FLP behaviour of cationic titanium complexes with tridentate Cp, O, N -ligands: Highly efficient syntheses and activation reactions of C-X bonds (X = Cl, F)

Article abstract of DOI:10.1039/c8dt04707c

The synthesis of cationic titanium complexes 4a,b with tridentate Cp,O,N-ligand frameworks, starting from the monopentafulvene complex Cp?Ti(Cl)(π-η⁵:σ-η¹-C₅H₄CR₂ (CR₂ = adamantylidene) (1) and bidentate O,N-ligand precursors CH₃C(O)CH₂CH₂NR₂ (R = Et, CH₂Ph) (L1a,b), in a high-yielding and efficient two-step synthetic pathway is described. The β-aminoketones L1a,b were synthesized by a herein reported solvent- and catalyst-free reaction. The reaction pathway involves insertion reactions, subsequent methylations and final activations with B(C₆F₅)₃. NMR investigations of the cationic titanium complex 4a in deuterated dichloromethane revealed an ongoing selective reaction under formation of the cationic complex 5a-d₂, which is the result of C-Cl bond cleavage. In addition to selective C-Cl bond activation reactions, C(sp³)-F bonds were activated by 4a,b, pointing out the special tm-FLP nature of 4a,b.

Full text of DOI:10.1039/c8dt04707c

Dalton
Transactions
View Article Online
View Journal
PAPER
FLP behaviour of cationic titanium complexes with
tridentate Cp,O,N-ligands: highly efficient syntheses
and activation reactions of C–X bonds (X = Cl, F)†
Cite this: DOI: 10.1039/c8dt04707c
Malte Fischer,
Kevin Schwitalla, Svenja Baues, Marc Schmidtmann
and
Ruediger Beckhaus
*
The synthesis of cationic titanium complexes 4a,b with tridentate Cp,O,N-ligand frameworks, starting from
the monopentafulvene complex Cp*Ti(Cl)(π-η5:σ-η1-C5H4CR2 (CR2 = adamantylidene) (1) and bidentate
O,N-ligand precursors CH3C(O)CH2CH2NR2 (R = Et, CH2Ph) (L1a,b), in a high-yielding and efficient two-
step synthetic pathway is described. The β-aminoketones L1a,b were synthesized by a herein reported
solvent- and catalyst-free reaction. The reaction pathway involves insertion reactions, subsequent methyl-
ations and final activations with B(C6F5)3. NMR investigations of the cationic titanium complex 4a in deute-
rated dichloromethane revealed an ongoing selective reaction under formation of the cationic complex
5a-d2, which is the result of C–Cl bond cleavage. In addition to selective C–Cl bond activation reactions,
C(sp3)–F bonds were activated by 4a,b, pointing out the special tm-FLP nature of 4a,b.
Received 28th November 2018,
Accepted 8th January 2019
DOI: 10.1039/c8dt04707c
rsc.li/dalton
ponent.4 This concept offers the advantage of versatile
tuning potentials of the Lewis acidity because of the nearly
Introduction
The rapidly expanding chemistry of frustrated Lewis pairs endless scope represented by ligand design prospects, which has
(FLPs) has flourished into one of the main topics in the inter- already allowed the combination of FLP-based small-molecule acti-
disciplinary field of inorganic and organic chemistry since the vation with the cornerstones of transition-metal based homo-
pioneering first description of this concept by Stephan and co- geneous catalysis.1–4 As recent outstanding examples with respect
workers1 several years ago.2 Initial reports were based on Lewis to tm-FLP chemistry, the activation of dinitrogen by late transition
acids and bases derived from main-group elements, in which metal dinitrogen complexes with B(C6F5)3 is especially worthy to
the classic adduct formation is prevented either by steric or note.5
electronic properties. Thus, unique reactivity patterns and
Many examples of tm-FLPs with Zr,6–20 Hf,14,19,21 Ru,22,23
manifold stoichiometric and catalytic transformations were and rare-earth metals24–29 as central metal atoms have found
unveiled, relying heavily on Lewis acid and Lewis base their entry into the literature, whereas only a handful of tita-
cooperation in a spatially confined reaction compartment.3 As nium-based tm-FLPs I–IV, which show only limited reactivity
a consequence, manipulating the Lewis acidity or basicity of (I–III), are reported (summarized in Scheme 1).30–33
the components can be crucial for further reactivity.
The scarceness of titanium-based FLPs is kind of ironic
Whereas main-group FLPs are usually focused on polyfluori- considering the early intramolecular example reported by
nated arylboranes as the Lewis acid component, which limits the Stephan et al. (Scheme 1, I).30
possibilities of tuning the Lewis acidity, a recent expansion was
In this context, it is highly desirable to focus on efficient
the development of transition metal based FLPs (tm-FLPs) where syntheses of titanium-based FLPs, because titanium being the
the electrophilic metal centres constitute the Lewis acidic com- second most earth-abundant transition metal, of cost and
environmental reasons, especially when compared to the
chemistry of late transition metals like ruthenium or rare-
earth metals like scandium.34–36 Additionally, titanium is easy
Institut fuer Chemie, Fakultaet für Mathematik und Naturwissenschaften, Carl von
to recycle, considered to be nontoxic and environmentally
benign.34–36
We have put recent effort into the development of efficient
and convenient syntheses of such systems by employing readily
accessible monopentafulvene titanium complexes and appropri-
ate bidentate ligand precursors (prepared in one step starting
Ossietzky Universitaet Oldenburg, Postfach 2503, 26111 Oldenburg, Germany.
