Kjeldahl Method for Total Nitrogen P. L. KIRK University of California Medical School, Berkeley, Calif.
The most important unsolved problems of the Kjeldahl method for total nitrogen are concerned with digestion conditions. The quantity of sulfuric acid used in comparison to the size of sample, and the amount of added salt, are factors of importance. Addition of extra sulfuric acid during digestion is often desirable. Mercury is the most effective catalyst for digestion of proteins and usually cannot be satisfactorily replaced by selenium, which must be kept under constant control. Oxidizing agents, except hydrogen peroxide, should be
used only with the greatest care. Length of digestion after clearing requires further attention. The milligram scale Kjeldahl method is recommended for general application. The ultramicromethod, in which the ammonia is separated by diffusion, is convenient and suitable for many small scale uses. Significant developments in reducing the size of sample into the submicrogram range are to be expected, through use of the spectrophotometer equipped with capillary cells, and improvements in spectrophotometric instrumentation.
ANY discoveries of far-reaching importance have been
The fact that the necessary conditions for reduction of nitrogen must coexist with oxidizing conditions for the remainder of the decomposed organic molecules gives an indication of the narrow oxidation-reduction range in which the decomposition must be carried out. Disregard for this fact has led to numerous unsatisfactory modifications in which strong oxidizing agents were used or other conditions were modified unfavorably. A4nydiscussion of the Kjeldahl method must be divided logically into consideration of the digestion, the separation of ammonia (or its omission), and the determination of ammonia. Except as applied to microchemical or other special techniques, the separation and determination of ammonia may be carried out by a number of very satisfactory procedures which are chosen almost solely on the basis of their rapidity, simplicity, or the personal preference of the analyst, rather than on any fundamental chemical basis. The digestion, on the other hand, which has been most investigated, also calls most imperatively for further fundamental and thorough study. I t is indeed probable, as indicated by the extensive work of Friedrich, Kuhaas, and Schnurch ( 9 ) , that no single digestion method of universal application will ever be found, but rather that the details of the digestion technique and conditions will have to be determined very largely by the nature of the material digested. This paper does not review in detail the modifications of digestion and the studies made upon them, but i t summarizes the most important generalizations that can be made on the basis of the large number of empirical studies that are available.
fundamentally extremely simple. The discovery that boiling many nitrogen-containing organic compounds in concentrated sulfuric acid liberates the nitrogen in the form of ammonium sulfate is probably one of the most significant analytical discoveries ever made. Attributed to Kjeldahl, !Tho announced the method in 1883, this method has probably been applied in one modification or another to every possible form of nitrogen, and in perhaps more laboratories than almost any other single type of analytical method. The method was first developed for and applied to the analysis of protein nitrogen, and in the analysis of this material some of the greatest difficulties inherent in the method are still only partially solved. Within the past 5 years, papers by Chibnall, Rees, and Williams ( 5 ) , Jonnard ( I I ) , and others have emphasized the difficulties of obtaining quantitative recovery of the nitrogen from proteins by the Kjeldahl method. Perhaps i t is more surprising that merely cooking one of the most refractory and difficultly manipulated biological materials with concentrated sulfuric acid could be made to yield dependable results than that it fails a t times to recover all of the nitrogen. In spite of the lapse of 65 years since the inception of the method, and the hundreds of papers published on its use, there is little accurate information on the nature of the breakdown taking place in the Kjeldahl digestion when different compounds are treated with hot sulfuric acid, and on the reactions, their mechanisms, or their rates. Hundreds of empirical studies on the effects of catalysts, oxidizing agents, added salts or other acids, and time and conditions of digestion have been published. Few if any accurate studies have been made on the intermediate chemistry of the digestion. Bradstreet ( 2 ) in 1940 extensively reviewed the subject of Kjeldahl analysis, quoting 148 references, most of which were empirical studies of the effects of alteration in digestion combinations. More recently, the application of the Kjeldahl method to the analysis of protein nitrogen was reviewed by Kirk (IS),who listed approximately 90 references, most of which were also concerned with similar empirical studies. All Kjeldahl digestions are subject to four simple facts: 1. Sulfuric acid is a very weak oxidant, which would not be expected to oxidize organic materials rapidly. 2. The conditions existing in the Kjeldahl digeotion are essentially favorable to reduction, as indicated by the fact that the ammonia is not oxidized to nitrogen, and that oxidized forms of nitrogen may be partly or wholly reduced during digestion, almost entirely by organic residues undergoing decomposition. 3. The high temperature of the digestion is conducive to pyrolytic decomposition of many compounds. 4. Because of the rapid removal of water-forming groups by the hot sulfuric acid, charring will always tend to occur.
