1-136
THE JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY
bpneblack, there would be a saving in overhead expenses, and, as revivifying processes develop, it may prove cheaper in revivification than char because of the smaller weight handled, thus diminishing operating expenses also. The great result to achieve is the production of an ash-
Vol. 14, No. 12
absorbing carbon, and from all present indications it will consist of a porous mineral foundation overlaid with a highly activated carbon, and such a material will have a great advantage if it can be handled in filter presses as the existing carbons are handled.
The Determination of Aluminium as Phosphate’,’ By G. E. F. Lundell and H. B. Knowles BUR&AU OF STANDARDS, WASHINGTON, D. C.
ANY methods of Determinations of aluminium as phosphate are generally in amounts (1 to 5 mg.) of error. Acceptable determinations are possible when onlr, a few analysis call for the aluminium are involved, it milligrams of aluminium are inooloed, and high values are the determination of is far from reliable where rule when more than 1 to 5 mg. of aluminium are precipitated by larger amounts of aluminaluminium by precipitation, ordinary procedures. immoderate washing of the precipitate or ium are in question, or when an4 weighing as phosphate, the use of a weak. acid wash leads to low values, as does also preiron or titanium or both either alone or together cipitation with only a moderate excess of the precipitant, or preare also present and precipiwith iron and titanium. cipitation in alkaline solution. Under these conditions iron betated as phosphates. The calculation of aluminhaves like aluminium, while titanium inoariably yields low oalues. ium is ordinarily made on the basis of the Eomposition EXPERIMENTAL AlP04, with corrections, The results of the investigation can best be summarized when necessary, for iron and titanium as separately determined and calculated on the basis of the normal phosphates FeP04 as in Tables I, 11, and 111. The aluminium, titanium, and and Ti3(P0&. Most procedures call for precipitation in acid iron solutions were all standardized by proceeding as in the solution, although a few specify an alkaline one; the array of method of W. Blum,6 with correction for silica in all cases. wash solutions is equally broad; hot water, hot or cold The phosphate solutions used for precipitation were standammonium nitrate, and hot dilute acetic acid are all men- ardized gravimetrically as magnesium pyrophosphate. All tioned. I n most cases the methods are given with no precipitates were carefully ignited and finally heated to constatement as to their accuracy; in a few instances the meth- stant weight at approximately 1000” C. In many cases ods are rated as of the highest accuracy; while in three re- macerated paper was used in an attempt to facilitate the searches, at least, attention has been called to the limita- filtering and washing of the precipitate and to obtain a finer tions of the phosphate method. OneS states that when grained ignited residue and greater ease in expelling any aluminium is boiled for 40 min. with an excess of sodium volatile matter. phosphate in acetic acid solution, the phosphate which is obIn Table I an aluminium chloride solution was used in tained approximates the formula 7A1203.6P205 more nearly Expts. 1 to 8 and 15 to 22, and an aluminium nitrate soluthan AI203.P2O5. Another4 claims that the phosphate pre- tion in Expts. 9 to 14. The precipitant in Expts. 1 to 8 cipitated by boiling a like solution for 5 min. contains P z O ~ was sodium ammonium hydrogen phosphate, in Expts. in excess of that called for by the formula A1208.P~Os. The 14, disodium hydrogen phosphate, and in the others, diamthird6 points out the difficulty in properly washing the monium hydrogen phosphate. Precipitation was in acetic precipitate. acid solution, except in Expt. 10 where the solution was made In view of the foregoing facts, it is considered desirable to just alkaline to methyl orange with ammonia, and in Expt. present the following data which embody tests carried out 11 where an excess of 5 cc. of ammonium hydroxide (sp. gr. a t this Bureau in order to determine the suitability of the 0.90) per 100 cc. was used. method for work in hand, and which demonstrate that, The data of Table I show that all the methods proposed while the method is quite satisfactory where only small for the determination of aluminium as phosphate are sub1 Received May 22, 1922. ject to error; those are of the least concern where only small 2 Published by permission of the Director of the Bureau o{ Standards. amounts (1 to 5 mg.) are involved. Low values are the rule: 8 T. M. Drown and Alex. G. McRenna, Chem. News, 64 (1891), 194. (1) when less than five times the theoretical requirement 4 Iq the “Report of the Committee on Research and Analytical Methods of the precipitant is used; (2) when the precipitation is per-Phosphate Rock,” THISJOURNAL, S (1911), 787, the conclusions are given: “In order that the ignited phosphates may contain enough phosphoric acid formed in alkaline solution; (3) when the precipitate is washed to form the normal phosphate, it is absolutely necessary that the second preimmoderately with hot or cold water or ammonium nitrate cipitation, either in the acetate, Glaser, or thiosulfate method, be made in solution; or (4)when dilute acetic acid is used as a washing the presence of an excess of phosphoric acid.” “It is impossible to exactly medium. High values are invariably obtained with moderate wash out the excess of phosphoric acid, leaving only the normal iron and aluminium phosphates.” “It is possible, though hardly practicable, to remove amounts of aluminium when over five times the theoretical by ignition all excess of phosphoric acid from a precipitate of aluminium of precipitant is employed in acetic acid solurequirement phosphate without reducing the normal aluminium phosphate.” “It is tion. This holds true regardless of whether the precipitate difficult, and the results are very uncertain at best, to remove the excess is washed moderately (with 300 cc. of solution) or until of phosphoric acid from a precipitate of iron phosphate without reducing, or, rather, driving off some of the phosphoric acid from the compound.” it is free from chlorides with hot water or hot or cold am“It is the opinion that any of the methods based on ignition of the phosphate monium nitrate solution, or of what ignition temperatures These qualiis subject to the above-mentioned errors and uncertainties.” are employed. THIS fications are not embodied, however, in the final committee report, J O U R N A7~ (1915), , 446. There seems to be little hope that a normal phosphate J. M. Camp, Iron Age, 6S (1900), 17, says: “One disadvantage of can be obtained by any method. The formula 3Al203.the phosphate method is that there is no end-point to the washing of the pre-
M
cipitate, i t being slightly soluble in the wash water.”
