Monographs on Inorganic and Physical Chemistry EDITED BY ALEXANDER FINDLAY, M.A., PH.D., D.Sc. Professor of Chemistry, University College of Wales, Aberystwyth To those engaged in guiding the reading of advanced students of Chemistry, the difficulty of obtaining ade- quately summarised accounts of the progress made in recent years, more especially along certain of the more actively pursued lines of advance, becomes ever more acutely felt. So great has now become the volume of chemical investigation, and so numerous the channels of its publication, that not only the Honours Student but also the worker desirous of undertaking Research in one or other department of his subject, feels it a growing difficulty to become au fait with the present state of the more important and more strenuously cultivated regions of his Science. To assist these two classes of students those reading for an Honours Degree, and those under- taking Research is the main aim of the present Series of Monographs. In this Series of Monographs it is hoped to place before advanced students of chemistry, accounts of certain sections of Inorganic and Physical Chemistry fuller and more extended in scope than can be obtained in ordinary text-books. Exhaustive treatment of the different sub- jects, however, so far as concerns work important in its time but now only of historical interest, will not be attempted ; the chief attention will be given to recent investigations. Arrangements have already been made to publish the following monographs, and should these prove themselves to be of value, others will be issued from time to time. THE CHEMISTRY OF THE RADIO- ELEMENTS. By FREDERICK SODDY, F.R.S., of the University of Glasgow. 8vo. 2s. 6d. net. PER-ACIDS AND THEIR SALTS. By T. SLATER PRICE, D.Sc., of the Municipal Tech- nical School, Birmingham. 8vo. 33. net. ELECTROLYTIC DISSOCIATION THEORY. By J. C. PHILIP, D.Sc., of the Chemistry Depart- ment, Imperial College of Science and Technology, South Kensington. OSMOTIC PRESSURE. ' By ALEXANDER FINDLAY, D.Sc., Editor of this Series. THE PHYSICAL CHEMISTRY OF FLAMES. By J. E. COATES, M.Sc., of the Chemistry Depart- ment, The University of Birmingham. CLAYS. By J. W. MELLOR, D.Sc. INTER-METALLIC COMPOUNDS. By CECIL H. DESCH, D.Sc., of the University of Glasgow. CATALYSIS OF GAS REACTIONS. By D. L. CHAPMAN, M.A., Jesus College, Oxford. THE ELECTRO-CHEMISTRY OF NON- AQUEOUS SOLUTIONS. By JAMES W. MACBAIN, of the Chemistry Department, The Uni- versity, Bristol. CATALYSIS IN LIQUID SYSTEMS. By GEORGE SENTER, D.Sc., of St. Mary's Hospital Medical School, London. LONGMANS, GREEN & CO., 39 PATERNOSTER Row, LONDON, E.G. NEW YORK, BOMBAY AND CALCUTTA. MONOGRAPHS ON INORGANIC AND PHYSICAL CHEMISTRY EDITED BY ALEXANDER FINDLAY, D.Sc. PER-ACIDS AND THEIR SALTS A COURSE OF PRACTICAL ORGANIC CHEMISTRY BY T. SLATER PRICE, D.Sc., PH.D., F.I.C., AND DOUGLAS F. TWISS, M.Sc., A.I.C., WITH 35 ILLUSTRATIONS. 8vo. 35. GD. LONGMANS, GREEN & CO., LONDON, NEW YORK, BOMBAY AND CALCUTTA. PER-ACIDS AND THEIR SALTS BY T. SLATER PRICE, D.Sc., PH.D., F.I.C. *< OF THE MUNICIPAL TECHNICAL SCHOOL, BIRMINGHAM LONGMANS, GREEN AND CO. 39 PATERNOSTER ROW, LONDON NEW YORK, BOMBAY AND CALCUTTA 1912 All rights reserved THE Author desires to express his thanks to Prof. Hugh Marshall, University College, Dundee, and to Prof. Melikoff, University of Odessa, who have kindly sent him reprints of their various papers, whereby the compilation of the literature references has been greatly facilitated. 257710 CONTENTS. CHAPTER PAGE I. INTRODUCTION .... i II. PERSULPHATES AND PERSELENATES ... 9 III. PERBORATES . . . 59 IV. PERCARBONATES 65 V. PERNITRIC ACID AND PERPHOSPHORIC ACID ... 74 VI. PERTITANATES, PERZIRCONATES AND PERSTANNATES . 79 VII. PERVANADATES, PERCOLUMBATES AND PERTANTALATES . 85 VIII. PERCHROMATES .92 IX. PERMOLYBDATES, PERTUNGSTATES AND PERURANATES . 101 LITERATURE REFERENCES . . . . .in INDEX 121 PER-ACIDS AND THEIR SALTS CHAPTER I. INTRODUCTION. IN his classical researches on the Periodic System of the Ele- ments, Mendeleeff has called attention to the fact that all the oxides of the type RO 2 cannot be considered as true peroxides. Some of them, as for example barium peroxide, BaO 2 , give hydrogen peroxide with acids, and may therefore be considered as derivatives of that compound, whilst others, such as manganese dioxide, MnO 2 , and lead dioxide, PbO 2 , do not so react with acids. To distinguish these two classes of oxides it is now customary to restrict the name peroxide to compounds of the first class, those of the second class being termed dioxides* * In various papers l Tanatar has expressed doubts as to whether there is really a constitutional difference between the " true " peroxides and the dioxides (" false " peroxides), he being of the opinion that the differences in behaviour of these two classes can be accounted for on thermochemical grounds. Amongst other experimental evidence brought forward, he states that ordinary nickel dioxide, prepared by the action of sodium hypobromite on nickelous hydroxide, yields hydrogen peroxide with acids, its presence being demonstrated by the oxidizing action of the solution on potassium iodide, and by the decolorization of potassium permanganate ; also, that hydrogen peroxide can be obtained from it in quantity by interaction with hydrocyanic acid in the presence of potassium cyanide. This evidence, if substantiated, would be of special importance, since Pellini and Meneghini 2 1 have prepared a greyish-green compound, NiO 2 ,xH 2 O, which gives all ^ the reactions of hydrogen peroxide and is quite different in its behaviour from ordinary nickel dioxide. 3 This would point to the existence of two structural isomerides, Ni whereas if Tanatar's observations were correct there could be no such difference in structure. Tubandt and Riedel 4 have recently shown, how- I k 2 : PE^- ACIDS AND THEIR SALTS Similarly, it is necessary to distinguish between the so-called per-acids. True per-acids may be defined as those which are either formed by the action of hydrogen peroxide on ordinary acids,* or else give rise to hydrogen peroxide on treatment with dilute sulphuric acid ; with concentrated sulphuric acid many of them evolve ozonized oxygen, thus behaving similarly to the metallic peroxides and to hydrogen peroxide itself. 5 In some cases the free acids are not known, but the same definition will apply to the formation or decomposition of their salts. There are other acids to which the prefix per has been applied, for example, perchloric, periodic, permanganic, and perruthenic acids. In the light of the above definition, however, these cannot be classed as true per-acids, since they are not formed by the action of hydrogen peroxide on the lower acids, nor is hydrogen peroxide produced by the action of dilute sulphuric acid on them or on their salts. In their case the use of the prefix per denotes only that they contain relatively more oxygen than chloric, iodic, manganic, and ruthenic acids respectively. Hydrogen peroxide is known to combine with many salts to form definite crystalline compounds. 6 In many cases it is tolerably certain that the hydrogen peroxide is present in a condition analogous to that of the water in salts containing water of crystallization ; such compounds are, for example, (NH 4 ) 2 S0 4 ,H 3 2 ; Na 2 SO 4 ,H 2 O 2 ,H 2 O.7 In other cases it is difficult to say whether the hydrogen peroxide is present as " hydrogen peroxide of crystallization," or whether it ever, that when ordinary nickel dioxide is treated with dilute sulphuric acid, no hydrogen peroxide can be detected in the solution by means of titanium dioxide, or chromic acid and ether. They ascribe the oxidizing action on potassium iodide to the formation, improbable as it may seem under the conditions of experiment, of either persulphuric acid or Caro's acid (in this connexion, see p. 45). The decolorization of the permanganate observed by Tanatar is only apparent, the colour of the nickel solution at first masking that of the permanganate. They also show that the formation of hydrogen peroxide in the experiments with nickel dioxide, hydrocyanic acid, and potassium cyanide, is due to the autoxidizing action of the complex nickel cyanogen compounds formed. The balance of evidence at present seems to be in favour of a real difference between peroxides and dioxides. * In the case of persulphuric acid, H 2 S a O 8 , chlorsulphonic acid has to be used, and not sulphuric acid. INTRODUCTION 3 has entered into combination with the acid radicle forming a per- acid. Examples of this class of compounds are those produced by the action of hydrogen peroxide on solutions of the alkali carbonates and bicarbonates, 8 and the difficulty of characterizing them is increased by their more or less complete hydrolysis in aqueous solution, hydrogen peroxide being one of the hydro lytic products. Other compounds, which are undoubtedly per-acids, undergo a similar hydrolysis, and this has caused considerable controversy as to whether such compounds are true per-acids, or so-called molecular compounds with hydrogen peroxide as one of the molecular constituents. This question will be discussed more fully when dealing with the percarbonates (q.v.) t in the case of which it has been found possible to obtain a criterion distin- guishing true percarbonates from carbonates containing hydrogen peroxide of crystallization. . The question still remains an open one in some cases, as, for example, with some of the compounds of boron. It is instructive to note that similar difficulties occur in the formulation ofonany of the compounds, other than per-acids, formed by hydns^en peroxide. Crystalline compounds of de- finite composition, tbxexample, Na 2 O 2 ,8H 2 O, Na 2 O 2 ,2H 2 O 2 ,4H 2 O, 2KOH,H 2 O 2 or K 2 O.2^LO, K 2 O 4 ,2H 2 O, etc., have been known for a long time, 9 but whether, for example in the potassium compound, the hydrogen peroxide is present as hydrogen per- oxide of crystallization (2KOH,H 2 O 2 ), or has given up its oxygen to the potassium, the water thereby formed remaining as water of crystallization (K 2 O 2 ,2H 2 O), remained undecided until Calvert 10 showed definitely by physico-chemical methods, that hydro- gen peroxide possessed acid properties in aqueous solution ; con- sequently the second method of formulation is probably the correct one.* A similar difficulty arises in connexion with compounds such as Na 2 O 2 ,2H 2 O 2 . If hydrogen peroxide is an acid, the salt in which one hydrogen has been replaced by sodium would be NaO 2 H, and Tafel n has isolated such a compound which he terms " sodyl hydroxide ".f If the compound Na 2 O 2 ,2H 2 O 2 * For a full discussion of the question Calvert's paper should be con- sulted. t Tafel inclines to the formulation of this compound as O:Na.OH, thus assuming tervalent sodium, for which there is very little evidence. 12 I * 4 PER- ACIDS AND THEIR SALTS were a derivative of this it could be formulated as 2NaO 2 H,H 2 O 2 ;* an alternative formula would be 2NaO 2 ,2H 2 O, in accordance with Calvert's results, although difficulties would then arise as to the valency of sodium or oxygen. These compounds are thus to be considered, not simply as molecular compounds, but as definite derivatives of hydrogen per- oxide, which are hydrolysed to a considerable extent in aqueous solution. When compared with the hydroxides of the metals, for example, ii /OH R.OH and R/ \OH they may be considered as metallic hydroperoxides, j- formed ac- cording to the types : ii /O.OH 15 RO.OH and R( \O.OH They should consequently form salt-like compounds with per-acids, and numerous such compounds have been prepared, for the most part by the Russian investigators, Melikoff and Pissarjewsky. As examples may be mentioned : (Na 2 O 2 ) 2 UO 4 ; Li 2 O 2 (UO 4 ) 2 ; (Na 2 O 2 )WO 4 ,H 2 O 2 ; (K 2 O 2 )MoO 4 ,H 2 O 2 , etc. As a general rule these compounds are not very stable, decomposing both as the dry substance, and on solution in water. Their formation shows that the per-acids from which they are derived are weak acids ; if it were not so, the hydroperoxides would be decomposed, instead of entering into combination. An examination of the elements which form per-acids shows that they belong to the third, fourth, fifth and sixth groups of the Periodic System. J In the following table the elements forming per-acids are printed in thick type. * This formula has been proved to be the correct one by the experiments of d'Ans and Friederich 13 on the interaction between anhydrous hydrogen peroxide and either sodium peroxide, sodium ethoxide or sodium. fThis nomenclature is due to Baeyer. 14 I It should be stated that Carrasco, 16 by the action of an ethereal solu- tion of hydrogen peroxide on zinc oxide, has prepared compounds having the formulae Zn 4 O 7 ,2H 2 O and Zn 3 O 5 ,H 2 O, which he claims to be perzincic acids. Until further evidence is forthcoming it is probably sufficient to consider these compounds as hydrated peroxides. 17 INTRODUCTION III IV V VI B c N Al< Si P s Sc Ti V Cr Ga Ge As Se Y Zr Cb Mo In Sn Sb T> La(?) Ce Yb Ta W Tl Pb Bi Th u Leaving out of consideration the elements in the short periods, that is, the elements in the first two rows, it will be noticed that the formation of per-acids is confined to groups IV, V and VI, * and more particularly, with the exception of tin and selenium, to the members of the even 18 series. The per-acids and per-acid salts of these elements are formed in all cases by the use of hydro- gen peroxide as one of the reacting substances, and their stability increases with increase in the atomic weight of the metal. This gradation in stability accounts for the ready decomposition of the perchromates, and explains why they have been isolated only in comparatively recent years, although their formation in solution was demonstrated as long ago as 1847, by Barreswil. 19 Also, it is in complete agreement with the rules concerning the position of the elements in the Periodic System, namely, that the elements of the even series are the more basic, and that with increasing atomic weight the basic properties of the oxygen compounds in- crease. For example, in the sulphur-group, peruranic acid is the most stable of the per-acids ; its salts with the metallic hydroper- oxides are prepared at the ordinary temperature and can be pre- served in an atmosphere free from carbon dioxide for a considerable length of time without undergoing decomposition. The analogous salts formed by permolybdic acid are obtained only at tempera- tures of- 10 to- 13, and decompose very rapidly at ordinary temperatures, with loss of active oxygen. The compounds formed by pertungstic acid are somewhat more stable than the permolyb- dates. 20 Of the elements of the short periods, carbon and sulphur are * The existence of per-acids of lanthanum is doubtful. The hydrated per- oxide apparently does not possess acid properties. 173 6 PZR-ACIDS AND THEIR SALTS distinguished from the remainder in that their per-acid salts can be readily obtained from the ordinary salts by electrolytic methods, whereas with the other elements the use of hydrogen peroxide is generally necessary. In formulating the constitution of the various per-acids as derivatives of hydrogen peroxide, the constitution of this com- pound will be taken as HO. OH. The reasons for this cannot be discussed here, the literature on the subject being very ex- tensive. 21 The constitution of the per-acids formed by the elements in the short periods will be considered when dealing with the re- spective acids, but reference will now be made to the methods used for expressing the formulae, which are often very com- plicated, of other acids and their salts. Pissarjewsky 22 has pointed out that the per-acids may be considered as salts derived from hydrogen peroxide, acting as an acid. Tungstic acid, WO 2 (OH) 2 , for example, may be assumed to be amphoteric in character ; * as a base it could then react with hydrogen peroxide according to the equations : WO 2 (OH) 2 + H(O 2 H) = WO 2 (O 2 H)OH + H 2 O ; W0 2 (OH) 2 + 2H(0 2 H) = W0 2 (0 2 H) 2 + 2 H 2 O. Theoretically it is possible^ for other per-acids to exist, which would be derived from the hypothetical orthotungstic acid W(OH) 6 . Assuming that this can act as a base as well as an acid, the highest per-acid would be W(O 2 H) 6 ; the acid derived from one molecule of the ortho-acid and four molecules of hydrogen peroxide would be W(O 2 H) 4 (OH) 2 , or WO(O 2 H) 4 . The salt K 2 O 4 ,WO 4 ,H 2 O, would be a salt of this acid, namely, WO(O 2 K) 2 (O 2 H) 2 ; similarly (Na 2 O 2 ) 2 UO 4 ,8H 2 O, is a salt of UO(0 2 H) 3 .OH, namely, UO(O 2 Na) 3 .ONa,8H 2 O. Pissarjewsky states expressly that these formulae must not be taken asyrepresenting constitutional formulae ; they have the same significance as, for example, the formula A1K(SO 4 ) 2 for alum, which simply denotes that the hydrogen of the sulphuric acid is partly replaced by potassium and partly by aluminium, without denoting anything else as to the constitution.f * Compare uranic acid, UO 2 (OH) 2 , which forms the uranates, and also the uranyl compounds, thus acting both as an acid and a base. + Pissarjewsky also considers the various ways in which these acids and salts may dissociate. For details, the original paper should be consulted. INTRODUCTION 7 For representing the constitution of these compounds, use is made of other considerations. MelikofT and Pissarjewsky 23 have found that when the salts of a per-acid with a metallic hydroperoxide are digested at the ordinary temperature with freshly precipitated aluminium hydroxide,* part of the active oxygen goes into solution as hydrogen peroxide, and part re- mains with the aluminium hydroxide. For example, with the salt (NagOg^UO^SHgO, in which the ratio of uranium to active oxygen is 1:3, two-thirds of the active (peroxidic) oxygen is found as hydrogen peroxide in the filtrate from the aluminium hydroxide, and one-third remains in the precipitate as peruranic acid, as determined by titration with sulphuric acid and potassium permanganate. Peruranic acid is UO 4 ,2H 2 O, or H 4 UO 6 , and since the ratio of uranium to active oxygen is I : I, its constitu- tion may be either HO.CK X) \U:0 or (HO) 4 =U< | . (H0) 3 ^ \0 The sodium salt above mentioned may be considered as formed from this acid by interaction with an alkaline solution of sodium hydroperoxide, and would be either f o, Na If the former formula is correct, there appears to be no reason why all the active oxygen should not be found as hydrogen peroxide in the filtrate from the aluminium hydroxide, whereas the latter formula will explain the actual experimental results. The formula for peruranic acid is therefore written as Further support for the above is found in the fact that carbon dioxide, which has no action on uranic acid, converts insoluble peruranates, for example, (BaO 2 ) 2 UO 4 ,9H 2 O, into metallic hydro- gen carbonates, hydrogen peroxide, and free peruranic acid. As will be seen in the sequel, the constitution of the per-acids of the other heavy metals may be expressed similarly. J * This substance was chosen because it is neither a strong base nor a strong acid, and cannot be further oxidized. t Compare, however, pp. 108-109. + As far as I am aware, Melikoff and Pissarjewsky have not stated expli- citly their reasons for adopting such formulae, but the arguments given above follow naturally from their experimental results. 8 PER-AC1DS AND THEIR SALTS The constitutions 'of the various per-acids and their salts, with the possible exception of those of chromium, can all be formulated on the assumption that the element forming the per-acid possesses the maximum valency towards oxygen corresponding with the particular group of the Periodic System in which it occurs. For example, the highest positive or negative valency which titanium can assume, according to its position in the Periodic System, is four, and it is therefore improbable that titanium is sexavalent in pertitanic acid and its derivatives, as has been assumed in some cases. 24 Again, the composition of the sodium peruranate mentioned above may be expressed either as (Na 2 O 2 ) 2 UO 4 , 8H 2 O, or as 2Na 2 O,UO 6 ,8H 2 O ; 25 the second formula would, however, assign to uranium a valency higher than six, and is therefore improbable. The evidence which is at present available seems to indicate that chromium is septavalent in the perchromates, 26 and there- fore occupies an exceptional position. The matter will be more fully discussed when the perchromates are dealt with. In the following chapters the per-acids formed by the elements of the two short periods will be dealt with first, in the order : sulphur, boron, carbon, nitrogen, phosphorus. This order is taken because of the importance of the persulphates, which were the first per-acid salts to be studied exhaustively. The re- mainder of the per-acids will then be dealt with in accordance with the periodic classification. The organic per-acids 27 will not be considered. CHAPTER II. THE PERSULPHURIC ACIDS AND PERSULPHATES. PERSELENATES. The Persulphuric Acids and Persulphates. Two per-acids of sulphur are known to exist, namely, per- monosulphuric acid (or Caro's acid), H 2 SO 5 , and perdisulphuric acid (commonly called persulphuric acid), H 2 S 2 O 8 . In the early days of their investigation no definite distinction was drawn between them, indeed it was not known that two different acids existed, with the result that the literature on the subject is very confusing. In dealing with the historical aspect of the question, it will make matters clearer if the relations which are now known to exist between these two acids are first indicated. When an aqueous solution of fairly concentrated sulphuric acid is electrolysed between platinum electrodes, using a high current density at the anode, perdisulphuric acid is first formed by the discharge of the HSO^-ions, in accordance with the equation : 2 HSO' 4 + 2 = H 2 S 2 8 . In the presence of sulphuric acid, however, this acid is not very stable, and changes more or less quickly, depending on the concentration of the sulphuric acid, into permonosulphuric acid, as shown by the equation : H 2 S 2 8 + H 2 = H 2 S0 5 + H 2 S0 4 . The permonosulphuric acid can then undergo further decom- position with the production of hydrogen peroxide : H a SO 5 + H a O = H 2 SO 4 + H 2 O 2 . The two acids are readily distinguished from hydrogen per- oxide by the fact that they do not decolorize permanganate, or give rise to the characteristic colorations produced by hydrogen peroxide with titanium salts, or with chromic acid ; but it is more difficult to differentiate them from each other in aqueous solution (compare the tests given on pp. 40, 46, 54-57). 9 10 PER- ACIDS AND THEIR SALTS As will be understood from the equations just given, hydrogen peroxide is often co-existent in solution with these acids, and this has caused difficulties in the quantitative experiments carried out, since, although the persulphuric acids do not react with permanganate, and only very slowly with hydrogen peroxide, an induced action is brought about between them and hydrogen peroxide when titrated with permanganate, whereby the estima- tion of hydrogen peroxide by means of permanganate is made untrustworthy unless special precautions are taken. In what follows, the customary terms "persulphuric acid" and "persulphate" will be used instead of perdisulphuric acid and perdisulphate, it being understood that they apply only to the acid H 2 S 2 O 8 and its salts. In 1832 Faraday observed that in the electrolysis of concen- trated sulphuric acid the oxygen evolved is considerably less than one-half of the hydrogen simultaneously produced. This observa- tion was confirmed by Meidinger, 1 who found further that, after the electrolysis, the electrolyte possessed the property of oxidizing hydriodic acid. He attributed this to the formation of hydrogen peroxide at the anode, but Brodie 2 showed that this substance could not be present, since the anolyte did not give the ordinary reactions for hydrogen peroxide with chromic acid or potassium permanganate. Indigo was, however, bleached by the warm solution, and ferrocyanide was oxidized, and these properties being those of a peroxide of an acid, Brodie suggested that a peroxide of sulphuric acid, to which he ascribed the formula H 2 SO 5 , was formed.* Berthelot, 3 as also McLeod, 4 confirmed Brodie's observation that hydrogen peroxide was not produced, and brought forward evidence that the oxidizing power of the solution was due to some oxygenized compound of sulph- uric acid, which he denoted as persulphuric acid, in analogy to permanganic acid. By submitting a dry mixture of equal volumes of oxygen and sulphur dioxide to the action of a silent electrical discharge he found that a viscid liquid was produced, which gave crystalline needles or lamellae in the neighbourhood of o. It had the composition S 2 O 7 and dissolved readily in water combining with it with a hissing noise and the evolution of much heat with the formation of a solution which liberated iodine from potassium * Brodie's paper is concerned chiefly with organic peroxides, and the formula HoSO 5 was suggested from analogy to these compounds. THE PERSULPHURIC ACIDS AND PERSULPtJATES If iodide, and oxidized sulphurous acid, ferrous sulphate, etc., but had no action on permanganate or chromic acid ; that is, it pos- sessed the properties of the solution obtained by the electrolysis of sulphuric acid, some of which he prepared for the purpose of comparison. By cautiously neutralizing with barium hydroxide and filtering, a solution was obtained which deposited barium sulphate on boiling, whilst free sulphuric acid, equivalent in quantity to the barium sulphate, remained dissolved. This seemed to prove that the substance in solution was really an acid, cor- responding with the anhydride, and Berthelot assigned to it the formula HSO 4 (S 2 O 7 + H 2 O = 2HSO 4 ), from analogy to per- chloric and permanganic acids, although it was found impossible to obtain the salts in the solid state. Mendeleeff, 5 however, considered that the new compounds were peroxides, correspond- ing with barium peroxide, hydrogen peroxide, etc., which would not form salts. Berthelot found also that a solution with similar properties could be obtained by the action of concentrated sulphuric acid on solutions of hydrogen peroxide. Further experiments on the electrolysis of sulphuric acid of varying concentrations at low temperatures, using a divided cell with a porous partition and platinum electrodes, led him to the conclusion that at a certain stage of concentration the reaction altered, and instead of pure persulphuric acid being formed, combination of that substance with hydrogen peroxide took place, tending to the formation of a com- pound of the composition S 2 O 7 ,2H 2 O 2 .* The solution of this substance gave an immediate liberation of iodine from potassium iodide (supposed to be caused by the S 2 O 7 ), followed by a further slow liberation of iodine (by the hydrogen peroxide), whereas the solution supposed to contain the acid HSO 4 liberated all the iodine immediately from potassium iodide. Berthelot' s interpretation of his results is now known to be in- correct ; his persulphuric acid had the properties of permonosul- phuric acid, H 2 SO 5 , whilst persulphuric acid, H 2 S 2 O 8 , corresponds * Richarz 6 also investigated the formation of hydrogen peroxide at the anode during the electrolysis of sulphuric acid, but did not agree with Berthelot that the ratio of hydrogen peroxide to persulphuric acid tends to become equal to 2. His observations led him to the conclusion, which cannot now be accepted, that hydrogen peroxide is formed by the oxidation of water by persulphuric acid. 12 PER- ACIDS AND THEIR SALTS with his compound containing combined hydrogen peroxide.