ANALYTICAL EDITION
118 Table I-Comparison
of Lead S u l f a t e a n d B a r i u m S u l f a t e Methods SAMPLE H2S0, SULFUR USED PbSO4 FORMED PbSO4 METHOD
Grams Heavy oil residue 1.4500 1 2910 Panuco crude oil Heavy oilresidue 1:5634 Upton-Crane crude oil 1.5932 Inalewood crude oil 1.2838 Silks ael with adsorbsd sulfur compounds 1.3130 Distillate, b. p. 195230’ C. 0,6340 Distillate, b. p. 155170’ C. 1.2500 Light oil 0.7633
Gram 0.2497 0.2089 0.1306 0,1087 0.0888
1
1 Per5.63cent I
;%
2.23 2.28
DEVIA-
AS:
Procedure
1
1 Per-0cent .9
Enough oil is burned in the oxygen bomb to produce approximately 0.03 t o 0.25 gram of sulfuric acid. The bomb
-
Note-A Parr oxygen bomb of illium alloy with a capacity of 375 cc. was used. The oxygen pressure was 35 atm. In exploding the sample the procedure of the A. S. T. M. Method D129-27 (2) was followed except that a larger sample was used.
ilff
5.68
2.30
I
fl.1 -1.8 -0.9 1.7
0,0321
0,800
0.831
-3.7
0.0151
0.780
0.820
-4.9
0.0129 0,0117
0.338
0.335 0.507
+0.9 -1.6
0.499
POTASSIUM IODIDE-Dissolve 50 grams of potassium iodide in water and make up to 50 cc.
Bas04
Per cent
VoI. 1, No. 3
Solutions Three solutions are necessary-an approximately 0.2 N aqueous solution of lead nitrate; a standardized, approximately 0.1 N solution of sulfuric acid to be used in standardizing the lead nitrate; a strong, almost saturated solution of potassium iodide t o be used as indicator. LEADNITRATE-Dissohe about 33 grams of lead nitrate in water and make up to 1 liter. Standardize by titration against the standard sulfuric acid as follows: Take about 20 cc. of the sulfuric acid and add water to a volume of 50 cc. Add 100 cc. of 95 per cent alcohol. The alcohol and water may be measured from a graduated cylinder. Add about 0.2 cc. of potassium iodide indicator. It is most convenient to measure this by drops and to take the same number of drops each time. Run in the lead nitrate until a permanent yellow color is produced. The following calculation gives the sulfur value of 1 cc. of the lead nitrate solution: Cc. H&O* X normality &So4 X 0.016032 = grams sulfur per cc. CC.Pb(N0a)s SULFURIC ACID-An approximately 0.1 N solution of sulfuric acid should be used. This may be conveniently made by diluting 3 cc. of the conrentrated acid (sp. gr. 1.84) to 1 liter and standardizing by any of the usual methods.
is washed out into a beaker keeping the volume of the solution as small as possible. The indicator is then added, the same amount being used as in the standardization of the lead nitrate solution. A small amount of powdered aluminum is then added (about 0.01 gram is enough ordinarily), and the solution is boiled down to a volume slightly less than 50 cc. If the solution still remains yellow, not enough aluminum has been added. A 300-cc. Erlenmeyer flask is marked to show when a volume of 50 cc. is reached. The solution is transferred to this flask and cooled by holding the flask under running cold water. The beaker is rinsed out into this flask with a small amount of water, enough t o make up to 50 cc., and the rinsing finished with the 95 per cent alcohol to be used in making the solution up to a 50-70 per cent alcohol solution. To do this, 100 cc. are measured out and added to the 50 cc. of aqueous solution in the flask, as much as desired being used to complete the rinsing of the beaker. The solution should be colorless. The presence of metallic iron or aluminum in the flask does not interfere. The lead nitrate is then run in until a permanent yellow color is produced. The calculation of the percentage of sulfur in the sample is as follows: Cc. Pb(NO& X sulfur value of 1 cc. X 100 = per cent sulfur Grams sample Literature Cited (I) Am. Soc. Testing Materials, Tentative Standards, 1927, p. 433, Method DDO-26”. (2) Ibid., Standards, 1927, Pt.11, p. 419, Method D129-27. (3) Nikaido, J . Am. Chem. SOC.,24, 774 (1902).
Rapid Method for Determination of Phenols‘
.