6
J . A m . Chem. Soc., 88 (1916), 1282.
-
T H E JOURNAL OF
Dec., 1922
37
TABI,E~-DETERMINATION OF ALUMINIUM A S PHOSPHATE PZOSAdded EXPT. Theory NO. A1z0a:Pz06 1 1 2 1 3 3 4 3
5 6 7 8
5 5
9
10 lo 10
10 11
10 10
12
AlzOs Taken AMOUN?. O F WASHING MEDIUM^ 500 cc. hot HzO 500 cc cold 2 per cent NHiNOa 500 cc. hot Hz0 500 cc. cold 2 per cent NHaNOa 500 cc hot HzO 500 cc cold 2 per cent “4N03 500 cc hot 1320 500 cc cold 2 per cent NHaNOs
0 0
0 0 0 0 0
0
G. 0861 0861 0861 0861 0861 0861 0861 0861
AhOa Found, Calcd. on Basts of AlPOa, G.’
-Brat- ’ G.
0.07701 0.0776’ 0.0846 0.0851 0,0874 0.0867 0 0890 0,0907
-0 0091 0.0085
-0 0015 -0 0010 + O 0013 +O +O
+O
Per cent
-10.8 1 9.9 1.7 1 - 1 2
0006
0029 0046
+
5.3
1
300 cc. hot 2 per Cent Acetic
0 0485
0.04251
-0 0060
-12.4
300 cc. hot Hz0 300 cc. hot Hz0
0 0485 0 0485
0.04541 0.03671
-0.0031 -0 0118
-64 -24.3
1
300 cr. hot 5 per cent hTHaNOs
0 0485
0.0446
-0 0039
- 8.0 1’
13 14
5 20
300 cc. hot 5 per cent NHaNOs 300 cc. hot 5 per cent NHaN03
0 0485 0 0485
0.0516 0.0523
15 16
10 10
500 cc. hot 5 per cent ”4h’Oa 1000 cc. hot 5 per cent NHdNOs
0 0471 0 0471
0.0482 0.0459
17 18
10 10
2000 cc. hot 5 per cent NHaNOa 2000 cc. hot HzO
0 0471 0 0471
0.04451 0.03942
-0,0026 - 0.0077
19 20 21 22
10 10 10 10
175 cc. hot 5 per cent NHiNOa 175 cc. hot 5 per cent NHiNOa
0 0019 0,0094
0.0019 0.0096 0.0964 0.1930
+o. 0000 +0.0002
J
0 0843
+O 0031 + O 0038
+o .001 1 -0.0012
+- 2.3 2.6 -
5.5 -16.4
‘I
+ 02.1 1\ + I
“Peters.’,’, Blair “Chemical Analysis of Iron 8 t h Ld., J. B. Lippincott Co., p.’ 274. Essentially the same in Ibbotson “Chemiypl Analysis of Steel Works’ Material, p. 104, Longmans Green & Company “Methods of Chemists of U. 3. Steel Cor oration for Sampling and Analysis of &rroalloys, etc.,” 1920 ed., p. 42 I b i d , alloy steel, 1921 ed., p. 75 Slightly more alkaline than t h e above Expt. 12 illustrates the effect of insufficient phosphate and Expts. 14 t o 22 embody slight ’ modifications of Dxpts. 13 a n d 14 Expts. 13 and 14 were performed by the method recommended b y the Fertilizer Division of t h e American Chemical Society, THIS JOURNAL,7 (1915),446. Scott “ S t a n d y d Methods of Chemlcal Analyses D. Van Nostrand Company, 2nd ed., p 320 T h e washings in Expt. 17 contained only traces of AIz03, while those of Expt. 18 contained less than 1 mg. of A1203 As in previous method, only precipitates were washed until entirely free from chlorides
+0.0021 + 2 2 550 cc. hot 5 per cent NHiNOa 4-0.0044 213 0 1886 600 cc. hot 5 per cent NHaNOs 1 Considerable aluminium in filtrate. 2 T h e weighed residue obtained in Expt. 17 gave when analyzed 45.33Per Cent - 4 1 2 0 3 and 53.88 per cent P p 0 1 while t h a t obtained in Expt. 18 gave 50 16 per cent A120aand 48.58 per cent Pa06 a s against t h e theoretical Values of 41.79 Per cent Alios and 58.21pdr cent PzOa for AlPOa. These values in conjunction with t h e nenligible alumina recbveries from t h e washings, illustrate t h e Progressive hydrolysis of the precipitate and the washing o u t of phosl phorii: acid.
