Mar., 1915
T H E J O U R N A L OF f N D l l S T R I A L A N D ENGINEERING C H E M I S T R Y
THE DETERMINATION OF AMMONIA 1N SOILS B y R . S. POTTERAND R. S. SNYDER Received October 28, 1914
Theoretically, t h e ideal method for ammonia in soils would give t h e absolute amount of ammonia present as such and as the ammonium radical, but in the light of our present knowledge, it cannot be said beyond all doubt t h a t any conceivable method would give this ideal result. This is true because a large part of the nitrogen of t h e soil is present in protein and protein degradation products, t h e extensive decomposition of which gives large amounts of ammonia. The uncertainty regarding just what products are present does not permit t h e finding of conditions which we can be certain will expel ammonia already present and not decompose any material t o give, among other products, ammonia. The problem is further complicated by t h e well-known absorptive and adsorptive powers of the soil. Therefore, since this ideal result is not attainable, in our opinion the value of any method will depend upon whether it fulfils the requirements given below. R E Q U I R E M E N T S OB A M E T H O D BOR AMMONIA
I-Closely agreeing duplicate results should be given and t h e same result obtained whether t h e reagent or reagents act, within reasonable limits, for a longer or shorter period. Of course, t h e reagent or reagents must not include any which are known t o decompose material contained in t h e soil t o give ammonia. For instance, it would not be permissible t o use strong, hot hydrochloric acid, for this reagent decomposes proteins t o give large amounts of ammonia. 11-Upon the addition of a known amount of ammonia, the method must give this added amount plus t h a t previously found in t h e soil. Not enough time, of course, should elapse between the addition of t h e ammonia and the determination for any bacterial action t o take place. I n soils already containing ammonia, and apparently all soils do, it is difficult t o see how any chemical reaction could, in a reasonable length of time, change t h e added nitrogen t o some other combination t h a n t h e ammonium. Because his method does not recover all added ammonia, Russel assumes t h e ammonia t o be chemically changed. This will be further discussed later. 111-For use in a soils laboratory, the method should permit one t o run several determinations within a reasonable length of time. DISCUSSION O F METHODS
The seven methods for ammonia are discussed below. I-Schloesing’s first method’ consists in leaving a mixture of t h e soil with strong sodium- hydroxide solution under a bell jar, together with a vessel containing some standard acid t o absorb t h e ammonia evolved. This method has fallen into complete disuse because of t h e recognition of the fact t h a t strong sodium hydroxide solution even in the cold would gradually decompose the organic nitrogen compounds t o give amnia. I.-In Schloesing’s second method1 t h e soil is treated A nelyse des &fuli$res Agvicdej, 1879.
221
with dilute hydrochloric acid until the liquid remains distinctly acid. The mixture is then well shaken, filtered and an aliquot portion distilled with alkali. Since this method complies with the third requirement named above, and since no evidences in t h e literature t h a t i t does or does not comply with t h e first two can be found, the method has been given a rather detailed examination by us, the results of whic‘h are reported and discussed in t h e experimental part of this paper. 111-In Boussingault’s method,l a mixture of about one part of soil t o two parts of water, together with a little magnesia, is distilled into standard acid. It is quite generally admitted t h a t this method does not give true results with material containing protein and protein degradation products as does t h e soil. Yet, because of its ease arid simplicity of manipulation, it is often used, it being stated t h a t t h e results, although not absolute, are comparable with one another. This method has also been examined by us and the results will be taken up in the experimental part of this paper. IV--The method of Wolf and Knop2 consists in treating the soil with sodium hypobromite, which reacts with the ammonia t o give nitrogen gas which is measured. Baumann2 has shown t h a t this method does not give accurate results. V--In Baumann’s method3 a hydrochloric acid extract is made as in Schloesing’s second method. T o this extract, magnesia is added and ozone is bubbled through the solution, and it is then treated with sodium hypobromite and t h e nitrogen collected a n d measured. I n view of t h e facts which are brought out below in regard t o t h e hydrochloric acid extract of soils, together with t h e fact t h a t t h e sodium hypobromite used contained an excess of sodium hydroxide