(11) Soluble starch solution, approximately 2 percent.

(12) Control serum, prepared from fresh milk with the same quantities of ammonia and acetic acid and precipitants as for 40 g of condensed milk, made up to 200 ml and filtered.

Procedure 1, oxidation of aldose sugars.-Pipette 10 ml of the condensed milk serum (zinc serum, page 207) and the same amount of the control serum into 250-ml Erlenmeyer flasks in such a way that the liquid does not flow on the sides of the flasks. To the condensed milk serum add 10 ml of water, and to the control serum add 10 ml of the sucrose solution. Add to each exactly 5 ml of the iodine solution and exactly 6 ml of the 2 N mixed alkali solution; mix gently, and allow the flasks to stand for 10 minutes at from 18° to 20° C. Acidify with 1.6 ml of 5 N sulfuric acid, and remove the liberated iodine, first, with 20-percent sodium sulfite solution, and finally, after adding 6 drops of starch solution, with the 2-percent sulfite solution. This operation should be conducted as rapidly as possible and with the precision of a titration, although the quantities of sulfite solution need not be measured. When all of the free iodine is eliminated, immediately add 1 drop of methyl orange solution and neutralize with 2 N mixed alkali solution. The time elapsing between acidifying with 5 N sulfuric acid and neutralizing with the mixed alkali should not exceed 2 minutes to avoid the danger of inversion of the sucrose.

Procedure 2, treatment with Luff solution.-To the contents of each flask add 20 ml of Luff solution; cover with a watch glass and heat the contents to boiling on a plain wire gauze over a burner regulated so that the boiling takes place in 2 minutes; impinging of the flame as hot gases on the sides of the flask should be prevented by an asbestos sheet, with a central hole of suitable dimensions, placed in contact with the wire gauze. When boiling takes place, transfer the flask to an asbestos-covered gauze already heated by a small Bunsen flame, attach a reflux condenser, and maintain gentle ebullition for exactly 10 minutes. Remove from the flame and cool in running water for 4 or 5 minutes.

Titration of reduced copper.-Add exactly 25 ml of the iodate-iodide solution and 20 ml of saturated potassium oxalate solution. Acidify carefully, while swirling, with 20 ml of 5 N sulfuric acid. Carefully shake with a rotating motion until the precipitate of cuprous oxide (which is partly converted into white cuprous iodide) has dissolved, and titrate with 0.05 N thiosulfate. No further addition of starch should be required. The end point is distinguished by a sharp change to a fine light blue (the color of the cupric salt).

Calculation of levulose.-The difference between the titrations of the sample serum and the control serum, as milliliters of 0.05 N thiosulfate solution, multiplied by 0.064, gives the percentage of levulose in the sample, uncorrected for the volume of the clarification precipitate. This factor is strictly correct only for a 20-percent serum, that is, if exactly 40 g of condensed milk is diluted to 200 ml in the preparation of the serum.

(c) SICHERT AND BLEYER MODIFICATION OF BARFOED COPPER ACETATE METHOD FOR MONOSES

Barfoed [65] found that copper acetate in the presence of acetic acid was reduced by monosaccharides but not reduced to any great

extent by the disaccharides, maltose and lactose. Steinhoff [66] modified the Barfoed solution by substituting sodium acetate for the acetic acid. This strongly buffered solution had a pH of 6.4, which remained unchanged during the reduction, whereas the Barfoed solution continually lost acetic acid.

Sichert and Bleyer [67], using Steinhoff solution, formulated a detailed method of analysis which, however, required the use of a specially designed filter. This unnecessary complication has been avoided in the procedure of the Corn Products Refining Co., which, through the courtesy of that company, has been brought to our attention.

Reagents. (1) Copper solution, 69.28 g of CuSO4.5H2O per liter. (2) Sodium acetate, 500 g of NaC2H3O2.3H2O per liter.

(3) Ferric sulfate, 120 g of Fe2(SO4)3. (NH4)2SÒ4.24H2O (or 50 g of Fe2(SO4)3) and 100 ml of concentrated H2SO, per liter.

(4) Potassium permanganate, 0.1 N, standardized against pure sodium oxalate.

Determination.-Into a wide-necked 250-ml Erlenmeyer flask introduce 10 ml of copper solution, 20 ml of sodium acetate solution, and 20 ml of a sugar solution containing less than 100 mg of dextrose. The flask, equipped with a Bunsen valve, is placed into a calmly boiling water bath for exactly 20 minutes. The flask should be out of line with steam bubbles and should be immersed up to the neck.

