Introduction

After the discovery of graphene by Geim and Noveselev in 2004, tremendous research was carried out on the field of this thinnest and flattest material that could ever be in the universe1,2,3,4. Graphene has offered a unique two-dimensional (2D) sp2-hybridized structure and great properties such as high mechanical flexibility, superior electrical and thermal conductivities, large specific surface area, and high chemical stability5,6,7,8,9,10,11. Because of these characteristics, graphene has various applications including in supercapacitors12,13, photovoltaics14,15,16, fuel cells17,18,19, sensors20,21 and nanofluids22,23. Moreover, 2D crystal structure of Graphene can make it more popular in order to load diverse materials or form various composites that enhance the beneficial features of both graphene and the added components24. For instance, it is observed that graphene/metal oxide composites have shown higher performance for energy storage25,26,27,28 and electrochemical detection29,30,31,32 in comparison with individual graphene or added components. Furthermore, research on functionalized graphene has shown that graphene composites exhibit remarkable nonlinear optical (NLO) responses33. In this respect, inorganic metal oxides can be good candidates for combining with graphene. Recently, they have attracted significant attention because of their wide utilization in catalysis, water purification, hydrogen production, lithium-ion batteries, and transparent electronics34,35,36,37,38,39. For example, Zinc oxide (ZnO) is an inorganic metal oxide with a wide band gap of 3.37 eV40,41 and a large exciton binding energy at room temperature (60 meV) has diverse potential applications such as light emitting diodes42, solar cells43,44,45,46, sensors47,48,49, photodetectors50, and nanogenerators1,5. Consequently, according to the super individual properties of graphene and ZnO, combining graphene with ZnO nanoparticles can enhance performances1. Obviously, proper solubility and processability are considered the first requirements for many applications of graphene-based materials33. The poor solubility of graphene limited its application in either organic solvent or inorganic solvent24. One of the possible methods to improve the solubility is the oxidation of graphene and modification of GO with some soluble materials. Since GO has large quantities of oxygen-containing groups such as carboxyl, carbonyl and hydroxyl/epoxy, it can easily provide various types of decoration ways with organic and inorganic materials by covalent/no covalent functionalization51. Graphene oxide (GO) can be synthesized by various methods such as Staudenmaier52, Hofmann53, Jaleh54 and Marcano55. Among them, the Hummers’ method is widely used for the production of GO today56.

As assumed, combination of graphite with metal oxides can enhance the optoelectronic properties. For this purpose, in this work we study the linear and nonlinear optical properties of GO-ZnO and RGO-ZnO hybrid. Firstly, GO was synthesized directly from graphite by a modified Hummers’ method. Then, the nanocomposite of RGO-ZnO was fabricated using the hydrothermal method. The structural properties of samples were investigated using powder X-ray diffraction (XRD), Energy-dispersive x-ray (EDX) and Fourier-transform infrared (FT-IR) spectra. Optical properties of samples were investigated by UV–Visible spectroscopy and Z-scan analysis.

The Z-scan technique studied the NLO properties of GO, RGO, GO-ZnO and RGO-ZnO. The results show that GO-ZnO and RGO-ZnO hybrids exhibit enhanced NLO properties compared to GO and RGO. Hence, GO-ZnO and RGO-ZnO nanocomposites can be good candidates for optical communication and optical storage33. It is worth saying that in this study a noble approach was developed to compare GO, RGO, GO-ZnO, and RGO-ZnO characteristics usefully which has not reported in many works before. In this efficient pathway all the properties of GO, RGO, GO-ZnO, and RGO-ZnO such as \(\upbeta \) and n2 values are provided in one table (Table 1).

Table 1 The calculated nonlinear refractive index and nonlinear absorption coefficient using the Z-scan method.
Full size table

Experimental

Materials

Spectroscopically pure (SP) graphite, sulfuric acid (H2SO4), hydrogen peroxide (H2O2), Potassium hydroxide (KOH), hydrazine hydrate (N2H4), Zinc acetate dehydrate Zn (CH3COO)·7H2O potassium permanganate (KMnO4) and Sodium hydroxide (NaOH) were purchased from Merck chemical company and used without any purification.

