Sensitive and Selective Detection of New Red Colorant Based on Surface-Enhanced Raman Spectroscopy Using Molecularly Imprinted Hydrogels

Featured Application: The prepared MIHs doped with positively charged Au NPs show excellent surface-enhanced Raman scattering activity, could be used as e ﬃ cient SERS substrates for the detection of synthetic colorants. Abstract: A polyacrylamide-based molecularly imprinted hydrogel (MIH) doped with positively charged gold nanoparticles (Au NPs) has been synthesized via a free radical polymerization of acrylamide (AM) aqueous solution containing positively charged Au NPs as a Raman active substrate, New Red colorant as a template molecule, N,N’ -methylenebis(acrylamide) as a crosslinking agent, and potassium persulfate as an initiator. The Au NPs-doped MIHs were subsequently explored as a Raman active substrate for the sensitive and selective detection of New Red colorant via surface-enhanced Raman spectroscopy (SERS). The logarithmic intensity of the characteristic peak of New Red at 1572 cm − 1 was proportional to the logarithmic concentration of New Red with a detection linear range of 1.64 × 10 − 6 to 1.64 × 10 − 4 M and a limit of detection (LOD) of 1.64 × 10 − 7 M. The recoveries ranged from 86.3% to 100.6% with a relative standard deviation (RSD) in the range of 2.3% to 7.7%. The RSD and recovery rates for the detection of New Red spiked in a sports drink sample were 1.8% to 7.7% and 91.0% to 97.1%, respectively. These results showed that SERS combined with MIHs as Raman active substrates could provide a sensitive, selective, and e ﬀ ective approach for the detection of the New Red colorant in beverage matrix.

characterization of positively charged Au NPs and MIHs, the selectivity of MIPs for New Red, and the limit of detection (LOD) of New Red as well as the corresponding linearity range were investigated in this work.

Instruments and Reagents
The morphology of colloidal gold particles and Au NPs-embedded MIHs were observed on a transmission electron microscope (TEM, Tecnai G2 F30 S-Twin, FEI Company, Hillsboro, OR, Netherlands) operating at an acceleration voltage of 300 kV. The TEM was equipped with an energy dispersive X-ray (EDX) analyzer (DPP-ІІ). The zeta potential and the size of Au NPs were measured using dynamic light scattering instrument (DLS, Nano-ZS90, Malvern Instrument, Worcs, UK) equipped with a 628 nm laser source. SERS spectra were collected on a DeltaNu 785 Raman spectrometer (DeltaNu Inc., Laramie, WY, USA). The laser power of the spectrometer is 120 mW with an excitation wavelength at 785 nm and a spectral range of 200 to 2000 cm -1 . The spectra were acquired with baseline off using NuSpec software (Copyright DeltaNu 2009) and analyzed using GRAMS/AI software (Ver 9.1, Thermo Fischer Scientific, Waltham, MA, USA).

Synthesis of Positively Charged Au NPs
Prior to experiments, all glasswares were bathed in freshly prepared aqua regia (v/v HNO3:HCl = 1:3) for 4 h, then rinsed thoroughly with distilled water, and dried prior to use. The positively charged gold nanoparticles were synthesized according to the literature [48]. Briefly, 50.0 mL of

Instruments and Reagents
The morphology of colloidal gold particles and Au NPs-embedded MIHs were observed on a transmission electron microscope (TEM, Tecnai G2 F30 S-Twin, FEI Company, Hillsboro, OR, Netherlands) operating at an acceleration voltage of 300 kV. The TEM was equipped with an energy dispersive X-ray (EDX) analyzer (DPP-II). The zeta potential and the size of Au NPs were measured using dynamic light scattering instrument (DLS, Nano-ZS90, Malvern Instrument, Worcs, UK) equipped with a 628 nm laser source. SERS spectra were collected on a DeltaNu 785 Raman spectrometer (DeltaNu Inc., Laramie, WY, USA). The laser power of the spectrometer is 120 mW with an excitation wavelength at 785 nm and a spectral range of 200 to 2000 cm −1 . The spectra were acquired with baseline off using NuSpec software (Copyright DeltaNu 2009) and analyzed using GRAMS/AI software (Ver 9.1, Thermo Fischer Scientific, Waltham, MA, USA).

