568 ANALYTICAL METHODS COMMITTEE : THE DETERMINATION OF [Vol. 75 Analytical Methods Committee REPORT PREPARED BY THE CAROTENE PANEL OF THE SUB-COMMITTEE ON VITAMIN ESTIMATIONS The Determination of Carotene in Green-Leaf Material Part I. Fresh Grass THE Analytical Methods Committee has received the following Report from the Carotene Panel of the Sub-committee on Vitamin Estimations, and its publication has been duly authorised. PROLEGOMENON In 1948 the Sub-committee on Vitamin Estimations appointed a Panel “to investigate the application to other green-leaf materials of the method outlined by Dr. W. A. G. Nelson at the April, 1947, meeting of the Society for the estimation of carotene in dried grass, and to advise the Sub-committee on the practicability of extending the method to animal and vegetable food products generally,” under the Chairmanship of Mr.A. L. Bacharach. The other members were Dr. R. G. Booth, Dr. V. H. Booth, Dr. J. Green, Mr. N. T. Gridgeman (Hon. Secretary), Mr. T. Barton Mann, Dr. F. E. Moon, Professor R. A. Morton, Dr. W. A. G. Nelson and Dr. S. Y. Thompson. Mr. D. C. M. Adamson served for a while but was unable to continue. The method of Nelson being inapplicable as it stands to fresh green-leaf material (for which it was not designed), a new method has been derived from the work of V. H. Booth ( J . SOC. Chem. I d . , 1945, 64, 162) and T. Barton Mann (Analyst, 19p4, 69, 34). This method is based solely on experimental work with fresh grass, which was used because of its interest per se, its constant availability and the comparatively small sampling error involved in analysing it.It is hoped to consider later any special problems that may arise in estimating the carotene content of certain green leaves in common human consumption, such as cabbage and lettuce, though it is realised that they present formidable sampling difficulties and that any figure purporting to express the carotene content of, for example, a whole cabbage or a whole lettuce is of doubtful practical and scientific value. Meanwhile, there is no evidence for believing that the analytical method for determining carotene in the fresh grass will not, apart from sampling difficulties, apply to green-leaf material generally. GENESIS OF METHOD The carotenoids that, with chlorophyll, make up the pigment of green-leaf material, are divisible into two groups according as they are biologically inert or exhibit provitamin A activity.The former comprise the carotenols and their esters; the latter are dominated byNov., 19501 CAROTENE IN GREEN-LEAF MATERIAL: PART I. FRESH GRASS 569 p-carotene, and usually include small quantities of congeners. Separation of these allied hydrocarbons from /?-carotene is difficult ; fortunately the relative quantities (in green leaf) justify their inclusion and estimation together with /?-carotene as simply “carotene,” and this is normal practice. Until about 1940 the pigments were usually partitioned by phase separation; subsequently the more selective method of chromatography has replaced it. Numerous variants of this method have appeared, all the more important of which have been studied, and many tried, by the Panel, which now recommends two alternative adsorbents.Efficient use of either demands, however, careful working, as indeed does the whole process of estimation, although it is in essence simple. Some forethought is needed for the drawing of the sample, which must be fresh and must be processed without delay, for the cutting of tissue initiates enzymic destruction of carotene; and throughout the operations the worker must be alert to the fact that excessive heat, light and air are inimical to accuracy. To avoid delays and consequent losses, it is advisable to mobilise all apparatus and reagents beforehand. There are four stages in the method: (1) comminution-extraction in cold light petroleum and acetone; (2) removal of acetone from the extract by washing with water; (3) chromato- graphic separation of the carotene on a column of bone-meal or alumina; (4) photometric or tintometric estimation of the carotene in solution.Three not uncommon features of other carotene methods have been avoided : mechanical homogenisation, which needs special apparatus and is more efficient than the method here recommended; alkali treatment, which promotes formation of epiphasic pigments ; and heat, which fosters isomerisation. The choice of solvents requires explanation. Acetone, one of the best extractors of carotene from moist vegetable matter, tends to hold particles in suspension ; this necessitates the operation, undesirable in this context, of filtration ; furthermore, acetone is unsuitable in chromatography.Light petroleum, although itself an inefficient extractor, yields clean decantable solutions and is eminently suitable for chromatography. It has been found that a mixture of equal parts of acetone and light petroleum retains the advantages of both components and the disadvantages of neither, provided the acetone is removed before chromatography. APPARATUS Pestles-Pestles 12 cm. long, made from 8-mm. glass rod flattened at one end, and provided with handles made from corks (see Fig. 1). Squat beakers-Heavy-gauge squat beakers, 50 to 60ml., to serve as mortars. The beaker bottoms