THE UN I V E R S I T Y OF M I C H I G A N COLLEGE OF LITERATURE, SCIENCE, AND THE ARTS Department of Chemistry Technical Report A VIBRATIONAL STUDY OF SOME LEWIS ACID-BASE COMPLEXES CONTAINING PHOSPHORUS AND BORON Richard J. Wyma ORA Project~ 04956 under contract with: NATIONAL SCIENCE FOUNDATION GRANT NO. G-21408 WASHINGTON, D. C. administered through: OFFICE OF RESEARCH ADMINISTRATION October 1964 ANN ARBOR

This report was also a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The University of Michigan, 1964.

ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Professor Robert C. Taylor for his encouragement and guidance throughout the course of this investigationo The author is also grateful for the financial support given to him by the National Science Foundation (G-21408), the E. I, Du Pont de Nemours & Co. and the Rackham School of Graduate Studies. The assistance of Miss Mary K, Schmidt in obtaining the Raman Spectrum of (CH3)2NPF2B10F3 is gratefully acknowledged. ii

TABLE OF CONTENTS Page LIST OF TABLES.......o.............. v LIST OF FIGURES,....,,oo..............,..,.. vii INTRODUCTION... o,,, o o o o. o, o 1 STATEMENT OF PROBLEM.........,, o.,... 6 HISTORICAL BACKGROUND O *.,.,,.......... o. o o. 9 EXPERIMENTAL PROCEDURES......,. o...... o.. 13 Preparation of Samples..o......o..o o o.... 13 Miscellaneous Procedures. o 000.. o 000 o.000.. o00 18 Spectroscopic Methods and Equipment o..,,,... 20 EXPERIMENTAL RESULTS AND INTERPRETATION O o........ 28 General Discussion of Spectra. o,...o........ 28 Polarization Results 3...........,,..o.ooOoOo, 30 Spectroscopic Results and Assignments 3,,....., 31 Trifluorophosphine, PF3.e.,.o.,.,,, 0.. o 32 Trifluorophosphine-Borane, F3PBH3 and F3PB1 H3..o.000.000.0. 0... 00000 36 Methylaminodifluorophosphine-Borane CH3NHPF2BH3.... o.. o o o o o o o. 49 Bis (methylamino)fluorophosphine-Borane (CHNH)2 PFBH3 0 0.00..0 0.0.00 o... o 64 Dimethylaminodifluorophosphine, ( CH3 )2NPF2.................. 0 0 74 Dimethylaminodifluorophosphine-Borane, (CH3)2NPF2BH3 and (CH3)2NPF2B1D3 000o o0 88 Dimethylaminodif luorophosphine-Boron Trifluoride, (CH3 )2NPF2BllF3 and (CH3)2NPF2B 0 o o 1 08 (C 2P -F3...................,.. 108 iii IIIb~

Page Bis(dimethylamino)fluorophosphine, [(CH3)2N]2PF 0.00...! 0000 118 [(CH3)2NJ2PF.....o o.......oooo 118 Bis (dimethylamino )fluorophosphine-Borane [(Cd3 )2N]2PFB11H3 o o 0 0 o oo oo 127 Trimethylphosphine, (CH3 )3P o o o o o o o o o o 136 Trimethylphosphine-Borane, (CH3)3PBH3. (CH3 )3PB10H and (CH3)3PBllD3oo... o o~ 145 SUMMARY 0 0 0 0 o 0 0 0 0 0... e0 o0 o 0 0 o o 0 0 164 APPENDIX A - Preparation and Purification of Materials,o,, o,. o, o o,, o o o... o 173 APPENDIX B - Preparation and Purification of Samples..0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 177 BIBLIOGRAPHY 1. 0.................0 83 iv

LIST OF TABLES TABLE Page 1 The Observed Raman and Infrared Frequencies of PF3....... o.. o.. o...8 34 2 The Fundamental Vibrations of F3PBH3, C3v Symmetry o.... 08o. 8..o. o..oo 40 3 The Observed Infrared and Raman Frequencies of F3PBH3 and F3PB10H3...............o 41 4 A Comparison of the Fundamental Frequencies of PF3 Ni(PF3)4, 0PF3, OCBH3 and F3PBH3 (cm-) o o o e 46 5 The Observed Infrared and Raman Frequencies of CH3NHPF2BH3...,.................. 56 6 The Observed Infrared and Raman Frequencies of (CH3NH)2PFBH3.... o o o 0... 69 7 The Fundamental Vibrations of (CH3)2NPF25 Cs Symmetry. 0 ~. o.. o o 79 8 A Comparison of the Observed Vibrational Frequencies of Ni(PF2CH3 )4, (CH3 )2NH, (CH3)2NPC12 and (CH3)2NPF2...............0 81 9 The Observed Infrared and Raman Frequencies of (CH3 )2NPF2........................... 83 10 The Fundamental Vibrations of (CH3 )NPF2BH3 C. Symmetry 3ooo o2o 2o o oooooooo oJ.o...o8 96 11 The Observed Infrared and Raman Vibrational Frequencies of (CH3)2NPF2BH3 o8..oo.....,,, 98 12 The Observed Infrared and Raman Vibrational Frequencies of (CH3)2 NPF2B D3OO110...... 100 13 A Comparison of Some Isotope Dependent Vibrations of (CH3 )2NPFBH3 and (CH3)2NPF2B D3 (cm ) 0 102 14 A Comparison of Observed Infrared and Raman Frequencies of (CH3)3NB 1F3 (CH3)2NPF2 and ( CH3 )2NPF2B F3 0 0 0 0 80 0 0 0 0 8. 0 0 0 0 0 0 0111 V

TABLE Page 15 The Observed Infrared and Raman Vibrational Frequencies of (CHN3 )2NPF2B10F3 and (CH3)2NPF2B F3..O..000o... 00.... 11 4 16 The Observed Infrared and Raman Frequencies of [(CH3)2N]2PF....... 0............ o 123 17 The Observed Infrared and Raman Frequencies of [ (CH3 )2N]2PFB11 3..,.,, o.... o o 131 18 The Fundamental Vibrations of (CH3)3P, C3v Symmetry,...... 0 ^.. 0 140 19 The Observed Raman Vibrational Frequencies of Liquid and Solid (CH3)3P......, 141 20 A Comparison of the Observed Vibrational Frequencies of Trimethylphosphine and Related Compounds (cm )., e,o o o.. 142 21 The Fundamental Vibrations of (CH3)3PBH3 C3 Symmetry.. 0 0 0 0 0 0 0 00 0 0 0 151 22 The Observed Raman and Infrared Frequencies of (CH3)3PBlOH3 0.. o............. 153 23 The Observed Raman and Infrared Frequencies of (CH3)3PBH3 000.,.00 0 0 0..0 0 0 0........~~ 155 24 The Observed Raman and Infrared Frequencies of (CH3 )3PB11D3 0 o..........o..... 157 25 Frequency Assignments for the Fundamental Vibrations of Trimethylphosphine-Borane,,,. 159 26 A Comparison of the Skeletal Vibrations of (CH3)3P and (CH3)3PBH3,,..o o o o.. 167 27 The P-B Stretching Vibrations of Several Phosphine-Borane Complexes................ 172 vi

LIST OF FIGURES FIGURE Page 1 The Low Temperature Infrared Cell 0........... 23 2 The Experimental Arrangement for Obtaining Raman Spectra of Liquids 0 O............. OOOOO 24 3 The Experimental Arrangement for Obtaining Raman Spectra of Solids at Low Temperatures o, 24 4 The Infrared Spectrum of Solid F3PB10OH (-180~C) o. 6 o 0...................... o 37 5 The Infrared Spectrum of Solid F3PBH3 (180~C) 38 3 ( 6 The Raman Spectra of Solid F PBH3 10 3 3 and FBPB10H3 (180~C)...0........ 39 and F3PB H3 (-18o0C) 000000060000000000000000 39 7 The Infrared Spectrum of Gaseous CH3NHPF2BH3 (6 mm)............................. 52 8 The Infrared Spectrum of Solid CH3NHPF2BH3 (-18 0C) 0 * 0 0 0 0 0 0 o 0 0 o o o 0 o o 0 0 o o o 0 O 0 O 0 O 0 0 O 0 53 9 The Raman Spectra of Solid (-180~C) and Liquid (-150C) CH3NHPF2BH3 0000000.00000000000. 54 Liquid ( 15 ~C ) CH3NHPF2BH3 o o o o o o c 6 o o o o o o o o o o o o o o 54 10 The Polarized Raman Spectra of Liquid CH3NHPF2BH3 (-4o0C) 00 00 1 0 0 00O00 55 11 The Infrared Spectrum of Liquid (CH3NH )2PFBH3 0 (25C00 ) 0 f 0 0 0 0 0 0o o o 0 ff o0 o 0 00 00o00 00 00 000 0 00 66 12 The Infrared Spectrum of Solid (CH3NH)2PFBH3 (-180~C) o............................ o....... 67 13 The Raman Spectrum of Liquid (CH3NH)2PFBH3 (o C ) O O O 0 0 0 0 0 O O O O o O.0 O O 0 00 00 00 68 14 The Raman Spectra of Solid (-l80~C) and Liquid (-320C) (CH3 ) 2NPF2 0000000000 o o 0000 77 15 The Polarized Raman Spectra of Liquid (CHi3) 2NPF2 ( 320C) 0000000000000000000000000 78 16 The Raman Spectra of Solid (-180~C) and Liquid (-24~ C ) (CH3 )2NPFP2BHi3 0e R S o 90 17 The Polarized Raman Spectra of Liquid (CH3)2NPF2BH3 (-15~C) 0.................. Q 91 vii

FIGURE Page 18 The Infrared Spectrum of Gaseous (CH3)2NPF2BD 3 (11 nmm)................... 92 19 The Infrared Spectrum of Liquid (CH3)2NPF2BllD3 (25~C) o........... 93 20 The Raman Spectra of Solid 1-l800C) and Liquid (-29~C) (CH3)2NPF2BlD 3.........0o o 94 21 The Polarized Raman Spectra of Liquid (CH3)2 NPF2B113 (-29~C)...........oooo o95 22 The Raman Spectra of (CH3)2NPF2B10F3 and (CH3)2NPF2B -F3 (-780C) oo00000000000000000000 110 23 The Raman Spectra of Solid (-180~C) and Liquid (-45~C) [(CH3)2N]2PF oo.............. 121 24 The Polarized Raman Spectra of Liquid [(CH3)2N]2PF (-45~C)........................0o 122 25 The Raman Spectra of Solid (-1800C) and Liquid (0 C) [(CH 3)2N] 2PFBllH3...........o 129 26 The Polarized Raman Spectra of Liquid (0~C) [ (CH3)2N]2PFB1 H3..... 6 0 o 0 o 00 0o o 0 130 27 The Raman Spectra of Solid (-141lC) and Liquid (-40~C) (CH3)3P 1o3000. ooo.,..o.oo.o 138 28 The Polarized Raman Spectra of Liquid (CH3)3P (-40~C) 0..................00000000., 139 29 The Infrared Spectrum of (CH3)3PB10H3 (-18o c) 3o O o3oo o147o oooo o 4 30 The Infrared Spectrum of (CH3)3PBH3 (-180~C).. 148 31 The Infrared Spectrum of (CH3)3PBllD3 (l180~C). 149 32 The Raman Spectra of (CH3)3PB10H39 (CH3)3PBH3, (CH3)3PB11D3 (-10~C) 0 0oo00 0eooooooooooo oooe 0 150 33 A Correlation Diagram for Some Fundamental Vibrations of PF3, F PBH, (CH3)2NH, (CH) NPF o (CH3)2NPF2BH3, [(CA3)2] 2NPF and [ (CH323]2PBH 166 34 A Correlation Diagram for the Borane Stretching Fundamentals of Various Borane Lewis Acid-Base Complexes..0, o 0 o, o o 0 0 0o o o 0 0 0 o. 0 0 0 o o0 o 0 0 0 0 o0 0 169 viii

FIGURE Page 35 A Correlation Diagram for the Borane Stretching Fundamentals of Some Phosphine-Borane Complexes 170 ix

INTRODUCTION In recent years a great deal of interest has been shown in the complexes which are formed from Lewis acids and bases. One reason for this interest is that the formation of the complex may stabilize either the Lewis acid or base, which may be unstable by itself. A good example of this is shown in the reactions of diborane with Lewis bases. Experimental evidence, primarily from kinetic studies, suggests that diborane dissociates into the borane:group,. BH3, which is thought to be the principal intermediate in reactions involving diborane. The borane: group.. has never been isolated and presumably has only a transitory existence. However, many complexes are known which contain the borane group joined to Lewis bases such as the amines, phosphines, ethers and the like. In all these complexes the BH3 fragment is stabilized by the dative bond to the base. In many cases the complex is formed without the necessity of a bond rupture on the part of the parent Lewis acid or base prior to reaction. The only energies which are involved in the process of complex formation are those which pertain to the possible rearrangement of the parent species plus that of the dative bond formation in the complex. A spectroscopic study of such compounds could afford valuable insight into their chemical bonding and physical properties. From a spectroscopic point of view one would expect the principal differences which might appear to be largely a 1

2 L- X consequence of the dative bond formed in the complex. A great deal of discussion has been given to the subject of bonding in Lewis acid-base complexes and to the interpretation of stabilities, heats of formation and other thermodynamic properties in terms of the electronic structures of the acid and base or the complex. Complexes containing phosphorus have received considerable attention particularly with respect to the role of the d-orbitals of the phosphorus. Traditional acid-base and coordination theories suggest that compounds of phosphorus(III) should serve as ligands in complex formation because of the lone pair of electrons on the phosphorus atom. It is evident that the nature of the groups attached to the phosphorus atom will greatly affect the availability of the lone-pair electrons. One would expect that electron withdrawing groups on the phosphorus, eog. the halogens, would tend to make the lone-pair electrons less available for contribution. On the other hand, electron releasing groups such as CH3-, CH3NH-, (CH3)2N-, etc. would tend to release electrons to the phosphorus, thereby making the lone-pair electrons less tightly bound by the phosphorus and more available for coordinate bond formation. Somewhat contrary to the above expectations based on simple theory, trifluorophosphine is known to form fairly stable complexes with transition metal ions of platinum (1,2) and nickel (3). In order to account for the unusual

stability of the complexes, Chatt and Williams (1) proposed that a different type of bonding was present. They invoked the concept of w-bonding between the d-orbital electrons on the metal ion and the vacant orbitals on the ligand. As a corollary, it was postulated that PF3 should not form a complex with Lewis acids such as BF3, BH3, or AlCl3 (1,2) since these have no d-orbitals. However, with the synthesis of F3PBH3 in this laboratory by Parry and Bissot (4), the ability of PF? to coordinate with the borane Lewis acid was demonstratedO Taylor and Bissot (5) have provided evidence from a Raman spectroscopic study that this complex probably has an ethane-like structure and contains a typical dative P-B bondo Although not in accord with Chatt s ideas, the formation of this complex was not entirely unexpected in view of the marked similarity of PF3 and CO and the known existence of the compound H3BCO (6). Attempts have been made to rationalize the nonexistence of F3PBF3 while still accounting for the existence of F PBH3o Graham and Stone (7) proposed that the borane complex existed because in addition to the adative bond formed from the lone-pair electrons on the phosphorus, there was additional w-bonding between the vacant 3d-orbitals of phosphorus and a "pseudo' orbital" provided by the delocalized electrons in the borane group. The non-existence of F3PBF3 was attributed to the

4 fact that the fluorine atoms on the boron attract electrons to such an extent that delocalization cannot take placeo Arhland, Chatt, and Davies (8) invoked a similar argument and attributed the existence of F3PBH3 to hyperconjugation effects. It should be mentioned here that several workers have reported previously the synthesis of C13PBBr3 (9) and C13PBF3 (10) and it is surprising to note that very little has been mentioned in the literature (2) to explain the bonding in these complexes. The synthesis of F3PAlC13 in this laboratory by Alton (11) necessitated a re-examination of the bonding question since the formation of this complex could not be explained adequately by using either the w-bonding or delocalization concepts. Alton suggested that perhaps this bonding issue was not as complex as the previous authors tried to make it but could be explained in the more conventional terms of steric and electronic effects. For example, he tried to show by approximate calculations that more energy is required to deform planar BF3 into the pyramidal configuration found in complexes than is needed for BH3 (11), thereby making the F3PBH3 complex thermodynamically favored over F3PBF3. Also, from steric considerations, the smaller BH3 is able to bring the bonding site on the boron closer to the phosphorus allowing for greater stability to be achieved upon complexationo The existence

5 of F;PAlC1 was attributed to the probable ease of deformation of A1C1l (11). Subsequent to the discovery of F3PBH3 a new series of related compounds was synthesized by Kodama (12) in this laboratory. His work involved reactions in which ammonia and the methyl amines were used to substitute for fluorine in F3PBH3 Sr. Fleming (13) in continuing this work, studied the dimethylamine series in greater detail and by means of infrared and NoM.R. spectra was able to show fairly conclusively that the borane group was bonded to the phosphorus in the complexes. Beyond this information, little or no structural data have been accumulated for these interesting complexes. Likewise no attempt has been made to gather information which might lead to a better understanding of the nature of the P-B dative bond.

STATEMENT OF PROBLEM Although a large number of complexes containing phosphorus and boron are known. a number of questions concerning these complexes still remain unanswered. Structural data for them are relatively meager and in many cases only those physical properties are known which are used in their characterization. A spectroscopic investigation of these complexes would be of interest to supplement the data already accumulated and help provide a better understanding of these complexes. Since the number of phosphorus-boron complexes is appreciable, a selection must be made for the present work. Several borane complexes, first investigated by Kokama (12), offer a very interesting series to study spectroscopically. These include the following: F3PBH3, CH3NHPF2BH3, (CH3NH) 2PPBH., (CH3) 2NPFBH3, [ (CH3 )N]2PFBH3 and (CH) 3PBH3. A Raman investigation of (CH3)2NPF2BF3 is also of interest for comparison with the preceding because of its unusual nitrogen-boron bonding. Detailed information on the vibrational spectra of the preceding compounds would be of interest in connection with the general question of bonding since the vibrational frequencies can be directly related to bond strength. The actual calculation of force constants for molecules of the size being considered may not be a simple process and may not lead to an unequivocal answer. However. this is a 6

7 problem in itself and is a step removed from the experimental one which is an essential prelude to any further theoretical work. In addition to the information which may be garnered concerning bonding, the vibrational data may also provide a basis for the comparison of the properties of the Lewis acids, bases and complexes formed from them. Such data would contribute to the general fund of knowledge accumulated for other Lewis complexes in this laboratory and in addition would be pertinent to the general question of stability, strength of the dative bond, and related characteristics. A somewhat secondary but still useful benefit from this investigation would be the identification of characteristic frequencies which might be associated with certain groups of atoms found in these molecules. The vibrational spectra of these complexes may be expected to be relatively complex due to the large number of atoms in the moleculeso The large number of vibrational modes resulting may create difficulties in making assignments and increase the possibility of modes being accidentally degenerate. Due to the low symmetry of many of these molecules none of the vibrational modes are truly degenerate and this will increase the difficulty of interpretation of the spectra. Since many of the bases of interest contain methyl groups, some simplification can be achieved by virtue of

8 the fact that the methyl group frequencies are well known and easily identified. To a good approximation, the frequencies remaining after elimination of the methyl modes can be considered to arise from a molecule in which the methyl groups act as point masses. The problem can be simplified further by examining the vibrational spectrum of the ligand whenever possible and then comparing this spectrum to that of the complex. Polarization information and data obtained from isotopic substitution will also facilitate making assignments.

HISTORICAL BACKGROUND Even though spectroscopic work on compounds containing phosphorus and boron is not voluminous, data for a large number of closely related compounds have been given in the literature. Trifluorophosphine has been examined spectroscopically by several workers (14,15,16,17). Despite the relative simplicity of the molecule, a surprising amount of confusion has existed regarding the spectra and the assignment of fundamentals. For example, Yost and Anderson (14) assigned the frequencies 531 cm-1 and 486 cm- to the respective totally symmetric and degenerate P-F bending vibrations whereas Gutowsky and Liehr (15) reported the reverse of these assignments. Somewhat later, Wilson and Polo (16) assigned the frequency 487 cm-l to the totally symmetric bending mode, a new band at 344 cm-1 to the doubly degenerate bending vibrations, and reported no evidence of -1 a band at 531 cm o A recent Raman investigation by Taylor (17) agreed with Wilson and Polo's findings, although polarization data were not obtained. The Wilson-Polo-Taylor assignment is probably correct and it is possible that the 531 cm band reported earlier is due to a PF2C1 impurity. The complexation of PF3 with a metal such as nickel (18 19) or Lewis acids like oxygen (15 20) and BH3 (5) produces noticeable shifts in the P-F stretching and bending 9

10 vibrations, the largest shifts occurring with P-F stretching motions. For example in comparing the frequencies of PF3 (17) and F3PBH3 (5), the symmetric and degenerate P-F stretching frequencies at 874 and 832 cm1 were found to shift to 944 and 958 cm 1 respectively, in the complex, while the symmetric and degenerate bending frequencies shift from 484 and 351 to 441 and 370 cm-1, respectively. A similar shift is noted for the Ni(PF3)4 complex (19). In this case, the symmetric and doubly degenerate P-F stretching frequencies are 859 and 898 cm-1 while the symmetric and doubly degenerate bending frequencies have been assigned at 503 and 386 cm-1 respectively. A Raman spectroscopic study of F3PBH3 and F3PBD3 carried out by Taylor and Bissot (5) allowed frequency assignments to be made for the eleven active fundamentals predicted for a C3 structure. Frequencies were assigned on the basis of correlation made with Raman spectra of PF3 and H3BCO (21)o An infrared spectroscopic study of (CH3)2NPF2, (CH3)2NPF2BH3, and (CH3)2NPF2BF3 has been reported by Sro Fleming (13)o Tentative assignments were made on the basis of correlations between the vibrational frequencies of these compounds and those of PF3 (17), F3PBH3 (5), (CH33N (22), (CH3)2NH (22), (CH3)2NPC12 (23) and (CH3)3NBF3 (24)o Data obtained from NoMoRo studies and supplemented by the infrared data indicated that the BH3

11 group bonded through the phosphorus atom while the BF3 bonded through the nitrogen atom in these complexes. The infrared evidence was based on the intensities of two characteristic group frequencies, one described as a symmetric C2-N-P -1 stretch around 1007 cm and the other as a N-P-F2 stretch around 693 cm1. In the BH, complex, the relative intensity of a band assigned to a N-P-F2 motion was lessened apopreciably compared to the free ligand, while in the BF; complex the intensity of a band assigned to a C2-N-P motion was diminished. The P-F symmetric and asymmetric stretching frequencies were in the range of 767 to 827 cm-1 and 814 to 880 cm-1 respectively. Some doubt remained concerning the correct assignment of the symmetric or asymmetric P-F stretching frequencies and also the assignments for the BH3 deformation and rocking modes in the borane complex. Further spectroscopic study appears desirable to confirm the results and conclusions ar-l to provide additional data. An infrared spectroscopic analysis of [(CH3)2N]2PF and [(CH3)2N]2PFBH3 has been reported in the literature (13), and N.M.R. data have indicated the existence of a B-P bond in the latter. Tentative assignments have been made for a large number of vibrationsbut a Raman investigation including a polarization study is desirable to observe some of the lower frequencies and to provide additional information about the symmetry of the vibrational bonds. No detailed spectroscopic data concerning the

12 CH3NHPF2BH3 and (CH3NH)2PFBH3 complexes have been reported, except for some unpublished results by Kodama (25). No structural data have been accumulated or reported. A great deal of information is available in the literature concerning trimethylphosphine. Its physical properties (26,27,28), spectral properties (29,30,31.32) and molecular parameters are known (33). Spectroscopic results include both infrared and Raman data although the earlier workers, Rosenbaum, Rubin and Sandberg (29), and Wagstaffe and Thompson (30) made no attempt to assign frequencies to the fundamental modes of vibration. Later Siebert (31) made some assignments using existing data and performed a few force constant calculations. Finally, Halmann (32) using infrared data made more complete assignments; however these latter assignments were made without the use of polarization data and no attempt had been made to reproduce the earlier Raman data of Rosenbaum, Rubin and Sandberg (29), Trimethylphosphine-borane is a very stable crystalline solid and its physical properties are known (34,35). However, its vibrational spectrum has not been published in the literature. Daasch and Smith (36) have studied an analogous complex, (CH3)3PO, and have reported both its infrared and Raman spectrum and have made assignments to its vibrational modes. These vibrational data should be of use in the present spectral study of (CH3)3PBH3. ^ 3.3

