Variable Angle Spectroscopic Ellipsometry Characterization of Graphene Oxide in Methanol Films

Abstract

It has been widely established that solvents modify the functional groups on the graphene oxide (GO) basal plane and, thus, modify its reactivity. Despite the increasing interest in GO films, a less studied aspect is the influence of methanol on the refractive index of GO films. Herein, the Variable Angle Spectroscopic Ellipsometry (VASE) technique has been used to characterize the optical response of GO in methanol films (0.4 mg/mL) dip-coated on glass substrates. The ellipsometric data have been modeled using a Lorentz oscillator model. We have found that the energy of the oscillator at ~3.9 eV for GO in water shifts to ~4.2 eV for GO in methanol films.

1. Introduction

The remarkable electrical, thermal, mechanical, optical, and long electron mean free paths properties of graphene make it compelling for various engineering applications [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18].

Despite considerable efforts from the scientific community, the large-scale availability of graphene samples is an significant limitation to the employment of the potentialities of graphene [19,20]. For this reason, while single layer graphene is unavailable for use in large-scale bulk applications, graphene-based materials, such as graphene oxide (GO), few layer graphene, multilayer graphene, and graphene nanoplatelets (GNPs) are innovative materials that can be exploited for several applications [21,22,23].

GO is a single-atomic-layered material made by the oxidation of graphite crystals, which are inexpensive and abundant [24].

GO, opposed to its non-oxidized counterpart, can form stable solutions in water and some organic solvents [25]. The solubility of GO in water and other solvents makes possible its uniform deposition onto different substrates in the form of thin films [26].

Until now, most studies reporting the solution-phase manipulation of GO have been carried out in aqueous media. Nevertheless, the preparation of GO dispersions in other solvents, mainly organic solvents, is worthwhile because it may crucially facilitate the practical use of this material [27].

Neklyudov et al. [28] demonstrated that the solubility of GO and the stability of the solutions are dependent not only on the solute and solvent cohesion parameters, but mainly on the chemical interactions at the GO/solvent interface. They found that in the systems, GO–water and GO–methanol hydrogen bonding is established between GO oxygen functional groups and solvent molecules. GO solvates with water have the strongest bonding, while the bonding with methanol is weaker.

Our work is based on the work of Pendolino et al. [29], who studied the structural change of GO in a methanol dispersion, demonstrating that GO functionalization is not completely attributable to the synthetic method but is dependent on the effect of the solvent, which modifies the functional groups on the graphene basal plane and, consequently, its reactivity. The article of Pendolino et al. [29] reports the influence of methanol on the functionalization of GO differently to water.

In this work, we focus on the fundamental issue regarding the refractive index of films dip-coated on glass substrates using a GO in methanol dispersion, with the intention of expanding the future practical uses of GO films. The films have been studied using VASE (Variable Angle Spectroscopic Ellipsometry).

In our previous work, we studied the refractive index of GO films on silicon, gold, silver and titanium substrates using VASE [30,31,32,33,34]. In addition, the optical properties of GO films using spectroscopic ellipsometry were studied by other research groups [35,36,37]. Despite the availability of literature on this subject, few studies have used VASE to determine the influence of methanol solvent on the refractive index of GO thin films. Our work intends to fill this gap, discussing the difference between the ellipsometric optical models of GO in water and of GO in methanol films.

Our results may help the manipulation and processing of GO films for several applications. It is important to study the optical properties of the system GO–methanol for the successful implementation of GO.

2. Materials and Methods

Corning glass substrates were cleaned in a mixture of sulfuric acid (H₂SO₄) and hydrogen peroxide (H₂O₂), known as “piranha solution”.

The samples were prepared using GO (Sigma Aldrich, St. Louis, MO, USA) 0.4 mg/mL in methanol.

The dip-coating process was employed to deposit GO thin films on glass substrates with a homemade apparatus at a speed of 3 cm/min.

Dip-coating consists of immersing a substrate into a tank containing coating material, removing the piece from the tank, and letting it drain. Then, the coated piece can be dried by force-drying or baking [38].

The optical characterization of the GO thin films on glasses was performed using VASE [39].

Spectra of the ellipsometric angles ψ and Δ were acquired by a M2000 F (Woollam Co., Lincoln, NE, USA) ellipsometer (Rotating Compensator Ellipsometer) in the (250–1000) nm wavelength range. The data were taken at 50°, 60° and 70° incident angles at room temperature.

