versão impressa ISSN 0103-5053
J. Braz. Chem. Soc. v.10 n.4 São Paulo jul./ago. 1999
Selected Topics from Recent NMR Studies of Organolithium Compounds
University of Siegen, FB 8, OCII
D-57068 Siegen, Germany
Após uma breve introdução à espectroscopia de RMN de metais alcalinos e alcalino terrosos, esta revisão concentra-se nas investigações de RMN em compostos organo-lítios. O método de impressão digital isotópica, que baseia-se no deslocamento das ressonâncicas de 6Li induzido por deutério, é apresentado e exemplificado com aplicações sobre o comportamento de agregação de sistemas de ciclopropil-lítio e formação de agregados mistos entre metil-lítio e sais de lítio. No capítulo seguinte discutem-se experimentos uni- e bidimensionais, tanto para sistemas de spin homonucleares quanto heteronucleares. Finalmente, descrevem-se os aspectos estruturais associados ao benzil-lítio e a formação de sistemas poli-lítio pela redução de bifenilas por lítio.
After a short introduction to NMR spectroscopy of alkali and alkaline earth metals the review concentrates on NMR investigations of organolithium compounds. The isotopic fingerprint method, which rests on deuterium-induced isotope shifts for 6Li resonances, is introduced and exemplified with applications from the aggregation behavior of cyclopropyllithium systems and mixed aggregate formation between methyllithium and lithium salts. In the following chapter, one- and two-dimensional pulse experiments, both for homo- and for heteronuclear spin systems are discussed. Finally, the structural aspects associated with benzyllithium are outlined and the formation of polylithium systems by lithium reduction of biphenylenes is described.
Keywords: NMR, 6Li-NMR, 15N-NMR, isotope shifts, isotopic fingerprints, pulse methods, spin-spin coupling, organolithium compounds, aggregation, benzyllithium structure, p-systems, polylithium systems, reduction
The alkaline and alkaline earth metals, a group of elements which comprises the four biologically most important cations (Na+, K+, Ca2+, Mg2+), provides us with an appreciable number of magnetic nuclei (Table 11,2). No wonder then, that NMR spectroscopy finds widespread applications, including such diverse topics like ion solvation in solution, investigations of ion binding to biological macromolecules and enzymes, solid state NMR of minerals and metal-doped fullerenes, as well as sodium NMR imaging.
With respect to investigations of structure and dynamics in organometallic chemistry, however, high-resolution NMR spectroscopy of most of these nuclides suffers from large quadrupole moments which lead to severe line broadening. Notable exceptions are beryllium, 9Be, cesium, 133Cs, and in particular the lithium isotopes 6Li and 7Li which can be successfully employed in various one- and two-dimensional NMR experiments. Especially 6Li, which has the smallest quadrupole moment of all stable nuclides and which has been classified ludicrously as an 'honorary spin-1/2 nucleus'3, is an important tool for the elucidation of structure and dynamics in lithiated carbon, nitrogen, and phosphorus compounds.
A concise and informative review on NMR of alkali and alkaline earth metals was lately given by Akitt2, who also lists the earlier progress reports for this field. Laszlo3,4 and Lutz5 provided additional articles, as did Drakenberg6 and just recently again Laszlo7. Several extensive progress reports dealing with 6,7Li-NMR have appeared8-13, a fact which underlines the continuous activity in this area. On the other hand, the biological importance of certain group I and II metals like sodium, magnesium, and calcium has initiated numerous NMR investigations of the respective nuclides in biological systems and results from this field, including accounts on sodium NMR imaging, have been summarized by several authors14-22. In addition, completely new areas for NMR investigations became accessable with the discovery of alkali anions23, the synthesis of alkali and alkaline earth intercalation compounds of fullerenes24-27, and studies on clay minerals used as catalysts in organic synthesis28. Two Specialists Periodical Reports29,30 summarize regularily the literature on NMR investigations involving alkali and alkaline earth metals.
For high-resolution NMR in organometallic chemistry, especially 6,7Li, but to some extent also 9Be and 133Cs are the nuclides of choice, while NMR of the remaining nuclei in Table 1 is less common for a number of reasons. As already mentioned, line broadening as a result of fast quadrupole relaxation renders the measurement of chemical shifts difficult if not impossible. For the same reason, scalar spin-spin coupling, which forms the basis of many modern NMR experiments, is not resolved or is even absent due to purely ionic bonding or the existence of solvent separated ion pairs. The majority of NMR investigations is thus confined to chemical shift and relaxation studies. In addition, because of effective quadrupolar relaxation, most of the heavier nuclei are not expected to show nuclear Overhauser effects which have proved so important in the structural elucidation of organolithium compounds. Aside from 1H, 6,7Li NOE effects, only for 133Cs NOE spectra have been reported31,32. Finally, from all organometallic systems of main group metals, the lithium compounds are by far the most important for synthetic applications, only rivaled in the field of carbon compounds by the Grignard reagents. Initial attempts to use 25Mg NMR in this area met with success33-35, but have not initiated further efforts in this direction, despite reported improvements in the experimental technique36.
It is thus quite understandable, that from the viewpoint of structural research on organometallic systems 6,7Li-NMR is much more attractive. In addition to 1H, 6Li and 1H, 7Li nuclear Overhauser effects, ample spin-spin coupling between 6,7Li and other nuclei like 1H, 13C, 15N, 29Si, 31P etc. exists and opens the doors to Alices wonderland of modern one- and two-dimensional NMR. The following account, therefore, deals exclusively with selected topics from recent NMR investigations of lithiated systems, where small linewidths and scalar spin-spin coupling paves the way for experiments which lead to a deeper understanding of structure and bonding.
Structure Determinations Via 2H-induced 6Li-NMR Isotope Shifts
small is beautiful
NMR isotope effects are long known37, but it was only after the introduction of high-field instrumentation that these parameters, which are often in the ppb region, became generally accessible. They soon were recognized as interesting data in connection with research on structure and bonding38-41. If an atom nX is replaced by its heavier isotope mX (m > n), a NMR shift, D(Y)(m/nX) is observed for the nucleus Y which may be directly attached to mX or several bonds away. For one-bond effects the shift is exclusively to high field (low frequency), while isotope shifts induced over several bonds may have the opposite direction. Two illustrative examples for one-bond effects from the literature are shown in Fig. 1.
The reason for the isotope shift lies in the bond lengths changes associated with the isotopic replacement, which lead to a slightly shorter bond for the compounds with the heavier isotope (Fig. 2). This is due to the anharmonicity of the X-Y bond potential and the lower zero-point vibrational energy of the mX-Y as compared to the nX-Y bond. This results in a shielding effect for Y but also for nuclei several bonds away. However, in the case of isotope shifts over more than one bond, the opposite sign (low-field or high-frequency shift) is often observed42.
In organic compounds, deuterium-induced shifts of 13C resonances have been studied extensively and correlations with carbon hybridization and substitution43, hyperconjugation44, dihedral angles45, and spin-spin coupling constants46 were found. Apart from these aspects which are related to physical organic chemistry, isotope shifts have also been used in a straightforward way as assignment aids in 13C-NMR47-49. 2H/1H isotope shifts decrease with the number of bonds between deuterium and the 13C nucleus which is observed and are usually too small to be detected if more than four bonds are involved. However, in unsaturated compounds effects over as much as twelve bonds have been reported50,51.