J. A. Shaw THEKOPPERSCOMPANY, MELLONINSTITUTE OF INDUSTRIAL RESEARCH, UNIVERSITY OF PITTSBURGH, PITTSBURGH, PA.
HE following method has been used in these laboratories in developing the gas recirculation process for the recovery of phenols from ammonia liquor. The development of this method has enabled one man to run a very greatly increased number of tests and as a consequence has facilitated research work upon phenol removal. The words “phenol” and “phenols” are used in the generic sense in this article and refer not only to CaH60Hbut also t o homologous and related compounds. Apparatus
T
One steam chamber (Figure 1) Two Pyrex test tubes, 8 X 1inch (20.3 X 2.5 cm.) Three test tubes, 4 X ’/zinch (10.2 X 1.2 cm.) One Liebig condenser, 10 inch (25.4 cm.) One set of graduated cylinders, 25, 50, 100, and 250 ml. Two No. 5 rubber stoppers, two holes, for 8 X 1 inch test tubes One rubber stopper for condenser inlet Presented before the Division of Water, 1 Received March 19, 1929. Sewage, and Sanitation Chemistry a t the 77th Meeting of the American Chemical Society, Columbus, Ohio, April 29 to May 3, 1929.
TWO 10-h~ MOIX pipets Two 300-ml. bottles with stoppers for standard solutions One 500-ml. bottle with stopper for standard solution Necessary glass and rubber tubing and iron stands Water and compressed air connections
Description of Steam Chamber One of the main advantages of this method of determining phenols is the short time required. Since it involves a distillation, it is necessary that the sample shall be very small or the rate of heat input high. There are various objections to using a high-temperature source of heat, one of which is the danger of charring the organic matter in the sample. It was therefore decided to employ a form of steam distillation. With a small sample, such as 10 ml., only R very small amount of either evaporation or condensation is allowable. The present system using a steam chamber was therefore adopted. The flow of hot vapor is easily and delicately controlled by the setting of the air valve. The temperature is kept almost exactly a t the boiling point by inclosing the
.
July 15, 1929
INDUSTRIAL AND ENGINEERING CHEMISTRY
119
train in a steam chamber. At first boiling water was used phenol up to 1 gram per liter in the sample. It has, however, as the immersion medium, but steam is much to be preferred. been found possible to distil samples of considerably higher The steam chamber consists of two parts-a box with a phenol content, but a longer time and more care are required U-shaped gutter around the outside edge a t the top, and a so that 1 gram per liter has been set a8 the practical working cover the edges of which rest in the gutter when it is in limit. When the concentration of phenol is higher than this place. The box has a drain a t the bottom to carry away the sample should be diluted to less than 1 gram per liter condensate. Steam is introduced into the side about 3 of phenol. The distillate is mixed by shaking and is now inches (8 em.) from the bottom, with the inlet turned in such ready for necessary dilution and treatment with bromine a way as to direct the jet of steam toward the bottom of the water. The actual estimation of the phenols depends upon the box. I n each end face of the top there is a slit which accommodates the inlet and the outlet tubes of the train. fact that when a solution containing phenols free from cerSteam also escapes through these slits keeping the pressure tain impurities is treated with bromine water, a colloidal approximately a t that of the atmosphere. This steam precipitate is formed in the solution and under carefully prechamber can conveniently be made of galvanized iron. Or- scribed conditions the cloudiness produced thereby bears a definite relation- to the dinary pipe fittings can be amount of phenol present u s e d f o r the steam inlet (I).* and the drain. The steam In the course of developing the gas recirculation For the purpose of dechamber is mounted on a method for removal of phenols from gas condensates, termining the amount of stand a t a convenient height The Koppers Company Laboratory has developed a phenol present, the exact (about 15 inches (38 cm.)) method for determination of phenols that has proved quantity of the previously above the table top so that to be very satisfactory, particularly with regard to the mentioned distillate is noted a small graduated cylinder time factor. The time required has been reduced from and small portions of it are can be placed under the con6-8 hours for the well-known Skirrow method to about successively diluted until denser. 