(b) DE WHALLEY METHOD FOR THE DETERMINATION OF INVERT SUGAR IN REFINED WHITE SUGARS [56]

Range, 0.001 to 0.015 percent, capable of extension to higher percentages by dilution with pure sucrose.

Select test tubes of white glass 6 by 4 inches having a uniform weight of about 9.4 to 9.6 g. Fit large rubber rings about the tops in order to support them in the water bath. The bath is of sheet copper, 7-inch cube with three holes 1 inch in diameter, the central hole being used for the test and the other two as steam vents. A constant water level is maintained 2 inches below the top of the bath. Heat is applied by a ring burner protected from draught. For accurate work, maintain a constant gas pressure of 3.5 to 3.75 inches of water by means of a pressure regulator.

Prepare accurately solutions of, (1) 0.20-percent methylene blue and (2) 3 N sodium hydroxide.

Grind the sample of sugar and transfer 7 g, weighed correctly within 0.05 g, to a clean, dry test tube. Add 6 ml of distilled water, 1 ml. of methylene blue (microburette or calibrated pipette), and 1 ml of the caustic soda solution.

Shake the test tube (closed with a rubber stopper) vigorously for 15 seconds, and immerse in the boiling water bath for exactly 2 minutes. Remove from the bath and compare immediately with the row of standard tubes.

Standards. Prepare a solution of 19.5 g of CuSO4.5H2O in boiled. distilled water and make to 500 ml. Measure accurately the following volumes into 50-ml volumetric flasks and add to each 10 ml of ammonium hydroxide (sp gr 0.880; 32.9 percent of ammonium hydroxide by titration) and make to volume.

TABLE 26.-Standards for use with the de Whalley method

When a solution of reducing sugar is added to a boiling alkaline copper solution, the potential set up at a platinum electrode during the first part of the titration is represented by the expression

The concentration of cupric ions is governed by the composition of the alkaline tartrate solution and by the removal of copper in the form of insoluble cuprous oxide. The concentration of cuprous ions is determined by the solubility product of cuprous hydroxide. When

all the cupric ions have been removed the potential changes rapidly. This sudden change in potential indicates the end of the reaction between cupric ions and reducing sugar.

A micro application of the electrometric method has been made by Niederl and Müller [59]. They used alkaline copper solutions composed of (A) 3.95 g of CuSO4 5H2O in 500 ml of solution; (B) 19.75 g of Rochelle salt and 7.4 g of NaOH in 500 ml of solution. For the analysis, they mixed equal volumes of A and B. Their apparatus consisted of two vessels, each containing an alkaline copper solution of the same concentration, in which was immersed a platinum electrode. The vessels were connected by an agar-potassium chloride bridge.

For analysis, 3 to 5 mg of substance is accurately weighed, dissolved in water (1 ml for each mg), and transferred to a microburette. One milliliter of the mixed copper reagent, containing 1 mg of copper is placed in each of the two vessels. A reading is taken on the potentiometer and again after the solution in one vessel is brought to boiling. A preliminary titration is carried out as follows: 0.1 ml of the sugar solution is added, the mixture boiled, and a reading taken again. Additions of the sugar solution are continued until 1.2 ml has been used. In the region of the end point the change of potential is very considerable. The maximum change per 0.01 ml marks the end point. The actual determination is carried out in the same way, but larger volumes of the sugar solution are used until the end point is approached.

The following method, in which the end point is determined by the change of potential between two thick copper wires serving as electrodes, has been devised by Tryller [60]:

Reagents and apparatus.-(1) Soxhlet solution (I), 34.639 g of CuSO4.5H2O made to 500 ml of solution; (II), 173 g of Rochelle salt and 50 g of NaOH made to 500 ml of solution.

(2) Sodium sulfate for cell, 39.415 g of anhydrous Na2SO, made to 1 liter of solution.

(3) An electrode of thick copper wire (about 2 mm in diameter). (4) A cell of Pyrex-glass tubing (8 mm inside diameter with a plug (8 mm long) of plaster of paris at one end. The plug should be washed with the cell solution in which it should be stored when not in use.

(5) A dead-beat moving-coil galvanometer with central zero. (6) A press-key.

(7) Burette stand, 250-ml Pyrex flasks, burettes, etc.

