Structural determination of four manufacturing impurities of D&C Red No. 33.

Four manufacturing impurities of D&C Red No. 33 isolated by counter-current chromatography were analyzed by NMR and ESI mass spectrometry. Three of these impurities were reported previously with minimal details of structural determination. All four are structurally related to the main component of the dye. The fourth exhibited an unusual discrepancy between the NMR structure and its chemical formula suggested by ESI-MS results. Structural determination and assignment of the main component and four impurities are discussed as well as resolution of the discrepancy between the NMR and ESI-MS results of the fourth impurity.

Identification, separation by spiral high-speed counter-current chromatography, and quantification of 7-chloro-5-methyl-2H-1,4-benzothiazin-3(4H)-one, an impurity in the thioindigoid color additive D&C Red No. 30 and its lakes.

An impurity in the color additives D&C Red No. 30 (R30) and D&C Red No. 30 lakes (R30L) was newly identified and characterized as 7-chloro-5-methyl-2H-1,4-benzothiazin-3(4H)-one (BTZ), and its extent and level in certified batches of these color additives was determined. BTZ was extracted from the dye with ethanol, resulting in a crude extract enriched to a concentration of over 60%. BTZ was then separated from a portion of the enriched extract by high-speed counter-current chromatography using a spiral-tube assembly column with intermittently pressed tubing of 60 ml capacity. It was the first reported use of such a column to separate a small, moderately hydrophobic compound. The two-phase solvent system was also moderately hydrophobic, consisting of hexane-ethyl acetate-methanol-water (5:2:5:2), and the retention of the organic stationary phase measured after the separation was 83.3%. The separation yielded BTZ of two purity grades, the higher of which (~95.5%) was used as a standard to quantify the impurity in 37 batches of R30 and R30L using an HPLC method developed and validated for that purpose. Analyses revealed a wide range of BTZ levels across batches, <0.05 - 0.84%, and suggested that BTZ contamination could be reduced by appropriate adjustments in the manufacturing process. An explanation of the likely source of BTZ - as a side-reaction product in a particular step of the manufacturing process - was also presented.

Determination of sulphonated quinophthalones in Quinoline Yellow and its lakes using high-performance liquid chromatography.

Quinoline Yellow (QY, Colour Index No. 47005) is internationally used as a colour additive in foods, drugs, and cosmetics. The manufacture of QY requires sulphonating quinophthalone, and depending on the degree of sulphonation, various forms of QY result, containing different proportions of quinophthalone mono-, di-, and trisulfonic acid sodium salts (monoSA, diSA, and triSA, respectively). Regulations on the specific composition and uses of QY differ across countries with associated differences in names for QY. The QY form certified for use in the U.S. in drugs and cosmetics is known as D&C Yellow No. 10 (Y10). The Code of Federal Regulations (CFR) specifies that Y10 and its lakes consist of predominantly monoSA's, the sum of whose levels is ≥ 75%, and that the sum level of diSA's is ≤ 15%, with one of them (6'8'diSA) at ≤ 3%. The present work reports the development of an HPLC method for determining those CFR-specified values and the level of a non-CFR-specified component, 6'8'5triSA. The selected analytes, 6'SA, 6'5diSA, 6'8'diSA, and 6'8'5triSA, were quantified by using five-point-calibration curves (R2 > 0.999) with data-point ranges of 9.96-96.53%, 0.54-21.69%, 0.10-5.00%, and 0.11-5.53% by weight, respectively. The method was found to be precise (relative standard deviation values, 0.55-0.80%) and accurate (recovery values, 91.07-99.45%). LOD and LOQ values, respectively, were as follows: 1.23 and 3.70%, 6'SA; 0.42 and 1.26%, 6'5diSA; 0.11 and 0.34%, 6'8'diSA; and 0.01 and 0.04%, 6'8'5triSA. The HPLC method was applied successfully to the analysis of 20 Y10 and eight Y10 lake samples. It can be extended to other QY forms such as E104 and Yellow 203 because it enables analysis of 6'8'5triSA. This paper also addresses the implications of the varying structure depictions and CAS numbers of the QY components that are due to the existence of three tautomeric forms of quinophthalone.

