The Electronic Properties of Ni(PNN) Pincer Complexes Modulate Activity in Catalytic Hydrodehalogenation Reactions

Article abstract of DOI:10.1002/ejic.202000721

Three chloronickel(II) complexes of PNN- pincer ligands with pyrazolyl and diphenylphosphino donors appended to different arms of diarylamido anchors were prepared and fully characterized. The three derivatives (1-OMe, 1-Me, 1-CF₃) differ only by the identity of the para-aryl substituent on the pyrazolyl arm with 1-OMe being 310 mV easier to oxidize than 1-CF₃. All three complexes are competent catalysts for hydrodehalogenation reactions of 1-bromooctane and a variety of aryl halides in dimethylacetamide using NaBH₄ as both base and hydride source. Comparative studies using diverse substrates showed that catalytic activity correlates with electron donor properties; 1-OMe was superior to the other two. Deuterium labeling studies verified NaBD₄ as the deuteride source and excluded solvent-assisted radical pathways.

Full text of DOI:10.1002/ejic.202000721

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
Catalytic Hydrodehalogenation | Very Important Paper |
The Electronic Properties of Ni(PNN) Pincer Complexes
Modulate Activity in Catalytic Hydrodehalogenation Reactions
[
a]
[a]
Denan Wang and James R. Gardinier*
Abstract: Three chloronickel(II) complexes of PNN- pincer li- halogenation reactions of 1-bromooctane and a variety of aryl
gands with pyrazolyl and diphenylphosphino donors appended halides in dimethylacetamide using NaBH as both base and
4
to different arms of diarylamido anchors were prepared and hydride source. Comparative studies using diverse substrates
fully characterized. The three derivatives (1-OMe, 1-Me, 1-CF3) showed that catalytic activity correlates with electron donor
differ only by the identity of the para-aryl substituent on the properties; 1-OMe was superior to the other two. Deuterium
pyrazolyl arm with 1-OMe being 310 mV easier to oxidize than labeling studies verified NaBD as the deuteride source and ex-
4
1
-CF . All three complexes are competent catalysts for hydrode- cluded solvent-assisted radical pathways.
3
Introduction
low valent nickel species, reactive nickel hydride intermediates
while also supporting catalytic activity.
Catalytic hydrodehalogenation (HDH) of organic halides is of
great interest in organic synthesis and environmental pollution
[
1–5]
remediation efforts.
that have been developed,
recently emerged as attractive, stable, homogeneous catalysts
Among the numerous HDH methods
[
6–15]
nickel pincer complexes have
[
16–21]
for such reactions.
In 2012, the Hu group reported that
[
16,22]
“
Nickamine”
(complex A in Figure 1) could be used as a
precatalyst for hydrodehalogenation reactions of both aryl and
alkyl halides by employing diethyloxymethylsilane as reductant
and sodium methoxide as base. Rettenmeier, Wadepohl and
Gade found that a chiral Ni(NNN) complex, B, (Figure 1) can be
transformed to a stable Ni(I) derivative via reaction with LiBEt3H
and could be used to effect stereoselective HDH of various alkyl
[
18]
halides via a radical mechanism. Later this group successfully
[
19]
used these catalysts for HDH of aromatic halides.
demonstrated that this pincer complex could even effect
Hydrodefluorination reactions.
Norton
[
21]
Pincer ligands with other do-
Figure 1. Representative nickel pincer precatalysts used in dehydrohalogena-
tion reactions of organic halides.
nor atoms were also found to be useful for HDH reactions. Thus,
[
17]
II
Enthaler and co-workers
used a (ONO)Ni PPh complex, C,
3
(
Figure 1) in concert with either Grignard or organozinc rea-
A fundamental question prompted by the above results was
whether the replacement of one (or both) of the “hard” flanking
donors with groups such as carbenes or organophospine do-
nors would still give catalytically active species or whether the
hydride or low valent nickel intermediates would be too stable
to allow for catalytic turnover. A recent contribution from the
gents to effect HDH of aromatic and alkyl halides. The reaction
was thought to proceed via the hydride from ꢀ-hydride elimina-
tion of the in-situ generated organonickel species; the expected
alkyl-aryl cross-coupling products were the minor by-products
of The HDH reaction. It is interesting that ligands with relatively
“
hard” nitrogen or oxygen donor sets allow for isolation of both
[20]
Sun group addressed this question. Their carbene-containing
R
pincer complexes (NNC)NiBr, D (Figure 1), can be converted to
(NNC)NiH via reactions with sodium tert-butoxide as a base and
[
a] Dr. D. Wang, and Prof. J. R. Gardinier
triethyloxysilane as a hydride source. These (NNC)NiH can effi-
ciently catalyze HDH reactions of various aryl and alkyl halides
where the DiPr variant proved superior to the other two D
Department of Chemistry, Marquette University,
Milwaukee, WI, USA 53201-1881
E-mail: james.gardinier@marquette.edu
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejic.202000721.
This manuscript is part of the Special Collection Pincer Chemistry and Catal-
ysis.
R
complexes.
We sought to extend these studies to pincer complexes with
organophosphine donors. Despite numerous (PNN)Ni pincer
Eur. J. Inorg. Chem. 2020, 4425–4434
4425
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
complexes being known as catalysts for a variety of reac- prove yields of this step by using drastically different reaction
[
23–29]
tions,
it was fairly surprising that there were no reports of times, alternative solvents, different catalysts, or adding co-cata-
0
their use as catalysts for HDH reactions. Several years ago, our lysts have not yet been proven successful. The ensuing Pd -
group introduced a PNN pincer type ligand, (N-2-diphenyl- catalyzed coupling reaction with diphenylphosphine afforded
phosphinophenyl)(N-2-pyrazolyl-p-tolyl)amine, H(PNN–Me), and high yields of the desired H(PNN-X) ligands as air stable color-
showed that the hemilabile pyrazolyl arm conferred enhanced less solids. The desired nickel complexes 1-X (X = OMe, Me, CF3)
reactivity and structural adaptability to stabilize various rho- were obtained by first mixing a CH Cl solution of the ligand
2
2
[
30]
dium(I/III) complexes. The modular nature of the ligand syn- with a methanol solution of NiCl ·6H O to generate a complex
2
2
thesis suggested a simple means to alter the electron donor in-situ prior to addition of a commercial MeOH solution of
strength of the ligand by varying para- aryl substituents. Such NEt (OH) to deprotonate the ligand which is indicated by the
4
variants would be attractive for nickel-catalyzed HDH reactions, appearance of a characteristic deep green color (vide infra) and
in that the stability of any hydride intermediates could be tuned partial precipitation. Compounds 1-X exhibit limited solubility
by substitution, thereby influencing the rates of reaction. in MeOH which allows a facile means of separation from the
Herein, we report the preparation and properties of three new soluble NEt Cl by-product. The compounds are soluble in THF,
4
nickel(II) pincer complexes, (PNN-X)NiCl, 1-X (X = MeO, Me, and CH CN, aromatic and halocarbon solvents but are only very
3
CF ) (Figure 1, bottom right) and disclose our initial findings slightly soluble pentane, hexane, Et O and lower alcohols.
3
2
regarding their ability to catalyze HDH reactions of aromatic
and alkyl halides.
Solid State. Crystals suitable for single-crystal X-ray diffraction
were obtained by vapor diffusion of either Et O (1-Me) or pent-
2
ane (1-OMe and 1-CF ) into benzene solutions of the com-
3
plexes. Views of a representative structure of 1-Me are given in
Figure 2, while other structures are provided in Figures S1 and
Results and Discussion
An overview of the syntheses of the ligands and nickel com- S2. Selected bond lengths and angles are given in Table 1. Com-
plexes is given in Scheme 1 with synthetic details provided in pound 1-Me was found to crystallize as both triclinic needles
the Supporting Information. The CuI-catalyzed amination reac- (P 1¯ ) and monoclinic plates (P2
/n). The triclinic polymorph has
1
tion between 1,2-diiodobenzene and the known 4-X-2-pyrazol- one molecule while the monoclinic form contains two mol-
X
[31]
1
-ylaniline, H(pzAn ), (X = OMe, Me, or CF ) compounds gives ecules in the asymmetric unit. The former is slightly more dense
3
3
3
a mixture of mainly the desired compound along with variable (1.48 g/cm ) than the latter (1.42 g/cm ) which may give rise to
X
amounts of unreacted H(pzAn ) and a di-aminated product that the more distorted square planar NiN
PCl coordination geome-
2
require separation by column chromatography. Attempts to im- try (τ = 0.16 dominated by N12–Ni1–P1 163.5°) vs. those in the
δ
Scheme 1. Synthesis of pincer ligands and their chloronickel(II) complexes.
