THE VOLUMETRIC DETERMINATION OF PEROXYDIPHOSPHATE By Thomas Chulski A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry THE VOLUMETRIC DETERMINATION' OF PEROXYDIPHOSPHATE By Thomas Chulski AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements .for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry Year Approved__ 1953 THE VOLUMETRIC DETERMINATION OF PEROXYDIPHOSPHATE Thomas Chulski' ABSTRACT Tetrapotassium peroxydiphosphate was prepared by the method of Fichter and Gutzwiller (1) by electrolyzing a so­ lution containing 302.2 g. monopotassium orthophosphate, 198 g. potassium hydroxide, 120 g. potassium fluoride, and 0.355 g. potassium chromate per liter using platinum elec­ trodes. Their findings were substantiated except for the claim that the final product was pure white. The tetrapo­ tassium peroxydiphosphate always contained some chromate. Ho method of crystallization could be found that would re­ move the. chromate. Tetrapotassium peroxydiphosphate was precipitated from the electrolyte of Fichter and Gutzwiller by adding solid potassium hydroxide. The product also contained chro­ mate but the precipitation served as an efficient method for the preparation of a rather impure chromate containing tetrapotassium peroxydiphosphate which was found to be suit­ able 4 s .a starting material for the preparation of tetralithium peroxydiphosphate tetrahydrate. A solution of lithium peroxydiphosphate was prepared f r o m a solution of the crude tetrapotassium peroxydiphosphate by precipitating the potassium, phosphate, and fluoride by adding a lithium perchlorate solution. The peroxydiphosphate was precipitated from the solution as tetralithiumperoxy­ diphosphate tetrahydrate by the addition of methanol. material was chromate free. The THE VOLUMETRIC DETERMINATION OF PEROXYDIPHOSPHATE Thomas Chulski A 0.1 normal solution of the tetralithium peroxydi­ phosphate tetrahydrate had a pH of 10.2. It decomposed to the extent of five parts per thousand in 49 days. At lower .pH values the solution was less stable. Peroxydiphosphate was determined by adding excess ferrous ions to a solution of peroxydiphosphate and back titrating the excess with potassium dichromate solution using diphenylamine sodium sulfonate as indicator. Reaction mixtures 0.25 to 1.0 molar in sulfuric acid and 0.25 to 2.0 molar in perchloric acid were found to give satisfactory results. In 1.0 to 2.0 molar hydrochloric acid solutions plausible results were obtained. rous ion was not critical. The amount of excess fer­ The determination could be car­ ried out in the presence of phosphate. The direct potentiometer titration of peroxydiphos­ phate with a ferrous ammonium sulfate solution was accom­ plished. Satisfactory results were obtained when the reactio mixture was 0.25 to u-. 5'0 molar in sulfuric acid or 0.50 molar in phosphoric acid. The equivalence point in the titration of peroxydi­ phosphate with a ferrous ammonium sulfate solution was suc­ cessfully determined by means of the dead stop end point technique. Satisfactory results were obtained if the peroxy­ diphosphate solution was 0.25 to 1.0 molar in sulfuric acid, 0.50 to 1.0 molar in phosphoric acid, or 0.25 to 0.50 molar in perchloric acid. THE VOLUMETRIC DETERMINATION OF PEROXYDIPHOSPHATE Thomas Chulski An iodometric determination of peroxydiphosphate was achieved. The rate of reaction between peroxydiphosphate and iodide ions was found to be influenced profoundly by the acidity of the reaction mixture; a small increase in.. acidity greatly increased the rate of reaction. The deter­ mination was successful both in the absence and presence of phosphate when the reaction mixture was 1.0 molar in sul­ furic acid. Cl) Fichter, F. and Gutzwiller, E., Helv. Chim. Acta., 11, 323 (1928). DEDICATION The author -wishes to dedicate this thesis to all those people who made his education possible. Special mention is made of Dr. Elmer Leininger under whose kind and efficient direction this work was done. TABLE OF CONTENTS PAGE I. II. III. IV. V. VI. VII. VIII. IX. X. INTRODUCTION............■...................... 1 HISTORICAL.............. 2 PREPARATION OF PEROXYDIPHOSPHATE IN SOLUTION .......... PREPARATION OF TETRAPOTASSIUM PEROXYDIPHOSPHATE............ 17 27 PREPARATION OF TETRALITHIUf-iPEROXYDIPHOSPHATE TETRAHYDRATE.. 33 STABILITY OF PEROXYDIPHOSPHATE SOLUTIONS .............. kk A. Tetrapotassium Peroxydiphosphate................... UU £ . Tetralithium Peroxydiphosphate Tetrahydrate........ U6 AN ANALYTICAL STUDY OF THE REACTION BETWEEN PEROXYDIPHOSPHATE AND FERROUS IONS................................ £L AN ANALYTICAL STUDY OF THE REACTION BETWEEN PEROXYDIPHOSPHiTTE A N jj IODIDE IONS IN ACID SOLUTION................ 82 SUMMARY.................................................... ' 97 LITERATURE CITED APPENDIX.............. ............. 100 I02 LIST OF FIGURES FIGURE ' PAGE 1. Per Cent of Phosphate Oxidized to Peroxydiphosphate upon Prolonged Electrolysis............. ........... 22 2. Titration of 25.00 ml. 0.1110 Normal Lithium Peroxydiphos­ phate Solution with 0.1000 Normal Ferrous Ammonium Sulfate Solution.......................... 72 3. Titration of 25.00 rnl. 0.1110 Normal Lithium P'eroxydiphosphate Solution with 0.1000 Normal Ferrous Ammonium Sulfate • ' Solution. Differential Plot............................... I. .INTRODUCTION Phosphorus forms two peroxy acids: peroxymonophosphoric acid, HgPOg } and peroxydiphosphoric acid, H4P2O s . Peroxymonophosphoric acid has been used as a reagent for the detection of manganese (23). No use has been made of peroxydiphosphoric acid or its salts in analytical chemistry. Before applications of either can be made more information is needed on their characteristic reactions as there is but little information on them in the literature. Hence the. purpose of this work was twofold: to work out methods for the determination of peroxydiphosphate, and to characterize it more fully. 2 II. HISTORICAL Petrenko (l5) reported, the preparation of a sodium peroxyphosp-hate by treating sodium phosphate with hydrogen peroxide. Pinerva-Alvarez (16) claimed to have isolated a compound having the formula NaP04 but Rudenko (20) showed, that it was an indefinite mixture of sodium phosphate and sodium peroxide . Rudenko also showed that the material reported by Petrenko was sodium phosphate containing hydrogen peroxide of crystallization. In 1910 Sclmiidlin and. Massini (23) prepared peroxymonophosphoric acid by treating phosphorus pentoxl.de' at a low temperature with hydrogen peroxide of thirty per cent strength or greater. They represented the re­ action as P 305+2H20 2+H20=2H3P05 . The concentrated solution decomposed to the extent of about thirty per cent in two days but diluted solutions retained their activity much longer. Acid solutions were much more stable than neutral, or alkaline solutions. The diluted reaction mixture had strong oxidizing properties. It liberated iodine instantly from iodides. Aniline was oxidized first to nitrosobenzene and then to nitrobenzene. Manganous ion was oxidized, presumably to permanganate, even at room temperature. Heavy metal salts such as silver, iron, and manganese did not precipitate peroxymonophosphates in acid solution but did so in neutral or alkaline solution. Decomposition started, immediately with the formation of orthophosphate and oxygen with ozone being formed in some cases . 3 By treating syrupy pyrophosphoric acid with 30 per cent hydrogen peroxide Schmidlin and Massini obtained the formation of both peroxydiphosphoric, H 4P 20 8, and peroxymonophosphoric acids. The peroxymonophosphoric acid was determined by titrating the iodine liberated immediately from the iodide ion with thiosulfate solution; the peroxydiphosphoric acid by the additional iodine liberated after standing for 2k hours. Before the addi­ * tion of the potassium iodide the excess hydrogen peroxide was determined f by titrating it with potassium permanganate solution. Schmidlin and Massini used the solution of peroxymonophosphoric acid to detect manganese. In this application hydrogen peroxide would interfere; hence it .was destroyed electrolytically. A clay cathode chamber filled with dilute phosphoric acid was used. Electrolysis was carried out until the anolyte no longer decolorized potassium permanganate solution. Schmidlin and Massini stated that the electrolysis of a dilute phos­ phoric acid solution with hydrogen peroxide present formed peroxymono­ phosphoric acid at the anode. No experimental conditions were given. With the exception of the unstable heavy metal peroxymonophosphates Sclimidlin and Massini did not attempt to prepare the salts of either peroxy acid. D'Ans and Friedrich (l) claimed that they independently did the same work as Sclimidlin and Massini (23) but that the latter published first. D*Ans and Friedrich stated that their results were the same as those of Sclimidlin and Massini. Siebald (2l|.) claimed to have formed a peroxyphosphoric acid by the action of 30 per cent hydrogen peroxide on phosphorus oxychloride and that t 1 k a true peroxyphosphate was formed by neutralizing the solution with alcoholic sodium hydroxide. However Husain and Partington (10) were not able to obtain peroxyphosphoric acid by this method. Fichter and Muller (6) developed a method for the simultaneous preparation of peroxymonophosphate and peroxydiphosphate by the electroly­ sis of a solution two molar in respect to dipotassium hydrogen phosphate, four molar in potassium fluoride and containing 0.32 g. potassium chromate per liter. Fichter and Muller studied the influence of temperature and found that the current efficiency increased with decreasing temperature . The range of temperature from 23 to 0° C . was studied. The current efficiency decreased with increasing current density in the range from 1.06 to 0.01 ampere per square centimeter. A current density of 0.03 ampere per square centimeter was taken as a practical optimum current density. Platinum electrodes were used and in this con­ nection they stated that the surface condition of the platinum anode had a great effect on the current efficiency but in an unexplainable manner. Prolonged electrolysis favored the formation of peroxymonophosphate. Ammonium phosphate solutions containing ammonium fluoride and ammon­ ium chromate were employed.; both peroxydiphosphate and peroxymonophosphate were formed. The same result was obtained using rubidium salts. Sodium phosphate was not soluble enough to be employed. Fichter and Muller stated that if a phosphoric acid solution contain­ ing hydrofluoric acid and a little monopotassium hydrogen phosphate were electrolyzed the solution gave a pronounced reaction with manganous ion. No experimental conditions were given. In ail cases the peroxymonophosphate was determined by titrating the iodine liberated immediately from acidified potassium iodide solution, the peroxydiphosphate plus the peroxymonophosphate by the iodine liberated upon standing for 2h pliosphates. hours. No attempt was made to prepare solid peroxy■ Fichter and Rius y Mirb (?) studied the electrolysis further. They repeated much of the work of Fichter and Muller and. obtained essentially the same results. They recommendeo a solution 2.2 molar in respect to fluoride ion, and containing 0.60 g. potassium chromate per liter. The practical optimum current density at 5° C. was 0.03 ampere per square centimeter. Decreasing the concentrations decreased the yield. It was shown that a high cathode density prevented the cathodic re­ duction of peroxydized material and that the chromate served as a cathodic depolarizer. A compartmental cell did not work as efficiently as having chromate present. The mole ratio of potassium to phosphate ions before the addition of I —3 the potassium fluoride was varied from 1.6? K :P04 -f- to i+ K :P04 — 3 . With a ratio of 3.5 K +:P04 J and higher there was no yield at all of peroxydized material. The maximum yield of peroxydized product occurred at a ratio + of 2.3 - 2.0 K :P04 -3 . The yield of peroxydiphosphate went tlirough a maximum at that point while the yield of peroxymonophosphate decreased steadily with' increasing amounts of .potassium . The function of the potassium fluoride in solutions having a potassium to phosphate ration of 2.5 - 2.6 : 1.0 was to- increase the potential on the anode. Oxidation occurred -without fluoride being present but the efficiency was only one half of that obtained with fluoride present. It 6 was claimed that in alkali poor solutions the potassium fluoride also served to suppress the ionization of the monopotassium phosphate thereby ' increasing the yield. Fichter and Rius y Miro were able to prepare pure tetrapotassium peroxydiphosphate, K4P 20 8 . A solution containing 30.22 g. potassium di­ hydrogen phosphate, 19.8 g. potassium hydroxide, 12 g. potassium fluoride, and 0.0395 g. potassium chromate in 100 ml. solution was electrolyzed in a large platinum crucible which served as the anode. of 0.02 to 0.03 ampere per square centimeter was used. was cooled to'0° C. in an ice bath. during electrolysis. A current density The electrolyte The solution was stirred mechanically Electrolysis was carried on until 1.0 ml. of the electrolyte required 18 - .19 ml. of 0.1 normal sodium thiosulfate for the iodine liberated from acidified potassium iodide solution. (Note: for 100 per cent conversion of the phosphate to peroxydiphosphate and with no loss of water during electrolysis 21.7 ml. would be required.) This high content of peroxydiphosphate was attained at the expense of good current efficiency. The electrolyzed solution was evaporated with constant stirring on a water bath. The first crystals to come down were largely unchanged phosphate. After further evaporation the bulk of the peroxydiphosphate precipitated as fine crystals. Two recrystallisations in this manner gave a white product which, when analyzed for phosphorus, potassium, and active oxygen, gave the empirical formula K4P 20 8 . Further recrystalli­ zation was not advised because decomposition to phosphate and oxygen occurred. 7 A solution of the purified product in •water was not stable. A $0 per cent decomposition occurred in five days. When the purified material was dissolved in potassium iodide solu­ tion, iodine X'jas not liberated. Acidification with acetic acid did not cause the liberation of iodine. However when sulfuric acid was used iodine was liberated Immediately'. It'was stated that an alkaline solu­ tion of tetrapotassium peroxydiphosphate precipitated the higher oxides of cobalt, nickel, and lead. Aniline was oxidized to nitro or nitroso- benzene followed by a brown color. Indigo was decolorized. Fichter and Rius y Miro further showed that their product did not contain hydrogen peroxide of constitution but was the salt of a true peroxy acid by its failure to give typical hydrogen peroxide reactions with chromic, per­ manganic , or titanic acid. The tetrapotassium peroxydiphosphate did not form a precipitate with magnesia mixture. In strong phosphoric acid solution the following disproportionation reaction took place: ^4^3^8+H20 — HgPOg+HgPO^. This reaction was suggested as a method for preparing a solution of peroxy­ monophosphoric acid to be used in testing for manganous ion. The electrolysis was believed to take place in two steps as follows: 2K2HP04 = K4P 207+H20 K4P a07+0 = K4P 20 8 The fact that peroxydiphosphate was formed when a solution of potassium pyrophosphate was electrolyzed was taken as verification of the above reactions. 