Determination of Phthalic Acid, 2-(3',5'-Dibromo-2',4'-Dihydroxybenzoyl) Benzoic Acid, and Tribromoresorcinol in the Color Additives D&C Red No. 21, D&C Red No. 22 (Eosin Y), and Their Lakes Using UHPLC.

Background: Certain impurities in the color additives drug and cosmetic (D&C) Red No. 21 (R21), D&C Red No. 22 (R22), and their lakes are limited to levels specified in the Code of Federal Regulations (CFR) and are quantified by the U.S. Food and Drug Administration in batches of these color additives submitted for certification. Currently, a lengthy and tedious method based on gravity flow elution column chromatography is used to quantify the following CFR-specified impurities: the intermediate, phthalic acid (PhthAc); the manufacturing by-products, 2-(3',5'-dibromo-2',4'-dihydroxybenzoyl)benzoic acid (Br2BBA); and brominated resorcinol. "Brominated resorcinol" implies the sum of all possible brominated resorcinols, but the current work focused on 2,4,6-tribromoresorcinol (Br3R) as the most probable side-reaction product. Objective: An improved method was needed to quantify PhthAc, Br2BBA, and Br3R in R21, R22, and their lakes. Methods: A rapid ultra-HPLC (UHPLC) method was developed to replace the gravity flow method for quantitative determination of PhthAc, Br2BBA, and Br3R. Results: PhthAc, Br2BBA, and Br3R were quantified by using five-point calibration curves with data point ranges of 0.11-1.55, 0.06-0.77, and 0.04-0.61% by weight, respectively. LODs for the analytes ranged from 0.01 to 0.03%. Recoveries of the analytes ranged from 90.6 to 99.9%. Conclusions: The UHPLC method is accurate and significantly more rapid than the gravity -flow method, requiring approximately 7 min as compared with 6 h to detect PhthAc, Br2BBA, and Br3R in one sample. Highlights: A rapid UHPLC method was developed to determine CFR-specified impurities in the color additives D&C Red Nos. 21 and 22 and their lakes.

Separation using high-speed counter-current chromatography and identification of 1,3-bis(4-phenylazophenyl)triazene, an impurity in the color additive D&C Red No. 17 (Sudan III).

The present work describes the application of high-speed counter-current chromatography to the preparative separation of a previously unreported impurity in the color additive D&C Red No. 17 (R17, Colour Index No. 26100, Sudan III). Due to the hydrophobic nature of the impurity, a hydrophobic two-phase solvent system (hexane-ethanol-water, 5:4:1) was used for its separation. The separated impurity was chemically characterized by spectroscopic methods as a disazo triazene, 1,3-bis(4-phenylazophenyl)triazene (PAPT). This impurity was synthesized and used as a reference material to quantify it in 15 batches of the color additive produced by various domestic and foreign manufacturers and certified by the U.S. Food and Drug Administration (FDA). Analysis of test portions by high-performance liquid chromatography showed a range of PAPT levels, from "not detected" (<0.006%) to 0.70%, across batches. The variability suggests that contamination by PAPT can be decreased or eliminated through manufacturing modifications. A chemical pathway for PAPT formation and an associated adjustment to minimize it during the process of manufacturing R17 are proposed.

Identification and quantification of the decarboxylated analogue of Pigments Red 57 and 57:1 in the color additives D&C Red No. 6, D&C Red No. 7, and their lakes, using a chelating agent and UHPLC.

