is made up to 200 ml with double-distilled water, the flask is stoppered, and the solution mixed.

For storage the standard buffer solutions are transferred to resistant glass bottles labeled with the corresponding pH. The bottles containing the most frequently used standards may be provided, for convenience, with clean one-hole rubber stoppers, each carrying a 10-ml pipette.

Indicator solutions. Indicators recommended for the colorimetric determination of hydrogen-ion determination are listed in table 36 under the common names of the dyes. The list includes the indicators originally selected by Clark and Lubs [32], while those marked with an asterisk have been added as a result of the work of Cohen [33]. All are sulfonphthaleins with the exception of methyl red, an azo dye, and cresol phthalein. Methyl red, although replaced in the revised indicator list by chlor phenol red, is included in table 36 because it is prescribed in the method of Gillespie, as described later. Opposite the name of the indicator are given the molecular weight, the number of milliliters of 0.01 N sodium hydroxide needed to form the monobasic salt of 0.1 g of dye, pK (=log 1/K, where K is the dissociation constant of the indicator), useful pH range, and the colors of the acid and alkaline forms of the dye.

The Determination of Hydrogen Ions, W. M. Clark, 3d ed. p. 94 (Williams & Wilkins Co., Baltimore, Md., 1928). *Added as result of later work by Cohen [33].

Stock solutions of the indicators are prepared by grinding 0.1 g of the dry powder in an agate mortar with the corresponding volume of 0.01 N sodium hydroxide given in column 3 of table 36. The alkali is added little by little, and when solution is complete it is diluted to 250 ml with distilled water to give a 0.04-percent solution. Methyl red (0.1 g) is dissolved in 125 ml. of neutral alcohol and diluted with water to 250 ml.

Performing the tests.-Test tubes 15 mm in diameter and 120 mm long are calibrated roughly for uniformity of bore by adding 10 ml of water from a pipette and choosing those in which the water stands

at approximately the same level. These are cleaned and afterward rinsed with distilled water, and drained. To separate tubes are added 10 ml. of each standard buffer mixture to cover a chosen pH range, and 5 drops of the appropriate 0.04-percent indicator solution. Meade [34] recommends that the corks used for closing the tubes be first boiled and thoroughly rinsed before inserting.

To another calibrated test tube is added 10 ml. of the solution to be tested and 5 drops of the indicator. The matching of colors is best done with the aid of a comparator, a simple form of which is the block comparator described by Clark as having six deep holes of such size as to accommodate neatly the test tubes, drilled parallel in pairs, each pair being as close together as possible without breaking the wood. Through the center line of each pair and at right angles to them are drilled three smaller holes completely through the block so that light may pass horizontally through the pairs of tubes when introduced. The block is then painted a dull black inside and out. The back of the illuminating holes is covered with a ground plate of colorless glass to diffuse the light. The sample tube with indicator is placed in one of the middle holes and backed by a tube of water. Standard tubes containing the indicator are placed in the holes on either side of the sample, and each is backed by a tube containing the sample without the indicator, thus compensating for color and turbidity in the sample. When the sample appears to fall midway between two standards that differ by 0.2 pH unit, let us say, for example, between pH 7.2 and pH 7.4, then the value 7.3 may be taken as the pH of the sample.

There are other forms of comparators arranged for quick and convenient exchange of color standards, such as that of Meade and Baus [35], which has a sliding rack carrying alternate tubes of color standard and water, and of such length as to cover the entire pH range of the indicator. A metal cover with suitably spaced apertures, and having sockets for holding tubes of the sample and the sample plus indicator, is provided. A white surface at 45° is placed behind the apertures and illuminated from above with a Daylite Mazda bulb.

Care must be exercised in maintaining like volume of solutions and of indicators and like thicknesses of light-transmitting layer. Alkaline and neutral buffer solutions should be protected from absorbing carbon dioxide and should not be allowed to come in contact with the hands.

(b) WITHOUT BUFFER SOLUTIONS

GILLESPIE METHOD [36]

This method for the determination of hydrogen-ion concentration without the use of buffer mixtures depends upon the ratio of the number of drops of a given indicator present in a solution wholly transformed into the alkaline form to the number existing wholly in the acid form.

The indicators used are those of the original selection of Clark and Lubs [32]. The stock solutions of the dyes are prepared, except in the case of methyl red, by grinding 0.1-g portions with 1:1 equivalents of alkali solution (1.5 equivalents in the case of cresol red) to give the monosodium salt of the indicator acid. The equivalents of sodium hydroxide may be obtained from table 36, column 3. When the dye is in solution the volume is made up to 250 ml with water. Methyl red is dissolved as explained under (a), p. 296.