The isotopic fingerprint method which we introduced as a tool to study aggregation of organolithium compounds52 uses for the first time 2H-induced isotope shifts of 6Li-NMR signals. The idea was, that, for example in the case of a tetramer like methyllithium in diethylether, a 1:1 mixture of deuteriated and non-deuteriated material, CD36Li and CH36Li, should yield different environments for the 6Li nuclei. As shown in diagrams 1a - 1d, the direct surrounding of a particular 6Li nucleus might consist of three, two, one, or no CH3 group, leading to the environments hhh, hhd, hdd, and ddd. Considering the statistical distribution of the deuteriated ligand, a quadruplet with an intensity ratio of 1:3:3:1 was expected and indeed observed (Fig. 3). Here, the isotope shift amounts to roughly 16 ppb per CD3 group and is of positive sign on the d-scale (low-field or high-frequency shift).
By a straightforward extension of this reasoning, a doublet is expected for a monomer and a triplet for a dimer (Fig. 4). Thus, clusters of different size are characterized by isotopic fingerprints, where the intensity ratio within the multiplets follows Pascal's triangle. In general, the number of observed lines is n + 1 and the intensity distribution is given by the expression (a + b)n, where n is the number of organic ligands around each lithium cation and a and b are the mole fractions of 2H-labeled and non-labeled material, respectively.
The argument developed above takes into account only next neighbors and corresponds to the local environment approximation introduced by Brown.53 Indeed, 2H-induced isotope shifts from deuterons residing in organic ligands not directly attached to the 6Li nucleus under study are mostly too small to be detected and have sofar been observed in simple alkyllithium compounds only in a few cases (see below). The remote neighbor thus normally does not effect the lithium resonance as long as we deal with aggregates which are static on the NMR time scale. Considering the magnitude of the shift effect (16 ppb or 0.9 Hz at 58.88 MHz for 6Li on a 400 MHz 1H instrument), the lifetime of the particular cluster should be in the order of 2 s or more. If the lifetime falls short of this limit, intra-aggregate exchange brings also the remote neighbors into play and for a tetramer, again with equal numbers of deuteriated and non-deuteriated ligands, five different environments exist: hhhh, hhhd, hhdd, hddd, and dddd. Now a quintuplet results, as observed for the fluxional phenyllithium tetramer (Fig. 5a).
Finally, with the inset of inter-aggregate exchange, line broadening starts and a singlet is found in the fast exchange limit (Fig. 5).
An advantage of the isotopic fingerprint method as compared to other NMR techniques which are used to study aggregation phenomena and which rely on the measurement of 13C spectra (chemical shift studies, observation of 13C,6Li scalar spin-spin coupling) is its high sensitivity due to double isotopic enrichment, which is easily achieved. 6Li is readily incorporated directly or via lithiation with [6Li]butyllithium, while numerous procedures for the deuteriation of organic ligands are available. Thus, even aggregates which coexist in low concentration may be detected and characterized54.
In order to illustrate the application of the isotopic fingerprint method further, we discuss below recent findings for lithiated cyclopropyl compounds and results of a study on the structure of mixed aggregates between methyllithium and lithium salts.
Aggregation behavior of 1-Lithio-trans-2,3-dimethyl-cyclopropane
For a study of cyclopropyllithium compounds we chose the trans-2,3-dimethyl system 2 which was synthesized from the corresponding bromide 3, obtained by tributyltinhydride reduction of the 1,1-dibromide 4, which in turn resulted from the addition of dibromocarbene to trans-2-butene. A salt-free sample was prepared via the mercury compound 5. Deuterium at C-1 was introduced by reduction of the dibromide with tributyltindeuteride.
For the 6Li-NMR investigation of the aggregation behavior, a sample of 2 and [D]2 (1:1) in diethylether/THF (1:1) and one mole equivalent LiBr was prepared, which showed a 6Li doublet (Dd = 4 ppb) indicating the formation of a mixed dimer [C5H9Li,LiBr] (Fig. 6a). The observed multiplicity is also compatible with the presence of the monomeric lithium species 2, but in this case a separate 6Li signal for LiBr should have been observed. The sole resonance at 0.8 ppm (rel. to 0.1 M ext. LiBr in THF) is thus due to the mixed dimer. This finding contrasts with the observation made for the unsubstituted parent compound cyclopropyllithium, where a mixed tetramer, [(C3H5Li)2,(LiBr)2], has been found in the crystal55 and in diethylether/THF solution56.
The two-dimensional 1H, 6Li HOESY spectrum57 (Fig. 6b) established nuclear Overhauser effects between 6Li and H(1) as well as H(2), but cross peaks between 6Li and 3-CH3 were not observed. This may be due to steric repulsion between the methyl group and the lithium double bridge which increases the Li-CH3 distance.
For a salt-free sample of 2 and [D]2 (1:1), prepared via the mercury compound 5 in the same solvent mixture, a 6Li triplet, which characterizes a dimer, (2)2 , is observed at 187 K as the dominating signal (Fig. 7a). Interesting lineshape changes are, however, found at lowering the temperature to 161 K. The structure of the 6Li signal resembles a quadruplet between 170 and 163 K, which is associated with a tetramer. Tetramer formation is, however, rather unlikely considering the relatively modest and steady change in chemical shift which is much better explained by a normal temperature gradient rather than by a change of aggregation state. A 6Li-NMR spectrum run at 73.5 MHz (500 MHz 1H instrument) revealed that two overlapping triplets are present at 161 K which deceived a quadruplet at lower field strength (Fig. 7b). Thus, two dimers are present in a ratio of 1.0 : 0.7, a consequence of the chirality of the monomer 2 which forms diastereomeric dimers of the (R,S); (S,R) and (R,R); (S,S) type, respectively.
From the intensity ratio of the signals one calculates K = 0.48 and DGo = 1.1 kJ/mol at 163 K, but it is not known which of the two diastereomers is the more stable one. It is interesting, however, that NOE effects are found between 6Li and H(2) as well as 3-CH3 in these dimers (Fig. 7c) which indicates that their structure, as far as the orientation of the lithium double bridge with respect to the three-membered rings is concerned, differs from that of the mixed dimer containing LiBr. In the 1H spectrum the signals for the diastereomers are not separated and it is not clear if the NOE effects result from the d,l or the meso compounds or if both are responsible.
Quite a different picture emerged from measurements of the salt-free sample of 2 in diethylether as the sole solvent. The 6Li spectrum now shows three signals at 2.02, 2.08, and 2.22 ppm of comparable intensity (1.15 : 1.23 : 1.00). The isotopic fingerprint method yielded a doublet, a triplet, and a quintuplet which characterizes these signals as belonging to the monomer, the dimer, and a fluxional tetramer (Fig. 8). The coexistence of these three different aggregates is unique and apparently a consequence of comparable energies for solvation of the lithium cation with the solvent and the organic ligand.
