320
REPORT ON T H E PKOOREBS O F AKALYTICAL CHEMIIGTRY.
REPORT ON
T H E P R O G R E S S OF .!iXAI,YTIC;AI,
CHEMISTRY,
A P R I L TO
JCNE:,1879. BY GIDEOK E. MOORE, PH. D. I. ATOMIC W E I G H T S O F T l l E ELEMENTS.
DR. F. KESSLER( " 1st das Atomgewicht des Antiinons, 120 oder 122?", Bochum. Ad. Stumpf., 1879) has subjected to a critical review t h e question of the tctwnic ioeight of'utditnoiiy. The figures which have been obtained by different analysts are : Schneider Dexter Dumas Kessler Cooke
(1856). . . . . . . . . . . . . . . . . . . . . . .Sb = 120.30 (1857) . . . . . . . . . . . . . . . . . . . . . " = 122.33 (1858). . . . . . . . . . . . . . . . . . . . . . . " = 122.00 (1861) . . . . . . . . . . . . . . . . . . . . . . . ' = 122.37 (1893-1857) . . . . . . . . . . . . . . . . " = 120.00
T h e author shows that the results of Schneider (Pogg. Annal., 97, 98, 293-305) were obtained on a material (native Stibnite, from Arnsberg, Westphalia) which contains impurities of which he took no notice, and that his figures are necessarily too low ; t h a t the results obtained hy J. P. Cooke, J r . (Proc. Am. h a d . Sci., 13, 1-71, and Sill. Jour. [3] 15, 41-49 and 107-124) are discordant, and the methods employed hy him liable to numerous sources of error which have not been duly taken into account in regard t o their influence on the question of accuracy. H e successfully refutes the objections which have been brought against t h e result8 of Dcxter, Dumas and himaelf, by Schneider, Cooke and others, and maintains that the true atomic weight of antimony is 122.0. LE-COQ D E HOISBAUDRAN (C. R., 86, 941-943, through Pogg. Annalen. [Beibl.] 3, 248) has found the atomic ioei,qht qf gallium t o be 70.032 and 69.688, mean, 69.865 (H = 1, 0 = 16). T h e determination was made by the ignition of gallium alum and of gallium nitrate. This atomic weight agrees with that of a body standing midway between A1 and I n (69.82). From the consideration of the relative positions of the spectral lines of Al,Ca,Zn on the one hand, and K,Rb,Cs on t h e other, the ntoinic weight is calculated at 69.86. K. SEUBERT (Inaug. Diss. Tubingen, 1878, through Pogg. Ann. [Beibl.] 3, 322) has determined t h e atomic weight qf iridium. T h e determinations were made on t h e double salts-iridium-ammonium chloride and iridium-potassium chloride. T h e mean result attained was 195.220 (H = 0.9975), or 192.744 (H = 1). 483-484 ;
REPORT ON THE PROQRESS OF ANALYTICAL CHEMISTRY.
321
A. TERREIT. (Bull. SOC.chim. de Paris, 31, 153) has redetermined the atomic weight of aluminum. T h e method employed by the author consists in decomposing gaSeous hydrochloric acid by the metal, and measuring the volume of hydrogen evolved. The determinations were made a t a red heat. The results obtained would indicate that the chloride, formed, possessed the composition : A1 20.26 79.74
c1
100.00
This composition corresponds to the atomic weight 9, if the formula be AlC1, or to 13.5, if the formula be Al,Cl,. T h e latter number alone satisfies the law of isomorphism, and accords with the vapor density determination8 of Derille and Troost (9,35; A1,ClS 2 ~01s.).
-
11. GENERAL INORGANIC ANALYSIS.
EMIL SCHOENE (Zeitschr. f. anal. Chem., 18, 133) contributes a voluminous paper on the quantitative estimation qfhydrogen peroxide. After a description of the author’s experience with the previously proposed methods, taken from a brochure printed in the Russian langnage (“ Experimental Investigation on Hydrogen Peroxide,” MOSCOW,1875), the author proceeds to the description of the colorimetric method employed by him in the determination of small quantities of hydrogen peroxide, snch as in atmospheric precipitates (rain, dew, etc.). The method with potassium permanganate rests, according to h c h o f f (J. f. prakt. Chem., 1850, 81, 404), on the reaction : 2KMn0,
+ 5H,O, + 3H,SO, =-=K,SO, + 2MnS0, + 8H,O + 50,
which was confirmed by the experiments of the author. The reaction can be employed as the basis either for a gasometric method, in which the volume of the oxygen resiilt.ing from the decomposition is directly meauured, or of a volumetric method, wherein a titrated solution of permanganate is employed. The author did not test the gasometric method, but considers it to be unreliable on account of the variations in the coefficient of absorption of oxygen, and other gases contained in the air ; a simple calculation showing that 1 mgm of hydrogen peroxide cannot by this method be determined with certainty in one liter of liquid. T h e volumetric method, conducted by adding the permanganate solution from a burette to the solution,
322
RKPORT ON THE PROGRESS O F ANALYTICAL CHEMIGTRT.
