Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies of difluorophosphoric acid and its alkali metal salts Reed, William 1965

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1965_A6_7 R4.pdf [ 2.67MB ]
JSON: 831-1.0062126.json
JSON-LD: 831-1.0062126-ld.json
RDF/XML (Pretty): 831-1.0062126-rdf.xml
RDF/JSON: 831-1.0062126-rdf.json
Turtle: 831-1.0062126-turtle.txt
N-Triples: 831-1.0062126-rdf-ntriples.txt
Original Record: 831-1.0062126-source.json
Full Text

Full Text

STUDIES OF DIFLUOROPHOSPHORIC ACID AND ITS ALKALI METAL SALTS by WILLIAM REED B.Sc, University of Durham, 1963 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in the Department of Chemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September, 1965 In presenting th i s thes i s in p a r t i a l f u l f i lmen t of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make i t f r ee l y a va i l ab l e fo r reference and study, I fur ther agree that per-mission for extensive copying of t h i s thes i s for scho la r l y purposes may be granted by the Head of my Department or by his representatives.. It i s understood that copying or p u b l i -ca t ion of t h i s thes i s for f i n a n c i a l gain sha l l not be allowed without my wr i t ten permiss ion. Department of C \ i £ M i S>i jg>/  The Un ivers i ty of B r i t i s h Columbia Vancouver 8, Canada Date q<? -Kv /Uf i ABSTRACT Difluorophosphoric acid was purified by a double distillation technique. Alkali metal difluorophosphates were prepared by reaction of the metal chlorides with purified difluorophosphoric acid; x-ray powder photographs and infra-red spectra of the salts were obtained. Electrical conductivity measurements were made on solutions of the alkali metal difluorophosphates in difluorophosphoric acid as solvent. The order of mobility of the alkali metal cations in this solvent was found to be Li>Na>K>Rb>Cs. Results indicate that the difluorophosphate ion does not conduct by a proton transfer process in this solvent. Conductimetric studies on solutions of a number of other compounds in difluorophosphoric acid are also reported. i i TABLE OF CONTENTS INTRODUCTION (a) Historical (b) Purpose of this work EXPERIMENTAL I. PREPARATION AND PURIFICATION OF MATERIALS (a) Difluorophosphoric acid (b) Difluorophosphates (c) Other materials II. PHYSICAL METHODS (a) Electrical conductivity (b) Infra-red spectra (c) X-ray powder photography RESULTS AND DISCUSSION (a) Electrical conductivity (b) Infra-red spectra (c) X-ray powder photography i i i LIST OF FIGURES FIGURE PAGE 1 Difluorophosphoric acid distillation apparatus 10 2. Fraction collector used on distillation apparatus 11 3. (a) Dropping funnel used for acid addition for distillation 12 (b) Weight dropper used fox addition of fluorosulphuric acid to the cell 4. Apparatus used for the preparation of alkali metal 13 difluorophosphates 5. Drying-train tester 16 6. Difluorophosphoric acid conductivity cell 18 7. Injector used for solute additions to conductivity cell 21 8. Specific conductances of some difluorophosphates at 25° 25 9. Specific conductances at low concentrations at 25° 28 10. Specific conductances of some electrolytes at 25° 31 11. Conductiraetric titration of HSOjF against N H 4 P 0 2 F 2 d t 2 S * 3 2 12. Infra-red spectra of CsPo2F2 and NaP02F2 38 13. X-ray powder films of some difluorophosphates 40 iv LIST OP TABLES TABLE PAGE I Analysis results for P and F in the difluorophosphates 14 IX Specific conductances of some difluorophosphates at 25° 24 III Specific conductances of some electrolytes at 25* 30 IV Infra-red data of the difluorophosphates 37 V ACKNOWLEDGEMENTS The author wishes to express gratitude to Dr. R.C. Thompson who first suggested the problem, and under whose supervision this work was done. Thanks are due to Messrs. R. Bellamy and S. Rak who constructed the glass apparatus, to Messrs. R. Green and A. Hardin for assistance with the operation of the Perkin-Elmer 421 spectrometer, and to Mr. R. Rao who assisted in the taking of the x-ray powder photographs. The generous gift of difluorophosphoric acid by Ozark-Mahoning Chemical Company is also gratefully acknowledged. Finally the author would like to thank Dr. H. Daggett of the Chemistry Department of the University of British Columbia for the use of the conductivity apparatus. INTRODUCTION As no previous survey on difluorophosphoric acid and its salts has been made i t was felt that a review of the literature would prove valuable, (a) Historical Difluorophosphoric acid was first prepared by Lange1 in 1927 -phosphoryl trifluoride was hydrolysed in cold water to give difluoro-2 3 phosphoric acid, the hydrolysis taking place in three stages: ' F OH OH OH 0»P—F ^ 0«P—F * 0=P —OH »r 0=P— OH ^ F ^ F OH If the hydrolysis of POFj is carried out in a weakly basic solution i t is possible to isolate the compounds HP02F2 and H 2 pOjF in the form of their 4 salts. The fluorine substitution products of meta- and orthophosphoric c acids are stable to water in much the same way as fluorosulphuric acid . However, even in neutral aqueous solution hydrolysis slowly occurs, the rate being much more rapid in the presence of acid or a l k a l i 6 . No further studies were reported on this acid for almost fifteen years. 7 In 1941 Tarbutton, Egan and Frary examined the reaction, CaF2 • P 20 5 * POFj • Ca(P0 3) 2 and found that when they varied the ratio of the reactants varying amounts of difluorophosphoric acid as a volatile by-product were obtained. The presence of the acid was explained by the partial hydrolysis of POF^  by small amounts of water present. 2 The volatile by-products were separated and the boiling point of the acid was given as 108-111°. This appears to have been the first isolation of the acid contrary to the view of Sidgewick that the free acid does not exist 3. 