вход по аккаунту


Modern NMR Spectroscopy of Organolithium Compounds.

код для вставкиСкачать
Modern N M R Spectroscopy of Organolithium Compounds
By Harald Giinther,* Detlef Moskau, Peter Bast, and Dietmar Schmalz
Dedicated to Professor Emanuel Vogel on the occasion of his 60th birthday
Modern methods of N M R spectroscopy, in particular the two-dimensional techniques,
offer new chances for structure determinations in the field of organolithium compounds,
where the combination of 'H-, I3C-, and '"'Li-NMR spectroscopy is a n especially useful
feature. Chemical shift correlations which also include the lithium nuclei allow a complete
assignment of the 'H-, I3C-,and 6Li-NMR spectra and thereby a better characterization of
the various aggregates and complexes present in solution. Spatial proximities of 6Li and 'H
can be detected by nuclear Overhauser experiments, and 6'7'Li-NMR exchange spectroscopy can provide new information with regard to the mechanisms and energetics of dynamic processes like aggregate interchange and complexation. After a short resume of the
experimental aspects of the NMR spectroscopy of organolithium compounds and a discussion of the NMR parameters of these systems, new experimental techniques are presented.
Areas of application of these newly conceived one- and two-dimensional N M R experiments are illustrated with selected examples. The results show that even more detailed information about the structure and reactivity of organolithium compounds, which are so
important for organic synthesis, can be expected in the future.
1. Introduction
In the last ten years a large number of new experimental
techniques have been developed for nuclear magnetic resonance spectroscopy, the most important of which already
belong to the standard repertoire of NMR investigations of
molecular structure!'] Structure determinations of dissolved organolithium compounds, which are valuable reagents for organic synthesis,"] also profit from this development. Numerous X-ray structure investigations performed
in recent years have shown that this class of compounds
presents a fascinating variety of structures in the solid
state."] The same is true even more in solution, where organoiithium compounds of the type RLi usually d o not exist
as monomers but rather form oligomeric aggregates (RLi),?.
Association in hydrocarbon solvents leads to dimers, tetramers, hexamers, and even nonamers, which very often
coexist in fast, reversible equilibria (Fig. 1). In basic solvents like ethers o r upon the addition of complex-forming
amines like tetramethylethylenediamine (TMEDA) oligomerization is repressed in favor of complex formation
with the solvent or the amine. Much more complicated situations can be expected for polylithium compounds,
where research o n structures is only at its beginning.[4'
The reactivity of dissolved organolithium compoundsL2']
also varies with the structure, consequently a knowledge of
the structure is very important for planning synthetic strategies. N M R spectroscopy, as the most important physical
method for the determination of molecular structures in
solution, has therefore long occupied a dominating position in this field,L5'all the more so because various nuclei
[*I Prof. Dr. H. Giinther, Dr. D. Moskau, DipLBiol. P. Bast,
Dipl.-Chem. D. Schmalz
Fachbereich 8, Organische Chemie I1 der Universitat Gesarnthochschule
Postfach 10 1240, D-5900Siegen (FRG)
0 VCH Verlagsgesellsehafl mbH. 0-6940 Weinheim. 1987
can be employed as probes. Besides the well-known nuclides ' H and I3C, which are used for the characterization
of the organic ligands, two additional attractive magnetic
nuclei are available with the isotopes 'Li and 'Li. Despite
their quadrupole moment, both are well suited for NMR
measurements.[61 Because of the above mentioned equilibrium reactions between aggregates, the dynamic NMR
spectroscopy of organolithium compounds is likewise of
great importance, since temperature-dependent NMR
spectra enable the elucidation of the exchange mechanisms as well as the determination of the thermodynamic
and kinetic parameters of the equilibria.
Fig. I . Schematic representation of the structures of dimeric, tetrarneric, and
hexameric aggregates of organolithium compounds (RLi),, in solution IS].
Two-dimensional N M R spectroscopy (2D-NMR) will
undoubtedly have a significant impact on further progress
in the fields mentioned above. Thus, on the one hand, the
'H- and 13C-NMR spectra of the ligands may be assigned
more precisely with the help of homo- and heteronuclear
shift correlations, and, on the other, polarization transfer
and multiple quantum spectroscopy enable the development of new measuring techniques incorporating the lithium isotopes. The detection of nuclear Overhauser effects
between 6Li and neighboring protons provides a new approach for the measurement of Li-H distances in solution
and thus allows conclusions to be drawn about the regioselectivity during second metalation reactions. Finally,
0570-0833/87/1212-1212 $ 02.50/0
Angew. Chem. In!. Ed. Engl. 26 (1987) 1212-1220
two-dimensional exchange spectroscopy should gain importance for the study of the dynamic processes of solute
organolithium compounds. This progress report outlines
and critically evaluates the potential of these new experimental techniques. Technical aspects are emphasized and
a number of experiments are described for the first time.
2. Nuclear Properties
Table I lists the nuclear properties of the nuclides 'H,
l3C, 'Li and 7Li which are most important for N M R spectroscopy. Those of 'H and I3C are well-known and serve
here only for comparision.
Table I . Nuclear properties of the lithium isotopes 'Li and 'Li in comparision to ' H and "C [a].
N [%]
Spin I
- 8 x lo-'
- 4 . 5 ~lo-'
R ('lC)
1 . 5 4 ~lo3
5.68 x lo3
[a] Abbreviations and symbols: N = natural abundance; y=gyromagnetic ratio in 10' rad T - ' s - ' ; Q=quadrupole moment in lo-'' m2;vo resonance
frequency (MHz) at field strength Eo of 2.35 T ; R("C)=signal strength (receptivity), y Z N I ( N + I ) , rel. to "C.
