T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
600
sulfur trioxide present when I O g. were used, but no conclusion can be drawn from so few analyses, and in fact the results obtained by Fieldner and Taylor' show apparently t h a t the per cent of nitrogen is independent of the amount of potassium sulfate, provided the ratio of grams of potassium sulfate t o cubic centimeters of sulfuric acid is not greater t h a n 0.5. Thirty cubic centimeters of acid were used in all of the analyses reported in this paper. Mercury equivalent to 0.7 g. of mercuric oxide was added and permanganate was added a t t h e end of the digestion. It is shown by our analyses t h a t either t h e anhydrous or t h e hydrated sodium sulfate may be used in the Kjeldahl-Gunning method, t h a t the time of clearing is not affected appreciably by the water of crystallization of the sodium sulfate, and t h a t as little as 5 g. of potassium sulfate is sufficient in t h e analysis of substances such as we used. No analyses were made with greater amounts of sodium sulfate than 4.07,Since t h a t amount gave the same result as 5 g. of potassium sulfate, and j g. of the potassium sulfate gave the same result as I O g., which is the amount used in the official method. It is realized t h a t our reasoning is not quite conclusive because of the lack of a sufficient number of analyses to compare t h e results when 5 g. of potassium sulfate are used with these when I O g. are used, but the analyses of Fieldner and Taylor2 seem t o leave no question on this point. OKLAHOMA EXPERIMSNT STATION STILLWATER, OKLAHOMA
THE STRUCTURE OF SCARLET S3R (B) AND' PONCEAU 3R(By)
NaSOs-
-N = N-OH -NaSOa
03
R-salt
Vol.
IO,
No. 8
CH~C>CH CHa
+ pseudo-cumidine Scarlet S3R
This investigation therefore classes the Badische Scarlet S3R as Ponceau 3R, No. 8 3 Schultz Farbstofftabellen. I n making comparative tests of the Scarlet S3R with several classified Ponceaus, slight discrepancies were noted in the case of Bayer's Ponceau 3R. This dyestuff islisted by Schultz under No. 83 as being of the same structure as t h a t determined for Scarlet S3R. An investigation established the interesting fact t h a t Ponceau 3R is entirely different in structure from t h a t given by Schultz. Both cleavage products were found t o be naphtholsulfonic acid derivatives. Difficulty was encountered a t this point in obtaining either product free enough of t h e other t o proceed with their identification, as both were only slightly soluble in water, neutral sodium sulfite, etc. Small quantities of both components were finally obtained in a pure state. Further investigation established t h e rather unusual use of amido Bayer acid (I : 2-amido-naphthol-8-sulfonic acid) as the diazo componenT, and gamma acid ( 2 : 8-amido-naphthol-6sulfonic acid) as the azo component, thus giving Bayer's Ponceau 3 8 the structure: OH
"
-
+
By H. W. STIEGLER Received May 21, 1918
Amido Bayer acid gamma acid Ponceau 3R (Bayer)
Scarlet S3R (B. A. S. F.) is one of the more important of the unclassified azo dyestuffs (U. S. Dyestuff Census), some 80,000 lbs. being imported in 1913. I t was thought t h a t a determination of its structure would be of interest. The sample of Scarlet S3R was decomposed by means of SnC12-HC1 solution and t h e cleavage products separated and purified. The azo component was identified as amido R-salt (I : 2-amido-naphthol-3 : 6 di-sodium-sulfonate). Steam distillation of t h e alkaline reduction liquid yielded a brownish oil of no definite boiling point. On standing for some time (cold), traces of crystallization were noted. Separation by further cooling yielded a white crystalline solid, identified as pseudo-cumidine (I : 2 : 4-trimethyl-5-amido-benzene;melting point, 63' C.). The presence of a n oil with the pseudo-cumidine crystals probably indicates the use of crude cumidine, which contains a considerable amount of one of its isomers, mesidine. Scarlet S3R then, being a monazo dyestuff, has the following structural formula:
This investigation indicates a n error in Schultz, in Rt h a t Bayer's Ponceau 3R is lzot crude cumidine salt as stated there, b u t amido Bayer acid gamma acid. I t also classifies Scarlet S3R (Badische) as Ponceau 3 R , No. 83 Schultz.
1
Bureau of Mines, Technical Po$er, 64, 10.
2
t o c . Cil.
