On the solution behaviour of benzyllithium·(-)-sparteine adducts and related lithium organyls - A case study on applying 7Li, 15N{1H} HMQC and further NMR methods, including some investigation into asymmetric synthesis

Article abstract of DOI:10.1002/chem.201202346

2D ⁷Li,¹⁵N heteronuclear shift correlation through scalar coupling has successfully been applied to several lithium organyls consisting of polydentate N ligands such as N,N,N,N-tetramethylethylenediamine (tmeda), N,N,N,N,N′-pentamethyldiethylentriamine (pmdta) and (-)-sparteine. Structural insights on the conformation of benzyllithium× pmdta (5) in a toluene solution and the strength of ion pairing in combination with PGSE NMR measurements, ¹H,¹H-NOESY and ¹H,⁷Li-HOESY experiments are presented. By studying in detail the formation of 5 in solution, a transient species has been observed for the first time and assigned to a pre-complex of nBuLi and pmdta. In addition, the solution behaviour of the complex formed between benzyllithium and (-)-sparteine (8) has been studied by PGSE and multinuclear NMR spectroscopy. The straightforward synthesis and first applications in asymmetric lithiations are also reported, which show that the new system benzyllithium×(-)- sparteine (8) provide poorer enantioselective induction than the classical nBuLi×(-)-sparteine (6). The results were supported by deprotonation experiments confirming that the formation of 8 relies on two relevant factors, namely temperature and lithiating reagent. The existence of 8 may thus interfere with the asymmetric induction when the system nBuLi× (-)-sparteine is used in the enantioselective deprotonations of N-Boc-N-(p-methoxyphenyl)- benzylamine conducted in toluene. To pair or not to pair: 2D ⁷Li,¹⁵N heteronuclear shift correlation has been applied to several lithiumorganyls consisting of polydentate N ligands such as N,N,N,N-tetramethylethylenediamine, N,N,N,N,N′- pentamethyldiethylentriamine and (-)-sparteine. In combination with other spectroscopic techniques, details on the formation of the lithiumorganyl- polyamine adducts the strength of ion pairing in solution have been explored (see figure). Copyright

Full text of DOI:10.1002/chem.201202346

FULL PAPER
DOI: 10.1002/chem.201202346
On the Solution Behaviour of Benzyllithium·(ꢀ)-Sparteine Adducts and
7
Related Lithium Organyls – A Case Study on Applying Li,15N{1H}HMQC
and Further NMR Methods, Including Some Investigation into
Asymmetric Synthesis
Marꢀa Casimiro,[a] Pascual OÇa-Burgos,[b] Jens Meyer,[b] Steffen Styra,[b] Istemi Kuzu,[b, c]
Frank Breher,*[b] and Ignacio Fernꢁndez*[a]
Abstract: 2D 7Li,15N heteronuclear
shift correlation through scalar cou-
pling has successfully been applied to
several lithium organyls consisting of
transient species has been observed for
the first time and assigned to a pre-
complex of nBuLi and pmdta. In addi-
tion, the solution behaviour of the
complex formed between benzyllithium
and (ꢀ)-sparteine (8) has been studied
by PGSE and multinuclear NMR spec-
troscopy. The straightforward synthesis
and first applications in asymmetric
lithiations are also reported, which
show that the new system benzyllithi-
um·(ꢀ)-sparteine (8) provide poorer
enantioselective induction than the
classical nBuLi·(ꢀ)-sparteine (6). The
results were supported by deprotona-
tion experiments confirming that the
formation of 8 relies on two relevant
factors, namely temperature and lithi-
polydentate
N ligands
such
as
N,N,N’,N’-tetramethylethylenediamine
(tmeda), N,N,N’,N’,N’’-pentamethyldi-
ethylentriamine (pmdta) and (ꢀ)-spar-
ACHTUNGTRNEaNUNG ting reagent. The existence of 8 may
teine. Structural insights on the confor-
mation of benzyllithium·pmdta (5) in a
toluene solution and the strength of
ion pairing in combination with
PGSE NMR measurements, 1H,1H-
NOESY and 1H,7Li-HOESY experi-
ments are presented. By studying in
detail the formation of 5 in solution, a
thus interfere with the asymmetric in-
duction when the system nBuLi·
(ꢀ)-sparteine is used in the enantiose-
lective deprotonations of N-Boc-N-(p-
methoxyphenyl)-benzylamine conduct-
ed in toluene.
Keywords: asymmetric synthesis ·
lithium · N ligands · NMR spec-
AHCTUNGTREGtNNUN roscopy · sparteine
Introduction
corporation of alkyl groups, either by salt metathesis or car-
bolithiation.[2] Numerous investigations have revealed the
influence of solvation and aggregation on the reactivity and
selectivity of lithium organyls,[3] where the effect of the sol-
vent appears to be highly variable depending on the nature
of the substrate being solvated. Most prominently, chelating
N donor ligands are employed in order to achieve a de-ag-
gregation of the oligomeric lithium organyls.[4] They have
been shown to improve product yields, alter product distri-
butions and increase reaction rates primarily through solva-
tion and chelation of the lithium cations.[5] However, not all
chelating N donors exert exactly the same effect. Under cer-
tain conditions, the tridentate ligand N,N,N’,N’,N’’-penta-
The use of carbanions in synthetic chemistry is well estab-
lished. Polar main group organyls such as organolithium
compounds provide excellent nucleophilic reactivity and
belong to the most important reagents for synthetic and in-
dustrial applications.[1] They are extensively used, for in-
stance, as deprotonating reagents or for metal halogene ex-
change reactions and possess significant relevance for the in-
[a] M. Casimiro, Prof. Dr. I. Fernꢀndez
Laboratory of Organic Chemistry, University of Almerꢁa
Carretera de Sacramento s/n 04120 Almerꢁa (Spain)
E-mail: ifernan@ual.es
A
ACHTUNGTRENyNUGN ldiethylentriamine (pmdta, 1) shows a distinctly differ-
[b] Dr. P. OÇa-Burgos, Dr. J. Meyer, S. Styra, Dr. I. Kuzu,
Prof. Dr. F. Breher
A
A
Institute of Inorganic Chemistry
are employed in the reaction of nBuLi with the specific sub-
strate, different regioselectivies have been observed, which
were explained by assuming different nuclearities of the
pre-complexes formed between the nBuLi and the polyden-
tate amine.[6]
Karlsruhe Institute of Technology (KIT)
Engesserstr. 15, 76131 Karlsruhe (Germany)
E-mail: breher@kit.edu
[c] Dr. I. Kuzu
Present address: Fachbereich Chemie
Philipps-Universitꢂt Marburg
(ꢀ)-Sparteine (3) is by far the most important of the
chiral N donor ligands used to control the enantioselectivity
of organolithium reactions. In fact, one of the pioneering
Hans-Meerwein-Str., 35043 Marburg (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202346.
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
691

achievements in organolithium chemistry was the discovery
that 3 and alkyllithium reagents allowed the efficient asym-
metric deprotonation of alkoxy and pyrrolidine substrates,[7]
which constituted the starting point of investigations on a
wide range of derivatives.[8] Related (+)-sparteine surrogates
have shown important enantioselective achievements, which
makes them excellent alternatives.[9] The role of organolithi-
um chemistry in asymmetric synthesis is well known,[10] and
configurationally stable organolithium intermediates partici-
pate in numerous enantioselective reactions. Asymmetric
transformations that employ 3 in catalytic amounts are gain-
ing increased recognition;[11] however, methods involving
dipole-stabilised carbanions still provide the main body of
knowledge.[12]
As a suitable entry point we have chosen to study the
dimer (nBuLi·tmeda)2 (4). nBuLi forms hexameric aggre-
gates in hydrocarbons,[16] dimers in the presence of
tmeda[17,18] and exists in a tetramer–dimer equilibrium in
THF solutions.[18a,c,19] Complex 4 was prepared by addition
of 19 mL of a 1.6m nBuLi solution in hexane to a 5 mm
NMR tube containing non-deuterated toluene (0.5 mL) and
tmeda (1 equiv). Figure 1 shows the natural abundance
7Li,15N{1H} HMQC spectra recorded at 173 K.
