In order to identify possible wear mechanisms for our alloys, we first sought to better understand the nature and role of the individual scratch. This was done by producing coupons in each alloy that could be scratched using a conical Rockwell C hardness tester with a diamond indenter under controlled loads. The samples were first ground plane-parallel on both sides and then polished on the side designated for testing, followed by scratching under both constant and increasing loads. A tape lift consisting of adhesive tape applied directly and uniformly to the scratch in order to embed and remove any spalled material allowed us to compare the susceptibility of the platinum and gold alloys to scoring damage. Tapes from the lift were subsequently examined with energy dispersive X-ray spectroscopy (EDX) to confirm composition of the metal chips as well as characterise the amount of chipping.

A key objective for our study is the simulation of typical human wear mechanisms as closely as possible. There are countless chemical environments and unique mechanical forces that jewellery items are subjected to during human wear, hence a standardised test that attempts to replicate such conditions can only be seen as an approximation of what actually happens in real-life conditions. Correlation with the anecdotal is therefore critical in terms of supporting experimental outcomes as representative of what may be experienced in the human population.

The wear testing performed consisted of three different tests. The first being an abrasion test that utilises a stone and sand media, the second a corrosion test in artificial human sweat and the third a polishing test employing a nutshell media. All media used were calibrated and laboratory grade. Cycles were done in sequence fashion with each of the first five cycles including abrasion, followed by corrosion, followed by polishing. Two subsequent cycles were performed that omitted abrasion and corrosion and only included polishing media. The total test duration amounted to 252.5 h.

Five cube-shaped and individually identified samples of each alloy were used for the testing as shown in Figure 2. Before and after each test in the sequence samples were weighed and characterised by optical microscopy and Vickers microhardness testing. Samples were cleaned in an ultrasonic bath with ethanol to assure any media that might be clinging to the surface was removed. The surfaces of select samples were also characterised by SEM.

The abrasion and polishing tests were based upon the European Industrial Standard DIN EN 12472. The apparatus consists of a motorised rotating drum (Figure 3) that is filled with either an abrasive blend of sand and stones (abrasion test) or nutshells (polishing test). According to the standard, the samples must be physically isolated from one another during testing in order to avoid mutual damage through sample-to-sample contact. Therefore, cubes were anchored along a nylon cord attached to both ends of the drum frame.

Fig. 3

The possible roles of corrosion and erosion corrosion, specifically in gold alloys that contain significant amounts of corrosion-prone base metal elements, were other areas we considered as possibly contributing to wear. The platinum alloys tested were pure platinum group metal (pgm) alloys that did not contain any base metals and are otherwise well-known for their high resistance to chemical corrosion. Therefore, while we did not expect this test to have any effect on pure pgm alloys we included them for the sake of completeness. The corrosion test was based upon the international standard ISO 3160-2. The test involves application of artificial human sweat to the test cubes followed by heating in a closed chamber at 40°C +/– 2°C for 24 h (Figure 4). This test was conducted for cycles one through five right after the abrasion test and prior to the polishing test. Table II gives the composition of the artificial sweat and Figure 4 shows the samples positioned in the chamber. Following the test, samples were cleaned in an ultrasonic bath of deionised water and documented by light optical microscopy.

Fig. 4

Tensile testing was performed in accordance with ISO 6892-1 and microhardness testing was done using a 100 g load (HV0.1) in accordance with DIN EN ISO 6507-1. Tensile properties for cast product were derived from the same casting processes as the test cubes and coupons with the exception of the gold-nickel alloys that were cast by the producer of these alloys. Details on tensile testing are described in (14).

Prior to testing, samples were documented by stereomicroscopy and light optical microscopy. Due to hand polishing the samples exhibit some deviation from the ideal shape as shown in the computer aided design (CAD) images. Selected samples were also documented to obtain details of the geometry, shape and surface condition (Figure 5). After the fourth and fifth cycles the surfaces of select samples were also investigated by SEM (Figure 6).

Fig. 5

Fig. 6

The cube dimensions were measured using a calibrated micrometre calliper. Mass was determined by an analytical balance with an accuracy of 10 μg. Density was determined with the same balance using the buoyancy method (Archimedes’ principle). The mass and volume losses were determined after the abrasion and polish tests and in order to compare the samples, mass loss was normalised with the sample surface area. Volume loss was calculated by dividing mass loss by density.

