Strategies to enhance the supercapacitance performance of 2D MXenes through defect engineering, doping, hybridization, and magnetic field assistance: recent progress and challenges

Abstract

The rapid global shift towards electrification has driven a growing demand for scalable, durable, and high-efficiency energy storage systems. Two-dimensional transition-metal carbides and nitrides known as MXenes are distinct materials due to their tunable surface chemistry, exceptional conductivity, and unique layered architectures. However, challenges like limited interlayer spacings and susceptibility to restacking effects and oxidation often limit their electrochemical performance. This review critically evaluates multiple engineering strategies, such as defect engineering, doping, hybridisation, and magnetic field assistance, enhancing their specific capacitance and improving their cycling stability. Specifically, this review integrates experimental breakthroughs with density functional theory insights into ion mobility and conductivity. Furthermore, it highlights their synergy with the emerging machine learning technology, where algorithms screen over 23 000 MXene structures to identify the optimal candidates for charge storage. This work provides a roadmap for the transition from lab-scale MXene synthesis to AI-driven, high-performance supercapacitor design by bridging multiscale simulations with laboratory data. Finally, prevailing challenges such as the complexity of defect–property correlations and future perspectives to accelerate the practical realization of MXene-based high-performance supercapacitors are discussed.

---

Dhilip kumar Chinnalagu

Dr Dhilip Kumar is currently a University Postdoctoral Fellow (UPDF) at the Centre for Interdisciplinary Research (CIDR), SRM University, AP, India. He completed both his PhD and MSc in Chemistry from the Department of Industrial Chemistry, Alagappa University, Karaikudi, India. His research focuses on the development of two-dimensional (2D) nanostructured materials and carbon-based materials derived from waste resources for electrochemical energy storage applications, particularly supercapacitors. His work also involves the design of advanced carbon materials through in situ doping and co-doping of heteroatoms to enhance their electrochemical performance.

Sathya Arumugam Thirumalai

Mr Sathya Arumugam Thirumalai is currently pursuing an MSc in Nanoelectronic Systems at Technische Universität Dresden, Germany, and works as a student assistant at the Chair of Materials Science and Nanotechnology, TU Dresden, and Leibniz Institute for Solid State and Materials Research Dresden. He completed his BTech in Engineering Physics at the Indian Institute of Technology Roorkee, India. His bachelor's thesis, conducted in collaboration with the Chair of Materials Science and Nanotechnology at TU Dresden, focused on lead-free perovskite photodetectors fabricated using e-beam lithography under the Saxon Student Mobility Scholarship. He gained hands-on nanofabrication experience at the Centre for Nano Science and Engineering (CeNSE), Indian Institute of Science (IISc), and conducted field emission studies using a 30 MeV electron accelerator and DFT studies at the Bhabha Atomic Research Center (BARC). His research integrates experimental techniques, COMSOL simulations, and DFT studies of the electronic and quantum properties of 2D nanomaterials for sensing and energy applications.

Surjit Sahoo

Dr Surjit Sahoo is an Assistant Professor at the Centre for Interdisciplinary Research (CIDR), SRM University, AP, India. He previously served as a DST-INSPIRE Faculty at IIT Jammu and as a Fulbright-Nehru Postdoctoral Research Fellow at Kansas State University, USA, supported by USIEF. He also worked as a National Postdoctoral Fellow at IIT Bhubaneswar. Dr Sahoo earned his PhD in Mechatronics Engineering from Jeju National University, South Korea, along with a master's degree from Jadavpur University and a bachelor's degree from Utkal University. His research focuses on nanostructured electrode materials for electrochemical energy storage devices such as supercapacitors and lithium–sulfur batteries, as well as energy harvesting technologies like nanogenerators for self-powered applications.

Brahmananda Chakraborty

Dr Brahmananda Chakraborty is a scientist in High Pressure & Synchrotron Radiation Physics Division of the Bhabha Atomic Research Centre, Mumbai, and Associate Professor in the Homi Bhabha National Institute. His current research interests include density functional theory (DFT) and molecular dynamics (MD) simulations of materials for energy materials, sensing of gas and bio-molecules, reactor fuel and structural materials, nanomaterials and behavior of materials under high pressure. Dr Chakraborty obtained his PhD degree from IIT Bombay, for his research topic on ‘Irradiation Effects and Hydrogen Storage on Carbon Nanotubes’. He did post-doctoral research for two years at North Carolina State University, USA. He received the International Association of Advanced Materials (IAAM) Scientist Medal in 2017. He has published around 350 papers in reputed international journals along with around 9150 Google Scholar citations. He has authored and edited a book titled “2D Metallic Transition Metal Dichalcogenides: Fundamentals and Applications,” published by NOVA Publisher (New York, USA, 2022), as well as several book chapters. Based on his performance in a single year, he was selected among top 2% influential researchers by Stanford University and Elsevier in 2025, 2024, 2023 and 2022.

---

1. Introduction

The modern world is undergoing a major transition in energy practices, prioritizing responsible energy generation and usage to combat climate change while supporting the escalating energy needs of global electrification.¹ The large-scale use of renewable energy sources like wind and solar power is at the heart of this transition. However, these sources are intermittent and pose challenges to maintaining grid stability and reliability. This advancement enables the development of high-performance, reliable energy storage systems which are critically important.² This growing need for efficient energy storage has fuelled the rapid progress of advanced energy storage systems (AESS) on the market, which are projected to expand by billions of dollars in the upcoming years, with applications ranging from utility-scale grid stabilization to powering electric vehicles (EVs) and portable electronics.³ The landscape of energy storage technologies is diverse, with each technology serving a specific role defined by its energy and power density. The Ragone plot visually illustrates this spectrum, with batteries occupying the high-energy-density domain, suggesting the capability of storing large amounts of energy over long periods, analogous to a marathon. However, conventional dielectric capacitors residing in the high-power-density domain are able to deliver or absorb energy very rapidly but in small quantities, analogous to a sprint. However, there is a significant performance gap for high-power operation and substantial energy storage applications like regenerative braking in EVs, grid frequency regulation, and powering high-pulse electronic devices.^4–6

Supercapacitors, also known as ultracapacitors, have become a pivotal technology to narrow the performance gap. They possess a unique combination of attributes that make them superior to batteries in high-power applications, such as fast charge–discharge rates (from seconds to minutes), with exceptionally long cycle life surpassing 500 000 cycles, wider operating temperature range, and enhanced safety profiles.⁷ The specific capacitance of a supercapacitor is primarily determined by the electrode material (C), and hence, its energy density is given by , where V is the operating voltage window. Energy storage in supercapacitors occurs through two main mechanisms: first, electrical double-layer capacitance (EDLC), which is a non-faradaic process that involves the physisorption of electrolyte ions at the electrode–electrolyte interface, and second, pseudocapacitance, a faradaic process that involves reversible and quick redox reactions occurring close to the electrode surface.^8,9 Pseudocapacitive materials can store much more charge than EDLC materials, making the strategic development of high-performance pseudocapacitive electrodes a key objective in supercapacitor research.^4,10 The invention of graphene in 2004 led to a breakthrough in materials science, especially in low-dimensional materials. The extraordinary potential of two-dimensional (2D) materials for a vast array of applications, with energy storage being a prominent focus has been exposed.¹¹ The distinctive structural and physical features of 2D materials, their high surface-to-volume ratio, make them highly suitable as supercapacitor electrodes. Their atomically thin structure offers an exceptionally large specific surface area (SSA), which maximizes the interface available for electrolyte ion interaction and charge storage.^12,13 This thinness also provides very short ion diffusion paths, enabling quick charge–discharge kinetics crucial for high power density. Additionally, many 2D materials combine high intrinsic electrical conductivity, strong mechanical stability, and chemical resilience, which are crucial for producing durable and efficient electrodes.¹⁴ Besides 2D materials, a wide range of nanocomposite architectures have been developed to enhance the electrochemical performance and multifunctionalities. These include composite nanoparticles such as ZnS@CNT, conducting polymer nanocomposites like Cu-MOF/PANI and Cu-MOF/PPy, and metal oxide–carbon nanocomposites, particularly CNT/metal oxide systems such as CNT/MnO₂, CNT/RuO₂, and CNT/NiO as pseudocapacitive electrodes.^15–17 Binary and ternary metal sulfide heterostructures such as NiCo₂S₄@SnS₂ (NCS@TS) core–shell architectures and CoNi₂S₄/Ni₃S₂ heterostructures have also been reported as effective pseudocapacitive materials. Subsequently, a binary metal sulfide coupled with a ternary metal oxide heterostructure such as structurally assembled MoS₂@CuCo₂O₄ has demonstrated significant promise for high-performance solid-state supercapacitors due to their synergistic redox activity, enhanced electrical conductivity, and improved structural stability.^18–20

The family of 2D materials investigated for supercapacitors has grown far beyond graphene, as illustrated in Fig. 1, encompassing a variety of compositions and structures. This class includes transition-metal dichalcogenides (TMDs), including MoS₂, covalent organic frameworks (COFs)-porous crystalline materials with high surface area and tunable pore structures, Nobel prize 2025-winning layered metal–organic frameworks (MOFs), black phosphorus (BP), and perovskites.²¹ The incorporation of these novel nanomaterials and their composites into supercapacitor electrodes has consistently resulted in significant performance enhancements including higher specific capacitance, greater power and energy densities, and excellent rate capability, highlighting the transformative effect of reduced dimensionality on energy storage. Among these emerging materials, 2D transition-metal nitrides and carbides offer a unique combination of properties that closely align with the demands of advanced pseudocapacitors.^4,12 A breakthrough in 2D materials occurred in 2011 when researchers at Drexel University, led by Yury Gogotsi and Michel Barsoum,²² synthesized the first 2D transition-metal carbide (Ti₃C₂). The new class of materials, named MXenes, reflect their origin from MAX phases and structural resemblance to graphene obtained from graphite. MXenes are produced from the MAX phase, which has the general formula of M_n+1AX_n, where ‘M’ is an early transition metal, ‘A’ is a group 13/14 element, ‘X’ is carbon/nitrogen, and n = 1, 2, or 3. The ‘A’ atomic layer (aluminum or silicon) is selectively etched from MAX phases using a top–down approach.²³ The etching process was performed using hydrofluoric acid (HF) or a mixture of a fluoride salt and hydrochloric acid (e.g., LiF/HCl) to remove the ‘A’ layer and produces 2D sheets with the formula of M_n+1X_n. During this aqueous synthesis, highly reactive MXene surfaces are passivated by functional groups, giving the general formula of M_n+1X_nT_x, where T_x refers to surface terminations such as chlorine (–Cl)/fluorine (–F), oxygen (–O), and hydroxyl (–OH) from the etchants.⁴

Fig. 1 Chronological evolution and discovery of modern 2D nanomaterials for energy storage applications. The timeline highlights the progression of major 2D nanomaterials.

Since the discovery of the first MXene, the family of these 2D transition-metal nitrides and carbides has grown rapidly, with over 40 compositions synthesized experimentally and many more predicted theoretically. This rapid expansion has opened a vast chemical space for tailoring 2D material properties to specific applications.²⁴ MXenes possess a distinct integration of ceramic and metallic characteristics that make them exceptionally suited for electrochemical energy storage, setting them apart from most other 2D materials.²⁵ One of the most striking features is their high metallic conductivity, which TMDs usually do not possess. Unlike many 2D materials, which are semiconductors (TMDs) or semimetals (graphene), the majority of MXenes exhibit metal-like electrical conductivity. For example, single flakes of Ti₃C₂T_x have demonstrated high conductivity (∼24 000 S cm⁻¹), a property that is critical for supercapacitor electrodes because it enables rapid electron transport from the current collector without any hindrance. This high conductivity allows for high power delivery without the need for conductive additives, which typically add weight and volume without contributing to capacitance.^26,27 The key advantages of MXenes lie in their hydrophilic surface chemistry. The presence of polar surface functional groups renders MXene surfaces inherently hydrophilic, allowing them to be dispersed in water to form stable colloidal solutions.^11,28 This property greatly simplifies electrode fabrication using scalable, solution-based techniques such as vacuum-assisted spin-coating, spray-coating, and screen printing.²⁹ Furthermore, the hydrophilic nature of MXenes ensures excellent wettability with aqueous electrolytes, promoting efficient ion transport at the electrode/electrolyte interface. MXenes also offer remarkable tunability in both composition and surface chemistry. The vast combinations of ‘M’ and ‘X’ elements, along with a variety of surface functional groups, provide unprecedented control over electronic structure, interlayer spacing, and electrochemical activity, enabling the optimization of their properties for specific energy storage applications. From the outset, MXenes have been recognized for their dominant pseudocapacitive behavior. Unlike purely EDLC-based materials, MXenes store charge primarily through reversible and quick faradaic reactions involving the transition-metal atoms on their surface, often complemented by ion intercalation between the layers. This mechanism allows MXenes to achieve exceptionally high volumetric capacitances, with some reports exceeding 1500 F cm⁻³, which surpasses traditional carbon-based materials. Despite these intrinsic advantages, pristine MXene electrodes face a significant practical limitation: strong van der Waals forces and hydrogen bonding between individual nanosheets allow them to restack into tightly packed multilayer structures during electrode fabrication. This restacking severely limits access to the electrochemically active surface area, particularly the inner layers, leading to reduced capacitance and poor rate performance.³⁰

To overcome this challenge and realize the full capabilities of MXenes, researchers have developed a range of enhancement strategies aimed at producing open, accessible electrode architectures while retaining the inherent benefits of MXene nanosheets. Surface functionalization and doping involve deliberate modifications of the surface termination groups (T_x) or the introduction of heteroatoms into the MXene lattice. These modifications can generate additional redox-active sites, increase the interlayer spacing, and tune the electronic properties to improve the conductivity and charge storage. The intentional construction of atomic vacancies or other structural defects offers another route to modulate the electronic properties and introduce new active sites for redox reactions and ion adsorption, thereby enhancing the electrochemical performance. Perhaps one of the most effective approaches is hybridization and composite formation, where MXene nanosheets are combined with other functional nanomaterials, such as zero-dimensional nanoparticles (metal oxides), one-dimensional nanotubes or nanowires (carbon nanotubes), and other 2D sheets (graphene). These intercalated “spacers” prevent restacking, create porous, ion-accessible three-dimensional architectures, and often introduce synergistic effects by combining multiple charge storage mechanisms. Collectively, these strategies form the foundation of modern MXene research, enabling the design of high-performance electrodes tailored for advanced energy storage applications.^14,31,32

Recently, numerous reviews have capably introduced the fundamentals of MXenes in energy storage. The articles often focus broadly on the synthesis (HF-based, electrochemical, molten salt, alkali-based approaches), general properties, and cover modification strategies in isolation, such as atomic-scale design and surface terminations.^11,33–37 There are reviews on specific aspects of MXenes like their surface terminal groups, MXene–carbon hybrids, MXene–metal oxide composites, and performance optimisation.^38,39 However, a comprehensive integrated analysis that systematically connects advanced experimental modification strategies with the underlying theoretical, computational, and machine learning frameworks, along with a comparative overview of emerging family MBenes, is still missing. The scope of this review is to bridge this critical gap and act as a comprehensive resource for material scientists and researchers to develop next-generation 2D materials for energy storage applications by providing a different perspective of experimental and theoretical knowledge. Furthermore, various structural modification strategies have been employed to fully exploit the potential of MXene structures. First, the fundamentals of MXene synthesis, properties, and charge storage mechanisms are discussed and quickly advanced to a detailed exploration of cutting-edge modification techniques. The significance and novelty of this work lie in its unique, holistic synthesis of three pillars: (1) recent experimental breakthroughs in material engineering, (2) deep theoretical and computational (DFT) insights into how these modifications alter electronic properties and electrochemical behavior, and (3) the emerging power of machine learning-guided predictive design. By integrating these diverse disciplines, the review provides not just a summary of past progress but a forward-looking roadmap for the rational, predictive design of next-generation and high-performance MXene supercapacitors.

2. Fundamentals of MXenes: synthesis and properties

Two-dimensional layered nanomaterials are highly explored for energy storage applications due to their wide lateral size, high aspect ratio, and ultrathin structure. These features offer a large specific surface area and rapid ion transport and provide abundant active sites that contribute to efficient charge storage. MXenes, a family of 2D transition-metal carbides, nitrides, and carbonitrides, were first introduced in 2011 with the discovery of Ti₃C₂.^40,41 This class of materials has been widely expanded and found applications in a broad range of areas including electromagnetic interference shielding, energy storage, water desalination, optoelectronics, catalysis, and biomedical uses. Ti₃C₂T_x can achieve electrical conductivity as high as 24 000 S cm⁻¹ and also offer liquid-phase processability, high mechanical strength, and tunable surface chemistry. Owing to these advantages, MXenes have become one of the most extensively studied 2D materials, leading to thousands of research publications over the past few decades.^42,43

MXenes follow the general chemical formula of M_n+1X_nT_x, which are composed of n + 1 layers (n = 1–4) of early transition metals (M = groups 3–6). These metal layers are alternated with n layers of carbon and nitrogen atoms, denoted as X. Furthermore, T_x indicates the surface groups (–F, –OH, and –O) attached to the outermost exposed M layers.⁴⁴ X atoms in MXenes occupy the octahedral sites of the metal atoms within a hexagonal lattice (M₆X). The surface functional groups depend on both the chemistry of the transition metal and the type of termination present. MXene synthesis involves three major steps: (1) Preparation of layered precursor materials, which are usually MAX phases (M_n+1AX_n), where A belongs to groups 11–16 (Si, Ga, or Al). Sometimes non-MAX layered structures are utilized that contain multiple A atomic layers (M_n+1A₂X_n) and carbide-based A layers (M_n+1A₃X_n). (2) Selective etching process to remove A layers in which weaker M–A bonds break and stronger M–X bonds remain intact. Consequently, undercoordinated M surfaces are generated that react with etchant species to form surface terminations to yield multilayered MXene sheets.^33,45,46 (3) Delamination of multilayered MXene sheets to generate single/few-layer nanosheets (Fig. 2). In the past decade, the chemical and structural diversity of MXenes has expanded significantly. Researchers have synthesized MXenes with either ordered arrangements or disordered solid solutions of multiple transition metals.

Fig. 2 Schematic of the synthesis of MXene sheets from the MAX phase. The MAX phase composed of layered ternary carbides, nitrides, and carbonitrides (M_n+1AX_n) undergoes HF treatment to selectively etch the “A” layers.

2.1 From MAX phases to MXenes: synthesis and delamination

Converting MAX phases to MXenes is a straightforward top-down synthesis. It is a controlled chemical engineering process that unlocks numerous new applications in biomedicine, energy storage, catalysis, and water purification. After synthesis, delamination is an important step to obtain single- or few-layer MXenes that are easier to access and work better. The first etching makes layered MXene sheets that are bound together by hydrogen bonding and van der Waals forces with small active surface area and ion accessibility (Fig. 3). To overcome this issue, intercalation agents such as dimethyl sulfoxide (DMSO), tetrabutylammonium hydroxide (TBAOH), and large organic cations are used to insert between the layers and weaken the interactions between them. After that, sonication or mechanical shaking breaks the layers apart into independent nanosheets that are only a few nanometers thick. Compared to multilayered nanosheets, delaminated MXenes offer better electrical conductivity, mechanical flexibility, and electrochemical activity.⁴⁷ For instance, delaminated Ti₃C₂T_x performs better as an electrode material in lithium-ion and sodium-ion batteries due to its fast electron transport and good ion intercalation pathways. The surface chemistry resulting from etching the balance of O, OH, and F groups is also crucial in determining the stability and utility of the delaminated nanosheets. Recent developments include electrochemical etching, thermal shock procedures, and even green chemistry methods to make delamination safer and more environmentally friendly.⁴⁸ Delaminated MXenes exemplify a complex interaction between materials chemistry and engineering design.

Fig. 3 Schematic of different MAX phase compositions (M₂AX, M₃AX₂, and M₄₄AX₃) and their conversion to corresponding MXenes (M₂XT_x, M₃X₂T_x, and M₄X₃T_x) through selective etching of A-layers. The MAX phase consists of early transition metals (M), A-group elements (A), and carbon/nitrogen (X).

2.1.1 Etching methods (HF and LiF/HCl).
2.1.1.1 Hydrofluoric acid (HF) etching. The initially reported approach for synthesizing MXenes entails the direct etching of “A” layer from MAX phases utilizing HF. This technique selectively breaks the weaker M–A bonds while keeping the stronger M–X bonds. HF is an effective etching agent, but it is very poisonous and corrosive, posing a big safety and environmental risk. MXenes that have been HF-etched usually have more flaws, and a higher concentration of fluorine terminations could make them less stable and conductive.⁴⁹
2.1.1.2 LiF/HCl etching. Researchers developed a LiF/HCl approach as a safer option. In this method, HCl and LiF are combined to make HF in situ. This method relies on direct contact with concentrated HF and makes it easier to regulate the etching process. Moreover, lithium ions from LiF get between the layers during etching, which makes it easier for the layers to separate into single/few-layer MXenes.⁵⁰ MXenes synthesized by this method usually have better conductivity, fewer flaws, and better electrochemical performance than those synthesized by direct HF etching.

2.1.2 Delamination techniques. After etching, MXenes are frequently found in multilayered stacks. Multilayered MXene nanosheets should be separated into one/few-layered MXene sheets, as represented in Fig. 4.

Fig. 4 Schematic of the synthesis and delamination process of MXenes from the MAX phase. The MAX phase (M₃AX₂) undergoes HF etching in which A-layers (Al/Si) are selectively removed by producing surface-terminated MXenes (M₃X₂T_x) with functional groups such as –OH, –O, or –F.

2.1.2.1 Intercalation method. One of the most common ways to make few/single-layer MXenes is an intercalation-based delamination approach. Small molecules/ions were arranged between the MXene sheets to make the interlayer forces weaker. Some common intercalants are TBAOH, lithium ions (Li⁺), and big organic cations. The quality and yield of the exfoliated MXenes hang on the intercalant. For instance, DMSO makes nanosheets that are well separated with good quality, while TBAOH can make colloidal dispersions of stable single-layer MXenes. However, intercalation needs to be carefully optimized since incomplete or uneven intercalation can cause structures to delaminate partially.⁵¹
2.1.2.2 Sonication-assisted delamination. Sonication-assisted delamination is a common method that uses ultrasonic vibrations to separate multilayered MXenes into thinner sheets. Ultrasonication generates bubbles in a liquid medium that create shear forces and shockwaves in a small area. These mechanical forces are strong enough to pull apart the enlarged MXene layers into single- or few-layer. The best thing about this approach is that it works very quickly, breaking down vast amounts of MXenes into stable colloidal solutions in a short period. Too much sonication can cause flaws like splitting nanosheets into smaller pieces and even lowering the conductivity of MXenes.^52,53 To fix this problem, researchers use controlled sonication with intercalants to obtain high delamination sheets without collapsing the structure.
2.1.2.3 Mechanical shaking/stirring. Compared to ultrasonication, mechanical shaking is a simple way to delaminate 2D nanosheets. Intense mechanical stress was applied between MXene sheets to maintain vast lateral dimensions and structural integrity.⁵⁴ The resultant MXene nanosheets have fewer flaws and cracks, which is good for applications that require excellent electrical conductivity and mechanical stability. This makes it especially useful for supercapacitors, conductive films, and composite reinforcements.
2.1.2.4 Comparative assessment and scalability considerations. Among the discussed strategies, the LiF/HCl in situ etching method combined with controlled intercalation-assisted delamination currently appears more advantageous in terms of scalability and cost-effectiveness. Compared to direct HF etching, the LiF/HCl approach offers improved safety, reduced handling risks, and better control over surface terminations. Additionally, Li-ion intercalation during etching facilitates easier delamination, which reduces the need for excessive mechanical treatment. From a scalability perspective, chemical etching routes are generally more feasible than intensive sonication processes, as prolonged ultrasonication increases energy consumption and may reduce material quality. Mechanical shaking preserved structural integrity but faced limitations in achieving uniform large-scale exfoliation. Therefore, mild in situ HF generation (LiF/HCl) coupled with optimized intercalation and low-energy delamination strategies represents the most practical balance between performance, cost, safety, and scalability for bulk MXene production. Nevertheless, the development of fluorine-free etching methods will be crucial for achieving truly sustainable industrial-scale synthesis.
2.1.2.5 Pros and cons of MXene delamination methods. The intercalation-based delamination method offers the advantage of producing high-quality, well-separated MXene nanosheets by weakening interlayer interactions through the intercalation of molecules and ions like Li⁺, TBAOH, and DMSO. This approach can yield stable single/few-layer dispersions, and requires careful optimization, while incomplete and uneven intercalation leads to partial delamination and poor structural uniformity. However, sonication-assisted delamination provides rapid and efficient exfoliation by generating shear forces and shockwaves capable of breaking multilayer MXenes into thin, stable colloidal nanosheets within a short time. However, excessive sonication may fragment the sheets, reduce the lateral size, decrease the electrical conductivity, and introduce defects. Furthermore, mechanical shaking and stirring offer a milder alternative that preserves the larger lateral dimensions and structural integrity of MXene sheets, resulting in nanosheets with fewer defects, which would be highly helpful for applications requiring excellent conductivity and mechanical stability, like conductive films and supercapacitors. However, mechanical methods are generally slower and exhibit a lower delamination efficiency compared to ultrasonication and intercalation techniques.
2.1.2.6 Removal of functional groups. The removal of functional groups from MXenes through irradiation, purification, and etching is a crucial step for tailoring surface chemistry, improving suitability, and improving electrical conductivity.⁵⁵ This enhances their suitability for energy storage, electromagnetic shielding, catalysis, and electronics. In traditional etching methods (HF), the resulting MXene surfaces are terminated with –OH, –F, and –O functional groups. These groups contribute to dispersion stability and hydrophilicity, which disrupt electronic conductivity, modify interlayer spacing, and alter reaction kinetics in electrochemical systems.^56,57 Irradiation-based treatments like UV, plasma, electron beam, and laser irradiation offer a controllable route to remove these functional groups. Particularly, UV irradiation can break bonds through photochemical and photothermal effects, thereby promoting the desorption of –O and –OH groups and partially converting MXene surfaces into more pristine metallic states. Plasma irradiation with inert gases effectively sputters away weakly bonded surface species while inducing mild defect engineering.^34,58 However, irradiation methods must be carefully optimized because excessive exposure may damage the MXene lattice, accelerate oxidation, and introduce unwanted vacancies in oxygen-rich environments. Furthermore, etching-based removal provides another avenue to tune surface terminations.⁵⁹ Alkaline solutions such as NaOH/KOH tend to remove –F groups, replacing them with –O and –OH. Molten salt treatments can generate nearly bare MXene surfaces by providing a high-temperature ionic environment that destabilizes surface terminations.^60,61 Thermal treatment under an inert gas or in vacuum can also desorb functional groups, with the temperatures tuned to target specific terminations around 300–800 °C. The drawback of etching-based methods is the risk of altering stoichiometry, inducing layer restacking, and promoting oxidation at high temperatures.

2.1.3 Other methods.
2.1.3.1 Electrochemical delamination. Electrochemical delamination is a novel approach for synthesizing high-quality MXene nanosheets that are more controllable and environmentally friendly. This approach uses an electric potential to weaken and separate the stacked MXene layers. In this technique, multilayer MXenes are introduced into the electrolyte solution, and a steady voltage/current is applied through the electrodes. Electrochemical reactions on the electrode surface help disrupt the bonds between the layers. This method has several benefits, including protecting nanosheets from physical damage, maintaining their lateral dimensions, and producing highly stable dispersions.⁶² Electrochemical delamination changes the degree of delamination by altering parameters like voltage, current density, electrolyte composition, and duration. However, the approach needs to be carefully optimized, as excessively high potentials may lead to the oxidation of MXene surfaces.
2.1.3.2 Thermal shock method. Another way to delaminate MXenes into thinner nanosheets is the thermal shock method. This approach depends on gases that are trapped between MXene layers. MXenes are first mixed with water and volatile compounds and heated in a preheated furnace. This generates an internal pressure that delaminates MXene sheets.^63,64 The process needs to be carefully managed since too much heat can destroy the MXene structure, cause oxidation, and lower the conductivity.⁶⁵ Unfortunately, improper heating of MXene sheets leads to partial exfoliation and uneven delamination. Despite these challenges, thermal shock has demonstrated significant potential for fabricating high-quality MXenes appropriate for large-scale industrial applications. The method is advantageous for significant MXene production due to its simplicity, speed, and compatibility with existing thermal processing facilities.
2.1.3.3 Green chemistry-based intercalants. Intercalation methods based on green chemistry are a safer and more sustainable way to accomplish MXene delamination than traditional procedures. This technique uses eco-friendly chemicals made from natural, biodegradable, and low-toxicity sources as intercalants to delaminate MXene layers.⁶⁶ The green intercalants occupy the spaces among the MXene layers, weakening the inter-sheet connections that facilitate separation of the layers upon gentle shaking. Bio-based intercalants contribute additional functional groups that make 2D nanosheets more hydrophilic, stable, and compatible with polymer matrices.^67,68 This method offers significant flexibility because various renewable and bio-inspired compounds can be adapted to suit specific MXene applications.
2.1.3.4 Comparative assessment and scalability considerations. The thermal shock method appears promising for scalable production among the emerging delamination strategies. Electrochemical methods produce high-quality nanosheets but face challenges in scale-up due to equipment complexity, energy consumption, and limited batch throughput. Subsequently, electrochemical delamination requires controlled electrochemical setups, optimized voltage, electrolyte composition, and reaction time. Similarly, green chemistry-based intercalation methods provide environmental benefits and improve sustainability but often require longer processing durations and post-treatment steps that may increase the overall production cost and variability. These approaches are attractive for eco-friendly and application-specific modifications, but achieving uniform large-scale exfoliation can be challenging. Contrarily, thermal shock offers shorter processing times, simpler operation, and the ability to treat larger material quantities, making it cost-effective for bulk fabrication. Thermal shock presents a more practical pathway for large-scale MXene delamination owing to its scalability, operational simplicity, and industrial feasibility. Electrochemical and green intercalation techniques remain advantageous for precision engineering and sustainable material customization.
2.1.3.5 Pros and cons of other methods. Thermal shock, electrochemical delamination, and green chemistry approaches each offer distinct advantages and limitations for MXene exfoliation. Electrochemical delamination provides highly controlled, environmentally friendly layer separation with minimal physical damage and excellent dispersibility. However, it requires careful optimization because excessive potential can oxidize and degrade MXene surfaces. The thermal shock method enables extremely rapid and scalable exfoliation by generating internal gas pressure between layers and is compatible with industrial thermal systems. However, improper heating can cause structural collapse, partial delamination, and oxidation. Green chemistry-based intercalants offer the safest and most sustainable route by using bio-derived, low-toxicity molecules that enhance hydrophilicity, polymer compatibility, and stability. The choice of intercalants must be tailored to specific MXene structures and may yield slower and weaker delamination than more aggressive chemical and electrochemical approaches.

2.1.4 Challenges in the synthesis and advanced path. The large-scale synthesis of MXene nanosheets remains challenging due to several key issues, particularly the dependence on hazardous etching agents like HF. Over-etching might cause flaws that weaken the conductivity and mechanical strength. Additionally, delamination procedures such as sonication and intercalation may damage the structures and reduce the size of the sheets. Another problem is maintaining control over surface terminations. The electrical, catalytic, and mechanical behaviors of MXenes are greatly affected by groups (–O, –F, and –OH) that are generated during etching. Alternative synthesis approaches like molten salt processes and CVD methods could synthesize high-quality MXenes without using harsh chemicals. Currently, computational modeling and machine learning are likely to accelerate the design of MXenes by predicting stable structures. Combining nanomaterials with polymers, metals, and other materials can also make them more stable and provide them with more uses.

2.2 Characterization of MXenes using different techniques

2.2.1 XRD analysis. XRD is a principal method employed to verify the effective synthesis of MXenes from MAX precursors. The etching process results in the loss of the A-layer, causing a significant shift of the (002) peak to lower angles. This signifies an augmentation in interlayer spacing that is attributed to the incorporation of surface terminations and intercalants. Farah Ezzah Ab Latif et al.⁶⁹ etched the Ti₃AlC₂ MAX phase to Ti₃C₂T_x MXene sheets using different concentrations of NaOH (5–30 M) as an etching agent. In Fig. 5(a), the XRD spectra of the MAX phase and MXenes etched with varying concentrations of NaOH are displayed. There is a subtle broadening and a reduction in intensity at the (104) plane, which demonstrates the aggressive etching effect. Eventually, the progressive decrease and eventual disappearance of the characteristic MAX phase peaks such as (104), (101), and (105), along with the emergence and the left shift of the (002) reflection, provide direct evidence of A-layer removal. The absence of residual sharp MAX reflections at higher angles confirms nearly complete conversion. Furthermore, the narrowing of FWHM values at different NaOH concentrations provides insights into the crystallinity and exfoliation quality. A semi-quantitative estimate of conversion efficiency can be obtained by comparing the relative intensity of the MXene (002) peak to that of the residual MAX peaks.

Fig. 5 Structural and chemical characterization of Ti₃C₂T_x MXenes synthesized under varying LiF/HCl etching conditions. (a) XRD patterns showing phase evolution with the increasing etchant concentration; the figure has been reproduced from ref. 69 with permission from Elsevier, Copyright 2025. (b) FTIR spectra indicating surface functional groups; the figure has been reproduced from ref. 69 with permission from Elsevier, Copyright 2025. (c) Raman spectra confirming MXene formation from the Ti₃AlC₂ MAX phase; the figure has been reproduced from ref. 70 with permission from Springer Nature, Copyright 2024. (d and e) SEM and TEM images revealing multilayered and few-layer morphologies: (d) SEM image; the figure has been reproduced from ref. 71 with permission from IOP Publishing, Copyright 2015. (e) TEM image; the figure has been reproduced from ref. 72 under a Creative Commons license, Elsevier, Copyright 2021. (f) XPS spectra showing the elemental composition and chemical states of Ti, O, and C in MXene samples; the figure has been reproduced from ref. 73 with permission from Springer Nature, Copyright 2021.