Mixed aggregate formation between Methyllithium and Lithium salts
NMR studies of mixed aggregate formation between alkyllithium compounds and lithium salts often suffer from low sensitivity of 13C measurements if the concentration of certain clusters falls below 0.1 M. Here, the isotopic fingerprint method with its high isotopic enrichment can be used to advantage. We investigated methyllithium in the presence of LiI and LiBr, systems studied earlier by Brown58 and Waak59 by 7Li chemical shift measurements.
In the case of the CH3Li/LiI system52, five 6Li resonances were observed in the slow exchange limit at 178 K, four of which give rise to typical fingerprints which characterize the 6Li environment in the different aggregates; the signal at highest field is due to LiI (Fig. 9). The assigments based on signal multiplicity were confirmed by NOE measurements for a non-deuteriated sample, where a constant intensity increase per CH3 neighbor was found. This result also agrees with the observation of four signals in the 1H-NMR spectrum.
From an analysis of the measured intensity distribution and the observed 1H, 6Li NOE effects, the presence of aggregates Li4(CH3)4 (6), Li4(CH3)3I (7), Li4(CH3)2I2 (8), and Li4(CH3)I3 (9) was derived. Due to facile crystallisation of cluster 8 and LiI, instead of a statistical distribution only a non-equilibrium distribution of the aggregates was observed which did not allow to calculate energy differences on the basis of signal integration.
A similar study of the system CH3Li/LiBr in diethylether yielded the 6Li-NMR spectra shown in Fig. 10a, where the NOE difference experiment (Fig. 10b) identifies those lithium sites which are adjacent to at least one CH3 group. This leaves a total of five signals, centred at three different chemical shift values (0.44, 1.04/1.08, 1.72/1.80 ppm rel. to LiBr).
The measured NOE effects suggested next neighbor environments [CH3BrBr], [CH3CH3Br], and [CH3CH3CH3] for these 6Li resonances and this is borne out by the isotopic fingerprints observed for a sample of composition CH3Li/CD3Li/LiBr (1:1:2) (Fig. 10c): there is a doublet for signal 2, two triplets for signals 3 and 4, and two quadruplets for signals 5 and 6. As in the case of the iodine containing clusters, we can distinguish the 6Li resonances with the environments [LiCH3CH3CH3]CH3 (signal 5) and [LiCH3CH3CH3]X (signal 6) with X = Br in the present case, but compared to the iodine case the chemical shift order for these two aggregates is reversed. Furthermore, we have here different signals for [LiCH3CH3Br]CH3 and [LiCH3CH3Br]Br. As shown by the highly resolved spectrum of the two triplets around 1.06 ppm, there is a small doublet splitting of 0.24 Hz for each line of the triplet corresponding to [LiCH3,CH3Br]CH3 (Fig. 10d). Thus, an isotope effect of 4.1 ppb from the remote CD3 group is present. The analysis of the signal intensities, the splitting patterns as well as the NOE effects shows that apart from the tetramer 6 (signal 5) the mixed aggregates 10 (signals 4 and 6) and 11 (signals 2 and 3) are present.
There is no clear indication of the presence of a significant concentration of cluster 12 which should yield a doublet as its isotopic fingerprint. A number of smaller lines around 0.3 ppm, which indeed yield doublets in spectrum c), are unidentified and may come from this source. The 6Li resonance of the Li[BrBrBr] environment could coincide with the LiBr signal at 0 ppm.
If THF is used as the sole solvent, dramatic changes in the number of lines and their intensities as well as multiplicities are observed (Fig. 11). Cluster 6 (signal 5) is now the dominating species with a small contribution of 10 (signals 3 and 4). Again a long range isotope effect is observed for environment [LiCH3CH3,Br]CH3 (Fig. 11c). A doublet at 0.73 ppm in spectrum b) (signal 2) indicates the presence of a next neighbor environment [CH3BrBr]. This cannot, however, originate from cluster 11, because this would require another signal of the same intensity at ca. 1.1 ppm for the next neighbor combination [CH3CH3Br]. An interesting information as to the origin of this signal comes from the dynamic behavior of the 6Li spectrum, which shows coalescence between this doublet with the singlet of LiBr while the remaining resonances are virtually unaffected (Fig. 12). The doublet thus arises from a mixed dimer 13. This is nicely born out by the 13C-NMR spectrum, which shows in addition the the septuplet of the tetramer 6 (J = 5.7 Hz) a quintuplet with J = 9.8 Hz, compatible with a dimer (Fig. 12b).
Compared to the results of the earlier investigations58,59, which were based on the temperature and concentration dependence of the 7Li chemical shifts, the isotopic fingerprint method thus established the additional existence of aggregates 8, 9 and 13.
One- and Two-Dimensional NMR Experiments Based on Scalar Spin-Spin Coupling and Nuclear Overhauser Effects
Two-D or not Two-D
Scalar spin-spin coupling is one of the fundamental NMR phenomena for chemical structure determinations and lends color to the otherwise rather dull singlet spectra of uncoupled spins. Even more important, scalar interactions form the basis for numerous one- and two-dimensional experiments which yield information on atomic connectivities in a given molecular structure.
In the case of organolithium compounds, the magnitude of spin-spin coupling involving lithium strongly depends on the coupling partner, with fairly large values (> 2 Hz) for 13C, 15N, and 31P and small values (< 1 Hz) for 1H and homonuclear 6,7Li,6,7Li coupling. The sensitivity of various new NMR techniques for small coupling constants is thus of considerable interest if 6,7Li,1H or 6,7Li,6,7Li coupling is to be detected.
It is important to remember that information about the structure of the various aggregates of RLi systems which are formed in solution in the presence or absence of stabilizing diamines and other ligands comes primarily from spin-spin coupling or nuclear Overhauser effects which involve lithium and to a lesser extent from chemical shift data of the ligands. X,1H coupling (X = 1H, 13C, 15N, 31P) within the ligands R does not yield information about aggregate size or structure and coupling between protons or X nuclei of different ligands R is normally not observed. Thus, in this context 6,7Li,X and homonuclear 6,7Li,6,7Li couplings are of fundamental importance.
From the various homonuclear 2D NMR experiments which can be applied to detect and measure scalar spin-spin coupling, the COSY, COSY-DQF, TOCSY and INADEQUATE experiment60 were used in this context successfully for 6Li,6Li spin systems. Experiments whith 6Li generally profit from the smaller linewidth of 6Li as compared to 7Li signals, but the larger 7Li,7Li coupling (factor 2.642 ~ 7 due to the ratio g(7Li)/g(6Li) = 2.64) is an attractive feature of 7Li,7Li experiments, since small splitting might lead to an elimination of cross peaks in the 2D spectra if antiphase components result. The identification of homonuclear 6,7Li,6,7Li coupling, which sofar has never been resolved in a normal 1D Li-NMR spectrum, is important in cases were several non-isochronous 6,7Li-NMR signals are observed.61 Those belonging to the same cluster can then be recognized if homonuclear coupling exists, which requires short lithium distances typical for Li-C-Li arrangements. It is not clear if coupling between the two Li nuclei is transmitted directly or as a geminal interaction via the carbon.
