Spontaneous implantation of gold nanoparticles on graphene oxide for salivary SERS sensing

Amin Zhang a, Jie Chang a, Yunsheng Chen b, Zhicheng Huang a, Gabriel Alfranca a, Qian Zhang a and Daxiang Cui *a
aInstitute of Nano Biomedicine and Engineering, Shanghai Engineering Research Center for Intelligent Instrument for Diagnosis and Therapy, Department of Instrument Science & Engineering, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, 800 Dongchuan Rd, Shanghai 200240, China. E-mail: dxcui@sjtu.edu.cn
bDepartment of Plastic and Reconstructive Surgery, Shanghai Ninth People's Hospital, School of Medicine Shanghai Jiao Tong University, 639 Zhizaoju Road, Shanghai 200011, P. R. China

Received 14th July 2019 , Accepted 29th July 2019

First published on 5th August 2019


Clean SERS substrates could be prepared without using polymer stabilizers and highly toxic reducing agents, and they play a pivotal role in trace amount detection. Herein, well-dispersed graphene oxide (GO)-supported Au (GO–Au) nanocomposites are synthesized using a hydrothermal method in an alkaline environment in the absence of traditionally used reductants and surfactants. These nanocomposites are subsequently used to manufacture a SERS sensor. The silk-like GO plays multiple roles as the reductant for the synthesis of Au nanoparticles (NPs), the surfactant for Au NPs and the support to uniformly distribute Au NPs. Moreover, the in situ obtained Au NPs could provide efficient ‘hot spots’, and therefore ensure favorable SERS analytical performance of the GO–Au nanocomposite-based sensor which featured an enhancement factor of 8.1 × 107. To further evaluate its practical clinical applications, the prepared SERS sensor was utilized to analyze eight salivary amino acids, which are gastric cancer (GC) biomarkers, in 44 saliva samples. The results revealed that the performance of the prepared sensor was satisfactory for salivary SERS sensing and the selectivity and sensitivity of the sensor for diagnosing GC were excellent. Therefore, this simple synthesis strategy could provide a novel idea for preparing clean SERS sensor substrates with potential application prospects in the clinical field.


Introduction

Currently, surface-enhanced Raman spectroscopy (SERS) is emerging as a powerful, sensitive, and non-invasive analytical tool in biological analysis, surface science, and environmental monitoring and offers elaborate molecular fingerprint information on analyzed species.1–5 The analytical performance of SERS is mainly influenced by the amplification of the electromagnetic field.6 A commonly used strategy to achieve a noteworthy enhancement in the efficiency of SERS sensors is to produce numerous nanogaps or ‘hot spots’7 (interstitial positions between coinage-metal nanoparticles (NPs), particularly Au,8 Ag9–14 or Cu NPs, or the locations around the sharp surfaces of protrusions) where the electromagnetic field is strengthened significantly. Consequently, the physicochemical properties of metallic nanomaterials could be regulated by changing their morphological structures, and many efforts have been devoted to generating inter- or intra-particle nano-gaps or synthesizing heteromorphic Au/Ag/Cu nanomaterials, such as spheres, stars with sharp tips, triangles, and rods.2,15–17 The Raman signals of analytes are greatly magnified when analytes are located at or close to nano-gaps or hot spots, and therefore nanogaps and hot spots improve the sensitivity of SERS analysis. Of all reported NPs, Au NPs exhibited chemical stability and a favourable Raman enhancement effect, and they were selected to generate numerous hot spots during SERS analysis in this study.

Surfactants and reductants play crucial roles in most conventionally prepared Au NP systems. However, surfactants that adhere to the surface of Au NPs might cause some undesired SERS signals, and hence could interfere with the Raman performance of analytes. Therefore, the absence of organic surfactants and chemical reducing agents during the preparation of Au NPs is regarded as a technologically important challenge for synthesizing ‘clean’ NPs. Recently, carbon nanomaterials, including carbon nanotubes,18 carbon nanospheres19,20 and graphene,21,22 have received significant attention, owing to their favorable stability and unique physicochemical properties. Among these carbon nanomaterials, graphene, which has a one atom-thick two dimensional (2D) structure, has been widely employed for manufacturing sensors23,24 and preparing novel nanocomposites,25,26 owing to its mechanically strong sp2 C-bonded honeycomb lattice and high carrier mobility with ballistic transport. While traditionally PEGylated or other surfactant-modified Au NPs have been extensively studied, only a few papers have been published on synthesizing Au NPs on the surface of graphene using GO and metal cations27 and no added surfactants and reductants. In this study, we innovatively synthesize Au NPs on the surface of GO to form GO–Au nanocomposites by directly using GO as the carrier, surfactant and reducing agent during the synthesis process. The synthesised clean GO–Au nanocomposites are then used to manufacture the SERS sensor.

Tumor biomarkers are chemical or biological substances that are overexpressed in cancer tissues and could be used for disease diagnosis, clinical research, and predicting disease recurrence,28 and thus, detecting their presence or observing the quantitative changes of tumor biomarkers is of paramount importance. Commonly used tumor biomarkers mostly contain oncofoetal proteins,29 tumor-associated antigens,30 enzymes,31 and hormones, which are usually present in patients' blood and urine. Traditional medical tumour diagnostic techniques involve blood tests and cumbersome approaches and present many disadvantages, such as high cost as well as time consuming and complex detection processes. Therefore, it is important to develop fast and non-invasive methods for tumor diagnosis. Notably, the determination of internal metabolites, particularly amino acids, in human saliva would offer an excellent non-invasive disease-surveillance method.32 Saliva, a complex oral fluid, could provide abundant information on people's health conditions and is regarded as a type of accessible and safe biomarker for diagnosing diseases. According to one of our previous studies, many amino acids in human saliva could be selected as gastric cancer (GC) biomarkers, and they were analysed using high performance liquid chromatography-mass spectrometry.33 However, in the above-mentioned study, we experienced some shortcomings, including the complex preparation procedure, low nanocomposite yield for SERS sensor preparation, and lower SERS enhancement factor (EF). Therefore, in this study, GO–Au nanocomposites which are simple, innovative, and green materials were synthesized and subsequently used to manufacture a SERS sensor featuring a high EF, which presented potential sensitivity for detecting salivary amino acids. Moreover, eight salivary amino acid biomarkers for GC, and significant differences in their concentrations, were observed in saliva samples from healthy control and GC patients.