As part of our recent efforts providing new NMR meth-
ods able to understand structure/reactivity relationships of
lithiated species, we have been investigating the application
of natural abundance experiments without the need to use
isotopically enriched 15N and 6Li reagents. In this line of
thought, we have previously shown the application of
7Li,15N HMQC experiments at natural abundance on zwit-
terionic lithium complexes consisting of multifunctional
polyACHTUNGTRENNUNG(pyACHTUNGTRENNUNGrazolyl) ligands of the general formula [E(3,5-
Me2pz)3]ꢀ (E=C, Si, Ge and Sn),[13] and on the well-known
lithium diisopropylamide LDA.[14] We describe herein the
application of such method to several known and recently
published organolithium species, including the new system
benzyllithium·(ꢀ)-sparteine. The strength of the 7Li,15N
HMQC method at natural abundance is particularly obvious
in cases adducts of lithium organyls with naturally occurring
alkaloids like (ꢀ)-sparteine are studied, for which enantio-
merically pure 15N-labeled samples are only hardly available.[15]
Results and Discussion
All NMR studies were carried out on a spectrometer
ACHTUNGTRENNUNGoperating at a proton frequency of 400 or 500 MHz equip-
ped with a TBI probehead which allowed the performance
of double and triple resonance experiments involving 1H,
7
13C, Li and 15N as reported previously.[14] Scheme 1 shows
the lithiated compounds considered in this study. All of
them have been previously characterised in solution and in
the solid state, except complex 8, which is described herein
for the first time.
Figure 1. 7Li,15N{1H} HMQC 2D NMR spectra of (nBuLi·tmeda)2 (4) at
173 K (0.15m in non-deuterated toluene). a) Recycle time of 1 s, b) recy-
cle time of 300 ms.
The 2D map clearly depicts a single cross-peak, which
leads unambiguously to the conclusion that each lithium
atom is attached to two magnetically equivalent nitrogen
donors, as has been nicely proved by Williard and co-work-
ers with the help of isotopically enriched [15N,15N’]-tmeda
and [6Li]-n-butyllithium.[17]
7
The natural abundance Li,15N{1H} HMQC experiment on
4 illustrates that even at temperatures close to the freezing
point of the solvent, the experiment can be performed in
less than 1 h using a recycling time of 1 s (Figure 1). The
axial signal observed at F1=50 ppm was experimentally re-
moved by increasing the recycling time up to 1 s instead of
Scheme 1. Organolithium species studied in this contribution featuring
different polyamines and nuclearities.
692
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 691 – 701

Solution Behaviour of Benzyllithium/(ꢀ)-Sparteine Adducts
FULL PAPER
300 ms (Figure 1). From the internal projections in both di-
mensions, one can clearly see how in Figure 1a all the mag-
netisation is employed in coherence selection, whether in
Figure 1b, only part is taken, resulting in a much poorer
signal-to-noise ratio in both dimensions.
While the bidentate N donor 2 usually forms dimers or
bridged tetramers, the tridentate amine pmdta (1) also pro-
duces monomeric organolithium species,[20] which usually
show an enhanced reactivity. We thus tried to characterise
at very low temperature the proposed monomeric complex
resulting from the reaction of n-butyllithium with 1 in tol-
uene solution.
It is important to note that the solid-state structure of the
complex formed by pmdta and nBuLi obtained from n-pen-
tane/n-hexane solutions has been described by Strohmann
and Gessner.[21] It was found that the complex consists of a
centrosymmetric dimer that can be formally understood as
two monomeric nBuLi·pmdta units coordinated to the lithi-
um atoms of two nBuLi molecules forming a dimer and con-
sisting of a central Li2C2 heterocycle. However, our investi-
gations revealed that the situation is somewhat different in
toluene solution. Although the aforementioned complex
nBuLi·pmdta is formed, we also observed—not astonishing-
ly—the reaction of the latter with the solvent. NMR samples
(~0.15m) were prepared in non-deuterated toluene as men-
tioned earlier, but this time replacing tmeda by pmdta. The
7Li NMR spectrum measured in the first minute of reaction
at 173 K exhibits two singlets located at dLi =1.77 and
0.53 ppm. The higher frequency signal rapidly disappeared
within a few minutes and converted exclusively into the
second one (Figure 2). The transient species has been tenta-
ised as complex
5
(benzyllithium·pmdta). Selective
1H gTOCSY 1D experiments, 13C and DEPT135 NMR
measurements supported the formation of this species. Com-
plex 5 has been recently synthesised by Stalke and co-work-
ers from the reaction of trimethylsilylmethyllithium in pres-
ence of pmdta and toluene.[22] Although complex 5 has been
characterised by X-ray crystallography, no special emphasis
was given in terms of solution NMR at low temperature;
therefore a lack of structural insights exists about this new
species.
Complex 5 was straightforwardly synthesised as men-
tioned above and isolated as crystalline blocks by layering
n-hexane into a cold toluene solution. The product was fur-
ther characterised by 1H and 13C NMR spectroscopy
(Table 1). Similar chemical shifts have been reported previ-
ously for related benzyllithium compounds.[18a,23,24]
1
Table 1. Selected H and 13C NMR chemical shifts (in ppm) of crystalline
5
(benzyllithium·pmdta) and
8
(benzyllithium·(ꢀ)-sparteine) in
[D8]toluene at 293 K.
5
8
1H
13C
1H
13C
ipso
–
160.7
117.1
128.9
106.0
36.3
–
156.8
115.9
129.5
106.9
34.9
o-Ph
m-Ph
p-Ph
CH2Li
6.66
7.00
6.18
2.06
6.94
6.57
6.12
2.26
The 13C NMR chemical shift of the anionic methylene
carbon atom appears at 293 K as a sharp singlet at dC =
36.3 ppm (W1/2 =3.0 Hz) with no evidence of coupling with
the lithium nucleus. At 173 K where chemical exchange
processes are slowed down with respect to the NMR time
scale, a similar spectrum is observed (singlet at dC =
38.4 ppm; W1/2 =57 Hz). The chemical shift is obviously
found at higher frequencies as compared to either toluene
(dC =20.3 ppm), or, more importantly, free benzyllithium
(dC =30.1 ppm).[24] However, the 13C NMR chemical shift of
the methylene carbon atom of 5 was found at lower fre-
quencies than the related benzyllithium compound 6 (dC =
39.9 ppm), which indeed does possess a 6/7Li,13C coupling
constant of 7.0 Hz.[25] Note that a similar trend was found
for the 13C NMR chemical shift of the ipso carbon atom in
phenACTHNUGRTNENUGylCAHUTNGTRENNlGU ithium. The degree of aggregation
Figure 2. 7Li NMR (194.37 MHz) spectra of the earlier stages in the lith-
iation of toluene with nBuLi at 173 K in non-deuterated toluene.
a) 5 min and b) 45 min.
of aryllithium compounds is well known
to correlate with the ipso carbon chemical
shift. An increasing aggregation causes a
shielding of the Cipso due to the existence
tively assigned to a pre-complex of the general formula
nBuLi·pmdta with a non-established degree of aggregation.
The proposed structure is supported by 1H NMR spectro-
of
AHCTUNGTRENNUNG
a
larger number of Li–C con-
tacts.[18a,26,27] As far as we are aware, com-
pound 6 represents the only example
scopic investigations showing
a
characteristic low-field
where the methylenic benzyllithium
signal at dH =ꢀ0.94 ppm, which can be clearly assigned to
the a-methylene protons of the butyl chain directly attached
to the lithium atom. This is the first evidence of such a reac-
tive species in toluene as solvent.
After several minutes, the transient species is completely
converted to a new species that has been further character-
shows 6/7Li,13C couplings. In this particular case, however,
the Li–C contacts are somehow imposed by the crown ether
moiety.
In addition to these findings, we noticed the low frequen-
cy shifts of the ortho and especially the para carbon atoms
of 5 indicating that considerable amounts of p-electron den-
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
693

F. Breher, I. Fernꢀndez et al.
sity are located on these atoms. Note that the chemical shift
of the p-phenyl carbon atom (dC =106.0 ppm) appears at
lower frequency than the corresponding value of benzyllithi-
um in benzene solution (dC =113.6 ppm).[24] Altogether, we
suggest a solvent-separated ion pair structure for 5, in which
the carbon atom has no direct Li–C contact.