2.2.2 FT-IR analysis. The FT-IR spectroscopy is used to identify functional groups and confirm the presence of surface terminations such as –O, –OH, and –F. These spectroscopic instruments collectively provide an extensive understanding of the bonding environment of MXenes. The FT-IR technique was utilized to analyze the alteration of MAX phase to MXene by identifying respective shifts and bonding vibrations in the surface-bonded functional groups. The research work done by Farah Ezzah Ab Latif et al.⁶⁹ about the synthesis of MAX phase to MXene demonstrates a positive shift of Ti–O towards a higher wavenumber (Fig. 5(b)). Additionally, the availability of the –OH functional group on the MXene surface was confirmed by the characteristic bands at 3753 and 3257 cm⁻¹. Notably, the MAX phase spectrum lacks the characteristic –OH and Ti–O surface-related bands, while these vibrations clearly emerge after etching. The detected functional groups originate from the surface terminations of MXenes rather than the unreacted MAX precursors. Additionally, the absence of Al-related vibrational features supports complete A-layer removal. MXene samples are properly washed and dried to minimize interference from adsorbed water, ensuring that the observed bands correspond to chemically bonded surface groups (–OH, –O, and –F). The non-appearance of such functional groups in the MAX phase FT-IR spectrum confirms the elimination and termination of elements on the MXene surface.

2.2.3 Raman analysis. Raman spectroscopy offers vibrational data, and is employed to investigate structural alterations in MXenes following etching or functionalization. The RAMAN spectra of exfoliated MXene sheets show a lack of ω₁ (250 cm⁻¹) compared to the MAX phase (Fig. 5(c)). Briefly, the MAX phase showcased three different peaks ω₁ (250 cm⁻¹), ω₂ + ω₃ (436 cm⁻¹), and ω₄ (613 cm⁻¹), which reflect the longitudinal oscillation and intrinsic shear of Al, Ti, and C atoms. The disappearance of the ω₁ mode confirms the removal of the Al layer in the MAX phase and its effective conversion to MXenes. The positional shift and reduced intensity of ω₂ + ω₃ and ω₄ modes indicate structural transformation and weakened Ti–Al interactions. The absence of Al-related vibrational contributions in the etched samples verified the successful conversion of the MAX phase into MXenes.⁷⁰

2.2.4 Scanning and transmission electron microscopy (SEM and TEM) analysis. SEM offers insights into the morphology, surface texture, and stratified structure of MXenes, elucidating characteristics such as flake dimensions and stacking dynamics. Fig. 5(d) shows the SEM images of the exfoliated Ti₃C₂-MXene through HF etching. MXene images displayed a clear separation of layers (<20 nm) with a large specific surface area.⁷¹ Fully converted MXenes typically exhibit an accordion-like, loosely stacked morphology, whereas the residual MAX phase retains a dense and compact layered structure. Therefore, the absence of unetched compact particles in the SEM images can qualitatively indicate high conversion efficiency. TEM provides high-resolution pictures to elucidate atomic configurations, flaws, and exfoliated MXene nanosheets. An increase in d-spacing relative to the parent MAX phase confirms A-layer removal and interlayer expansion. If both compact MAX domains and expanded MXene layers are observed simultaneously, partial conversion can be inferred. The selected area electron diffraction (SAED) patterns developed during TEM investigation further validate the crystallinity and lattice orientation. The HR-TEM images displayed in Fig. 5(e) show uniform contrast sheets with a very thin foil-like structure.⁷² The images also clarified that MXene sheets maintained a hexagonal structure, and the d-spacing value correlates with the (2110) and (0110) lattice planes. The lattice plane values suggest the generation of MXene sheets with the elimination of Al from its MAX phase.

2.2.5 XPS analysis. XPS is essential for determining the surface terminations (–O, –F, and –OH) and the chemical states of transition-metal atoms. This approach is crucial as surface chemistry directly influences the electrical, catalytic, and electrochemical performance of MXenes. The quantitative examination of termination groups aids in customizing their functional features. Fig. 5(f) displays the survey spectrum of Ti₃C₂T_x, where C 1s, Ti 2p, O 1s, Ti 2s, and F 1s elements are clearly detected. The absence of Al at% in the MXene XPS spectrum serves as a quantitative indicator of high conversion efficiency from MAX to MXene. The high-resolution spectrum of Ti 2p can be deconvoluted into Ti–C bonds, Ti–O, and Ti–F corresponding to the intrinsic MXene framework and surface termination formed during etching. Ti–O and Ti–F bonds are absent in pristine Ti₃AlC₂, but appear prominently after etching.⁷³ The presence of such groups confirms that these functional groups are chemically bonded to MXene sheets. Similarly, the O 1s and F 1s spectra verify the formation of –O, –OH, and –F terminations. Quantitative atomic percentage analysis from the survey spectrum enables the estimation of the Ti : C ratio and residual Al content, providing a semi-quantitative measure of conversion efficiency. In this way, the XPS technique highly validates the purity, chemical states, and the surface functional groups available in the MXene sheets.

2.3 Properties of MXenes

2.3.1 Crystal structures and mechanical properties. MXenes remain primarily synthesised by etching an A-layer from the MAX phase. The MXene crystal structure is adopted from its MAX-phase precursors, as displayed in Fig. 6(a). MAX phase materials are a family of ternary layered carbide and nitride compounds that have the combined properties of both metal and ceramics. They have a general chemical formula: M_n+1AX_n, where M represents early transition metals (Nb, Ti, V, Mo, and Cr). “A” represents A-group elements, mostly Si, Al, Ge, or S; “X” is either a carbon/nitrogen element; “n” says about the number of layers (thickness).^74,75 These MAX phase materials crystallise in a coated hexagonal close-packed structure, belonging to the space group P6₃/mmc, hence MXenes also adopt the same with the general chemical formula M_n+1X_n. Carbon/nitrogen layers occupy the octahedral sites among the adjacent M layers. The bonding between M and X is a strong covalent or ionic bond, whereas the weak metallic bond between M and A helps in etching away the A layer from the MAX-phase to synthesize MXenes. For MXenes, the lattice parameter ‘a’ typically lies in the range of 3 Å, and ‘c’ usually has a larger value, approximately from 15 to 25 Å. This is because of the etching of ‘A’ layer from the MAX phase, that forms 2D sheets. The presence of surface termination affects the crystal structure of MXenes, which breaks the hexagonal symmetry and induces lattice distortion. These surface terminations modify the MXenes' chemical formula as M_n+1X_nT_x, where T signifies the surface terminal group. High electronegative –O groups compact the lattice slightly, while –F and –OH expand both ‘a’ and ‘c’ lattice constants, leading to increased interlayer spacings and reduced symmetry from P6₃/mmc to P3m1 or even lower.^76,77Fig. 6(b) presents the atomic structure of the Ti₃C₂ MXene monolayer.

Fig. 6 (a) Schematic of the exfoliation step from the MAX phase (Ti₃AlC₂) to MXenes (Ti₃C₂); the figure has been reproduced from ref. 78 with permission from Elsevier, Copyright 2021. (b) Ball-stick atomic structure of Ti₃C₂; the figure has been reproduced from ref. 79 with permission from Springer Nature, Copyright 2021. (c) Formation energy of various compositions of MXenes; the figure has been reproduced from ref. 80 with permission from the American Chemical Society, Copyright 2016.

MXenes may adopt two different stacking configurations depending on their composition and surface terminations. The first is the ABA (hexagonal) stacking, corresponding to the 2H phase, where the second M layer is positioned directly above the first, resulting in trigonal prismatic coordination. This arrangement can be either metallic or semiconducting, depending on M or X. MXenes with late transition metals (Mo and W) as the M component generally form an ABA stacking structure. In contrast, the ABC (trigonal) stacking, also known as the 1T phase, arises when the second metal layer is shifted relative to the first, resulting in an octahedral coordination environment around the X atoms. This structure typically exhibits metallic behavior and a more symmetric arrangement. MXenes containing early transition metals such as Ti, V, and Nb tend to favor the ABC stacking configuration.³³ Compared to ABC, ABA stacking is found to be energetically favoured. The designed sliding energy barriers for the alteration of ABC near ABA expose actual slight energy barriers down to 0.12 eV per formula unit for some MXenes, and higher ones 1.12 eV per formula unit, for heavier MXenes.^81,82 Strong hydrogen bonding and van der Waals interactions between functional groups on neighboring sheets cause individual layers to restack and agglomerate. This restacking reduces the interlayer spacing, restricting ion intercalation between the layers and ultimately impairing the performance of energy storage devices.⁸³

The structural stability of a material is commonly assessed through its formation energy and cohesive energy. Formation energy reflects the energetic feasibility of converting the parent MAX phase into an MXene structure. Generally, a more negative formation energy corresponds to a higher thermodynamic stability. The formation energy (eV per atom) is calculated using the following equation:

where E_tot (MXene) refers to the total energy of the relaxed MXene structure, which is usually calculated using DFT (density functional theory) simulation. µ_i refers to the chemical potential of each element in MXene (M, X, and T), and n is the number of atoms of that element existing in MXenes. Ashton, Michael, et al.⁸⁰ computed the formation energies of all 54 MXene compounds with oxygen as surface termination and predicted that five MXene compounds that have not been synthesized have their formation energies less than 0.1 eV per atom, and hence, their synthesis is straightforward. Fig. 6(c) provides the formation energies of M_n+1X_nO₂ MXenes for different combinations of elements. This analytically helps us determine the compounds that have the potential to be experimentally synthesized.⁸⁴ Cohesive energy is another important parameter that determines the structural stability of a compound. This reflects the bonding strength among atoms within the MXene structure, which also refers to the energy required to break the material into isolated free atoms. The cohesive energy formula is given as follows:where E_i represents the energy of an isolated atom of each element from MXenes. The higher the cohesive energy, the stronger the bonding and structural stability. Cohesive energies of stable MXenes are around 5–8 eV per atom, similar to other strongly bonded 2D materials (like MoS_2, 5.02 eV per atom, or graphene, 7.85 eV per atom).⁸⁵ For instance, a recent study has shown that 2D TiC with a tetragonal sandwich structure Ti₃C_3, which has an elevated cohesive energy of 6.72 eV per atom, was found to be more stable than standard Ti₃C₂ with a cohesive energy of 6.13 eV per atom.⁸⁶

Similar to other two-dimensional materials, defects and structural disorders are commonly observed in MXenes, particularly when exfoliation is involved in their synthesis. During the selective etching of MAX phases, partial dissolution of surface metal atoms often generates M-site vacancies while incomplete etching or high-temperature treatments can create X-site (C/N) vacancies, both of which locally distort the M–X octahedral coordination and modify lattice parameters. In practice, MXene surfaces are randomly terminated with –O, –F, and –OH groups, leading to surface termination disorder that reduces the overall symmetry and causes lattice corrugation. Furthermore, few-layer MXenes frequently exhibit stacking faults, restacking, and turbostratic disorder, where adjacent layers are misaligned or rotationally offset, resulting in broad asymmetric XRD reflections and diffuse SAED streaks. Raman spectroscopy observes the M–X layer stretching vibration strongly damped with surface termination.⁸⁷ Substitutional or non-stoichiometric disorder arising from the partial replacement of metal atoms or deviation in M : X ratios can further induce strain and local symmetry breaking. Collectively, these structural imperfections broaden diffraction peaks, alter interlayer spacing, and strongly influence the crystallographic stability and anisotropy of MXenes. The comprehensive key structural properties of various MXenes are shown (Table 1).⁸⁸

Table 1 Key structural properties of MXenes

MXene	Lattice constant a (Å)	Lattice constant c (Å)	Cohesive energy (eV per atom)	Young's modulus (GPa)	Elastic constant C11 (GPa)	Tensile strength (GPa)	References
Ti3C2Tx	3.07	19.2 (HF-etched), 18.3 (intercalated)	−6.91	330 ± 30	473	15.4	89–94
Ti2CTx	3.04	24.3	−6.64	610	609	30.52 (arm chair)	95–100
V2CTx	2.9	13.5	−6.26	—	—	—	95, 101 and 102
Mo2CTx	2.7	20.6	—	312	—	20.8	103–105
Nb2CTx	3.13	23.68	−7.257 (–O terminal)	448 (3D crystal)	390	28.5 (Vickers hardness)	106–109
Cr2CTx	2.81	—	−4.2	450.6	350.8	—	97, 110 and 111
Ta4C3Tx	3.12	—	—	296	571	21.92 (hardness)	112 and 113
Mo2TiC2Tx	—	25.8	—	—	—	—	114

---

One of the greatest advantages in dealing with 2D materials is that they are extremely strong, ultra-thin, yet flexible, enabling durable flexible electronics. Theoretical calculations ascertained the mechanical properties of MXenes and suggested that it is very difficult to calculate Young's modulus for a MXene monolayer owing to the limitation of compaction head tip size in the transverse local test area of MXene nanosheets, which usually leads to highly inhomogeneous stress and the generation of a strain field. Tensile tests performed using TEM on 40 nm thick MXene multilayers may not accurately represent the structural properties and performance of a single MXene monolayer. Even though AFM nanoindentation was used to study the mechanical properties of graphene and h-BN, monolayer materials take a single atomic layer. Although a monolayer MXene typically consists of approximately five atomic layers, the arrangement of MXene sheets can cause the AFM tip to slip or deviate from the aligned atomic structure, resulting in unreliable measurements with errors up to ∼30%. To obtain reliable results, Rong, Chao, et al. carried out in situ tensile experiments utilizing an exactly controlled attentive ion beam (FIB) fabrication technique, and Young's modulus obtained has only slight deviation from the theoretical values.^54,89

2.3.2 Electronic properties. The potential of MXenes lies in their ability to tune properties according to their targeted applications. Pristine MXenes are almost metallic, and DFT calculations show MXenes with the DOS at the Fermi level ranging from 1.5 to 4.0 states per eV per formula unit, and experimentally almost all pristine MXenes have a linear association between voltage and current at lower potentials; the representative metallic character due to the dominance of transition-metal (M) d-orbital near the Fermi level is revealed in the total density of state plot in Fig. 7(b). Therefore, the tunability comes from its surface terminal groups –F, –OH, and –O, making MXenes semiconducting. Among the given terminations, MXenes with an –O surface termination have the lowest conductivity. One can say that F is the most electronegative; –F termination pulls charge most strongly, which will lower the Fermi level. However, it remains metallic or semi-metallic. However, with –O termination, the 2p orbitals of O hybridise strongly with the d orbitals of the transition metal, creating bonding and anti-bonding states with a noticeable energy gap between them, making the system semiconductive. For M₂CO₂ (M = Ti, Hf, and Zr) MXenes, the band gap widens with the period of M metal as the metallicity decreases. However, an increased M atom concentration suppresses the contribution of O 2p orbitals, restoring metallic character.^115,116 The near-free electron states in MXenes are a distinctive electronic feature that sets them apart from many other 2D materials. The NFE states refer to the electronic state that behaves like a free electron and has parabolic dispersion, which resides near the surface or interface of the material. Again, surface termination leads to the formation of NFE states in MXenes. This is predominantly observed with a hydroxyl surface terminal group as hydrogen atoms create a positive surface potential, which can trap or localise the weakly bound electrons outside the surface, giving rise to NFE states, as shown in Fig. 7(a). They appear in the band structure as low-dispersion (parabolic) states slightly above or crossing the Fermi level near the Γ point. The corresponding charge density plots show electron clouds outside the surface plane (in vacuum), which is shown in Fig. 7(c). Comparable to the –OH terminal group, O– and F– terminated MXenes also display a trend of NFE states, though located at high energy. These NFE states have unique implications for electron emission, catalysis, gas sensing, and charge storage. These factors contribute to the structural degradation of MXenes under gas adsorption, doping, and elevated pressure conditions.¹¹⁶

Fig. 7 (a) Band structure of Hf₂C(OH)₂ with the NFE state indicated by the orange circle, where E_F is set at 0 eV; the figure has been reproduced from ref. 117 with permission from the American Physical Society, Copyright 2016. (b) Total density of states of the Ti₃C₂ MXene nanosheet; the figure has been reproduced from ref. 79 with permission from Springer Nature, Copyright 2021. (c) Electron localization function (ELF) contour plot for Hf₂C(OH)₂ along with the averaged total potential (blue line); the green line represents the average charge density for the lowest NEF state at the Γ point; the figure has been reproduced from ref. 117 with permission from the American Physical Society, Copyright 2016.

Intercalation offers a tunable handle for controlling the MXene electronic behavior from metallic to semiconducting by modulating the interlayer spacing and inter-flake electron transport rather than changing the intrinsic band structure. With intercalation, the interlayer spacing increases; hence, inter-flake resistance increases (inter-flake hopping dominates), causing MXenes to behave like a semiconductor, and the temperature-dependent resistance (dR/dT) becomes negative. In contrast, with de-intercalation, the inter-flake metallic nature is regained, resistance decreases and dR/dT changes from negative to positive, marking a semiconductor-to-metal transition. Though intercalation expands the layers, which enhances ionic access and increases capacitance, too much intercalation can disrupt electronic connectivity, which leads to lower conductivity and rate performance. This enables electronic property engineering for applications like sensors, energy devices, and flexible conductors.³⁷

While pristine MXenes like Ti₃C₂ and V₂C exhibit metallic or semiconducting behavior (with surface termination) and are topologically trivial, recent theoretical studies have predicted that certain heavy metal MXenes (e.g., W₂HfC₂O₂) with O termination can host a 2D topological protecting state owing to spin–orbit coupling-induced band inversion, yielding to protected edge states which are robust against scattering and defects.¹¹⁸Fig. 8(b) shows the Dirac cone in Mo₂TiC₂O₂, which is a linear crossing of valence and conduction bands forming a cone-shaped dispersion around a Dirac point where the bands merely touch each other, which sometimes stay very close to the Fermi level. These Dirac cones indicate massless Dirac fermions with high carrier mobility and unusual quantum transport phenomena, which in 2D materials are associated with exotic electronic, topological, and potentially spintronic behavior.¹¹⁹ Dirac cones are even witnessed in graphene, which makes it a special 2D material to study quantum electrodynamics.¹²⁰ Recent first-principles work has explicitly reported a Dirac-cone-like crossing at the Γ point in Mo₂TiC₂O₂, highlighting its nontrivial topological signatures and reinforcing the significance of this feature for MXene-based electronic and energy applications.⁹⁷

Fig. 8 (a) Band structure of the biaxial strain on the Ti₂CO₂ MXene; the figure has been reproduced from ref. 121 with permission from the Royal Society of Chemistry, Copyright 2017. (b) Dirac cone in Mo₂TiC₂O₂; the figure has been reproduced from ref. 122 under a Creative Commons license, AIP Publishing, Copyright 2024. (c) Absorption spectra of Ti₃C₂T_x ranging from the visible region to the IR region; the figure has been reproduced from ref. 123 with permission from John Wiley and Sons, Copyright 2021. (d) Dielectric constants of Sc₂C(OH)₂-I and Sc₂CF₂-I; the figure has been reproduced from ref. 124 under a Creative Commons license, Springer Nature, Copyright 2017.

It is also possible to tune between the direct and indirect band gaps in MXenes while maintaining the same composition, primarily through strain engineering. Ti₂CO₂ naturally exhibits an indirect bandgap of 0.32 eV with the valence band maximum (VBM) at the Γ point and the conduction band minimum (CBM) at the M point, as shown in Fig. 8(a). Under a biaxial tensile strain of approximately 4–6%, this material transforms into a direct bandgap semiconductor with the VBM and CBM located at the Γ point. This is due to the orbital modification, which happens when in-plane tensile strain is applied; the lattice expands in the xy-plane while the material thickness decreases slightly, reducing Ti–O and Ti–C distances. This enhances O–Ti/C interactions and significantly increases oxygen's contribution to the CBM, causing the Ti 3d and O 2p states to redistribute. The MXene again stands out in demonstrating how the same material composition can be engineered for different optoelectronic applications, like indirect band gaps for thermoelectric materials and direct band gaps for light-emitting devices, simply by controlling the mechanical strain.¹²⁵

2.3.3 Optical properties. MXenes demonstrate rich and highly tunable optical properties, characterized by strong interactions with electromagnetic radiation across an ultra-broadband, from the vacuum ultraviolet (VUV) to the mid-infrared (mid-IR), visible, and NIR regions. This broadband optical activity is a direct consequence of their electronic structure. The (Table 2) summaries the key electronic properties of various MXenes along with its corresponding observed optical properties. The high concentration of free charge carriers (electrons), which is responsible for their metallic conductivity, gives rise to prominent collective electron oscillations known as surface plasmons (SPs). These plasmonic resonances are particularly strong in the NIR range and are responsible for the strong absorption of light in this region, a property that has been effectively harnessed for applications such as photothermal therapy, thermal camouflage, and electromagnetic interference (EMI) shielding. For a Ti₃C₂ stacked sheet, surface plasmons dominate the screening process at energies 0.3 eV up to 45 nm stack. This dominance arises from the combined effects of effective free-electron dynamics, the Begrenzung effect, and decreased inter-band damping. The SP energies can be tuned within the mid-infrared range (0.2–0.7 eV) by altering the functionalization and/or thickness of the sheets.¹²⁶ Therefore, Ti₃C₂ emerges as a promising building block for plasmonic applications operating in the NIR region. The multisampling analysis (MSA) of the spectroscopic ellipsometry (SE) data provided the calculated real (ε₁), imaginary (ε₂), and components of the complex dielectric function for MXene thin films. The imaginary dielectric functions for Sc₂C(OH)₂-I and Sc₂CF₂-I are shown in Fig. 8(d).¹²⁷

Table 2 Key electronic and optical properties of MXenes

MXene	Bandgap (eV)	Work function (eV)	Conductivity (S m^−1)	Quantum capacitance (µF cm^−2)	Open circuit voltage (OCV) (V)	Diffusion energy barrier (eV)	Optical absorption peaks	References
Ti3C2Tx	Metallic tunable from 0.92–1.75	4.47	2.4 × 10^6	398.19 @ −0.072 V	Ti3C2O2 for Li: 0.88	Ti3C2O2 for Li: 1	260 nm, 320 nm	128–134
							Characteristic plasmonic resonance peak at ∼780 nm
Ti2CTx	Metallic (0 eV)	6.5 for Ti2CO2	5.2 × 10^5	1084.7 for Ti2C(OH)2	Ti2CO2 for Li: 1.4	Ti2CO2 for Li: 0.5	IR to UV range characteristic plasmonic resonance peak at ∼500 nm	96 and 135–139
	Ti2CO2 1.1
V2CTx	Metallic (0.02)	4.5	3 × 10^5	377 F g^−1	—	—	Near-IR to visible	140–143
Mo2CTx	Metallic	—	—	1129.58			Visible to UV	144–147
	Mo2CF2 0.278
Nb2CTx	0.81	4.1	6 × 10^5	1828.4	—	Nb2C for Li: 35 meV and for Nb2CO2: 250 meV	IR to visible	148–150
Cr2CTx	Metallic (0 eV)	—	10^6	894.5 F g^−1	—	—	Visible to UV	151–153
	Cr2CF2 3.5 eV
Hf3C2Tx	0.952	—	Semiconducting	—	—	—	Visible to UV (bandgap material)	154–156

---

The interesting nature of MXenes is that they have intrinsically small mid-IR emissivity (below 10%) reported for 2D Ti₃C₂T_x MXenes, along with high solar absorptance (up to 90%). This holds the greatest spectral selectivity between the reported intrinsic black solar-absorbing materials, making it visible black but infrared white, as shown in Fig. 8(c). This special property can be very useful in solar thermal energy harvesting (higher photothermal conversion efficiency, providing lower emissivity), anti-counterfeiting, thermal insulation, and multispectral camouflage. Additionally, with first-principles calculations, it is observed that the lesser emissivity is found in those films with well-aligned nanoflakes, similar to the substrates, and –F and –OH terminal groups. The behaviour with –O terminal groups does not hold good, and there is no effect of MXenes' thickness.^123,157 This indicates that with oxidation or aging, metallic plasmonic features reduce. However, oxygen-terminated MXenes are promising for optoelectronic applications due to their optical bandgap (∼1.29–1.92 eV) that exists in the visible or IR region.¹⁵⁸

Recently, a new subclass of MXenes called Janus bimetal MXene monolayers (MM′CT₂) have attracted considerable attention owing to their tunable electronic, intrinsic structural asymmetry, and optical properties. The calculated exciton binding energies of these Janus MXenes (187 to 520 meV), which are comparable to those of typical 2D materials, indicate strong excitonic effects that must be considered for an accurate description of their optical bandgap, a key factor in energy conversion processes. The optical gap values typically fall between 0.2 and 0.5 eV below the fundamental band gap due to strong excitonic effects. Furthermore, the power conversion efficiency (PCE), estimated using the Shockley–Queisser limit, spans 23.44% to 32.55%, while under the spectroscopy-limited maximum effective approach (considering quasi-particle effects), it falls between 16.48% and 28.82%. These findings establish Janus MM′CT₂ MXenes as hopeful candidates for next-generation photovoltaic methods.¹⁵⁹ The Ti₃C₂T_x MXene flakes exhibit multicolor photoluminescence (PL), yellow-green, emitting blue and red light below diverse excitation wavelengths. A clear redshift in the PL emission wavelength is observed as the excitation wavelength increases from violet to red light. Furthermore, surface modification involving defect-related anatase TiO₂ sites on the MXene surface enables additional tuning of the PL peak toward the near-infrared region, accompanied by a reduction in their intensity.¹⁶⁰

2.3.4 Magnetic properties. While the most widely studied MXenes, such as those based on titanium, niobium, and molybdenum, are nonmagnetic, the vast compositional flexibility of the MXene family allows for the incorporation of magnetic transition metals like chromium (Cr), manganese (Mn), and vanadium (V). This has opened an exciting new frontier to these 2D materials, offering a rich spectrum of behaviors that are highly tunable and distinct from conventional bulk magnets. Magnetism in the above-mentioned magnetic transition metals originates from the unpaired d-electrons residing on the transition-metal ‘M’ atoms, with the intricate interplay between these d-orbitals, crystal field effects, and surface chemistry governing the final magnetic state, exhibiting robust local magnetic moments. For instance, vanadium-based MXenes have been extensively studied, while manganese-based compositions are predicted to show even stronger magnetism due to Mn's higher d-electron count and tendency toward high-spin configurations. This intrinsic magnetism emerges from competing exchange interactions including direct metal–metal (M–M) coupling and superexchange mediated by the carbon or nitrogen (‘X’) atoms. The delicate balance of these interactions can stabilize a wide variety of ground states, including nonmagnetic, ferromagnetic (FM), antiferromagnetic (AFM), and even complex ferrimagnetic or noncollinear arrangements. This diversity means that different MXenes can be tailored for specific needs. Ferromagnetic systems with parallel-aligned moments are attractive for magnetic recording, while antiferromagnetic systems, characterized by antiparallel spins and zero net magnetization, are highly valued for spintronic applications due to their ultrafast dynamics and insensitivity or frustration to external fields. Further compositional complexity, such as in solid-solution MXenes (e.g., V_2−xMn_xC) or ordered Janus structures (M′M″C₂), introduces additional degrees of freedom through chemical and magnetic disorders, allowing for even finer control over the material's properties, specifically in their magnetic domains.^161,162 Mainly, the Mo-containing MXenes such as Mo₂Ti₂C₃T_x are found to be semiconducting; unlike the metallic behavior seen in the typical Ti₃C₂T_x MXene, with a –OH terminal group, the Mo-MXene has a narrow bandgap of 0.05 eV. This semiconducting nature had been confirmed experimentally by observing the system's resistivity vs. temperature. The resistivity of Mo-MXenes increases with the decrease in temperature from 250 to 10 K. The magnetoresistance at 10 K displayed that they take opposite signs, signifying basically different transport processes. Fig. 10(d) displays the spatial spin-density distribution of Mo₂Ti₂C₂(OH)₂.¹⁶³

Given the extreme sensitivity of these quantum-level interactions, the magnetic properties of MXenes are profoundly influenced by their structural and chemical environments, most notably their surface terminations as observed in electronic or optical properties as well. These functional groups are not passive onlookers but active participants that can dramatically alter the magnetic landscape, as they directly modify the local electronic environment, crystal field splitting, and metal–metal distances, which, in turn, impact the magnitude of magnetic moments and the strength of exchange coupling. This effect is highly dependent on the system's composition; in some MXenes, oxygen termination may quench magnetism by promoting hybridization, while in others, it can enhance it by increasing electron localization.^99,164 This tunability via surface chemistry is one of the most powerful tools available for engineering MXene magnetism, allowing properties to be optimized through synthesis control or post-synthetic treatments. Because of this complexity, first-principles calculations, particularly DFT, have been indispensable in mapping the vast magnetic phase diagram of the MXene family. These theoretical studies predict the stability, magnetic moment magnitude, and ordering of countless compositions, guiding experimental efforts. A particularly exciting theoretical prediction for spintronic applications is the existence of significant spin polarization at the Fermi level.¹⁶⁵ Certain nitride-based or ordered double-transition-metal MXene systems are even predicted to exhibit half-metallic behavior, where one spin channel is metallic and the other is insulating, leading to a theoretical 100% spin polarization of charge carriers.¹⁶⁶ While computations predict magnetic ordering temperatures spanning from cryogenic levels to potentially above room temperature for specific compositions, the experimental validation of these ground states and high-temperature ordering in pristine materials remains a significant and active research frontier.

The most novel and transformative application, however, lies at the intersection of magnetism and electrochemistry. Recent computational studies, such as those on V_2−xMn_xCO₂ solid solutions, have revealed a direct and substantial coupling among the material's magnetic order and its electrochemical performance. It has been shown that both chemical disorder and magnetic structure can modulate the total capacitance, and more strikingly, that the surface magnetic structure can actively enhance the redox charge transfer.¹⁶⁶ This discovery opens the door to magneto-electrochemical systems and magnetic-field-assisted energy storage, where an external magnetic field could theoretically be used to tune capacitance, charge/discharge rates, or cycling stability in a supercapacitor. While the experimental realization of long-range magnetic ordering in pristine MXenes remains challenging, the field is actively progressing. To bridge the gap, many current studies employ hybrid or composite approaches, such as embedding MXenes with robust ferrite nanoparticles (e.g., CoFe₂O₄ or NiFe₂O₄).¹⁶⁷ These composite electrodes successfully merge the high conductivity and electrochemical activity of the MXene with the strong magnetic response of the nanoparticles, paving the way for multifunctional supercapacitors that respond to both electrical and magnetic stimuli and providing a practical route to harnessing magnetic effects in energy devices.

Most MXene family members remain non-magnetic, which restricts their potential use in spintronic applications. Consequently, attaining controlled magnetism in 2D materials has become a major research goal. A few pristine MXenes, such as Cr₂C, Mn₂C, Ti₂C, Ti₂N, and Cr₂N, are projected to exhibit intrinsic magnetism. Interestingly, Cr₂C, Ti₂N, and Ti₂C displayed ferromagnetic behaviour, whereas Mn₂C and Cr₂N are anti-ferromagnetic. Nowadays, special attention is given to magnetic MXenes with half-metallicity which behaves as a metal in one spin channel and as a semiconductor in the other, producing fully spin-polarized electrons at the Fermi level.^168,169 In recent years, special attention has been paid to magnetic MXenes with half-metallicity. A half-metal behaves as a metal in one spin channel and as a semiconductor in the other, producing fully spin-polarized electrons at the Fermi level. Cr₂C was the first MXene projected to display a half-metallic characteristic. Notably, the half-metallic gap defined as the energy alteration among the Fermi level and the top of the engaged spin-down band (2.9 eV). Such a wide gap ensures nearly 100% spin filter efficiency across a broad range. Later studies predicted half-metallicity in Ti₂N and Ti₂C. The electronic and magnetic properties of bare MXenes are very sensitive to surface functionalization.

Magnetism in these systems originates from the unpaired electrons of the transition-metal atoms on the surface. When functional groups passivate these surfaces, the magnetic behaviour of most MXenes tends to diminish.¹⁷⁰ For example, Cr₂C functionalized with Cl, F, H, and OH undergoes a transition from a ferromagnetic half-metal to an antiferromagnetic semiconductor with a significant band gap. Contrarily, Cr₂N passivated with oxygen changes from an antiferromagnetic metal to a ferromagnetic half-metal. Recent research works discussed the magnetic properties of MXenes synthesized via different synthesis routes. For instance, Pavla Eliášová et al.¹⁷¹ calculated the magnetic properties of V₂CT_x MXenes (T = OH, F, and O) in both theoretical and experimental methods. The PXRD patterns of all the films except parent V₂AlC displayed similar intensities with the same profiles and the (002) peak. The STEM images of V₂AlC displayed a delaminated sheet-like morphology, confirming the successful delamination process (Fig. 9(a)). The theoretical analysis was calculated with DFT levels utilizing the Vienna Ab initio Simulation package (VASP) and the PBE exchange correlation for the MXene samples. Fig. 9(b) displays the difference in temperature-dependent molar magnetic susceptibility, which suggests that V₂AlC samples demonstrated paramagnetic behaviour dependent on temperature. Furthermore, the M(H) curves of V₂AlC samples recorded at 4 and 300 K temperature exhibit identical behaviour similar to Pauli paramagnetism evaluated at independent temperature (Fig. 9(c)). Similarly, A. Kumar et al.¹⁷² synthesized NiFe₂O₄-MXene (NFO-MX) nanocomposites through the solution-dispersion method and analyzed the magnetic properties using an X-ray magnetic circular dichroism (XMCD) system and a vibrating sample magnetometry (VSM) analyzer. The results of NFO-MX NCs demonstrated a ferromagnetic characteristic with an Fe spin magnetic moment (5.81 ± 0.02–3.26 ± 0.09µ_B) and a Ni spin magnetic moment (3.10 ± 0.04–2.76 ± 0.01µ_B).

