Open Access Article
Nan-Fu Chiu
* and
Cheng-Du Yang
Laboratory of Nano-photonics and Biosensors, Institute of Electro-Optical Science and Technology, National Taiwan Normal University, Taipei 11677, Taiwan. E-mail: nfchiu@ntnu.edu.tw
First published on 23rd March 2018
The development of a stepwise deoxidized process and real-time monitoring of the large-scale mass production of electrochemically reduced graphene oxide (ErGO) sheets are important issues. In this study, we have shown that graphene oxide (GO) sheets can be quantitatively monitored in real-time and controlled in a stepwise manner using electrochemical-surface plasmon resonance (EC-SPR), due to the fact that the oxygen functional groups can be tuned through a deoxidization procedure. The SPR signal can then be detected quantitatively in real-time by changes in the dielectric constant of the GO film during the EC stepwise removal of oxygen functional groups. This is because the refractive index of the GO sheets is affected by the oxygen-containing groups, so that monitoring the SPR angle shift provides a real-time measure of changes in the concentration of the residual oxygen functional groups of the GO sheets. In this study, we demonstrated GO and 100 CV cycles of ErGO at X-ray photoelectron spectroscopy carbon-to-oxygen ratios of 4.1 and 31.57 respectively, and Raman spectra of the D/G intensity ratio of 0.85 and 1.89, respectively. The 100 CV cycles of ErGO at SPR angle shifts were −227.13 mdeg for GO at a concentration of 0.275 mg ml−1, and −263.47 mdeg for GO at a concentration of 1 mg ml−1. The photoluminescence emission bands of the GO and the CV 100 cycles of ErGO were 615 to 470 nm. These results may be beneficial for future studies on GO fluorescence characteristics in the field of optoelectronic and biosensor applications.
O and O
C–OH, which have a close relationship and are involved in the reaction,10,15 leading to various applications including biosensors,17–20 and fluorescence spectra.21 In addition, GO sheets can be used as effective fluorophores10,14,22,23 and in quenching24,25 to enhance the efficiency of the excitation energy transfer (Förster) process, and this has been widely used in advancing biosensors for cell imaging applications. Moreover, the GO surface has residual oxygen-containing functional groups, which may lead to poor crystallinity with no uniformity.26,27 Other applications of reducing GO to reduced graphene oxide (rGO) film-based carbon materials include increasing the stability of electrodes,28 improving charge carrier mobility,7,29 tuning its dielectric and optical band-gap properties,1,10,14 enhancing binding interactions and improving biosensor sensitivity,30,31 modulating surface plasmon resonance (SPR) energy,32,33 increasing the efficiency of gas molecule adsorption34,35 and drug delivery loading,36 all of which can contribute to advances in science and technology. We previously demonstrated the first electrochemical-surface plasmon resonance (EC-SPR) immunosensor incorporating an electrochemically reduced graphene oxide (ErGO) film, and showed the potential of the specific affinity properties of ErGO in electrochemical-enhanced biosensing.37
In this study, we propose an alternative method to chemical and thermal reduction by simultaneously using SPR real-time monitoring of changes in the GO deoxygenation process and electrochemical (EC) stepwise reduction of GO to ErGO. This EC-SPR technique can monitor and control the oxygen-containing functional group in a stepwise manner on the GO surface in real time, and can improve GO surface crystallinity defects, as GO surface crystallinity defects are closely related to the preparation methods and reduction technologies depending on the size of the GO sheet layers. These phenomena will directly affect the characteristics of the fluorescence emission. There are currently no relevant studies on the use of the EC-SPR technique for the real-time monitoring and stepwise reduction of GO fluorescence emission, and no discussion of the related issues. Our results show that EC-SPR could be successfully applied to the development of GO for the simultaneous detection of residual oxygen functionality in ErGO leading to changes in the refractive index (carbon-to-oxygen (C/O) ratio and band-gap) resulting in angle shifts. Using real-time monitoring and the stepwise reduction of GO are conducive to the future development of luminescent semiconductor GO materials and the future of sensing materials. More importantly, the emission spectra of GO sheets can effectively be tuned, and therefore have the potential to advance the field of fluorescence in various applications, as well as biosensing technology. Of these applications, assays for naked-eye biosensors are the most common due to their simplicity, rapid screening ability, semi-quantitative analysis, and low cost.
ErGO chips were obtained from GO chips electrochemically with different reduction conditions as shown in Fig. 1b. Adsorption of functional oxygen in the form of epoxy, hydroxyl, hydrocarbon, carbonyl, carboxyl and ether groups created in the GO sheets is shown in Fig. 1c.