previously acidified with sulphuric acid, nntil a permanent pink tinge appears, was found by the author to be extremely perfect. One millegramrne in a liter of liquid is susceptible of perfectly accurate determination, and :o millegramrrie with tolerable acciiracg. Accwrtling t o all the authorities, hydrogen peroxide acts on hydrogen iodide according to the equation : Hz02 2111 = 3II,O J- I,. This rea(’tiori has been made the basiq of two yrinntitntivr methods, viz: the iodometric (Brodie) arid the acidimetric (IIouzeauj. In the former, t h e amount of iodine, separated, is determined by titration witlt sodium hyposulphite ; in the latter, the aniourit of undecomposed hydriodic acid is determined by titration with centinormal alkali, after expulsion of the separated iodine by boiling. The former method is inferior to the permanganate method. The acidimetric method may he used with advantage where an error of 1 rnpm i n the liter i h immaterial. Thenard’s gasornetric method is based on the decompobition of the hydrogen peroxide h y heat, or b y contact with manganese peroxide, blood-fibrine, etc., and measurement of the volume of oxygen produced. This nietliotl is inferior to the permanganate method, and to that of Houzeau. Methodo: based on the change of color produced in indigo and indigo-carmine, by oxidation a n d reduction, have also been proposed by J . Assmus and H. Struve, but it is still doubtful whether these methods are capable of yielding accurate results. I n the author’s determinations of the percentage of hydrogen peroxide in atmospheric precipitates, the quantities present rarely exceeded 1 mgm in the liter. The only one of the preceding methods, viz., that with permanganate, which was sufficiently delicate, was inapplicable, owing t o the presence of other substances which also reduce permanganate. The author has, therefore, been led t o devise a method, based on the coloration produced by hydrogen peroxide in a neutral solution of potassium iodide and starch. The reaction is so delicate that 0.00008 g m H,O, in a liter, produces a weak violet coloration, which attains its maximal intensity in about 5 or 6 hours. T h e reaction is only applicable to solutions of not more than .001 g m to the liter. When this limit is exceeded, the color changes are n o longer appreciable. T h e color scale is prepared by ascertaining, by means of the permanganate method, the strength of a solution of pure hydrogen peroxide, and diluting, so as to obtain eolutions of 0.1, 0.2, O.:l, 0.4, 0 . 5 , 0.6, 0.7, 0.8, 0.9 and 1 mgm, t o the liter. These solutionh are treated with potassium iodide and starch solution in the same manner as the liquid t o be tested. When preserved in the dark, and in well stoppered
+
REPORT ON THE PROORE88 OF ANALYTICAL CHEMIBTRY.
328
glasses, these solutions remain unchanged for three to four weeks. T h e author prefers to prepare a new scale every ten to fourteen days.
J. VOLEIARD (Ann. d. Chem., 190,1, through Zeitschr. f. anal. Chem., 18,271) has communicated a voluminous paper on the application of the silver titration with arnmoiaium ,wZphocyardide to the determination of the halogens, copper and mercury. T h e method involves the use of two standard solutions, viz.: a deci-normal silver solution and a solution of ammonium sulphocyanide of the same strength. T h e indicator consists of a cold saturated solution of 'ferric ammonium alum. The titration of chlorine is as follows : the aubstance is dissolved in 200 to 300 cc of water ; 5 cc of the iron solution is added, and then nitric acid until the color of the ferric salt disappears. A moderate excess of silver solution is then added from the burette, and then, without the necessity of filtration or even shaking or heating, to cause the silver chloride to agglomerate, the potassium sulphocyanide is added, the flask being continually shaken, to insure the rapid mixture of the sulphocyanide solution with the contents of the flask. As soon as the liquid assumes a light yellowishbrown color which, after shaking and letting stand for 10 minutes, doe8 not disappear, the titration is finished. T h e difference between the volume of silver solution originally taken and that corresponding t o the sulphocyanide added, is the measure of the chlorine in the substance. The titration of homine is conducted in the same manner a~ that of chlorine. The titration of iodhe requires certain precautions owing t o the facility with which silver iodide carries down with it potassium iodide, silver nitrate, or both,which then enter very slowly into reciprocal decomposition with the silver or sulphocyanide solutions. The iodide is dissolved in two to three hundred times its weight of water, and the silver solution added from the burette. The silver iodide is, a t first, in a state of extreme subdivision, and the liquid appears like a yellowish-white milk ; as soon, however, as a slight excess of silver has been added, the liquid appears t o curdle, and the precipitate separates out after a short shaking ; o: to to cc more of the silver solution is then added, and the liquid cleared by shaking. 5 cc of the iron solution are introduced and the titration is conducted with sulphocyanide solution, as above indicated. When the color first appears it is dispelled by shaking, owing t o the silver nitrate occluded in the precipitate. The addition of small quantities of the sulphocyanide solution, with prolonged shaking, is continued until the eolor no longer disappears on shaking. T h e process is well adapted to the determination of the halogens in organic compounds.
324
REPORT UK TIIE: PH0QKE:SS
O F ANALYTICAL CFI EYISTRY.
