Determination of synthetic food colourant in tea drink by absorption

spectrophotometry

JIHUA ZHANG, ZHIMIN ZHAO,*, RENDONG JI, LEXIN WANG
*Corresponding author
College of Science, Nanjing University of Aeronautics and Astronautics, Nanjing, 210016, China
KEYWORDS: absorption spectra; first derivative of ratio spectra; synthetic food colourant, tea drink
ABSTRACT: In this paper, two kinds of simple, precise and sensitive approaches are developed for the
determination of synthetic food colourant in the tea drink based on absorption spectrophotometry. One method
focuses on the analysis of the intensity at 480nm wavelength of original absorption spectra. The other one aims at the
technology of calculating the intensity at 462nm and 524nm wavelength of the first derivative of ratio spectra. In
addition, both of correlation coefficients are found to be near to 0.999. Validated by assessing the precision, limit of
quantification, limit of detection and sensitivity, both of them show accurate, simple, cheap and less time-consuming
feature. Therefore, they can be suitably applied in the estimation of the synthetic food colourant in tea drink.
INTRODUCTION
Tea drink is one of the most popular drinks in the world. China, the birthplace of tea, exports a great quantity of tea
leaves to various countries in the world. Owing to tea has anticarcinogenic action (1-3) and resists radiation damage
(4, 5). Tea has become the favourite for more and more people. So safety detection of tea, such as identification of
tea adulterations, analysis of pesticide residues and determination of microbial, fluroine and heavy metal
contamination, becomes important nowadays.
Food colourant addition is one of the problem in tea adulterations. Some synthetic colourants may be pathogenic,
especially if they are consumed in excess. Therefore, safety data for every synthetic colourant food additive have
been repeatedly determined and evaluated by the Food and Agricultural Organization and World Health Organization.
The use of synthetic colourants as food additives is common in most countries. Different methods have been
employed for the determination of colourants in diverse matrices, including chromatographic methods, such as
column chromatography, HPLC and TLC (6-9).Numerous procedures are based on the use of electroanalytical
methods (10, 11), such as polarographic analysis (12), voltammetry (13) and ion-selective electrode analysis (14).But
spectrophometrical method is also used to determine food colourant (15, 16), such as original absorption spectra
method (13) and derivative spectrometric method (18).
Because of absorption spectrometry can indicate information of energy levels, it is applied to determination of
molecular structure (19) and quantitative analysis for substance (20). Nowadays, molecular absorption
spectrophometric method of analysis is extensively used in the determination of colourants (21-23) due to its high
sensitivity and cost-effectiveness. A variety of mathematics algorithms are combined with molecular absorption
spectrometry in substance analysis, for instance partial least squares regression(PLS) (24), principal component
analysis(PCA) (25), derivation (26), calculus of finite differences (27) and so on. The original absorption
spectrophometic method is based on the Lambert-Beer Law, which states the linear relationship between absorbance
of a solution and its concentration.Inoriginal measurement methods such as derivative spectrophotometry and ratio
spectra derivative spectrophotometric approaches have been the focus of the determination of colourant (28-30).
Two methods, original absorption spectra and the first derivative of the ratio spectra, are used in the present paper.
The objective of the current study is to develop rapid, accurate, reproducible, validated and economical methods for
the determination of the food colourant in tea drinks.
EXPERIMENTS
Instruments and apparatus
SHIMADZU UV-VIS-NIR spectrophotometer,UV-3600, connected to a computer and loaded with UV-probe software
was used. For all absorbance measurements,1-cm matched quartz cuvettes were used. The absorbance from 200nm
to 700nm was recorded by UV-3600.
The material for testing was a tea-drink and synthetic colourant mixed solution.
The reference substance was deionized water.
Preparation of synthetic tea colourant
All colourants were purchased from the local market. The food colourant sample was obtained by mixing 2.5mg apple
green and 15mg sunset yellow, whose colour is similar to the tea. Apple green is a blend of two food dyes, lemon
yellow (CAS:1934-21-0) and brilliant blue (CAS:3844-45-9), and sunset yellow (CAS:2783-94-0) is a single chemical
compound whose chemical formula is C16H10N2Na2O7S2.
Preparation of standard stock solution
The stock solution of colourant was prepared by dissolving the tea colourant ( mixed by 2.5mg apple green and 15mg
sunset yellow ) in 600ml water.
The stock solution of tea was prepared by brewing 1g tea leaves in 125ml boiling water and collecting the
supernatant liquor after 30 minutes for testing.
All working samples were divided into two sets: calibration set and validations set, as Table 1 showing.
RESULTS AND DISCUSSIONS
The wave band for the determination of the synthetic tea colourant
Figure 1 shows the absorbance of the synthetic tea colourant from 200nm to 700nm. It can be found that there are
four peaks at 234nm,314nm,480nm and 630nm, respectively. Peaks at 234nm,314nm and 480nm are engendered by
sunset yellow and 630nm is caused by brilliant blue in apple green.
Figure 2 shows the absorbance of the tea drink from 200nm to 700nm. Apparently, there is no peak in the spectra
and the noise appears in the ultraviolet band that may results from the scattering of some colloid and suspension.
In figure 3, the absorption spectra of the colourant and the tea are showed together. Although there are four peaks of