+
+
LOWELL TEXTILEORGANIC LABORATORIES LOWELL, MASSACEUSBTTS
AMMONIA AND NITRIC NITROGEN DETERMINATIONS IN SOIL EXTRACTS AND PHYSIOLOGICAL SOLUTIONS' By B. S. DAVISSON Received January 8, 1918
INTRODUCTION
Studies in soil biology dealing with the transformations of the soil nitrogen require frequent and exact determinations of ammonia and nitric nitrogen. T h e unreliability of the methods in vogue among soil biologists renders necessary a study of the means by which the true value for ammonia and nitric nitrogen can be obtained. The error due t o t h e hydrolyss of nitrogenous organic compounds is quite appreciable, and should be reduced t o a minimum. The often 1 An abstract of a dissertation presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in the Graduate School of the Ohio State University.
Aug., 1918
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
limited amount of material available for analysis renders i t desirable t o obtain both ammonia and nitric nitrogen upon t h e same sample. The method should be applicable for both large and small quantities of t h e two forms of nitrogen and t h e error should be of t h e order of magnitude of z per cent or less. HISTORICAL
The determination of ammonia in urine and in other animal fluids has received the attention of many biological chemists. The aim of t h e proposed methods has been t o obtain the true ammonia value by reducing t o a minimum the error due t o t h e hydrolysis of the nitrogenous organic matter. The methods employed in determining ammonia in soils are more or less modifications of those used by biological chemists for determining ammonia in urine. These methods, until recently, have been adopted by soil investigators without any study having been made of their applicability t o soil investigations. In I ~ O I ' Folin offered objections t o the Schlosing method of determining ammonia in urine because of t h e uncertainty of the time necessary for the transference of the ammonia t o the standard acid. He proposed diluting the urine sample with 400 t o 5 0 0 cc. of water, distilling the solution with MgO, collecting t h e ammonia in standard acid and titrating. The urine solution was then diluted back t o the original volume, distilled for a second period, and t h e ammonia collected in a new portion of standard acid. The ammonia obtained during the second distillation represents t h e urea which was decomposed. The decomposition for this period is assumed t o be the same as t h a t for t h e first distillation and t h e difference between t h e two values represents the preformed ammonia of t h e urine. Shaffer2 made a critical study of the methods used b y . biological chemists for determining ammonia in urine and found t h a t the earlier method of Schlosing (consisting in allowing t h e urine and t h e alkali t o stand under a bell jar with standard acid for absorbing t h e ammonia) and the Boussingault method (distilling with an alkali t o dryness in a vacuum a t 30' t o 40") gave dependable results when the directions of t h e original workers were carefully followed. With certain modifications, satisfactory results were obtained. Folin's3 method was found t o be unreliable because t h e ammonia from t h e second distillation did not represent the decomposition of urea during t h e first distillation. More urea was decomposed during t h e first distillation t h a n was represented by t h e ammonia recovered, consequently the second boiling gave results which were too high, thus reducing the value for the preformed ammonia. Folin4 outlined a second method which, with a few modifications, is now largely used for determining ammonia in urine and in other animal fluids. This method consists in the transference of t h e ammonia f r o m 2 ; t o 50 cc. of urine into standard acid by means Z. physiol. C h e m , 32
(1901), 515. 2 A m J . Physiol., 8 (1903), 330. 1
3 LOG. 4
Lit.
