Frustrated Lewis Pair Chemistry with Magnesium Lewis Acids

Article abstract of DOI:10.1002/ejic.201700787

Anilido imine–magnesium bromide complexes, (L^HMgBr)₂ and (L^MeMgBr)₂, have been prepared [L^H = DIPP-N(C₆H₄)C(H)N-DIPP and L^Me = DIPP-N(C₆H₄)C(Me)N-DIPP (DIPP = 2,6-iPr₂C₆H₃)]. Their dimeric crystal structures are compared to the dimer (DIPP-nacnacMgBr)₂ [DIPP-nacnac = DIPP-NC(Me)C(H)C(Me)N-DIPP]. The anilido imine ligands show some degree of charge delocalization but are clearly more asymmetric than the DIPP-nacnac complex: one of the N atoms has considerable anilido character, whereas the other one is close to an imine. Calculated natural population analysis (NPA) charges (B3PW91/6-311++G**) and an atoms-in-molecules (AIM) study on model systems quantify the asymmetry in charge distribution. Quantification of the Lewis acidity of L^HMgBr and L^MeMgBr by the Gutmann–Beckett test gave acceptor numbers of 58.9 and 58.3, respectively. The complexes themselves are inert to a number of electrophiles; however, in combination with PPh₃ the substrate 1-butene oxide was ring-opened. The resulting betaine-type alkoxide complex is unstable and decomposes at room temperature through a Wittig-type reaction. At higher temperatures also a reverse Wittig-type decomposition is observed.

Full text of DOI:10.1002/ejic.201700787

A Journal of
Accepted Article
Title: Frustrated Lewis Pair chemistry with magnesium Lewis acids
Authors: Sjoerd Harder, Steffen Brand, Jürgen Pahl, and Holger Elsen
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201700787
Link to VoR: http://dx.doi.org/10.1002/ejic.201700787

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
Frustrated Lewis Pair chemistry with magnesium Lewis acids
Steffen Brand, Jürgen Pahl, Holger Elsen and Sjoerd Harder*
Inorganic and Organometallic Chemistry, University Erlangen-Nürnberg, Egerlandstrasse 1,
9
1058 Erlangen, Germany
E-mail corresponding author: sjoerd.harder@fau.de
Homepage: www.harder-research.com
Keywords: Alkaline-earth metals – Magnesium – Frustrated Lewis Pair – Ring opening
Abstract
H
Me
Anilido imine magnesium bromide complexes, (L MgBr) and (L MgBr) , have been prepared:
2
2
H
Me
L = DIPP-N(C H )C(H)N-DIPP and L = DIPP-N(C H )C(Me)N-DIPP (DIPP = 2,6-iPr-phenyl). Their
6
4
6 4
dimeric crystal structures are compared to the dimer (DIPP-nacnacMgBr)2 (DIPP-nacnac =
DIPP-NC(Me)C(H)C(Me)N-DIPP). The anilido imine ligands show some degree of charge delocalization
but are clearly more asymmetric than the DIPP-nacnac complex: one of the N atoms has considerable
anilido character whereas the other is close to an imine. Calculated NPA charges (B3PW91/6-
3
11++G**) and an Atoms-in-Molecules study on model systems quantify the asymmetry in charge
H
Me
distribution. Quantification of the Lewis acidity of L MgBr and L MgBr by the Gutmann-Beckett test
gave acceptor numbers of 58.9 and 58.3, respectively. The complexes itself are inert to a number of
electrophiles, however, in combination with PPh the substrate 1-butene oxide was ring opened. The
3
resulting betaine type alkoxide complex is unstable and decomposes at room temperature via a
Wittig type reaction. At higher temperatures also a reverse Wittig type decomposition is observed.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
2
Introduction
[
1]
Since their discovery in 2006 by Stephan and coworkers, frustrated Lewis Pairs (FLPs) have emerged
as a highly prominent field in main group chemistry. The unquenched combination of Lewis
acid/base pairs enables molecule activations that were previously only possible with transition
[
1-5]
[6]
[7]
metals.
Also catalytic transformations, like hydrosilylation of CO , alkyne hydroamination and
2
[
2], [8-10]
hydrogenation,
could be performed with FLP catalysts.
[
11]
[12]
[13]
Although the majority of systems is limited to P/B or N/B pairs, the use of Al, C, or Si based
Lewis acids emerged. Especially FLPs with Al Lewis acids have been able to add unique metal-based
[
14]
reactivities to the growing repertoire: e.g. the complexation of alkali-metal hydrides,
[
15]
[16]
[17,18]
stoichiometrical and catalytic dehydrocoupling of amine-boranes, activation of alkenes,
[
16,19]
[19]
[20]
[18]
[20-22]
alkynes,
carbodiimides, isocyanates, hydrogen, CO and other carbonyl groups.
Our
2
expertise in alkaline-earth metal chemistry stimulated us to expand the scope of Lewis acids with
[
23]
group 2 metal Lewis acids. The low-cost, biocompatible and abundant metal magnesium
is
2+
especially attractive for FLP applications. Although the Mg ion shows significant Lewis acidity it is
3+
3+
clearly less Lewis acidic than Al (or B ). Vice versa, the Mg-X bonds of intermediates from FLP
activation should be much more reactive than comparable Al-X or B-X bonds. Consequently, the
advantage of using group 2 metal cations in FLP type chemistry may lie in the much richer chemistry
of molecules activated by such an FLP.
Langer et al. reported very recently a Mg complex with an intramolecular coordinating phosphine
[
24]
group that acts as a masked FLP (Scheme 1). Due to a hard-soft mismatch, the P-Mg bond is weak
and can be easily cleaved by THF coordination. This triggers the Lewis acid assisted nucleophilic ring
opening of THF. Herein, we describe a continuation of our studies on FLP reactivity with Lewis acidic
Mg complexes. In contrast to the complex in Scheme 1, we aim to focus on intermolecular P/Mg FLP
reactivity. Our goal is to synthesize Lewis acidic complexes of the type L-Mg-X (in which L is a non-
reactive spectator ligand and X is a heavier halogenide) and test its FLP reactivity with Ph P, a simple
3
air-stable phosphine that recently was found to be active as a FLP in combination with a cationic Al
[
11]
complex.
Finding suitable ligands (L) for the target complex L-Mg-X is a challenge. The ligand should act as a
spectator ligand and itself not react with the small molecule substrates to be tested in FLP reactivity.
For example, Mg amide or amidinate complexes are not suitable for FLP activation of CO , due to
2
[
25],[26]
insertion reactions of carbon dioxide into the Mg-N bond giving formation of carbamates.
Since
the commonly used -diketiminate ligands (nacnac’s) show strong bidentate metal coordination and
are easily tunable, they seem attractive candidates. Unfortunately, these systems have a major
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
3
drawback: the methine carbon in the ligand backbone shows nucleophilic behavior towards
substrates like CO , a reactivity that can be easily understood on account of its resonance structures
2
[
27-29]
(Scheme 2).
[
24]
Scheme 1. Ring opening of THF with [Mg(dppmflu) ] by Langer et al.
2
Scheme 2. Comparison of the mesomeric structures of anionic nacnac and anilido imine ligands.
A promising alternative are the anilido imine systems, which were introduced by Hayes et al. and
[
30]
have been used previously in group 2 metal chemistry. In contrast to the nacnac anion, the
resonance structures B and C of these systems are highly unfavored, thus strongly reducing the
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
4
[
31]
nucleophilicity of the backbone carbon. However, similar to the nacnac ligand, the steric and
electronic properties of these systems can be easily tuned by variation of substituents. We herein
report magnesium anilido imine bromide complexes that in combination with PPh act as a FLP.
3
Results and discussion
Two anilido imine ligands with bulky DIPP substituents (DIPP = 2,6-iPr -C H ) have been synthesized.
2
6 3
H
Me
The ligands differ by their backbone substituents which can be either H (L H) or Me (L H); see
[
30]
Scheme 3. These ligands have been prepared by modifying the original procedures by Hayes et al.
[
31]
and Reynolds et al., respectively. Reaction with MeMgBr in Et O gave access to the respective
2
anilido imine magnesium bromide complexes which after treating under high vacuum could be
H
Me
isolated as ether-free products L MgBr (1) and L MgBr (2) in the form of bright-yellow fluorescent
powders. Crystals suitable for X-ray diffraction could be obtained by cooling a concentrated solution
of these complexes in benzene to 6 °C.
