A Multifunctional Pincer Ligand Supports Unsaturated Cobalt: Five Functionalities in One Pincer

Article abstract of DOI:10.1002/chem.201700859

A pyridyl pincer ligand was developed to incorporate steric bulk, through a PtBu₂ arm, and proton responsivity, through a pyrazole pincer ligand arm, together with reactivity at benzylic hydrogen and redox activity within a 1,4 diazabutadiene moiety. Binding it to CoCl₂ yielded square-pyramidal [(PNNH)CoCl₂], which was deprotonated by Li[N(SiMe₃)₂] to form [{Li(THF)₂PNN}CoCl₂]. Reduction of this LiCl adduct with KC₈ under CO atmosphere led to formation of Co^I mono- and dicarbonyl complexes, which can be protonated but also further deprotonated at the benzylic CH group to give a dearomatized pyridyl group. The ligand was characterized in its neutral, monoanionic, and dianionic forms, and the anions were shown to exist as intimate ion pairs with Li⁺ bound to pyrazolate N and chloride bound to Lewis acidic cobalt. X-ray photoelectron spectroscopy was used to assay both Li content and cobalt oxidation states. The general character of binding of LiCl to a metal complex acidic at metal and nucleophilic at ligand (pyrazolate Nβ) is discussed, as are potential catalytic applications of the concept.

Full text of DOI:10.1002/chem.201700859

A Journal of
Accepted Article
Title: A Multifunctional Pincer Ligand Supports Unsaturated Cobalt:
Five Functionalities in One Pincer
Authors: Alexander V Polezhaev, Chun-Hsing Chen, Yaroslav Losovyj,
and Kenneth G. Caulton
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.201700859
Link to VoR: http://dx.doi.org/10.1002/chem.201700859
Supported by

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
A Multifunctional Pincer Ligand Supports Unsaturated Cobalt:
Five Functionalities in One Pincer
Alexander V. Polezhaev, Chun-Hsing Chen, Yaroslav Losovyj and Kenneth G. Caulton*[a]
Abstract: A pyridyl pincer ligand is introduced to incorporate steric
bulk, via a PtBu2 arm, and proton responsivity, via a pyrazole pincer
ligand arm, together with reactivity at benzylic hydrogen and redox
activity within a 1,4 diazabutadiene moiety. Binding it to CoCl2 yields
square pyramidal (PNNH)CoCl2, that is deprotonated with
Li[N(SiMe3)2] forming [Li(THF)2PNN]CoCl2. Reduction of this LiCl
adduct with KC8 under CO atmosphere leads to formation of Co(I)
mono- and dicarbonyls, which can be protonated but also be further
deprotonated, at the benzylic CH, leading to a dearomatized pyridyl
functionality. The ligand is characterized in its neutral, monoanionic
and dianionic forms, and the anions are shown to be intimate ion pairs
with Li+ bound to pyrazolate N and chloride bound to Lewis acidic
cobalt. X-ray photoelectron spectroscopy is used to assay both Li
content and cobalt oxidation states. The general character of binding
of LiCl to a metal complex acidic at metal and nucleophilic at ligand
(pyrazolate-N) is discussed, as are potential catalytic applications of
the concept.
FLP is that steric bulk prevents direct interaction between Lewis
acid and Lewis base but substrate can bind between FLP partners.
In H2L-based complexes, Lewis acidic (metal) and Lewis basic
(pyrazolyl-β-nitrogen) sites cannot quench intramolecularly, but
intermolecular binding occurs rapidly, leading to dimers or
polymers. To prevent dimerization, steric bulk must be introduced
in the ligand.
There is currently an immense amount of activity on pincer ligands
containing two bulky phosphine arms connected by an aryl or
pyridyl central donor (Fig. 1), and their modular character allows
systematic modification, for various applications[6]
, including
activation of small molecules[7]. We were especially interested in
combining the features of these two pincer ligands by creating a
ligand with one bulky phosphine arm and one pyrazolate arm
(Fig. 1) in an effort to prevent dimerization.
N
R
N
N
N
NH
PR2
PR2
HN
Introduction
We are targeting multielectron and multiproton reduction
processes for transformation of hard-to-reduce molecules, such
as N2 and CO2 into valuable chemicals[1]. It was earlier shown by
Floriani[2] for CO2 and recently by Holland[3] for N2 that alkali metal
cations can play an important role in such a process, acting as a
templating element in multimetallic reactions. We have been
exploring a bis-pyrazole pyridine ligand (H2L) (Fig. 1) and have
begun to establish the general characteristics of this diprotic acid
as follows: It binds well in the tridentate meridional fashion, and is
proton responsive via the protons on the  nitrogens[4]. When
deprotonated, the anionic pyrazolate becomes a peripherally
directed nucleophile, coordinating to alkali metal cations, but also
N
PtBu2
N
N
H
PNNH
Figure 1. Combining two functionalities in one ligand.
Additional features of phosphorus incorporation can be easily
recognized including 31P NMR spectroscopy for reaction
monitoring of diamagnetic species, and the two tBu groups allow
sensing any mirror plane of symmetry containing the pincer plane,
a feature absent in the bis-pyrazolyl pyridine ligand. We hoped to
synthesize metal complexes in varied oxidation states/geometries
and the flexibility of the phosphine arm could help to achieve
geometries impossible for rigid bis-pyrazolyl pyridine. An unusual
feature of the PNN-1 ligand is that an anionic center is located in
the arm, not in a core of the ligand like in majority of monoanionic
pincers, positioning it cis to incoming substrate.
The PNNH ligand contains two acidic protons: one on the -
nitrogen of pyrazole ring and the second one on the benzylic
position. These have very different acidity, so could be removed
stepwise (Scheme 1a). Removal of pyrazolic protons creates a
negative charge on the ligand and also a strong donor site on the
-nitrogen that could carry alkali metal cation or interact with a
second transition metal with formation of M2 or MM’ species. The
benzylic hydrogens which link the phosphorus to the pyridine ring
are particularly reactive and removal of one of these as a proton
unmasks anionic charge which is also stabilized at the pyridine
readily bridging to
a second transition metal, leading to
aggregation. In the event that the metal in a bis-pyrazolate pyridyl
complex is unsaturated, then it combines, in close proximity, both
Lewis acid and Lewis base character, with the potential reactivity
of a single molecule Frustrated Lewis Pair (FLP) (i.e binding both
nucleophile and electrophile); this concept has already been
recognized for some metal-ligand interactions[5]. The idea behind
[a]
Dr. A. V. Polezhaev, Dr. C.-H. Chen, Dr. Yaroslav Losovyj,
Prof. Dr. K. G. Caulton
Department of Chemistry
Indiana University Bloomington
47405, Bloomington, IN (USA)
E-mail: caulton@indiana.edu
Supporting information containing CV, XPS, NMR, IR spectra, and
crystallographic information for this article is given via a link at the
end of the document
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
nitrogen (Scheme 1b). In sum, the electronegative nitrogen of
pyridine is much more potent for encouraging these cooperative
or bifunctional behaviors than would be true if that were a phenyl
ring[8]. Considering also the P and two types of N lone pair donors,
the PNNH ligand thus incorporates an unusually large number of
functionalities. Both of these deprotonations are redox neutral at
the metal, but have the added benefit of creating additional
unsaturation at the metal, via chloride abstraction, as envisioned
in Scheme 1a.
Seeing one signal for the two tBu groups on phosphorus suggests
(incorrectly) a mirror plane relating those two. An ion of m/z 453.1
with characteristic isotope distribution appears in the positive
mode of both ESI and APCI mass spectrum, corresponding to
[(PNNtBuH)CoCl]+. Metallation of PNNPhH was performed under
the same conditions and spectroscopic data (Figure 2) are very
similar for both [PNNtBuH]CoCl2 2a and [PNNPhH]CoCl2 2b.
CoCl2
N
N
R
R
tBu2P
N
N
tBu2P
N
N
THF, r.t. 12h
Co
1a, 1b
H
H
1 eq. base
- HCl
1 eq. base
-HCl
R
R
R
N
N
N
Cl
2a R = tBu; 2b R = Ph
Cl
R2P
N
R2P
N
R2P
N
N
NH
N
M2+
M2+
Cl
M2+
a)
Cl
Cl
Scheme 3. Metallation of PNNRH 1a and 1b with CoCl2.
- H+
N
N
R
R
tBu2P
b)
N
N
tBu2P
N
N
R
R
Single crystal X-ray diffraction of a crystal of (PNNtBuH)CoCl2 1a
grown by slow diffusion of pentane vapors in concentrated THF
solution at -45oC enhances what is learned from spectroscopy.
The molecule crystallizes with one lattice guest molecule of THF,
whose oxygen hydrogen bonds to the NH proton (Figure 3). The
molecule is square pyramidal, 5 = 0.3, where a value of 0 is
square pyramid and 1 is trigonal bipyramidal[9]. The two chlorides
are therefore chemically inequivalent, with one located trans to
the empty coordination site. The two tBu on P are also
Scheme 1. Proposed stepwise deprotonation of PNNH ligand.
We report here on our initial observations with such a ligand on
cobalt in different oxidation states. We thus challenge a three-
coordinate Co(I) complex to additionally coordinate both
nucleophile and electrophile and explore the redox and acid/base
chemistry of designed ligand.
t
inequivalent. One phosphine Bu group is directed towards that
empty site, but the Co/C distance, >3.7 Å, shows no bonding, but
simply filling otherwise empty space. Another indication of no
agostic interaction is that both Co-P-C(quaternary) angles are
similar at ~ 118o. The Co/N distance to pyrazole is shorter than
that to pyridine.
Results and Discussion
Synthesis and characterization of CoII Complexes
The ligands with phenyl and tert-butyl groups on C5 pyrazole
carbon were synthesized to manage solubility; all phenyl
derivatives showed lower solubility than their tert-butyl analog.
