Lady V.
Quispe-Garrido
ab,
Angélica María
Baena-Moncada
c,
Guto
Garcia
d,
Carla
Real
d,
Irlan S.
Lima
b,
Karen
Magno da Silva
ef,
Hudson
Zanin
d,
Walter
Estrada
*a,
Lucio
Angnes
*b and
Josué M.
Gonçalves
*ef
aFacultad de Ciencias de la Universidad Nacional de Ingeniería, Av. Túpac Amaru 210, Rímac, Lima, Peru. E-mail: westrada@uni.edu.pe
bDepartment of Fundamental Chemistry, Institute of Chemistry, University of São Paulo, Av. Prof. Lineu Prestes 748, São Paulo, SP, Brazil. E-mail: luangnes@iq.usp.br
cLaboratorio de Investigación de Electroquímica Aplicada, Facultad de Ciencias de la Universidad Nacional de Ingeniería, Av. Túpac Amaru 210, Rímac, Lima, Peru
dAdvanced Energy Storage Division, Carbon Sci-Tech Labs, Center for Innovation on New Energies, School of Electrical and Computer Engineering, University of Campinas, Av Albert Einstein 400, Campinas, SP 13083-852, Brazil
eMackenzie Institute for Research in Graphene and Nanotechnologies (MackGraphe), Mackenzie Presbyterian Institute, Rua da Consolação, 896, São Paulo, SP 01302-907, Brazil. E-mail: josue.martins@mackenzie.br
fSchool of Engineering, Mackenzie Presbyterian University, Rua da Consolação, 896, São Paulo, SP 01302-907, Brazil
First published on 8th April 2026
Sodium-ion batteries (SIBs) have emerged as a cost-effective and sustainable alternative to lithium-ion-based energy technologies, intensifying the search for high-performance cathode materials. Among the available candidates, layered transition-metal oxides with the general formula NaxTMO2 stand out due to their high energy density and structural versatility. Recently, entropy engineering, ranging from low-to high-entropy oxide designs (LEOs, MEOs, and HEOs), has been reported as an effective strategy to enhance structural stability, ionic transport, and electrochemical performance. In parallel, spray-drying has gained increasing attention as a scalable, industrially relevant synthesis route that ensures homogeneous cation distribution, controlled particle morphology, and reproducible microstructures. This review provides a critical and systematic assessment of spray-drying technologies applied to the synthesis of layered NaxTMO2 cathodes for SIBs, with particular emphasis on entropy-driven material design. Fundamental aspects of the spray-drying process, including atomization methods, are discussed in detail. In addition, statistical information was obtained based on the type of precursor used in the spray-drying synthesis of NaxTMO2 and the post-thermal processing treatments applied. Recent advances in spray-dried LEOs, MEOs, and HEOs are analyzed, correlating composition trends and strategies with structural features and electrochemical performance. Moreover, common misclassifications of entropy levels in the literature are addressed through a rigorous discussion of configurational entropy criteria. Finally, current challenges and future perspectives for spray-drying-assisted entropy engineering of layered cathodes are outlined, highlighting its potential for large-scale manufacturing of next-generation SIBs.
Several families of cathode materials have been investigated to develop low-cost and high-performance SIBs, including polyanionic compounds (such as Na Superionic Conductor – NASICON), Prussian blue analogues (PBAs), and layered transition-metal oxides (LTMOs).4 Among these, LTMOs with the general formula NaxTMO2 (where TM = Ni, Fe, Mn, Cu, Ti, or other transition metals) have attracted particular attention owing to their high energy density and structural versatility.1 Besides, recent studies have demonstrated that incorporating non-transition-metal elements, such as Mg2+, Al3+, Zn2+, Ca2+, Li+, and B3+, can significantly improve the structural stability and electrochemical behavior of layered oxides. For instance, substitutions with electrochemically inactive elements such as Mg2+ and Al3+ enhance structural stability by effectively suppressing irreversible structural degradation and inducing more reversible phase transitions during repeated charge–discharge processes,5,6 thereby stabilizing the layered framework. In addition, NaxTMO2 doped with lower-valence cations (Mg2+ and Al3+) exhibits limited volume expansion, thereby maintaining lattice integrity and enhancing cycling stability.7 On the other hand, chemical substitutions with Li+, Zn2+, Ca2+, and Ti4+ have been reported to enhance air stability and suppress detrimental side reactions in NaxTMO2.8 Fortunately, recent reviews have comprehensively outlined the functional roles of both active and inactive metal cations in NaxTMO2 for SIBs.3,9
The introduction of an emerging class of materials based on the high-entropy concept has recently been employed to regulate both the structure and the electronic properties of layered Na-based cathode materials, thereby enhancing their overall electrochemical performance.10 In fact, the incorporation of multiple cations in equimolar or near-equimolar ratios within the transition-metal lattice has been a principal strategy for obtaining the so-called high-entropy oxides (HEOs). In general, the high-entropy design can result in a single-phase structure introducing characteristic effects, such as (i) high configurational entropy, (ii) lattice distortion, (iii) sluggish diffusion, and (iv) cocktail effect. Interestingly, in layered HEOs (e.g., high-entropy NaxTMO2), their configuration can stabilize layered phases and promote more efficient ionic-transport pathways. These materials may be classified according to their composition into low-entropy materials (LEMs), medium-entropy materials (MEMs), and high-entropy materials (HEMs), based on the calculation of configurational entropy (aspect discussed in topic 6). The high-entropy design in NaxTMO2 was first adopted by Zhao et al. (2019), who synthesized a nine-transition-metal-containing HEO, (NaNi0.12Cu0.12Mg0.12Fe0.15Co0.15Mn0.1Ti0.1Sn0.1Sb0.04O2) showing remarkable cycling stability and excellent rate performance.11,12 In general, a high-entropy NaxTMO2 typically introduces multiple (≥5) TM cations with differing physical and chemical properties, thereby significantly increasing the material's configurational entropy, thereby enhancing the stability of the crystal structure.9
As expected for a rapidly emerging topic in materials chemistry, recent review articles have summarized advances in the use of HEOs across different energy storage technologies, particularly layered HEOs for SIBs. For example, Dong et al.13 provided an in-depth evaluation of the advantages of high-entropy design, elucidating the underlying factors that enhance the electrochemical performance of layered HEO-based cathodes. Moreover, their work addresses key challenges and constraints related to the development and characterization of HEMs, offering valuable insights and directions for future studies and real-world implementation. In addition, Gao et al.14 summarized recent progress on layered high-entropy cathodes for SIBs, addressing the underlying high-entropy mechanisms, material design strategies, and prospective challenges, while highlighting their relevance for next-generation energy storage and conversion systems. Conversely, other recent reviews provide a broader and more detailed assessment of progress in layered HEO as electrode materials for Li/Na/K-ion batteries, systematically addressing synthesis strategies and their impact on structure and performance.15,16
These HEO-based cathode materials have been synthesized using various methodologies, including sol–gel, high-energy ball milling, coprecipitation, spray pyrolysis, spray drying followed by thermal treatment, and others. These techniques enable fine control over cation mixing, particle morphology, and crystallinity, all essential factors for stabilizing layered phases and optimizing electrochemical performance. Solution-based methods such as sol–gel and coprecipitation promote atomic-scale homogeneous cation distribution, while techniques like spray pyrolysis and spray drying can offer scalable routes to producing uniform particles with high reproducibility.17 For instance, Xiangnan Li et al.18 reported, for the first time, the synthesis of HEOs for SIBs using a spray-drying route. However, the authors referred to the process as a “prilling” method, focusing primarily on preparing the slurry before atomization. This terminology and emphasis obscure the central role of spray drying itself, hindering a clear evaluation of how this technique specifically contributes to the formation and optimization of HEOs. Consequently, their study does not provide a focused or detailed assessment of spray drying as an independent and scalable synthesis strategy for layered HEO-based materials.
Interestingly, evidence across metal-ion chemistries shows that spray-dried cathode active materials (CAMs) can match or exceed the practical energy, power, and cycling performance of many commercial cathode families when microstructure, conductive architecture, and composition are co-optimized, with the added advantages of scalability and potential cost reduction. For instance, spray-dried LMFP-type materials with optimized dispersant/surfactant and carbon coating can deliver as high as 159 mAh g−1 at 0.2C, with good high-rate capability.19 The doping approach in spray-dried Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2 further expands the Na interlayer distance and facilitates faster Na+ diffusion, resulting in a reversible specific capacity of 156.86 mAh g−1 and a capacity retention of 88.81% after 600 cycles at 1C.20 These metrics are competitive with many cathode materials considered for commercial energy storage devices.
Despite significant advances in material design via spray drying across a wide range of energy conversion and storage applications, to the best of our knowledge, only two review articles specifically address progress in the design of cathode active materials (CAMs). In this study, Vertruyen et al.,17 reviewed the growing literature on spray-drying methodologies, covering various processing routes and highlighting the roles of solution and suspension chemistry in different CAMs for LIBs and SIBs. The authors also addressed post-synthesis treatments and the associated granule morphologies and provided an extensive compilation of more than 300 studies, arranged by composition, precursor selection, and carbon-related aspects. More recently, Zheng et al.21 reported progress in the spray drying of micro-nanostructured materials for LIBs. The review outlines the fundamentals, evolution, and industrial status of SD in LIBs, highlighting strategies to improve electrode materials, while discussing its advantages, challenges, and potential for designing high-performance micro-nanostructured electrode materials.21
Despite the important findings highlighted in the review by Vertruyen et al.,17 the work covers only a limited number of NaxTMO2 compounds for SIBs, including publications up to 2018. Besides, to the best of our knowledge, no review has systematically addressed the promising results and performance comparisons among LEOs, MEOs, and HEOs (low-, medium-, and high-entropy oxides, respectively) used in SIBs prepared via spray drying. This cost-effective approach ensures homogeneous mixing and precise particle-size control, both of which are critical for improving the electrochemical performance of the resulting materials.22 This aspect is particularly critical in high-entropy systems, where elemental uniformity is essential to stabilize a single phase and to enable the emergence of characteristic configurational effects.
In this context, the present work aims to fill the existing gap in the literature while also highlighting recent advances in the field. Therefore, the focus of this work is to highlight recent studies on LEOs, MEOs, and HEOs synthesized via spray drying, to correlate their synthesis features with electrochemical properties, and to outline emerging trends in the design of entropy materials as new classes of cathode materials for SIBs, as summarized in Scheme 1. In addition, this review provides a comprehensive overview of spray-drying as a versatile and scalable synthesis strategy for cathode materials, with emphasis on its fundamentals, atomization techniques, and formulation chemistry. Furthermore, it introduces entropy-engineering concepts applied to layered NaxTMO2, highlighting how entropy design enhances structural stability and electrochemical performance. This review also critically analyzes spray-drying-assisted synthesis routes, precursor selection, and post-treatment effects on microstructure and performance of layered NaxTMO2. Finally, the conclusions are presented, along with perspectives and outlooks for future research directions.
![]() | ||
| Scheme 1 Summary of the main strategies and the different types of layered NaxTMO2 (low-, medium-, and high-entropy oxides) obtained by spray drying, which are reviewed in this work. | ||
Presenting the same charge and discharge mechanism as LIBs (based on ion insertion and de-insertion), SIBs utilize sodium ions as charge carriers between the anode and cathode.24 However, even though Li and Na belong to the same group of chemical elements and exhibit similar physicochemical properties, Na+ is significantly larger, with an ionic radius of 1.02 Å compared to 0.76 Å for Li+ 25. Despite the development of compounds with open frameworks capable of accommodating Na+, this size difference primarily results in greater volume expansion and contraction of the electrode's crystallographic structure.26 This leads to a lower energy density and a shorter lifespan compared to LIBs, as the insertion of Na+ can alter the phase and lattice of the host materials.27
Additionally, the working voltage window of a battery is a critical parameter that influences electrode stability, capacity retention, and battery pack design. SIBs generally operate at a lower average voltage and exhibit a lower cut-off voltage compared to LIBs, resulting not only in reduced energy density but also in lower power output, particularly at a lower state of charge.28 The literature on SIBs reports a typical working voltage window ranging from 1.5 V to 4.0 V, often characterized by voltage profiles with multiple plateaus that contribute to a relatively high specific capacity.29 In contrast, commercial LIBs generally operate with a higher lower cut-off voltage, typically between 2.5 V and 3.5 V, with either single or double plateaus depending on the electrode material.30 Increasing the lower cut-off voltage of SIBs raises their average operating voltage, which can enhance battery pack architecture and power output but at the cost of reduced available capacity.28
On the other hand, this lower voltage also brings potential benefits, such as enhanced safety by reducing the likelihood of thermal runaway. Xie et al.31 examined the thermal runaway characteristics of SIBs using accelerating rate calorimetry (ARC). Their results showed that SIBs equipped with a NaxNi1/3Fe1/3Mn1/3O2 cathode can reach a maximum self-heating temperature of 312.24 °C during a thermal runaway event, whereas LIBs using analogous layered oxides can reach 800 °C. Zhao et al.32 conducted a comparative evaluation of the thermal stability of the most used sodium salts, NaX (X = ClO4 and PF6), in EC/DMC and PC, comparing them with their lithium counterparts. Their findings demonstrated that Na-based electrolytes exhibit higher onset temperatures and lower heat release, suggesting that SIBs could potentially be safer battery systems than LIBs.
Moreover, the configuration of SIBs offers a potentially safer option for battery storage and transportation. Unlike LIBs, which require the use of Cu foil to collect the current at the anode (due to the formation of Li–Al alloy), SIBs can utilize Al foil instead. In LIBs, over-discharging below approximately 30% state of charge (typically 2.0–2.8 V) can lead to copper dissolution, which negatively affects cycling performance and cell safety. In contrast, the Al current collectors in SIBs remain stable even under overpotential conditions, including at 0 V, as demonstrated by Rudola et al.33 in their review of safe shipping requirements for recently commercialized SIBs. Additionally, it is worth mentioning that the use of Al as a current collector for SIBs is also an important cost advantage since Al foil is cheaper than Cu foil. This substitution can reduce total battery cost by approximately 3%.25 Among emerging battery chemistries beyond lithium, sodium-ion batteries (SIBs) represent a promising alternative, benefiting from more affordable and widely available raw materials while delivering comparable performance. They offer distinct advantages -such as the use of a cheaper (e.g., anode current collector), and improved safety – as well as some disadvantages, including lower cathode voltage and reduced energy density, as previously discussed. Consequently, SIBs have already reached commercialization, with both startups and established companies, such as Faradion, Tiamat, HiNa, Natron, Altris AB, and CATL, producing commercial cells at the amp-hour scale along with their respective modules.34
However, even though SIBs are already a reality, there remains significant potential for further development, particularly from an industrial perspective, to optimize the mass production of cathode materials.35,36 Current synthesis processes for layered oxides, polyanionic compounds, and Prussian blue analogues often involve multi-step, lengthy, and energy-intensive methods that hinder scalability. Commonly used techniques, such as solid-state, sol–gel, and coprecipitation methods, each pose distinct limitations. Solid-state and sol–gel processes frequently produce irregular, agglomerated particles with poor tap density and high surface area, resulting in performance that falls short of the materials' true potential.37 Meanwhile, the coprecipitation method involves a complex interplay of variables, particularly when incorporating transition metals, which demand precise control to achieve consistent quality.37
Achieving consistent particle morphology, high purity, optimal crystallinity, and economic viability at an industrial scale remains a critical challenge, as these attributes are directly linked to battery performance.38 In this context, the spray-drying method has emerged as a promising alternative due to its simplicity, cost efficiency, and scalability, making it well-suited for large-scale production, which can reach ton-scale quantities.39–41 This innovative approach offers the potential to enhance efficiency and commercial appeal for SIB manufacturing.
