_,_‘.. ‘M —- . ., 'I' Dr] ‘-—:V;,',‘ 4'??? . - .' .L“.1'¥l -tzf.‘\_{.~ mu”; I In. ‘*irv’?3"“‘::‘ 114‘“?!.tffieigpgqfi,5:;.r',',u"T. "CW7” $Wffifi a: ., 'Il . '.'. ‘1 . 3.. ,5 ‘ , s . ' ._ .. . 0 ‘v ,'_ “t. f' . T”I:.:"_(II:_ ‘::.’:’$": :-"::l..}l:"f’f¢\‘ i..'!.- 1;? U. "O‘_’éofi‘!f'..’_dfr. .’ . “ ir;f‘:-é'"?‘t 3:: 1‘? ‘ 140.;b"r'P\-"-“?’I1.:1’&'1;£' .h gum! Ir .".;1,~3’-'£'»=,r$ é'wi'bq'q’i'n’ 3-" n§»$~.,g.':+'f{?vu‘ffl‘ 2"“?4 --7'=“°:‘v’9’.';:-'}‘I< ‘ ‘t " "3 ;-,~"§'n'#:‘ unit ‘. ‘ if“! "’r 1 1-‘35 , .n , ' .~."v'w_ j. ‘ I" 'I: . 5.3;?“ ‘It .I.’.l‘. ’Iniq‘t‘ ' 93. "N“ '92., “1‘ ‘AM‘ :- I’I‘ ' "i ‘9‘. ’}‘.I ’ ’1‘1*.;:?1 l t ' rd“ 1.‘ H . 1.. *- ! “ ~ I J . ' "'11. . - ‘0’. .40"; - Y , ‘1 .1“: i_Wc‘,-. .35 ‘1! . .3 5 I, :{w i H "5. 14“; 6'5.”fi‘g afr?,k"!fl¥’31:°;'9.‘ffuhit“. ”0"“!34‘5“m T‘;"’ \Ifi“ 9"‘Jg‘..{;. 'gév 'L‘;':“’l . . 335“.““"-’f¥‘1”"‘3‘“i '. :'-.‘:"'r' '31 ’ .3513“ “I“ 961' '73" 3 . 39"» “hiM'fithG‘t '.'.';-‘.' ‘51:, .1 'm. “LLB"; r. "5"" .4. ‘w. : T- .,..‘v .I . ‘.' o: ‘0‘1 ‘ ‘l'l‘ulhi‘ \‘ ‘n 5 “.1'!’ a ‘J¢".'."‘.h HIM-1 I! "1‘~"..~:y'&§=:£av. " “‘1. art-75‘.“ ‘ ' new ‘ f .‘ k‘w'm'“? ‘IJr-p-‘I'f‘h'l‘kgq, P 04:. xi 3; A . : 9‘ I, J? «mo 1"" .4} '2‘ "’7" ‘3" “$11.5? ;,’*“‘3“{7 r ",;b‘fiqa ’{5’13‘ 79W": r“; Var-1;_:a,«,:,4,y_!9.‘.i- «Md-""4 345'“ “55‘ fi 5:3. .. ’av‘tf’tf?“ru. ’ '- . .u, '1 ""1 ":E1.~i.-‘3‘3‘-'-" - "as“. w-"-'Ii’r‘ ‘I’sfl -" ’1 " x-‘tn- “I ~;-.~;»¢~ I’f‘ ‘ “"22: 5:6;‘4' .9 32‘ WWW. “w!" %a.'u:.'.8'-'."-sf}f'vé;A; 4m .LW.:}': ”ma-”v3; ~- on?! *3 w? I I... " . "‘OI51:.I ' ‘ I u. "\..- ‘ ' .’ "N" 5”: "I I: if. t, ‘5' ‘.J?¥tb‘. ‘3'?" 'éztn‘t.,.l“ In, 1‘“ Twitrjg‘ I? 5:1}; 'ii‘x‘!’?'kla"zr’.:affl’j‘i?s lap, :. 2‘3“...“ ' ‘Swi *‘~ I Y ". ‘1‘}. ‘2: ii; *1. ‘ .3‘. ~“"’v .1 13“.." o I. ' m --"*l t' ‘ 'W'fiW-‘lnfl‘ gr:- '3! ‘6" if? .3": ft; M""""‘3mm293‘”‘ ' ‘ 2"“ l " ~"(; (3.. ‘-.':’ Nix?» I ."U ‘ 0:1,"? ." ‘3.“ ‘ ‘ I: -. t J m‘l’b “I a}, ‘1‘“ M 1-." :ézm‘.‘ “4.4%“? W t..- -v.. 1' ~. -. 2' m J“ (3" H1 - r awn: ~ .A?»r§;.-:$’»'{6;“‘1»~.zz.w“ “M a» :3. Al. .,. '~ . -'-'.-'"- . 4".“ 5:1"! '! ‘F ' 35““ ”at“: 'u':"‘.':r$ ”$5?“ IIIL‘? H ,‘él‘l '5}: . . 5 {‘ I‘M“ ‘1‘: 7r... . . .., “Na“ p, .1"; “-«m, v ‘1}- ”bum”. ‘ . o. N. h.‘ $3.1“ 3?“; rt:- ‘ J Mm“? . K” ‘9 I‘? Piggy!" '3‘ S‘l“'2.'.~.d , ..‘¢.\.~ k W‘ cm M“. ‘E. ."-'-.\\f.‘-."‘ 7‘. ,‘I'. ‘g. um“ J' .1; ”-1,“:‘32'1 u . " .- ‘._ . a.g "4:": u 9‘ A . “n ,-.‘§N."';'~.-_ 1‘). 9g” :1,?--.‘~x13'..-£.. - \ ' '..‘.‘ ’~ Jr. 4‘. ~ ', 'U. ." "'1‘; ~ £9x.\.ru,' . 0.x. v. ' . ..5 \" V "- . 3 .I...:.~“ V'- ‘?er~ :5 ‘mm" .-.: 95*: ,Q- 03-“ ’1‘... {IL . u ’ - ‘. .1“".'r. ‘ tflr .- \. .N' ;‘ ”.f“;r‘:‘, 5.31 s. s V I . a- v ..«. Hui ‘ {\l’LM. ‘1‘: {.I.\.H..H“H. ’ .‘ . . _ L l o‘ . - .. . . 4 . .fl.-. . . . (r u VIII‘ ‘ . 3r {Jab :ol “' pl THE USE OF SELENIUI‘ COHPOUHDS IN THE IJILDLHL NITROGEN DETERHIHATION A THESIS Submitted to the Faculty of Michigan Btnte College of Agriculture and Applied Science in Pertiel Fulfillment of the Requiremente for the Degree or flute: of Science by Clerk Sherman Chamberlain September 6, 1933 f . ‘ ACKNOWLEDGEMENT C The writer eiehee to expreee hie eincere eppreciet ion to llr. Elmer Leininger, Aeeietant Profeeeor of Chemietry, whoee eble guidance and friendly, helpful eug eetione lade possible the completion of th e project. 94629 ¢ TABLE OF CONTENT - IntrOdUOtion e e e e e e e History of Nitrogen.Determination Original Kjeldahl Process Theory of Reactions in Original Process The Gunning Madifioation e e e e e Reactions of the Gunning Modification Review of WOrk with Selenium . . . Chemistry of Selenium- .Preparation of Selenious Dioxide Used as in this Work 0 Experimental Results . . . . . . . . Figure 1 - Digestion Room Efficiency of Various Catalysts O Catalysts Use of Selenium Compounds Under Adverse Conditions Results on Flour Analysis Results on Milk Analysis . . ' Figu Distillation of Blank Samples. Volatilization.of Selenious Acid . . Figure 3 - Distillation Apparatus Used . . Distillation of NH4Cl Solution Various Catalysts. re 3 - Apparatus for Delivery of NH Cl Distillation of NH4Cl Solution Using Se 3 Distillation of Blank Samples Using 860 Distillation of NH4Cl Solution Using Cu Distillation of Blank Samples Using CuSO4 304 Sample Distillation and Digestion NH4Cl Various Catalysts Three Hour Digestion of Cotton Seed Meal . . . . . Use of Selenium Compounds under Favorable Conditions Determination of Soil Samples. Determinations of Cotton Seed Meal Samples . . Determination of Rice Flour Samples Determination of Commercial Fertilizer Samples Summary Of EXPerimental Work e e e e e e e e e e e Bibliography e e e e e e e e e e e e e e e e e e e Page INTRODUOT ION the determination of nitrogen is one of the most common quantitative determinations made in an analytical laboratory. Because of the importance of nitrogen in all matters pertaining to nutrition, the determinationof this element is one of the most important, if not the most important, to every food analyst. this determination is also of great importance in Soils and Farm Crops labora- tories. The organic chemist is constantly in demand of ‘ the exact amount of nitrogen present in synthetic compounds. Since the value of fert ilisers is based on nitrogen content, we find that fertilizer laboratories are constantly making use of nitrogen determinations. Due .to the universal appli- cation of the Kjeldahl-Gunning-Arnold official method, any change which would shorten the length of the process with- out affecting the accuracy of the determination or increas- ing the cost more than the amount that the labor item is de— creased, would prove a benefit to thousands of analytical chemists. There has been enough work done in the last few years to prove that the use of selenium compounds, materially shor- tens the process. Iith this in mind the author in the work here outlined has endeavored to carry on a detailed investiga- tion using various selenium compounds, in an effort to deter- mine the extent of their efficiency over the official method, and also to determine whether or not the accuracy of the process is affected in any way. To the knowledge of the author this is the first work available as to the use of selenium dioxide as a catalyst in the nitrogen determination. Before dealing with the adaptability of new catalysts it will be beneficial to review the history of the original Kjeldahl process and study its chemistry. HISTORY OF NITROGEN DETERHINATION In the past there have been three different types of methods used. First, the absolute method of Dumas, which involves the dry combustion and reduction of the gaseeous products by a copper foil and measurement of the nitrogen foned. Second, the method of Will and Iarrentrapp in which the material is heated with soda lime and the ammonia formed either titrated or weighed as ammonium platinic chloride. ‘i'hird, the method of Kjeldahl, devised by Kjeldahl in 1883. The first two methods are the oldest. They are highly accurate, however, very time consuming and laborous. The Kjeldahl method with its modifications is so much more simple and short that it is at the present time almost universally employed by commercial chemists. ORIGINAL KJELDAHL PROCESS The process was first introduced in 1883. It con- sisted of the production of ammonium sulphate by moist combustionof the nitrogenous substance with concentrated sulphuric acid. The salt thus formed was subsequently distilled with an excess of caustic alkali. The free ammonia so produced is absorbed'in a known volume of stand- ard acid solution present in excess, the risidual amount being determined by titration. The method as originally introduced was applicable only to nitrogenous organic com- pounds in the absence of nitrates. During the moist combustion in the original method, . dry powdered potassium permanganate was added little by little to the hot liquid in the digestion flask until a permanent green color was produced; the object being to hasten the process of decomposition. It is extremely in- teresting to note that while potassium permanganate is a reagent upon which, to a large extent, the Kjeldahl method was originally based, the process has in the course of time been so modified that the desirability of its use is ex— tremely doubtful, and in some cases it has been found to give low results owing to the destruction of a portion of the ammonia. THEORY OF REACTIONS IN ORIGINAL PROCESS‘ I. Sulphuric acid abstracts from the organic matter the elements of water. II. The sulphur dioxide produced by the action.of the risidual carbon on sulphuric acid exercises a reduc- ing effect on the nitrogenous bodies present. III. From the nitrogenous bodies produced by the above reduction, ammonia is formed‘by the action of an oxidizing body. IV. The ammonia thus fbrmed is at once fixed by. the strong acid as ammonium sulphate. Nitrogenous Body § EéSO4 —-—> (NHA)BSO4 (11345304 «5 uses -.-) N33 a Na2804 s son 1133 4..ch «as 113401 THE GUNNING MODIFICATION Since Kjeldahl's article“I first appeared, a number of investigations have been made by various chemists and a seemingly endless number of articles have been.written.on the subject.f A great deal of good work has'been done, al- ‘Volumetric Analysis 11th. Ed., Sutton, pp. 85. I"’Zeitschrift fur analytische Chemie - 83, 366.