Z . physiol. C h e m . , 37 (1902-3),
of a rapid air current. The ammonia is liberated by I t o 2 g. of sodium carbonate and 8 t o 16 g. of sodium chloride. An air current of 600 liters per hr. for a period of I t o 1'/2 hrs. is necessary for the complete removal of the ammonia a t room temperature. The author found t h a t an appreciable amount of alkali is carried by a rapid air current and a trap, inserted between the aeration cylinder and t h e standard acid, is necessary t o arrest this alkali. A special absorption tube was devised t o insure complete absorption of t h e ammonia. This method was modified by Steel,' who used 0.5 g. of sodium hydroxide as the alkali. The hydroxide decomposes any triple phosphate present in the urine but does not decompose such nitrogenous organic compounds as urea, leucine, tyrosine, glycocoll, uric acid, hippuric acid, creatine, creatinine, and taurine. Russell2 investigated the Schlosing method for determining ammonia in soils by allowing t h e latter t o stand in contact with a strong alkali. To remove the danger of re-adsorption of ammonia b y the soil, he prepared a hydrochloric acid extract of t h e soil. Russell found t h a t distillation with magnesium oxide and alcoholic potash gave reliable results and did not decompose the nitrogenous organic compounds. The most reliable results were obtained when t h e soil was distilled under reduced pressure with either of these alkalies. Only 50 t o 7 0 per cent of the ammonia added t o a soil could be recovered. I n 1915 Potter and Snyder3 employed Folin's4 aeration method for determining ammonia in soils. The sample of 2 5 g. of soil was suspended in 50 cc. of water and aerated with z g. of sodium carbonate for $period of 19 hrs. at a rate of 2 5 0 liters of air per hr. The apparatus is essentially the same as t h a t used by Folin with the exception t h a t no t r a p was employed t o stop any entrained alkali. Remarkably concordant results were obtained and in nearly all cases the added ammonia was recovered. The work of Potter and Snyder is a step toward obtaining more reliable methods for determining ammonia in soils. The method of attack in this laboratory differs somewhat in t h a t we are working with solutions instead of with the soil direct. A serious objection t o the method as used b y Potter and Snyder is t h e time necessary for the removal of t h e ammonia. The employment of large volumes of solution and t h e reduction of the time of aeration present difficulties not encountered by these authors. Increasing t h e rate of aeration increases t h e error from entrained alkali and lack of absorption of the ammonia. It has been shown in this laboratory6 t h a t amounts of ammonia up t o 2 5 mg. can be recovered from 2 5 0 cc. of solution by aerating with magnesium oxide for a period of 3 hrs. at a rate of 1080 liters of air per hr. With such a rapid air current it was found t h a t complete absorption could not be obtained without t h e use of a scrubbing tower t o thoroughly wash the J . B i d . Chem., 8 (1910). 365. J Agr S c i , 3 (1910), 233. a THISJOURNAL, 7 (1915), 221. 1 2
4
161.
60I
LOC. cit
STHI3 JOURNAL,
8 (1916), 896.
602
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E a I S T R Y
Vol.
IO,
No. 8
passing air. After aeration is complete the absorbing acid is washed from t h e tower into a 500 cc. Kjeldahl flask and the ammonia determined by distillation with magnesium oxide. This distillation overcomes the error from entrained alkali.
upon the same sample. Also, an examination of the hydrolysis of nitrogenous organic matter under varying conditions is necessary t o establish the justification of such methods.
NITRIC NITROGEN
WATER-The water used in this work was distilled over sulfuric acid and potassium dichromate and the steam scrubbed before condensation. AMMONIA-BREE REAGENTS-A~~ the reagents were made ammonia-free before using. INDICATOR-Methyl red, prepared by dissolving 0 . 0 2 g. of methyl red in IOO cc. of double-distilled alcohol, was used. The solutions were carbon dioxidefree when titrated. AXMONIA soLuTIqNs-standard ammonium sulfate solutions were made from chemically pure ammonium sulfate and standardized by distilling with magnesium oxide. NITRATE soLuTIoNs-standard nitrate solutions were prepared from pure sodium nitrate and standardized b y t h e Valmari-Mitscherlich-Devarda method under ideal conditions. S T A N D A R D A C I D S A N D ALKALIES-The standard acids were prepared from chemically pure sulfuric acid and carbon dioxide-free water. The solutions were standardized by the sodium carbonate method, which, according t o Mitscherlich,l is the most accurate. Twentyfive cc. burettes provided with 3-way stopcocks and connected with reservoir bottles were used. These burettes are of regular 50 cc. burette length with a correspondingly smaller internal diamdter and graduated t o 0.05 cc. These burettes were standardized by the U. s. Bureau of Standards for 2 0 ' C. and the temperature was maintained as nearly as possible at t h a t point. Slight deviations from this