Although many of the current micro-nanostructured cathode materials are fabricated primarily using the coprecipitation-calcination, solid-state, sol–gel, and solvo/hydrothermal methods, as highlighted below, the use of spray drying to produce precursors for battery active materials may become a major trend in the battery market. In fact, these conventional synthesis routes are often limited by their complexity, high energy requirements, and other specific challenges.21,43 For instance, co-precipitation is the most widely used industrial method for producing cathode precursors; however, this process is sensitive to various parameters, including reaction temperature, pH, Ksp of different substances, solution concentration, ligand type, and stirring rate.44,45 Besides, co-precipitation has disadvantages, such as tight process tolerances, challenges with composition and impurity control, and scale-up-related inhomogeneities, all of which can degrade final cathode performance and raise manufacturing costs and complexity. Furthermore, the co-precipitation process generates a significant volume of alkaline ionic wastewater upon completion, undermining sustainability.44 While spherical particles designed by co-precipitation methods are generally preferred due to their efficient packing and reduced electrode porosity, and because they improve overall performance for both SIBs and LIBs,46 maintaining precise control over particle growth and composition is challenging due to concentration gradients.47
In the sol–gel method, organic acids (e.g., citric acid) are typically introduced as chelating agents, along with additional energy input for solvent evaporation and gel formation,43 as well as metal alkoxide precursors with high-purity (purity
as high as 99.99%).21 Moreover, the resulting particles are generally in the submicrometer size range.43 On the other hand, solvo/hydrothermal methods are long batch processes, hard-to-monitor parameter spaces, require complex multi-step processing, use costly high-pressure equipment, and pose solvent/waste handling issues. Besides, in some cases, solvo/hydrothermal methods impose higher environmental and cost burdens than dry or simpler wet routes. Solid-state synthesis typically involves high-temperature calcination of mechanically mixed precursors, often resulting in poor control over morphology. Moreover, high temperatures and prolonged calcination are required to obtain an active cathode material.48 Besides, the materials synthesized by this method do not achieve complete atomic-level uniformity, and the morphology and particle size are difficult to control.49
Fortunately, spray-drying is one of the most promising routes for industrial cathode production thanks to its morphology control, ease of carbon integration, scalability, and strong electrochemical performance, but it requires rigorous control of solution formulation and thermal treatment to avoid phase, microstructural, and reproducibility issues. In addition, spray-drying equipment is already widely commercialized by several companies for various applications, indicating its ease of implementation and adaptation for large-scale design of cathode active materials. The precursor tends to exhibit high porosity due to the rapid solvent evaporation during spray drying. Consequently, achieving a high tap density is challenging, and morphological changes are expected during calcination.43 On the other hand, the design of porous active materials can shorten the ion diffusion path and inhibit interfacial side reactions, thereby mitigating volume expansion and improving rate performance.21 In addition, some recent studies have focused on developing spray-drying methods to design electrode materials with high tap density,50,51 suggesting that even more significant advances can be expected for this highly versatile synthesis approach.
Interestingly, some recent studies have indeed demonstrated the prospects and advantages of using spray drying in the design of cathode active materials compared to other synthesis methods. A comparative study by Canini et al.22 investigated the influence of the synthesis route on Mg-doped P2-Na0.67Fe0.5Mn0.5O2, contrasting spray-drying with conventional solid-state synthesis. While both samples exhibited comparable overall morphology, the spray-dried material showed a narrower particle-size distribution and smaller grain size, both of which are advantageous for shortening Na+ diffusion pathways. In fact, the estimated Na+ diffusion coefficient (DNa+) for P2-Na0.67Mn0.5Fe0.3Mg0.2O2 synthesized via spray drying is over two orders of magnitude greater than that obtained through solid-state synthesis (10−8vs. 10−10 cm2 s−1). More importantly, distinct differences in phase composition were identified through Rietveld refinement. The sample prepared by solid-state contained an inactive spinel-like MnFe2O4 secondary phase, whereas the sample designed by spray-drying favored the formation of an electrochemically active layered O3 phase. This difference highlights the superior capability of spray drying to control phase purity and crystallographic structure. As a result, the spray-dried sample delivered enhanced electrochemical performance, achieving 81% capacity retention after 100 cycles at 1C, compared to 72% for the solid-state counterpart.
![]() | ||
| Fig. 1 (a) Two-fluid nozzle-type atomizer representation. Reproduced with permission from ref. 54. Copyright © 2010 The Society of Powder Technology Japan. Published by Elsevier B.V. All rights reserved. SEM images of precursor samples: (b) pristine sample; (c) SDCNCA1 (1% of Al2O3), (d) SDCNCA2 (2% of Al2O3), (e) SDCNCA5 (5% of Al2O3). SEM images after spray drying treatment: (f) pristine sample; (g) SDCNCA1, (h) SDCNCA2, and (i) SDCNCA5. Reproduced with permission from ref. 56. Copyright © 2016 American Chemical Society. (j) Rotary disk nozzle-type atomizer representation. Reproduced with permission from ref. 54. Copyright © 2010 The Society of Powder Technology Japan. Published by Elsevier B.V. All rights reserved. SEM images of prepared materials after spray drying technique (k) Na2FePO4F/CNT powder, (l) sprayed Na2FePO4F/C powder without CNT, and after thermal treatment at 600 °C under Ar atmosphere (m) Na2FePO4F/CNT, (n) Na2FePO4F/C. Reproduced with permission from ref. 63. Elsevier has partnered with Copyright Clearance Center's RightsLink service to offer a variety of options for reusing this content. (o) Ultrasonic nozzle-type atomizer representation. Reproduced with permission from ref. 54. Copyright © 2010 The Society of Powder Technology Japan. Published by Elsevier B.V. All rights reserved. SEM images of the LiFePO4 prepared at different temperatures using the spray drying technique (p) 200 °C, (q) 250 °C, (r) 300 °C, and (s) 350 °C. Reproduced with permission from ref. 62. Copyright © 2021 by the authors. Licensee MDPI, Basel, Switzerland. | ||
The rotary nozzle atomizer consists of a horizontal disk, which can be either straight or curved.57 To minimize nozzle clogging, rotary nozzle atomizers operate based on centrifugal force (typically 5–60k rpm), propelling the feed solution through the nozzle channel. In this process, the liquid feed is pumped through the atomizer's center and spreads radially before disintegrating into droplets at the periphery (Fig. 1j).57 Due to air turbulence and liquid feed properties, rotary nozzle atomizers generate spherical and highly homogeneous droplets.58 This type of atomizer allows the production of large quantities (up to 200 tons per hour) and enables the handling of highly viscous feed solutions.57 Moreover, droplet size can be precisely controlled by adjusting the wheel's rotation speed.58 However, drawbacks include high energy consumption and significant maintenance requirements.58,59
The rotary disk atomizer has also been utilized in battery material synthesis, such as in the preparation of Na2FePO4F/CNT composites. The synthesis involved mixing iron powder, citric acid, glacial acetic acid, and water, followed by combining NH4H2PO4, NaF, and NaOH in equal proportions. Multi-walled carbon nanotubes were then incorporated in a 10
:
2 ratio to form Na2FePO4F/CNT. The final step involved spray-drying at 140 °C, followed by thermal treatment at 600 °C for 12 h under an argon atmosphere. SEM images (Fig. 1k–n) reveal that the material consists of collapsed particles (∼15 mm in size), whereas well-defined particles are observed in the absence of CNTs.60
Ultrasonic nozzle atomizers employ ultrasonic vibrational energy to atomize liquid feeds (Fig. 1o).57 For industrial applications, high solid concentrations in the feed are preferred to reduce solvent usage and conserve energy.57 However, increased solid concentration also raises feeding viscosity, increasing the likelihood of clogging and nozzle blockages.61 A major advantage of ultrasonic nozzles is their larger orifice, which reduces clogging risks compared to conventional nozzle atomizers that require smaller orifices to pressurize the flow.58 Furthermore, ultrasonic nozzles offer precise control over droplet morphology and size distribution due to their large probe and exit orifice.61 Despite these benefits, ultrasonic nozzles have limited throughput and may face viscosity limitations in certain applications. For instance, in the work of Lan et al.,62 LiFePO4 precursor particles were prepared for Li-ion batteries by the spray drying method using an ultrasonic nozzle. To achieve this, the authors mixed LiH2PO4, FeCl2·4H2O, LiOH·H2O, with HCl, sucrose, and deionized water. The atomization was performed at 1.7 MHz and then treated at different temperatures ranging from 200 to 350 °C under a 5% H2/Ar atmosphere. After that, a second thermal treatment was carried out under the same atmosphere at 650 °C for 8 h. The obtained particles (Fig. 1p–s) were spherical for all prepared samples, with particle sizes below 1 µm.
After the atomization process, the droplets undergo solvent evaporation and particle formation, facilitated by heat treatment, with parameters such as carrier gas flow, chamber geometry, and temperature carefully optimized for uniform particle formation.64 Particle collection is carried out using equipment such as cyclones, filter bags, or electric field precipitators, ensuring recovery of solvent-free particles with desired structural and compositional characteristics.65–67 To illustrate the aerosol-assisted spray-drying process, Fig. 2a and b provides a schematic representation of the spray-dryer and the method's mechanism. The process begins with atomization, where the initial solution or slurry (precursor) is converted into fine droplets. A carrier gas transports these droplets into the drying chamber, where solvent evaporation occurs. This step transforms the droplets into solid particles through rapid heat exchange. Finally, the dried particles are collected using suitable equipment, such as a cyclone or filter bag. The main steps involved in the process were atomization, droplet-to-particle conversion (by solvent evaporation), and particle collection. Fig. 2b illustrates the flow and integration of the components in the spray-drying system. A visual representation reinforces the versatility and precision of the method, indicating how heat and mass transfer phenomena are used to achieve efficient particle formation.
![]() | ||
| Fig. 2 (a) Diagram of a spray-dryer, demonstrating a co-current setup with bi-fluid nozzle atomization. Reproduced with permission from ref. 17. Copyright © 2018 by the authors. Licensee MDPI, Basel, Switzerland. (b) Conceptual illustration of the spray-drying technique for synthesizing nanostructured particles, highlighting the operational mechanism of the apparatus. Reproduced with permission from ref. 54. Copyright © 2010 The Society of Powder Technology Japan. Published by Elsevier B.V. All rights reserved. | ||
The aerosol-assisted spray-drying method presents significant advantages, making it highly attractive for material synthesis and industrial applications. Its primary benefits include its versatility, allowing compatibility with a wide range of materials and precursors.42 The method is economical and enables continuous, high-throughput processing,64 making it suitable for large-scale production. It can produce particles with high purity, uniform spherical shapes, and minimal agglomeration, which are essential characteristics for applications requiring consistent quality.68 Additionally, the method offers precise control over particle size, ranging from the nanoscale to the micron scale, facilitating the fabrication of nanostructured powders.69 The adaptability of spray-drying to various configurations and atomization techniques further enhances its usability across multiple industries, including food processing, pharmaceuticals, and advanced ceramics.70 The aerosol-assisted spray-drying method, a variant that integrates aerosol-assisted self-assembly with drying, has gained traction for producing high-purity, agglomeration-free spherical particles with relatively monodispersed sizes. This method is widely applied in industries such as food processing, fertilizers, ceramics, and pharmaceuticals, supported by over 15
000 industrial-scale spray dryers in operation globally.70 However, aerosol-assisted spray-drying method also has limitations. The process consumes a considerable amount of energy due to the heat treatment required for solvent evaporation.68 Optimizing process parameters, such as temperature, carrier gas flow, and residence time can be at the same time critical and complex, once requires careful calibration of all these parameters.71,72 Certain configurations may limit the processing of heat-sensitive materials, which can degrade during solvent evaporation.73 Additionally, the choice of atomizer and the properties of the precursor, such as viscosity and surface tension, significantly influence droplet formation and particle uniformity.74 These factors may pose challenges for certain applications or materials with stringent requirements.
| Aspect | Conventional spray-drying | Improved spray-drying techniques (e.g., rapid-solvent evaporation, additive-assisted spray) |
|---|---|---|
| Particle size control | Controlled by droplet size and precursor concentration75 | Enhanced control via rapid solvent evaporation or additive addition76,77 |
| Production rate | High production rate | Varies depending on technique; often lower for nanoparticle focus17,71 |
| Nanoparticle production | Limited by minimum droplet size (∼4.5 µm)71,78 | Enables smaller particle sizes through mechanisms like pressure-driven agglomeration breakup76–78 |
| Mechanism for nanoparticles | Controlled by reducing precursor concentration79 | Rapid nucleation/crystal growth (denoted as R1) or additive use to prevent agglomeration (denoted as R2)76,77 |
| Adaptability for industrial use | Requires additional optimization for nanoparticle production40 | High adaptability with specific equipment improvements76,77 |
| Cost of implementation | Relatively low due to standard equipment use80 | Potentially higher due to specialized techniques and materials81 |
Enhanced techniques offer greater adaptability for industrial applications by optimizing precursors and process parameters to achieve efficient nanoparticle production. Despite higher costs, their ability to scale up fabrication, particularly using rapid-solvent evaporation/nucleation (Route 1, R1) or the additive-assisted spray method (Route 2, R2), makes them valuable for advanced material synthesis.76
The rapid-solvent evaporation technique (R1) promotes quick nucleation and crystal growth through fast solvent elimination. The short drying time limits particle agglomeration.78 Additionally, thermal reactions and a high drying rate generate vapors of solvents or gases, creating huge internal pressures that break agglomerates and disperse nanoparticles efficiently.76 However, the optimization of the precursor is essential for maximizing dispersion, to the point that inadequate selection can significantly reduce the effectiveness of the process.82 The additive-assisted spray method (R2) provides a cost-effective and simple approach to nanoparticle production. By adding compounds such as salts,77 polymers,83 or low-boiling-point chemicals84 before spraying, standard spray-drying equipment can be used, reducing capital and processing expenses. The additive (R2) facilitates composite particle formation, preventing nanoparticle agglomeration and ensuring individual dispersion. After solvent dissolution or heat treatment, the nanoparticles are extracted. However, the need for additive removal adds complexity and may require additional purification steps.