Synthesis of Positively Charged Au NPs
Prior to experiments, all glasswares were bathed in freshly prepared aqua regia (v/v HNO 3 :HCl = 1:3) for 4 h, then rinsed thoroughly with distilled water, and dried prior to use. The positively charged gold nanoparticles were synthesized according to the literature [48]. Briefly, 50.0 mL of HAuCl 4 ·4H 2 O solution (1.42 mM) and 500 µL of cysteamine solution (213 mM) were added to a clean round-bottomed flask. The solution was mildly stirred with a magnetic bar at a stirring speed of 100 rpm for 1 h at room temperature. Subsequently, 12.5 µL of NaBH 4 solution (10.0 mM) was added to the mixture, and it was vigorously stirred at a stirring speed of 500 rpm at room temperature in dark for 1 more hour. Finally, a wine-red solution of positively charged gold nanoparticles with an average diameter of 49.0 ± 0.9 nm was obtained and stored in a refrigerator at 4 • C for later use.

Preparation of Positively Charged Au NPs-embedded MIHs and Non-Imprinted Hydrogels (NIHs)
A typical procedure for the synthesis of positively charged Au NPs-embedded MIHs is as follows. First, 1.0 g of acrylamide was dissolved in 4.0 mL above positively charged Au NPs solution in a 20 mL glass tube, and then 0.1 g of N,N'-methylenebis(acrylamide) and 1.0 mg of New Red were added to the solution. The solution was degassed with nitrogen gas and kept in an ultrasonic bath for 10 min to achieve complete homogenization at room temperature. Then, 0.1 mL of potassium persulfate solution (0.15 M in water) was injected into the reaction tube. The glass tube was tightly sealed, and the polymerization was carried out at 50 • C for 12 h. After the polymerization, a red, jelly-like, Au NPs and New Red-containing MIH was obtained.
In order to remove the New Red template from MIHs, the obtained MIHs were dried and crushed in a mortar and ground to pass through a 300-mesh stainless-steel sieve. Then, the MIH powders were put into a Soxhlet extractor and were washed with a solution of methanol:acetic acid (9:1, v/v) for 60 h until the New Red could not be detected in the MIHs by SERS.
By comparison, positively charged Au NPs-containing non-imprinted hydrogels (NIHs) were synthesized according to the same procedure in the absence of New Red. All hydrogels were vacuum dried overnight before they were used in re-binding studies.

Detection of New Red using Positively Charged Au NPs-Embedded MIHs as a Substrate
For the re-binding studies of New Red, 5.0 mg of positively charged Au NPs-embedded MIHs were mixed with 1.0 mL of New Red aqueous solutions at different concentrations (1.64 × 10 −7 , 1.64 × 10 −6 , 1.64 × 10 −5 , 8.18 × 10 −5 , 1.64 × 10 −4 M), and the mixture was incubated at room temperature under shaking for 2 h. Then, the mixture was centrifuged at 8000 rpm for 10 min and the supernatant was removed. The precipitated tiny pieces of swelled hydrogels were washed with distilled water and separated by centrifugation. Then, the swelled hydrogels were transferred to a concave glass slide with pipettes and observed under the microscope of Raman spectrometer. The spectrum was recorded, and each spectrum are an average of 10 independent results taken at an integration time of 10 s. The logarithmic intensity of the characteristic peak of New Red at 1572 cm −1 was used to plot against the logarithmic concentration of New Red to build a calibration curve. The detection of New Red using positively charged Au NPs-embedded NIHs as a substrate was performed according to the same procedure described above except that the MIHs were replaced with NIHs.