must have plane inside surfaces. Separating funnels-About 250-ml. capacity, of narrow conical type with outlet tubes 7 cm.long. The taps should be ungreased or lubricated with a non-fatty compound. It is wise to cover the sides with opaque paper. Continuous washing device (see Fig. 2)-The glass drip tube is sleeve-joined with heavy- gauge rubber tubing-clamps being thereby obviated-to the outlet tube of a separating funnel. A bigger separator, or other suitable reservoir, steadily releases large drops of water. Chromatograph tubes (see Figs. 3 and 4)-The recommended dimensions are approximate. Water pump, Buchner JEask, measwing cylinders, etc. A spectrophotometer or absorPtiometer or colorimeter. REAGENTS, ETC. Light #etroZeum-Boiling range 40” to 60” C. A cetone-Redis t illed. Quartz ~owder-“Medium fine,” average particle size 0.17 mm. Quinol. Sodium sulphate-Anhydrous. fi-Carotene standard-For calibration of absorptiometer or colorimeter.For routine purposes commercial crystalline carotene is satisfactory. Bone-meal-Bone-meal specially prepared for carotene chromatography can now be purchased. It is advisable to ensure that bought material contains nothing that will pass a 120-mesh sieve, because “fines” retard the flow of carotene. Preparation, from steamed bone flour (synonyms, sterilised bone-meal, feeding bone flour) passing 80-mesh, but retained by 120-mesh sieves, is not difficult. Extract by boiling under570 ANALYTICAL METHODS COMMITTEE THE DETERMINATION OF Pol. 75 reflux about 100 g. of the sieved meal in a Tate tube for about 24 hours with a 3 + 1 + 1 mixture of light petroleum, acetone and ether (ether meth. sp.gr. 0.720); dry overnight at 100" C.; immediately before use wash with light petroleum to remove any traces of the mixed solvents. Ahminu and sodium suZ@ate-This mixture is an alternative to the bone-meal. Mix equal weights of alumina (as sold for chromatography; 200 mesh) and anhydrous sodium sulphate (100 mesh) and dry in 1 to l*-cm. layers, at 150" C. for 12 to 16 hours. The heating conditions may have to be altered for particular materials-brands differ in adsorptive properties. A good preparation should (a) allow light petroleum to percolate evenly down a standard column at a rate of about 1 cm. per second-slower flow is usually due to the presence of too many fines, and faster or uneven flow to too much coarse material (which I I I Dark Paper To Fig. 1. Pestle Fig.2. Continuous washing apparatus grinding will remedy), and (b) adsorb as a band at the top of a standard column all the pigment from a light-petroleum extract of 0.5 g. of green leaves; from this band the carotene should be elutable with 15 ml. of a 2 + 98 mixture of acetone and light petroleum as a well-defined evenly descending orange zone. The drier the adsorbent, the poorer and slower will elution be. Slight over-dryness can be compensated by the use of more acetone, e.g., a 3 + 97 mixture, but a considerably over-dry adsorbent will have to be diluted with unheated sodium sulphate and alumina. The wetter the adsorbent the greater the likelihood of elution of non-carotene pigments with the carotene as a thin sharp line; over-wet material must be further heated.Once prepared and tested, a batch of the adsorbent can be stored in tightly- stoppered bottles and used at any time without further treatment.Nov., 19501 CAROTENE IN GREEN-LEAF MATERIAL: PART I. FRESH GRASS 571 t5 cm.' \ Anhyd. Na2S04 4 -Cotton Woof Bone Meal - Fig. 3. Bone-meal column 5 cm. f 7; F I n t . Diam. = 1.3 cm. -Alumina + Na2SOd -Cotton Wool Fig. 4. Alumina column872 ANALYTICAL METHODS COMMITTEE THE DETERMINATION OF [Vol. 75 SAMPLING AND PREPARATION Quantities of the order of 1 to 2 g. are required for one determination. Not less than two, and preferably three, simultaneous samples should be taken and assayed independently. Care should be taken to minimise leaf damage. Weigh a grinding beaker. If the leaves are wet, gently press them between blotting- or filter-paper to remove extraneous moisture, quickly cut off pieces or strips from various parts or bunches, drop them into the beaker, and weigh.Add approximately the same weight of quartz, and at once cover with 5 to 8 ml. of a mixture consisting of equal parts of acetone and light petroleum and containing 0.1 per cent. of quinol. EXTRACTION . Firmly grind the mixture in the solvent for a minute or so, allow to settle, and then decant into the separator containing water. Add fresh solvent to the grist, regrind, settle, and decant as before. Not less than five, and sometimes up to ten, extractions will be required. If in doubt about completion, decant the last extract into a small separator half-full of water and shake gently; any pigment will then easily be visible through the width of the thin layer.If any pigment collects below the spout of the beaker it can be recovered on a scrap of filter-paper, which is subsequently placed in the beaker. If the grist becomes dehydrated during extraction, the efficiency of extraction and settlement of the particles will be impaired ; a drop or two of water should therefore be added. REMOVAL OF ACETONE Repeat until no more coloux can be extracted. The separator containing the extract, over water, now takes its place as the lower, paper-covered