In the second method, the GO solution-based reduction of GO was performed in phosphate buffered saline (PBS), NaCl and directly reduced with the aforementioned electrochemical method. ErGO solutions were prepared from the GO aqueous suspensions at different concentrations (0.01, 0.275, 1, and 2 mg ml−1) in a volume of 1 cm3 of the ErGO solution. In the ErGO aqueous suspensions, there was no need to link the GO sheets for immobilization on the Au surfaces, as shown in Fig. 1d. The scanning electron microscope (SEM) images of different concentrations of GO films after ErGO are shown in Fig. 1e–g. The transmission electron microscope (TEM) image showed that the GO sheets exhibited a configuration of a few 2D layers with a typical wrinkled flake structure (Fig. 1h).
Fig. 2b and c show the CV curve of the reaction current and proper procedure potential of a triangular wave, respectively. Fig. 2b shows a GO reduction peak of the current curve at −0.596 mA for point “c”, and Fig. 2c shows an electrochemical reduction voltage of GO of around −1.085 V for point “b” during the first CV cycle. The first irreversible oxidation reaction exhibited a pronounced peak, showing a maximum current at a potential of 1.085 V. Fig. 2c shows that point “a” had a CV scanning potential of −1.1 V, and Fig. 2a shows that point “d” had an SPR angle shift of −36.18 mdeg in the first CV cycle. The potential started at 0 V, however, the GO reduction process resulted in a gradual decrease in the SPR angle shift in each cycle. As the instability of the SPR angle shift represents the changes of instantaneous double layer interface charge density, electrochemical oxidation and reduction currents may cause such a shift in response to a potential perturbation in the dielectric properties of the reduction of GO.45
Fig. 2d shows that the reduction current began to drop significantly in the first voltammetric scanning cycle, and that ErGO showed a lower reduction potential of −1.085 V, yielding a current peak at −0.596 mA. This result indicated that the ErGO film that was formed at −1.085 V had a lower reduction peak than that obtained in the first cycle of the electrochemical reduction. In later cycles, the negative shifting of the applied reduction potentials shrank the reduction peak of the resulting ErGO films. A reduction current was observed in the ErGO film that was prepared using a reduction potential of −1.085 V, showing the efficient stepwise electrochemical reduction of the oxygen groups under this condition. The reduction current continued to fall until it disappeared, and the deoxidization processes exhibited irreversible properties. The plots showed the relationship between the current and the shift in the SPR angle, as shown in Fig. 2d–f. The stepwise cyclic voltammograms and real-time SPR curves were recorded in the first CV cycle of the electrochemical redox-reaction. Fig. 2e shows the SPR angle shift and complete, stepwise deoxidization of GO. The first CV at a potential of −1.085 V increased the reduction current to −0.417 mA and generated an SPR angle shift of −101.9 mdeg. The SPR angle could be observed in the sharp reduction in the current and potential-dependent change in the deoxygenation process, and the largest shift was produced at an SPR angle (θSPR) of −201.0 mdeg. In a related report, the deoxy-reduction of GO to ErGO reduced the thickness from 1.2 nm to 0.8 nm and increased the refractive index from 2.24 to 3.5. Therefore, the influence of the shift in the SPR resonance angle in the stepwise deoxygenation process that changed the refractive index was far stronger than that in the deoxygenation reaction that reduced the thickness of the 0.275 mg ml−1 ErGO film.46,47
Fig. 2f shows the real-time SPR evaluation of the deoxidization process at various scan cycles at a scan rate of 50 mv s−1. During several stepwise CV cycles, the oxygen groups in the GO were progressively reduced, and as the number of CV cycles increased, the C/O ratio increased and the number of residual oxygen functional groups decreased. The SPR angle shift that was caused by the reduction process in the first cycle was obvious. Fig. 2e shows the plot of SPR response curves that were obtained by real-time monitoring of the residual oxygen-containing functionality of the ErGO film and changes in the refractive index which caused an angle shift. The results showed that long-term monitoring of SPR angular displacement was obviously affected by environmental temperatures, resulting in gradual changes in the drift angle. Table 1 shows the SPR angle shifts upon electrochemical reduction for 10 (720 s), 50 (3600 s) and 100 (7200 s) CV cycles of the deoxidization process were −168.45, −214.22 and −227.13 mdeg for a GO concentration of 0.275 mg ml−1, and −196.01, −260.94 and −263.47 mdeg for a GO concentration of 1 mg ml−1, respectively.
