Gallic acid-derived carbon dots intercalated 3D porous Ti₃C₂T_x MXene for high-capacity supercapacitors

1. Introduction

MXenes are two-dimensional (2D) materials that have been widely investigated in many fields due to their compelling surface properties and unique structures. For instance, MXene-based materials exhibit high conductivity and good hydrophilicity, indicating their promising potential for catalysis, energy storage, shielding electromagnetic waves, and sensing applications [1,2]. However, the van der Waals forces between the monolayers in MXene structures lead to layer stacking or undesirably narrow layer spacing. When MXenes are used as electrochemical materials, this results in poor electrochemical performance because ion intercalation/deintercalation is inhibited [3-5].

To address this daunting challenge, researchers have investigated a variety of interesting approaches. For instance, a guest material can be introduced as an intermediate layer to generate a porous structure between the MXene layers [6-8]. Jiang’s team reports an innovative restricted hydrothermal strategy that combines chemical crosslinking to construct three-dimensional layered porous MXene/reduced graphene oxide (RGO) composite fibers. This well-designed architecture effectively prevents MXene stacking while creating efficient ion transport channels. It is worth noting that thiourea serves as both a chemical crosslinking agent and selective reducing agent, while reduced graphene oxide nanosheets have the physical barrier effect, which synergically enhances the oxidation resistance of the composite. The optimized composite fiber has excellent electrical conductivity (862.2 S cm⁻¹), mechanical strength (93.1 MPa), and oxidation resistance (90.5% conductivity after 60 days) [9]. Assembling 2D sheets into three-dimensional (3D) morphologies is another effective method for preventing MXene layer stacking [10,11]. Moreover, the self-assembly of 3D porous structures generates an abundance of electrochemically active sites that can be accessed by electrolyte ions [12,13]. Yang’s team designed and implemented a layered heterostructure of conductive mesoporous hollow carbon spheres (MHCS) and MXene composite electrodes through simple vacuum filtration without further template removal processes. This direct preparation strategy not only effectively improves the specific surface area of the electrode, but also enhances the abundant surface pores and good electrical conductivity of MHCS, improves the permeability of electrolyte solution, shortens the ion transport path, and thus significantly improves the specific capacitance (395F g^-1 at 2 mVs^-1) and rate performance (70.9% at 1000 mV s^-1) of the electrode. The introduction of carbon nanotubes (CNTs) further improves electrical conductivity and stability while maintaining good flexibility. As a result, MXene/MHCS/CNT films offer high specific capacitance, excellent magnification performance, and excellent cycle stability [11]. However, although three-dimensional structures are conducive to electrolyte ion transport in Ti₃C₂T_x MXene electrodes, these structures are not conducive to electron transport between the Ti₃C₂T_x MXene nanosheets. Due to these factors, new strategies should be developed to prepare three-dimensional porous MXene structures that inhibit the interlayer stacking of Ti₃C₂T_x MXene nanosheets while guaranteeing rapid electron transport between the nanosheet layers. Due to their small size, carbon dots formed in situ can be easily embedded into the interlayer of MXene, playing a pillar-like role and thus inhibiting the interlayer stacking of MXene. At the same time, due to its high conductivity, it can also serve as an electron transmission channel between MXene layers, thereby enhancing the electron transmission efficiency between MXene layers. Zhang’s team proposed a carbon dot-intercalated strategy to prepare flexible MXene thin film electrodes with large ion-accessible active surfaces and high density by gelating and carbonizing calcium alginate (CA) within MXene nanosheets. CA hydrogels are formed in MXene nanosheets and accompanied by evaporation drying, resulting in high density of MXene/CA films. During the carbonization process, CA-derived carbon dots can be embedded into MXene nanosheets, increasing the layer spacing and promoting electrolyte diffusion within MXene films. Thus, MXene films embedded with carbon dots exhibit high capacitance at 3 M H₂SO₄ (1244.6F cm⁻³ at 1 A g⁻¹). Excellent rate performance (662.5 F cm⁻³ at 1000 A g⁻¹) and excellent cycle stability (93.5% capacitance maintained after 30000 cycles). However, it is difficult for polymer materials to completely enter the interlayer of MXene, resulting in an uneven distribution of carbon dots between the MXene layers, and carbon dots derived from small molecular materials can solve this problem [14].

