Journal of Materials Science & Technology  2019 , 35 (9): 1996-2002 https://doi.org/10.1016/j.jmst.2019.05.012

Orginal Article

Defective graphene as a high-efficiency Raman enhancement substrate

Tong Zhaoab, Zhibo Liu, Xing Xinab, Hui-Ming Chengabc, Wencai Renab*

a Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China
b School of Material Science and Engineering, University of Science and Technology of China, Shenyang 110016, China
c Shenzhen Geim Graphene Center, Tsinghua-Berkeley Shenzhen Institute (TBSI), Tsinghua University, 1001 Xueyuan Road, Shenzhen 518055, China

Corresponding authors:   *Corresponding author at: Shenyang NationalLaboratory for Materials Science, Institute of MetalResearch, Chinese Academy of Sciences, Shenyang110016, China.E-mail address: wcren@imr.ac.cn (W. Ren)

Received: 2019-04-9

Online:  2019-09-20

Copyright:  2019 Editorial board of Journal of Materials Science & Technology Copyright reserved, Editorial board of Journal of Materials Science & Technology

More

Abstract

Pristine graphene (PG) has been demonstrated to be an excellent substrate for Raman enhancement, which is called graphene-enhanced Raman scattering. However, the chemically inert and hydrophobic surface of PG hinders the adsorption of molecules especially in aqueous solutions, and consequently limits the Raman enhanced efficiency. Here, we synthesized defective graphene (DG) films by chemical vapor deposition on Au, which has a defect density of ∼2.0 × 1011 cm-2. The DG shows a much better wettability than PG towards dye solution. Combining with the strong adsorption ability of defects to molecules, DG shows greatly enhanced efficiency than PG with perfect lattice. For example, the detection limit for rhodamine B can reach 2 × 10-9 M for DG while it is on the order of 10-7 M for PG. In addition, DG has high enhancement uniformity and the Au substrate can be reused after electrochemical bubbling transfer. These advantages suggest the great potential of the DG grown on Au for practical applications in environmental monitoring.

Keywords: Defective graphene ; Raman enhancement ; Chemical vapor deposition ; Bubbling transfer

0

PDF (2129KB) Metadata Metrics Related articles

Cite this article Export EndNote Ris Bibtex

Tong Zhao, Zhibo Liu, Xing Xin, Hui-Ming Cheng, Wencai Ren. Defective graphene as a high-efficiency Raman enhancement substrate[J]. Journal of Materials Science & Technology, 2019, 35(9): 1996-2002 https://doi.org/10.1016/j.jmst.2019.05.012

1. Introduction

Surface-enhanced Raman scattering (SERS) is a promising analytical method with swift response, nondestructive detection, and fingerprint recognition, which has a wide range of applications in environmental monitoring, surface science research, biomolecular detection, and food safety [[1], [2], [3]]. Electromagnetic enhancement (EE) and chemical enhancement (CE) are the two major mechanisms for SERS [4]. EE occurs when the surface plasmons excited by the incident light magnify the electromagnetic field [5]. The SERS substrates based on EE are all noble metals with rough surfaces [6], such as nanoparticle arrays or thin films with subwavelength patterns of Au and Ag [[7], [8], [9], [10]]. However, the fabrication process of such substrate is too complex to control and reproduce. Moreover, the enhanced efficiency is varying from metal to metal, and the poor biological compatibility prevents these substrates being used in biomolecular detection [4,11]. CE is based on a charge transfer process between the adsorbed molecules and the substrate, which makes the positive and negative charges in the molecule become more separated and leads to an increase in the polarizability of the molecule, and thus the cross section of the Raman scattering is enhanced [12].

Graphene, a two-dimensional (2D) honeycomb lattice of carbon atoms, with the highest known carrier mobility, record thermal conductivity and extremely high mechanical strength [[13], [14], [15], [16]], has attracted great interest in both fundamental studies and technological applications [[17], [18], [19], [20]]. Previous studies show that pristine graphene (PG) is an excellent substrate for Raman enhancement, known as “graphene-enhanced Raman scattering” (GERS) [3,4,21]. However, the chemically inert and hydrophobic surface of PG hinders the adsorption of molecules especially in aqueous solutions, and consequently limits the Raman enhanced efficiency [22]. Even though the defects in graphene are adverse to electronic and optoelectronic applications [23], they might be favorable for GERS since the defects have strong adsorption ability to molecules and might modify the surface chemistry of graphene [24].

