Raman signature Figure 1(a) represents the evolution of zone-center phonon E12gand A1gmodes of the monolayer 2H MoS2[0 Minute photo exposure] with a wavenumber difference of 17.32 cm-1. An increase in the frequency of the E12gmode from the reported monolayer 2H MoS2 is arising from either Coulomb interlayer forces or stacking induced changes in the intra-layer bonding via CHL-a intercalation17,18. However while compared the Raman spectra of 2H MoS217 [Fig 1(a)] and CHL-a/2H MoS2, we observed little broadening of E12g modes which indicates nonuniform shear strain distribution,17,18 as nochange in A1g mode is visible. Thus the presence of CHL-a affects the equilibrium lattice parameter, resulting in the reduction of the lattice symmetry from D3h. This can lift the degeneracy of the E12gmodes, hence the broadening and shifting of respective modes. This data also corroborates the observation of TEM showing lattice displacement of 0.04Å from pristine 2H MoS2 (3.16 Å)19 for CHL-a exfoliated MoS2 [Fig.1 (e)]. The presence of CHL-a can provide a significant substrate preventing random orientation, thus 2H Mos2can be formed with CHL-a assisted exfoliation [supporting information 1]. This has also been corroborated by XPS data in latersection20 where the shifting is maximized in CHL-a exfoliated sample.
Interestingly with increased light exposure, the A1g mode becomes sharp from exfoliated 2H counterpart, and E12g line width increases [Fig. 1 (b)] implying enhanced electron-phonon coupling (EPC) in Photo-excited CHL-a with the occupation of the anti-bonding states.Electron doping leads to the occupation of the bottom of the conduction band of MoS2, [also further corroborated via DFT calculation] making the bonds weaker.The A1g mode, which preserves the symmetry of the lattice, softens [Fig 1 (b)] because of the strengthening of electron-phonon coupling (λi) with doping. When the photo-exposure time is being elevated to 10 minutes the appearance of the reduced intensity of E12g peak from that of 5 minutes photo-exposure found, and additional peaks, 156 cm-1, 226 cm-1, and 330 cm-1 evolve, which can be explained in terms of the existence of a superlattice, as distorted octahedral structure with D3d symmetry17. The nature of the superlattice determines the particular point of the Brillouin-zone boundary that will be folded into the zone center and hence determines the phonons that will be observed after 10 minutes [the 156 cm-1, 226 cm-1, and 330 cm-1 peaks of Fig [1 (a)]. These peaks correspond to frequencies at the M point of MoS2,17 which appeared as a 2a0Xa0superlattice in TEM. This distorted superlattice formation also evident from TEM and discussed in the following section with TEM. The formation of 1T phase is observed, from a strong Raman band in 20 minutes photo exposure at 146 cm-1, attributed to Mo–Mo stretching vibrationsalong with. 219, 283, and 326 cm-1 supporting the TEM data18. We also found an E12g peak, which is anomalous as no peak corresponding to trigonalbipyramidal symmetry of the 2H phase is expected to be found. From XPS it is evident that the conversion efficiency is 85% [Fig. 2 (m)]remaining 2H phase could have contributed to this signal, but the intensity ratio, in that case, would have been the opposite. We found out that the A1g peak intensity reduces to such an extent due to electron-phonon coupling after photo-exposure that a small E12g peak corresponding to residual 2H peak becomes prominent. Due to the flattening of the A1g peak with electron donation discussed above, the E12g peak intensity becomes prominent. As a result, the intensity ratio cannot determine the degree of phase transformation which has been evaluated as 85% from the XPS data in the later XPS section.
