the HP/LT sample corroborated the simple visual observation in figure 4 (refer to DeCarlo [39] for a complete description of kurtosis). From the results the following generalizations about the structure of the SiO₂ sputtered layers were made: HP/LT conditions produced a more open loosely packed porous structure resulting in a surface with relatively higher RMS roughness and a broader roughness distribution; LP/HT conditions produced a closer packed structure resulting in a surface with relatively lower RMS roughness and a narrow roughness distribution. From this one can extrapolate and propose that the density of the SiO₂ layer dictates the level of electrolyte (i.e. NaCl salt spray test) ingress and/or the surface's effective reactivity to corrosive attack. The HP/LT plaques with their open porous structure were susceptible to chemical damage, resulting in the delamination of the stack. Contrastingly the LP/HT plaque resisted ingress and/or its effective reactivity to corrosive attack was higher. For a more complete discussion regarding the corrosive results the reader is directed to reference [30].

Zoom In Zoom Out Reset image size

When subjected to mechanical scuffing, the first surface chromium layer did not possess sufficient abrasion resistance to pass the industry-standard testing. The addition of dopant materials (in this case by means of co-sputtering) has been shown [40] to enhance the abrasion resistance of chromium (alloys). The addition of dopant materials such as borides, carbides and nitrides alter the spatial distribution and packing density of the resulting chrome alloys and produce a hard material with enhanced resistance to mechanical wear [41–44]. For example nitriding (i.e. CrN) has long been used to improve the wear properties of cutting tools [45]. However, two specific problems arose with respect to the deposition of a chrome layer for the plastic reflectors discussed herein. Firstly, many alloying techniques are carried out at elevated temperatures, which makes the process incompatible with the plastic substrates used in this project. Secondly, many of the commonly used dopants that were initially tried, while producing acceptable wear/scuff results, rendered the finished reflective surface with a yellow/orange hue that was deemed cosmetically/commercially unacceptable by the industry partner. Failure to adequately address the first issue is self evident and warrants no further commentary. Failure to adequately address the second (non-technical) issue, however, is very rarely given due recognition by academics charged with finding a technical solution, yet it can easily result in the early termination of a project.

A suite of alternate dopants were investigated, consisting primarily of transition metals. Thin films of chromium alloy were deposited by magnetron co-sputtering onto the PC/HC/SiO₂ stack using a previously described method [46]. A rotating disk setup was used to create a multi-pass deposition of chrome followed by the dopant of choice, with each pass yielding sub-nanometre layers. Co-sputtered films of Cr doped with a secondary metal dopant (Zr, Mo, W, Ti, or Co) were trialled where the atomic percentage of the dopant was varied from 1% to 20%. Dopants similar to those used here have been shown to alter the structure of Cr alloy coatings by replacing Cr atoms within the body centred cubic (BCC) lattice structure, yielding a hexagonally close packed (HCP) arrangement [42, 47]. Changes in the inter-atomic spacing (or crystal lattice parameter) were measured using electron diffraction as a means of tracking the changes in the atomic structure of the material. This lattice parameter was correlated against a macroscopic abrasion enhancement factor (i.e. Bayer abrasion ratio test) for pure Cr and alloys consisting of Mo, W, Ti, Co and Zr. Variations in the atomic concentration and dopant metal resulted in changes to the lattice parameter with a concomitant change in the abrasion enhancement. Figure 5 shows the effect of dopant material with respect to the abrasion enhancement, and based on the results obtained, the CrZr alloy was selected for further testing. The Zr:Cr ratio was varied from ca. 0.03 to 0.4, producing atomic structures of BCC, BCC + HPC and BCC + amorphous. A maximum in material abrasion was achieved when the atomic structure was a combination of BCC + HCP (in the range ca. 0.03–1.0 Zr:Cr ratio).

Zoom In Zoom Out Reset image size

The optical thin-film stack was subjected to the relevant automotive industry performance tests, which are reported in table 2. Small-scale pilot production under controlled production process conditions relevant to the industry partner was initially carried out at the university. The implementation of this was required to prove the protocol process and product performance to a level appropriate for the automotive component sector. Further progress, which is commitment to a full production facility, was undertaken based on the successful completion of these initial tests. To benchmark the new plastic reflector, a 'control'glass-based product was tested at the same time. For all the relevant tests the plastic reflector at least matched the specifications recorded by the 'in-current-use' glass-based reflector and, unsurprisingly, it was able to record a superior impact resistance result. While the impact test resulted in crazing of the plastic reflector coating, the same test on the glass reflector resulted in reflector shattering. This test alone affords the plastic reflector a sizable marketing edge in a very competitive market place.

