On the Challenges to Develop Hybrid Faradaic-Capacitive Electrodes Incorporating a Sacrificial Salt for Lithium-ion Capacitors: The Case of Li3V1.95Ni0.05(PO4)3-AC-Li2C4O4

On the Challenges to Develop Hybrid Faradaic-Capacitive Electrodes Incorporating a Sacrificial Salt for Lithium-ion Capacitors: The Case of Li3V1.95Ni0.05(PO4)3-AC-Li2C4O4 Miguel Granados-Moreno,[a, b] Maria Arnaiz,[a] Emanuele Gucciardi,[a] Nahom Enkubahri Asres,[a, b] Eider Goikolea,[b] and Jon Ajuria*[a] The low capacity of activated carbon (AC) electrodes remains as one of the major limiting factors for the development of high energy density lithium-ion capacitors (LICs). Hybridization of capacitive AC electrodes by incorporating faradaic materials into the electrode formulation could be performed to enhance the capacity of the overall device. However, this strategy requires an accurate electrode design to maximize the performance. In this work, Li3V1.95Ni0.05(PO4)3(LVNP) was selected as faradaic material due to its compatibility with AC, showing high capacity, fast ionic diffusion, and relatively high conductivity. Various formulations and mass loadings have been studied to analyze the impact of incorporating LVNP into the positive electrode on the performance of the hybrid electrode. Moreover, for practical LIC applications, a sacrificial salt -dilithium squarate, Li2C4O4was included in the hybrid electrode as a prelithiation additive, developing a ternary electrode. The sacrificial salt oxidized releasing lithium ions, while the electrochemical performance of the hybrid positive electrode remained almost unaltered. Finally, a cycle life test combined with a post-mortem analysis allows understanding the failure mechanisms of the electrode, suggesting the need of further improvements of the electrolyte and electrode-electrolyte interface to develop long lifetime hybrid faradaic-capacitive electrodes based on LVNP-AC active materials. 1. Introduction The high dependence of the modern industrial society on the use of fossil fuels is the main reason for the increase of CO2 emissions and the accelerated climate change. An energy transition towards a net zero emission society relying on renewables is a priority challenge that civilization is facing today.[1,2] The electrification of our energy ecosystem needs efficient and robust energy storage systems (ESSs) capable of assuming multiple key roles within the next few years. On the one hand, ESSs will become a cornerstone in the expansion of renewable energies, providing large-scale energy storage and stability to the electrical grid. On the other hand, ESSs will be critical enablers of the portable use of energy, which is essential for consumer electronics and electric vehicles. Among the different energy storage technologies, lithiumion batteries (LIBs) are the most extended electrochemical ESSs, and they are currently leading the revolution toward electric mobility. LIBs are the preferred systems due to their high energy density provided by faradaic reactions in the electrode. However, their low power density requires non-cost-effective solutions, such as the use of oversized batteries or externally hybridized LIBs and electrochemical capacitors, which require complex battery management systems (BMSs). Unlike LIBs, electrochemical capacitors are characterized by their high power density but low energy density. In order to bridge the energy-power gap between LIBs and electrochemical capacitors, a hybrid device combining a battery-type electrode and a capacitor-type electrode was first proposed in 2001.[3] That hybrid device, called lithium-ion capacitor (LIC), triggered the development of metal-ion capacitor (MIC) technology. Currently, LICs are drawing the attention of both academia and industry, with an exponentially increasing number of yearly published articles as well as new start-up companies.[4] Despite the progress and the development degree achieved to date, LICs present unresolved challenges that hinder their commercial expansion and limit their market share. The most important one is to increase the energy to values close to those offered by power batteries.[5] In the early years, the research focused on the development and improvement of the faradaic negative electrode. Initially, in the first LICs, negative electrodes were crafted from graphite.