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4426
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
monoclinic form (τδ = 0.11 (Ni1), 0.05(Ni2)). Despite the angular
differences, the bond lengths about nickel in each form are
quite similar. The greatest deviation occurs in the Ni1–P1 dis-
tance where the more distorted complex has a slightly longer
bond (2.142 Å) than the average (2.130 Å) found in the mono-
clinic case. The next biggest difference occurs in the Ni1-Cl1
distance: the triclinic form measures 2.181 Å, slightly smaller
than the average in the monoclinic form 2.185 Å. The Ni1-Cl1
distances are at the longer end while the Ni1–P1 distances are
at the shorter end of the ranges found in other
Table 1. Selected bond lengths [Å] and angles (°) for (PNN-OMe)NiCl (1-OMe),
two forms of (PNN–Me)NiCl (1-Me), and for (PNN-CF )NiCl (1-CF3).
3
1-OMe
1-Me (P1¯) 1-Me
(P2
/n)[a]
1-CF
3
Distances[Å]
1
Ni1-Cl1
Ni1a–Cl1a
Ni1–P1
Ni1a–P1a
Ni1–N1
2.1825(4)
2.1285(4)
2.1808(5) 2.1854(8), 2.1718(10)
2.1855(8)
2.1421(5) 2.1271(8), 2.1226(9)
2.1343(8)
1.8992(13) 1.8895(15) 1.894(2), 1.912(3)
Ni1a–N1a
Ni1–N12
Ni1a–N12a
1.887(2)
1.9412(13) 1.9306(16) 1.943(2), 1.941(3)
1.923(2)
[
26,28,29]
[32,33,41]
(
PNN)NiCl
or (PNP)NiCl
complexes. In 1-Me, the
five-membered NC PNi chelate ring adopts an envelope confor-
2
Angles(°)
mation (fold angles = 16.3° for P 1¯ , 11.0° and 20.1° for rings
involving Ni1 and N1a, respectively, in the monoclinic form)
and the six-membered NC N Ni chelate ring adopts a half-boat
Cl1–Ni1–N1 Cl1a–Ni1a–N1a 176.64(4)
173.37(5) 175.36(8) 175.97(8)
77.43(8)
163.50(5) 169.38(7), 172.44(8)
176.02(8)
1
N12–Ni1–P1
172.52(4)
93.62(4)
2
2
N12a–Ni1a–P1a
Cl1–Ni1–N12
Cl1a–Ni1a–N12a
Cl1–Ni1–P1
conformation to give an overall structure that distinguishes the
92.04(5)
92.50(7), 93.80(8)
91.83(7)
phenyl rings of the PPh group (Figure 2, right), with a pseudo-
2
axial ring (type A) being closer to the toluidinyl ring than the
pseudoequatorial ring (type B). In most respects, the structure
of 1-OMe is similar to those of 1-Me with nearly identical
square planar (τδ = 0.08) coordination geometry and bond
93.103(17) 92.74(2)
89.73(3), 91.78(4)
91.18(3)
Cl1a–Ni1a–P1a
N1–Ni1–N12
N1a–Ni1a–N12a
N1–Ni1–P1
89.58(6)
83.76(4)
90.14(7)
86.89(5)
91.96(10), 89.86(11)
90.71(10)
86.11(7), 84.72(8)
86.30(8)
lengths about the PN NiCl kernel. The envelope fold angle in 1-
2
N1a–Ni1a–P1a
OMe is more severe (28.0°), which is likely influenced by crystal
packing as indicated by the polymorphs of 1-Me. The short C4–
O1 bond of the anisidine unit (1.375(2) Å) in 1-OMe does not
impart any significant changes in C-C or C-N bond lengths in
the aromatic ring from those displayed by 1-Me. On the other
[
a] Two molecules in asymmetric unit.
single resonance near δ = 30 ppm that is significantly shifted
downfield from the resonance for the appropriate free H(PNN-
X) ligand (δ = –19.5, –18.3, and –16.5 ppm for X = OMe, Me,
and CF , respectively) due to binding the nickel center. A com-
bination of 2D NMR techniques (COSY, HMQC, HMBC) was used
to unravel the rather complex H and C NMR spectra (see
Supporting information). The complexity arises both due to de-
P
hand, the structure of 1-CF displays significantly shorter Ni1–
P
3
Cl1 (2.17 vs. 2.18 Å), Ni1–P1 (2.12 vs. 2.13 Å) and N1–C1 (1.36
vs. 1.40 Å) bonds than the other two derivatives. Additionally,
the Ni1–N1 distance of 1.91 Å is longer than 1.90 Å for 1-OMe,
and 1.89 V (avg.) for 1-Me. These trends may be explained if
the inductive effect of the CF group weakens the Ni–N
3
1
13
1
4
1
13
31
tectible J to J coupling of H and C nuclei with the
P
3
amido
nucleus and because of the two sets of multiplet resonances
for distinguishable phenylphosphino groups (types A and B, as
in the right of Figure 2) that overlap and sometimes mask other
resonances. A few points are worthy of mention regarding the
bond, rendering the metal more electrophilic which is then
compensated for by shortening bonds with more polarizable P
and Cl atoms. The Ni1–N12 (Ni–N ) distance of 1.94 Å is statisti-
pz
cally identical to the other compounds in the series, which is in
line the observation that metal-nitrogen (pyrazolyl) bonds are
generally more sensitive to oxidation and spin state changes
1
NMR analysis. First, in the H NMR spectra, the resonances for
pyrazolyl hydrogen atom are readily identified at δ = 8.1 (H ),
H
5
[
34–37]
8.0 (H ) and 6.6 (H ) ppm. The assignment of the latter is
than to inductive effects.
3 4
straightforward due its characteristic chemical shift, but the
former two are tentative (vide infra), being based on both dif-
1
3
1
13
ferent C resonance multiplicities and H- C HMQC cross
3
peaks with corresponding resonances near δC = 128.6 ( J
C-P
doublet, C ) and 143.0 (singlet, C ) ppm. Of the aromatic reso-
3
5
nances, those of the PPh group are the least electron rich and
2
appear most downfield. Of these, the resonances for ortho-hy-
3
3
drogens appear as two doublet-of-doublets ( J ca. 12 Hz, J
H-P
HH
ca. 8 Hz) near 7.97 and 7.87 ppm; the assignment to type B and
A rings, respectively, was arbitrary and could be reversed. Their
relationship to the corresponding para- and meta-hydrogens (as
Figure 2. Left: structure of (PNN–Me)NiCl (1-Me, P1¯ form) with partial atom
labelling. Hydrogens and some phenyl carbon ellipsoids were removed for
1
well as to particular ipso-carbons) was established by both H-
clarity; right: view parallel with the C
formation of the C NPNi chelate ring and the distinguishable phosphino-
phenyl groups (A and B).
6
H
4
moiety showing the envelope con-
1
H COSY, HMQC, and HMBC experiments. The resonances for
2
the three ring hydrogens of the pyrazolyl-aniline group are
readily identified by their characteristic multiplicities in the δ =
H
1
3
Solution. The solution structures of the three 1-X compounds 6.7 to 7.4 ppm region. For 1-OMe and 1-Me, the C NMR reso-
elucidated by NMR methods match expectations based on the nances of this group are distinguished because they are too
solid-state structures. First, each 31P NMR spectrum contains a remote to display C- P coupling. For 1-CF , the CF and C -,
13
31
3
3
3
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4427
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
C -, and C nuclei of the aromatic aniline ring give characteristic
4
5
quartet resonances due to 13C- F coupling. Finally, the 2-phos-
19
phinoaniline group is the most electron rich aromatic ring and
1
the H corresponding resonances are generally furthest upfield,
below about δ = 7.05 ppm, with the H - (ortho- to N) multiplet
H
3
resonance near δ = 6.58 ppm being the most distinguishable.
H
1
3
Similarly, in the C NMR spectrum the associated C - doublet
3
resonance is uniquely upfield (δ = 119 ppm) while that for the
C
C2-ring (4° center, ipso-N) is the most downfield aromatic dou-
blet resonance at ca. δ = 163 ppm. Interestingly, the C ring
C
6
nuclei (ortho- to P) that was expected to appear as a doublet
2
with J coupling ca. 25 Hz (similar to the C - resonance) ap-
C-P
2
pears as a singlet. Its identity was verified by both HMQC and
the strong three-bond HMBC cross peak with the H - resonance.
