Analyst, November, 1973, Vol. 98, PP. 797-801 797 The Determination of Chlorhydroxyquinoline in Medicated Pig Feeds BY J. E. FAIRBROTHER AND W. F. HEYES (Squibb raternational Development Laboratory, Moreton, Wirral, Cheshire) A method has been developed for the determination of chlorhydroxy- quinoline (Halquinol) in medicated pig feeds. Because of the interference by feed constituents in simple spectrophotometric and polarographic assay procedures, a spectrofluorimetric procedure is recommended. Spectrofluori- metric measurements are taken in a methanolic solvent containing 5 per cent. of chloroform, and the fluorescence of chlorhydroxyquinoline, as its mag- nesium chelate, is measured at 500 nm, with an excitation wavelength of 402 nm. Cyanide is used to suppress interference from copper and zinc salts that are commonly added to these feeds.The procedure is not affected by the presence of other feed additives, such as dimetridazole and arsanilic acid. CHLORHYDROXYQUINOLINE, I, is a mixture of three isomeric chloro-8-hydroxyquinolines that is claimed1 to possess antibacterial and ant ifungal activities greater than those of its individual components : Y OH I ( Halquinol) where X = C1, Y = C1 (57 to 74 per cent. m/m); X = H, Y = C1 (23 to 40 per cent. nz/m); and X = C1, Y = H (not more than 3 per cent. m/m). One of the veterinary uses of chlorhydroxyquinoline is in the treatment of bacterial diarrhoea in weanling pigs. The drug is marketed for this purpose as a pre-mix (“Quixalud,” E. R. Squibb & Sons Ltd.) that is diluted with pig feed.Chlorhydroxyquinoline is readily soluble in chloroform and shows an absorption maximum at 335 nm (&& = about 150). It forms chelates with many cations, in a number of instances (iron, copper, vanadium, titanium and molybdenum) giving coloured products suitable for spectrophotometric determination. The green chelate formed with iron(II1) (with an absorp- tion maximum at 685nm) is soluble in a mixture of chloroform and acetone and can be used (J. E. Fairbrother and W. F. Heyes, unpublished work) for the spectrophotometric determination of chlorhydroxyquinoline in pre-mixes that contain inorganic diluents. The determination of chlorhydroxyquinoline in medicated pig feeds, however, is difficult for two reasons. Firstly, simple solvent extraction in order to remove the active substance from the feed also removes feed constituents that produce a high level of background absorp- tion in both the ultraviolet and visible-light regions of the spectrum.Secondly, in polaro- graphic procedures (J. E. Fairbrother and W. F. Heyes, unpublished work), constituents of medicated feeds either interfere in the assay or prevent the quantitative extraction of chlorhydroxyquinoline. We theref ore decided to examine a spectrofluorimetric procedure for the determination of chlorhydroxyquinoline in a medicated pig feed that contained 600 g ton-l of chlorhydroxyquinoline (about 590 p.p.m.). The procedure had to be capable of determining the active substance in the presence of other common feed additives, such as copper and zinc salts, dimetridazole and arsanilic acid.@ SAC and the authors.798 FAIRBROTHER AND HEYES: THE DETERMINATION OF [AnaZyst, Vol. 98 EXPERIMENTAL The chlorhydroxyquinoline is extracted from the sample of medicated feed by being shaken with chloroform. After the chloroform extract has been filtered an aliquot is treated with a solution of potassium cyanide in methanol, in order to complex any co-extracted copper and zinc. A methanolic magnesium acetate reagent solution is added to the treated aliquot to convert the chlorhydroxyquinoline into its fluorescent chelate with magnesium and the solution is diluted with neutralised methanol containing phenolphthalein indicator. The alkalinity of this solution is adjusted to a defined point by the dropwise addition of methanolic potassium hydroxide solution, then the fluorescence-emission intensity of the solution is measured and compared quantitatively with that of a similarly treated chlorhydroxyquinoline standard.METHOD APPARATUS- A Baird Atomic SF1 spectrofluorimeter was used, with exit and entrance slits set at 8 pm. The scale used (coarse gain, 1000 at 1 s; fine gain, 10; photomultiplier setting, 3 or 4) was adjusted to give a meter reading of 60 for the chlorhydroxyquinoline standard solution. The spectrofluorimeter was coupled with an Advance Electronics X Y pen recorder, Model HR 100, and recorded spectra rather than direct meter readings were used. REAGENTS- All reagents were of analytical-reagent grade, unless otherwise stated. Chlorhydroxyquinoline-A sample of the batch used to medicate the feed, or a suitable reference standard.Chloroform. Magnesium acetate solution-A 0.1 per cent. m/V solution in methanol. Methanol, neutralised-With 0.1 M methanolic potassium hydroxide solution, neutralise 500 ml of methanol containing 0.75 ml of 1.0 per cent. m/V phenolphthalein solution to a faint pink colour. Phenolphthalein solution-A 1.0 per cent. m/V solution in ethanol. Potassium