[6,7] This material continues to be extensively employed in both commercial and research cells, thanks to its well-established attributes such as high capacity, chemical stability, and mechanical robustness. However, its rising cost, the supply deficit for natural graphite feedstock, the [a] M. Granados-Moreno, M. Arnaiz0000-0001-8800-0643, E. Gucciardi, N. Enkubahri Asres, J. Ajuria Centre for Cooperative Research on Alternative Energies (CIC energiGUNE), Basque Research and Technology Alliance (BRTA), Alava Technology Park, Albert Einstein 48, 01510 Vitoria-Gasteiz, Spain E-mail: [email protected] [email protected] Homepage: 0000-0001-8800-0643 [b] M. Granados-Moreno, N. Enkubahri Asres, E. Goikolea Department of Organic and Inorganic Chemistry, Faculty of Science and Technology, University of the Basque Country UPV/EHU, 48940 Leioa, Spain Supporting information for this article is available on the WWW under https://doi.org/10.1002/celc.202400117 © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 29/41] 1 ChemElectroChem 2024,11, e202400117 (1 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem www.chemelectrochem.org Research Article doi.org/10.1002/celc.202400117 low power response and the lithium plating risk, prompted the search of alternative candidates. Other carbonaceous materials have also been proposed as negative electrode, such as hard carbon (HC),[8,9] soft carbon (SC)[10] and graphene,[11] among others. In general, those alternative-to-graphite carbons present lower capacity, providing slightly lower energy density, but offering better rate capability, what overall, provides considerably higher power density. Beyond carbons, metal oxide based electrodes such as Li4Ti5O12 (LTO)[12] or Fe2O3,[13] with fast kinetics have also been widely studied. However, their high potential plateau, usually above 1 V vs. Li+/Li, limits the output voltage of the final cell. Last, several metalloids and metals such as Si or Sn[14–16] that operate through alloying/dealloying reactions have caught the interest of researchers owning to their very high specific capacity (ca. 994 mAhg1or 3579 mAhg1for Sn and Si respectively[17,18]). Unfortunately, these materials suffer from large volumetric changes upon lithiation and delithiation processes, presenting a poor solid electrolyte interphase (SEI) stability and mechanical fractures, and consequently, requiring composite electrodes[19,20] or nanostructuring[15,21] to improve the lifetime of the cells, among others. Intermetallic compounds have been developed as a strategy to increase the stability of those metals, maintaining very high capacity output as in the case of TiSb2 [22,23] or Sn4P3,[24,26] although lifetime is still limited. Yet, despite all the research on the search of advanced materials, carbon is the first option when considering market available products.[27] Regarding the positive electrode, activated carbon (AC) has been the most widely used material, owing to its excellent high-rate performance as well as long cyclability. AC has been widely studied and improved: from inexpensive and ecofriendly biowaste derived ACs to superfast and highly stable nanostructured ACs.[28,29] Functionalization is another strategy that can be used to improve the conductivity and capacity of the AC electrodes.[30,32] However, despite the exceptional features of ACs, their low capacity of ca. 40–60 mAhg1, limited by its capacitive nature, is a major drawback in terms of energy density. The total or partial replacement of ACs by faradaic materials in the positive electrode has been studied as an alternative to increase their capacity and the overall energy density of LICs.[33,36] Fast and well-known faradaic materials such as LiFePO4 [19,37] or LiNixMnyCo1-x-yO2 [38] have been combined with AC to fabricate composite positive electrodes in LICs. Although the hybrid faradaic-capacitive materials show 3–5 times greater capacities with respect to ACs, the capacity retention at high current densities is, in general, lower and the cycle life of devices is also negatively affected. Li2V1.95Ni0.05(PO)4(LVNP) was developed by Secchiaroli et al.,[39,40] as a high-capacity fast insertion/deinsertion faradaic material. The Ni doping increased the capacity with respect to Li2V2(PO)4(LVP), while maintaining the high electronic conductivity and ionic diffusion. The excellent properties and suitable electrochemical stability window (ESW) make it an interesting choice as positive electrode material for LICs. Further investigations were focused on the use of a binary LVNP-AC electrode, taking advantage of both the high capacity of LVNP and the high rate capability of the AC.[36] The composite electrode successfully improves capacity retention, increasing both the gravimetric and volumetric capacity at high current density. The excellent properties of LVNP pushed the development of full devices constituted by LTO-AC negative electrodes and LVNP or LVNP-AC positive electrode.[41] The devices showed high energy density, high power, and long cycle life, highlighting the viability of using LVNP-based positive electrodes also in LICs. However, the development of real life LICs requires a prelithiation step to: i) compensate lithium losses during the first cycle irreversibility and SEI formation; and ii) adequate the cell potential to maximize the energy density output. Currently, industrial pre-lithiation strategies are based on the use of metallic lithium.