SOME REACTEONS AND PROPERTEE3 CF TEYRALIYHEUM FEROXYDE?HOSPHATE Thesis for “19 Degree of ph. D. MICE-{EGAN STATE UNIVERSITY Clyde HOE: Heen Chong 1958 This is to certify that the thesis entitled SOME REACTIONS AND PROPERTIES OF TETRALITHIUM PERomlPI-IOSPHATE presented by CLYDE HOK HEEN CHONG has been accepted towards fulfillment of the requirements for Eh. D. degree inflHEl-IISTRY “ Major professor Z; Date 0-169 LIBRARY Michigan State U . . H some REACTIONS AND PROPERTIES OF TETRALITHIUM PEROXYDIPHOSPHATE By Clyde Hok Heen Chong A TEEIS Submitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial filiflJment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1958 ll ‘1 I1 " ~ - Wet-3""w cur ACKNOWLEDGMENTS The author wishes to express his deep and sincere appreciation to all those people who have made his education possible. Special mention is made of Dr. Elmer Leininger under whose kind and efficient guidance this work was done. The author is indebted to the Socony Mobil Company whose program provided for the establishment of a Socony Mobil Fellowship in Analytical chemistry. Acknowledgment is also made to Dr. Clayton F. Callis of the Inorganic Chemical Division of Monsanto Chemical Company, St. Louis, Missouri, for the preparation of the nuclear magnetic resonance spectrum of tetralithium peroxydiphosphate and for his helpful interpretation of its structure. W—X—kfi ii VITA Name: Clyde H. H. Chong Born: March 6, 1933 in Honolulu, Territory of Hawaii Academic Career: Western Military Academy, Alton, Illinois 19145-1950 Wabash College, Crawfordsville, Indiana 1950-1951; Michigan State University, East Lansing, Michigan 1951:4958 ' Degrees Held: A. B., Wabash College, 1951; iii SOME REACTIONS AND PROPERTIE. OF TETRLLITHIUM PEROXIDIPHOSPHATE By Clyde Hok Heen Chong AN ABSTRACT Subndtted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MTOR OF PHILOSOPHY Department of Chemistry Year 1958 n Approved fl 773,51, %( )Kflff/ L/A «r --_ pcwer . Ml th: as areas 5 to 30 1 £51? he; he he may I \b) Sulfate ; £811er 1.01. ABSTRACT Tetrapotassium peroxydiphosphate was prepared by electrolysis of an alkaline phosphate solution and COnVeI'ted to the lithium Salt by the method 01' Chulski (1). This method was modified to increase the yield to 69 per cent from 33 per cent based on the amount of Phosphate originally taken. It was found that basic or neutral peroxydiphosphate solutions exposed to diffuse light for 12 hours at room temperature or heated at 100°C. for 30 to 60 seconds showed no appreciable loss in oxidizing power. A method was developed for the determination of manganese based upon the oxidation of manganous ion to manganese dioxide by means of an excess of peroxydiphosphate. The time required for the oxidation of 5 to 50 mg. of manganese was approximately 12 hours at room temperature, four hours at 50-5500. and 30 to 60 seconds at 100°C. Two methods were used to complete the determination. (a) The excess peroxydiphosphate in the filtrate ,after removing the manganese dioxide , was determined by reaction with ferrous ion (1). (b) The manganese dioxide was dissolved in excess ferrous ammonium sulfate and the excess ferrous ion determined by means of potassium dichromate. This method required the use of a correction factor of 1.01. m- “'7 ll peSpite their interference with the analysis of the filtrates ’ mercuri-c’ Chronic, arsenic(l'II), chloride and ammonium ions were found to give no adverse effects when the precipitates were analyzed for manganesa Tungstate, thallous, cobaltous, ferric, aluminum, silver, barium, lead, zinc, and nitrate ions interfered in either MOdification. me mechanism of the PeroxydiphOSPhate reaction is believed to involve the oxidation of manganous ion to a higher valent Species, Probably permanganate, whiCh then reacts with manganous ion to form manganese dioxide. A spectrophotometric method for determining peroxydiphosphate in the ultraviolet region was developed. No definite absorption peaks were obtained, only characteristic curves dependent on concentration and pH. These curves were found to adhere to Beer‘s Law at 2&0 nu and at a pH of 6.0. The method was used for the determination of excess peroxydiphosphate in the filtrate following oxidation of divalent manganese and removal of the precipitate formed. Individual oxidation of thallous, chronic, and cobaltous ions as well as selenious acid was attempted under various conditions. Although oxidation occurred to some extent, none of these reactions were found suitable for quantitative use. In the presence of a known amount of manganous ions, it was possible to effect quantitative oxi- dation of thallous and chronic ions. In these reactions, the oxidation of the manganese appears to induce the oxidation of the vi ‘ery—mu i" .. . ‘1 s "( thallium or the chromium. The structure of peroxyfliphosphate was investigated. It Was shown that decomposition of perOxydiphOSphate resulted in the formation of orthophOSPhate’ but no pyrophosphate was detected. Potentiometric titration of perowfiphosphoric acid revealed that there was one Strong hydro gen per phosphorus atom. This Was further verified by a conducto_ metric titration. A nuclear magnetic resonance spectrum revealed only one peak shifted 47.), parts per million relative to 85 per cent orthophosphoric acid which suggested the symmetrical structure. The foregoing series of investigations indicated that the structure of peroxydiphOSphate must be h o .. | T _ O - P - O - O - P - 0 ¢ _ 0 and not 9 ‘3- h-P-o—P-of. J t O O REFERENCE 1. Chulski, T., Doctoral Dissertation, Michigan State College, East Lansing, Michigan (1953). vii Page I WCTIONODoaoocoOOOO0000......000......o.................. 1 II mwmoooootooooooocooooo0000.0...Ooooooooooocoo-co....... 2 III 1,me OF W PEROIYDD’HOSPHATE TETRAHYDRATE... . 3 IV W.........OOIOOC'OOCOOO000.000.00.003.coo-0.00000... 10 L. Reagmts and ChMSoooooococooolooooooooo-0.0.00.0... 10 B. ‘ppmms......O.00......OOCIOOOOO000.00.00.000000000... 15 0. Stability Studies 01’ TGtralithium Peroxydiphosphate SOlutiOHSoooO.00OO’OOOOoooooooooooooooooooooooo-ooooo. 16 D. Analytical Study of the Reaction Between Peroxydiphos.. phate and Wese(n)000000'000000000000000000...on 20 1. Introduction”....oo.............................. 20 2. Identification of the Precipitate Formed.......... 2h 3. Titration Of the Filtrateooooocoooooooococo-00o... 26 1;. Effect of pH on the Rate of =Forrmtion of Manganese DiOXideooooooooocoo-0000.00.00.00...coco-coo..- 5. Proposed Hecba-n-im Of Ofidationoooooonoooooocoo... 3O 6. Time Required to Effect Oxidation at Room Tmemmreooooooooooooococoa-cococoono coo-one 3h 7. Time Required to Effect Oxidation at 50-55 C...... 38 8. midation at Elevated Taupera‘hlreS....u-.uu.u. 9. Recommended Volumetric Procedure.................. 1&5 10. Effect Of Foreigl Ionsoooooooooooooooooo-coo-coco. )4? E. Bpectrophotometric Determination of Peroxydiphosphate.. . 51 1. Determination of Excess Peroxydiphosphate in the “@1636 ondationoooooooo-oooooooon-ooooooco. 61 F0 ‘ttwpts at mdation Of Other IOHSoooooooo-oococoon-coo 1. Oxidation Of Miml)oooooooooooooooacocoa-coo- 2. mation or Chromi IH)-ooooooonoooooooouoooooo 72 30 mation Of vanadim IV)oooooooooco-oooooooococo. 79 ’ he Oxidation Of Cobalt..l‘q...IQ.' . Q 0 .III‘010'.|.Q‘D"..'VQO.I..I.D‘D‘O'OCI'O‘IIOIIOOO tea-n . I ..‘§notc.oo..z.....'.ogoooorooaonowso-ouoooooocollau c c~- g.-.....o‘.-:oooQ-II9‘O"‘I-OccootfiofioclcnontoonOAIOI o -.o¢x~.¢-. \w';.51.viCID..‘)O."OO¢I£~‘UI.‘ Di ’(1 '7", ( I'V o.“ x... I. DITRDDUCTION Little use of peroxydiphosphate as an oxidizing agent has been mde. Despite methods for the successful Preparation (12:16:17,18,35) and determination of perOUdiPhOSPhate (1.1), an analytical applications have been made of peroxydiphosphoric acid, ”203: or its salts in analytical chemistry. It was the purpose of this work, therefore, to undertake a more extensive investigation on the Possible analytical use of peroxydiu- phosphate as an oxidizing agent and to extend the studies, initiated primarily by chulsld (11), on the characterization of peroxydiphosphate. The problem was undertaken by preparing the stable tetralithium peroxydiphosphate tetrahydrate, Ligzoe'hego, and using this peroxy salt as the original starting mterial throughout the course of the work. sue} II. HISTORICAL A Considerable amount of work has been done on peromiphosphates prior to 1953. Chulski (11) has 5110111de 3. complete bibliography of the work on both peromnono" and Peroxydiphosphate reported through 1953. The preparations, reactions and properties of peroxyphosphates were considered. survey of the literature from 1953 up to and including 1957 indi— cated no additional studies have been made with peroxyphOSphates except perhaps for the amine-peroxide reaction of Schoenemann (33,36) . The reaction involved the detection of organorphosphorus compounds. An alkaline peroxide solution containing an oxidizable primary amine such as o-dianisidine reacted with the organophosphorus compound forming a perdxyphosphate which then reacted with the primary amine to form a highly colored azo dye. The amine-peroxide reaction was stated to be applicable to all quinquevalent phosphorus containing compounds having a residual positive charge on the phosphorus atom allowing nucleophilic displacement of a labile anionic group by the perhydrmqyl ion. j, PREPARATION OF mmnHIUM Pmml‘PHOSPHATE TEI'RAHIDRATE (31'11llski (11) "as able 17° Prepare pure tetralithium permdiphosphate, 1.13208, by adding lithium perchlorate 1"0 a cold solution of tetra.- potassium peroxydiphosphate. The latter was prepared by the method of Fichter and Gutzwiller (16)- This consisted of electrolyzing a mono- 'po‘baSSiu-m phosphate solution containing potassium hydroxide, potassium fluoride , and potassium chromate. The potassium perchlorate, lithium Phosphate and lithium fluoride formed upon the addition of lithium perchlorate were not very soluble in water and were filtered off. The lithium peroxydiphosphate was then removed from the filtrate by the addition of methanol. The recommended procedure as suggested and outlined by Chulski for preparing hydrated lithium peroxydiphosphate was foILowed. Sixty grams monopotassium phosphate, LLO g. potassium hydroxide, 2],; g. potassium fluoride djhydrate and 0.07 g. potassium chromate were dissolved in 150 ml. of water. Cooling was necessary as considerable heat was evolved. After cooling the solution, it was filtered through a number two Whatman filter paper on a Buechner funnel and then diluted to 200 ml. The solution was transferred to a 210 ml. volume platinum dish. The solution was cooled in an ice bath during the electrolysis; the temperature of this bath was maintained at o—3°c. The platinum dish served as the anode and a rapidly rotating bent platinum wire was used F-v—v-r—u '- “2.2 the cathode. A cuI‘I‘en’C of 1-9 EU'IpeI‘eS Was obtained b y applyin g 9-2 aQPrO’jmateJ-y 6-5 hours using a Sargent-610mm electro-analyzer e volts across the electrodes. Th eleCtrOlysis was carried 011?: for Since the sharp odor of ozone was detected in the Course of the electrol YSis, the preparation of the peroxydiphosphate was carried out in a well ventilated hood. Following the stated electrolysis period the platinum dish was covered with a watch glass and the electrolyte allowed to stand in the dish at room temperature for at least 12 hours. Longer standing was not detrimental since it was found that the peroxydiphosphate in this solution was stable for months. Electrolysis was carried out a second time for another 6.5 hours. Following this electrolysis period the electrolyte was transferred to a rubber stoppered 300 ml. Erlenmeyer flask. The platinum dish was used for the next run without rinsing. After standing for at least 12 hours the electrolyte was heated to 50-5500. Sixty grams of potassium hydroxide was then added to this warm solution. The flask was stoppered and shaken until all the potassium hydroxide dissolved. After standing overnight or for a longer period of time, the super- natant liquid was decanted off and the microcrystalline impure tetra- potassium peroxydiphosphate filtered on a medium porosity sintered glass filter funnel. In this method the mother liquor was then dis— carded - oxydipho The crude tetraPOtaSSium Per Sphate was th - en dlSSOlv-ed in N3" of Water in a 600 "’1' beaker. 200 . tlon was cooled in an ice ham while 1&0 ml- Of a lithium perchlorate SOlution containing 5h go lithium perchlorate trihy'drate per 100 ml- of solution Was added in a droprwise rate from a dropping i‘urmel. HGChanical stirring was used. After stirring for 15-20 nfimtes the material was filtered througx a number tm mutmn filter paper employing a Buechner funnel. Completeness of the precipitation was tested by the addition of a little more lithium perchlorate solution to the cold filtrate. Since qualitative tests showed that carbonate was present, the solution of lithium peroxydiphosphate was freed from carbonate by acidifying it to a pH of 6.0—6.5 with 35 per cent perchloric acid solu- tion and removing the carbon dioxide by aeration under reduced pressure. Approximately 10-15 ml. of the acid was required. The solution was aerated by placing it in a 500 ml. filtering flask fitted with a rubber stopper bearing a capillary tube which extended to the bottom of the flask. A water aspirator was used to draw air through the capillary tube and solution for about five minutes. Before adding the methanol for the removal of the lithium peroxydi- phosphate the pH was adjusted to a value of 9.5-10 .0 by the addition of lithium hydroxide. It was observed that if the pH was higher or . lower than the indicated range the material precipitated by the methanol was less crystalline in nature. Approximately 0.5-0.8 g. lithium hydroxide was required. to 9 05-10 After the pH was adJUSted '0 the solution was Cooled 00. in an ice bath. Any additional potassium Perchlorate that to Precipitated was removed by filtrati on and the Solution transferred to a 600 ml. beaker. . It was found that oftentjmes a viscous material was obtained when the methanol was added at room temperature. Precipitation at ice bath taupemmres always produced the viscous mterial, However, When the precipitation was carried out at hO—hSOC. the formation of the Viscous material was avoided. Horeover, by carrying out the precipitation with methanol at the specified temperature range, a smaller volume of the alcohol was required than when precipitation was done at room temperature. As a result of the above findings the solution containing the lithium peroxydiphosphate was heated to 140-1500. and placed in a water bath at the same temperature. While the solution was stirred mechanically 70 ml. methanol was poured in from a graduated cylinder. No precipitate was obtained innuediately but after five minutes of stirring a precipitate formed. Then 30 additional ml. of methanol was added at a rapid dropwise rate from a dropping funnel. The solution was stirred for a total of 15 minutes. The lithium perondiphosphate was filtered on a Buechner funnel