4P205. 18H20represents very closely the composition of aluminium phosphate precipitates, obtained by the use of an excess of phosphate in either acetic acid or ammoniacal solution, with preliminary drying with blotting paper, final drying in the air, and correction for all impurities. Washing this compound progressively removes phosphorus, with little, if any, aluminium, even after the “AlPOI” stage has been reached. Expts. 19 to 22 show that desirable modifications of the official method’ consist in precipitating in the presence of macerated paper, and final washing of the phosphate until it is free from chlorides. TABLE 11-BEHAVIOR OF IRON IN THE OFFICIALMETHOD (Determinations made with a solution of ferric sulfate a s in the official method,’ except as t o varying the amount of diammonium hydrogen phosphate precipitant a s noted in Column 1 ) FelOa Found PzOa Added Fez03 Calc. on Basis Taken of FePOa -Error-FezOcEXPT. Theory G. G. Per cent No. A ~ z O ~ : P Z O ~G. -5.1 0.0530 0.0503 -0,0027 1 1 5 5 20 20
0.0533 0.0582 0,0505 0.0537 0.0564
0.0503 0.0607 0.0519 0.0565 0.0594
-0.0030 +0.0025 +0.0014 +0.0028 +O. 0030
-5.6 $4.3 +2.8 +5.2 +5.3
The data show that iron behaves like aluminium, and that it will increase the errors in any aluminium determinations where both are precipitated and aluminium then obtained by difference after a separate determination of the iron and deduction as calculated Pepo4. For example, in such a procedure a rock containing 10per cent of Fez03and no aluminium would appear to have somewhere in the neighborhood of 0.20 per cent of A203 on the basis of Expts. 5 and 6. In the official method, a known amount of iron is added to the solution of the material and a like amount carried through as a blank, which is afterward subtracted together with the theoretical FeP04 equivalent of the iron in the maferial as separately determined. As the actual blank is quite small, a more satisfactory procedure would lie in computing the 7 Recommended b y t h e Fertilizer Division of the American Chemicsl Society, THISJOURNAL, 7 (1915),446. See Table I, Expts. 13 and 14 of this article.
FeP04 equivalent of the iron in the material on the basis of the actual performance of the known amount of iron in the blank. TABLE 111-BEHAVIOR OF TITANIUM BY THE OFFICIAL METHOD (Determinations made b y precipitating an acetic acid solution of titanic sulfate as in the official method,’ except as t o varying the amount of diammonium hydrogen phosphate precipitant, as noted in Column 1) Ti02 Found Calc. P z O Added ~ TiOz on Basis of Theory Taken Tis(PO4)a -Error TiOzA1203.PzOfi G. G. G. Per cent 1
1 5 5 20 20
0.0528 0.0528 0.0528 0 0528 0.0.528 0.0528
0.0424 0.0422 0.0456 0.0456 0,0485 0.0473
-0.0104 - 0,0106 0.0072
-0,0073 -0 0043 - 0.0055
-19.7 -20.1 -13 6 -13 8 - 8.1 -10.4
The data here show that titanium consistently yields low values in the phosphate method, and that when aluminium is obtained by difference after a separate determination of titanium and deduction as calculated Ti,(PO&, small amounts of titanium will offset the normally high values obtained for aluminium, as in Tables I and 11, while large amounts of titanium will cause low values. Where titanium is present ill of course add to the and ignored, as is usually the case, it w high value obtained for aluminium. The data presented in Tables I, 11, and 111dem that values for aluminium which are obtained by such methods as that of Peters, and the one recommended by the SOCIETY Fertilizer Division of the AMERICANCHEMICAL will normally be high in case more than a few milligrams of aluminium are involved; that the values will be higher still when iron is also involved, as in the latter method; and that the values will be lowered by titanium (provided it is considered) in proportion to the amount present. The degree of Bachelor in Colour Chemistry has been created a t the Manchester College of Technology, Manchester, England. Dr. F. M. Rowe, editor of the Colouv Index now being issued by the Society of Dyers and Colourists, who is in charge of this work, states that students are being encouraged t o carry agt some of their experiments in a miniature plaqt, because although the dyestuffs industry is now staffed to capacity with the large numbers who entered it during the war. new research workers will be needed as dye making develops in Great Britain.
