Environmental Impact Assessment of Na3V2(PO4)3 Cathode Production for Sodium‐Ion Batteries

Sodium‐ion batteries (NIBs) are key enablers of sustainable energy storage. NIBs use Earth‐abundant materials and are technologically viable to replace lithium‐ion batteries in the medium term. Na3V2(PO4)3, as a popular cathode for NIBs, requires further improvements to boost its electrochemical performance, particularly regarding the rate capability and operational lifetime. These strategies involve the incorporation of carbonaceous materials, heteroatom doping, morphology modification, or biopolymer incorporation. Considering the circular economy actions to foster environmentally sustainable battery industries, there is an urgent need to disclose the environmental impacts of battery production. A cradle‐to‐gate life cycle assessment methodology is used to quantify, analyze, and compare the environmental impacts of ten representative state‐of‐the‐art Na3V2(PO4)3 cathodes. Impacts are disclosed for 18 indicators normalized to 1 kg of cathode considering laboratory‐scale approaches. Global warming potential values of 423.9–1380.0 kg CO2‐equiv. kgcathode −1 and 539.8–1622.1 kg CO2‐equiv. kWhcathode −1 are obtained considering Na3V2(PO4)3/Na half‐cell configuration. Simple carbon additives mixed with NVP provide a good CO2 footprint‐to‐storage capacity balance, although the sacrificed capacity retention hinders reuse strategies. A sensitivity analysis demonstrates a 16.9–38.0% reduction transitioning from fossil‐based to renewable‐based energy mix. Herein, it is aimed to support battery developers and assist future advances in the development of sustainable cathodes applied into beyond‐Li‐ion technologies.

Sodium-ion batteries (NIBs) are key enablers of sustainable energy storage. NIBs use Earth-abundant materials and are technologically viable to replace lithium-ion batteries in the medium term. Na 3 V 2 (PO 4 ) 3 , as a popular cathode for NIBs, requires further improvements to boost its electrochemical performance, particularly regarding the rate capability and operational lifetime. These strategies involve the incorporation of carbonaceous materials, heteroatom doping, morphology modification, or biopolymer incorporation. Considering the circular economy actions to foster environmentally sustainable battery industries, there is an urgent need to disclose the environmental impacts of battery production. A cradle-to-gate life cycle assessment methodology is used to quantify, analyze, and compare the environmental impacts of ten representative state-of-the-art Na 3 V 2 (PO 4 ) 3 cathodes. Impacts are disclosed for 18 indicators normalized to 1 kg of cathode considering laboratory-scale approaches. Global warming potential values of 423.9-1380.0 kg CO 2 -equiv. kg cathode À1 and 539.8-1622.1 kg CO 2 -equiv. kWh cathode À1 are obtained considering Na 3 V 2 (PO 4 ) 3 / Na half-cell configuration. Simple carbon additives mixed with NVP provide a good CO 2 footprint-to-storage capacity balance, although the sacrificed capacity retention hinders reuse strategies. A sensitivity analysis demonstrates a 16.9-38.0% reduction transitioning from fossil-based to renewable-based energy mix. Herein, it is aimed to support battery developers and assist future advances in the development of sustainable cathodes applied into beyond-Li-ion technologies.
the need for CRMs such as cobalt, lithium, manganese, or graphite often needed in LIBs. Thanks to the cheaper cathode active materials and the avoidance of Cu at the current collector, NIBs at battery cell level are clearly cheaper than LIB cells. [11] This lower cost together with the chemistry and technology similarities with conventional lithium-ion-based batteries facilitates the replacement of LIBs by NIBs in the medium term. In addition, the environmental sustainability of NIBs was confirmed by Peters et al. in 2016, who used the life cycle assessment (LCA) methodology to conclude that in comparison with LIBs, NIBs present reduced impacts in global warming, fossil depletion potential, freshwater eutrophication, and human toxicity potential per 1 kWh of storage capacity. [12] In spite of these benefits, the lower theoretical capacity of 1165 mAh g À1 of Na as opposed to 3829 mAh g À1 of Li results in lower specific energy densities, limiting their practical implementation. [13,14] As a critical battery component determining its electrochemical performance, [15] diverse cathode materials have been developed to make NIBs competitive against LIBs. Among them, the Na 3 V 2 (PO 4 ) 3 (NVP) cathode having a fast Na þ -transportable NASICON framework is a popular choice given their balance between discharge capacity of 117.6 mAh g À1 and platform voltage of %3.4 V, outperforming the electrochemical performance of many NIB cathodes reported in the literature. [16] The initial promising results obtained with Na 3 V 2 (PO 4 ) 3 have sparked notable scientific efforts aimed at improving