DIGESTION CONDITIOSS
Amount of Sulfuric Acid Used. This question is related to the amount of potassium sulfate or other salt added to raise the boiling point. At no time must the composition of the mixture approach that of potassium acid sulfate, or ammonia will be lost, in fact, almost quantitatively a t the composition of potassium acid sulfate. This question was investigated extensively by Self (E%), who states that a 15-gram excess of sulfuric acid should remain if 25 ml. were used initially with 10 grams of potassium sulfate. Inasmuch as the utilization of sulfuric acid varies from 7.3 grams for 1gram of carbohydrate to 17.8 grams for 1gram of fat, and 6.7 grams of the acid go to the formation of acid sulfate with 10 grams of potassium sulfate, extra sulfuric acid should often be added in the middle of the digestion, particularly when the latter is long continued, as is frequently recommended for protein analysis. This conclusion is confirmed by various investigators. Amount and Kind of Added Salt. The addition of neutral salt to raise the boiling point, first introduced by Gunning ( I O ) in 1899, is effective in increasing the rate of digestion and is almost universally used. I t can lead to the loss of very significant quantities of ammonia when used in too great excess, or when the
354
V O L U M E 2 2 , NO. 2, F E B R U A R Y 1 9 5 0 amount of sulfuric acid is allowed to diminish. In the Kjeldahl micromethod particularly, the use of added salts has been less common, and the amounts are smaller in proportion to the macromethods. In fact, it is not uncommon to use added salts merely as diluents for the catalyst, in order to keep the amount of the latter conveniently small. Of the various salts used in this manner, potassium sulfate has been shown to be most effective (19) with the possible exception of dipotassium hydrogen phosphate (8). The latter compound must be used with great care, and never alone. When mixed with potassium sulfate, up to a little over half of the total. considerably more rapid digestibns have been shown to result; greater proportions caused loss of nitrogen. This addition is clearly equivalent to the addition of the phosphate in the form of phosphoric acid, provided the total salt content is maintained with potassium sulfate. Phosphoric acid has been used by many investigators, particularly with the Kjeldahl micromethod, with good results when the proportion of phosphoric to sulfuric acid was carefully controlled. The role of the phosphoric acid is obscure, but its beneficial effect appears to be well established, particularly in the digestion of proteins. Catalysts. There is most confusion and lack of agreement, and a most voluminous literature, on the role of catalysts. Osborn and Wilkie ( Z 4 ) , who studied 39 different metals, found mercury superior for protein digestion, followed in order by tellurium, titanium, iron, and copper. Selenium, molybdenum, vanadium, tungsten, and silver were useful under less vigorous conditions. The superiority of mercury over all ot,her single metallic catalyst,s seems well established by the work of Osborn and Krasnitz (E?),hIilbauer ( g o ) , and others. Copper, though xvidely used because of its convenience, is knon-n to be far less effective, though adequate for many purposes. Selenium, introduced by Lauro, has been the cause of most disagreement between investigators. Certain conclusions drawn from the large amount of work that has been done in studying this catalyst seem to be warranted. There is little doubt of the effectiveness of selenium as a catalyst, though there is evidence that its chief effect is in shortening the clearing time of the digest. It has been shown repeatedly that the clearing time is no adequate measure of the digestion time, and recent work (6, 7 , 11, 29) indicates that a digest may clear a long time before the decomposition is complete. There is little doubt that the quantity of selenium must be kept as small as possible, for an excess tends strongly to lead to the ~ O S Rof nitrogen. Selenium apparently causes loss of nitrogen, which increases considerably with the length of digestion time and with the quantity of selenium. The reason for this effect is obscure in spite of studies of the mechanism of selenium catalysis carried out by Sreenivasan and Sadisivan (SO) and more recently by Patel and Sreenivasan (65),who confirm in all respects the above-mentioned effects and recommend against general use of selenium. Maintenance of a