[4] Nevertheless, metallic lithium compromises the safety of the technology and requires an inert atmosphere environment during the processing of Li, which increases the production price. Organic lithium sacrificial salts are a cheap, safe and air stable alternative that has been studied for carbonaceous electrodes.[42,43] The viability of using non-conductive sacrificial salts in electrodes containing faradaic materials has been evaluated and validated, with most of the sacrificial salt being decomposed during the first cycle. Thus, dilithium squarate (Li2C4O4)[44] was selected as a pre-lithiation additive and included in the formulation of the positive electrode, developing for the first time -to the best of our knowledge-, a ternary hybrid capacitive-faradaic electrode including a pre-lithiation agent. Previous results disclose the potential of LVNP for the development of high energy and high power devices. However, from fundamental studies towards prototypes, electrodes must be adapted to meet different standards and requirements. Besides technical aspects such as gravimetric or volumetric capacity, the price or scalability of the processes involved in the fabrication also need to be considered. In this work, our objective is the development of a realistic proof-of-concept, optimizing the electrode formulation and including a sacrificial salt as an effective pre-lithiation strategy. First, LVNP electrodes are optimized to understand its performance limitations. Second, LVNP is incorporated on an AC electrode to develop a hybrid LVNP-AC electrode. Third, the sacrificial salt (Li2C4O4, -Li) is incorporated, developing a ternary LVNP-AC-Li electrode. Finally, a proof-of concept LIC is demonstrated facing the LVNPAC-Li positive electrode with a negative HC electrode. This work aims to bridge fundamental science and technology, clarifying the critical steps of the optimization and adaptation process, from fundamental studies towards the starting point of prototypes development. The constraints of prototype cells push electrodes towards their limits, revealing blind spots that might be ignored at lab-scale, but are fundamental for their suitability in real life applications. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 30/41] 1 ChemElectroChem 2024,11, e202400117 (2 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Experimental Synthesis of Li3V1.95Ni0.05(PO4)3and Li2C4O4 Li3V1.95Ni0.05(PO4)3(LVNP) was synthesized by ball-milling and carbon-thermal reduction methods. First, stoichiometric amounts of lithium carbonate (>99.0%, Sigma-Aldrich), ammonium metavanadate (99.99%, Sigma-Aldrich), ammonium dihydrogen phosphate (99.999%, Sigma-Aldrich) and nickel(II) nitrate hexahydrate (99.999%, Sigma-Aldrich) were mixed with 40 ml of ethanol. Subsequently, poly(acrylic acid) (Sigma-Aldrich) and D-(+)-glucose (>99.5%, Sigma-Aldrich) were introduced as carbon sources in a 4:1 mass ratio, respectively. The mixture was transferred to an agate jar with agate balls of 1 cm diameter in a 20:1 ball to sample mass ratio. The mixture was ball milled in a Pulverissete 5 planetary milling at 350 rpm for 8 h. Then, the mixture was transferred to a rotary evaporator at 60°C for 30 min to remove the solvent. The obtained powder was manually grinded in an agate mortar for 1 h to reduce the particle size and ensure the close contact between the precursors. Afterwards, the precursors were thermally treated in two steps: the first one at 350°C for 5 h, and the second one at 800°C for 8 h, both steps were performed under dynamic Ar atmosphere. Between the first and the second annealing, the powder was manually grinded for 30 minutes.[40] Dilithium squarate (Li2C4O4) was synthesized from 3,4-dihydroxy-3cyclobutene-1,2-dione (99.0%, Merck) and lithium carbonate (> 99.0%, Merck). Within a 1:1 molar stoichiometric relation, each precursor is dissolved in the required deionized water. Once both are well-dissolved, Li2CO3solution is added slowly into one of squaric acid and stirred overnight. Afterwards, the mixture was transferred to a rotary evaporator at 50°C to remove the water. The Li2C4O4was recovered and dried at 120°C under vacuum overnight before being used.