4
Thus, assignments based on the magnitude of JC-P coupling,
such as those for the H(C)3/5 nuclei above, are tenuous and
could be reversed.
The electronic properties of the three 1-X compounds were
studied by electronic absorption spectroscopy, electrochemistry
Figure 3. Selected frontier orbitals for 1-OMe calculated at the M06/def2-
SV(P)/PCM (CH Cl ) level of theory with orbital number, energies, and main
contributions shown.
2
2
(square-wave and cyclic voltammetry) and by computational
methods (DFT and TD-DFT). A summary of calculated and ex-
perimental properties is given in Table 2. It is useful to first
describe frontier orbitals of the complexes to facilitate discus-
sion of electronic properties. The frontier orbitals of 1-Me and
parameter[38] (Figure S18). That is, 1-OMe is easier to oxidize
(
(
E′(1-OMe)+/1-OMe = 0.51 V vs. Ag/AgCl, σpara = –0.27) than 1-CF3
E′(1-CF3)+/(1-CF3)= 0.82 V vs. Ag/AgCl σpara = +0.54). It is also
noted that the calculated oxidation potentials give excellent
agreement with the experimental values, providing another
measure (along with excellent agreement of bond lengths) to
validate our choice of using this DFT model. Finally, as can be
elucidated by inspection of the HOMO in 1-OMe or of the Mul-
1
-CF are nearly identical to 1-OMe (except for relative ener-
3
gies), so only those of the latter will be discussed. The frontier
orbitals of 1-OMe calculated at the M06/def2-SV(P) level are
provided in Figure 3. Full computational details including more
complete diagrams for the series of 1-X are provided in the
supporting information (Figures S23–S25). As indicated in Fig-
ure 3, the HOMO of 1-OMe is mainly a pincer ligand π-orbital
mixed in a π* manner with a minor contribution from nickel's
+
liken spin-densities of atoms in (1-OMe) (Figure 4, right), the
oxidation event of 1-OMe is mainly (88 %) ligand-based such
+
II ·+
III
+
that (1-OMe) is best described as Ni -L , rather than (Ni -L) .
3
d orbital. The 3d orbital is mainly involved in dπ–pπ interac-
yz yz
tions with the chloride p orbitals where the π* combination
z
found in HOMO(–2) and the corresponding π-bonding compo-
nent comprises HOMO (–11 and –13), Figure S23. The
HOMO(–1) is mainly nickel's 3dz orbital involved in σ*-interac-
2
tions with P and anilino-N orbitals with minor contributions
from the ligand π system and chloride p orbital. The LUMO
y
has the Ni 3d
x2–y2
involved in a classical σ* interaction with li-
gand group orbitals, while the LUMO (+N) (N = 1–7) are exclu-
sively ligand based π* orbitals.
Table 2. Summary of experimental and calculated (M06/def2-SV(P)/PCM
(
CH
Compound
-OMe exp.
2
Cl
2
)) properties for 1-X (X = OMe, Me, CF
3
).
E'1+/1, V[a]
UV/Vis (CH
Cl
2
) λmax, nm, (ε, M–1 cm–1
)
2
1
+0.51[b]
+0.52
660 (930), 480 sh (1100), 429 (3200), 334 (8340)
820 (175), 604 (540), 391, (7940), 339 (9710)
656 (910), 478 sh (1100), 419 (3100), 334 (9040)
832 (154), 604 (480), 391, (7900), 339 (11200)
630 (780), 470 (830), 395 (3100), 355 (7700)
841 sh(140), 606 (425), 373 (8032), 328 (14900)
Figure 4. Left: cyclic voltammograms obtained for 1-X (X = OMe, Me, CF
3
) at
(
1
(
1
Calcd)
-Me exp.
Calcd)
-CF exp.
Calcd)
[
b]
a scan rate of 200 mV/s in CH Cl with NBu PF as a supporting electrolyte;
2
2
4
6
+0.59
+0.61
+0.82
+0.91
+
right: Mulliken spin densities and isosurface (isovalue 0.05) of (1-OMe) .
[
b]
3
An overlay of the UV/Visible spectra of the three 1-X com-
pounds is shown in Figure 5 while data are summarized in
Table 2. Details regarding calculated (time-dependent density
functional theory, TD-DFT) spectra along with calculated
(
[
2 2 4 6
a] V vs. AgAgCl. [b] In CH Cl , NBu PF as supporting electrolyte, average of
potentials acquired at scan rates of 100, 200, 300, and 400 mV/s.
The three 1-X complexes exhibit reversible oxidations spectra are provided in the Supporting information (Figures
Figure 4) whose E′(1-X)+/(1-X) reduction potentials scale linearly S26–S27). The lowest energy and least intense (ε ca. 800 to
(
–1
–1
with the electron donating nature of the para-anilino substitu- 1000
M
cm ) band in the 600 to 750 nm range gives rise to
ent as quantified using the substituent's Hammett σpara the green or green blue color of the complexes. The modest
4428 © 2020 Wiley-VCH GmbH
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
intensity and the hypo- and hypso-chromic shifts on ligand significant losses due to inadvertent evaporation. A reaction
substitution of OMe with more electron withdrawing substitu- time of 3 h was arbitrarily chosen for the initial survey as it was
ents are suggestive of ligand-to-metal charge transfer (LMCT) reasonably short to allow rapid screening. Fortuitously, this time
[
34]
character.
This assertion is also obtained via analysis of period provided useful window into reactivity differences, so no
TD-DFT results (Tables S5 – S7), albeit not straightforwardly. further effort was made to optimize the reaction time. A sum-
The lowest energy experimental band is resolved by TD-DFT mary of results from HDH reaction optimization studies is found
–
1
–1
into disparate bands: a low-intensity (ε ca. 100
M
cm ), low- in Table 3, with the optimized conditions found in entries 1–4.
–1
energy (ca. 12,000 cm , λ
ca. 830 nm) component A and a That is, all three complexes were excellent at catalyzing the
cm ), higher energy (ca. HDH reaction of 4-bromobiphenyl at 8 mol-% loading, with 1-
ca. 605 nm) component B. These bands in- OMe being slightly superior to the other two complexes. The
max
–1
–1
more intense (ε ca. 400–500
M
–
1
16,500 cm , λ
max
volve admixtures of HOMO(+N) → LUMO transitions where yields of product decrease with lower catalyst loading (4 mol-
N = 0–2 (component A) and N = 3, 5 (component B). Thus all %, entries 5–7) and the superiority of 1-OMe vs. 1-Me becomes
have d–d character but component A HOMO(+N) has significant evident. The 1-X complexes all outperform simple nickel salts
contributions from pincer ligand π system and chloride p (and and the reaction does not proceed in the absence of nickel
z
to a lesser extend p ) orbitals whereas the HOMO(+3 and +5) (entries 8–12). A control reaction with 1-OMe in the presence
x
of component B have substantial contributions from chloride of a drop of elemental mercury (entry 2), showed no significant
px-orbitals (Figure S23) with little or no contribution from pincer reduction in efficiency suggesting that nickel colloids are not
ligand orbitals. As such, the energy of the component A band responsible for the catalytic activity. A second control (not tabu-
is more ligand dependent than that of component B. In general, lated) showed that no reaction occurred in the presence of the
transitions involving the ligand-based HOMO as the origin of a stable radical 2,2,6,6-tetramethylpiperidinyl-N-oxyl (TEMPO),
transition will experience ligand-dependent hypsochromic suggesting either a radical pathway or an irreversible reaction
shifts along the series 1-OMe to 1-CF , which is exhibited by with a non-radical intermediate. In initial temperature screen-
3
the next two higher energy bands in the 400–500 nm range of ings, 80 °C was found to be the optimal reaction temperature
the experimental spectra. These latter two bands are due to to give a maximum yield of biphenyl during the 3 h reaction
overlapping HOMO → LUMO (+0 and +1) transitions for the time; there was no improvement in reactions performed at
lower energy shoulder (λexper ca. 470–480 nm, λTDDFT ca. 100 °C. Finally, screening different solvents (compare entries 15
391 nm) and to overlapping HOMO → LUMO (+2, +3, and +4)
transitions for the higher energy band (λexper ca. 395–430 nm,
Table 3. Optimization of conditions for hydrodehalogenation of 4-bromobi-
λTDDFT ca. 333 ± 6 nm). The more intense band (ε ca.
[a]
phenyl using NaBH as a reductant.
