Bora-amidinate as a cooperative ligand in group 2 metal catalysis

Article abstract of DOI:10.1039/c7dt02136d

Syntheses and crystal structures of the monomeric bora-amidinate (bam) complexes ^DIPPNBN-Mg·(THF)₃ and ^DIPPNBN-Ca·(THF)₄ are presented; ^DIPPNBN = HB[N(2,6-iPr₂-C₆H₃)]<

Full text of DOI:10.1039/c7dt02136d

Dalton
Transactions
View Article Online
View Journal
PAPER
Bora-amidinate as a cooperative ligand in group
2
metal catalysis†
Cite this: DOI: 10.1039/c7dt02136d
a
a
a
a
Benjamin Freitag, Christian A. Fischer, Johanne Penafiel,
Gerd Ballmann,
a
a
b
a
Holger Elsen, Christian Färber, Dirk F. Piesik and Sjoerd Harder
*
DIPP
Syntheses and crystal structures of the monomeric bora-amidinate (bam) complexes
NBN-Mg·(THF)
3
DIPP
DIPP
1
and
NBN-Ca·(THF)
4
are presented;
NBN = HB[N(2,6-iPr
-C
2 6
H
3
)]
2
. The simplicity of their H NMR
8
spectra in THF-d suggest that their monomeric solid state structures are retained in solution.
DIPP
NBN-Mg·(THF)
3
in C
6
D
6
, however, is in equilibrium with a dimeric species. Calculations (B3PW91/6-
DIPP
Received 12th June 2017,
Accepted 17th July 2017
311++G**) reveal a very high localized negative charge (NPA: −1.103) on the N atoms in
NBN-Mg. The
strongly basic properties of the bam ligand are in agreement with catalytic activity of these complexes in
the intramolecular alkene hydroamination. A mechanism is proposed in which the bam ligand is non-
innocent and cooperative, playing an active role in substrate deprotonation and product protonation.
DOI: 10.1039/c7dt02136d
rsc.li/dalton
Although heteroleptic complexes of the type (bam)LnR are
Introduction
isoelectronic to alkene polymerization catalysts like Cp LnR
2
+
The first description of a metal complex bearing a bora-amidi- (or Cp
2
TiR ), attempts to use an yttrium complex of type 6 in
nate ligand (bam) is nearly 40 years old: the serendipitous ethylene polymerization were not met with great success. This
9
reaction of LiN(SiMe )BMe with SnCl produced, upon elimin- is due to the fact that the “open” coordination site initiates
3
2
2
ation of LiCl and BMe
3
,
the dimeric tin complex dimerization and the yttrium complex is in the solid-state as
1
{
MeB[N(SiMe )] Sn} (1, Scheme 1). Despite this long history, well as dissolved in benzene a tightly bound, inactive, dimer
3
2
the chemistry of bam complexes is still in a developing stage. (7). We are currently not aware of any major application of
The only review article on the bam ligand, which was written a bam complexes in catalysis, however, use of a combined bam–
2
decade ago, describes its complexes with transition metals am ligand as spectator ligand in polymerization catalysis has
1
0
and metals from the s- and p-blocks. More recently, its chem- been documented.
3
3
4
5
5
istry has been extended to the f-block (2, 3, 4, 5, 6 ) and the
As part of our endeavors to develop the rapidly growing
field of early main group metal catalysis, we recently reported
The N,N-chelating bora-amidinate ligands (bam) are isoelec- on the importance of the metal cation. The hydroamination
tronic to the ubiquitously used amidinate ligands (am). The of carbodiimides can be catalyzed by a range of group 1 and
bam ligand, however, is dianionic and this has two important 2 metal amides (LiN″, NaN″, CaN″ ; N″ = N(SiMe ) ) but also by
consequences. The high concentration of electron density in a the metal-free catalyst [Me N ][Ph N ]. In the intramolecular
single ligand leads to intriguing redox chemistry and several alkene hydroamination, however, substrate activation by coordi-
examples of complexes with the radical anion bam are nation to the Lewis-acidic metal cation is crucial. These investi-
6
11
heavier alkaline-earth metals (4, M = Sr).
1
2
7
13
2
3 2
+
− 11
4
2
14
•
−
12
8
known. Secondly, fewer ligands are required to charge- gations evolved to a group of novel inorganic/organic hybrid cata-
balance the positively charged metal center thus leaving ample lysts that consists of a neutral strong organic base (cf. the
space at the metal for coordination/activation chemistry.
Schwesinger base P4) and a metal salt like CaI (Scheme 2). The
2
poor solubility of the salt CaI suggested the use of Lewis-acidic
2
complexes of type 4 (M = Ca) in which the bam ligand would be
an inactive spectator ligand and the larger part of the coordi-
nation sphere would be available for substrate coordination.
However, during the course of our investigations, we found that,
when using (bam)Ca complexes, the organic P4 base is not
a
Inorganic and Organometallic Chemistry, Friedrich-Alexander-Universität
Erlangen-Nürnberg, Egerlandstraße 1, 91058 Erlangen, Germany.
E-mail: sjoerd.harder@fau.de
b
Inorganic Chemistry, Universität Duisburg-Essen, Universitätsstrasse 7,
4
5151 Essen, Germany
†
Electronic supplementary information (ESI) available: Crystallographic details,
15
needed. Unlike the monoanionic am spectator ligands, the
1
13
selected H, C and DOSY NMR spectra, details for the NMR characterization of
bam ligand apparently takes an active part in catalysis by depro-
tonating the amine. Herein, we describe the first studies on
group 2 metal catalysts with cooperative bam ligands.
DIPP
(
2 2 2 2 x
NBN-H)Mg-N(H)CH CMe CH CHvCH ·(THF) and details for theoretical
calculation. CCDC 1551537 and 1551538. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c7dt02136d
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 2 Intramolecular alkene hydroamination with the inorganic/
2
organic hybrid catalyst CaI /P4.
