‘ME REACHQMS 22f? D EME‘CHYWKQ NQ‘FE‘EAFLQQEQGF?€@$?H€}MNE WETH AN—H‘Y’BRQUs {WEIRQGEEQ HALEDéS Emu ‘Eav ‘Z‘Ew Dogma @‘3 M. 5. fitCEEfiAN S'Ffi‘li UM?E¥§E?Y Ronaid Michaei Rogowski £968 ..os in»; L 1 E R A R. Y Michigan State University ‘ Tnasxs ABSTRACT THE REACTIONS OF DIMETHYLAMINOTETRAFLUORO- PHOSPHORANE WITH ANHYDROUS HYDROGEN HALIDES by Ronald Michael Rogowski The reactions between dimethylaminotetrafluorophosphorane and anhydrous hydrogen halides have been investigated. By means of these reactions, it has been possible to prepare tetrafluorochloro- phosphorane and the previously unknown tetrafluorobromophosphorane. 19 These compounds have been characterized by molecular weight, F and 31F nmr, mass and infrared spectral measurements. In contrast to the behavior of hydrogen chloride and hydrogen bromide, hydrogen iodide gives only PF3 and 12 as volatile products. The stoichiometry of these reactions, structure, and symmetry of the prepared compounds is considered. THE REACTIONS OF DIMETHYLAMINOTETRAFLUORO- PHOSPHORANE WITH ANHYDROUS HYDROGEN HALIDES By Ronald Michael Rogowski A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1968 ._ / - q '3 "1‘ .‘l g .b,,o-m'* q,../.»4/a? / ACKNOWLEDGMENTS The author is especially indebted to Dr. Kim C. Cohn for his guidance and personal interest throughout the course of this research. The author is also grateful to Dr. J. Heeschen of the Dow Chemical Company, Midland, Michigan, for his assistance in obtaining the 19F nmr spectra, and to Mr. F. Parker of the University of Michigan, Ann Arbor, for the 31? nmr data. ii TABLE OF CONTENTS Page I 0 INTRODUCTION 0 I O 0 O O 0 O O 0 O O O 0 0 O O O O O 1 II. EXPERIMENTAL METHODS . . . . . . . . . . . . . . . . 4 A. Sample preparation . . . . . . . . . . . . . . 4 B. Instruments employed . . . . . . . . . . . . . 5 C. Reaction with HCl . . . . . . . . . . . . . . . 7 D. Reaction with HBr . . . . . . . . . . . . . . . 13 E. Reaction with HI . . . . . . . . . . . . . . . 16 III. DISCUSSION 0 O O O O O O 0 O O O O O O O O O O O O O 20 A. Reaction stoichiometry . . . . . . . . . . . . 20 B. Hydrolysis reactions . . . . . . . . . . . . . 27 C. Suggestions for future research . . . . . . . . 28 BIBLIOGRAPIIIY O O O O O O O O 0 0 O O O O O O O O 0 O G I O 29 iii LIST OF FIGURES FIGURE Page I. Reaction bulb used in experimental work . . . . . . . 6 II. it Of PF C1 0 O O O O O O O O O O O O O O O O O O O O 8 4 19 III. F nmr Of PF4C1 O O O O O O O O O 0 O O O O O O O O 11 IV. 1: Of PF4Br O O O O O O O O O O O O O O O O O O O O O 14 19 V. F nmr Of PF4Br O O O O O O O O O O O O O O O O O O 22 VI. 31F nmr Of PFaBr O O O O O O O O O O O O 0 O O O O 0 24 iv LIST OF TABLES TABLE Page I. Infrared peaks and assignments for PF4CI . . . . . . . 9 II. Mass spectral data for PF4C1 . . . . . . . . . . . . . 12 III. Infrared peaks and assignments for PFABr . . . . . . . 15 IV. Mass spectral data for PFABr . . . . . . . . . . . . . l7 I. INTRODUCTION While the simplest fluorOphosphorane, the parent compound, PPS, has been known since 1876, well before the discovery of elemental fluorine, a fluorophosphorane (defined as any compound of the variety RnPFS-n’ where R is any alkyl, aryl, amino, or halo derivative) was first referred to in the patent literature ten years agol. Fluorophosphoranes were not actually isolated and characterized completely until 1958 when A. B. Burg and his codworkersz, and w. C. Smith3 reported the preparation of some perfluoroalkyl derivatives and phenyl- and isooctenyltetrafluorophosphorane, respectively. The first known method for the synthesis of fluorophosphoranes was based upon the fluorination of complexes between PCl3 or alkyl- dichlorophosphines, alkyl halides, and AlCl34’ 5. [R PCl ]A1Cl 451+ an (n - 1, 2) n 4-n 4 Sb-F3 n S-n Later, methods for the synthesis of fluorophosphoranes were developed using various types of fluorinating agents3’ 6-9. These were followed by redox methods10 and a number of synthetic methods involving phosphorous-oxygen compoundsll, as well as the use of organometallicslz’ 13. Most of the known fluorophosphoranes are distillable liquids at room temperature. The lower alkyl or perfluoroalkyl phosphoranes are gases at room temperaturelé. Most fluorophosphoranes exhibit 15, 16 typical covalent prOperties; ionic forms are seldon observed This behavior is different than that of the halofluorophosphoranes. 