6 11.-On the Arseniates of Baryta Lime and Magnesia and the Separation of Arsenic porn other Elemeds. BY FREEERICK FIELD,M.R.I.A. F.C.S. ALLchemists must have experienced the difficulty of separating arsenic from other elements with whichbit is associated. It is scarcely necessary to say that the method usually adopted is to precipitate the arsenic as sulphide with the other metals which are thrown down by sulphide of hydrogen from their acid solutions and subsequently digest the precipitate with the sulphides of ammonium or potassiiim. Many other bodies however are re- dissolved as well as the sulphide of arsenic the separation of which in this stage of the process becomes exceedingly perplexing and difficidt. Even in the earlier operations of the analysis the arsenic has to be entirely converted from arsenic acid into arsenious acid in order to insure its entire precipitation and this is usually effected by sulphurous acid or a sulphite.When tEe arsenic exists in a very large proportion repeated deoxidations and precipitations are necessary which not only involve the expendi- ture of much time and labour but are attended with loss of the substance under aiialysis. Much of the arsenic it is true can be eliminated by calcination with occasional additions of carbon ;and probably in the establishments where nickel and cobalt are obtained on a large scale eight or nine-tenths of the arsenic are got rid of by this method. Even then the manipulator finds it necessary to pass a powerful and long-continued stream of sul-phuretted hydrogen through the solution of the oxides and to expel the excess by ebullition previous to the separation of iron nickel and cobalt.Calcination however is inadmissible in chemical analysis for-independently of the mechanical loss which is likely to occur-some metals especially silver are volatile in arsenical fumes. Moreover presuming that all the operations mentioned above have been carried out successfully the arsenic cannot be separated entirely by the alkaline sulphides. Mr. Bloxam has shown" that Then the proportion of copper is large * Quarterly Journal of the Chemical Society vol. v p. 104. ARSENIATES OF LIME BARYTA AND MAGNESIA.. much arsenic is left with its sulphide even after long digestion with sulphide of ammonium and indeed that 1 per cent.of arsenic becomes insoluble iii the alkaline sulphide when the remainder consists of sulphide of copper. In the analysis of the ores of nickel and cobalt I have found the formation of arseniate of potash highly useful as a means of separating arsenic from these metals. When the arseniates of nickel and cobalt in solution of hydrochloric acid are boiled with excess of potash they are entirely decomposed the whole of the arsenic wid uniting with the alkali. Ammonia cannot be employed as the arseuiate of nickel is very slowly decomposed by this reagent. When native arseniate of nickel is dissolved in hydrochloric acid and ammonia added a gelatinous white precipitate is produced and the super- natant liquid is colourless; and in order to obtain the blue ammo- niacal solution it is first necessary to free the nickel from arsenic.Iron is also present in this class of minerals and sesqui-arseniate of iron is highly soluble in ammonia so that if the substance con- tain arsenic it is impossible to separate nickel from iron by that alkali. When an ars&cal ore of cobalt (containing iron) is digested in nitro-hydrochloric acid and ammonia added in excess no precipitate is formed provided there be sufficient arsenic to convert the whole of the iron into arseniate and chloride of ammo-nium to form a soluble compound with cobalt. The solution has a fine brown colour and is decomposed by sulphate of magnesia sesquioxide of iron and ammonio-arseniate of magnesia being formed.When only iron and arsenic are present both are preci- pitated in this manner and the supernatant liquid contains merely ammoniacal salts. The employment of sulphate of magnesia in conjunction with ammonia for the determination of arsenic acid has been pro- posed by Level;* and Rose in his Handbook of quantitative analysis strongly recommends this process as very accurate yro-vided certain precautions are adopted in preparing the arseniate of magnesia and ammonia for the determination of its weight. Rose also applies this method to the separation of arsenicfrom antimony and states that it is the best process known furnishing very accurate results if conducted with care. I was led in the first instance without previous acquaintance with the results of Levol and Rose to institute experiments with a view to the application of magnesia to the determination * Ann.Ch. Phys, [3] xvii 501. FIELD ON THE of arseaic in the above manner and to submit to a comparative examination the precipitates produced by baryta lime and msg-nesia in an ammoniacal solution of arsenic acid for the purpose of ascertaining the relative values of the alkaline earths in this branch of analysis. The present communication contains a sum-mary of these experiments Arseniate ofBaryta.