its relatively poor electronic conductivity of 1.63 Â 10 À6 S cm À1 and limited structural stability that result in low discharge capacities (at high rates) and poor operating lifespans. [17] Generally speaking, these synthetic approaches have been mainly focused on the improvement of electronic conductivity and structural stability of Na 3 V 2 (PO 4 ) 3 , so an efficient Na þ and e À transference through the cathode could be obtained with limited volume changes in the process of Na þ extraction/insertion. [17] Notwithstanding the many different synthetic approaches reported, these can be roughly categorized into the incorporation of diverse carbonaceous materials to enhance electric conductivity, [18][19][20] structural arrangement, heteroatom doping, [21][22][23] particle downsizing, and morphology modification to shorten transport distances of Na þ and electrons [24,25] or biopolymer incorporation. [26,27] Specifically, it is possible to improve the reversibility and cycling stability (at high rates specially) of the battery upon dehybridization of NVP with carbon nanostructures, which not only improve the cathode's electronic conductivity but also its structural stability. [28,29] In addition, the presence of mesoporosity in the cathode enhances sodium storage performance in terms of achieved reversible capacity, rate performance, and cycle life. [30] However, these approaches are not designed according to green chemistry principles and often involve a synthetic process with environmentally harmful reagents and large amounts of energy. In addition, Na 3 V 2 (PO 4 ) 3 contains vanadium and phosphorus, which are CRMs due to their pressure over finite natural resources, supply chain risks, and economic relevance. [31] Considering these issues and the circular economy actions to foster the development of environmentally sustainable battery industries, [32] the assessment of the environmental sustainability of Na 3 V 2 (PO 4 ) 3 cathode fabrication becomes necessary. The LCA methodology can be applied to that end because it enables the quantification of the environmental impacts of a given product or a service over its whole (or specific) life cycle stage, covering raw material extraction, manufacturing, distribution, use, and end-of-life. [33,34] Following widely recognized calculation methods, LCA can provide information on a wide variety of impact categories including global warming, fossil resource scarcity, ecotoxicity, eutrophication, land use, acidification, ozone depletion, or water use. The public disclosure of the LCA results enables selecting not only the environmentally preferred Na 3 V 2 (PO 4 ) 3 cathode, but also gives insights for the prospective reduction of the impacts through the implementation of ecodesign strategies. [35] Accordingly, the environmental impacts originating from Na 3 V 2 (PO 4 ) 3 cathode materials designed to show improved electrochemical performance are exhaustively quantified, analyzed, and compared. Ten representative state-of-the-art Na 3 V 2 (PO 4 ) 3 cathode materials at laboratory scale are selected to provide a representative overview of the currently available alternatives. Although the environmental competitiveness of several NIB cathodes has been recently reported, no works have been devoted to specifically analyze Na 3 V 2 (PO 4 ) 3 cathodes. [36] Using a cradle-to-gate approach, the impacts are normalized to 1 kg of cathode material. As energy storage field is a performance-driven area, the environmental impacts are then normalized to 1 kWh of (cathode-only) storage capacity. Overall, this work is aimed to provide support for battery developers and assist future advances in the development of sustainable cathodes applied into beyond-Li-ion technologies.

Goal, Scope, and Life Cycle Interpretation
LCA was applied to quantify the cradle-to-gate environmental impacts of ten representative state-of-the-art Na 3 V 2 (PO 4 ) 3 cathodes. The lack of accurate information regarding the NIB end-of-life and its nonmature character make difficult cradleto-cradle studies. Accordingly, raw material acquisition and Na 3 V 2 (PO 4 ) 3 cathode synthesis is considered for laboratory-scale NIB cathodes. As shown in Figure 1, the laboratory-scale synthetic approaches involved different strategies to upgrade the electrochemical performance (discharge capacity, cycling stability and rate capability) of the Na 3 V 2 (PO 4 ) 3 cathodes, including the incorporation of carbonaceous structures, heteroatom doping, nanostructuring, or biopolymer incorporation (see Figure S1-S10, Supporting Information, for the followed synthetic approaches and obtained morphologies). As lab-scale LCA is recognized as an effective environmental advisory tool for emerging technologies and materials, [37] the outcome facilitates future follow-on works in the field of sustainable NIBs. In any case, we acknowledge that the obtained impacts may be considerably larger than actual industrial-scale batteries.