high level of sulfuric acid concentration apparently reduces this loss to some degree, as does the use of a combined mercury-selenium catalyst (66). The use of selenium in small quantities in combination with mercury has been approved by a majority of investigators who have.tested it (23, 86,Z9, SO), though the conclusion is far from unanimous. The same can be said for the combination of copper and selenium ( 7 ,2 7 , S l ) . It is doubtful if majority opinion favors a combination of all three catalysts, though many investigators also favor this. Milbauer ( d o ) , who has investigated this question with considerable thoroughness, recommends the combination of mercuric sulfate and selenium in the ratio of 4 to 1, as the best catalyst. This recommendation should be combined with that of Bradstreet ( S ) , who specified an upper limit of 0.25 gram of selenium for a macrodigestion regardless of other conditions of the digestion, and with consideration of the necessity of maintaining an adequate amount of sulfuric acid as compared with both added salt and catalyst. The question of catalysts, though extensively investigated, has rarely if ever been studied in combination with complete control and study of the other factors of the digestion which may be equally important and are probably interrelated with the catalyst effect. Oxidizing Agents. Perhaps no other factor in Kjeldahl digestion has been the subject of as much irrational treatment as oxi-
355 dizing agents. The desire to speed up the digestion by adding an oxidizing agent is understandable, but so is the fact that elementary nitrogen is a more stable form than ammonia nitrogen, and treatment with oxidizing agents may well lead to oxidation of ammonia to nitrogen. Only when there is present a sufficient amount of other reducing agents, which will remove oxidizing agents preferentially as compared with the removal by ammonia or other reduced nitrogen, is the use of an oxidizing agent considered safe. ilny violent oxidizing agent such as persulfate, permanganate, or perchlorate is of dubious applicability (2, 12, 6 4 , S S ) . In the early 1900’s various highly capable investigators, such as Osborn and S$rensen, employed potassium permanganate successfully. Their success may well be a tribute to their chemical judgment, rather than an endorsement of the procedure. Judicious use of such agents in small quantity, and pith digests still rich in carbonaceous reducing agents, is probably a safe procedure. The assumption of such judicious use may well be inapplicable to the routine analyst, who is likely to perform the greater number of Kjeldahl analyses. Of all the oxidizing agents that have been used or recommended, only hydrogen peroxide seems to have received no criticism whatever. Whether this is due to the inherent merits of this reagent or solely to the inadequacy of investigation is still not clear. As generally used, additions of small amounts of this reagent in the Kjeldahl micromethod do not appear to cause lorn results (17, 18, 26). The reagent frequently contains nitrogen in the form of a preservative, and correction for this must be made by means of a properly determined blank, or high values will result ( 1 5 ) . Whether this is a factor in the favorable reports on the use of hydrogen peroxide is not clear, though some investigators (21) so reporting have been aware of this source of extra nitrogen and corrected their results accordingly. The proper method of addition of the reagent with respect to the stage of the digestion appears to be of importance ( d l ) . The study on which this was based contained considerable inherent error and variability, but the conclusions are probably valid because they rest on comparative data treated statistically. The influence of hydrogen peroxide on Kjeldahl digestions should be studied further. Reducing Agents. Oxidized forms of nitrogen require reduction before they may be satisfactorily determined by the Kjeldahl method. There are several satisfactory reducing agents for all but perhaps nitrate nitrogen, and this can with some difficulty be quantitatively reduced also. More difficult is the question of determining certain ring forms of nitrogen, such as pyridine and pyrrol derivatives. These forms of nitrogen appear to respond to reducing agents, but better to combined reducing and hydrolyzing agents such as hydriodic acid. The classical work of Friedrich, Kuhaas, and Schnurch (9) on this subject has probably not been bettered. Jonnard