2 These compounds can be ionic or covalent. They also lack the thermal stability characteristic of most phosphoranes. The halofluorophosphoranes undergo hydrolysis readily. While all the chlorofluorophosphoranes, and a few bromofluoro- phosphoranes are known, none of the iodofluorOphosphoranes have been isolatedls. Halofluorophosphoranes are usually prepared by the direct addition of the appropriate halogen to the trihalide, as shown by the following reaction17. PF +c1 33; PF C1 +PC1+PF-+PC1+F- 3 2 3 2 4 6 4 Carter and Holmes recently prepared PF4C1 using a controlled low-temperature fluorination of the molecular form of PCle318. The mixed halofluorophosphoranes PF3012 and PF3Br2 were the subject of investigation by Salthouse and Waddingtonlg. Their data indicated that these molecules possess C v symmetry, and that the 2 chloro or bromo groups tend to substitute in equatorial positions on the trigonal bipyramidal structurelg. The nmr data indicated that both PF3Cl2 and PF3Br2 have a trigonal bipyramidal structure with one equatorial and two axial fluorine atomszo’ 21. The ease with which halodifluorophosphines may be formed from dialkylaminodifluorophosphinezz-24 and the ease with which halodifluorophosphoryl and halodifluorothiophosphoryl compounds may be obtained from the corresponding dialkylaminodifluorOphosphoryls and dialkylaminodifluorothiophosphorylszs’ 26 has prompted this investigation of the action of anhydrous hydrogen halides upon dimethylaminotetrafluorophosphorane27-30. These previous results suggested that the dimethylamino group of (CH3)2NPF4 could be replaced by a halide atom to give the compounds PF4C1, PFéBr, and PFaI. 3 The last member of the PF5_nCln series, PF4C1, was completely characterized by Carter and Holmesls' 31. The characterization of PF4C1 would provide a check on the reaction of (CH3)2NPF4 with anhydrous hydrogen chloride. Neither PF4Br nor PF4 the interaction of (CH3)2NPF4 and hydrogen halides would result in a synthesis of these compounds. I had been reported and it was hoped that II. EXPERIMENTAL METHODS Standard high vacuum techniques were employed throughout. The dimethylaminotetrafluorophosphorane was prepared as described by Brown, Fraser, and Sharng. Their original process was modified for vacuum line methods and larger amounts of reactants. A 50 g sample of anhydrous dimethylamine was distilled in 19593 onto 150 m1 of freshly distilled dry toluene. Phosphorous pentafluoride was then passed over the stirred mixture which was maintained at -780 for two hours. The pressure was monitored using a manometer. The reaction was assumed to be complete when phosphorous pentafluoride was no longer absorbed by the mmine. The toluene was then removed from the white suspension by distillationIinggggg_to a -78° trap. The adduct, (CH3)2NH:PF remained behind as a white solid. After 5 the removal of all toluene, the adduct was heated to 1300 and the volatile products separated by distillation ignzgggg through traps held at -45, -78, and -196°. The -78° fraction was then re-distilled on a spinning band column under an atmosphere of dry nitrogen. The fraction which boiled between 60-630 was collected. The identity of the phosphorane was established by a comparison of its infrared spectrum with the previously reported spectrung. The proton nmr and glpc suggested that the material was more than 951 pure. The water white liquid was stored at -600 until used. The anhydrous hydrogen chloride (Matheson) was distilled ignzgggg