GA is a natural compound containing a catechol structure based on mussel adhesion protein, which shows great potential in the surface coating, modification, and functionalization of different substrates. The small molecule compounds containing phenolic hydroxyl groups can form intermolecular hydrogen bonds with the functional groups (-O,-OH, -F) on the surface of MXene nanosheets. Meanwhile, the phenolic hydroxyl groups on the surface of small molecule compounds and the hydroxyl groups on the surface of MXene nanosheets can be physically cross-linked under the action of divalent metal cations. Therefore, GA coating can effectively maintain the structural stability of MXene and achieve controllable surface functionalization.

In this paper, a carbon dot intercalation strategy is proposed for preparing three-dimensional porous carbon dot-intercalated MXene materials. First, the MXene nanosheets are cross-linked with GA. The hydroxyl groups of GA and the MXene surface are cross-linked under the action of Zn²⁺, which forms a three-dimensional structure. Then, a carbonization process is employed to generate carbon dots from the GA. These carbon dots are embedded into the MXene nanosheet structure, which increases the layer spacing and promotes electrolyte ion diffusion into the MXene. Thus, the Ti₃C₂T_x@C electrode achieves excellent electrochemical performance, good retention of capacitance at higher scanning rate values, and excellent cycle stability. Moreover, a symmetric supercapacitor prepared using carbon cloth and the Ti₃C₂T_x@C electrode also demonstrates superb energy density. The results in this paper offer a compelling and efficient pathway for obtaining strong electrochemical performance.

2. Materials and Methods

3. Results and Discussion

The process for preparing porous Ti₃C₂T_x@C aerogel has been shown in Figure 1. First, the Al layer in the Ti₃AlC₂ phase is removed using a LiF/HCl mixture to obtain multilayer Ti₃C₂T_x nanosheets. In this etching step, hydrophilic groups (-F, -O, and -OH) are used to terminate Ti₃C₂T_x [16]. The Ti atoms have different electron gravities than the O/F atoms, which means that the end of the Ti₃C₂T_x nanosheets carry more charge. Consequently, the Ti₃C₂T_x nanosheets are negatively charged. This provides the Ti₃C₂T_x nanosheets with good water dispersion properties due to electrostatic repulsion [17]. Next, gallic acid solution (GA) is added to the Ti₃C₂T_x MXene suspension and evenly mixed, followed by the addition of ZnCl₂ solution. The introduction of Zn²⁺ into the Ti₃C₂T_x MXene@GA suspension breaks its electrostatic balance. At the same time, Zn²⁺ acts as a cross-linker, enabling the formation of a three-dimensional porous Ti₃C₂T_x MXene@GA structure. The obtained Ti₃C₂T_x@GA hydrogel then undergoes freeze-drying to obtain a Ti₃C₂T_x@GA aerogel. Finally, the Ti₃C₂T_x@C aerogel containing intercalated carbon dots is obtained by carbonizing the Ti₃C₂T_x@GA aerogel at 500°C.

Close

Figure 1.

Preparation process for obtaining the three-dimensional porousTi₃C₂T_x@Caerogel.