| Films | XPS spectra (O1s) | XPS spectra (C1s) | SPR (θ, mdeg) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
O C (%) |
O–C (%) | C–C, sp2 (%) | C–C, sp3 (%) | C–O (%) | C O (%) |
O–C O (%) |
C/O ratio | 0.275 mg mL−1 | 1 mg mL−1 | |
| GO | 85.6 | 14.4 | 76.59 | 3.79 | 17.47 | 1.38 | 0.78 | 4.1 | 0 | 0 |
| ErGO (10 cycles) | 86.9 | 13.1 | 78.59 | 12.5 | 7.38 | 0.95 | 0.58 | 10.22 | −168.45 | −196.01 |
| ErGO (50 cycles) | 94.5 | 5.5 | 79.31 | 15.11 | 4.2 | 0.85 | 0.53 | 16.92 | −214.22 | −260.94 |
| ErGO (100 cycles) | 95.8 | 4.2 | 80.88 | 16.05 | 2.32 | 0.54 | 0.21 | 31.57 | −227.13 | −263.47 |
We previously used the multilayer reflection model theory for Fresnel's law to calculate the GO and ErGO film at the SPR angle shift (θsp) versus thickness (d) to verify the calculated relationship between the refractive index and thickness of GO and ERGO at the SPR angle shift.37 These results suggest that the effect of the electrochemical reduction GO can effectively remove interlamellar water layers, resulting in an increased refraction index of ErGO and reduced oxygen content of sheet layers, which then results in a significant shift of the SPR angle in a real-time response.
The combination of SPR and electrochemical stepwise deoxidization process enabled large changes in SPR angle shifts as a result of reducing the oxygen content in the conductivity of the GO film on a gold electrode surface. The electrochemical stepwise deoxidization process changed the oxygenated functional groups of the GO sheets and reduced the number of oxygen bonds. This then led to an increase in the refraction index of the ErGO sheets, and possibly also resulted in a reduction in their thickness. Partial removal of intercalated water and oxygen may also have affected the measured thicknesses, thereby resulting in a significant shift of the SPR angle in a real-time response.
O bonds (288.2 eV, 1.38%), and carboxylate carbon O
C–O bonds (289.2 eV, 0.78%). Fig. 3b shows the O 1s XPS spectra of the GO film, which showed two components corresponding to the oxygen atom groups including O
C (531.8 eV, 85.6%) and O–C (533.4 eV, 14.4%). Fig. 3c–h show the ErGO films at different electrochemical reduction CV cycles, and the composition and ratio of the carbon and oxygen atoms are shown in Table 1. The ratio of the original GO film for carbon (C–C, sp2 and sp3) was 80.38%. The increase in the full width half maximum (FWHM) of the peaks in Fig. 3g show a clear trend. The main effect in XPS is a slight increase in the fraction of sp2 carbon atoms, however, the proportion of sp3 carbon atoms showed a dramatic increase.
The results showed that the ErGO film at CVs of 10, 50 and 100 cycles contained significant amounts of sp2 carbon atoms (78.59%, 79.31%, and 80.88%), signifying increases in most carbon atoms, but decreases in oxygen atoms on the GO due to the electrochemical reduction. This result showed that the C–C bonds were electrochemically reduced from GO sp3 to the structure of graphene sp2. In contrast, the C/O ratio of the GO film and that of the CV 100-cycle ErGO film were 4.1 and 31.57, respectively. Table 1 shows the C/O atomic ratios of GO before and after electrochemical reduction, which were obtained by analyzing the C 1s XPS spectral peaks. The C 1s XPS spectral peaks of the ErGO films yielded C/O ratios of 10.22, 16.92 and 31.57 after 10, 50 and 100 cycles, respectively.1,7,10,37
The relationship between the SPR angle and the XPS of the residual oxygen functionality of the ErGO film was further investigated. According to the XPS data, the C/O ratio of ErGO exceeded that of GO, confirming the effectiveness of electrochemical deoxygenation. This implied that the ErGO film from the electrochemical reduction process contained far less oxygen, thereby confirming the tunable band-gap and high quality of the ErGO.37
Fig. 4a shows the FTIR spectra of GO and ErGO films. The absorption peaks at approximately 860 and 1200 cm−1 were from the C–O–C of the epoxy stretching vibrations and the C–O of alkoxy stretching vibrations at approximately 1080 cm−1, respectively. The peak at around 1650–1750 cm−1 was caused by the carboxyl C
O stretching vibration of the COOH group. The peak O–H deformation vibrations in C–OH were seen at approximately 1305 cm−1, and the peak at around 1500–1600 cm−1 was attributed to the C
C skeletal vibration of the graphene sheets. The peak at around 2950–2850 cm−1 was attributed to C–H stretching vibrations due to pendant alkyl chains, and the peak at approximately 3410 cm−1 was due to –OH stretching vibrations. FTIR analysis showed an increase in C
C and decreases in C–O–C and O–H. Raman analysis of the same carbon lattice products revealed a G band, which represented the formation of the graphene sheet.