Fig. 9 Structural, morphological, and magnetic characterization of V₂C-based MXenes synthesized via different delamination and surface treatment routes. (a) Schematic of type I–III surface terminations, XRD patterns confirming phase formation, and STEM images showing sheet morphology and lattice fringes. (b) Temperature-dependent magnetic susceptibility (χ–T) plots revealing variations in magnetic behavior among the samples. (c) Field-dependent magnetization (M–H) curves at 4 K demonstrating differences in magnetic ordering and anisotropy associated with surface terminations and delamination conditions; the figure has been reproduced from ref. 171 under a Creative Commons license, the Royal Society of Chemistry, Copyright 2024.

2.3.5 Electrochemical properties. The electrochemical properties of MXenes have been established as premier materials in high-performance energy storage research, primarily for supercapacitors, mostly due to their high volumetric capacitance. This family checks all the boxes for an excellent energy-storage material, from high metallic-level electrical conductivity and surface redox activity to 2D layered structures that facilitate the intercalation of ions. Moreover, the high intrinsic density and efficient layered packing in MXenes enable volumetric capacitance values reaching and even exceeding 1500 F cm⁻³, which surpasses nearly all other carbon-based and 2D material electrodes. This perfect mix enables a hybrid charge storage mechanism that fluidly merges fast, non-faradaic EDLC with high-capacity faradaic pseudocapacitance.^26,30 Delocalized d-electron bands enable the inherent metallic conductivity in MXenes, which is a fundamental advantage over most other electrode materials like conducting polymers or metal oxides. It minimizes the need for conductive additives and ensures rapid electron transport for high-rate performance without any leakage. However, this conductivity is highly sensitive to surface chemistry, particularly oxidation, as well as their surface termination, which plays an important role as discussed later. The formation of insulating oxide phases (like TiO₂) on the surface can severely degrade both conductivity and pseudocapacitive performance. This degradation is not necessarily permanent. Recent studies have shown that high-frequency nanoscale vibration can reverse this damage, restoring nearly pristine pseudocapacitance and demonstrating the critical importance of surface integrity.^173,174 The most remarkable feature of MXene-pseudocapacitive behavior is not simple ion adsorption but a true, fast, and reversible surface redox reaction. In Ti₃C₂T_x, this mechanism has been precisely identified through in situ Raman spectroscopy as the reversible bonding (during charging) and debonding (during discharge) of hydronium ions (H₃O⁺) from an acidic electrolyte with the material's surface oxygen terminations, a process that is coupled with a corresponding change in the titanium oxidation state.^175,176

The tunability of MXene electrochemical performance is governed by a complex interplay between their structure, surface, and environment. The 2D layered van der Waals structure is not a rigid framework but a dynamic one that allows for the intercalation of various ions by expanding its interlayer spacing, and thus, improving the actual surface area accessible to the electrolyte. A wide array of cations, from small inorganic ions like H⁺ and Na⁺ to large organic molecules like tetrabutylammonium (TBA⁺), can be inserted between the layers. The intercalation of TBAOH, for example, can produce a massive interlayer expansion (up to 16.58 Å), which fundamentally shifts the storage mechanism from a diffusion-limited process to a predominantly surface-controlled one. This structural modification results in both high capacitance (391 F g⁻¹) and extraordinary cycling stability with 96.3% retention after 10 000 cycles.¹⁷⁷ While intercalation controls physical access, the surface chemistry, arguably the most important factor, dictates the electrochemical reactivity. The surface terminations (T_x), a mixture of –O, –F, and –OH groups, determine the hydrophilicity, ion-binding affinity, and density of redox-active sites. Oxygen-rich surfaces are highly desirable as they provide the active sites for the proton-based pseudocapacitance. Conversely, fluorine terminations, a common byproduct of HF etching, are far less redox-active. This provides a clear optimization strategy: replacing –F with –O or –OH groups, though this must be done with precision, as uncontrolled, excessive oxidation creates the insulating oxides that degrade performance. The choice of electrolyte is the final piece of this puzzle, as it must be matched to the surface chemistry. The high pseudocapacitance of Ti₃C₂T_x, for instance, is most prominent in acidic electrolytes (H₂SO₄) that provide the H₃O⁺ ions necessary for its redox mechanism.^178–180

This intricate engineering of material, surface, and electrolyte has yielded exceptional performance metrics that position MXenes at the forefront of supercapacitor technology. Reported specific capacitances span a wide range, from ∼100 to 700 F g⁻¹ for highly optimized systems. Specific examples from the recent literature highlight this potential; the vacancy-engineered V_1.9CT_x has shown a capacitance up to 760 F g⁻¹,¹⁸¹ while Ti₃C₂T_x–polyaniline (PANI) composites have achieved 430 F g⁻¹ in a full symmetric device, delivering a high specific energy of 38 Wh kg⁻¹.¹⁸² Critically, MXenes do not sacrifice stability or rate capability for this high capacitance. Their metallic conductivity and surface-controlled kinetics enable excellent performance at elevated charge/discharge rates, and well-optimized electrodes routinely demonstrate >90% capacitance retention after 10 000 cycles. When compared to traditional electrode materials, MXenes occupy a unique and advantageous position. They offer substantially higher capacitance than activated carbons (which rely on lower-capacity EDLC), superior electrical conductivity to pseudocapacitive metal oxides (RuO₂/MnO₂), and far greater cycling stability than conducting polymers (which suffer from mechanical degradation).^174,177,182 Computational studies on solid-solution MXenes like V_2−xMn_xCO₂ suggest a direct coupling between the material's magnetic order and its electrochemical properties, where surface magnetic structure may actively enhance redox charge transfer. This opens the tantalizing possibility of developing novel, multifunctional energy storage devices that could one day be tuned or controlled by an external magnetic field.¹⁸³

2.3.6 Vibrational properties. Vibrational spectroscopy, which comprises Raman and IR techniques alongside theoretical density functional perturbation theory (DFPT) phonon calculations, provides an indispensable, non-destructive toolkit for interrogating the fundamental properties of MXenes. These methods are pivotal for characterizing the composition, crystal structure, stability, defect density, and, most critically, surface chemistry. Raman spectroscopy, in particular, has emerged as a go-to technique due to its high sensitivity to both the M–X (metal–carbide/nitride) lattice vibrations and the crucial surface termination groups. For the typical Ti₃C₂T_x MXene, characteristic peaks are typically found between 200 and 800 cm⁻¹, as shown in Fig. 10(a).^184–186 The most transformative application of this technique has been in operando studies, which monitor the material in real time during electrochemical cycling. A landmark operando Raman study of Ti₃C₂T_x in an acidic H₂SO₄ electrolyte provided the first direct, molecular-level evidence of its pseudocapacitive charge storage mechanism with reversible shifts and intensity changes in vibrational modes associated with Ti–O bonds and O–H groups, which correlated perfectly with the charge–discharge current. This elegantly demonstrated that pseudocapacitance arises from the reversible binding and release of hydronium ions (H₃O⁺) at the surface oxygen termination sites, coupled with a simultaneous change in the titanium oxidation state.^{178,184,187,188}

Fig. 10 (a) Raman spectra of the Ti₃C₂T_x monolayer at a low frequency; the figure has been reproduced from ref. 189 under a Creative Commons license, the American Chemical Society, Copyright 2024. (b) XPS showing the recovery of the oxidised MXenes through high-frequency vibration; the figure has been reproduced from ref. 190 under a Creative Commons license, Springer Nature, Copyright 2023. (c) Phonon dispersion curve for the Ti₃C₂ system; the figure has been reproduced from ref. 185 with permission from the Royal Society of Chemistry, Copyright 2015. (d) Spatial spin–density plot of the Mo₂Ti₂C₂(OH)₂ semiconducting MXene; the figure has been reproduced from ref. 163 with permission from the Royal Society of Chemistry, Copyright 2016.

Beyond mechanistic insights, Raman spectroscopy is a frontline diagnostic for material stability. The oxidation of MXenes, a common degradation pathway, produces distinct and strong Raman peaks corresponding to metal oxide phases, such as anatase TiO₂ (e.g., at 144, 399, 516, and 639 cm⁻¹). Monitoring the appearance of these peaks allows for the assessment of degradation, while their disappearance, as shown in studies using high-frequency vibration to remove surface oxides, can quantitatively confirm the recovery of the MXene structure and its electrochemical performance, as shown in Fig. 10(b).^189,190 IR spectroscopy is a critical complement to Raman spectroscopy, as it excels at probing polar bonds and functional groups that involve dipole moment changes. While Raman spectroscopy is more sensitive to the symmetric M–X framework, IR spectroscopy provides a clear window into surface chemistry, which is paramount for controlling the electrochemical behavior. Key IR features include broad O–H stretching bands (3000–3600 cm⁻¹) indicating hydroxyl groups or adsorbed water, M–O stretching modes (400–800 cm⁻¹), and M–F modes from fluorine terminations. This makes IR an ideal tool for tracking post-synthesis modifications, such as the intentional replacement of less-active fluorine terminations with more redox-active oxygen or hydroxyl groups. The combined spectroscopic data confirm that surface terminations are not merely passive functional groups but are active participants in the electrochemical processes, as the operando Raman studies proved.^185,191,192

This experimental work is deeply reliant on theoretical DFT phonon calculations. These calculations are essential for first assessing the dynamical stability of a proposed MXene structure. The absence of imaginary (negative) frequencies in the phonon dispersion curve confirms that the structure is stable, as shown in Fig. 10(c). Furthermore, DFT is used to unambiguously assign experimental Raman and IR peaks by computing the specific atomic displacement patterns (eigenvectors) for each vibrational mode, allowing researchers to confidently identify which bonds and atoms are responsible for each spectral feature. These calculations can systematically predict how different terminations or dopants will alter the vibrational spectrum, guiding the synthesis and interpretation of new materials.¹⁸⁵ In multilayer MXenes, interlayer shear and breathing modes appear at low frequencies and serve as a direct measure of the van der Waals coupling strength between layers. The frequencies of these modes are highly dependent on the number of layers, providing a non-destructive means to assess the degree of delamination and the MXene layer count. This sensitivity is powerfully exploited when studying intercalation, that is, when large intercalants are inserted, they pry the layers apart and weaken the interlayer coupling. For example, the intercalation of TBAOH expands the interlayer spacing to 16.58 Å, causing a significant softening (a shift to a lower frequency) of these interlayer modes, providing clear evidence for successful intercalation. This structural change, confirmed by spectroscopy, is directly correlated with the superior electrochemical performance. Vibrational spectroscopy is also used to characterize defect-engineered MXenes; studies on V_1.9CT_x showed that vanadium vacancies induce specific broadening and shifts in the Raman peaks, and these spectral signatures were correlated with a massive enhancement in capacitance up to 760 F g⁻¹, demonstrating how spectroscopy can guide the optimization of defects for performance.^177,181

2.4 Comparison of properties between MXenes and MBenes

The fundamental distinction between MXenes and MBenes stems from their parent compounds and core elemental compositions, the differences that cascade into profound divergences in their structures, syntheses, and physical properties. MXenes, which have been a major focus of research since 2011, are derived from the MAX phases, which possess a well-defined M_n+1AX_n stoichiometry. Their surface chemistry and etching protocols (detailed in Section 2.1) are well established.^179,193 In sharp contrast, MBenes are born from MAB phases, in which boron (B) replaces carbon or nitrogen, as in the MAX phase. This substitution is not a simple one; it introduces significant structural complexity. MBenes do not follow any single universal formula as MXenes due to boron's unique bonding characteristics. Therefore, they form multiple stable crystal polymorphs, such as the H- and T-type configurations observed in Mo₂B. Fig. 24(a–d) show the ball-stick structure for H- and T-type Mo₂B configurations along with the plots for its band structure (e and f) and density of states in Fig. 24(h and i). Moreover, Fig. 24(g) show the relationship between their total energy and lattice parameter to determine the stable lattice parameter for each configuration. This polymorphism creates a much richer, but far less understood, structural landscape in the research world.^194,195 Furthermore, the M–B bonds are significantly tougher than their M–C counterparts, making the careful etching of the “A” layer and the subsequent delamination into monolayers far more challenging. While this complexity poses a major synthesis hurdle, it also unlocks a new realm of potential properties.¹⁹⁶ MBenes are predicted to possess superior thermal conductivity. For instance, Mo₂B has thermal conductivities of 141 and 146 W m⁻¹ K⁻¹, respectively, for the H- and T-type configurations,¹⁹⁷ significantly enhanced oxidation resistance, and ultralow ion-diffusion barriers, making them theoretically ideal for battery applications.

Currently, these materials occupy the opposite ends of the technology readiness spectrum. MXenes boost extensive experimental validation across dozens of materials and hundreds of device demonstrations.¹⁹⁸ Their processing versatility, enabled by hydrophilic surface terminations and metallic conductivity, allows for easy dispersion into inks and films for printed electronics. However, MXenes face a persistent susceptibility to oxidation and hydrolysis in humid environments, which limits their long-term stability. MBenes, by comparison, remain in a nascent research phase, where most compositions exist only as computational predictions (Table 3). Yet, the few that have been synthesized, such as Mo₂B and Mo₂B, exhibit tantalizing properties that directly address MXene weaknesses. Their exceptional oxidation resistance could enable operation in harsh environments, where MXenes would fail.¹⁹⁹ They also show promise for niche applications, including record-high volumetric capacities (e.g., Mo₂B at ∼2424 mA h cm⁻³) and a 50% reduction in friction for solid lubrication. The challenges associated with MBenes such as difficult synthesis and limited data are primarily due to their recent development. While the next decade will likely see MXenes achieve widespread commercial integration, MBenes are poised to transition from computational promise to experimental validation. By overcoming current scalability constraints and closing the theory-experiment gap, MBenes will strategically carve out specialized roles where their unique boron chemistry and superior stability provide a decisive, mission-critical advantage.^200–203

Table 3 Comprehensive comparison between MXenes and MBenes

Property/aspect	MXenes	MBenes	References
General formula	Mn+1XnTx (M = transition metal, X = C or N, n = 1–4, Tx = surface terminations like –O, –OH, –F, –Cl)	No universal formula; multiple stoichiometries (e.g., M2B, M2B2, MB2) with various stacking arrangements	194, 195 and 200–202
Precursor phase	MAX phases (Mn+1AXn) – layered ternary carbides/nitrides. A = group 13–16 elements – typically Al, Si, Ga, In, Pb, S, Sn	MAB phases (M2AB, MAB2, etc.) – layered ternary borides. A = group 13–16 elements – typically Al
Crystal structure	Layered hexagonal structure with van der Waals gaps	Multiple polymorphs (H-type, T-type)
Discovery/research maturity	Discovered in 2011, extensive experimental literature with thousands of publications – mature field	Emerging field (significant research from ∼2016 onwards), limited experimental demonstrations – mostly theoretical predictions
Number of compositions explored	More than 30 synthesized compositions and more than 100 theoretically predicted	Less than 10 experimentally synthesized and many more theoretically predicted
Synthesis method – etching	Targeted etching of A-layer from MAX phases using HF, LiF/HCl, electrochemical, alkaline, or Lewis acid methods	Targeted etching of A-layer from MAB phases, similar to MXenes but more challenging due to stronger M–B bonds
Delamination/exfoliation	Relatively straightforward with TMAOH, DMSO, or sonication	More challenging due to polymorphism and stronger B–M bonding
Scalability	Multiple scalable routes demonstrated with growing commercial interest	Limited scalability demonstrated to date
Surface terminations	–F, –O, –OH, –Cl (depending on etchant) well-characterized and tunable	–O, –OH, B–O species reported but less characterized termination control less developed
Young's modulus	Composition-dependent; generally 200–400 GPa range for common MXenes	High mechanical strength	89 and 204
	Ti3C2 – 484 GPa	Mo2B – 399 GPa
Electrical conductivity	High, varies by composition and termination	Competitive with metallic MXenes	197 and 205
	Ti3C2 ∼ 2.4 × 10^6 S m^−1	Mo2B ∼10^6 S m^−1
Bandgap	Most are metallic (zero bandgap); some can be tuned to semiconducting state with terminations or composition	Mostly metallic, some theoretical predictions of tunable electronic properties	194 and 201
Thermal conductivity	Composition and termination dependent generally moderate; some compositions show good thermoelectric properties	Substantially higher than MXenes	197 and 206
	Ti3C2 – 50.48 W m^−1 K^−1	H-type Mo2B – 197.9 W m^−1 K^−1 @ 300 K
Thermal stability	Moderate, oxidation susceptible at elevated temperatures in air; varies by composition	Enhanced thermal stability reported, better high-temperature performance in some studies	194, 201 and 207
Magnetic properties	Generally non-magnetic; some compositions can be magnetic with specific terminations or defects	Theoretical predictions of magnetic behavior in some compositions
Theoretical capacity	Ti3C2: 320 mA h g^−1 for Li	V2B2: 968 mA h g^−1 for Li higher than many MXenes	208 and 209
Ion diffusion barriers	Low to moderate, composition-dependent; good for fast charging	Ultralow for some MBenes (Mo2B: 0.0372 eV for Li); exceptional for fast charging	197 and 208
	Ti3C2 – 0.07 eV for Li
Supercapacitors	Leading 2D material for supercapacitors, pseudocapacitive behavior, hydrophilic surface, high power density	Explored for high-power, long-cycle supercapacitors, early stage but promising	203 and 210–213
	PPy/MXene (Ti3C2)/PMFF: 1295 mF cm^−2	Multilayer thin film MoAl1−xB electrode: 2006.60 mF cm^−2
		MBene-MoB: 4025.60 mF cm^−2
Sensor applications	Extensive use in gas sensors, biosensors, electrochemical sensors; high sensitivity	Standalone MBene and MXene/MBene hybrid sensors demonstrated; requires more development	203 and 214–216
		MBene-MOB studied for detecting Cd^2+, Pb^2+, Cu^2+, and Hg^2+, with high sensitivity and selectivity
Biosensing	Well-developed; used for DNA, protein, glucose, miRNA detection	MXene/MBene platforms reported for miRNA detection
Colloidal stability	Form stable dispersions in water and organic solvents	Good colloidal stability; suitable for spray coatings	217 and 218
Biocompatibility	Generally good; some compositions show low cytotoxicity; used in biomedical applications	Limited data. Early studies suggest potential biocompatibility	219 and 220
Toxicity	Composition-dependent; generally low acute toxicity	Limited toxicological data	219 and 221
Advantages	• Tunable surface chemistry	• Enhanced oxidation resistance	194–196, 198, 200–202 and 207
	• Excellent electrical conductivity	• Better thermal stability
	• High capacitance for energy storage	• High thermal conductivity
	• Good mechanical properties	• Ultralow ion diffusion barriers
	• Hydrophilic nature	• Very high theoretical battery capacity
	• Commercial interest and investment	• Excellent solid lubrication properties
		• Promising for harsh environments
		• Novel properties due to boron chemistry
Disadvantages	• Poor oxidation stability and degradation in humid air	• Synthesis is very challenging due to less developed MAB chemistry
	• Limited shelf life	• Limited experimental data
	• HF etching raises safety concerns	• Difficult delamination
	• Environmental impact of synthesis	• Polymorphism complicates characterization
	• Some compositions are difficult to synthesize	• Scalability issues
	• Challenges in termination control	• Precursors are less available
		• Processing methods are underdeveloped
		• Theory-experiment gap is large

---

3. Supercapacitance mechanisms in MXene electrodes

MXenes possess a layered architecture that can undergo partial delamination. This structure features two mechanisms such as ion intercalation and electric double layer (EDL) formation.^178,222

3.1 Fundamentals of electrochemical capacitors

Supercapacitors, often called electrochemical capacitors, remain energy storage devices that correlate high specific power with moderate energy density. This makes them better than regular capacitors and batteries.^223,224 Their operation is essentially based on two mechanisms: EDLC, which comes from electrostatic ion accumulation at the electrode/electrolyte interface, and pseudocapacitance, which involves quick and reversible surface redox processes. The electrode material significantly influences the performance of supercapacitors. Surface area, conductivity, ion accessibility, and chemical stability are all important factors.²²⁵ MXenes remain a group of 2D transition-metal nitrides, carbides, and carbonitrides that show promising characteristics for next-generation supercapacitor electrodes. These materials exhibit distinctive structural features, including metallic conductivity, adjustable surface terminations (–O, –F, and –OH), and hydrophilicity, which promote fast electron mobility and efficient ion intercalation. MXenes can exhibit both double-layer capacitance and pseudocapacitance, enabling a greater specific capacitance and better energy storage ability than traditional carbon-based EDLC materials. The layered structure of MXenes makes it easy for a wide range of electrolyte ions to penetrate. This makes MXenes useful for both aqueous and organic electrolyte systems. Additionally, MXenes can change the spacing between layers and add functional groups, which open up new ways to improve charge storage processes.²²⁶ However, challenges like structural restacking, restricted cycling stability, and environmental susceptibility must be addressed through techniques including surface modification, composite formation, and heteroatom doping.

3.1.1 Electric double-layer capacitance. EDLC is a key part of supercapacitors. The EDLC effect works especially well with MXenes, 2D transition-metal nitrides, carbides, and carbonitrides because of their unique surface and structural properties. When a potential is applied, electrolyte ions move and form a thin layer on the MXene surface that resembles a structure similar to a capacitor without any charge transfer reactions.^227,228 The capacitance is expressed as follows:

C is the capacitance, ε_r is the dielectric constant of the electrolyte, ε₀ is the permittivity of free space, A is the effective surface area of MXenes, and d is the thickness of the double layer. The stored energy follows the classic relation:

where V is the applied voltage. MXenes are useful owing to their high electrical conductivity, which speeds up the movement of electrons. Their hydrophilic surface terminations (–O, –OH, and –F) also allow them to come into close contact with aqueous electrolytes, which lowers d and increases capacitance. The layered architecture of MXenes has interlayer gaps that accommodate ions, significantly improving the available surface area (A) for double-layer formation.^229,230 The performance of EDLC can be improved even more by changing the gap between layers through ion pre-intercalation and surface modification. MXenes are better than regular carbon-based EDLC electrodes because they possess both elevated volumetric capacitance and metallic conductivity. This makes MXenes capable applicants for compact high-power energy storage devices.

3.1.2 Pseudocapacitance (faradaic reactions). MXenes display noteworthy pseudocapacitive activity due to rapid and reversible faradaic redox reactions occurring on their surface. MXenes have a lot of surface terminations (–O, –F, and –OH) and are hydrophilic, which easily interact with electrolyte ions. When ions move into the space between layers and transition metals undergo surface redox reactions (like Ti³⁺/Ti⁴⁺ in Ti₃C₂T_x). This mechanism has both high capacitance and fast charge and discharge rates, which is an advantage over traditional battery-type materials. An essential aspect of MXene pseudocapacitance is that charge storage is not constrained by diffusion rates.²³¹ Instead, it occurs via faradaic reactions confined to the surface, thereby resulting in rapid kinetics. Electrochemical measures like cyclic voltammetry (CV) provide almost rectangular curves with clear peaks exhibiting faradaic contributions (Fig. 11). MXenes exhibit strong conductivity, reduced internal resistance, and enhanced cycling stability, which accelerates the electrochemical process. The dimensions of the ions and electrolyte type also influence the pseudocapacitive response. For example, smaller protons (H⁺) or Li⁺ ions can easily fit between the MXene layers, while larger ions may only be able to stick to the surface. MXenes can be used in both water-based and non-water-based systems owing to their excellent flexible nature.

Fig. 11 Schematic of different charge storage mechanisms in supercapacitors: (a) EDLC based on ion adsorption at the electrode/electrolyte interface, (b) pseudocapacitors involving fast surface redox reactions, and (c) hybrid capacitor combining EDLC and pseudocapacitive behaviors for enhanced energy and power performance; the figure has been reproduced from ref. 232 with permission from Springer Nature, Copyright 2022.

In aqueous electrolytes, ions like H⁺, Li⁺, K⁺, and Na⁺ move into the interlayer galleries of MXenes. Upon addition, surface interactions occur with functional groups (–O, –F, and –OH) and redox transformations of the metal atoms. For instance, the Ti atoms in the Ti₃C₂T_x MXene change from Ti³⁺ to Ti⁴⁺ while keeping the charge balanced by adding protons or cations. MXenes' metallic conductance and narrow diffusion routes make the reactions occur quickly.^233,234 In organic or ionic-liquid electrolytes, bigger cations like Li⁺, K⁺, Na⁺, and tetraalkylammonium ions fit between the MXene layers. The redox activity of surface metal atoms remains significant; however, ion size, solvation, and desolvation processes all influence the mechanism. Li⁺ and other small ions can easily intercalate, maintaining battery-like redox behavior at pseudocapacitive rates. However, larger organic cations may mostly stick to the MXene surface with partial intercalation, which nevertheless gives faradaic pseudocapacitance. In systems devoid of water, pseudocapacitance arises via a mixture of ion intercalation and surface redox reactions, with the dimensions of the ions limiting their penetration depth into the layers. MXenes have a large range of interlayer tunability and strong electrical conductivity, which leads to high capacitance, wide voltage windows, and long-term stability.

3.2 Dominant charge storage mechanisms in MXenes

The most common ways to store charge are EDLC and pseudocapacitance. MXenes can exhibit battery-type intercalation, conversion processes, and surface adsorption/ion separation effects depending upon their structure, content, and electrolyte environment.

3.2.1 Battery-type intercalation reactions. Ions from the electrolyte penetrate further into the interlayer gap, allowing MXenes to function as insertion-type battery electrodes. This approach is crucial for electrolytes devoid of water that comprise Li⁺, Na⁺, and K^{+ 233}. Battery-type intercalation involves diffusion-controlled ion storage and redox transitions of transition metals such as Ti³⁺/Ti⁴⁺.

Ti₃C₂T_x + xLi⁺ + xe⁻ ↔ Li_xTi₃C₂T_x (5)

The battery type intercalation reaction could be improved by altering the distance between successive MXene sheets. Zhiqian Cao et al.²³⁵ described a strategy to enhance the interlayer spacing of MXene sheets by injecting low-valent Zn atoms on MXene's O-termini. Moreover, a battery-type voltage plateau was observed in the fabricated pseudo-symmetric micro-redox capacitor.

3.2.2 Conversion reactions. MXenes store charge through conversion-type charge storage. This means that electrochemically active surface groups or integrated species react with ions to make new phases.²³⁶ For instance, MXenes terminating in oxygen can produce Li₂O when subjected to extensive lithiation in non-aqueous electrolytes:

Ti–O + 2Li⁺ + 2e⁻ → Ti + Li₂O (6)

For instance, Qunbo Hui et al.²³⁷ coated a hollow sphere-structured Sn₄P₃ around MXene nanosheets. The hollow nanostructure provides an inwards volumetric expansion, whereas the MXene coating improves the rate and cycling stability. The coated Sn₃P₄ acts as a conversion-type anode modulating the energy storage mechanism of Sn₃P₄-coated MXene sheets. The electrochemical performance displayed excellent capacitance retention of 92.75% and 95% for Li- and Na-ion storage.

3.2.3 Surface adsorption and ion sieving effects. In addition to EDLC, MXenes can have strong interactions with electrolyte ions on their surfaces. This frequently occurs in ionic liquids and organic electrolytes when large cations or anions interact with surface terminations but are unable to fully intercalate. Such processes resemble ion sieving, wherein only ions of specific sizes or solvation energies can occupy the interlayer galleries. This selective adsorption can make hybrid capacitors have more capacitance and help them stay stable over time.²³⁸ Furthermore, modifying the interlayer space of MXenes by introducing polymers and metal ions facilitates ion accessibility (Fig. 12). Zhennan Chen et al.²³⁹ designed a titanium-based lithium-ion sieve (HTO) membrane by incorporating MXenes to overcome the recyclability challenges of powdery LIS materials. The reports suggested that the incorporation of MXenes enhanced Li⁺ migration and absorption capacity with an excellent stability and selectivity.

Fig. 12 Schematic of different electrochemical lithium storage mechanisms: (a) 50 battery-type intercalation reactions in MXene-based electrodes; the figure has been reproduced from ref. 240 with permission from the American Chemical Society, Copyright 2023. (b) Conversion-type reactions involving reversible redox transformation between metal compounds and Li₂O; the figure has been reproduced from ref. 241 under a Creative Commons license, MDPI, Copyright 2020. (c) Surface adsorption and ion-sieving effects dominated by non-aqueous electrolyte diffusion during charge/discharge processes; the figure has been reproduced from ref. 242 with permission from John Wiley and Sons, Copyright 2023.

3.2.4 Intercalation pseudocapacitance. Intercalation pseudocapacitance is a unique technique of charge storage that links surface-controlled capacitive mechanisms with bulk diffusion-limited battery-like processes. Ions from the electrolyte infiltrate the stratified or porous structure of an electrode material. The kinetics remains rapid and reversible, which is similar to the operation of a capacitor rather than a battery. Intercalation pseudocapacitance differs from conventional battery intercalation since it does not entail slow ion diffusion through the bulk or significant structural alterations. Instead, it is characterized by: (1) intercalation sites: they are located near or adjacent to the surface, allowing ions to access just superficial regions, thereby minimizing the distance they must traverse. (2) Reversible faradaic reactions: during ion intercalation, the oxidation states of transition-metal centers change, which enables increased charge storage. (3) No phase changes or lattice collapse: the host material can hold ions without large changes in its structure, which makes it very stable over time.^243,244 Intercalation pseudocapacitance combines the best features of capacitors (fast kinetics and high specific power) and batteries (high specific energy). This mechanism is essential in MXenes and other two-dimensional layered materials, supporting their exceptional performance in supercapacitors. The Trasatti and Dunn plot is a viable graph to understand the capacitance and diffusion percentage of the MXene and its composite. João V. M. Lima et al.²⁴⁵ developed MXenes doped with PANI and polypyrrole polymers. Fig. 13(a) displays the respective Trasatti and Dunn plot of MXenes, which suggests 65.5% surface capacitance and 52.2% diffusion-controlled mechanism. Subsequent doping of 10% of PANI and PPy with MXenes alters the contribution percentage: 10% of PANI-doped MXenes showcased 22.4%, and 10% PPy-doped MXenes displayed 60.1% of diffusion-controlled contribution. Manopat Depijan et al.²⁴⁶ reported an interfacial engineered MXene with g-C₃N₄ nanosheets for an elevated-performance supercapacitor electrode. The nanocomposite synthesized with different percentages of g-C₃N₄ demonstrated that increasing g-C₃N₄ nanosheets within MXenes increased the percentage of capacitive-controlled contribution (Fig. 13(b)). A research article on VS₄ with MXenes and CNTs as composites (VS-MX100-CNT5) reported by Rahul S. Ingole et al. showcased increased diffusion contribution (69%), as evident from Fig. 13(c).²⁴⁷ Lu Li et al.²⁴⁸ synthesized a Ti₃C₂T_x/PEDOT:PSS (Ti₃C₂T_x/P-100-H) hybrid film with 99.6% surface-controlled capacitance (Fig. 13(d)).

Fig. 13 Capacitive- and diffusion-controlled charge storage kinetics of MXene-based composite electrodes. (a) Quantitative separation of surface-controlled and diffusion-controlled specific capacitance contributions for the pristine MXene, 10% PANI-MXene, and 10% PPy-MXene electrodes; the figure has been reproduced from ref. 245 with permission from Elsevier, Copyright 2024. (b) Variation in capacitive-controlled contribution with scan rate for a-MXene and a-MCN composites; the figure has been reproduced from ref. 246 under a Creative Commons license, the American Chemical Society, Copyright 2024. (c) Comparative capacitive- and diffusive-controlled contribution analysis of VS-MX100 and VS-MX100-CNT5 electrodes at different scan rates; the figure has been reproduced from ref. 247 with permission from John Wiley and Sons, Copyright 2025. (d) Relative percentage contribution of intercalation pseudocapacitance and surface capacitance in Ti₃C₂T_x and Ti₃C₂T_x/P-100-H electrodes as a function of scan rate; the figure has been reproduced from ref. 248 with permission from the Royal Society of Chemistry, Copyright 2019.

3.2.5 Surface redox reactions of functional groups. MXenes (M_n+1X_nT_x, where T refers to the surface terminations such as –O, –F, and –OH) have numerous functional groups that are crucial for their behavior in electrochemistry.²⁴⁹ These surface terminations not only affect the hydrophilicity of the material and the distance between the layers, but also participate in surface redox processes, which are crucial for the pseudocapacitance of MXenes.
3.2.5.1 Oxygen-terminated groups (–O). Surface O groups that are attached to transition metals (like Ti–O in Ti₃C₂T_x) are highly redox-active. They undergo proton- or cation-coupled electron transfer processes during electrochemical cycling. This interaction can go both ways and helps store charge in water-based electrolytes. Similarly, alkali cations (Li⁺, Na⁺, and K⁺) utilize oxygen terminations to provide sites for ion storage and facilitate redox transitions of transition metals (Ti³⁺/Ti⁴⁺).

Ti–O + H⁺ + e⁻ ↔ Ti–OH (7)

3.2.5.2 Hydroxyl-terminated groups (–OH). The –OH groups can also take part in redox reactions that are linked to protons on the surface. Their activity is very much affected by the pH and ionic content of the electrolyte. This contributes to the pseudocapacitive contribution, especially in acidic and neutral electrolytes.

Ti–OH ↔ Ti–O⁻ + H⁺ + e⁻ (8)

3.2.5.3 Fluorine-terminated groups (–F). The strong Ti–F link makes –F groups electrochemically inactive, unlike –O and –OH groups. The redox reactions occurring on the surface functional groups in MXenes confer their pseudocapacitive properties. –O and –OH terminations are the main players in proton- and cation-coupled redox reactions, while –F groups mostly stay out of the way.^250–252 A crucial design strategy for improving the performance of MXenes in supercapacitors and other electrochemical devices is to modify the surface chemistry by increasing the oxygen terminations and decreasing the fluorine content.