claims that some of the nitrogen of proteins, presumably that contained in tryptophan, histidine, or proline, is not completely recovered without the use of some agent such as hydriodic acid. Although this statement has not been refuted, we can a t this stage a t least hope that this investigator is wrong. Considering the large number of protein analyses that are performed, it would be little less than tragic if each required p r e liminary reduction and hydrolysis before being subjected to the Kjeldahl digestion. Time and Conditions of Digestion. There seems to be little general agreement as to the time necessary for digestion, particularly in the analysis of protein nitrogen. Clearly this factor is modified greatly by the conditions of the digestion, such as the catalyst, and addition of oxidizing agents. Most striking are such recent studies as those of Jonnard and of Chibnall, Reese, and Williams, who find that it is necessary to continue digestion for as much as 8 to 16 hours to obtain maximum yield of nitrogen, and, in fact, to employ this long time even with a stringent set of digestion conditions, including the use of selenium, reducing agents, and added salts. On the other end of the scale are the recent studies of Miller and Miller ( 2 1 ) , who, using nothing but
ANALYTICAL CHEMISTRY
356 stnight sulfuric acid a n d hydrogen peroxide, no catalyst, sslt., n r other aids, claim complete recovery of nitrogen from even the most refrxtoctory proteins in 0 to 15 minutes. The number of clinical methods in which Somc 20 minutes' digestion is used is very considerable, and probablJ- includes most of the clinici analyses porfoormed. Whether thesc methods by virtue of their low accuracy simply fail to show the crrors caused by the short. digestion is a question. Even allob\-ing for this effect,,it remains difficult,to explain the enormous discrepancies between the digcs. t,inn t,imca xilopted hp various investigators.
With qu:mtitios in the milligram range thcre are several ritpid, accurate, and highly satisfactory steam-distillation procedures. Thc author naturally prefers tho equipment and technique daveloped by him (f4),although no chim is made for other than ourclv . "tpchnical advantaees, Tho apparatus shown in Figuro I is made as a singlc pioec of glass, ,,.ith a surrounding the distillittion flask, and an internal o d d finger condenser, which gives s rugged and compact, unit that is newly autumatic in its operation. Figure 2 shovs a convenient type oi mounting for the distillation apparatus. which carries a movahlo burner. reeciver holder on a, rack
ties Co., Berkclcy, Calif.). About 8 minutes are required for thc complete distillation with this equipment, a time somem,hat, in ~ X C O S Sof that specified for some ot,hcr t.ypes of appsratus, bermse of the necessity of heating t,hesmall amount of cold a s t e r to boiling. The simultaneous use of t,xn scts of appsratus by tlro analyst not only cancels this increased time hut increascs considerably the dnily output of nnalgscs over mast other types of equipment, because most of them do not readily a.llou. for operation of more than one unit a t it time. The convenience and the saving in time and reagents of tho rnilligmm scale Kjeldnhl over all Kjeldahl macroprocodulrs me so g r p t that it is difficult far one fsmiliar with bot,h to understand why any Kjeldahl macroan.zlyscsshould ever bc perforrncrl. Thore is no loss in ~ C C U I ~ L C Vand t,ho ezuense of inst:dl:Ltion is u fratinn of th:it of thi
tion Apparatus
V O L U M E 22, NO. 2, F E B R U A R Y 1 9 5 0
351 volume sulfuric acid containing a little copper selenite. Mcrcury catalyst and other combinations of digestion fluid have been used a t times. The digest is neutralized by sodium hydroxide layered under the acid digest. The- small glass cup attached to the stopper carries the standard acid in which the ammonia is collected. The volume of standard acid is only 50 X (1 X == 10-0 liter = 0.001 m1.L and it is contained in the small cun whose rim
~~0
~
~
acid that has condensed on the walls Of the vessel during the digest,ion. Threo hours' diffusion a t room temperature or incubator tomperature or 0.5 hour in vacuo Serves to transfer the ammonia quantitatively to the standard acid in the receiver. The excess ncid is titrated as usual, using a capillary buret which can he read to a sensitivity of 0.02 h. The method is accurate to about 1%and ha!,an approximate lower limit of 0.5 microgram of total amr 11.
Figure 3.