In this study, we first synthesized GO–Au nanocomposites without using traditional chemical reductants and surfactants, resulting in highly clean of GO–Au nanocomposites. And the nanocomposites presented favourable stability and an excellent SERS enhancement effect. Subsequently, a clean SERS sensor based on green GO–Au nanocomposites was developed for rapid determination of salivary amino acids to distinguish between healthy controls and GC patients (see Scheme 1). Owing to the clean synthesis procedure of GO–Au and presence of numerous Au NPs exposed on the surface of GO, the developed SERS sensor exhibited favourable selectivity and good sensitivity. Moreover, the large number of Au NPs on the surface of the nanocomposite provided numerous ‘hot spots’, which significantly enhanced the Raman signals of amino acid metabolites when amino acid molecules were adsorbed into the nanogaps between Au NPs by GO. The SERS sensor discussed above showed favourable analytical performance for amino acid determination, and therefore could represent a novel and non-invasive method of diagnosing GC.


image file: c9ay01500k-s1.tif
Scheme 1 Schematic illustration of the surface-enhanced Raman scattering sensor manufacturing process.

Results and discussion

Synthesis and characterization

In this study, GO–Au nanocomposites were fabricated by in situ generation of numerous Au NPs on the surface of a GO substrate in the absence of traditional chemical reductants and surfactants. Both TEM and SEM analyses revealed the 2D and silk-like layered structure of GO (Fig. 1A and B), which provided favourable and large-area growth sites for Au NPs. In addition, the numerous in situ synthesized Au NPs (ca. 20 nm) were well distributed on the surface of GO (Fig. 1C), which indicated that large numbers of hot spots were generated to enhance the analytical performance of the manufactured SERS sensor. Furthermore, the SEM image of the GO–Au nanocomposites (Fig. 1D) also demonstrated the co-existence of Au NPs and GO. Additionally, energy-dispersive X-ray spectroscopy analysis was utilized to verify the presence of Au, C, and O in the structure of the GO–Au nanocomposites (Fig. 1E), which confirmed the successful preparation of the GO–Au nanocomposites.
image file: c9ay01500k-f1.tif
Fig. 1 (A) Transmission electron microscopy (TEM) and (B) scanning electron microscopy (SEM) images of graphene oxide (GO). (C) TEM and (D) SEM images of GO–Au nanocomposites (inset: enlarged TEM image of GO–Au). (E) Energy-dispersive X-ray spectroscopy analysis of GO–Au nanocomposites.

Elemental analysis of the GO–Au nanocomposites was performed by recording their X-ray photoelectron spectroscopy (XPS) spectra (Fig. 2). The XPS data of the GO layer and GO–Au nanocomposites correspond to curves a (black) and b (red), respectively, in Fig. 2A. As shown in Fig. 2A, the presence of C 1s and O 1s in curve a indicated the presence of GO in the sample. Compared with the XPS spectrum of GO (curve a), additional peaks for Au 4s, Au 4p, Au 4d, Au 4f, and Au 5s were observed in curve b, which suggested the presence of Au NPs on the surface of GO (red curve in Fig. 2A). Comparing the high-resolution C 1s spectra of GO and GO–Au (Fig. 2C and D), the intensity of the C–O signal sharply decreased while the intensity of the C–C signal increased during the formation of GO–Au nanocomposites, which indicated that some O-containing functional groups had been consumed while participating in the synthesis of GO–Au nanocomposites.


image file: c9ay01500k-f2.tif
Fig. 2 (A) Full XPS spectra of GO (curve a) and GO–Au nanocomposites (curve b). (B) The UV-Vis absorption spectroscopy of GO (a) and GO-Au (b). (C and D) High-resolution C 1s spectra of GO and GO–Au. (E) The high-resolution Au 4f spectrum of GO–Au.

Moreover, the fabricated nanocomposites were characterized using UV-Vis absorption spectroscopy. As illustrated in Fig. 2B, bare GO (curve a) displayed an obvious absorption peak (λmax) at 246 nm, and GO–Au (curve b) presented a new absorption peak at 525 nm (which was the typical absorption peak of Au NPs) while the absorption peak of GO exhibited a slight red shift, which further demonstrated the successful manufacture of GO–Au nanocomposites and partial reduction of GO in the process.