The formulation of solutions and suspensions for spray-drying is a pivotal step in synthesizing precursor materials, which involves careful consideration of both inorganic and organic/carbon components, particularly for applications like electrode materials in energy storage devices. Solutions offer a high degree of homogeneity because soluble salts and precursors distribute evenly in the liquid phase. However, they are constrained by solubility limits, and the use of highly acidic media can pose risks to spray-drying equipment. On the other hand, suspensions provide flexibility to incorporate insoluble or partially soluble precursors, enabling the use of materials such as oxides and carbonates.85,86
Solvent selection significantly influences the process. Water, being inexpensive and environmentally friendly, is the most widely used solvent.85,87 Alcohols, such as ethanol, are also employed, either alone or mixed with water, to expand solubility ranges; however, they require additional safety measures due to their flammability and potential for hydrolysis reactions with certain precursors.88–90 Soluble organic molecules, such as citric acid, have multiple functions, acting as complexing agents, reducing agents, or binders.17,91–94 Synthetic polymers such as polyethylene glycol (PEG), polyvinyl alcohol (PVA), and polyvinylpyrrolidone (PVP) act as dispersants and binders, improving particle stability and mechanical integrity.95–97 These compounds also serve as precursors for carbon coatings during heat treatment in an inert atmosphere, enhancing the electrical conductivity of the resulting material.17
A key motivation for incorporating carbon is to address the low intrinsic conductivity of many electrode-active materials, enabling better electron transport and minimizing the volume changes during electrochemical cycling. Citric acid, saccharides like glucose, and other carbon precursors are commonly used to form graphitic carbon coatings. Carbon nanotubes (CNTs) and graphene oxide (GO) are also integrated into formulations to improve conductivity and mechanical properties. Concentration optimization is critical for both solutions and suspensions. While higher concentrations reduce the volume of solvent to be evaporated and can enhance energy efficiency, they may also introduce operational challenges, such as gel formation, precipitation, or nozzle clogging.98–100 Complexation agents, such as citrates, are frequently used to prevent unwanted precipitation and enhance the homogeneity of multicomponent solutions.98,101 For suspensions, maintaining stability is essential. Dispersants can enhance particle dispersion and prevent aggregation, while methods such as ball milling can be used to achieve finer particle sizes.98,100 However, challenges arise with multicomponent suspensions, where varying particle sizes or differences in chemical properties may lead to distribution gradients in spray-dried granules.102–104
Additives such as binders are often included to improve the cohesion and mechanical strength of the granules, but their presence increases viscosity, complicating the spray-drying process.90,104 Additionally, considerations such as the hygroscopic nature of some precursors (e.g., nitrates and acetates) or the aging of as-sprayed materials require careful control of post-spray-drying storage conditions.105
Solid-state reactions are widely employed due to their operational simplicity and suitability for large-scale industrial production; however, they proceed primarily through cation diffusion between solid particles at elevated temperatures. In complex systems such as layered HEOs, this diffusion-controlled mechanism often results in incomplete reactions, compositional heterogeneity, and the formation of undesired secondary phases rather than a uniform single-phase material. Moreover, solid-state processing typically leads to the formation of dense, irregular agglomerates, which hinder the achievement of well-dispersed particles with controlled morphological features.107 Attaining the atomic-scale compositional uniformity required for HEOs via this approach generally requires prolonged high-temperature calcination combined with repeated mechanical milling. This process is highly energy-intensive and may still fail to ensure complete compositional homogeneity.37
Co-precipitation is often favored because it allows intimate mixing of different metal species at the molecular level within a liquid phase. Nevertheless, applying this strategy to HEO systems presents substantial chemical challenges. Achieving the simultaneous precipitation of five or more transition metals, such as Ni, Fe, Mn, Ti, and Co, requires careful control of pH, temperature, and reagent concentrations, due to the differences in the solubility products of their respective hydroxides or carbonates. As highlighted in previous studies, the co-precipitation of multiple metal ions in HEO systems is far from a simple process, and the resulting materials do not achieve the targeted equimolar composition.107 Even relatively small deviations in precipitation behavior can lead to sequential or fractional precipitation, in which certain cations precipitate preferentially, resulting in compositional inhomogeneities in the final material.37 Such heterogeneity directly undermines entropy-based design by introducing local variations in configurational entropy, which can ultimately drive phase separation during subsequent calcination.
Spray drying provides an effective means of addressing several of the limitations associated with conventional synthesis routes. The process begins with the preparation of a homogeneous solution or suspension containing all required metal precursors, typically in the form of nitrates or acetates. In this case, the precursor salts are fully dissolved in the liquid phase, and the atomized droplets retain the same elemental proportions as the initial feed solution. As solvent evaporation proceeds rapidly, these droplets transform into solid particles with homogeneous mixing that preserve the uniform initial cation distribution. This rapid drying process reduces compositional segregation, which often complicates co-precipitation methods. As a result, the retention of solution-phase homogeneity allows single-phase HEOs to form after significantly shorter thermal treatments than those generally required for conventional solid-state synthesis.107 Importantly, spray drying can be viewed as a conceptual link between solution-based chemical processing and scalable solid-state manufacturing, as it preserves molecular-level mixing throughout the dried precursor matrix.54 In the literature, it is reported that spherical secondary particles form from the aggregation of nanoscale primary crystallites. This aggregation phenomenon produces hierarchical structures with decreased Na+ diffusion pathways and improved electrochemical performance.108,109 The combination of a high degree of elemental uniformity inherent to the spray-drying process with high control over morphology is especially beneficial for layered HEOs. As a consequence, the maximization of configurational entropy and the stabilization of a single-phase framework rely critically on homogeneous cation distribution.110,111
Despite these advantages, spray drying has some limitations. The rapid solvent evaporation characteristic of this method, in some cases, produces hollow microspheres, which can reduce tap density relative to materials prepared for conventional routes. In addition, many studies provide limited details related to key parameters during the synthetic route, making it difficult to compare across different reports.112,113 However, recent efforts have focused on optimizing the spray drying process, with several strategies proposed to enable rapid production of spherical precursors and electrode materials with high tap density and minimal structural collapse.44,50
While the underlying principles of spray drying are well documented, its application to battery cathode precursors still needs to be explored further, particularly regarding how specific processing variables affect physicochemical properties of the material. In fact, only one review specifically addresses the general chemical aspects of solution/suspension formulation for spray drying, taking into account the post-processing of the resulting precursors and the granule morphologies obtained for battery electrode materials. More comprehensive details can be found in the review by Vertruyen et al.17
As this review reveals, the level of detail in reporting these parameters varies widely across studies, and only a handful provide sufficiently complete synthesis descriptions to enable meaningful cross-comparisons. For instance, Mahmoud and colleagues108 reported a comprehensive synthesis of data, specifying an inlet temperature of 140 °C, a feed rate of 25 mL min−1, and post-treatment at 600 °C for 12 h under argon to produce Na2FePO4F-carbon composites. In contrast, other works omit key spray-drying conditions. Kim et al.,113 for example, reported a calcination temperature of 750 °C but omitted the drying parameters entirely. Similarly, Häringer et al.112 systematically varied processing parameters without reporting their exact values, while Boutelle et al.114 emphasized adjustments to the milling protocol and carbon source without specifying the spray-drying conditions.
The specific influence of spray-drying inlet and outlet temperatures on the synthesis of the cathode precursor has not yet been systematically investigated. Nevertheless, comparisons among studies employing similar parameter ranges reveal general trends relating processing conditions to particle morphology and electrochemical performance. In particular, spray-drying inlet temperatures in the range of 140–265 °C mainly affect droplet evaporation kinetics and particle formation, typically yielding spherical particles with diameters between 2 and 30 µm.108,115,116 For example, several studies have reported similar operating conditions, with inlet temperatures around 180–200 °C and outlet temperatures between 80 and 108 °C.109,115,116 In the work of Schmidt et al.,109 four P2-type cathode materials with hierarchical secondary particles (aggregates of smaller primary crystallites) were synthesized: Na0.6Al0.11Ni0.11Mn0.66Fe0.11O2 (NANMFO), Na0.6Al0.11Ni0.22Mn0.66O2 (NANMO), Na0.6Al0.11Fe0.22Mn0.66O2 (NAFMO), and Na0.6Mn2/3Fe1/3O2 (NMFO). Spray drying was conducted at inlet and outlet temperatures of 200 °C and 108 °C, respectively, followed by calcination at 950 °C. The resulting materials displayed spherical particles with diameters ranging from 3 to 30 µm (Fig. 3a). All compositions crystallized in the P2 structure with hexagonal space group P63/mmc, and lattice parameters increased with Fe substitution. Among them, NANMFO exhibited the best electrochemical performance, delivering 189 mAh g−1 at 20C and 148 mAh g−1 at 1C, with 86.5% capacity retention after 100 cycles and approximately 50% at 10C (Fig. 3c). This behavior was attributed to a reduced Na+ diffusion barrier associated with decreased structural ordering. Notably, these electrochemical improvements arise primarily from the post-spray-drying calcination step rather than from the droplet-drying process itself. In contrast, the relatively narrow range of reported spray-drying temperatures suggests that these parameters primarily influence particle formation kinetics rather than phase crystallization.
![]() | ||
| Fig. 3 (a) SEM of the particles corresponding to Na0.6Al0.11Ni0.22Mn0.66O2 (NANMO), Na0.6Al0.11Ni0.11Mn0.66Fe0.11O2 (NANMFO), Na0.6Al0.11Fe0.22Mn0.66O2 (NAFMO), and Na0.6Mn2/3Fe1/3O2 (NMFO) materials after sintering at 950 °C; (b) particle porosity (Φ, %) and specific surface area (ABET, m2 g−1) complement the high magnification SEM images. (c) Cross-sectional view of the electrode's secondary particles reveals that porosity decreases as the composition shifts from nickel-rich (NANMO) to iron-rich (NMFO). Reproduced with permission from ref. 109. Copyright © 2024 the Authors. Advanced Energy Materials published by Wiley-VCH GmbH. | ||
On the other hand, the work by Schmidt et al.109 provides a representative example of how the precursor composition of the active material can influence the microstructure of hierarchically structured particles, particularly their porosity and specific surface area. When Ni is partially substituted by Fe, both the porosity and surface area decrease. This behavior is attributed to the higher sintering activity associated with increased iron content, which also promotes grain growth. As a result, NMFO exhibits the largest grain size and the lowest internal porosity within the granules, leading to poor rate capability and the lowest specific capacity.
Additional parameters such as feed rate, atomization pressure, nozzle geometry, and spray duration also influence droplet size and evaporation dynamics. For instance, Mahmoud et al.108 used a feed rate of 25 mL min−1, while R. R. Pothi, K. Rajasekar & B. Raja117 investigated mass flow rates between 3 and 3.9 kg h−1. In contrast, Koo et al.118 focused on suspension concentration, varying it between 0.5 and 8.0 g L−1. Their findings revealed a morphological shift from compact particles at lower concentrations to porous, sponge-like architectures at higher concentrations, with average pore diameters expanding from 12.5 µm to 23.7 µm.118 This structural evolution was attributed to the manner in which solid constituents accumulate during droplet drying. Although atomization pressure is seldom detailed in the literature, some studies have begun to address its role. K. Rajasekar & B. Raja,117,119 for instance, examined pressures spanning 4 to 7 bar, whereas Koo et al.118 employed a notably lower pressure of roughly 0.3 bar. In the same work, R. R. Pothi, K. Rajasekar & B. Raja117 also varied nozzle diameters between 0.2 and 0.4 mm, spray durations from 10 to 40 s, and applied a heating interval of 120 s, with an inlet air velocity of 0.88 m s−1. Elevating the pumping pressure enhanced atomization efficiency and boosted volumetric heat transfer coefficients to the range of 1.5–3.5 kW m−3 K, thereby accelerating solvent evaporation and shaping particle formation dynamics.119
On the other hand, it is well known that precursor chemistry (raw materials and/or reactants), especially the choice of carbon sources, along with carbon additives, binders, and calcination temperatures, plays a decisive role in shaping the morphology of spray-dried particles. These variables dictate whether the resulting structures take the form of irregular aggregates, dense spheres, hollow morphologies, or yolk–shell architectures, each with a marked impact on electrochemical performance. A case in point is the work by Tang et al.,120 who synthesized Na2Fe0.6Mn0.4PO4F/C using various carbon sources, namely oxalic acid, ascorbic acid, citric acid, and glucose. SEM from Fig. 4a revealed that ascorbic acid led to the formation of uniform hollow spheres approximately 1–2 µm in size, while the other precursors resulted in irregular particle morphologies. Despite this structural consistency, their electrochemical behavior varied considerably. Citric acid and glucose yielded the poorest performance, with capacities of just 30 and 24 mAh g−1, respectively. Oxalic acid delivered a moderate 65.7 mAh g−1, whereas ascorbic acid stood out with a capacity of 95.1 mAh g−1 and an impressive 91.7% capacity retention after 100 cycles. The improved performance was attributed to the hollow particle architecture, which reduced charge-transfer resistance and facilitated Na+ transport. The study demonstrates that careful selection of carbon sources during spray drying can precisely control particle morphology, with hollow spherical structures providing optimal electrochemical performance through enhanced ion transport and reduced resistance.120
![]() | ||
| Fig. 4 (a) SEM images of Na2Fe0.6Mn0.4PO4F/C materials derived from different sources: (a1) oxalic acid; (a2) ascorbic acid; (a3) citric acid; (a4) glucose. Reproduced with permission from ref. 120. Copyright © 2021 Tang, Zhang, Sui, Wang, Li, and Wu. (b) SEM images of (b1) the precursor of Na0.67Ni0.11Cu0.11Fe0.3Mn0.48O2 (NCFM), and the materials after calcination at different temperatures: (b2) 700 °C, (b3) 750 °C, (b4) 800 °C, (b5) 850 °C, and (b6) 900 °C; and SEM images of (b7) CFM-750 (Na0.67Cu0.22Fe0.3Mn0.48O2), (b8) FM-750 (Na0.67Fe0.41Mn0.59O2), and (b9) NFM-750 (Na0.67Ni0.22Fe0.3Mn0.48O2). Reproduced with permission from ref. 1. Copyright © 2024, American Chemical Society. | ||
More complex morphologies, such as yolk–shell structures with nitrogen-modified carbon coatings121 and hollow core–shell structures,122 have also been reported. For instance, Zhou et al.121 synthesized a yolk–shell Na3V2(PO4)2F3 cathode with Cu substitution, delivering 117.4 mAh g−1 at 0.1C with 91.3% retention after 5000 cycles. This behavior is explained by the synergistic modulation of crystalline structure and interfacial properties, which enhanced intrinsic and interfacial conductivity, while the yolk–shell morphology with nitrogen-modified carbon coating reduced polarization potential through improved Na+ diffusion.121 Similarly, Qi et al.122 reported hollow core-shelled Na4Fe2.4Ni0.6(PO4)2P2O7 particles formed through viscosity differences between citric acid and polyvinylpyrrolidone, enabling 86.4 mAh g−1 at 25 °C and stable cycling at −20 °C. These studies illustrate that hierarchical architecture requires precise control of precursor chemistry and drying dynamics, which partly explains why simpler spherical morphologies remain predominant in large-scale synthesis. Calcination temperature also strongly affects phase composition and particle size evolution after spray drying. For example, Liang et al.1 prepared Na0.67Ni0.11Cu0.11Fe0.3Mn0.48O2 (NCFM) by spray drying at 170 °C inlet and 80 °C outlet temperatures, followed by calcination at 700–900 °C. It was observed that calcination temperature governs particle size evolution from nanoscale to microscale, with nanoscale flake-like morphology at intermediate temperatures (750 °C) outperforming phase-pure microscale materials (Fig. 4b). The NCFM-750 sample showed the best performance due to an optimal balance between crystallinity and diffusion path length, delivering 160.3 mAh g−1 at 0.1C with 68.8% retention after 300 cycles, while maintaining 102.6 mAh g−1 at −20 °C.1
In general, although the previously reported complex morphologies exhibit appealing properties, it is well established that hollow spherical structures not only lower the tap density but also cause significant irreversible capacity loss due to pronounced interfacial side reactions.44 Besides, lowering the tap density can negatively impact tap density and, consequently, volumetric energy density. To address this issue, Liu et al.44 reported a rapid method for producing spherical precursors with high tap density and minimal structural collapse (Fig. 5). The optimized spray drying approach resulted in precursors with improved structural integrity, achieved through kinetic control of the process by incorporating suitable binders and dispersants. The authors emphasized that incorporating suitable binders can regulate solute interactions, induce pre-compression among nanoparticles, mitigate Darcy stress arising from solvent diffusion, and improve the mechanical robustness of the shell layer. Meanwhile, dispersants can enhance shell permeability and reduce concentration gradients. As a result, the drying process proceeds without structural collapse due to improved mechanical stability and without significant solute accumulation owing to enhanced permeability. Consequently, the shell layer progressively thickens toward the particle core, ultimately leading to the formation of dense, solid spherical particles.44 This result clearly demonstrates that the composition of the precursor solution or suspension is a key factor governing the morphological properties of the dried material, which in turn strongly influences the final active material obtained after calcination.
