Metal???F?C Bonding in Low-Coordinate Alkaline Earth Fluoroarylamides

Article abstract of DOI:10.1002/chem.201901262

A set of calcium and barium complexes containing the fluoroarylamide N(C₆F₅)₂^? is presented. These compounds illustrate the key role of stabilising M???F?C secondary interactions in the construction of low-coordinate alkaline earth complexes. The nature of Ca???F?C bonding in calcium complexes is examined in the light of structural data, bond valence sum (BVS) analysis and DFT computations. The molecular structures of [Ca{N(C₆F₅)₂}₂(Et₂O)₂] (4′), [Ca{μ-N(SiMe₃)₂}{N(C₆F₅)₂}]₂ (5₂), [Ba{μ-N(C₆F₅)₂}{N(C₆F₅)₂}?toluene]₂ (6₂), [{BDI^DiPP}CaN(C₆F₅)₂]₂ (7₂), [{N^N^DiPP}CaN(C₆F₅)₂]₂ (8₂), and [Ca{μ-OB(CH(SiMe₃)₂)₂}{N(C₆F₅)₂}]₂ (9₂), where {BDI^DiPP}^? and {N^N^DiPP}^? are the bidentate ligands CH[C(CH₃)NDipp]₂^? and DippNC₆H₄CNDipp^? (Dipp=2,6-iPr₂-C₆H₃), are detailed. Complex 6₂ displays strong Ba???F?C contacts at around 2.85 ?. The calcium complexes feature also very short intramolecular Ca?F interatomic distances at around 2.50 ?. In addition, the three-coordinate complexes 7₂ and 8₂ form dinuclear structures due to intermolecular Ca???F?C contacts. BVS analysis shows that Ca???F?C interactions contribute to 15–20 % of the bonding pattern around calcium. Computations demonstrate that Ca???F?C bonding is mostly electrostatic, but also contains a non-negligible covalent contribution. They also suggest that Ca???F?C are the strongest amongst the range of weak Ca???X (X=F, H, C_π) secondary interactions, due to the high positive charge of Ca²⁺ which favours electrostatic interactions.

Full text of DOI:10.1002/chem.201901262

A Journal of
Accepted Article
Title: Metal···F–C bonding in low-coordinate alkaline-earth
fluoroarylamides
Authors: Yann Sarazin, Hanieh Roueindeji, Antsa Ratsifitahina, Thierry
Roisnel, Vincent Dorcet, Samia Kahlal, Jean-Yves Saillard,
and Jean-François Carpentier
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: Chem. Eur. J. 10.1002/chem.201901262
Link to VoR: http://dx.doi.org/10.1002/chem.201901262
Supported by

10.1002/chem.201901262
Chemistry - A European Journal
Metal···F–C bonding in low-coordinate alkaline-earth fluoroarylamides
Hanieh Roueindeji, Antsa Ratsifitahina, Thierry Roisnel, Vincent Dorcet, Samia Kahlal,
Jean-Yves Saillard,* Jean-François Carpentier and Yann Sarazin*[a]
[a] Univ Rennes, CNRS, ISCR (Institut des Sciences Chimiques de Rennes) – UMR 6226, F-35000
Rennes, France.
Corresponding authors: yann.sarazin@univ-rennes1.fr; jean-yves.saillard@univ-rennes1.fr
Keywords: calcium • alkaline earth metals • low coordinate complexes • DFT computations • metal-
fluorine bonding
1
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Abstract
‒
A set of calcium and barium complexes containing the fluoroarylamide N(C6F5)2 is presented. These
compounds illustrate the key role of stabilising M···FC secondary interactions in the construction of
low coordinate alkaline earth complexes. The nature of Ca···FC bonding in calcium complexes is
examined in the light of structural data, bond valence sum analysis (BVSA) and DFT computations. The
molecular structures of [Ca{N(C6F5)2}2.(Et2O)2] (4’), [Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52), [Ba{µ-
N(C6F5)2}{N(C6F5)2}.toluene]2 (62), [{BDIDiPP}CaN(C6F5)2]2 (72), [{N^NDiPP}CaN(C6F5)2]2 (82), and
[Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 (92), where {BDIDiPP}‒ and {N^NDiPP}‒ are the bidentate ligands
CH[C(CH3)NDipp]2 and DippNC6H4CNDipp (Dipp = 2,6-iPr2-C6H3), are detailed. Complex 62 displays
strong Ba···FC contacts at ca. 2.85 Å. The calcium complexes feature also very short intramolecular
Ca-F interatomic distances, at around 2.50 Å. Besides, the three-coordinate complexes 72 and 92 form
dinuclear edifices due to intermolecular Ca···FC contacts. BVSA shows that Ca···FC interactions
contribute to 15-20% of the bonding pattern around calcium. Computations demonstrate that Ca···FC
bonding is mostly electrostatic, but also contains a non-negligible covalent contribution. They also
suggest that Ca···FC are the strongest amongst the range of weak Ca···X (X = F, H, C) secondary
interactions, owing to the high positive charge of Ca2+ which favours electrostatic interactions.
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Introduction
Well defined, soluble complexes of the abundant large alkaline earths calcium, strontium and barium (=
Ae metals) have of late emerged as viable alternative to mainstream homogenous catalysts based on
expensive late transition metals.[1] A panel of molecular Ca-Ba precatalysts featuring a controlled
coordination sphere tailored around bulky, multidentate and often monoanionic ligands are now
available for the catalysis of a growing range of transformations such as hydroelementations,[2]
dehydrocouplings[3] or polymerisations.[4] Complexes of these metals also promote remarkable cases of
stoichiometric or subcatalytic functionalisations of small molecules.[5] The pronounced electrophilicity
of the metal in low coordinate Ae complexes is often the main driving force behind the efficiency of
these systems. Ae metals are large (rionic = 1.00, 1.18 and 1.35 Å for Ca, Sr and Ba in a coordination
number of 6), electropositive elements (with their respective Pauling electronegativity of 1.00, 0.95 and
0.89) that generate complexes where the bonding pattern around the dication Ae2+ is governed by
electrostatic and steric factors, with little propensity for orbital contributions. The perennial search for
low coordinate complexes offering the best compromise between stability and reactivity has triggered
the design of multidentate ligands based on O- or N-donors that match the hard, oxophilic nature of Ae
metals. The ubiquitous, bidentate β-diketiminate {BDIDiPP}‒ (where BDIDiPP = CH[C(CH3)NDipp]2,
Dipp = 2,6-diisopropylphenyl, Figure 1) has proved exceptionally successful in enabling access to
highly reactive, low coordinate Ae species,[1-6] with the four-coordinate [{BDIDiPP}CaN(SiMe3)2.(thf)]
(1) proving a notably versatile and potent precatalyst.[1] Related complexes were subsequently devised,
‒
and the sterically hindered iminoanilido ligand {N^NDiPP
}
for instance afforded
[{N^NDiPP}CaN(SiMe3)2.(thf)] (2), an excellent precatalyst for the cyclohydroamination of α,ω-
aminoalkenes.[7] Yet, because it reduces electrophilicity of the metal ion and might inhibit coordination
of an incoming substrate, the presence of coordinated thf in 1 and 2 and related complexes often proves
detrimental to catalytic turnovers or stoichiometric reactivity studies.[1] This is perhaps best epitomised
by Hill’s recent demonstration that [{BDIDiPP}CaH]2 (made from solvent-free [{BDIDiPP}CaN(SiMe3)2]
(3)[8]) inserts non-activated α-olefins to generate [{BDIDiPP}CaR]2 Ca-alkyl species, while
[{BDIDiPP}CaH.(thf)]2 is comparatively inert.[2j,6e]
Figure 1 Examples of low-coordinate calcium complexes bearing chelating N-based ligands.
