J. MATER. CHEM., 1991, 1(3), 327-330 6Li Magic Angle Spinning Nuclear Magnetic Resonance Spectroscopy: A Powerful Probe for the Study of Lithium-containing Materials Stephen P. Bond, Andrew Gelder, John Homer, William R. McWhinnie* and Michael C. Perry Department of Chemical Engineering and Applied Chemis trx As ton Universitx Aston Triangle, Birmingham 84 7ET, UK A narrow chemical-shift range has been established for a variety of lithium compounds via study of their 6Li magic angle spinning nuclear magnetic resonance (MAS NMR) spectra. Comparative studies of 6Li and 7Li spectra (MAS, and 'off-angle' spinning) establish that, for solid-state (and even solution) analytical purposes, 6Li is the preferred nucleus, since the advantage of narrow absorption lines outweighs the poorer sensitivity of 6Li relative to 7Li.6Li MAS NMR spectra have been obtained for laponite clay, which had been thermally treated at 200,400, 600, 800 and 1300 "C; at the lower temperatures (<800"C) both dehydrated and rehydrated specimens were considered. The data are consistent with mobility of lithium ions from the trioctahedral clay sites at 600 "C. Both conventional and microwave methods were used to prepare lithium-exchanged laponite. The superior resolution achievable in 6Li MAS NMR in comparison with 7Li MAS NMR is demonstrated with the microwave specimen where use of 6Li spectroscopy revealed two lithium sites. On storage of the sample for 3 months, the two sites give way to a single lithium environment. Possible causes are discussed.Keywords: Magic angle spinning nuclear magnetic resonance spectroscopy; Laponite; Mobility in clays; Lithium6 ; Ion exchange; Microwave heating The use of 7Li NMR spectroscopy is well established in solution studies. The nucleus has a high natural abundance (92.5%) and a favourable receptivity (1 500, 13C =1.OO); how-ever, the quadrupole moment (-4.5 x e m2) can give rise to broad lines in non-cubic environments. It is, therefore, sometimes profitable in solution studies to examine 6Li spectra despite the lower abundance (7.5%) and the less favourable receptivity (3.58, 13C =1.00) as the quadrupole moment is also significantly lower (-8 x e m'), and narrower spectral lines can The use of 6Li rather than 7Li for solid-state MAS NMR studies has been less widely explored.In this paper information is presented which supports the view that, for solid-state work, 6Li is the preferred nucleus for study because the narrower lines may enable resonances of similar frequency to be resolved using 6Li MAS NMR when an apparent singlet may be seen in the 'Li MAS NMR spectrum. Experimental Lithium compounds were obtained from commercial sources: e.g. LiCl, LiBr, LiI (Aldrich), lithium metasilicate (Pfalz and Bauer). Lithium (12-crown-4) bromide was prepared by a literature method' (Found: C, 36.5; H, 6.03%. CsH16BrLi04 requires C, 36.5; H, 6.08%). Materials 8-Lithoxoquinoline sesquihydrate. 8-Hydroxyquinoline (15 g, 0.103 mol) was treated with an ethanolic solution of lithium ethoxide (0.103 mol).A precipitate was formed which was filtered from the yellow solution and Soxhlet extracted with ethanol. The resulting yellow solution was evaporated to afford yellow crystals, which were washed with cold ethanol and dried in vacuo; yield 60%; m.p. >300 "C (Found: C, 67.4; H, 4.54; N, 8.62; 0, 15.2%. C9H7LiN01.5 requires C, 67.5; H, 4.38; N, 8.76; 0, 15.0%). Laponite RD (laponite in this paper) was obtained from Laporte Industries Ltd. The thermal treatment, and rehy- dration (200, 400, 600, 800, 1300 "C) was carried out exactly as described in an earlier paper,6 using a Carbolite furnace. The clay was ion exchanged with lithium both using a conventional method7 and by treating 1 g laponite with an aqueous solution (10 cm3) of 1 mol dmP3 LiCl in a screw- capped Teflon container (Savillex Corporation, Minnetonka, Minnesota, USA) and heating for 5 min in a Sharp Carousel domestic microwave oven (650 W, med.