8 Lange investigated the reaction between phosphoric oxide and hydrogen fluoride to yield monofluorophosphoric acid. P 20 5 • 2HF • H20 » 2H2POjF Difluorophosphoric acid or a mixture of H^ PO^ F and HPO^were produced depending on the HF concentration. The same products were obtained when 100% orthophosphoric acid was allowed to react with aqueous hydrofluoric acid containing 41% HF.*1* When liquid, anhydrous hydrogen fluoride was o used in a 1:1 molar ratio with HjPO^, 67% of the HjP04 was transferred into the fluorinated derivative. However, the monofluorophosphoric acid, formed in the reaction, entered into a new equilibrium reaction with unreacted HF, forming difluorophosphoric acid. H3P04 + HF 3 = * H2P03F • H20 H2P0jF • HF 7 * HP02F2 + H20 Analysis of the equilibrium product showed that 33% H^ PO^  was unchanged, 60% was transferred to the monofluorophosphoric acid but only 7% to the difluorophosphoric acid. Variations on previous methods were attempted in order to synthesize the acid. Anhydrous monofluorophosphoric and difluorophosphoric acids were prepared** by allowing orthophosphoric acid and phosphoryl trifluoride to react at 70°-80*. However, the usual problem of a mixture of products was 12 13 obtained . Further work by Lange and Livingston consisted of a study 3 of the fluorophosphoric acids and a preparation of anhydrous difluoro-phosphoric acid. The hydrolysis of phosphoryl trifluoride was investigated and i t was found that pure acid could be obtained by the reaction: POFj • POP(OH)£ * 2POF2(OH) Experimental details are given in the article together with various 25° properties of the acid (d » 1.583, m.p. «* -96.5° * 1* and b.p. • 115.9°). 14 The system H^ O-HF-P^ O^  was further examined by Shaposhnikova ; difluorophosphoric acid was prepared by adding P 20 5 to an ice eold solution of HF in fluorosulphuric acid. The product was distilled in vacuo and fractionated; difluorophosphoric acid was found to decompose at its boiling point of 108°, its molecular heat of evaporation is given as 9125-360 cals. and Trouton's constant as 23.7-24.6. The acid was found to attack silicates at room temperature, the rate increasing with rise of temperature. Nuclear magnetic resonance studies of solutions in the homogeneous region of the system H20«HF-P205 confirm the presence of a mixture of acids. Ames, Ohashi, Callis and Van Wazer15 detected the presence of HPO^ F*,, H P Pg» HjPO^ and H J P O J F in the system and estimated their relative concentrations 19 31 by examining the F and P nuclear magnetic resonance spectra. An actual total of nine structural entities were found; in addition to the above acids, free water, end- and middle-phosphate structure units and a new structure unit containing one fluorine per phosphorus atom which is believed to be a monophosphate end group, were also found. Various other workers have investigated the fluorophosphoric acids by nuclear magnetic resonance; multiple magnetic resonance lines have been observed for F 1 9 and P 3 1 nuclei in H P O J P J , PFj, PH3, H J P O J F and HPFfi in the 4 liquid state. At a field strength of 6385 gauss a splitting of 0.244 gauss was obtained by Gutowsky, McCall and Slichter 1 6 for the P 1 9 resonance 17 in HPOjFj* Quinn and Brown measured the nuclear magnetic resonance 19 splittings for F in HP02F2 in weak fields. At a magnetic field strength of approximately 550 gauss they obtained a splitting of 0.240 gauss, but 19 the doublet structure obseived for the F resonance in HP02F2 coalesced as the field was decreased until a single broad resonance was obtained at 18 approximately 200 gauss. However, Roux and Bene obtained spectra at 35-15 gauss and found a doublet; the results were in general agreement with experiments at higher fields (at a field strength of 37 gauss a splitting of 0.245 gauss was observed), but as no singlet was observed i t was concluded that the singlet observed by Quinn and Brown in the field of 200 gauss cannot be explained in terras of the size of the main field only. Further information regarding indirect coupling of nuclear spins in 19 molecules containing P-F and P-H bands has been reported by Frank ; the relative magnitude of J p _ p and J H p are given for HP02F2 and other phosphorus compounds. For the series (i) FjPO, (ii) F2P0(0H), ( i i i ) FP0(0H)2 and (iv) 20 PO(OH)3 Gutowsky and McCall found phosphorus resonance in (ii) and (iv), with phosphorus shielding greater in ( i i ) ; the fluorine shielding increases with F-substitution. 21 Several salts of difluorophosphoric acid are known; the ammonium salt can be prepared by the fusion of phosphoric oxide with ammonium fluoride, the product being extracted with dry alcohol to remove the ammonium difluoro-phosphate which is then recrystallised from hot water. Dilute solutions of the ammonium salt treated with a soluable salt of nitron give a crystalline 2 precipitate of nitron-difluorophosphate. Lange prepared a number of salts 5 by treating hot aqueous solutions of nitron difluorophosphate with metallic nitrates, nitron nitrate was filtered off and the solution evaporated to dryness to recover the difluorophosphate. Most of the salts were found to be very soluble in water but the potassium and the cesium salts only moderately so. The difluorophosphates are stable in neutral solution, but less so in alkaline or acid solutions. Their general behaviour is similar to that of the perchlorates (and also the fluoroborates and fluorosulphates), the less soluble salts of both acids being the potassium and cesium salts. The alkali salts are isomorphous with the perchlorates, fluorosulphates and fluoroborates; this resemblance is due to these a l l 3 being strong monobasic acids with tetrahedral anions of nearly the same size . 