4. Advanced modification strategies for the improvement of charge storage performance

4.1 Atomic-scale engineering

4.1.1 Doping. Doping is a broadly utilized strategy to improve the properties of carbon-based materials and graphene (Fig. 14). The incorporation of atoms such as B, N, P, S, and O significantly alters their structure and chemical characteristics. Based on this approach, researchers have also explored doping in MXenes for a varied range of applications. Rare-earth elements are often introduced to enhance the magnetic behaviour, while semiconductor doping mainly targets the improvement of mechanical strength. However, MXenes are typically modified with transition metals, non-metals, and heteroatoms to optimize their performance in electrochemical applications.^253,254 Notably, the overall capacitance of MXenes originates from two main mechanisms: capacitive-controlled and diffusion-controlled processes. The capacitive contribution arises from the formation of the EDLC along with minor surface redox reactions occurring on the electrode surface. However, the diffusion-controlled contribution is associated with a redox reaction primarily driven by ion intercalation/de-intercalation within the bulk of the electrode. Electron-rich atoms such as nitrogen possess a higher electronegativity, generating defects within the lattice that expand the interlayer spacing, which improves inter/de-intercalation.^255,256 Pseudocapacitance is associated with the redox activity of elements within the MXenes phase, and theoretical studies have shown that nitrogen doping highly alters these valence states.²⁵⁷ Moreover, metal atom substitution in MXenes introduces hetero-metal centers with different atomic radii and electronic configurations compared with the parent transition metal. This replacement induces lattice distortion and structural defects among MXene layers, resulting in increased interlayer separation and the formation of extra active sites, as described in detail below.

Fig. 14 Schematic of elemental doping in MXenes via metal and non-metal substitution, highlighting improvements in electrochemical performance and modulation of electrical and magnetic properties.

4.1.1.1 Non-metal substitution (B, N, S, and P). Non-metal substitution, like heteroatom doping, has become a known technique to enhance the specific capacitance, electrical conductivity, and stability of MXenes. Doping the MXene framework with heteroatoms such as N, B, P, and S improves its electrochemical properties. Doping with these heteroatoms induces controlled lattice defects, generates additional electrochemically active sites, and alters charge distribution. These modifications promote rapid electron transport, enhance charge carrier concentration, and improve pseudocapacitive performance. However, excessive doping levels can significantly enhance the electrical conductivity and structural robustness; excessive doping may lead to charge localization and ultimately compromise long-term cycling stability. Among them, B-atom doping provides distinct benefits like boosting the electrical conductivity of MXenes, thereby facilitating faster charge transfer during the electrochemical process. Additionally, B-atom doping improves the hydrophilic nature of MXenes, which strengthens their collaboration with the electrolyte and supports efficient ion transport at the electrode/electrolyte interface.²⁵⁸ Doping of N-atom has become a common strategy to modify the properties of MXenes. The incorporation of N-atoms improves their electrical conductivity by facilitating more efficient charge transfer. Furthermore, N-doping promoted electrolyte penetration, improved surface wettability, and provided greater accessibility of ions. The incorporation of N into the MXene lattice occurs through substitutional doping in which the N-atom partially replaces carbon atoms within Ti₃C₂T_x through surface modification via ammonium treatments. NH₄⁺ ions intercalate electrostatically into the negatively charged MXene interlayer galleries during thermal annealing under an ammonia/N atmosphere. The resultant N-doped MXene exhibits a modified electrochemical behaviour by undergoing the following surface redox reaction during charge/discharge cycling:

Ti₃C₂T_x–N + H⁺ + e⁻ ↔ Ti₃C₂T_x–NH (9)

The reversible proton-coupled electron transfer process provides additional pseudocapacitance beyond the typical EDLC charge storage. N-incorporation generates delocalized electronic states close to the Fermi level by reducing the electron-transfer energy barrier and enhancing the overall charge transport. Particularly, pyridinic-N and pyrrolic-N species generate excessive electrochemically active sites, and graphitic-N improves the electrical conductivity of the basal plane. The capacitance of the N-doped MXene can increase by about 4.6 times compared with the pristine MXene. However, the incorporation of excessive nitrogen may lead to the removal of surface functional groups like fluorine terminations, which can ultimately deteriorate the electrochemical performance. Interestingly, phosphorous atoms induce more significant lattice distortion in the MXene framework than N-atom owing to their large atomic radius and different electronegativities compared with carbon. During thermal treatment with phosphoric acid and other phosphate-based precursors, phosphorous species can either substitute carbon sites within the lattice or attach to the surface functional groups. The incorporation enlarges the interlayer spacing of Ti₃C₂T_x, increasing the d-spacing from 1.265 nm to 1.477 nm. The expanded structure lowers ion diffusion resistance and facilitates improved electrolyte filtration. The electrochemical reaction mechanism associated with phosphorous doping can be described as follows:

Ti₃C₂T_x − PO₄ + nA⁺ + ne⁻ ↔ Ti₃C₂T_x–PO₄(A⁺)_n (10)

where A⁺ denotes intercalating cations such as Na⁺, Li⁺, and H⁺. The presence of phosphate-related functional groups provides reversible adsorption sites that facilitate efficient ion insertion and extraction. Sulfur incorporation also aims to enlarge the interlayer spacing and suppress the restacking of MXene sheets. Similar to P- and N-atoms, S-atom replaces existing surface terminations or partially substitutes carbon atoms in the lattice. This process generates defect-rich sites that enhance pseudocapacitive activity. The electrochemical mechanism associated with S-doping involves the reversible participation of thiol or sulfonate functional groups, described as follows:

Ti₃C₂T_x–SO₃H + e⁻ ↔ Ti₃C₂T_x–SO₃⁻ + H⁺ (11)

The faradaic charge storage occurs through the reversible redox activity of sulfur-containing functional groups. Additionally, S-incorporation reduces electron–hole recombination, facilitates charge transfer at the electrode/electrolyte interface, and improves the overall rate capability of the electrode.²⁵⁹ As a result, these modifications lead to improved electrochemical performance in rate capability and capacitance.²⁶⁰ Doping MXenes with phosphorus and sulfur is a good way to improve their electrochemical characteristics. The incorporation of phosphorus atoms enhances the material's electron density, thus boosting its electrical conductivity. Phosphorus doping also makes MXenes more hydrophilic and improves their pseudocapacitive response, which increases charge storage and overall performance. The incorporation of sulfur improves the conductivity of MXenes and generates new active sites that participate in redox processes, hence improving the pseudocapacitive characteristics of MXenes. For instance, Yangyang Wen et al.²⁶¹ reported the synthesis of a nitrogen-doped two-dimensional MXene (N-Ti₃C₂T_x) via post-etch annealing of Ti₃C₂T_x in ammonia, demonstrating its potential as a high-performance supercapacitor electrode (Fig. 15(a)). The nitrogen content could be precisely tuned from 1.7 to 20.7 at% by adjusting the annealing temperature between 200 and 700 °C. The incorporation of nitrogen expanded the c-lattice parameter from 1.92 nm in pristine Ti₃C₂T_x to 2.46 nm in N-doped MXenes treated at 200 °C. Under optimized conditions, the N-Ti₃C₂T_x electrodes exhibited significantly improved electrochemical performance by achieving specific capacitances of 192 F g⁻¹ in 1 M H₂SO₄ and 82 F g⁻¹ in 1 M MgSO₄. These findings highlight that nitrogen doping effectively modifies the electronic structure and interlayer spacing of the Ti₃C₂T_x MXene, offering a straightforward and efficient strategy to enhance its energy storage performance. Pengcheng Sun et al.²⁶² made a printable ink based on the nitrogen- and sulfur-doped Ti₃C₂T_x MXene (N, S-MXene). DFT analysis revealed that the incorporation of N and S increased the electronic density of states near the Fermi level, enhancing the metallic conductivity and facilitating the rapid electron transport. Morphologically, the N, S-MXene nanosheets kept a layered two-dimensional structure with an expanded interlayer spacing of ∼1 nm and an ultrathin thickness of ∼3.4 nm, which enables efficient electrolyte penetration and short ion diffusion paths. This conductive ink was used to inkjet print planar micro-supercapacitors (MSCs) with a 300 µm electrode spacing (Fig. 15(b)). The produced MSCs had great electrochemical performance with a volumetric capacitance of 710 F cm⁻³, an energy density of 8.9 mWh cm⁻³ at a power density of 411 mW cm⁻³, and 94.6% capacitance retention over time. The enhanced performance resulted from simultaneous nitrogen and sulfur doping, which improved the reaction kinetics, redox activity, and H⁺ adsorption.

Fig. 15 (a) Electrochemical performance of various MXene-based electrodes in 1 M H₂SO₄ and 1 M MgSO₄ electrolytes showing CV and specific capacitance behavior at different scan rates; the figure has been reproduced from ref. 261 with permission from Elsevier, Copyright 2017. (b) Schematic of the N, S co-doped MXene synthesis via annealing, intercalation, and solvothermal processes followed by fabrication steps including inkjet printing and H₂SO₄/PVA coating for flexible supercapacitor assembly; the figure has been reproduced from ref. 262 with permission from Elsevier, Copyright 2022.

4.1.1.2 Metal substitution/adatom. Metal substitution in MXenes involves replacing the native transition-metal atoms with alternative metals to modify the material's electrical, magnetic, and catalytic properties. Mo, V, Cr, Nb, W, Ta, and Ti are all examples of early and late transition metals that can be used as substituents.^263–265 Each one adds its own unique properties to the MXene lattice. For example, replacing Ti with Mo in Ti₃C₂ to form Ti₂MoC₂ alters the electronic density around the Fermi level, thereby enhancing the electrical conductivity and catalytic activity for hydrogen evolution events. W-substituted MXenes like W₂C-based systems have stronger metal–carbon bonds and better thermal and mechanical stability, which is good for high-temperature uses.²⁶⁶ Late transition metals like Co, Ni, and Fe are usually added to surfaces as adatoms instead of replacing lattice sites. This forms localized active sites that make catalytic performance better, especially for reactions that reduce or evolve oxygen. These adatoms change the electrical structure of the surface, which encourages selective reactant adsorption and speeds up the reaction kinetics. The addition of Cr to Ti–Cr MXenes gives them magnetic moments, which could be useful for spintronic applications.^267–269 Mixed-metal MXenes like Ti–Mo or Ti–V systems also use the synergistic effects of the metals to improve conductivity, redox activity, and the ability to change band gaps. In recent days, double-transition-metal MXenes are emerging as a new class of MXenes that incorporate two different transition metals within the layered structures, offering greater compositional and structural tunability than that of single-metal MXenes. The synergistic interaction between two metal components enables improved electronic conductivity, redox activity, and catalytic performance for various electrochemical applications.²⁷⁰

4.1.2 Vacancy and defect engineering. Vacancy and defect engineering has become a useful way to fine-tune the atomic structure of MXenes and improve their physical and chemical properties. Vacancy engineering is the process of making controlled atomic vacancies in the MXene lattice, either at metal or at non-metal locations.²²⁷ In the absence of transition-metal atoms (Ti, Mo, and V), the electronic density is redistributed, resulting in the formation of localized states. This alters the band structures and enhances the catalytic capabilities. Vacancies in carbon or nitrogen sites behave as active centers that speed up the electrochemical reaction kinetics and promote adsorption. For example, carbon vacancies in Ti₃C₂T_x MXenes make ion binding stronger, which greatly increases the amount of lithium or sodium that can be stored. Metal-site vacancies also progress the processes of oxygen reduction and hydrogen evolution by making the adsorption–desorption dynamics of reaction intermediates work better.^271,272 Defect engineering encompasses not just vacancies but also grain boundaries, antisite defects, edge terminations, and heteroatom replacements. These structural abnormalities disrupt lattice symmetry, resulting in chemically active regions that affect surface reactions. Point defects and dislocations facilitate some pathways in overcoming energy barriers, hence enhancing the catalytic selectivity.^273,274 For instance, adding sulfur or nitrogen to vacancy sites has been demonstrated to improve the conductivity and redox activity. This makes MXenes very interesting for use in supercapacitors and metal–air batteries. Certain structural defects can serve as conduits for ion migration, thereby shortening pathways and facilitating the transport of electrolytes to their intended locations. This is very important for electrochemical devices, as fast ion kinetics improve the rate performance.⁸⁸ Moreover, controlled insertion of flaws helps stop sheet restacking by breaking up long-range ordering that keeps porous structures intact and maximizes the available surface area.
4.1.2.1 Inducing and controlling defects. The predominant method for introducing defects in MXenes involves altering the coordination of transition metals or carbon/nitrogen atoms post-synthesis. Ion irradiation is one of these technologies that offers a precise and changeable path. The number of defects can be controlled by changing the ion fluence and energy. This process creates vacancy clusters and lattice distortions, which change the density of catalytic sites. Plasma treatments like Ar, O₂, N₂, and H₂ plasma treatments are other flexible methods that selectively remove surface terminations or sub-lattice atoms to create dangling bonds that serve as reactive centers. Thermal annealing under reducing or oxidizing circumstances also speeds up the movement of carbon or nitrogen atoms at high temperatures.^271,275,276 Too many vacancies can produce electronic localization and lower conductivity, whereas random defect patterns often make reactivity less reliable. To fix this, modern regulatory tactics use heteroatom doping and co-doping to change local bonding networks and stabilize vacancies. For example, adding sulfur or phosphorus to Ti₃C₂T_x helps hold defect sites in place, which keeps them from disappearing and moves the d-band center to improve the catalytic activity. Additionally, intrinsic defect control can be attained during the synthesis of the MAX phase.²⁷⁷ A novel approach is to consolidate defects in a singular location rather than distributing them uniformly. Researchers can generate catalytic “hotspots” by forming isolated defect regions. This does not impact the conductivity of the surrounding pristine domains. This selective technique is possible because of methods like patterned ion-beam lithography and site-specific plasma exposure. Strain engineering also gives you a way to change fault structures without using chemicals.^278,279 Substrate-induced strain or controlled crumpling can affect the stability and movement of defects, allowing for dynamic changes in catalytic performance while the system is running.

4.2 Interface and structural engineering (pillaring and crumpling)

MXenes, a fast-increasing group of two-dimensional transition-metal nitrides, metal carbides, and carbonitrides, have attracted attention because they are very good at conducting electricity and have tunable surface chemistry. Even while MXenes have these benefits, they have problems such as nanosheet restacking and aggregation produced by strong van der Waals interactions. These problems make it harder for ions to diffuse and make the effective surface area smaller. To solve these problems and improve their physicochemical properties, researchers have frequently used structural and interfacial engineering methods, such as pillaring and crumpling. To increase the spacing between MXene layers, nanospacers, polymers, organic compounds, or metal oxides are introduced between them.^280,281 This modification not only avoids undesirable restacking but also enhances ion transport, increases the accessibility of active sites, and improves structural stability during electrochemical cycling. Because of faster charge transfer kinetics and bigger surface areas, pillared MXenes have worked much better in applications including supercapacitors, lithium-ion batteries, and catalysis. Crumpling is the intentional creation of three-dimensional wrinkles or deformations in MXene nanosheets.²⁸² These convoluted frameworks prevent sheets from stacking by creating porous, high-surface-area structures that facilitate the accommodation of ions and molecules. Pillaring emphasizes the augmentation of interlayer spacing and the engineering of interfaces, whereas crumpling introduces three-dimensional structures that are more accessible and possess more durability.²⁸³Fig. 16(a and b) demonstrate the pillaring of MXene sheets with SiO₂ and amine, which also act as spacers in between successive MXene nanosheets.²⁸² Danyang Zhao et al.²⁸⁴ report crumpled MXene sheets engineered with porous structures by alkaline activation and NiCoP deposition. These methods address the challenges associated with pristine MXenes, promoting their applicability in energy storage devices, electrocatalysis, sensing technologies, and electromagnetic interference shielding.

Fig. 16 (a) Conceptual illustration of MXene interlayer regulation via organic–inorganic cooperative intercalation of amines and TEOS, followed by hydrolysis and calcination to generate robust SiO₂-pillared MXenes with hierarchical porosity; the figure has been reproduced from ref. 285 under a Creative Commons license, the Royal Society of Chemistry, Copyright 2021. (b) Schematic of the structural evolution from the Ti₃AlC₂ MAX phase to Ti₃C₂T_x nanosheets by selective etching, alkali-induced crumpling, hydrothermal deposition of NiCo–LDH, and subsequent phosphorization to form Ti₃C₂/NiCoP composites, along with its Na⁺ storage mechanism in a sodium-ion battery; the figure has been reproduced from ref. 284 with permission from the Royal Society of Chemistry, Copyright 2019.

4.2.1 MXene-based hybrids and composites. MXenes have attracted increasing attention for supercapacitor applications owing to their excellent electrochemical properties. However, MXene's inherent tendency to restack face to face and aggregate randomly is an issue common to most 2D materials. To overcome this challenge, various strategies have been developed, with one of the most effective being the incorporation of nanomaterials as spacers within MXene films. Commonly used spacer materials include carbon-based components such as conducting polymers (CPs), carbon nanotubes (CNTs), graphene derivatives (GO/rGO), and activated carbons (ACs), which help to prevent restacking and improve the ion accessibility.^286–288
4.2.1.1 MXene/carbon nanomaterial hybrids (graphene oxides and CNTs). Graphene is very useful for fixing the problems with supercapacitors since it is very strong, conducts electricity well, and is stable at high temperatures. These qualities make it a good choice for making innovative energy storage devices, which have caught the attention of researchers (Fig. 17).²⁸⁹ Graphene has better intersheet van der Waals interactions than CNTs, making the electrode structures stronger and more durable. Graphene and CNTs have a lot in common that make them good for storing energy.^290,291

Fig. 17 Graphical representation of the integration of MXenes with different dimensional carbon materials (0D, 1D, 2D, and 3D) through physical and chemical synthesis routes for supercapacitor applications.

In this regard, Zhimin Fan et al. developed flexible, free-standing MXene/holey graphene films by first filtering dispersions of alkalized MXenes and holey graphene oxides, trailed by a mild annealing process.²⁹² Annealing-induced removal of terminal groups (–F/–OH) increases the density of exposed Ti active sites, thereby enhancing pseudocapacitive performance through intensified Faradaic redox activity and improved charge storage. Concurrently, the incorporation of holey graphene nanosheets prevented severe restacking of MXene layers and introduced interconnected nanoporous channels within the layered structure. Structural analyses revealed an expanded interlayer spacing of ∼7.2° to ∼6.3° and increased accessible surface area, which facilitate electrolyte penetration and shorten ion diffusion pathways. MXenes have proved ultrahigh volumetric capacitance reaching values up to 1445 F cm⁻³ at a scan rate of 2 mV s⁻¹ with excellent rate capability and the ability to accommodate high mass loadings. Moreover, symmetric supercapacitor devices assembled with these MXene electrodes have exhibited remarkable volumetric energy densities with reported values up to 38.6 Wh L⁻¹, representing the elevated performance achieved for MXene-based electrodes in aqueous electrolytes to date. These findings underscore the potential of MXenes not only for high-performance energy storage but also as a versatile platform for the design of next-generation supercapacitors (Fig. 18(a)). Qiang Wang et al.²⁹³ reported the development of a zinc-ion hybrid supercapacitor (ZHSC) using a porous three-dimensional Ti₃C₂T_x MXene-reduced graphene oxide (rGO) aerogel as the cathode and zinc foil as the anode (Fig. 18(b)). This work represents one of the earliest examples of a ZHSC based on 3D MXene architectures. The device demonstrated excellent electrochemical performance, achieving a high specific capacitance of 128.6 F g⁻¹ at 0.4 A g⁻¹ and an impressive energy density of 34.9 Wh kg⁻¹ at a power density of 279.9 W kg⁻¹. Remarkably, the ZHSC retained over 95% of its initial capacitance after 75 000 charge–discharge cycles at a high current density of 5 A g⁻¹. These findings highlight the synergistic advantages of coupling MXene's high conductivity with the porous, mechanically resistant structure of the 3D aerogel, presenting a possible route for high-performance zinc-ion energy storage devices.

Fig. 18 Electrochemical performance and device characterization of MXene-based electrodes and flexible supercapacitors. (a) CV and GCD curves of MXenes and MXene/rGO composites at different scan rates and current densities, with corresponding specific and volumetric capacitances; the figure has been reproduced from ref. 292 under a Creative Commons license, John Wiley and Sons, Copyright 2018. (b) CV, GCD, and EIS of MXene–rGO/ZnSO₄/Zn hybrid supercapacitors, illustrating the rate performance and charge-transfer resistance; the figure has been reproduced from ref. 293 with permission from John Wiley and Sons, Copyright 2019. (c) CV and GCD curves of single and multiple interconnected devices showing areal capacitance and rate capability; the figure has been reproduced from ref. 294 with permission from the Royal Society of Chemistry, Copyright 2021. (d) Photographs of flexible MXene-based micro-supercapacitor arrays with the corresponding electrochemical performance in series and parallel configurations; the figure has been reproduced from ref. 295 with permission from Elsevier, Copyright 2021.

Iijima first wrote about CNTs with one-dimensional nanostructures in the 1990s. CNTs have great thermal, electrical, and mechanical capabilities. CNTs also have a high surface-to-weight ratio, a huge surface area, and a great ability to store charge.^296–298 Extensive research has been conducted on CNT-based materials for several types of energy storage systems. Wenyu Liang et al.²⁹⁴ documented the inaugural manufacture of an asymmetric supercapacitor (ASC) employing negative MXene/MCNT electrodes alongside positive PPy-coated MCNT electrodes. This work investigated novel approaches to improve the electrochemical performance of high-active-mass MXene electrodes (Fig. 18(c)). The conductive MCNT network increased electron transport channels and reduced internal resistance by building an interconnected conductive framework within the layered MXene structure. Simultaneously, the tubular shape of MCNTs worked as spacers between Ti₃C₂T_x nanosheets, inhibiting restacking and providing a porous architecture that promotes electrolyte penetration and ion diffusion even at large active mass loadings (35 mg cm⁻²). Cyclic voltammetry analysis in a Na₂SO₄ electrolyte showed that the Ti₃C₂T_x–MCNT electrodes had an areal capacitance of 1.93 F cm⁻². This is far higher than that previously reported in the literature for pure Ti₃C₂T_x composites with lower electrical resistance. The asymmetric device has a high capacitance of 0.94 F cm⁻² in a 0–1.6 V operating window, which is better than that of the symmetric supercapacitors. Ruochong Wang et al.²⁹⁵ wrote about making an on-chip mSC with a Ti₃C₂T_x–CNT composite as the electrode material. The Ti₃C₂T_x–CNT electrodes that come out of this process have flexible layered channels and customizable electronic architectures. In this structure, CNTs work like tiny spacers tucked between the Ti₃C₂T_x nanosheets, keeping them from clumping together and stacking too tightly. This creates a well-organized, layered network with more breathing room between the sheets and a greater overall surface area. The result is a web of open pathways that ions can travel through more freely, making it much easier for the electrolyte to actually reach and interact with the active MXene surface. At the same time, weaving CNTs into the mix does something important for the electrode's electrical behavior – they form a continuous, well-connected conductive network throughout the material (Fig. 18(d)). The completed mSC device, which is based on a layer-by-layer (LbL) Ti₃C₂T_x–CNT structure, had an areal capacitance of 61.38 mF cm⁻² at 0.5 mA cm⁻². This is higher than that of most MXene- or carbon-based micro-supercapacitors. The gadget can also be connected in series or parallel, which enables storing a lot of energy or obtaining a lot of power, making it promising for real-world applications.

4.2.1.2 MXene/metal oxide composites. Intercalation with pseudocapacitive materials is a successful way to get around the low areal capacitance of MXenes and stop them from self-restacking. MXene–metal oxide hybrids exhibit significant potential for very effective supercapacitors. It is crucial to identify the optimal combination of these components to achieve superior performance as energy storage is significantly influenced by surface area and oxidation states. Combining two or more components is often necessary to improve the performance of metal oxides. This is attributed to the fact that binary oxides generally exhibit electrical conductivity about twice that of single oxides. Therefore, researchers pay more attention to MXene–metal oxide hybrids to improve the capacitive behavior and overall performance of supercapacitors. For instance, Yuming Zhang et al.²⁹⁹ wrote about an easy way to make a Co₃O₄ nanoparticle–MXene (Co-MXene) composite through self-assembly using hydrothermal methods. The incorporation of Co₃O₄ nanoparticles into the MXene framework successfully inhibited nanosheet restacking and reduced ion and electron transport routes by enhancing the electrode's electrochemical accessibility. The Co-MXene electrode had a specific capacitance of 1081 F g⁻¹ (0.5 A g⁻¹), which is far higher than the 89 F g⁻¹ observed for the pristine MXene under the same conditions. The composite electrode showed a high energy density of 26.06 Wh kg⁻¹ at a power density of 700 W kg⁻¹ when used in ASCs (Fig. 19(a)). It also showed good cycling stability retaining 83% of its original capacitance after 8000 cycles at 2 A g⁻¹. Additionally, Qi Xun Xia et al.³⁰⁰ wrote about an ease way to make Ti₃C₂T_x MXene electrodes with NiO nanosheets on them using hydrothermal methods (Fig. 19(b)). In this method, NiO nanosheets were formed on a carbon-supported TiO₂ layer from Ti₃C₂T_x MXenes during thermal annealing. The NiO-grown TiO₂/C-Ti₃C₂T_x MXene (Ni-dMXNC) has high specific capacities of 92.0 mA h cm⁻³ and 53.9 mA h cm⁻³. The device produced a specific energy of 1.04 × 10⁻² Wh cm⁻³ at a specific power of 0.22 W cm⁻³ when utilized in an ASC with Ni-dMXNC as the positive electrode and Ti₃C₂T_x MXene as the negative electrode. The ASC also showed 72.1% of its original capacity after 5000 cycles. This study shows that mixing transition-metal oxides with MXenes can improve the electrochemical performance by adding more redox-active sites and keeping the structure stable throughout repeated cycling. The enhanced electrochemical performance is attributed to the high-surface-area, multilayered nanostructures formed by the combination of metal oxides into MXenes. These studies indicate that such nanostructured composites hold great promise as electrode materials for the progress of advanced, high-performance energy-storage devices.

Fig. 19 Electrochemical characterization of the MXene-based hybrid electrodes and asymmetric supercapacitors. (a) Schematic of the CV and GCD profiles of Co–MXene//PANIC asymmetric devices, demonstrating high rate capability and superior energy density. The Ragone plot compares energy and power densities with previously reported systems, while the cycling performance indicates excellent stability, inset figure shows GCD plot before and after 8000 cycles; the figure has been reproduced from ref. 299 with permission from Elsevier, Copyright 2021. (b) CV and GCD curves of Ni-doped MXene-based hybrid devices at different scan rates and current densities with corresponding specific capacity and cycling stability. The schematic and Ragone plot highlight the structural configuration and enhanced energy–power characteristics of the developed supercapacitor; the figure has been reproduced from ref. 300 under a Creative Commons license, the Royal Society of Chemistry, Copyright 2017.

4.2.1.3 MXene/conducting polymer composites. Conducting polymers have gathered substantial interest in supercapacitor research owing to their intrinsic electrical conductivity. PPy, poly(3,4-ethylenedioxythiophene) (PEDOT), PANI, and polythiophene (PTh) are the most studied polymers. These polymers can conduct electricity when they are doped in different ways because their band gaps (1–3 eV) are substantially lower than those of regular polymers (10 eV).^182,301,302 Conducting polymers also have different shapes, different ways of doping, and reasonably quick charge and discharge kinetics, which makes them very good for storing energy. Yue Li et al.³⁰³ developed a hydrothermal method to synthesize MXene layers anchored with chain-like PANI wrapped by PANI nanofibers to produce a hybrid electrode. The layered MXene framework offers a large surface area and structural robustness, facilitating efficient ion insertion and extraction. The MXene/PANI electrode with a mass ratio of 1 : 3 achieved a high specific capacitance of 563 F g⁻¹ at 0.5 A g⁻¹ and demonstrated excellent rate performance. The electrode showed remaining cycling stability, maintaining 95.15% of its initial capacitance after 10 000 cycles. Expanding to device-level performance, an asymmetric supercapacitor was assembled using MXene/PANI hybrids and activated carbon as the positive and negative electrodes (Fig. 20(a)). The fully packaged device delivered an energy density of 22.67 Wh kg⁻¹ and a power density of 0.217 kW kg⁻¹, underscoring the potential of MXene/PANI composites for high-performance energy storage. Furthermore, Binxia Chen et al.³⁰⁴ demonstrated the use of PANI NFs as conductive interlayer spacers to help the self-assembly of Ti₃C₂T_x MXene-based electrodes (Fig. 20(b)). Electrostatic interactions and hydrogen bonding among the positively charged PANI NFs and negatively charged Ti₃C₂T_x nanosheets enabled the formation of a sandwiched porous structure. This architecture provides numerous accessible sites for ion transport and enhances electrolyte diffusion throughout the electrode. Consequently, the flexible freestanding Ti₃C₂T_x/PANI NFs composite exhibited a high electrical conductivity of 1373.3 S cm⁻¹ and an impressive gravimetric capacitance of 645.7 F g⁻¹. The incorporation of PANI NFs also reinforced the structure, resulting in outstanding cycling stability with 98% capacitance retention after 5000 cycles.

Fig. 20 Electrochemical performance comparison of MXene/PANI composite electrodes with varying composition ratios. (a) GCD and CV curves of MXene/PANI hybrids at different current densities and scan rates, showing enhanced capacitive behavior with the optimized PANI content. Corresponding specific capacitance values highlight the effect of composition on rate capability; the figure has been reproduced from ref. 303 with permission from Elsevier, Copyright 2021. (b) CV, GCD, and EIS plots of MXene–PANI electrodes with different mass percentages, revealing improved conductivity and charge-transfer kinetics. The plot of specific and areal capacitance versus composition demonstrates optimal electrochemical performance at an intermediate PANI loading; the figure has been reproduced from ref. 304 with permission from John Wiley and Sons, Copyright 2021.

4.2.1.4 MXene/TMD composites. Combining MXenes with TMDs like MoS₂, WS₂, and MoSe₂ has become a good way to make multifunctional 2D heterostructures that work better for electronics, catalysis, and electrochemistry. MXenes have strong electrical conductivity and tunable surface chemistry. TMDs possess semiconducting characteristics and several catalytically active sites. TMD nanosheets are often synthesized as MXene/TMD composites using hydrothermal, solvothermal, and chemical vapor deposition techniques.^305,306 Robust interactions at the interface attributable to van der Waals forces or chemical bonding facilitate charge transfer between MXenes and TMDs. This can lower charge recombination in photocatalysis and improve electron mobility in energy storage applications. For example, Balakrishnan Kirubasankar et al.³⁰⁷ described the creation of a heterogeneous 2D layered MoS₂/MXene (MMX) nanohybrid produced by a straightforward surfactant-assisted interstratification technique. This method works well to stop MXene and MoS₂ nanosheets from restacking on their own, which makes ions more accessible and improves the electrochemical performance (Fig. 21). The MMX electrode has a hybrid-type capacitance with a high specific capacitance of 583 F g⁻¹ at 1 A g⁻¹, a decent rate capability of 82.5%, and an outstanding cycling stability of 96.5% for 5000 cycles at 5 A g⁻¹. The device has a high specific capacitance of 153 F g⁻¹ at 1 A g⁻¹ and a great cycling stability with 90% capacitance retention over 10 000 cycles at 5 A g⁻¹. These results show that mixing MXenes with stacked TMDs has a synergistic impact that leads to high-performance electrochemical behavior for enhanced supercapacitor applications.

Fig. 21 Electrochemical performance of the MoS₂/Ti₃C//Ni(OH)₂ asymmetric supercapacitor. (a) CV curves of the MoS₂/Ti₃C negative and Ni(OH)₂ positive electrodes at 10 mV s⁻¹, showing a potential window of 1.6 V. (b) CV curves of the assembled device at different scan rates. (c) GCD curves at different current densities. (d) Ragone plot comparing energy and power densities with other reported devices (inset: lighting of an LED). (e) Cycling stability and coulombic efficiency over 10 000 cycles at 5 A g⁻¹. (f) Nyquist plots before and after cycling (inset: equivalent circuit model); the figure has been reproduced from ref. 307 with permission from Elsevier, Copyright 2020.

4.2.1.5 MXene/metal clusters (or others). MXene/metal cluster composites are a flexible type of hybrid nanomaterial that merges the unique capabilities of 2D MXenes with the catalytic, optical, and electrical properties of metal clusters. MXenes like Ti₃C₂, Nb₂C, or V₂C have a lot of surface terminations (–OH, –O, and –F) that make it easier to anchor and stabilize metal clusters. They also have a lot of surface area and strong electrical conductivity. Metal clusters, which are usually noble metals like Pd, Au, Pt, and Ag and transition metals like Ni, Co, and Cu, have quantum-size effects, high surface energy, and many active sites. This makes them great for catalysis, sensing, and energy applications. For instance, Tehseen Nawaz et al.³⁰⁸ synthesized an Au₂Cu₂ nanocluster-based Mo₂TiC₂ MXene composite for lowering the restacking effects. The incorporation of a Au₂Cu₂ nanocluster into Mo₂TiC₂ MXene sheets reduced the restacking effect. The assembled SSC manages to achieve a 0–1 V potential range with a high energy density of 13 W kg⁻¹ at a power density of 501 Wh kg⁻¹. In recent days, MOFs with MXenes have emerged as an effective strategy to design high-performance hybrid electrode materials for energy storage devices. The incorporation of MXene nanosheets into MOF frameworks generates a hybrid architecture (MXene/Ni–Co MOF, ZIF-8/Ti₃C₂T_x, and MOF-801@MXene), which improves the electrochemical kinetics by suppressing the restacking of MXene layers.³⁰⁹ Subsequently, MOFs and MOF-based gels have recently attracted significant attention in electrochemical energy storage due to their tunable porosity, high surface area, mechanical flexibility, processability, and structural versatility. Furthermore, MOF-based hydrogels and gel composites combine the intrinsic porosity of MOFs with the aforementioned flexibility and interconnected network of gel matrices. This improves ion transport and mechanical stability, which makes it a promising electrode material for supercapacitor devices.³¹⁰ For instance, Yaxiong Ji et al.³¹¹ designed a bimetallic organic framework supported with an MXene (Ni₁Co₁-MOF@MXene) as an electrode material for supercapacitor applications. The results of electrochemical testing demonstrated that Co- and Ni-based composites exhibited a high specific capacitance of 1493.6 F g⁻¹ at a current density of 1 A g⁻¹. The assembled solid-state flexible supercapacitor achieved 73.9 Wh kg⁻¹ energy density at 750 W kg⁻¹ power density. The strong contacts between the metal and the support stop agglomeration and keep the surface-to-volume ratio high, which is important for better catalytic activity.