8 Rius y Miro published the information in reference (7) above independent of Fichter in a journal of limited circulation (17) . In further work (18) he showed that peroxymonophosphoric acid reacts with hydrogen peroxide to form oxygen and orthophosphate. Fichter and Bladergroen (U) prepared a solution of peroxymonophos­ phoric acid by treating a 5 molar solution of orthophosphoric acid with fluorine. When the solution of peroxymonophosphoric acid prepared in this manner was added to a dilute manganous sulfate solution a violet color was produced. Schmidlin and Massini (23) interpreted this as the formation of permanganic acid. Fichter and Bladergroen concluded that the color was due to the formation of the very stable manganic phosphate as the five characteristic absorption bands of permanganic acid were never observed in a spectrometer; only a uniform darkening from green to violet was observed. Fichter and Bladergroen suggested that the peroxymonophosphate was formed as the result of two successive reactions. The first yielded peroxydiphosphoric acid by the removal of hydrogen: 2H3P04+F2 = H4P 208+ 2HF. The peroxydiphosphoric acid then hydrolyzed immediately to peroxy­ monophosphoric acid because of the high acid concentration (17): H4P 20 8+H 20 = H 3P05+H3P04 . No solid compounds containing peroxidized phosphorus were obtained. Fichter and Gutzwiller (3) Improved the method of Fichter and Rius y Miro (7) for preparing tetrapotassium peroxydiphosphate. The composi­ tion of the electrolyte and the conditions of electrolysis were essen­ tially the same as those of Fichter and Rius y Miro. However the electroly­ sis was carried out at three intervals so that any peroxymonophosphate 9 formed decomposed, in the alkaline solution. The electrolyte contained 302.2 g. monopotassium orthophosphate, 198 g. potassium hydroxide, 120 g. potassium fluoride and 0.355 g. potassium chromate per liter of solution. Two hundred fifteen milliliters •of this solution was electrolyzed in a platinum crucible. The crucible served as the anode and a bent rapidly rotating platinum wire served as the cathode. The temperature was kept below 1)4° C . by external cooling. The anode current density was 0.023 ampere per square centimeter, but neither the total current nor the total anode surface area was mentioned. After three hours the electrolysis was interrupted and the electrolyte allowed to stand quietly overnight at room temperature. The peroxymono­ phosphate formed during the electrolysis decomposed in the alkaline solu­ tion to orthophosphate, oxygen, and some peroxydiphosphate. The amount of peroxydiphosphate in solution was determined by treating an aliquot with potassium iodide solution acidified with two normal sulfuric acid, heating gently, and titrating the liberated iodine with 0.1 normal sodium thiosulfate solution. A 145.53 per cent conversion of the phosphate to peroxydiphosphate was obtained. After a second electrolysis of two hours duration the resulting solution was allowed to stand overnight. centage conversion was 72.1i5. The per­ A third electrolysis of one horn- duration followed by standing overnight increased the per cent converted to 79.65. By further electrolysis a greater conversion coaid be obtained but only . with a disproportionately large expenditure of current; maximum conversion was stated to be about eighty five per cent. 10 After the third electrolysis the solution was evaporated in a porcelain vessel on a water bath. Mechanical stirring was used. A cur­ rent of air was blown across the surface so that the temperature of the solution did not exceed 80° C . After cooling the mother liquor was removed from the crystals by centrifugation in a porcelain centrifuge. Crystallizing three times in this manner yielded a product 96.U to 99.8 per cent pure. No analytical data were given. Barium peroxydiphosphate tetrahydrate was^prepared by adding a solu­ tion of the tetrapotassium peroxydiphosphate to an aqueous solution of barium chloride. It was insoluble in water, but soluble in acetic, hydro­ chloric, and nitric acids.. The concentrations of the acids in which the barium peroxydiphosphate dissolved were not stated. Zinc peroxydiphosphate tetrahydrate was formed by the interaction of potassium peroxydiphosphate and zinc, sulfate solutions. It was described as being somewhat soluble in water. Anhydrous lead peroxydiphosphate was precipitated from a lead nitrate solution in the same manner. The lead peroxydiphosphate dissolved in warm dilute nitric acid. Analytical data were given for the above salts of peroxydiphosphoric acid. In all cases the experimental values were close to the calculated ones. Fichter and Gutzwiller also prepared rubidium peroxydiphosphate. It was prepared by electrolyzing a rubidium phosphate solution containing fluoride and chromate. The electrolyte was concentrated under vacuum. Unchanged rubidium phosphate precipitated first, then rubidium 11 peroxydiphosphate. It was stated that rubidium peroxydiphosphate was too soluble to prepare easily. No analytical data were given. Ammonium peroxydiphosphate was obtained in solution by the electroly­ sis of an ammonium phosphate solution containing fluoride and chromate. The solution of the ammonium salt was not stable, apparently due to the oxidation of the ammonium ion as nitrite and nitrate could be detected in the solution. Peroxydiphosphate did not form a precipitate with magnesia mixture . A separation from phosphate was suggested but no experimental work was recorded. Rius y Miro (19) showed that peroxymonosulfuric acid could be titrated potentiometrically with a standard solution of ferrous sulfate. A well defined break occurred in the potential curve if manganous ion was present. Peroxydisulfuric acid did not interfere unless its concentration was high. Rius y Miro stated that peroxymonophosphoric acid behaved in a manner entirely analogous to peroxymonosulfuric acid but gave no data. Rius y Miro also titrated peroxymonosulfuric acid potentiometrically in the presence of peroxydisulfuric acid with a sodium sulfite solution containing five per cent of 95 per cent ethanol. He stated that peroxymono­ phosphoric acid could be determined in the same manner but gave no data. The peroxymonophosphoric acid was prepared by the hydrolysis of tetra­ potassium peroxydiphosphate in 50 per cent phosphoric acid. Husain and Partington (10) studied the action of hydrogen peroxide upon lithium, sodium, potassium, rubidium, cesium, calcium, strontium, barium, and thallium phosphates. In all cases no peroxy compound was 12 formed; only phosphates containing hydrogen peroxide of crystallization were obtained. As criteria for phosphates containing hydrogen peroxide of crystallization the following were used: (l) the compounds did not oxidize acidified manganous ion to permanganate ion; (2) acidified aniline solution was not oxidized to nitrosobenzene and nitrobenzene, e.g., a color change from green to brown did not take place; (3) silver nitrate did not give a yellow precipitate which soon turned brown (except when ammonium ion was present); (I4) the compound gave the reactions characteristic of hydrogen peroxide, (A) decomposition of permanganic acid, (B) blue color with chromic acid and ether, (C) yellow color with titanic acid. Phosphorus oxychloride with hydrogen peroxide gave no peroxidized material. The electrolysis of lithium, sodium, or thallium phosphate solution having present fluoride and chromate gave no peroxyphosphate. When rubi­ dium, cesium, or ammonium phosphate was used peroxyphosphate. was formed. A detailed study of the electrolysis of a secondary ammonium phosphate solution containing ammonium fluoride and ammonium cliromate was made. The influence of anodic current density, duration of electrolysis, temperature, and the concentrations of 'ammonium phosphate, fluoride, and chromate was studied. The most favorable condition for the formation of ammonium peroxydiphosphate was found to be the electrolysis of a four molar secondary ammonium phosphate solution three molar in ammonium fluoride and containing 0.32 g. ammonium- chromate per liter. The...optimum current density was O.Oli+6 ampere per square centimeter at 0° C . The longer the electrolysis the greater the yield. 13 The peroxymonophosphate plus the peroxydiphosphate content of the electrolyte was determined by adding an aliquot immediately after the cessation of the electrolysis to an acidified potassium iodide solution and after standing titrating the liberated iodine' with standard sodium thiosulfate solution. The peroxydiphosphate was determined by treating an aliquot similarly after the electrolyte had stood overnight. The acid used to acidify the potassium iodide solution and its concentration in the reaction mixture was not stated, neither was the time of standing. Husain and Partington attempted to prepare pure ammonium peroxydi­ phosphate by evaporating the electrolyzed solution of secondary ammonium phosphate at room temperature in vacuum over anhydrous calcium chloride and over sodium hydroxide . ammonium phosphate. Even the fourth crop of crystals was chiefly The residual mother, liquor was evaporated to dryness. The solid material so obtained contained peroxydiphosphate. No analytical data were given. Saturating the electrolyzed solution with secondary ammonium phosphate did not deposit crystals of ammonium peroxydiphosphate. ■ When freslily electrolyzed solution was treated with 95 per cent ethanol and silver nitrate a precipitate was formed which,, contained both the phosphate and the peroxyphosphates. No precipitate was obtained when the electrolyte was acidified with dilute or glacial acetic acid. Barium carbonate did not precipitate the phosphate. Barium hydroxide precipitated phosphate, fluoride, and chromate; the filtrate contained active oxygen. The filtrate was freed of ammonia Ik by a current of air under reduced pressure over concentrated sulfuric acid. The excess barium was precipitated by passing in carbon dioxide. Husain and Partington concluded that "the perphosphate probably decomposed during the lengthy process." The electrolysis of a concentrated solution of secondary rubidium phosphate or secondary cesium phosphate with chromate present gave both peroxymonophosphate and peroxydiphosphate even without fluoride being present. By evaporating the electrolyzed solution at ordinary temperature and reduced pressure over sulfuric acid a solid product containing active oxygen was obtained. Neither pure rubidium nor cesium peroxydiphosphate could be prepared in a pure condition because of their extreme solubilities. Toennies (27) suspended phosphorus pentoxlde in acetonitrile and to the suspension added a solution of hydrogen peroxide and water in aceto­ nitrile. An energetic reaction took place; peroxymonophosphoric acid being formed. Exbernal cooling was necessary to keep the reaction mixture from exceeding room temperature. Rather small quantities of material were used. In one instance 5.55 millimoles of phosphorus pentoxlde was suspended in 1.5 ml. of acetonitrile. To the suspension was added 12.9b millimoles of hydrogen peroxide (80 to 90 per cent hydrogen peroxide was used) and ij..03 millimoles of water in l.ilO ml. .of acetonitrile. No attempt to isolate the peroxymonophosphate from the reaction mix­ ture was made. Phosphorus pentoxide and hydrogen peroxide did not react in ether or isoamyl alcohol. 15 Schenk and Rehaag (22) prepared the anhydride, P a06 , of peroxydiphos­ phoric acid. By heating phosphorus pentoxide in a stream of oxygen at 'approximately 190° C . and eight millimeters of mercury pressure a violet solid sublimed to the cooler portion of the apparatus. A solution of'the violet colored material did not form a color -with titanic sulfate solution. It liberated iodine 'from"potassium iodide solution. Aniline in acid solu­ tion was oxidized in the cold to nitrosobenzene and nitrobenzene. With manganous sulfate solution a violet color was formed. It was noted that the peroxydiphosphoric anhydride was insoluble in chloroform but that phosphoric anhydride was soluble. Maxted (13) studied the detoxification of platinum hydrogenation catalysts in the hydrogenation of crotonic acid in acetic acid solution. Toxifying agents used were cysteinei thionaphthol, and thiophene. Detoxi­ cating agents studied were potassium persulfate, perchromic acid, and peroxymonophosphoric acid. The peroxymonophosphoric acid was the most efficient detoxifier with all three of the toxifying agents studied. The peroxymonophosphoric acid was made by treating phosphoric pent­ oxide with 30 per cent hydrogen peroxide. The entire reaction mixture was added to the hydrogenation pipette. In further work (llj.) Maxted showed that the double bond of crotonic acid was not attacked by the peroxymonophosphoric acid under the conditions used. Machu (12) credits Degussa (2) with the preparation of pure ammonium peroxydiphosphate, (NH4)4P20 e *2Hs0. A hot saturated solution of ammonium chlorate was added to a cold saturated solution of potassium 16 peroxydiphosphate. The potassium chlorate was filtered.off and the ammonium peroxyphosphate precipitated from the filtrate by the addition of methyl alcohol. Degussa (3) also prepared ammonium peroxyd.iphosph.ate by the electroly­ sis of a triammonium phosphate solution containing a small amount of chromate. By evaporating in vacuum or by cooling below 0° C . pure ammon­ ium peroxydiphosphate was obtained. 17 III. PREPARATION OF PEROXTDIPHOSPHATE IN SOLUTION Fichter and Gutzwiller (5) electrolyzed 215 ml. of a solution contain­ ing 302.2 g. monopotassium orthophosphate, 198 g. potassium hydroxide, 120 g. potassium fluoride, and 0.355 g. potassium chromate per liter in a platinum dish. * The platinum dish served as the anode and a rapidly rotat- ing bent platinum wire served as the cathode. An anode current density of 0.023 ampere per square centimeter was used. They did not specify the anode area, the total current, nor the voltage applied across the electrodes. In duplicating their work 200 ml. of solution of the above composition was electrolyzed in a platinum dish of approximately 210 ml, capacity and lh3 square centimeter inner surface area-. The platinum dish served as the anode and a rapidly rotating bent platinum wire was used as the cathode. A voltage of 7.5 volts across the electrodes gave a current density of 0.015 ampere per square centimeter at the anode. During electrolysis the electrolyte was cooled in an ice bath. Since Fichter and Muller- (6) stated that the surface condition of the ¥ anode had a marked effect on the efficiency of the electrolytic oxidationbut did not explain how the anode should be conditioned to obtain maximum efficiency work was done on the effect of various treatments of the anode in connection with the interruption technique of Fichter and Gutzwiller (5). In the run designated herein as run number 1 the platinum dish was conditioned by polishing it with sea sand and then with a cerium oxide optical polish. Immediately before it was used it was heated to a dull 18 red heat over a Fisher burner. Electrolysis was carried out for three hours using the conditions described above. Upon cessation of the electrolysis the electrolyte was allowed to stand quietly in the platinum dish at room temperature for overnight. The next day a two ml. aliquot of the electrolyte was pipetted into an iodine flask. It was diluted with SO ml. of water; then three grams of potassium iodide, one gram boric acid, and five milliliters of six normal sulfuric acid were added. After standing for at least 12 hours the liberated iodine was titrated with 0.1 normal standardized sodium thiosulfate solution. The amount of tetrapotassium peroxydiphosphate in solution was calculated from the amount of iodine liberated. Electrolysis was carried out again but for only two hours. The electrolyte was allowed to stand overnight as before. The tetrapotassium peroxydiphosphate in a two ml. aliquot of the solution that had been electrolyzed for five hours was determined as above. Electrolysis was carried out again for one hour. The data are given in Table I. In run 2 the platinum dish was conditioned by polishing it again with the cerium oxide optical polish followed by ignition at a bright red temperature. Two hundred milliliters of the electrolyte was processed in the same manner as in run 1. In.run 3 the platinum dish was conditioned by polishing it with the optical polish. The nitric acid was boiled in the dish. In run 1+ the platinum dish was conditioned by boiling concentrated hydrochloric acid in it .■ 19 In runs $ and 6 the platinum dish was conditioned by filling it with concentrated hydrochloric acid and allowing it to stand at room tempera­ ture for three days before using it. TABLE I EFFECT OF VARIOUS TREATMENTS OF THE ANODE UPON THE AMOUNT OF PHOSPHATE OXIDIZED TO PEROXYDIPHOSPHATE Run Number % KH2P04 Oxidized to K4P 208 Hours Electrolysis 3 $ 6 1 22.3 1,0.8 18.5 2 27.9 la .7 51.2 3 15.1 .