H
Scheme 3. Synthesis of complexes L MgBr (1)
Me
and L MgBr (2).
Figure 1. Crystal structures anilido imine
magnesium bromide complexes (H atoms omitted
for clarity). Selected bond lengths (Å) and angles (°)
H
are given. (a) (L MgBr) (1): Mg-N1 1.9954(18), Mg-
2
N2 2.0508(18), N1-Mg-N2 92.42(7), C1-C6-C7-N2
Me
8
.4(3). (b) (L MgBr) (2): Mg-N1 1.9929(16), Mg-
2
N2 2.0468(16), N1-Mg-N2 91.02(7), C1-C6-C7-N2
5.1(3).
1
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
5
Their crystal structures reveal very similar, centrosymmetric, dimeric complexes with bridging Br
ligands (Figure 1) that are comparable to the dimeric complex: [(DIPP-nacnac)MgBr] ; DIPP-nacnac =
2
[
32]
HC[C(Me)N-DIPP]2. Scheme 4 shows a comparison of bond lengths and angles in the crystal
H
H
Me
structures of the ligand L -H, (L MgBr) (1), (L MgBr) (2) and [(DIPP-nacnac)MgBr] . This led us to
2
2
2
conclude the following:
H
(
i) Deprotonation of L -H results in slight lengthening of the imine C=N bond (1.278(4) → 1.299(3) Å)
which signifies some delocalization of the electron density over the N-C-C-C-N backbone. This is
supported by shortening of the C-N(anilido) bond (1.379(4) → 1.360(3) Å) and elongation of the CC
bond in the aromatic ring (1.391(5) → 1.433(3) Å).
(ii) Delocalization of negative charge, however, is much less pronounced as in DIPP-nacnac which
H
shows a mirror symmetric N-C-C-C-N backbone with C-N bonds (1.332(3) Å) between those in the L
anion.
H
(
iii) This imbalance in charge distribution in the L anion, i.e. the observed anilide/imine character,
results in different Mg-N bond lengths. The Mg-N(anilide) is considerably shorter than the
Mg-N(imine) bond whereas Mg-N for the fully delocalized DIPP-nacnac ligand falls nicely in between.
(iv) The DIPP-nacnac ligand features a significantly wider bite angle than the anilido imine ligand. This
is likely due to higher flexibility in the DIPP-nacnac backbone (the anilido imine backbone is part of an
aromatic ring).
(v) Further differences in Mg coordination are only marginal. The Mg-Br bond lengths and Br-Mg-Br’
angles are essentially equal.
Scheme 4. A comparison of bond lengths (Å)
and angles (°) in the crystal structures of the
H
H
ligand L H (black), (L MgBr) (1; red),
2
Me
(
(
L MgBr) (2; green) and [(DIPP-nacnac)MgBr]
2
2
blue).
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
6
The conclusions from the structural comparison of nacnac and anilido imine ligands is in good
agreement with theoretical calculations on model magnesium complexes containing these ligands
and a simple hydride (Scheme 5; see Supporting Information for methods and details). The charges
on the N atoms in the nacnac ligand (-0.754) are in between those on the anilido N (-0.797) and the
imine N (-0.676). The high electron density on the methine C(H) in the nacnac backbone (-0.415) is
nearly absent in the anilido imine ligand (-0.076). Although anilido imine ligands show a considerable
asymmetry in their charge distribution, this has no further consequences for the charges on Mg and
H, which are for both complexes essentially equal. The boxed numbers in Scheme 5 are the electron
densities in the bond-critical-points (BCP’s) from an Atoms-in-Molecules (AIM) analysis. These are a
measure for bond strength: while being higher for multiple bonds, low values are typical for ionic
bonds. This analysis clearly shows the expected asymmetry in anilido imine Mg-N coordination. The
values for the Mg-N bond in the nacnac complex lie in between those for Mg-N(anilido) and
Mg(imine) bonds. Also noteworthy is the value for the C-N bonds in the nacnac ligand which lies in
between those for the C-N(anilido) and C-N(imine) bonds. The higher electron density in the C-
N(imine) bond signifies a higher double bond character.
Scheme 5. Calculated NPA
charges for the model
complexes
substituents at the B3PW91/6-
11++G** level. The boxed
numbers represent the
with
N-Ph
3
electron densities in the bond-
critical-points ( ) from AIM
BCP
analysis.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
7
In order to compare the reactivities of anionic nacnac and anilido imine ligands towards
electrophiles, solutions of their magnesium bromide complexes in C D were reacted with 1.5 bar of
6
6
1
1
CO at room temperature and were monitored by H NMR. The H NMR spectrum of [(DIPP-
2
nacnac)MgBr] with CO revealed decomposition of the complex and the formation of several
2
2
different species. As earlier described by Chisholm et al., this is likely due to nucleophilic attack of the
[
27]
methine CH group on the CO carbon atom. In contrast, pressurizing a C D solution of the
2
6 6
H
magnesium anilido imine complex (L MgBr) (1) with CO gave even after heating to 60 °C for several
2
2
1
days no changes in the H NMR spectrum. This observation is at first sight surprising. Although the
backbone carbon is protected by integration in an aromatic ring, charge localization on the anilido N
Me
(
N1 in Figure 1a) would favor nucleophilic attack. Indeed, in the case of (L MgBr) (2), slow insertion
2
Me
of CO into the Mg-N1 bond and formation of the carbamato complex (L CO MgBr) (3) was
2
2
2
observed; see Figure 2. This reaction, however, does not proceed at room temperature but needs
heating to 60 °C over several days. The slightly higher nucleophilicity of complex 2 versus complex 1
may be explained by the -donor effect of the Me group in 2. These experiments show that, despite
Me
slow CO insertion in the Mg-N(anilido) bond of (L MgBr) , anilido imine ligands are in general
2
2
substantially more stable towards electrophilic attack than nacnac ligands which makes them ideal
candidates as Lewis acids in FLP chemistry.
Me
Figure 2. Crystal structure of the carbamato complex (L CO MgBr) (3); Hydrogen atoms and iPr
2
2
substituents are omitted for clarity. Selected bond lengths (Å) and angles (°): Mg-N2 2.161(2), Mg-O1
1
1
.9220(17), Mg-O2’ 1.9598(17), Mg-Br 2.4227(7), O1-C33 1.259(3), C33-N1 1.380(3), C33-O2
.270(3), N1-C1 1.440(3), C1-C6 1.409(3), C6-C7 1.508(3), C7-N2 1.289(3), O1-Mg-N2 90.55(8), O1-
Mg-Br 122.89(6), O1-Mg-O2’ 108.52(8), O2’-Mg-Br 104.28(5), O2’-Mg-N2 123.57(8), N2-Mg-Br
08.30(6), C1-C6-C7-N2 89.1(3), C1-C6-C7-C32 -99.4(3).
1
H
Me
The anilido imine complexes (L MgBr) (1) and (L MgBr) (2) were screened as Lewis acids in FLP
2
2
reactivity. For the Lewis base we chose the very simple air-stable Ph P which in combination with
3
[
11]
Lewis-acidic cationic Al complexes displayed FLP reactivity. Addition of Ph P to solutions of 1 and 2
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
8
1
31
in C D did not give changes in the H NMR spectra and the P NMR spectrum shows a signal for free
6
6
H
Me
Ph P. Various electrophiles were added to a solution of (L MgBr) or (L MgBr) and Ph P in C D at
3
2
2
3
6 6
1
31
room temperature and samples were investigated by H and P NMR spectroscopy (Scheme 6). In
case no reaction was observed, the samples were heated to 60°C overnight.
H
Me
Scheme 6. Screening of [L MgBr] /PPh and [L MgBr] /PPh for FLP reactivity.
2
3
2
3
Me
Except for the already mentioned insertion of CO into the Mg-N1 bond of (L MgBr) , no reactivity
2
2
could be observed for the substrates CO , Ph-N=C=O, PhC≡CH and Me NHBH , even after the sample
2
2
3
was heated to 60°C overnight. Addition of PhC(H)=NMe, Me C=O and iPrN=C=NiPr to the
2
R
1
(
L MgBr) /Ph P mixtures led to a shift and broadening of signals in the H NMR spectrum, but no
2
3
31
change in the P resonance could be observed. From this we deduce Mg-substrate coordination but
even after heating to 60 °C no further nucleophilic addition of Ph P took place.
