вход по аккаунту


Structure Identification of Precatalytic Copper Phosphoramidite Complexes in Solution.

код для вставкиСкачать
Copper Catalysts
DOI: 10.1002/anie.200601880
Structure Identification of Precatalytic Copper
Phosphoramidite Complexes in Solution**
Hongxia Zhang and Ruth M. Gschwind*
Copper salts and organocopper(I) compounds are widely
used for stoichiometric and catalytic additions and substitutions.[1–3] The high synthetic potential of copper(I) complexes
or metal organocuprate clusters and the difficulty in designing
new active copper systems are both based on the ability of
these complexes to form supramolecular clusters, which
represent an intricate example of molecular recognition.[1, 4]
The complexity in determining the structures of these
supramolecular species in solution,[4] the existence of dynamic
[*] MSc H. Zhang, Prof. Dr. R. M. Gschwind
Institut f&r Organische Chemie
Universit+t Regensburg
Universit+tsstrasse 31, 93053 Regensburg (Germany)
Fax: (+ 49) 941-943-4617
[**] This work has been supported by the Fonds der Chemischen
Supporting information (including experimental details) for this
article are available on the WWW under
or from the author.
Angew. Chem. Int. Ed. 2006, 45, 6391 –6394
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
equilibria between several species,[5] and the resulting sensitivity of the reaction to solvent and salt effects[1] [2] have so far
been a hindrance to a rational design capable of tapping the
full potential of copper reagents.
In the case of stoichiometric, diastereoselective lithium
diorganocuprate reagents, we were able to shed some light on
their supramolecular structures in solution as well as the
resulting structure–reactivity correlations.[6] However, in the
vast field of catalytic, asymmetric conjugate addition reactions that use copper salts and chiral ligands, very little is
known about the structures of the precatalytic copper
complexes or the active catalytic species.[1] For example, for
the ground-breaking system for enantioselective additions to
enones (by use of phosphoramidite ligands for coppercatalyzed additions of dialkylzinc),[2, 7] only two crystal
structures of copper complexes with less selective phosphor-
explains for the first time the observations of synthetic
optimization procedures and supports a rational design for
further catalytically active ligands.
To investigate the structures of copper phosphoramidite
complexes in solution, two phosphoramidite ligands that are
highly efficient in 1,4-additions to enones were selected
Scheme 1. Selected phosphoramidite ligands as representatives for the
binaphthol- and biphenol-based ligands developed in the groups of
Feringa[3] and Alexakis.[10]
(Scheme 1): the binaphthol-based ligand 1 developed by
Feringa and co-workers[3, 8] and the biphenol-based ligand 2
developed by Alexakis et al.[10, 11] CuICl was chosen as the
copper salt since it gives complete conversion and high
ee values in several solvents[10] and produces very stable and
diamagnetic NMR samples.
The 31P NMR spectra of 2 and CuCl in different solvents
are presented in Figure 2 a. A 2:1 ratio between ligand and
CuCl was selected, which was found to be the optimal ratio in
synthetic procedures.[3, 8, 10] Under standard conditions, only
one complex species (C2) and free ligand are observed in
CD2Cl2 and CDCl3, thus indicating C2 to be the active
precatalytic complex. In [D8]THF and [D8]toluene, the
P NMR spectra show broad signals that correspond to
Figure 1. Known crystal structures of copper phosphoramidite complexes. a) [{CuBr(O,O’-(R)-(1,1’-spirobiindane-7,7’-diyl)-N,N’-dimethylphosphoramidite)2}2],[9] b) [CuI(O,O’-(S)-(1,1’-dinaphthyl-2,2’-diyl)-N,N’dimethylphosphoramidite)3].[8] For schematic drawings, see 8 and 7 in
Scheme 2.
amidite ligands are known (Figure 1).[8, 9]
Herein, we present the first structural information about a
precatalytic copper complex with phosphoramidite ligands in
solution under standard conditions as used in synthetic
protocols. The proposed structure is derived from a combination of 31P NMR spectra, NMR spectroscopic diffusion
experiments, elemental analyses, mass-spectrometric investigations, and crystal structures of related compounds. The
presented binuclear copper complex with a mixed trigonal/
tetrahedral coordination environment represents a new
structural type of precatalytic copper complexes. This finding
Figure 2. 31P NMR spectra of the complexes built by 2 and CuCl
(0.02 m) at a ratio of 2:1 (a) and 1:1 (b) in different solvents at 220 K
(the two predominant complex species are labelled as C1 and C2).
