INDUSTRIAL A N D ENGINEERING CHEMISTRY
744
prepared by the alkali process, when bleached, shows an increase in the lignin determination, although it is quite evident that the bleaching pl'ocess cannot manufacture lignin from some other constituent of wood pulp.6 Another instance of a mistaken conclusion which might be made from comparative determinations of cellulose is found in the pulping of wood with neutral sodium sulfite.' On cooking aspen wood with sodium sulfite solution a t low temperatures] the amount of cellulose based on the weight of the original wood is actually increased, if one is to credit the evidence afforded by the chlorination method of cellulose determination. Summary 1-Although the principal effect of decay on wood, so far as removal of constituents is concerned, is apparently the 8
Hawley and Wise, op. cit., p. 259.
' Rawling and Staidl, Paper Trade J . , 81, No. 8, 49 (1923).
Vol. 19, No. 6
same as a hydrolysis, yet these two processes differ in their effect on the alkali-solubility of the residue. 2-The total alkali-solubility of partly decayed wood is much greater than that of wood which has been hydrolyzed to the same extent (as shown by equal loss in weight). 3-The solubility of the lignin in alkali is very slightly increased by partial hydrolysis, whereas decay may render the lignin 50 per cent soluble. 4-The solubility of the residual cellulose is considerably increased by partial hydrolysis, but probably not to the same extent as it is by decay. &There is in partially hydrolyzed wood a material,.probably a degradation product of the cellulose, which is not determined as cellulose, lignin, or pentosans. This material is soluble in 1 per cent sodium hydroxide.
Manganese Interference in the o-Tolidine Test for Available Chlorine' By Edward S. Hopkins DEPARTMEST O F PUBLIC WORKS,BUREAUOF WATERSUPPLY,BALTIMORE, MD
I
T IS an established fact that manganese salts produce the characteristic yellow chlorine color with the o-tolidine reagent. Buswell and Boruffl,* state that salts of this metal in as concentrated a solution as 11 p. p. m. manganese do not produce color. Solutions of a concentration of 100 p. p. m. manganese will not produce color provided the salt used is a stable one, such as sulfate or chloride] as any easily reduced salt of this and other metals will produce color with this reagent. As a matter of record a concentration of 0.01 p. p. m. manganese as potassium permanganate will give a color intensity corresponding to 0.03 p. p. m. chlorine and this color will increase in direct proportion to the concentration of the salt. Two mols of potassium permanganate in alkaline solution produce 3 mols of available oxygen and, since chlorine is a univalent element and oxygen a bivalent one, 1 mol of potassium permanganate will equal 1.5 mols of oxygen or 3 mols of chlorine. Plotting the intensity of color produced, mol for mol in terms of chlorine readings] as in Figure 1, it is seen that this ratio is true and that the color is in direct proportion to the loosely bound oxygen. Such a condition would be expected from the studies of Ellms and Hauser.2 This would indicate by analogy that the chlorine reaction is a similar one and is not a question of a color produced under definite pH conditions. If a non-color-producing solution of manganese] sulfate, for example, is converted to the hydroxide by addition of a base, upon acidifying and adding the o-tolidine reagent 0.07 p. p. m. manganese will produce color. This phenomenon is true for all stable salts of manganese by actual experiment. Experimental
Remembering the reaction involved in the Winkler3 method for the determination of dissolved oxygen, it seemed quite Presented before the Division of Water, 1 Received January 28, 1927. Sewage, and Sanitation at the 73rd Meeting of the American Chemical Society, Richmond, Va , April 11 to 16, 1927. Numbers in text refer to bibliography at end of article
*
likely that the absorption of oxygen by the manganese hydroxide was to be expected and that the reduction of these loosely bound oxides to manganous hydroxide or a stable compound by the o-tolidine produced the color. To prove this theory, manganous hydroxide was prepared with air excluded and the usual white precipitate was produced. By carefully keeping this precipitate under anaerobic conditions it was not possible to produce color with the o-tolidine reagent. Oxygen was absorbed very rapidly, the slightest addition of air changing the white hydroxide t o a yellow hydroxide with subsequent production of color with the reagent. Oxidation of this white precipitate with slight amounts of hydrogen peroxide gave color increase in proportion to the amount of oxygen present. Excessive amounts of oxygen produced a heavy black almost insoluble precipitate of hydrated manganese oxides containing excess oxygen. Upon addition of the reagent t o these precipitates an intense yellow color was p r o d u c e d . In other words, the more oxygen absorbed the greater yellow color 5 os p r o d u c e d b y t h e 0- S tolidine. Owing to the rapid oxidation of the mang a n e s e hydroxide by air, it was not possible t o weigh a c c u r a t e amounts and p r e s e n t 0 10 quantitative data. A PPM KMNO~ second experiment was Figure 1--o-Tolidine Color (Chlorine p e r f o r m e d using the Scale) Produced b y Mol Equivalents of Oxygen f r o m Alkaline Potassium Persame amounts of re- manganate agent as above, but the hydrogen peroxide and manganous hydroxide were allowed t o stand together for 10 minutes. A greater color was produced and a heavier precipitate obtained after this time of contact than in the previous experiment. 0-
