DOI: 10.1002/chem.201601494
Communication
& Organometallic Chemistry |Hot Paper |
Solution Structure of Turbo-Hauser Base TMPMgCl·LiCl in
[D8]THF**
Roman Neufeld and Dietmar Stalke*[a]
Abstract: Turbo-Hauser bases are very useful and highly
reactive organometallic reagents in synthesis. Especially
TMPMgCl·LiCl 1 (TMP = 2,2,6,6-tetramethylpiperidide) is an
excellent base for converting a wide range of (hetero)aromatic substrates into highly functionalized compounds
with a broad application in organic synthesis. The knowledge of its structure in solution is of essential importance
to understand the extraordinary reactivity and selectivity.
However, very little is known about the aggregation of
this prominent reagent in solution. Herein, we present the
THF-solution structure of 1 by employing our newly elaborated DOSY NMR method based on external calibration
curves (ECC) with normalized diffusion coefficients.
The regio- and stereoselective deprotonation and functionalization of aromatic compounds using amide bases or alkyl organometallics is one of the most versatile and applied synthetic transformations. As a drawback, strong organic bases such
as alkyl amides require relatively low temperatures (¢78 8C)
and show competing addition reactions (e.g. Chichibabin reactions). A better selectivity was achieved by Hausers’ magnesium bases R2NMgX.[1] These are analogues to Grignards’ reagents, but instead of an alkyl substituent, the magnesium
halide is connected to an amido moiety. Unfortunately, most of
them show a poor solubility in THF and, as a consequence, the
metalation rates are very slow. It is well-known that numerous
metallic salts are more soluble when LiCl is present.[2] This feature has led to the design of LiCl-solubilized TMPMgCl·LiCl
1 and its less bulky Turbo analogue DAMgCl·LiCl 2 (DA = diisopropylamide, Figure 1 a).[3] Equipped with enhanced kinetic basicity, these commercially available reagents display a high reactivity even at room temperature (RT) as well as excellent regioselectivity and high functional-group tolerance for a large
number of aromatic and heteroaromatic substrates.[4] Although
there is a great deal of information on the utility of these reagents, very little is known regarding the nature of TurboHauser base 1 in solution. Knochel et al.,[3] Garc†a-Ýlvarez and
Mulvey et al.[5] suggested that added LiCl deaggregates RMgX
[a] Dr. R. Neufeld, Prof. Dr. D. Stalke
Institut fìr Anorganische Chemie, Georg-August-Universit•t
Tammannstraße 4, 37077 Gçttingen (Germany)
E-mail: [email protected]
[**] TMP = 2,2,6,6-tetramethylpiperidide.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201601494.
Chem. Eur. J. 2016, 22, 12624 – 12628
Figure 1. a) Typical notation of Turbo-Hauser bases TMPMgCl·LiCl 1 and
DAMgCl·LiCl 2. b) The most plausible solution structures of 1 and 2 in
[D8]THF solution.[8]
oligomers to furnish magnesiate character to the Grignard reagent by a solvent-separated ion pair (SSIP) [Li(thf)4] +
[RMg(thf)Cl2]¢ .[6] At least, in the solid state the Turbo-Hauser
bases TMPMgCl·LiCl[7] 1 C and DAMgCl·LiCl[5] 2 A (Figure 1 b)
supported the existence of a contact ion pair (CIP). Recently,
we showed that redissolved crystals of 2 A do not form SSIPs
in THF solution either.[9] Instead, the dimeric crystal structure
2 A equilibrates with its monomeric counterpart 2 C at room
temperature. In both structures, LiCl co-coordinates to both
Hauser bases. Lowering the temperature triggers the formation
of more oligomers in the equilibrium. First LiCl gets separated
from 2 A and forms a well-known [(thf)2Li(m-Cl)2Li(thf)2]-dimer
12624
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
(Li1).[10] While the Hauser base keeps the dimeric aggregation
(2 B) its interaction with Li1 inhibits a Schlenk-equilibriumdriven transformation to the homoleptic side and the anticipated formation of DA2Mg and MgCl2. Since both magnesium
amides 1 and 2 show a different reactivity and regioselectivity,[3] it is very important to understand how TMPMgCl·LiCl 1 organizes and interacts, especially in solution. The method of
choice to identify species in solution is NMR spectroscopy.