1
7
6
weak base (pK
deprotonation of
DMAT) Sr·(THF)
crystals of
the similar benzylic Ca base (DMAT) Ca·(THF)
a
(Me
3
Si)
2
NH: 25.8). As reported previously,
DIPP
NBN-H with the stronger benzylic base
2
(
2
2
(DMAT
NBN-Sr·(THF)
=
2-Me
2
N-α-Me
3
Si-benzyl) gave
DIPP
4
in 71% yield. Deprotonation with
was success-
NBN-Ca·(THF) in a yield of 54%.
4
1
8
2
2
DIPP
ful and gave crystalline
DIPP
The Mg complex
NBN-Mg·(THF)
3
was obtained in a yield of
7
6% using the strong base (nBu) Mg.
Crystallographic studies on
2
DIPP
NBN-Mg·(THF)
3
and
DIPP
NBN-Ca·(THF)
tures (Fig. 1). The crystal structure of
similar, but not isomorphous, to the earlier reported structure
4
reveal in both instances monomeric struc-
DIPP
NBN-Ca·(THF)4 is
DIPP
6
of
4
NBN-Sr·(THF) . Table 1 shows a comparison of selected
geometric parameters.
The average Mg–N (2.045(2) Å) and Ca–N (2.321(5) Å) bond
distances in the bora-amidinate complexes are significantly
shorter than those in homoleptic bis-formidinate
1
9
Scheme 1 Bora-amidinate (bam) complexes 1–7.
complexes: [HC(N-DIPP)
]
2
Mg·THF (2.143(2) Å)
and
2
2
0
[
HC(N-DIPP) Ca·THF (2.383(3) Å). This is likely due to the
2 2
]
higher negative charge on N in the bam ligand. Charge deloca-
lization over the N–B–N backbone is less extensive than that in
the N–C–N fragment (see below for a theoretical comparison).
Indeed, the N–Mg and N–Ca bond distances in the bam com-
Results and discussion
Synthesis and structures of bam complexes
The sterically protected bam ligand DIPP(H)N-BH-N(H)DIPP plexes compare closely to those in Si-bridged bis-amide com-
DIPP
(
DIPP = 2,6-iPr-phenyl), abbreviated in here as
2
NBN-H , is plexes in which the negative charges are localized on the
2
1
easily accessible via a one-pot-synthesis reacting two equiva- N atoms: Ph(H)Si(N-DIPP) Mg·(THF) N–Mg = 2.076(3) Å,
lents of 2,6-diisopropylaniline with one equivalent of Ph(H)Si(N-DIPP)
2
3
2
1
2
Ca·(THF)
3
N–Ca
=
2.312(3) Å,
2
Me Si
1
6
22
Me
Attempts to prepare alkaline-earth metal complexes by double
deprotonation with the readily available metal amides MN″
2
S·BH
3
while using catalytic amounts of (nBu)
2
Mg.
(N-DIPP)
2
Ca·(THF)
3
N–Ca = 2.301(3) Å.
The N–M–N′ bite angles decrease from 72.28(9)° (Mg) to
60.3(1)° (Sr). This is directly related to an increase of the M–N
2
(
M = Mg, Ca, Sr) failed. This is presumably due to an unfavor- bond distances from 2.045(2) Å (Mg) to 2.460(3) Å (Sr). These
−
able acid–base equilibrium: the (Me Si) N anion is a rather very acute bite angles leave a large part of the metal cation
3
2
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
DIPP
exposed. The space-filling model for the
NBN-Ca unit
(
Fig. 1c) shows a remarkable amount of space for further
coordination of neutral ligands. The ligand geometries are not
significantly different and therefore not influenced by the
nature of the group 2 metal: the B–N bonds in the three
complexes are similar and also the N–B–N′ bond angles show
only a slight increase from 116.2(2)° (Mg) to 119.5(3)° (Sr).
Although not isomorphic, the crystal structure of
DIPP
NBN-Ca·(THF)
NBN-Yb·(THF)
4
is
closely
similar
to
that
of
DIPP
4
4
whereas the structure of the Sr complex
4
2+
compares well to the similar Sm and Eu complexes. The Ca /
2
+
2+
2+
2+
Yb and Sr /Sm /Eu relationships, that originate from
2
3
similarities in their ionic radii, are well-documented.
DIPP
The group 2 metal bam complexes,
NBN-Mg·(THF)3,
DIPP
DIPP
NBN-Ca·(THF)
4
and
NBN-Sr·(THF)
4
1
, dissolve in aro-
matic solvents after slight heating. The H NMR spectra of
DIPP
NBN-Ca·(THF)4 in C D or THF-d8 show only one heptet
6
6
3
and one doublet signal for the iPr CH and CH fragments,
respectively. This indicates that the monomeric solid state
structure is likely retained in solution and that the DIPP sub-
stituents can rotate freely around the N-DIPP axis. Giving the
small size of the hydride substituent in the boron backbone,
this is not surprising. A similar rotational freedom is observed
DIPP
1
in
NBN-Sr·(THF)
4
(ESI†). The H NMR spectrum of
, however, is much more compli-
DIPP
NBN-Mg·(THF) in C D
3
6 6
cated and at least two different species can be recognized (see
ESI†). Two-dimensional DOSY measurements suggest a second
species with nearly twice the molecular weight as
DIPP
NBN-Mg·(THF) . We propose this species to be the dimer
NBN-Mg·(THF)] . Similar as found in lanthanide chem-
2
3
DIPP
[
istry (see 7), this dimer could be bound through Mg⋯Ar con-
tacts. Theoretical calculations (B3PW91/6-311++G**) show that
such dimers may exist and are formed through Mg⋯N contacts
with minor Mg⋯Ar interactions (see ESI, Fig. S24†).
DIPP
DIPP
Dimerization of
NBN-Mg to form (
NBN-Mg) is exother-
2
mic by 76.9 kcal mol− but since THF ligands were not con-
sidered in the calculations, the true value is likely significantly
lower. Increasing the temperature shifts the equilibrium to the
dimer. The counterintuitive increase of entropy upon dimeri-
zation can be understood by incorporating THF ligands: 2
1
DIPP
DIPP
NBN-Mg·(THF) ⇄ [
NBN-Mg·(THF)] + 4 THF.