22 Jonas was able to prepare difluorophosphates by allowing salts of hexa-fluorophosphoric acid to react with oxides such as: Si0 2, B 20j, WO^ ; CaO or sodium metaphosphate. Quantitative yields were obtained. Ryss and 23 Tul'chinskii investigated a new method to prepare the sodium salt free from other fluorophosphates. NaHF2 arid P20g were allowed to react in a platinum crucible, the product being extracted with absolute methanol. Rates of hydrolysis of the sodium salt were investigated and i t was found that in neutral aqueous solution, hydrolysis was slow even at 100*. However, on heating in 0.1 N NaOH solution at a 3:1 ratio of OH* to NaP02F2 for 10 minutes at 70* quantitative hydrolysis occurred by the reaction: P0 2F 2" • 20H" — POjF2" • F" • H20 On heating with excess of 0.1 N. NaOH in a stainless steel ampoule for about 2 hours at 160* complete decomposition occurred by the reaction P0 2F 2" + 30H" HPoJ" + 2F~ • HjO 6 Of the salts, only the potassium and the ammonium salt have been 24 investigated spectroscopically; Corbridge and Lowe examined the infra-red spectra of ammonium difluorophosphate in the region 5,000 - 6S0 cm."1, and an assignment of the frequencies was made. Buhler and Bues investigated the vibration spectra of fluorophosphate melts and crystals; the infra-red and Raman spectra of crystalline KP02F2, KPFg and ^POjF are reported and assignments made. The force constants and bond orders in these and in related anions are discussed. The only conductivity studies on difluorophosphoric acid have been made by Barr, Gillespie and Robinson , who measured the electrical conductivity of solutions of HC104, HSOjF, H P O J F J, and HSOjCl in sulphuric acid. They found that HSOjF and HSOjCl behaved as acids, whereas HP02F2 and CHJSOJH were bases in the H2S04 system. Difluorophosphoric acid has found l i t t l e use in inorganic chemistry*, 27 28 however, Stolzer and Simon * have used the acid extensively in organic 29 reactions. Hood has shown that the treatment of difluorophosphoric acid with aliphatic alcohols yields alkyl hydrogen phosphorofluoridates; HP02F2 • ROH > (RO) (OH) P0F2 30 & \) The fluorophosphoric acids have been used as polymerisation ' , condensation and alkylatlon catalysts, and also as anhydrous acids in the non-oxidising refining -Spoils. The salts of difluorophosphoric acid have been used industrially: 32 Na, K, L i , Ba and Pb salts stabilize chloroethylene polymers ; Zn, Co, Pb, 33 Fe and Cd salts are used as catalysts in the preparation of 8-lactones , and, substituted aluminium chlorides, e.g., A1C12P02F2 serve as alkylatlon 34 catalysts . 7 00 Purpose of the present work The purpose of this work is to investigate the properties of solutions in difluorophosphoric acid and in so doing extend the range of studied protonic solvent systems. As no complete study of a l l the alkali metal difluorophosphates has ever been made by a single author these compounds have been investigated further by infra-red and x-ray powder diffraction methods. EXPERIMENTAL I. Preparation and purification of materials (a) Difluorophosphoric Acid Commercial difluorophosphoric acid, supplied by Ozark-Mahoning Chemical Company, was purified by five double distillations at 9.0 cms. of 26 mercury and a temperature range of 45*-49*C, (Gillespie used 15 mms and 35 44tt_46°) in the apparatus shown in Fig. I. The procedure was as follows. The apparatus was evacuated and flamed out with a hot bunsen flame. Dry air was allowed to enter through tap H until atmospheric pressure was attained. A dry, dropping funnel (Fig. 3a) was fi l l e d , in a dry box, with difluorophosphoric acid and was then fitted into the distillation apparatus at L. The acid was allowed to drain into the flask A. The dropping funnel was removed and the thermometer well replaced. The system, which was connected to a vacuum pump at N via liquid nitrogen traps and an acetone-dry ice trap, was evacuated to about 9 cms. of mercury. The acid was then refluxed for about 30 minutes to remove hydrogen fluoride which was condensed out in the liquid nitrogen traps. The acid was distilled into tube C until a temperature of 45°C. was reached. At this temperature the distillate was directed into flask B by the fraction collector E (E is shown in more detail in Fig. 2). Tap M was then closed and the vacuum pump was disconnected from N and attached to 0. The second distillation was carried out under the same conditions, but in this case separation was achieved by rotating F within the B.19 ground glass socket at G so that the acid could be directed into either D or J as 8 9 desired. Flask D containing the required acid was quickly detached and capped; i t was then transferred to the dry box and the acid poured into the dropping funnel. Clean, dry apparatus was set up and the distillation repeated. In the first two or three double distillations considerable attack of glass occurred. In an attempt to reduce this, the acid was first refluxed under vacuum at temperatures slightly above room temperature for 24 hours in a stainless steel condenser and flask; however, on subsequent distillation in the glass apparatus no improvement was observed. Carrying out the same procedure as above but at room temperature had no effect on the result. Periodic checks of acid purity were made by examination of the fluorine 14 nuclear magnetic resonance spectrum ' (b) Difluorophosphates Ammonium, lithium, sodium, potassium, rubidium and cesium difluorophosphates were prepared by the reaction: M Cl • H P0 2F 2 > M P0 2F 2 + H Cl where M is the metal cation. Fig. 4 shows the apparatus used. The dry metal chloride was placed in the reaction vessel D which was then attached to the acid distillation apparatus at K in Fig. 1. The apparatus was evacuated and then flamed out in the normal manner. Difluorophosphoric acid was distilled at the usual pressure and temperature onto the chloride which immediately reacted with the acid resulting in the evolution of HCl. When sufficient acid had been added to dissolve a l l of the solid the reaction vessel was removed from the distillation apparatus and a ground glass B 19 FIG. I. DIFLUOROPHOSPHORIC ACID DISTILLATION APPARATUS 12 FIG. 3(a) DROPPING FUNNEL (b) WEIGHT DROPPER APPARATUS USED FOR THE PREPARATION OF THE ALKALI METAL DIFLUOROPHOSPHATES 300 14 cap was quickly placed on Q. Tube R was then removed and the reaction vessel was attached