Garc†a-Ýlvarez and Mulvey et al. analyzed crystals of 1 in
[D8]THF solution[5] by diffusion-ordered spectroscopy (DOSY)
and the diffusion coefficient formula weight (D-FW) analysis
that was pioneered by Li and Williard et al.[11] This method is
based on internal calibration curves (ICC), for which many internal standards that may interfere with the reactive metal
complexes are required. Because of peak overlap problems the
authors had to use inappropriate internal standards,[12] so the
molecular weight (MW) determination was prone to a relatively
high error of approximately œ 30 %. Consequently Garc†a-Ýlvarez and Mulvey et al. stated that they were not able to “clearly
establish the exact nature of the solution species”.[5] Predominantly it could not be established whether or not lithium chloride is co-coordinated to the magnesium amide. In the end, it
was concluded that a SSIP situation appeared to be the most
feasible. In the following section we will shed light on the solution structure of 1 and prove that LiCl does indeed coordinate
to Turbo-Hauser base 1. Recently, we developed an advanced
DOSY NMR technique that enables the identification of reactive species in solution.[13] This method for accurate MW-determination rests on external calibration curves (ECC) with normalized diffusion coefficients. Only one internal reference is
sufficient and the addition of multiple internal references is
dispensable. Furthermore, it is independent of diversities in
temperature and viscosity and offers an easy and robust approach to determine accurate MWs in THF solution with an
error of less than œ 9 %.[14] Employing this method, we were already able to characterize the complex oligomeric mixture of
lithium diisopropylamide (LDA) in toluene solution[15] and the
structure of DAMgCl and DAMgCl·LiCl 2 in [D8]THF solution.[9]
In the current paper, we describe DOSY-ECC-MW-determinations of the TMP-Turbo-Hauser base TMPMgCl·LiCl 1 in [D8]THF
solution.
Primarily it seems advisable to a priori rationalize which species are feasible to be present in the solution of 1. The most
obvious question to be addressed in s-block organometallics
of course is the amount of coordinating THF molecules to be
present in the solution structure of 1. Even from a vast
number of solid-state structures it is known that polar solvents
like THF are necessary to coordinate such highly ionic compounds. Following the preparation of Mulvey et al.[7] we synthesized the Turbo-Hauser base 1 by reaction of equimolar
LiTMP with a suspension of MgCl2 in THF and then stirring the
mixture for one day at room temperature. Removing the solvent in vacuo and recrystallization (2 Õ) in a 1:1 THF/hexane
mixture at ¢45 8C afforded 1 as colorless crystals in 17 %
yield.[16] A redissolved crystal of 1 in neat THF could on the
one hand retain its solid state structure (1 C) or even aggregate
further to combine to dimers (1 A, 1 B) as well as dissociate to
Chem. Eur. J. 2016, 22, 12624 – 12628
www.chemeurj.org
a number of smaller molecules (1 C*–1 E, Figure 1 b). LiCl can
either coordinate to the magnesium amide or dissociate as
a well-known [(thf)2Li(m-Cl)2Li(thf)2] dimer[10] Li1. A solvent-separated ion pair (SSIP) containing [Li(thf)4] + Li2 would promote
the formation of an ate-complex 1 E in which two chloride ions
coordinate a single magnesium cation. From the crystal structure of 1 it is known that the lithium cation is located at an
average distance of 4.68 æ to the closest CH3-protons of the
TMP ligand.[7] This relatively close distance should be detectable in a 1H–7Li-HOESY experiment for which the structure is retained in solution.[13e] For NMR spectroscopic measurements,
we used diluted solutions of 1 (20 mm) by dissolving the crystals in [D8]THF.[17] The 1H NMR spectra of 1 show at all temperatures (25 8C to ¢75 8C) only one single type of TMP ligand (d =
1.19/1.22/1.62 ppm for CH3/b-CH2/g-CH, Figure 2), whereas the
Turbo-Hauser base 2 displays several oligomers.[18] Indeed the
Figure 2. Superposition of 1H NMR spectra of redissolved crystalline 1 in
[D8]THF at various temperatures.