2
3
DIPP
DIPP
1
DIPP
Fig. 1 Crystal structures of (a)
NBN-Mg·(THF)
3
, (b)
NBN-Ca·(THF)
4
The H NMR spectrum of
3 8
NBN-Mg·(THF) in THF-d is
DIPP
and (c) the space filling model for the
NBN-Ca fragment.
much simpler and comparable to that of its heavier congeners.
Table 1 Selected geometric parameter for bam alkaline-earth metal complexes; range and average values between < > (distances in Å and angles
in degrees)
DIPP
DIPP
DIPP
a
NBN-Mg·(THF)
3
NBN-Ca·(THF)
4
NBN-Sr·(THF)
4
M–N
M–O
B–N
N–M–N′
N–B–N′
Distance M to NBN-plane
Interplanar Angle Ar/NBN
2.026(2)–2.063(2) <2.045(2)>
2.077(2)–2.095(2) <2.084(2)>
1.417(4)–1.424(4) <1.421(4)>
72.28(9)
116.2(2)
1.7(3)
2.317(5)–2.325(5) <2.321(5)>
2.411(4)–2.471(4) <2.436(4)>
1.429(7)–1.436(8) <1.433(8)>
64.23(16)
119.0(6)
0.8(6)
2.435(3)–2.475(3) <2.455(3)>
2.536(3)–2.622(3) <2.568(3)>
1.427(6)–1.437(6) <1.433(6)>
60.0(1)–60.5(1) <60.3(1)>
118.5(2)–120.5(4) <119.5(3)>
0.1(4)–0.5(4) <0.3(4)>
66.6(3)–72.7(3) <69.7(3)>
59.9(6)–62.7(6) <61.3(6)>
59.0(2)–78.0(2) <66.7(2)>
a
From ref. 6.
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
2 2 2
It only shows one set of signals (one iPr heptet and doublet) CH -bridged di-amide anion [H C(N-DIPP) ]
2−, a dianion with
that may be assigned to a monomeric species with free localized negative charges. In addition, their Mg complexes
DIPP
rotation of the DIPP substituents.
have been optimized: HC(N-DIPP)
C(N-DIPP) -Mg (THF ligands were not considered for practi-
cal reasons). Scheme 3 contains selected geometric para-
2
-MgH,
NBN-Mg and
H
2
2
Theoretical consideration
2
4
Although the amidinate and bora-amidinate anions are iso- meters, charges calculated by Natural Population Analysis
25
electronic, there are, apart from their different total charges, (NPA) and an Atom-In-Molecules analysis (AIM).
also distinct discrepancies in charge delocalization and metal
The anions and complexes show a variety of geometries.
bonding properties. In order to compare the ligand–metal The DIPP substituents and the NCN framework in the formidi-
bonding and charge distribution in these complexes, Density nate anion are nearly oriented perpendicular. The interplanar
−
Functional Theory (DFT) calculations at the B3PW91/6- NCN/DIPP angle of 63.4° in [HC(N-DIPP)
11++G** level have been performed (see ESI† for details). The than that of 17.9° in the CH
]
is much larger
2
3
2
-bridged di-amide anion
2−. The near-planarity of the latter anion is due
] , in to charge-delocalization of the N p-type lone pair into the DIPP
structures of three different ligand anions have been opti- [H
2
C(N-DIPP)
2
]
−
mized (Scheme 3): the formidinate anion [HC(N-DIPP)
2
which the negative charge is delocalized over the N–C–N substituent which indeed is significantly negatively charged
DIPP
2−
framework, the bam anion [
NBN] , in which delocaliza- (−0.451). The charge analysis also shows that, despite the
−
tion of negative charge is potentially a possibility, and the different
C(N-DIPP)
nearly the same (−0.640 and −0.658, respectively). The bam
total
charges
on
[HC(N-DIPP)
]
and
2
2
−
[H
2
2
] , the negative charges on their N atoms are
DIPP
2−
anion [
NBN] shows a much higher charge of −0.878 on
the N atom. With an interplanar NBN/DIPP angle of 52.2°,
there is partial delocalization of charge into the DIPP substitu-
ent (−0.326). The high negative charge on N also arises from a
strong σ-donor effect of the HB-bridge which is much more
2
positively charged than the CH -bridge (+0.413 vs. +0.218).
The charge distributions in their Mg complexes show in all
cases highly positive charges of +1.343 to +1.460 on the metal
atom which signifies largely ionic bonding (circa 70% ionic
character). Complexation of the anionic ligands with the posi-
+
2+
tively charged MgH or Mg cations causes a strong localiz-
ation of negative charge on the N atoms. The bridging units
CH, BH or CH2 become more positively charged while less
electron density is delocalized over the DIPP substituents. The
interplanar NCN/DIPP angle in H
larger than that in the free anion [H C(N-DIPP) ]
2 2
C(N-DIPP) Mg is much
2
−
.
2
2
Interplanar angles in the amidinate and bam complexes do
not change much upon Mg coordination. Again, the highest
negative charge (−1.103) is observed on the N atoms in
DIPP
NBN-Mg. As the calculated NCN/DIPP interplanar angle of
DIPP
46.8° in
NBN-Mg is somewhat smaller than that of 69.7(3)°
DIPP
in the crystal structure of
NBN-Mg·(THF) , it is likely that
3
the concentration of negative charge on N in the THF complex
is even larger.
All complexes have been analyzed by the Atom-In-Molecules
program (AIM). Of particular interest are the electron densities
(
ρ) in the bond-critical-points (BCP’s) of the C–N and B–N
bonds in the N–C–N and N–B–N frameworks. These are indi-
cators for bond multiplicity and therefore indirectly also
related to bond delocalization. Using the same method of
theory (B3PW91/6-311++G**), the electron densities in the BCP
for C–C (ethane) and CvC (ethylene) are 0.240 and 0.343 au,
respectively. Table 2 summarizes the values for the ligands of
interest and, for comparison, also those for some model com-
pounds calculated at the same level of theory.