4
–
1
–1
9000
M
cm ) or split bands (1-CF ) near λ
ca. 350 nm
exper
3
have MLCT character, being lowest energy for 1-CF and high-
3
est for 1-OMe. Accordingly, TD-DFT predicts these bands
(
λTDDFT ca. 280 ± 4 nm) to be due to mainly overlapping transi-
tions between metal-based HOMO(–1, –2, or –5) to LUMO(+1),
a pincer π* orbital.
Entry
Pre-catalyst
mol-%
solvent
[°]C
yield[b]
1
2
3
4
5
6
7
8
9
1-OMe
1-OMe[
1-Me
8
8
8
8
4
4
4
4
4
4
4
4
4
4
4
4
4
4
8
4
4
4
4
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMF
80
80
80
80
80
80
80
80
80
80
80
80
100
100
60
40
20
60
80
60
60
60
60
98
96
97
89
76
70
54
17
14
7
c]
1-CF
3
1-OMe
1-Me
1-CF
Ni(dppe)Cl
Ni(DME)Cl
NiCl ·6H
3
2
2
10
11
12
13
14
15
16
17
18
2
2
O
Ni(OAc)
none
2
6
0
1-OMe
1-Me
1-Me
1-Me
1-Me
1-Me
1-OMe
1-Me
1-Me
1-Me
1-Me
78
67
55
25
7
32
76
33
28
25
0
Figure 5. Overlay of the UV/visible spectra of 1-X (X = OMe, thick solid red
line; X = Me, dashed blue line; X = CF , thin solid green line) in CH Cl .
3 2 2
Catalysis. With the series of electronically diverse nickel pin-
cer complexes in hand, their utility in hydrodehalogenation re-
19
DMF
THF
dioxane
iPrOH
toluene
2
0
actions was investigated. NaBH was selected as the hydride
4
[
39,40]
21
source because it is conveniently handled
and is, by far,
2
2
2
3
the least expensive of the commonly used reductants. The com-
pound 4- bromobiphenyl was chosen as an initial substrate for
reaction optimization because it and its product, biphenyl, are
readily available, easily-handled solids that can be reliably quan-
tified both on the spectroscopic and synthetic scale without
[
a] Typical conditions: 0.2 mmol of 4-bromobiphenyl (46.6 mg), 0.016 mmol
of Ni catalyst (8 mol-%), NaBH (15.0 mg, 0.4 mmol, 2.0 equiv.) and 2 mL
dimethylacetamide (DMA) or other solvent, heat 3 h. [b] GC/MS yield, average
of three experiments, error ± 3 %. [c] In the presence of Hg (l).
4
0
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4429
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
and 18–23) showed that DMA was superior to other solvents;
Table 4. Summary of catalytic hydrodehalogenation reactions of aryl and alkyl
halides.[
a]
the insolubility of NaBH being the critical factor in the lack of
4
activity in toluene.
Next, the scope of the HDH reactions with a variety of aro-
matic and alkyl halides was explored, with results provided in
Table 4. The conversion of halobenzenes to benzene catalyzed
by 1-OMe (under conditions optimized for bromobiphenyl
above) decreased up the periodic table being quantitatve for
iodobenzene and zero for fluorobenzene (entries 1–4). This
trend roughly parallels the expected stability of the correspond-
ing halo radical, but biphenyl was never observed as a by-prod-
uct. Substrate electronic effects were probed in HDH reactions
of para-R-substituted bromobenzenes catalyzed by 1-OMe (R =
Ph: Table 3, entry 1; R = H, Me, OMe, CN: Table 2, entries 2, 5,
Entry
R
Y
X
pre-cata-
lyst
Special Condi- [%]
tions
yield[b]
1
2
3
4
5
6
7
8
9
H
H
H
H
Me
MeO
MeO
MeO
MeO
CN
CN
CN
H
H
H
H
H
H
H
H
H
H
H
H
H
F
I
1-OMe
1-OMe
1-OMe
1-OMe
1-OMe
1-OMe
1-Me
99
85
8
Br
Cl
F
0
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Cl
H
61
55
45
29
53
97
99
93
94
99
27
19
6, and 9). The yields of the dehalogenated product are sup-
1-CF
A
3
pressed for the electron releasing substituents, R = Me and MeO
whereas they are quantitatve for electron withdrawing CN. If
one considers Hammett σpara to be indicative of electron releas-
ing capabilities of the R group, the yield for the conversion of
bromobenzene (R = H) is lower than expected; a phenyl is more
electron releasing than H (σpara (Ph) = –0.01 vs. σpara (H) = 0.00)
yet the yield for R = Ph was higher than the R = H case (98 %
vs. 85 % yield, respectively). Other factors such as resonance
stabilization of radical species or simply redox potentials may
also be (minor) contributing factors in determining conversion.
Next, the electronic effects of the 1-X catalysts were evaluated
for 4-bromoanisole (Table 4, entries 6–8) and 4-bromobenzonit-
rile (Table 4, entries 10–12) to validate the generality of results
observed for 4-bromobiphenyl (Table 3, entries 1 and 3–7) on
electronically diverse substrates. All three catalysts were excel-
lent at catalyzing the HDH reaction of the electron-poor bromo-
10
11
12
13
1-OMe
1-Me
1-CF
3
1-OMe
1-OMe
1-OMe
1-OMe
14
15
CF3
H
Cl
F
Cl
Cl
1 h, 4 mol-%
16
Other substrates:
substrate
pre-cata-
lyst
Special Condi- [%]
tions
yield[b]
17
2-bromonaphthalene
2-bromofluorene
9-bromoanthracene
2-bromopyridine
benzyl bromide
1-OMe
1-OMe
1-OMe
1-OMe
1-OMe
1-OMe
98
96
98
99
99
98
18
19
20
21
22
1 h, 4 mol-%
1 h, 4 mol-%
1-bromooctane
[
(
a] Reaction scale: 0.2 mmol of organic halide, 0.016 mmol of Ni catalyst
8 mol-%) (or 0.008 mmol, 4 mol-%, if specified), NaBH (15.0 mg, 0.4 mmol,
benzonitrile, with 1-CF being marginally less efficient than the
3
4
other two (93 % vs. quantitative conversion). With the more
electron rich anisole substrate, the yield increased inversely
with (pre)catalyst oxidation potential: 1-CF3 (29 %) < 1-Me
2.0 equiv.) and 2 mL dimethylacetamide (DMA), 80 °C 3 h (unless specified).
[b] GC/MS yield, average of three experiments, error ± 3 %···.
(
45 %) < 1-OMe (55 %). Interestingly, 1-OMe performed as well
While detailed experimental and computational analysis of
reaction kinetics will be the subject of a future report, some
as “nickamine”, A, (Table 4, entries 8 and 9). Such results sug-
gest that further modifying 1-X, nickamine or other nickel pin-
cer complexes with electron donor groups might be key to im-
proving catalytic performance toward electron rich aromatics.
Next di-halogenated arenes were examined. With 1-OMe as a
catalyst, the substrates 1-bromo-2-fluorobenzene and 1-bromo-
plausible mechanistic pathways can be envisioned based on
[16–20]
previous proposals with similar compounds
and on the
current results. Scheme 2 and Figure S28 give some possible
mechanisms. The first step (step A) in any of the paths is the
reaction between 1-X and NaBH to give the (PNN-X)NiH inter-
4
2
-fluoro-4-trifluoromethylbenzene (Table 4, entries 12 and 13) mediates, 2-X (X = OMe, Me, CF ). Given the reported propen-
3
st
underwent exclusive debromination with the more electron-de- sity for 1 row transition metals to be involved in single-elec-
ficient tri-substituted derivative reacting faster, giving quantita- tron transfer (SET) chemistry, the previous observations of the
tive conversion with half the catalytic loading and only 1 h hemilability of the current ligand, and the observed trends with
heating. In accord with the increased activity with number of electronic effects in the catalysis we currently favor the pathway
electron withdrawing substituents, ortho-dichlorobenzene was that includes steps A–C, as outlined in the top of Scheme 2.
a more receptive substrate (27 % total (ca. 26:1 ClC H :C H ), That is, 2-X, could undergo homolytic cleavage to give
6
5
6 6
I
Table 4 entry 14) than the meta-dichlorobenzene (19 % total (PNN-X)Ni , 3-X, and hydrogen, as hypothesized for other hydri-
[
16,18,19] [20]
(
ca. 18:1 ClC H :C H ), Table 4 entry 15) which, in turn, was donickel(II) (ZNN) pincer complexes (Z = N,
C
), and as
6
5
6 6
better than chlorobenzene (8 %, entry 3). Finally, it was found supported by the observation of H in the NMR spectra during
2
that 1-OMe was an excellent catalyst for HDH reactions with a reaction monitoring. Such dissociation may be favored for more
polyaromatic bromides (entries 16–18), heterocyclic 2-bromo- electron rich ligands as suggested by DFT calculated Ni–H
pyridine (entry 19), and was especially potent for transforma- stretching frequencies (1857, 1861, and 1874 cm-1 for 2-OMe,
tions of benzyl bromide and 1-bromooctane (entries 20 and 2-Me, and 2-OMe, respectively). In the ensuing step B, the 3-X
21).
intermediate would presumably undergo oxidative addition
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4430
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
Scheme 2. Potential mechanisms for hydrodehalogentation reactions of organic halides (R-X) catalyzed by [PNN]Ni pincer complexes.
