INDUSTRIAL AND ENGINEERING CHEMISTRY
184
If they ale both present, this titration would represent the total alkalinity due to alkali metals.
ON KNOWNMIXTURES TABLEI. RESULTS
Weight of Sample Grams
Mixture Ammonium sulfate, calcium hydroxide and sodium oarboAate Fused sodium chloride Potassium nitrate
Sodium Carbonate In mixture Found
Sodium In mixture Found
%
%
%
0.176 0.240 0.366 0.146
35.2 37.1
35.2 37.2
'.
... .
15.3 16.1 39.4
0.406 0.230
,. ..
.. ..
%
15.3 16.2 39.1 40.0 Potassium 38.6 38.6 38.6
.. ..
The sulfated residue is dissolved as far as possible in water. Solid barium hydroxide is added in excess to precipitate sulfates as barium sulfate, and to convert sodium and/or potassium sulfate to the hydroxide. Na2S04 Ba(OH)* +2NaOH Bas04 J. The precipitate, which may include other undissolved alkalineearth sulfates, is filtered off and washed well with hot water, and ammonium carbonate is added to the boiling combined filtrate and washings to precipitate soluble alkaline-earth compounds as carbonates. The carbonates are filtered off, washed well with hot water, and the filtrate is boiled down to one-fifth of its volume to decompose and get rid of excess ammonium carbonate. (There should be no odor of ammonia at this point.) After cooling, any slight recipitate of alkaline-earth carbonates is filtered off, washed witl water at room temperature, and the combined filtrate and washings are titrated to a methyl orange end point with 0.1 N acid in the usual manner. The acid consumed is calculated to per cent sodium or potassium, as the case may be.
+
+
VOL. 7, NO. 3
Since the estimation is based on getting the alkali metals in the form of soluble compounds, it is not necessary to burn off all carbon or ignite the ash to constant weight. This reduces the loss by volatility of alkali metal compounds, which is a troublesome factor in the usual gravimetric determinations of the alkali metals. It is hoped that this method may be useful in estimating alkali metals in other fields of analytical procedure.
Results on Unknown Mixtures Neutrally bound alkali metals in the ash may be calculated from the difference between the result obtained by the method as outlined and the result obtained by determining the alkalinity of the ash to methyl orange. Results of tests on ash from soda-base grease, containing 0.93 per cent of total sulfur, showed the following. Alkalinity of ash as.sodium, uncorrected Total sodium by this method, per cent Neutrally hound aodium, per cent
30.8 34.6 3.8
30.9 34.8 3.8
Acknowledgment The author wishes to thank Harry Levin and Karl Uhrig for helpful criticisms and suggestions. RECEIVED February 4, 1935.
Detection of Benzene in Alcohol A. C. LANSING, Commercial Solvents Corporation, Terre Haute, Ind.
B
ENZENE may occur in alcohol in small amounts, and is undesirable in alcohol used for some purposes. The presence of benzene may be due to faulty process, or to accidental contamination. When the benzene content is too small for any of the turbidity tests with water, indications of its presence are usually obtained through taste and odor. Unfortunately, taste and odor tests do not give uniformly positive selective indications of the presence of benzene in low concentration. A direct and sensitive chemical test for small amounts of benzene in alcohol is therefore of value in the examination of alcohol. A reliable method was desired for detecting benzene in concentrations of the order of 1 part by volume in 10,000, in alcohol of good quality otherwise. After unsatisfactory results by other methods a procedure involving concentration of the benzene by extraction, nitration of the extract, and color reaction of the nitration product was devised and applied with satisfactory results.
Reagents Required Carbon tetrachloride, reagent grade. Acetone, reagent grade. Amyl alcohol (redistilled crude fusel oil) refluxed with strong caustic soda solution, and distilled from the alkali, collecting for use the dry fraction boiling at 128" to 132' C. Nitrating acid: 70 grams of reagent 20 per cent oleum, 45 grams of reagent sulfuric acid of specific gravity 1.84, and 43 grams of reagent nitric acid of specific gravity 1.42. Sodium sulfate solution, containing 10 grams of the anhydrous salt in 100 cc. of solution. Caustic soda solution, 140 grams of reagent (sticks) made up to 250 cc. at room temperature.
Procedure If the alcohol contains much above 0.01 per cent benzene by volume, as indicated by test on the undiluted sample, it is diluted
to approximately that value with alcohol free from benzene, A 40-cc. portion is placed in a 100-cc. glass-stoppered cylinder with 6 cc. of carbon tetrachloride. Distilled water is added to the 90 cc. mark, followed by 10 cc. of sodium sulfate solution. The cylinder is stoppered, shaken thoroughly, and allowed to stand until the layers separate sharply. Five cubic centimeters of the bottom layer are transferred by pipet to a test tube. Three cubic centimeters of nitrating acid are measured in a small cylinder and added to the test tube, which is shaken carefully but thoroughly. During 10 minutes the tube is shaken twice more, and at the end of that time, 20 cc. of distilled water are added, preferably rapidly from a cylinder. After mixing by pouring into another test tube and back again, the bulk of the water layer is decanted and discarded. The lower layer with a small amount of water layer is placed in an evaporating dish on an electric hot plate. When the carbon tetrachloride is gone, the dish is emptied into the same test tube just used and is rinsed into this tube with 1 cc. of amyl alcohol. The dish is then rinsed into the tube using 4 cc. of caustic soda solution. The tube is mixed by swirling and then 1 cc. of acetone is added, the tube is swirled again, and placed in a rack for observation of color produced in the top layer.