Dig
KJELDAML MICROGRAM ANALYSIS
~MOEint,erestingperhaps than the milligram mnge is t,lieultramicro or. mieraermn ranee of annlvsis. . . in which the nitroecn in the samlde may amount only to 1 01 9 few micmgremq. As early as 1934, x'irk (25)demonstrated that I or 2 microgrsms of ammonia nitrogen could he distilled quant,itatively by a combinn~
~
tion of boiling and seration. Ot,hcr invcstigatars hivc used awnlion to separate very minute amounts ofanmonia, but it. is gene?. ally recognized a t present that, the diffusion technique originated by Conway and Byme ( 8 ) is the most SB.tisfactory method of removing ammonia from a micro-Xjeldahl digest (@j. Thispmcedure is generally used today wherever Iijtddahl nitrogen ultramioroanalysis is performed, particularly in t,he Carlsberg 1,:horztory by Briiel, Halter! Lindcrstrplm-Lang, tmd Razits (41, irhose procedure will detect amounts as smdl ai3 0.005 microgram of ammonia nitmgen and may he used with a total sample ns small 1 m:nr.."-"m TnnLninnl,.. ihn "."+Lnrl o1 ...1v.v6~u -.A. is complex and cxitcting. A total time of about 2 days clapses during the digestion alone, a factor which effeatively crcludes the method from general or rautine use, and makcs it, vnlual,le only for very special research purposes. It has the distinction of representing probably the mast delicate titration procedure so far described, 5s mell as the quantitative transfer of the smnllest quxnbities of ammonia by diffusion. Slightly less sensitive hut more pr;Lct.ieial in general use is the method of Tompkins and Kirk ( S f ) , which has not been applied to less than 0.5 microgram of total nitrogen and has a sensitivity of about 0.01 microgram, or roughly one half the sensitivity of the Carlsberg Laboratory method discussed above. The total elapsed time of a determination is of the order of 4 hours, and with enough equipment a single analyst can complete up to 30 complete analyses in a single working day. "l"
....
..... .. "..-
y..yu
The apparatus consists of a stoppored hubshaped vessel with a tubular extension terminst,ing in small digestion hulh, as shown in Figure 3, which includcs a rubher-stoppered vessel for use a t ordinary pressures and a glass-stoppered vessel for use in vacuum diffusion. The sample i s digested Tvith 0.1 ml. of 50% by
...
"
".
...."..,
-r
"
."
1 .
"I."
ll.y"y."y.
11"
l..r
u,,,I~-
micro or miorogram Iijeldnhl methods, inasmuch as i t is relatively rapid, rcquires a minimum of difficult manipulation, and is aerurate and simple. I t s application to the analysis of proteins and other difficultly digestible forms of nitrogen has not been eomplctely studied, though it has been shown that tho microgram met,hod checked aceumtely TTith the milligram method on the Same protein solutions, and in fact, yielded a superior precisian as compared with themilligram method. It does not a .. m e a r that the lower limits of sensitivitv of microgram lijeldabl analysis have been reached, though i t is probable that the accuracy in the present, range nil1 not be greatly improved by future developments. To do so would require substsntid improvements in tbe aceunicy of the Kjeldahl mncmmethod, because thwe are errom of unknown or only part,ially understood nature in d l Xjeldnhl digestions which mill not bo eliminated by working on 5 smaller scale. It can bo safely predicted that, the limits of analysis mill be lowered by at least u factor of 10 and possihly as much a8 100- to 1000-fold. Such devolopments are now being studied and some progress has been made. The only methods of determining ammonia wliich offer . . ... . . . greauy iiiCreRsefl sonsn~vlayme spocaropnocomet,nc ~n nature. For spectrophotometrv tlic Xe'e~elcrmclhod is not completely suitable becausc of thc colloirlxl form of the colored compound m d other inherent diffioultics. V;m Slyke and Hiller (32), about 1033, and later Borsook ( 1 ) applied the sodium phenate and hypochlorite colorimetric met,hod far determining ammonia r:olorimct~ricall,v, a method which apparently has several points of superiority over the Nessler method, including a t least ns great sensitivity and 5 truly soluble colored pmdueb. Under proper eondit,ions, this color apparently obeys B e d s law. Because t.he spectrophotometer may be made to yield enormous sensitivity by proper design of the absorption cell and proper modifiestion of technique, factors of as much as 500-foId increased sensitivity over standard operation UWP achieved by Kirk, Roaenicls, nnd Hanahm (16), who used cspillsry absorption cells 5 cm. in length nnd containing as little as 160 X of liquid. Further incre:mes in sensitivity are possihle by extending the same principles, and spectrophotomehers capable of sonsitivities a ~ ~ smi~liua~u ~ 1 ~ uperacmn ~ ~~ ~, , . .,. millionfald greater ~, n a, npresena are umnm the realm of possibility. Thcse developments may w e l l make it possible ultimately to determine as litt,leas 0.001r micragram of ammonia nitrogen, or even less, with accuracies of hetter than 5%. If this e m be realized in practice, there is no reason why thesinglelivingcell snditscomponentpartsmaynot besubiect to nnalysis and treatment as an experimental object much as whole animals me now used. Thc possibilities inherent in ihe range of ultramicroenalysisi.,far from being exhausted, are just now beginning to be realized and t,heir application to the field of experimental biology can he only vaguely surmised. One of the most important methods used will certainly be the Kjeldahl procedure
..