EXPERIMENTAL PROCEDURES Preparation of Samples Most of the samples used for the present spectroscopic study were prepared in this laboratory. The methods of preparation have been described in the literature, in previous dissertations or represent minor modifications of existing methods. Specific details of the preparation and purification of the samples are given in Appendix Bo Throughout the preparative work vacuum line techniques were used extensively since many of the materials involved are moisture or air sensitive and also possess toxic properties. Trifluorophosphine-Borane, F3PBH3 and F3PB1 H3 Trifluorophosphine-borane, F3PBH3, containing the natural isotope mixture of boron and hydrogen, was prepared by direct reaction of B2H6 and a five fold excess of PF3. Due to the fact that a large quantity of this material was needed not only for the spectroscopic work but also for preparing other samples, the method of preparation as.-. reported by Parry and Bissot (4) was modified slightly and scaled up. The pure F3PBH3 exhibited a vapor pressure of 23 mm at -t11.8~C and melted at -116Co. Trifluorophosphine-borane enriched in the boron-10 isotope was prepared similarly using somewhat smaller amounts of starting material and retaining a four to five 13

fold excess of PF3o The method of purification was identical to that of F3PBH3o Methylaminodifluorophosphine-Borane, CH NHPF2BH?. A procedure for the preparation of CH3NHPF2BH3 has been described by Kodama (12) in which CH3NH2 was allowed to react with F3PBH3 under mild conditionso The conditions of the reaction which Kodama describes were followed with few modifications. However, the amounts of the reactants were doubled so that a larger quantity of product would be obtained. The pure liquid CH3NHPF2BH3 obtained was stable at room temperature (melting point -65~C) and showed a vapor pressure of 9 mm at 250C. This value compared favorably with 8.7 mm at 250C reported by Kodamao Bis(methylamino)fluorophosphine-Borane, (CH3NH) PFBH3 Kodama (12) has reported that the disubstituted complex, (CH3NH)2PFBH3 is formed by the reaction of CH3NH2 with F3PBH3 under relatively severe conditions. The procedure which he described was used for the present study. Pure (CH3NH)2PFBH3, a colorless nonvolatile liquid, exhibited an infrared spectrum which was identical to that reported by Kodama (25). Dimethylaminodifluorophosphine, (CH3)2NPF2 Kodama (12) reported the preparation of (CH3)2NPF2 by the reaction of (CH3)2NPF2BH3 with (CH3)3N as shown by the equation

15 (CH3) NPF2BHI + (CH3)3N -+ (CH3)3NBH + (CH) NPF2 However, Sro Fleming (13) has reported the formation of (CH3)2NPF2 by the direct reaction of PF3 and (CH3)2NH and this method was used in the present study. The purified compound had a melting point of -87~C and showed a vapor pressure of 94 mm at 0~C which compared favorably with the value of 93.4 mm reported by Kodama and Sr. Fleming. The purity of the purified product was checked further by examining the infrared spectrum and comparing it to that reported by Sro Fleming. Dimethylaminodifluorophosphine-Borane, (CH3)2NPF2BH3 and (CH3) 2NPF2B11D3..... 3* The synthesis of (CH3)2NPF2BH3 has been reported in the literature (12) by allowing F3PBH3 to react with (CH3)2NH and also by the direct reaction of (CH3)2NPF2 and B2H6o Sro Fleming (13) found the latter method more satisfactory if sufficient amounts of (CH3)2NPF2 were availableo The latter method of preparation was employed for this worko The purified product (melting point -56.7~C) exhibited a vapor pressure of 17 mm at 25~C which compared favorably with 16o7 mm at 25o5~C reported by Kodama (12)o An infrared spectrum of the product was identical to that recorded by Kodama (25) and Sro Fleming (13)o A sample of (CH3)2NPF2B11D enriched in boron-ll and deuterium was prepared by the direct reaction of excess

16 (CH3)2NPF2 with B211D6 using the same procedure. Dime thylaminodifluorophosphine-Boron Trifluoride.L (CH3)2NPF2BlF and (CH3)2NPF2B10F3.2...... _ 3............ l..... __.......The method of preparation for (CH3)2NPF2BF3 has been reported by Sro Fleming (13) and involved the direct reaction of (CH3)2NPF2 with BF3o This method was utilized for the present study with the exception that BF3 enriched in boron-10 and 11 was. used instead of the natural isotope of boron. The purified product exhibited a dissociation vapor pressure of 56 mm at 24~C and its infrared spectrum was identical to that reported by Sro Fleming. Bis(dimethylamino)fluorophosphine, [(CH3)2N]2PFo..., _......'.... ~_.., ~.....,..,,. The method used to synthesize the [(CH3)2N]2PF ligand has been discussed by Sro Fleming (13) and involved the reaction of (CH3)2NPF2 with excess (CH3)2NH in a sealed tube. The purified liquid [(CH3)2N]2PF exhibited a vapor pressure of 4 mm at 0~C and its infrared spectrum was the same as that reported by Sr0 Flemingo Bis (dimethylamino)fluorophosphine-Borane [ (CH) 2N]2PFB 1H3o Kodama (12) has reported the formation of the [(CH3)2N]2PFBH3 complex by the reaction of (CH3)2NH on (CH3)2NPF2BH3. A more straightforward reaction also was reported by him and later by Sro Fleming (13) in which [(CH3)2N]2PF was allowed to react with B2H6 yielding the desired complexo The latter procedure was utilized for the

17 present work, the only modification being that B2H6 enriched in the boron-ll isotope was used. The vapor pressure of the pure sample was not measurable at room temperatureO Kodama (12) has reported the freezing point of the disubstituted complex as -15.2~C. The infrared spectrum of the sample was identical to that reported by Sro Fleming (13). Trimethylphosphine, (CH3)3P o A sample of (CH3)3P was obtained from the Evans Chemical Laboratory at Ohio State University through the generosity of Professor Sheldon Shore and Mr. Gerald Mc Achrano The (CH3)3P had been prepared by the procedure of Mann and Wells (27) as described by the following reactions: CH3I + Mg ether CH3MgI 3 CH3MgI + PC13 ether (CH ) P + 3 MgClI (CH )3P + AgI KSti 32PA(CH3 ) 3PAgI (CH3)3P + AgI The sample was purified in this laboratory by distilling it from a -25~C trap into a -196~C trap. The pure (CH3)3P exhibited a vapor pressure of 15.9 cm at O~Co Rosenbaum and Sandberg (28) also report a value of 15.9 cm for the vapor pressure of (CH3)3P at O~Co

18 Trimethylphosphine-Borane, (CH3)3PBH3, (CH3)3PB1OH3 and (CH3)PB11D3.. The preparation of (CH3)3PBH3 has been, reported in the literatureby Burg and Wagner (34) and also by Hewitt and Holliday (35). The method involves the direct reaction of (CH3)3P and B2H6 with the formation of a volatile, white, crystalline solid. Burg and Wagner (34) have characterized the solid and have reported the-following information: melting point is 103~-103.5~C, vapor pressure at 45.7~C is 2.2 mm, and log10 Pmm = 9.531-2933/T. A sample of (CH3)3PBH3 containing the natural isotopes of boron and hydrogen was generously donated by Professor Sheldon Shore, Before use in this laboratory, the sample was freshly sublimed and stored in an evacuated tube. A sample of (CH3)3PB10H3 containing the boron-10 enriched isotope was prepared by condensing about 1 ml of 10 (CH3)3P into a reaction tube with excess B H6. Upon warming the tube to room temperature a white crystalline solid, (CH3)3PB10H3, formed. The reaction tube was placed in a O~C bath and the volatile components were distilled away. The product was purified by sublimation. A sample of (CH3)3PB D3, enriched in boron-1l and deuterium, was prepared and purified in the same fashion. Miscellaneous Procedures Some discussion of general experimental procedures appears appropriate. As was mentioned earlier, vacuum line

19 techniques were employed for the preparation and handling of the materials under investigation. It was found that some high vacuum greases were more compatible with these materials than were others, Kel-F Grease was found to be particularly suitable in the presence of BF3 and the (CH3)2NPF2BF3 complex, Dow Corning High Vacuum Silicone Grease and Apiezon N Grease were employed to a large extent in the handling of the borane complexes. (CH3)2NPF2 and [(CH3)2N]2PF seemed to be reactive to the Dow Corning Silicone Grease so Apiezon N Grease was employed while handling these samples, The vacuum line was cleaned periodically with alcoholicKOH or K2Cr207- H2S04 depending on whether Silicone or Apiezon Ni Grease was used. In every case, after the line was clean the glass was treated with dilute HC1 or HFo This was done because the samples under investigation are sensitive to the presence of bases, such as traces of the alcoholic-KOH cleaning solutionO It was also noted that mercury was readily absorbed by the samples, rendering them somewhat opaque. Care was taken to limit the exposure of the samples to mercury only for vapor pressure measurements. In every case, the samples were stored in evacuated tubes which were pre-treated with acid or made of new pyrex, Also all samples were stored as solids at dry ice or liquid nitrogen temperatures to retard decomposition

20 until they were ready for useo Special precautions were taken to minimize hydrogen contamination of the deuterated samples during the preparation and handling of the samples. The vacuum system and all attachments were cleaned prior to use and then allowed to equilibrate with heavy water so as to reduce hydrogen exchange with water absorbed in the glass surface. Spectroscopic Methods and Equipment The infrared spectra of the compounds studied were obtained on a Perkin-Elmer Model 21 Spectrophotometer equipped with CaF2, NaCl or KBr prisms. The spectra were 2.. calibrated in the appropriate regions from 4000-400 cm1 using indene, H20 vapor, NH3, HC1, and HBr. Gaseous samples were examined in a ten cm cell having KBr windows and liquids were studied as a liquid film between KBr plates. The infrared spectra of solid samples were observed in transmission through a thin film by employing a modified version of the cold cell described by Wagner and Hornig (37). This cold cell was fitted with a bulb in which the sample could be kept at liquid nitrogen temperature until use0 To prepare the film, the compound was allowed to evaporate at a low pressure and the vapor was sprayed directly on a KBr window kept in thermal contact with a liquid nitrogen reservoiro Nonvolatile liquid samples were examined in their solid states by

21 placing a drop between two KBr plates and then inserting the assembly in the low temperature cello The low temperature infrared cell is shown schematically in Figure lo In essence, the technique of Raman spectroscopy involves irradiating the sample with monochromatic light and observing the spectrum of scattered light. Since the Raman scattering phenomenon is intrinsically inefficient, an intense light source and high aperture spectrographic equipment are desirableo Special precautions must be taken to eliminate light scattered from sources other than the sample and to avoid fluorescence and Tyndall scattering from within the sampleo For the Raman spectroscopic study, samples were distilled into pyrex Raman tubes having an rIDo of 2 to 4 mm and an OoDo of 6 mm and then sealed off under vacuumo These tubes were fitted with optically clear flat ends so that with liquids and some solids, the Raman scattered light could be observed at right angles to the incident radiation produced by the Toronto Arco These Raman tubes were used for a spectroscopic study of both liquid and solid sampleso The light source used for obtaining the spectra of liquids was a helical, mercury arc having large, water cooled mercury pools as electrodes as described by the workers at the University of Toronto (38.39)o Solutions of ethyl violet in alcohol and KN02 in water were used to

22 filter the mercury radiation so that mercury lines other than the 4358 A were greatly reduced in intensity, In obtaining a spectrum, the sample tube was placed in a specially constructed dewar flask in the center of the arc. The temperature of the sample was controlled by passing a stream of cold air in at the bottom of the dewar and up along the sample. The temperature of the air stream was monitored by a thermocouple and recording potentiometer and could be adjusted if desired, by a small heater arrangement to within + 5~C of the desired temperature. The experimental setup for the Toronto arc, filters, and temperature control is given in detail elsewhere (40); however, the main features are shown in Figure 2. Polarization characteristics of liquid samples were studied by taking two equal exposures, during which times the incident radiation was polarized respectively parallel and perpendicular to the axis of observation by using appropriate polaroid filters. The experimental arrangement used to obtain Raman spectra of solid samples is shown in Figure 3. Mercury light'from two General Electric AH-4 lamps was passed through monochromator systems which contained condensing lenses, absorption filters and two interference filters which were adjusted to allow the maximum transmission of 0 4358 A light. With this arrangement about 33% to 50% of the 4358 A light was transmitted'while reducing the intensity of the other mercury lines and the continuous background to

23...........:-iiii..................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... IFIGUI 1. - TIlIE LOW TEMPEiATUiRE INFI{ARIED CELI,

24-J FI:.GUR-:iE 2 THE E41XPR:IMENTAL. AlRRANGENMEiNT FOR 0BTAIA:INING RAMAN SPECTRA OF LIQUIDS FIGURE 3... TE EXPE.RIMENTAL ARRANGEMENT FOR OBTAINING RAMAN SPECTRA OF SOIJDS AT LOW TEMPE.RATURES

25 negligible values In a few cases the Raman spectra of solid samples were taken with the Toronto arc serving as a light source and these spectra exhibited a higher general background, although below 400 cm-1 the background was lower. Since the solid samples were polycrystalline masses, it was not possible to determine the polarization characteristics of the observed frequencieso When examining the spectra of solid samples a great deal of the incident mercury radiation is reflected from the crystal faces and therefore is scattered towards the spectrograph. This scattered light usually obscures the lower frequency Raman lines. This effect was reduced by allowing the scattered radiation to be reflected four times from two nearly parallel interference filters, The characteristics and the arrangement of these filters has been described by Dahl (41). By correctly adjusting the angles of the two filters with respect to each other and the incident beam, one is able to properly "tune" the interference filters to allow the passage of a maximum of scattered Raman radiation into the spectrograph and a minimum of scattered mercury lighto With this arrangement it was estimated that about 65% of the Raman light above 200 cm 1 was transmitted while only about 0~4% of the mercury exciting 4358 A light was transmitted (4l)o In some instances a Corning yellow filter, Noo 3387, was placed in front of the slit to reduce further the mercury 4358 A light.

26 The previously mentioned multireflection filter system also has been used with success in connection with liquidso The use of this arrangement allowed a Raman spectrum to be taken of slightly turbid liquid samples. In the present work, this proved quite helpful because a slight sediment appeared in many of the samples during exposure. The low temperature cell used to obtain the Raman spectra of solid samples has also been described by Dahl (41)o In this cell, the sample tube was placed in an evacuated chamber and brought into thermal contact with a low temperature reservoir by means of a copper rod. By utilizing this cell samples could be maintained at temperatures as low as -180~C during an exposureo The spectrograph used for this investigation was a Gaertner two prism instrument having a dispersion of 180 cm- per millimeter in the blue region and a camera aperture of f:3o50 The Raman spectra were recorded photographically on Eastman Kodak IIa-0 plates with antihalation backingo These plates were sensitized by heating them in a vacuum desiccator at 55~C for 12 to 14 hours prior to use, The plates were developed for four minutes in Eastman Kodak D-19 developer at 20~C and were fixed with F-6 solution for ten minuteso Exposure times ranged from 1/4 to 10 hours for liquid samples and 1 to 72 hours for solid samples

27 The position of the Raman lines were measured directly on the plates using a Mann comparator and also on tracings of the plates. The latter were made with a Leeds and Northrup, Knorr-Albers recording microphotometer and also by a Joyce-Loebl microphotometerO The respective tracings represented approximately 25 and 20 fold enlargements of the plates. The Raman frequencies were calculated from the plate measurements by a computer program based on a quadratic interpolation of the dispersion curve for the spectrometero Mercury lines at 4339.24 and 4916.04 A were used as references for the Raman spectra of liquid sampleso For spectra of solid samples argon reference lines were supplied by placing an argon spectrum above and below the Raman spectrum.

EXPERIMENTAL RESULTS AND INTERPRETATION General Discussion of Spectra Since the present study deals with an infrared and Raman investigation of samples in the gaseouss liquid and solid states, a brief discussion of -the differences in spectra with physical state. is appropriate at this point. The change of state of a sample often results in a significant modification in the vibrational frequencies of the individual molecules or the intensities of the vibrational bandso The extent to which these modifications occur depends on intermolecular separation, the dielectric constant of the surrounding medium, the degree of molecular interaction and other less important factorso The vibrational spectra of gaseous samples frequently show resolved rotational structure. This rotational structure is normally "washed out" in the spectra of liquids but the vibrational bands tend to be somewhat broadened, perhaps due to increased molecular interactionso However, the vibrational spectra of solids, especially at low temperatures, tend to be much more complex. This increased complexity can be attributed to one or more of the following factorso splittings of degenerate bands as a result of decreased molecular symmetry, multiplets arising from the coupling of the vibrations of one molecule with others in the same unit cell, combinations of the molecular funda 28

29 mentals with the lattice modes or resolution of otherwise accidentally degenerate fundamentals because of the greater sharpness of bands. The last named cause may be important in the present case due to the presence of different isotopic species which otherwise would contribute to the intensity of a single unresolved band, Other changes in intensity, both in fundamentals as well as the combinations or overtones frequently occur between liquid or gas and the solid state and may contribute to the problem of interpretationo Differences between the infrared and Raman spectrum of a compound are frequently observed. Molecular vibrations which absorb infrared radiation are often active in the Raman effect, However, the intensities of the vibrational bands may be markedly different in each case, This phenomenon is brought about by the fact that a particular vibration may not involve the dipole moment to the same extent as it does the polarizibility, thereby making the transition moment different in each caseo Minor differences in the vibrational frequencies are also observed. Such variations may occur as a result of a difference in the physical state of the sample, experimental error in the measurements, crystal symmetry effects and intermolecular coupling, differences in resolving power of the spectrographic equipment, or the noncoincidence of the band maximum with the actual vibrational frequency,

30 Polarization Results An important and useful feature in the analysis of Raman spectra is the study of the polarization characteristics of the scattered lighto Theory predicts that if the incident light is polarized. with its electric vector parallel to the direction of observation, then the intensities of the Raman bands are less than if the incident light is perpendicularly polarized. The ratio of the parallel to perpendicular intensities can be shown theoretically to depend on the nature of the vibrationo For vibrations which preserve the symmetry of the molecule the depolarization ratio may have any value between zero to 6/7 while for asymmetric vibrations the ratio is 6/7. This provides an important way of distinguishing between the totally symmetric and asymmetric vibrations. Unfortunately it is difficult experimentally to make accurate depolarization *ratio measurements because of numerous instrumental and other factors which must be taken into account. If photographic plates are used, one has the additional handicap of a detector with a nonlinear response In the present work, this factor made it difficult to identify unequivocally polarized lines whose depolarization ratio was close to 6/7, whose intensity was quite low or which were superimposed on a high intensity background caused by fluorescence or Tyndall scatteringo

31 Spectroscopic Results and Assignments Throughout the following section, vibrational frequencies and their assignments for a number of compounds will be discussed in detail, For clarity, it is desirable to define several terms which will be used in the following discussion, The terms symmetric and asymmetric most generally are employed in describing whether or not a particular fundamental vibration preserves or destroys some or all symmetry elements of the molecule. This strict definition will be used whenever possible; however, for molecules which have two or more chemically equivalent groups or which possess no molecular symmetry, further clarification is necessaryo For compounds such as (CH3)3P and (CH3)2NPF2 which contain two or more chemically equivalent methyl groups, chemical knowledge tells us that the methyl group vibrations are influenced to a larger degree by local interactions than by the coupling of the motions of one methyl group with those of anothero For this reason it appears more meaningful to define the methyl vibrations as being symmetric or asymmetric motions within the methyl group, ioe, with respect to local symmetry, and then compare the motions of one methyl group with another by using the terms "in phase" or "out-of-phase"o This treatment is particularly useful in discussing the motions of molecules which have no symmetry, such as CH3NHPF2BH3, for in this case the use of symmetrical or asymmetrical

32 vibrations can refer only to the local symmetry of the group vibration. In addition, the numerous tabulated frequencies, band intensities and vibrational assignments necessitate an abbreviated notation in the tableso Therefore the following symbolism will be employed in the tables when referring to band intensities, structure and polarization characteristicsv = very, s = strong, m = medium, w = weak, sh = shoulder, br = broad, p = polarized and dp = depolarizedo Also, the symmetry and the type of vibration will be represented asO s = symmetric, a = asymmetric, v = stretch, 5 = deformation, p = rock, C = wag and T = tors ion, Trifluorophosphine, PF In order to carry out a vibrational analysis of the compounds of principal interest in this work, it was desirable to record the vibrational spectrum of PF3 and confirm previous assignments. Although PF3 is a stable compound, the information concerning its vibrational spectrum is somewhat limited and the assignments of the fundamental modes of vibration have been subject to some uncertainty. It is known that the PF3 molecule has a regular triangular pyramidal configuration belonging to the C3v point group (42,43,44). This model has four fundamental modes of vibration, all active in the infrared and Raman effect, Two of these fundamentals belong to the A1 class

33 and the other two belong to the doubly generate E classo The four vibrational modes may be described as vl and vw the symmetric P-F stretching and deformation modes, and w3 and v4,the doubly degenerate asymmetric P-F stretching and bending vibrations respectivelyo Yost and Anderson (14), the first to examine PF3 spectroscopically, reported the vibrational frequencies given in Table lo They assigned vl and v3 to 890 cm- and 840 cm-1 and v2 and w4 to 531 cm'1 and 486 cm1l respectively. Later, Gutowsky and Liehr (15) reported essentially the same results but reversed the assignments of v2 and v4 on the basis of band contour considerations0 Wilson and Polo (16) observing the gaseous infrared spectrum found similar results for vl, v2, 3 but assigned the doubly degenerate bending vibration, v4, to a new band at 344 cm1 -1 and found no evidence of a band at 531 cm o A more recent Raman investigation by Taylor (17) agreed with the findings of Wilson and Polo, Although polarization data were not obtained, the contours of bands in the spectrum of the gas supported their assignment. The previously reported band at 531 cm-1 possibly is due to an impurityo Due to the conflicting earlier work, additional spectroscopic work was carried out to confirm the correct assignments of w2 and W4o In particular, the polarization properties of the bands were determined for liquid PF3o The results clearly showed that the bands at 874 and 484 cm-1

TABLE 1 THE OBSERVED RAMAN AND INFRARED FREQUENCIES OF PF3 (14) (14) Gas Liquid (17) Gas (20~C ) RAMAN (17)0 )a Liquid (-900C)a INFRARED Ass ignment Solid (-180~C) (15)(16) Gas Gas Gas (4 mm) 893 m 890 s 891 vs 874 s,p 858 vw,sh 920 w,sh 891 892 891 ssh R branch v1 867 s,sh 851 m 840 s 846 ssh 832 vs,dp 809 w 862 vs 844 860 851 s,sh V3 833 833 m,sh 693 -=r V2 + w4 351 w 532 499 vw R branch 487 ms 486 w 486 m 484 m,p 479 vw 486 487 485 w w2 351 w,dp 344 a s = strong m = medium w = weak v = very sh = shoulder br p dp = broad = polarized = depolarized

35 were polarized while the bands at 832 and 351 cm 1 were depolarizedO The assignments of w1 to 874, v2 to 484, v3 to 832 and w4 to 351 cm-1 are in agreement with the assignments and frequencies reported by Wilson and Polo. In addition, the Raman spectrum of solid PF3 at -180~C and also the infrared spectrum of gaseous PF3 at 20~C and 4 mm pressure were studied, As a consequence of this study, several features of the infrared and Raman spectra were observed that are worth mentioningo First of all, the band at 513 cm 1 as reported by Yost and Anderson and also by Gutowsky and Liehr was not observed which substantiates Wilson and Polo's findings. Secondly, in the infrared -1 spectrum of gaseous PF3, the bands at 902 and 499 cm are proposed to be the R branches of v1 and w2 respectivelyo The very strong band at 862 cm1 and the strong shoulder at 851 cm 1 remain to be assigned. The question ariseso which of these two frequencies should be assigned to v 3 It is evident that neither of these frequencies can be attributed to an overtone or combination bandnor is Fermi resonance likely in this case, Two possible explanations can be given. Either the higher band at 862 cm"1 can be attributed to the P branch of w1 and the band at 851 cmassigned to w3, or these two bands are unresolved P and R branches of V3, the Q branch being absent. A study of band shapes and intensities for both the gaseous infrared and Raman spectra (17) indicates that perhaps the latter

36 case is more likely correct. A simple doublet band outline for a degenerate Raman band has been observed in at least two other cases (39,45)o Trifluorophosphine-Borane, F3PBH and F PB H 0o Electron diffraction studies of trifluorophosphine (43,44) and phosphoryl fluoride (46) have shown that these molecules are of C3v symmetry having the fluorine atoms arranged at the corners of an equilateral triangleo In view of the known structure of OCBH3 (47), the trifluorophosphineborane complex would be expected to have an ethane-like configuration with C3 symmetry, the threefold axis coinciding with the phosphorus-boron bondo For this configuration, there would be twelve vibrational frequencies with the symmetry distribution 5A1 + A2 + 6Eo Four of these are localized in the borane group, four in the trifluorophosphine group and four motions arise from the formation of the phosphorus-boron bondo A description of the fundamental vibrations and the symmetry species to which each belongs is given in Table 2o All of the vibrational modes are active both in the infrared and Raman effect except for the inactive (A2) torsional mode, The infrared and Raman spectra of solid F3PBH3 and F3PB1H3 at -180~C are shown in Figures 4, 5 and 60 The observed vibrational frequencies and their assignments for the Raman and infrared spectra of solid F3PBH3 and F3PB10H3 are given in Table 3o The Raman data reported by Taylor

37 I I I I I I I I I 4000 3000 2000 1500 1500 1000 500 CM-' FIGURE 4 - THE INFRARED SPECTRUM OF SOLID F3PB1H3 (-80~)

38 I l I I I I I I I I I I 4000 1500 3000 2000 1500 FIGURE 5 - THE INFRARED SPECTRUM OF SOLID F3PBH3 (-180~C) FIGURE 5 - THE INFRARED SPECTRUM OF SOLID F3PBB3~~~~~~~~~~~~~~~~~~~~~~~

39 PF3BH3 (SOLID) PF B10H 3 (SOLID) 0 1000 2000 3000 CM-' FIGURE 6 - THE RAMAN SPECTRA OF SOLID F3PBH3 AND F3PB1H3 (-180~C) 3 3

40 TABLE 2 THE FUNDAMENTAL VIBRATIONS OF F3PBH3, C3v SYMMETRY Symmetry A1 Number 1 2 3 4 5 A2 6 Description B~H stretch BH3 deformation P-F stretch P-B stretch PF3 deformation P-B torsion B-H stretch BH3 deformation P-F stretch BH rock 3 PF3 deformation PF3 rock E 7 8 9 10 11 12

TABLE 3 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF F3PBH3 AND F3PB10H3 F3^PBH^3~ F^3~PB10 FPBH F PB H 3.3 3 3 Raman (5) liquid (-800C) Raman solid (-180~C) Infrared solid (-180~C) Raman solid (-180oC) Infrared solid (-1800C) Assignment 2655+5 vvw 2530 vvw 2655 vw 2643+2 (2459 2450 w vs,sh) vs 2654 vw v7 + v12 diborane?.V7 (2469 m,sh) 2455 vs,dp 2460 s 2471 ms 2456 2394 vs m 2385 vs,p 2328+4 vvw 2247+2 vw 2392 vs 2336 w,sh 2239 vvw 2391 ms,sh 2397 s 2340 w,sh,br 2337 vw, sh 2231 w 2243 vvw 2341 w,sh 2239 w 2212 w v1 2v3 + w5 2v8 w1 - 12? v1 - v12o 2w4 + 73 2v2 diborane? 4= 1p 2197 vw 2173 w 2137 vw 2140+4 vw 2112 vvw 2146 vvw 2113 vvw 2146+5 vw 1818 vvw 1831 vw,br? V8 + o10 1703 vwbr,:h 1704 m w2 + w4 (B-10) 1681 w 1676 vw v2 + v4 (B-1) )9 + v10 1668 w, sh?