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 6391 –6394
several complex species. Nevertheless, two key complexes can
be identified, which are labeled as C1 and C2. Upon reducing
the ligand/CuCl ratio to 1:1 or lower, only C1 is observed in all
solvents (Figure 2 b), independent of temperature and concentration (for details see the Supporting Information). Thus,
in chloroform and dichloromethane, the two key complexes
C1 and C2 can be investigated separately.
Figure 3 summarizes the appearance of complexes C1 and
C2 as a function of the ligand/CuCl ratio.[12] For an excess of
Table 1: Diffusion coefficients D[a] (10 10 m2 s 1) of the free ligands and
complexes C1 and C2 (CuCl and 1 or 2).
Free ligand
[a] Experimental error 2 %, 0.02 m in CDCl3 at 220 K.
ratios between 1:1 and 2:1. Schematic drawings of model
complexes 3–6 with typical configurations of CuI complexes[13]
are shown in Scheme 2. Furthermore, the two known crystal
structures are included as model complexes 7 and 8. The
volumes of C1, C2, and the model complexes 3–8 are given in
Figure 3. 31P NMR spectra of 2 and mixtures of 2 and CuCl at varying
ratios in CDCl3 at 220 K.
CuCl and up to a ratio of 1:1, only C1 (d = 121.7 ppm) is
Scheme 2. Mono-, bi-, and trinuclear model complexes including the
proposed structures of C1 and C2.
observed. Starting from 1.05 equivalents of 2, the signal for C2
emerges at d = 126.6 ppm. Between 1.05 and 1.5 equivalents
Table 2 (for experimental details, see the Supporting Inforof 2, the 31P NMR signal of C2 increases continuously, and the
mation). A comparison of the volumes of 3–8 shows
signal of C1 decreases. At ratios higher than 1.5 equivalents of
significant differences for complexes with 2, 3, or 4 ligands;
2, C2 and increasing amounts of free ligand are observed.
however, varying the number of CuCl units induces only
Identical measurements using the ligand 1 instead of 2 led to
marginal changes. Thus, a comparison of the volumes of 3–8
very similar 31P NMR spectra in CDCl3, thus indicating that
with those of C1 and C2 reveals that: a) both complexes C1
similar complexes are formed (spectra not shown).
and C2 are composed of three ligands, independent of the
Because of the highly efficient quadrupole relaxation of
kind of ligand; b) diffusion-ordered NMR spectroscopy
copper and dynamic processes within the complexes, very
(DOSY) does not allow for determining the absolute
broad 31P NMR signals (n1/2 120 Hz) without 1JCu,P or 2JP,P
number of CuCl units attached; and c) slightly lower volumes
coupling patterns were observed, and diffusion measurements
of C2 compared to C1 hint at a lower amount of CuCl in C2.
on 31P NMR resonances were thus not applicable. Therefore,
The missing information about the ratio between ligand
H diffusion experiments were performed to obtain informaand CuCl in the complexes C1 and C2 is found in Figure 3.
tion about the hydrodynamic radius of C1 and C2. The
Since we found that the solubility of pure CuCl in CHCl3 or
viscosity-corrected diffusion coefficients of 1, 2, and their
complexes C1 and C2 are listed in Table 1. The ligand 1 and
CH2Cl2 is very low (about 1 mm at room temperature), the
the complexes containing 1 show lower diffusion coefficients
given ligand/CuCl ratio represents the composition of the
compared to the compounds with 2, which is consistent with
copper complex. [12] Thus, C1 has a ligand/CuCl ratio of 1:1
the larger molecular size of 1 and its complexes. Surprisingly,
and the formula [(LCuCl)3], and C2 has a ratio of 1.5:1 and
the D values of C1 and C2 with the same ligand do not differ
the formula [L3Cu2Cl2]. The crystal structures of similar
complexes with the identical stoichiometry, [(PPh3)2Cu2X2The information about the molecular
size and the composition of the complexes Table 2: Experimental (C1, C2)[a] and calculated (3–8) volumes [G3] of complexes composed of CuCl and
can be combined to elucidate the complex ligands 1 or 2.