3
H
Figure 3. Crystal structure of L MgBr∙OPEt ; Hydrogen atoms are omitted for clarity. Selected bond
3
lengths (Å) and angles (°): Mg-Br 2.4647(7), Mg-O 1.9267(15), Mg-N1 2.0201(18), Mg-N2 2.0718(18),
N1-C1 1.362(3), C1-C6 1.438(3), C6-C7 1.423(3), C7-N2 1.300(3), P-O 1.5132(14), N1-Mg-N2 92.04(7),
O-Mg-Br 108.04(5), O-Mg-N1 110.78(7), O-Mg-N2 114.48(7).
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
9
The lack of FLP reactivity is likely due to a combination of low nucleophilicity (PPh ) and low Lewis
3
acidity. We determined the Lewis acidity of the magnesium complexes by the Gutmann-Beckett
31
method in which the pertubation of the P chemical shift of Et PO upon coordination to an acceptor
3
[
34,35]
molecule is measured.
Addition of the Lewis base Et PO to 1 and 2 led to Mg-O bond formation
3
and dissociation of dimeric into monomeric magnesium bromide complexes. Crystals of one of these
adducts could be obtained by storing the solution at room temperature over several weeks. The
H
crystal structure of L MgBr∙OPEt is shown in Figure 3. Addition of one equivalent of Et PO to a C D
3
3
6 6
H
Me
31
solution of (L MgBr) (1) or (L MgBr) (2) resulted in a shift of the P NMR signal from 45.5 ppm
2
2
(
free Et PO) to 67.6 ppm (1∙OPEt ) and 67.3 ppm (2∙OPEt ). This corresponds to acceptor numbers of
3
3
3
5
8.9 (1) and 58.3 (2), respectively. The Lewis acidity of the magnesium bromide complexes 1 and 2 is
therefore substantially lower than that of B(C F ) , which in C D has an average acceptor number of
6
5 3
6 6
[
36]
7
8.3. It is even slightly lower than that of the unfluorinated borane BPh which has an average
3
[
36]
acceptor number of circa 63.4.
R
Despite the low Lewis acidity/basicity of the (L MgBr) /PPh Lewis pair, in one of the cases FLP type
2
3
R
reactivity could be observed. Addition of 1-butene oxide to (L MgBr) /Ph P mixtures in C D resulted
2
3
6 6
31
at room temperature within seconds in a color change from bright yellow to orange. The P NMR
H
spectrum showed two new peaks at 22.8 and 23.3 ppm (for (L MgBr) /Ph P) or at 22.5 and
2
3
Me
1
2
4
2.2 ppm (for (L MgBr) /Ph P). Similar, the H NMR spectra gave two new sets of signals in a ratio of
2
3
6:54 (4) and 47:53 (5) (see Supporting Information). We attribute these new signal sets to the
formation of the ring opening product that may occur as two different diastereomers with chiral
centers at Mg and the alkoxide ligand (Scheme 7). The magnesium alkoxide complexes 4 and 5 are
not stable at room temperature and have to be stored at −30°C.
Crystals of complex 5 could be obtained by vapor diffusion of pentane into a toluene solution at
−30°C over a period of four days. Crystal structure determination confirmed formation of the
proposed product (Figure 4). The R,R- and S,S-enantiomers crystallized in a monoclinic unit cell with
the centrosymmetric space group P2 /c. Complex 5 crystallized as a monomeric species with
1
chelating anilido imine and bromide ligands and a neutral phosphonium-alkoxide ligand. Formation
of this complex is explained by phosphine → epoxide attack at the least sterically hindered carbon
atom giving a secondary alkoxide species. Due to the small bite angle of the anilido imine ligand
(N1-Mg-N2 89.29(6)°), the magnesium center has a distorted tetrahedral conformation. The P-C and
C-C bond lengths in the opened epoxide moiety, C33-C34 1.539(3) Å and C34-P 1.8070(18) Å, are
[
37]
typical for C-C (1.542 Å) and P-C (1.793 Å) single bonds. The C-O bond of 1.372(2) Å, however, is
[
37]
significantly shorter than the value for a C-O single bond (1.413 Å) but this C-O contraction is
[
38]
characteristic for polar metal alkoxide complexes. Interestingly, the observed structure reveals a
short P∙∙∙O distance (3.006(1) Å) which is substantially shorter than the sum of both van der Waals
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
0
[
39]
[40]
radii (3.32 Å)
but much larger than a P-O single bond (1.78(1) Å).
In addition, the
phosphonium-alkoxide moiety has a synclinal conformation (O-C33-C34-P 45.71(2)°) instead of an
antiperiplanar O-C-C-P conformation which would be expected on the basis of steric interactions.
Both, short P-O distances and a gauche O-C-C-P conformation have also been observed in the cation
+
[
tBu(Me)PH-CH -C(CF ) -OH ] and are explained by attractive Coulomb forces between the negatively
2
3 2
-
+
[41]
charged O and the positively charged P atoms. This could be seen as a first step to the
formation of a oxaphosphetane ring and is closely related to the facile decomposition of this
complex.
R
Scheme 7. Ring opening of 1-butene oxide by (L MgBr) /PPh at r.t. in C H and decomposition
2
3
6 6
pathways at different temperatures.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
1
Figure 4. Crystal structure of complex 5 (only the S,S-enantiomer is shown). Hydrogen atoms and iPr
groups are omitted for clarity. Selected bond lengths (Å) and angles (°): P-O 3.006(1), Mg-O
1
1
.8874(13), Mg-N1 2.0512(15), Mg-N2 2.0512(15), O-C33 1.372(2), C33-C34 1.539(3), C34-P
.8070(18), O-Mg-Br 110.27(4), N1-Mg-N2 89.29(6), O-Mg-N1 122.39(6), N2-Mg-Br 100.33(4), C37-P-
C49 105.88(8), C37-P-C43 112.99(9), C37-P-34 111.23(9), C43-P-C49 106.43(9), C43-P-C34 111.74(9),
C34-P-C49 108.16(8), C33-C34-P 113.56(12), O1-C33-C34 111.31(15), O-C33-C34-P 45 45.71(2).
The phosphonium alkoxide complexes 4 and 5 are not stable at room temperature and decompose
within five days. Representative for the decomposition of both complexes, the decomposition of 5
31
has been investigated in detail (see Supporting Information for details). The P NMR spectrum of the
decomposition products revealed the formation of only one new phosphorus-containing species with
31
a chemical shift of 38.4 ppm. In contrast to the P NMR spectrum, two new species are observed in
1
H NMR spectrum. One of the species shows a set of signals at 5.0 and 5.8 ppm with a typical pattern
for a terminal C=C double bond. After degassing the sample, these signals could not be observed
1
13
anymore. Analysis of this species by H TOCSY and C NMR revealed the formation of 1-butene
[42]
during the reaction. This may be explained by a Wittig-type reaction mechanism. The preorganized
short P∙∙∙O contact leads to formation of a four-membered oxaphosphetane ring (Scheme 7). Such
[
43]
species have been isolated already since 1968 by the chemical company Monsanto.
Following this, the oxaphosphetane intermediate decomposes by a retro-[2+2]-cycloaddition to give
H
1
-butene and Ph PO. Like in complex (L MgBr) ∙OPEt (Figure 3), Ph PO is likely coordinated to the
3
2
3
3
magnesium center. This would generate a chiral center at Mg and nicely explains the observation of
four septets and eight doublets for the CH and Me groups of the iPr substituents, respectively. The
proposed Wittig-type reaction mechanism is supported by an additional experiment. Addition of one
Me
1
equivalent of Ph PO to a solution of (L MgBr) (2) resulted in an identical H NMR spectrum and a
3
2
31
shift of the P NMR signal of Ph PO from 25.3 to 38.4 ppm.
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
2
At room temperature, decomposition of the phosphonium alkoxide complex 5 takes several days but
heating to 60°C accelerated this process and full conversion was observed already after 6 hours.
Interestingly, decomposition at the higher temperature gave an additional side product. Beside the
1
31
H NMR signals for 1-butene and Ph PO, a new doublet at 2.71 ppm was observed. Also the P NMR
3
spectrum revealed formation of a new species with a resonance at 21.8 ppm. Increasing the
decomposition temperature apparently led to lower selectivity and the formation of a side product in
circa 20% yield. Cooling the reaction mixture gave single crystals that could be identified as
+
[
Ph PMe ][Brˉ] (see Supporting Information).