When the halogcnn are present with sulphocyanides, the last named acid niiist be destroyed ; if chlorine aloire is sought, the acids are prec*ipitated as silvcr wmpourids, and t h e srilphoc.y~rti&of silver is decomposed, irk :L fl:tsk, by warininy with a mixture of 2 p:trts sulphiiric acitl arid I part water, :mi addition of :L f e w tlrops of' iiitric acid ; i f bromine mid iotline are prtwirt, the nir'tliod of ('arius, or me!tjng with alkaline nitrate and carbtmate, should be employed. Chloride of silvcr may be separated from iodide of silver by treatmerit in arnnioniacal stjiiition with airinioriiuni nulphocyanidc. ?'he clllorirle is completely ti~snd'ormcdto sulpliocyariide ; t h e iodide reniains u n a f f t ~ m l . I ' y o ~ w / ~ may / he dctcrmincd by precipitation with a11 excess of silver, filtration and H hing of the precipitate, and titl.atiO11 of t h e VS(~C'SS of silvcii3 in the filtrate with sulphocyanide. C'oppr is determined b y heating the soliitioii t o h i l i n g , and then adding an excess of siilpliiiroiis :wid ; 0 1 1 the first addition of sulphocyanide solution, the liquid is cGo1orc.d of a dirty green, owing t o the formation of cupric srilphc,cyanidr ; on shaki~ig,howttver, the liquid becomes c:olorle.ss, :tiid R white precipitate of cuprutin sulphocyanide is tlirowri down. I Y h w the copper is (Intirely precipitated, the last additiou of siilphocyariide produces no change of color. T h e liquid is allowed to cool, diluted to 300 cc, filtered through a d r y filter, and 100 cc of the filtrate titrated with silver solution, to determine t h e excess of sulpliocyanidc~. Merwr!/, when present an it salt of an oxygen acid, may also be approximately determined irr the same manner as silver. T h e author finally calls attentivn t o another application of ammonium sulphocyanide in volumetric analysis, viz : as a measure of the strength of staiidard solutions of putassium permailgimatc, Tlic oxidatioii of ~ulphocyanhydricacid proceeds :iccording t o the equation : S,C," 6 0 = 2s0, CJIE. The solution of ammonium sulphocyanide possesses the great requisite of entire stability. T h e author ias minutely indicated the precaaiitions to be observrcl iri each of the foregoing methods, and upon which the accuracy of the results largely depend. W.J. SELL(Jour, Cheni. Soc., 1 8 i 9 , 292) has devised t h e following method for the volumetric determinution oj. chroniiunr. T h e solution of a salt of chromium is acidified with sulphuric acid, heated to boiling, and treated with solution of potassiuni permanganate, in small portions, until the liquid retains a distinct purplish tint, after boiling for three minutes. T h e chromium is t h u s oxidized to chromic acid, according t o t h e equation :
+
5Crg(SO1k
+
+ S&Yn,Oa + 6H80 - 10CrOa f 2K,so4 + 6MnSO4 + 6HzSO4.
REPORT ON T H E PROGRESS OF ANALYTICAL CHEMISTRY.
325
T h e solution is then rendered slightly alkaline by addition of sodium carbonate ; alcohol is added, to reduce the excess of pernianganrte, aud the manganese removed by filtration. The chromic acid in the filtrate is determined by titration with iodine and sodium thiosulphate. I n the analysis of chrome-iron ore, the ore is rendered Rohble b y fusion with potassium, or, preferahly, sodium acid ~ulphate. Thr-ee determinations of chromic oxide in the bame sample of chrome-iron ore gave the author 47.11, 47.23, and 4 i . 1 1 p. c. of Cr20,. The solution of the ore may be more quickly effected by fusion with potassium fluoride, or with a mixture of one molecule of sodium acid sulphate, and two of sodium fluoride. War. GALBRAITH (C'hem. News, 39, 276) claims priority in the discovery of the foregoing method, and refers to his article in the Chemical News, 25, p. 151, wherein it is recommended for the determination of chromium in iron and steel. The process there described is in daily use, and gives satisfactory results.
ALEXANDER CLASSRX(Zeitschr. f. anal. Chem., 18,1 7 5 ) communicates a piem method for the sepnrutioa of ferric oxide and nlumiita .from manganese. The author has previously (Berichte d. deutsch. chem. G e s e l l d . , IO, 1315) communicated a method for the separation of iron from manganese (nickel, cobalt and zinc), based on the following facts: If to a solution of a salt of manganese, there be added, drop by drop, a solution of neutral potassium oxalate, there is formed a white precipitate of manganous oxalate, which dissolvee in an excess of the precipitant as potassio-manganous oxalate. If acetic acid be added to this solution, the double salt is decomposed, with separation of manganous oxalate, which is absolutely insoluble in acetic acid, although soluble in acid oxalates, and in oxalic acid. Ferric salts form, with a sufficient excess of potassium oxalate, potassio-ferric oxalate, which is not decomposed by acetic acid. The presence of the alkaline chlorides exerts an unfavorable influence on the precipitation of the manganous oxalate. When, however, the solution contains a sufficient quantity of a metal, the oxalate of which is not soluble in alkaline chlorides, the entire amount of manganous oxalate precipitates with the oxalate of the metal in question, even when a large quantity of alkaline chlorides is present. Zinc was found by the author to satisfactorily fulfill this condition. T h e process of the author is to be employed in two modifications, according as the material analyzed has been dissolved in hydrochloric acid, with addition of bromine, or in nitric acid. In the first ca8e the solution is evaporated t o dryness, the dry residue digested a few min-
326
REPORT ON THE PROGRESS OF AXALYTICAL CHEMISTRY.