TABLE 3 - CONT'Do F3PBH3 Raman (5) Raman Infrared liquid (-80oC) solid ( 1800C) solid (-1800C) 1648 vvw 10 F3PB H3 Raman Infrared solid (-180~C) solid (-180~C) 1650 vw 1636 vvw 1560 vvw 1542 vvw 1523 1447 1417 1403 1396 VVw vvw vvw, sh vw, sh? 1460 1418 vvw w w 1391 vvw? Assignment V3 + V10 v3 + v10i V4 + vg 3 + 4 ~2 + V5 v5 + v9" 2v10 or "5 + w9 2v10 or 5 + w9 vg + ll 9 V11 w3 + vli V8 + V12 or V4 + v10 2v4 (B-10) 2V4 (B-ll) w3 + w12 V8 V2 (B-10) [=) PO 1328 1314 vvw vw 1335 1308 vv'w? vw 1227 1208 vw w 1227 1208 1127 1123 1093 m vvw s, sh? w 1117 s, dp 1124 ms 1121 s 1127 1092 m vvw

TABLE 3- CONT'D. Raman (5) liquid (~80oC) 1077 wp F3PBH3 Raman solid (-180~C) 1079 w F3PB10H 3 3 1040+3 957+3 944 920 886+5 w,p m, dp m,p? w,p vw, p 1030 969 939 921 vw, sh m m m Infrared solid (-1800C) 1079 w,sh 1068 w 1032 vvw 957+7 s,sh? 942 vs 919 s 892+5 w,sh? Raman solid (-1800C) Infrared solid (-1800C) 1078 vw 1051 vw 972 944 925 w w w 988+5 943 916 902 m, sh? Vs s s,sh? 830+5 vwvbr, sh Assignment V2 (B-ll) 10 + v11 4 + V5 9 v3 or w97 W3~ 2v5 or 5 v10 + l12 v5 + "l 2v11 l10 (B-10) V10 (B-ll) Impurity 74 (B-10) 4 (B-ll) 11 12 4= uj 799 vw 736 709 700 vw w, sh ms 697+2 607 441 370 197 vw s,p m,p vw m,dp 701 608 438 vw m w 736 709 702 662 613 606 439 w s w, sh vvw s s,sh S 616 606 -436 s,sh s s 5 5 619 m 449 vw

and Bissot (5) for liquid F3PBH3 are also tabulated for purposes of comparison~ The agreement between the frequencies reported here and those found previously by Taylor and Bissot (5) is quite satisfactory and most of the small differences which are observed can be attributed to experimental error or to a change of state of the sampleo The symmetric and asymmetric B-H stretching vibrations appear as a very characteristic and intense doublet in the region between 2350 and 2500 cm* The asymmetric B-H stretching vibration, the higher band in the doublet, shows a single strong peak in the Raman spectrum, but in the solid infrared spectrum using CaF2 optics a strong shoulder is observed on the high-frequency side of the fundamentalo The shoulder and the fundamental are observed at 2459 cm l and 2450 cm~l respectively for F3PBH3, and 2469 cm'l and 2456 cm 1 respectively for F3PB10H3 Since the shoulder cannot be identified with an isotope effect or a combination band, it probably arises from crystal splitting effectSo The symmetric band occurring at 2392 cml1 has a high intensity in the Raman effect but appears only as a medium intensity shoulder in the infraredo The asymmetric and symmetric deformation modes of the borane group are found at 1124 cm'l and 1079 cm1 respectively for the compound containing the natural boron isotopeo The asymmetric deformation mode appears as a strong band in both the infrared and Raman spectra, but the symmetric

45 vibration is much weaker, A weak band at 1092 cm 1 is also observed in the spectrum of the B-10 compound and is assigned to the symmetric B-10 borane deformation mode, The frequencies associated with the PF3 group are similar to those of free PF3 (16,17) and F3PO (15,20),and a comparison of these P-F vibrations is shown in Table 4, In the region from 1000 cm1 to 900 cm1 three distinct bands are observed in the Raman spectrum- a medium intensity band at 969 cm1 and two bands of lesser intensity at 939 cm-1 and 921 cm o In the infrared spectrahowever, two very strong bands are observed at 942 cm 1 and 919 cm 1 each having a strong shoulder around 957 cm 1 and 892 cm 1 respectively, The shoulder at 892 cm, appearing only in the infrared spectrum, is probably due to the overtone 2v5 or the combination v0 + W12~ The two higher frequencies, which are observed at 969 cm 1 and 939 cm 1 in the solid Raman spectrum and 957 cm"1 and 942 cm-l in the solid infrared spectrumalso appear in the Raman spectrum of the liquid and are assigned to the asymmetric and symmetric P-F stretching motions respectively (5), The lower band at 921 cm is also observed in the liquid spectrum and is attributed to a difference band, V8 - 12- However, in view of the fact that this band is very intense in the solid infrared spectra taken at -180~C, this assignment does not appear tenableo It must therefore be either a fundamental or a combination bando It is possible that the bands at

46 TABLE 4 A COMPARISON OF THE FUNDAMENTAL FREQUENCIES OF PFOPF3, OCBH3 and F3PBH3 (CM-1) PF3(16) Ni(PF )4(19) F3PO(15) OCBH (21) F3PBH3(5) 2434 2455 2380 2385 2169 1415 1101 1073 1117 1077 957 944 697 39 Ni(PF3)41 Assignment asymo B-H stretch symo B-H stretch (symo C-0 stretch) (symo P-0 stretch) asymo BH3 symo BH3 deformation asymo — P-F stretch symo P-F stretch asymo BH3 rock (symo B-C stretch) symO P-B stretch symo PF deformation asym, PF deformation (asymo BCO deformation) asymo PF3 rock 860 892 859 898 990 873 816 692 607 441 487 344 386 473 485 503 370 317 345 197

47 921 cm1 and 939 cm are the P-F stretching fundamentals while the higher 969 cm1 band is the combination. In this region only two possible combinations appear reasonable~ w4 + 11 at about 977 cm and 2vll + v12 at about 937 cm The intensities of these proposed bands probably would be relatively weak unless resonance occurred with one of the P-F stretching fundamentals, If this were true, the band intensity might be appreciable and the band maximum might be shifted from the expected position. Although the intensities of the observed bands at 969 cm 1 and 939 cmil might be explained in terms of Fermi resonance, little or no apparent shifting is observed from the calculated frequencies of the combinations and no strong preference for either choice is indicated Another possibility is that the observed band at 921 cm may be due to an overtone or combination of the inactive torsional mode with an E fundamental, Since the frequency of the torsional mode is not known, one cannot predict where such a band might be observed, None of the other bands observed in the spectra, however, require the assumption of a torsional frequency, At the present time, Taylor s assignments for the asymmetric and symmetric P-F stretching modes are considered to be correct,but the band at 921 cm1 remains to be assigned, A normal coordinate treatment might help in resolving the uncertainties connected with these assignments,

48 Since the PF3 deformation modes occur at much lower frequencies, only the symmetric PF3 deformation vibration at 438 cm was observed in infrared absorption, the asymmetric deformation frequency at 370 cm 1 (5) being too low to be recorded experimentally, The broadness Of the exci ting line also prevented it from being observed in the Raman' effect. The four fundamentals which remain are the BH3 and PF3 rocks, the P-B stretch, and the inactive P-B torsional mode, Of these, only the BH3 rocking and P-B stretching vibrations are observed. The BH3 rocking vibration is easily identified in the infrared spectrum and shows a typical boron-10 isotope effect. The observed frequencies for the boron-10 and 11 enriched samples are 709 cm 1 and 700 cm 1 respectively. Similarly, the P-B stretching vibration is shifted from 616 cm-1 to 606 cm-1 in the appropriately enriched samples, The weaker bands shown in Figures 4, 5 and 6 are due to overtones and combinations, and the assignments for many of these are given in Table 3, Due to the fact that these assignments are somewhat arbitrary, they are not as certain as those for the fundamentals, It should be mentioned here that many of the infrared tracings of solid F3PBH3 and F3PB 103 exhibit bands which do not appear in the Figures 4 or 5. These bands are attributed to volatile impurities, perhaps diborane, since their relative intensity depended on the treatment of the

49 sampleo If the sample of F3PBH3 is fractionated immediately before use and the less volatile fraction is used for examination, no evidence of these bands is observed in the infrared spectra, For purposes of record the following impurity bands are observed for solid F3PBH3 a strong band at 2340 cm-1 with a shoulder at 2358 cm 1, a very weak band at 2261 cm-1 and a strong band at 663 cm 1 with a shoulder at 656 cm-1 Similar bands are observed in the infrared spectrum of solid F3PB10H3 The band at 2340 cm-1 also appears in the Raman spectra and Taylor (5) has attributed this band to the combination 2v3 + V5, but in the light of the present study it probably should be ascribed to an impurity. Methylaminodifluorophosphine-Borane, CH3NHPF2BH30 Virtually no experimental data are available which provide information about the structure of this compoundo However, nuclear magnetic resonance studies indicate that the borane group bonds through the phosphorus atom in the similar compounds, (CH3)2NPF2BH3 and [(CH3)2N]2PFBH3 (13)o Since these dimethylamine complexes are very similar chemically to the present methylamine complex, one might similarly expect P-B bonding in this compound. In addition, chemical arguments support the hypothesis that the pyramidal configurations of CH3NH2 and PF3 are retained in the CH3NHPF2BH3 complex. The application of these conditions leads to-a structure which has,no molecular symmetry or

50 at best a "pseudo" Cs symmetry in which the carbon, nitrogen, phosphorus and boron atoms lie on a vertical reflection plane, Because the former configuration is more probable, the motions of the molecule will be discussed in terms of group or site symmetry rather than in terms of possible molecular symmetry, Since this molecule contains thirteen atoms, there are thirty-three Raman and infrared active vibrational modes. Of these vibrations, nine are confined to the vibrations of the methyl group, nine involve predominantly the motions of the borane group, three are essentially N-H vibrations and the remaining twelve vibrations are primarily skeletal motions of the molecule, regarding the CH3, BH3 and NH groups as point masses. Although this vibrational classification is employed mostly for the sake of convenience, nevertheless such a classification is validated by the experimental observation that the hydrogen motions are well separated from the skeletal vibrations and probably are not coupled to the skeletal modes to any appreciable extent, Both the infrared and Raman spectra of CH3NHPF2BH3 for various sample states have been observed and are represented in Figures 7 through 10o The Raman polarization spectra of the liquid sample at -40~C have also been included in the List of Figures (see Figure 10) for assistance in making assignments. The large number of mercury lines which are observable in the Raman spectrum of the liquid sample (see

51 Figures 9 and 10) are a result of inadequate filtering of the incident mercury light. The observed infrared and Raman frequencies and their tentative assignments are listed in Table 5o The very strong band appearing in the region from 3350 cm1 to 3454 cm 1 in both the infrared and Raman spectra is obviously the N-H stretching vibration and correlates very well with the similar vibration at 3355 cm1 for (CH3)2NH (22). The frequency of this band shifts from 3454 cm1, to 3369 cm1, to 3362 cm 1 in the spectra of the respective gaseous, liquid and solid samplesO Such shifting probably is indicative of appreciable hydrogen bonding in the condensed statesO In the C-H stretching region, two polarized bands at 2845 cm-1 and 2957 cm"1 and one depolarized band at 3004 -1 cm are clearly resolved in the Raman effect but are not completely resolved in the infrared spectrum. The assignment of the lower band at 2845 cm"1 to the symmetric C-H stretch and the two higher bands to asymmetric C-H stretching vibrations is consistent with the observations of CH3NH2 in which the corresponding bands are found near 2820 cm-1 2961 cm-1 and 2985 cm-1 (48,49)o Other bands which are observed in this region of the spectrum are probably overtones or combinations of lower frequency fundamentals probably in resonance with themselves or with the C-H valency fundamentalsO

52 4000 3000 2000 1 -, I I I I I I I I I 1500 1000 FIGURE 7 - THE INFRARED SPECTRUM OF GASEOUS 500 500 CM-I CH3NHPF2BH3 ( 6 mm)

53 4000 3000 2000 1500 1500 1000 FIGURE 8 - THE INFRARED SPECTRUM OF SOLID CH3NHPF2BH3 (-180 C) 500 CM-'

54 0 1000 2000 3000 CM-I FIGURE 9 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (-150C) CH3NHPF2BH3

55 *. E(1) *r * e E(ll) * H9 0 1000 2000 3000 CMA FIGURE 10 - THE POLARIZED RAMAN SPECTRA OF LIQUID CH3NHPF2BH3 (-40~C)

TABLE 5 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF CH3NHPF BH3 3 2 3 Inf ra: gas (6mm), 3454+10 w 3028+10 vw,sh? 2961+10 w 2929+10 w sh 2827+10 vw, sh 2438+10 ms 2365+10 w,sh 2233+10 vw red solid (-180"C) 3353 vs 2997 wsh Raman liquid (-15~C) solid (-1800C) 3369 vsp 3362 s 3004 s,sh,dp 3015 s,sh,br 2957 vs,p 2956 vs 2938 w 2839 vw,sh 2794 vw,sh? 2413 s 2370 msh 2911 wsh? 2845 s,p 2425 vs,p 2369 vw,p 2253 w,sh? 2848 ms 243 2 vs 2407 vs,sh 2369 vs 2312 m,sh 2264 w,sh Ass ignment N-H stretch asym. C-H stretch asym. C-H stretch? 1454 + 1479 = 2933 2 x 1458 = 2916 sym. C-H stretch 2 x 1405 = 2810 asym. B-H stretch 917 + 1489 = 2406? sym. B-H stretch 1135 + 1181 = 2316 2 x 1125 = 2250? 2 x 1125 = 2250? 705 + 1455 = 2160 2 x 1077 = 2154 646 + 1479 = 2125 2 x 773 = 1546 369 + 1125 = 1494 2242 wsh 2186 vvw,sh? 2162 vw 2143 vw 2115+5 vvw,sh? 2114 vvw? 1536 w,sh 1545 vw,sh? 1494 w,sh?

TABLE 5 - CONTID. gas (6mm) 1400 m 1323 w,sh 1186 w,sh 1120 ms 1059 m,sh 1008 m,sh 908 ssh 883 vs,b 845 s,sh 770 m,sh Infrared solid (-180~C) 1479 ms 1454 m,sh 1439 m,sh 1405 vs 1335 w 1185 w,sh 1125 s 1076 s,sh? 1066 s 1018 m L? 915 vs r 867 vs 849 vs,sh 773 w,sh Raman liquid (-15~C) solid (-180~C) 1483 msh,dp 1489 ms,sh 1458 m,sh,dp 1455 ms 1438 m 1393 w,sh 1410 w,sh 1182 vvw,sh?? 1131 s,p? 1064 m,p 1020 vvw?? 914 w,dp 877 m,sh.,dp? 843 s,p 796 vvwsh? 708 vvw? 1181 vw,sh'?? 1135 vw 1077 w 917 w 851 vs 705 vvw? Assignment asymo CH3 deformation asymo CH3 deformation asymo CH3 deformation? sym. CH3 deformation 465 + 867 = 1332 CH3 rock asym. BH3 deformation and/or CH rock? sym. BH3 deformation and/or C -N-P stretch C-N-P stretch BH3 rock? asym. P-F stretch symo P-F stretch NH rock and/or N-P-F2 stretch NH rock and/or N —P-F2 stretch 211 + 465 = 676 kJi1 -KI 668 vwsh

TABLE 5 - CON1T'D. Infra red Ra man gas (6mm).....solid (-180~C) liquid (-i5~C) solid (-180~C) 652 m,sh 636 m 646 s 640 s, 647 s 584- vw,br 465 m 401 m 596 w,p 467 sp 400 w 369 vvw? 255 w,sh,p? 211 w, sh,p? 169 vwi, sh 591 vw? 4TF 5 vw As signr- en t P-B stretch (B-10)? P-B stretch (B-ll) skreleta 1 PF2 deformation skeletal C-N,-P deformation? CH- torsion PF, rock? skele ta 1 i 1 CO

59 The very strong doublet which is observed in both the Raman and infrared spectra around 2369 cm-1 and 2425 cm1 is most assuredly the symmetric and asymmetric B-H stretching vibrations in view of a similar doublet at 2385 cm1 and 2455 cm1 in the Raman spectrum of F3PBH3 (5) Since the lower band is strongly polarized while the upper is only weakly polarized, the lower band is assigned to the symmetric B-H stretch and the upper to the asymmetric vibration. The influence of the rest of the molecule apparently is not sufficient to cause an observable splitting of the two asymmetric modes which form a degenerate pair in the case of full Cv symmetry, The strong shoulder appearing between the two lines of the doublet in the solid Raman spectrum is probably an overtone or combination, perhaps 917 + 1489 = 2406. In the methyl deformation region from 1400 cm1 to 1500 cm"1 several clearly resolved bands are observed in the solid infrared spectrum but are not well resolved in the Raman effect. The very strong infrared band at 1405 cm1 and the weak Raman shoulder at 1393 cm"1 are probably the totally symmetric methyl deformation vibration while the asymmetric methyl deformations are most likely found as one or more of the bands in the triplet at 1438 cm 1 1458 cm1 and 1483 cml The tentative assignments of the asymmetric CH3 deformations to the two higher bands at 1458 cm and 1483 cm are preferred because these

60 correlate better with the respective assignments in methylamine at 1459 cm 1 and 1476 cm-1 (49). The great intensity change of the symmetrical deformation vibration between the infrared and Raman spectra may be somewhat surprising but Barcelo and Bellanato (22) report similar observations for the symmetric methyl deformation of CH3NH2 at 1410 cm lo Because the methyl rocking vibrations are found in the same region of the spectrum as the asymmetric borane deformations, their assignments are not certain and deuteration of the borane group will be necessary to make unambiguous assignments. Since methyl rocks are found at 1130 cm and 1195 cm1 in the infrared spectrum of CH3NH2 (48), one or both of the bands at 1131 cm-l and 1182 cm 1 in the spectrum of CH3NHPF BH3 may be methyl rocking vibrationso On the other hand, the 1131 cm-1 band is quite likely the asymmetric BH3 deformation mode on the basis of its intensity in the infrared and Raman spectra and its agreement with the asymmetric BH3 deformation observed around 1120 cm 1 in F3PBH3, CH3NH2BH3 and for other amine boraneso Many of the observable bands below 1100 cm 1 are very difficult to identify and therefore their assignments are not at all certaino Two such bands are observed around 1020 cm-1 and 1064 cm l The C-N stretch for methylamine is generally agreed to be the band at 1044 cm 1 (22,48,49). Bellamy has cited evidence for a coupled C"N-P motion in

61 many dimethylaminophosphinas (51)o The extrapolation of this hypothesis to the present case, results in the assumption that the normal vibration for the C-N stretch probably contains some N-P stretching as well, thereby yielding a C-N-P couple'd motion. With the present information at hand. one is not able to make an unequivocal assignment of either of these frequencies to the proposed C-N-P stretch, although the higher frequency at 1064 cm is favored because experimental evidence tends to show that the C-N stretching frequency oftentimes increases with added substitution on the carbon or nitrogen atoms (13,52)o However, the higher 1064 cm frequency may be the symmetric BH3 deformation vibration because similar vibrations have been observed at 1073 cm-1 for 0CBH3 (21), at 1077 cm-1 for F3PBH3 (5) and at 1070 cm-1 for (CH3)3PBH3. Isotopic substitution of deuterium in the borane group will be necessary:for further clarification of these assignmentso Three very strong infrared bands at 849 cm, 869 cm1 and 915 cm are also observed in the Raman effect, but the intensities of the upper two Raman bands are considerably diminished. Polarization measurements indicate that the two higher bands are depolarized while the lower more intense band is strongly polarized. Although all three of these bands are observed in the P-F stretching region from 720 cm 1 to 950 cm 1 (53,54), the bands at 849 cm 1 and 867 cm 1 are probably the respective symmetric and asymmetric

62 P-F stretching vibrations. These frequencies may be correlated with those observed for Ni(PF3)4 at 859 cm 1 and 898 cm 1 (19), for Ni(PF2CH3)4 at 723 cm 1 and 781 cm 1 (19) and for (CH3)2NPF2 at 743 cm and 792 cm o The higher -1 band at 915 cm may be the asymmetric P-F stretch but is more likely a BH3 rock. Even though borane rocking modes are sometimes found at lower frequencies, this assignment is consistent with that observed for OCBH3 at 816 cm- (21) and for various amine-boranes around 900 cm 1 (50)o Two bands are observed in the 700-800 cm 1 region which may be the N-H rocking and N-P-F2 stretching vibrations. -1 One of the bands, at 773 cm, appears as a very weak to medium intensity shoulder in the infrared and Raman spectra, The other is observed only as a very weak Raman line at 708 cm- 0 Evidence for the fact that the N-P stretch in aminophosphines is probably coupled to neighboring vibrations has been reported by Bellamy (51), Sro Fleming (13) and others (23). Because such a group vibration, eogo N-P-F2 stretch, probably has its greatest amplitude along the N-P bond, its frequency should not be too different from that of P-N stretching motion which is oftentimes found in the region from 680 cm 1 to 750 cm"1 (53,54)o Since the similar group vibration is observed at 705 cm in the spectrum of (CH3)2NPF2, the assignment of the N-P-F2 stretch to the lower frequency at 708 cm-1 is preferred, but the higher assignment to the 773 cm band cannot be ruled outo

63 The NH rocking motion of CH3NH2 has generally been assigned to a band around 724-780 cm l (22,48)o A similar band for the N-H rock has been reported at 724 cm1 for (CH3)2NH (22) and around 730 cml for various dialkylamines (52)o Therefore the assignment of the 773 cm- frequency to a NH rocking vibration is consistent with the present data on the alkylamine s, The strong polarized band at 646 cm1 is assigned to the P-B stretching vibration, The medium intensity shoulder seen at 652 cm1l in the solid infrared spectrum may be the boron-10 component of the P-B vibration. Although this assignment is somewhat higher than that observed for F3PBH3 at 607 cm, for (CH3)2NPF2BH3 at 591 cm 1 and for (CH3)3PBH3 at 571 cm l the intensity of the band in question is comparable to that observed for similar phosphineboranes, A lower band at 596 cmi1 has a frequency which is expected more for a P-B stretch, but its intensity is so weak in the infrared and Raman spectra that it is not considered to be the P-B stretch but rather an unassigned skeletal vibration Several observed vibrations remain to be assigned, The strong, polarized band at 467 cm l is probably the PF2 deformation mode which may be regarded as a remnant of the symmetric PF3 deformation of F3PBH3 at 441 cm o The very weak band at 369 cm 1 has been tentatively assigned to the C-N-P deformation because of its close proximity to the C-N-C