3
A possible explanation for the generation of this phosphonium salt is a mechanism in which, as
previously described, an oxaphosphetane is formed. However, in contrast to the aforementioned ring
opening to give 1-butene and Ph PO, the oxaphosphetane intermediate could be opened
3
alternatively to form propionaldehyde and the phosphonium ylide Ph PCH (Scheme 7). The ylide
3
2
could be basic enough to deprotonate the aldehyde in α-position, to form a magnesium enolate and
+
[
Ph PMe ][Brˉ]. Reverse Wittig reactions, which involves breakage of the P-O bond in the
3
[
44, 45]
intermediate P-O-C-C ring, are rare but known.
+
Although we have definite proof for the formation of [Ph PMe ][Brˉ], hitherto the proposed
3
magnesium enolate species could not be isolated. However, the following experiment supports this
mechanism. Addition of the Wittig reagent Ph PCH to a solution of acetaldehyde in C D , expectedly
3
2
6 6
led to formation of Ph PO and propene. However, repeating the same experiment in the presence of
3
H
+
a stoichiometric amount of (L MgBr) (1), resulted in the formation of [Ph PMe ][Brˉ] as the sole P-
2
3
containing species. Although we were not able to isolate and purify the Mg-enolate product, the
latter experiment nicely shows that the presence of a Lewis acidic magnesium complex can change
the course of the classical Wittig reaction. Formation of a Mg∙∙∙O=C(H)CH CH coordinative bond
2
3
reduces the nucleophilic properties of the ketone O and increases the acidity of the -CH group. This
2
+
prevents formation of Ph PO and triggers deprotonation to give [Ph PMe ].
3
3
Conclusions
The anilido imine ligands in the magnesium complexes (L MgBr) (1) and (L MgBr) (2) are excellent
H
Me
2
2
non-nucleophilic alternatives to nacnac ligands. Whereas (DIPP-nacnac)MgBr reacts with CO at the
2
backbone CH group, complex 1 is even at 60 °C inert to CO and complex 2 only slowly reacts with
2
CO by insertion in the Mg-N anilido bond.
2
The inertness of these complexes towards electrophiles makes them suitable candidates for Lewis
acids in FLP chemistry. In combination with the weak Lewis base Ph P, no FLP reactivity was observed
3
with CO , Ph-N=C=O, PhC≡CH and Me NHBH . Addition of PhC(H)=NMe, Me C=O or iPrN=C=NiPr to a
2
2
3
2
R
(
L MgBr) /PPh mixture led to Mg-substrate coordination but the magnesium complex was not Lewis
2
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
3
acidic enough to activate the substrate for further nucleophilic attack by PPh . Indeed, the
3
Guttmann-Becket method estimated Lewis acidities that are far inferior to that of B(C F ) and even
6
5 3
slightly lower than that of BPh3.
Despite the low Lewis acidity/basicity of the pair (L MgBr) /Ph P, addition of 1-butene oxide gave
R
2
3
typical FLP type reactivity. Ring opening resulted in a phosphonium alkoxide species in which the
P∙∙∙O distance (3.006(1) Å) is substantially shorter than the sum of both van der Waals radii (3.32 Å).
This interaction can be seen as a first step to the formation of an oxaphosphetane and leads to low
complex stability (the temperature should be kept below -30 °C). At room temperature, facile
decomposition to Ph PO and 1-butene by a Wittig type reaction mechanism is observed. At higher
3
+
temperatures [Ph PMe ][Brˉ] is formed as an additional side product (presumably together with an
3
Mg-enolate). Formation of this phosphonium bromide salt can be explained by a reverse Wittig type
mechanism. Indeed, addition of a Mg Lewis acid changes the course of the classical Wittig reaction
(
e.g. Ph PCH + O=C(H)Me → Ph PO + H C=C(H)Me) to a mechanism in which the Wittig reagent acts
3
2
3
2
+
as a base: Ph PCH + *Mgꢀ∙∙∙O=C(H)Me → Ph PCH + *Mgꢀ∙∙∙‾O-CH=CH .
3
2
3
3
2
These findings represent a rare example of a magnesium/phosphorous FLP and demonstrate the
possibilities of this research area. It is highly likely that the recently reported conversion of
[
46]
epoxide/CO mixtures to cyclic carbonates by a MgBr /PPh catalyst also proceeds through an FLP
2
2
3
mechanism in which Mg activates the epoxide and PPh , instead of Br‾ as proposed by the authors, is
3
responsible for nucleophilic ring opening. The introduction of metal centers like Mg in FLP chemistry
2+
3+
3+
seems attractive. The rather modest Lewis acidity of Mg , at least when compared to B or Al , also
means that the product after FLP activation is much more reactive thus opening possibilities for
further transformations or even catalytic application. The latter is currently under investigation.
Experimental section
All experiments were carried out under inert atmosphere using either standard Schlenk techniques
or a glove box (MBraun, Labmaster SP). THF (THF AnalaR Normapur, VWR) was dried over molecular
sieves (3Å) and distilled from sodium. All other solvents were degassed with nitrogen, dried over
activated aluminum oxide (Innovative Technology, Pure Solv 400-4-MD, Solvent Purification System)
and stored over molecular sieves (3Å) under inert atmosphere prior to use. Starting materials were
used as delivered unless noted otherwise. Following chemicals were dried according to standard
[
47]
procedures and stored over molecular sieves (3 Å): phenyl isocyanate, phenylacetylene, N-methyl-
-phenylmethanimine, and diisopropylcarbodiimide. 2,6-Diisopropylaniline was dried over CaH2,
1
distilled and stored over molecular sieves (3 Å) prior to usage. 1-Butene oxide was dried over
molecular sieves (3 Å), distilled and stored under inert atmosphere prior to usage. Acetone was dried
H
[30]
Me [31]
over molecular sieves (3 Å) and stored under inert atmosphere prior to use. L H and L H were
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
4
synthesized according to modified routes of Hayes et al. and Reynolds et al. (using an excess of the
lithium anilido salts gives better yields). The precursors for this synthesis, N-(2,6-diisopropylphenyl)-
1
-(2-fluorophenyl)-methanimine and N-(2,6-diisopropylphenyl)-1-(2-fluorophenyl)ethan-1-imine,
[
30,31]
were prepared according to the original procedures.
NMR data were recorded with a Bruker
Avance DPX 300 NB, a Bruker Avance DRX400 WB, a Bruker Avance III HD 400 MHz or a Bruker
Avance III HD 600 MHz spectrometer. Deuterated solvents were dried over molecular sieves (3 Å).
The chemical shifts are expressed in parts per million (ppm) and referenced to the residual signals of
the respective solvent. Crystal structures were measured on a SuperNova diffractometer with dual
Cu and Mo microsources and an AtlasS2 detector (Rigaku-Oxford Diffraction).
Synthesis of lithium (2,6-diisopropylphenyl)amide. A 2.5 M solution of n-BuLi (26 ml, 65 mmol) in
n-hexane was added dropwise via a syringe to a solution of 2,6-diisopropylaniline (12 ml, 64 mmol) in
THF (20 ml) at −78°C. The resulting orange solution was allowed to warm to room temperature and
stirred for 4 h. After removal of all volatiles in vacuo the off-white solid was washed two times with
pentane (30 ml, 10 ml). Drying in vacuo afforded the clean white crystals of the lithium salt with one
coordinating THF molecule. Yield: 14.04 g (54.99 mmol, 86%).
1
3
3
H NMR (600 MHz, C D ): δ/ppm 7.10 (d, J = 8 Hz, 2H, ArH), 6.67 (t, J = 8 Hz, 1H, ArH), 3.44-3.23
6
6
HH
HH
3
(
m, 2H, CHMe ), 2.99-2.93 (m, 4H, THF), 2.83 (s, 1H, NH), 1.41 (d, J = 7 Hz, 12H, CHMe ), 1.06-0.99
2
HH
2
13
1
(
m, 4H, THF). C{ H} NMR (151 MHz, C D ): δ/ppm 156.4 (NC-Ar), 132.0 (C-Ar), 123.2 (CH-Ar), 112.2
6
6
(
CH-Ar), 68.2 (THF), 28.5 (CHMe ), 25.15 (THF), 23.9 (CHMe ). Anal. Calcd. for C H LiNO: C, 75.27; H,
2
2
16 26
1
0.26; N, 5.49. Found: C, 75.20; H, 10.11; N, 5.61.