just prior to use. Commercial anhydrous hydrogen bromide (Matheson) contained large amounts of hydrogen chloride. It was therefore necessary to prepare anhydrous HBr and HI via the reaction of bromine and iodine, respectively, with l,2,3,4 tetrahydronapthaleneaz. l. 5 The reaction bulb used for all experimental work was patterned after the model used by Treichel to prepare HPF413, and is shown in figure I. The bulb was baked at 1500 for two days, evacuated on the vacuum line and flamed before use. The infrared spectra were obtained on a Perkin-Elmer 237B grating spectrophotometer. For the region below 600 cm-1, a Perkin- Elmer 301 spectrophotometer was employed. For all volatile materials a gas cell with a 7.5 cm path length and CsI windows was used. Proton nmr spectra were observed on a Varian Model A-6O nuclear magnetic resonance spectrometer Operating at the ambient temperature of the instrument. Fluorine nmr spectra.were obtained on a Varian Model 56/60 nuclear magnetic spectrometer Operating at 56.4 Mc. For the proton spectra tetramethyl silane and methylene chloride were used as external standards. For fluorine magnetic resonances, fluorotrichloromethane was used as an external reference, by the tube interchange technique. Whenever possible, samples were run as neat liquids. Phosphorus nmr absorptions were obtained on a Varian Model HA—lOO spectrometer with 85% phosphoric acid used as a reference. All mass spectra were obtained on a Consolidated Electrodynamics Corporation Model 21-103C Spectrometer Operating with an ionizing voltage of 56V. A F&M Research Chromatograph was used for glpc, with helium as the carrier gas and a flame ionization detector. Glpc of the products was performed on a 0.125 in x 20 ft stainless steel column packed with 202 silicon gum rubber SE-30 on Chromosorb W. The column was Operated at 70°. GAS PHASE REACTION BULB Qw— \ ‘.)-|-J Figure l 7 The Reaction of Dimethylaminotetrafluorgphosphorane with Anhydrous Hydrogen Chloride. Because of the difficulties encountered when this reaction was allowed to proceed in a condensed state, the reaction between (CH3)2NPF4 and anhydrous HCl was carried out entirely in the gas phase using a system previously described by Treichel, Goodrich, and Pierce33. Two reaction bulbs of about 1 liter capacity and about 250 ml capacity were connected together with an intervening stopcock. In a typical reaction, a 6.28 mmol sample of anhydrous HCl was condensed ig_yaggg,at -l96o into the smaller bulb. A 3.51 mmol sample of (CH3)2NPF4 was then condensed ighggggg into the larger bulb at -196°. Both reactants were allowed to warm to 230 until no liquid (CH3)2NPF4 was observed. When the stopcock between the two bulbs was opened, the HCl expanded into the larger bulb. An immediate reaction was observed by the formation of a fine white solid. The volatile products of this reaction were condensed into a ~196° trap on the vacuum system and then fractionated through traps held at 0, -78, and -l96°. This fractionation was repeated three times to insure complete separation. The —780 fraction NPF contained 0.30 mmol of unreacted (CH identified by its 3)2 4’ gas-phase ir spectra. The -1960 trap contained 3.12 mmol of PF Cl. 4 The PFacl was identified by its gas-phase ir spectrum. This spectrum was identical with the previously reported spectrum of PF4C131, except those bands which were attributed to POF3 are of lower intensity. The spectrum obtained in the region from 300 to 1600 cm"1 is shown in figure II. The infrared peaks and some band 31 assignments based on the work of R. R. Holmes are listed in table I. vaaosmm mo asuuooam saunumaH any HH enemas cow oom oco com OOOH OONH coca coca . p P b L . . L AHIEOV huoosvoum 9 Table I Gas-Phase Infrared Data on PF Cl 6 Absorption (cm-1) Assignment (sz) 480(vs) v8(b1)v3(al) 530(m) PF5 565(9) v11(b2) 625(m) ? 670(5) PCle3 691(m) v2(al) 700(m) ? 