From the 2D techniques cited above, the INADEQUATE62 and the COSY-DQF63 experiment have the additional advantage of a built-in double quantum filter which eliminates any true singlet from the observed spectrum. For the detection of 6Li,6Li coupling, the INADEQUATE experiment is most easily applied with appreciable time saving in its 1D version,64 recently performed also for 7Li (Fig. 13). But it is also an attractive choice for other applications and was used to measure for the first time a homonuclear 15N,15N coupling in a mixed aggregate of lithiated amides. As is well known from investigations by Collum et al.,12 lithium diisopropylamide (LDA) forms dimers in THF65 and cyclic trimers and higher cyclic oligomers in hydrocarbon solvents66. For a 1:1 mixture of 15N and 6Li labeled LDA and lithium di(3-pentyl)amide (LDPA) in THF we observed the expected four 15N signals - two of them nearly degenerate - stemming from the symmetric aggregates 14 and 16 and the mixed aggregate 15 (Fig. 14). All signals show quintuplet splittings due to coupling to two 6Li with a coupling constant of 5.0 Hz (Fig. 14a).
The 1D 15N,15N INADEQUATE experiment (Fig. 14b), which selects coupled AX systems identifies the two signals belonging to 15NA and 15NB of the mixed aggregate which now show an additional antiphase splitting of 1.6 Hz due to the 15NA,15NB coupling. A further observation of interest is the different intensity of the two 15N signals which results from different nuclear Overhauser effects in the two amide residues, where the larger number of protons in the LDPA part of the mixed cluster enhances the intensity of the 15NB signal. The 15N assignment which follows from this effect is in agreement with the assignment derived from substituent increments where a b-methyl group leads to a downfield shift67.
In changing the solvent to hexane the number of 15N signals increases which indicates also an increase of coexisting structures (Fig. 15a). The INADEQUATE experiment (Fig. 15b) selects two 15N AX systems which we assign to cyclic aggregates (RLi)n with n = 3 or 4 (17, 18), because the homonuclear 15N,15N coupling now drops to 1.0 Hz. This is a strong indication of a structural change to a cyclic trimer or a higher cyclic aggregate where only one coupling path is available between the two nitrogens.
Analysis of ligand 1H spectra
For the analysis of ligand structures, 1H-NMR plays an important part in many cases and the possibility to start magnetization transfer selectively is an attractive feature of several 1D versions of well-known 2D NMR experiments. For example, the selective homonuclear 1H TOCSY experiment60,68-70, improved by trim pulses (TP)71 and a z-filter69 (pulse sequence (1)) can be employed to unravel the strongly coupled 1H spectrum of cyclohexyllithium. As shown in Fig. 16, starting the magnetization transfer at the tertiary proton adjacent to the metal, which has a resonance well separated from the remaining signals by its low-field shift, axial and equatorial protons at subsequent ring positions are differentiated by the variable mixing time.
Another application of selective excitation is indicated if deuteriated isotopomers of certain solvents are not easily available and 1H signals of interest might overlap with large solvent peaks. Selective excitation of the particular spin system then allows elimination of the solvent signals and the inspection of spectral regions which were before masked by the solvent lines. An example is shown in Fig. 17 with the application of a 1D COSY experiment70 to the 1H spectrum of isopropyllithium, a compound that forms tetramers and hexamers in hydrocarbon solvents.72,73 At around 200 K in pentane the tetramer/hexamer ratio is ca. 10:1. Here we start with a selective 90° pulse, thereby defining the origin of the magnetization transfer which follows (pulse sequence (2)), The detection of the methyl resonances of both aggregates, which are hidden under the huge solvent lines, allows the chemical shifts of the methyl protons and the vicinal 1H,1H coupling constants to be measured. An advantage is the antiphase character of both doublets, which facilitates the extraction of the NMR parameters by discriminating artefacts.
Heteronuclear shift correlations
Heteronuclear shift correlations have been used quite frequently with success to correlate 1H, 13C, 15N, and 31P signals with the relevant 6Li resonances of the aggregates of interest and in many cases the appropriate experiments with 2H as a spin-1 nucleus have paved the way74-79. Over the years, these experimental techniques have been considerably improved and especially the so-called inverse techniques based on multiple quantum coherences80 (the HMQC experiment, pulse sequence (3)), where the experimental success relies on the suppression of uncoupled I magnetization, have profited from hardware developments, the application of additional pulses81 like the BIRD sequence82, and more recently from the introduction of linear field gradients60,83.
The inverse 13C,6Li experiment with 6Li detection (pulse sequence (3), I = 6Li, S = 13C), which was used for the first successful realization of a two-dimensional 13C,6Li shift correlation,76 not only yields correlation information, but also allows 13C,6Li coupling constants to be determined since usually the S nucleus is not decoupled during I signal acquisition (Fig. 18a). This experiment can also be performed most effectively and time saving by the corresponding 1D version (pulse sequence (4)), as demonstrated in Fig. 18c.
A recent addition to the list of X,6Li correlations is the 29Si,6Li experiment based on sizable scalar 29Si,6Li coupling84. An example is shown in Fig. 19 with the result for the dimer of (E)-1-lithio-2-(o-lithiophenyl)-1-trimethylsilylethene (Fig. 19).
Heteronuclear Overhauser spectroscopy
Finally, turning to 1H,6Li nuclear Overhauser spectroscopy, the 2D 1H,6Li HOESY experiment85 is one of the important tools in structure elucidation of organolithium compounds57. Recent developments in this field have led to the proposal of the inverse experiment with 1H detection which has the advantage of higher spectral dispersion in the 1H domain. The idea was originally put forward already in 1990 by Bauer and Schleyer86a, but only the introduction of linear Bo field gradient techniques paved the way for a practical solution of the experimental difficulty to eliminate the dominating proton magnetization which is not due to a heteronuclear 6Li ®1H NOE86b. It was found that the experiment is best performed with the 7Li,1H spin pair, apparently due to the stronger dipolar interactions and the faster relaxation rate of 7Li as compared to 6Li.
In an attempt to transfer these ideas to the one-dimensional version of the NOE measurement, we based our experiments on results reported by Keeler et al.87 for gradient enhanced 1H,1H nuclear Overhauser (GOESY) spectroscopy and introduced two frequency channels along the lines of the 2D 1H,6Li HOESY experiment. This leads to a pulse sequence shown in Fig. 20, which takes advantage of a later version of the GOESY experiment88. Here, the first part up to the gradient pulse G4 serves for the selection of the desired 7Li magnetization, I(7Li)sel , of a particular lithium resonance, 7Lik, which is to be transferred to the protons. Therefore, the conditions G1 = G2 and G3 = G4 refocus I(7Li)sel because the two selective 180° pulses change the sign of the coherences. The 90° 1H pulse produces transverse proton magnetization which is destroyed by the gradient pulse G5 , leaving for detection only the Overhauser enhancement which builds up during the mixing time through transfer from the selected nucleus 7Lik.
Experimental results for the well characterized dimer of (Z)-2-lithio-1-(o-lithiophenyl)ethen89 are shown in Fig. 21. In spectrum b) we see strong NOE's between Li(2) and both olefinic protons, in spectrum c) between Li(2') and H(3) at the aromatic ring. Weaker responses are coming in spectrum b) for H(3) and in spectrum c) for H(1) and H(2). In contrast, the 2D 1H,6Li HOESY spectrum (Fig. 21d) shows only the relationships Li(2)/H(2) and Li(2')/H(3) which are also the strongest in the 1D experiment.