The optical model and the best fitting values were calculated with WVASE32 (J.A. Woollam) application by means of the nonlinear Levenberg–Marquardt algorithm [40], which determines the minimum value of the Mean Square Error (MSE) [41].

Moreover, in the Supplementary Materials (Figure S1), the size distribution of the GO nanoparticles is studied using dynamic light scattering measurements.

3. Results and Discussion

Variable Angle Spectroscopic Ellipsometry Measurements

Glass substrates were modeled by means of the Cauchy dispersion law with the Urbach law [41] to characterize the absorbing properties of the substrates.

In Figure 1 the experimental and generated values of ψ (a) and Δ (b) for GO in methanol films made using dip-coating after 1 immersion in the tank are reported. An excellent agreement between the model generated data (red curves) and experimentally acquired data (circles) is observed.

GO films in methanol were modeled as the sum of Lorentz oscillators to keep consistency with the Kramers–Kronig relations [42]. The complex dielectric function is characterized by the relation:

$$\tilde{\mathsf{\epsilon }}\left(\mathrm{hv}\right)={\mathsf{\epsilon }}_1+\mathrm{i}{\mathsf{\epsilon }}_2={\mathsf{\epsilon }}_{\infty }+{{\displaystyle \sum }}_{\mathrm{k}=1}^{\mathrm{N}}\frac{{\mathrm{A}}_{\mathrm{K}}}{{\mathrm{E}}_{\mathrm{k}}{}^2-E^2-\mathrm{i}{\mathsf{\Gamma }}_{\mathrm{k}}\mathrm{E}}$$

(1)

where $\mathrm{E}$ is the energy of the incident photons, ${\mathsf{\epsilon }}_{\infty }$ is the real part of the dielectric function when $\mathrm{E}\to \infty$, ${\mathrm{A}}_{\mathrm{k}}$ is the strength expressed in ${\mathrm{eV}}^2$, ${\mathsf{\Gamma }}_{\mathrm{k}}$ is the broadening in eV end ${\mathrm{E}}_{\mathrm{k}}$ is the central energy of the k-th oscillator. ${\mathrm{A}}_{\mathrm{k}}$ also indicates the contribution of each oscillator k to the whole system.

Table 1. Lorentz oscillators parameters for graphene oxide in methanol films (this work) and for graphene oxide in water films (Ref [43]). Amplitude ${\mathrm{A}}_{\mathrm{k}}$ has unit of ${\mathrm{eV}}^2$, center energy ${\mathrm{E}}_{\mathrm{k}}$ and broadening ${\mathsf{\Gamma }}_{\mathrm{k}}$ have units of eV; d is the thickness of film in nm; the high-frequency dielectric constant ${\mathsf{\epsilon }}_{\infty }$ is dimensionless.

	GO in Methanol	GO in Water Ref [43]
$\mathbf{d}$ (nm)	5.5 ± 1.6	19 ± 1
${\mathsf{\epsilon }}_{\infty }$	1.62 ± 0.04	1.62 ± 0.04
${\mathbf{A}}_1\left(\mathbf{e}{\mathbf{V}}^2\right)$	25.1 ± 5.5	1.8 ± 0.2
${\mathsf{\Gamma }}_1\left(\mathbf{e}\mathbf{V}\right)$	1.3 ± 0.3	1.05 ± 0.09
${\mathbf{E}}_1\left(\mathbf{e}\mathbf{V}\right)$	5.8 ± 0.4	2.8 ± 0.1
${\mathbf{A}}_2\left(\mathbf{e}{\mathbf{V}}^2\right)$	6.8 ± 3.6	7.0 ± 0.2
${\mathsf{\Gamma }}_2\left(\mathbf{e}\mathbf{V}\right)$	2.3 ± 0.2	0.69 ± 0.05
${\mathbf{E}}_2\left(\mathbf{e}\mathbf{V}\right)$	4.2 ± 0.1	3.22 ± 0.01
${\mathbf{A}}_3\left(\mathbf{e}{\mathbf{V}}^2\right)$		2.3 ± 0.1
${\mathsf{\Gamma }}_3\left(\mathbf{e}\mathbf{V}\right)$  ${\mathbf{E}}_3\left(\mathbf{e}\mathbf{V}\right)$		0.59 ± 0.08 3.90 ± 0.02

shows the parameters from the best fit with a low MSE for GO in methanol films on glass substrates. The model provides excellent data fittings with a MSE ~9. Moreover, we report in the same table the Lorentz oscillators parameters for GO in water films that we published in Ref. [43] to make a comparison.