The EPC of the Raman mode at the Γ-point of MoS2 exhibits a strong dependence on doping, similar to K-point phonons of graphene21. The electron-phonon coupling constant associated with a particular mode can be estimated from the Allen formula which relates the line width of the phonon mode to the dimensionless electron-phonon coupling constant21 :
where N(ϵf) is the electronic density of states at the Fermi level per eV per spin per unit cell (obtained from DFT calculation), gi is the mode degeneracy, with gi = 1 for an A1g mode and ωbi is the bare frequency in the absence of electron-phonon interaction.The full-width half maxima and frequency shift of two major bands A1g and E12g is represented in [Fig.1 (c) and (d) ] respectively We estimated the scaling of the EPC parameter in 20 min photo-exposure to be λi=6.084x10-4 [calculation in supporting information 2]22 which justifies the strong coupling and the flattening of the A1gpeak.
The distortion in a lattice with photo exposure in 20 minutes can also be related to strain which helps in lower frequency shifting of A1g mode. The presence of CHL-a in the intercalated state with 6th coordination as evident further in CHL-a Raman zone [supporting information 2; Fig.2s] can also induce strain23.
In TEM Data we found diffraction pattern (EDP) with sharp Bragg spots [6 fold symmetry] alternating with dim ones in the (100) orientation under observation along the [010] zone-axis direction, which is following the 2H MoS224. A notable alteration of crystal symmetry is evident, after the light exposure. The 2x2 superstructure reflection appears along with the symmetry (010) direction where the intensity of Bragg’s spot becomesprominent[Fig. 1 (f)]. At 10 min photo exposer in Figure 1(g), the basic diffraction spots are in different degrees of brightness. The Bragg spots within green lines are more intense than yellow lines, indicating a phase transition from 2H to 1T MoS2. At 20 min photo exposer,[Fig. 1 (h)] the electron diffraction pattern moves from hexagonal 2x2 superstructure to orthogonal 2x√3 superstructure analogues to sodium intercalated MoS2 24,25. This is due to enhanced electron transfer from CHL-a to MoS2 with time resulting in 2x√3superstructure with lattice expansion. The electron transfer and probable Dynamics of photo conversion have been substantiated by TCSPC and optical anisotropy data and well supported by DFT calculation.[supporting Information 3&4]
2.1. Quantitative and qualitative evidence of phase conversion from XPS:
The XPS spectra have been divided into three high-resolution regions corresponding to Mo 3d,S 2p, and N1s to obtain a clear view of the 1T phase evolution and heterostructure formation. The binding energy of Mo 3d in 2H-MoS2[Fig. 2 (a)]features two principal peaks at around 229.5 and 232 eV that correspond to Mo4+3d5/2 and Mo4+ 3d3/2 components, respectively for pristine 2H MoS2.26After photo exposure of 5 mint [Fig. 2 (b)]a new peak at 236.2 eV and a deconvoluted peak at 233.0 eV emerge in the XPS spectrum of MoS2/CHL-a, corresponding to Mo6+ 3d5/2 and 3d3/2 respectively which intensifies with photo- exposure time[Fig. 2(c)-(d)]. Down-shift of bonding energies also appears in the S 2p1/2 and S 2p3/2 peaks as compared to doublet peaks of 2H-MoS2[S 2p1/2 at 163.2 eV and S 2p3/2 at 162.2 eV in the core-level S 2p] [Fig. 2(e)] with increased photo-conversion indicating the emergence of 1T phase20.[Fig. 2(f)-(h)]
In the 2H MoS2/CHL-a hybrid [without photo-exposure], Mo 3p peak at 395.2eV partially overlaps with N1S spectra of CHL-a, at a binding energy of 398.1–398.9 eV [Fig. 2(i)] which is characteristic of the pyrrolidine nitrogen atoms of the porphyrinmacro-cycle27. We found a second peak (N-2), a shoulder, at 400 eV after photoexposure [Fig. 2 (j)], most likely due to the protonated nitrogen produced as a result of a small degree of de-metalization of CHL-a during its exposure to X-rays in the course of the experiment. After gradual photo exposure[Fig. 2 (k)-(l)] peak N-3 gets intense at around 408.4 eV with respect to 404 ev , indicating positively charged nitrogen of CHL-a as a result of the photo-electron transfer, the expulsion of the core electrons from the CHL-a nitrogen has become more difficult in this oxidized form thereby needing comparatively higher energy, i.e. 408.4 eV.in the photo- exposed CHL-a form. The positively charged nitrogen evolved is due to the transfer of non-bonding electrons on the nitrogen of CHL-a to MoS2 after photo-exposure and 1T MoS2 oxidized CHL-a gets stabilized. The presence of Mo6+ ion in 1T form also supports bandgap renormalization through Mo 4d state as mentioned further in PDOS of DFT.