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
It has been shown on repeated occasions that the removal of coordinated solvent molecules was a
valuable strategy in order to obtain low-coordinate Ae compounds. The implementation of secondary
interactions has been a driving force in this purpose,[9] and a variety of unusual complexes stabilised by
β-agostic Ae···HSi,[10] Ae···C(π)[11] or, perhaps most importantly, Ae···FC[12] non-covalent
interactions, is now available. Many of them have been used to great effect as molecular precatalysts.[1]
In this context, fluorinated amines have been used to obtain low coordinate, homoleptic synthetic
precursors of electrophilic elements (Figure 2). Sm[N(C6F5)(SiMe3)]3 exhibited three intramolecular
Sm···FC contacts.[13] The Schelter group prepared U(III), U(IV) and Ce(III) complexes bearing various
‒
fluorinated arylamides, notably N(C6F5)2 and N(SiMe3)(o-F-C6H4)‒, that feature several stabilising
metal···FC contacts.[14] Sundermeyer and coworkers pioneered this theme in main group metal
chemistry. They prepared volatile magnesium and alkali salts of the mixed N(C6F5)(C(CF3)3)‒ amide,[15]
and also reported on the highly Lewis acidic Al[N(C6F5)2]3 and Ga[N(C6F5)2]3.[16] Prior to this,
Westerhausen had utilised 2,6-difluoroanilide to synthesise multinuclear complexes that exhibit intra-
and intermolecular Ae···FC contacts (Ae = Sr or Ba), such as the one-dimensional polymer
[(thf)2Ba{N(H)-2,6-F2-C6H3}]∞.[12b]
Figure 2. Selected examples of homoleptic fluorinated arylamido complexes.
One topical field of studies in main group chemistry aims at devising low-coordinate alkaline earth
molecular (pre)catalysts that present an optimal compromise between stability and reactivity. Improving
our understanding of the nature of the bonding, especially with respect to secondary interactions, is key
in this purpose. We are presenting here the outcome of our investigations concerning the syntheses,
structures and bonding patterns in calcium and barium complexes, including heteroleptic ones,
‒
supported by the perfluorinated arylamide N(C6F5)2 . Spectroscopic and crystallographic data are
discussed in the light of complementary DFT computations and bond valence sum analysis.[17]
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Results and Discussion
Synthesis of complexes
Our investigations started with the attempted preparations of calcium-based starting materials devoid of
coordinated Lewis bases. The homoleptic complex [Ca{N(C6F5)2}2] (4) was obtained in a non-optimised
82% yield by treatment of [Ca{N(SiMe3)2}2]2 with a stoichiometric amount of the perfluorinated
[18]
arylamine HN(C6F5)2 in toluene (Scheme 1). Complex 4, the nuclearity of which could not be
established, is only sparingly soluble in aliphatic and aromatic hydrocarbons; single crystals of 4 suitable
for X-ray diffractometry could not be obtained in spite of multiple attempts. Layering a petroleum ether
solution of [Ca{N(SiMe3)2}2]2 with a solution of HN(C6F5)2 in this solvent also proved inadequate.[14b]
On the other hand, 4 is soluble in ethers and colourless crystals of the diethyl ether adduct
[Ca{N(C6F5)2}2.(Et2O)2] (4’) were grown upon recrystallisation from a diethyl ether solution.
Scheme 1. Synthesis of fluoroarylamide-containing starting materials [Ca{N(C6F5)2}2] (4),
[Ca{N(C6F5)2}2.(Et2O)2] (4’), [Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52) and [Ba{µ-N(C6F5)2}{N(C6F5)2}.toluene]2
(62), with representation of their molecular solid-state structures (structure for complex 4 unknown).
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
19
The F NMR spectrum of complex 4’ in benzene-d6 at room temperature is characterised by three
resonances: a doublet centred on δ ‒153.80 ppm for o-F atoms, and two overlapping triplets in the region
1
δ ‒162.60 to ‒163.00 ppm for m- and p-F. Its H NMR spectrum solely indicates the presence of
coordinated diethyl ether. By comparison, the 19F NMR spectrum of ether-free 4 in thf-d8 exhibits three
broad resonances at δ ‒159.45, ‒168.88 and ‒178.23 ppm integrating respectively for 8, 8 and 4F atoms.
The heteroleptic dinuclear complex [Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52) was synthesised by reacting
[Ca{N(SiMe3)2}2]2 with two equiv of HN(C6F5)2 in toluene. It was isolated in 64% yield as colourless
crystals upon recrystallisation from a concentrated solution in 1,2-difluorobenzene, and its solid-state
structure was established (vide infra). Its solubility in hydrocarbons, Et2O and thf is high. The 1H NMR
spectrum of 52 (benzene-d6, 25 °C) expectedly features a single resonance, a singlet at δ 0.01 ppm. Three
sharp and well separated resonances are detected in its 19F NMR spectrum, respectively for o-, m- and
p-F atoms: a doublet at δ ‒156.14 ppm, and two triplets at δ ‒164.58 and ‒172.65 ppm. The high-field
19F chemical shifts observed in 52, compared to those in the parent amine (δ –154.42, –163.19 and –
163.48 for o-, p- and m-F atoms, respectively), are consistent with the accumulation of a negative charge
on the Namide atom on passing from HN(C6F5)2 to 52 where bonding is predominantly ionic.
To assess the possibility of combining the fluoroarylamine with a larger alkaline earth metal, the
stoichiometric reaction of [Ba{N(SiMe3)2}2]2 with HN(C6F5)2 afforded the dinuclear barium complex
[Ba{µ-N(C6F5)2}{N(C6F5)2}.toluene]2 (62). This complex is moderately soluble in petroleum ether and
aromatic hydrocarbons. The presence of one coordinated toluene molecule per Ba2+ (which could not be
removed, even by heating under dynamic vacuum) was established by 1H and 13C{1H} NMR in benzene-
d6 and by XRD analysis. Coordination of toluene testifies to the electrophilic nature of the Ba2+ ions in
complex 62. Its 19F NMR spectrum recorded at 25 °C show broad resonances of expected intensities for
o-, m- and p-F atoms centred on δ ‒157.66, ‒165.85 and ‒176.06 ppm, close to those of 52.
‒
The heteroleptic 52, with its basic N(SiMe3)2 amide (pKa = 25.8 in thf[19]), was used as a precursor
to other heteroleptic species. NMR monitoring of the reaction of 52 with the β-diketimine {BDIDiPP}H
(2 equiv) in toluene-d8 showed quantitative formation of the dinuclear [{BDIDiPP}CaN(C6F5)2]2 (72) upon
selective release of HN(SiMe3)2 (Scheme 2). The reaction subsequently performed on larger scales
1
afforded the formally three-coordinate 72 as a colourless solid in reproducible 80-90% yields. Its H
NMR spectrum recorded in toluene-d8 showed the formation of a pure product, characterised notably by
a singlet at δ 4.77 ppm diagnostic of the {BDIDiPP}‒ backbone methine CH hydrogen atom, and the
‒
absence of residual N(SiMe3)2 . The 19F NMR spectrum of the complex exhibits a doublet at δ ‒158.11
and two triplets at δ ‒164.85 and ‒175.62 ppm for o-, m- and p-F atoms, respectively. Complex 72 can
be seen as the derivative of [{BDIDiPP}CaN(SiMe3)2] (3).[8] It is more easily accessed than this latter
complex and in greater yields, but the fluoroarylamide in 72 is less basic and less reactive than the
–
N(SiMe3)2 residue in 3. In a similar way to 72, the dinuclear complex [{N^NDiPP}CaN(C6F5)2]2 (82) was
obtained near-quantitatively by reacting 52 with two equiv of the iminoaniline {N^NDiPP}H. As for 72,
with which it shares comparable organisation in the solid state (vide infra), the 1H and 19F NMR spectra
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
of 82 recorded at 25 °C in toluene-d8 confirms the presence of a single environment. Well-resolved
diagnostic resonances are located at δ 7.92 ppm in the 1H spectrum for the CH=N hydrogen atom, and
at δ –158.11, –164.78 and –175.95 ppm (for o-, m- and p-F atoms, respectively) in the 19F NMR
spectrum. Complexes 72 and 82 are only moderately soluble in petroleum ethers and toluene.
Scheme 2. Synthesis of fluoroarylamide-containing [{BDIDiPP}CaN(C6F5)2]2 (72), [{N^NDiPP}CaN(C6F5)2]2 (82),
and [Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 (92), with representation of their molecular solid-state structures.
