physical chemistry - American Chemical Society


physical chemistry - American Chemical Societypubs.acs.org/doi/pdf/10.1021/j100462a001Similarby GR Miller - ‎1980 - â€...

0 downloads 109 Views 308KB Size

THE J O U R N A L OF

PHYSICAL CHEMISTRY Registered in U.S.Patent Office 0 Copyright, 1980, by the American Chemical Society

VOLUME 84, NUMBER 25

DECEMBER 11, 1980

LETTERS The Identification of PF6- Ions in Intercalated Graphttes by lgF and "P NMR Spectroscopy Gerald Ray oepertmnr of Chmbtry, Unlverslty of Mamnd,college Park, Maryland 20742, and Code 6120, Naval Research labcatmy, Washlcyton, D.C. 20375

H. A. Reslng,' Code 6120, Naval Research laboratory, Wadwbgton, D.C. 20375

F. L. Vogel,$ A. Pion, 1.C. WU,* and

D. Billaud'

The Moore &thwl, and the hpartmenf of Chdstry, University of Pennsylvania, Phliadeghle, Pennsylvania 19 104 (Received: Ju& 15, 1980; In Flnal Form: September 23, 1980)

A graphite intercalation compound resulting from the reaction of graphite with N02PF6has been examined by 19Fand 31PNMR spectroscopy. The 'gF spectrum (doublet, JpF = 708 Hz)and the 31Pspectrum (septet, JPF= 708 Hz)show conclusively that the only mobile fluorophosphorus species present is the PF6- ion. This finding leads to the assignment of a positive charge density on the graphite lattice equal to the concentration

of PF6- ions found in the graphite.

Introduction Intercalation of molecules and ions in graphite raises the electrical conductivity of the graphite by a mechanism not yet well understood.lJ In many cases the actual chemistry of intercalation has yielded only to such classical techniques as maw balance and elemental analyais. Even when 'Address correspondence to this author at the University of Maryland address. *The Moore School. 8 Department of Chemistry. 0022-3654/80/208~3333$01.OO/O

the stoichiometry is known, there is usually not enough information to determine the chemical structure of the species present in the interlamellar region of the solid. NMR spectroscopy is, in general, a powerful tool for determining the structure of molecules or complex ions The containing suitable nuclei such as 'H,W, 'BF, and 31P. '?F NMR spectra of a number of graphites with fluorinecontaining intercalants have previously been The small relative chemical shifts between SbF6and SbF,and between AsF6 and AsF6- have not permitted the unambiguous differentiation between the presence of the 0 1980 American Chemical Society

3334

The Journal of Physical Chemlstty, Vol. 84, No. 25, 1980

molecular and the ionic species in these graphites. The indirect spin-spin coupling of nuclear spins by the electrons in the molecule or complex ion is another interaction which can reveal the nature of the species being observed in an NMR experiment. The magnitude of the coupling constant, J, between directly bonded nuclei depends sensitively on the electronic structure of the species. The number of lines in the resulting multiplet is dependent upon the nuclear spin quantum number, I, and upon the number of nuclei which are coupled with the same coupling constant to the nucleus being observed. However, the relaxation time of each of the coupled nuclei must be longer than the reciprocal of the coupling constant in Hz in order to observe the spin-spin coupling between the nuclei. Unfortunately, nuclei with I > usually have very short spin-lattice relaxation times due to quadrupolar interactions unless they are at sites of cubic symmetry, Even slight distortions from cubic symmetry can cause rapid spin-lattice relaxation of such nuclei. Although molecular motions render the 19FNMR lines of graphites which have reacted with AsF5, SbF,, and NOzSbF6quite narrow? the expected splittings due to coupling of fluorine to arsenic (75As,I = 3 / z ) or antimony (I2lSb, I = 5/2; 123Sb, I = 7/2) are not observed, even though in the case of an isolated SbF, ion the antimony site has cubic symmetry, It has been found r e ~ e n t l that 9 ~ NOzPF6 can react with graphite according to the reaction 1, where n 48 (second NOzPF6 + nC NO2 (gas) + Cfl+PF6(I)

-

+

stage; solvent molecules are also intercalated, see below). Since this intercalated graphite also shows high conductivity, it is worthwhile to take advantage of the fact that both 31Pand 19Fare spin l/z nuclei in order to establish, in one case at least, the nature of the species intercalated in the graphite. Although reaction 1seems unambiguous, the side reaction of graphite fluorination, reaction 2, can Cfl+PF6- C,F +- PF5 (2)