using a number two Hhatman filter paper and sucked as dry as possible. The residual chromate was removed by washing the precipitate with 1:1 water-methanol mixture at hO-hSOC. until the wash liquor came through colorless; generally four or five 20 ml. portions of the wash solution 1‘3 “fieient. The absence °f ;:0::: :8 the peroxydiphOSphate was oneox‘ed by the use Of 5411311311le ) as Well as spectrophoto- metrically. . The compound was air dried until free flowing, It was Shown by chulski that the material was tetrallithium PerOJQ'diphosphate tetrahydrate. The yields obtained from this method of preparation averaged ZLJ; g. Based on the amount of phosphate taken this would correspond to a 38 per cent yield; based on the conversion of the potassium PerquiphOSphate to the lithium salt this would correspond to a 52 per cent yield. The possibility for improving the yield of tetralithium PerOXVdiphosphate was investigated. It had been shown previously that prolonged electrolysis for total periods of time greater than 12 hours resulted in no material increase in the amount of phosphate which was converted to the peroxydiphosphate form (11). It was claimed a 73 per cent conversion was usually obtained. This would immediately imply that complete recover of the tetrapotassium peroxydiphosphate from the electrolyte was not achieved in the previous work. The consistently low yields of tetralithium peroxydiphosphate tetrahydrate can be attributed mainly to this factor. The following modification was made in the conversion and recovery procedure for tetralithium peroxydiphosphate tetrahydrate. It essentially consisted of a resaturation of the mother Liquor. The filtrate follow- :Lng the removal of tetrapotassium peroxydiphosphate was returned to the 300 ml. rubber stoppered Erlenmeyer flask and left overnight at room temperature. Following this period of standing the solution was gated to 50.55%. and an additional 60 g. y; o f Solid potassium obi-b.1011. Wide was added to the "am 5 The flask was restopp ered 33d owed shortly thereafter. Had the resaturation been carried out Men until all the potassium hydroxide dissolved. A precipitate i Wadi-3W1? after removal of the original precipitate, no aPPI'GCiable precipitation of tetrapotassium peroxydiphosphate would have occurred. The precipitate of tetrapotassium peroxydiphosphate was filtered off after an overnight standing period and the subsequent treatment previously described for the preparation of the lithium salt from tetra— potassinm peroxydiphosphate was followed. Resamration With potassium hydroxide in the nanner described ultimately resulted in an average total yield of U4 g. tetralithium peroxydiphosphate tetrahydrate. Based on the amount of phosphate taken the yield was increased to 69 per cent or based on the ammmt of tetrapotassium peroxydiphosphate used, the yield was 95 per cent. The oxidizing power of the perondiphosPhate from each of the two fractions obtained was determined by the ferrous ammonium sulfate- potassium'dichromate method (ll). From several representative prepara— tions 3. series of approximately 0.09 H lithium pemxydiphosphate solutions was prepared from each fraction by dissolving approximately 1.600 g. tetralithium peroxydiphOSphate tetrahydrate in water and dilut— ing to 100 m1. Twenty milliliter portions of the peroxydiphosphate solutions were pipetted into 250 m1. iodine flasks containing 30 ml. water. The concentration of the peroxydiphosphate was calculated from the amount of ferrous ion that had not reacted attributing all the 01} de-ing properties of the solution to the perowphosphate. Triflicate determinations were made With a precision of two parts pe mouSand' The concentrations of the peroxydiphOSphate samples Were r expressed in.terms of normality and all were calculated to the Same relative sample weight of 1.600 g. The results are tabulated in Table I o TABLEI COMPARISON OF THE OIEDIZDIG POWER OF PERDXIDTPI-DSPHATE . FROM REOVERED FRACTIONS Run . . . Fraction Normality LiaPzOe* 3 a ' 0.0883 b 0.08149 6 . a 0.0850 0.08h9 10 a 0.0868 0.0863 *Calculated to the same sample weight of 1.600 g.5 1.600 g. - 000883 N The above data show that there is no appreciable loss in oxidizing power in the second fraction obtained by saturating the filtrate again with solid potassium hydroxide. VI, EXPERIMENTAL A. Reagents and Chemicals All the chemicals which were used in the Course of this Work were of sufficient purity to meet the specifications of the American chemical Society for reagent chemicals. An approximately 0.1 N tetralithinm peroxydiphosPhate solution was prepared by dissolving 16 g. of tetralithium peroxydiphosphate tetrahydrate, which was made by the method previously described, in water and diluting to one liter. Other concentrations of peroxydiphos... phate solutions were prepared by taking the appropriate quantity of lithium salt and diluting with water. Preparation of peroxydiphosphate solutions nude in this moner were basic in nature and stable for a considerable period of time (11). Ferrous ammonimn sulfate solution approximately 0.2 N and 0.15 M in sulfuric acid was prepared by dissolving 80 g. of ferrous ammonium sulfate hexahydrate in 600 m1. of water acidified with 18 ml. of concentrated sulfuric acid and diluting to one liter. Potassium dichromte solution, 0.1000 R, was prepared by weighing out n.9352 g. Hallinckrodt reagent grade potassium dichromate, dissolv- ing it in water, and diluting quantitatively to one liter. The potassium dichromate was dried at 110°C. for two hours before it was weighed out . lO llnl' NLV‘II'UW 1 II .n u. ~~ y . solut' The ferrous ammonl um sulfate 1011 Was Standardized each day . . . . the fOILOWing £0 «35 used by mtratmg it 311 "unner- with the 0.1000 N e sodium Sulfonate as ’ ' ' the ferrous . mdicator' Twenty milliliters of Womum Sulfate SOlution otaSSium dichroma‘be solution using diphenylamin ? was pipetted into a 250 ml. Erlenmeyer flask followed by the additio n of 30 ml. water and 6 ml. 6 N sulfuric acid. Then 3 ml. of 85 per Cent phosphoric acid was added for each 50 ml. of volume expected at the end. Three—tenths milliliter, six drops, of 0.01 M diphenylamine sodium sujfonate solution was added as indicator. Potassium dichromte solu- tion was added until the first tinge of purple or violet-blue color appeared. The concentration of the lithium peroxvdiphosphate solution was found by determining the total equivalent weight of oxidizing material present and attributing all the oxidizing properties of the solution to peroxydiphosphate. The amount of oxidizing material present was determined by adding a measured excess of standardized ferrous ammonium sulfate solution to a known volume of the acidified solution of peroxy— diphosphate. After standing for a minute the excess ferrous ion was back titrated with 0.1000N potassium dichromate solution using diphenyl- amine sodium Slflfonate as indicator. The difference between the volume of 0.1000N potassium dichromate solution used in the standardization of a given volume of ferrous ammonium sulfate solution and the volnme of potassium dichromate solution used in the back titration of excess unreacted ferrous ion following reaction with peroaqdiphosphate 90 ‘0. 3» {11°11 rePresented! the amount of ferrous ion calculfi’oed from this difference obtained. The concentration of per oxy... diphoSPh‘te “3 expressed in terms °f ”muty- Triplicate detemin- ations agreed to three parts per thousand. A mnganous sulfate solution containing a‘p'prmtely 10 mg. manganese per ml. was prepared by dissolving 32.60 g, Baker’s manganous sulfate monohydmte in 100 ml. water and diluting to one liter. Approximately 0.01m potassium permanganate solution was prepared and standardized agfinst pure sodium oxalate by the usual procedure (19). Someof the solution was also standardized against ferrous ammonium sulfate solution. The ferrous ammonium sulfate was previously standard- ized against 0.100011 potassimn dichromte solution using diphenyimjne sodium sulfonate as indicator. Standardization of the permanganate solution by the ferrous ammonium sulfate method gave an average value of 0.01412)! for three determinations with a deviation of three parts per thousand; by the oxalate method an average value of 0.0141114 was obtained for three determinations with a deviation also of three parts per thousand. The potentiometric titration method of Lingane and Karplus (26) was used to standardize the mnganous sulfate stock solution. Twenty milliliters of the manganous sulfate was pipetted into a 1.00 ml. beaker followed by 150 ml. of saturated sodium pyrophosphate solution. f which reacted with the ' same 01111113 0 ferrous ro‘dlehOSphate' The v amonim sulfate was tion of th 56a in both cases. The concan‘bra e perOJUdiphosphate was l2 . o . o . . n e . . o . I n o e a n a I o _ n '- e n u u a _ _ . _ o I u : I . - , . . «1...: . ‘ V. . n I i .. ~ . . L. o . a .I _ u _ . . p . I :4 . u . . . u I . .. . i q 411» | 1 x 1 \‘I \l‘ln'. «ml. 1|. _ Io“. . )1!- I II . .3 13 Two milliliters of 6N sulfuric acid was added to give a solution with a pH of 6.7. A Becknan saturated calomel electrode was placed in the solution to serve as a reference electrode. As the indicating electrode, a. Bechmn platinum electrode was used. The titration was carried out with a Fisher titrimeter. Standardized potassium permanganate solution was added from at- burette bearing an offset tip in one milliliter increments until a rapid change in potential was noted; then it was added in 0.1-0.2 ml. increments. During the titration the solution was stirred with a magmtic stirrer. Standardization of the manganous sulfate solution in this manner gave values of 0.37146; 0.3759 and 0.3769]! for three determinations or an average value of 0.375811. The normlity was expressed in toms of an oxidation change of two for manganese. A 0.1003)! thallous sulfate solution was prepared by dissolving exactly 12.621 3. of recrystallized thallous sulfate in 200 ml. of distilled water and diluting quantitatively to 500 ml. (314). This stock solution, at a pH of 6.1;, was used throughout the investigations to follow. A stock solution of chronic sulfate, Cr2(SO4)3, approximately 0.111, was prepared by dissolving approximately 35 g. of Cr2(m4)3-18H20 in 500 m1. of water and diluting to one liter. Heating was necessary to effect dissolution of the chronic sulfate. The standardization of the chronic sulfate stock solution was effected by oxidation to chromate, followed by the addition of a ”in We measured excess of ferrous ammonium sulfate solution and titration with standard potassium dichronmte (23). The oxidation was carried out in acid solution with ammonium persulfate in the presence of silver nitrate (14,3) . Upon standardization of the chronic sulfate stock solu- tion in this manner, the nonnality was found to be 0.14227. An approximately 0.111 vanadyl sulfate solution was prepared by dissolving 8.0 g. of vanadyl sulfate, VOSO4oxH20, in 200 ml. of water and diluting to 500 ml. The dark blue vanadium solution was standardized with potassium pemnganate (37). The average of four detenninations gave a value of 3.62 i 0.001 mg. vanadium per m1. An approximately 0.0514 solution of sodium nitrite was prepared by dissolving 1.777 g. Hallinckrodt analytical reagent sodium nitrite in 300 ml. of water and diluting to 500 ml. The solution prepared in this manner give a pH of 9.1. Similarly, 3.15 g. Hallinckrodt analytical reagent sodium sulfite was dissolved in 300 m1. of water and diluted to 500 ml. to give an approximately 0.05}! sulfite solution. The result- ant solution was at a pH of 10.1. A 0.1]! sodium sulfide solution was prepared by dissolving 1.78 g. sodium sulfide in 500 ml. of water. Standardization of the sodium sulfide solution was made with iodine (37). The solution was found to be 0.088711 as determined by this procedure. A selenious acid solution was prepared by adding 5.6 g. of selenium dioxide to 500 m1. of water and diluting to one liter. The solution gave a pH of 2.3. The selenious acid solution was standardized by the 15 addition of a measured excess of standardized potassium permanganate and back titrating with standard ferrous ammonium sulfate (ML). The solution prepared and standardized in this manner was 0.0718N. A cobaltous sulfate solution was prepared by dissolving 7.030 g. of Mallinckrodt analytical reagent cobalt sulfate heptahydrate, CoSO4o7H20, in 100 ml. of water and diluting to 250 ml. The solution was standardized by an electrodeposition method (37) and was found to contain 6.092 mg. cobalt per ml. other solutions and reagents were prepared whenever necessary. These are described more fully at appropriate placed in the work carried out . B . Apparatus A Beckman Model H-2 pH Meter, equipped with a glass electrode and a sleeve type saturated calomel electrode pair, was used in adjusting solutions to the required pH values. Spectrophotometric measurements were made with a Beckman Model DU quartz spectrophotometer equipped with a photomultiplier attachment and hydrogen discharge lamp. Preliminary absorbence and per cent trans- mittancy measurements were made with a Beckman Model DK-Z Ratio Recording Spectrophotometer. Generally the Beckman DK-2 Spectrophotometer was used primarily for rapid scanning of the spectrum and for qualitative work; the Beckman DU Spectrophotometer was used for quantitative work. Matched 1-cm.Corex and silica cells were used in the visible and ultra- violet regions of the spectrum respectively. l6 Weights, burettes, pipettes and other glassware were calibrated before use and corrections were applied wherever necessary. C. Stability Studies of Tetralithium Peroxydiphosphate Solutions Some stability studies have been carried out for peroxydiphosphate solutions both at room tenpemture and at 50°C. (ll). It was found that at room temperaimre a 0.11! solution of tetralithinm peroxydiphos- phate was stable for )49 days in basic medium. At lower pH values the solutions were less stable. Peroxydiphosphate solutions of pH 2.0-lO.2 were found to be stable at 50°C. for at least four hours. In addition, it was found that the effect of light increased the rate of decomposition of the peromdiphosphate solution. It was desired to study the extent of this decomposition. , Peroxydiphosphate solutions of various pH values were prepared and exposed to strong sunlight as well as to diffused artificial light approximating working conditions. Twenty milliliter portions of approxi- mately 0.055 tetralithiun peroxydiphosphate solution were pipetted into 250 ml. iodine flasks containing 25 ml. water. The desired pH was adjusted by the addition of 17 per cent phosphoric acid solution. Four pairs of solutions at various pH values were prepared. One pair was exposed to extremely bright sunlight for 1.2 hours at room tempera- ture, another pair was placed in an actual working laboratory containing artificial light, and the remaining pairs were left in the dark for the same period of time as those exposed to the light. After the requisite period of exposure to light, the normality of each solution was determined. diphosphate stock solution was used as the basis of comparison and as a measure of the extent of decomposition of the other perquiphosphate solutions. The results are shown in Table II. TABLEII The original normality of the peroxy~ INFLU'DICE OF LIE-IT ON THE IEOHPOSITION 0F TETRALITHIUM WEEDSPHATE SOLUTIONS OF VARIOUS pH VALUES AT ROOM TEMPERATURE Time of Exposure: 12 Hours Nomfilfiryfiof Nom1ity of Deviation pH LingoB Deviation 1.141920: Parts/Thousand ' in the Dark Parts/Thousand In bright sunlight 0.06180 22.2 11.0 0.061155a 0.0 0.05895 86.8 8.0 0.06h65 1.6 0.05902 85.7 7.5 0.06155 0.0 0 .05895 86.8 7 .0 0 .061.58 0 .5 0 .057h7 109 .7 11.0 o .o 61158 o .5 o .056ho 126.3 3 .0 o .06h15 6.2 In diffused light b 0 .06210 0 .0 10 .0 0 .06210 0 .0 0.06198 2.2 8.5 0.06202 1.3 0.06198 2.2 7.0 0.06195 2.5 0.06198 2.2 6.5 -- - 0.06190 3.2 6.0 0.06192 2.9 0.06165 7.1; h.o 0.06195 2.5 a 0.061.591 1.141530,3 taken as the standard of comparison in the b sunlight studies . 