Food Additives and Contaminants, December 2005; 22(12): 1163–1175
Method development and HPLC analysis of retail foods and beverages for copper chlorophyll (E141[i]) and chlorophyllin (E141[ii]) food colouring materials
MICHAEL J. SCOTTER, LAURENCE CASTLE, & DOMINIC ROBERTS Central Science Laboratory, Department for Environment, Food and Rural Affairs, Sand Hutton, York, UK (Received 11 July 2005; revised 3 August 2005; accepted 7 August 2005)
Abstract An analytical method using high performance liquid chromatography with photodiode array and fluorescence detection has been developed and applied to the determination of the food colour additives copper chlorophylls and copper chlorophyllins (E141[i] and [ii]) in foods and beverages. The analytical procedures from previously reported methods have been refined to cover a range of food colour formulations and retail foods. The method was single-laboratory validated. Recoveries of the polar copper chlorophyllins from spiked samples (at 14.5 mg/kg in all but one case) were in the range 79–109%, except for jelly sweets (49%). Recoveries of relatively non-polar copper chlorophylls were in the range 77–107% (except for ‘made’ jelly at 50%). The %RSD for recoveries was generally below 12%. Quantitative estimates of the total copper chlorophyll/chlorophyllin content of a small range of food commodities are reported, based on the use of trisodium copper chlorophyllin as a surrogate standard. The majority of E141-containing foods and colour formulations analysed exhibited a multiplicity of components due to the various extraction and purification processes that are used to obtain these colour additives. This was confounded by the presence of overwhelming amounts of native chlorophylls in certain samples (e.g. mint sauce). Food commodities containing significant amounts of emulsifiers (i.e. ice cream), gelatine or fats were problematic during extraction hence further development of extraction regimes is desirable for such products. All of the samples analysed with added E141, had estimated total copper chlorophyllin contents of below 15 mg/kg (range 0.7–13.0).
Keywords: Food, additives, colours, analysis, chlorophyll, chlorophyllin, copper, E141
Introduction The food additives E141 are copper complexes of chlorophylls (E141[i]) and chlorophyllins (E141[ii]) and are listed in Directive 94/36/EC on food colouring materials (EC 1994). E141(i) and (ii) are permitted in a wide range of food commodities quantum satis. Nevertheless, these additives have an acceptable daily intake of 15 mg/kg bw/day (MAFF 1993). However, there is very little intake data for these additives, and estimates have been difficult to obtain due to reported low usage (MAFF 1993). There is a need therefore to have intake estimates for E141(i) and E141(ii) which can be assessed within the context of the ADI. The chlorophylls are a family of naturallyoccurring pigments present in the photosynthetic tissues of all living plants, including algae, and in some photosynthetic bacteria. As an integral part
Correspondence: Michael J. Scotter. E-mail:
[email protected] ISSN 0265–203X print/ISSN 1464–5122 online ß 2005 Taylor & Francis DOI: 10.1080/02652030500306885
of vegetable foodstuffs, they have formed a constant component of the human diet. Chlorophyll derivatives form a significant and growing element in the range of natural pigments used as food colorants. These are almost entirely derived from lucerne or alfalfa (Medicago sativa), nettles (Urtica dioica) or several high-yield pasture grasses. The green colour is due to the pigments chlorophyll a (Chla, bluegreen) and chlorophyll b (Chlb, yellow-green) that occur together in a ratio of about 3 : 1 (Hendry and Houghton 1992). Removal of magnesium from the chlorophylls gives the corresponding phaeophytins a and b, both of which are olive brown. Replacing the Mg2þ with copper (Cu2þ) retains the green colour. Purified food-grade forms of the oil-soluble Cu2þ complexes of chlorophylls (E141[i]) are produced after further washing and purification. The watersoluble copper complexes of chlorophyllins (E141[ii]) are prepared by saponification of crude mixtures