H
[30]
Synthesis of L H. The ligand was synthesized by a slightly modified literature procedure. Lithium
(2,6-diisopropylphenyl)amide∙(THF) (2.76 g, 10.8 mmol) and N-(2,6-diisopropylphenyl)-1-(2-
fluorophenyl)-methanimine (1.50 g, 5.29 mmol) were dissolved in 20 ml THF. The resulting red-
orange solution was stirred for 3 d at room temperature. The reaction mixture was then quenched
with 10 ml H O. The product was extracted four times with 10 ml n-hexane. The combined organic
2
phases were dried over MgSO and all volatiles were removed in vacuo. The resulting yellow-brown
4
oil was dissolved in circa 200 ml of hot MeOH and then slowly cooled to 6°C to give 1.67 g of pale
yellow crystals. Reducing the volume of the mother liquor by half and cooling it to 6°C yielded in a
second batch of crystals (0.32 g). The crystals were dried in vacuo. Total yield: 1.99 g (4.52 mmol
1
[30]
8
5%). H NMR data are in agreement with those in the literature.
1
H NMR (300 MHz, C D ): δ/ppm 10.87 (s, 1H, NH), 8.22 (s, 1H, CH=NAr), 7.32-7.12 (m, 6H, ArH), 7.08
6
6
3
4
3
(
dd, J = 8 Hz, J = 1 Hz, 1H, ArH), 7.0-6.91 (m, 1H, ArH), 6.61-6.51 (m, 1H, ArH), 6.40 (d, J =
HH
HH
HH
3
3
3
8
Hz, 1H, ArH), 3.40 (sept, J = 7 Hz, 2H, CHMe ), 3.27 (sept, J = 7 Hz, 2H, CHMe ), 1.20 (d, J = 7
HH
2
HH
2
HH
3
3
Hz, 12H, CHMe ), 1.15 (d, J = 7 Hz, 6H, CHMe ), 1.12 (d, J = 7 Hz, 6H, CHMe ).
2
HH
2
HH
2
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
5
Me
[31]
Synthesis of L H. The ligand was synthesized by a slightly modified literature procedure. Lithium
(2,6-diisopropylphenyl)amide∙(THF) (4.43 g, 17.4 mmol) and N-(2,6-diisopropylphenyl)-1-(2-
fluorophenyl)ethan-1-imine (2.44 g, 8.22 mmol) were dissolved in 60 ml THF resulting in a nearly
black solution which was heated to reflux for 40 h. It was allowed to cool to room temperature and
quenched with 50 ml H O. The product was extracted four times with 10 ml n-hexane. The combined
2
organic phases were dried over MgSO and all volatiles were removed in vacuo. The resulting dark
4
brown oil was dissolved in ~600 ml hot MeOH and cooled slowly to 6°C to yield 2.20 g pale yellow
crystals, which were filtered, washed with cold MeOH and dried in vacuo. Reducing the volume of
the mother liquor by half and cooling it to −20°C yielded another batch of 0.68 g crystals. Total yield:
1
[31]
2
.88 g (6.33 mmol, 77%). H NMR data are in agreement with those in the literature.
1
3
4
H NMR (300 MHz, C D ): δ/ppm 11.57 (s, 1H, NH), 7.49 (dd, J = 8 Hz, J = 1 Hz,1H, ArH), 7.32-
6
6
HH
HH
7
.09 (m, 6H, ArH), 7.05-6.95 (m, 1H, ArH), 6.63-6.54 (m, 1H, ArH), 6.53-6.43 (m, 1H, ArH), 3.45 (sept,
3
3
J = 7 Hz, 2H, CHMe ), 3.06 (sept, J = 7 Hz, 2H, CHMe ), 1.97 (s, 3H, MeC=NAr), 1.22-1.07 (m, 24H,
HH
2
HH
2
CHMe2).
H
Synthesis of (L MgBr) (1). A 3M solution of MeMgBr (34.8 µL, 1.04 mmol) in Et O was added to a
2
2
H
pale yellow solution of L H (0.51 g, 1.16 mmol) in n-hexane (4 ml). The green-yellow-fluorescent
emulsion was heated to 60°C for 1 h, resulting in gas evolution and the formation of a yellow
precipitate. The solution was removed by cannula filtration and the bright yellow solid was washed
two times with 1 ml n-hexane and dried in vacuo. Yield: 0.41 g (0.38 mmol, 73%);
1
3
H NMR (300 MHz, C D ): δ/ppm 7.84 (s, 2H, CH=NAr), 7.29-7.11 (m, 8H, ArH), 7.02 (d, J = 8 Hz, 4H,
6
6
HH
3
ArH), 6.83-6.69 (m, 4H, ArH), 6.30-6.17 (m, 4H, ArH), 3.31 (sept, J = 7 Hz, 4H, CHMe ), 3.09 (sept,
HH
2
3
13
1
J = 7 Hz, 4H, CHMe ), 1.10-0.95 (m, 48H, CHMe ). C{ H} NMR (75 MHz, C D ): δ/ppm 173.9
HH
2
2
6 6
(
CH=NAr), 160.7 (CH=NC), 146.5 (NC-Ar), 144.7 (C-Ar), 143.4 (NC-Ar), 141.5 (C-Ar), 138.6 (NC-Ar),
1
1
34.6 (CH-Ar), 127.1 (CH-Ar), 126.0 (CH-Ar), 124.8 (CH-Ar), 124.4 (CH-Ar), 118.8 (CH-Ar), 115.1 (C-Ar),
13.4 (CH-Ar), 29.1 (CHMe ), 28.4 (CHMe ), 25.9 (CHMe ), 25.7 (CHMe ), 24.8 (CHMe ), 23.2 (CHMe ).
2
2
2
2
2
2
Anal. Calcd. for C H Br Mg N : C, 71.45; H, 7.29; N, 4.50. Found: C, 71.04; H, 7.52; N, 4.40.
74
90
2
2
4
Me
Synthesis of (L MgBr) (2). A 3M solution of MeMgBr (45.6 µL, 1.37 mmol) in Et O was added to a
2
2
Me
yellow solution of L H (0.62 g, 1.37 mmol) in n-hexane (11 ml). The yellow emulsion was heated to
5°C for 6 h, resulting in gas evolution and the formation of a yellow precipitate. The mixture was
6
centrifuged and the solution was removed via a syringe. The bright yellow solid was washed two
times with 1 ml toluene and dried under vacuum. Yield: 0.62 g (0.56 mmol, 82%);
1
3
H NMR (300 MHz, C D ): δ/ppm 7.33-7.08 (m, 10H, ArH), 6.99 (d, J = 8 Hz, 4H, ArH), 6.80-6.70 (m,
6
6
HH
3
3
2
H, ArH), 6.34-6.19 (m, 4H, ArH), 3.32 (sept, J = 7 Hz, 4H, CHMe ), 2.89 (sept, J = 7 Hz, 4H,
HH
2
HH
3
CHMe ), 1.76 (s, 6H, CMe=NAr), 1.14 (d, J = 7 Hz, 12H, CHMe ), 1.04-0.96 (m, 24H, CHMe ), 0.93 (d,
2
HH
2
2
3
13
1
J = 7 Hz, 12H, CHMe ). C{ H} NMR (101 MHz, C D ): δ/ppm 178.6 (CMe=NAr), 159.8 (CMe=NC),
HH
2
6 6
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
6
1
1
2
44.9 (C-Ar), 144.3 (NC-Ar), 143.3 (NC-Ar), 141.0 (C-Ar), 133.8 (CH-Ar), 132.9 (CH-Ar), 127.7 (CH-Ar),
27.3 (CH-Ar), 125.8 (CH-Ar), 124.9 (CH-Ar), 120.1 (CH-Ar), 116.2 (C-Ar), 112.6 (CH-Ar), 28.9 (CHMe2),
8.4 (CHMe ), 26.0 (CHMe ), 25.2 (CHMe ), 24.9 (CHMe ), 24.8 (CHMe ), 23.7 (CMe=NAr). Anal. Calcd.
2
2
2
2
2
for C H Br Mg N : C, 71.76; H, 7.45; N, 4.40. Found: C, 71.55; H, 7.56; N, 4.99.
