Size tuneability of highly efficient li-rich cathode materials using an emulsion-based synthesis route

Size tuneability of highly efficient li-rich cathode materials using an emulsion-based synthesis route Saul Rubio a , Ana M. Beltr´ an a , Cristina Ar´ evalo b , Gerardo T. Martinez c , Francisco J. Garcia-Garcia a , Eva M. P´ erez-Soriano a , Isabel Montealegre-Mel´ endez a , Juan G. Lozano a,* a Departamento de Ingeniería y Ciencia de los Materiales y del Transporte, Escuela Polit´ ecnica Superior, Universidad de Sevilla, 41011, Sevilla, Spain b Departamento de Ingeniería y Ciencia de los Materiales y del Transporte, Escuela T´ ecnica Superior de Ingeniería, Universidad de Sevilla, 41092, Sevilla, Spain c Department of Materials, University of Oxford, OX1 3PH, Oxford, United Kingdom GRAPHICAL ABSTRACT ARTICLE INFO Keywords: Li-rich cathodes Li 1.2 Ni 0.2 Mn 0.6 O 2 organic route cathodes size control electrochemistry ABSTRACT Lithiumand manganese-rich transition metal oxides exhibit excellent specific capacities, making them strong candidates for the development of the next generation of Co-free lithium-ion batteries. In this study, the synthesis of size-tunable Li 1.2 Ni 0.2 Mn 0.6 O 2 using a synthetic route based on the formation of an emulsion, which is ultrafast, cost-effective, and easily scalable to an industrial level is presented. We demonstrate that variations in the concentrations of hydrophobic, hydrophilic, and surfactant components, which lead to micelle formation within the emulsion, have a significant impact on the average particle size and size distribution of the synthesized material, and subsequently, on their electrochemical performance. Specifically, increasing the concentration of oleic acid as a surfactant results in an optimal average particle size, with discharge specific capacities exceeding 317 mAh g −1 in the first cycle and 230 mAh g −1 after 100 cycles, demonstrating an excellent battery performance comprising state-of-the-art lithiumand manganese-rich transition metal oxide materials. * Corresponding author. E-mail address: [email protected] (J.G. Lozano). Contents lists available at ScienceDirect Journal of Colloid And Interface Science journal homepage: www.elsevier.com/locate/jcis https://doi.org/10.1016/j.jcis.2025.139182 Received 1 August 2025; Received in revised form 2 October 2025; Accepted 3 October 2025 Journal of Colloid and Interface Science 703 (2026) 139182 Available online 4 October 2025 0021-9797/© 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/bync-nd/4.0/ ). 1. Introduction Nowadays, the growing demand for energy and sustainable production underscores the urgent need for efficient energy storage technologies adaptable to various applications, including portable electronic devices, smart grids, and electric and hybrid vehicles. In this context, lithium-ion batteries (LIBs) are widely chosen due to their high energy and power density, long cycle life, and relatively high safety. However, cathode materials remain a significant challenge, and the development of greener, cost-effective, and safer alternatives is a key research focus in the field of LIBs. Extensive research has been dedicated to identifying cathode materials that enhance battery performance. One of the critical factors for optimal battery operation is a cathode with a crystalline structure capable of accommodating a large number of Li-ions. Currently, LiCoO 2 - based cathodes are the most widely used in the industry. While they offer reliable performance, they also present several drawbacks, particularly due to cobalt-related issues such as scarcity, ethical concerns regarding mining conditions, toxicity, and recyclability challenges [1,2]. Lithium-rich layered oxides have shown great potential as viable replacements for current cathode materials. Specifically, Li 1.2 Ni 0.2 Mn 0.6 O 2 emerges as a strong candidate, as it surpasses the energy density limitations imposed by conventional transition-metal redox reactions [3–5]. Moreover, manganese is both environmentally friendly and cost-effective, further reinforcing its suitability for next-generation LIBs. However, the primary challenge limiting the utilization of these materials is their significant capacity loss and voltage fade after the first cycle. This degradation is mainly attributed to oxygen-loss from the lattice upon charge which leads to the formation of spinel and rock-salt phases on the particle surfaces, partially blocking the reinsertion of lithium in subsequent cycles, and to the volume changes during lithiation and delithiation which can induce defects and cracks over repeated cycles, ultimately compromising battery performance [6–12]. Regarding the last issue, one of the proposed solutions is the control of particle shape and size. In similar systems, it has been observed that nanosized particles offer