-2 5 .3 32.7 21, .7 38.2 16.3 5 39.7 57.9 67.3 6 21.3 37.7 16.1 For anode.treatment , see text. - The data in Table I cannot be compared directly to that of Fichter and Gutzwiller as they did not specify the total current or voltage used. However platinum dishes of approximately the same volume of necessity must have approximately the same inner surface area. If the percentages of phosphate converted to peroxydiphosphate as' listed in Table I were to be multiplied by the factor, current density used by Fichter and Gutzwiller: current density used herein, about the same yield of peroxydiphosphate in solution would be indicated as that obtained by Fichter and Gutzwiller except for runs 3 and 5. See Table II. 20 TABLE II AMOUNT OF PHOSPHATE OXIDIZED TO PEROXYDIPHOSPHATE AFTER SIX HOURS ELECTROLYSIS WHEN CALCULATED TO THE CURRENT DENSITY USED BY FICHTER AND GUTZWILLER (5) Run Number 1 % KH2P04 Oxidized to K4P20 8 72.8 ' 2 76.8 3 1*9.1 h 69.5 5 100.0 6 69.2 For anode treatment, see text. It is seen from the data in Table I that conditioning the surface with hot nitric acid lowered the yield of peroxydiphosphate in solution. However of all the other procedures tried none -was significantly better than the others. As a matter of convenience the hydrochloric acid treat­ ment -was taken as the procedure to be followed in the immediate subsequent work. The separation of pure potassium peroxydiphosphate from- the electro­ lyte would be facilitated by having present only a small amount of phosphate. Hence an attempt was made to convert all the phosphate to peroxydiphosphate. The platinum dish described above was used. It was conditioned by filling it with concentrated hydrochloric acid and allowing it to stand for three days before being used. 21 In run 7 , 200 m l . of solution of the above composition was electro­ lysed for six hours. During electrolysis the solution was cooled in an ice bath. A potential of 7.5 volts across the electrodes gave a current of 2.2 amperes. Upon cessation of the electrolysis the electrolyte was • allowed to stand quietly at room temperature in the platinum dish for overnight. The next day a two milliliter aliquot was pipetted into an iodine flask. Twenty-five milliliters of water and six milliliters six normal sulfuric acid were added. Immediately after the addition of the sulfuric a c i d '50 ml. of standardized 0.1 normal ferrous ammonium sulfate solution was added by pipette. After standing for five minutes the excess ferrous ion was titrated witin 0.1000 normal potassium dichromate solution using diphenylamine sodium sulfonate as indicator. The amount of peroxydiphos­ phate in solution was calculated from the amount of ferrous ion oxidized by the two milliliter aliquot of the electrolyte . The percentage of the phosphate oxidized was calculated assuming no diminution in volume.' The . volume of the electrolyte was then measured in a graduated cylinder. Assuming that there had been no mechanical loss of solution and that the diminution in volume was due to the electrolysis of water the percentage of phosphate oxidized to peroxydiphosphate was calculated on the basis of the measured volume of the electrolyte. Electrolysis was carried out again for six hours. This procedure was repeated a number of times. The data are presented graphically in Figure 1. % PHOSPHATE OXIDIZED 100 PER CENT OF PHOSPHATE OXIDIZED TO PEROXYDIPHOS­ PHATE UPON PROLONGED ELECTROLYSIS 50 RUN 7 ASSUME NO DIMINUTION OF VOLUME RUN 7 CALC. ON BASIS OF ACTUAL VOLUME RUN 8 ORIGINAL VOLUME MAINTAINED 0 0 20 HOURS OF E L E C T R O L Y S IS FIGURE I 30 I Y> ro 23 Run 8 -was a parallel experiment but the volume of the electrolyte was measured after each six hour period of the electrolysis and the loss in volume made up by adding water before sampling. The data are presented graphically in Figure 1 and tabulated in the data appendix. Duplicates of nans 8 and 9 verified the results obtained. Fichter and Rius y Mirb (7) and Fichter and Gutzwiller (5) claimed that the electrolytic oxidation of the phosphate to peroxydiphosphate could be made nearly complete. Figure 1 suggests that their data were in error inasmuch as they did not consider the diminution in volume during the electrolysis. From Figure 1 it is seen that the optimum time of electrolysis is 12 hours. Information was sought to explain why more than 70 per cent of the phosphate could not be oxidized. The apparent pH of the solution before electrolysis and during elec­ trolysis was determined with a Beckman pH meter, model G. The initial pH was 11.1 and it did not change during the electrolysis. The potential on the anode va.s followed during electrolysis by using a saturated calomel electrode as a reference electrode. A potentiometer was. used to measure the potential existing between the calomel electrode and the platinum dish. It was found that the measured potential varied with the position of the calomel electrode. However when the calomel electrode was placed very near the platinum dish reproducible readings were obtained. 2k Starting with a fresh portion of electrolyte the measured, potential increased gradually from 1.91 volts to 2.07 volts in six hours. standing overnight the electrolysis was continued. After The initial potential was 1.89 and increased gradually to 2.07 volts in six hours. Decreasing the applied voltage so as to get the original potential on the anode did not increase the yield of tetrapotassium peroxydiphosphate in solution. An attempt was made to decrease the potential on the anode by adding dipotassium phosphate, potassium hydroxide, and phosphoric acid. All in­ creased the anode potential. The yield of peroxydiphosphate was not increased. With the above information in mind the effect of various prior treat­ ments of the anode on the efficiency of the electrolytic oxidation was studied further. In runs 10, 11, 12, 13, and lH, 200 ml. of solution of the composi­ tion used previously was electrolyzed for six hours in the platinum dish as above. .After the electrolysis the volume of the electrolyte was de­ termined by means of a graduated cylinder whereupon the electrolyte was transferred to a Pyrex flask. Enough water, five to ten milliliters, to bring the volume of the electrolyte to the original volume was used to rinse the platinum dish. the Pyrex flask. The rinsings were added to the electrolyte in The platinum dish was thoroughly rinsed with distilled water and filled with concentrated hydrochloric acid. The next day the platinum dish was thoroughly rinsed with distilled water and the electrolysis icarried out again for six hours. The data are given in Table III. 25 Between runs the platinum dish was kept filled with concentrated hydrochloric acid. Runs 15, 16, 17, and 18 were the same as runs 10 through li; except for the treatment of the anode. In these runs the anode was thoroughly rinsed with distilled water after being used and stored inverted in a clean drawer. Runs 19, 20, 21, 22, and 23 were the same as runs 10 through lh and 15 through 18 except that the dish was not rinsed after being used. The ■electrolyte was poured out after the cessation of the electrolysis and the platinum dish kept in a clean drawer. The data in Table III indicate that the three' anode treatments tried are of comparable value. Storing the platinum dish without rinsing it apparently is beneficial to a slight extent and is the recommended anode treatment. No pero;xydiphosphate was formed when a lead dioxide anode was used. 26 TABLE III .'EFFECT OF VARIOUS TREATMENTS OF THE ANODE UPON THE AMOUNT OF PHOSPHATE OXIDIZED TO PEROXTDIPHOSPHaTE Run Number % KH3P04 Oxidized to K4P a08 10 70.It Filled -with 11 69.1 cone. 12 68.8 hydrochloric 13 66.3 acid except lh 69.3 when in use. 13 67.0 Rinsed with 16 70.9 distilled 17 71.6 water and 18 71.2 stored dry. 19 7It.6 Electrolyte 20 72.1 poured out and 21 72.8 dish stored witl: 22 7lt.lt. residual electro 23 70.2 lyte in i t . Anode Treatment 2? I V . PREPARATION OF TETRAPOTASSIUM PEROXZDIPHOSPHATE A . By the Method of Fichter and Gutzwiller Fichter and Gutzwiller (5) prepared pure tetrapotassium peroxydiphos­ phate, K4P 208, by electrolyzing a potassium phosphate solution containing potassium fluoride and potassium chromate and evaporating the electrolyte. The potassium peroxydiphosphate concentrated in the material that separated first. Two recrystallizations•in the same manner gave 9 6 .b to 99.8 per cent pure tetrapotassium peroxydiphosphate. In duplicating their work the following procedure was used. Two hundred milliliters of solution containing 302.2 g. monopotassium phosphate, 198 g. potassium hydroxide, 120 g. potassium fluoride, and 0 .3$5 g. potassium chromate per liter was electrolized in the platinum dish described previously. The platinum dish served as the anode and a rapidly rotating bent platinum wire as the cathode. The temperature of the electrolyte was kept between 8 and 12° C. by cooling in an ice bath. Electrolysis was carried out for three hours. After standing over­ night at room temperature the solution was electrolyzed again for two hours. The third day the electrolysis was carried out for one hour. Approximately 60 per cent of the phosphate was converted to peroxydi­ phosphate . The electrolyte was evaporated in a porcelain evaporating dish on a steam bath. The solution was stirred mechanically so that the tempera­ ture never exceeded 80° C . Evaporation was continued until a considerable 28 amount of crystals was obtained, The material was filtered on a Buechner funnel through a number tiro Whatman filter paper. Filtration was slow. The volume of the mother liquor was U7 ml. The yellow colored crystals were hygroscopic and could not be air dried. They were dissolved in 70 ml. of water and recrystallized in the same manner. . The second batch of crystals contained nearly all the chromate of the first crystals. The volume of the mother liquor was 23 ml. A second recrystallization was carried out. The crystals obtained again contained nearly all the chromatej i.e., the crystallization did not remove chromate efficiently. After drying in vacuo over anhydrous calcium chloride the crystals weighed 20 g. liquor was 9 ml. The volume of the mother ' Based on the amount of potassium phosphate taken the yield was 2? per cent. Based upon the amount of potassium peroxydiphosphate in solu­ tion the yield was Uh per cent. The material was analyzed for active oxygen by titrating the iodine liberated by weighed samples from potassium iodide solution 0.6 normal in sulfuric acid upon 12 hours of standing using sodium thiosulfate solution. The total phosphorus was determined by reducing the peroxydiphosphate with warm 1:1 hydrochloric acid and precipitating the phosphate as magnesium ammonium phosphate. To insure that all the phosphorus was in the ortho­ phosphate form the solution was evaporated to near dryness on a steam bath in the presence of nitric acid prior to the precipitation of the magnesium ammonium phosphate. The magnesium ammonium phosphate was I 29 reprecipitated, ignited to magnesium pyrophosphate, and weighed as such. The data are tabulated below. Calc. for K4P 20 8 Found % P20e 39 J 6 39M ho.99 I4.568 % Active Oxygen h.$68 | A c t l v e 0W n Found . % Active Oxygen Calc, for K4P 20 8 x 1Q0 . ?8 b.619. % ^ ^ The amount of chromate in the purified potassium peroxydiphosphate was not determined. However by a visual comparison it was estimated that half of the chromate originally taken was contained in the potassium •peroxydiphosphate. The amount of chromate present was not accurately de­ termined as the yellow color of the chromate present was sufficient evidence that the material contained too much chromate to be suitable for elucidating the analytical chemistry of peroxydiphosphate. Further work was done to obtain a material that was chromate free. Evaporation of the electrolyte under reduced pressure yielded a yellow crystalline mate2'ial that stuck to the walls of the porcelain evaporat­ ing dish. The mother liquor was colorless indicating that the chromate was also precipitated. Attempts to recrystallize the potassium peroxydiphosphate from dilute methanol or ethanol always led to two liquid phases . The addition of either alcohol to the electrolyzed electrolyte also produced two liquid phases. The non aqueous layers did not contain active oxygen. 30 B. By Precipitation with Potassium Hydroxide It was discovered that the addition of a considerable quantity of solid potassium hydroxide to the electrolyzed solution (see Section III) caused the precipitation of a microcrystalline material that contained active oxygen. When a £0 per cent solution of potassium hydroxide was added no crystals were formed. The addition of less than 12 g. potassium hydroxide per 100 ml. of electrolyte did not cause the formation of a precipitate. The addition of increasing amounts of potassium hydroxide brought down more and more material. However after 30 g. potassium hydroxide had been added per 100 m l .• electrolyte the further addition of potassium hydroxide removed very little additional peroxydiphosphate from the solution. When the potassium hydroxide was added to the electrolyte at room temperature the material3 presumably tetrapotassium peroxydiphosphate, did not settle or filter very well. Cooling was not beneficial. By heat­ ing the electrolyte to $0-$$° C . and adding the solid potassium hydroxide all at once the settling and filtering properties of the precipitate were greatly improved. However even with this improvement standing overnight before filtering was necessary. Wien the precipitation was carried out in an open beaker carbonate was coprecipitated. If the precipitation was carried out in an Erlenmeyer flask stoppered with a rubber stopper the amount of carbonate coprecipitated was minimized. In all cases most of 'the chromate-present in the electrolyte was co­ precipitated. 31 The procedure used for the preparation of tetrapotassium peroxydi•phosphate by this method was as follows. The solution described in Section III was electrolyzed for six hours on each of two successive days. The day after the second period of electrolysis the electrolyte was transferred to a 300 ml. Erlenmeyer flask and heated to 30-35° C . . Sixty grams potassium hydroxide pellets were added as rapidly as possible. The flask was stoppered with a rubber stopper and shaken until all the potassium hydroxide was dissolved. Considerable heat was evolved as the .temperature rose to 70-75° C . After standing for at least 12 hours at room temperature the super­ natant solution was decanted and the crystals collected on an all glass Buechner funnel which had a medium porosity glass frit as the filtering medium. After drying in vacuo over anhydrous calcium chloride until it was free flowing the material weighed 36 g. The active oxygen in it was de­ termined by adding excess standardised ferrous ammonium sulfate to a weighed portion dissolved in one normal sulfuric acid and back titrating the excess ferrous ion with 0.1000 normal potassium dichromate solution using diphenylamine sodium sulfonate as indicator. The total phosphorus was determined as above. The data are given beloxir. Calc. for K4P208 Found 36.U 2 % Active Oxygen % Active Oxygen Found % Active Oxygen Calc., "or K4P 30 8 U.05U x 36.I42 U.053 U0.99 U.619 100 = 87.77 % pure 32 The yield of tetrapotassium peroxydiphosphate based on the amount of phosphate taken was b2 per cent. Based upon the amount of potassium peroxydiphosphate in solution the yield was 72 per cent. Recrystallization in the same manner gave a material that contained only a slightly greater percentage of active oxygen and nearly all of the chromate present in the original crystals. Since a chromate free material uas desired no further crystallizations were carried out. 33 V. PREPARATION OF TETEALITHI'UM PEROXYDIPHOSPHATE TETRAHYDRATE The purification of tetrapotassium peroxydiphosphate was greatlyhampered by its high solubility in water. As previously mentioned attempts to recrystallize it from dilute methanol or ethanol always led to two liquid phases. The possibility that lithium peroxydiphosphate would have solubility characteristics more suitable was- investigated. It appeared that a lithium peroxydiphosphate solution containing but little potassium, phosphate, and fluoride could be obtained by adding lithium perchlorate to a solution of the tetrapotassium peroxydiphosphate obtained in Section IV-B inasmuch as potassium perchlorate, lithium phos­ phate, and lithium fluoride are not very soluble in water. The idea was investigated as follows . The damp crude tetrapotassium peroxydiphosphate obtained by adding 60 g. potassium hydroxide to 200 ml. of the electrolyte.as described above was dissolved in 200 ml. water. a glass electrode was 11.5. The pH of the solution as determined by To the solution of the tetrapotassium peroxy-' diphosphate was added at a dropwise rate lh,.0 ml. of a lithium perchlorate solution which contained of solution. used. g. lithium perchlorate trihydrate per 100 ml. Mechanical stirring and external cooling in an ice bath was A thick slurry was obtained. It filtered rapidly through a number two Whatman filter paper bn a Buecliner funnel. Additional lithium per­ chlorate solution was added to the filtrate until the further addition 3b caused no further precipitation, approximately 10 ml. being required. The additional precipitate -was collected on a Buechner funnel as before. Then 350 ml. methanol -was added at a dropwise rate while the solution was stirred mechanically. A white crystalline material containing active oxygen, presumably lithium peroxydiphosphate, was obtained. The material was not pure white. However if the crystals, were washed on the filter with 1:1 water-methanol solution at bO-b$° C. a pure white . material was obtained. Qualitative tests showed that carbonate was present. The solution of lithium peroxydiphosphate was freed of carbonate by acidifying it to a pH of 6.0 to 6.5 with a 35 per cent perchloric acid solution and remov­ ing carbon dioxide by aeration under reduced pressure. Before adding the methanol the pH was adjusted to a pH of 9.5 - 10.0 by adding solid lithium hydroxide. It was observed that if the pH of the solution was either lower or higher than the indicated range the material precipitated by the methanol was less crystalline in nature. Often times a viscous material was obtained when the methanol was added at room temperature. Precipitation at ice bath temperatures always produced the viscous material. However when the precipitation was carried out at bO-kB0 C. the formation of the viscous material was avoided. By carrying out the precipitation with methanol at I4.O—ij.50 C . a smaller volume of methanol was required than when the precipitation was carried out at room temperature. It was found that adding more than 100 ml. methanol did not materially increase the yield of precipitated material. 