It has recently been established that boryloxides could yield barium complexes of the type
[Ba(OBR2)2] with coordination numbers as low as two.[3d] In an attempt to extend this reactivity to
calcium and combine it with fluoroarylamides, the colourless [Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2
(92) was isolated in high yield following the stoichiometric reaction of 52 with the borinic acid
HOB{CH(SiMe3)2}2. Complex 92 is the first reported example of calcium boryloxide.[20] It is only mildly
1
soluble in hydrocarbons, but dissolves upon heating. Its H NMR spectrum in benzene-d6 is
characterised by two sets of resonances of matching intensities for CH(SiMe3)2 and CH3 hydrogens in
19
non-equivalent boryloxides: one at δ 0.21 and 0.15 ppm, and the other at 0.20 and 0.05 ppm. The F
NMR spectrum also exhibits two sets of equal intensities for o-, m- and p-F atoms: at δ –155.79 (d, 8F),
–163.59 (t, 8F) and –173.47 (t, 4F) ppm for one, and at δ –158.73 (d, 8F), –164.74 (t, 8F) and –174.45
(t, 4F) ppm for the other. Typically for this boryloxide,[3d] a single resonance at δ 53.71 ppm is seen in
the 11B NMR spectrum. This pattern agrees with the C2-symmetry axis seen in the solid-state structure
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
of the complex, with two distinct pairs of C6F5 groups due to hindered rotation along the Ca–N axes,
although an dynamic association-dissociation of the dimer in solution cannot be ruled out.
Structural studies – X-ray diffraction crystallography
The molecular solid-state structure of [Ca{N(C6F5)2}2.(Et2O)2] (4’) determined by XRD analysis is
depicted in Figure 3. The coordination sphere comprises two NC6F5 and two Oether atoms, and the metal
centre rests in a distorted tetrahedral environment. The CaNC6F5 bond lengths in 4’ (2.370(1) and
2.380(1) Å) are substantially longer than the corresponding ones in the related bis(amido) complex
[Ca{N(SiMe3)2}2.(thf)2]2 (2.294(3) and 2.309(3) Å).[21] This reflects the greater electron withdrawing
‒
ability of the N(C6F5)2 moiety and, as a result, the greater ionicity on the CaN bond in 4’. The
electrophilicity of the calcium atoms in 4’ is partly compensated by CaOether bonds (2.340(1) and
2.342(1) Å) stronger than in [Ca{N(SiMe3)2}2.(thf)2]2 (2.369(3) and 2.385(3) Å). In addition, further
electronic density is provided to the metal by four Ca···FC intramolecular interactions in the range
2.493(1)-3.074(1) Å. Considering that the accepted limit for Ca···F contacts is commonly set at 3.13
Å,[22] two of them, at 2.493(1) and 2.523(1), can be regarded as extremely strong. They for instance
compare well with those found in an unusual homoleptic β-diketiminato calcium complex bearing CF3
substituents in the backbone of the η1-bonded ligands.[12a] These Ca∙∙∙F‒C interactions may also be a
factor to the long Ca‒NC6F5 bond length.
Figure 3. ORTEP representation of the molecular solid-state structure of [Ca{N(C6F5)2}2.(Et2O)2] (4’). Ellipsoids
at the 50% probability level, with only the main component of the disordered Et2O molecule. H atoms omitted for
clarity. Ca···F interactions depicted in dashed bonds. Selected bond lengths (Å) and angles (deg): Ca1-O61 =
2.3402(12), Ca1-O71 = 2.3421(12), Ca1-N31 = 2.3705(14), Ca1-N1 = 2.3805(14), Ca1-F23 = 2.4935(10), Ca1-
F38 = 2.5226(10), Ca1-F53 = 2.7713(11), Ca1-F8 = 3.0740(12); O61-Ca1-O71 = 87.71(4), O61-Ca1-N31 =
100.65(5), O71-Ca1-N31 = 144.30(5), O61-Ca1-N1 = 143.24(5), O71-Ca1-N1 = 98.38(5), N31-Ca1-N1 =
95.23(5).
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
The solid-state structure of [Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52), with its crystallographic C1
symmetry, shows a dinuclear arrangement (Figure 4). The more electron-rich hexamethyldisilazide
‒
groups N(SiMe3)2 occupy the bridging positions. The corresponding CaNHMDS bond lengths, in the
range 2.404(2)-2.439(2) Å, are shorter than their homologues in [Ca{µ2-N(SiMe3)2}{N(SiMe3)2}]2,
where they are found between 2.431(7) and 2.521(7) Å.[23] The distances to the terminal NC6F5 groups,
at 2.385(2) and 2.383(2) Å, are comparable to those in 4’. Taking the three N atoms into account, each
calcium in 52 lies in a perfect trigonal planar geometry (Σθ(Ca1) = 359.95 ° and Σθ(Ca2) = 359.90 °).
Besides, each also engages in a strong Ca···FC interaction, as indicated by short Ca1F43 and
Ca2F62 distances (2.423(2) and 2.456(2) Å). Complex 52 is a rare example of charge-neutral three-
coordinate calcium complex free of coordinated solvent; other representative examples include
e
[Ca{N(SiMe3)2}2]2 and [{BDIDiPP}CaN(SiMe3)2] (3) already mentioned,[6 ,8,23] as well as [Ca(O-tBu2-
2,6-C6H3)2]2,[24] the N-heterocyclic carbene adducts [Ca{N(SiMe3)2}2.NHCaryl] (aryl = mesityl or 2,6-
iPr2C6H3),[25] and the imine adduct [Ca{N(SiMe3)2}2.PhCH=NtBu].[26,27]
Figure 4. ORTEP representation of the molecular solid-state structure of [Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52).
Only one of the two independent but similar molecules in the asymmetric unit represented. Ellipsoids at the 50%
probability level. H atoms omitted for clarity. Ca···F interactions depicted in dashed bonds. Selected bond lengths
(Å) and angles (deg): Ca1-N21 = 2.385(2), Ca1-N2 = 2.404(2), Ca1-F43 = 2.4227(18), Ca1-N1 = 2.433(2), Ca2-
N51 = 2.383(2), Ca2-N2 = 2.411(2), Ca2-N1 = 2.439(2), Ca2-F62 = 2.4560(18); N21-Ca1-N2 = 125.47(8), N21-
Ca1-N1 = 145.63(8), N2-Ca1-N1 = 88.85(8), F43-Ca1-N1 = 85.00(7), N51-Ca2-N2 = 125.06(8), N51-Ca2-N1 =
145.99(8), N2-Ca2-N1 = 88.55(8).
9
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
The centrosymmetric barium complex [Ba{µ-N(C6F5)2}{N(C6F5)2}.toluene]2 (62) is dinuclear
(Figure 5). The BaN interatomic distances to terminal and bridging N-atoms, respectively of 2.740(3)
and 2.871(4)-2.951(3) Å, are longer than in the related [Ba{µ-N(SiMe3)2}{N(SiMe3)2}]2 (2.576(3) and
‒
2.798(3)-2.846(4) Å).[28] This again results from lower electron availability in N(C6F5)2 and greater
ionic bonding in 62 due to the presence of very electron-withdrawing substituents on the N-atoms. In
addition to the fluoroarylamides, the coordination sphere of each Ba2+ cation in 62 is completed by a
capping η6-coordinated toluene molecule, with a Ba-centroid distance of 3.065 Å and corresponding Ba-
Ctoluene distances between 3.287(5) and 3.436(6) Å. Coordinative unsaturation is further relieved by five
close Ba···FC contacts, with BaF distances in the range 2.825(2)-3.014(3) Å. The formation of a
dinuclear species, the high number of Ba···FC contacts (e.g. compared to 4’ and 52) and additional
coordination of toluene molecules are seen as the reflection of the great electrophilicity of the Ba2+
cations in 62. Ba···Cπ interactions are common in complexes of barium, a metal softer than calcium.[9,29]
Ba···FC interactions are known to stabilise electron-deficient barium complexes. The intramolecular
BaF distances are for instance of 2.871(2) and 2.901(3) Å in [(thf)2Ba{N(H)-2,6-F2-C6H3}]∞,[12b] while
they range from 2.918(2) to 2.992(3) Å in discrete monocationic complexes bearing multidentate
aminoether-fluoroalkoxides.[4b,30]
Figure 5. ORTEP representation of the molecular solid-state structure of [Ba{µ-N(C6F5)2}{N(C6F5)2}.toluene]2
(62). Ellipsoids at the 50% probability level. H atoms omitted for clarity. Ba···F interactions depicted in dashed
bonds. Selected bond lengths (Å): Ba1-N1 = 2.740(3), Ba1-F11 = 2.825(2), Ba1-F32 = 2.832(2), Ba1-N2 =
2.871(4), Ba1-F2 = 2.910(3), Ba1-F22iv = 2.923(3), Ba1-N2iv = 2.951(3), Ba-F22 = 3.014(3).