Both lithium nuclei have a nuclear spin I > 1 / 2 and
therefore a quadrupole moment, which in the case of the
isotope 6Li, however, is so small that this nucleus behaves
nearly like a spin-1/2 nucleus and thus is well suited for
N M R investigations o n organolithium compound^.^'^ Line
broadening, a disturbing feature often found in the case of
nuclei with I > 1/2 because of fast quadrupole relaxation,
is rarely observed in .'Li-NMR spectra, and is also always
within reasonable limits in 7Li-NMR spectra. The resonance signals of both nuclei can easily be measured; however, scalar spin-spin couplings J to neighboring nuclei are
usually better resolved in the case of 'Li. This is especially
true for low-temperature measurements, which are indispensable for investigations dealing with the dynamics of
solute organolithium compounds, because faster 'Li quadrupole relaxation leads to line broadening and thereby
masks line splittings.181Thus, it is only 'Li-labeling which
in general paves the way for the measurement of the important structural parameter 'J(C,Li), as the pioneering
work of Fraenkel et
and Seebach et al.l'O1has shown. If
high sensitivity is important, additional '3C-labeling is advisable, which then provides optimal NMR conditions for
the spin system in question. Since 'Li labeling can be
achieved relatively easily-in the majority of the synthetic
routes metallic lithium is used directly o r indirectly
through a practicable lithium reagent like n-butyllithiumfzal-and since this isotope is easily obtainable," 'I the
reduced receptivity['"' of 'Li as compared to 7Li presents
no problems for the NMR experiment. Furthermore, 'Li
has the advantage that its resonance frequency is close to
that of the deuteron (58.86 as compared to 61.40 MHz at a
magnetic field strength of 9.4 T), so that with many spectrometers the 'H-lock channel can be used for 'Li-NMR
experiments, as for instance I3(J6Li) decoupling.
Angew. Chem. In!. Ed. Engl. 26 (1987) 1212-1220
2.1. Spin Lattice Relaxation Mechanisms
Because of its small quadrupole moment, other mechanisms besides the mechanism of quadrupole relaxation
must contribute to a considerable extent to the spin lattice
relaxation of 'Li. This was first shown by Wehrli,[7'who
found u p to 35% dipolar relaxation through 'Li-IH interactions in the case of alkyllithium compounds and who
was able to measure 'Li( 'H} nuclear Overhauser effects
(NOES) for the first time. NOE measurements provide a
valuable supplement to structural information obtained
through spin-spin couplings. 7Li relaxes almost completely via the quadrupole mechanism and exhibits much
shorter spin lattice relaxation times TI than 'Li, for which
values of 125 s are possible[121even in organolithium compounds. In case of solvated 'Li@ ions T, values of u p to
1000 s have been r n e a s ~ r e d ! ~Therefore,
for the determination of the 90" excitation pulse length, which is indispensible in FT-NMR spectroscopy, the steady-state meth~ dis suitable
[ ~ in~ the ~case of 6Li because it allows high
repetition rates despite long TI values. The different contributions of 'Li and 'Li to the spin lattice relaxation of
neighboring nuclei also provide a means for the determination of Li-C and Li-H distances in aggregates of solute
organolithium compounds through the use of the isotopic
substitution method,["] as has been shown by Jackman et
al. for phenyllithium.[''] If one replaces a nucleus "X (spin
quantum number I) with strong dipole-dipole interaction
by its isotope "'X (spin quantum number S) with weak dipole-dipole interaction, the difference in the spin lattice relaxation rate R1 of the neighboring I3C nucleus is given by
If the correlation' time z, is known, the distance r(I-x can be
determined according to
z, is obtained from the dipolar relaxation rate R P D of another 13C nucleus in the same molecule measured under
identical conditions and its respective C-H distance.
2.2. Chemical Shifts
The chemical shifts of the lithium nuclides have thus far
been the subject of very few investigations. Calc~lations[''~
and measurements o n aryllithium compounds['si have
shown that the diamagnetic anisotropy effect of the organic ligands strongly influences the 7Li resonance frequency in these systems. Since the chemical shift range is
extremely small (ca. 2-3 ppm, i.e. ca. 65 Hz for 'Li and ca.
120 Hz for 'Li at 2.35 T according to Table 1) and since the
resonance frequencies strongly depend on the medium and
concentration effects owing to aggregate formation, chemical shifts are not very informative for structure investigations on organolithium compounds. Furthermore, measurements are usually made with external references, and, in
addition, various reference compounds are used, thus
causing further uncertainties. Nevertheless, a diamagnetic
ring current effect has been detected for the 6Li resonance
of lithium salts of charged x-electron systems.["] In these
cases the 7Li shift can also be used for the study of ion pair
structures.12o1The 'H and I3C shifts in the organic groups
follow the well-known trends, where the charge density
plays no major role in the case of aryllithium compounds
RLi. Much more important are the anisotropy of the C-Li
bond (for 'H)12'J and the local paramagnetism caused by
n-+x* transitions (for 13C).1223
Typically, the situation for
phenyllithium is similar to that in ~ y r i d i n e . [ ~ ~ l
The 13C resonance frequencies of alkyllithium compounds RLi have recently been analyzed in more detail.Iz4]
Compared to the data for the corresponding hydrocarbons
RH, low-field shifts of 5.3f0.7 and 7.7k0.5 ppm have
been observed for the 0- and y-carbon atoms, respectively.
Steric interactions in branched alkyl groups diminish the
y-effect and can even give rise to high-field shifts. The resonance of the &carbon atom is not influenced by the metalation, while the effects on the a-carbon atom vary
strongly. For primary alkyl groups a shielding of 5.9k0.9
ppm was found which increases on a branching, i.e. in the
case of secondary and tertiary alkyllithium compounds.
For secondary systems a relationship between 6('3C,) and
the degree of aggregation was found: the shielding is larger
for tetramers than for hexamers.
tions. The magnitude of the 13C,6Li coupling constants
measured and the multiplicity of the I3C signal are directly
related to the degree of aggregation, n."'"] For such dynamic systems the empirical rule IJI= 18/n holds true. In
the case of fast exchange between different aggregates the
line splitting disappears. At room temperature the 6(7)LiN M R spectra of organolithium compounds therefore typically show a singlet, because the chemical shifts are averaged as well. "Static" spectra are obtained, on the other
hand, if the exchange rates are so small that 13C,6'7'Licoupling is limited to next neighbors within one aggregate.