To shed some more light on the structure of 5 at 173 K in
toluene,
we
performed
a
natural
abundance
7Li,15N{1H} HMQC experiment (Figure 3). As expected, the
model compounds A to F at the RI-DFT/BP86/def2-TZVP
level of theory (Figure 4). We have used two sets of model
compounds differing in the orientation of the benzyl moi-
AHCTUNGTREGeNNUN ties relative to the pmdta chelate (Figure 4 top and
bottom) as well as the relative orientations of the ethylene
bridges to each other (A!C and D!F in Figure 4). Only
the structures A and C were found to be minimum struc-
tures with no imaginary frequencies on the potential energy
surface. The model structure A, which was found to be of
lowest energy, features the expected endo/exo arrangement
of the ethylene bridges (corresponds to the schematic draw-
ing I). Only for this isomer one would expect three different
15N NMR signals (see Figure 3). We also noticed that all
other calculated structures (either minimum structures or
transition states) are higher in energy. It is reasonable to
assume that these (related and not calculated) structures are
in equilibrium at room temperature and that this equilibri-
um can only be slowed down at very low temperature.[29]
This assumption would be in accord with our experimental
findings that three different 15N NMR chemical shifts were
observed only at T<173 K. We note in the passing that the
endo/exo arrangement of the ethylene bridges found for the
calculated species A was also observed by Stalke and co-
workers in the solid state,[22] although the orientation of the
benzyl moiety differs from A.
The last polyamine employed in this study was (ꢀ)-spar-
teine, which is a naturally occurring tetracyclic quinolizidine
alkaloide[15] that has been extensively used by Hoppe and
Beak in numerous stereoselective lithiation/substitution re-
actions as chiral inducer, together with various organometal-
lic reagents.[8a,28,43] Interestingly, Beak and co-workers have
reported on a surprisingly large solvent dependence on the
selectivity—a result indicating that the equilibrium between
different complexes of alkyllithium with (ꢀ)-sparteine is
strongly solvent dependent.[28d]
Figure 3. Natural abundance 7Li,15N{1H} HMQC 2D NMR spectra of 5
(benzyllithium·pmdta; 0.15m in non-deuterated toluene). a) 173 K,
b) 163 K.
7Li NMR signal located at dLi =0.57 ppm correlate well with
two different nitrogen centres at dN =21.5 and 25.9 ppm.
Lowering the temperature down to 163 K, the low frequency
signal splits into two cross-peaks (dN =21.1 and 22.2 ppm),
which were intuitively assigned to magnetically non-equiva-
lent dimethylamino nitrogen atoms, considering a restricted
conformation at this low temperature.
The specific conformation, in which the three nitrogen
atoms are magnetically non-equivalent, can be reasonably
explained by taking the relative orientations of the ethylene
bridges into account. Three different, that is, magnetically
non-equivalent nitrogen atoms are only present in struc-
ture I.
NMR spectroscopic monitoring of toluene solutions con-
taining equimolar amounts of nBuLi and (ꢀ)-sparteine re-
vealed the clean formation of a single complex, which has
previously been described by Hoffman and Collum using
6Li-enriched samples to be the dimer (nBuLi·(ꢀ)-sparteine)2
(7; see Scheme 1).[30,31] Our 7Li NMR spectrum displayed
two resonances in a nearly 1:1 ratio with almost identical
chemical shifts, consistent with two possible orientations of
the sparteine ligand, as previously reported.[30] Under our
conditions of natural abundance ingredients, we performed
7
the optimised 2D shift correlation between Li and 15N. Un-
fortunately, no success was obtained, probably due to the
small relaxation times of the lithium atoms on the NMR
time-scale (most likely due to exchange dynamics). Broad
To refine the picture a little further, we have performed
some density functional theory (DFT) calculations on the
694
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 691 – 701

Solution Behaviour of Benzyllithium/(ꢀ)-Sparteine Adducts
FULL PAPER
complex was characterised as
benzyllithium·(ꢀ)-sparteine (8;
7
see below); identical Li NMR
spectra were obtained from
pure and isolated samples of 8
(see the Experimental Sec-
tion). To the best of our knowl-
edge, no experimental data
have been reported for this
species so far. The main results
of these NMR investigations
appear to be consistent with
the findings of Beak and co-
workers on the solvent de-
pendence on the selectivity of
alkyllithium/(ꢀ)-sparteine mix-
tures in stereoselective reac-
tion mentioned above.[28d] Our
7Li NMR study provide evi-
dence that this dependence
may arise from an equilibrium
between aggregates containing
the same lithiumorganyls, or
between lithiumorganyls of dif-
ferent nature, that is, contain-
ing benzyllithium.[32]
Figure 4. DFT calculated structures A to F at the RI-DFT/BP86/def2-TZVP level of theory and relative ener-
gies (A=0 kJmolꢀ1) including ZPE corrections. Note the different orientation of the benzyl entities in the two
different sets A–C and D–F. Structures marked with an asterisk are no minimum structures and consist of one
imaginary frequency.
We were able to isolate com-
1H and 13C resonances further supported this assumption.
Interestingly, we found that when freshly prepared samples
were slightly “heated”, that is, changes in the temperature
from ꢀ1008C up to ꢀ808C, the appearance of several new
plex 8 as crystalline material by layering the resulting tol-
uene solution with n-hexane and storing the solution at
ꢀ408C after several days (see the Experimental Section).
7
The Li NMR resonance for 8 at room temperature appears
7
1
signals are visible in the Li NMR spectra (marked with as-
at dLi =1.13 ppm, with a H NMR spectrum evidently show-
terisks in Figure 5b). We also noticed an almost quantitative
formation of a new species if the sample was kept at room
temperature for some minutes (Figure 5c). The resulting
ing the presence of a new benzyl entity with the lithiated
methylene resonance located at dH =2.26 ppm (see Table 1).
1
Again, H, 13C and DEPT135 NMR spectra were acquired
confirming the structure of 8 (Table 1). As discussed for the
case of 5, a delocalised benzyl moiety was found for 8. By
comparing the chemical shifts of the ortho (dC =115.9 ppm)
and especially the para carbon atom (dC =106.8 ppm) with
the values obtained for 5 it becomes evident that the charge
delocalisation in 8 is slightly reduced, but significantly more
pronounced if compared with amine-free benzyllithium.[24]
Compared to free (ꢀ)-sparteine, the N donor ligand in 8
1
shows well-separated H and 13C NMR signals.[33] Again, the
lack of 6/7Li,13C couplings in the 13C NMR spectrum confirms
ꢀ
the existence of a solvent-separated ion pair with no Li C
7
bond. The natural abundance 2D Li,15N{1H} HMQC experi-
ment at 173 K worked out this time showing a correlation of
the lithium atom with two different nitrogen centres at dN =
30.6 and 54.0 ppm (Figure 6).
The difference between the nitrogen chemical shifts of ap-
proximately 20 ppm provides clear evidence for the unsym-
metrical binding capabilities of both nitrogen donors. This
might be one of the reasons why (ꢀ)-sparteine has gained
an enormous success in enantioselective reactions. Interest-
ingly, only a single species is formed in solution and no evi-
dence of conformational equilibrium seems to be occurring,
Figure 5. a) 7Li NMR spectra (194.4 MHz) of the system nBuLi·(ꢀ)-spar-
teine in [D8]toluene at a) ꢀ1008C, b) ꢀ808C and c) ꢀ1008C after warm-
ing the sample up to room temperature for some minutes.
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
695

F. Breher, I. Fernꢀndez et al.
Table 2. Diffusion coefficient (D [10ꢀ10 m2 sꢀ1)]) and hydrodynamic
radius (rH) values for complexes 5, 8 and free polyamines pmdta and
(ꢀ)-sparteine at room temperature in [D8]toluene.
Complex Ligand Entity
Nucl. D[a]
rH
rXꢀray
[ꢄ][b] [ꢄ]
10.197 3.6
4.7[c]
5
8
1
3
PhCH2
[Li(pmdta)]
1H
1H
7Li
1H
G
9.179 4.0
9.194 4.0
7.156 5.2
6.881 5.4
6.894 5.4
13.661 2.7
11.801 3.1
PhCH2
5.7[d]
[Li(ꢀ)-sparteine] 1H
7Li
1
3
–
–
–
–
1H
1H
–
–
[a] The experimental error in the D values is ꢁ2%. [b] The viscosity h
used in the Stokes–Einstein equation was taken from Perryꢀs Chemical
Engineersꢀ
Handbook
8th Edition
(www.knovel.com)
and
is
0.5819·10ꢀ3 kgmꢀ1 sꢀ1
.