separator in the washing apparatus shown in Fig. 2. Adjust the tap of the upper and larger “water reservoir” separator to deliver 100 to 200 large drops of ordinary tap water per minute directly on to the solution surface; if the water runs down the inside of the funnel the washing action is lessened.Open the lower tap fully and slowly oscillate the separator in the plane of the S-tube (whose top should reach just over half-way up the separator bowl) until the tube is full of water. Surplus water will overflow, taking with it acetone, quinol and other water-soluble substances. About 14 litres of water will be needed to remove all the acetone, any trace of which will interfere with the subsequent chromato- graphy. The depth of water below the solvent level must at all times be sufficient to prevent mechanical loss of pigmented layer; about 6 to 8 cm. is usually necessary. CHROMATOGRAPHY Either a bone-meal column (Fig. 3) or an alumina-Na,SO, column (Fig. 4) may be employed. With bone-meal, use a “straight-through” method, i.e., pass the light-petroleum solution of the pigments through the column (previously moistened with light petroleum) by gentle suction; the carotene will elute directly, leaving all other fractions in a zone at the top.Colourlessness of the eluate indicates completion; in case of difficulty of decision on the end- point, a test tube placed in the collecting flask, after most of the carotene has obviously come through, can be used to collect fractionally; bulkiness of the final solution can thus be avoided. The column can be used many times without cleaning. When the adsorbed chlorophylls and irrelevant carotenoids eventually choke the column its useful life may be extended as follows: wash out with acetone and remove the acetone with light petroleum; the column is then ready for further use.Moisten the column with light petroleum, pour the solution of the pigments on top, rinse with small amounts of the same solvent, applying suction as required. All the pigments should adsorb at the top of the column. Suck through 10 to 25 ml. of a solution of 2 per cent. of acetone in light petroleum; carotene will be seen to free itself from the other pigments and to descend the column. The trailing edge should be fairly clear; a diffuse band indicates that the adsorbent is too strong (possibly having caused isomerisation) and that satisfactory elution of the carotene cannot be ensured. If the adsorbent is too moist, or if the acetone has been in- completely removed, elution will be too rapid and unspecific.To keep down the bulk of the final solution do not begin to collect the eluate until the leading edge of the descending carotene zone is near the bottom of the column, With the alumina column use an adsorption-elution procedure.NOV., 19501 CAROTENE IN GREEN-LEAF MATERIAL: PART I. FRESH GRASS 573 ESTIMATION If a spectrophotometer is used, work on the basis of Amax. = 450 mp., at which E:&, x 400 = p.p.m. If a light-filter absorptiometer or acolorimeter is used, calibrate with a solution of the carotene standard in light petroleum. Carotene obtained by the bone-meal method appears to be all-trans ,%carotene. On the reasonable assumption that the alumina product is similar, both results may alter- natively or additionally be expressed in the form of International units of Provitamin A per gram; one such unit is 0.6 pg.of ,&carotene. PRECISION Some typical duplicate results on ten pairs of samples of fresh grass are given below- Iaboratory Instrument Carotene, p.p.m. A Lovibond colorirneter . . .. .. 110, 112; 112, 115 B Pulfrich absorptiometer . . .. . . 117, 118; 119, 123 C Spekker absorptiometer . . .. . . 105, 120; 107, 112 D Visual spectrophotometer . . .. 128, 132; 144, 145 E Photo-electric spectrophotometer . . 131, 132; 144, 153 N.B.-The ten samples are all different. From these and all other available replicated results the coefficient of variation of one determination was found to be approximately 4. This means that in any given set of replicated results, two-thirds of them will lie within the range 96 to 104 per cent. of the over-all mean, and only one result in twenty will be more than 1 8 per cent. from the over-all mean. This does not, however, take into account inter-laboratory differences, which mainly arise from imperfect standardisation of instruments rather than manipulative errors. No comparisons between laboratories were carried out in the trials on which this report is based, because of the difficulty of transporting fresh grass samples without loss of moisture or stimulation of enzymic action. Evidence from another comparable investigation may, however, be quoted. The Carotene Committee of the Grass Driers’ Association recently organised collaborative trials of a method of determining carotene in dried grass, and the variation between laboratories was found to average 4.6 (coefficient of variation). Those laboratories that used spectrophotometers showed better agreement than those that used absorptiometers or colorimeters-which is to be expected. Taking these things into con- sideration it can be said that the coefficient of variation of single estimates of the carotene content of fresh grass is probably about 6, and that of duplicate estimates about 4.5.