Scalar couplings to 'Li are, according to the relation
~ ( ~ L i ) / y ( ~ Llarger
i ) , by a factor of 2.64 than coupling to
6Li. Besides coupling to I3C scalar interactions of the lithium nuclides with other nuclei have also been demonstrated; thus, for instance, 'J(7Li,3'P) (39-40 Hz),[27,281
1J(6'7'Li,29Si)(20-29, 51 HZ),~'~'
and for highly diluted solutions (0.01 M ) of the complex 1
(CS(CH3),)irHLi-PMDETA 1
PMDETA = pentamethyldiethylenetriamine
in the temperature range between 20 and -40°C even
J('Li,'H) (8.4 HZ).~~']
3. Applications of New Measuring Techniques
2.3. Scalar Spin-Spin Coupling
The measurement of '3C,6'7)Li coupling constant^^^^^ and
the determination of 13C-NMR signal multiplicities has in
recent years considerably increased our knowledge of the
structure of solute organolithium c o m p ~ u n d s . ~261~ *An
overview of the rules which have been established, based
on the data collection published by Seebach et aI.,['oal is
presented in Table 2. The values for the 'J('3C,6Li) coupling constant range from 2 to 18 Hz, whereby the smaller
values are a consequence of exchange processes within individual aggregates which are still fast on the NMR time
scale even at low temperature. Under these conditions, o n
time average, each I3C nucleus has contact with all lithium
atoms of the aggregate in question so that the coupling
with the 6Li nuclei is averaged effectively over all posi-
For the N M R spectroscopy of the organic ligands in organolithium compounds the entire arsenal of one and twodimensional pulse techniques1" available for 'H- and I3CN M R spectroscopy can be employed. This has already
been demonstrated in a few cases for the important homoand heteronuclear shift correlation^.^^'.^^^ Absolute assignments of I3C resonances are obtained most reliably
through the two-dimensional INADEQUATE experimentY3' which, however, because of its insensitivity-the
I3C satellites in the I3C NMR spectrum are detectedrequires concentrated samples. The experiment fails if the
relative chemical shift between neighboring I3C resonances falls below a critical value which depends on the
field strength and the resolution
In the case of
organolithium compounds a further limit must be seen in
the fact that it is difficult or even impossible to detect the
Table 2. Compilation of '3C,6Li, spin systems for aggregates of organolithium compounds of the type (RLi),, with characteristic data for the dependence of the
' I ('3C,6Li) coupling constants on the degree of aggregation n (n-propyllithium 136, 26d], fert-butyllithium [70a], phenyllithium (101, trichloromethyllithium [lo]).
dynamic [a]
static [a]
[a] "Dynamic" implies interaction with all lithium atoms of the aggregate in the fast exchange region; "static" implies selective interactions with a reduced number
of lithium atoms in the sense of a multicenter bond. The "C-multiplicity is given by 2.+ 1 (x=number of lithium nuclei); for 'Li by 3x+ I .
Angew. Chem. Int. Ed. Engl. 26 (1987) 1212-1220
resonance of the metalated carbon atom which is broadened because of coupling to 6Li,7Li. If the assignment of
this carbon is uncertain, the experiment should be repeated with 7Li decoupling or, in the case of 'Li-labeled
samples, with 6Li decoupling. The scope and the limits of
the INADEQUATE method are demonstrated in Figure 2
with the I3C-NMR spectrum of the l-naphthyllithiumTMEDA complex.[341
spectrum have to be detected. This measurement is facilitated by I3C labeling of the metalated position. In general,
for aggregates of organolithium compounds with labeled
a-carbon atom y = 1, since spin-spin coupling between individual organic groups in the aggregate is absent. In this
case, therefore, the simple first order multiplet patterns for
spectra of the 13C6Li, type are observed for the 13C resonances (Table 2). This situation is met with aggregates of
monolithium compounds of type (RLi),, where the labeled
carbon atom on time average can have contact with up to
nine lithium atoms.[361In the case of static spectra one
finds coordination with up to three 'Li nuclei. On the
other hand, more complicated signal patterns result if different lithium neighbors are present, a situation encountered in the case of polylithium compounds (see Section
2.3). In contrast, in the 'Li-NMR spectrum simple multiplets can always be expected, because homonuclear 'Li,6Li
couplings are negligibly small (see Section 2.2) and the metalated carbon atoms have no homonuclear coupling
partner if, as is usual, only this position in the molecule is
For the recognition of spin multiplets the J-modulated
spin echo techniques (attached proton test[371)have proven
of value in the case of I3C-NMR spectroscopy. They yield
for CX, groups (X = spin- 1/2 nucleus) signal phases of - 1
1 (CH,), and - 1 (CH,). Through gated I3C-decoupling a similar phase selection can be achieved for 'Li
signals which are coupled to 13C, as is shown in Figure 3a
with two signals from the 'Li-NMR spectrum of dilithio(phenylsulfonyl)trimethylsilylmethane (2).[3x1
Fig. 2. Two-dimensional I NADEQUATE-experiment [33] for I-naphthyllithium-TMEDA in the region of the resonances of C-2 to C-I0 (contour diagram; measuring frequency 100.61 MHz; reference TMS; ca. l M in C6D6
total experimental time 37 h; additional signals N = naphthalene, L=solvent). The resonances at 6= 148.6 and 133.7 have been identified as those of
quaternary carbons by a DEPT-experiment [35]. The assignment starts at C10, since only this carbon can show three correlation signals with direct
neighbors (a, b, c). The C,C connection proceeds from signal c via C-9 as the
only quarternary neighbor to C-4. It is then interrupted, because the strongly
coupled AB-system of C-3 and C-4 (A6 24.3 Hz, 5 ~ ~ H6z)0,degenerated to a
singulet, cannot be detected 1331. The remaining pair of signals d, e must
belong to C-2 and C-3 and the absolute assignment follows from the fact that
only the signal at 6= 124.39 can yield a strongly coupled AB system (B part
at 6 = 124.15). The results of this experiment are in agreement with the results
of a combined 'H- and "C,'H-shift correlation [31].
In the following we deal with experimental techniques
which have been developed on the basis of the new pulse
methods especially for the spectroscopy of the metalated
carbon atom and the metal, that is for the spin systems
I3C,6Li and I3C,'Li. Because of the far more favorable
NMR properties of 6Li the system I3C,6Li stands to the
3.1. Multiplicity Selection and J,6 Spectroscopy
The signal multiplicity of scalar coupled I3C,Li, spin
systems can be determined best in the I3C-NMR spectrum
(determination of x) for 6Li labeled compounds. If y is also
to be determined, the I3C satellite signals in the 'Li-NMR
Angew. Chem. I n [ .
Ed. Engl. 26 11987) 1212-1220
C6H5-S02-CLi2-Si(CH3)3 2
Fig. 3. a) Triplet (5=4.1 Hz) and doublet (J=5.6 Hz) for a "C2"Li- and a
"C'Li-spin system in the 58.86-MHz 'Li-NMR spectrum of 'Li- and ''Clabeled dilithio(phenylsulfonyl)trimethylsily~methane (2) (measuring frequency 58.86 MHz; ext. ref. 1 M LiBr/[Ds]THF). h) Phase selection with the
technique of gated "C-decoupling (measuring time I h). c) Two-dimensional
'J("C,6Li)-resolved 6Li-NMR spectrum (measuring time 5.7 h).