![]() | ||
| Fig. 5 Schematic diagram illustrating the morphological changes induced by varying parameters in the tailored spray drying processes. Reproduced with permission from ref. 44. Copyright © 2025 Wiley-VCH GmbH. | ||
Despite variations among studies, three consistent observations emerge: (1) spray drying reliably produces spherical secondary particles in the 2–30 µm range;108,109,115,116 (2) hierarchical structures can be generated but usually require tailored precursor formulations,118 and (3) crystalline phase formation depends primarily on post-spray calcination temperature rather than the spray-drying temperature itself, with critical temperatures varying among materials.
Future studies should systematically vary spray-drying parameters while maintaining constant calcination conditions. Such experiments would clarify the mechanistic role of spray-drying parameters in particle formation and electrochemical behavior, helping distinguish true processing effects from variables such as precursor chemistry or crystallization conditions.
From an industrial perspective, co-precipitation processes are routinely implemented in continuous stirred-tank reactors operating at production scales of approximately 6500 kg per day of cathode precursor materials, followed by calcination in rotary or tunnel kilns.123 In contrast, techno-economic analyses indicate that spray-based processes can be scaled from pilot production of ∼4 kg per day to comparable industrial capacities of ∼6500 kg per day.124 Moreover, spray-based approaches can significantly reduce manufacturing complexity by integrating mixing, particle formation, and precursor structuring into a single step, compared to the more than ten unit operations typically required in co-precipitation routes.124,125
Techno-economic analyses of flame-assisted spray pyrolysis provide important benchmarks for aerosol-based processing of cathode materials. Reported minimum cathode material selling prices are approximately $19 kg−1, corresponding to ∼17% lower costs than co-precipitation, primarily due to simplified process flows and reduced operational expenditures.124,125 Despite comparable capital expenditures (CAPEX), driven by the cost of high-temperature reactors and thermal treatment units, operational costs (OPEX) are substantially reduced, with non-material-related expenses estimated at ∼43% of those of co-precipitation due to simplified equipment chains and reduced labor requirements.124 From an environmental perspective, spray-based synthesis also offers significant benefits, including reductions of up to ∼60% in CO2 emissions and ∼30% in water consumption, reinforcing its potential for more sustainable manufacturing.125 Spray drying, for instance, suppresses the generation of large amounts of alkaline ionic waste solution, which is generally produced at the end of the preparation process by the co-precipitation method.44
Although dedicated techno-economic analyses for spray drying in cathode production remain scarce, this analysis is not the focus of the present review. However, insights from related industrial sectors indicate that spray drying is an energy-intensive process, with specific energy consumption typically in the range of ∼3000–5500 kJ kg−1 of evaporated solvent.126 Compared to co-precipitation, the overall energy demand can be lower, as conventional co-precipitation routes involve multiple processing steps and prolonged high-temperature calcination.125 For instance, compared with the hydroxide co-precipitation route, flame synthesis can reduce water consumption and lower CO2 emissions while avoiding the generation of solid or liquid waste. This approach, therefore, offers significant advantages in terms of energy efficiency and emission reduction in the production of high-nickel cathode materials (layered oxides).125
Despite its relatively high thermal demand, spray drying shares several characteristics with spray pyrolysis, notably its continuous operation and the ability to convert liquid precursors into dry powders in a single step, thereby reducing process complexity and equipment requirements. However, the need for a subsequent calcination step partially offsets these benefits. As such, spray drying is expected to offer competitive techno-economic performance, although quantitative assessments specific to cathode manufacturing are still required.
The immediate industrial adoption of spray-drying technologies remains challenging in the energy storage sector. Current battery manufacturing infrastructure is highly optimized for conventional processes such as co-precipitation and solid-state synthesis, which are deeply integrated into continuous production lines and supported by mature supply chains. Transitioning to spray-based methods would therefore require significant capital investment and process redesign, including modifications to precursor handling and thermal processing. Consequently, spray-based technologies can be more realistically positioned as complementary or hybrid approaches in the near term rather than direct replacements for established industrial synthesis routes. In practice, spray drying is already integrated as a precursor-forming step prior to calcination, particularly in the production of polyanionic cathodes such as LiFePO4 (LFP) and analogous sodium-ion materials.127,128 For example, the large-scale synthesis of sodium vanadium phosphate fluoride (NVPF) and NASICON-type compounds (e.g., NFPP) commonly relies on spray-drying-assisted solid-state processes, underscoring their compatibility with existing manufacturing workflows.128
Beyond these implementations, spray-based processing is increasingly employed in upstream powder engineering to tailor particle morphology and enhance the mixture of metal-based precursors.129 This capability is particularly advantageous for advanced material systems, such as medium- and high-entropy layered oxides, where strict compositional homogeneity and tailored particle architecture are essential. Continued advances in reactor design, process integration, and scale-up strategies will be critical to facilitate broader industrial adoption and to fully unlock the potential of spray-based synthesis for next-generation sodium-ion batteries.
![]() | ||
| Fig. 6 (a) The structural evolution among P2, P3, O3, O2, and OP4 phases is illustrated based on the gliding of TMO2 slabs and the resulting changes in the oxygen stacking sequence. (b) and (c) These transitions modify the coordination environment of Na+ ions between prismatic and octahedral sites, and a schematic illustration of sodium migration paths in different structures. Panel (a–c): adapted with permission from ref. 8. Copyright © 2023 The Authors. InfoMat published by UESTC and John Wiley & Sons Australia, Ltd. Panel (b and c): adapted with permission from ref. 133. Copyright © 2022 Wiley-VCH GmbH. | ||
In comparison, the Na+ ions in the O3-type are located at octahedral sites (edge-sharing) with an AB, CA, and BC layers because the ionic radius of sodium ions 1.02 Å is much larger than those of 3d transition-metal ions with a trivalent state <0.7 Å (Fig. 6).26,132 The structure is therefore classified as a 3R phase with a space group of R3m.26 In terms of cycling stability and rate capabilities, the P-type structure exhibits superior performance to the O-type due to the low diffusion barrier of Na ions between the two trigonal prismatic sites. However, the O-type presents a higher initial capacity due to its higher initial content of Na ions.133Fig. 6a illustrates that when a certain amount of Na+ ions is extracted from P2-NaxTMO2, the prismatic sites devoid of Na+ become thermodynamically unstable. The structural change occurs reversibly through the gliding of TMO6 octahedral sheets, leading to the generation of octahedral sites and the transformation into an O2-type phase with AB and AC oxygen layers. Besides the reversible P2–O2 transition, the P2-type structure can also change into an OP4 phase under certain conditions. This phase has stacked faults and a low-energy configuration. In the OP4 structure, octahedral and prismatic layers stack alternately along the c-axis, providing a mixed environment for Na+ ions. This arrangement helps to reduce lattice strain and improve structural stability during Na+ insertion and extraction. Sodium extraction from the O3-type phase typically triggers phase transitions. Partial “desodiation” leads to the formation of vacancies, making Na+ ions at prismatic sites energetically favorable. The formation of these prismatic sites occurs through the gliding of TMO2 slabs without breaking TM–O bonds. Consequently, the oxygen stacking sequence changes from O3 (ABCABC) to P3 (ABBCCA), and the resulting structure is classified as a P3-type phase.134 In contrast, the direct transformation from the O3 phase to the O2 phase is not readily achievable under electrochemical Na+ insertion/extraction at ambient temperature. While the O3 → P3 transition occurs through simple gliding of layers, the O3 → O2 transition requires the breaking and reformation of TM–O bonds, as it involves a change in the oxygen packing from a cubic close-packed (ccp) to a hexagonal close-packed (hcp) configuration. Based on the oxygen stacking arrangement, the O2-type structure is regarded as an intergrowth between ccp and hcp arrays.26 Notably, two glide vectors of the TMO2 slabs, (1/3, 2/3, z) and (2/3, 1/3, z), are responsible for the structural reorganization leading to the formation of the O2 phase.26Fig. 6b and c shown the P-type structures (P2 and P3), Na+ ions occupy prismatic sites, where the diffusion channels are straight and open, facilitating faster ionic transport along the interlayer direction. Conversely, in the O-type structures (O3 and O2), Na+ ions reside in octahedral sites, where the diffusion path becomes curved and less direct due to the geometry of the octahedral coordination and stronger electrostatic interactions. Thus, the difference in Na+ coordination (prismatic vs. octahedral) directly explains the distinct diffusion mechanisms, correlating with the structural transitions depicted in Fig. 6b and c.
Consequently, during the charge–discharge process, the partial extraction of Na+ ions causes the TMO2 layer to adopt a relatively low potential, altering the Na+ coordination environment due to Na–Na electrostatic repulsion, which generates a diffusion barrier for Na+ ions.135 This change in Na+ coordination can lead to Na-vacancy ordering, which in turn triggers the O3 → P3 phase transition. The continuous sliding of the TMO2 layers during this process severely deteriorates the storage capacity of the electrode material.131,136
In addition to the well-documented O3 phase, an O′3 phase has also been identified in NaxTMO2 systems. This distorted variant arises when the octahedra at the TM sites in the O3 structure experiences lattice distortion, producing TM–O bonds of varying lengths and resulting in the O′3 phase.137 The prime symbol (′) is conventionally used to denote a structurally distorted phase.8
On the other hand, during charge–discharge cycling, P2-type NaxTMO2 materials typically undergo phase transitions from the P2 phase to O2 or OP4 phases.136,138 These transformations, frequently observed in both binary and ternary compositions, can adversely affect capacity retention. Specifically, when the charging voltage exceeds 4.1 V vs. Na+/Na, the TMO2 layer facilitates Na+ migration from tetrahedral to octahedral sites. This process induces significant contraction along the c-axis, leading to a gradual structural transition from the P2 phase to the O2 phase with ABAC oxygen stacking.
It is essential to note that the spray-drying process is typically the primary step in the methodology for preparing precursors of cathode active materials (p-CAMs), which can often be preceded or followed by mechanical treatments, with ball milling being the most reported. In fact, the use of ball milling has been shown to have a significant impact on the electrochemical performance of cathode active materials (CAMs), refining particle size, facilitating the attainment of spherical morphology, and providing improved homogeneity of the mixture prior to spray drying.139 Additionally, the p-CAM often requires one or more additional thermal treatment steps to achieve the final CAM, as summarized in Table 2 and discussed in more detail below.