1
H–7Li-HOESY-spectra at all temperatures display a cross peak
between lithium and the CH3 groups confirming the lithium
coordination to the magnesium amide (see Figures S3 and S4
in the Supporting Information).[19] The 7Li NMR spectra show
one singlet at about d = 0.2 ppm in the whole temperature
range (see Figure S2 in the Supporting Information). The presence of remaining LiTMP can be excluded since it resonates at
d = 0.7 (monomer) or at 1.3 ppm (dimer), respectively.[20] Additionally, the lithium cation [Li(thf)4] + Li2 in a SSIP can also be
excluded to be present at significant concentrations since it is
known to resonate at negative field (d = ¢1.1 ppm),[21] curtailing the plausible present species to the lithium-containing
dimer 1 A and the monomers 1 C/1 C*. Both aggregates should
clearly be distinguishable by DOSY NMR spectroscopy. The 1H
and 7Li-DOSY-ECC-MW-determination of 1 gives a MW of
MWdet = 403 g mol¢1 for the 1H and 387 g mol¢1 for the 7Li nucleus (Figure 3).[22] This low MW discriminates both the lithium
free aggregates 1 B (MWerr = 26 %), 1 D (MWerr = ¢17 %), 1 E
(MWerr = ¢31 %), and also dimeric 1 A (MWerr = 48 %).[23] Most interestingly, the crystal structure 1 C does not keep its full integrity in [D8]THF solution reflected by an unacceptably high error
of the MW (MWcalcd = 459 g mol¢1, MWdet = 403 g mol¢1, MWerr =
12 %). The smaller contact ion pair 1 C* derived from 1 C by
the loss of a single THF molecule matches best the determined
12625
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Our investigations confirm that lithium chloride
indeed co-coordinates to the magnesium amides in
solution as observed in the Turbo-Hauser base 2.[9] A
detectable population of the solvent separated lithium cation Li2 can also be excluded for Turbo-Hauser
base 1 (MWcalcd = 295 g mol¢1, MWdet = 387 g mol¢1,
MWerr = ¢31 %). Changing the steric bulk and the
electronic properties of the amide strongly controls
the structural features, both in the solid state as well
as in solution. While 1 is a monomer in the solid
state and in solution, 2 is a dimer in the crystal structure and forms predominately dimeric aggregates in
THF solution.[9] The deviation in the aggregation
state and steric demand is also reflected in the different reactivity[3] and regioselectivity[5] of both TurboHauser bases (Scheme 1). From our measurements,
the different reactivities can be rationalized as follows: The TMP ligand in TMPMgCl·LiCl 1 is bulky
enough to prevent dimerization (Figure 3) and to
promote a LiCl solubilized monomer 1 C*. Similarly
the appreciable bulk provided by the fixed methyl
groups facilitates the cleavage of a labile THF ligand
and induces an unsaturated magnesium site. In DAMgCl·LiCl 2 the rigid TMP ligand is replaced by two
floppy diisopropyl groups and the methyl groups are
not fixed in a static position relative to the metal.
This flexibility permits both a direct THF coordination
at the Mg cation and a dimerization of the lithiumstabilized monomer 2 C to form the tetranuclear
dimer 2 A. The latter is, at high concentrations, the
main aggregate in solution.[5, 9] This would also explain the limited solubility of dimeric 2 in THF (0.6 m)
compared to the monomeric TMP-Turbo-Hauser base
1 (1.2 m). In general, monomeric species display the
most active kinetic species in organolithium chemistry.[27] This could explain why reactions of the TMPTurbo-Hauser base are much faster than that of diFigure 3. 1H- and 7 Li-DOSY-ECC-MW-determination[8] of crystalline 1 redissolved in
meric DA-Turbo-Hauser base (Scheme 1a). In addi[D8]THF (20 mm). The accuracy of this method is in the range of MWerr ‹ œ 9 %.[14]
tion, we suggest that the highly regioselective ortho
deprotonation reactions could stem from a sufficient
MW (MWcalcd = 387 g mol¢1, MWdet = 403 g mol¢1, MWerr = ¢4 %).
complex-induced proximity effect (CIPE)[28] between the bimetMulvey et al. already mentioned the labile THF coordination at
allic aggregate 1 C* and the functionalized (hetero)aromatic
1 C as the expected integrated 1H NMR intensity. THF/amide
substrates 7 and 8 in Scheme 2. Compared to the TMP ligand,
decreases from 3:1 to 2:1 when the crystals are dried in
the diisopropyl amide ligand provides lower steric hindrance
vacuo.[7] From this they concluded that the powerful regioseand facilitates the approach to the carbonyl function (see the
lective magnesiating ability of 1 might be a consequence of
reaction in Scheme 1b). Further, due to the lower basicity (DA
vs. TMP ligand: pKa = 34 vs 38)[29] the addition to the nucleoa coordinately unsaturated Mg that “could facilitate the pre-coordination of the functionalized aromatic substrate prior to
philic carbon center is favorable (thermodynamic product). In
magnesiation”.[7] Our DOSY-ECC-MW-determination fully supcontrast, due to the high steric demand of the TMP ligand and
ports this hypothesis. The labile THF ligand at the magnesium
its higher basicity, the deprotonation reaction is favored over
atom could be a result of steric overload from the rigid TMP
the addition reaction (kinetic product).