Pigment Red 57 (Colour Index No. 15850, PR57) and Pigment Red 57:1 (Colour Index No. 15850:1, PR57:1) are certifiable in the USA as the color additives D&C Red No. 6 (R6) and D&C Red No. 7 (R7) for use in drugs and cosmetics. In the EU, PR57:1 is permitted in cosmetics and also as a food additive (E180) for colouring edible cheese rinds. The USFDA batch-certifies R6, R7, and their lakes in accordance with limiting specifications for impurities stated in the Code of Federal Regulations (CFR). In the current work, an impurity not specified in the CFR was studied because of its consistent presence in samples of R6 and R7 submitted for certification. Using spectroscopic methods, the impurity was tentatively identified as 4-[(4-methyl-2-sulfophenyl)azo]-3-naphthalenol (DPR57), the decarboxylated analogue of PR57 and PR57:1. Its identity was confirmed by synthesising DPR57 and determining that the UHPLC retention time, UV/visible spectrum and mass spectrum of the synthetic material were identical to those of the impurity. Using the synthesised DPR57 as a reference material, the impurity was quantified in 43 batches of R6, R7, and lakes produced by eight different manufacturers. Calibration curves ranging from 0.02% to 1.00% (w/w) were prepared by plotting the UHPLC area of DPR57 at 485 nm against its concentration. DPR57 levels ranged from < 0.02% to 0.50%. To facilitate dissolution of the color additive samples for DPR57 analysis, a relatively simple procedure was developed by adapting a previously published method that involves use of a basic solution of the chelating agent EDTA and the organic solvent N,N'-dimethylformamide. A source for DPR57 contamination of the color additives is also proposed.

Determination of fluorescein and brominated fluoresceins in the color additive D&C Orange No. 5 and its lakes using high-performance liquid chromatography.

The colour additives D&C Orange No. 5 (O5) and its lakes (O5L) are subject to batch certification by the U.S. Food and Drug Administration (FDA) to ensure compliance with specifications in the Code of Federal Regulations (CFR). The present study reports the development of a high-performance liquid chromatography (HPLC) method for the quantitative determination of seven CFR-specified components in O5 and O5L - fluorescein and six brominated fluoresceins. The analytes were quantified using six-point calibration curves with data points (w/w) that ranged as follows: 20.0-70.0% for 4',5'-dibromofluorescein; 9.8-44.1% for 2',4',5'-tribromofluorescein; 1.01-15.2% for 2',4',5',7'-tetrabromofluorescein; 0.10-3.12% for 2',4'-dibromofluorescein; 0.10-3.06% for 2',5'-dibromofluorescein; 0.11-2.85% for 4'-bromofluorescein; and 0.10-2.02% for fluorescein. For all seven analytes, the HPLC instrument response was linear (R2 > 0.999) over the tested concentration ranges and the limits of detection (0.01-1.55%) were well below the CFR-specified levels. Other validation data showed good analyte recovery (87.91-101.73%) as well as method precision measured by the relative standard deviation (0.53-1.56%). The new method was applied to the analysis of test portions from 15 batches of O5 and eight batches of O5L submitted to FDA for certification by domestic and foreign manufacturers. Compared to the thin-layer chromatography/spectrophotometric procedure currently used for routine batch-certification analyses, the new method was found to be more sensitive, simpler to implement, and significantly faster, requiring 25 minutes rather than six hours to analyse one sample.

Preparative purification of 4-hydroxy-1-naphthalenesulfonic acid sodium salt by high-speed counter-current chromatography.

Among the many applications of 4-hydroxy-1-naphthalenesulfonic acid sodium salt (HNSA, CAS No. 6099-57-6) is its use as a common intermediate in the manufacture of dyes used as color additives. In order to develop an HPLC method to analyze HNSA in such additives, it was necessary to obtain pure HNSA as a reference material. This compound is only commercially available in "technical" or "practical" grade ( approximately 70-85% purity). Unsatisfactory results were encountered during the attempts to eliminate impurities using published preparative purification methods. The current work demonstrates the successful application of high-speed counter-current chromatography (HSCCC) to the preparative purification of "practical" grade HNSA. The purifications of 0.55 g and 1.00 g portions were achieved by using a solvent system that consisted of n-butanol/water (1:1), acidified with 0.2% trifluoroacetic acid. In contrast to the procedure involved in previous HSCCC separations of aromatic sulfonates, it was not necessary to add a ligand and a retainer to the solvent system for these separations. The resulting yields were 320 mg and 485 mg of HNSA, respectively, of over 99% purity. Elemental analysis, HPLC (at 240 nm), UV and high-resolution MS analyses confirmed the identity and purity of the HNSA obtained.