temperature were neglected as they were found t o cause no appreciable change in the volume of t h e liquid. When portions of the solutions had stood in the burettes for 1 2 hrs. or more they were discarded. For small quantities of nitrogen N / s o acid was used, and for larger quantities N / I Oacid. Artificial light, having been found more satisfactory t h a n daylight because of its being constant a t all times, was used for all titrations. Large electric bulbs were used as a source of light. These bulbs were enclosed outside of the laboratory window with three panes of glass separating them from the titrating table. Thin paper was then placed over the window t o shade the eyes while t h e full light from the bulbs fell upon the table. D I S T I L L I N G APPARATUS-The apparatus used for distilling over magnesium oxide is shown in Fig. I, in which A is a quartz or Pyrex glass tube, B contains the ammonia solution and C is t h e Erlenmeyer flask containing t h e standard acid. Quartz Erlenmeyer flasks and quartz tubes were first used for the distillations. Later it was found t h a t the Pyrex glass could be substituted for t h e expensive quartz. Nitrate reductions were made in the apparatus
A method has been developed in this laboratory' for determining nitric nitrogen in soil extracts. The reduction methods were studied and the combination of the best features of the Devarda, Valmari-Devarda, and Mitscherlich-Devarda methods resulted in a method designated as the Valmari-Mitscherlich-Devarda method. The nitrates are reduced in a N / I O sodium hydroxide solution with I g. of Devarda's alloy. By using a minor modification of t h e Mitscherlich2 distilling apparatus very accurate results are easily obtained. The reduction is carried on for a period of 40 min. after the solution begins boiling. It
I I FIG.I
has been found t h a t the hydrogen evolved a t the boiling temperature is much more effective for reducing nitric nitrogen t h a n t h a t evolved at a lower temperature. Solutions containing decomposable nitrogenous organic matter are boiled for 30 min. with the alkali, previous t o the reduction of the nitrates. This preliminary boiling was intended t o destroy such nitrogenous compounds, b u t it has since been found t h a t this is not of universal application, as some of these organic compounds continue t o yield ammonia for several distillations. The methods for determining ammonia and nitric nitrogen have been developed t o give reliable results under conditions admittedly extreme, t h a t is, large volumes of solutions and small quantities of nitrogen. However, a further study, refinement, and modification of the methods seemed desirable in order t h a t both these forms of nitrogen might be determined 1
2
THISJOURNAL, 7 (19151, 521. Landw. Jahrb., 38 (1909), 279.
EXPERIMENTAL AMMONIA-FREE
1
Landw. Jahrb., 39 (1910),345.
T H E JOL'RNAL OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
Aug., 1918
shown in Fig. 11. Flask A contains the nitrate solution, B a pinch of magnesium sulfate and one of magnesium oxide with a small amount of water, and C contains t h e standard acid.
/
603
Formamide has undergone decomposition on aeration with the alkalies but the other substances do not show an appreciable decomposition when one considers the large sample of substance taken. It is doubtful if as easily decomposable a substance as formamide can exist in the soil as such for any length of time. Consistent results could not be obtained by h o t distillation with MgO. The results were in,consistent when the same gas burner or electric heater was used for all distillations. A soil extract rich in organic matter after standing inoculated with Aspergillus niger for several days was subjected t o analysis. T w o hundred cc. of the extract were used. T h e results in Table I1 show the justification of aeration mzthods for determining a m monia nitrogen in soil extracts. Boiling with magnesium oxide has given an error of 2 1 . 8 per cent. TABLEI1
\
Milligrams of Nitrogen in a Soil Extract Boiling over MgO Aeration over MgO 1.15 0,948 1.11 0.945
.. ..
FIG. I1
0.948 0.871
Av., 0.928
Av., 1.13
The aeration apparatus is shown in Fig. 111. The two towers, I and J, are used as scrubbers, one containing sulfuric acid a n d the other sodium hydroxide, when aeration< is made over magnesium oxide. If sodium carbonate is used as t h e alkali both towers may be filled with acid. R A T E O F AERATION-The rate of aeration was 1080 liters of air per hr. measured as previously described.l A Crowell pump was used for drawing air through the solutions.
The application of the methods for determining ammonia and nitric nitrogen was studied upon an extract prepared from a greenhouse soil which had been heavily manured for several years. The extract was prepared by extracting one part of soil with five parts of water. After agitation for 4 hrs. the extract was clarified with a laboratory centrifuge, some dextrose added, and, when nitrate-free, t h e extract was sterilized with chloroform and preserved in a closed bottle.