Effects of pH and temperature on the manufacture of GO–Au nanocomposites

To investigate the optimal reaction conditions, the effects of the solution pH and hydrothermal reaction temperature were analyzed. The UV-Vis spectra obtained in this study could provide compelling evidence of the generation of Au NPs in the presence of GO. As presented in Fig. 3A, GO–Au nanocomposites were synthesized at room temperature in the dark for seven days at different pH values, and the corresponding UV-Vis spectra were obtained and monitored. The specific peak intensity of Au NPs at 525 nm in the UV-Vis curves in Fig. 3 increased as the solution pH increased from 3.0 to 12.0 for the same reaction time, which indicated that the quantities of generated Au NPs increased as the pH values of the reaction systems increased. Moreover, the peak intensity of GO at approximately 240 nm did not change as the pH changed, which suggested that the concentration of GO in all reaction systems was the same. Therefore, it was concluded that the higher pH values of the reaction system were beneficial for the formation of Au NPs. For our protocol, we selected pH 11 as the optimal pH value because higher pH values could have caused the reduction of GO and therefore would have increased its hydrophobicity, thus resulting in the precipitation of GO.37 Additionally, the GO–Au solutions displayed distinct color differences owing to the gradually increasing number of Au NPs (see the inset of Fig. 3A: bottles 1–10, pH 3.0–12.0).
image file: c9ay01500k-f3.tif
Fig. 3 Ultraviolet-visible absorption spectra of graphene oxide–Au (GO–Au) nanocomposites at (A) different solution pH levels and (B) various reaction temperatures. (C) Surface-enhanced Raman scattering (SERS) spectra of rhodamine 6G solutions of different concentrations: (a) 10−4 M, (b) 10−5 M, and (c) 10−6 M on the GO–Au-based SERS sensor and (d) 10−3 M on the GO-based SERS sensor; (e) Raman spectrum of bulk R 6G on the Au film substrate (excitation wavelength: 532 nm, excitation intensity: 2.0 mW, and acquisition time: 1.0 s). (D) Thermal gravimetric analysis curves of (a) GO sheets and (b) GO–Au nanocomposites.

The hydrothermal temperature also played an important role in the formation of Au NPs by affecting the size and amount of Au NPs. As displayed in Fig. 3B, as the reaction temperature increased, a remarkable increase in the UV-Vis absorption intensity of Au NPs was observed, which indicated that higher temperatures could shorten the reaction time and effectively accelerate the generation of Au NPs on the surface of GO. However, much higher temperatures could destroy the silk-like structure of GO and therefore could result in the aggregation of Au NPs (Fig. S1), which could decrease the adsorption capacity of the obtained nanocomposite materials and reduce the number of hot spots. Therefore, 120 °C was selected as the optimal temperature for manufacturing GO–Au nanocomposites. Therefore, GO–Au nanocomposites were successfully obtained using the optimal experimental conditions.

Chemosynthesis mechanism of GO–Au nanocomposites

To elucidate the chemosynthesis mechanism of the prepared GO–Au nanocomposites, the redox potential of GO at various solution pH values was tested. As illustrated in Fig. S2, increasing the pH of the reaction solution could effectively reduce the reduction potential of GO, which would increase the redox potential gap between GO and HAuCl4, and therefore accelerate the in situ growth of Au NPs resulting in a higher generation rate of Au NPs on the surface of GO. Moreover, comparing the C 1s binding energies of GO–Au and GO (Fig. 2C and D), we observed that the intensity of the peak ascribed to the C–C bonds (284.6 eV) increased and the peak signal of the C–O bonds (286.6 eV) significantly decreased, which revealed that numerous O-containing functional groups of GO were consumed during the synthesis of GO–Au nanocomposites and played a crucial role in generating Au NPs. Additionally, as presented in Fig. 3A, GO–Au nanocomposites could be prepared at room temperature without using additional energy sources, such as heating, irradiation, or ultrasonication. This suggested that the redox reaction between HAuCl4 and GO under alkaline conditions was spontaneous. The reaction mechanism for preparing GO–Au nanocomposites was similar to the previously reported mechanisms, such as those for the preparation of Pt NPs on single-walled carbon nanotubes38 and direct generation of Pd NPs on GO.39 The process included two steps: (i) generation of small Au nucleation sites on the surface of GO and (ii) growth of Au nucleation sites into large Au NPs. The growth of Au NPs after nucleation could be a galvanic reaction-like process, where numerous Au3+ ions are reduced on the surface of small Au nuclei by the electrons that were transported from GO and accompanying consumed O-containing functional groups of GO.

Raman enhancement effect of the prepared sensor

The EF of the manufactured SERS sensors, which was closely associated with the detection sensitivity of SERS sensors, was analysed. In this study, R 6G40 was used as the SERS analyte to simulate the adsorption of amino acids to investigate SERS measurements. As depicted in Fig. 3C, the Raman enhancement of various materials was studied after they were modified using R 6G. 2 μL R 6G solutions of varying concentrations (10−4, 10−5, and 10−6 M) were added onto the surface of the developed GO–Au SERS sensor. In addition, 2 μL R 6G solution (10−3 M) is added onto the surface of the GO modified base. Following drying at ambient temperature, the obtained sensors were analysed using a Raman spectrometer equipped with a 532 nm laser. As shown in Fig. 3C, as the R 6G concentration increased from 10−6 to 10−4 M, the characteristic signal intensity of R 6G (curves c to a) gradually increased owing to the increased number of R 6G molecules adsorbed on the surface of the SERS sensor. When 10−3 M R 6G was added onto the surface of GO (curve d) and bulk R 6G was placed on the bare Au base (curve e), a higher signal intensity of R 6G was recorded from the GO-modified base, which indicated that GO could strengthen the Raman signal owing to its large specific area and chemical enhancement properties. Compared with the R 6G signal collected from GO-modified sensor, the signal intensity of 10−6 M R 6G recorded using the SERS sensor was conspicuously higher than that of 10−3 M R 6G recorded using GO, which revealed the low detection limit for R 6G and favourable enhancement efficacy of the constructed SERS sensor owing to the numerous hot spots generated by Au NPs on the surface of GO–Au nanocomposites. According to the spectra described above, the EF was calculated to be 8.1 × 107, which indicated that the developed SERS sensor exhibited favourable SERS sensitivity and could be further employed for determining salivary amino acids. The ISERS value used to calculate the EF was the signal intensity of 10−6 M R 6G collected at the surface of the GO–Au film base while the IBulk value was the signal intensity of bulk R 6G collected at the bare base.