which would gradually decompose nitrogenous organic compounds to give ammonia, Methods IV and V may be considered as unreliable. VI--In Russel’s first m e t h ~ d I, 50 ~ g. of soil are distilled a t the pressure obtainable with a water pump with 2 g. of magnesia suspended in I O O cc. of water. The distillation flask is kept a t 40’ C., and t h e distillation continued for six hours. VII-Russel’s second method4 is just like his first method, except t h a t t h e 2 g. of magnesia are replaced by 0.7 g. potassium hydroxide and the I O O cc. of water by an equal volume of alcohol, and t h e temperature of the distillation flask is kept a t 2 5 ’ C. instead of 40’ C. Russel states t h a t in so far as he has tested his methods, they give practically identical results. He also points out t h a t with the use of magnesia there is some decomposition. Indeed, with t h e soil upon which he reports he obtained both on t h e second and t h e third distillation as much ammonia as was obtained on the first. Although Russel shows t h a t his potash method gives concordant results and no ammonia is evolved after a reasonable length of time (first requirement), yet upon addition of ammonia t o t h e soil, only from 50 t o 84 per cent is recovered by either method. Hence, ’
Agronomie, 3 (1864), 206. Chem. Centvalbl., 1860, pp. 243, 253. a Landw Vers. Stat., 33 (1886), 247. 4 Jour. Agv. Sci.,[3 (1910). 233. 1
a
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 CHEMISTRY
222
they do not comply with the second requirement. He explains this by stating t h a t the ammonia enters into a stable combination which is not a n ammonium compound. I t could be argued with equal force t h a t the ammonia forms a relatively insoluble ammonium compound with the complex organic or inorganic material of the soil, or t h a t it was physically adsorbed. I t does not seem surprising t h a t all the ammonia was not expelled when j o cc. of water were left with the soil after t h e distillation was stopped, and the high solubility of ammonia' and the low solubility of magnesium hydroxide2 are considered. E X P E R I hl E iYT A L
I n the following work, the ammonia-free water which was used was prepared by boiling ordinary t a p water for a few minutes with about two grams of potassium acid sulfate t o one liter of water, and then distilling the solution and condensing the vapor in block-tin tubes which had previously been steamed for 1 2 hours. The distillate in every case was found t o be neutral t o alizarin red and gave but a slight indication of ammonia when tested with Nessler's solution. Whenever magnesia was used, it was always freshly calcined. Alizarin red was the indicator used in all ammonia titrations. I t has been found by far the most satisfactory indicator for ammonia titrations, and it is now exclusively used for t h a t purpose in this laboratory. Blanks were always run and the suitable corrections made. Duplicates were not run, except as reported. The following soils were used in this work: Soil Soil Soil Soil
No. A-0. No. No.
1-Southern Iowa loess 2-Ioma drift 3--Wisconsin drift 4-Mississippi loess
Soil No. 5-Missouri loess Soil No. 6-Alkali I Soil S o . 7-Alkali I1
Soils Nos. I t o j were typical samples from the five principal soil areas of Iowa. Soils Kos. 6 and 7 were from the Wisconsin drift area.
machine, the mixture was filtered through a doublc folded filter, a clear filtrate being obtained in all cases. For the magnesia distillation, zoo cc. of the filtrate, representing r o o g. of soil, were distilled after the addition of about I O g. of magnesia until 150 cc. of distillate had been collected. In some cases, 0.1iV acid and alkali were used in t h e titrations, and in others 0 . 0 2 S. The results of the magnesia distillations are given in Table I . h'orTime Soil Lab. mality shaken No. No. of acid Min. 1 1,2 0.2 20 20 3,4 0.3 60 5, 6 0.2 60 7, 8 0.3 9, 10 0.2 30 11,12 0.3 30 3 13 0.2 20 20 14 6.3 60 15 0.2 60 16 0.3 20 17 0.2 20 18 0.3 19 0.2 60 20 0.3 60 (a) In computing these t o 8 were used.
30.7 g. per 100 g. water a t 40' C. and 760 mm. pressure. A t the pressure Russel used, t h e ammonia would be soluble t o t h e extent of approximately 1 g. per I00 g. water. 2 Ibid., p. 181. 0.008-0.009 g . Mg(0H)z per liter of water a t 18' C.
TABLEI AMgs.hTas ammonia FOUND
h- recovered __---
Added DuDlicates .4veraee Mes. Per cent Sone I .26 1.19 1.23 . . . . .... 1.26 .... Sone 1.26 1.26 . . . . 1.26 .... 1.23 None 1.25 . . . . 1.26 . . 1.26 Ii'one 1.26 2.60 2.60 2.25 2 . 6 0 i:is(a! 6 0 . 0 2.60 2 . 6 2 1 . 3 i ( a ) 60.9 2.64 2.25 0.772 , . . , .... .... .... h'one 0.842 . . . . .... . . . . .... None 0.814 . . . . .... .... .... None 0.482 .... .... .... None 2.25 2.32 ... , . , . 1.55 68.0 2.25 2.22 1.38 61.3 64.0 2.25 2.25 1.44 67.6 1.32 2.25 2.36 values the average of the averages of Nos. 1 ,
.
.