utes with 5 t o 10 cc bromine water, in a covered dish, on the water bath ; then about 'i times as much potassiuni oxalate, as there is of oxides present, is added, and the mass digested 1 hour on the water bath. Most o f t h e iron ant1 all of the manganese thus dissolve as oxalates. The undissolved residue of ferric: oside is dissolved by the gradual addition of dilute acetic o r hydrochloric acid. I'he intensely green solut,ion is concentrated until green crystal8 of ferric oxalate begin to separate ( t o allout 20-25 cc. wlien 0.4 to 0 . 5 c;f iron has been taken), poured hot into a beaker of about 100 vc contents, the dish washed with hot water so that the bulk of the liquid is ahout, 4 0 t o 45 cc, a few oc alcohol added, to reduce :my higher osicle of manganese which may have formed, and then zinc chloride solntion, frorii a burette, until to 1 part of mangmoits oxide, 3 to .i parts of zinc oxide are present. If too little potassium oxalate has been employed, more is added until the precipitate of zinc oxalate, which then forms, is, for the most part, dissolved. An equal volume of concentrated acetic acid is now added t,o the liquid, and the covered beaker allowed t o stand a short time a t 50-60' C., until the precipita.te has becoinc crystalline, and the supernatant liquid perfectly c:lcar. F o r exact determinations 6 hours standing is reqiiirtd 'rhe precipitate is tiltered, wa.shed, by decantation, with a mixture of equal volumes of concentrated acetic acid, alcoliol and water, until the filtrate gives no iron reaction w i tli potassiiirn sul phot y anide (3-4 decal1tat ions, using each time 10 cc liquid, are sufficient), brought on the filter, dried and ignited, wit,h free access of air, so as t o form thv cwmpoiind of Mn20s and ZnO, iiientii)iied by I3iinsrin and Krieger (Ann. 11. ('liein., 87,2 5 5 , et seq.). T h e ignited precipitate is finally decwrnposed with st#ronghydrochloric acid, and the chlorine thus liberated, titrated with sodium hyposulphite. T h e precipitate is very apt to retain a S I K I ~ I I proportion of potassium oxalate, which, by ignition, passes first t o carbonate, and then to permanganate. T o avoid the excess whirh would result from this carise, the precipitate should first, IJe heated for a short time only, then washed out with hot water, and again ignited, until the mangariese ceases t o absorb oxygen, before t h e chlorimetric titration. If the ferric oxide resulting from t h e solution of the residue of the first evaporation, has been dissolved in hydrochloric acid, it is necessary to carefully neutralize the excess of oxalic acid thus induced, before precipitation with acetic acid. If the substance analyzed has been dissolved in hydrochloric acid, with addition of nitric acid and evaporation t o dryness, nitric acid must he added t o bring the residue entirely into solution. The excess of acid must in
REPORT ON THE PROGRESS OF ANALYTICAL CHEMISTRY.
327
this case, also, be neutralized with potassa or soda solution before precipitation. The method also permits the simultaneous separation of Co, Ni, Cu and Ca, from a large excess of iron, and these metals may be ~eparatelydetermined in the liquid from the chlorimetric determination. Phosphoric acid, when present, is also thrown down with the metals above mentioned. The author reserves for further investigation the question, whether the separation of this last-named substance is quantitatively exact. The ree.ults obtained by the author in the separation of quantities of manganous oxide, varying from 0.0022 t o 0.4970 gm, from quantities of ferric oxide, varying from 1.51 t o 2.85 gms, show the method to be extraordinarily exact, as well as extremely easy and rapid in execution. ALEX.CLASSEN (Zeitsch. f. anal. Chem., 18, 189) describes a method for the quantitative estimation of cobalt, nickel and zinc, by prec@itation as oxalates, based upon the same principles and manipulation as in the preceeding method, except that no addition of zinc salts is required. T h e advantages of this method, compared with those previously in use, are, that the oxalates of these metals are more easily obtained in a Rtate of purity than the hydroxides and carbonates precipitated by the older methods, and that these metals may be separated from ferric oxide and alumina without previous precipitation of the latter. The oxalates are converted into oxides by ignition-at first gentle and in a covered platinum crucible, and then intense, with free access of air. Nickel and zinc are weighed a8 oxides (NiO, ZnO) ; cobalt is reduced to the metallic state by ignition in a current of hydrogen. The analytical results obtained by the author demonstrate that in respect to accuracy and constancy the method leaves nothing to be desired. The presence of ammonia or potassium chloride causes the method to give too low results with cobalt, and of ammonium chloride, with zinc ; sodium acetate is without effect on the results.