-When chloride of barium is added to a solution of arsenic acid in ammonia the latter being in excess a copious precipitate of trisarseniate of baryta is formed consisting of 3Ba0,As05. According to Graham this substance attracts a small quantity of carbonic acid from the air.B er z eliu s says it is very slightly soluble in water somewhat more soluble in aqueous ammonia; and Laugier informs us that the solubility does not seem to be increased by the presence of ammonia potash or soda salts.* My own experiments differ considerably fiorn the above. The arseniate of baryta was found to consist after very careful analysis of- Baryta . . 66.39 Arsenic Acid 33.32 99*71 or 3BaO,AsO Calculated. 3Ba0 . . 66.65 Arsenic Acid 33.35 100~00 The baryta was precipitated from the solution of the arseniate in hydrochloric acid by sulphate of soda. After deoxidation by sulphurous acid the arsenic was estimated as tersulphide-10.00grs.3BaO,AsO,= 6.639 BaO and 3.55 ASS,-3-33AsO,.10*00grs. arsenious acid were converted into arsenic acid and ammonia added in excess. The addition of chloride of barium caused a precipitate which after washing with weak ammonia until no chlorine could be detected in the filtrate gave on desicca- tion at 300’ Fahr.-34.52 grs. 3BaO,AsO,. 34*57grs 3BaO,AsO,. Calculated. Arseniate of baryta loses all its water at a temperature a little above 212O. After diying in a water-bath for several hours 18.90 grs. only lost 0.05 on ignition. After drying upon a sand- * Gmelin’s Handbook vol. iv 304. ARSENIATES OF LIME BARYTA AND MAGNESIA. bath at a temperature which never exceeded 300°,in a platinum crucible no loss was experienced on heating to redness.With regard to the solubility of this salt the following experi- ments were performed 10-00grs. 3BaO,AsO digested with 2000 grs. cold distilled water for forty-eight hours lost 1-10gr. The filtrate gave 0*9gr. sulphate of baryta on addition of sulphate of baryta. 10.00id.digestedwith 2000gru. solution of chloride of ammonium (containing 100 grs. ofthe dry salt) lost 39352. The filtrate gave 341 sulphate of baryta. 10.00id. digested with solution of ammonia formed by adding 200 grs. ammonia sp. gr. *880 to 1800 grs. water lost 0.06. Sulphate of soda hardly produced any precipitate of sulphate of baryta in the filtrate. From this it appears that although very soluble in chloride of ammonium and moderately so in water arseniate of baryta is very insoluble in aqueous ammonia And even (as in most cases in analysis) when chloride of ammonium is present the addition of excess of ammonia prevents in a great measure the solubility of the baryta salt.And it may be mentioned here that ammonia determines the insolubility both of the magnesia and lime arseniates in the same manner precipitating the ammonio-arseniates from their solution in chloride of ammonium Chloride of barium can be advantageously employed for the detection of small quantities of arsenic acid when in combination with copper and other metals. 100*@0 grs. of copper free from sulphur were placed in a flask together with 0'20 grs. arsenious acid. After solution in nitric acid and addition of excess of ammonia sulphate of magnesia caused a precipitate which settled at the bottom of the vessel after 12 hours.On filtering the small precipitate arsenic was readily detected by the usual test. Minerals containing sulphur arsenic and nickel can be very neatly and correctly analysed by the employment of a baryta salt. By adding chloride of barium in excess to the acid solution sulphate of baryta is precipitated and after filtra-tion and addition of a large excess of ammonia all the arsenic is precipitated provided sufficient chloride of barium has been employed. It is necessary to add this reagent to the acid solution and therefore to decompose the arseniate of nickel in hydro- chloric acid. There is no fear of any nickel remaining with the FIELD ON THE mseniate of baryta which is a dense white powder and can be thoroughly and expeditiously washed.For this purpose of course tt weak solution of ammonia is employed instead of pure water. Arseniate of Lime and Ammonia.