high capacities in the first cycle but suffer from more pronounced capacity and voltage fading compared to slightly larger particles [13–17]. In contrast, micrometer-sized particles are generally inactive [18]. Therefore, it is necessary to find an optimal particle size that can accommodate most of the structural changes induced by charge and discharge cycles while still providing sufficient channels for lithium insertion and extraction. Synthesis routes based on microemulsion fabrication, unlike conventional techniques such as coprecipitation, ball-milling, or even solgel, offer greater control over the size of the resulting particles [19,20]. Microemulsions are complex systems consisting of a hydrophilic and a hydrophobic phase, stabilized by surfactants to form thermodynamically stable and isotropic dispersions. The procedures that use microemulsions stabilized as reverse micelles involve reactions carried out within the volume defined by water droplets dispersed in a nonaqueous medium. Aggregates of surfactant molecules, with their hydrophilic groups oriented toward the micelle core, define the volume of the microreactor and control the growth, size distribution, and morphology of the products resulting from the reactions inside the droplets [21–24]. When the internal volume of the reverse micelles contains aqueous solutions of suitable reagents, three possible reaction routes can occur: (a) the interaction between two micelles containing different reagents within the same emulsion, (b) the reaction of micelles with reagents that diffuse through the oil medium and penetrate the micelle interface, and (c) the activation of reactions (e.g., hydrolysis) within a set of identical micelles by means of an external stimulus such as light or heat. Interestingly, methods (a) and (b) typically lead to the precipitation of insoluble inorganic salts. Moreover, salts obtained using the reverse micelle approach can also serve as precursors of mixed oxides. A detailed review on the preparation of metal nanoparticles in water-in-oil (w/o) microemulsions can be found in [25]. Depending on the proportions of the components and the hydrophilic–lipophilic balance value of the surfactant used, the microdroplets formed can be either oil-swollen micelles dispersed in water (o/w microemulsion) or water-swollen micelles dispersed in oil (w/o microemulsion, i.e., reverse microemulsion). In addition, the shape of micellar aggregates and the formation of the microemulsion can be rationalized and tuned using the packing parameter of the emulsifier molecule in the micellar assembly, v/a⋅l, where v is the hydrocarbon volume, a is the polar headgroup area, and l is the fully extended chain length of the emulsifier [25]. In the context of lithium-ion batteries, a series of materials with different morphology have been prepared by microemulsion, including Liand Mn-rich cathodes [26–28] or LiFePO 4 [29,30]. Although the precise influence of the hydrophilic, hydrophobic, and surfactant components on micellar systems remains challenging to fully characterize, it is well established that parameters such as solvent nature, surfactant headgroup size, chain length, and temperature, among others, play a crucial role in modulating the microemulsion composition. Regarding the surfactant, F. J. Garcia-Garcia et al. demonstrated previously [31] the importance of the nature of the surfactant structure (chain length, headgroup and so forth) in stabilizing the reverse micelles and influencing nucleation/growth kinetics in the fabrication of Na 2/3 Ni 1/3 Mn 2/ 3 O 2 cathodes for Na-ion batteries. They demonstrated that the use of short-chain amines as co-surfactants lower the interfacial tension and help in changing the curvature of the reverse micelles. As an example, the use of octylamine was demonstrated to trigger the formation of the precursor nuclei and the formation of particles with new morphology and excellent dispersibility. A much shorter-chain amine, i.e. methylamine, promoted a significant smaller crystallite size and thinner particle responsible for enhanced sodium diffusivity. A similar strategy can be found in [32] for the fabrication of Fe 3 O 4 nanoparticles for a different application. Building on this understanding, we have employed a modified emulsion-based synthesis adapted within our research group [31]. The synthesis process employs a nonpolar organic liquid immiscible with water. In this study, Vaseline oil was selected due to its absence of known health hazards compared to industrial mineral oils. Given the hydrophilic nature of the metal ions used for Li 1.2 Ni 0.2 Mn 0.6 O 2 synthesis, they are initially dissolved in water. Oleic acid serves as the surfactant due to its amphiphilic nature, possessing a long carbon chain with distinct hydrophilic and hydrophobic regions. At a specific concentration, and