[44] Physicochemical Characterization The morphology and composition of the samples was characterized by scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX), respectively, using a Quanta200 FEI (3 kV, 30 kV) microscope. Both Everhart-Thornley (ETD) and backscattered electron (BSED) detectors were used. For a better understanding of the electrode morphology, top view and cross section SEM images were recorded. Cross section samples were prepared using a Hitachi 4000 Plus ion milling which utilizes an Ar+ion beam (0– 6 kV acceleration voltage) milling method. Textural properties of active materials were determined by nitrogen adsorption/desorption isotherms. Isotherms were registered at 196 °C using an ASAP 2460 instrument from Micromeritics. The samples were outgassed at 250°C for 12 h under vacuum prior to the analysis. Specific surface areas (SSA) were calculated from the Brunauer–Emmett– Teller (BET) equation using the Rouquerol procedure for the monolayer capacity.[45] Pore size distributions (PSD) were calculated using SAIEUS software, applying the 2D-NLDFT model to the adsorption branches data.[46] The elemental composition of LVNP-based electrodes was determined by inductively coupled plasma optical emission spectroscopy (ICP-EOS) using a 5800 Agilent ICP-OES. The following operating conditions were selected: 1.20 kW of RF power, 12.0 L min1of plasma-gas flow and 1.00 L min1of auxiliary flow. Solutions were introduced into the plasma torch using a concentric glass nebulizer and a double pass spray chamber at a flow rate of 0.70 L min1. The wavelengths employed in the ICP-OES analysis are 214.914 nm for phosphorous, 231.604 nm for nickel and 309.310 nm for vanadium determination. Different wavelengths for each single element were also tested to ensure no spectral interferences are affecting the results. Electrochemical Characterization Electrodes were prepared by dispersing the LVNP and/or AC (Norit) active materials, Super C65 carbon as a conductive material (Imerys, C-NERGY) and polyvinylidene fluoride (PVdF, Solef) as a binder in Nmethyl-2-pyrrolidinone (NMP, Merck). Dilithium squarate (Li2C4O4) sacrificial salt was included in the electrode formulation as the prelithiation additive. The dispersion was vigorously stirred, and the NMP-based slurries coated onto aluminum foil. Laminates were dried at 80°C for 12 h under vacuum. Electrode discs of 12 mm were further punched out of the laminates and dried at 120 °C overnight under vacuum prior to cell assembly. Different electrode formulations were used: i) LVNP:C65:PVdF with a mass ratio of 80:10:10, 90:5: 5 and 90: 7:3; ii) LVNP:AC:C65:PVdF with a mass ratio of 45: 45 :5:5; and, iii) LVNP:AC:Li2C4O4:C65:PVdF with a mass ratio of 27.5: 27.5 :35:5:5. Electrode formulations are summarized in Table 1. Following the same procedure, HC carbon electrodes have been fabricated using 90 % of HC (Kuraray), 5 % Super C65 (Imerys, CNERGY) and 5 % PVdF (Solef). In the case of HC electrodes, the NMP-based slurries were coated in cupper foil. Electrochemical performance of LVNP-based electrodes was evaluated by using 3-electrode Swagelok cells in half-cell configuration. Electrodes of ca. 2.5 mg cm2mass loading were used unless otherwise stated. Oversized self-standing AC electrodes of >25 mg cm2mass loading and metallic lithium discs of 10 mm in diameter were used as counter and reference electrode, respectively. Whatman D-type glass fiber discs of 13 mm in diameter were selected as separators. 1 M LiPF6in EC : DMC (ethylene carbonate : dimethyl carbonate, 50: 50 vol.) (99%, Sigma-Aldrich) was used as electrolyte. Galvanostatic charge-discharge (GCD) and cyclic voltammetry (CV) measurements between 3–4.3 V vs. Li+/Li were performed in a VMP3 generator from Biologic. In GCD, during both charge and discharge the same current density is applied with a 5 s hold step but no constant current constant voltage (CCCV) step[47,48] is included. Dilithium squarate sacrificial salt is decomposed by cycling the electrode at CLi2C4O4/10 for 10 cycles, being CLi2C4O4 = 425 mAhg1. Both the applied current density (A g1) and the calculated specific capacity (mA h g1) are presented per mass of active material in the electrodes unless another metric is indicated. Table 1. Summary of the different electrode formulations used in this work. Sample LVNP (%) C65 (%) PVdF (%) AC (%) Li2C4O4 (%) LVNP80:10:10 80 10 10 – – LVNP-90:5:5 90 5 5 – – LVNP-90:7:3 90 7 3 – – LVNP-AC 45 5 5 45 – LVNP-AC-Li 27.5 5 5 27.5 35 Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 31/41] 1 ChemElectroChem 2024,11, e202400117 (3 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2. Results 2.1. Electrode Formulation and Mass Loading Electrodes are usually composed of three main components: active material, conductive additive and binder. The ratio between those three components determines many of the overall properties of the electrode and impacts the final device performance. The conductive additive interconnects active material particles, increasing the overall electrode conductivity and it also helps to homogeneously distribute the binder through the electrode. At last, the binder ensures the mechanical properties of the electrode, thus, an insufficient amount of it could lead to electrode fractures