4.3 External field-assisted engineering

4.3.1 Radiation-induced. Radiation-assisted methods have become effective ways to create defects and produce MXenes utilizing energy sources such as electron beams, X-rays, gamma rays, and ion irradiation. Ion-beam irradiation allows for the selective removal of A-layer atoms (Al, Si, and Ga) from the MAX structure.³¹² This generates holes analogous to those produced by chemical etching, although devoid of caustic acids. Researchers can control etching rates, defect density, sheet thickness, and lateral dimensions by carefully changing the ion fluence and energy. This method also allows you to generate MXene in patterns that are better suited for device integration, while reducing contamination from external ions that typically occur during chemical etching. Gamma irradiation and high-energy electron beams are two more types of radiations that can be used to treat MXenes. These types of radiations generate solvated electrons and reactive radicals in solutions, which break down precursors, and the A-layer bonds weaken. Localized electron irradiation under a transmission electron microscope can directly turn MAX phases into ultrathin MXene sheets.^313–315 Exposing MXene precursors to gamma rays in liquid media can also cause exfoliation, surface functionalization, and defect creation at the same time. This makes the nanosheets more hydrophilic with better catalytic performance. One unique benefit of radiation-based synthesis is that it can combine etching with fault engineering. Irradiation can intentionally create vacancies at locations of metal, carbon, or nitrogen, which changes the electronic band structure and makes catalytic activity stronger. Moreover, the radiation environment can sustain a typical surface termination (–O, –N, or –S), which is hard to get through chemical means. This makes MXenes more useful for electrocatalysis, energy storage, and sensing. Radiation-driven technologies also offer the possibility for scalable and eco-friendly MXene manufacturing from a practical perspective. It is possible to use large irradiation platforms, such as gamma-ray facilities and electron accelerators, for continuous synthesis without using dangerous fluoride-based etchants. This not only enhances the purity of the materials but also reduces their environmental impact.^316,317 However, obstacles including high energy needs, limited access to irradiation equipment, and the difficulty of attaining homogeneous large-scale manufacturing are still major problems that need to be solved.

4.3.2 Magnetic field-assisted. Magnetic field-assisted synthesis uses the inherent or introduced magnetic properties of MXenes to regulate the alignment and aggregation of nanosheets during exfoliation.³¹⁸ If MXenes contain magnetic transition metals such as Cr, V, Ti, and Mn, static or alternating magnetic fields can be employed to align the flakes along the field lines. This alignment reduces the likelihood of material restacking and facilitates the movement of ions and electrons in divergent directions. This is good for energy storage devices like supercapacitors and lithium-ion batteries. Researchers may fine-tune interlayer spacings, improve surface accessibility, and build hierarchical designs without changing the strength, direction, and duration of the field. Magnetic fields influence the production and development of MXenes during solvothermal or etching-assisted synthesis processes. The Lorentz force that acts on charged particles can make ions move more easily, leading to even etching of the A-layer and faster peeling of MAX precursors. Additionally, magnetic fields can interact with paramagnetic ions or radicals that are generated during chemical etching. This stabilizes intermediate structures and stops defects from gathering. Magnetic fields assist not only in structural control but also in surface functionalization and the creation of composites.³¹⁹ When MXene sheets are formed or subjected to hydrothermal treatments in the presence of a magnetic field, they exhibit preferential orientations that facilitate the incorporation of magnetic nanoparticles or heteroatom dopants into the material. This method allows the generation of MXene hybrids that exhibit magnetic responsiveness applicable in EMI shielding, magneto-electrocatalysis, and spintronic devices. Alignment in magnetic fields also promotes anisotropic stacking and pore development, making it easier for ions to move around and for electrolytes to get into energy storage devices. Nevertheless, challenges remain in optimizing field parameters like scaling the process and integrating it with conventional etching techniques.^320–322 Effective utilization of nanosheets often necessitates high-strength fields, which may pose challenges for large-scale production.

5. Recent research progress in enhancing the supercapacitor performance of MXenes

5.1 Recent work on substitutional doping

Substitutional doping, also known as elemental or lattice substitution is a potent way to change the basic properties of MXenes with the formula M_n+1X_nT_x. Substitutional doping alters the MXene lattice directly, whereas surface functionalization introduces terminal groups such as –O, –F, and –OH to the surface. In this method, host atoms are swapped out at either the transition-metal site (M-site) or the non-metal site (X-site).³²³ Substituting atoms at the lattice level changes the material's overall crystal structure and electrical structure. Specifically, doping at the M-site involves substituting transition metals like Ti, Mo, V, and Nb with other metals. This alters the material's conductivity, magnetic characteristics, and catalytic performance of MXenes.^324,325 Uzair Ahmed et al.³²⁶ reported an HF-free in situ exfoliation strategy and doped Ta in the Nb₂CT_x sample (Fig. 22(a and b)). Structural analysis indicated that Ta doping significantly improved the exfoliation of Nb₂CT_x and increased its interlayer spacing. This was evidenced by the shift of the (002) diffraction peak toward lower angles, with the interlayer distance expanding to approximately 40.54 Å in the 6 wt% Ta-Nb₂CT_x sample. The SEM images further showed the formation of well-separated layered nanosheets, resulting in a larger accessible surface area (∼42.26 m² g⁻¹) and wider interlayer channels that enable more effective electrolyte penetration and faster ion transport. Moreover, the incorporation of Ta atoms into the Nb₂CT_x lattice modified the electronic structure, leading to enhanced electrical conductivity and the construction of additional redox-active sites. The presence of favorable surface terminations (O⁻/OH⁻) also contributed to improved electrochemical behavior by facilitating pseudocapacitive reactions at the electrode–electrolyte interface. Notably, the 6 wt% Ta-Nb₂CT_x sample delivered a high specific capacity of 454.4C g⁻¹ along with excellent cycling durability, retaining 88% of its initial performance after 12 000 cyclic voltammetry cycles. Furthermore, the ASC assembled using 6 wt% Ta-Nb₂CT_x, and AC attained a high energy density of 47.5 Wh kg⁻¹ at a power density of 500 W kg⁻¹. Similarly, Uzair Ahmed et al.³²⁷ doped Ce in Nb₂CT_x by following the same procedure (Fig. 22(c–e)). The incorporation of cerium into the Nb₂CT_x MXene significantly expanded the c-lattice limitation to approximately 29 Å for the 6% Ce-Nb₂CT_x composition, leading to an enlarged interlayer spacing and an increased specific surface area of 36.34 m² g⁻¹, offering more electroactive sites and enabling easier electrolyte ion intercalation within the layered structure. The specific capacitance for 6% Ce-Nb₂CT_x at a scan rate of 1 mV s⁻¹ further reached 1073 F g⁻¹ at a current density of 1 A g⁻¹. The material also demonstrated excellent long-term cycling stability with 86% retention of its initial capacitance after 10 000 charge–discharge cycles. In addition, high energy and power densities of 45 Wh kg⁻¹ and 275 W kg⁻¹ were reported. The recent works on the substitutional doping in MXene sheet-based supercapacitors are given in Table 4.

Fig. 22 (a) FE-SEM images of pristine Nb₂CT_x and doped Nb₂CT_x MXenes at different magnifications. (b) Electrochemical performance of 6 wt% Ta-Nb₂CT_x compared with activated carbon (AC): CV curves at different scan rates, GCD profiles at different current densities, specific capacity as a function of current density, cycling stability, and Ragone plot illustrating energy and power densities; the figure has been reproduced from ref. 327 with permission from Elsevier, Copyright 2025. (c) FE-SEM images of Ce-doped Nb₂CT_x MXenes with different Ce contents. (d) Nitrogen adsorption–desorption isotherms and BET surface area analysis of theNb₂CT_x and Ce-Nb₂CT_x samples (e) GCD curves and corresponding specific capacitance comparison of Nb₂CT_x and Ce-Nb₂CT_x electrodes at different doping levels and current densities, inset shows CV and GCD plots of initial and 10 000^th cycle; the figure has been reproduced from ref. 327 with permission from Elsevier, Copyright 2025.

Table 4 Comparative electrochemical performance of substitutionally doped and composite MXene-based materials synthesized via various approaches

Catalyst	Synthesis method	Surface area (m^2 g^−1)	Specific capacitance	Cyclic stability (cycling number, current density)	Specific power and specific energy	References
Ta–Nb2CTx	In situ exfoliation approach	—	826.2 F g^−1 at 5 mV s^−1	88% retention after 1200 CV cycles	500 W kg^−1 and 47.5 Wh kg^−1	326
Ce-Nb2CTx	Nonhazardous in situ exfoliation	36.34 m^2 g^−1	1073 F g^−1 at 1 A g^−1	86% retention after 10 000 cycles at 1 A g^−1	275 W kg^−1 and 45 Wh kg^−1	327
Nb-doped Mo2TiC2Tx MXene	Selective etching of aluminium layers with HF	—	398 F cm^−3	100% capacitance retention after 6000 cycles at 10 A g^−1	82.6 kW L^−1 and 48.1 Wh L^−1	328
V1.8Nb0.2AlC	Self-propagation high-temperature synthesis with HF	—	1698 F cm^−3	86% after 8000 cycles at 10 A g^−1	—	329
Nb2CTx MXene	Delamination through HF treatment	6.611 m^2 g^−1	307 F g^−1 at 1 A g^−1	90% capacitance retention after 4000 cycles at 2 A g^−1	1.41 kW kg^−1 and 33.2 Wh kg^−1	330
VMNiO@MXene	Co-precipitation approach	—	1009 F g^−1 at 0.5 A g^−1	86.3% capacitance retention after 5000 cycles of CV at 100 mV s^−1	—	331

---

5.2 Recent work on vacancy and defect engineering

Vacancy and defect engineering has been an effective method for improving the electrochemical performance of MXenes in supercapacitors. This is because it increases the number of active redox/adsorption sites, making it easier for ions to diffuse, and improves electronic conductivity. This strategy affects the attainment of equilibrium between structural stability and electrochemical activity. Recent research reveals three primary insights: (1) controlling the creation of vacancies greatly improves capacitance and rate performance; (2) precursor/MAX phase design or gentle post-treatment methods are better for getting consistent control than harsh etching; and (3) combining advanced characterization with computational modeling is very important for linking vacancy type and density to device performance.^332,333 A synergistic methodology integrating experimental characterisation and theoretical modeling is essential. High-resolution STEM/HAADF allows analysing metal and C/N vacancies directly, while XPS and EPR provide information about surface terminations and paramagnetic defect centers. Impedance spectroscopy, GCD, and CV are all examples of electrochemical instruments that can help tell if the performance increases are due to surface reactions or diffusion processes. On the theoretical side, calculations of vacancy formation energy are becoming more common as predictive tools. They can help scientists figure out which MXenes can hold stable vacancies and how these defects affect electronic properties and ion interactions. This makes it possible to do effective pre-screening of materials before testing them in the lab.³³⁴

The objective for MXene-based supercapacitors is to maintain a balanced vacancy density. This can be achieved by meticulously modifying the precursors or gradually lowering the temperature or chemical concentrations instead of employing aggressive techniques such as irradiation or excessive etching which may compromise the stability of the structure. Employing vacancy engineering in conjunction with techniques such as incorporating interlayer spacers or integrating pseudocapacitive components (oxides or metal–organic frameworks) can enhance structural integrity while augmenting the availability of electroactive sites. There are still some big problems that need to be solved. For instance, there are no standardized methods to quantify the vacancy density, and the defects can accelerate oxidation and solubility in water. Recent reviews and experimental reports have given us methodological guidelines, and the expanding use of “vacancy formation energy” calculations is a promising way to design and improve materials in a logical way.³³⁵

5.3 Recent work on MXene-based composite materials

MXenes are a rapidly expanding category of 2D transition-metal carbides, nitrides, and carbonitrides that have garnered significant interest due to their metallic conductivity, hydrophilic surfaces, adjustable terminations, and exceptional mechanical strength.³³⁶ Since they were found in 2011, they have been studied in many areas such as energy storage, catalysis, biomedicine, and environmental uses. Pristine MXenes have problems such as layer restacking being prone to oxidation, and not lasting very long, which makes them less useful right away. To address these challenges, researchers have increasingly concentrated on the engineering of MXene-based composites by integrating them with polymers, metals, metal oxides, carbon nanostructures, and even biomaterials.³³⁷ Recent advancements indicate that these composites significantly improve electrochemical energy storage in supercapacitors, lithium/sodium-ion batteries, and flexible electronics. MXene–metal oxide/sulfide hybrids work very well in catalysis for hydrogen evolution, oxygen reduction, and water splitting. Incorporating MXenes into polymer matrices has enabled the development of multifunctional materials with high electrical conductivity, mechanical strength, and suitability for EMI shielding, flexible electronics, and wearable devices³³⁸ (Fig. 23). The quick progress of research demonstrates the versatility and potential of MXene-based composites in addressing significant challenges across various domains. To maximize their potential for next-generation technologies, synthesis, structural engineering, and surface modification must continue to improve.

Fig. 23 Schematic of the MXene-based composite hybrid structures and their representative architectures: (a) MXenes/metal oxides; the figure has been reproduced from ref. 339 with permission from Elsevier, Copyright 2025. (b) MXenes/TMDs; the figure has been reproduced from ref. 340 under a Creative Commons license, Springer Nature, Copyright 2022. (c) MXenes/polymers; the figure has been reproduced from ref. 341 with permission from the American Chemical Society, Copyright 2025. (d) MXenes/CNT/r-GO; the figure has been reproduced from ref. 342 under a Creative Commons license, Springer Nature, Copyright 2020.

Fig. 24 (a and c) Top and (b and d) side views of Mo₂B MBene H-type and T-type, (e) and (f) band structure of H-type and T-type Mo₂B. (g) Relationships between the total energy and the lattice parameter in the basal plane. (h) and (i) Density of states for H-type and T-type Mo₂B MBene systems; the figure has been reproduced from ref. 197 under a Creative Commons license, the Royal Society of Chemistry, Copyright 2020.

5.3.1 MXenes/metal oxides. Recent advancements in MXene/metal oxide composites have significantly enhanced supercapacitor performance by utilizing the high conductivity and surface functionality of MXenes in conjunction with the considerable capacitance and redox activity of metal oxides. For example, Shinde et al.³⁴³ wrote about how they made an MXene/MnCo₂O₄ nanocomposite electrode utilizing a simple and cheap method. This composite confirmed superior electrochemical performance to the individual MXene and MnCo₂O₄ electrodes. The enhancement was attributed to the cooperative interaction between MXene sheets and MnCo₂O₄ nanoparticles, which improved electrical conductivity and created more effective channels for ion transport. Gautam et al.³⁴⁴ reported the creation of a SnO₂/Na-SnO₂@MXene hybrid electrode in another important study. Adding sodium to the SnO₂ nanoparticles made a composite with a high specific capacitance and great cycling stability. At a scan rate of 5 mV s⁻¹, the Na-SnO₂@MXene electrode had a specific capacitance of 91.2 F g⁻¹ and kept about 89% of its original capacitance after 3000 cycles. This great performance was due to the combination of MXene conductive support and the sodium-doped SnO₂ NPs, which increased electrochemical activity and structural strength. Tholkappiyan Ramachandran et al.³⁴⁵ synthesized a novel Fe–SnO₂/Ti₃C₂T_x MXene composite using a hydrothermal method that exhibited a well-organized hierarchical structure. The structure contains rod-like or granular Fe-SnO₂ nanoparticles uniformly distributed on the surface of accordion Ti₃C₂T_x MXene nanosheets. The architecture exhibited a highly mesoporous structure (Type-IV) with a SSA of 220 m² g⁻¹, which can effectively promote electrolyte penetration and minimize ion diffusion resistance. Morphological analysis indicates that the strong interfacial interaction between crystalline SnO₂ NPs and the MXene layers helps preserve a conductive layered framework, which also suppresses MXene restacking during electrochemical cycling. Electrochemical measurements performed in 1 M KOH revealed that the composite demonstrates excellent charge storage performance, delivering 1225.6 F g⁻¹ at 1 A g⁻¹, thereby maintaining 277.2 F g⁻¹ at a high current density of 30 A g⁻¹. The pseudocapacitive behaviour arises from the reversible redox reaction of tetragonal rutile SnO₂ and bcc metallic Fe phase. The Fe-SnO₂/Ti₃C₂T_x MXene composite maintains 92.7% capacitance retention after 10 000 cycles. Ansari et al.³⁴⁶ also did a thorough assessment of the advances made in MXene/transition-metal oxide heterostructures for use in supercapacitors. The writers said that MXenes have great conductivity and a huge surface area, but their electrochemical performance can be hampered by problems like restacking and surface oxidation. Subsequently, transition-metal oxides have a lot of capacitance, usually with low conductivity. It was suggested that making heterostructures out of MXenes and transition-metal oxides would be a good way to deal with these problems. The review went into great depth about different ways to make these composites and stressed their benefits, such as better charge storage, better rate capability, and better cycling stability. In general, these results show that MXene/metal oxide composites have a lot of potential for improving the supercapacitor technology. The way the two components work together not only recovers electrochemical performance but also gives us a way to design energy storage devices that are more efficient and last longer.

5.3.2 MXenes/carbon nanomaterials. Recent advancements in MXene/carbon nanomaterial composites have significantly enhanced supercapacitor performance by using the synergistic features of both components. For instance, Liang and Zhitomirsky et al.²⁹⁴ made MXene–CNT composite electrodes for high active-mass asymmetric supercapacitors. Adding CNTs to MXene nanosheets made the electrodes more electrically conductive and structurally sound, which improved their electrochemical activity. The composite had a high specific capacitance and was very stable over time, which shows that MXene–CNT hybrids could be useful for enhanced energy storage applications. Huang et al.³⁴⁷ conducted a significant work detailing the manufacturing of free-standing MXene/CNT@Fe₂O₃ films for supercapacitor applications. Adding Fe₂O₃ nanoparticles to the MXene/CNT framework made it pseudocapacitive which increased the overall capacitance even further. The composite showed a high volumetric energy density and great mechanical flexibility, which makes it a good candidate for application in flexible energy storage systems. Collectively, these investigations underscore the considerable promise of MXene/carbon nanomaterial composites for advanced supercapacitors. The integration of MXenes with carbon nanomaterials such as CNTs boosts electrode conductivity and mechanical stability while also providing supplementary pseudocapacitive contributions. Nevertheless, challenges such as scalable manufacturing and long-term operational stability persist, necessitating continued research and development.

5.3.3 MXenes/TMDs. Recent studies have underscored the significant potential of MXene/transition-metal dichalcogenide (TMD) composites in enhancing the supercapacitor performance. For example, Hemanth et al.³⁴⁸ studied how to synthesize MXene/TMD heterostructures and showed that the interactions between MXenes and TMDs work together to greatly improve the electrochemical performance. Adding TMDs to MXenes endowed them with better storing charge and cycling stability, making these composites good candidates for high-performance supercapacitor applications. Sikdar et al.³⁴⁹ suggested a method for making strong 3D freestanding MXene hydrogels with hierarchically porous architectures. These MXene-based hydrogels exhibited improved electrochemical performance when mixed with TMDs, which can be attributed to their larger surface area and more effective ion transport channels. Adding TMDs to MXene hydrogels is a promising way to make high-performance supercapacitors with better energy and power densities. The recent studies reported in the literature that suggest MXenes and TMDs for supercapacitor applications are listed in Table 5.

Table 5 Electrochemical performance of MXene–metal oxide and MXene–TMD composite materials synthesized via different routes

Catalyst	Synthesis method	Surface area (m^2 g^−1)	Specific capacitance	Cyclic stability (cycling number, current density)	Specific power and specific energy	Ref.
MXene in CeO2–La2O3	Microwave-assisted method	79.3 m^2 g^−1	1254 F g^−1 at 2 A g^−1	77% capacitance retention after 25 000 cycles at 5 A g^−1	3000 W kg^−1 and 31.33 Wh kg^−1	350
MXene copper selenide (MCuSe)	Hydrothermal method	5.4 m^2 g^−1	570 F g^−1 at 1 A g^−1	98% even after 1000 cycles 5 A g^−1	2055.46 W kg^−1 and 59.38 Wh kg^−1	351
MoS2/Ti3C2 MXene	Hydrothermal method	22.510 m^2 g^−1	46.6 F g^−1 at 1 A g^−1	88% after 5000 cycles at 1 A g^−1	—	352
MXene/WS2	Chemical and hydrothermal methods	—	373 F g^−1 at 0.4 A g^−1	—	—	353
WS2/MoO3/MXene	Hydrothermal and ultrasonication routes	—	1241 F g^−1 at 1 A g^−1	95.23% after 5000 cycles at 1.8 A g^−1	—	354

---

5.4 Comparison of the supercapacitance performances of various 2D materials with those of functionalized MXenes

Recent studies have methodically evaluated the supercapacitive performance of many 2D materials, focusing specifically on functionalized MXenes. Their unique structural characteristics and electrochemical capabilities have garnered significant interest, establishing them as formidable contenders for next-generation energy storage systems. Gautam et al.³⁴⁴ examined SnO₂/Na–SnO₂@MXene hybrid electrodes, as reported in the New Journal of Chemistry. Their research showed that adding sodium to SnO₂ nanoparticles in the MXene framework greatly increases their electrochemical performance reaching a specific capacitance of 91.2 F g⁻¹ at a scan rate of 5 mV s⁻¹. The hybrid also has great cycling stability, keeping about 89% of its original capacitance after 3000 cycles. The comparison of the supercapacitor performance of varying 2D materials with functionalized MXenes is illustrated in Table 6. The data show that functionalized MXenes, especially when used in hybrid configurations, are a viable type of electrode material for next-generation supercapacitors. They have a good balance of high power density, energy density, and long-term operating stability.

Table 6 Electrochemical properties of the MXene-based hybrid and composite electrode materials synthesized through various fabrication routes

Catalyst	Synthesis method	Surface area (m^2 g^−1)	Specific capacitance	Cyclic stability (cycling number, current density)	Specific power and specific energy	References
Nb2CTx/CNT//AC	Surface etching with LiF	—	200 F g^−1 at 1 A g^−1	73% of cyclic stability after 2000 at 10 A g^−1	0.38 kW kg^−1 and 16.5 Wh kg^−1	355
NiCo2O4@Nb2CTx//AC	Hydrothermal-calcination coupling method	75.0 m^2 g^−1	1873 F g^−1 at 1 A g^−1	97.1% capacitance retention after 5000 cycles	4.1 kW kg^−1 and 22.7 Wh kg^−1	356
MXene/graphene oxide nanocomposite film	CO2 laser irradiation	—	662.9 mF cm^−2 at a current density of 5 mA cm^−2	Maintained 40 000 cycles at 25 mA cm^−2	—	357
MXene/GO/LDH	Etching by LiF/HCl	140.66 m^2 g^−1	241.9 mA h g^−1 at 1 A g^−1	∼94.9% after 4000 cycles at 5 A g^−1	800 W kg^−1 and 55.3 Wh kg^−1	358
MXene-graphene hydrogel	Layer-by-layer self-assembly	55 m^2 g^−1	653.7 F g^−1 at 2 mV s^−1	98% capacitance retention after 8000 cycles at 3 A g^−1	10.4 kW kg^−1 and 4.65 Wh kg^−1	359

---

6. Theoretical and computational studies

MXene is superior to graphene owing to its capacity to control the properties of the material. This can be completed using processing, doping, and functionalization, while graphene allows only functionalization. This section discusses the properties and performance of MXenes when modified with various techniques like doping, vacancy, and functionalisation. Computational studies made this exploration at a cheaper cost than the experiments to be made on each modification. Most of the computational studies discussed in the subsections below are based on DFT in combination with a solvation model. It is possible to investigate the wonder of electrode–electrolyte interaction at the electrode surface and its significance on the electrochemical properties.³⁶⁰ The following sub-sections focus on different ways to modify the structure of MXenes and get the best out of them. Before delving into the modification studies of MXenes, one must understand how capacitance is dependent on the density of situations, which changes with the modification of the structure or the system. The total capacitance in low-dimensional materials is determined not only by the electrochemical double layer but also by the electronic structure of the electrode itself. This intrinsic contribution is called quantum capacitance (C_Q), which is given as follows:where e is the charge, D(E) is the density of states at the energy level E, Φ is the local potential derived from the electrochemical potential µ = eΦ, and F_T is the thermal broadening function represented as follows:where k_B is the Boltzmann constant, and usually the temperature T is set at 300 K.³⁶¹ Basically, C_Q is the measure of how many electronic states are available at the Fermi level to accommodate additional charges. The reason why traditional metallic electrodes do not consider C_Q is that D(E_F) is extremely large, and it has less effect on surface terminations or any other adsorbates, and the total capacitance relationship is given as follows:

Due to the large value of D(E_F) in the bulk materials, the reciprocal of the quantum capacitance term gets eliminated, and the total capacitance is mainly determined by the electrolyte (double-layer). However, in 2D materials like MXenes, where D(E_F) varies a lot with composition, surface termination, and layer thickness. C_Q plays a key function in determining the charge storage capability of the material. Since MXenes in their pristine state exhibit high C_Q owing to their metallic nature, –O and –F termination lowers the DOS at E_F, reducing C_Q. In contrast, –OH terminations in some cases enhance C_Q due to near-free-electron states that provide additional surface charge accommodation.^362,363

6.1 Modification of MXenes through doping

Heteroatom doping means deliberately replacing some atoms in the MXene structure with different elements to change its properties. It is a common and effective method to improve the material performance. Researchers mainly use DFT calculations to study this process and predict how adding different dopants will affect the material's stability and electronic behavior before doing any complex experiments. Doping changes the electronic properties by modifying the band structure and the DOS around the Fermi level (E_F), as shown in Fig. 26(a). When a dopant atom has a different number of valence electrons than the atom it replaces, it can either add or remove electrons from the system, shifting the Fermi level and creating new energy states. DFT studies revealed that the effect of doping is highly site-specific in MXenes and many other 2D materials. Heteroatom nitrogen doping at different sites of Ti₃C₂O₂ MXene (Fig. 25(a)) reveals that the highest capacitance is achieved when doping occurs at functional sites (FS), where nitrogen replaces oxygen, rather than at lattice sites (LS). This is attributed to the increased electron count, which shifts the band structure downward and enhances the density of states (DOS) near the E_F. Higher DOS near E_F enhances the quantum capacitance (C_Q). It was found that the increase in capacitance resulting from nitrogen doping was primarily due to enhanced pseudocapacitance. The improvement was attributed to increased surface redox activity induced by nitrogen atoms. In contrast, the substituted systems exhibited degraded performance compared to the pristine material. These results suggested that better charge storage capacity in the Ti₃C₂O₂ electrode could be achieved through doping rather than substitution. The FS-doped Ti₃C₂O₂ exhibited an approximately fourfold increase in integrated quantum capacitance (C_Qint) compared to pristine Ti₃C₂O₂ in the −1 V range, while the LS-doped system showed a smaller enhancement. Fig. 26(b) shows the C_Qint of the pristine and doped Ti₃C₂O₂ system.³⁶⁰ A similar trend was also observed for V₂CO₂.³⁶⁴ With Nb doping in the Ti₃C₂ system, it was found to increase the DOS again. The introduction of Nb atoms modified the electron density and strengthened the bonding within the structure. This effect arises because Nb possesses a filled 3d shell and a partially filled 4d shell, whereas Ti has only three electrons in its 3d shell. As a result, the specific capacitance increased significantly from 124 F g⁻¹ for pristine Ti₃C₂ to 442.7 F g⁻¹ for Nb-doped Ti₃C₂.³⁶⁵

Fig. 25 Schematic of (a) N-doped Ti₃C₂O₂ system along with defects; the figure has been reproduced from ref. 366 with permission from John Wiley and Sons, Copyright 2023. (b) Structure of defect-free Ti₂CO₂ along with Ti (grey), C (black), and O (red) vacancies; the figure has been reproduced from ref. 367 with permission from the American Physical Society, Copyright 2022. (c) N-Doped graphene/V₂C hybrid system; the figure has been reproduced from ref. 368 with permission from the Royal Society of Chemistry, Copyright 2018.

Fig. 26 (a) Density of states (DOS) of pristine and N-doped Ti₃C₂O₂ systems; the figure has been reproduced from ref. 366 with permission from John Wiley and Sons, Copyright 2023. (b) Integrated quantum capacitance of pristine and doped Ti₃C₂O₂; the figure has been reproduced from ref. 360 with permission from IOP Publishing, Copyright 2022. (c) DOS of pristine and various vacancy-defect Ti₂CO₂ monolayers; the figure has been reproduced from ref. 369 under a Creative Commons license, the Royal Society of Chemistry, Copyright 2019. (d) DOS of N-doped graphene and N-doped graphene/V₂C MXene hybrid systems; the figure has been reproduced from ref. 368 with permission from the Royal Society of Chemistry, Copyright 2018.

6.2 Modification of MXenes through vacancies

Every material, when synthesized, comes with atomic-scale imperfections where an atom is missing from its lattice position. Far from being intrinsic defects, these flaws can be strategically designed to intentionally modify the structural, electronic, and magnetic properties of the materials. These modifications directly reconfigure the material's charge distribution and ionic landscapes, offering a pathway to transcend the intrinsic limitations of pristine MXenes for energy storage applications. Defect engineering, particularly the intentional creation or control of atomic vacancies, represents another sophisticated strategy for tuning the properties of MXenes. DFT calculations are essential for understanding the energetics and electronic impact of these point defects. The vacancy formation energy (E_f) is a key parameter calculated via DFT that determines the energy cost to remove a specific atom from the lattice, and thus, predicts the likely concentration and type of vacancies that may form during synthesis or processing.³⁷⁰ For instance, systematic computational studies across 84 different MXene surfaces have shown that in Ti₂CO₂, carbon vacancies are energetically much more favorable to form than titanium or oxygen vacancies.³⁶⁹ These calculations also reveal the profound influence of surface terminations on defect stability; O-termination significantly increases the energy required to form a metal vacancy, thus stabilizing the lattice, whereas F-termination can lower it.³⁶⁷ Vacancies are of different types in MXenes, first M-type vacancies, where a transition metal commonly forms during liquid etching, and X-type, where these are vacancies of the carbon or nitrogen atoms that make up the inner “X” layer. Carbon vacancy defects may occur in situ arising from inherent carbon deficiencies in the parent MAX phase, as T-type vacancies associated with terminal groups (oxygen), or as double vacancies involving multiple vacancy types within the system.³⁶⁷Fig. 25(b) shows the structure of defect-free Ti₂CO₂, along with Ti(grey), C(black), and O(red) vacancies, and their corresponding density of states shown in Fig. 26(c).

The formation of Ti vacancies is highly suppressed on O-functionalized surfaces, with a formation energy significantly higher (∼5 eV) than that of the other surfaces. Conversely, Ti vacancy formation is prevalent and energetically favorable on pristine or OH-functionalized surfaces. This suggests that Ti vacancy formation is a crucial first step in the oxidation of MXenes. The trend is the opposite for X vacancies. The formation of carbon and nitrogen vacancies is more favorable (lower formation energy) on O-functionalized surfaces and less favorable on OH-functionalized surfaces.⁸⁸ Structurally, the atoms surrounding a vacancy are typically pushed away from the vacant site. This is because the missing atom no longer provides its charge to balance (or ‘screen’) the forces around it and the remaining atoms repel each other more strongly.³⁶⁷ Intentional vacancy engineering, such as the repeated etching of V₂C, can create a disordered and porous architecture. This defect-laden structure results in a more expansive surface area, which is highly beneficial for energy storage. Carbon vacancies have been theoretically shown to increase the conductivity and flexibility of Ti₂CT₂.^370,371

In semiconducting Ti₂CO₂, a single Ti-vacancy causes a semiconductor-to-metal transition.³⁷² Similarly, a Sc vacancy in semiconducting Sc₂CF₂ also induces a semiconductor-to-metal transition. This transition is caused by the vacancy introducing localized states in the vicinity of the Fermi level (E_F).³⁷³ Not all vacancies cause metallicity, in Ti₂CO₂, carbon and oxygen vacancies retain the semiconducting behaviour but significantly reduce the band gap from 1.3 eV in pristine to 0.4 eV and 1.1 eV, respectively, for carbon and oxygen vacancies.³⁷² In V₂C MXenes, DFT calculations showed that introducing V-vacancies (from single to triple) progressively narrows the band gap, thereby enhancing the material's conductivity.³⁷⁴ Vacancies enhance quantum capacitance (C_Q) by modulating the DOS near the E_F value, as mentioned before, and C_Q depends directly on the DOS near E_F. The introduction of localized states and the resulting substantial charge redistribution around the defect site are directly linked to a higher C_Q. In Ti₃C₂, Ti-vacancy enhanced the C_Q value to 799 µF cm⁻², while a C-vacancy showed a remarkable C_Q of 824.19 µF cm⁻². Even greater enhancement was seen with substitutional defects (where one atom replaces another, creating a vacancy for the original atom). A carbon atom replacing a titanium atom yielded the highest C_Q of 934.81 µF cm⁻² at positive potentials, whereas in Sc₂CF₂, the introduction of a Sc-vacancy significantly improved the C_Q at negative bias, making both the pristine and vacancy-defected systems excellent candidates for cathode materials.¹³⁰ The primary challenge lies in achieving an optimal balance. The goal is to introduce enough vacancies to enhance the electronic and ionic conductivity, without creating an excessive number of defects that would trap ions or cause the material's structure to collapse during cycling.