-When diarseniate of am-monia potash or soda is added to a solution of chloride of calcium trisarseniate of lime is formed and the supernatant liquid becomes acid. When chloride of ammonium trisarseniate of ammonia and lime-water are mixed together crystals of arseniate of ammonia and lime 2CaO,NH,O,AsO are found." I have met with two distinct precipitates during the course of my experiments which possessed the following characteristics 10.00 grs. arsenious acid were converted into arsenic acid ammonia added in very slight excess and subsequently solution of chloride of calcium.The precipitate after prolonged washing with water and dessication at 300° weighed 15.50 grs. and on ignition lost only 0*80gr. yielding 14-70of a compound of arsenic acid and lime. A large quantity of arsenic was detected in the filtrate. 10.00 grs. of arsenious acid were similarly treated and the preci- pitate washed with weak ammonia; after drying at 300° it was found to weigh 18*24grains and 17.02 on ignition. The precipi- tates in both instances seemed to consist of different substances one of a very slightly crystalline powder the other of tolerably large needles united together in stellated masses. Although in the latter experiment the 10 grs. of arsenious acid yielded 17.02 of lime salt and by calculation it should yield 17-10 diarseniate of lime (2CaO,AsO,) the loss by ignition was so small and so much less than it would have been had an atom of ammonia been expelled as to lead me to doubt very much if the precipitate were the arseniate of lime and ammonia notwithstanding that it evolved much ammonia on ebullition with potash.In order to analyze the residue left upon ignition advantage was taken of the fact that arseniate of lime is perfectly decom- posed by a boiling solution of oxalate of ammonia. It is better however to dissolve the arseniate in hydrochloric acid add the oxslate and finally excess of ammonia and boil for some time. The whole of the arseniate is found in the filtrate and the lime in the precipitate as oxalate of lime.The 14.70 grains boiled in this manner gave on ignition 8.00 carbonate of lime= 5.48 lime and the 17-02 grs. gave it Gmdin's Handbook vol. iv p. 304. ARSENIATES OF LIME BARYTA AND MAGNESIA. 1lW56CaO,CO,= 6.47 CaO. Sulphate of magnesia was added to the filtrate and after standing some time the ammonio-mapesian ameniate collected and the arsenic estimated. Composition of Diarseniate of Lime 2CaO,AsO Lime Salt Calculated. Found. Lime . . 4.82 Lime . . 5-48 Arsenic Acid 9-88 Arsenic Acid 9.27 14.70 14-75 Lime . . 5-57 Lime . . 6.47 Arsenic Acid 11.45 Arsenic Acid 10.44 17.02 16.91 It is evident from the above analyses that the salt under con-sideration is not diarseniate of lime but probably a mixture of diarseniate and triarseniate the former produced by the decom- position of the ammonia compound thrown down from the original solution in company with the triarseniate.The loss by ignition as well as the quantities found of arsmic acid and lime leave no doubt that this is the case. 3Ca0 As05+2Ca0 NH,0,As05 upon ignition=3CaO,AsO5 + 2Ca0 AsO,. And the relative numbers are as follows Found. As 396 370 : 18*24 17.04 .. 17.02 , 396 370 : 15-50 14.47 -.. 14.70 And with regard to the quantities of lime and arsenic acid 3CaO,AsO,+ 2CaO,AsO, Calculated. Found. Lime . . 5.56 Lime . . 5-48 Arsenic Acid 9-14! Arsenic Acid 9-27 14.70 1475 Lime . . 6.44 Lime . . 6-47 Arsenic Acid 10.58 Arsenic Acid 10.44 17.02 16.91 FIELD ON THE It is not at all surprising that when ammonia is not in great excess that trisarseniate of lime should be formed as even diarseniate of ammonia produces this compound.It is however rather singular that when in combination with the diarseliiate of lime the trisarseniate does not lose arsenic acid upon ignition which it does when heatedper se.* The arseniate of lime and ammonia can be readily formed by adding ammonia and trisarseniate of ammonia to a solution of chloride of calcium. The precipitants must be in considerable excess This salt crystallises from weak solutions in large needle- shaped crystals and from more concentrated solutions it separates as a white crystalline mass. Like the baryta compound it is more soluble in water than in aqueous ammonia and soluble to a great extent in chloride of ammonium.It retains an atom of water at 212O and consists at that temperature of 2Ca0 NH,O AsO + HO. 10.00 grs. in 200OdO0 grs. water lost 0.40 grs. 10-00 grs. , , grs. weak ammonia lost 0.02 grs. 10.00 grs. , , grs. chloride of ammonium contain- ing 100 grs. dry salt lost 8.300.f This salt loses its atom of water at a temperature a little beyond 212O. Dried at 280’. Found. Calcnlsted. 