The synthesis was performed in 3 steps with excellent yields
using
Claisen
condensation
with pinacolone
of
or
2-methyl-6-methoxycarbonylpyridine
acetophenone followed by hydrazine treatment with formation of
pyrazole ring and subsequent treatment of doubly deprotonated
heterocycle by tBu2PCl (Scheme 2). The structure of products was
confirmed by 1H, 31P and 13C NMR.
1) BuLi (2.2 eq.)
1) NaH / RCOCH3
O
N
N
N
R
2) tBu2PCl
3) H2O
R
2) HCl
3) N2H4*H2O / EtOH
tBu2P
1a,
N
N
N
O
NH
H
R = tBu
1b, R = Ph
Scheme 2. Synthesis of PNNH ligands.
Figure 2. 1H NMR spectra of (PNNtBuH)CoCl2 2a and (PNNPhH)CoCl2 2b in
CD2Cl2.
Reaction of equimolar CoCl2 and PNNtBuH 1a in THF occurs in 12
h at r.t. After concentration of the reaction mixture and addition of
Et2O, a deep violet precipitate was separated (Scheme 3). Its 1H
NMR spectrum in CD2Cl2 shows, with correct intensities, signals
for all ring proton environments, including that of the NH proton,
The ground state structure found shows that the nonrigidity
characteric of five coordinate species pertains here: a simple
intramolecular rearrangement which puts the axial Cl trans to
pyridine nitrogen while the second chlorine moves to an apical
position is a low energy process, and dynamically averages the
t
and 18:9 intensity signals for two types of Bu groups. Given the
chemical shift range -2 to 74 ppm, the molecule is paramagnetic.
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
t
environments of the two Bu groups on phosphorus, giving only
The 1H NMR spectrum of 3a in THF-d8 shows, with correct
one tBu 1H NMR chemical shift. Another feature which
distinguishes this ligand from its C2-symmetric variants with either
two pyrazolates or two phosphines is its nonplanarity: cobalt does
not lie in the plane of the three donor atoms, although it does lie
in the plane of the pyridine and pyrazole rings. This is also evident
in the nonplanarity of the ring containing phosphorus and also the
small angle N3Co1P1, which has a value of only 95o. A THF
molecule hydrogen bonds to a pyrazole β-nitrogen proton (3.139
Å N1/O1) typical of moderate hydrogen bonds[10] and also
previously observed between THF and other H-pyrazole
intensities, signals for all ring proton environments and 18:9
intensity signals for two types of Bu groups in the chemical shift
range -2 to 67 ppm. 1H NMR of 3b in the same solvent shows a
very similar pattern with exception of an additional set of three Ph
group signals in 2:2:1 ratio and the absence of an intensity 9H
signal.
t
Single crystals of [Li(THF)2(PNNtBu)]CoCl2 were grown by slow
diffusion of pentane vapors into a concentrated THF solution at -
45oC and analyzed with X-ray diffraction (Figure 4). The metal
coordination sphere is similar to the parent compound
(PNNH)CoCl2 with 5 = 0.39 showing geometry closer to square
pyramidal. Both metals occupy the plane of the aromatic rings of
the ligand. The phosphorus lies out of this plane and has
inequivalent tBuP groups with no signs of agostic interactions with
Co1. The distance Co1-N2 is significantly shorter (by 0.044 Å) and
the distance Co1-Cl2 is longer (by 0.064 Å) than in (PNNH)CoCl2.
Lithium is coordinated to pyrazolyl nitrogen N1, to Cl2 and to two
THF molecules, together forming distorted tetrahedral geometry
around the alkali metal. The main distortion from tetrahedral Li+ is
containing complexes[4, 11]
.
t
caused by steric clash between pyrazolyl Bu and one of the
coordinated THF. Overall, this shows that the first product in
Scheme 1a is oversimplified and dehydrohalogenation is not the
result. Instead a “proton to lithium metathesis” occurs and LiCl is
retained, to satisfy both Lewis acid (Co) and Lewis basic
(pyrazolyl β-nitrogen) sites.
Figure 3. Mercury view (50% probability) of the nonhydrogen atoms of
(PNNtBuH)CoCl2 2a, showing selected atom labelling; unlabeled atoms are
carbon. Selected structural parameters: Co1-Cl1, 2.307; Co1-Cl2, 2.287; Co1-
P1, 2.480; Co1-N2, 2.079; Co1-N3, 2.207; Cl1-Co1-Cl2, 111.46; Cl1-Co1-P1,
104.43; Cl2-Co1-P1, 103.57; Cl1-Co1-N2, 91.26; Cl2-Co1-N2, 108.30; Cl1-
Co1-N3, 154.61; Cl2-Co1-N3, 92.97. tBu groups on phosphorous atom are
shown as wireframe, for clarity.
Pyrazole Deprotonation.
Treatment of violet slurry of (PNNRH)CoCl2 2a or 2b with
equimolar LiN(SiMe3)2 (LiHMDS) in THF at -78oC results in
immediate color change from violet to blue. After warming up to
room temperature and removal of the solvent, blue residue was
obtained in nearly quantitative yield. The ESI+ mass spectrum of
THF solution of the product 3a shows ions at 905.2 and 911.1 m/z
corresponding to dimers of formula ([(PNNtBuH)CoCl]2+H)+, and
([(PNNtBuH)CoCl]2+Li)+, respectively and a small peak at 949.1
m/z corresponding to ([((PNNtBuH)CoCl]2+LiCl+H)+ (all with
characteristic isotope patterns due to 35Cl and 37Cl natural
abundance) proving that LiCl is a part of the complex. The product
is not simply the dehydrohalogenation species (PNNtBu)CoIICl.
Dicobalt species in the mass spectrum indicate an unexpected
tendency of lithium to aggregate multiple ligand/cobalt moieties.
Figure 4. Mercury view (50% probabilities) of the nonhydrogen atoms of
(Li(THF)2(PNNtBu))CoCl2 3a, showing selected atom labelling; unlabeled atoms
are carbon. Selected structural parameters: Co1-N2 2.034; Co1-N3 2.163; Co1-
P1 2.466; Co1-Cl1 2.301; Co1-Cl2 2.363; Li1-Cl2 2.367; Li1-N1 1.678. Two THF
molecules coordinated with Li1 and tBu groups on phosphorous atom are shown
in wireframe, for clarity.
The molecule (LiPNN)CoCl2 can be presented in two resonance
structures (Scheme 5) depending on which pyrazole ring nitrogen
is considered as amidic. Chlorine Cl2 is equidistant from the two
metals (Co1-Cl2 2.363; Li1-Cl2 2.367).
LiHMDS
N
N
R
R
tBu2P
Cl
N
N
tBu2P
Cl
N
N
THF, -78oC
Co
Co
H
t
3
2
a
R = Bu
N
N
Li(THF)2
R
Cl
Cl
R
tBu2P
Cl
N
b R = Ph
tBu2P
Cl
N
Co
N
Co
N
Li
Scheme 4. Deprotonation of (PNNRH)CoCl2 2a and 2b.
Cl
Li
Cl
Scheme 5. Two resonance forms of (LiPNN)CoCl2, as pyrazolate bridges CoII
and Li+.
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
Deprotonation could also be performed with KHMDS and
NaHMDS, with formation of the K and Na analogs than show
lower solubility in THF but no difference in NMR spectra
compared to Li analogs. Deprotonation of (PNNPhH)CoCl2 2b with
LiHMDS was performed under the same conditions, and the
product showed spectroscopic data in agreement with the
proposed structure (Fig. S14).
at 203 ppm with no resolved coupling to phosphorus. We
anticipated the green solid to be
a square-planar Co(I)
monocarbonyl, as typical for Co(I) complexes with monoanionic
pincer ligands[12] and the red product to be dicarbonyl adduct
(Scheme 6).
N
N
N
CO/KC8
vacuum
CO
R
R
R
tBu2P
N
Redox Chemistry of 2 and 3.
tBu2P
OC
red
N
N
tBu2P
N
Co
CO
N
Co
Co
N
THF, rt
- KCl
A goal of multielectron catalysis stimulated brief study of the
electrochemical behavior of both 2a and 3a. Cyclic voltammetry
of (PNNtBuH)CoCl2 2a showed (Fig. 5) one quasireversible
oxidation at 0.55 V associated with Co2+/Co3+ and two irreversible
1ē reductions at -1.99 and -2.81 V vs. Fc/Fc+. CV of complex 3a
was measured in 0.1 M TBAPF6 electrolyte containing 0.5 M of
LiCl to resolve reduction peaks and showed quasireversible
oxidation and irreversible reduction waves. The oxidation peak is
shifted by 0.35 V to more positive potential, in agreement with
weakened donation from -Cl-. Reduction of 3a occurs at almost
0.5 V less negative potentials than for 2a which made compound
3 our candidate for chemical reduction.
CO
5a,b
Cl
Cl
Li(THF)2
green
4a,b
3a,b
Scheme 6. Reduction of [Li(THF)2PNN]CoCl2 3a and 3b under CO atmosphere.
Single crystal X-ray diffraction of the green monocarbonyl 4b
established (Figure 6) that it is (PNNPh)Co(CO) and confirms four
coordinate planar (4 = 0.1) unsaturated Co(I) with no significant
short intermolecular contacts involving the metal or the carbonyl.
It also confirms the retention of two benzylic hydrogens, and the
pyrazolate CNN angle, 106.4o confirms that this nitrogen is
deprotonated; moreover, there is no evidence for hydrogen
electron density near N in the final difference map. The Co-N
distance to pyrazolate is shorter than that to pyridine, and the Co-
C and C-O distances, 1.704 and 1.154 Å, respectively, reveal
significant weakening of the CO bond. The phenyl ring has an
angle of 29.8o to the pyrazolate ligand.