..
I~
~
1
~~~I I ~
.
~
.
358
ANALYTICAL CHEMISTRY
and i t is probably one of the first that will be extended into the range of cellular studies because of its tremendous importance in biological investigation. SUMMARY
The present status of the Kjeldahl method leaves much to be desired in spite of its wide use and tremendous value to industry, nutrition, physiology, biochemistry, and medicine. Fundamental research is still needed to elucidate and improve the digestion process for use with different forms of nitrogen. The question of reducing and oxidizing agents other than sulfuric acid, and the matter of the proper catalyst, particularly require further fundamental investigation. Most significant is the rapid development and wide adoption of the micro and ultramicro forms of the Kjeldahl analysis. It is here that many of the most striking and significant developments are still to be expected, particularly in lowering the range of application and applying it to the exploration of minute biological systems, perhaps the living cell itself. LITERATURE CITED
Borsook, H., J . BioZ. Chem., 110, 451 (1935). Bradstreet, R. B., Chem. Rev., 27, 331 (1940). Bradstreet, R. B., IND. ENG.CHEN.,ANAL.ED.,12, 657 (1940). Brael, D., Holter, H., Linderstr$m-Lang, K., and Rbzits, K., Compt. rend. trav. lab. CarZsherg, Ser. chim., 25, No. 13, 289 (1946). Chibnall, A. C., Rees, M.W., and Williams, E. F., Biochem. J . , 37, 354 (1943). Conway, E. J., and Byrne, A, Ibid., 27, 419 (1933). Dalrymple, R. S.,and King, G. B., ISD.ESG.CHEM.,- ~ N . L L . ED., 17, 403 (1945). Friediich, A , , .Uikrochem., 13, 114 (1933).
Friedrich, A . , Kuhaas, E., and Schnurch, R., 2. physiol. Chem., 216, 68 (1933). Gunning, J. W., 2. anal. Chem., 28, 185 (1599). Jonnard, R., IXD.ENG.CHEM.,ANAL.ED., 17, 246 (1945). Kaye, I. A., and Weiner, N.,Ibid., 17, 397 (1945). Kirk, P. L., “Chemical Determination of Proteins,”in “Advances in Protein Chemistry,” Vol. 111, New York, Academic Press, 1947. Kirk, P. L., IND. ENG.CHEM.,ANAL.ED.,8, 223 (1936). Kirk, P. L., Mikrochem., 16, 13 (1934). Kirk,P. L., Rosenfels, R. S., and Hanahan, D. J., ANAL.CHEM., 19, 355 (1947). Koch, F. C., and Meekin, T. L., J . Am. Chem. Soc., 46, 2066 (1924). Lundin, H., Ellburg, J., and Riehm, H., 2. anal. Chem., 102, 161 (1935). Margosches, B. M.,and Vogel, E., Ber., 55B, 1350 (1922). Milbauer, J., 2. anaZ. Chem., 111, 397 (1935). Miller, G. L., and Miller, E. E., ANAL.CHEM.,20, 481 (1948). Needham, J., and Boell, E. J., Bwchem. J . , 33, 149 (1939). Osborn, R. A,, and Krasnitr, A., J . Assoc. Ofic.Agr. Chembte, 16,107 (1933) ; 17,339 (1934). Osborn, R. A,, and Wilkie, J. B., I b i d . , 18, 604 (1935). Patel, S.M., and Sreenivasan, A., ANAL.CHEM.,20, 63 (1948). Poe, C. F., and Schafer, R., J . Dairv Sci., 18, 733 (1936). Schwoegler, E. J., Babler, B. J., and Hurd, L. C., J . BwZ. Chem., 113,749(1936). Self, P. A. W., Pharm. J . , 88,384 (1920). Shirley, R. L.. and Becker, W.W., IND. EXG.CHEX.,ANAL.ED., 17, 437 (1945). Sreenivasan, R., and Sadisivan, V., Ibid., 11, 314 (1939). Tompkins, E. R., and Kirk, P. L., J . BioZ. Chem., 142, 477 (1942). Van Slyke, D. D., and Hiller, A., Ibid., 102, 499 (1933). Wicks, L. E., and Firminger, H. I., ISD. EKG.CHEM.,AXAL. ED.,14,760 (1942). RECEIVEDMay 16, 1949. Presented before the Division of Analytical and I I i c r o Chemistry, Symposium on Determination of Sitrogen in Organic Compounds, a t the 114th Meeting of the A?IIERICASCHEMICAL SOCIETY St. Louis, 110.