76
94
2
2
4
Me
Me
Synthesis of (L CO MgBr) (3). A solution of (L MgBr) (2) (111.4 mg, 0.0998 mmol) in 4 ml
2
2
2
benzene was pressurized with 0.5 bar CO and thoroughly mixed. The mixture was then heated to
2
Me
8
0°C without stirring for 4 days. The resulting pale yellow crystals of (L CO MgBr) (3) were suitable
2
2
for single crystals X-ray diffraction and contain two cocrystallized benzene molecules per dimer. The
rest of the batch was washed thoroughly with benzene (3x1 ml) and dried under vacuum. Yield:
8
8.7 mg (0.07 mmol, 70%).
1
H NMR (400 MHz, C D ): δ/ppm 7.37-6.86 (m, ArH, benzene, solvent), 6.81 (d, J = 8.1 Hz, 2H, ArH),
6
6
3
.53 (sept, J = 6.5 Hz, 2H, CHMe ), 3.16 (sept, J = 6.6 Hz, 2H, CHMe ), 3.01 (sept, J = 6.6 Hz, 2H,
2
2
CHMe ), 2.60 (sept, J = 6.4 Hz, 2H, CHMe ), 1.87 (s, 6H, CMe=NAr), 1.40 (d, J = 6.6 Hz, 6H, CHMe ),
2
2
2
1
6
.32-1.19 (m, 18H, CHMe ), 1.10 (d, J = 6.8 Hz, 6H, CHMe ), 0.93 (d, J = 6.7 Hz, 6H, CHMe ), 0.42 (d, J =
2
2
2
1
3
1
.7 Hz, 6H, CHMe ), 0.31 (d, J = 6.5 Hz, 6H, CHMe ). C{ H} NMR (101 MHz, C D Br): δ/ppm 186.2
2
2
6 5
(
CMe=NAr), 162.2 (C-Ar), 147.5 (NCO ), 144.7 (C-Ar), 142.6 (C-Ar), 140.8 (C-Ar), 140.2 (C-Ar), 139.6 (C-
2
Ar), 137.0 (C-Ar), 135.7 (C-Ar), 130.9 (CH-Ar), 130.1 (CH-Ar, superimposed by solvent signal), 129.0
CH-Ar), 128.7 (benzene), 128.4 (CH-Ar), 126.5 (CH-Ar, superimposed by solvent signal), 126.0 (CH-
Ar), 126.0 (CH-Ar), 125.3 (CH-Ar), 124.9 (CH-Ar), 124.8 (CH-Ar), 29.5 (CHMe ), 29.2 (CHMe ), 28.5
(
2
2
(
CHMe ), 28.2 (CMe=NAr), 27.7 (CHMe ), 26.6 (CH ), 25.3 (CH ), 25.1 (CH ), 24.9 (CH ), 24.5 (CH ),
2
2
3
3
3
3
3
2
6
4.1 (2x CH ), 23.9 (CH ). Anal. Calcd. for C H Br Mg N O : C, 68.88; H, 6.97; N, 4.12. Found: C,
3
3
78 94
2
2
4
4
9.14; H, 6.96; N, 4.01.
H
Reaction of (L MgBr) /PPh with 1-butene oxide to give 4. 1-Butene oxide (8.4 mg, 0.12 mmol) was
2
3
H
added to a suspension of (L MgBr) (1) (62.4 mg, 0.0574 mmol) and PPh (31.1 mg, 0.119 mmol) in
2
3
toluene (500 μl). Vapor diffusion with pentane within four days resulted in the formation of a yellow,
oily phase, which solidified. The supernatant was removed and the yellow solid was washed with
1
00 μl pentane and briefly dried. Longer drying procedures result in decomposition and the product
was obtained with 0.29 molecules pentane and 0.11 molecules toluene. Yield: 67.5 mg (0.08 mmol,
7
0%). D1= Diastereomer 1, D2= Diastereomer 2. Ratio D1:D2 = 46:54
1
H NMR (300 MHz, C D ): δ/ppm 8.18 (s, 1H, HC=NAr (D1)), 8.12 (s, 1H, HC=NAr (D2)), 7.46-6.75 (m,
6
6
3
3
ArH (D1 and D2), toluene, solvent signal), 6.57 (d, J = 9 Hz, 1H, ArH (D2)), 6.49 (d, J = 9 Hz, 1H,
HH
HH
3
ArH (D1)), 6.34-6.27 (m, 2H, ArH (D1), ArH (D2)), 4.41 (sept, J = 7 Hz, 1H, CHMe (D2)), 4.25-
HH
2
3
.89 (m, 9H, 3x CHMe (D1), 3x CHMe (D2), 2x CHO(CH ) (D1 and D2), PCHH(CH) (D2)), 3.79 (sept,
2
2
2 2
3
J = 7 Hz, 1H, CHMe (D1)), 3.22-3.08 (m, 1H, PCHH(CH) (D1)), 2.80-2.67 (m, 1H, PCHH (CH) (D1)),
HH
2
3
2
.67-2.53 (m, 1H, PCHH(CH) (D2)), 2.11 (s, toluene), 1.75 (d, J = 7 Hz, 3H, CHMe (D2)), 1.69 (d,
HH
2
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
7
3
3
J = 7 Hz, 3H, CHMe (D1)), 1.59 (d, J = 7 Hz, 3H, CHMe (D2)), 1.57-1.52 (m, 6H, CHMe (D1),
HH
2
HH
2
2
CHMe (D2)), 1.44-0.82 (m, 6x CHMe (D1), 5x CHMe (D2)), (CH)CHH(CH ) (D2), (CH)CH (CH ) (D1),
2
2
2
3
2
3
3
3
(
(
(
CH)CHH (CH ) (D2), pentane), 0.56 (t, J = 7 Hz, 3H, CH CH (D1)), 0.47 (t, J = 7 Hz, 3H, CH CH
3
HH
2
3
HH
2
3
31
1
13
1
D2)). P{ H} NMR (122 MHz, C D ): δ/ppm 23.3, 22.8. C{ H} NMR (151 MHz, C D ): δ/ppm 172.1
6
6
6 6
HC=NAr), 172.1 (HC=NAr), 161.0 (HC=NC), 161.0 (HC=NC), 150.1 (2x NC-Ar), 148.0 (NC-Ar),
48.0 (NC-Ar), 146.3 (C-Ar), 146.3 (C-Ar), 145.0 (C-Ar), 145.0 (C-Ar), 143.7 (C-Ar), 143.7 (C-Ar),
1
1
1
1
1
1
3
3
42.6 (C-Ar), 141.9 (C-Ar), 138.9 (CH-Ar), 138.7 (CH-Ar), 134.2 (d, J = 10 Hz, CH-Ar), 134.0 (d, J =
PC
PC
4
4
0 Hz, CH-Ar), 133.8 (d, J = 3 Hz, CH-Ar), 133.7 (d, J = 3 Hz, CH-Ar), 133.3 (CH), 133.2 (CH),
PC
PC
2
2
32.9 (CH), 132.9 (CH), 129.7 (d, J = 13 Hz, CH-Ar), 129.7 (d, J = 12 Hz, CH-Ar), 128.4 (CH),
PC
PC
25.8 (CH), 125.6 (CH), 125.1 (CH), 124.8 (CH), 124.6 (CH), 124.3 (CH), 124.2 (CH), 124.2 (CH),
1
1
24.2 (CH), 123.7 (CH), 123.1 (CH), 123.0 (CH), 120.8 (d, J = 85 Hz, PC-Ar), 120.8 (d, J = 85 Hz,
PC
PC
2
PC-Ar), 119.2 (CH-Ar), 119.2 (CH-Ar), 116.1 (C-Ar), 116.1 (C-Ar), 111.2 (CH-Ar), 71.2 (d, J = 21 Hz,
PC
2
3
3
OCH(CH )), 71.2 (d, J = 22 Hz, OCH(CH )), 36.4 (d, J = 12 Hz, CH CH ), 36.2 (d, J = 11 Hz,
2
PC
2
PC
2
3
PC
1
1
CH CH ), 35.3 (d, J = 42 Hz, PCH (CHO)), 34.5 (pentane), 33.9 (d, J = 39 Hz, PCH (CHO)), 28.6 (CH),
2
3
PC
2
PC
2
2
8.6 (CH), 28.5 (CH), 28.4 (CH), 28.4 (CH), 28.3 (CH), 28.3 (CH), 28.3 (CH), 26.3 (CH ), 26.2 (CH ), 26.2
3
3
(
CH ), 26.1 (CH ), 26.0 (3x CH ), 25.7 (CH ), 25.6 (CH ), 25.6 (CH ), 25.3 (CH ), 24.9 (CH ), 24.9 (CH ),
3
3
3
3
3
3
3
3
3
2
4.6 (CH ), 23.4 (CH ), 23.2 (CH ), 22.7 (pentane), 14.3 (pentane), 11.1 (CH CH ), 10.8 (CH CH ).