Inadequate GO amounts would cause aggregation of the nanocomposite, which would result in a smaller number of Au NPs being used to generate hot spots. In this study, TGA was performed to analyse the mass percentages of GO and Au NPs of the GO–Au nanocomposites. As displayed in Fig. 3D, the TGA curves of GO and GO–Au (curves a and b, respectively) displayed three weight loss sections, between 30 and 140, 140 and 300, and 300 to 600 °C, which were attributed to the removal of water, disappearance of O functional groups and C combustion of GO, respectively. Moreover, comparing these two TGA curves, we concluded that more than 70 wt% Au NPs were present in the GO–Au nanocomposites, which ensured a significant increase in the Raman signal of the analyzed amino acids when using the manufactured SERS sensor.

Uniformity analysis of developed SERS sensor

Signal uniformity of the developed Raman sensor substrates was of pivotal importance for the analysis of the saliva samples, and ensured the favourable reproducibility and accuracy of the SERS sensor. As illustrated in Fig. 4A, a randomly selected 120 μm × 120 μm area of the used SERS sensor was investigated by measuring the signal intensity distribution of R 6G molecules. Additionally, all SERS spectra corresponding to the selected points are depicted in Fig. 4B, and revealed the great uniformity and excellent reproducibility of the developed sensor. Moreover, the intensities of the characteristic peaks at 1504 cm−1 of the R 6G molecules collected using the Raman sensor were selected to generate a Raman map. The uniform and strong Raman signal of R 6G can be observed in the Raman map in Fig. 4C, which suggested that the manufactured SERS sensors presented great potential applications for detecting salivary amino acids.
image file: c9ay01500k-f4.tif
Fig. 4 (A) SERS bright-field microscopic picture of the partial GO–Au base. (B) Surface-enhanced Raman scattering (SERS) spectra of rhodamine 6G (R 6G) corresponding to selected positions. (C) Raman map of the constructed SERS sensor based on the 1504 cm−1 band of R 6G (the step size of the x, y stage was 3 μm at 532 nm excitation wavelength, 0.2 mW excitation intensity and 5.0 s accumulation time for every point).

Analytical application of the SERS sensor for amino acid samples

To test the applicability and reliability of the proposed SERS sensor for clinical analysis, eight salivary amino acids were selected as biomarkers to distinguish between healthy controls and GC patients. In one of our previous studies,33 38 salivary amino acids were selected as GC diagnostic biomarkers owing to the significant differences in concentration between healthy controls and GC patients. Saliva samples were collected from 220 individuals (both GC patients and healthy controls) and were analysed using mass spectroscopy. All selected amino acids used as biomarkers presented specific Raman peaks, and their intensities were linearly proportional to their concentrations. In this study, receiver operating characteristic analyses were carried out to evaluate the reported 38 amino acids (Fig. S3). Finally, eight amino acids were selected, and the selected amino acids presented areas under curves higher than 0.5 (the value of the state variable was set to 1), which suggested the excellent practicability of the selected amino acids for distinguishing between GC patients and healthy controls and diagnosing GC patients. The eight amino acids (PEtN, His, Gly, Glu, Val, Ile, Asp, and Leu) ultimately acted as biomarkers and fingerprints to distinguish between GC patients and healthy controls owing to the high concentrations of these amino acids closely correlated with patients' health status and significant concentration differences between GC patients and healthy controls (Fig. S4). Additionally, from Fig. S4, we can observe that the concentrations of PEtN, His, Gly, Glu, and Asp in the saliva samples of the GC patients were higher, which could be attributed to the cancerous tissues of GC patients generating these amino acids. Furthermore, the other three amino acids, Val, Ile, and Leu, were utilized as biomarkers for healthy controls.

The SERS spectra of the selected amino acid standards are presented in Fig. 5 and were used to test the practicability and sensitivity of the prepared SERS sensor. Compared with the Raman spectra of the bulk amino acids (black curves), the positions of the characteristic peaks (red curves) of the amino acids (10 mM) obtained using the manufactured SERS sensors were almost the same, while the intensities of the specific peaks of amino acids in the SERS spectra were significantly increased, which indicated the favourable practicability and excellent Raman enhancement effect of the manufactured SERS sensor, which was ascribed to the superficial hot spots. Moreover, all tested amino acids presented specific fingerprint peaks. Owing to the presence of GO, additional Raman peaks of GO at approximately 1600 and 1340 cm−1 were observed in the SERS spectra of the amino acids.


image file: c9ay01500k-f5.tif
Fig. 5 Raman spectra of bulk biomarker standards (black curves) and surface-enhanced Raman scattering spectra of biomarker standards (10 mM, red curves) (excitation wavelength: 532 nm, excitation intensity: 20 mW, and accumulation time: 1.0 s). PEtN, His, Gly, Glu, Val, Ile, Asp and Leu represent O-phosphorylethanolamine, histidine, glycine, glutamic acid, valine, isoleucine, asparagine, and leucine, respectively.

To demonstrate the real-life applications of SERS sensors, 44 saliva samples from 20 healthy controls and 24 GC patients were analyzed using the manufactured SERS sensor. The SERS spectra of saliva samples from GC patients (curve c), healthy controls (curve b) and the blank base (curve a) were recorded (Fig. 6A). Compared with the blank SERS spectra (curve a), different fingerprint bands were observed in the SERS spectra of GC patients and healthy controls, which suggested that the manufactured SERS sensor exhibited favourable selectivity and high sensitivity for distinguishing GC patients from healthy controls by identifying the fingerprint peaks of salivary amino acids. Moreover, as illustrated in Fig. 6A, 12 types of fingerprint bands associated with the eight amino acid biomarkers (summarised in Table 1) could be selected as SERS specific peaks to differentiate between healthy controls and GC patients. Additionally, the areas of the selected 12 fingerprint bands in the SERS spectra of the 44 saliva samples were investigated using principal component analysis (Fig. 6B). As presented in Fig. 6B, two clear clusters were well differentiated using principal components 1, 2, and 3 (PC 1, 2, and 3, respectively), which indicated the excellent application potential of this sensor for distinguishing between GC patients and healthy controls.