Noteworthy is the enormous time advantage of the 1D sequence: these spectra were recorded within 35 min, while for the HOESY experiment 13 h had to be invested!
The Benzyllithium Story
Ach wie gut daß niemand weiß,
daß ich Rumpelstilzchen heiß
Despite the power of modern NMR experiments, even in the case of simple organolithium systems all attempts to determine their solution structure may fail due to a variety of reasons, among which are low solubility and fast exchange dynamics. One of these small molecules, which preserved until recently the secret of its detailed structure in solution, is benzyllithium. Even today, all facets of this structural problem may not have been uncovered.
Early X-ray crystallographic studies for solids consisting of benzyllithium and donors like triethylamine90 or diethylether91 ain struas ligands revealed chctures with Li-Ca distances of 217 and 221 pm, respectively, and different orientations of the Li cation with respect to the benzyl residue. With respect to the solution structure, the results of calculations by various semi-empirical and ab-initio methods92 are of interest, which suggested that in principle three alternative structures (19 - 21) may be discussed for solvated benzyllithium and its a-substituted derivatives (L = solvent or complexing ligand). Following MNDO results for the heat of formation93, the energy difference between the h1 and h3 structure is rather small (~ 4 kcal / mole, see next page).
Indeed, experimental evidence for a h1 h3 equilibrium was presented in the case of a-(dimethylamino)benzyllithium in THF solution94.
For solids, Boche et al. focused again attention to this structural problem by reporting the results of an X-ray investigation for the [benzyllithium,THF,TMEDA] complex93. These workers found a monomeric h1 structure with pyramidal Ca and a Li-Ca distance of 221 pm. Among the various a-substituted benzyl systems which have been studied by X-ray diffraction95, the trimethylsilyl substituted system [C6H5CHSi(CH3)3Li].TMEDA is of particular interest. Here, a h2 structure with pyramidal Ca and a Li-Ca distance of 213 pm was found in the crystal96 and according to 1H,6Li and 13C,6Li HOESY measurements97, the same or a closely related structure prevails in solution.
The earliest solution studies98 had already indicated that benzyllithium is monomeric, but in the NMR spectrum of 6Li labeled material no 13C,6Li coupling constant was found.99 This is also true for a-substituted derivatives100, with the single exception of a phenylsubstituted cyclopropyl system101. Consequently, for solute samples one could expect solvent separated ion pairs or fast equilibria between contact ion pairs or dynamic processes between both structural alternatives. The existence of rapid dynamic processes was also indicated by the low configurational stability of a-substituted derivatives102.
Findings reported recently carry our understanding of these structural aspects a step further. By chaining up the lithium cation to Ca via a crown ether 'necklace', Hoffmann and Boche et al.103 were able to observe for the first time a scalar 13Ca,6,7Li coupling constant in compounds 22 and 23 in THF at 203 K (22: J(13Ca,7Li) = 7.0 Hz, 23: J(13Ca,6Li) = 3.1 Hz which corresponds to J(13Ca,7Li) = 9.0 Hz). 1H,6Li HOESY experiments suggest an arrangement such as that shown in 24 for these monomeric systems.
At the same time, using a similar strategy, Fraenkel and Martin104 found 13Ca,6Li couplings of 2.8 and 3.4 Hz in THF at 250 K for the two compounds 25a and 25b and we observed 13C, 6Li couplings (2.7 Hz) for the lithium cation trapped in the organic ligand of the Schlenk dimer 26 where the second lithium exists as a solvent separated ion105. Even more interesting, Fraenkel and Martin104 were able to measure a 13Ca,6Li coupling of 3.8 Hz for the parent compound itself in the presence of TMEDA, using 13C and 6Li labeling, low concentration (0.005 M in THF), and low temperature (180 K). Thus, finally the conversation of the Ca and 6,7Li spin was tapped, but the low value of the coupling constants is unexpected in view of the findings that monomers usually show 13C, 6Li coupling close to 17 Hz99. Fraenkel assumes that this suggests the existence of a continuum of covalency between 'classical' monomers with large coupling and solvent separated ion pairs with no coupling at all. But, as mentioned above, the benzyllithium system is probably still good for a number of surprises.
Compounds with p- and s-Bound Lithium
some like it hot
The strong shielding and deshielding effects exerted by cyclic p-systems on surrounding protons has long fascinated annulene chemists who coined the terms diatropic and paratropic to describe the diamagnetic and paramagnetic shielding properties of (4n+2) and (4n)p-systems, respectively.106 In particular the oxidation or reduction of neutral hydrocarbons to the corresponding dications or dianions, respectively, has generated interesting systems where the difference in the number of p-electrons results in spectacular shielding and deshielding effects in proton NMR spectra107. An example from our laboratory is the generation of methanoannulene dianion (28) from methanoannulene (27) by reduction with lithium metal.108
The transformation 27 ® 28 was achieved after earlier unsuccessful attempts by using ultrasonic radiation or by simply attaching the sealed NMR tube with the parent hydrocarbon and lithium sand in THF to the rod of a vibrational mixer. This technique was also successful in the case of biphenylene, were we had found109 that the dianion 29, formed initially via reduction by potassium in THF, reacts with protic solvents to yield benzocyclo- octatetraene (31) via the Woodward- Hoffmann allowed ring opening reaction of the primarily formed 4a,8b-dihydrobiphenylene (30). In the absence of proton donors, 29 has a half-life of 1.7 h and opens the four-membered ring to yield o,o`-dilithiodiphenyl (32)110, a compound with a lithium double bridge111.
The reduction of benzoannelated Biphenylenes
In an attempt to study this reaction in the case of benzoannulated biphenylenes, we treated benzo[b]biphenylene (33) with lithium sand in diethylether. Instead of the dianion we found immediate formation of 2-o-lithiophenyl-3-lithionaphthalene (34), which yields a purple solution and is characterized by two ABCD systems and two singlets in the 1H-NMR spectrum. Two carbon resonances at 174.2 and 176.0 ppm and a 6Li singlet at 2.30 ppm (rel. to ext. 0.1 M LiBr in THF) complete the information which is significant for the structure.
In contrast, the oxidation of the hydrocarbon 33 with the 'Olah mixture' SbF5/SO2ClF at -30 °C yielded the 14p electron system 35 which is, at that temperature, perfectly stable. Its Q-value112 of 1.43 as determined from the bond orders derived on the basis of the vicinal H,H coupling constants 3J(6,7) and 3J(7,8) via analysis of the 1H-NMR spectrum, is typical for a benzoannelated diatropic system, in the present case the 10p electron system of biphenylene dication.
Completely different NMR spectra were observed when the reduction of 33 was carried out in THF. In particular the 1H- spectrum showed spectacular high-field shifts for some of the protons, which resonate at 2.25 and 2.60 ppm. Similar spectacular high-field shifts for 'aromatic' protons were found before in the case of dilithionaphthalenediide113. Most conclusive evidence for the structure of the new product came from the deuteron NMR spectrum of the deuterolysis product, which showed four signals, two in the allylic, one in the olefinic, and one in the aromatic region. All spectroscopic data, including the 13C and 6Li spectra, where in accord with the tetralithiated structure 36 which has p- and s-bound lithium as a unique feature114. Additional proof for the proposed structure came from an experiment where 34 was prepared in diethylether and further reduction was carried out after replacing this solvent by THF.