3
3
3
2
3
2
3
Toluene was not observed due to its low concentration. Anal. Calcd. for C55.2H66.4Br Mg N O P : C,
1
1
2
1 1
7
2.94; H, 7.36; N, 3.08. Found: C, 73.68; H, 7.16; N, 2.70.
Me
Reaction of (L MgBr) /PPh with 1-butene oxide to give 5. 1-Butene oxide (6.5 mg, 0.090 mmol)
2
3
Me
was added to a suspension of (L MgBr) (2) (50.0 mg, 0.0448 mmol) and PPh (23.5 mg,
2
3
0
.0896 mmol) in toluene (500 μl). A mixture of diastereomers could be crystallized with a co-
crystallized equivalent toluene at −30°C from solution by vapor diffusion with pentane within four
days. Yield: 55.1 mg (0.06 mmol, 67%). D1 = Diastereomer 1, D2 = Diastereomer 2. Ratio D1:D2 =
4
7:53
1
H NMR (600 MHz, C D ): δ/ppm 7.55-7.50 (m, 2H, ArH (D1 and D2)), 7.46-7.43 (m, 1H, ArH (D2)),
6
6
7
7
.45-7.43 (m, 1H, ArH (D2)), 7.41-7.38 (m, 1H, ArH (D2)), 7.34-7.17 (m, 21H, ArH (D1 and D2)), 7.15-
.11 (m, 2H, toluene), 7.07-7.00 (m, 8H, ArH (D1 and D2), toluene), 6.97-6.88 (m, 14H, ArH (D1 and
D2)), 6.69-6.66 (m, 1H, ArH (D2)), 6.61-6.57 (m, 1H, ArH (D1)), 6.38-6.34 (m, 2H, ArH (D1 and D2)),
3
3
4
.47 (sept, J = 7 Hz, 1H, CHMe (D2)), 4.29 (sept, J = 7 Hz, 1H, CHMe (D1)), 4.07-3.82 (m, 6H, 2x
HH
2
HH
2
CHMe (D1), 2x CHMe (D2), 2x CHO(CH ) (D1 and D2)), 3.79-3.70 (m, 2H, 2x CHMe (D1 and D2)),
2
2
2 2
2
3
.66-3.58 (m, 1H PCHH(CH) (D2)), 3.21-3.10 (m, 1H, PCHH(CH) (D1)), 2.87-2.77 (m, 1H, PCHH (CH)
D1)), 2.74-2.64 (m, 1H, PCHH(CH) (D2)), 2.12 (s, 3H, CMe=NAr (D1)), 2.11 (s, 3H, toluene), 2.07 (s, 3H,
CMe=NAr (D2)), 1.74-1.70 (m, 6H, 2x CHMe (D1)), 1.57-1.51 (m, 9H, 2x CHMe (D1), CHMe (D2)),
(
2
2
2
1
.48-1.44 (m, 6H, 2x CHMe (D1)), 1.42-1.37 (m, 9H, CHMe (D1), 2x CHMe (D2)), 1.32-1.15 (m, 27H ,
2
2
2
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
8
2
1
0
x CHMe (D2), CHMe (D1), 3x CHMe (D2), 2x (CH)CHH(CH ) (D1 and D2), pentane), 0.95-0.90 (m,
2
2
2
3
3
H, (CH)CHH(CH ) (D1), 0.87 (t, J = 7 Hz, 6H, pentane) 0.72-0.64 (m, 1H, (CH)CHH(CH ) (D2), 0.44-
3
HH
3
31
1
13
1
.38 (m, 6H, CH CH (D1 and D2)). P{ H} NMR (162 MHz, C D ): δ/ppm 22.5, 22.2. C{ H} NMR (151
2
3
6 6
MHz, C D ): δ/ppm 175.1 (MeC=NAr), 175.0 (MeC=NAr), 160.3 (MeC=NC), 160.2 (MeC=NC),
6
6
1
1
1
49.1 (NC-Ar), 149.0 (NC-Ar), 146.8 (NC-Ar), 146.8 (NC-Ar), 146.1 (2x C-Ar), 145.3 (C-Ar), 145.0 (C-Ar),
43.1 (C-Ar), 143.0 (C-Ar), 141.2 (C-Ar), 141.0 (C-Ar), 137.9 (toluene), 134.0-133.8 (m), 131.6 (CH-Ar),
2
2
31.6 (CH-Ar), 129.9 (d, J = 12 Hz, CH-Ar), 129.9 (d, J = 12 Hz, CH-Ar), 129.3 (toluene), 128.7
PC
PC
(toluene), 128.4 (CH-Ar), 125.7 (toluene), 125.6 (CH-Ar), 125.5 (CH-Ar), 124.7 (CH-Ar), 124.7 (CH-Ar),
1
1
1
1
3
2
2
2
1
7
24.5 (CH-Ar), 124.3 (CH-Ar), 124.1 (CH-Ar), 123.9 (CH-Ar), 123.7 (CH-Ar), 123.6 (CH-Ar),
1
1
23.6 (CH-Ar), 120.8 (d, J = 85 Hz, PC-Ar), 120.7 (CH-Ar), 120.7 (d, J = 85 Hz, PC-Ar), 120.4 (C-Ar),
PC
PC
2
2
18.4 (C-Ar), 118.3 (C-Ar), 110.7 (CH-Ar), 110.6 (CH-Ar), 71.3 (d, J = 13 Hz, OCH(CH )), 71.1 (d, J =
PC
2
PC
1
1
3 Hz, OCH(CH )), 36.3 (d, J = 62 Hz, PCH (CHO)), 36.3 (d, J = 61 Hz, PCH (CHO)), 34.5 (pentane),
2
PC
2
PC
2
3
3
4.5 (d, J = 4 Hz, CH CH ), 34.3 (d, J = 4 Hz, CH CH ), 29.1 (CH), 28.7 (CH), 28.5 (CH), 28.4 (CH),
PC
2
3
PC
2
3
8.4 (CH), 28.2 (CH), 28.1 (CH), 28.1 (CH), 27.3 (CH ), 26.9 (CH ), 26.9 (CH ), 26.8 (CH ), 26.2 (CH ),
3
3
3
3
3
6.1 (CH ), 26.0 (CH ), 25.9 (CH ), 25.1 (CH ), 25.0 (CH ), 25.0 (CH ), 24.9 (CH ), 24.9 (2xCH ),
3
3
3
3
3
3
3
3
4.8 (CH ), 24.2 (CH ), 23.9 (CH ), 23.8 (CH ), 22.7 (pentane), 21.4 (toluene), 14.3 (pentane),
3
3
3
3
1.0 (CH CH ), 10.8 (CH CH ). Anal. Calcd. for C H BrMgN OP: C, 75.10; H, 7.93; N, 2.65. Found: C,
2
3
2
3
66 83
2
5.04; H, 7.28; N, 2.79.
[
34,35]
Lewis acidity according to Gutmann-Becket.
A portion of 19 μmol of the respective complex
H
Me
(
(L MgBr) or (L MgBr) ) was dissolved in 600 μL C D in a NMR tube. Et PO (24.0 μL of a 0.75 M
2
2
6
6
3
solution in C D , 18 μL) was then added and NMR spectra were recorded after 10 min.
6
6
[
48]
Crystal structure determinations. Using Olex2,
crystal structures have been solved by Direct
[
49]
[50]
Methods (ShelXT) and were refined with ShelXL using Least Squares minimisation. The hydrogen
atoms have been placed on calculated positions and were refined isotropically in a riding model.
Crystals of complex 3 contain cocrystalized benzene molecules. Two were refined with large
anisotropic displacement factors and one benzene molecule is disordered over 3 positions with one
part on a special position (30:30:40). Crystal data and experimental methods can be found in the
Supporting Information. Crystallographic data (excluding structure factors) have been deposited with
the Cambridge Crystallographic Data Centre as supplementary publication CCDC 1559452 (1),
+
H
1
559453 (2), 1559454 (3), 1559455 (5), 1559456 ([Ph PMe ][Brˉ]) and 1559457 (L MgBr∙OPEt ).