E-mail: ruediger.beckhaus@uni-oldenburg.de; https://www.uni-oldenburg.de/ac-beck-
haus/; Fax: (+49) 441-798-3581
†Electronic supplementary information (ESI) available: Additional experimental
details, further spectral and crystallographic data. CCDC 1871463, 1871464,
1871466, 1871467, 1871469 and 1871470. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c8dt04707c
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
targeted for the subsequent synthetic protocol in analogy to
the β-keto phosphines, which were established as ligand pre-
cursors previously.38
β-Amino ketones can be prepared via the Mannich reaction
through a carbon–carbon bond formation, starting from eno-
lates and imines. However, this reaction requires equimolar
amounts of base for the preparation of enolates, and often
harsh reaction conditions accompanied by long reaction
times.39 The Aza-Michael reaction offers another way to
prepare β-amino ketones through conjugated addition of
amines to α,β-unsaturated carbonyl compounds, having the
advantage of being much more atom economic than the
Mannich reaction. Despite this benefit, Michael reactions in
general require basic reaction conditions,40,41 the use of either
stoichiometric or catalytic amounts of Lewis acids (e.g. Bi
(NO3)3,42 CeCl3·7H2O,43 SmI2(THF)2 44), and often (toxic)
organic solvents, so that subsequent purification steps are
indispensable. Another approach known from the literature
suggests the use of micellar solutions.45
Inspired by the preparation of the aforementioned β-keto
phosphines by a solvent- and catalyst-free hydrophosphanation
of methyl vinyl ketone46 the β-amino ketones L1a,b were pre-
pared by simple addition of the corresponding alkyl substi-
tuted secondary amines to methyl vinyl ketone (Scheme 2).
The diethyl substituted derivative L1a is obtained in nearly
quantitative yield and on a multigram scale, and the exother-
mic reaction is complete after five minutes as shown by 1H
NMR measurements. The dibenzyl substituted ligand precur-
sor L1b is obtained after 36 h of stirring at room temperature
and removal of a small excess of the starting material as a col-
ourless solid in 96% isolated yield (crystals suitable for single
crystal X-ray diffraction of L1b where obtained from a saturated
n-hexane solution of L1b at −26 °C; Fig. S11†47). For the
analytical data of L1a,b see the ESI.†47
Scheme 1 Overview of titanium-based FLPs I–IV and motivation.30–33
from commercially available starting materials) to give the corres-
ponding cationic compounds in optimized two-step syntheses. It
could be shown that complexes like IV show remarkable sub-
sequent reactions in a tm-FLP fashion (Scheme 1).33,37,38 The
fact, that the subsequent reactions take place at the Ti–N bond,
emerged the question of introducing a nitrogen functionality at
the hemilabile donor position (Scheme 1, bottom). This general
concept includes the characteristics of hemilabile O,N-ligands, in
which the unoccupied coordination site can become available for
further reactions by dissociation of a donor atom to create an
open coordination site for further reactions.61 The highly functio-
nalized tridentate Cp,E,P-ligand frameworks (E = O, N) are built
directly in the coordination sphere of the central metal atom,
which underlines the efficient approach. This is in stark contrast
to typical syntheses of tridentate ligand systems, which require
multiple purification steps for the formation of comparable
densely functionalized ligands.
In effect, a highly desirable, solvent- and catalyst-free, and
environmentally benign Aza-Michael reaction is possible by
simple addition of secondary amines to methyl vinyl ketone.
Synthesis, characterization and reactivity of the complexes
The reactions of the monopentafulvene titanium complex
Cp*Ti(Cl)(π-η5:σ-η1-C5H4vCR2) 1, featuring the sterically
demanding adamantylidene substitution pattern at the exocyc-
lic carbon atom (Cexo), with the bidentate O,N-ligand precur-
sors L1a,b in n-hexane at room temperature were accompanied
by colour changes of the reaction mixtures from yellow-brown
to yellow-orange. Removal of all volatiles under reduced press-
With this in mind we investigated the reactivity of suitable
bidentate O,N-ligand precursors, accessible by
a herein
reported solvent- and catalyst-free addition of various second-
ary amines to methyl vinyl ketone, their successful application
in the two-step synthesis of cationic group 4 metal complexes
with novel tridentate Cp,O,N-ligands, starting from a mono-
pentafulvene titanium complex, and the subsequent tm-FLP
chemistry of these compounds.
Results and discussion
Preparation of the β-amino ketones L1a,b
In order to start our investigations, an efficient synthesis of
Scheme 2 Solvent-and catalyst-free synthesis of the β-amino ketones
β-amino ketones as potential suitable ligand precursors was L1a,b.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Table 1 Comparison of selected bond lengths (Å) of 2a,b and 3a,b
ures yield the products 2a,b with Cp,O σ-π-chelating ligands, as
the result of the insertion of the carbonyl functional groups of
L1a,b into the polarized Ti–Cexo bond as yellow or orange
solids in isolated yields of up to 88% (Scheme 3).
