Effect of hydrothermal and chemical treatment on the optical and electrical properties of reduced graphene oxide deposited on ITO glass

In order to tune the optical and electrical properties of graphene oxide, graphite was highly oxidized to yield highly oxidized graphene oxide [28], the resulting HOGO was reduced by hydrothermal [30] and hydrazine method seperately and then spray deposited. The effect of different reduction approach is utilized in tuning the optical and electrical properties of the films The optical and electrical properties of the film for each RGO sample is determined and compared with literature. For the hydrothermal treatment, HOGO solution, (3 mg ml⁻¹) was placed in a 60 ml teflon lined autoclaved. The autoclave was placed in a beaker containing silicon oil to allow for good thermal contact

and maintained at 180 °C for 24 h under constant stirring using an electric hotplate (figure 1). Thereafter, it was allowed to cool to room temperature and the synthesized RGO was filtered using a filter paper and air dried. This is referred to as RGO_W. A second RGO sample was obtained via the chemical method by reducing HOGO using hydrazine monohydrate. HOGO, (300 mg) was dispersed in 100 ml of distilled water and stirred continuously at ambient temperature for 30 min. Then, 0.2 ml of Hydrazine Monohydrate was added to the mixture of HOGO and distilled water and placed in oil bath maintained at 80 °C for one hour. The obtained RGO was washed with deionized water and filtered with a filter paper to remove residual N₂H₄. This is referred to as RGO_H.

Zoom In Zoom Out Reset image size

The thin film of RGO was obtained by deposition using spray pyrolysis technique (SPT) arrangement as shown in figure 2. ITO glass was utilized as the substrate while the carrier gas used was compressed air. The ITO glass did not require any additional cleaning apart from wiping the surface with a clean and new handkerchief in order to retain the conductivity on the glass. RGO powder, (0.25 g) was dispersed in 15 ml Ethylene Glycol and deposited on the substrate at a temperature of 300 °C. The distance between the nozzle and substrate was maintained at 30 cm for uniform deposition. A dark colour on the substrate after deposition showed that the precursor was successfully deposited. The film thickness of the RGO samples were estimated to be 100 nm.

Zoom In Zoom Out Reset image size

Infrared spectrum of RGO was obtained using Fourier Transform Infrared Spectrophotometer (SHIMADZU IRTracer-100). The surface morphology characterization was carried out using Zeiss Ultra Plus 55 Field Emission Scanning Electron Microscope (FE-SEM) while the elemental composition was carried out with the Energy Dispersive Xray (EDX) setup. Optical absorbance of the RGO measured using Jenway UV-Visible Spectrophotometer (Model 6405). The Raman Spectroscopy analysis was carried out using micro-Raman Spectrometer (Jobin-Yvon Horiba T64000) equipped with a triple monochromator system to eliminate contributions from Rayleigh scattering while the electrical property of the film was carried out using a four point probe set up.

Figures 3(a)–(d) shows the FTIR spectra obtained for highly oxidized graphene oxide (HOGO), reduced graphene oxide obtained by; chemical method using hydrazine (RGO_H), hydrothermal method (RGO_W) and solvothermal method (RGO_S). In figure 3(a), the HOGO spectrum show bands at 3423 cm⁻¹ for O–H stretching, at 3064 cm⁻¹ and 3236 cm⁻¹ the bands are due to sp² C–H stretching frequency, at 2881 and 2804 cm⁻¹ the bands typify sp³ C–H. Between 1040 cm⁻¹ and 1170 cm⁻¹ are peaks that are attributed to the alkoxy C–O stretching vibration. All these bands are typical of GO and are similar to what we obtained in our earlier work [28].

Zoom In Zoom Out Reset image size

In figure 3(b), RGO hydrazine (RGO_H) has bands for O–H at 3429 cm⁻¹; presence of sp³ C–H band at 2854 cm⁻¹ and 2924 cm⁻¹; C–C stretching at 1564 cm⁻¹ and epoxides group at 1224 cm⁻¹. C–O stretching vibrations strongly decrease to sharp vibration after reduction of HOGO to RGO. Also, the peak at 860 cm⁻¹ present in graphite flakes but absent in HOGO has been restored and this is attributed to aromatic C–H bands. The broad peak between 900–1200 cm⁻¹ present in HOGO has disappeared and resulted in a stretching vibration in RGO_H.

The spectrum for RGO obtained by the hydrothermal method (RGO_W) is shown in figure 3(c). In this spectrum, bands at 3392 cm⁻¹ is due to O–H stretching frequency, sp³ C–H stretching frequency occurs at 2854 cm⁻¹ and 2920 cm⁻¹. At 1398 cm⁻¹, there is a bending vibration frequency and the intensity of the C–O band in HOGO has been reduced to a lower intensity at 1039 cm⁻¹ in RGO_W.