Dumas Nitrogen Determination Using Nickel Oxide
‘
WOLFGAYG KIRSTEK, Institute of Medical Chemistry, University of Lippsala, Sweden
HE Dumas nitrogen determination apparatus previously deTscribed by the present author ( 4 ) has now been used in this laboratory and in several industrial and scientific laboratories during a longer period, and experience has indicated improvements in certain constructional details, which lower the risk of breakage and add to the convenience in using the apparatus. A schematic view of the apparatus is shown in Figure 1. I n order to avoid the risk of breakage, most of the tapered ground joints have been replaced by ball and socket ground joints, or flat ground joints, held together by clamps. The ball and socket ground joints were satisfactory when kept in a fixed position during use, as is the case with the joints on the combustion tube. The joints between the four-way stopcock and the nitrometers are sealed together with Kronigs glass cement. tl and tP are flexible metal tubes, connected with the stopcocks through flat ground joints of metal and glass. t B and tr are glass tubes shaped in a U to provide flexibility. sz, a special four-way stopcock, replaces two three-way stopcocks in the old apparatus. A laboratory for mineral oil research, which had to determine small quantities of nitrogen in oil products, wanted a tube filling with a greater oxidation capacity. The part of the tube which is situated in the hot furnace, k , was therefore made wider. As this seemed to have no bad influence on the microanalyses, every spparatus is now provided with this wider tube, which makes it possible to carry out more analyses with one tube filling and gives
more assurance of complete combustion. Samples of 50 mg. of motor oil are burned with good results, using a combustion time of about 20 minutes Although the suitability of the nickel-nickel oxide tube filling may be doubted because of the experiments of Bell (Z), this tube filling has always worked well, in good agreement with the experiments performed by Kurtenacker (6). The part of the quartz tube situated within furnace k must have a wall thickness of at least 1.5 mm. During use a thin layer of nickel silicate is formed on the inside of the tube, which seems to protect the quartz from further chemical attack. Jf the quartz walls are too thin this layer will cause cracking of the tube because of the difference in thermal expansion. The score in the ground joint of the capsule can be dispensed with, as it has no perceptible influence on the blanks and the necessary washing times, and it is unnecessary to have a plug in the capsule during the combustion. Instead of the plug rod 1 is now used, which is taken out of the tube after the capsule is put in place. The capsule and the rod are provided with a better supporting device, the use of which IS shown in Figure 1. Outside the apparatus the capsule is handled with the tweezers, p and I, and the stand, d. The stopper, m, for closing the combustion tube has been redesigned (Figure 1); it should be lubricated with silicone grease. The nitrometers, n1 and w,are more rigidly constructed than formerly. The capillaries leading doR-n into the
