5.3 Basic Material Response
159
5.3.4 Raman Signatures
Raman spectroscopy employs typical PL equipment and a very narrow-line excitation laser, which typically impinges the sample surface after passing a well-focusing
microscope objective. In Raman measurements, the back-scattered light from the
sample is collected and analysed in order to identify the Raman shifts due to scattering processes between incident photons (wavelength λ 0 ) and lattice vibrations.
Depending on the absorption or emission of a phonon’s energy by the photons, the
back-scattered photon either obtains more or less energy, respectively. This gives the
lines referred to as anti-Stokes (AS) and Stokes (S) lines, respectively. Their lines
appear at characteristic wave-numbers k = 2π(λ
−1
0 − λ
−1
(A)S ) which are related to
the phonon energies involved. The intensity spectrum as a function of the wavenumber acts as a fingerprint for a sample’s material.
Commercial Raman microscopes are widely used in the field of materials characterisation. In lab applications, self-built confocal microscopes for μ-PL spectroscopy
can be equipped with a Raman functionality. Since the modes are located very closely
around the excitation laser’s position, a very narrow linewidth of the laser light and a
very strong filtering of it with a sharp-edged filter are necessities in this spectroscopy
mode. Holographic and interference filters are commonly employed to suppress the
back-scattered Rayleigh line. Compared to the Rayleigh signal, the targeted Raman
modes are orders of magnitude weaker and long integration times are typically needed
to obtain sufficient signal-to-noise ratios in Raman spectra. Moreover, the optical
power has to be kept below the sample damage threshold and cannot be freely used
for intensity scaling. Further details about Raman spectroscopy can be found for
instance in laser spectroscopy textbooks such as [100].
In combination with 2D materials, Raman spectroscopy can provide useful information about the layer thickness and strain (e.g., used in [5, 7, 31]). When for instance
verification is needed that the data was obtained for sample locations which exhibit
monolayer or few-layer material, the characteristic modes can be analysed. This alloptical combination of PL and Raman spectroscopy can even compensate for the
lack of atomic-force microscopy (see Fig. 5.13 and cf. [7, 8]), which is alternatively
used to determine sub-nm height steps occurring at the transition from one layer
terrace to the next (e.g., used in [2, 3, 6]). For WSe 2 samples, a shift of the main
(monolayer material specific) Raman mode towards higher wave-numbers can be
clearly observed on various substrates with increasing layer number, i.e. thickness
[7, 31].
For instance, exfoliated WSe 2 monolayer samples feature one important Raman
peak around 252 cm
−1 with a weak shoulder slightly higher than 260 cm
−1 . However,
a bilayer or multilayer of this TMDC exhibits the weak shoulder below 260 cm
−1
[101]. The monolayer Raman mode for CVD-grown monolayers can be slightly
shifted to higher wave-numbers when compared to the exfoliated counterparts, which
can be an indicator for strain [102]. Such shift’s magnitude can be comparable or
larger than the shift induced by a transition from monolayer to bilayer TMDC.
159
5.3.4 Raman Signatures
Raman spectroscopy employs typical PL equipment and a very narrow-line excitation laser, which typically impinges the sample surface after passing a well-focusing
microscope objective. In Raman measurements, the back-scattered light from the
sample is collected and analysed in order to identify the Raman shifts due to scattering processes between incident photons (wavelength λ 0 ) and lattice vibrations.
Depending on the absorption or emission of a phonon’s energy by the photons, the
back-scattered photon either obtains more or less energy, respectively. This gives the
lines referred to as anti-Stokes (AS) and Stokes (S) lines, respectively. Their lines
appear at characteristic wave-numbers k = 2π(λ
−1
0 − λ
−1
(A)S ) which are related to
the phonon energies involved. The intensity spectrum as a function of the wavenumber acts as a fingerprint for a sample’s material.
Commercial Raman microscopes are widely used in the field of materials characterisation. In lab applications, self-built confocal microscopes for μ-PL spectroscopy
can be equipped with a Raman functionality. Since the modes are located very closely
around the excitation laser’s position, a very narrow linewidth of the laser light and a
very strong filtering of it with a sharp-edged filter are necessities in this spectroscopy
mode. Holographic and interference filters are commonly employed to suppress the
back-scattered Rayleigh line. Compared to the Rayleigh signal, the targeted Raman
modes are orders of magnitude weaker and long integration times are typically needed
to obtain sufficient signal-to-noise ratios in Raman spectra. Moreover, the optical
power has to be kept below the sample damage threshold and cannot be freely used
for intensity scaling. Further details about Raman spectroscopy can be found for
instance in laser spectroscopy textbooks such as [100].
In combination with 2D materials, Raman spectroscopy can provide useful information about the layer thickness and strain (e.g., used in [5, 7, 31]). When for instance
verification is needed that the data was obtained for sample locations which exhibit
monolayer or few-layer material, the characteristic modes can be analysed. This alloptical combination of PL and Raman spectroscopy can even compensate for the
lack of atomic-force microscopy (see Fig. 5.13 and cf. [7, 8]), which is alternatively
used to determine sub-nm height steps occurring at the transition from one layer
terrace to the next (e.g., used in [2, 3, 6]). For WSe 2 samples, a shift of the main
(monolayer material specific) Raman mode towards higher wave-numbers can be
clearly observed on various substrates with increasing layer number, i.e. thickness
[7, 31].
For instance, exfoliated WSe 2 monolayer samples feature one important Raman
peak around 252 cm
−1 with a weak shoulder slightly higher than 260 cm
−1 . However,
a bilayer or multilayer of this TMDC exhibits the weak shoulder below 260 cm
−1
[101]. The monolayer Raman mode for CVD-grown monolayers can be slightly
shifted to higher wave-numbers when compared to the exfoliated counterparts, which
can be an indicator for strain [102]. Such shift’s magnitude can be comparable or
larger than the shift induced by a transition from monolayer to bilayer TMDC.