June, 1927
INDUSTRIAL -4-ND E,VGINEERING CHEMISTRY
Manganese dioxide suspended in water will produce color with o-tolidine, which color can be also obtained in the clear filtrate from such an acidified suspension. Oxidation of this compound to MnOOH by chlorine gave ztn increase in color greater than that obtained from the salt originally. It seemed desirable to learn if manganic hydroxide would produce color. This salt was carefully prepared as described in a standard textbook4 and tested with the o-tolidine reagent, color being produced. On dissolving this hydroxide in sulfuric acid (1 to 4) and converting to manganous hydroxide by addition of sodium hydroxide, a greater intensity of color was produced than from the original manganic hydioxide owing to the absorption of oxygen from the air. This condition is shown in Figure 2 . These curves are somewhat inaccurate since varying drops of the precipitate were placed in the o-tolidine solution and no exact n-eieht could be obt a i n e d . Considering this condition and rem e m b e r i n g that the curves in Figure 2 are plotted on a logar i t h m i c s c a l e , they agree very well with the one obtained from the c a r e f u l l y controlled potassium permanganate solution shown in Figure 1. It was noted that if hydrogen peroxide was added in excess to either .I I I ! 3I I the manganous or manDROPS OF OXIDE g a n i c hydroxide preFigure --Tolidine Color (Chlorine Scale) Produced by Absorbed Oxygen in cipitates, t h e s e comManganous Hydroxide pounds were promptly reduced by the process of double oxidation to a stable compound and no color could be obtained with o-tolidine. To conform with data obtained in relation to chlorine color' the pH's of all solutions were kept between 3.0 and 8.0, this test being made before the addition of the acid o-tolidine reagent. It is of interest to note that the maximum intensity of color, measured quantitatively, developed by the manganese oxides was produced within 10 minutes and faded after 20 minutes. This corresponds with results obtained from chlorine tests. To be certain that the color intensity was not affected by any possible production of chlorine from use of hydrochloric acid as a solvent for the o-tolidine, the solution was made with an equivalent amount of sulfuric acid since comparative tests produce similar color for the same amount of manganese compounds. It is fairly evident that hydrated oxides of manganese containing more oxygen than manganous hydroxide are easily reduced and that the oxygen is loosely bound.5 What form of manganese containing more oxygen than manganous hydroxide is necessary to produce this color? Since chemical compounds combine in definite ratio, the first consideration was to vary the sodium hydroxide equivalents for forming the hydrated manganese oxides in order to learn if a complex oxide was formed. Two mols of sodium hydroxide combine with l mol of manganese to form 1 mol of manganous hydroxide, which upon standing 10 minutes gave color reading of 3.3 p. p. m. on the chlorine scale. A concentration of 50 mols sodium hydroxide to 1 mol l f n did not increase the color obtained, which shows that it is produced by oxygen absorption in the manganous hydroxide and is not a complex compound.
-
746
Discussion
T-arious formulas have been assigned to the hydrated manganous hydroxide precipitate when used as an oxygen absorbent. Studies of the Winkler method for the determination of dissolved oxygen by Theriault6 present three separate formulas. Other te~tbooks-Smith,~Scott,7 Dennis and Whittelseys-assign different combinations. They all agree that the formulas indicate easily displaced oxygen atoms. This is confirmed by the experimental data obtained, and the reduction of such an oxide by the o-tolidine produces color. One characteristic was noted. Upon acidifying the manganous hydroxide only those solutions giving a slight pink color, brown with some acids, or those containing definite brown or black particles, produced color with o-tolidine. When these particles were present they were insoluble in dilute sulfuric acid. On testing these colored solutions in a diffusion cell, it was found that they were colloidal solutions. That such a precipitate could be expected is shown by Bancroftlo and dnargyros.ll Since it is known that the higher oxides of manganese are insoluble in sulfuric acid and that it is possible for manganous
t-
I
I I
0-TOLIPINE COLOQ
Figure 3-Plant
"!