The spectrum for RGO_S obtained by the solvothermal method using ethylene glycol as solvent is shown in figure 3(d). The intensity of C=O bond increases in the spectrum, however, there are noticeable decrease in the peaks of other functionalities within the region 900 cm⁻¹–1400 cm⁻¹. However, because the hydrothermal and solvothermal reduction involved extensive heat treatment, the intensity of C=C peak at 1616 cm⁻¹ in HOGO showed a decline in RGO_W and RGOs unlike in RGO_H. The presence of water in the reduction process via the hydrothermal method strengthens the C–OH peak occurring at 1398 cm⁻¹ unlike in RGO_H and RGOs where there seems to be a near disappearance of the C–OH peak. The C=C band typical at 1570 cm⁻¹ was observed to occur in RGO_W and between 1550–1565 cm⁻¹ in RGOs and RGO_H.

The microstructure and surface analysis of RGO_H (hydrazine) and RGO_W (hydrothermal) were carried out using FE-SEM and are shown in figure 4. RGO_W was utilized for FE-SEM and other analysis since our FTIR result has shown that hydrothermal reduction favoured the reduction process compared to solvothermal process. The micrograph of RGO using hydrazine as shown in figures 4(a) and (b) shows that the sheets converged and cleaved to form networks which gave rise to sharp and interlinked three dimensional graphene sheets forming a porus arrangement [30]. The RGO sheets in figures 4(a) and (b) appear like folded structure [31]. Furthermore, RGO hydrazine exhibits typical wrinkled structure that causes sheet folding.

Zoom In Zoom Out Reset image size

The FE-SEM microstructure of RGO obtained via hydrothermal method as shown in figures 4(c) and (d) shows a layered structure with crumples and sheet which overlay rather than aggregating [32]. Also, RGO hydrothermal typify prominent overlapping and increase in the number of sheets compared to RGO hydrazine. This may be attributed to the material transformation that took place in the autoclave at high temperature and pressure during synthesis.

The microstructure of RGO hydrothermal gave a spongy outlook at lower magnifications as shown in figure 4(c). However, at higher magnifications figures 4(b) and (d), the FE-SEM micrograph of both RGO samples amplified the wrinkles and folding that were not so obvious at lower magnifications.

The EDX spectrum of HOGO, RGO (Hydrazine) and RGO (hydrothermal) are shown in figures 5(a)–(c). The defined spectrum showed peaks corresponding to carbon and oxygen. In the case of HOGO, the percentage of carbon was 22.0% while oxygen was 78.0%. In RGO (Hydrazine), carbon increased to 29.8% while oxygen decreased to 70.2% while in RGO (hydrothermal), carbon accounted for 37.3% and oxygen 62.7%. The extended period of oxidation brought about by our synthesis method of HOGO introduced more oxygen functionalities to the HOGO thus indicating a high degree of oxidation [28]. This accounts for higher percentage of oxygen in HOGO and our RGOs when compared to literature [13, 33]. The presence of carbon and oxygen is typified in the spectrum, however, RGO from both hydrazine and hydrothermal method showed an increase in the amount of carbon and a decrease in oxygen content thus indicating that both chemical and hydrothermal methods have successfully reduced HOGO.

Zoom In Zoom Out Reset image size

Figure 6 shows the Raman spectra of HOGO, RGO hydrazine and RGO hydrothermal. The Raman spectra of the samples were compared with the signatory broad D and G peaks centered at a wavenumber of 1354 and 1598 cm⁻¹ respectively. The characteristics peak of 1354 cm⁻¹ for D-band and 1598 cm⁻¹ for G-band were observed for HOGO. In RGO hydrothermal, these peaks red shifted to 1351 cm⁻¹ and 1595 cm⁻¹ for D-band and G-band rspectively. A decrease in the D and G band to 1346 cm⁻¹ and 1592 cm⁻¹ respectively was also obtained for RGO hydrazine. The decrease in the wavenumber of the D peak towards a lower wavenumber is consistent with previous studies [34, 35]; indicating a successful reduction of graphene oxide and a restoration of the sp2 network .

Zoom In Zoom Out Reset image size

The structural properties of RGO can be observed and interpreted from the relative intensity of the D and G peak of the Raman Spectra [36]. In the RGO hydrothermal sample, a very disordered graphitic structure is observed by a pronounced D-peak. A conspicuous D band in RGO hydrothermal indicate disorder in the Raman of the RGO springing up from defects associated with vacancies, grain boundaries and amorphous carbon species [17]. This is complemented by an increase in the I_D/I_G ratio (from 1.01 to 1.10) of RGO hydrothermal thus accounting for a decrease in the sp² domain size which also indicates an increase in the surface defects [18]. This suggests that new or more graphitic domains are formed with increased restoration of the sp² cluster number thus indicating good reduction efficiency of the hydrothermal method, [36] and excellent recovery of the conjugated networks in graphene. The prominent D and G peak as well as the increase in relative intensity in RGO hydrothermal typify small crystal sizes were obtained.

The hydrazine treatment of HOGO offers inconsequential changes in the I_D/I_G ratio of RGO. The I_D/I_G ratio of highly oxidized graphene oxide was reduced from 1.01 to 0.99 for RGO Hydrazine.