Figure 5. CV (1.5 to -3.2 V) of (PNNtBuH)CoCl2 2a (bottom) in THF/ 0.1 M
[NBu4]PF6 at 100 mV/s. and [Li(THF)2PNNtBu]CoCl2 3a (top) in THF/ 0.1 M
[NBu4]PF6, 0.5 M LiCl at 100 mV/s. (CV without LiCl addition: see Fig. S60)
We used CO as a surrogate substrate for Co(I) here and as a
spectroscopic probe to evaluate donor properties of different new
ligands in their Co(I) complexes. Addition of 1 eq. of KC8 to
[Li(THF)2PNN]CoCl2 3a in THF under 1 atm of CO results in
immediate color change from blue to red. After stirring for an
additional 4 h at r.t., vacuum removal of unreacted CO caused
immediate color change from red to green. Filtration from graphite
Figure 6. Mercury view (50% probabilities) of the nonhydrogen atoms of
(PNNPh)CoCO, showing selected atom labelling; unlabeled atoms are carbon.
tBu groups on phosphorous atom are shown without probability ellipsoids for
clarity. Selected structural parameters: Co1-N2 1.899; Co1-N3 1.949; Co1-P1
2.156; Co1-C24 1.704; C24-O1 1.154; P1-Co1-N2 166.3; N3-Co1-C24 179.00;
Co1-C24-O1, 178.76.
1
and removal of the solvent yields green glassy solid. H NMR of
Surprisingly, IR spectra of freshly isolated 4a or 4b in THF each
show two CO stretches, at 1885 and 1933 cm-1, and always in
different intensity ratio. When 4a was stirred in Et2O for 3 days,
the stretch at 1885 disappears completely and formation of white
precipitate was observed. When crystals of 4b from the same
batch that was analyzed by X-ray diffraction were dissolved in
THF, only the 1933 cm-1 band was observed, consistent with the
absence of LiCl. Altogether this suggests the reversible formation
of a LiCl adduct, with equilibrium populations influenced by partial
removal of LiCl during multiple solvent manipulations.
the green solid shows signals in a range 1 – 8 ppm, indicating
diamagnetism and no evidence of paramagnetic species from -
200 to 200 ppm; 31P NMR shows one broad singlet at 105 ppm
and 13C NMR showed a set of signals from PNN ligand but no
signal from carbonyl carbon, probably because of broadening by
quadrupolar 59Co. Addition of 1 atm of CO to a green solution
causes immediate color change back to red and 31P NMR shows
a 15 ppm downfield shift of the singlet, all confirming the
equilibrium in Scheme 6. 1H NMR indicates minor changes in
chemical shifts of PNN ligand protons and 13C NMR shows a set
of signals from PNN ligand and one broad signal from CO carbon
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
To test this hypothesis, an excess of solid LiCl was stirred with
compound 4a in THF and monitored by solution IR for 48h (Fig.
LiHMDS to form only a single carbonyl complex which has almost
the same CO stretching frequency, but is due to pure 4.
7).
If the equilibrium mixture is instead treated with HCl∙Et2O, it
converts to one monocarbonyl complex with frequency 1883 cm-
(which rapidly precipitates) and this can be deprotonated with
1
KC8/ CO
LiHDMS to give a species which shows primarily the 1885 cm-1
carbonyl frequency. This is best understood with the reactions in
Scheme 7, where replacement of NH by NLi, with retention of
chloride, leaves the CO stretching frequency nearly unchanged.
All of this suggests that the two cobalt(I) monocarbonyl species
are isomeric four coordinate Co(I) with free chloride counterion,
and coordinated chloride in the five-coordinate saturated Co(I).
Binding of chloride is apparently promoted in this case by the
cationic charge of the four-coordinate cobalt complex, and
certainly its lower CO frequency is consistent with better  basicity
in the saturated cobalt, but it is interesting that the adduct
formation constant is modest, able to be influenced by
concentration and sample history. The ready loss of CO from
(PNN)Co(CO)2 5 in Scheme 6 also shows the limited Lewis acidity
of unsaturated Co(I) here. It is noteworthy that the CO frequencies
of H and Li analogs are extremely similar. The low solubility of the
protonated chloro complex (PNNH)CoCl(CO) 4a*HCl is attributed
to intermolecular hydrogen bonding between pyrazole H and
chloride.
Cationic complex 4a*HBF4 (Scheme 7) was isolated and
characterized with IR, NMR and X-ray diffractometry. The IR of
the product shows one stretch at 1935 cm-1 and 1H NMR shows a
set of signals for PNNtBuH ligand including an acidic NH proton at
13 ppm (Scheme 8); one 31P NMR singlet confirms formation of a
single product. The identity of the NH proton was proven by 1H/13C
HMBC NMR experiment, where the signal at 13 ppm in 1H
spectrum has a cross-peak with the 13C NMR signals at 101.4,
154.6 and 159.7 ppm, that were independently assigned as
pyrazole ring carbons. The carbonyl 13C NMR signal appears at
201.8 ppm, and BF4 appears in 19F NMR as a singlet at -156.79
ppm. Single crystals of (PNNtBuH)Co(CO)BF4 were grown by slow
diffusion of pentane vapors into a concentrated THF solution at -
45oC and analyzed with X-ray diffraction (Figure 8).
- LiCl
+LiCl
N
N
N
R
R
R
tBu2
P
N
tBu2
P
N
N
tBu2
P
N
N
Co
Co
N
-1 KCl
Co
4*LiCl
4
3
Cl
Cl
Li(THF)2
Li(THF)2
Cl
CO
1885 cm-1
CO
1933 cm-1
Chloride-free
path
LiHMDS
HCl/Et2O
LiHMDS
-LiBF4
HBF4
Chloride
present
THF
Et2O
BF4
N
N
R
R
tBu2P
N
N
tBu2P
N
N
H
Co
Co
H
Cl
4*HCl
4*HBF4
CO
1883 cm-1
CO
1935 cm-1
Scheme 7. Protonation and deprotonation of (PNNtBu)Co(CO) 4 with two
different acids.
The band at 1933 cm-1 significantly decreases during reaction
time and 1885 cm-1 stretch increases, in agreement with the
equilibrium in Scheme 7 shifting to the right. When a THF solution
of compound 4a was mixed with preformed 0.5 M LiCl THF
solution, IR detected the equilibrium established immediately and
no changes of the spectra were observed after 12h. After addition
of 1 atm of CO to a green solution of pure (LiCl-free) 4a or 4b, to
form the red dicarbonyl, two IR bands at 1940 and 1994 cm-1 were
observed, in agreement with the predicted structure. When LiCl
adduct 4a∙LiCl was exposed to CO, three bands were observed
at 1884, 1940 and 1994 cm-1. This indicates that two species 5a
and 4a*LiCl are present in solution and in fast equilibrium. The
fact that the mixture of 4 with 4*LiCl adduct or even together with
5 under CO atmosphere has only a single phosphorus NMR
signal and also a single set of proton NMR signals shows that this
equilibrium is rapid on the NMR timescale and is certainly rapid
on the synthetic and laboratory timescales.
Figure 7. IR monitoring of reaction mixture of 4a and LiCl in THF.
Figure 8. Mercury drawing (50%) of the nonhydrogen atoms of
[(PNNH)Co(CO)]BF4●THF 4a*HBF4, showing selected atom labelling.
Unlabeled atoms are carbons, THF molecule is omitted, tBu groups on
phosphorous atom are shown without probability ellipsoids for clarity. Selected
structural parameters: Co1-P1, 2.1738(19); Co1-N1, 1.954(5); Co1-N2,
1.939(5); Co1-C22, 1.731(7); O1-C22, 1.130(7); P1-Co1-N1, 84.32(15); P1-
Co1-N2, 164.18(16); N1-Co1-N2, 80.5(2); P1-Co1-C22, 94.2(2); N1-Co1-C22,
178.3(3); N2-Co1-C22, 100.8(3).
To further support the LiCl coordination hypothesis, a series of
experiments (Scheme 7) were designed to create contrasting
chloride-free and chloride-rich environments. Treatment of the
mixture of 4a*LiCl (nearly 1:1 ratio of 1885 and 1933 cm-1 bands)
with HBF4 yields a solution which contains a single carbonyl
frequency, 1935 cm-1. This species was then deprotonated with
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
Cobalt is square planar with 4 = 0.07. All metal heteroatom
distances are shorter (Co-P by 0.306 Å, Co-Npy by 0.253 Å, and
Co1-Npz by 0.139 Å) compared to [PPNH]CoCl2 2a, consistent
with a low-spin state of Co(I). The phosphine has two equivalent
tBu groups with no signs of agostic interactions with Co1. The NH
hydrogen forms a hydrogen bond with F4 of the counterion
(N3•••F4 = 2.761 Å).
in Et2O shows one band at 1851 cm-1, a significant red shift, in
agreement with pyridine nitrogen (trans to carbonyl) now being
amidic compared to imine in previous examples: push/pull
interaction increases back donation to CO.
N
KHMDS
Cationic 4a*HBF4 reversibly binds another CO molecule when
exposed to 1 atm of CO (Scheme 8). The reaction mixture turns
from deep green to red in the same manner as for [PNNtBu]Co(CO)
N
R
tBu2P
N
N
R
tBu2P
Cl
N
Co
Co
N
Et2O
-KCl
Li(THF)2
4a and IR showed two new bands at 1944 and 2007 cm-1
.
CO
6a,b
Li(THF)2
CO
4*LiCl
Removal of CO atmosphere reverses the process. 1H NMR of the
CO adduct showed significant broadness of all peak
corresponding to compound 5a*HBF4 (but no broadness of
solvent residual peaks) indicating that CO binding rate is closer to
NMR timescale than in case of neutral 4a.
Scheme 9. Dearomatization of PNN ligand in its Co(I) carbonyl complex.
Single crystal X-ray diffraction study of 6a (Fig. 9) confirmed what
was established by spectroscopy but with the surprise that the
alkali metal is lithium (not K from the KHMDS), and the lithium
binds to pyrazolate nitrogen (2.005 Å), but also to carbonyl
oxygen (2.394 Å) and to two THF molecules (1.922; 1.916 Å).