image file: c9ay01500k-f6.tif
Fig. 6 (A) Surface-enhanced Raman scattering (SERS) spectra of (a) the blank base and saliva samples from (b) healthy controls and (c) gastric cancer (GC) patients. (B) Principal component analysis (PCA) of saliva samples of GC patients and healthy controls. The data points in the PCA image correspond to the areas of the selected 12 fingerprint bands in various SERS spectra (excitation wavelength: 532 nm, excitation intensity: 2 mW, and accumulation time: 30 s).
Table 1 Relationship between 12 fingerprint bands and selected biomarkers
Band no. Band position (cm−1) Biomarkers Band no. Band position (cm−1) Biomarkers
1 345 PEtN, His, Gly, Asp 7 808 Val, Ile
2 388 Glu 8 840 Val, Ile, Leu
3 398 Val, Leu 9 926 Val, Ile, Leu
4 448 Asp 10 1000 Ile, Asp
5 524 PEtN, Asp 11 1125 His, Asp
6 745 Asp 12 1455 Leu


Experiments and methods

Reagents and materials

Hydrogen tetrachloroaurate(III) trihydrate (HAuCl4·3H2O), histidine (His), glycine (Gly), glutamic acid (Glu), o-phosphorylethanolamine (PEtN), valine (Val), isoleucine (Ile), asparagine (Asp), and leucine (Leu) were purchased from Sigma-Aldrich. Sodium hydroxide (NaOH), sodium nitrate (NaNO3), sulfuric acid (H2SO4) and potassium permanganate (KMnO4) were obtained from Sinopharm Chemical Reagent Co., Ltd. (China). All chemical reagents were of analytical grade and were used as received. Milli-Q water (18.2 MΩ) was obtained using a Milli-pore system.

Our study was approved by the Current Ethical Committee of Shanghai Jiao Tong University (Shanghai, China). All saliva samples collected from healthy controls and GC patients in this study were provided by the Shanghai Renji Hospital, First Hospital of Jilin University, Shanghai Tongren Hospital, Shanghai Ninth People's Hospital and Chinese PLA General Hospital and informed consent was obtained. The saliva samples were collected from volunteers who brushed their teeth and refrained from eating for more than 1 h. After centrifuging the samples for 40 min at 11[thin space (1/6-em)]000 rpm and 4 °C, approximately 0.25 mL aliquots of the supernatant liquors were collected and maintained at −70 °C for further use.

Apparatus

The morphologies and structures of various types of nanomaterials were analysed using a transmission electron microscopy (TEM, Tecnai G2 SpiritBiotwin, USA) instrument and scanning electron microscopy (SEM, Zeiss Ultra5) apparatus. Elemental analysis of the prepared nanomaterials was performed using an X-ray photoelectron spectroscopy (XPS, AXIS Ultra DLD, Japan) device. The components of nanocomposites were characterised using a thermal gravimetric analyser (TGA, Pyris 1). The surface changes of GO and GO–Au nanocomposites were investigated using an ultraviolet-visible (UV-Vis) spectrophotometer (Varian Inc., Palo Alto, CA, USA).

Synthesis of GO–Au nanocomposites

The strategy devised to grow Au NPs on the surface of GO included two steps: (i) the synthesis of GO and (ii) the in situ growth of Au NPs on the surface of GO. A slightly modified Hummers method described in one of our previous studies was used to prepare GO.27 In brief, 1 g graphite and 0.5 g NaNO3 were carefully mixed with 25 mL concentrated H2SO4 in a flask under magnetic stirring at 0 °C. Then, 3.5 g KMnO4 was gradually added to the above mixture over 60 min to maintain a low temperature (below 20 °C). After the chemicals reacted for 2 h at 35 °C, 45 mL Milli-Q water was slowly added, which quickly increased the reaction temperature to 70 °C. Then, 3.5 mL H2O2 (30%) and 30 mL Milli-Q water were added to the flask to generate a bright yellow product. After purification using centrifugal ultrafiltration (Amicon Ultra-15, Millipore, 3 K MWCO), the solid powder was washed with diluted HCl (4%) and centrifuged (8000 rpm, 20 min) three times. Then, the obtained GO was washed with pure water several times. Following purification by dialysis in water for 5 days, the GO precipitate collected by centrifugation (a small amount of HCl was added to the sample) was dried at 60 °C for 8 h. Subsequently, 2.0 mg purified GO powder was dissolved in 38 mL Milli-Q water under ultrasonication to generate a homogeneous GO solution. Then, 132 mL HAuCl4 (242.81 mM) solution was added to the GO solution under magnetic stirring. After 10 min of mixing, the obtained solution was adjusted to an alkaline pH (pH 11) using 1 M NaOH solution. Subsequently, the obtained solution was transferred into a 50 mL Teflon-sealed autoclave which was placed in an oven at 120 °C for 24 h. When the reaction was complete, the obtained product was collected by centrifugation and was washed with pure water three times to remove the excess NaOH and unreacted HAuCl4. Lastly, the obtained GO–Au nanocomposites were re-dissolved in 100 μL water for further use.

SERS sensing devices based on GO–Au for amino acid determination

Firstly, 2.5 μL of the GO–Au nanocomposite solution obtained above was added onto the surface of the Au substrate followed by drying in a N2 atmosphere for 30 min. Then, the GO–Au SERS sensor was used to determine different types of analytes, such as amino acids in saliva samples. The Au substrates used in this study were prepared as follows: a 300 nm thick Au film was uniformly sputtered onto the surface of a piece of 3 inch glass, and the obtained sample was subsequently cut into 2.5 cm × 2.5 cm pieces, which were ultimately used as substrates for the preparation of the Raman sensor.

