Food Sci Biotechnol (2018) 27(3):877–882

https://doi.org/10.1007/s10068-018-0308-2

Simultaneous determination of synthetic food additives in kimchi

by liquid chromatography-electrospray tandem mass
spectrometry
Ho Jin Kim1 • Mi Jin Lee1 • Hye Jin Park1 • Hye Jin Kim1 • Soon Kil Cho1 •

Min Hee Jeong1

Received: 21 August 2017 / Revised: 15 November 2017 / Accepted: 3 January 2018 / Published online: 16 January 2018
Ó The Korean Society of Food Science and Technology and Springer Science+Business Media B.V., part of Springer Nature 2018

Abstract A new analytical method was developed for the The prevention of excessive fermentation by Lacto-
simultaneous determination of seven food additives (Pon- bacillus is a major concern of the kimchi industry [3]. Food
ceau 4R, Allura Red AC, Amaranth, 4-hydroxymethyl additives such as preservatives, nutrients, colors, flavors,
benzoic acid, ethyl-4-hydroxybenzoate, butyl-4-hydroxy- texturizing agents, and natural additives, are used to
benzoate, and saccharin sodium) in kimchi using high- improve taste or storage life [8]. Food companies use
performance liquid chromatography-electrospray ioniza- synthetic food additives for economy and efficiency, and
tion tandem mass spectrometry. The linearity, sensitivity, kimchi is no exception [7].
selectivity, precision, and accuracy of the method were Food additives used in kimchi include synthetic dyes,
validated. The limit of detection was 0.00004–0.24 lg/mL, preservatives such as para-hydroxybenzoic acid, and sac-
and the limit of quantification was 0.00012–0.8 lg/mL. charin, an artificial sweetener. However, synthetic dyes can
Recoveries ranged from 85.65 to 120.82%. The method cause cancer [8, 9] if taken in excess, and may also cause
was successful and may help to ensure food safety. allergies, asthma [10–12], DNA damage [13], hyperactivity
[14], and other disorders. Therefore, restrictions on syn-
Keywords Food additives
Kimchi
HPLC–MS/MS
thetic colors and food additives apply internationally
Food analysis
Food safety [15, 16]. Simultaneous analysis of multiple food additives
in foods is important in order to comply with regulations on
the import and export of processed foods, including kim-
Introduction chi, to investigate the intake of food additives, to provide a
basis for risk evaluation, and to respond to the increasing
Kimchi is a traditional Korean fermented food that is rich demand for safe food.
in vitamin C, dietary fiber, b-carotene, minerals, phyto- A number of methods for the determination of preser-
chemicals, and Lactobacillus species. Most Koreans eat vatives have been reported, including high-performance
kimchi every day as a side-dish [1]. Consumption of liquid chromatography (HPLC) [17, 18], gas chromatog-
kimchi is increasing worldwide [2, 3]. Various benefits of raphy (GC) [19], thin-layer chromatography (TLC)
kimchi, including anticarcinogenic [4], antioxidant [5], and [20, 21], capillary electrophoresis (CE) [22, 23], liquid
immune-stimulating effects [6], have been reported. chromatography-mass spectrometry (LC–MS) [24, 25], and
gas chromatography-mass spectrometry (GC–MS) [26].
In this study, we developed the first highly sensitive and
accurate LC–MS/MS method for the simultaneous analysis
Ho Jin Kim and Mi Jin Lee have contributed equally to this work. of 7 food additives from kimchi: Ponceau 4R, Allura Red
AC, Amaranth, 4-hydroxymethyl benzoic acid, ethyl-4-
& Min Hee Jeong hydroxybenzoate, butyl-4-hydroxybenzoate, and saccharin
miniya33@korea.kr
sodium. We also validated analytical parameters such as
1
National Agricultural Products Quality Management Service, sensitivity, precision, mass accuracy, selectivity and lin-
Kimchun 740-871, Republic of Korea earity. This method has been successfully applied to

123
878 H. J. Kim et al.