The interesting structural properties of 36 with p- and s-bound lithium initiated related studies for dibenzo[b,h]biphenylene (37) and naphtho[b]biphenylene (38)115. In both cases reduction with lithium sand in diethylether yielded the dianions, 39 and 40, respectively, which could be fully characterized by their NMR spectra. Surprisingly, however, both are paratropic systems with high-field shifted 1H resonances (2.81, 4.49, and 4.98 ppm for 39, and 2.40, 3.43 and 3.96 ppm for 40), despite the total number of 22 p-electrons. The Q-values are 0.963 for 39 and 0.982 for 40 and point in the same direction. The NMR data of 39 closely resemble those of dilithionaphthalenediide which suggests a mesomeric structure 39a «39b, while 40 resembles a phenylene-annelated anthracene dianion. In contrast to the dianions of the linear annelated systems, the dianion of the angular annelated benzo[a]biphenylene (41) is perfectly stable. If 37 and 38 are reduced in THF, again four-membered ring opening and a second reduction to the new tetralithio compounds 42 and 43 is observed.
The topics discussed show how a variety of high-resolution NMR techniques can be used in structural research in the field of organolithium compounds. Isotope shifts as well as homo- and heteronuclear shift correlations and nuclear Overhauser spectroscopy provide detailed informations about the aggregation behavior of lithiated carbon compounds which are important synthetic aids. In particular techniques which utilize the nuclides 6Li and 7Li yield valuable insights into the course of lithiation reactions, aggregate formation and dynamics.
As far as our own efforts in the field of organolithium chemistry and its NMR spectroscopy are concerned, I am deeply obliged to my coworkers for their enthusiastic collaboration. In particular I thank Werner Andres, Rainer Benken, Klaus Bergander, Bernd Böhler, Oswald Eppers, Thomas Fox, Maria-Eugenia Günther, Heike Hausmann, Runxi He, Dietmar Hüls, Martin Kreutz, Hans-Egbert Mons, Detlef Moskau, and Cirsten Zaeske for their fine contributions. Many ideas and stimulations came from my colleague, Prof. Dr. Adalbert Maercker, Siegen, and his coworkers, which is gratefully acknowledged. Thanks for financial support is due to the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie, and the Chemetall Company, Frankfurt.
1.Harris, R.K. In NMR and the Peridic Table; Harris, R.K.; Mann, B.E., eds.; Academic Press; London, p. 1, 1978. [ Links ]
2.Akitt, J.W. In Multinuclear NMR; Mason, J., ed.; Plenum Press; New York, p. 189, 1987. [ Links ]
3.Laszlo, P. In The Multinuclear Approach to NMR Spectroscopy; Lambert, J.B.; Riddell, F.G., eds.; D. Reidel; Dordrecht, Holland, p. 261, 1983. [ Links ]
5.Lutz, O. In The Multinuclear Approach to NMR Spectroscopy; Lambert, J.B.; Riddell, F.G., eds.; D. Reidel, Dordrecht, Holland, p. 297, 1983. [ Links ]
6.Drakenberg, T. Ann. Rep. NMR Spectrosc. 1986, 17, 231. [ Links ]
7.Laszlo, P.; Sodium-23 NMR, In Encyclopedia of NMR;. Grant, D.; Harris, R.K., eds.; Wiley; Chichester, v. 7, p. 4451, 1996. [ Links ]
9.Thomas, R.D. In Isotopes in the Physical and Biomedical Sciences; Buncel, E.; Jones, J.R., eds.; Elsevier, Amsterdam, p. 367, 1991. [ Links ]
10.Jackman, L.M. and Bortiatynski, J. In Advances in Carbanion Chemistry; Snieckus, V., ed.; Jai Press, Greenwich, Connecticut, p. 45, 1992. [ Links ]
11.Bauer, W.; Schleyer, P.v.R. In Advances in Carbanion Chemistry; Snieckus, V., ed.; Jai Press, Greenwich, Connecticut, 1992, p. 89. [ Links ]
12.Collum, D.B. Acc. Chem. Res. 1993, 26, 227. [ Links ]
13.a) Günther, H., Lithium NMR, Encyclopedia of NMR; Grant, D.M.; Harris, R.K., eds.; Wiley, 1996, v. 5, p. 2807; [ Links ]b) H. Günther, High-resolution 6,7Li-NMR of Organolithium Compounds, In Advanced Applications of NMR to Organometallic Chemistry, Gielen, M.; Willem, R.; Wrackmeyer, B., eds., Wiley & Sons, Chichester 1996. [ Links ]
14.Forsén, S.; Drakenberg, T.; Wennerstroem, H.; NMR Studies of Ion Binding in Biological Systems; Q. Rev. Biophys. 1987, 19, 83. [ Links ]
15.Johansson, C.; Drakenberg, T. Metal Ion NMR Studies of Ion Binding; Ann. Rep. NMR Spectrosc. 1990, 22, 1. [ Links ]
16.de Freitas, M. Alkali Metal Nuclear Magnetic Resonance; Methods Enzymol. 1993, 227, 78. [ Links ]
17.Miller, S.K.; Elgavish, G.A. Biol. Magn. Reson. 1992, 11, 159. [ Links ]
18.Allis, J.L. Ann. Rep. NMR Spectrosc. 1993, 26, 211. [ Links ]
19.Forsén, S.; Johansson, C.; Linse, S. Methods Enzymol. 1993, 227, 107. [ Links ]
20.Forsén, S.; Calcium Binding Proteins, In Encyclopedia of NMR; Grant, D. and Harris, R.K., Eds.; Wiley, Chichester, v. 2, p. 1092, 1996. [ Links ]
21.Kohler, S.J.; Kolodny, N.H. Sodium Magnetic Resonance Imaging and Chemical Shift Imaging, Progr. NMR Spectrosc. 1992, 24, 411. [ Links ]
22.Joseph, P.M. Sodium-23 Magnetic Resonance Imaging, In Encyclopedia of NMR; Grant, D.; Harris, R.K., eds.; Wiley, Chichester, v. 7, p. 4445, 1996. [ Links ]