-high setting in five 1 min bursts). Physical Measurements 6Li and 7Li MAS NMR spectra for powdered specimens were measured using a Bruker AC(E) 300 MHz spectrometer.The 'magic angle' was set with KBr and samples were packed into Delrin rotors and spun at ca. 5 kHz. The observation frequen- cies were 44.168 MHz (6Li) and 116.644 MHz (7Li). Apparent chemical shifts for both nuclei were measured with respect to a saturated aqueous solution of LiC1. Some spectra were obtained by spinning off the 'magic angle', and for 8-lithox- oquinoline sesquihydrate comparisons of line width for MAS NMR spectra were carried out for 6Li and 7Li. Results and Discussion Table 1 contains comparative data for the 6Li and 7Li nuclei, Table 2 reports 6Li chemical-shift data for some lithium compounds, Table 3 records data obtained from laponite heated to the indicated temperatures, and Table 4 gives data for Li-exchanged laponite.Following heat treatment and cooling to room temperature, half of each sample was treated with distilled water for 1 h, after which, following decantation of excess water, it was air dried for 1 week. Chemical-shift data for both dehydrated and rehydrated laponite are given in Table 3. The purpose for considering 6Li MAS NMR as an analytical tool arose from current work with lithium-containing minerals Table 1 Measurements of linewidth for 6Li and 7Li resonance lines under a variety of experimental conditions compound/measurement conditions G(ppm)” FWHM/Hzb 84ithoxoquinoline sesquihydrate 6Li MAS (dipolar decoupling, ‘H) 6Li MAS (no dipolar decoupling) 1.397 128 135 %i ‘off angle’ spinning 7Li MAS -1.881 -0.437 750 926 7Li static 5150 lithium (12-crown-4) bromide in D20 6Li solution (linewidth measurements made with high digital resolution) 0.504 0.39 7Li solution (linewidth measurements -0.169 0.46 made with high digital resolution) ‘‘Apparent’ chemical shift, i.e.at experimental peak maximum; FWHM =full width at half maximum. Table2 Some 6Li chemical-shift data obtained from MAS NMR studies of lithium compounds compound G(ppm us.satd. aq. LiCl) LiI -2.312 --2.152LiBr -0.359 laponi te -0.735 Li,Si03 0.200 Li( 12-crown-4)Br 0.659 Li(C,H6NO) * )H20 1.397 Table3 6Li MAS NMR data for thermally treated laponite and for Li +-exchanged laponite G(ppm us. satd. aq. LiCl) voc heated specimen rehydrated -room temp. -0.735 200 -0.63 1 -0.8 1 400 -0.395 -0.52 600 -0.227 -0.12 800 -0.145 0.05 1 1300 -0.077 e.g. laponite. It was considered prudent to survey some solid-state 6Li data for a variety of ‘simple’ compounds. Accordingly, the range of materials listed in Table 2 was considered. Before the experimental data in Tables 2 and 3 can be addressed it is necessary to point out that comparisons When the nucleus is located in a ligand field of less :.spin of J.MATER. CHEM., 1991, VOL. 1 between 6Li and 7Li spectra are not simple. 7Li has a nuclear than cubic symmetry [i.e. when the electric field gradient (EFG) is >O], the maximum of the -+++ transition observed in the MAS spectrum does not correspond to the isotropic chemical shift since spinning at the ‘magic angle’ does not average second-order quadrupole effects to zero. The situation for 6Li is different. The nuclear spin is integral (I= 1) and the quadrupole effect would be expected to produce a doublet for an orientated 6Li sample. The effect of magic angle spinning on such spectra is not well documented and this paper attempts to address this problem.In Table 1 some comparisons are made between 6Li and 7Li resonances in the same compounds. For MAS NMR spectra the dipolar broadening should be eliminated so that remaining effects should be dominated by quadrupole interac- tions, with 6Li and 7Li subject to the same EFG. Spinning off the ‘magic angle’ will introduce dipolar broadening but because the 6Li has a lower gyromagnetic ratio than 7Li the dipolar effect should be less on 6Li than on 7Li. 8-Lithoxo- quinoline sesquihydrate was used to investigate these points. Although the structure of this compound is not known, the symmetry of the lithium environment must be low. The 6Li MAS NMR spectrum of the compound has a FWHM= 128 Hz which increases to 135 Hz in the absence of proton dipolar coupling (a result not inconsistent with the co-ordi- nation of water).Spinning slightly off the ‘magic angle’ does introduce dipolar broadening. The 7Li MAS NMR linewidth is greater (some 6.86 times), as expected, than that for 6Li in the identical chemical environment (the ratio 7Li :6Li of the quadrupole moments is 56.25). For 7Li, the MAS experiment produces a 5.56-fold narrowing of the static linewidth. Attempts were made to obtain solution spectra for lithium in a low-symmetry environment. Solubility considerations narrowed the choice to lithium (12-crown-4) bromide; how- ever, it