0.062101! Lingo, taken as the standard of comparison in the diffused light studies. 18 The data in Table II clearly show that tetralithium peroxydiphos- phate solutions are unstable when exposed to very bright sunlight. Decomposition of the peroxydiphosphate solutions occurred over the pH range 3.0 to 11.0 increasing narkedly with decrease in pH. Peroxydi- phosphate solutions left in the dark are also included in the table for comparative purposes. These solutions appeared to be essentially stable over the mtire pH range studied except in strongly acid solu- tion of pH 3.0 where decomposition occurred. Table II also indicates exposure of peroxydiphosphate solutions to diffused light results in no appreciable-loss in relative oxidizing power. Maximum decomposition occurred at a pH of 14.0 where a deviation of seven parts per thousand was obtained when compared with the original concentration of peroxydiphosphate stock solution. Thus, when compared with samples of the same pH that had been left in the dark, it can be seen that peroxydiphosphate solutions at various pH values can be left in diffused light for some time before any measurable decomposition occurred, provided a neutral or basic medium was maintained. Studies have been made on the stability of peroxydiphosphate solu- tions at 50°C. It was believed advantageous to try even higher tempera- tures. Preliminary qualitative investigations revealed that boiling of peroxydiphosphate solutions for periods of time greater than two to three minutes resulted in rapid decomposition. However, shorter periods of boiling were apparently without effect. Permcydiphosphate solutions of various pH values and at different concentrations were prepared in 19 a manner previously described. These solutions were then heated to 10000. for five seconds, cooled for one hour and subsequently the amount of peroxydiphosphate was determined following dilution with water back to its original volume. Triplicate determinations were made and the precision obtained was one part per thousand. The amount of peroxydi- phOSphate was expressed in terms of normality. The results are given in Table III along with the normality of the original peroxydiphosphate stock solution used as a means of comparison. TABLE III STABILITY OF 0.1N TETRALITHIUM PEROXIDII’HOSPHATE SOLUTIONS OF VARIOUS pH VALUES WHEN BDILED FOR FIVE SECONDS .— v —-———~ v———- T v_._ __._ wfi— ‘Vfiv — ——-V rr—v— ‘— W Normality of 13413203 i—w— ‘— vv—v— "' v—f w Temperature ‘ I '— pH of [Solution w 1. a ‘ ' __ 9.9 i 7.9 W 33.0 __ h.o 22°C 0.1055 0.0863 0.1058 0.09h6 100°C 0.1055 0.0862 0.1057 0.09112 W -_._w The data in Table III show that peroxydiphosphate solutions of PH 14.0 to 9.9 are stable at 100°C. when heated for five seconds. There apparently is no loss in oxidizing power. This suggests the strong POSSibility of carrying out reactions with tetralithium peroxydiphos- Phate at 100%. without fear of appreciable loss in oxidizing power of the solution. . o The results obtained by increasing the time of heating at 100 C. . o are compiled in Table IV. Heating was maintained at 100 C. for a 2O TABLEIV STABILITY OF 0.111 'JETRALITHIUM PEROXYDIPHOSPHATE SOLUTIONS OF VARIOUS pH VALUES WHEN RELIED FOR SIXTY SECONDS Normlity of M42208 Temperature pH of Solution 3 .0 6.0 6. 8 7 .3 9 .5 23°C 0 .0867 0 .0939 0 .0877 0 .0 819 0 .1163 100°C 0.08511 0 .0930 0.0876 0 .0819 0.1161; maximum period of 60 seconds. Under these conditions of relatively excessive heating the stability of the peroxydiphosphate solution again decreased with decreasing pH. Decomposition of the solution became evident immediately in the acid region. At pH 6.0 the deviation was eight parts per thousand while at pH 3.0 decomposition increased to 15 parts per thousand when compared with their respective pemxydiphosphate stock solutions. Solutions in basic medium showed no decomposition upon heating to boiling. These solutions were stable. I). in Analytical Study of the Reaction Between Perondiphosphate and Manganese(fl) 1. Introduction Numerous anaJytical methods for the precipitation of manganese as the dioxide and the volumetric determination of this compound with a standard reducing agent have been proposed in the literature. From the practical point of view the most important procedures are those in which the precipitation of the dioxide takes place in acid medium; 2l oxidation in neutral or alkaline medium by means of bromine, hypobromite, chlorine, hypochlorite or ferricyanide is unsatisfactory, in general, when iron is present, and even in the absence of iron results tend to be variable. Manganese may also be precipitated by ammonium persulfate in an maniacal solution, or by potassium chlorate in the presence of zinc chloride in a neutral solution (32). In acid solution manganese is oxidized to the dioxide by boiling with ammonium or potassium persulfate. Von Knorre (112) has based a determination on this particular reaction. The precipitated manganese dioxide is collected by filtration, washed, and dissolved in standard ferrous sulfate or hydrogen peroxide and the excess ferrous ion or peroxide is titrated with standard potassium permanganate. The method does not give theoretical values; an empirical factor must be applied. The method has also been applied by Lu‘dert and others (20 ,2? ,31) with Some slight modifications. A method that has enjoyed considerably greater popularity than that of von Inorre is the procedure based on the precipitation of manganese dioxide by long boiling with potassium chlorate in strong nitric acid solution. Beilstein and Jawein (2) first made use of this method of precipitation for a gravimetric determination of the element; Hampe (22) and others based a volumetric estimation on the same reaction. Basque found that potassium bromate could be used in place of the Chlorate but preferred the chlorate. Kolthoff and Sandell (25) recom- mended the use of potassium bromate as the reagent for the oxidation O . . i . O . I . . . ' u .. . . . . J I o . . . . J u I I 22 of manganese to the dioxide claiming reproducible results were obtained when an empirical factor was employed. In every case cited above a considerable excess of the oxidizing agent is used to effect the oxidation of divalent manganese to the dioxide. The results are based on the precipitate formed; the excess oxidizing agent is not determined. The excess chlorate or bromate added to effect the oxidation can not be determined following the reaction because of destructive loss of much of the excess chlorate or bromate in the method used. The bromine or chlorine in excess of that needed to oxidize the manganese in an ammoniacal solution oxidized part of the ammonia to nitrogen and therefore can not be used as a suitable means of determination (104) . Excess persulfate can not be determined because under the conditions used, decomposition occurred. Similar results are encountered with other oxidizing agents. The Volhard method (141) is based on the principle that when potassium permanganate is added to a hot, neutral solution of manganous salt, the latter is oxidized to manganese dioxide and the permanganate reduced to the same form. The complete oxidation of the divalent manganese is evident by the pink color of permanganate persisting in solution. Since acid _is formed in the reaction, zinc oxide is added in order to keep the solution neutral. Fichter and Eladerga'oemdlh) have shown that when peroxydiphosphoric acid is added to a dilute manganous sulfate solution, a violet color is produced. Schmidlin and Hassini (3S) interpret this as the formation 23 of permanganic acid, HMnO4. Fichter and Bladergroenconoludszthat the color is due to the formation of the very stable manganic phosphate (15) as the five characteristic absorption bands of permanganic acid are never observed in a spectrometer; only a uniform darkening from green to violet is observed. From the work cited so far it becomes quite evident that either the manganese dioxide formed or the amount of oxidizing agent required to effect the oxidation of divalent manganese was determined but not both° There has been no case found in the literature where both the amount of oxidizing agent used and the manganese dioxide formed as a result of the oxidation of divalent manganese are determined in a given solution. It was found that when excess slightly basic tetralithium peroxy- diphosphate solution was added to a slightly acidic manganous sulfate solution at room temperature, oxidation of the divalent manganese occurred in the resultant near neutral solution. Evidence for such a reaction taking place was indicated by the immediate appearance of a pale pink color imparted to the solution which became more intense upon standing, ultimately leading to the formation of a brown-black precipitate believed to be manganese dioxide. In addition, it was also observed that in the presence of sufficient excess of peroxydiphosphate a drop of 1.1-1.3 pH units accompanied the oxidation. It was found possible to remove the precipitate and determine the amount of unreacted peroxyh diphosphate using an excess of standard ferrous ammonium sulfate solu— tion and back titrating the excess ferrous ion with O.lOOON potassium dichromate solution. 2h The possibility of using this reaction as a quantitative method for determining manganese was investigated. 2. Identification of the Precipitate Formed As the first step in, the study of the oxidation of divalent manganese the identity of the precipitate formed in the near neutral solution was made. It was strongly suspected to be manganese dioxide. Five milliliters 0.0972N mangancus sulfate solution (with respect to an oxidation change of two) corresponding to 13.35 mg. manganese was pipetted into a 250 ml. iodine flask containing 25 ml. of water. Twenty milliliters of approximately 0.09N tetralithium peroiqrdiphosphate was added and the flask stoppered. Initial pH of the solution was 7.5. An immediate flesh pink color developed in the solution increasing in depth to a deep red then to a dark brown and ultimately ending in the formation of a dense brown-black precipitate. Four such samples were prepared in this manner as well as four blanks. After an overnight period of standing in the dark at room tempera- ture, the pH of the solutions dropped to _6.h-6.5. The precipitates formed were removed by filtration through Gooch crucibles with suction. The precipitates were readily filterable and each was washed thoroughly with'lSO ml. of water in 20 ml. increments. Completeness of washing was tested on the last 50 ml. portion of wash for the presence of any orthophosphate by the addition of ammonium molybdate reagent (36) in a nitric acid medium. No yellow precipitate of ammonium phospho- molybdate was found . 25 Addition of nitric acid did not dissolve the precipitate. Hydrochloric acid, however, did dissolve the precipitate with the evolu— tion of chlorine gas. Acidic ferrous ammonium sulfate solution dissolved the precipitate forming a nearly colorless solution. The precipitate, after dissolving with ferrous ammonium sulfate solution, was found to contain no phOSphOI'uS in the form of orthophosphate when treated with ammonium molybdate . The oxidation state of the manganese precipitate was determined by means of an indirect titration procedure in the following manner. The precipitate was dissolved in 19.98 ml. 0.1032N ferrous ammonium sulfate solution. Nine milliliters 6N sulfuric acid was added and the solution was diluted to a total volume of 150 ml. with water in a 1400 ml. beaker. A magnetic stirrer was used to aid in a more rapid dissolu- tion of the precipitate. After complete reaction, as evidenced by the absence of any particles in solution, the excess ferrous ion was titrated With 0.100011 potassium dichromate solution in the presence of 9 ml. 85 per cent phosphoric acid using diphenylandne sodium sulfonate as indicator. The back titration of the excess ferrous ion contained in two samples required 15.83 ml. and 15 .87 ml. of the standard potassium dichromate solution. The mlmber of milliequivalents of ferrous iron used to reduce the precipitate to a soluble form was obtained from the difference between the number of milliequivalents of ferrous iron originally added and the number of milliequivalents of ferrous iron found after reaction -2 c c. t we . a. a a a W a c a .n m u m. m «(J “a h: A. M“ oulg . . A... C n . hi ~ fl 4 f . +9 , on 0 “D; ‘M has an i! a m p 0 m 6%. I I C . O l O C O C . 7 O C . O I O O C .. . I Q Q p Jurrhuvg; «I . lua‘ s11 ,-Il.ik'll...,r‘rl . unmaslfirfljm futon! . . .:..m, 1f.w.au\fl _ I31. . ...i . u . 4 . . W. , {fps . M. . .... B. A . A. L a, , .. .,, . 2.. . . — \w ,1. L. V . .. . .y ‘. I .c. . . U wit? u . I . . .. 26 with the precipitate. This difference in milliequivalents would also correspond to the number of milliequivalents of manganese present. The average of two determinations revealed that 0.}479 meq. of ferrous iron was used to effect reduction of the precipitate. Since 0.h86 meq. of manganese was taken initially with respect to an oxidation change of two; 1.6., manganese (II) to manganese (IV), the results indicated that the original assumption of having manganese dioxide formed as the product of oxidation was indeed valid. 