I
The results in Table I show t h a t within reasonable limits no ammonia enters into solution after 2 0 min., and t h a t also within the limits of the experiment the amount of ammonia dissolved is independent of the strength of acid. I t is also shown t h a t not nearly all the ammonia which was added was recovered. The rather wide variation of the percentages recovered in the case of Soil 3 is due t o the error in the determination of such a small amount of ammonia as was in this soil. Since the hydrochloric acid extract undoubtedly contains some organic nitrogenous material, i t was of some interest t o subject the residues in some of the distillation flasks t o a second distillation. Accordingly, I j o cc. of ammonia-free water were added t o each of the flasks of Nos. I t o 4 and I j o cc. distilled as before. The results are given in Table 11. TABLEI1 Lab. h-0. Residues from 21 . . . . . . . . . . . . No. 1 2 2 . . . . . . . . . . . . h-0. 2 2 3 . . . . . . . . . . . . h-0. 3 24. . . . . . . . . . . . No. 4
AXIMONIA IK T H E H Y D R O C H L O R I C A C I D E X T R A C T O F S O I L S
T o find out whether the Schloesing method would give reliable results, i t was examined as follows: I-The strength of the hydrochloric acid used for extraction was varied. a-The time of extraction was varied. 3-Known amounts of ammonia as sulfate were added t o t h e soil and i t was then extracted with different strengths of hydrochloric acid for different lengths of time. All of the acid extracts were obtained by shaking in a mechanical shaking machine for the indicated length of time a mixture of one part of air-dry soil with two parts of ammonia-free acid. When ammonia was added, it was done by means of standard ammonium sulfate solution immediately before the addition of acid of suitable strength t o give t h e desired normality. Thus, a n y appreciable bacterial change of the ammonia was precluded. After removing from the shaking * Seidell, "Solubilities of Inorganic and Organic Substances," 1911, p. 17.
Vol. 7, No. 3
i l k s . Ii- as ammonia found 0.28 0.27 0.42 0.39
I t is seen t h a t an appreciable amount of ammonia is given in each instance, and this no doubt comes from a decomposition of organic material. T o further test this point and t o test the applicability of Folin's' method for ammonia t o the hydrochloric acid extract of soils, I O O cc. portions of the same filtrates of which 2 0 0 cc. were used in experiments Nos. I t o 4 were aerated for four hours, after the addition of 4 g. of sodium carbonate, in a n apparatus which will be described later. The results are given in Table 111, and are for I O O g. of soil. For purposes of comparison, the results obtained by distillation with magnesia are repeated here. TABLEI11
From same filtrate Lab. No. as was used in 25 . . . . . . . . . . . . Nos. 1 and 2 2 6 . . . . . . . . . . . . Nos. 3 and 4 2 7 . . . . . . . . . . . . h-os. 5 a n d 6 28. . . . . . . . . . . . Nos. 7 and 8
Mgs. N as ammonia found by Aeration Distillation 1.26 1.23 1.26 1.26 1.26 1.25 1.29 1.26
The aeration was continued for one hour more, b u t no ammonia was recovered. The results obtained by the 1
Z. pkysiol. Ckem., 97 (1902), 161.
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 M I S T R Y
Mar., 1915
aeration method do not vary from those found by distillation by more t h a n t h e experimental error. Owing t o t h e possibility of decomposition with magnesia a t 100' C. one would expect t h e results obtained by aeration t o be slightly lower. The results show t h a t this method can be applied t o t h e hydrochloric acid extract and its desirability over t h e distillation method is recommended both from t h e standpoint of accuracy and ease of manipulation. I n Table IV are given t h e results obtained b y aerating t h e acid filtrates from soils Nos. 2 , 4 a n d j as outlined above. For purposes of comparison, t h e results are computed for I O O g. of soil, although as above they were obtained on I O O cc. of filtrate representing j o g. of soil. Soil Lab. Normality Shaken Min. N O . No. of acid 29 0.2 20 5 30 0.3 20 31 0.2 60 32 0.3 60 33 0.2 20 34 0.3 20 35 0.2 60 36 0.3 60 37 0.2 30 2 38 0.2 30 4 39 0.2 30 40 4.2 30
TABLEIV Mgs. N as ammonia Added Found None 1.10(a) None 1.21 None 1.24 None 1.24 2.25 2.63 2.25 2.63 2.25 2.59 2.25 2.63 None 1.24 2.25 2.80 None 0.53 2.25 1.93
Ammonia recovered Mgs. Per cent
.... .... ....
.... ....
1.40 1.40 1.36 1.40
62: 2(b) 62.2(b 60.4(b 52.2(b
1.56
69.3
1.40
62.2
....
.... ....