10
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
The solid-state structure of [{BDIDiPP}CaN(C6F5)2]2 (72) is displayed in Figure 6. It shows the
complex to exist as a dimer in the solid state through the presence of two intermolecular Ca···F‒C
contacts with m-F atoms carried by a neighbouring C6F5 moieties. The dimer possesses a
crystallographic C2 symmetry axis that goes through the two calcium atoms. Compound 72, where the
metal centres are formally three-coordinate, is most closely related to [{BDIDiPP}CaN(SiMe3)2] (3): this
complex, with three-coordinate metal centres, forms tetramers in the solid phase through intermolecular
Ca···H3C interactions, and also displays intramolecular Ca···H3C contacts.[8] The CaN interatomic
distances to NBDIiPP-atoms in the bidentate ligand in 72 (2.306(2) and 2.346(2) Å) match those in 3
(2.323(3) and 2.331(3) Å). On the other hand, the CaNC6F5 bond length in 72 is noticeably longer than
the corresponding CaNHMDS one in 3 (2.388(2) vs 2.299(3) Å), showing that the more basic and
electron-rich amide induces a tighter bond with calcium. Albeit weaker than in 4’ and 52, the
intramolecular (2.556(1), and 2.712(2) Å) and intermolecular (2.532(1) Å) Ca···FC interactions in 72
remain on the whole particularly strong in regard of those found in many other calcium complexes.[12]
The dinuclear configuration is held together solely through the intermolecular Ca···FC contacts. The
two C6F5 groups located between the two Ca+2 ions are parallel, but this is due to the symmetry in 72;
there is no substantial π-stacking interaction between them, as they are in offset positions.
Figure 6. ORTEP representation of the molecular solid-state structure of [{BDIDiPP}CaN(C6F5)2]2 (72). Ellipsoids
at the 50% probability level. H atoms and non-interacting benzene molecule omitted for clarity. Ca···F interactions
depicted in dashed bonds. Selected bond lengths (Å): Ca1-N1 = 2.3066(19), Ca1-N2 = 2.3465(18), Ca1-N3 =
2.3880(19), Ca1-F40i = 2.5325(15), Ca1-F31 = 2.5559(15), Ca1-F37 = 2.7120(17).
11
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Perhaps unsurprisingly on account of the similarities of their ligand backbones, the molecular structure
of [{N^NDiPP}CaN(C6F5)2]2 (82) is very similar to that of 72; it is given in the Supporting Information,
without further discussion here.
The heteroleptic boryloxide/fluoroarylamide complex [Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 (92)
is dinuclear in the solid state, with formally three-coordinate calcium centres surrounded by boryloxides
in bridging positions and amides in terminal ones (Figure 7). This is the opposite situation of that in the
‒
barium complex [Ba{µ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2, where N(SiMe3)2 amides, more electron-
‒
rich than N(C6F5)2 , occupy the bridging positions.[3d] Hence, the overall electron availability and
‒
‒
donating ability in these monoanionic ligands increases with N(C6F5)2 < OB{CH(SiMe3)2}2
<
‒
N(SiMe3)2 . Delocalisation of N-electrons into antibonding σ(SiC*) orbitals that takes place in
‒
N(SiMe3)2 is of limited effect compared to partial delocalisation of O-electrons into the empty pz orbital
‒
at boron in the boryloxide, while the presence of two strongly electron-withdrawing groups in N(C6F5)2
bears the strongest effect on electron-sharing properties.
Figure 7. ORTEP representation of the molecular solid-state structure of [Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2
(92). Ellipsoids at the 50% probability level. H atoms omitted for clarity. Ca···F interactions depicted in dashed
bonds. Selected bond lengths (Å): Ca1-O2 = 2.269(2), Ca1-O1 = 2.271(2), Ca1-N1 = 2.370(3), Ca1-F32 =
2.544(2), Ca1-F38 = 2.597(2), Ca2-O1 = 2.243(2), Ca2-O2 = 2.259(2), Ca2-N2 = 2.356(3), Ca2-F50 = 2.498(2),
Ca2-F44 = 2.553(2); O2-Ca1-O1 = 82.58(8), O2-Ca1-N1 = 138.45(9), O1-Ca1-N1 = 124.32(10), O1-Ca2-O2 =
83.40(8), O1-Ca2-N2 = 143.23(10), O2-Ca2-N2 = 126.74(9).
12
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
By contrast with 52, where each three-coordinate Ca2+ is in a trigonal planar geometry, the arrangement
around Ca1 and Ca2 in 92 is irregular (Σθ(Ca1) = 345.35 ° and Σθ(Ca2) = 353.37 °, taking N and O atoms
into account). This may reflect high steric congestion around the metals, or can also be the consequence
of two Ca···FC interactions per metal, as opposed to one in 52. The strength of the various Ca···FC
interactions in 92, with interatomic distances of 2.544(2) and 2.597(2) for Ca1 and 2.498(2) and 2.553(2)
for Ca2, appears to be weaker than in 52. The geometry around the boron, oxygen and nitrogen atoms
approaches perfect trigonal planarity in all cases (Σθ(B1) = 359.90 ° and Σθ(B2) = 359.90 °; Σθ(O1) =
359.96 ° and Σθ(O2) = 359.96 °; Σθ(N1) = 359.80 ° and Σθ(N2) = 359.90 °).
Bond valence sum analysis
The contribution of Ca···FC secondary interactions towards the coordination sphere of the Ca2+ ions
in our low-coordinate [Ca]-N(C6F5)2 complexes was gauged using bond valence sum (BVS) analysis.
This empirical treatment, pioneered by Brown and further developed by O’Keefe and Brese,[17] makes
use of interatomic distances measured by single-crystal X-ray diffraction to evaluate the relative
contributions of each neighbouring atom towards the coordination sphere of a given metallic centre. It
has been applied in recent years to analyse the bonding patterns of a variety of s-block complexes, in
particular to evaluate the role of non-covalent interactions.[12d,31] For each complex, bond valences (ν)
were calculated using eq 1, with experimental bond lengths dCa-X and tabulated (empirical) bond valence
parameters RCaX
:
ν = exp[(RCa–X – dCa−X)/B]
(1)
in which X = O, F, N; B = 0.37; RCaO = 1.967, RCaF = 1.842 and RCaN = 2.140.[17] The bond valence
value, ν, provides a quantitative appraisal of the contributions of secondary Ca···FC interactions to the
global coordination sphere of each Ca2+ cation. The calculated sums of bond valence values (Σ(νCa–X
)
for X = N/F/O) for each Ca2+ in the calcium complexes 4’, 52, 72 and 92 are collated in Table 1. As
anticipated for a divalent cation, the values approach 2, although the deviations observed in 52 and 92
may suggest that BVS analysis is not ideally suited to these very coordinatively unsaturated species.
Qualitative analysis of the data indicates that Ca···FC interactions contribute to around 15% of
the total bonding pattern, and culminate in the value of 20% estimated for complex 4’. The remaining
of the coordination sphere is filled out by CaN and, where relevant, CaO bonds. In the dinuclear
complexes in which the two Ca2+ centres are not identical (that is, 52 and 92), the bonding patterns remain
very comparable around the two cations. The contributions of Ca···FC interactions to the bonding in
4’, 52, 72 and 92 outclass that established for a dinuclear complex [{ROF}CaN(SiMe3)2]2 where each
five-coordinate Ca2+ was surrounded by an η3-coordinated aminoether-fluoroalkoxide {ROF}‒, and in
which a single Ca···FC contact counted for ca. 4% of the coordination sphere around each ion.[12d] This
translates the growing importance of secondary interactions as the coordination number decreases.