(OA) to aryl halide or 1-bromooctane, by an SET pathway to an unidentified brown black paramagnetic species (presumably
produce a species with the composition (PNN-X)Ni(R)(Br), 4-X. Ni metal and other species) over the course of a couple hours
Intermediate 4-X, would likely be tetracoordinated nickel(II) at room temperature. Freshly prepared samples of 2-X give
2
2
bound to a κ -ligand radical (to give a 16 e-complex) due to characteristic hydride doublet resonances at δ = –20.6 ( J
=
H
HP
2
2
internal electron transfer from the ligand to the presumptive 109 Hz), –20.7 ( J = 110.0 Hz), and –20.9 ( J = 112 Hz) ppm
highly oxidizing nickel(III) center and pyrazolyl- arm dissocia- with corresponding doublet P NMR resonances at δ = 44.7,
HP
HP
3
1
P
tion. This step B would also be favored for more electron rich 44.6, 44.1 ppm for 2-OMe, 2-Me, and 2-CF , respectively. Ap-
3
1
groups. This latter step rather than step A is thought to be the propriately, the H NMR resonances for 2-X are in between
rate-limiting step given the lack of strong kinetic isotope effect those reported for Liang's [PNP-(o-PPh C H ) N]NiH (δ =
2
6
4 2
H
[
42]
[16]
(
KIE, vide infra) for reactions involving deuteride, and the varia- –18.3 ppm)
and Hu's [NNN]NiH (δ = –22.8 ppm).
More-
H
ble rates of reaction with changing organohalide or para-aryl over, the Ni–H stretch was observed by IR spectroscopy at ν
NiH
–1
substituents. In a subsequent step C, NaBH would transfer a (THF) = 1852, 1853, and 1855 cm , for 2-OMe, 2-Me, and 2-
4
·
+
II
hydride to 4-X to give (PNN-X )Ni (R)(H), 5-X. Reductive elimi- CF , respectively. These frequencies are similar to Sun's
3
–1 [20]
nation would give formally a nickel(0) bound to a ligand cation [CNN]NiH (νNiH = 1894 cm )
derived from compound D
radical which would immediately reorganize to the nickel(I) pin- (Scheme 1) or Lutz's [NNN]NiH (from compound B, Scheme 1;
–
1 [18]
cer, 3-X. An alternate pathway is possible where (PNN-X)Ni(R), νNiH = 1858 cm )
6
but higher energy than Hu's [NNN]NiH
–1 [16]
-X, and a bromo radical are produced by OA between 3-X (νNiH = 1768 cm ). Next, in the presence of TEMPO, no reac-
and organo bromide (perhaps also via dissociation of 4-X). The tion occurred. This result was not particularly informative as
bromo radical could abstract a proton from solvent to generate it only indicates that TEMPO reacts irreversibly with a reactive
HBr. Any HBr present would then react with (PNN)NiR to give intermediate. Therefore, deuterium-labeling studies were inves-
the dehalogenated organic product and (PNN-X)NiBr. Com- tigated. In the reaction of 4-bromobiphenyl with NaBD and
4
pound 2-X would be regenerated by reaction with NaBH . This 8 mol-% 1-OMe (3 h 80 °C) 81 % yield (implying KIE ≥ 1.2 by
4
alternate mechanism was excluded on the basis of deuterium comparison with Table 3, entry 1) of 4-d -biphenyl as the only
1
labeling studies described in the next section. The possibility of product, as indicated by its mass spectrum and NMR data. If a
a Ni(II)/Ni(0) cycle is described in the Supporting information, solvent-assisted radical pathway were operative one might ex-
but such a cycle does not have literature precedence and the pect a majority of fully hydrogenated biphenyl. A similar reac-
heterolytic cleavage of the Ni–H bond in 2-X appears less likely tion in DMF gave 59 % 4-d -biphenyl (implying KIE = 1.3 by
1
than homolytic cleavage.
Several experiments were performed to address the above biphenyl as the only products (Figure S22). The reaction of
possibilities. First, it was possible to generate unstable orange- NaBH in [D ]DMF neither showed an isotope effect nor gave
comparison with Table 3, entry 19) and 41 % unreacted bromo-
4
7
red (PNN-X)NiH intermediates, 2-X, via the reaction between deuterium incorporation. Thus, the path involving solvent radi-
green 1-X and either NaBH in THF or NaHBEt in THF/benzene cals is excluded. Regardless, it is evident that the non-innocence
4
3
mixtures (Supporting Information). The hydrides decompose to (redox or chemical) of the pincer ligand plays a crucial role in
small amounts of H gas (δ = 4.47 ppm, C D ), free ligand, and the catalytic activity. The strategy of decorating diarylamido-
2
H
6 6
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4431
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
based pincer ligands with electron donor groups to improve this set up, the ferrocene/ferrocenium couple matched the literature
[
43,44]
value
00 mV/s.
with E1/2 = +0.52 V in CH Cl2 at a scan rate of
2
the hydrodehalogenation activity may be general to nickel pin-
cer complexes of all donor types.
2
Nickel Complexes.
PNN-OMe)NiCl, 1-Me. A pale green solution of NiCl ·6H O
(
2
2
Conclusion
(0.237 g, 1.00 mmol) in 4 mL of methanol was added slowly to a
magnetically stirred solution of H(PNN-OMe) (0.449 g, 1.00 mmol)
in 4 mL of dichloromethane, whereupon the solution immediately
became darker green. After 5 min of stirring, a solution of tetraeth-
Three new chloronickel(II) complexes of PNN-pincer ligands
have been prepared and shown to be excellent pre-catalysts for
the hydrodehalogenation of aryl iodides, as well as aryl and
ylammonium hydroxide (0.70 mL of 1.43
M in methanol, 1.0 mmol)
alkyl bromides by employing NaBH as a hydride source. The
4
was added causing copious green precipitate. After, the resulting
dark green suspension had been stirred for 3 h, the green solid was
collected by filtration. The filtrate was reduced to 3 mL by rotary
evaporation and a second portion of green precipitate was isolated
catalytic activity towards aryl chlorides is significantly lower
than the bromides but increased with more electron withdraw-
ing substituents in the aryl ring. Deuterium labeling studies
demonstrated that NaBH is the sole hydride source, and the by filtration. The combined green precipitate was rinsed with 3 mL
4
involvement of solvent radicals could be excluded. Substitution methanol and then was dried under oil pump vacuum (10– Torr)
3
at the pincer ligand para-aniline position with electron donat- at room temperature 1 h to give analytically pure 1-OMe (0.512 mg,
yield 94 %) as an olive green powder.
ing groups leads to an increase in reactivity due to the in-
creased basicity of the diarylamido nitrogen lone pair. Further
Mp, > 260 °C (dec.). Anal. Calcd. (Found) for C H ClN ONiP: C,
substitutions may allow for improvements in activity toward 61.98 (62.12); H, 4.27 (4.27); N, 7.74 (7.50). H NMR (400 MHz, CDCl )
2
8
23
3
1
3
B
chloro- or fluoro-aromatics, a direction of current study in our
group.
δ
H
= 8.12 (s, 1 H, H
5
-pz), 7.99 (s, 1 H, H
), 7.55–7.40 (m, 6 H, m- + p-PPh
3
-pz), 7.98 (m, 2 H, o-PPh
2
),
A
7.87 (m, 2 H, o-PPh
), 7.34 (d, J =
2
2
9
.0 Hz, 1 H, H -tolyl), 7.05–7.01 (m, 3 H, H -,H -, and H -C H ), 6.69
6 4 5 6 6 4
(
6
d, J = 2.8 Hz, 1 H, H -tolyl), 6.62 (d, J = 9.0, 2.8 Hz, 1 H, H -tolyl),
3
5
1
3
.58 (m, 1 H, H -pz), 6.54 (m, H -C H ), 3.78 (s, 3 H, OCH ) ppm.