| Materials | Raw materials | Synthesis steps | T inlet (°C) | T oulet (°C) | Chelating agent/solvent | Thermal treatment (thermal rate) | Ref. |
|---|---|---|---|---|---|---|---|
| a The authors do not specify whether the temperature mentioned refers to the inlet or outlet. b The thermal treatment is performed prior to spray-drying. c The authors did not specify the type of precursor used. d The article synthesizes the materials at different temperatures. | |||||||
| O3-NaCrO2 | CH3COONa (CH3COO)3Cr | Spray-drying/calcination/sintered/calcination | 290 | — | H2O | 450 °C for 2 h (1 °C min−1)/900 °C for 10 h (2 °C min−1) | 147 |
| O3-NaCrO2−x | Na2CrO4·4H2O | Spray-drying/sintered | 260 | — | H2O | 850 °C for 6 h | 148 |
| Na0.44Mn0.97Bi0.01Zr0.02O2/CNT | Na2CO3 | Ball milling/solid state reaction/calcination/ball milling/spray-drying | 250 | 80 | Polyethylene glycol | 400 °C for 3 h | 149 |
| MnO2 | |||||||
| Bi(NO3)3·5H2O | Ethanol | ||||||
| ZrO2 | |||||||
| CNT | |||||||
| Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2 | NaNO3 | Calcination/ball milling/spray-drying | 160a | — | H2O | 450 °C for 5 h/750 °C for 12 hb | 109 |
| Al(NO3)3·9H2O | |||||||
| Fe(NO3)3·9H2O | |||||||
| Ni(NO3)2·6H2O | |||||||
| Cu(NO3)2·3H2O | |||||||
| Zn(NO3)2·6H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| P2-Na2/3Fe1/2Mn1/2O2 | CH3COONa | Spray-drying/calcination | 170 | 100 | H2O | 650 °C for 8 h/800 °C for 8 h | 150 |
| Fe(NO3)3·9H2O | |||||||
| C4H6MnO4·4H2O | |||||||
| Na0.62Ca0.025Ni0.28Mg0.05Mn0.67O2 | Na2CO3 | Ball milling/spray-drying/calcination | — | 80 | H2O | 550 °C for 5 h/900 °C for 12 h | 151 |
| MnO2 | |||||||
| NiO | |||||||
| CaCO3 | |||||||
| MgO | |||||||
| Na0.6Al0.11−xNi0.22−yFex+yMn0.66O2 (x = 0 and y = 0.11) | CH3COONa | Spray-drying/calcination/ball milling/spray-drying/calcination | 200 | 108 | H2O | 1000 °C/950 °C for 6 h (5 °C min−1) | 109 |
| Al(NO3)3·9H2O | |||||||
| Ni(OCOCH3)2·4H2O | |||||||
| Fe(NO3)3 9H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| P2-Na0.67Mn0.5Fe0.3Mg0.2O2 | Fe(NO3)3·9H2O | Spray-drying/calcination | 200 | 143 145 |
H2O | 775 °C for 8 h/800 °C for 10 h | 22 |
| Mn(CH3COO)2·4H2O | |||||||
| Mg(NO3)2·6H2O | |||||||
| NaCH3COO·3H2O | |||||||
| Na0.67NixFe0.52−xMn0.48O2 (x = 0.22) | CH3COONa | Spray-drying/calcination | 170 | 80 | H2O | 750 °C for 8 h | 152 |
| Ni(NO3)2·6H2O | |||||||
| Fe(NO3)3·9H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| NaNi0.4Fe0.2Mn0.4O2/MoS2 | Na2CO3 | Sanding/spray-drying/calcination/coating | — | — | H2O | 850 °C for 10 h | 153 |
| NiO | |||||||
| MnO2 | |||||||
| Fe2O3 | |||||||
| NaNi1/3Mn1/ 3Fe1/3O2-Lu (0.02) | CH3COONa | Spray-drying/calcination/ground/calcination | 150a | — | H2O | 450 °C for 6 h (2 °C min−1)/900 °C for 15 h (5 °C min−1) | 154 |
| Ni(CH3COO)2·4H2O | |||||||
| Fe(NO3)3·9H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| Lu2O3 | |||||||
| NaNi1/3Mn(1/3−x)Fe1/3TixO2 (x = 0.005) | Na2CO3 | Ball milling/spray-drying/calcination | — | — | H2O | 550 °C for 4 h/920 °C for 12 h | 155 |
| NiO | |||||||
| Fe2O3 | |||||||
| MnO2 | |||||||
| TiO2 | |||||||
| NaNi1/3−xMn1/3Fe1/3ZnxO2 (x = 0.005) | Na2CO3 | Ball milling/spray-drying/calcination | — | — | H2O | 550 °C for 5 h/900 °C for 12 h | 156 |
| NiO | |||||||
| MnO2 | |||||||
| Fe2O3 | |||||||
| ZnO | |||||||
| NaNi1/3Mn(1/3−x)Fe1/3AlxO2 (x = 0.005) | Na2CO3 | Ball milling/spray-drying/calcination | — | — | H2O | 550 °C for 4 h/900 °C for 12 h | 157 |
| NiO | |||||||
| MnO2 | |||||||
| Fe2O3 | |||||||
| Al2O3 | |||||||
| Na(Ni1/3Fe1/3Mn1/3)0.99Mo0.01O1.99F0.01 | Na2CO3 | Spray-drying/calcination | 180a | — | H2O | 900 °C for 12 h | 158 |
| NiO | |||||||
| MnO | |||||||
| Fe2O3 | |||||||
| NaF | |||||||
| Mo-based compoundc | |||||||
| NaNi1/3Mn1/3−xFe1/3O2-Sc (x = 0.01) | MnO | Ball milling/spray-drying/calcination | — | — | H2O | 900 °C for 12 h | 159 |
| NiO | |||||||
| Fe2O3 | |||||||
| Sc2O3 | |||||||
| Na2CO3 | |||||||
| NaNi1/3−x/3Mn1/3−x/3Fe1/3−x/3O2-Tax (x = 0.01) | Na2CO3 | Ball milling/spray-drying/calcination | — | — | H2O | 900 °C for 12 h (3.65 °C min−1) | 160 |
| NiO | |||||||
| MnO | |||||||
| Fe2O3 | |||||||
| Ta2O5 | |||||||
| NaNi1/3Fe1/3−xMn1/3ZrxO2 (x = 0.01) | Na2CO3 | Ball milling/spray-drying/calcination | — | — | H2O | 550 °C for 4 h/900 °C for 12 h | 161 |
| NiO | |||||||
| Fe2O3 | |||||||
| MnO2 | |||||||
| ZrO2 | |||||||
| NaNi1/3Mn1/3−xFe1/3NbxO2 (x = 0.01) | Na2CO3 | Ball milling/spray-drying/calcination | — | 80 | H2O | 550 °C for 5 h/900 °C for 12 h/25 °C (5 °C min−1) | 162 |
| MnO2 | |||||||
| NiO | |||||||
| Fe2O3 | |||||||
| Nb2O5 | |||||||
| Na0.67Ni0.11Cu0.11Fe0.3Mn0.48O2 | CH3COONa | Spray-drying/calcination | 170 | 80 | H2O | 700, 750, 800, 850, 900 °C for 8 hd | 1 |
| Ni(NO3)2·6H2O | |||||||
| Cu(CH3COO)2·H2O | |||||||
| Fe(NO3)3·9H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| NaNi1/3−xFe1/3Mn1/3MgxO2 (x = 0.03) | Fe2O3 | Ball milling/spray-drying/calcination | — | — | H2O | 900 °C for 12 h (10 °C min−1) | 163 |
| MnO | |||||||
| NiO | |||||||
| MgO | |||||||
| Na2CO3 | |||||||
| NaNi0.32Fe0.32Mn0.32Al0.02Cu0.02O2 | CH3COONa | Spray-drying/calcination/manual grinding/sintering | — | — | H2O | 230 °C for 3 h (1 °C min−1)/850 °C for 20 h (3 °C min−1) | 164 |
| Ni(CH3COO)2·4H2O | |||||||
| Fe(NO3) 3·9H2O | |||||||
| Mn(CH3COO)2·4H2O | |||||||
| Al(NO3)3·9H2O | |||||||
| Cu(CH3COO)2·H2O | |||||||
| NaNi0.25Fe0.15 Mn0.3Ti0.1Sn0.05Co0.05Li0.1O2 | Fe2O3 | Spray-drying/calcination | — | — | H2O | 900, 950, 980 °C for 10 hd | 18 |
| Co2O3 | |||||||
| NiO | |||||||
| SnO | |||||||
| TiO2 | |||||||
| Na2CO3 | |||||||
| Li2CO3 | |||||||
In the synthesis of NaxTMO2 compounds via the spray-drying approach, different sodium sources are employed, depending on the adopted synthesis strategy, with Na2CO3, NaCH3COO, and NaNO3 being the most reported (Fig. 7). Among these, Na2CO3 is the predominant sodium precursor, which may be attributed to its wide industrial availability and versatility, arising from its use as a common raw material for the synthesis of other sodium salts.140 Additionally, carbonates are widely used because they decompose during heat treatment in air, leaving no residual byproducts.17 As expected, the greater use of Na2CO3 is also related to the type of material being dried. In fact, Na2CO3 is commonly used as a raw material in suspension drying processes, particularly for the drying of a mixture of transition metal oxides.
![]() | ||
| Fig. 7 Distribution of sodium sources used in the spray-drying synthesis of NaxTMO2 compounds, with Na2CO3 as the most used precursor. | ||
Regarding the selection of transition-metal precursors for the TM site in the synthesis of NaxTMO2, a clear predominance of metal oxide–based materials is observed (Fig. 8a). Metal acetates and nitrates are also commonly employed, whereas carbonate-based precursors are used far less frequently. Nitrates and acetates are common metal precursors due to their low decomposition temperatures,9 as well as their relatively high solubility in water, facilitating the preparation of homogeneous solutions.
Metal oxides are predominantly associated with synthesis routes involving high-temperature calcination, which is linked to their superior thermal stability and the absence of decomposition byproducts during calcination.141,142 These characteristics, combined with their lower cost compared to other raw materials, make them attractive precursors.142 Additionally, as shown in Fig. 8c, a wide range of metal oxide precursors has been employed, as expected for a class of materials encompassing diverse compositions and structures. For instance, two different manganese oxide phases (MnO and MnO2) have been explored in the synthesis and property tuning of NaxTMO2 compounds. Interestingly, drying processes involving oxide-based precursors are typically preceded by ball milling, whereas those involving metal nitrates or acetates generally do not require this step. In fact, ball milling is used to increase suspension stability by reducing particle size in a liquid medium.9 The uniform dispersion of fine particles is maintained in the spray-dried material, facilitating the formation of the final phase as diffusion lengths during heat treatment are consequently reduced.9
On the other hand, Ni, Fe, and Mn dominate the cation chemistry of the NaxTMO2 cathode design because, together, they deliver high capacity and voltage, fast Na+ transport, and good structural stability. Additionally, these metals are abundant, relatively low-cost, and less toxic elements143 (Fig. 8b). Other 3d metals either underperform electrochemically or undermine the cost and sustainability advantages central to SIBs technology. In fact, NiFeMn-containing CAMs are the most favorable redox-active elements for commercial NaxTMO2 when cost and supply risk are included.144,145
The operating conditions of the spray-drying process vary significantly among the studies analyzed, with inlet temperatures ranging from 150 °C to 290 °C. This parameter is particularly important, as a clear correlation has been established between the particle size of spray-dried materials and the temperature employed during the synthesis process.64 However, outlet temperature and other important parameters such as feed rate, atomization pressure, and nozzle diameter are not consistently reported. In several cases, only a few or none of the spray-drying operating conditions are presented, which limits direct comparison between different synthesis methods.
After spray-drying, the obtained powders (p-CAM) are generally subjected to thermal treatment as a final step. This subsequent high-temperature annealing/calcination step is typically required to promote the formation of the desired homogeneous crystalline phase.17,22 Studies report a single calcination step with temperatures ranging from 400 °C to 980 °C and durations varying from 3 to 12 hours. For instance, Liang et al.18 reported the impact of calcination temperature on the morphology of the formed materials, showing that higher temperatures led to larger particle sizes and more pronounced layered structures, which directly affected the specific discharge capacity and capacity retention of the batteries. On the other hand, many other studies employ a two-step calcination process, in which the first step is carried out at relatively mild temperatures, ranging from 230 °C to 775 °C, followed by a second calcination at higher temperatures between 800 °C and 900 °C. Multiple calcination steps are commonly applied to ensure the formation of the desired crystalline phase, enhance crystallinity, and promote improved cathode material performance.146 However, the authors do not clearly justify the need for two (or more) consecutive thermal treatment steps in the preparation of NaxTMO2, given that these processes are energy-intensive and increase the overall production cost.
Nevertheless, to prevent ambiguity in both the definition and conceptual understanding of HEMs, the literature has placed particular emphasis on quantifying configurational entropy (ΔSconfig). Thus, considering the calculation of ΔSconfig based on eqn (1), the materials that exhibit ΔSconfig ≥ 1.5R are classified as high-entropy. On the other hand, compounds exhibiting configurational entropy values in the range of 1R ≤ ΔSconfig < 1.5R are categorized as medium-entropy materials (MEMs), while those with ΔSconfig < 1R are classified as low-entropy materials (LEMs).166,167
![]() | (1) |
Among the various categories of HEMs, high-entropy oxides (HEOs) have attracted significant attention. Nevertheless, it is essential to note that metal oxides (MOs) comprise a diverse class of materials that can exhibit distinct compositions and structural frameworks. As a result, their physicochemical properties vary considerably, enabling applications across multiple fields, including sensors,168 electrocatalysis,169 supercapacitors,170 batteries,171 and many related energy-conversion and storage technologies.172 Interestingly, the library of HEOs encompasses a wide range of MOs, among which the most representative crystal structures include rocksalt, spinel, perovskite, and layered frameworks. As expected for compounds incorporating multiple metal species, these MOs typically feature distinct crystallographic sites with diverse local chemical environments. These multielemental-based compounds not only enable a rich compositional and property space but also introduce a substantial level of structural complexity.
Interestingly, since many HEOs may contain multiple crystallographically distinct sites, the configurational entropy should, in principle, be evaluated separately for each unique site. However, this aspect is not always considered, particularly because the same element can occupy different crystallographic positions. For instance, transition metals with different oxidation states may simultaneously reside in tetrahedral and octahedral sites in spinel-based HEOs or other spinel MOs. Similarly, layered oxides with the general formula NaxTMO2 (subject of this review) typically exhibit well-defined crystallographic sites, consisting of repeating TMO6 slabs, with Na+ ions located between the oxide layers.10 In these layered structures, different elements may populate several crystallographically distinct positions, providing considerable versatility for compositional tuning at the Na, TM, and O sublattices. While substitution at the Na and O sites is also feasible, the spectrum of elements that can be incorporated at these positions is comparatively restricted compared to the TM site.13
In summary, the Na, TM, and O sites can all contribute to the total configurational entropy, and the overall ΔSconfig of NaxTMO2 can be determined by considering the contributions from each crystallographic site, as described by eqn (2):
![]() | (2) |
As previously highlighted, the TM site offers substantially greater compositional freedom, enabling a broader selection of substituent elements and different proportions. In fact, proper element selection and component stoichiometry optimization are the two key aspects to consider when designing high-performance HEM-based cathode materials.173 These species may include electrochemically active metals (e.g., Ni2+, Fe3+, Mn2+, Cr3+, and Co3+), which function as redox-active centers and directly contribute to charge storage, as well as electrochemically inactive species (e.g., Nb5+, Zn2+, Zr4+, Mg2+, Sn4+, Sb5+, and Ti4+), which primarily enhance structural robustness without engaging in redox processes.13 Besides, mobile cations (e.g., Li+, Mg2+, and Zn2+) can trigger anionic redox processes, which may contribute extra capacity and broaden the operating voltage window.9 Consequently, the elemental composition and relative proportions can be adjusted to meet specific design objectives.9
Interestingly, incorporated elements can perform different roles, particularly when present in specific oxidation states. For instance, the active Mn2+ cation can increase reversible capacity by providing charge compensation,9 while inactive Mn4+ cations enhance cycling stability by stabilizing the structure with suppressed Jahn–Teller distortion.174 Besides, the same cation can perform different functions depending on the designed composition. For example, while the Co3+ cation is expected to provide charge compensation for capacity in the NaNi0.12Cu0.12Mg0.12Fe0.15Co0.15Mn0.1Ti0.1Sn0.1Sb0.04O2-based cathode,12 the same ion plays a stabilizing role in the host structure, suppressing Jahn–Teller distortion during Na+ insertion/extraction in Na2/3Li1/6Fe1/6Co1/6Ni1/6Mn1/3O2.174 Fortunately, many review articles have summarized the roles of various incorporated ions in both HEMs and lower-entropy systems,3,9,10 and should be consulted for specific information.
As previously exemplified and shown in the next subtopic, the design of layered NaxTMO2 cathodes can be significantly improved through entropy engineering of the TM sublattice. Unlike conventional low-entropy counterparts (such as unary, binary, or other doped LEMs), multicomponent systems introduce substantial configurational entropy, which stabilizes the layered framework.110 Configurational entropy represents a key thermodynamic principle of high-entropy materials, originating from the large number of accessible atomic arrangements in the crystal lattice.175 In fact, from a thermodynamic perspective, the increase in configurational entropy has a direct and critical impact on the Gibbs free energy of the system. According to the relationship expressed by eqn (3), a higher entropy term reduces the Gibbs free energy, thereby enhancing the thermodynamic stability of the material.10
| ΔGmix = ΔHmix − TΔSmix | (3) |
The stabilization induced by configurational entropy leads to homogeneous solid solutions, suppressing the formation of ordered intermetallic or phase-separated structures.111,175 Besides, it is important to understand the four main core effects in HEMs, which were as follows: (i) thermodynamics: high entropy effects, (ii) structure: lattice distortion, (iii) kinetics: sluggish diffusion, and (iv) properties: cocktail effects.176
Since thermodynamic stabilization, lattice distortion, sluggish diffusion, and cocktail effects all depend on a uniform distribution of multiple metals into the entropy materials, selecting an appropriate synthesis method is crucial to fully realize entropy engineering in layered NaxTMO2. In this regard, spray-drying provides an effective strategy, offering homogeneous mixing of metal precursors and precise control over composition. This technique enables the scalable production of entropy-engineered NaxTMO2 cathodes with optimized structural and electrochemical properties for SIBs, as shown below.