ALEX.CLASSEN(Zeitschr. f. anal. Chem., 18,194) has investigated the method of Tamm (Fresenius, Quant. Anal., 6 Aufl., p. 575) for the separation of mangatwe f r o m zinc, by the precipitation of t h e former as manganous carbonate by ammonium carbonate, in the presence of ammonium chloride, and finds that the method gives too high results, owing to the fact that the precipitate always contains more or less zinc carbonate, which cannot be removed b y washing with ammonium carbonate. JOHN PATTINSON (Jour. Chem. SOC., 1879, 365) communicates a method of precipitating manganese entirely as dioxide, and its appli-
crctiojc to the volumetric cleternLitii7tioii of n t u u g ~ ~ ) i e s e After . a careful study of the conditions under which manganese ruay he precipitated as dioxide, the author has attained the samv c*oncliision :is Rlessrs. Wright and TAuff (idern, 33, 1). 513), vix., that cwtain methods usually prescrilwrl for prodiic.ing hydrated nixnganese (lioxide, do not yield this oxide alone, lirct :i mixture of dioxide>with varioils lowei. oxides. T h e author has trip11 other methods usu;1lly I Y A conmiended for this purpose, with similar results. If(, finds, however, that the inanganese may lie invariably prc~tipitatedentirely a* dioxicle~, it' (( certtciic c o i i o i o c t o ~ ' . ~ > ro,rkZ/J r i ~ hr. ~ ) w . s c i i f iii tkc, , s o l i / t ; o i i , whem a sufficent excess of rxlcium hypochloritc~ or bromine water, aiid, a f t e r heating the solution to from ti()' t o io" C., exress of calcirini carbonate is added, and the mixture well stirred. 'I'he prcsence of half as mnch iron as there is of rnariganrsc, is snflie~ientt o determine t h e precipitatioii of the latter wholly as diositlc. Iri practice, however, it is preferable t o have equal quantities of each metal-a large excess of iron is no disadvantage. T h e experiments of the author showed that from 99.95 to 100.12 1)er cent. o f the nianganesc is precipitated as dioxide under the above wnclitions. In tile alia1 an iron ore, the author dissolves 10 grains in a 20 ox. twaker, in 100 fl. gra hydrochloric hcid (sp. gr. 1 , l W ) : cdcium carlmiate is added until the liquid is slightly red, and then six or seven drops o f hydrochloric acid : ~ , O ( ) I ) grs of a solutioii of bleaching powder (10,000 grs water to 1st) bleaching ponder) is now added, a i d boiling water poured in until the temperature is 60' tu 70" C. : 2 5 grs calcium carbonate is introduced, and the solution well stirred until the ciffervescence ceases. T h e dark brown precipitate settles readily. If the supernatant liquid lias a pink color, owiiig t o the formation of permanganic acid, a few drops of alcohol sliould be added to decolorize it. Instead of the bleaching powder, 500 grs of saturated bromine water may he used. l'he precipitate is t i l t e d , wasllcitl, and placed in 1,000 grs of a solution of s:~I) grs ferrous sulphate i l l 10,000 grs of a mixture of 1 part sulpliuric acid and 3 parts water, i n which i t readily dissolves, even in t l i i cold, converting its equivalent of ferrous sulphate into ferric sulphate. Cold water is n o w added, and t h e liquid titrated with potassium dichromate, to determine the excess of ferrous srilphate. From the ferrous sulphate oxidized, the percentage of manganese may readily b e caaleulated. If too little iron be present, pur(' ferric chloride (free from niailganestl) is added. The method is ap1)lical)lt. t o tlit. analysis of spiegeleisen, ferroniaiiganse, steel, slags, ctc. The special precautioiis to be used in rac*h case are fully dc-
R E P o R r ON THE PROGRESS OF ANALYTICAL CHEMISTRY.
329
scribed in the paper. T h e author claims for this method not only a qreat saving of time, but, in many cases, more accurate results than the gravimetrical method is capable of affording. The author further states that it is now possible t o determine manganese volJmetrically by this method, with as great exactitude and almost as speedily, as iron may be determined by Penny’s method.
C. ROESSLER (Herl. Ber., 12,925) communicates a new method for the voltrrnetric deternaiiiution of inanyanese, with the use of Volhwrd’s method of silver titration. T h e method is based on the formation of a compound, first observed by Woehler, and which H. Rose found t o possess the composition Ag,O,\Zn,O,. T h e solution of the manganw e compound (which must be free from the halogens and from organic compounds capable of reducing silver salts) is placed in a to liter Aask, treated with an excess of a solution, of known htrength, of silver nitrate, and heated on the water bath. Sutiicient Nodium carbonate is added to completely precipitate t h e silver. A moderate excess of ammonium hydrate is then added, the flask is well shaken, cooled, filled to the mark, and the solution filtered through a d r y filter. The precipitate on the filter possesses the above composition. Ily the determination by titration, by Volhard’s method, of the amount of silver remaining in the filtrate, the proportion contained in the precipitate, and hence the weight of the latter, is readily ascertained. T h e reaction takes place even in the presence of a large proportion of ferric oxide ; on account, however, of the organic compounds resulting from the solution of spiegel, and other varieties of iron, it is necwsary to separate the iron as basic acetate. T h e filtrate from this precipitate is then free from reducing substances, and suited for the foregoing method.