via S to the vacuum pump. The excess difluorophosphoric acid was removed at a pressure of about 0.5 irons, of mercury, with occasional warming of the vessel D with a bunsen flame. When a l l the acid had been removed the product was washed with ether and then recrystallized from dry methanol (except the sodium salt which was found to be exceedingly soluble and was, therefore, recrystallized from dry ethanol). Finally the difluorophosphates were washed with ether, dried and stored over phosphoric oxide in a vacuum desiccator. Aqueous solutions of the salts showed no precipitation on addition of silver nitrate solution (negative test for chloride) and no precipitation 36 on addition of lead and barium nitrate solutions (negative tests for monofluorophosphate and fluoride). Fluorine and phosphorus microanalysis were obtained in the A. Bernhardt Microanalytical Laboratories, Germany, and the results are shown in Table I below. TABLE I calc. Li obt. calc. Na obt. calc. K obt. %P 28.81 28.7 24.99 24.82 22.12 22.1 %F 35.21 35.5 30.65 30.80 27.14 27.29 calc. Rb obt. calc. Cs obt. calc. NH. 4 obt. %P 16.61 16.53 13.24 13.14 26.03 25.83 %F 20.38 20.42 16.25 16.42 31.93 32.23 %N - 11.76 11.63 15 (c) Other Materials sodium monofluorophosphate: Commercial Na2POjF obtained from Alfa Inorganics Inc. was recrystallized once from water. sodium fluoride: Chemically pure NaF was dried in a drying pistol at 80° and a pressure of 10 rams, of mercury for three days. fluorosulphuric acid: Commercial HS03F obtained from the Allied Chemical Co. was double-distilled at 164*. potassium fluorosulphate: KSO^ F was prepared by the method used by Gillespie et a l . 3 7 dry air: The air allowed to flow into the vacuum distillation system was passed first through calcium chloride, then magnesium perchlorate and finally through a liquid nitrogen trap followed by an acetone/dry ice trap. As some compounds, as well as the acid, were handled in the dry box, compressed air was passed first through three gas wash-bottles containing concentrated sulphuric acid, then through a tube containing calcium chloride and finally through a tube containing magnesium perchlorate. The air, dried in this way, was always tested for traces of water by means of the m ,drying-train M tester, shown in Fig. 5. The air was allowed to enter the tester at A from the outlet of the dry box, then pass over 30% oleum contained in B and exit via C. If the air caused no fuming in the vessel i t was considered to be dry enough for use. Air, dried in this manner was passed through the dry box for several hours before use. Several dishes of phosphoric oxide were placed at various locations in the dry box to ensure as dry an atmosphere as possible. 17 1 1 • Physical Methods (a) Electrical Conductivity The design of the cell used to measure the conductivities of solutions in difluorophosphoric acid is shown in Fig. 6. The cell could be attached to the distillation apparatus at K by means of the B.19 ground glass cone L. The cell has three electrodes and was designed so that the cell constant, when using electrodes B and C, was approximately 5 while the constant, when used electrodes B and A was approximately 15. Thus, accurate conductivity measurements could be made on weakly conducting solutions using the electrodes A and B, while measurements on more strongly conducting solutions were made using electrodes B and C. The capacity of the cell was about 400 mis. The cell was cleaned with aqua-regia and the electrodes were plated with platinum black by electrolyzing a chloroplatinic acid solution prepared 38 according to Jones and Bollinger . The solution consisted of a 0.3% solution of chloroplatinic acid in 0.025N. hydrochloric acid with 0.02% lead acetate added, A current of 10 milliaraps. was passed for 15 mins. with a reversal of current every 10 seconds. The cell was steamed out, dried and then calibrated using aqueous potassium chloride solution according to the 39 method of Lind, Zwolenik and Fuoss . The cell was replated and recalibrated after every four or five experiments. All measurements were made with the cell immersed in an o i l bath regulated by means of a mercury-thallium regulator at 25i 0.002°. The temperature of the thermostat was measured by Beckmann thermometers which had been calibrated against a platinum resistance thermometer. FIG. 6. DIFLUOROPHOSPHORIC ACID CONDUCTIVITY CELL 19 The apparatus used to make so l id additions to the c e l l i s shown in F ig. 7. It consisted of a "T"-shaped glass tube with B.19 ground glass sockets at the ends 0 and P, and a B.24 ground glass cone with an extension at M. The corks at 0 and P were made of teflon and they were t ight ly f i t ted with stainless steel pistons A and B respectively. The f l a t "runners" which were also made from tef lon interlocked at Q and lay on the bottom of the tubes. The compound to be added to the c e l l was weighed into small, preweighed, dry, glass boats which were inserted through 0 of the sidearm. Approximately eight boats could be accommodated in the sidearm. The loaded injector was then connected at M tr a rotary vacuum pump v ia l iquid nitrogen traps, warmed and evacuated. This was done to remove any water absorbed during the weighing-out process. After several hours the injector was detached from the pump and stored unt i l use in the dry box. A conductivity run was carried out in the following manner; the conductivity c e l l was attached by means of a B.19 inner ground glass joint d irect ly to the d i s t i l l a t i on apparatus at K and flushed out with dry a i r . Difluorophosphoric acid was d i s t i l l e d d irect ly into the c e l l ; acid obtained -4 -4 in this way usually had a conductivity between 2.41 x 10 and 2.51 x 10 ohm. cm. ~*. At a l l times in handling the acid great care was taken to exclude water. It was found that there was a gradual increase in the specif ic conductivity with time; over a period of 4-5 hours a 1% increase in K was observed. Solutions