13
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Table 1. Bond valence sum analysis for the calcium complexes 4’, 52, 72 and 92
a
a
a
b
b
b
Complex
Σ(νCa–X)
dCaN
dCaF
dCaO
νCaN
νCaF
νCaO
[Ca{N(C6F5)2}2.(Et2O)2] (4’)
One 4-coordinate Ca atom
Ca1
2.380
2.370
-
-
2.493
2.523
2.771
3.074
2.340
2.342
-
-
0.52
0.54
-
0.17
0.16
0.08
0.04
0.45
20
0.36
0.36
-
-
-
Σ(νCa1–X)
%(νCa1–X)
1.06
47
0.73
33
2.24
100%
[Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52)
Two distinct 3-coordinate Ca atoms
Ca1
Ca2
2.433
2.404
2.385
2.423
-
-
-
-
-
0.45
0.49
0.52
1.46
87
0.45
0.48
0.52
1.44
88
0.21
-
-
0.21
13
0.19
-
-
-
-
-
-
-
-
-
-
-
Σ(νCa1–X)
%(νCa1–X)
1.67
100
2.439
2.411
2.383
2.456
-
-
-
-
-
-
Σ(νCa2–X)
%(νCa2–X)
0.19
12
1.63
100
[{BDIDiPP}CaN(C6F5)2]2 (72)
Two identical 3-coordinate Ca atoms
Ca1
2.307
2.346
2.388
2.532
2.556
2.712
-
-
-
0.64
0.57
0.51
1.72
81
0.15
0.14
0.09
0.39
19
-
-
-
-
-
Σ(νCa1–X)
%(νCa1–X)
2.11
100
[Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 (92)
Two distinct 3-coordinate Ca atoms
Ca1
Ca2
2.370
-
2.544
2.597
2.271
2.269
0.54
-
0.54
32
0.56
-
0.56
31
0.15
0.13
0.28
16
0.17
0.15
0.32
18
0.44
0.44
0.88
52
0.47
0.45
0.93
51
Σ(νCa1–X)
%(νCa1–X)
1.70
100
2.356
-
2.498
2.553
2.243
2.259
Σ(νCa2–X)
%(νCa2–X)
1.81
100
2.356
a Measured interatomic distances to X = N, F or O atoms, given in Å; see Figures 3, 4, 6 and 7 for detail. b Bond valence contribution for atom X. c Sum of all contributions
for the different atoms: X = N, F and O.
14
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
It can also be seen from Table 1 that the relative contributions to the bonding pattern around Ca2+ is
directly proportional to the number of Ca···FC contacts detected in the solid state. Although caution
must be applied in the appreciation of quantitative BVS analysis with these electrophilic complexes,
there can be little doubt that Ca···FC interactions bear a great influence in their overall stability. The
structural data and BVS analysis for these compounds can be added to the growing body of evidences
that testifies to the significance of secondary interactions in the coordination chemistry of s-block
metals.[1,9]
Bonding analysis by DFT calculations
DFT calculations were performed on series of model systems at the PBE0/Def2-TZVP and PBE0/Def2-
TZVP-D3 levels (see Computational details in the SI) to shed some light on the nature and strength of
the Ae···FC bonding in our complexes. Since, in these compounds, both Ae and F belong to the same
molecule, one might have to consider the potential role of the pinch intramolecular effect that could
force the two atoms to get close to each other. In a first approach, we ruled out this possibility by
investigating bi-molecular systems, i.e. the interaction between one F-containing molecule or ion,
namely C6H5F, CH3F and F–, and a simplified calcium complex. Since most of the molecular structures
described above exhibit an Ae2+ ion in a trigonal planar coordination mode (not considering the
Ae···FC interactions), the model complex chosen for this initial study was the simple
[{NH(CH)3NH}Ca(NH2)] (A). The optimised geometries of the three A(C6H5F), A(CH3F) and AF–
bimolecular systems are shown in Figure 8. Optimisations were performed with and without considering
the 3-parameter Grimme’s empirical corrections (D3)[32] for dispersion forces. Such corrections are
expected to take satisfyingly into account the van der Waals component of the Ca···F bonding. The
CaF distances obtained with and without D3 corrections are reported at the top of Figure 8 in black
and blue colours, respectively. A weak shortening of the CaF distances (lower than 1%) occurs when
introducing dispersion forces, indicating that van der Waals bonding does not play a crucial role in the
interaction. The CaF distance is only slightly shortened when going from A(C6H5F) to A(CH3F) and
Ca remains in an approximate planar CaN3 coordination. The AF‒ structure is different, with a very short
CaF distance (2.09 Å) and a pyramidalised (sp3) Ca atom. Clearly, AF‒ exhibits strong Ca···F bonding.
A more detailed understanding of the nature of the Ca···F(C) bonding in the above series is
obtained by an energy decomposition analysis (EDA) of the interaction between two fragments,
according to the Morokuma-Ziegler procedure.[33] The decomposition of the total bonding energy (TBE)
between A and the C6H5F, CH3F and F– fragments in the three optimised systems of Fig. 8 is provided
in Table 2 (D3 corrections considered). TBE is expressed as the sum of four components, the Pauli
repulsion (EPauli), the electrostatic interaction energy (Eelstat), the orbital interaction energy (Eorb) and the
component associated with the dispersion forces (Edisp). Unsurprisingly, EPauli increases with a decrease
of the Ca···FC separation. It is overbalanced by the three other stabilizing components, of which Eelstat
15
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
prevails, indicating predominance of ionic bonding. In A(C6H5F), the Eorb component is about half of
Eelstat, indicating weak, although non-negligible, covalent interactions. Consistently with the above
discussion on bond distances, the Edisp contribution is minor, corroborating weak van der Waal
interactions. While the EDA analysis of A(CH3F) provides rather similar values and conclusions as those
for A(C6H5F), different results were obtained for AF‒. Clearly, this other model contains a real and strong
CaF single bond of ionocovalent nature (note the significant Eorb component), whereas in the two other
models, the Ca···FC interaction is about six-fold weaker.
Figure 8. Optimised geometries of the A(C6H5F), A(CH3F), AF‒, B(C6H5F), B(CH3F) and BF‒ models. Interatomic
distances are given in Å; in black, with D3 corrections; in blue, without D3 corrections.
Table 2. Morokuma-Ziegler energy decomposition analysis in the model compounds 7’2 and those derived from
[{NH(CH)3NH}Ca(NH2)] (= A) and [{NH(CH)3NH}2Ca] (= B). a
A(C6H5F) A(CH3F)
AF–
A(CH4) A(C6H6) A(C2H4) B(C6H5F) B(CH3F)
BF–
7’2
0.42
0.50
2.14
0.98
0.29
0.48
0.52
0.39
0.47
2.16
EPauli
Eelstat
Eorb
‒0.53
‒0.28
‒0.15
‒0.55
‒0.71
‒0.28
‒0.11
‒0.60
‒4.46 ‒0.95
‒1.31 ‒0.62
‒0.02 ‒0.80
‒3.65 ‒1.39
‒0.24
‒0.19
‒0.09
‒0.24
‒0.53
‒0.41
‒0.21
‒0.66
‒0.54
‒0.27
‒0.11
‒0.43
‒0.44
‒0.21
‒0.17
‒0.43
‒0.61
‒0.22
‒0.11
‒0.48
‒4.22
‒1.43
‒0.02
‒3.51
Edisp
TBE b
a all values in eV. b Total bonding energy = EPauli + Eelstat + Eorb + Edisp
.
The weak Ca···FC covalent contribution in A(C6H5F) and A(CH3F) is in line with the natural
atomic orbital (NAO) population analysis. In the two models, the computed F→Ca charge transfer is
virtually nil (lower than 10–2), whereas it is 0.07 in AF‒. Surprisingly, the fluorine charge in A(C6H5F)
and A(CH3F) (‒0.36 and ‒0.40, respectively) is more negative than in the free C6H5F and CH3F ligands
16
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
(‒0.31 and ‒0.35, respectively). This is diagnostic of strong electrostatic interaction with Ca, which
attracts electron density towards the fluorine, but without any significant F→Ca electron transfer due to
the poor bond covalence.
Since in the solid state, compounds 72 and 82 are dimers of mononuclear three-coordinate Ca
complexes paired through intermolecular Ca···FC contacts, an EDA analysis of the interaction
between the two monomers was also carried out. Compound 72 was slightly simplified by replacing its
Dipp and methyl substituents by hydrogen atoms. The geometry of this model (7’2) was fully optimised
at the PBE0/Def2-TZVP-D3 level. It showed the presence of Ca···FC intermolecular contacts with
interatomic distances of 2.584 and 2.612 Å, whereas the intramolecular contact was 2.581 Å. The EDA
data obtained for 7’2 on the basis of two monomeric fragments are given in Table 2. Considering that
there are four inter-monomer Ca···FC contacts (but somewhat longer that in A(C5H5F), see below and
Table 3), and that other van de Waals contacts are also present (see the structure of 72 in Fig. 6), the
interaction energy component of 7’2 are fully consistent with that of the simple A(C5H5F) model.