6.3 Modification of MXenes due to hybridization

Creating hybrid systems or heterostructures by combining MXenes with other nanomaterials, such as graphene, CNTs, or conductive polymers, is one of the most successful experimental strategies for enhancing the charge storage performance by modifying their electronic properties. With DFT, these hybrid systems can be explained well for their enhancement in performance by exploring mechanisms at a molecular level, and sometimes this computational method offers an explanation for the experimental finding, which does not properly explain the underlying process/mechanism change with the hybrid systems, as shown in Fig. 25(c). Especially at the junction between two dissimilar materials, electronic redistribution occurs to align their Fermi levels, resulting in interfacial charge transfer. The direction and magnitude of this charge transfer, which can be precisely calculated using DFT, fundamentally alter the electronic properties of both components.^14,24 A prime example is the hybridization of Ti₃C₂T_x with graphene (M/G).

DFT calculations show that this passivation forms strong covalent Ti–C bonds at the interface, which not only enhances the thermodynamic and mechanical stability of the MXene but also fundamentally alters its electronic properties. This hybridization facilitates a charge transfer, with electrons moving from the Ti₃C₂ MXene layer to the graphene layer. This charge redistribution significantly enhances the DOS at the Fermi level for the hybrid structure, boosting its intrinsic electrical conductivity, as shown in Fig. 26(d). This electronic modification has a direct, positive impact on charge storage as the quantum capacitance (C_Q) depends on the DOS at the Fermi level. DFT calculations show that the theoretical C_Q value of the defect-free M/G hybrid is 1305.76 µF cm⁻² at zero potential, which is substantially higher than that of pristine MXene (1171.63 µF cm⁻²) or graphene alone (0.91 µF cm⁻²).⁹³

Computational models also reveal that defect engineering at this hybrid interface is a critical factor for optimization, and it allows for different combinations of defects on MXenes, other 2D materials or both. For instance, N-doping in graphene introduces a new, high-density peak in the DOS spectrum near the Fermi level, which results in a maximum calculated C_Q value of 1594.22 µF cm⁻² (at −0.07 V) in the MXene/graphene hybrid structure. This value surpasses that of both the pristine MXene and the defect-free hybrid, demonstrating that hybridization creates new pathways for tuning the electronic properties.⁹³ Similar electronic enhancements are predicted in the theoretical studies of Ti₃C₂T_x hybridized with polyaniline (h-MXene/PANI). The DFT calculations for this system identify a large difference in the work functions of MXene (6.25 eV) and PANI (3.79 eV). This disparity drives a spontaneous flow of electrons from PANI to the MXene, creating a built-in electric field at the heterojunction interface. This charge redistribution, as in the graphene hybrid, results in a h-MXene/PANI hybrid possessing a significantly higher DOS at the Fermi level, which in turn creates a wider electroactive region that promotes quicker and easier electron transfer during electrochemical processes.^375,376

Besides modifying the electronic structure for better capacitance, computational studies show that hybridization fundamentally improves the ion kinetics. For the h-MXene/PANI system, DFT calculations of the diffusion barrier energy revealed that the energy required for a proton (H⁺) to move through the material is significantly lowered to 2.44 eV, a substantial reduction from the 3.29 eV calculated for the pristine MXene. This substantially predicts that ions can migrate much more easily in the hybrid, leading to faster kinetics. This mechanism is not unique to PANI.³⁷⁶ The theoretical modeling of Ti₃C₂T_x/g-C₃N₄ heterostructures also shows that g-C₃N₄ acts as an electron donor, transferring charge to the MXene nanosheet and introducing new local states near the Fermi level, which is predicted to enhance the quantum capacitance.^377,378 In another computational study on a Ti₃C₂T_x/TiO₂ hybrid, DFT combined with the climbing-image Nudged Elastic Band (ci-NEB) method was used to understand the underlying kinetics. The calculations, performed on simplified Ti₃C₂, revealed an ultra-low energy barrier of just 0.037 eV for Li-ion migration on the MXene surface. This theoretical finding explains the mechanism of the hybrid's high performance. The TiO₂ nanoparticles function as crucial spacers that prevent the MXene nanosheets from restacking, thereby ensuring that the electrolyte ions can access the MXene surface and take full advantage of the intrinsically fast, low-barrier migration pathways that DFT calculations confirmed.^379–381 Among the diverse electronic modulation strategies for enhancing the supercapacitive performance of MXenes, each approach presents unique advantages accompanied by important practical considerations. For instance, heteroatom doping emerges as a relatively straightforward and cost-effective approach owing to precise tuning of the band structure and DOS near the Fermi level. This enhances quantum capacitance and surface redox activity without drastically altering the structural framework. DFT studies consistently demonstrate that appropriate site-selective doping significantly increases the electronic conductivity and capacitance, making doping highly attractive for scalable supercapacitor design. Moreover, vacancy engineering offers another powerful pathway by introducing localized states near the Fermi level, improving charge redistribution, and narrowing band gaps. However, careful control of vacancy concentrations is essential to avoid structural instability and excessive defect trapping that compromises long-term cycling performance. Hybridization provides the most substantial enhancement in both quantum capacitance and ion-transport kinetics due to synergistic interfacial charge transfer and restacking suppression. Nevertheless, hybrid systems often involve additional synthesis complexity and higher material costs. Collectively, heteroatom doping can be considered the most practical and economically viable pathway for designing high-performance MXene-based supercapacitors. Vacancy engineering and hybridization require precise defect control and involve additional synthesis complexity and higher material costs. Heteroatom doping also showed considerable performance improvement, cost-effectiveness, and feasibility for large-scale electrode fabrication.

7. Machine learning for predicting and designing MXene structures for charge storage

The chemical space for MXenes is exceptionally vast, defined by a combinatorial explosion of M, X, and T_x. Additionally, there are ordered double transition metals, solid solutions, doped systems, hybrid systems, defect-engineered systems, and mixed terminations, which make traditional experimental or first-principles (DFT) screening approaches intractable. Although DFT calculations provide almost accurate results, they are computationally expensive and too slow to traverse the entirety of the enormous MXene space. Machine learning (ML) has emerged as a transformative tool to navigate this challenge, enabling the high-throughput screening, prediction, and rational design of new MXene materials for specific applications, including charge storage.^382–386 By training on data from DFT calculations or physical experiments, ML models can rapidly predict material properties, identify complex structure–property relationships, and accelerate the discovery of novel MXene materials, along with doping or defect engineering, or hybridization. Fig. 27 shows the general workflow diagram from DFT results to model evolution and ML-based predictions.

Fig. 27 A conceptual diagram illustrating the workflow from MXene chemical space to DFT database construction, feature generation, ML prediction and charge storage property outputs; the figure has been reproduced from ref. 387 with permission from American Chemical Society, copyright 2024.

7.1 Various algorithms

A wide array of ML algorithms have been deployed to model MXene compositions along with their properties, ranging from classical regression models to sophisticated deep learning architectures. Among the classical supervised learning methods, which are commonly used to predict specific target properties from a curated set of features (descriptors), tree-based ensembles are prominent. Algorithms like Random Forest (RF) and Gradient Boosting (GBR) (including XGBoost) are frequently cited for their high accuracy and robustness against overfitting, especially with tabular data. RF has been successfully used to predict the Gibbs free energy of hydrogen adsorption (ΔG_H) for the HER as well as the gravimetric capacity, voltage, and induced charge for energy. Regression Trees (RT) are also used for their high interpretability.^388–390 Alongside these, Neural Networks offer powerful tools for capturing complex, non-linear relationships. Artificial Neural Networks (ANNs), including Multilayer Perceptrons (MLPs), have been extensively applied to predict the specific capacitance of supercapacitors based on experimental data. Other classical models, such as Support Vector Regression (SVR), Support Vector Machines (SVM), Multiple Linear Regression (MLR), Bayesian Ridge Regression (BRR), LASSO, K-Nearest Neighbors (KNN), and Adaptive Neuro-Fuzzy Inference System (ANFIS), are also widely used for baseline comparisons, feature selection, and prediction. However, in several comparative studies, ANNs demonstrated superior performance over MLR, SVR, and RF for predicting supercapacitor properties. Representing a significant advancement, Graph Neural Networks (GNNs) can directly take a material's atomic structure as a graph input, bypassing the need for manual feature engineering. The Atomistic Line Graph Neural Network (ALIGNN), an equivariant GNN, has been specifically used to screen MXenes for hydrogen storage. In this context, the GNN was trained to predict the hydrogen adsorption geometry (H–H bond length and distance to the metal) directly from the MXene structure, which is a key indicator of Kubas-type interaction for H₂ storage.^391,392 Finally, Active Learning (AL) is a goal-oriented strategy that iteratively guides the computational or experimental workflow by using an ML model to identify the most informative data points to calculate or measure next. This “robot-in-the-loop” approach, as demonstrated by Shrestha et al. through 8 active learning cycles with robotic sample preparation, drastically reduces the number of expensive DFT calculations or experiments needed to build an accurate ANN model.³⁹³

7.2 Training the data

Ultimately, the quality and structure of the training data decide the performance of any ML model. Training data are primarily sourced from DFT calculations by providing results like relaxation energies, band structure, bandgap, and Gibbs free energy, and collated from experimental data reported in the literature. The input to these models consists of features or descriptors that quantify the material's properties. For a supercapacitor, the key features include material properties, e.g., specific surface area, pore size, pore volume, and operational parameters, e.g., scan rate, current density, potential window, electrolyte concentration, and the mobility of cations and anions. For the HER, the primary descriptor is ΔG_H, and ML models have identified correlated elemental properties like the valence electron number and electron affinity of the termination groups, as well as the material's work function (Φ). To identify the most impactful features and reduce model complexity, methods like Recursive Feature Elimination (RFE), LASSO regression, and SHapley Additive exPlanations (SHAP) are employed. To prevent overfitting and ensure that the model generalizes to new, unseen data, rigorous validation is performed. This involves splitting datasets (often 80 : 20 train : test) and using k-fold cross-validation (typically 5-fold or 10-fold). Model performance is then assessed using standard metrics for regression tasks, which are R-squared (R²), mean absolute error, and root mean square error, while classification tasks use accuracy, precision, recall, and the receiver operating characteristic (ROC) curve.^{388–390,392,393}

7.3 Machine learning-based interatomic potentials

A transformative application of ML in materials science led to the development of Machine Learning Force Fields (MLFFs), also known as machine learning interatomic potentials (MLIPs). Standard molecular dynamics (MD) simulations are crucial for understanding dynamic processes, but they face a dilemma that classical force fields lack quantum accuracy, while ab initio MD (AIMD) is too computationally demanding for the large system sizes and long timescales needed to model electrochemical interfaces or thermal properties. MLFFs bridge this gap by training on a database of DFT-calculated forces and energies, allowing them to drive MD simulations with near-DFT accuracy at a fraction of the cost. Some methods utilize “on-the-fly” training, where the MLFF is generated dynamically during an AIMD simulation, only calling DFT when the model encounters a new configuration with high uncertainty. This approach has been successfully applied to MXenes, for instance, by Thanasarnsurapong et al., used an on-the-fly MLFF to calculate the second-order and third-order interatomic force constants needed for lattice thermal conductivity calculations in Ti₂C and Ti₃C₂. This MLFF-based method was found to be tens to thousands of times faster and cheaper than the conventional DFT method. A key limitation of standard MLFFs, however, is their assumption of a constant number of electrons (n_e), which is invalid at an electrochemical interface held at a constant external potential (U_ext).^394,395 This has spurred the development of Constant Potential MLFFs (CP-MLFFs). Wang et al. developed a CP-MLFF based on an equivariant GNN (MACE) that takes both the atomic structure and the number of electrons as input to predict the forces and the Fermi level. Furthermore, Lin et al. developed a Heteroatomic Constant Potential Method (HCPM) specifically for modeling heteroatomic MXene electrodes like Ti₃C₂(OH)₂. This model accounts for the different electronegativities and charge localizations of the Ti, C, O, and H atoms, which conventional CPM ignores. When used in MD simulations of a supercapacitor, the HCPM provided a more realistic charge distribution on the MXene and predicted increased cation attraction to the surface.^396,397

7.4 Using databases to design new materials through machine learning

The foundation of ML-driven materials discovery is the availability of large, high-quality databases. Several key databases serve as the starting point for screening. The aNANt database is a comprehensive resource focused on MXenes, containing over 23 000 structures. It is frequently used as the initial pool for high-throughput screening studies for hydrogen storage, photocatalysis, and transistors.^390,398 The C2DB (Computational 2D Materials Database) is often used as a source for transfer learning, where a model is pre-trained on its large and diverse dataset before being fine-tuned on a smaller, specific MXene dataset.^399,400 Other databases like MatHub-2d (∼1900 2DM entries)⁴⁰¹ and MXene-db (over 4000 structures) are also utilized.³⁸⁹ ML models are integrated into high-throughput screening (HTS) workflows to rapidly sift through these large databases. A common approach is a DFT-ML hybrid HTS, where DFT is used to calculate a small, random subset of a large database (e.g., 1125 of 4500 MM′XT₂ MXenes).⁴⁰² An ML model is then trained on this subset to predict a key descriptor and rapidly screen the remaining 75% of the database, identifying top candidates for final DFT validation. A more rigorous multi-stage filtering workflow, as demonstrated by Guha et al. for transistor design, involves screening the >23 000 aNANt structures through sequential filters such as PBE bandgap/magnetism (203 candidates), structural integrity/effective mass (38 candidates), and finally matching metallic phases/Schottky barriers (16 candidates), which were then passed to expensive quantum transport simulations.³⁹⁸ This concept has been advanced to Robotic HTS (Self-Driving Labs), where Shrestha et al. merged ML with a collaborative robotic platform. An automated pipetting robot prepared 264 aerogel mixtures, and an SVM classifier was trained on their structural integrity to define a “feasible parameter space,” allowing an ANN to be trained via active learning to predict the mechanical and electrical properties.³⁹³

7.5 Recent work on machine learning-based MXene design

Several recent studies have successfully applied these ML frameworks specifically to design MXenes for charge storage applications. In the area of electrochemical supercapacitors, Shariq et al. developed MLR, SVR, RF, and ANN models to predict the specific capacitance (C_sp), conductivity, and sheet resistance R_s of MXene and graphene nanoplatelet (GNP) composites based on experimental data, finding that the ANN model performed best.³⁹¹ This trained ANN was then used to predict long-term cycling stability, identifying the optimal 20 wt% GNP composite (MG-80) and forecasting its retention at 10 000 cycles. Similarly, Krishna and Mir focused specifically on Ti₃C₂ electrodes, training KNN, BRR, and ANN models on experimental data obtained from the literature to predict C_sp.⁴⁰³ While KNN achieved the highest R² value (0.928), the ANN model was analyzed with SHAP, which provided crucial explainable AI (XAI) insights, identifying cation mobility and scan rate as the most critical features influencing capacitance. Advancing the simulation front, Lin et al. used their HCPM-MD method to simulate a Ti₃C₂(OH)₂-based supercapacitor, revealing that the more realistic charge model resulted in increased Li⁺ cation attraction to the MXene surface compared to conventional CPM. In the domain of chemical storage (hydrogen), Abraham et al. performed a large-scale ML-accelerated screening of 4500 MM′XT₂ MXenes for the hydrogen evolution reaction (HER).⁴⁰⁴ An RFR model trained on a 25% DFT subset (MAE of 0.374 eV) was used to screen the rest, identifying that O-functionalized, carbon-based MXenes with Nb, V, Mo, Cr, and Ti were highly active HER catalysts. For H₂ storage, Cheng et al. used a GNN (ALIGNN) to screen 23 857 compounds from the aNANt database. The GNN predicted adsorption geometries, identifying 46 candidates, with GCMC simulations confirming the top candidate, ScYCH₂, had a high predicted gravimetric storage capacity of 5.7 wt% at 230 K.^391,397,402 Besides materials discovery, ML has been applied to optimize MXene-based energy harvesting and storage systems; for example, recent work has highlighted ML-assisted design strategies for MXene-MOF-chalcogenide hybrid triboelectric nanogenerators coupled with supercapacitors, enabling predictive optimization of material selection, device architecture, and performance in self-powered energy storage systems, where supervised learning models such as Random Forest, SVR, and XGBoost are used to correlate material descriptors (e.g., dielectric constant, surface roughness, and filler fraction) with TENG output parameters such as open-circuit voltage, transferred charge, and power density, enabling predictive optimization of TENG-supercapacitor architectures.⁴⁰⁵

7.6 Closed-loop AI-driven inverse design and autonomous experimentation in MXene research

The integration of ML and MXene research is rapidly evolving, moving from simple property prediction to autonomous materials design. The future lies in physics-informed models; the development of CP-MLFFs and HCPM-MD is a critical step toward simulating the dynamic charge storage process at the electrochemical interface, explicitly modeling the electrolyte, terminations, and applied potential simultaneously. This helps eradicate the traditional approaches to material discovery, which rely on iterative and resource-consuming steps often limited by human intuition and experimental facility limitations. Although the ML is primarily applied in the design phase, its predictions must often be retrospectively validated against experimental results to refine the model and enhance its practical relevance, as poor agreement with experimental observations, significantly limits its applicability to real-world systems.

The ultimate goal is inverse design, where a researcher specifies desired properties and the ML model generates a list of the most promising compositions (M, X, and T) to achieve them. To achieve this, models will increasingly leverage transfer learning, using knowledge from large, general databases (like C2DB) to improve accuracy on smaller, specific MXene datasets. Finally, as models become more complex (e.g., GNNs and ANNs), it is crucial to understand why they make certain predictions. The use of Explainable AI (XAI) tools like SHAP will become standard. This will allow researchers to extract scientifically interpretable design rules and fundamental physical insights from the “black box” of the ML model, truly closing the loop on accelerated, intelligent materials design. Adaptive inverse design refers to the iterative framework in which models are refined through the integration of both simulation and experimental data. In each iteration, the model proposes candidate materials that progressively converge towards the desired target properties, guided by feedback from prior evaluations. The interaction feedback loop between the design model and materials property predictions or evaluation is typically implemented through forward calculations or experiments, informed by computational simulations, empirical datasets or automated experiments.

This dynamic, closed-loop feedback is now powered by laboratory automation, which enables autonomous high-throughput experimentation and data collection within a fully automated workflow, including precursor handling, robotic synthesis, and X-ray diffraction, all controlled by a closed-loop system where experimental outcomes are continuously analysed and optimized using active learning algorithms. This approach ensures efficient coupling between physical validation and computational modelling. Recent advances further illustrate the potential of autonomous, AI-driven experimentation. For instance, Epps R. W., et al. developed an autonomous quantum dot synthesis bot, while Volk A. A., et al. introduced AlphaFlow, a self-driven fluidic laboratory that enables the autonomous discovery and optimization of multistep chemical processes using reinforcement learning. Both studies demonstrate that AI-guided, closed-loop autonomous laboratories can efficiently explore vast, high-dimensional chemical spaces, enabling the discovery and optimization of advanced nanomaterials with greater speed, precision, and reproducibility than traditional trial-and-error approaches. The experimentally validated data generated through such platforms can be integrated into existing machine learning models trained on computational datasets, thereby enabling a more robust and ultimately realizable inverse design framework.^406–408 Ultimately, this framework paves the way for self-evolving AI-driven systems capable of dynamically refining the generative models computationally, while optimising the synthesis pathways and experimental workflows in the laboratory.

7.7 Critical evaluation and challenges

ML has become a powerful tool for screening and designing MXenes for charge storage. Its current usage is accompanied by several important limitations that must be acknowledged. Most MXene-focused ML models are trained on relatively small datasets, severely limiting their ability to generalize and predict the novel properties of MXenes. This includes MXene chemistries, terminations, and defect landscapes. MXenes can differ in many ways at once (such as composition, surface chemistry, defects, structure, and processing), creating a very large and complex set of possible variations. Complex models such as deep neural networks and graph neural networks are prone to overfitting when trained on limited data. The above-mentioned common methods, such as k-fold cross-validation or 80 : 20 train : test splits, are widely used but often fail to test a model's true predictive ability on new cases. This happens because similar chemistries may appear in both training and testing data. Stronger evaluation methods include leave-one-chemical-family-out validation, time-based splits for literature data, and testing under distribution shifts (new electrolytes/synthesis methods).^409–412 In addition, uncertainty estimation (e.g., Bayesian neural networks or ensemble methods) is rarely reported owing to its importance for high-throughput screening and active learning. ML models may select poor candidates or miss promising MXenes without reliable uncertainty estimates. Many models trained on DFT data implicitly inherit the assumptions and limitations of the underlying theory, which include treating the system at absolute zero, bandgap underestimation, approximate treatment of solvation, and simplified interface models. However, when such models are implemented in practical electrochemical systems, factors including pH variations, ion-specific adsorption, interlayer water structuring, and surface reconstruction can significantly affect the system, often leading to reduced prediction accuracy. A critical bottleneck in ML-guided MXene design is experimental validation. Only a small fraction of ML-predicted candidates are synthesized and rigorously tested, limiting feedback for model refinement and hindering the identification of systematic failure modes. Even with experimental validation, discrepancies may arise from variations in synthesis conditions (e.g., etchant type, temperature, and intercalants), along with differences in flake size distribution, restacking, and surface contamination, all of which strongly impact capacitance, rate performance, and cycling stability. This makes it challenging to disentangle genuine model errors from experimental variability.^413–416 Despite these limitations, careful dataset curation, uncertainty-aware modeling, physics-informed architectures, and thorough experimental validation can enable ML to act as a reliable complementary tool that accelerates MXene discovery while working alongside traditional theoretical and experimental approaches.

8. Summary and perspectives

MXene nanosheets have developed as highly adaptable and high-performance 2D materials that demonstrate significant potential in various fields including energy storage, sensing, environmental remediation and catalysis. Their distinctive combination of metallic conductivity, adjustable surface chemistry, hydrophilicity, and mechanical durability sets them apart from conventional 2D materials like graphene or transition-metal dichalcogenides. Ti₃C₂T_x is the most extensively investigated MXene for supercapacitor electrodes that display remarkable gravimetric capacitance values of 1500 F cm⁻³ in aqueous electrolytes and great cycling stability beyond 100 000 cycles. Other MXenes like Nb₂CT_x, Mo₂CT_x, and V₂CT_x demonstrate significant electrochemical activity attributable to their diverse oxidation states and complex surface chemistry. Structural changes like interlayer expansion, heteroatom doping, and hybridization with conductive polymers have significantly improved their electrochemical performance. Despite significant advancements in synthesis, some critical obstacles hinder the large-scale implementation and long-term stability of MXene-based technology. Addressing these limits is crucial to achieving commercially viable devices that integrate high energy and power density with mechanical flexibility and environmental sustainability.

Recent technologies for surface modification and passivation have been developed to prevent oxidation. Coating MXenes with conductive polymers (polyaniline and polypyrrole), carbon materials, and metal oxides enhances the properties of MXene layers. These coatings not only inhibit oxygen penetration but also augment redox activity, thereby enhancing pseudocapacitance. A further interesting technique entails heteroatom doping (N, S, and P), which can modify the electron density and impede surface oxidation. Mitigating nanosheet restacking is essential for optimizing ion accessibility. This can be accomplished by inserting organic molecules, conductive polymers, and nanoparticles between MXene layers to form stable porous structures. The advancement of 3D MXene aerogels, foams, and hydrogels has demonstrated significant efficacy. These structures provide interconnected pathways for fast ion transport and enhanced electrolyte wettability. These designs integrate elevated volumetric capacitance with exceptional rate performance and mechanical durability. The hybridization of carbon-based materials (carbon nanotubes, graphene, and carbon cloth) with metal oxides (MnO₂ and NiCo₂O₄) produces synergistic effects that integrate the superior conductivity of MXenes with the elevated pseudocapacitance of alternative active materials. Moreover, the development of wide-potential, solid-state electrolytes represents a potential route for refining the energy density of MXene supercapacitors. Gel polymer electrolytes (PVA/H₂SO₄, PVA/LiCl, and PVA/Na₂SO₄) have been successfully combined with MXene electrodes to generate flexible and leak-proof devices appropriate for wearable electronics. Solid-state electrolytes enhance the operational voltage range and augment mechanical durability by facilitating the progress of flexible and transparent MXene-based supercapacitors.

Despite the remarkable progress in enhancing the supercapacitor performance of 2D MXenes through various strategies, several critical limitations still hinder their practical implementation. Hazardous and uncontrolled synthesis routes (HF-etching) raise concerns regarding safety, environmental impact, and termination inconsistency. Additionally, oxidative and structural instability affect long-term durability, while restacking-induced surface loss limits ion accessibility and active surface area. Challenges related to scalability and reproducibility further restrict large-scale deployment. Moreover, the limited mechanistic understanding of magnetic field-assisted charge storage and ion transport mechanisms remains a significant knowledge gap. To overcome these constraints, future research should focus on the development of fluorine-free green etching routes and controlled defect engineering strategies to enable safer and more precise synthesis. Surface functionalization, heteroatom doping, and protective hybrid coating are essential to improve structural stability and suppress oxidation. Interlayer engineering through defect-induced spacing regulation, 3D porous architectures, and pillaring approaches can effectively mitigate the restacking issues. Furthermore, machine learning-assisted screening, electrolyte–electrode interface modelling, and autonomous optimization strategies can accelerate scalable material design. In situ electrochemical characterization, multiphysics modelling, and DFT-based spin polarization studies are necessary to elucidate the fundamental mechanisms governing magnetic-field-enhanced charge storage. Collectively, integrating these targeted research pathways with existing enhancement strategies will enable the rational design of next-generation MXene-based supercapacitors with improved stability, scalability, and performance (Fig. 28).

Fig. 28 Conceptual illustration of the critical limitations and future perspectives of MXene-based materials.

Prospects include the normal design of MXene heterostructures incorporating synergistic materials like MOFs, conductive polymers, and 2D semiconductors, which will produce multifunctional electrodes with optimal ion and electron transport routes. MXenes possess mechanical flexibility and high conductivity, which render them optimal for flexible, stretchable, and transparent energy storage systems. Integrating MXenes with biocompatible polymers (chitosan, sodium alginate, PVA, and silk fibroin) will facilitate their application in implantable and wearable power systems. In addition, an extensive array of anticipated MXenes comprise elements including Mo, Nb, V, and Ta, which may provide enhanced stability, conductivity, and redox activity. The incorporation of MXene supercapacitors into hybrid energy systems that integrate batteries, photovoltaics, and piezoelectric generators would enhance their practical application in powering advanced electronics, sensors, and electric cars.

In summary, MXenes exhibit remarkable potential as electrode materials for supercapacitors, effectively bridging the gap between batteries and conventional capacitors. Their unique mixture of metallic conductivity and adjustable surface chemistry presents unmatched prospects for attaining elevated energy and power densities. Nonetheless, achieving their complete potential necessitates overcoming the current obstacles via sustainable synthesis, structural innovation, and advanced materials engineering. Future advancements will require interdisciplinary collaboration and computational experts to convert laboratory successes into scalable technologies. The incorporation of MXenes into electric vehicles, flexible electronics, and grid-scale storage systems could signify a pivotal advancement in efficient, lightweight, and sustainable energy storage technologies. Ultimately, the present investigation of MXenes and their composites will enhance our comprehension of 2D material electrochemistry and facilitate the development of advanced, environmentally friendly supercapacitor technologies that address the increasing requirements of contemporary energy systems.

Conflicts of interest

There are no conflicts to declare.

Data availability

This review does not include any primary research findings, software, or code, nor does it generate or analyze any new data. All data supporting the conclusions of this review are available in the cited published references.

Acknowledgements

Surjit Sahoo gratefully acknowledges the financial support received from the Department of Science and Technology (DST), New Delhi, India, through the DST-INSPIRE Faculty Award [DST/INSPIRE/04/2023/000200]. He also acknowledges the SEED Research Grant (SRMAP/URG/SEED/2025-26/052) provided by SRM University-AP and appreciates the university's supportive research environment.

