866 DYER AND HAMENCE THE DETERMINATION OF NITROGEN 1 9 The Determination of Nitrogen in Mixed Fertilisers containing Nitrates and Chlorides BY BERNARD DYER D.Sc. F.I.C. AND J. HUBERT HAMENCE PH.D. M.Sc. F.I.C. (Read at the Meeting November 2 1938) THE Jodlbauer modification of the Kjeldahl method (using phenol or salicylic acid) laid down in the Statutory Rules and Orders (1932) of the Fertilisers and Feeding Stuffs Act for the determination of the total nitrogen in mixed fertilisers containing organic ammoniacal and nitric nitrogen is satisfactory in the absence of chlorides. When however the fertiliser contains a substantial quantity of potassium chloride or sodium chloride as is often the case this method is apt to givelow results owing apparently to the loss of some of the nitric nitrogen in the form of nitrosyl chloride which is not trapped by the phenol or salicylic acid even though the operation is carried on in an ice-bath.The following experiments illustrate this somewhat drastically : TABLE I Composition of mixture Nitrogen present Xtrogen found Per Cent. Per Cent. Per Cent. Potassium nitrate 100.0 13-85 13.75 Sodium chloride nil Potassium nitrate 8745 12.12 11.95 Sodium chloride 12.5 Potassium nitrate 75.0 10-38 9.91 Sodium chloride 25-0 Potassium nitrate 50 6-92 5.07 Sodium chloride 50 In the old pre-Kjeldahl days of the soda-lime combustion method of deter-mining nitrogen in fertilisers J. RuHe in 1881 devised a means of dealing with fertilisers containing nitrate by the addition of a liberal quantity of sodium thio-sulphate sulphur and powdered charcoal to the soda lime.This gave correct results even in presence of chlorides but it entailed the use of a large iron com-bustion tube and though satisfactory as far as it went was cumbrous and tedious. The soda-lime combustion method for determining nitrogen has long become a method of antiquity and there are probably few surviving analysts who can look back to it (as one of the authors can) as a matter of daily routine. The need for a modified “wet” method has been realised for some time. The problem resolves itself into a search for a reducing agent capable of con-verting the nitrate present into ammonia before the digestion of the organic matter with sulphuric acid MI s I i IT IITI 1-1 5 I;. 11s cox .r;i I s I s c; N IT ii .i-r 1.:.s .ix L c LO K I DES 8ti7 Combinations o f tlie lvlscli nic'tliocl of nitrate reduction with the Kjeldalil Iiit.tliod have been siigg:.1>stecl biit lia\*e not given satisfactorj. results owing to the I x k of uniformity i n rcdiiccd iron \vliicli (in our sornen.lixt estensive expericnce) Aways contains cnoiigh nitrogen to neccssit;itc. :i considc~rablc correction a c-orrection that moreover cannot lw relied on for any onc portion of the iron. l'reliminary :tlk:tliric~ rediictioii ivitli I)c\.ai-tla nietiil iolloivctI by distillation o f the ammonia formctl a i i t l su1,s;ecliicrit Iijcldahliiig of tlic rcsitluc gives xcurate rwults; but the manipulation in\-ol\wl is c-umbrous ;ind h r c ~ i l ~ s clown when much organic matter is present oxviiig to irotliing in tlic initial (listillation.Xftcr ;L good clcd of t.spc-rinicntal work ;L c-onil)in:it ion of the Devarda and I i jeldahl methods in onc op~r;itioii was :irri\wl at and foiincl to be satisfactory. The app;iratu> einployetl consist s of ;L 5OO-nil. I< jcl(1;ihl flask having a doubly perforated rubber bung fitted with a tap-funnel and ;L U-tubc with bulbs (see lig. 1). Before conirncincing n determination the U-tiibe is charged with 10 ml. of 10 per cent. viv siilpliiiric w i t l and ;L small plug of filter-paper is placed in the to 868 DYER AND HAMENCE THE DETERMINATION OF NITROGEN IN of the tube in order to catch any spray that might be produced during the bubbling of gas through the dilute sulphuric acid. Two g. of the sample and 3 g .of finely powdered Devarda metal are trans-ferred to the Kjeldahl flask and the sides of the flask are washed down with 50 ml. of water. The flask is closed with the rubber bung (with its tap-funnel and absorption tube) and 5 ml. of sodium hydroxide solution of sp.gr. 1-4 (about 500 g. of caustic soda per litre) are added through the tap-funnel. Composition of mixture Per Cent. Potassium nitrate 33-33 Ammonium sulphate 33.33 Mixture A: Sodium chloridi Mixture A Dried blood Mixture A Superphosphate Mixture A Dried blood Superp hos p h at e Mixture A Dried blood Superphosphate Potassium nit rate Dried blood Superp hospha t e Sodium chloride Dried blood Potassium nitrate Superphosphate Sodium chloride 33.33 50.00 60*00 50900 50 *OO 33.33 33.33 33.33 50.00 25.00 25-00 25.00 25.00 25-00 25.00 40.00 15.00 25-00 20.00 TABLE I1 Nitrogen found Jodlbauer New Nitrogen present method method Per Cent.Per Cent. Per Cent. 4861 } 11-68 Nitric Ammoniacal 7-07 10.78 11.68 12.65 12.95 Organic 7-09 Nitric Ammoniacal Nitric 2*31} 3.54 5.85 5.37 5-84 8.47 8.61 Organic 4.72 Nitric 9-27 9.41 Organic 3-54 Nitric Nitric Organic 3-54 Nitric Organic 6-33 7.01 7.63 7-82 The flask is allowed to stand