In the above-computed models, the calcium atom lies in an approximate trigonal planar geometry
(except for AF-), thus having its accepting 3pz orbital essentially pointing in the direction of the electron-
rich fluorine atom. A point of interest is the importance of the (weak) covalent interaction when calcium
is tetravalent, being in a distorted sp3 hybridisation. To probe this aspect, three other models, namely
B(C6H5F), B(CH3F) and BF in which B is [{NH(CH)3NH}2Ca], were investigated. Their optimised
geometries are shown in Figure 8 (bottom). On going from A to B systems, only a small lengthening of
the CaF distances is noticed. Consistently, the EDA components of the various B models do not differ
much from that of their A homologues. Thus, on switching from a sp2 to a sp3 calcium, its (moderate)
accepting properties do not vary significantly, owing to its propensity for hypervalence. In the weakly
bonded B(C6H5F) and B(CH3F), the Ca-centred CaN4 polyhedron maintains its distorted tetrahedral
configuration. On the other hand, BF‒ is structurally different from the two other B systems. With a real
CaF single bond, Ca is in a hypercoordinated square-pyramidal environment in BF.
The real intramolecular Ca···FC interactions detected in the molecular structures of compounds
4’, 52, 72 and 92 are also suitably reproduced in the structure of the fully optimised molecules (dispersion
forces included, Table 3). There is however one noticeable exception, namely the long Ca···FC contact
(3.074(12) Å) observed in the X-ray structure of 4’, which is found to be much shorter (2.769 Å) in the
nearly C2-symmetric optimised structure. This discrepancy is attributed to the packing interactions in
the solid state, which distort the molecule away from ideal C2 symmetry. The smaller deviation of ca.
0.10 Å found for the inter-monomer contact in 72 is likely the result of the simplification made on going
from 72 to the less crowed 7’2 model. As a whole, these data are fully consistent with those obtained in
our simple intermolecular models, providing full confidence on the validity of our analysis of the
bonding scenario. The above-observed electro-attracting effect of calcium onto the fluorine without any
noticeable F→Ca electron transfer is also present in these compounds. With a NAO charge comprised
17
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
between ‒0.30 and ‒0.38, all fluorine atoms involved in Ca···FC bonding are more negatively
polarised that the other fluorine atoms, the NAO charge of which being comprised between ‒0.27 and ‒
0.29. The major ionic character of the interaction is also consistent with the large positive charge
computed for calcium in these compounds (between +1.81 and +1.84).
Table 3. Comparison of experimental and computed Ca···FC interatomic distances for the calcium
complexes 4’, 52, 72 and 92
dCaF (XRD) a
dCaF (DFT) b
Complex
[Ca{N(C6F5)2}2.(Et2O)2] (4’)
One 4-coordinate Ca atom
Ca1
2.4935
2.5226
2.7713
3.0740
2.553
2.552
2.761
2.769
[Ca{µ2-N(SiMe3)2}{N(C6F5)2}]2 (52)
Two distinct 3-coordinate Ca atoms
Ca1
Ca2
Ca1
2.4227
-
-
2.477
2.477
2.4560
-
-
[{BDIDiPP}CaN(C6F5)2]2 (72)
Two identical 3-coordinate Ca atoms
2.5325
2.5559
2.7120
2.581 c
2.584 c
2.612 c
[Ca{µ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 (92)
Two distinct 3-coordinate Ca atoms
Ca1
Ca2
2.544
2.597
2.549
2.637
2.498
2.553
2.551
2.583
a Measured interatomic distances to X = N, F or O atoms, given in Å; see Fig. 3, 4, 6 and 7 for detail.
b PBE0/Def2-TZVP-D3-optimised distances.
c Values calculated on the simplified 7’2 model (see above).
Owing to the dominant electrostatic nature of the Ca···FC interactions, the propensity for Ca to
also trigger Ca···HC non-ionic bonds raises the question of the nature of these interactions. Therefore,
for the sake of comparison, the non-fluorinated equivalent of A(CH3F), i.e. A(CH4), was also
investigated. Its optimised structure (Figure 9) exhibits a 3-coordinated methane, with one shorter
(2.621 Å) and two longer (2.817 Å) contacts. These values are commensurate with experimentally
determined agostic Ca···HC distances, e.g. 2.8933(6) Å in [Ca{N(SiMe3)(mesityl).(thf)2][34] and
2.7741(15) Å in [Ca{N(SiMe3)(mesityl).tmeda].[34] The EDA data (Table 2) indicate substantially
18
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
weaker bonding in the A(CH4) case (by more than 50%), but still with the same energy component order.
Thus, the Ca···HC interaction also has a major electrostatic and a weaker van der Waals character.
Interestingly, the optimised structure of the A(C6H6) model (Figure 9) does not exhibit any Ca···HC
contact, but rather features a weakly 6-coordinated benzene ligand (CaC = 3.054-3.165 Å). This
indicates prevailing Ca···C(π) over Ca···HC interactions, due to stronger covalent and van der Waals
bonding (see Table 2). The interaction of calcium with a single unsaturated C=C bond is exemplified in
the A(C2H4) model (Figure 9).The CaC distances (2.929 and 2.974 Å) are shorter than in A(C6H6). Its
EDA values (Table 2) are very close to those of A(CH3F), except for Eelstat which indicates, as
anticipated, weaker ionic bonding. Thus, amongst the various weak Ca···X (X = F, H, C()) interactions
reported in the literature, our computations suggest that the Ca···FC one appears to be the strongest,
due to the high positive charge of Ca2+ which in any case favours electrostatic interactions. This is in
agreement with our earlier experimental findings, using structurally characterised alkyl-substituted
complexes that display a range of Ca···FC and Ca···C(π) interactions (as well as Ca···HSi agostic
contacts).[11a,11b,12e]
Figure 9. Optimised geometries of the A(CH4), A(C6H6) and A(C2H4) models. Interatomic distances are given in
Å; in black, with D3 corrections; in blue, without D3 corrections.
Concluding remarks
Structural analysis of the several new [Ae]-N(C6F5)2 complexes prepared herein for Ae = Ca and Ba
demonstrate that Ae···FC interactions are a convenient synthetic tool for the preparation of low
coordinate complexes of the alkaline earths. The case of [{BDIDiPP}CaN(C6F5)2]2 (72), where each Ca2+
ion is formally three-coordinate, is particularly illustrative. Unlike the benchmark congener
[{BDIDiPP}CaN(SiMe3)2.(thf)], 72 does not contain coordinated thf; this would in most cases be
beneficial for applications in catalysis. Besides, 72 is more readily accessed than the thf-free analogue,
[{BDIDiPP}CaN(SiMe3)2] (3), which has proved to be a most valuable synthetic precursor.[2j,6e,8] The
molecular structures of the new complexes, with CaF interatomic distances around 2.50 Å or even
shorter, indicate strong Ca···FC intramolecular interactions, sufficiently so to allow the smooth
formation of, for instance, compounds such as thf-free 72 and 92. Complementary DOSY analysis
19
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
performed in benzene-d6 at 25 °C (SI) suggests that intermolecular Ca···FC interactions, and hence
the dinuclear arrangement of these complexes, are retained in solution.
Although it is hence a useful synthetic tool in alkaline earth organometallic chemistry, the main
‒
drawback of the N(C6F5)2 amide is its low basicity compared to that of more classical amides, e.g.
‒
‒
N(SiMe3)2 or even N(SiMe2H)2 . This limits the utilisation of [Ae]-N(C6F5)2 complexes in molecular
catalysis to the reactions where substrates are relatively acidic, e.g. phosphines in hydrophosphination,
or where Brønsted acidity considerations are of little importance. This is illustrated for instance by the
performances displayed by 72 in the hydroelementation catalysis. It proved inactive in the attempted
intermolecular hydroamination of styrene with benzylamine, and in cyclohydroamination of 2,2-
dimethylpent-4-en-1-amine, i.e. two reactions that [{BDIDiPP}CaN(SiMe3)2.(thf)] catalyses very
effectively.[2a,2d] On the other hand, 72 is an efficient precatalyst for the hydrophosphination of styrene
with HPPh2, affording near-quantitative formation of the anti-Markovnikov product of addition
PhCH2CH2PPh2 under experimental conditions (60 °C, 12 h, 10 mol-% precatalyst) that compare
favourably with those required by the ubiquitous [{BDIDiPP}CaN(SiMe3)2.(thf)] (75 °C, 20 h, 10 mol-%
‒
Ca).[35] On the whole, it can be concluded that although it is a convenient synthetic tool, N(C6F5)2 is
overall not ideally suited to Ae-mediated catalysis.
