2. Interlayer Chemical Modulation of Electronic Phase Transitions
2.1. Superconducting Transition
Superconductivity is a macroscopic quantum phenomenon, which features elusive electron pairing with a zero-resistance state, presenting a unique platform to study strongly correlated physics and to construct advanced electronic devices. The modulation of superconducting states is beneficial for understanding the mechanism of superconducting transitions and for exploring new exciting physical properties. Numerous layered 2D-MCs have demonstrated superconductivity behaviour, providing an ideal model system to modulate the superconducting properties
[24]. As is known, superconductivity is sensitive to the charge carrier density and magnetic interactions. It can be expected that interlayer chemical modulations enable stronger host–guest interaction in the confined vdW space, giving rise to enhanced electronic coupling and producing new phenomena.
Recently, Wang et al. found the tunable superconducting transition temperature for cetyltrimethylammonium (CTA
+)-intercalated 2H-TaS
2 [25]. A series of TaS
2-(CTA
+)
x (x denotes the intercalated amount of CTA
+) compounds were obtained by controlling the charge time using cetyltrimethylammonium bromide (CTAB) as the electrochemical intercalation agent. Through resistivity and Hall measurements, they discovered that the superconducting transition temperature (T
c) of TaS
2-(CTA
+)
x first increased and then decreased with the increase of x, displaying a dome-like behaviour. As a result, the T
c enhanced from 0.8 K in pristine 2H-TaS
2 to a maximum of 3.7 K in TaS
2-(CTA
+)
x (x value is 0.6). Furthermore, as the intercalation progressed, the charge carrier concentration of 2H-TaS
2 near T
c increased, indicating that the improvement of T
c was closely associated with the increase in the carrier concentration in the superconducting layers. Thus, the charge transfer from CTA
+ to the TaS
2 layers may be primarily responsible for the enhanced superconductivity in TaS
2-(CTA
+)
x. Recently, Li et al. developed a method called interlayer-space-confined chemical design (ICCD) that utilises the co-intercalation of organic molecules and Co
2+ ions to obtain a 2H-TaS
2 inorganic–organic molecular superlattice with quasi-monolayer properties
[26]. Magnetic measurements show that the TaS
2 hybrid superlattices exhibit enhanced superconductivity with the T
c of 3.8 K, which is higher than that of the previously reported monolayer TaS
2 [27]. They believe that the intercalation of organic macromolecules leads to electronic decoupling between the layers of TaS
2, which makes the bulk TaS
2 molecular superlattice exhibit quasi-monolayer characteristics, and this molecular superlattice structure has higher environmental stability than single-layer TaS
2. In addition, Wu et al. also used a gated intercalation method to insert lithium ions into a 2H-TaSe
2 flake, achieving a significant increase in T
c from 0.15 K to nearly 2 K
[28][29]. The electron doping and interlayer coupling introduced by lithium ion intercalation enhance the electron–phonon interaction, resulting in the enhancement of superconductivity. This demonstrates that interlayer chemical modulation is a facile and effective approach for controlling the electrical properties of 2D materials. It may open up a new avenue for discovering high-temperature superconductors.