Scheme 3 Calculated
(B3PW91/6-311++G**)
formidinate
and bis-amide [H C
[HC
−
2−
(
N-DIPP)
N-DIPP)
2
] , bora-amidinate [HB(N-DIPP)
2
]
2
2
−
(
2
]
anions and their Mg complexes. NPA atomic charges or
The formidinate anion shows a ρC–N of 0.349 au which is
close to the value of 0.388 au for the CvN double bond in the
group charges (in black, e), geometric parameters (in blue; distances in
Å, angles in degrees) and ρ at the BCP (in red; au).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
2
9
Table 2 The electron density (ρ) and ellipticity (ε) in the bond critical
The last column in Table 2 shows the delocalization index
−
points (BCP) of C–N and B–N bonds in the anions [HC(N-DIPP)
2
] ,
which is a measure of the number of electrons that are shared
or exchanged between two atoms or basins. The delocalization
index can be used to estimate the formal bond order of chemi-
cal bonds in an alternative way. It is close to the formal bond
order for covalent bonds (CC or CN) but deviations are found
when atoms are involved with electronegativities considerably
different from that of C (e.g. C–O, C–Si, Si–Si). Within one bond
DIPP
2−
2−
[
NBN] and [H
2
C(N-DIPP)
2
]
and Mg complexes thereof. The last
column shows the delocalization index (DI). For comparison, also values
for simple benchmark compounds are given at the same level of theory
30
(
B3PW91/6-311++G**)
Compound
Bond
ρ (au)
ε
DI
]−
DIPP
[
[
[
HC(N-DIPP)
2
C–N
B–N
C–N
C–N
B–N
C–N
C–C
0.349
0.195
0.277
0.347
0.186
0.258
0.240
0.343
0.265
0.388
0.102
0.207
0.197
0.174
0.008
0.050
0.151
0.067
0.070
0.000
0.335
0.034
0.193
0.000
0.075
0.062
1.256
DIPP
2−
DIPP
DIPP
DIPP
DIPP
DIPP
NBN]
C(N-DIPP)
MgH
0.555 type, however, there is a linear relationship between DI and the
2
−
H
2
2
]
0.987
1.257
0.529
30
formal bond order. Values in Table 2 are in agreement with
single C–N bonds for the bis-amide ligand and delocalized C–N
0.959 bonds for the amidinate ligand. The DI for the B–N bond in
NBN] is considerably lower than that for the C–N bonds
on account of a mismatch in B and N electronegativities which
1.765 results in a bond with significant ionic character. Its value of
HC(N-DIPP)
2
DIPP
NBN-Mg
H
H
H
H
H
H
H
2
3
2
3
2
3
2
C(N-DIPP)
2
Mg
C–CH
CvCH
C–NH
CvNH
3
1.018
1.906
1.026
DIPP
2−
[
2
CvC
C–N
CvN
B–N
B–N
B–N
2
B–NH
B–NH
3
0.351
0.644
0.528
0
3 3
.555 lies between those for the DI’s in H B–NH (0.351) and
2
/H
2
BvNH
2
H BvNH (0.644) and is similar to that in borazine (0.528).
Borazine
2
2
These data show that there is only minor delocalization of
electron density over the NBN framework in the bam ligand.
Its charge density and bonding properties more closely match
simplest imine. As expected, the ρC–N of 0.277 au found in that of a bridged diamide with localized charges on N. Similar
2
−
[
H
2
C(N-DIPP)
2
]
is similar to that of a single C–N bond in conclusions can be drawn from electron-density plots shown
DIPP
2−
H C–NH (0.265 au). The B–N bond in [
NBN] shows the in Fig. S25 (ESI†).
3
2
smallest BCP electron density: ρB–N = 0.195 au. The low elec-
tron density along the B–N axis hints to a low bond order but
Catalysis
comparison with a formal single B–N bond (H N–BH : 0.102 The three Mg, Ca and Sr bam complexes have been tested in
3
3
2 2
au), a formal double BvN bond (H NvBH : 0.207 au) and the the intramolecular alkene hydroamination of three different
B–N bond in borazine (N : 0.197 au) shows that it is close benchmark aminoalkenes (Table 3). The latter differ in their
3 3 3
B H
to the latter two. It should be noted, however, that the substitution in the chain and are known to cyclize with
suggested aromaticity in borazine is still a matter of debate. Its increasing difficulty along the row: Ph,Ph < Cy < Me,Me.
2
6
DIPP
experimentally determined electron distribution and its reac-
tivity are strongly in favour of significant localization of elec- the longest reaction times and, at room temperature, only con-
The magnesium based catalyst
3
NBN-Mg·(THF) exhibits
2
7
tron density on N. The latter is due to the large difference in verts the Ph,Ph substrate for which cyclization is favored by
electronegativity between B and N that causes a mismatch
between p-orbital energies of B and N. A similar electron local-
a
Table 3 Catalytic intramolecular alkene hydroamination
ization on N is in line with the rather high negative charge on
DIPP
2−
the N atoms in [
NBN] (vide supra). The BCP electron den-
sities for the C–N and B–N bonds in the Mg complexes follow
the same trends as observed for the anions but are generally
slightly lower. The electron densities in the Mg–N bonds are
quite low, as expected for largely ionic bonds. The values
Catalyst
Substrate Temp. (°C) Time (h) Conversion (%)
DIPP
found for the Mg–N bonds in
2
NBN-Mg (0.060 au) and H C
DIPP
NBN-Mg·(THF)
3
Me,Me
Cy
Ph,Ph
Ph,Ph
60
60
25
25
24
4
12
12
—
(
N-DIPP) Mg (0.061 au) are very similar and significantly
2
>99%
>99%b
79%
higher than those observed in HC(N-DIPP) MgH (0.042 au).