18.25 on ignition gave 15-80 15.83 11.90 I> 9 10.25 10.32 13.20 >Y 9 11.48 11-45 The residues were analysed separately and found to consist of 2Ca0 AsO, as the following numbers will show :-Calculated. Found. Lime . . 5.18 5.01 Arsenic Acid 10.62 10.49 15.80 15.50 * The cold solution of this salt in chloride of ammonium evolves ammonia copiously on ebullition with formabion of diarseniate of lime which still remains dissolved in the liquid.According ta Wach diarseniate of lime when in solution of ammoniacal salts becomes converted into arseniate of lime and ammonia but this is decomposed on boiling with a re-formation of diarseniate of lime. t The trisarseniate of lime is decomposed when ignited alone (Simon) Gmelin’a Handbook vol. iv p. 301. ARSENIATES OF LIME BARYTA AND MAGNESIA. Calculated. Found. Lime . . 3.41 3.38 Arsenic Aoid 6-84 6-75 10.25 10.13 Lime . . 3-76 3.64 Arsenic Acid 7.72 7-69 11-48 11.33 Arseniate of Magnesia and Ammonia.-This salt is thrown down in great purity when arseniate of ammonia and excess of ammonia are added to a soluble salt of magnesia.In applying this salt to the estimation of arsenic Levol recom- mends its ignition and the determination of the arsenic as pyro- arseuiate of magnesia (2Mg0,AsOJ. But Rose has shown that a portion of arsenic acid is reduced at a high temperature by ammonia and volatilised so that a loss in weight due to that reaction is sustained in proportion to the necessary duration of the ignition. It is stated by the latter chemist that the weight of the double salt may be determined either by its desiccation in wacuo in which case its composition is represented by the formula 2Mg0 NH,O AsO +l2H0 or at 212' F. when it consists of- 2Mg0 NH,O AsO +HO. Like the corresponding lime salt it loses its water at a slight increase of temperature; at 180" F.it retains three atoms of water and it looses its ammonia between 500" and 600'. Its formula may be expressed at various temperatures as follows :-When dried in vaeuo 2Mg0,NH40,&0 +l2HO. From 180" to 200' 2Mg0,NH40,As0,+3H0. At 212 .. 2Mg0,NH40,As0 +HO. At 300 .. 2Mg0,NH40,As05. At 600 .. 2RXgO,AsO,. In the estimation of arsenic from this compound it is therefore very necessary to be extremely cautious in its desiccation. Speci-mens dried upon a filter placed above a sand-bath with a thermo- meter suspended at the same distance from the heated surface (the mercury never rising above 300') were found to have lQat the whole of their water in from four to five hours. 14 FIELD ON THE The following experiments were tried regarding the solubility of this double salt :-10.00 grs.digested with 2000 grs. water lost 0.28grs. 10.00 , , ammonia , 0.14 grs. JJ 10.00 9 , , NH,Cl , 1.90 grs. Ammonia precipitates the salt from its solution in chloride of ammonium. The following diagram will give in a tabular form the solu- bility of the arseniates of lime and ammonia magnesia and am- monia and arseniate of baryta in water ammonia and chloride of ammonium. 10 gw. 10 grs. 10 gra. 2MgO,NH,O,AsO, HO. 2CaO,NH,O,AsO,,HO. 3BaO,AsO,. ---I_--1 Water .... 2000 .. 0.28 2000 .. 0.40 2000 .. 1-10 Ammonia.. . 2000 .. 0-14 2000 .. 0.02 2000 .. 0.06 Chloride of Ammonium 2000 .. 1.90 2000 .. 830 2000 .. 3-85 From the foregoing experiments the relative advantages and disadvantages in the employment of the salts of the alkaline earths may be estimated and- 1.The precipitation of arsenic acid by a soluble salt of lime. This is disadvantageous when sulphuric acid is present in the liquid. Sulphate of lime would be precipitated and require long and protracted washing for its entire separation. Neither coiild the arseniates be freed from the sulphate by dilute hydrochloric acid 3s the latter is especially soluble in that menstruum. Another objection is the uncertainty of the compound precipitated which would render its analysis indispensable when the quuntity of arsenic as well as its abstraction from other bodies is desired. On the other hand its great insolubility in ammonia renders the employment of a lime-salt in certain circumstances very advan- tageous.2. The employment of a soluble baryta salt. The objection regarding the sulphuric acid cannot be urged as in the case of the lime-salt Sulphate of baryta is easily separated from its arseniate and indeed the whole of the sulphuric acid could be ARSEPU’IATES OF LIME BARYTA AND MAGNESIA. removed by chloride of barium previously to the introduction of the ammonia. When chloride of ammonium does not exist in any very great quantity in the solution the employment of baryta may be resorted to advantageoixsly. 