monitoring and ensuring the safety of kimchi samples from Method validation
a local market.
The analytical method was validated in terms of linearity,
accuracy, precision, limit of detection (LOD) and limit of
Materials and methods quantification (LOQ). Quantitative analysis was carried out
using external standard calibration. Calibration solutions
Chemicals were prepared by appropriate dilution of intermediate
mixed standard solutions in acetonitrile.
Ponceau 4R, Allura Red AC, Amaranth, 4-hydroxymethyl Accuracy was evaluated in terms of percentage recov-
benzoic acid, ethyl-4-hydroxybenzoate, butyl-4-hydroxy- ery. For recovery studies, blank kimchi samples were
benzoate, saccharin, ammonium formate, and formic acid spiked prior to the extraction step (sample extraction). A
were purchased from Sigma-Aldrich (St. Louis, MO, weighed sample was added to a small volume of working
USA). HPLC-grade methanol and acetonitrile were solution of analyte. The extraction was then carried out as
obtained from Fisher (Pittsburgh, PA, USA). Membrane previously described. For precision, three concentrations
filters (0.22 lm) were purchased from Whatman (Maid- (2, 10, and 20(lg/mL) were investigated. Each experiment
stone, UK). was conducted three times, and the data were averaged.
The average recovery and relative standard deviation
Sample extraction (RSD) were calculated. The sensitivity of the method was
evaluated by estimating LOD and LOQ at signal-to-noise
Kimchi samples were purchased from local supermarkets ratios of 3 and 10, respectively. The intra- and inter-day
in Korea. Approximately 5 g was weighed into a 50 mL precisions of the methods were examined by analysis of six
flask; 25 mL of 60% acetonitrile were added, followed by replicates of standard solution for each of three concen-
sonication for 10 min. The extract was centrifuged at trations (2, 10 and 20 lg/mL) over 3 days. Intra-day and
3000 rpm (10 min) then filtered through a 0.22 lm mem- inter-day was assessed by analyzing standard solutions.
brane prior to LC–MS/MS.

LC–MS/MS analysis Results and discussion

LC–MS/MS analysis was performed on an Agilent 1260 Method validation

HPLC system coupled with an Agilent 6460 triple quad-
rupole mass spectrometer (Agilent Technologies, Diegem, Individual analyte solutions were directly infused into the
Belgium). We used a Cadenza CD C18 column, mass spectrometer in positive and negative ion scan modes.
100 mm 9 2 mm with a 3-lm particle diameter (Imtakt, Under the selected ESI conditions, all compounds showed
Kyoto, Japan). The injection volume was 3 lL, and the higher sensitivity in negative than in positive mode; the
oven was held at room temperature. Mobile phase A was most abundant ion was [M–H]- for all analytes. Therefore,
20 mM ammonium formate and 0.1% formic acid in water, the deprotonated form of each compound (the [M–H]- ion)
mobile phase B was 7:3 (by volumn) acetonitrile:methanol was selected as the parent ion for Q1 spectra and used as
with 0.1% formic acid. The optimized chromatographic the precursor for Q3 product ion spectra. The detailed data
conditions were as follows: 10% B to 0.1 min, linear gra- was shown in Fig. 1. Total ion chromatogram (TIC) and
dient to 100% B between 0.1 and 10.0 min, 100% B until MRM chromatogram of the compounds are shown in
13.0 min, and then return to 10% B in 1.0 min. The run Fig. 1. Figure 1(A) is the TIC of the seven food additives
time was 20.0 min, and the flow rate was 0.4 mL/min. The standards in acetonitrile and Fig. 1(B) is the seven food
triple quadrupole tandem mass spectrometer operated additives standards in blank kimchi sample.
under multiple reaction monitoring mode (MRM) for To evaluate the feasibility of HPLC–MS/MS for quan-
quantitative and qualitative analysis. The optimized elec- titative analyses of food additives in kimchi, the analytical
trospray ionization(ESI) conditions were as follows: ESI performance of the proposed method was examined. Linear
was carried out at atmospheric pressure in negative ion range, correlation coefficient (r2), LOD, LOQ, and recov-
mode; gas temperature, 150 °C; gas flow, 9 L/min; sheath ery are listed in Table 2. To test linearity, three concen-
gas temperature, 350 °C; sheath gas flow, 11 L/min; cap- trations of food additives within the range 2–10 lg/mL
illary voltage, 3000 V. Detailed MRM settings are illus- were prepared. Analysis was performed in triplicate.
trated in Table 1. Excellent linearity was achieved for each food additive,
with a linear regression coefficient of r2 C 0.9947. There
was no difference of primary slope and intercept of the