Determination of Aniline, 4-Aminoazobenzene, and 2-Naphthol in the Color Additive D&C Red No. 17 Using Ultra-High-Performance Liquid Chromatography.

Specifications in the U.S. Code of Federal Regulations for the color additive D&C Red No. 17 (R17, Colour Index No. 26100) limit the levels of the dye's intermediates, aniline (AN), 2-naphthol (ß-naphthol, BN), and 4-aminoazobenzene (4AAB), to 0.2, 0.2, and 0.1%, respectively. The present work reports the development and application of an ultra-HPLC method for the quantitative determination of these impurities in R17. A 1.7 µm particle size C-18 column was used with 0.2 M ammonium acetate and acetonitrile as the eluents. AN, BN, and 4AAB were quantified by using six-point calibration curves with data points (w/w) ranging from 0.01 to 0.25% for AN, 0.01 to 0.24% for BN, and 0.01 to 0.19% for 4AAB. The correlation coefficients ranged from 0.9992 to 0.9999. Limits of detection for the analytes ranged from 0.002 to 0.01%. Recoveries of the analytes ranged from 99.5 to 102%. Relative standard deviations ranged from 0.482 to 1.262%. The new method was applied to analyze portions from 22 batches of R17 submitted to the U.S. Food and Drug Administration for certification. It was found to be simpler to implement, faster, and more sensitive than the older gravity-elution column chromatography method, which it has replaced.

Identification of 1',5'-naphthyridinophthalone and its quantification in the color additive D&C Yellow No. 10 (Quinoline Yellow) using high-performance liquid chromatography.

The present work reports the identification and characterization of a contaminant, 2-(2'-(1,5-naphthyridinyl))-1,3-indanedione (1',5'-naphthyridinophthalone, 1,5NP), in the color additive D&C Yellow No. 10 (U.S.-certifiable form of Quinoline Yellow), together with its quantification in batches of the color additive certified by the U.S. Food and Drug Administration (USFDA). The impurity, which is a compound not previously reported in the literature, was synthesised and characterised for use as a reference material. Test portions from 26 certified batches of D&C Yellow No. 10 submitted to USFDA by four domestic and four foreign manufacturers were analyzed for 1,5NP using high-performance liquid chromatography. The results revealed a wide range of 1,5NP levels across batches, with 18 (69.2%) of the test portions containing amounts from 0.32 to 169.94 µg g-1 while the remaining test portions contained no detectable (<0.07 µg g-1) amounts. Samples of the European and Japanese forms of Quinoline Yellow were also analyzed and found to contain a wide range of 1,5NP levels. The varying levels of 1,5NP in all three forms of Quinoline Yellow suggest that contamination can be significantly decreased or eliminated through manufacturing adjustments. Since 1,5NP is closely related to a D&C Yellow No. 10 contaminant (quinophthalone) that has a USFDA-specified limit of 4 µg g-1 and is a known allergen, assessment of the possible allergenicity of 1,5NP is warranted.

Preparative separation of isomeric and stereoisomeric dicarboxylic acids by pH-zone-refining counter-current chromatography.

This work involves the preparative separation of some isomeric dicarboxylic acids using pH-zone-refining counter-current chromatography (CCC), a relatively new preparative technique for the separation of ionizable compounds. The paper concentrates especially on the separation of a synthetic mixture of closely related cis and trans pairs of 1-methyl- and 1,3-dimethyl-1,3-cyclohexanedicarboxylic acids. The elution sequence of the isomers is discussed in terms of their relative acidities (pK(a) values) in solution and gas phase, hydrophobicities, and steric configuration. Two possible explanations are suggested for the mechanism of separation. They both involve the amount of retainer acid used, as it affects the separation and plays a role in the chemohydrodynamic equilibrium of the dicarboxylic acids in the column.

Determination of Sudan I and a newly synthesized Sudan III positional isomer in the color additive D&C Red No. 17 using high-performance liquid chromatography.

