December, 1944
Table
IV. Determination of Starch b the Proposed M e t h o d in the Presence of AddeJSubstances Starch Found, Air-Dry Basis Control plus Control substance
substance Added4 Raffinose (commercial) Sucroae (romrnerci~l) Gum arahic (ronimerrial) E g g albuniin (commercial) Inulin (ci,ninierrial) Gluten (laborat.orv puri6ed) 1-Cystine (comme;ciaI) I-proline (commercial) White potato dextrin (commercial) Levulose (com inertial) Maltose (comniercial) Malic acid (comnierrial) Sweet potato pectin (laboratory preparation)
%
%
8b 9
86 0 85.3
85 9 86 5 86.6 86 3 86.4 86.1 86.1 86.1 86 6 86.6
86.5 86 6 86.6 86.6 86.6
86.5 87.3 86.7 86.8 86.8 86.2
86.4
86.0
200 mg. of each substance added t o 1-gram samples of sweet potato starch B. a
Table
V.
Recovery of A d d e d Starch’ A B C
From 1.0 gram of cottonseed meal Starch added Starch recovered From 1.0 gram of orange rind Starch added Starch recovered 0
739
ANALYTICAL EDITION
Me.
Ma.
Me.
D Mg.
28.1 29.4
68.7 69.6
112.1 112.9
156.5 155.8
24.9 26.0
69.2 70.6
115.5 118.1
159.0 158.8
Calculated as empirically pure starch.
Bnal solution is essentially correct. As an illustration of the reproducibility of the method, duplicate determinations on sample 12 gave 59.3 and 59,3y0 starch on the dry basis, while duplicate determinations made 5 months later gave 59.8 and 59.8%. An analyst, who had not used the method previously, obtained 59.3 and 59.5% on the same sample. LITERATURE CITED
(1) Assoc. Official Agr. Chem., “Official and Tentative Methods of Analysis”, 5th ed., pp. 125, 359 (1940). (2) Brown, C. A., and Zerban, F. W., “Physical and Chemical Methods of Sugar Analysis”, p. 861, New York. John Wiley & Sons, 1941. (3) Clendenning, K. A., Can. J. Research, C20, 403-10 (1942). (4) Denny, F. E., Contrib. Boyce Thompson Inst., 6, 139-47 (1934). (5) Ihid., 6, 381-95 (1934). (6) Earle, F. R., and Milner, R. T., in manuscript. (7) Hopkins, C. Y., Can. J . Research, 11, 751-8 (1934). (8) Mannich, C., and Lens, K., 2. Untwsuch. Nahr. u. Genussm., 40, 1-11 (1920). (9) Pucher, G . W., and Vickery, H. B., IND.ENO.CKEM:., ANAL. ED., 8, 92-7 (1936). (10) Sullivan, J. T., J. A ~ E O Oi?iciaZ C. Agr. Chem., 18, 621-36 (1935).
Colorimetric A s s a y of Quaternary Ammonium Salts
The only substance listed that showed evidence of interference was white potato dextrin. This material gave a red color with iodine and interfered slightly. Higher polymer dextrins, which give a violet or violet-blue color with iodine and which may be found in slightly degraded starch, do not precipitate completely when the starch is precipitated with iodine in the method and if present may be readily detected in the supernatant liquid. They are partially carried down with the starch iodide and consequently behave in an anomalous manner in the method. Consequently the results on samples containing higher dextrins are of doubtful validity. Glycogen acts in much the same way. When it 1s analyzed alone by the method, it is almost completely removed, but in the presence of starch it is carried down with the starch iodide and included as starch. RECOVERY OF A D D E D STARCH
Sweet potato starch was added in various quantities t o 1.0gram samples of cottonseed meal and orange rind and the mixture analyzed by the proposed method. Controls without starch were also run and the polarimetric readings subtracted from the polarimetric readings of the samples to which starch had been added. The results given in Table V show that the method works well in the presenve of large amounts of pectin and protein. PRECISION AND A C C U R A C Y
Duplicates on I-gram samples of relatively pure starches usually agree within 0.4%, or 4 mg. of starch. The agreement is dependent to a great. extent on the precision obtainable from a polarimeter. In most instruments this precision is +0.01 mgular degree. The deviation between duplicates, then, due to the instrument alone, for a 1-gram sample :n a 4-dm. tube may be expected to be as great a s 0.25%, even if the highest precision of the instrument is attained. This variation becomes increasingly important as the starch content decreases. Based on experiments with 1-gram samples containing large amounts of pectin and protein, i t may be concluded that the method gives values that are within 10 mg. of the true starch content even on samples containing large quantities of the substances which interfere in most starch methods. The la.ter statement assumes that the factor 200.9 for the specific rotation of starch in the
M. E. AUERBACH Research Laboratories, Winthrop Chemical Co., Inc., Rensselaer, N. Y.
R
ECENTLY a method was described for the determination of germicidal quaternary ammonium salts in dilute solution (I). Since publication of this paper, comments from various correspondents have made it clear that some workers find it difficult to clarify-i.e., dry-the ethylene dichloride dye solution without affecting its color intensitv Re-examination of this point led to the decision to use benzent instead of ethylene dichloride. Benzene is not so good a solvent for the colored salt, but, being lighter than water, it can conveniently be clarified by centrifugation, thus avoiding all danger of contamination. At the same time, to compensate for the loss of solvent power, two other changes were made: (1) The amount of sample was reduced to one fourth-50 micrograms. (2) The light filter was changed from one transmitting a t about 540 mp to one transmitting a t about 600 mp. The method, as now used, is as follows:
In a 125-ml. Squibb separatory funnel, take 50 ml. of water containing 50 to 75 micrograms of the quaternary compound. Ordinary stopcock grease should be avoided. Starch-glycerol lubricant is satisfactory ( 2 ) . Add 5 ml. of 10% sodium carbonate solution, 1 ml. of aqueous 0.04% bromophenol blue indicator solution, and exactly 10 ml. of benzene. (The indicator solution should be prepared on the day it is to be used. Dissolve 40 mg. of bromophenol blue powder in 100 ml. of water containin 1 ml. of 0.1 N sodium hydroxide.) Shake steadily for 2.5 to minures, let the layers separate roughly (20 to 30 seconds), and then swirl the funnel contents. Let stand several minutes or until well separated. Rinse a 15-ml. centrifuge tube with a portion of the lower aqueous layer, discard this layer entirely, and then run the colored benzene layer into the tube. Stopper the tube with a clean rubber diaphragm stspper and centrifuge f0r.a few minutes a t about 1000 r.p.m., if necessary to clarify. Transfer to a dry Klett-Summerson colorimeter tube, and read, using filter No. 60.
8
Changes in technique involve no change in rationale of method. Limit of error is about *2%, with occasional errors 15%. LITERATURE CITED
(1) Auerbach, M. E., IND. ENQ.CHEM.,ANAL.ED., 15, 492 (1943). (2) Herrington and Starr, Ibid., 14, 62 (1942).