There are no other short packing contacts with the carbonyl
oxygen. The coordination geometry around cobalt(I) is planar,
and only the lithium lies slightly out of that plane. The pattern of
CC and CN distances (Fig. 10) confirms that the benzylic
hydrogen has been removed by N(SiMe3)2-1, and that pyridine
aromaticity has been interrupted. The resulting amide nitrogen
lone pair can only interact constructively with a d orbital by a
push/pull interaction involving one CO * orbital with the two
occupied Co and amide  orbitals. The Co-C distance is indeed
0.021 Å shorter than in [PNNPh]Co(CO) 4b and 0.048 Å shorter
than in [(PNNtBuH)Co(CO]BF4 4a*HBF4 and the Co-N(amide)
distance is not particularly short. The CO distance is 0.017 Å
longer in [Li(THF)2PNN*]Co(CO) 6a than in [PNNPh]Co(CO) 4b
and 0.041 Å longer than in 4a*HBF4, in agreement with a higher
degree of back donation and consequently 84 cm-1 lower CO
frequency. Lithium binds two THF molecules and pyrazolate N,
allowing additional interactions between cobalt-bound substrate
and Li+.
N
N
tBu
CO
tBu
tBu2P
N
tBu2P
N
Co
N
Co
N
vacuum
OC CO 5a
4a
CO
Et2O*HBF4
LiHMDS
Et2O*HBF4
BF4
LiHMDS
-
-
BF4
N
N
tBu
tBu
CO
tBu2P
N
tBu2P
N
Co
Co
N
H
N
H
vacuum
OC CO
CO
5a*HBF4
4a*HBF4
Scheme 8. Protonation and deprotonation of mono and dicarbonyl Co(I)
adducts.
The 1H and 31P NMR spectra show broad peaks shifted upfield
compared to the parent monocarbonyl, attributed to equilibrium
between mono and dicarbonyl species. Treatment of both mono
and dicarbonyl complexes with 1 eq. of LiHMDS (Scheme 8)
removes HBF4 with formation of neutral [PNNtBu]Co(CO) 4a and
[PNNtBu]Co(CO)2 5a, respectively; with no signs of LiBF4 retention,
protonation is thus fully reversible. The absence of 19F NMR
signals in the isolated product after deprotonation indicates that
LiBF4 has been removed completely.
Benzylic Deprotonation.
Methylene protons on the phosphine arm are known to be acidic
because of resonance stabilization of conjugate base by the
pyridyl nitrogen. After treatment of 4a or 4b with 1 eq. of KHMDS
in Et2O or THF at r.t., a color change from green to violet was
observed and 31P NMR indicates consuming of starting material
and formation of a single product (singlet at 99.2 ppm for 6a
1
Scheme 9). 1H NMR shows a singlet with intensity of H at 2.96
Figure 9. Mercury drawing (50%) of the nonhydrogen atoms of
[Li(THF)2P*NNtBuCo(CO)] 6a, showing selected atom labelling. Unlabeled
atoms are carbons. tBu groups on phosphorous atom and two coordinated THF
molecules are shown without probability ellipsoids for clarity. Selected structural
parameters: Co1-P1 2.185; Co1-N2 1.908; Co1-N3 1.926; Co1-C22 1.683; C22-
O1 1.171; N1-Li1 2.005; Li1-O1 2.294; C8-C9 1.369; C8-N3 1.390; P1-Co1-N2,
166.0; N3-Co1-C22, 176.4; N2-Co1-C22, 96.2.
ppm indicating removal of one methylene proton (we designate
this CH2-deprotonated dianionic ligand as PNNR*) accompanied
by significant downfield shift of pyridyl protons associated with
loss of aromaticity. 13C NMR and 1H/13C HMBC and HMQC
confirms the structure proposed for [Li(THF)2PNNR*]Co(CO). IR
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
Collectively, these results show remarkable selectivity for
retention of the Li+ in competition with the larger alkali metal K+.
This may foretell a thermodynamic preference which could be
used productively in future studies with this alkali metal/ligand
transition metal partnership, enabled by pyrazolate.
are interested in detecting the presence or absence of LiCl in 4
and 4*LiCl; we measured XPS of two samples, one
a
homogenous mixture of 4a and 4a*LiCl (based on solution IR)
and one a crystalline sample of pure 4b. To facilitate analysis of
BE, compounds 2a and 3a were also analyzed. Spectra of 4b
showed both Cl and Li are nearly absent, consistent with
molecular structure data. For 4a*LiCl, the measured Co/Cl ratio
was almost 2:1 (see table 1) that agrees with 1:1 mole ratio in the
mixture of 4a:4a*LiCl. The measured Co/Li ratio is almost 4:1 but
much lower sensitivity of XPS to Li atoms compared to Cl makes
this number less reliable than the Co/Cl ratio, and the presence
of Li in the sample was confirmed. Potassium is below detection
limit on this sample, confirming that the reducing agent is not
incorporated in this sample in the reduction process. For all
samples, the measured Co to Cl, P, and N atomic ratios confirmed
the proposed formulas (see Table 1).
Comparison of the bond lengths within the NNPCoCO unit of
[Li(THF)2PNN*tBu]Co(CO)
6a
to
those
in
[(PNNtBuH)Co(CO)]BF4●THF show the largest changes to be
shortening of the CoC and lengthening of the CO distances on
going from Li+ to H+ as electrophile on pyrazolate N.
-0.026
0.071
0.036
0.047
-0.148
N
-0.079 -0.027
N-0.02
X
P
N
Co -0.031
-0.048
C
Table 1. Atomic ratios in % measured (calculated) for cobalt complexes 2a,
3a, 4a*LiCl and 4b
O0.041
Co/P
Co/N
Co/Cl
Co/Li
(50:50)
(25/75)
Figure 10. Significant differences (more than 0.020 Å) of bond lengths of
[Li(THF)2PNNtBu*]Co(CO) 6a and ([PNNtBuH]Co(CO)BF4 4a*HBF4. Numbers
obtained by subtraction of bond lengths in Å of 4a*HBF4 from corresponding
bond length in 6a.Lewis structure is from the benzylic deprotonated PNN* form,
showing bond length alternation after dearomatization.
2a
3a
52/48
55/45
52/48
52/48
26/74
26/74
28/72
27/73
36/63 (33/67)
30/70 (33/67)
66/34
47/52 (50/50)
80/20
4a*LiCl
4b
93/7 (100/0)
91/9 (100/0)
Proton transfer Reactivity: Identifying basic sites.
Treatment of [Li(THF)2PNNtBu*]Co(CO) 6a with H2O in Et2O or
THF results (Scheme 10) in immediate color change from violet
to green and IR indicates one stretch at 1935 cm-1 corresponding
to [PNN]Co(CO) 4a. The selectivity of this proton transfer was
established by the analogous addition of D2O, which initially
shows half intensity CHD signal, but subsequent disappearance
of methylene proton due to slower multiple H/D exchange (Fig.
S57-58)
In addition (Table 2), cobalt binding energies for compound 2a
and 3a are in the range known for Co(II) complexes[15] and
accompanied by shake-up peaks characteristic of Co(II). Binding
energies of monovalent cobalt are expected to be ~1 eV lower
than for Co(II). We obtained Co(I) BE by deconvolution of broad
spectra (see SI for details) and Co2p3/2 BE of 4a and 4b agrees
with this prediction. Interestingly, the N1s BE of 2a appears as
two peaks in 2:1 ratio, assigned as NH pyrazolic nitrogen having
higher BE. After deprotonation, N1s peak of 3a shifts to lower BE
(~0.3 eV), in agreement with ligand negative charge. Compound
3a contains two inequivalent Cl atoms according to
crystallographic data; they cannot be resolved in XPS but both
Cl2p components are significantly (~0.7 eV) shifted to higher BE
compared to 2a. Presumably two THF molecules coordinated to
Li were lost in the XPS instrument high vacuum, resulting in a
polymeric structure with bridging lithium coordinated to several Cl
atoms. Vacuum dried 3a is almost insoluble in non-coordinating
solvents; NMR spectra in CD2Cl2 following vacuum treatment
showed evidence of aggregation (multiple species) and the
absence of THF signals in the diamagnetic region, but after
removal of CD2Cl2 and redissolving in THF, a single species was
detected, supporting the occurrence of THF loss in UHV.
H
D2O
N
D
N
R
R
tBu2P
N
N
tBu2P
N
N
Co
Co
-LiOD
6a
Li(THF)2
4a-D
CO
CO
Scheme 10. Reaction of [Li(THF)2PNN*]Co(CO) with D2O/H2O.
X-ray photoelectron spectroscopy (XPS): elemental
composition and oxidation states
XPS measures the energy and intensity of core electrons of
selected elements ejected by monochromatic X-rays. Intensity of
ejected photoelectrons allow quantitation of each element, in
terms of atom ratios, and binding energy, BE (ionization potential),
is related to the electron richness, hence oxidation state of the
element involved: higher BE correlates with higher oxidation state,
hence more positive local charge at that element[13]. XPS was
recently applied for better understanding of ligand-noninnocent
behavior in pincer complexes of iron[14]. In the present case, we
Table 2. Binding energies for cobalt complexes 2a, 3a, 4a*LiCl and
4b (see Fig. S61-64 for details).
Co2p3/2
eV
N1s eV
P2p
eV
Cl2p3/2; Cl2p1/2
eV
O1s
eV
2a
780.31
399.49+400.96
in 2:1 ratio
130.87
197.87; 199.35
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
3a
4a*LiCl
4b
780.20
779.31
779.44
399.19
399.05
399.16
130.77
130.72
130.75
198.54; 200.22
197.89. 199.44
Several cobalt pincer complexes with monoanionic pincer ligands
were reported in the last decade (Fig. 11)[12b-g, 17]. All of them
contain amidic nitrogen trans to carbonyl, resulting in relatively
low carbonyl bands 1863-1901 cm-1. When neutral pyridine
nitrogen is trans to CO as in Chirik’s (PDI)Co(CO), almost 100 cm-
1 higher stretch was observed. In our case in monoanionic PNN1-
the negatively charged pyrazolyl α-N is cis to carbonyl and
moderate 1933 cm-1 shift of CO stretch was observed but second
deprotonation to PNN*2- dramatically increases back donation
from metal to carbonyl to 1856 cm-1. This makes PNN ligand able
to manage the degree of substrate-complex interaction by
addition or removal of proton on the ligand.