References

1. M. Mann, S. Babinec and V. Putsche, Energy Storage Grand Challenge: Energy Storage Market Report, 2020.
2. A. Ashraf and M. Sagheer, Environ. Prog. Sustainable Energy, 2025, e70071.
3. A. Chmielewski, J. Kupecki, Ł. Szabłowski, K. J. Fijałkowski, J. Zawieska, K. Bogdziński, O. Kulik and T. Adamczewski, Currently Available and Future Methods of Energy Storage, WWF Poland, 2020.
4. M. Tomy, A. Ambika Rajappan, V. Vm and X. Thankappan Suryabai, Energy Fuels, 2021, 35, 19881–19900.
5. S. Ghosh, S. K. Behera, A. Mishra, C. S. Casari and K. K. Ostrikov, Energy Fuels, 2023, 37, 17836–17862.
6. N. Parvin, D. Merum, M. Kang, S. W. Joo, J. H. Jung and T. K. Mandal, J. Mater. Chem. A, 2025, 13, 24320–24386.
7. V. S. Bhat, J. M. Shivanna, A. Shetty, V. Molahalli, S. G. Krishnan, S. Sahoo, R. K. Pai, T. M. Aminabhavi and G. Hegde, Energy Fuels, 2025, 39, 16737–16767.
8. S. A. Kumar, S. Nadavurmath, S. Sahoo, G. K. Laxminarayana and C. S. Rout, ACS Appl. Electron. Mater., 2025, 7, 8850–8860.
9. Parul, S. Sahoo, M. Amrutha, S. Ratha, B. Chakraborty and S. K. Nayak, Mater. Adv., 2026, 7, 2068–2085.
10. A. Akhoondi, M. Y. Nassar, B. Yuliarto and H. Meskher, Synth. Sinter., 2023, 3, 200–212.
11. L. De Sousa Silva, E. E. Fileti and G. Colherinhas, ACS Mater. Au, 2025, 920–939.
12. P. Forouzandeh and S. C. Pillai, Mater. Today: Proc., 2021, 41, 498–505.
13. G. Saeed, T. Kang, J. S. Byun, D. Min, J. S. Kim, S. V. Sadavar and H. S. Park, Energy Mater., 2024, 4, 400023.
14. S. Liu, H. Zhang, J. Chen, X. Peng, Y. Chai, X. Shao, Y. He, X. Wang and B. Ding, Energies, 2025, 18, 1223.
15. P. Dubey, V. Shrivastav, S. Sundriyal and P. H. Maheshwari, ACS Appl. Nano Mater., 2024, 7, 18554–18565.
16. A. Tundwal, H. Kumar, B. J. Binoj, R. Sharma, G. Kumar, R. Kumari, A. Dhayal, A. Yadav, D. Singh and P. Kumar, RSC Adv., 2024, 14, 9406–9439.
17. A. Tamilselvan, S. Sengupta, A. Pramanik, P. M. Ajayan and M. Kundu, J. Mater. Sci., 2025, 60, 1661–1674.
18. S. Sengupta, S. Reza, R. Arulraj, R. Thapa and M. Kundu, Chem. Eng. J., 2025, 522, 167421.
19. P. A. K. Reddy, H. Han, K. C. Kim and S. Bae, ACS Appl. Energy Mater., 2022, 5, 13751–13762.
20. A. M. Bogale, T. Ramachandran, L. T. Tufa, B. B. Badassa, M. E. Suk, R. Pitcheri, J. Lee, S. K. Jilcha, A. Y. Tiky, B. A. Zenebe, N. K. Amare, M. M. Solomon and F. B. Tesema, Mater. Sci. Semicond. Process., 2025, 200, 109958.
21. T. Ramachandran, F. Hamed, Y. A. Kumar, R. K. Raji and H. H. Hegazy, J. Energy Storage, 2023, 73, 109299.
22. M. Naguib, M. Kurtoglu, V. Presser, J. Lu, J. Niu, M. Heon, L. Hultman, Y. Gogotsi and M. W. Barsoum, Adv. Mater., 2011, 23, 4248–4253.
23. N. K. Shaw, B. Maharana and S. Chatterjee, Phys. Scr., 2025, 100, 035952.
24. S. Saharan, U. Ghanekar and S. Meena, ChemistrySelect, 2022, 7, e202203288.
25. 2D Metal Carbides and Nitrides (MXenes): Structure, Properties and Applications, ed. B. Anasori and Y. Gogotsi, Springer International Publishing, Cham, 2019.
26. M. Sun, W. Ye, J. Zhang and K. Zheng, Inorganics, 2024, 12, 112.
27. H. Zhu, Appl. Comput. Eng., 2025, 124, 45–53.
28. R. Mangiri, T. Ramachandran, Y. A. Kumar, A. Ghosh, A. G. Al-Sehemi, A. K. Yadav and D. Mani, Mater. Today Chem., 2025, 48, 102864.
29. A. M. Aravind, M. Tomy, A. Kuttapan, A. M. Kakkassery Aippunny and X. T. Suryabai, ACS Omega, 2023, 8, 44375–44394.
30. B. Shen, R. Hao, Y. Huang, Z. Guo and X. Zhu, Crystals, 2022, 12, 1099.
31. G. Behzadi Pour and L. Fekri Aval, Results Eng., 2024, 24, 103045.
32. S. Yi, L. Wang, X. Zhang, C. Li, Y. Xu, K. Wang, X. Sun and Y. Ma, Nanotechnology, 2023, 34, 432001.
33. J. Zhou, M. Dahlqvist, J. Björk and J. Rosen, Chem. Rev., 2023, 123, 13291–13322.
34. J. Björk and J. Rosen, Chem. Mater., 2021, 33, 9108–9118.
35. Y. Gogotsi, Chem. Mater., 2023, 35, 8767–8770.
36. Y. Anil Kumar, V. K. Mishra, T. Ramachandran, A. M. Fouda, H. H. Hegazy, E. Ravindran, A. Maity and S.-C. Kim, Sustainable Energy Fuels, 2026.
37. J. L. Hart, K. Hantanasirisakul, A. C. Lang, B. Anasori, D. Pinto, Y. Pivak, J. T. Van Omme, S. J. May, Y. Gogotsi and M. L. Taheri, Nat. Commun., 2019, 10, 522.
38. V. Natu and M. W. Barsoum, J. Phys. Chem. C, 2023, 127, 20197–20206.
39. Md. R. Alam, M. Gharami, B. Dev, M. A. Rahman and T. Islam, Adv. Mater. Interfaces, 2025, 12, 2400681.
40. V. dos, P. de Souza, L. M. G. da Silva, R. K. Nishihora and S. F. Santos, Surf. Coat. Technol., 2025, 132583.
41. M. Gao, F. Wang, S. Yang, A. Gaetano Ricciardulli, F. Yu, J. Li, J. Sun, R. Wang, Y. Huang, P. Zhang and X. Lu, Mater. Today, 2024, 72, 318–358.
42. T. Li, W. Qiang and B. Lei, Nanoscale, 2025, 17, 4854–4891.
43. J. Ma, L. Zhang and B. Lei, ACS Nano, 2023, 17, 19526–19549.
44. S. Venkateshalu, M. Shariq, B. Kim, M. Patel, K. S. Mahabari, S.-I. Choi, N. K. Chaudhari, A. N. Grace and K. Lee, J. Mater. Chem. A, 2023, 11, 13107–13132.
45. M. A. Zaed, R. Saidur, A. K. Pandey, M. Kadhom, K. H. Tan, J. Cherusseri and N. Abdullah, Sep. Purif. Technol., 2025, 354, 129055.
46. W. Coley, A.-A. Akhavi, P. Pena, R. Shang, Y. Ma, K. Moseni, M. Ozkan and C. S. Ozkan, Nanomaterials, 2025, 15, 1089.
47. A. Eskandarli, M. Khazaei and M. R. Mohammadizadeh, Mater. Solidif., 2025, 1, 9580012.
48. T. S. Mathis, K. Maleski, A. Goad, A. Sarycheva, M. Anayee, A. C. Foucher, K. Hantanasirisakul, C. E. Shuck, E. A. Stach and Y. Gogotsi, ACS Nano, 2021, 15, 6420–6429.
49. X. Zhang, W. Zhang and H. Zhao, Int. J. Energy Res., 2022, 46, 15559–15570.
50. X. Zhang, W. Zhang and H. Zhao, Mater. Today Commun., 2022, 33, 104384.
51. P. Sarkar, K. Chatterjee, P. Pal and K. Das, Mater. Sci. Semicond. Process., 2025, 185, 108881.
52. S. Shi, R. Zhong, L. Li, C. Wan and C. Wu, Ultrason. Sonochem., 2022, 90, 106208.
53. A. Raman, J. S. Jayan, B. D. S. Deeraj, M. Srivastava, K. Joseph and A. Saritha, Polym. Compos., 2025, 46, 3193–3207.
54. W. Cai, Z. Li, T. Cui, X. Feng, L. Song, Y. Hu and X. Wang, Composites, Part B, 2022, 244, 110204.
55. J. Zou, J. Wu, Y. Wang, F. Deng, J. Jiang, Y. Zhang, S. Liu, N. Li, H. Zhang, J. Yu, T. Zhai and H. N. Alshareef, Chem. Soc. Rev., 2022, 51, 2972–2990.
56. M. Mozafari and M. Soroush, Mater. Adv., 2021, 2, 7277–7307.
57. S. Patra, N. U. Kiran, P. Mane, B. Chakraborty, L. Besra, S. Chatterjee and S. Chatterjee, Surf. Interfaces, 2023, 39, 102969.
58. J. L. Hart, K. Hantanasirisakul, Y. Gogotsi and M. L. Taheri, ACS Nanosci. Au, 2022, 2, 433–439.
59. J. Wu, M. Yang, P. Lu, K. Wei, Y. Qu, Y. Zhang and D. Li, Appl. Phys. Lett., 2022, 121, 191602.
60. F. H. Fagerli, Z. Wang, T. Grande, H. Kaland, S. M. Selbach, N. P. Wagner and K. Wiik, ACS Omega, 2022, 7, 23790–23799.
61. D. Gan, Q. Huang, J. Dou, H. Huang, J. Chen, M. Liu, Y. Wen, Z. Yang, X. Zhang and Y. Wei, Appl. Surf. Sci., 2020, 504, 144603.
62. S. Nouseen and M. Pumera, J. Mater. Chem. A, 2025, 13, 34055–34084.
63. G. Zhong, H. Yang, M. Zeng, S. Liu, S. Chen, Z. Fan, C. Zhu, J. Xu and J. Yu, Adv. Funct. Mater., 2024, 34, 2313845.
64. Z. Chi, C. Wang, Y. Dong, Y. Zhou, H. Xu, Z. Islam, C. Qian and Y. Fu, Compos. Sci. Technol., 2022, 225, 109505.
65. J. Yoon, Y.-J. Kim, J.-Y. Song, A. Jamal, I. Gereige, C. Kim and H.-T. Jung, J. Mater. Chem. A, 2023, 11, 22295–22303.
66. A. S. Rana, N. Raza, M. J. Anwar and M. F. Nazar, Mater. Res. Bull., 2026, 193, 113720.
67. M. Shen, W. Jiang, K. Liang, S. Zhao, R. Tang, L. Zhang and J. Wang, Angew. Chem., 2021, 133, 27219–27224.
68. S. Kim, H. Jo, J. Yun, J.-W. Lee, J. Cho, K. Kang and H.-D. Lim, Nano Convergence, 2025, 12, 39.
69. F. E. A. Latif, M. Khalid, A. Numan, N. A. Manaf, N. M. Mubarak, H. A. Zaharin and E. C. Abdullah, J. Mol. Struct., 2025, 1329, 141407.
70. A. G. Juandito, D. S. Khaerudini, S. Priyono, G. T. M. Kadja, D. Djuhana and M. Khalil, J. Nanopart. Res., 2024, 26, 110.
71. F. Wang, C. Yang, C. Duan, D. Xiao, Y. Tang and J. Zhu, J. Electrochem. Soc., 2015, 162, B16–B21.
72. H. Alnoor, A. Elsukova, J. Palisaitis, I. Persson, E. N. Tseng, J. Lu, L. Hultman and P. O. Å. Persson, Mater. Today Adv., 2021, 9, 100123.
73. L. Yao, X. Tian, X. Cui, R. Zhao, X. Xiao and Y. Wang, J. Mater. Sci.: Mater. Electron., 2021, 32, 27837–27848.
74. A. J. Borah, V. Natu, A. Biswas and A. Srivastava, Oxford Open Mater. Sci., 2025, 5, itae017.
75. X. Lei and N. Lin, Crit. Rev. Solid State Mater. Sci., 2022, 47, 736–771.
76. C. Shi, M. Beidaghi, M. Naguib, O. Mashtalir, Y. Gogotsi and S. J. L. Billinge, Phys. Rev. Lett., 2014, 112, 125501.
77. R. Garg, A. Agarwal and M. Agarwal, Mater. Res. Express, 2020, 7, 022001.
78. P. V. Shinde, P. Mane, B. Chakraborty and C. Sekhar Rout, J. Colloid Interface Sci., 2021, 602, 232–241.
79. A. Sharma, A. Patra, K. Namsheer, P. Mane, B. Chakraborty and C. S. Rout, J. Mater. Sci., 2021, 56, 20008–20025.
80. M. Ashton, K. Mathew, R. G. Hennig and S. B. Sinnott, J. Phys. Chem. C, 2016, 120, 3550–3556.
81. J. D. Gouveia, F. Viñes, F. Illas and J. R. B. Gomes, Phys. Rev. Mater., 2020, 4, 054003.
82. B. Akgenc, Solid State Commun., 2019, 303–304, 113739.
83. U. Kalsoom, S. Khan, M. Kashif, H. S. Yaseen, S. A. Hussain, S. Azizi and M. Maaza, Ionics, 2025, 31, 10053–10073.
84. H. Zhu, Z. Liang, S. Xue, X. Ren, X. Liang, W. Xiong, L. Gao and A. Liu, Ceram. Int., 2022, 48, 27217–27239.
85. M. Sun and U. Schwingenschlögl, J. Phys. Chem. C, 2021, 125, 4133–4138.
86. M. Sun and U. Schwingenschlögl, Adv. Energy Mater., 2021, 11, 2003633.
87. J. Plaickner, T. Petit, P. Bärmann, T. Schultz, N. Koch and N. Esser, Phys. Chem. Chem. Phys., 2024, 26, 20883–20890.
88. R. Ibragimova, P. Rinke and H.-P. Komsa, Chem. Mater., 2022, 34, 2896–2906.
89. C. Rong, T. Su, Z. Li, T. Chu, M. Zhu, Y. Yan, B. Zhang and F.-Z. Xuan, Nat. Commun., 2024, 15, 1566.
90. A. Lipatov, H. Lu, M. Alhabeb, B. Anasori, A. Gruverman, Y. Gogotsi and A. Sinitskii, Sci. Adv., 2018, 4, eaat0491.
91. J. Fatheema, M. Fatima, N. B. Monir, S. A. Khan and S. Rizwan, Phys. E Low-dimens. Syst. Nanostruct., 2020, 124, 114253.
92. M. Sadeghi and B. Khoshnevisan, RSC Adv., 2024, 14, 20300–20311.
93. M. Ali and S. M. Alqahtani, ACS Appl. Energy Mater., 2023, 6, 7535–7544.
94. N. Zhang, Y. Hong, S. Yazdanparast and M. Asle Zaeem, 2D Mater., 2018, 5, 045004.
95. G. Gao, G. Ding, J. Li, K. Yao, M. Wu and M. Qian, Nanoscale, 2016, 8, 8986–8994.
96. G. Ying, A. D. Dillon, A. T. Fafarman and M. W. Barsoum, Mater. Res. Lett., 2017, 5, 391–398.
97. N. García-Romeral, Á. Morales-García, F. Viñes, I. D. P. R. Moreira and F. Illas, J. Phys. Chem. C, 2023, 127, 3706–3714.
98. Y. Ibrahim, A. Mohamed, A. M. Abdelgawad, K. Eid, A. M. Abdullah and A. Elzatahry, Nanomaterials, 2020, 10, 1916.
99. P. Chakraborty, T. Das, D. Nafday, L. Boeri and T. Saha-Dasgupta, Phys. Rev. B, 2017, 95, 184106.
100. M. M. Hassan, J. Islam, W. R. Sajal, M. N. H. Noman and M. A. Rahman, Heliyon, 2024, 10, e25913.
101. T. Nguyen-Minh Le, T. Le Minh Pham, T. B. Phan and Y. Kawazoe, RSC Adv., 2025, 15, 301–311.
102. P. J. Pöllmann, R. Sahu, M. Fečík, C. Scheu and J. M. Schneider, J. Mater. Chem. A, 2025, 13, 25590–25598.
103. X.-H. Zha, J. Yin, Y. Zhou, Q. Huang, K. Luo, J. Lang, J. S. Francisco, J. He and S. Du, J. Phys. Chem. C, 2016, 120, 15082–15088.
104. J. Lei, A. Kutana and B. I. Yakobson, J. Mater. Chem. C, 2017, 5, 3438–3444.
105. H. Zhou, Z. Chen, E. Kountoupi, A. Tsoukalou, P. M. Abdala, P. Florian, A. Fedorov and C. R. Müller, Nat. Commun., 2021, 12, 5510.
106. R. Santoy-Flores, H. N. Fernández-Escamilla, J. I. Páez-Ornelas, E. G. Perez-Tijerina, J. Guerrero-Sánchez, R. Ponce-Pérez, N. Takeuchi and Ma. G. Moreno-Armenta, ACS Omega, 2024, 9, 28903–28911.
107. K. Wang, H. Jin, H. Li, Z. Mao, L. Tang, D. Huang, J.-H. Liao and J. Zhang, Surf. Interfaces, 2022, 29, 101711.
108. X. Sha, N. Xiao, Y. Guan and X. Yi, RSC Adv., 2017, 7, 33402–33407.
109. X. Lv, S. Zhang, J. Wang, M. Wang, J. Shan and S. Zhou, J. Mol. Graphics Modell., 2022, 110, 108056.
110. Y. Ajaj, A. Basem, M. H. Khaddour, A. Yadav, M. Kaur, R. Sharma, M. Alsubih, S. Islam and R. Zainul, J. Mol. Graphics Modell., 2024, 130, 108774.
111. J. Zhang, K. Tao, L. Ma, Y. Yang, L. Yang and W. Duan, AIP Adv., 2022, 12, 055019.
112. X.-X. Yu, C. R. Weinberger and G. B. Thompson, Acta Mater., 2014, 80, 341–349.
113. M. Liu, J. Wu, C. Wang, Z. Sun, Z. Fan and C. Xin, Solid State Commun., 2022, 341, 114585.
114. B. Anasori, Y. Xie, M. Beidaghi, J. Lu, B. C. Hosler, L. Hultman, P. R. C. Kent, Y. Gogotsi and M. W. Barsoum, ACS Nano, 2015, 9, 9507–9516.
115. M. Khazaei, M. Arai, T. Sasaki, C. Chung, N. S. Venkataramanan, M. Estili, Y. Sakka and Y. Kawazoe, Adv. Funct. Mater., 2013, 23, 2185–2192.
116. Y. Wang, Y. Xu, M. Hu, H. Ling and X. Zhu, Nanophotonics, 2020, 9, 1601–1620.
117. M. Khazaei, A. Ranjbar, M. Ghorbani-Asl, M. Arai, T. Sasaki, Y. Liang and S. Yunoki, Phys. Rev. B, 2016, 93, 205125.
118. M. Khazaei, A. Ranjbar, M. Arai, T. Sasaki and S. Yunoki, J. Mater. Chem. C, 2017, 5, 2488–2503.
119. J. Wang, S. Deng, Z. Liu and Z. Liu, Natl. Sci. Rev., 2015, 2, 22–39.
120. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson, I. V. Grigorieva, S. V. Dubonos and A. A. Firsov, Nature, 2005, 438, 197–200.
121. S. A. Khan, B. Amin, L.-Y. Gan and I. Ahmad, Phys. Chem. Chem. Phys., 2017, 19, 14738–14744.
122. D. Parajuli and K. Samatha, AIP Adv., 2024, 14, 035011.
123. Y. Li, C. Xiong, H. Huang, X. Peng, D. Mei, M. Li, G. Liu, M. Wu, T. Zhao and B. Huang, Adv. Mater., 2021, 33, 2103054.
124. K. Xiong, P. Wang, G. Yang, Z. Liu, H. Zhang, S. Jin and X. Xu, Sci. Rep., 2017, 7, 15095.
125. X.-H. Li, R.-Z. Zhang and H.-L. Cui, ACS Omega, 2020, 5, 18403–18410.
126. V. Mauchamp, M. Bugnet, E. P. Bellido, G. A. Botton, P. Moreau, D. Magne, M. Naguib, T. Cabioc’h and M. W. Barsoum, Phys. Rev. B:Condens. Matter Mater. Phys., 2014, 89, 235428.
127. S. L. Murray, S. Serajian, S. I. Gnani Peer Mohamed, S. Robinson, R. Krishnamoorthy, S. R. Das, M. Bavarian, S. Nejati, U. Kilic, M. Schubert and M. Ghashami, ACS Appl. Mater. Interfaces, 2024, 16, 70763–70773.
128. R. Ibragimova, M. J. Puska and H.-P. Komsa, ACS Nano, 2019, 13, 9171–9181.
129. N. Liu, Q. Li, H. Wan, L. Chang, H. Wang, J. Fang, T. Ding, Q. Wen, L. Zhou and X. Xiao, Nat. Commun., 2022, 13, 5551.
130. G. Nandkumar, S. Z. Karazhanov and P. Selvarajan, J. Power Sources, 2025, 657, 238135.
131. M. P. Shilpa, S. J. Shetty, S. S. Bhat, N. B. Gummagol, S. Surabhi, J.-R. Jeong and S. C. Gurumurthy, Carbon, 2025, 235, 120025.
132. S. Mathew, M. Ramachandra, S. Devi K R, D. Pinheiro, S. Manickam, C. H. Pang and S. H. Sonawane, Mater. Today Sustain., 2023, 24, 100568.
133. E. M. D. Siriwardane, I. Demiroglu, C. Sevik and D. Çakır, ACS Appl. Energy Mater., 2019, 2, 1251–1258.
134. P. Komen, L. Ngamwongwan, S. Jungthawan, A. Junkaew and S. Suthirakun, ACS Appl. Mater. Interfaces, 2021, 13, 57306–57316.
135. N. García-Romeral, Á. Morales-García and F. Viñes, J. Phys. Chem. C, 2025, 129, 826–836.
136. T. Wang, C. Yao, R. Gao, M. Holicky, B. Hu, S. Liu, S. Wu, H. Kim, H. Ning, F. Torrisi and A. A. Bakulin, Nano Lett., 2024, 24, 16333–16341.
137. T. Sruthi, N. Kavyasree, A. K. Sneha and V. Mathew, ChemRxiv, 2025 DOI:10.26434/chemrxiv-2025-8q56d.
138. Q. Zhang, X. Zhang, Y. Xiao, C. Li, H. H. Tan, J. Liu and Y. Wu, ACS Omega, 2020, 5, 29272–29283.
139. L. Wang, W. Liu, F. Bai, X. Zheng, C. Yin, J. Wei, J. Ma, H. Bai and B. Dong, RSC Adv., 2024, 14, 19945–19952.
140. A. Rafique, U. Naeem, A. Marques, I. Ferreira, S. Rizwan and A. C. Baptista, ACS Omega, 2025, 10, 7621–7634.
141. G. Ying, S. Kota, A. D. Dillon, A. T. Fafarman and M. W. Barsoum, FlatChem, 2018, 8, 25–30.
142. P. Gao, M. Song, X. Wang, Q. Liu, S. He, Y. Su and P. Qian, Nanomaterials, 2022, 12, 556.
143. S. M. Majhi, A. Ali, Y. E. Greish, H. F. El-Maghraby and S. T. Mahmoud, Sci. Rep., 2023, 13, 3114.
144. D. T. K. Anh, L. V. Mui, P. H. Minh, N. T. Binh and M. Cadatal-Raduban, Acta Crystallogr., Sect. B:Struct. Sci., Cryst. Eng. Mater., 2022, 78, 714–720.
145. Y. Guo, D. Liu, B. Huang, L. Wang, Q. Xia and A. Zhou, J. Phys. Chem. Solids, 2023, 176, 111238.
146. S. Anwar, M. Rafique, M. Irshad, M. I. Khan, S. S. A. Gillani, M. Shakil, M. A. Nawaz, S. M. Shaheen and M. A. Assiri, Electrochim. Acta, 2025, 512, 145508.
147. H. Pazniak, A. S. Varezhnikov, D. A. Kolosov, I. A. Plugin, A. D. Vito, O. E. Glukhova, P. M. Sheverdyaeva, M. Spasova, I. Kaikov, E. A. Kolesnikov, P. Moras, A. M. Bainyashev, M. A. Solomatin, I. Kiselev, U. Wiedwald and V. V. Sysoev, Adv. Mater., 2021, 33, 2104878.
148. A. Sreedhar and J.-S. Noh, J. Water Proc. Eng., 2025, 75, 107996.
149. L. Gao, C. Ma, S. Wei, A. V. Kuklin, H. Zhang and H. Ågren, ACS Nano, 2021, 15, 954–965.
150. M. Liu, T. Wang, X. Cui, X. Li and X. Ju, J. Energy Storage, 2025, 120, 116508.
151. I. A. M. Ibrahim, S. Abdel-Azeim, A. M. El-Nahas, O. Kühn, C.-Y. Chung, A. El-Zatahry and M. F. Shibl, J. Phys. Chem. C, 2022, 126, 14886–14896.
152. R. Madhushree, K. R. Sunajadevi, K. P. Chaithra, T. P. Vinod and B. Saravanakumar, Nanoscale Adv., 2025, 7, 6791–6795.
153. Y. Cui, J. Zhu, H. Tong and R. Zou, iScience, 2023, 26, 105824.
154. S. Qiao, Y. Zhang, S. Li, L. Wei, H. Wu and F. Li, Phys. Chem. Chem. Phys., 2025, 27, 513–519.
155. J. Zhou, X. Zha, X. Zhou, F. Chen, G. Gao, S. Wang, C. Shen, T. Chen, C. Zhi, P. Eklund, S. Du, J. Xue, W. Shi, Z. Chai and Q. Huang, ACS Nano, 2017, 11, 3841–3850.
156. N. Kumar, M. Kolos, S. Bhattacharya and F. Karlický, J. Chem. Phys., 2024, 160, 124707.
157. Z. Kandemir, P. D'Amico, G. Sesti, C. Cardoso, M. V. Milošević and C. Sevik, Phys. Rev. Mater., 2024, 8, 075201.
158. Z. Kandemir, E. Torun, F. Paleari, C. Yelgel and C. Sevik, Phys. Rev. Mater., 2022, 6, 026001.
159. B. D. Aparicio-Huacarpuma, E. Marinho, J. A. S. Laranjeira, W. F. Giozza, A. M. A. Silva, A. C. Dias and L. A. Ribeiro, J. Phys. Chem. C, 2025, 129, 13568–13580.
160. L. Zhang, W. Su, H. Shu, T. Lü, L. Fu, K. Song, X. Huang, J. Yu, C.-T. Lin and Y. Tang, Ceram. Int., 2019, 45, 11468–11474.
161. N. G. Zamkova and V. S. Zhandun, Inorg. Chem., 2024, 63, 24295–24305.
162. A. Champagne and J.-C. Charlier, JPhys Mater., 2021, 3, 032006.
163. B. Anasori, C. Shi, E. J. Moon, Y. Xie, C. A. Voigt, P. R. C. Kent, S. J. May, S. J. L. Billinge, M. W. Barsoum and Y. Gogotsi, Nanoscale Horiz., 2016, 1, 227–234.
164. Z. Tan, Z. Fang, B. Li and Y. Yang, ACS Omega, 2020, 5, 25848–25853.
165. Y. Niu, Y. Jiang, F. Zou, W. Song, Y. Zhao, H. Zhang, Q. Li and Y. Pan, RSC Adv., 2025, 15, 13442–13452.
166. G. Wang, J. Phys. Chem. C, 2016, 120, 18850–18857.
167. X. Chen, J. B. Sallach, W. Ling, X. Zhao, T. Borch and Y. Gao, Appl. Catal., B, 2024, 350, 123953.
168. K. R. B. Singh, S. Thapa, J. Singh, S. S. Pandey and R. P. Singh, MXenes: Fundamentals and Applications, 2024, pp. 1–22.
169. S. Karki, R. Bhuyan, S. R. Geed and P. G. Ingole, MXene Reinf. Polym. Compos., 2024, 359–387.
170. B. Vénosová and F. Karlický, Nanoscale Adv., 2023, 5, 7067–7076.
171. P. Eliášová, B. Šmíd, J. Vejpravová, S. Li, F. Brivio, M. Mazur, D. N. Rainer, M. I. H. Mohideen, R. E. Morris and P. Nachtigall, J. Mater. Chem. C, 2024, 12, 5431–5441.
172. A. Kumar, S. K. Reddy, V. K. Verma, P. Patel, S. K. Sahoo, U. K. Goutam, K. Amemiya, A. Kandasami and V. R. Singh, Colloids Surf., A, 2025, 725, 137547.
173. S. Zhang, Y. Huang, J. Wang, X. Han, G. Zhang and X. Sun, Carbon, 2023, 209, 118006.
174. S. Ullah, T. Najam, A. U. Rehman, S. S. Alarfaji, M. A. Ahmad, S. Riaz, B. Akkinepally, S. S. A. Shah and M. A. Nazir, J. Alloys Compd., 2024, 1001, 175172.
175. D. Johnson, K. Hansen, R. Yoo and A. Djire, ChemElectroChem, 2022, 9, e202200555.
176. J. Hao Ran Huang, S.-W. Tseng and I.-W. Peter Chen, Chem. Eng. J., 2025, 503, 158232.
177. J. Xu, T. Peng, Q. Zhang, H. Zheng, H. Yu and S. Shi, ACS Appl. Nano Mater., 2022, 5, 8794–8803.
178. M. R. Lukatskaya, S. Kota, Z. Lin, M.-Q. Zhao, N. Shpigel, M. D. Levi, J. Halim, P.-L. Taberna, M. W. Barsoum, P. Simon and Y. Gogotsi, Nat. Energy, 2017, 2, 17105.
179. M. Ghidiu, M. R. Lukatskaya, M.-Q. Zhao, Y. Gogotsi and M. W. Barsoum, Nature, 2014, 516, 78–81.
180. L. Liu, E. Raymundo-Piñero, S. Sunny, P. Taberna and P. Simon, Angew. Chem., Int. Ed., 2024, 63, e202319238.
181. L. Qin, R. Samal, J. Jiang, J. Halim, N. Chen, F. Chabanais, P. O. A. Persson and J. Rosen, arXiv, 2025, preprint, arXiv:2505.04226,  DOI:10.48550/ARXIV.2505.04226.
182. S. M. Varghese, V. V. Mohan, S. Suresh, E. Bhoje Gowd and R. B. Rakhi, J. Alloys Compd., 2024, 973, 172923.
183. M. Das, H. Murari, S. Ghosh and B. Sanyal, arXiv, 2023, preprint, arXiv:2305.17462,  DOI:10.48550/ARXIV.2305.17462.
184. A. Sarycheva and Y. Gogotsi, Chem. Mater., 2020, 32, 3480–3488.
185. T. Hu, J. Wang, H. Zhang, Z. Li, M. Hu and X. Wang, Phys. Chem. Chem. Phys., 2015, 17, 9997–10003.
186. A. Champagne, L. Shi, T. Ouisse, B. Hackens and J.-C. Charlier, Phys. Rev. B, 2018, 97, 115439.
187. K. Prenger, Y. Sun, K. Ganeshan, A. Al-Temimy, K. Liang, C. Dun, J. J. Urban, J. Xiao, T. Petit, A. C. T. Van Duin, D. Jiang and M. Naguib, ACS Appl. Energy Mater., 2022, 5, 9373–9382.
188. M. Brunet Cabré, D. Spurling, P. Martinuz, M. Longhi, C. Schröder, H. Nolan, V. Nicolosi, P. E. Colavita and K. McKelvey, Nat. Commun., 2023, 14, 374.
189. S. Adomaviciute-Grabusove, A. Popov, S. Ramanavicius, V. Sablinskas, K. Shevchuk, O. Gogotsi, I. Baginskiy, Y. Gogotsi and A. Ramanavicius, ACS Nano, 2024, 18, 13184–13195.
190. H. Ahmed, H. Alijani, A. El-Ghazaly, J. Halim, B. J. Murdoch, Y. Ehrnst, E. Massahud, A. R. Rezk, J. Rosen and L. Y. Yeo, Nat. Commun., 2023, 14, 3.
191. T. Parker, Y. Zhang, K. Shevchuk, T. Zhang, V. Khokhar, Y.-H. Kim, G. Kadagishvili, D. Bugallo, M. Tanwar, B. Davis, J. Kim, Z. Fakhraai, Y.-J. Hu, D. Jiang, D. V. Talapin and Y. Gogotsi, ACS Nano, 2025, 19, 22228–22239.
192. T. Parker, D. Zhang, D. Bugallo, K. Shevchuk, M. Downes, G. Valurouthu, A. Inman, B. Chacon, T. Zhang, C. E. Shuck, Y.-J. Hu and Y. Gogotsi, Chem. Mater., 2024, 36, 8437–8446.
193. M. Downes, C. E. Shuck, R. J. Wang, P. P. Michałowski, J. Shochat, D. Zhang, M. Shekhirev, Y. Yang, N. J. Zaluzec, R. Arenal, S. J. May and Y. Gogotsi, J. Am. Chem. Soc., 2024, 146, 31159–31168.
194. V. G. Nair, M. Birowska, D. Bury, M. Jakubczak, A. Rosenkranz and A. M. Jastrzębska, Adv. Mater., 2022, 34, 2108840.
195. N. Kitchamsetti, A. L. F. De Barros, H. Han and S. Mhin, Adv. Sustainable Syst., 2025, 9, e00581.
196. M. Jakubczak, A. Szuplewska, A. Rozmysłowska-Wojciechowska, A. Rosenkranz and A. M. Jastrzębska, Adv. Funct. Mater., 2021, 31, 2103048.
197. X.-H. Zha, P. Xu, Q. Huang, S. Du and R.-Q. Zhang, Nanoscale Adv., 2020, 2, 347–355.
198. B. Zhang, J. Zhou and Z. Sun, J. Mater. Chem. A, 2022, 10, 15865–15880.
199. P. Aghamohammadi, F. Karakaya Mert, E. T. Akgul, N. Aghabalapoor Keshtiban, O. C. Altıncı, A. Gelir, C. Sanga, N. Nayir, H. Aydın and M. Demir, ACS Omega, 2025, 10, 8202–8212.
200. M. Ozkan, MRS Energy Sustain., 2024, 11, 181–190.
201. J. Chen, W. Zhang, R. Chen, Y. Dai, J. Zhang, H. Yang, W. Zong, Z. Jiang, Y. Zhong, J. Wang, X. Zhang and G. He, Adv. Energy Mater., 2025, 15, 2403757.
202. M. Ozkan, K. A. M. Quiros, J. M. Watkins, T. M. Nelson, N. D. Singh, M. Chowdhury, T. Namboodiri, K. R. Talluri and E. Yuan, Chem, 2024, 10, 443–483.
203. M. S. Javed, X. Zhang, T. Ahmad, M. Usman, S. S. A. Shah, A. Ahmad, I. Hussain, S. Majeed, M. R. Khawar, D. Choi, C. Xia, W. Al Zoubi, M. A. Assiri, A. M. Hassan, S. Ali and W. Han, Mater. Today, 2024, 74, 121–148.
204. D. Zhou, J. Wang, Q. Cui and Q. Li, J. Appl. Phys., 2014, 115, 113504.
205. A. Shayesteh Zeraati, S. A. Mirkhani, P. Sun, M. Naguib, P. V. Braun and U. Sundararaj, Nanoscale, 2021, 13, 3572–3580.
206. M. Wang, Y. Liu, H. Zhang, Y. Wu and L. Pan, Int. J. Heat Mass Transfer, 2022, 194, 123027.
207. Y. Yang, H. Wang, C. Wang, J. Liu, H. Wu, N. Liu, Q. Wang, Y. Shang and J. Zheng, Small, 2024, 20, 2405870.