Export to PPT

In the Figure 2, the surface topography changes caused by pore formation in the Ti₃C₂T_x MXene that have been shown in Figure 2(a). Large-sized MXene nanosheets with a typical layered structure can be observed. The MXene sheets are oriented in a face-to-face manner, so the Ti₃C₂T_x MXene has a dense structure, which may limit the availability of active sites for electrolyte ions. In contrast, a three-dimensional macroporous structure (macropore diameters of 5–10 μm) can be observed in the Ti₃C₂T_x@C and Ti₃C₂T_x@GAaerogel (Figure 2b and 2c). Both Ti₃C₂T_x and Ti₃C₂T_x@C have type-IV nitrogen–desorption isotherms with H₃ hysteresis loops (Figure S1a), indicating mesoporosity. Ti₃C₂T_x@C has a larger specific surface area than pure Ti₃C₂T_x ((35.57 vs. 29.69 m²g^-1). The barret-joyner-halenda (BJH) pore size distribution of Ti₃C₂T_x@C (Figure S1b) confirms its mesoporosity [18]. The presence of the intercalated carbon dots causes the MXene nanosheets in Ti₃C₂T_x@C to become more disordered. The 3D macroporous structure of Ti₃C₂T_x@C is attributed to the Zn²⁺ acting as a cross-linking agent for the -OH on the surface of Ti₃C₂T_x and the -OH on GA. This structure facilitates electrolyte penetration and ion accessibility on the surface of the MXene, which is expected to provide superior capacitance and rate performance. The microstructure of the Ti₃C₂T_x@C powder was observed by transmission electron microscope (TEM) (Figure 2e). The carbonized Ti₃C₂T_x@C nanosheets contain a distribution of many nanometer-sized particles. These nanoparticles are carbon, as indicated by the 0.43 nm lattice spacing (Figure 2f). Thus, Ti₃C₂T_x@C contains intercalated carbon particles. Element mapping demonstrates the uniform distribution of Zn in Ti₃C₂T_x@C (Figure 2g).

Close

Figure 2.

SEM micrographs of (a) Ti₃C₂T_x, (b) Ti₃C₂T_x@GA and (c) Ti₃C₂T_x@C. TEM micrographs of (d) Ti₃C₂T_x and (e) Ti₃C₂T_x@C. (f) HRTEM micrographs of Ti₃C₂T_x@C. (g) EDX elemental mapping images of C, Ti, O, F, and Zn for Ti₃C₂T_x@C. EDX: Energy dispersive X-ray spectroscopy.

Export to PPT

XRD was used to evaluate the atomic structures of Ti₃C₂T_x@C and Ti₃C₂T_x, as displayed in Figure 3(a). After etching with LiF-HCl, the (002) peak of the Ti₃AlC₂ MAX phase is shifted to 7.28° (indicating 12.13 Å d-spacing), and the (104) peak vanishes, proving that 2D Ti₃C₂T_x was successfully synthesized [19]. Compared with Ti₃C₂T_x, the (002) peak of the Ti₃C₂T_x@C aerogel shifts to a lower 2θ value by 6.36°(indicating 13.85 Å d-spacing). Thus, Ti₃C₂T_x@C has wider d-spacing, which can be ascribed to the carbon dots intercalated between the Ti₃C₂T_x MXene nanosheets. This is favorable for making the structure more accessible to the electrolyte [15]. The XPS survey spectra of Ti₃C₂T_xandTi₃C₂T_x@C demonstrate the successful addition of Zn²⁺ to Ti₃C₂T_x@C (Figure 3b). The Ti 2p spectra (Figure 3c) show deconvoluted peak pairs at 454.76 and 460.56 eV (Ti-C), 455.46 and 461.24 eV (Ti²⁺), 456.56 eV and 462.71 eV (Ti³⁺), and 459.03 eV and 464.21 eV (TiO₂) [16]. Due to the presence of Zn²⁺ as a cross-linker, Ti₃C₂T_x@C exhibits stronger Ti²⁺ peaks. The C1s spectrum of Ti₃C₂T_x@C (Figure 3d) has a stronger C-C bond peak than Ti₃C₂T_x, which confirms the large number of carbon dots in Ti₃C₂T_x@C. The O 1s spectrum of Ti₃C₂T_x@C contains a slightly smaller C-Ti-OH bond peak compared to that of Ti₃C₂T_x (Figure 3e), which can also be ascribed to Zn²⁺ cross-linking the -OH on the MXene surface with the -OH on GA. The Zn 2p spectrum of Ti₃C₂T_x@C confirms that Zn2+ has been successfully inserted into the Ti₃C₂T_x layers (Figure 3f).

