Pairwise H2/D2 Exchange and H2 Substitution at a Bimetallic Dinickel(II) Complex Featuring Two Terminal Hydrides

Article abstract of DOI:10.1021/jacs.7b08629

A compartmental ligand scaffold HL with two β-diketiminato binding sites spanned by a pyrazolate bridge gave a series of dinuclear nickel(II) dihydride complexes M[LNi₂(H)₂], M = Na (Na·2) and K (K·2), which were isolated after reacting the precursor complex [LNi₂(μ-Br)] (1) with MHBEt₃ (M = Na and K). Crystallographic characterization showed the two hydride ligands to be directed into the bimetallic pocket, closely interacting with the alkali metal cation. Treatment of K·2 with dibenzo(18-crown-6) led to the separated ion pair [LNi₂(H)₂][K(DB18C6)] (2[K(DB18C6)]). Reaction of Na·2 or K·2 with D₂ was investigated by a suite of ¹H and ²H NMR experiments, revealing an unusual pairwise H₂/D₂ exchange process that synchronously involves both Ni-H moieties without H/D scrambling. A mechanistic picture was provided by DFT calculations which suggested facile recombination of the two terminal hydrides within the bimetallic cleft, with a moderate enthalpic barrier of ～62 kJ/mol, to give H₂ and an antiferromagnetically coupled [LNi^I₂]^- species. This was confirmed by SQUID monitoring during H₂ release from solid 2[K(DB18C6)]. Interaction with the Lewis acid cation (Na⁺ or K⁺) significantly stabilizes the dihydride core. Kinetic data for the M[L(Ni-H)₂] → H₂ transition derived from 2D ¹H EXSY spectra confirmed first-order dependence of H₂ release on M·2 concentration and a strong effect of the alkali metal cation M⁺. Treating [LNi₂(D)₂]^- with phenylacetylene led to D₂ and dinickel(II) complex 3^- with a twice reduced styrene-1,2-diyl bridging unit in the bimetallic pocket. Complexes [LNi^II₂(H)₂]^- having two adjacent terminal hydrides thus represent a masked version of a highly reactive dinickel(I) core. Storing two reducing equivalents in adjacent metal hydrides that evolve H₂ upon substrate binding is reminiscent of the proposed N₂ binding step at the FeMo cofactor of nitrogenase, suggesting the use of the present bimetallic scaffold for reductive bioinspired activation of a range of inert small molecules.