2
Another parameter of interest for the evaluation of charge
2
8
delocalization is the ellipticity (ε) which measures the extent
DIPP
DIPP
NBN-Ca·(THF)
4
Me,Me
Cy
Ph,Ph
Ph,Ph
60
25
25
25
1.4
1.4
0.1
12
>99%
>99%
>99%b
63%
of anisotropy in the electron density in a plane perpendicular
to the bond at the BCP: ε = (λ /λ ) − 1. Whereas pure σ bonds
1 2
have ε values close to 0, the π bond in ethylene is considerably
elliptic with ε = 0.343 (see Table 2). Whereas the ellipticity of
the CvN bond in the formidinate anion (0.174) is similar to
NBN-Sr·(THF)4 Me,Me
60
25
25
25
25
1.6
2.7
1.8
12
>99%
>99%
Cy
DIPP
2−
Ph,Ph
Ph,Ph
Ph,Ph
>99%b
that in an imine (0.193), the B–N bond in [
little π character but ε in
NBN] shows
90%
>99%
DIPP
c
NBN-Mg (0.067) is similar to
3.3
that in borazine (0.062). The small anisotropy in the B–N
bond electron density, however, is in favor of a bond with
mainly σ character.
a
Substrate concentration: 0.011 M in C
6 6
D ; 10 mol% cat. The time for
b
>
99% conversion has been determined by proton NMR. Reaction run
c
8
in THF-d . Reaction run with 2.5 mol% catalyst.
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
3
1
the Thorpe-Ingold effect. The other substrates need either a
higher reaction temperature (Cy) or do not react at all (Me,Me).
The sluggish conversion with Mg-based catalysts has been
3
2
observed earlier in transformations with the catalyst MgN″
The bam Mg catalyst, however, is noticeably less active than
MgN″ . The latter is able to convert the Me,Me-substrate,
albeit very slow: 10 mol% cat., 80 °C, 5 days, 60% conversion.
The heteroleptic Mg catalyst (DIPP-nacnac)MgMe·THF (DIPP-
nacnac = HC[C(Me)N-DIPP] ) shows higher conversion rates;
e.g. full conversion of the Ph,Ph substrate is smooth: 5 mol%
2
.
2
3
2
2
3
3
cat., 25 °C, 0.25 h.
The catalytic alkene hydroamination with the calcium
DIPP
complex,
catalyst also the Me,Me substrate can be fully converted. The
activity of the bam Ca complex compares to that of CaN″ ·(THF)
full conversion of the Ph,Ph substrate: 5 mol% cat., 20 °C, 1 h.
It does, however, not match the high activity found for
4
NBN-Ca·(THF) , is significantly faster and with this
2
2
;
3
4
33
(
DIPP-nacnac)CaN″·THF. This catalyst is able to cyclize the Me,
Me substituted substrate: 10 mol% cat., 0.25 h, 25 °C.
33
Although the reactivity of the alkaline-earth metal com-
plexes increases down the group, the catalytic activities found
DIPP
for
4
NBN-Sr·(THF) are in all cases less than those observed
for the Ca catalyst. This trend in activity, Mg < Ca > Sr, has
1
4,33,35
been observed previously
and can be explained by
assuming two anti-parallel effects. The reactivity of the M–N
Scheme 4 Proposed catalytic cycle for the intramolecular alkene
hydroamination with bora-amidinate alkaline-earth metal catalysts.
bond increases along the row: Mg < Ca < Sr. The metal Lewis
1
2
acidity necessary for alkene activation, however, decreases
along the same row: Mg > Ca > Sr. The balance between high
reactivity and strong Lewis acidity seems to be optimal for Ca
catalysts which are often found to be superior to catalysts
incorporating the other group 2 metals.
In order to find support for the non-innocence of the bam
ligand along the catalytic pathway, we took advantage of the inac-
DIPP
It is important to note that changing the solvent from tivity of
benzene to THF gave in all cases a significant decrease in stituted substrate (Table 3). Reaction of
NBN-Mg·(THF) in the cyclization of the Me,Me-sub-
3
DIPP
3
NBN-Mg·(THF) with
activity. This is likely related to the fact that substrate acti- one equivalent of H NCH CMe CH CHvCH
2
2
2
2
2
in C gave at
6 6
D
vation needs an open coordination site at the metal. room temperature a clean conversion to a new species.
1
Monitoring catalytic conversions by H NMR shows that at the Numerous crystallization attempts failed but with the assistance
end of the reaction the bam catalysts are formed back again. of various NMR methods it was possible to define the product as
DIPP
Repeated additions of substrate gave smooth conversions but
(
NBN-H)Mg-N(H)CH CMe CH CHvCH ·(THF) (see ESI†).
2 2 2 2 x
catalyst activity decreases after each cycle which is likely due to Consistent with the inactivity of this Mg catalyst, prolonged
catalyst poisoning by the product. The latter is an amine that heating of the sample to 100 °C gave no further conversion.
may compete with the substrate for metal coordination.
Importantly, also no Schlenk equilibria were observed. Therefore,
We propose a mechanism in which the bam ligand reacts we propose a catalytic cycle in which the bam ligand plays an
as a cooperative ligand that, in contrast to spectator ligands, active role in shuttling protons from the substrate back to the
actively takes part in the catalytic cycle (Scheme 4). After sub- product.