Specifications in the Code of Federal Regulations for the color additive D&C Red No. 17 (Colour Index 26100) limit the levels of two subsidiary colors, 1-(phenylazo)-2-naphthol (Sudan I) and 1-[[2-(phenylazo)phenyl]azo]-2-naphthalenol (Sudan III o-isomer), to 3% and 2%, respectively. The present work reports the development of a high-performance liquid chromatography (HPLC) method for the quantitative determination of these subsidiary colors. Since Sudan III o-isomer needed to be synthesized for use as a reference material, a two-step procedure was devised: (i) preparative-scale synthesis of the intermediate 2-aminoazobenzene (2AAB) and its purification by counter-current chromatography and (ii) diazotization of 2AAB and coupling with 2-naphthol. Characterization of the newly synthesized Sudan III o-isomer is also reported. Sudan I and Sudan III o-isomer were quantified by using five-point calibration curves with data points ranging from 0.108 to 3.240% and 0.077 to 2.227% by weight, respectively. The HPLC method is rapid (14 min for the total analysis cycle) and simple to implement. It was applied to the analysis of test portions from 25 batches of D&C Red No. 17 submitted to the U.S. Food and Drug Administration (USFDA) for certification, and it has recently been implemented by USFDA for routine batch certification of that color additive.

Identification of the decarboxylated analog of tetrabromotetrachloro-fluorescein and its quantification in the color additives D&C Red Nos. 27 and 28 (phloxine B) using high-performance liquid chromatography.

The present work describes (a) the identification and characterization of an impurity, 2,4,5,7-tetrabromo-6-hydroxy-9-(2,3,4,5-tetrachlorophenyl)-3H-xanthen-3-one (BCPX), in the color additives D&C Red Nos. 27 and 28 (phloxine B) and (b) the determination of the extent and level of BCPX contamination in certified lots of these colors. For these purposes, BCPX (a compound not previously reported in the literature) was synthetically prepared. Test portions from 42 certified lots of D&C Red Nos. 27, 28 and 27 lakes were analyzed for BCPX using an HPLC method that included gradient elution and UV-vis photodiode array detection. Those lots were submitted for certification by both domestic (six) and foreign (six) manufacturers during the past 4 years. Of the test portions analyzed, 32 (76.2%) contained BCPX in amounts ranging from 0.01 to 3.21%. The remaining 10 test portions (23.8%) contained no detectable BCPX or less than 0.008%, which is the limit of quantification for the present method. The analyses revealed substantial differences in the level of BCPX across different manufacturers. The wide range of BCPX levels found in the analyzed lots suggests that the presence of BCPX in D&C Red Nos. 27 and 28 may be avoided or significantly reduced during the manufacturing process.

Hydrogen-shift isomerism: mass spectrometry of isomeric benzenesulfonate and 2-, 3- and 4-dehydrobenzenesulfonic acid anions in the gas phase.

The isomeric 3- and 4-dehydrobenzenesulfonic acid anions b and c were prepared by collision induced dissociation (CID) of the [M - H](-) ions of isomeric sulfobenzoic acids obtained by negative electrospray ionization (ESI). The CID spectra (MS(3)) of anions b and c are different from each other, and both are different from that of the isomeric benzenesulfonate anion a, obtained from benzenesulfonic acid. The stability of ions b and c shows that 1,2-proton transfer does not take place in this system under the conditions of the CID experiment. Density functional (DFT) calculations at B3LYP/6-31+G(2d,p) level of theory show that benzenesulfonate anion a is the most stable isomer, and the energies of isomers b and c are higher by more than 65 kcal mol(-1). The calculated energies of the transition states involved in the 1,2-hydrogen migration leading to the interconversion of the isomeric anions are very high (>120 kcal mol(-1)relative to ion a, barrier energies >55 kcal mol(-1)), much higher than those of transition structures leading to fragmentation. This situation does not allow isomerization of ions b and c to a, under the conditions of the CID experiments. The isomeric 2-dehydrobenzenesulfonic acid anion isomerizes to the benzenesulfonate anion a by a facile proton transfer from the SO(3)H group to the adjacent position 2. The results of this work indicate that the gas phase deprotonation of meta- and para-sulfobenzoic acids is a kinetically controlled process.