for conductivity measurements were prepared as follows: difluorophosphoric acid was d i s t i l l e d d irect ly into the c e l l which was weighed before and after addition of the acid. The B.24 stopper was removed and the injector was quickly inserted into the c e l l at F. Mercury was poured into the glass tubes holding the platinum electrodes, care being taken 20 to remove a l l the air bubbles. The cell and the injector were then placed on a support in the o i l bath. To make an addition, a glass boat was pushed by the piston A from the sidearm into the main tube, the boat was then moved by piston B along N and pushed into the acid. After each addition of solute the cell was well shaken to ensure good mixing, and returned to the thermostat. After sufficient time had elapsed to allow for temperature equilibrium (IS to 20 minutes) the resistance measurements were made. The cell was then removed from the thermostat, shaken again, and the resistance measurements repeated. In this manner errors due to insufficient mixing were eliminated. Fluorosulphuric acid was added to the cell by means of the weight dropper shown in Fig. 3b. As both adds hydrolyze in air, the weighed sample was added to the cell in the dry box. The cell was returned to the thermostat and the above procedure for measuring resistance was followed. Resistances of solutions were measured on a precision a-c resistance bridge which has been previously described by Daggett40. A 2,000 c/s oscillator was employed as the source and a telephone headset was used as the null-detector. Throughout this work,specific conductance will be referred to by the symbol K. (b) Infra-red spectra All spectra were recorded from 4,000 to 250 cm. on a Perkin-Elraer 421 Double Beam Spectrophotometer under "normal" operating conditions. The materials were examined as fine powders spread on cesium iodide plates. This was accomplished by finely grinding the sample and then dissolving i t in dry 21 22 methanol. A thin layer of the resulting solution was obtained on the cesium iodide, which was then placed on a hot plate to drive off the methanol. Specimens were obtained as a thin layer, thicker samples being used when searching for weak absorption. A decrease in transmittance above about 2,000 cm. "* due to scattered radiation was observed in many spectra; this was not removed on decreasing the particle size. By taking spectra with Nujol mulls in the higher frequency region, sharper absorptions were obtained. (c) X-ray powder photographs X-ray powder samples of the difluorophosphates were prepared in 0.3 41 mm. quartz capillaries by the method described by Azaroff and Buerger . The x-ray photographs were taken using a General Electric Camera of 14.32 cm. diameter. This camera employs Straumanis loading. Nickel filtered (using a 0.089 cm. thick Ni-foil) Cu-Ko radiation (X«- 1.5418 A) was used as the source. The x-ray tube was operated at 35 kilovolts and 15 milliamps. The camera employed a s l i t collimator, for which the exposure time required was between 3-6 hours depending on the sample. RESULTS AND DISCUSSION (a) Electrical Conductivity The results of the conductivity measurements on solutions of metal difluorophosphates in difluorophosphoric acid at 25± 0.002" are given in Table II. As a l l the solutions were made up by weight the concentrations are expressed in molal units (m); due to the lack of accurate density data no attempt was made to express the concentrations in molar units. In each case a plot of ie against molality was made (Fig. 8). By analogy with other protonic solvent systems HPO.^ would be expected to undergo autoprotolysis according to the equation: 2HP02F2 ^ = ± H2P02F2* • P02F2- 1 In this system then, bases may be defined as substances which, when dissolved in difluorophosphoric acid, increase the concentration of the difluorophosphate anion P02F2~, and acids may be defined as any substance which increases the concentration of the difluorophosphoric acidium ion HjPO^P^*/ It i s , therefore, expected that the alkali and alkaline earth difluorophosphates will behave as strong bases in this system. M P0 2F 2 M* + P0 2F 2" 2 In solvents where the mobilities of the autoprotolysis ions are very much greater than the mobilities of other ions (due to a proton transfer mechanism of conduction for the former ions), strong bases exhibit almost identical conductivity curves at low concentrations with small deviations 37 noticeable only in the more concentrated solutions . As the conductivity curves for the alkali metal difluorophosphates deviate from each other at even the lowest concentrations measurable, i t must be concluded that P0„F„~ ion does not show abnormal conduction. 24 TABLE II SPECIFIC CONDUCTANCES OF SOME DIFLUOROPHOSPHATES AT 25°C. LiP0 oF o KPO-F- RbPO-F, 9 * * 4 2 4 2 4 10 in 10\ 10 m 10 tc 10 m 10% ohm."* em."" ohm."* cm."* ohm. * cm. * 0.000 2.482 0.000 2.410 0.000 2.499 0.306 2.673 0.1S8 2.557 0.639 2.662 1.020 3.464 0.566 2.740 2.051 3.318 2.190 4.528 . 1.336 3.153 4.056 4.314 4.429 6.379 2.485 3.889 7.084 5.825 6.773 8.086 4.043 - 4.925 11.22 7.829 9.733 10.15 6.054 5.939 16.84 10.56 13.91 12.73 9.310 7.658 23.63 13.81 19.69 16.59 13.83 10.10 31.88 17.72 27.66 20.46 18.53 12.79 37.20 24.48 23.88 15.21 CsPOJ* 49.84 29.18 29.65 17.77 1 Q2 * 1 Q4 60.08 32.21 34.37 19.82 -1 -1 70.55 34.56 40.28 22.46 ohm. cm. 48.26 24.27 0.000 2.443 NaP02F2 NH4P02F2 J;Jg 22f2l 10 m 10 ic 10"m 10K 3.191 3,708 - 1 - 1 . -1 -1 6.155 5.041 ohm. cm. ohm. cm. 9.751 6.665 0.000 2.503 0.000 2.473 13.48 8.395 0.222 2.557 1.977 3.546 17.92 10.51 0.761 2.868 5.864 5.851 24.29 13.64 2.859 4.500 11.44 9.110 6.008 6.794 18.75 13.35 NH4P07F,(at low concentration) 8.220 8.266 27.15 18.71 11.16 10.12 35.60 23.91 15.13 12.26 20.95 15.15 25.36 17.96 31.37 , 20.67 36.88 22.81 42.46 24.75 47.79 26.40 7 ,«4 10 m * -1 ohm. cm. 0.000 2.499 0.380 2.693 1.090 3.088 1.842 3.546 2.724 4.076 5.758 4.691 4.834 5.507 6.718 6.619 8.940 7.740 26 The conductivity of the alkali