LCA studies were carried out with the OpenLCA software coupled with ecoinvent v3.8 dataset (released on September 2021). The environmental impacts and life cycle inventory (LCI) were provided to enable future comparison. The environmental impacts were grouped into 18 categories based on ReCiPe 2016 Midpoint (H). As a globally accepted method, [38] this method provides additional metrics in comparison with other approaches such as the CML-baseline. Given the relevance of greenhouse gas (GHG) emission to meet the Paris Agreement, the global warming potential (GWP) impact indicator, measured in kg·CO 2 -equiv., was used for comparison. The definition of an adequate functional unit (FU) represents a challenge; while FU closer to the actual service of the battery is desired (storage capacity, for instance), additional assumptions should be made, which entail risks of increasing the level of uncertainty. [39] Contrarily, using the mass of cathode as a FU has no direct relationship to the service of a battery but provides relevant information for straightforward comparison across multiple studies. Following the precedent established by Ellingsen et al., [40] 1 kg of synthesized cathode was used as a FU. However, as the energy field is an electrochemical performance-driven area, a second FU focused on the energy storage capacity was also considered. This second FU was set as the gravimetric energy density (based on cathode active mass; Wh kg cathode À1 ), particularly relevant toward practical implementation. [41] This standardization facilitated a simple and accurate comparison with other electrochemical energy-storage technologies.

Life Cycle Inventory
The LCI for Na 3 V 2 (PO 4 ) 3 cathodes is shown in Table 1 (data extracted from secondary sources). This table summarizes the inventory of input and output flows for the cradle-to-gate fabrication of the cathodes. The resources of water, energy, and raw materials needed for the synthesis were computed as the inputs, while the releases/emissions to air, soil, and water originating from the fabrication process were computed as the outputs. The LCI was constructed for the specific amounts of cathode synthesized in each of the works, which ranged from %15 to 500 mg.
When accounting for the electricity consumed upon fabrication, specific processes were estimated (generally involving stirring, heating, drying, or annealing). Additional details for the inventory are disclosed in the Supporting Information as Scheme S1-S10 and Table S1-S23, Supporting Information. The electricity mix of the European Network of Transmission System Operators (ENTSO-E) accounting for the electricity from 35 countries across Europe was used in the ecoinvent v3.8 database.
As representative state-of-the-art examples, ten Na 3 V 2 (PO 4 ) 3 cathode designs applied into NIBs were selected (all works have been published from 2015 onwards). A short description for each cathode is given hereafter. One of the first examples to upgrade the performance of Na 3 V 2 (PO 4 ) 3 was carried out by Fang et al., who wrapped the active material by a highly conductivity and interconnected hierarchical carbon framework (cathode 1, denoted as "hierarchical carbon-NVP"). [18] The chemical vapor deposition process affords graphene-like coating layers and interconnected carbon nanofibers onto highly crystallized Na 3 V 2 (PO 4 ) 3 particles, boosting e À transport and accommodating volume changes upon Na þ insertion/extraction. As a result, the half cell delivered 110 mAh g À1 at 2C (1C ¼ 117 mAh g À1 ) with a remarkable 54% capacity retention after 20 000 cycles at 3C rate. Following the combination of Na 3 V 2 (PO 4 ) 3 with carbonaceous structures, Na 3 V 2 (PO 4 ) 3 nanoparticles were grown in between reduced graphene oxide (rGO) layers upon the modification of the surface charge of the gel precursor (cathode 2, denoted as "rGO-LbL NVP"). [19] The obtained layer-by-layer structure concomitantly increased the electronic/ionic conductivity, offering a rapid Na þ diffusion and electron transport pathway, while the physical structure upon charge/discharge cycles was maintained. As a result, %110 mAh g À1 at 2C with a remarkable 70% capacity retention after 15 000 cycles at 5C rate was achieved (half-cell). The simple combination of sol-gel and annealing under Ar-H 2 atmosphere rendered Na 3 V 2 (PO 4 ) 3 nanoparticles   7: "3D NVP nanofiber" 8: "Nanoplatelet NVP" 9:"Agarose carbon NVP" and 10: "Glucomannan NVP". www.advancedsciencenews.com www.advenergysustres.com being embedded into porous carbonaceous microspheres (cathode 3, denoted as "μPorous NVP"). [20] Thanks to the favored nanoparticle connectivity and improved electron transfer by the carbon phase, %82 mAh g À1 was achieved at 2C (half cell). Heteroatom doping (with nitrogen, boron, sulfur, or phosphorus) is a widely pursued approach to enhance the electrochemical performance of carbonaceous materials as it introduces additional active sites for electrochemical reactions, enhances electronic conductivity, and improves surface wettability. In particular, nitrogen doping has been proven efficient to increase the intrinsic low electronic conductivity of Na 3 V 2 (PO 4 ) 3 and thus improve its cyclability and rate capability. Kim et al. used polydopamine as a nitrogen-containing source to obtain N-doped mesoporous carbon-wrapped Na 3 V 2 (PO 4 ) 3 (cathode 4, denoted as "N-doped carbon NVP"). [21] Thanks to the adhesive properties of polydopamine, intimate contact between the doped carbon and Na 3 V 2 (PO 4 ) 3 was achieved, reaching 94.8% capacity retention after 1000 cycles at 2C (half cell). In addition, dual-atom doping involving nitrogen and boron was pursued to obtain a codoped carbon-coated 3D Na 3 V 2 (PO 4 ) 3 composite (cathode 5, denoted as "N,B-doped carbon/NVP"). [22] The doping enabled a quick Na þ and electron transport and the flower-like morphology shortens the electronic transport distance while protects the cathode structural integrity, resulting in NIB with a long cycle life as proved by 38 mAh g À1 after 5000 cycles at 2C (half cell). Not only the carbon phase but also the Na 3 V 2 (PO 4 ) 3 itself was also doped to improve its electrochemical performance. In particular, lanthanum (La) was incorporated into the Na 3 V 2 (PO 4 ) 3 structure (Na 3 V 2Àx La x (PO 4 ) 3 /C) upon a combination of solÀgel and carbon-thermal reduction methods (cathode 6, denoted as "La 3þ -doped NVP"). [23] La 3þ doping expanded the lattice structure, increasing the Na þ mobility and providing a capacity retention of 93.5% after 3000 cycles at 2C (half cell).