+
C. MANN (Zeitschr. f. anal. Cheni., 18, 162) has devised u new method f o r the uoltrnietric estiinutiort of z i m , based upon t h e reciprocal deconiposition of hydrated zinc sulphide and silver chloride, and the estimation of the zinc from its equivalent contents in chlorine by means of Volhard’s method of silver titration with alkaline rhodanateu, using ferric nitrate or ferric ammonium sulphate as an indicator. T h e zinc is precipitated by hydrogen sulphide in acetic acid solution, the excess of t h e precipitant expelled by violent ebullition, the precipitate washed by decantation and filtration. ‘ h e precipitate, with the filter, is then introduced into a tlask, water and silver chloride added, and the whole boiled until t h e mpernatant liquid is quite clear; finally 5-6 drops of dilute (1-6) sulphuric acid is added. T h e precipitate is then filtered off, washed, and the solution, after adding a
measured voliime of silver solution, titrated, as abovc. The determination eonaumes, from the moment of tlie introduction of the suiphuretted hydrogen to the enti of t h c p r o c e ~ \ ,about 1 horirb. ‘I’he results obtained b y tlie author arv t*rc*twiinylyiatisfactory ~ T H U 0. K ~ ~ A I ) I ) oh( ((’heni. scans, 39, 156) h a h found that 111 t h e titrtrtiorr 01‘su4)fi~wic u ~ i d7ritk h t r h [ i i t chloride, a* also in other cases where a iniiiute quantity of soli(i prw*ipitate I - t o be observed in a drop of liquid, the observation is greatly facilitated 1)j placing the liquid on a mirror, instead of O I I tlir black glass h ~ t h e r t orecommended.
+
b’. ~ ~ I L S T E Iand N L. JAWEIN(Berl. &I-., 12, 446) communicate a method f o r the quurititutiae detetrrnuhutioit qf zanr by electrolysis. T h e solution is prepared by adding sodic hydrate to a solution of zinc iiitrate or sulphate, until a precipitate is formed, and then adding potassium cyanide until t h e solution is again clear. W i t h a current from 4 Bunsen’s elements and platinurn electrodes (Zeitschr. f. anal. Chem., 8, 28 ; 11, 6 ) the precipitation is s t the rate of abont 0.1 g m zinc per hour. When the precipitation is finished, the electrode with the adhering zinc is removed, washed with water, then with alcoliol, and finally with ether, and dried i n the air bath. After weighing, the zinc is dissolved off, the electroJe re-weighed and again immersed in the solution t o ascertain if the precipitation be complete. T h e method gives exact results. The authors made an analysis of brass by solution in nitric acid, evaporation to dryneah, solution In water, and precipitation of the copper by electrolybis. T h e platinum spiral forming t h e positive electrode waq coated with lead dinoxide, from which t h e percentage of lead was determined. T h e zinc was then determined, aa above. T h e results of the analysis were : Cu Zn Fb
67.13 32.93 0.01
100.0;
F. BEILSTEIN and L. JAWEIN ( B e d Uer., 12,759) have been led by reason of the insufficiency of t h e methods previously in use, to attempt the electrolytic detervninotion of cadmium. T h e authors employed a method similar t o the one used b y them in t h e electrolytic determination of zinc. T h e solution of cadmium chloride or nitrate was neutralized with potassium hydrate, treated with a n excess of potasuium cyanide, and subjected t o electrolysis, using a spiral of thick platinum wire as the anode,
REPORT ON THE PROQRESS OF ANALYTICAL CHEMISTRY.
331
and a cylinder of platinum foil as the cathode. The following results were obtained with electrolytic cadmium, using the current from 4 15-cm Bunsen's cells. TAKEN.
FOUXD.
0.2356 0.2 5 6 5 0.2344 0.2547
0.2344 0.2547 0.2324 0.2532
TIME.