30 minutes for the new method. The dimensions and dethe conbentration falls beThe method is simple, requiring no unusual reagents tails of construction are tween 30 and 35 p. p. m. or apparatus though certain forms of apparatus are given in Figure 1. Three carefully sized 4 X desirable. The procedure consists of a rather special'/z inch (10.2 X 1.2 em.) ized distillation followed by a determination of the Reagents test tubes are then held in phenols in the distillate turbidimetrically. The turthe hand side by side. Into Sulfuric acid, c. P., sp. gr. bidity is produced by the addition of bromine water. the middle tube is poured 1.84 Certain empirical conditions must be fulfilled and cerBromine water about 1 to 1.5 inches (2.5 to tain classes of substances must be eliminated. These Stock solution of phenol 3.8 cm.) of the sample and are discussed in detail. (1.000 gram of C~HSOHper into the tubes on either side 1i t e r ) . Standard solutions The method has been quite thoroughly checked both is poured about the same containing 30 and 35 p. p. m. in the laboratory and by use in the plant. The method phenol are prepared from the a m o u n t of t h e standard is sensitive to 30-35 p. p. m . of phenols. above stock solution by dilutphenol samples containing, ing 30 ml. and 35 ml.. rer e s p e c t i v e l y , 30 and 35 spktively, to 1000 ml. 'with p. p. m. phenol. The soluwater. T h e s e s h o u l d b e prepared fresh daily. 250 ml. of each are therefore usually tions in the three tubes are adjusted to equal depth by shaking sufficient. out a few drops from those containing the greatest amounts. Methyl orange indicator solution The temperatures of the solutions in the tubes are brought to the same point by shaking for a few minutes in a water bath. Procedure The desirable temperature is 18" to 20" C. These solutions The two Pyrex test tubes are connected by means of glass are then treated with sufficient bromine water to color each tubing and rubber stoppers as shown in Figure 2. The a light straw yellow of about the same depth of color. Usufirst tube is two-thirds filled with water. Ten milliliters ally three to four drops of bromine water is the right amount. of the sample to be analyzed are placed in the second tube The tubes are then closed with a clean finger and given a and acidified with a few drops of 1: 3 sulfuric acid, using few violent shakes to mix. methyl orange as indicator. The test tubes are then stopThe comparison is made immediately by holding the tubes pered and placed in the steam chamber. The outlet from about a foot (30 cm.) above a plain, dull, light-colored surface in the second tube is connected with the condenser. The inlet a good light so that each tube will get the same amount of light. of the first tube is connected with the air line. It is now By a little experimenting the analyst will soon determine the possible to blow air through the water, then through the most suitable intensity of light for his individual eyesight. sample and into the condenser. A gentle current of air (in Upon looking down through the tubes (not crosswise)the degree distinct hubbles a t the start) is turned on and then steam of turbidity (not the color) may be observed. If the dilution is turned into the steam chamber, the condenser first having is correct, it will be found that the sample will fall between been set to deliver into a 25-ml. graduated cylinder. The the two standards, and by interpolation the sample may be air rate should be increased during the latter part of the estimated as having 30, 31, 32, 33, 34, or 35 p. p. m. phenol distillation. concentration. The air passes through the hot water and becomes satuCalculation rated and the hot vapor thus formed passes through the sample and distils out the phenols without materially changP. p. m. phenol in standard matched X total dilution factor = ing the volume of the sample. As large a volume of vapor p. p. m. phenol in sample. as desired may thus be passed and, since there can be neither Discussion of Method charring nor dilution, very small volumes of sample can be The separation of phenols from contaminating substances distilled and the distillation accomplished in a very short time. When 25 ml. of distillate have been collected, the by distillation as described above is empirical, and the prodistillation has been found complete for concentrations of * Italic numbers in parenthesis refer to literature cited at end of article.
ANALYTIC A L EDITION
120
cedure should be followed exactly as described, except when it is necessary to remove certain interfering substances as described below. The method of distillation described has advantages not possessed by a dry or an ordinary steam distillation. A discussion of this point would require too much space here. The 10-ml. sample distilled should be slightly acid and should not have a phenol concentration of more than 1000 p. p. m.