Vickers hardness of each sample was measured in the as-polished condition and after completion of each cycle (abrasion + corrosion + polish). One measurement was done on each side of the cubes with the exception of the side bearing the sample ID. Table III gives the average hardness value of each sample.

Through SEM analysis (Figures 7 and 8) we see the evidence that the depth of the scratch is impacted by the hardness of the alloy. As one might expect, the softer the alloy, the deeper the scratch and the more material is displaced. In the case of the soft alloy 950PtIr, the displaced material was concentrated at the edges and the tip of the scratch (Figure 8), which is typical for micro-ploughing. Local overload also resulted in cracking of the displaced material at the edge of the scratch that appears to be loosely connected. In comparison, the gold alloys showed not only cracking, but also significant chipping along the cracks. This was especially true for the 585AuNi, which has a stronger tendency for micro-cutting.

Fig. 7

Fig. 8

We noted that the alloys appeared to show different levels of porosity after polishing with the platinum alloys exhibiting low levels and the gold alloys exhibiting higher levels characterised as finely dispersed microshrinkage. From previous studies on the tensile properties of platinum alloys (13) it was established that the ductility values of elongation and reduction of area are significantly impacted by porosity levels. Therefore, increased chipping in the gold alloys may be not only a result of intrinsically lower ductility for these alloys, but also porosity-related decreases.

3.1.1 Tape Lift

Corrosion was qualitatively assessed by optical microscopy after each test. The presence of corrosion was most visible after the first cycle because the surface had less scratching from the wear tests than subsequent cycles. As expected for pure pgm alloys, both 950 platinum alloys (Table I) showed no visible changes following corrosion testing (Figure 10).

Fig. 10

The alloy that demonstrated the least amount of resistance to corrosion was the 585AuNi containing high amounts of nickel, copper and zinc (Figure 11). Following wear testing porosity was exposed to the surface, suggesting that corrosion was further promoted by microshrinkage pores that had been revealed. Such pores act as crevices where a concentration of corrodents is able to accelerate the corrosion process. This being the case, the casting quality level may be a contributor to reduced (or improved) wear resistance, particularly in alloys demonstrated to have low corrosion resistance such as the 585AuNi.

Fig. 11

The 750AuNi and both the 585AuPd and 750AuPd alloys did not exhibit visible corrosion after any of the five cycles. While higher corrosion resistance is expected with the greater noble metal content of these alloys, the potential effects of corrosion cannot be ruled out given their base metal content and the limited scope of our testing. Moreover, the corrosion testing performed was of a static nature, omitting the potential for an erosion corrosion dynamic that is likely present in human wear conditions. This topic is recommended for further testing to better understand the potential for effects on wear resistance in gold alloys.

The goal of this series of tests was quantitative determination of mass loss and volume loss through a combination of abrasion testing and polish testing. The total testing time can be segregated into abrasion time (sand + stone media) and polish time (nutshell media). Mass loss and volume loss were normalised with the surface area of the sample, allowing us to compare data from samples with a different geometry. The plotted values show the mass loss and volume loss per surface area of the sample. For simplicity, the terms ‘mass loss’ and ‘volume loss’ are used for normalised values in the text of this paper. Mass and volume loss were plotted against abrasion and polishing time and total wear time, respectively. The plots show the average loss of the five samples per alloy that were tested. This allowed for a segregation of data for the amount of wear measured in each of the different tests.

During the abrasion test portion of our assessment the mass and volume losses show a non-linear increase with increasing abrasion test time (Figure 12) in the beginning of the tests, which turns into a linear trend with increasing testing time. No remarkable difference between the alloys is observed and overall mass loss during abrasion testing is extremely small. The 585AuPd does show slightly higher wear compared to other alloys in this phase of the cycle, but mass loss was only 0.00216 g, or 0.08% of original mass.