-

be postulated without any change in stoichiometry. If reaction 1 dominates, it is clear that one positive charge is deposited on the graphite lattice per PF,, there to act as a hole carrier or to be trapped. Dominance of reaction 2, on the other hand, would give no mechanism of charge carrier formation. Experimental Section The intercalated graphite used for these experiments was prepared by heating N02PF6(Ozark-Mahoning) for 30 min at 100 "C to remove volatile products of hydrolysis of the salt, vacuum distilling enough dried CH3N02 (Fisher) to make a saturated solution, and then adding degassed graphite powder. After 12 h the graphite powder was washed for 1 h with dry nitromethane, transferred under an inert atmosphere to glass sample tubes, and sealed after pumping on a vacuum line for 20 min. Specimens prepared in the same way from HOPG graphite achieved maximum loading at stage two with a c spacing' of 11.10 A and a composition6 of C4,PF6.2CH3N02. The c spacings for higher stages were also determined? When these data were used, X-ray diffraction (Mo Ka) showed our powder sample to be a mixture of stage two and stage three. Using a Bruker SXP spectrometer equipped with fluorine-free probes, we have examined the ?F (56.4MHz) and the 31P (24.3 MHz) NMR spectra of this intercalated graphite. A Nicolet 1080 system was utilized to accumulate off-resonancefree-induction decays (FIDs) and to perform the Fourier transformation. Alternate n / 2 and a-r-a/2 (r = 100 ps) pulse sequences were used to minimize probe

Letters

400

200

0

-200

8, P P ~ Flgure 1. The ''F NMR spectrum of the graphite intercalation compound of N02PFB.

TABLE I : NMR Parameters of Fluorine Comnounds compd SiF, PF,

PF I

6 Fa

160,170 35 76 65

6 Pb

- 97 80d

JPFc

1441 93fjd 710, 705 2 5e

PF, 118 observed main signal 76 128 708 20f impurity 176 a Referenced to 6 ~ ~ =~0;1data , from ref 13, Tables A38, A39, and A4 2. Referenced to 6 H,PO, = 0; data from ref 14, p 1055. In Hertz; data from ref 14, p 881. Data from ref 15. e Data from ref 9. f The observed coupling constant is 708 Hz in the I9F spectrum and in the 31Pspectrum, though the exact agreement is a result of utilizing the same experimental conditions for digitizing the data and Fourier transforming the FIDs.

ringing and base line effects.

Results and Discussion The 19FNMR spectrum (Figure 1) shows a doublet at 76 ppm with a splitting of 708 i 20 Hz and a small signal due to a trace of SiF4at 176 ppm (cf. Table I). The line width at half-height of each component of the doublet is about 270 Hz. The indirect spin coupling constant, JPF, varies from 710 Hz in PF6-to 938 Hz in PF6and 1441 Hz in PF,. The observed splitting of 708 Hz indicates that PF6-ions are present in the sample. The narrow line width results from (a) rapid tumbling about the center of the mass and (b) translational diffusion. (On the basis of the 12.2 G2 value of the rigid lattice second moment of an isolated PF6-ion: the line width should be on the order of 15 OOO Hz if molecular tumbling were slow on an NMR timescale.) It should be noted that the phase of the weak SiF4 signal in Figure 1is different from that of the PF6ions. This would be expected if the SiFl is present as a gas and the PF, ions are between highly conducting layers of graphite. Since it is possible that other species containing P-F bonds could have a coupling constant of 708 Hz, we have obtained the 31Pspectrum of the specimen used in the ?F NMR study. Figure 2b shows the spectrum obtained by Fourier transformation of 237 000 accumulated FIDs. Figure 2c shows the calculated spectrum of the 1:615:20156:1 septet expected for the coupling of six equivalent fluorine nuclei with a single phosphorus nucleus. The uniform separations of 708 Hz and the intensity distribution in the spectrum observed show conclusively that the species observed is the PF6- ion. There is no evidence in either the l9F or the 31PNMR spectra of this material for any other fluorine- or phosphorus-containing species. Specifically, neither the 938

The Journal of Physical Chemistry, Vol. 84, No. 25, 1980 3335

Letters

b

C

I v

2000

0

-2000

HZ

Flgure 2. Exptrnded scrnle NMR spectra of the graphite intercalation compound of NOpPFB. (All spectra have the same Hz scale as the abscissa.) (a) The F ‘ s ectrum, experimental. (b) The !“P spectrum, experimental. (c) The “P spectrum, theoretical.