(1885) -5- though the results reported in some instances would indi- cate that the experimental work was not of a satisfactory nature. In these articles many modifications are suggested and the use of a great number of chemicals has been inves- tigated, however, only a few of these have introduced features which are of importance to the K3 eldahl method. Of all these the modification of I. I. Gunning introduced in 1889 deserves first mention“. This modification was based upon the observation that in the ordinary KJeldahl process the excess sulphur trioxide in the beginning of the operation soon escapes or unites with water in a form not easily decomposed. During this process the acid diminishes in strength and in this diluted condition the oxidation takes place more slowly. To remedy this difficulty Gunning proposed the use of potassium sulphate. The salt forms with the acid, acid salts which on heating lose water more easily than the acid alone; the acid salts supplement the decomposing and oxidizing power of the sulphuric acid in a valuable Imare Upon heating the mixture of sulphuric acid and potassium sulphate with organic matter, not only the water *Agricultural Analysis Vol. II - Iiley - pp. 370. -5- originally present, but also that which is formed during the oxidation, is driven off without loss of the acid. For this reason instead of the oxidising mixture becoming weaker, the acid becomes stronger, the boiling point of the mixture rises and this combined with the fluidity of the mass, favors the decomposition and oxidation of the ‘ organic matter in a constantly increasing ratio. REACTIONS OF THE GUNNING MODIFICATION The various reactions which take place during the combustion as tabulated by Van Slyke follow". The first reaction to take place is the union of sulphuric acid and potassium sulphate in accordance with the following equation: 1 18 sec ems () o‘fz4—-> o 3 4 Ihen heated the potassium acid sulphate decomposes forming potassium disulphate and water, thus: ' (e) :amo4 ---> 128207|9H30 The potassium disulphate at higher temperatures decomposes . forming normal potassium sulphate and sulphur trioxide, thus: (3) 138207 -—-> 13304 e 803 *Division of Chemistry, Bulletin'35, 1892268 - 7 - .At a sufficiently high temperature the two succeeding re- actions may take place as one, thus: BKHSO ---> K 804 {- 80:5 4 a At the temperature at which these reactions take place the water that is set free does not recombine with the sulphur trioxide nor with the sulphuric acid present in excess, but is expelled from the mixture, hence the mixture becomes more concentrated. 'The sulphur trioxide acts on the organic matter in a powerful manner and the potassium sulphate formed in the last reaction above unites with another molecule of sulphuric acid, and the same round of reactions is repeated continuously as long as there is an.excess of sulphuric acid present. The credit for the introduction.metallic oxides into the process to serve as catalysts is shared Jointly by Arnold? and lilfarth**;copper and mercury oxides have been used commonly. lhen mercuric oxide or metallic mercury is used it is necessary to add sufficient potassium sulfide solution to precipitate the mercury before distillation. This pre- vents loss of ammonia by formation of nonrvolatile mercuro ammonium compounds. "Je ‘e 0e ‘e Ce " 10’ 507 (193?). ‘*Agricultural Analysis, Vol. II, Wiley, pp. 356. -3- REVIEW OF WORK WITH SELENIUM It was first discovered by Lauro’ that selenium and its compounds might be used to advantage as catalysts in the Kjeldahl nitrogen determination. In 1931 Mr. Lauro was working on an entirely different project; the chemistry of the rare earths, when he stumbled upon the fact that metallic selenium might be a good catalyst. He found that it was possible to shorten the period of digestion to one— fifth of the time required in the original method. Also this eliminated the necessity of precipitating the mercury before distillation of the nitrogen. Laure recommends the use of finely powdered metallic selenium rather than selenium oxychloride, as it is much more pleasant to work with. llr. C. 11. Rich" in working with cereals found that the best combination of catalysts for cereal work was metallic cOpper and fuming selenium oxychloride (dispensed by means of a medicine dropper, calibrated to deliver .3 c.c. ). There is a two fold reason for using both catalysts; the combination accelerat es the reaction more than when either is used alone, and for the indicator properties of the copper in turning the solution a deep blue when an excess I'Ind. and Eng. Chem. Anal. Ed. - 3, 4Cl-3 (1931) He Fe Laum , "Modification to the Kjeldahl Method with Selenium“. "Cereal Chemistry - 9, 118-130 (1933) C. E. Rich ‘ . . 'SeOClz in the K3 eldahl Determination“. «9.. of alkali has been added prior to distillation. Mr. Rich found that with this method the solutions were clear in fifteen minutes and digestion was complete in thirty minutes. This constitutes a time saving of thirty-five minutes per sample over the original method. llr. R. M. Sanstedt“ carried out an investigation as to the catalytic properties of metallic selenium, as compared to copper and mercury, working with wheat and bran flour. He made the following conclusions: “Metallic selenium acts more rapidly than metallic copper, and.about the same as mercuric oxide. The digestion with selenium . and with mercuric oxide is complete in thirty minutes, while with copper sixty minutes is necessary. It appears that there is greater danger in losing nitrogen by extremely long digestion with selenium than with the other catalysts. The cost of catalyst per determination was found to be .15 cents as against .48 cents for .7 gram mercuric oxide or about .002 for .1 gram copper“. Mr. Barry C. Messman“ working in an elevator company in.Enid, Oklahoma conceived the idea of using a flux in the *Cereal Chemistry - s, 156-157 (1932) Eh I. Sandstedt . . 'Selenium as a Catalyst in the Kjeldahl Method“. nCereal Chemistry - 9, 357 (1933) Barry C. Messman . 'Hetallic Selenium in the stldahl Method'. Ejeldahl determination made up as follows: 90 parts NaZSO4, 7 parts HgSO4, 1% parts CuSO4, 11} parts powdered Se. He used 8 grams ‘of this flux in'each determination and found that constant and accurate results were obtainable after fifteen minutes digestion. The selenium modification has also found favor in the laboratories of the New York State Hospital'“ where it is necessary to determine the nitrogen content of various aqueous extracts as a method of standardization. By use Of powdered metallic selenium they were able to do the same amount of work in one-half the length of time originally consumed. Results were good. R. A. Osborn and Alexander Krasnitz‘" working in the Bureau of Chemistry and Soils in Washington, contributed additional work on the use of selenium and its compounds. They did considerable work on the effects of the use of selenium on the Kjeldahl determination; Their work was done with 400 watt, calibrated burners, working on the time required to heat the samples after the time of clearing. They concluded that if the samples were heated for a period *Ind. and Eng. Chem. Anal. Ed. - 4, 410 (1932) J. Tennant, H. L. Harrell and A. Stull “Selenium in the Kjeldahl Method“. "J. A. o. A. c. - 16, 110 (1933) R. A. Osborn and Alexander Krasnitz “Comparison of Selenium, Mercury and Copper as Catalysts“. of ten to fifteen minutes after clearing it was sufficient to obtain.good.resulte. Periods of digestion were tried all the way from twenty-five minutes to three hours and it was found that results after three hours were the same as after twenty-five minutes. This disproved the old belief that long periods of heating with selenium resulted in losses in the nitrogen content. They concluded that the use of SeOClz had a slight advantage over CuSO4 but no advantage over HgO. Metallic selenium was found to be more suitable and economical than the 890013. They found that a combination of Se and HgO and CuSO had an advantage 4 over any of the three when used alone. Mr. L. V. Taylor’ of the Missouri Department of Agri- culture, carried out an investigation on the efficiency of metallic selenium,:mercuric oxide, and.a combination of the two. His work was done on animal feeds. He found that ac- curate results were obtained after thirty minutes digestion in the presence of the mercuric oxide-selenium combination. However, with mercury alone accurate results were not obtained until after sixty minute digestion.. He reports that there is no significant differences between the relative efficiencies of mercuric oxide and selenium when each is used alone. *Ind. and E . Chem. Anal. Ed. - 5, 363 (1933) L. V. Tay or, Jr. “Use of a Selenium-Mercuric Oxide Combination in Determina- tion of Nitrogen in Feed Materials“. Iith this brief but comprehensive survey of the work done with selenium compounds, we have all'the information that is available. Thus, we find that the use of selenium compounds is a comparatively new subj ect and is still a matter to be carefully investigated. In the work that is to follow a new compound of selenium, which has, here-to-fore never been reported as a catalyst in the nitrogen determina- tion, is used. It is well to study the chemistry of selenium and its compounds before outlining the work. CHEMISTRY OF SELENIUN The element selenium was discovered by Berzelius in 1817, in the flue dust of a sulphuric acid plant. Selenium belongs in the chemical family of sulphur and