35 Microscopic examination showed that it was impossible to air dry the material without some dehydration of the crystals occurring. Dehy­ dration occurred even when the material was dried under conditions of very high humidity and limited air circulation. Drying over partially dehydrated material proceeded too slowly to be practical. After ill weeks of drying in this manner the crystals were still damp. The crystals which had been dried under conditions of high humidity and limited circulation were analyzed for phosphorus pentoxide, active oxygen, lithium oxide, potassium oxide, water, and chromate as follows. Approximately four grams of the air dried crystals were weighed out by difference, dissolved in water, and diluted quantitatively to 500 ml. All determinations except the determination of .water were carried out using aliquots of this stock solution. Immediately after preparing the stock solution the active oxygen was determined by adding an excess of standard ferrous ammonium sulfate solu­ tion to 25 ml. aliquots acidified with sulfuric acid and back titrating the excess ferrous ion according to the procedure developed in Section VIII. "The lithium and potassium were separated from the peroxydiphosphate by means of an ion exchange column using the method of Helrich and Rieman (9) for separating calcium and phosphate. An ion exchange column contain­ ing 19'g. Amberlite IR-100H ion exchange resin in the hydrogen form was used. 15 cm. The diameter of the resin column was 2.5 cm. and the height was Ten milliliters of the stock lithium peroxydiphosphate solution was diluted to 30 ml. The lithium and potassium ions were placed on the 36 resin by allowing the solution to flow into the resin column. The peroxy­ diphosphate ion was eluted by washing the column with 500 m l . water. A flow rate of 0.01 to 0.05 ml./minute/ml. exchanger column (ll); i.e., 1 to 5 ml./minute was used. The last 250 ml. of eluate was neutral to litmus paper. The lithium and potassium ions were then eluted by passing 500 ml. of 1.2 normal hydrochloric acid through the column using the same flow rate as when the peroxydiphosphate was eluted. The eluate containing the lithium and potassium was evaporated to near dryness on a steam bath and then transferred quantitatively to a tared platinum dish. A drop of concentrated sulfuric acid was added and the solution evaporated in a Hildebrand evaporator to fumes of sulfur trioxide followed by ignition at 500° C . for 30 minutes in a muffle furnace. After cooling in a desicator over anhydrous calcium chloride the material was weighed as lithium plus potassium sulfate. In order to separate the lithium from the potassium it was necessary to convert back to mixed chlorides. This was done by precipitating the sulfate with barium chloride and removing the excess barium ion with ammonium carbonate. The ammonium ion was destroyed by adding nitric acid to the final filtrate and evaporating to dryness on a steam bath. The salts were dissolved in water and transferred to a porcelain casserole. One milliliter concentrated hydrochloric acid was added and the solution again evaporated to dryness. To ensure the complete removal of ammonium salts the material was ignited at J4OO0 C. for 30 minutes. The potassium in the mixed chlorides was precipitated as the chloroplatinate (28) and weighed as such. The lithium content was then calculated by difference from the weight of the mixed sulfates. 37 The amount of water in the prepared material was determined by igniting a weighed sample in a stream of oxygen at 500° C . and weighing the liberated water in a Pregl type absorption tube filled with Dehydrite. The stock solution of lithium peroxydiphosphate was tested qualita­ tively for chromate with s-diphenylcarbazide (25). that there was no chromate present. The test indicated That the test was valid in the presence of peroxydiphosphate was determined by adding a small amount of chromate to the solution being tested and observing that the pink color characteristic of the test developed. The analytical data are collected in Table IV. It is shown that the material is tetralithium peroxydiphosphate tetrahydrate. 'As previously stated microscopic examination of the crystals revealed some dehydration had occurred. Yet the analyses show more water to be present than that required for the formation of a tetrahydrate. These facts are not considered incompatible; it merely means that the material was heterogenous. That is , some crystals dehydrated before others were fully dry. Drying over partially dehydrated material was shown to be too slow to be practical. The variable water content greatly influences the calculated formula as shown by the data in Table V which are the same data as in Table IV but calculated to an anhydrous basis. The recommended procedure for preparing hydrated tetralithium peroxy­ diphosphate is as follows. Sixty grams monopotassium phosphate, bO g. potassium hydroxide, 2b g. potassium fluoride dihydrate, and 0.07 g. potassium chromate are dissolved TABLE IV ANALYSIS Ob' LITHIUM PEROXYDIPHOSPHATE PRECIPITATED BY ADDING METHANOL TO A LITHIUM PEROXYDIPHOSPHATE SOLUTION Per cent p 2o5 Run 1*2 Run 1+3 Calculated for Li4P 30 B -l*H20 Li4P 30 8 -5H30 I48.86 1+8 .8.6 1+7 .79 1*7.66 1*8.98 U6.10 5.36 5.36 5.35 5.35 5.52 5.20 20.37 20.67 20.27 20.11 20.62 19.1*2 K30 0.66 0.1*1* 1.67 1.81 h 3o 27.17 26.88 25.91 25.99 21+ .88 29.27 99.79 98.26 98.26 Active Oxygen Li 30 Purity' Summation 100.2 102.1*2 102.21 100.99 100.92 Active oxygen found calc, to anhydrous basis Active oxygen calc, for Li4P 30 8 TABLE V ANALYSIS OF THE LITHIUM PEROXTDIPHOSPH&TE PRECIPITATED BY ADDING METHANOL TO A LITHIUM PEROXYDIPHOSPHATE, SOLUTION, CALCULATED TO AN ANHYDROUS BASIS Per cent P 20s Active oxygen Li20 K 20 Run 1*2 Run 1*3 Calculated for Li4P 20 a Li3P04 67.09 66.82 61*.50 61* .31* 65.63 7.36 7.33 7.22 7.22 7 .35 27.97 27.62 27.36 27.15 27 .1*1* 0.91 0.60 2.23 99.79 98.26 Purity" 100.2 Summation 103.3 102.3 101.3 "Active oxygen f o u n d Active oxygen calc. for Li4P 20 8 2 .1*1*. 98.26 101.2 " - 100 = * TOiritv ^ ^ 61.30 38.70 i|0 in 1^0 ml. water. Considerable heat is evolved. After cooling the solu­ tion, it is filtered through a number two Whatman paper on a Buechner funnel and then diluted to 200 ml. The filtration step is necessary to remove the paraffin introduced by the potassium fluoride. Potassium fluoride is commonly obtained in paraffin lined glass bottles and invari­ ably it is contaminated by the paraffin. ■ The solution is transferred to a 210 ml. volume platinum dish. The solution is cooled in an ice bath during electrolysis. The platinum dish serves as the anode and a rapidly rotating bent platinum wire is used as the cathode. A current of 2.2 amperes is obtained by applying 7.5 volts across the electrodes. Electrolysis is carried out for six hours. A Sargent-Slomin electrolytic analyzer is ideal for this electrolysis. Upon cessation of the electrolysis the platinum dish is covered with a watch glass and the electrolyte allowed to stand in the dish at room temperature for at least overnight. Longer standing is not detrimental as the peroxydiphosphate in this solution is stable for months. Electrolysis is carried out for a second time for six hours. This time the electrolyte is transferred to a rubber stoppered 300 ml. Erlenmeyer flask. The dish is used for the next run without rinsing. After standing for at least overnight the electrolyte is heated to 50-55° C. Sixty grams potassium hydroxide is added. The flask is stoppered and shaken until all the potassium hydroxide dissolves. After standing overnight (longer standing is not detrimental) the supernatant is decanted off and the microcrystalline impure tetrapotassium peroxydiphosphate filtered on a medium porosity sintered glass filter. The mother liquor is discarded. ill The crude tetrapotassium peroxydiphosphate is dissolved in 200 ml. water in a 500 ml. beaker. The solution of crude tetrapotassium peroxy­ diphosphate is cooled in an ice bath while li|0 ml. of a lithium perchlorate solution containing $h g. lithium perchlorate trihydrate per 100 ml. of solution is added at a dropwise rate from a dropping funnel. Mechanical stirring is used. After stirring for 15 minutes the material is filtered through a number two Whatman filter paper on a Buechner funnel. A lithium perchlorate solution of the specified concentration can be prepared by slurrying 52 g. lithium carbonate in 100 ml. water and neutraliz­ ing it with 2i|0 ml. of 35 per cent perchloric acid. It is necessary to remove carbon dioxide by aerating under reduced pressure. After filtration the completeness of precipitation is tested for by adding a little more lithium perchlorate solution. perchlorate solution is added if necessary. Additional lithium Rarely is more, than 15 ml. required. The pH of the solution is adjusted to 6 .0-6.5 by the addition of 35 per cent perchloric acid. solution. A pH meter is used to determine the pH of the Approximately 10 ml. of the acid is required. The solution is aerated by placing it in a 500 ml. filtering flask fitted with a rubber stopper bearing a capillary tube that extends to the bottom of the flask. A water aspirator is used to draw air through the capillary tube and solution for five minutes. The pH of the solution is adjusted to 9.5-10.0 by adding solid hy­ drated lithium hydroxide. Approximately 0.5 g. is required. bZ After adjusting the pH to 9.5-10.0 the solution is cooled to 0-5° C. in an ice bath. Any additional potassium perchlorate that precipitates is removed by filtration. The solution is transferred to a 600 ml. beaker. It is heated to ii0-ii5° C . and placed in a water bath at the same temperature . While the solution is stirred mechamically 70 ml. methanol is poured in from a graduated cylinder. Wo precipitate forms immediately but after stirring for five minutes a considerable quantity of crystals is obtained. Then 30 additional m l . of methanol is added at a rapid dropwise rate from a dropping funnel. The solution is stirred for a total of 15 minutes. The lithium peroxydiphosphate is filtered on a Buechner funnel using a number two Whatman filter paper and sucked as dry as possible.The residual chromate is removed from the crystals by washing them onthe filter with 1:1 water-methanol mixture at ItO-i^0 C. until the wash liquor comes through colorless. Four or five 20 ml. portions of the wash solu­ tion are usually sufficient. The phere of crystals are air dried, until free flowing. Drying in anatmos­ high relative humidity is recommended as the material is easily dehydrated. The yield is 27.0 g. ' The percentage yield of pure product based on the amount of phosphate taken is per cent. Based on a 65 per cent conversion of the phosphate to peroxydiphosphate by the electrolytic oxidation the percentage yield is 61; pei’ cent. Additional but less pure product can be obtained by combining the mother liquor and the methanol-water wash liquor. Sufficient methanol is k3 added to bring the methanol concentration to 50 per cent by volume. While mechanically stirred the temperature of the solution is maintained at 50-55° C . for 30 minutes, About three grams of material is obtained. All the chemicals used should be of sufficient purity to meet the specifications of the American Chemical Society for reagent chemicals. uu VI. STABILITY OF PJi£OXYDIPHOSPHiiTE SOLUTIONS A. Tetrapotassium Peroxydiphosphate The only information recorded in the literature on the stability of peroxydiphosphate is that of Fichter and Rius y -Mirb (7). They.stated that a solution of purified tetrapotassium peroxydiphosphate decomposed to the extent of 50 per cent in five days. Yet both Fichter and Pius y Miro (7) and Fichter and Gutzwiller (5) were able to crystallize tetra­ potassium peroxydiphosphate from aqueous solution. More exact information on the stability of tetrapotassium peroxydi­ phosphate solution was desired. For this work the tetrapotassium peroxydiphosphate prepared by the method of Fichter and Gutzwiller (5) was used even though it was colored yellow by the presence of chromate. Six grams of it was weighed out, dissolved in water, and diluted to 250 ml. to give a 0.05 molar solution. Since there is one atom of -active oxygen per molecule of peroxydiphosphate a 0.05 molar solution is 0.10 normal in a redox reaction. The pH of the solution was 7.5 as determined by the glass electrode. The concentration of the peroxydiphosphate was determined periodically by determining the total equivalent weights of oxidizing material present and attributing all the oxidizing properties of the solution to peroxydi­ phosphate . The amount of oxidizing material present was determined by adding two milliliters of three molar sulfuric acid to a 10 ml. aliquot followed immediately by 20 ml. of standardized 0.1 normal ferrous ammonium b$ sulfate solution which was 0.1 molar in respect to sulfuric acid. After standing for five minutes the excess ferrous ion was back titrated with 0.1000 normal potassium dichromate solution using diphenylamine sodium sulfonate as indicator. VII. Details of this titration are given in Section A 10 ml. burette was used. The number of milliequivalent weights of oxidizing material present in the 10 ml. aliquot was determined by calculating the amount of ferrous ion that had disappeared. Duplicate determinations gave a precision of two parts per thousand. The data are tabulated in Table VI. TABLE VI STABILITY OF 0.1 NORMAL POTASSIUM PEROXYDIPHOSPHATE SOLUTION AT pH 7.5 AT ROOM TEMPERATURE Days Standing Normality of k4p 2o 8 0 0.1227 1 0.1222 2 0.1221 7 0.1222 17 0.1215 The data in Table VI show that the solution of tetrapotassium peroxy­ diphosphate did not decompose to a 50 per cent extent in five days as claimed by Fichter and Rius y Miro (7) . In fact it is perfectly stable for seven days if the decomposition occurring during the first day is U6 disregarded. Over a 16 day period the peroxydiphosphate concentration de­ creased only 0.6 per cent. The pH of the solution did not change during the 1? day period. Fichter and Rius y Miro did not state the concentration of the tetra­ potassium peroxydiphosphate solution that decomposed to such a large extent in five days. Perhaps a more concentrated solution would be less stable but a $0 per cent decomposition of a concentrated solution in five days seems unlikely. In this connection it is worthy to note that in the concentrated alkaline solutions of potassium peroxydiphosphate obtained by electrolysis as described in Section III, the concentration of oxidiz­ ing material did not decrease even after standing for six weeks at extreme summer temperatures. Inasmuch as the material prepared by the method of Fichter and Gutzwiller contained too much chromate to be useful in the study of the reactions of peroxydiphosphate of analytical importance the stability of solutions of various pH's was not studied. B . Tetralithium Peroxydiphosphate Tetrahydrate In the study of the reactions of a particular substance it is desir­ able to have information on the stability of solutions of the material under various conditions . Since the tetralithium peroxydiphosphate pre­ pared in Section V was to be used in elucidating the analytical chemistry of peroxydiphosphate the stability of solutions of it under various con­ ditions was studied. ^7 A liter of 0.1 normal tetralithium peroxydiphosphate solution was made by dissolving 1 6 .£ g. of tetralithium peroxydiphosphate tetrahydrate prepared as described in Section V in water and diluting to one liter. The pH of the solution was 10.2 as determined by the glass electrode. The solution was stored in a glass stoppered Pyrex bottle at room tempera­ ture in the dark. Solutions with various pH values were prepared by adding various amounts of 85 per cent phosphoric acid to solutions prepared in the above manner. The addition of one milliliter of phosphoric acid gave a pH of 7.0; eight milliliters gave a pH of I4.O; and 16 ml. gave a pH of 2.0. A solution 0.1*5 molar in respect to phosphoric acid was prepared by adding 30 ml. of the 85 per cent phosphoric acid to the liter of solution; the pH of the solution acidified in this manner was 1.5. The concentration of the peroxydiphosphate was determined periodically by determining the total number of equivalent wights of oxidizing materi­ als present and attributing all the oxidizing properties of the solution to peroxydiphosphate as was done with the tetrapotassium peroxydiphosphate solution above. A 25 m l . aliquot was taken. It was acidified with six milliliters of three molar sulfuric acid immediately before adding 50 ml. of standard ferrous ammonium sulfate solution. The excess ferrous ion was back titrated with 0.1000 normal potassium dichromate solution. Duplicate determination gave a precision of two parts per thouisand. The data are given in Table V I I . hB TABLE VII STABILITY OF 0.1 NORMAL LITHIUM PEROXYDIPHOSPHATE SOLUTIONS OF VARIOUS pH VALUES AT ROOM TEMPERATURE Days of Standing 10.2 Normality of Li4P20 8 pH of the solution 7.0 I4.O 2.0 0 .1*5 molar h *p o a 0 0.1123 o .1072 0.1090 0.1160 0.1183 1 0.1120 0.1069 0.1086 0.1156 0.1178 2 0.1115 0.1067 0.108)4 0.1156 0.117I4 3 0.1117 0.1068 0.1086 0.1156 0.1166 h 0.1121 0.1068 0.1088 0.115)4 0.1159 3 0.1117 0.1067 0.1086 0.11)4.9 0.1153 8 0.1116 0.1068 0.108)4 0.11I48 0.11)41 lb 0.1118 0.1067 0.108)4 0.11)42 0.1117 21 0.1120 0 .1069 0.108)4 0.1129 0.1092 28 0.1119 0.1068 0.1082 0.1118 0.1068 35 0.1118 0.1066 0.1077 . 