SEPARATION O F AMMONIA AND ORGANIC NITROGEN
Before a positive method for separating ammonia a n d organic nitrogen can be developed i t is necessary t h a t we have some knowledge of the hydrolytic action of t h e alkalies upon some pure nitrogenous organic compounds somewhat similar t o those found in t h e soil. The action of magnesium oxide a n d sodium carbonate was studied upon some pure compounds. T h e averages of several determinations on each substance are reported in Table I. ThAse substances were dissolved in water and aerated with magnesium oxide a n d with sodium carbonate. D a t a are also given for boiling with magnesium oxide for a period of 2 0 min. Substances containing a n amide group show considerable hydrolysis on boiling with magnesium oxide while the amino groups have not been appreciably attacked. T'ABLE I
NITROGEN OBTAINED Boiling Aeration Aeration Wt. of Substance Used with MgO with MgO w i t h NaKOa Mg Mg Mg Mg
Substance Forma'mide . . . . . . . . . . . 200.0 Acetamide.. . . . . . . . . . . 100.0
Urea... . . . . . . . . . . . . . . Asparagine. . . . . . . . . . .
....... .............
Aspartic acid.. Tyrosine. . . . . . . . . . . . . Leucine. 1 LOC. cit.
100.0 100.0
50.0
41.7 33.3
5.933 0.127 1.094 0,243 0.023 0.027 0.034
7.726 0.008 0.022 0.012 0.013 0.034 0.029
3.166 0.032 0.017 0.008 0.030 0.039 0.032
FIG.I11
Soil extracts rich in organic matter offer some difficulty in determining nitrates by reduction with Devarda's alloy in an alkaline solution. The preliminary boiling in N / I O alkali was found not t o destroy all decomposable nitrogenous compounds. The nitrate determination under such conditions has, therefore, a plus error. . A volume of 2 j o cc. of the extract continued t o give ammonia after two distillations of 30 min. each with 2 cc. of a 50 per cent sodium hydroxide solution. It was found t h a t much of this decomposable organic matter could be removed by using 2 cc. 'of a saturated lead acetate solution and subsequently boiling the filtrate with 4 cc. of 5 0 per cent sodium hydroxide. Table I11 contains results on the soil extract, one with organic matter removed and the other having it present. The amounts 0 . 0 2 8 and 0.017 are practically negligible.
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
604
TABLE111 NITROGEN OBTAINED Organic Matter Organic Matter
Dis tillatiou
No.
............... ................. ...................
l... 2. 3
Present Me N o t determined 0.109 0.057
Removed Mg. Not determined 0.028 0.017
After boiling 30 min. the solution is diluted back to cc., and 0.9 cc. of concentrated sulfuric acid is added, leaving approximately a N / I O alkaline solution for reducing nitrates. When sodium carbonate is used as the alkali it is necessary to use some substance other than lead acetate for removing the organic matter. Stutzer’sl reagent was prepared and found very satisfactory. At the outset of the work with the greenhouse soil, it was found that added nitric nitrogen could be recovered by reduction with Devarda’s alloy, but in no case could all of the added ammonia be recovered by aeration with magnesium oxide or sodium carbonate. The amount of ammonia remaining in t h e solution was usually less than 0 .5 mg. This retention of the ammonia has been attributed to the formation of the difficultly soluble magnesium ammonium phosphate2 or t o the presence of a considerable number of calcium and magnesium3 ions. Although the cause of the difficulty in this case was not exactly clear the addition of sodium oxalate overcame the difficulty, as shown in Table IV. A very heavy crystalline precipitate was obtained upon adding the sodium oxalate. 250
TABLEI V Wt. of Re- Nitrogen Reagent agent Used Takeu Used G. Mg. 11.27 0.5 MgO. 11.27 11.27 11.27 11.27 Na:COs. . . . . . . . 10 11.27 11.27 11.27 0:5 5.57 MgO, 5.57 5.57 5.57 10’ 5.57 NazCOs. 5.57 2 NazCzOa. . . . . . . . 5.57 .. 5.57
........
..
.. ..
.........
..
..
.......
..
Nitrogen Recovered Mg. 11.02 10.94 10.97 10.71 11.00 10.55 11.08 11.08 5.24 5.29 5.38 5.16 5.53 5.56 5.60 5.53
Brror -0.05 -0.33 -0.30 -0.56 -0.27 -0.72 -0.19 -0.19 -0.33 ---0 . 2 8 -0.19 -0.41 -0.04 -0.01 $0.03 -0.04
A number of determinations were next made using cc. volumes of the extract with added ammonia and nitric nitrogen. The extract was free from ammonia and nitric nitrogen. The results of the determinations are found in Table V , and the recovery of both forms of nitrogen is complete. 250
TABLE V Ammonia Nitrogen 4.73 M g . taken Mg. found Error +0.02 4.75 0.00 4.73 t0.05 4.78 -0.03 4.70 -0.01 4.72 -0.01 4.72 +0.02 4.75 .-0 .os 4 68 0.00 4.73 -0.03 4.76 Average 4 . 7 3 i.0.02 i0 . 0 2 Probable e r r o r , . 0.42 Per cent error..