Ti–O
O–Cq
C
q,exo–Cq,ipso
Cq,exo–OCq
2a
2b
3a
3b
1.8713(8)
1.8522(4)
1.8654(7)
1.8527(16)
1.4326(13)
1.4298(7)
1.4272(11)
1.432(3)
1.5235(15)
1.5226(7)
1.5234(13)
1.530(3)
1.6250(15)
1.6386(8)
1.6312(13)
1.635(3)
Complexes 2a,b are moderately soluble in aliphatic hydro-
carbons like n-hexane and show good solubilities in aromatic
and polar solvents (e.g. toluene, benzene or tetrahydrofuran).
They can be stored in the solid state for months under inert
conditions without indication of decomposition. Subsequent
reactions to yield the corresponding singly methylated com-
plexes 3a,b proceed via reactions of 2a,b with methyllithium in
tetrahydrofuran at room temperature. After purification 3a,b
were isolated as pale yellow or pale orange solids in good
yields, showing the same solubility properties as 2a,b.
Furthermore, 3a,b can also be prepared in a one-pot procedure
starting from 1 with the advantage of one less purification and
isolation step in consistently good yields of up to 74%.
Complexes 2a,b and 3a,b were thoroughly characterized by
NMR spectroscopy and single-crystal X-ray diffraction.
Due to strong similarities of the most important and
characteristic structural parameters of the crystallized com-
plexes, the X-ray data are exemplary discussed for complexes
2b and 3b and summarized for all in Table 1 (ORTEP presenta-
tions of 2a and 3a are presented in the ESI (S13,14)).†47 The
single crystals were obtained from saturated solutions in
n-hexane or toluene at temperatures of −4 °C or −26 °C.
Fig. 1 Molecular structure of 2b in the crystal. Thermal ellipsoids are
drawn at the 50% probability level (hydrogen atoms have been omitted
for clarity). Selected bond lengths (Å) and angles (deg): Ti1–Cl1 2.39082
(19), Ti1–O1 1.8522(4), O1–C26 1.4298(7), N1–C28 1.4775(7), C11–C16
1.5226(7), C16–C26 1.6386(8), C26–C27 1.5388(8), C26–C29 1.5311(8),
C27–C28 1.5278(8), Cl1–Ti1–O1 97.643(15), Ct1–Ti1–Ct2 134.0, Σ∠C26
322.5 (O1–C26–C27 + O1–C26–C29 + C27–C26–C29), ∑∠N1 333.9
(Ct1 = centroid of C1–C5; Ct2 = centroid of C11–C15).
ˉ
2b crystallizes in the triclinic space group P1 and complex
3b crystallizes in the monoclinic space group P21/c. The mole-
cular structures of 2b and 3b (Fig. 1 and 2) display distorted
tetrahedral coordination environments at the titanium atoms
(e.g. Cl1–Ti1–O1 97.643(15)°, Ct1–Ti1–Ct2 134.0° for 2b).
The Ti1–Cl1 bond length of 2.39082(19) Å in 2b and the
Ti1–C44 bond length of 2.185(3) Å in 3b are typical of single
bonds and lie in the same range as in related complexes with
Cp,O σ-π-chelating ligands.37,38,48–50 The newly formed bonds
Ti1–O1 (1.8522(4) Å (2b), 1.8527(16) Å (3b)) and C16–C26
(1.6386(8) Å (2b), 1.635(3) Å (3b)) confirm the successful inser-
tion reaction and are characteristic for complexes derived from
insertion reactions of monopentafulvene complexes with mul-
tiple bond substrates.33,37,38 The Ti1–O1 bond lengths, which
are slightly shortened compared to Ti–O single bonds, are
indicative for attractive Ti(dπ)–O(pπ) interactions.49,50 The
strong ring strains, resulting from the formation of the Cp,O σ-
π-chelating ligands cause exceptional elongations of the newly
formed C16–C26 bonds between the former carbonyl carbon
atom and the Cexo atom when compared to typical C(sp3)–C
(sp3) single bonds,51 but are in the same range as reported for
highly sterically crowded alkanes.52 Despite these significant
elongations, the former carbonyl carbon atoms are sp3-hybri-
dized as shown by the sum of angles around C26 in both cases
Scheme 3 Three-/two-step synthesis of cationic titanium complexes 4a,b.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
field compared to the corresponding ligand precursor L1b
(δ1H = 3.36 (s, 4H) ppm/δ13C{1H} = 58.5 ppm). The 13C{1H}
NMR chemical shifts of the quaternary Cexo (δ13C{1H} =
54.8 ppm) and Cipso (δ13C{1H} = 156.9 ppm) atoms of the
former pentafulvene ligand are in the same range as for
related complexes reported previously,37,38 and show signifi-
cantly different chemical shifts compared to the starting
material 1 (δ13C{1H} = 130.1 (Cq,exo) and 131.3 (Cq,ipso) ppm).53
1
The four H NMR signals for the C5H4 moiety range from 5.02
to 6.44 ppm. The signal of the terminal methyl group of the
corresponding methylated complex 3b is observed at δ1H =
0.10 ppm (δ13C{1H} = 34.8 ppm), which is in the same range as
reported for other terminal methylated titanium com-
plexes.33,38,54 The 15N NMR chemical shift at δ15N = 48.4 ppm
is only marginally different to L1b, clearly confirming identical
chemical environments at the nitrogen atoms.