C
Experimental Section
4
3
6
4
3
2
NMR (101.52 MHz, CDCl ) δ : = 162.90 (d, J = 26.1 Hz, C -C H ),
3
C
C-P
2
6 4
Experimental Details.
1
53.07 (s, C -tolyl), 142.89 (s, C -pz), 137.61 (s, C -tolyl), 133.98 (d,
4 5 1
2
A
2
JC-P = 10.3 Hz, o-PPh ), 133.10 (s, C -C H ), 132.90 (d, JC-P
1
=
Syntheses.
2
6
6 4
B
4
0.4 Hz, o-PPh ), 132.35 (d, J = 1.9 Hz, C -C H ), 131.80 (C -
2 C-P 4 6 4 2
General Considerations.
4
A
4
tolyl), 131.50 (d, J = 2.5 Hz, p-PPh ), 131.13 (d, J = 2.4 Hz, p-
C-P
2
C-P
PPh ), 129.45 (d, 1JC-P = 47.1 Hz, ipso-PPh ), 129.25 (d, JC-P
B
B
1
=
Chemicals. Solvents for syntheses, spectroscopic characterization or
electrochemical studies were dried by conventional means and dis-
tilled under Argon prior to use. Solvents used in organic workup or
2
2
A
3
A
59.1 Hz, ipso-PPh ), 129.13 (d, J = 10.7 Hz, m-PPh ), 128.70 (d,
2
C-P
2
3
B
3
JC-P = 11.4 Hz, m-PPh ), 128.59 (d, J = 2.2 Hz, C -pz), 123.89 (s,
2
C-P
3
1
3
chromatographic separations were used as received from commer- C6-tolyl), 123.00 (d, JC-P = 54.7 Hz, C -C H ), 120.63 (d, JC-P
=
1
6 4
3
cial sources. The compound H(PNN) was prepared by the literature
method.[ All other chemicals were used as received from com-
mercial sources. The new ligands H(PNN-OMe) and H(PNN-CF3)
were prepared according to the procedures outlined in the Sup-
porting Information.
11.7 Hz, C -C H ), 118.54 (d, J = 7.3 Hz, C -C H ), 114.20 (s, C -
tolyl), 108.68 (d, JC-P = 1.6 Hz, C -pz), 107.94 (s, C -tolyl), 56.08
(OCH ) ppm. P NMR (162 MHz, CDCl ) δ = 30.59 (s) ppm.
5 6 4 C-P 3 6 4 5
30]
4
4
3
31
3
3
P
Crystals of 1-OMe were grown by vapor diffusion of pentane into
a solution of 30 mg 1-OMe in 2 mL C H .
6
6
Instrumentation and characterization. Melting point determinations
were made on samples contained in glass capillaries using an Elec-
trothermal 9100 apparatus and are uncorrected. Midwest MicroLab,
LLC, Indianapolis, Indiana 45250, performed all elemental analyses.
(
PNN-Me)NiCl, 1-Me. In a manner identical to the above, 0.237 g
(1.00 mmol)NiCl ·6H O in 5 mL of methanol, 0.433 g H(PNN-Me)
2
2
(1.00 mmol) in 5 mL of dichloromethane, and 0.70 mL of 1.43
M
NEt (OH) in methanol (1.0 mmol) gave 0.478 g (91 % yield) 1-Me
as a green powder.
1
13
19
31
4
H, C, F, and P NMR spectra were recorded on either a Varian
00 or 400 MHz spectrometer. Chemical shifts were referenced to
solvent resonances at δH = 7.26 and δC = 77.23 for CDCl or to
3
Mp, > 260 °C (dec.). Anal. Calcd. (Found) for C H ClN NiP: C, 63.86
(63.64); H, 4.40 (4.77); N, 7.80 (8.00). H NMR (400 MHz, CDCl ) δ =
3
28 23
3
1
external references 85 % H PO (aq.) δ = 0 and 1.0
M
CF CO H in
3
4
P
3
2
3
B
H
CDCl at δ = –78.5 ppm. Electronic absorption (UV/Vis/NIR) meas- 8.11 (s, 1 H, H -pz), 8.00 (s, 1 H, H -pz), 7.96 (m, 2 H, o-PPh ), 7.87
3
F
5
3
2
A
urements were made on a Cary 5000 instrument. Abbreviations for (m, 2 H, o-PPh ), 7.55–7.49 (m, 2 H, p-PPh ), 7.49–7.39 (m, 4 H, m-
2
2
NMR and UV/Vis data: br (broad), sh (shoulder), m (multiplet), ps
pseudo-), s (singlet), d (doublet), t (triplet), q (quartet). FTIR spectra
PPh ), 7.34 (d, J = 8.5 Hz, 1 H, H -tolyl), 7.07– 6.97 (m, 3 H, C H ),
6.96 (s, 1 H, H -tolyl), 6.81 (d, J = 8.5 Hz, 1 H, H -tolyl), 6.57 (br m,
3 5
2 6 6 4
(
1
3
were recorded for either solid samples or THF solutions (KBr plates)
in the 4000–500 cm– region on a Thermo Scientific Nicolet iS5 IR
spectrometer equipped with an iD3 Attenuated Total Reflection
2 H, H -C H + H -pz), 2.25 (s, 3 H, CH ) ppm. C NMR (101.52 MHz,
3 6 4 4 3
1
2
CDCl ) δ : = 162.83 (d, J = 25.7 Hz, C -C H ), 142.59 (s, C -pz),
3
C
C-P
2
6
4
5
2
A
141.63 (s, C -tolyl), 133.98 (d, J = 10.4 Hz, o-PPh ), 133.03 (s, C -
1
C-P
2
6
2
B
4
(ATR) accessory. Electrochemical measurements were collected un- C H ), 132.87 (d, J = 10.3 Hz, o-PPh ), 132.29 (d, J = 2.1 Hz,
6
4
C-P
4
2
C-P
B
der a nitrogen atmosphere for samples as 0.1 mm solutions in
p-PPh ), 132.15 (d, J = 2.1 Hz, C -C H ), 131.22 (C -tolyl), 131.12
2
C-P
4
6
4
2
4
A
1
A
CH Cl with 0.1
M
NBu PF as the supporting electrolyte. A three-
(d, J = 2.1 Hz, p-PPh ), 129.38 (d, J = 50.2 Hz, ipso-PPh ),
2
2
4
6
C-P 2 C-P 2
1
B
3
electrode cell comprised of an Ag/AgCl electrode (separated from
the reaction medium with a semipermeable polymer membrane
filter), a platinum working electrode, and a glassy carbon counter 11.7 Hz, m-PPh ), 128.48 (d, J = 2.1 Hz, C -pz), 123.34 (d, J
129.26 (d, J = 47.0 Hz, ipso-PPh ), 129.12 (d, J = 10.4 Hz, m-
C-P 2 C-P
PPh ), 129.09 (s, C -tolyl), 128.91 (s, C -tolyl), 128.69 (d, JC-P
B
3
=
=
2
5
4
A
3
1
2
C-P
3
C-P
electrode was used for the voltammetric measurements. Data were 54.4 Hz, C -C H ), 122.75 (s, C -tolyl), 122.73 (s, C -tolyl), 120.98 (d,
1
6
4
3
6
3
3
collected at scan rates of 50, 100, 200, 300, 400, and 500 mV/s. With
JC-P = 11.7 Hz, C -C H ), 118.82 (d, J = 7.2 Hz, C -C H ), 108.48
5 6 4 C-P 3 6 4
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4432
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
(
d, 4JC-P = 1.6 Hz, C -pz), 20.84 (CH ) ppm. 31P NMR (162 MHz, CDCl3)
decay during collection in each case. Direct methods structure solu-
tions, difference Fourier calculations and full-matrix least-squares
4
3
–
1
δ = 30.32 (s) ppm. FTIR (cm ; solid): 3114 (w), 3047 (w), 2863 (w),
1
P
2
[46]
579 (s), 1500 (s), 1456 (s), 1434 (s), 1290 (s).
refinements against F were performed with SHELXTL. Numerical
absorption corrections based on Gaussian integration over a multi-
faceted crystal model were applied to the data in each experiment.
All non-hydrogen atoms were refined with anisotropic displacement
parameters. Hydrogen atoms were placed in geometrically idealized
positions and included as riding atoms. The X-ray crystallographic
parameters and further details of data collection and structure re-
finements are given in Table S1.
Crystals of 1-Me were grown by vapor diffusion of Et O into a solu-
tion of 30 mg 1-Me in 3 mL C H .