0.1102 o.iola h9 0.1118 0 .1065 0.1071 Examination of the data shows that the solutions are stable over the pH range 10.2 to 2.0 for a number of days. At pH 10.2 the concentration of peroxydiphosphate in the solution did not change in 1*9 days. At pH 7.0 the solution was stable for 28 days and in h9 days decomposed only four parts per thousand. Since the experimental procedure used was precise to two parts per thousand for all practical purposes the solution of tetra­ lithium peroxydiphosphate at pH 7.0 was stable for b9 days. At pH it.O the 59 solution was stable for 21 days and in 59 days decomposed five parts per thousand. At pH 2.0 the solution was stable for four days and in 35 days decomposed 53.5 parts per thousand; i.e., 5.35 per cent. The solution of tetralithium peroxydiphosphate which was 0.55 molar in respect to phosphoric acid was not stable at all. In four days it decomposed 19 parts per thou­ sand and in 35 days it decomposed 120 parts per thousand; i.e., 12 per cent. The pH of each solution remained constant during the period of stand­ ing . Thus it is seen that no special procedures or precautions are necessary to obtain a stable 0.1 normal solution of lithium peroxydiphosphate. Merely dissolving the tetralithium peroxydiphosphate tetrahydrate in the requisite quantity of water gives a solution that is perfectly stable. Although a solution of peroxydiphosphate in 0.55 molar phosphoric acid annot be considered stable it is stable enough to carry out reactions in such a medium involving a few hours of standing. o The stability of these same solutions was determined at $0 C. Twenty-five milliliter aliquots were pipetted into 250 ml. iodine flasks and placed in an oven at 50° C. for various periods of time. After cool­ ing the concentration of the peroxydiphosphate ion was determined as above. The data are given in Table VIII. The data in Table VIII shows that tetralithium peroxydiphosphate solutions of pH 10.2 to 2.0 are stable at 50° C. for at least four hours. This points out the possibility of carrying out reactions with peroxydi­ phosphate at 50° C . 5o TABLE VIII STABILITY OF 0.1 NORMAL LITHIUM PEROXYDIPHOSPHAT E SOLUTIONS OF VARIOUS pH VALUES AT '50° C . Normality Li4P 308 •I4H 30 pH of the solution 7.0 li.O Hours at 50° c. 10.2 0 0.1119 0.1068 0.1082- 0.111+2 0.1117 1 0.1121 0.1068 0.1080 0.11^0 0.1106 2 0.1120 0.106lj 0.1077 0.1138 0.1093 k 0.1118 0.1066 0.1080 0 .llitO 0.1097 2.0 0.1+5 molar h 3p o 4 I 51 VII. AN ANALYTICAL STUDY OF THE REACTION BETWEEN PEEOXYDIPHOSPHATE AND FERROUS IONS Wb.en a ferrous ammonium sulfate solution was added to an acidified chromate free tetralithium peroxydiphosphate solution the color of the ferric ion appeared immediately. The possibility of using this reaction analytically was investigated by making a systematic study of the more important variables. An approximately 0.1 normal tetralithium peroxydiphosphate solution was prepared by dissolving 30 g. of tetralithium peroxydiphosphate tetrahydrate, which was made by the method described in Section V, in water and diluting to two liters. Ferrous ammonium sulfate solution approximately 0.08 normal and 0.12 molar in respect to sulfuric acid was prepared by dissolving 32 g. offerrous ammonium sulfate hexahydrate In 300 ml. water acidified with seven milliliters concentrated sulfuric acid and diluting to one liter with water. Potassium dicliromate solution 0.1000 normal was prepared by weighing out 9 .87OI4 g. Mallinckrodt reagent grade potassium dicliromate, dissolving it in water, and diluting the solution quantitatively to two liters. The potassium dicliromate used had been compared to Bureau of Standards standard sample number I36 and found to be the same as the standard sample. The o potassium dicliromate was dried for one hour at 110 weighed out. C. before it was All the chemicals used met the American Chemical Society's specifications for reagent chemicals. 52 The ferrous ammonium sulfate solution was standardized each time it was used by titrating it in the following manner with the 0.1000 normal potassium dicliromate' solution using diphenylamine sodium sulfonate as indicator. Twenty-five milliliters of the ferrous ammonium sulfate solu­ tion was pipetted into a 250 ml. iodine flask. Twenty-five milliliters water and 10 ml. three molar sulfuric acid were added. Then three milli­ liters of 85 per cent phosphoric acid was added for each $0 ml. of volume expected at the end point. Sarver and Kolthoff (21) have shown that the presence of the phosphoric acid is necessary to obtain a satisfactory end point but that the amount present is not critical. Hence even though the ejected amount of ferrous ion were not present a satisfactory end point would be obtained. Three tenths of a milliliter, six drops, of 0.01 molar diphenylamine sodium sulfonate solution was used as indicator. The potassium dicliromate solution was added until the first tinge of purple or violet-blue color appeared. As an indicator blank, 0.05 ml. was sub­ tracted from the volume of dichromate solution used (21). The burettes and pipettes were Kimble Normax grade and were calibrated before use. Corrections were applied wherever necessary. As the first variable studied in the reaction between peroxydiphos­ phate and ferrous ions it was decided to determine the' effect of carrying out the reaction in sulfuric acid of various concentrations. To do so would necessitate the arbitrary standardization of the amount of excess ferrous ion in the reaction mixture as no information was available on the amount of excess ferrous ion necessary to drive the reaction to com­ pletion. Consequently the effect of having present various amounts of. 53 excess ferrous ion was combined, with the study of the reaction in sulfuric acid of various concentrations. This was carried out in the following manner. In one series of determinations five milliliter portions of the lithium peroxydiphosphate solution were pipetted into five different 250 ml. iodine flasks. Sufficient nine molar sulfuric acid to give final sulfuric acid concentrations of 0.25, 0 .5 0 , 1.0, 2.0, and 3.0 molar was added from a burette. Immediately 50 ml. standardized 0.08 normal ferrous ammonium sulfate solution 0.12 molar in respect to sulfuric acid was added by means of a pipette. After standing for at least five minutes but for no more than 25 minutes the excess ferrous ion was back titrated with 0.1000 normal potassium bichromate solution using diphenylamine sodium sulfonate as indicator. The concentration of the peroxydiphosphate was calculated from the amount of ferrous ion that had disappeared attributing all the oxidizing properties of the solution to peroxydiphosphate. Dupli­ cate determinations gave a precision of two parts per thousand. In a second series of experiments the amount of peroxydiphosphate solution was increased to 10 ml. Sufficient nine molar sulfuric acid was added to give the designated acid concentrations. Fifty milliliters of ferrous- ammonium sulfate solution was again added so that the ratio of peroxydiphosphate ion to ferrous ion was halved. The procedure was repeated using 20 and 1*0 ml. of the lithium peroxydiphospliate solution. The data are tabulated in Table IX. 5k TABLE IX DETERMINATION OF PEROXYDIPHOSPHATE WITH FERROUS ION IN SULFURIC ACID SOLUTION Ml. Li4P 20 8 Mole ratio P 30 8~4 Fe*2 5 1.0*17.6 0.1018 0 .1000 0.1006 o.iook 0 .100k 10 1.0:8.8 0.1023 0.1012 o .101k 0 .101k 0.1011 20 l.Otk.h 0.1016 0.1015 0 .1013 0.1010 0.1010 ko 1.0:2.2 0.101k o .1015 o .101k 0.1010 0.1003 Normality Li4P 20 8 Molarity of H 2S04 of the Reaction Mixture 0.25 0.50 1.0 2.0 3.0 To arrive at the concentration of peroxydiphosphate in the solution it is obvious that the 20 different determinations reported in Table IX should not be averaged. The results obtained with reaction mixtures 2.0 and 3.0 molar in sulfuric acid are lower than the rest of the determina­ tions . This is attributed to decomposition of the peroxydiphosphate in the strong acid solution even though the ferrous ammonium sulfate solution was pipetted in immediately after the addition of the acid. The results obtained using five milliliters of sample are not as accurate as the others as in a determination involving the back titration of a large excess of reagent the constituent being determined is obtained by subtracting a large volume from a second somewhat larger volume. Hence a small percentage of error in the two large volumes is reflected in the determination as a proportionately larger error. The anomalous results obtained with the 10 ml. sample at 0.25 molar sulfuric acid cannot be explained. Additional determinations gave a 55 normality of 0.1023, 0.1023, 0.1025, and 0.1020. An error due to faulty calibrations of the volumetric ware would be reflected in the other determinations involving the ten milliliter samples. In the above work the factor of time of standing was not rigidly con­ trolled. Inasmuch as there are no discrepancies in the data of Table IX it can be concluded that the time of standing is not critical. However to determine unequivocally whether the time of standing is a critical factor or not an experiment was designed for that expressed purpose as follows. Twenty milliliters of the lithium perosydiphosphate solution was pipetted into a 250 ml. iodine flask. Sufficient 9 molar sulfuric acid was added to give a final sulfuric acid concentration of 0.25 molar followed immediately by 50 ml. ferrous ammonium sulfate solution added by means of a pipette. The excess ferrous ion was back titrated as expedi­ tiously as possible with potassium dichromate solution. Although there was a short time interval before all the excess ferrous ion was titrated this procedure by definition was designated as zero time of standing. The procedure was repeated but the reaction mixture was allowed to stand for various times before back titrating the excess ferrous ion. Timing was started when the pipette containing the ferrous ammonium sulfate solu­ tion was half, empty and stopped when the first drop of potassium di­ cliromate solution was added. The data are given in Table X. The data in Table X substantiates the conclusion made before that the time of standing before the excess ferrous ion is back titrated is not critical,* at least up to 60 minutes. 56 TABLE X DEI EEMIN.&TION OF PEROXYDIPHOSPHA.TE WITH FERROUS ION IN 0.25 MOLAR SULFURIC ACID AS A FUNCTION OF TH4E Minutes of Standing Normality Li4P 30 8 0 0.1018 10 0.1018 20 0.1018 ho 0.1018 60 0.1017 Mole ratio P 308~4 :Fe+3 = 1.0:1*.!; The effect of substituting hydrochloric acid for the sulfuric acid was studied. The procedure used was the same as was used when the re­ action mixtures were acidified with sulfuric acid. Twelve molar hydro­ chloric acid was added to give the desired acid concentration in the re­ action mixtures . The data are given in Table X I . The data in Table XI were gathered without rigid time control. As when sulfuric acid was used it was desirable to determine whether the time of standing was a factor or not. In addition more information was sought on the reaction in 0.2$ molar hydrochloric acid. The procedure was the same as above except 25 ml. of the lithium peroxydiphosphate solution was used. The data are given in Table XII. The low results with the reaction media 0.25 and 0.50 molar in hydro­ chloric acid can be explained by the reaction between peroxydiphosphate 57 TABLE XI DETERMINATION OF PEROXYDIPHOSPHATE WITH FERROUS ION IN HYDROCHLORIC ACID SOLUTION Mole ratio P 2O 8 4 Ml. Li4p 2°a Normality Li4P 308 Molarity of HC1 of the Reaction Mixture 0.25 0.50 1.0 2.0 1.0:17.6 0.1028 0.1020 0.1030 0.1028 10 1.0:8.8 0.1019 0.1020 0.1017 0.1019 20 1.0 sii.it 0.1011 0.1018 0.1019 0.1020 Uo 1.0:2.2 0.09985 0.1011 0.1018 0.1019 5 TABLE XII DETERMINATION OF PEROXYDIPHOSPHATE WITH FERROUS ION IN HYDROCHLORIC ACID AS A FUNCTION OF TIME Minutes Standing Normality Li4P 2Q8 1.0 M HC1 0.25 M i-ICl 0 0.1008 0.1015 5 0.1012 0.1016 10 0.1012 0 .101U 20 0.1012 0.1015 bo 0.1012 0 .101U 60 0.1013 0.1012 Mole ratio P 308_4:Fe+2 = 1.0:U-U 58 and ferrous Ions being sluggish under such conditions. There is no apparent explanation for the somewhat high and erratic results with the higher acid concentrations. In all cases the end point of the titration was sharp and permanent. The reaction between ferrous and peroxydiphosphate ions was carried out in a medium acidified with perchloric acid. The procedure used was the same as was used previously when the reaction media were acidified with sulfuric and hydrochloric acids. The perchloric acid was added as a 35 per cent solution. The data are given in Tables XIII and XIV. TABLE XIII • DETERlilNiiTION OF PEROXJDlPKOSPHATE WITH FERROUS ION IE p e r c r I o r i c ACID SOLUTION Ml. Li4P 20e Mole ratio P 20e~4 Fe+a Normalit■y Li4P 20e Molarity of HC104 in the Reaction Mixture 1.0 2.0 0.25 0.50 1.0:17.6 . 0 .1002* 0.1016 0.1016 0.101b 10 1 .0 :8.8 0 .1009* 0.1016 0.1016 o .1015 20 1 ,0 :1;.h o .1015 0.1015 0 .1015 0.1016 1|0 1 .0 :2.2 0.1016 0.1016 0.1016 0.1015 £ End point not distinct. The data show that peroxydiphosphate is quantitatively reduced by an excess of ferrous ion in perchloric acid media and that the time interval between adding the excess ferrous ion and the back titration of the excess is not a factor up to 60 minutes of standing. 59 TABLE XIV DETERMINATION OF PEROXXDIPHOSPHATE WITH FERROUS ION IN 0.50 MOLAR PERCHLORIC ACID AS A FUNCTION OF TIME Minutes Standing Normality LI4P gOg 0 0.1015 10 0.1017 20 0 .101U uo 0 .101b 60 0.1017 Mole ratio P 20 8~4 :Fe+3 = 1 .0:U The results obtained using i|0 ml. samples of the peroxydiphosphate are compiled in Table XV. It is seen that approximately the same normality wa.s obtained with reaction media 0 .25, 0 .50, and 1.0 molar in sulfuric acid as was obtained with reaction media 0 .25, 0 .50, 1.0, and 2.0 molar in perchloric acid. Sufficient data were not gathered for a statistical treat­ ment . However the average normality of the cited determinations is 0.1015. The recommended acidities for carrying out the determination are 0.25 to 1.0 molar sulfuric acid solutions or 0.25 to 2.0 molar perchloric acid solutions. The use of hydrochloric acid is not recommended. The reaction between ferrous and peroxydipiiospha.te ions was carried out in phosphoric acid solutions of various concentrations. The same solution of lithium peroxydiphosphate was used as was used for the work 60 TABLE XV DETERMINATION OF PEROXXDIPHOSPHATE WITH FERROUS ION IN SOLUTIONS OF SULFURIC, HYDROCHLORIC, AND PERCHLORIC ACIDS Reaction Medium Normality Li4P 308 0.25 M H 2304 0 .lOllt. 0.50 M 2s o 4 0.1015 1.0 M H 2S04 O.lOlIj. 2.0 M H 2S04 0.1010 3.0 M H 3S04 0.1008 h 0.25 M HC1 0.09985 o vn o ■ M HC1 0.1011 1.0 M HC1 0.1018 2.0 M HC1 o .1019 0.1016 o vn O 0.25 M HC104 M hgio4 o .1016 1.0 M hcio4 0.1016 2.0 M o H O kC o .1015 Sample size, ii0..00 ml. Mole ratio P 208~4 :Fe+ = 1.0:2.2 with reaction mixtures containing sulfuric, hydrochloric, and perchloric acids. The procedure employed was identical to that used, previously. phosphoric acid was added as an 85 per cent solution. in Table XVI. The The data are given 61 TABLE XVI DETERMINATION OF PEROXYDIPHOSPHATE WITH FERROUS ION IN PHOSPHORIC ACID SOLUTION Ml. Li4P2Q8 Mole ratio P^0S“4 t1 +3 • Fe 5 0 £ Normality Li4P20a Molarity of H3P0^. 1.0 1.0:17.6 ' 0.1038 • .0.1052 10 1.0:8..8 0.1027 0.103$ o.iokl 20 l.Otb.h O.lOlh 0.1022 . d.1027 ho 1.0:2.2 0.1017 0.1016 0.1019 2.0 0 .1051* The data show that the reaction between peroxydiphosphate and ferrous ion is not as suited for analytical purposes when carried out in the described manner in a phosphoric acid solution as when carried out in sulfuric, hydrochloric, or perchloric acid solution. Since the back titra­ tion of an excess of a reagent was carried out the high results tabulated in Table XVI indicate a consumption of ferrous ion in addition to that required by the peroxydiphosphate. This could be attributed to the fact that ferrous ion is a stronger reducing agent in phosphoric acid than in the other acids employed.(21) thereby accelerating air oxidation. The data verifies this hypothesis inasmuch as the solutions containing the most phosphoric acid gave the highest results when the amount of excess ferrous ion was kept constant. In addition with solutions containing the' same amount of phosphoric acid but different amounts of excess ferrous ion the ■solution containing the greatest excess gave the highest results. 62 Various experiments were designed to further verify the hypothesis that the air oxidation of the ferrous ion was responsible for the high results reported in Table XVI. Twenty-five milliliter portions of approximately 0.08 normal ferrous ammonium sulfate solution which was 0.12 molar in respect to sulfuric acid were pipetted into 2$0 m l . iodine flasks. Sufficient 85 per cent phosphoric acid was added to give the desired phosphoric acid concentrations. After standing for 15 minutes in diffused daylight six milliliters of three molar sulfuric acid was added, and the concentration of the ferrous ion determined by titration with 0.1000 normal potassium dichromate solution as above. The data are given in Table XVII. The deviations tabulated, are based upon the results obtained with aliquots treated similarly but having no phos­ phoric acid present prior to the actual titration with the potassium di­ chromate solution. The determinations were carried out in duplicate. A precision of one part per thousand was attained. TABLE X/II DECOMPOSITION OF FERROUS AMMONIUM SULFATE SOLUTIONS CONTAINING VARIOUS AMOUNTS OF PHOSPHORIC ACID Molarity H 3PO4 Normality Fe+2 Deviation Parts /Thousand. 