T H E QUANTITATIVE DETERMINATION O F SELENIUM IN ORGANIC COMPOUNDS. BY R.
E. L Y O N S
AND
Received August
F.
L.S H I N N .
20. 1902.
?'HE difficulties encountered in the estimation of selenium in organic compounds led t o a study of the analytical methods in use. W e have found these methods impracticable in their application to selenium compounds which are to be decomposed with nitric acid. T h e results obtained were invariably too low. I t is stated in the literature,' and has been verified in this laboratory, that when metallic selenium is oxidized by nitric acid, the product is always selenious acid without a trace of selenic acid. This statement holds good in the case of organic compounds containing selenium when they are heated with red nitric acid in a sealed tube for several hours at temperatures between 180' and 260" C. Advantage has been taken of this deportment for the determination of selenium by methods based upon the reducibility of selenious acid by sulphur dioxide, hydrogen sulphide, sodium thiosulphate, hydriodic acid, etc. The determination of selenium in solutions of selenious acid free from nitric acid is comparatively easy. Several of the methods proposed permit of a high degree of accuracy. The two following have best served our purpose : ( I ) The old gravimetric method of Rose,? which Fonsists in reducing the selenious acid, in the presence of hydrochloric acid, by means of sulphur dioxide, collecting the precipitate on a weighed paper, drying at 100' C., and weighing. ( 2 ) T h e iodometric method of Norris and Fay,a whereby the selenious acid is reduced, in the presence of hydrochloric acid, by an excess of 9/10 sodium thiosulphate solution and the excess determined by titrating back with a standard iodine solution. Both methods when properly executed yield excellent results. The volumetric process is preferable when a number of estimations are to be made. I t is simple, rapid and accurate. Other methods deserving mention are those of Gooch and 1 Wohlwill : A n n . Chon. Phavm.. 114, 173 ; Dammer: Band I, 683: Roscoe and Schorlemmer : Vol. I, 362. 2 Pogg. A n n . , 113, 472 ; Zlschr. anal. Chem., I, 73. 3 A m . Chcm.J., 1 8 , 7 0 3 ; 23, 119.
IOY8
R . E . 1,SOSS .\SI)F . I.. S H I S S .
1'iercc.I antl llnthnian and Scliaeffer.' T h e latter niethotl, lio\vever, has been studied by Gooch and and found to be i I i acc 11rat e'. Xof these nietliotls are directly applicable t o solutions of selenious acid containing free nitric acid, a condition obtaining froin tlic customary osidation c)f seleniuiii and its compounds with nitric acid. To frec such solutions from nitric acid ivithout loss of selenium has afforded great difficult!- from the tiine of the first attempts to estimate this inetal. Rose.' states that solutions of selenious acid can not be lieatetl ahovc I 00' c'. \vitliout volatilization of selenium dioxide and provides for the removal oi the nitric acid l)y evaporation on a waterbath. It has been our experieiicc. and tlie experience of others, that with the most carefull!. coiicluctetl evaporatioiis a loss of selciiitim invariably occurs. I hI.c.1-s antl Schiinosez attribute the lolv results to the formation of selenic acid during tlie evaporation \\it11 nitric acid. They recoiiinien(1 boiling the residue \\it11 hydrochloric acitl for tlic pirposc~t ~ I-cciuciiig f an!. selcnic acid to selenious acitl. This ~ I I - V cautioii i3 iiiaiiiicstlj, iiiiiiecessary \\.licii nitric acitl a l o i i c . has I W C I I used io fli~colll],ose the c0lnpO"ntl. Rat1il;c." reconiniciitletl the atltiitioii of li!.drochluric acid tu the selenious acici solution previous tc 1 c~vaporaticin. 1,ater.' he suggcstctl tlic addition of sodium or potassium chloride to the solution ty prevent volatilization of selenium chloride during the repeatcd evaporations with hydrochloric acid. The results of Rattikc's analyses show a loss of from I to j per cent. hlichaclis,' finding that evaporation with hydrochloric acid a l \ ~ a y sresuitrtl in the loss of selenium. sought to avoid cvapor:ltioii I)>- I)oiliiig ilic solution with a great excess of concentratctl Ii>.(ir,icIiloric acitl until the nitric acid should I I C coinpletcl!. ( I F coiiiposc~(i. 'I'lic~ operation was contluctctl i n ;I flask provided \\,itIi ;L vc'tui-ii con(lenser. llichaclis used this sclicme success-
' 'i j
.I
,f>ll.j..>"!., I ,
(41, j r .
Rc, d cilptll. f;c.! , 26, 1 ( , ' 1 \ . : 1 i u , ] . S r i , s o 3 254 Zlsrlz,..
~ l i l f l l Cll?I>l . , I. -;
J r/te)lr. sOL, 57, 1,
. ~ ~ U > LC/ZPIN. . Pimr-iir.,
i .j,
152, 194 and 201.
f r u k l . Cheiii., 108, 249.
S E L E N I U M IS O R G A N I C C O X P O U N D S .