upon mixing with the use of a shear mixer, the surfactant facilitates the formation of reverse micelles, where the nonpolar tails orient outward while the polar heads interact with the aqueous phase containing the metal ions. These micelles establish the foundation for a microemulsion, which undergoes thermal treatment to remove organic components, yielding the final Li 1.2 Ni 0.2 Mn 0.6 O 2 particles. This approach presents several advantages over conventional techniques such as hydrothermal synthesis, sol-gel processing, or dry-milling: (i) it relies on organic precursors, minimizing the formation of contaminant byproducts; (ii) it significantly reduces processing times compared to conventional dry-milling or sol-gel methods [33–35], decreasing the mixing stage prior to heat treatment from 90 min, or longer, to just 4 min, thanks to the use of a shear mixer; and (iii) it offers substantial potential for upscaling to industrial production [36]. Since the stability and dimensions of the reverse micelles directly impact the final particle size, this study systematically investigates the effect of varying water, vaseline oil, and oleic acid concentrations on the size, distribution, morphology, structure, and electrochemical performance of Li 1.2 Ni 0.2 Mn 0.6 O 2 particles. Crucially, we demonstrate that the Li 1.2 Ni 0.2 Mn 0.6 O 2 particle size and distribution can be effectively tuned by varying the hydrophobic-to-hydrophilic-to-surfactant ratio. This optimization enables first-charge capacities of, comparable to some of the best reported values in the bibliography, along with excellent capacity retention over 100 cycles. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 2 2. Materials and methods 2.1. Synthesis of Li 1.2 Ni 0.2 Mn 0.6 O 2 Stoichiometric amounts of LiOCOCH 3 , Ni(OCOCH 3 ) 2 ⋅4H 2 O and Mn (OCOCH 3 ) 2 ⋅4H 2 O from Sigma Aldrich, were mixed with distilled water, oleic acid and vaseline inside a Teflon container. The specific quantities used in each synthesis are summarized in Table 1, corresponding to different formulations: a synthesis with a high oleic acid concentration (O-rich), one with increased water content (W-rich), and another with a high hydrophobic component content (V-rich). The mixture was processed for 4 min using a high speed shear mixer. The resulting emulsion was poured into an alumina crucible and placed inside an oven with a heating ramp of 5 ◦C/min with a target temperature of 800 ◦C. It is worth noting that, during the heating ramp, at a temperature of approximately 100 ◦C the H 2 O and OH − groups are eliminated, and at approximately 450 ◦C the calcination of the undesired organic groups will take place. Finally, once reached the temperature of 800 ◦C, the product is maintained inside the oven for 20 h and left to cool down to room temperature in order to achieve the Li 1.2 Ni 0.2 Mn 0.6 O 2 particles with the expected crystalline structure. 2.2. Electrochemical characterisation The electrochemical experiments on Li half-cells were carried out in Swagelok™-type cells at ambient temperature. The cathode consisted of 80 wt% of active material, 10 wt% of carbon black, and 10 wt% of polyvinylidene Fluoride (PVDF). The mixture was homogeneously dispersed into N-methyl pyrrolidine (NMP) and then spread onto Ti disks (Sigma-Aldrich, 99.7 % purity and 0.127 mm thickness). To evaporate the NMP solvent, the electrode was dried for at least 2 h in a vacuum oven at 120 ◦C. Glass fiber discs were used as separators. 1 M LiPF 6 EC:DEC and lithium metal discs were used as the electrolyte and counter electrode, respectively. Swagelok™-type cells were assembled in a glovebox under an argon atmosphere (H 2 O, O 2 <0.1 ppm). The cells were cycled between 4.8 and 2 V at 50 mA g −1 , and galvanostatic charge-discharge tests were conducted on an Ivium-n-Stat battery test system. 2.3. Materials characterisation To study the crystallinity and purity of the samples, X-ray diffraction (XRD) was employed. XRD patterns were scanned at 0.02◦/s between 10 and 70◦(2θ-degree) on a BrukerD8 Discover A25 diffractometer equipped with Cu K radiation and a graphite monochromator. Rietveld refinements were performed using the Topas-32v6 software. For the analysis of the crystallinity, morphology and size of the nanoparticles via transmission electron microscopy (TEM) and related techniques, the TEM samples were prepared by dispersing the particles in ethanol, followed by sonication, and drop casting on lacey carbon copper grids. Scanning-transmission electron microscopy (STEM) and energy dispersive X-ray spectroscopy (EDX) were performed using a FEI Talos F200S microscope operating at an accelerating voltage of 200 keV. Further STEM analyses with atomic resolution were carried out at aberration corrected (S)TEM FEI Titan Cubed Themis 60–300 and JEOL ARM200F microscopes, both operating at 200 kV. In order to minimize damage and correct for drift and scan distortions, sequences of fast STEM images were recorded and subsequently aligned and averaged using a non-rigid registration algorithm [37]. 