Fig. 12

Volume loss was calculated by dividing mass loss by density. Due to the considerably different densities of the tested alloys three groups can be distinguished. The platinum alloys have a density of ca. 20 g cm⁻³; 750 gold alloys are at ca. 15 g cm⁻³; and 585 gold alloys are at 13–14 g cm⁻³. While mass loss is very similar for all alloys, the volume loss differs more due to these distinctly different density levels. The platinum alloys showed the lowest volume loss, followed by the 750 gold alloys and the 585 gold alloys. Total volume loss in the abrasion test was very low with a maximum value at only 0.0005 mm³, or 0.03% of the original volume.

For the polishing test the mass and volume loss rate (i.e., the mass and volume loss per unit of time) was comparable to the loss rate abrasion test. Mass loss was demonstrated to increase linearly with increasing polishing time. The platinum alloys again show the lowest mass loss with total mass loss after 244 h of combined testing at less than half that of the 750AuPd, which showed the highest mass loss in the group. The mass loss of the 585AuPd and the 585AuNi lies in between the two 750 gold alloys. The total mass loss after 244 h of testing was 0.013 g for the 950PtRu (lowest value) and 0.031 g for the 750AuPd (highest value). These are still very small amounts equal to 0.03% for the 950PtRu and 1.1% for the 750AuPd. However, when we consider volume losses these differences take on much greater significance. The volume loss of both 950 platinum alloys is a factor of three times lower compared to 750AuPd, and a factor of about two times lower compared to 585AuPd and both 750AuNi alloys.

Figure 13 shows the total mass and volume loss after all cycles of wear testing were completed. Since the absolute mass loss in the abrasive test was much lower than that in the polishing test, the abrasive test was omitted in the last two cycles of wear testing. The result in Figure 13 is very similar to that of Figure 14. Error bars indicate the results from the samples with the lowest and highest mass loss in one group of alloys, while the full symbols indicate the averaged mass loss of the five samples. The error bars confirm that the difference between the alloys remains significant. The mass loss curves demonstrate a linear trend that was fitted for select alloys. The slope indicates the mass loss per hour of wear, i.e., the rate of wear.

Fig. 13

Fig. 14

The assessment of surface quality focused on the rounding of corners and edges, which was qualitatively determined by stereo microscopy. Figure 5 demonstrates the samples with the lowest and highest volume loss, which are 950PtRu and 750AuPd respectively. Figures 5(a) and 5(c) were taken after completing the first two cycles of 10 h total wear testing. After 10 h very little difference can be detected in comparison with the as-polished condition of the samples. The mass loss after two cycles was only 0.0004 g, therefore this result is expected. Figures 5(b) and 5(d) demonstrate the sample surface after completing seven cycles. 950PtRu displays a very well-defined cube shape after the second cycle and only a very slight rounding of the corners following the seventh and final cycle. The absolute mass loss after the complete series of testing was 0.0131 g, or 0.3% for the 950PtRu.

All five of the 750AuPd samples displayed a less-defined cube shape in the as-polished condition as a result of hand polishing prior to testing. The surface also appears somewhat uneven (Figure 5). Nevertheless, a continuing deterioration of the cube geometry was demonstrated through testing. Following the second cycle edges and corners present with increased rounding, and this condition is even more pronounced after the seventh cycle, indicating mass loss had occurred during testing. Absolute mass loss for the 750AuPd after the completion of wear testing was 0.0307 g, or 1.1%.

The surface of select samples and conditions was captured by SEM imaging. Figure 6 depicts the samples with the lowest and highest volume loss after abrasion testing (Figures 6(a) and 6(c)) as well as subsequent corrosion and polish tests (Figures 6(b) and 6(d)). Following the abrasion test both sample surfaces are quite rough and exhibit deep dents and scratches. After the polish test both samples display a levelling of the topography of the sample. Notably, despite its lower hardness, (or perhaps because of it) 950PtRu exhibits a smoother surface finish compared to 750AuPd.

Tensile testing was performed to determine whether strength and ductility measures might play a role in mass loss. Table III shows the average results of tensile testing from four as-cast bars in each alloy. The hardness values are the average values that were measured on a set of five cube samples of each alloy.