As mentioned above, complex 2b was obtained diastereo-
merically pure and its X-ray structure reveals, that the titanium
atom and the former carbonyl carbon atom are (R,R) and (S,S)
configured.
The prepared methylated complexes 3a,b were reacted with
the highly Lewis acidic borane B(C6F5)3 in toluene at room
temperature and led to the isolation of the activated cationic
complexes 4a,b in isolated yields of up to 80% after purifi-
cation as pale orange solids (Scheme 3).
Fig. 2 Molecular structure of 3b in the crystal. Thermal ellipsoids are
drawn at the 50% probability level (hydrogen atoms have been omitted
for clarity). Selected bond lengths (Å) and angles (deg): Ti1–C44 2.185
(3), Ti1–O1 1.8527(16), O1–C26 1.432(3), N1–C28 1.472(3), C11–C16
1.530(3), C16–C26 1.635(3), C26–C27 1.547(4), C26–C29 1.536(3), C27–
C28 1.530(3), C44–Ti1–O1 98.60(9), Ct1–Ti1–Ct2 136.0, ∑∠C26 320.7
(O1–C26–C27 + O1–C26–C29 + C27–C26–C29), ∑∠N1 330.1 (Ct1 =
centroid of C1–C5; Ct2 = centroid of C11–C15).
The complexes are insoluble in aliphatic hydrocarbons
such as n-hexane, which proves to be ideal for purification
(322.5° (2b), 320.7° (3b). The transition from the pentafulvene purposes. In aromatic solvents such as benzene or toluene
ligand in 1 to substituted cyclopentadienyl ligands in 2b and only a marginal solubility is observed, which, during the reac-
3b is confirmed by the C11–C16 bond lengths, which are now tions of 3a,b with B(C6F5)3, is reflected by the separation of
typical of C(sp3)–C(sp2) single bonds with 1.5226(7) Å (2b) and two phases as soon as the stirring process is stopped. The
1.530(3) Å (3b) respectively.51
previously reported complexes (e.g. derivatives of IV,
The solution NMR data of 2a,b and 3a,b are in accordance Scheme 1) show good solubilites in dichloromethane and
with their solid-state structures. Due to the prochirality of the were stable without any indication of decomposition in di-
bidentate O,N-ligand precursors L1a,b, a double set of signals chloromethane solutions.33,37,38 Therefore CD2Cl2 was also
is observed (2b is the exception) for these complexes, exhibit- chosen as solvent for multinuclear NMR spectroscopy of the
ing two stereogenic centres at the titanium atoms and the current complexes.
former carbonyl carbon atoms. The ratios of the diasteroi-
Surprisingly, the 1H NMR spectrum of the cationic complex
somers were determined by integration of appropriate signals 4a shows initially a double set of signals, for which the relative
1
in the H NMR spectra and lie between 4 : 1 and 10 : 1. Similar ratios of the integrals change after one hour, indicating either
ratios of the diastereoisomers were reported for complexes an ongoing rearrangement or a subsequent reaction. After one
with the Cp,O,P-ligand frameworks.37,38 The NMR data are dis- day, only one set of signals remains. 2D- and 1D NMR spec-
cussed for the obtained diastereomerically pure derivative 2b.
troscopy reveals, that a subsequent clean reaction between
In the high field region of the 1H NMR spectrum the CD2Cl2 and the cationic complex 4a took place to yield the acti-
methyl group (δ1H = 1.29 ppm/δ13C{1H} = 31.2 ppm) located at vation product 5a-d2 (summarized in Fig. 3).
the quaternary carbon atom of the former carbonyl group,
In consequence of this unexpected reaction with CD2Cl2, we
together with the signals of the linking methylene groups attempted to obtain NMR data of 4a in THF-d8 and C6D5Br as
between the oxygen and nitrogen atoms, and the signals of the the solvents, which also proved to be inadequate because of
adamantyl substituent are observed together with the signal of poor solubility of 4a. However, 4a,b are characterized by
the methyl groups of the Cp* ligand (δ1H = 1.76 ppm/δ13C{1H} elemental analysis and the clean detection of the M+ signals
= 12.6 ppm).
with high-resolution ESI mass spectrometry. We started to
Of high diagnostic value are the signals of the methylene prepare 4a,b in situ for investigations of their further reactivity.
groups of the benzyl substituents, which show two doublet
The reaction of in situ synthesized 4a with non-deuterated
signals at δ1H = 3.63 and 3.98 ppm (δ13C{1H} = 60.1 ppm) due dichloromethane was repeated and the corresponding acti-
to the diastereotopicity with characteristic coupling constants vation product 5a was obtained rapidly (reaction took less than
2
of JH,H = 14 Hz each. The signals are shifted toward lower five minutes, as indicated by a colour change to clear orange)
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Fig. 3 Excerpt of the 1H NMR spectrum of the reaction of 4a with
CD2Cl2 (bottom: 300 MHz, CD2Cl2, rt; middle, top: 500 MHz, CD2Cl2,
rt).
