Recently, food and pharmaceuticals safety has attracted significant attention as food additives have become increasingly common [1, 2]. Pigment additives are added to food to improve appearance, enhance aesthetic appeal, and maintain the natural color during processing or storage [3]. However, illegal synthetic dyes, such as auramine O, are used as food and herb pigments, posing a threat to public health. Auramine O has been found in beans, meats, and condiments in China [4, 5]. Auramine O, bis [4-(dimethylamino)phenyl] methaniminium chloride, is a yellow dye that is widely used as colorant in the production of paper, textiles, leather, incense, and paint [6, 7]. Auramine O has several advantages over natural dyes, including lower price, more intense color, increased colorfastness, and greater stability [8]. However, it is also carcinogenic and toxic to humans and has been classified as a group 2B carcinogen (possibly carcinogenic to humans) by the International Agency for Research on Cancer (IARC) [9]. Therefore, auramine O is forbidden as an additive in foods by many national and international food regulations. The European Community has emphasized that food additive legislation didn't allow the use of colors other than those specifically authorized by Directive 94/36/EC [10]. Consequently, it is important to identify auramine O in food and herb samples to regulate auramine O abuse.

It is also important to analyze impurities in auramine O samples because when the specific impurities are identified, their generation can be prevented. In addition, the synthetic method can be inferred and the source of the impurities proposed according to information obtained from the impurities detected in the samples. In recent years, the high performance liquid chromatography-ion trap-time of flight mass spectrometry (HPLC-IT-TOF-MS) method to identify impurities in drugs has been reported. Li et al. [11] identified the ten impurities of drug 2C-E and deduced its synthetic route by HPLC-IT-TOF-MS. Ma et al. [12] established a HPLC-IT-TOF-MS method for the identification of the two impurities in the drug o-chlorophenyl cyclopentyl ketone.

For the quantitative determination process of illegal food additives using HPLC, the related reference materials are usually playing an important role in accurate quantification [13]. However, no domestic auramine O reference materials exist that can meet laboratory accreditation requirements at present. This will lead to uncertainty in detection and prevent accurate determination. Therefore, the preparation of auramine O standards is extremely urgent.

Preparative HPLC (prep-HPLC) is an effective method for isolation and purification of specific compounds [14-16]. Therefore, this study aims to establish a method for preparing reference standards for auramine O by prep-HPLC and identify common impurities in auramine O samples.

1 Experimental

1.1 Chemicals and reagents

All organic solvents used for preparative and analytical HPLC analysis were of HPLC grade. Methanol and trifluoroacetic acid were purchased from J & K Chemical Ltd. (Beijing, China). Ultrapure water from Wahaha Ltd. (Hangzhou, China) was used during both preparative and analytical HPLC analyses. The auramine O sample was obtained from Heowns Biochem Technologies Ltd. (Tianjin, China).

1.2 Sample preparation

First, 1 mg of auramine O was dissolved in 1 mL of methanol for analytical HPLC analysis. For preparative HPLC analysis, a suitable amount of the auramine O sample was dissolved in 10 mL of methanol-ultrapure water (1 : 1, v/v). All sample solutions were placed in an ultrasonic bath for 5 min and subsequently filtered using a 0.45 μ m nylon membrane filters before analytical and preparative HPLC analyses.

1.3.1 HPLC

The analytical HPLC equipment used was as follows: Shimadzu LC-20AVP system equipped with two LC-20AT solvent pumps; SPD-M20AVP UV-VIS photodiode array detector (DAD); Model 7 725 injection valve with a 20 μ L loop and a Shimadzu SIL-20A auto-sampler; SCL-20AVP system controller; Class-VP-LC work station (Shimadzu, Kyoto, Japan). An Agilent Eclipse XDB-C18 column (150 mm×4.6 mm, 5 μ m) was used for analysis. A mobile phase consisting of methanol-0.5% (v/v) trifluoroacetic acid aqueous solution (55 : 45, v/v) was used at a flow rate of 1.0 mL/min. The column temperature was maintained at 30 ℃ and the detection wavelength set to 254 nm.

1.3.2 HPLC-IT-TOF-MS

HPLC experiments were performed using a Shimadzu (Kyoto, Japan) HPLC system consisting of a solvent delivery pump (LC-20 AD), autosampler (SIL-20 AC), DGU-20A3 degasser, photodiode array detector (SPD-M20A), communication base module (CBM-20A), and column oven (CTO-20A). An ODS-SP column (150 mm×4.6 mm, 5 μ m) was used for the analysis with a mobile phase consisting of methanol-0.5% (v/v) trifluoroacetic acid aqueous solution (55 : 45, v/v) at a flow rate of 0.5 mL/min. The sample chamber in the autosampler was maintained at 4 ℃, while the column was set to 40 ℃ and the sample injection volume was 5 μ L.