Synthesis of graphene oxide (GO)

GO was prepared using the Hummers’ Method and by oxidation of graphite. First, 0.2 g graphite was mixed with concentrated H2SO4 (50 ml) and then stirred for 12 h. Subsequently, the mixture was cooled in an ice bath under vigorous stirring, and at the same time, KMnO4 (2.5 g) was added slowly to the suspension and stirred for 2.5 h. The solution was diluted with distilled water (50 ml) and the stirring process was continued for 1 h. The mixture was further allowed to cool down to room temperature and finally treated with 50 ml distilled water followed by 100 ml H2O2 30%. Then, it was purified by centrifuging and washing with excess water until the pH reached 7 to obtain GO56.

Synthesis of reduced graphene oxide (RGO)

Graphite oxide was dispersed in deionized water by sonication for 2 h. Then, 0.6 g KOH (purity 99.5%) and 4 ml hydrazine hydrate (concentration 80%) were added to the suspension and the temperature increased up to 100 °C and refluxed for 24 h. In the end, the obtained RGO was washed and dried52.

Preparation of GO-ZnO and RGO-ZnO nanocomposite

For the preparation of GO-ZnO nanocomposite with the mass ratio of 10% GO, 2.7 g of Zn (CH3COO)·7H2O was added to 30 ml deionized water and then, 0.05 g GO was added under low-speed stirring. After that, NaOH 3 M was added to the above solution slowly until pH reached 12 and the resulting mixture was ultrasonicated for 30 min. The final mixture was transferred to a Teflon-lined autoclave for hydrothermal synthesis at 160 °C in an oven for 20 h. The precipitate (GO-ZnO) was washed three times with deionized water and acetone and dried at 80 °C. In order to obtain RGO-ZnO, the mentioned stages were repeated with 0.05 g of RGO.

Results and discussion

Structural and elemental properties

Crystal structure quality and orientation of Graphite, GO, RGO, GO-ZnO, and RGO-ZnO nanocomposite are shown in Fig. 1. The XRD pattern of GO shows the sharp diffraction peak at 9° corresponds to the basal spacing of GO (8.97 Å) and due to the intercalation of oxygen-containing groups, it is higher than Graphite (4 Å)5. In graphite pattern, the sharp peak at 2θ = 26.4 and wide peak at 24.42 in RGO pattern are attributed to the (002) plane of a carbon atom. This wide diffraction peak in RGO is related to the poor ordering of the sheets along the plane direction and many defects in the carbon lattice. The GO-ZnO and RGO-ZnO nanocomposite patterns indicate highly crystalline peaks at 2θ = 31.9°, 34.7°, 36.5°, 47.7°, 56.8°, 63.0°, 66.6°, 67.9°, 69.2°, 72.7°. They can be attributed to the high crystallinity of ZnO during the hydrothermal method led to the disappearing characteristic peak of GO and RGO. Moreover, the results of Energy-dispersive X-ray (EDX) spectroscopy are provided in in Fig. 2 which show the present of Zn, C, S, Mn, K, and O atoms in our samples. Also, the elemental distribution and EDS spectrum indicates a clear change of Oxygen groups ratio from 34.08 (at%) in GO to 20.43 (at%) in RGO samples5,30.

Figure 1
figure 1

XRD patterns of Graphite, GO, RGO, GO-ZnO, and RGO-ZnO nanocomposite.

Full size image
Figure 2
figure 2

Scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM–EDX) analysis (a) GO, (b) RGO, (c) GO-ZnO, and (d) RGO-ZnO.

Full size image

FT-IR

The results of FTIR spectroscopy of GO, RGO, GO-ZnO, and RGO-ZnO are depicted in Fig. 3. The pattern of GO shows an O–H group stretching vibration band at 3340 cm−1. The sp2 structure of C=C and carbonyl functional groups (C=O stretching) are observed at 1630 cm−1 and 1730 cm−1 respectively24. The absorption peak at 1220 cm−1 and 1044 cm-1 are related to the C–O stretching vibration band. It must be mentioned, the vibrations bands of O–H, C=O, and C–O have been severely reduced, attenuated and slightly shifted to a lower wavenumber due to deoxygenation in RGO. Finally, the absorption peaks at 480 cm−1 in GO-ZnO and RGO-ZnO patterns can be attributed to ZnO stretching vibration.