Fig. 4 Molecular structure of 5b in the crystal. Thermal ellipsoids are
drawn at the 50% probability level (hydrogen atoms except H44A, H44B
and co-crystallized solvent have been omitted for clarity). Selected bond
lengths (Å) and angles (deg): Ti1–Cl1 2.3856(4), Ti1–O1 1.8749(8), O1–
C26 1.4233(13), N1–C28 1.5332(16), N1–C44 1.5176(17), Cl2–C44 1.7644
(13), B1–C63 1.6333(19), C11–C16 1.5271(16), C16–C26 1.6277(16), C26–
C27 1.5543(16), C26–C29 1.5350(16), C27–C28 1.5256(17), Cl1–Ti1–O1
96.10(3), Ct1–Ti1–Ct2 133.3, ∑∠C26 321.8 (O1–C26–C27 + O1–C26–
C29 + C27–C26–C29), ∑∠N1 329.1 (C28–N1–C30 + C28–N1–C37 +
C30–N1–C37) (Ct1 = centroid of C1–C5; Ct2 = centroid of C11–C15).
in 71% isolated yield as an analytically pure pale yellow solid,
which shows similar solubility properties as 4a.
Complex 5a is the result of addition of a chloride,
abstracted from CH2Cl2, to the titanium centre and addition of
the remaining CH2Cl moiety to the nitrogen, which then
forms an ammonium functional group. The same reaction
took place starting from 4b to give the corresponding complex
5b (Scheme 4).
5a,b are fully characterized by NMR measurements, ESI
mass spectrometry (clean detection of the M+ signal), and for
5b, single crystal X-ray diffraction, for which crystals were
obtained by slow diffusion of n-hexane into a bromobenzene
solution of 5b. The molecular structure is shown in Fig. 4.
the ammonium moiety.51 Due to the formation of the
ammonium moiety, the N–C bond lengths are generally
elongated compared to their non-cationic precursor complexes
(e.g. N1–C28: 1.5332 (16) Å (5b), 1.4775(7) Å (2b), 1.472(3) Å
(3b)). The molecular structure of 5b proves, that the borate
anion is noncoordinating. Additionally, the central metal atom
and the C26 carbon atom both show S, or by applying crystal
symmetry, R configuration, so that consequently 5b is obtained
diastereomerically pure as the (S,S) and (R,R) pair of enantio-
mers, which is confirmed by the single set of signals in the
NMR spectra.
ˉ
5b crystallizes in the triclinic space group P1. The central
titanium atom is in a distorted tetrahedral coordination
environment (Cl1–Ti1–O1 96.10(3), Ct1–Ti1–Ct2 133.3). The
newly formed Ti1–Cl1 bond length of 2.3856(4) Å and the Ti1–
O1 bond length of 1.8749(8) Å are typical of single bonds and
remain nearly unchanged compared to those of 2b and 3b.49,50
The second newly formed bond N1–C44 (1.5176(17) Å) is
characteristic of a N–C(sp3) single bond and clearly confirms
The signal of the methyl group (δ1H = 0.49 ppm) located at
the boron atom is characteristically broadened. This is caused
by the high magnetic quadrupole moment of the boron atom,
so that the corresponding chemical shift of the corresponding
carbon atom had to be assigned by the 1H/13C-HMQC spec-
trum. The 11B{1H} (δ11B{1H} = −14.9 ppm) and 19F{1H} NMR
chemical shifts (δ19F{1H} = −167.8 (m-FArB), −165.2 (p-FArB),
−133.1 (o-FArB) ppm; Δδ(m,p-F) = 2.6 ppm) are in accordance
to other cationic complexes with this specific anion. Here, the
Δδ(m,p-F) value is a good probe for the evaluation, whether the
−
MeB(C6F5)3 anion is coordinating or not.13,19,33,37,38,55 As
Scheme 4 Reactions of 4a,b with dichloromethane.
reported in the literature, the Δδ(m,p-F) value of 2.6 ppm (5b)
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 2 Selected NMR parameters of complexes 2b, 3b, and 5ba
2b
3bb
5b
δ1H/δ13C{1H} ECH3
—
0.10/34.8c
55.0/151.5
3.57/3.69//60.2
2JH,H = 13.9/14.0
—
0.49/10.0d
55.3/156.8
δ
13C{1H} Cq,exo/Cq,ipso
54.8/156.9
3.63/3.69//60.1
2JH,H = 13.9/14.0
—
δ1H/δ13C{1H} NCH2Ph
2JH,H
4.49/4.55/5.00//63.2/64.0
2JH,H = 13.4
δ1H/δ13C{1H} NCH2Cl
4.81/4.90//66.2
2JH,H = 10.3
109.6
2JH,H
δ
13C{1H} OCq
112.1
5.03, 5.44 (2H), 6.44
107.6
4.85, 5.13, 5.28, 6.19
δ1H C5H4
5.37, 5.45, 5.98, 6.39
a Values are given in ppm and J values in Hz. Measurements were carried out in C6D6 (2b, 3b) or CD2Cl2 (5b) at room temperature. b Product is
obtained as a mixture of diastereoisomers; therefore signals of the main diastereoisomer are given. c E = Ti. d E = B.
is characteristic for the borate anion being noncoordinating in multinuclear NMR spectroscopy.47 Complexes, which are
solution, which is also represented by its structure in the solid capable of activating (inactivated) aliphatic C(sp3)–F bonds are
state.56,57,62 The other NMR data of 5b are in the expected scarce,6,9,58 so that the herein presented complexes might
ranges and summarized together with those of 2b and 3b in allow possible applications in defluorination or hydrodefluori-
Table 2 (for the data of the congener 5a see the ESI†47).