3. Titration of the Filtrate The filtrates plus washings which were retained following the removal of the manganese dioxide were analyzed for excess unreacted peroxydiphosphate by the addition of a. measured excess of standard ferrous ammonium sulfate solution. After standing for one minute, the excess ferrous ion was titrated with 0.100011 potassium dichromate solu- tion. Blanks, prepared in exactly the same manner as those containing manganous sulfate, were treated with the same amount of ferrous ammonium sulfate solution as those used for the samples. After a standing period of one minute, the excess ferrous ion was back titrated with O.lOOON potassium dichromate solution. The difference in volume of potassium dichromate used in the back titration of the excess ferrous ion in the actual sample and that in the blank represented the milliliters of peroxydiphosphate. Since the dichromate solution used was exactly 0.100011, the milliequivalents of peroqdiphosphate required for oxidation of divalent manganese was readily determined. The number of milli- equivalents of peroxydiphosphate used in the oxidation would also a 3&1: 27 correspond to the number of milliequivalents of divalent manganese oxidized. Any unreacted divalent manganese present in the filtrate did not interfere with the titration. Triplicate deteminations were made. The results on the amount of perom'diphosphate required to effect the oxidation of a known quantity of divalent manganese to some higher oxidation state are shown in Table V. TLBLEV mmUIVAIflTS TETRALITHIUH PEROIIDIPHOSPHATE REQUIRED TO OIIDIZE 13.35 KIILIGRAMS MANGANESE Heq. Heq. Meq. Heq. 1.142203 ‘ Iii-41,205 1.12208 Mn Meq. Mn Calculated for Added ' F U ound sed Present WW) 1 .1421 o .938 o .h83 0 MB 1.1;21 0.93h 0.1m o.h87 0.219 0 .h86 mm 0.226 0&8; 0.385 Av. 0.936 o.h85 0.1;85 The results indicate that, froman average of three analyses, O.h85 meq. of permdiphosphate was required to effect oxidation of the divalent manganese present in solution. This value compared favorably with the theoretical amount of 0.1486 meq. manganese originally added in which the assumption was made that an oxidation change of two occurred resulting in the oxidation to manganese dioxide. The results also sub- stantiated the previous findings for the precipitates. Measurement of the filtrate for excess peroxydiphosphate provided a. satisfactory method for determining the amount of divalent manganese 28 oxidized. Analysis of the manganese dioxide formed also provided a satisfactory method of determining the extent of oxidation. Utilization of both methods were employed as checks against one another in the studies to follow. h. Effect of pH on the Rate of Formation of Mangese Dioxide Preliminary investigation revealed that both the rate of oxidation of divalent manganese by tetralithium peroxydiphosphate and the character of the precipitate formed were greatly influenced by the initial pH of the solution. A series of experiments was conducted to determine the effect of pH on the time required for precipitation to occur. Five milliliter portions of 0.09TON manganous sulfate solution were pipetted into 125 ml. iodine flasks containing 25 ml. water. To each was added 19.98 ml. of 0.07h2N tetralithium peroxydiphosphate solution and the various pH values were obtained by the addition of the required amount of 1N sulfuric acid. Solutions from pH 1.5-7.7 were , prepared in this manner. The length of time required for the appearance of a brown-black precipitate following the addition of the oxidizing agent was measured. The results of pH influence on the manganese oxidation at room temperature, 23°C., at 50°C. and at 100°C. are con- tained in Table VI. The data in Table VI reveal that appearance of a precipitate was immediate in the pH range 5.3-7.7 under all temperature conditions studied accompanied by a drop in pH. The precipitates formed in this indicated pH range were dense in nature, settled relatively fast and .t. Wu .. . a r.“ pres I 3.36. f 29 TABLEVI EFFECT OF pH ON THE OEDATION 0F DIVALENT MANGANEE BY TETRALITHIUH PERDXIDIPHOSPHATE 1311 Time Required for the Formation of an Observable Precipitate Qiinutes) Room empera me, 23W."'13§-55°C'. 100°C. 1.5 360 10 2.0 210 120 < 5 3 .1 60 1&0 immediate h.5 30 < 5 immediate 5.3 < S imediate immediate 6.0 immediate immediate ' immediate 7 .2 imediate immediate immediate 7 .5 immediate immediate immediate 7 .7 immediate immediate immediate Above pH 5. 3 precipitates dense and settled rapidly; easily filtered. Below pH 5.3 precipitates finely divided and colloidal in character. were easily filtered, particularly those that had been heated. However, decreasing the pH below 5.3 at room temperature progressively, resulted in correspondingly longer periods of time before the formation of precipitates were observed. - In conjunction with this work, it was found that various shades of color were obtained prior to the ultimate form- ation of a precipitate. The precipitates formed in this region were finely divided making for difficult filtration. By selecting emerimental conditions in the acid range it was possible to follow the fomtion of the precipitate more closely. Addition of variable amounts of perwqdiphosphate to fixed amounts of 30 manganese were also studied qualitatively. A faint pink appeared in the initially colorless solution deepening in color and intensity with standing to a dark red, then brown and ultimately resulting in precipi- tation of manganese dioxide. No studies were made at pH values greater than 7.7 because brown colloidal manganese hydroxide formed at these higher values. Addition of peroxydiphosphate resulted in an increased darkening of the solution brought about by the partial formation of manganese dioxide. These precipitates were gelatinous in character; apparently incomplete oxi- dation existed. Furthermore, the contimlal decrease in pH until a constant value was reached in the oxidation of manganese can be explained by the reaction Li4Pan + M3504 + 2H20 a 1412804 + 141102 + adj—HEPO4 in which the acidic character of the solution was increased by the formation of manganese dioxide. 5. Proposed Megianism of Oxidation When manganous sulfate was treated with excess tetralithium peroxy- diphosphate solution in an acidic medium at room temperature, the formation of a precipitate resulted after standing for one to two hours. In the time interval from the preparation of the sample solution to the appearance of a precipitate, a series of color changes occurred. By adding an insufficient quantity of peromdiphosphate solution to a 31 fixed amount of divalent manganese, it was possible to retard the color changes somewhat, particularly in the pink-red region. A sample containing 25.59 mg. divalent manganese, 10 ml. 0.1628N tetralithium peromdiphosphate solution, 30 ml. water and 2 ml. 6N sulfuric acid was prepared. The sample prepared in such a manner gave a pH of 1.6 and under these conditions a faint pink color, suggestive of permanganate, was observed in the solution approximately 1.5 hours after preparation. It was found that chromate-free tetralithium peroxy- diphosphate solution did not absorb in the 500- to BOO—mu region of the spectrum. The possibility of obtaining a spectrum for the pink—red color which could then be used as a possible aid in characterizing the mechanism of oxidation was investigated. A sample containing 13.21; mg. manganese, 25 ml. water, 2 ml. 6N sulfuric acid and 19.98 ml. 0.1036]! tetralithium peroxydiphosphate solution was prepared. The solution prepared in this manner gave a pH of 1.5. After a 50 minute waiting period at room temperature, the development of faint pink color was observed in the previously colorless solution. Using the Beckinan 110-2 a per cent transmittAncy spectrum in the 350- to 800-mp region was obtained with a portion of the colored solution ten minutes after color development occurred. Other spectra were made at five minute intervals until the formation of manganese dioxide resulted. A total of eight spectra :was obtained (Figure 1). Examination of the series of spectra obtained revealed absorption characteristics in the region approximating that of permanganate. 32 _ mane... 1.. £52394... 08 com . com 8.. own _ .2.... 3:0 .2... 3.1. .2... 8 no .2... can... .2... no 1. .2! coin .2... 2|“ .2... 2.1 Exo¢ua 2...; 29.5%.. «mt... $2402.... .6 3.6.3 2059.523: .232”... 100 NOISSIWSNVHL .LNBDHBd LOB non . 33 Ihiwever, no definite, well—defined peaks at 525- or 5hO—mu were obtained; lnerely indications of such breaks were observed. The spectrum was shown not to be due to manganic absorption by comparing it with a man— ganic pyrophosphate Spectrum. The manganic spectrum did not show absorption peaks in the same spectral region as the sample. The continual increase in the color intensity with time may be partially responsible for the poor spectrum recorded together with some possible interference from trivalent manganese. It was possible to follow the increase in color intensity with time as shown in Figure 1. This would suggest that if permanganate were present and thus responsible for the color observed, it formed initially as a Very small concentration which increased with increase in time to Some maximum concentration. Failure to observe whether there was a decrease in the concentration, if a maximum concentration value was once reached, must be attributed to the formation of manganese dioxide which obscured the absorption of the solution thereafter. Apparently what must be occurring in a solution of divalent man- ganese and peroxydiphosphate can be postulated in the following manner. Divalent manganese reacts with the peroxydiphosphate to give permanganate followed by the permanganate reacting with the divalent manganese to yield a precipitate of manganese dioxide. The reaction probably involves a rate determining step evident by the persistent pink-red color prior to the formation of a precipitate. It appears the concentration of permanganate must first reach a definite maximum concentration before 3h precipitation occurs. This was evident in the studies conducted previously where various amounts of peromdiphosphate were added to a fixed quantity of manganese and the time required for the formation of a precipitate recorded. These studies on the appearance of a possible permanganate color were trade in acid solution where the pink-red color persisted for a short but measurable period of time. In neutral and slightly basic solutions, divalent manganese oxidized by peroxydiphosphate yielded a light brown color almost immediately. This color development probably obscured the pink-red permanganate color which may have formed initially. As was indicated in previous studies, the rate of reaction depended, to a great extent, on the pH of the solution. 6. Time Regiired to Effect Oxidation at Room Temperature A series of experiments was designed for studying the time required to effect complete oxidation at room temperature. Six pairs of samples containing 13.21; mg. manganese were prepared by pipetting L97 ml. 0.0970N manganous sulfate solution into 250 ml..iodine flasks. Twenty milliliters 0.0860}! peroxydiphosphate solution was added to each flask and diluted to a total volume of 50 ml. with water. In similar fashion, solutions containing 10.23 mg. manganese were prepared by the addition of 1.00 ml. 0.372611 standard manganous sulfate solution to each of twelve 125 ml. iodine flasks. Ten milliliters 0.1022N peroxydiphosphate Solution was added followed by sufficient water to give a total volume of 50 ml. to each sample. A pair of blanks was also prepared for each set at a pH of 5.9. 35 The samples and blanks were stoppered and stored in the dark at room temperature for varying lengths of time to effect oxidation. Temperature was not maintained at a particular value since it was not critical in this region. The solutions possessed initial pH values ranging from 7.2 to 7.5. Formation of manganese dioxide occurred shortly after the addition of excess oxidizing agent passing through the color changes noted and described previously. At various times after preparation, pairs of samples containing the precipitate of manganesa dioxide and excess peroxydiphosphate were analyzed for complete- ness of oxidation. The precipitate was removed by filtration through a Gooch crucible with suction and washed thoroughly with distilled water. Both the combined filtrate and washings and the precipitate were analyzed. The results of the analysis at various time intervals denoting the quantity of divalent manganese oxidized are shown in Table VII. The data presented show that, as expected, increased periods of time gave progressively better results. After ten hours complete oxi— dation of the manganese was effected as indicated in Table V11. The average of the last six determinations for the 10.23 mg. manganese samples in the time interval 12-2h hours résulted in values of 10.21 mg. and 10 .13 mg. for the filtrate and precipitate respectively. The average of these six precipitates was approximately one per cent below the theoretical value, strongly suggesting that an empirical factor be applied to the precipitates (25). Incomplete oxidation was observed visually by the faint brown imparted to the filtrate following removal of the precipitate. The absence of any color imparted to the filtrate and the 0 u. o u I e - e , v u c v _e , o . . .. e e o I ‘. r I l - TIME STUDY ON THE OXIDATION OF DIVALFNT MANGANESE TABLEVII AT ROOM TEMPERATURE 36 Time Hg. Mn )1 . Hn Found Per Cent Error Hrs. Added Filtrate Precipitate filtrate Precipitate l 13.21; incomplete 2 13.21; 11.62 10.82 ~12.211 -18.28 11.59 10.78 -12.116 -18.58 h 13.21-'- 12022 1101-60 - 7070 “'13089 12.16 11.311 - 8.16 -1h.35 5 13.211 12.71 11.70 - 11.00 -11.56 12.77 11.75 - 3.55 -11.25 6 10023 9.88 9031-1» "' 30142 - 8.70 9.86 9.28 - 3.61 "' 9029 8 10.23 9.91. 9.61 - 2.83 - 6.06 9091-1- 19061-1, "' 2083 " 5-77 9 10.23 10.00 9.69 4.2.25 - 5.26 9.914 9.89 - 2.83 - 3.32 10 13.26 13.21; - 0.15 13 .18 - 0.60 12 13.26 13.26 0.00 13.21 0.38 10.23 10.21. 10.16 + 0.09 - 0.68 10.211 10.11 + 0.09 - 1.17 15 10.23 10019 10.11-L "‘ 0039 - 0.88 10.16 10.10 -' 0.68 "’ 1.27 214 10.23 10.22 1001.3 - 0.09 " 0.98 10022 10.16 " (1009 " 0.68 constant pH values of 6.0-6.3 indicated complete reaction had occurred. Duplicate determinations were therefore run under a rigid time schedule to ndnimize continued oxidation of the manganese left in solution. The same procedure as above was employed for various concentrations of manganese up to approximately 50 mg. A period of 12 hours was arbitrarily selected as the time required for complete oxidation. larger SiZBt complete reaction was increased to 11; hours. Table m e OXIDATION OF VARI'JNG GIANTITIES 0F MANGANESE AT ROOM TEMPERATURE FOR TWELVE FDURS TABLEVIII Results are given in samples of manganese, the time of standing to effect Mg. Mn 14 . Kn Found Per Cent Error Added m ra. e rechpi 51.38 50 .38 149 .98 -1.98 -2 .72 50 053 119.98 ”1.614 ‘2072 11.1052 141.38 160.56 '00314 ”'2 .31 141.33 30 .50 -0 .116 -2 .116 25.68 25119 25.22 -0.7h -1.79 20 .146 20 .32 20 .10 -0 .68 -1.76 20 .35 20 .19 -o .511 -1.32 13 .211 13 .13 13 .25 -0.83 +0 .08 13.21. 13.18 0.00 -0.hS 13 029 13 .08 4'0 0’45 "1 .21 10 .23 10 .19 10.10 -0 .39 ~1.27 10 .13 10.02 "0097 '2 .05 5.12 5.11 5.08 -0.19 -0.78 5.08 5.20 -0.78 +1.57 Sol-h 14.96 +0 039 -3 012 As evident from the table, the results obtained were fairly good. Quantitative oxidation was not obtained for the largest size sample uSed. The form of the precipitates in the larger samples was more 38 finely divided in nature resulting in extremely prolonged periods of time for effective removal of the precipitate. Increased time of standing did not materially improve the results obtained. The long period of standing required to effect the complete oxidation suggested that a search be made for a more rapid method of oxidation. 