Full text of DOI:10.1021/jacs.7b08629

Subscriber access provided by Eastern Michigan University | Bruce T. Halle Library
Article
2
2
2
Pairwise H/D Exchange and H Substitution at a Bimetallic
Dinickel(II) Complex Featuring Two Terminal Hydrides
Dennis-Helmut Manz, Peng-Cheng Duan, Sebastian Dechert, Serhiy
Demeshko, Rainer Oswald, Michael John, Ricardo A. Mata, and Franc Meyer
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08629 • Publication Date (Web): 16 Oct 2017
Downloaded from http://pubs.acs.org on October 16, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 13
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Pairwise H2/D2 Exchange and H2 Substitution at a Bimetallic Dinick-
el(II) Complex Featuring Two Terminal Hydrides
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DennisꢀHelmut Manz§‡, PengꢀCheng Duan§‡, Sebastian Dechert§, Serhiy Demeshko§, Rainer Oswald†,
Michael John§, Ricardo A. Mata*,† and Franc Meyer*,§
§ Universität Göttingen, Institut für Anorganische Chemie, Tammannstr. 4, Dꢀ37077 Göttingen, Germany
† Universität Göttingen, Institut für Physikalische Chemie, Tammannstr. 6, Dꢀ37077 Göttingen, Germany
‡ These authors contributed equally to this work.
ABSTRACT: A compartmental ligand scaffold HL with two βꢀdiketiminato binding sites spanned by a pyrazolate bridge gave a
series of dinuclear nickel(II) dihydride complexes M[LNi2(H)2], M = Na (Na∙2), K (K∙2), which were isolated after reacting the
precursor complex [LNi2(µꢀBr)] (1) with MHBEt3 (M = Na and K). Crystallographic characterization showed the two hydride ligꢀ
ands to be directed into the bimetallic pocket, closely interacting with the alkali metal cation. Treatment of K∙2 with dibenzo(18ꢀ
crownꢀ6) led to the separated ion pair [LNi2(H)2][K(DB18C6)] (2[K(DB18C6)]). Reaction of Na∙2 or K∙2 with D2 was investigated
by a suite of 1H and 2H NMR experiments, revealing an unusual pairwise H2/D2 exchange process that synchronously involves both
NiꢀH moieties, without H/D scrambling. A mechanistic picture was provided by DFT calculations which suggested facile recombiꢀ
nation of the two terminal hydrides within the bimetallic cleft, with a moderate enthalpic barrier of ~62 kJ/mol, to give H2 and an
antiferromagnetically coupled [LNiI2]− species. This was confirmed by SQUID monitoring during H2 release from solid
2[K(DB18C6)]. Interaction with the Lewis acid cation (Na+ or K+) significantly stabilizes the dihydride core. Kinetic data for the
1
for the M[L(NiꢀH)2] → H2 transition derived from 2D H EXSY spectra confirmed first order dependence of H2 release on M∙2
concentration and a strong effect of the alkali metal cation M+. Treating [LNi2(D)2]− with phenylacetylene led to D2 and dinickel(II)
complex 3− with a twice reduced styreneꢀ1,2ꢀdiyl bridging unit in the bimetallic pocket. Complexes [LNiII (H)2]− having two adjaꢀ
2
cent terminal hydrides thus represent a masked version of a highly reactive dinickel(I) core. Storing two reducing equivalents in
adjacent metal hydrides that evolve H2 upon substrate binding is reminiscent of the proposed N2 binding step at the FeMo cofactor
of nitrogenase, suggesting the use of present bimetallic scaffold for reductive bioinspired activation of a range of inert small moleꢀ
cules.
preceded by charging of the Fe/S active site with four elecꢀ
INTRODUCTION
Transition metal hydride complexes hold an important place in
trons and four protons, and is accompanied by the obligatory
release of one molecule of H2.10,11 It has recently been shown
inorganic chemistry, and they represent key intermediates in
organometallic catalytic processes involving the transfer of H
atoms, protons or hydride to substrates.1,2 In recent times, a
focus in molecular metal hydride chemistry has been the
search for catalysts mediating the reversible interconversion of
protons and H2, spurred by the interest in using H2 as an enerꢀ
gy carrier.3 Bioinorganic chemistry has provided much inspiꢀ
ration for the field, as natural hydrogenase enzymes use earthꢀ
abundant iron and nickel for H2 formation and oxidation.4 In
particular, the mechanistic interrogation of [NiFe] hydrogenꢀ
ase,5 and the development of efficient bioinspired H2 evolution
catalysts based on nickel,6,7 has led to an impressive progress
in nickel hydride chemistry.8
that photolysis of the charged state, denoted E4(4H) according
to the LoweꢀThorneley kinetic scheme for the nitrogenase
mechanism,11a generates an intermediate E4(H2;2H) described
as an H2 complex of the doubly reduced Fe/S cluster.12 It has
further been suggested that this H2 complex may be a thermalꢀ
ly populated intermediate on the trajectory of reductive elimiꢀ
nation of H2 from, and reaction of N2 with, the E4(4H) state
(Scheme 1).12
Biology is also offering blueprints for the use of metal hydride
species in the reductive binding and activation of inert subꢀ
strates. In this case, reducing equivalents are stored as hyꢀ
drides, preferably at multimetallic sites, which upon reductive
elimination of H2 unmask the lowꢀvalent metal species. This
strategy avoids strong reducing agents and may bypass highly
unfavorable oneꢀelectron reduced substrate intermediates. The
prominent metallobiosite exploiting this mechanism is the
FeMo cofactor of nitrogenase,9 where binding of inert N2 is
Scheme 1. Cartoon of the proposed H2 reductive elimination
scenario that leads to the doubly reduced FeMoco intermediate
prior to N2 binding.
The nitrogenase background provides a strong impetus for
synthetic efforts targeting the use of transition metal hydrides
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 13
as masked lowꢀvalent complexes capable of reductively actiꢀ
vating small molecules.13
ꢀdiketiminato ligands have proven