More interestingly, interlayer chemical modulation can also induce superconductivity in non-superconducting materials, offering an alternative opportunity to create new types of superconductors. For example, superconductivity was successfully achieved in SnSe
2 by the intercalation of Co(Cp)
2 molecules, wherein the pristine SnSe
2 is a semiconductor
[30]. The intercalation of Co(Cp)
2 molecules into SnSe
2 was carried out by a solution-processed approach, forming a SnSe
2-Co(Cp)
2 organic–inorganic hybrid superlattice. Through resistivity and magnetic measurements, Li et al. observed that the superconducting T
c of the SnSe
2-Co(Cp)
2 superlattice was around 5 K, and it exhibited type-II superconducting behaviour. Angle-resolved photoemission spectroscopy (ARPES) measurements showed that the original Fermi energy level of SnSe
2 rose across the conduction band, suggesting that the superconductivity of SnSe
2-Co(Cp)
2 was due to an increase in the density of electronic states near the Fermi energy by virtue of the electrons transferring from the interlayer Co(Cp)
2 molecules to the SnSe
2 layers. In addition, the isothermal magnetisation curve (M−H curve) was an “S” shape, accompanied by a clear magnetic hysteresis loop, reflecting the ferromagnetic character in the SnSe
2-Co(Cp)
2 superlattice. Since the pristine SnSe
2 is nonmagnetic and Co(Cp)
2 molecules are paramagnetic, the ferromagnetic behaviour results in the formation of coupled magnetic interactions in SnSe
2-Co(Cp)
2 superlattices. Electron spin resonance (ESR) spectra showed that the spatial confinement effect weakened the coordination field of the Co(Cp)
2 molecules, giving rise to the spin-state transition from a low spin state to a high spin state, thereby resulting in ferromagnetism. This is the first freestanding coexistence of superconductivity and ferromagnetism constructed by non-superconducting and non-ferromagnetic materials of SnSe
2 and Co(Cp)
2 molecules, which results in strong interfacial electronic coupling. As a result, they found a unique correlated behaviour of the Kondo effect at low temperature. This coexistence of superconductivity and ferromagnetism in a SnSe
2-Co(Cp)
2 superlattice provides an efficient material platform to study the interplay of superconductivity and ferromagnetism, as well as application in superconducting spintronics.
Similarly, the superconductivity of bulk SnSe
2 could also be achieved by introducing charge into SnSe
2 via inserting tetraoctylammonium (TOA
+) cations
[31]. The SnSe
2-(TOA
+)
x superlattice was constructed by the electrochemical intercalation of TOA
+ into bulk SnSe
2. Resistivity and magnetic measurements showed that the SnSe
2-(TOA
+)
x superlattice exhibited a superconducting transition at 6.6 K, whilst it showed metallicity above the critical temperature. In addition, the temperature dependence and the magnetic-field dependence of the resistance are quite different for different magnetic field orientations, suggesting that the SnSe
2-(TOA
+)
x superlattice behaviour shows anisotropic superconductivity. Furthermore, below 10 K, as the temperature decreases, the exponential α of the voltage–current–power-law fitting gradually exceeds 1 and increases rapidly, indicating the occurrence of a Berezinskii–Kosterlitz–Thouless (BKT) transition, which evidences the quasi-2D superconducting properties of the SnSe
2-(TOA
+)
x superlattice. The red shift of the in-plane vibrational mode (E
g) and the out-of-plane vibrational mode (A
1g) of SnSe
2 and the increase in the E
g/A
1g intensity ratio in Raman spectra indicated that the intercalation of TOA
+ cations into SnSe
2-induced electron doping thereby led to superconductivity. Moreover, the interlayer spacing of SnSe
2 was enlarged after intercalation, which greatly weakened the interlayer coupling interactions. However, since the coherence length was larger than the interlayer spacing of the superconducting SnSe
2 layer, the SnSe
2-(TOA
+)
x superlattice exhibited a combination of 2D and 3D superconducting properties. More interestingly, the quantum Griffith singularity appeared in the SnSe
2-(TOA
+)
x superlattice. The random distribution of intercalated TOA
+ led to the non-uniform charge transfer in the system, which introduced charge disorder into SnSe
2 layers and triggered the quantum Griffith singularity in the superconducting–metal transition. The organic–inorganic superlattices provides a promising potential platform to investigate the interaction between disordered systems and quantum phase transitions.
As a widely studied layered material, MoS
2 can also exhibit superconductivity by intercalation with alkali or alkali-earth metals. For instance, Grigorieva et al. discovered that potassium (K) intercalation endowed two metallic 1T polytypes of MoS
2 with superconductivity
[32]. K intercalation was achieved by putting the platelet-shaped 2H-MoS
2 into a liquid ammonia solution of K metal, and 1T or 1T′ phase KMoS
2 was obtained depending on the intercalation times. Through energy-dispersive spectroscopy (EDS), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) characterisation, and magnetisation measurements, they found that, with the increase in K content, MoS
2 exhibited three kinds of superconducting phases, evolving from 2H-K
0.4MoS
2 to 1T-KMoS
2 and 1T’-KMoS
2. The superconducting T
c of these three phases were 6.9 K, 2.8 K, and 4.6 K, respectively, and all three phases showed type II superconductivity. The appearance of superconductivity is induced by the charge doping arising from the electron doping by K atoms. Interestingly, the well-separated superconductivity indicated the coexistence of local order and multi-structural phase. This may provide a basis for the construction of miniature Josephson junctions.