Close

Figure 3.

(a) X-ray diffractograms of Ti₃C₂T_x and Ti₃C₂T_x@C. (b) XPS survey, (c) Ti 2p, (d) C 1s, (e) O 1s, and (f) Zn 2p spectra of Ti₃C₂T_x and Ti₃C₂T_x@C.

Export to PPT

The porous, three-dimensional morphology and larger layer spacing of the Ti₃C₂T_x@C aerogel favors the more rapid transport of electrolyte ions in shorter ion diffusion channels [17,20]. In addition, the numerous carbon dots in Ti₃C₂T_x@C provide excellent electron transport pathways between the Ti₃C₂T_x nanosheets. The CV analysis of the prepared electrodes is exhibited in Figure 4(a). Reversible Faraday pseudocapacitive behavior mainly contributes to the capacitance of Ti₃C₂T_x and Ti₃C₂T_x@C, as demonstrated by the pair of redox peaks for both materials [21,22]. Ti₃C₂T_x@C has a higher C_g than Ti₃C₂T_x, as indicated by the larger CV curve area. The GCD analysis of the Ti₃C₂T_x and Ti₃C₂T_x@C electrodes is displayed in Figure 4(b). These curves are not linear, indicating that both materials exhibit contributions from Faradaic pseudocapacitance and electric double-layer capacitance [23]. The higher C_g of Ti₃C₂T_x@C is confirmed by its longer discharge time. Ti₃C₂T_x@C was further studied by obtaining multiple CV curves at varying scan rate values (Figure 4c). Within 2–100 mVs^-1, no significant distortion in the curve shape is observed. Thus, Ti₃C₂T_x@C has superb rate performance and low internal resistance [16]. Ti₃C₂T_x (Figure S2a) shows a similar CV curve trend. Moreover, GCD curves acquired at various current density values are also consistent for both samples (Figure 4d and Figure S2b). The calculated C_g values of both electrode materials are displayed in Figure 4(e). Ti₃C₂T_x@C displays significantly higher C_g values than Ti₃C₂T_x at all scan rates (393.7 vs. 315.5 Fg^-1 at 2 mVs^-1). This difference is attributed to the three-dimensional porous structure of Ti₃C₂T_x@C, which hinders the stacking of layers, opens more electrodynamically active sites, and promotes faster electrolyte ion transport [18,24]. Additionally, the carbon dots in Ti₃C₂T_x@C act as fast electron transfer channels that accelerate electron transfer between the Ti₃C₂T_x nanosheets. At 200 mVs^-1, Ti₃C₂T_x@C still retains a C_g of 329.3 F g^-1, while that of Ti₃C₂T_x is only 272.3 Fg^-1. The GCD curves acquired in the current density range of 1–100 Ag^-1 display a similar trend (Figure 4f). At 1 A g^-1, Ti₃C₂T_x@C exhibits a notably high C_g of 436 Fg^-1, and when the current density increases to 100 A g^-1, this value only decreases to 292.5 Fg^-1. However, the C_g of the Ti₃C₂T_x electrode at 100 Ag^-1 declines to 257.5 Fg^-1. Therefore, Ti₃C₂T_x@C has better rate performance, which can be explained by three main reasons. First, the self-assembly of the Ti₃C₂T_x sheets with GA via Zn²⁺ cross-linking forms a three-dimensional porous structure, which inhibits Ti₃C₂T_x layer stacking. This enables the exposure of more electrochemically active sites. Second, the formed Ti₃C₂T_x@C structure facilitates electrolyte ion transport, which ensures excellent rate performance. Third, the carbon dots in Ti₃C₂T_x@C accelerate electron transport between the Ti₃C₂T_x nanosheets by providing electron transport paths. A C_g value comparison between this work and other MXene-based electrode materials is shown in Table S1. Compared to these other materials, Ti₃C₂T_x@C has a superb C_g and exceptional rate performance [11,25-27].