SERS analysis of the prepared salivary sensor

After manufacturing the salivary sensors, they were analysed using a Raman spectrometer featuring a 10× magnification objective. Spectra were recorded from 300 to 2000 cm−1 at a laser excitation wavelength of 532 nm.

EF calculation using rhodamine 6G

Rhodamine 6G (R 6G), which exhibits a high quantum yield of fluorescence, was employed to measure the EF of the manufactured SERS sensor featuring macro-Raman configuration. Several characteristic peaks of R 6G were detected. The peak at 610 cm−1 was attributed to the C–C–C in-plane bending, and the specific peaks at 771 and 1181 cm−1 were ascribed to the bending of the C–H bonds in the aromatic structure. In addition, the 1574 cm−1 peak resulted from the C[double bond, length as m-dash]C bond stretching, and 1649, 1574, 1504 and 1310 cm−1 were attributed to the C–C bond stretching. Additionally, the EF value of SERS was estimated using the following equation:
 
image file: c9ay01500k-t1.tif(1)
where ISERS and IBulk are the intensities of the specific peaks at 1504 cm−1 for SERS and bulk R 6G, respectively, NBulk is the number of molecules present in the bulk R 6G sample and NSERS is the number of R 6G molecules adsorbed on the SERS substrate. Here, NSERS was calculated by considering that the R 6G layer operates in monolayer-adsorption mode, which resulted in the theoretical maximum quantity of R 6G molecules being adsorbed on the surface of the SERS sensor. This indicated that the EF values measured here were smaller than the actual EF value. Moreover, the NBulk and NSERS values could be estimated using the following relationships:
 
image file: c9ay01500k-t2.tif(2)
 
image file: c9ay01500k-t3.tif(3)
where Slaser is the illumination focus area, and the diameter of the laser area for the used Raman system is approximately 1 μm. Moreover, the area occupied by one R 6G molecule (S_(R 6G)) was approximately 1 × 104 nm2 according to the literature.34–36 The laser penetration depth, h, of the 532 nm laser beam was approximately 2 μm and A is the Avogadro constant (6.02 × 1023). In addition, the density of R 6G, ρ, was 0.79 g cm−3 and the molar mass, MR 6G, of R 6G was 479.01 g mol−1.

Considering eqn (1)–(3), the EF could be calculated using the following formula:

 
image file: c9ay01500k-t4.tif(4)

Conclusions

In this study, novel GO–Au nanocomposites have been synthesized based on the spontaneous generation of Au NPs on the surface of GO using a reductant- and surfactant-free preparation process. The demonstrated synthesis process was completely different from the traditional electrostatic modification method, and GO–Au nanocomposites were synthesized using a direct redox reaction between Au3+ ions and GO in an alkaline environment, which resulted in highly pure nanocomposites and efficient hot spots provided by the Au NPs. Additionally, the auxiliary heating approach could effectively increase the reaction rate and therefore shorten the reaction time. Then, the as-prepared GO–Au nanocomposites were used to manufacture the SERS sensor. Afterwards, the developed salivary SERS sensor was used to identify eight salivary amino acids in 44 saliva samples including 24 samples from GC patients and 20 from healthy controls with good selectivity. The results revealed the excellent performance of the SERS sensor in practical clinical diagnosis. The non-invasive and selective salivary diagnostic method proposed in this study could provide a new and fast strategy for diagnosing GC. Moreover, this straightforward and green preparation approach could offer a novel strategy for developing surfactant-free nanocomposites for clinical applications.

Conflicts of interest

There are no conflicts to declare.

Abbreviations

NPsNanoparticles
GO–AuGraphene oxide–gold
SERSSurface-enhanced Raman scattering
2DTwo dimensional
Ag NPsSilver nanoparticles
EFEnhancement factor
HAuCl4·3H2OHydrogen tetrachloroaurate(III) trihydrate
TEMTransmission electron microscopy
SEMScanning electron microscopy
XPSX-ray photoelectron spectroscopy
TGAThermal gravimetric analyser
R 6GRhodamine 6G
PEtN O-Phosphorylethanolamine
HisHistidine
GlyGlycine
GluGlutamic acid
ValValine
IleIsoleucine
AspAsparagine
LeuLeucine
GCGastric cancer
PCAPrincipal component analysis
UV-VisUltraviolet-visible

Acknowledgements

This study was supported by the China National Key Basic Research Program (Project 973) (grant no. 2017FYA0205301 and 2015CB931802), the National Natural Science Foundation of China (grant no. 81571835) and the China Postdoctoral Science Foundation (grant no. 2017M620159). The sponsorships provided by the Shanghai Sailing Program (grant no. 19YF1422300), Shanghai Jiao Tong University (grant no. 18X100040044), and Shanghai Engineering Research Center for Intelligent Diagnosis and Treatment Instrument (grant no. 15DZ2252000) are also acknowledged.