If the above sample is exposed to air, oxygen and water react with the intercalated K, causing the K to de-intercalate, which is similar to the results of previous studies
[33]. Surprisingly, they found that, after K
xMoS
2 was exposed to air for de-intercalation, the XPS results showed that the structural changes induced by K intercalation still existed, but the superconducting signal disappeared. This seems to indicate that the structural changes brought by the intercalation do not contribute to the generation of MoS
2 superconductivity. However, the inevitable multiphase-coexistence structure of MoS
2 in the above experiments makes it impossible to clearly study the contribution of a single structural change to the superconducting phase transition. Therefore, it is of great significance to obtain pure 1T or 1T′-phase MoS
2. Guo et al. discovered intrinsic superconductivity in pure-phase 1T’-MoS
2 produced by the structural phase transition of intercalated 2H-MoS
2 [34]. They fabricated highly crystalline LiMoS
2 by intercalating lithium uniformly into 2H-MoS
2 through modified high-temperature solid-state reactions. The charge transfer caused by the lithium between the layers induced a valence change of Mo in 2H-MoS
2, resulting in a uniform octahedral coordination structure. This highly ordered intercalation structure obtained by lithium intercalation is a prerequisite for the formation of 1T’-MoS
2. Following liquid-phase exfoliation in distilled water and lithium removal to obtain monolayer MoS
2, the Mo
3+ in the original octahedra was oxidised to Mo
4+, which transformed all of the diamond-like Mo-Mo chains into zig-zag chains and formed a long-range ordered distorted octahedral coordination structure, eventually yielding pure 1T′-MoS
2. The resistivity and magnetic measurements showed that 1T′-MoS
2 had type II superconducting properties and a superconducting transition at 4.6 K, which were absent in pure 2H phase. These results demonstrated that the twisted octahedral structure of 1T′-MoS
2 brings about extraordinary electronic interactions, which in turn trigger superconductivity. In addition, Fang et al. also successfully prepared pure bulk 1T-MoS
2 using the intercalation compound of LiMoS
2 and found that it also possessed intrinsic superconductivity with a superconducting T
c of 4 K
[35]. These findings provide a new avenue for exploring the mechanism of the superconducting transition and also showed that the structural phase transition induced by intercalation played a critical role in the study of the intriguing properties of various layered materials.
As discussed above, various structural phases, such as metastable 1T/1T′ phases and the stable 2H phase, have been discovered in transition-metal dichalcogenides (TMDs). The physical and electronic properties of TMDs are substantially governed by their structural phases. 2M-WS
2 is a new structural phase of TMDs recently reported. Pristine 2M-WS
2 exhibits a metallic behaviour and is a metastable p-type intrinsic superconductor. The superconducting T
c of 2M-WS
2 was 8.8 K, which was the highest intrinsic superconducting T
c in TMDs
[36]. Afterwards, Che et al. found a reversible superconducting–insulating transition in electron-doped 2M-WS
2 [37]. The 2M-WS
2 flakes were obtained by mechanical exfoliation, and the lithium ions were intercalated into the layers of 2M-WS
2 by applying gate voltage (V
g). Raman spectroscopy and Hall measurements revealed that the carrier concentration of the sample decreased from 9.05 × 10
21 cm
−3 to −6.04 × 10
20 cm
−3 as the V
g increased. Meantime, the resistance measurement showed that, with the increase in V
g, the superconducting T
c of the sample decreased, the resistance increased rapidly, and, finally, the sample transformed from a superconducting state to a completely insulating state, which can be clearly seen in the phase diagram. This is because the electron doping arising from lithium ion insertion leads to a significant decrease in the concentration of majority carriers (holes) in 2M-WS
2, thereby suppressing superconductivity and producing an insulating state. In addition, the decrease in T
c caused by a decrease in the majority carrier density is similar to a phenomenon recently found in Cu-based superconductors
[38]. In Cu-based superconductors, the change in the superconducting plane buckling angle caused by doping elements would affect T
c [38][39]. Therefore, it is speculated that the buckling angle on the 2M-WS
2 superconducting plane also might be adjusted by lithium ion doping, which affects its superconducting T
c. It is worth mentioning that the superconducting–insulating transition in 2M-WS
2 was reversible. By applying gate voltage to de-intercalate lithium ions, the original superconductivity of the sample could be restored. This indicates that gate-controlled intercalation is of great importance for the study of the complex electronic states induced by charge doping.