Close

Figure 4.

Electrochemical analysis ofTi₃C₂T_x and Ti₃C₂T_x@Celectrodes: (a) CV (5 mV s^–1) and (b) GCD (1 A g^–1) curves obtained in 3 М H₂SO₄ electrolyte. (c) CV curves of Ti₃C₂T_x@Cat various scan rate values. (d) GCD curves of Ti₃C₂T_x@Cunder different current density values. Rate performance of Ti₃C₂T_x and Ti₃C₂T_x@Celectrodes at different (e) scanning rate and (f) current density values.

Export to PPT

A log (i) vs. log(ν) plot was developed to study the charge storage kinetics and distinguish the charge storage mechanisms of Ti₃C₂T_x@C [17]. The peak current was obtained using the following Eq. (5):

Where the variables a and b are obtained by plotting log(i) and log(ν). A capacitatively controlled charge storage mechanism is revealed by b = 1, and a diffusion-controlled charge storage process is revealed by b = 0.5 [28]. As displayed in Figure 5(a), the anode peak b value of Ti₃C₂T_x@C is 0.94, while the cathode peak b value is 0.89. Consequently, the storage of charges in Ti₃C₂T_x@C is mainly related to its fast capacitance behavior. At 20 mVs^-1, capacitive control accounts for 82.45% of overall charge storage in the Ti₃C₂T_x@C electrode (Figure 5b). From 2–50 mVs^-1, the capacitive control contribution increases from 72.31% to 87.87% with increasing scan rate (Figure 5c). This indicates that the storage of charge in Ti₃C₂T_x@C is primarily governed by Faraday pseudocapacitance behavior rather than a diffusion-controlled process [29]. The ion diffusion kinetics of the synthesized electrodes were evaluated using EIS. Figure 5(d) exhibits the Nyquist plots of Ti₃C₂T_x and Ti₃C₂T_x@C. Ti₃C₂T_x@C has a negligible semicircle, indicating low charge transfer resistance. The equivalent circuit fitting parameters in Table S2 and Table S3 show that Ti₃C₂T_x@C has a lower charge transfer resistance than Ti₃C₂T_x. This is mainly because the intercalation of carbon dots in Ti₃C₂T_x@C facilitates electron transfer between the Ti₃C₂T_x nanosheets. In addition, the low-frequency linear region of the Ti₃C₂T_x@C Nyquist plot has a higher slope. This is because the 3D porous structure of the Ti₃C₂T_x@C electrode accelerates the transport of electrolyte ions [30]. Bode plots of the Ti₃C₂T_x and Ti₃C₂T_x@C electrode samples are displayed in Figure 5(e). These electrode materials both exhibit phase angles close to -90°, indicating that their capacitive properties are close to the ideal state [31]. The relaxation time constant τ₀ of the Ti₃C₂T_x@C electrode (1.25 s) is shorter than that of the Ti₃C₂T_x electrode (1.84 s), indicating that electrolyte ions are more rapidly transported within Ti₃C₂T_x@C. The Ti₃C₂T_x@C electrode also has excellent cyclic stability (Figure 5f), retaining more than 90.3% of its C_g even after 5,000 cycles, while the C_g of Ti₃C₂T_x only retains 87.6%. To further confirm the stability of the MXene structure, the Ti₃C₂T_x@C electrode that underwent cyclic stability tests was characterized by SEM. After long-term cycling, the porous structure of Ti₃C₂T_x@C collapses to a certain extent, but the basic morphology is still retained (Figure S3).

Close

Figure 5.