123
Simultaneous determination of synthetic food additives in kimchi 879

Table 1 Tandem mass spectrometry parameters of seven food additives

Compound Retention time Precursor ion Product ion Fragmentation Collision energy
(min) (m/z) (m/z) voltage (V) (V)

Allura Red AC 5.2 451 207 190 15

451 371 190 35
Ponceau 4R 4.2 537 302 150 10
537 429 150 15
Amaranth 3.3 537 317 230 25
537 457 230 25
Saccharin 3.5 182 42 130 25
182 106 130 25
4-hydroxymethyl benzoic acid 6.7 151 92 90 15
151 136 90 15
Ethyl-4-hydroxybenzoate 7.6 165 92 90 20
165 137 90 20
Butyl-4-hydroxybenzoate 9.2 193 92 110 25
193 137 110 25
The first transition was used for quantification and the second as a qualifier

calibration equation y = ax ? b of the between kimch

matrix matched calibration and acetonitrile calibration.
The sensitivity of the method was assessed by deter-
mining the LOD and LOQ. The LOD and LOQ were
estimated with concentrations giving a signal-to-noise ratio
of 3 and 10, respectively. As shown in Table 2, the LOD
ranged from 0.00004 to 0.24 lg/mL, while the LOQ ran-
ged from 0.00012 to 0.8 lg/mL.
Precision is expressed as percent relative standard
deviation (% RSD). Overall intra- and inter-day RSDs were
less than 5.0%. The results showed good reproducibility
and precision (Table 3). RSD is less than 5% may be found
more satisfactory than in SANCO guidelines, it can be said
that the precision of this method is acceptable. Recovery
was tested by adding mixed standard solutions at three
different concentrations (2, 10 and 20 lg/mL) to kimchi
samples. The results are shown in Table 3: recoveries were
in the range of 85.65–120.82%. It can be concluded that the
recovery values for the food additives in kimchi were
acceptable. In addition, in the SANCO guideline, the
recovery of the simultaneous multi component analysis is
acceptable 70–120% with slight difference range [27].
Therefore, the proposed analytical method was sufficiently
accurate. The validation data showed that the method
provides good linearity, sensitivity, selectivity, precision,
and accuracy for simultaneous analysis of seven food
additives in kimchi.

Fig. 1 HPLC-MS/MS total ion chromatogram (A) TIC of standards Application to real samples
in acetonitrile, (B) TIC of standards in blank kimchi sample,
(C) fragment ion (1 Amaranth, 2 Saccharin, 3 Ponceau 4R, 4 Allura
Red AC, 5 4-hydroxymethyl benzoic acid, 6 Ethyl-4-hydroxyben- In Korea, synthetic food additives are not permitted in
zoate, 7 Butyl-4-hydroxybenzoate) kimchi. To detect illegal food additives in kimchi and to

123
880 H. J. Kim et al.

Fig. 1 continued

123
Simultaneous determination of synthetic food additives in kimchi 881

Table 2 Linear equation, correlation coefficient, LOD, and LOQ for detection of seven food additives
Compound Linear equation Correlation coefficient LODa (lg/mL) LOQb (lg/mL)

Allura Red AC Y = 35,306.53X ? 57,300.64 0.9961 0.00042 0.00140

Ponceau 4R Y = 20,443.12X ? 4974.43 0.9963 0.24000 0.80000
Amaranth Y = 28,053.05X - 9582.95 0.9986 0.24000 0.80000
Saccharin Y = 48,924.87X ? 25,177.49 0.9968 0.03000 0.10000
4-hydroxymethyl benzoic acid Y = 146,657.51X ? 40,424.35 0.9974 0.00070 0.00230
Ethyl-4-hydroxybenzoate Y = 104,848.55X ? 42,716.31 0.9947 0.00004 0.00012
Butyl-4-hydroxybenzoate Y = 214,361.07X ? 202,419.14 0.9978 0.00011 0.00037
a
Limit of detection
b
Limit of quantification

Table 3 Recoveries and relative standard deviations (%) of seven food additives (n = 6, n = 18)
Compound Concentration (lg/mL) Precision (% RSDa) Recovery (%)b ± SDc
Intra-day (n = 6) Inter-day (n = 18)
Day Day 1 Day 2 Day 3

