Raman spectroscopy is a powerful, non-destructive analytical tool extensively used to characterize two-dimensional (2D) materials such as graphene, transition metal dichalcogenides (TMDCs), and related layered structures. The technique provides critical insights into layer thickness, strain, doping, defects, and interlayer interactions by probing vibrational modes unique to these materials. Unlike bulk crystals, 2D materials exhibit distinct Raman signatures due to quantum confinement and reduced dimensionality, making Raman spectroscopy indispensable for their study.
Graphene serves as the archetypal 2D material for Raman analysis due to its well-defined vibrational modes. The most prominent features in its spectrum are the G peak near 1580 cm⁻¹, representing in-plane bond stretching of sp² carbon atoms, and the 2D peak (also called G') around 2700 cm⁻¹, arising from a second-order double-resonance process. The G peak position and intensity remain relatively stable, while the 2D peak is highly sensitive to layer number. Monolayer graphene exhibits a sharp, symmetric 2D peak that can be fitted with a single Lorentzian, whereas bilayer graphene shows a broader, asymmetric peak requiring four Lorentzian components due to interlayer coupling. As layer count increases beyond two, the 2D peak further broadens and shifts slightly but becomes less diagnostic for precise layer counting. The D peak near 1350 cm⁻¹ appears only in the presence of defects or edges, serving as a measure of disorder. The intensity ratio of D to G peaks (I_D/I_G) correlates with defect density, though interpretation requires care as different defect types influence this ratio differently.
TMDCs, such as MoS₂, WS₂, MoSe₂, and WSe₂, also exhibit strong layer-dependent Raman signatures. Unlike graphene, their Raman spectra involve both in-plane (E') and out-of-plane (A') vibrational modes. For MoS₂, the E' peak near 385 cm⁻¹ and the A' peak near 405 cm⁻¹ show a clear separation that increases with decreasing layer number due to enhanced dielectric screening in thicker layers. Monolayer MoS₂ exhibits the largest separation (~20 cm⁻¹), while bulk MoS₂ shows a smaller difference (~5 cm⁻¹). Additionally, the frequency of the A' mode softens with increasing layer count due to interlayer van der Waals interactions. The intensity ratio of these peaks can further assist in layer number determination. Similar trends are observed in other TMDCs, though exact peak positions vary depending on the constituent elements.
Strain engineering is another critical application of Raman spectroscopy in 2D materials. Both graphene and TMDCs exhibit peak shifts under mechanical deformation. In graphene, uniaxial strain splits the G peak into two components (G⁺ and G⁻) due to anisotropic stress, with shifts of approximately 10–30 cm⁻¹ per 1% strain depending on crystallographic orientation. Biaxial strain shifts the G peak uniformly without splitting. The 2D peak also shifts but with higher sensitivity (~60 cm⁻¹ per 1% strain). For TMDs, strain induces linear shifts in the E' and A' modes, typically at rates of 2–5 cm⁻¹ per 1% strain. Compressive strain redshifts these peaks, while tensile strain blueshifts them. The exact shift rates depend on material composition and the direction of applied strain relative to crystal axes.
Doping, whether intentional or unintentional, significantly alters Raman spectra. In graphene, electron or hole doping shifts the G peak position due to changes in the Fermi level. At low doping levels, the G peak hardens (shifts to higher wavenumbers) due to adiabatic phonon renormalization, but at high doping, it softens due to non-adiabatic effects. The 2D peak also shifts but is less sensitive to doping than the G peak. The G peak's full width at half maximum (FWHM) narrows with doping due to reduced electron-phonon coupling. For TMDCs, doping affects both peak positions and intensities. In n-doped MoS₂, the A' peak softens, while p-doping causes hardening. Doping can also activate normally forbidden modes or alter the relative intensities of existing peaks due to changes in electron-phonon coupling or symmetry breaking.
Beyond graphene and TMDCs, Raman spectroscopy is valuable for other 2D materials like hexagonal boron nitride (hBN) and black phosphorus (phosphorene). hBN exhibits a strong E mode near 1365 cm⁻¹, which shifts slightly with layer number but lacks the pronounced layer dependence seen in TMDCs due to its insulating nature. Phosphorene, with its anisotropic puckered structure, shows distinct Raman modes (A_g¹, B_2g, A_g²) that are sensitive to layer number and crystallographic orientation. The A_g¹ mode near 360 cm⁻¹ and the A_g² mode near 465 cm⁻¹ are particularly useful for thickness determination.
Raman spectroscopy also reveals interlayer interactions in heterostructures. When 2D materials are stacked, new modes may emerge due to interlayer vibrations or modified phonon dispersion. For example, in graphene-hBN heterostructures, the graphene G peak may shift slightly due to altered dielectric screening. In twisted bilayer graphene, additional peaks appear due to moiré-induced phonon replicas. Similarly, TMDC heterostructures exhibit modified Raman spectra depending on stacking angle and interlayer coupling strength.
Despite its strengths, Raman spectroscopy has limitations. Peak overlaps can complicate analysis in complex heterostructures or heavily doped samples. Laser-induced heating may unintentionally modify samples, particularly for materials with low thermal conductivity. Careful calibration and power optimization are necessary to avoid artifacts. Additionally, Raman cannot directly probe electronic properties like carrier mobility, requiring complementary techniques such as transport measurements for a complete picture.
In summary, Raman spectroscopy is an essential tool for characterizing 2D materials, providing detailed information on layer number, strain, doping, and interlayer interactions. Its non-destructive nature and high sensitivity make it indispensable for both fundamental research and industrial applications involving graphene, TMDCs, and related layered structures. Future advancements in hyperspectral Raman imaging and tip-enhanced techniques may further expand its capabilities for nanoscale analysis of 2D materials.