species C2 and C1. The first step is to Ligand C1
4 (2:1)
5 (1:1)
6 (1.5:1)
7 (3:1)
8 (2:1)
3 (1:1)
examine whether the number of ligands
and the amount of CuCl in the complexes 1
can be determined from the diffusion coef- 2
1525[f ]
ficients. For this purpose, four model com[a] Calculated from the experimental diffusion coefficients in Table 1. [b] For each model complex the
plexes were selected which represent all ligand/CuCl ratio is given in brackets. [c] + 383, 314. [d] + 371, 303. [e] + 264, 216. [f ] + 254,
possible Lm(CuCl)n stoichiometries with
207. Calculated from 5 % error in the diffusion coefficient ( 2 % experimental error plus 3 %
two or three ligands as well as ligand/CuCl possible systematic error).
Angew. Chem. Int. Ed. 2006, 45, 6391 –6394
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
(PPh3)] (X = Cl, Br, I)[14] and a [L3Cu2Br2] complex with a
cyclic phosphorous ligand,[15] both exhibit the typical configuration of binuclear complexes with three- and four-coordinate copper centers. Therefore, a binuclear complex with a
mixed trigonal/tetrahedral copper geometry is proposed for
the precatalyst C2 (see 6 in Scheme 2).
Trinuclear copper complexes like C1 are much less
common than binuclear and tetranuclear complexes of CuI,
and there is a lack of X-ray structures of trinuclear copper
complexes with monodentate, nonchelating phosphorous
ligands. Thus, it is only possible to suppose for C1 a planar
hexagonal ring structure (see 5 in Scheme 2) or a triangular
structure from related copper complexes.[13] In agreement
with these proposed structures, the electrospray mass spectra
of a 1:1 mixture of 2 and CuCl in CH2Cl2 show a dominant
fragment-ion peak at m/z 605 ([2·Cu2Cl]+) as the base peak,
whereas a 2:1 mixture produces the fragment-ion peak at m/z
941 ([22·Cu]+) as the base peak and the molecular-ion peak of
free ligand 2 at m/z 438 (see the Supporting Information).
With the identification of C1 and C2, it is now possible to
understand some observations of synthetic optimization
procedures, to propose a refinement of the catalytic cycle,
and thus to support a rational design of further catalytically
active ligands. 1) In conjugate additions to enones, ratios of
ligand to copper salt of less than 1.5:1 were found to be
detrimental,[2] thus suggesting that C2 rather than C1 is the
precursor for the actual enantioselective catalyst (see
Figure 3). 2) The substitution pattern of the amine moiety
of the ligands was found to be decisive for the ee values;
substituents that are too small (for example, methyl groups)
and groups that are too large (for example, 1-naphthyl instead
of the phenyl groups in 1) both lead to reduced ee values.[3] A
comparison of C2 with the crystal structures in Figure 1 shows
that amine moieties that are too small may lead to tetrahedral
coordination of both copper atoms, whereas substituents that
are too large may hinder the formation of a binuclear
structure. 3) In the catalytic cycle for copper-catalyzed 1,4additions of R2Zn reagents, as based on the mononuclear
crystal structure (Figure 1 b), an initial alkyl transfer from zinc
to copper is proposed to be followed by a substitution of one
of the ligands by the p-coordinated enone.[16] Alternatively,
based on the 2:1 ratio of ligand to copper salt that was used in
the synthesis and the observed negative nonlinear effects,
dimeric [L2CuX]2 or mononuclear [L2CuX] precatalysts were
proposed,[1, 3, 7] the latter of which has a coordination number
of three, which is quite uncommon for mononuclear
CuI complexes.[13]
The mixed trigonal/tetrahedral configuration of C2 seems
to answer several advantages of the catalysts studied. The
open coordination site on one of the copper centers may
explain the strongly ligand-accelerated catalysis. The three
ligands may account for the observed negative nonlinear
effect,[3] the bridging anions may explain the strong sensitivity
to the copper salt used,[2] and the aggregation level may
illustrate the influence of the solvent.[2] Furthermore, p–
p stacking within the {L2Cu} moiety together with nitrogen–
zinc interactions may orient the amine moieties towards the
free coordination site, thus explaining the decisive role of its
configuration and the steric demand of the amine fragment.