7. Time Reflired to Effect Oxidation at 50—55%. Chulski (11) suggested the possibility of carrying out oxidation reactions at 50-5500. with perom'diphosphate. The quantitative oxi- dation of manganese was studied in this temperature range. Six 1L.97 ml. portions of 0.07ON manganous sulfate solution (13.23 mg. manganese) were pipetted into 250 ml. iodine flasks containing 25 ml. water. Twenty milliliters of 0.1112111 tetralithium peroiqrdiphosphate Solution and 0.05 ml. 1N. sulfuric acid were added to each flask. Preparation in this marmer gave solutions with pH values of 6.6. The iodine flasks were stoppered and placed in a constant temperature oven set at 50° i 5°C. Six blanks at a pH of 6.0 were also prepared in enctly the same manner as the samples. After an hour at 50°C. , two samples and blanks were removed and allowed to cool to room temperature. Thirty minutes cooling at room tempera’mre was sufficient. At this point the samples were at a pH of 6.1. The manganese dioxide precipitates were filtered through Gooch crucibles with suction. The precipitates were washed with 150 m1. of Vater in 30 m1. portions. Both the filtrate and washings were retained. The precipitates were analyzed by the ferrous ammonium sulfate-potassiwn 39 fichromate method described earlier. Similarly, the filtrate and washings were determined for the amount of peroxydiphosphate used in the oxidation from which the amount of manganese could then be calculated. The same procedure was carried out after two hours at 50°C. with another set of samples and blanks. The precipitates that had formed were much more voluminous. The solutions were at a pH of 5.8. Both the precipitate, which was recovered, and the filtrate and washings were analyzed in the usual manner for the amount of manganese which had been oxidized. The third set of samples and blanks were removed from the constant temperature oven after four hours. The cooled solutions gave a pH value of 5.6. Both the combined filtrate and washings and the precipi- tate of each sample were analyzed for manganese in the usual manner. The results of the six samples analyzed at the above prescribed time intervals are contained in Table II. TABLE I: _’ OXIDATION. OF matures AT 50-55°c. FOR VARIOUS PERIODS 0F mm ¥ Time Hg. Mn )1 .' Mn Found Per Cent Error Hrs . Present filtrate Frecipi$ Filtrate PrecipiEte 11. 130211 12.06 12.01. ‘8091 .9029 12 .1]. 12 .06 -8 .53 “'8 09:1. 2 13.21. 12.1w 12.32 -5 .82 -6.95 1211 12 .111 -—6.25 , -6.25 1. 13 .21. 13 .21 13 .13 4) .15 -0 .83 13.23 13.15 -0.08 ~0.68 ho Examination of the data reveal that oxidation was not complete after two hours at 50°C. but that quantitative oxidations were obtained for samples maintained at 50°C. for four hours. The progressive decrease in pH with time was also taken as an indication of incomplete oxidation. The results show that treatment for four hours at 50°C. was adequate for complete oxidation of the manganese samples. A series of additional runs containing varying quantities of man- ganese and sufficient excess peroxydiphOSphate solution were prepared and allowed to remain at a constant temperature of 50°C. for four hours. The solutions were within a pH range of 6.8-7.2 initially. Both filtrate and precipitate were analyzed for manganese in the same manner as previously described. The results of 11 runs containing various amounts of manganese are compiled in Table X. From the numerous samples of manganese which have been run both at room temperature and at 50°C., the amounts of manganese in the precipitates were found to be all approximately one per cent low, strongly indicative of the need for an empirical factor. Utilization of a factor equal to 1.01 was made on the precipitates to bring them more in line With the values found for the filtrate. This was done for the results in Table I where the corrected values of the manganese obtained in the precipitates are shown together with the actual values obtained from the precipitates. The latter was included primarily for comparative Purposes. It is readily seen that application of the empirical factor on the precipitates results in good agreement with the theoretical. The factor was used hereafter in all the work to follow. TABLEX ommom 0F MANGANESE AT 50°C. mm FOUR HOURS Mg. Kn Hg. Mn Found Present F' tra Precipitate Precipita e with Factor = 1.01 10.23 10.16 10.08 10.18 10 .16 1o .16 10.26 10.21; 10.00 10.10 10.13 10.13 10.23 10.13 10.08 10.18 10.27 10.10 10.20 --- 10.13 10.23 Av. 10.19 10.10 10.21 13 .214 13 .211 13 .11 13 .211 13 .18 13 .07 13 .20 Av. 13.21 13.09 13.22 20.1.6 20.51 20.25 20.115 20.110 20.20 20.110 20.38 20:22, 20.112 Av. 20.113 20.22 20.112 By maintaining a temperature of 50°C. for four hours, the time required for oxidation was materially reduced. The precipitates formed at this elevated temperature were observed to be much more dense in nature and subsequently found to be readily filterable. As seen from Table I the results on the filtrate and precipitate for manganese agreed quite well with the theoretical value. 8. Oxidation at Elevated Tmemtures Additional studies were made on the oxidation of manganese at temperatures of 80—8500. Six 1.00 ml. portions of the standard manganous 112 sulfate stock solution containinglO.23 mg. manganese per ml. were pipetted into 250 ml. iodine flasks followed by sufficient excess peroxydiphosphate solution to effect oxidation; 9.97 ml. 0.1011111 was deemed sufficient. The total volume of each sample was then made up to 50 ml. with water. The solutions were initially at a pH of 6.8. The samples were then heated over Tirrill burners until a temperature of 80-8506. was reached. Treatment of the samples in this manner resulted in the immediate fomation of the brown-black manganese dioxide. The samples were allowed to cool to room temperature for a period of time ranging from 20 to 30 minutes. Blanks were also prepared and treated in exactly the same manner as the samples. The blanks were adjusted to an approximate pH value of 6.0 with 111 sulfuric acid. Following this period of cooling, during which time the pH of the solutions dropped to 5.11-5 .6, the precipitates were filtered with suction on Gooch crucibles and thoroughly washed with water. The time required for this operation was usually 25 to 30 minutes. The total quantity of water used for each precipitate was approximately 100 ml. The procedure of analysis on the filtrate, supplemented by the blank, and the precipi~ tate have been discussed previously. In similar fashion a series of samples containing different amounts Of manganese were also prepared and treated with sufficient excess Perondiphosphate to effect oxidation. Five samples containing 20 .116 mg. manganese, four samples containing 30.70 mg. manganese and four samples containing 111.214 mg. manganese were studied in exactly the same manner “521153 LLB as previously described for the 10.23 mg. manganese sanples except the cooling time after heating to 80-8500. was increased to 1.0 to 1.5 hours. The results found for all these samples of different manganese concentrations are compiled in Table XI. Similar studies were made at a slightly increased temperature. Two samples containing 10.23 mg. manganese and two sanples containing 15.140 mg. manganese were treated with 19.98 ml. 0.0876N perondiphos— phate solution and were diluted to a total volume of 50 ml. with water. The samples were heated to 100%. and were maintained at that temperature for ten seconds. The solutions were cooled to room tenperature for 30-15 minutes. Both the filtrates and precipitates were determined in the usual manner for the amount of peroxydiphosphate used in the oxi- dation of manganese and the quantity of nanganese present as manganese dioxide respectively. The results are also included in Table XI. Examination of the data in Table II revealed that the amount of manganese found from the filtrates compared favorably with the theoreti- cal in all cases studied. This also proved to be true for the manganese in the precipitates following an appropriate correction factor. Increased length of standing at room temperature did not affect the results obtained on the filtrates nor on the precipitates. The data in the table also reveal that there was little to be gained by increased heating to 10000., heating to 80-8500. was adequate for quantitative results. TABLE II OXIDATION OF HANGANEE AT ELEVATED TEMPERATURES Hg. Hn H . Hn Found Added Ffitrate Precipitate Frecipitate with Factor 1.01 it 80-85%. 10.23 10.211 10.13 10.23 10.19 10.16 10.26 10. 10.16 10.26 10.19 10.16 10.26 10.19 10.10 10.20 10.21 10 .10 10.20 Av. 10.22 10.11. 10.211 20.146 20.h6 20.26 20.116 20. 20.21; 20 .11; 20.118 20.32 20 .52 20 .113 20.30 20.50 20.1.6 20.26 20.411_6_ 1v. 20.115 20.27 20 .147 30.70 30.62 30.51; 30.811 30.70 30.51. 30.811 30.73 30.56 30.86 3943 30.28 30.58 Av. 30 .70 30 .118 30 .78 141.211 111.211 110.70 111.10 111.35 ho. 1.1.21 111.13 110.83 111.211 111.211 ho.86 111.27 1"- Ill-2,410.80 141-205 At 100°c. 10.23 10.211 10.16 10.26 10.27 10 .13 10.23 1v. 10.26 10.15 10.211.5 15.1w 15 .113 , 15.211 15.39 5.110 15 .30 lS-_h§ Av. 15 .111.5 15.27 15.112 115 9. Recommended Volumetric Procedure The results of all these preceding experiments provided sufficient background data for the development of a suitable volumetric determin— ation of manganese following oxidation of divalent manganese with excess tetralithium peroxydiphosphate solution. The recommended procedure for the determination of manganese is !— as follows. The requisite volume of solution containing divalent manganese, in the form of the sulfate and acidified to a pH of B-h, is added to I a 250 ml. iodine flask. Convenient quantities for the determination range from S to 50 mg. of manganese. A 2.5-to}.0-fold excess tetra.- lithium perondiphosphate solution is then added to effect the oxidation of the manganese to manganese dioxide followed by the addition of sufficient distilled water to give a total volume of 50 ml. Addition of the oxidizing agent results in a faint pink color imparted to the solution, turning progressively darker with time. The solution sample oxidized in this manner should have an initial pH in the range 7.0 to 7.5. Heating over a Tirrill burner is then carried out until the first Sign of boiling occurs. it this elevated temperature the immediate formation of a hydrous brown-black precipitate of manganese dioxide was Obtained. Blanks, containing the same quantity of peroxydiphosphate as added in the sample, are also prepared and diluted to the same total VOlume of 50 m1. These blanks are adjusted to a pH of 6.0 with 1N Sulfuric acid and carried through the entire procedure in the same manner as the samples. 1:6 Following a. 145 minute cooling period, the precipitate is filtered on a Gooch crucible previously washed with dilute sulfuric acid and thoroughly with water. The precipitate is washed with 150 ml. of water in 30 ml. portions. Both the filtrate and washings are retained. The precipitate, after washing, is transferred to a 1400 ml. beaker, dis- solved in a known quantity of excess ferrous ammonium sulfate followed by 9 ml. 1811 sulfuric acid and diluted to 200 ml. with water. Magnetic stirring is used in hastening the dissolution of the precipitate. The excess ferrous ion is titrated with 0.1000N potassium dichromate solu- tion in the presence of 9 ml. 85 per cent phOSphoric acid and using diphenylamine sodimn sulfonate as the indicator. The weight of manganese present in the precipitate can readily be found from the difference required to titrate the same volume of ferrous solution in the absence of the precipitate. The combined filtrate and washings of 200 ml. is analyzed for the amount of peroxydiphosphate remaining after the oxidation 0f divalent Mganese to manganese dioxide in the following manner. Nine milliliters 18H sulfuric acid is added to the filtrate and washings followed immediately by a measured quantity of ferrous ammonium sulfate. After a minute 9 ml. 85 per cent phosphoric acid is added and the excess ferrous ion is back titrated with O.lOOON dichromate solution using diphenyl- amine sodium sulfonate as indicator. The blank is titrated in exactly the same manner following the addition of the same quantity of ferrous solution as used in the sample. FI‘om the difference in volumes of 3,0 0.1000N potassium dichromate used in the blank and the sarrple the amount of manganese can be calculated. 10. Effect of Foreig; Ions It was desired to study the effect of other ions on the manganese oxidation. The solutions used in studying the effect of various ions on the manganese oxidation were prepared from reagent grade salts which were essentially free from manganese. The solutions generally contained 1 or 2.5 mg. of the desired ion per ml. Wherever possible salts contain- ing sulfate as the anion were used. Solutions containing a fixed amount of manganese in the divalent form, a measured volume of the standard solution of the foreign ion and sufficient excess tetralithium peroqdiphosphate to effect oxidation of the manganese, were diluted to 50' ml. These solutions were adjusted to pH values of 6 or 7. These series of solutions prepared in such a manner were heated to boiling for five seconds and then allowed to cool for 145 minutes. The precipitate formed was filtered off and the amount of manganese was determined. From the amount of peroxydiphosphate used to effect oxidation of the manganese, the manganese could be calcu- lated from the filtrate. The results are tabulated in Table III. The data in Table III indicate that the determination of manganese can be made in the presence of quite a number of ions. The amount of mnganese found in the presence of these various individual ions agreed quite closely with the theoretical quantity added initially. TABLE III DETERMINATIDN OF MANGANESE IN THE PRESENCE OF OTHER ELEMENTS 148 Forei Ion Ion fig. Added Present Hg. Mn Per Cent Error e Ffiltrate Precipitate Vanadium(v) 1 1 1 vanadyl h to Cupric 10 Calcium 10 Cadmhnn 10 20 30 lickelous S 10 10.23 15 .35 25 .59 10.23 10.23 10.23 10.23 10.23 10.23 10.23 10.23 20.18 0 En FOImd ra e rec pi with Factor 1.01 10 .22 10 .26 10.27 10.22 15 .32 15 .32 15 .32 15 .17 15 .3). 15 .51 25 .57 25.59 25 .59 25 .57 25 .59 25 .514 10 .16 10 .21; 10 .22 10 .10 10 .16 10 .23 10 .21; 10 .21 10.16 10.2h 10.19 10.23 10.21; 10.18 10.30 10 .23 10.21 10.15 10 .27 10 .20 10 .21 10 .23 10.27 10 .23 10.21 10.23 10 .21. 10 .29 1o .27 10 .26 10 .27 10 .26 10.21 10.23 10 .21 10.26 10.26 10.26 10 .08 10 .15 10 .27 10 .10 20 .16 20 .10 20.29 20.02 -0.09 +0.39 -0.19 40.19 -0.07 -0.08 0.00 0.00 -0.68 -0.09 -0.68 +0.09 —0.68 -0.39 +0.09 +0.69 -0.19 +0.39 -0.19 +0.39 -0.19 +0.09 +0 039 +0.39 -0.19 -0.19 +0.29 '1 0’4? +0.39 -0 .09 +0.5h . .1.-. 