Sodium ion batteries
Several works have attempted to modify the Na 3 V 2 (PO 4 ) 3 morphology to enhance the rate-capacity and cycling stability. In this context, a self-sacrificed approach was followed to change the morphology of Na 3 V 2 (PO 4 ) 3 itself into a 3D connected nanofiber network (cathode 7, denoted as "3D NVP nanofiber"). [24] Such a structure combined the multichannel ion diffusion paths with continuous electronic conduction and a physically stable architecture, reaching 95.9% capacity retention after 1000 cycles at 1C when evaluated into a half cell. Similarly, Li et al. obtained highly crystalline and porous Na 3 V 2 (PO 4 ) 3 nanoplatelets via hydrothermal method and postcalcination under Ar-H 2 atmosphere (cathode 8, denoted as "Nanoplatelet NVP"). [25] The pores within the Na 3 V 2 (PO 4 ) 3 enlarged the specific surface and shortened the Na þ and e À diffusion path, while the thin nanoplatelet character provided easy accessibility of the electrolyte to the active sites, reaching a remarkable rate performance of 76.5 mAh g À1 at a rate as large as 10C (half cell).
Given the inherent polar character and carbon content of natural biopolymers (polymers originating from the cells of living organisms), [42] these macromolecules could be used to improve the electrochemical properties of Na 3 V 2 (PO 4 ) 3 cathodes. For instance, a soft templating approach using biopolymers substantially enhanced the electrochemical performance of NIBs. In this context, agarose dissolved in boiled water was applied as a carbon source to coat 20-200 nm Na 3 V 2 (PO 4 ) 3 nanoparticles with a 3D carbonaceous skeleton (cathode 9, denoted as "Agarose carbon NVP"). [26] When mounted into a half cell, the cathode obtained after Ar-H 2 sintering offered a discharge capacity of 113 mAh g À1 at 1C, 60 mAh g À1 at 5C, showing 87.5% capacity retention after 8000 cycles at 2C (half cell). From the other side, the konjac glucomannan biopolymer was applied to develop an adhesive and stable solid permeable interface in Na 3 V 2 (PO 4 ) 3 cathodes (cathode 10, denoted as "Glucomannan NVP"). [27] Developed interface did not hinder Na þ transfer but limited the inherent oxidation of the cathode by the electrolyte at high voltages. In addition, the abundant polar groups facilitated Na þ transport, yielding a cathode material with 74.1% capacity retention after 10 000 cycles at 5C (half cell). This strategy of surrounding the active NVP by an amorphous carbon phase acting as a binder to enhance the structural stability upon charge/discharge has been also recently reported by Zhou et al. [43] 3. Results and Discussion

Global Warming Potential Per 1 kg of Cathode
Given its predominant relevance to meet the legally binding international treaties on climate change, we first studied the GWP impact category. The results, provided in kg CO 2 -equiv., account for the emissions from GHGs on the basis of their global warming potential. To do so, a conversion is carried out to the equivalent amount of CO 2 with the same GWP. Figure 2 shows the GWP values for the fabrication of 1 kg of cathode material for ten different laboratory-scale designs (cradle-to-gate approach). Overall, GWP values ranging from 423.9 to 1380.0 kg CO 2 -equiv. kg cathode À1 are achieved. Given the particularly striking 12810.5 kg CO 2 -equiv. kg cathode À1 value of the "Nanoplatelet NVP" cathode arising from the low quantity of synthesized material (15.6 mg is achieved in comparison with 469-516 mg obtained in the other nine processes; see Scheme S1-S10, Supporting Information, for further information), this cathode is estimated not representative and will not be considered for discussion. In general, the cathodes bearing biopolymers (average: 548.5 kg CO 2 -equiv. kg cathode À1 ) and consisting of the simple incorporation of carbonaceous nanostructures (average: 571.5 kg CO 2 -equiv. kg cathode

À1
) are the environmentally friendlier options given the simpler character of the synthetic processes and the use of Earth-abundant materials. On the other side, nanostructuring (633.1 kg CO 2 -equiv. kg cathode À1 for "3D NVP nanofiber") and especially doping (average: 705.9 kg CO 2 -equiv. kg cathode À1 ) entail the largest GHG footprint due to the need for doping agents and additional energy for production.