3. Results and discussion XRD was employed to investigate the crystal structures of the Li 1.2 Ni 0.2 Mn 0.6 O 2 powder samples (O-rich, W-rich, and V-rich), as presented in Fig. 1a. Rietveld refinement (see Table 2 for figures of merit, unit cell parameters and phase fractions) shows that diffraction signals are consistent with a mix of rhombohedral layered phase, assignable to the R3‾m space group; and the monoclinic Li 2 MnO 3 phase (Li[Li 1/3 Mn 2/ 3 ]O 2 ), which crystallizes in the C2/m space group, in agreement with previous results in the literature [38–40]. The two minor peaks located at approximately 20.9◦and 21.7◦2θ correspond to the (020) and (110) reflections are attributed to the ordered Li – Mn arrangements (LiMn 6 units) typically found in the transition metal layers of Li 2 MnO 3 -like nano-domains, a known feature of Li-rich layered materials [6,41]. The absence of additional peaks confirms the high phase purity of the synthesized samples, free from undesirable secondary phases that generally yields in worst battery performances. Additional XRD patterns were collected after the first charge at 4.8 V and after the first discharge, showing that both structures remain in all samples after the first full cycle (Fig. 1b). To evaluate the particle size in the fabricated Li 1.2 Ni 0.2 Mn 0.6 O 2 samples, conventional TEM images at low magnification were obtained, and the results are summarized in Fig. 2. A statistical analysis of the average particle size and size dispersion allowed us to draw some preliminary conclusions: although particles with similar morphology are observed in all samples, both the average particle sizes and the size distributions vary significantly among them. From the study of the Orich sample (Fig. 2a), it was concluded that an increase in oleic acid concentration leads to a decrease in both the average particle size and the size distribution, reaching a value of 88 ±23 nm. However, in the W-rich sample (Fig. 2b), an increase in water concentration results in a larger average particle size (150 ±100 nm), with a very random size distribution. Finally, an increase in the amount of vaseline (V-rich sample, Fig. 2c) causes a drastic rise in particle size, reaching dimensions close to the micrometer range, with an average value of 720 ±300 nm. The first two observations can be attributed to the stabilizing effect that a higher surfactant concentration has on inverse micellar systems [42,43]. In the W-rich sample, an insufficient concentration of oleic acid results in an unstable emulsion, leading to phase separation between the micellar system and the aqueous phase containing dissolved ions. This instability promotes uncontrolled particle formation, resulting in a broader distribution of particle sizes. In contrast, the O-rich sample, prepared with a higher surfactant concentration, exhibits a significantly more stable microemulsion characterized by well-defined reverse micelles with homogeneous size and spatial distribution. This homogeneity facilitates more consistent nucleation events and promotes uniform particle growth during the thermal treatment at 800 ◦C. In the V-rich sample, we hypothesize that the lower water content leads to reverse micelles being more widely spaced, increasing the possibility of their expansion. Moreover, the higher mobility of the micelles raises the likelihood of fusion events during the calcination process, ultimately resulting in a larger final particle size. From the EDX maps obtained by STEM (Fig. 3), it can be concluded that the observed variations in particle size also affect the chemical distribution of the elements within the particles. In the case of the O-rich sample, the distribution of Mn, Ni, and O is homogeneous. However, Ni shows a tendency to diffuse toward the particle surface, attributed to the formation of Ni-rich layers with a spinel-like structure on specific particle facets, as previously reported in the literature [41]. In contrast, in the larger particles of the W-rich sample—and even more markedly in Table 1 Proportion of components for the synthesis of the different samples. Li [g] Ni [g] Mn [g] H 2 O [g] Oleic acid [g] Vaseline [g] Sample 1 (Orich) 1.47 0.90 2.63 5 32 14 Sample 2 (Wrich) 32 7 14 Sample 3 (Vrich) 5 7 32 S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 3 the V-rich sample—the distribution of Ni is significantly more heterogeneous. This can be explained by the presence of a larger internal grain boundary area within each particle, resulting from the coalescence of smaller particles into larger ones during sintering. Additionally, due to the higher mobility of Ni, it tends