The results recorded in Table I V bring out about t h e same points as those in Table I: namely, t h a t t h e amount of ammonia extracted is within t h e limits of the experiment independent of t h e strength of acid or t h e time of extraction a n d t h a t not all of t h e added ammonia is recovered. For t h e soils tested, and these represented quite a range of soil types, we are forced to t h e conclusion t h a t a n y method which depends upon a n examination of t h e hydrochloric acid extract is not reliable. I t might indeed be contended t h a t treatment with fresh portions of hydrochloric acid would finally recover all t h e added ammonia. It is doubted if this point could easily be proven, for no doubt long-continued treatment of t h e soil with even this weak acid would gradually decompose organic matter with t h e production of ammonia. It was not thought worth while t o investigate this point, for even if a method could be elaborated, it probably would not comply with t h e third requirement, i. e . , t h a t of practicability. As t o t h e cause of t h e failure t o recover all t h e added ammonia, any answer would be purGly speculative. I n t h e course of some work which we shall publish later, we have found t h a t t h e use of too rigorous flocculating agents t o clear t h e solution for t h e determination of nitrates in t h e soil causes a decided lowering in t h e amounts of nitrates found. It is just. possible t h a t t h e phenomenon of t h e incomplete recovery of ammonia is due t o t h e flocculating action of t h e acid. OF
TABLBV-MGS. Soil No.
Lab.
SOIL AT ORDINARY PRESSURE W I T H MAGNESIA
It has been shown' t h a t successive distillation with magnesia of such substances as beef, eggs, dried blood a n d cottonseed meal gives for several distillations small b u t appreciable amounts of ammonia, b u t so far as we know, this method has never been critically L Trescott. U. S. Dept. of Ag., Bur. of Chem., Bull. 131, p. 20.
No. 41 42
.... I
n computing these results, the average of the results obtained in xperiments Nos. 30, 31 and 32 was used.
DISTILLATION
examined with regard t o soil. We have accordingly carried out t h e following tests: I O O g. of soil were placed in a copper flask, together with zoo cc. of ammonia-free water, a small piece of paraffin and about I O g. of magnesia a n d distilled until I j o cc. of distillate had been collected in standard acid. The receiving flask was removed a n d another one with a suitable quantity of standard acid was put in its place: I j o cc. of ammonia-free water were then put in t h e copper flask a n d ~ j cc. o of distillate were again collected. This was repeated a third, a n d in the-case of one soil, a fourth time. The results are given in Table V.
....
....
223
48 49 50
1 3.51 3.16 3.79 3.89 2.67 2.67 2.46 2.53 3.16 3.26
N AS AMMONIA Di.jtillation 2 1.54 1.54 1.26 1.26 1.12 1.12 0.98 0.84 0.77 0.77
3 0.91 0.91 0.84 0.84 0.81 0.84 0.70 0.70 0.98 0.91
4
....
.... .... .... .... ....
0.70 0.63
....
....
From t h e results recorded in Table V, i t is apparent t h a t t h e amount of ammonia obtained b y distillation is dependent upon t h e duration of t h e distillation which is, of course, dependent upon t h e amount of water a n d soil used and upon t h e magnitude of t h e heat applied. The results of t h e following experiments emphasize t h e later point in a striking way. For each experiment, I O O g. of soil, zoo cc. of ammonia-free water, a small piece of paraffin a n d I O g. of magnesia were used. The full flame of a large burner was used on each of t h e flasks of Nos. jI, j2, jj a n d 56. T h e flames played directly upon the flasks, while in Nos. j 3 , 54, 5 7 a n d 58, a low flame was used, a n d each flask was protected with a wire gauze. The results are given in Table VI. Soil No.
Lab. No. 51 52 53 54 55 56 57 58
TABLEVI Time, Min. 40
411 _.
150 150 45 40 140 160
Mgs. N as ammonia 2.80 2.67 4.63 4.63 3.37 3.09 4.49 5.34
From a consideration of t h e results in Table VI, although t h e conditions have been made extreme, there seems to be room for doubt as t o whether even comparable results can be obtained b y distillation with magnesia, Although one might regulate t h e gas flame during the distillation of a single series so t h a t each gave about the same heat, yet t h e results obtained on different days or at a different time on t h e same d a y might be appreciably variable. It is, therefore, seen from t h e d a t a presented above a n d from t h e work of others which has been mentioned t h a t there is no very reliable method available for t h e determination of ammonia in soils. T h a t it is highly desirable t o have such a method is apparent from t h e general importance of t h e soil nitrogen problem, a n d particularly in t h e study1 of t h e influence 1 Ehrenburg, Landw. Ztg., 60 (1911), 441 and 479. Von Wlodeck, Jour. Chem. SOC.(Eng.), 102, 2, 85; 0. Lemmermann and L. Fresenius,
Landw. Jahrb., 46 (1913), 127.