for 30 minutes during which time a vigorous reaction is produced if nitrate is present in substantial quantity. The flask is then heated with a small flame to just short of the boiling-point for a further 1 hour, being gently shaken at intervals.It is then allowed to cool and 20 ml. of 50 per cent. v/v sulphuric acid are added through the tap-funnel in such a manner that the sides of the Kjeldahl flask are well washed by the acid. The flask is again allowed to cool after which the rubber bung is removed and rinsed into the flask, into which the contents of the U-tube are washed. Twenty-five ml. of conc. sulphuric acid are now added the water is boiled off and the residue is Kjeldahle MIXED FERTILISERS CONTAINING NITRATES AND CHLORIDES 869 in the usual way with a drop of mercury as accelerator and the amount of nitrogen is arrived at in one distillation. The total nitrogen in mixtures prepared in the laboratory was determined by the procedure which has just been described and also by the official Jodlbauer process and the results are given in Table 11.These results demonstrate that the total nitrogen may be determined accurately by the new procedure; they also indicate the type of low result that may be obtained by the official Jodlbauer process. CALCIUM CYANAMIDE .-cakium cyanamide is sometimes put on the market with a small admixture of nitrate which is added in order to convert it into a satisfactory granular form. The total nitrogen in such mixtures may be determined satisfactorily by the new method as is shown by the following results (Table 111) : TABLE I11 Composition of mixture Per Cent. Nitrogen present Nitrogen found Per Cent. Per Cent. 19.50 19-85 Potassium nitrate 15.13 Nitric 2-00 } 19.53 Calcium cyanamide 84-87 Organic 17-53 Potassium nitrate 9-67 Nitric 1-34 } 19*89 Calcium cyanamide 90.33 Organic 18.55 It is important that the quantities of reagents given in the description of the method should be adhered to.These quantities have been found to produce the best conditions for the reduction of nitrate to ammonia by Devarda metal in alkaline solution. One interesting observation was made during this investigation of the action of Devarda metal. If insufficient Devarda metal is present to convert all the nitrate into ammonia no initial heat of reaction is experienced and the nitrate is converted mainly into nitrite although on the addition of more Devarda alloy the reaction becomes exothermic and all the nitrate is reduced to ammonia. The nitrogen in sodium and potassium nitrates may be determined satis-factorily by this method but the quantity of Devarda alloy should be propor-tionally larger than with a mixed fertiliser.Operating on 1 g. of nitrate 4 g. of Devarda alloy should be used. The Devarda alloy powder used in our experiments had a fineness of 90 per cent. through a 60-mesh wire sieve. 17 GREAT TOWER STREET Sovember 1938 LONDON E.C.3 DISCUSSION Mr. W. S. CLARK observed that the present-day fertilisers contained very little chloride. He was grateful for having been invited to listen to such an excellent paper which was of great interest both to his firm and to the fertiliser industry. Mr. BAILEY said that in his experience it was not a very frequent occurrence for fertilisers to contain chlorides as well as nitrates. After hearing the authors’ method of dealing with the difficulty he realised how frequently a very simple method proved to be the correct one 870 EVANS A NEW VOLUMETRIC PROCESS FOR VANADIUM Mr.J. R. MICHOLLS said that it was true that the fertiliser and feeding stuffs regulations did not make provision for determining total nitrogen in those infrequent cases in which nitrates and chlorides were present in organic matter. The official methods for the determination of nitrogen present as nitrates were satisfactory in the presence of chlorides and Dr. Dyer had coupled one of these very neatly with the Kjeldahl process in order to obtain the total nitrogen. There was no reason why the other official method-that using reduced iron-should not be followed by the Kjeldahl treatment. As a matter of fact some years ago having a sample in which chlorides were present he had used such a method and found it quite satisfactory. Reduced iron had in the past caused a certain amount of difficulty being irregular in action and containing variable amounts of nitrogen. Nowadays it was possible to obtain satisfactory reduced iron which was practically nitrogen-free. This being so he thought it preferable to use a method whereby reduction was carried out in acid solution thus avoiding the necessity of trapping the ammonia evolved during the alkaline reduction as proposed. Mr. F. L. OKELL said that apart from nitrogen reduced iron contained many other impurities; if however one prepared it oneself it was possible to obtain reduced iron that was quite pure. Mr. G. TAYLOR said that a further objection t o the use of reduced iron was that it was difficult to dissolve the large amount of ferric sulphate that it formed in the Kjeldahl process