We did not find any significant correlation with tensile properties or hardness and mass loss. As other studies showed before (2), it appears that high hardness is not an indicator for low mass or volume loss. However, the opposite also cannot be concluded. Rather, the situation appears to be more complex and depends upon the mechanism of mass loss during wear testing. The alloys exhibited very different hardness levels with one series of samples (585AuPd) showing a spread of more than 10%, indicating an inhomogeneous microstructure, due to porosity for example. Micropores were visible on the polished coupons of the 585 gold alloys (Figure 8).

It has been demonstrated in platinum alloys that the reduction of area value (ROA) is strongly reduced by microporosity (14). If this is the case, then the microstructure of the samples plays an important role on wear behaviour. Micropores along scratches will act as points of stress concentration and may cause the chips to break free. Increased levels of microporosity are likely to favour micro-chipping over micro-ploughing, suggesting increased mass loss due to metal chips. Further investigations will be necessary to prove such a hypothesis.

Industries face mounting challenges in the paradigm shift to a more circular economy. Research and development is increasingly focused on finding ways to turn waste into resources, recover energy and materials and make better use of resources extracted from the natural environment. At the same time industry and consumers seek to cause less harm in the form of pollution or CO₂ emissions. In this issue of the Johnson Matthey Technology Review, we look at current and future technologies that may be used by industries including energy, fuels, chemicals, pharmaceuticals and transport to create the products we need while meeting the United Nations (UN) 2030 goals for sustainable development (1): “development that meets the needs of the present without compromising the ability of future generations to meet their own needs” (2).

Continuing the theme from our previous issue (3), several articles present different approaches to future fuels and chemicals. These approaches include water electrolysis, utilisation of biomass and waste and CO₂ reduction.

Electrification will provide alternatives to fossil fuel use in many areas of industrial science and technology but some areas like long-haul aviation will likely continue to need liquid fuels. These fuels will be provided through one or more of the technologies being developed today. For example, a technology recently commercialised by Johnson Matthey and bp uses a Fischer-Tropsch process to create sustainable jet and diesel fuels from waste, biomass or existing CO₂ emissions. The challenges involved in achieving a commercially viable process at scale are explained.

The technoeconomics and life cycle assessment of producing sustainable commodity chemicals from waste biomass using aerobic fermentation at scale are explored in another Johnson Matthey collaboration, this time with the University of Nottingham and Northumbria University, UK. Rigorous process modelling has determined at what point the production of commodity chemicals from a lignin source will become commercially viable. The future of this promising technology looks bright, with the authors concluding that their platform has promise as a best-in-class technology for the production of a broad spectrum of renewable commodity chemicals.

Activated carbon can be produced and characterised from biomass waste for applications in environmental protection, clean energy and catalysis. The work is presented by Gebze Technical University, Turkey, in collaboration with Gasification Consultancy Ltd, UK. Waste biochar from the gasification of biomass is the feedstock, and removal of contaminants is key to its successful use.

Reducing CO₂ emissions from iron metallurgy will become increasingly important. The electrification of primary iron production in a carbon-free process is presented in a collaborative research article from National Technical University of Athens, Laboratory of Metallurgy and Mytilineos SA, Metallurgy Business Unit-Aluminium of Greece. The technology is demonstrated at an early stage with additional optimisation recommended by the authors. Catalytic hydrogenation of CO₂ to methane using power-to-gas combined with biomass gasification is another option to reduce the CO₂ emissions of the steel industry, presented by Montanuniversität Leoben, Johannes Kepler Universität Linz and K1-MET GmbH, Austria.

This journal has long championed sustainable technologies involving the precious metals. Metals are inherently recyclable and none more so than the platinum group metals (4). Today’s focus on electrification of transport and energy means that elements such as lithium, nickel, cobalt and manganese join their precious cousins as critical materials for the clean energy revolution. Clean and efficient extraction of these minerals from spent lithium-ion cathodes is an emerging area of study that will become increasingly important in the coming years when batteries begin to reach end-of-life. Recycling techniques need to be developed for the sustainable development of the lithium-ion batteries industry as discussed in this issue. Meanwhile life cycle assessment of the entire lithium-ion batteries production process from both primary ore and recycled material is provided in the output from an Innovate UK project involving Johnson Matthey and the Warwick Manufacturing Group, UK.