tellerium; it is a non-metallic element and exists, as does sulphur, in several allotropic forms: amorphous, crystalline and metallic. A red form may be obtained from reduction of selenious acid (328.03); the black form is obtained by melting the red. Selenium compounds in general resemble those of sulphur in composition as well as in properties. Frequently metallic ' selenium is found in small quantities in natural sulphur, and quite often in various metallic ores. It may be separ- ated from a pulverized ore by treating with hydrochloric acid to dissolve earthy carbonates. The washed and dried residue is then ignited with potassium carbonate and char- -13eie coal, this treatment converts the selenium to potassium selenide which upon treatment with boiling water is dis- solved away from the oxides formed at the same time. The water solution when exposed to the oxygen of the air yields selenium as a grey deposit which may be purified by washing. Another source is in the combustion of seleniferous ‘pyrites in sulphuric acid manufacture. Selenic oxide (Seos) being thereby formed and reduced'by sulphurous acid to free selenium which may be recovered and purified. The most common.source of selenium in America is its recovery from the sludge of COpper slimes. In 1935 the annual yield was 135,000 pounds. Much more could have been produced had there been sufficient demand.for it. During the late World War an important commercial use was found for selenium. Due to the war American importation of manganese was cut off, selenium replaced manganese dioxide (Mnoz) as a glass decolorizer, and it has now been adopted as a standard by the glass industry of the world. Selenium colors glass a rose red.and so may be used to neutralize the green tint of ferrus iron impurities. -14... PREPARATION OF SELENIUM DIOXIDE USED AS CATALYST IN THIS WORK Selenious oxide ($903), the compound featured as a catalyst in this work, is"a white crystalline solid. This compound was prepared by dissolving selenium in boiling BN03 and evaporating the solution.to dryness be- neath an inverted funnel. The selenious oxide sublimes on.the funnel as a white solid, in needle-like crystals. It may also be prepared by burning selenium in a stream of oxygen. Selenium dioxide dissolved in.boiling water yields selenious acid. A $903 + hot HOH -—--> H38903 EXPERIMENTAL RESULTS One of the first things considered at the beginning of this work, was the proper indicator to use in the titra- tion of the excess acid, thus determinging the dissolved ammonia present. After careful consideration methyl red was chosen.to be the most applicable. Methyl red is red in acid solution and lemon yellow in alkaline solution. One small drop will effect a distinct change of color. In this work an alcoholic solution of the indicator was used and it was made of such strength that two or three drops were sufficient for each determination. Since in the standard method the use of either sodium sulphate or’potassium sulphate is Optional, it was desirable to note whether either in its action brought about more speedy digestion than the other. Using the standard.method’ and working on samples of wheat flour, varing the procedure only as to the use of 10 grams of K3804 or 10 grams of Na 80 the follows 3 4 ing results were obtained: [.3804 clear 45 minutes 1.3904 clear 60 minutes These figures give average results of duplicate samples of same weight and same rate of heating. Thus it was found advisable to use 1180 in all work through- 3 4 out this investigation. Throughout this work all digestions were carried out on.digestion.raCks in the nitrogen room pictured on the following page. The system of drawing off fumes did not work as efficiently as might be desired. The first pre-requisite for the use of selenium come pounds is the fact that their use materially shortens the time required.to clear a solution. Results follow giving a comparative study of the efficiencies of various catalysts in clearing solutions of equal weight samples of flour. ’Official Methods of Analysis, 1930, pp. 31. - 15a- Figure 1 - Digestion Rack Used in the Digestion of All Samples. - 16 a Table I - Efficiency of Various Catalysts Number Average Catalysts E 'of clearing ' .5 samples time(min.) .3 gram Metallic Capper 3 45.0 .5 gram 011804 a 31.5 .1 gram Powdered Se 5 16.0 .1 gram Powdered Se ; .5 gram CuSO4 ! 3 13.0 .3 gram 9903' i 5 10.0 .5 gram Rare Earth é a 47.5 i . . .5 gram‘Vzo5 i 3 44.0 .5 gram Cu88 4 13.0 .5 gram Powdered Se I 3 § 11.0 Equivalent guantities; ' Seoz and uSO4 3 * 9.5 Equivalent Quantities; CuSO4 and Powdered Se 3 1 14.5 Equivalent quantities; : ‘ ' ' CuSO4 and 9.03 and Se 2 1 17.0 .2 c.c. 9.0012 g f .3 gram.netal”ic Copper f 9 . 15.7 .3 c.c. SeOCl; 5 7 .5 gram CuSOg. 9 17.? I0 grams [335‘ used in each sampIe. .The results indicate that .3 gram 9.03 reduce the time required to bring solutions clear, more'than any of the other combinations tried. More results throughout this paper will tend to substantiate this primary conclusion. In the case of the use of .5 gram of powdered selenium we find that the time is less than with a .1 gram sample, however, later investigation demonstrated that it was im- practical to use so large an amount. Running determinations on 3 gram samples of wheat flour, it was first noted that there was a tendency to obtain results slightly lower where selenium was used than in the standard method. Distillation of these samples was carried out in the nitrogen laboratory with the block tin condensers. Table 3 — Results on Flour Analysis numb er We ight ' We ight Average Average .of of of clearing per cent samples sample catalyst time min.) nitrogen .zg. 890 k 10 3 grams .5g. CuS 4 15.3 1.33 ‘ . Standard 10 3 grams method 33.6 1.33 .5g. CuSO4 From Table 3 we find that the results with the selenium method are lower by .03 per cent than those with the standard method. 1.33 - 1.30 = .03 The following figures represent the percentage loss over the standard method: .03 r 100 / 1.39 = 1.51% -13... Before attempting an explanation for this error, we will study results on the analysis of a milk sample. In the experimentation with milk samples they were delivered from a calibrated 10 c.c. pipette and the weight _ obtained by multiplying the specific gravity, as determined by lactometer, by the corrected volume of milk delivered. The pipette was allowed to drain for one minute each time it was used thus a uniform volume was always delivered. Table 3 - Results on Milk Sample Number 1 lumber Is ight leight Average Average I of of of clearing per cent samples sample catalyst t ime(m in. ) nitrogen 8 1°. 30 8e e 38s 8.03 13a 5 e 4652 e 33. 3.0 g ‘ 8 10a 30 8e e 58s 0‘1884 14e9 e4713 8 10s 30 8e I e 53e 611304 3°e 6 ! e ‘72? 1 n i From Table 3 we find that the results with the selenium method are lower by .0075 per cent than the re- sults with the standard method: ' .4737 - .4653 9 .0075 This represents a percentage loss of 1.58 per cent over the standard method: ' .0075 x 100 / .4727 = 1.59% a19- Iithout explanation at this point we look at the results on another suple of milk. Table 4 «- Results on Milk Sample Number 3 lumber weight Weight *Average .AveragC- of of of clea ing per cent samples sample catalyst time min.) nitrogen 10 10.30 gs e3 8e 8603 13.0 .4700 9 10.30 80 .5 8e 01184 15e0 .4779 9 1°e3° 8e , e5 8e 011804 SOeO . 47 76 {W __ -.-_-__. From Table 4 we find that the results with the selenium method are lower by .0076 per cent than those with the standard method: .4776 - .4700 8 .0076 This represents a percentage loss of 1.59 per cent over the standard method: .0076 x 100 / .4776 = 1.59% Tables 3, 3 and 4 all show that low results were obtained when selenium was used. This did not seem to be in keeping with results reported by other workers, hence in an effort to determine if this was really an unavoidable error when using selenium the author started working with a standard solution of ammonium chloride. _ 30 - A solution of ammonium chloride (NH401) was made up by carefully weighing out 10 grams of the dried salt and dissolving it in a clean beaker and transferring it into a calibrated 1000 c.c. graduate flask. Care was taken to wash the beaker at least five times so that there would be no danger of loss of ammonium chloride in transferring to the flask. It was then made up to 1000 c.c. and in determinations a 10 c.c. sample was used. ‘ fl . p Calibrated 50 c.c. Pipette ' Calibrated 10 c.c. Pipette K eldahl Flask 5 0 c.c. Erlenmeyer Flask Box shaped to hold round bottom b flasks in verticle position. HUOmb lllll l d Fig. 3 - Apparatus for delivery of E401 sample and for measure of standard acid. -31.. The samples were always measured from a calibrated pipette and allowed to drain into the Kjeldahl flasks from a vertical position. A funnel rack was especially constructed to hold the pipette in a vertical position inside the neck of the flask, so that it might drain each time from a stationary position without touching the edge of the flask. The pipette was allowed to drain for sixty seconds on each sample then the tip was touched to the side of the flask neck and any solution collected in the tip would thus drain to the same point each time. Figure I illustrates the apparatus used for pipette delivery of HH‘Cl samples and also for standard acid. In these samples 10 grams of K3304, roughly weighed, was added with