1164
M. J. Scotter et al.
of chlorophylls and phaeophytins thereby allowing the hydrophobic phytyl group to be displaced by sodium or potassium, and the carboxyl moiety of the cyclopentanone ring to be converted to the Na or K salt. After further fractionation and washing to remove much of the lipophilic contamination, conversion to a stable green colorant is achieved by acidification in the presence of Cu2þ salts. Almost any type of food processing, alone or combined with another treatment or storage, causes some deterioration of native chlorophyll pigments (Belitz and Grosch 1986; Davidek et al. 1990; Hendry and Houghton 1992). The Cu-derivatives of E141(i) and (ii) are stable to both moderate heat, light and mineral acids but there remains the possibility for a multiplicity of coloured components in each group. These are likely to be present as a result of the various extraction and purification processes and could include native chlorophylls, phaeophytins (chlorophyll minus the Mg2þ), phaeophorbides (phaeophytins minus the phytol group), rhodochlorins (free carboxyl form of phaeophorbides) as well as the principal colouring components (Belitz and Grosch 1986; Coultate 1989; Hendry and Houghton 1992). The main objective of this project was to develop and validate an HPLC-based method for the determination of these additives in the range of foodstuffs in which they are permitted, which could then be used to monitor the levels of these additives in foods thereby facilitating an estimation of dietary exposure. Review of published analytical methodology The E141 copper chlorophylls and chlorophyllins are chemically well-defined species for which purity criteria are laid down that include identification and assay tests (EC 1995). There are many literature references available on the determination of naturally-occurring chlorophylls, which have largely concentrated on their determination in fresh and processed fruit and vegetables (Schwartz et al. 1981; Saag 1982; Roy 1987; Suzuki et al. 1987). The main strategies of methods for chlorophylls have focused on solvent extraction, clean up using liquid–liquid partition and measurement using reverse-phase HPLC. However, analytical methods for their determination in foodstuffs, especially the water soluble forms E141(ii), are very poorly documented. Yasuda et al. (1995) identified copper chlorin e4 (CuCe4) as a suitable indicator for the analysis of sodium copper chlorophyllin in foods. The colouring matter was extracted from a small range of samples (boiled bracken, agar–agar and chewing gum) with diethyl ether after pH adjustment (3–4). Following
removal of the solvent, the residue was dissolved in methanol and analysed using reverse-phase HPLC on an octadecylsilane (C18) column with a mobile phase comprising methanol : water (97 : 3) containing 1% acetic acid. A monitoring wavelength of 405 nm was used and the compounds characterized using photodiode array detection. Inoue et al. (1994) prepared and separated the components of copper chlorophyllin, consisting of copper pheophorbide a (CuPPa), copper chlorin e6 (CuCe6), copper rhodin g7 (CuRg7) and copper chlorin e4, using semi-preparative RP-HPLC. Separation was achieved using a mobile phase of methanol–water (97 : 3, v/v) containing 1% (v/v) of acetic acid. Linear calibration plots were obtained for copper chlorophyllin in the concentration range of 0–30 mg/ml with UV-VIS detection at 407 or 423 nm. The detection limits of CuPPa, CuCe6, CuRg7 and CuCe4 were 3.5, 1.5, 3.3 and 1.4 ng/ml respectively. The reversed-phase HPLC method proposed was demonstrated to be useful for the determination of the components of sodium copper chlorophyllin in food colour formulations. Five samples of commercial copper chlorophyllin preparations were analysed by Chernomorsky et al. (1997) using gradient elution C18 RP-HPLC with methanol/ammonium acetate/acetone mobile phase and photodiode array detection. Analysis revealed several significant differences in the porphyrin compositions of the samples and copper isochlorin e4 was identified as the major component in most commercial materials. Copper complexes of Ce6 (i.e. 131-carboxyl-), PPa and unidentified porphyrins with either chlorin or non-chlorin type PDA spectra were found in some samples, which eluted within 15 min. Almela et al. (2000) used similar chromatographic conditions to screen chlorophyll derivatives produced during the ripening of fruit. In that study, chromatograms were monitored at 660 nm (PDA) and by using fluorescence (Ex ¼ 440 nm, Em ¼ 660 nm). This method was particularly successful in separating a series of compounds exhibiting a broad range of polarities. This causes some problems because the chlorophyllins are dissociated even at neutral pH and can thus interact hydrophobically with the C18 stationary phase. The high concentration of ammonium acetate used in the mobile phase was essential for decreasing proton equilibration times, especially for the ionogenic chlorophyllides and phaeophorbides. In a later study, Inoue et al. (1988) used non-aqueous RP-HPLC to separate and characterize several copper (II) chlorophyll derivatives. This work was later extended to the analysis of iron (III) derivatives of chlorophyllin using ion-pair RP-HPLC (Nonomura et al. 1996). We took these published procedures as the basis for further method
HPLC analysis for E141(i) and E141(ii) food colouring
1165
development in order to be able to analyse foodstuffs for E141 content.
720 4 nm bandwidth (2) Ex ¼ 400 nm and Em ¼ 640 nm.
Materials and methods
Sample extraction
Reagents
One to 5 g of sample was weighed into a centrifuge tube and 200 ml internal standard (Solvent Green 3, 50 mg/kg), 15 ml citrate/phosphate buffer (pH 2.6) and 10 ml ethyl acetate : acetone (5 : 1 v/v) were added. Samples such as jelly, boiled sweets and jelly sweets were dissolved in warm water prior to extraction. Certain samples such as dried soup mixes and flour confectionery required homogenization in the buffer solution using a 5 mm micro dispersion tool for ca. 30 sec and a longer solvent extraction time of 2–6 hours on an orbital shaker. Prior to extraction, biscuit samples were defatted by homogenization with 2 10 ml hexane followed by centrifugation (hexane discarded). The tube contents were either homogenized using the micro dispersion tool or a vortex mixer for 1 min, depending upon sample type. The extraction mixture was centrifuged at 1600 g for 5 min and the upper solvent layer removed to a clean centrifuge tube. [Note: (1) If the upper layer did not separate clearly or was gel-like, emulsifiers and/or gels were deemed to be present. In this case, 0.5 ml of ethanol was carefully added (no shaking), the tube centrifuged and the upper solvent layer removed as above. (2) Where a coloured interfacial (emulsion) layer persisted, the lower aqueous layer was carefully removed using a pipette, and discarded. Acetone (3 ml) was added and the tube contents vortex mixed for ca. 15 sec, centrifuged and the supernatant combined with the pooled ethyl acetate : acetone extract. This step was repeated if necessary]. A further 5 ml of 5 : 1 ethyl acetate: acetone solution was added to the buffer mixture in the original tube and the contents vortex mixed for 1 min, centrifuged and the upper layer pooled with that obtained from the first extraction. This step was repeated until no more colour was observed in the upper layer. The solvent was removed by gentle blowdrying under nitrogen at