seems from the very narrow lines observed that the D20 solutions contain Li(D,O); species, thus the data are not comparable to those obtained from solid materials reported in Table 2.Despite this, measurement of linewidths under conditions of high digital resolution indicate that, even in the liquid phase, 6Li shows lines sharper than the corre- sponding ones from 7Li, and in the example cited the 7Li line is 18% broader than the 6Li line. Thus the data of Table 1 establish that, under all experimental conditions considered, 6Li gives narrower lines. Since neither series of spectra will give isotropic chemical shifts directly from the observed resonance maximum, the direct comparison of 6Li and ’Li ‘chemical shifts’ is not valid. However, comparisons of 6Li ‘chemical shifts’ measured under the same conditions should provide chemical information. Prior to using 6Li for this purpose information was sought regarding the range of 6Li chemical shifts. Table2 presents Table4 Some 6Li and 7Li NMR data on Li+-exchanged laponite material (experimental conditions) 6Li MAS (microwave method; fresh specimen) 6Li ‘off anglelc (microwave method; fresh specimen; small departure from magic angle) 6Li ‘off angle’ (microwave method; fresh specimen; large departure from magic angle) 6Li MAS (microwave method; aged specimen, 3 months) 6Li MAS (conventional method7) 7Li MAS (microwave method)d ’Li ‘off angle’ (microwave method;d small departure from magic angle) ‘Chemical shift, i.e.position of resonance maximum; FWHM =full width at half maximum; was used for both 6Li and 7Li spectra; freshly prepared specimens. S”(PPm) FWHMb/Hz -0.192 25.0 -0.438 (dv= 11.3 Hz) Av=11.9Hz 37.1 broad singlet 86.8 -0.179 37.7 0.232 -0.316 86.5 broad singlet 267 the same small departure from the ‘magic angle’ J.MATER. CHEM., 1991, VOL. I some data obtained for this purpose. The data of Table 2 do imply that the lithium is most shielded when it is more covalent (LiI); interestingly, the most 'ionic' of the samples considered was Li(CgH6ON) *3H2) which must imply that co- ordination of the nitrogen lone pair is, at best, extremely weak. Table 2 establishes a narrow chemical-shift range for Ti. The data for LiBr are of some interest. In the initial stages of the accumulation of the spectrum, a single resonance was observed. However, as the experiment proceeded a split resonance was seen with the parameters reported in Table 2.The new, weaker, component was the more deshielded and is attributed to LiBr (as) reflecting the hygroscopic nature of the salt; the more shielded resonance arises from LiBr. Laponite, a synthetic smectite clay (the closest natural counterpart is hectorite) contains ca. 0.62% lithium when obtained in the normal sodium-exchanged form: Nao.67 (Lio.67Mg5,33)Si8020(oH)4.The thermolysis of this mineral was recently the subject of a combined NMR (29Si, 23Na) and X-ray diffraction study, and it was thus of interest to examine the 6Li NMR spectra of specimens subjected to the same heat treatment. Despite the low lithium content, the natural- abundance MAS NMR spectra were of excellent quality and were obtained without excessive demands on instrument time; the FWHM of the spectral lines was of the order of 62 Hz.The 29Si MAS NMR study of laponite thermolysis sug- gested that even at 400 "C some loss of crystallinity and even breakdown of the silicate structure could occur.6 However, the effects seemed almost reversible on rehydration. The 6Li data in Table 3 reflect this quite well. Thus, on heating to 400 "C there is a deshielding of the lithium, possibly mainly caused by dehydration of the interlamellar sodium ions, since, on rehydration, the lithium undergoes an upfield shift to values closer to the -0.735 ppm characteristic shift of untreated laponite. The earlier study6 suggested that, on heating laponite to 600 "C, the lithium ions migrated from the trioctahedral sites to edge sites.In this instance, rehy- dration leads to a downfield shift consistent with the formation of aquated lithium ions on edge or interlamellar sites. By 800 "C the silicate network has essentially broken down and crystalline enstatite (MgSiO,) is the dominant phase. It was suggested that the lithium formed a silicate phase also; 6Li data are not inconsistent with this view but, at this juncture, are insufficient to identify the phase (no distinctive XRD lines were seen, the 800 "C and 1300 "C XRD traces were dominated by the enstatite polymorphs6). It was decided to ion exchange laponite with lithium to evaluate the possibility of distinguishing the structural and interlamellar ions.One method of ion exchange used was the classic but lengthy method of Posner and Q~irk.