THE PHOSPHATES OF MAGNESIUM AND IRON' - BY F. K. CAMERON AND J. M. BELL The behavior of phosphates of calcium in phosphoric acid solutions has been discussed in former papers from this labora- toryZ. Two crystalline phosphates of calcium have been shown to exist in equilibrium with solutions containing free phosphoric acid: monocalcium phosphate, which at 25' exists in equilibrium with solutions containing over 3 I 7 grams of phosphoric acid (PzO,) per liter; and dicalcium phosphate, which is stable below this concentration. At very low con- centrations (up to about 5 grams phosphoric acid per liter) the stable solid was found to be a solid solution consisting of lime and phosphoric acid. Both these crystalline phosphates of lime suffer decom- position when water is added to the solid, the resulting solu- tion becoming richer in phosphoric acid and the solid poorer in phosphoric acid than the original compounds. The phos- phates of magnesium are also decomposed by water, yielding solutions which are more strongly acid than the original solid, and the composition of the various acid solutions which can exist in contact with magnesium phosphates have been determined at 25O. In a similar way the phosphates of iron suffer decomposition when treated with pure water and com- positions of various acid solutions in contact with ferric phosphate have been determined at 25O. The composition of the solid in contact with these solutions is shown to be variable, that is, the solids are solid solutions. Phosphates of Mag.nesium.--Solid dimagnesium phosphate was added in excess to solutions of phosphoric acid in varying concentrations. These mixtures were constantly agitated in a thermostat at 25O, and at the expiration of two months 'Read before the American Chemical Society, Ithaca, June, 1906,by per- mission of the Secretary of Agriculture. Cameron, Seidell, and Bell : Jour. Am. Chem. SOC.,27, 1503, 1512 (1905). 364 F. K. Canteroa and J. M. Bell the solutions were analyzed for magnesia and for phosphoric acid. Magnesia was determined by adding ammonium chloride to the solution for analysis and then ammonia was slowly added to the solution until the solution became alkaline; the resulting precipitate was filtered, washed, and weighed as magnesium pyrophosphate. To determine phosphates ammonium chloride was added to the solution for analysis, then ammonia was added slowly with constant stirring until the solution became alkaline, then magnesia mixture was added in excess until all the phosphoric acid was precipitated as magnesium ammonium phosphate. The composition of the solutions is given in Table I and Fig. I. MgO per liter. P,O, per liter. Grams Grams 0.207 0.280 0.553 1.438 2.23 4.73 16..84 1.017 11.19 38.59 1.042 17.33 61.21 I .069 26.09 93.09 1.109 37.40 130.7 1. I44 75.5 281.8 1.285 109.5 439.0 - 122.6 49 8.4 1.470 129.9 546.5 - 140.0 584.0 - 146.8 623.3 1.595- 147.3 625.9 - 150.3 645-8 $1- 155.5 680.7 87. I 779.6 1.626 77.1 809.6 1.644 70.6 835.1 1.654 From the above table it appears that there are three series of solutions which exist in contact with the phosphates \ 10, per liter 0 800 Fig. I of magnesia. With the solutions very concentrated with respect to phosphoric acid the content of magnesia phosphate decreased with increasing quantities of phosphoric acid, but with more dilute solutions both magnesia and phosphoric acid increase regularly. The stable solid in contact with the more dilute solutions was dimagnesium phosphate (MgHP0,.3H20) and in concentrated solutions monomagnesium phosphate (MgH,(PO,),,xH,O). The constant solution contains about 700 grams of phosphoric acid (P20,), and about 160 grams magnesia (MgO) per liter. In all but the last three solutions described in the table the solid phase is dimagnesium phos- phate and in the last three, described in the table, mono- magnesium phosphate. A magnesium phosphate more basic than the diphosphate exists at 25' only in very dilute solutions of phosphoric acid, and over a range of concentration too narrow to be determined with precision. Phosphates of Iron. -The action of water upon the phosphates of iron resembles in a striking manner the action of water upon the phosphates of calcium and magnesium. They undergo the same decomposition whereby the acid is removed at a different rate from the base, resulting in the formation of a more basic precipitate and the formation of a .366 F. K. Cameron and]. M. Be22 solution which is acid. Lachowiczl has investigated the decomposition action of water upon ferric phosphates, and it is shown that the continued leaching of a substance corre- sponding closely to the formula FePO, gave a solution which was acid. Relatively very minute quantities of iron were found in the acid solution. The results of Lachowicz have been confirmed by Cameron and Hurst,’ who have shown that the action of water upon ferric phosphate cannot be considered as one.of mere solution only. If it were a simple solution phenomenon the solution resulting from a mixture of phosphate and water would have the same composition regardless of the quantity of the solid residue, and also the ratio of iron to phosphoric acid in the solution would be the same as in the original solid. In recent experiments in this laboratory solid ferric phosphate, whose composition was not determined, was placed in contact with phosphoric acid solution of varying concen- tration up to 5 percent; after four months constant agitation at 25O, the solutions were analyzed for phosphoric acid and for iron and the density having been determined, the percentage of each component could be calculated. The iron in solution was determined as ferrous sulphide by precipitation from the solution which had previously been made alkaline by an excess of ammonia which dissolves the precipitate formed when the solution is just neutral. The iron was weighed as ferric oxide. Phosphoric acid in the filter was determined by precipitation with magnesium mixture. The solid phase was freed as far as possible from the solution and a weighed quantity was dissolved with hydrochloric acid. Aliquot portions were analyzed by the above method for both phos- phoric acid and lime. The results of these analyses are given in Table I1 and have been plotted in Fig. 2. The lines joining corresponding points in the diagram do not intersect but are approximately parallel, and hence Monatshefte, 13, 357 (1892). Jour. Am, Chem. SOC., 26,885 (1904). The Phosphates of Magizeszzcnz and IYO~Z 367 Phosphoric Pho~~~ic1 Ferric oxide acid Ferric oxide Fe,O,. Density % Fe,O,. p20,. ' Per cent 25 p2°b. Per cent Per cent Per cent -~ 0.942 0.0105 I .0074 13.81 15.11 1.984 0.0205 1.0162 17-90 19.35 2.838 0.0384 I .0244 22.54 23. I I 3.770 0.0611 1.0310 17-73 15-13 4.706 0.0849 1.0383 23.37 20.43 Fig. 2 the solid phase is a solid solution, for it is evident that its composition varies continuously. The ratio of ferric oxide to phosphoric acid in the limiting solid solution is very close to that represented by FePO,, and if any one of these solid solutions be subjected to continued leaching by water, the resulting solid will not be far in composition from FePO,. It is also apparent from the table that phosphoric acid is washed from the solid at a much more rapid rate than is iron, and continual leaching will cause very little loss of iron but I 368 F. K. Camevan and /. M. 3elL will reduce the phosphoric acid content of the solid very rapidly until a limiting solid solution results. With concentrations higher in phosphoric acid than those of the above table definite chemical compounds may be formed, as in case of lime and magnesium. The behavior of iron phosphate in these more concentrated solutions will be in- vestigated ,further. Bureau ofsoils, U.S. Department of Agriculture, Washington, D. C. .