* Even when Marshall 7 isolated the persulphates in 1891 Berthelot considered that they were salts of his persulphuric acid, not taking any notice of the fact that Marshall's salts liberated iodine only slowly from potassium iodide, whereas his acid did so instantaneously. Already in 1889, Traube 8 had analysed the solution of persulphuric acid prepared by electrolysis, after first precipitating the sulphuric acid with barium hydrogen phosphate, and found therein a different ratio between sulphuric acid and active oxygen from that assumed by Berthelot. Berthelot's formula, HSO 4 , demands the ratio H 2 SO 4 : active oxygen = 2 : i , whereas Traube found the ratio i : I, and hence he deduced the formula SO 4 . He considered the substance to be a peroxide and called it sulphuryl holoxide.f We now know that it contains a molecule of water, and is H 2 SO 5 . The confusion became worse confounded when Traube repeated his experiments in 1893 8 and found the ratio 2 : i, that is, the same as that assumed by Berthelot. Both Berthelot and Traube were right. In the freshly electrolysed 40 per cent sulphuric acid used by Traube, persul- phuric acid, H 2 S 2 O 8 , is present, whilst on keeping for some time, permonosulphuric acid, H 2 SO 5 , is formed, with the result that the above-mentioned ratio alters. Traube, in repeating his ex- periments, used the freshly electrolysed acid, whereas in the first set of experiments, the acid had been kept for some time. The isolation of the persulphates by Marshall and the proof (see p. 41) that they were derived from the acid H 2 S 2 O 8 , together with the fact that Traube could not confirm his earlier experi- ments, seemed to lead to the acceptance of the existence of one persulphuric acid only, although Berthelot's experi- ments had pointed to the existence of two such, and there were numerous discrepancies between the reactions given by the persulphates and the solutions obtained by the electrolysis of * In these experiments persulphuric acid would be first formed, and then hydrolysed more or less quickly, giving first permonosulphuric acid and then hydrogen peroxide, according to the conditions of concentration, temperature, etc. t Traube 9 proposed the name holoxide for all derivatives of hydrogen peroxide, since they were considered to be built up from an undivided molecule of oxygen. THE PERSULPHURIC ACIDS AND PERSULPHATES 13 sulphuric acid. In 1898, however, Caro 10 discovered that a solution prepared by the action of concentrated sulphuric acid on ammonium persulphate and subsequent neutralization with ammonium carbonate, had a different action on aniline from a solution of ammonium persulphate alone, thus pointing to the existence of a persulphuric acid other than the acid H 2 S 2 O 8 . This reopened the whole question and led to numerous investiga- tions, with the result that the existence of the two acids, H 2 SO 5 and H 2 S 2 O 8 , has been definitely proved. The further treatment of the subject will be best carried out by dealing separately with the respective acids, describing first ordinary persulphuric acid and its salts. Ordinary Persulphuric Acid (Perdisulphuric Acid) and its Salts. In 1891, whilst engaged in an investigation on the electrolytic oxidation of cobalt salts, Marshall incidentally obtained a small quantity of a colourless potassium salt, which analysis proved to have the composition represented by the empirical formula KSO 4 . The aqueous solution gave only a very faint precipitate with barium chloride, liberated iodine from potassium iodide, and showed other oxidizing actions which were similar to those of the persulphuric acid solutions obtained by Berthelot. Successful attempts were made to prepare further quantities of the substance by electrolysing solutions of potassium hydrogen sulphate in a divided cell. 7 A platinum dish of 200 c.c. capacity contained the anolyte, which consisted of a saturated solution of potassium sulphate in dilute sul-' phuric acid ; the dish formed the anode and was cooled externally by a stream of water. Dipping into the anolyte was a porous pot containing a platinum wire cathode, and dilute sulphuric acid as the catholyte. A current of about 3 amperes was used, and solid potassium persulphate began to separate after a day or two, depositing thereafter fairly rapidly. When a considerable amount had collected, it was drained off and dried on a porous plate, and the crude salt thus obtained was recrystallized and purified by dissolving it in boiling water, filtering rapidly and cooling the filtrate. The solid ammonium salt was prepared similarly, re- placing potassium sulphate by ammonium sulphate. Some time later n Marshall, wishing to prepare larger quan- tities of the ammonium salt, substituted the small platinum 14 PER- ACIDS AND THEIR SALTS basin by a larger one and used a higher current; no persul- phate separated, however, even after some days. Berthelot, in his earlier experiments on persulphuric acid, had found it pre- ferable to use a small platinum wire anode in place of a platinum dish, his idea being that the possible decomposition of persulphuric acid in contact with platinum would thereby be reduced to a minimum. Consequently, on repeating Marshall's experiments on the preparation of potassium and ammonium persulphates, he placed the sulphate solution in the porous pot and used a platinum wire anode; the yields of the salts were thus considerably in- creased. By adopting a similar arrangement, the anode con- sisting, however, of a thin platinum tube through which cold water flowed, Marshall readily obtained ammonium persulphate in quantity. The impossibility of preparing persulphates when a very large anode, which necessitates a low current density, is used, indicates that persulphuric acid is not formed simply by the discharge of the HSO' 4 -ions conveying the current, these, and not SO" 4 -ions, forming the majority of the anions in solutions of sulphuric acid or of acid sulphate of the concentration employed. The formula HSO 4 for the acid would therefore seem to be doubtful. If, however, the true molecular formula is H 2 S 2 O 8 , persulphuric acid would be formed by two discharged HSO' 4 -ions uniting together, and there would be greater opportunity for such union the closer the ions were packed together at the moment of their discharge, that is, the higher the current density. 11 The evidence (see pp. 41-44) is now conclusive that the molecular formula is H 2 S 2 O 8 , and assuming that the mechanism of its formation is as given, it is evident that to obtain good yields of the acid we should have : (l) an anode solution containing the largest possible proportion of HSO' 4 -ions, (2) high current density, and (3) low temperature. The third condition is necessary because persulphuric acid readily decomposes when warmed. The same conditions will hold in the preparation of the salts, since it is probable that persulphuric acid is the substance first formed in the electrolysis of the acid sulphates, the strong solutions of which all contain HSO' 4 -ions; the potassium or other salt would then result by double decom- position with the' potassium or other sulphate present, and crystallize out when the solution becomes saturated. 11 THE PERSULPHURIC ACIDS AND PERSULPHATES 15 Persulphuric Acid. (a) Influence of Concentration of the Sulphuric A rid. Berthelot 3 first investigated the influence of concentration of the sulphuric acid on the yield of per-acid, and an extended series of observa- tions was made by Richarz, 6 who showed that the yield was a maximum with 40 per cent sulphuric acid ; also, that a high current density at the anode was necessary. A very complete investigation was then carried out by Elbs and Schonherr, 12 who, at the time, were not aware of the extent of Richarz' s ob- servations. The apparatus employed consisted of a divided cell formed by a porous pot standing in a beaker. The cathode was a cylinder of lead surrounding the pot, the anode consisting of platinum wire or foil, dipping into the pot. The temperature was maintained at 5-10, and a current density of 100 amperes per sq. dcm. at the anode was used. From time to time 5-10 c.c. of the anolyte were pipetted out and run into 200-300 c.c. of cold water, excess of ferrous sulphate solution added, and the excess titrated with permanganate in order to determine the amount of persulphuric acid and hydrogen peroxide present (for methods of estimating persulphates, see p. 40) ; in another portion of the anolyte the amount of hydrogen peroxide was determined by direct titration with permanganate.* The electrolysis of very dilute sulphuric acid (density and not to hydrion. In what way the addition of hydrofluoric acid raises the anodic potential has not hitherto been explained, but the increased yield of persulphate is readily accounted for. In the acid solutions used, permonosulphuric acid is formed by the hydrolysis of the persulphuric acid and acts as a depolarizer (cf. p. 17); the higher the anode potential the more readily will the permonosulphuric acid be destroyed, according to the reaction already given, namely, H 2 SO 5 + 2OH' + 2 = H 2 SO 4 + O 2 + H 2 O, and therefore the greater the yield of persulphate. The above results, and also those obtained by the addition of hydrochloric acid, sulphur dioxide, etc., in the electrolysis of sul- phuric acid (cf. p. 1 8) are discussed by Muller and Schellhaas 13 in connexion with the technical manufacture of persulphates. The addition of fluoridion is of use only on a small scale ; on pro- longed electrolysis the quantity of permonosulphuric acid must keep increasing, until there comes a point where just as much persulphuric acid is formed at the anode as is changed into permonosulphuric acid, that is, the current yield of persulphate becomes zero. Destruction of the permonosulphuric acid with hydrochloric acid, sulphur dioxide, etc., can only be of use in certain cases, as may be seen from the following considerations : Supposing the S 2 O" 8 -ions formed remain in solution, that is, that solid persulphate is not precipitated, there will come a time when the persulphuric acid changes into permonosulphuric acid as fast as it is formed, and destruction of the latter will not increase the current yield of the former. The reactions are otherwise when the persulphate separates from solution before the concentration of the S 2 O" 8 -ions has been reached at which the velocity of change to permonosulphuric acid is equal to the velocity of formation of persulphuric acid. The velocity of transformation into the permono-acid then attains a maximum value which is less than that of the formation of persulphuric acid, and the difference between these two velocities is all the greater the more sparingly soluble the persulphate is. In this case the permonosulphuric 28 PER-ACIDS AND THEIR SALTS acid can be destroyed with advantage by the addition of sulphur dioxide. The velocity of change of the one acid into the other increases with the hydrion concentration (cf. p. 21), and in slightly acid solutions of ammonium persulphate the velocity is so small that even in saturated solutions it does not reach the limiting concentration where the velocity of change is equal to the velocity of formation ; consequently ammonium persulphate is readily prepared from these weakly acid solutions. For the formation of sodium and potassium persulphates it is neces- sary to have strongly acid solutions in order to obtain the necessary concentration of the HSO' 4 -ions. Potassium persulphate is so slightly soluble, however, that the conditions for its formation are favourable. They are also favourable in the case of sodium persulphate, because it so happens that its solubility is considerably diminished in strongly acid solutions of sodium sulphate. The solubility is still great enough, how- ever, to lead to the formation of considerable quantities of permonosul- phuric acid, and in order to maintain a good current efficiency this substance must be removed by the addition of sulphur dioxide. Thus, in one experiment the electrolysis was continued for ten hours, sulphur dioxide in excess being continuously passed through the electrolyte, which consisted of 36*2 per cent sulphuric acid saturated with sodium sulphate ; the current density was 50 amperes per sq. dcm., and the temperature was 15. The current yield was 77-2 percent (solid + per- sulphate in solution), the amount of solid being 236 grams, containing 80 *4 per cent Na 2 S 2 O 8 . With such concentrated solutions, however, it is difficult to filter the paste of crystals obtained. The results already given form the basis of a series of patents for the manufacture of sodium, potassium, and ammonium per- sulphates. 32 According to other patents 33 similar beneficial results an 84 per cent yield are obtained in the manufacture of sodium persulphate by the addition of simple or complex cyan- ides, for example K 4 FeC 6 N 6 , in small quantities to the electro- lyte. 34 Alkali cyanates, thiocyanates, and cyanamides produce a similar effect. The addition of small quantities of potassium salts causes the sodium persulphate to separate in a granular form, and not as a paste which is purified only with difficulty.* To what extent the above processes are used commercially it is difficult to say, since according to a patent of the " Konsort. % * On account of the great solubility of sodium persulphate and the ready decomposition of its solution on evaporation, its manufacture is attended with difficulties. A process patented by Lowenhen; 35 for their preparation has now lapsed. THE PERSULPHURIC ACIDS AND PERSULPHATES 29 elektrochem. Ind. " for the manufacture of ammonium persulphate, very satisfactory results are obtained in an undivided cell, with continuous working and without extraneous additions, if the electrolyte is kept acid and a high current density at the cathode is employed. Thus, in a saturated solution of ammonium sul- phate, with a current density of 10 amperes per sq. dcm., the current yield is 25 per cent, with 50 amperes it is 50 per cent, and with 300 amperes it becomes 70 per cent.* The best method, then, for the manufacture of potassium persulphate would prob- ably be by double decomposition of the ammonium salt with potassium carbonate. Sodium persulphate could, if necessary, be prepared similarly. 37 Other Persulphates. Rubidium per sulphate , Rb 2 S 2 O 8 , and cczsium persulphate Cs^Og have both been prepared by the electrolysis, in a divided cell, of sulphuric acid solutions of rubidium and caesium sulphates re- spectively. 38 Marshall 39 has also obtained them from am- monium persulphate by double decomposition. It is noteworthy that the crystals are not isomorphous with those of the potassium salt (triclinic), but with those of the ammonium salt (monoclinic). Mixtures of the potassium salt with the others have, however, been obtained in well-formed monoclinic crystals, notwithstand- ing a great preponderance of the potassium salt. The pure thallium salt could not be prepared by electrolysis, 38 but Marshall 39 obtained mixed crystals of thallous persulphate with ammonium persulphate, which were isomorphous with the monoclinic crystals mentioned above. Otin 40 has obtained an impure lithium salt by neutralization of a solution of per- sulphuric acid with lithium carbonate, and evaporation of the resulting solution to dryness in a vacuum. To prepare barium persulphate^ BaS 2 O 8 , 4H 2 O, a saturated aqueous solution of ammonium persulphate is well rubbed up in a mortar with an excess of crystallized barium hydroxide, and then treated with a current of air to drive off the major portion of the ammonia, the remainder of which is removed by placing the liquid in a vacuum over sulphuric acid. Carbon dioxide is passed through the solution so * Miiller and Friedberger found that a high current density at the cathode reduced the persulphate formed ; according to the above patent this was probably because their solution became alkaline during the electrolysis. 30 PER- ACIDS AND THEIR SALTS obtained to remove the excess of barium hydroxide, any barium bicar- bonate formed being subsequently decomposed by placing the solution in a vacuum again for a short time. After collecting the barium carbonate, the filtrate is evaporated in a vacuum, the sulphuric acid formed by decomposition of the solution being neutralized from time to time with barium hydroxide. As soon as crystallization commences, the evapora- tion is discontinued and the crystals redissolved by the addition of the minimum quantity of water ; the solution is filtered and then cooled with ice to bring about the separation of crystals. 7 The crystals so ob- tained are purer than those formed by the gradual evaporation of the solution * ; they form beautiful, small prisms and are apparently mono- clinic. In the course of a few days they become milky from separation of barium sulphate,- finally crumbling to pieces, forming a moist, powdery mass. To preserve the crystals as long as possible it is best to keep them in a moist atmosphere. At o, 100 parts of water dissolve 52-5 parts of the crystallized salt. Pure solutions gradually decompose ; if a fairly dilute solution is heated it generally remains clear till near the boiling point, when sulphate rapidly precipitates. It requires rather prolonged boiling, however, to destroy the last traces of persulphuric acid. When the finely powdered crystals are digested for several days with succesive portions of absolute alcohol, the latter then removed with anhydrous ether, and the product dried in a current of dry air, the com- pound BaS 2 O 8 , H 2 O, is obtained : it cannot be further dehydrated without suffering decomposition. Other simple f persulphates have not been isolated in a pure condition, although impure lead and zinc salts have been obtained by Marshall. 7 * Large, transparent, interlocking prisms, generally deeply striated. ^'tjorgensen 41 has obtained hexamminecobaltisulphatepersulphate, [Co(NH 3 ) 6 ] 2 (SO 4 ) 2 (S a O 8 ) in the form of yellow, rhombic plates, by the inter- action, under pressure, of an ammoniacal solution of cobalt sulphate and ammonium persulphate. The following complex compounds have also been described by Barbieri and Calzolari. 42 They were prepared by interaction of the metallic sulphate, ammonium persulphate, and the base, in cold concentrated solution. Ammonia compounds: ZnS 2 O 8 ,4NH 3 ; CdS 2 O 8 ,6NH 3 ; NiS 2 O 8 ,6NH 3 . Pyridine compounds : RS-A^CsHsN, where R = Zn, Cd, Ni, or Cu. Hexamethylenetetramine compounds : RS 2 O 8 ,2C 6 Hi2N 4 ,8H 2 O, where Mg, Mn, Co, or Ni. THE PERSULPHURIC ACIDS AND PERSULPHATES 31 Properties of the Persulphates. Dry potassium persulphate can be kept indefinitely at the ordinary temperature without undergoing decomposition ; the ammonium salt is not obtained pure so readily as the potassium salt, but Marshall records that a dry specimen which he had pre- served for five and a half years still contained 97 per cent of the persulphate. The data with respect to the other salts of the alkali metals are very scanty, but in all probability the pure salts may be considered to be stable when preserved in the dry con- dition. Salts which contain water of crystallization, as, for example, the barium salt, and an impure lead salt mentioned by Marshall, are less stable, the barium salt decomposing in the course of a few days. All the salts are soluble in water, the solubility increasing with the atomic weight of the metal in the series, K, Rb, Cs and Tl. The lithium and sodium salts are, however, very soluble. The solubility of potassium persulphate is less in solutions of potassium salts than in pure water, and greater in solutions of sodium salts, in the latter case being a function of their concen- tration in sodium. 43 Evidently the increased solubility is due to the transformation of the potassium into the sodium salt Solutions of persulphates gradually undergo decomposition in accordance with the equation : KgSaOs + H 2 = 2 KHS0 4 + O a . The reaction is unimolecular and is much accelerated by the addition of acids. 44 ' 46 At 30. there is no decomposition, even after two days ; the rate at which the reaction proceeds at higher temperatures may be seen from the following velocity constants. 45 A represents the strengths of the solution in gram- molecules per litre, and the time is reckoned in minutes : Temp. A. j * for N a2 S 2 8 . Salt. A. k (at 80). -0 2k o-i25V 0-0016 NaA0 8 0-226 0-00541 NaA0 8 0-125 0-00577 80 0-126 0-0055 Na 2 S 2 8 0-127 0-00533 90 0-130 0-0161 K a S 2 8 0-108 0-00541 1 (NH 4 ) 2 S 2 8 0-229 0-0061 The unimolecular formula does not hold strictly for the ammonium salt, since side reactions (oxidation of the ammonia, 366 P- 35) ta ke place. 32 PER- ACIDS AND THEIR SALTS The reaction itself produces acid sulphate, and is, as a matter of fact, followed throughout its course by measuring the increasing acidity ; it is therefore strange that it is not accelerated autocatalytically when the pure salts are used, since the addition of acid has an accelerating effect. Green and Masson 45 account for this by assuming that the acid sulphate produced ionizes only into the metallic ion and HSO' 4 -ion, producing practically no hydrion, the reaction then being formulated by the equation : S 2 O" 8 + H 2 O = 2HSO' 4 + -JO 2 . Evidence in favour of this assumption was obtained by experiments with solutions of persul- phuric acid and of the barium salt. In the former case the reaction, which is represented by the equation : H 2 S 2 O 8 + H 2 O = 2H 2 SO 4 + -JO 2 , is only apparently unimolecular, since its velocity coefficient varies with the concentration. At the dilutions employed it is assumed that the persulphuric acid is practically completely dissociated into 2H and S 2 O" 8 , whilst the sulphuric acid gives H and HSO' 4 only. Thus, although the acidity doubles during the reaction, the concentration of hydrion remains constant. The velocity coefficient will then be the sum of a true unimolecular constant (k) depending on the decomposition of the S 2 O" 8 -i ns > an d a quantity (^jA) which depends on the concentra- tion, and expresses the catalytic effect of the hydrion. The velocity equation would then be given by dx\dt = (k + X A)(A - x)* which is the same as the unimolecular equation dx\dt K(A - x), where K = k + /fcjA. By comparison of experiments with different values of A, k and k^ can be evaluated and thus a value of K calculated and com- pared with that obtained by direct experiment. The results obtained by Green and Masson give a very good agreement between these two sets of values. The addition of alkalis does not alter the unimolecular character of the decomposition of the sodium or potassium salt, and hardly changes the value of the constant. The most striking chemical property of the persulphates is their oxidizing action. For example, the halogens are liberated * Green and Masson deduce this equation in a slightly different manner. Although it gives a quantitative explanation of the results obtained, the validity of the assumption that sulphuric acid gives only H-and HSO' 4 -ions at the dilu- tions employed is somewhat doubtful, as also of the assumption that the HSO' 4 -ion of the acid sulphate formed from salts of persulphuric acid during the reaction, provides practically no hydrion. Some support for their assump- tions was obtained by experiments in which known quantities of sulphuric acid, sodium hydrogen sulphate, or sodium sulphate, were added beforehand to solutions of persulphuric acid or sodium persulphate, but even here quite anomalous results were obtained when sodium sulphate was added to a solution of sodium persulphate. For details, the original paper should be consulted. THE PERSULPHURIC ACIDS AND PERSULPHATES 33 from chlorides, bromides and iodides ; thus with potassium iodide the reaction is given by the equation : K 2 S 2 8 + 2KI = 2K 2 S0 4 + I 2 . This reaction proceeds slowly at the ordinary temperature, more quickly on warming, and has been investigated by Price. 46 Ac- cording to the chemical equation it should be one of the third order, but investigation by the ordinary methods of chemical dynamics showed it to be of the second order. This would agree with the equation KSO 4 + KI = K 2 SO 4 + I, but all other evi- dence is in favour of the formula K 2 S 2 O 8 for potassium persul- phate (compare pp. 41-44). The usual explanation given in such cases is that the reaction takes place in stages, one of which governs the order of the reaction, the others proceeding very quickly in comparison. In this case it is difficult to formulate the probable stages. It may be that the potassium salt of permonosulphuric acid is first formed according to the equation : K 2 S 2 O 8 + H 2 O = KHSO 5 + KHSO 4 , and that this is a slow reaction, the iodine then being immediately liberated from the potassium iodide by the potassium permonosulphate (KHSO 5 ). In solutions of the strength used (N/3o N/3oo) there is no evidence, however, that potassium permonosulphate is thus formed.* Nor, on account of the dilution, is it probable that the anomaly is due to the oxidation, by the persulphate, of the liberated iodine to iodic acid, in accordance with the equation : I 2 + 5K 2 S 2 O 8 + 6H 2 O = 2HIO 3 + ioKHSO 4 . 48 The reaction is very sensitive to the catalytic influence of iron and copper salts, 46 a marked acceleration being produced even in the presence of M/64OOO solutions of either ferrous sulphate or copper sulphate ; iron salts are the more active. In each case the acceleration is directly proportional to the concentration of the catalyst, and in the case of iron salts, independent of whe_ther^ it is added as the ferrous or ferric salt /One very notable pheno- menon is that when the catalyst is a mixture of ferrous and copper sulphates, the acceleration produced is more than twice as great ^ _ s+ as would be the case if the effect were simply additive; the acceleration produced by a M/2 560000 solution of ferrous sulphate * Merk 47 uses the same explanation to account for the action of potas- sium persulphate on a solution containing potassium iodide and bromide, but gives no evidence in support of the intermediate formation of potassium permonosulphate. 3 34 PER-ACIDS AND THEIR SALTS or copper sulphate in the presence of M/32OOO copper sulphate or ferrous sulphate can readily be detected. Federlin 49 has made use of this reaction in connexion with the question of " induced reactions " (cf. Mellor's " Statics and Dynamics," p. 333), and with the explanation of catalysis by means of the formation of intermediate products. This explan- ation, as has been emphasized by Ostwald, is only valid when it can be shown that the velocities of the intermediate reactions are such that they account for the velocity of the main reaction. Phosphorous acid is oxidized by potassium persulphate with an immeasurably slow velocity, but with a measurable velocity by iodine, whilst iodine is liberated from potassium iodide at a rate which is readily measured. Federlin was able to show that the velocity of oxidation of phosphorous acid by persulphate in the presence of potassium iodide was equal to that calculated from the velocities of the individual reactions. This was not true in the presence of copper and iron salts, which act as catalysts only on the reaction between persulphate and iodide. Ferrous salts are oxidized very rapidly to the ferric condition ; the reaction is best .carried out in acid solution, and is made use of in the estimation of persulphates (see p. 40). In neutral solutions of potassium persulphate and ferrous sulphate a pre- cipitate of a" basic salt is deposited. 