nation reactions.59,60 Above that, 4a,b are the first titanium-
Similar reactions with dichloromethane have been based cationic complexes, showing this reactivity. In addition,
reported,6,9,30 for example for the intermolecular Ti+/P system the chemistry concerning C–F bond activations by strong
reported by Stephan and coworkers (Scheme 1, I),30 but in the Lewis acids has been most successful in breaking (activated)
light of to the entirety of FLP systems, those examples remain C(sp2)–F bonds, employing late transition metals.63 In the last
rare. Especially, until now, no main group based FLPs were years significant improvements have been reported in the field
capable of cleaving carbon–halogen bonds.6 On top of that, of activating (inactivated) C(sp3)–F bonds by e.g. strong main
analogous titanium complexes with Cp,O,P- or Cp,N,P-ligands group Lewis acids.64a,b Noteworthy, no C(sp2)–halogen cleavage
were completely inert toward dichloromethane, emphasizing reactions with aryl substituted substrates like chloro- or bro-
the unique features of the Cp,O,N-ligand framework.33,37,38
mobenzene were observed, even when 4a and 4b were dis-
With this in mind, we further explored the reactivity toward solved in excess of these substrates. These observations are in
other carbon–halogen substrates (Scheme 5).
good agreement to previous reports.6,9
The reaction of 4a with benzyl chloride followed the same
As demonstrated, the pendant amine functionalities, in
reactivity pattern, yielding complex 6 in 74% isolated yield as a contrast to the previously reported phosphorus containing
pale yellow solid, which has been thoroughly characterized.47 counterparts,37,38 enable subsequent carbon–halogen bond
Even when alkyl fluorides are reacted with 4a or 4b under mild activation reactions. This can be attributed to the higher basi-
reaction conditions, an analogous reaction pathway is city of the amine functionalities in 4a,b compared to the phos-
observed. Thus, treatment of 4a and 4b with stoichiometric phine moieties, which, together with the Lewis acidic titanium
amounts of 1-fluorodecane resulted in clean conversion to 7a centres, are capable of cooperatively activating these substrates
(70%) and 7b (81%), respectively. They were also characterized at the Ti⋯N bond. In addition, a persistent covalent Ti–P bond
by detection of the M+ signals by ESI mass spectrometry, and is present in the previously reported complexes,37,38 which
hampers subsequent reactions. This is further supported by
the two other tm-FLP systems (Zr+/P,6 Zr+/N9) capable of acti-
vating carbon–halogen bonds, in which no or only a weak
contact between the Lewis acidic metal centres and the Lewis
basic functionalities are required for subsequent activation
reactions. Thus, a higher grade of frustration between the
amine functionalities and the titanium centres in 4a,b is an
additional reason for the observed enhanced reactivity.
Conclusions
In summary, we report a novel solvent- and catalyst-free access
to the β-aminoketones L1a,b, starting from methyl vinyl
ketone and various secondary amines. The β-aminoketones
L1a,b were successfully employed as ligand precursors in reac-
tions with the titanium monopentafulvene complex 1 (Cp*Ti
Scheme 5 Enhanced reactivity studies of 4a,b.
(Cl)(π-η5:σ-η1-C5H4CR2: CR2
= adamantylidene), and sub-
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
sequent adoption in the previously established and optimized
two-step synthetic pathway to yield the corresponding cationic
complexes 4a,b. The two-step synthetic pathway involves inser-
tion of the carbonyl functional group of L1a,b into the polar-
ized Ti–Cexo bond, methylation, and final activation with the
strong Lewis acid B(C6F5)3 (insertion and methylation can be
done in a one-pot procedure). The cationic complexes 4a,b
feature the first tridentate Cp,O,N-ligand system, which is
6 A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am.
Chem. Soc., 2011, 133, 18463–18478.
7 O. J. Metters, S. J. K. Forrest, H. A. Sparks, I. Manners and
D. F. Wass, J. Am. Chem. Soc., 2016, 138, 1994–2003.
8 A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am.
Chem. Soc., 2011, 133, 8826–8829.
9 X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, J. Am. Chem.
Soc., 2015, 137, 4550–4557.
directly synthesized in the coordination sphere of the metal. 10 A. T. Normand, C. G. Daniliuc, B. Wibbeling, G. Kehr, P. Le
This in an expansion of the series of tridentate ligands of the
Gendre and G. Erker, J. Am. Chem. Soc., 2015, 137, 10796–
previously reported Cp,O,P- and Cp,N,P-congeners. The nitro-
10808.
gen, now localized at the hemilabile donor position, provides 11 Y.-L. Liu, G. Kehr, C. G. Daniliuc and G. Erker,
the ability of activating carbon–halogen bonds, whereas the Organometallics, 2017, 36, 3407–3414.
Cp,O,P- and Cp,N,P-derivatives are completely inert toward di- 12 Z. Jian, C. G. Daniliuc, G. Kehr and G. Erker,
chloromethane, the cationic complexes 4a,b react cleanly and Organometallics, 2017, 36, 424–434.
readily with dichloromethane to give the corresponding com- 13 A. T. Normand, C. G. Daniliuc, B. Wibbeling, G. Kehr, P. Le
plexes 5a,b via C–Cl bond cleavage, addition of chloride to the
Gendre and G. Erker, Chem. – Eur. J., 2016, 22, 4285–
titanium centres, and formation of ammonium functional
4293.
groups by formal addition of CH2Cl+ to the nitrogen atoms. 14 X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, J. Am. Chem.