Energy efficiency will be a key enabler for a transition to a low carbon future. High technology industries like electronics, energy and medical applications require novel materials and processes. Cooling is a challenge, especially at the microscale. Nanofluids containing titania offer a potential solution and are investigated in this issue.

Conventional technologies will continue to be used alongside newer ones. To help define the next generation of emissions legislation to clean up the air in China, a portable emissions measurement system was used to investigate on-road tailpipe volatile organic compounds emissions in diesel trucks compliant with Euro III–V. The results with recommendations from the authors are presented in this issue.

It will become apparent from reading this issue that collaborations both within and between industry and academia are vital to progress. The research projects described here are just a selection. Many more advances can be expected in the coming years and decades as fruitful collaborations continue apace, with industry and academia working together to meet the challenges of the present and the future.

Growing environmental concerns have made it imperative to reduce global climate change and this has resulted in prolific development of various energy storage technologies for different applications ranging from portable electronic devices (PED) to electric vehicles (EVs) (1, 2). The most common chemical energy storage devices are batteries for applications requiring high energy density and electrochemical capacitors (ECs) for applications with high power density requirements (3–5). LICs which have the combined desirable properties of batteries (high energy density) and ECs (high power density) are increasingly being investigated as high-performance energy storage devices that have a significant role in the decarbonisation of the transport sector (6, 7).

While there are many promising negative electrode materials for LICs, the lithium titanium oxide (Li₄Ti₅O₁₂, LTO) based anode offers high stability towards charge-discharge cycles, faradaic efficiency and lower costs (8–10). As the envisaged use of the LTO based LIC is in hybrid and EVs to assist in decarbonising the transport sector, it becomes pertinent to conduct a LCA for the production of a LIC using primary ore minerals and make comparisons to a manufacturing process that relies on recycling end-of-life LIC. LCA is defined as a process to evaluate the environmental burdens associated with a product, process or activity by identifying and quantifying energy and materials used and wastes released to the environment (11). The assessment includes the entire life cycle of a product, process or activity, encompassing extracting and processing raw materials, manufacturing, transport and distribution, use, reuse, recycling and final disposal.

LCA facilitates informed decision making as comparative analysis of competing processes or products can be conducted based on environmental impact. At the early stages of R&D activities, LCA is an invaluable tool as it can inform process and material choices that support sustainability goals in addition to promoting innovation for designing products that are more amenable to recycling when they reach end-of-life (12–14). Increasingly, LCA is also being utilised to engage with stakeholders as an evolving green marketing tool through brand competitive differentiation on the basis of sustainability as well as regulatory compliance purposes (15–17). Besides the multifaceted benefits of LCA, its utilisation is not without limitations with uncertainties in inventory data, methodology and application of the weighting technique often being cited as major weaknesses of the approach (18, 19).

While the LCA methodology has been widely applied to energy storage systems this has mostly been for lithium-ion batteries (LIBs), with most studies having focussed on comparative analysis of LIBs to internal combustion engine (ICE) or sustainability of the different battery chemistries (20–22). There is a scarcity in the literature of LCA studies that have analysed production of energy storage devices using primary ore materials in comparison to manufacture of a similar product using recycled materials and specifically for LICs. This study objective is to take a comparative approach with the aim of utilising LCA to inform early phase R&D activities to improve the sustainability of the various process and reagents choices in the production of a LIC module.

The LCA study was conducted as part of the Advanced Lithium Ion Capacitors Electrodes (ALICE) project whose objective was to develop a 48 V LIC module for use in automotive, e-bus and materials handling equipment. The project consortium had industrial and academic partners for developing and scaling-up materials production including application of novel coating techniques to electrode structure to improve performance. The 48 V module built in the project was tested based on end user requirements and physics based numerical modelling applied at different stages of the project to interlink sophisticated layer structure characterisation results with cell performance.