and delamination, but an excess will affect conductivity and specific capacity since binders are usually insulating non-electrochemically active polymeric materials. The active material is the main player, storing charge by faradaic, pseudocapacitive and/or capacitive mechanisms. Therefore, it is of utmost importance to properly design the electrode formulation to allow a high mass loading of the active material without penalizing the electrochemical performance and mechanical properties. LVNP electrodes reported in literature were mostly fabricated containing 80% of active material, 10% of conductive additive and 10% of PVdF binder (named as LVNP-80:10:10). Thus, some optimization is necessary to develop a final device. With the objective of increasing the content of the active material in the electrode, a new formulation with 90% of active material, 5% of C65 and 5% of PVdF (LVNP-90:5:5) was developed. Galvanostatic charge-discharge (GCD) curves of LVNP80:10:10, LVNP-90:5:5 electrodes are shown in Figure 1. It is worth mentioning that no constant current constant voltage (CCCV)[47,48] step is included neither during charge nor discharge, since LICs are intended to operate under constant current only. During charge and discharge, LVNP undergoes three Li+insertion/deinsertion reactions as stated in Equation (1) to (3).[39] The reactions correspond to the three plateaus observed in Figure 1a. at 3.6, 3.7 and 4.11 V vs. Li+/Li. As expected, reducing the amount of C65 has a detrimental effect on the electrode performance due to its higher resistivity, especially at high current density, as proved when comparing LVNP-80:10:10 and LVNP-90 :5:5. Under the same applied current density, the discharge time of LVNP-90:5: 5 is reduced, accounting for less capacity, and the overpotential almost tripled the one of the reference LVNP-80:10: 10, as reported in Table 2. The overpotential is defined as the additional potential beyond the thermodynamic requirements needed to drive a reaction.[49] Low conductivity faradaic materials tend to show high overpotential, which promotes active material degradation and leads to low energy efficiency if incorporated in a full cell. This is a key point to consider since low energy efficiency means that energy losses are transformed into heat, requiring more complex and expensive heat management systems. Li3V1:95Ni0:05ðPO4Þ3$Li2:5V1:95Ni0:05ðPO4Þ3 þ0:5Liþat 3:6V vs:Liþ=Li (1) Li2:5V1:95Ni0:05ðPO4Þ3$Li2V1:95Ni0:05ðPO4Þ3þ 0:5Liþat 3:7V vs:Liþ=Li (2) Li2V1:95Ni0:05ðPO4Þ3$Li1V1:95Ni0:05ðPO4Þ3þ 1Liþat 4:11 V vs:Liþ=Li (3) In view of this result, an advanced formulation was developed in order to maintain the promising performance shown by LVNP when formulated with high content of conducting agent and binder. To this aim, the amount of binder was reduced while the amount of C65 was increased to reach a formulation consisting of LVNP-90:7: 3. GCD of LVNP-90:7:3 electrodes, observed in Figure 1, shows intermediate behavior Figure 1. Charge/discharge curves of LVNP-80:10:10 (red), LVNP-90:5: 5 (black) and, LVNP-90:7 : 3 (blue) electrodes at (a) 0.5 A g1, (b) 1 A g1and (c) 3 A g1. Table 2. Overpotential of LVNP-80:10:10, LVNP-90 :5:5 and LVNP-90:7: 3 electrodes at 1 A g1and 3 A g1current densities calculated by the potential difference existing between the upper plateau during charge and discharge. Formulation (LVNP:C65:PVdF) Overpotential 1 A g13 A g1 80:10:10 0.041 V 0.081 V 90:5:5 0.11 V 0.285 V 90:7:3 0.05 V 0.114 V Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 32/41] 1 ChemElectroChem 2024,11, e202400117 (4 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License between LVNP-80:10:10 and LVNP-90 :5:5, with relatively good capacity retention at high current density, and similar overpotential to LVNP-80:10:10. Figure 2a reports the gravimetric capacity of the three different formulations per mass of LVNP. The optimization of the electrode formulation is not trivial as can be observed in Figure 2; LVNP-80 :10:10 displays higher capacity than LVNP90:5:5 and LVNP-90:7 :3 and has an excellent rate capability. In Figure 2b, the gravimetric capacity is reported per mass of electrode, considering the weight of the total electrode coating (coat), including the conducting agent and the binder and excluding the current collector. The metric reported in Figure 2b standardizes the capacity of the three electrodes, enabling an accurate comparison between the three formulations. In that case, LVNP-90:7: 3 clearly outperforms LVNP80:10:10 due to the higher ratio of active material in the total mass, which increases the specific capacity of the cell. Thus, when reporting with respect to merely LVNP, capacity is overestimated as the reported metric neglects the weight of the C65 and PVdF. Figure 2c reports the volumetric capacity of the