cyanide solution-A 0.5 per cent. mm/V solution in methanol. Potassium hydroxide solution, 0.1 M-weigh 0-561 g of potassium hydroxide into a 100-ml calibrated flask. Dissolve it in, and dilute to volume with, methanol. PREPARATION OF STANDARD- Weigh accurately 120 mg of chlorhydroxyquinoline into a 100-ml calibrated flask. Dis- solve it in, and dilute to volume with, chloroform.Mix the solution thoroughly and transfer 10ml by pipette into another 100-ml calibrated flask, dilute to volume with chloroform, and again mix thoroughly. PREPARATION OF SAMPLE- Weigh accurately 20 g of medicated feed (or an amount containing approximately 12 mg of chlorhydroxyquinoline) into a 250-ml conical flask. Add, by use of a pipette, 100 ml of chloroform, then stopper the flask and shake it thoroughly for 5 minutes. Immediately filter about 20ml of this solution through a fluted Whatman No. 4 filter-paper, rejecting the first 10ml of filtrate. REACTION PROCEDURE- Transfer, with a pipette, 5 ml of each of the standard and sample solutions into separate 100-nil calibrated flasks and treat each in the following manner. With care add, by use of an autonutic pipette, 5 ml of potassium cyanide solution followed by 10 ml of magnesium acetate solution.Mix the solutions and add about 60 ml of neutralised methanol containing phenolphthalein indicator. Then, by using a Pasteur pipette, add 0.1 M methanolic potassium hydroxide solution dropwise until the solution just turns pink. Add a further 1.5 ml of 0.1 M methanolic potassium hydroxide solution from a pipette and dilute each solution to volurne with more neutralised methanol. Mix them thoroughly, then filter, if necessary, through a Whatman No. 1 filter-paper, rejecting the first 1 O m l of filtrate.November, 19731 CHLORHYDROXYQUINOLINE IN MEDICATED PIG FEEDS 799 SPECTROFLUORIMETRIC MEASUREMENT- Scan the fluorescence-emission spectrum of a 1-cm layer of both the sample and standard solutions between 485 and 570 nm, with an excitation wavelength of 402 nm.(The emission peak should occur at approximately 500 nm.) Record the peak heights given by the sample and standard solutions, and calculate the chlorhydroxyquinoline content of the sample. If necessary, a sample of unmedicated feed should be put through the procedure as described for the sample, and a suitable blank correction shouid be made to the reading for the sample of feed. RESULTS AND DISCUSSION REACTION CONDITIONS- Inadequate recoveries of chlorhydroxyquinoline obtained in initial experiments were found to have resulted from a reduction in fluorescence yield, rather than from inefficient extraction from the feed. This result was found to correlate with the co-extraction of acidic components from the feed, and the problem was overcome by controlling the alkalinity of the solution (see Fig.1). The change in fluorescence yield brought about by the addition of alkali was first measured by use of external acid - base indicators. It was then found that the inclusion of the indicator in the reaction mixture did not significantly affect the results, and in view of the necessity to retain methanol as the main solvent, the use of the internal indicator was adopted. Fig. 1. Fluorescence (emission) spectrum of a methanolic solution of chlorhydroxyquinoline - magnesium com- plex neutralised with potassium hydroxide : A, to bromothymol blue external indicator ; B, to phenolphthalein; and C to H, to phenolphthalein with 0.26 (C), 0.6 (D), 0.76 (E), 1.0 (F), 1.6 (G) and 2-0 (H) ml excess of 0.1 M methanolic potassium hydroxide solution per 100 ml of solution. (Subsequent spectra were recorded off-set with respect to the wavelength scale so as to permit comparison) A study of the extraction procedure showed that recoveries of chlorhydroxyquinoline from medicated feeds were dependent on time, values reducing to about 70 per cent.recovery after 30 minutes’ extraction. An examination of the chloroform extracts from medicated feeds by use of atomic-absorption spectroscopy showed that copper and zinc were the main cations co-extracted with the active drug. These ions were found to compete with magnesium800 FAIRBROTHER AND HEYES: THE DETERMINATION OF [Analyst, Vol. 98 acetate reagent for chelation with the chlorhydroxyquinoline, thereby reducing the fluores- cence intensity of the chlorhydroxyquinoline - magnesium chelate solution.A small amount of extracted manganese ion had no significant effect on the fluorescence. When potassium cyanide reagent was added to the chloroform extract the copper and zinc ions complexed with it rather than with chlorhydroxyquinoline, thus overcoming the extraction problem and that of poor recoveries. FLUORESCENCE PHENOMENA- The fluorescence spectra of certain metal chelates of 8-hydroxyquinoline and its deriva- tives have been extensively reported.2-s The fluorescence spectrum of chlorhydroxyquinoline is similar to that of S--hydro~yquinoline,~~8 but its excitation and emission maxima are shifted slightly towards longer wavelengths.The possibility