DIPP
2−
strate–metal coordination, the
NBN anion deprotonates
The bam catalysts introduced in here are of a new type.
the amino alkene. Product formation could proceed in two Hitherto group 2 metal catalyst design was mainly based on
steps; i.e. ring closure followed by protonation of the metal heteroleptic complexes L–M–R in which L is an inactive specta-
alkyl species. More likely is a proton-assisted pathway that tor ligand and R the active part of the catalyst. In this classical
circumvents formation of highly energetic metal-alkyl inter- design, the spectator ligand L may not only control the sterics
mediates which for Ca and Sr hitherto have been unobserved. and electronics of the catalyst but also influences important
In this case, the reaction proceeds through a transition state physical properties like solubility. Indeed, although
with simultaneous ring closure and protonation in which homoleptic catalysts of the type MR
2
, i.e. complexes with two
DIPP
NBN-H− takes part as a proton source. This avoids the active groups, often show higher activities as their heteroleptic
3
7
necessity for coordination of a second substrate molecule equivalents, in some cases catalysis failed due to precipi-
which alternatively has been proposed as a proton-source in tation of intermediates. For example, the catalytic alkene
3
6
order to prevent a metal-alkyl intermediate.
hydrophosphination with the homoleptic CaN″ ·(THF) cata-
2 2
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
lyst is particularly slow on account of precipitation of the Ca degassed solvents (all solvents were dried over a column except
38
(
PPh ) intermediate. In this case, a heteroleptic L–M–R cata- for THF which was dried over Na and redistilled). Following
2 2
lyst with a solubilizing spectator ligand is advantageous. Also reagents were obtained commercially: (nBu)
2
Mg (1.0 M solu-
in enantioselective transformations heteroleptic L*–M–R cata- tion in heptane, Sigma-Aldrich). The starting materials HB[NH
1
6
19
DIPP
6
lysts are a prerequisite: the chiral spectator ligand L* controls (DIPP)]2, (DMAT) Ca·(THF)
the stereoselectivity.
and
Catalyst selectivity, however, can be been prepared according to literature procedures. NMR spectra
disturbed by the Schlenk equilibrium in which larger amounts were measured on Bruker Avance III HD 400 MHz and Bruker
NBN-Sr·(THF)4 have
2
2
3
4,39
3
4
of achiral, but catalytically active, MR may be formed. The Avance III HD 600 MHz spectrometers. Elemental analysis was
2
challenge in enantioselective group 2 metal catalysis is the performed with an Hekatech Eurovector EA3000 analyzer.
difficulty to control such Schlenk equilibria. In the new type of Crystal structures have been measured on a SuperNova
catalysts introduced here, the role of spectator and active (Agilent) diffractometer with dual Cu and Mo microfocus
DIPP
group is combined in one ligand which implies that Schlenk sources and an Atlas S2 detector (
3
NBN-Mg·THF ) and on
equilibria are obsolete.
Bruker SMART diffractometer with APEXII detector
DIPP
(
NBN-Ca·THF ). Crystal data and experimental details are
4
included in the ESI.† Crystallographic data have been de-
posited with the Cambridge Crystallographic Data Centre as
Conclusions
supplementary
publication
) and 1551538 (
no.
CCDC
NBN-Mg·THF
Schlenk-flask was
NBN-H2 (2.32 g, 6.37 mmol) suspended in
, as monomeric entities in 6.4 mL heptane. To this suspension (nBu) Mg (1.0 M in hep-
1551537
DIPP
DIPP
Bora-amidinate complexes of Mg and Ca have been prepared.
(
NBN-Ca·THF
4
3
).†
DIPP
DIPP
Crystal structures of their THF adducts,
NBN-Mg·(THF)
3
Synthesis of
3
NBN-Mg·(THF) . A
DIPP
DIPP
and
NBN-Ca·(THF) , show that these compounds crystal- charged with
4
DIPP
6
lize, similar to
NBN-Sr·(THF)
4
2
which the THF ligands occupy the larger part of the coordi- tanes, 6.40 mL, 6.40 mmol) was added. Stirring overnight at
1
nation sphere. The very simple H NMR spectrum of the Ca room temperature, yielded a slightly turbid solution. All vola-
6 6
and Sr complexes in C D suggest that the monomeric struc- tiles were removed under reduced pressure resulting in a
ture found in the crystal is retained in solution and that the bright brown solid. The product was extracted with 20 mL of
DIPP substituents are rotating freely. However, the Mg THF and crystallized from a mixture of 20 mL THF and 15 mL
complex dissolved in C
species for which we propose a dimeric constitution.
6 6
D is in equilibrium with a second hexane. The product was obtained as clear colourless crystal-
1
line rods. Yield: 2.92 g, 4.87 mmol, 76%. H NMR (600 MHz,
3
Calculations show that although the bam ligands are iso- THF-d , 25 °C): δ = 1.08 (d, JHH = 6.9 Hz, 24H, CH(CH3)2),
8
electronic to amidinate ligands, the electron distribution in 1.69–1.79 (m, 12H, THF), 3.55–3.61 (m, 12H, THF), 3.76 (sept,
3
the ligand framework is different. Whereas the negative charge
3 2
JHH = 6.9 Hz, 4H, CH(CH ) ), 4.56 (s, br, 1H, BH), 6.39 (t,
3
3
in amidinates is delocalized by resonance, the twofold negative
JHH = 7.5 Hz, 2H, CH
), 6.74 (d, J = 7.5 Hz, 4H, CHarom.)
arom HH
DIPP
2−
13
1
charge in the
NBN anion is mainly located on both N ppm. C{ H} NMR (100 MHz, THF-d
), 25.4 (THF), 27.0 (CH(CH
than that on the N atoms in [H C(N-DIPP) ] , a bis-amide (Carom), 121.0 (Carom), 141.1 (Carom), 156.3 (Carom) ppm. B{ H}
8
, 25 °C): δ = 24.0
atoms. The negative charge on these N atoms is even higher (CH(CH
3
)
2
3 2
) ), 67.2 (THF), 115.2
2
−
11
1
2
2
with localized charges. This is partly also due to the +I effect NMR (193 MHz, THF-d
of B which releases electrons to N. lysis: calculated for C36
The Mg, Ca and Sr bam complexes are all active in the intra- found: C 71.29, H 9.81, N 4.63.
8
, 25 °C): δ = 30.9 ppm. Elemental ana-
H
59BMgN : C 71.71, H 9.86, N 4.65;
2 3
O
DIPP
molecular alkene hydroamination with highest activities found
Synthesis of
NBN-Ca·(THF)
NBN-H
2
4
.