[c] Deduced from the X-ray structure (see
Ref. [22]) by considering the volume of the crystallographic cell divided
by the number of contained asymmetric units, the latter one is assumed
to be spherical in shape. [d] Theoretical estimation obtained using Chem-
Bio3D Ultra 12.0. Estimated from the critical volume (Vc) and assuming
a spherical shape.
Figure 6. Natural abundance 7Li,15N HMQC 2D NMR spectra of 8 (ben-
zyllithium·(ꢀ)-sparteine; 0.15m in non-deuterated toluene) at 173 K.
opposite to the behaviour evidenced by us and by others[30]
1
in the reaction of n-butyllithium with several amines. H,15N
gHMQC 2D NMR experiments carried out at room temper-
ature allowed us to assign the two different nitrogen atoms
(see the Supporting Information).
nificant difference between 5 (3.6 ꢄ) and 8 (5.2 ꢄ). From
the X-ray structure of 5[22] or the estimated geometry for 8
(see the Supporting Information) one can gauge rotational
radii of 4.7 and 5.7 ꢄ, respectively. For 5, the radius deter-
mined by X-ray crystallography is much larger than the cal-
culated rH values of 3.6 and 4.0 ꢄ for the benzyl fragment
and the lithium–pmdta entity, respectively. This leads to the
assumption that no contact ion pair exists in solution and
that 5 most likely represents a mononuclear species partially
separated by solvent molecules. This solution behaviour
would be in sharp contrast to the solid-state structure,
where a clear h1 monomer is found. In 8, the closer experi-
mental D values for both the cationic and anionic entities
indicate again partial ion pairing, but this time appearing to
be tighter than in 5 (larger radii).
Single crystals of 8 could be grown (see the Experimental
Section) but in all cases they were unfortunately not suitable
for X-ray diffraction studies. Nonetheless, the crystalline
material was dissolved in [D8]toluene and employed in the
following studies.
1
7Li and H PGSE diffusion methods at room temperature
have been used to study ion pairing and/or aggregation phe-
nomena for the two crystallised species 5 and 8. The meas-
urement of diffusion constants through pulsed gradient spin-
echo (PGSE) NMR methods[34,35] has recently attracted in-
creasing interest as this technique provides data on molecu-
lar volumes, and thus indirectly, on structural insights.[36] The
combination of 6/7Li NMR and PGSE methods is gaining in-
creased recognition.[22,37,38] In addition, extensive experimen-
tal work relating the structures of organolithium compounds
with their reactivities has been carried out by Reich and co-
workers.[39]
PGSE NMR diffusion measurements in combination with
homo- and hetero-nuclear nOe spectroscopy afforded com-
plementary data. The existence and structure of the ion pair
can be deduced from NOESY and HOESY experiments,
while the size and nature of the adducts can be explored by
1
The data reported here, in combination with the recently
described molecular structure of 5,[22] provide novel insights
into both the question of the degree of aggregation and if
(ꢀ)-sparteine (3) promotes smaller aggregates as compared
to pmdta (1). Although the viscosity of the toluene solutions
varies with concentration, we have used pure solvent viscosi-
ty for radii calculation using the Stokes–Einstein equation.
The 1H and 7Li PGSE results for toluene solutions
(ca. 0.1m) of 5 and 8 and 60 mm samples of the polyamines 1
and 3, all at ambient temperature, are given in Table 2.
It was found that for both complexes the coordination of
the larger amines to the lithium cations slows the translation
PGSE measurements. H NOESY measurements allow the
1
recognition of H spins which are located in the same region
of space and, most importantly, identify close inter
contacts.
ACHTUNGTRENNUNGliACHTUNGTRENNUNGgand
Figure 7 depicts expansions of the magnitude mode 2D
1H gNOESY spectra of 5 and 8 at 293 K. As can be seen
from the numbers and intensities of cross-peaks, a quite dis-
tinct degree of ion paring can be estimated, in particular for
8, which would be in accord with the PGSE studies. Main in-
teractions were observed for both complexes between the o-
Ph entity and the closest protons of both polyamines. In 5,
very weak interactions are found with the strongest one be-
tween the o-Ph proton and the pmdta methyl moieties (Fig-
ure 7a). The 1H NMR spectrum of 8 is more complicated
due to the more complex structures of the (ꢀ)-sparteine
ligand. Consequently, a spectrum with at least sixteen differ-
7
of the Li+ ions and produces a decrease in the Li D values.
Once coordinated, the movements of 1 or 3 are slowed
down as well, which can be seen from the compiled 1H
D values in Table 2. The benzyl fragment experiences a sig-
696
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 691 – 701

Solution Behaviour of Benzyllithium/(ꢀ)-Sparteine Adducts
FULL PAPER
7
served between the Li nucleus
and both CH2 and CH3 protons
of the pmdta subunit, whereas
only a single and weak cross-
peak is detected arising from
the CH2 benzyl moiety (Fig-
ure 8a). These data again sup-
ported
a picture where not
much ion pairing exists for 5 in
solution. No HOESY cross
peaks were observed for the
interaction of 7Li and any of
the aromatic protons of the
benzyl fragment. The internu-
clear distance between the Li
nucleus and the ortho proton
of the phenyl group is approxi-
mately 5.5 ꢄ. Hence, the
reason why at this distance no
Figure 7. Section of 1H gNOESY NMR spectra of a) 5 and b) 8 at 400.13 MHz (ca. 0.1m); nOes are shown
mainly between o-, m-Ph and pmdta and (ꢀ)-sparteine moieties. Asterisks represent partially deuterated tol-
uene. See the Supporting Information for assignment details.
ent aliphatic signals results, many of these arising from over-
lapping resonances (Figure 7b). However, the 2D map
shows more dipolar interactions and stronger ones support-
ing a tighter ion pair than in 5.
observable cross-peak is obtained has to be due to ion sepa-
ration, as has been deduced from diffusion studies.
In the case of 8, strong HOESY cross-peaks were ob-
7
served between Li and the nitrogen-bound CH2 entity H15
Heteronuclear nOe methods, while inherently very insen-
sitive, have also received attention for structural elucida-
tion.[41] 2D heteronuclear X-1H Overhauser spectroscopy
(HOESY) techniques[40a,41] have been applied to a wide
range of different compounds and a variety of X, that is,
nuclei with spin quantum numbers of 1/2 and ꢂ1 as well,
being all predominantly X-detected.[42]
(see the Supporting Information for numbering) as well as
between 7Li and the ortho proton of the benzyl fragment
(Figure 8c). This observation would be in accord with a spe-
cies forming a stronger ion pair than 5. The lack of a cross-
peak with the corresponding CH2 moiety H2 (see Figure 7)
indicates a fluxional conformation where parts of the biden-
tate ligand is partly bent, inhibiting the dipolar interaction
with the lithium.
7Li,1H HOESY experiments obtained from [D8]toluene
solutions of 5 or 8 indicated that the structures of both com-
plexes are somewhat different. The correlations found in the
2D NMR maps indicated that the solution structures of 5
are quite similar to the solid
Having established by means of multinuclear NMR the
features of 8 as a new potential lithiating reagent, we per-
formed some enantioselective assays on N-Boc-N-(p-me-
state.[22] However, the nOe ef-
fects obtained in this experi-
ment with only a single mixing
time cannot be used for quan-
titative distance determina-
tions.
The relative intensities of
the cross-peaks in a HOESY
experiment is determined by
the competition between the
nOe build-up and the decay of
signal
through
relaxation.
Owing to the rapid relaxation
7
rates of most Li nuclei in most
lithium organyls very often
precludes the observation of
useful nOes. However, for 5
and 8, these values are large
enough to be adequate for
their observation. In the
NOESY experiments on 5,
strong cross-peaks were ob-
Figure 8. Sections of 7Li,1H-HOESY NMR spectra of a,b) 5 (benzyllithium·pmdta) and c) 8 (benzyllithi-
um·(ꢀ)-sparteine) at 155 MHz. Intermolecular nOes are shown between F2 and the CH3, CH2 and CH2Li in 5,
and H15, and the o-Ph protons in 8. The 1D projections correspond to 7Li and 1H NMR spectra. Spectral ac-
quisition parameters are the following: 0.5 s for mixing time, 1871 Hz in F2 (7Li) dimension, and 128 spectra
(64 scans each) were accumulated with a t1 increment to make 4068 Hz in the F1 dimension. See the Support-
ing Information for numbering.