the catalyst which was also weighed.into the Ejeldahl flask; the 10 c.c. of NH4Cl, and 35 c.c. concen- trated.H3804 were added and the mixture was gently warmed to bringrabout a clear solution. When properly cooled the samples were diluted with 360 c.c. of tap water; a pinch of granulated.zinc was added; 80 c.c. 40%»NaOH was carefully added, allowing it to flow slowly down the side of the flask so that it forms a layer in the bottom of the flask without mixing with the acid layer. Flasks were immediately trans- ferred.to distillation racks and 335 c.c. of distillate was collected in Erlenmeyer flasks containing 50 c.c. of care- fully measured standard hydrochloric acid. These distilla- - 33 - tions were carried.out through the block tin condensers in the nitrogen room. The stills were carefully cleaned.be— fore each distillation by washing with boiling distilled water. During the distillation the condenser tips were allowed to dip below the surface of the acid in the receiv- ing flask. Blank determinations were run in.exactly the OMB manner. The results which follow are not as concordant as might be desired, however, sufficient numbers were run so that the average results give a true picture of the condi- tions. During later work it was found that more or less erratic results were due to the use of the block tin conden- sers in the nitrogen room. These condensers were being used. by many students; the processes involved by each was slightly different, hence it was difficult to know Just when the cone densers were clean enough to give accurate results. Some students used paraffin in distilling flasks to eliminate frothing. Upon distillation paraffin collects on the inner surfaces of the block tin;later due to negligence on the part of operators the content of Kjeldahl flasks boils over through the condenser, and into the receiving flasks resulting in strong alkali accumulating on the paraffin which makes it impossible to obtain true results. Table 5a.- Distillation of - 33,. NH Cl Solution Without 4 Digestion Using 8603 as a Catalyst number weight Acid ‘fif Alkali Milligrams of of (c.c.) (c.c.) nitrogen samples catalyst added used recovered 1 .3 g. 8e02 50.00 20.13 25.78 2 I ' - 20.13 26.78 3 u 8 30.36 25.48 4 I a 20.04 25.99 5 . a 19.90 26.32 6 u a 19.92 26.28 7 w a 20.05 25.97 8 u a 20.00 26.09 9 u a 18.92 25.85 10 a a 18.80 25.90 11 n a 18.76 26.34 12’ a a 18.79 26.27 13 - a 18.85 26.13 14 . a 18.75 26.36 15 . . 18.78 g 26.29 16 a - g 18.81 E 26.22 17 a g u g 18.76 . 26.34 18 u i a 18.76 3 26.34 Average milligrams nitrogen recovered - 36.096 Each sample contained 36.183 milligrams of nitrogen. Standard H01 used, samples 1-8 incl. Standard NaOH used, 9-18 1-8 9-18 .10405 Hermal .lOO7l ' .16710 ' .16813 ' -34.. Table 5b - Distillation of Blank Samples Without Digestion Using Seoz as a Catalyst Number Weight Acid.‘ 1 Alkali Milligrams of of (c.c.) (c.c.) samples catalyst added used nitrogen 1' .3 g. 8e08 50.00 31.05 .18 2 I ' I 31.05 .18 3 I I g 31.03 . .23 4 I I 5 31.07 ’ .14 5 I I 2 29.88 .16 6 I I E 29.87 .18 7 i I I § 29.85 .23 8 I I I E 29.86 , .21 9 I I I ' 29.84 § .16 10 I E I | 29.88 .16 Average milligrams nitrogen - .133 no NH‘Ul solution added. Standard 301 used, samples 1-4 incl. - .10405 normal ' 5-10 ' - .10071 ' Standard NaOH ' ' 1-4 ' - .16710 ' ' 5-10 ' - .16813 ' Tables 6a and 6b will show results of the same procedure when Ou804 is used as a catalyst instead.of 8002. Table 6a - Distillation of NH4Cl Solution Without Digestion Us ing CuSO as a Catalyst 4 Number Weight Acid Alkali Milligrams:: of of (c.c.) (c.c.) nitrogen samples catalyst added used recovered 1 .5 g. 0u804 50.00 19.70 26.79 2 I - I 19.90 26.32 3 I I 19.75 26.67 4 I I 19.93 26.25 5 I' I 18.72 26.32 6 I I 18.69 26.50 7 I I 18.63 26.65 8 I I 18.68 26.53 9 I I 18.67 26.55 10 I I 18.41 27.16 11 I I 18.30 27.42 12 I I 18.70 26.48 13 I I 18.80 26.25 14 I I 18.80 26.25 15 I g I 18.78 26.29 16 I I 18.90 26.01 Ln i :verage milligrams nitrggten recovered - 36.532 Each sample contained 36.183 milligrams of nitrogen. Standard H01 used, samples Standard NaOH " , 1.4 incl e 5—16 1-4 ' 5-16 . 10405 No rmal . 10071 . 1671 .16813 - 36 - Table 6b - Distillation of Blank Samples Without Digestion Using CuSO4 as a Catalyst number Weight Acid Alkali ' Milligrams - of of (0.0.) (0.3.) samples catalyst added used ‘ nitrogen 1 .5 g. ouso4 50.00 31.05 E .18 3 I ' “I 31.05 i .18 3 I I 31.00 g .30 4 - I I 30.95 i .43 5 I I 29.87 1 .30 6 I I 29.80 E .35 7 I I 39.75 E .47 8 3 I I E 29.78 i .40 9 . I I E, 29.70 g .58 10 ; I g I E 39.85 E .33 A_Average milligrams nitrogen - .341 lb NH4Cl solution added. 5 Standard 301 used, samples 1-4 incl. .10406 Normal I 5-10 I - .10071 I Standard NaOH I , I 1-4 I - .16710 I I 5.10 I "' a 16813 . - 37 - A study of Tables 5a and 5b, 6a and 6b, reveals the following figures: ‘ Average gross nitrogen recovered with 3302 - 36.096 mg. Average blank with $603 - I ,193 I Actual net recovery nitrogen - 35.903 mg. Average gross nitrogen recovered with CuSO4- 36.53? mg. Average blank with CuSO4 - ’ ,341 I Actual net recovery nitrogen - 36.186 mg. Thus, we find further proof that results with selenium dioxide are lower than results of the standard method. This is in keeping with results as previously reported in Table 3 on flour analysis, and in Tables 3 and 4 on milk analysis. Remembering the equation for the reaction.between selenious oxide and hot water, it was thought that perhaps selenious acid was being distilled. I 8603 4 hot HOH --> 323303 Since selenious acid is a strong acid resembling sulphurcus acid in its reactions, if any was to be passed. over into the receiving flask; it would add to the standard acid already present; upon titration more than the amount of standard alkali actually necessary to titrate the excess acid not used in absorbing the ammonia liberated, would have to be used. This would make the number of mil-equivalents of NaOH used abnormally high and upon deduction from the a 33 - number'of mil-equivalents of HCl present, it would give a figure for the number of mil-equivalents of standard acid used in absorbing the ammonia, which would be ab— normally low, hence upon multiplying by the mil-equiva— lent weight of nitrogen the results would still be ab- normally low. It did not seem possible that selenious acid could volatilize from a solution strongly alkaline with caustic soda, hence further investigations were made. In an effort to determine the amount of selenious acid distilling over the following experiment was outlined: Reagents including 10 grams K3804 — 35 0.0. 33804 and catalyst as indicated in Table 7, were mixed together and digested for twenty minutes; they were cooled, and diluted with 360 0.0. BBB and made alkaline with 80 0.0. NaOH and distilled into receiving flasks containing 50 c.c. distilled HOH and 10 c.c. standard acid. A ‘ Table 7 - Distillation of Blank Samples iimber' Weight wAcid. Alkali ExcessiwaOH Acid of Of (c.c.) (c.c.) (c.c.)=3301 samples catalysts added used acid used used 1 Nona lOe 00 5e 73 ""'- e 36 e 434 3 e5 8e 011804 ' 5e65 "I" ‘ e34 e567 3 .5 g. 8e02 I 5.90 --- . .09 .150 4 .5 g. coca I 5.90 --- i .09 .150 5 ‘3. g. 8602 I ; 6.04 .083 : --. --. 10 c.c. 501:1: 5.99 NaOH -39“ There is always a certain amount of nitrogen present in the HQSO4, tap water, and other reagents, hence the amount indicated'by a blank determination is a measure of the nitrogen present in reagents used. Table 7 indicates that where selenium dioxide is used it is necessary to use more NaOH in titration and the amount of HCl used in dis— solving the ammonia is lower than it actually should be. In sample 5, 8 grams of 8903 was used to see if this would further increase the amount of NaOH used. It was found . necessary to use more'NaOH in titration than was equivalent to the amount of acid added in the beginning. This was not surprising, however, but further confirmed.previous data that perhaps 338603 was distilling into the standard acid. In this case no acid was used in absorbing the nitrogen and there was an excess equivalent to .083 c.c. acid.more than was added in the beginning. It was thought advisable to run a.more detailed investigation along this line. Tables 8a,‘b, c, and d, were carried.out on exactly the same procedure as Table 7. 1 30 - Table 8 - Demonstrating the Volatilization of Selenious Acid fiumber Weight Acid Alkali ‘ "Fab HIA‘cid Milligram§_ of g of' (0.0.) (0.0.) *THCl- (c.c.) samples ! catalysts addedg_ used . used used. nitro en a— I jg. 9903 10.007 8.44 ; .174 .303; .305 3 I I § 8.38 . .334 .373; .413 3 I I ‘ 8.33 .384. .339 .498 4 I I 8.50 .114 .133 .300 5 II ' ' . 8e49 e134 a;:‘4 e318 _ *Average acid used.- .3158 c.c. - Average nitrogen - .3368 mg. b "' 1 e5 e 80 IGeGU Beau 4 I e 6; e 4 3 I I 8.45 .164f .190? .387 4 I I 8.55 .064 .074! .113 5 I I 8.43 .134 .335; .345 6 I l L 8e 4 e 551 e 35 Average acid‘used- .10990.c. Average nitrogen - .166 mg. 0 - l 715 g. CuSO4; 10.0 8.35 §’.4647 .538 .815 3 i . I I 8.37 a .444. .515 .780 3 f I I 8.87 5 .444- .515 .780 4 L. I I 8,35 1g.464; .538 .815 Average as d used - 7536 c.c. “ Average nitrogen - .798 mg. d -—1 None , 10.00 8. 34 I .274 .318 .48—1 " 3 I I 8. 33 j; .384 § .338 .498 3 I I 8.45 ; .164 .190 .387 4 I I 8.33 .394 1.457 .693 5 I I 8.30 I .314 3.364 Average acid use .33 0.0.. Average nitrogen - .503 mg. .551 10 c.c. acid::& 8.614 c.c. alkali _ 31 a Results of Tables 8a, b, c and d may be summarized in the following succinct manner. .5 g. 8803 Average acid used - . Average nitrogen - .5 g. Se Average acid used.