‒
In view of these considerations, we have considered using the more basic N(o-F-C6H4)2 as another
stabilising, yet more reactive fluorinated arylamide. Preliminary results are encouraging. Following
similar protocols to those implemented here, we have for instance been able to obtain [{BDIDiPP}CaN(o-
F-C6H4)2.(thf)], a mononuclear complex that displays two strong intramolecular Ca···FC contacts.[36]
However, although we have spectroscopic evidence for its formation, we have not yet been able to
crystallise the thf-free derivative [{BDIDiPP}CaN(o-F-C6H4)2]n that is required to carry out pertinent
comparative catalytic studies. We have also been able to reproducibly isolate small crops of crystals of
the tetranuclear [Ca4{µ2-N(o-F-C6H4)2}6{N(SiMe3)2}2],
a
by-product in the reaction of
[Ca{N(SiMe3)2}2]3 with HN(o-F-C6H4)2. This unusual compound, certainly reminiscent of the trinuclear
[Ca{N(SiMe2H)2}2]3,[10c] is also stabilised by very strong Ca···FC interactions (SI), but the poor quality
of the data and refinement preclude an informative discussion of metric parameters. The identity of the
other compounds in this reaction have not yet been established, but optimisation of other reactions with
HN(o-F-C6H4)2 and related amines are still in progress in our laboratory and will be reported elsewhere.
DFT calculations have reproduced faithfully the molecular structures observed experimentally. A
detailed bonding analysis of the Ca···FC interactions indicate that its electrostatic component is largely
predominant over of the covalent one, whereas dispersion forces are merely marginal. The Ca2+ ion is
not an efficient electro-acceptor centre and no significant F→Ca charge transfer occurs in the Ca···FC
interactions. On the other side, its highly charged nature confers it the ability of attracting substantial
electronic density towards the interacting F atoms. Importantly, upon comparison of the relative strength
of the three types of secondary interactions Ca···F(C), Ca···H(C) and Ca···C(π) in model calcium
20
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
complexes, it emerges clearly that Ca···FC is the strongest one; this is consistent with our experimental
observations made in heteroleptic calcium fluoroaminoalkoxides.[11a,11b,12e] The combination of
experimental and computational data presented here is an additional contribution to previous works that
demonstrates the potential offered by non-covalent interactions in alkaline earth chemistry. We believe
it is likely to open new opportunities to synthetic chemists with an imaginative mind, for instance to
stabilise low-coordinated catalytic species.
Supporting Information
Full experimental details; NMR spectra and crystallographic data for CCDC 1895996-1896001 and
1900852-1900853; details of DFT computations.
Acknowledgements
H. R. is grateful to Erwann Le Coz for his assistance in the synthesis and crystallisation of complex 62.
The GENCI French national computer resource centre is acknowledged (grant A0050807367).
Conflicts of interest
The authors declare no conflict of interest.
Notes and references
[1] a) S. Harder, Chem. Rev. 2010, 110, 3852; b) M. R. Crimmin, M. S. Hill, Top. Organomet. Chem.
2013, 45, 191; c) M. S. Hill, D. J. Liptrot, C. Weetman, Chem. Soc. Rev. 2016, 45, 972; d) Y.
Sarazin, J.-F. Carpentier, Chem. Rec. 2016, 16, 2482.
[2] See for instance: a) M. R. Crimmin, I. J. Casely, M. S. Hill, J. Am. Chem. Soc. 2005, 127, 2042; b)
F. Buch, J. Brettar, S. Harder, Angew. Chem. Int. Ed. 2006, 45, 2741; c) J. Spielmann, F. Buch, S.
Harder, Angew. Chem. Int. Ed. 2008, 47, 9434; d) A. G. M. Barrett, C. Brinkmann, M. R. Crimmin,
M. S. Hill, P. Hunt, P. A. Procopiou, J. Am. Chem. Soc. 2009, 131, 12906; e) P. Jochmann, J. P.
Davin, T. P. Spaniol, L. Maron, J. Okuda, Angew. Chem. Int. Ed. 2012, 51, 4452; f) B. Liu, T.
Roisnel, J.-F. Carpentier, Y. Sarazin, Angew. Chem. Int. Ed. 2012, 51, 4943; g) 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; h) C. Qi, V. Gandon, D. Lebœuf, Angew. Chem. Int. Ed. 2018, 57, 14245;
i) H. Bauer, M. Alonso, C. Fischer, B. Rösch, H. Elsen, S. Harder, Angew. Chem. Int. Ed. 2018,
57, 15177; j) A. S. S. Wilson, C. Dinoi, M. S. Hill, M. F. Mahon, L. Maron, Angew. Chem. Int. Ed.
2018, 57, 15500.
[3] See for instance: a) J. Spielmann, G. Jansen, H. Bandmann, S. Harder, Angew. Chem. Int. Ed. 2008,
47, 6290; b) M. S. Hill, D. J. Liptrot, D. J. MacDougall, M. F. Mahon, T. P. Robinson, Chem. Sci.
21
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
2013, 4, 4212; c) C. Bellini, J.-F. Carpentier, S. Tobisch, Y. Sarazin, Angew. Chem. Int. Ed. 2015,
54, 7679; d) E. Le Coz, V. Dorcet, T. Roisnel, S. Tobisch, J.-F. Carpentier, Y. Sarazin, Angew.
Chem. Int. Ed. 2018, 57, 11747.
[4] a) S. Harder, F. Feil, K. Knoll, Angew. Chem. Int. Ed. 2001, 40, 4261; b) Y. Sarazin, B. Liu, T.
Roisnel, L. Maron, J.-F. Carpentier, J. Am. Chem. Soc. 2011, 133, 9069; c) C. Bellini, C. Orione,
J.-F. Carpentier, Y. Sarazin, Angew. Chem. Int. Ed. 2016, 55, 3744.
[5] See for instance: a) J. Spielmann, S. Harder, J. Am. Chem. Soc. 2009, 131, 5064; b) P. Jochmann,
T. S. Dols, T. P. Spaniol, L. Perrin, L. Maron, J. Okuda, Angew. Chem. Int. Ed. 2010, 49, 7795; c)
D. J. Liptrot, M. Arrowsmith, A. L. Colebatch, T. J. Hadlington, M. S. Hill, G. Kociok-Köhn, M.
F. Mahon, Angew. Chem. Int. Ed. 2015, 54, 15280; d) V. Leich, T. P. Spaniol, L. Maron, J. Okuda,
Angew. Chem. Int. Ed. 2016, 55, 4794; e) A. Causero, H. Elsen, J. Pahl, Sjoerd Harder, Angew.
Chem. Int. Ed. 2017, 56, 6906; f) B. M. Wolf, C. Maichle-Mössmer, R. Anwander, Angew. Chem.
Int. Ed. 2016, 55, 13893; g) M. D. Anker, C. E. Kefalidis, Y. Yang, J. Fang, M. S. Hill, M. F.
Mahon, Laurent Maron, J. Am. Chem. Soc. 2017, 139, 10036; h) X. Shi, C. Hou, C. Zhou, Y. Song,
J. Cheng, Angew. Chem. Int. Ed. 2017, 56, 16650; i) S. Brand, H. Elsen, J. Langer, W. A.
Donaubauer, F. Hampel, S. Harder, Angew. Chem. Int. Ed. 2018, 57, 14169; j) B. M. Wolf, C.
Stuhl, C. Maichle-Mꢀssmer, R. Anwander, J. Am. Chem. Soc. 2018, 140, 2373; k) D. Mukherjee,
T. Hꢀllerhage, V. Leich, T. P. Spaniol, U. Englert, L. Maron, J. Okuda, J. Am. Chem. Soc. 2018,
140, 3403; l) H. Bauer, K. Thum, M. Alonso, C. Fischer, S. Harder, Angew. Chem. Int. Ed. 2019,
58, 4248; m) X. Shi, G. Qin, Y. Wang, L. Zhao, Z. Liu, J. Cheng, Angew. Chem. Int. Ed. 2019, 58,
4356.
[6] a) S. Harder, J. Brettar, Angew. Chem. Int. Ed. 2006, 45, 3474; b) C. Ruspic, S. Nembenna, A.