If the '3C,6Li-coupling constants over one bond differ
strongly, several experiments are necessary for optimal
phase selection. If one carries out a two-dimensional J.6NMR experiment,[391 the 'J('3C,6Li)-resolved 'Li-NMR
spectrum shown in Figure 3c is obtained.
Phase selection in the I3C-NMR spectrum through gated
6Li decoupling corresponds to the spin echo methods developed for I3C,'H spin systems,'401whereby the intensities
of the CX,, signals (X=spin-1 nucleus) with an evolution
time t , = 1 / 2 J are given by Equation (c):~~"]
For practical applications this procedure is less well
suited, since the signal intensity is strongly reduced already for n = 3 and the I3C signal disappears practically
completely for n > 3.
Besides the determination of multiplicity the two-dimensional J,6 spectroscopy allows a simplification of complicated spectra, because coupling to other nuclei only appears in the F2 dimension.[4” In this way the 3’P,6Li-coupling constants have been seperated from the homonuclear
3’P,3iP-couplingconstants in the 3’P-NMR spectra of lithiated organophosphorus compounds.[281
Since scalar spin-spin coupling constants-and only
these can be measured in isotropic media-are transmitted
in general through covalent bonds, it seems reasonable to
ask whether the observation of 6(7)Li,X coupling allows
conclusions to be drawn about the nature of the corresponding Li-X bonds (ionic or covalent). In this connection Streirwieser et al.1421
pointed out for the first time that
scalar spin-spin coupling between carbon and lithium
does not necessarily imply covalent bonding. The mechanism which they termed spin polarization and which is
known in N M R spectroscopy as through-space coupling1431
is believed to operate in several cases where I9F nuclei are
involved. In general it is more an exception than the rule.
It requires non-bonded interactions of neighboring orbitals
which are close in space. Also in the case of a purely ionic
Li-X bond scalar Li,X coupling could therefore operate, if
only the orbital contact is sufficient.
Several arguments can be cited in favor of this interpretation. It appears, that 13C,6(7)Li
couplings over more than
one bond have, to the best of our knowledge, so far not
been observed. In case of a through-bond mechanism however, also geminal and vicinal ‘3C,6‘7)Licouplings are to be
expected. Furthermore, many well resolved 7Li couplings
are not compatable with a high covalent character of the
carbon lithium bond, since the unsymmetrical charge distribution should increase the ’Li quadrupole relaxation
rate and broaden the I3C-NMR
Considering the
mechanistic dichotomy for the scalar spin-spin interaction
(through-bond o r through-space) it cannot so far be decided if the experimentally measured scalar spin-spin couplings of the lithium nuclides are a consequence of a covalent contribution to the respective Li-X bonds or if they
are transmitted by a pure through-space mechanism. In
view of the dominancy of the through-bond mechanism
with other nuclei the second possibility would certainly be
3.2. Homonuclear Shift Correlations
Shift correlations are without doubt the most important
two-dimensional pulse techniques. One distinguishes between scalar and dipolar correlated spectra. The origin of
the cross peaks observed in the 2D-NMR spectrum is
scalar spin-spin coupling in the first case and the nuclear
Overhauser effect in the second. That is, information is obtained, on the one hand about the bonding in the molecule
under study, and, on the other, about the spatial surroundings.
Homonuclear 6,6 o r COSY
for “Li or ’Li rest
on corresponding homonuclear spin-spin coupling. So far
this type of coupling has not been observed conventionally, that is by line splitting. Since most organolithium compounds, for reasons of symmetry, show only one “Li- o r
7Li-NMR signal for the individual aggregates, even at low
temperature (see Fig. l), such measurements had to be performed on systems with a 6Li/7Li isotope ratio of ca. 1 : 1.
From the line width in the 7Li-NMR spectrum of methyllithium, which forms a t e t ~ a m e r , [ ~Brown
et al.1471concluded that in this case a 6Li,7Li coupling over one bond of
about 0.3 Hz exists.
With the COSY
we were able to detect a
homonuclear 6Li,“Li coupling for the first time for 3,4-dilithio-2,5-dimethyl-2,4-hexadiene
( 3).1491
The crystal structure of this compound (Fig. 4)[”] revealed two different Li
o c
o Et,O
Fig. 4. Crystal structure of 3,4-dilithio-2,5-dirnethyl2,4-hexadiene (3) 1501.
positions. In T H F as solvent, 3 shows two “Li-NMR signals at -78°C; the COSY experiment (Fig. 5 ) proves that
a scalar spin-spin coupling exists between the two “Li signals. This suggests that 3 has the same structure in both
the solid state and in solution. Both “Li-NMR signals must
be assigned to the same aggregate, which, therefore, like
the molecule in the crystal, must contain two types of lithium atoms. The 6Li,6Li coupling should be of the order of
0.1 Hz. This is in good agreement with the value estimated
by Brown et al. for the 6Li,7Licoupling, if the corresponding y factors are taken into account. So far, it is unknown
if the coupling is a one-bond or a geminal interaction
which includes the metalated carbon in the coupling
COSY experiments for quadrupole nuclei have so far
been successfully performed for ”B,[”l s’V,[521
and in
a few cases for 14N.1543
Broad lines and/or small scalar
couplings have negative consequences and may prevent
the detection of cross peaks. According to the product operator formalism[551the dependency of the cross peak intensity, Zk, on the evolution time t , for COSY experiments
with spin-I nuclei is governed by the following equation:
Angew. Chem. Inl. Ed. Engl. 26 (1987) 1212-1220
between 6Li or I3C detection, are then applicable. Such experiments form the basis for a successful simulation of the
respective 13C,Ti spin systems.
0.0 -0.2 -04 -0.6
Fig. 5. 6Li,6Li COSY spectrum of 3.4dilithi0-2.5-dimethyI-2.4-hexadiene (3)
[49], recorded with the pulse sequence
proposed for emphasizing small couplings 90:-t,-A,
909, A, FID(t2) [45]
(measuring frequency 58.86 MHz; measuring time 1.2 h).