electrode coating in mAh cmcoat3. In volumetric terms, the low density of C65 and PVdF compared with LVNP severely affects the capacity of the electrode, thus, higher amount of LVNP enhances the electrochemical performance. Overall, LVNP-90:7:3 electrodes have shown the best performance, with high-capacity values of ca. 82 mAh gAM1, 79 mAh gcoat1and 31 mA cmelectrode3, good capacity retention and low overpotential of 0.05–0.114 V at 1–3 A g1, reaching an optimum balance between the active material, the conductive additive and the binder. Next, the impact of the mass loading of the electrodes was studied. It is expected that higher loading electrodes will suffer from increased internal resistance and slower ion diffusion through the bulk electrode. Despite the good performance of LVNP-90:7:3, a formulation containing only 3% of PVdF did not show the required mechanical stability for the fabrication of high mass loading electrodes, which collapsed during the cutting step. Then, LVNP-90:5: 5 electrodes (i.e., with higher amount of binder, 5%) were developed instead. Mass loadings of 2.5, 4.5, 6 and 9 mg cm2were fabricated to evaluate the influence of the mass loading in the electrochemical performance. GCD curves of LVNP-90 :5:5 electrodes of 2.5, 4.5 and 6 mg cm2are shown in Figure 3. The effect of the mass loading is almost negligible at low current density, while at high current density, the larger diffusion times and resistance, increased the overpotential of the electrodes as shown in Table S1. When using 6 mg cm2electrodes, the second Li+insertion reaction occurring in the third plateau could not take place below 4.3 V vs Li+/Li, which is considered to be the potential limit to avoid electrolyte degradation.[50] In half-cell configuration, the capacity of 6 and 9 mg cm2electrodes dramatically drops as observed in Figure S1. Thus, even if LVNP conductivity and ionic diffusion are remarkably high compared with other faradaic materials,[39,51] it is not enough to operate high mass loading electrodes. Increasing the mass loading of LVNP-based electrodes would require the addition of a constant voltage step when charging the material.[47] This is a critical point since LICs aim to operate in constant current charge-discharge mode only. Instead, as described in the next section, the active material will be modified to enhance the conductivity. 2.2. LVNP-AC Composite Electrode Finally, once LVNP boundaries are well defined and understood, its hybridization with an AC to develop a hybrid faradaiccapacitive electrode as a high energy positive electrode for LICs is envisaged. As first approach, the composite electrode is designed by combining both components in a 1:1 mass ratio with the overall electrode formulation being 45% LVNP, 45% AC, 5% C65 and 5% PVdF (from now on called LVNP-AC), and an average mass loading of 2.5 mgAM cm2. The textural Figure 2. (a) Specific capacity reported respect to the active material (AM) weight; (b) specific capacity reported respect to the electrode coating weight; and (c) volumetric capacity of the electrode coating of LVNP80:10:10 (red), LVNP-90:5:5 (black), and LVNP-90:7:3 (blue) electrodes at different current densities. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 33/41] 1 ChemElectroChem 2024,11, e202400117 (5 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License properties of LVNP powder and LVNP-AC mixture were determined from nitrogen adsorption/desorption isotherms and included in Figure 4a. As expected, the non-porous LVNP poses a very low BET specific surface area (SSA)[40] of 29 m2g1, while the addition of AC increases the BET SSA of the composite up to 586 m2g1. Pore size distributions (PSD) of both samples are Figure 3. Charge/discharge curves of 2.5 (black), 4.5 (green) and 6 (red) mg cm2electrodes at (a) 0.5 A g1, (b) 1 A g1and (c) 3 A g1. Figure 4. (a) Nitrogen adsorption/desorption isotherms and (b) pore size distribution of LVNP and LVNP-AC active materials. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 34/41] 1 ChemElectroChem 2024,11, e202400117 (6 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License depicted in Figure 4b. The LVNP-AC mixture is composed of micropores (<2 nm) and small mesopores (2–50 nm), which are attributed to the AC. Top view and cross section SEM images of LVNP-AC electrodes are presented in Figure 5. The homogeneous distribution of LVNP and AC in the electrode ensures the overall electronic conductivity of the electrode owing to the close contact of low conductive LVNP particles with the highly conductive AC. Moreover, LVNP particles of ca. 1μm are sandwiched in between AC particles, with an average size of 5μm, contributing to decrease the internal resistivity. In order to conduct the electrochemical evaluation of LVNPAC, LVNP and AC electrodes were