0.0 0.08812 -- 0.5 0.08792 2 1.0 0.08780 k 2.0 0.0871*8 8 Time of standing, 15 minutes 63 The data show that ferrous ammonium sulfate solution containing phosphoric acid is not stable. The effect of varying the time of standing was studied by pipetting 25 ml. portions of the ferrous ammonium sulfate solution into 2$0 ml. iodine flasks, adding 1.7 ml. 85 per cent phosphoric acid to make the solution 1.0 molar in phosphoric acid, and titrating the ferrous ion present with 0.1000 normal potassium dicliromate solution after various times of standing. Timing was started when the phosphoric acid was added and stopped when the first drop of potassium dichromate solution had been added. Hence the designation zero time of standing is somewhat in error as there was necessarily a few minutes of time between the addition of the phosphoric acid and the completion of the titration with the potassium dicliromate solution. The data are given in Table XVIII. TABLE XVIII DECOMPOSITION OF FERROUS AMMONIUM SULFATE SOLUTIONS 1.0 H O U R IN PHOSPHORIC ACID DURING VARIOUS TIME INTERVALS Minutes Standing Normality Fe+2 Deviation Parts /Thousand 0 0.08856 c 0.08860 0 10 0.0881*0 2 20 O.O882L1 U 30 0.08820 h — 6k The data show that the air oxidation of ferrous ion in phosphoric acid solution is not instantaneous. Also the rate of reaction is appar­ ently greatly decreased after 20 minutes. The effect of having present both phosphoric acid and lithium ion was studied. Twenty-five milliliter aliquots of the ferrous ammonium sulfate solution were pipetted into 250 ml. iodine flasks. To each was added 10 m l . of 0.1 molar lithium hydroxide solution followTed immediately by the requisite amount of 85 per cent phosphoric acid. After standing' for 20 minutes the concentration of the ferrous ion was determined as before. The data are given in Table XIX. TABLE XIX DECOMPOSITION OF FERROUS AMMONIUM SULFATE SOLUTIONS 0.025 MOLAR IN LITHIUM ION CONTAINING VARIOUS AMOUNTS OF PHOSPHORIC ACID Molarity H3P04 Normality F e +Z Deviation Parts /Thousand 0.0 0.0888k -- 0.5 0.08860 3 1.0 0.0882b 8 2.0 0.08808 10 Time of standing, 20 minutes Comparison of the data in Table XIX with that of Table XVII wherein no lithium ion was present indicates that the presence of lithium ion, in the concentration used, does not have a marked effect on the stability of ferrous ammonium sulfate solutions containing phosphoric acids. 65 The stability of ferrous ammonium sulfate solutions containing lithium ions, ferric ions, and various amounts of phosphoric acid was studied. The procedure was the same as employed above. To the 25 ml. of ferrous ammonium sulfate solution was added an equal volume of 0.1 molar ferric sulfate solution which was 0.2 molar in lithium sulfate. The requisite amount of 85 per cent phosphoric acid was added immediately. The data are given in Table XX. TABLE XX DECOMPOSITION OF FSREOUg AMMONIUM SULFATE SOLUTIONS 0.05 MOLAR IN FERRIC ION AND 0.1 MOLAR IN LITHIUM ION CONTAINING VARIOUS AMOUNTS OF PHOSPHORIC ACID Molarity h 3p o 4 Normality Fe+3 Deviation Parts/Thousand 0.0 0.09008 -- o.5 0.08996 1 1.0 0.08980 3 2.0 0.089hh 7 Time of standing, 20 minutes The data in Table XX indicate that the presence of lithium and ferric ions in the concentration used does not render a ferrous ammonium sulfate solution containing phosphoric acid less stable than when they are absent. The data of Tables XVII, XVIII, XIX and XX support the hypothesis that the determination of peroxydiphosphate in a phosphoric acid medium by 66 the addition of excess ferrous ion and the back titration of the excess gives high results because' of the air oxidation of the ferrous ion. The use of phosphoric acid in the determination as described is not recommended. The determination of peroxydiphosphate by means of the reaction between it and ferrous ion was carried out in the presence of various amounts of phosphate. The reaction media were acidified with sulfuric, hydrochloric, and perchloric acid. The procedure used was as follows. Twenty milliliters of the same lithium peroxydiphosphate used previous­ ly was pipetted into a 250 ml. iodine flask. 2.0 molar monopotassium phosphate solution. respect to sulfuric acid was desired so was added. To it was added 50 ml. of A final solution 0.5 molar in 9 .h ml. of nine molar sulfuric acid Immediately after the addition of the acid 5° ml. of standard­ ized ferrous ammonium sulfate solution was added by means of a pipette. The excess ferrous ion was back titrated as before. The 20 ml. of 0.05 molar lithium peroxydiphosphate solution contained 1.0 millimole of peroxydiphosphate. The total amount of phosphate added was 100 millimoles. Hence the peroxydiphosphate to phosphate ratio was 1:100. That is to say the reaction mixture so prepared was comparable to the solution of a sample of phosphate material containing one mole per cent of peroxydiphosphate. The amount of ferrous ammonium sulfate solution added represented a 2.2 fold excess. Solutions containing smaller amounts of phosphate were prepared by adding smaller amounts of the 2.0 molar monopotassium phosphate solutions. 67 In each case enough water was added to bring the volume to 79 ml. The data are given in Table XXI. The effect of phosphate upon the reaction between peroxydiphosphate and ferrous ions in hydrochloric acid media was also determined. The « procedure used was the same as that above except that 6.h ml, of 12 molar hydrochloric acid was added to make the reaction mixture 1.0 molar in respect to hydrochloric acid. The data are also given in Table XXI. With another series of samples 11+ ml. of six molar perchloric acid was added to each sample. TABLE XXI DETERMINATION Ob PEROXYDIPHOSPHATE IN THE PRESENCE OF PHOSPHATE Normality Li4P208 1 .0 :0.0 Mole ratio P 30 8“4:P04. 3 1 .0:20 1 .0 :1+0 1^0:80 1 .0:100 0.50 M H 3S04 0.1010 0.1012 0 .1012 0.1012 0.1017 1.0. M HC1 0.1012 0.1010 0.1009 0 .1009 0.1013 1.0 0.1010 0.1011 0 .10lU o.ioiA 0 .1011+ Reaction Medium M HC10*. Mole ratio P 20 8“4 :Fe+2 = 1 .0:U.k The determination of peroxydiphosphate in the presence of potassium phosphate cannot be compared directly to the determination in reaction media acidified with phosphoric acid by comparing the data of Tables XVI and XXI. In Table XVI the concentration of the phosphate is stated but the ratio 66 between the phosphate and peroxydiphosphate is not. In Table XXI the converse is true. However in Table XXII both the molar concentration of the phosphate in the reaction mixture and the ratio existing between the peroxydiphosphate and the phosphate are given. Data was taken from Table XVI which utilized a mole ratio of peroxydiphosphate ion to ferrous ion of 1.0 to U.U and from Table XXI which utilized reaction media 0.50 molar in sulfuric acid. TABLE XXII DETERMINA.TION OF PEROXZDIPHOSFHATE■IN THE PRESENCE OF PHOSPHATE Cone. P04~3 Mole .ratio P 20 a~4 :P04"3 0.00 M 1.0:0.00 0.5 M h 2s o 4 0.1010 0.15 1.0:20 0.5 M II2S04 0.1012 0.31 1.0:37 0.5 H3P04 0.1012 0.50 1.0:1*0 0.5 M H 2304 0.101k 0.62 1.0:80 0.5 M h 3s o 4 0.1012 0.77 1.0:100 0.5 h 2s o 4 0 .1011| 1.0 1.0:75 1.0 M H 3P04 0.1022 2.0 1.0:162 2 .0 M H3P04 0.1027 Acid Cone. m m Normality ■^d4P208 Mole ratio F 30 8"4:Fe+2 = l.Oth.k It is seen that the concentration of phosphate in the reaction mixture has a greater influence than the value of the ratio between the I } 69 peroxydiphosphate and phosphate. Hence in the determination of peroxy­ diphosphate in the presence of phosphate the reaction mixtures should be diluted until the phosphate concentration is 0.9 molar or less. The data of Table XXI was obtained without rigid control of time. Hence a series of determinations were made in which the time of standing before the back titration of the excess ferrous ion was the variable studied. Twenty milliliters of the lithium peroxydiphosphate solution was pipetted into a 290 ml. iodine flask. Twenty-five milliliters of 2.0 molar monopotassium phosphate, 29 ml. water, and 9.k ml. of nine molar sulfuric acid were added in the order named. Then 90 ml. of standardized ferrous ammonium sulfate solution was added by means of a pipette. For zero time of standing the excess ferrous ion was back titrated as expedit­ iously as possible. In the other cases timing was started when the pipette containing the ferrous ammonium sulfate solution was half empty and stopped when the first drop of potassium dichromate had been added. The data are given in Table XXIII. The procedure was repeated acidifying with 12 molar hydrochloric acid and six molar perchloric acid instead of the sulfuric acid. The data are also given in Table XXIII. It is seen from the data of Table XXIII that the time of standing after the addition of the ferrous ammonium sulfate solution is not critical except when the reaction medium is 1,0 molar in hydrochloric acid. With a reaction medium 1.0 molar in hydrochloric acid it is necessary to wait five minutes before back titrating the excess ferrous ion. 70 TABLE XXIII DETERMINATION o f p e r o x x d i f h o s p h a t e hi t h e p r e s e n c e PHOSPHATE AS A FUNCTION OF TIME Reaction Medium 0 Normality Li4P 30 a Minutes Standing 10 20 of 60 bo 0.5 M H 2S04 0.1010 0.1010 0.1010 0.1010 0.1012 1.0 M HC1 * O.IOO7 0.1010 0.1010 0.1005 1 .0 M H C 104 0.1008 0 .1011 0.1010 O.lOll* 0.1013 Mole ratio P 30s‘"4 :P04"3 = 1.0:1*0 Mole ratio P 30 a‘"4 :Fe+2 = 1.0:1*.1* * No end point obtainable. gave O.lOO^ normal. Five minutes of standing The above work shows that the reaction between peroxydiphosphate and ferrous ions is rapid. This fact suggested the direct-potentiometric titration of peroxydiphosphate with standardized ferrous ammonium sulfate solution. Twenty-five milliliters of approximately 0.1 normal lithium peroxydi­ phosphate solution was pipetted into a 150 ml. beaker. Seven tenths of a milliliter of nine molar sulfuric acid was added to make the solution 0.5 molar in respect to sulfuric acid. A Beckman saturated calomel electrode was placed in the solution to serve as a reference electrode. As the indicating electrode a Beckman platinum electrode was used. The calomel electrode was connected to the negative binding post of a Sargent po­ tentiometer. The potential existing between the two electrodes was 71 determined by means of the potentiometer. After adding one milliliter of standardized 0.1 normal ferrous ammonium sulfate solution 0.12 molar in sulfuric acid from an Exax quality burette bearing an offset tip, the potential was again determined. The ferrous ammonium sulfate solution was added in one milliliter increments until a rapid change in potential was noted; then it was added in one tenth milliliter increments. During the titration the solution was stirred with a magnetic stirrer. It was found that the magnetic stirrer did not affect the potentiometer readings. The data are shown graphically in Figure 2 and tabulated in the Data Appendix. The titration was repeated except that the ferrous ammonium sulfate solution was added as rapidly as possible to the acidified lithium peroxy­ diphosphate solution to within one milliliter of the predetermined end point. It was then added in one tenth milliliter increments . The data are shown graphically in Figure 2 and tabulated in the Data Appendix. The equivalence point in the titration was taken as the volume of 0.1000 normal ferrous ammonium sulfate solution required to reduce the peroxydiphosphate in a 25 ml. aliquot of the solution when an excess of the reducing agent was added and the excess back titrated with potassium bichromate solution using diphenylamine sodium sulfonate as indicator. The reaction medium was 0.50 molar in respect to sulfuric acid. It was • found that 25.00 ml. of the lithium peroxydiphosphate solution required 27.75 ml. of 0.1000 normal ferrous ammonium sulfate solution. A different platinum electrode was used and the potentiometric titra­ tion carried out again by adding the ferrous ammonium sulfate solution in one milliliter increments. The data are shown graphically in Figure 2. ELECTRODE 100 800 SLOW TITRATION, ELECTRODE I 500 SLOW TITRATION, ELECTRODE 2 RAPID TITRATION, ELECTRODE I EQUIVALENCE POINT 2 7.75 ML. MV. Pt VS. SAT. CALOMEL TITRATION OF 25.00 ML. O.IIION LITHIUM PEROXY­ DIPHOSPHATE SOLUTION WITH 0.1000 N FERROUS AMMONIUM SULFATE SOLUTION 200 20 ML. 0 . 1 0 0 0 N FERROUS AMMONIUM SULFATE FIGURE 2 40 30 SOLN. —o ro 73 It is seen from Figure 2 that the shape of the curve depends upon external factors . The reaction between peroxydiphosphate and ferrous ions would be expected to be irreversible which is borne out by the data plotted in Figure 2. That is, the curve results from the combination 'of two half cell reactions; (l) the irreversible reduction of peroxydiphos­ phate and (2) the reversible oxidation of ferrous ion and as such would be expected to be nonreproducible. However .if the cliange in potential per one tenth of a milliliter of ferrous ammonium sulfate solution, A E/ Zi ml., is plotted against total volume of ferrous ammonium sulfate solution the end point can be readily obtained as shown by Figure 3. The precision was t 0.05 ml. The titration was repeated with reaction media of 0.50, 1.0, 2.0, and 3.0 molar sulfuric acid. Both the slow, i.e., adding the ferrous ammonium sulfate solution in one milliliter increments, and the rapid titration were carried out at each acidity. The data are given in Table XXIV. . In no case was exactly 0.1000 normal ferrous ammonium sulfate solu­ tion used. Since a solution of ferrous ammonium sulfate is not stable all volumes were calculated to equivalent volumes of exactly 0.1000 normal splution. Thus the data are readily compared. Reaction media of phosphoric, hydrochloric, and perchloric acid were tried, Wien hydrocliloric acid was used the odor of chlorine was detected. The data are also given in Table XXIV'. If the volume of ferrous ammonium sulfate solution required to react with 25.00 ml. of the lithium peroxydiphosphate solution determined by - A E /A M L . Pt VS. SAT. CALOMEL ELECTRODE 130 TITRATION OF 2 5 .0 0 ML. O.IIION LITHIUM PEROXY­ DIPHOSPHATE SOLUTION WITH 0.1000 N FERROUS AMMONIUM SULFATE SOLUTION SLOW TITRATION, ELECTRODE I SLOW TITRATION, ELECTRODE 2 80 - RAPID TITRATION, ELECTRODE I 30 - EQUIVALENCE 2 7 .7 5 ML. POINT 0 u 20 25 30 35 --0 4=r ML. 0 .1 0 0 0 N FERROUS AMMONIUM SULFATE SOLN. FIGURE 3 75 TABLE XXIV POTENTIOMETRIC DETERMINATION OF THE END POINT IN THE TITRATION OF 25.00 ML. OF LITHIUM PEROXYDIPHOSPHATE SOLUTION WITH O.IOOON FERROUo AMMONIUM SULFATE SOLUTION Reaction Medium Ml. O.IOOO N Ferrous Ammonium Sulfate Solution to Reach the End Point Slow titration Rapid titration 0.25 M li2S04 27.75 27.70 27.80 0.50 M H 2S04 27.70 27.65 27.90 1.0 M H 2S04 27.15 27.80 2.0 M H 2S04 26.55 27.80 3 .0 M H s S04 25.75 21.hS 0.50 M H 3P04 27.75 27.75 27.75 27.85 1.0 M Il3P04 27.65 27.65 27.85 27.85 2.0 M H 3P04 27.85 27.85 27.85 27.95 0.25 M HC1* 25.95 26.30 0.50 M HC1# 25J40 26.10 0.25 M HC104 .17 .ho 0.50 M HC104 27.65 27.85 1.0 M HC104 - 27.55 27.85 2.0 M HC104 27.10 27.55 Equivalence point at 27.75 milliliters * Odor of chlorine detected adding an excess of the reducing agent and back titrating the excess with potassium dichromate solution is taken as the volume required to reach the equivalence point the data show that in the potentioraetric titration the 76 sulfuric acid concentration should be between 0.25 and 0.$0 molar for results that check to within three parts per thousand of the designated standard. However, if the amount of peroxydiphosphate present is approxi­ mately known, sulfuric acid concentrations up to 2 .0 molar can be used by adding the ferrous ammonium sulfate solution rapidly until the end point is nearly reached and then finishing the titration slowly. With acid concentrations greater than 0.50 molar with the slow titration and greater than 2.0 molar with the fast titration low results are obtained, presum­ ably due to the decomposition of the peroxydiphosphate because of the high acidity. If the criterion described above is applied to the titrations utiliz­ ing phosphoric acid satisfactory results are obtained only when the phos­ phoric acid concentration is 0.5 molar. Greater acid concentrations give high results. When hydrochloric acid is used low results are obtained because of the oxidation of the chloride ion. When the reaction media acidified with perchloric acid are used, satisfactory results are not obtained with acid concentrations from 0.25 to 2 .0 molar . In summary the end point in the titration of peroxydiphosphate with ferrous ammonium sulfate solution can be satisfactorily determined potentiometrically if the peroxydiphosphate solution is 0.25 to 0.50 molar in sulfuric acid or 0.50 molar in phosphoric acid. If the amount of peroxydiphosphate is approximately known sulfuric acid concentrations up to 2.0 molar can be used. 77 When a solution of peroxydiphosphate is titrated with ferrous ammonium sulfate there are essentially no ferrous ions in the solution being titrated until after the equivalence point has been reached. The possibility of detecting the appearance of ferrous ions by means of the dead stop end point technique (26) was. investigated as follows. Platinum electrodes identical to those used by Stone and Scholten (26) were employed. Also a Fisher ELecdropode was used to apply the potential'across the electrodes and to observe the current flowing in the. electrode circuit. Twenty-five milliliters of lithium peroxydiphosphate solution was pipetted into a 1$0 ml. beaker. The platinum electrodes were introduced and a potential of fifty millivolts was applied across the electrodes. The sensitivity control of the Elecdropode was set at IX. The peroxydi­ phosphate solution was made 0.25 molar in sulfuric acid by adding 0,7 ml. of nine molar sulfuric acid. Standardized 0.1 normal ferrous ammonium sulfate solution was added at a rapid rate until the galvanometer indi­ cated that a current was flowing in the electrode circuit. The addition of the ferrous ammonium sulfate solution was then ceased until the system reached equilibrium which was indicated by the galvanometer reading being the same as before the titration was started. The titration was carried out in this manner until a permanent deflection of the galvanometer was obtained. As the end point was approached smaller and smaller volumes of the ferrous ammonium sulfate solution caused a temporary deflection of the galvanometer. However even when very close to the end point equilibrium was attained rapidly. 