0'1: x v I. . v . v- ‘5 r o o .C “5. x in! . . TKELE‘XII - Continued Hg. Mn Forei Ion Ion Eg..Aaded Present H . Mn Found Fiitrate P recipitate Filtrate Precipitate With Factor 1.01 Per Cent Error v_1—_ Sodium 20; Potassium 10: Zinc 2 Kolybdate 6 Phosphate, dihydrogen 50 Chromic h 5 Arsenic(III)25 cmmnme 25 Mercuric 25 Ammonium 10 10.23 10.23 10.23 10.23 10.23 20.26 20.18 20.18 20.26 10.23 10.27 10.32 10.21 10.22 10.20 10.26 10.21 10.19 10.17 10.26 10.29 10.21 10.26 10.2h 10.19 10.11 10.17 10.20 10.16 110.26 10.16 10.21 20.30 20.21 20.10 19.88 20.15 20.11 20.32 20.32 10.18 10.15 \ +0.39 +0.88 -0.19 ~0.09 -0.29 +0.29 -0.19 .0 .39 —0.59 +0.29 +0.59 -0.19 +0.29 +0.09 -O.39 "l .17 -0.59 -O.29 -0.68 +0.29 —0.68 -O.l9 +0.22 -0.25 -0.39 -l.h8 ~0.15 ~O.35 -0.29 -0.29 -0.h9 -0.78 No interference was observed in the deterndnation of manganese based on the filtrate or the precipitate when the following ions were added in the amounts indicated. 5o vanadium (V) 1 mg. calcium 10 mg. cadmium ho mg. sodium and potassium 20 mg. + 10 mg. vanadyl ho mg. zinc 2 mg. cupric 10 mg. molybdate 6 mg. nickelous 10 mg. phosphate, dihydrogen 50 mg. Proportionally larger quantities of the foreign ions resulted in incomplete oxidation evident by colloidal precipitates formed in most cases and erratic and irregular titration data. High concentrations of copper and cadmium, 50 and 100 mg. respectively, resulted in highly colored filtrates following removal of the precipitates. These colors were bright red suggestive of permanganate. All the ions which inter— fere with the ferrous ammonium sulfateupotassium dichromate back titration in the analysis for peroxydiphosphate used in the oxidation must be absent. It was this apparent limitation on the analysis on the filtrate that led to Spectrophotometric methods of analysis to be described in the next section. In cases where no analysis was possible on the filtrate due to oxidation of the foreign ion or interference of the ion with the back titration procedure, the precipitate provided a means for determining the amount of manganese oxidized. Ions which interfered in the analysis of the filtrate but not in the precipitate included: 'mercuric 25 mg. arsenic(III) 25 mg. chromic 5 mg. chloride 25 mg. ammonium 10 mg. 51 The ions which definitely interfered in the'oxidation and hence, effected both the filtrate and precipitate were: tungstate, thallous, cobaltous, ferric, aluminum, silver, barium, and nitrate. Tungstate, ferric, and aluminum ions actually retarded oxidation. D1 these cases no definite precipitations occurred; only a persistent red color was observed in solution marked by an apparent colloidal su3pension of undetermined nature. Lead, barium, silver and zinc ions formed insoluble peroxydiphos- phates (16) which would effectively reduce the amount of peroxydiphos— phate in solution. Furthermore, the insoluble peroxydiphosphate salts would be carried over in the manganese dioxide precipitate resulting in large positive errors when the precipitates were determined for man- ganese. It was found, however, that zinc in small quantities, l-3 mg. , did not interfere. E. Spectrophotometric Determination of Peroxydiphosphate It was noted that solutions of peroxydiphosphate absorbed in the ultraviolet region. This opened up the possibility of a spectrophoto— metric method for determining the excess peroxydiphosphate in solution following a given reaction. This would be particularly true of the manganese oxidation where the manganese dioxide could be filtered off and the excess peroxydiphosphate in the filtrate determined spectro- photometrically. Preliminary absorbance and per cent transmittancy measurements were made with a Beckman IK-Z Spectrophotometer. Lithium peroxydiphosphate 52 solutions at various concentrations were prepared and absorbance measure- ments made in the ultraviolet region from 220- to BhO-mu using water as the reference standard. No definite absorption peaks were obtained in this region of the spectrum, merely a strong dependence on concen- tration. Figure 2 shows the characteristic type of curve obtained. The following sets of experiments were carried out to establish definitely that the curves obtained in Figure 2 were truly due to the peroxydiphosphate and not to the decomposition products. Spectra of monohydrogen phosphate, dihydrogen phosphate, pyrophosphate and phosphoric acid were obtained in the 220- to 3hO-mu range. The different phosphates and pyrophosphate used were in the form of sodium salts. The solutions on which such measurements were made contained approximately 8-10 mg. phosphate ion in its various forms and 1.7 mg. pyrophosphate ion per ml. at pH values ranging from 8.1 to 8.3. These solutions of anions were transparent throughout the spectral range studied in agreement with the work of Buck and COdworkers (6). Furthermore, the deliberate addition of phOSphate or pyrophosphate to a peroxydiphosphate solution caused no change in the characteristic shape of the absorption curve. The phos- phate was added in this instance either as the disodium or monosodium phosphate. The fact that no absorption peaks were found in the ultraviolet region for peroxydiphosphate made it necessary to arbitrarily select a suitable wavelength for the Spectrophotometric determination. The selection of a suitable wavelength was achieved by preparing a series 53 m maze... 1: .Eozunmzi 9.» can com com com 03 can 08 _ _ _ _ u :nb. xodlv 2923.7» 2 3:3.qu z .b. .2... n. «d In .mzoEEzmozoo 39.33 E uk ._.< uhx0mum c0... mza<¢0 20....(«0120 0.0 0“ O 00.0 33NVBHO 88V 56 In conjunction with the above work, it was found that these solu- tions of varying concentrations adhered to Beer's Law in every instance. Since many of the reactions with peroxydiphosphate were to be carried out under near neutral conditions (pH between 5.5-7.5) a working cali- bration curve was prepared at a pH of 6 at 2110 mu and at a slit width of 0.70 mm. (Figure h). The concentration region covered by this plot was from 0.000 to 0.03111], sufficient for the method of determination of peroaqrdiphosphate. The peroaqndiphosphate solutions, previously standardized by the ferrous ammonium sulfate method, were heated to boiling for five seconds, cooled for 16 minutes and 20 minutes after adjustment to pH 6.0 the measurements were made. Water at a pH of 6.0 was used as the reference blank. It was qualitatively observed that decreasing the pH to lower values resulted in absorbance measurements which varied appreciably as evi- denced by the change in slope of the spectra obtained. This was mentioned previously in connection with the work involved in selecting a suitable Working wavelength. Further studies were made to determine the effect 0f pH on the Spectrophotometric analysis of peroxydiphosphate. A solution of lithium perozqdiphosphate was prepared by dissolving approximately 11 g. of tetralithium peroxydiphosphate tetrahydrate in water and diluting to 250 ml. The pH of the solution prepared in this InIE‘I-Tlner was 9.6. The solution was standardized by the ferrous ammonium Bl-llLfate method and found to be 0.121511. Solutions with various pH values were prepared by adding various a“Haunts of l? per cent phosphoric acid to 10 ml. portions of this above 57 00.00 . c mmzoc n0. x 5.35232 .zofifiFzmozoo 00.0.v 00.0w 00.0w 00.0. 00.0 _ . _ 1 l 1 8 x... 00N.0 0 0 ¢.0 0 0 0.0 O O m 0 8V 000.. 00 N. 00¢.— ww 092 ‘aonvauos 00 0.. 000.. miramozaa>xomuu ...0 +£5.10 zo....x0mun. ....0 uoz....0.0< I a 0.0 0.0 o... 0.0 0.0 of 0.» 0.~ 0.. _ . . . _ . . 0.0.. 8 32:09. cut! .53..» . 0.04 .0 20....an cur... .czuna 10050 0.0.. 8 20534 SE: .530... 1000.0 L800 l 000.. 3ONVBUOSBV 61 1. Determination of Excess Pero ' ho hate in the Man anese Oxidation The excess peroxydiphosphate following the reaction between lithium peroxydiphosphate and divalent manganese was determined Spectro- photometrically. This involved the addition of a measured excess of standard peroxydiphosphate solution to a fixed quantity of manganous sulfate, heating at 100°C. for five seconds and then allowing to cool to room temperature for 30 minutes. Following the removal of the man- ganese dioxide formed by the reaction using either a 60F fritted funnel or a 60M fritted funnel covered with asbestos, the filtrate was diluted to a known volume, generally 200 ml. , such that the absorbance obtained would fall somewhere in the calibrated portion of the previously prepared graph (Figure 1;). The diluted filtrate was adjusted to a pH of 6.0 and, after a 20 minute period of standing, the absorbance was measured. Subtraction of the peroxydiphosphate found in this manner from the quantity of peroxydiphosphate originally added to the reaction gave the amount of manganese which had been oxidized. Samples containing from 10 to 100 mg. manganese were oxidized with excess peroxydiphosphate. The results, corrected in the manner described above, are contained in Table XIII along with the method used to effect the removal of the precipitates formed following the oxidations. The original peroxydiphosphate solution used in the above work was standardized by the ferrous ammonium sulfate method and also spectro- photometrically by the standard calibration curve. Comparison of the two methods showed no appreciable difference. Standardization of a — 62 TABLE XIII INDIRET WHON 0F MANGANESE SPECTROPHOTOMETRIGAILY Determinations Hg. Mn Added Hg. Mn Found Per Cent Error L. Filtered with 6014 Fritted Funnel Covered with Asbestos. L; 10.10, 10.08 -0.25 b 20.18 20.30 +0.5h 2 3o.h8 30.51; +0.19 3 to .67 ho.9h +0.67 3 50.66 51.13 +0.93 B. Filtered with 60F Fritted Funnel. h 10.105 10 .10,5 0.00 h 20 .18 20.11; -o.19 b, 30.}48 30.55 +0.23 6 1.0.67 ho.91 +0.59 2 So .66 So .73 +0 .114 2 101.7h 101.56 -0.18 peroxydiphosphate solution of pH 6.0 by the chemical method gave an average vflue of 0.1911511 for three determinations with a deviation of one part per thousand. This same peroxydiphosphate solution was quanti- tatively diluted lo-fold and the absorbance measured. From the standard curve prepared previously in terms of the chemical method, the normality was determined. The average of three such measurements resulted in a value of 0.1910)! with a deviation of four parts per thousand. Comparison of the normality of other peroxydiphosphate solutions by the two methods were found to give similar results. 63 The removal of precipitates by filtration through asbestos covered 60H fritted funnels or through 60F fritted funnels did not effect the results obtained. This was shown in Table XIII where the calculated per cent error of various amounts of manganese taken for oxidation studies ranged for 40.25 to +0.93. As generally expected, increasing the amount of manganese for oxidation gave progressively poorer results. The indirect determination of peroxydiphosphate used in the reaction provided a means for estimating the amount of manganese oxidized. The results were satisfactory. To further verify the fact that the results were reproducible, a series of peroxydiphosphate solutions was prepared and determined spectrophotometrically under exactly the same experimental conditions as the manganesedperoxydiphosphate samples. In addition, the effect of temperature on the absorbance readings of the individual samples was simultaneously studied. Ten milliliters portions of 0.1056N lithium peroxydiphosphate standardized by chemical methods were pipetted into 50.00 ml. volumetric flasks and diluted quantitatively to ark. The pH of each sample was adjusted to 6.0 by the addition of 1 ml. 17 per cent phosphoric acid. The samples were heated at 100°C. for 10 seconds and then allowed to cool for 30 minutes. The peroxydiphosphate content was determined at the preselected wavelength of 2h0 mu and slit width of 0.70 mm. The solutions were read 1.5 hours after preparation. The results are tabu- lated in Table XIV at a temperature of 23° i 3°C. It is seen from the table that the results of 10 determinations Inhllilxlrdl ‘ III I 6b TABLEXIV REPRODUCIBILITY OF DETERIUINATION OF PEROIIDIPHOSPHATE SOLUTIONS Difference Normality M42205,“1 Heq. LiJ’an d Faun Heq. 1.14on Foun Absorbance (Corrected) (Found - Taken 0 .852 0 .02132 1.062 +0 .006 0.838 0.02097 1.01.6 -0.010 0.832 0.02082 1.01:1 -0.015 0.81.0 0.02102 1.051 -0.005 0.839 0.02100 1.01.6 -0.010 0 .8h0 0 .02102 1.050 -0 .006 0.800 0.02102 1.050 -0.006 0.838 0.02097 1.01;? -o.009 0.810 0.02102 1.0h8 -0.008 0.8h1 0.02105 1.0h8 -0.008 Av. 1.0h8 Av. dif. 0.008 0" 0.005b a'Normality 1:142:03 taken: 0.02112n Heq. Lingo, taken: 1.056 bO’ - average- oun n with no regard to temperature gave an average value of 0.10118 meq. for lithium peroxydiphosphate. The average difference behreen the samples was only 0.008 meq. and the standard deviation was 0.005 meq. The series of deteminations indicated that temperature variance was of no appreciable consequence in the absorbance measurements. 65 The successful determination of manganese by the indirect method of spectral analysis was accomplished in the absence of any interfering ions. Several ions were studied for possible interference. It was possible to immediately eliminate the study of ions which absorbed at 21.10 mu (6). Moreover, ions that manually interfered with the manganese oxidation were not studied except for those which showed negligible effect when present in small amounts. As a result of these considera- tions, the following ions were studied as major constituents; nickelous, zinc, cobaltous and magnesium. Both cadmium and tungsten ions were considered as minor constituents. Ellie ions were added as the metallic sulfates except tungsten which was added as sodium tungstate. Solutions used in the interference studies were prepared by adding the diverse ion or a combination of ions to a solution containing A approximately 10 mg. manganese. The solutions were oxidized in the same manner as before with excess peroxydiphosphate. Filtration was effected using medium porosity fritted funnels covered with asbestos. Lbsorbance measurements were mde at 210 mp 20 minutes after adjustment of the pH to 6.0 with 17 per cent phosphoric acid. Blanks containing peroxydiphosphate and the foreign ion or ions were also prepared and carried out in similar fashion. Data for these determinations of man- ganese are given in Table IV. Duplicate determinations were made. The data in Table IV show that the indirect determination of man— ganese spectrophotometrically yielded good results despite the presence of the foreign ion'or ions listed. The manganese dioxide‘formed in all n\ A I J J . J . ) a . . _ c . . .. I 1 u . . . o . J .. o J 1 o . ) J . . . J 1 _ 1 Q ‘. I ‘vh‘nr I. i‘ u ' .I - ‘I l I, v I. [III-- TABLEXV INDIRECT DETERMINATION OF HANME'SE SPECTROPHOTOETRICAILY IN THE Pm OF FOREIGN IONS. 10 .105 ICELLIGRAH MANGANESE ADDED Foreign Ion Added Hg. Foreigi Ion Hg. Mn Found Tungstate 0.08 10.2h Zinc 5 .0 10 .13 Nickelous S .0 10 .16 nagnesium S .0 10 .ll Cobaltous S .0 --- Cadmium O .08 10 .18 a 10 .13 b 10 .06 c 10 .08 d 10 .01 80.08 mg. each of Cd(II) and tungstate. b5.0 mg. each of Zn(II), Ni(II), and Hg(II). °5.0 mg. each of Ni(II), A1(III), Zn(II), Hg(II) and 0.08 mg. each of Cd(II) and tungstate. d12.5 mg. each of 111(11), A1(III), 2:1(11), ng(II) and 0.20 mg. each of Cd(II) and tungstate. cases was dense and easily filterable except when cobalt was present. The manganese dioxide precipitate formed in the presence of cobalt was gelatinous and too finely divided to effect removal. As a consequence, no suitable analysis of this particular filtrate could be made. In samples c and d, the precipitates were also determined by the previously described method for manganese. The amount of manganese in 6? these individual precipitates was found to agree quite well with that found in their corresponding filtrates from the spectral analysis work. F. Attempts at Oxidation of Other Ions It was noted in the course of investigating the interferences associated with the manganous oxidation that thallous as well as chromic ions were oxidized by peroxydiphosphate. In the case where chromic ions were present, a yellow color characteristic of chromate was evident in solution following removal of the manganese precipitate. Correspondingly, the results on the detemination of the manganese precipitate and the excess permdiphosphate in solution were erratic in the presence of thallous ionS. An investigation was carried out to study the possibility of oxidiz- ing thallous and chromic ions individually with peroxydiphosphate. 