Figure 2 shows the dispersion laws, estimated by ellipsometry characterization in the (250–1000) nm wavelength range, of GO in methanol films.

The discussion of the parameters for GO in water can be found in Ref. [43].

The GO absorbance spectrum is very broad, with a peak at 5.4 eV and a smaller shoulder at 3.9 eV. The main peak is thought to be due to the graphene π/π* transition, while the shoulder has been attributed to the n–π* transition of C=O [44]. The peak at 5.8 eV for GO in methanol may be assigned to the broad peak of GO absorbance spectrum. The oscillator at ~3.9 eV for GO in water, which is assigned to the small shoulder of GO in the absorbance spectrum, shifts to ~4.2 eV for GO in methanol films. This shift may be due to the influence of methanol on the functionalization of GO differently to water, as reported by Pendolino et al. [29]. Therefore, the methanol solvent has an impact on the Lorentz oscillators parameters of GO films.

A large mapping system allows thin film measurements at every location on the film.

We have chosen to dip-coat only one half of the glass to highlight in the mapping the difference between the part of the glass where there is nothing and the other part covered by the film.

Figure 3 shows the best-match model calculated maps of film thickness for GO in methanol films dip-coated on glass substrates (1 immersion in the tank).

The thickness map in Figure 3 reveals that the central area of the GO films is quite disordered.

Using the same built optical model (Table 1), we report the mapping carried out on the films obtained using the same dispersion, (0.4 mg/mL of GO in methanol), but after 5 immersions in the tank (Thickness ~17 nm). After each immersion in the tank, we air dried the substrate with a drier.

Figure 4 shows the best-match model calculated maps of GO in methanol films thickness for 5 immersions in the tank. It is worth noticing that the Lorentz parameters reported in Table 1 do not change with the increase in the film thickness.

In both Figure 3 and Figure 4, it is possible to see a clear edge area between clean glass to that covered with the GO film.

The transmittance spectra of GO in methanol films in the (250–1000) nm wavelength range are shown in Figure 5 (1 immersion) and Figure 6 (5 immersions).

Lambert’s law is used to identify the optical absorption coefficient (α) [45]

$$\mathsf{\alpha }=\frac{1}{d}\ln \left(\frac{1}{T}\right)$$

(2)

where d is the film thickness and T is the transmittance. The absorption coefficient α is associated with the optical band gap in a direct-transition semiconductor material by the following equation [45]

$${\left(\alpha hv\right)}^2=A\left(hv-E_g\right)$$

(3)

where h is Planck’s constant; ν is the frequency of the incident photon; A is a constant that depends on the mobility of the electrons and holes in the material; E_g is the optical band gap. The optical band gap of GO in methanol films is obtained by Tauc plot. The linear part of the ${\left(\alpha hv\right)}^2$ curve is extrapolated toward the energy hν axis at ${\left(\alpha hv\right)}^2=0$. The determined band gap of GO in methanol films (1 immersion) is about 4.4 eV, as shown in Figure 7. The band gap for GO in methanol films (5 immersions) is about 4.3 eV, as shown in Figure 8. The found band gap values are in agreement with previous research [46].

The variation in the optical band gap energies reported in the literature is dependent on dissimilar processing factors such as time, degree of oxidation and chemical functionalization, which may influence structural, electronic and optical properties of GO [47].

4. Conclusions

Previous research has reported the influence of methanol on the functionalization of GO compared to water.

In this work VASE has been used to characterize the optical response of GO in methanol films dip-coated on glass substrates in the (250–1000) nm wavelength range. A Lorentz oscillator model was used to analyze the ellipsometric data. In particular, the oscillator at ~3.9 eV for GO in water films, which is assigned to the small shoulder of GO in absorbance spectrum, shifts to ~4.2 eV for GO in methanol films. This shift has potentially interesting consequences for future applications.

The optical band gap was determined by the extrapolation of Tauc plot and the extracted E_g are ~4.4 eV (1 immersion in the tank) and ~4.3 eV (5 immersions in the tank) in accordance with the existing literature.

Herein, we have shown that methanol influences the optical properties of the GO films studied using VASE.

Future research on this topic may involve other experimental techniques such as Raman spectroscopy and X-ray photoelectron spectroscopy to study the defect level in GO [48].