metal difluorophosphates at any given concentration decreases in the order Li>Na>NH ~K>Rb>Cs. As a l l these salts 4 have the common anion P02F2" the difference in conductivity must be due to differences in the mobilities of the cations. This order af cation mobility is opposite to that found by Gillespie et al. in their conductance measure-37 42 ments in HSOjF and H2S04 in which they found the order Cs>Rb>NH4~K>Na>Li 43 prevailed. Gillespie has suggested that the lighter members have the larger solvated ion size, therefore, accounting for their lower mobilities. Our results suggest that solvation of cations in HP02F2 is weak and, therefore, the mobility is determined by the unsolvated ion size. As may be seen the potassium and the ammonium salts give conductance curves which are almost collinear at low concentrations (Fig. 9). This agrees with other workers' findings that the potassium and ammonium ions are of similar 37 42 size and hence have similar mobilities * On extrapolation of the linear portion to zero concentration the curves do not pass through the origin nor through the in i t i a l point corresponding to the solvent conductivity. There must, therefore, be some curvature of the conductivity curve at the lowest concentrations and this appears to be substantiated on close examination (Fig. 9). The conductivity of the acid may be attributed at least partly to its autoprotolysis. However, i f the conductivity of the pure solvent is due only to the ions produced in the autoprotolysis reaction the linear portion of the conductivity curve should pass through the origin on extrapolation. It appears that on extrapolation -4 -1 the conductivity at zero concentration is approximately 2.3x10 ohm. cm. This may be attributed to ions other than those produced in reaction (1). 27 These ions presumably arise from impurities such as monofluorophosphoric acid, hydrogen fluoride and water. Small traces of water would cause hydrolysis of the acid as given by the following equations: HP02F2 • H20 -^—^ HF • H2POjF H2P03F + H20 HF + HjP04 Ions would arise from the ionization of HF which would be expected to behave as a weak acid: HP02F2 • HF H2 P 02 F2* * F " and from the ionization of H2POjF which would be expected to behave as a weak base: H2P03F • HP02F2 — H 3P0/ • P02F2^ No evidence of any impurity was obtained on examination of the fluorine nuclear magnetic resonance spectra (a doublet with a splitting of 0.241 gauss (969.5 c.p.s.) was obtained which agrees well with that of Gutowsky*6, Quinn 1 7 and Roux18). The original acid appeared to contain 50% monofluoro-phosphoric acid and although the final acid showed no monofluorophosphoric acid impurity in the N.M.R. spectrum the concentrations of impurity with which we are concerned in conductivity work are too low to be detected by N.M.R. In order to investigate salts which contain anions other than P0 2F 2" solutions of NaF, Na2P03F, KSOjF and HSOjF were studied (Table III). Potassium fluorosulphate is insoluble; however, conductance results were obtained for the other solutes and these are given in Fig. 10. At low concentrations the slope of the ic-tn curve for Na?PO_F is more 29 than twice that of NaP02F2 and this may be explained by assuming that Na2POjF reacts as is given by equation below: Na2P03F + 2HP02F2 > 2Na + 2P02F~ * H 2 P 0 3 F As a result of this NajPOjF should have a conductivity curve with twice the slope of the curve for NaP02F2. The fact that i t is slightly more than twice can be explained by assuming that monofluorophosphoric acid is a base in this solvent and is prctonated by the difluorophosphoric acid: H2P03F • HP02F2 — V H 3P0 3F + • P O ^ The slope of the conductivity plot for sodium fluoride is considerably lower than that of NaP02F2, while this could be due to a much lower mobility of the F~ ion compared to that of the P02F2~ * o n * t * s m o r e probably due to incomplete dissociation of NaF through ion pair formation Na* F" ===^  Na* F~ This idea is supported by the fact that at low concentration where ion pair formation is weakest the slope of the NaF curve is very similar to the curve for NaP02F2, while at higher concentrations the slope of the NaF curve decreases whereas that of NaP02F2 remains essentially constant. This tendency for ion pair formation is consistent with the apparent low solvating power of the medium. It is not immediately obvious why the difluorophosphates themselves show less tendency to form ion pairs than other salts in difluorophosphoric acid. Conductivity data for solutions of fluorosulphuric acid in HP02F2 are also given in Fig. 10. Solutions of HSOjF may be neutralized by addition of the base NH^PO^ giving rise to precipitation of the salt NH^ SOjF and a resulting decrease in conductivity (Table III, Fig. 11) of the solution. These results prove that HSOjF is an acid in this system, probably ionizing according to the equation: 30 TABLE III SPECIFIC CONDUCTANCES OF SOME ELECTROLYTES AT 25°C. Na,PO,F HSO.F ? 3 4 2 4 10 m 10\ 10 m 10 K ohm. 1 cm. 1 ohm. 1 cm. 0.0000 2.443 0.000 2.580 0.0754 2.455 11.81 5.02 0.1990 2.567 17.70 7.46 0.5426 3.216 23.62 9.79 1.131 4.280 29.50 12.19 2.064 6.368 41.35 18.86 3.684 8.896 53.10 24.60 6.908 14.38 -1 11.83 21.30 Addition of NH.PO-F, to the 0.531 molal. HSO_F/HPO,F, solution - conductimetric titration* 1 NaF NH4P02F 2 102m 10 4K 2 lO'm 104K ohm. 1 cm. 1 ohm. 1 cm. 0.0000 2.300 6.79 19.33 0.4115 2.441 22.9 10.25 1.811 3.239 45.6 4.48 4.166 4.263 61.8 8.16 8.315 5.685 13.81 7.520 30.15 11.68 1 33 HSOgF • HP02F2 > H2 P 02 F2 * S 0 3 F " The K versus m curve for HSOjF shows an i n i t i a l flat portion followed by a linear increase in conductivity with concentration. On extrapolation of the linear portion to zero concentration the curve does not pass through the ini t i a l point corresponding to solvent conductivity but appears to pass through the origin. If the impurity in the solvent is in fact basic then addition of HSOjF should neutralize this impurity. In fact the flat portion of the HSOjF curve may be attributed to titrating the impurity. Since the slope of the linear portion of the HSOjF curve is similar to the slopes of the curves for the difluorophosphates i t may be concluded that fluorosulphuric acid is undergoing dissociation to roughly the same extent as occurs with the alkali metal difluorophosphates. Hence fluorosulphuric acid appears to be a strong acid in this solvent, a result which is consistent 26 with ;the conclusion of Gillespie et al. that HSOgF is a very much stronger acid than HPOjFj. 