Obtained  [44] the cradle-to-gate %21 kg·CO 2 -equiv. kg cathode À1 of the LiCoO 2 cathode when prepared by solid-state synthesis, [45] the %32 kg CO 2 -equiv. kg cathode À1 showed by LiNi x Mn y Co z O 2 cathode, and the %28 k CO 2 -equiv. kg cathode À1 obtained for LiNi x Co y Al z O 2 cathode. [46] However, it should be considered that the impacts of the last two cathodes (which are in the same range of the other reports) correspond to cradle-to-gate results obtained from representative manufacturers of LIBs in China. Therefore, the 13-to-45 fold increase on the GWP category shown by this work can be explained by the notably more complex synthetic approaches here analyzed, which require energy-intensive multistep processes that are not designed according to atom-efficiency principles, and use relatively hazardous chemicals with notable embodied energy and carbon impacts. For example, the production of 0.516 g of "μPorous NVP" requires 51. In this sense, comparing the impacts of Na 3 V 2 (PO 4 ) 3 cathode production with the synthesis of high-performance nanomaterials may be more appropriate as the involved precursors and synthetic processes have a certain similarity. Accordingly, values of 1060-2360 kg CO 2 equiv. kg À1 for reduced graphene oxide production, or 217-501 kg CO 2 equiv. kg À1 for graphene oxide synthesis via the Hummers and Marcano methods, have been reported. [47] Importantly, these large values do not only apply for inorganic nanomaterials but also for nanoparticles having a renewable origin such as cellulose nanofibers (190-1160 kg CO 2 equiv. kg À1 ), [48] cellulose nanocrystals extracted from cotton (112 kg CO 2 equiv. kg À1 ), or the cellulose nanocrystals obtained from unripe coconuts (1086 kg CO 2 equiv. kg À1 ). [49] The underlying reasons for these impact differences will be explained in the forthcoming sections. In any case, it is worthy to note that according to the conclusions drawn by Piccinno et al., who reported that the environmental impacts of bio-based nanoparticle production can be lowered by a factor of 6.5 transitioning from the laboratory-scale (10 g) production to the industrial scale (50 kg), [50] we estimate that the carbon footprint of Na 3 V 2 (PO 4 ) 3 cathode production could be lowered in the near future.

Environmental Impacts in 18 Categories
A more detailed analysis of the environmental impacts facilitates the identification of the environmental hotspots during Na 3 V 2 (PO 4 ) 3 fabrication while enables future optimization through ecodesign approaches. Accordingly, the analysis has been expanded to all the impact categories considered by the ReCiPe 2016 Midpoint (H) LCA and the results are shown in Table 2. As occurring with the GWP value, the "Glucoammam NVP" cathode shows the lowest impacts in 16 of the 18 categories analyzed. These results may originate from the relatively simple synthetic process involving heating, stirring/evaporation, grounding, and subsequent heating at 800 ºC under N 2 atmosphere. In addition, Earth-abundant or organic precursors such as NaOH and citric acid have been used. However, it should be considered that all the cathode synthesis processes studied, "Glucoammam NVP" included, present highly hazardous inorganic vanadium-containing compounds as the V source for the cathode. Examples include vanadium(V) oxide (V 2 O 5 ), vanadium(III) acetylacetonate (V(C 5 H 7 O 2 ) 3 ), or ammonium metavanadate (NH 4 VO 3 ) depending on the design. Therefore, all the cathodes are subjected to certain levels of toxicity.
In spite of their multistep character involving several of the following treatments (ultrasound treatment, ball milling, stirring, autoclave heating, (freeze)drying, grinding, sintering under Ar/H 2 atmosphere, or heating under Ar atmosphere), the "hierarchical carbon NVP" "rGO-LbL NVP" and "μPorous NVP" cathode are the choices with intermediate impacts in most of the categories. These results highlight that the incorporation of carbonaceous (nano)structures into Na 3 V 2 (PO 4 ) 3 cathodes results in an environmentally sound alternative. Similarly, the "La 3þ -doped NVP" doping approach seems a sustainable alternative as solely 9 mg of the lanthanum-containing precursor (La(NO 3 ) 3 ) is required to obtain 484 mg of cathode material. In addition, the "3D NVP nanofiber" strategy also results in an environmentally viable alternative. Finally, "N-doped carbon NVP" "N,Bdoped carbon NVP" and specially "nanoplatelet NVP" bear the largest environmental burdens.