50 Persulphates cause more or less complete precipitation of dioxides when added to solutions of certain metallic salts ; for example, of manganese, silver and cobalt. In other cases, as those of lead and nickel, precipitates of the dioxides are obtained only after the addition of alkali. In the case of manganese the oxidation can be carried still further by digesting the mixture on the water bath for some time, when a pink solution of perman- ganic acid or of alkaline permanganate is slowly formed. Simi- larly, chromic salts can be converted into chromates by means of potassium persulphate, either in acid or alkaline solution 7> u> 51 . The reactions of persulphates in the presence of silver salts are of great interest and importance. Marshall showed that solutions of potassium persulphate and silver nitrate give a black precipitate which was supposed to be silver peroxide, Ag 2 O 2 , formed in accordance with the equation : Ag 2 S 2 8 + 2H 2 = Ag 2 2 + 2H 2 S0 4 . Analysis has shown it to consist in all probability of a mixture THE PERSULPHURIC ACIDS AND PERSULPHATES 35 of silver peroxide and silver persulphate 52 .* When ammonium persulphate is used in place of the potassium salt, less peroxide is precipitated, but there is, nevertheless, considerable decom- position of the persulphate, as indicated by the formation of sulphate and free sulphuric acid. When ammonia is added to the solution containing ammonium persulphate and the silver salt, there is no deposition of peroxide, but a rapid evolution of nitrogen is produced by the oxidation of the ammonia by the silver peroxide; in concentrated solutions the action is very violent and may become uncontrollable. The silver salt acts catalyticaily, a small amount being sufficient to determine the decomposition of large quantities of persulphate and ammonia in accordance with the equation : 3 (NH 4 ) a S 2 8 + 8NH 3 = 6(NH 4 ) 2 SO 4 + N 2 . In the absence of ammonia, an aqueous solution of ammonium persulphate containing silver salts slowly decomposes without the evolution of gas, the change taking place according to the equation : 8(NH 4 ) 2 S 2 O 8 + 6H 2 O = 7(NH 4 ) 2 SO 4 + gH 2 SO 4 + 2HNO 3 .f This reaction has been studied by Marshall and Inglis. 54 v * * ' The catalytic action of silver salts on persuiphates gives rise to many interesting cases o'f oxidation. Minute quantities of manganese may be detected by gently warming the solution with potassium or ammonium persulphate, a moderate quantity of sulphuric acid or nitric acid, and a drop of silver nitrate solution. A distinct permanganate coloration is obtained in half a c.c. of solution containing O'OOi mgm. of manganese, and it is easy to demonstrate the presence of manganese in marble or in a deep sea deposit 51 . When solutions of potassium chloride, bromide, or iodide, slightly acidfied with nitric acid, are oxidized by warm- ing with ammonium persulphate in the presence of silver nitrate, *Tarugi 44 characterizes it as silver permonosulphate, Ag 2 SO 5 , which is very impi$>bable. + Levi and Migliorini 53 state that aqueous solutions of ammonium per- sulphate when heated at 50-100 undergo marked oxidation even in solutions which are originally neutral, or almost so, and become more and more acid owing to the sulphuric acid liberated. Less than the theoretical amount of oxygen is evolved and nitric acid is formed in solution. The addition of sodium hydroxide in increasing proportions to the persulphate solution causes the oxidation first to diminish to a minimum and then to increase in- definitely. 3* 36 PER-ACIDS AND THEIR SALTS chlorates and bromates are formed to some extent, whilst iodides can be converted completely into iodates without much diffi- culty. 55 Ammonium persulphate bleaches indigo and methyl orange in the presence of silver nitrate ; 54 a solution of potas- sium persulphate containing a small quantity of silver nitrate and acidified with nitric acid, oxidizes benzene to quinone, the latter then being further oxidized to maleic acid and carbon dioxide, together with formic acid and carbon monoxide. 56 Oxalic acid can be oxidized quantitatively to carbon dioxide. Similar re- actions take place also in the absence of nitric acid, toluene being oxidized to benzaldehyde and benzoic acid, and thymol to dithymol. 52 When potassium cyanide and persulphate react in alkaline solution in the presence of excess of ammonia, 75 per cent of the cyanide is con- verted into carbamide by way of potassium cyanate, this proportion re- maining constant, no matter how the proportions of the two salts are varied ; 44 this reaction may be used as a means of estimating potas- sium cyanide, an advantage of the method being that it can be carried out in alkaline solution. Complex cyanides, for example, potassium ferro- and ferri-cyanides, are decomposed by persulphates in the presence of a mineral acid, hydrogen cyanide being evolved and small amounts of ammonium salts formed ; 57 on this reaction a method for the analysis of ferro- and ferri-cyanides is based. An interesting reaction of persulphates is that with thio- sulphates. When a slight excess of persulphate is used, tetra- thionate is formed, in accordance with the equation : M 2 S 2 O 8 + 2M 2 S 2 O 3 = 2M 2 SO 4 + M 2 S 4 O 6 , thus affording a ready method for the preparation of tetrathion- ates ; the reaction is accompanied by a considerable evolution of heat. 11 When an excess of thiosulphate is used, however, trithionates are formed, owing to the removal of sulphur from the tetrathionate by the excess of thiosulphate or its decomposition products (the solution becomes faintly acid). In this way Mar- shall 58 has prepared the trithionates of potassium, rubidium and caesium. Some time ago potassium persulphate was put on the market, under the name of " anthion," as a hypo-eliminator in photography, but such a use is open to two serious objections directly contradictory of the advantages claimed : firstly, the tetrathionate which results from the action is just as objectionable as the original thiosulphate, since it de- THE PERSULPHURIC ACIDS AND PERSULPHATES 37 composes easily, giving free sulphur and other deleterious products ; secondly, the persulphate is certain to attack the silver image itself 59 (cf. p. 38). Marshall n first called attention to the action of solutions of persulphates on the metals, many of which dissolve more or less rapidly ; for example, if strips of copper are placed in a test tube and covered with a solution of persulphate, the solution soon be- comes blue, showing that the metal is attacked ; if a fairly con- centrated solution is employed the liquid becomes quite warm. There is no evolution of gas,* the two sulphates being the only product : (NH 4 ) 2 S 2 O 8 + Cu = (NH 4 ) 2 SO 4 + CuSO 4 . The interaction of persulphates and metals has also been studied by other investigators, 62 in particular by Levi, Migliorini, and Ercolini 63 who state that all the metals (elements) examined (Cu, Ag, Au, Mg, Zn, Cd, Hg, Al, Ti, Sn, Pb, As, Sb, Bi, Cr, Se, Te, U, Mn, Fe, Ni, Co, Pd, Pt), except gold and platinum, react with persulphate solutions, either passing directly into solution, or remaining undissolved in the form of oxides or basic salts ; generally, the reaction is slower for ammonium than for potassium persulphate. Magnesium is exceptional in its reaction, which is very violent with ammonium persulphate, causing evolution of ammonia. The results show that those metals go into solution as anions which in their general chemical behaviour exhibit a marked non- metallic character, for example, chromium, manganese, selenium, arsenic, molybdenum, etc. ; some metals of this type, however, such as anti- mony, are transformed into insoluble oxides. Elements which are distinctly metallic in character pass into solution as cathions, the per- sulphate being decomposed, sometimes with evolution of gas. With zinc, mercury, ( cadmium, aluminium, nickel, cobalt, etc., no gas is * According to Levi, Migliorini and Ercolini 60 there is a slow evolution of gas. This reaction may be used for etching copper or silver, thus dispensing with nitric acid. 11 Copper, and copper alloys in which the percentage of copper is predominant, can be coloured black by heating the metal for some minutes at 100 in 5 per cent sodium hydroxide to which I per cent potassium persulphate is added from time to time until the desired colour is obtained. For brass and aluminium bronze, 10 per cent sodium hydroxide must be used. 61 t Tarugi 62 claims that a mercury ammonium persulphate, NH 4 S 2 O 8 .Hg.2NH 3 , is formed by the action of an ammoniacal solution of ammonium persulphate on mercury, or by the action of ammonium persulphate on mercuric chloride. It is said to be decomposed by water, giving ammonium mercurous permono- 38 PER-ACIDS AND THEIR SALTS evolved, and, as for example, in the case of zinc, the reaction may be represented by Zn + M 2 S 2 O 8 = ZnSO 4 + M 2 SO 4 . When gas is evolved, it is found to be hydrogen, at any rate for the concentrated persulphate solutions. The velocity with which these reactions take place does not always bear any evident relation to the oxidizability of the metal, or to its readiness of attack by sulphuric acid. With gold and platinum, catalytic reactions take place very slowly, resulting in the decomposition of the persulphate. Solutions of potas- sium or ammonium persulphate are not affected by colloidal plati- num. 64 Potassium persulphate and hydrogen peroxide react very slowly in aqueous solution 19 ' 65 , the reaction probably being represented by the equation: H 2 O 2 + K 2 S 2 O 8 = 2KHSO 4 + O 2 . Colloidal platinum accelerates the reaction, the hydrogen per- oxide undergoing additional decomposition. The reaction in the absence of platinum is probably unimolecular, 65 due to the for- mation of an unstable intermediate compound, which was ob- tained by Friend in an impure condition by the evaporation of a mixture of hydrogen peroxide and potassium persulphate over sulphuric acid. Price 66 has also found that potassium per- sulphate causes a smaller molecular depression of the freezing point of solutions of hydrogen peroxide than it does of water, indicating that combination between the two compounds takes place to some extent in solution. The action of ammonium persulphate on silver, combined with the catalytic effect of dissolved silver on a solution of ammonium persulphate, affords an explanation of the use of this salt as a density reducer for negatives. Its solvent action will be all the greater the greater the con- centration of the silver salt in solution ; this (local) concentration will be proportionally greater at the denser portions of the negative. 59 This explanation is supported by the observation of Liippo Cramer and von Pinnow that the addition of silver nitrate increases the density- reducing action. 67 The . oxidizing properties of persulphates have been made use of extensively in analytical processes, some of which have been mentioned already. Thus, in the estimation of manganese in irons and steels, the manganese may be obtained in the form of hydrated peroxide by the use of persulphate alone, or, as is now usually the case, it is converted sulphate, NH 4 .SO 6 .Hg. The evidence for the existence of these salts is un- satisfactory ; Levi, Migliorini, and Ercolini characterize the product formed as a basic sulphate. THE PERSULPHUR1C ACIDS AND PgRSULPffATES 39 into permanganate by means of persulphate and silver nitrate, the amount of permanganate being then determined by any of the usual methods. For details, the literature should be consulted. 68 Similarly, chromium may be oxidized to chromate and so estimated in irons and steels. 69 For other uses and reactions of the persulphates the reader is referred to the literature. Mention has already been made of the oxidizing action of persul- phates on organic compounds in the presence of silver nitrate. Numerous such oxidizing actions are known, however, which do not necessitate the use of the latter compound. Many organic colouring matters, as, for example, indigo, litmus, turmeric, are slowly bleached by acid or alkaline persulphate solutions, whilst paper and cloth become quite rotten after being dipped in a solution of potassium persulphate. 7 When heated with a persulphate at 70-80, ethyl alcohol is rapidly con- verted into aldehyde 7 ' 71 , whilst many organic compounds capable of oxidation, especially closed chain compounds, give up hydrogen and yield sulphonic derivatives. Thus, quinol forms an insoluble sulpho- compound, and diaminophenol, /-phenylenediamine, /-aminophenol, each give characteristic colour reactions. 71 On adding a persulphate to a cold solution of aniline hydrochloride a dark green precipitate is formed, which is insoluble in water or alcohol, and when treated with sodium hydroxide or sodium carbonate solution turns dark blue, but remains insoluble. 71 The ammonium salt introduces the hydroxyl group directly into the benzene nucleus, a reaction which is applied technically. Thus 0-nitrophenol in alkaline solution is oxidized to nitrohydroquinone, 72 salicylic acid gives either 2:5-or 2!3-dihydroxybenzoic acid, 73 hydroxy- anthraquinone gives alizarine, alizarine gives purpurine, etc. 74 In acid solution, aniline black (emeraldin) is formed from aniline, whilst in alkaline solution azobenzene, phenylquinonediimide, etc., are formed; in neutral solution, however, a crystalline, orange-brown precipitate is pro- duced, which dissolves in hydrochloric acid with a yellow colour, chang- ing to violet on heating. 10 Many of the alkaloids give characteristic colour reactions, whilst some of them are precipitated as insoluble salts, of which strychnine persulphate,* (C 21 H 22 O 2 N 2 ) 2 H 2 S 2 O 8 ,H 2 O, may be mentioned. 75 * Ammonium persulphate has consequently been recommended as an antidote in cases of strychnine poisoning. For references to the physiological action of persulphates, see Gmelin-Kraut's " Handbuch der anorg. Chem." I. P-l P- 565. 40 PER- ACIDS AND THEIR SALTS Tests for Persulphates. Any of the various characteristic reactions which have already been mentioned may be used as tests for persulphates, as, for example, the various oxidizing actions on inorganic salts and on aniline sulphate ; the formation of the strychnine salt, etc. In some cases these reactions are similar to those obtained with hydrogen peroxide, but persulphates may readily be distinguished from the former by the fact that they neither decolorize perman- ganate, nor give a yellow coloration with solutions of titanium sulphate, nor give perchromic acid with bichromates and sulphuric acid. Furthermore, cerous salts are oxidized to the eerie condi- tion, the yellow solution thus produced not being decolorized by excess of persulphate in the presence of sulphuric acid, as is the case when hydrogen peroxide is used. 76 With a two per cent alcoholic benzidine solution, one part per million of persulphate can be detected by the blue colour produced ; with concentrated solutions of persulphate a brownish-yellow colour is obtained, and also a precipitate, which then dissolves to a dark yellow solution. Hydrogen peroxide reacts similarly only in the presence of proteins, for example, in the presence of milk. Estimation of Persulphates. Numerous methods have been proposed for the estimation of persulphates, but of these it will be sufficient to mention the following, of which the first is the one in most general use. 1. The persulphate solution, acidified with sulphuric acid, is treated with excess of ferrous sulphate solution at 60-80, and the excess of the latter determined by titration with standard permanganate. At ordinary temperatures the reaction is com- paratively slow, and erroneous results may be obtained. 78 2. Two or three grams of the salt are dissolved in 100 c.c. of cold water, and then heated at 60-80 in a stoppered bottle with 0-5 gram of potassium iodide for ten minutes. The liberated iodine is titrated with standard thiosulphate. 79 In some cases the solution is allowed to stand at the ordinary temperature for twelve hours or longer, and the liberated iodine then titrated. This method is useful in certain cases, but it is of doubtful accuracy, since oxidation of some of the iodine to iodic acid may take place, especially at high temperatures. 48 THE PERSULPHURIC ACIDS AND PERSULPHATES 4* 3. The persulphate solution is boiled for at least twenty-five minutes in order to bring about decomposition, and the acidity then determined by titration with standard alkali. 44 The time necessary for the decomposition can be considerably shortened by the addition of a little methyl alcohol. 80 For other methods, or criticisms of the above the literature should be consulted. 81 Mention has been made of the fact that persulphates do not de- colorize permanganate, and extended use has been made of this in the estimation of hydrogen peroxide in the presence of persulphates. Friend 82 has found, however, that unless certain precautions are taken the volume of permanganate solution necessary in the titration of a mixture of hydrogen peroxide and potassium persulphate can vary con- siderably, being always less than that required theoretically for the hydrogen peroxide in the mixture. The longer the time of titration, the diluter the solution to be titrated, and the less sulphuric acid it contains, the greater is the deficiency in the permanganate solution required. This is due to a reaction taking place between the hydrogen peroxide and the persulphat^M^the titrated solution ; for every molecule of hydrogen peroxide not accounted for by titration with the permanganate, a molecule of persurpBate disappears, which is in accordance with the equation : H 2 O 2 + K 2 S 2 Og = K 2 SO 4 + H 2 SO 4 + O 2 . Price and Denning 19 have shown that in a solution containing hydrogen per- oxide and potassium perl^Kate the concentration of both compounds gradually diminishes, the^miinution in that of the persulphate being appreciable, however, only after long keeping (30 hours). Under the conditions observed by Friend the reaction takes place quickly, probably because of the catalytic effect of some manganese compound * produced in the solution by the reduction of the permanganate, f Under certain limitations with respect to the concentration of the persulphate, Friend was able to devise a method for the estimation of hydrogen peroxide in the presence of persulphate (cf. also, Skrabal and Vacek ^J. Formula and Constitution of the Persulphates. * Formula. So far, it has been assumed that the formula of persulphuric acid is H 2 S 2 O 8 , but no evidence in support of this has been given. When Marshall first obtained potassium per- sulphate, Walker measured the equivalent conductivity of the * Friend 82 finds that manganese sulphate, and therefore the manganese ion, does not catalyse the reaction. t Skrabal 83 prefers to regard the reaction as " induced," rather than as catalytically accelerated. PER- ACIDS AND THEIR SALTS solutions and found that the difference between the values at dilutions of 1 024 and 64 litres was I57(4 1024 = 1407;^= 125-0).* Since Ostwald had found that the corresponding difference in the case of potassium perchlorate was 1 5'Ojt it was concluded that 85 re- the molecular formula of the salt was KSO 4 . Bredig peated the conductivity measurements, with the following results : Dilution in litres ... 32 64 128 256 512 1024 Equiv. conductivity in mer- cury units . . . 118-4 !26-3 132-7 137-1 140-7 143-5 Equiv. conductivity in recip. ohms.J .... 126-7 135-1 142-0 146-7 150-5 153-5 In the case of other potassium salts of known basicity, it had been found that (-4 1024 - A^)/ basicity = IO'8. The basicity of per- sulphuric acid is therefore : (A im - 4J/I0.8 = (143-5 - i i8-4)/io-8 = 2-3, that is, it is dibasic, the formula of the potassium salt being therefore K 2 S 2 O 8 . Additional evidence in favour of this formula was obtained by Moeller 86 from a comparison of electrical con- ductivity and freezing-point measurements. In the following table, g is the weight of potassium persulphate in grams, G is the weight of water, A is the observed depression of the freezing point, M is the molecular weight of the potassium persulphate, calculated from the freezing-point measurements, and A v is the equivalent conductivity (in mercury units) : 8- G. A. M = 100 x 18-9 x -^ Atj A* I 0-5969 130-2960 0-084 103-07 118-20 2 0-6831 130-8327 0-093 106-11 II5'32 3 1-3602 130-3869 0-183 107-74 109-74 4 0-4375 I30-3273 0-064 99-135 126-93 Now M (rea i) = i M (found) , where i is van't Hoff's factor. Also, if a is the dissociation of the salt, * Expressed in mercury units. t This difference has since been found to be incorrect, which invalidates conclusions drawn as to the formula of the persulphates. J These numbers are recalculated from Bredig's figures. Similar measurements, but not so complete, were made by Lowenherz. 87 THE PERSULPHURIC ACIDS AND PERSULP HATES 43 where n is the number of ions into which the salt dissociates. The dissociation can be calculated from the conductivity measurements by the relation a = A v /A 00 , A^ being taken as 145-89, and hence / can be calculated, assuming different values for n (= 2 if the salt is KSO 4 ; = 3 if the salt is K 2 S 2 O 8 ). A comparison of the values of i so obtained is given in the following table, assuming that the salt is KSO 4 in the one case, and K 2 S 2 O 8 in the other : KS04. K 2 S 2 8 . t (from A). (from A). (from A). i (from A). I 2 3 4 I'3I3I I-2758 1-2589 1-3648 I-8I02 1-7905 1-7522 1-8700 2-6262 2-5516 2-5I78 2*7296 2-6205 2*5810 2-5 44 2*7400 The values of i give the best agreement on the assumption that the salt is K 2 S 2 O 8 , which may therefore be considered to be the molecular formula. Similar results were obtained with ammonium persulphate.* Constitution. It has already been pointed out that sulphuric acid of a strength above 30 per cent (approx.), and a high current density at the anode are necessary for the formation of per- sulphuric acid. Helmholtz in his Faraday lecture 89 put forward the hypothesis that sulphuric acid dissociates not only into H- and SO" 4 -i ns > but also into H- and HSO' 4 -ions, and Richarz 90 made use of this hypothesis, suggesting that persulphuric acid is formed at the anode by the combination of two HSO' 4 -ions of opposite sign, the reaction being expressed by the equation : 2HSO' 4 + 2 = HSO' 4 + HSO 4 " = H 2 S 2 O 8 . This explanation is substantially the one which is now adopted, but instead of making the improbable assumption that the HSO' 4 - ions may be both positive and negative, the mechanism of the reaction is assumed to be, that two HSOVions lose their charges at the anode, and the neutral residues then combine with the formation of H 2 S 2 O 8 , thus: 2HSO' 4 + 2@ = H 2 S 2 O 8 . This for- mulation of the reaction affords an explanation of the necessity for a high anodic current density, and of sulphuric acid above a certain minimum concentration ; if there are no HSO' 4 -i ns * Additional evidence has been given by Lowenherz. 88 44 PER-ACIDS AND THEIR SALTS present in the solution it would not be possible for persulphuric acid to be formed. The constitution ordinarily assigned to sulphuric acid is \ S / OH . O^ \OH ' that of the HSO' 4 -ion would therefore be, , O/' \OH ' and consequently that of persulphuric acid is O^ \OH HO/' (compare Melikoff and Pissarjewsky 91 ), that is, it is a derivative of hydrogen peroxide (HO. OH) in which each hydrogen has been replaced by the group HSO 3 . This constitution is supported by the synthesis of persulphuric acid from chlorsulphonic acid and hydrogen peroxide, which would take place according to the equation : H;O.O;H ci;\ /o o> \-, 2 / \ ' : )S0 2 = 2 S/ )S0 + 2HC1. OH HO/ \OH HO/ It should be mentioned that another formulation of the mechanism of the production of persulphuric acid has been given. Friessner 92 found that the electrolytic oxidation of sulphites to dithonates proceeds according to the equation : 2SO" 3 + O + H 2 O = S 2 O" 6 + aOH', and suggested that persulphates were formed similarly : 2SO" 4 + O + H 2 O = S 2 O" 8 + 2 OH'. This method of formulation, however, offers a number of difficulties in the explanation of the conditions necessary for the production of persulphuric acid and its salts. Kastle and Loewenhart 93 have assigned the constitution /OH SA< \0 2 H to persulphuric acid ; if this were so, it would probably be monobasic, owing to very slight acidity of the hydrogen atom in the O 2 H-group (cf. p. 54). THE PERSULPHURIC ACIDS AND PERSULPHATES 45 Permonosulphuric Acid (Caro's Acid). In the course of an investigation of the action of persulphates on aniline, Caro 10 added pure ammonium persulphate to con- centrated sulphuric acid as long as solution took place. After a few minutes a small portion of the solution was diluted with water and then an aqueous solution of aniline added. The mixture re- mained clear and colourless, no aniline black being formed, as is ordinarily the case with a persulphate (see p. 39). On neutrali- zation with ammonium carbonate the liquid became green and nitrosobenzene was precipitated, but no aniline black. A similar result was obtained with potassium persulphate. It was t evident that under the action of the concentrated sulphuric acid the persulphate had undergone a change resulting in the forma- tion of a new compound possessing different oxidizing properties. The change was found to depend chiefly on the concentration of the sulphuric acid, but it depended also on the quantities taken and on the time of action ; with ordinary concentrated sulphuric acid the change was complete in a few minutes, but with acid of density I '2 it was not complete even after several months. It was further found that a similar compound is produced by electrolysis of sulphuric acid of density 1*45. This latter observation con- firmed the results of Traube s which have already been referred to (p. 1 2). In 1900, Baeyer and Villiger 94 ascribed the formula H 2 SO 3 to this new acid, which they denoted as Caro's acid.* The evid- ence for this formula was given in 1901, when the same authors published an important paper on the subject. 17 ( They pointed out that three methods could be used for its preparation, namely: (i) Treatment of potassium persulphate with concentrated sul- phuric acid. (2) Electrolysis of a tolerably concentrated solution of sulphuric acid. 95 (3) Action of concentrated sulphuric acid on concentrated hydrogen peroxide, a reaction first discovered by Berthelot. 