44 Gillespie has suggested that the strengths of inorganic oxyacids are determined largely by the number of equivalent oxygen atoms in the anion over which the negative charge may be spread. The strength of the acid increases with the number of equivalent oxygen atoms, therefore, HSO^ F, HSOgCl and HSOjOH are of the same type, having the anions FSOj", CISOj" and HOSOj" with three equivalent oxygen atoms. C104~ should, therefore, be the anion of the stronger acid HCIO^ , which has four equivalent oxygen atoms, and (H0)2P02~ and F2P02~ anions of weaker acids, having only two equivalent oxygen atoms. However, i t seems reasonable to suppose that some of the charge on the anion is accommodated on the halogen atom and i f the sharing of the charge between 3 4 the oxygen atoms and halogen are equal HSOjF and HSO^Cl would be in the same class as HCIO^. A similar effect for P O 2 F 2 ~ w o u l d P u t t n e a c * d *-n t h e s a m e class as HC10A and i t would be expected to be stronger, rather than weaker , t n a n 2 6 H J S O J . G i l lespie et a l . have made comparisons of acid strengths of di lute solutions of various acids with that of H S^O^ in bulk, and obtained the order: H 2 S 2 0 7 >HSO s F>HSOjCl>HC10 4 >HP0 2 F 2 ; they have indicated that the acid strength of H 2SOA in bulk may, because of co-operative hydrogen bonding, be considerably greater than that of H 2 S 0 4 in di lute solution. This suggests that HP0 2F 2 may in fact be a stronger acid in bulk but our results do not substantiate th i s , since even in bulk HP0 2F 2 i s a very much weaker acid than HSOjF Conclusions Difluorophosphoric acid i s a solvent of weak solvating ab i l i t y as shown by the low so lub i l i ty of salts such as KSOjF and the order of mobilities of the a lka l i metal cations. Since one of the requirements for proton transfer conduction in a solvent is association between the autoprotolysis ions and the solvent molecules through strong hydrogen bonds, then the absence of proton transfer conduction in this solvent may be due to a large extent to the inab i l i ty of the solvent to solvate the ions strongly. Indeed the relat ive low boi l ing point of difluorophosphoric acid compared to the boi l ing points of HSO,F ( 1 6 3 * ) and H 2 S 0 4 ( 2 9 0 - 3 1 7 ° ) for example indicates that hydrogen bonding in this solvent i s very weak. 35 (b) Infra-red spectra Results of measurements of the infra-red spectra of the difluorophosphates in the region 4,000-300 cm."1 are given in Table IV. The position of the peaks is In good agreement with the reported spectra for difluorophosphates; the absence of any absorption in the monofluorophosphate region indicates l i t t l e contamination of the compounds used in this work by monofluorophosphate impurity. The spectrum for the lithium salt differs from that of the other salts studied; this could be due to a basic difference in the crystal structure. Spectra of CsPQ2F2 and NaPO*2F2 are given in Fig. 12 as typical spectra. P-0 stretching vibration region: The ionic phosphate vibration has already been connected with absorption .44 .45 .46 at 1040-1000 cm. , 1110-1050 cm. , and 1,170-1,000 cm. . In the series of compounds studied only absorption in the last range of values was observed. The difference in position of the ionic phosphate absorption bands of the difluorophosphates compared with other phosphates has been connected with the presence of highly electronegative F atoms bonded to the 47 24 phosphorus atom . Corbridge and Lowe examined the spectra of ammonium difluorophosphate in the region 5,000-650 cm.**1 and their values are given in the table. It can be seen that they obtained an extra peak at 1,005 cm."1 and i t appears likely that this may be due to monofluorophosphate (which absorbs in the region 1,070-1,000 cm."1) as no peak was detected in any difluorophosphate studied in this work. A difference of 30 cm."1 was found for the asymmetric PO stretching frequency from that found by Corbridge. 25 However, the value obtained by Buhler and Bues for the asymmetric P0 stretch 36 in KP02F2 agrees well with that found in this work. Their values for K P O J F J are also given in the Table; i t is felt that the peak they report at 535 cm."1 is due to monofluorophosphate which absorbs at 530 cm."1 in 48 Robinson has recently discussed P-0 stretching frequencies in a number of phosphorus compounds. He found a linear relationship existed between the symmetric and asymmetric stretching frequencies of the P02 group. In ammonium difluorophosphate 1,125 cm."1 was assigned to the symmetric PO stretching frequency and 1,262 cm."1 to the asymmetric. However Robinson used only the ammonium difluorophosphate in his plot of symmetric against asymmetric PO stretches; the straight line correlation is given by the equation: *sym. ' °* 6 5 W • 2 7 0 Making use of the information obtained here on the difluorophosphates a better value for the straight line correlation i s : v««« " °* 7 0 v a e«n, • 2 1 0 sym. asym. P-P stretching vibration region. P-F stretch has been assigned to the region 990-840 cm."1 in PFj, P0F3 and PF,.4** and 980-740 cm."1 in organophosphorous compounds47. The absorption at 835-720 cm."1 found in a l l the monofluorophosphates has been 24 assigned by Corbridge and Lowe as probably due to P-F stretching; they also found that ammonium difluorophosphate absorbs in this region. However examination of Table indicates that the P-F stretch of the difluorophosphat lies in the region 940-818 cm."1. I t 5 0 has been shown that the symmetric PFj deformation occurs around 500 cm."1 and this agrees well with the values 25 obtained. However Buhler and Bues have also assigned the symmetric P02 deformation to about the same region, namely 535 cm."1. 