Furthermore, 4a,b selectively activate the C(sp3)–F bond of
n-fluorodecane at room temperature under formation of 7a,b 15 S. Frömel, G. Kehr, R. Fröhlich, C. G. Daniliuc and
with Ti–F and ammonium functional groups, being the first G. Erker, Dalton Trans., 2013, 42, 14531–14536.
titanium-based cationic complexes showing this reactivity. 16 X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, Angew. Chem.,
Soc., 2014, 136, 12431–12443.
With the herein presented study, the special tm-FLP nature of
4a,b is demonstrated.
Int. Ed., 2013, 52, 13629–13632, (Angew. Chem., 2013, 125,
13874–13877).
17 X. Xu, G. Kehr, C. G. Daniliuc and G. Erker,
Organometallics, 2013, 32, 7306–7311.
18 X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, J. Am. Chem.
Soc., 2013, 135, 6465–6476.
Conflicts of interest
There are no conflicts of interest to declare.
19 A. M. Chapman, M. F. Haddow and D. F. Wass,
Eur. J. Inorg. Chem., 2012, 2012, 1546–1554.
20 R. C. Neu, E. Otten, A. Lough and D. W. Stephan, Chem.
Sci., 2011, 2, 170–176.
21 M. J. Sgro and D. W. Stephan, Chem. Commun., 2013, 49,
2610–2612.
22 M. J. Sgro and D. W. Stephan, Angew. Chem., Int. Ed., 2012,
51, 11343–11345, (Angew. Chem., 2012, 124, 11505–
11507).
Acknowledgements
Financial support by the DFG Research Training Group 2226 is
kindly acknowledged.
23 K. F. Kalz, A. Brinkmeier, S. Dechert, R. A. Mata and
F. Meyer, J. Am. Chem. Soc., 2014, 136, 16626–16634.
Notes and references
1 G. C. Welch, R. R. S. Juan, J. D. Masuda and D. W. Stephan, 24 P. L. Arnold, I. A. Marr, S. Zlatogorsky, R. Bellabarba and
Science, 2006, 314, 1124.
R. P. Tooze, Dalton Trans., 2014, 43, 34–37.
2 Frustrated Lewis Pairs
I
and II, Uncovering and 25 A. Berkefeld, W. E. Piers, M. Parvez, L. Castro, L. Maron
Understanding/Expanding the Scope, in Topics in Current
Chemistry, ed. G. Erker and D. W. Stephan, Springer,
and O. Eisenstein, J. Am. Chem. Soc., 2012, 134, 10843–
10851.
Heidelberg, New York, Dordrecht, London, 2013, vol. 332/ 26 A. Berkefeld, W. E. Piers, M. Parvez, L. Castro, L. Maron
334. and O. Eisenstein, Chem. Sci., 2013, 4, 2152–2162.
3 Recent reviews: (a) D. W. Stephan, Science, 2016, 354, 27 P. Xu, Y. Yao and X. Xu, Chem. – Eur. J., 2017, 23, 1263–
aaf7229; (b) D. W. Stephan and G. Erker, Angew. Chem., Int. 1267.
Ed., 2015, 54, 6400–6441, (Angew. Chem., 2015, 127, 6498– 28 K. Chang, X. Wang, Z. Fan and X. Xu, Inorg. Chem., 2018,
6541); (c) D. W. Stephan and G. Erker, Angew. Chem., Int.
Ed., 2010, 49, 46–76, (Angew. Chem., 2010, 122, 50–81).
4 S. R. Flynn and D. F. Wass, ACS Catal., 2013, 3, 2574–2581.
5 A. Simonneau, R. Turrel, L. Vendier and M. Etienne, Angew.
57, 8568–8580.
29 K. Chang and X. Xu, Dalton Trans., 2017, 46, 4514–4517.
30 L. Cabrera, E. Hollink, J. C. Stewart, P. Wei and
D. W. Stephan, Organometallics, 2005, 24, 1091–1098.
Chem., Int. Ed., 2017, 56, 12268–12272, (Angew. Chem., 31 A. M. Chapman and D. F. Wass, Dalton Trans., 2012, 41,
2017, 129, 12436–12440).
9067–9072.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
32 A. T. Normand, P. Richard, C. Balan, C. G. Daniliuc,
G. Kehr, G. Erker and P. Le Gendre, Organometallics, 2015,
34, 2000–2011.
33 M. Fischer, D. Barbul, M. Schmidtmann and R. Beckhaus,
Organometallics, 2018, 37, 1979–1991.
J. E. P. Dahl, R. M. K. Carlson and A. A. Fokin, Nature,
2011, 477, 308–311.
53 A. Scherer, D. Haase, W. Saak, R. Beckhaus, A. Meetsma
and M. W. Bouwkamp, Organometallics, 2009, 28, 6969–
6974.