2
6
6
(
(
(
PNN-CF )NiCl, 1-CF . In a manner identical to the above, 0.237 g
3 3
1.00 mmol)NiCl ·6H O in 4 mL of methanol, 0.487 g H(PNN-CF )
2
2
3
1.00 mmol) in 4 mL of dichloromethane, and 0.70 mL of 1.43
M
NEt (OH) in methanol (1.0 mmol) gave 0.541 g (94 % yield) 1-CF
as a teal green powder.
4
3
Computational Details.
Mp, > 260 °C (dec.). Anal. Calcd. (Found) for C H ClF N NiP: C,
5
δH = 8.14 (s, 1 H, H -pz), 8.07 (s, 1 H, H -pz), 7.95 (dd, J = 11.4,
7
7
1
7
6
2
8
20
3 3
1
DFT calculations were performed using the Minnesota M06 meta-
7.92 (57.52); H, 3.47 (3.73); N, 7.24 (7.18). H NMR (400 MHz, CDCl )
3
[
47]
hybrid GGA functional
because it has been found to be useful
5
3
B
A
for affording accurate solutions to a wide variety of computation
problems. Geometry optimizations employed the def2-SV(P) dou-
.9 Hz, 2 H, o-PPh ), 7.86 (dd, J = 12.2, 7.9 Hz, 2 H, o-PPh ), 7.58–
2
2
.51 (m, 2 H, p-PPh ), 7.50–7.42 (m, 4 H, m-PPh ), 7.49 (d, J = 8.4 Hz,
2
2
[
48]
[34,49]
ble-zeta basis set
because we previously found
(and find
H, H -tolyl), 7.39 (s, 1 H, H -tolyl), 7.18 (d, J = 8.4 Hz, 1 H, H -tolyl),
6
3
5
again here) that this method provides excellent agreement (within
0.2 Å) with solid-state structures. Solvent (dichloromethane) effects
were accounted for by using the polarizable continuum model
.15– 7.05 (m, 3 H, H -, H -, and H -C H ), 6.68 (m, 1 H, H -C H ),
4
5
6
6
4
3
6 4
1
3
.63 (m, 1 H, H -pz) ppm. C NMR (101.52 MHz, CDCl ) δ : = 161.94
4
3
C
2
(
d, JC-P = 24.8 Hz, C -C H ), 143.09 (s, C -pz), 148.04 (s, C -tolyl),
2 6 4 5 1
[
50]
[51]
2
A
PCM/UFF,
as implemented in Gaussian 16.
Analytical vibra-
1
34.01 (d, J = 10.4 Hz, o-PPh ), 133.17 (s, C -C H ), 132.86 (d,
C-P 2 6 6 4
2
B
4
tional frequency calculations were carried out to verify that opti-
mized geometries were stationary points. Time-dependent (TD) DFT
J
= 10.4 Hz, o-PPh ), 132.63 (d, J = 2.2 Hz, C -C H ), 131.45
C-P
4
2
C-P
4
6
4
B
A
4
(d, J = 3.0 Hz, p-PPh ), 131.42 (d, J = 3.1 Hz, p-PPh ), 130.58
C-P
2
C-P
2
[
52–54]
3
A
3
methodology
was used for excitation energy calculations of
(C2-tolyl), 129.31 (d, JC-P = 10.7 Hz, m-PPh ), 129.00 (d, JC-P
=
=
2
3
B
1
the first 25 excited states. Each excitation was fitted to a Gaussian
curve with standard deviation of σ = 0.3 eV and spectra were calcu-
lated by summing individual contributions. Tables S4-S7 summari-
zes the results of these studies. Cartesian coordinates are contained
within the PNNX_NiCl_DFT.xyz file found in the Supporting Infor-
mation. Reduction potentials were calculated by using the
1
5
(
(
.8 Hz, C -pz), 128.83 (d, J = 11.9 Hz, m-PPh ), 128.66 (d, J
3 C-P 2 C-P
A
1
B
2.1 Hz, ipso-PPh ), 128.60 (d, J = 48.2 Hz, ipso-PPh ), 124.96
2
C-P
1
2
3
q, J = 3.3 Hz, C -tolyl), 124.51 (d, J = 53.8 Hz, C -C H ), 122.51
C-F 5 C-P 1 6 4
3
1
s, C -tolyl), 121.30 (d, J = 11.5 Hz, C -C H ), 122.16 (q, J
=
=
6
C-P
5
6
4
C-F
97 Hz, CF ), 120.43 (d, 3JC-P = 7.3 Hz, C -C H ),120.10 (q, JC-F
2
1
3
3
6 4
3
3
3.6 Hz, C -tolyl), 119.72 (q, JC-F = 4.0 Hz, C -tolyl), 109.12 (d,
4
3
[
55,56]
4
31
method outlined by Batista
but applying corrections
JC-P = 1.7 Hz, C -pz) ppm. P NMR (162 MHz, CDCl ) δ = 29.80 (s)
4
3
P
[
57]
as outlined by Bühl,
where the ultimate potential obtained
ppm.
o
by –ΔG (red,solv - ox,solv)/F was scaled to the Ag/AgCl electrode by
subtraction of 4.403 eV (i.e., it was referenced against the absolute
SHE potential (CH Cl ) at 4.60 V and then to Ag/AgCl +0.197 V vs.
Crystals of 1-CF were grown by vapor diffusion of pentane into a
solution of 30 mg 1-CF in 2 mL C H .
3
3
6 6
2
2
SHE).
Catalysis.
Deposition Numbers 2027563, 2027564, 2027565 and 2027566 con-
tain the supplementary crystallographic data for this paper. These
data are provided free of charge by the joint Cambridge Crystallo-
graphic Data Centre and Fachinformationszentrum Karlsruhe Access
Structures service www.ccdc.cam.ac.uk/structures.
General Procedure. A 100 mL round-bottomed flask equipped with
a magnetic stir bar was charged with 0.047 g (0.20 mmol) of 4-
bromobiphenyl (or 0.20 mmol of another aryl halide), 8.4 mg
(
(
0.016 mmol, 8 mol-%) catalyst 1-X (or other nickel salt) 15 mg
0.40 mmol, 2.0 equiv.) of NaBH as hydrogen source and 2 mL of
4
dimethylacetamide (DMA) as solvent under argon atmosphere.
Next, the magnetically stirred mixture was heated with 80 °C oil
bath. After 3 h stirring at 80 °C, the mixture was cooled down and
Acknowledgments
passed through silica gel with rinsing 2 mL of dichloromethane. JRG thanks the donors of the Petroleum Research Fund
Then, 1,4-Bis(trimethylsilyl)benzene (11.1 mg) was added as an in- (#58705-ND3) and Marquette University for support. We thank
ternal standard and the resulting mixture was subjected to GC-MS
for the product analysis. The yields of product and unreacted start-
ing material were also calibrated against standard solutions of bi-
phenyl and 4-bromobiphenyl.
Brandon Kizer for initial work with syntheses and characteriza-
tion of 1-Me.
Keywords: Dehydrohalogenation · Nickel Pincer · PNN
ligands · Nickel hydride · Deuterium labelling
Crystallography. X-ray intensity data from a dark green prism of 1-
OMe (CCDC 2027563), a green plate (P1¯, CCDC 2027564) and a
green needle (P2 /n, CCDC 2027565) of 1-Me, and a green needle of
1
1
-CF (CCDC 2027566) were collected at 100.0(1) K with an Oxford
[1] F. Alonso, I. P. Beletskaya, M. Yus, Chem. Rev. 2002, 102, 4009–4092.
[2] C. Menini, C. Park, E.-J. Shin, G. Tavoularis, M. A. Keane, Catal. Today 2000,
3
Diffraction Ltd. Supernova diffractometer equipped with a 135 mm
Atlas CCD detector. Mo(K ) radiation was used for the experiments
with 1-OMe and monoclinic 1-Me while Cu(K ) radiation was used
for the other experiments. Raw data frame integration and Lp cor-
rections were performed with CrysAlis Pro (Oxford Diffraction,
Ltd.).[ Final unit cell parameters were determined by least-squares
refinement of 12291, 13164, 10260, and 12849 reflections of 1-OMe,
6
2, 355–366.
3] G. Chelucci, S. Baldino, G. A. Pinna, G. Pinna, Curr. Org. Chem. 2012, 16,
921–2945.
α
[
α
2
[
[
[
4] V. D. Shteingarts, J. Fluorine Chem. 2007, 128, 797–805.