96 J They give, at the same time, the explanation, * This name will now be used until the evidence for the formula has been given, since confusion will thereby be avoided. t Other papers on the same subject had been published previously, but these will be referred to later. I According to Tubandt and Riedel 97 freshly precipitated nickel di- oxide gives a solution containing Caro's acid when treated either with sul- 46 PER-ACIDS AND THEIR SALTS which is now accepted as the correct one, of the formation of Caro's acid by the electrolysis of concentrated sulphuric acid, persulphuric acid being the primary product, which then under- goes further hydrolysis (cf. p. 9). In their experiments it was necessary for them to devise methods for the estimation of persulphuric acid and Caro's acid, and they made use of the fact that the latter compound liberates iodine from acid solutions of potassium iodide practically instan- taneously, even in dilute solutions, whereas with persulphuric acid the reaction is comparatively slow, being complete in dilute solu- tion only after twelve to twenty-four hours. Hydrogen peroxide was estimated by titration with permanganate, a process which is of doubtful accuracy (cf. p. 41); the presence of this compound was tested for with titanium sulphate, which gives no colour reac- tion with either of the persulphuric acids. The method of preparation of a solution of Caro's acid was as follows : Ten grams of potassium persulphate were triturated at a low tem- perature with 20 grams of concentrated sulphuric acid, and the mixture allowed to stand for one hour,* after which time it was diluted by pour- ing on to crushed ice. The sulphuric acid present was then precipitated by the addition of the calculated quantity of a solution of barium hydrogen phosphate,f freshly prepared by the addition of phosphoric acid to the calculated quantity of hot barium hydroxide solution, and the liquid filtered through a porous pot filter. The filtrate, having a volume of about 1500 c.c., was treated with a current of air under diminished pressure until it no longer had the odour of ozone, or of bleach- ing powder (cf. p. 55) ; it contained no hydrogen peroxide and was fairly stable, depositing only a trace of barium sulphate after standing for months ; it contained about 16 per cent of undecomposed persulphuric acid. A determination of the amount of iodine immediately liber- ated by this solution from potassium iodide, and of the sulphuric acid thus formed gave the ratio of active oxygen to sulphuric acid phuric acid (dilute or strong) or with a concentrated solution of potassium hydrogen sulphate. This may indicate that the higher oxide of nickel has a peroxide constitution, but the evidence, as yet, is inconclusive. * If a longer time were used, the mixture began to decompose, with evolu- tion of oxygen and ozone. t Barium carbonate or hydroxide decomposes Caro's acid. 8 THE PERSULPHURIC ACIDS AND PERSULPHATES 47 formed as I : 1*158, and I : 1*154 for the Caro's acid. This pointed to the formula H 2 SO 5 , for which the theoretical ratio is I : I . In 1900, Lowry and West 98 attacked the question of the per- sulphuric acids in a different manner. In connexion with the part which persulphuric acid plays in the lead accumulator 20 and with the possible effect of the production of these acids on the conduc- tivity of solutions of sulphuric acid " the theory had been put forward that in very concentrated acid a persulphuric acid of the formula H 2 O 2 ,4SO 3 , is formed, and that with increasing dilution this gradually breaks down in the stages H 2 O 2 ,3SO 3 , H 2 O 2 ,2SO 3 and finally H 2 O 2 ,SO 3 . To test this theory, Lowry and West carried out an extended investigation of the interaction between sulphuric acid and hydrogen peroxide ; this was all the more necessary, since Berthelot in his experiments on the same subject had not come to the conclusion that the proportions of hydrogen peroxide and persulphuric acids formed were interdependent. The reaction mixtures were made, for the most part, by the careful mixing of sulphuric acid and hydrogen peroxide of different strengths. In some cases electrolysis of sulphuric acid was resorted to, and in one case the mixture was made by the action of concentrated sulphuric acid on ammonium persulphate. After keeping the mixtures at 18 until equilibrium had been obtained, they were diluted by running on to ice, and in the solution so obtained, the hydrogen peroxide was first estim- ated by titration with permanganate, a little manganese sulphate being previously added to catalyse the reaction, and then the " persulphuric acid " in the same solution by adding an excess of ferrous sulphate solution, and titrating back in the usual way. It was found that the rate at which equilibrium is attained depends on the concentration of the acid, and that equilibrium results only in the presence of a large excess of acid. The con- version of hydrogen peroxide into persulphuric acid takes place within very narrow limits of concentration ; for example, on in- creasing the concentration of the acid from 70 to 80 per cent the proportion of " persulphuric acid oxygen " increases from 31 to 78 per cent ; 85 per cent of the change takes place between the limits H 2 SO 4 ,H 2 O and H 2 SO 4 ,4H 2 O. From the results obtained a curve was plotted showing the relation between the percentage concentration of the sulphuric acid and the percentage of " persulphuric acid oxygen " in the mixture. This curve could be represented approximately by the formula CJC. 2 = ^(C 3 /C 4 ) 2 , 48 PER-ACIDS AND THEIR SALTS where C lf C 2 , C 3 and C 4 , represent the molecular proportions of the persulphuric acid, hydrogen peroxide, sulphuric acid and water respec- tively in the mixture ; that is, the reaction taking place would be represented by the equation H 2 O 2 + 4 H 2 SC) 4 <^ H 2 S 4u + 4 H 2 O. The value of k did not remain constant, however, gradually increasing from 1 1*48 to 15*81 as the concentration of the sulphuric acid diminished from 82-81 to 65-64 per cent in the equilibrium mixture. The curve was very accurately represented by the equation : k^ and / 2 having the values 0-63 and 10-96 respectively. The authors consequently drew the conclusion that the mixture consists mainly of a pertetrasulphuric acid, H 2 S 4 O 14 , which would be Caro's acid, and of per- sulphuric acid, H 2 S 2 O 8 , this latter acid accounting for the term (C 3 /C 4 ) 2 in the above equation, the reaction being : H 2 O 2 + 2H 2 SO 4 ^ H 2 S 2 O 8 + 2H..O. The formula H 2 SO 5 is not in accordance with their results. In connexion with the above results and conclusions it should be noted that the amount of hydrogen peroxide in the equilibrium mixture was estimated by titration with permanganate, a procedure which has since been shown to be untrustworthy (cf. p. 56). Since, at present, there is no other evidence for the existence of a pertetrasulphuric acid, it is probable that Lowry and West's re- sults were vitiated by the above source of error.* A different method from any of the above was adopted by Price i02 for investigating the formula of Caro's acid. Its formation from potassium persulphate and concentrated sulphuric * Armstrong and Lowry 10 think it is possible that pertetrasulphuric acid may exist in the concentrated solution before dilution. The results of Lowry and West cannot, however, be used in support of this supposition, since although equilibrium was attained in concentrated solution, the solutions were diluted before the components were estimated by titration. Ahrle m has also shown, as follows, that pertetrasulphuric acid does not exist in the diluted solution. Caro's acid was made from 30 per cent hydrogen peroxide and 98-5 per cent sulphuric acid and diluted by pouring on to crushed ice. The sul- phuric acid in the solution was precipitated with barium phosphate and any hydrogen peroxide present destroyed with a few drops of permanganate and a known quantity of sulphuric acid. To a portion of the resulting solution potas- sium iodide was added and the liberated iodine titrated to determine the active oxygen ; in another portion the active oxygen was destroyed with hydrochloric acid and potassium iodide and the sulphuric acid determined as barium sul- phate. The ratio of active oxygen to barium sulphate was 1:1-2, which is in total disagreement with the formula H2S 4 O J4 . THE PERSULPHURIC ACIDS AND PERSULPHATES 49 acid can be represented by the following equations: (i) If the formula is H 2 SO 5 , we have : K 2 S 2 O 8 + H 2 SO 4 = H 2 S 2 O 8 + K 2 SO 4 . The H 2 S. 2 O 8 is then hydrolysed by the water present according to the equation : H 2 S 2 O 8 + H 2 O = H 2 SO 5 + H 2 SO 4 . For every molecule of Caro's acid formed, the acidity (expressed in grams H 2 SO 4 , and assuming H 2 SO 5 to be dibasic) will increase by the equivalent amount, in this case by one molecule. (2) For the formula H 2 S 4 O 14 the first equation would be : 2K 2 S 2 O 8 + 2H 2 SO 4 = 2H 2 S 2 O 8 + 2K a 97 P 61 " cent acid after eight days contained 827 per cent H 2 SO 5 with 15*3 per cent H 2 S 2 O 8 . It reacts with organic com- pounds similarly to persulphuric acid (v. p. 22), but the reaction is not so violent; for example, paraffin is hardly changed in the cold, and neither ether nor alcohol give explosions. 22 No pure metallic salts have been prepared, but the aniline * It could also be obtained by mixing the calculated quantities of anhy- drous hydrogen peroxide and persulphuric acid. 22 54 PER-ACIDS AND THEIR SALTS salt has been obtained by the interaction of aniline and the acid in ethereal solution. 22 When freshly prepared it is colourless, but it rapidly becomes coloured, especially in moist air. The anhydrous acid reacts with a further molecule of chlor- sulphonic acid, giving persulphuric acid. H 2 SO 5 + C1.SO 3 H = H 2 S 2 O 8 + HC1. The Constitution of Permonosulphuric Acid is given by its syn- thesis from chlorsulphonic acid and hydrogen peroxide. /;C1 Hio.OH /O.OH 2 S( ' = 2 S( + HC1. \OH \OH The hydrogen in the hydroxyl group directly attached to the sulphur is presumably the one replaced by metals in the forma- tion of salts. The remaining hydrogen atom would possess only very weak acid properties, if any at all, as it is contained in the group -O.OH, derived from H 2 O 2 , which is only a very weak acid. Also, from analogy to other dibasic acids, the acidity would be very much diminished by the presence of the much more strongly acid hydrion in the OH-group directly attached to the sulphur. The potassium salt would therefore have the formula : /O.OH 2 S( \OK and this is supported by the formation of the benzoyl and ben- zene sulphonyl derivatives, and also by the decomposition they undergo. For example, in alkaline solution the former is readily decomposed with the formation of benzoic acid and permonosul- phuric acid, whilst when treated with the equivalent quantity of sulphuric acid in ethereal solution, perbenzoic acid and sulphuric are formed. 105 These reactions are explained by the following equations : /o.o:coc 6 H 5 i /O.OH O 2 S( , + HOlH = O 2 S( + C 6 H 5 .COOH. \OH " \OH /;o.o.coc 6 H 5 2 S/ I + HjOH = H 2 S0 4 + C 6 H 5 .CO.O.OH. \6"H"" The formation of permonosulphuric acid from persulphuric acid by hydrolysis takes place thus : H-OH /o; o\ /OH /O.OH o 2 s( >so 2 = o,s< + o 2 s<; \OH HO/ \OH \OH THE PERSULPHURIC ACIDS AND PERSULPHATES 55 Properties of Permonosulphuric Acid (in Solution). Even when the anhydrous substance is dissolved in ice-cold water, slight decomposition takes place with the formation of hydrogen peroxide : H 2 SO 5 + H 2 O = H 2 SO 4 + H 2 O 2 . A cryo- scopic determination. of the molecular weight in aqueous solution after allowing for the sulphuric and persulphuric acids present, gave the value 55 ; 22 H 2 SO 5 demands 104, so that the acid is prac- tically completely dissociated into H'- and HSO' 5 -ions. Baeyer and Villiger 17 found that the aqueous solution is fairly stable in the presence of 8 per cent sulphuric acid. In a solution containing 87*05 per cent H 2 SO 5 and 12-95 per cent H 2 S 9 O 8 in 8 per cent sulphuric acid, the percentage of permono- sulphuric acid was 88*98 * after keeping seven days at the ordinary temperature ; hydrogen peroxide could just be detected at the end of that time.f The solution is much more stable than is generally supposed. Thus Price 104 found that 10 c.c. of a solution prepared from potassium hydrogen sulphate and hydrogen peroxide were equiva- lent to 27*8 c.c. of a thiosulphate solution on 23 May, to 27*1 c.c. on 15 August, and to 26*8 c.c. on 14 September. Also, a dilute solution prepared from potassium persulphate and sulphuric acid showed very little diminution in strength after keeping for two years, and there was no formation of hydrogen peroxide. The solution generally has a strong odour resembling that of bleaching powder I 17 which is also, accompanied by that of ozone 106 ; the impure salt obtained by Price possessed no such odour, but when dissolved the solution acquired it again after a short time. A slightly alkaline solution rapidly decomposes, but the solu- tion is much more stable in the presence of a large excess of potassium hydroxide, 105 possibly because of the formation of a dipotassium salt. Price 103 noticed that in titrating the * The increase in the percentage is due to the decomposition of some per- sulphuric acid with the formation of the permono-acid. t It should be mentioned that a dilute solution, freshly prepared from potassium persulphate and concentrated sulphuric acid, generally contains traces of hydrogen peroxide, as shown by the titanium sulphate test, but that on keeping, the hydrogen peroxide gradually disappears. I Baeyer and Villiger 17 attribute this odour to the presence of a com- pound of the formula SvO 8 ( = 2H 2 SO 5 - 2H 2 O). 56 PER-ACIDS AND THEIR SALTS solution with potassium hydroxide, using phenolphthalein as indicator, the colour change was not sharp, and pointed out that this was probably due to the slight acidity of the H in the - O.OH group. The concentrated solutions (92 per cent) prepared by Ahrle 101 have a strong odour of ozone and rapidly decompose at body tem- perature. The decomposition is explosive in the presence of smooth or finely divided platinum, manganese dioxide, or silver. Zinc dust, lead dust, magnesium powder, and wood charcoal have no violent action, and the solution is not decomposed by finely divided iron. Wool and cellu- lose are carbonized almost immediately, whilst with cotton there is at first no action, and then after a few seconds a violent action, accom- panied by a yellow flame. A dilute solution of permonosulphuric acid prepared from the persulphate and sulphuric acid undergoes decomposition very slowly in the presence of colloidal platinum ; when mixed with hydrogen peroxide, however, there is a rapid evolution of oxygen, which is probably due to the reaction : H 2 S0 5 + H 2 2 = H 2 S0 4 + H 2 + O 2 , which is catalysed by the platinum. 102 ' 107 If potassium permanganate solution is added to a dilute solution of permonosulphuric acid which contains no hydrogen peroxide, there is no action between the two, even after several hours ; on adding a little manganese sulphate, however, a slow reaction takes place, the perman- ganate being reduced and the acid decomposed. After a short time, a deposition of a higher oxide of manganese takes place, and the evolution of gas becomes more marked, the supernatant liquid retaining a red colour, which is not due to permanganate. Manganese sulphate alone also decomposes the acid. 102 - 108 It has been mentioned several times that the usual method of estimating hydrogen peroxide in the presence of permono- sulphuric acid has been by titration with permanganate, a little manganese sulphate being added to catalyse the reaction between the hydrogen peroxide and the permanganate. Price 102 has shown, however, that the method is inaccurate, even in dilute solution, as is indicated by the following. Varying quantities of a solution of permonosulphuric acid, containing only a small amount of hydrogen peroxide, were titrated, the same amount of manganese sulphate being added in each case. The titrations were : THE PERSULPHURIC ACIDS AND PERSULPHATES 57 c.c. H 2 SO 5 ... 5 10 15 20 30 I 5 25 c.c. KMn6 4 . . . 0-30 0-42 0-60 0-69 o'94 1-75 6-15 In another case, a mixture of hydrogen peroxide and per- monosulphuric acid was made and then titrated ; 10 c.c. required 5 '60 c.c. of permanganate, whilst 25 c.c. required only ryoSS c.c. of the permanganate. It follows that the longer the titration ' *? takes, the less, proportionally, is the amount of permanganate ) necessary (cf. p. 41). This is probably due to an induced chem- \ ical reaction between the hydrogen peroxide and the permono- / sulphuric acid, or to the catalytic effect on this reaction of some manganese compound formed in the solution. These results would account for the observations of Bach, which led to a prolonged controversy 108> 109 , that when a solution containing permonosulphuric acid and hydrogen peroxide was titrated with perman- ganate, the amount of oxygen evolved was always in excess of that which would be expected from the permanganate used. With titanium sulphate a solution of permonosulphuric acid gives no coloration.* Chlorine is liberated from aqueous or gaseous hydrogen chloride, and bromine from hydrogen bromide, but fluorine could not be obtained from hydrogen fluoride. 110 A neutral solution of a permonosulphate froths violently on the addition of silver nitrate, an ozonized gas being evolved. A similar reaction is produced by other catalysts, such as the dioxides of man- ganese and lead, platinum sponge, etc. If a copper salt and then sodium hydroxide are added to a solution of permonosulphuric acid, a deep brownish-black precipitate, probably copper peroxide, is produced, which soon decomposes with evolution of oxygen ; this precipitate is totally different in appearance from that produced with hydrogen per- oxide under the same conditions. 106 Baeyer and Villiger 108 recommend a mixture of potassium permanganate, permonosulphuric acid and dilute sulphuric acid as a very strong oxidizing agent. The method used for the Estimation of Permonosulphuric Acid has already been mentioned (p. 46). Permonosulphuric acid has been extensively used as an * Bach says that the undiluted solution made from potassium persulphate and 100 per cent sulphuric acid gives a yellow colour with titanium sulphate, and that hydrogen peroxide could not be present since 100 per cent sulphuric acid was used. No proof, however, is given that the sulphuric acid was really joo per cent. 58 PER-ACIDS AND THEIR SALTS oxidizing agent in organic chemistry. It is distinguished from persulphuric acid by the fact that the amino-group of primary amines is directly oxidized to the nitroso-, and finally to the nitro- group. 10 It has found extended use as an oxidizing agent, chiefly in the hands of Baeyer and of Bamberger. 111 The formation of hydrogen peroxide from permonosulphuric acid by hydrolysis in the presence of sulphuric acid is used on a large scale for the manufacture of hydrogen peroxide. The permonosulphuric acid is prepared either by the electrolysis of sulphuric acid 112 or by treatment of potassium persulphate with sulphuric acid. 113 The hydrogen peroxide is obtained by distillation under diminished pressure.* The chief difficulty met with in connexion with this method of manufacture is the catalytic decomposition of the hydrogen peroxide by traces of impurities chiefly platinum in the solution. This can be over- come by depositing the impurity electrolytically by means of a subsidiary cathode in the anolyte, or simply by hanging a rod of aluminium in the electrolyte, the platinum then depositing on it. Malaquin 1U has shown that ozone is produced by the action of concentrated nitric acid on ammonium persulphate at 6o-70 ; the gas evolved contains 3-5 per cent, of ozone and 94-95 per cent of oxygen, together with some nitrogen. Perselenates. Perselenates have not yet been obtained in a pure condition. An impure potassium perselenate, analogous to potassium per- sulphate, was obtained by Dennis and Brown 1 by the electro- lysis in a divided cell of a saturated solution of potassium selenate containing a little free selenic acid. Platinum electrodes were used, and the temperature maintained at 4. A white solid substance was obtained which was not free from selenate, the highest percentage of perselenate being 74*44 per cent. When the aqueous solution of the salt is warmed, oxygen is evolved. It rapidly oxidizes ferrous and thallous sulphates in the cold, and when hot, converts manganese dioxide into per- manganate. * In some of the patent specifications it is proposed to extract it by means of a suitable solvent. CHAPTER III. PERBORATES. The perborates differ from the persulphates in that they are most readily obtained from the borates by direct interaction with hydrogen peroxide or with the peroxides of the alkali metals, and not by the electrolysis of solutions of the borates. The best known salts are derivatives of metaboric acid, HO.B:O ; since they are very readily hydrolysed in aqueous solution, with the formation of hydrogen peroxide, their constitution may be repre- sented as MO.OB:O, where M is a univalent metal. The existence of perborates was first demonstrated by Etard, 1 who found that by the interaction of a saturated solution of boric acid and hydrated barium peroxide, a compound was pro- duced which possessed oxidizing properties, liberating chlorine from hydrochloric acid, and giving oxygen with dilute acids ; it was not till 1898, however, that perborates were obtained in a pure condition by Melikoff and Pissarjewsky. 2 The preparation and properties of these compounds will be best understood by reference to sodium perborate. Sodium perborate, NaBO 3 ,4H 2 O, is readily obtained when a saturated solution of borax containing an equivalent quantity of sodium hydroxide is treated with an excess of hydrogen peroxide (double the amount necessary for the production of NaBO 3 ) ; after some time, large, transparent, prismatic crystals of the salt separate. 3 Na2B 4 O 7 + 2NaOH + 4 H 2 O 2 = 4 NaBO 3 + sH 2 O. Tanatar 4 states that perborates are produced at the anode by the electrolysis of concentrated solutions of sodium ortho- borate, but Constam and Bennet 5 were not able to confirm Tanatar's results. Bruhat and Dubois, 6 however, support Tanatar's observations, and Pouzenc 7 has patented a process for the manufacture of sodium perborate by electrolysis. 59 60 PER- ACIDS AND THEIR SALTS The mechanism of the electrolytic formation of sodium per- borate is not clear. It may be formed by direct oxidation at the anode, or, as Tanatar 8 suggests, by oxidation of the borate by hydrogen peroxide formed at the anode. Constam and Bennet 9 object to the latter explanation, because such a dilute solution of hydrogen peroxide is formed under the experimental condi- tions that it is not capable of converting borates into perborates. This objection is not necessarily valid, however, since the local concentration of the hydrogen peroxide produced at the anode may be sufficient for the purpose. The hydrogen peroxide would be produced by the discharge of OH'-ions contained in the solu- tion of sodium orthoborate used. It is true that Riesenfeld and Reinhold, 10 owing to the ready decomposition of hydrogen peroxide in the presence of sodium hydroxide, were unable to obtain it at the anode by the electrolysis of solutions of sodium hydroxide, although it was formed in quantity when potassium hydroxide was used. In the electrolysis of the solution of the orthoborate the conditions are probably more favourable for the existence of hydrogen peroxide than in solutions of pure sodium hydroxide. Sodium perborate, NaBO 3 ,4H 2 O, forms large, transparent, monoclinic prisms, which are stable in the air.* By careful dehydration the monohydrate, NaBO 3 ,H 2 O, can be obtained ; the remaining molecule of water can be removed in a vacuum over phosphoric oxide. 6 One litre of water dissolves about 25-5 grams at 15, 26-9 at 21, 28*5 at 26, and 37-8 at 32 ; 12 the solubility is greater in solutions containing boric, tartaric, or citric acids, or glycerol 6 ; it is also increased by small quantities of magnesium sulphate or ammonium sulphate. 12 The heat of solution in water at 16-1 is - 1 1-564 Cal., and in N/2 sulphuric acid at 17-29, -8-950 Cal. 4 The aqueous solution is distinctly alkaline and contains hydrogen peroxide ; at temperatures above 40 it begins to de- compose with evolution of oxygen. The solution possesses oxid- izing properties, liberating iodine from potassium iodide, chlorine from hydrochloric acid, and oxidizing ferrous to ferric salts, chromic to perchromic acid, manganous salts to manganese * Jaubert " states that the salt is not affected by atmospheric carbon dioxide, whereas, according to Tanatar 3 , perborates are rapidly decomposed by carbon dioxide. PERBORATES 6 1 dioxide, etc. Other characteristic reactions are given by Chris- tensen, 13 and Bruhat and Dubois. 6 By the addition of the powdered salt to 50 per cent sulphuric acid, boric acid is precipitated and a concentrated solution of hydrogen peroxide is obtained.* 11 The distribution coefficient of hydrogen peroxide between water and ether is affected by the presence of borates in the aqueous phase, the relative concentration in this phase increasing ; this indicates that perborates exist as such in aqueous solu- tion. By extraction experiments with ether, Pissarjewsky u has shown that in a solution of sodium perborate in water at 25 at least 60 per cent of the peroxide oxygen is present as free hydrogen peroxide. He considers that at 25 sodium perborate, NaBO 3 , does not exist as such in solution, but decomposes (hydrolyses) according to the equation : NaO- 2 BO + H 3 O = NaO 2 H + HBO 2 . The sodium hydroperoxide thus formed hydrolyses to some ex- tent into hydrogen peroxide and sodium hydroxide, 15 and a portion of the sodium hydroxide thus formed combines with the boric acid, so that there are present in solution sodium borate, boric acid, sodium hydroperoxide, hydrogen peroxide, sodium hydroxide, and traces only of sodium perborate. At lower temperatures sodium perborate is increasingly formed ac- cording to the equation : NaO 2 H + HO.BO = NaO 2 BO + H 2 O. At o practically only sodium perborate is present, as can be de- duced from conductivity measurements made by Constam and Bennet 5 ; if only 5 per cent hydrogen peroxide had been present in the solution at o they could not have carried out their measure- ments at dilutions of 32 and 64 litres owing to the catalytic de- composition of the hydrogen peroxide by the electrodes. These conductivity measurements show, at the same time, that sodium perborate has the formula NaBO 3 and not Na 2 B 2 O 6 , since Aio-24 - AJQ = 40'I - 30-7 = 9-4 which, according to Ostwald's rule, indicates that the acid is monobasic. * In the presence of sulphuric acid of this strength permonosulphuric acid would be formed, so that to obtain solutions containing hydrogen peroxide alone it would be necessary to use a diluter acid. 62 PER-ACIDS AND THEIR SALTS Jaubert 11 describes the preparation of a sodium perborate having the composition Na 2 B 4 O 8 ,ioH 2 O. It crystallizes from a solution made by adding gradually an intimate mixture of 248 grams of boric acid and 78 grams of sodium peroxide to two litres of cold water. Its solubility at 11, 22, and 32 is 42, 71, and 138 grams per litre of solution. The solution, which is fairly stable at ordinary temperatures, is alkaline and contains hydrogen peroxide. The salt cannot be recrystallized from water, the first crystals which separate being richer in active oxygen than those separating later, the last fraction not containing any active oxygen. Ordinary sodium perborate crystallizes when a quantity of hydro- chloric acid equivalent to half the sodium in the perborate is added to the solution. Melikoff and Pissarjewsky 2 have obtained indications that a higher perborate than NaBO 3 , namely NaO.BO 3 , exists, but the substance could not be isolated in the pure condition. Other Perborates. Ammonium perborate, NH 4 BO 3 ,-JH 2 O, is obtained when boric acid is dissolved in aqueous hydrogen peroxide (2-5 per cent) and ammonia added to the solution ; it is deposited in colourless isotropic crystals on the addition of alcohol. The freshly pre- cipitated salt contains 3H 2 O, but when kept for twenty- four hours over concentrated sulphuric acid the salt containing ^H 2 O is ob- tained. 