The principles established by Shaffer and Hartmann are the basis of the cupric titration given in detail below. The cuprous titration is generally made directly in the reaction mixture, and examples are given in several of the special reducing-sugar methods described under "volumetric processes" page 185. (See also Methods of Analysis of AOAC [27].)

Reagent. Standard thiosulfate solution.-Prepare a solution_containing 39 g of pure Na2S2O3.5HO2 in 1 liter. Weigh accurately 0.2 to 0.4 g of pure Ĉu and transfer to a 250-ml Erlenmeyer flask roughly graduated by marks at 20-ml intervals. Dissolve the Cu in 5 ml of a mixture of equal volumes of HNO3 and H2O, dilute to 20 or 30 ml, boil to expel the red fumes, add a slight excess of strong Br water, and boil until the Br is completely driven off. Cool, and add NaOH solution with agitation until a faint turbidity of Cu(OH)2 appears (about 7 ml of a 25-percent solution is required). Discharge the turbidity with a few drops of acetic acid and add 2 drops in excess.

Prepare a solution of 42 g of KI in 100 ml of solution made very slightly alkaline to prevent formation of HI and its oxidation.

It is essential for the thiosulfate titration that the concentration of KI in the solution be carefully regulated. If the solution contains less than 320 mg of Cu, 4.2 to 5 g of potassium iodode should have been added at the completion of the titration for each 100 ml of total solution. If greater quantities of Cu are present, the amount of KI should be proportionately greater. The KI solution should be added slowly from a burette, with constant agitation.

Observe the volume of the Cu solution and add 1 ml of KI solution for each 10 ml of the solution undergoing titration. Titrate at once with the thiosulfate solution until the brown color becomes faint. Again observe the volume and add an additional volume of KI to make the required concentration, noting from the volume of the thiosulfate the approximate Cu content of the solution. Add sufficient starch indicator to produce a marked blue coloration. Continue the titration cautiously until the color changes toward the end to a faint lilac. As the end point is approached, add the thiosulfate in fractions of drops, allowing the precipitate to settle slightly after each addition. One ml of thiosulfate solution equals about 10 mg of Cu.

Determination.-Wash the precipitated Cu2O, cover the Gooch crucible with a watch glass, and dissolve the oxide by means of 5 ml of HNO3(1+1) directed under the watch glass with a pipette. Collect the filtrate in a 250-ml Erlenmeyer flask roughly graduated by marks at 20-ml intervals, and wash the watch glass and Gooch crucible free. from Cu. Proceed as directed under "Reagent," beginning with "boil to expel the red fumes."

Foote and Vance [28] have studied the titration and have proposed that an addition of 2 g of ammonium thiocyanate be made when the tiration has proceeded nearly to the end point. In the usual procedure the cuprous iodide is not white but slightly discolored, apparently as a result of the adsorption of iodine. Because of the greater insolubility of cuprous thiocyanate, the surface particles of cuprous iodide are changed to thiocyanate, releasing the adsorbed iodine, which thus consumes a slight additional volume of thiosulfate. The end point is exceedingly sharp, the precipitate turning completely white. This modification can be introduced into the procedure described above,

provided both standardization and determination are conducted in the same manner.

Foote and Vance performed the titration in the presence of 5 ml of sulfuric acid and in the presence of the same reagent buffered by 3 g of ammonium acetate and obtained the same analytical results. Even a small volume of nitric acid (1 ml) produced no measurable deviation. In view of these results, it is probable that the procedure given on page 181 could be materially simplified.

(e) PERMANGANATE METHOD

When cuprous oxide is dissolved in a ferric sulfate solution the latter is reduced to ferrous sulfate and can be titrated with standard permanganate. The reduced copper can then be computed on the basis of the stoichiometric relations. This method of determination of reduced copper was originally proposed by Mohr [29], but has erroneously been attributed to Bertrand [16]. Mohr prescribed that the ferric sulfate be dissolved in sulfuric acid, and this procedure was adopted by Bertrand, who based his tables (see page 587) upon the volume of permanganate consumed, instead of referring his sugar to copper. Subsequently, it was discovered that the permanganate volumes, when converted to copper, gave results which were invariably about 1.4 percent too low, and many authors advocated a blanket correction by this amount. Schoorl and Regenbogen [30] ascribed the low results to the rapid oxidation of ferrous sulfate by air. The oxidation is more rapid than is ordinarily the case with ferrous sulfate, but in the presence of copper, which increases the rate of oxidation, and in the presence of the asbestos, which carries finely subdivided air, the amount of oxidation appears to be accounted for.

Schoorl and Regenbogen found that if the cuprous oxide was dissolved in neutral ferric sulfate, or better, ferric alum, before addition of sulfuric acid, correct results were obtained. A brownish-red clear solution is obtained in which the Fe2O3 apparently remains dissolved in the form of a basic salt. The authors suggest that the following reaction occurs:

Cu2O+2Fe2(SO4)3 →2CuSO4+2FeSO4+Fe2O3+SO3.