Indications Pure benzene a t the concentralion 1 in 10,000 by volume in a rectified alcohol of ordinary purity gives by this test a red color with a purple quality in the top layer. This color holds for some hours, and finally fades to a dull orange-red. As the benzene content rises above 0.01 per cent, the color produced soon becomes too dark for identification, and dilution with alcohol free from benzene is necessary. Pure toluene gives a slightly brownish yellow and reagent xylene (chiefly m-xylene) produces a definite though not intense green, which fades in 30 minutes, giving way to a dull orange. Blank runs should always be run in duplicate with a test series; they give a light clear lemon yellow. Thorough cleaning of the apparatus between tests is necessary t o maintain blank indications. Among other hydrocarbons tested, denaturing gasolines gave yellow colors of somewhat orange cast and a cer-
MAY 15, 1935
ANALYTICAL EDITION
tain klsrosene gave a reddish color, readily distinguished from the benzene color.
Precautions If suitable concentration methods and adequate precautions are used, concentrations of benzene well below 1 in 10,000 should be detected. Exclusion of chlorides from the extraction is necessary, as the subsequent nitration seems sensitive to their presence. A lower salt concentration than that used gave erratic results and a high salt concentration precipitated solids into the bottom layer. Exclusion of water from the nitration mixture is imperative. The pres-
185
ence of oxidizable material, such as excessive amounts of alcohol from the extraction or stopcock lubricant, interfere probably by their effect in the nitration. Care in preventing this effect is fully as important as efficient extraction of the benzene in the carbon tetrachloride.
Acknowledgment The writer wishes to express appreciation for the interest taken by A. Izsakin this work, and to the Commercial Solvents Corporation for permission to publish the results. RECEIVED January 10,1936.
Destruction of Organic Matter in Plant Material by the Use of Nitric and Perchloric Acids J. E. GIESEKING, H. J. SNIDER, and C. A. GETZ Department of Agronomy, University of Illinois, Urbana, Ill,
T
HE destruction of organic matter by some form of ashing generally precedes the determination of the mineral constituents in plant material. Two undesirable features of the ashing process suggest the need of a wet oxidation method similar to the Kjeldahl digestion: (I) The method of ashing commonly used for calcium and magnesium determinations is not applicable for potassium and phosphorus determinations on account of the loss of the more volatile compomds of potassium and phosphorus while ashing; and (2) the residues after ashing may be in a slowly soluble form. Perchloric acid either alone or in mixtures with other acids has been widely used in the oxidation of organic materials of animal origin previous to the determination of the mineral constituents. Kahane and associates (4) have made the most comprehensive study of these methods and have developed a method for determining silica in plant materials ( 5 ) . Winter and Bird (11) have used perchloric acid similarly for determining aluminum in plants. The properties of perchloric acid, as well as those of the perchlorate ion, make it a very desirable oxidizing agent for the analysis of organic substances. No water-insoluble perchlorates of the metals have been reported. Furthermore, cold perchloric acid, either dilute or concentrated, is not affected by ordinary reducing agents. The dehydrating action of perchloric acid on silica, shown by Willard and Cake ( l o ) ,aids in the quantitative separation of silica. Since hot concentrated perchloric acid may react violently with organic substances, the reaction intensity must be controlled. It has been found advisable to pretreat samples of plant material with nitric acid before adding perchloric acid. With substances very high in fat it may be necessary to pretreat the sample several times with nitric acid before it can be oxidized with perchloric acid without a loss of a portion of the sample. The perchloric acid should be diluted with water and nitric acid.
Experimental Procedure The following method of wet oxidation was applied to a wide variety of plant materials, including sweet clover (roots and tops), alfalfa hay, red clover hay, alsike clover hay,
timothy hay, redtop hay, wheat straw, cornstalks, corncobs, corn (grain), and soy beans: Place a 4-gram sample of the material to be oxidized in a 400ml. beaker and add 10 ml. of nitric acid (RP.gr. 1.42). Cover the beaker with a watch glass and heat gently until any rapid initial reactions have subsided. Then heat to boiling and boil until the contents of the beaker are almost dry. Remove the beaker from the hot plate and add 10 ml. of dilute nitric acid (1 to 1) and 10 ml. of perchloric acid (70 to 72 per cent). Replace the cover glass and heat very gently to a low boiling temperature (avoid superheating). Maintain this temperature until all organic material has been removed from the sides of the beaker and from the solution, which will be indicated by a colorless or slightly colored solution. Remove the cover glass, allow the beaker to cool a few minutes, and wash any adhering salts into the beaker, (If the cover glass is washed with perchloric acid, the contents of the beaker need not be cooled.) Evaporate t o dryness at a temperature just below the boiling point in a clean hood. If potassium is to be determined on the residue the ammonium salts should be removed at this oint. After the removal of ammonium salts, add 5 ml. of hydrociloric acid (I to 1) and 10 ml. of water. Heat until all salts are dissolved. Filter into a suitable volumetric flask. Wash the silica residue thoroughly with hot water and make the filtrate up to volume. Aliquot portions of the filtrate may be taken for subsequent analyses. The above method was applied to the plant materials studied and calcium, magnesium, potassium, and phosphorus were determined. Known amounts of calcium from a standard solution of calcium acetate, magnesium from a standard solution of magnesium sulfate, and potassium and phosphorus from a standard solution of potassium dihydrogen phosphate were then added to a duplicate sample of the material and the determinations repeated, using the same procedure. The acids and salts used were taken from the usual laboratory stock of c. P. reagents. Calcium was precipitated as the oxalate and titrated with permanganate as directed by Wiley (9). Magnesium was determined by the method of Handy (2) as modified by Truog and Chucka (8). The method of Schueler and Thomas (6) was used for potassium. Phosphorus was precipitated as the phosphomolybdate and titrated with sodium hydroxide according to the method given by Treadwell and Hall (7). Table I shows the amounts of calcium, magnesium, potassium, and phosphorus recovered.