3
3
Copies of the data can be obtained free of charge on application to CCDC, 12 Union Road, Cambridge
CB21EZ, UK (fax: (+44)1223-336-033; E-mail: deposit@ccdc.cam.ac.uk).
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
1
9
References
[
[
[
[
[
[
[
[
[
1]
2]
3]
4]
5]
6]
7]
8]
9]
G. C. Welch; R. R. S. Juan, J. D. Masuda, D. W. Stephan, Science 2006, 314, 1124-1126.
D. W. Stephan, Dalton Trans. 2009, 3129-3136.
D. W. Stephan, G. Erker, Angew. Chem., Int. Ed. 2015, 54, 6400-6441.
D. W. Stephan, J. Am. Chem. Soc. 2015, 137, 10018-10032.
D. W. Stephan, Org. Biomol. Chem. 2008, 6, 1535-1539.
A. Berkefeld, W. E. Piers, M. Parvez, J. Am. Chem. Soc. 2010, 132, 10660-10661.
T. Mahdi, D. W. Stephan, Angew. Chem., Int. Ed. 2013, 52, 12418-12421.
D. W. Stephan, G. Erker, Angew. Chem., Int. Ed. 2010, 49, 46-76.
D. W. Stephan, S. Greenberg, T. W. Graham, P. Chase, J. J. Hastie, S. J. Geier, J. M. Farrell, C.
C. Brown, Z. M. Heiden, G. C. Welch, M. Ullrich, Inorg. Chem. 2011, 50, 12338-12348.
G. Erős, H. Mehdi, I. Pápai, T. A. Rokob, P. Király, G. Tárkányi, T. Soós, Angew. Chem., Int. Ed.
[
[
10]
11]
2
010, 49, 6559-6563.
T. E. Stennett, J. Pahl, H. S. Zijlstra, F. W. Seidel, S. Harder, Organometallics 2016, 35, 207-
17.
2
[
[
[
12]
13]
14]
B. Inés, S. Holle, R. Goddard, M. Alcarazo, Angew. Chem., Int. Ed. 2010, 49, 8389–8391.
M. Reißmann, A. Schäfer, S. Jung, T. Müller, Organometallics 2013, 32, 6736-6744.
C. Appelt, J. C. Slootweg, K. Lammertsma, W. Uhl, Angew. Chem., Int. Ed. 2012, 51, 5911-
5
914.
[
[
15]
16]
G. Mꢁnard, D. W. Stephan, J. Am. Chem. Soc. 2010, 132, 1796-1797.
C. Appelt, H. Westenberg, F. Bertini, A. W. Ehlers, J. C. Slootweg, K. Lammertsma, W. Uhl,
Angew. Chem., Int. Ed. 2011, 50, 3925-3928.
[17]
G. Ménard, L. Tran, J. S. J. McCahill, A. J. Lough, D. W. Stephan, Organometallics 2013, 32,
6
759-6763.
[
[
18]
19]
G. Ménard, D. W. Stephan, Angew. Chem., Int. Ed. 2012, 51, 4409-4412.
T. Holtrichter-Rößmann, C. Rösener, J. Hellmann, W. Uhl, E.-U. Würthwein, R. Fröhlich, B.
Wibbeling, Organometallics 2012, 31, 3272-3283.
[
20]
F. Bertini, F. Hoffmann, C. Appelt, W. Uhl, A. W. Ehlers, J. C. Slootweg, K. Lammertsma,
Organometallics 2013, 32, 6764-6769.
[
[
21]
22]
J. Boudreau, M.-A. Courtemanche, F.-G. Fontaine, Chem. Commun. 2011, 47, 11131-11133.
G. Ménard, T. M. Gilbert, J. A. Hatnean, A. Kraft, I. Krossing, D. W. Stephan, Organometallics
2
013, 32, 4416-4422.
[
[
[
23]
24]
25]
P.-A. Huhn, M. Dornheim, T. Klassen, R. J. Bormann, J. Alloys Compd. 2005, 404-406, 499-502.
J. Langer, I. Kosygin, R. Puchta, J. Pahl, S. Harder, Chem. - Eur. J. 2016, 22, 17425-17435.
K.-C. Yang, C.-C. Chang, C.-S. Yeh, G.-H. Lee, S.-M. Peng, Organometallics 2001, 20, 126-137.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
2
0
[
[
[
[
26]
27]
28]
29]
Y. Tang, L. N. Zakharov, A. L. Rheingold, R. A. Kemp, Organometallics 2004, 23, 4788-4791.
M. H. Chisholm, K. Choojun, J. C. Gallucci, P. M. Wambua, Chem. Sci. 2012, 3, 3445-3457.
C. Camp, J. Arnold, Dalton Trans. 2016, 45, 14462-14498.
M. D. Anker, M. Arrowsmith, P. Bellham, M. S. Hill, G. Kociok-Köhn, D. J. Liptrot, M. F. Mahon,
C. Weetman, Chem. Sci. 2014, 5, 2826-2830.
[
30]
(a) P. G. Hayes, G. C. Welch, D. J. H. Emslie, C. L. Noack, W. E. Piers, M. Parvez,
Organometallics 2003, 22, 1577-1579; (b) P. Bellham, M. D. Anker, M. S. Hill, G. Kociok-Köhn,
M. F. Mahon, Dalton Trans. 2016, 13969-13978; (c) C. Bellini, V. Dorcet, J.-F. Carpentier, S.
Tobisch, Y. Sarazin, Chem. - Eur. J. 2016, 22, 4564-4583.
[
[
31]
32]
A. M. Reynolds, B. F. Gherman, C. J. Cramer, W. B. Tolman, Inorg. Chem. 2005, 44, 6989-6997.
A. P. Dove, V. C. Gibson, P. Hormnirun, E. L. Marshall, J. A. Segal, A. J. P. White, D. J. Williams,
Dalton Trans. 2003, 3088-3097.
[
[
[
33]
34]
35]
D. Zhang, Q. Yue, J. Wang, G. Shigeng, L. Weng, Inorg. Chem. Commun. 2009, 12, 1193-1196.
V. Gutmann, Coord. Chem. Rev. 1976, 18, 225-255.
M. A. Beckett, D. S. Brassington, M. E. Light, M. B. Hursthouse, J. Chem. Soc., Dalton Trans.
2
001, 1768-1772.
[
[
36]
37]
I. B. Sivaev, V. I. Bregadze, Coord. Chem. Rev. 2014, 270-271, 75-88.
F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen, R. Taylor, J. Chem. Soc.,
Perkin Trans. 2 1987, S1-S19.
[
[
[
[
38]
39]
40]
41]
K. B. Wiberg, J. Am. Chem. Soc. 1990, 112, 3379-3385.
A. Bondi, J. Phys. Chem. 1964, 68, 441-451.
M. Appel, S. Blaurock, S. Berger, Eur. J. Org. Chem. 2002, 1143-1148.
A. S. Ionkin, W. J. Marshall, B. M. Fish, M. F. Schiffhauer, F. Davidson, J. Am. Chem. Soc. 2007,
1
29, 9210-9215.
[
[
[
[
[
[
42]
43]
44]
45]
46]
47]
G. Wittig, G. Geissler, Justus Liebigs Ann. Chem. 1953, 580, 44-57.
G. Chioccola, J. J. Daly, J. Chem. Soc. A 1968, 568-577.
E. Ciganek, J. Org. Chem. 1970, 35, 1725-1729.
E. Vedejs, P. L. Fuchs, J. Am. Chem. Soc. 1973, 95, 822-825.
Y. Ren, J.-J. Shim, ChemCatChem 2013, 5, 1344-1349.
th
W. L. F. Armarego, C. L. L. Chai, Purification of Laboratory Chemicals, 7 ed. 2013, Kidlington,
Elsevier
[
48]
O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H. Puschmann, J. Appl. Cryst.
2
009, 42, 339-341.
[
[
49]
50]
G. M. Sheldrick, Acta Cryst. 2015, A71, 3-8.
G. M. Sheldrick, Acta Cryst. 2008, A64, 112-122.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201700787
2
1
Table of Contents
The FLP type reactivity of (Mg anilido imine)/PPh combinations has been investigated. Whereas the
3
majority of substrates show no reaction, 1-butene oxide gave ring opening. The resulting
phosphonium alkoxide complexes are unstable and show temperature dependent decomposition
pathways.
This article is protected by copyright. All rights reserved.