A plot of Equation (d) (Fig. 6) shows that for small couplings and short transverse relaxation times T2 the cross
peak intensity has a maximum if the condition t , = T2
holds. Experiments with the quadrupole nucleus ’H show
good results with the pulse sequence (I),[“’’ which was developed in order to emphasize small couplings and to suppress diagonal signals.
90e-tt -A, 45; , A -FID(/2)
where t , (max)+A=T2
Fig. 7. Pulse sequences for two-dimensional heteronuclear shift correldtions
on the basis of polarization transfer (a) [56]and multiple quantum coherences (b) 1571.
The dilithiosulfone 2 already mentioned above offered
the possibility of testing the use of the 6Li,13Cshift correlation, because the 13C- as well as the 6Li-NMR spectrum of
this compound displays a number of simple as well as
complicated multiplets (Fig. 8).[’*] Considering the spectrometer configuration available, the two-dimensional experiments could be performed in either case (a) or (b) only
with 6Li detection. Method (a) failed, apparently because
of fast transverse relaxation, and difficulties with the correct choice of the A I interval, while method (b) was successful (Fig. 8d). All correlations obtained were confirmed
by selective I3C decoupling. The double-resonance technique yielded two additional correlations which were not
found in the 2D-NMR experiments, apparently because of
small scalar couplings.
Fig.6. Cross-peak intensity for COSY experiments with nuclei of spin I= 1
and small couplings as a function of the transverse relaxation time T, (0.5,
0.1, 0.05 s) and the evolution time I , .
As expected, the COSY experiment for ’Li with the
larger quadrupole moment is less favorable. Nevertheless,
homonuclear 7Li,7Licouplings for 3 were also successfully
detected, even if less clear then in the case of 6Li.1541
3.3. Heteronuclear Shift Correlations
In principle, there are two pulse sequences available today for the measurement of heteronuclear shift correlations: the standard technique, which is based on polarization transfer (Fig. 7a),[561and a variant (Fig. 7b),[571which
uses the possibiIities of multiple quantum coherences. In
the case of organolithium compounds the 6Li,”C correlation is of primary interest, because the identification of
‘Li,13C spin pairs via the multiplicities in the 13C- o r 6LiNMR spectra is no longer possible in more complicated
cases. Only selective decoupling or one of the two-dimensional methods mentioned above, where one can choose
Angew. Chem. Int. Ed. Engl. 26 (1987) 1212-1220
50 -
LO -
Fig. 8. a) ‘H.6Li decoupled 100.61-MHz “ C - N M R spectrum of the dilithtosulfone 2 ( - 103°C in [D,]THF); b) “C multiplets with ‘Li coupling; c)
58.86-MHz % - N M R spectrum of ”C-labeled 2 with ”C,6Li coupling (ext.
reference, see Fig. 3); d ) Two-dimensional 6Li,’3C-shiftcorrelation for 2 with
the pulse sequence of Figure 7b; the correlations X have been detected with
selective ‘Li{”CJ-spin decoupling [58] (measuring time 4.7 h).
The potential of the 6Li I3C-shift correlation experiment
is also demonstrated by its application to the dilithium
compound 3.is91Here, the “C-NMR spectrum shows all
signals of the tetramethylbutadiene structure twice, and the
task was to assign the individual resonance lines to the “inner” and “outer” ligands (Fig. 4). This was achieved
through the finding that one of the metalated carbon
atoms couples to two different lithium nuclei, namely to
one of the “inner” and to one of the “outer” positions. The
two-dimensional 6Li,’3Cdouble quantum spectrum (Fig. 9)
shows two cross peaks for the high-field I3C and for the
high-field 6Li signal. Considering the direction of the carbanion orbitals at C-3 and C-4 given by the geometry of
the ligands as found in the crystal (Fig. 4), only the carbon
of the “outer” ligand can take part in two different scalar
interactions and thus show two correlation signals. The results of dynamic N M R spectroscopy, 2D-’H exchange and
2D-l3C,’H-6,6-NMR spectroscopy completed the assignment of the resonance signals.
6Li (F,)
for 3’P,‘H twofor the pair I3C,’H and by Yu and
dimensionally with the pulse sequence (2) (HOESY[631);
( I = ‘H, I9F, 3’P; S = i3C, I5N, 6Li):
90:-t, -9O:-t,,, -BB
In this way hydrogen bridges between lithium and boron
as well as aluminum, indium or gallium in hydrido[tris(trimethylsilyl)methyl]metalates could be detectedia1. 2DN M R experiments with the pulse sequence (2) and a 1 : 1
aggregate of methyl- and l - n a p h t h y l l i t h i ~ r nas
~ ~well
’ ~ as
2-lithio- l - p h e n y l p y r r ~ l e allowed
‘ ~ ~ ~ the localization of spatially closely related 6Li and ‘H nuclei ( < c a . 350 nm).
Since short 6Li-’H distances correlate with the activation
of the respective C H bond for a second metalation, NOE
spectroscopy also yields information about the reactivity
of the organic ligands. Quantitative measurements of NOE
build-up rates should here, as in the case of ‘H-NMR
spectroscopy,1661enable a more exact determination of nuclear distances. Figure 10 shows the result of a lD-6Li(’H)NOE difference experiment for the complex 1 -naphthyllithium-TMEDA. Compared to the 2D-NMR experimentI3”
the selective detection of 6Li- ‘H neighborhoods is simpler,
because a better digital resolution can be achieved.
Fig. 9. Two-dimensional “Li,’%shift correlation based on multiple quantum
(3) [59] (measuring
coherences for 3,4-dilithio-2,5-dimethyl-2,4-hexadiene
time 8.7 h).
3.4. Nuclear Overhauser Effects
The nuclear Overhauser enhancement (NOE)i601is based
on dipolar cross relaxation between closely spaced nuclei,
which have no chemical bond in common. It is therefore
an important supplement for the identification of chemical
structures on the basis of scalar spin-spin coupling constants, which are transmitted nearly without exception
through bonding pathways. Consequently, N M R spectroscopy is the only method which allows one to determine
three-dimensional molecular structures in solution.
For the N M R spectroscopy of solute organolithium
compounds it is important that neighboring protons can
make a significant contribution to the relaxation of ‘Li
through dipolar interaction^'^] (see Section 2.1). The proximity of 6Li and ’ H can, therefore, be detected by heteronuclear 6Li(‘HJ-NOEspectroscopy, where a maximum signal enhancement of 340% would result for the ‘Li nucleus
if no competing relaxation mechanisms were available. For
7Li such experiments are impossible because of its different relaxation behavior.