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
697

F. Breher, I. Fernꢀndez et al.
thoxyphenyl)-benzylamine 9. The solvent of choice was tol-
uene, where we have observed the formation of 8, either in
quantitative yield or in small amounts. First we reproduced
the reaction conditions developed by Beak and co-work-
ers[43] with an exquisite control of the temperature, in this
case ꢀ788C (Scheme 2a). When we used analogous condi-
was worked up. The second and third assays (b) and (c),
were conducted in the same way, but this time raising the re-
action temperature up to ꢀ708C and ꢀ608C, respectively.
NMR spectroscopic monitoring[44] of the product distribu-
tion nicely confirmed the increased formation of
PhCH2SnMe3 (11) when the temperature is slightly raised
(a: 2%!b: 7%!c: 11%). In the last experiment (d), the
temperature of the reaction mixture was shortly raised to
208C before the addition of Me3SnCl, exclusively generating
11.
This set of experiments makes evident that at tempera-
tures slightly above ꢀ788C complex 8 does exist and can,
therefore, participate in the lithiation process of the sub-
strate of choice. In this sense, there are literature reports
where the systems nBuLi·(ꢀ)-sparteine (or sBuLi·(ꢀ)-spar-
teine)[45] is used in asymmetric deprotonation reactions con-
ducted in toluene at ꢀ788C,[8a,28d,46] and the existence of 8 in
these reaction possibly may interfere with the asymmetric
induction. We believe that the stereoselectivity can be sur-
prisingly altered if the formation of 8 is not avoided, either
by precisely controlling the temperature or the lithiating re-
agent (see above).
Scheme 2. Asymmetric lithiation/stannylation of N-Boc-N-(p-methoxy-
phenyl)benzyl-amine 9 with a) nBuLi/L* (with L*=(ꢀ)-sparteine) and
b) 8 (benzyllithium·(ꢀ)-sparteine).
tions as in the control experiment (Scheme 2a), but allowing
the solution to reach room temperature for 30 min, and
therefore quantitatively forming the benzyllithium complex
8, we obtained a significant erosion of the enantiomeric
excess (43% ee, Scheme 2b). These experiments prove the
scarce ability of 8 in achieving good enantiodiscrimination
for this specific substrate. Based on these results it is impor-
tant to mention that good control in the temperature has to
be achieved when using nBuLi·(ꢀ)-sparteine (6) in toluene
solutions. In doing so, the formation of the unfavourable
complex 8, and consequently a decrease in the enantiomeric
induction, can be precluded.
To further prove this final statement, we performed four
lithiation experiments without any substrate under various
reaction conditions (Scheme 3a–d and Figures S7–S10 in the
Supporting Information).
In the first reaction (a), after 30 min of stirring of a mix-
ture of (ꢀ)-sparteine and nBuLi in toluene at ꢀ788C, the
electrophile Me3SnCl was added and the reaction mixture
Conclusion
2D 7Li,15N heteronuclear shift correlation through scalar
coupling has successfully been applied to n-butyllithium and
benzyllithium bearing polydentate N ligands such as tmeda,
pmdta and (ꢀ)-sparteine under conditions that are typically
employed in standard lithiations, that is, low temperatures
and non-deuterated solvents. This method is supposed to
serve to routinely obtain useful information in relatively
short time by employing facilely accessible, non-isotopically
enriched reagents. Thus, this technique is a valuable tool for
studying compounds containing lithium–nitrogen bonds in
solution. This 2D experiment has been able to provide us
with the right conformation of the recently described com-
plex 5 as a species where the three nitrogen donor atoms of
pmdta are magnetically different at 163 K. Based on PGSE
data for 5 and 8, we conclude that toluene promotes differ-
ent degree of ion pairing, which has been supported by
homo- and heteronuclear nOe studies. The new benzyllithi-
um·(ꢀ)-sparteine system (8) has been applied in asymmetric
synthesis proving the scarce ability of 8 in achieving good
enantiodiscrimination for the specific substrate used in this
study. By employing the system nBuLi·(ꢀ)-sparteine (6) in
deprotonation experiments in toluene we could show that
complex 8 is generated as a function of temperature and can
therefore participate in the lithiation process of any depro-
tonable substrate.
Experimental Section
Scheme 3. Exploring the lithiation of toluene with nBuLi in the presence
of (ꢀ)-sparteine under different reaction conditions.
General considerations: All reactions and manipulations were carried
out in a dry nitrogen gas atmosphere using standard Schlenk procedures.
698
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 691 – 701

Solution Behaviour of Benzyllithium/(ꢀ)-Sparteine Adducts
FULL PAPER
Solvents were distilled from sodium/benzophenone immediately prior to
use, or via elution through a solvent column drying system.[47]
then layered with n-hexane and stored at ꢀ408C, affording, after 3 days,
air-sensitive yellow blocks of 5. 1H NMR ([D8]toluene, 298 K,
500.13 MHz): d=1.74 (brs, 8H), 1.92 (brs, 15H), 2.06 (s, 2H), 6.18 (tt,
1H, J=6.9 Hz, 1.3 Hz), 6.66 (m, 2H, J=8.4 Hz), 7.01 ppm (m, 2H, J=
8.4, 6.9 Hz); 13C NMR ([D8]toluene, 298 K, 125.76 MHz): d=36.4
(CH2Li), 44.2 (CH3), 45.5 (CH3), 53.5 (CH2), 57.3 (CH2), 105.9 (p-Ph),
117.2 (o-Ph), 128.9 (m-Ph), 160.7 ppm (Cipso). See Ref. [22] for an alterna-
tive synthetic access.
Multinuclear magnetic resonance spectra were measured in Bruker
Avance 400 and 500 spectrometers using indirect TBI 1H/31P/BB probe-
heads. The 908 pulse widths and operating frequencies were: 11.4 ms (1H,
500 MHz), 26.7 ms (7Li, 194.4 MHz) and 31 ms (15N, 50.7 MHz). The at-
tenuation levels used were 0 dB for proton and lithium, and ꢀ3 dB for
nitrogen. The spectral references used were TMS for 1H, LiBr (1m) in
D2O for 7Li, and NH3 for 15N. For the 7Li,15N{1H} HMQC NMR experi-
ments the outer coil was doubly tuned for 1H and 7Li, and the inner coil
tuned for 15N, with 908 pulse widths and attenuation levels of 22.7 ms/0 dB
for lithium, and 95 ms/ꢀ3 dB for nitrogen. Unless otherwise stated, stand-
ard Bruker software routines (TOPSPIN and XWINNMR) were used for
the 1D and 2D NMR measurements.
Benzyllithium·(ꢀ)-sparteine (8): A solution of nBuLi (0.545 mL of a 1.6m
solution in hexanes, 0.871 mmol) at ꢀ908C was added to a toluene solu-
tion (8 mL) containing (ꢀ)-sparteine 3 (0.2 mL, 0.871 mmol). The reac-
tion mixture was stirred at the same temperature for 30 min and then let
warm up to room temperature. The reaction mixture was slowly concen-
trated under vacuum and then layered with n-hexane and stored at
ꢀ408C, affording, after 5 days, air-sensitive crystals of 8 as yellow nee-
dles. 1H NMR ([D8]toluene, 298 K, 500.13 MHz): d=0.71 (brd, 1H, J=
9.4 Hz), 0.92 (brd, 1H, J=12.9 Hz), 1.03 (m, 4H), 1.13 (tt, 1H, J=12.9,
1.4 Hz), 1.28–1.52 (m, 8H), 1.58 (m, 2H), 1.67 (brd, 1H, J=13.8 Hz),
1.87 (m, 2H), 2.25 (m, 1H), 2.26 (s, 2H), 2.58 (brd, 1H, J=11.5 Hz), 2.73
(m, 3H), 2.82 (brd, 1H, J=9.6 Hz), 3.10 (brt, 1H, J=12.0 Hz), 6.12 (tt,
1H, J=7.0, 1.2 Hz), 6.57 (m, 2H, J=8.3 Hz), 6.95 ppm (m, 2H, J=8.3,
7.0 Hz); 13C NMR ([D8]toluene, 298 K, 125.76 MHz): d=23.8 (CH2), 24.5
NMR samples of 4, 5, 7 and 8 were prepared according to the following
protocol: to a 0.5 mL toluene solution prepared in a dried 5 mm NMR
tube containing 0.03 mmol of the corresponding polyamine (tmeda for 4,
pmdta for 5 and (ꢀ)-sparteine for 8) were added 19 mL (0.03 mmol) of
nBuLi (1.6m solution in n-hexane), with the tube immersed in liquid ni-
trogen. The frozen solution is placed into a refrigerated bath (N2(l)/
MeOH) at approximately ꢀ908C and then introduced inside the bore of
the magnet formerly pre-cooled at the same temperature.