Due to availability and price issues, certain materials applied into conventional LIBs (cobalt, lithium, graphite) are considered as CRMs by the European Union. As one of the most critical impact in the energy storage field, the "mineral resource scarcity" category is further discussed. [51] Low values in this category reflect an enhanced potential availability of the materials needed to fabricate Na 3 V 2 (PO 4 ) 3 cathodes, smoothing possible supply chain issues associated with key raw materials. In this sense, cathodes using abundant (sodium carbonate) or organic (oxalic acid, a biopolymer, citric acid) precursors, "hierarchical carbon NVP" and "glucoammam NVP" more precisely, show the lowest results with values of 2.34 and 2.43 kg Cu-equiv., respectively. In contrast, the "N-doped carbon NVP" and "N,B-doped carbon NVP" present the largest burdens (5.12 and 8.30 kg Cu-equiv., respectively). These results are explained by the relatively large quantities of reagents having vanadium (such as NH 4 VO 3 , V 2 O 5 ), boron (NH 4 HB 4 O 7 ), or phosphorus (NH 4 H 2 PO 4 , NaH 2 PO 4 ), highlighting the notable contribution of doping processes to the "mineral resource scarcity" impact category.
Global warming potential (kg·CO 2 -equiv.)  Increasing the solution concentration could be one of the possibilities to reduce the resulting environmental impacts. This particularly applies to certain processes, where the generated wastewater strongly contributes to toxicity estimations. For instance, 1162 g of H 2 O-waste and 1810 g of DMF-waste are generated per gram of "Nanoplatelet NVP" cathode. In addition, we encourage pursuing shorter reactions as these often require large amounts of energy (to power the furnaces) and large quantities of gases such as argon, nitrogen, or hydrogen (to ensure inert atmospheres). Cathode fabrication processes should be also designed to obtain high atom-efficiency values, which represent one of the cornerstones of the 12 Green Chemistry principles. [52] In particular, the process efficiency regarding material use (the ratio of the targeted products to the total mass of used products) [53] should be as high as possible. This could be accomplished by carefully selecting the reaction stoichiometry and avoiding excess reactant from one side and the recycling of the solvents and other reagents from the other.

Environmental Impacts and Electrochemical Performance
As the energy storage is a performance-driven field, the environmental impacts cannot be fully understood without the analysis of the electrochemical behavior. Accordingly, LCA studies have been completed considering the Na 3 V 2 (PO 4 ) 3 /Na half-cell configuration. Although full-cell configurations (i.e., NVP|| carbon) result closer to practical implementation, most of the works have focused on half cells as it provides further information on the working electrode (NVP cathode in this case). Thereby, the discharge capacity and gravimetric energy density are considered for the half-cell configuration. In fact, the impacts are normalized to 1 kWh of (cathode-only) storage capacity defined as Gravimetric energy density ðWh · kg À1 Þ ¼ nominal voltage ½V Â discharge capacity ½Ah cathode weight½kg (1) Given that the information provides by the published works does not provide enough information on cell weight, 1 kWh of cathode-only storage capacity is considered as the FU. This standardization allows considering the electrochemical performance for environmental impact analysis and avoids uncertainties associated with the lack of information (the exact amount of cathode material is provided in the form of mass loading and cell size). The gravimetric energy density values have been extracted from the published works, while a C-rate of 0.5C (a 2 h discharge) has been selected for all the batteries to provide a common ground for comparison. For the sake of clarity, characteristic galvanostatic charge-discharge curves of Na 3 V 2 (PO 4 ) 3 /Na half cells are shown in Figure 3a. [54] Generally, the Na 3 V 2 (PO 4 ) 3 Table 2. Environmental impacts for Na 3 V 2 (PO 4 ) 3 cathode fabrication considering 1 kg of material as FU. The Na 3 V 2 (PO 4 ) 3 cathode code is: 1: "hierarchical carbon-NVP" 2: "rGO-LbL NVP" 3: "μPorous NVP" 4: "N-doped carbon NVP" 5: "N,B-doped carbon/NVP" 6: "La 3þ -doped NVP" 7: "3D NVP nanofiber" 8: "Nanoplatelet NVP" 9:"Agarose carbon NVP" and 10: "Glucomannan NVP". www.advancedsciencenews.com www.advenergysustres.com cathode presents a flat plateau centered at nearly 3.4 V versus Na þ /Na during sodiation/desodiation processes (Na 3 V 2 (PO 4 ) 3 ↔ NaV 2 (PO 4 ) 3 reaction), [16] which coupled with a discharge capacity of up to 117.6 mAh g À1 (theoretical value) [55] renders gravimetric energy densities exceeding 300 Wh kg cathode À1 .