Application of the Positively Charged Au NPs-Embedded MIHs in Food Samples
A beverage sample (Gatorade, orange flavored) used in this work was purchased from a local supermarket in Hangzhou, Zhejiang, China. New Red was not detected in the origin sample. To obtain a calibration curve for the detection of New Red, the sample solution was centrifuged at 10,000 rpm for 10 min and then filtered through a 0.25 µm filter membrane. A series of aliquot samples (1.0 mL) were spiked with New Red at concentrations of 1.64 × 10 −6 , 8.18 × 10 −5 , and 1.64 × 10 −4 M, respectively. Then, the New Red-spiked samples were handled and determined by the procedures described in Section 2.5. Recoveries were calculated and three parallel analyses were performed for each beverage sample. The data are expressed as the mean ± standard deviation (SD). Figure 2a shows the TEM image of positively charged Au NPs. Obviously, the Au NPs have a size range of about 10-50 nm with a number-average diameter of 44.3 ± 6.8 nm (measured by Nano measurer 1.2). The zeta potentials of the Au NPs were determined to be 46.6 ± 1.9 mV by DLS ( Figure 2b), indicating that the Au NPs were highly positively charged, and the Au NP solution was stable. Figure 2c demonstrates that the hydrodynamic diameter of the positively charged Au NPs detected by DLS was monomodal with an average size of 49.0 ± 0.9 nm, which was comparable to the average diameter of Au NPs measured by TEM. The EDX spectrum of the Au NPs (Figure 2d) displays Au signal peaks from Au nanoparticles and Cu peaks as well as C peaks from the ultrathin carbon-coated copper grid. Figure 2a shows the TEM image of positively charged Au NPs. Obviously, the Au NPs have a size range of about 10-50 nm with a number-average diameter of 44.3 ± 6.8 nm (measured by Nano measurer 1.2). The zeta potentials of the Au NPs were determined to be 46.6 ± 1.9 mV by DLS (Figure 2b), indicating that the Au NPs were highly positively charged, and the Au NP solution was stable. Figure 2c demonstrates that the hydrodynamic diameter of the positively charged Au NPs detected by DLS was monomodal with an average size of 49.0 ± 0.9 nm, which was comparable to the average diameter of Au NPs measured by TEM. The EDX spectrum of the Au NPs (Figure 2d) displays Au signal peaks from Au nanoparticles and Cu peaks as well as C peaks from the ultrathin carbon-coated copper grid.

Characterization of the Positively Charged Au NPs in MIHs
Dry MIHs were ground into fine powders, ultrasonically dispersed in tetrahydrofuran (THF) at room temperature, and transferred onto a carbon-coated copper grid by dipping. The elemental distribution of MIHs was then determined by EDX mapping technique in high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) mode. In a selected area (outlined in red in Figure 3a

Characterization of the Positively Charged Au NPs in MIHs
Dry MIHs were ground into fine powders, ultrasonically dispersed in tetrahydrofuran (THF) at room temperature, and transferred onto a carbon-coated copper grid by dipping. The elemental distribution of MIHs was then determined by EDX mapping technique in high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) mode. In a selected area (outlined in red in Figure 3a), the corresponding element mapping of gold (Figure 3c

SERS Selectivity of the Positively Charged Au NPs-Embedded MIHs
To investigate the SERS selectivity of the positively charged Au NPs-embedded MIHs, both the MIHs and the positively charged Au NPs-embedded NIHs were used to adsorb New Red from an aqueous solution of New Red at a concentration of 8.18×10 −5 M, followed by SERS measurements using the two swelled hydrogels as substrates. Figure 4a presents the SERS spectrum of New Red solid powders. Figure 4b shows the spectrum obtained using the Au NPs-embedded MIHs as a substrate. Strong characteristic signals of New Red at 1572 cm −1 , 1513 cm −1 , 1439 cm −1 , 1359 cm −1 , 1238 cm −1 , and 1144 cm −1 were clearly observed. By contrast, Figure 4c displays the SERS spectrum using Au NPs-embedded NIHs as a substrate, and no notable signal of New Red could be detected. These results demonstrate that the selectivity of the MIHs towards the detection of New Red was quite satisfactory.