also evaluated and used as references (all of them with a 90% AM,5% C65 and 5% PVdF formulation). Thus, GCD curves of LVNP-90:5 :5, LVNP-AC and AC electrodes characterized between 3–4.3 V vs. Li+/Li are illustrated in Figures 6a, 6b and 6c. The AC displays the typical linear shape indicative of a capacitive charge storage mechanism, while LVNP -described beforehandshows three welldefined plateaus. The hybrid LVNP-AC electrode clearly shows a combination of both mechanisms: a capacitive linear profile over the whole potential window with slope changes at 3.6, 3.7 and 4.1 V vs. Li+/Li. As the current density increases, the capacitive mechanism of LVNP-AC becomes more and more dominant, and at 3 A g1there are less inflection points in the triangular-shaped curve. Overall, LVNP electrode shows the highest capacity output within all the applied current density range, as shown in Figure 6d. However, the overpotential of the LVNP electrode at 3 A g1is very high, which could lead to shorter cycle life and decrease the energy efficiency when full cells are assembled. On the contrary, LVNP-AC exhibits double the capacity of the AC and the symmetric charge-discharge profile with low internal resistivity, what anticipates high energy efficiency. The same trend is observed in the voltammograms of LVNP, LVNP-AC and AC electrodes depicted in Figure S2 showing the displacement of the redox peaks of LVNP with respect to LVNP-AC. 2.3. Incorporating Dilithium Squarate in the LVNP-AC Electrode as Pre-Lithiation Agent LICs require a pre-lithiation step to compensate for the Li losses during the first cycle, corresponding mainly to the SEI formation in the negative electrode. In addition, the pre-lithitiation step adjusts the cell voltage, lowering the potential of the negative electrode to maximize the capacity and energy density of the full cell.[52] To this aim, dilithium squarate (Li2C4O4) is used, since it has already been validated as pre-lithiation agent combined Figure 5. (a, b) Top view SEM and (c, d) cross-section SEM images of LVNP-AC electrode by using ETD (a, c) and BSED (b, d) detectors. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 35/41] 1 ChemElectroChem 2024,11, e202400117 (7 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License with AC in the positive electrode. However, Li2C4O4is needed in a high amount, between 10–40% of the electrode weight for pre-lithiation, depending on LIC chemistry and cell mass balance. Such an amount of additive might negatively impact the overall performance of the electrode when it incorporates a low conductivity faradaic component. When Li2C4O4sacrificial salt decomposes, it leaves some holes in the electrode that could detach particles from each other, increasing the internal interparticle resistivity of the electrode. When it is incorporated on carbonaceous electrodes, this feature is a minor issue due to the high conductivity of ACs. Nevertheless, as presented in Section 3.1 and 3.2, an increase in the internal resistivity could lead to poor rate performance, detrimental for the use of those electrodes in high power devices. A positive electrode with formulation 27.5% LVNP, 27.5% AC, 35% Li2C4O4, 5% C65 and 5% PVdF was developed (hereafter called LVNP-AC-Li) and its electrochemical response was evaluated. Dilithium squarate decomposes during the first cycles as shown in Figure S3. The capacity delivered during the first cycle corresponds predominantly to the Li2C4O4decomposition (CLi2C4O4 =425 mAhg1), due to most of the Li2C4O4is successfully decomposed during the first cycle. Furthermore, the Li2C4O4decomposition has been confirmed by XRD as illustrated in Figure S4, where the signals related to Li2C4O4 disappear after the first GCD cycle. After Li2C4O4decomposition, the electrode formulation is 42.31% LVNP, 42.31% AC, 7.69% C65 and 7.69% PVdF, which has been used for the capacity calculations. The mass loading of the electrodes was 2.5 mgAM cm2, considering that the active material is the mixture of LVNP and AC. After 10 initial cycles at CLi2C4O4/10 between 3–4.3 V vs Li+/Li, a rate capability test was performed to the electrodes. GCD curves of LVNP-AC and LVNPAC-Li at 0.5, 1 and 3 A g1are illustrated in Figures 7a, 7b and 7c. Both electrodes show similar charge-discharge profiles, despite the incorporation of the sacrificial salt. However, at the highest applied current density, i.e. a discharge time of only 60 s, the overpotential increases up to 0.12 V when the Li2C4O4 is added to the electrode, as observed in Figure 7c. As abovementioned, the addition of Li2C4O4might have induced a detrimental effect on the electronic conductivity of the electrode due to the loss of interparticle contact between LVNP particles or with AC particles. Furthermore, capacity at different current densities is illustrated in Figure 7d. The addition of Li2C4O4results in a higher capacity decrease and slightly higher overpotential at high current density, while increases the capacity and maintains negligible overpotential at low current density. The capacity increase caused by Li2C4O4sacrificial salt has already been observed in carbonaceous electrodes and is now under study.