the colourant, two peaks (234nm and 314nm) are covered by the UV noise of the tea. In addition, intensity of 480nm
is larger than the intensity of 630nm.Therefore, range of 420nm-560nm is intended to be the wave band for the
determination of the synthetic tea colourant.
Two methods for the determination of the synthetic tea colourant in the tea drinks
The original absorption spectrophotometric method and the first derivative of the ratio spectra method were
developed for the determination of the synthetic tea colourant. Absorption spectra is a simple, economic and nature
friendly analytical choose for the determination of the synthetic tea colourant.
The original absorption spectrophotometric method used in the determination of the synthetic tea colourant is based
on the linear relationship between absorbance of the solution and its concentration depicted by Lambert Beer Law.
For the max of the tea colourant was found to be 480nm (Figure 4), the absorbance intensities at 480nm of all
samples in calibration set was selected to fit linear with the concentration of the colourant. The regression line
equation is : y = 0.01968*x + 0.13582, x is the concentration of the colourant and y is the absorbance intensity at
480nm. According to the regression line equation, the concentrations of samples in validations set can be determined
and the prediction result and evaluation are presented in the Table 2.
The absorption spectra of tea drinks with synthetic colourant of the increasing concentrations was divided by the
spectrum of the pure tea drink to get the ratio spectra. From the ratio spectra, first-derivative spectra (Figure 5) were
calculated. As showed in Figure 5, it was found that the measured signals have a maximum at 462nm and a
minimum at 524nm.Further, it was found that the measured signals at these wavelengths were proportional to the
concentrations of the synthetic colourant.Hence, the values of the first derivative of the ratio spectra of 462nm and
524nm of all samples in calibration set was selected to fit linear with the concentration of the synthetic colourant,
respectively.The regression equations are :y = 0.00344*x - 0.00933 at 462nm and y = -0.00654*x + 0.02245 at
524nm. Based on the two equations, the concentrations of the synthetic colourant in the samples of validations set
can be calculated and the recoveries are also figured out (Table 3).
The evaluation of the two methods (The original absorption spectrophotometric method and the first derivative of the
ratio spectra method)
A critical evaluation of the proposed methods was performed by the statistical analysis of the experimental data. The
obtained slopes, intercepts and correlation coefficients are summarized in Table 4.Limit of detection (LOD) was
measured as the lowest amount of the analyte that could be detected to produce a signi?cant response. It was
approved by calculations based on the standard deviation of the response ( ) and the slope (S) of the calibration
curve at the levels approaching the limits according to equation LOD = 3.3 ( /S) and LOQ = 10 ( /S).
The evaluations of the prediction results by two methods are in Table 5.The mean error of prediction, the sum of
squares of prediction error, the mean recovery is revealed.
Both of the two methods have their advantages and disadvantages, respectively. The original absorption spectra
method has the higher sensitivity (Table 4) and does without further calculation of the spectra.However, the mean
prediction error of original method is -3.695, which is much larger than this of first derivative ratio spectra (0.2519).
Besides, the sum of error squares of original method is 2.9609, larger than 0.86808 (the sum of error squares of first
derivative ratio spectra). So it indicates that the precision of original absorption spectra method is lower than that of
the first derivative of the ratio spectra method, which perhaps due to the ratio spectra that can reject some noises. On
the other hand, the first derivative of the ratio spectra method has higher precision but lower sensitivity. And division
and derivation have to be done in the method, which is more complex than the original absorption spectra method.
CONCLUSIONS
In both of methods, to get the concentration of the synthetic food colourant in tea drink, the whole process should be
divided into three step firstly to collect different samples of absorption spectra, secondly to establish measurement
model of line fitting, and lastly to predict the result. In addition, the comparison between two methods has been done
with considering the mean error, the sum of squares of error and the mean recovery, revealing that first derivative of
the ratio spectra method is much better in the prediction accuracy but much worse in sensitivity. Therefore which one
should be selected depends on the application of scene in real. Finally, with considering some important assessing
parameters of LOD, LOQ, RSD, etc, two kinds of analytical methods proposed prove feasible to predict the
concentration of the synthetic food colourant in tea drink with being rapid, accurate and non-pollution because of the
media of clean water without any chemistry.
ACKONWLEDGEMENTS
This work is supported by the Innovation of Graduate Student Training Project in Jiangsu Province (CXLX13-146),
the National Natural Science Foundation of China (NO.10172043), Specialized Research Fund for the Doctor
Program of Higher Education of China (NO.20093218110024), International Science and Technology Cooperation
Grant (NO.BZ2010060), Program Sponsored for Scientific Innovation Research of College Graduate in Jiangsu
Province (CXLX12_0143), the Innovation and Excellence Fund for Doctoral Dissertation of Nanjing University of
Aeronautics and Astronautics (BCXJ12-10), Postdoctoral Grant and the Fundamental Research Funds for the Central
Universities (NS2012062).