64 deformation vibration at 383 cm 1 for dimethylamine. The weak shoulder at 255 cm 1 is considered to be the methyl torsion because of its low intensity and its nearness to the 269 cm 1 torsional motion of CH3NH2 (55). Although the assignment of the 211 cm1 frequency to a PF2 rock is uncertain, it is justified on the basis that a similar PF3 rock is observed in the spectrum of F3PBH3 at 197 cm"1 (5). The remaining bands at 169 cm-1 and 400 cm- are probably skeletal vibrations but they have not been given specific assignments because of the lack of knowledge concerning many of the lower fundamental frequencies. Bis(methylamino)fluorophosphine-Borane, (CH3NH)2PFBH3. The structure of (CH3NH)2PFBH3 is unknown; however, a phosphorus-boron bond is postulated by analogy to the bis(dimethylamino)fluorophosphine-borane complex. In addition, the assumption that the nitrogen and phosphorus atoms remain at the apexes of the generic NH3 and PF3 pyramids is based largely upon chemical knowledge of similar compounds. It is reasonable to assume that this molecule has no molecular symmetry, but a symmetrical orientation of the CH3NH-groups about the phosphorus atom could lead to a configuration having C symmetry. Since the methylamine groups probably experience some rotation around the N-P bonds, it is likely that more than one rotational isomer exists at room temperature. Therefore no attempt will be made to define the orientation of one methylamine group with

65 respect to the other. The vibrational spectrum of this compound is expected to be very complex due to the existence of these rotational isomers and because a total of forty-eight fundamentals may be observed in both the infrared and the Raman effect. Although some resolution may be achieved between the eighteen methyl group vibrations, the nine borane group vibrations, the six NH group vibrations and the fifteen skeletal vibrations, extensive band overlapping and vibrational coupling present severe difficulties in establishing assignments for the fundamentals. Infrared spectra of liquid and solid films were studied at 25 C and -1800C respectively and are represented in Figures 11 and 12, Also the Raman spectrum of the liquid sample at 0~C was observed along with the polarized spectra, but only the former is presented in Figure 13. The observed infrared and Raman frequencies are listed in Table 6 along with their tentative assignments. For the most part, the infrared spectrum of (CH3NH)2PFBH3 closely resembles that of the difluoro-complex, However, striking differences are observed between the infrared and Raman spectra of the present compound. The unusually weak scattering ability of this complex along with the presence of fluorescent impurities presented severe difficulties to observing the Raman spectrum. On the other hand, the infrared spectrum exhibited many strong absorption bands, several of which were

66 4000 3000 2000 I I I I I I I I I 1500 1000 5 FIGURE 11 - THE INFRARED SPECTRUM OF LIQUID (CH3NH )2PFBH3 (25 C) 1500 00 CM-1

40< 15 00 3000 2000 1500 I I I I I I I I I I i00 1000 500 CM-I FIGURE 12 - THE INFRARED SPECTRUM OF SOLID (CH3NH)2PFBH3 (-180~C)

68 LIQUID (OC) *Hg 0 1000 2000 300-0 CM-I FIGURE 13 - THE RAMAN SPECTRUM OF LIQUID (CH'NH)2PFBH3 (0~C)

69 TABLE 6 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF (CH3NH)2PFBH3 Infrared liquid(200C) solid(-1800C) 3365 vs 3343+10 s 2940 m 2954+10 w,sh 2905 m,sh 2910+10 w Raman liquid(0 C ) 3366 vw 2947 w,p? 2912 w,sh,p? 2830 w,p 2391 m,sh,dp 2350 ms,p Assignment 2828 w 2388 vs 2359 s,sh 2265 m,sh 2121 vw,sh? 1595 vvw,sh? 1476 m,sh,br? 2802+10 vw,sh? 2380+10 s N-H stretch asym, C-H stretch asym, or sym, C-H stretch sym, C~H stretch asymn B-H stretch sym, B-H stretch 2 x 1136 = 2272 2 x 1147 = 2294 2 x 1063 = 2126 2 x 800 = 1600 asym, CH deformation 2288+10 w,sh 1473 m 1461 m,sh,br? 1390 vs,br 1465 w,sh 1453 m 1422 w,sh 1460 vw 1432 vw asym. CH deformation asym, CH deformation? sym, CHI def omation 1404 s 1169 vw,sh 11.47 w,sh 1136 w, dp 1098 vw,sh? CH3 rock asymo BH deformation ard/or CH3 rock? sym, BHI deformation and/or out of phase (C-NN)2-P stretch? 1088 vs,br 1117 s,sh? 1100 s

70 TABLE 6 Infrared liquid(200C) solid(-180~C) 1063 vs,br 1062 s CONT ~D Raman liquid(0~C) 1056 vvwsh? 874 vvw 836 vw,sh 801 vw,p? Assignment 875 vs,br 837 vs, br? 800 vs,br 729 w,sh 651 w,sh? 629 m 590 w,sh,br? 877 m,sh 842 s 793 s 749 w,sh 720 vw,sh in phase (C-N)2-P stretch BH3 rock? P-F stretch? P-F stretch NH rock and/or N2 -P~F stretch NH rock and/or N2 P-F stretch skeletal? P-B stretch skeletal PF rock skeletal? skeletal skeletal 631 s 532 w, vbr?? 462 vw 623 w,p? 589 vw9,shp? 467 w,p? 424 vvw? 407 vvw?,p? 347 vvw?

71 very broad in the liquid spectrum, By cooling the sample to -180OC, the infrared bands were considerably sharpened, causing components of the band envelopes to be resolved. The increased resolution in the solid spectrum may result partly from the curtailment of internal rotation within the molecule as a consequence of increased barriers in the solid state as well as reduction of the thermal motion, The very strong infrared band and the very weak Raman band around 3365 cm-1 is most certainly the N-H stretching vibration in view of the similar vibration found near 3369 cm in the spectra of CH3NHPF2BH3 The methyl stretching and deformation vibrations in the 2900 cm1l and 1450 cm 1 regions show no peculiarities and are in fact nearly identical to those of the monosubstituted complex. Reference may be made to the previous section for specific information concerning these vibrations~ As in the spectra of CH3NHPF2BH3 and similar fluorophosphine-borane complexes, the B-H stretching vibrations are observed as a very characteristic doublet around 2359 cm-1 and 2388 cm l1 The lower band is definitely polarized and is assigned to the symmetric stretch while the higher 2388 cm,1 depolarized band is the asymmetric B-H stretching vibration, The medium intensity shoulder found in the infrared at 2265 cm-1 is probably an overtone, perhaps 2 x 1147 = 2294, rather than a B-H stretching fundamental,

72 The frequencies below 1200 cm 1 are very difficult to assign because of extensive overlapping of the fundamentals, Many of these assignments are uncertain and several alternate possibilities may exist. The first of such difficulties arises in identifying the methyl rocking vibrations, The very weak shoulder at 1169 cm"1 in the solid infrared spectrum is in all probability a methyl rock,but a lower depolarized band at 1136~1147 cml may either be a methyl rock or the asymmetric BH3 deformationo The assignment of this band to the BH3 fundamental is preferred because similar absorption is observed for (CH3)2NPF2BH3 at 1128 cm, for [(CH3)2N]2PFBH3 at 1131 cm 1 and for (CH3)3PBH3 at 1137 cm o Similar difficulties are encountered in establishing the identity of bands between 1000-1100 cm l1 Two partially resolved, very strong infrared bands are observed in this region- one at 1063 cm 1 and the other at 1088 cm 1o These bands are resolved further in the solid infrared spectrum and both are present in the Raman effecto Since the C-N stretch for CH3NH2 is observed at 1044 cm1, either or both of these bands may be attributed to the proposed C-N-P group vibrationo The modification of the C-N-P group to the C ~P entity may cause the frequency to be split, C 0Nperhaps to 1063 cml1 and 1088 cm l but no definite evidence for such splitting has been observedo Justification for such a postulate may be found in the 942, 991 cm C2=-N-P

73 stretching doublet of [(CH3)2N]2PF and the 969, 999 cml doublet of [(CH3)2N]2PFBH3o In addition to the previous possibilities the higher, 1088 cm1, band may be the symmetric BH3 deformation vibration. No definite choice is made concerning these assignments and isotopic substitution of the borane group will be required to resolve the difficultyo Three broad, very strong infrared bands are observed at 800 cm 1, 837 cm"1 and 875 cm l. They also appear in the Raman effect and the 800 cm 1 band seems to be polarizedo Although all of these bands are found in the P-F stretching region, the assignment of the 800 cm 1 band to the P-F stretching fundamental is preferred because of its greater band intensity. The 875 cm 1 band may be a BH3 rocking vibration, but the 837 cm l band remains to be assigned, The assignments for the very weak shoulders at 720 cm" and 749 cm'1 are uncertain but are probably the NH rocking vibration and/or the N2-P-F group frequency with no preference being given to either assignmento A definite possibility exists that both of these bands are components of the N2p PF group frequency in view of the fact that this frequency appears to be split at 644 cm l, 689 cm 1 and at 685 cm A 706 cm 1 in the spectra of [(CH3)2N]2PF and of [(CH3)2N]2PFBH3 respectivelyo A medium to strong intensity band at 629 cm 1 is observed in the liquid and solid infrared spectra and is seen as a weakly polarized, weak band in the Raman effecto

74 This vibration is assigned to the P-B stretch, because of its strong band intensity in the infrared, Although the assignment of this vibration to a lower, very weak, band at 590 cm1 correlates better with the observed P-B vibrations of similar fluorophosphine-borane complexes, the very low intensity of the latter band removes it from serious consideration and instead it may be an unassigned skeletal vibration. The very weak band at 462 cm1" is assigned to a PF rocking motion to agree with the related PF2 deformation vibration of CH3NHPF2BH3 assigned at 467 cm l The remaining bands at 651 cm-1, 590 cm, 424 cm, 407 cm-1 and 347 cm-1 are all very weak to weak in intensity and are probably skeletal vibrationso Further classification of these is impossible at this time, The weak, very broad 532 cm band has not been termed a skeletal mode because of its doubtful origino The frequencies which are assigned to combinations or overtones are considered to be tentative and may be seen in Table 6, Dimethylaminodifluorophosphine, (CH3 )2NPF2. Although no structural data have been reported in the literature for (CH3)2NPF2, it is logical to assume that the pyramidal configurations of (CH3)2NH and PF3 are retained with the formation of (CH3)2NPF2o Sr, Fleming (13) has invoked theoretical arguments along with experimental evidence to show that (CH3)2NPF2 probably has a structure in which the methyl groups and the fluorine atoms are in a trans configu

75 ration, although a gauche configuration cannot be completely ruled out. Assuming the trans configuration, one would expect that this molecule possesses only one element of symmetry, a vertical reflection plane, vV, and would belong to the Cs point group. The Cs symmetry of the molecule requires that it have thirty active vibrational modes, of which sixteen are of the A' symmetry species and fourteen are of the A" symmetry species, Eighteen of the total number of vibrations are localized primarily in the C2NPF2 framework of the molecule, A complete numerical listing of the fundamental vibrational modes along with a description of each mode is given in Table 7o Since the atomic masses in the (CH3)2NPF2 skeleton are very similar, one may expect the skeletal modes to involve extensive coupling of the individual bond stretching and deformation motions, This is particularly true in the case of w9 and W 11 which appear to be modes involving the symmetric stretching of all the bonds in the C2-N-P and N-P-F2 groups, This is indicated by the fact that they do not occur at quite the positions at which "pure" C-N and N-P frequencies occuro The observed Raman spectra of liquid and solid (CH3)2NPF2 at -32~C and -180~C respectively are shown in Figure 14 and the polarization spectra for the liquid sample are shown in Figure 15o As an aid in making assignments to the vibrational

76 modes in the molecule the observed Raman frequencies of the sample are compared to published infrared frequencies of Ni(PF2CH3)4 (19), (CH3)2NH (22) and (CH3)2NPC12 (23) in Table 8. Also, the observed infrared frequencies of (CH3)2NPF2, reported by Sr, Fleming (13) and Schmutzler (56), are tabulated along with the present Raman results in Table 9, An inspection of this table reveals that the observed Raman frequencies agree with the infrared values, within the range of experimental error. This table also includes the vibrational assignments which were deduced from the present investigationo Although most of the assignments for the skeletal modes appear fairly certain, the assignments for the overtones, combinations and methyl group vibrations are less certain and should be considered only tentative. The infrared spectrum of the sample for the most part resembles that of the Raman effect, however, some striking differences also appear. Such differences are present in the C-H stretching region from 2800 cm-1 to 3000 cm-1 where only three bands were observed in the infrared spectrum (13) while at least five bands are clearly resolved in the Raman spectrum. These bands are probably components of the six symmetric and asymmetric C-H stretching vibrationso However, there is a distinct possibility that one or more of these bands may be due to overtones of the deformation modeso The methyl deformation frequencies in the region from 1300 cm1 to 1500 cm1 are well characterized and tentative

77 0 1000 2000 3000 CM-4 FIGURE 14 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (-32 c) (CH )2NPF2

78 * E(1) E UI) 0 1000 2000 3000 CM-, FIGURE 15 - THE POLARIZED RAMAN SPECTRA OF LIQUID (CH3)2NPF2 (-320C)

79 TABLE 7 THE FUNDAMENTAL VIBRATIONS OF (CH3)2NPF2, Cs SYMMETRY Sy metry Number Description Al 1 in phase asymmetric C-H stretch 2 in phase asymmetric C-H stretch 3 in phase symmetric C-H stretch 4 in phase asymmetric CH3 deformation 5 in phase asymmetric CH3 deformation 6 in phase symmetric CH3 deformation 7 CH3 rock (out of C-N-C plane) 8 CH3 rock (in C-N-C plane) or CH3 wag 9 C2-N-P stretch 10 P-F stretch 11 N —P-F2 stretch 12 F-P-F deformation 13 C-N-C deformation 14 NC2 rock 15 PF2 rock 16 CH torsion 17 18 19 20 21 22 out out out out out out of phase asymmetric C-H stretch of phase asymmetric C-H stretch of phase symmetric C-H stretch of phase asymmetric CH3 deformation of phase asymmetric CH3 deformation of phase symmetric CH3 deformation

80 TABLE 7 - CONT'Do Symmetry Number A'l 23 24 25 26 27 28 29 30 CH3 CH4 C-N P-F NC2 PF2 CH3 P-N Description rock (out of C-N-C plane) rock (in C-N-C plane) or CH3 wag stretch stretch wag wag torsion torsion

TABLE 8 Ni(PF2CH 3)4( ) Infrared 19) A COMPARISON OF THE OBSERVED VIBRATIONAL FREQUENCIES ( Ni(PF2CH3)4, (CH3)2NH, (CH3)2NPCl AND (CH3)2NPF2 (CH3)2NH(g) (CH3)2NPC12( ) (CH3)2NPF2(e, -300C) InfraInfrared(232). Raman. 3355 2967 2840 ms 2998 s,sh,dp 2945 vs,dp? 2912 2855 2802 1496 1466 1404 2790 mw 1478 ms 1450 ms 2912 2861 2814 1487 1439 1414 1306 vs,p? s,p s,p m,dp s,dp w,sh.dp? m,p DF Assignment and Symmetry [N-H stretch (a')] asym. C-H stretch (a") sym. or asym. C-H stretch (a" or a') asym, C-H stretch (a')? asym. C-H stretch (at)? sym, C-H stretch (a' and a"?) asym. CH deformation (a") asym, CH deformation (a' ) sym. CH3 de3 formation (a")? sym, CH deformation (a' and a"?) [CH3 rock (a")] CH rock (a' and/ 3or a") CO 1 —i 1287 ms 1245 1155 1175 ms 1103 vw

Ni(PF2CH 3)L4() Infrared(1)_ (CH3)2NH(g) Infrared(22) 1024 TABLE 8 - CONT'D. (CH3)2NPC12(2) (CH3)2NPF2(e, -30~C) Infrared 23) Raman 1063 m 1071 w,dp? 978 vs 989 m,dp? Assignment and Symmetry 930 781 723 792 743 m,p? m, dp? 724 705 vs,p 690 s C-N stretch (a") C2-N-P stretch (a') [C-N stretch (a')] P-F stretch (a") P-F stretch (a') [NH bend (a")] N-P-F2 stretch (a') [N-PC12 or C2N-P stretch (a')] PF2 deformation (a') C-N-C deformation (a') skeletal [CH3 twist (at)] [CH3 twist (a")] PF2 rock (a') and/ or PF2 wag (a") ro R) 485 442 495 w 393 vw (397) 336 w (290) (250) 168 vvw

TABLE 9 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF (CH3)2NPF2 Gas (13) Infrared Liquid(56) Solid(13) (-180~c) 3002 vw 2932 mw 2925 m 2908 s 2865 sh 2820 w 1507 2808 s 1691 w 1484 s 1468 w 1450 s 1432 w 2817 vw 1498 w 1490 m 1473 w 1463 w 1455 m 1433 w 1430 vw 1313 s 1233 vw 1206 m 1191 s 1145 vw Raman Liquid (-3o~C) 2998 s,sh,dp 2945 vs,dp? 2912 vs,p? 2861 s,p 2814 s,p 14887 m,dp 1439 s,dp 1414 wshdp? 1306,p Solid (-_800c) 2996 s 2935 vs 2906 vs,sh 2867 s 2815 s 1495 m 1432 s 1315 m p17' V18 v19? or 1 vl 27 7v3' 19? 705 + 990 = 1695 728 + 764 = 1492 v20' 721? 707 + 764 = 1471 393 + 1073= 1466 21? 74? 707 + 728 = 1435 v5 or v22 v6,v722? 499 + 728 = 1227 ~23o w7, 23 or 25 V8 or v24 Assignment 1307 ma 1195 m 1308 vs 1182 vs

TABLE 9 - CONT'D. Gas (13) Infrared Liquid(56 Raman Solid(13) (-180oc) Liquid....(-30~ ) 1103 vw 1071 w,dp? Solid (-18o0c). 1080~Ci 1107 m 1075 m Assignment 1073 w 989 s 814 s 770 s 704 m 501 w,br 1068 s 990 vs 800 vs 743 vs 705 s 1073 m 1002 sh 994 s 764 s 728 m 707 vs 499 w 486 vvw V8 or v24 w25 2 x 499 = 998 989 792 743 705 495 393 336 168 m, dp? m,p? m, dp? vs,p w 999 m 759 m 704 s 494 w v9 v26 vlO.11 vll w12 skeletal? 13? skeletal? v15 or 728? OC vw w vVW

85 frequency assignments have been made for these vibrations with the assistance of polarization data. Although considerable vibrational data have appeared in the literature concerning dimethylamine and dimethylaminoderivatives, there seems to be some disagreement about the assignments for the methyl rocking and asymmetric C-N stretching vibrations. The methyl rocking and the asymmetric C-N stretching frequencies most generally have been assigned to bands in the regions 1125-1260 cm-1 and 1015-1090 cm-1 respectively for (CH3)2NH (22,57,58), for [(CH3)2NH2]C1 (58), and for [(CH3)2NH2]I (59). On the other hand, evidence for a higher frequency asymmetric C-N stretch around 1020-1220 cm(for secondary aliphatic amines) has been cited by Bellamy (51) and Stewart (52). In addition, Goubeau, Rahtz and Becher (60) and also Banister and co-workers (61) have assigned the asymmetric C-N stretch in (CH3)2NBC12 to a band around 1143 cm1 and the methyl rocks to bands near 1199 cm and 978-1068 cm. The higher C-N frequency assignment is substantiated further by the observation of an asymmetric C-C stretching frequency near 1170 cm-1 in the 2-methylalkanes (51). With the present information at hand, one is not able to make definite assignments to the methyl rocks and asymmetric C-N stretch without further spectroscopic data of (CD3)2NH and some of its derivatives. -1 A very strong band around 1190 cm is observed in the infrared spectrum of (CH3)2NPF2 (13,56) but not in the Raman.

86 The assignment of this frequency to a methyl rock (13), perhaps the methyl rocking motions perpendicular to the C-N-C plane, is consistent with similar assignments for (CH3)2NHo A very weak band at 1103 cm-1 is..observed.in.the "Raman spectrum but not in the infrared and is thought to be one or both of the in C-N-C plane methyl rocking motions. The remaining torsional frequencies were not observed in the Raman spectrum but are probably found near 260 cm-1 as evidenced by far infrared studies of (CH3)2NH (62), The tentative assignments for the asymmetric C-N stretching and the symmetric C2-N-P group vibrations have been made identical to those of Sr. Fleming (13). The respective motions -1 are observed in the Raman as a weak band at 1071 cm and as a medium band at 989 cm- 1 Polarization data for these bands are of little value because the band intensities are too small to determine depolarization ratios accurately. The asymmetric C-N stretch frequency at 1071 cm1 correlates well with that reported for (CH3)2NH (22) and for (CH3)2NPC12 (23), while both the position and existence of the C2-N-P group vibration at 989 cm1 is cited by Bellamy (51) and also by Holmstedt and Larsson (63). The intense infrared bands at 743 cm1 and 792 cm 1 without doubt are the symmetric and asymmetric P-F stretching vibrations, because similar bands are found at 874 cm 1 and 832 cm1 for PF3 (17), at 944 cm1 and 957 cm1 in F3PBH3 (5), and at 723 cm1 and 781 cm1 in Ni(PF2CH3)1) (19). However,

there is some question as to which mode is higher in frequency, the asymmetric or symmetric vibration. In PF3 the symmetric mode is higher while in F3PBH3 the asymmetric mode is highero Raman polarization results favor the lower frequency as the asymmetric mode but the bands are sufficiently weak in the Raman effect to make accurate determination of the depolarization ratio impossible. Therefore the higher frequency was assigned to the asymmetric mode in accordance with the present P-F stretching assignments for F3PBH3 and fluorophosphine derivatives, Several workers have attempted to establish a frequency region for the P-N stretching vibration (51,53,54,63)o Although the intensity and position of the N-P stretching band seem to be influenced by neighboring groups, it is quite frequently found in the region from 680 cm1 to 750 cm- -1 Therefore the strong band at 705 cml very likely is the symmetric N-P-F2 stretching group frequency in agreement with the assignment proposed by Sr. Fleming (13)o It is also interesting to note that a similar group vibration is mknoan to occur in the infrared spectrum of (CH3)2NPC12 at 690 cm1 (23)o A weak band at 495 cm1 is found in both the infrared and Raman spectra and is probably the symmetric PF2 deformation frequency because the comparable band in PF3 is found around 484 cm 1 and at 482 cm-1 in Ni(PF2C6H5)4 (19). The very weak absorption band at 393 cm1 is considered to be the NC2 deforma

88 tion frequency in agreement with the assignment of a weak Raman band at 390 cm-1 (57) and a far infrared band at 383 cm l (62) for (CH3)2NH. Stewart also reports that this deformation mode appears in the infrared around 427 + 14 cm-1 for a large number of dialkylamines: (25), Three bands remain to be assigned- a very weak infrared band at 486 cm1 (13) and two weak Raman bands at 336 cm-1 and 168 cm1. The low band at 168 cm-1 may be one of the PF3 rocking modes because the corresponding degenerate vibration in F3PBH3 is found near 197 cm1 (5)~ No attempt has been made to characterize the other two bands because little or no data. is avilable concerning the possible frequencies of the NC2 rocks and the P-N torsional vibrationO Dimethylaminodifluorophosphine-Borane, (CH3 )2NPF2BH3 and (CH3)2NPF2B D3 This molecule at most may have Cs molecular symmetry assuming a similar symmetry for the free ligand, (CH3)2NPF2, and knowing that a phosphorus-boron dative bond is present as shown by nuclear magnetic resonance studies (13)o The hypothesis that this molecule belongs to the Cs point group requires that it have forty-two fundamental vibrations with the symmetry distribution 23A' + 19A"o Of these modes, thirty may be considered as belonging to the (CH3)2NPF2 ligand, six involve the vibrations of the borane group and six motions arise as a consequence of the dative bond formed between the

89 Lewis base and the borane Lewis acid, A complete listing of the fundamental vibrations and descriptions of their motions are tabulated in Table 10o The observed Raman spectra of liquid (-24~C) and of solid (-180~C) (CH3)2NPF2BH3 are shown in Figure 16 and the polarized Raman spectra may be seen in Figure 17o In addition, infrared absorption spectra of gaseous and liquid (CH3)2NPF2BllD3 were observed and are represented in Figures 18 and 19 3 respectively, Raman spectra of the liquid and solid deuterated samples were also recorded and are shown in Figures 20 and 21. The observed Raman vibrational frequencies of (CH3)2NPF2BH3 agree, within the limits of experimental error, with the infrared values reported in the literature (13)o The observed infrared and Raman frequencies and their tentative assignments, for both the natural hydrogen and the deuterated borane complexes, are given in Tables 11 and 12 respectively. For purposes of comparison, several vibrational frequencies are exhibited in Table 13 which are particularly sensitive to isotopic substitution in the borane groupo A more detailed discussion of these figures and tables followso The vibrational spectra of (CH3)2NPF2BH3 and of the borane 11-d3 complex are remarkably similar to that of the free ligand, The most evident similarities are found in the methyl group stretching and deformation regions around 2900 cm and 1450 cm 1 respectively. Because the intensities,