Hofmeister, J. Magull, S. Harder, H. W. Roesky, J. Am. Chem. Soc. 2006, 128, 15000; c) S.
Nembenna, H. W. Roesky, S. Nagendran, A. Hofmeister, J. Magull, P.-J. Wilbrandt, M. Hahn,
Angew. Chem. Int. Ed. 2007, 46, 2512; d) M. P. Blake, N. Kaltsoyannis, P. Mountford, J. Am.
Chem. Soc. 2011, 133, 15358; e) A. S. S. Wilson, M. S. Hill, M. F. Mahon, C. Dinoi, L. Maron,
Science 2017, 358, 1168.
[7] B. Liu, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Chem. Eur. J. 2013, 19, 13445.
[8] M. R. Crimmin, M. S. Hill, P. B. Hitchcock, M. F. Mahon, New J. Chem. 2010, 34, 1572.
[9] W. D. Buchanan, D. G. Allis, K. Ruhlandt-Senge, Chem. Commun. 2010, 46, 4449.
[10] a) K. Yan, B. M. Upton, A. Ellern, A. D. Sadow, J. Am. Chem. Soc. 2009, 131, 15110; b) Y. Sarazin,
D. Roşca, V. Poirier, T. Roisnel, A. Silvestru, L. Maron, J.-F. Carpentier, Organometallics 2010,
29, 6569; c) N. Romero, S.-C. Roşca, Y. Sarazin, J.-F. Carpentier, L. Vendier, S. Mallet-Ladeira,
C. Dinoi, M. Etienne, Chem. Eur. J. 2015, 21, 4115.
[11] a) S.-C. Roşca, C. Dinoi, E. Caytan, V. Dorcet, M. Etienne, J.-F. Carpentier, Y. Sarazin, Chem.
Eur. J. 2016, 22, 6505; b) S.-C. Roşca, V. Dorcet, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Dalton
22
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Trans. 2017, 46, 14785; c) B. Freitag, J. Pahl, C. Färber, S. Harder, Organometallics 2018, 37, 469;
d) L. Garcia, M. D. Anker, M. F. Mahon, L. Maron, M. S. Hill, Dalton Trans. 2018, 47, 12684.
[12] See references [4b], [10a], [11a], [11b] and: a) A. G. M. Barrett, M. R. Crimmin, M. S. Hill, P. B.
Hitchcock, P. A. Procopiou, Angew. Chem. Int. Ed. 2007, 46, 6339; b) M. Gärtner, H. Görls, M.
Westerhausen, Dalton Trans. 2008, 1574; c) G. B. Deacon, P. C. Junk, R. P. Kelly, Aust. J. Chem.
2013, 66, 1288; d) S.-C. Roşca, T. Roisnel, V. Dorcet, J.-F. Carpentier, Y. Sarazin, Organometallics
2014, 33, 5630; e) S.-C. Roşca, E. Caytan, V. Dorcet, T. Roisnel, J.-F. Carpentier, Y. Sarazin,
Organometallics 2017, 36, 1269.
[13] D. R. Click, B. L. Scott, J. G. Watkin, Chem. Commun. 1999, 633.
[14] a) H. Yin, A. J. Lewis, U. J. Williams, P. J. Carroll, E. J. Schelter, Chem. Sci. 2013, 4, 798; b) H.
Yin, A. J. Lewis, P. Carroll, E. J. Schelter, Inorg. Chem. 2013, 52, 8234; c) H. Yin, P. J. Carroll,
E. J. Schelter, Inorg. Chem. 2016, 55, 5684; d) H. Yin, A. V. Zabula, E. J. Schelter, Dalton Trans.
2016, 45, 6313.
[15] J. F. Kꢀgel, L. H. Finger, N. Frank, J. Sundermeyer, Inorg. Chem. 2014, 53, 3839.
[16] J. F. Kꢀgel, D. A. Sorokin, A. Khvorost, M. Scott, K. Harms, D. Himmel, I. Krossing, J.
Sundermeyer, Chem. Sci. 2018, 9, 245.
[17] a) I. D. Brown, D. Altermatt, Acta Crystallogr. Sect. B 1985, B41, 244; b) N. E. Brese, M. O’Keefe,
Acta Crystallogr., Sect. B 1991, B47, 192; c) I. D. Brown, Chem. Rev. 2009, 109, 6858.
[18] R. Koppang, Acta chem. Scand. 1971, 25, 3067.
[19] J. Eppinger, E. Herdtweck, R. Anwander, Polyhedron 1998, 17, 1201.
[20] M. P. Coles, Coord. Chem. Rev. 2016, 323, 52.
[21] M. Westerhausen, M. Hartmann, N. Makropoulos, B. Wieneke, M. Wieneke, W. Schwarz, D.
Stalke, Z. Naturforsch., B: Chem. Sci. 1998, 53, 117.
[22] H. Plenio, Chem. Rev. 1997, 97, 3363.
[23] M. Westerhausen, W. Schwarz, Z. Anorg. Allg. Chem. 1991, 604, 127.
[24] T. J. Boyle, B. A. Hernandez-Sanchez, C. M. Baros, L. N. Brewer, M. A. Rodriguez, Chem. Mater.
2007, 19, 2016.
[25] A. G. M. Barrett, M. R. Crimmin, M. S. Hill, G. Kociok-Köhn, D. J. MacDougall, M. F. Mahon,
P. A. Procopiou, Organometallics 2008, 27, 3939.
[26] H. Bauer, M. Alonso, C. Färber, H. Elsen, J. Pahl, A. Causero, G. Ballmann, F. De Proft, S. Harder,
Nat. Catal. 2018, 1, 40.
[27] A search on the CSD database (version 5.40) for 3-coordinate charge neutral calcium compounds
returned 14 hits: CSD entries ABAGOA, CUDZUX, CULMAX, EPENIY, FISWEK, JOFPID,
LIDFAF, LOSPOY, PUVGAP, TERGIH, VEMNAF, YOGYAV, YOGYEZ and YUMBAT.
[28] B. A. Vaartstra, J. C. Huffman, W. E. Streib, K. G. Caulton, Inorg. Chem. 1991, 30, 121.
23
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
[29] a) S.-O. Hauber, F. Lissner, G. B. Deacon, M. Niemeyer, Angew. Chem. Int. Ed. 2005, 44, 587; b)
O. Michel, S. König, K. W. Törnroos, C. Maichle-Mössmer, R. Anwander, Chem. Eur. J. 2011, 17,
11857.
[30] B. Liu, T. Roisnel, Y. Sarazin, Inorg. Chim. Acta 2012, 380, 2.
[31] a) W. D. Buchanan, E. D. Nagle, K. Ruhlandt-Senge, Main Group Chem. 2009, 8, 263; b) J. S.
Sum, L. Tahsini, J. A. Golen, C. Moore, A. L. Rheingold, L. H. Doerrer, Chem. Eur. J. 2013, 19,
6374; c) W. D. Buchanan, K. Ruhlandt-Senge, Chem. Eur. J. 2013, 19, 10708.
[32] S. Grimme, S. Ehrlich, L. Goerigk, J. Comp. Chem. 2011, 32, 1456.
[33] a) K. Morokuma, J. Chem. Phys. 1971, 55, 1236; b) T. Ziegler, A. Rauk, Inorg. Chem. 1979, 18,
1558; c) F. M. Bickelhaupt, E. J. Baerends, Rev. Comput. Chem.; K.B. Lipkowitz and D.B. Boyd,
Eds.; Wiley, New York 2000, 15, 1.
[34] M. Gillett-Kunnath, W. Teng, W. Vargas, K. Ruhlandt-Senge, Inorg. Chem. 2005, 44, 4862.
[35] M. R. Crimmin, A. G. M. Barrett, M. S. Hill, P. B. Hitchcock, P. A. Procopiou, Organometallics
2007, 26, 2953.
[36] The molecular structure of [{BDIDiPP}CaN(o-F-C6H4)2.(thf)] is given in the Supporting
Information. The data set and final refinement we have so far been able to achieve are of insufficient
quality to discuss the details of interatomic distances and angles (Rint = 0.1638, R1 = 0.0978), but
the connectivity was established without ambiguity.
24
This article is protected by copyright. All rights reserved.

10.1002/chem.201901262
Chemistry - A European Journal
Graphical Abstract
25
This article is protected by copyright. All rights reserved.