Figs. 5.10 - 5.13 display the room-temperature experimental and best-fit calculated data of the triazine-doped and undoped graphene thin films prepared by CVD and ECE. The parameters of the stacked layer model used to fit the raw ellipsometry data are listed in Table 5.4. The independently measured experimental data at 60º and 70º incidence angles and the modeled curves are in good agreement. It is worth noting that the values of surface roughness of four thin films are one order of magnitude smaller than those deduced from AFM measurements. This discrepancy may arise from the surface contaminations as the AFM experiments were performed on the aged thin films. The optical constants derived from the ellipsometric parameters of Ψ and
∆ are shown in Figs. 5.14 – 5.15. We notice that the shape of refractive index dispersion of the triazine-doped and undoped graphene thin films for two different prepared methods are quite similar despite a constant 5% lower level difference for the doped thin films. Optical transitions can be identified in the spectrum of resonance features that appear in the extinction coefficient, with detailed analysis shown below.
80
Table 5.4: Parameters of a stacked layer model fit for the undoped and nitrogen-doped graphene thin films. All units are in nm.
CVD-undoped CVD-doped ECE-undoped ECE-doped Silicon
substrate
1 (mm) 1 (mm) 1 (mm) 1 (mm)
SiO2 297.5 297.4 297.4 297.5
Film 0.35 3.62 1.16 4.87
Roughness 0.08 0.29 0.83 1.42
Figure 5.16 shows the room-temperature optical absorption spectra of the triazine-doped graphene thin films compared with that of the undoped analog. As we can see the absorption gradually increases, manifests a sharp rise from 3 eV, reaches a maximum vale about 4.8 eV, and then showing the asymmetric line shape [95,96,98,136,137]. Such an asymmetry is fitted with the formula of Fano [138], I(ω) = I0 (q+ε)2/ (1+ε2), where ε = (ω-ω0) / γ, I0 is the normalized intensity, ω0 is the resonant (exciton) energy, γ is the effective linewidth, and q is the Fano asymmetry parameter. Our fitting curves are shown in Fig. 5.16. A list of fitting parameters is given in Table 5.5. From the results shown in Fig. 5.16 and the fit parameters in Table 5.5, three interesting optical properties in triazine-doped graphene thin films have been revealed by spectroscopic ellipsometry. First, the overall absorption profiles are similar, the observed 4.8 eV electronic excitation of the undoped graphene thin films prepared by CVD and ECE, assigned as exciton-dominated charge transfer feature [71,72], is blue shifted by 0.2 - 0.3 eV in the triazine-doped thin films. The blue shift of this resonant exciton excitation is a consequence of the exothermic nature of triazine molecule adsorption.
81
Table 5.5: Parameters of a Fano fit for the room-temperature optical absorption data.
Triazine is an electron-rich aromatic molecule due to the incorporation of N atoms in the aromatic ring, and some negative charges are transferred onto the graphene, leading to the modifications of its electronic structures. Second, the linewidth of the excitonic resonance for the triazine-doped graphene thin films broadens, which is attributed to the effects of defects and disorders. Third, the line shape of the optical absorption peak in our CVD-grown graphene thin films shows good agreement with that of Nelson et al. [98]. The Fano resonance can be interpreted as interference between the transition into the continuum and to the discrete state [138]. The negative value of q further indicates an interaction between the exciton and an electronic continuum, extending over an energy interval below the resonance frequency of the electronic transition. Notably, the line shape of the ECE-grown graphene thin films displays less asymmetric. The 1/q ~ 0 feature assumes a symmetric line shape and recovers Lorentz model.
Such behavior could be due to the changes of the charge distributions in the graphene thin films prepared by different growth methods.
Sample ω0 (eV) γ (eV) I0 q
CVD-undoped 4.81±0.03 5.39±0.02 20.5±0.02 -1.90±0.02
CVD-doped 5.12±0.03 5.49±0.02 20.6±0.02 -1.90±0.02
ECE-undoped 4.80±0.03 6.19±0.02 32.5±0.02 -138±0.02
ECE-doped 5.03±0.03 6.21±0.02 30.5±0.02 -138±0.02
82
When the sample is heated from 200 K to 350 K, we do not observe any sharp changes in absorption, but rather a continuous evolution. Figs. 5. 17 – 5.18 and Figs. 5. 19 – 5.20 illustrate the peak energy, damping, normalized intensity, and asymmetry parameter of the excitonic resonance of the undoped and triazine-doped graphene thin films as a function of temperature. It becomes clear that with increasing temperature the excitonic resonance shows a shift of the peak position to lower energies. The lattice thermal expansion could be responsible for this redshift.
Further temperature dependence of x-ray diffraction experiments currently under way could help to find the correlation. The linewidth does not exhibit clearly temperature-dependent trend. The oscillator strength of the CVD-grown graphene thin films increases when the temperature is raised. It is likely due to a change of local symmetry breaking at high temperatures. Further details of theoretical investigations are needed to confirm this speculation.