Allura Red AC 0.002 5.6 4.8 4.2 6.1 105.79 ± 1.4

0.02 4.2 4.8 4.7 5.1
0.1 5.3 5.0 5.2 3.8
Ponceau 4R 1 4.6 3.8 3.1 4.0 85.65 ± 3.5
10 4.8 3.5 4.5 4.3
50 4.9 4.1 4.0 4.3
Amaranth 1 3.6 3.8 3.1 4.0 96.78 ± 3.0
10 4.0 3.5 4.5 4.3
50 4.2 4.1 4.0 4.3
Saccharin 0.2 0.6 3.8 3.9 4.1 110.99 ± 2.4
2 0.9 3.2 4.0 3.9
10 1.1 4.1 3.3 4.3
4-hydroxymethyl benzoic acid 0.005 1.5 1.5 1.4 1.4 101.15 ± 4.2
0.05 1.6 1.6 1.3 1.3
0.25 1.4 1.1 1.4 1.1
Ethyl-4-hydroxybenzoate 0.0002 1.4 1.5 1.5 1.3 111.40 ± 1.5
0.002 1.1 1.7 1.8 1.4
0.01 1.5 1.4 1.3 1.2
Butyl-4-hydroxybenzoate 0.0005 1.4 1.9 2.0 1.2 120.82 ± 1.3
0.005 1.4 2.0 2.5 1.8
0.025 1.6 2.0 2.3 1.7
a
Relative standard deviation (%)
b
Average recoveries at three spike levels
c
Standard deviation

evaluate the effectiveness of the proposed method, the hydroxybenzoate, butyl-4-hydroxybenzoate, and saccharin
developed HPLC–MS/MS method was applied to the sodium) in kimchi purchased from local supermarkets in
analysis of seven food additives (Ponceau 4R, Allura Red Korea. Samples were analyzed in triplicate. In most cases,
AC, Amaranth, 4-hydroxymethyl benzoic acid, ethyl-4- the seven food additives were not found.

123
882 H. J. Kim et al.

Acknowledgements We are no funding received or any financial or community: a randomised, doubleblinded, placebocontrolled
material support. trial. LANCET 370: 1560–1567 (2007)
15. GSFA. Codex General Standard for food additives. Codex Stand
192- 1995 (1995)
16. Ministry of Health, P.R.C. Hygienic Standards for Uses of Food
References Additives, National Standard of the P.R.C., Beijing, 90 (2007)
17. Saad B, Bari MF, Saleh MI, Ahmad K, Talib MKM. Simulta-
1. Cheigh HS, Park KY. Biochemical, microbiological and nutri- neous determination of preservatives (benzoic acid, sorbic acid,
tional aspects of kimchi (Korean fermented vegetable products). methylparaben and propylparaben) in foodstuffs using high-per-
Crit. Rev. Food Sci. 34: 175–203 (1994) formance liquid chromatography. Chromatogr. A. 1073: 393–397
2. Hong Y, Yang HS, Li J, Han SK, Chang HC, Kim HY. Identi- (2005)
fication of lactic acid bacteria in salted Chinese cabbage by SDS- 18. Mikami E, Goto T, Ohno T, Matsumoto H, Nishida M. Simul-
PAGE and PCR-DGGE. Sci. Food Agric. 94:296–300 (2014) taneous analysis of dehydroacetic acid, benzoic acid, sorbic acid
3. Kim J, Bang J, Beuchat LR, Kim H, Ryu JH. Controlled fer- and salicylic acid in cosmetic products by solid-phase extraction
mentation of kimchi using naturally occurring antimicrobial and high-performance liquid chromatography. Pharm. Biomed.
agents. Food Microbiol. 32:20–31 (2012) Anal. 28: 261–267 (2002)
4. Choi MW, Kim KH, Kim SH, Park KY. Inhibitory effects of 19. Pan Z, Wang L, Mo W, Wang C, Hu W, Zhang J. Determination
kimchi extracts on carcinogen-induced cytotoxicity and trans- of benzoic acid in soft drinks by gas chromatography with on-line
formation in C3H/10T1/2 cells. Korean Soc. Food Sci. Nutr. 2: pyrolytic methylation technique. Anal. Chim. Acta. 545: 218–223
241–5 (1997) (2005)
5. Lee YM, Choi JS, Kwon MJ, Kim JK, Suh HS, Song YO. Iso- 20. Thompson AC, Hedin PA. Separation of organic acids by thin-
lation and identification of active principle in Chinese cabbage layer chromatography of their 2,4-dinitrophenylhydrazide
kimchi responsible for antioxidant effect. Korean J. Food Sci. derivatives and their analytical determination. Chromatogr. A.
Technol. 36:129–133 (2004) 21:13–18 (1966)
6. Kim JY, Lee YS. The effects of Kimchi intake on lipid contents 21. Khan SH, Murawski MP, Sherma J. Quantitative high perfor-
of body and mitogen response of spleen lymphocytes in rats. mance thin layer chromatographic determination of organic acid
Korean Soc. Food Sci. Nutr. 26:1200–1207 (1997) preservatives in beverages. Liq. Chromatogr. 17: 855–865 (1994)
7. Sofos JN, Beuchat LR, Davidson PM, Johnson EA. Naturally 22. Pant I, Trenerry VC. The determination of sorbic acid and ben-
occurring antimicrobials in food. Regul. Toxicol. Pharmacol. 28: zoic acid in a variety of beverages and foods by micellar elec-
71–72 (1998) trokinetic capillary chromatography. Food Chem. 53:219–226
8. IARC. Monographs on the Evaluation of the Carcinogenic Risk (1995)
of Chemicals to Humans. IARC, Lyon. 35: 125p (1985) 23. Nevado JJB, Robledo VR, Callado CSC. Monitoring the
9. Robens JF, Dill GS, Ward JM, Joiner JR, Griesemer RA, Douglas enrichment of virgin olive oil with natural antioxidants by using a
JF. Thirteen-week subchronic toxicity studies of Direct Blue 6, new capillary electrophoresis method. Food Chem. 133:497–504
Direct Black 38, and Direct Brown 95 dyes. Toxicol. Appl. (2012)
Pharmacol. 54:431–442 (1980) 24. Panusa A, Gagliardi L. Counterfeit homeopathic medicinal
10. Ibero M, Eseverri JL, Barroso C, Botey J. Dyes preservatives and products: Syrups A simple and rapid LC–ESI-MS method to
salicylates in the induction of food intolerance and/or hypersen- detect preservatives not declared in label. Pharm. Biomed. Anal.
sitivity in children. Allergol. Immunopathol. 10: 263–268 (1982) 47:786–789 (2008)
11. Miller K. Sensitivity to tartrazine. Br. Med. J. 285: 1597–1598 25. Kenny O, Smyth TJ, Hewage CM, Brunton NP. Antioxidant
(1982) properties and quantitative UPLC-MS analysis of phenolic
12. Settipane GA, Chafee FH, Postman IM, Levine MI, Saker JH, compounds from extracts of fenugreek (Trigonella foenum-
Barrick RH. Significance of tartrazine sensitivity in chronic graecum) seeds and bitter melon (Momordica charantia) fruit.
urticaria of unknown etiology. Allergy Clin. Immunol. Food Chem. 141:4295–4302 (2013)
57:541–546 (1976) 26. Tranchida PQ, Franchina FA, Zoccali M, Pantò S, Sciarrone D,
13. Combes RD, Haveland-Smith RB. A review of the genotoxicity Dugo P, Mondello L. Untargeted and targeted comprehensive
of food, drug and cosmetic colours and other azo, triphenyl- two-dimensional GC analysis using a novel unified high-speed
methane and xanthene dyes. Mutat. Res. Genet. Toxicol. 98: triple quadrupole mass spectrometer. Chromatogr. A.
101–243 (1982) 1278:153–159 (2013)
14. McCann D, Barrett A, Cooper A, Crumpler D, Dalen L, Grim- 27. European Commission. Guidance on analytical quality control
shaw K, Kitchin E, Lok K, Porteous L, Prince E, Sonuga-Barke and validation procedures for pesticide residues analysis in food
E, Warner JO. Stevenson J. Food additives and hyperactive and feed. SANCO/12571/2013, 19 November 2013 rev. 0(2013)
behaviour in 3-year-old and 8/9-year old children in the

123