868(m) v2(POF3) 895(vs) vl(a1) 903(vs) v10(b2) 925(vs) v1(b1) 945(m) PFS - POF3 954(m) POF3 995(3) POF3 1019(vs) PF5 1025(vs) PF5 1415(wm) POF 10 The product was also identified by a gas-phase molecular weight of 141.1 g/mol (theoretical 142.4 g/mol) and by its 19F nmr at -600 which exhibited a doublet pattern at +24.3 ppm (literaturelS; +23.5 ppm) from CClSF. The coupling constant (JPF) was 1000 i;10 cps (literature18; 1000 cps). The nmr spectrum exhibited no peaks which can be attributed to POF or PF C1 but did exhibit peaks which 3 3 2’ can be attributed to minor amounts of PF5 at +68 ppm with a coupling constant (JPF) of 930 cps (literature: 6 +72.5 ppm34; JPF - 916 cpszo). The 19F nmr is shown in figure III. Mass spectral data obtained at 56V also supported the formulation of PF4C1. Peaks attributed to the product with mass number and relative abundance respectively are shown in table II. NO parent peak ions were detected at 56V, which is consistent with other phosphorane results33. No vapor pressure data is reported because the sample contained traces Of HCl which were difficult to separate from PFQCl due to the similarity in vapor pressure. Carter and Holmes report vapor pressure data obtained at -123° and -54018, however all their spectral data reveal amounts of impurities in their samples. 11 Ea m3? u a A was 83 1f ace. 9 Aavaosmm mo escuuoam mzz use HHH assume :32; 12 Table 11 Mass Spectrum Of PF4C1 mZe Relative Intensity Assignment 2+ 17.5 42 0135 2+ + 18 40 11.0135 - H20 2+ 18.5 14 0137 2+ 19 5 Hc137 25 6 222+ + 2 8 57 N2 31 7 p+ + 32 12 02 33 4 90F2+ + 35 72 0135 36 100 HCl + 35 + 37 24 0137 + 38 32 H0137 44 8o PF32+ 50 10 PF+ 57 1.5 9023++ 66 1.5 P0F+ 69 62 PF2+ + + 85 68 2022 - 230135 + 87 13 PF0137 88 18 PF3+ + + 104 22 POF3 - 9220135 4. 106 2.0 PF2C137 107 18 pr + 13 The Reaction of Dimethylaminotetgggluorgphosphorgye with Anhydrous Hydrogen Bromide. This reaction was run in a manner identical with that previously described for the reaction involving HCl. The same precautions were taken with the reaction bulb to ensure anhydrous conditions. A 7.22-mmol sample of freshly distilled anhydrous HBr was condensed iguggggg at -l96o into the smaller reaction bulb. A 3.35dmmol sample of (CH3)2NPF4 was then condensed ig_x§ggg_at -l96o into the larger bulb. After warming the bulbs to ambient temperature (23°) the stopcock between the bulbs was opened to allow the 7.22 mmol 3)2NPF4' The reaction proceeded with the formation of a white powder. After sample of HBr to interact with the 3.35 mmol sample of (CH allowing the reaction to proceed for two minutes, the volatile products were condensed in a -l96o trap and then fractionated three times through traps held at -78, ~126, and ~196° respectively. Again the -780 trap contained traces of unreacted (CH3)2NPF4, identified by its gas-phase ir spectrum. A 3.22 mmol sample of a gas, later identified as a mixture of about 802 PF Br and about 202 PFS’ was 4 obtained from the -1260 fraction. A 0.35 mmol sample of HBr, identified by its infrared spectrum, was recovered from the -1960 trap. Characterization of PFABr is presented in subsequent portions of this thesis. The infrared gas-phase spectrum of the -1260 fraction exhibited peaks attributed to PF Br, POF3, and PF Several tentative 4 5' assignments on PF Br were made on the basis of assignments carried 4 out for PF4C131 assuming that PFABr and PFACl are of the some sz symmetry. Figure IV shows a typical ir spectrum and table III lists the peaks and some tentative band assignments based on PF4C131. 