90 0 1000 2000 3000CM-I FIGURE 16 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (-24 ~) (CH 3) NPF2BH3

91 * E(II) E (I) *Hg 0 1000 2000 3000 CM FIGURE 17 - THE POLARIZED RAMAN SPECTRA OF LIQUID (CH3)2NPF2BH3 (-15~C)

92 4000 3000 2000 15C I I I I I I I I I I 1500 1000 5C FIGURE 18 - THE INFRARED SPECTRUM OF GASEOUS (CH3)2NPF2B D3 (11 mm) )0 )0 CM-I

93 4000 3000 2000 1500 I I I I I.I I I 11 1500 1000 600 CM-I FIGURE 19 - THE INFRARED SPECTRUM OF LIQUID ( CH3)-NPF2B D (25 OC)

94 0 1000 2000 3000 CMFIGURE 20 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (-29~C) (CH3)2NPF2B D3 32 2 3

95 * E ) EII1) *Hg 0 2000 0000 CM FIGURE 21 - THE POLARIZED RAMAN SPECTRA OF LIQUID (CH3)2NPF2B1lD3 (-29~C)

96 TABLE 10 THE FUNDAMENTAL VIBRATIONS OF (CH3)2NPF2BH3, Cs SYMMETRY Symmetry Number A' 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Desc ript ion in phase asymmetric C-H stretch in phase asymmetric C-H stretch in phase symmetric C-H stretch asymmetric B-H stretch symmetric B-H stretch in phase asymmetric CH3 deformation in phase asymmetric CH3 deformation in phase symmetric CH3 deformation asymmetric BH3 deformation symmetric BH3 deformation CH3 rock (out of C-N-C plane) CH3 rock (in C-N-C plane) or CH3 wag BH3 rock C2-N-P stretch P~F stretch N —P-F2 stretch P-B stretch N-P-B deformation F-P-F deformation C-N-C deformation C2N rock PF2 rock CH3 torsion 3

97 TABLE 10 - CONTVD, Symmetry Number Description A'T 24 out of phase asymmetric C-H stretch 25 out of phase asymmetric C-H stretch 26 out of phase symmetric C-H stretch 27 asymmetric B-H stretch 28 out of phase asymmetric CH3 deformation 29 out of phase asymmetric CH3 deformation 30 out of phase symmetric CH3 deformation 31 asymmetric BH3 deformation 32 CH3 rock (out of C-N-C plane) 33 CH3 rock (in C-N-C plane) or CH3 wag 34 BH3 wag 35 C-N stretch 36 P-F stretch 37 F-P-B deformation 38 C2N wag 39 PF2 wag 40 CH3 torsion 41 BH3 torsion 42 C2N torsion or P-N torsion

TABLE 11 THE OBSERVED INFRARED Infrared gas(25' C) liquid( 25C) 3014 vw AND RAMAN VIBRATIONAL FREQUENCIES OF (CH3 )2NPF2BH3 Raman As s ignment - - - l- - - - - - - %, liquid ( -24 vC) solid(-180C ) 2945 m 2885 w, sh 2820 w 2620 vwbr 2440 s 2375 w 2245 vw 14701490 sh 1465 m 1320 ms 1187 m 1121 w 2937 w 2917 w 2810 vw 2413 ms 2348 w, sh 2228 vw 1486 w 1455 w 1323 m 1180 ms 1125 w 3011 s,sh,dp 2971 vs,p 2941 vs,sh,p 2917 s,sh,p 2875 msh,p 2827 msh,p 2421 vs,dp 2366 vs,p 2249 vw,p 2144 vwp 1486 w,sh,dp 1445 m,dp 1418 w,sh,dp 1325 w,p 1128 m,dp 1108 w, sh 3015 m, sh 2972 ms 2942 ms 2917 ms,sh 2864 m, sh 2813 m, sh 2419 vs 2364 vs 2311 vw 2251 vw 1492 m 1439 ms 1413 w,sh 1331 m 1130 ms V24,) V25 2 x 1486 = 2972 v26? W? w3,'26? 2 x 1325 = 2650 V4, v27 v5 1128 + 1180 = 2308 2 x 1128 = 2256 2 x 1074 = 2148 V28 w29'w6? w8 w30 lly32 or V35? "9 v31 12 9 33 co

TABLE 11 - CONTUD. Infrared gas(25 ~C) liquid (25 C) 1071 m 1070 mw 0110 vs 1006 s 891v 880 873 vs,br liquid (-240 C 1074 wp? 1013 w, dp? 862 vw,dp? 753 s 701 w 851 S 765 ms 694 vw 586 441 mw vvw, br 586 mw 443 vvwbr 768 705 660 591 447 408 376 344 270 211 m,p vw ww m,p m,p? vvw vvw wp? w, dp? w dp? Raman * ) solid(-1l80C) 1072 m 1008 m 877 w 853 w 767 ms 705 vw 659 ww 590 ms 445 w 346 vw w35 and v10 V14'36 v15 v13 V16 skeletal As signment v17 l19 V20 skeletal skeletal v23 and/or V40 skeletal kl',D

TABLE 12 THE OBSERVED INFRARED Infrared gas(llmm) liquid(25~C) 3020+10 vw,sh 3008+10 w,sh AND RAMAN VIBRATIONAL FREQUENCIES OF (CH3 )2NPF2BllD3 2948+10 s 2880+10 m,sh 2829+10 w,sh 2405+10 w 2355+10 w,sh 1827 s 1792 vvw,sh 17521732 vw,br 17171697 vvw,br 1492 m,sh 1460 im 2942+10 ms 2867+10 m,sh 2816+10 m,sh 2385+ 5 w 2357+ 5 w,sh 1819 s 1746 w,sh 1704 w,sh 1649 vvw 1616 vw 1569 vw 1507 w,sh? 1488 ms 1454 ms Raman liquid(-300C) solid(-1800C) 3009 w,sh,dp 3024 vw 2974 m,sh,p 2980 w 2945 m,p 2948 w 2923 m,sh,p 2918 w 2872 w,p 2867 vw 2826 w,p 2819 vw 2395 vw,p 2388 vvw Assignment p24,' 25 2 x 1488 = 2976 2? Vw3 26? 1070 + 1322 = 2392 5 (BH3) v4' 27 J 0 0 1818 vs, dp 1743 ms,p 1705 vw,p 1652 vw,sh,p? 1618 vw,sh,p? 1485 w,dp 1446 m, dp 1822 s 1750 m 1708 ms 1656 vvw 1496 w 1445 w 2 x 873 = 1746 or vw7 5 737 + 912 = 1649 2 x 814 = 1628 252 + 1322 = 1574 183 + 1322 = 1505 w28 w29,6 7

Infrared gas (llmm) liquid(250C) 1419 vw, sh 1321 s 1322 s 1189 s 1178 s TABLE 12 - CONTV'D Raman liquid(~30~C) solid(-180~C) 1420 w,sh,dp 1323 w,p 1434 w Assignment 1071 w 1007 vs 913 s 873 vs 1070 ms 1007 s 1101 vvw 1069 vvw 1017 w 980+5 ms,sh? 916 vvw,p? 1104 1070 1019 vvw VW vw V89 30 vll 32 or 35o w12 v33 v35 V14 252 + 737 = 989 V10 V36 912 s, br 894 s,sh? 870850 s,br 826 s,sh 810 s 747 s,sh 737 s 860 vw,sh,dp? 814 w,dp? 911 vvw 862 vw 818 w 743 m O 0 H 817 745 738 mssh s, sh s V9' 31 V15 skeletal 584 w 552 ms 736 651 588 549 423 ms,p vvw? w,p m,p s,p 596 552 435 vw w VW 398 vvw,sh 340 v,p 252 w 183 m,dp V13 V17 l19 V20 skeletal w23 and/or V40 skeletal

102 TABLE 13 A COMPARISON OF SOME ISOTOPE DEPENDENT VIBRATIONS OF (CH3 ) 2NPF2BH3 AND (CH3 ) 2NPF2B1DD 3 (CM1 ) Assignment* w49 27 B-H v5 vB-H 5 9 *w31 aBH S o10 B BH3 13 P BH3 133 734 " BH3 w41 BH3 w17 y PB (CH )2NPF2BH3 2418 2362 1128 (1072) 767 (CH) 2NPF2B D3 1821 1706 815 913 589 Frequency Ratio 1033 1.38 1038 (1017) 1030 589 551 1007 *v = stretch, T = torsion 5 = deformation, p = rock, cl = wag and

103 shapes and frequencies of these bands are nearly identical to those of the ligand and to a lesser degree those of dimethylamine, their assignments are based upon those observed for the latter two compoundso Although these are somewhat arbitrary, they are probably as reliable as those reported for dimethylamine and its derivativeso Reference may be made to Tables 11 and 12 for specific assignments to these vibrations The very strong doublet which appears near 2362 cm1 and 2418 cm 1 in the Raman spectrum of (CH3)pNPF2BH3 is most certainly due to components of the symmetric and asymmetric borane stretching fundamentals, since they are not present in the spectrum of the ligand and because they shift to 1706 cm1 and 1821 cm 1 upon deuteration. The weak to medium intensity band at 1746 cm 1, found in the infrared and Raman spectra of the deuterated sample, is considered to be an overtone of 36 rather than the symmetric B-H stretching fundamental because its intensity is considerably weaker than the other two bandso The relatively low ratio of the 2362 cm"1 and 1746 cm 1 frequencies, 135, also gives credence to this assignment. Since only one asymmetric B-H stretching vibration is observed, the influence of the ligand is probably not sufficient to cause an observable splitting of wvl and v27o A strong band is observed near 1185 cm' in the infrared spectra of both isotopic samples, but is absent in the Raman effect, Sro Fleming has assigned this band to a methyl rock (13)o

104 However, in view of the fact that some workers have assigned the asymmetric C-N stretch of various secondary aliphatic amines to bands in this region, this assignment remains uncertain~ Of the two possibilities, the former is given more weight because it agrees with the majority of the assignments reported in the literatureo A lower band at 1105 cm, appearing only in the Raman effect, probably is a methyl rock since a similar band is observed in the Raman spectrum of the free ligando A moderately intense, depolarized band is observed near 1128 cm-1 in both the infrared and Raman spectra of the nondeuterated sample. Since this band is absent in the spectrum of the deuterated sample, it is assigned to the asymmetric borane deformation modes, v9 and v31 This assignment correlates quite well with the known asymmetric borane deformation of (CH3)3NBH3 at 1169 cm 1 (69), of (CH3)3PBH3 at 1136 cm-1 and of F3PBH3 at 1117 cm 1 (5). Upon deuteration this vibration shifts to 815 cm 1 yielding an isotopic frequency ratio of lo38, The assignments of the C2-N-P group vibration and of the asymmetric C-N stretch to bands at 1008 cm 1 and 1072 cm 1 respectively are made in agreement with those proposed by Sr. Fleming (13)o The assignment of the higher frequency bears further explanation because other possibilities existo Although the asymmetric C-N stretch in many dialkylamines has been reported to be near 1160 cm 1 (52), the majority of

105 workers place it in the 1025-1100 cm regiono Therefore, in keeping with the majority, the assignment of the C-N fundamental to the 1072 cm1 band is preferredo Because of the closeness of this band to the observed symmetric borane deformation vibrations in OCBH3 (21), F3PBH3 and (CH3)3PBH3 at 1073 cm, 1077 cm- and 1070 cml respectively, it is likely that the band envelope also contains the borane deformation fundamental. If this were the case, deuteration of the sample would cause this frequency to shift to around 850 cm-1 as is observed for many borane complexeso Several vibrational bands are observed in the region 725-925 cm 1 for both isotopic varieties of the complex, In the infrared spectrum of the non-deuterated sample three bands are observed: a medium strong band at 767 cm-1 and a strong doublet at 851 cm-l and 880 cm-1 The identical bands appear in the Raman effect but their intensities are reversed~ Although Sro Fleming assigned the two lower bands to P-F stretching fundamentals and the higher to a borane rock (13), it is believed that these bands are better assigned by reversing the order of the previous assignmentso These are reversed on the basis that the P-F stretching fundamentals of (CH3)2NPF2BF3, tentatively assigned to bands at 850 cm 1 and 880 cm - coincide exactly with those of the borane complex. Even though polarization data are inconclusive, the higher vibration at 880 cm-1 is assigned to the asymmetric P-F stretch and the lower to the symmetric fundamental by analogy

106 to similar compounds. The remaining polarized band at 767 ciml assigned to the borane rock, shifts to 589 cm-1 upon deuterationo This frequency shift is larger than expected (ratio of the frequencies = o130), however it may be justified in view of the strong vibrational coupling observed in the spectrum of the deuterated sample, In the region in question, the vibrational spectrum of the deuterated sample is not well understood because of its complexity. Four bands around 738 cm-, 815 cm-1 865 cm-1 and 913 cm-1 are clearly observed along with several shoulderso Of these bands, the two near 815 cm-1 and 865 cm-1 appear to be depolarized, while the 738 cm-1 band is definitely polarized and the polarization of the 913 cm-1 band is questionableo The 815 cmr1 band has been previously assigned to the asymmetric borane-d3 deformationo The band at 865 cm-1 is not affected by isotopic borane substitution and probably is the asyrmmietric P-F stretch. The two bands at 738 cm-1 and 913 cm 1 cannot be explained on the basis of Fermi resonance or overtones, and their strong band intensities suggest that they are fundamentalso Because both the expected symmetric borane-d3 deformation and the symmetric P-F stretch are found near 850 cm- 1 these modes may be expected to couple very strongly with each other, thereby causing their frequencies to splito If this is true, then the 738 cm-1 band may be the symmetric P-F stretch and the 913 cm1 band may very well be the symmetric borane-d3 deformation vibrationo The exceptionally low frequency ratio of the borane deformation funda

107 mentals, (1072)/913 = 117, is probably the result of such couplingo The observed shoulders remain unassignedo Assignments to the remaining observed frequencies require little discussion and wll only be enumerated. A weak band at 700 cm is observed in both the infrared and Raman spectra of the non-deuterated sample, but is absent in the spectra of the deuterated compound. The assignment of the N-P-F2 stretch to this vibration by Sr. Fleming (13) remains unchanged in the present work, The polarized band near 589 cm1 in the hydrogen compound shifts to 551 cm1 upon deuteration, and is most assuredly the P-B stretching vibration, This assignment correlates very well with the observed P-B stretch in F3PBH3 at 607 cm-, The polarized band near 444 cm-1 and 430 cm1 in the respective natural hydrogen and deuterated samples is assigned to the PF2 deformation in view of a similar vibration of the ligand near 393 cm o The weak Raman bands at 408-398 cm 1 and at 270-252 cm-1 are assigned to the C2N deformation and methyl torsional modes respectively by analogy to similar vibrations in dimethylamine at 383 cm and 257 cm 1 (62), Several skeletal vibrations are observed below 600 cm-1 and no attempt has been made to characterize them. The assignments for combinations and overtones which are listed in Tables 11 and 12 are quite arbitrary because of uncertainties involved with the selection of the fundamentals

108 Dimethylaminodifluorophosphine-Boron Trifluoride, (CH3)2NPF2B1OF3 and (CH )2NPF2BllF30 Nuclear magnetic resonance studies have indicated that the boron trifluoride Lewis acid does not bond through the phosphorus atom of (CH3)2NPF2 as does the borane group, but rather through the nitrogen atom (13)o Although the bonding site in the boron trifluoride complex is different than in the borane complex, the molecule is expected to have essentially the same C molecular syrzmmetry, Analogous to the previously mentioned borane complex, this molecule possesses forty-two Raman and infrared active fundamental vibrations, six of which may be regarded as involving primarily motions of the boron trifluoride group, six result from the formation of the B-N dative bond and the remaining fundamentals may be considered as belonging to the Lewis base, Raman spectra of (CH3)2NPF2BF3 at -78~C, enriched in the boron-10 and 11 isotopes, were studied and are represented in Figure 22~ Although this compound is reported to have a very intense infrared absorption spectrum (13), its Raman spectrum is very weak, The weak scattering ability of this compound is probably the primary factor contributing to the low intensity of the Raman spectrum. The observed infrared and Raman frequencies of (CH3)2NPF2B F3 are compared with those of (CH3)2NPF2 and of (CII3)3NB1F3 in Table 14, In addition, the reported infrared frequencies, their observed

109 Raman counterparts and their tentative assignments, are listed in Table 15 for both the boron-10 and 11 trifluoride complexeso For the most part, the present assignments are identical to those reported for the infrared results (13). Therefore, only those frequencies and their assignments which are different will be mentioned in the following discussion, The bands associated with the methyl stretching motions in the 2800-3050 cm-1 region are decidedly different in the infrared and the Raman spectra and also between the spectra of the ligand, of the borane complex and of the boron trifluoride complex. The C-H1 stretching frequencies are extremely broad and low in intensity in the solid infrared spectrum. However, in the Raman effect a medium strong doublet is observed at 2980 cm- and 3044 cm- The lower band correlates very well with the asymmetric C-H stretching vibrations of the ligand at 2998 cm-1 and of the borane complex at 3011 cm-1, but the assignment for the 3044 cm'1 band is more doubtful. The frequency of the latter band coincides exactly with a frequently observed weak mercury emission line of the source,'however, its surprising intensity in the present spectra leads to the conclusion that this band is due to an asymmetric C-H stretching fundamental rather than the mercury line. The very weak infrared band at 2907 cm (13) is absent in the Raman effect and is also probably a methyl stretching fundamental, The characteristic symmetric C-H stretching vibration, found near 2815 cm1 in similar dimethylamino-compounds, is clearly missing in both the

110 (CH )2N PF 2BIO0 (CH3)2NPF2B"1F3 * *Hg 1000 2000 3000 CM" FIGURE 22 - THE RAMAN SPECTRA OF (Ci)2NPF2B1 F AND (CH3)2NPF2B1 FO (-780C) 1F (-78 C)

TABLE 14 A COMPARISON OF OBSERVED INFRARED AND RAMAN FREQUENCIES OF (CH ) NB1F3, (CH3)2NPF2 AND (CH3) NPF2B1F3 (C3)3 3 2B23I2' (CH3)3NB1 F3(24) Infrared Raman (CH3)2NPF2(liquid) Inf rare (56) Raman (CH3) 2NPF2Bll F3 Infrared(13) Raman Assignment and Symmetry 3Q33 w 3013 m 3030 s 3043 ms asym. C-H stretch (a") asymn C-H stretch (a") 2998 s,sh,dp 2978 ms 2971 sh 2954 s 2924 vw 2967 s 2945 vs,dp 2908 s 2912 vs,p? 2908 vw,br asym. C-H stretch (a')?.-1 h-j H - 2890 w 2889 m 2868 vw 2861 s,p 2871 vw? asym" C-H stretch (a')? 2845 vw 2807 vw 2802 sh 2808 s 1691 w 2814 s,p 1506 vw 1496 vw 1486 w 1487 m 1484 s 1487 m,dp 417+1101=1518 650+850=1500 asym. CH deformation 3(a") asym. CH deformation 3(a) 1480 w 1474 w 1474 w, sh

TABLE 14 - CONT'D. (CH ) 3NB11 F3 ( 24 (CH3) NPF2(liquid) Infrared Raman Infrared(56) Raman 1469 w 1473 vs 1468 w 1454 w (CH3) 2NPF2B11F3 Infrared (13) Raman 1465 vw 1456 vw Assignment and Symmetry asym. CH deformation 3(a') 561+907=1468 asym. CH deformation 3(a') 1450 s 1439 s,dp 1445 vw 1445 m 1431 vw 14.24- vw 1414 w,sh,dp? 1414 mw,br 561+880=1441 417+1008=1425 1403 vw,sh?sym. CH deformation3(a' or at")? 1412 vw 1413 vw 1308 vs 1306 m, p 1271 vw 1255 w 1253 w 1208 w H 1-\) rol 1235 w 1200 mw 1175 sh 1178 vvw1182 vs 1171 ms,br 1151 vs CH3 rock (a")? CH rock or wag (af 3and/or a")? CH3 rock (at and/or a")or C-N stretch (a")? asym. B-F stretch (a' and/or a") 2 x 561 = 1122 CH3 wag (a' and/or 3a" )? 1142 sbr 1121 mw 1101 mw 1104 m 1105 w 1103 vw 1068 s 1071 w,dp? 1017 vw

TABLE 14 - CONT'D. (CH3 )3NB1F3 (24) (CH3) 2NPF2(:liquid) Infrared Raman Infrared_(56) Raman 991 w 988 s 841 s 842 s 990 vs 989 m,dp? (CH3)2NPF2B1F3 Infrare d(13) Raman Assignment and Symmetry 932 s 935 w, sh 800 vs 743 vs 792 m,p? 743 m,dp? 705 vs,p 1008. w 921 vw,sh 907 m 880 s 850 s 827 s 692 s 650 ms 1003 w 879 w,sh? 852 m 827 w,sh 690 vw 653 ms 551 vw, 509 vvw? 705 s C2 —N-P stretch (a') B(10)-F stretch (a') B(ll)-F stretch (a') P-F stretch (a") P-F stretch (a') 2 x 417 = 834? N-P-F2 stretch (a') B-N stretch (a') BF3 deformation (a') ske e tal? PF2 deformation (a') 2 692 m. 694 s 640 vw 547 w. 545 vw 519 vw,sh 513 vw 1-C HHj 561 m 495 w 473 mw 464 vvw? 440 vw ske le ta 1 430 w 428 vvw 372 vw 393 vw 417 vw C2N deformation (a')? 340 w 320 w 336 w 168 vvw

TABLE 15 THE OBSERVED INFRARED AND RAMAN VIBRATIONAL FREQUENCIES OF (CH3 )NPF2B1F3 AND (CH3 )2NPF2B 1F3 (CH3) 2NPF2B10F3 Infrared(.13) Raman 3044 w 2981 ms 2965 m.,sh? 2948 w,sh? 2907 vw 2873 vw? (CH3 )2NPF2BllF3 Infrared(13) Raman 3043 ms 2978 ms Assignment and Symmetry asyrnm C-H stretch (a") asym, C-H stretch (a") 2908 vw,br 2871 vw? 1507 vw 1497 vw 1485 w 1473 w 1466 vw 1457 vw 1474 m,sh 1506 vw 1496 vw 1486 w 1474 w 1465 vw 1456 vw 1445 vw 1431 vw 1424 vw 1414 mw,br 1235 w 1445 m 1410 vw, sh? 1474 w,sh 1445 rm 1403 Vw^, sh? asym, C-H stretch (a')? asym, C-H stretch (a,')? 417 + 1101 = 1518 650 + 850 = 1500 asym. CH3 deformation (a") asym. CH3 deformation (a") asym, CH3 deformation (a') 561 + 907 = 1468 asym, CH3 deformation (a') 561 + 880 =1441 417 + 1008 = 1425 sym, CH3 deformation (a' or a")? CH3 rock (a")? 562 + 654 = 1216? CH3 rock or wag (a' and/or a")? 3 H 4-P 1242 m 1216 m 1201 s 1200 vw

(CH3 )2NPF2B10F3 Infrared(13) Raan 1189 s TABLE 15 - CONT'D. (CH3 ) 2NPF2B11F3 Infrared(13) Raman 1171 ms,br 1162 rn 1152 mw, sh 1151 vs 1010 w 921 s 908 w 884 s 853 s 830 s 699 s 654 m 562 ms 511 vw 480 m 445 vw 416 vw 1006 w 878 w,sh 852 m 827 w,sh 698 vw 656 ms 605 vvw? 558 vw 509 vvw? 1121 mw 1101 mw 1008 w 921 vw,sh 907 m 880 s 850 s 827 s 692 s 650 ms 561 m 1003 w 879 w,sh? 852 m 827 w,sh 690 vw 653 ms 551 vw 509 vvw? Assigrnment and Synmetry CH3 rock (a' and/or a") or C-N stretch (a") asym, B(10)-F stretch (a' and/ or a ) asym. B11)-F stretch (a' and/ or a ) 2 x 562 = 1122 CH3 wag (a' and/or a")? C2 —N-P stretch (a') B(10)-F stretch (a') B(11)-F stretch (a') P-F stretch (a") P-F stretch (a') 2 x 417 = 834? N-P-F2 stretch (a') B-N stretch (a') BF3 deformation (a') skeletal? PF2 deformation (a') skeletal C2N deformation (a')? H uJ 473 mw 440 vw 417 vw

infi:r;are;d an..-id Raman spectra of the present boron trifluoride complex, The vibrations associated with the methyl defom-ation modes show no peculiarities except for the fact that the totally symmetric methyl deformation vibration near 1315 cml1 in the spectra of (CII3)2NPF, and (CIH3) NPF BH is absent in the present spectra. The assignment of the 1235-1242 cmr mrediuml infrared band to the symmetric methyl deformation mode by Sr. Fleming (13) is probably incorrect, and is better assigned to a methyl rocking fundamental in view of a similar vibration found near 1255 cm 1 in the spectrum of (CHi3)3NBF3 (24). Several infrared bands are observed in the 1150-1200 cmn region for the different isotopic samples, but are absent in the Raman effect (see Table 15). The previous assignment of the asymmetric B(11)-F stretch to the 1152 -l cm band appears to be correct (13). However, the assignment of the boron-10 fundamental to the 1189 cm-1 band in the spectrum of (CH3)2NPF2B F3 may be more correctly assigned to the, 1162 cm1 band in view of a similar vibration of (CH3)3NB1OF3 at 1165 cnm1 (24). The remaining band at 1171 cm" or 1189 cm -l is probably a methyl rocking vibration or the asyimmretric C-N stretch by analogy to the observations for (CH3)2NPF2 and (CH3)2NPF2BH3, with no preference being given to either assignment, Vibrational assignments to bands appearing in the 800900 cm1 region also bear explanation. In this region, three