Method. The arrangement consists of a galvanometer wired directly to the electrodes through a press-key. Any convenient

number of flasks may be wired to the same galvanometer. Figure 39 shows the flask and electrode assembly. The 250-ml, flat-bottomed flask is fitted with a cork (7) which holds the thick copper-wire electrode, the tip of the burette (6), the cell (2), and the steam vent (5) in position. A copper wire makes contact between the inner cell and the lead to the galvanometer. The liquid in the inner cell is composed of 5 ml of Soxhlet solution II, 5 ml of sodium sulfate solution, and 40 ml of water. This solution may be made up and stored for short periods of time, the liquid in the cell being changed after every six determinations.

5

A plaster-of-paris plug lasts for 25 to 30 determinations, after which time its sensitivity is reduced. The changing of the cell is easily carried out by making the hole for the cell a sliding fit and placing the rubber collar (9) around the top and adjusting it so that the plug is immersed to a suitable depth. With this arrangement, the cell may be changed while waiting for the solution to boil. A perforated asbestos slab resting on the shoulder of the flask deflects the hot gases away from the measuring apparatus.

23

4

7

In carrying out the determination, the method of Lane and Eynon is followed to the point where methylene blue is added. At this time the press-key is depressed; and, if the galvanometer needle is deflected, the sugar solution is added dropwise. The swing of the needle becomes less and less until finally 1 drop will cause a deflection in the opposite direction. Toward the end of the titration it is advisable to wait 3 to 5 seconds between the addition of 1 drop of sugar solution and the next, as at this point there appears to be a slight lag before the system attains equilibrium.

FIGURE 39.-Assembly for electrometric determination of reducing sugars.

1, Copper electrode; 2, cell; 3, cell liquid; 4, plasterof-paris plug; 5, steam vent; 6, burette tip; 7, cork; 8, asbestos; and 9, rubber collar.

8. SELECTIVE METHODS

(a) JACKSON AND MATHEWS MODIFICATION OF NYNS METHOD FOR LEVULOSE

Elaborating an observation made by Biourge [61], that at 50° C levulose had, in the presence of Ost reagent, a reducing power 10 times as great as that of dextrose, Nyns [62] determined the copperlevulose equivalents over a wide range of sugar concentrations by heating the reaction mixtures at 48.6° C for 21⁄2 hours. Jackson and Mathews [63] modified the Nyns procedure, shortening the time of reaction by raising the temperature to 55° C and specifying an increased concentration of copper, thus enlarging the range of sugar concentrations which could be analyzed and supplying more rapid methods of copper analysis.

(1) STANDARD SOLUTIONS

Potassium dichromate. Prepare a standard solution N

0.1573, by dissolving 7.7135 g of the pure dry salt and filling to 1 liter. One milliliter is equivalent to 10.00 mg of copper.

Ferrous ammonium sulfate.-Dissolve 61.8 g of the hexahydrate, add 5 ml of concentrated sulfuric acid, and complete to a volume of 1 liter.

Ost solution. Dissolve 250 g of potassium carbonate (anhydrous) in about 700 ml of hot water and add 100 g of pulverized potassium bicarbonate. Agitate until completely dissolved. Cool and add, with very vigorous agitation, a solution of 25.3 g of pure CuSO4.5H2O in 100 to 150 ml of water. Adjust to room temperature, make to 1 liter, and filter.

(2) ANALYTICAL PROCEDURE.-Transfer 50 ml of Ost reagent to a 150-ml Erlenmeyer flask. Add by means of an accurately graduated pipette a volume of the solution to be analyzed. This should contain not more than 92 mg of levulose or its equivalent of a levulosedextrose mixture, remembering that dextrose has about one-twelfth the reducing power of levulose. Add enough water to make the total volume 70 ml. Immerse in a water bath regulated preferably to within 0.1° C at 55° C. Digest for exactly 75 minutes, agitating with a rotary motion at intervals of 10 or 15 minutes.