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 M I S T R Y
224
of lime and various other substances added t o soil upon its ammonia content; in the s t u d y of t h e ammonia evaporation from soils; in t h e study’ of t h e direct assimilation of ammonia by higher plants and in t h e s t u d y of the ammonia problem in many other practical and theoretical phases. We have, therefore, attempted t o find a method for t h e determination of ammonia in t h e soil which would comply with t h e three requirements named in t h e introduction t o this paper. Our results as given below will show t o what extent we have succeeded. AMMONIA I N T H E S O I L B Y A E R A T I O N
In 1902, Folin2 outlined the method which, with some modifications, is largely used for t h e determination of ammonia in urine a n d various other physiological products. Originally, t h e method involved drawing air a t t h e rate of 600 t o 7 0 0 liters per hr., for I: t o 1 1 / ~ hrs. through 2 ; cc. of urine containing I g. of sodium carbonate and 8 t o I O g. of sodium chloride. From t h e cylinder containing t h e urine t h e air passed through standard acid, which absorbed t h e ammonia. Steel a n d Gies3 found t h a t with urines containing magnesium ammonium phosphate crystals, low results were given because this salt was b u t slowly decomposed b y sodium carbonate. It was later shown4 t h a t all urines upon t h e addition of sodium carbonate gave copious deposits of magnesium ammonium phosphate and, therefore, low results for ammonia. Steel,5 therefore, recommends t h e use of 0.5 g. sodium hydroxide in place of t h e one gram of sodium carbonate. He found t h a t with t h e use of these reagents (sodium hydroxide and sodium chloride) all t h e ammonia was readily driven off from triple phosphate crystals, and t h a t apparently they did n o t decompose any organic materials, as no ammonia was obtained after four hours aeration. He also found t h a t aeration of solutions of each of t h e following substances gave no ammonia: glycine, urea, uric acid, leucine, tyrosine, hippuric acid, guanine, allantoin, creatine and creatinine. Of these compounds, creatinine6 a n d guanine7 have been isolated from t h e soil by such methods as would make i t appear t h a t they were present as such a n d not in combination. Tyrosine8 a n d leucine9 have been isolated from the solution obtained by long-continued boiling of t h e soil with hydrochloric acid. Since t h e aeration methods have been so successfully used in connection with ammonia determinations in urine, which contains a wide variety of organic nitrogenous products, we decided t o make a n attempt t o modify i t for use with soil. Since ammonia has been shown by Russell0 and ourselves t o be rather tenaciously held by t h e soil, t h e stronger reagent used in Steel‘s modification was first tried. I n t h e course of some preliminary investigations, various forms of aeration
* Hutchinson a n d
Miller, Jour. Agr. Sci., 3, 179. 2. physiol. Chem., 37 (1902), 161. 8 Jour. Bid. Chem., 6 (1909), 71. 4 Benedict a n d Osterburg. Biochem. BdZ., 3 (1913). 41. 6 Jour. B i d . Chem., 8 (1910), 365. 6 Shorey, U. S. Dept. of Agr., Bur. of Soils, Bull. 83 (191 1). 7 Lathrop, Jour. A m . Chem. Soc., 34 (1912). 1260. 8 Suzuki, Bull. Coll. Tokyo, 7 (1908), 513. 9 Robinson, Mich. Tech. Bull., 7 (1911). 10 LOC. cit. 2
Vol. 7 , S o . 3
apparatus were tried, b u t the form which was found most satisfactory was similar t o the apparatus used by Kober’ in his “Ammonia Distillation by Aeration” method. We use a 16 02. bottle for t h e absorption bottle, a n d a 500 cc. round-bottomed Kjeldahl flask for t h e aeration flask. The air a n d ammonia enter t h e absorption bottle through a specially made absorption tube. The directions for making this tube are given by F o l i n 2 For any further details as t o the setting up of t h e apparatus, t h e accompanying illustration should be consulted. The titration flask must always be set perpendicular t o the table, and t h e long tube in t h e aeration flask should reach t o within not more t h a n of a n in. of t h e bottom of t h e flask. Both of these last precautions are necessary t o secure adequate stirring of t h e mixture. Only four units are shown in the cut, b u t as many as fourteen determinations in series on one pump have been run. For all t h e work reported in this paper, a current of air of about 2 5 0 liters per hour was used. I n our trial of t h e Steel method, the following technique has been used: The “Alkali” was prepared by saturating ammonia-free water with sodium chloride. In order t o be always sure t h a t a saturated solution
was obtained, the calculated amount of the salt was added t o a known volume of water. Sodium hydroxide was added at t h e rate of 2 g . per I O O cc. of water: 2 j g. of soil were placed in a Kjeldahl flask. The absorption bottle was half-filled with ammonia-free water3 a n d I O cc. of 0 . 0 2 N sulfuric acid added. T h e apparatus was then connected up, the stopper in t h e Kjeldahl flask raised, jo cc. of “Alkali” added, t h e stopper replaced and t h e flask shaken and the suction started. The results are tabulated in Table V I I , and are given for I O O g. of air-dry soil. For purposes of comparison t h e results obtained for the corresponding soils on the first distillation with magnesia are given. TABLSV I 1
Soil Lab. Aeration Mgs. N as NHs No. No. Hrs. by aeration 1 59 4l/2 3.87 GO 4]/2 3.93 61 6 4.49 62 6 4.63 63 a 5.61 5.56 64 8 7.02 65 15 7.30 66 15 67 19 8.70
Soil Lab. Aeration Mgs. N as KO. No. Hrs. Aeration 1 68 19 8.99 3 69 15 4.21 70 15 4.49. 71 19 6.60 72 19 6.46 5 73 15 7.58 74 15 7.86 75 19 9.13 76 19 9.41
NHs b y MgO 3.33
,... ,...