Moving from laboratory testing to industrial scale-up involved a significant overlap between the academic material scientists involved in the project, the process equipment engineers in configuring and commissioning the system, and the industrial engineers in producing a robust final product. A combined scale-up and technology transfer was first carried out with the supplier of the vacuum deposition equipment, Leybold Optics Germany, with both academic material scientists and industrial engineers that had been involved with the project since inception. This enabled an equipment specification to be created and a level of confidence to be established prior to proceeding with the process equipment build. An industry proven Leybold Optics PylonMET batch-machine normally used for the metallization of plastic headlight reflectors (in use for more than a decade) was chosen and modified to suit the process specific requirements that had been established during laboratory trials. To achieve stable high rate deposition of SiO₂ a dual-rotating cylindrical cathode sputter source was chosen. Such an arrangement is not commonly used and it was the first time Leybold had attempted this configuration. Issues relating to venting the targets after each run, the influence of degassing water from the plastic components and the target environment, were significant engineering tasks that needed addressing. Laboratory established process protocols were transferred to Leybold process engineers and this knowledge was incorporated into the final configuration batch-machine, which was commissioned in dedicated clean room facilities built in Australia by the industrial partner (shown in figure 6).

Zoom In Zoom Out Reset image size

For high-rate metal deposition, Leybold Optics has developed an inter-pole target (IPT) magnetron configuration which is standard on their aluminium metallizer. In this configuration the magnets (used for plasma confinement) are placed external to the target. Magnet cooling is simplified and target life-time is greatly increased, an important consideration for commercial usage. Additionally, as the target cathode wears, the effective magnetic field containment does not significantly change, which yields sustained high deposition rates over the life of the target. To fit this configuration a custom alloy CrZr sputter target was fabricated (i.e. Leybold Optics developed the IPT configuration and Plasmaterials developed the IPT sputter target), which is shown in figure 7. To establish the capability of the process equipment and the process itself, an established systems approach involving a pre-acceptance test (PAT) program and a final acceptance test (FAT) program was undertaken. The PAT and FAT programs were jointly established between UniSA material scientists and the industry-based engineers.

Zoom In Zoom Out Reset image size

The process commissioning began in Leybold Optic's factory located in Alzenau, Germany. When the process was transferred to the PylonMET there was a subsequent change in the performance of the plastic reflector, which had been established in preliminary benchmark testing. Whilst performance in the corrosion resistance test (i.e. salt spray) was actually enhanced, a concurrent drop in abrasion resistance was noted. As a result, collaborative development was undertaken between Leybold engineers, the academic material scientists and the industrial engineers. A thorough experimental evaluation of the process conditions was unable to raise the level of abrasion resistance to that obtained using the (scale-up) university system. Research during the early stages of the project had identified an increase in abrasion resistance when a thin (~5–10 nm) primary capping layer was added to the stack. This concept was transferred to the stack deposition protocol used by the PylonMET with the addition of a secondary ~10 nm thick plasma polymer capping layer. This additional layer resulted in a hydrophobic surface that enhanced the corrosion resistance of the primary capping layer and reduced surface friction. This solution was deemed necessary even though it increased the cycle time and complexity of the process. The combined effect of the primary/secondary capping layers increased the abrasion resistance well above the first surface glass reflector target.

Having passed, the PAT focus shifted to the production site where a class 10 000 cleanroom facility was commissioned to house the high volume dip coater, oven and PylonMET (see figure 6). As part of this commissioning process, academic material scientists were relocated to the production site to ensure a smooth transition. The FAT was conducted whilst training, engineering support, infrastructure and production protocols (i.e. quality systems) were implemented by the academic/industry team members. After system commissioning and successfully passing the FAT, the production phase for the supply of spotter reflectors for the US-built Ford F250 series truck reflector assemblies commenced. The final reflector assembly is shown in figure 8.

Zoom In Zoom Out Reset image size

As part of the original project scope to ensure that any related teething-issues were dealt with in an expedient manner, a dedicated academic material scientist was assigned to the role of production support co-ordinator, a somewhat unusual appointment within the academic setting but one that reaped dividends. Over the next 12 months, any production process related issues that were encountered were resolved by the (academic) production support co-ordinator in conjunction with the process engineers. By way of example, issues arising included such things as defect identification and mitigation work, and adhesion related issues due to improper material handling. During this time both planned strategic training and ad hoc training (typically resulting from the necessity of remediating an immediate production issue) was implemented. The upshot of this 'hand-holding' phase was that production and engineering personnel developed a high level of industrially relevant scientific knowledge and process expertise, becoming experts in their own right. This phase of the project cannot be undervalued, as engineering a reliable production process can pose levels of complexity similar to those experienced at the beginning of the project. The facility now runs 24 hours a day 5 days a week, with engineering personnel possessing the skill-sets necessary for self-sufficient operation.