5] M. Hu, Y. Liu, Z. Yao, L. Ma, X. Wang, Front. Environ. Sci. Eng. 2017, 12, 3.
6] J. Ke, H. Wang, L. Zhou, C. Mou, J. Zhang, L. Pan, Y. R. Chi, Chem. Eur. J.
45]
2019, 25, 6911–6914.
monoclinic 1-Me, triclinic 1-Me, and 1-CF , respectively, with
[7] C. Lei, F. Liang, J. Li, W. Chen, B. Huang, Chem. Eng. J. 2019, 358, 1054–
3
I > 2σ(I) for each. Analysis of the data showed negligible crystal
1064.
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4433
© 2020 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejic.202000721
EurJIC
European Journal of Inorganic Chemistry
[
8] V. Agarwal, Z. D. Miles, J. M. Winter, A. S. Eustáquio, A. A. El Gamal, B. S.
Moore, Chem. Rev. 2017, 117, 5619–5674.
9] R. F. Howe, Appl. Catal. A 2004, 271, 3–11.
[36] L. G. Lavrenova, Russ. Chem. Bull. 2018, 67, 1142–1152.
[37] D. L. Reger, J. R. Gardinier, J. D. Elgin, M. D. Smith, D. Hautot, G. J. Long,
F. Grandjean, Inorg. Chem. 2006, 45, 8862–8875.
[
[
10] Y. Wang, Y. Wei, W. Song, C. Chen, J. Zhao, ChemCatChem 2019, 11, 258–
68.
11] Z.-Z. Zhou, J.-H. Zhao, X.-Y. Gou, X.-M. Chen, Y.-M. Liang, Org. Chem. Front.
019, 6, 1649–1654.
[38] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
[39] A. M. Beaird, T. A. Davis, M. A. Matthews, Ind. Eng. Chem. Res. 2010, 49,
9596–9599.
[40] H. I. Schlesinger, H. C. Brown, A. E. Finholt, J. R. Gilbreath, H. R. Hoekstra,
E. K. Hyde, J. Am. Chem. Soc. 1953, 75, 215–219.
2
[
2
[
[
12] T.-H. Ding, J.-P. Qu, Y.-B. Kang, Org. Lett. 2020, 22, 3084–3088.
13] K. V. Murthy, P. M. Patterson, M. A. Keane, J. Mol. Catal. A 2005, 225, 149–
[41] O. V. Ozerov, C. Guo, L. Fan, B. M. Foxman, Organometallics 2004, 23,
5573–5580.
160.
[
[
14] A. Scrivanti, B. Vicentini, V. Beghetto, G. Chessa, U. Matteoli, Inorg. Chem.
Commun. 1998, 1, 246–248.
15] C. J. E. Davies, M. J. Page, C. E. Ellul, M. F. Mahon, M. K. Whittlesey, Chem.
[42] L.-C. Liang, P.-S. Chien, P.-Y. Lee, Organometallics 2008, 27, 3082–3093.
[
43] D. Bao, B. Millare, W. Xia, B. G. Steyer, A. A. Gerasimenko, A. Ferreira, A.
Contreras, V. I. Vullev, J. Phys. Chem. A 2009, 113, 1259–1267.
44] I. Noviandri, K. N. Brown, D. S. Fleming, P. T. Gulyas, P. A. Lay, A. F. Masters,
L. Phillips, J. Phys. Chem. B 1999, 103, 6713–6722.
Commun. 2010, 46, 5151–5153.
[
[
[
16] J. Breitenfeld, R. Scopelliti, X. Hu, Organometallics 2012, 31, 2128–2136.
17] M. Weidauer, E. Irran, C. I. Someya, M. Haberberger, S. Enthaler, J. Orga-
nomet. Chem. 2013, 729, 53–59.
[
[
[
[
[
45] CrysAlisPro, Agilent Technologies, 2010.
46] G. M. Sheldrick, Acta Crystallogr., Sect. C 2015, 71, 3–8.
47] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215–241.
48] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305.
49] J. R. Gardinier, J. S. Hewage, B. Bennett, D. Wang, S. V. Lindeman, Organo-
metallics 2018, 37, 989–1000.
[
[
[
[
18] C. Rettenmeier, H. Wadepohl, L. H. Gade, Chem. Eur. J. 2014, 20, 9657–
9
665.
19] C. A. Rettenmeier, J. Wenz, H. Wadepohl, L. H. Gade, Inorg. Chem. 2016,
5, 8214–8224.
20] Z. Wang, X. Li, H. Sun, O. Fuhr, D. Fenske, Organometallics 2018, 37, 539–
44.
21] C. Yao, S. Wang, J. Norton, M. Hammond, J. Am. Chem. Soc. 2020, 142,
793–4799.
5
[
[
50] G. Scalmani, M. J. Frisch, J. Chem. Phys. 2010, 132, 114110.
51] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Na-
katsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.
Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Broth-
ers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghava-
chari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
Revision C.01, Gaussian, Inc., Wallingford CT, 2016.
5
4
[
[
22] R. Shi, Z. Zhang, X. Hu, Acc. Chem. Res. 2019, 52, 1471–1483.
23] Y.-P. Zhang, M. Zhang, X.-R. Chen, C. Lu, D. J. Young, Z.-G. Ren, J.-P. Lang,
Inorg. Chem. 2020, 59, 1038–1045.
[
[
24] D. Wu, Z.-X. Wang, Org. Biomol. Chem. 2014, 12, 6414–6424.
25] Y. Kim, J. Lee, Y.-H. Son, S.-U. Choi, M. Alam, S. Park, J. Inorg. Biochem.
2020, 205, 111015.
[26] S. Das, V. Subramaniyan, G. Mani, Inorg. Chem. 2019, 58, 3444–3456.
[27] X. Yang, Z.-X. Wang, Organometallics 2014, 33, 5863–5873.
[28] L.-C. Liang, W.-Y. Lee, Y.-T. Hung, Y.-C. Hsiao, L.-C. Cheng, W.-C. Chen,
Dalton Trans. 2012, 41, 1381–1388.
[
[
[
[
[
[
29] C. P. Yap, Y. Y. Chong, T. S. Chwee, W. Y. Fan, Dalton Trans. 2018, 47, 8483–
8488.
[
[
[
[
[
[
52] G. Scalmani, M. J. Frisch, B. Mennucci, J. Tomasi, R. Cammi, V. Barone, J.
Chem. Phys. 2006, 124, 094107.
53] M. E. Casida, C. Jamorski, K. C. Casida, D. R. Salahub, J. Chem. Phys. 1998,
30] S. Wanniarachchi, J. S. Hewage, S. V. Lindeman, J. R. Gardinier, Organome-
tallics 2013, 32, 2885–2888.
31] B. J. Liddle, R. M. Silva, T. J. Morin, F. P. Macedo, R. Shukla, S. V. Lindeman,
1
08, 4439–4449.
54] A. D. Laurent, D. Jacquemin, Int. J. Quantum Chem. 2013, 113, 2019–
039.
55] L. E. Roy, E. Jakubikova, M. G. Guthrie, E. R. Batista, J. Phys. Chem. A 2009,
13, 6745–6750.
J. R. Gardinier, J. Org. Chem. 2007, 72, 5637–5646.
32] L.-C. Liang, P.-S. Chien, J.-M. Lin, M.-H. Huang, Y.-L. Huang, J.-H. Liao,
Organometallics 2006, 25, 1399–1411.
33] S. Lapointe, E. Khaskin, R. R. Fayzullin, J. R. Khusnutdinova, Organometal-
2
1
lics 2019, 38, 1581–1594.
56] S. J. Konezny, M. D. Doherty, O. R. Luca, R. H. Crabtree, G. L. Soloveichik,
V. S. Batista, J. Phys. Chem. C 2012, 116, 6349–6356.
57] L. Castro, M. Bühl, J. Chem. Theory Comput. 2014, 10, 243–251.
34] J. S. Hewage, S. Wanniarachchi, T. J. Morin, B. J. Liddle, M. Banaszynski,
S. V. Lindeman, B. Bennett, J. R. Gardinier, Inorg. Chem. 2014, 53, 10070–
10084.
[
35] T. J. Morin, S. Wanniarachchi, C. Gwengo, V. Makura, H. M. Tatlock, S. V.
Lindeman, B. Bennett, G. J. Long, F. Grandjean, J. R. Gardinier, Dalton
Trans. 2011, 40, 8024–8034.
Received: July 29, 2020
Eur. J. Inorg. Chem. 2020, 4425–4434
www.eurjic.org
4434
© 2020 Wiley-VCH GmbH