2.1 Goal and Scope

2.2 Functional Unit and System Boundary

2.3 Methods and Databases

The elementary flows of material required to make the cells for a 48 V LIC module are based on pilot plant data. The recycled process data is based on a laboratory flowsheet that is simulated using a process model with appropriate scaling of model parameters from gram scale to a full-scale production plant. The main product and process stages for the primary (ore) and recycled materials manufacture of LIC are as follows: (a) anode powder material preparation; (b) anode preparation; (c) cathode preparation; (d) electrolyte preparation; (e) cell formation; and (f) recycling of LTO powder (recycled material process only).

The detailed breakdown of materials for the assemblies and product stages of the two LCA comparative projects is in Table I. The only difference between the two comparative studies is in the source of lithium carbonate and titania for making the anode LTO powder material. For the primary (ore) process, the information for the environmental footprint associated with lithium carbonate and titania is obtained from the ecoinvent 3.5 database based on ore extraction and salt formation environmental impact values. However, in the case of the LIC module made from recycled materials, lithium carbonate and titania are obtained from the recycling product stage and only have process environmental impact values associated with reagents and energy consumption demand to recycle the end-of-life LTO anodes.

The supplementary data which contains the flowcharts and inventory to produce a lithium ion capacitor module and the list of assumptions used in the study is located with the online version of this article.

3.1 LTO Powder Synthesis

3.2 Anode and Cathode Preparation

3.3 Electrolyte, Cell and Lithium-Ion Capacitor Module Formation

3.4 Modelling of LTO Recycling Process

While results for several environmental impact categories were available for analysis, for purposes of this study climate change (kilogram of CO₂ equivalent) and terrestrial acidification (kilogram of SO₂ equivalent) were analysed in greater detail for comparing the LIC module manufacture from primary ore materials against the recycled material process. The calculations are based on the ReCiPe Midpoint (H) with European Normalisation (24). The ReCiPe method was utilised because of its environmental relevance to the scope of the study, transparency and reproducibility. However, other methods which are also compatible with ISO standards could have been applied to the study.

Aluminium had the highest climate change and terrestrial acidification burdens to the extent of overshadowing contributions from other materials. To facilitate detailed analysis of environmental burdens of the other materials and processes, visual graphics of the results were plotted without the contribution from aluminium. Aluminium has established recycling processes but the decision if the quality of this recycled aluminium was of specifications sufficient for direct use in LIC manufacture was indeterminate and therefore the LCA credit process was not applied towards aluminium used. Figure 4 compares the climate change impact for making a 48 V LIC module using primary ore material and recycled LTO. Overall, utilising recycled LTO materials reduces the climate change impact by 12%. The order of decreasing climate change for the LIC module manufacture using primary (ore) materials is titania > lithium hexafluorophosphate > ethylene carbonate. For LIC module manufacture using recycled LTO, the order of decreasing climate change is lithium hexafluorophosphate > formic acid > ethylene carbonate. The highest contributor towards climate change for primary (ore) case is titania while for the recycled LTO it is the lithium hexafluorophosphate electrolyte. Lithium hexafluorophosphate and ethylene carbonate are both part of the electrolyte system and have significant contributions which are equal for LIC manufacture using either primary (ore) or recycled LTO materials. Therefore, significant reductions in climate change for LIC manufacture using recycled LTO can only be achieved by reducing the quantities of formic acid used. Table II shows the climate change impact over the various stages of manufacturing a 48 V LIC module. The anode preparation stage has the highest contribution towards climate change for the two comparative cases. However, using recycled LTO lowers the climate change impact by 21 kg_CO2eq compared to using primary (ore) during the anode preparation stage. The cathode preparation and cell formation stages have the same values as the two cases only differ in source of materials used the anode preparation stage.

Fig. 4

Comparative analysis of terrestrial acidification for producing a 48 V LIC module using primary (ore) and recycled LTO materials is shown in Figure 5. Usage of recycled LTO for the anode manufacture product stage results in 18% reduction in terrestrial acidification compared to using primary ore materials. The major contributors towards terrestrial acidification in decreasing order are lithium hexafluorophosphate > titania > ethylene carbonate for LIC module manufacture using primary (ore) materials. For LIC manufacture using recycled LTO, the major contributors towards terrestrial acidification in decreasing order are lithium hexafluorophosphate > formic acid > ethylene carbonate. Table III shows the terrestrial acidification associated with the various stages of manufacturing a 48 V LIC module for the two cases. The anode preparation stage has the highest contribution towards terrestrial acidification when the primary (ore) and recycled LTO material sources are compared. By utilising recycled LTO for the anode preparation process the terrestrial acidification impact is lowered by 0.21 kg_SO2eq compared to using primary (ore) materials.