1. Oxidation of Tha111umg12 The time required for the appearance of a precipitate at room temperature indicative of oxidation in a solution containing h.97 ml. 0.1003]! thallous sulfate (51.1 mg. thallium) and 19.98 ml. of approxi- mately 0.0811 tetralithium perondiphosphate was greater than 14.5 days. The solution was at an initial pH of 8.1;. Increasing the amount of peroxydiphosphate and also increasing the acidity with dilute sulfuric acid did not materially reduce the time required for precipitate formation. Solutions of thallium and peroxydiphosphate were also heated at elevated temperatures. Samples were heated at 50° :1: 5°C. for four 68 hours and also at 100°C. for 30 seconds. As normally would be expected with an increase in temperature, the time required for the appearance of a precipitate was reduced somewhat. The time required for the appear- ance of a precipitate when a sample was heated at 50°C. for four hours and then allowed to stand at room temperature was 21; hours. And for a sample heated at 100°C. for 30 seconds and allowed to stand at room temperature, the time required for a precipitate to form was eight hours. In no instance was complete oxidation of thallium nude. That this was true was readily seen by the removal of the precipitates after a 12 hour period following the initial appearance of the precipitates and allowing the filtrates to stand whereupon additional precipitation occurred. Attempts at dissolving the brown precipitates with 6N sulfuric or hydrochloric acid were without success. Addition of excess reducing agent, ferrous ammonimn sulfate, caused the imediate dissolution of the thallium precipitates. It was further shown that the precipitates did not contain any orthophosphate by qualitative testing'with ammonium molybdate solution. The conclusion was drawn that the brown precipitate was thallic oxide, T1203. The preliminary work described above indicated that the reaction between peroxydiphosphate and thallium(I) proceeded very slowly. ’ However, it was found that the reaction was greatly accelerated by the simultaneous occurrence of the reaction between manganese(II) and peroxydiphosphate. The possibility of achieving rapid and complete oxidation of thallium(I) in the presence of manganese(II) was further investigated. A series of samples containing measured quantities of thallous sulfate and manganous sulfate was taken from their respective stock solutions in various milliequivalent' ratios and treated with a 3.0- to 35-fold excess of perquiphosphate solution to effect oxidation. Blanks containing the same amount of manganese and peroxydiphosphate were also prepared. The method of preparation, temperature conditions, . period of standing and pH were essentially those used in the manganese- peroxydiphosphate oxidation procedure. The oxidations were studied at room temperaimre, at 50° i 5°C. for four hours and at boiling for approximately 30 seconds. Removal of the precipitates formed in all cases was achieved by employing asbestos covered Gooch crucibles with suction. Thorough washing of the precipitates was made with water. The individual filtrates and washings were retained and subsequently were analyzed for excess unreacted perquiphosphate. The mnganese—peroxy— diphosphate blanks were treated similarly. The precipitates were dissolved in a measured excess of ferrous ammonium Sulfate solution approximately 0.111 in sulfuric acid and the excess subsequently back titrated with 0.1000N potassium dichromate solution. The difference between the milliequivalents of standard dichromate required to oxidize an equivalent quantity of ferrous ammonimn sulfate and the milliequivalents used in the back titration represented the milliequivalents of manganese and thallium which had 70 been present in the precipitate. Under the conditions used, complete oxidation of manganese was assumed in the reaction. Hence, subtraction of the milliequivalents of manganese originally added to the sample from the milliequivalents found gave the milliequivalents of thallium which had been oxidized. The results, compiled in this manner, for the oxidation of thallium with perowdiphosphate in the presence of manganese are contained in Table XVI. The filtrate and washings from the individual samples and blanks were analyzed for excess peroxydiphosphate by the ferrous ammonium sulfate- potassium dichromate method. The difference in milliequivalents between the sample and blank represented the milliequivalents of thallium oxidized. The results on the amount of thallium oxidized by this indirect method of determination are also contained in Table XVI. It is readily seen from the table that the presence of manganese exerted a pronounced influence on the rate of oxidation of thallium. Approximately an equivalent amount of manganese(II) was needed to effect the oxidation of thallinm(I). Ratios of milliequivalents of manganese(II) to milliequivalents thallium( I) less than one caused incomplete oxidation. More important, however, was the effect of temperature and the manner. in which heating and cooling was carried out. As expected, an increase in temperature resulted in an increase in the amount of thallium(I) oxidized. In light of the studies made in the presence of mnganese(II), the mechanism of the thallium(I) oxidation with peroxydiphosphate can be ‘vll..I1 i - 71 TABLE XVI THALLIUM(I) OXIDATION WITH PEROXYDIPHOSPHATE IN THE PRESENCE OF MANGANESE(II) AT VARIOUS TEMPERATURES Mg. Tl Mg. Mn Time of Per Cent Tl Oxidized Added Added Meq. Mn/Meq. Tl Standing Filtrate Precipitate Room temperature 20.2h 0.66 0.12 12 hours h5.95 h8.96 Heated for four hours at 50°C. 20.2h 0.66 0.12 1 hour 79.25 77.27 20.2h h.60 0.8h 1 hour 95.72 9h.h2 20.2h 5.5h 1.01 1 hour 95.96 97.31 Boiled for 30 seconds 20.2h 5.5h 1.01 30.mun. 85.82 83.ho 20.2u 5.5h 1.01 us min. 92.95 92.1h 20.2h 5.5h 1.01 Cooled to 50°C.98.7o 98.h6 30 min. 20.2h ' 5.5h 1.01 Cooled to 60°c.99.20 100.98 0 min. 20.2h 5.5h 1.01 2 hours 99.8h 99.72 20.2h 5.5h 1.01 2 hours 97.6h 96.65 30.h6 6.90 0.8h 2 hours 96.29 9h.oo 50.90 h.15 0.30 2 hours 8h.87 85.76 tentatively postulated. The immediate appearance of a pink color in the original sample solution followed by the subsequent brown-black precipitate characteristic of manganese dioxide was indicative that a reaction was occurring. Manganese(II) was probably oxidized to a higher valent species which reacted not only with the manganese(II) but also with the thallium(I). Indications pointed to the fact that manganese(II) 72 was oxidized to permanganate and it was this latter species or some intermediate which effected the oxidation of the thallous ion. This would necessarily imply a couple type reaction (28). The peroxydiphos- phate was the actor since it acted on both the other two constituents; manganese(II) was the inductor and thallium(I) was the acceptor. The reaction between peroxydiphosphate and manganese(II) was the primary reaction and that between peroxydiphosphate and thallium(I) was the induced reaction. 2. Oxidation of ChromiumfiIII) Preliminary investigations had shown that attempts at oxidizing green chromium(III) with peroxydiphosphate in near neutral solutions at room temperature did not result in a visual color change to yellow. It was virtually impossible to distinguish visually whether a slight color change to yellow occurred in the green solution. Varying the temperature, pH, concentrations, and periods of standing also gave incon— clusive results. However, boiling a chromic sulfate solution with excess peroxydiphosphate for 10 minutes and cooling for 30 minutes resulted in the appearance of a yellowish cast to the solution. The initial pH of this solution was 14.1 and after two hours dropped to 3.6. The yellow color did not increase in intensity with increased time of standing. It became obvious that oxidation of chromium(III) with peroxydi- phosphate was too slow for a suitable quantitative procedure. It was found that the amount of chromium(III) oxidized by peroxydiphosphate 73 to chromium(VI) was materially increased in the presence of manganous sulfate. A more thorough study of this reaction was made. Two milliliters of the chromic sulfate stock solution was pipetted into a 125 ml. iodine flask followed by 1.98 ml. 0.3?leN manganous sulfate solution. Twenty milliliters of 0.1027N peroxydiphosphate solution was added and the entire solution diluted to a total volume of 50 ml. with distilled water. The solution was then boiled for approximately three minutes whereupon the appearance of a yellow color developed simultaneously with the formation of a brown-black manganese dioxide precipitate. After cooling for three hours, the precipitate was removed by filtering through an asbestos covered 60M fritted funnel. The yellow colored filtrate was retained for further study. A Spectrophotometric method of analysis for determining the presence of chromium in the hexavalent form was used. This procedure was used because there is essentially no other way of determining chromate or dichromate in the presence of excess peroxydiphosphate. Both gravi— metric and volumetric methods of analysis for hexavalent chromium would fail for this particular situation. In addition, other factors influenced the selection of a spectro— metric method of analysis. In acid solution dichromate has two absorp- tion maxima, one at 257.5 mu and the other at 350 mm, which are well defined (13). It was found that peroxydiphosphate did not absorb at 350 mu in strong acid solution and other ions such as phosphate, sulfate, chromic and divalent manganese were transparent in the ultraviolet region (6). Convenience of the method was also an advantage. Art—“aim '- 7b A.complete absorption curve can be obtained in a matter of minutes and compared with a standard working curve prepared.under exactly the same manner as the samples. In order to have some reference source as a means of estimating the extent of chromic oxidation in the presence of manganous sulfate, it was necessary to prepare a suitable working curve, i.e., a Beer‘s Law plot of absorbance as a function of concentration. A 0.1000N potassium dichromate solution was prepared by dissolving b.9352 g. of Baker's analytical reagent grade potassium dichromate in 300 ml. of water, adjusting the pH to 1.5 by the addition of 0.1N sulfuric acid, and diluting quantitatively to exactly one liter. A series of solutions at various concentrations was prepared by taking aliquots of the standard dichromate solution and diluting with water to the desired concentra- tions. Dichromate solutions containing 0.88 x 10—3, 3.h6 x 10‘3, 6.92 x 10‘3, 8.68 x 10'3 and 17.30 x 10"3 mg. chromium(III) per ml. were prepared in such a manner. These solutions were all adjusted to a pH of 1.5 with dilute sulfuric acid. The absorbances of the individual solutions were measured at 350 mm at a constant slit width of 0.h0 mm. using a Beckman DU Spectrophoto- meter. The reference cell was a matched l-cm. silica cell containing water which had been adjusted with dilute sulfuric acid to a pH of 1.5. A plot (Figure 7) of absorbance vs. milligrams of chromium resulted in a straight line from 0.000 to 0.0173 mg. chromium per ml. As a check on the validity of the working curve, 1.98 ml. 0.1055N chromic sulfate stock solution was pipetted into a 250 ml. beaker and .‘[0 ‘8‘ ‘0'?! (.1 .5). I .iu . .2 . , . .\\.. .1. .i.“ . . .n‘ ,c ab 5 mum—40.... =>.23_20¢:0 m0 20Fd 23.00m ..._0 20_._.<._wm mkdxdwOIQOEa 23.000 Pzwoxwd 0nd 0¢.0 0N0 0 _.0 00.0 l 9.0 BONVBHOSEV 0N0 9L1 had been done in the preparation of the standard curve. Duplicate determinations on each of the above solutions were made along with their corresponding blanks. The results taken from the calibration curve are shown in Table XVIII. The absorbance readings of the two determinations for each solution agreed with each other. The values found indicated that there was essentially no pyrophosphate present in any of the peroxy- diphosphate solutions used. The amount present was too mall to be of any measurable consequence and may be attributed to experimental errors. TABLEIVIII DETERMINATION OF PYROPHOSPHATE DI TETRALITTEEUM PEROXYDIPHOSPHATE Initial Absorbance Per Cent Sample pH Reading Pyrophosphate Present 10 day old solution 7.1 0.160 0.00 10 day old solution 7.1 0.162 0.00 Freshly prepared solu- tion 8 .6 o .158 o .016 Freshly prepared solu- tion 8.6 0.160 0.00 Three day old solution 3 .8 0 .159 < 0 .012 Three day old solution 3 .8 0.157 < O .027 c. Titration of Perogydiphosphoric Acid and Its Salt Van Wazer and Holst (I40) have shown by titrations of various phos- phoric acids with alkali. that all the phosphoric acids, regardless of -...-. . . 