![]() | ||
| Fig. 4 Analysis of 0.275 mg ml−1 GO films at different electrochemical reduction conditions for (a) FTIR spectra, and (b) Raman spectra of the GO films, showing D and G bands. | ||
The sp2 carbon lattices were all common and produced by the stretching of C–C bonds. The GO peak near 1595 cm−1 was due to first order scattering of E2g phonons of the sp2 carbon atoms.48 However, whereas the D-band peak intensity revealed the plane vibrations attributed to the presence of the graphene structure defects,49–51 the G peak represented the ordered sp2 hybridization of the in-plane vibrations of the carbon–carbon bonds in graphene.52 The peak ratios of the intensity of the D and G peaks showed that rGO exhibited a significant increase compared to GO. The relative intensity ratio (ID/IG) is a measure of the defects present on a graphene structure. The results showed that the D-band was higher, meaning that sp2 bonds were broken, thereby resulting in more sp3 bonds. Therefore, the reduced GO had a higher ID/IG, meaning that there was a defect. The XPS showed a slight increase in the fraction of sp2 carbon atoms, however the proportion of sp3 carbon atoms increased dramatically. At the same time the D-band Raman intensity also increased, suggesting that the reduced GO had more defects than the original GO. This proved that the XPS and Raman were consistent with the sp3 carbon atoms being located at the defect sites, which is consistent with previous studies.53–56
Therefore, we analyzed the GO sheets under different electrochemically reduced conditions, which showed two characteristic Raman D and G bands at 1350 cm−1 and 1595 cm−1, respectively (Fig. 4b). The Raman spectra of the GO sheets showed the D/G intensity ratio (ID/IG = 0.85). The ErGO sheets also contained both D and G bands in Raman spectra, with D/G intensity ratios of 1.18, 1.52, and 1.89 for 10, 50 and 100 CV cycles, respectively, which is larger than that of GO sheets (ID/IG = 0.85). As a comparison, the electrochemically reduced ErGO sheets exhibited a much higher D/G intensity ratio of 1.89.
O bonds.57 In addition, with the increase in the number of reductions, the absorption peaks at 227 and 300 nm slowly disappeared, whereas ErGO from CV 1 to 100 cycles had a characteristic absorption band shift at approximately 270 nm, which corresponded to the π–π* transition of C–C bond shifts to 270 nm, indicating the reduction of GO and restoration of C
C bonds in the ErGO sheets.58
In this normalized PL measurement, the spectra showed GO and ErGO at different concentrations of aqueous solution and electrochemical reduction conditions (Fig. 5b–d). The PL optical behavior of the photo-excited electrons in GO and ErGO was due to non-radiative relaxation and radiative recombinations from discrete sp2-related states and continuous-defect states.59,60 GO consists of many disordered defect states within the π–π* gap and exhibits PL spectra with long wavelengths and broad optical frequencies. However, ErGO had a lower number of disordered inducible defect states in the π–π* gap and an increased number of clusters of newly formed small isolated sp2 domains.59 Therefore, the results showed that the PL spectra of GO and ErGO solutions at room temperature shown in Fig. 5b exhibited a broad PL response from 450 nm to 800 nm. Fig. 5b shows the PL spectra of GO solution at room temperature at concentrations of 0.275, 1, and 2 mg ml−1 with an excitation wavelength of 405 nm. It can be seen that the PL intensity tended to increase with increasing concentrations of the solution, which is similar to previous studies.61–63
Fig. 5c and d represent the PL spectra at λex = 405 nm of rGO at four different reduction cycle conditions for solutions with concentrations of 0.275 and 1 mg ml−1, respectively. The rGO sheets exhibited quenching of PL emission spectra resulting in a blue-shift due to an increase in the number of sp2 clusters after reduction.64 Therefore, the tunable PL spectra during the reduction of GO could be attributed to changes in the relative intensities of PL emission of the two different types of electron excitation states. The PL emission spectra of ErGO due to the disappearance of functional oxygens atoms was due to restoration of more sp2 clusters, and the newly formed sp2 clusters in rGO could provide percolation pathways between the sp2 clusters already present.64,65 This result showed the reduction of the sp2 and sp3 hybridization of the GO and ErGO heterostructures.59,60
The pure GO sheets exhibited a PL band with the maxima at 615 nm, as shown in Fig. 5b. In contrast to the GO sheets, the ErGO sheets, with an increased number of electrochemical reduction cycles, showed a gradual shift in the spectrum to the blue band. Fig. 5c shows that the GO sheets at a high concentration of 1 mg ml−1 had electrochemical reduction condition indices of 1, 10, 50, and 100 CV cycles at center wavelengths of 600, 530, 520 and 472 nm, respectively, compared to 565, 511, 504, and 470 nm, respectively, for the 0.275 mg ml−1 GO sheets (Fig. 5d). The PL experiments showed the PL emission spectra for the direct transition types of GO and ErGO semiconductor materials,14,18,59 and confirmed that the real-time and stepwise deoxidization process of the electrochemical reduction of GO sheets could effectively tune the PL emission spectra.
| This journal is © The Royal Society of Chemistry 2018 |