Fig. 5

While there are environmental benefits from using recycled LTO, the existing recycling process flowsheet has a lot of optimisation opportunities especially regarding the quantities of formic acid used which have a significant contribution towards both climate and terrestrial acidification.

The application in early phase R&D activities is demonstrated in this section as applied to process development choices for the recycling stage. Before the hydrometallurgical treatment detailed in Section 3.4, the LTO has to be decoated (removed) from the aluminium foil to which it is bound. The widely used binder for coating the LTO and most active materials to current collectors is polyvinylidene fluoride (PVDF) because of its adhesive capabilities and electrochemical stability (25). The PVDF binder presents a challenge to the decoating process as it is only partially soluble in most common solvents. The common solvent for dissolving PVDF is NMP which is also used during the slurry coating process. However, NMP has high environmental and toxicity burdens which has resulted in stringent legislative restrictions of its usage (26).

To improve the sustainability metrics of the recycling process, several alternative solvents were investigated for their capabilities to remove the LTO from the aluminium foils. Acetone and polyethylene glycol (PEG) have similar properties to dipolar aprotic solvents like NMP and were identified as greener alternatives (27). Several other organic reagents such as acetic acid, formic acid, ethylene glycol and methanol were also screened as potential candidates for the process. Figure 6 shows images of LTO anodes after stirring in different reagents for 1 h at room temperature. From Figure 6, formic acid has higher technical performance compared to the other solvents as it removed all of the visible traces of carbon and LTO from the aluminium foil.

Fig. 6

The decision-making process considered the environmental footprint of these reagents in addition to their technical performance for removing the LTO from the aluminium foils. Figures 7(a) and 7(b) show the contribution towards climate change and terrestrial acidification respectively based on using 1 kg of these reagents. NMP has the highest environmental burden for climate change and terrestrial acidification, while methanol has the least environmental footprint. However, methanol efficiency in decoating LTO was low and therefore a trade-off of technical performance, environmental impact and costs resulted in formic acid as the alternative reagent choice for the decoating process stage. The bulk purchase price for these reagents is in Table IV.

Fig. 7

While several product and process research activities are focussing on novel binders that are less toxic and low costs compared to PVDF (29), the electrodes bound with PVDF that have already been manufactured will still require a more environmentally sustainable process to recover the active material at their end-of-life. The approach applied in this case study of early phase R&D process development activities demonstrated sustainable choices in alternate reagent selection in alignment to the triple bottom line approach (30). The choice of formic acid when compared to NMP results in intersection of people, planet and profit (3Ps) requirements of sustainability. For this case study, formic acid had reduced environmental impact, toxicity and meets the profit criteria through high decoating technical efficiency at lower costs compared to NMP.

The LCA methodology was applied to quantitatively determine the environmental burdens associated with manufacturing a 48 V LIC module. The prospective LCA compared the environmental impact of manufacturing a LIC module using primary ore materials versus LIC manufacture using recycled materials from end-of-life LICs. The anode preparation stage is associated with most of the environmental burden for manufacturing the LIC module for both processes due to the source of precursors used in production of the active LTO material. Utilisation of LTO precursors from recycled end-of-life LICs reduced both climate change and terrestrial acidification environmental impact categories for the LIC module manufacture. However, the sustainability metrics of the recycled process route of production could potentially be improved further by optimised application of formic acid which is used in the process stage for separating the LTO from the aluminium current collector foils.

The application of the LCA methodology in early phase R&D activities was demonstrated for the process development reagent choice case study. The LTO decoating reagent decision-making process considered the environmental footprint, technical performance and costs. The decision to utilise formic acid as a decoating agent was a sustainable choice which balanced environmental, economic and social performance. For the demonstrated case study, the choice of formic acid as decoating reagent reduced climate change and terrestrial acidification, lowered human toxicity values and met the profit criteria through high separation efficiency at lower costs.