2 It can be completely dehydrated by keeping over phosphoric oxide. 5 Tanatar 4 describes an ammonium perborate with one molecule of water of crystallization. In 1 897, Melikoff and Pissarjewsky 16 described the pre- paration of an ammonium derivative of hydrogen peroxide, namely (NH 4 ) 2 O 2 ,H 2 O 2 , or perhaps, NH 4 O.OH. If this is considered to be a metallic hydroperoxide it should form salts with perboric acid, for example, NH 4 O.O.OBO, or NH 4 BO 4 . Petrenko 17 has isolated such a compound, having the com- position NH 4 BO 3 ,NH 4 BO 4 ,2H 2 O, by the action of hydrogen peroxide on ammonium perborate and precipitation with alcohol at a low temperature ; it may also be obtained with one mole- cule of water of crystallization. Potassium perborates may be prepared similarly to the sodium salts, but their constitution appears to be somewhat doubtful at present. The following salts have been described : PERBORATES 63 KB 2 O 5 ,2H 2 O ; 6 2KBO 3 ,KBO 4 ,5H 2 O, or perhaps 3KBO 3 ,H 2 O 2 , 4 H 2 O ; 13 2KBO 3 ,H 2 O and aKBOg.H^O-j. 18 Rubidium perborate, RbBO 3 ,H 2 O, and caesium perborate, CsB0 3 ,H,0, have also been prepared. 13 Barium perborate Ba(BO 3 ) 2 ,7H 2 O, is a sparingly soluble salt formed by double decomposition of barium chloride and sodium perborate. 2 The calcium, stron- tium, copper, nickel and cobalt salts are also obtained under similar conditions, but they are unstable, rapidly decomposing with evolution of oxygen. Uranyl perborate (UO)BO 3 , is a yellow, stable compound obtained by the action of a perborate solution on uranium dioxide. 6 In addition to the perborates above mentioned, potassium and ammonium fluoroperborates of complex composition have been obtained by the action of hydrogen peroxide on fluoro- borates. Their aqueous solutions gradually decompose, evolving oxygen. 17 * 19 Free perboric acid has not been isolated, and from extrac- tion experiments with ether, Pissarjewsky u has come to the conclusion that it does not exist in aqueous solution (at 25) even in the presence of excess of hydrogen peroxide ; it does exist, however, to some extent in ethereal solution. Estimation of Perborates. The usual method employed to determine the amount of active oxygen present in a sample of perborate is direct titration with potassium permanganate in acid solution. Farrar 20 re- commends adding about 0*1 gram of the sample to an acid solu- tion of 2 grams of ferrous ammonium sulphate and then titrating the excess of ferrous iron with titanous chloride, using potassium thiocyanate as indicator. Or the perborate may be added to a ; definite volume of titanous chloride, the excess of which is then determined in the usual manner. 21 Manufacture and Use of Perborates. Sodium perborate is best manufactured by mixing the necessary materials (borax, hydrogen peroxide, etc.) in small batches, which are then united and slowly cooled, with stirring, the concentration being so chosen that, after mixing, the liquor is at about 25 and crystallization of the per- 64 PER- AC IDS AND THEIR SALTS borate begins only after some time ; crystals are thus obtained which are comparatively large and stable. The introduction of catalytic metallic compounds into the liquor is avoided by using wooden vats and stirrers, and all metal taps, pipes, etc., are well tinned. The product is best separated from the mother liquor centrifugal ly, and drying is effected by passing the material though an apparatus in which it meets a current of air, which may be safely heated to 5o. 22 There are numerous patents for making the sodium and other salts. 23 Sodium perborate is used industrially for bleaching purposes. It possesses the advantage that it can be incorporated with soaps and washing powders, and several preparations of this kind are on the market, as, for example, "persil," which is a mixture of soap, soda ash, sodium silicate and sodium perborate ; " clarax," con- sisting of borax, sodium phosphate and sodium perborate ; " ozonite," which is similar to persil in composition. These pre- parations are sold under the general name of " perborin pro- ducts," " perborin " itself being sodium perborate, and " perborin M " a mixture of soap, alkali and sodium perborate, which can be used for laundry purposes as a combined washing and bleaching agent. 24 ADDENDUM. Since the above chapter was printed, Bosshard & Zwicky (Zeitsch. angew. Chem. 1912, 25, 993) have published a paper on the constitution of the perborates. They find that when sodium or potassium perborate is gently dis- tilled under diminished pressure at 5o-6o, or in a current of dry, carbon di- oxide-free air at 53-6o, no hydrogen peroxide is found in the distillate ; water is lost, and the percentage of active oxygen in the residue increases. From these and other experiments they come to the conclusion that in the perborates the active oxygen cannot be present as hydrogen peroxide of crystallization, / and that sodium perborate, for example, must be represented as NaO.B/ | \0 and not as NaO.O.BO. This constitution is supported by the fact that they have obtained a stable compound having the composition KBO 4 ,H 2 O, which evolves only traces of oxygen on solution in water. It may be considered to be a salt of potassium hydroperoxide with perboric acid, and, since it is / stable, its constitution may be represented as KO.O.B/ | ; it is very pro- \0 bable that if it had the constitution KO.O.OBO, the chain of three oxygen atoms would make it very unstable. As mentioned on page 62, Melikoff and Pissarjewsky 2 have obtained indications of the existence of a corres- ponding sodium salt, which, however, is very unstable. CHAPTER IV. PERCARBONATES. The attempt to prepare percarbonates by electrolytic methods was a natural corollary to the preparation of the persulphates. Acting on the assumption that the dissociation of the normal alkali carbonates would take place in stages, Constam and Hansen l electrolysed their concentrated solutions in order to determine whether percarbonates were formed at the anode, in accordance with the equation : 2 MCO' 3 +2= M 2 C 2 6 where M represents an alkali metal. The carbonates ot sodium and ammonium were found to possess too small a solubility at the low temperatures necessary, success being attained only when potassium and rubidium carbonates were used. Using an apparatus similar to that described for the preparation of ammonium persulphate (p. 23), the anolyte and catholyte consisting of saturated solutions of potassium carbonate, it was found that with con- tinuously diminishing temperatures the anodic evolution of oxygen became less and less, until at - 10 it practically ceased, and at the same time a bluish amorphous precipitate of potassium percarbonate formed. The progress of the formation of percarbonate could be followed by decomposing an aliquot portion of the electrolyte with dilute, sulphuric acid, whereby hydrogen peroxide is liberated (vide p. 67) in equivalent quantity to the percarbonate present, and subsequent titration with permanganate. It was found 2 that if the electrolysis is commenced at - 15, using a very concentrated electrolyte, small variations in the temperature are of little effect ; the temperature may even rise to o without the yield being seriously affected. If, however, the concentra- tion of the potassium carbonate solution falls below D = 1*52, rise in temperature diminishes the yield. It follows that the temperature should be gradually lowered as the electrolysis proceeds, owing to the diminution in the concentration of the carbonate. In order to obtain precipitation of the solid percarbonate, which is very soluble in water, it is necessary to keep the anolyte practically saturated with respect to 65 5 66 PER- ACIDS AND THEIR SALTS potassium carbonate, a condition which is best maintained by running continuously a concentrated solution of potassium carbonate into the bottom of the anolyte through a funnel ; being heavier than the electro- lysed anolyte the latter is displaced and overflows, carrying the suspended percarbonate with it on to a vacuum filter. At- 10 a potassium carbonate solution, D = 1*56, gave a yield of 25-55 per cent of potassium percarbonate, using a current density of 0*5 to 2 amperes per sq. dcm. ; when, however, the current density rose to 30-60 amperes per sq. dcm., the yield be- came 85-95 per cent. Platinizing the anode diminishes the yield of percarbonate very considerably, 3 possibly owing to the increase in surface, whereby the current density is decreased ; the potential of a smooth anode is also greater than that of a platinized one. Excess of OH'-ions, or the presence of bicarbonate also diminishes I the yield. 3 Generally there is a pronounced odour of ozone at the commencement of the electrolysis, but this disappears after a short time. Iron, nickel, copper or silver cannot be used as anodes. 2 The potassium percarbonate thus prepared is obtained as a sky-blue, extremely hygroscopic powder, which is best dried by spreading on a porous plate and exposure to a current of dry air, which, towards the end of the operation, may be warmed to 40. The dry salt so obtained is slightly blue in colour and contains 87-93 P er cent f percarbonate. It cannot be re- crystallized from water, but may be purified from carbonate by digestion with excess of a concentrated solution of potassium hydroxide at - 5 to - 10 for some time; after collecting the percarbonate it is washed with alcohol to remove potassium hydroxide, whereby a 95-99 per cent product is obtained. 2 Riesenfeld and Reinhold, 4 by the electrolysis at - 30 to - 40 of a concentrated solution of potassium carbonate containing 82-5 grams of K 2 CO 3 ,2H 2 O, to 100 grams of water, claim to have prepared an anhydrous, 100 per cent, product. Rubidium percarbonate was obtained similarly to the pot- assium salt as a white, extremely hygroscopic powder. 1 Properties of Potassium Percarbonate, On warming gently, potassium percarbonate decomposes ac- | cording to the equation : 2K 2 C 2 O 6 = 2K 2 CO 3 + 2CO 2 + O 2 , but for decomposition to be complete and quantitative in a short PERCARBONATES 67 time it is necessary to heat at 2OO-3OO. In ice-cold water it dissolves almost without decomposition, but at ordinary tem- peratures, and especially at 45-5o, it decomposes with the evolution of oxygen, which really proceeds from the decomposi- tion of the hydrogen peroxide formed at an intermediate stage : K 2 C 2 O 6 + 2^0 = 2KHCO 3 + H 2 O 2 . The rate of decomposition of the aqueous solution at 25, as determined by measurement of the gas evolved, and also by titration methods, gives good constants in accordance with a unimolecular equation. 5 The constants vary with the dilution, however, the ratio of the velocity constant to the initial concen- tration of the solution being approximately constant. When, however, the order of reaction is calculated by van't Hoff s method, 6 it is found to be bimolecular. The exact chemical process which takes place is difficult to define, potassium per- carbonate being in this respect similar to ammonium nitrite. 7 With dilute potassium hydroxide, even at - 2, potassium per- carbonate decomposes into potassium carbonate and hydrogen peroxide, the latter then rapidly decomposing in the alkaline solution, with evolution of oxygen : K 2 C 2 O 6 + 2KOH = alCjCOj, + H 2 O 2 . Dilute sulphuric acid, in the cold, gives a quantitative liberation of hydrogen peroxide : K 2 C 2 O 6 + H 2 SO 4 = K 2 SO 4 + 2CO 2 + H 2 O 2 . As an oxidizing agent, potassium percarbonate oxidizes lead sulphide to lead sulphate, decolorizes indigo, and bleaches cotton, silk and wool. Towards manganese dioxide, lead dioxide, and silver oxide, it acts as a reducing agent, the carbonates of the respective metals being formed and oxygen evolved. A very important reaction, as will be seen in the sequel, is that with a solution of potassium iodide. When the solid per- carbonate is added to a strong, neutral solution of 'potassium iodide, there is an immediate, quantitative liberation of iodine, 4 showing that hydrogen peroxide, which liberates iodine only very slowly from potassium iodide, has not been formed. If, however, a solution of the percarbonate is used instead of the solid salt, the amount pf iodine liberated diminishes with the length of time the percarbonate solution has been kept before being added to the solution of potassium iodide, owing to its 5* 68 PER- ACIDS AND THEIR SALTS gradual decomposition into potassium hydrogen carbonate and hydrogen peroxide. The reaction of the solid percarbonate is not affected to any considerable extent in the presence of hydrogen peroxide, even when percarbonate and peroxide are present in molecular proportions. Also the addition of sodium carbonate in the proportion of 0-5, I, i'5 and 2 mols. Na 2 CO 3 to one mol. of percarbonate has not much effect, the iodine liberated varying from 67-9 to 88'O per cent, calculated on the active oxygen present in the percarbonate. 8 The estimation of percarbonates is best carried out by add- ing a known weight of the solid salt to a dilute solution of sulphuric acid, and titrating the hydrogen peroxide liberated with potassium permanganate. 2 - 9 Oxidation of ferrous sulphate and titration of the excess of ferrous salt with permanganate does not give satisfactory results. 1 ' 2 The constitution follows from the method of preparation, two KCO 3 -residues coupling together to form potassium percarbonate, in accordance with the scheme 10 : /o /o o\ 2 O:C( = O:C( >C:O \OK \OK KO/ Free percarbonic acid has not been isolated, although it is possible that it exists in ethereal solution. According to Bach, 10 if, at low temperatures, 0-5 gram of solid phosphoric acid, etheliy'and a few drops of water are added to two grams of potassium percarbonate, there is a violent reaction. On pouring off the ether and adding it to alcoholic potassium hydroxide a bluish- white precipitate of potassium percarbonate forms, thus showing the presence of percarbonic acid in the ether. Wolffenstein and Peltner 10 state, however, that they could not con- firm Bach's observation. In 1899, Tanatar 11 found that when sodium carbonate is dissolved in cold, 3 per cent, hydrogen peroxide and the solution precipitated with alcohol, a white precipitate is obtained which, after collecting, washing with alcohol and ether, and drying over sulphuric acid, has a composition corresponding with the formula Na 2 CO 4 , i^H 2 O. The aqueous solution of this substance possesses oxidizing properties. An alternative formula for this compound would be Na 2 CO 3 ,H 2 O 2 ,^H 2 O, which would represent it as a carbonate with hydrogen peroxide of crystallization, that is, similar to salts of other acids containing hydrogen peroxide of crystallization. 12 Against this latter view, Tanatar evidences PERCARBONATES 69 the fact that the heat of decomposition of a tenth-normal solution by tenth-normal nitric acid is 7'2i-7-26 Cals., whereas that of an equivalent solution of sodium carbonate is 5-33 Cals.; and also that the measurement of the partition coefficient of hydrogen peroxide between ether and an aqueous solution of sodium carbonate shows Khat the carbonate binds some 30 per cent of the peroxide. 13 These facts, however, do not necessarily prove more than the formation of a double compound between the carbonate and the hydrogen peroxide, as distinguished from a true percarbonate. Ether will not extract hydrogen peroxide from potassium percarbonate, whereas it will do so from Tanatar's salt ; moreover, when the latter is added to a neutral solution of potassium iodide there is practically no liberation of iodine, 8 whereas if the one active oxygen atom were present as the per- carbonate, Na 2 CO 4 , there should be an immediate quantitative liberation of iodine, as has been shown to be the case with potas- sium percarbonate. The use of a neutral solution of potassium iodide to distinguish between true percarbonates and carbonates containing hydrogen peroxide of crystallization has been a subject of much controversy, 14 but the balance of evidence seems to be in favour of its use for such a purpose, and further support may be found in the fact that persulphates and sulphates containing hydrogen peroxide of crystallization may be similarly differentiated. 8 We may therefore conclude that Tanatar's salt is not a true percarbonate, a conclusion which holds also for various compounds prepared from potassium, rubidium and ammonium carbonates and hydro- gen peroxide, and which have hitherto been classed as percar- bonates. 15 In 1908, Wolffenstein and Peltner 16 obtained a new series of percarbonates by the action of carbon dioxide on sodium peroxide or on sodium hydroperoxide, the reactions taking place according to the equations : + CO 2 = Na 2 CO 4 ; Na 2 O, + 2CCX = g. 2NaO 2 H + CO 2 = Na2CO 5 ,H. 2 O ; NaO.OH + CO 2 = NaHCO 4 . Examination of the properties of the two latter compounds, especially with respect to their action on a neutral solution of potassium iodide, has shown, however, that they are to be considered as hydrogen peroxide addition products of the first 70 PER- ACIDS AND THEIR SALTS two compounds, that is, as having the formulae Na 2 CO 4 ,H 2 O 2 and Na 2 C 2 O 6 ,H 2 O 2 , respectively. 17 The compound Na 2 C 2 O 6 is best prepared as follows 17 : 2-3 grams of sodium peroxide are slowly added to 100 c.c. of ice-cooled absolute alcohol, whereby sodium hydroperoxide is obtained, according to the equation : * C 2 H 5 OH + Na-A = C,H 5 Na + NaO.OH. Dry carbon dioxide is then passed into the reaction mixture, which is well stirred meanwhile, the temperature being maintained between o and 5. The absorption of carbon dioxide is complete after five to six hours, the end of the reaction being denoted by the fact that the solution no longer reacts alkaline to phenolphthalein. The mass of glistening miscroscopic crystals which has then formed is collected, washed with absolute alcohol and ether, and dried in the air for a short time on a porous plate. The analysis of this compound was made by determining the active oxygen with potassium permanganate, and the sodium and carbon dioxide by titration with hydrochloric acid, using phenolphthalein and methyl orange as indicators. For the formula Na 2 C 2 O 6 the ratio Na : CO 2 : active O should be I : I : 0*5, the ratio determined experimentally being I : 0.93 10.49. Actu- ally the salt so obtained contains half a molecule of alcohol of crystallization, so that its formula is Na 2 C 2 O 6 ,^C 2 H 6 O. When moist sodium peroxide is used to absorb the carbon dioxide, instead of the above reaction mixture, crystals are obtained having the composition Na2C2O 6 ,H 2 O. 16 If, in the above preparation, 5 c.c. of 30 per cent hydrogen peroxide (perhydrol) are (i) added to the reaction mixture before carbon dioxide is passed in, or (2) added after the absorption of * Tafel 18 claims that the compound prepared by this reaction, and which he terms " sodyl hydroxide," has the constitution O:Na.OH, and Wolffenstein and Peltner 16 claim to have obtained a different product by the action of carbon dioxide on this compound, from that resulting from the interaction of carbon dioxide and the sodium hydroperoxide prepared from sodium ethoxide and hydrogen peroxide : Riesenfeld and Mau's results, 8 however, which have since been confirmed by d'Ans and Friederich, 19 show that the only difference between these sodium hydroperoxides is that the compound prepared by the second method contains half a molecule of hydrogen peroxide of crystallization ; they show the same behaviour chemically, and there is no valid evidence for the existence of isomerides. PERCARBONATES 7t carbon dioxide is complete, stirring being continued for another ten minutes, crystals are obtained having the composition Na 2 C 2 O 6 ,H 2 O 2 . That this is the correct formula, and not 2NaHCO 4 , is shown by the reaction with a neutral solution of potassium iodide. The active oxygen in the salt Na 2 C 2 O 6 liberates 50 per cent of the theoretical amount of iodine from neutral potassium iodide solution, the remaining active oxygen escaping as gas. The iodine liberated by the compound Na. 2 C 2 O 6 ,H 2 O 2 , corresponds with 25 percent of its active oxygen, since hydrogen peroxide has no action on the neutral potassium iodide, whereas if the formula were NaHCO 4 , the iodine liberated should correspond witti 50 per cent of its active oxygen. The compound Na 2 CO 4 ,iH 2 O, is prepared as follows * 16 : In a dish cooled in a freezing mixture, 38 grams (0*5 mol.) of sodium peroxide are hydrated by trituration with 50 grams of ice, after which carbon dioxide is led in, with continuous stirring. In proportion as water is split off and the mass becomes moist, a further quantity of 38 grams of sodium peroxide is added in small portions at a time. The passage of carbon dioxide is stopped when the increase in weight is 44 grams, or better still, 17 when a sample of the substance requires twice as much hydrochloric acid for neutralization with methyl orange as indicator as with phenolphthalein as indicator. The product is washed with alcohol and ether and dried in a vacuum. It loses oxygen slowly in a desiccator, being completely decomposed in a few months ; the decomposition takes place much more rapidly when the salt is moist. It can be analysed by a method similar to that described for the com- pound Na 2 C 2 O 6 . The compound Na 2 CO 4 ,H 2 O 2 is prepared in a similar manner by hydrating 1875 grams of freshly prepared sodium hydroperoxide f with 5 -2 grams of water and passing in carbon dioxide till the increase in weight is 6 -8 grams. That it should not be formulated as Na 2 CO 5 ,H 2 O is shown by comparing its action with that of the compound Na 2 CO 4 on a neutral solution of potassium iodide, the results being analogous to those described for the compounds Na 2 C 2 O 6 and Na 2 C 2 O 6 ,H 2 O 2 . An unstable compound, BaCO 4 ,xH 2 O, has also been obtained by the action of carbon dioxide on hydrated barium peroxide. 21 It is not probable that it has a composition represented by * This method of preparation forms the basis of various German patents. 20 t Made by the interaction of I mol. of sodium ethoxide with half a mol. of hydrogen peroxide, and washing the resultant precipitate with alcohol. 72 PER-ACIDS AND THEIR SALTS BaCO 3 ,H 2 O 2 , since neither alcohol nor ether will extract hydrogen peroxide from it. It will have been noticed that the compounds Na 2 C 2 O 6 and K 2 C 2 O 6 react differently towards a neutral solution of potassium iodide in so far as the amount of iodine liberated is concerned. It is, therefore, probable that they are different in constitution, and the probability has been strengthened by the preparation of a compound, having the composition K 2 C 2 O 6 , from potassium peroxide, absolute alcohol, and' carbon dioxide, by a method similar to that described for the compound Na 2 C 2 O 6 . 