5-3 Summary
In summary, the THz absorption and spectroscopic ellipsometry spectra of the triazine-doped graphene provide important insight into the charge dynamics and electronic structures in this system. First, the THz conductivity consists of free carrier absorption at zero frequency and an additional finite frequency peak at about 155 cm-1 for CVD-doped graphene thin film. As the temperature is lowered, the Drude plasma frequency (~ 21 THz) decreases, whereas the carrier relaxation time (~ 13 fs) does not show much temperature variation. These scenarios show the semiconducting behavior of the CVD-doped graphene thin film. Most importantly, a finite-frequency peak, which coexists with a Drude contribution, is likely associated with the
83
significant disorder induced by triazine doping. A comparison of our measured conductance spectra and the theoretical predictions in turn illuminates the importance of the several scattering mechanisms present in the material. Second, the Drude plasma frequency of the ECE-grown graphen thin films is three times larger than that of CVD-grown ones. In contrast, the carrier relaxation time of the ECE-grown graphen thin films is shorter than that of the CVD-grown samples. Third, the optical absorption spectrum of the CVD-grown thin films exhibits an asymmetric Fano resonance in the ultraviolet frequency region. The excitonic-dominated charge transfer band in the triazine-doped graphene thin films shows a blueshift in comparison with that of undoped analog, reflecting changes in electronic structures. Furthermore, the linewidth of the excitonic resonance for the triazine-doped graphene thin films broadens, which is likely due to the effects of defects and disorders. Additionally, the line shape of the ECE-grown graphene thin films displays less asymmetric. Such behavior could be attributed to the changes of the charge distributions in the graphene thin films prepared by different growth methods.
84
Fig. 5.1: AFM images of (a) CVD-undoped thin film, (b) CVD-doped thin film, (c) ECE-undoped thin film, and (d) ECE-doped thin film.
(a) (b)
(c) (d)
20 nm 20 nm
0 nm 0 nm
0 nm 0 nm
10 nm 10 nm
85
Fig. 5.2: Raman-scattering spectra of the CVD-undoped, CVD-doped, undoped, and ECE-doped graphene thin films.
1000 1500 2000 2500 3000
CVD-doped
Raman Shift (cm-1)
Intensity (arb. units)
CVD-undoped
ECE-undoped ECE-doped
86
Fig. 5.3: Room-temperature optical transmission spectra of a bare silicon substrate, CVD-undoped, and CVD-doped graphene thin films.
Fig. 5.4: Real part of the room-temperature optical conductivity spectrum of the CVD-doped graphene thin film (solid line). The various terms in the fits are also shown (dashed line): the Drude band and two Lorentz oscillators.
0 2 4 6 8
87
Fig. 5.5: Temperature dependence of the Drude plasma frequency and the carrier relaxation time.
Fig. 5.6: Temperature dependence of the Drude conductivity.
0 100 200 300
3500 4000 4500 5000
Temperature (K)
pD (cm-1 )
10 11 12 13 14
Relaxation time (fs)
0 100 200 300
500 600 700 800 900
Drude (-1 cm-1 )
Temperature (K)
88
Fig. 5.7: Measured optical sheet conductance spectra (solid line) of the undoped and CVD-doped graphene thin films at 10 K along with the theoretical spectrum (dashed line).
0 2 4 6 8
0.0 0.5 1.0 1.5
2.00 10 20 30 40
CVD undoped CVD doped Photon Energy (meV)
G 1[units of (/2)e2 /h]
Frequency (THz) 10 K
Theory
89
Fig. 5.8: Room-temperature optical transmission spectra of a bare silicon substrate, CVD-undoped, CVD-doped, ECE-CVD-undoped, and ECE-doped graphene thin films.
Fig. 5.9: Real part of the room-temperature optical conductivity spectrum of the CVD-undoped, CVD-doped, ECE-undoped, and ECE-doepd graphene thin film (solid line). The fits term with an Drude band are also shown (dashed line).
0.0 0.5 1.0 1.5 2.0 2.5 3.0
90
Fig. 5.10: Room temperature experimental (symbols) at 60º and 70º incidence angles and fitted (dashed lines) values of ellipsometric parameters of Ψ and ∆ of the undoped graphene thin films prepared by CVD.
Fig. 5.11: Room temperature experimental (symbols) at 60º and 70º incidence angles and fitted (dashed lines) values of ellipsometric parameters of Ψ and ∆ of the triazine-doped graphene thin films prepared by CVD.
91
Fig. 5.12: Room temperature experimental (symbols) at 60º and 70º incidence angles and fitted (dashed lines) values of ellipsometric parameters of Ψ and ∆ of the undoped graphene thin films prepared by ECE.
Fig. 5.13: Room temperature experimental (symbols) at 60º and 70º incidence angles and fitted (dashed lines) values of ellipsometric parameters of Ψ and ∆ of the triazine-doped graphene thin films prepared by ECE.
92
Fig. 5.14: Room temperature refraction index for the triazine-doped graphene thin films compared with those of the undoped thin films prepared by CVD and ECE.
Fig. 5.15: Room temperature extinction coefficient for the triazine-doped graphene thin films compared with those of the undoped thin films prepared by CVD and ECE.
0 1 2 3 4 5 6 7
93