to segregate more readily toward the particle surfaces and along grain boundaries [9,41]. More importantly, the differences observed in the average particle size and size distribution, as well as in compositional homogeneity, have a direct impact on their electrochemical performance, as summarized in Fig. 4. Fig. 4a shows the initial charge–discharge curves of the, O-rich, W-rich, and V-rich Li 1,2 Ni 0.2 Mn 0.6 O 2 samples at room temperature in the voltage windows from 2.0 to 4.8 V at 50 mA g −1 . In all cases, there are two distinct regions for the initial charge process: the sloping region below 4.5 V and the flat region at around 4.5 V. The sloping region is attributed to the de-intercalation of Li + from the structure with space group R3 ‾m and the accompanying oxidation of Ni from Ni 2+ to Ni 4+ . The flat region at around 4.5 V could be attributed to the activation of the Li 2 MnO 3 component, accompanying with the oxidation of O 2− to facilitate the extraction of Li + and the following structural rearrangement. This process appears only in the initial cycle, corresponding to the sharp oxidation peak at about 4.6/4.7. When the discharge process is proceeded, there are two reduction peaks. The first reduction peak at about 3.25 V is attributed to the reduction of Mn 4+ /Mn 3+ and the other one is suggested to be the reduction reaction of Ni 4+ /Ni 2+ . However, the values of charge and discharge in the first cycle - O-rich (354–317 mAh g −1 ); W-rich (276–274 mAh g −1 ); and V-rich (275–181 mAh g −1 ) - are Fig. 1. Rietveld-refined XRD spectra of the Li 1.2 Ni 0.2 Mn 0.6 O 2 particles synthesized using high-oleic acid content (O-rich), high-water content (W-rich), and highvaseline content (V-rich). At the bottom, simulated spectra of the C2/m and R3‾m structures. b) Ex situ XRD patterns of the O-rich, W-rich and V-rich samples after full charge and subsequent discharge. At the bottom, simulated spectra of the C2/m and R3‾m structures. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 4 remarkably different, with the capacity of the O-rich sample being comparable to some of the best values reported in the literature for this type of material (See Section 1 in the Supplementary Information for a comparative table of the electrochemical properties of the O-rich sample with previously reported materials synthesized by other methods). Fig. 4b shows the retention capacity profiles of the three samples. It can be observed that the sample O-rich presents the best retention capacity after 100 cycles, with discharge values of 230 mAh g −1 and a coulombic efficiency of 99 %. The rest of samples present a value of capacity after 100 cycles of 200 and 75 mAhg −1 for W-rich, and V-rich respectively. Rate capability tests were performed cycling between 2 V and 4.8 V vs Li + /Li for kinetics from 0.5C-10C and returning to 0.5C, are shown in Fig. 4c. All samples show good capacity and capacity retention even at high kinetics such as 10C, with the O-rich sample showing the best performance at different current densities. When the charge and discharge current densities are reduced from 10C to 0.5C, (1C =100 mA g −1 ) the capacity can recover almost 100 % (Fig. 4c). Although the specific capacities achieved—particularly for the Orich sample—are considerably high, their behavior after the first charge varies. While the O-rich and W-rich samples recover high capacities in the first discharge, with coulombic efficiencies around 90 % and 82 % respectively, a difference mainly attributed to the already reported oxygen loss, the V-rich sample exhibits comparatively much poorer coulombic efficiency during the first charge–discharge cycle, with over Table 2 Rietveld refinement parameters for crystallite size of the monoclinic and rhombohedral phases exhibited for all samples. Phase Parameters Sample O-rich W-rich V-rich Monoclinic a[Å] 4.9433(11) 4.9602(16) 4.9424(8) C2/m b[Å] 8.5554(16) 8.5545(16) 8.5457(8) c[Å] 5.032(2) 5.0419(13) 5.0372(7) β[◦] 109.06(5) 109.336(15) 109.311(6) R-Bragg 0.538 0.245 0.441 Phase[%] 75.0(12) 80.1(14) 73.5(7) Rhombohedral a[Å] 2.8911(9) 2.8877(10) 2.9219(11) R-3 m c[Å] 14.183(13) 14.34(2) 14.313(7) R-Bragg 0.516 0.224 0.283 Phase[%] 25.0(12) 19.9(14) 26.5(7) R wp 3.12 2.77 2.63 R p 2.31 2.20 2.08 GOF( χ 2 )1.30 1.07 1.10 Fig. 2. Low-magnification TEM micrographs of the a) O-rich, b) W-rich and c) V-rich samples, showing the difference in average size and distribution. The scale bar is common to all micrographs. Fig. 3. ADF images and corresponding Mn, Ni and O EDX maps of the O-rich, W-rich and V-rich samples. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 5 34 % of the capacity lost. This results in low capacities from the second cycle onwards. Since the XRD results do not show significant changes in the structure of the materials after the first charge–discharge cycle, it is reasonable to assume that the differences occur at the nanometric or even atomic scale, below the spatial resolution of the technique. To corroborate this hypothesis, STEM measurements were analyzed in greater detail for the samples that exhibited the most distinct behavior—namely the O-rich and V-rich compositions—in their pristine state, and after one charge-discharge cycle. The analysis of low-magnification STEM images of the two samples after discharge reveals an initial difference. While the particles of the OFig. 4. Galvanostatic first charge/discharge curve of the O-rich, W-rich and V-rich. b) shows the retention capacity profiles over 100 cycles and c) and the rate performance at room temperature of the three samples. To reveal the mechanisms for the improved electrochemical properties achieved by the Li₁ 1.2 Ni 0.2 Mn 0.6 O 2 , using O-rich, W-rich and V-rich respectively, electrochemical impedance spectroscopy (EIS) was employed to trace variations in the interfacial resistance between the electrodes and electrolyte. The obtained Nyquist plots, in the range between 1 MHz and 1 mHz, are illustrated in Fig. 5 (see also Section 2 of the Supplementary Information for the equivalent circuit model (Fig. S1), together with Table S2, reporting the resistance values obtained from impedance fitting after the first discharge). S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 6 rich sample more clearly retain their structural stability (Fig. 6a), the larger particles of the V-rich sample display (i) a transition to more irregular surfaces (marked by blue arrows in Fig. 6b), indicative of structural transformation, and (ii) the formation of circular regions with lower contrast (marked by orange arrows in Fig. 6b). These features have recently been associated with the presence of trapped O 2 bubbles, Fig. 5. Impedance spectra, represented as Nyquist plots, recorded on O-rich, W-rich and V-rich electrodes subject to a) the first charge and c) the first discharge. b) and d) display, for clarification, the amplified region corresponding to high frequencies in a) and c), respectively. The EIS of half-cells were tested at the first charged state of 4.8 V and the first discharge state of 2 V respectively, at 50 mA g −1 current rate. Their spectra profiles are characterized by two semicircles at high and low frequencies respectively, both fitted with a resistance and a phase constant element ion parallel. Both semicircles, in agreement with reported research manuscripts [44,45], can be associated with the internal electrolyte (Re), the surface layer (Rs), and charge-transfer reaction (Rct) at the electrode/electrolyte interphase respectively. Interestingly, it is noticeable that a Warburg element in series with Rct is needed to consider the fact that these two processes can often occur in sequence [46]. Initially, the Li + ions undergo a charge transfer at the interface. Subsequently, the ions diffuse into the bulk electrode material. Herein, the Warburg element would represent the impedance due to Li + -ions diffusion. It is worth mentioning that V-rich sample led to the highest resistance values, suggesting that the whole process is limited by the internal impedance of the cell. On the contrary, O-rich sample led to the lowest resistance values, in line with the capacitance values obtained in both long cycling and C-Rate (Fig. 4b and c). Fig. 6. Low magnification STEM images of the a) O-rich and b) V-rich samples after the first charge-discharge cycle. The scale bar is common to both micrographs. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 7 linked to the well-known oxygen loss (O-loss) occurring during the first charge [47]. Because the V-rich particles exhibit a comparatively lower surface-to-volume ratio, this sample is more prone to the formation of such O 2 -filled voids since the formed oxygen O 2 cannot be as easily released as in the O-rich smaller particles. These voids eventually lead to the development of cracks and particle fragmentation during extended cycling, leading to battery failure [47]. To analyse the differences at the atomic scale between the two samples in the pristine and discharged state, further aberrationcorrected high-resolution STEM images were recorded. The results for the pristine samples are summarized in Fig. 7. The micrographs taken along the [010] zone axis (Fig. 7a and c) show the laminar structure characteristic of Li-rich oxides in all cases. This feature can be attributed either to the monoclinic phase (Li₂MnO₃) or to the trigonal phase (LiMO₂, where M represents the transition metals present in the compound), in accordance with the XRD results. Additional micrographs acquired along the [100] zone axis (Fig. 7b and d) reveal the characteristic “dumbbell” structure formed by pairs of transition metal atoms—a distinctive feature exclusive