Figure 3
figure 3

FTIR spectra of GO, RGO, GO-ZnO, and RGO-ZnO.

Full size image

UV–Vis absorption spectra

As shown in Fig. 4a, the UV–Vis spectra show that GO exhibits two absorption peaks: one at about 230 nm, presumably due to the π → π* transition of the C–C bonds, and another shoulder at about 300 nm corresponds to the n → π* transition of the C = O bonds57,58,59,60. Whereas the peak of the π → π* transition shifts to 260 nm for RGO, suggesting that some groups on the GO surface are removed and the conjugated structure is restored, reflecting increased π-electron concentration and structural ordering, which is consistent with the restoration of sp2 carbon and possible rearrangement of atoms61,62. Two absorption peaks were observed in the spectrum of GO-ZnO at 230 nm and 366 nm related to the GO absorption peak and the main absorption peak of ZnO respectively63. RGO–ZnO hybrids exhibit band edge absorption at 361 nm, almost 9 nm less blue-shifted than that of the band gap absorption of bulk ZnO at 370 nm, which might be explained by the quantum confinement effect of the smaller feature size of ZnO. Also, it can be seen a slight blue shift in absorption peak of RGO-ZnO from 260 to 258 nm comparing pure RGO.

Figure 4
figure 4

(a) Absorbance spectra of GO, RGO, GO-ZnO, and RGO-ZnO nanocomposite and (b) band gap determination according to absorption measurements of GO, RGO, GO-ZnO, and RGO-ZnO.

Full size image

The optical absorption coefficient (\(\mathrm{\alpha }\)) can be calculated using Eq. (1),

$$\mathrm{\alpha h\nu }=\mathrm{D}{\left(\mathrm{h\nu }-{\mathrm{E}}_{\mathrm{g}}\right)}^{\mathrm{n}},$$
(1)

where D is constant, \(\mathrm{h\nu }\) is the incident photon energy and \({\mathrm{E}}_{\mathrm{g}}\) is the optical band gap and n illustrates type of optical transition. In the present case, n = 1/2 is considered, which is attributed to direct bandgap of prepared material; hence, the plot of \({\left(\mathrm{\alpha h\nu }\right)}^{2}\) versus \(\mathrm{h\nu }\) is depicted in Fig. 4b. Finally, the bandgap of prepared materials was calculated by extrapolation used on X-axis.

The band gaps of GO and RGO are ~ 3.36 eV and 3.18 respectively because on reduction, some of the oxygen groups are removed and bandgap can therefore be adjusted further by managing the oxygen present in RGO. The Energy bandgap for pure ZnO found 3.37 eV64 and with GO, it increased around 3.63 eV. It is well-known that the excitation energy of ZnO nanoparticles will increase with the decrease of grain diameter according to the Kubo theory65; therefore, the blue shift of ZnO bandgap in GO-ZnO structure can be attributed to the reduction of ZnO nanoparticles size. The band gap energy of RGO-ZnO reduced to 3.25 eV due to the increase in the surface charge between ZnO and RGO led to the optical band gap shifting to a higher wavelength.

Nonlinear optical measurement

As the NLO properties are of great importance for high performance all-optical photonic devices, in this part, the third-order optical nonlinearities of the GO, RGO, GO-ZnO and RGO-ZnO samples have been investigated using the Z-scan techniques. Therefore, the second harmonic of a Q-switched Nd: YAG laser (532 nm, 4 ns) was used as the laser source. The concentrations of the sample solutions GO, RGO, GO-ZnO and RGO-ZnO are 0.2 mg/ml, which were placed in 1 mm quartz cells. After entering the sample, the laser beam was divided by a beam splitter. The measurements were carried out employing open-aperture and closed-aperture configurations. The reflected beam was used as an open-aperture signal and the transmitted one passed through a small hole (s = 0.3) as a close-aperture signal.

Nonlinear optical properties

As Graphene has excellent nonlinear optical properties and optical limiting performance, it was of significant interest to assess the NLO properties of metal oxides combined with graphene. Graphene has a unique atomic and electronic 2D sheet structure including sp2 hybridized carbon atoms. On the other hand, GO is a mainly 2D network which contains a large number of sp3 hybridized carbon atoms along with some sp2 domains and its concentration can be increased by chemical reduction. It is reported that highly reduced GO with a larger percentage of sp2 carbon domains exhibit good SA, while partially reduced GO exhibit good RSA characteristics66.