Data--1925
hydroxide to "rapidly oxidize to a hydrated Mn304and then be slowly converted to a colloidal hydrated Mn203,''9 the theory may be advanced that the form of hydrated manganous hydroxide is the usual one present in natural wacers. Knowledge of these hydrated manganous oxides is essential to water purification', for Baylis12 has shown the conditions under which manganese may occur in a water supply. Another factor would be the presence of manganous carbonate, which is soluble in free carbon dioxide. This was determined by test in the laboratory and confirmed by Robinson, Gardner, and Holmes.13 This soluble manganese would be converted to the hydrated oxide by the action of air or dissolved oxygen in the water and produce color with o-tolidine, giving a false chlorine reading in the raw water. Such a condition is always present in this plant, the color ranging from a minimum of 0.02 to more than 0.3 p. p. m. Figure 3, showing the total manganese content of the raw
I-VD USTRIAL A Y D ENGINEERING CHEMISTRY
746
water and the color resulting from same, is plotted from plant data for 1925. Since there is no relation between the trend of these curves it shows that the manganese is present as a combination of stable and unstable salts. Even after filtration soluble manganese will be present from such a supply. Upon addition of alkali to reduce pipe corrosion by water containing free carbon dioxide, this manganese is promptly converted to the hydroxide with production of color by o-tolidine. This condition will give higher readings for chlorine than actually exists and therefore indicate security from a chlorine residual test, when as a matter of fact the actual amount of free chlorine present may be negligible. Such a condition may easily exist in a small purification plant, with meager or no laboratory facilities, where greater reliance is placed upon the residual chlorine test than in the
Vol. 19, No. 6
larger ones under constant bacteriological control. To call attention to these possibilities is the object of this paper. Bibliography I-Buswell and Boruff, J . A m . W a t e v W o r k s Assoc., 14, 384 (1925). 2-Ellms and Hauser, J . Ind. Eng. Chem., 6, 553 (1914). 3-American Pub. Health Assoc., Standard Methods of Water Analysis, p. 59 (1925). 4-Smith, “General Chemistry for Colleges,” p. 620 (1920). 5-Taylor, J . Phys. Chem., SO, 145 (1926). 6--Theriault, U.5’. Pub. Health Bull. 161 (1925). 7-Scott, “Standard Methods of Chemical Analysis,’’ p. 302 (1918). 8-Dennis and Whittelsey, “Qualitative Analysis,” p. 58 (1902). g-wilborn, Farben-Zlg., 31, 338 (1926). l@-Bancroft, “Applied Colloidal Chemistry,” p. 175 (1921). 11-Anargyros, Comfit. rend., 161, 419 (1925). 12--Baylis, J . A m . W a f e r Works Assoc., 12, 211 (1924). 13--Robinson, Gardner, and Holmes, Science, 60, 423 (1919).
Adsorption of Vapors b y Ferric Hydroxide Gel’ By J . H. Perry 1211 DELAWARE AVE., WILMINGTON, DEL.
The eficiency
of the adsorption cjf fourteen
capors by ferric hydroxide gel has been studied by a dynamic method. preliminary data indicate that ferric hydroxide gel can be used f o r the recovery of most vapors ac. ejectiueh as alumina and silica gels.
T
HERE are in the literature considerable data on the
adsorption of vapors by the gels of silicic acid and alumina. There appear to be no such data, however, for a pure ferric hydroxide gel, although there are a number of papers on the adsorption of ions and dyes by this gel. The similarity of silica, alumina, and ferric hydroxide gels led to the belief that the last would have approximately the same efficiency of adsorption from vapor-air mixtures and approximately the same saturation capacities for different vapors as the first two gels. This paper describes results obtained in a general investigation of the adsorption of fourteen vapors by ferric hydroxide gel. General Procedure
PREPARATION OF GEL-C. P. ammonium hydroxide was added to a n aqueous solution of C. P. ferric nitrate. The resulting ferric hydroxide was then washed with hot distilled water until the washings showed no trace of nitrates. 1
Received January 20, 1927.
These
The washed precipitate was dried, first a t 60” C., at-which temperature most of the shrinkage took place; then a t 100” C. until a hard, glass-like material was obtained. This was broken up and screened. The material used for this series of experiments was 10-12 mesh. ACTIVATION-unless otherwise stated the gel was activated by passing through it dry carbon dioxide-free air heated to 230’ C., the gel being heated in the same bath as the air. During the cooling from 230” to 25” C. no air was allowed to come in contact with the gel until the vaporair mixture was started through it. This activation was repeated after each experiment, until the gel returned to its original weight, before being used in the adsorption experiments with another vapor. The duration of the activation varied but little after each experiment and was about 2 hours. ADsoRPTIoN-The general method of the experiments was the same as that described in a previous paper2 and consisted in passing a vapor-air mixture at a definite rate (50 cc. 2
Perry, J . Phys. Chem., 29, 1462 (1925).