By combining information from NMR spectroscopy, mass
spectrometry, and elemental analysis, we were able to propose
for the first time the structure of a precatalytic copper
complex with phosphoramidite ligands in solution, which
represents a new structural motif for catalytically active
copper complexes. The identification of a mixed trigonal/
tetrahedral precatalyst not only provides an explanation for
the known synthetic optimization procedures and the reduction of the ligand/copper ratio from 2:1 to 1.6:1, but also
offers the basis for designing new improved catalysts.
Received: May 12, 2006
Published online: August 23, 2006
Keywords: asymmetric catalysis · copper · NMR spectroscopy ·
phosphorous · structure elucidation
[1] N. Krause, Modern Organocopper Chemistry, Wiley-VCH,
Weinheim, 2002.
[2] A. Alexakis, C. Benhaim, Eur. J. Org. Chem. 2002, 3221.
[3] L. A. Arnold, R. Imbos, A. Mandoli, A. H. M. de Vries, R.
Naasz, B. L. Feringa, Tetrahedron 2000, 56, 2865.
[4] E. Nakamura, S. Mori, Angew. Chem. 2000, 112, 3902; Angew.
Chem. Int. Ed. 2000, 39, 3750.
[5] N. Krause, A. Gerold, Angew. Chem. 1997, 109, 194; Angew.
Chem. Int. Ed. Engl. 1997, 36, 186.
[6] a) M. John, C. Auel, C. Behrens, M. Marsch, K. Harms, F.
Bosold, R. M. Gschwind, P. R. Rajamohanan, G. Boche, Chem.
Eur. J. 2000, 6, 3060; b) R. M. Gschwind, X. Xie, P. R.
Rajamohanan, C. Auel, G. Boche, J. Am. Chem. Soc. 2001,
123, 7299; c) X. Xie, C. Auel, W. Henze, R. M. Gschwind, J. Am.
Chem. Soc. 2003, 125, 1595; d) W. Henze, A. Vyater, N. Krause,
R. M. Gschwind, J. Am. Chem. Soc. 2005, 127, 17 335.
[7] B. L. Feringa, Acc. Chem. Res. 2000, 33, 346.
[8] A. H. M. de Vries, A. Meetsma, B. L. Feringa, Angew. Chem.
1996, 108, 2526; Angew. Chem. Int. Ed. Engl. 1996, 35, 2374.
[9] W. Shi, L. Wang, Y. Fu, S. Zhu, Q. Zhou, Tetrahedron:
Asymmetry 2003, 14, 3867.
[10] A. Alexakis, C. Benhaim, S. Rosset, M. Humam, J. Am. Chem.
Soc. 2002, 124, 5262.
[11] A. Alexakis, S. Rosset, J. Allamand, S. March, F. Guillen, C.
Benhaim, Synlett 2001, 1375.
[12] The ratio stated in Figure 2 is identical with the actual ratio in
solution. This was verified on selected samples by elemental
analyses of the filtered solutions.
[13] G. Wilkinson, R. Gillard, J. McCleverty, Comprehensive Coordination Chemistry, Vol. 5, Pergamon, New York, 1987.
[14] a) J. T. Gill, J. J. Mayerle, P. S. Welcker, D. F. Lewis, D. A. Ucko,
D. J. Barton, D. Stowens, S. J. Lippard, Inorg. Chem. 1976, 15,
1155; b) J. C. Dyason, L. M. Engelhardt, C. Pakawatchai, P. C.
Healy, A. H. White, Aust. J. Chem. 1985, 38, 1243.
[15] Juan Frutos, Ph.D. thesis, UniversitM Pierre et Marie Curie, Paris,
[16] B. L. Feringa, R. Naasz, R. Imbos, L. A. Arnold in Modern
Organocopper Chemistry (Ed.: N. Krause), Wiley-VCH, Weinheim, 2002, p. 234.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 6391 –6394
Без категории
Размер файла
178 Кб
solutions, structure, precatalytic, identification, complexes, coppel, phosphoramidate
Пожаловаться на содержимое документа