......
.......
1
2 8
Nitric Nitrogen 4.03 M g . taken Mg. found Error -0.03 4.00 -0.04 3.99 $0.03 4.06 0.00 4.03 4.04 +0.01 0.00 4.03 -0.02 4.01 +0.02 4.05 +0.03 4.07 -0.03 4.00 4.03 *0.02 10.02 0.49
.. ..
Bureau of Chemistry, Bulletin 107 (1903), 38 Steel, J . B i d . Chew., 8 (1910), 365. Kober, J . A m . Chem. SOC.,SO (1908), 1279.
Vol.
I O , No.
8
The probable error obtained by calculation, using the method of least squares, is satisfactorily low. No determinations were made using larger quantities of the two forms of nitrogen. The order and magnitude of the error remains practically constant while the percentage error decreases with increasing amounts of the two forms of nitrogen. DISCUSSION
The data presented in this paper show t h a t the aeration method for determining ammonia in small volumes of urine can be successfully used for determining ammonia in large volumes of soil extracts and physiological solutions. Some modifications of the method as originally used by Folin’ were necessary for its application t o conditions encountered in soil biology studies. The hydrolytic action of magnesium oxide and sodium carbonate upon such nitrogenous organic compounds as occur in the soil is very small when the soil extract is aerated in the cold with either of the alkalies. The organic and ammonia nitrogen of the soil extract are easily separated by aeration in the cold with sodium carbonate or magnesium oxide. Hot distillation of soils or soil extracts with magnesium oxide gives unreliable results. This procedure has been widely used and the results obtained are of questionable value. When blank and check distillations are made the results are unreliable because of the unequal hydrolysis. The slowly decomposable protein-like substances which yield ammonia when making nitric nitrogen determinations are easily removed by using basic lead acetate or Stutzer’s reagent. The subsequent boiling with N / j sodium hydroxide destroys the remaining simpler substances which are likely t o decompose during the reduction of the nitric nitrogen. PROCEDURE
FOR
DETERMINING
XITROGEN
XITRIC
AND
AMMONIA
O N T H E SAME SAMPLE
I n the absorbing towers are placed 2 5 cc. of N / 2 sulfuric acid. Two hundred t o 2 j o cc. of the ammonia and nitrate solutions are placed in the aeration flask, a few drops of oil, 2 g. of ammonia-free sodium oxalate, and I O g. of pure sodium carbonate added, and the flask connected with the aerating apparatus. The solutions are then aerated for 3 hrs. a t a rate of 1080 liters of air per hr. After aeration is complete the ammonia is determined by washing the acid from the towzrs with 3 or 4 portions of 50 cc. each of distilled water and distilling with magnesium oxide in the apparatus shown in Fig. I. The aeration flasks are removed and the aerating tubes washed with distilled water. The tube is washed on the inside by forcing water into it by means of jet from a wash bottle and allowing the water t o drain into the aeration flask. The sodium carbonate is destroyed by adding j cc. of concentrated sulfuric acid. The solution is heated to boiling and 5 cc. of the copper hydroxide suspension are added and the boiling continued for about one minute. The solution is then 1
Lac. cit.