ligand. Searching the Cambridge Crystallographic Database[24]
emphasizes Mg typically to be coordinated by four to six liExperimental Section
gands. Enlarging the bulk of the various substituents, however,
results also in 3-coordinated magnesium amides. This is often
Dry [D8]THF stored with 4 æ molecular sieves under argon was
the case when bulky SiMe3 groups[25] and especially when TMP
used. The NMR samples were prepared by dissolving crystalline
ligands are involved.[26]
1 and the DOSY reference 1-phenylnaphtaline (PhN) in equimolar
Chem. Eur. J. 2016, 22, 12624 – 12628
www.chemeurj.org
12626
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
Acknowledgements
We acknowledge funding from the Danish National Research
Foundation (DNRF93) funded by Centre for Materials Crystallography (CMC).
Keywords: diffusion · metallation · organometallic chemistry ·
NMR spectroscopy · reactive species
Scheme 1. a) DAMgCl·LiCl 2 displays a much lower reactivity than its TMP
counterpart 1. That difference in reactivity was shown by the deprotonation
of isoquinoline in THF solution. Whereas TMPMgCl·LiCl 1 required only 2 h
and 1.1 equivalents, 2 needed 12 h and two equivalents for comparable
metalation.[3] b) Whereas 1 easily metalates ethyl-3-chloro-benzoate 5 in the
C2-position to give, after iodation, benzene 6, the same reaction carried out
with 2 results in no metalation at all. Instead, an addition–elimination reaction occurs with the formation of m-chloro-N,N-diisopropylbenzamide 4.[5]
Scheme 2. Complex-induced proximity effect (CIPE):[28] Proposed transition
states for 1 C* in some regioselective ortho deprotonation reactions with
functionalized (hetero)aromatic compounds. FG = coordinating functional
group; X = heteroatom.
ratio (each 20 mm) in [D8]THF. The diffusion coefficients of the
amide species were normalized to the fixed diffusion value of the
reference PhN (logDref,fix(PhN) = ¢8.8812; for more information see
the Supporting Information). NMR spectra were recorded on
a Bruker Advance 400 spectrometer equipped with an obverse
broadband probe with z-axis gradient coil with maximum gradient
strength of 57 G cm¢1. All spectra were acquired using 5 mm NMR
tubes, which were not spinning during the measurements. All
DOSY experiments were performed using a double-stimulated
echo sequence with bipolar gradient pulses and three spoil gradients with convection compensation (dstebpgp3s).[30] The duration
of the magnetic field pulse gradients was adjusted for every temperature in a range of d/2 = 400–3500 ms. The diffusion time was
D = 0.1 s. The delay for gradient recovery was 0.2 ms and the eddy
current delay 5 ms. In each PFG NMR experiment, a series of 16
spectra on 32 K data points were collected. The pulse gradients
were incremented from 2 to 98 % of the maximum gradient
strength in a linear ramp. After Fourier transformation and baseline
correction, the diffusion dimension was processed with the Topspin 3.1 software. Diffusion coefficients, processed with a line
broadening of 2 Hz, were calculated by Gaussian fits with the T1/
T2 software of Topspin.
Chem. Eur. J. 2016, 22, 12624 – 12628
www.chemeurj.org
[1] C. R. Hauser, H. G. Walker, J. Am. Chem. Soc. 1947, 69, 295 – 297.
[2] a) D. Stern, N. Finkelmeier, D. Stalke, Chem. Commun. 2011, 47, 2113 –
2115; b) F. M. Piller, P. Appukkuttan, A. Gavryushin, M. Helm, P. Knochel,
Angew. Chem. Int. Ed. 2008, 47, 6802 – 6806; Angew. Chem. 2008, 120,
6907 – 6911; c) J. P. Snyder, J. Am. Chem. Soc. 1995, 117, 11025 – 11026;
d) T. Stemmler, J. E. Penner-Hahn, P. Knochel, J. Am. Chem. Soc. 1993,
115, 348 – 350.
[3] A. Krasovskiy, V. Krasovskaya, P. Knochel, Angew. Chem. Int. Ed. 2006, 45,
2958 – 2961; Angew. Chem. 2006, 118, 3024 – 3027.