Reagents. Prepare a potassium permanganate solution, about 0.1573 N, containing 4.98 g per liter. After several days' aging, filter through asbestos or sintered glass. Standardize by either of the following methods.

(a) Transfer 0.35 g of sodium oxalate (dried at 103°C) [31] to a 600-ml beaker. Add 250 ml of sulfuric acid (5+95) previously boiled for 10 minutes and cooled to 27+3°C. Stir until the oxalate is dissolved. Add 29 to 30 ml of permanganate solution at a rate of 25 to 35 ml per minute while stirring slowly. Let stand until the pink color disappears (about 45 seconds). Heat to 55° to 60°C, and complete the titration by adding permanangate solution until a faint pink color persists for 30 seconds. Add the last 0.5 to 1 ml dropwise, allowing each drop to become decolorized before the next is added. Determine the excess of solution (usually 0.03 to 0.05 ml) required to impart a pink color to the same volume of acid boiled and cooled to 55° to 60°C.

In potentiometric titrations the correction is negligible if the end point is approached slowly.

(b) Transfer [32] about 0.3 g of As2O3 (dried at 110°C) to a 400-ml beaker. Add 10 ml of a cool solution of sodium hydroxide (20 percent) and allow to stand until dissolved, stirring occasionally. Add 100 ml of water, 10 ml of HCl (sp gr 1.18), and 1 drop of 0.0025 M KIO, or KI. Titrate with permanangate solution until a faint pink color persists for 30 seconds, adding dropwise the last 1 to 1.5 ml, and allowing each drop to become decolorized before adding the next. Determine by a blank test with all reagents except As2O3, the volume of KMnO4 (usually about 0.03 ml) required to duplicate the pink color of the end point. The end point can also be determined with ferrous phenanthroline indicator, in which case 1 drop of a 0.025 M solution of the indicator is added as the end point is approached. Determine blank correction.

The titration can also be conducted potentiometrically. Ferric sulfate.-Dissolve 135 g of ferric ammonium alum or 55 g of Fe2(SO4)3 (anhydrous) and dilute to 1 liter. Determine Fe2(SO4)3 in the stock supply by strong ignition to Fe2O3. Ferric sulfate, particularly after exposure to light, contains a small quantity of reducing substance which must be oxidized before the solution is used. Titrate 50 ml of the ferric sulfate solution, acidified with 20 ml of 4 N sulfuric acid, with the permanganate and use the titer as a zeropoint correction in subsequent titrations.

4 N sulfuric acid.

Ferrous phenanthroline indicator.-Dissolve 0.7425 g of orthophenanthroline monohydrate in 25 ml of 0.025 M ferric sulfate solution (6.95 g of FeSO4.7H2Ŏ in 1 liter).

Procedure. Filter the cuprous oxide in a Gooch crucible and wash the beaker and precipitate thoroughly. Transfer the asbestos film to the beaker with the aid of a glass rod. Add 50 ml of the ferric sulfate solution and stir vigorously until the cuprous oxide is completely dissolved. Examine for complete solution, holding the beaker above the level of the eye. Add 20 ml of 4 N sulfuric acid and titrate with standard permanganate. As the end point is approached, add 1 drop of ferrous phenathroline indicator. At the end point the brownish solution changes to green. One milliliter of 0.1573 N permanganate equals 10 mg of copper.

The concentration of the permanganate solution in the method described above is such that a single filling of the burette supplies enough reagent for the determination of the maximum amount of copper precipitated by any of the usual methods of analysis. For many purposes the reagent is too concentrated. Thus, for the small quantities of copper precipitated by high-grade cane- or beet-sugar samples, a N/30 permanganate solution is more commonly used. An outstanding disadvantage of the method is that the larger weights of cuprous oxide dissolve with great difficulty in the ferric sulfate solution.

(f) DICHROMATE METHODS

(1) ELECTROMETRIC.-Jackson and Mathews [33] have described a rapid and convenient method for the determination of reduced copper, in which cuprous oxide is oxidized in hydrochloric acid solution by an excess of standard dichromate, the excess being determined by back titration with ferrous sulfate to an electrometric end point. The electrometric apparatus can be obtained by purchase or it may be

assembled easily. Its arrangement is shown diagrammatically in figure 38. The current from two dry cells, E, flows continuously through a resistance, R, of about 3,000 ohms. The potassium chloride bridge, C, of a calomel cell and a bare platinum wire, P, dip into the solution undergoing analysis. The sliding contact on the rheostat is so adjusted as to produce a zero reading on the galvanometer, G, when the electrodes dip into an acidified ferrous-ferric system containing a slight excess of dichromate. As ferrous sulfate is added, there is but slight change in the position of the galvanometer needle until the end point is reached. At the end point the deflection of the needle is so large that no actual measurements of electromotive force are required. The galvanometer must be fairly sensitive.