37 TABLE IV INFRA-RED DATA OP THE DIFLUOROPHOSPHATES (FREQUENCIES IN cm.*1) LiP0 2F 2 NaP02F2 K P 02 F2 RbPOjFj 1273s 1309s (1332s (1330s 1164 s 1152 s [1310 s \l310 s 940 s (868 s 1148 s 1145 s 890 s 1844 s j 850 s (846 m ("525 s [502 s L832 s 1827 s 1498 s 1458 m [503 s (505 ra (426 s 360 w 1495 s 1492 s 1415 s-("357 s 1342 s CsPO,F- NH-P0,F, NH.PO-F, (Corbridge KPO-F, (Buhler , c 3 MO a 2680 . vb 4 2 2 and Lowe") 2 2 and Bues25) 3,380-2,2680 m,vb. 2 ) 9 0 0 . 2 > S 5 0 (1321 s [1443 s (1445 w,sh 1330 w U299 s (1410 s 11414 s 1311 s 1137 s 1292 s 1262 s 1145 s (843 1138 s 1125 s f857^ 8 1818 s (860 m,sh 1005 ww 834 s (503 m 1842 s (870 w,sh. 535 w 1489 s 500 s (.832 s (512 m 1481 s 286 w Note: Brackets indicate incompleted resolved bands (see Fig.12) s * strong; m '*> medium; w • weak; sh » shoulder FIG. 12. INFRA-RED SPECTRA 1.000 I.S00 1,000 500 WAVE NUMBER (cm."*1) (b) C S P 0 2 F 2 2.000 1,500 1,000 500 200 WAVE NUMBER (cm. ) 39 The influence of the cations on the spectra do not appear to be very marked, except in the case of the lithium salt. A pronounced shift of a l l peaks to lower frequency was observed, excepting the asymmetric PO stretching frequency which occurs at 1,273 cm.-1. These shifts are probably bound up with the tendency of lithium salts to exhibit some covalency. With the remaining alkali metal salts, a tendency for the shift of certain peaks to higher frequency with increasing mass is notice-able. The progressive shift is most evident with the symmetric PO stretching frequency, i.e. in LiP0 2F 2 i t is 1,164 cm."1, and drops gradually to 1,137 cm. in CsPOjPj. (c) X-ray powder photographs Inspection of the x-ray powder films (Fig. 13) indicates that with the exception of LiP0 2F 2 and possibly NaPOjFj the alkali metal and ammonium difluorophosphates are isoroorphous. The anomality of the lithium salt may be due to the tendency of the lithium atom to attain only 4 co-ordination in the crystal while the other larger alkali metal cations attain 6 co-ordination. The presence of numerous lines in the NaP02F2 powder film indicates either that considerable impurity occurred in the sample used or that the NaP09F, is in fact not isomorphous with the other salts. ray Powder Fi1ms 4;1 REFERENCES 1. W. Lange, Ber. 60B, 962 (1927). 2. W. Lange, Ber. 62B, 786 (1929). 3. N.V. Sidgwick, Chemical Elements and their Compounds, Vol. 1, Oxford (1950). 4. W. Lange, Ber. 61, 799 (1928). 5. F. Ephraigm, Inorganic Chemistry, Gumey and Jackson (1943). 6. T. Moeller, Inorganic Chemistry, Wiley (1955). 7. G. Tarbutton, E. Egan and S. Frary, J. Am. Chem. Soc. 63, 1782 (1941). 8. W. Lange, Chem. Abs. Al, 572f. (1947). 9. W. Lange and R. Livingston, J. Am. Chem. Soc. 69, 1073 (1947). 10. W. Lange, Ber. 62, 1084 (1929). 11. W. Lange and R. Livingston, Chem. Abs. 42_, 1711h (1948). 12. W. Lange, Fluorine Chemistry, Vol. I, ed. J. Symons, Academic Press (1950) J.R. Van Wazer, Phosphorus and its Compounds, Vol. I, Interscience (1958). 13. W. Lange and R. Livingston, J. Am. Chem. Soc. 72, 1280 (1950). 14. M. Shaposhnikova and Alliloeva, Zh. Prikl. Khim. 35, 760 (1962). 15. D. Ames, S. Ohashi, C. Callis and J. Van Wazer, J. Am. Chem. Soc. 81, 6350 (1959). "~ 16. H.S. Gutowsky, D. McCall and C. Slichter, J. Chem. Phys. 22, 162 (1954). 17. W. Quinn and R. Brown, J. Chem. Phys. 21_, 1605 (1953). 18. D. P. Roux and G. Bene, J. Chem. Phys. 26, 968 (1957). 19. P.J. Frank, Helv. Phys. Acta. 31_, 54 (1958). 20. H.S.,Gutowsky and D. McCall, J. Chem. Phys. 2j2, 162 (1954). 21. W. Lange, Inorganic Synthesis, Vol. II, ed. Fernelius, McGraw-Hill (1946). 22. H. Jonas, Chem. Abs. 46, 11603a (1952). 23. I. Ryss and B. Tul'chinskij, Russ. J. Inorg. Chem. 7, 677 (1962). 24. D. Corbridge and 6 . Lowe, J . Chem. Soc. 4S55 (1954). 25. K. Buhler and W. Bues, Z. anorg. u. allgem. Chem., 308, 62 (1961). 26. J . Barr and R.J. Gi l lespie and E . Robinson, Can. J . Chem., 39, 1266 (1961). 27. C. Stolzer and A . Simon, Ber. 93, 2578 (1960.) 28. C. Stolzer and A. Simon, Natur. 47, 229 (1960). 29. A. Hood, Chem. Abs. 51, 459a (1957). 30. a. J . Brooks, A. 0 'Kel ly and R. Work, Chem. Abs. 41, 2608d. (1947). b. Topchiev and Andronov, Dokl. Akad. Nauk, III, 365 (1956). 31. 6. Johnson and B. Rope, Chem. Abs. 43, 5583c. (1949). 32. C.B. Havens, Chem. Abs. 52, 21, 244 (1958). 33. J.R. Caldwell, Chem. Abs. 44, 10,732g. (1950). 34. H.R. Appell, Chem. Abs. 52, 19107f (1958). 35. R.C. Thompson, Ph.D. Thesis, McMaster University (1962). 36. R.P. Claridge and A. Maddock, Radiochimica Acta. I_, 80 (1963). 57. J . Barr, R. Gi l lespie and R, Thompson, Inorg. Chem. 3_, 1149 (1964). 38. G. Jones and D. Boll inger, J . Am. Chem. Soe. 57, 280 (1935). 39. J . Lind, J . Zwolenik and R. Fuoss, J . Am. Chem. Soc. 81_, 1557 (1959). 40. A.C. Harkness and H.M. Daggett, Can. J . Chem. 43, 1215 (1965). j 41. L. Azaroff and M. Buerger, Powder Method in X-ray Crystallography, McGraw H i l l , New York (1958). 42. S.J. Bass, R . Flowers, R. G i l lesp ie , E . Robinson and C. Solomons, J . Chem. Soc. 4315 (I960). 43. R.H. Flowers, R.J. G i l lesp ie, E . Robinson and C. Solomons, J . Chem. S o c , 4327 (1960). 44. R.J. G i l lesp ie , J . Chem. Soc. 2537 (1950). 45. N.B. Colthup, J . Opt, Soc. Amer. 40, 397 (1950). 46. L .J . Bellamy and Beeeher, J . Chem. Soc. 1701 (1952). 47. D. Corbridge and E . Lowe, J . Chem. Soe. 493 (1954). 43 48. L. Daasch and D. Smith, Analyt. Chem. 23, 853 (1951). 49. E.A. Robinson, Can. J . Chem. 41, 173 (1963). 50. W. Gerrard, J , Chem. Soc. 1454 (1940). H.S. Gutowslcy and A. Lieher, J . Chem. Phys. 20_, 1652 (1952). 51. E.A. Robinson, Can. J . Chem. 40, 1725 (1962). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items