This work also shows that Li+ is not simply a proton equivalent (cf.
study of the “methane analog” CLi4) but has productive capability,
and not merely involving aggregation to (LiX)n[18]. The retention of
LiCl described here can be rephrased by considering the LiCl as
a temporary placeholder ligand, subject to replacement by any
substrate which binds better than chloride of LiCl. Our lack of
recognition of this in Scheme 1 apparently originates from the fact
that four coordinate PNNCoIICl and three coordinate (PNN)CoI
are sufficiently unsaturated that they retain chloride nucleophile,
and then Li+ is retained for charge balance. The frequent retention
of Li+ raises the question of its selectivity vs. other alkali metal
cations, specifically K+ here. The preference for Li+ vs.
deprotonating base cation K+ in the conversion to
[Li(THF)2PNNtBu*]Co(CO) 6a is remarkable: potassium is both
larger than Li+ and demands a higher coordination number. But it
is also surprising that (PNN)Co(CO)*LiCl retains LiCl, since this
four coordinate planar species 4 seems an attractive alternative
which might not be Lewis acidic towards LiCl. It remains to be
seen whether pyrazolate-bound K+ or Li+ has greater transforming
impact on substrate bound locally. CO here serves as a surrogate
to future substrates, and reveals that an electrophile on pyrazolate
N indeed interacts with ligands bound trans to pyridine N. Said
differently, pyrazolate here demonstrates its high affinity for
bridging two metals, including the present heterobimetallic Co/Li+
pair, and this can be a productive feature to further interact with
substrate bound to a neighboring transition metal (Fig. 12).
531.6
531.9
Conclusions
Dehydrohalogenation of alkyl halides is the reaction which leads
to olefins, hence unsaturation, and this is the logical link to use of
the term unsaturation in metal complex chemistry, where
IrHCl2(CO)(PR3)2 was treated with base to give the 16 valence
electron, hence “unsaturated,” Vaska complex, IrCl(CO)(PR3)2.[16]
When base is reacted with a metal ligand complex which contains
an acidic proton on a ligand, then net removal of H and Cl yields
a metal complex of unchanged oxidation state, but unsaturated at
metal, due to removal of halide. This envisioned product is of
special interest since the unsaturation at M is accompanied by
Lewis basicity at ligand. If that lone pair is remote from M, then
dehydrohalogenation has created a metal complex which is
simultaneously a Lewis acid, at M, and a Lewis base, via the
conjugate base lone pair (Scheme 1a). This is an attractive
bifunctional species, but hinges on whether the chloride will
actually be removed from the metal, or will be retained, in order
to avoid Lewis acidity.
The acidity of the benzylic hydrogens here is exceptional because
metal-coordinated P and pyridine are both electron withdrawing
substituents on CH2. In particular, coordinated phosphines should
be compared to phosphonium substituents (PV), hence highly
stabilizing towards the conjugate base carbanion, allowing
considerable precedent. Both benzylic arms have been singly
deprotonated on a PNP pincer complex of Ni(II), where the
buildup of negative charge on the dianionic ligand was determined,
yet the CO2 electrophile attacked benzylic carbon, with no bond
formation between CO2 and nickel[7e]
. The singly benzyl
deprotonated product failed to react with CO2. In contrast, a singly
deprotonated benzylic arm example on Re(I) augments this
analogous CC bond formation with a Re/O bond[7d]. Other
examples of pincer benzylic deprotonation have shown that Li+
associates closely with the system of the dearomatized ring of
the conjugate base[7a-c]
.
N
R
N
2iPrP
N
PiPr2
N
N
N
CO
R'2P
N
iPr
iPr
iPr
iPr
N
iPr
M
iPr
P
P
N
N
Co
P
P
Co
iPr
Co
R
R
iPr
Co
CO
1901 cm-1
OC CO
CO
M'
Sub
OC
CO
1906, 1963 cm-1
Kirchner, 2014
1975 cm-1
1957, 1893 cm-1
Mindiola, 2006
Chirik, 2010
Mindiola, 2006
Figure 12. Pyrazolate as binding motif for bimetallic species.
N
Si
Si
N
N
N
Co
R
Ph2P
PPh2
OC CO
P
R
R
P
tBu
tBu
Co
tBu
tBu
Co
P
Co
Co
tBu
tBu
P
tBu
tBu
P
P
tBu
tBu
tBu
tBu
P
P
R
CO
CO
The collective evidence from this work is that the general synthetic
tool of dehydrohalogenation must confront the question of
whether, when the base employed is an anion, its alkali metal
cation will separate from the resulting conjugate base or bind to it.
This is a question which always exists when deprotonating an acid
HX with M+[Base-]: will the conjugate base X- exist free, or only
ion paired with M+? In the present work, this is tied up with the fate
of chloride on Lewis acid Co: will it remain coordinated, or depart
CO
R = tBu 1863 cm-1
Schneider 2016
R = iPr 1882 cm-1
Arnold 2014
CO
1885 cm-1
Caulton 2007
1882 cm-1
1890 cm-1
Schneider 2016
1894, 1916 cm-1
Waas, 2014
Schneider 2016
Figure 11. Cobalt(I) mono- and dicarbonyl complexes with monoanionic pincer
ligand[12b-g, 17]
.
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
2-methyl-6-(5(3)-phenyl-1H-pyrazol-3(5)-yl)pyridine
with the alkali metal? Clearly this option is eliminated if a neutral
base is employed, and the resulting cationic protonated base is
unreactive: P and sulfur ylides are examples of such bases.
Following the above procedure, using acetophenone instead of pinacolone,
2-methyl-6-(5-phenyl-1H-pyrazol-3-yl)pyridine was obtained as off-white
solid in 75% yield. Spectroscopic data are in agreement with literature
values[19]. 1H NMR (500 MHz, CDCl3 298K): δ(ppm) δ 7.85 (dd, JH,H = 7.7
Hz, 1Hz, 2H), 7.63 (t, JH,H = 7.7 Hz, 1H), 7.50 (d, JH,H = 7.7 Hz, 1H), 7.42
(t, JH,H = 8.3 Hz, 2H), 7.37 – 7.28 (m, 1H), 7.09 (d, JH,H = 7.2 Hz, 1H), 7.01
(s, 1H), 2.58 (s, 3H). 13C NMR (126 MHz, CDCl3 298K): δ(ppm) 152.57,
146.00, 141.81, 138.71, 131.28, 126.93, 122.85, 122.11, 119.82, 116.74,
111.17, 94.27, 18.58.
Experimental Section
General. All manipulations were carried out under an atmosphere of ultra-
high purity nitrogen using standard Schlenk techniques or in a glovebox
under N2. Solvents were purchased from commercial sources, purified
using Innovative Technology SPS-400 PureSolv solvent system or by
distilling from conventional drying agents and degassed by the freeze-
pump-thaw method twice prior to use. Glassware was oven-dried at 150 °C
overnight and flame dried prior to use. THF, including d8-THF was stored
over activated 4 Å molecular sieves or/and sodium metal pieces. NMR
spectra were recorded in various deuterated solvents at 25 °C on a Varian
Inova-400 or 500 spectrometers (1H: 400.11 MHz, 500.11 MHz,
respectively). Proton chemical shifts are reported in ppm versus solvent
protic impurity, but referenced finally to SiMe4. Mass spectrometry
analyses were performed in an Agilent 6130 MSD (Agilent Technologies,
2-(5(3)-(tert-butyl)-1H-pyrazol-3(5)-yl)-6-((di-tert-
butylphosphanyl)methyl)pyridine 1a
In glovebox in 25 ml scintillation vial 2.5 equivalents of nBuLi (2.5 M in
hexane) were added dropwise to a solution of 2-methyl-6-pyrazolylpyridine
(645 mg, 3 mmol) in THF (10 mL) at -78oC. The reaction mixture
immediately changed its color to red-brown. The mixture was slowly
warmed to room temperature during which formation of red precipitate was
observed, stirred overnight and the red precipitate was formed. Then the
solution was again cooled to -78oC and 1.2 equivalent of ClP(tBu)2 (550
mg, 3.1 mmol) was slowly added; after warming to room temperature
stirring was continued for another 12 hrs. The reaction mixture was
transferred to a Schenk flask containing a Teflon stirbar, 15 ml of Et2O was
added followed by degassed water (4 mL). The mixture was stirred for 15
min and then the organic layer was transferred by metal cannula and argon
flow to an empty Schlenk flask. Evaporation of the solvent under reduced
pressure gave almost pure product in ~75-85% yield (and 90% purity) as
a yellowish oil. Purification of 1a by recrystallization from Et2O/pentane
was performed and lead to significant loss of material due to a high
solubility. Purity of crude material is enough for metallation and purification,
if necessary, can be performed at the next stage. There are two pyrazole
H-tautomers. 1H NMR (400 MHz, C6D6 298K): δ(ppm) isomer 1 11.55 (br.s;
1H, NH), 7.25 (d, JH,H = 7.7 Hz, 1H), 7.09 (t, JH,H = 7.7 Hz, 1H), 6.96 (d, JH,H
= 7,7 Hz, 1H), 6.51 (s, 1H), 2.97 (d, JH,P = 9 Hz, 2H), 1.53 (s, 9H), 1.06 (d,
JH,P = 11 Hz, 18H). isomer 2 10.70 (br.s; 1H, NH), 8.24 (m, 1H), 7.36 (m,
1H), 7.26 (m, 1H), 3.18 (bs, 2H), 1.12 (m, 27H). 31P NMR (162 MHz, C6D6
298K): δ(ppm) isomer 1 34.31; isomer 2 35.59. 13C NMR (100 MHz, C6D6
298K): δ(ppm) isomer 1 + isomer 2 163.46, 162.63, 162.48, 153.77,
153.41, 147.81, 142.34, 136.67, 128.29, 128.17, 128.05, 127.93, 127.81,
123.05, 122.95, 116.64, 100.84, 99.68, 32.60, 32.31, 31.81, 31.04, 29.91,
29.77.