78 During the titration the solution was stirred by a motor stirrer. Stirring with a magnetic stirrer caused a deflection of the galvanometer which was proportional to the speed of the stirrer. The motor stirrer did not affect the galvanometer. As before the equivalence point in the titration was taken as the volume of 0.1000-normal ferrous ammonium sulfate solution required to reduce the peroxydiphosphate in a 25 m l . aliquot of the solution when an excess of the reducing agent was added and the excess back titrated with potassium dichromate solution. It was found that 25.00 ml. of the lithium peroxydiphosphate solution required 26.95 ml. of 0.1000 normal ferrous ammonium sulfate solution. The titration was repeated with the reaction media acidified with various amounts of sulfuric, phosphoric, and perchloric acid. The data are given in Table XXV. In all cases the end point was determined without recourse to graphical methods . The shai*pness of the end point is shown by the data in Table XXVI. If the equivalence point is taken as 26.95 ml. of 0.1000 normal ferrous ammonium sulfate solution the determination of the end point by the dead stop technique is satisfactory for sulfuric and solutions ranging from 0.25 to 1.0 molar. The criterion taken for satisfactory results is that the end point in the titration does not differ by more than tiiree parts per thousand from the equivalence point. On the same basis reaction media 0.50 to 1.0 molar in phosphoric acid and 0.25 to 0.50 molar in perchloric acid are satisfactory. 79 TABLE XXV DETERMINETION OP THE END POINT IN THE TITRATION OF LITHIUM PEROXYDIPHOSPHATE WITH FERROUS AMMONIUM SULFATE SOLUTION BY MEANS OP THE DEAD STOP END POINT Titration Medium M l . of 0.1000 N Ferrous Ammonium Sulfate Solution Required for 25 ml. Li4P 20 8 0.25 M h 2so4 26.90 26.90 0.50 M H 2S04 26.88 26.86 1.0 M H 2so4 26.86 2.0 M h 2so4 26.82.1. 3.0 M h2so4 26.67 0.50 M h 3po4 26.97 1.0 M h3po4 26.97 2.0 M h3po4 27.05 27.03 0.25 M hcio4 26.91 26.87 0.50 M hcio4 26.90 26.83 1.0 M hgio4 26.81+ 2.0 M hcio4 26.66 26.9k Equivalence point at 26.95 milliliters. As with the direct potentiometric determination of the end point higher concentrations of sulfuric or perchloric acid gave low results which substantiates the hypothesis that decomposition of the peroxydiphosphate occurred. 80 TABLE XXVI SHARPNESS OF THn DEAD STOP END POINT IN THE TITRATION OF LlTHIUi-i P pJROXIDIPHOSPHh TE WITH FERROUS AMMONIUM SULFATE Titration Medium Ml. 0.1000 Fe+S Solution Galvanometer Reading 0.25 M H 2S04 26.85 26.90 3 70 1.0 M fl2S04 26.85 26,87 6 h2 0.50 M H3P04 26.95 26.97 3 27 1.0 h H3P04 26.97 27.03 3 12 0.25 M HC104 26.87 26.90 5 81 2.0 26.6I1 26.66 M HC104 5 100 Examination of the data in Tables XXIV and XXV reveals that when re­ action mixtures containing phosphoric acid gave unsatisfactory results it was because too much of the ferrous ammonium sulfate solution was used to reach the end point, i.e., additional ferrous ion was oxidized. The phenomenon can be explained by the ferrous ion being a stronger reducing agent in the presence of phosphoric acid than in sulfuric or perchloric acids thereby accelerating the air oxidation of the ferrous ion during the titration. This hypothesis is substantiated by the fact that the titration was satisfactory when the concentration of phosphoric acid was low (21). 81 Although much the same results were obtained with the potentiometric determination of the end point as with the dead stop technique} the dead stop technique is recommended as equilibration of the electrode system was attained much more rapidly. In addition the dead stop technique gives the end point directly; graphical methods are not necessary. 82 VIII. AN ANALYTICAL STUDY OF TH d REACTION BETWEEN PEROXYDIPHOSPHATE AND IODI.UE IONS IN ACID SOLUTION The liberation of iodine by peroxydiphosphate from acidified potassium iodide solution was used by previous workers for the estimation of peroxydi­ phosphate . However an analytical study of the reaction had never been made . Fichter and Gutzwilier (5), used two normal sulfuric acid to acidify the reaction mixture but did. not state how much acid was used. None of the other authors even stated what acid they used to acidify the solution. The time of standing necessary to complete the reaction had never been studied. With the exception of Fichter and Gutzwilier (5) previous workers allowed the reaction mixture to stand overnight without ascertain­ ing if a shorter period of time would be sufficient. Instead of allowing the reaction mixture to stand overnight Fichter and Gutzwiller gently warmed the reaction mixture and then titrated the liberated iodine. Their procedure cannot be considered good analytical practice because of the probability of losing iodine by reason of its high vapor pressure. Also none of the previous workers considered the possibility of the formation of iodine by air oxidation of the iodide. All of these factors have been considered in this study. The reaction mixture was acidified with various amounts of various acids and. allowed to stand for various periods of time before titrating the liberated iodine. It was desirable to be able to relate the results obtained by determin­ ing peroxydiphosphate by the titration of the iodine liberated by it from 83 acidified iodide solution to the results obtained by the reduction with ferrous ion as described in Section VII. In order to better relate the results obtained by the two methods the sodium thiosulfate and sodium arsenite solution used to titrate the liberated iodine were standardized with the same potassium diehromate as was used in Section VII. Approximately 0.1 normal sodium thiosulfate solution was made bydissolving 2b.65 g. sodium thiosulfate pentahydrate in 100 ml, water containing 0.1 g. sodium carbonate and diluting this solution to one liter.Approximately 0.1 normal sodium arsenite solution was prepared by dissolving k.9b60 g. arsenious oxide previously dried at 110° C. for one hour in a solution composed of three grams sodium hydroxide dissolved in 10 ml. water (8) . The solution was diluted to 1+00 ml. Three drops of phenolphthalein solution was added and the alkali neutralized by adding three molar sulfuric acid until the indicator was decolorized. T+^enty-five grams sodium bicarbonate was added and the solution diluted quantitative!y to one liter. Approximately 0.1 normal iodine solution was prepared by dissolving 6.5 g. iodine in a solution of 10 g. potassium iodide in 20 ml. water and diluting with water to 500 m l . All the chemicals used met the American Chemical Society*s specifica­ tions for reagent chemicals. The sodium thiosulfate solution was standardized by means of the potassium diehromate solution as follows. Twenty-five milliliters of the diehromate solution was introduced into a 250 ml. iodine flask by means of a pipette. Two grams potassium iodide and four milliliters 1:1 8i[ hydrochloric acid was added. The flask was stoppered and a freslily prepared concentrated potassium iodide solution was placed in the gutter of the flask. The flask was allowed to stand in the dark at room tempera­ ture for one hour. After the period of standing the potassium iodide solution in the gutter was allowed to run into the flask. After diluting with 100 m l . water the iodine was titrated with the sodium thiosulfate solution using three milliliters of a 0.5 per cent soluble starch solu­ tion as indicator (29) . Triplicate standardizations agreed to one part per thousand. , The iodine solution was standardized by titrating 25 ml. aliquots acidified with two milliliters 1:1 hydrochloric acid with the sodium thio­ sulfate solution using starch solution as indicator. The sodium arsenite solution was then standardized by titrating 25 ml. aliquots of the iodine solution in the presence of sodium bicarbonate with the arsenite solution. After the sodium arsenite was standardized a 25 ml. aliquot of the iodine solution was titrated with the sodium thiosulfate solution to insure that there had been no change in the concentration of the iodine solution. By this method the normality of the sodium arsenite solution.was found to be 0.09905. . The same sodium thiosulfate and sodium arsenite solutions were used tliroughout. In studying a reaction as a function of time it is always desirable to be able to quench the reaction at will. The reaction between peroxydi­ phosphate and iodide ions could be quenched merely by neutralizing the acid present provided that the peroxydiphosphate- would not liberate iodine from a neutral potassium iodide solution. Hence it -was necessary to determine if peroxydiphosphate and iodide ions would react in a neutral solution. A neutral solution containing peroxydiphosphate and iod.ide ions was prepared as follows. Ten milliliters of 0.1 normal tetrapotassium peroxy­ diphosphate solution was pipetted into a 2f>0 ml. iodine flask. Three grams sodium bicarbonate, 10 ml. -water, one milliliter three molar sulfuric acid, and three grams potassium iodide were added in the order named. The flask was stoppered and a freshly prepared concentrated potassium iodide solution was placed in the gutter of the flask. After standing for four hours in the dark the potassium iodide solution in the gutter was allowed to run into the flask. was added. Three milliliters of a 0.5 per cent starch solution A very faint blue color was observed. The color was too faint to warrant titration of the iodine present. The experiment was repeated but the time of standing was increased to 2k hours. In this instance the color of the starch-iodine complex was much deeper. One tenth of a milli- , . liter of 0.1000 normal arsenite solution was required to discharge the color. At the most 10 minutes is required for the titration of iodine with arsenite solution using starch solution as indicator. Since peroxydiphos­ phate does not liberate an appreciable amount of iodine from a neutral potassium iodide solution in 10 minutes the extent of reaction between the two ions in acid solution for a given period of time can be determined by neutralizing the solution with sodium bicarbonate and titrating the liberated iodine with neutral arsenite solution provided that peroxydi­ phosphate and arsenite do not react to an appreciable extent in 10 minutes in a neutral solution. 86 Experiments were carried out to determine if peroxydiphosphate and arsenite react in a neutral solution. Ten milliliters 0.1 normal tetrapotassium peroxydiphosphate solution was pipetted into a 250 ml. iodine flask. Three grams solid sodium bi­ carbonate was added and. 20 ml. of 0.09905 normal sodium arsenite solution 0.06 molar in respect to sulfuric acid was added by means of a pipette. A blank x-@-s prepared by replacing the peroxydiphosphate solution with -10 ml. water. After standing for one hour the amount of arsenite remaining was determined by titrating with 0.1 normal iodine solution using a 0.5 per cent starch solution as indicator. The blank and the solution contain­ ing the peroxydiphosphate required the same volume of iodine solution in­ dicating that no reaction had taken place between the arsenite and peroxy­ diphosphate . The experiment was repeated with the time of standing in. • creased to 2k hours. In 2h hours one per cent of the peroxydiphosphate -f was reduced by the arsenite. Thus it is seen that the extent of the reaction between peroxydiphos­ phate and iodide ions in an acidified potassium iodide solution for various times of standing can be determined by neutralizing the reaction mixture with sodium bicarbonate after the period of standing and titrating the liberated iodine with neutral arsenite solution. The reaction between peroxydiphosphate and potassium iodide was studied in reaction mixtures acidified with sulfuric} phosphoric, and perchloric acids. The reaction was not studied in the presence of hydrochloric acid because of the oxidation of the chloride ion by the peroxydiphosphate as previously noted. Also the effect of having phosphate present was studied. 87 The effect of varying the time of reaction and the concentration of sulfuric acid was ascertained as follows. Twenty-five milliliters of lithium peroxydiphosphate solution approximately 0.1 normal was pipetted into a 2$0 ml. iodine flask. Three grams potassium iodide was added followed by sufficient nine molar sul­ furic acid to give the requisite concentration of acid in the reaction mixture. The flask was immediately stoppered and a freshly prepared con­ centrated potassium iodide solution was placed in the gutter of the iodine flask to guard against the loss of iodine. After standing at room tempera­ ture in the dark for the designated, time the potassium iodide solution in the gutter was allowed to run into the flask. Excess sodium bicarbonate was added and the iodine present was titrated with standardized approxi­ mately 0.1 normal neutral sodium arsenite solution using starch solution as indicator. Timing was started upon the addition of the sulfuric acid and stopped when the sodium bicarbonate was added. Blanks were prepared by omitting the lithium peroxydiphosphate solu­ tion and substituting an equal volume of water. In all cases only a small amount of iodine was liberated; e.g., in all cases 0.02 ml. of 0.1 normal neutral arsenite solution was sufficient to discharge the blue starchiodine color. The data for various concentrations of sulfuric acid and for various times of reaction are given in Table XXVII. The concentration of tetralithium peroxydiphosphate in the solution being used was determined by adding excess ferrous ammonium sulfate solution 88 TABLE X3C7II DETERMINATION OF PEROXYDIPHOSPHATE BY THE TITRATION OF THE IODINE liberated from potassium iodide solution acidified wi t h SULFURIC ACID AS A FUNCTION OF TIME Reaction Time 0.25 M , Normality Li4P 30a-£ Acidity 0.30 M 1.0 M 3 mini 0.1012 0.1078 10 min. 0.1073 0.1078 0.1079 0.1078 15 m i n . 0.09^22 30 min. . 0.1037 0.1079 0.1079 if.3 min. 0.1067 0.1079 0.1080 60 min. 0.1073 0.1080 0.1080 2 hours 0.1077 3 hours 0.1077 5 hours 0.1077 7 hours 0.1079 8 hours 0.1077 •K-By the ferrous ammonium sulfate procedure, O.IO78. to a measured volume of the pero;xydiphosphate solution acidified to 0.5>0 molar with sulfuric acid and back titrating the excess reducing agent with. 0.1000 normal potassium diehromate solution as previously described. The concentration of the lithium peroxydiphosphate solution determined in this manner was 0,1078 normal. 89 The data show that the rate of reaction between peroxydiphosphate and iodide ions is markedly affected by the concentration of sulfuric acid in the reaction mixture. With the reaction mixture 0.25 molar in sulfuric acid a reaction period of 120 minutes is required. When the acidity is 0.50 molar a reaction period of 15 minutes is sufficient; when 1.0 molar, a five minute period is adequate. Thus it is seen that if previous workers carried out the reaction between peroxydiphosphate and iodide ions in a medium acidified with sul­ furic acid it was not necessary to allow the solution to stand overnight for the reaction to go to completion if the reaction mixture was 0.25 molar or greater in sulfuric acid. The determination was attempted without a period of standing; e.g., the liberated iodine was titrated with standardized sodium thiosulfate solution immediately after the addition of the sulfuric acid. mixture 1.0 molar in sulfuric acid was used. A reaction Approximately two milliliters from the equivalence point a false end point was obtained. After five false end points were obtained without obtaining a true end point the titration was terminated as it was deemed that a determination carried out in such a manner had no advantage over allowing a five minute reaction period prior to the titration and then titrating to a true end point with no complications arising from false end points. Adding three drops of 0.25 per cent osmium tetroxide solution was not beneficial. _The reaction between peroxydiphosphate and iodide ions wasstudied as a function of tine with solutions acidified with six molar perchloric acid. 90 The procedure employed -was the same as when sulfuric acid .was used. The data are listed in Table XXVIH. TABLE XXVIII DEL1ERMINATION OF PEROXYDIPHOSPHATE BY TH e TITRATION OF THE IODINE LIBERATED FROM POTASSIUM IODIuE SOLUTION ACIDIFIED WITH PERCHLORIC ACID AS A FUNCTION OF TIME Reaction Time 0 .09I4I8 0.1078 10 Min. O.lOlfL 0.1079 15 M i n . o.ioia 0.1077 30_ 0.1078 0.1080 hS 0.1078 0.1082 60 0.1078 0.1083 S Min. .. Normality Li4P 30a“ Acidity 0.30 M 1.0 M * By the ferrous ammonium sulfate procedure, 0.1078. The data of Table XXVIII again show the pronounced influence of acid­ ity on the reaction between peroxydiphosphate and iodide ions. Also with the greater acid concentration the reaction time should not be longer than 30 minutes. The determination was attempted without a period of standing. action mixture 1.0 molar in acid was used. A re­ Once again the procedure was not considered feasible because of the occurrence of false end points. The addition of three drops of 0.25 per cent osmium tetroxide solution was not beneficial. 