117 infrared and Raman bands-.are observed;.at: 827 cm, 852 ccm and 880 cm, and their frequencies are not affected by isotopic substitution of the boron atom. Previously, the 880 cm-1 band was assigned to the asymmetric P-F stretching vibration and 827 cm 1 850 cm-1 bands were considered to be a Fermi doublet, with the lower band being assigned to the symmetric P-F stretch (13). The present work suggests that the assignment to the 880 cm-1 band is correct, but the symmetric -1 P-F fundamental may be better assigned to the 850 cm band because of its stronger intensity in the Raman effect and because the fundamental is also observed in the spectrum of the borane complex at 852 cm-1 The 827 cm-1 vibration may be an overtone of the 417 cm1 bando Although the previous assignment of the B-N stretch and the N-P-F2 group frequency to the respective infrared bands at 654 cm-1 and 690-699 cm-1 seem to be correct (13), further explanation is desirable. The N-P-F2 group motion is assigned to the higher frequency because a similar vibration is observed in the spectrum of the ligand at 705 cm1 It is interesting to note that the infrared intensity of this fundamental is about the same as that found in the spectrum of the ligand, but is considerably diminished in the infrared spectrum of the borane complex. The identical vibration in the Raman effect appears much weaker in the boron trifluoride and borane complexes than in the ligand. Even though the B-N stretching vibration of (CH3)3NBF3 is also observed near 690-700 cm 1 (24), it is

118 believed to shift to 653 cl in the present complex. The 561 cm-1 band, observed in both the infrared and Raman effect, is reassigned to the symmetric BF3 deformation rather than the asyizmetric vibration on the grounds that the symmetric mode is found at 545 cm-1 in the spectrum of (CH3)3NBF3 (24). The infrared band at 416 cm 1 is assigned to the C2-N deformation mode by analogy to the 393 cm vibration of the ligand. The remainder of the absorptions appearing in the spectra of the boron-10 and 11 trifluoride complexes do not require discussion since their relationships to the vibrations of the free ligand are quite evident. The uncertainties of these assignments are considered to be the same as those of the free ligand, Bis (dimethylamino) fluoproplpsphine,- [ (CH3N ):2N ]. PF,. No experimental data are available bearing on the structure of this compound although chemical arguments favor a bond configuration around the phosphorus similar to that in (CH3)2NPF2. The structure of [(CH3)2N]2PF may be one of no symmetry or, at most, the molecule may have a plane of symmetry and belong to the Cs point group. However, even if this molecule were assumed to have C symmetry, the orientation of the dimethylamine groups around the central phosphorus atom would remain unknown, There are several difficulties in interpreting the vibrational spectrum of this compound. This molecule has a

119 total of fifty-four Raman and infrared active vibrations which may be subdivided into thirty-six methyl group vibrations and eighteen skeletal modes. Also, many of the vibrations, especially those of the methyl groups, may be expected to be accidentally degenerate. In addition, many fundamentals of the same symmetry are found in the same region of the spectrum, thereby increasing the probability of interaction or coupling between various groups of the molecule. In any case, detailed assignments to numbered vibrations, such as those made for dimethylaminodifluorophosphine, have much less meaning here and therefore no attempt will be made to present a systematic numbering and description of the vibrational modes. However, it is possible to give assignments to the more characteristic group frequencies and to define the symmetry of the vibrations in terms of the local symmetry of the group. The observed Raman spectra of [(CH3)2N]2PF in the liquid (-45~C) and solid (-180~C) states are shown in Figure 23 and the supplementary polarization spectra of the liquid are represented in Figure 24. An inspection of these tracings reveals several interesting features, As is true for most of the aminophosphines in this work, it is very difficult to prepare a sample of [(CH3)2N]2PF which is completely devoid of fluorescent impurities and minute suspended particles. These impurities have the general effect of increasing the background of the Raman spectra, Fluo

120 rescence and increased light scattering from the crystal faces is particularly evident in the spectrum of the solid sample while the Tyndall scattering, seen in the liquid spectra, was kept to a minimum in Figure 23 by the use of an interference filter rejection unit. In general, the spectrum of the disubstituted ligand (Figure 23) closely resembles that of (CH3)2NPF2 (Figure 14) and the noticeable differences and similarities will be discussed later. The observed Raman frequencies and the reported infrared values (13) along with their present assignments are tabulated in Table 16. The observed vibrational bands of the methyl stretching and deformation modes closely resemble those of other dimethylaminophosphines and no peculiarities are observed in these regions of the spectrum. However, it is interesting to note that at least five Raman bands are clearly resolved in the methyl stretching region while the infrared spectrum shows only the absorption envelope, On the other hand, the infrared spectrum of the methyl deformation region shows much more structure than does the Raman effect, perhaps due to the appearance of more overtones. Because the bands overlap extensively in these regions, their assignments are somewhat arbitrary and they may be modified when deemed necessary, Reference may be made to Table 16 for specific vibrational assignments in these regions, Several bands are observed in the methyl rocking region from 1100 cm1 to 1250 cm-1 The very weak band at 1257 cm1

121 SOLID (-180 C) LIQUID (-45~C) * *Hg 0 1000 2000 3000 CM-1 FIGURE 23 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (-45~C) [(CH3)2N]2PF

122 * E (1) E(Il) *Hg 0 1000 2000 3000 CMFIGURE 24 - THE POLARIZED RAMAN SPECTRA OF LIQUID [(CH3)2N]2PF (-45 C)

123 TABLE 16 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF [(CH3 )2N]PF Infrared(13) Raman Assignment gas liquid(-45~C) solid(-180~C) 2992 s,sh,dp 2994 vvw out of phase asym, C-H stretch 2925 vs,p 2936 w out of phase sym, C-H stretch 2882 s 2885 vs,p 2902 w in phase asym. C-H stretch? 2842 sh 2837 s,sh,p 2837 w in phase asym, C-H stretch? 2795 m 2790 s,sh,p 2789 w in phase sym, C-H stretch 2542 vvw? 2503 vw? 1504 w 1494 w 1485 w 1479 w 1479 s,sh,dp 1491 vw,sh out of phase asym, CH3 deformation 1469 mw,br 1463 m out of phase asym. CH3 deformation 1453 mw 1448 w 1434 vvw 1432 s,dp 1438 w in phase asym. CH3 deformation 1402 m,sh,dp 1401 vw,sh in phase asym, CH3 deformation

124 TABLE 16 - CONT'D. Infrared(13) gas 1280 m Raman liquid(-45~C) solid(-180~C) 1284 m,p 1304+10 vvw 1257 w,sh,dp? 1242 vvw Assignment 1199 m 1164 mw, sh 1195 vw 1141 vw 1209 vvw 1147 vvw 1098 vw 1057 mw 985975 s,sh 949 vs 745 s 677 s 496 w, br 403 w, br 1059 m 980 ms 944 m,sh 817 vvw? 716 m,sh,p? 689 s,sh,p 644 s,p 612 vvw? 497 w,sh? 486 w 408; vw 386 vw 329 w,br 165 vvw? 1064 vw 991 vw 942 vw 702 ms 672 m 490 vvw in phase sym. CH3 deformation asym. CHI rock (out of C3N-C plane) CH3 rock (out of C-N-C plane) or asym, C-N stretch? CHl rock (in or ou; of C-N-C plane) CHl rock or asym. -N stretch? asym, C-N stretch out of ph se (C2-N)2-P stretch in phase (C2N)2-P stretch P-F stretch asym, N2-P-F stretch sym, N2-P-F stretch P-F rock or wag C2-N deformation C2-N deformation? skeletal skeletal

125 is probably a methyl rocking vibration. However, the methyl rocking assignment to the 1195 cm 1 band is questionable in view of the present uncertainty connected with the location of the asymmetric C-N stretching frequency in dimethylamine derivatives, In addition, the medium weak shoulder at 1164 -1 cm in the infrared and the very weak Raman line at 1141 -l cm are probably methyl rocking vibrations on the basis of similar bands found for (CH3)2NH (22) and (CH3)2NPC12 (23), The assignment for the very weak Raman line at 1098 cm'1 is uncertain but it may be a methyl rock or less likely a component of the asymmetric C-N stretch. The tentatively assigned asymmetric C-N stretching motion in (CH3)2NPF2 shifts from 1071 cm 1 to 1059 cm 1 in the disubstituted compound. The Raman band at 989 cm1, C which was attributed to the N-P group stretch in C (CH3)2NPF2,clearly splits in the Raman spectrum of [(CH3)2N]2PF yielding a doublet at 944 cm 1 and at 980 cm This doublet is also observed as a strong band at 949 cm1 and shoulder at 973-985 cm 1 in the infrared spectrum. It is interesting to note that Evleth, Freeman and Wagner (64) also report a very strong infrared doublet at 958 cm-1 and 978 cm1 which is characteristic of the C2-N-P group in [(CH3)2N]2PC1. It is very probable that this doublet arises from the in and out of phase C2-N-P group motions in the C C CN-P-N skeleton of the moleculeo C7 C Sr. Fleming observed a strong infrared band around 745 cm-1 which she attributed to the P-F stretch (13). No such

126 band at this frequency is observed in the Raman effect, but a medium intensity shoulder did appear at 716 cm on the high side of a strong doublet at 664 cm 1 and 689 cm 1 in the liquid spectrum. Although this frequency is somewhat lower than the expected P-F stretching range, 750 cm l-900 cm, reported by Bellamy (51), this band shows the same downward shift in frequency as was observed for PF3 and (CH3)2NPF2 when the sample was condensed to a liquid or a solid, -1 A strong band at 677 cm- is observed in the infrared spectrum and is assigned'to the N-..P-F group vibration. However, in the Raman spectrum this band appears clearly as a polarized, strong doublet at 664 cm"l and 689 cm 1.This splitting is most likely the result of symmetric and asymmetric motions of the N2-P-F group. Similar observations are reported for (CH3)2NPOC12 and [(CH3)2N]2POC1 where the N-P-C12 frequency at 723 cm 1 in (CH3)2NPOC12 splits to 672 cm-1 and 755 cm-1 in the monochloro- compound (23). Since this frequency does not shift appreciably from compound to compound, the motion is probably localized to a large extent in the N-P bondso A broad, weak absorption band is reported at 496 cm1 in the infrared spectrum (13). Similar bands, appearing in the liquid and solid Raman spectra, are found at 486 cm, 497 cm-1 and 490 cm-1 respectively. Since the absorption occurs at virtually the same frequency as the P-F2 deformation in (CH3)2NPF2, it seems logical to assume that this

127 absorption is the result of the P-F rocking or wagging motion, a remnant of the PF2 deformation in the difluoro-compound, A broad, weak infrared band at 403 cm 1 is also seen in the Raman effect as a very weak band at 408 cm-1 and has been assigned by Sr. Fleming to a (CH3)2N- deformation mode (13)o The assignment of this band is consistent with the findings of Stewart in which he observed that the corresponding vibration occurred at 427 + 14 cm 1 for a large number of dialkylamines (52)o A lower band at 386 cm-1 is also observed in the Raman spectrum and no doubt is a skeletal motion. Due to its closeness to the 408 band it might also be assigned to a (CH3)2N- deformation, but this postulate is only a conjecture, The weak Raman band at 329 cm1 and the very weak band at 165 cm-1 are probably skeletal vibrations, however, no attempt has been made to characterize them, Bis (dimethylamino)fluorophosphin6-Borane, [ (CH3 )2N]2PFB11 The structure of [(CH3)2N]2PFBH3 is unknown; however, nuclear magnetic resonance data have indicated that the borane group is bonded to the phosphorus atom in this complex (13)o Since no attempt was made to discuss the vibrations of the ligand in relation to possible symmetry elements of the molecule, the same procedure will be used in the present discussion, If this molecule has no symmetry, all sixty-six of its fundamental vibrations may be observed in the infrared and Raman effect. Of these vibrations thirty-six involve the

128 motions of the methyl groups, nine are localized primarily in the borane group and the remaining twenty-one vibrations are predominantly skeletal vibrations of the molecule. As was seen for (CH3)2NPF2 and its borane adduct, the Raman and infrared spectra of [(CH3)2N]2PFBH3 closely resemble that of the ligand, [(CH3)2N]2PF. The observed Raman spectra of liquid and solid samples are represented in Figure 25, while the polarized spectra are shown in Figure 26. Similar to the observations of [(CH3)2N]2PF, the Raman spectrum of liquid [(CH3)2N]2PFBH3 contains an abnormally high background due to Tyndall scattering within the sample. The observed Raman frequencies, band intensities, polarization data and assignments are listed in Table 17 along with the reported infrared data (13). In view of the similarities in the spectra of the ligand and the complex, a detailed discussion of the methyl frequencies is omitted here and reference may be made to preceeding sections for further detailso A broad infrared band of medium intensity is observed at 1194-1167 cm-1 but does not appear in the Raman spectrum. The assignment of this band to a methyl rocking motion by Sr, Fleming (13) is supported by the observation that rocking motions oftentimes exhibit bands which are much weaker in the Raman effect than in the infrared. Although this vibration may be correlated with the observed methyl rocking vibrations of the ligand at 1164 cm-1 and 1199 cm 1, the correctness of the present

129 0 1000 2000 3000 CMIFIGURE 25 - THE RAMAN SPECTRA OF SOLID (-180~C) AND LIQUID (OOC) [ (CH3 )N] 2PFB lH 3'- 2 3

130 * E(1) E(11) *Hg O 1000 2000 3000 CMr FIGURE 26 - THE POLARIZED RAMAN SPECTRA OF LIQUID [(CH ).N]2 PFB lH3 (00C)

Infrar liquid 2991 s 2924 r 2855 s 2810 w 2389 s 131 TABLE 17 THE OBSERVED INFRARED AND RAMAN FREQUENCIES OF [(CH3 )2N]2PFB11H3'ed (13) Raman Assignr 1 (25 C) liquid(0~C) Solid(-180~C) ih 3006 m,sh,dp 3001 w,sh out of phc C-H str( i 2945 s,p 2946 m out of phi C-H strf 2905 s,sh,p 2904 m,sh in phase E C-H str<,h 2854 m,sh,p 2856 w,sh in phase C-H strE r 2810 m,sh,p 2807 w,sh in phase' C-H str( 2387 s,sh, dp 2379 s,sh asym. B-H 2343 vs,p 2343 vs symo B-H 2286 vw 977 + 131: Mw 2249 vw 2246 vw 2 x 1131: nent ase asymo etch ase symn etch asymo etch asymo etch sym, etch stretch stretch L = 2288 =2262 31 = 2191 = 2136 ase asymo ormation ase asymo ormation asymo ormation asymo ormation symo ormation 2249 v 2199 vw 2134 vvw 14891475 br,sh 1460 m 1416 vvw 1310 ms 2137 vvw 1481 m,sh,dp 1441 m,dp 1411 w,sh,dp 1305 vvw 1493 1467 1437 1408 1311 w,sh w, sh? m vw, sh vw 1060 + 11l 2 x 1068 = out of phc CH3 def( out of phc CH3 def( in phase CH3 deft in phase c CH3 def in phas e E CH3 deft 3

132 TABLE 17 - CONT Do Infrared (13) liquid (250C) liquid(00o 1283 vvw 11941167 m,br 1134 w Raman ) Solid(-180~C) 1280 vvw 1126 w 1064 vw 1000 vw 977 vw 1131 w,dp Assignment 1068 m 998 1060 999 w,dp? vw? 973 s 969 vw? 821 s 775 ms 750 s 706 mw 685 w 811 773 vvw? vvw 809 771 vvw vvw CH3 rock (in C-N-C plane ) CH3 rock (in and/ or out of C-N-C plane) asym. BHI deformation asymo C-N stretch out of phase (C2-N)2-P stretch in phase (C2-N)2-P stretch BH3 rock BH3 rock or skeletal P-F stretch asym. N2-P-F stretch symo N2-P-F stretch P-B stretch PF rock or C -N deformation skeletal skeletal skeletal skeletal 747 w,p 683 vvw? 743 vw 680 vvw 570 s 438 vvw 574 w 569 438 m,p w,p? 340 243 180 140 vw vw vvw vvw?

133 assignment is no better than the assignments of the free baseo Therefore, the assignment of this band to a methyl rock is questionable in view of the present uncertainty involved with the location of the asymmetric C-N stretching vibration in the ligand and in related dimethylamine derivatives. The borane stretching motions are easily identified in the Raman spectrum as a very strong, well resolved doublet at 2343 cm1 and 2387 cm- respectively. The lower band is most assuredly the symmetric stretching motion because it is strongly polarized while the upper band is depolarizedo Sr. Fleming (13) observed the asymmetric B-H stretching motion at 2389 cm1 in the infrared. However, her assignment of the symmetric stretch to a weak band at 2249 cm-1 does not appear correct in view of the Raman results. A weak depolarized band at 1131 cm, observed in both the infrared and Raman effect, is believed to be the asymmetric borane deformation vibration. Even though this band is found very near to a methyl rocking band at 1141 cm in the spectrum of the ligand, it is not assigned to the rocking mode because of its relatively strong intensity in the Raman effecto Three strong infrared bands appear at 750 cm, 775 cm, and 821 cm-1 while the respective bands are observed in the Raman effect as a weak, polarized line at 747 cm, a very weak line at 773 cm-1 and an extremely weak line at 811 cm-l Since the borane rocking and P-F stretching motions are expected in this region (50,53), a choice must be made con

134 cerning these assignmentso Although. Sr Fleming has assigned the P-F stretch to the 821 cm band (13), it is believed that this vibration is better assigned to the 747 cm 1 band,because the latter is definitely polarized and has a somewhat stronger intensity than the others in the Raman effect. In addition, this frequency correlates reasonably well with the corresponding Raman frequency of the ligand at 716 cm-1, The highest band at 821 cm-1 or 811 cm-1 is most likely a borane rocking vibration and this assignment is substantiated by the fact that no similar band is observed in the spectrum of the ligand. The band at 773 cm1 is identified with less certainty but is probably a skeletal vibration or borane rock, perhaps a motion out of the hypothetical symmetry plane of the molecule, Many of the skeletal motions of the borane complex may be identified by their analogous vibrations in the free ligando The vibrational assignment to the depolarized, weak to medium intensity band around 1060 cml is such an example. Since a similar vibration is found at 1059 cm in the spectrum of the ligand,which was tentatively assigned to the asymmetric CAN stretching motion, the present band is also thought to be the C-N stretching fundamental, It should be noted that any uncertainty in the original assignment would certainly be carried over to the present situation~ In another case, the in and out of phase (C2-N)2~P group stretching motion, observed as a medium intensity doublet

135 at 944 cm and 980 cm 1 in the Raman spectrum of the ligand, also is clearly observed in the solid Raman spectrum of the -1 -1 complex at 977 cm and 1000 cm respectively. However, these motions remain unresolved in the infrared spectra,both of the ligand and its borane complex, The symmetric and asymmetric N2-P-F coupled motions of the complex are more difficult to observe in the Raman effect, because their band intensities are so weak. The asymmetric motion is probably the medium weak infrared band at 706 cm1 (13), but no such band is observed in the Raman effect, The corresponding symmetric motion is probably the weak, infrared band at 685 cm 1 and the very weak Raman band at 683 cm o The intensities of these Raman bands decrease markedly upon borane complexation, as was observed in the infrared, and their frequencies shift from 664 cm-1 and 689 cm-1 in the ligand to 683 cm-1 and 706 cm-1 in the complex. The weak infrared band at 574 cm and the polarized, medium intensity Raman band at 569 cm-1 are probably the P-B stretching vibration. This assignment correlates well with the similar vibration in F3PBH3 at 607 cm and in (CH3)2NPF2BH3 at 591 cm-1 The weak, slightly polarized Raman line at 438 cm1 is believed to be a P-F rocking vibration or a (CH3)2-N deformation. Since the respective vibrations in the ligand are probably those found near 490 cm and 408 cm-1 one is not able to make a more definite assignment to this frequency,

136 Four very weak Raman lines which appear at 340 cm-1, -1 1 -1 243 cm, 180 cm and 140 cm1 are probably skeletal vibrations, but more specific assignments are impossible because of a general lack of knowledge concerning many of the deformations and torsional modes of the molecule, Trimethylphosphine, (CH3)3P. Trimethylphosphine has been shown by electron diffraction and microwave studies to have a pyramidal structure (33,65,66). It is assumed in this study that it has real or effective C3 symmetry in that the methyl groups are in their most symmetrical orientation or are rotating sufficiently fast to have an effective cylindrical symmetry. This assumption is consistent with the spectroscopic data reported in the literature (29,30,31,32), although the question of whether the symmetry is C3v or only C3 has never been investigated critically, The C symmetry requires that trimethylphosphine have twenty-two fundamental vibrations with the symmetry distribution 7A1 + 4A2 + 11E. Of the eighteen active vibrations, only four are skeletal vibrations (two in the A1 class and two in the E class) while the remaining fourteen are methyl group vibrations. The numbering used for the fundamentals and the descriptive notation are listed in Table 18, The Raman spectra of liquid (-40~C) and solid (-141~C) trimethylphosphine are shown in Figures 27 and 28. The observed vibrational frequencies and their assignments are

137 given in Table 19. The observed frequencies agree, within experimental error, with those reported in the literature (29,30,31,32), and their assignments for the most part are identical to those of the earlier workers. Variations in the assignments from the previous to the present work will be cited later. The spectroscopic results of this investigation are compared to those of other workers in Table 20. Since the bands associated with the methyl group vibrations are well separated from the skeletal frequencies, the two groups are relatively easy to identify. The very strong bands at 2958 cm 1 and 2891 cm-1 are assigned to asymmetric and symmetric C-H stretching vibrations on the basis of intensity and polarization measurements. Although six fundamentals must be present in these two bands, it is not possible to resolve them. However, it is observed that the asymmetric and symmetric methyl deformation bands at 1417 cm 1 and 1288 cm 1 respectively are split in the solid Raman spectrum due to the increased line sharpness at low temperatures. The methyl rocking and wagging vibrations are tentatively assigned to the bands appearing at 940 cm1 and 823 cm- respectively. These assignments are made on the basis that the higher band is weakly polarized while the lower band is depolarized. In addition neither of the bands can be satisfactorily described as an overtone or combination, Although Siebert (31) and Halmann (32) assign somewhat higher frequencies to these vibrations, it is felt that these assign

138 SOLID -141 *C) * LIQUID t- 40'C) 3000 CM' I 200 O SOLID (-1410C) AND FIGURE 27 - THE RAMAN SPECTRA OF SOLID - ) AND LIQUID (-40 C ) (CH3) 3P

(ooI0-) Od( HO) aInbI do vs Iassg NvwvY aGZIvoa0d Hi - 92.Hnoiz,-W O 000O 00O 0001 0 -- -- - -u"10 I I I I I OHO (11)3 (T)3 n vIA i r I t 11/1711 I I I i i i 8 I m 6ZT

140 TABLE 18 THE FUNDAMENTAL VIBRATIONS OF (CH3)3P. C3 SYMMETRY Symmetry Number Description A1 1 in phase asymmetric C-H stretch 2 in phase symmetric C-H stretch 3 in phase asymmetric CH3 deformation 4 in phase symmetric CH3 deformation 5 CH3 rock 6 C-P stretch 7 C3P deformation 3 A2 8 9 10 11 out of phase asymmetric C-H stretch out of phase asymmetric CH3 deformation CH3 rock CH3 torsion E 12 13 14 15 16 17 18 19 20 21 22 out of phase asymmetric C-H stretch out of phase asymmetric C-H stretch out of phase symmetric C-H stretch out of phase asymmetric CH3 deformation out of phase asymmetric CH3 deformation out of phase symmetric CH3 deformation CH3 rock CH3 wag C-P stretch C P deformation CH3 torsion