From XPS spectra the amount of conversion has been evaluated as 85.2% after 20 minutes of photo exposure[Fig. 2 (m)] 27 by peak area contributions of 1T S 2p1/2 and S 2p3/2 peaks in S 2p high-resolution region of MoS2.
2.2. Transport properties:
To evaluate the macroscopic transport phenomenon along with the applicability of this material in device conformation wesubjected this material via simple drop-casting to construct two-terminal device architecture [details of device fabrication is in supporting information 5]. The 1T MoS2/CHL-a current-voltage curve is non-ohmic [Fig 3 (a)]. Another interesting phenomenon that is observed here is the temperature-dependent negative differential resistance (NDR) effect [Fig 3 (a)]. The current-voltage characteristics have been discussed in detail in the following section. The NDR effect has been evaluated in the later section by correlating MD simulation.
2.2.1. Current-voltage characteristics:
Considering the power-law dependence [IVβ+1] of current-voltage characteristics, we correspond this transport to non-Fermi liquid or Luttinger liquid behavior. Angle-integrated studies28 of quasi-2D organic metals have always reported a power-law dependence of the density of states, suggestive of Luttinger liquid behavior like in the quasi-1D organic metals29. The T and V dependence for tunneling into a 1D LuttingerLiquidvia Fermi-liquid metal contacts is given by
Where α=(g-1-1)/4, β=(g+g-1-2)/8 J0 is a constant, and the Luttinger parameter g=ϑF/ϑρis a fitting parameter that accounts for the voltage drop over the circuit30.
To validate thisLuttinger behavior a collapse diagram of the transport characteristic is obtained by plotting I/Tα+1 against eV/kT. Where α is the slop of zero-voltage conductivity against temperature [Fig 3 (b)]. We found the α =7.26 (and γ-1≈1000, β=12.007); Plotting the entire data set I /Tα+1 against eV/kT according to the Luttinger Liquid prediction.In our case, the data collapse quite well onto a single curve confirming Luttinger LiquidFig. 3 (b)31.
2.3. Mechanism of anisotropic transport via DFT:
. The two-dimensional (2D), highly dispersive interface states of π-conjugated organic molecules and a metal surface have been described theoretically and experimentally as strongly dispersive anisotropic, introducing some effective 1D potential32. Analogous to this, we tried to investigate the underlying mechanism of Luttinger transport with DFT calculation.