91 The effect of having present various amounts of lithium peroxydiphos­ phate was studied in reaction mixtures acidified with sulfuric and perchloric acid. The procedure used was the same as above except that the liberated iodine was titrated with standardised sodium thiosulfate solution after the period of standing. The timing was started, when the acid was added and stopped with the first addition of the sodium thiosulfate solution. The data are given in Table XXIX. TABLE XXIX DETERi.-iIN.aTION OF PEROXYDIPHOSPHATE BY THE TITRATION OF THE IODINE LIBERATED FROM POTASSIUM IODIDE SOLUTION Ml. Li4PgOe Taken 0.25 M a H 2S04 Normality Li4P 20 8# Acidity 1.0 M b H 3SO4 1.0 M b HC104 5' 0.1089 0.1081 0.1086 10 0.1078 0.1080 0.1078 20 0.1077 0.1080 ■ 0.1078 25 0.1075 0.1078 0.1078 ihO 0.1081 0.1077 0.1077 By the ferrous ammonium sulfate procedure} 0.1078 a 60 minute reaction time b Five minute reaction time The data of Table XXIX indicate that the determination of peroxydi­ phosphate by the titration of the iodine liberated from a potassium iodide solution 0.25 molar in sulfuric acid is satisfactory when the iodine is 92 titrated with sodium thiosulfate solution after 60 minutes of standing. Wien the reaction mixture is 1.0 molar in sulfuric acid or 1.0 molar in perchloric acid satisfactory results are obtained with a reaction time of five minutes. The reaction between peroxydiphosphate and iodide ions in 0.50 and 1.0 molar phosphoric acid was studied as a function .of time. The procedure used was the same as the determinations wherein sulfuric and perchloric acids were used. The data are listed in Table XXX. TABLE XXX DETERMINATION OF PEROXYDIPHOSFHaTE BY THE TITRATION 01' THE lODHfe LIBERATED FROM POTASSIUM IODIDE SOLUTION ACIDIFIED WITH PHOSPHORIC ACID AS A FUNCTION.OF TIME Reaction Time Normality Li4P 20 s* Acidity 0.50 M ' 1.0 M 0.25 Hr. 0 .01*706 0.05301 0.50 0.07219 0.1019 0.75 0.07761 0.1058 1.0 0.0881*6 0 .101*9 2.0 O.lOlli 0.1073 3.0 0.1055 0.1072 5.0 0.1078 0.1078 "3y the ferrous Ammonium Sulfate procedure} 0.1078. The sluggishness of the reaction between peroxydiphosphate and iodide ions in the solutions acidified with phosphoric acid is attributed to the 93 fact that phosphoric acid is a weaker acid than sulfuric or perchloric acid. This hypothesis was substantiated by carrying out the reaction in a 10 per cent acetic acid, solution. In a reaction time of one hour per cent of the peroxydiphosphate reacted as determined by the amount of iodine liberated and in a three hour reaction time 7h per cent of the peroxydiphosphate reacted. That is, in a reaction medium acidified with an acid weaker than phosphoric acid the extent of the reaction was less than in media acidified with phosphoric acid. The reaction between peroxydiphosphate and.iodide in the presence of phosphate in solutions acidified with sulfuric and perchloric acid was studied as follows. Twenty milliliters of the lithium peroxydiphosphate solution was pipetted into a 250 ml. iodine flask. Fifty milliliters of 2.0 molar monopotassium phosphate solution was added. This amount of phosphate gave a molar ratio of 1:100 between the peroxydiphosphate and the ortho­ phosphate. After the addition of three grams potassium iodide, 8.3 ml. of nine molar sulfuric acid was added to give a reaction mixture 1.0 molar in sulfuric acid. A freshly prepared concentrated potassium iodide solution was placed in the gutter. After standing for five minutes the liberated iodine was titrated with 0.1032 normal sodium thiosulfate solu­ tion using starch solution as indicator. In all cases a well defined end point was obtained. Smaller volumes of the potassium phosphate solution were used to give different peroxydiphosphate to phosphate ratios. The volume of the re­ action mixture -was kept constant by the addition of water. 9h The data are given in Table XXXI. The procedure was repeated with the time of standing increased to 15 minutes. The reaction was also carried out in solutions acidified with perchloric acid. TABLE XXXI DETERMINATION OF PEROXYDIPHOSPHATE IN THE PRESENCE OF PHOSPHATE BY THE TITRATION OF THE IODIu E LIBERATED FROM POTASSIUM IODIDE SOLUTION Normality Li4P 20 8 Acidity Hole ratio ' P p08"4 :P0^“3 1.0 M H sS04 h 3so4 1.0 M ° ' 1.0 M b HC14 1 .0:0 .0 0.1076 0.1079 0.1083 1 .0:20 0.1078 0.1081 0.1080 1 .0 :i|.0 0.1078 0.1078 0.1078 1 .0:60 0.1053 0.1068 0.06069 1 .0:100 0.09728 0.1057 0.07751 .* By the ferrous ammonium sulfate procedure , 0.1078. a Five minute reaction time. b lp minute reaction time. The data of Table XXXI again illustrates the profound influence that the acidity has upon the reaction between peroxydiphosphate and iodide That is, a large amount of phosphate decreases the acidity of the medium by the formation of phosphoric acid. This is shown more clearly by Table XXXII which contains the same data as Table XXXI but expressed as indicated 95 TABLE XXXII EXTENT OF REACTION BETWEEN PEROXYDIPHOSPHATE AND IODIDE IN THE PRESENCE OF VARIOUS AMOUNTS OF PHOSPHATE Li4P 20g Reacted in 5 minutes in 15 minutes 1..0:0..00 100.0 100.1 1.0:026 100 ;0 100.2 1 .0 :0.52 100.0 100.0 Mole ratio H 2S04:KH2P04 a ' 1.0:1.0 97.68 99.30 1.0:1.3 9 0 .2U 98.05 HI:KH2P04 b 1.0:0.00 100.3 1.0:0.26 100.1 1 .0 :0.52 100.0 1.0:1.0 7-5.85 1.0:1.3 71.90 a Acidity 1.0 molar H 3S04 b Acidity 1.0 molar HC14 The data presented show that the reaction between peroxydiphosphate and iodide is greatly influenced by the acidity of the reaction medium. Even in the presence of phosphoric acid a small change in the concentration of the acid exerts a large influence on the rate of reaction. The Jf 96 dependence of the rate of reaction between peroxydiphosphate and iodide upon the acid concentration is an important consideration in the deter­ mination of peroxydiphosphate by the iodoroetric procedure} especially in the presence of phosphate. The recommended acidity in the absence of phosphate is 1.0 molar in sulfuric acid or perchloric acid. The minimum recommended reaction time is five minutes. A reaction time of up to 60 minutes is not detrimental when sulfuric acid is used. When perchloric acid is used the maximum reaction time is 30 minutes. In the presence of phosphate the acid concentration should be 1.0 molar in sulfuric acid. Wien less than 2.3 mole per cent of peroxydiphosphate is present the minimum reaction time should be increased to 13 minutes. Since a primary standard quality peroxydiphosphate compound could not be prepared the iodometric determination must be compared to the results obtained by the ferrous ammonium sulfate procedure. Sufficient data were not gathered for a statistical treatment of the iodometric determination. However it is worthy of note that the average concentration of the lithium peroxydiphosphate in Table XXVII obtained oy titrating the iodine liberated from reaction mixtures 1.0 molar in sulfuric acid was 0.1079 while the normality determined by the method involving the back titration of an excess of ferrous ion gave a value of 0.1078. With the reaction mixture acidified with perchloric acid to the extent of 1.0 molar the average normality of the lithium peroxydiphosphate solution was found to be 0.1079. Although not on a statistical basis it can be said that the ferrous ammon­ ium sulfate procedure gives results that differ from the results obtained iodometrically by one part per thousand. 97 I X . SUMMARY Tetrapotassium peroxydiphosphate was prepared by the method of Fichter and Gutz-wilier (5). Their findings were substantiated except for the claim that the final product was pure white. The tetrapotassium peroxydiphosphate always contained some chromate. No method of crystalli­ zation could be found that would remove the chromate. Tetrapotassium peroxydiphosphate was precipitated from the electrolyte of Fichter and Gutzwiller by adding solid potassium hydroxide. The product also contained chromate but the precipitation served as an efficient method for the preparation of a rather impure chromate containing tetrapotassium peroxydiphosphate which was found to be suitable as a starting material for the preparation of tetralithium peroxydiphosphate tetrahydrate. A solution of lithium peroxydiphosphate was prepared from a solution of the crude tetrapotassium peroxydiphosphate by precipitating the potas­ sium phosphate, and fluoride by adding a lithium perchlorate solution. The peroxydiphosphate was precipitated from the solution as tetralithium peroxydiphosphate tetrahydrate by the addition of methanol. The material was chromate free. A 0.1 normal solution of the tetralithium peroxydiphosphate tetra­ hydrate had a pH of 10.2. thousand in ij.9 days. It decomposed to the extent of five parts per At lower pH values the solution was less stable. Peroxydiphosphate was determined by adding excess ferrous' ions -to a solution of peroxydiphosphate and back titrating the excess with potassium 98 dichromate solution using diphenylandne sodium sulfonate as indicator. Reaction mixtures 0.25 to 1.0 molar in. sulfuric acid and 0.25 to 2.0 molar in perchloric acid were found to give satisfactory results . Higher acid concentrations gave low results. When reaction mixtures containing hydrochloric acid were used somewhat erratic results were obtained. In 1.0 to 2.0 molar hydrochloric acid solutions plausible results were ob­ tained. Ho time of standing was necessary. ion was not critical. The amount of excess ferrous The determination could be carried out in the presence of phosphate. When phosphoric acid was used to acidity the re­ action mixture high results were obtained. acid concentrations gave higher results. In this instance the higher With reaction mixtures 0.5 to 2.0 molar in phosphoric acid the results were a f-onction of the amount of excess ferrous ion present. The largest excess gave the highest results. The direct potentiomeiric titration of peroxydiphosphate with a ferrous ammonium sulfate solution was accomplished. Satisfactory results were obtained when the reaction mixture was 0.25 to 0.50 molar in sulfuric acid or 0.50 molar in phosphoric acid. Higher sulfuric acid concentra­ tions gave low results while higher phosphoric acid concentrations gave high results. Reaction mixtures 0.25-to 2.0 molar in perchloric acid gave low results . The use of hydrochloric acid gave low results because of oxidation of the chloride ion. The equivalence point in the titration of peroxydiphosphate with a ferrous ammonium sulfate solution was successfully determined by means of the dead stop end point technique. Satisfactory results were obtained if the peroxydiphosphate solution was 0.25 to 1.0 molar in sulfuric acidj 99 0.50 to 1.0 molar in phosphoric acid, or 0.25 to 0.50 molar in perchloric acid. Higher concentrations of sulfuric or perchloric acid gave low re­ sults . Higher concentrations of phosphoric acid gave high results. An iodometric determination of j>ero:xydiphosphate was achieved. The rate of reaction between peroxydiphosphate and iodide ions was found to be influenced profoundly by the acidity of the reaction mixture,- a small increase in acidity greatly increased the rate of reaction. In the ab­ sence of phosphate the' reaction between peroxydiphosphate ana iodide in 1.0 molar sulfuric acid or 1.0 molar perchloric acid, was complete in five minutes. The determination was successful in the presence of phosphate . However when less than 2.5 mole per cent of peroxydiphosphate was present only the 1.0 molar sulfuric acid medium was satisfactory and a 15 minute period of standing was necessary. 100 X. LITERATURE CITED (1) D JAns, J. and Friedrich, W., Ber., k 3 1880 (1910) . (2) Degussa, A.P., D. R. P. 6U8715. (3) Degussa, A.P., D. R. P. 672299. (10 Fichter, F. and Bladergroen, ¥., Helv. Chim. Acta., 10, 5U9 (1927). (5) Fichter, F. and Gutzmller, E., Helv. Chim. Acta., 11, 323 (1928). (6) Fichter, F.and Muller, J., Helv.Chim. Acta.,1,297(1918). (7) Fichter, F.and Rius y Miro,A., Helv.Chim.Acta.,2, 3 (1919). (8) Furman, N. H., "Standard Methods'of Chemical Analysis", Fifth Edition, p. 1201+, New York, D. Van Nostrand Co., Inc. 1939. (9) Helrich, K. and Rieman, ¥., Ind. Eng. Chem., Anal. Ed., 19 651 (19U7). (10) Husain, S. and Partington, J. R., Trans.Faraday Soc., (11) Kunnin, R. and Myers, R. J., "Ion ExchangeResins", John Wiley and Sons., Inc., 1950. 2ij., 235 (1928). p. 130, New York, (12) Machu, W., "Das Wasserstoffperoxyd und die Perverbindungen" Zweiten Auflage, p. 27k, Vienna, Springer-Verlag, 1951. (13) Maxted, E. B., J. Chem. Soc., 19k3, 20k. (ill) Maxted, E. B., J. Chem. Soc., 19U5, 763. (15) Petrenko, G. I., Journ. Russ. Phys. Chem. Soc., 3k, 20k (1902). (16) Finerva-Alvarez, E., Chem. News, 9k, 269 (1906). (17) Rius y Miro, A., Anal. Fis. Quim., 16, 573 (1918); C. A., lU, 3605 (1920). (18) Rius y Miro, A., Anal Soc: . expon. fis quim., 18, 35 (1920); C. A., 15, 3596 (1921) . (19) Rius y Miro, A., Trans. Am. Electrochem. Soc., 3k, 3k7 (1928). (20) Rudenko, Y. P., J. Russ. Phys. Chem. Soc., kk, 1209 (1912); C. A., 7, 738 (1913). 101 (21) Sarver, L. A. and Kolthoff, I. M., J. Am. Chem. Soc., 53, 2906 (1931). (22) Schenk, P. W. and Rehaag, H., Z. anorg. Chem., 233, 5-03 (1937) . (23) Sc’ nmidlin, J.and Massini, P., Ber., 53, 1162 (1910) . (25) Siebold, A. D. R. P. 279, 306. (25) Snell, F. D. ana Hall, C. A. "Colorimetric Methods of Analysis", Third Edition, p. 275, Hew York, D. Van Nostranc Co., Inc., 1959. (26) Stone, K. G.,and Scholten, H. G,, Anal. Chem., 25, (27) Toennies, G., J. Am. Chem. Soc., 59, 555 671 (1952) . (1937). (28) Willard, H. H. and Diehl, H., "Advanced Quantitative Analysis", p. 258, New York, D. Van Nostrand Co., Inc., 1953. (29) Willard, H. H. and Furman, N. H., "Elementary Quantitative Analysis", Third Edition, p. 266, New York, D. Van Nostrand Co., Inc., 1950. 102 APPENDIX The data plotted in Figure 1 are tabulated in Table XXXIII. TABLE XXXIII PER CENT OF PHOSPHATE OXIDISED TO PEROXYDIPHOSPiiiTE UPON PROLONGED ELECTROLYSIS . % Phosphate Oxidized to Peroxydiphosphate Hours Electrolysis Calc, on Basis of Actual Volume Assume No Diminution of Volume Run 7 - 5 ^3 h3 ii 67 71 18 73 60 2U 73' 83 30 73 66 Run 8~ 6 h9 12 70 16 70 22 72 28 73 30 71 -if Original volume maintained by adding water. 103 The data plotted in Figure 2 are tabulated in Table XXXIF. TABLiii XXXIV TITRATION OF 25.00 M L . 0.1110 W LITHIUM PEROXTDIPHOSPHATE SOLUTION WITH 0.1000 N FERROUS AMMONIUM SULFATE SOLUTION Slow Titration, Electrode 1 Milliliters Millivolts 0.0 1.1 2.2 : . 3.b 1+.5 5.6 6.7 7.6 9.0 10.1 11.2 12.3 13. b lb.6 15.7 ' 16.8 17.9 19.0 20.2 21.3 22.h 23.5 2b.6 25.8 26.9 27.0 27,2 27.3 27.5 27 .6 27.7 27.S 27 .9 28.0 901 901 901 901 901 901 901 905 905 905 905 905 905 905 905 905 905 905 905 905 888 880 8 61 £35 . 813 810 810 602 795 795 668 5U3 515 501 Slow Titration, Electrode 2 Milliliters Millivolts 0.0 l.i 2.2 3.3 b .1+ 5.6 6.7 7.6 8.9 10.0 11.1 12.2 13.3 11+.2+ 15.5 16.7 17.8 18.9 20.0 21.1 22 .2 23 .3 2k.k 25.5 26.6 27.0 27.2 27.327.b 2 7 ,5 27.0 27.8 27.9 28.1 ' 91+3 975 983 998 998 1005 1005 1005 1000 1000 1000 999 995 986 969 957 952 9U0 927 910 90l+ 873 660 822 789 ?i+o 727 711 697 691 593 563 527 518 Rapid Titration, Electrode 1 Milliliters Millivolts 26.9 27.0 27.1 27.2 27.3 27 .U 27.6 27.7 27.8 2?.9 28.0 28.1 28.2 28.3 28.6 29.1 (Continued next page) 63U 61+0 6247 6I47 61+9 61+9 61+8 61+5 61+2 572 52U 511 1+96 1+92 162 1+66 TABLE XXXIV - Continued Titration, Electrode 1 Milliliters Hi.llivolts SIovj 28.6 29.1 30.2 31 .1 32.5 33.7 31.8 35.9 ' 37.0 38.2 39.3 175 li.62 119 k31 1 27 I22 I20 Slow Titration, Electrode 2 Millilite;rs Millivolts 28.2 28.1 28.9 .30.0 31.1 33.3 38.9 Rapid•Titration, Electrode 1 Milliliters Millivolts 510 197 182 I63 152 119 A01 ■115 ■ H5 110 107 The data plotted in Figure 3 tabulated in Table XXXV. 105 TABLE XXX? TITRATION OF 25.00 ML. 0.1110 N LITHIUM PEROXYDIPHOSPHATE SOLUTION WITH 0.1000 N FERROUS AMMONIUM SULFATE SOLUTION Slow Titration Electrode 1 Milliliters 20.8 21.9 23.0 2h.l 25.2 26.3 2?.0 27.3 27.6 27.7 27.8 27.9 28.0 28.3 29.h 29.7 30.8 31.9 33.2 3U.3 El. ral. 0 1.7 0.8 1.9 2.6 2.3 3.0 8.0 0.0 127 . 125 28 lb 9 1.3 1.3 1.2 1.0 0.5 0.2 Slow Titration Electrode 2 Milliliters 20.6 21.7 22.8 23.9 25.0 26.1 26.8 27.1 27.3 27 .U 27.5 27.6 27.8 27.9 28.1 . 28.2 26.3 28.7 29.5 30.6 El. ml. 0.2 0.1 2.0 1.0 3.8 3.3 12.0 13.0 16.0 lii.o 6.0 2.0 30.0 36.0 9.0 8.0 .6.0 '■'1.6 2.0 1.0 Rapid Titration Electrode 1 Milliliters 27.0 27.1 27.2 27.3 27.ii 27.6 27.7 27.8 27.9 28.0 28.1 28.2 28.3 28.5 28.9 E ml. 6 7 0 2 0 1 3 3 70 b8 13 lit it. 3 3