Moreover, higher-entropy systems generally exhibit improved electrochemical performance while also diversifying their compositions, particularly through the incorporation of more abundant and less expensive elements. In fact, ∼80% of the compositions achieving a capacity retention of ≥80% after 200 or more charge–discharge cycles are mainly based on NiFeMn-containing NaxTMO2 (Fig. 9), revealing a clear trend in the development of cathode materials for SIBs. Interestingly, the capacity retention values of spray-dried NaxTMO2 are as high as those of many Ni-rich NaxTMO2 analogues prepared by other synthesis methods (solid-state and coprecipitation), as recently summarized by Gonçalves et al.143 Furthermore, a common strategy among these studies involves the introduction of a fourth element as a dopant, accounting for about 86% of the NaxTMO2 compounds, as presented in Fig. 9, demonstrating that doping contributes to enhanced electrochemical stability.
The following sections discuss the main strategies in the design of Na-LEOs and Na-MEOs, considering the configurational entropy calculation of the compositions reported in the respective papers as the classification criterion for low- and medium-entropy materials (LEMs and MEMs).
In fact, to the best of our knowledge, the first report on the use of the spray-drying process for the synthesis of NaxTMO2 was dedicated to the preparation of a binary P2-Na2/3[Ni1/3Mn2/3]O2 composition. This finding aligns with the fact that the development stage of NaxTMO2 materials was already relatively advanced, and it is well established in the literature that unary systems exhibit certain advantages but, more importantly, significant drawbacks, as highlighted below. This may partly explain the limited number of studies highlighting the use of spray-drying in the synthesis of unary and even binary NaxTMO2 compounds.
Considering the aforementioned observations, only a few Na-LEOs have been reported, specifically two unary Na-LEOs (O3-NaCrO2)147,148 and one binary Na-LEO (P2-Na2/3Fe1/2Mn1/2O2).150 In fact, all other Na-LEOs are based on multimetal NaxTMO2 compounds, as expected for a common strategy in the design of active materials, particularly through doping with one or more active or inactive elements (Tables 2 and 4). In fact, the spray-drying approach allows for homogeneous atomic-level mixing while relying on a simple and well-controlled processing route, thereby enabling efficient and uniform dopant incorporation.164
Although comparison with other unary Na-LEOs prepared by spray-drying is not possible, O3-NaCrO2-based cathodes have recently been reported to exhibit highly reversible redox reactions and excellent cycling stability. In fact, as already mentioned, it is well established that the practical capacities of this class of cathode materials remain below their theoretical limits, while their cycling stability at high voltages is notably compromised by irreversible transition-metal migration once more than 0.5 mol of Na+ (per mol of CrO2) is extracted.148 In this sense, Liu's group147 reported improved electrochemical properties by in situ constructing oxygen vacancy defects in O3-NaCrO2 and accurately regulating the cut-off voltage, which effectively averts irreversible chromium migration and ensures a highly reversible phase transition during cycling.
On the other hand, despite the recent advances demonstrated for spray-dried engineered O3-NaCrO2, overall, unary LEOs already prepared by other synthesis methods present significant challenges for real-world applications. For instance, while NaxCoO2, unary LEOs NaxMnO2, and NaxFeO2 offer acceptable theoretical capacities and sodium abundance,177 their practical use is limited by structural instability, phase transitions, poor cycling performance, and, in the case of cobalt, both cost and toxicity.178–180 In fact, for other low-entropy systems, some alternative strategies (or combinations of strategies) have been widely employed to overcome these drawbacks and to guide the design of NaxTMO2 materials with enhanced performance. For instance, doping strategies, specifically co-doping and multimetal doping with one or more active or inactive elements, have been recently used in designing active Mn-rich Na-LEO-based multimetal compounds, thereby lowering the concentration of Jahn-Teller-active Mn3+ sites and mitigating the associated structural impact.
In one of these studies, Gong et al.149 developed an effective synergistic design strategy for Mn-rich ternary NaxTMO2 cathodes by combining Bi/Zr co-doping with the incorporation of carbon nanotubes (CNTs). The dual doping enhances Na+ diffusion kinetics and stabilizes the crystal framework, while CNTs improve electronic conductivity and reduce polarization. As a result, the Mn-rich Na0.44Mn0.97Bi0.01Zr0.02O2/CNT composite delivers 102.26 mAh g−1 at 5C and maintains 87.2% and 72.4% of its initial capacity after 1000 and 2000 cycles, respectively, significantly outperforming pristine and singly doped analogues.
In addition to Mn-rich ternary NaxTMO2 design by spray-drying process, multielemental doping is one of the most recent and promising approaches, and in the view of some authors, this strategy may encompass the concept of high entropy (but this does not reflect the view of the present authors, as discussed below). A recent study introduced a Mn-rich senary P2-type composition, Na0.67Al0.02Fe0.02Ni0.02Cu002Zn0.02Mn0.9O2,20 in which both redox-active and lattice-stabilizing cations were incorporated to address sodium deficiency while enabling low-strain electrochemical operation (Fig. 10a–c). The incorporation of Ni2+, Fe3+, and Cu2+ (redox-active) provides effective charge compensation, whereas Al3+ and Zn2+ (redox-inactive) reinforce the metal–oxygen framework, diminishing the population of Jahn–Teller-active Mn3+ sites and improving structural robustness. These hetero-cations strengthen M–O bonding, limit lattice distortion, and prevent particle cracking, ultimately yielding minimal volumetric variation during cycling. In situ XRD analyses reveal that this multimental doping strategy substantially alleviates lattice strain (≈1.58%) and enlarges the Na-layer spacing, promoting rapid Na+ transport (Fig. 10d and e). As a result, the cathode delivers a reversible capacity of ∼157 mAh g−1 and preserves 88.81% of its capacity after 600 cycles at 1C, with an average coulombic efficiency of 99.90%. Impressively, the capacity retention value surpasses that of many other Mn-rich NaxTMO2 materials (Na0.5Ni0.2Co0.15Mn0.65O2, with a capacity retention of 87.6% after 100 cycles at 0.1C46), as well as other active materials typically used in commercial lithium-ion batteries, such as LiFePO4 (LiFePO4, with a capacity retention of 85.5% after 300 cycles at 1C181), LiNixCoyMn1−x−yO2 (LiNi0.9Co0.05Mn0.05O2, with a capacity retention of 77.0% after 400 cycles at 1C182), and LiNixCoyAl1−x−yO2 (LiNi0.865Co0.095Al0.04O2, with a capacity retention of 79% after 200 cycles at 1C183).
![]() | ||
| Fig. 10 (a) Schematic representation of the Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2 crystal structure. (b) Conceptual illustration depicting how the HENM achieves high capacity and minimal lattice strain. (c) SEM-EDS elemental mapping results for Na, Mn, Al, Fe, Ni, Cu, and Zn. (d) Contour maps of the in situ XRD measurements collected for the Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2. (e) Comparative analysis of volume changes across different cathode materials. Reproduced with permission from ref. 20. Copyright © 2025 Wiley-VCH GmbH. | ||
It is important to emphasize that, although the previously reported multielement composition (Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2) was described as employing a “high-entropy doping” strategy, the configurational entropy calculated for the TM site of NaxTMO2 clearly corresponds to a low-entropy system. In this context, the use of the term “high-entropy” for doped materials can lead to conceptual ambiguity and inadvertently disseminate a definition that does not align with the established criteria for high-entropy materials (HEMs). As aforementioned, the designation of high-entropy compounds in this work follows the conventional threshold based on configurational entropy (ΔSmix > 1.5R). Consequently, compositions dominated by a single transition metal (rich in a metal) or systems obtained through simple multimetal doping should not be classified as HEMs, even when they contain five or more elements. Therefore, greater caution is required when applying such terminology to avoid misinterpretation within the field.
It is important to emphasize that ternary materials rich in a given element should generally not be categorized as MEMs; therefore, the major element should not exceed approximately 45 mol% for ternary compounds in which the other two elements are equimolar. In fact, it is possible to obtain multimetallic compositions with low entropy; thus, calculating configurational entropy is essential for accurate classification, as demonstrated in Table 4. Thus, for medium-entropy ternary NaxTMO2 compositions (Na-MEOs), equal proportions of 1/3 for each transition-metal species yield the highest configurational entropy (whereas a 1/4 distribution maximizes entropy in quaternary systems). Besides, Na-MEOs (that is, typically composed of three to four elements) are expected to exhibit a ΔSconfig in the range of 1.0 to 1.5R.166
Among the ternary NaxTMO2 compounds, and as expected, compositions with transition metals based on Ni, Fe, and Mn (NaxNiyFezMn1−(y+z)O2) have attracted particular attention, not only considering conventional synthesis routes but also when analyzing compositions obtained through spray-drying techniques. In general, the predominance of studies focusing on ternary NaxNiyFezMn1−(y+z)O2 (Na-NFM) arises from their favorable compromise between reliable electrochemical performance and the simplicity and reproducibility of their synthesis routes.3 And, more specifically, in syntheses employing spray-drying, this trend is no different. Besides, as expected for higher-entropy Na-MEOs, most cathode materials synthesized via spray-drying that exhibit capacity retention ≥80% after 200 or more charge–discharge cycles are based on Na-NFM or related doped analogues, in which the transition-metal species are present in equimolar proportions (or in close equimolar composition).
Interestingly, almost all the Na-NFM compositions described incorporate inactive cation dopants, representing one of the dominant design trends for MEOs prepared via spray-drying, whereas electrochemically active dopants are relatively less prevalent. In general, a broad range of dopant species has been reported, including alkali (Li+) and alkaline-earth metals, d-block elements (Co2+, Cu2+, Zn2+, Sc3+, Ta5+, Zr4+, Mo4+, Ti4+, Nb5+), p-block elements (Al3+), and even rare-earth cations (Lu3+). It is also worth noting in this review that the term doping is only used when the introduced element is present at concentrations typically ≤ 5% in mass or atomic fraction, as recently discussed by Gonçalves et al.184 Above this level, the incorporated metal species should be regarded as part of the host lattice (structural components) rather than as a true dopant.
Metal doping is one of the simplest and most effective strategies for the design of the Na-NFM cathode materials because it stabilizes the layered structure by accommodating more Na+ ions, improving its capacity retention.185,186 Besides, metal ion incorporation additionally promoted phase optimization by modulating strain distribution and redox activity.3 In fact, Namazbay et al.3 reported that in general, all doped Na-NFM cathode materials show improved capacity retention and rate capability, although a slight reduction in capacity is observed for some compositions.
Interestingly, in the context of doped Na-NFM materials designed via the spray-drying process, a large fraction of the dopant metals is aliovalent species, more specifically, highly valent metal cations. An optimal concentration of high-valence metal cations within the transition metal (TM) layer aids in maintaining charge balance, thereby inhibiting the oxidation of Ni2+ to Ni3+ during thermal treatment. As a result, a larger fraction of Ni2+ remains electrochemically active during charge–discharge processes, resulting in an enhanced capacity.3
Despite the limited number of studies addressing the same dopant metal, the design of aliovalent-containing medium-entropy Na-NFM cathodes reported in the literature indicates that small dopant concentrations are preferred (Table 3), notably at atomic fractions of 1% or 0.5% (atomic percent). As expected, cathode materials typically consist of secondary particles with porous spherical or porous hollow-spherical morphologies, ranging in size from 2 to 10 µm, while the primary particles are generally smaller than 1 µm and have a polyhedral shape. Some of these examples are summarized in Fig. 11, showing 1 at% of the aliovalent dopant metal uniformly distributed throughout all particles.