A
Schlenk-flask was
(610.3 mg, 1.68 mmol) and
catalysts, the activities are modest and generally lower than (DMAT) Ca·(THF) (1.00 g, 1.68 mmol; DMAT = 2-Me N-
DIPP
DIPP
for
NBN-Ca·(THF)
4
. Compared to other group 2 metal charged with
2
2
2
those of the homoleptic amides MN″
2
3
. A catalytic cycle is α-Me Si-benzyl) and this mixture was dissolved in 5.0 mL of
proposed in which the bam ligand is cooperative and plays an heptane and 0.5 mL of THF. The clear solution was stirred
active role in shuttling protons from the substrate back to the overnight, yielding a slightly turbid solution. All volatiles were
product.
removed under reduced pressure resulting in a bright brown
We believe that this new type of catalysts in which spectator solid. The product was extracted with 10 mL of THF and after
and active group are connected by a bridge may give impetus slowly cooling the THF solution to +7 °C, the product crystal-
to the field of group 2 metal catalysis and currently we evaluate lized in the form of colourless rods. Yield: 0.624 g, 0.90 mmol,
1
3
these possibilities.
54%. H NMR (600 MHz, THF-d
8
, 25 °C): δ = 1.04 (d, JHH =
6
.9 Hz, 24H, CH(CH ) ), 1.72–1.76 (m, 16H, THF), 3.56–3.61
3
2
3
(
m, 16H, THF), 3.79 (sept, JHH = 6.9 Hz, 4H, CH(CH
3
)
2
), 4.71
3
(
s, br, 1H, BH), 6.26 (t, JHH = 7.4 Hz, 2H, CHarom), 6.69 (d,
Experimental section
General considerations
3
13
1
JHH = 7.5 Hz, 4H, CHarom.) ppm. C{ H} NMR (100 MHz THF-
, 25 °C): δ = 24.8 (CH(CH ), 25.4 (THF), 26.7 (CH(CH
67.2 (THF), 113.6 (Carom), 120.8 (Carom), 140.2 (Carom),
d
8
3
)
2
3 2
) )
All experiments were carried out in dry glassware under N
2
11
1
using standard Schlenk-techniques and freshly dried and 159.3 (Carom) ppm. B{ H} NMR (193 MHz, THF-d , 25 °C):
8
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.

View Article Online
Paper
Dalton Transactions
δ = 30.0 ppm. Elemental analysis: calculated for C40
C 69.54, H 9.78, N 4.05; found: C 68.93, H 9.49, N 4.34.
H
2 4
67BCaN O :
12 J. Penafiel, L. Maron and S. Harder, Angew. Chem., Int. Ed.,
2015, 54, 201–206.
2
Experimental procedure for catalysis. In the inert N atmo- 13 (a) T.-G. Ong, J. S. O’Brien, I. Korobkov and D. S. Richeson,
sphere of a glove box, a given amount of the catalyst (between
Organometallics, 2006, 25, 4728–4730; (b) J. R. Lachs,
A. G. M. Barrett, M. R. Crimmin, G. Kociok-Kohn,
M. S. Hill, M. F. Mahon, P. A. Procopiou and A. Panayiotis,
Eur. J. Inorg. Chem., 2008, 26, 4173–4179.
DIPP
DIPP
6
.4 and 10.6 μmol of
NBN-Mg·(THF)3,
) was weighed into a NMR tube that was
rigorously dried by heating under high vacuum. The catalyst
NBN-Ca·(THF)4,
DIPP
or
NBN-Sr·(THF)
4
was dissolved in the respective amount of C D (between 704 14 M. R. Crimmin, I. J. Casely and M. S. Hill, J. Am. Chem.
6
6
and 1166 μL) to get a concentration of 0.0011 M in solution.
Soc., 2005, 127, 2042–2043.
After addition of the respective substrate (2,2-dimethylpent-4- 15 J. Penafiel and S. Harder, unpublished results.
en-1-amine), (1-allylcyclohexyl)methanamine, or 2,2-diphenyl- 16 J. Spielmann, M. Bolte and S. Harder, Chem. Commun.,
pent-4-en-1-amine (10–40 equivalents, between 64 and
2009, 6934–6936.
2
56 μmol), the substrate to product conversion was monitored 17 R. R. Fraser, T. S. Mansour and S. Savard, J. Org. Chem.,
1
by H NMR spectroscopy.
1985, 50, 3232–3234.
1
1
8 S. Harder and F. Feil, Organometallics, 2001, 20, 4616–4622.
9 M. L. Cole and P. C. Junk, Z. Anorg. Allg. Chem., 2015, 641,
2
624–2629.
Conflicts of interest
2
0 M. L. Cole and P. C. Junk, New J. Chem., 2005, 29, 135–140.
The authors declare no competing financial interest.
21 M. S. Hill, D. J. Liptrot, D. J. MacDougall, M. F. Mahon and
T. P. Robinson, Chem. Sci., 2013, 4, 4212–4222.
2
2
2 D. Yang, Y. Ding, H. Wu and W. Zheng, Inorg. Chem., 2011,
0, 7698–7706.
3 S. Harder, Angew. Chem., Int. Ed., 2004, 43, 2714–2718.
5
Acknowledgements
We acknowledge the Deutsche Forschungsgemeinschaft for 24 A. E. Reed, R. B. Weinstock and F. Weinhold, J. Chem.
financial support (HA 3218/8-1).
Phys., 1985, 83, 735–746.
2
2
5 R. F. W. Bader, Atoms in Molecules: A Quantum Theory,
Oxford University Press, Oxford, UK, 1990.
6 R. Boese, A. H. Maulitz and P. Stellberg, Chem. Ber., 1994,
References
127, 1887–1899.
1
2
3
4
H. Fußstetter and H. Nöth, Chem. Ber., 1979, 112, 3672– 27 (a) P. v. R. Schleyer, H. Jiao, N. J. R. van Eikema-Hommes,
3
681.
C. Fedorchuk, M. Copsey and T. Chivers, Coord. Chem. Rev.,
007, 251, 897–924.