208. Q. Tang, Z. Zhou and P. Shen, J. Am. Chem. Soc., 2012, 134, 16909–16916.
209. J. Jia, B. Li, S. Duan, Z. Cui and H. Gao, Nanoscale, 2019, 11, 20307–20314.
210. V. Krishnasamy, in Novel Energy Storage and Conversion Technologies for Two-Dimensional MXenes and MBenes, ed. K. Kannan and V. Tari, IGI Global, 2025, pp. 11–36.
211. S. Wei, X. Lai and G. M. Kale, ACS Appl. Mater. Interfaces, 2023, 15, 33560–33570.
212. X. Li, J. Hao, R. Liu, H. He, Y. Wang, G. Liang, Y. Liu, G. Yuan and Z. Guo, Energy Storage Mater., 2020, 33, 62–70.
213. S. Wei, G. Kale and X. Lai, Small, 2024, 2401573.
214. A. Taghavi-Kahagh, F. Behboodi-Sadabad and M. Salami-Kalajahi, J. Mater. Chem. B, 2025, 13, 7226–7248.
215. G. Sanyal, A. Vaidyanathan, A. T. Sathya and B. Chakraborty, Langmuir, 2025, 41, 22525–22534.
216. S. Wei, G. Kale and X. Lai, Sens. Actuators, B, 2025, 429, 137258.
217. C. J. Zhang, S. Pinilla, N. McEvoy, C. P. Cullen, B. Anasori, E. Long, S.-H. Park, A. Seral-Ascaso, A. Shmeliov, D. Krishnan, C. Morant, X. Liu, G. S. Duesberg, Y. Gogotsi and V. Nicolosi, Chem. Mater., 2017, 29, 4848–4856.
218. S. Iravani, A. Zarepour, A. Khosravi and A. Zarrabi, Nanoscale Adv., 2025, 7, 670–699.
219. M. Jakubczak, D. Bury, D. Moszczyńska, M. Naguib and A. M. Jastrzębska, J. Environ. Chem. Eng., 2025, 13, 116971.
220. M. Ramezani Farani, D. Mirzaee, A. Hatami, K. Kumar, S. M. Ghoreishian and Y. S. Huh, Bioact. Mater., 2026, 55, 546–567.
221. T. R. Dmytriv and V. I. Lushchak, Chem. Rec., 2024, 24, e202300338.
222. C. I. Idumah, J. Mater. Sci., 2022, 57, 14579–14619.
223. L. Liu, X. Zhang, Y. Liu and X. Gong, ACS Appl. Electron. Mater., 2025, 7, 2233–2270.
224. A. Huseyin and A. J. S. Salih, J. Alloys Compd., 2025, 1037, 181972.
225. H. Li, Y. Ma, Y. Wang, C. Li, Q. Bai, Y. Shen and H. Uyama, Renewable Energy, 2024, 224, 120144.
226. E. Kim, S. Kim, H. M. Jin, G. Kim, H.-H. Ha, Y. Choi, K. Min, S.-H. Cho, H. Han, C. W. Ahn, J. Roh, I.-K. Oh, J. Lee and Y. Lee, Nano-Micro Lett., 2025, 17, 86.
227. G. Chen, Y. Xie, Y. Tang, T. Wang, Z. Wang and C. Yang, Small, 2024, 20, 2307408.
228. C.-C. Hsiao, J. Kasten, D. Johnson, B. Ngozichukwu, R. M. S. Yoo, S. Lee, A. Erdemir and A. Djire, ACS Nano, 2024, 18, 7180–7191.
229. H. Wang and L. Pilon, J. Phys. Chem. C, 2011, 115, 16711–16719.
230. J. Zhao and A. F. Burke, J. Energy Chem., 2021, 59, 276–291.
231. J. Zhao, Q. Li, T. Shang, F. Wang, J. Zhang, C. Geng, Z. Wu, Y. Deng, W. Zhang, Y. Tao and Q.-H. Yang, Nano Energy, 2021, 86, 106091.
232. K. Wasnik, M. D. Pawar, L. R. Raphael, A. Pullanchiyodan, M. V. Shelke and P. Raghavan, J. Mater. Res., 2022, 37, 3865–3889.
233. Z. Yuan, S. Ju, W. Li, H. Guo, K. Chen, M. Yue, X. Yu and Y. Wang, Chem. Eng. J., 2022, 450, 138453.
234. S. Zhao, Z. Liu, G. Xie, X. Guo, Z. Guo, F. Song, G. Li, C. Chen, X. Xie, N. Zhang, B. Sun, S. Guo and G. Wang, Angew. Chem., Int. Ed., 2021, 60, 26246–26253.
235. Z. Cao, G. Liang, D. Ho, C. Zhi and H. Hu, Adv. Funct. Mater., 2023, 33, 2303060.
236. Z. Ye, Y. Jiang, L. Li, F. Wu and R. Chen, Adv. Mater., 2021, 33, 2101204.
237. Q. Hui, W. Fan, X. Xia and H. Liu, Adv. Sustainable Syst., 2025, 9, 2500087.
238. B. Meng, G. Liu, Y. Mao, F. Liang, G. Liu and W. Jin, J. Membr. Sci., 2021, 623, 119076.
239. Z. Chen, J. Du and J. Shi, Sep. Purif. Technol., 2025, 354, 129316.
240. N. P. Shetti, A. Mishra, S. Basu, T. M. Aminabhavi, A. Alodhayb and S. Pandiaraj, Energy Fuels, 2023, 37, 12541–12557.
241. M. Ghiji, V. Novozhilov, K. Moinuddin, P. Joseph, I. Burch, B. Suendermann and G. Gamble, Energies, 2020, 13, 5117.
242. A. Jilani, Z. Awan, S. A. A. Taqvi, F. Khan and T. Alshahrani, ChemBioEng Rev., 2024, 11, 95–114.
243. A. K. Tomar, T. Kshetri, N. H. Kim and J. H. Lee, Energy Storage Mater., 2022, 50, 86–95.
244. K. Liang, R. A. Matsumoto, W. Zhao, N. C. Osti, I. Popov, B. P. Thapaliya, S. Fleischmann, S. Misra, K. Prenger, M. Tyagi, E. Mamontov, V. Augustyn, R. R. Unocic, A. P. Sokolov, S. Dai, P. T. Cummings and M. Naguib, Adv. Funct. Mater., 2021, 31, 2104007.
245. J. V. M. Lima, H. G. Lemos, R. A. Silva, J. H. H. Rossato, M. H. Boratto and C. F. O. Graeff, J. Alloys Compd. Comm., 2024, 3, 100007.
246. M. Depijan, K. Hantanasirisakul and P. Pakawatpanurut, ACS Omega, 2024, 9, 22256–22264.
247. R. S. Ingole, S. L. Kadam, K. Kim, M. Kim, Y. T. Kim, J. S. Lee and J. G. Ok, Small, 2025, 21, e07971.
248. L. Li, N. Zhang, M. Zhang, X. Zhang and Z. Zhang, Dalton Trans., 2019, 48, 1747–1756.
249. X. Yang, Y. Yao, Q. Wang, K. Zhu, K. Ye, G. Wang, D. Cao and J. Yan, Adv. Funct. Mater., 2022, 32, 2109479.
250. N. Xue, X. Li, L. Han, H. Zhu, X. Zhao, J. Zhuang, Z. Gao and X. Tao, J. Mater. Chem. A, 2022, 10, 7960–7967.
251. X. Ma, X. Zhu, C. Huang and J. Fan, J. Membr. Sci., 2022, 647, 120334.
252. J. T. Lee, B. C. Wyatt, G. A. Davis, A. N. Masterson, A. L. Pagan, A. Shah, B. Anasori and R. Sardar, ACS Nano, 2021, 15, 19600–19612.
253. G. B. Pour, L. F. Aval and H. Nazarpour-Fard, Results Chem., 2025, 13, 101965.
254. Y. Liu, L. Zhang, K. Guo, J. Zhao, Y. Xing, Z. Wang, Y. Tian and X. Cui, Chem. Eng. J., 2025, 513, 162751.
255. L. Pu, J. Zhang, N. K. L. Jiresse, Y. Gao, H. Zhou, N. Naik, P. Gao and Z. Guo, Adv. Compos. Hybrid Mater., 2022, 5, 356–369.
256. K. Nasrin, M. Arunkumar, N. Koushik Kumar, V. Sudharshan, S. Rajasekar, D. Mukhilan, M. Arshad and M. Sathish, Chem. Eng. J., 2023, 474, 145505.
257. M. Cai, X. Wei, H. Huang, F. Yuan, C. Li, S. Xu, X. Liang, W. Zhou and J. Guo, Chem. Eng. J., 2023, 458, 141338.
258. N. Prabhakar, M. Lakshmi Narayanan, H. Mana-ay, N. Ponpandian, C. Viswanathan and P.-Y. Chen, Fuel, 2025, 398, 135489.
259. T. Ramachandran, R. M. N. Kalla, R. Khan, A. G. Al-Sehemi, Y. A. Kumar, A. K. Yadav and J. Lee, J. Power Sources, 2026, 664, 238945.
260. A. Chen, C. Wang, O. A. Abu Ali, S. F. Mahmoud, Y. Shi, Y. Ji, H. Algadi, S. M. El-Bahy, M. Huang, Z. Guo, D. Cui and H. Wei, Composites, Part A, 2022, 163, 107174.
261. Y. Wen, T. E. Rufford, X. Chen, N. Li, M. Lyu, L. Dai and L. Wang, Nano Energy, 2017, 38, 368–376.
262. P. Sun, J. Liu, Q. Liu, J. Yu, R. Chen, J. Zhu, G. Sun, Y. Li, P. Liu and J. Wang, Chem. Eng. J., 2022, 450, 138372.
263. H. He, J. Wang, Q. Xia, L. Wang, Q. Hu and A. Zhou, Appl. Surf. Sci., 2021, 568, 150971.
264. B. Shen, X. Liao, X. Zhang, H.-T. Ren, J.-H. Lin, C.-W. Lou and T.-T. Li, Electrochim. Acta, 2022, 413, 140144.
265. R. Shafique, M. Rani, A. Mahmood, R. A. Alshgari, K. Batool, T. Yaqoob, N. K. Janjua, S. Khan, S. Khan and G. Murtaza, Int. J. Energy Res., 2022, 46, 6689–6701.
266. U. Shreenag Meda, O. Madan Raikar, C. Adaguru Rudregowda, D. Rangappa, N. Rani, S. S. Ranga and A. Pandey, Chem.–Asian J., 2025, 20, e202401678.
267. I. Ali, M. Yousaf, I. H. Sajid, M. W. Hakim and S. Rizwan, Mater. Today Chem., 2023, 34, 101766.
268. X. Du, L. Wang, Y. Fu, H. Wang, M. Yuan, Q. Xia, Q. Hu and A. Zhou, Ceram. Int., 2023, 49, 19737–19745.
269. Y. Wang, J. Sun, X. Qian, Y. Zhang, L. Yu, R. Niu, H. Zhao and J. Zhu, J. Power Sources, 2019, 414, 540–546.
270. Y. A. Kumar, R. M. N. Kalla, R. Khan, A. G. Al-Sehemi, A. Ghosh, T. Ramachandran and J. Lee, J. Power Sources, 2026, 665, 239027.
271. Y. Tang, C. Yang, X. Xu, Y. Kang, J. Henzie, W. Que and Y. Yamauchi, Adv. Energy Mater., 2022, 12, 2103867.
272. X. Li, G. Wang, Q. Li, Y. Wang and X. Lu, Chem. Eng. J., 2023, 453, 139488.
273. Y. Sun, T. Li, X. Liu, Y. Han, Y. Liu, A. Zada, W. Deng, Z. Yuan and A. Dang, Chem. Eng. J., 2024, 494, 152911.
274. Y. Li, L. Ding, Z. Liang, Y. Xue, H. Cui and J. Tian, Chem. Eng. J., 2020, 383, 123178.
275. K. Rajavel, X. Yu, P. Zhu, Y. Hu, R. Sun and C. Wong, ACS Appl. Mater. Interfaces, 2020, 12, 49737–49747.
276. H. Zheng, Y. Wang, B. Niu, R. Ge, Y. Lei, L. Yan, J. Si, P. Zhong and X. Ma, J. Phys. Chem. C, 2021, 125, 15210–15222.
277. Y. Jiang, D. Baimanov, S. Jin, J. Cheuk-Fung Law, P. Zhao, J. Tang, J. Peng, L. Wang, K. S.-Y. Leung, W. Sheng and S. Lin, Proc. Natl. Acad. Sci. U. S. A., 2023, 120, e2210211120.
278. X. Hui, X. Ge, R. Zhao, Z. Li and L. Yin, Adv. Funct. Mater., 2020, 30, 2005190.
279. P. Wang, D. Zhao, X. Hui, Z. Qian, P. Zhang, Y. Ren, Y. Lin, Z. Zhang and L. Yin, Adv. Energy Mater., 2021, 11, 2003069.
280. X. Bai, S. Hou, X. Wang, D. Hao, B. Sun, T. Jia, R. Shi and B.-J. Ni, Catal. Sci. Technol., 2021, 11, 5028–5049.
281. Y. Shi, C. Liu, Z. Duan, B. Yu, M. Liu and P. Song, Chem. Eng. J., 2020, 399, 125829.
282. S. Palei, G. Murali, C.-H. Kim, I. In, S.-Y. Lee and S.-J. Park, Nano-Micro Lett., 2023, 15, 123.
283. P. Yi, H. Zou, Y. Yu, X. Li, Z. Li, G. Deng, C. Chen, M. Fang, J. He, X. Sun, X. Liu, J. Shui and R. Yu, ACS Nano, 2022, 16, 14490–14502.
284. D. Zhao, R. Zhao, S. Dong, X. Miao, Z. Zhang, C. Wang and L. Yin, Energy Environ. Sci., 2019, 12, 2422–2432.
285. P. A. Maughan, L. Bouscarrat, V. R. Seymour, S. Shao, S. J. Haigh, R. Dawson, N. Tapia-Ruiz and N. Bimbo, Nanoscale Adv., 2021, 3, 3145–3158.
286. T. Prasankumar, K. Manoharan, N. K. Farhana, S. Bashir, K. Ramesh, S. Ramesh and V. K. Ramachandaramurthy, Mater. Today Sustain., 2024, 28, 100963.
287. S. Hussain, D. Vikraman, M. T. Mehran, M. Hussain, G. Nazir, S. A. Patil, H.-S. Kim and J. Jung, Renewable Energy, 2022, 185, 585–597.
288. T. Xu, Y. Wang, K. Liu, Q. Zhao, Q. Liang, M. Zhang and C. Si, Adv. Compos. Hybrid Mater., 2023, 6, 108.
289. Y. Liu, J. Yu, D. Guo, Z. Li and Y. Su, J. Alloys Compd., 2020, 815, 152403.
290. Y. Zhou, K. Maleski, B. Anasori, J. O. Thostenson, Y. Pang, Y. Feng, K. Zeng, C. B. Parker, S. Zauscher, Y. Gogotsi, J. T. Glass and C. Cao, ACS Nano, 2020, 14, 3576–3586.
291. D. Wen, G. Ying, L. Liu, Y. Li, C. Sun, C. Hu, Y. Zhao, Z. Ji, J. Zhang and X. Wang, J. Alloys Compd., 2022, 900, 163436.
292. Z. Fan, Y. Wang, Z. Xie, D. Wang, Y. Yuan, H. Kang, B. Su, Z. Cheng and Y. Liu, Adv. Sci., 2018, 5, 1800750.
293. Q. Wang, S. Wang, X. Guo, L. Ruan, N. Wei, Y. Ma, J. Li, M. Wang, W. Li and W. Zeng, Adv. Electrode Mater., 2019, 5, 1900537.
294. W. Liang and I. Zhitomirsky, J. Mater. Chem. A, 2021, 9, 10335–10344.
295. R. Wang, S. Luo, C. Xiao, Z. Chen, H. Li, M. Asif, V. Chan, K. Liao and Y. Sun, Electrochim. Acta, 2021, 386, 138420.
296. X. Shi, F. Guo, K. Hou, G. Guan, L. Lu, Y. Zhang, J. Xu and Y. Shang, Energy Fuels, 2023, 37, 9704–9712.
297. S. Li, Q. Zhang, L. Liu, J. Wang, L. Zhang, M. Shi and X. Chen, J. Alloys Compd., 2023, 941, 168963.
298. N. K. P. Siddu, S. M. Jeong and C. S. Rout, Energy Adv., 2024, 3, 341–365.
299. Y. Zhang, J. Cao, Z. Yuan, L. Zhao, L. Wang and W. Han, J. Colloid Interface Sci., 2021, 599, 109–118.
300. Q. X. Xia, J. Fu, J. M. Yun, R. S. Mane and K. H. Kim, RSC Adv., 2017, 7, 11000–11011.
301. S. Bashir, L. Gurbanova, I. A. Shaaban, M. S. Javed, M. R. Karim, S. S. A. Shah and M. A. Nazir, J. Energy Storage, 2025, 132, 117588.
302. S. A. Kumar, M. S. Yadav, S. Karmakar, A. Kundu, V. Etacheri, S. Ratha, V. K. Pal and S. Sahoo, J. Mater. Chem. A, 2026, 14, 9177–9206.
303. Y. Li, P. Kamdem and X.-J. Jin, J. Alloys Compd., 2021, 850, 156608.
304. B. Chen, Q. Song, Z. Zhou and C. Lu, Adv. Mater. Interfaces, 2021, 8, 2002168.
305. M. Ali, A. M. Afzal, M. W. Iqbal, S. Mumtaz, M. Imran, F. Ashraf, A. Ur Rehman and F. Muhammad, Int. J. Energy Res., 2022, 46, 22336–22364.
306. W. Hou, Y. Sun, Y. Zhang, T. Wang, L. Wu, Y. Du and W. Zhong, J. Alloys Compd., 2021, 859, 157797.
307. B. Kirubasankar, M. Narayanasamy, J. Yang, M. Han, W. Zhu, Y. Su, S. Angaiah and C. Yan, Appl. Surf. Sci., 2020, 534, 147644.
308. T. Nawaz, M. Ahmad, I. Hussain, X. Chen, B. M. Abraham, S. Zhuang, K. H. Low, K. Zhang and J. He, Small Struct., 2025, 6, 2400664.
309. T. Ramachandran, R. K. Raji and M. Rezeq, J. Mater. Chem. A, 2025, 13, 12855–12890.
310. T. Ramachandran, R. M. N. Kalla, R. Khan, A. Ghosh, Y. A. Kumar, J. Lee and K. V. V. C. Mouli, Surf. Interfaces, 2025, 72, 107420.
311. Y. Ji, W. Li, Y. You and G. Xu, Colloids Surf., A, 2024, 699, 134680.
312. Y. Chen, F. Wu, S. Zhang, S. Chang, X. Hu, X. Zhang, P. Zhang and H. Zhang, Compos. Commun., 2024, 50, 102027.
313. J. Jiang, D. Wen, W. Zhao and L. Zhao, Langmuir, 2023, 39, 13890–13896.
314. W. Zhao, J. Jiang, W. Chen, Y. He, T. Lin and L. Zhao, Chem. Eng. J., 2023, 468, 143660.
315. S. Li, X. Que, X. Chen, T. Lin, L. Sheng, J. Peng, J. Li and M. Zhai, ACS Appl. Energy Mater., 2020, 3, 10882–10891.
316. X. Ren, M. Huo, M. Wang, H. Lin, X. Zhang, J. Yin, Y. Chen and H. Chen, ACS Nano, 2019, 13, 6438–6454.
317. Y. Hu, L. Wang, T. Lin, N. Zhao, M. Shi, J. Peng, J. Li, W. Shi and M. Zhai, Adv. Mater. Interfaces, 2020, 7, 1901839.
318. N. C. Frey, A. Bandyopadhyay, H. Kumar, B. Anasori, Y. Gogotsi and V. B. Shenoy, ACS Nano, 2019, 13, 2831–2839.
319. Y. Gan and Y. Xiong, RSC Adv., 2025, 15, 9555–9568.
320. Y. Yang, Z. Xiu, F. Pan, H. Liang, H. Jiang, H. Guo, X. Wang, L. Li, B. Yuan and W. Lu, Adv. Funct. Mater., 2025, 35, 2406133.
321. F. Qiu, Z. Wang, M. Liu, Z. Wang and S. Ding, Ceram. Int., 2021, 47, 24713–24720.
322. J. Ren, Z. Zhu, Y. Qiu, F. Yu, T. Zhou, J. Ma and J. Zhao, Chemosphere, 2021, 284, 131284.
323. R.-Z. Zhang, X.-H. Cui, H.-L. Cui and X.-H. Li, Appl. Surf. Sci., 2022, 581, 152360.
324. X. Wei, M. Cai, F. Yuan, D. Lu, C. Li, H. Huang, S. Xu, X. Liang, W. Zhou and J. Guo, Appl. Surf. Sci., 2022, 606, 154817.
325. I. C. Onyia, S. O. Ezeonu, D. Bessarabov and K. O. Obodo, Comput. Mater. Sci., 2021, 197, 110613.
326. U. Ahmed and G. Nabi, J. Power Sources, 2025, 653, 237745.
327. U. Ahmed and G. Nabi, J. Environ. Chem. Eng., 2025, 13, 115949.
328. J. Zhao, W. Yan, Z. Liu, X. Liu, Y. Tian and X. Cui, Nano Res., 2024, 17, 7174–7181.
329. J. Xu, Z. Liu, Q. Wang, J. Li, Y. Huang, M. Wang, L. Cao, W. Yao, H. Wu and C. Chen, ACS Appl. Mater. Interfaces, 2023, 15, 15367–15376.
330. L. Malyala, S. Karingula, T. Bhookya and G. K Vengatajalabathy, Energy Storage, 2024, 6, e70012.
331. U. Fatima, B. Basha, M. M. Baig, N. Alomayrah, S. G. Lee, M. N. Khan, M. S. Al-Buriahi, M. F. Warsi, M. Akhtar and I. Shakir, J. Alloys Compd., 2025, 1040, 183584.
332. W. Bao, H. Shen, G. Zeng, Y. Zhang, Y. Wang, D. Cui, J. Xia, K. Jing, H. Liu, C. Guo, F. Yu, K. Sun and J. Li, Nanoscale, 2025, 17, 6204–6265.
333. S. Gokul Eswaran, M. Rashad, A. Santhana Krishna Kumar and A. F. M. EL-Mahdy, Chem.–Asian J., 2025, 20, e202401181.
334. A. M. Malik, K. Albe and J. Rohrer, npj 2D Mater. Appl., 2025, 9, 56.
335. M. N. Lakhan, A. Hanan, Y. Wang, H. K. Lee and H. Arandiyan, Chem. Sci., 2024, 15, 15540–15564.
336. L. Qian, F. Rahmati, F. Li, T. Zhang, T. Wang, H. Zhang, S. Yan and Y. Zheng, Nanoscale, 2025, 17, 8975–8998.
337. B.-Z. Guo, S.-B. Kim, S.-Y. Lee and S.-J. Park, Carbon, 2025, 231, 119732.
338. J. D. Gouveia and J. R. B. Gomes, Phys. Chem. Chem. Phys., 2025, 27, 18760–18769.
339. L. Guo, B. Yan, J. Hou, H. Zheng, C. Sun, Y. Xu and Q. Wang, Electrochim. Acta, 2025, 543, 147560.
340. S. Wei, Y. Fu, M. Liu, H. Yue, S. Park, Y. H. Lee, H. Li and F. Yao, npj 2D Mater. Appl., 2022, 6, 25.
341. J. Pan, W. Liu, C. Zhan, Y. Liu, Z. Zhao, J. Sun, H. Li, Q. Liu and L. Xiong, ACS Appl. Energy Mater., 2025, 8, 8532–8542.
342. T. Zhou, C. Wu, Y. Wang, A. P. Tomsia, M. Li, E. Saiz, S. Fang, R. H. Baughman, L. Jiang and Q. Cheng, Nat. Commun., 2020, 11, 2077.
343. N. M. Shinde and M. Pumera, ACS Appl. Electron. Mater., 2024, 6, 7339–7345.
344. N. K. Gautam, A. Sharma, M. Varshney, J. P. Singh, H.-J. Shin, S. Kumar, R. Brajpuriya, B. Lee, K.-H. Chae and S.-O. Won, New J. Chem., 2025, 49, 11203–11217.
345. T. Ramachandran, M. P. Pachamuthu, R. K. Raji and F. R. M. S. Raj, Mater. Chem. Phys., 2026, 350, 131860.
346. S. N. Ansari, M. Saraf, Z. Abbas and S. M. Mobin, Nanoscale, 2023, 15, 13546–13560.
347. Y.-L. Huang and S.-W. Bian, ACS Appl. Energy Mater., 2024, 7, 10358–10366.
348. N. R. Hemanth, T. Kim, B. Kim, A. H. Jadhav, K. Lee and N. K. Chaudhari, Mater. Chem. Front., 2021, 5, 3298–3321.
349. A. Sikdar, F. Héraly, H. Zhang, S. Hall, K. Pang, M. Zhang and J. Yuan, ACS Nano, 2024, 18, 3707–3719.
350. P. E. Lokhande, V. Kadam, C. Jagtap, U. Rednam, B. A. Al-Asbahi and A. A. Aziz, Diamond Relat. Mater., 2025, 159, 112918.
351. M. Manikandan, M. Muthulakshmi, S. Abuthahir, E. Papanasam, D. D. Dubal, S. A. Ansari and E. Manikandan, J. Energy Storage, 2025, 126, 117038.
352. S. Hepsibha, C. M. Magdalane, T. Keerthana, G. Ramalingam and N. Manivannan, J. Solid State Electrochem., 2025, 29, 4763–4774.
353. P. T. Pınar, M. Gülcan and Y. Yardım, J. Alloys Compd., 2025, 1010, 177656.
354. N. Shaheen, S. Zulfiqar, M. E. El Sayed, A. Samir, M. Shahid, M. F. Warsi and E. W. Cochran, J. Alloys Compd., 2025, 1010, 178072.
355. J. Xiao, J. Wen, J. Zhao, X. Ma, H. Gao and X. Zhang, Electrochim. Acta, 2020, 337, 135803.
356. B. Shen, X. Liao, X. Hu, H.-T. Ren, J.-H. Lin, C.-W. Lou and T.-T. Li, J. Mater. Chem. A, 2023, 11, 16823–16837.
357. S. Park, S. H. Choi, J. M. Kim, S. Ji, S. Kang, S. Yim, S. Myung, S. K. Kim, S. S. Lee and K. An, Small, 2024, 20, 2305311.
358. M. Chen, J. Chen, X. Tan, W. Yang, H. Zou and S. Chen, J. Energy Storage, 2021, 44, 103456.
359. P. Dutta, A. Sikdar, A. Majumdar, M. Borah, N. Padma, S. Ghosh and U. N. Maiti, Carbon, 2020, 169, 225–234.
360. M. Das and S. Ghosh, J. Electrochem. Soc., 2022, 169, 090525.
361. Q. Xu, G. Yang, X. Fan and W. Zheng, ACS Omega, 2019, 4, 13209–13217.
362. S.-H. Yin, X.-H. Li, R.-Z. Zhang and H.-L. Cui, FlatChem, 2024, 44, 100632.
363. M.-L. Qin, S.-Y. Wu, T.-H. Guo, M.-Q. Wu, Q.-S. Zhu and M.-Q. Kuang, Surf. Interfaces, 2024, 46, 104073.
364. T. Sruthi, M. Das and V. Mathew, ChemRxiv, 2025 DOI:10.26434/chemrxiv-2025-0l0k0.
365. M. Fatima, J. Fatheema, N. B. Monir, A. H. Siddique, B. Khan, A. Islam, D. Akinwande and S. Rizwan, Front. Chem., 2020, 8, 168.
366. G. Chen, Y. Xie, Y. Tang, T. Wang, Z. Wang and C. Yang, Small, 2024, 20, 2307408.
367. J. D. Gouveia and J. R. B. Gomes, Phys. Rev. Mater., 2022, 6, 024004.
368. S. Zhou, X. Yang, W. Pei, N. Liu and J. Zhao, Nanoscale, 2018, 10, 10876–10883.
369. L. Xiao-Hong, S. Xiang-Ying and Z. Rui-Zhou, RSC Adv., 2019, 9, 27646–27651.
370. Y. Tang, Z. Bi, Y. Xie, X. Xuan and C. Yang, Front. Chem., 2025, 13, 1656521.
371. B. Wu, M. Li, V. Mazánek, Z. Liao, Y. Ying, F. M. Oliveira, L. Dekanovsky, L. Jan, G. Hou, N. Antonatos, Q. Wei, M. Li, B. Pal, J. He, D. Koňáková, E. Vejmělková and Z. Sofer, Small Methods, 2024, 8, 2301461.
372. T. Sakhraoui and F. Karlický, ACS Omega, 2022, 7, 42221–42232.
373. R.-Z. Zhang, X.-H. Cui, S.-S. Li, X.-H. Li and H.-L. Cui, J. Mol. Liq., 2022, 345, 118263.
374. Y.-M. Li, Y.-L. Guo and Z.-Y. Jiao, Curr. Appl. Phys., 2020, 20, 310–319.
375. K. Li, X. Wang, S. Li, P. Urbankowski, J. Li, Y. Xu and Y. Gogotsi, Small, 2020, 16, 1906851.
376. Y. Wang, Y. Yuan, X. Chen and W. Yang, ACS Appl. Mater. Interfaces, 2025, 17, 35457–35467.
377. S. Zhang, Y. Huang, J. Wang, X. Han, C. Chen and X. Sun, Appl. Surf. Sci., 2022, 599, 154015.
378. T. Su, Z. D. Hood, M. Naguib, L. Bai, S. Luo, C. M. Rouleau, I. N. Ivanov, H. Ji, Z. Qin and Z. Wu, Nanoscale, 2019, 11, 8138–8149.
379. K. Li, J. Zeng, Y. Wang, J. Zhang and Y. Zhou, Phys. Rev. Appl., 2025, 23, 014011.
380. M. Noman, M. Mahmood Baig, Q. Muhammad Saqib, S. R. Patil, C. S. Patil, J. Kim, Y. Ko, E. Lee, J. Hwang, S. Goo Lee and J. Bae, Chem. Eng. J., 2024, 499, 156697.
381. B. Koneru, J. Swapnalin, S. Natarajan, A. Franco Jr and P. Banerjee, ACS Omega, 2022, 7, 20369–20375.
382. G. R. Schleder, A. C. M. Padilha, C. M. Acosta, M. Costa and A. Fazzio, JPhys Mater., 2019, 2, 032001.
383. A. Merchant, S. Batzner, S. S. Schoenholz, M. Aykol, G. Cheon and E. D. Cubuk, Nature, 2023, 624, 80–85.
384. O. N. Oliveira and M. C. F. Oliveira, Front. Chem., 2022, 10, 930369.
385. J. Fang, M. Xie, X. He, J. Zhang, J. Hu, Y. Chen, Y. Yang and Q. Jin, Mater. Today Commun., 2022, 33, 104900.
386. K. Kim, L. Ward, J. He, A. Krishna, A. Agrawal and C. Wolverton, Phys. Rev. Mater., 2018, 2, 123801.
387. S. Manna, A. Das, S. Das and B. Pathak, ACS Mater. Lett., 2024, 6, 572–582.
388. E. W. Vertina, N. Aaron Deskins, E. Sutherland and O. Mangoubi, in 2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE, Nassau, Bahamas, 2022, pp. 1573–1578.
389. D. Ontiveros, S. Vela, F. Viñes and C. Sousa, ACS Catal., 2025, 15, 14403–14413.
390. J. D. Gouveia, T. L. P. Galvão, K. Iben Nassar and J. R. B. Gomes, npj 2D Mater. Appl., 2025, 9, 8.
391. M. Shariq, S. Marimuthu, A. R. Dixit, S. Chattopadhyaya, S. Pandiaraj, M. Muthuramamoorthy, A. N. Alodhyab, M. Khaja Nazeeruddin and A. N. Grace, Chem. Eng. J., 2024, 484, 149502.
392. S. Iravani, A. Khosravi, E. Nazarzadeh Zare, R. S. Varma, A. Zarrabi and P. Makvandi, RSC Adv., 2024, 14, 36835–36851.
393. S. Shrestha, K. J. Barvenik, T. Chen, H. Yang, Y. Li, M. M. Kesavan, J. M. Little, H. C. Whitley, Z. Teng, Y. Luo, E. Tubaldi and P.-Y. Chen, Nat. Commun., 2024, 15, 4685.
394. C. Cui, Y. Zhang, T. Ouyang, M. Chen, C. Tang, Q. Chen, C. He, J. Li and J. Zhong, Phys. Rev. Mater., 2023, 7, 033803.
395. T. Thanasarnsurapong, S. K. Jana, P. Detrattanawichai, W. Namunmong, W. Hirunpinyopas, P. Iamprasertkun and A. Boonchun, ACS Mater. Au, 2025, 5, 823–830.
396. R. Wang, S. Fang, Q. Huang and Y. Liu, J. Chem. Theory Comput., 2025, 21, 7628–7635.
397. X. Lin, S. R. Tee, P. R. C. Kent, D. J. Searles and P. T. Cummings, J. Chem. Theory Comput., 2024, 20, 651–664.
398. S. Guha, A. Kabiraj and S. Mahapatra, npj Comput. Mater., 2022, 8, 202.
399. M. N. Gjerding, A. Taghizadeh, A. Rasmussen, S. Ali, F. Bertoldo, T. Deilmann, N. R. Knøsgaard, M. Kruse, A. H. Larsen, S. Manti, T. G. Pedersen, U. Petralanda, T. Skovhus, M. K. Svendsen, J. J. Mortensen, T. Olsen and K. S. Thygesen, 2D Mater., 2021, 8, 044002.
400. S. Haastrup, M. Strange, M. Pandey, T. Deilmann, P. S. Schmidt, N. F. Hinsche, M. N. Gjerding, D. Torelli, P. M. Larsen, A. C. Riis-Jensen, J. Gath, K. W. Jacobsen, J. Jørgen Mortensen, T. Olsen and K. S. Thygesen, 2D Mater., 2018, 5, 042002.
401. M. Yao, J. Ji, X. Li, Z. Zhu, J.-Y. Ge, D. J. Singh, J. Xi, J. Yang and W. Zhang, Sci. China Mater., 2023, 66, 2768–2776.
402. J. Cheng, T. Li, Y. Wang, A. H. Ati and Q. Sun, Appl. Surf. Sci., 2023, 641, 158560.
403. S. Krishna and A. Mir, Energy Adv., 2024, 3, 2986–2998.
404. B. M. Abraham, P. Sinha, P. Halder and J. K. Singh, J. Mater. Chem. A, 2023, 11, 8091–8100.
405. T. Ramachandran, L. Zheng, H. Butt and M. Rezeq, Mater. Today Adv., 2026, 29, 100675.
406. R. W. Epps, M. S. Bowen, A. A. Volk, K. Abdel-Latif, S. Han, K. G. Reyes, A. Amassian and M. Abolhasani, Adv. Mater., 2020, 32, 2001626.
407. A. A. Volk, R. W. Epps, D. T. Yonemoto, B. S. Masters, F. N. Castellano, K. G. Reyes and M. Abolhasani, Nat. Commun., 2023, 14, 1403.
408. M. Cheng, C.-L. Fu, R. Okabe, A. Chotrattanapituk, A. Boonkird, N. T. Hung and M. Li, Nat. Mater., 2026, 25, 174–190.
409. M. D. Witman and P. Schindler, Digital Discovery, 2025, 4, 625–635.
410. Z. Xiong, Y. Cui, Z. Liu, Y. Zhao, M. Hu and J. Hu, Comput. Mater. Sci., 2020, 171, 109203.
411. K. M. Jablonka, D. Ongari, S. M. Moosavi and B. Smit, Chem. Rev., 2020, 120, 8066–8129.
412. E. Stach, B. DeCost, A. G. Kusne, J. Hattrick-Simpers, K. A. Brown, K. G. Reyes, J. Schrier, S. Billinge, T. Buonassisi, I. Foster, C. P. Gomes, J. M. Gregoire, A. Mehta, J. Montoya, E. Olivetti, C. Park, E. Rotenberg, S. K. Saikin, S. Smullin, V. Stanev and B. Maruyama, Matter, 2021, 4, 2702–2726.
413. A. J. Cohen, P. Mori-Sánchez and W. Yang, Chem. Rev., 2012, 112, 289–320.
414. J. A. Gauthier, S. Ringe, C. F. Dickens, A. J. Garza, A. T. Bell, M. Head-Gordon, J. K. Nørskov and K. Chan, ACS Catal., 2019, 9, 920–931.
415. J. Schmidt, M. R. G. Marques, S. Botti and M. A. L. Marques, npj Comput. Mater., 2019, 5, 83.
416. Y. Gogotsi and B. Anasori, ACS Nano, 2019, 13, 8491–8494.

---