With GWP values ranging from 539.8 to 1622.1 kg CO 2equiv. kWh cathode À1 (see Figure 3b, the full impact analysis is provided in Table S24, Supporting Information), the lowest carbon footprint is obtained by the "μPorous NVP." The "3D NVP nanofiber" cathode also scores low (696.6 kg CO 2equiv. kWh cathode À1 ), while the "La 3þ -doped NVP" and "Nanoplatelet NVP" present the largest impact (1622.1 and 6137.3 kg CO 2 -equiv. kWh cathode À1 , respectively). Comparing these results with those previously shown for 1 kg of cathode, it remains apparent that the environmentally friendlier choice by weight does not necessarily translate into the most adequate design when considering battery performance. For example, although the GWP impacts of "hierarchical carbon-NVP" "rGO-LbL NVP" and "μPorous NVP" are roughly equal in terms of mass, the greater energy density and cathode mass loading can halve the GWP value, from 1102.5 kg CO 2 -equiv. kWh cathode À1 of the "hierarchical carbon-NVP" cathode to 539.8 kg CO 2equiv. kWh cathode À1 of the "μPorous NVP" cathode when considering 1 kWh of storage capacity as FU.
Considering the notably 13-45 larger CO 2 footprint of the cathodes here analyzed in comparison with the impacts originating from the industrial-scale fabrication of Na 3 V 2 (PO 4 ) 3 cathodes, [46] and the fact that the discharge capacity differences during the initial charge/discharge cycles are below 20% in most of the cases (see "cycle capacity" column in Table 3), the following dilemma appears: how worthwhile results, environmentally speaking, are improving the electrochemical performance at the expense of multistep, energy-consuming, and toxic processes? Nowadays this question remains open to debate, although future work may help to clarify this matter. As recently considered by Prozio and Scown, [39] performing LCA studies focusing on the battery use phase is highly recommended, where parameters such as cycle capacity, rate capacity, Coulombic efficiency, or operation lifespan should be considered. The cathodes containing graphene ("hierarchical carbon-NVP" and "rGO-LbL NVP") are among those showing the largest discharge capacity values at 1C, so graphene results in a suitable additive to provide superior electronic conductivity to the cathode and enhance the rate capability (which is consistent with the literature). [56,57] Moreover, the "nanoplatelet NVP" cathode delivers 76.5 mAh g À1 at 10C, meaning that it will only take 36 s to charge/discharge the battery. So this design, in spite of its large environmental impacts, may be useful and environmentally efficient for fast charging electric buses.
In particular, capacity retention should be considered as one of the primary drivers toward sustainability (both economic and environmental) because it ensures the reuse strategy, the tightest loop in the circular economy diagram. The right column in Table 3 summarizes the capacity retention (%) after a given number of charge/discharge cycles. The comparison however is complex because of the different C-rates used. Overall, the majority of the cathodes show large capacity retention values, with specially remarkable values for "rGO-LbL NVP," "N-doped carbon NVP" "N,B-doped carbon/NVP," "3D NVP nanofiber" and "Nanoplatelet NVP", which keep more than the %80% of their initial discharge capacity after more than 5000 cycles at very high C rates. These cathodes are encouraged for longlasting applications where conventional Na 3 V 2 (PO 4 ) 3 cathodes fail. Conversely, the "μPorous NVP" cathode seems to have the poorest cyclability, so in spite of its lower GHG emission per 1 kWh, it could bear larger environmental burdens when considering use phase. Conventional cathodes fail in these two specific applications (long operation lifespan and high charge), so the increased environmental footprint may be offset as conventional Na 3 V 2 (PO 4 ) 3 cathodes can hardly function beyond 100 cycles at 2-3C rate. [58] In any case, it should be also considered that these values remain notably above when comparing with the GWP values (for the whole cell) obtained for other electrochemical energy storage systems such as lithium-O 2 batteries (average of 55.8 kg CO 2 -equiv. kWh À1 ), [59] lithium-sulfur (average of 127.4 kg CO 2 -equiv. kWh À1 ), [60] and LIBs (average of 120 kg CO 2 -equiv. kWh À1 ). [61] In addition, Na 3 V 2 (PO 4 ) 3 /Na half cells also surpass the 140.3 kg CO 2equiv. kWh À1 reported by Peters et al. for batteries composed of a hard carbon anode, Na 1.1 Ni 0.3 Mn 0.5 Mg 0.05 Ti 0.05 O 2 cathode, an organic solvent with NaPF 6 electrolyte, and a polyethylene/ polypropylene separator. [12] The prime cause for these larger impacts is that the laboratory-scale synthetic processes here studied have been not optimized in regard with their environmental performance, requiring both large amounts of reagents and energy. In addition, certain cathodes present remarkable capacity retention values, a critical aspect as it directly affects the durability of the batteries. In times when enduring goods can keep the material and energy resources in the loop for longer, the search for cathodes combining low GHG emissions, high energy density, and durability is encouraged.