17 In its reaction towards neutral potassium iodide it behaves similarly to Na 2 C 2 O 6 and differently from ordinary potassium percarbonate, this difference in behaviour being maintained when solutions of the two compounds are compared, instead of the solids being added directly to the solution of potassium iodide. It has already been shown that the constitution of ordinary potassium percarbonate is probably /O - o O : C( ^C : O. \OK KO/ This would leave the formula : \O.OK KO/ for the isomeric salt just described, and the constitution of the compound Na 2 CO 4 would be represented probably by the for- mula /O.ONa 0:C( \ONa in analogy to that of permonosulphuric acid, /O.OH 2 S( \OH Riesenfeld and Mau 17 suggest that salts derived from the acid H 2 CO 4 should" be termed monoperoxy carbonates, those from the acids H 2 C 2 O 6 being monoperoxydicarbonates. In analogy to the nomenclature of the persulphuric acids it would be better to term them permonocarbonates and perdicarbonates respectively. Ordinary potassium percarbonate is used in photographic PERCARBONATES 73 practice under the name " Antihypo" for removing the last traces of sodium thiosulphate. It has also been recommended for bleaching purposes. Sodium permonocarbonate is used for medical and technical purposes, especially as a disinfectant and in the preparation of hydrogen peroxide. . "V f CHAPTER V. PERNITRIC ACID AND PERPHOSPHORIC ACID. Pernitric Acid. THE evidence for the existence of pernitric acid and pernitrates is, at present, somewhat unsatisfactory. An unstable oxide of nitrogen, possibly having the composition NO 3 or N 2 O 6 , and which would correspond with some higher acid of nitrogen, has been described by Hautefeuille and Chappuis l and by Berthe- lot. 2 The former investigators obtained it by the action of a silent electric discharge on a mixture of nitrogen and oxygen, and also state that it is mixed with the ozone produced by the action of a silent electrical discharge on a mixture of nitrogen and nitrogen dioxide. The substance produced by either of these methods has a remarkable absorption spectrum characterized by fine, very dark lines in the red, orange and green ; it is de- composed by water. By the action of nitrogen pentoxide on well-cooled, anhydrous hydrogen peroxide, d'Ans and Friederich 3 have obtained a colourless liquid possessing a penetrating odour similar to that of bleaching powder. The aqueous solution of this liquid has the specific properties of a per-acid similar in character to per- monosulphuric acid, liberating iodine and bromine from potassium iodide and bromide respectively, and oxidizing aniline to nitroso- benzene, so that it probably contains pernitric acid, formed pos- sibly according to the equation : * N a 5 + H 2 2 = HN0 3 + 2 N.0 2 H. It is also possible that a pernitric acid similar to persulphuric acid is formed, since in addition to nitrosobenzene, an orange brown oxidation product is formed from aniline in neutral solu- tion, and a black product in sulphuric acid solution (cf. persul- phuric acid). In dilute aqueous solution, hydrolysis to nitric acid and hydrogen peroxide takes place rapidly. * The concentrated substance is explosive, and its properties and com- position have not been fully investigated as yet. 74 PERNITRIC ACID AND PERPHOSPHORIC ACID 75 Raschig 4 states that pernitric acid is formed in very dilute aqueous solution by the oxidation of nitrous acid with 3 per cent hydrogen peroxide, according to the equation : HNO 2 + 2H 2 O 2 = HNO 4 + 2H,O. It cannot be obtained when nitric acid is used, indeed, the dilute solution slowly decomposes thus : HN0 4 + H 2 = HN0 3 + H 2 2 . The solution possesses the characteristic property of liberating bromine from solutions of potassium bromide, whereas neither hydrogen peroxide, nitrous acid, nor nitric acid is capable of doing this. Schmidlin and Massini 5 find that the solution does not oxidize manganous salts to permanganates (cf. perphos- phoric acid, p. 77) ; they consider that Raschig's compound is a hydroperoxide ester of nitrous acid, having the constitution O:NO.OH,* and formed according to the reaction : O:N.OH + HO.OH = OrN.O.OH + H 2 O. Schellhaas 6 has found that in the electrolysis of pure nitric acid some oxygen is fixed at a platinum anode, possibly owing to the formation of a pernitric acid in accordance with the equations : 2 NO' 3 + 2 = N 2 6 ; N 2 6 + H 2 O = HNO 4 + HNO 3 . When a 5-25 percent solution of silver nitrate is electrolysed at o in a divided cell, using a current density of 1*5 to 14 amperes per sq. dcm., black octahedral crystals are deposited at the platinum anode, which have a composition corresponding with the formula Ag 7 NO u . They decompose slowly at ordinary temperatures, and on heating in a current of dry air at 52- 185, 5 atoms of oxygen are lost. The elimination of oxygen apparently takes place in at least two stages, 3 atoms being lost at 52-6o, and the remainder at a much higher temperature. The behaviour towards heat of these black crystals, their action as an oxidizing agent, and the fact that their composition always corresponds with the formula Ag 7 NO n , has led to their formulation as a double compound of silver dioxide and a silver pernitrate, 3Ag 2 O 2 , AgNO 5 , and the name silver peroxynitrate has been given to them. An alternative formula is 2Ag 3 O 4 ,AgNO 3 , and there has been much discussion as' to which of these is correct 7 * It is difficult to see in what way this is different from a per-acid, since permonosulphuric acid may also be considered as a hydroperoxide ester of sulphuric acid. 76 PER- ACIDS AND THEIR SALTS However, similar compounds have been prepared by the elec- trolysis of solutions of silver sulphate * 8 and silver fluoride, 9 and Bose 10 has found that the anodic decomposition curves of solutions of silver nitrate and silver sulphate show a marked break at 1-57 and 1-53 volts respectively, at which potential the anode becomes covered with a coating of the higher oxide of silver. Since the break occurs at practically the same point in solu- tions of the nitrate and sulphate, it follows that the same compound is deposited in both cases ; this is a higher oxide of silver, and the compounds Ag 7 NO n and Ag 14 SO 17 are double compounds of this oxide with silver nitrate and silver sulphate respectively. Further support is found for this in the experiments of Luther and Pokorny, 11 who have shown that the so-called silver peroxynitrate, silver peroxysulphate, and a higher oxide of silver prepared electrolytically, all give the same electrolytic reduction curve ; also, the potential of these compounds is not altered by treatment with sodium hydroxide, whereas if, for example, a salt of a silver peroxynitrate were present, together with higher oxides of silver, the action of sodium hydroxide should give a soluble sodium salt, leaving the higher oxides of silver, together with silver oxide, on the electrode, whereby the potential should alter. Finally, Baborovsky and Kuzma 12 have investigated not only the composition of the anodic deposit, but also the changes in concentration which take place in the anolyte of a divided cell, when solutions of silver nitrate or silver sulphate are electro- lysed. These changes in concentration can be calculated accord- ing to the processes assumed to take place at the anode, and experiment and observation are in agreement only when the anode deposit is taken to be 2Ag 3 O 4 ,AgNO 3 . It may therefore be considered that, the existence of a per- nitrate of silver, AgNO 5 , has not yet been proved. Pifierua-Alvarez 15 claims to have prepared a potassium pernitrate, KNO 4 , by the action of sodium peroxide on an aqueous- alcoholic solution of potassium nitrate, but no analyses are given and no proof that the compound produced is not a mixture of potassium nitrate with a peroxide of potassium or sodium. * The deposit formed in this case has the composition Ag 14 SO l7 and has been called silver peroxysulphate. PERNITRIC ACID AND PERPHOSPHORIC ACID 77 Perphosphoric Acid. Electrolysis of solutions of phosphoric acid and its salts does not give rise to perphosphoric acid, nor does orthophosphoric acid react with hydrogen peroxide. When, however, phosphoric oxide, or meta- or pyro-phosphoric acid is treated at low tempera- tures with 30 per cent, or stronger, hydrogen peroxide, and the reaction mixture afterwards diluted with ice-water, a solution is obtained which possesses oxidizing properties similar to those of permonosulphuric acid. 1 ' 2 It will even oxidize manga- nous salts to permanganic acid in the cold, a reaction which is not shown by Caro's acid. The analysis of the per-acid thus obtained was carried out by determining the maximum amount of oxygen with which a molecule of phosphoric acid could combine ; the excess of hydrogen peroxide was determined by titration with potassium permanganate and the amount of per-acid formed by measuring the iodine liberated immediately from potassium iodide. 1 The results point to the formula H 3 PO 5 , or PCXOH^O.OH, the permonophosphoric acid thus being derived from orthophos- phoric acid, no matter whether phosphoric oxide, meta- or pyro- phosphoric acid be used in the preparation. No salts could be obtained, but neutral solutions give precipitates with salts of the heavy metals, which rapidly change to the ordinary phosphates with evolution of ozonized oxygen. The neutral solution remains neutral after boiling to decompose the perphosphoric acid (cf. permonosulphuric acid, p. 50). From hydrogen peroxide and excess of pyrophosphoric acid a perphosphoric acid, probably having the composition H 4 P.,O 8 , is formed in small quantity, in addition to permonophosphoric acid. Electrolysis of a dilute solution of phosphoric acid containing hydrogen peroxide gives rise to permonophosphoric acid at the anode, this oxidation probably depending on the formation of perhydroxylanions, HO.O'-. Petrenko 3 has described a compound, PO(ONa)(O 2 Na) 2 ,6H 2 O, made by the action of hydrogen peroxide on a solution of normal sodium phosphate. It is readily decomposed by water, with the formation of hydrogen peroxide, and is probably better charac- 78 PER-ACIDS AND THEIR SALTS terized as sodium phosphate containing hydrogen peroxide of crystallization, namely PO(ONa) 3 ,2H 2 O 2 ,4|H 2 O. Pifierua-Alvarez 4 claims to have prepared a sodium per- phosphate, NaPO 4 , by the action of an aqueous-alcoholic solution of sodium peroxide on ordinary sodium phosphate. No analyses are given. A similar claim is made for the preparation of a sodium perarsenate, NaAsO 4 . Somewhat similar compounds of bismuth have been described by Hanus and Kallauner, 5 who do not, however, claim that they are per-salts. CHAPTER VI. PERTITANATES, PERZIRCONATES, PERSTANNATES. Pertitanic Acid and the Pertitanates. IN 1873 Schone 1 observed that the addition of hydrogen per- oxide to neutral or acid solutions of titanium salts produces a yellow coloration. The production of this yellow colour, which forms a very sensitive test for either titanium salts or hydrogen peroxide, 2 is due to the formation of a peroxide of titanium having the composition TiO 3 ,aq., 3 which is best isolated as follows 4 : Titanium chloride is added drop by drop to dilute alcohol, and the clear and very dilute solution is treated with a large excess of hydrogen peroxide. Ammonia, ammonium carbonate, or aqueous potassium hydroxide, is then added to the solution, with the production of a yellow, or in case of ammonia, of a reddish-yellow liquid, which after some time yields a yellow precipitate. This is repeatedly washed by decantation, collected, and dried on a porous plate, when a product is obtained approximating to the composition TiO 3 ,3H 2 O. The exact composition of the solid is, however, somewhat doubtful, since it has been variously stated by different investi- gators. The want of agreement is probably due to the fact that the action of hydrogen peroxide is very slow, and the precipitate decomposes while being dried. Levy 5 endeavoured to deter- mine the composition of the particular oxide formed, by the method of approximation ; a solution of hydrogen peroxide of known strength was mixed with definite quantities of a solution of titanic oxide in sulphuric acid. After a long time (some 10 days or so) the mixtures were examined in order to ascertain which of them still contained hydrogen peroxide. It was found that the formula lay between TiO 2 . 98 and TiO^, that is, in all pro- bability the formula is TiO 3 , the reaction therefore taking place according to the equation : TiO 2 ,aq. + H 2 O 2 = TiO 3 ,aq. + H 2 O. 79 80 PER- ACIDS AND THEIR SALTS It is, of course, possible that the hydrogen peroxide may be simply added on to the titanium dioxide, with the formation of an additive compound, TiO 2 ,H 2 O 2 , but the existence of the fluoroxypertitanates (sea later) is against this view, as also the fact that the solution does not show the presence of hydrogen peroxide when tested with chromic acid. Mazzucchelli and Barbero 6 state that the potential of the solution is the same as that of hydrogen peroxide, which would indicate the existence of the latter compound in the solution. Since, however, the reaction between titanium dioxide and hydrogen peroxide is a very slow one, it is possible that the re- action was not complete before the potential was measured. The action of the solution on a photographic plate is said to be different from that of a solution of hydrogen peroxide. 7 The solution is a strong oxidizing agent, behaving similarly to one of hydrogen peroxide ; the velocity of reaction is, how- ever, somewhat slower. Higher oxides, such as the dioxides of lead and manganese, are reduced by the solution. It differs from a solution of hydrogen peroxide in that it gives a red pre- cipitate with potassium ferrocyanide, a white precipitate with acetaldehyde, and a yellow precipitate with acetcuie 6 ; also chromic acid is not oxidized to perchromic acid. The solid substance, which generally forms a brownish-yellow, horny mass, is decomposed on heating, or on boiling with water, oxygen being evolved. Hydrochloric acid dissolves it with evolution of chlorine, and it is decomposed by sulphuric acid. Dilute sulphuric acid dissolves it, giving a reddish-yellow solution, which decomposes with evolution of oxygen on being concentrated. The aqueous solution may be considered to contain a per- titanic acid, although it does not show any acid reaction towards indicators ; the freshly precipitated solid is soluble in alkalis, giving a solution which yields precipitates with alkaline solutions of copper, zinc, cobalt and lead salts. Hitherto no salts have been prepared which are derived from a pertitanic acid by simple replacement of hydrogen by a metal. Melikoff and Pissarjewsky, 8 however, have been able to prepare a number of compounds which they consider to be salts formed from pertitanic acid and a metallic hydroperoxide (cf. p. 4). The general method of preparation is to add well-cooled hydrogen peroxide to the compound TiO 3 ,aq., and * then sufficient well-cooled solution PERTITANATES, PERZIRCONATES, PERSTANNATES 8 1 of the alkali to give complete solution ; the addition of alcohol then gives a precipitate which, after washing with alcohol and ether, is dried on a porous plate. The following compounds were thus obtained : Nao0 2 ,Ti0 3 , 3 H 2 0; (NH 4 ) 2 O 2 ,TiO 3 ,H 2 O 2 ; BaO 2 ,TiO 3 , 5 H 2 O ; the sodium and barium salts being yellow powders of indefinite crystalline structure, the ammonium salt forming slender yellow prisms. On solution in dilute acids, hydrogen peroxide is formed, in contradistinction to pertitanic acid, which gives no hydrogen peroxide under like conditions. Another series of compounds has been obtained, the members of which form colourless crystals. They are prepared similarly to the above, except that a larger excess of hydrogen peroxide and alkali is used, the latter being added until the yellow solution first formed becomes colourless. The compounds prepared have the formulae : K 2 O 2 ,TiO 3 ,ioH 2 O and (Na2O 2 ) 4 Ti 2 O 7 ,ioH 2 O. On solution in water they both decompose with evolution of oxygen, and the latter salt is characterized by the fact that it does not liberate iodine from potassium iodide. The first com- pound deliquesces on exposure to the air, at the same time losing oxygen and becoming yellow in colour, while the second under- goes no change during several hours' exposure to the air. The constitution of these compounds is formulated by Melikoff and Pissarjewsky as follows 9 : Pertitanic acid, TiO 3 ,aq., can be assumed to have the constitution : O, since the solution possesses all the properties of a peroxide. The above compounds are the salts of this acid and the basic metallic hydroperoxides. The compound (Na 2 O 2 )TiO 3 ,3H 2 O, will then be : * The constitution may also be /O.OH Ti(T. , i.e. TiO 3 ,2H 2 O. \\\(OH) 3 In either case titanium is tetravalent. According to its position in the periodic system it should not assume a higher valency. ' 6 82 PER-ACIDS AND THEIR SALTS NaO.CK /O >Ti( |, 3 H 2 0; NaO/ \O /O (NHJaO^TiOg.HaOg, or (NH 4 O 2 ) 2 TiO 2 ,H 2 O is (NH 4 O.O) 2 Ti( | ,H 2 O. \O The salt K 2 O 4 ,K 2 O 2 ,TiO 3 , ioH 2 O, is considered to be KO.CK /o >Ti( | ,K 2 4) ioH 2 0, KO/ \O its behaviour on exposure to the air being exactly what one would expect if potassium tetroxide were present. In the compound (Na 2 O 2 ) 4 Ti 2 O 7 , ioH 2 O, Melikoff and Pissar- jewsky are inclined to assume a higher valency than four for titanium, although this would not be in accordance with its position in the periodic system. If, however, pertitanic acid is represented as (HO) 3 TiO.OH, that is, the acid is TiO 3 ,2H 2 O, the above salt may be formulated as NaCK \ONa The evidence for these formulae is not quite convincing at present. The various compounds can be represented as double compounds of a pertitanate with hydrogen peroxide, for example, (NaO) 2 TiO 2 ,H 2 O 2 ,2H 2 O. The fact, however, that the one series is colourless, whereas per- titanic acid is coloured yellow, seems to indicate that they are not salts of ordinary pertitanic acid, and is in favour of the formu- lation of Melikoff and Pissarjewsky. Pertitanic acid is characterized by forming a number of com- plex compounds with various acids. For example, when hydro- fluoric acid or sulphuric acid solutions of titanium are oxidized with hydrogen peroxide, and an alkali or alkaline earth fluoride is added, precipitates of the fluoroxypertitanates are obtained, generally in the form of crystals. 10 Such compounds are: Ti0 2 F 2 , 3 NH 4 F; TiO 2 F 2 , 2 KF ; TiO 2 F 2 ,BaF 2 . They are thus derived from pertitanic acid, TiO 3 ,aq., by the re- placement of one atom of oxygen by two atoms of fluorine. The fact that they are decomposed by hydrofluoric acid with the formation of hydrogen peroxide and titanium tetrafluoride, as, for example, Ti0 2 F. + 2HF = TiF 4 + H 2 2> shows that the O 2 contained in them is a divalent group, being PERTITANATES, PERZIRCONATES, PERSTANNATES 83 equivalent only to 2F, and affords further support for the formula- tion of pertitanic acid as For other complex salts the original literature should be con- sulted. 11 Perzirconates. In analogy to titanium, zirconium gives a hydrated peroxide, ZrO 3 ,nH 2 O, by the action of hydrogen peroxide on a neutral or ammoniacal solution of a zirconium salt, 1 or by the action of sodium hypochlorite on the hydroxide. 2 In the presence of sodium or potassium hydroxide, however, no precipitate is obtained, owing to the formation of soluble perzirconates. These perzirconates are best obtained by dissolving freshly precipitated zirconium peroxide in an alkaline solution of hydrogen peroxide and precipitating with alcohol . 3 The sodium and potassium salts have the formulae Na 4 Zr 2 O n ,9H 2 O and K 4 Zr 2 O u ,9H 2 O, but it is doubtful if they have "been obtained quite pure. With dilute sulphuric acid they give hydrogen peroxide, and with concen- trated sulphuric acid ozonized oxygen is evolved. Hydrated peroxides of cerium and thorium have been obtained, 4 but no per-salts corresponding with them. Perstannic Acid and Perstannates. By the action of 30 per cent hydrogen peroxide on stannic acid which had been precipitated from a solution of stannic chloride by sodium carbonate, Tanatar obtained a white, amorphous powder, having a composition corresponding with the formula HSnO 4 ,2H 2 O. It undergoes with water partial decomposition into stannic acid and hydrogen peroxide, and when dried at 100 it is converted into the acid H 2 Sn 2 O 7 ,3H 2 O. 1 The potassium salts corresponding with these perstannic acids were obtained by the action of 30 per cent hydrogen peroxide on a 30 per cent solution of potassium stannate.* The perstannate, KSnO 4 ,2H 2 O, slowly deposits, and more can be obtained by precipitation with alcohol. It is a white amorphous powder, which, on prolonged drying, gives a salt having approximately the composition * Coppadoro 2 states that perstannates are formed at the anode in the electrolysis of concentrated alkaline stannate solutions. 6* 84 PER-ACIDS AND THEIR SALTS K 2 Sn 2 O 7 ,3H 2 O. The aqueous solution is alkaline and contains hydrogen peroxide, which can be extracted with ether ; it gives precipitates with various salts of the heavy metals, which pre- cipitates consist of mixtures of the hydroxides of these metals with their perstannates ; with the exception of the barium and strontium salts, they decompose rapidly, evolving oxygen. The lead salt, when first produced, is pure white, but it grad- ually becomes coloured, owing to the formation of lead dioxide. It is this behaviour of the lead salt which causes Tanatar to con- sider these compounds as perstannic acids and perstannates, rather than molecular compounds, or hydrates in which H 2 O has been replaced by H 2 O 2 . Sodium perstannate, NaSnO 4 ,2H 2 O, is prepared similarly to the potassium salt. CHAPTER VII. PERVANADATES, PERCOLUMBATES, PERTANTALATES. Pervanadic Acid and Pervanadates. In 1 86 1 Werther 1 found that acidified solutions of salts of vanadic acid turn red on the addition of hydrogen peroxide, and that the red colour is not extracted by ether (compare chromium). Scheuer 2 then showed that on dissolving vanadium pentoxide in a solution of hydrogen peroxide containing sulphuric acid, a red solution is formed, from which, on concentration in a vacuum, dirty yellow crystals were obtained, which dissolved in water to a red solution, but could not be analysed. The solution so obtained evolved chlorine and oxygen under the action of hydro- chloric acid. According to Pissarjewsky, 3 vanadium pentoxide dissolves in an aqueous solution of pure hydrogen peroxide, giving a red solution which gradually evolves oxygen and reacts acid towards Congo red. At ordinary temperatures the forma- tion of the red compound is accompanied by vigorous evolution of oxygen, the vanadium pentoxide acting catalytically on the hydrogen peroxide; it is better, therefore, to carry out the reaction at low temperatures. Partition experiments 3 with ether indicate that each atom of vanadium fixes one molecule (0*87, 1*26, 1*12 mols.) of hydro- gen peroxide, so that the reaction is represented by the equation : V,0 5 + 2 H 2 2 = 2 HV0 4 + H 2 0. Titration of the resulting solutions with sodium hydroxide, using Congo ted as indicator,* show that approximately one molecule of sodium hydroxide is used up for each atom of vanadium, but the titrations are very much affected by hydrolysis, not quite the equivalent of sodium hydroxide being used. The pervanadic acid formed is thus derived from metavanadic acid, HVO 3 . * This indicator does not show any action between sodium hydroxide and hydrogen peroxide. 85 86 PER-ACIDS AND THEIR SALTS Pissarjewsky also observed that the addition of hydrogen per- oxide to a solution of vanadic acid gives the red colour of per- vanadic acid, but that the formation of the red colour does not always take place immediately ; in some cases it takes a few minutes, in others some hours ; occasionally, no red colour is produced. The solution of metapervanadic acid gradually decomposes with evolution of oxygen, a yellow solution remaining, which contains hexavanadic acid, H 4 V 6 O 17 , 4 and no hydrogen per- oxide. Metapervanadates corresponding with the above acid were prepared by Scheuer. 2 The potassium salt, KVO 4 , is prepared by dissolving potassium metavanadate in a solution of hydrogen peroxide acidified with sulphuric acid, and is obtained as a micro- cystalline precipitate by the addition of alcohol to the solution ; it is stable when dry. Barium metapervanadate, Ba(VO 4 ) 2 , is ob- tained as an amorphous, yellow precipitate on adding a solution of barium chloride to a saturated solution of ammonium meta- vanadate in 30 per cent hydrogen peroxide. Other salts, all of which are anhydrous, have been prepared by methods similar to the above, or by methods involving double decomposition and precipitation ; some are amorphous, and others microcrystalline salts of a beautiful bright yellow to deep orange colour. The monobasicity of metapervanadic acid has been confirmed by measurements of the equivalent conductivity of solutions of the potassium salt. 5 ^ 1024 - A m = 140*36 - 134*8 = 5*56, and therefore the acid is monobasic, since Ostwald's numbers for potassium salts of monobasic acid vary between 4 and 6*8. The conductivities could not be measured at greater concentrations, ow- ing to the catalytic effect of the electrodes on the decomposition of the hydrogen peroxide present in the solutions. This hydrogen peroxide results from the hydrolysis of the potassium metaper- vanadate, KVO 4 +H 2 O^KVO 3 + H 2 O 2J and amounts to 5'i per cent in N/I5 solutions at 25, as determined by partition experi- ments with ether. 6 The presence of hydrogen peroxide cannot be accounted for on the assumption that potassium metapervana- date is a double compound of potassium metavanadate and hydrogen peroxide, namely KVO 3 ,H 2 O 2 , since the salt is anhy- drous and does not contain hydrogen. The heat developed when one molecule of potassium meta- PERVANADATES, PERCOLUMBATES, PERTANTALATES 87 vanadate reacts with hydrogen peroxide, is shown in the following table : 7 Mols. of H 2 O 3 ..... i 23-5 4 Cals. evolved ..... 9-02 17* 77 r 7'73 17*69 The figures show that besides the compound KVO 4 , in which the ratio of vanadium to active oxygen is I : I, there exists in aqueous solution a more highly oxidized compound in which the ratio is I : 2. This may have the composition KVO 5 , or KVO 4 ,H 2 O 2 . The constitution of the acid HVO 4 may be represented as HO v = o Since it possesses acid properties it should form salts with the metallic hydroperoxides (v. p. 4). Such salts have been pre- pared by Melikoff and Pissarjewsky, 8 but they are derived from pyropervanadic acid. The ammonium salt, for example, is obtained by dissolving ammonium metavanadate in an aqueous solution of hydrogen peroxide, adding ammonia until the solution acquires a distinct ammoniacal odour, and then precipitating the salt with alcohol. The reaction may be represented as follows : Ammonium metapervanadate, NH 4 VO 4 , is first formed, and this then gives the pyro-salt : o o o o o o V NH 4 O\ V V /ONH 4 2 NH 4 O.V = O + H 2 O = )V O V< HO/ \OH Ammonium hydroperoxide, NH 4 O.OH, is formed in solution at the same time (cf. p. 62), and this forms a salt, (NH 4 ) 4 V 2 O n , with the ammonium pyropervanadate : O O O O O O O O NH 4 O\ V V /ONH 4 NH 4 O\ V V /ONH 4 >V O V( = \V O V< +2H 2 H : 0/ \0;H NH 4 0.0/ \O.ONH 4 HO:ONH 4 NH 4 O:OH This salt crystallizes in minute, slender, yellow, rhombic prisms, which remain undecomposed for some time when dry, and then slowly decompose with evolution of oxygen. With dilute sul- phuric acid it yields hydrogen peroxide, and with concentrated sulphuric acid, ozonized oxygen. The aqueous solution smells of ammonia, owing to hydrolysis. It contains four atoms of 88 PER-ACIDS AND THEIR SALTS active oxygen * to two atoms of vanadium, which is in agreement with the above constitutional formula, one active oxygen being connected with each vanadium atom, as in the metapervanadates, the other two active oxygens being in the ammonium hydroper- oxide radicles. A potassium salt, K 8 V 5 O 26 ,2H 2 O, is obtained by treating a saturated solution of potassium metapervanadate with hydrogen peroxide and potassium hydroxide until a yellow solution is formed ; after some time, yellow, rhombic prisms separate, which are similar in their properties to the ammonium salt just described. Melikoff and Pissarjewsky formulate this compound as a double salt, 3K 4 V 2 O 12 ,4KVO 4 ,4H 2 O, the constitution of the component K 4 V 2 12 being o_o o o V V /O.OK (KO.O) 2 =V O \OK Support for this is found in the fact that when treated with hydrogen peroxide and potassium hydroxide at o, an unstable compound is formed having the composition K 4 V 2 O 13 ,7H 2 O, that is, o o o o V V (KO.O) 2 V O V (O.OK) 2 If this reaction is carried out at ordinary temperatures an unstable compound containing one atom less of oxygen, that is, only three KO.O-groups is probably obtained, namely, K 4 V 2 O 12 ,4H 2 O. Measurements of the equivalent conductivity of solutions of the double salt did not lead to any definite results. 6 Orthopervanadates have been obtained by Melikoff and Jehlchaninoff 9 by making use of 30 per cent hydrogen perox- ide (perhydrol). For example, ammonium orthopervanadate, (NH 4 ) 3 VO 6 ,2|H 2 O, is prepared by dissolving ammonium vana- date in excess of concentrated ammonia, cooling to o, adding well-cooled 30 per cent hydrogen peroxide, and precipitating with alcohol. The solution first obtained is blue, and the precipitated product, after washing with ether and drying, is pale blue. There are two active oxygens present to each atom of vanadium, as determined by titration with potassium permanganate, so that the constitution may be represented by : * Determined by decomposition with sulphuric acid or acetic acid and measurement of the evolved oxygen. PERVANADATES, PERCOLUMBAJES, PERTANTALATES 89 NH 4 O.O\ yO \yX | (NH 4 O) 2 9l 15 for the preparation of the compound CrO 4 ,3NH 3 , or chromium tetroxide triammine, the following may be mentioned 14 : Five c.c. of 30 per cent hydrogen peroxide are added drop by drop to a mixture of 25 c.c. of 10 per cent ammonium hydroxide with 5 c.c. of 50 per cent chromic acid, the temperature being maintained at o. The solution is then kept in a freezing mixture for one hour, after which, to- gether with the crystals of the red ammonium perchromate which have deposited, it is heated to 50 until the rapid evolution of gas ceases, and the salt .has almost completely dissolved. The solution is then filtered and cooled to o, when the long, brown needles of the chromium tetroxide triammine separate. The same compound is formed very quickly when the red ammonium perchromate is treated with 10 per cent ammonia at 40, but only slowly at the room temperature, separating either in needles, or as rectangular or rhombic plates. These three forms have the same density at 1 5*8, namely I '964, and when examined crystallographically appear to be identical, the form of the crystals seeming to depend on the concentration of the ammonia ; u it is not probable that they constitute two isomeric compounds, as stated by Hofmann and Hiendlmaier. 