SERS Selectivity of the Positively Charged Au NPs-Embedded MIHs
To investigate the SERS selectivity of the positively charged Au NPs-embedded MIHs, both the MIHs and the positively charged Au NPs-embedded NIHs were used to adsorb New Red from an aqueous solution of New Red at a concentration of 8.18×10 −5 M, followed by SERS measurements using the two swelled hydrogels as substrates. Figure 4a presents the SERS spectrum of New Red solid powders. Figure 4b shows the spectrum obtained using the Au NPs-embedded MIHs as a substrate. Strong characteristic signals of New Red at 1572 cm −1 , 1513 cm −1 , 1439 cm −1 , 1359 cm −1 , 1238 cm −1 , and 1144 cm −1 were clearly observed. By contrast, Figure 4c displays the SERS spectrum using Au NPs-embedded NIHs as a substrate, and no notable signal of New Red could be detected. These results demonstrate that the selectivity of the MIHs towards the detection of New Red was quite satisfactory.

SERS Selectivity of the Positively Charged Au NPs-Embedded MIHs
To investigate the SERS selectivity of the positively charged Au NPs-embedded MIHs, both the MIHs and the positively charged Au NPs-embedded NIHs were used to adsorb New Red from an aqueous solution of New Red at a concentration of 8.18×10 −5 M, followed by SERS measurements using the two swelled hydrogels as substrates. Figure 4a presents the SERS spectrum of New Red solid powders. Figure 4b shows the spectrum obtained using the Au NPs-embedded MIHs as a substrate. Strong characteristic signals of New Red at 1572 cm −1 , 1513 cm −1 , 1439 cm −1 , 1359 cm −1 , 1238 cm −1 , and 1144 cm −1 were clearly observed. By contrast, Figure 4c displays the SERS spectrum using Au NPs-embedded NIHs as a substrate, and no notable signal of New Red could be detected. These results demonstrate that the selectivity of the MIHs towards the detection of New Red was quite satisfactory.

SERS Specificity of the Positively Charged Au NPs-Embedded MIHs
To evaluate the specificity of MIHs for the detection of New Red, New Red and its structural analogues (Amaranth, Erythrosine B sodium, and Allura Red) were chosen, and 5.0 mg of the MIHs were incubated with 1.0 mL of aqueous solution of each colorant at a concentration of 1.64 × 10 −5 M for 2 h. Then, the MIHs were used as substrates to measure SERS spectrum of each colorant. Figure 5a-d demonstrates the SERS spectra using the MIHs as substrates after the MIHs were, respectively, incubated with the solutions of New Red, Erythrosine B sodium, Amaranth, and Allure Red at the concentration of 1.64 × 10 −5 M. Obviously, the four colorants can be readily distinguished according to their corresponding spectrum fingerprint. Strong characteristic signals of New Red at 1572 cm −1 , 1513 cm −1 , and 1439 cm −1 were observed in Figure 5a. Fairly weak signals of Erythrosine B sodium were detected in Figure 5b. Both Amaranth (Figure 5c) and Allura red (Figure 5d) show no signals at 1572 cm −1 , 1513 cm −1 , and 1439 cm −1 . The results indicate that the three structural analogues did not interfere the detection of New Red when using the positively charged Au NPs-embedded MIHs as substrates and the MIHs indeed possessed imprinted cavities. MIHs were prepared using New Red as a template and, thus, MIHs contained a lot of specific cavities after New Red molecules were removed by washing. These three-dimensional cavities were complementary to New Red molecules in terms of shape recognition, hydrogen bonding, and electrostatic interactions. As a result, Erythrosine B sodium, Amaranth, and Allura Red molecules were very difficult to diffuse into the cavities, and accordingly, cause no interferences to the detection of New Red.