An ion trap-time of flight instrument (Shimadzu Corp., Kyoto, Japan) equipped with an electrospray ionization (ESI) source in positive ion mode was used for the identification of the impurity in the auramine O sample. The optimized MS conditions were as follows: positive electrospray voltages, 4.5 kV; flow rate of nebulizing gas (N₂), 1.5 L/min; drying gas (N₂) pressure, 0.1 MPa; curve dissolution line (CDL) temperature, 200 ℃; block heater temperature, 200 ℃; ion accumulation time, 60 ms; detector voltage, 1.75 kV; pressure of TOF and IT, 1.7×10^-4 Pa and 1.8×10^-2 Pa. Mass spectrometric analyses were performed by full-scan MS with a mass range of m/z 50-1 000 and data-dependent MS/MS acquisition on the suspected impurity ions.

Mass calibration of the HPLC-IT-TOF-MS was achieved using a trifluoracetic acid sodium solution (2.5 mmol/L) from 50 to 1 000 (m/z) at a flow rate of 5 mL/min. The tuning operation was set in autotuning mode and the result was saved as a tuning file. All calculated mass errors were less than 10 ppm (10×10^-6) after mass calibration with the reference standard.

1.4 Prep-HPLC separation

The prep-HPLC was performed using a FLEXA purification system (Agela, China). Firstly, a Venusil XBP C18 preparative column (250 mm×50 mm, 10 μ m) was used to separate auramine O. Mobile phase: (A) ultrapure water and (B) methanol. Linear gradient elution program: 0-30 min, 10%B-40%B. The flow rate was 80 mL/min and the monitored wavelength was 254 nm. Subsequently, a Waters Prop C18 column (250 mm×20 mm, 5 μ m) was used to secondarily separate auramine O. A mobile phase consisting of methanol-ultrapure water (55 : 45, v/v) was used at a flow rate of 8.0 mL/min and the injection volume was 500 μ L.

1.5 Identification of auramine O

The auramine O was identified by UV, IR, LC-MS, ¹H NMR, and ¹³C NMR at the Analysis Center of the Beijing University of Chemical Technology.

2 Results and discussion

2.1 Identification of the impurity from auramine O

During the separation and purification process, a single impurity from the auramine O sample was difficult to remove. HPLC-IT-TOF-MS analysis was used for the identification of this impurity in the auramine O sample. As shown in Fig. 1a, the impurity is observed at an elution time of approximately 5 min. In positive ion mode, the MSⁿ data of the impurity from the auramine O sample are shown in Fig. 1b. The protonated molecular ion [M+H]⁺ of the impurity was observed at m/z 254.141 7 (C₁₆H₂₀N₃⁺), which is 14.038 6 Da lower than that of auramine O. This indicates that the impurity is homologous to auramine O, differing by a single CH₂ unit. MS² analysis of the precursor ion m/z 254.141 7 (C₁₆H₂₀N₃ ⁺) yielded a product ion at m/z 133.073 8 (C₈H₁₀N₂⁺). The MS³ fragment ion at m/z 134.077 4 was formed through an obtained H from the ion at m/z 133.073 8. By elemental analysis and MS³ analysis according to the exact mass data, the most likely molecular formula of the impurity was determined to be C₁₆H₂₀ClN₃, 4-(imino(4-(methylamino)phenyl) methyl)-N, N-dimethylaniline hydrochloride. The likely structure of the impurity is shown in Fig. 1c.

Fig. 1

Fig. 1 (a) Chromatogram of auramine O sample, (b) Mass spectra of the impurity from the auramine O sample and (c) fragmentation pathway of the impurity

2.2 Analysis of the synthetic route

Many routes could be used to synthesize auramine O. A synthetic route was proposed according to the impurity analysis shown above, and the proposed synthetic route is shown in Fig. 2a, as was also described in the literature [17, 18]. The impurity was an intermediate product of the proposed synthetic method. Some possible sources of this impurity will also be discussed. N, N-dimethylaniline, as a raw material for the synthesis of auramine O, may contain N-methylaniline impurities. This results in the inclusion of the above impurity in the final product, auramine O. In addition, when N, N-dimethylaniline was used as a raw synthesis material, its chemical component could react to produce the abovementioned impurity [19]. Depending on the nature of the material, the latter scenario was more likely because N, N-dimethylaniline and N-methylaniline exhibited large differences in polarity and are easily separated and purified. However, the difference in the polarity of the intermediates is small, so the intermediates were not easily separated.

Fig. 2

Fig. 2 Synthetic route of the (a) impurity and (b) auramine O

2.3 Preparation of auramine O using prep-HPLC

The prep-HPLC method described in section 1.4 was used to isolate pure auramine O from the auramine O samples. At a retention time of approximately 25 min, a fraction was collected in a brown bottle, as shown in Fig. 3a. The methanol solvent was removed by nitrogen blowing in a dark hood, and directly lyophilized in the lyophilizer. The purity of the prepared auramine O standard was 97.02%, as shown in Fig. 3b. Therefore, a smaller particle size column was selected for secondary purification and the prep-HPLC chromatogram is shown in Fig. 4a. The purity of the secondary auramine O standard was 99.52%, as shown in Fig. 4b. The purity was determined using the analytical HPLC area normalization method. In the mass balance method, the content of all impurities including nonvolatile impurities, moisture, volatile impurities, and ash was subtracted from 100% to calculate the purity of auramine O [20]. The mean moisture content was 0.34%, as determined by Karl Fisher method. The contents of the volatile impurities and ash were determined by thermal gravity analysis and loss upon drying in a muffle furnace. The ash content was determined to be 0.13%, while the volatile impurity content was very low (< 0.001%) and was omitted from the purity calculation. As a result, the purity of the auramine O standard was determined to be 99.05% by the mass balance method.