Our DFT calculations indicate several important differences between bare 1T MoS2 and 1T MoS2/ CHL-a [Fig. 3 (e) & (f)]. Compared to shallow electron pocket in bare 1T MoS2, 1T MoS2/ CHL-shows a deep electron pocket at Γ high symmetry point. The anisotropic dispersion of band on either direction of Γ point is evident which arises from the cross over at point-(ii)[ Fig. 3 (e), (f)] towards Γ-M direction and dispersion towards Γ-k direction at the point-(ii). In bare 1T MoS2the highly dispersive band arising from a spin-orbit coupling on either side of Γ forming a small electronic pocket at -250 mv via large energy splitting33 [Fig. 3 (e) point-(ii)]in small momentum space while compared to 1T MoS2. This indicates spin-orbit interaction and lesser inter-orbital interaction in 2H MoS2 since orbitals are deformed by the atomic bonding. On the flip side, 1T MoS2/CHL-a shows lesser dispersion and band crossing in larger momentum space [complying strong interorbital interaction on either side of symmetry points Γ [Fig. 3 (e)]. PDOS around Fermi level [Fig. 3 (c), (d)] points out that this evolved through the s3p and Mo 4d orbital with lifted degeneracy in DOS at “zero” bias [Fig. 3 (c), (d)]. Resulting in type-II Dirac points like dispersion and electronic pocket indicating strong finite range correlation [Fig 3 (f)point-(ii)]. This type of band which is not prominent in bare counterpart generally evolved with a topologically ordered state. Though it calls for further experimental verification this theoretical input gives the first glimpse of the probable existence of the exotic state in 1T MoS2/CHL-a VDWH. Doping induced via CHL-a interaction, as verified via the experimental section enhances the energy value of the electronic pocket around Γ point up to -400 mev specifically indicating strong finite range correlation. Another interesting feature arises around Γ point 1mev away from Fermi level. The two-fold band degeneracy in bare 1T MoS2 at 1mev from Γ to M and Γ to K points is lifted in 1T MoS2/CHL-a and a new set of orbital dispersion combination evolve via splitting and band lift off around 300 mev with strong asymmetry at Γ point which indicates strong spin-orbital coupling interaction34, this results in new inter-band asymmetry formation at 600 mev closer to the 499mev minima of electron pocket, opening a gap probably showing directionality for quasi-particle movement [Fig. 3 (f) point-(ii)]. Thus at Γ symmetry point highly dispersive band with enhanced spin-orbit coupling is prominent in the 1T MoS2/CHL-a system. This entire feature especially electronic pocket signifying quantization of carrier is one of the signatures of the Luttinger Liquid phenomenon we observed here. The 1D nature can be further illustrated via anisotropic effective mass distribution [Fig. 3 (g) (h) ]35. The calculated effective mass of 1T MoS2/CHL-a [for details, Supporting information 5] shows, charge carriers are in lower order that is 0.58 X m0 (where m0=free electron effective mass) for the electrons moving in the Γ to M direction than Γ to K direction (1.32 X m0) corresponding to highly dispersive and flat curve in the K space respectively. Lower dispersion could be attributed to the different confinement of quasi 1D structure while flattening indicates localization. The band flattening indicated by the α parameter depicting nonparabolicity is higher in Γ to K direction due to increased transport effective mass. The localization may have occurred due to the proximity of CHL-a π-electrons forming Van der waal interaction which also creates lattice changes in the supercell. The band edge dispersion states strong anisotropy with α value as 0.30434 and 1.00706 while compared to bare counterpart showing highly dispersive confinement of quasiparticle.
A clear visual of the finite contribution of each of the s3p and 4d orbital [Fig. 3 (d)] is evident while in bare 1T MoS2 the s3p and Mo 4d orbital contribute the degenerative higher value of density of state [Fig. 3 (c)]. The lifting of degeneracy in DOS around “zero” bias in 1T MoS2/CHL-a gives rise to asymmetric effective mass distribution corresponding to the shape of the orbital emphasizing directionality. The particle-hole asymmetry is prominent in 1T MoS2/ CHL-a. The asymmetry of the density of states is caused by the nontrivial interplay of the spin and charge degrees of freedom. The pseudo-gap-like structure shed light on the possibility of the directional movement of electrons. This can be related to the lift-off orbital degeneracy near the electron pocket above the Fermi level opening a gap at the Brillouin zone boundary which is evident in 1T MoS2/ CHL-a hole pocket [Fig. 3 (e), (f) (iii)].36.