- . - Average nitrogen - .5 g. Cu804 Average acid used - . Average nitrogen - no Catalyst Average acid used - Average nitrogen Here, again, we find that results .2158 0.0. .3368 mg. 01099 0000 .166 mg. 0536 0000 0798 mg. .331 0.00 0503 mg. in samples where selenium or selenious oxide were used, are lower than in the samples where capper sulphate or no catalyst were used. The apparent amount of acid used in the case of selenium is found.t0 be lower than that This is due to the fact that .5 grams of Sec lent to .5 grams of selenium. 8e03 --- selenium 79.3 oxygen 33 ITiTé' 79.3 . 100 / 111.3 = .713 . X = .5 . 1 X .703 grams in the case of selenious oxide. 3 is not equiva- 71.2% Se - 33 - Thus it would take .703 grams of 880 to be equivalent in selenium to .5 grams of powdered selfinium. This will ac— count for the difference between the apparent amount of acid.used in samples of’SeO3 and Se. As usual the amount of acid used in samples where CuSO4 was used is greater than in those samples where no catalyst was used. This may be attributed to the fact that the copper sulphate has a small nitrogen.content which necessitates the use of slightly greater amount of acid. These figures may be explained the same as all previous data. Since much difficulty had been experienced in getting good results by using the distilling racks in the nitrogen room, it was decided to make another attempt to obtain more accurate results before drawing final conclusions. At one time during this work it was observed that the use of an excessive amount of zinc, results in erratic re- sults in samples where GuSO4 was used. It was decided that from this point on, all reagents used in determinations should.be weighed. 0ne-tenth.gram of granulated zinc was used in each sample. At this point use of the distilling racks and.block tin condensers was abandoned and a set-up of glass condensers was arranged in my laboratory. The photo on the following page shows details of the distilling set-up which accommodated four distillations at one time. An effec- tive trap was used to prevent a spray of NaOH from passing ll" .. n'M. . ‘ \flt/n- ‘ ;sr .'_v’\_1 .. 7* ,_, _v _ _ Vflf-__‘_r ,v_i,___-A-l!—__. Figure 3 - Glass Condensers Used to Replace Block Tin Condensers for all Distillations in the Latter Part of Experimental Work. «.33.. into the receiving flask. A downeslope from this trap into the condenser was used; thus there was no excess liquid allowed to condense and.to be trapped before reach- ing the condenser. Connection from the tip of the conden- ser dipped below the surface of the liquid in Erlenmeyer flasks. After distillation was complete the stopper COD! necting distilling flask and condenser was disconnected and the entire apparatus was lifted and a rest inserted below the base of the standard; this permitted the tip of delivery tube to be well above the surface of the liquid. A stream of distilled water from a wash bottle was played over this tip then the inside of the condenser was rinsed with distilled water and allowed to drain into receiving flask. After thus making sure to get all dissolved ammonia in contact with standard acid, flasks were removed and in- side sprayed with distilled water just prior to titration. Standard alkali was added from a calibrated burette and.all corrections were made in readings before recording in the data. Many results on determinations carried.out in the nitrogen room were not of sufficient concordance to receive‘ mention in this paper. For a time mossy zinc was used in flasks to prevent bumping during distillation; the pieces were so large that it is Quite probable that as much as one gram was used per determination; this did not affect the selenium results but in the case of copper sulphate,results _ 34 _ seemed to vary quite extensively. This is in keeping with‘ the finds of O. M. Shedd*. He reported, that where large amounts of zinc were used, sodium hydroxide was carried over as a spray in the excessive hydrogen gas generated. This error will occur even when an effective trap is used. He concluded that when cupric sulphate was in solution there was a greater tendency to carry over sodium hydroxide, be- cause in each instance when it was precipitated.by potassium polysulfide, less NaOH was found in the distillate. Shedd recommends the use of not to exceed 100 milligrams of zinc per determination and also the precipitation of either mercury or copper with potassium polysulfide, previous to distillation. ' In the final work befbre drawing conclusions as to the reason for persistant low results where selenium or selenious oxide were used, a solution of ammonium chloride was prepared exactly twice as strong as had been used in foregoing determinations. Thus theoretically each sample contained 53.181 milligrams of nitrogen. Tables 9a, b, c and d are results on distillation of samples without any digestion; reagents were just dissolved then diluted, made alkaline and distilled. An.amount of selenium equivalent to .5 grams of SeOz was used in selenium samples. *J. A. o. A. c. - 10, 507 (1937) Table 9 - Distillation of angel Solution lithout Digestion Inger Height Nitrogen Alkali 11111ng c samples catalyst added added used recovered a'- l .5 g. Seoz 53.181 mg. 49.94 33.50 51.63 3 I I I 33.43 51.74 3 I I I 33.40 51.76 4 I I ' 33.35 51.81 5 I I I 33.35 51.81 6 I I I 33.45 51.66 g I Blank _' 57.83 .183 I I I 56. 3 .38 Average - average blank - mg. , b - 1 @5551}. s 523.191 mg. 49.94 22.49 51.65 3 , , I .‘ I . I 33.46 51.67 3 ' I I 33.41 51.76 __£L. ' _ Blank ' 137.00 _.334 . Average - an . 7 mg, “" e - 1 .5 g. CuSO 53.181 mg. 49.94 33. 33 51.86 a , . I . 33.38 51.93 3 I 33.35 51.97 Average - mm mg. . .. d - 1 .7 g. Hg 53.181 49.94 I 33.18 53.07 a I ., I I 3 93.21 53.19 3 i ' I ' l 2.3030 53.30 4 l ' ' ' ' 56.80 .463 'Average - Blank - . «mg. W _ Standard acid - .13176 N Standard alkali - .1065 H a 35 . Tables 9a, b, c and d may be summarized by the following averages; 8002 Average - Average - Blank - 51.50 mg. Se - Average - Blank - 51.47 " CuSO4 Average - Blank - 51.74 " - 51.69 ' Hg Average - Blank From these tables it is observed that when equiva-. lent amounts of selenium and selenious oxide are used the results are alike in that they are both lower than the standard method by the same amount. In Tables 10a, b, c and d, .5 gram.of sugar was added to each sample. 'This made it necessary to digest the sample until clear. This digestion was carried out at a moderate temperature and samples were all heated for a period.of fifteen minutes after they were clear. If nitrogen were being lost in the selenium and selenious oxide samples during the digestion, we would expect the results in Tables 10a and 10b to be considerably lower than the results in Tables 9a and 9b. If, however, they are not lower we are forced.tc conclude that losses occur during distillation rather than during digestion. n - 37 - Table 10 - Distillation of NH4Cl Solution With Digestion lumber Weight Nitrogen Acid Alkali Milligrams of of milligrams (c.c.) (c.c.) nitrogen samples catalyst added added used recovered ... l .5 g. Seoz 53.181 49.94 33.55 51.53 3 . I ,I I 33.70 51.39 3 I “ I I 33.38 51.77 4 I .I I 33.43 51.70 5 I I. I 33.44 51.68 6 I I I 33.44 51.68 1 I Blank I 57.00 .168 3 '. ' I 56e93 .352 Average -—Average - Blan - 51.40 mg. b - l .3561 g.Se 53.181 49.94 33.38 51.77 3 . I .I _ I . 33.44 51.68 3 I I I 33.37 51.79 4 I I I 33.46 51.65 5 I I I 33.44 51.68 6 I I I 33.43 51.70 1’ I Blank I 53.83 .332 3 I I I 5 . . Average - Average - Blank -‘BIT%5’ mg. c - 1 .5 g. 01180;] 52.191 49.94 22.20 52.04 3 . I A,I . I 33.18 53.08 3 I I I 33.35 51.83 4 I I I '33.18 53.07 5 I I I 33.30 53.04 6 I I I 33.38 53.04 1 I Blank I 57." ..l68 3 I , I I 56.97 .196 Average - Average - Blank - . mg. d - 1 07 Se Hg 53e181 49e94 23e01 53s 33 3 I ._I - I . 33.09 53.31 3 I I I 33.05 53.36 4 I I I 33.01 53.33 5 I I I 31.98 53.36 6 I I I 31.85 53.50 1 I Blank I 56.87 ,.350 3 I I I 56.75 .533 verage - Average - Blank - 51. 9 mg. ‘— W Standard acid - .13176 N. Standard alkali - .1065 N. - 38 _ The following comparison of averages of results from Tables 9 and 10 will show the effect of digestion and distillation as compared with distillation only. Average Net Catalyst Nitrogen Recovery N0 DIGESTION DIGESTION 05 go 8803 51050 mg. 51.40 mg. .3561 g. Se 51.47 I 51.45 I .5 g. CuSO4 51.74 I 51.83 I .7 g. Hg 51.69 I ' 51.89 I The differences as shown here between samples di- gested and samples not digested are so small and insignifi- cant that they may be taken as conclusive evidence that losses occur during distillation rather than during diges- tion. They also prove conclusively that results in samples containing selenium and selenium dioxide are considerably lower than results obtained by the standard methods. Working with cotton seed meal it was decided to try a more lengthy digestion period to see if this would over- come this tendency for low results. A three hour digestion was carried out on the samples summarized in the following Table. _ 39 - Table 13 - Three Heur Digestionggf Cotton Seed Meal —: Number Wei ht Weight Time Acid Alkali Per cent of 0% of to (c.c.) (c.c.) samples catalyst sample clear’ added used nitrogen a - 1 .5 g. Seoz 1.0460 15' 49.94 3.68 6.03 3 I 1.0155 15' I 3.74 6.19 3 I 1.0095 15' , I 3.90 6.31 4 I .8768 13' i I 7.90 6.38 5 I .8581 13' ‘ I 11.10 5.85 6 I .8887 10' l I L_9.64 5.90 Average clearing time - 13' Average per cent nitrogen - 6.07% b - 1 .356 g. Se .9990 18' 49.94 1.80 6.44 3 I 1.0065 30' . I 1.30 6.46 3 I .9734 30' I s 4.88 6.13 4 ' e 8533 12’ II I 9.91 6009 5 . I .8111 15' I i 8.87 6.01 6 I .8058 15' I 3L13.70 5.93 Average oIearTng time - 16 . ST Average per cent nitrogen — 6.17% c - 1 .5 g. GuSO .9335 35' 49.94 3.88 6.71 3 . . ' 1. 