Santa Clara, CA) quadrupole mass spectrometer equipped with
a
Multimode (ESI and APCI) source. All starting materials have been
obtained from commercial sources and used as received without further
purification. Cyclic voltammetry was done with Pt as the working electrode,
Pt as the counter electrode, and Ag wire as the reference electrode. 0.1 M
TBAPF6 was employed as a supporting electrolyte, in THF solvent. In this
medium, ferrocene has a peak-to-peak separation of 0.50 V. 1 mmol of
each complex was dissolved in 10 mL of electrolyte solution. All CVs are
referenced to internal Fc/Fc+ as the standard, added at the end of a study
of the experimental sample. Electrodes were polished when a new
molecule was studied.
2-methyl-6-(5(3)-(tert-butyl)-1H-pyrazol-3(5)-yl)pyridine.
Sodium hydride (60 wt % in mineral oil, 0.612 g, 16 mmol) was added to a
250mL Schlenk flask and evacuated then refilled with N2 3 times. This solid
was suspended in 20 mL THF and stirred at RT for 10 min. Into this
suspension was syringed in 1.70 mL (15 mmol) of 3,3-dimethylbutan-2-
one over the course of 5 min. This mixture was allowed to stir at RT for 20
min and then heated to reflux, during which bubbling was observed and
the mixture turned a dull yellow color. To this boiling mixture was added
1.89 g (12.5 mmol) of 2-methyl-6-methoxycarbonylpyridine dissolved in 12
mL THF. This addition was done over the course of 10 min and the mixture
was then allowed to stir for an additional 4 h at reflux. The reaction mixture
was cooled to RT and treated with 1.0 M aqueous HCl until the pH was
between 6-7. This mixture was then extracted with 4 x 15 mL diethyl ether
and washed with brine. The combined organic layer was dried over MgSO4,
filtered then solvent removed under vacuum to yield a yellow powder. This
solid was then dissolved in 25 mL EtOH and brought to a boil. To this
boiling solution was added 2 mL of N2H4•H2O in 15 mL EtOH over the
course of 10 min. This combined solution was refluxed for 2 hours with no
additional color change. Solvent was removed under reduced pressure to
leave a golden oil which was dissolved in 10 mL of THF and the solvent
was removed again and the yellowish solid was formed. The solid was
dissolved in a minimal amount of Et2O and 5 fold excess of hexane was
added. The white precipitate was collected by filtration after 48 h at -20oC.
Colorless crystals (2.28 g, 85%). 1H NMR (400 MHz, CDCl3 298K): δ(ppm)
7.55 (t, JH,H = 7.5 Hz, 1H), 7.49 (d, JH,H = 7.5 Hz, 1H), 7.01 (d, JH,H = 7.5
Hz, 1H), 6.58 (s, 1H), 2.54 (s, 1H), 1.34 (s, 1H). 13C NMR (126 MHz, CDCl3
298K): δ(ppm) 158.17, 152.08, 136.80, 122.06, 116.81, 99.52, 31.68,
30.39, 24.50. ESI-MS(+) in THF: m/z = 216.2 [M+H]+. ESI-MS(-): m/z =
214.2 [M-H]-.
2-(5(3)-(phenyl)-1H-pyrazol-3(5)-yl)-6-((di-tert-
butylphosphanyl)methyl)pyridine 1b.
Following the above procedure, compound 1b was obtained in 75-85%
yield and 90% purity. Recrystallization from THF/pentane was performed
and lead to significant loss of material due to a high solubility. Purity of
crude material is enough for metallation and purification, if necessary, can
be performed at the next stage. Single tautomer: 1H NMR (400 MHz,
CD2Cl2 298K): δ(ppm) δ 7.86 (d, JH,H = 7.9 Hz, 2H), 7.67 (t, JH,H = 7.9 Hz,
1H), 7.50 (d, JH,H = 7.6 Hz, 1H), 7.44 (t, JH,H = 7.6 Hz, 2H), 7.36 (d, JH,H
=
7.6 Hz, 1H), 7.03 (s, 1H), 3.10 (d, JH,P = 3.3 Hz, 2H), 1.17 (d, JH,P = 11.2
Hz, 18H). 31P NMR (162 MHz, CD2Cl2 298K): δ(ppm) 37.25. 13C NMR (126
MHz, CD2Cl2, 298K) δ 162.48 (d, JC,P = 14.0 Hz), 136.92 (s), 128.62 (s),
127.81 (s), 125.46 (s), 123.29 (d, JC,P = 8.5 Hz), 116.43 (s), 99.49 (s),
31.85 (d, JC,P = 23.5 Hz), 31.66 (d, JC,P = 20.5 Hz), 29.46 (d, JC,P = 13.6
Hz).
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
Bu
[PNNtBuH]CoCl2 2a
[PNNt ]Co(CO) 4a and [PNNtBu]Co(CO)*LiCl 4a*LiCl.
A solution of 359 mg (1 mmol) of 1a in 10 mL of THF was added to the
130 mg (1 mmol) of CoCl2. After stirring for 12 h at r.t, the reaction mixture
became homogenous and turns to deep violet. The solution was
concentrated to 2 mL under vacuum and 15 mL of Et2O was added. The
reaction mixture was placed at -45oC overnight and violet solid precipitates
from the solution. The mother liquor was decanted, the residue washed 1-
2 times with 10 mL of Et2O and dried under vacuum. Yield: 460 mg (96%)
of violet microcrystalline powder. Additional recrystallization from
THF/Et2O can be performed for purification if necessary. Crystals suitable
for X-ray diffractometry were grown by slow diffusion of pentane vapors
into a concentrated THF solution at -35oC. 1H NMR (400 MHz, CD2Cl2
298K): δ(ppm) 73.27 (br.s, 1H), 61.25 (s,1H), 59.01 (s, 1H), 57.00 (br.s,
2H), 47.96 (s, 1H), 22.20 (s, 1H), 3.26 (s, 9H), -1.82 (br.s, 18H). ESI-
MS(+): m/z = 453.15 [M-Cl]+; APCI-MS(+): m/z = 453.15 [M-Cl]+; XPS (BE
eV): Co2p3/2 780.31; Cl2p 197.87; N1s 399.45; 400.91; P2p 130.87
A solution of 64 mg (0.1 mmol) of LiPNNCoCl2 in 5 mL of THF was added
to a 25 mL Schlenk flask and freeze-pump-thawed 3 times to remove N2
and then charged with 1 atm of CO. The flask was transferred to a
glovebox and 14 mg of KC8 was quickly added and the stopcock was tightly
closed. Reaction mixture immediately turns brown and, in 30 min, deep
red. The suspension was stirred for 12 h at r.t. and then CO was removed
by vacuum. When vacuum was applied, the red reaction mixture
immediately turns green and green suspension was filtered through Celite
to remove graphite. The volatiles were removed yielding a green oil; 10 mL
of Et2O was added and the mixture was stirred for 12-48 h. Most of the
residue was dissolved except for small amount of white solid and this
solution was filtered through Celite, concentrated and pentane was added
causing precipitation of green amorphous solid. If IR indicates presence of
a second band at 1884 cm-1 repeated extraction with Et2O is necessary.
Yield: 80%. 1H NMR (400 MHz, THF-d8, 298K): δ(ppm) δ 7.52 (t, JH,H = 7.7
Hz, 1H), 7.15 (d, JH,H = 7.7 Hz, 1H), 7.01 (d, JH,H = 7.1 Hz, 1H), 6.24 (s,
1H), 3.63 (d, JH,P = 9.3 Hz, 2H), 1.32 (d, JH,P = 11.3 Hz, 18H), 1.25 (s, 18H).
31P NMR (162 MHz, THF-d8, 298K) δ(ppm) 103.2.). 13C NMR (126 MHz,
tol-d8 298K): δ(ppm) 164.99 (s), 161.23(s), 156.07(s), 151.99(s), 138.79(s),
[PNNPhH]CoCl2 2b
Following the above procedure, compound 2b was obtained in 90% yield
as magenta powder. Solubility of 2b in THF is lower than for 2a; additional
recrystallization could be performed from DCM/Et2O if necessary. 1H NMR
(400 MHz, CD2Cl2 298K): δ(ppm) 68.35 (s, 1H), 63.21 (s,1H), 58.63 (s,
1H), 54.63 (s,2H), 49.89 (s, 1H), 22.55 (s, 1H), 9.12 (s, 2H), 6.77 (s, 2H),
5.49 (s, 1H), -1.42(s, 18H). ESI-MS(+): m/z = 473.12 [M-Cl]+; APCI-MS(+):
m/z = 473.12 [M-Cl]+
116.87(s), 114.19(s), 98.56(s), 34.70 (d, JC,P = 16.2 Hz), 34.50 (d, JC,P
=
19.3 Hz), 32.51, 31.31, 28.78 (d, JC,P = 2.4 Hz). IR (cm-1) 4a 1927 (KBr);
1933 (THF); 1940 (Et2O); 4a*LiCl 1881 (KBr); 1885 (THF or Et2O); XPS
(BE, eV) 4a*LiCl Co2p3/2 779.31 eV (FWHM 1.76 eV); N1s 399.05 eV
(FWHM 1.79 eV), P2p 130.72 eV (FWHM 1.80 eV); O1s 531.60 eV
(FWHM 2.05 eV); Cl 2p3/2 197.89 eV (FWHM 1.51eV) Cl 2p1/2 199.44 eV
(FWHM 1.60eV); Li1s 55.3 eV (weak).