Electrochemical evaluation of Ti₃C₂T_x@Celectrode: (a) Peak current vs. scan rate log-log plot. (b) Capacitance contributions at 20 mV s^-1. (c) Contributions of capacitive control and diffusion control to total capacitance at different scan rate values. (d) Nyquist and (e) Bode plots of Ti₃C₂T_x and Ti₃C₂T_x@C. (f) Cyclic stability of Ti₃C₂T_x@C and Ti₃C₂T_x at 100A g^-1.

Export to PPT

To investigate the potential for using Ti₃C₂T_x@C in practical applications, a symmetric supercapacitor was prepared [32-39]. CV curves of this symmetric supercapacitor from 2–50 mVs^-1 are shown in Figure 6(a). With the increase in scan rate, no significant distortion was observed at 20 mVs^-1; however, a significant distortion occurred at 50 mV s^-1, indicating that the device has good reversibility and rate performance. A similar trend can be seen in the GCD curves acquired from 1–5 Ag^-1 (Figure 6b). The calculated C_g values have been shown in Figure 6(c). This symmetric supercapacitor achieves a C_g of 153.25 Fg^-1 at 2 mVs^-1. Moreover, a C_g of 45.29 Fg^-1 is still reached at 50 mVs^-1. Next, a cycling test was performed at 50 mVs^-1 to study the stability of the symmetric supercapacitor (Figure 6d). After 5000 continuous cycles of charging and discharging, 90.1% of the original C_g value is still retained, confirming the good cycle stability of the prepared device. A Ragone plot (Figure 6e) is used to show the gravimetric E_g and P_g values of the symmetric supercapacitor. The highest E_g of 21.28 Whkg^-1 is attained at a P_g of 153.25 Wkg^-1. Moreover, an E_g of 14.63 Whkg^-1 is still retained when P_g increases to 526.75 Wkg^-1. The practical use of this symmetric supercapacitor design was demonstrated by connecting four symmetric supercapacitor devices in series. These devices are capable of successfully lighting an light emitting diode (LED) (Figure 6f).

Close

Figure 6.

Performance of Ti₃C₂T_x@C-based symmetric supercapacitor. (a) CV curves at different scanning rate values. (b) GCD curves obtained at different current density values. (c) Rate performance at various scanning rate values. (d) Cycling of Ti₃C₂T_x@C and Ti₃C₂T_x at 50 mV s^-1. (e) Comparative Ragone plot. (f) Application of four Ti₃C₂T_x@C-based symmetric supercapacitors in series.

Export to PPT

4. Conclusions

In summary, a three-dimensional porous carbon dot-intercalated Ti₃C₂T_x MXene material was prepared by a simple carbon dots intercalation strategy. The -OH on the Ti₃C₂T_x MXene nanosheet surface was cross-linked with the -OH on GA under the action of Zn²⁺, forming a three-dimensional structure. Next, Ga-derived carbon dot particles were embedded into the Ti₃C₂T_x MXene via carbonization, leading to larger layer spacing and an active surface accessible to ions. Moreover, these carbon dots acted as excellent electron transport channels, which accelerated electron transport between the Ti₃C₂T_x MXene nanosheets. Due to this favorable design, the Ti₃C₂T_x@C electrode had a C_g of 393.7 Fg^-1 at 2 mVs^-1 and maintained 83.7% of this value at 200 mVs^-1. Moreover, the electrode showed 90.3% C_g retention after 5000 charging and discharging cycles, indicating superb cycle stability. A Ti₃C₂T_x@C/carbon cloth symmetric supercapacitor displayed a high E_g of 21.28 Whkg^-1 under a P_g of 153.25 Whkg^-1. The results and analysis presented in this study offer a useful approach for efficiently preparing MXene electrodes with high performance. However, due to the limitation of the MXene material preparation process, large-scale manufacturing of Ti₃C₂T_x@C composites has not been possible. In addition, due to the high activity of MXene, it is also necessary to continue to improve the long-term cyclic stability of Ti₃C₂T_x@C composites.