In determining the data upon which table 37 is based, Gillespie

made use of the Clark and Lubs buffer mixtures, as given in table 35, which were checked electrometrically. The data were then smoothed by use of the equation pH-k log (alkaline form)/(acid form), where pH is the hydrogen-ion exponent, k is the apparent or total dissociation constant of the dye, and "alkaline form" and "acid form" designate, respectively, the concentrations of the indicator completely transformed into the alkaline or acid form by excess of base or acid. In table 37 the drop ratios are given in the first column, the first figure of the ratio being the alkaline form and the second figure the acid form of the dye. The pH values corresponding to these ratios are found in the succeeding columns headed by the indicator used.

TABLE 37.-pH values corresponding to various drop ratios 1

The following is the Gillespie procedure for the preparation and use of the color standards: Test tubes, preferably without flanged tops, 15-mm bore and 150-mm length, selected for uniformity of bore as already described under (a), are cleaned, rinsed, and drained. The tubes may be held in pairs by means of a rubber band wound around them in the form of a figure 8. It is convenient to use test-tube racks, one for each indicator, each holding two rows of tubes, accommodating one tube of each pair in front and one in back. For any desired indicator a set of color standards is prepared by placing from 1 to 9 drops in the back row of tubes, and 9 to 1 drops in the front row. A drop of 0.05 N sodium hydroxide is then added to each of the tubes in the back row (2 drops in the case of thymol blue) to develop the alkaline color. A 0.2-percent solution of stick soda is sufficiently accurate for this solution. To each of the tubes in the front row is added the kind and quantity of acid indicated at the bottom of each column in table 37. According to Clark, the 0.05 N hydrochloric acid is prepared with sufficient accuracy by diluting 1 ml of concentrated hydrochloric acid (sp gr 1.19) to 290 ml. The volume is at once made up in all the tubes to a constant height, corresponding to 5.5 ml with distilled water. Each pair of tubes thus constitutes a colorimetric standard and is to be labeled with the corresponding pH value given in table 37. These dilute standards are not considered stable and daily renewal is recommended.

A third

A block comparator for matching solutions and standards, as shown in Gillespie's article, is illustrated in figure 60. This is a modification of the comparator described in section (a) in that there are three holes. in series instead of two to accommodate the extra tube. series of vertical holes with a third observation hole would permit comparison with two adjacent standards in the indicator range.

To perform a test, 10 drops of the desired indicator solution is added to a clean test tube, and the test solution is added in amount to match the height of the standard solutions in their tubes. The sample tube with the indicator is placed in the comparator and is backed by two tubes containing distilled water. Pairs of the indicator tubes are placed in two of the holes of the other series and are backed by a tube containing the sample without the indicator. For colorless test solutions, one compensating tube may be omitted from each series. The pH value of the pair of standards most nearly matching the sample containing the indicator is taken as the pH value of the sample. The same

precautions pointed out in the previous section are to be observed in regard to volume and concentration relations.

5. REFERENCES

[1] F. J. Bates and Associates, Intern. Sugar J. 22, 654 (1922) (Abstract). [2] J. F. Brewster and W. G. Raines, Louisiana Planter 69, 167 (1923); Intern. Sugar J. 25, 88 (1923).

[3] W. M. Clark, The Determination of Hydrogen Ions, 3d Ed. (The Williams & Wilkins Co., Baltimore, Md., 1928).

[4] J. Reilly and W. N. Rae, Physico-Chemical Methods, 2d Ed. (D. Van Nostrand Co., Inc., New York, N. Y., 1932).

[5] E. Biilmann, Bul. soc. chim. 41, 276 (1927).

[6] F. Haber and Z. Klemenziewicz, Z. physik. Chem. 67, 385 (1909).

[7] W. S. Hughes, J. Am. Chem. Soc. 44, 2860 (1922).

[8] P. M. T. Kerridge, Biochem. J. 19, 611 (1925).

[9] D. A. MacInnes and M. Dole, Ind. Eng. Chem., Anal. Ed. 1, 57 (1929).

[10] D. A. MacInnes and M. Dole, J. Am. Chem. Soc. 52, 29 (1930).

[11] G. R. Robertson, Ind. Eng. Chem., Anal. Ed. 3, 5 (1931).

[12] M. R. Thompson, J. Research NBS 9, 833 (1933) RP511.

[13] G. A. Perley, Trans. Am. Inst. Chem. Eng. 29, 257 (1933).