2.4. Resonance tunneling (RT) in Luttinger liquid via Van der waal screening:
Another important feature we can find in the IV curve is the Negative differential resistance which gradually becomes diffusive with lower temperature. Considering the VDWH is a molecular interface we explain this owing to conformational heterogeneity of CHL-a to different temperatures as evident from MD simulation. Resonance tunneling occurs with the level alignment of CHL-a molecular orbital with that of 1T MoS2. In higher temperature 323K, we observed the onset energy is little negative or very close to zero with a sharp resonance peak, implying that tunneling direction is from 1T MoS2 to CHL-a [considering CHL-a intercalated within MoS2 layer]. As the temperature reduces [from 323K to 293K] the onset energy gradually shifts to positive energy with diffusive NDR peaks. The shifts may be correlated to the directional flip of electron transfer, that is electron is now tunneled from CHL-a to 1T MoS2 where higher energy is needed to bring the molecular orbital resonantly accessible to each other in the high-temperature range [Fig. 4 (a)]14. This is because the conformation of CHL-a at this ascending temperature range gradually becomes side by side attractive as evident from Fig. 4 (b) and (c). Now we correlate this phenomenon with the Luttinger Liquid resonance tunneling theory. [supplementary 4&5]
And found out [ Fig. 4 (d)] the reduction of repulsive scattering[g] as transmission probability (Γi) enhanced with temperature stemming from the scaling of Van der waal distance between CHL-a and 1T MoS2.This is preserved upon CHL-a conformational heterogeneity. At a temperature above 293 K the system shows NDR Fig. 4 (f)-(i) and below, the system conductance becomes linear [supporting information 6, Fig. 6S].
As the temperature increases the Van der waal distance also increases, but Van der waal energy of the whole system first decreases then increases, and finally goes down to a minimum [Fig. 4 (b)]. This is because our proposed system can act as an electrostatic double barrier configuration of polarizable CHL-a and 1T MoS2, hence system potential energy is predominated by effective interaction of Van der waal attraction energy which is caused by the formation of induced dipoles between polarizable materials which are much stronger on a short distance [between 4.5 to 9.5Ǻ]. On the other hand polarizability of CHL-a, in turn, depends upon the molecular orientation with refference to 1T MoS2plane. Head-to-tail interaction [molecular Z-axis perpendicular to xy-plane of MoS2] is attractive and side-by-side interaction [the deviation of Z-axis from perpendicular position xy-plane] becomes repulsive37. Starting from 263K as the distance increases with temperature, the configuration gets tilted and changes to head to tail [Fig. 4 (b) and (c)], hence energy is minimized and a local minimum is formed at temperature 273k. Further enhancement of temperature increases the distance, molecular orientation changes to the side by side repulsive configuration, and energy increases [283K to 293K]. It attains the maxima when torsion energy is 35.5 kcal/mol and temperature is 293k. After 293K the orientation becomes head to tail, with gradual reduction of Van der waal energy, finally finding a second minimum at temperature 323K with reduced torsional energy 31 kcal/mol [Fig. 4 (e)]. This energy minimization is caused by the almost perpendicular molecular z-axis orientation of CHL-a with 1T MoS2 on the xy-MoS2plane.
The cross-over point where the Van der waal distance increases with temperature and the Van der waal energy of the system decreases is around 300k. Experimentally around this point (crossover point), we started observing the NDR effect.
The NDR effect as well as the transmission probability directly related to Van der waal distance and the level alignment of the molecular orbital. Only the level alignment depends upon configuration which changes upon temperature resulting in a temperature-dependent NDR effect. Hence at a temperature above 293K, the configuration is such that Van der waal energy minimizes, but at the same time distance and transmission probability increased and level alignment maximizes, cumulatively giving rise to a stable structure having an NDR effect in high temperature. We observed from the simulation that the system became stabilized at temperatures 263K [point A] and 323K[point B].;
Though we find the same energy regime in low distance without NDR peak because Van der waal screening is higher due to hybridization in the lower distance with no tunneling probability and the molecules enter into the off-resonance situation. At a higher distance, the coupling between CHL-a and MoS2 reduces which in turn minimizes the screening effect and the system enters into a highly resonating regime 38. Thus this material can act in two energy regimes where energy is minimized first in low temperature where the Luttinger phenomenon is prominent and second high temperature where the NDR effect is profound [Fig. 4 (a) scheme].