Hz doublet expecteld for mobile PF5 in the 19Fspectrum nor the 1:5:10:105:1 multiplet with the same splitting expected in the 31Pspectrum is observed. We would expect that the PF5 would be mobile and have a narrow line width as is observed for other highly symmetrical fluorine-containing species intercalated in graphite~.~ Reaction 2 requires fluorinating the graphite as well as the presence of PF6. The nature of the l9F spectrum of the fluorinated graphite would depend upon whether the graphite was sparsely fluorinated, in which case the lines would be expected to be narrow, or was densely fluorinated, having regions approximating carbon monofluoride. The latter compound yields very broad linesl0J1and under the same experimentall conditions utilized for our lV NMR studies of the intercalated graphite, we observe no signal from a sample of carbon monofluoride, presumably because of the unfavorable combination of short T2 and long Tl. We would expect to observe the narrow line or lines expected from sparse1:y fluorinated graphite, particularly if the Tl

is short for the C-F fluorines due to the presence of reorienting PF5 molecules or PFs- ions nearby. No such signal is observed. Since the PF6- ions retain their identity on intercalation, we assign a positive charge density to the graphite lattice equal to the concentration of PF6-ions found in the graphite. A measurement of carrier density would then determine by difference the fraction of positive charges trapped at defects and unavailable for contributing to the electrical conductivity. It is true that in some graphite compounds12cation “co-intercalation” has been found; this process lowers the positive charge density attributable to the graphite. However, elemental analysis of a first stage intercalate prepared similarly by reaction of NOzSbF6with graphite has shiswn NO2+to be absent (although the elements of -1.7 (:H3NO2 per SbF6- ion were found)! Since our reaction conditions were similar, we expect no excess NOz+in the PF8-intercalated specimen and, thus, that our assignment of positive charge density to the graphite is well founded. In summary, our experiments show that only one mobile fluorine-containingspecies exists in our intercalated graphite and that this species is the PF6- ion.

References and Notes (1) F. L. Vogel, 0. M. T. Foley, C. Zeller, E. R. Falardeau, and J. &an, Mater. Scl. Eng., 31, 261 (1977). (2) Ian L. Spain and F. Llncoln Vogel, Abstracts of the Second Inter(3) (4)

(5) (6) (7) (8) (9)

(IO) (1 1) (12) (13) (14) (15)

national Conference on Intercalation Compounds of Graphite, Provlncetown, Mass., May 19-23, 1980. L. B. Ebert and H. Selig, Mater. Sci. Eng., 31, 177 (1977). B. R. Welnberger, J. Kaufer, A. J. Heeger, E. R. Falardeau, and J. E. Flscher, SclM State Commun., 27, 163 (1978). H. A. Resing, .:1 L. Vogel, and T. C. Wu, Mater. Sci. Eng., 41, 113 (1979). W. C. Forsman and Helen Mettwoy, submitted for publicatlon. D. Blllaud, A. Pron, and F. L. Vogel, “Synthetic Metals”, In press. A. Pron, private communication. Gerald R. Mlllor and H. S. Gutowsky, J . Chem. Phys., 38, 1983 11963). i.B. Ebert, J. I. Brauman, and R. A. Huggins, J . Am. Chem. SOC., 86, 7841 (1974). Peter Kamarclhick, Jr., and John L. Margrave, Acc. Chem. Res., $1, 296 (197a~). J. Besenhart amd H. P. Frltz, Z. Nafurforsch. B, 27, 1294 (1972). J. W. Emsley find L. Phillips in “Progress In Nuclear Magnetic Resonance Spectroscopy”, Vol. 7, J. W. Emsley, J. Feeney, and L. H. Sutcliffe, Eds., Pergamon Press, Oxford, 1971. J. W. Emsley, J. Feeney, and L. H. Sutcliffe, “High Resonance Nuclear Magnetic Resonance Spectroscopy”, Pergamon Press, 1965. L. Maler and R. Schmutrler, J . Chem. SOC.D,961 (1969).