3 hours. 3 " 2 " 3 "
The authors attribute the slight loss in these experiments to the contamination of the electrolytic cadmium with platinum. The end of the precipitation is ascertained by testing with hydrogen sulphide. T h e precipitated metal is washed, first with water, thcn with alcohol, and dried by placing in a heated platinum dish. E. F. SMITH(Am. Jour. Sci., 117,60) communicates the following method for the electrolytic determination of cadntiuvn. T h e solution of acetwte of cadmium is placed i n a platinum crucible, which it should fill to about one-half, and placed in contact with the negative pole of a two-cell Bunsen's battery, the positive pole is formed by a piece of platinum foil dipping into the crucible. I n about 3 hours the cadmium (0.1270 gm) was entirely deposited, as a crystalline greyish-white coaung on the crucible. T h e precipitated metal was washed with water, then with alcohol, and finally with ether. T h e conditions for success are : lst, to work with concentrated solutions, and Ond, to have the current of sufficient strength for rapid and energetic action. PH. DE CLERMONT(Comptes Rendus, 19,May 12, 1879, through Chem. News, 39. 251) has investigated the action of ammoniacal salts on certain metallic sukhides. Bismuth, cadmium, copper, mercuric and mercurous sulphides, are not affected by boiling with solution of sal-ammoniac. Antimony trisulphide yields antimonou8 chloride, which dissolves, and ammonium sulphide, which escapes. Stannic and stannous sulphides yield respectively stannic and stannous oxides. Metals precipitable by ammonium sulphide, but not by hydrogen sulphide in acid solutions, behave in a peculiar manner with sal-ammoniac. ManganeRe, iron, nickel and cobalt sulphides dissolve, the last two more slowly. Zinc sulphide resist8 longer, b u t ultimately dissolves. Alumina and chromic oxide are insoluble. On these reactions the author founds a method for the separation of certain metals. T h e solution containing the precipitate produced by ammonium sulphide with salts of cobalt, nickel, manganese, iron, aluminum, chromium and zinc, is added t o a boiling solution of sal-ammoniac,
ani1 hoiletl f o r a siifiicierit tiiiie. Oil tiltering, tlir ti1tr:ttt. coiitains all tlic i i m :tiit1 ~iiaiigaiiese,a i d 1);trt of the c,)lxilt, riic.kel ; i i i c I zinc., wliile the uiic1iw)Ivwl lwrtioii contailis all t l i o ;ilriiriina ani1 c~lironiiuntwith t h e rcssi(lirc of tlica nickel, c.ol);ilt :inil ziii(+. Wie analysit. is ( ~ o 1 i i ~ ) l c ~ t ( ~ ~ l b y t h e i i s i i i i l rrietht)tlu. If i r o i i :iritl nianganesc ;tlonc havia to t w .;elmratetl frorii aluniiniini ai111 clironiiuni, the ~ I X J C P S S is ~.i)iril)lctea11(1
F. HEII,STEIS(13erl. Her., XI, I , i l S , through Zeitschr. f. anal. ,(ic*keljtio,u z i i i c Cliein., 18,'Lti'L) has found that t h e septrrcctioii may be effected b y precipitating t h e latter with hydrogen sulphide fiviri wliitioiis cwntaining citric acid 01' soliihle citriitcbs. 'l'he precipitation of tlie zinc is complete after staiirlirig 3 ours i i i t,he cold, while the nic*kel C a l i be determined t)y electrul of the tiitrate, after evalmr:it,ioii :iritl riaturation with ammonia. i i i i i i i i o n i i i i i i c*hloritlc Iiinclers the elcv*trol?;tic precipitation of the nickel, the riic.tals should he i i i solution as nitrates. With clue obstwaticv, of tlie precautions intlicatetl hy the author, the inetliod gives extremely :tc.cwr:rtc rcaults. J L I , I U S ' l h o m m (Uerl. Uer., XI, %,~4:3, throriyli Zeitschr. f. anal. Cliein., 18,2 6 3 ) has found that the /irwiliitcctt; j i i - o ( I J ( m {/I!/ / i y < h o p t t s d p f i i t k c i i ~solirtiuns of cupric. .ericoloration, and. 011 dilutioti with Tv:it(xr, :L soiiitioti , i f :t yellowisli-green w l o r . cSlt3:tr
I)R. J . PICTRI (Zeitscli. f. iilia1. ( ‘ l i m i . , 18, 2 1 1) wmniunisatt~x the results of Iiis c~x1)eiiinentso t i ? / t c m?rosrvq)ic. f k t w t i o i , 1q‘m p t i t t fliio.. Piircl ergot i s estretiiely c~iisyl i t ‘ itlctititicaticitt. \Vlicii e x m i ined undc,r tlici niicrosco1)c, it nliows t h c I i w s e i i c e of the riiinutc oval granules, usually .with two nwleii, wliich, as yellow c.01ora,tion 1JdUCftl b y tlir acticin of citnstic alkalich, is a l s o unreliable. Tlie xiitlior gives tlic preference to the following inodificatiori of Jacoby’s rrietliod : About 20 ginti of the flour are boiled for .i minutes with alcohol, and, after subsidence, the yellow solution decatited, the operation r .
REPOOHT ON THE PROGRESS OF ANALYTICAL CHEMISTRT.
341
being repeated until the alcohol above the powder is colorless ; 20 drops of dilute sulphuric acid are then added, the m w i well stirred and allowed t o suhside ; the red solution is then filtered and examined in the spectroscope. About one-qaarter test-tube full of the solution is then mixed with 1 to 2 volumes of water, and the milky liquid s!inken with 1 to 2 cc amylic alcohol, a second portion with chloroform, a third with benzole, and a fourth with ether. If ergot be present, these Rollitions will be reddish, and will show the characteristic spectrnm. By this method, 0.2 per cent, of ergot may easily be detected in flour, using 20 gmR. B y means of colorimetric coinparisons with solutions of pure ergot, the author has devised a satisfactory method of quantitative determination. The author’s measurements of the ergot spectrum agree satisfactorily with those of C. H. Wolf (Zeitschr. f. anal. Chem., 18,119 ; and this Journal, I, 104).