.... 2.67
,... ,...
3.21
,... Since after aerating fifteen hours ammonia was being given off in relatively large amounts, and because t h e 1 2
Jour. A m . Chem. Soc., 36 (1913). 1594. LOG.cit.
T h e reason for using ammonia-free water here was t h a t a sharper end point was obtained with i t t h a n when ordinary distilled water was used in the titration. This was, no *doubt, due to a slight trace of amphoteric organic matter in t h e distilled water. 3
Mar., 1915
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 M I S T R Y
amounts evolved are far in excess of those obtained by distillation with magnesia, there is no other conclusion b u t t h a t there is a decided decomposition of organic nitrogenous matter t o give ammonia. The m7ork of Steel1 shows t h a t probably this excess ammonia is not due t o a decomposition of amino acids. Another class of nitrogenous compounds less stable in the presence of alkalies t h a n the amino acids is t h e acid amides. Kelly and Thompson2 in analyses of t h e alkali extract of nine soils found an average of 14.3 per cent amide nitrogen. J ~ d i d i ,Robinson4 ~ a n d others have found from 2 0 t o 30 per cent of amide nitrogen in t h e solution obtained by boiling the soil for several hours with strong acid. But because of their inherent instability, it seems highly improbable t h a t acid amides are present in soils in much larger, if as large, percentages as ammonia. Many acid amides would slowly decompose in the presence of the Steel reagent. For instance, we found t h a t 0.1 g. acetamide, purified by recrystallization from ether, in two hours aeration with the Steel reagent gave 1.54 mg. ammonia and in four hours 2.56 mg. It seems quite probable, therefore, t h a t a t least part of the excess ammonia is due t o t h e decomposition of acid amides. It was next decided t o t r y t h e use of sodium carbonate as the alkali. Accordingly, as a preliminary test, 0.1 g. of acetamide was dissolved in 50 cc. of water, 2 g. of sodium carbonate added a n d the solution aerated. At the end of two hours, 0.08 mg., and a t t h e end of twenty hours 0 . 0 2 mg. ammonia was given. No doubt t h e production of even t h e 0.08 mg. during t h e first two hours was from a trace of ammonium salts in the acetamide. As is stated above, it has been found t h a t all urines upon t h e addition of sodium carbonate give triple phosphate crystals, and on aeration low results are given for ammonia. Since magnesium, ammonium and phosphate ions are present in the $oil solution, it would seem t h a t the same thing would take place as with urines, namely, low results with sodium carbonate. I n 1908, Folinh found t h a t this objection in the case of urine could be overcome by t h e addition of potassium oxalate t o the solution before the addition of sodium carbonate. If t h e magnesium ammonium phosphate was already there as a precipitate, t h e solution should first be made acid, which would dissolve the crystals and then t h e potassium oxalate added. This same procedure could, of course, be applied directly t o any solution besides urine. We have, therefore, tried t h e use of sodium carbonate with and without t h e use of potassium oxalate on seven soils, and in the case of one soil, we have tried sodium chloride a n d sodium carbonate. The results are all grouped together in Table V I I I , and are for I O O g. of soil. I n all cases, 2 5 g. of soil, 50 cc. of ammonia-free water and about 2 g. of sodium carbonate, a n d where indicated 8 t o I O g. of potassium oxalate a n d 1 5 t o 1 7 g. of sodium chloride were used. Also when potassium 1 LOG.
cit.
Jour. A m . Chern. Soc., 96 (1914). 438. s Mjch. Tech. Bull., 4 (1909); I o w a Research Bull., 1 and 8 (1911). 4 Mich. Tech. Bull., 7 (1911). 6 JOMT. Biol. Chem., 8 (1910). 497. 2
22j
oxalate was used, t h e mixture was always made slightly acid before t h e addition of t h e sodium carbonate. An examination of Table VI11 brings out many interesting points. For the soils tested, sodium carbonate used alone causes no appreciable decomposition of organic matter. It is possible t h a t with a soil high in organic matter, i t might be advantageous t o use sodium chloride, which, of course, acts t o weaken TABLEVIII-ABRATION
METHOD WITH
VARIOUS
Milligi‘ams N a s ammonia
,
FOUND
-
ALKALI RBAGENTS NHa recovered M g s . Per cent
Added
Duplicates
Av.