14 com com vaumqmm mo Esuuooom woumumaH one 000 com >H magmas OOOH OONH coca P AaIEOV mucosvouw coed 15 Table III Gas-Phase Infrared Data on FF Br Absorption (cm-1) 387(w) 461, 470, 480(8) 522(ms) 532, 542(8) 573, 584(ms) 617(mw) 675, 678(3) 725(vw) 855, 865(m) 855(vs) 899(vs) 906, 915(3) 945(vs) 955(8) 985, 990, 1000(5) 1020, 1025, 1035(3) 1365(w) L405, 1415, 1426(mw) 1570(mw) 1760(vw) 4 Assignment (02v) 06, (PBr stretch) .Br in-plane v and v 8 2 bend) 3, (PF2 and PF PF5 v (PFZBr out-of—plane bend) 11’ ? ? v (PF2 stretch) 2, ? POF3 v 1 (PF2 stretch) v (PF2 stretch) 10’ v7, (PF2 stretch) POF3 or PF5 POF3 POF3 PF5 (vlo-f-ve - 1370) POF3 ' v1 + v2 = 1560) (201 - 1770) 16 A mass spectrum obtained at 56V on the -1260 fraction demonstrates that parent ions again were not detected. The results of this data with mass numbers and relative abundances are shown in table IV. The large amount of HBr which appears is apparently formed by hydrolysis of PFéBr at the injection part of the mass spectrometer. No bands which could be attributed to HBr were observed in the infrared spectra of PFéBr at 80 mm pressure. The gas-phase molecular weight of the -1260 fraction was 178.5 g/mol (theoretical for PF Br; 186.9 g/mol). 4 The Reaction of Dimethylaminotetrafluorophosphorane with Anhydrousgflydrogen Iodide. The reaction was run as described previously. Into the throughly dried reaction bulb was condensed iggzgggg_at -l96o a sample Of anhydrous HI which had been freshly distilled. A sample of (CH3)2NPF4 was then condensed 13,33229’ into the larger bulb. The contents were then warmed to 230 and the stopcock was Opened. The reaction proceeded with the initial formation of a white powder, which turned a deep purple within seconds after forming. This mixture was fractionated through traps held at -78, -126, and -l96°. The ~78° fraction contained unreacted (CH NPFa, while the ~196° 3’2 rap contained PF3, which was identified by its gas-phase infrared f spectra and a gas-phase molecular weight of 88.3 gimol (theoretical; 88.0 gfmol). The solids remaining in the reaction bulb were a deep purple. A portion of these solids could be dissolved in 001 The visible 4. spectrum of this solution was identical with that of a solution of 31 39.5 40 40.5 41 44 50 69 80 81 82 88 104 10? 131 148 150 17 Table IV Mass Spectrum of PFaBr Relative Intensity 6.3 10.7 12.5 13.8 12.5 13.8 18.4 12.5 40 86.5 100 86 100 46 22.6 38.7 58 2.7 2.7 4.0 4.0 Assignment PF Br J" PFZBr81 18 Table IV (continued) gig Relative Intensity Assignment 189 2.3 28:2 (79)+ 191 4.7 PBr79Br81+ 193 2.3 PBr2(81)+ 208 1.0 PFBr2(79)+ 210 2.0 PFBr7gBr81+ + 212 1.0 PFBI;Z (81) 19 elemental iodine dissolved in CCla. The amount of I2 produced in this reaction was determined using standard techniquesas. These results suggest that the experimental stoichiometry for the overall reaction is 1.00(CH3 NPF + 1.99 HI-9 0.94PF + 0.92I + Solids )2 4 3 2 The solid remaining in the reaction flask may be identified, on the basis of this data, as (CH3)2NHZ+F-. Each of these gas-phase reactions was repeated several times with the same qualitative results Observed for each trial. III. DISCUSSION Anhydrous hydrogen bromide or hydrogen chloride react with dialkylaminotetrafluorophosphorane to yield tetrafluorobromophosphorane and tetrafluorochlorOphosphorane according to the equation (CH NPF + ZHX + PF X + [(CH 3)2 4 4 3)2 in contrast to this behavior, hydrogen iodide reacts with (CH3)2NPF4 +- NH2] X according to the equation (CH NPF + 2111 + PF + I + [(CH3)2NH2]+F- 3)2 4 3 2 Cavell and Charlton36 have shown that hydrophosphoryl difluoride, OPF H, is formed when (CH NPOF is allowed to interact with HI. 2 3)2 2 The fact that PFAH is not produced from the interaction of (CH3)2NPF4 and HI may be a result of the inability of PFAH to exist in the presence of reducing agents such as HI. 0f the two tetrafluorohalophosphoranes prepared by this method, PF4C1 is much more stable and easier to isolate and characterize. It has been previously reported18 that a sample of PF401 could be stored in the gaseous state for four days before deposition of a white solid on the walls of the storage container could be Observed. This observation was also corroborated in these experiments. In contrast to this behavior, when a sample of PFABr was stored in the gaseous state (in a sealed glass ampoule), it was observed that a yellow-red