34 A. J. Hunt, T. F. Farmer and J. H. Clark, in Element Recovery 54 C. Adler, A. Bekurdts, D. Haase, W. Saak, M. Schmidtmann
and Sustainability, ed. A. J. Hunt, Royal Society of
Chemistry, Cambridge, U.K., 2013, ch. 1, pp. 1–28.
and R. Beckhaus, Eur. J. Inorg. Chem., 2014, 2014, 1289–
1302.
35 A. J. Hunt and T. J. Farmer, in Sustainable Catalysis: With 55 C. Adler, N. Frerichs, M. Schmidtmann and R. Beckhaus,
Nonendangered Metals, Part 1, ed. M. North, Royal Society
of Chemistry, Cambridge, U.K., 2015, ch. 1, pp. 1–14.
36 A. Odom and T. J. McDaniel, Acc. Chem. Res., 2015, 48,
2822–2833.
37 M. Fischer, R. Schaper, M. Jaugstetter, M. Schmidtmann
and R. Beckhaus, Organometallics, 2018, 37, 1192–1205.
38 M. Fischer, M. Jaugstetter, R. Schaper, M. Schmidtmann
and R. Beckhaus, Eur. J. Inorg. Chem., 2018, 5146–5159.
Organometallics, 2016, 21, 3728–3733.
56 A. D. Horton, J. de With, A. J. van der Linden and H. van de
Weg, Organometallics, 1996, 15, 2672–2674.
57 A. D. Horton and J. de With, Organometallics, 1997, 16,
5424–5436.
58 A. G. M. Barrett, M. R. Crimmin, M. S. Hill, P. B. Hitchcock
and P. A. Procopiou, Angew. Chem., Int. Ed., 2007, 46, 6339–
6342, (Angew. Chem., 2007, 119, 6455–6458).
39 S. Kobayashi and H. Ishitani, Chem. Rev., 1999, 99, 1069– 59 J. Choi, D. Y. Wang, S. Kundu, Y. Choliy, T. J. Emge,
1094.
K. Krogh-Jespersen and A. S. Goldman, Science, 2011, 332,
1545–1548.
60 B. M. Kraft, R. J. Lachicotte and W. D. Jones, J. Am. Chem.
Soc., 2001, 123, 10973–10979.
40 S. D. Bull, S. G. Davies, S. Delgado-Ballester, G. Fenton,
P. M. Kelly and A. D. Smith, Synlett, 2000, 1257–1260.
41 S. G. Davies and T. D. McCarthy, Synlett, 1995, 700–702.
42 N. Srivastava and B. K. Banik, J. Org. Chem., 2003, 68, 2109– 61 C. S. Slone, D. A. Weinberger and C. A. Mirkin, Prog. Inorg.
2114. Chem., 1999, 48, 233–350.
43 G. Bartoli, M. Bosco, E. Marcantoni, M. Petrini, L. Sambri 62 Coordination of a fourth substituent to the boron atom of
and E. Torregiani, J. Org. Chem., 2001, 66, 9052–9055.
44 I. Reboule, R. Gil and J. Collin, Tetrahedron Lett., 2005, 46,
7761–7764.
45 H. Firouzabadi, N. Iranpoor and A. A. Jafari, Adv. Synth.
Catal., 2005, 347, 655–661.
46 F. Alonso, Y. Moglie, G. Radivoy and M. Yus, Green Chem.,
2012, 14, 2699–2702.
47 For a detailed description see the ESI.†
48 J. Stroot, D. Haase, W. Saak and R. Beckhaus, Z. Anorg. Allg.
Chem., 2002, 628, 755–761.
49 P. Pyykkö and M. Atsumi, Chem. – Eur. J., 2009, 15, 12770–
12779.
50 A. G. Orpen, L. Brammer, F. H. Allen, O. Kennard,
D. G. Watson and R. J. Taylor, J. Chem. Soc., Dalton Trans.,
1989, S1–S83.
51 M. B. Smith and J. Marchn, Advanced Organic Chemistry,
Wiley, New York, 6th edn, 2007.
52 P. R. Schreiner, L. V. Chernish, P. A. Gunchenko,
E. Y. Tikhonchuk, H. Hausmann, M. Serafin, S. Schlecht,
B(C6F5)3 leads to the pyramidalization of the boron center,
thus the loss of planarity hampers the resonance between
the empty 2p orbital localized at the boron center and the
p-fluorine atoms, whereas the m-fluorine atoms are not
influenced. As a consequence, a significant high-field shift
for the p-fluorine atoms is observed. Thus, the Δδ19Fm,p
value is comparably large in neutral tricoordinated per-
fluorinated boranes and becomes smaller in tetracoordi-
nated borate anions. For more details, see: T. Beringhelli,
D. Donghi, D. Maggioni and G. D’Alfonso, Coord. Chem.
Rev., 2008, 252, 2292–2313.
63 Recent review on metal-catalyzed C–F bond activation:
H. Amii and K. Uneyama, Chem. Rev., 2009, 109, 2119–
2183.
64 (a) J. Zhu, M. Pérez and D. W. Stephan, Angew. Chem., Int.
Ed., 2016, 55, 8448–8451, (Angew. Chem., 2016, 128, 8588–
8591). Recent review on main group Lewis acids for C–F
bond activation: (b) T. Stahl, H. F. T. Klare and
M. Oestreich, ACS Catal., 2013, 3, 1578–1587.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019