Preparative separation and identification of novel subsidiary colors of the color additive D&C Red No. 33 (Acid Red 33) using spiral high-speed counter-current chromatography.

Three low-level subsidiary color impurities (A, B, and C) often present in batches of the color additive D&C Red No. 33 (R33, Acid Red 33, Colour Index No. 17200) were separated from a portion of R33 by spiral high-speed counter-current chromatography (HSCCC). The separation involved use of a very polar solvent system, 1-BuOH/5mM aq. (NH4)2SO4. Addition of ammonium sulfate to the lower phase forced partition of the components into the upper phase, thereby eliminating the need to add a hydrophobic counterion as was previously required for separations of components from sulfonated dyes. The very polar solvent system used would not have been retained in a conventional multi-layer coil HSCCC instrument, but the spiral configuration enabled retention of the stationary phase, and thus, the separation was possible. A 1g portion of R33 enriched in A, B, and C was separated using the upper phase of the solvent system as the mobile phase. The retention of the stationary phase was 38.1%, and the separation resulted in 4.8 mg of A of >90% purity, 18.3mg of B of >85% purity, and 91 mg of C of 65-72% purity. A second separation of a portion of the C mixture resulted in 7 mg of C of >94% purity. The separated impurities were identified by high-resolution mass spectrometry and NMR spectroscopic techniques as follows: 5-amino-3-biphenyl-3-ylazo-4-hydroxy-naphthalene-2,7-disulfonic acid, A; 5-amino-4-hydroxy-6-phenyl-3-phenylazo-naphthalene-2,7-disulfonic acid, B; and 5-amino-4-hydroxy-3,6-bis-phenylazo-naphthalene-2,7-disulfonic acid, C. The isomers A and B are compounds reported for the first time. Application of the spiral HSCCC method resulted in the additional benefit of yielding 930 mg of the main component of R33, 5-amino-4-hydroxy-3-phenylazo-naphthalene-2,7-disulfonic acid, of >97% purity.

Attachment of neutrals during tandem mass spectrometry of sulfonic acid dyes and intermediates in an ion trap.

Several positional isomers of 2-(2-quinolinyl)-1H-indene-1,3(2H)-dione mono- and disulfonic acids prepared as reference materials for development of analytical methods involved in FDA certification of D&C Yellow No. 10 (Quinoline Yellow) were found consistently to show [MH + 14](+) ions when their electrospray- or atmospheric pressure chemical ionization-prepared MH(+) ions were subjected to collisional activation. The source of these ions was found to be the methanol used as solvent in these procedures which combined with their [MH - H(2)O](+) ions under chemical ionization conditions. The reaction was found to be sensitive to their isomeric and chemical structures and other examples of this process are reviewed.

Identification of 2-bromo-3,4,5,6-tetrachloroaniline and its quantification in the color additives D&C Red Nos. 27 and 28 (phloxine B) using solid-phase microextraction and gas chromatography-mass spectrometry.

The present work describes (a) the identification and characterization of a contaminant, 2-bromo-3,4,5,6-tetrachloroaniline (2BTCA), in the color additives D&C Red Nos. 27 and 28 (phloxine B) and (b) the determination of the extent and level of 2BTCA contamination in certified lots of these colors. For these purposes, 2BTCA (a compound not previously reported in the literature) and its positional isomer 4-bromo-2,3,5,6-tetrachloroaniline (4BTCA) were synthetically prepared. 4BTCA was used as the internal standard for the quantification of 2BTCA in the colors. Test portions from 35 certified lots of D&C Red Nos. 27 and 28 were analyzed for 2BTCA using a solid-phase microextraction-GC-MS method. Those lots were submitted for certification by both domestic (seven) and foreign (four) manufacturers during the past 4 years. Of the test portions analyzed, 22 (62.9%) contained 2BTCA in amounts ranging from 0.15 to 435.7 ppm with an average value of approximately 131.7 ppm. The remaining 13 (37.1%) test portions contained no detectable 2BTCA or less than 0.01 ppm, which is the limit of quantification of the present method. The analyses revealed substantial differences in the level of 2BTCA across lots from the same manufacturer as well as among different manufacturers. The wide range of 2BTCA levels found in the analyzed lots suggests that the presence of 2BTCA in D&C Red Nos. 27 and 28 may be avoided or significantly reduced during the manufacturing process. A direct correlation was observed between the presence of 2BTCA and that of 3,4,5,6-tetrachlorophthalic acid in analyzed batches of D&C Red Nos. 27 and 28. A chemical pathway that could explain the presence of 2BTCA in these color additives, and ways to avoid its formation, are also proposed.