Fig. 3

Fig. 3 (a) Preparative and (b) analytical chromatograms of the auramine O after the first purification step

Fig. 4

Fig. 4 (a) Preparative and (b) analytical chromatograms of the auramine O after the second purification step

The auramine O standard chemical structure was confirmed by UV, IR, MS, and NMR. The maximum UV absorption peak was observed at 430 nm. The IR spectrum showed a ν _max (KBr) of 1 375, 2 870, 2 960, 1 597, 1 500, 1 450, and 830 cm^-1 indicating that the product was auramine O upon comparison with standard infrared spectral data from the BioRad/Sadtler IR Data Collection. The mass spectra [M-Cl] showed a peak at m/z 268.33 in positive ion mode, consistent with the literature [4]. The¹H NMR (500 MHz, CD₃COCD₃), δ ppm values were as follows: 7.59 (d, J=8.8 Hz, 4H, H-3), 6.69 (d, J=8.8 Hz, 4H, H-4), 2.98 (s, 12H, H-6);¹³C NMR (125 MHz, CD₃COCD₃), δ ppm: 175.31 (C-1), 117.38 (C-2), 113.36 (C-3), 137.39 (C-4), 150.80 (C-5). The NMR data also indicated that the purified compound was auramine O.

3 Conclusions

In this study, a HPLC-IT-TOF-MS method for the identification of the impurity in auramine O samples was established. The impurity was assigned by analysis of its MSⁿ spectra. A useful and reliable prep-HPLC method for preparing reference standard quality auramine O from crude auramine O samples was established. The established method is suitable for preparing useable quantities of pure auramine O as a reference standard.

References

[1]	Zhang H, Li Z, Chen T, et al. Spectrochim Acta A Mol Biomol Spectrosc, 2017, 184: 335. doi: 10.1016/j.saa.2017.05.017
[2]	Qi F F, Jian N G, Qian L L, et al. Anal Bioanal Chem, 2017, 409(24): 5697. doi: 10.1007/s00216-017-0510-8
[3]	Feng F, Zhao Y S, Yong W, et al. J Chromatogr B, 2011, 879(20): 1813. doi: 10.1016/j.jchromb.2011.04.014
[4]	Li J, Ding X M, Zheng J X, et al. J Sep Sci, 2014, 37(17): 2439. doi: 10.1002/jssc.v37.17
[5]	Mall I D, Srivastava V C, Agarwal N K. J Hazard Mater, 2007, 143(1): 386.
[6]	Wang Z B, Zhang L Y, Li N, et al. J Chromatogr A, 2014, 1348(12): 52.
[7]	Jafari M, Rahimi M R, Ghaedi M, et al. Chem Eng Res Des, 2017, 125: 408. doi: 10.1016/j.cherd.2017.07.011
[8]	Tung J C, Huang W C, Yang J C, et al. Environ Toxicol, 2017, 32(11).
[9]	Qi P, Liang Z A, Wang Y, et al. J Chromatogr A, 2016, 1437: 25. doi: 10.1016/j.chroma.2016.02.005
[10]	Zhai H Y, Li J M, Chen Z G, et al. Microchem J, 2014, 114: 223. doi: 10.1016/j.microc.2014.01.006
[11]	Li Y N, Wang M C, Li A X, et al. Anal Methods, 2016, 8: 8179. doi: 10.1039/C6AY02162J
[12]	Ma X M, Lan T, Li Y N, et al. Chinese Journal of Chromatography, 2018, 36(3): 268.
[13]	Guo Z, Zheng H, Lu Y Z, et al. Forensic Sci Int, 2012, 221(1/3): 120.
[14]	Matusheski N V, Wallig M A, Juvik J A, et al. J Agric Food Chem, 2001, 49(4): 1867. doi: 10.1021/jf0013860
[15]	Kumar V, Shitiz K, Chauhan R S, et al. J Plant Biochem Biotechnol, 2016, 25(2): 125. doi: 10.1007/s13562-015-0317-7
[16]	Searle P A, Glass K A, Hochlowski J E. J Comb Chem, 2004, 6(2): 175. doi: 10.1021/cc0340372
[17]	Jager W D, Prakken B J, Bijlsma J W, et al. J Immunol Methods, 2005, 300(1): 124.
[18]	Tan C J, Tong Y W. Anal Chem, 2007, 79(1): 299.
[19]	Zheng Y, Mao J C, Chen J, et al. RSC Adv, 2015, 5(62): 50113. doi: 10.1039/C5RA06773A
[20]	Ma K, Wang H F, Zhao M, et al. Anal Chim Acta, 2009, 650(2): 227.