In this study, we report the synthesis of monolayer defective graphene (DG) on Au foils by chemical vapor deposition (CVD), which shows a defect density of 12 times higher than that of PG grown on Cu and much better wettability than PG for dye solution. Such high density defects and good wettability greatly improve the Raman enhanced efficiency. For example, the detection level of DG for rhodamine B (RhB) can reach 2 × 10-9 M, which is two orders of magnitude higher than that of PG. Moreover, the electrochemical bubbling transfer enables the reuse of Au foil and the DG film has very uniform structure. These advantages suggest the great potential of DG grown on Au foil as a promising candidate for SERS substrate.

2. Experiment

2.1. CVD growth of DG on Au substrate

50-μm-thick high-purity Au foil (99.99%) was used as substrate. In order to remove impurities and organics adsorbed on the surface, the Au foil was annealed in air at 1000 °C for 1 h before first use. After that, it was placed in the center of a tube furnace chamber and further annealed at 1000 °C for 20 min in argon (240 SCCM) and hydrogen (10 SCCM) atmosphere. Then methane gas was introduced with a flow rate of 8 SCCM for 20 min. After growth, the Au foil was quickly removed from the high-temperature zone under argon and hydrogen for rapid cooling (Supplementary Fig. S1).

2.2. CVD growth of PG on Cu substrate

We grew PG on Cu substrate by CVD as previously reported [25]. 25-μm-thick Cu foil (99.9%) was used as substrate. It was first annealed at 1000 °C in hydrogen (5 SCCM) for 30 min to remove impurities and organics adsorbed on the surface and then exposed to a mixture of hydrogen (5 SCCM) and methane (35 SCCM) for 30 min to grow graphene at a total pressure of 100 Pa. After that, it was slowly cooled down to room temperature.

2.3. Transfer of graphene by electrochemical bubbling

The graphene films were transferred onto SiO2/Si substrate or quartz by electrochemical bubbling [26]. Poly(methyl methacrylate) (PMMA, Mw = 960,000, 4 wt% in ethyl lactate) was first spin-coated on the surface of graphene grown on Au or Cu foil at a rate of 2000 rpm for 1 min and then baked at 180 °C for 30 min. After that, the PMMA-coated graphene on Au or Cu foil was dipped into a NaOH (1 M) aqueous solution and used as the cathode under a constant current of 0.18 A or 0.05 A, respectively. After PMMA-coated graphene was separated from Au or Cu foil by the hydrogen bubbles, it was stamped onto SiO2/Si substrate or quartz and finally PMMA was removed by hot acetone at 50 °C.

2.4. Characterizations of graphene films grown on Au and Cu substrates

The morphology of graphene films grown on Au and Cu substrates and those transferred onto SiO2/Si were characterized by using a scanning electron microscope (SEM, Nova Nano SEM 430) and an optical microscope (Nikon LV100D). The optical transmittance of the transferred graphene films on quartz was measured by UV-vis-NIR spectrometer (Varian Carry 5000) with wavelength from 400 to 700 nm. Raman spectra were recorded with a Raman spectrometer (JY HR800, 784, 633 and 532 nm laser wavelength, 1 μm spot size). Raman mapping was performed with a step of 1 μm across a 20 × 20 μm2 area. The laser power was below 1 mW to avoid laser-heating-induced damage on the sample. Wettability of the graphene films was measured using a contact angle dropmeter at room temperature.

2.5. Preparation of samples for SERS detection

The dye molecules were deposited on the surface of graphene by simply soaking the graphene on SiO2/Si substrate in the solution of dye molecules. Three kinds of molecules with a series of concentrations were used. One is the methyl orange (MO) solution (2 × 10-5 M in deionized water), the second one is the methylene blue (MB) solution (2 × 10-5 M in deionized water), and the third one is the rhodamine B (RhB) solution (2 × 10-9 M to 2 × 10-5 M in deionized water). The RhB solutions with different concentrations were prepared by diluting the concentrated solution step by step. The soaking time was 1 h for all the samples. After soaking, the samples were washed with deionized water to remove free molecules and then dried under gentle nitrogen flowing.