[14] D. A. MacInnes and L. G. Longsworth, Trans. Electrochem. Soc. 71, 73 (1937).

[15] K. Horowitz, Z. physik. Chem. 115, 424 (1925).

[16] L. Michaelis, Naturwissenschaften 14, 33 (1926).

[17] M. Dole, J. Am. Chem. Soc. 53, 4260 (1931); 54, 2120, 3095 (1932).
[18] D. A. MacInnes and D. Belcher, J. Am. Chem. Soc. 53, 3315 (1931).

[19] K. H. Goode, J. Am. Chem. Soc. 44, 26 (1922).

[20] L. W. Elder and W. H. Wright, Proc. Nat. Acad. Sci. 14, 936 (1928). [21] H. M. Partridge, J. Am. Chem. Soc. 51, 2 (1929).

[22] R. Müller, Z. physik. Chem. 155, 451 (1931).

[23] A. DuBridge, Phys. Rev. 37, 392 (1931).

[24] C. Morton, J. Sci. Inst. 9, 289 (1932).

[25] E. C. Gilbert and A. Cobb, Ind. Eng. Chem., Anal. Ed. 5, 69 (1933). [26] L. G. Longsworth and D. A. MacInnes, J. Bact. 29, 595 (1935); 31, 287 (1936); 32, 567 (1936).

[27] N. Crites, Rep. Assn. Hawaiian Sugar Tech. Chem. Eng. Sect. 14, 41 (1935). [28] R. T. Balch and H. S. Paine, Planter Sugar Mfr. 75, 347 (1925). [29] G. A. Perley, Ind. Eng. Chem., Anal. Ed. 11, 316 (1939). [30] G. A. Perley, Ind. Eng. Chem., Anal. Ed. 11, 319 (1939). [31] W. M. Clark and H. A. Lubs, J. Biol. Chem. 25, 497 (1916). [32] W. M. Clark and H. A. Lubs, J. Bact. 2, 1, 109, 191 (1917). [33] B. Cohen, U. S. Pub. Health Service Rep. 41, 3051 (1927).

[34] G. L. Spencer and G. P. Meade, Handbook for Cane Sugar Manufacturers and Their Chemists, 7th Ed. (John Wiley & Sons, Inc., New York, N. Y., 1929).

[35] G. P. Meade and R. Baus, Planter Sugar Mfr. 74, 509 (1925).

[36] L. J. Gillespie, J. Am. Chem. Soc. 42, 742 (1920); Soil Sci. 9, 115 (1920). [37] A. S. Brown, J. Am. Chem. Soc. 56, 646 (1934).

[38] H. S. Harned, J. Am. Chem. Soc. 51, 416 (1929).

[39] C. K. Rule and V. K. La Mer, J. Am. Chem. Soc. 58, 2339 (1938).

[40] E. R. Smith and J. K. Taylor, J. Research NBS 20, 837 (1938) RP1108. [41] H. S. Harned and R. W. Ehlers, J. Am. Chem. Soc. 55, 652 (1933). [42] L. R. Bacon, J. W. Hensley and T. H. Vaughn, Ind. Eng. Chem., 33, 723 (1941).

XIX COLORIMETRY

1. INTRODUCTION

The maintenance and improvement of the quality and appearance of sugar products leads naturally to the recognition of the importance of colorimetry and of the need for adequate apparatus and methods for the measurement of sugar color. The early literature on this subject dealt with methods used in connection with the refining of sugar with bone char; thus Ventzke [1] in 1860-61 published the description of a "decolorimeter," by means of which colors of solutions were compared before and after char treatment. In these articles Ventzke referred to the work done by Payen about 25 years previously. In 1861 Stammer [2] published a description of the instrument and method for sugar colorimetry that bears his name, and which, little changed, is still in use.

In 1873 von Vierordt studied the measurement of color in diluted molasses with a spectroscope, the entrance slit of which was divided into upper and lower halves by independent jaws actuated by micrometer screws. To each screw was attached a drum which bore a graduated scale reading from 0 to 100. The opening and closing of the slits thus served to vary the light intensity in a measurable manner in either half of the field of view. The description of this apparatus is contained in a reprint in the Bureau's possession. This was reprinted in 1873 by Schmidt & Haensch, Berlin, from an article by Vierordt, but the journal source is not cited. The table in the reprint gives log transmittancy corresponding to scale readings, and layer thickness and dilution were taken into consideration in the text, the latter on a volumetric basis. This approach to a spectrophotometric method apparently received little attention from sugar technologists until many years later.