| Doping element (ionic radius) | Optimized at% of the dopant element | Key effects induced by aliovalent dopants | Ref. |
|---|---|---|---|
| Ta5+ (0.64 Å) | 1% | • Strong bonding energy effect of Ta–O | 160 |
| • Ta5+ doping increased the inter-slab spacing | |||
| • Ta5+ doping decreased the slab thickness | |||
| • Ta5+ doping reduced the Na+ activation barrier energy | |||
| Sc3+ (0.74 Å) | 1% | • Sc3+ doping introduced strong Sc–O bonds | 159 |
| • Sc3+ decreased the band gap of the material | |||
| • Sc3+ decreased the Na+ activation barrier energy | |||
| • Sc3+ doping decreased the slab thickness | |||
| • Sc3+ doping increased the inter-slab spacing | |||
| Nb5+ (0.64 Å) | 1% | • Nb5+ doping can enlarge Na+ diffusion channels | 162 |
| • Nb5+ promotes the conversion of Mn4+ to Mn3+ in the cathode material | |||
| • Nb5+ doping increased the D-spacing and slab thickness | |||
| • Nb5+ doping inhibited phase transitions | |||
| Zr4+ (0.79 Å) | 1% | • Zr4+ doping reduced oxygen losses | 161 |
| • Zr4+ doping decreased the slab thickness | |||
| • Zr4+ doping increased the inter-slab spacing | |||
| • Zr4+ doping inhibited phase transitions | |||
| Al3+ (0.53 Å) | 0.5% | • Al3+ doped samples shortened the TM–O bond length | 157 |
| • Al3+ doping decreased the diffusion barrier of Na+ | |||
| • Al3+ doping increased inter-slab spacing | |||
| • Al3+ doping reduce the lattice strain or volume change | |||
| Ti4+ (0.605 Å) | 0.5% | • Ti4+ doping increased the inter-slab spacing | 155 |
| • Acts as a “pillar” to reduce structural damage | |||
| • Ti4+ doping can inhibited of phase transition of O3–P3 during the cycle | |||
| • Ti4+ doping can reduce the kinetic hindrance | |||
| • Ti4+ doping can decrease the extent of polarization of the material throughout the cycling process |
| Material | Entropy | (Secondary/primary particles morphology) | Phase | Capacity retention (%) | First discharge capacity (mAh g−1) | Mass ratio/electrolyte/mass loading | Window potential (V vs. Na/Na+) | Ref. |
|---|---|---|---|---|---|---|---|---|
| a DEC = diethyl carbonate, DMC = dimethyl carbonate, EC = ethylene carbonate, FEC = fluoroethylene carbonate, PC = propylene carbonate. | ||||||||
| O3-NaCrO2 | 0 | Irregular platelike/— | O3 | 81.9 (after 1000 cycles@5C) | 122.0@0.2C | 80 : 10 : 10/1 M NaClO4 in EC : PC (1 : 1) + 5% FEC/— |
2.3–3.7 | 147 |
| O3-NaCrO2−x | 0 | — | O3 | 80.2 (after 400 cycles@5C) | 134.5@1C | 80 : 10 : 10/1 M NaClO4 in EC : PC (1 : 1) + 5% FEC/— |
1.5–3.8 | 148 |
| Na0.44Mn0.97Bi0.01Zr0.02O2/CNT | 0.154R | Regular rod/— | NMO-BZ/C Bi2Mn4O10 ZrO2 | 87.2 (after 1000 cycles@5C) | 123.6@0.2C | 70 : 20 : 10/1 M NaClO4 in EC : DMC (1 : 1)/— |
— | 149 |
| Na0.67Al0.02Fe0.02Ni0.02Cu0.02Zn0.02Mn0.9O2 | 0.486R | Spherical particles/— | P2 | 88.8 (after 600 cycles@1C) | 156.86@0.1C | 80 : 10 : 10/1 M NaPF6 in PC : EMC:FEC (45 : 37 : 0.5) — |
2.0–4.0 | 20 |
| P2-Na2/3Fe1/2Mn1/2O2 | 0.693R | Irregular particles/irregular particle | P2 | 100 (after 130 cycles@0.1C) | 217.9@0.1C | 80 : 10 : 10/1 M NaClO4 in PC + 5% FEC/— |
1.5–4.2 | 150 |
| Na0.62Ca0.025Ni0.28Mg0.05Mn0.67O2 | 0.867R | Porous spherical structure/hexagonal shape | P2 | 85.6 (after 100 cycles@1C) | 105.3@1C | 70 : 20 : 10/1 M NaClO4 in EC : DMC (1 : 1)/— |
2.2–4.35 | 151 |
| Na0.6Al0.11−xNi0.22−yFex+yMn0.66O2 (x = 0 and y = 0.11) | 1.003R | Porous spherical particles/flake-like particles | P2 | 73.7 (after 100 cycles@1C) | 226@0.05C | 80 : 10 : 10/1 M NaPF6 in PC + 5% FEC/3.3 mg cm−2 |
1.5–4.6 | 109 |
| P2-Na0.67Mn0.5Fe0.3Mg0.2O2 | 1.029R | — | P2 | 81.0 (after 100 cycles@1C) | — | 80 : 10 : 10/1 M NaPF6 in EC : DMC (1 : 1) + 2% FEC/2.0 mg cm−2 |
1.5–4.2 | 22 |
| Na0.67NixFe0.52−xMn0.48O2 (x = 0.22) | 1.046R | Multilayer stacking of nanosheets/nanosheets particles | P2/O3 | 72.7 (after 300 cycles@0.1C) | 174.2@0.1C | 80 : 10 : 10/1 M NaClO4 in EC : DMC (1 : 1)/2.0 mg cm−2 |
1.5–4.1 | 152 |
| NaNi0.4Fe0.2Mn0.4O2/MoS2 | 1.055R | Microsheet structure/lamellar particles | O3 | — | 148.4@0.1C | 70 : 20 : 10/1 M NaClO4 in PC + 5% FEC/2.5 mg cm−2 |
2.0–4.2 | 153 |
| NaNi1/3Mn1/ 3Fe1/3O2-Lu (0.02) | 1.096R | Spherical-like particles/flake-like particles | O3 | 82.4 (after 500 cycles@0.1C) | 151.36@1C | 80 : 10 : 10/1 M NaClO4 in EC : PC (1 : 1) + 5% FEC/2.0–3.0 mg cm−2 |
2.0–4.0 | 154 |
| NaNi1/3Mn(1/3−x)Fe1/3TixO2 (x = 0.005) | 1.125R | Spherical particles/— | O3 | 86.2 (after 200 cycles@1C) | 130.8@1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm−2 |
2.0–4.3 | 155 |
| NaNi1/3−xMn1/3Fe1/3ZnxO2 (x = 0.005) | 1.125R | Porous spherical particle/— | O3 | 86.53 (after 200 cycles@1C) | 123.3@1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm−2 |
2–4.0 | 156 |
| NaNi1/3Mn(1/3−x)Fe1/3AlxO2 (x = 0.005) | 1.125R | Spherical particle/polyhedral particle | O3 | 87.8 (after 200 cycles@1C) | 115.1@1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm−2 |
2–4.0 | 157 |
| NaNi1/3Mn1/3−xFe1/3O2-Sc (x = 0.01) | 1.144R | Porous spherical aglomerates/— | O3 | 86.5 (after 200 cycles@1C) | 93.4@5C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/6 mg cm−2 |
2.0–4.0 | 159 |
| NaNi1/3−x/3Mn1/3−x/3Fe1/3−x/3O2-Tax (x = 0.01) | 1.144R | — | O3 | 80.2 (after 200 cycles@1C) | 145@0.1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/6 mg cm−2 |
2.0–4.0 | 160 |
| NaNi1/3Fe1/3−xMn1/3ZrxO2 (x = 0.01) | 1.144R | Spherical particle/polygonal particle | O3 | 87.1 (after 200 cycles@1C) | 137.6@0.2C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/— |
2.0–4.0 | 161 |
| NaNi1/3Mn1/3−xFe1/3NbxO2 (x = 0.01) | 1.144R | Porous hollow spherical/polyhedral structure | O3–P3 | 83.8 (after 200 cycles@1C) | 140@0.2C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm−2 |
2.0–4.0 | 162 |
| Na(Ni1/3Fe1/3Mn1/3)0.99Mo0.01O1.99F0.01 | 1.175R | Plate like particle stack/polygonal plate-like shape | O3 | 91.97 after 100 cycles@1C | 149.4@0.1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1) + 5% FEC/5.0 mg cm−2 |
2.0–4.0 | 158 |
| Na0.67Ni0.11Cu0.11Fe0.3Mn0.48O2 | 1.199R | Layer structure stacking/spherical particles | P2/P3 | 71.1 (after 300 cycles@1C) | 160.3@0.1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/2.0 mg cm−2 |
1.5–4.1 | 1 |
| NaNi1/3−xFe1/3Mn1/3MgxO2 (x = 0.03) | 1.199R | — | O3 | 87.5 after 200 cycles@1C | 133.9@0.1C | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/6.0 mg cm−2 |
2.0–4.0 | 163 |
| NaNi0.32Fe0.32Mn0.32Al0.02Cu0.02O2 | 1.250R | Flake like stack particles/flake-like particles | O3 | 83.0 (after 200 cycles@2C) | 113@5C | 80 : 10 : 10/1 M NaClO4 in PC + 5% FEC/2.0–3.0 mg cm−2 |
2.0–4.0 | 164 |
| NaNi0.25Fe0.15 Mn0.3Ti0.1Sn0.05Co0.05Li0.1O2 | 1.522R | Spherical structure/layered structure | O3 | 80.7 after 200 cycles@0.5C | 150.1@0.1C | 70 : 20 : 10/1 M NaClO4 in PC + 5% FEC/— |
2.0–4.3 | 18 |
![]() | ||
| Fig. 11 SEM images at different magnifications and corresponding EDS elemental mappings of the medium-entropy Na-NFM cathodes doped with 1 at% Ta5+ (a–c), Nb5+ (d–f), and Zr4+ (g–i). Panel (a–c): reproduced with permission from ref. 160. Copyright © 2024 Elsevier Ltd. All rights reserved. Panel (d and e): reproduced with permission from ref. 162. Copyright © 2023 Elsevier B.V. All rights reserved. Panel (g–i): reproduced with permission from ref. 161. Copyright © 2023 Elsevier B.V. All rights reserved. | ||
As previously mentioned, in addition to the similar morphology readily observed in Fig. 11, these studies report optimized compositions at an aliovalent cation concentration of 1 at% (atomic percent). Interestingly, in all cases presented in Fig. 12, the incorporation of high-valence dopant cations leads to enhanced capacity retention, even after 200 charge–discharge cycles, for compositions optimized at 1 at% of dopant. In addition, most medium-entropy Na-NFM cathodes exhibit noticeable improvements in rate capability. Remarkably, many reported studies show that doping with such aliovalent cations yields electrodes with higher specific capacities than their undoped counterparts, an unexpected result given that these cations are electrochemically inactive. Nevertheless, the origin of the enhanced initial specific capacity is strongly dependent on the nature of the incorporated dopant. For instance, in samples containing 1 at% Ta5+, the capacity increase was attributed to a higher relative fraction of Ni2+.160 In the case of Nb5+ – containing medium-entropy Na-NFM cathodes, the discharge profile shows a broader voltage plateau and reduced polarization compared to pristine Na-NFM, which helps explain the higher discharge capacity achieved with optimal Nb5+ doping at 1 at%.162 Additionally, the improvement in initial capacity may be associated with increased Mn3+ content, which enhances electronic conductivity. By contrast, NFM doped with 1 at% Zr4+ exhibits an overall broader discharge plateau than the undoped material.161 Overall, these results indicate that the enhancement of initial specific capacity is governed not only by the dopant concentration but also by the specific aliovalent cation introduced into the lattice. In contrast, at higher dopant levels (≥2 at%), the observed decline in capacity retention and rate performance is primarily attributed to increased crystal lattice distortion,159,160 the Jahn–Teller effect arising from a higher proportion of Mn3+ (which compromises structural reversibility),160 and an increase in grain size.162
![]() | ||
| Fig. 12 Cycling performance for 200 cycles and rate performance at various rates for medium-entropy Na-NFM cathodes doped with different at% of (a and b) Ta5+, (c and d) Sc3+, (e and f) Zr4+, and (g and h) Nb5+. Panel (a and b): reproduced with permission from ref. 160. Copyright © 2024 Elsevier Ltd. All rights reserved. Panel (c and d): reproduced with permission from ref. 159. Copyright © 2024 Elsevier B.V. All rights reserved. Panel (e and f): reproduced with permission from ref. 161. Copyright © 2023 Elsevier B.V. All rights reserved. Panel (g and h): reproduced with permission from ref. 162. Copyright © 2023 Elsevier B.V. All rights reserved. | ||
Despite the different dopants employed, Table 3 reveals optimization mechanisms or effects that are generally observed in this type of doping strategy. In fact, some common effects induced by aliovalent dopants are similar: (1) the reinforcement of TM–O bonds; (2) the increase of the inter-slab spacing (sodium layer spacing); (3) reduction of the Na+ activation barrier energy; (4) the inhibition of phase transitions.
In general, introducing high-valence metal cations strengthens TM–O interactions. This enhanced coupling between TM cations and O2− ions weakens the Na+–O2− interaction, resulting in a contraction of the TM slab and an expansion of the sodium layers. The enlarged sodium-layer spacing favors Na+ diffusion, ultimately improving the rate capability of the material.159,160 In fact, the marked enhancement in rate capability arises not only from Na+ diffusivity but also from the influence of high-valence cation doping on the Na+ migration activation energy and the electronic transport properties of the electrode material (Fig. 13). Compared with pristine NFM, the high-valence cation-doped NFM exhibits a substantially narrower band gap. This band-gap reduction lowers the energy needed for electronic excitation from the valence to the conduction band, thereby improving the electronic conductivity of the cathode.160 Moreover, the contraction of the slab thickness (TM layer spacing) plays a key role in enhancing the structural stability of the cathode material.161 Despite the significant stabilizing effect and enhancement of phase transition reversibility induced by the incorporation of high-valence cations in the optimized compositions, the phase transition processes during charge–discharge cycling are relatively complex, involving structural changes and/or the coexistence of biphasic systems, as revealed by ex situ XRD experiments (Fig. 13i–l). Moreover, distinct phase evolution pathways are observed, even when originating from the same O3 phase. In fact, stabilization mechanisms differ depending on the dopant. For instance, Ta5+ incorporation can retard the structural transition from the hexagonal O3 phase to the hexagonal P3 phase.160 In contrast, Zr4+ doping “slows down the phase transition” while preserving the integrity of the crystal lattice, without inducing significant structural damage.161 Furthermore, the Nb5+-doped sample displays a broader operating voltage window than pristine medium-entropy Na-NFM in the pure P3 phase, indicating enhanced structural robustness and a more effective reversibility of the phase transformation.162
![]() | ||
| Fig. 13 3 × 3 × 1 supercell model (a and e), calculated diffusion barriers (b and f), and band structure (c, d, g and h) for medium-entropy NaNFM cathodes with and without high-valence cation doping strategy. Ex situ XRD pattern of (i and j) medium-entropy NaNFM cathodes (pristine NFM) and medium-entropy NaNFM cathodes doped with (k) Nb5+ (named NFM-Nb 0.01) and (l) Zr4+ (named NFM-Zn 0.01) during the first charging/discharging process. Panel (a–d): reproduced with permission from ref. 160. Copyright © 2024 Elsevier Ltd All rights reserved. Panel (e–g): Reproduced with permission from ref. 159. Copyright © 2024 Elsevier B.V. All rights reserved. Panel (i and k): reproduced with permission from ref. 162. Copyright © 2023 Elsevier B.V. All rights reserved. Panel (j and l): reproduced with permission from ref. 161. Copyright © 2023 Elsevier B.V. All rights reserved. | ||
Other aliovalent (and isovalent) cations have also been introduced, with the reported optimized compositions at 0.5 at%. Despite the relatively low dopant content, the incorporation of Al3+ has been shown to effectively suppress the Jahn–Teller effect and improve the structural stability of the materials.157 The underlying mechanism of Al3+ doping under high-voltage cycling was further clarified through ex situ XRD analyses, which revealed an O3–P3–O3′ phase transition sequence during cycling (Fig. 14a and b), closely resembling the behavior observed in other medium-entropy Na-NFM cathodes previously reported in the literature. In contrast, Ti4+ incorporation can suppress the O3–P3 phase transition during cycling155 (Fig. 14c). By restraining this phase evolution, the presence of non-electrochemically active Ti4+ reduces kinetic limitations and mitigates polarization, thereby improving the electrochemical response of the material throughout repeated charge–discharge cycles.
![]() | ||
| Fig. 14 (a) Evolution of the O3–P3 phase during the charge. Ex situ diffractograms of medium-entropy NaNFM cathodes doped with aliovalent cation: (b) Al3+ and (c) Ti4+ during the initial charge and discharge cycle. Panel (a and b): reproduced with permission from ref. 157. Copyright © 2022 Elsevier B.V. All rights reserved. Panel (c): reproduced with permission from ref. 155. Copyright © 2024 Elsevier Inc. All rights reserved. | ||
Interestingly, another meaningful comparison can be made among the Ti-,155 Zn-,156 and Al-157 Na-NFM-based cathodes (with optimized compositions at 0.5 at%), since these materials were evaluated under nearly identical electrochemical conditions. All three compounds exhibit the same configurational entropy value (ΔSconfig = 1.125R), crystallize in the layered O3 phase, and were tested under similar conditions, including a cathode composition of 80
:
10
:
10 (active material
:
carbon
:
binder), a 1 M NaClO4 electrolyte in EC
:
DEC, and a similar mass loading. Under these controlled conditions, the Ti-doped NaNFM electrode material exhibits the highest initial discharge capacity (130.8 mAh g−1 at 1C), followed by the Zn-doped (123.3 mAh g−1) and Al-doped (115.1 mAh g−1) compounds. Since the configurational entropy remains constant among these materials, the observed electrochemical variations are mainly attributed to the distinct structural and electronic effects introduced by the dopant species, as previously highlighted. Despite the differences in specific capacity, all materials display comparable cycling stability, with capacity retentions of 86.2%, 86.5%, and 87.8% after 200 cycles at 1C for the Ti-, Zn-, and Al-doped samples, respectively. These results indicate that although configurational entropy contributes to stabilizing the multicomponent structure, the specific dopant chemistry plays a crucial role in determining the electrochemical behavior of layered metal oxide cathodes.