Notes and references

  1. F. Tian, J. Conde, C. Bao, Y. Chen, J. Curtin and D. Cui, Gold nanostars for efficient in vitro and in vivo real-time SERS detection and drug delivery via plasmonic-tunable Raman/FTIR imaging, Biomaterials, 2016, 106, 87–97 CrossRef CAS PubMed.
  2. X. Su, Y. Wang, W. Wang, K. Sun and L. Chen, Phospholipid Encapsulated AuNR@Ag/Au Nanosphere SERS Tags with Environmental Stimulus Responsive Signal Property, ACS Appl. Mater. Interfaces, 2016, 8(16), 10201–10211 CrossRef CAS PubMed.
  3. H. J. Butler, L. Ashton, B. Bird, G. Cinque, K. Curtis and J. Dorney, et al., Using Raman spectroscopy to characterize biological materials, Nat. Protoc., 2016, 11(4), 664–687 CrossRef CAS PubMed.
  4. K. Haruna, T. A. Saleh, J. Al Thagfi and A. A. Al-Saadi, Structural properties, vibrational spectra and surface-enhanced Raman scattering of 2,4,6-trichloro- and tribromoanilines: a comparative study, J. Mol. Struct., 2016, 1121, 7–15 CrossRef CAS.
  5. K. Haruna, T. A. Saleh, M. K. Hossain and A. A. Al-Saadi, Hydroxylamine reduced silver colloid for naphthalene and phenanthrene detection using surface-enhanced Raman spectroscopy, Chem. Eng. J., 2016, 304, 141–148 CrossRef CAS.
  6. H. Xu, J. Aizpurua, M. Käll and P. Apell, Electromagnetic contributions to single-molecule sensitivity in surface-enhanced Raman scattering, Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top., 2000, 62(3), 4318 CrossRef CAS PubMed.
  7. K. N. Kanipe, P. P. Chidester, G. D. Stucky and M. Moskovits, Large Format Surface-Enhanced Raman Spectroscopy Substrate Optimized for Enhancement and Uniformity, ACS Nano, 2016, 10(8), 7566–7571 CrossRef CAS PubMed.
  8. T. A. Saleh, M. M. Al-Shalalfeh, A. T. Onawole and A. A. Al-Saadi, Ultra-trace detection of methimazole by surface-enhanced Raman spectroscopy using gold substrate, Vib. Spectrosc., 2017, 90, 96–103 CrossRef CAS.
  9. T. A. Saleh, M. M. Al-Shalalfeh and A. A. Al-Saadi, Graphene dendrimer-stabilized silver nanoparticles for detection of methimazole using surface-enhanced Raman scattering with computational assignment, Sci. Rep., 2016, 6, 32185 CrossRef CAS PubMed.
  10. M. Kamran, M. Haroon, S. A. Popoola, A. R. Almohammedi, A. A. Al-Saadi and T. A. Saleh, Characterization of valeric acid using substrate of silver nanoparticles with SERS, J. Mol. Liq., 2019, 273, 536–542 CrossRef CAS.
  11. M. M. Al-Shalalfeh, A. T. Onawole, T. A. Saleh and A. A. Al-Saadi, Spherical silver nanoparticles as substrates in surface-enhanced Raman spectroscopy for enhanced characterization of ketoconazole, Mater. Sci. Eng., C, 2017, 76, 356–364 CrossRef CAS PubMed.
  12. T. A. Saleh, M. M. Al-Shalalfeh and A. A. Al-Saadi, Silver loaded graphene as a substrate for sensing 2-thiouracil using surface-enhanced Raman scattering, Sens. Actuators, B, 2018, 254, 1110–1117 CrossRef CAS.
  13. A. T. Onawole, S. A. Popoola, T. A. Saleh and A. A. Al-Saadi, Silver-loaded graphene as an effective SERS substrate for clotrimazole detection: DFT and spectroscopic studies, Spectrochim. Acta, Part A, 2018, 201, 354–361 CrossRef CAS PubMed.
  14. M. M. Al-Shalalfeh, T. A. Saleh and A. A. Al-Saadi, Silver colloid and film substrates in surface-enhanced Raman scattering for 2-thiouracil detection, RSC Adv., 2016, 6(79), 75282–75292 RSC.
  15. A. B. Serrano-Montes, J. Langer, M. Henriksen-Lacey, D. Jimenez de Aberasturi, D. M. Solís and J. M. Taboada, et al., Gold Nanostar-Coated Polystyrene Beads as Multifunctional Nanoprobes for SERS Bioimaging, J. Phys. Chem. C, 2016, 120, 20860–20868 CrossRef CAS.
  16. A. W. Zhang, G. Q. Tao and J. Wang, Assembly of bioconjugated rod-nanotags and multilayer plasmonic nanorod-array for ultrasensitive SERS detection of S-aureus bacteria, J. Nanopart. Res., 2018, 20(4), 97 CrossRef.
  17. Y. H. Luo, G. Q. Wen, L. Ma, A. H. Liang and Z. L. Jiang, A Sensitive SERS Quantitative Analysis Method for Amino Acids Using Ruhemann's Purple as Molecular Probe in Triangle Nanosilver Sol Substrate, Plasmonics, 2017, 12(2), 299–308 CrossRef CAS.
  18. E. Sanchez-Tirado, A. Gonzalez-Cortes, P. Yanez-Sedeno and J. M. Pingarron, Magnetic multiwalled carbon nanotubes as nanocarrier tags for sensitive determination of fetuin in saliva, Biosens. Bioelectron., 2018, 113, 88–94 CrossRef CAS PubMed.
  19. D. D. Yang, M. Liu, J. Xu, C. Yang, X. X. Wang and Y. B. Lou, et al., Carbon nanosphere-based fluorescence aptasensor for targeted detection of breast cancer cell MCF-7, Talanta, 2018, 185, 113–117 CrossRef CAS PubMed.
  20. T. A. Saleh and I. Ali, Synthesis of polyamide grafted carbon microspheres for removal of rhodamine B dye and heavy metals, J. Environ. Chem. Eng., 2018, 6(4), 5361–5368 CrossRef CAS.