that interfering substances are co-extracted from the feed, as well as the limited availability of solvents of suitable purity, limited the choice of solvents to the chloroform - methanol mixture used. Aluminium, magnesium and lithium were considered as possible chelating cations, and magnesium was selected. Although aluminium was known6 to give the highest fluorescence yield, it was difficult to select an aluminium salt soluble in a chloroform - methanol mixture and also available in a sufficiently pure form for spectrofluorimetric work. Lithium had a lower fluorescence intensity than magnesium and was therefore not examined further. Under the conditions selected , chlorhydroxyquinoline showed an excitation maximum at 402 nm and a corresponding emission maximum at 500 nm.The linearity of the relation- ship between the emission intensity and the concentration of the chlorhydroxyquinoline solution was demonstrated over the concentration range 3-75 to 7.25 pg ml-l, and the fluorescence intensities of the sample and standard solutions remained reproducible for at least 1 hour after their preparation. It was considered valuable to examine the fluorescence characteristics of the three isomeric components of chlorhydroxyquinoline. The changes in fluorescence characteristics caused by variations in the ratio of the component isomers were also examined (see Table I). TABLE I FLUORESCENCE CHARACTERISTICS OF CHLORHYDROXYQUINOLINE AND ITS COMPONENT SUBSTANCES Wavelength of excitation maximumlnm 5,7-Dichloro-8-hydroxyquinoline .. 402 5-Chloro-8-hydroxyquinoline . . .. 402 7-Chloro-8-hydroxyquinoline . . .. 388 Chlorhydroxyquinoline-Batch A . . 402 B .. 402 c .. 400 D .. 402 Wavelength of emission maximumlnm 495 510 487 600 500 497 498 Relative fluorescence intensity 79.3 35.7 60-2 64.6 64-6 62.6 64.2 Although the fluorescence characteristics of the three component isomers are different , the over-all effect of these differences on the fluorescence yield of different batches of chlor- hydroxyquinoline is not critical. However, it is recommended that a sample of the same batch of chlorhydroxyquinoline that was used to medicate the feed be used as a standard in the assay procedure. RESULTS Feed blanks did not appear to make any major contribution to the fluorescence intensity, except for those from feeds that contained a high proportion of meat meal.However, because different feed bases were not studied extensively, such interferences cannot be ruled out for all grades of feed. Recoveries of chlorhydroxyquinoline from different types of feeds medicated in the labmatory are shown in Table 11.November, 19731 CHLORHYDROXYQUINOLINE IN MEDICATED PIG FEEDS TABLE I1 RECOVERY OF CHLORHYDROXYQUINOLINE FROM FEEDS MEDICATED IN THE LABORATORY AT LEVELS BETWEEN 60 AND 140 PER CENT..OF THE THEORETICAL CONTENT (600 g ton-l) 801 Halquinol Feed added/ sample g ton-l 1 600 360 840 2 600 360 840 3 600 360 840 Feed blank, per cent. of contribution of active drug 3.3 5.4 2.4 3.3 5.4 2.4 3-3 5.3 2.4 Recovery of chlorhydroxyquinoline, per cent.(corrected for blank contribution) 97.8, 97.9, 99.0 106.2, 104.2 93.0, 93.8 99.2, 96-6, 96.6 98.8, 98.9 97-9, 96.6 100.2, 102.1, 97.1 102.5, 105.5 94.4, 93.8 The feeds were chosen as being representative of the different cereals and additives used in the manufacture. Deviations from a recovery of 100 per cent. are apparent at levels of 360 and 840 g ton-l of chlorhydroxyquinoline. However, as the procedure described was developed specifically for the assay of feeds containing 600 g ton-l, the deviations encountered (+6 per cent. at 360 g ton-l and -7 per cent. at 840 g ton-l) were considered acceptable. Satisfactory recoveries of chlorhydroxyquinoline (600 g ton-l) were obtained from feeds containing the following additives- Additive Content, p.p.m. Copper . . .. . . .. .. 100 Zinc . , .. .. .. .. 100 Calcium .. .. * . . . 1000 Dimetridazole , . ,. .. .. 100 Arsanilic acid . . .. . . .. 250 The procedure described is not suitable for the determination of Halquinol in pre-mixes and concentrates. Simpler and more precise procedures for these determinations that make use of the spectrophotometric determination of the iron(II1) chelate of Halquinol in a chloroform - acetone solvent can be used (J. E. Fairbrother and W. F. Heyes, unpublished work). 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES British Patent 909,359, 1961. Udenfriend, S., “Fluorescence Assay in Biology and Medicine,” Academic Press, New York, 1962, p. 388. Ohnesorge, W. E., in Hercules, D. M., Editor, “Fluorescence and Phosphorescence Analysis,’’ J. Wiley & Sons, New York, 1966, p. 155. Parker, C. A., “Photoluminescence of Solutions,” Elsevier, London, 1968, p. 475. Stevens, H. M., Analytica Chim. Acta, 1959, 20, 389. Schulman, S. G., Analyt. Chem.. 1971, 43, 285. Schulman, S. G., and Rietta, M. S., J . Pharm. Sci., 1971, 60, 1762. Bratzel, M. P., Aaron, J. J., and Winefordner, J. D., Analyt. Chem., 1972, 44, 1240. Received May 4th, 1973 Accepted June 27th, 1973