Determination of 2,4,6-tribromoaniline in the color additives D&C Red Nos. 21 and 22 (Eosin Y) using solid-phase microextraction and gas chromatography-mass spectrometry.

The present work demonstrates the presence of an impurity, 2,4,6-tribromoaniline (TBA), in the color additives D&C Red Nos. 21 and 21 lake (21L) and describes the determination of TBA in certified lots of D&C Red Nos. 21, 21L and 22 (Eosin Y). A method was developed using solid-phase microextraction with [13C6]TBA as an internal standard followed by gas chromatography-mass spectrometry analysis. Test portions from 23 lots of US-certified color additives D&C Red Nos. 21, 21L and 22 were analyzed for TBA using the new method. These lots represent domestic (four) and foreign (four) manufacturers that requested certification for the color additives during the past 2 years. Of the test portions analyzed, 12 (52.2%) contained TBA in amounts ranging from 19.9 to 638.9 ppm with an average value of approximately 278.7 ppm. The remaining 11 (47.2%) test portions contained no detectable TBA or less than 0.01 ppm, which is the limit of quantification of the present method. The wide range of TBA levels found in lots submitted for certification suggest that the contamination with TBA may be avoided or significantly decreased through appropriate changes in the color-manufacturing process.

Preparative separation of isomeric sulfophthalic acids by conventional and pH-zone-refining counter-current chromatography.

Two modes of high-speed counter-current chromatography (HSCCC) were applied to separate 3- and 4-sulfophthalic acid from a mixture. Conventional HSCCC was useful for the separation of up to several hundred milligram quantities of these positional isomers, while pH-zone-refining CCC was implemented successfully to separations at the multigram level. The conventional HSCCC separations were performed with a standard J-type HSCCC system that has a superior resolution but a lower level of retention of the stationary phase of the biphasic solvent system used (acidified n-butanol-water). The pH-zone-refining CCC separations were performed with an X-type HSCCC system (a cross-axis system) that has a higher capability for retention of the stationary phase. The purified positional isomers (over 99% pure as determined by HPLC) were characterized by 1H NMR and negative ion electrospray ionization mass spectrometry.

Determination of synthetic by-products and an intermediate in the colour additives D&C Red Nos 27 and 28 (phloxine B) and their lakes using conventional HPLC.

Specifications in the US Code of Federal Regulations (CFR) for the colour additives D&C Red No. 27 (R27) and D&C Red No. 28 (R28) limit the levels of the synthetic by-products 2,3,4,5-tetrachloro-6-(3´,5´-dibromo-2´,4´-dihydroxybenzoyl)benzoic acid (SBBA) and "brominated resorcinol" and of the intermediate 3,4,5,6-tetrachlorophthalic acid (TCPA). The present study reports the development and application of a conventional HPLC method for the quantitative determination of these impurities in R27, R28 and their lakes. Because the CFR-listed "brominated resorcinol" refers to a group of synthetic by-products, six variously brominated resorcinols were examined as possible impurities. Due to their rapid decomposition in the presence of the dye, the existence of any brominated resorcinols in R27 and R28 is highly unlikely, and this supposition is supported by the results obtained in this study. SBBA, 2,4,6-tribromoresorcinol (Br3R, the brominated resorcinol most likely to be found as an impurity) and TCPA were quantified by using five-point calibration curves with data points (w/w) that ranged from 0.066% to 0.820% for SBBA, from 0.050% to 0.499% for Br3R, and from 0.061% to 1.410% for TCPA. The HPLC method was applied to the analysis of test portions from batches of R27, R28 and their lakes submitted to the USFDA for certification by domestic and foreign manufacturers.