The formation of superconductivity by charge doping via gate electric field is usually effective for non-superconductors with low charge carrier density. However, in terms of materials with a high carrier concentration, it is difficult to control their phase transitions, especially the superconducting transition. For example, 1T-TaS
2 is a layered TMD with complex charge ordered states. Due to its high carrier concentration, it is difficult to control its phase transition by conventional gate modulation. To address this challenge, Yu et al. reported multiple phase transitions from Mott insulator to superconductor and then to metal in a high-charge-doped 1T-TaS
2 flake with the aid of a new intercalation method
[40]. The high-charge-doped 1T-TaS
2 was obtained by an ionic field effect transistor (iFET) based on 1T-TaS
2 in the case of gate-controlled lithium ion intercalation. On the one hand, they found that several charge density wave (CDW) phase transitions of undoped 1T-TaS
2 were modulated by the sample thickness. On the other hand, in the bulk limit (thickness > 10 nm), with the increase in V
g, the charge doping gradually increased, the Mott state in 1T-TaS
2 disappeared with the disappearance of the commensurate CDW (CCDW) phase, and a superconducting phase with T
c of 2 K appeared at the phase transition intersection of the nearly commensurate CWD (NCCDW)/the incommensurate CDW (ICWD) phase. In the quasi-2D (thickness < 10 nm) 1T-TaS
2, the CCDW phase did not exist due to the decrease in dimensionality, but the superconducting–metal transition still existed in the sample as in the bulk. Whilst in the 2D limit (~3 nm), the charge doping partially suppressed the original insulating state of the 2D sample, which may be due to disorder. In the charge-doped 1T-TaS
2, the microscopic phase separation at the junction of the NCCDW/ICDW phase transition may provide electron–phonon coupling for the formation of Cooper pairs, which may be the cause of superconductivity. It can be expected that gate-controlled-ion intercalation techniques provide powerful tools for interaction studies in the presence of extreme carrier doping and the design of 2D devices.
In general, the charge doping and interlayer coupling interaction caused by different intercalation elements are different, which will have different effects on the electronic properties of the host materials. However, Zhang et al. recently found almost identical superconducting T
c and electronic properties in a series of alkali and alkali-earth metal intercalated black phosphorus (BP) crystals
[41]. Their study found that the phosphorene layer could always maintain almost the same electron doping density as the intercalation of Li, K, Rb, Cs, Ca metal atoms. This extraordinary superconducting behaviour indicated a more complex interaction mechanism, which was very different from that of MCs. These findings illustrate the great potential of the intercalation method in the regulation of superconductivity and provide a new method for further studying the underlying mechanism of abnormal superconductivity.
2.2. Charge Density Wave Transition
As a quantum-ordered state, CDW phase transition often appears in low-dimensional materials with strong electron correlation interactions and has complex interactions with the superconducting phase, Mott insulating phase, and other ordered phases
[19][42][43]. Studying the CDW phase transition and its interactions with other electronic states are advantageous for the in-depth understanding of various complex correlation interactions, which is pivotal for the effective regulation of the electronic properties of low-dimensional materials and the application of low-dimensional electronic devices. At present, charge density waves have been found in a variety of material systems, especially in low-dimensional layered MCs, providing a versatile platform for studying the ordered origin of CDW phases. It is well known that layered compounds are susceptible to molecular or atomic intercalation, which is of paramount importance for the generation and regulation of CDW transitions.