95 their compositions, contained one strongly acidic hydrogen atom per phOSphorus atom. This hydrogen was neutralized at pH 3.8 to 14.2 depend— ing on the acid taken. Examination of structures (I) and (II) would indicate, therefore, that peroxydiphosphoric acid should exhibit two strongly acidic hydrogens when titrated with alkali and these hydrogens should be neutralized somewhere in the pH range of 14.0 '1' 0.2. A series of pH experiments was designed to verify these statements. The conversion of tetralithium perox‘ydiphosphate to peroxydiphos— phoric acid was achieved by an ion exchange procedure. An ion exchange column containing 52 g. of Amberlite 1R-100H ion exchange resin in the hydrogen form was used. The diameter of the resin column was 5.0 cm. and the height was 27 .0 cm. A tetralithium peroxydiphosphate solution, prepared by dissolving 0.569 g. of tetralithium peroxydiphosphate tetra- hydrate in 50 m1. of distilled water, was placed on the resin by allow- ing the solution to flow into the resin column. The peroxydiphosphoric acid was eluted by washing the column with 300 m1. of water. A flow rate of 0.01 to 0.05 ml./mimlte/ml. exchanger column, i.e., 1 to 5 ml./ minute was used (11). The first 70 m1. fraction of eluate was found to contain essentially none of the acid. A second fraction of 200 ml. was found to contain the bulk of the acid, while the third fraction of 150 ml. was much like the first in that no appreciable acid was detected when treated with alkali. Selection of the second fraction of eluate was, therefore, used in all the subsequent studies to follow. This fraction was retained 96 {cum—ing titration with alkali to determine the number of milliequi- valents of acid present in the sample solution titrated. Duplicate phosphorus determinations were made 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 orthophosphate form the solution was evaporated to near dryness on a steam bath in the presence of nitric acid prior to the precipitation of the magiesium ammonium phosphate. The magnesium ammonimn phosphate was reprecipitated, ignited to magnesium pyrophosphate, and weighed as such. The average value of the two determinations for phosphorus indi- cated 1.336 millimoles peroxydiphosphate had been converted to the acid form. This would correspond to approximately 65 per cent of the total peroxydiphOSphate originally taken which was present in the second fraction. The eluate containing the bulk of the acid was titrated with 0.1183N potassium hydroxide. The change in pH upon the addition of alkali was measured with a Beckman model H-2 pH meter equipped with a glass electrode. The potassium hydroxide solution was added in one milliliter increments until a masurable change in pH occurred, then in smaller increments. Magnetic stirring was used during the titration. It was found that no adverse effects were experienced in the readings provided slow stirring was used. The data are plotted in Figure 9 and tabulated in the Data Appendix. The initial pH of the acid solution was 2.0. Four breaks in the titration curve were obtained. From the appearance of the titration 0 umaofi .2400 m0_x0¢0>.._ 23_.wm<.r0n_ 23:0 .42 97 8.8 code 80» 8.8 8.0. , oo.o _ _ _ _ A .0.» 10.... d H -3. -o.m -9. uexomer 55633.. 3.... 99.. exoramoreexomma .6 20.55.... In . \(‘I‘ IL: .I Ill-ilfl- .' all r l L— 98 curve it can be seen that the first two hydrogens were nearly of equal strength in fairly good agreement with the statement of Van Wazer and Holst (ho) . Moreover, the neutralization occurred at approximately pH 7 3.6. As expected the third and fourth titration breaks corresponded to much weaker hydrogens being neutralized. The approximate pK values obtained from the half-titration points of the individual breaks are: p1!1 of 2.1, pK2 2.5, pK3 h.2, and pK4 7.0. The equivalence point at pH 9.0 corresponded to the complete neutralization of the acid to the tetra- potassium peroxydiphosphate salt. Assuming 1.336 millimoles of peroxy- diphosphate were in the sample that had been titrated and four replace- able hydrogens are available for neutralization with alkali, 5.3L;6 milli- equivalents of the acid must have been present. This would require 15 .11 m1. of 0.1l83N potassium hydroxide for complete neutralization of the acid. The equivalence point for the fourth break required 15.00 m1. of standard base. This experimental value was in good agreement with the theoretical 145 .ll m1. of base required. The other titration breaks required approximately 11.0 ml, 21.6 ml. and 32 .0 m1. of base to neutral- ize the first, second and third hydrogens, respectively. A replacement type of titration with acid was performed to see if it was possible to obtain a titration curve the reverse of the one obtain- ed by the direct titration of the peroxydiphosphoric acid with alkali. L tetralithium peroxydiphosphate solution was prepared by taking 0.579 g. (two millimoles) of tetralithium peroxydiphosphate tetrahydrate and di5501ving it in 150 ml. of water. The pH titration of the paroxy- diphosphate was made immediately thereafter in much the same manner as 99 described above for the acid. Standardized sulfuric acid, 0.1007N, was used as titrant. The data are plotted in Figure 10 and tabulated in the Data Appendix. The initial pH of the tetralithimu perozqdiphosphate solution was 8.140 which corresponded fairly well with the pH of 9.0 found for the complete neutralization of the acid in the previous work. Only one break in the titration curve was obtained. This break, at 20.16 ml., did not correspond to any one of the breaks obtained in the direct titration of the acid with base. 110 conclusions pertaining to the relative strengths of the peroxydiphosphoric acid and its acid salts can be deduced from the information revealed by this titration. From the appearance of the titration curve in Figure 9, it is apparent that the first and third breaks are rather poorly developed. It was decided to establish with certainty whether these two were actual breaks by employing a conductometric method of titration. A peroxydiphosphoric acid sample, prepared in the same nanner as previously described in the pH titration, was titrated conductometrically with base. A Serfass Conductivity Bridge Model RI: 1115 equipped with platinized platinum electrodes was used for the measurements. The cell constant was 0.10. Potassium hydroxide, 0.1183N, was added in one milliliter increments to the sample containing the peroxydiphosphoric acid. The conductance was expressed in terms of micromhos/cm. specific conductance. Hagnetic stirring was employed during the titration. The results are plotted graphically in Figure 11 and tabulated in the Data Appendix. i ' r A . . .. ‘ u o - . ' v , ' I . .. . _ . . A > . o 4. ‘ - u u w . . - 7 _ c 1' \ ‘ v . ~ . . I , l _ o _ V . - e V ‘ i - . - \ . _\ . O n _ 100 00.Nn O. UCDGE .znom 90¢ 032.330 2500.0 J! 00.8 00.¢N 00.0w 00.0. 00.N_ 00.0 09¢ 00.0 i a _ _ _ _ _ 80 00.N 00* 00.0 00.0 90¢ 0.x:u43m It? ubkrmmOIEn—Cnomua 2:.I...3<¢._.u.r “.0 29.25:. In. Hd 101 = manor.— .ZJOm uo_x0¢o>1 23.003.09.— znozd J! 0.00 0.9.. 00¢ can 0.0a 0.0. 0.0 J _ _ _ _ _ 00.0 :00._ O [cod loan . 00d uo_x0¢o>1 iQmmea 1.53 0.04 o_¢0:amoxa.0>xomua a0 zo....<¢.:._. mozfibnozoo .0: x Inc/mow” ‘aomuonouoo Four breaks in the conductance titration curve were obtained, the first one was poorly developed while the next three were sharp and well defined. It was found that 10.60 ml., 22.15 ml., 32.90 ml. and £5.00 ml. of 0.1183N potassium hydroxide were required to effect the neutral- ization of the first, second, third and fourth hydrogens, respectively. Comparison of Figures 9 and 10 revealed almost identical neutralization points for approximately the same size sample of acid taken. The con- ductance titration essentially substantiated the work done previously with the pH titration. The curves obtained indicated all four breaks are present and that the first two hydrogens are of approximately equal strength. In summary, titration of peroxydiphosphoric acid with alkali revealed neutralization of all four hydrogens was possible, two of which were approximately of equal strength substantiating the statement of Van Wazer and Holst (ho), the third and fourth were much weaker. Attempts at a displacement type titration with the salt of peroxydiphosphate were made. lo satisfactory results were obtained when the salt was treated in this manner. Additional evidence for the work on the pH titration of the acid was obtained. A conductometric titration of the acid substantiated the previous statement of two equivalent hydrogens in peroxydiphosphoric acid. The investigation strongly suggested the presence of a -P-0-0-P- type structure rather than a ~P-0—P- type where nonequivalent hydrogens would be much more evident. if. \3.:. a1 103 2. Playgical Methods a. Structural Proof by Infrared The infrared spectra have been recorded for many inorganic compounds most of which are salts containing polyatomic ions (29). No graphical spectrum in the infrared region has been made for tetralithium peroxydi- phosphate tetrahydrate. It was believed that such a spectrum would serve as an aid in distinguishing whether the peroxy group, -0-0~, was linked between two phosphorus atoms (structure I) or between a phosphorus and a lithium atom (structure II). The latter would result in a pyro~ phosphate type structure, ~P-0—P, which has indeed been observed and recorded tentatively in the literature 04,5). The infrared spectrum was obtained with a Perkin-Elmer Model 21 double-beam infrared spectrophotometer equipped with a sodium chloride prism. The sample was prepared by mulling a small portion (approximately 0.3 an.) with liquid petrolatum (Nujol) and placing the mull paste be- tween sodium chloride plates. The spectrum of the sample so prepared was recorded from 2.0 to 114.5 microns, using air as the reference. The spectrum of tetralithium peroxydiphosphate tetrahydrate is shown in Figure 10 together with the conditions used. The marked influence of the cation (lithium) as well as the hydrated water present is noticeable when compared with other phosphate containing compounds (29). It can readily be seen from Figure 10 that an ill- defined spectrum of peroxydiphosphate was obtained. The reason for this was not clear, but it may be due to lack of a single well—ordered crystal 10h structure (29). The absence of sharp clearly defined peaks in both the 930-970 cm"1 and 2560—2700 cm"1 (regions of symmetric and asymmetric -P-O-P type structure and -P-O-H structure respectively) portions of the spectrum in Figure 12 made the evaluation of any characteristic type linkage or molecular groups for peroxydiphosphate virtually impossible. That is, the lack of spectral data on reference compounds such as those structures with the type ~P~O-O-H and ~P-O-O-P, and the inadequacy of information concerning characteristic frequencies of molecular groups containing phosphorus made it impossible to reach a conclusion on the structure of the perowdiphosphate from infrared studies. b. mmral Proof by Eclear Magnetic Resonance - It has been shown that correct structural formulas for phosphorus compounds can be ascertained by nuclear magnetic resonance measurements (8,39). Interaction of a nucleus with its electronic environment, especially the valence electrons, influences the magnetic resonance absorption of the nucleus (1). At resonance, a change in the electronic environment within the atom so as to reduce the magnetic field at the nucleus necessitates an increase in the applied magnetic field. This increase is called a positive chemical shift. These shifts, which may be positive or negative, are generally measured in parts per million (p.p.m.) of the applied magnetic field relative to a chemical compound of the element arbitrarily chosen as a reference. Since the only naturally occurring isotope of phosphorus, P31, has a spin of one-half and a high magnetic moment (9), this tool would be especially appropriate for investigating phosphorus compounds. 105 N. muse—n. mzomoi .Eozmiuzi u. m m . m — — _ _ of .. 226 _" . ”320.52. a " ouuam ¢ “mommumgm . Em "225.63,. .5tz z. mk<¢o>Ixomua SEIPDE “.0 Saxkown—m om¢\ 260 A270 N x 103 Slit Width, mm. 0.80 0.76 0.70 0.6u 0.60 0.51 105.h ‘ -- -- -- 1.81 1.01 0.609 52.7 -- - 1.82 0.99 0.5h5 0.312 26.35 2.00 1.31 0.98h 0.511 0.28h 0.163 10.51 0.850 0.55h 0.120 0.220 0.121 0.072 . 5.27 0.130 0.275 0.208 0.115 0.066 0.039 o c ..“‘.EJ 1:. .. .....I I DATA APPENDIX The data plotted in Figure 5 are contained in Table IX. TABLEXX EFFET OF pH ON THE SPECTROPHOTOMEI‘RIC DETEEUINATION OF PEROXEIPHOSPHATE Addition of 17 per cent phOSphoric acid Concentration of peroxydiphosphate solutions - 0.0209ON* Absorbance at 2110 rrnrL Slit Width of 0.70 mm. pH 1 Hour 2 Hours 9.6 0.635 0.635 9.1 0.6h2 0.612 8.0 0.670 0.670 7.3 ‘ 0.729 0.729 7.2 0.7hh 0.7hh 7.1 0.760 0.760 7.0 0.790 0.790 6.1 0.910 0.910 60 09m) 09% 5.6 0.9h0 0.938 5.5 0.938 0.936 h.3 0.880 0.820 3.0 0.790 0.778 2.3 0.790 0.776 1.8 0.795 0.775 *By the ferrous amonimn sulfate method. Inl' 118 DATA APPENDIX The data plotted in Figure 6 are contained in Table XXI. TABLE XXI EFFET OF pH ON THE SPEETROPHOTOMETRIC DETERMINATION OF PEROXYDIPHOSPHATE Addition of 3N sulfuric acid Concentration of perosydiphosphate solutions - 0.026M1N* Absorbance at 2110 mu at Slit Width of 0.70 mm. PH 0.5 Hour 1 Hour 2 Hours 8.11 0.680 0.678 0.678 7.7 0.715 0.712 0.711 6.7 0.900 0.900 0.900 6.1 1.02 1.02 1.02 5.25 1.02 1.01 1.00 5.1 1.01. 1.01. 1.01 b.1 0.895 0.885 0.810 3.0 0.830 0.812 0.778 1.8 0.810 0.800 0.760 1.55 0.810 0.800 1.55 0.809 0.798 0.760 *By the ferrous ammonium sulfate method. o . . e n . a a o a u a . _ . o e . a o o n e 0 o n _ . . . v . . . . . _ . . I . o o s u u . o n . . a o . a . e . ,1 1.0.11. 1 I v 1 l1 . _ II c . ,I. . . _ _ a! I . a l.- IvJIIIIII u DATA APPENDIX The data plotted in Figure 9 are tabulated in Table HII. TABLE XXII 119 pH TITRATION OF PEROXIDIPHOSPHORIC ACID WITH 0.1183N POTASSIUM HYDROXIDE (200 ml. Fraction From Ion Exchange) Milliliters pH Milliliters pH Milliliters pH 0.00 2.01 20.00 2.89 32.00 5.71 1.00 2.01 21.00 3.15 31.00 6.25 2.00 2.05 21.20 3.21 36.00 6.68 1.00 2.09 21.10 3.30 38.00 7.00 6.00 2.11 21.60 3.10 39.00 7.11 8.00 2.15 21.80 3.19 39.50 7.16 10.00 2.20 22.00 3.60 10.00 7.20 11.00 2.30 23.00 3.99 11.00 7.18 12.00 2.12 23.60 1.08 12.00 7.70 13.00 2.15 21.00 1.20 11.00 8.99 11.00 2.15 21.10 1.28 16.00 10.35 11.50 2.17 21.60 1.30 18.00 10.70 16.00 2.19 21.80 1.31 50.00 10.95 17.00 2.51 25.00 1.39 52.00 11.10 18.00 2.59 26.00 1.52 51.00 11.12 18.20 2.60 28.00 1.85 56.00 11.21 18.80 2.69 30.00 5.19 The data plotted in Figure 10 are contained in Table XXIII. pH TITRATION OF 0.579 GRAN TETRALITHIUM PEROXYDIPHOSPHATE WITH 0.1007N SULFURIC ACID 120 Milliliters pH Milliliters pH 0.00 8.10 18.50 5.21 1.00 7.70 19.00 5.05 2.00 7.35 19.80 1.75 3.00 7.10 20.00 1.50 1.00 6.96 20.30 1.30 5.00 6.80 21.00 1.15 7.00 6.58 22.00 3.90 8.00 6.18 21.00 3.60 10.00 6.30 26.00 3.30 11.00 6.20 30.00 2.80 15.00 5.75 35.00 2.38 17.00 5.19 36.01 2.16 The data plotted in Figure 11 are contained in Table XXIV. DATA APPENDIX TABLEXXIV 121 CONDUCTOMETRIC TITRATION OF PEROXIDIPHOSPHORIC ACID WITH 0.1183N KOH FOLLOWING CONVERSION OF THE LITHIUM SALT BY ION—EXCHANGE 0.589 GRAN TETRALITHIUM PEEOXYDIPHOSPHATE TAKEN Milliliters A— 103 umhos/cm. Hilliliters A- 103 umhos/Cm. 0.00 1.00 2.00 3.00 1.00 5.00 6.00 7.00 h038 1.21 1.09 3.92 3.78 3.62 3.50 3.31 3.19 3.05 2.98 2.90 2.82 2.80 2.78 2.70 2.68 2.6 2 2 2 2 UL 1.51 21.50 21.80 22.00 23 .00 21.00 25.00 26.00 27.00 28.00 29 .00 30.00 31.00 31.50 32.00 32.10 32.80 33.00 33.50 31.00 36.00 38.00 10.00 12.00 13.00 11.00 16.00 18.00 50.00 52.00 51.00 56.00 .- U‘LUI O I O ODCDKIN \omooOI—Jooo NMHHi—‘HHI—‘Hl—‘Hl—‘HH O I \0 0\ U1 0 o o o 0000 \OWHO wwwwmmmmmmmmmmmm O ['00 U1 WNNH ggowgwgfigwwwmbg N a. r5 . , o . .r»..u~ . .r P. c o o I .. .. . ._ . 1...... :2... .51.... 2 . DATA APPENDIX The nuclear magnetic resonance spectrum in Figure 13 was obtained using the following operational conditions: 31 RF Frequency 16.2 MC. Nuclei P Reference: zero H3PO4 Resolu: H Instrument data: Peak cps RF atten: 18_db A -119 RF current: _hp_ua or —7.1 ppm Sw. Freq.: 10 6F Sw. Field: _1_C 2F Attenuation _2_.X Ch. Sp. Sum/sec. Probe Insert 15 mm .. 1. ”IR-1.1 . JUL 1 2 '01 cam/1mm! LIBRARY