8 Chromium tetroxide triammine is soluble in ammonia Jo a brown solution; also in water, but with partial decomposition. When heated, it detonates at comparatively low temperatures. When treated with alkalis, chromates are formed, oxygen and ammonia being evolved ; hydrogen peroxide is said to be an intermediate product. 16 W T hen treated with acids a portion of the oxygen is liberated in the gaseous form, but some of it remains in solution as hydrogen peroxide, chromic salts being formed at the same time ; the amount of hydrogen peroxide formed diminishes with increase in concentration of hydrion. 16 Analysis shows that the ratio of hydrogen to nitrogen is 3:1, so that an ammonium salt cannot be present, 8 which is in agreement with the non-existence of potassium or sodium compounds ; the molecular weight in water by the cryoscopic method agrees with the formula CrO 4 ,4NH 3 . 16 The compound is probably best formulated as a complex chromium ammonia, 96 PER-ACIDS AND THEIR SALTS 1,N . J the peroxide-oxygen group occupying one co-ordination position only, 17 which would account for the formation of hydrogen peroxide on decomposition with acids. Support for this formula is found in the fact that when treated with an aqueous solution of potassium cyanide the three ammonia groups are readily replaced by three potassium cyanides, giving the compound CrO 4 ,3KCN, 7 and also that when reduced with hydrochloric acid in acetic acid solution, dichloroaquotriamminechromium chloride, rCl OH 2 -i Cr Cl, LCI (NH 3 ) 3 J is formed ; 15 Riesenfeld has also been able to prepare a number of complex chromium ammonias from it. 16 Compounds of chromium tetroxide with ethylenediamine and hexamethylenetetramine, namely, CrO 4 ,C 2 H 8 N 2 ,2H 2 O and CrO 4 ,C 6 H 12 N 4 , have been described. 18 From the foregoing it follows that the above four classes of compounds are closely related chemically ; they are formed under analogous conditions from chromic acid and hydrogen peroxide, and are easily transformed one into the other. If an aqueous solution of (NH 4 ) 3 CrO 8 is treated with an acid, oxygen is evolved and the salt (NH 4 )H 2 CrO 7 separates, a blue colour being developed at the same time, since this latter salt is blue. If an excess of pyridine is added to the aqueous paste of (NH 4 ) 3 CrO 8 , or to NH 4 H 2 CrO 7 , the salt C 5 H 5 N,HCrO 5 separates. On treating either of these compounds with excess of ammonia, the most stable of all the oxidation products of chromium is produced, namely CrO 4 ,3NH 3 . These changes can be explained by the addition or elimination of hydrogen peroxide, thus : 18 2CrO 4 + H 2 O 2 = 2HCrO 5 ; 2CrO 4 + 3H 2 O 2 = 2H 3 CrO 7 ; 2CrO 4 + 5H 2 O 2 = 2H 3 8 + 2H 2 O. The constitution of the red and blue perchromates has been investigated by Riesenfeld, 12 and the results of the cryoscopic and electrical conductivity measurements have already been men- tioned. Measurements of the "electrochemical equivalent of the chromium in these compounds, using a chromium anode in an aqueous solution of sodium peroxide, and a divided cell, were not PERCHROMATES 97 satisfactory, since the results could be interpreted in different ways according to the process assumed to take place in the cell. Although hydrogen peroxide cannot be detected among the pro- ducts of the decomposition of the perchromates, yet alkaline gold ' chloride, or potassium permanganate is reduced during the change, which reaction is characteristic of hydrogen peroxide ; hydrogen peroxide is also used in their preparation, and it may therefore be assumed that they contain peroxidic oxygen, that is, the biva- lent O 2 -group. The attempt to determine the number of such groups from the amount of hydrogen peroxide liberated in the presence of permanganate, was not possible with the blue per- chromates, either in acid or in alkaline solution, since they de- compose so very rapidly. In strongly acid solution the red perchromates do not react with permanganate ; in alkaline solu- tion the amount of permanganate reduced increases with the alkalinity of the solution to a maximum of 5*5 equivalents. It is probable, therefore, that one molecule of the salt gives rise to three molecules of hydrogen peroxide, but that some of the peroxidic oxygen escapes without acting on the permanganate. This conclusion is justified by the fact that when added to a strongly acid solution of permanganate the red perchromates de- compose with evolution of oxygen and without appreciable reduc- tion of the permanganate ; also, the blue perchromates, when added to either acid or alkaline permanganate, undergo decom- position with evolution of oxygen, whilst traces only of the per- manganate are reduced. 12 The formula of the acid H 3 CrO 8 may then be written : 0' O^ /O.OH ^ Cr ^-O.OH ^O.OH By analogy, that of the acid corresponding with the blue per chromates is * 12 : O^Cr O.OH It will be noticed that according to these formulae chromium is septavalent in the perchromates, possessing a higher valency than in the chromates, where it is sexavalent. Such a method of formulation is not in accordance with the position of chromium 7 98 PER-ACIDS AND THEIR SALTS in the Periodic System (cf. p. 8) and makes the perchromates exceptions to the other per-acids. If one assumes that the acids forming the basis of the red and blue perchromates have the double formulae H 6 Cr 2 O 16 and H 2 Cr 2 O 10 , they can still be formulated as derivatives of sexavalent chromium namely : HO.CK x r O O Cr and O== Cr O O Cr =O V) As far as the red perchromates are concerned there are no serious difficulties in connexion with this method of formulation. Determina- tions, by the cryoscopic method, of the molecular weight in aqueous solution of the potassium salt, K 3 CrO 8 , gave the values 68-41, 81*81, 61*55 and 7872 for solutions containing respectively 0-0318, 0-0451, 0*0173 an d 0*0451 gram of the substance in 20 grams of water. 12 These numbers are in approximate agreement with the theoretical value of 74*3, assuming complete dissociation ; they vary very much among themselves, however, although the solutions are very dilute and the concentrations not very different. The experimental error in the determination of the freezing point was + 0*003, an ^ would not alone account for the variations. It is conceivable that in aqueous solution the salt is not so stable as Riesenfeld 12 supposes it to be. In the paper describing the preparation and properties of this compound, 11 it is stated that the aqueous solution is stable for days in the presence of alkali at the ordinary temperature ; on boiling, the solution decomposes with formation of chromate, the rate of decomposition increasing with decreasing concentration of the OH'- ion ; in acid solution the salt is unstable. Possibly, therefore, the large variations in the values of the molecular weights may be accounted for by the aqueous solution having undergone more or less decomposition. This decomposition would also account for the observed molecular weights being less than the theoretical required for complete dissociation if 'the formula were K 6 Cr 2i6 namely 84*9. With the red ammonium salt, (NH 4 ) 3 CrO 8 , the molecular weight determinations were made in o*3N ammonium hydroxide solution, the salt not being sufficiently stable in pure water. With solutions con- taining respectively 0*1064 and 0*0532 gram of the salt in 20 grams of the solvent the values obtained for the molecular weight were 129 and 98, the theoretical value, assuming complete dissociation, being 58*5.* The values are much greater than the theoretical and can be accounted * Riesenfeld 12 gives 78 as the theoretical value. An error has evidently been made by assuming that three ions only are formed on complete dissocia- tion. PERCHROMATES 99 for on the assumption that the dissociation of the ammonium salt is not as great as that of the potassium salt,* the lesser dissociation being partly accounted for by the partial suppression of the dissociation by the solvent ammonium hydroxide. In any case, the results are also in agreement with the formula (NH 4 ) 6 Cr 2 O 16 , which would require the molecular weight 67, if complete dissociation were assumed. In the reaction with alkaline permanganate, each molecule of the red perchromate, assuming the double formula, should give rise to seven molecules of hydrogen peroxide ; the experimental results do not con- tradict this, and are readily explained on the assumption that more peroxidic oxygen escapes without acting on the permanganate than was assumed to be the case by Riesenfeld. 12 As regards the acid, HCrO 5 , the experimental evidence, at present, is against the assumption of the double formula, H 2 Cr 2 O 10 . The molecular weight of the pyridine salt, C 5 H 5 N,HCrO 5 , has been determined in benzene solution and found to be 220, 216 and 208 in different ex- periments, the theoretical value being 2i2. 7 12 It is improbable that dis- sociation would take place in benzene solution, and there are no data which would indicate that fission into the components has taken place during the molecular weight determinations, or that a solid solution is formed between the solvent and the solute. It seems, therefore, that chromium, at all events in the blue per- chromates, is septavalent, and forms a real exception to the other elements forming per-acids. If the acid \ O=Cr.O.OH is considered to be the hydrogen peroxide derivative of an acid, \ 0=|Cr.OH, O^ this acid would be analogous to permanganic acid, % 0=Mn.OH, so that a further analogy between the compounds of chromium and manganese would be brought out, other than the isomorphism of the chromates and manganates, and of the chromium and manganese alums, etc. Hitherto nothing has been said as to what compound, or com- pounds, is the cause of the blue colour in the ordinary test for * Conductivity measurements confirmed this, 7 * 100 PER- ACIDS AND THEIR SALTS hydrogen peroxide or chromic acid. According to Riesenfeld, 19 when chromic acid is in excess, the blue colour is due to the acid HCrO 5 , and when hydrogen peroxide is in excess, to the acid H 3 CrO 8 , which, although it gives red salts, is itself blue in colour. These conclusions are based on the isolation of the various perchromates and on experiments dealing with the de- composition of hydrogen peroxide in the presence of chromic acid, and have led to much controversy with Spitalsky, who has studied the kinetics of the reaction just mentioned. The matter cannot be taken as definitely settled, and the reader is referred to the literature 20 to form his own conclusions. CHAPTER IX. PERMOLYBDATES, PERTUNGSTATES AND PERURANATES. Permolybdic Acid and Permolybdates. IN acid solutions the molybdates give a yellow coloration with hydrogen peroxide, 1 which is not extracted by ether, 2 nor is it affected by heating the liquid. 3 The colour is due to the formation of permolybdic acid, or acids, 'in solution. Fairley * found that molybdenum trioxide reacts with hydrogen peroxide giving a compound which dissolves in acid solution to form a yellow or deep orange liquid, which gradually deposits an in- soluble compound of a yellow colour, whilst Pechard 5 isolated a permolybdic acid of the empirical formula HMoO 4 ,2H 2 O, either by decomposing the barium salt with the equivalent quantity of sulphuric acid, or by dissolving hydrated molybdic acid, MoO 3 ,2H 2 O, metallic molybdenum, or the blue oxide of molyb- denum, in a solution of hydrogen peroxide. The solution obtained by either of these methods is evaporated in a vacuum at the ordinary temperature, and yields a yellow crystalline powder. It can be boiled without undergoing decomposition and is not affected by strong acids, with the exception of hydrochloric acid which reduces it to molybdic acid with evolution of chlorine. The solid loses 2H 2 O at 100, and at a higher temperature it decom- poses with loss of water and oxygen. Muthmann and Nagel state that Pechard 's analytical re- sults are untrustworthy, and Pissarjewsky, 7 on repeating the preparation of permolybdic acid according to Pechard's method, obtained a yellow product having the composition H 2 MoO 5 ,2H 2 O, in which the ratio of molybdenum to active oxygen, as determined by titration with potassium permanganate in sulphuric acid solu- tion, was i : i. On the other hand, Cammerer 8 has described a product of the composition 2MoO 3 ,H. 2 O 2 ,H 2 O, obtained by heating molybdic acid with a two per cent solution of hydrogen 101 102 PER-ACIDS AND THEIR SALTS peroxide ; with the exception of the combined water, this com- pound has the same empirical formula as that described by Pechard. This latter investigator has also described the preparation of a number of crystalline salts having the general composition RMoO 4 , where R is a univalent metal. 9 For example, potas- sium trimolybdate dissolves in hydrogen peroxide forming an orange-yellow solution which, when concentrated at a gentle heat, deposits beautiful yellow crystals of the composition KMoO 4 , 2H 2 O ; when gently heated in a vacuum they lose water and oxygen. Similarly ammonium molybdate, when evaporated with hydrogen peroxide at ioo, yields yellow crystals of the composi- tion NH 4 MoO 4 ,2H 2 O. Analogy to the persulphates would point to these compounds having the formula M 2 Mo 2 O 8 , 10 where M is a univalent metal, and Moeller n has shown, by cryoscopic measure- ments, that in solution the potassium and ammonium salts have molecular weights greater than would correspond with the single formulae, although the solutions are good electrolytes. Owing to the uncertainty 6 as to the accuracy of Pechard's analytical re- sults, however, the existence of the acid H 2 Mo 2 O 8 and corres- ponding salts cannot be considered as definitely proved. In addition to the yellow powder, H 2 MoO 5 ,2H 2 O, an orange- red, amorphous compound having the composition H 2 MoO 5 , xH 2 O(? x= i-J), and possessing the properties of a per-acid, has been obtained by digesting molybdic trioxide with 25 per cent hydrogen peroxide, first at the ordinary temperature and finally on the water bath, filtering, and concentrating the filtrate under diminished pressure. 12 It reduces potassium permanganate, silver oxide, and the hypochlorites, and liberates the halogens from their compounds with hydrogen. Pissarjewsky 13 has also shown that when molybdic trioxide is dissolved in aqueous hydrogen peroxide the following heat effects occur : Mols. of H ? O 2 per mol. of MoO 3 i 2 3 5 Heat developed (Cals.) .... S'o8 12-33 12-33 12-44 These figures indicate the existence in solution of a compound MoO 3 ,2H 2 O 2 , which is scarcely dissociated into its components. Such a conclusion is in agreement with Brode's observations on the distribution of hydrogen peroxide between ether and aqueous solutions of molybdic acid. 14 These results, together with the isolation of the compound PERMOLYBDATES, PERTUNGSTATES, PERURANATES 103 H 2 MoO 5 ,xH 2 O, indicate the existence of at least two permolybdic acids, MoO 3 ,H 2 O 2 and MoO 3 ,2H 2 O 2 , the constitution of which may be written as : // HO.CK (HO) 2 Mo=0 and (HO) 2 Mo- O or \Mo = O ^O \0/ H0/ ^ L respectively. 7 Barwald 2 has given very complex formulae to various potassium, ammonium and barium salts prepared by him, but his results are contradicted by Muthmann and Nagel. 6 These latter experimenters describe the preparation of an orange-red and a lemon-yellow ammonium salt by the action of 20 per cent hydrogen peroxide on ordinary ammonium molybdate, the latter salt being obtained by concentrating the mother liquor from the crystals of the former salt. The formula of the orange-red salt is 3(NHJ 2 O,7MoO 4 ,i2H 2 O, and that of the lemon-yellow salt, 3(NH 4 )2O,5MoO 3 ,2MoO 4 ,6H 2 O, the constitutional formulae as- signed to them being : f o_ o / o o \ o o o o / \ o o Mo ! O Mo O Mo and Mo f O Mo ) O Mo A,A A I 3 Oxygen is evolved when the substance is ignited, but different observers give different tem- peratures at which the evolution of oxygen commences. Ac- cording to Brunck, 4 when heated at 150 in an atmosphere of carbon dioxide, no water is lost, decomposition beginning only above 1 50. The substance contains peroxidic oxygen, since it decolorizes permanganate in dilute sulphuric acid solution, which reaction affords a method for determining the active oxygen ; it liberates chlorine from hydrochloric acid. The ratio of uranium to active oxygen is I : I, 3 so that the formula may be written as : (HO) a U = O ^ or perhaps : According to Fairley, 1 the above precipitate, when dried at ordinary temperatures, has the composition UO 4 ,4H 2 O, but it is some- what doubtful whether the product is a definite chemical compound. 2 - 5 The same author has also described an anhydrous compound, UO 4 , obtained as a heavy crystalline precipitate, almost white in colour, when uranyl nitrate solution is added to a mixture of hydrogen peroxide and a large excess of sulphuric acid, and the solution kept for a week or more ; other investigators 2 have been unable to obtain this compound. The solubility in water of the compound precipitated by hydrogen peroxide from a solution of uranyl acetate, expressed in gram? of UO 3 per litre of solution, is o f Oo6i at 20 and 0-0084 at 90. 6 Peruranates of the alkali metals, including ammonium, are generally obtained by the action of hydrogen peroxide on alkaline solutions of uranyl nitrate, the addition .of alcohol being necessary in many cases to precipitate the salt. 1 ' 7 ' 8 ' 9 For example, with ex- cess of sodium hydroxide, yellow needles of the salt 2Na 2 O,UO 4 , 108 PER-ACIDS AND THEIR SALTS 8H 2 O, separate, whereas with less sodium hydroxide, a red salt, Na 2 O 2 ,2UO 4 ,6H 2 O,* is obtained. 1 An orange-yellow ammonium salt, (NH 4 ) 2 O 2 ,UO 4 ,8H 2 O, was also obtained by Fairley. By double decomposition with solutions of the sodium salt, peru- ranates of the heavier metals, for example, of barium, calcium, copper, nickel, and lead, have been prepared. 7 These salts are more stable than the corresponding salts of molybdenum and tungsten, but they gradually decompose with loss of oxygen. Almost all of them yield ozonized oxygen when treated with con- centrated sulphuric acid. Dioxides, such as manganese and lead dioxides, give rise to a vigorous evolution of oxygen, which takes place at the expense of the metallic hydroperoxide contained in the salts. The solutions of the salts of the alkali metals contain hydrogen peroxide, and the total amount of active oxygen can be determined by titration with potassium permanganate in sulphuric acid solution. The action of freshly precipitated aluminium hydroxide on the salts of the alkali metals converts them into soluble metallic hydro peroxides and insoluble peruranic acid, 7 and this sup- ports the formulation of, for example, the sodium salt as (Na2O 2 ) 2 UO 4 ,8H 2 O, and not as 2Na 2 O,UO (5 ,8H 2 O. 1 Also, carbon dioxide, which has no action on uranic acid, converts insoluble peruranates, for example, (BaO 2 ) 2 UO 4 ,9H 2 O, into metallic hydrogen carbonates, hydrogen peroxide, and free peruranic acid. MelikofT and Pissarjewsky 10 have investigated these reactions quantitatively in order to formulate constitutions for these salts. With the sodium salt, (Na 2 O 2 ) 2 UO 4 ,8H 2 O, for example, two-thirds of the active oxygen is found as hydrogen peroxide in the filtrate from the aluminium hydroxide, and one-third in the precipitate, results which would be in accordance with the formula : (NaO.O^/ O \0 although Melikoff and Pissarjewsky prefer, for what seem to be insufficient reasons, the formula : * Aloy 8 has prepared sodium and potassium salts having the composition Na 2 UO 5 ,5H 2 O and K a UO 5 ,5H,O, respectively, and Oechsner de Coninck 9 claims to have prepared anhydrous K 2 UO 5 . PERMOLYBDATES, PERTUNGSTATES, PERURANATES 109 NaO.O\ \U = O NaO/ ^ The salt Na 2 O 2 (UO 4 ) 2 ,6H 2 O, and similar salts, may be considered as derived from a pyroperuranic acid, thus : O O II II NaO.O U O U ONa /\ X\ O O O O The conductivities of aqueous solutions of the salt (Na20 Q ) 2 U0 4 ,8H 2 0, and the distribution coefficient of hydrogen peroxide between ether and the aqueous solution of the salt, have led Pissarjewsky n to the conclusion that a very complex hydrolysis takes place, with hydrogen peroxide, sodium hydroperoxide and sodium hydroxide as hydrolytic products. As is the case with molybdenum and tungsten, fluoroxy- peruranates have been described, 12 as also complex compounds formed by the interaction of uranyl nitrate, hydrogen peroxide, and, for the most part, salts of organic acids. 13 These latter compounds are decomposed by water with precipitation of per- uranic acid. LITERATURE REFERENCES. INTRODUCTION. 1 Tanatar, Ber., 1900, 33, 205 ; 1903, 36, 1893 ; 1909, 42, 1516. 2 Pellini and Meneghini, Zeitsch. anorg. Chem., 1908, 60, 178. 3 Cf. Bellucci and Clavari, Atti R. Accad. Lincei, 1905 [v.], 14 (ii.), 234. 4 Tubandt and Riedel, Ber., 1911, 44, 2565. 5 Riche, Jahresber. Chem., 1860, 66. 6 See Gmelin-Kraut's Handbuch der anorg. Chem., yth edition, I (i.), 146. 7 Willstatter, Ber., 1903, 36, 1828. 8 Tanatar, Ber., 1899, 32, 1544 ; Wolffenstein and Peltner, Ber., 1908, 41, 275, 280 ; 1909, 42, 1777. 9 Schone, Annalen, 1878, 192, 257; 1878, 193, 241; de Forcrand, Compt. rend., 1900, 130, 716, 778, 1250. 10 Calvert, Zeitsch. physikal. Chem., 1901, 38, 513. 11 Tafel, Ber., 1894, 27, 816, 2297. 12 Cf. Riesenfeld and Mau, Ber., 1911, 44, 3595. 13 Cf. d'Ans, Zeitsch. Elektrochem., ign, 17, 849 ; d'Ans and Friederich, Zeitsch. anorg. Chem. 1911, 73, 325. 14 Baeyer, Ber., 1900, 33, 247g. 15 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., i8gg, 19, 416. 16 Carrasco, Gazzetta, ign, 41 (i.), 16. 17 Cf. Ebler and Krause, Zeitsch. anorg. Chem., ign, 71, 150. 1? a Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., i8gg, 21, 70. 18 Mendele"eff, Principles of Chem., 3rd edition, p. 21. 19 Barreswil, Ann. Chim. Phys. 1847, 20, 364. 20 Pissarjewsky, Zeitsch. anorg. Chem., igoo, 24, 108. 21 Cf. Gmelin-Kraut's Handbuch der anorg. Chem., 7th edition I (i.), 135. 22 Pissarjewsky, Zeitsch. physikal. Chem., igo3, 43, 160. 23 Melikoff and Pissarjewsky, Ber., i8g7, 30, 2go2. 24 Cf. Abegg's Handbuch der anorg. Chem., III. (ii.), 455; also Faber, Zeitsch. anal. Chem., igo7, 46, 277. j 55 Fairley, Trans. Chem. Soc., 1877, 31, 127. -" Riesenfeld, Ber., igo8, 41, 3g4i. 27 Cf. Pechmann and Vanino, Ber., i8g4, 27, 1510 ; Baeyer and Villiger, Ber., igoo, 33, i56g, 247g ; igoi, 34, 762 ; Clover and Richmond, Amer. Chem. J., igo3, 29, i7g ; Clover and Houghton, ibid., igo4, 32, 43, etc. PERSULPHURIC ACIDS AND PERSULPHATES. 1 Meidinger, Pogg. Annalen, 1853, 88, 57, 77 ; cf. also Schonbein, ibid., 1845, 65, 161 ; Bunsen, ibid., 1854, 91, 621 ; Hoffmann, ibid., 1867, 132, 607 ; Rund- spaden, ibid., 151, 306. 2 Brodie, Trans. Chem. Soc., 1864, 17, 2g3. 3 Berthelot, Compt. rend., 1878, 86, 20, 71, 277 ; 1880, 90, 26g, 331 ; i8gi, 112, 1481 ; i8g2, 114, 875. 4 McLeod, Trans. Chem. Soc., 1886, 49, sg4. in 112 PER-ACIDS AND THEIR SALTS 5 Mendeldeff, Bull. Soc. Chim., 1882, 38, 168.- 6 Richarz, Wied. Annalen, 1885, 24, 183 ; 1887, 31, 912 ; Ber., 1888, 21, 1669, 1682. 7 Marshall, Proc. Roy. Soc. Edin., 1891, 18, 63 ; Trans. Chem. Soc., 1891, 59, 771. 8 Traube, Ber., 1889, 22, 1518, 1528; 1891, "24, 1764; 1892, 25, 95 ; 1893, 26, 1481. 9 Traube, Ber., 1886, 19, 1115. 10 Caro, Zeitsch. angew. Chem., 1898, n, 845, s " Marshall, J. Soc. Chem. Ind., 1897, 16, 396. " 12 Elbs and Schonherr, Zeitsch. Elektrochem., 1894-5, *> 4 X 7> 468 ; 1895-6, 2, 245. 13 Miiller and Schellhaas, Zeitsch. Elektrochem., 1907, 13, 257. 14 Cf. also Schellhaas, Zeitsch. Eloktrochem., 1908, 14, 121. 15 Moldenhauer, Zeitsch. Elektrochem., 1905, u, 307. 16 Petrenko, J. Russ. Phys. Chem. Soc., 1904, 36, 1081. & Baeyer and Villiger, Ber., 1901, 34, 853. 18 Mugdan, Zeitsch. Elektrochem., 1903, 9, 718, 980. 19 . Price and Denning, Zeitsch. physikal. Chem., 1903, 46, 89. 20 Robertson, Proc. Roy. Soc., 1891, 50, 105 ; Darrieus and Schoop, Zeitsch. Elek- trochem., 1894-5)^1, 293 ; Elbs and Schonherr, ibid., 1894-5, i, 473 ; 1895-6, 2, 471 ; Schoop/ZW, 1895-6, 2, 273. 21 d'Ans and Friederich, Ber., 1910, 43, 1880. 22 d'Ans and Friederich, Zeitsch. anorg. Chem., 1911, 73, 325 ; cf. also d'Ans, Zeitsch. Elektrochem., 1911, 17, 849. 23 Elbs, J. pr. Chem., 1893, 48, 185. 24 Fock, Zeitsch. Krystall., 1893, 22, 29. 25 Elbs, Zeitsch. Elektrochem., 1895-6, 2, 162. 28 Deissler, Ger. Pat., 105,008 (1898). 27 Miiller, Zeitsch. Elektrochem., 1901, 7, 398 ; 1904, 10, 776. 28 Muller and Friedberger, Zeitsch. Elektrochem., 1902, 8, 230. 29 Cf. Biiltemann, Dissertation, Dresden, 1905. 30 Levi, Zeitsch. Elektrochem., 1903, 9, 427. 31 Konsortium elektrochem. Ind., Ger. Pat., 195,811 (1907). 32 Konsortium elektrochem. Ind., Ger. Pat., 155,805, 170,311 (1904) ; 173,977 (1905) ; French Pat., 351,613 (1905) ; Teichner and Askenasy, Eng. Pat., 2823 (1906). 33 Vereinigte Chemische Werke, Ger. Pat., 205,067-69 (1908). 34 Cf. also Blumer, Zeitsch. Elektrochem., 1911, 17, 965. 35 Lowenherz, Ger. Pat., 81,404. 36 Konsortium elektrochem. Ind., Ger. Pat., 195,811 (1907). 37 Lowenherz, Ger. Pat., 77,340. 38 Foster and Smith, J. Amer. Chem. Soc., 1899, 21, 934. 39 Marshall, J. Amer. Chem. Soc., 1900, 22, 48. 40 Otin, Zeitsch. Elektrochem., 1911, 17, 919. 41 Jorgensen, Zeitsch. anorg. Chem., 1898, 17, 459. 42 Barbieri and Calzolari, Zeitsch. anorg. Chem., 1911, 71, 347. 43 Pajetta, Gazzetta, 1906, 36, (ii.), 298. 44 Tarugi, Gazzetta, 1902, 32, (ii.), 383 ; Levi and Migliorini, ibid., 1906, 36, (ii.), 599- i5 Green and Masson, Trans. Chem. Soc., 1910, 97, 2083. 46 Price, Zeitsch. physikal. Chem., 1898, 27, 474. 47 Merk, Pharm. Zeit., 1906, 50, 1022. 48 Marshall, Proc. Roy. Soc. Edin., 1898, 22, 388. 49 Federlin, Zeitsch. physikal. Chem., 1902, 41, 256. 60 Berthelot, Ann. Chim. Phys., 1895 [7], 4, 429. 51 Marshall, Chem. News, 1901, 83, 76. 52 Austin, Trans, Chem. Soc., 191 1, 99, 262. LITER A TURE REFERENCES 1 1 3 53 Levi and Migliorini, Gazzetta, 1908, 38, (ii.), 10 ; cf. also Kempf, Ber., 1905, 38, 3972. Kempf and Oehler, ibid., 1908, 41, 2576. 54 Marshall, Proc. Roy. Soc. Edin., 1900, 23, 168 ; Marshall and Inglis, ibid., 1902, 24,88. 55 Dittrich and Bollenbach, Ber., 1905, 38, 747. 56 Kempf, Ber., igos, 38, 3g63. 57 Dittrich and Hassel, Ber., 1903, 36, 1929. 58 Marshall, Trans. Chem. Soc., 1908, 93, 1726. 59 Marshall, Trans. Edin. Photog. Soc., igo2, 2, 117. 60 Levi, Migliorini and Ercolini, Gazzetta, 1903, 33, (i.), 583. 61 Goschuff, Deutsch. Mechan. Zeit., 1910, 134, 141. 62 Namias, L'Orosi, 1900, 23, 218 ; Tarugi, Gazzetta, 1903, 33, (i.), 127 ; Turrentine, J. Physical Chem., 1907, n, 623. 63 Levi, Migliorini and Ercolini, Gazzetta, 1908, 38, (i.), 583. 64 Price, Ber., 1902, 35, 291. 2. Friend, Trans. Chem. Soc., igo6, 89, iog2. 66 Price, Trans. Chem. Soc., 1907, 91, 531. 