At the expiration of the prescribed time, filter the precipitated copper on a closely packed Gooch crucible and wash the flask and filter thoroughly without attempting to transfer the precipitate quantitatively. (It is well, however, to transfer all of the loose cuprous oxide, leaving in the flask only the small portion which adheres to the walls.) Remove the asbestos mat by means of a glass rod and transfer to a 400-ml beaker. Add 5 or 10 ml of water and disintegrate the asbestos mat. Add a carefully measured volume of standard potassium dichromate (NX0.1573) in excess of the volume required to oxidize the cuprous oxide. In many cases the expected amount of precipitated copper will be roughly known, and it will be possible to gage the volume of dichromate which will supply a 3- to 4-ml excess. If the amount of precipitated copper is not even roughly known, it is preferable to add an amount which will supply an assured excess. A very large excess introduces no uncertainty, provided its volume is accurately measured. Of this volume about 1 ml is added to the original reaction flask in order that the residual cuprous oxide may be dissolved and subsequently added to the remainder of the solution. Add to the Erlenmeyer flask by means of a graduated cylinder about 50 ml of hydrochloric acid (1+1). Pour the acidified solution slowly into the 400-ml beaker with constant stirring, and continue to stir until the cuprous oxide is completely dissolved. Wash the Erlenmeyer flask with a jet from the wash bottle, receiving the rinsings in the beaker. Examine the asbestos critically by looking through the bottom of the beaker, which is held above the eye. (If any undissolved cuprous oxide remains, it can be clearly discerned as dark-colored particles.) Immerse the crucible in the acidified solution to dissolve such cuprous oxide as remained in it. Remove the crucible with a glass rod, washing it free of solution. Dilute the solution thus prepared to about 250 ml and electrometrically titrate the excess of dichromate with ferrous sulfate.

The excess of dichromate can also be determined satisfactorily

with the use of the internal indicator, orthophenanthroline, as described on page 184. Indeed, any other method of copper analysis will serve equally well, but in all cases control analyses with pure levulose should be made and a correction applied to bring the copper equivalents into correspondence with table 93, p. 597.

In table 93 is given the correspondence between the reduced copper and the levulose in the sample.

(3) ANALYSIS OF DEXTROSE-LEVULOSE MIXTURES.-Jackson and Mathews found that the method, when applied to dextrose-levulose mixtures, lacked perfect selectivity and that it was necessary to apply a correction for the reducing power of dextrose. Throughout the entire range of concentrations 12.4 mg of dextrose proved to have the same reducing power as 1 mg of levulose.

For the analysis of an unknown mixture of dextrose and levulose. two equations are necessary for a solution, and Jackson and Mathews recommended a combination of the Lane and Eynon titration for total reducing sugar and the modified Nyns method. The Lane and Eynon titration (25 ml of Soxhlet reagent) is corrected to correspond with the Lane and Eynon table, as described on page 187, and multiplied by the number of milligrams of "apparent" levulose, that is, the levulose equivalent of the copper from table 93. The product, divided by 100, is referred to table 94, p. 599 and the ratio of levulose to total sugar is read from the appropriate column.

(b) HINTON AND MACARA METHOD FOR LEVULOSE IN CONDENSED MILK

Hinton and Macara [64] devised a method which they applied to the determination of levulose in sweetened condensed milk, but which obviously is capable of a more general application. In outline, the method consists in the oxidation of aldose groups to aldonic acids, leaving the levulose unaltered. The levulose is then determined by reducing-sugar analysis. A control solution free of levulose is analyzed simultaneously.

Reagents. (1) Sucrose solution, approximately 9 g of sucrose per 100 ml (freshly prepared).

(2) Iodine solution, 13 g of iodine and 15 g of KI per 100 ml.

(3) Mixed alkaline solution, equal parts of 2 N Na2CO3 and 2 N NaOH.

(4) Sulfuric acid, approximately 5 N.

(5) Sodium sulfite solution, 20 percent by volume.

(6) Dilute sodium sulfite solution, 2 percent freshly prepared; or diluted from 20-percent solution.

(7) Luff solution. Dissolve 25 g of CuSO4.5H2O in 100 ml of water; 50 g of citric acid in 50 ml of water; 338 g of Na2CO3.10H2O in 300 to 400 ml of lukewarm water. Add the citric acid solution to the sodium carbonate solution, and then add the copper solution. Mix, cool, make up to 1,000 ml and filter. This solution should be accurately prepared, and 10 ml of the finished solution should require approximately 45 ml of 0.5 N sulfuric acid for neutralization to methyl

orange.

(8) Iodate-iodide solution, 2.7 g of KIO,, 30 g of KI and 10 ml of 0.5 N NaOH solution per liter.

(9) Potassium oxalate solution.

(10) Sodium thiosulfate solution, 0.05 N.

323414°-42-15