Filter the precipitated cuprous oxide through washed asbestos in a Gooch crucible, and wash the flask three times with hot distilled water. It is not necessary to transfer the precipitate quantitatively. Transfer the crucible to a 150-ml beaker, which bears a mark at 60 ml. Wash the precipitation flask with exactly 20 ml of ferric sulfate solution divided in three portions. Add all washings to the beaker. Care must be taken to dissolve all of the red precipitate. Wash the flask with hot water. Remove and wash the crucible and add water to the 60-ml mark. Bring the solution to boiling on a hot plate, allow to stand 3 minutes, and titrate with permanganate to a pinkgray color, which persists for about 20 seconds. The addition of 1 ml of sirupy phosphoric acid at a late stage of the titration facilitates reading the end point. Refer the volume of permanganate to table 102, p. 607.

(d) MONIER-WILLIAMS MODIFIED BARFOED METHOD FOR MONOSE SUGARS IN CONDENSED MILK [68]

Copper reagent.-Dissolve 60 g of crystallized sodium acetate in water, add 105 ml of N acetic acid and make up to 1 liter. Transfer to a dry bottle, add 52 g (or more) of finely powdered crystallized copper acetate, and shake to saturation, and filter.

Ferric sulfate solution.-Dissolve 50 g of ferric sulfate in about 400 ml of water to which 109 ml of concentrated sulfuric acid has been added. Make to 1 liter and filter. Before use, this solution should be treated with 0.1 N permanganate until the color of the latter ceases to be discharged.

Clarification. Transfer to a 100-ml beaker an accurately weighed quantity, approximately 40 g of the well-mixed sample, add 50 ml of hot distilled water (80° to 90° C), mix, and transfer to a 200-ml measuring flask, washing in with successive quantities of water at 60° C until the total volume is 120 to 150 ml. Mix, cool to air tem

perature, and add 5 ml of dilute ammonia solution (1+9, approx. 1.43 N). Again mix, and allow to stand for 15 minutes. Add a sufficient quantity of dilute acetic acid (approx. 1.43 N) exactly to neutralize the ammonia added and again mix. Add with gentle mixing 12.5 ml of zinc-acetate solution (21.9 g of Zn (C2H3O2)2.2H2O+3 ml of glacial acetic acid per 100 ml) and subsequently 12.5 ml of potassium-ferrocyanide solution (10.6 g per 100 ml). Adjust the temperature to 20° C and fill to 200 ml. In all operations avoid the formation of air bubbles. Mix thoroughly, allow to stand for a few minutes, and filter, rejecting the first 25 ml of filtrate.

Determination.-Introduce 25 ml of milk serum (40 g of condensed milk in 200 ml, clarified with zinc ferrocyanide) into a thin-walled boiling tube (8 by 1/4 inches), add 70 ml of the copper solution, mix, cover with a watch glass, and immerse to the level of the liquid in the tube in a large water bath maintained at 80° C for 20 minutes. Remove, cool in running water, filter on asbestos with suction, wash the tube and filter rapidly a few times with freshly boiled water. Dissolve the cuprous oxide in the tube and filter in 20 ml of the ferric sulfate. solution. Wash the asbestos pad with cold water and add the washings to the ferric sulfate filtrate. Titrate with 0.1 N permanganate to faint permanent pink.

Table 27 is applicable to 25 ml of the milk serum containing 5 g of condensed milk.

TABLE 27.-Determination of monose sugar in condensed milk

(e) IODOMETRIC DETERMINATION OF ALDOSES

In the presence of iodine in alkaline solution, aldose sugars under suitable conditions are oxidized quantitatively to their respective aldonic acids. The postulated reactions for this oxidation are

I2+2NaOH NaIO+NaI+H2O

RCHO+NaIO+NaOH–RCOONa+NaI+H,O.

(49)

(50)

Equations 49 and 50 may be combined to represent the reaction,

which is usually written.

RCHO+I+3NaOH-RCOONa+2NaI+2H2O.

(51)

It was early recognized that the sodium hypoiodite may also react to form sodium iodate:

3NaIO

NaIO3+2NaI.

(52)

Since sodium iodate cannot oxidize the sugar in alkaline solution, some active iodine is lost from the sugar oxidation reaction, but all the unreduced iodine, that is, all that remains in the form of iodate or iodite, is recovered when the solution is acidified and titrated with standard thiosulfate, thus

NaIO+NaI+H2SO1→I1⁄2+Na2SO4+H2O

NaIO3+5NaI+3H2SO→→312+3Na2SO4+3H2O

I2+2Na2S2O3-2NaI+Na2S4O6.

The difference between the total iodine added and the excess iodine as found by the thiosulfate titration represents the amount of iodine. used up in the oxidation of the sugar according to eq 51.

Slator and Acree [69] showed that a check on the iodine value could be obtained by titrating with standard alkali the free acid left after completing the iodine titration.

HCI+NaOH-NaCl-H2O.

Sugar acid lactone+H2ORCOOH.