to the monoclinic phase, which cannot be observed in any orientation of the trigonal phase [41]. Additionally, the images taken along the [100] direction show that the particle surfaces exhibit 1–2 nm thick layers with brighter contrast and hexagonal patterns on certain facets, as described in the STEM-EDX maps (Fig. 3) due to the formation of a nickel-rich thin layer with spinel structure and a specific orientation relationship: (100)ₘ || (110)ₛ. After the first charge–discharge cycle, two relevant phenomena are observed. First, the monoclinic structure is preserved (Fig. 8b and d), in good agreement with the XRD results displayed in Fig. 2b. Second, surface restructuring occurs as a result of the aforementioned oxygen loss (O-loss), which induces the formation of spinelor rock-salt–like structures on the particle surface. The thickness of the surface reconstructed layer is 2–5 nm in the O-rich particles (Fig. 8a and b) but is much more pronounced in the V-rich sample, where the thickness increases up to 10–12 nm. Additionally, bending of the (002) planes near the surface is also observed (see Fig. 8d). These two facts can be attributed to the larger particle size, which limits the ability of the structure to accommodate local stresses arising from uneven delithiation. Additionally, the formation of O₂-filled voids, as previously discussed, further contributes to structural instability. These structural changes are likely responsible for the poor coulombic efficiency observed during the first cycle in the V-rich sample, as they hinder lithium reinsertion. In contrast, the smaller particles in the O-rich sample allow for better accommodation of local stresses during the charge–discharge process, resulting in less pronounced structural changes and a greater number of accessible and wellpreserved lithium diffusion channels. Finally, this long-term behavioral difference between the samples is further supported by the additional aberration-corrected HR-STEM measurements performed on the O-rich and V-rich samples after 100 charge–discharge cycles, as presented in Fig. 9. From the analysis of Fig. 9a, it can be deduced that in the case of the O-rich sample, surface restructuring is more pronounced than after the first cycle, with the thickness of the rock-salt layer increasing to about 5 nm. However, as shown in Fig. 9b, the bulk of the particle still retains high crystallinity and the monoclinic structure. In contrast, for the Vrich sample, the transformation into the rock-salt structure is complete in the area visible in the HR-STEM image, which is consistent with the significant capacity loss observed for this sample after 100 charge–discharge cycles. 4. Conclusions To summarize, we demonstrate that a rapid and straightforward emulsion-based synthesis method enables the production of Li 1.2 Ni 0.2 Mn 0.6 O 2 with tuneable particle size, allowing for the optimization of particle dimensions that can accommodate the volumetric changes induced by charge–discharge cycles while maintaining effective lithium diffusion pathways. The use of large amounts of water leads to the formation of unstable emulsions, resulting in particles with poorly controlled size and morphology. Conversely, increasing the hydrophobic component leads to the formation of significantly larger particles, although with inferior electrochemical performance. High surfactant concentrations, however, Fig. 7. Aberration-corrected HRSTEM images of a pristine O-rich particle taken along the a) [010] and b) [100]/[110] zone axis. c) and d) show images of a V-rich particle also recorded along the [010] and [100]/[110] zone axis in pristine state, respectively. The scale bar is common to all micrographs. Fig. 8. HR-STEM micrographs of a discharged O-rich particle along the a) [010] and b)[100]/[110] direction; and a discharged V-rich particle along the c) [010] and d)[100]/[110] direction The scale bar is common to all micrographs. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 8 promote the formation of stable micelles that drive the growth of more uniform particles with optimal size and narrower particle size distribution, ultimately resulting in cathode materials with excellent electrochemical performance (317 mAh g −1 initial discharge capacity and 230 mAh g −1 after 100 cycles). Aberration-corrected STEM imaging confirms that larger particles undergo more extensive surface restructuring, which hinders lithium reinsertion in subsequent cycles and contributes to their low coulombic efficiency. In contrast, smaller particles can more effectively accommodate the stresses associated with volumetric fluctuations, preserving lithium diffusion channels and promoting more stable cycling performance. While further studies are needed to fully elucidate the mechanisms underlying micelle stabilization and to refine the optimal surfactant