Sensitivity Analysis
Overall, the synthetic approaches here analyzed have not been designed following green chemistry principles. Particularly relevant is the contribution of electricity to diverse environmental impacts including GWP, air pollution, water pollution, or solid waste disposal. In fact, the manufacture energy requirements have been highlighted as a major cause of GWP in battery www.advancedsciencenews.com www.advenergysustres.com cathode production. [44] To evaluate the potential for the environmental impact reduction of the proposed cathodes, a sensitivity analysis is performed by shifting from a standard energy mix (high voltage | electricity, high voltage | Cutoff, U -ENTSO-E) to a fully renewable electric power supply. A new electricity source has been modeled considering the current renewable source proportion from Germany, Denmark, and the Netherlands and further modifying the ecoinvent 3.8 electricity mix to convert it to 100% renewable (76% wind, 12% biomass, 7% biogas, and 5% hydro-power), so representative values over the European renewable energy mix are reached (see Figure S11, Supporting Information). [62] The "μPorous NVP" cathode is selected as a representative design comprising common synthetic steps (ultrasonic treatment, stirring, evaporation, drying, grounding, heat treatment, and annealing) and average energy and material inputs. As summarized in Figure 4, when considering 1 kWh of storage capacity as an FU, the environmental impacts could be reduced by 16.9 to 38.0% depending on the category, where the GWP value is lowered from 539.8 to 341.7 kg CO 2 -equiv. kWh cathode À1 . However, renewable energy is time dependent, so a realistic scenario may give intermediate results. [63] However, it remains clear that additional efforts should be carried out to optimize reaction conditions (temperature and time) so the energy consumption could be reduced and environmentally sound Na 3 V 2 (PO 4 ) 3 cathode fabrication processes can be established.

Conclusion
The cradle-to-gate environmental impacts originating from the fabrication of Na 3 V 2 (PO 4 ) 3 cathodes are quantified, analyzed, and compared using LCA. Ten different cathode designs aimed to enhance the energy density, operation lifespan, and rate capability of sodium-ion batteries are selected to provide a representative overview in the sodium-ion battery landscape. The analysis of laboratory-scale batteries provides insights during the earlydesign step and subsequent practical implementation for future research on environmentally friendlier cathodes in particular and sodium-ion batteries in general. The impacts of 1 kg cathode production are accounted for 18 indicators. Global warming values from 423.9 to 1380.0 kg CO 2 -equiv. kg cathode À1 are obtained, where cathodes using biopolymer precursors or those incorporating carbonaceous structures bear the lowest impacts. On the contrary, doping approaches present larger impacts. To get the bigger picture, the analysis has been expanded to Na 3 V 2 (PO 4 ) 3 /Na half cells. When considering the impacts per 1 kWh of cathode-only storage capacity, the global warming contribution was found to be 539.8-1622.1 kg CO 2equiv. kWh cathode À1 . A sensitivity analysis demonstrates the potential to reduce the environmental impacts by 16.9-38.0% by transitioning to a renewable energy mix. According to the electrochemical performance in Na 3 V 2 (PO 4 ) 3 /Na half-cell configuration, the "μPorous NVP" cathode shows the lowest carbon footprint per 1 kWh of storage capacity. However, when considering the use phase, heteroatom doping and morphology modification are good alternatives. However, considering the larger impacts of the cathodes here analyzed against industrial Na 3 V 2 (PO 4 ) 3 cathodes, the use of electrochemically enhanced cathodes is recommended for specific cases requiring long operation lifespans or fast charging applications. In the future, LCA studies could be accompanied by life-cycle costing analyses to evaluate the economics of battery cathodes and shed further light on the industrial feasibility of sodium-ion batteries. Overall, this www.advancedsciencenews.com www.advenergysustres.com work highlights that not only material weight or storage capacity but also further operating features (C-rate, lifespan, and discharge capacity) should be considered to design truly environmentally sustainable batteries. Table 1 summarizes the material and energy input inventory obtained for secondary data. The Supporting Information provides additional flowcharts for each cathode, together with a brief explanation to guide the reader. The materials and energy inventory modeling according to the ecoinvent v3.8. database is also provided. Additional calculations conducted to evaluate the environmental impacts of the cathodes into Na 3 V 2 (PO 4 ) 3 /Na cells are available from the corresponding author upon reasonable request.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.