23.Edwards, P.P.; Ellaboudy, A.; Holton, D.M.; Pype, N.C. Ann. Rep. NMR Spectrosc. 1988, 20, 315. [ Links ]
24.Tycko, R. Solid State Nucl. Magn. Reson. 1994, 3, 303. [ Links ]
25.Maniwa, Y.; Mizoguchi, K.; Kume, K. Kotai Butsuri 1994, 29, 839 (Japanese). [ Links ]
26.Tycko, R. J. Phys. Chem. Solids 1993, 54, 1712. [ Links ]
27.Lauginie, P.; Messaoudi, A.; Conard, J. Synth. Meth. 1993, 56, 3002. [ Links ]
28.Grandjean, J.; Laszlo, P. ACS Symp. Ser. 1989, 415, 396. [ Links ]
29.Chemical Society Specialists Reports: NMR Spectroscopy 1980 ff. [ Links ]
30.Chemical Society Specialists Reports: Spectroscopic Properties of Organometallic Compounds 1980 ff. [ Links ]
31.Bauer, W.; Magn. Reson. Chem. 1991, 29, 494. [ Links ]
32.Bauer, W.; Lochmann, L. J. Am. Chem. Soc. 1992, 114, 7482. [ Links ]
33.Benn, R.; Lehmkuhl, K.; Mehler, K.; Rufinska, A. Angew. Chem. Int. Ed. Engl. 1984, 23, 534. [ Links ]
34.Benn, R.; Rufinska, A. Angew. Chem. Int. Ed. Engl. 1986, 25, 861. [ Links ]
35.Lehmkuhl, K.; Mehler, K.; Benn, R.; Rufinska, A.; Kriger, C. Chem. Ber. 1986, 119, 1054. [ Links ]
36Lin, Y.-Y.; Ge, N.-H.; Hwang, L.-P. J. Magn. Reson. Ser. B, 1994, 103, 189. [ Links ]
37.For an early review see Batiz-Hernandez, H.; Bernheim, R.A. Progr. NMR Spectrosc. 1967, 3, 63. [ Links ]
38.Hansen, P.E. Ann. Rep. NMR Spectrosc. 1983, 15, 106 [ Links ]
39.Forsyth, D.A. In Isotopes in Organic Chemistry; Buncel, E.; Lee, C.C., eds.; Elsevier, Amsterdam, v. 6, p. 1, 1984. [ Links ]
40.Hansen, P.E. Progr. NMR Spectrosc. 1988, 15, 105. [ Links ]
41.Berger, S. In NMR - Basic Principles and Progress; Diehl, P.; Fluck, E.; Günther, H.; Kosfeld, R.; Seelig, J., eds.; Springer, Heidelberg, v. 22, p. 1, 1990. [ Links ]
42.For the latest reviews on theory and experiment of NMR isotope shifts see Jameson, C. In Encyclopedia of NMR; Grant, D.; Harris, R.K., eds.; Wiley, Chichester, 1996, v. 4, p. 2638; [ Links ]de Dios, A.C.; Jameson, C. Ann. Rep. NMR Spectrosc. 1994, 29, 1 [ Links ]
43.Wesener, J.R.; Moskau, D.; Günther, H.; J. Am. Chem. Soc. 1985, 107, 7307. [ Links ]
44.Wesener, J.R.; Günther, H. Tetrahedron Lett. 1982, 2845; [ Links ]Ernst, L.; Eltamany, S.; Hopf, H. J. Am. Chem. Soc. 1982, 104, 299; [ Links ]Ernst, L.; Hopf, H.; Wullbrand, D. J. Am. Chem. Soc. 1983, 105, 4469; [ Links ]Schaefer, T.; Peeling, J.; Wildman, T.A. Can. J. Chem. 1983, 61, 2777; [ Links ]Forsyth, D.A.; Botkin, J.H.; Osterman, V.M.; J. Am. Chem. Soc. 1984, 106, 7663; [ Links ]Forsyth, D.A.; Yang, J.-R. J. Am. Chem. Soc. 1986, 108, 2154; [ Links ]Siehl, H.; Walter, H. J. Chem. Soc. Chem. Commun. 1985, 76; [ Links ]Schaefer, T.; Peeling, J.; Sebastian, R. Can. J. Chem. 1986, 65, 534; [ Links ]Berger, S.; Diehl, B.W.K.; Künzer, H.; Chem. Ber. 1987, 120, 1059. [ Links ]
46.Aydin, R.; Günther, H. J. Am. Chem. Soc. 1981, 103, 1301 [ Links ]
47.Günther, H.; Jikeli, G. Chem. Ber. 1973, 106, 1863. [ Links ]
48.Reuben, J.; J. Am. Chem. Soc. 1983, 105, 3711. [ Links ]
49.Christofides, J.C.; Davies, D.B. J. Am. Chem. Soc. 1983, 105, 5099. [ Links ]
51.As a shift phenomenon, the measurement of NMR isotope effects profits from high B0 fields and the use of the lately introduced 750 or even 800 MHz instruments may well lead to the detection of hitherto unobserved isotope shifts.
52.Eppers, O. and Günther, H. Helv. Chim. Acta 1990, 73, 2071. [ Links ]
53.Seitz, L.M.; Brown, T.L. J. Am. Chem. Soc. 1966, 88, 2174. [ Links ]
54.Attempts to measure 2H-induced isotope shifts for 7Li sofar failed because of the larger linewidth as compared to 6Li, which furthermore increases at low temperature in the slow exchange limit due to increasing quadrupole relaxation.
55.Schmidbaur, H.; Schier, A.; Schubert, U. Chem. Ber. 1983, 116, 1938. [ Links ]
57.Bauer, W. In: Lithium Chemistry - A Theoretical and Experimental Overview; Saspe, A.M.; Schleyer, P.v.R., eds.; Wiley, Chichester, p. 125, 1995. [ Links ]
58.Novak, D.P.; Brown, D.L. J. Am. Chem. Soc. 1972, 94, 3793. [ Links ]
59.Waak, R.; Doran, M.A.; Baker, E.B. Chem. Commun. 1967, 1291. [ Links ]
60.For an outstanding review on experimental aspects of one- and two-dimensional NMR experiments see: a) Hull, W.E. In Two-Dimensional NMR Spectroscopy - Applications for Chemists and Biochemists; Croasmun, W.R.; Carlson, R.M.K., Eds.; VCH Publishers, Weinheim, p. 67, 1994; [ Links ]b) Berger, S.; Braun, S.; Kalinowski, H.-O. 150 and More Basic NMR Experiments, 2nd ed., Wiley-VCH, Weinheim, 1998. [ Links ]
61.Günther, H.; Moskau, D.; Dujardin, R. and Maercker, A. Tetrahedron Lett 1986, 27, 2251; [ Links ]Bauer, W.; Feigel, M.; Müller, G. and Schleyer, P.v.R. J. Am. Chem. Soc. 1988, 110, 6033; [ Links ]Gallagher, D.J.; Kerrick, S.T.; Beak, P. J. Am. Chem. Soc. 1992, 114, 5872. [ Links ]
63.Bax, A.; Freeman, R. J. Magn. Reson. 1981, 44, 542. [ Links ]
64.Eppers, O.; Fox, T. and Günther, H. Helv. Chim. Acta 1992, 75, 883. [ Links ]