None 2.25 None 2.25
2.08 2.08 4.29 Lost 1 . 9 7 2.08 4.29 4.27
2.08 4.29 2.03 4.28
2.25
None 2.25
2.08 4.27
1.97 4.32
2.03 4.30
2.27
None 2.25 2.25
2.08 4.32 4.32
2.08 4.27 4.32
2.08 4.30 4.32
2.22 2.24
None 2.25 None 2.25
2.59 4.84 2.53 4.60
2.53 4.78 2.55 4.78
2.56 4.81 2.54 4.69
2.25
None 2.25 2.25
2.59 4.84 4.71
2.59 4.78 7.84
2.59 4.81 4.78
2.22 2.19
None 2.25 None 2.25
1 . 2 3 1.23 3.48 3.54 1.23 1.18 3.48 3.48
1.23 3.51 1.21 3.48
2.28
ioi :3
2.27
100:9
None 2.25 2.25
1.23 3.48 3.48
1.18 3.54 3.48
1.21 3.51 3.48
2.30 2.27
102: 1 100.9
None 2.25 None 2.25
1.39 1 . 3 9 3.59 3 . 6 5 1 . 2 9 1.34 3 . 5 9 3.59
1.39 3.62 1.32 3.59
2.23
.... ....
99.1
2.27
1k:9
None 2.25 2.25
1.34 3.59 3.54
1.34 3.59 3.59
1.34 3.59 3.57
2.25 2.23
None 2.25 None 2.25
2.35 4.44 2.25 4.44
2.19 4.49 2.25 4.44
2.27 4.47 2.25 4.44
2.20
97.8
2.19
97.3
None 2.25 2.25
2.19 4.44 4.49
2.25 4.49 4.49
2.23 4.47 4.49
2.24 2.26
None 2.25 None 2.25
0.79 3.03 0.79 3.03
0.79 3.03 0.84 3.14
0.79 3.03 0.82 3.09
2.24 2.25
None 2.25 2.25
0.90 3.03 3.09
0.79 3.09 3.09
0.85 3.06 3.09
2.21 2.24
None 2.25 h’one 2.25
1.12 3.37 1.23 3.48
1.12 3.26 1.18 3.37
1.12 3.32 1.21 3.43
None 2.25 2.25
1 . 1 2 1.23 3.37 3.37 3.37 3.48
1.18 3.37 3.43
....
2.21
....
.... ....
....
....
2.15
....
.... .... ....
....
*...
.... ....
....
....
.... ....
...
98.2
... 100
16o:9
...
98.7 99.5
... ... 95.6
100
...
98.7 97.3
... ...
100 99.1
...
...
...
99.5 100.5
...
99.5 I
.
.
100
...
98.2 99.5
...
2.20
97.8
2.22
98.7
....
.... 2.19 2.25
...
...
97.3 100
the base. The use of potassium oxalate, it is seen, gives no more ammonia than sodium carbonate alone. This is a t first thought somewhat surprising, but no doubt the reason for this is t h a t t h e phosphate a n d ammonia are present in such high dilution in t h e soil solution t h a t t h e magnesium ammonium phosphate does not precipitate. I n urine’ t h e phosphate content averages about 2000 parts per million, and the ammonia2 1000 parts per million, while in t h e soil King3 found as a maximum amount of water-soluble phosphate about 2 0 parts per million, and we have found as a maxi1 Hammarsten, “Text Book of Physiological Chemistry,” 6th English Ed.. p. 646. 2 Hawk’s “Practical Physiological Chemistry,” 4 t h Ed., p., 313. 8 King, U. S. Deut. of Am., Bur. of Soils, Bull. 26.