Heteronuclear NOE experiments can be performed onedimensionally (1 D), best employing the method of difference spectroscopy‘“] or, as was first shown by Rinaldi1621
Fig. 10. Selective 58.86-MHz 6Li{’HJnuclear Overhauser difference spectrum
for I-naphthyllithium-TMEDA with natural 6Li abundance; a) irradiation at
2-H, b) irradiation at 8-H, c) control experiment with off-resonance decoupling (measuring time 4 h each). The NOE effects are 8.5% and 7%, respectively. The decoupler power was calibrated with the SP1-experiment for chloroform’68’;for the assignment of the ’H-resonances see Ref. [3 I].
3.5. 2D-6Li,6Li- and 7Li,7Li-NMR Exchange Spectroscopy
Whether homonuclear NOE effects for 6Li,6Li nuclear
pairs can be detected or not has so far not been investigated. However, Wehrli estimated only 6% for the dipolar
contribution to the 6Li relaxation in 6Li,7Li pairs in met h y l l i t h i ~ r n . ~For
~ ] 6Li,6Li pairs this contribution reduces
further to less then 1% and is thus practically negligible.
O n the other hand, 2D-6Li,6Li and 2D-7Li,7Li-NMR exchange spectra for dynamic systems can be measured without difficulty in the region of slow exchange with the pulse
sequence (3),[“”
as is shown in Figure l l a for 6Li in the case of 3 and in
Figure l l b for 7Li in the case of an interaggregate exchange process. The investigation of dynamic processes of
solute organolithium compounds by N M R line shape analysi~[’~]
thus receives valuable support through 2D-NMR
experiments, since not only can exchange mechanisms be
Angew. Chem. Inr. Ed. Engl. 26 (1987) 1212-1220
recognized, but, in addition, quantitative analysis of cross
peak intensities allows the determination of rate constant~.[~']
mation bridging the gap between the results of crystal
structure analysis on the one hand and NMR spectroscopy
in solution on the other. The investigation of methyllithium and dilithi~methane[~~]
mentioned above shows that
6Li decoupling is also possible in solid state NMR spectroscopy and that the quality of the I3C-NMR spectra
thereby increases considerably.
4. Outlook
0.0 -0.2 -0.4 -0.6
Fig. 1 I . a) 2 D-'Li,"Li-NMR exchange spectrum for the intra-aggregate 6Li
(3) ([D8]THF, - 83"C,
exchange in 3,4-dilithio-2,5-dimethyl-2,4-hexadiene
measuring frequency 58.86 MHz, ext. reference, see Fig. 3, mixing time 1 s,
measuring time 1 h); above, the ID-NMR spectrum for comparision). b)
Two-dimensional 'Li,'Li-NMR exchange spectrum for the inter-aggregate
'Li exchange between the diethyl ether complex of the dimer of 8-dimethylamino-I-naphthyllithium (signal b) and the ether-free aggregate (signal a)
([D,]toluene [69], - 15°C; measuring frequency 155.51 MHz; ext. reference
1 M LiCI/DzO; mixing time 0.1 s; measuring time 45 min).
3.6. High Resolution Solid State NMR Spectroscopy
This timely and promising technique[721is without doubt
also of interest for the field of organolithium compounds,
despite the fact that so far only one investigation has been
The reason for this may be seen in the relatively large amount of material necessary for such
measurements (200-300 mg) and in the more complicated
sample preparation required for compounds that are sensitive to air and moisture. I3C- as well as 6Li-NMR measurements on solids should furnish important structural inforAngew. Chem. I n t . Ed. Engl. 26 (1987) 1212-1220
The experiments discussed in the foregoing sections demonstrate how the structure of organolithium compounds
in solution may be determined even more confidently and
in more detail in the future. Our knowledge about the reactivity of this important class of compounds can thus be
deepened and defined more exactly. More emphasis
should be devoted to the elucidation and the quantitative
investigation of the mechanisms of the dynamic processes
present in solution. 2D-NMR methods can certainly yield
new insights in this respect. From the standpoint of bonding theory further investigations of the mechanism of
'3C:(7)Li spin-spin coupling as well as the dependence of
this parameter on structure are desirable. This is even more
true for couplings between lithium nuclei. Finally, interesting results can be expected from the solid state NMR spectroscopy of organolithium compounds, especially since
this method can also detect dynamic phenomena in the
~ r y s t a l . ~Organolithium
compounds will therefore remain exciting subjects for demanding NMR studies in the
We are indebted to Prof. Dr. A . Maercker, Siegen. for a
number of important references. Our thanks are due to him
and Prof. Dr. H.-J. Gais, Freiburg, as well as to their coworkers Dr. R . Dujardin and Dip1.-Chem. F. Brauers, Siegen, and Dip1.-Chem. J. Vollhardt, Freiburg, for gifts of valuable compounds and exciting collaboration. Generous support of our investigations by the Fonds der Chemischen Industrie, is gratefully acknowledged.
Received: July 24, 1987:
revised: August 21, 1987 [A 644 IE]
German version: Angew. Chem. 99 (1987) 1242
[I] a) J. K. M. Sanders, B. K. Hunter: Modern NMR Spectroscopy, Oxford
University Press, Oxford 1987; b) R. R. Ernst, G. Bodenhausen, A. Wokaun, Principles of Nuclear Magnetic Resonance in One and Two Dimensions. Clarendon Press, Oxford 1987; c) R. Benn, H. Gunther, Angew.
Chem. 95 (1983) 381; Angew. Chem. Int. Ed. Engl. 22 (1983) 351; d) G.
A. Morris, Magn. Reson. Chem. 24 (1985) 371; e) review on metal NMR
spectroscopy: R. Benn, A. Rufihska, Angew. Chem. 98 (1986) 851; Angew. Chem. In[. Ed. Engl. 25 (1986) 861.
[2] a) L. Brandsma, H. Verkruijsse: Preparative Polar Organometal~icChemistry, Vol. 1, Springer, Berlin 1987; b) B. J. Wakefield: The Chemistry of
Organolithium Compounds, Pergamon Press, Oxford 1974; c) M.
Schlosser: Polare Organornetol/e. Springer, Heidelberg 1973.
131 a) P. von R. Schleyer, Pure Appl. Chem. 56 (1984) 151 ; b) W. N. Setzer, P.
von R. Schleyer, Adu. Orgonomer. Chem. 24 (1985) 353.