  21. V. Perumal, M. S. M. Saheed, N. M. Mohamed, M. S. M. Saheed, S. S. Murthe and S. C. B. Gopinath, et al., Gold nanorod embedded novel 3D graphene nanocomposite for selective bio-capture in rapid detection of Mycobacterium tuberculosis, Biosens. Bioelectron., 2018, 116, 116–122 CrossRef CAS PubMed.
  22. Q. Drmosh, Z. Yamani, A. Hendi, M. Gondal, R. Moqbel and T. Saleh, et al., A novel approach to fabricating a ternary rGO/ZnO/Pt system for high-performance hydrogen sensor at low operating temperatures, Appl. Surf. Sci., 2019, 464, 616–626 CrossRef CAS.
  23. G. Mathew, P. Dey, R. Das, S. D. Chowdhury, M. P. Das and P. Veluswamy, et al., Direct electrochemical reduction of hematite decorated graphene oxide (alpha-Fe2O3@erGO) nanocomposite for selective detection of Parkinson's disease biomarker, Biosens. Bioelectron., 2018, 115, 53–60 CrossRef CAS PubMed.
  24. Y. Shoja, A. Kermanpur and F. Karimzadeh, Diagnosis of EGFR exon21 L858R point mutation as lung cancer biomarker by electrochemical DNA biosensor based on reduced graphene oxide/functionalized ordered mesoporous carbon/Ni-oxytetracycline metallopolymer nanoparticles modified pencil graphite electrode, Biosens. Bioelectron., 2018, 113, 108–115 CrossRef CAS PubMed.
  25. M. Arvand and M. Dehsaraei, Amperometric determination of diazinon by gold nanorods/ds-DNA/graphene oxide sandwich-modified electrode, Ionics, 2018, 24(8), 2445–2454 CrossRef CAS.
  26. Q. Z. Wang, H. Y. Yue, J. J. Zhang, X. Gao, H. Zhang and X. Y. Lin, et al., Electrochemical determination of uric acid in the presence of ascorbic acid by hybrid of ZnO nanorods and graphene nanosheets, Ionics, 2018, 24(8), 2499–2507 CrossRef CAS.
  27. Y. Chen, Y. Zhang, F. Pan, J. Liu, K. Wang and C. Zhang, et al., Breath Analysis Based on Surface-Enhanced Raman Scattering Sensors Distinguishes Early and Advanced Gastric Cancer Patients from Healthy Persons, ACS Nano, 2016, 10, 8169–8179 CrossRef CAS PubMed.
  28. J. Lin and H. Ju, Electrochemical and chemiluminescent immunosensors for tumor markers, Biosens. Bioelectron., 2005, 20(8), 1461–1470 CrossRef CAS PubMed.
  29. P. E. Garrett and S. R. Kurtz, Clinical utility of oncofetal proteins and hormones as tumor markers, Med. Clin. North Am., 1986, 70(6), 1295–1306 CrossRef CAS PubMed.
  30. Y. Wang, H. Xu, J. Luo, J. Liu, L. Wang and Y. Fan, et al., A novel label-free microfluidic paper-based immunosensor for highly sensitive electrochemical detection of carcinoembryonic antigen, Biosens. Bioelectron., 2016, 83, 319–326 CrossRef CAS PubMed.
  31. L. Deng, Y. Du, J. J. Xu and H. Y. Chen, An off-on-off electrochemiluminescence approach for ultrasensitive detection of thrombin, Biosens. Bioelectron., 2014, 59, 58–63 CrossRef CAS PubMed.
  32. C. Andreou, M. R. Hoonejani, M. R. Barmi, M. Moskovits and C. D. Meinhart, Rapid detection of drugs of abuse in saliva using surface enhanced Raman spectroscopy and microfluidics, ACS Nano, 2013, 7(8), 7157–7164 CrossRef CAS PubMed.
  33. Y. Chen, S. Cheng, A. Zhang, J. Song, J. Chang and K. Wang, et al., Salivary Analysis Based on Surface Enhanced Raman Scattering Sensors Distinguishes Early and Advanced Gastric Cancer Patients from Healthy Persons, J. Biomed. Nanotechnol., 2018, 14(10), 1773–1784 CrossRef CAS PubMed.
  34. Z. Yi, X. Tan, G. Niu, X. Xu, X. Li and X. Ye, et al., Facile preparation of dendritic Ag–Pd bimetallic nanostructures on the surface of Cu foil for application as a SERS-substrate, Appl. Surf. Sci., 2012, 258(14), 5429–5437 CrossRef CAS.
  35. T. Ding, L. O. Herrmann, B. de Nijs, F. Benz and J. J. Baumberg, Self-aligned colloidal lithography for controllable and tuneable plasmonic nanogaps, Small, 2015, 11(18), 2139–2143 CrossRef CAS PubMed.
  36. Z. Yi, S. Chen, Y. Chen, J. Luo, W. Wu and Y. Yi, et al., Preparation of dendritic Ag/Au bimetallic nanostructures and their application in surface-enhanced Raman scattering, Thin Solid Films, 2012, 520(7), 2701–2707 CrossRef CAS.
  37. X. Fan, W. Peng, Y. Li, X. Li, S. Wang and G. Zhang, et al., Deoxygenation of Exfoliated Graphite Oxide under Alkaline Conditions: A Green Route to Graphene Preparation, Adv. Mater., 2008, 20(23), 4490–4493 CrossRef CAS.
  38. H. C. Choi, M. Shim, S. Bangsaruntip and H. J. Dai, Spontaneous reduction of metal ions on the sidewalls of carbon nanotubes, J. Am. Chem. Soc., 2002, 124(31), 9058–9059 CrossRef CAS PubMed.
  39. X. Chen, G. Wu, J. Chen, X. Chen, Z. Xie and X. Wang, Synthesis of “clean” and well-dispersive Pd nanoparticles with excellent electrocatalytic property on graphene oxide, J. Am. Chem. Soc., 2011, 133(11), 3693–3695 CrossRef CAS PubMed.
  40. T. A. Saleh and A. A. Al-Saadi, Surface characterization and sorption efficacy of tire-obtained carbon: experimental and semiempirical study of rhodamine B adsorption, Surf. Interface Anal., 2015, 47(7), 785–792 CrossRef CAS.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ay01500k

This journal is © The Royal Society of Chemistry 2019