0034 30' 500 94 -I'“ 60 81 3 I ' .9878 35' 50.33 .10 6.81 4 ' 08170 85' t 49094 7014 6088 5 I .8333 . 38' i I 6.89 6.88 6 I .8105 38' i. I 7.38 6.89 Average clearing time — 38.5' Average per cent nitrogen - 6.83% Standard acid - .09585 N. Standard alkali - .1084 N. -40... The following averages were obtained from Table 13. SeOB Average per cent nitrogen - 6.07% 3 Average clearing time a 13.0 Se Average per cent nitrogen - 6.17% Average clearing time — 16.6' CuSO4 Average per cent nitrogen - 6.83% Average clearing time — 38.5' These results indicate that more lengthy digestion does not overcome the tendency for low results in the case of selenium and selenious oxide, but that it increases the amount by which they are lower than the standard results. This confirms the findings of other workers on the selenium modification, who have reported that lengthy digestion of selenium samples has a tendency to lower the results. Conclusive evidence, that selenious acid is being carried over into the standard acid in the receiving flask, may be seen in comparing sample 3 in Table 13a, with sample 3 in Table 13c. In sample 3 Table 13c, we find that with a slightly less weight of cotton seed meal, 49.94 c.c. of standard acid is not sufficient to dissolve the ammonia liberated and in- stead of adding standard alkali to neutralize the excess acid which is not present, it is necessary to add more acid, until 50.94 c.c. were used. In sample 3 Table 13a, the weight of cotton seed meal is slightly more hence there is present a -41.. slightly larger amount of nitrogen, however, after dis- tillation there was still present an excess of standard acid equivalent to 3.90 c.c. of standard alkali. Since it is absolutely impossible for the ammonia liberated from 1.0095 grams of cotton seed meal to be dissolved in less standard acid than the ammonia liberated from 1.0034 grams of cotton seed meal, the only conclusion.to follow is that more acid must have been added to the standard acid in the receiving flask. The most logical assumption to be made is that selenium in the form of the dioxide has-been dissolved in hot water and carried.over as selenious acid. $603 + hot water 4...? HzSeO3 After digestion is complete there is always a red covering of selenium around the inside of the neck or the Kieldahl flasks; this is in the form of a scum. It is in? soluble in cold water, as it is not washed down into the flask upon dilution. It can be wiped.out on a piece of absorbent paper and imparts a red color to the paper. A1- ways when the strong alkali was added to flasks prior to distillation it was added carefully down.cne side of the flask; with this manner of addition, strong alkali did not contact all portions of the neck of the flask; however, it was noticed that the portion which it did contact was left colorless, making a break in the red scum Just where this alkali had passed. .. 43 .. It was known that in these experiments much more selenium and selenium dioxide was being used than was necessary to produce the characteristic marked catalytic effect. This was being done intentionally for the purpose of determining the effect of excessive amounts of selenium compounds upon results. Also previous results have shown the effect of lengthy periods of digestion. In all proceeding work selenium compounds have been used under adverse conditions. From this point on the ex- perimental work was outlined in such a manner that the effect of selenium compounds might be studied under favor- able conditions. In samples containing selenious oxide, not to exceed .15 gram per determination was used, and in the samples containing powdered selenium not to exceed .1 gram per determination was used. The strong alkali added, innediately before distillation was carefully poured down the neck of the flask, which was being slowly rotated. In this manner strong alkali contacted all portions or the flask neck. Different lengths of digestion were tried so that we would be able to see how long a digestion was neces- sary to obtain maximum results. This would give an exact comparison of the efficiency of the selenium modification as compared to the standard method, providing the accuracy of the method was not impared. The first material, on which the efficiency of selenium compounds was tested under favorable conditions, was a sample of Brookston loam soil. The Official Method - 43 - for Soil" was used as a check. Approximately 10 gram samples were used, and 35 c.c. sulphuric acid was added per determination. It was impossible to prevent the sainples from'bumping around during the digestion as well as during the distillation. Table 13 - Determination of Soil Samples '_fi- weight Time ' Time Acid. Alkali Per cent Number of I of to of (c.c.‘z (c.c.) samples sample clear digestion adde used nitrogen a - 1 11.7936 18' 30' 49.94 36.39 .331 3 9.9870 18' I . I , 38.34 .341 3 10.6399 18' I I 37.38 .341 #4 10. 7069 18 ' ,__ I I 37.51 . 336 . Average per cent nitrogen - .337? 6 10.9939 18' I . I, 37.15 .335 7 8.5408 30' I I 30.68 .340 8 10.4331 16' I I 39 . l4 . 330 ' Average per cent nitrogen - . , . _ , c - 9 13.3803 18' 130' 49.94 36.30 .333 10 10 .3308 16 ' I I 38 . 60 . 339 11 11.3156 18' I I 37.10 .331 . 13 9.8760 30' I I 38.78 .337 ' Average per cent nitrogen - .33 Weight of Catalyst - .15 gram 8e03 Average clearing time - '18 min. Standard acid - .09585 N. Standard alkali - .1082' N. I'Official Methods of Analysis - 1930 - pp. 7. Table 13 (Cont.) - Determination of Soil Samples EEEEEEL Weight Time Time Aaidf' Alkali Per cent of of to of (0.0.) (c.c.) samples sample clear digestion. added used nitrogen d é_—l 10.3979 20' 30. 49.94 29.62 .212 3 10.6085 33' I I 39.33 .314 3 10.6338 33'- I I . 39.10 .316 4 11.4355 20' I I ‘g27.90 .219___ Average per cent nitrogen - .31‘% e a 5 11.9520 227 60' 49.94 ’ 36.63 .225 6 10.4510 31 I I . 38.50 .338 7 9.2404 21: I I 29.97 .234 8 9.9148 30 I I 39 38 .338 Average per cent nitrogen.- .3 s r 4——9 ”'9.7179 20' 120I 49.94 ‘ 20.65 .227 10 9.0951 30' “I . I 30.60 .338 11 10.4615 33' I I 38.50 .337 ' 13 9.9159 33' I I 9 40 .336 ‘ Average per cent nitrogen.- .3 Weight of Catalyst - .10 gram Se ‘— Average clearing time - 31 min. 3 - B _ 10.0222 60' 607 49.94 30.60 .206 0 10.9027 50' I I , 29.00 .254 D 10.5309 50' I I 39.16 .317 ‘Ins f ciently concordant to average h - 1 5 9.7496 55' 120I 49.94 A 21.00 ’ .229 3 9.3958 55' .I . I . 30.11 .338 4 9.6885 55' I I 39.90 .334 _ Average per cent nitrogen - .3 ‘ i - A 9.5680 A 50' 190I 49.94 30.30 .221 B 9.7185 55' I . I . 39.81 .335 0 10.8361 _ 60' ,_1 I I 27.91 .229 Average per cent nitrogen - .335%fii Weight of Catalyst .- .70 gram Hg Average clearing time - 55 min. Standard acid — Standard alkali - .1083 .09595 n. N. Before discussing the preceding tables on the soil determination it is best to summarize the results in the following concise manner. Time % N. 3% N. % N. at N. Catalyst to 30' 60' 130' 180' clear digestion digestion digestion digestion .15 808.02 18' e337 0231 0339 --._- .10 g. 89* 21' . .215 .229 .227 -_-- .7 . H 54' -4- .lbt 2‘7 .22 6 8 ooncordan ' 3' 5 The first feature to be noticed is that the results in samples where selenium and selenium dioxide were used are as high as the standard method; they also check well with each other. Thus from the work on.soil samples it is apparent that selenium dioxide may be used.to advantage without affecting the accuracy of the determination. It will also be noted that the maximum results are obtained with the selenium dioxide samples after a digestion period of only thirty minutes. The samples determined by the standard method required a period of one-hundredrtwenty minutes before maximum results were obtained. Thus by the use of the selenium dioxide modification.under controlled conditions as outlined it is possible to complete the diges- tion in.one-quarter of the time previously required. Results on the standard method after sixty minutes digestion were not of sufficient concordance to average. This is due to insuffi— cient length of digestion. In samples where powdered selenium was used it is found that maximum results were not obtained until sixty minutes digestion. This is perhaps a bit misleading as it is thought that the optimum.period of digestion is be— tween thirty and sixty minutes. The selenium dioxide samples were clear in eighteen minutes while the selenium samples required twenty-one minutes for clearing, thus it would be expected that a longer time of digestion would'bs necessitated, however, it seems most logical that the optimum time be around forty-five minutes rather than sixty. It is unfortunate that time did not permit a more accurate assertation of the optimum time of digestion. The results on samples determined with selenium dioxide as a catalyst, show a tendency to decrease upon.1ong periods of digestion, however, this is not a very marked decrease. The next step in the outlined procedure was to test the accuracy of the controlled use of selenium compounds on the determination.of cotton seed meal. Blank determinations were made to test for the amount of nitrogen in the reagents ‘used, and it was found that .0003 gram nitrogen was present per sample. However, the amount of nitrogen added in the cotton seed meal per sample was about .067 gram thus the blank is sufficiently small to become negligible in the deal- ing with these figures. Table 14 - Determination of Cotton Seed Meal Samples 3 47 - ___“__ A:— m t .. Number Weight Time Time Acid. Alkali Per cent of of to of (c.c.) (c.c.) lamplea sample clear digestion added used nitrogen 3 .9333 17' I . I . 