A. M. Corrente, M. Andrea and T. Chivers, Inorg. Chem.,
010, 49, 2457–2463.
V. G. Malkin and O. L. Malkina, J. Am. Chem. Soc., 1997,
119, 12669–12670; (b) B. Kiran, A. K. Phukan and
E. D. Jemmis, Inorg. Chem., 2001, 40, 3615–3618;
(c) R. Miao, G. Yang, C. Zhao, J. Hong and L. Zhu, J. Mol.
Struct. (THEOCHEM), 2005, 715, 91–100.
2
2
S. Harder and D. Naglav, Eur. J. Inorg. Chem., 2010, 18, 28 C. Silva López and A. R. de Lera, Curr. Org. Chem., 2011, 15,
2
836–2840.
3576–3593.
5
6
S. Harder, Dalton Trans., 2010, 39, 6677–6681.
S. Harder, B. Freitag, P. Stegner, J. Pahl and D. Naglav,
Z. Anorg. Allg. Chem., 2015, 641, 2129–2134.
For reviews, see: (a) J. Barker and M. Kilner, Coord. Chem.
Rev., 1994, 133, 219–300; (b) F. T. Edelmann, Coord. Chem.
29 (a) Y. G. Wang and N. H. Werstiuk, J. Comput. Chem., 2003,
24, 379–385; (b) R. F. W. Bader, A. Streitwieser, A. Neuhaus,
K. E. Laidig and P. Speers, J. Am. Chem. Soc., 1996, 118,
4959–4965; (c) R. F. W. Bader and M. E. Stephens, J. Am.
Chem. Soc., 1975, 97, 7391–7399.
7
Rev., 1994, 137, 403–481; (c) F. T. Edelmann, Adv. 30 C. Firme, A. O. C. Antunes and P. Esteves, Chem. Phys. Lett.,
Organomet. Chem., 2008, 57, 183–352. 2009, 129–133.
8
(a) T. Chivers, D. J. Eisler, C. Fedorchuk, G. Schatte, 31 M. E. Jung and G. Piizzi, Chem. Rev., 2005, 105, 1735–1766.
H. M. Tuononen and R. T. Boere, Inorg. Chem., 2006, 45, 32 M. Arrowsmith, M. R. Crimmin, A. G. M. Barrett, M. S. Hill,
2
119–2131; (b) T. Chivers, D. J. Eisler, C. Fedorchuk,
G. Kociok-Köhn and P. A. Procopiou, Organometallics, 2011,
30, 1493–1506.
33 M. R. Crimmin, M. Arrowsmith, A. G. M. Barrett,
I. J. Casely, M. S. Hill and P. A. Procopiou, J. Am. Chem.
Soc., 2009, 131, 9670–9685.
G. Schatte, H. M. Tuononen and R. T. Boere, Chem.
Commun., 2005, 3930–3932.
S. Harder, C. Petrisor, C. Boulho and O. Staal, unpublished
results.Attempted ethylene polymerization.
9
1
0 T. Chivers, C. Fedorchuk and M. Parvez, Organometallics, 34 F. Buch and S. Harder, Z. Naturforsch., B: Chem. Sci., 2008,
005, 24, 580–586. 63, 169–177.
1 (a) S. Harder, Chem. Rev., 2010, 110, 3852–3876; 35 (a) N. Romero, S. Rosca, Y. Sarazin, J.-F. Carpentier,
2
1
(b) M. S. Hill, D. J. Liptrot and C. Weetman, Chem. Soc.
L. Vendier, S. Mallet-Ladeira, C. Dinoi and M. Etienne,
Rev., 2016, 45, 972–988.
Chem. – Eur. J., 2015, 21, 4115–4125; (b) B. Liu, T. Roisnel,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Dalton Transactions
Paper
J.-F. Carpentier and Y. Sarazin, Chem. – Eur. J., 2013, 19, 38 M. R. Crimmin, A. G. M. Barrett, M. S. Hill, P. B. Hitchcock
1
3445–13462; (c) A. G. M. Barrett, M. R. Crimmin,
and P. A. Procopiou, Organometallics, 2007, 26, 2953–
M. S. Hill, P. B. Hitchcock, G. Kociok-Köhn and
2956.
P. A. Procopiou, Inorg. Chem., 2008, 47, 7366; 39 (a) F. Buch and S. Harder, Z. Naturforsch., B: Chem.
(
d) A. G. M. Barrett, I. J. Casely, M. R. Crimmin, M. S. Hill,
J. R. Lachs, M. F. Mahon and P. A. Procopiou, Inorg. Chem.,
009, 48, 4445–4453.
6 (a) M. R. Gagne and T. J. Marks, J. Am. Chem. Soc., 1992,
14, 275–294; (b) B. D. Stubbert and T. J. Marks, J. Am.
Sci., 2008, 63, 169–177; (b) X. Zhang, T. J. Emge and
K. C. Hultzsch, Angew. Chem., Int. Ed., 2012, 51, 394–
398; (c) P. Horrillo-Martínez and K. C. Hultzsch,
2
3
3
Tetrahedron
Lett.,
2009,
50,
2054–2056;
1
(d) V. Rodriguez-Ruiz, R. Carlino, S. Bezzenine-Lafollée,
R. Gil, D. Prim, E. Schulza and J. Hannedouche,
Dalton Trans., 2015, 44, 12029–12059; (e) J. S. Wixey
and B. D. Ward, Chem. Commun., 2011, 47, 5449–
5451; (f) T. D. Nixon and B. D. Ward, Chem.
Commun., 2012, 48, 11790–11792.
Chem. Soc., 2007, 129, 6149–6167; (c) S. Tobisch, Chem. –
Eur. J., 2015, 21, 6765–6779.
7 (a) F. Buch, J. Brettar and S. Harder, Angew. Chem., Int. Ed.,
2
006, 45, 2741–2745; (b) J. Spielmann, F. Buch and
S. Harder, Angew. Chem., Int. Ed., 2008, 47, 9434–9438.
This journal is © The Royal Society of Chemistry 2017
Dalton Trans.