[Li(THF)2PNNtBu]CoCl2 3a
[PNNPh]Co(CO) 4b and [PNNPh]Co(CO)*LiCl 4b*LiCl
A suspension of 100 mg (0.2 mmol) of 2a in 5 mL of THF was cooled to -
78oC and 2 mL of the precooled solution of LiHMDS (1.1 eq, 37 mg, 2.1
mmol) in 2mL of THF was added dropwise. The reaction mixture was
stirred for 3 h at -78oC and then slowly warmed to r.t. during which the
color changed from violet to blue. The solvent was removed under vacuum,
the residue was washed with Et2O until the washings remain colorless,
redissoved in THF, filtered through Celite, concentrated to 1 mL and
addition of 5 mL of Et2O caused precipitation of blue powder that was dried
under vacuum. Yield: 80%. Single crystals suitable for X-ray diffractometry
were grown by slow diffusion of pentane vapors into a concentrated THF
solution at -45oC. 1H NMR (400 MHz, THFd8 298K): δ(ppm) 66.63 (br.s,
2H), 55.83 (s, 1H), 36.41 (s, 1H), 27.75 (s, 1H), 9.50 (s,1H), 8.93 (s, 9H),
-1.84 (br.s, 18H). ESI-MS(+): m/z = 905.2 [((PNN)CoCl)2 + H]+; 911.1
[((PNN)CoCl)2 + Li]+; 947.1 [((PNN)CoCl)2 + LiCl + H]+ APCI-MS(-): m/z =
487.0 [(PNN)CoCl2. XPS Co2p3/2 780.20 eV (FWHM 1.75eV, asymmetric,
shake-up 786.18 eV); N1s 399.19 eV (FWHM 1.67 eV, symmetric), P2p
(FWHM 1.8 eV, asymmetric, spin-orbit components are not resolved), Cl
2p3/2 198.54 eV (FWHM 1.77eV, symmetric) Cl 2p1/2 200.22 eV (FWHM
1.55eV, symmetric); Li1s 55.36eV (FWHM 1.21, symmetric); Potassium
and sodium adducts were obtained using KHMDS and NaHMDS,
respectively, instead of LiHMDS, both showing lower solubility in THF
compared to Li one.
Following the above procedure, compound 3b was obtained in 75% yield
as a green powder. Pure samples of 3b without LiCl coordinated were
obtained by slow crystallization using pentane diffusion into concentrated
THF solution and crystals suitable for X-ray diffractometry as well as
sample for XPS analysis were collected from this batch. 1H NMR (400 MHz,
THF-d8, 298K): δ(ppm) δ 7.76 (d, JH,H = 7.5 Hz, 2H), 7.50 (m, 1H), 7.14 (t,
JH,H = 7.5 Hz, 2H), 7.07 (d, JH,H = 7.5 Hz, 1H), 7.69 (m, 2H), 7.76 (d, JH,H
=
7.5 Hz, 2H), 6.58 (s, 1H), 3.57 (d, JH,P= 8.8 Hz, 2H), 1.33 (d, JH,P = 13.3 Hz,
18H). 31P NMR (162 MHz, THF-d8, 298K) δ(ppm) 105.44. 13C NMR (126
MHz, THF-d8 298K): δ(ppm) 162.50 (d, JC,P= 8.5 Hz ), 155.60 (s), 155.57
(s), 153.98 (s), 153.95 (s), 153.00 (s), 139.59 (s), 136.08 (s), 127.67 (s),
125.15 (s), 124.53 (s), 117.66 (d, JC,P= 9.1 Hz ), 114.21 (s), 98.42 (s),
34.68 (d, JC,P= 18.12 Hz ), 34.61 (d, JC,P= 17.8 Hz ), 28.57 (d, JC,P= 4.0
Hz ). IR (cm-1) 1935 (THF); XPS (BE, eV) Co2p3/2 779.44 eV (FWHM 1.80
eV, asymmetric); N1s 399.16 eV (FWHM 1.67 eV), P2p 130.75 eV (FWHM
1.86 eV), O1s 531.9 eV (FWHM 2.2 eV).
[PNNtBu]Co(CO)2 5a. [PNNtBu]Co(CO) 4a (22 mg, 0.05 mmol) was
dissolved in THF-d8 and added to J. Young NMR tube; after 3 pump-thaw
operations the tube was pressurized with 1 atm of CO and reaction mixture
turned from green to red. The dicarbonyl adduct exists only under CO
atmosphere and turns green immediately after CO removal. 1H NMR (400
MHz, THF-d8, 298K): δ(ppm) 7.57 (t, JH,H = 7.6 Hz, 1H), 7.30 (d, JH,H = 7.3
Hz, 1H), 7.04 (d, JH,H = 7.1 Hz, 1H), 6.41 (s, 1H), 3.73 (d, JH,H = 9.6 Hz,
2H), 1.38 (d, JH,P = 12.7 Hz, 18H), 1.31 (s, 9H). 31P NMR (162 MHz, THF-
d8, 298K) δ(ppm) 114.2. 13C NMR (125 MHz, THF-d8, 298K) δ(ppm):
200.14, 164.08 (bs), 159.35 (s), 152.85 (s), 149.22 (s), 135.38 (s), 114.94
(s), 97.68 (s), 36.00 (d, JC,P = 14.7 Hz), 35.37 (d, JC,P = 20.1 Hz), 32.10,
30.69, 28.52. IR (cm-1) 1944, 1996 (Et2O); 1936, 1994 (THF.)
[Li(THF)2PNNPh]CoCl2 3b
Following the above procedure compound 3b was obtained in 85% yield
as blue powder. 1H NMR (400 MHz, THFd8, 298K): δ(ppm) 64.57 (s, 2H),
58.28 (s, 1H), 42.81 (s, 1H), 35.31 (s, 1H), 15.75 (s, 2H), 13.70 (s, 1H),
13.09 (s, 2H), 7.04 (s, 1H), 2.44 (s, 18H). ESI-MS(+): m/z = 925.2
[((PNN)CoCl)2 + H]+; 931.1 [((PNN)CoCl)2 + Li]+; 967.1 [((PNN)CoCl)2
LiCl + H]+ APCI-MS(-): m/z = 507.0 [(PNN)CoCl2]-.
+
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
[PNNPh]Co(CO)2 5b
K[PNNPh*]Co(CO) 6b
Following the above procedure, yields a red solution in THF. 1H NMR (400
MHz, THF-d8, 298K): δ(ppm) 7.78 (d, JH,H = 8.0 Hz, 2H), 7.51 (t, JH,H = 7.5
Hz, 1H), 7.34 (d, JH,H = 7.5 Hz, 2H), 7.14 (t, JH,H = 8.0 Hz, 2H), 6.98 (m,
2H), 6.90 (s, 1H), 3.67 (d, JH,P= 10.3 Hz, 2H), 1.27 (d, JH,P = 13.6 Hz, 18H).
31P NMR (162 MHz, THF-d8, 298K) δ(ppm) 116.35. 13C NMR (126 MHz,
THF-d8 298K): δ(ppm) 200.65 (bs), 160.21 (d, JC,P= 4.6 Hz ), 154.83 (d,
JC,P= 4.0 Hz ), 152.88 (s), 151.62 (s), 137.19 (s), 136.24 (s), 128.46 (s),
125.83 (s), 125.28 (s), 116.48 (d, JC,P= 4.0 Hz ), 116.39, 99.68 (s), 37.14
(d, JC,P= 17.2 Hz ), 36.31 (d, JC,P= 21.0 Hz ), 29.33 (d, JC,P= 3.0 Hz ). IR
(cm-1) 1937, 1995 (THF.).
Following the above procedure, compound 6b was obtained in 75% yield
as a violet powder. 1H NMR (400 MHz, THFd8, 298K): δ(ppm) 7.58 (d, JH,H
= 7.5 Hz, 2H), 7.22 (t, JH,H = 7.4 Hz, 2H), 7.06 (t, JH,H = 7.4 Hz, 1H), 6.33
(s, 1H), 6.26 (m, 1H), 5.78 (d, JH,H = 8.3 Hz, 1H), 5.41 (d, JH,H = 6.2 Hz,
1H), 2.96 (s, 1H), 1.32 (d, JH,P = 12.9 Hz, 18H). 31P NMR (162 MHz, THFd8,
298K) δ(ppm) 101.48. 13C NMR (126 MHz, THFd8, 298K) δ(ppm): 169.81
48 (d, JC,P= 19.2 Hz ), 157.68 (s), 154.17 (s), 151.57 (s), 137.28 (s), 132.87
(s), 128.79 (s) 126.16 (s), 125.81 (s), 111.68 (d, JC,P= 20.5 Hz ), 97.63 (s),
95.21 (s), 59.90 (d, JC,P= 54.5 Hz ), 36.71 (d, JC,P= 23.4 Hz ), 29.80 (d,
JC,P= 4.0 Hz ). IR (cm-1) 1852 (Et2O).
([PNNH]Co(CO))BF4 4a*HBF4
Acknowledgements
[PNN]Co(CO) 4a (45 mg, 0.1 mmol) was dissolved in 5 mL of Et2O and 1
eq. of Et2O*HBF4 (13.6 μL) in 1 ml of Et2O was added dropwise. Immediate
formation of green precipitate was observed. The liquid phase was
decanted and the residue was washed with 2x1 mL of Et2O and dried
under vacuum. Yield quantitative. 1H NMR (500 MHz, THFd8, 298K):
δ(ppm) 12.99 (s, 1H), 8.07 (t, JH,H = 7.8 Hz, 1H), 7.93 (d, JH,H = 7.8 Hz,
1H), 7.90 (d, JH,H = 7.8 Hz, 1H), 6.99 (s,1H), 4.11 (d, JH,P = 9.9 Hz, 2H),
1.47 (d, JH,P = 17.7 Hz, 18H), 1.43 (s, 9H). 31P NMR (162 MHz, THFd8,
298K) δ(ppm) 103.46. 19F NMR (376 MHz, THF-d8, 298K): δ(ppm) 156.79.
13C NMR (125 MHz, THFd8, 298K): δ(ppm) = 201.75, 164.85, 159.72,
This work was supported by the Indiana University’s Office of the
Vice President for Research and the Office of the Vice Provost for
Research through the Faculty Research Support Program and the
National Science Foundation, Chemical Synthesis Program
(SYN), by grant CHE-1362127. ChemMatCARS Sector 15 is
principally
supported
by
the
National
Science
Foundation/Department of Energy under grant number
NSF/CHE-1346572. Use of the Advanced Photon Source was
supported by the U. S. Department of Energy, Office of Science,
Office of Basic Energy Sciences, under Contract No. DE-AC02-
06CH11357.