[44] GCD experiments were accompanied by cyclic voltammetry (CV) to better understand the dominant mechanisms. As Figure 6. Charge/discharge curves of LVNP (black), LVNP-AC (red) and AC (blue) electrodes at (a) 0.5 A g1, (b) 1 A g1and (c) 3 A g1. (d) Specific capacity per gram of active material of the three compositions at different current densities. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 36/41] 1 ChemElectroChem 2024,11, e202400117 (8 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License observed in Figure S4, similar to LVNP-AC, LVNP-AC-Li electrodes showed mixed capacitive and faradaic mechanisms, with quadratic shape interrupted by faradaic peaks at 3.6, 3.7 and 4.1 V vs. Li+/Li. LVNP-AC-Li shows wider CV in the 3.5–4.1 V vs. Li+/Li potential window, leading to higher capacities. Thus, it can be concluded that a hybrid positive electrode containing a LVNP-AC composite improves the capacity of the traditional AC electrode. In addition, Li2C4O4was successfully incorporated into the composite electrode formulation, which will be of utmost importance when assembling a LIC cell without a sacrificial Li electrode. As a final validation test, cyclability is shown in Figure 8. The lifetime of LVNP-AC and LVNP-AC-Li electrodes is evaluated between 3–4.3 V vs. Li+/Li at a current density of 1 A g1(i.e. discharge time of ca. 1–2 minutes). As shown in Figure 8a, it results in an unexpected short lifetime of only 550 cycles for both LVNP-AC and LVNP-AC-Li electrodes before they reach 20% loss of initial capacitance. GCD curves of LVNP-AC-Li presented in Figure 8b show progressive reduction of 3.6, 3.7, and 4.1 V vs. Li+/Li insertion/deinsertion semi-plateaus upon cycling, which completely disappear after 800 cycles. The fast capacity decay observed is primarily caused by the decrease of the electrochemical activity of LVNP. In order to shed some light into this unexpected performance as well as the degradation mechanism behind the failure of the LVNP-ACLi electrodes, post-mortem analysis of the electrodes was performed. SEM and EDX characterization of the pristine and cycled electrodes is presented in Figure 8c and 8d and Table S2. SEM images of pristine and cycled LVNP-AC-Li electrodes show a morphology change after cycling that could be related with a phase transition that might impact the electrochemical activity of LVNP. Additionally, EDX measurements were also performed: in the case of pristine electrodes only at the surface while in the case of post-mortem electrodes both the surface and the internal part of the electrode -after scratching the surfacewere analyzed. The vanadium and phosphorus content of the electrodes obtained by EDX is reported in Table S2. Pristine electrodes show similar vanadium and phosphorus amounts, but the contents change after cycling. The surface of the electrodes is enriched in vanadium while the internal part is vanadium poor, suggesting a migration of vanadium to the surface of the electrode. The elemental content of a LVNP-AC-Li electrode cycled for 1000 cycles at 1 A g1has also been characterized by inductively coupled plasma optical emission spectroscopy (ICP-OES). In order to confirm and quantify the amount of vanadium dissolved during the ageing test, a pristine LVNP-AC-Li electrode was also characterized. The amount of P, Ni and V in the samples is shown in Table S3. LVNP-AC-Li pristine sample shows a similar content of P and V, and a small content of Ni, as expected from the stoichiometry of the material. In the case of LVNP-AC-Li cycled, the P and Ni content remains almost unaltered with respect to LVNP-AC-Li pristine. A small difference in the amount of Ni is observed, nevertheless it might be caused by experimental uncertainty since it only Figure 7. Charge/discharge curves of LVNP-AC (red) and LVNP-AC-Li (dark yellow) electrodes at (a) 0.5 A g1, (b) 1 A g1and (c) 3 A g1. (d) Specific capacity per gram of active material of LVNP-AC and LVNP-AC-Li electrodes at different current densities. Wiley VCH Mittwoch, 25.09.2024 2419 / 364655 [S. 37/41] 1 ChemElectroChem 2024,11, e202400117 (9 of 13) © 2024 The Authors. ChemElectroChem published by Wiley-VCH GmbH ChemElectroChem Research Article doi.org/10.1002/celc.202400117 21960216, 2024, 19, Downloaded from https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.202400117 by Universidad Del Pais Vasco, Wiley Online Library on [14/10/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License