1089
fully, in connection with the gravimetric method, \\.hen ordinary nitric acid was employed t o bring about the decomposition of the organic compound. I t is our experience that when as little as 4 or j cc. of red fuming nitric acid are used, several hours' boiling under an inverted cooler is not sufficient to bring about complete decomposition and removal of the nitric acid. The method, even should it be successful, is a very great time-consumer and is troublesome from a mechanical standpoint. Further, the solution of selenious acid so treated would be unfit for direct titration, according to the method of Norris and Fay, on account of the large and unknown amount of hydrochloric acid present. It has heen demonstrated that for accurate results with the method of Sorris and Fay the amount of hydrochloric acid employed must fall within a certain, well-defined 1imit.l T h e method suggested in this paper for the elimination of the nitric acid has the advantage of easy execution, requires less time and attention, and gives very accurate results when used in connection ivith either the gravimetric or volumetric methods for determining the selenium in solutions of selenious acid. In the attempted estimations of selenium in a number of compo~inds,some of which contained mercury, n-e observed that in the case of a mercuric chloride compound the results always came nearer the theoretical than was true with the purely organic compounds. This deportment is verified in the analytical results reported by Rathke.* In a body, C,H,Se,OK. the selenium found was 58.21 per cent. (theory 62.36 per cent.), while in the compound 2Se( C,H,),,CI ZnCI,, containing the hen^!- metal zinc, the selenium found \vas 28.59 per cent. (theory 29.44 per cent.). The plausible explanation of this deportment is that the heavy metal present forms with the selenious acid a selenite which is stable at the temperature of the water-bath, thus retaining all of the selenious acid in combination during the evaporation. Accordingly, exactly equivalent portions of mercuric nitrate were weighed out and dissolved in the selenious acid solution previous to evaporation. The results were encouraging rather than satisfactory. The quantity of selenium usually lost by '
+
1 2
Nortoii : Ani. f.S c i . , 157, 267 : Sorris a i d Fay : A m . C l v m .f ,23, 123. . 4 n n . Chcnr. Pharnt.,152,mj-ziz.
1090
R. E. LYONS Ah-D F. L . S H I N S .
evaporation in the absence of a heavy metal was found much diminished. It would be impracticable, for obvious reasons, to have in the solution always a quantity of mercury salt exactly equivalent to the selenious acid. It became evident that the metal addetl to the solution to retain the selenious acid d u r i n g the evaporation must forin a selenite perfectly stable at 100’ C., and sufficiently insoluble in colt1 water to permit of washing free from iiitrates antl nitric acid. X number of simple selenites, e. g.. cadmium. mercury, niagnesiuin and silver, were trial but without success. A l l “ere found too soluble for quantitative purposes. A further search i n the literature was rewartled by finding the statement of 13outzoureaiio’ concerning the double selenite ammonia compounds. A number of selenites dissolve i n ammonium hydroxide and upon evaporation of the solution on a water-bath, or in a vacuum, form double compounds. e. g.,:ig,SeO,,NH,,. Silver and zinc selenite ammonia may lie formed in this way. These compounds are insoluble in water. soluble i n acids and perfectl!, stable at rooo C. I t was foresecn from this, that if to a solution of selcnioiis acid containing nitric acid an excess of silver or zinc nitrate lie added ant1 the solution evaporated to dryness t o expel the nitric acid, the residiie, containing all of the selenium as selenite, coultl then be evaporated Lvith ammonium hydroxide to form the tlouble selenite ammonia. This conipouiid. it was hoped. n.ould permit of washing free from nitrates without appreciable loss. The residue of silver or zinc selenite ammonia could then be decomposed by means of hylrochloric acid, and the usual gravimetric or volumetric nietho(Is for the estimation of selenious acid applied. The methotl \\.as tried antl proved entirely successful. ESI’ERIlfEXT.\L PART.
f’repczrxfioii of PILI.(, Seleiiiiirri Dioside.-T’ulverized selenium was oxidized with concentrated nitric acid and the reaction mix-
ture evaporated to dryness on a water-bath. The residue was moistened with nitric acid and sublimed in a small beaker according to the method suggested by Norris and Fay.? T o ensure greater purity the above product was resublimed in a current of 1
2
Iahvesb. 1 1 . Fo?tsr/wiLfed. C h m i . , 4a, 398. A m . Chem.J., 18,704.
1091
S E L E N I U M I N O R G A N I C COMPOUNDS.
dry oxygen in a combustion tube. The selenium dioxide so obtained, in the form of delicate, white needles, was dried to constant weight in a vacuum desiccator over phosphorus pentoxide. &An accurately weighed quantity of pure selenium dioxide was dissolved in water, the solution acidulated with a drop of concentrated hydrochloric acid and diluted to 1000cc. I n all of the following experiments the volumetric method of Norris and Fay' was used for the final estimation of the selenious acid. Direct Titratioit of the Selenioiis Acid Solution.-Portions of the selenious acid solution, 25 cc. each, were titrated directly with the following result : Selenium present. Gram.
............ ............ (c) ............
(a)
(b)
0.05468 0.05468 0.05468
"IO
Na2S20yused. cc.
27.69 27.63 27.65
Selenium found. Gram.