8.39 3.33: 3 .9531, 17' I I 7. 3 . 3‘_ Average per cent nitrogen - 6.733% 6 .9969 17' I _ I . 5.39 6.691 7 .9136 17' j: I ‘_9.16 6.675 " Average per cent nitrogen - 6.67 a c - 9 .9524 ’5‘ 17' 120I 49.945‘ 7.75 6.627 10 .9299 17' ,I . I . 9.69 6.640 11 .9490 17' I I 9.99 6.592 Average per cent nitrogen - 6.61% . . Weight of Catalyst - .15 gram S903 Average clearing time -'17 min. di- 1 .9955 25: 60' 49.94 4.97 6.780 3 1.0163 35 I . I . 4.88 6.640 9 1.0079 25: I I 4.70 6.720 4 .9578 35 I I 7.40 6.645 Average per cent nitrogen - 6.697%5 '5“ 0 I" 5 09553 35' 130' 490 94 7 e 50 So 648 6 .9630 35' ,I - I . 6.90 6.689 7 .9554 35' I I 7.08 6.713 8 .9969 35' I I __ .64 6.653 Average per.cent nitrogen - 6.678%’. . ‘." Weight of Catalyst - .70 gram Hg Average clearing time — 35 min. A Standard acid ~ Standard alkali - .10711 N. .10845 N. Table 14 (Cont.) - Determination of Cotton Seed Meal Samples —__* ;— Alkali Fae. cent Number Weight 5 Time Time Acid of of to of (c.c.) ,(c.c., samples_gsample_i clear fAdigestiogfitadd601:4; usedv- nitrogen l “'1 .9496 i 20' 90I '49.94 7 7.91 ~ 6.645 3 f .9333 30' I I 1 I ' 7.48 , 6.808 9 3 .9160 20' . I ; I , 9.66 i 6.574 __4 gj .9687 30' I .L, I 8.33 L 6.443 Insufficiently concordant to average 5 .8780 30' 60' 49.94 6.733 6 .9017 30' I I 6.739 7 .9007 30' I 6.758 8 .9519 jfix A I 6.711 Average per cent nitrogen - h - 9 .9456 20' 120I 149.94 7.51 6.714 10 .9377 30' I . I 8.34 6.734 11 .8035 33: I I 14.30 2.63% 13 .9733 3 I I 6.56 .6_ Average per cent nitrogen - 6W . ,— Weight of Catalyst - .10 gram Se Average clearing time - 30 min. 1 I" 1 09708 50' 60. 49094 6.00 6e760 2 .9229 45' I I . 9.19 6.771 3 .9794 40' I I 6.40 6.716 4 .9765 45' I I 5.80 6.767 Average per cent nitrogen - 6.75 3 a 9 .9696 I 45I 3 120I ;49.94 6.90 6.744 10 .9925 I 45' I I . I 7.70 6.777 11 .8537 45' 3 I I 11.70 6.691. 13 ._.8854 45' I I L, I 10.90 6.692 Average per.cent nitrogen - 6. a - Weight of Catalyst - .50 gram CuSO4 Average clearing time —‘ 45 min. —: ‘ __ Standard acid _ Standard alkali - .10711 N. 0 10845 N. ..49 - The following summary of the four methods used in Table 14, again shows the advantage of the selenium dioxide modification. - Time 4 N. % N. 4 N. Catalyst to 30' 60' . 130' clear digestion digestion digestion .15 3. sec; 17 6.792 6.679 6.619 . ' . 20 not’ 6.733 6.699 1 g 88 concordant . .5 g. CuSO4 45 -—- 6.753 6.799 .7 g. Hg 25 --- 6.696 6.676 In samples where selenious oxide was used as a catalyst, maximum results were obtained after a thirty minute digestion. In all other cases maximum results were obtained after a diges- tion period of one hour. Thus the use of selenium dioxide has increased the efficiency of the method 50 per cent without af- fecting the results in any way. The samples in which selenium was used as a catalyst required'but three minutes longer to clear, however, results after thirty minutes digestion were not sufficiently concord? ant to average. It is believed that the optimum period of . digestion in these cases is much less than sixty minutes. Time did.not permit a more accurate determination, however, it is known to lie between thirty and sixty minutes. Slight differences between duplicate samples determined in the same manner may be due to the heterogenioity of the cotton seed meal sample. Table 15 - Determination of Rice Flour Samples. Number Weight Time Time Acid. Alkali Per cent of of to of (0.0. (c.c. samples sample clear digestion added used .nitrogen a _ 1 "" "2.1249 17' 90I 49.94 95.15 1.0120 3 3.3948 17' I . I 34.35 .9975 9 1.9929 17' I I 36.33‘ .9944 4 3.0738 17' I I 35.88 .9843 .'3' .Average per cent nitrogen.- .9975% . . b - 5 2.0420 ‘ 17' 60' 49.94 95.74 + 1.0166' '6 1.9673 17' - I I 36.46 .9937 7 3.3378 17' I I 34.40 1.0130 ___9 2.1709 17' _I I pg94.70 1.0290 ,Average per cent nitrogen - 1.01.% 0 -7“9 1.7291 17I 130. 49.94 99.20 .9754 10 3.1393 17' ,I ..I . 48.35 .068. 11 2.0219 17' I I 49.20 .090 12 32.5957 17' __ggg_,_ _ I 49.30 .070_ * InsTl'Tf cl ently concordant to average Weight of Catalyst .15 gram 8803 Average clearing time - '17 min. 3 3.1694 31' I .I .. 34.97 1.004 3 l. 9313 31' I I 36. 61 1. 005 4 + 2.4799 92' . I w. 92,70 1.021 . ,Average per cent nitrogen - 1.013%“: e «A's 8 2.0599 92' #120! 49.94 95.95 1.099—' 6 1.9790 33' .I . I 96.01 1.021 7 3.1630 33' I ' I 34.65 1.030 ._£L: _3-4013 393 I I -99.90 1.019 "Average per cent nitrogen.- 1.03 %' '5‘ F’ Weight of Catalyst .5 gram CuSO4 Average clearing time -' 31.4 min.’ -===:_ r . _ _ Standard 30 1d. " 0 10711 No Standard alkali - N. .10845 .._._‘-._‘ - 51 a Since all results thus far obtained have proven selenium dioxide to-be a better catalyst than powdered selenium, the use of selenium was discontinued through- out the remainder of the experimental work. A review of the work on rice flour follows: Time % N. 7. N. 7. N. Catalyst to clear digestion digestion digestion 1 not .15 g. $902 17 .9975 1.011 concordant .5 g. 0u90, 91.4I --- 1.019 1.024 It is evident that a thirty minute digestion of the rice flour is not sufficient to give maximum results when selenium dioxide is used, therefore, the optimum time of digestion is betWeen thirty and sixty minutes. The time required to obtain maximum results with the standard method is about two hours or perhaps slightly less. Even though the time required for maximum results with selenium dioxide is more than it has been on other materials tested it is, however, still 50 per cent more efficient than the standard mathOde In the one—hundred-twenty minute digestion of rice flour the results appear very bad. Referring to Table 150 it is observed that sample 9 is slightly lower in nitrogen content than the maximum results. This may be attributed to the general tendency for loss of nitrogen on prolonged digestion where selenium dioxide is used. Samples 10, 11 and 12 all were so reduced in volume of acid that they crystalized on the inner surface of the flasks the instant the heat was turned off. This crystalized.mass was rather insoluble in cold water; so to bring about a solution after dilution it was necessary to heat them again. They were heated in an upright position over a bunson burner, the mouths of the flasks being open. Although all this took place before samples had been made alkaline, the nitrogen was lost in some manner. It is unnecessary to carry the digestion anywhere near two hours in the selenious oxide modification thus this danger will never be encountered in actual determinations. It was thought advisable to test the applicability of the selenious oxide modification on a series of fertil- izer samples. Table 16 will give results of determinations carried out on the cOmmercial fertilizer selling under the name of ILoma'. The certified analysis as given on the box gives the nitrogen content as 5 per cent. Table 16 - Determination of Commercial Fertilizer Samples _‘ 1 Number Weight § Time Time Acid Alkali Per cent of of ' to of (0.0.) (0.0.) samples L sample ;iclear digestion added used nitrogen a a 1 .9590 f 19' 90I 49.94 17.95 5.040 2 .9995 2 19' I I 21. 02 5.124 3 1. 3670 . 13: I I 3 11.01 5.117 4 .9199 19' I I I 22. 05 5.054 . Average per cent nitrogen - 5. 084%‘5 5 1 1.0799 19' I I 19.55 5.049 7 5 .8866 13' I I 19.65 5.081 __s ._ .9955 19' I 5.057 ,g Average per cent nitrogen - o - 9 7’ .9455 i 19' 120I 49.94 17.47 5.114 10 1.3738 g 13’ ,I . I . 7.09 5.058 11 1. 0046 13' I , I 16.83 4.911 12 1.021941 19' I i I 14.95 5.121 Average per cent nitrogen - 5. 05 _‘ weight of Catalyst — .15 gram 9503 Average clearing time - '13 min. d - 1 1.0346 31' 60I 49.94 15.36 4.999 3 1.0135 31' I I 16.97 4.981 9 .9557 21' I I 17.50 4.999 4 1.1942 21' I I 12.20 4.970 Average per cent'nitrogen - 4.98 e - 5 ‘7’ 1.0905 21' j 120I 749.94 15.19 5.090"—' 6 .8440 31' i _I I . 5.033 7 § .9397 31' I 5.016 __8 if .8575 31' I 4.993 Average per cent nitrogen - Weight of Catalyst I- .7 gram Hg Average clearing time - 31 min. Standard acid - .10711 N. Standard alkali .10845 N. €554.- Table 16(Cont.)-Determination of Commercial Fertilizer Samples Number weight Time Ir 71m; fifi‘TAcica Alkali Per cent of of to 4 of °c.0. c.c.,. samples sample clear mdagestionj_added ;93°Q nitrogen to 9 A“ ‘ .9299 27' - 50I 49. 94 19.97 1 5.009 10 1.0767 . 37' ' I i 13.77 » 5.014 11 . 1. 0951 i 27' ‘ I e I , 15.21 i 5.000 12 g: .9955 : 27' .. i I #1 15.90 5.050_ Average per cent nitrogen - 5. 031,0 4“ g — 1 .9140 27' 120I 49:94 22.40 5.022 2 1.1474 27' ,I . I . 10.90 5.099 9 1.0402 27' I I 14.91 5.029 4 1.2909. 