141 TABLE 19 THE OBSERVED RAMAN VIBRATIONAL FREQI LIQUID AND SOLID (CH3)3P (-40~C) Solid (-141~C) Liquid 3155 2958 2891 2841 2808 2749 2537 1600 - vvw vs,dp vs,p w, shp w, sh,p vvw vvw vvw ~VVWI 14i7 sdp 1288 vw p 940 w,p 900 vvw,sh? 823 vw,dp 705 vs,dp 650 vs,p 591 vvw? 549 vvw? 297 w,sh,p 258 m,dp 210 vw,sh,dp 2959 2897 2856 2806 2538 1444 1427 1407 1292 1272 943 900 827 703 652 368 307 272 212 159 136 vs vs w, sh w, sh vvw w, sh m w, sh vw vvw, sh vvw? vvw m m vvw? vw vw vvw vvw? vvw? JENCIES OF As s ignment V2 + v21? or w22 + w12? v1 V12, V13 V2 v 14 2 V15 or 2 v16 2 V3 V22 + 2 w17 2 v17 w3 + v22? v15- vV16 3 V4 V17 V5, V18? V5, V6 + V21? V19 020 V6 2 v7 2 V21 V7 V21 V22

TABLE 20 A COMPARISON OF THE OBSERVED VIBRATIONAL FREQUENCIES OF TRIMETHYLPHOSPHINE AND RELATED COMPOUNDS (CM ) (CH 3)3P IR(g) R) Raman() R aman) Rama( s ) (32) (29.,31) (CH3)3P0(36) IR Raman (CH3 )3N(22) IR(g) A1 Species vw C-H asym. stretch v2 C-H sym. stretch v3 CH3 asymo deformation ^4 GCH3 sym, deformation gV5 CH3 rock 5 3'6 C-P (C-N) stretch V7 C3P (C3N) deformation A2 Species "8-"ll Forbidden in Infrared E Species V12 w13 C-H asym. stretch v14 C-H sym, stretch 2970 2850 1417 1310 960 652 2969 2894 1421 1312 973 653 263 2958 2891 1288 940 650 297 2959 2897 1406 1292 943 652 307 2999 2923 1415 1340? 1340 872 2967 2777 1466 1402 1183 826 365* ro r\) 671 256 and Raman Spectra 2970 2920 CH3 asymo deformation 1430 2969 2954 1421 1293 1072? 2958 2891 1417 1288 940 2959 2897 1427 1406? 12929 1272 943 2999 2923 2967 2822 V17 CH3 sym. deformation 1298 1437 1420 1305 1292 950 1402 V18 CH3 rock 3 1067 937 1272

TABLE 20 - CONT'Do (CH3 )3P IR(g) Raman (32) (29,31), Raman ( )+ Raman(s)+ (CH3)3P0(36) (CH3)3N(22) IR Raman IR(g) "19 CH3 wag 947 715 707 V20 v21 v22 C-P (C-N) stretch C3P (C3N) deformation CH3 torsion 948 708 305 823 705 258 210 827 703 272 212 866 866 750 756 311 (-275) 1104 1043 425* 264* + This work Values obtained from Raman data w

144 ments are more consistent with the methyl rocking and wagging frequencies reported in the literature for (CH3)3P0 (36), (CH3)2PH (67), and (CH3)3SiCl (68). The remaining methyl group vibration, the degenerate methyl torsion, is assigned to a weak shoulder at 210 cm 1 in the Raman spectrum, This frequency is in good agreement with the value previously reported for (CH3)3P at 223 cm1 (66) and is consistent with the values reported for (CH3)3N at 264 cm-1 (22) and for (CH3)2PH at 236 cm-1 (67)o The four remaining bands in the Raman spectrum of (CH3)3P are due to the skeletal motions of the molecule: the asymmetric and symmetric C-P stretching vibrations and the asymmetric and symmetric C3P deformation vibrations. The components of the very strong doublet appearing at 705 cm1 and 650 cm'1 are assigned to the asymmetric and symmetric C-P stretching vibrations respectively. The two deformation modes are found at 297 cm 1 and 258 cm. Since the stronger, low frequency band at 258 cm is clearly depolarized while the weaker band appears to be weakly polarized, the higher frequency is assigned to the symmetric deformation and the lower to the asymmetric deformation mode. The presently observed skeletal frequencies agree very well, within experimental error, with those appearing in the literature (29,30,31.32) and the assignments for the C-P stretching motions are identical to those reported earlier (31,32), However, the assignments for the skeletal deformation modes are the reverse

145 of those reported by Siebert (31) based on the Raman data published by Rosenbaum, Rubin and Sandberg (29). Although this previous work included some polarization studies, no polarization data were obtained for the bands in question, The literature assignments were apparently based on analogy with (CH3)3N, where the symmetric deformation mode is lower. The present assignments for the C3P deformation modes are further substantiated by microwave studies (66) in which the deformation frequencies are determined from relative intensity measurements on the vibrational satellite lines in the microwave spectrum. It should be noted here that there is a suggestion of two very low lines at 159 cm-1 and 136 cm"1 in the solid Raman spectrum. However, since these are very weak in intensity and possibly are caused by lattice vibrations, they are neglected in the present discussion, Trimethylphosphine-Borane, (CH3 )3PB10H3 (CH3 )3PBH3, and.... 3.....3~ (CH,,,P,, B11D. Such compounds as PF3 and PC13 are known to retain their C3v symmetry upon formation of their respective oxides, F3PO and C13PO (42,43,44,46). Therefore, it is not illogical to assume that the trimethylphosphine skeleton retains its C3v symmetry with the formation of trimethylphosphine-oxide or the isoelectronic trimethylphosphine-borane, The species and qualitative description of the fundamental modes of vibration for trimethylphosphine-borane are

given in Table 21, It is noted that for a C3v configuration, this molecule must have thirty vibrational frequencies classified as 10A1 + 5A2 + 15E. Of these frequencies, the A1 and E are active in both the infrared and Raman effect, while those belonging to the A2 class are inactive in both, For convenience, as was done in the case of trimethylphosphine, the fundamentals are divided into three groups: the methyl group vibrations, the vibrations involving the borane group and the skeletal vibrations, Since in reality interaction between the skeletal, methyl and borane vibrations certainly exists, this model is only an approximate one. Three isotopic varieties of trimethylphosphine-borane, (CH3)3PB10H3, (CH3)3PBH3 and (CH3) PB11D3, were examined spectroscopically, Each of these species was studied in the infrared as a solid film at -180~C and the observed spectra are shown in Figures 29, 30 and 31. Since infrared studies involving the use of KBr pellets at room temperature yielded essentially the same spectra, they are not shown here, In addition, the Raman spectrum of each of these samples was studied at -10~C and -l45~C, Only the spectra at -10~C are shown in Figure 32, since the differences from the spectra at -1450C are slight. The observed Raman and infrared frequencies along with their tentative assignments are given in Tables 22, 23 and 24o A comparison of the observed fundamentals for the different isotopic varieties is given in Table 25,

147 4000 3000 2000 1500 I I I I I I I 1500 1000 500 CM-' FIGURE 29 - THE INFRARED SPECTRUM OF (CHL) PB10-i (-i80~C) PD H? — l^

148 4000 3000 2000 1500 1500 1000 500 CM FIGURE 30 - THE INFRARED SPECTRUM OF (CHi;) PBtI2 (-180~C)

149 4000 3000 2000 1500 I I I I I I I I II 1500 1000 500 CM-1 FIGURE 31 - THE INFRARED SPECTRUMI OF (CH;)2PBllD (-180~C)

150 o 1000 2000 3000 CM-I FIGURE 32 - THE RAMAN SPECTRA OF (CtI) PBI10HI, (CIHL)PBHi, (CII ) PB1D3 (-10oC)

TABLE 21 THE FUNDAMENTAL VIBRATIONS OF (CH3)3PBH3, C3v SYMMETRY Symmetry Number Description Al 1 in phase asymmetric C-H stretch 2 in phase symmetric C-H stretch 3 B-H stretch 4 in phase asymmetric CH3 deformation 5 in phase symmetric CH3 deformation 6 B-P stretch 7 BH3 deformation 8 CH3 rock 9 C-P stretch 10 C3P deformation A2 11 out of phase asymmetric C-H stretch 12 out of phase asymmetric CH3 deformation 13 CH3 rock 14 CH3 torsion 15 BH3 torsion (around molecular axis) E 16 out of phase asymmetric C-H stretch 17 out of phase asymmetric C-H stretch 18 out of phase symmetric C-H stretch 19 B-H stretch 20 out of phase asymmetric CH3 deformation 21 out of phase asymmetric CH3 deformation

152 TABLE 21 - CONT'D Symmetry E Number 22 23 24 25 26 27 28 29 30 Description out of phase asymmetric CH3 deformation BH3 deformation CH3 rock CH3 wag C-P stretch BH3 rock C3P deformation CH3 torsion C3P rock or C-P-B deformation

153 TABLE 22 THE OBSERVED RAMAN AND INFRARED FREQUENCIES OF (CH3 )3PB 1i3 Infrared -180~C Raman -145~C Assignment -15~C 3045 vvw 3042 3028 2988 s 2988 2915 s 2917 2860 vvw 2863 2814 vvw 2818 vvw vvw s s vvw 3052+10 2977+10 2898+10 vw w w w3 + w9? v1' w16' 717 V29 718 2w20,21 v10 + 2v26 2373 2338 2258 m ms vvw 2374 2339 2253 2141 vvw vvw 2788+10 2564+10 m 2360+5 s 2340+5 vvw 2258+5 2215+5 2142+5 2070 2040 1990 1838 1785 1655 1645 1633 1592 1517 1508 1466 1457 1435 m 1426 1383 w,sh? w s 2v VS V19 vs,sh w3 s,sh 223 w,sh v7 + v23 w,sh 2v7 or vg + v20,21 vvw V8 + V23 vw V5 + w26 vw v5 + v9 or w9 + w22 vw /7 + ~26 vw w7 + w9 or 2v27 w w8 9+vg or 20,21 + w30 W., sh. p 27 WSh V 426 +'27 vw,sh + VW5h 7W8 + w9? vw,sh w9 + V27 vw V5 + V29 vw 2v26 w V9 + V26 w,sh v6 + V27 s v20' w21 5s /4 VVW 1418 w 1421

154 TABLE 22 - CONT'Do Raman -150C Infrared -145~C -180~C 1338 w 1305 m, 1294 s, 1288 s 1175 vw 1137 w 1138 s 1082 s 1076 vvw? 1072 ms 970 vs 945 vw 948 vs 885 vvw 891 s Assignment 1134 vw sh sh,sh,sh V7 + V28 v22 v5 v10 + v27 V23 vw (B-10) v7 (B-ll) V24 V8 V27 V6 + v30? 6 + v29? 945 886 829 808 755 707 647 580 288 261 209 169 147 vvw vvw vvw? vvw? w vw vvw m vw vw vw vvw, sh vvw, sh 757 709 644 581 288 260 232 208 m w 786 762 757 714 701 vw s,sh? s s m sh m vvw s w w 583 V26 V9 V6 (B-10) vlo V10 v28 w30? V29 vw w

155 TABLE 23 THE OBSERVES RAMAN AND INFRARED FREQUENCIES Raman -100C -15( 2984 s 2987 2914 2854 2808? 2366 2335 2258 S vvw vvw ms S vw 2915 2855 2809? 2365 2336 2258 OF (CH3)3PBH3 Infrared 3~C -180~C 3057 vvw s 2983 w 2960 w,sh s 2915 w vvw 2851 vvw vvw 2803 vw 2561 w ms 2371 vs s 2328 vs vw 2257 s 2200 w 2148 vvw, 2126 w 2075 vvw 2046 vw 2000 vvw 1827 vw 1774 vw 1651 vw,s] 1633 w 1586 vw,s] 1516 vw 1507 vw 1465 w 1456 w 1432 s w 1425 s 1327 w, sh 1303 m,sh sh As s ignment h V3 + w9? V1 v 16, v17 W1. ~2' w18 2720,21 10 + 2v26 2v5 ^5 "19 v3 2v23 w7 + 323 2v7 or 29 + 20,21 74 + 79 w8 + 723 v5 + v26 W5 + w9 Or v9 + v22 77 + V26 W7 + w9 or 2.v27 V8 + w9 or V20,21+Y30 V8 + V9? v9 + 27 75 + w29 2w26 79 + /26 w6 + w27 v20' v21 74 77 + w28.22 1419 w 1420

156 TABLE 23 - CONT D. Raman -150~C -10~C Infrared -180~C As s ignment 1294 1134 vw 1136 vw 944 892 vvw vvw 948 889 vw vvw 1288 1172 1137 1082 1070 973 948 889 854 787 761 757 735 711 695 vs, sh vs m s m, sh ms vs vs s vvw, sh w s,sh? s vvw s, mnsh.5 10 + W27 w23 17 (B-10) v7 (B-11) w24 V8 V27 +6 + 10 V6 + V30? v6 + v29? v26 w9 808 vvw? 808 vvw, br', 757 w 705 vw 650 vw 570 w 286 vw 261 vw 288 vw,sh 211 vw 176 vvw,sh 145 vvw,sh 757 w 707 vw 647 571 287 259 230 204 169 146 vw 584 571 m, sh ms w vw vw vw VW VW vw vvw vvw V6 (B-10) V6 (B-11) w10 w28 v307 w29

157 TABLE 24 THE OBSERVED RAMAN AND INFRARED FREQUENCIES OF (CH3)3PB lD3.Raman -100C Infrared -180~C Assignment 3053 vvw? 3034 2986 2916 2344 1769 1713 1660 1622 1503 vvw? s 2986 s 2915 2860 ms ms vvw 2982+10 2903+10 vw ms ms w w w 2347 1769 1717 1662 1625 2801+10 vvw 2338+5 2040 1980 m 1774 w 1720 w 1669 w 1630 1432 1419 vw 1422 w 1422 1361 vvw? 1370 vvw? 1371 1301 1293 1283 986 934 vvw 942 vvw 949 873 851 820 vw 819 vvw? 830 785 749 w 753 w 757 706 vw VW vw W vvw VVW w vvw vvw vs m w m m, sh m vvw? m, sh ms m, sh ms s m ms vvw, sh vvw m w, sh v1 J 16, V v2' 18 2V20, 21 W10 + 2v26 v3 (B-H) 5 + 26 W5 + v9 or vl9 w3 7 + V25 o: g8 9+ 9? o: 2v26 p20, w21 V4 2v9'22 r 2V25? r w23 + v26 17 v9 + w22 v5 w24 V8 V23 I7 w25? V6 + w29? w26 w27

158 TABLE 24 - CONT'D. Raman -10C 685 w 645 vvw? --- - -- - -145~C 688 vw 641 vvt Infrared -180~C 696 m Ass ignment r? 562 527 vw w 521 282 256 209 188 144 101 m w w 523 283 255 223 w vw vw vw v9 2v10 V6 "10 v28 30? w29 vvw,s h?? vw 186 vw vw vvw, sh?

159 TABLE 25 FREQUENCY ASSIGNMENTS OF THE FUNDAMENTAL VIBRATIONS OF TRIMETHYLPHOSPHINE-BORANE Symmetry No. Notationa Class A1 1 v-CH 2 v-CH 3 v-BH 4 -CH3 5 6-CH3 6 v-BP 7 5-BH3 8 p-CH3 9 w-CP 10 5-C3P A2 11-15 Inactive (CH3 )PB1OH0 2988 2913 2339 1422 1288 582 1082 946 710 288 Assignmentb (CH3)3PBH3 2985 2915 2333 1421 1288 571 1070 947 708 287 (CH3)3PB D 3. 3 2986 2915 1717 1421 1283 524 851 942 690 283 E 16 v-CH 2988 17 w-CH 2988 18 v-CH 2913 19 w-BH 2371 20 6-CH3 1435 21 6-CH3 1435 22 5-CH3 1294 23 5-BH3 1136 24 p-CH3 970 25 (-CH3 829 26 w-CP 756 27 p-BH3 887 28 5-C3P 260 29'r-CH3 209 30 p-Me3P 232 aSymbols used: v = stretching, 5 c = wagging, T = torsion bAll frequencies in cm-l 2985 2985 2915 2367 1432 1432 1294 1136 973 757 890 260 205 229 = deformation, 2986 2986 2915 1771 1432 1432 1293 873 986 823 753 706 256 209 223 p = rocking,

160 The vibrational modes of (CH3)3PBH3 are assigned to the observed frequencies using the isotopic data and also by comparison with the assignments for (CH3)3P (discussed in the previous section) and those for (CH3)3P0 (36), (CH3)3N (22), (CH3)3NBH3 (69) and for other amine-boranes (50). The absorption bands associated with the methyl vibrations are easily identifiable by their band positions and their strong band intensities. Since the polycrystalline nature of the samples prevented polarization measurements in the present case, some uncertainty in the assignments must result. Nevertheless, tentative methyl stretching, deformation, rocking and torsional modes are assigned to the observed frequencies and these are shown in Tables 22 to 25. A detailed discussion of the methyl stretching and deformation frequencies is omitted since they are virtually the same as those reported earlier for (CH3)3P. The methyl rocking and wagging modes are not easily identified because their frequencies occur in the same region as the borane rocking vibration and the isotopic borane-d3 deformation vibrations. However, the methyl rocking and wagging vibrations are assigned to bands observed at 947 cm 1, 973 cm1 and 829 cm- respectively while the borane rocking mode is seen at 890 cm 1 and the borane-ll-d3 deformation modes are placed at 851 cm 1 and 873 cm l The methyl vibrations are distinguished from those of the borane group, because their frequencies are relatively insensitive to

161 isotopic substitution in the borane group. The remaining degenerate methyl torsional mode is found at 223 + 20 cm for (CH3)3P (66) and is assigned to a very weak band at 205 cm-1 in the borane complex. The basis for many of the assignments of the borane group is not immediately apparent and some discussion is necessary. The B-H symmetric and asymmetric vibrations, w3 and v19 respectively, appear as a very strong doublet in both the infrared and Raman spectra at about 2333 cm and 2367 cm 1 The strong band in the infrared at 2258 cm-1 appears to be a combination or an overtone, possibly 2v23o For the deuterated sample, four bands appear between 1600 cm 1 and 1800 cm1 in both the infrared and Raman spectra (see Figures 31 and 32), Because only two bands are expected in this region, the more complex spectrum must be due to the presence of overtones or perhaps combinationsO Since fundamentals normally are more intense than the combinations, the bands at 1717 cm 1 and 1771 cm 1 are assigned to the symmetric and asymmetric B-D stretching vibrations, w3 and v19, respectively. The isotope frequency ratios calculated on the basis of these assignments are l136 and lo34 lending additional support to the correctness of these assignments. The remaining two bands at 1662 cm 1 and 1625 cm-1 for the deuterated sample involve combinations or overtones, perhaps 7 +25 or 2v25 and 2V8 + v9 or v23 + v26 respectively; the fundamentals remain somewhat uncertain~ The borane defor

162 mation modes, V7 and v23, are observed at 1070 cm 1 and 1136 -1 -1 cm in the hydrogen species and shift to 851 cm and 873 cm1 in the deuterated sample. Lastly, the degenerate borane rocking vibration, v27, is observed at 890 cm 1 as a strong band in the infrared and a very weak band in the Raman and shifts to 706 cm-1 upon deuteration, The skeletal vibrations are perhaps the most interesting to study spectroscopically, since they are pertinent to the discussion of bonding and stability of the borane complex. The skeletal motions of the molecule involve only the motions of the C3P-B framework under the assumption that the methyl and borane groups behave as point masses. Such a molecular model has six vibrational modes, four of them originating in the trimethylphosphine ligand and two of them arising from the formation of the phosphorus-boron bond. Since the configuration of trimethylphosphine is not expected to change significantly upon complexation with the borane Lewis acid, the C-P stretching and C3P deformation frequencies of the complex are not expected to be very different from those of the free base. In reality the symmetric and asymmetric C-P stretching frequencies shift from 650 cm-1 and 705 cm-1 in the free base to 708 cm-1 and 757 cm 1 respectively in the complex. However the C3P deformation frequencies show almost no effect upon complexation, the symmetric and asymmetric frequencies moving respectively from 297 cm1 and 258 cm-1 to 287 cm-1 and 260 cm- The two remaining vi

163 brations are the symmetric P-B stretching motion and the degenerate C3P rocking or the C-P-B angle deformation motion. The P-B stretching frequency, v6, at 571 cm-1 in the natural borane compound is easy to recognize because of its relatively strong band intensity in the infrared and its large isotopic dependence. The assignment for the degenerate (CH3)3P- rocking vibration is not altogether certain, but its frequency is expected to be quite low. In fact Rice, Galiano and Lehmann (69) have assigned the 226 cm 1 frequency to the corresponding vibration for (CH3)3NBH3. Several low lying bands are found in the Raman spectrum of (CH3)3PBH3 and the very weak band at 229 cm 1 is tentatively assigned to the rocking mode by analogy to that of (CH3)3NBH30 Although all of the active vibrations are assigned frequencies, several bands are observed to which no assignments can be given. Of these, two or three very weak bands are observed below 200 cm.1 in the Raman spectrum. Because these lines are uncertain and are of dubious origin,: they are not used in the assignments. The very broad infrared band which appears from 3200 cm-1 to 3260 cm-1 in some of the figures is attributed to an 0-H band from the KBr windows. The assignments to the fundamental vibrations, for the most part, appear reasonably certain. However, the assignments of the overtones and combinations are much less certain and oftentimes several possibilities exist,

SUMMARY Raman and infrared spectra of a number of phosphine Lewis bases and their borane complexes have been studied in this work, An attempt has been made to describe the vibrational modes of these molecules by means of symmetry analyses and the group vibration approximation. On this basis, the observed vibrational frequencies have been assigned to the fundamental modes of vibration and in some cases to overtones or to combinations of the fundamentals, Since the group vibration approximation has been employed extensively throughout this investigation, it is worth while at this point to review its validity. This approximation has been used largely in connection with organic compounds, and it has been particularly successful in describing the motions of small isolated groups which are not incorporated into the molecular framework, For example,, the methyl group stretching and deformation vibrations for a large number of compounds have been found in the 28003000 cml and the 1400-1500 cm"l regions respectively, and they do not appear to shift appreciably from one compound to another. In the present case this has also been observed, the -BH3, -CH3 and -N-H group frequencies being easy to identify and remaining relatively constant in all the complexes studied, In general, the group vibration approximation has been studied less extensively in connection with inorganic than

165 with organic compounds, largely because it has proved less successful when applied to low frequency vibrations arising from various individual oscillators comprising a heavy molecular framework. However, these more complex skeletal modes often retain a certain degree of similarity in related molecules having the same general structure, thus frequency comparisons may still be informative even though a particular mode cannot be localized within a certain bond, A correlation diagram of the skeletal vibrations for the substituted fluorophosphine compounds, studied in this work, is represented in Figure 33, It is clearly evident from the diagram that certain vibrations do not shift appreciably from one compound to another. However, others, such as the P-F stretching and deformation vibrations, appear to be influenced by neighboring groups and these motions may involve a sensitive coupling with other skeletal vibrations. In the case of (CH3)3P, a comparison of the skeletal modes of the ligand with those of (CH3)3PBH3 in Table 26 reveals that the C3P deformation frequencies do not change a great deal when (CH3)3P complexes with the borane group However, both the symmetric and asymmetric C-P stretching vibrations shift appreciably upon complexation. This frequency shift probably is, to a great extent, the result of coupling between the C-P and P-B vibrations, but it is not possible to separate this from the effect of a possible change in the C-P force constant without a normal coordinate calculation v

OC2N 6PF2 vP-B yN-P-F2 va-F vP-F vC2-N-P vaC-N,, 11 -— 111 —------- --- — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — I -- -~~~~~~~~~~~~~~~~~~ 1 I I I r I ~ ~ Irr ~ ~ - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ II r 3~~~~~~~~~~. / \ I I \ / \ / I \\ I I l /'\ I l I I I I I I I I I I I I I IlI <.f. / /7,, " "' j.- -, / / /0 I //! I I \ \ \\ \ \ \ N II \ I f I I I I I I I I I I I F3PBH3 (5) (CH3)2NH (57,62) (CH3 )2NPF2 (CH3 )2NPF2BH3 [(CH3)2N]2PF [(CH 3)2N]2PFBH3 PF3 O-' 0~ (17)'i \\ \ \'\ \ II I I I I I I I I I' -. I - -- - ~ ~ I -'13 ~ L t -I ~ H I ~ 300 400 500 600 700 800 900 1000 1100 1200 CM 1 FIGURE 33 A CORRELATION DIGRAM FOR SOME FUNDAMENTAL VIBRATIONS OF PF3, F3PBH3, (CH3)2NH, (CH3)2NPF2, (CH3)2NPF2BH35 [(CH3)2N]2PF AND [(CH3)2N]2PFBH3

167 TABLE 26 A COMPARISON OF THE SKELETAL VIBRATIONS OF (CH3)3P AND (CH3)3 PB3 (CH3P (CH3 3PBH3 Description of Vibration 650 708 C-P symmetric stretch (a1) 297 287 C3P symmetric deformation (a1) 571 P-B stretch (a1) 705 757 C-P asymmetric stretch (e) 258 260 C3P asymmetric deformation (e) 229 C3P rock (e) The borane stretching vibrations of the various phosphineboranes studied appear to be good group vibrations by virtue of their separation from other molecular frequencies, As a consequence, one may expect that the effect of the dative bond on the borane group will be reflected directly in the magnitude of the B-H frequencies with relatively little complication by coupling effects. Previous work, in this laboratory and elsewhere, has shown that these vibrations can be correlated qualitatively with the strength of the Lewis base which is complexed with the borane group. If these B-H stretching vibrations are arranged in systematically decreasing order of magnitude, one obtains the effect shown in Figure 34, A more uniform trend is shown by the weighted average of the borane stretching fundamentals, since the use of the average eliminates coupling effects