F. SF,I.MI (Berl. Ber., 11, 1,692, thronqh Zeituchr. f. anal. Chem., strychnhe .- When moistened with a solution of iodic acid in sulphuric acid, it becomes colored, first yellow, then brick-red, and very gradually passes to brilliant violet-red. B QUANTITATIVE OTTO HEHNER(Cheni. News, 39, 197) has subjected to a critical examination the method of Dr. Prrusyaforthe ~ e ~ e r m i i ~ ( i t i o ~ i q f g 1 u c o s e by m e a i ~of ammoniated FehZkg’s solittion (Chem. News, 39, 7 7 ; and this Jour., I, 109). The author finds that the method gives variable results, except when the proportion of caustic soda in the test solution, is between 120 to 150 gms per liter, when the results obtained accord with those of Dr. Pavey. W h c n this proportion is observed, the method gives excellent results, the end reaction ia beantifully sharp, and the titrations of the same liquid never differ more than 0.1 cc of the sugar solution. It must not be taken for granted that the ratio between the ordinary and ammoniated Fehling’s solutions, iR valid also in the case of other reducing saccharine su6stances. Dextrose and laevulose act equally, but milk-sugar and maltose do not. T h e siithor’s investigation on these points is still in progress. DR. E’. W. PAVEY (Chem. Wews, 39, 249) has been led by the foregoing article of Hehner, on the subject of the titration of glzicoee with anirnoii inted Fehling’s eolution, t o make an explanation of the causes of the discrepancies between the remlts obtained by the lastnamed observer and himself. These reside in differences in manipulation, which are corrected, as Hehner has found, by the use of a larger
18,292) communicates the following rwctiou
or
342
mpmr
O N THE PROGRESS OF ANAI,YTICAL CHEMIY~’KY.
proportion of alkali. The author uses nearly double the aniount of alkali existing i i r the usual Fehling’s sulution, his formula being : Cupric sulphate :34.tj5 gnis I’ut:tssio-sud i u ni . 17Y.UO “ Potassium hytlratc,. . . . . . . . . . . . . . . . . . . . . . l60-0o I‘ IVater to 1 liter. 120 cc of this solution are mixed with 300 cc ammonia (sp. gr. 0 . 8 8 0 ) , rind water addctl to 1 liter ; 20 cc of tlie liquid correspond to 0.01ti gni glucose. l‘lie author lias made the highly impurtaiit observation that ~ r i Ic K ; ~ exerts :L reducing action oii cupric acid, equally as precise as glucose, and finds tliat its qwntitative deterniination may be easily and speedily made with ammoniated cupric liquid. Three molecules of cupric. oxide are reduced by one of uric acid. I
F. Soxiiixr (Chem. Centralbl., 1878, 2 I8 and 236, through Zeitschr. f. anal. Chern., 18, 348) lias investigated the grrciictitutioe ~ ~ I t w i ) i g e f w t (fthe d i f m t t t Xirctls cg‘ sugci~’ o i t ct1k:ali~rew p p r s d i i t i w z s . The author finds that the effect cannot be expressed by a definite equivalent proportiuir, an one part of sugar will reduce very tlifferrnt, quantities of copper, accorcling to the strength of the cwpper solution. ‘The results obtained by titration with Fehling’s solation are purely empirical, and have nothing to (10 with a fixed stcwhioinetric ratio. The mct.hoti is, lwwever, correct up t u u.2 per ctsnt., when the tratioii is ubserved i i i standardizing the solution id in proper. 1 lie differcanre between a deterirriii:ttioii \vlierein a barely perceptible excess of cup1)er solution is used, a i i d oiie in which tlie amount is tloubled, ilia? re:ic.li 6.5 to 7 . j per cent. 1Ie believes, tlrcrefore, tliat a graviirietriu:rl estimation is absolutely impossible. R. CLBRICIIT (Cliein. Centralbl., 1878, 3 9 2 ) and JI. Maercker (ideni, 584, through Zeitschr. f. anal. Clien~,,18, 3 4 9 ) ~ o i i f i r n ithe foregoills observations of Soxhlet. Maercker, ho\vever, contest?: the correctness of Soxhlet’s conclusion of the impossibility of a gravimetric determination. ‘rlie following prucess i s very exact : :j4.6:12 grns of pure c*upricsulphate, on the one hand, and 6 3 gms caustic sot1:t witli 1 i 8 gin?: potassic-sodic tartrate, on the other, are each tiissolved in 1 liter of water, and preserved separately. The sugitr determination is made as folloirs : 25 cc of each of these solutions are placed in a be;tker, t h e sugar solutioii (containing not more than 0.12 g m dextrose) is added, and then water to a msrk on the beaker i n d i c d r i g 100 cc. ‘I’lie beaker i H then allowed tu stand in boiling \viit,er for 20 minutes. The liquid is then filtered, and the percepitate washed r ,
REPORT ON THE PROQRESS OF ANALYTICAL CKEYIEITRY.
343
with 300 cc of boiling water, The filter is then burned in a platinum crucible (sic), and the copper reduced to the metallic state by a stream of hydrogen, with the aid of the Rose's crucible cover. The circumstance that the filter paper is liable to absorb some copper, was not found to materially effect the accuracy of the method. The fact that copper, wheii reduced by hydrogen, retains a certain proportion of that gas, is immaterial, for the reason, that it equally affects the deterniination of the reduction equivalent. This latter datum was determined by the use of pure dextrose. Proceeding as above, the authors obtained the following results : 0.1 1 1 1 gm dextrose g a v e . . . . . . . . . . . 0.196 g m Cu. 0.1000 0.0667
L