Figure 7 gave the spectrum of absorbance against wavelength for HOGO, RGO_W, and RGO_H. In HOGO, a major peak typical of GO is observed at 230 nm which is due to π–π* transition of C=C in amorphous carbon arrangement [37, 38] A shoulder peak is slightly observed at ∼290 nm for HOGO which suggests that there exists a bond excitation value corresponding to n -π* transition of the C=O chromophores for aldehyde and ketone molecule [28, 39]. The red shift in the peak position of our RGO to 240 nm is due to reduction of HOGO and the partial restoration of sp² network after removing certain functionalities [31, 38, 40]. The high degree of oxidation in our graphite brought about by our synthesis method has impacted on the degree of reduction of graphene oxide; hence, the lower degree of red shift in wavelength of our RGO compared to literature [34, 41].

Zoom In Zoom Out Reset image size

In figure 8, the transmittance of HOGO, RGO hydrazine and RGO hydrothermal are shown. RGO hydrothermal absorbs light better than RGO hydrazine and HOGO in the visible region of the EM radiation. However, HOGO typified better transmittance with a transmittance of eighty percent at 600 nm and above. This is because there are few bonds to absorb the incident light in the sample at that wavelength. However, at wavelength below 600 nm, RGO hydrothermal and RGO hydrazine proved to allow more light to pass through them than HOGO.

Zoom In Zoom Out Reset image size

The absorption coefficient is related to the photon energy by the Tauc relation [42]. This equation as shown in equation (1) has been used to evaluate the energy band gap of our RGO film.

$$\begin{eqnarray}&&{\left(\alpha hv\right)}^{\displaystyle \frac{1}{n}}=\beta \left(hv-{E}_{g}\right)\end{eqnarray} \tag{ 1 }$$

Where β is a constant, h is Planck's constant, v is frequency, ${E}_{g}$ is the energy band gap and n is the power factor of the transition mode. For direct allowed band gap, $n=\tfrac{1}{2}$ and a graph of absorption square against photon energy is plotted.

For our thin film samples, the absorption coefficient is given by

$$\begin{eqnarray}&&\alpha =\displaystyle \frac{1}{d}\,\mathrm{ln}\,\displaystyle \frac{1}{T}\end{eqnarray} \tag{ 2 }$$

where d is the thickness of the film, T is the transmittance which is equal to 10^{− A} and A is the absorbance.

Figure 9 shows the plot of absorption square (α²) against the energy of the photon for HOGO, RGO_H, and RGO_W. A band gap of 3.0 eV, 2.4 eV and 2.1 eV respectively were obtained by extrapolating the linear part of the graph. The reduction in bandgap of RGO compared to HOGO may lead to an increase in conductivity and improved optical properties. Since RGO_W has a lower band gap compared to RGO_H, RGO_W might be more conductive than RGO_H.

Zoom In Zoom Out Reset image size

The electrical characterization of RGO film was carried out using a four point probe. The sheet resistance was calculated using equation (3)

$$\begin{eqnarray}&&{R}_{s}=K\displaystyle \frac{V}{I}\end{eqnarray} \tag{ 3 }$$

Where K is a constant which is equal to

$$\begin{eqnarray}&&K=\displaystyle \frac{\pi }{\mathrm{ln}\,2}=4.5324\end{eqnarray} \tag{ 4 }$$

The sheet resistance of RGO_H and RGO_W thin film were determined to be $3\times {10}^{6}$ Ω sq⁻¹ and $0.9\times {10}^{6}$ Ω sq⁻¹ respectively. The resistivity of RGO_H and RGO_W thin films were calculated using the sheet resistance and the thickness of the films. The thickness of the thin films using the color code chart [43] was estimated to be 100 nm while the resistivity of the film was calculated using equation (5)

$$\begin{eqnarray}&&\rho ={R}_{s}t\end{eqnarray} \tag{ 5 }$$

where $\rho$ is the resistivity of the film, and t is the thickness of the thin film. The resistivity of RGO_H and RGO_W was calculated to be 0.3 Ω m and 0.09 Ω m respectively. The conductivity of RGO_H and RGO_W was also calculated to be 3.3 S m⁻¹ and 11.1 S m⁻¹ respectively. The increase in the conductivity of RGO_W compared to RGO_H is as a result of better restoration of the sp² carbon chain which were recovered during the reduction of HOGO. This creates more conductive pathways for carrier transport in the carbon plane of the RGO sheet [29]. However, the disparity in value of electrical conductivity obtained in our work compared to literature [26, 27] can be attributed to factors such as high degree of oxidation in our graphene oxide, type of reductant used and the difference in substrate used for deposition. This also confirm that the recovery of conductivity in highly oxidized graphene oxide is difficult owing to increased oxygen content. Our results have shown that the optical and electrical properties of RGO deposited on ITO glass will yield better result when the degree of oxidation of graphene oxide is moderate. This observation is consistent with literature where it was proven using density functional calculations that incomplete oxidation of GO is thermodynamically favorable compared to complete oxidation [12, 44].