sci d was deposited on the walls of the ampoule within 15 minutes. A liquid sample of water white PFaBr deposits a deep red solid on the walls of the container in less than one minute at room temperature. On the other hand, no change was observed when a liquid, water white 20 21 sample of PF401 was held at room temperature for 30 minutes. This thermal instability of PF Br suggests that the compound decomposes 4 when allowed to strike the warm tubing between the cold traps on the vacuum system. That this is the case is also suggested by the fact that attempts to purify PF4Br by recycling iggggggg the partially purified material four times through traps held at -78, -126, and -l96o did not result in a tensiometrically homogeneous sample. In spite of the fact that PF Br could not be obtained in a state of high 4 purity, physical measurements presented below, strongly suggest that PFaBr in about 80% yield is obtained from the interaction of (CH NPFQ 3)2 and anhydrous HBr. First, the 19F nmr spectrum (shown in figure V) of a sample presumed to be PF4Br consisted at -600 of a simple doublet pattern at +9.6 ppm upfield from CCl3F. The coupling constant was 1085 1.10 cps. The chemical shift and the (JPF) coupling constant for PF4Br is intermediate between the values for PF5 and PF3Br237. The 19F nmr spectrum also exhibited absorptions which can be attributed to PF5 at 36607 PPm With JPF of 940'3.10 cps (literature; 34’ 20 6 - +72.5 ppm; JPF - 916 cps). On the basis of the area under the peaks, the sample contained about 20% PFS’ No absorptions attributed to PF3Br2 were observed. These data are consistent with molecular weight determina- ions which always yielded values lower then the theoretical values. As mentioned previously, we were unable to remove the PF5 by fractional distillation igngggug, At the lowest temperature obtainable on the nmr spectrometer employed, -60°, no line broadening was observed, suggesting that the molecule was exchanging intramolecularly at a rate faster than could be observed by nmr techniques at -60°. 22 Boa o.m+ I m coon e havumemm mo asuuuosm mzz use > cosmos 36 22 ||I||¥é 23 Second, the 31F nmr spectrum at -70° of this same sample (shown in figure VI) consisted of a simple quintet of 1:4:6:4:1 intensity at +72.6 ppm with a JPF Of 1075 1.15 cps. No other absorp- tions were observed probably because Of the low magnetogyric ratio Of phosphorous; the signal is about an order of magnitude less intense than that of fluorine. The lack of sensitivity accounts for the fact that the expected septet due to the 20% impurity of PF5 and the outer members of the PFABr quintet were not observed; but the 31F spectral data can only be rationalized by the existance of a tetrafluorophos- phorus group. Third, the formula PFABr is also supported by the infrared data. Several assignments may be made on the basis of assignments carried for PF40131, assuming that PFABr and PF4C1 are of sz symmetry. In the P-F stretching region of the PFABr spectrum, the very intense bands at 885, 899, and 915 cm”1 are associated with the P-F stretching modes. In PF4C1 these are found31 at 895, 903, and 921 cm-1. The PF2 Symmetric axial stretch has been assigned31 to bands at 691 cm.1 in PF4C1, and is probably observed at 675 cm-1 in PFaBr. An out-of—plane bending motion appears at 560 cm“1 in PF4C131, and is probably associated with the intense bands at 532 and 542 cm"1 in PF4Br. An intense band centered at 470 cm.1 in PFaBr and at 490 cm.1 in PF4C1 may be assigned to either PF2 in-plane bending or a PF X in-plane bending motion. Medium weak bands in 2 P64Cl which are easily ascribed to P-Cl stretching vibration appear at 427 and 434 cm.1 in PF401. These are absent in the spectrum of PFABr. However, a band at 387 cm“1 in PF4Br may be associated with a P-Br stretching motion; other bands in the infrared absorption spectra of PFABr may be ascribed to impurities. 