[4] a) D. Tilly, F. Chevallier, F. Mongin, P. C. Gros, Chem. Rev. 2014, 114,
1207 – 1257; b) T. Klatt, J. T. Markiewicz, C. S•mann, P. Knochel, J. Org.
Chem. 2014, 79, 4253 – 4269; c) F. Mongin, A. Harrison-Marchand, Chem.
Rev. 2013, 113, 7563 – 7727; d) N. M. Barl, V. Werner, C. S•mann, P. Knochel, Heterocycles 2013, 88, 827 – 844; e) H. Ila, O. Baron, A. J. Wagner, P.
Knochel, Chem. Lett. 2006, 35, 2 – 7; f) H. Ila, O. Baron, A. J. Wagner, P.
Knochel, Chem. Commun. 2006, 583 – 593.
[5] D. R. Armstrong, P. Garc†a-Ýlvarez, A. R. Kennedy, R. E. Mulvey, J. A. Parkinson, Angew. Chem. Int. Ed. 2010, 49, 3185 – 3188; Angew. Chem. 2010,
122, 3253 – 3256.
[6] a) A. Krasovskiy, B. F. Straub, P. Knochel, Angew. Chem. Int. Ed. 2006, 45,
159 – 162; Angew. Chem. 2006, 118, 165 – 169; b) A. Krasovskiy, P. Knochel, Angew. Chem. Int. Ed. 2004, 43, 3333 – 3336; Angew. Chem. 2004,
116, 3396 – 3399.
[7] P. Garc†a-Ýlvarez, D. V. Graham, E. Hevia, A. R. Kennedy, J. Klett, R. E.
Mulvey, C. T. O’Hara, S. Weatherstone, Angew. Chem. Int. Ed. 2008, 47,
8079 – 8081; Angew. Chem. 2008, 120, 8199 – 8201.
[8] Deuterated species diffuse like their protonated counterparts. This is
why we used MWcalc(THF) = 72.11 g mol¢1 in the calculation of all THF
coordinated species. For more information see ref. [13].
[9] R. Neufeld, T. L. Teuteberg, R. Herbst-Irmer, R. A. Mata, D. Stalke, J. Am.
Chem. Soc. 2016, 138, 4796 – 4806.
[10] a) A. V. Yakimansky, A. H. E. Mìller, M. Van Beylen, Macromolecules 2000,
33, 5686 – 5692; b) W. Ming Keong Wong, A. I. Popov, J. Inorg. Nucl.
Chem. 1972, 34, 3615 – 3622; c) S. H. Bauer, T. Ino, R. F. Porter, J. Chem.
Phys. 1960, 33, 685 – 691.
[11] a) J. Guang, R. Hopson, P. G. Williard, J. Org. Chem. 2015, 80, 9102 –
9107; b) D. Li, I. Keresztes, R. Hopson, P. G. Williard, Acc. Chem. Res.
2008, 41, 270 – 280.
[12] The error of the DOSY molecular weight (MW) analysis depends highly
on the geometry of the used references as well on that of the analyte
(see ref. [14]). Hauser bases have an ellipsoidal geometry. Therefore ellipsoid references should be used for an accurate MW analysis. However, the authors from ref. [5] used references with different geometries
(spherical, ellipsoidal, and flat discs) that dramatically decreased the accuracy of the MW determination.
[13] a) T. Niklas, D. Stalke, M. John, Chem. Commun. 2015, 51, 1275 – 1277;
b) A.-C. Pçppler, M. Granitzka, R. Herbst-Irmer, Y.-S. Chen, B. B. Iversen,
M. John, R. A. Mata, D. Stalke, Angew. Chem. Int. Ed. 2014, 53, 13282 –
13287; Angew. Chem. 2014, 126, 13498 – 13503; c) A.-C. Pçppler, S.
Frischkorn, D. Stalke, M. John, ChemPhysChem 2013, 14, 3103 – 3107;
d) A.-C. Pçppler, M. M. Meinholz, H. Faßhuber, A. Lange, M. John, D.
Stalke, Organometallics 2012, 31, 42 – 45; e) M. Granitzka, A.-C. Pçppler,
E. K. Schwarze, D. Stern, T. Schulz, M. John, R. Herbst-Irmer, S. K. Pandey,
D. Stalke, J. Am. Chem. Soc. 2012, 134, 1344 – 1351; f) A.-C. Pçppler, H.
Keil, D. Stalke, M. John, Angew. Chem. Int. Ed. 2012, 51, 7843 – 7846;
Angew. Chem. 2012, 124, 7963 – 7967.