Table1. The calibration set and the validation set of the experiment



Table 2. The prediction result by original absorption spectra method



Table3. The prediction result by first derivative of the ratio spectra method



Table 4. The evaluation of the proposed two methods



Table5. The evaluation of the prediction results by the two methods



Figure 1. Absorption spectrum of colourant



Figure 2. Absorption spectrum of tea drink



Figure 3. Absorption spectrum of colourant and tea drink



Figure 4. Absorption spectrum of samples in calibration set



Figure 5. The ratio spectra, first-derivative-spectra of samples in calibration set
REFERENCES

1. H. Fujiki, The Chemical Record, 5, 119-132 (2005).
2. N.T. Zaveri, Life Science, 78, 2073-2080 (2006).
3. S. Shankar, Frontiers in bioscience, 12, 4881-4899 (2007).
4. Y.L. Wu, J.K. Lin, Journal of Tea, 37,4, 213-217 (2011).
5. Y.F. Wang, S.Z. Liang, S.K. Zhang, Journal of Tea Science, 5, 31-36 (2011).
6. X. Wang, G. Song and W. Wu, Chromatographia, 68, 659-662 (2008).
7. S.C. Rastogi, V.J. Barwick, S.V. Carter, Chromatographia, 45, 215-228 (1997).
8. A.W. Strigl, E. Leitner, W. Pfannhauser, Lebensmittel Untersuchung und Forschung, 201,266-268(1995).
9. F. Soponar, A.C. Mot and Costel Sarbu, Journal of Chromatography A, 1188. 2, 295-300 (2008).
10. H. Liu, T. Zhu and Y. Zhang, Journal of Chromatography A, 718, 448-453 (1995).
11. J. Zhao, Y. Zhang and K. Wu, Food Chemistry, 128. 2, 569-572 (2011).
12. N. Zhao, B. Xu, M. Wang and K. Jiao, Journal of the Iranian Chemical Society, 5. 1, 115-121 (2008).
13. J. Zima, J. Barek and J.C. Moreira, Fresenius Journal of Analytical Chemistry, 369, 567-570 (2001).
14. E.V. Nizhnilova and A.P. Podterob, Journal of Analytical Chemistry, 60. 1, 61-64 (2005).
15. T. Pocock, M. Krol and N.P.A. Huner, Photosynthesis Research Protocols, 274, 137-148 (2004).
16. A.E. Solovchenko, O.B. Chivkunova, M.N. Merzlyak and I.V. Reshetnikova, Russian Journal of Plant
Physiology, 48. 5, 693-700 (2001).
17. N. Shahabadi, M. Maghsudi and S. Rouhani, Food Chemistry, 135. 3, 1836-1841 (2012).
18. C. Cruces Blanco, A.M.Garcia Camparia and F. Ales Barrero,Tanlata, 43. 7, 1019-1027 (1996).
19. W. Zhou, Z.H. Sun, W.X. Cao, G.F. Wang.Spectroscopy and Spectral Analysis, 32.12, 3347-3352 (2012).
20. X.J. Ma, H.W. Zhao, G.F. Liu. Spectroscopy and Spectral Analysis, 29.11, 2885-2888 (2009).
21. P.R. Robinson, K. Griffith, J.M. Gross and M.C.O' Neill, Vision Research, 39. 9, 1707-1712 (1999).
22. P. Jaramillo, K. Coutinho, B.J.C. Cabral and S. Canuto, Chemical Physics Letters, 516, 250-253 (2011).
23. E. Torrecilla, D. Stramski and R.A. Reynolds, Remote Sensing of Environment, 115. 10, 2578-2593 (2011).
24. S. Scedilahin, E. Sarinodotburun and C. Demir, Analytical Methods, 1. 3, 208-214 (2009).
25. J.Q. Wu, X.Y. Luo, J.P. Geng and X.Y. Li, Computers and Applied Chemistry, 29. 2, 211-214 (2012).
26. M.M. Seleim, M.S. Abu-Bakr, E.Y. Hashem and A.M. El-Zohry, Journal of Applied Spectroscopy, 76. 4, 554-
563 (2009).
27. L.T. Shao and S.M. Wang, Instrument Technique and Sensor, 10, 86-90 (2010).
28. S. Sayar and Y. Ozdemir, Food Chemistry, 61. 3, 367-372 (1998).
29. M.H. Sorouraddin, A. Rostami and M. Saadati, Food Chemistry, 127. 1, 308-313 (2011).
30. S.Altinoz and S. Toptan, Journal of Food Composition and Analysis, 16. 4, 517-530 (2003).