24 All .8 m2: IIV All who who." .IIII' cos: a AHVumemm mo asuuooam mzz a H> was as an 25 Fourth, mass spectral data Obtained at 56V of PF Cl and 4 PF4Br are quite similar, as expected. See tables II and III for the mass spectral data of PFACl and PFABr respectively. In both cases parent ions were not detected, which is consistent with other fluorOphosphorane results33. Intense peaks appear at mle 4. 2 C1 and PFaBr further augment the formulation as corresponding to PF4+, PF5+, PF , and 1+. The similarity of the 4 PFaBr. The presence of molecular ions containing two bromine atoms in the spectrum of PFABr suggests that small amounts of PF3Br2 are or that some rearrangement of ion mass spectra of PF formed when PF4Br decomposesls, particles in the mass spectrometer occurs. If PF3Br2 does form, it appears to be, on the basis of the 19F nmr data, substantially removed upon distillation _i_n m. The symmetry of PFAX molecules was shown by R. R. Holmes using low temperature Raman studies to be C2v31. A comparison of the dipole moment of PFACl with the values of other phosphorous(V) chlorofluorides shows the assignment of C v symmetry to PF4C1 is 2 self-consistent. The observed value of 0.78 2_for PF4C1 is close to the gas state value observed for PC12F3 (0.68 2) shown to be a trigonal bipyramid with the two chlorine atoms located in equatorial positions38. The fact that the dipole moments are only 0.1 Q_apart supports the contention that they have the same symmetry. Small changes in electronic distribution and degrees of distortion between the two molecules could easily account for the difference. The dipole moment of the trigonal bipyramidal structure of PF401 having an axial chlorine atom (03v) was vectorially considered in terms of an axial P-Cl opposing an axial P-F bond dipoleag. 26 These calculations led to a theoretical value for the gas-state dipole moment of about 0.2 2”. Experimentally, as previously mentioned, the value was determined to be 0.78'2, suggesting that a 03v symmetry for PFACI is incorrect. For PCl4F theoretical calculations of the gas-state dipole 39 moment yielded a value of about 0.2‘2_ for a C symmetry of this 3V molecule. Experimentally the value observed was 0.21239 suggesting that the molecule does possess C3v symmetry, with the fluorine atom in an axial position. Holmes also did pure chlorine nuclear quadrupole measurements on PF4C1 and found further evidence supporting the fact that the chlorine atom is located at an equatorial site in PF40131, which is consistent with sz symmetry. WOrk done by Bartell and Hansen"o using electron diffraction techniques concluded the molecule CH3PF4 is a distorted trigonal bipyramid with methyl groups occupying equatorial positions. A microwave spectrum of CH3PF4 obtained by Cornwell and Cohen revealed that the molecule is an asymmetric top and hence cannot be either a tetragonal pyramid or an axially substituted bipyramidéo. E. L. Muetterties and coaworkersl’1 completed a study of five-coordinate phosphorous(V) fluorides. They reported that at 25° the 1’ F nmr spectrum of (C2H5)2NPF4 comprises two peaks of equal intensity. On cooling the sample, the doublet gradually broadens and eventually is resolved into two doublets‘l, This data establishes a structure in which there are two fluorine atom environments, each of which contains two fluorine atoms. Because geometry closely approximating a trigonal bipyramid prevails in 27 (CZHS)2NPF4’ there are only two possible geometrical isomershl. Apical substitution may be ruled out since in no way could a 2:2 fluorine atom environment be generated. Thus we expect that PFéBr, a RPFa type molecule, does possess 02v symmetry like PF4C1. The larger size of the bromine atom may suggest more axial distortion than found in PF Cl. 1. The hydrolysis of PF X has been observed; a sample was I, placed in the infrared gas cell and the results of an addition of moisture were observed. The spectrum shows an increase in the bands