Aug., 1918
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
filtered, while hot, through a coarse I j cm. filter paper into another Kjeldahl flask. T h e aeration flasks are washed with hot water and the washings poured into the other flasks without filtration, since practically all the residue has been transferred t o the filter paper. T h e residue is washed 4 or j times with small quantities of boiling water. A small piece of paraffin a n d a n ebullition tube are placed in the flask t o prevent frothing and bumping, 4 cc. of a 50 per cent sodium hydroxide solution added, and t h e solution boiled for 30 min. The solution is diluted back t o 250 cc., and 0 . 9 cc. of concentrated sulfuric acid added t o reduce the alkalinity of the solution t o about N / I o . The nitric nitrogen is then determined by reducing with I g. of Qevarda’s alloy a n d boiling the solution for 40 min. after bringing it t o boiling in minimum time. When the solution reaches the boiling point it is advisable t o reduce the flame. The reaction is quite vigorous a n d may result in foaming over if this precaution is not taken. As soon as the vigorous action ceases, t h e flames are turned up and boiling continued. S U M MARY
The work reported in this paper justifies the following conclusions: I-Organic and ammonia nitrogen can be separated by aerating t h e solutions in t h e cold over magnesium oxide or sodium carbonate. 11-Ammonia determinations obtained by boiling soil suspensions or soil extracts rich in organic matter with magnesium oxide are unreliable. 111--Ammonia a n d nitric nitrogen can be accurately determined upon t h e same sample by the method reported in this paper. The author takes this opportunity t o t h a n k Dr. E. R. Allen for his helpful criticisms in this investigation. LABORATORY OF S O I L BIOLOGY OHIOAGRICULTURAL EXPERIMENT STATION WOOSTER. OHIO
STUDIES IN SYNTHETIC DRUG ANALYSIS’ V-ESTIMTION OF THEOBROMINE B y W. 0.EMERY A N D G. C. SPENCER Received April 25, 1918 INTRODUCTION
Questions having quite recently arisen relative t o the actual therapeutic strength of certain diuretic combinations of theobromine a n d theophylline, notably with sodium acetate a n d sodium salicylate, a n investigation of such products seemed desirable. In t h e present paper, however, consideration will be given only t o experiments involving theobromine and looking t o t h e utilization of its periodide as a basis for evaluation. A description of very similar work on theophylline and its combinations is reserved for a future communication. T h e quantitative estimation of theobromine in admixture with other agents of medicinal value, or with materials of a more or less inert nature, is complicated by the great insolubility of this compound in the more 1
THIS JOURNAL, 6 (1914), 665.
605
common organic reagents. A somewhat similar difficulty, encountered in developing a procedure for the estimation of acetanilide a n d phenacetin (acetphenetidine) in admixture, was met by t h e employment of glacial acetic acid. A number of preliminary trials soon demonstrated t h e adaptability of this solvent also for theobromine when applied t o periodide formation. It was further found t h a t t h e solubility is favorably affected by t h e presence of sodium acetate. While it is by no means difficult t o prepare several periodides of theobromine, its quantitative separation in the form of a n iodine addition-product of unvarying composition, suited t o t h e purposes of titrimetric control, is manifestly beset with difficulties naturally inherent in operations of this character, such as I;lomogeneity, freedom from other periodides a n d salts, losses by decomposition, evaporation, etc. I n t h e method presently t o be described, advantage is taken of,the fact t h a t , when a n acetous solution of theobromine containing sufficient iodized potassium iodide is treated with a mineral acid, a grayish black crystalline precipitate separates, which, judged by its iodine content, has t h e composition C7HsN402.HI.14. The separation of this hydriodo-tetriodide becomes quantitative if t h e iodine solution is reenforced with a small quantity of sodium chloride. If, therefore, t h e precipitation is effected in a measured volume of standard iodine and the insoluble additionproduct removed by filtration, t h e volumetric determination of t h e unconsumed iodine is readgy accomplished, and therefrom t h e quantity of theobromine involved as readily calculated by means of t h e appropriate factor. EXPERIMEKTAL
For purposes of identification, t h e above-mentioned periodide was prepared by dissolving theobromine in a feN cubic centimeters of hot glacial acetic acid, transferring t h e clear liquid to a flask containing sufficient iodized potassium iodide, adding a little concentrated hydrochloric acid with constant agit?tion, and then shaking the mixture vigorously. After standing some hours, t h e periodide was isolated by pouring t h e product onto a small suction plate provided with a suitable filter, washing the mass several times with a saturated aqueous iodine solution, a n d exposing the crystals in t h e open air until apparently dry. Protracted exposure of t h e substance t o atmospheric influences, however, is inadvisable, such treatment inducing a n appreciable lightening in the color of t h e crystals with consequent loss of chemically combined iodine. T h e “exterior” iodine was determined by titration with sodium thiosulfate in alcoholic solution, a n d in t h e presence of sodium bicarbonate. Total iodine, on the other hand, was estimated by first treating the substance in acetic a c i d . with a saturated solution of sulfur dioxide in water, followed by precipitation with silver nitrate. Calcd. for C7HsN40e.HI.L: 14, 62.2; Is, 77.8. Found: IP,59.9, 61.7, 62.2; Is, 76.5, 7 7 . 0 , 77.2.
Thus i t appears t h a t , even with the greatest care, the operations of washing and drying are likely t o be