3. By a magnesia salt. Magnesia possesses many advantages over lime and baryta which can be easily appreciated in practice.The ammonia-magnesian arseniate after standing some hours settles down in a heavy crystalline mass so hard and dense that the supernatant liquid may be frequently decanted off and the crystals drained thus rendering the subsequent washing very expeditious and easy. There is no fear of sulphuric acid being in the precipi- tate which possesses the advantage of being very definite in com-position and illsoluble in weak ammonia. From my own experi- ence I should recommend magnesia above baryta and lime if it were only for the greater facility of its management in chemical analysis. The following estimations were invariably made with a magnesian salt :-10.00 gr. arsenious acid converted into arsenic acid and preci- pitated with a magnesian salt gave 18.87 2MgO,NH,O AsO +HO.Calculation 19.19. 10.00 gr. gave 18.94. 10.00 gr. AsO, and 1gr. copper dissolved in nitric acid gave 19-14 2Mg0 NH,O AsO +HO =9.97 AsO,. 1.00gr. AsO +10.00gr. Cu gave 1-90 2MgO,NH,O AsO +HO; and when 0.19 gr. AsO,=0*076 arsenic was boiled with NO, and 100.00 of copper a crystalline deposit of the ammonia-magnesian itrseniate was observed on the sides of the flasks after standing for 24 hours. On drawing off the copper and washing the crystals with weak ammonia they mere dissolved in HC1 and a clear yellow pentasulphide of arsenic obtained on addition of a few drops of sulphide of ammonium and afterwards excess of hydrochloric acid. It has been before observed that nickel can be easily and per- fectly separated from arsenic by the introduction of a msgnesiau salt.In the following experiments with a given weight of ore (?) the arsenic was not estimated as the ore contained iron which was precipitated w ith the ammonio-magnesian arseniates. The filtrate was precipitated by sulphide of ammonium the sulphide dissolved and the nickel thrown down as oxide by potash FIELD ON THE The first experiment gave . . 6.127 oxide of nickel. second , 9 6.00 9 third , Jf 5-95 77 The Separation of Arsenic from Antimony by means of a soluble magnesian salt in the presence of ammonia is very strongly recommended by Rose; and my own experiments fully confirm the statements of that chemist. The method described in detail in his Handbook succeeds perfectly; I may state however that as the tartaric acid added to the solution of the oxidised metals for the purpose of retaining the antimony in solution is sometimes liable to occasion the separation with the arseniate of a small quantity of tartrate of magnesia and ammonia I have found it advisable to substitute sulphate of ammonia for the chloride of ammonium added previous to the ammonia.The double tartrate is very readily soluble and the double arseniate particularly in- soluble in the sulphate of ammonia. The following analyses will show the correctness of the method :-Taken. Found. Arsenious Acid 5.00 2MgO,NH,O,AsO,+ HO 9*54=4*97AsO,. Antimony . 5.00 93 Arsenious Acid 5.00 >> 9*52=4-95 AsO,. Antimony . 5.00 The antimony was not determined in the above analyses; in those below the arsenic was estimated as before sulphide of ammonium added in excess to the filtrate until the sulphide of antimony was completely re-dissolved.It was afterwards precipitated by hydro- chloric acid. The precipitate was dried at 212O weighed and a portion examined quantitatively for sulphur the weight of which was deducted and the quantity calculated from the whole precipitate. Taken. Found. Arsenious Acid . 10.00 = Arsenic 7-57 .. 7*39 T Antimony . . . -6-80 .. 7.04 -14.37 14.43 Arsenious Acid 12.40 = Arsenic 9.39 .. 9-12 -Antimony -12.40 .. 12-23 -21.79 21.35 ARSENIATES OF LIME BARYTA AND MAGNESIA. Taken. Found. Arsenious Acid 1-00 . = Arsenic 0.75 .. 0-68 Antimony .. . .= 10.00 .. 9-89 -10.75 10.57 Arsenious Acid 10.00 . = Arsenic 7.57 .. 7-42 Antimony . . . .-1.00 .. 0.98 -8.57 .. 8.40 Arsenious Acid OS1O gave a crystalline precipitate of 2MgO,NH,O,AsO +HO on the APLtim"ny 5*00 sides of the glass Arsenious Acid 5-00 filtrate gave an orange-red preci- Antimony . O*lO} pitate of sulphide of antimony. Arsenic cannot be separated from tin by means of magnesia. Ammonia produces no precipitate in the solution of the bichloride of tin when tartaric acid is present but a precipitate is produced by the introduction of magnesian salts which appears to contain both tin and magnesia but which so far as I know has not been examined. Hopes may be entertained that antimony may be separated from tin by taking advantage of this fact.The subject is now under consideration and promises satisfac. tory results though rat her difficult and laborious.