The gated intercalation mentioned above is an effective and controllable ion intercalation method, which can be used to realise the modulation of CDW in 2D-MCs and study the microscopic interaction process caused by intercalation. However, for 2D materials, the electric double layer formed by the electrostatic gating effect may interfere with the influence of ion intercalation
[44]. Therefore, it is very important to identify the change in material properties caused by electric double layer and ion intercalation. Recently, Wu et al. developed a gating ion intercalation technology combined with lithography technology, studied the microscopic process of lithium ions intercalation 2D 2H-TaSe
2, and realised the gating intercalation modulation of CDW
[45]. By covering a part of the crystal surface with a photoresist, lithium ions were intercalated from the uncovered side, thus distinguishing the electrostatic gating effect from the intercalation effect. The similar gate voltage dependence of channel resistance in the covered and uncovered regions indicated the dominance of ion intercalation. Furthermore, they found that the hump of the R-T curve gradually disappeared with increasing V
g, indicating that the CDW transition was suppressed by the lithium ion intercalation. In addition, the anomaly that the channel resistance increased, then decreased, and then increased again illustrated the complexity of the intercalation process. They suggested that the modulation of CDW and resistance by lithium ion intercalation was not achieved by simple carrier doping but may have been closely related to the position distribution of lithium ions after intercalation.
Generally, the CDW in TMDs is considered to be generated by electron–phonon coupling near the Fermi surface
[46]. The superconductivity in most materials is also closely related to electron–phonon coupling. Therefore, there exists a competition between the CDW and the superconductivity in certain materials, which has also been confirmed by many studies. For example, Fang et al. discovered competition between superconductivity and CDW in the Na-intercalated crystal of Na
xTaS
2 grown directly through a solid-state reaction
[47]. The Na content in the intercalated 2H-TaS
2 was analysed by the energy-dispersive X-ray microanalysis (EDX) images of a scanning electron microscope (SEM), and both magnetic and resistive properties were measured. It was found that the CDW phase transition gradually disappeared with the increase in Na. Meanwhile, the superconducting phase appeared and increased with the increase in Na, and the maximum superconducting T
c reached 4.4 K. This result suggested a competitive relationship between CDW states and superconductivity. Superconductivity appeared mainly due to the increase in the density of states near the Fermi surface when the CDW state was suppressed.
Although the competition between superconductivity and CDW has been found in many materials, the physical mechanism is still poorly understood. The in-depth study of the competition between superconductivity and CDW can help
uspeople further understand the physical properties of CDW in 2D materials, guiding the regulation and functionality of materials. Like 2H-TaS
2, 2H-TaSe
2 also presents a typical CDW phase transition with intrinsic superconducting properties. The metal/ICDW phase transition, the ICDW/CCDW phase transition, and the CCDW/superconducting transition of 2H-TaSe
2 can occur with a decrease in temperature
[48]. Bhoi et al. found multiband superconductivity and direct competition between CCDW and superconductivity in a Pb-intercalated 2H-TaSe
2 crystal. 2H-Pd
xTaSe
2 crystals were synthesised through a solid-state reaction
[49]. The powder X-ray diffraction pattern (PXRD) showed that the c-direction spacing increased due to the intercalation of Pb. By measuring the resistivity and magnetic properties of Pb
xTaSe
2 with different x, they found that, with the increase in x, the transition temperature of the CCDW phase was rapidly suppressed, whilst the transition temperature of the ICDW phase decreased slowly. The superconducting transition temperature exhibits a dome shape with the increase in Pb. When x = 0.9, the superconducting temperature reaches a maximum of 3.3 K, and the CCDW phase collapses. This indicates that the superconducting phase seems to be directly competitive with the CCDW phase but not directly related to the ICDW. The results of specific heat measurement and calculation analysis indicate that the increase in superconductivity is due to the increase in effective electron–phonon coupling and density of states at the Fermi level due to the embedding of Pb. In addition, the electron-specific heat and the temperature dependence of the critical magnetic field indicate that there are multiple superconducting energy gaps in 2H-Pb