E

G

Reagents. Potassium dichromate. Dissolve 7.7135 g of pure C crystals, preferably pulverized and dried at 150° C, and make to a volume of 1 liter. One milliliter of this solution, which is 0.1573 N, is equivalent to 10.00 mg of copper.

FIGURE 38.-Apparatus for electrometric determination of copper.

Ferrous ammonium sulfate.-Dissolve 61.8 g of the hexahydrate crystals, add 5 ml of sulfuric acid, and make up to 1 liter. Equally suitable is 43.8 g of FeSO4.7H2O, acidified and made to 1 liter. The ferrous solutions are oxidized slowly by air and must occasionally be titrated against the standard dichromate solution.

Procedure. Collect the precipitated cuprous oxide on a Gooch crucible and wash thoroughly. Detach the mat with a glass rod and transfer to the reaction beaker. Add a small volume of water and disintegrate the mat. Pipette accurately a volume of standard dichromate in excess of the quantity required to oxidize the cuprous oxide. In general, the approximate weight of copper will be known or can be roughly estimated, but in any case a sufficient volume must be added to supply an assured excess. Add from a graduated cylinder 50 ml of 1+1 bydrochloric acid with stirring, and continue to stir until all cuprous oxide is dissolved. Immerse the crucible in the solution and be sure that the adhering cuprous oxide is dissolved. Remove the crucible with the glass rod, washing it free of solution. Dilute the solution to about 250 ml and titrate the excess dichromate with ferrous sulfate to an electrometric end point. Determine the ratio of ferrous sulfate to dichromate, and thence compute the volume of dichromate required for the oxidation of cuprous oxide. This volume multiplied by 10 gives directly the number of milligrams of copper reduced.

(2) COLORIMETRIC.-Many laboratories lack the necessary equipment for electrometric titration, but a number of suitable internal indicators are known which make possible a colorimetric end point. A particularly serviceable one, orthophenanthroline,10 was shown by Walden, Hammett, and Chapman [34] to be applicable in oxidationreduction reactions.

Jackson and McDonald [35] have applied the colorimetric dichromate method of titration to the determination of copper in cuprous oxide.

10 Manufactured by the G. Frederick Smith Chemical Co., Columbus, Ohio.

Reagents. Potassium dichromate.-Standard solution, 0.1573 N, containing 7.7135 g of pure dry crystals in 1 liter. One milliliter is equivalent to 10 mg of copper. Ferrous ammonium sulfate.-Dissolve 61.9 g of the hexahydrate, add 5 ml of sulfuric acid, and complete the volume to 1 liter. Hydrochloric acid. Approximately 6 Ñ. Phenanthroline-ferrous complex.-Dissolve 0.725 g of orthophenanthroline monohydrate in 25 ml of 0.025M ferrous sulfate solution (6.95 g of FeSO4.7H2O in 1 liter).

Procedure. Estimate the volumes of hydrochloric acid and water which at the end of the titration will yield a concentration of HCl of about 2 N in about 200 ml of final volume. An error of from 5 to 10 percent can be tolerated.

Collect the precipitated cuprous oxide on a Gooch crucible and wash thoroughly. Detach the mat with a glass rod and transfer to the reaction beaker. Add a small volume of water and disintegrate the mat. Pipette accurately a volume of standard dichromate in excess of the quantity required to oxidize the cuprous oxide. In general, the approximate weight of copper will be known or can be roughly estimated, but in any case a sufficient volume must be added to supply an assured excess. Add rapidly the whole required volume of hydrochloric acid with continuous stirring and continue to stir until all cuprous oxide is dissolved. Immerse the crucible in the solution and be assured that the adhering cuprous oxide is dissolved. Remove the crucible with the glass rod and wash it with the water from the graduate. Add 1 drop of phenanthroline solution and titrate with ferrous sulfate to the permanent appearance of the brown ferrousphenanthroline complex. As the end point is approached, the brown color appears and fades as each of the last few drops is added. The ferrous sulfate must be added until the color is permanent, the additions finally being in fractions of drops.

Determine the ratio of concentrations of ferrous sulfate and dichromate and then compute the volume of dichromate required for the oxidation of cuprous oxide. This volume multiplied by 10 gives directly the number of milligrams of copper reduced.

The electrometric and colorimetric dichromate methods described above are still in an experimental stage in the laboratories of this Bureau. A comparison has been made of the values of copper reduced by the Munson and Walker method and determined by the thiosulfate and by colorimetric dichromate titration, respectively. At low concentrations of copper, identical values were obtained by the two methods. Above 200 mg of copper, systematic differences occurred, which rose to a maximum of 0.2 percent for levulose and 0.3 percent for invert sugar. This discrepancy would disappear if a special table were used for the dichromate methods. Further studies are in progress, since the methods recommend themselves on account of their great rapidity and convenience.

4. VOLUMETRIC METHODS

(a) LANE AND EYNON METHOD

The most convenient, most expeditious, and frequently the most accurate method for the determination of reducing sugars is the volumetric method of Lane and Eynon [36]. In principle, the method involves the determination of the volume of sugar solution required