(CH2), 24.8 (CH2), 25.1 (CH2), 27.9
(CH2Li), (CH3), 37.4 (CH), 45.5 (CH2), 52.6 (CH2), 56.9 (CH2), 59.1(CH),
60.5 (CH2), 66.4 (CH), 106.9 (p-Ph), 115.9 (o-Ph), 129.5(m-Ph),
156.8 ppm (Cipso).
ACHTUGNTREN(NUNG CH2), 30.2 (CH2), 34.6 (CH2), 34.9
Diffusion measurements were performed using the Stimulated Echo
Pulse Sequence[35] without spinning. The shape of the gradient pulse was
rectangular, and its strength varied automatically in the course of the ex-
periments. The D values were determined from the slope of the regres-
ACHTUNGTRENNUNG
sion line lnACHTUNGTRENNUNG
(I/I0) versus G2, according to Equation (1). I/I0 =observed spin
echo intensity/intensity without gradients, G=gradient strength, D=
delay between the midpoints of the gradients, D=diffusion coefficient,
d=gradient length.
I
I0
d
2
lnð Þ ¼ ꢀðgdÞ ðD ꢀ ÞDG2
ð1Þ
Acknowledgements
3
This work was supported by the Ministerio de Ciencia e Innovaciꢅn
(MICINN) and fondos FEDER (projects CTQ2008–117BQU). Financial
support by Junta de Andalucꢁa for a travelling grant (I.F.) is gratefully ac-
knowledged. M.C. and S.S thank MICINN and the Studienstiftung des
Deutschen Volkes, respectively, for doctoral fellowships and P.O.B.
thanks the Alexander von Humboldt foundation for a postdoctoral grant.
The calibration of the gradients was carried out via a diffusion measure-
ment of HDO in D2O, which afforded a slope of 1.99·10ꢀ4 [48]
We estimate
.
the experimental error in the D values to be ꢁ2%. All of the data lead-
ing to the reported D values afforded lines whose correlation coefficients
were >0.999 and 8–12 points have been used for regression analysis. To
check reproducibility, three different measurements with different diffu-
sion parameters (d and/or D) were always carried out. The gradient
strength was incremented in 8% steps from 10% to 98% and the recov-
1
7
[1] L. S. Hegedus, B. H. Lipshutz, J. A. Marshall, E. Nakamura, E. Ne-
gishi, M. T. Reetz, M. F. Semmelhack, K. Smith, H. Yamamoto in
Organometallics in Synthesis: A Manual (Ed.: M. Schlosser), Wiley,
New York, 2002.
ery delay set to 5 times T1. In the H and Li diffusion experiments d was
set to 2 and 4 ms, respectively, and the number of scans were 8–64 per in-
crement with a recovery delay of 10 to 60 s. Typical experimental times
were 1–4 h.
[2] a) J. Clayden, Organolithiums: Selectivity for Synthesis, Pergamon,
Oxford 2002, Chapter 7. For some recent examples, see: b) B. Stꢆ-
fane, Org. Lett. 2010, 12, 2900; c) S. Yasuda, H. Yorimitsu, K.
Oshima, Organometallics 2009, 28, 4872; d) T. L. Rathman, W. F.
Bailey, Org. Process Res. Dev. 2009, 13, 144, and references therein.
[3] a) J. L. Wardell in Comprehensive Organometallic Chemistry, Vol. 1
(Eds.: G. Wilkinson, F. G. A. Stone, F. W. Abel), Pergamon Press,
New York, 1982, Chapter 2; b) M. Szwarc in Ions and Ion Pairs in
Organic Reactions, Vols. 1 and 2, Wiley, New York, 1972; c) P. G.
Williard in Comprehensive Organic Synthesis, Vol. 1, (Eds.: B. M.
Trost, I. Flemming), Pergamon Press, New York, 1991, Chapter 1.
[4] For selected reviews, see: a) G. Boche, Angew. Chem. 1989, 101,
286; Angew. Chem. Int. Ed. Engl. 1989, 28, 277; b) E. Weiss, Angew.
Chem. 1993, 105, 1565; Angew. Chem. Int. Ed. Engl. 1993, 32, 1501;
c) T. Stey, D. Stalke in The chemistry of organolithium compounds
(Eds.: Z. Rappoport, I. Marek), Wiley, Chichester, 2004; d) V. H.
Gessner, C. Dꢂschlein, C. Strohmann, Chem. Eur. J. 2009, 15, 3320;
e) A. Torvisco, K. Ruhlandt-Senge, Inorg. Chem. 2011, 50, 12223.
[5] a) D. Seebach, Angew. Chem. 1988, 100, 1685; Angew. Chem. Int.
Ed. Engl. 1988, 27, 1624; b) V. Snieckus, Chem. Rev. 1990, 90, 879;
c) D. B. Collum, Acc. Chem. Res. 1992, 25, 448.
Computational details: The calculations were performed with the TUR-
BOMOLE program package.[49] The geometries were optimised without
symmetry restrictions at the BP86[50] level in the def2-TZVP basis[51]
using sharp convergence criteria. Analytical frequency calculations[52]
were performed at the BP86/def2-TZVP level, and the species A and C
(Figure 4) represent true minima on the respective potential-energy sur-
face. B, D, E and F were found to possess one imaginary frequency. The
coordinates of the calculated structures are compiled at the end of the
Supporting Information.
Materials and reagents: Me3SnCl and nBuLi were used as obtained from
commercial sources without further purification. Polyamines tmeda,
pmdta and (ꢀ)-sparteine were purchased from commercial sources and
distilled from KOH prior use. Low temperatures were achieved with a
N2(l)/MeOH baths, except ꢀ788C that was obtained using dry ice in ace-
tone. N-Boc-N-(p-methoxyphenyl)benzylamine 9 was prepared as de-
ACHTUNGTRENNUNG
scribed before.[43]
Synthesis of the compounds:
Benzyllithium·pmdta (5): A solution of nBuLi (0.545 mL of a 1.6m solu-
tion in hexanes, 0.871 mmol) at ꢀ908C was added to a toluene solution
(8 mL) containing pmdta 2 (0.181 mL, 0.871 mmol). The reaction mixture
was stirred at the same temperature for 30 min and then let warm up to
room temperature. The reaction mixture was slowly concentrated and
[6] F. Faigl, K. Fogassy, A. Thurner, L. Toke, Tetrahedron 1997, 53,
4883.
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
699

F. Breher, I. Fernꢀndez et al.
[7] a) D. Hoppe, F. Hintze, P. Tebben, Angew. Chem. Int. Ed. 1990, 29,
1422; b) P. Beak, S. T. Kerrick, S. Wu, J. Chu, J. Am. Chem. Soc.
1994, 116, 3231.
h) M. A. Nichols, P. G. Williard, J. Am. Chem. Soc. 1993, 115, 1568;
i) N. D. R. Barnett, R. E. Mulvey, W. Clegg, P. A. OꢇNeil, J. Am.
Chem. Soc. 1993, 115, 1573.
[8] a) D. Hoppe, T. Hense, Angew. Chem. 1997, 109, 2376; Angew.
Chem. Int. Ed. Engl. 1997, 36, 2282; b) B. Wolfe, T. Livinghouse, J.
Am. Chem. Soc. 1998, 120, 5116; c) G. A. Weisenburger, N. C. Faib-
ish, D. J. Pippel, P. Beak, J. Am. Chem. Soc. 1999, 121, 9522; d) A.