27' I I 2.90 5.051 Average per cent nitrogen - 5.54§% . Weight of Catalyst - .5 gram Cu804 Average clearing time -' 37 min. Lr Standard acid - .10711 N. Standard alkali - .10945 N. W The eXperimentation on commercial fertilizer is succinctly reviewed in the following table: Time % N. % N. i N. Catalyst to 30' 60' 130' clear digestion. digestion digestion .7 g. Hg - ‘ 21I 4—4' 4.997 5.0155 5 .. ‘05 :2 011334 7 27. """ 5. 031 50049 It is evident that the use of selenious oxide is appli- cable, and highly efficient in fertilizer determination. Maxi- mum results are obtained after thirty minutes digestion while ~in the standard mercury and standard 00pper sulphate methods the maximum results are not obtained until after a two hour di- gestion. Thus the use of the selenious oxide modification per~ mits completion.of digestion in 35 per cent of the time required by the standard methods. - 55 _ S UM DIARY The use of potassium sulphate is recommended over sodium sulphate as a salt to raise the boiling point of acid solution. It was found the use of 10 grams of potas- sium sulphate per determination would bring about a more speedy digestion than an equal amount of sodium sulphate. ’ It was found advisable to use 10 grams per determination. A study of the efficiency of various catalysts and combinations of catalysts, revealed selenium dioxide to be the most efficient and'best adapted to the purpose. It is 'recommended that never more than .15 gram be used. With the use of .15 gram the length of digestion time required to give maximum nitrogen results is only 35 to 33 per cent of that required with the standard copper sulphate, or standard mer— cury methods. The cost of this amount of selenium dioxide is much less than the cost of .7 gram of mercury, required in the standard method. The use of selenium dioxide permits completion of determinations in.one-third to one-half the time originally required. All experimental data on the use of excessive amounts of selenium dioxide portrays the fact that low results are obtained. A possible explanation as gathered from experi- mental results, hinges on the fact that during digestion a red scummy layer of selenium collects on the neck of the Kjeldahl flask. The more selenium dioxide that is used the more of this red deposit will be formed. Upon distilla- tion hot water in the form of steam passes over this selenium dust and a certain amount of it is dissolved and carried over into the standard acid in the receiving flask. The selenium dioxide reacts with hot water forming selenious acid, an acid resembling sulphurcus acid in its reactions and nature. 8502 + hot H30 -——4 H38903 This adds to the acid already present thus upon titration the amount of standard sodium hydroxide used will be more than the amount actually necessary to titrate the excess acid not used in dissolving the ammonia liberated. This causes results to be lower in samples where excessive amounts of selenium dioxide were used. In order to avoid this error it is recommended that not to exceed .15 gram of selenium dioxide, or if selenium powder is used, not to exceed .1 gram, be used. As the samples are made alkaline just prior to distillation, the strong caustic soda should be added carefully down the neck of the flask; as it is being added the flask should be care- fully rotated so that all portions cf the neck of the flask have been in contact with the strong alkali. If these pre- cautions are followed no selenious acid will be formed. The results obtained by the selenium dioxide modification were - 57 - found to check as well as did the results with standard methods. All experimental work performed under favorable conditions as above outlined gave results comparable to standard method results. It is recommended that no paraffin be used during distillation of samples. If paraffin is used to prevent frothing some of it is carried over into the condenser and solidifies on the inside. If any samples are then allowed to boil over through condenser or if any acid from receiv- ing flask is drawn back into the condenser, foreign matter in the form of alkali or acid is incorporated in the paraffin; this is hard to clean especially where block tin condensers are used. Repeated distillations may wash off enough of this foreign matter to make results non concordant. Where excessive amounts of zinc were used during dis- tillation, poor results were obtained especially in samples determined with copper sulphate as a catalyst. This confirms the work of 0. M. Shedd whose work was referred to on page 34. Even though an effective trap is used small amounts of sodium hydroxide are carried over into the receiving flask in the ex- cessive hydrogen gas generation.. Thus it is recommended that not to exceed 100 milligrams of zinc be used per determination. All experimental data has proven selenium dioxide to be a better catalyst than powdered metallic selenium. The time required to clear solutions is less and the total time of di- gestion reQuired to obtain maximum results is considerably less. Hence, since it is no more expensive, its use is advocated in preference to powdered selenium. In the selenium dioxide modification it is never necessary to carry the digestion longer than sixty minutes. A long digestion tends to decrease the amount of nitrogen, hence lengthy digestions with selenium dioxide must not be made . There is an optimum length of digestion time for each substance tested, at this point maximum results are obtained. Regardless of the catalyst used this point should be carefully determined for the material to be tested before a series of samples are run. It will be found that this optimum diges- tion time where selenium dioxide is used as a catalyst will be as small as 30 to 33 per cent of the time reQuired in the standard method. Hence it is recommended, that (l) in the use of selenium dioxide the optimum time of digestion for each substance tested be carefully determined, and then (3) that digestion never be continued longer than this time. The time saving will be great enough to more than make up for the extra trouble of locating the optimum digestion time. After completion of experimental work in connection with the use of selenium dioxide in the Kjeldahl Nitrogen Determination, its use under favorable conditions is most _ 59 - heartily recommended. Distillations may be carried out more evenly with less bumping or frothing, and liquid during distillation is clear and transparent; much less time is consumed in a complete determination and the cost of the catalyst is so small that it may be considered negligible. It was found that determinations were more pleasantly performed with the selenium dioxide modification than with the standard method. Ind. Ind. Ind. Ind. Ind. B I B L I 0 G R A P H Y Use of Selenium and Eng. Chem. Anal. Ed. - 9, 401-2 (1991) Me Fe LauI'O IModification to the Kjeldahl Method with Selenium". and Eng. Chem. Anal. Ed.-- 4, 410 (1992) J. Tennant, H. L. Harrell, and A. Stull ISelenium in the Kjeldahl Method“. and Eng. Chem. Anal. Ed. - 4, 914 (1992) E. S. West and A. L. Brandon "Use of Selenium in the Kjeldahl Method". and Eng. Chem. Anal. Ed. - 5, 250 (1999) Floyd Ervin Kurtz "Selenium in the Determination of Phosphorus and Nitrogen in Phospholipides". and Eng. Chem. Anal. Ed. - 5, 259 (1999) L. V. Taylor, Jr. IUse of a Selenium-Mercuric Oxide Combination in Determination of Nitrogen in Feed.MaterialsI. Cereal Chemistry - 9, 118-130 (1933) C. E. Rich ISeOClz in the Kjeldahl Determination“. Cereal Chemistry - 9, 155-157 (1992) R. M. Sandstedt ISelenium as a Catalyst in the Kjeldahl MethodI. Cereal Chemistry - 9, 357 (1933) Harry G. Messman IMetallic Selenium in the Kjeldahl Method“. Journal Association Official Agricultural Chemists - 16, 110 (1933) R. A. Osborn and.Alexandria Krasnitz IComparisonof Selenium, Mercury and Copper as Catalysts". -61... B I B L I 0 G R A P 'H Y Kjeldahl Method Zeitschrift fur analytische Chemie - 33, 366 (1883) IOriginal Method of KjeldahlI. Agricultural Analysis by Wiley - pp. 349 IThe Method of Kjeldahl'. Volumetric Analysis by Sutten.- 11 Ed. — pp. 87 IThe Kjeldahl Nitrogen Determination“. Methods of Analysis - A.O.A.C. - pp. 21 (1990) IThe Gunning - Arnold Modification". American Association Cereal Chemists - (1938) IMethods of Analysis of Cereal and Cereal Products". Standard Methods of the Association Official Agricultural Chemists, Second Edition - (1925) Chemical Analyst - 20, No. 5, 19-20 (1991) F. Marsh and W. Lepper ”The Introduction of Copper Sulphate as a Catalyst”. Journal Association Official Agricultural Chemists - 16, 107 (1933) R. A. Osborn and Alexandria Krasnitz “A Study of the Kjeldahl Method“. Journal Association Official Agricultural Chemists - 16, 316 (1933) A. E. Paul and E. H. Berry IKjeldahl Method and It's Modifications”. Journal Association Official Agricultural Chemists - 10, 507 (1937) '0. M. Shedd IEffect of Rapid Boiling to Shorten the Digestion Period“. Journal Association Official Agricultural Chemists - 9, 455 (1925) IA Comparative Study of Kjeldahl-Gunning-Arnold Method and the L. W. Winkler Boric Acid Modification". Journal Association.0fficial Agricultural Chemists-5, 108 (1931—32) A. E. Paul and E. H. Berry IThe Kjeldahl Nitrogen Method and It's Modifications“. - ~nv- — -v-._. .- -.—~—‘1‘v —r— -- 1.- - m‘w .— T545 7545 0449 ' 94529 Chamberlain The use of selenium com- T“f\1‘Ih/Qfi 1°” +kn 04-1 4-1.1 0443 94639 Chamberlain 2 \\\\\\\\\\\\\\\\\\\\\ M2: 873 \ \ 02 \ \ II“ \‘I‘I \‘I‘\II \ ‘I‘I“ \| \‘I‘II “\I|Il I:‘|l\l‘| \ \ |II“I‘ \I‘l‘.‘|| 3 9 m0... 3 \\\ \ mm