due to POF3, and traces of HCl are noted with the PF4C1, while peaks due to HBr are observed upon the hydrolysis of PFABr. The suggested stoichiometry of the hydrolysis reaction is HOH + PFAX + [PF40H + H1] +-HF + BK 28 Suggestions for Future Research Much work has been done on the phosphorus(V) chlorofluorides, of which PF4C1 was the last.member to be synthesized. However, much less is known about the phosphorus(V) bromofluorides, and phosphorus(V) iodofluorides. Perhaps dipole moments and low temperature Raman studies could be done on PF4Br in a manner similar to the work of Holmes with PF4C131. The ionic and covalent behavior of these compounds is also of interest. The preparation of PFAI may be feasible and should be further investigated. The P-Br bond should be studied and perhaps its lability may be utilized in the development of synthetic routes for the preparation of new fluorophosphoranes. It may be possible to synthesize and characterize FAPPFA, PFACN, and FaPOPFA by a series of reactions analogous to the reactions which were developed by Rudolph to prepare FzPPFz, PFZCN, and FZPOPF2 from PFZI"2 and the pentavalent phosphorus reactions developed in this thesis. 10. ll. 12. 13. 14. 15. 16. 17. 18. 20. 21. 22. BIBLIOGRAPHY R. P. Carter and H. Coates, U.S. Patent 2, 853, 515 (1958). A. B. Burg and G. Brendel, 3. Am, Chem,.§g§., 80, 3198 (1958). W. C. Smith, U.S. Patent 2, 904, 588 (1959). H. Coates and R. P. Carter, British Patent 734, 187 (1955). I. P. Komkov, 2h. Obshch. Khim., 32, 301 (1962). R. Schmutzler, £2229.l§9° (London), 1868 (1962). R. Schmutzler, Inggg..ghgm., 3, 410 (1964). W. Mahler, ,1. Am. Chem. _S_g§_., 84, 4600 (1962). J. F. Nixon, ;. 13933. 33,151. 9m” 27, 1281 (1965). R. Schmutzler, lggrg..ghgm., 3, 421 (1964). w. 0. Smith, _._7_. Ag. glam. _s£_.-;_., 82, 6176 (1960). P. M. Treichel, Chem, Egg,‘Ngg§., Sept. 27, p. 53 (1965). P. M. Treichel and R. A. Goodrich, lgggg,.ghgm., 4, 1424 (1965). R. Schmutzler,‘édg§n. in Fluorine Chem., 5, 83 (1965). R. R. Holmes,.g. Chem. Educ., 40, 125 (1963). L. Kolditz, g, Chem,, 2, 291 (1963). T. Kennedy and D. S. Payne, 3, Chem,'§gg., 1228 (1959). R. P. Carter and R. R. Holmes, Inorg. Chem., 4, 738 (1965). J. A. Salthouse and T. C. Waddington, Spectrochem. Acta., Vol. 23A, 1069 (1967). E. L. Muetterties, W. Mahler, and R. Schmutzler, Inorg. Chem., 2, 613 (1963). R. R. Holmes, R. P. Carter, and G. E. Peterson, Inorg. Chem., 3, 1784 (1964). R. W. Rudolph, J. G. Morse, and R. W. Parry, Inorg. Chem., 5, 1464 (1966). 29 27. 28. 29. 30. 31. 32. 33. 41. 42. 30 R. G. Cavell, 3. Chem. Soc., 1992 (1964). J. G. Morse, K. Cohn, R. W. Rudolph, and R. W. Parry, Inorg. Syn., 10, 147 (1967). R. G. Cavell, Can. 3. Chem., 45, 1309 (1967). A. Muller, H. G. Horn, and O. Glemser, Z. Naturforsch., 20b, 1150 (1965). R- Schmutzler, Angew. Chem., 76, 893 (1964). O. H. Brown, C. W. Fraser, and D. W. A. Sharp, Chem. Ind. (London), 367 (1964). D. H. Brown, C. W. Fraser, and D. W. A. Sharp, J. Chem. cc., A 171 (1966). G. C. Demitras and A. G. MacDiarmid, Inorg. Chem., 6, 1903 (1967). R. R. Holmes,.l. Chem. Phys., 46, 3718 (1967). C. J. Hoffman and E. A. Heintz, Inorg. Syn., 7, 180 (1963). P. M. Treichel, R. A. Goodrich, and S. B. Pierce, J. Am, Chem Soc., 89, 2017 (1967). R. R. Holmes and W. P. Gallagher, Inorg. Chem., 2, 433 (1963). R. B. Fisher, "Quantitative Chemical Analysis," W. B. Saunders CO., Philadelphia, Pa., 1956 pp. 296-299. T. L. Charlton and R. G. Cavell, Inorg. Chem., 6, 2204 (1967). W. Mahler and E. L. Muetterties, Inorg. Chem., 4, 1520 (1963). R. R. Holmes and R. P. Carter Jr., J, Chem. Phys., 43, 1645 (1965). A. A. Maryott, S. J. Kryder, and R. R. Holmes, J, Chem. Phys., 43, 2556 (1965). L. S. Bartell and Kathryn W. Hansen, Inorg. Chem., 4, 1777 (1965). E. L. Muetterties, W. Mahler, K. J. Packer, and R. Schmutzler, Inorg. Chem., 3, 1298 (1964). R. W. Rudolph, Ph.D. Dissertation, University of Michigan, Ann Arbor, Michigan, 1966.