168 within the borane group itself It can be seen that the trend established by the B-H frequencies, shown in Figure 34, agrees with the generally accepted order of base strengths as established from other experimental results, In like manner, the borane stretching fundamentals of the phosphineborane complexes reported in this work, or preferably their weighted averages, may be arranged in systematic order, Such a listing is depicted in Figure 35 Chemical knowledge generally agrees that dimethylamine is a stronger base than methylamine and that the methoxy group is more basic than the methyl group. In addition, the strength of the fluorophosphines, as Lewis bases, is expected to increase with increased substitution of amine groups for fluorine atoms on the phosphorous, Analogous to the previously mentioned trend in Figure 34 and in accordance with chemical intuition concerning the basicity of phosphine Lewis bases, the borane stretching vibrations of phosphine-borane complexes are seen to shift systematically to lower frequencies with increased base strength of the ligando Such a trend may be useful in predicting the stabilities of these complexes and predicting displacement reactions. An attempt to correlate the B-H stretching fundamentals of the ether and the amine complexes in Figure 34 with those of the phosphine series in Figure 35 was not successful, This may be because oxygen and nitrogen are first row elements while phosphorus is a second row elemento A correlation involving borane complexes of

169 Lewis Base I_, OC ( 2 5)20 (CH3 )2 C4H8 H3N CH3TNH2 (CH3 )2NH (CH3 )3N H (21) (70) (70) (70) (50) (50) (50) (50) (71) I I / / / / / I I I I i / / / I / I I I I I I I I I I I I IX I I I I11 - Borane Fundamental -- Weighted Average I I I I I 2200 2300 2400 2500 cmm FIGURE 34 - A CORRELATION DIAGRAM FOR THE BORANE STRETCHING FUNDAMENTALS OF VARIOUS BORANE LEWIS ACID-BASE COMPLEXES

170 Lewis Base F3P CH NHPF. (CH3 )2NPF2 (CH3NH)2PF [(CH3 )2N]2PF I. I ~ ~ I 1 1 / / I I / I I I II I I I I I I I i /l / (C6H5 )3 (72) (CH3 )3P I (NH )3P (CH30)3P (CH3NH)3 P (73) (74) (25) I I! I I I I I - Borane Fundamental Weighted Average I I. I. 2200 2300 2400 2650 cm1 FIGURE 35 - A CORRELATION DIAGRAM FOR THE BORANE STRETCHING FUNDAMENTALS OF SOME PHOSPHINE-BORANE COMPLEXES

171 phosphorus and sulfur might be informative in this respect, The extension of the previously mentioned trends to the phosphorus-boron bond vibrations is of interest, since these vibrations are relevant to the discussion of bonding and to the stability of the complexes. In the analogous amine-borane series, Cluff has found that the B-N dative bond frequency decreases in a continuous and regular fashion with the number of methyl groups attached to the nitrogen (50). Taylor has carried out a normal coordinate analysis of the vibrational frequencies of H3NBH3 and (CH3 BH3O and he has found that this frequency decrease is due to a decrease in the B-N force constant accompanied by a simultaneous increase in the coupling between the B-N stretching and the C-N stretching and deformation modes (75). The observed phosph'orus-boron stretching vibrations of several phosphineborane complexes are listed in Table 27. Although vibrational data concerning the trisubstituted aminophosphineboranes are not available, a similar trend appears to be presen; here as is found for the amine-boranes For example, the dative bond vibrations of the disubst-ituted amin.ophosph~ne-boranes are observed at lower frequencies than those of the monosubstituted complexes. Attempts to correlate the stretching vibrations of the phosphorus-boron bonds with the base strength of the ligands are only partly successful. In general the dative bond vibrations in phosphine-boranes appear to shift to lower frequenries with increased base

172 strength of the ligand, with the noted exception of the methylaminophosphine-borane s TABLE 27 THE P-B STRETCHING VIBRATIONS OF SEVERAL PHOSPHINE-BORANE COMPLEXES Compound F3PBH3 CH3NHPF2BH3 (CH3)2NPF2BH3 (CH3NH)2PFBH3 [ (CH3)2N ]2PFBH3 (C6H5 )3PBH3 (CH3 )3 PBH3 P-B Stretching Frequency (cm_ ) 607 642 589 628 571 608 571 (72) With the presently available information, very little more can be said about the strength or the stability of the dative bond in the phosphine-borane complexes, A complete vibrational analysis involving a normal coordinate treatment will bQ a necessary preliminary to further conclusions, since the phosphorus-boron stretching mode appears to be mixed with those of the ligand skeleton, and the extent and the nature of the mixing is uncertain. In order to explain fully the chemical properties of the phosphine-borane complexes, thermodynamic data will be necessary in addition to vibrational data,

APPENDIX A Preparation and Purification of Materials Methylamine, Anhydrous methylamine was purchased from the Matheson Company and purified by fractionating it through a -78~C trap into a -196~C trap to remove any trace of water. The vapor pressure of the amine was within 2 or 3 mm of the value of 100 mm at -43.7~C reported by Stull (76), Dime hylamineo A commercial cylinder of anhydrous dimethylamine was purchased from the Matheson Company. Before use, traces of water were removed by fractionating the dimethylamine through a -78QC trap into a -196~C trapo Previous work in this laboratory has shown that the dimethylamine is relatively free of other impurities. The vapor pressure of the purified dimethylamine gave a pressure of 40 mm at -39 o8C (76). Trifluorophosphine. Trifluorophosphine was prepared by fluorinating PC13 with ZnF2 using the method described by Williams (77) as modified by Sr. Fleming (13). The impure PF3 was purified by passing it through Dowex-3 to remove traces of HC1 followed fractionation from a -1500C trap to one held at -196~Co 173

174 Potassium Fluoborate. Potassium fluoborate 92,0% enriched in boron-10 was obtained from AoEoCo stock held by the Michigan Chemical Company. A sample 99 4% enriched in boron-ll was obtained from the Union Carbide Nuclear Companyo Boron Trifluoride The isotopically enriched boron trifluoride was prepared by thermally decomposing a 50% mixture of LiF and the appropriate isotopically enriched KBF o The decomposition was accomplished by heating the mixture in a stainless steel tube to around 6500C under vacuum and condensing the products in a liquid nitrogen trap, The crude BF was purified by condensing it on NaF at -126~C 3 to remove traces of SiF4 followed by fractionation from a -150~C trap into one held at -1960Co All ground glass connections were lubricated with Kel-F grease, Diethyl Ethero Analytical, reagent grade, anhydrous diethyl ether was obtained from the Mallinckrodt Chemical Works. This was dried and stored over calcium hydride. Boron Trifluoride Etherate. Boron-10 and -11 trifluoride etherates were prepared by direct reaction of the isotopically enriched BF3 and anhydrous ether, This reaction was brought about by allowing BF3 to transfer slowly from a -126~C trap on the

175 vacuum line into a -78~C trap containing a large excess of ether. The latter trap was stirred continuously with a magnetic stirrer until the reaction was completeo The excess ether was stripped from the etherate by evaporation and the crude boron trifluoride etherate was purified by fractional distillation at 530C and 30 mm pressure0 Diborane. Diborane containing the natural boron isotope ration (boron-lO0boron-ll1l: 4) was obtained from Callery Chemical Company and was purified by fractionation from a ~126~C trap, through a -150~C trap, to a -1960C trap. The purified material was recovered from the -1960C trap and stored in liquid nitrogen until use. A cylinder of diborane containing boron-10 in 93% enrichment was obtained from Arthur Do Little, Inc. through the courtesy of Amos J. Leffler, Another sample of diborane, enriched to 92% in boron-10, was given to the author by Ao Jo Dahl of this laboratory. Both of these samples were purified by fractional distillation. Isotopically pure B2H6 (99.4% enriched in boron-ll) was prepared by using the method described by Shapiro, Wiess, et al, (78) and more recently by A. R. Emery (40) and Co Lo Cluff (50) in which an ether solution of B1F300(C2H5)2 was reduced by LiAlH4 according to the equation 3 LiAlH4+ 4 (C2H5 )20BllF3 - 2 B2H6 + 3 LiF + 3 A1F3+4 (C2H5)2g.

In a typical reaction, 10 ml of B1F300(C2H5)2 were reacted with a slurry of 2.3 g of LiAlH4 in 30 ml of ether to give about one ml of B2 H6. The diborane was purified by passing it through a -150~C trap into a trap held at -196~C. The purity of the diborane was checked by comparing its vapor pressure at -111,8~C with the value reported in the literature. The preparation of B2 lD6 required a slightly different procedure because the deuterium could most easily be obtained as LiD. The method of preparation was essentially the same as that described by Finholt (79), and later by Emery (40) and Cluff (50) in which LiAlD4 was prepared in Situ by the addition of an ether solution of aluminum chloride etherate to a slurry of LiDo The B21 D obtained in this reaction was purified by passing it through a -150~C trap into a -1960C trap and its vapor pressure was checked with that reported in the literature. Metal Hydrides~ The LiA1H4, LiAlD4, and LiD were obtained from Metal Hydrides, Incorporated, Beverly, Massachusetts. The isotopic purity of the latter two compounds was stated to be 98% and 96,,6% respectively,

APPENDIX B Preparation and Purification of Saples F3PBH o A typical preparation involved the reaction of 861l mM of PF3 (5.7 fold excess) with 15.7 mM of B2H6 in a high pressure stainless steel reaction vessel which could withstand pressure up to 1500 lbo/sqo in. The reaction tube was allowed to stand for 4 days at room temperature and then was connected to the vacuum line to strip the product F3PBH3 from the unreacted PF3 and B2H6o The fractionation was accomplished by distilling the F3PBH3 through a -128~C trap, into a -155C trap and collecting the more volatile components in a -1960C trap, The less volatile impurities remained in the -128~C trap while the pure F3PBH3 remained in the -155~C trap. CH3NHPF2BH3 In a typical preparation, a 3.4 mM of F3PBH3 was dissolved in about 20 ml of diethyl ether at -llloC by first condensing the sample of F3PBH3 and then the ether into a reaction tube using liquid nitrogen. The tube was allowed to warm very slowly, with the pressure in the tube being monitored until the F3PBH3 and ether had melted and had gone into solution completelyo At no time were the contents of the tube allowed to warm above -lll0C These precautions were necessary because F3PBH3 decomposes 177

178 rapidly in the gaseous phase but is somewhat more stable in solution After the solution of F3PBH3 was complete, CH3NH2 vapor was introduced over the surface of the solution while stirring, until 9O0 mM had been completely absorbedo After stirring for 30 minutes at -111~C, the solution was gradually warmed to -780C at which temperature a white precipitate slowly formedo The mixture was allowed to stand for 16 hours at -78~C after which the volatile components were distilled from the tube through a trap at -78oC into a trap at -1960C. The reaction tube was then warmed slowly to 0~ and the distillation was stoppedo Diethyl ether and CH3NH2 were trapped in the -196~C trap and a mixture of diethyl ether and CH3NHPF2BH3 was retained in the -78~C trapo The final mixture of diethyl ether and CH3NHPF2BH3 was separated by fractionating it from a trap at -35~C through a trap at -65~C into a trap at -196~Co The CH3NHPF2BH3 was retained in the -650C trap, About 2~23 mM of CH3NHPF2BH3 was collected giving a yield of 65o7% based on the F3PBH3 usedo (CH3NH) 2PBH3o A sample of F3PBH3 (3.8 mMr) was dissolved in 15 ml of diethyl ether at -111~Co The same procedure for bringing about the solution of F3PBH3 in diethyl ether was used as was mentioned previously for the preparation of CH3NHPF2BH3O

179 An excess of CH3NH2 (47.4 mM) was frozen into the reaction tube with liquid nitrogen, the reaction tube sealed off and the system allowed to warm to -ll1Co The solution was stirred for 20 minutes with a magnetic stirrer and then warmed to -78~Co The mixture was maintained at this temperature and stirred for 10 to 12 hours Finally the reaction tube was allowed to warm to room temperature and maintained for 2 to 3 dayso After opening the tube to the vacuum system and removing the volatile components, 10 ml of fresh diethyl ether were condensed into the tube and the ether insoluble CH3NH2 IIF salt was removed by vacuum line filtration through a glass frito The diethyl ether was then distilled from the filtrate leaving the nonvolatile liquid, (CH3NH)2PFBH3o The yield of crude product was almost quantitative based upon the amount of PF3BH3 usedo The sample of (CH3NH)2PFBH3 prepared this way had a slight yellowish color, The sample was passed through charcoal followed by an ether wash. The diethyl ether was removed by evaporation and the (CH3NH)2PFBH3 remained as a colorless liquid, (CH3)2NPF2 A 13o5 mM sample of PF3 was condensed with liquid nitrogen in a thin layer inside a reaction tube and then 20~3 mM of (CH3)2NH was condensed on top of the PF3o The reaction vessel was repeatedly warmed to room temperature

180 and then cooled down with liquid nitrogen until no further drop in pressure was noted in the system, The product was purified by distillation from a -6~CC trap through traps held at -95~C and -196~C, the (CH3)2NPF2 remaining in the -95~C trapo Traces of (CH3)2NH and [(CH3)2N]2PF were removed from the sample by condensing the mixture on a small amount of anhydrous CoBr20 The system was warmed to room temperature and the resulting green solution allowed to stand for one hour before fractionating through a -95~C trap to a -196~C trap The purified ligand was recovered from the -95~C trap while (CH3)2NH2 and [(CH3)2N]2PF remain behind as complexes of CoBr20 Using this method of preparation a yield of about 85% was usually obtained for the (CH3)2NPF2o (CH3) NPF2BH3 In the usual preparation, an excess of the ligand was condensed in a tube with a few millimoles of B2H6 at -1960Co The mixture was alternately warmed to room temperature and then cooled with liquid nitrogen several times to insure complete reaction of the reagents after which it was fractionated through a -780C trap to a -196~C trap. Unreacted B2H6 was collected in the -196~C trap while crude (CH3)2NPF2BH3 was retained in the -78~C trap, The complex was purified further by distilling it from a trap held at -45~C into a trap held at -78~Co

181 (3 ) 2NPF2B2 1F3 About 2.0 mM of (CH3)2NPF2 were condensed in the bottom of a reaction tube along with 2.5 mM of B F30 The reaction tube was warmed slowly to room temperature. As the gaseous BlF3 reacted with the liquid (CH3)2NPF2, (CH3)2NPF2B1F3 a white crystalline solid formed, When the system had reached equilibrium the reaction vessel was placed in a -64~C slush bath and the volatile components were distilled from the tube through traps held at -950C and -1960Co Unreacted (CH3)NPF2 and B 11F were collected in the -95~C and -196~C trap respectively. The impure (CH3)2NPF2B lF3 was purified by sublimation from a trap held at -23~C, [(CH3)2N]2PF. A sample of (CH3)2NPF2 (4.2 mM) was condensed into a thick walled reaction vessel along with 12.6 mM of (CH3)2NHo The tube was sealed off and allowed to warm to room temperature and was maintained at this temperature for three dayso The tube was then cooled down in liquid nitrogen, the tube opened to the vacuum line, and the contents fractionated through traps held at -78~C, -95~C and -1960Co Dimethylamine was recovered from the -196~C trap, (CH3)2NPF2 from the -95~C trap, and [(CH3)2N]2PF was recovered from the -780C trap, leaving a whites nonvolatile, solid residue in the reaction tubeo The sample of [(CH3)2N]2PF initially gave a vapor

182 pressure of 14 mm at 0~C as compared to 3.6 mm reported by Kodama (25) and Sro Fleming (13). An infrared analysis indicated that the sample of [(CH3)2N]2PF contained (CH3)2NPF2 as an impurity. The more volatile (CH3)2NPF2 impurity was removed by vaporization while the sample tube was maintained at -63,5~C, until the vapor pressure and the infrared spectrum of the sample was identical to that reported by Sr. Fleming (13). (CH3)2N2PFB 11iH A sample of [(CH3)2N]2PF (5.0 mM) was condensed into a reaction tube along with 6.0 mM of B21 H (boron-ll enriched). The reaction tube was allowed to warm to room temperature and after the reaction mixture had equilibrated the contents of the tube were fractionated through a -35~C trap to one held at -1960Co The unreacted diborane was recovered from the -196~C trap and impure [(CH3)2N]2PFB11H3 remained in the -35~C trap, The impure product was purified by distilling it from a 0~C trap into a -196~C trapo The product was recovered from the -196~C trap and any less volatile impurities remained behind in the 0~C trap,

BIBLIOGRAPHY 1o J, Chatt and A, A, Williams, J. Chem, Soc,, 3061, (1951). 2o J, Chatt, Nature, 165, 637 (1950)o 3. G, Wilkinson, J, Am, Chemo Soc,, 73, 5501 (1951), 4. R, W, Parry and T, A, Bissot, J, Am, Chemo Soc,, 78 1524 (1956), 5. R, C, Taylor and T., C, Bissot, J. Chem, Phys,, 2, 780 (1956). 6, A, Bo Burg and Ho Io Schlesinger, J, Am, Chem, SoCo, 59 780 (1937 ) 7. W, Ao G, Graham and F, G, A, Stone, J, Inorg, Nucl, Chem,, 3, 164 (1956). 8. S. Ahrland, J, Chatt, and N. R, Davies, QuartO Rev,, 12, 265 (1958)o 9. E, Wiberg and K, Schuster, Z Anorgo Allgem, Chemo, 213, 94 (1933), 10. Po Baumgarten and W, Bruns, Chem, Ber,, 80, 517 (1947), 11. E, R, Alton, Jr., Doctoral Dissertation, University of Michigan, 1960, 12, R, W. Parry, et al,, "The Chemistry of Boron Hydrides and Related Hydrides, WADD Technical Report 60-262, June, 1960, 13o Sister Mo A, Fleming, IHM,, Doctoral Dissertation, University of Michigan, 1963, 14, D, M, Yost and T. F. Anderson, J, Chemo Phys., 2, 624 (1934), 15o Ho So Gutowsky and Ao D, Liehr, Jo Chem, Phys 20, 1652 (1952), 16, Mo Ko Wilson and S. R, Polo, J, Chem, Phys,, 20, 1716 (1952), 17o Ro C, Taylor, Unpublished Results, University of Michigan, 1956. 18o L, Ao Woodward and J. R. Hall, Nature, 181, 831 (1958). 183

184 19o F, Seel, K. Balreich, and R. Schmutzler, Chem. Ber,, 24, 1173 (1961). 20. M. L. Delwaulle and M, F, Francois, Compt, rend,, 222, 550 (1946). 21, R. C. Taylor, J. Chem. Phys., 26, 1131 (1957). 22, J. R. Barcelo and J. Bellanato, Spectrochim, Acta,, 8, 27 (1956). 23, R. B, Harvey and J, E. Mayhood, Can, Jouro Chem,, 33, 1552 (1955). 24. R, L. Amster and R, C. Taylor, Spectrochim, Acta,, In Print. 25, Go Kodama, Unpublished Results, University of Michigan, 1960. 26. F. G. Mann, A. F. Wells and D. Purdie, J, Chem, Soc,, 1828 (1937). 27. F, G. Mann and A. F. Wells, J. Chem, Soc., 708 (1938). 28, E, J. Rosenbaum and C, R. Sandberg, J, Am, Chem, Soc,, 62, 1622 (1940). 29, E. J, Rosenbaum, D, J, Rubin and C, R. Sandberg, J, Chem, Phys,, 8., 366 (1940), 30, F. J. Wagstaffe and H. W. Tompson, Trans. Faraday Soc,, 40, 41 (1944). 31. H, Siebert, Z, Anorg, Allgem, Chem,, 273, 161 (1953). 32, M, Halmann, Spectrochim. Acta,, 16, 407 (1960). 33. L. S. Bartell and L, 0. Brockway, J. Chem, Phys,, 32, 512 (1960), 34, A, B, Burg and R, I, Wagner, J, Am, Chem, Soc., 75, 3872 (1953). 35, F, Hewitt and A, K, Holliday, J. Chem, Soc., 530 (1953). 360 Lo N, Daasch and D, Co Smith, J, Chem, Phys., 1, 22 (1951)o 37. E o Lo Wagner and D, F, Hornig, J, Chem, Phys,, 18, 296 (1950).

185 38, I, H, Rank and J. S. McCartney, J, Opt. Soco Amero, 38, 279 (1948). 39, H, L, Welsh, M. F, Crawford, T, R, Thomas and G, R. Love, Can, Jour. Phys., 30, 577 (1952)o 40, A. R, Emery, Doctoral Dissertation, University of Michigan, 1957. 41, A, J. Dahl, Doctoral Dissertation, University of Michigan, 1963o 42, 0. R, Gilliam, H. D, Edwards and W, Gordy, Phys, Rev,, 5., 1014 (1949). 43. Lo 0. Brockway and F, T. Wall, J, Am. Chem, Soc., 56, 2373 (1934). 44, 0. L. Hersh, Doctoral Dissertation, University of Michigan, 1963. 45o R. C. Taylor, Unpublished Results, University of Michigan, 1956o 46, L. 0 Brockway and J. Y. Beach, J. Am. Chem, Soco,, 60, 1836 (1938), 47, W. Gordy, H. Ring and A, B, Burg, Phys, Rev,, 7, 512 (1950). 48. A, P. Gray and R. C, Lord, J. Chem, Phys,, 26, 690 (1957). 49, H, Wolff and D, Staschewski, Ber, Bunsenges, Physik. Chem,, 68, 135 (1964)o 50, C, L, Cluff, Doctoral Dissertation, University of Michigan, 1962, 51, L. J, Bellamy, "The Infrared Spectra of Complex Molecules," Second Edition, John Wiley and Sons, Inc,, New York, 1958. 52. Jo E, Stewart, J, Chem. Phys,, 30, 1259 (1959), 53, Do E, C. Corbridge, J. Appl, Chem,, 6, 456 (1956)o 54, Lo Co Thomas, Chemo and Ind,, 198 (1957)o 55o Foil A, Miller, Private Communication, Mellon Institute, Pittsburgh, Pennsylvania, 1964, 56. Ro Schmutzler, Inorg, Chem,, 3, 415 (1964).

186 57. K. W. F. Kohlrausch, Monatsh. Chem., 68, 349 (1936). 58, J. To Edsall, J, Chem, Phys., 5, 225 (1937). 59. E, A, V. Ebsworth and N. Sheppard, Spectrochim, Acta,, 13, 261 (1959)o 60, J, Goubeau, M, Rahtz and H. Becher, Z. Anorg. Allgem, Chem,, 275, 161 (1954). 61, A, J. Banister, N, N, Greenwood, B, P, Straughan and A, J. Walker, J. Chem, Soc., 995 (1964), 62. W, G. Fateley and F. A. Miller, Spectrochim, Acta,, 18, 977 (1962). 63, B, Holmstedt and L, Larsson, Acta, Chem, Scand,, 5,. 1179 (1951)o 64, E, M. Evleth, Jr,, L D. Freeman and R, I. Wagner, J. Org. Chem,, 27, 2192 (1962). 65, H. D, Springall and L. 0, Brockway, J. Am, Chem. Soc., 60, 996 (1938), 66, D, R. Linde, Jr. and D, E, Mann, J, Chem, Phys, 29, 914 (1958). 67, H, C. Beachell and B. Katlafsky, J. Chem, Phys., 2_, 182 (1957), 68. K, Shimizu and H. Murata, J. Mol, Spectr. 4L, 201 (1960). 69, B. Rice, R. J. Galiano and W, J, Lehmann, J. Phys, Chem, 61, 1222 (1957)o 70, B, Rice and H, HI Uchida, J, Phys, Chem,, 5, 650 (1955). 71, A, R. Emery and R, C. Taylor, J. Chem, Phys., 28, 1029 (1958). 72, W. Van Doorne, Unpublished Results, University of Michigan, 1964. 730 R. Wyma, Unpublished Results, University of Michigan, 1961. 74. C, W, Heitsch and J, G. Verkade, Inorg, Chem, 1, 863 (1962). 75. R, C, Taylor, "Boron-Nitrogen Chemistry, Advances In Chemistry Series, Vol, 42, American Chemical Society, Washington, D. C,, 1964, po 59.

187 76. Do R, Stull, Ind, Eng, Chem., 39, 517 (1947). 77. A. A. Williams, "Inorganic Synthesis," Vol, 5, To Moller, Ed., McGraw-Hill Book Co., New York, 1957 p. 95, 78, I. Shapiro, H, G. Weiss, et a2l,, J. Am. Chem, Soc., 74, 901 (1952) 79. A. E. Finholt, A, C, Bond and H, I. Schlesinger, J, Am, Chem, Soc., 69, 1199 (1947),