Despite the promising results highlighted above, some strategies already employed in other medium-entropy Na-NFM cathode systems or in related NaxTMO2 compounds still require further investigation. For instance, the design of dual-doped (or co-doped) medium-entropy NaxTMO2 (at different sites within the material) or quaternary NaxTMO2-based cathodes incorporating electrochemically active metal cations, as designed by the spray-drying process, remains to be more thoroughly explored. Indeed, both approaches demonstrate strong potential for enhancing capacity retention. For instance, Feng et al.164 designed a 2 at% Al- and Cu-co-doped medium-entropy NaNi0.32Fe0.32Mn0.32Al0.02Cu0.02O2 (named A2C2-NFM). Cu-doping, enabled by the reversible Cu2+/Cu3+ redox couple, promotes additional capacity while improving air stability. The synergistic interaction between Al and Cu significantly enhances the kinetics of Na+ transport and structural robustness in A2C2-NFM, resulting in excellent cycling stability with 81% capacity retention after 200 cycles and superior rate performance with 113 mAh g−1 at 5C. In parallel, both the average operating voltage and resistance to air exposure are improved. Accelerated Na+ diffusion in A2C2-NFM promotes a more homogeneous distribution of sodium and suppresses the formation of unstable interfacial phases, thereby enabling more reversible phase transitions. Comprehensive structural and electrochemical characterizations (Fig. 15a–d) reveal that the reinforced layered framework and mitigated Jahn–Teller distortion effectively inhibit crack formation and unfavorable phase transitions, enhancing resistance to electrolyte-induced degradation. Moreover, at higher current densities (1C), the contrast between uniform and non-uniform Na+ distribution underscores that improved Na+ transport kinetics not only boost rate capability but also facilitate homogeneous charge distribution and reversible structural evolution. In fact, XPS analysis reveals that as-prepared A2C2-NFM contains Ni predominantly in the Ni2+/Ni3+ states (Fig. 15e and f). Upon charging to 3.5 V, Ni is mainly oxidized to Ni3+ at a low rate (0.1C), whereas higher-rate cycling (1C) induces the coexistence of Ni2+ and a noticeable fraction of Ni4+, indicating a more heterogeneous Na+ distribution. During discharge at 0.1C, Ni4+ is fully reduced back to its original valence state, while a portion of Ni4+ remains stabilized at the end of discharge under 1C conditions. These results demonstrate that the appearance and persistence of Ni4+ depend on the reaction rate. These differences arise from kinetic limitations in the migration of Na+ relative to the electrochemical reaction. At high rates, insufficient Na+ redistribution leads to non-uniform intercalation/deintercalation and triggers unfavorable phase transitions, such as O′3 and OP2. Overall, the outstanding rate and cycling performance of co-doped medium-entropy A2C2-NFM highlights the critical role of a stable layered framework and widened Na+ migration pathways, providing valuable insights for enhancing structural stability and accelerating the development of practical SIB cathodes.
![]() | ||
| Fig. 15 Ex situ XRD data illustrating the phase evolution during the first cycle at 0.1C for (a) A2C2-NFM and (b) NFM. (c) Schematic representation of the crystal structure changes associated with the phase transitions. (d) First and second voltammograms for p-NFM and A2C2-NFM at a scan rate of 0.1 mV s−1. Ni 2p XPS spectra at various charge and discharge states of A2C2-NFM measured at (e) 0.1C and (f) 1C. Reproduced with permission from ref. 164. Copyright © 2023 Elsevier B.V. All rights reserved. | ||
In addition to performance comparisons in half-cell systems, information on electrochemical performance in full cells was also collected. Interestingly, only a few studies report the assembly of full cells, making it difficult to obtain more in-depth comparisons among NaxTMO2 designed via the spray-drying process. Table 5 compiles papers reporting electrochemical characterization results for medium-entropy Na-NFM as cathode materials, with hard carbon (HC) employed as the anode active material in all studies. Although the results generally indicate that doped medium-entropy Na-NFM cathode materials are superior to their pristine Na-NFM counterparts, the limited data reported for full-cell configurations are not sufficient to establish a clear trend or to clearly identify which dopants incorporated into medium-entropy Na-NFM deliver better performance at a given configurational entropy. For instance, considering the same configurational entropy values (ΔSconfig = 1.125R) and electrochemical characterization procedures, the incorporation of 0.5 at% Zn2+ (ref. 156) or Al3+ (ref. 157) did not show significant differences, as the capacity retention was 82.08% and 82.20%, respectively, after 100 cycles at 1C. On the other hand, the capacity retention of Na-NFMs doped with 1.0 at% Zr4+ (ref. 161) and Nb5+ (ref. 162) (ΔSconfig = 1.144R) was 79.78% and 71.16%, respectively, after 200 cycles at 10C. Despite the difficulties in making such comparisons, the incorporation of 1.0 at% of high-valence cations once again stands out for full-cell applications, as they exhibited good capacity retention under quite severe electrochemical testing conditions, such as 200 cycles at 10C.
| Material | Mass ratio/electrolyte/mass loading | Capacity retention (%) | Discharge capacity (mAh g−1) | Energy density (Wh kg−1) |
Window potential (V vs. Na/Na+) | Ref. |
|---|---|---|---|---|---|---|
| a DEC = diethyl carbonate, DMC = dimethyl carbonate, EC = ethylene carbonate, FEC = fluoroethylene carbonate, PC = propylene carbonate. | ||||||
| NaNi1/3Mn1/ 3Fe1/3O2-Lu (0.02)‖HC | 80 : 10 : 10/1 M NaClO4 in EC : PC (1 : 1) + 5% FEC/2.0–3.0 mg cm−2 |
80.91 (after 500 cycles@5C) | 146@0.1C | 281.3 | 1.5–4.0 | 154 |
| NaNi1/3−xMn1/3Fe1/3ZnxO2 (x = 0.005)‖HC | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm −2‖2.95 mg cm−2 |
82.08 (after 100 cycles@1C) | 75.4@5C | — | 1.5–4.0 | 156 |
| NaNi1/3Mn(1/3−x)Fe1/3AlxO2 (x = 0.005)‖HC | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm−2 |
82.2 (after 100 cycles@1C) | 72.8@5C | — | 1.5–4.0 | 157 |
| NaNi1/3Fe1/3−xMn1/3ZrxO2 (x = 0.01)‖HC | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/— |
79.78 (after 200 cycles@10C) | — | 203.86![]() |
1.5–4.0 | 161 |
| NaNi1/3Mn1/3−xFe1/3NbxO2 (x = 0.01)‖HC | 80 : 10 : 10/1 M NaClO4 in EC : DEC (1 : 1)/5.31 mg cm −2‖2.95 mg cm−2 |
71.16 (after 200 cycles@10C) | 79.4@10C | — | 1.5–4.0 | 162 |
This review provides a comprehensive analysis of the main trends in the design of low-, medium-, and high-entropy NaxTMO2 materials prepared via spray drying. First, the fundamental concepts and key characteristics of spray-drying-derived materials are introduced, with an emphasis on both the advantages and limitations of this synthesis approach. Subsequently, the principal structural features of NaxTMO2 and the associated challenges are discussed, supported by an overview of the properties and conceptual framework of entropy-driven design in NaxTMO2 systems, as expected for oxides containing multiple distinct crystallographic sites. A detailed analysis was also presented to highlight, more specifically, the main advances in the preparation of NaxTMO2.
Spray-drying processes are widely used across various industrial sectors to promote scalability and homogeneity in material preparation. They have demonstrated versatility in producing micro- and nanostructured materials, enabling precise control over particle morphology and size, with high purity when parameters are optimized and additives are employed. Due to its high reproducibility, compatibility with diverse precursors, and ability to generate homogeneous products, the spray-drying process is a promising strategy for the large-scale production of materials for energy storage.
As highlighted in this review, suspensions are typically used to produce precursors, which often require mechanical treatments (e.g., ball milling) and subsequent thermal steps to yield the final electrode active materials. Na2CO3 and metal oxides are the most widely used sodium and transition-metal precursors, respectively, mainly due to their availability, stability, and cost-effectiveness. Ternary medium-entropy NiFeMn-based chemistries dominate current designs, balancing electrochemical performance, sustainability, and supply security. Despite these advances, large variations in spray-drying parameters and insufficient reporting of key operating conditions hinder meaningful comparison across studies. Furthermore, the widespread use of multi-step calcination lacks clear justification, highlighting the need for more systematic studies and standardized reporting to optimize efficiency, cost, and scalability.
In general, over the past five years, spray-drying has been widely used to design mainly low- and medium-entropy NaxTMO2 cathodes (referred to as Na-LEOs and Na-MEOs), with true high-entropy systems (Na-HEOs) remaining relatively rare. Most compositions with ≥80% capacity retention after 200 cycles are NiFeMn-based, highlighting a clear development trend for SIB cathodes. Besides, doping with a fourth element is a dominant strategy, appearing in ∼80% of reported systems and contributing to enhanced electrochemical stability. In fact, spray drying enables homogeneous atomic-level mixing and efficient incorporation of dopants. Interestingly, unary and binary Na-LEOs prepared by spray drying are scarce, as multimetal systems better exploit synergistic effects and mitigate intrinsic drawbacks, such as structural instability and phase transitions. Recent advances focus on Mn-rich multimetal and co-doped systems, which improve Na+ diffusion, structural robustness, and cycling life. However, some multielement compositions described as “high-entropy” do not meet the configurational entropy criteria and should instead be classified as low-entropy systems. This highlights the need for stricter and more consistent use of entropy-based definitions in the field.
Na-MEO-based cathodes synthesized by spray drying have also gained attention, following general trends established for LiTMO2 in LIBs. Ternary NiFeMn-based NaTMO2 (Na-NFM) dominates this field due to its balance between electrochemical performance and synthetic simplicity, with most high-performing systems retaining ≥80% capacity after 200 cycles. A prevalent strategy is aliovalent metal doping at low levels (≈0.5–1 at%), which enhances capacity retention and rate capability despite the dopants being electrochemically inactive. High-valence cations strengthen TM–O bonds, expand Na-layer spacing, reduce Na+ diffusion barriers, and suppress detrimental phase transitions. Nevertheless, the specific stabilization mechanism depends on the nature of the dopant, which influences phase evolution and voltage profiles. Excessive dopant content, however, leads to lattice distortion and performance degradation. Besides, emerging co-doping strategies further improve Na+ transport, structural robustness, and cycling stability, highlighting promising directions for practical SIB cathodes. On the other hand, the design of Na-HEOs by spray-drying is still in its infancy, representing an emerging strategy for SIBs, with only one report to date demonstrating this approach.
Despite these promising studies, challenges remain to fully exploit the potential of spray-drying in the development of advanced NaxTMO2-based cathodes, as summarized in Scheme 2. Considering these results, future research directions should focus on:
(1) Considering the ease of scalability as well as the precise control over synthesis parameters, such as temperature and precursor composition, the spray-drying process applied to precursors of NaxTMO2 cathodes or other active materials for SIBs should be further explored from a coordination chemistry perspective. In particular, this approach may enable the preparation of high-entropy coordination compounds (HE-CCs) and their lower-entropy analogues, which have emerged as promising alternatives for direct use or as precursors to active materials across a wide range of applications.187,188 In this context, the synthesis of NaxTMO2 cathode precursors in the presence of organic ligands and chelating agents may yield more attractive porous morphologies or even serve as a carbon source for the development of other active materials, such as polyanionic compounds. Moreover, the rational design of HE-CCs could also be extended to the synthesis of Prussian blue analogues (PBAs), a class of compounds widely employed as SIB cathodes. Therefore, the development of HE-CCs via spray drying represents a synthetically straightforward, highly scalable, and attractive strategy for the next generation of SIB cathode active materials. Additionally, from the perspective of coordination chemistry, by incorporating suitable binders and dispersants and by controlling the process kinetics, there remains significant room for exploration to obtain cathode materials with higher tap density.
(2) The spray-drying synthesis of precursors for Na-HEOs is still in its early stages, leaving substantial room for further exploration. Beyond the widely reported NiFeMn-containing compositions, the incorporation of other electrochemically active elements, such as Cr, V, and Co, requires more systematic investigation. Moreover, high-entropy systems involving the doping of a sixth or seventh element may lead to enhanced stabilization by synergistically combining entropic effects with conventional doping strategies. Conversely, Ni-rich medium-entropy systems prepared via spray drying also remain underexplored, despite their high promise as cathode materials, as already evidenced in the literature.189
(3) Despite the significant advances reported in the synthesis of NaxTMO2 cathodes using spray drying, there remains considerable room for process optimization. Improvements are needed not only in the selection of raw materials, but especially in the systematic investigation of readily controlled experimental variables in spray-drying equipment. A major challenge in this context is the limited ability to perform meaningful comparisons or draw robust conclusions, as many studies do not explicitly report key experimental parameters, such as outlet temperature, atomization nozzle diameter, feed rate, and related conditions. Therefore, it is crucial that authors provide more comprehensive descriptions of their experimental procedures, and that editors and reviewers emphasize the importance of detailed reporting of synthesis steps and operating parameters. A standardized reporting framework for spray-drying processes should encompass key parameters such as inlet and outlet temperatures, atomization mode, feed concentration and viscosity, pump speed, nozzle type and diameter, drying gas flow rate, and residence time, as well as post-processing conditions including calcination temperature and duration, heating rate, and atmosphere. In addition, particle size distribution should be reported. Systematic documentation of these parameters is essential to ensure reproducibility and enable meaningful comparisons across studies. Besides, although a few studies have reported the possibility of obtaining high-tap-density electrode materials via spray-drying for energy storage,50,190 in-depth research on the synthesis of high-tap-density NaxTMO2 cathodes that combine enhanced kinetics and high volumetric energy density via this route has not yet been reported. For this purpose, synthetic conditions are crucial, as they enable improvements in electronic conductivity and Li+ diffusivity by controlling the crystal and morphological structures of electrode materials.50
(4) Beyond the lack of detailed experimental information on the spray-drying process, the subsequent thermal treatment steps are also often insufficiently explored or explained. In fact, although many studies report the use of two-step heat treatments, the advantages of adopting this methodology are rarely discussed. Therefore, further studies are needed to clearly justify the necessity of dual-step thermal treatments, explicitly addressing their benefits and drawbacks, especially when cathode precursors are based on metal–organic compounds, which typically undergo thermal degradation at relatively low temperatures, usually below 400 °C. Such experiments should be accompanied by in situ characterization techniques during heat treatment, such as X-ray diffraction and electron microscopy.
Thus, considering the conclusions and perspectives highlighted above, we have no doubt that this review is a significant contribution to the advancement of this field. The compilation of recent findings on the spray-drying synthesis of cathodic materials constitutes a contribution to future investigations. The need for continuous research to achieve reproducibility, scalability, sustainability, and high efficiency in the development of SIBs for energy storage is clear.
| This journal is © The Royal Society of Chemistry 2026 |