ratio, our results represent a significant step toward the controlled design of Li₁ .2 Ni 0.2 Mn 0.6 O 2 particles for safe and high-performance lithium-ion battery cathodes. The proposed synthesis method is not only fast and cost-effective but also readily scalable, making it highly promising for industrial implementation. CRediT authorship contribution statement Saul Rubio: Writing – review & editing, Investigation, Formal analysis. Ana M. Beltr´ an: Writing – review & editing, Investigation, Formal analysis. Cristina Ar´ evalo: Writing – review & editing, Investigation, Formal analysis. Gerardo T. Martinez: Writing – review & editing, Investigation, Formal analysis. Francisco J. Garcia-Garcia: Writing – review & editing, Investigation, Formal analysis. Eva M. P´ erez-Soriano: Writing – review & editing, Investigation, Formal analysis. Isabel Montealegre-Mel´ endez: Writing – review & editing, Investigation, Formal analysis. Juan G. Lozano: Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Juan G. Lozano, Saul Rubio, Francisco J. Garcia Garcia, Ana M. Beltran, Cristina Arevalo, Eva M Perez Soriano, I Montealegre Melendez has patent #202430516 pending to Spanish Patent and Trademark Office. N/A If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgement Financial support was provided by the Grant PID 2020-113108RBI00 funded by MICIU/AEI/ 10.13039/501100011033 and, as appropriate, by “ERDF A way of making Europe”, by “ERDF/EU”, by the “European Union” or by the “European Union NextGenerationEU/ PRTR”. Financial support was also provided by the Grant CNS2022-135432 funded by MICIU/AEI/10.13039/501100011033 and, as appropriate, by “ESF Investing in your future”, by “ESF+” or by “European Union NextGenerationEU/PRTR”. The authors are also grateful to the Universidad de Sevilla for the use of the research facilities at CITIUS. The authors acknowledge the use of (S)TEM instrumentation provided by the Spanish National Facility ELECMI ICTS (“Divisi´ on de Microscopía Electr´ onica”, Universidad de Cadiz, DME-UCA). The authors also acknowledge the use of characterisation facilities within the David Cockayne Centre for Electron Microscopy, Department of Materials, University of Oxford, alongside financial support provided by the Henry Royce Institute (Grant ref. EP/R010145/1). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.jcis.2025.139182. Data availability Data will be made available on request. References [1] R. Konar, S. Maiti, N. Shpigel, D. Aurbach, Reviewing failure mechanisms and modification strategies in stabilizing high-voltage LiCoO2 cathodes beyond 4.55V, Energy Storage Mater 63 (2023), https://doi.org/10.1016/j.ensm.2023.103001. [2] J. Jyoti, B.P. Singh, S.K. Tripathi, Recent advancements in development of different cathode materials for rechargeable lithium ion batteries, J Energy Storage 43 (2021), https://doi.org/10.1016/j.est.2021.103112. [3] G. Assat, D. Foix, C. Delacourt, A. Iadecola, R. Dedryv` ere, J.M. Tarascon, Fundamental interplay between anionic/cationic redox governing the kinetics and thermodynamics of lithium-rich cathodes, Nat. Commun. 8 (2017), https://doi. org/10.1038/s41467-017-02291-9. [4] K. Luo, M.R. Roberts, R. Hao, N. Guerrini, D.M. Pickup, Y.S. Liu, K. Edstr¨ om, J. Guo, A.V. Chadwick, L.C. Duda, P.G. Bruce, Charge-compensation in 3d-transition-metal-oxide intercalation cathodes through the generation of localized electron holes on oxygen, Nat. Chem. 8 (2016) 684–691, https://doi.org/10.1038/ nchem.2471. [5] K. Redel, A. Kulka, K. Walczak, A. Plewa, E. Hanc, M. Marzec, L. Lu, J. Molenda, Origin of extra capacity in advanced Li–Rich cathode materials for rechargeable Li–Ion batteries, Chem. Eng. J. 424 (2021), https://doi.org/10.1016/j. cej.2021.130293. [6] K. Luo, M.R. Roberts, R. Hao, N. Guerrini, E. Liberti, C.S. Allen, A.I. Kirkland, P. G. Bruce, One-Pot Synthesis of Lithium-Rich Cathode Material with Hierarchical Morphology, Nano Lett. 16 (2016) 7503–7508, https://doi.org/10.1021/acs. nanolett.6b03296. [7] R.A. House, U. Maitra, L. Jin, J.G. Lozano, J.W. Somerville, N.H. Rees, A.J. Naylor, L.C. Duda, F. Massel, A.V. Chadwick, S. Ramos, D.M. Pickup, D.E. McNally, X. Lu, T. Schmitt, M.R. Roberts, P.G. Bruce, What Triggers Oxygen Loss in Oxygen Redox Cathode Materials? Chem. Mater. 31 (2019) 3293–3300, https://doi.org/10.1021/ acs.chemmater.9b00227. [8] Z. Cai, S. Wang, H. Zhu, X. Tang, Y. Ma, D.Y.W. Yu, S. Zhang, G. Song, W. Yang, Y. Xu, C. Wen, Improvement of stability and capacity of Co-free, Li-rich layered Fig. 9. HR-STEM micrographs of an O-rich particle along the a) [010] and b) [100]/[110] direction; and a discharged V-rich particle along the c) [010] and d) [100]/[110] direction, both after 100 charge-discharge cycles. The scale bar is common to all micrographs. S. Rubio et al. Journal of Colloid And Interface Science 703 (2026) 139182 9