T H E J O U R N A L O F I i V D U S T R I A L ,4ND E N G I N E E R I N G C H E M I S T R Y
226
m u m amount of ammonia about 30 parts per million. Since by our method two parts of water t o one part of soil are used, these values should be halved. I t is, therefore, apparent why there is no interference of t h e triple phosphate. From t h e d a t a presented in Table VI11 we feel safe in recommending t h e aeration method with sodium carbonate as t h e alkali f o r ammonia determinations in normal and Iowa alkali soils. T h e method complies with the three requirements laid down as fundamental. Excellent duplicates are given; no appreciable amount of ammonia is given of1 after fifteen hours aeration, and within the experimental error IOO per cent of any added ammonia was always recovered; also the method is eminently practical. For purposes of comparison t h e results obtained by t h e three methods are assembled in Table JX. Soil No. 1
TABLE IX-MGS.NITROGEN AS AMMONIA HC1 extraction MgO distillation Na2COB aeration
...... . . . . . . , . ,
1.25
3.33 3.84 2.67 2.50 3.21
2.03 2.56 1.23 I .39 2.27
It is seen t h a t there is very little co-relation in t h e amounts of ammonia found by t h e three methods, except t h a t extraction with t h e hydrochloric acid gives lower results, while distillation with magnesia gives higher results t h a n t h e aeration method. This is t o be expected as we have shown t h a t hydrochloric acid holds back ammonia, while distillation with magnesia effects some decomposition. Contrary t o our findings, Kelley and McGeorgel report in speaking on t h e ammonia determinations in t h e hydrochloric acid extract, t h a t t h e “results were very similar t o those obtained b y direct [magnesia] distillation.” Table IX emphasizes t h e fact t h a t in reporting ammonia determinations in soil, t h e method used should i n all cases be given. F o r the convenience of those who wish t o use t h e aeration method for ammonia i n soils, we will give t h e technique in detail which we have used. After all t h e apparatus is a t hand, t h e procedure is as follows: Prepare t h e absorption bottles as was described above for t h e Steel method. Weigh z j g. of soil into t h e j o o cc. Kjeldahl flasks, a d d 50 cc. of ammonia-free water a n d a few drops of a heavy oil t o prevent foaming, a n d then p u t in the rubber stopper bearing t h e t u b e for t h e entrance a n d exit of the air. Have t h e end of t h e long glass tube come within not more t h a n ’/s of a n inch of t h e bottom of t h e flask. See t h a t no water is between t h e rubber stopper and t h e mouth of t h e flask t o collect ammonia. The best way t o eliminate this danger is t o have t h e rubber stopper fit very tightly. Now, connect up t h e whole series, b u t d o not start t h e pump. As many as fourteen determinations can be r u n in series. The air before entering t h e system must, of course, be passed through a wash bottle containing dilute sulfuric acid. Xow loosen all t h e rubber stoppers in all t h e Kjeldahl flasks, s t a r t t h e pump, add about two grams of sodium carbonate t o t h e flask closest t o t h e pump, shake t h e flask a n d then insert t h e rubber stopper and then take t h e flasks in succession in just t h e same way. After a 1
Hawaiian Sta., Bull. SO (1913), 31.
Vol. 7 ? No. 3
series is once started, t h e p u m p should not be stopped. If anything happens to a determination, do not t r y t o remove it. After the aeration has run as long as desired, with t h e p u m p still going, remove t h e flasks one by one, starting with the one farthest removed from t h e pump, This same apparatus has been used by us for t h e determination of total nitrogen in soil. The digestion is carried out in t h e usual way, and t h e aeration carried out in much the same way as Kober recommends. A water bath as he advises was not used. but instead, add about one-third of t h e required amount of alkali; shake and allow t h e mixture t o cool and then add the remainder of t h e alkali by t h e same procedure he uses, except t h e absence of t h e water bath. To prevent spattering, t h e Kjeldahl flask is tilted in a plane perpendicular t o t h e line joining t h e absorption bottles. ,4s t o accuracy a n d ease of manipulation, Kober’sl statements have been confirmed. As has been stated, all aeration work reported in this paper is for a current of air of about 2 jo liters per hour. If a pump is available which moves more air t h a n this, no doubt t h e time of aeration could be correspondingly lessened. coscLusIOJY
I--The amount of ammonia extractedby hydrochloric acid is within t h e limits of the experiment independent of the strength of t h e acid and t h e period of extraction. 2I n t h e five soils tested, hydrochloric acid removes approximately from 60 to 7 0 per cent of t h e ammonia added. 3-The Folin aeration method can be advantageously applied directly t o t h e hydrochloric acid extract. 4-The amount of ammonia obtained by distillation of t h e soil directly with magnesia is dependent upon t h e duration of t h e distillation. j-The Steel method of aeration is not suitable for t h e determination of ammonia in soils. 6-The Steel reagents slowly decompose acetamide. 7-The Folin method of aeration is suitable for t h e determination of ammonia in soils, for the same result is obtained mhether the reagent acts for a shorter or longer period; all added ammonia is recovered and t h e method is practicable. 8-In t h e soils tested, there is no interference through formation of triple phosphate 9-For the soils tested, there is no need of using sodium chloride with the sodium carbonate. Ie-Acetamide is not decomposed by 4 per cent sodium carbonate. I I--The results for ammonia obtained by examination of t h e hydrochloric acid extract are lower t h a n those obtained by t h e aeration method. This is due t o a n occlusion of t h e ammonia by t h e soil, t h e nature of which is not clear. I 2-The ammonia results from the direct distillation of t h e soil with magnesia are higher t h a n those obtained by t h e aeration method. The difference is due t o a partial decomposition of t h e organic material by t h e magnesia, which gives ammonia. S O I L CHEMISTRY LABORATORY STATECOLLSGE EXPERIMENT STATION AMES~IOWA LOC. C i t .
‘