67 Cf. Stenger and Heller, Zeitsch. wiss. Photog., Photophysik,lPhotochem., 1911, 9, 389 ; Dodgson, Phot. J., igu, 51, 302. 68 Namias, L'Orosi, 1905, 27, 1550 ; Dakin, J. Soc. Chem. Ind., 1902, 21, 848 ; Dittrich and Hassel, Ber., 1902, 35, 3266; 1903, 36, 284, 1423 ; Chem. Zeit., 1903, 27, 853 ; Zeit. anal. Chem., 1904, 43, 382 ; von Knorre, Zeitsch. angew- Chem., 1891, 59, 771; 1903, 16,905; Chem. Zeit., 1903, 27, 53; Zeitsch. anal. Chem., 1904, 43, i ; 1905, 44, 88 ; Baubigny, Compt. rend., 1902, 135, g65, mo; igo3, 136, 44g, 1325, 1662 ; Stehmann, J. Amer. Chem. Soc., 1902, 24, 1204; Walters,'}. Amer. Chem. Soc., 1905, 27, 1550; Brichant, Ann. Chim. anal., 1906, n, 124 ; Gottschalk, Zeitsch. anal. Chem., 1908, 47, 237; Karslake, 1 J. Amer. Chem. Soc., 1908,30,905 ; Orthey, Zeitsch. anal. Chem., 1908, 47, 547 ; Rubricius, Stahl u. Eisen, 1905, 25, 890 ; Kunze, Chem. Zeit., 1905, 29, 1017 ; Stahl u. Eisen, 1908, 28, 175 ; Wdowiszewski, Stahl u. Eisen, 1908, 28, 1067. 69 Namias, L'Orosi, 1900, 23, 218 ; Dittrich and Hassel, Ber., 1903, 36, 284 ; von Knorre, Zeitsch. angew. Chem., 1903, 16, 1097 ; Zeitsch. anal. Chem., 1904, 43, i ; Ibbotson and Howden, Chem. News, 1904, 90, 320 ; Walters, J. Amer. Chem. Soc., 1905, 27, 150. 70 Seyewitz and Trawitz, Compt. rend., 1903, 137, 130. Dittrich and Reise, Ber., I 95. 38, i82g; Dede, Chem. Zeit., igu, 35, 1077 > Kempf and Oehler, Ber., igo8, 41, 2576 ; Brunner and Mellet, J. pr. Chem., igo8, (u.),77, 33 ; see also Chem. Zeit., igo2, 26, goo ; Ger. Pat, 134,301. 71 Namias, L'Orosi, igoo, 23, 21 8. 72 Elbs, Zeitsch. angew. Chem., i8g7, 10, 195. 73 Schering, Ger. Pat, 8i,2g7. 74 Bayer and Co., Ger. Pat. 75 , Vitali, Boll. Chim. Farm., igo3, 42, 273, 321 ; cf. also Wolff and Wolffenstein, Ber., igo4, 37, 3213 ; igo8, 41, 717. 76 von Knorre, Zeitsch. angew. Chem., i8g7, 10, 719. 77 Rothenfusser, Zeitsch. Nahr. Genussm., igo8, 16, s8g. 78 Le Blanc and Eckhardt, Zeitsch. Elektrochem., i8gg, 5, 355. 79 Mondolfo. Chem. Zeit, 1899, 23, 699. 80 Marie and Bunel, Bull. Soc. Chim., 1903, (iii.), 29, 930. 81 Griitzner, Arch. Pharm., i8gg, 237, 705 ; Peters and Moody, Amer. J. Sci., 1901 [4], 12, 367; Moreau, Bull. Soc. Pharm., igoi, 3, 179; Allard, J. Pharm. Chim., 1901, [6], 14, 506 ; Vitali, Boll. Chim. Farm., 1903, 42, 273, 321 ; Moreau, Apoth. Zeit., igoi, 16, 383 ; Pannain, Gazzetta, igo4, 34, (i.), 500 ; Rimini, Atti R. Accad. Lincei, igo6, (v.), 15, (ii.), 320; Knecht and Hibbert, Ber., 1905, 38, 3318 ; Pozzi-Escot, Bull, de 1'Assoc. des Chim. de Sucr. et Dist, 1908, 26, 267. 8 1.14 PER-ACIDS AND THEIR SALTS 82 Friend, Trans. Chem. Soc., 1904, 85, 597. *533 ; 1905, 87, 738, 1367 ; 1906, 89, 1092. 83 See Friend, Proc. Chem. Soc., 1910, 26, 88. 84 Skrabal and Vacek, Oesterr. Chem. Zeit., 1910, (ii), 13, 27. 85 Bredig, Zeitsch. physikal. Chem., 1893, 12, 230. 86 Moeller, Zeitsch. physikal. Chem., 1893, 12, 555. 87 Lowenherz, Chem. Zeit., 1892, 16, 838. 88 Lowenherz, Zeitsch. physikal. Chem., 1895, 18, 70. 89 Helmholtz, Trans. Chem.. Soc., 1881, 39, 289. 90 Richarz, Ber., 1888, 31, 1672 ; cf. Starck, Zeitsch. physikal. Chem., 1899,129, 385. 91 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1898, 18, 59. 92 Friessner, Zeitsch. Elektrochem., 1904, 10, 265 ; cf. also Muller, ibid., 1904, 10, 780 ; Skirrow, Zeitsch. anorg. Chem., 1903, 33, 25. 93 Kastle and Loewenhart, Amer. ,Chem. J., 1903, 29, 563. 94 Baeyer and Villiger, Ber., 1900, 33, 124, 858, 1569. 95 Bad. Anilin. Soda Fabrik, Ger. Pat., 110,249 (1898). 96 Berthelot, Ann. chim. phys., 1878 [5], 14, 360. 97 Tubandt and Riedel, Ber., 1911, 44, 2565 ; Zeitsch. anorg. Chem., 1911, 72, 219. 98 Lowry and West, Trans. Chem. Soc., 1900, 77, 950. 99 Armstrong, Proc. Chem. Soc., 1893, 9, 190; Trans. Chem. Soc., 1895. 67, 1156. 100 Armstrong and Lowry, Proc. Roy. Soc., 1902, 70, 94. 101 Ahrle, Dissertation, Darmstadt, 1908, p. 141 ; J. pr. Chem., 1909, 79, 129 ; Zeitsch. angew. Chem., 1909, 22, 1713. 102 Price, Ber., 1902, 35, 292 ; Trans. Chem. Soc., 1903, 83, 543. 103 Price, Trans. Chem. Soc., 1906, 89, 54. 104 Price, Trans. Chem., Soc., 1907, 91, 535. 122- Willstatter and Hauenstein, Ber., 1909, 42/1839. 106 Bamberger, Ber., 1900, 33, 1959. 107 Price and Friend, Trans. Chem. Soc., 1904, 85, 1526. 108 Baeyer and Villiger, Ber., 1900, 33, 2488. 109 Bach, Ber., 1900, 33, 1506, 3111 ; 1901, 34, 1520, 3851 ; 1902, 35, 158, 872, 3940 ; Armstrong, Proc. Chem. Soc., 1900, 16, 134 ; Ramsay, Trans. Chem. Soc., 1901, 79, 1324. 110 Wedekind, Ber., 1902, 35, 2267. 111 Baeyer and Villiger, Ber., 1899, 32, 3625 ; 1900, 33, 124, 858, 2479 ; Bamberger and others, Ber., 1899, 32, 1676 ; 1900, 33, 533, 1781 ; 1901, 34, 2023 ; 1902, 35, 1082 ; Cross, Bevan and Briggs, Chem. News, 1900, 82, 163 ; Albitzky, Ber., 1900, 33, 2909 ; Seel, Ber., 1900, 33, 3212 ; Springer, Pharm. Zeit., 47, 157 ; Wolffenstein and others, Ber., 1899, 32, 432 ; 1901, 34, 2423 ; 1904, 37, 3215, 3221 ; Ditz, Chem. Zeit., 1907, 31, 833, 844, 857 ; J. pr. Chem., 1908, (") 78, 343 ; Erban, Chem. Zeit., 1908, 32, 829 ; Tarugi, Gazzetta, 1902, 32, (ii.), 380 ; Migault, Chem. Zeit., 1910, 34, 337 ; Dittrich, Ber., 1903, 36, 3385, etc. 112 Konsortiuni elektrochem. Ind., Ger. Pat., 199,958, 217,538, 217,539 ; Fr. Pat., 358,806 (1905); Urbasch, Fr. Pat., 371,043 (1906); Teichner, Eng. Pat., 24,507 (1905) ; Oesterr. Chem. Werke, A. G., and L. Lowenstein, Fr. Pat., 422,460 (1910). 113 Pietzsch and Adolph, Fr. Pat., 421,164 (1910), 430,538 (1911) ; Eng. Pat., 23,660 (1910). 114 Malaquin, J. Pharm. Chim., 1911 [7], 3, 329. PERSELENATES. 1 Dennis and Brown, J. Amer. Chem. Soc., 1901, 23, 258. LITER A TURE REFERENCES 1 1 5 PERBORATES. I Etard, Compt. rend., 1880, 91, 931. > 2 Melikoff and Pissarjewsky, Ber., 1898, 3*, 678, 953. 3 Tanatar, Zeitsch. physikal. Chem., 1898, 26, 132 ; 1899, 29, 162; Jaubert, Compt. rend., 1904, 139, 796 ; Bruhat and Dubois, ibid., 1905, 140, 506; Jaubert and Lion, Rev. gen. Chem. pure appl., 1905 [7], 8, 163 ; Christensen, Danske Vidensk. Selsk. Forh., 1904, No. 6. 4 Tanatar, Zeitsch. physikal. Chem., 1898, 26, 132. 5 Constam and Bennet, Zeitsch. anorg. Chem., 1900, 25, 265. 6 Bruhat and Dubois, Compt. rend., 1905, 140, 506. 7 Pouzenc, Fr. Pat., 411,258 ; cf. Beltzer, Le Moniteur Scient., 1911, p, 10. 8 Tanatar, Zeitsch. anorg. Chem., 1901, 26, 343 ; cf. Pissarjewsky, ibid., 1902, 32, 341. 9 Constam and Bennet, Zeitsch. anorg. Chem., 1901, 26, 451. 10 Riesenfeld and Reinhold, Ber., 1909, 42, 2977. II Jaubert, Compt. rend., 1904, 139, 796. 12 Jaubert and Lion, Rev. gen. Chem. pure appl., 1905 [7], 8, 163. 13 Christensen, Danske Vidensk. Selsk. Forh., 1904,1 No. 6. 14 Pissarjewsky, Zeitsch. physikal. Chem., 1903, 43, 170. 15 Calvert, Zeitsch. physikal. Chem., 1901, 38, 513. 16 Melikoff and Pissarjewsky, Ber., 1897, 30, 3144; 1898, 31, 446. 17 Petrenko, J. Russ. Phys. Chem. Soc., 1902, 34, 37. 18 v. Girsewald and Wolokitin, Ber., 1909, 42, 865. 19 Melikoff and Lordkipanidze, Ber., 1899, 32, 3349, 3510 ; J. Russ. Phys. Chem. Soc., 1900, 32, 77- 20 Farrar, J. Soc. Dyers, 1910, 26, 81. 21 Cf. also Rupp and Mielck, Arch. Pharm., 245, 5 ; Lenz and Richter, Zeitsch. anal. Chem., 1911, 50, 537- 22 Fuhrmann, Chem. Zeit., 1911, 35, 1022. 23 Cf. Saccharin Fabrik Aktienges. vorm. Fahlberg, List and Co., Fr. Pat., 425,958 (1911) ; Klages and Sommer, Salbke-Westerhiisen an Elbe, U.S. Pat., 996,773 (1911) ; Chem. Werke vorm. Dr. H. Byk, Charlottenburg, U.S. Pat., 999,497 (1911) ; Eng. Pat., 1626 (1911) ; Chem. Fabrik Grunau, Landshoff and Meyer A.-G., and A. Brauer, Ger. Pat., 237,096 (1910) ; Konsort. electrochem. Ind., Ger. Pat., 237,608 (1911). 24 Matthews, J. Ind. Eng. Chem., 1911, 3, 191. PERCARBONATES. 1 Constam and Hansen, Zeitsch. Elektrochem., 1896-7, 3, 137. 2 Hansen, Zeitsch. Elektrochem., 1896-7, 3, 445. 3 Salzer, Zeitsch. Elektrochem., 1902, 8, 900. 4 Riesenfeld and Reinhold, Ber., 1909, 42, 4377. Biltz and Gahl, Zeitsch. Elektrochem., 1905, II, 412. Cf. Mellor's Statics and Dynamics, p. 59. Cf. Abegg's Handbuch der anorg. Chem., Ill (Hi.), p. 308. . Riesenfeld and Mau, Ber., 1911, 44, 3589 : Riesenfeld, Ber., 1910, 43, 566. Brown, J. Amer. Chem. Soc., 1905, 27, 1222. 10 Bach, J. Russ. Phys. Chem. Soc., 1897, 29, 373 ; Wolffenstein and Peltner, Ber., 1908, 41, 280. 11 Tanatar, Ber., 1899, 22, 1544. 12 Tanatar, Zeitsch. anorg. Chem., 1901, 28, 255 ; Willstatter, Ber., 1903, 36, 1828. 13 Tanatar, J. Russ. Phys. Chem. Soc., 1902, 34, 952. 8* Il6 PER- ACIDS AND THEIR SALTS 14 Riesenfeld and Reinhold, Ber., 1909, 42, 4377; 1910, 43, 566, 2594; 1911, 44, 3589; Tanatar, Ber., 1910, 43, 127, 2149; Wolffen stein, Ber., igio, 43, 639- 15 Kasanezky, J. Russ. Phys. Chem. Soc., 1902, 34, 202, 388 ; 1905, 35, 57 ; Peltner, Ber., 1909, 42, 1777- 18 Wolffenstein and Peltner, Ber., 1908, 41, 280. 17 Riesenfeld and Mau, Ber., ign, 44, 3595. is Tafel, Ber., 1894, 27, 2297. 19 d'Ansand Friederich, Zeitsch. anorg. Chem., 1911, 73, 325. 20 Bauer, Ger. Pat., 145,746 (1904) ; Merck, Ger. Pat., 188,569 (1907). 21 Wolffenstein and Peltner, Ber., 1908, 41, 275. PERNITRATES. 1 Hautefeuille and Chappuis, Compt. rend., 1881, 91, 134 ; 1881, 92, 80; 1882, 94* mi, 1306; Chappuis, ibid., 1882, 94, 946. 2 Berthelot, Compt. rend., 1881, 92, 82 ; Ann. Chim. Phys., 1881 [5], 22, 432. 3 d'Ans and Friederich, Zeitsch. anorg. Chem., 1911, 73, 344 ; d'Ans, Zeitsch. Elektrochem., 1911, 17, 850. 4 Raschig, Zeitsch. angew. Chem., 1904, 17, 1419 ; Ber., 1907, 40, 4585. 5 Schmidlin and Massini, Ber., 1910, 43, 1162. 6 Schellhaas, Zeitsch. Elektrochem., 1908, 14, 121. 7 Ritter, Gehlen's neues Journ., 1804, 3, 561 ; Wallquist, J. pr. Chem., 1842, 31, 179 ; Fischer, ibid., 1844, 33, 277 ; Grotthus, Jahresber. Chem., 1852, 423 ; Mahla, Annalen, 1852, 82, 289 ; Berthelot, Compt. rend., 1880, 90, 653 ; Novak, Rozpravy Cesk6 Akad., 1896, 5, No. 6 ; ulc, Zeitsch. anorg. Chem., 1896, 12, 89, 180 ; 1900, 24, 305 ; Mulder and Heringa, Rec. Trav. Chim., 1896, 15, (i.), 235 ; Mulder, ibid., 1897, 16, 57 ; 1898, !? 129 ; 1899, 18, 91 J 1903, 22, 385, 388, 405 ; Tanatar, Zeitsch. anorg. Chem., 1901, 28, 331 ; Kuzma, Rozpravy Ceske Akad., 1905, 19, No. n ; Watson, Trans. Chem. Soc., 1906, 89, 578 ; Brauner and Ku2ma, Ber., 1907, 40, 3371. 8 Mulder, Verh. Kon. Akad. Wetensch. Amsterdam, 1898 ; Rec. Trav. Chim., igoo, I9 II 5- 9 Tanatar, Zeitsch. anorg. Chem., 1901, 28, 331. 10 Bose, Zeitsch. anorg. Chem., 1905, 44, 237. 11 Luther and Pokorny, Zeitsch. anorg. Chem., 1908, 57, 290. 12 Baborovsky and Kuzma, Zeitsch. physikal. Chem., 1909, 67, 48 ; Bose, ibid., 1909, 68, 383. 13 Pinerua- Alvarez, Chem. News, 1906, 94, 269 ; (Paper read at the 6th Internal. Congress of applied Chem., Rome, 1906). PERPHOSPHATES, ETC. 1 Schmidlin and Massini, Ber., 1910, 43, 1162. 2 d'Ans and Friederich, Ber., 1910, 43, 1880 : Zeitsch. anorg. Chem., 1911, 73, 343 ; d'Ans, Zeitsch. Elektrochem., 1911, 17, 850. 3 Petrenko, J. Russ. Phys. Chem. Soc., 1902, 34, 204. 4 Pinerua-Alvarez, Chem. News, 1906, 94, 269. 5 Hanus and Kallauner, Zeitsch. anorg. Chem., 1911, 70, 232. PERTITANATES. 1 Schone, Dingl. Journ., 1873, 210, 317. 2 Haber and Grinberg, Zeitsch. anorg. Chem., 1898, 18, 37; cf. also Jackson, Chem. News, 1883, 47, 157 ; tHillebrand, J. Amer. Chem. Soc., 1895, 17, 718 ; Richard and Lonnes, Zeitsch. physikal. Chem., 1896, 20, 145 ; Reichard, Chem. Zeit., 1904, 28, 16; Faber, Zeitsch. anal. Chem., 1907, 46, 277. LITER A TURE REFERENCES 1 1 7 3 Piccini, Gazzetta, 1882, 12, 151 ; 1883, 13, 57 ; Ber., 1888, 21, 1391 ; Weller, Ber., 1882, 15, 2599 ; Classen, Ber., 1885,21, 370, i65o ; Lvy, Compt. rend., 1889, 108, 294. 4 Classen, Ber., 1888, 21, 370; Melikoffand Pissarjewsky, Ber., 1898, 31, 679. 5 Le"vy, Compt. rend., 1889, 108, 294. 6 Mazzucchelli and Barbero, Atti R. Accad. Lincei, 1906 [5], 15, (ii.), 35, 109. 7 O. and A. Dony-Henault, Bull. Soc. chim. Belg., 1908, 22, 224. 8 Melikoffand Pissarjewsky, Ber., 1898, 31, 678, 953. 9 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1899, 19, 413. 10 Piccini, Compt. rend., 1884, 97, 1064; Gazzetta, 1887, 17, 479. 11 Mazzucchelli and others, Atti R. Accad. Lincei, 1907 [5], 16, (ii.), 265, 349 ; 1909 [5], 18, (i.), 518, 608; Faber, Zeitsch. anal. Chenru, 1907, 46, 277. PERZIRCONATES, ETC. 'Cleve, Bull. Soc. Chim., 1885 [2], 43, 453 ; Bailey, Trans. Chem. Soc., 1886, 49, 149, 481 ; 1889, 58, 705 ; Proc. Roy. Soc., 1889, 46, 74 ; Pissarjewsky, Zeitsch. anorg. Chem., 1900, 25, 378; Geisow and Horkheimer, ibid., 1902, 32, 372; Wedekind, ibid, 1902, 33, 83 ; Hauser, ibid., 1905, 45, 190. 2 Pissarjewsky, Zeitsch. anorg. Chem., 1902, 31, 359. 3 Pissarjewsky, J. Russ. Phys. Chem. Soc., 1900, 32, 609 ; Zeitsch. anorg. Chem., iooe, 25, 37- 4 Cf. Abegg's Handbuch der anorg. Chemie, III (i.), 218; (ii.), 835. PERSTANNATES. 1 Tanatar, Ber., 1905, 38, 1184 ; cf. also Spring, Bull. Soc. Chim., 1889 [3], (i.), 180. 2 Coppadoro, Gazzetta, 1908, 38, (i.). 489. PERVANADATES. 1 Werther, J. pr. Chem., 1861, 83, 364- 2 Scheuer, Zeitsch. anorg. Chem., 1898, 16, 284. 3 Pissarjewsky, Zeitsch. physikal. Chem., 1903, 43, 173 ; J. Russ. Phys. Chem. Soc., 1902, 34, 472 ; Cammerer, Chem. Zeit., 1891, 15. 957. 4 Diillberg, Zeitsch. physikal. Chem., 1903, 45, 170. 5 Pissarjewsky, Zeitsch. physikal. Chem!, 1903, 43, 168 ; J. Russ. Phys. Chem. Soc., 1903, 35, 42. 6 Pissarjewsky, Zeitsch. anorg. Chem., 1902, 32, 341. 7 Pissarjewsky, J. Russ. Phys. Chem. Soc., 1902, 34, 210 ; Zeitsch. physikal. Chem., 1902, 40, 368. 8 Melikoffand Pissarjewsky, Zeitsch. anorg. Chem., 1899, 19, 405. 9 Melikoffand Jehlchaninoff, Ber., 1909, 42, 2291. 10 Abegg's Handbuch der anorg. Chem.yill (Hi.), pp. 745, 781. * For potential measurements on metapervanadic acid, see Mazzucchelli and Barbero, Atti R. Accad. Lincei, 1906 [5], 15, (ii.), 35, 109. PERCOLUM BATES. 1 Melikoffand Pissarjewsky, Zeitsch. anorg. Chem., 1899,20,340; cf. also Meli- koff and Kazanetzky, J. Russ. Phys. Chem. Soc.. 1903, 35, 457; Hall and Smith, Proc. Amer. Phil. Soc., 1905, 44, 177. 2 Balke and Smith, J. Amer. Chem. Soc., 1908, 30, 1637. 3 Piccini, Zeitsch. anorg. Chem., 1892, 2, 21 ; cf. also, Hall and Smith, loc. cit. Il8 PER-ACIDS AND THEIR SALTS PERTANTALATES. 1 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1899, 20, 340. 2 Balke, J. Amer. Chem. Soc., 1905,27, 1140; Balke and Smith, ibid., 1908, 30, 1637- 3 Piccini, Zeitsch. anorg. Chem., 1892, 2, 21. PERCHROMATES. 1 Barreswil, Ann. Chim. Phys., 1847 [3], 20, 364. 2 Schonbein, J. pr. Chem., 1859, 79, 69 ; 1860, 80, 257 ; Aschoff, ibid., 1860, 8l, 401 ; Werther, ibid., 1861, 83, 195 ; Storer, ibid., 1860, 80, 44 ; Proc. Amer. Acad. Arts Sci., IV, 338 ; Brodie, Phil. Trans., 1850, 140, 759 ; Trans. Chem. Soc., 1863, 16, 326 ; Fairley, Chem. News, 1876, 33, 238 ; Moissan, Compt. rend., 1883, 97, 96 ; Martinon, Bull. Soc. Chim., 1886, 45, 862 ; Carnot, Compt. rend., 1888, 107, 948, 997, 1150 ; Berthelot, ibid., 1889, 108, 24, 157, 477 ; Baumann, Zeitsch. angew. Chem., 1891, 4, 135 ; Marchlewski, ibid., 1891, 4, 392 ; Griggi, L'Orosi, 1892, 295 ; Grosvenor, J. Amer. Chem. Soc.," 1895, 17, 41 ; Bach, Ber., 1902, 35, 872 ; Patten, Amer. Chem. J., 1903, 29, 385- 3 Moissan, Compt. rend., 1883, 97, 96. 4 Pechard, Compt. rend., 1891, 113, 39. 5 Wiede, Ber., 1897, 30, 2179. 6 Haussermann, J. pr. Chem., 1893 [ 2 ]> 48, 70. 7 Wiede, Ber., 1897, 30, 2178 ; 1898, 31, 516, 3139 ; 1899, 32, 378. 8 Hofmann and Hiendlmaier, Ber., 1904, 37, 1663, 3405 ; 1905, 38, 3059, 3066. 9 Riesenfeld and others, Ber., 1905, 38, 1885, 338o, 3578, 4068 ; 1908, 41, 2826, 3536, 3941 ; 1911, 44, 147 ; Ber. Naturforsch. Ges. Freiburg, 1906, 17, I. 10 Byers and Reid, Amer. Chem. J., 1904, 32, 503. ,- 11 Riesenfeld, Wohlers and Kutsch. Ber., 1905, 38, 1885. 12 Riesenfeld, Ber., 1908, 41, 3941. 13 Riesenfeld, Ber., 1905, 38, 3380. 14 Riesenfeld, Kutsch and Ohl, Ber., 1905, 38, 4068. 15 Werner, Ber., 1906, 39, 2659. 16 Riesenfeld, Ber., 1908, 41, 3536. 17 Hofmann and Hiendlmaier, Ber., 1905, 38, 3059. 18 Hofmann and Hiendlmaier, Ber., 1906, 39, 3181. 19 Riesenfeld, Ber., 1908, 41, 2826. 20 Riesenfeld, Ber., 1908, 41, 2826 ; 1911, 44, 147 ; Zeitsch. anorg. Chem., 1912, 74, 48 ; Spitalsky, Zeitsch. anorg. Chem., 1907, 53, 184 ; 1907, 54, 265 ; 1907, 56, 72 ; 1910, 69, 179 ; Ber., 1910, 43, 3187. PERMOLYBDATES. 1 Werther, J. pr. Chem., 1861, 84, 198 ; cf. also Schonn, Zeitsch. anal. Chem., 1870, 9. 4i, 3ii- 2 Barwald, Chem. Centr., 1885, 424. 3 Deniges, Compt. rend., 1890, no, 1007. 4 Fairley, Trans. Chem. Soc., 1877, 31, 141. 5 Pochard, Compt. rend., 1892, 114, 1481 ; 1892, 115, 227. 8 Muthmann and Nagel, Zeitsch. anorg. Chem., 11898, 17, 73. 7 Pissarjewsky, Zeitsch. anorg. Chem., 1900, 24, 108. 8 Cammerer, Chem. Zeit., 1891, 15, 957. 9 Pochard, Compt. rend., 1891, 112, 720 ; 1892, 114, 1358. 10 Cf. Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1898, 18, 59. 11 Moeller, Zeitsch. physikal. Chem., 1893, 12, 555. 12 Muthmann and Nagel, Ber., 1898, 31, 1836. LITER A TURE REFERENCES 1 1 9 18 Pissarjewsky, J. Russ. Phys. Chem. Soc., 1902, 34, 210 ; Zeitsch. physikal. Chem., 1902, 40, 368. 14 Erode, Zeitsch. physikal. Chem., 1901, 37, 299. 15 Melikoff and Pissarjewsky, Ber., 1898, 31, 632. 16 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1899, 19, 414. 17 Melikoff and Pissarjewsky, Ber., 1898, 31, 2448. 18 Piccini, Zeitsch. anorg. Chem., 1892, I, 51 ; Kazanezky, J. Russ. Phys. Chem. Soc., Z902, 34, 383. 19 Mazzucchelli and others, Atti R. Accad. Lincei, 1907 [5], 16, (i.) 963 ; 1909 [5], 18, (ii.), 259 ; Gazzetta, 1910, 40, (ii.), 49, 241. * For potential measurements on solutions containing permolybdic acid, see Mazzuc- chelli and Barbero, Atti R. Accad. 'Lincei, 1906 [5], 15, (ii.), 35, 109. PERTUNGSTATES. 1 Fairley, Trans. Chem. Soc., 1877, 31, 141. 2 Cammerer, Chem. Zeit., 1891, 15, 957. 3 Pissarjewsky, Zeitsch. anorg. Chem., 1903, 24, 108. 4 Pissarjewsky, J. Russ. Phys. Chem. Soc., 1902, 34, 210 ; Zeitsch. physikal. Chem., 1902, 40, 368. 5 Erode, Zeitsch. physikal. Chem., 1901, 37, 299 ; Pissarjewsky, J. Russ. Phys. Chem. Soc., 1902, 34, 472 ; Zeitsch. physikal. Chem., 1903, 43, 160. 6 Pechard, Compt. rend., 1891, 112, 1060 ; cf. also Pinerua-Alvarez, Chem. News, 1906, 94, 269. 7 Pissarjewsky, Zeitsch. anorg. Chem., 1898, 18, 59. 8 Thomas, J. Amer. Chem. Soc., 1899, 21, 373. 9 Melikoff and Pissarjewsky, Ber., 1898, 31, 632. 10 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1899, 19, 414. u Mazzucchelli and others, Atti R. Accad. Lincei, 1908 [5], 17, (ii.), 30 ; Gazzetta, 1910, 40, (H.), 241. 12 Piccini, Zeitsch. anorg. Chem., 1892, 2, 21. * For potential measurements on solutions containing pertungstic acid, see Mazzuc- chelli and Barbero, Atti R. Accad. Lincei, 1906, [5], 15, (ii.), 35, 109. PERU RAN ATES. Fairley, Trans. Chem. Soc., 1877, 31, 127. Alibegoff, Annalen, 1886, 233, 117. Pissarjewsky, Zeitsch. anorg. Chem., 1900, 24, 108 ; cf. also Oechsner de Coninck, Bull. Soc. Acad. roy. Belg., 1909, 692. Brunck, Zeitsch. anorg. Chem., 1895, 10, 246. Zimmermann, Annalen, 1886, 232, 324. Mazzucchelli, Atti R. Accad. Lincei, 1906 [5], I5,'(ii.), 429, 494. 7 Melikoff and Pissarjewsky, Ber., 1897, 30, 2902. 8 Aloy, Bull. Soc. Chem., 1902 [3], 27, 734; 1903 [3], 29, 292. 9 Oechsner de Coninck, Bull. Soc. roy. Belg., 1907, 173 ; Compt. rend., 1909, 148, 1769. 10 Melikoff and Pissarjewsky, Zeitsch. anorg. Chem., 1899, 19, 405. 11 Pissarjewsky, Zeitsch. physikal. Chem., 1903, 43, 160. 12 Lordkipanidze, J. Russ. Phys. Chem. Soc., 1900, 32, 283. 13 Mazzucchelli and Bimbi, Atti R. Accad. Lincei, 1907 [5], 16, (ii.), 57^. * For potential measurements on solutions containing peruranic acid, see Mazzuc- chelli and Barbero, Atti R. Accad. Lincei, 1906 [5], 15, (ii.), 35, 109. INDEX. ALKALOIDS and persulphates, 39. Ammonium hydroperoxide, 62, 87. perborate, 62. perchromates, 93, 94. permolybdates, 103. - persulphate, 13, 23. influence of cathode material on preparation of, 26. of current density on prepara- tion of, 26. of temperature on preparation of, 25. pertitanates, 81, 82. peruranate, 108. pervanadates, 87, 88, 89. Analogy between chromium and man- ganese, 99. Anthion, 36. Antihypo, 73. BARIUM metapervanadate, 86. perborate, 63. percarbonate, 71. persulphate, 29. pertitanate, 81. peruranate, 108. Bismuth, compounds of, 78. CESIUM perborate, 63. permolybdates, 103. - persulphate, 29. Calcium perborate, 63. peruranate, 108. Caro's acid, 9, 13, 45. action of heat on neutralized solu- tions of, 50. aniline salt of, 54. constitution of, 54. depolarizing effect of, 16-17. estimation of, 46, 57. formula of, 45-54. interaction with potassium iodide, 46, 50. organic salts of, 52-3. potassium salt of, 50-1. preparation of anhydrous, 53. of solution, 45, 46. properties in solution, 55-7. use as an oxidizing agent, 57-8. Catalysis, explanation of, 34. in reaction between Caro's acid and hydrogen peroxide, 56-7. Catalysis, in reaction between potassium persulphate and potassium iodide, Cerium, hydrated peroxides of, 83. Chromate, influence of addition of, 24-6. Chromium, dichloroaquotriamminechro- mium chloride, 96. estimation of, in irons and steels, 39. tetroxide triammine, 95, 96. Cobalt, hexamminecobaltisulphateper- sulphate, 30. perborate, 63. Complex chromium ammonias, 96. percolumbates, 90. permolybdates, 104. persulphates, 30. pertantalates, 91. pertitanates, 82, 83. pertungstates, 106. peruranates, 109. pervanadates, 89. Conductivity of solutions of persulphates, 42. of potassium metapervana- date, 86. of sodium perborate, 61. Constitution of Caro's acid, 54. of per-acids and their salts, 6-8. of perborates, 59, 64. of percarbonates, 68. of persulphates, 43-4. Copper perborate, 63. peruranate, 108. Cryoscopic measurements on chromium tetroxide triammine, 95. on permonosulphuric acid, 55. on persulphates, 38, 42. on pyridine perchromate, 94, 99. on red perchromates, 93, 98. Cyanides, influence of, on preparation of persulphates, 28. DIOXIDES, i, 34. Dithionates, formation from sulphites, 44. ELEMENTS forming per-acids, 5. Equilibrium in formation of Caro's acid, 47-8, 52. Estimation of Caro's acid, 46, 57. of hydrogen peroxide in presence of persulphuric acids, 10, 41. of perborates, 63. 121 122 PER- ACIDS AND THEIR SALTS Estimation of percarbonates, 68. of persulphates, 40-1. Ether, compound with persulphuric acid, FLUOROPERBORATES, 63. Fluoroxypercolumbates, go. Fluoroxypermolybdates, 104. Fluoroxypertantalates, 91. Fluoroxypertitanates, 82. Fluoroxypertungstates, 106. Fluoroxyperuranates, 109. Fluoroxypervanadates, 89. HEAT of solution of ammonium persul- phate, 24. of molybdenum trioxide, 102. of potassium metavanadate,87. persulphate, 24. of sodium perborate, 60. of tungsten trioxide, 105. Hydrochloric acid, influence of, on pre- paration of persulphates, 27. Hydrofluoric acid, influence of, on pre- paration of persulphates, 26-8. Hydrogen peroxide, constitution of, 6. interaction with Caro's acid, 56. persulphates, 38. of crystallization, 2, 3, 64, 69. Hydroperoxides, metallic, 4. salts with per-acids, 4. INDUCED reactions, 34, 41. LANTHANUM, per-acids of, 5. Lead perstannate, 84. persulphate, 30. peruranate, 108. Lithium persulphate, 29. MANGANESE, detection of, 35. estimation in irons and steels, 38. Metallic sulphates, influence of, on pre- paration of persulphuric acid, 18. Metapervanadates, 86, 87. Metapervanadic acid, 85. Monoperoxycarbonates, 72. Monoperoxydicarbonates, 72. NICKEL dioxide, i, 45, perborate, 63. peruranate, 108. Nitrogen, unstable oxide of, 74. ORDER of reaction between potassium persulphate and potassium iodide, 33 Orthopercolumbates, go. Orthopertantalates, 91. Orthopervanadates, 88, 8g. PER-ACID salts, constitution of, 6-8. Per-acids, constitution of, 6-8. definition of, 2. organic, 8. Perborates, 59. constitution of, 59, 64. p erborates, estimation of, 63. hydrolysis of, 61. manufacture and use of, 63. oxidizing action of, 60. 3 erboric acid, 63. i'erborin products, 64, ^ercarbonates, action on potassium iodide of, 67. constitution of, 68, 72. estimation of, 68. preparation of, 65. properties of, 66-8. use of, 73. 3 ercarbonic acid, 68. Perchromates, 5, 92. blue, 93. derived from H CrO 5 , 94. red, 93. relation between, 96. types of, 92. Perchromic acids, 92, 98, gg, 100. constitution of, g7, 98. Percolumbates, go. Percolumbic acid, 89. constitution of, go. Perdicarbonates, 72. Perdisulphuric acid, g, 13. Periodic System, 5. Permolybdates, 103-4. constitution of, 104. Permolybdic acids, 5, 101-3. Permonocarbonates, 72. Permonosulphuric acid, vide Caro's acid. Pernitric acid, 74-5. Peroxides, i. Perphosphoric acids, 77. Perselenates, 58. Perstannates, 83-4. Perstannic acid, 83. Persulphates, 22. action on metals, 37. constitution of, 43-4. estimation of, 40-1. formula of, 41-3. interaction with hydrogen peroxide, 38. manufacture of, 27-g. oxidation of organic compounds with, 36, 39- oxidizing action of, 32-5, 39. properties of, 31-40. reactions in presence of silver salts, 34. 35- tests for, 40. velocity of decomposition of, 31-2. use in analysis, 38, 39. Persulphuric acid, 13-22. anhydride, 10, n. anhydrous, 22. decomposition of, 21. influence of colloidal platinum on, 21. in lead accumulators, 21. properties in solution, 19-21. INDEX 123 Persulphuric acid, pure solutions of, ig. acids, relation between, g. Pertantalates, gi. Pertantalic acid, gi. Pertitanates, 81-2. Pertitanic acid, 79. constitution of, 81. formula of, 79. properties of, 80. Pertungstates, 105, 106. constitution of, 105, 106. Pertungstic acid, 5, 104, 105, 106. constitution of, 105, 106. Peruranates, 107, 108. constitution of, 7, 108, log. Peruranic acid, 5, 107. constitution of, 107. Pervanadates, 86-g. Pervanadic acid, 85. basicity of, 86. - constitution of, 87. Perzirconates, 83. Photography and percarbonates, 73. and persulphates, 36, 38. Potassium perborates, 62, 63, 64. percarbonate, 65-8, 72. percolumbates, go. permolybdates, 103, 104. permonosulphate, 50-1. pernitrate, 76. perselenate, 58. perstannates, 83, 84. persulphate, 13, 24. pertitanates, 81, 82. pertungstate, 106. peruranate, 108. - pervanadates, 86, 88, 8g. perzirconate, 83. Potential of solutions of permolybdic acid, ng. of pertitanic acid, 80. of pertungstic acid, ng. of peruranic acid, ng. Preparation of persulphuric acid, influence of concentration of sul- phuric acid on, 15-7. of current density on, 18. Preparation of persulphuric acid, influence of nature of electrodes on, ig. -of temperature on, 18. Pyropercolumbates, go. Pyropervanadates, 87. RUBIDIUM perborate, 63. percarbonate, 65. permolybdates, 103. persulphate, 2g. SILVER peroxy nitrate, 75. peroxysulphate, 76. salts, action on persulphates, 34, 35. Sodium perarsenate, 78. perborate, sg-62, 64. percarbonates, 68-71. persulphate, 28, 2g. pertitanates, 81, 82. pertungstate, 105, 106. peruranates, 108, log. perzirconate, 83. Sodyl hydroxide, 3, 70. Solubility of ammonium persulphate, 24. of barium persulphate, 30. of peruranic acid, 107. of potassium persulphate, 24. of sodium perborate, 60. Strontium perborate, 63. Sulphur dioxide, influence of, on prepara- tion of persulphates, 27. Sulphuric acid, action on hydrogen per- oxide, ii. electrolysis of, 10. Sulphuryl holoxide, 12. THALLIUM persulphate, 2g. Tests for persulphates, 40. Tetrathionates, formation of, 36. Thorium, hydrated peroxides of, 83. Trithionates, formation of, 36. URANYL perborate, 63. ZINC, hydrated peroxides of, 4. persulphate, 30. ABERDEEN \ THE UNIVERSITY PRESS 9* UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. , 9 17Mar54KV REC'D LD QCT2 REC'D LD JUN261957 REC'D LD JUN EC'D LD AN 1 1 1963 LD 21-100m-9,'47(A5702sl6)476 70 -11AM 5 UNIVERSITY OF CALIFORNIA LIBRARY