xTaSe
2. The different interactions of superconductivity with CCDW and ICDW indicate that the two types of CDW sequences in 2H-TaSe
2 seem to have different origins. This is similarly demonstrated by recent studies on 2D 2H-TaSe
2 intercalated by lithium ions. Wu et al. found that Fermi surface nesting and electron–phonon interaction in 2H-TaSe
2 do not seem to be the main causes of the ICDW order
[29]. The controlled lithium ion intercalation of 10 nm thick 2H-TaSe
2 was performed by an iFET device. The flake 2H-TaSe
2 was mechanically exfoliated from the 2H-TaSe
2 crystal prepared by CVT. Through Hall measurement at low temperature, it is found that the lithium-intercalated 2H-TaSe
2 has a negative Hall coefficient at low gate voltage (V
g) and exhibits a nonlinear dependence of the magnetic field. At higher V
g, the Hall coefficient becomes positive and has a linear relationship with the magnetic field. This indicates that the intercalation of lithium ions causes the topological change of the Fermi surface from a multi-pocket to a single-pocket in the ICDW phase. Moreover, the change in the Fermi surface topology and the suppression of the ICDW phase appear in different V
g, which indicates that the nesting of the Fermi surface is independent of the appearance of ICDW. The Raman spectra of 40 nm-TaSe
2 before and after intercalation also show that the intercalation has more influence on the characteristic phonon spectrum, which corresponds to the change in the Fermi surface topology rather than the inhibition of ICDW. In addition, through electrical transport measurement and calculation fitting, they found that the electron–phonon scattering in the intercalated sample decreases with the increase in V
g, which indicates that the combined effect of carrier doping and interlayer coupling caused by lithium ion intercalation leads to the enhancement of electron–phonon interaction. Meanwhile, the obvious increase in the superconducting temperature and the small change in the ICDW phase transition temperature also shows that the electron–phonon interaction in 2H-TaSe
2 may not be the main cause of ICDW order. Therefore, superconductivity and CCDW in 2H-TaSe
2 are regulated by the same type of electron–phonon coupling, whilst ICDW order may come from another type of electron–phonon coupling or other unknown mechanisms. In a word, these findings give a deep insight into the physical origin of the interaction between electron phase transitions in 2H-TaSe
2. It is also demonstrated that interlayer modulation can be an effective means to regulate the delicate balance between various electronic phase transitions.
2.3. Semiconductor-to-Metal Transition
In recent years, the study of the semiconductor–metal phase transition of 2D materials has attracted much attention and has demonstrated great prospects in the construction of integrated devices by using their homojunction structures. TMDs usually have an abundance of structural phases, which exhibit different electrical properties. For example, MoS
2 has a semiconducting 2H phase, a metallic 1T phase, and a 1T’ phase. It has been found that several of these structural phases can be transformed into each other by ion intercalation
[50], gated charge doping
[51], annealing
[52], or plasma induction
[53]. Cheng et al. fabricated an ideal odd-symmetric memristor using 1T-MoS
2 nanosheets
[54]. 1T-MoS
2 nanosheets were prepared by lithium ion intercalation. Through I-V measurements, the Ag/MoS
2/Ag device exhibited an alternating on–off state and an asymmetric pinched hysteresis loop, indicating that it is an asymmetric switch with memristive behaviour. Moreover, the 1T-MoS
2 nanosheets have stable memristive behaviour without perceptible changes after 1000 I-V cycles. In addition, an ideal memristor Ag/MoS
2/Ag/MoS
2/Ag with odd symmetry was obtained by combining two asymmetric devices. Its I-V curve is an odd-symmetric “pinched” loop with excellent stability. 1T-MoS
2 nanosheets exhibit memristive behaviour because the electric field can cause the lattice distortion of nanosheets. When the external electric field is large, Mo and S ions in the lattice shift, resulting in lattice distortion. This lattice distortion enhances electron delocalisation and leads to an increase in conductivity. In contrast, when a reverse electric field is applied, the displaced Mo and S ions are relocated, resulting in a decrease in conductivity.