3. Results and discussion

Fig. 1(a) shows a typical SEM image of the graphene film grown on Au substrate. Uniform contrast was observed on the same grain of Au foils and no exposed Au areas were observed in the whole area, indicating that a continuous and uniform graphene film was synthesized on Au substrate. The optical image of the graphene transferred on SiO2/Si substrate is shown in Fig. 1(b). Note that the transferred film remains its original integrity, indicating that no breakage was generated by bubbling transfer. A rough analysis of the optical image shows that more than 95% of the area has the same contrast, which confirms the high uniformity of the graphene film in the number of layers. Fig. 1(c) shows that the optical transmittance of the graphene film transferred onto quartz is 97.1% at 550 nm, indicating that it is monolayer [27]. We further characterized the structure of the graphene film by using Raman spectroscopy. Fig. 1(d) shows Raman spectra taken from 5 randomly selected positions in Fig. 1(b) (blue lines). The intensity ratio of 2D peak to G peak, I2D/IG, is ranging from 1.2 to 2, confirming the monolayer structure of the film. Note that Au has a very low carbon solubility of 0.06%, which is similar to that of Cu (0.04%). We suggest that such low carbon solubility enables the growth of uniform monolayer graphene on Au by the surface adsorption mechanism [28].

Fig. 1.   (a) SEM image of the graphene film as-grown on Au substrate. The different contrasts represent the different orientations of Au grains. (b) Optical image of the graphene film transferred on SiO2/Si substrate. (c) Optical transmittance spectrum of the graphene film transferred on quartz. (d) Raman spectra taken from randomly selected 5 positions in (b) and the as-grown graphene film on Au.

More importantly, the graphene grown on Au shows prominent D peak and D′ peak (Fig. 1(d)), indicating the presence of a great number of defects. Note that the values of ID/ID′ taken from different positions are all larger than 4. It has been reported that the intensity ratio of D to D′ peak, ID/ID′, does not depend on the defect density, but only on the nature of defect. It is ∼13 for defects associated with sp3 hybridization, ∼7 for vacancy-like defects, and ∼3.5 for grain boundary (GB)-like defects [29]. The large ID/ID′ with a value over 4 indicates the coexistence of vacancy-like defects and GBs. It is worth noting that the as-grown graphene on Au shows prominent D and D′ peaks as well (Fig. 1(d), red line), indicating that the defects are the intrinsic structure features rather than generated by transfer. Moreover, the D, G and 2D peaks are respectively located at 1366, 1598 and 2731 cm-1 on Au while at 1345, 1591 and 2685 cm-1 on SiO2/Si substrate. It is apparent that D, G and 2D peaks of as-grown graphene on Au are upshifted compared to those of graphene after transfer onto SiO2/Si substrate and the upshift of 2D peak is larger than that of G peak, indicating that the graphene on Au are compressed during growth [30,31]. This is attributed to the large lattice mismatch between graphene and Au. Therefore, the formation of such defective graphene (DG) is mainly due to the presence of strain caused by big lattice mismatch between Au and graphene and the insufficient catalytic activity of Au foil [32,33].