RCOOH+NaOH→RCOONa+H2O.

As shown by eq 51, for every two equivalents of iodine used, three equivalents of acid are produced. The final acidimetric titer should then be three halves as great as that of the iodine consumed.

Later,

In its earliest form, the method was first devised by Romijn [70], who used buffer salts, such as alkaline carbonates, bicarbonates, phosphates, and borax, to supply the alkalinity in order to avoid the overoxidation which he found occurred with caustic alkali. many investigators who had used buffer salts found that extended periods of time were required to complete the oxidation of the sugar in these weakly alkaline media and that consistent results were not always obtainable. In the most widely used methods, caustic alkali is employed.

The difficulties which must be overcome in standardizing the analytical procedure can be summarized briefly. If the whole required volumes of alkali and iodine are admitted to the sample simultaneously, much of the iodine is transformed to iodate by eq 52. This may leave a deficiency of iodine for the oxidation reaction. If the iodine is present in too great excess, overoxidation can occur. Kline and Acree [71] showed that while the ratio of alkali to iodine consumed by the aldehyde group was 3:2, the ratio in the oxidation of sucrose or levulose proved to be 5:4, indicating that the primary alcohol group was slowly oxidized to a carboxyl group. Conceivably the overoxidation of an aldose proceeds similarly. The rate of addition of the reagents has an important bearing on the results. In the analysis of a mixture of aldose and ketose, such as dextrose and levulose, it is important that the levulose remain unoxidized. If too great an excess of alkali is present the levulose can undergo the Lobry de Bruyn-van Ekenstein rearrangement to dextrose and man

nose, which are rapidly oxidized by iodine. The time of reaction has undergone endless variation in the specifications, from 24 hours, according to Romijn, to 2 minutes, according to Kline and Acree. Observers agree, however, that in the presence of sodium hydroxide the reaction is rapid, and that short periods of time are sufficient. Space permits the inclusion of only four of the many procedures that have been described.

Method of Willstäter and Schudel [72].-To the aldose solution containing from 1 to 100 mg of sugar, add about twice (1.5 to 4 times) the volume of 0.1 N iodine required for oxidation. At room temperature and with effective stirring, add dropwise 1.5 as much 0.1 N NaOH as iodine (from alcohol-free sodium hydroxide) and allow the mixture to stand at room temperature for 12 to 15 minutes (for small amounts of sugar, 20 minutes). Add dilute sulfuric acid to slight acidity and titrate the excess of iodine with 0.1 N thiosulfate, using starch indicator. Deduct the excess iodine from the volume originally added. The amount of reducing sugar is calculated from the difference where 1 ml of 0.1 N I2-9.00 mg of hexose or 7.50 mg of pentose.

Kline and Acree [71] have criticized the procedure of Willstäter and Schudel on the grounds that the great excess of iodine in general tends to cause overoxidation. Occasional low results are probably caused by the formation of iodate, which removes iodine from the reacting system. Apparently, as Goebel [73] showed, the rate of addition of alkali requires delicate adjustment. If added over a period of 2 to 4 minutes, correct results were obtained. If added more rapidly, the recovery was low.

Method of Kline and Acree [71].-Kline and Acree have devised a procedure by which any considerable excess of either alkali or iodine is avoided, thus eliminating the errors due to iodate formation and overoxidation.

Transfer to an Erlenmeyer flask a weighed sample or aliquot containing approximately 180 mg of aldohexose or 150 mg of aldopentose and neutralize exactly (phenolphthalein, 1 drop only). Add 5 ml of 0.1 N iodine from a burette; then add drop by drop from a burette 7.5 ml of 0.1 N NaOH. Repeat this process until 22 ml of iodine and 35 ml of alkali solution have been run in, the operation requiring about 5 to 6 minutes. Allow a 2-minute interval for the completion of the oxidation. Acidify with 0.1 N (or 0.2 N) HCl and titrate the liberated iodine with 0.1 N thiosulfate (starch indicator). Add 2 to 3 drops of phenolphthalein and titrate the excess acid with 0.1 N NaOH. Deduct the thiosulfate titer from the number of milliliters of iodine added and deduct the hydrochloric acid titer from the number of milliliters of alkali added. The results will be the number of milliliters of iodine and alkali, respectively, consumed by the oxidation reaction. Then 1 ml of iodine equals 9.0 mg of hexose or 7.5 mg of pentose, and 1 ml of alkali equals 6.0 mg of hexose or 5.0 mg of pentose.

NOTES. The ratio of volumes of alkali to iodine should be 3:2 if the reaction is confined to the aldehyde group alone. A departure from this ratio shows that other groups or other substances, such as lignins, are being oxidized. If assurance is had that only sugars are oxidized, and if it is desired to diminish the labor of analysis, either one of the two titrations may be omitted.