VOI. 1, No. 3
Ordinary tap water is usually as satisfactory for the dilutions as distilled water. Upon the addition of bromine water the tubes should be compared immediately. Not more than about 15 to 20 seconds should elapse between the bromine addition and the reading of the tests. Since the method depends upon judging the depth of cloudiness in the solution due to the formation of brominated phenols, it is suggested that the analyst practice reading the turbidity formed in known solutions before attempting the analysis of unknown mixtures. It is believed that the foregoing is sufficiently specific to enable any chemist to make a satisfactory determination of phenols in water solution within the apparent h i t s of accuracy of the test. It has been found readily possible to interpolate within two points between 30 and 35 p. p. m. This method has been checked against synthetic solutions, against the Skirrow method (b), and has stood the test of use in engineering research calculations. The following results were obtained by an independent analyst on samples containing known amounts of phenol: PHENOL ACTUALLY PRESENT Grams per liler 5.26 4.78 2.01
1.83
0.72 0.67
0.100 0.200
Figure 1-Steam
Chamber
Alcohols, amines, aldehydes, organic bases, oils, and inorganic salts interfere with the bromine test. It is therefore necessary to separate them from the phenols. Oil alone may often be separated sufficiently by filtering the sample through a wet filter paper. A better method, which will eliminate many other substances as well, is to render the sample alkaline with sodium hydroxide, using enough of the caustic to make a solution containing 10 per cent sodium hydroxide, distil off 15 ml. as in the phenol determination, discard the distillate and wash the condenser, cool the sample, acidify carefully with sulfuric acid and distil off the phenol as above catching 25 ml. of distillate, etc. If, however, very large quantities of ammonium salts are present, as in saturator liquor, it is necessary to make a distillation from acid solution first and then render the distillate alkaline with sodium hydroxide (10 per cent free NaOH), distil off the pyridine, etc., then acidify and again distil off the phenols, catching 25 ml. of distillate from 10 ml. of sample. Ammonia liquors that have stood for some time will often exhibit foaming during the distillation. This may usually be overcome by filtering a quantity before measuring out the sample. As has been stated, the bromine method depends upon the formation of a colloidal precipitate. In general, anything affecting the colloidal condition interferes with this test. Examples of this are temperature, acidity or alkalinity, and salts or other substances in solution such as protective colloids or coagulants. I t is necessary that the temperature of the solution in all three comparison tubes be the same-preferably about 18" to 20" C.; hence the use of the bath of water. It must also be remembered that while the coloration due t o excess bromine should be nearly the same, it is the cloudiness, not the color intensity, that is a measure of the phenols iprment.
PHENOL FOUND BY M ~ P ~ H O D Grams $0liln 8.46 4.72 1.94 1.70 0.78
0.67
0.097 0.184
The method has cut the time required for phenol determinations from 6 or 8 hours for a Skirrow method to about 20 minutes for this method. One man working with two steam chambers has made twenty-five determinations of phenol in 7 hours and could doubtless have made more with better equipment. It is evident that 30 p. p. m. marks the lowest concentration that can be determined directly by this method. Solutions having lower phenol concentrations, however, can be analyzed by this method if they are concentrated by evaporating in such manner as to have a concentration of not less
Figure 2-Set-Up of Apparatus
than 10 per cent sodium hydroxide in the hot phenol solution at all times. It has also been suggested to the writer by Erich Laue that samples of low phenol content may be analyzed by this method if the phenol present is fortified by the addition of a measured amount of a stronger phenol solution, the phenol concentration of which is definitely known. This procedure, of course, operates a t the expense of accuracy and must be used with discretion. Phenols dissolved in oils may be analyzed by this method if the phenol is first transferred to water by washing the oil with three successive portions of 10 per cent sodium hy-
July 15, 1929
INDUSTRIAL AND ENGINEERING CHEMISTRY
droxide solution, each portion representing about 20 per cent of the volume of the oil. Several variations have been proposed for increasing the accuracy obtainable by this test, but almost without excep tion they tend to complicate the test, increase the time required, or interfere with the immediate comparison of the colloidal bromophenols. In certain special instances a little more refined technic has been employed and a somewhat higher degree of accuracy has been obtained at the expense of the time factor. The above procedure, however, was
decided upon as the best for the purpose in hand-the of phenols in the by-product coking system.
121
study
Acknowledgment
Grateful acknowledgment is made to Erich h u e for making check analysis relative to establishing the accuracy of the method. Literature Cited (1) Rose and Sperr, Am. Gas Assocn. MOnlhl7, 2, 117, 328 (1920). (2) Skirrow, J . SOC.Chem. I n d . , 27, 58 (1908).
Determination of Nitrate Nitrogen in Tobacco' Hubert Bradford Vickery and George W. Pucher BIOCHEMICAL LABORATORY, CONNECTICUT AGRICULTURAL EXPERIMENT STATION,
N E W HAVEN, CONN.