For comparison, we also synthesized PG film on Cu substrate by a low pressure CVD. Compared to the graphene grown on Au foils, the PG film grown on Cu is dominantly monolayer but with more multi-layer islands (Fig. 2(a, b)), which might due to the higher catalytic activity of Cu and the resulting increased carbon supply. These islands lead to a decrease in the optical transmittance (Fig. 2(c)). More importantly, it shows a much better quality with a very low D peak (Fig. 2(d)). According to the previous reports, the intensity ratio of D peak to G peak, ID/IG, is closely related to the density of defects, ID/IG ∝ 1/Ld2, where Ld is the mean distance between two neighboring defects [34]. Note that ID/IG of the graphene grown on Au is ∼0.9, while that of graphene grown on Cu is only ∼0.07 (Fig. 2(d)). We estimated the defect density in the two samples using the above equation. The extracted defect density of the graphene grown on Au is 2.0 × 1011 cm-2, which is 12 times larger than that of graphene grown on Cu (1.6 × 1010 cm-2). Because of the strong adsorption ability of defects, the high-density defects may cause more molecules being adsorbed in graphene in solution, especially for low concentration solution. We also measured the wettability of DG and PG for water and RhB solution, a representative dye solution. As shown in Fig. 3(a, b), the water contact angle (CA) of DG on SiO2/Si is ∼58° and that of PG on SiO2/Si is ∼60°, which are consistent with previous report [22]. More importantly, the contact angle of DG for RhB solution (2 × 10-5 M) is only ∼37°, while that of PG is ∼47°, indicating that the wettability of DG is superior to that of PG for RhB solution (Fig. 3(c, d)). Therefore, the defective graphene is expected to have a better performance as a GERS sensor.

Fig. 2.   (a) SEM image of PG film grown on Cu substrate. (b) Optical image of the PG film transferred on SiO2/Si substrate. (c) Optical transmittance spectrum of PG film transferred on quartz. (d) Raman spectra taken from randomly selected 5 positions in (b).

Fig. 3.   (a, b) Photos of a water droplet on (a) DG and (b) PG. (c, d) Photos of a RhB solution (2 × 10-5 M) droplet on (c) DG and (d) PG.

To evaluate the GERS performance of DG, the DG films were transferred onto SiO2/Si substrate and soaked in 2 × 10-5 M concentration of RhB, MB, and MO solution for 1 h, and then washed with deionized water to remove free molecule and dried under gentle N2 flowing. After that, Raman measurements were performed with lasers of 532 nm (2.33 eV), 633 nm (1.96 eV), and 784 nm (1.58 eV) to detect the signals of these molecules. As shown in Fig. 4(a-c), the three main Raman features of DG (D, G, and 2D peaks) and several other peaks from the dye molecules are observed for 532 nm and 633 nm lasers. Moreover, the main Raman features of RhB and MO are prominent with 532 nm laser while those of MB are more distinct with the 633 nm laser, which means that the Raman intensity exhibits a resonance condition for each dye molecule on DG with a specific laser excitation energy. However, neither graphene nor dye molecule signal can be detected with 784 nm laser possibly due to the very low energy (1.58 eV). In sharp contrast to the DG/SiO2/Si substrate, no any visible signals of the three molecules can be detected on bare SiO2/Si substrate for all the laser energies used (Fig. 4(d-f)). These results demonstrate that DG film can enhance the Raman scattering of dye molecules.

Fig. 4.   (a-c) Raman spectra of DG films that have been soaked in (a) RhB, (b) MB and (c) MO excited with different lasers. (d-f) Raman spectra of bare SiO2/Si that have been soaked in (d) RhB, (e) MB and (f) MO excited with different lasers.

To compare the Raman enhancement effect of DG and PG, we measured the Raman spectra of DG and PG soaked in a series of RhB solutions with different concentrations, ranging from 2 × 10-9 M to 2 × 10-5 M, and the results are shown in Fig. 5. Note that the peak at 1650 cm-1 is most distinct for RhB, therefore we used the intensity ratio of the 1650 cm-1 peak of RhB to the G peak of graphene (I1650/IG) to evaluate the Raman enhancement effect. It is clear that I1650/IG of DG substrate is larger than that of PG substrate and the difference in I1650/IG for these two samples becomes more pronounced with decreasing the concentration of RhB. This indicates that the GERS effect of DG is superior to PG as a sensing substrate. With decreasing the concentration, the coverage (or adsorption probability) of the RhB molecules on graphene film is decreased [35]. As a result, fewer RhB molecules are involved in the Raman process, which leads to a descent of I1650/IG for both samples except for the DG samples soaked in RhB solution of 2 × 10-5 M. As for the DG samples soaked in high concentration RhB solution of 2 × 10-5 M, the adsorbed RhB molecules are not monolayer and the clusters formed on the surface of graphene decreases the Raman efficiency [36]. The stronger adsorption ability of DG and its better wettability for RhB solution enable it a larger I1650/IG than PG, and difference is more distinct for the samples soaked in low concentration RhB. As shown in Fig. 5(d, e), there is no visible signal of RhB being detected for the PG substrates soaked in 2 × 10-8 M RhB solutions while the signal still can be easily distinguished even for the DG substrates soaked in 2 × 10-9 M. This indicates that the detection level of DG substrate for RhB is two orders of magnitude higher than that of PG, which makes DG film more practical in environmental monitoring for low concentration of dye molecules.