~This produced a material giving a single 6Li resonance, FWHM ~80Hz, centred on 6 =0.232 ppm. However, use of a new microwave method (see Experimental) gave a completely exchanged clay which was examined by both 6Li and 7Li MAS NMR spectroscopy (Table 4). Excellent spectra were obtained in both cases (Fig. l), but whereas a single resonance was obtained with 7Li (Table 4), clear resolution into a doublet is seen in the 6Li spectrum. However, re-examination of the spectrum after a lapse of 3 months revealed a single 6Li resonance (Table 4). A fresh preparation involving the micro- wave procedure again produced material which showed a doublet spectrum.For both preparations, the lines have a 1 : 1 intensity ratio. The spectrum observed may be that of a single lithium site giving rise to a quadrupole split line. However, we believe two pieces of evidence establish that the components of the doublet are chemically shifted. First, spinning 'off angle' by a small amount causes dipolar broaden- ing of the lines, but the separation of the components remains L I I I I I I I 1 I I 150 100 50 0 -50 150 100 50 0 -50 6 (PPW 6 (PPm Fig. 1 Comparison of the 'Li MAS NMR (a) and 6Li MAS NMR spectra (b) of lithium-exchanged laponite (microwave method) constant (within the error of the measurement, f0.5 Hz) at 11.3 Hz (MAS) and 11.9 Hz ('off angle').Secondly, there is no obvious reason why a quadrupole split resonance should change with time. It should be noted that spinning at large deviations from the 'magic angle' introduced sufficient dipolar broadening for resolution to be lost (Table 4). The 1 : 1 intensity ratio is consistent with expectation if the signals arise from interlamellar (exchanged) lithium ions and from structural lithium ions. The more deshielded resonance is assigned to interlamellar lithium ions (aquated) and the peak at 6= -0.438 ppm is assigned to the structural trioc- tahedral lithium ions; the slight deshielding relative to sodium laponite reflects the sensitivity of the lithium chemical shift to the identity of the interlamellar cations. It was of interest to note that investigation of an aged specimen failed to reveal two sites.Also, if the lithium was exchanged by a conventional method7 (i.e. washing several times with 1 mol dm-3 LiCl, stirring for 36 h at pH 4, followed by extensive washing and dialysis of the clay, a procedure of several weeks) a single resonance was noted (Table4). The data are consistent with a slow migration of lithium from trioctahedral sites into a common environment; the acceleration of the ion-exchange process in the microwave method enables the distinct lithium sites to be detected, but only if 6Li MAS NMR spectroscopy is used, the 7Li spectrum is a singlet. During the preparation of this manuscript Eckert et al. published their investigation of Li2S-P2S5 glasses by multi- nuclear NMR methods.8 Included were some 6Li MAS NMR data.The potential value of the superior resolution achievable in the 6Li spectra was noted in that work also. A. G. thanks the Royal Aircraft Establishment, Farnborough for support. S. P. B. thanks SERC and James River Graphics Corporation for a CASE award. This work has been carried out with the support of Procurement Executive MOD. References 1 S. Harder, J. Boersma, L. Brandsma, J. A. Kantas, W. Bauer and P. von R. Schleyer, Organometallics, 1989, 8, 1696. 2 D. R. Armstrong, D. Barr, W. Clegg, S. M. Hodgson, R. E. Mulvey, D. Reed, R. Snaith and D. S. Wright, J. Am. Chem. SOC., 1989, 111, 4719. 3 C. Brevard and P. Granger, Handbook of High Resolution Multi- nuclear NMR, Wiley, New York, 1981. 330 J. MATER. CHEM., 1991, VOL. 1 4 J. H. Gilchrist, A. T. Harrison, D. J. Fuller and D. B. Collum, 7 A. M. Posner and J. P. Quirk, Proc. R. SOC. London, Ser. A, J. Am. Chem. SOC., 1990,112,4069. 1964, 278, 35. 5 N. Poonia, J. Am. Chem. SOC., 1974,96, 1012. 8 H. Eckert, Z. Zhang and J. H. Kennedy, Chem. Muter., 1990, 2, 6 A-P. S. Mandair, P. J. Michael and W. R. McWhinnie, Poly-273. hedron, 1990,9, 517. Paper 0/04383D; Received 28th September, 1990