Basu, S. Thayumanavan, Angew. Chem. 2002, 114, 740; Angew.
Chem. Int. Ed. 2002, 41, 716.
[9] a) M. J. Dearden, C. R. Firkin, J.-P. Hermet, P. OꢇBrien, J. Am.
Chem. Soc. 2002, 124, 11870; b) M. J. Dearden, M. J. McGrath, P.
OꢇBrien, J. Org. Chem. 2004, 69, 5789; c) D. C. Ebner, R. M. Trend,
C. Genet, M. J. McGrath, P. OꢇBrien, B. M. Stoltz, Angew. Chem.
2008, 120, 6467; Angew. Chem. Int. Ed. 2008, 47, 6367; d) P. OꢇBrien,
Chem. Commun. 2008, 655; e) J. L. Bilke, S. P. Moore, P. OꢇBrien, J.
Gilday, Org. Lett. 2009, 11, 1935; f) D. Stead, G. Carbone, P.
OꢇBrien, K. R. Campos, A. Sanderson, J. Am. Chem. Soc. 2010, 132,
7260; g) G. Carbone, P. OꢇBrien, G. Hilmersson, J. Am. Chem. Soc.
2010, 132, 15445.
[19] a) J. Heinzer, J. F. M. Oth, D. Seebach, Helv. Chim. Acta 1985, 68,
1848; b) J. F. McGarrity, C. A. Ogle, Z. Brich, H. R. Loosli, J. Am.
Chem. Soc. 1985, 107, 1810; c) W. Bauer, T. Clark, P. v. R. Schleyer,
J. Am. Chem. Soc. 1987, 109, 970.
[20] a) U. Schꢈmann, J. Kopf, E. Weiss, Angew. Chem. 1985, 97, 222;
Angew. Chem. Int. Ed. Engl. 1985, 24, 215; b) L. M. Engelhardt, W.-
P. Leung, C. L. Raston, G. Salem, P. Twiss, A. H. White, J. Chem.
Soc. Dalton Trans. 1988, 2403.
[21] C. Strohmann, V. H. Gessner, Angew. Chem. 2007, 119, 4650;
Angew. Chem. Int. Ed. 2007, 46, 4566.
[22] T. Tatic, S. Hermann, M. John, A. Loquet, D. Stalke, Angew. Chem.
2011, 123, 6796; Angew. Chem. Int. Ed. 2011, 50, 6666.
[23] a) I. P. C. M. van Dongen, H. W. D. Dijkman, M. J. A. Bie, Recl.
Trav. Chim. Pays-Bas. 1974, 93, 29; b) A. Maercker, R. Stotzel, J.
Organomet. Chem. 1983, 254, 1.
[24] K. Takahashi, Y. Kondo, R. Asami, Org. Magn. Reson. 1974, 6, 580.
[25] T. Ruhland, R. Hoffmann, S. Schade, G. Boche, Chem. Ber. 1995,
128, 551.
[10] M. R. Luderer, W. F. Bailey, M. R. Luderer, J. D. Fair, R. J. Dancer,
M. B. Sommer, Tetrahedron: Asymmetry 2009, 20, 981.
[11] a) O. Chuzel, O. Riant, Top. Organomet. Chem. 2005, 15, 59;
b) M. J. McGrath, P. OꢇBrien, J. Am. Chem. Soc. 2005, 127, 16378;
c) C. Genet, S. J. Canipa, P. OꢇBrien, S. Taylor, J. Am. Chem. Soc.
2006, 128, 9336; d) J. Granander, F. Secci, S. J. Canipa, P. OꢇBrien,
B. Kelly, J. Org. Chem. 2011, 76, 4794, and references therein.
[12] a) D. Hoppe, F. Hintze, P. Tebben, M. Paetow, H. Ahrens, J.
Schwerdtfeger, P. Sommerfeld, J. Haller, W. Guarnieri, S. Kolczew-
ski, T. Hense, I. Hoppe, Pure Appl. Chem. 1994, 66, 1479; b) P.
Beak, A. Basu, D. J. Gallagher, Y. S. Park, S. Thayumanavan, Acc.
Chem. Res. 1996, 29, 552; c) C. Serino, N. Stehle, Y. S. Park, S.
Florio, P. Beak, J. Org. Chem. 1999, 64, 1160.
[13] Review: a) I. Kuzu, I. Krummenacher, J. Meyer, F. Armbruster, F.
Breher, Dalton Trans. 2008, 5836; see also, for instance b) F.
Armbruster, I. Fernandez, F. Breher, Dalton Trans. 2009, 5612; c) I.
Krummenacher, H. Rꢈegger, F. Breher, Dalton Trans. 2006, 1073;
d) H. R. Bigmore, J. Meyer, I. Krummenacher, H. Rꢈegger, E. Clot,
P. Mountford, F. Breher, Chem. Eur. J. 2008, 14, 5918; e) I. Kuzu, I.
Krummenacher, I. J. Hewitt, Y. Lan, V. Mereacre, A. K. Powell, P.
Hçfer, J. Harmer, F. Breher, Chem. Eur. J. 2009, 15, 4350; f) M. G.
Cushion, J. Meyer, A. Heath, A. D. Schwarz, I. Fernandez, F.
Breher, P. Mountford, Organometallics 2010, 29, 1174; g) D. Krat-
zert, D. Leusser, D. Stern, J. Meyer, F. Breher, D. Stalke, Chem.
Commun. 2011, 47, 2931.
[14] I. Fernꢀndez, P. Ona-Burgos, F. Armbruster, I. Krummenacher, F.
Breher, Chem. Commun. 2009, 2586.
[15] a) J. Rana, D. J. Robins, J. Chem. Soc. Perkin Trans. 1 1986, 1133;
b) M. Wink, Planta Med. 1987, 53, 509; c) T. Buttler, I. Fleming, S.
Gonsior, B.-H. Kim, A.-Y. Sung, H.-G. Woo, Org. Biomol. Chem.
2005, 3, 1557.
[16] a) D. Margerison, J. P. Newport, Trans. Faraday Soc. 1963, 59, 2058;
b) T. L. Brown, Acc. Chem. Res. 1968, 1, 23; c) S. Bywater, P. La-
chance, D. J. Worsfold, J. Phys. Chem. 1975, 79, 2148; d) See Sup-
porting Information for E. M. Arnett, F. J. Fisher, M. A. Nichols,
A. A. Ribeiro, J. Am. Chem. Soc. 1990, 112, 801.
[17] a) D. Waldmꢈller, B. J. Kotsatos, M. A. Nichols, P. G. Williard, J.
Am. Chem. Soc. 1997, 119, 5479; see also b) D. Hꢈls, H. Gꢈnther, G.
van Koten, P. Wijkens, J. T. B. H. Jastrzebski, Angew. Chem. 1997,
109, 2743; Angew. Chem. Int. Ed. Engl. 1997, 36, 2629.
[26] H. J. Reich, D. P. Green, M. A. Medina, W. S. Goldenberg, B. O.
Gudmundsson, R. R. Dykstra, N. H. Phillips, J. Am. Chem. Soc.
1998, 120, 7201.
[27] a) W. Bauer, W. R. Winchester, P. v. R. Schleyer, Organometallics
1987, 6, 2371; b) R. J. Wehmschulte, P. P. Power, J. Am. Chem. Soc.
1997, 119, 2847.
[28] a) S. T. Kerrick, P. Beak, J. Am. Chem. Soc. 1991, 113, 9708; b) A.
Basu, P. Beak, J. Am. Chem. Soc. 1996, 118, 1575; c) S. Wu, S. Lee,
P. Beak, J. Am. Chem. Soc. 1996, 118, 715; d) N. A. Nikolic, P. Beak,
Org. Synth. 1997, 74, 23.
[29] Note that these are gas-phase structures, not considering important
effects of the solvent (see the PGSE, NOESY and HOESY sec-
tions).
[30] a) D. Hoffmann, D. B. Collum, J. Am. Chem. Soc. 1998, 120, 5810;
b) For another important work on this topic see Ref. [9g].
[31] 6Li NMR data for dimer 7: dLi =1.77 and 1.74 ppm. See Ref. [30] for
more details.