Due to the existence of prior graphitic nanoislands which are sp2-hybridized carbon clusters, the GO possess some properties of graphene. For example, ultrafast carrier dynamics and Pauli blocking, cause fast SA in ultra-broad spectra region. As a result, after being excited by a 532 nm laser, the SA which originates from Pauli blocking overcomes the NLO absorption at low pump intensities. According to the small amount of the sp2 configurations in GO, the contribution of excited state absorption (ESA) originating from small localized sp2 configurations to the nonlinear absorptive valley should be minor compared to the TPA67.

The nonlinear absorption (NLA) of GO is mostly derived from two-photon absorption (TPA), originating from the sp3 domains; it dominates the NLA at high pump intensities which is because of the high energy gap of sp3 bonded carbon (2.7–3.1 eV)68.

Figure 5 shows open-aperture Z-scan results of GO, RGO, GO-ZnO and RGO-ZnO. It is known that the valley depth of the Open aperture Z-scan curve reflects the optical limiting properties of the material. If the valley was deeper, its optical limiting performance would be better68.

Figure 5
figure 5

Open aperture measurements (at 532 nm) of (a) GO, (b) RGO, (c) GO-ZnO, and (d) RGO-ZnO at excitation intensity of 532 nm. Symbols represent experimental data and solid lines represent theoretical fit.

Full size image

As the figure shows, the normalized transmittance curve of GO, RGO, GO-ZnO, and RGO-ZnO samples include valleys, which means that they exhibit reverse saturable absorption (RSA). Obviously, the depths of the valleys in the GO and RGO curves are slightly different which demonstrates that reduction process in RGO can improve NLO properties of GO.

In addition, the nonlinear absorptive valley of RGO–ZnO is obviously deeper and broader than that of GO–ZnO, which suggests an increase in NLO properties68. It can be mostly attributed to the thermal reduction of the GO moiety to RGO. Following the reduction, the small localized sp2 configurations may increase greatly in number but cannot interconnect to form new sp2 carbon clusters in RGO moiety67. The nonlinear absorption coefficient β of GO, RGO, GO-ZnO, and RGO-ZnO was investigated. The curves of materials show different trends for nonlinear absorption coefficient β, which should be owing to their complicated NLO response mechanisms. After covalent functionalization with ZnO, the GO–ZnO hybrid exhibits a much higher value of β than that of GO. Moreover, the RGO–ZnO hybrid shows the significantly larger value of β regarding that of the GO–ZnO hybrid and yields the highest nonlinear absorption coefficient β of 31.9 × 10−10 cm/w, which can be related to the effective reduction of GO moiety to RGO. The larger value of β observed for RGO–ZnO suggests that it should have competitively better optical limiting performance67.

To evaluate the nonlinear optical properties of GO, RGO, GO-ZnO and RGO-ZnO quantitatively, we fit the experimental data with the following equations:

$$\mathrm{T}\left(\mathrm{z}\right)=\sum_{\mathrm{m}=0}^{\infty }\frac{{\left({-\mathrm{q}}_{0}\left(\mathrm{z}\right)\right)}^{\mathrm{m}}}{{\left(\mathrm{m}+1\right)}^\frac{3}{2}},$$
(2)
$${\mathrm{q}}_{0}\left(\mathrm{z}\right)=\frac{\upbeta {\mathrm{I}}_{0}{\mathrm{L}}_{\mathrm{eff}}}{\left(1+\frac{{\mathrm{z}}^{2}}{{\mathrm{z}}_{0}^{2}}\right)} ,$$
(3)

where T is the open aperture normalized transmittance, \({\mathrm{z}}_{0}=\frac{\uppi {\mathrm{W}}_{0}^{2}}{\uplambda }\) is the Rayleigh range, z is the sample position, W0 is the beam waist at the focal point (Z = 0), \(\uplambda \) is the laser wavelength, I0 is the peak intensity, and \({\mathrm{L}}_{\mathrm{eff}}=\frac{1-\mathrm{exp}(-\mathrm{\alpha L})}{\mathrm{\alpha }}\) is the effective thickness of the sample where α is the linear absorption coefficient of the sample and is calculated from the UV spectrum68.