[4] A. Maercker, M. Theis, Top. Curr. Chem. 138 (1987) I.
[S] T. L. Brown, Ace. Chem. Res. I (1968) 23.
[6] a) B. Lindman, S. Forsin in R. K. Harris, B. E. Mann (Eds.): NMR and
the Periodic Table, Academic Press, London 1978, p. 129-194; b) P.
Laszlo in J. B. Lambert, F. G. Riddell (Eds.): The MultznuclearApproach
to NMR Spectroscopy, Reidel, Dordrecht 1982, p. 261-308; c) F. W.
Wehrli, Annu. Rep. NMR Spectrosc. 9 (1979) 125.
171 F. W. Wehrli, Org. Magn. Reson. I 1 (1978) 106.
181 G.E. Hartwell, A. Allerhand, J. Am. Chem. SOC.93 (1971) 4415.
19) G . Fraenkel, A. M. Fraenkel, M. J. Geckle, F. Schloss, J. Am. Chem. SOC.
I01 (1979) 4745.
[lo] a) D. Seebach, R. Hlssig, J. Gabriel, Helv. Chim. Acfa 66 (1983)
308; b)
D. Seebach, H. Siegel, J. Gabriel, R. Hassig. ibid. 63 (1980) 2046.
[I I] Martin Marietta Energy System Inc., Oak Ridge, T N (USA).
1121 R. D. Thomas, D H. Ellington, M. T. Clarke, unpublished, cited in
[I31 F. W. Wehrli, J . M a p . Reson 23 (1976) 527.
[I41 J. R. Wesener, H. Giinther, J . M a p . Reson. 62 (1985) 158.
( I S ] L. M. Jackman, L. M. Scarmoutzos, J Am. Chem. SOC. 106 (1984)
1161 L. M. Jackman, J. C. Trewella, J. Am. Chem. SOC.98 (1976) 5712; ibid.
101 (1979) 6428.
1171 P. A. Scherr, R. J. Hogan, J. P. Oliver, J. Am. Chem. SOC.96 (1974)
[I81 J. A. Ladd, J. Parker, J. Chem. Soc. Dalton Trans. 1972. 930.
1191 R. H. Cox, H. W. Terry, Jr., L. W. Harrison, Tetrahedron Lett. 1971.
[20] R. H. Cox, H. W. Terry, Jr., L. W. Harrison, J . Am. Chem. Sor. 93 (1971)
[21] G. Fraenkel, D. G. Adams, R. R. Dean, J . Phys. Chem. 72 (1968) 944; G.
Fraenkel, S. Dayagi, S . Kobayashi, h i d . 72 (1968) 953.
[22] A. J. Jones, D. M. Grant, J. G. Russell, G. Fraenkel, J. Pfiys. Cfiem. 73
(1969) 1624.
I231 V. M. S. Gil, J. N. Murrell, Trans. Faraday SOC.60 (1964) 248.
[24] R. D. Thomas, M. T. Clarke, T. C. Young, J. Organomet. Chem. 328
(1987) 239; see also J. P. C . M van Dongen, H. W. D. van Dijkman, M.
J. A. d e Bie, Recl. Trav. Chim. Pay-Bas 93 (1974) 29.
1251 L. D. McKeever, R. Waack, M. A. Doran, E. B. Baker, J. Am. Chem
Soc. 90 (1968) 3244.
[26] a) G. Fraenkel, H. Hsu, B. M. Su in R. 0. Bach (Ed.): Lithium: Current
Applications in Science. Medicine. and Technology, Wiley, New York
1985, p. 273; b) G. Fraenkel, P. Pramanik, J . Chem. Soc. Chem. Commun. 1983. 1527; c) R. Hassig, D. Seebach, Helv. Chim. Acta 66 (1983)
2269; d ) D. Seebach, J. Gabriel, R. Hassig, ibid. 67 (1984) 1083; e) R. D.
Thomas, R. M. Jensen, T. C. Young, Organometallics 6 (1987) 565; f) W.
Bauer, W. R. Winchester, P. von R. Schleyer, ibid. 6 (1987), in press.
(271 1. J. Colquhohn, H. C. E. McFarlane, W. McFarlane, J. Chem. SOC.
Chem. Commun. 1982, 220
[28] 1. J. Colquhoun, H. C. E. McFarlane, W. McFarlane, Phosphorus Sulfur
18 (1983) 61.
[29] U. Edlund, T. Lejon, T. K. Venkatachalam, E. Buncel, J . Am. Chem. SOC.
107 (1985) 6408.
(301 T. M. Gilbert, R. G. Bergmann, J. Am. Chem. Soc. 107 (1985) 6391.
[ 3 l ] W. Bauer, T. Clark, P. von R. Schleyer, J. Am. Chem. SOC.109 (1987)
[32] A. Rajca, L. M. Tolbert, J . Am. Chem. SOC.109 (1987) 1782.
(331 A. Bax, R. Freeman, T. A. Frenkiel, J. Am. Chem. SOC.103 (1981) 2102;
A. Bax, R. Freeman, T. A. Frenkiel, M. H. Levitt, J. Magn. Reson. 44
(1981) 212; review: J. Buddrus, H. Bauer, Angew. Chem. 99 (1987) 642;
Angew. Chem. Int. Ed. Engl. 26 (1987) 625.
(341 Synthese: W. Neugebauer, T. Clark, P. von R. Schleyer, Chern. Ber. 116
(1983) 3283.
1351 D. M. Doddrell, D. T. Pegg, M. R. Bendall, J. M a p . Reson. 48 (1982)
323; D. T. Pegg, D. M. Doddrell, M. R. Bendall, J . Chem. Phys. 77
(1982) 2745.
[36] G. Fraenkel, M. Henrichs, M. Hewitt, M. J. Geckle, B. M. Su, J. Am.
Chem. Soc. 102 (1980) 3345.
[37] S. L. Patt, J. N. Shoolery, J . Magn. Reson. 46 (1982) 318.