SERS Specificity of the Positively Charged Au NPs-Embedded MIHs
To evaluate the specificity of MIHs for the detection of New Red, New Red and its structural analogues (Amaranth, Erythrosine B sodium, and Allura Red) were chosen, and 5.0 mg of the MIHs were incubated with 1.0 mL of aqueous solution of each colorant at a concentration of 1.64 × 10 −5 M for 2 h. Then, the MIHs were used as substrates to measure SERS spectrum of each colorant. Figure 5a-d demonstrates the SERS spectra using the MIHs as substrates after the MIHs were, respectively, incubated with the solutions of New Red, Erythrosine B sodium, Amaranth, and Allure Red at the concentration of 1.64 × 10 −5 M. Obviously, the four colorants can be readily distinguished according to their corresponding spectrum fingerprint. Strong characteristic signals of New Red at 1572 cm −1 , 1513 cm −1 , and 1439 cm −1 were observed in Figure 5a. Fairly weak signals of Erythrosine B sodium were detected in Figure 5b. Both Amaranth (Figure 5c) and Allura red (Figure 5d) show no signals at 1572 cm −1 , 1513 cm −1 , and 1439 cm −1 . The results indicate that the three structural analogues did not interfere the detection of New Red when using the positively charged Au NPs-embedded MIHs as substrates and the MIHs indeed possessed imprinted cavities. MIHs were prepared using New Red as a template and, thus, MIHs contained a lot of specific cavities after New Red molecules were removed by washing. These three-dimensional cavities were complementary to New Red molecules in terms of shape recognition, hydrogen bonding, and electrostatic interactions. As a result, Erythrosine B sodium, Amaranth, and Allura Red molecules were very difficult to diffuse into the cavities, and accordingly, cause no interferences to the detection of New Red.       Table 1 summarizes the recovery and the relative standard deviation (RSD) of the detection of New Red. It can be seen that the recoveries were found to be 86.3% to 100.6% and the values of RSD were in the range of 2.3% to 7.7%. The high recovery and low RSD values achieved in this study illustrated good precision and accuracy of the detection of New Red based on SERS using the MIHs as substrates.

Detection of New Red Spiked in a Sports Drink Using the MIHs as a Substrate
The MIHs were further applied to detect New Red spiked in a sports drink (Gatorade, orange flavored). As shown in Table 2, the recoveries of samples were found to be in the range of 91.0% to 97.1% and the values of RSD were between 1.8% and 7.7%. The good recovery and RSD indicated that the positively charged Au NPs-embedded MIHs were suitable for the detection of New Red in real samples. As the data collection time for a single SERS measurement is 10 s, the total detection time for a real sample would be less than 2.5 h even when the incubation (2 h) and centrifugation (10 min) pretreatments are included.  Table 1 summarizes the recovery and the relative standard deviation (RSD) of the detection of New Red. It can be seen that the recoveries were found to be 86.3% to 100.6% and the values of RSD were in the range of 2.3% to 7.7%. The high recovery and low RSD values achieved in this study illustrated good precision and accuracy of the detection of New Red based on SERS using the MIHs as substrates.

Detection of New Red Spiked in a Sports Drink Using the MIHs as a Substrate
The MIHs were further applied to detect New Red spiked in a sports drink (Gatorade, orange flavored). As shown in Table 2, the recoveries of samples were found to be in the range of 91.0% to 97.1% and the values of RSD were between 1.8% and 7.7%. The good recovery and RSD indicated that the positively charged Au NPs-embedded MIHs were suitable for the detection of New Red in real samples. As the data collection time for a single SERS measurement is 10 s, the total detection time for a real sample would be less than 2.5 h even when the incubation (2 h) and centrifugation (10 min) pretreatments are included. The general detection parameters reported in the literature for the determination of New Red are outlined in Table 3. Obviously, the linear range of current method is comparable with other reported methods, and the LOD of current method is better than that of HPLC-UV and HPLC-MS/MS.

Conclusions
A new method for the detection of New Red based on SERS using positively charged Au NPs-embedded MIHs as substrates has been established. The MIHs which were synthesized via a free radical polymerization of acrylamide aqueous solution containing positively charged Au NPs, New Red, and N,N'-methylenebis(acrylamide) exhibited high selectivity and specificity to New Red. The linear range was 1.64 × 10 −6 to 1.64 × 10 −4 M and the LOD was 1.64 × 10 −7 M for the detection of New Red spiked in water, and the corresponding recoveries ranged from 86.3% to 100.6% with RSD in the range of 2.3% to 7.7%. The RSD and recovery rates for the detection of New Red spiked in a sports drink sample (Gatorade, orange flavored) were 1.8% to 7.7% and 91.0% to 97.1%, respectively.