In the situation of third order nonlinearity, the refraction index of material can be expressed in terms of light intensity:

$$ {\text{n }}\left( {\text{I}} \right) \, = {\text{ n}}_{0} + {\text{ n}}_{{2}} {\text{I}}, $$
(4)

where n0 is the linear refractive index and I is the intensity of the incident laser light. The z-scan measurement with an aperture (close aperture) was performed for the investigation of nonlinear refraction of GO, RGO, GO-ZnO, and RGO-ZnO nanocomposite.

Typical peak–valley (valley–peak) transmittance curve is achieved when the nonlinear refractive index of the medium is negative (positive). For determining the nonlinear refractive index, one can monitor the transmittance change through a small circular aperture which is placed at the far-field position.

In order to obtain the relation between the normalized transmittance T(z) and z position, the samples move along the axis of the incident beam (z-direction) concerning the focal point. The variation of this quantity as a function of ΔΦ0 is given by:

$$ \Delta {\text{T}}_{{{\text{P}} - {\text{V}}}} = \, 0.{4}0{6}\left( {{1} - {\text{ S}}} \right)^{0.25} \Delta \Phi_{0} , $$
(5)

where S is the linear transmittance of the far-field aperture. |ΔΦ0| relates to n2 through the following expression:

$$ \Delta \Phi_{0} = {\text{ k}} L_{eff} {\text{n}}_{{2}} {\text{I}}_{0} , $$
(6)

where I0 is the intensity of the laser beam at focus z = 0, \({\mathrm{L}}_{\mathrm{eff}}=\frac{1-\mathrm{exp}(-\mathrm{\alpha L})}{\mathrm{\alpha }}\) is the effective thickness of the sample, α is the linear absorption coefficient and L is the thickness of the sample. The nonlinear refractive index n2 (cm2/w) can be obtained from Eqs. (5) and (6)69.

The close-aperture Z-scan results of GO, RGO, GO-ZnO, and RGO-ZnO are provided in Fig. 6. In this figure, symbols show experimental transmission data, while solid lines are taken by fitting the experimental data to the non-linear transmittance. Moreover, the nonlinear refractive index (n2) is taken to be a fitting parameter.

Figure 6
figure 6

Closed aperture measurements (at 532 nm) of (a) GO, (b) RGO, (c) GO-ZnO, (d) RGO-ZnO. Symbols represent experimental data and solid lines represent theoretical fit.

Full size image

Closed aperture experimental data of GO, RGO, GO-ZnO, and RGO-ZnO for excitation intensity of 532 nm are found to fit well for typical values of n2. It is clearly evident from Fig. 6 that the samples exhibit prefocal peak and postfocal valley characteristics, which is a direct indication of negative n2 (positive lens) and it suggests that RGO-ZnO can also be used as self-focusing materials around 532 nm. The difference between normalized peak and valley transmittance ΔTP-V (denoting TP−TV) can be directly measured by z-scan technique70. The calculated values related to ΔTP-V and n2 are provided in Table 1.

Conclusions

We have reported the synthesis, structure and nonlinear optical properties of GO, RGO, GO-ZnO, and RGO–ZnO nanocomposite. The results of XRD, EDX, FT-IR, and UV–Vis confirm the successful fabrication of RGO–ZnO. The samples were separately characterized and tested for nonlinear optical properties. The results of open-aperture Z-scan testing on GO, RGO, GO-ZnO, and RGO-ZnO showed the significantly enhancing nonlinear absorption coefficient β values of RGO-ZnO (8.4 × 10−3 cm/w) compared to pure GO and GO-ZnO. It can be attributed to the combination of different NLO mechanisms in RGO–ZnO including the SA from the sp2 clusters in the RGO moiety and the RSA originating from the ZnO moiety. Furthermore, the Z-scan curve of RGO-ZnO displays a deeper RSA valley. the close-aperture Z-scan results of GO, RGO, GO-ZnO, and RGO-ZnO suggest a notable increase in nonlinear refractive index n2 of RGO-ZnO sample in comparison with GO.

Considering the easy-to-prepare and low-cost RGO-ZnO nanocomposite and its excellent NLO properties, this work may provide some insight into the design of other novel graphene-based materials for optoelectronic devices such as optical limiting, optical switches, and optical sensors.