Trimetaphosphoric acid is obtained as the sodium salt Na3P3O9 when any other metaphosphate of sodium is fused and slowly cooled, then dissolved in a slight excess of warm water, and the resultant solution evaporated. The crystals contain 6 mol. H2O, and dissolve in four parts of water. An acid reaction is only obtained, as with the preceding salt, after prolonged boiling with water. The acid is a true analogue of nitric acid, because all its metallic salts are soluble.
Hexametaphosphoric acid. Fleitmann so named the ordinary metaphosphoric acid (glacial) which attracts moisture. The deliquescent sodium salt is obtained, like the trimetaphosphate, only by rapid cooling. It is also formed by fusing silver oxide with an excess of phosphoric acid. The sodium salt is soluble in water, and gives viscous, elastic precipitates with salts of Ba, Ca, and Mg. Lubert (1893) obtained salts of Ag, Pb, &c.
Jawein and Thillot (1889), who investigated the sodium salts of metaphosphoric acid by Raoult's method, came to the conclusion that the salts of di- and tri-metaphosphoric acid behave in such a manner that their molecule must be represented as non-polymerised NaPO3, whilst those of hexametaphosphoric acid behave as (NaPO3)4. At all events, the series of salts which Fleitmann and Henneberg regard as monometaphosphates—i.e. as non-polymerised—are most probably the most polymerised, because they are insoluble.
According to Tamman's researches, vitreous metaphosphoric acid contains a mixture consisting chiefly of two varieties, differing in the solubility and degree of stability of their salts. The least stable corresponds to Fleitmann's hexa-acid, and gives three isomeric salts. Tamman came to the conclusion that there exist polymers also in the form of penta-, ortho-, and deca-metaphosphoric acids. Without going into details upon this subject, I do not think it superfluous to point out that the undoubted capability of metaphosphoric acid to polymerise should be connected with its faculty of combining with water, whilst the degree of polymerisation and the number of polymeric forms cannot yet be considered as sufficiently explained.
[21 bis] The bibasity of H3PO3, established by Würtz, has been proved by many direct experiments (see, for instance, Note [22]), among which we may mention that Amat (1892) took a mixture of the aqueous solutions of Na2HPO3 and NaHO and added absolute alcohol to it. Two layers were formed; the upper, alcoholic, contained all the excess of NaHO, whilst the lower only contained the salt Na2HPO3, which was therefore unable to react with the excess of NaHO. Amat also obtained NaH2PO3 by saturating H3PO3 with soda until he obtained a neutral reaction with methyl-orange. The replacement of one atom of H by sodium here, as in phosphoric acid (Note [16]), gives more heat than the replacement of the second atom. For the third atom there is no formation of a salt, and therefore no evolution of heat. The monometallic salts—for example, NaH2PO3—or the ammonia salts, when heated to 160°, give, as Amat had previously shown, a salt of bibasic pyrophosphorous acid, Na2H2P2O5.
[22] Phosphorous acid, when subjected to the action of nascent hydrogen (zinc and sulphuric acid), evolves phosphine, and when boiled with an excess of alkali it evolves hydrogen (PH3O3 + 3KHO = PK3O4 + 2H2O + H2); owing to its liability to oxidation, it is a reducing agent—for instance, it reduces cupric chloride to cuprous chloride, and precipitates silver from the nitrate and mercury from its salts.
These reactions are perhaps connected with the fact that in this acid one atom of hydrogen should be considered as in the same condition as in phosphuretted hydrogen, which is expressed by the formula PHO(OH)2, if we represent it as PH4X, with the substitution of two of the hydrogen atoms by oxygen and of HX by two of hydroxyl. The direct passage of phosphorous chloride into phosphorous acid would, however, indicate that all the three atoms of hydrogen in it occur in the form of hydroxyl, because no difference is known between the three atoms of chlorine in PCl3—they all react alike, as a rule. However, Menschutkin, by acting on alcohol, C2H5OH, with phosphorous chloride, obtained hydrochloric acid and a substance P(C2H5O)Cl2, and from it by the action of bromine he obtained ethyl bromide, C2H5Br, and a compound PBrOCl2, which proves, to a certain extent, the existence of a difference between the three atoms of chlorine in phosphorous chloride. If we turn our attention to the formation of phosphine by the ignition of phosphorous acid, we see that 4PH3O3 only evolve 3H in the form of PH3, and therefore the residue—that is, 3PH3O4—will still contain one hydrogen of the same nature as in phosphine, because in 4PH3O3 we should recognise four such hydrogens as in phosphine. We arrive at the same conclusion by examining the decomposition of hypophosphorous acid, 2PH3O2 = PH3 + PH3O4. In the two molecules of the monobasic hypophosphorous acid taken, there are only two atoms of hydrogen replaceable by metals, whilst in the molecule of the resultant phosphoric acid there are three. Perhaps relations of this nature determine the relative stability of the dimetallic salts of orthophosphoric acid.
[23] Calcium hypophosphite is used in medicine. According to Cavazzi, a mixture of sodium hypophosphite, NaH2PO2, and sodium nitrate explodes violently.
[24] Fluorine and bromine give PX3 and PX5, like chlorine. With respect to iodine PI5 is, in a chemical sense, a very unstable substance, and generally phosphorus tri-iodide only is formed (from yellow or red phosphorus and iodine in the requisite proportions). It is a red crystalline substance, fuses at 55°, is easily decomposed by water, forming phosphorous and hydriodic acids, and when heated it evolves iodine vapours and forms phosphorus di-iodide, PI2. This substance may be obtained in the same manner as the preceding by taking a smaller proportion of iodine (8 parts of iodine to 1 part of phosphorus, whilst the tri-iodide requires 12·3); it also forms red crystals, which melt at 110°. When decomposed by water it not only gives phosphorous and hydriodic acids, but also phosphine and a yellow substance (a lower oxide of phosphorus). In its composition di-iodide of phosphorus corresponds with liquid phosphuretted hydrogen, PH2, and probably its molecular weight is much higher: P2I4 or P3I6, &c. As the iodine compounds of phosphorus give hydriodic and phosphorous acids with water, and as both these substances are reducing agents in the presence of water (and hydrates), iodide of phosphorus also acts as a reducing agent.
[25] In a liquid state the density of phosphorous chloride at 10° = 1·597, and therefore its molecular volume = 137·5/1·597 = 86·0, and that of phosphorus oxychloride is equal to 153·5/1·693 = 90·7; hence the addition of oxygen has produced considerable increase in volume, just as in the conversion of sulphur dichloride, SCl2, into sulphuryl chloride, SOCl2, the volume changes from 64 to 71. It is the same with the boiling-points; phosphorus trichloride boils at 70°, the oxychloride at 100°, sulphur dichloride at 64°, and sulphuryl chloride at 78°—that is, the addition of oxygen raises the boiling points.