1381 J. Vollhardt, H.-J. Gais, K. L. Lukas, Angew. Chem. 97 (1985) 695; Angew. Chem. Int. Ed. Engl. 24 (1985) 696.
[39] a) G. Bodenhausen, R. Freeman, D. L. Turner, J. Chem. Phys. 75 (1976)
839; b) R. Freeman, S. P. Kempsell, M. H. Levitt, J. M a p . Reson. 34
(1979) 663; c) see also H. Giinther, P. Schmitt, Kontakfe Darmstadt
1985. No. 2, p. 3.
[40] a) P. Schmitt, J. R. Wesener, H. Giinther, J . Magn. Reson. 52 (1983) 5 11;
b) J. R. Wesener, H. Giinther, Org. Magn. Reson. 21 (1983) 433; c) J. R.
Wesener, P. Schmitt, H. Giinther, J . Am. Chem. Soc. I06 (1984) 10.
I411 L. D. Hall, S. Sukumar, J. Am. Chem. SOC. 101 (1979) 3120; see also
[42] A. Streitwieser, Jr., J. E. Williams, Jr., S. Alexandratos, J. M. McKelvey,
J . Am. Chem. SOC.98 (1976) 4778.
I431 J . Hilton, L. H. Sutcliffe, Prog. NMR Specfrosc. 10 (1975) 27.
[44] E. Buncel, T. K. Venkatachalam, U. Edlund, Can. J. Chem. 64 (1986)
[45] A. Bax, R. Freeman, J. Magn. Reson. 44 (1981) 542; A. Bax: Two-Dimensional NMR in Liquids, Reidel, Dordrecht 1983; see also H. Giinther, D.
Moskau, Kontakte Darmsfadt 1986. No. 2, p. 41.
[461 L. D. McKeever, R. Waak, M. A. Doran, E. B. Baker, J. Am. Chem. SOC.
91 (1969) 1057.
1471 T. L. Brown, L. M. Seitz, B. Y . Kimura, J. Am. Chem. SOC.90 (1968)
[48] A. Maercker, R. Dujardin, Angew. Chem. 97 (1985) 612: Angew Chem.
I n f . Ed. Engl. 24 (1985) 571.
1491 H. Giinther, D. Moskau, R. Dujardin, A. Maercker, Tetrahedron Letf. 27
(1986) 2251.
[SO] A. Maercker, R. Dujardin, B. Engelen, W. Buchmeier, M. Jung, unpub-
lished results.
[51] T. L. Venable, W. C. Hutton, R. N. Grimes, J. Am. Chem. SOC. 106
(1984) 29.
1521 P. J. Domaille, J . Am. Chem. Sor. 106 (1984) 7677.
[53] D. Moskau, H. Giinther, Angew. Chem. 99 (1987) 151 ;Angew. Chem. Int.
Ed. Engl. 26 (1987) 156.
1541 D. Moskau, Dissertation, Universitat-Gesamthochschule Siegen 1987.
[SS] 0. W. Sorensen, G. W. Eich, M. H. Levitt, G. Bodenhausen, R. R. Ernst,
Prog. NMR Spectrosc. 16 (1984) 163.
[56] A. A. Maudsley, L. Miiller, R. R. Ernst, J . Magn. Reson. 28 (1977) 463;
G . Bodenhausen, R. Freeman, ibid. 28 (1977) 47 1
I571 L. Miiller, J Am. Chem. SOC.I01 (1979) 4481; A. Bax, R. H. Griffey, B.
L. Hawkins, J . Magn. Reson. 55 (1983) 301.
[58] H.-J. Gais, J. Vollhardt, H. Giinther, D. Moskau, H. J. Lindner, D.
Braun, J . Am. Chem. press.
1591 D. Moskau, F. Brauers, H. Giinther, A. Maercker, J . Am. Chem. SOC.109
(1987) 5532.
1601 J. H. Noggle, R. E. Schirner: The Nuclear Overhauser Effekt, Academic
Press, New York 1971.
[61] J. K. M. Sanders, J. D. Mersh, Prog. NMR Spectrosc. 15 (1983) 353.
(621 P. L. Rinaldi, J . Am. Chem. Soc. 105 (1983) 5167.
1631 C. Yu, G. C. Levy, J . Am. Chem. SOC.105 (1983) 6995.
1641 A. G. Avent, C. Eaborn, M. N. A. El-Kheli, M. E. Molla, J. D. Smith, A.
C. Sullivan, J . Am. Chem. SOC.108 (1986) 3854.
[65] W. Bauer, G. Miiller, R. Pi, P. von R. Schleyer, Angew. Chem. 98 (1986)
1130; Angew. Chem. Int. Ed. Engl. 25 (1986) 1103.
1661 A. Kumar, G. Wagner, R. R. Ernst, K. Wiithrich, J . Am. Chem. SOC.103
(1981) 3654.
[67] B. H. Meier, R. R. Ernst, J . Am. Chem. SOC. I01 (1979) 6441; J. Jeener
B. H. Meier, P. Bachmann, R. R. Ernst, J. Chem. Phys. 71 (1979) 4546
review: R. Willem, Prog. NMR Spectrosc. 20 (1987) 1.
1681 K. G. R. Pachler, P. L. Wessels, J . Magn. Reson 12 (1973) 337.
1691 J. T. B. H. Jastrzebski, G. van Koten, K. Goubitz, C. Arlen, M. Pfeffer, J .
Organomet. Chem. 246 (1983) C75.
(701 Recent examples: a) R. D. Thomas, M. T. Clarke, R. M. Jensen, T. C.
Young, Organometallics 5 (1986) 1851; b) J . Heinzer, J. F. M. Oth, D.
Seebach, Helu. Chim. Acta 68 (1985) 1848.
[71] E. W. Abel, T. P. J. Coston, K. G. Orrell, V. Sik, D. Stephenson, J.
M a p . Reson. 70 (1986) 34, and references cited therein.
(721 C. A. Fyfe: Solid Stare NMRfor Chemisfs.C. F. C. Press, Guelph, Cana d a 1983.
1731 J. A. Gurak, J. W. Chinn, Jr., R. J. Lagow, H. Steinfink, C. S. Yannoni,
Inorg. Chem. 23 (1984) 3717.
(741 J. R. Lyerla, C. S. Yannoni, C. A. Fyfe, Arc. Chem. Res. I5 (1982) 208.
[75] a) B. H. Meier, C. B. Storm, W. L. Earl, J. Am. Chem. SOC.I08 (1986)
6072; b) R. Benn, H. Grondey, G. Erker, R. Nolte,
Angew. Chem. Int. Ed. Engl. 26 (1987) 1212- I220
Без категории
Размер файла
914 Кб
spectroscopy, nmr, compounds, organolithiums, modern
Пожаловаться на содержимое документа