Fig. 5.   Comparison of the enhanced Raman scattering effects of DG and PG that have been soaked in RhB solutions with different concentrations. (a) 2 × 10-5 M, (b) 2 × 10-6 M, (c) 2 × 10-7 M, (d) 2 × 10-8 M, and (e) 2 × 10-9 M. (f) I1650/IG vs the concentration of RhB solution for DG and PG. Inset, photograph of a series of RhB solutions with concentration from 2 × 10-5 M to 2 × 10-9 M.

Besides high detection level, the enhancement uniformity is also very important for the real applications of graphene as Raman enhancement substrate. We first studied the structural uniformity of DG films by Raman mapping. As shown in Fig. 6(a), ID/IG of DG is narrowly distributed in the range of 0.7-1.2 across an area of 20 × 20 μm2, corresponding to the defect density of 1.6 × 1011 cm-2 to 2.7 × 1011 cm-2. Such uniform defect distribution is expected to result in uniform GERS effect and avoid the appearance of random hot spots that are usually observed for metal substrate [3]. We then performed Raman mapping on the DG films that have been socked in RhB solution with a concentration of 2 × 10-7 and 2 × 10-9 M. It can be seen that I1650/IG is dominantly distributed in the range of 0.9-1.1 and 0.2‒0.4, respectively (Fig. 6(b, c)), confirming the high uniformity of the GERS effect of DG film.

Fig. 6.   Raman maps of (a) ID/IG of DG and I1650/IG of DG films that have been socked in RhB solution with a concentration of (b) 2 × 10-7 and (c) 2 × 10-9 M. (d) Photograph of the completely separated PMMA/DG and Au foil. (e) Optical image of the graphene film grown on the reused Au foil and then transferred on SiO2/Si substrate. (f) Raman spectra taken from randomly selected 5 positions in (e).

Production cost of DG films is the third issue that needs to be considered for practical applications. Here, we used electrochemical bubbling method to transfer the DG film grown on Au foil. Supplementary Fig. S2 shows that PMMA/DG is separating from the Au foil by the H2 bubbles produced at the cathode. Different from the etching-based transfer method, this electrochemical bubbling transfer enables the repeated use of Au foil. After bubbling transfer, the PMMA/DG is completely separated without cracking and the Au foil remains its original structure (Fig. 6(d)). Moreover, the graphene grown on the reused Au foil shows the same optical image and Raman spectra with those grown on the fresh Au foil (Fig. 6(e, f)). The reuse of Au substrate greatly reduces the cost of DG films, which paves the way for the industrial applications of DG films grown on Au foil as GERS-based sensor in the future.

4. Conclusion

In summary, we synthesized uniform DG films on Au by CVD. The high-density defects and good wettability of DG greatly improve the enhanced efficiency. Moreover, Au substrate can be reused after electrochemical bubbling transfer, which decreases the cost of scalable production of DG. The high enhanced efficiency, good enhancement uniformity and low production cost make DG a great potential in environmental monitoring. Moreover, the synthesis of DG also opens up the possibilities for the use of graphene in electrochemistry sensors and biosensors.

Acknowledgements

This work was supported by the National Key R&D Program of China (No. 2016YFA0200101), the National Natural Science Foundation of China (Nos. 51325205, 51290273, 51521091, and 51861135201), Chinese Academy of Sciences (No. 174321KYSB20160011), and the Strategic Priority Research Program of Chinese Academy of Sciences (No. XDB30000000).

The authors have declared that no competing interests exist.


/