154.58, 150.47, 142.00, 124.02, 123.96, 118.49, 101.42, 35.58 (d, JC,P
=
20.2 Hz), 35.31 (d, JC,P = 21.3 Hz), 31.60, 29.10, 28.66, 26.67. IR (cm-1)
1935 (THF). MS(APCI+) m/z = 446.18 ([PNNtBuH]Co(CO))+. MS(ESI+) m/z
= 446.18 ([PNNtBuH]Co(CO))+.
([PNNH]Co(CO)2)BF4 5a*HBF4
Keywords: PNN pincer complex • cobalt • proton-responsive
ligand • single-molecule FLP • metal-ligand cooperativity
[[PNNtBuH]Co(CO))BF4 4a (24 mg, 0.05 mmol) was dissolved in THF-d8
and added to a J. Young NMR tube; after 3 pump-thaw operations, the
tube was filled with 1 atm of CO and the reaction mixture turned from green
to red. The dicarbonyl adduct exists only under CO atmosphere and turns
green immediately after CO removal. 1H NMR (500 MHz, THF-d8, 298K):
δ(ppm) 13.07 (bs, 1H), 8.02 (bs, 2H), 7.83 (bs, 1H), 7.10 (bs,1H), 4.16 (bs,
2H), 1.54 (s, 9H), 1.38, 1.35 (bs, 18H). 31P NMR (162 MHz, THF-d8, 298K)
δ(ppm) 122.55. 19F NMR (376 MHz, THF-d8, 298K): δ(ppm) 156.78. IR
(cm-1) 1944, 2007 (THF.).
[1] J. G. Andino, S. Mazumder, K. Pal and K. G. Caulton, Angewandte Chemie
International Edition 2013, 52, 4726-4732.
[2] S. Gambarotta, F. Arena, C. Floriani and P. F. Zanazzi, Journal of the
American Chemical Society 1982, 104, 5082-5092.
[3] a) S. F. McWilliams and P. L. Holland, Accounts of Chemical Research 2015,
48, 2059-2065; b) G. P. Connor and P. L. Holland, Catalysis Today.
[4] B. J. Cook, C.-H. Chen, M. Pink, R. L. Lord and K. G. Caulton, Inorganica
Chimica Acta 2016, 451, 82-91.
[5] M. T. Whited, Beilstein J Org Chem 2012, 8, 1554-1563.
K[PNNtBu*]Co(CO) 6a
[6] a) G. van Koten, D. Milstein and SpringerLink (Online service) in
Organometallic Pincer Chemistry, Vol. Springer Berlin Heidelberg :
Imprint: Springer,, Berlin, Heidelberg, 2013, p. 1 online resource; b) K. J.
[PNNtBu]Co(CO) 4a*LiCl (calculated based on pure 4a, 45 mg, 0.1 mmol)
was dissolved in 5 mL of Et2O and 1.1 eq. of KMHDS (21 mg) in 1 ml of
Et2O was added dropwise. The reaction mixture immediately turned from
green to deep violet. After stirring for 2 h, the solution was filtered through
Celite, concentrated and pentane was added, causing precipitation of dark
solid. The solution was stored at -45oC for 12 h and then decanted. Solid
material was dried in vacuum. Yield: 40 mg of dark-violet solid. Single
crystals suitable for X-ray diffractometry were grown by slow diffusion of
Szabo, Pincer and pincer-type complexes
synthesis and catalysis, p. xv, 302 pages.
: applications in organic
[7] a) M. Feller, E. Ben-Ari, Y. Diskin-Posner, R. Carmieli, L. Weiner and D.
Milstein, Journal of the American Chemical Society 2015, 137, 4634-
4637; b) M. Feller, U. Gellrich, A. Anaby, Y. Diskin-Posner and D. Milstein,
Journal of the American Chemical Society 2016, 138, 6445-6454; c) E.
Fogler, J. A. Garg, P. Hu, G. Leitus, L. J. W. Shimon and D. Milstein,
Chemistry – A European Journal 2014, 20, 15727-15731; d) M. Vogt, A.
Nerush, Y. Diskin-Posner, Y. Ben-David and D. Milstein, Chemical
Science 2014, 5, 2043-2051; e) M. Vogt, O. Rivada-Wheelaghan, M. A.
Iron, G. Leitus, Y. Diskin-Posner, L. J. W. Shimon, Y. Ben-David and D.
Milstein, Organometallics 2013, 32, 300-308.
pentane vapors into
a precooled saturated solution of 6a in 5:1
heptane/THF mixture at -18oC. Crystals are extremely air-sensitive and
must be handled in inert atmosphere. 1H NMR (400 MHz, THFd8, 298K):
δ(ppm) 6.23 (t, JH,H = 7.6 Hz, 1H), 5.85 (s, 1H), 5.74 (d, JH,H = 7.7 Hz, 1H),
5.27 (d, JH,H = 7.7 Hz, 1H), 2.96 (s, 1H), 1.36 (d, JH,P = 13.3 Hz, 18H), 1.28
(s, 9H). 31P NMR (162 MHz, THFd8, 298K) δ(ppm) 99.2. 13C NMR (126
MHz, THFd8, 298K) δ(ppm): 206.88, 206.17, 169.08, 168.94, 162.23,
155.47, 151.89, 132.34, 110.34, 110.23, 94.35, 93.30, 58.56 (d, JC,P = 51.3
Hz), 35.63 (d, JC,P = 23.4 Hz), 35.56, 31.78, 30.95, 28.99. IR (cm-1) 1851
(Et2O).
[8] a) J. R. Khusnutdinova and D. Milstein, Angewandte Chemie International
Edition 2015, 54, 12236-12273; b) H. Grützmacher, Angewandte Chemie
International Edition 2008, 47, 1814-1818.
[9] A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn and G. C. Verschoor,
Journal of the Chemical Society, Dalton Transactions 1984, 1349-1356.
[10] T. Steiner, Angewandte Chemie International Edition 2002, 41, 48-76.
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
[11] W. C. M. Duivenvoorden, Y.-n. Liu, G. Schatte and H.-B. Kraatz, Inorganica
Chimica Acta 2005, 358, 3183-3189.
[14] B. Butschke, K. L. Fillman, T. Bendikov, L. J. W. Shimon, Y. Diskin-Posner,
G. Leitus, S. I. Gorelsky, M. L. Neidig and D. Milstein, Inorganic
Chemistry 2015, 54, 4909-4926.
[12] a) S. S. Rozenel, R. Padilla, C. Camp and J. Arnold, Chemical
Communications 2014, 50, 2612-2614; b) A. C. Bowman, C. Milsmann,
C. C. H. Atienza, E. Lobkovsky, K. Wieghardt and P. J. Chirik, Journal of
the American Chemical Society 2010, 132, 1676-1684; c) A. R. Fout, F.
Basuli, H. Fan, J. Tomaszewski, J. C. Huffman, M.-H. Baik and D. J.
Mindiola, Angewandte Chemie International Edition 2006, 45, 3291-
3295; d) P. O. Lagaditis, B. Schluschaß, S. Demeshko, C. Würtele and
S. Schneider, Inorganic Chemistry 2016, 55, 4529-4536; e) M. J.
Ingleson, M. Pink, H. Fan and K. G. Caulton, Inorganic Chemistry 2007,
46, 10321-10334; f) S. Murugesan, B. Stöger, M. D. Carvalho, L. P.
Ferreira, E. Pittenauer, G. Allmaier, L. F. Veiros and K. Kirchner,
Organometallics 2014, 33, 6132-6140; g) M. A. Kent, C. H. Woodall, M.
F. Haddow, C. L. McMullin, P. G. Pringle and D. F. Wass,
Organometallics 2014, 33, 5686-5692.
[15] a) J. K. Kouba, J. L. Pierce and R. A. Walton, Journal of Organometallic
Chemistry 1980, 202, C105-C107; b) C. A. Strydom and H. J. Strydom,
Inorganica Chimica Acta 1989, 159, 191-195.
[16] L. Vaska and J. W. DiLuzio, Journal of the American Chemical Society 1962,
84, 679-680.
[17] M. L. Scheuermann, S. P. Semproni, I. Pappas and P. J. Chirik, Inorganic
Chemistry 2014, 53, 9463-9465.
[18] a) E. D. Jemmis, J. Chandrasekhar, E. U. Wuerthwein, P. v. R. Schleyer, J.
W. Chinn, F. J. Landro, R. J. Lagow, B. Luke and J. A. Pople, Journal of
the American Chemical Society 1982, 104, 4275-4276; b) F. J. Landro,
J. A. Gurak, J. Chinn, R. Newman and R. Lagow, Journal of the American
Chemical Society 1982, 104, 7345-7346.
[19] A. Chadghan, J. Pons, A. Caubet, J. Casabó, J. Ros, A. Alvarez-Larena
and J. Francesc Piniella, Polyhedron 2000, 19, 855-862.
[13] S. Hüfner, Photoelectron spectroscopy
: principles and applications,
Springer, Berlin ; New York, 2003, p. xv, 662 p.
This article is protected by copyright. All rights reserved.

10.1002/chem.201700859
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
A pyridyl pincer ligand is introduced to
Author(s), Corresponding Author(s)*
incorporate
functionalities via
five
different
PtBu2 and
Page No. – Page No.
a
pyrazole pincer ligand arms. Ligand
properties in neutral mono- and
dianionic form were investigated on
Co(I) and (II) complexes. Concept of
single-molecule frustrated Lewis pair
(FLP) was discussed and alkaline
metal chlorides was used as a model
substrate to show FLP behavior.
Title
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))
Layout 2:
FULL PAPER
Author(s), Corresponding Author(s)*
((Insert TOC Graphic here; max. width: 11.5 cm; max. height: 2.5 cm))
Page No. – Page No.
Title
Text for Table of Contents
This article is protected by copyright. All rights reserved.