A modification of the Jones method for determining HE nitrate content of absolutely and r e l a t i v e l y nitrate nitrogen is described which is especially adapted plant tissue is usually greatly reduced, and no espefor the investigation of tobacco or its extracts. The determined by titracial precautions are necessary method consists in the preliminary quantitative retion of the ammonia produced to carry out exactly duplicate moval of nicotine by steam distillation of a suspension after subjecting an aqueous distillations of sample and suspension of the tissue to the *of the tobacco in an alkaline solution. Nicotine may be blank. A determination of determined in this distillate and nitrate in the residue action of a reducing agent. the nicotine content may be by reduction with acid and reduced iron powder. It is Owing to the presence of ammade on the steam distillate. necessary to conduct a blank determination omitting monia, or of substances yieldSteam distillation with alkali the reducing agent, but the relative magnitude of this ing it, in most tissues, it is removes any p r e e x i s t i n g blank is small and it is not highly variable. Data are ammonia from the sample, necessary to conduct a congiven illustrating the wide variability of the proportion trol determination in the aband also hydrolyzes and reof nitrate nitrogen in tobacco and its dependence on sence of the reducing agent, moves most, a t least, of the the type of nitrogenous fertilizer employed in growing and the nitrate content is amide nitrogen together with the Plantcalculated from the difference some ammonia of-other secin the amounts of ammonia ondary origin-e. g., from found. arginine. The occasional use of urea as a fertilizer for toThe wide variation in the nitrate content of samples of bacco introduces a possibility that this substance may be normal tobacco renders the determination of this constituent present in the sample. Although the writers have not yet a matter of considerable importance. Unfortunately, how- found urea in tobacco extracts, control experiments with ever, accurate analyses are difficult, since by the usual extracts to which urea had been added have shown that it is methods a considerable proportion of the nicotine present decomposed to ammonia and removed during the alkaline in the sample comes over during the distillation of the am- distillation. Control experiments also showed that the premonia from alkaline solution. A method for the determina- liminary distillation had no effect upon the nitrate present tion of nitrate in tobacco was therefore sought which should in, or added to, a tobacco extract and that added nitrate be free from inaccuracy due to the volatility of nicotine. could be recovered quantitatively. Many methods have been described for the reduction of Identification of Substance Yielding Ammonia nitrate to ammonia as a step in the quantitative estimation of nitrate nitrogen. Sbydell and Wicher (6)*investigated a As a preliminary step i t was thought essential definitely number of these methods and concluded that the most satisto establish the identity of the substance occurring in tobacco factory results were secured when the reduction was carried out in an acid solution. Methods for the analysis of the which, on reduction, yields ammonia. Accordingly a sample nitrate content of fertilizers, based on this procedure, have of an extract of fresh tobacco leaves was distilled with steam been in common use for several years. The presence of such in the presence of an excess of calcium hydroxide until the ingredients as cyanamide or urea in fertilizers introduces a nicotine was removed. The residue was evaporated to drypossibility of error which can be controlled only by properly ness, powdered, and extracted in a Soxhlet apparatus with conducted blank experiments, and Jones (6) has described a 95 per cent alcohol, in which calcium nitrate is soluble, until modification of the acid reduction method which can be the diphenylamine reaction for nitrate could no longer be applied in such cases. His method has recently been em- obtained upon the residue in the extractiou thimble. The extract was evaporated, taken up in cold water, filtered from ployed for the analysis of tobacco (8). The present method is a modification of the Jones method. a little fat, and treated with nitron acetate solution. Nitron I n order to avoid the high blank determinations due to the nitrate crystallized almost immediately and, when recrystalpresence of nicotine, this substance is removed from the lized from water, decomposed a t 259-261' C. (not corrected). sample by steam distillation from an alkaline solution. This Busch (3) gives the decomposition point as 260' C. The dehas several advantages. The blank determinations are both composition point of a mixture of this preparation with nitron nitrate prepared from the pure potassium salt was 1 Received April 3, 1929. The expenses of this investigation were 261" C. The sample of tobacco leaf extract so treated conshared by the Connecticut Agricultural Experiment Station and the Carnegie tained 0.149 gram of nitrate nitrogen, as indicated by the Institution of Washington, D. C. * Italic numbers in parenthesis refer to literitture cited at end of article, iron reduction method described below. The nitron ni-
T