[32] A detailed investigation of the transient species formed in the way
of getting 8, most probably mixed aggregates between benzyllithium
and n-butyllithium, is currently carried out in our laboratories.
[33] W. Boczon, B. Koziol, J. Mol. Struct. 1997, 403, 171.
[34] C. S. Johnson, Jr., Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203.
[35] P. Stilbs, Prog. Nucl. Magn. Reson. Spectrosc. 1987, 19, 1.
[36] a) Y. Cohen, L. Avram, L. Frish, Angew. Chem. 2005, 117, 524;
Angew. Chem. Int. Ed. 2005, 44, 520; b) T. Brand, E. J. Cabrita, S.
Berger, Prog. Nucl. Magn. Reson. Spectrosc. 2005, 46, 159; c) P. S.
Pregosin, P. G. A. Kumar, I. Fernandez, Chem. Rev. 2005, 105, 2977.
[37] a) F. Breher, J. Grunenberg, S. C. Lawrence, P. Mountford, H. Rꢈeg-
ger, Angew. Chem. 2004, 116, 2575; Angew. Chem. Int. Ed. 2004, 43,
2521; b) I. Fernꢀndez, E. Martinez-Viviente, P. S. Pregosin, Inorg.
Chem. 2004, 43, 4555; c) I. Fernꢀndez, E. Martinez-Viviente, P. S.
Pregosin, Inorg. Chem. 2005, 44, 5509; d) I. Fernꢀndez, F. Breher,
P. S. Pregosin, Z. F. Fei, P. J. Dyson, Inorg. Chem. 2005, 44, 7616;
e) I. Fernꢀndez, E. Martinez-Viviente, F. Breher, P. S. Pregosin,
Chem. Eur. J. 2005, 11, 1495.
[38] a) I. Keresztes, P. G. Williard, J. Am. Chem. Soc. 2000, 122, 10228;
b) D. Li, I. Keresztes, R. Hopson, P. G. Williard, Acc. Chem. Res.
2009, 42, 270; c) G. Kagan, W. Li, R. Hopson, P. G. Williard, Org.
Lett. 2010, 12, 520; d) W. Li, G. Kagan, R. Hopson, P. G. Williard,
ARKIVOC 2011, 180.
[39] a) H. J. Reich, J. E. Holladay, J. D. Mason, W. H. Sikorski, J. Am.
Chem. Soc. 1995, 117, 12137; b) H. J. Reich, W. H. Sikorski, J. Org.
Chem. 1999, 64, 14; c) H. J. Reich, A. W. Sanders, A. T. Fiedler,
M. J. Bevan, J. Am. Chem. Soc. 2002, 124, 13386.
[18] For previous spectroscopic studies confirming the structure of (nBu-
Li·tmeda)2 (3), see: a) D. Seebach, R. Hassig, J. Gabriel, Helv.
Chim. Acta 1983, 66, 308; b) W. Bauer, P. v. R. Schleyer, J. Am.
Chem. Soc. 1989, 111, 7191; c) J. F. McGarrity, C. A. Ogle, J. Am.
Chem. Soc. 1985, 107, 1805; d) G. Fraenkel in Lithium: Current Ap-
plications in Science Medicine, and Technology (Ed.: R. O. Bach),
Wiley, New York, 1985; e) W. Baumann, Y. Oprunenko, H. Gunther,
Z. Naturforsch. A 1995, 50, 429; f) G. Crassous, M. Abadie, F. Schue,
Eur. Polym. J. 1979, 15, 747; g) J. M. Saꢀ, G. Martorell, A. Frontera,
J. Org. Chem. 1996, 61, 5194. For crystallographic studies see:
[40] a) W. Bauer, P. v. R. Schleyer in Adv. Carbanion Chem., Vol. 1 (Ed.:
V. Snieckus), JAI Press, Greenwich, 1992, p. 89; b) H. Gꢈnther in
Advanced Applications of NMR to Organometallic Chemistry (Eds.:
700
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 691 – 701

Solution Behaviour of Benzyllithium/(ꢀ)-Sparteine Adducts
FULL PAPER
M. Gielen, R. Willem, B. Wrackmeyer), John Wiley, New York,
1996, Chapter 9, pp. 247–290.
[41] W. Bauer, Magn. Reson. Chem. 1996, 34, 532.
J. Reupohl, R. Frçhlich, D. Hoppe, Angew. Chem. 2004, 116, 1447;
Angew. Chem. Int. Ed. 2004, 43, 1423; e) W. Zeng, R. Frçhlich, D.
Hoppe, Tetrahedron 2005, 61, 3281; f) S. J. Lee, P. Beak, J. Am.
Chem. Soc. 2006, 128, 2178.
[42] Modified pulse field gradient (PFG)-enhanced inverse (1H)-detected
2D HOESY pulse sequences have been introduced for the acquisi-
tion of 1H–7Li heteronuclear correlations. See, for instance: T. M.
Alam, D. M. Pedrotty, T. J. Boyle, Magn. Reson. Chem. 2002, 40,
361.
[47] A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen, F. J.
Timmers, Organometallics 1996, 15, 1518.
[48] H. J. V. Tyrrell, K. R. Harris, Diffusion in Liquids, Butterworths,
London, 1984.
[43] Y. S. Park, M. L. Boys, P. Beak, J. Am. Chem. Soc. 1996, 118, 3757.
[44] 1H NMR experiments were performed by using a zg standard se-
quence with a pulse width of 308, a relaxation delay of 10 s and an
accumulation of 16 scans. Typical experimental time was about
5 min. Compounds PhCH2SnMe3 (11) and nBuSnMe3 (12) have
been previously prepared and structurally characterised. For
PhCH2SnMe3, see: a) W. Adcock, B. D. Gupta, W. Kitching, D. Dod-
drell, M. Geckle, J. Am. Chem. Soc. 1974, 96, 7360; b) Z.-Q. Zhao,
B. Sun, L.-Z. Peng, Y. Li, Y.-L. Li, Chin. J. Chem. 2010, 22, 1382.
For nBuSnMe3, see: I. L. Marr, D. Resales, J. L. Wardell, J. Organo-
met. Chem. 1988, 349, 65.
[49] a) TURBOMOLE, V6.0, 2009, a development of University of
Karlsruhe (TH) and Forschungszentrum Karlsruhe GmbH, 1989–
2007, TURBOMOLE GmbH, available from http://www.turbomo-
le.com since 2007; b) R. Ahlrichs, M. Bꢂr, M. Hꢂser, H. Horn, C.
Kçlmel, Chem. Phys. Lett. 1989, 162, 165; c) O. Treutler, R. Ahl-
richs, J. Chem. Phys. 1995, 102, 346; d) M. Sierka, A. Hogekamp, R.
Ahlrichs, J. Chem. Phys. 2003, 118, 9136; e) R. Ahlrichs, Phys.
Chem. Chem. Phys. 2004, 6, 5119.
[50] a) J. C. Slater, Phys. Rev. 1951, 81, 385; b) S. H. Vosko, L. Wilk, M.
Nusair, Can. J. Phys. 1980, 58, 1200; c) A. D. Becke, Phys. Rev. A
1988, 38, 3098; d) J. P. Perdew, Phys. Rev. B 1986, 33, 8822.
[51] a) F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297;
b) F. Weigend, Phys. Chem. Chem. Phys. 2006, 8, 1057.
[45] We also noticed that, as expected, the system sBuLi·(ꢀ)-sparteine
reacts much faster at ꢀ788C furnishing 11 in >95% yield after
30 min (Figure S11 in the Supporting Information). We thank one
reviewer of the manuscript for helpful suggestions.
[52] P. Deglmann, F. Furche, R. Ahlrichs, Chem. Phys. Lett. 2002, 362,
511.
[46] a) G. S. Gil, U. M. Groth, J. Am. Chem. Soc. 2000, 122, 6789; b) C.
Schultz-Fademrecht, B. Wibbeling, R. Frçhlich, D. Hoppe, Org. Lett.
2001, 3, 1221; c) D. J. Pippel, G. A. Weisenburger, N. C. Faibish, P.
Beak, J. Am. Chem. Soc. 2001, 123, 4919; d) M. Seppi, R. Kalkofen,
Received: July 2, 2012
Revised: October 13, 2012
Published online: November 23, 2012
Chem. Eur. J. 2013, 19, 691 – 701
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
701