65.Galiano-Roth, A.S.; Michaelides, E.M.; Collum, D.B. J. Am. Chem. Soc. 1988, 110, 2658; [ Links ]Galiano-Roth, A.S.; Collum, D.B. J. Am. Chem. Soc. 1989, 111, 6772; [ Links ]Romesberg, F.E.; Collum, D.B. J. Am. Chem. Soc. 1992, 114, 2112; [ Links ]Romesberg, F.E.; Gilchrist, J.H.; Harrison, A.T.; Fuller, D.J.; Collum, D.B. J. Am. Chem. Soc. 1991, 113, 5751. [ Links ]
66.Kim, Y.-J.; Bernstein, M.P.; Galiano-Roth, A.S.; Romersberg, F.E.; Williard, P.G.; Fuller, D.J.; Harrison, A.T.; Collum, D.B. J. Org. Chem. 1991, 56, 4435. [ Links ]
67.Levy, G.C.; Lichter, R.L. Nitrogen-15 Nuclear Magnetic Resonance Spectroscopy, Wiley; New York, p. 37ff, 1979. [ Links ]
68.Davis, D.G.; Bax, A. J. Am. Chem. Soc. 1985, 107, 7197. [ Links ]
69.Subramanian, S.; Bax, A. J. Magn. Reson. 1987, 71, 325. [ Links ]
70.Kessler, H.; Anders, U.; Gemmecker, G.; Steuernagel, S. J. Magn. Reson. 1989, 85, 1. [ Links ]
71.Bax, A.; Davis, D.G. J. Magn. Reson. 1985, 65, 355. [ Links ]
72.Lewis, H.L.; Brown, T.L. J. Am. Chem. Soc. 1970, 92, 4664. [ Links ]
73.Thomas, R.D.; Jensen, R.M.; Young, T.C. Organometallics 1987, 6, 565. [ Links ]
74.Bauer W. and Griesinger, C. J. Am. Chem. Soc. 1993, 115, 10871 (1H). [ Links ]
75.Mons, H.-E.; Günther, H.; Maercker, A. Chem. Ber. 1993, 126, 2747 (1H). [ Links ]
76.Moskau, D.; Brauers, F.; Günther, H.; Maercker, A. J. Am. Chem. Soc. 1987, 109, 5532 (13C). [ Links ]
77.Gais, H.-J.; Vollhardt, J.; Günther, H.; Moskau, D.; Lindner, H.J.; Braun, S. J. Am. Chem. Soc. 1988, 110, 978 (13C). [ Links ]
78.Gilchrist, J.H.; Harrison, A.T.; Fuller, D.J.; Collum, B.D. Magn. Reson. Chem. 1992, 30, 855 (15N). [ Links ]
79.Romesberg, F.E.; Bernstein, M.P.; Gilchrist, J.H.; Harrison, A.T.; Fuller, D.J.; Collum, D.B. J. Am. Chem. Soc. 1993, 115, 3475 (31P). [ Links ]
82.Garbow, J.R.; Weitekamp, D.P.; Pines, A. Chem. Phys. Lett. 1982, 93, 504. [ Links ]
83.Hurd, R.E.; John, B.K. J. Magn. Reson. 1991, 91, 468; [ Links ]Vuister, G.W.; Boelens, R.; Kaptein, R.; Hurd, R.E.; John, B.K.; van Zijl, P.C.M. J. Am. Chem. Soc. 1991, 113, 9688; [ Links ]Tyburn, J.-M.; Brereton, I.M.; Doddrell, D.M. J. Magn. Reson. 1992, 97, 305. [ Links ]
84.Böhler, B; Günther, H. Tetrahedron Lett. 1996, 37, 8723. [ Links ]
86.a) Bauer, W.; Schleyer, P.v.R.; Poster A.3 at the 10th European Experimental NMR Conference (EENC), Veldhoven, May 28 - June 1, 1990; [ Links ]b) Bauer, W. Magn. Reson. Chem. 1996, 34, 532. [ Links ]
87.Stonehouse, J.; Adell, P.; Keeler, J.; Shaka, A.J. J. Am. Chem. Soc. 1994, 116, 6037. [ Links ]
88.Stott, K.; Stonehouse, J.; Hwang, T.L.; Keeler, J.; Shaka, A.J. J. Am. Chem. Soc. 1995, 117, 4199. [ Links ]
89.Günther, H.; Eppers, O.; Hausmann, H.; Hüls, D.; Mons, H.-E.; Klein, K.-D. and Maercker, A. Helv. Chim. Acta 1995, 78, 1913. [ Links ]
90.Patterma, S.P.; Karle, T.L.; Stucky, G.D. J. Am. Chem. Soc. 1970, 92, 1130. [ Links ]
91.Beno, M.A.; Hope, H.; Olmstead, M.M.; Power, P.P. Organometallics 1985, 4, 2117. [ Links ]
92.Lipkowitz, K.B.; Uhegbu, C.; Naylor, A.M.; Vance, R. J. Comput. Chem. 1985, 6, 662; [ Links ]Sygula, A.; Rabideau, P.W. J. Org. Chem. 1987, 52, 3521; [ Links ]Sygula, A.; Rabideau, P.W. J. Am. Chem. Soc. 1992, 114, 821; [ Links ]Bühl, M.; v. Eikema Hommes, N.J.R.; Schleyer, P.v.R.; Fleischer, K.; Kutzelnigg, W. J. Am. Chem. Soc. 1991, 113, 2459. [ Links ]
93.Zarges, W.; Marsch, M.; Harms, K.; Boche, G. Chem. Ber. 1989, 122, 2303. [ Links ]
94.Ahlbrecht, H.; Harbach, J.; Hauck, T.; Kalinowski, H.-O. Chem. Ber. 1992, 125, 1753. [ Links ]
95.For a review of the relevant literature see Ref. 96.
96.Zarges, W.; Marsch, M.; Koch, W.; Frenking, G.; Boche, G. Chem. Ber. 1991, 124, 543. [ Links ]
97.Berger, S.; Müller, F. Chem. Ber. 1995, 128, 799. [ Links ]
98.West, R.; Waack, R. J. Am. Chem. Soc. 1967, 89, 4395. [ Links ]
99.Seebach, D.; Hässig, R. and Gabriel, J. Helv. Chim. Acta 1983, 66, 308. [ Links ]
100.see for example Schade, P.; Schäfer, T.; Müllen, K.; Bender, D.; Knoll, K.; Bronstert, K.; Chem. Ber. 1991, 124, 2833. [ Links ]
101.Hoell, D.; Lex, J.; Müllen, K. J. Am. Chem. Soc. 1986, 108, 5983. [ Links ]
103.Ruhland, T.; Hoffmann, R.W.; Schade, S.; Boche, G. Chem. Ber. 1995, 128, 551. [ Links ]
104.Fraenkel, G.; Martin, K.V. J. Am. Chem. Soc. 1995, 117, 10336. [ Links ]
105.Böhler, B.; Hüls, D.; Günther, H. Tetrahedron Lett. 1996, 37, 8719. [ Links ]
106.Sondheimer, F.; Acc. Chem. Res. 1972, 5, 81. [ Links ]
109.Günther, M.-E.; Aydin, R.; Buchmeier, W.; Engelen, B.; Günther, H. Chem. Ber. 1984, 117, 1069. [ Links ]
110.Benken, R.; Finneiser, K.; v. Puttkamer, H.; Günther, H.; Eliasson, B.; Edlund, U. Helv. Chim. Acta. 1986, 69, 955. [ Links ]
111.Schubert, U.; Neugebauer, W.; Schleyer, P.v.R. J. Chem. Soc. Chem. Commun. 1982, 1184. [ Links ]
112.Cremer, D.; Günther, H.; Liebigs Ann. Chem. 1972, 763, 87; [ Links ]Günther, H.; Vicinal1H,1H Coupling Constants in Cyclic p-Systems, Encyclopedia of NMR; Grant, D.M.; Harris, R.K., eds.; Wiley, v. 8, p. 4923, 1996. [ Links ]
113.Benken, R.; Günther, H. Helv. Chim. Acta. 1988, 71, 694. [ Links ]
Received: March 29, 1999