[14] a) R. Neufeld, D. Stalke, Chem. Sci. 2015, 6, 3354 – 3364; b) S. Bachmann,
R. Neufeld, M. Dzemski, D. Stalke, Chem. Eur. J. 2016, 22, 8462 – 8465.
12627
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Communication
[15] R. Neufeld, M. John, D. Stalke, Angew. Chem. Int. Ed. 2015, 54, 6994 –
6998; Angew. Chem. 2015, 127, 7100 – 7104.
[16] The crystal structure was proven by X-ray diffraction.
[17] The herein used external calibration curves work best with diluted solutions; this is why we chose 20 mm solutions of crystalline 1 C, considering the loss of one molecule of THF of 1 C in the drying process.
[18] A small amount of protonated amine TMP(H) is also present in solution.
The ECC-MW-determination predicts an accurate MW at temperatures
between 0 and ¢75 8C (in av.: MWcalcd = 141 g mol¢1, MWdet =
140 g mol¢1, MWerr = 1 %). Interestingly, at room temperature, the MW is
slightly overestimated (MWdet = 167 g mol¢1, MWerr = ¢18 %), indicating
an exchange between the amide and the amine. Anyway, this interaction is neglectible at temperatures below 0 8C, see Table S4 in the Supporting Information.
[19] Similar results were observed for Hauser base 2, see ref. [9].
[20] P. Renaud, M. A. Fox, J. Am. Chem. Soc. 1988, 110, 5702 – 5705.
[21] C. Elschenbroich, F. Hensel, H. Hopf, Organometallchemie, Vol 6, B. G.
Teubner Verlag, Wiesbaden, 2008.
[22] All MWdet values are dispayed for each species as an average value, derived from DOSY-ECC-MW measurements at different temperatures, see
Tables S5 and S6 in the Supporting Information.
[23] The deviation is calculated by MWerr = [1¢MWdet/MWcalc]·100 %, in which
MWdet is the experimentally determined and MWcalcd is the calculated
molecular weight.
Chem. Eur. J. 2016, 22, 12624 – 12628
www.chemeurj.org
[24] F. H. Allen, Acta Crystallogr. Sect. B 2002, 58, 380 – 388; CSD version 5.36
(Updated Nov 2014).
[25] a) J. Francos, B. J. Fleming, P. Garcia-Alvarez, A. R. Kennedy, K. Reilly,
G. M. Robertson, S. D. Robertson, C. T. O’Hara, Dalton Trans. 2014, 43,
14424 – 14431; b) B. Baishya, J. Organomet. Chem. 2014, 769, 112 – 118.
[26] a) A. J. Mart†nez-Mart†nez, D. R. Armstrong, B. Conway, B. J. Fleming, J.
Klett, A. R. Kennedy, R. E. Mulvey, S. D. Robertson, C. T. O’Hara, Chem.
Sci. 2014, 5, 771 – 781; b) E. Hevia, A. R. Kennedy, R. E. Mulvey, S. Weatherstone, Angew. Chem. Int. Ed. 2004, 43, 1709 – 1712; Angew. Chem.
2004, 116, 1741 – 1744.
[27] a) J. Liang, A. C. Hoepker, R. F. Algera, Y. Ma, D. B. Collum, J. Am. Chem.
Soc. 2015, 137, 6292 – 6303; b) J. Liang, A. C. Hoepker, A. M. Bruneau, Y.
Ma, L. Gupta, D. B. Collum, J. Org. Chem. 2014, 79, 11885 – 11902; c) L.
Gupta, A. C. Hoepker, Y. Ma, M. S. Viciu, M. F. Faggin, D. B. Collum, J.
Org. Chem. 2013, 78, 4214 – 4230.
[28] M. C. Whisler, S. MacNeil, V. Snieckus, P. Beak, Angew. Chem. Int. Ed.
2004, 43, 2206 – 2225; Angew. Chem. 2004, 116, 2256 – 2276.
[29] H. Ahlbrecht, G. Schneider, Tetrahedron 1986, 42, 4729 – 4741.
[30] a) A. Jerschow, N. Mìller, J. Magn. Reson. 1997, 125, 372 – 375; b) A. Jerschow, N. Mìller, J. Magn. Reson. 1996, 123, 222 – 225.
Received: March 31, 2016
Published online on June 17, 2016
12628
Ó 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim