3.1. Optimization of HPLC condition
The excitation and emission wavelengths of RhB and EsY were measured using a spectrofluorimeter, as shown in Supplementary Fig. 1 (a and b), and applied to HPLC-FLD. Several trials have been conducted to identify the best excitation and emission wavelengths for the simultaneous quantification of RhB and EsY. RhB and EsY exhibited native fluorescence characterized by excitation wavelengths of 555 nm and 535 nm with emission wavelengths of 578 nm and 545 nm, respectively, as shown in Supplementary Fig. 1 (a and b). The obtained values of the excitation and emission wavelengths of RhB and EsY show a wide difference in excitation and emission wavelengths between the two dyes, which clearly appears in the difference in peak area and intensity values, as shown in Fig. 2 and Supplementary Fig. 1, 2, 3, and 4. To overcome these issues in HPLC-FLD, an excitation wavelength of 535 nm for RhB and EsY was selected as a relatively suitable wavelength. However, the difference in the emission wavelength between RhB and EsY can be overcome by using the zero-order emission mode in the presented FLD (Model Dionex Ultimate 3000) subsequent to excitation at a chosen wavelength of 535 nm. The proposed detector has two advantages, including the use of dual photomultiplier tube detectors (dual-PMT), which have benefits over the single PMT represented in emission wavelength expansion without any influence on the detection sensitivity. Furthermore, all wavelengths (not just a single wavelength) emitted from the sample were collected and reflected onto the detector at a fixed excitation wavelength[28]. Therefore, the advantages of the zero-order emission mode were exploited to achieve a significant peak area and intensity for RhB and EsY after excitation at the chosen wavelength of 535 nm, as depicted in Fig. 2 and Supplementary Fig. 2. Furthermore, the sensitivity of the proposed method significantly improved, as indicated by the LOD values for RhB and EsY (Table 1).
Table 1
Results of assay validation parameters of the proposed HPLC method for the simultaneous determination of RhB and EsY.
Parameter
|
RhB
|
EsY
|
Range
|
0.5–300 ng/ mL
|
0.5–300 ng/ mL
|
Slope
|
288.03
|
279.23
|
Intercept
|
-194.85
|
286.1
|
r
|
0.9994
|
0.9991
|
LOD
|
0.344 ng/mL
(0.355 ng/g)
|
0.215 ng/mL
(0.222 ng/g)
|
LOQ
|
1.145 ng/mL
(1.181 ng/g)
|
0.715 ng/mL
(0.737 ng/g)
|
Recovery
|
98.38-101.86
|
100.15-101.09
|
RSD%
|
0.83–1.66
|
1.10–1.66
|
The separation efficiency of RhB and EsY was investigated at three columns included Thermo Scientific ACCLAIM™ 120 C8 (4.6 x 150 mm, 5 µm), Thermo Scientific ACCLAIM™ 120 C18 (4.6 x 150 mm, 5 µm) and Agilent Technologies ZORBAX Eclipse Plus C18 (4 x 100 mm, 5 µm). The last column, maintained at 40°C, was chosen for further optimization because it resulted in valuable separation of RhB and EsY in a reasonable time. Several mobile phases, comprising of 0.1% acetic acid in water/ acetonitrile and methanol, 0.1% phosphoric acid in water/ acetonitrile and methanol, and 0.1% formic acid in water/ acetonitrile and methanol, were tested. A 0.1% formic acid in water (A) and methanol (B) with ratio of 30:70 v/v in an isocratic elution mode at flow rate 0.8 mL/min revealed valuable resolution peaks of RhB and EsY. Under these conditions, the total time required to separate the RhB and EsY mixtures was 7 min, with retention times of 1.78 and 5.67 min, respectively.
3.2. The optimization of the extracting conditions
Several solvents (water, acetonitrile, methanol, acetone, ethanol, and isopropanol) were examined as extraction solvents for RhB and EsY in bottled chili sauce. A mixture of water and methanol (80: 20% v/v) was found to be suitable solvent for the extraction of studied dyes with minimal interference, as shown in Supplementary Fig. 5. In addition, the aforementioned solvents were tested as elution solvents for dyes from the Sep-Pak® C18 cartridge. An Effective extraction and good recoveries for the studied dyes were obtained using 10 mL of a solution composed of ultrapure water and acetonitrile (80: 20% v/v), without the need for any additional clean up procedures.
3.3. Method validation
The linearity, limit of detection (LOD), limit of quantification (LOQ), recovery, and relative standard deviation (RSD) were evaluated to validate the HPLC-FLD method based on preliminary research specificity matrix effects. The external standard method was applied to quantify RhB and EsY, and the concentration of each dye was tested in the range of 0.5–300 ng/mL. Calibration curves were constructed using the direct proportional relationships between the peak areas and dye concentrations. The proposed method has excellent linearity for RhB and EsY in the concentration range of 0.5–300 ng/mL with correlation coefficients (r) ≥ 0.9991, as shown in Table 1. The limits of detection (LOD) and quantification (LOQ) were defined as three and ten times the intercept standard deviation of a regression line over the average slope of the calibration curve, respectively, repeated in triplicate. The LODs for RhB and EsY were 0.344 and 0.215 ng/ml, respectively. In contrast, the LOQs for RhB and EsY were 1.145 and 0.715 ng/mL, respectively. The recovery % and relative standard deviation (RSD %) are terms referring to accuracy and precision and are calculated using the following formulas: measured value/theoretical value ×100 and SD of measured values/mean measured value × 100, respectively. At three concentration levels, average recoveries of RhB and EsY ranged from 98.38 to 101.86% and 100.15 to 101.09%, respectively, as shown in Table 1. While The average RSD % was in the range of 0.83–1.66% for RhB and 1.10–1.66% for EsY, as shown in Table 1). The aforementioned sensitivity, accuracy, and precision values of the method were sufficient for the confirmation and determination of RhB and EsY in the real samples. The specificity of the method was examined by analyzing eight bottled chili sauce samples, in addition to using the spike method. As shown in Figs. 3 and 4, there was no interferences in the retention time region in which RhB and EsY were eluted.
3.4. Analysis of samples and recovery
The manufactured chili sauce (eight samples) was used as a real sample to verify the performance and efficiency of the HPLC-FLD-zero-order emission mode method and the Sep-Pak® C18 cartridge for solid-phase extraction and preconcentration of RhB and EsY. RhB and EsY were detected only in one sample, as shown in Fig. 3 and Supplementary Fig. 5, at concentrations of 0.388 and 0.0156 ng/ml, respectively. Therefore, the sample was fortified with two concentration levels of RhB and EsY to ensure its positivity, as shown in Fig. 3. Spiked manufactured chili sauce samples with RhB and EsY at three different levels were prepared for additional tests to evaluate the accuracy of the proposed HPLC-FLD-zero-order emission mode method and the efficiency of the extraction cartridge. Figure 4 shows the chromatograms of the blank and spiked chili sauce samples. The amount of RhB and EsY in the spiked samples was measured. The recovery percentage and RSD values for two dyes varied from 85.04 to 96.97% and from 1.83 to 2.99%, respectively. These findings show that the proposed method and extraction cartridge are applicable for the measurement of RhB and EsY in real samples.
3.5. Comparison with other methods
The values of the linear range and LOD for the proposed approach for the simultaneous quantification of RhB and EsY were compared with those of previously reported approaches, as shown in Table 2. The LOD values of RhB obtained in this study were far lower than those of on-chip SPE spectrophotometry [14], MD µ-SPE spectrophotometry [15], SPE spectrophotometry [29], polyamide-SPE-HPLC-UV [20], C18-SPE- LC-UV [21], Fe3O4 @COF-MSPE HPLC-UV [19], HPLC-UV [27], HPLC-FLD [2] and UHPLC-HRMS [26]. However, it was higher than that of MSPE-fluorescence [17], HPLC–ESI–MS/MS [22], UHPLC-MS/MS [23] and HPLC-MS/MS [25]. In general, the presented method has satisfactory sensitivity for the estimation of RhB in real samples compared with the reported method. The LOD of EsY obtained using this approach is far lower than that of all reported methods. The results obtained for the simultaneous determination of RhB and EsY showed that the developed Sep-Pak® C18 SPE-HPLC-FLD-zero-order emission mode method was satisfactory.
Table 2
Comparison of the proposed method with some reported methods
Analyte
|
Method
|
Linear range
|
LOD
|
Ref.
|
Rhodamine B
|
On-chip SPE- Spectrophotometry
|
10–800 ng/mL
|
7 ng/mL
|
[14]
|
MD- µ-SPE -Spectrophotometry
|
5-200 ng/mL
|
1.6 ng/mL
|
[15]
|
SPE- Spectrophotometry
|
0.25-3.0 µg/mL
|
3.14 ng/mL
|
[29]
|
MSPE-Fluorescence
|
0.40–140.00 ng/mL
|
0.06 ng/mL
|
[17]
|
polyamide-SPE- HPLC-UV
|
0.2–50 µg/mL
|
0.107 µg/mL
|
[20]
|
C18-SPE- LC-UV
|
0.05–10 mg/ mL
|
0.14 µg/g
|
[21]
|
Fe3O4 @COF-MSPE- HPLC-UV
|
0.05–5 µg/mL
|
3.8 ng/mL
|
[19]
|
HPLC-UV
|
0.025 − 20 µg/mL
|
0.14 µg/g
|
[27]
|
HPLC-FLD
|
0.5–10 ng/mL
|
3.7 ng/g
|
[2]
|
UHPLC-HRMS
|
0.5 − 100 ng/mL
|
1.67 ng/g
|
[26]
|
HPLC–ESI–MS/MS
|
0.2–10 ng/mL
|
0.05 ng/g
|
[22]
|
UHPLC-MS/MS
|
0.5 to 100 ng/mL
|
0.1 ng/mL
|
[23]
|
HPLC-MS/MS
|
0.05–100 µg/g
|
0.05 µg/g
|
[25]
|
SPE-HPLC-FLD-zero order emission mode
|
0.5–300 ng/ mL
|
0.344 ng/mL
(0.355 ng/g)
|
This work
|
Eosin Y
|
Spectrophotometry
|
0.115–2.53 µg/mL
|
0.03 µg/mL
|
[13]
|
HPLC-UV
|
0.025 − 20 µg/mL
|
0.14 µg/g
|
[27]
|
SPE-HPLC-FLD-zero order emission mode
|
0.5–300 ng/ mL
|
0.215 ng/mL
(0.222 ng/g)
|
This work
|