0.05469 0.05457 0.05461
Detervzilzatioii of Selenitm iii Solzitions Containiizg Nitric Acid. - ( a ) Portions of 25 cc. of the selenious acid solution were treated with a slight excess of silver nitrate and about 2 cc. of fuming nitric acid. The evaporation to dryness and repeated treatment with ammonium hydroxide was then carried out according to the method to be described later. Following are the results : Selenium present. Gram.
( a ) ............ ( b ) ............ (c)
............
( d ) ............ (e)
............
0.05468 006491 0.06491 0.06491 0.06491
Nlio Na&O, iised. cc.
Selenium fonnd. Gram.
27.83 32.78 32.88 32.75 32.77
0.05498 0.06472 0.06489 0.06468 0.06470
( b ) An excess of zinc nitrate was used instead of silver nitrate in a series of experiments similar to ( a ) , with the following result : Seleninm present. Gram.
( a ) ............ ( b ) ............
0.06491 0.06491
W I ONa2SzOsused. cc.
Selenium found. Gram.
32.71
0.06460 0.06468
32.75
( c ) Similar experiments were made using cadmium nitrate in place of silver or zinc nitrate. The results were very much too low. Cadmium selenite ammonia, CdSeO,.NH,, does not withstand the washing. 1
A m . Chem.J., 88,704 ; 23, 119.
I092
SE1,ENIUM I N O R G A N I C C O 3 I P O V S I ) S .
( ii ) The selenium in the following aromatic seleiiiuni coinp i i n t l s \\'as determined by this process : Evaporated S , I O SanSvO:. Weight of with zinc or used. ~ubstance.l silver nitrate. cc. Se preseilt.
0.10745 0.21445
(I) ( 2 )
( 3 ) 0.09065 ( 4 ) 0.094.5
Agx'O:, %n(KO:i)l AgNO., .igNo:;
Se ioiind.
'Theor)-. Fouiid. i'er cent. Percellt.
27.52 54.7s 21.23
0.05444 o.roS66
0.05436
50.67
0,10819
50.67
0.04212
0.0419.;
14.84
0.02950
C.).OZg31
46.47 31.22
50.59 50.46 46.25 ;I.Ur
Lye recommend the following procedure for the determination of selenium in organic compounds :
Heat a weighed quantity of the substance in a sealed tube with red fuming nitric acid one hour, or longer, at a temperature of 235' or 240' C. Transfer the contents of the tube, together with the wash-water, to an evaporating dish of about 400 cc. capacity. Akltl about 1/4 more silver or zinc iiitratc than is theoretically reqtiiretl to unite with the selenious acid present. Evaporate to dryness on a ivater-bath.? V-ash down the sides of the dish with a little water and again evaporate to drytiess. Treat the residue twice with about jo cc. dilute ainmonium hydroxide and evaporate after each addition. Add water and evaporate twice to ensure complete removal of free ammonia. \\lien cool, wash the residue \\.ith water, tlecantiiig the wash-water on a filter, until free from nitrates. l'lace the filter-paper i n the evaporating dish, decoinIlose the silver or zinc selenite ammonia ivith I O cc. hydrochloric ;xid (sp. gr. I . I ~ ) dilute , the solution to about 300 cc. \\-it11 water and add a few pieces of ice. Run in N/ro sodium thiosulphate solutioii iii slight excess. ' allow the mixture to stailtl cxic lioiir' aiitl tleterininc thr excess o i thiosulphate by titration with standard iodine solution. 1 cc. S/IOSa,S,O,, solutio11 Io.o01(,7: gram of metallic c c 1en i uin . T o estimate the selenium gravimetrically : In case silver nitrate is used : Filter after tiecomposing the ((2) I
iiide
'The rubstance used i n
(I)
and ( z j is ~li-pheiiyl-ili-..rIt~~i(le ; i n 1.3') ili-beiizyl-di-sele-
: and in ( 4 ) a coinpound, C-;H&e,.
2 The water-bath must not he allowed to go dry during this or : I I I ~sulisequent evaporation. We have found this method most accurate \vhc~ithe excess of S , i o thiosulphate e n ployed does not exceed 5 to 7 cc. Conipare Sortnn : 1 ; ~ .I .Yo', 157, 2 8 ; : nlso Norris and F a y : A N I . Ch'n2.J.. 23, 123. 4 The mixture must he kept iiear tlir t c i i i p r r n t u r e 01' i i i e l t i ~ ~ ice g during the redirctioll with thiosulphate.
S U L P H U R A X D PIIOSPHOKI'S IN PL.\NT S C B S T h N C E S .
1093
silver selenite ammonia with hydrochloric acid, wash the paper free from chlorine and reduce the selenious acid in the filtrate by ineans of sodium bisulphite. ( b ) In case zinc nitrate is used : Filtering after the decomposition with hydrochloric acid may be omitted. INDIANA
[CONTRIRUTIOX FROM THE
LABORATORS O F T H E ~'ENSSYI.VANI.4 STATE EXPERIMENT STATiON. ]
COLLEGE
T H E DETERMINATION OF SULPHUR A N D PHOSPHORUS IN P L A N T SUBSTANCES. BY
c. P. 13EISTI.E.
i