extremely valuable in this context, and a particularly rich
1
2
β
3
chemistry has evolved from the iron and nickel type A comꢀ
4
plexes with M(ꢀꢀH)2M core mainly developed by the groups
of Holland and Limberg, respectively (Figure 1; including
variants thereof with other aryl and backbone substituents).14,15
These bimetallic hydrides were shown to readily eliminate H2
when treated with external donors or upon heating, leading to
a variety of iron(I) and nickel(I) complexes.16,17 A dinucleating
5
6
7
8
9
Scheme 2. Possible mechanisms of H2/H exchange for a hyꢀ
scaffold composed of two
pyridine afforded a hydrideꢀbridged mixedꢀvalent NiI(
species.18 Mechanistic studies on the reaction of a
diketiminato supported Fe( ꢀH)2Fe complex with azobenzene,
β
ꢀdiketiminato subunits bridged by
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀH)NiII
dride complex which in case of D2/H exchange leads to H/D
ꢀ
scrambling and formation of HD.23,24
β
ꢀ
ꢀ
Here we present a new ligand scaffold that provides two βꢀ
which leads to N=N bond cleavage, indicated that substrate
binding occurs prior to rate limiting HꢀH bond formation and
triggers reductive H2 release.19
diketiminate binding pockets spanned by a pyrazolate bridge.
This system was designed to feature two terminal MꢀH units in
close proximity within a bimetallic pocket, potentially prone
to H2 elimination and twoꢀelectron reductive activation of
substrates while capitalizing on metalꢀmetal cooperativity. In
contrast to type A, B, and related systems, however, any bridgꢀ
ing hydrides, ꢀꢀH, should be prevented because of the larger
metalꢁꢁꢁmetal separation enforced by the pyrazolate bridge. We
report a family of dinickel(II) dihydride complexes of that new
ligand, their unusual H2/D2 exchange reactivity that proceeds
without any scrambling and synchronously exchanges both
terminal NiꢀH, the effect of Lewis acids on hydride stability
and reactivity, and we demonstrate that the dinickel(II) dihyꢀ
dride indeed serves as a masked dinickel(I) synthon capable of
reductive substrate binding.
Figure 1
complexes supported by
.
Selected examples of iron(II) and nickel(II) hydride
ꢀdiketiminato ligands.
β
RESULTS AND DISCUSSION
A cyclophane type scaffold containing three
binding pockets, developed recently by Murray and coworkꢀ
βꢀdiketiminato
Synthesis and characterization of dinickel dihydride com-
plexes. A multiꢀstep synthetic route starting from readily
available 3,5ꢀdisubstituted pyrazole derivatives was developed
to afford the new proligand H3L in good yields and gram
quantities; details of the reaction sequence and experimental
protocols are provided in the Supporting Information (SI).
Deprotonation of H3L with nꢀBuLi and subsequent addition of
NiBr2(dme) (dme = dimethoxyethane) at room temperature, all
in THF solution, gave the beige bromidoꢀbridged complex
[LNi2(µꢀBr)] (1) (Scheme 3). 1 is a diamagnetic compound
with apparent C2v symmetry according to NMR spectroscopy.
Single crystals were obtained from CH2Cl2 or CHCl3 soluꢀ
tions, and the molecular structure of 1 determined by Xꢀray
diffraction confirmed that the two nickel(II) ions are hosted in
the two {N3}ꢀtridentate binding sites of the trianionic ligand
scaffold, bridged by the pyrazolate and an exogenous bromide
(Figure 2). The metal ions are found in roughly squareꢀplanar
coordination environment with d(NiꢁꢁꢁNi) = 3.8066(5) Å.
However, the bromide appears to be a bit small for the bimeꢀ
tallic pocket and is thus pulled into the cleft, which is reflected
by the angles N3ꢀNi1ꢀBr and N5ꢀNi2ꢀBr (~168°) which deviꢀ
ate significantly from linearity.
ers, gave access to tris(ꢀꢀhydrido) trimetallic clusters such as
the triiron(II) complex B (Figure 1).20 The latter showed COꢀ
induced reductive elimination of H2 to produce a lowꢀvalent
FeI2FeII species that reversibly regenerates the trihydride comꢀ
plex under H2 atmosphere.21
Exchange of the hydrides with D2 in type A diiron(II) dihyꢀ
drides produced all isotopologues with H/D scrambling,
though the mechanism could not be fully elucidated. It was
suggested, however, that H2/D2 exchange does proceed at the
dimeric molecule and not via a transient monomer with a
terminal FeꢀH.22 It should be noted that hydride complexes
that can bind H2 (or D2) cis to the hydride usually undergo a
facile low barrier (~5 kcalꢁmolꢀ1 or less) ligand exchange.23
Such effective intramolecular site exchange of H atoms beꢀ
tween H2 ligands and hydride ligands leads to a single hydride
1
resonance in H NMR, and to the rapid incorporation of deuꢀ
terium accompanied by the formation of HD when the protio
form of the hydride complex is treated with D2 gas. Mechanisꢀ
tically the H2/H exchange can be considered as an oxidative
addition/reductive elimination route via a trisꢀhydride interꢀ
mediate, or as a nondissociative single step transfer of a hyꢀ
drogen atom involving quantum mechanical exchange pheꢀ
nomena (Scheme 2).23,24
ACS Paragon Plus Environment

Page 3 of 13
Journal of the American Chemical Society
suitable for Xꢀray diffractometry for both compounds were
N
N
1
2
3
4
5
6
7
8
obtained by slow diffusion of pentane into THF solutions at
room temperature. The molecular structures of K∙2 and Na∙2
are shown in Figure 3 (the asymmetric unit of K∙2 contains
two crystallographically independent but very similar moleꢀ
cules; only one is shown). The two hydride ligands in each
complex are directed into the bimetallic pocket, as anticipated,
and the coordination spheres of all nickel ions are only slightly
distorted from squareꢀplanar. In both cases, the alkali metal
cation is closely associated with the anionic [LNi2(H)2]− core,
but the location of Na+ and K+ is distinct. In K∙2, the K+ ion is
hold between the two aryl rings of the DIPP substituents via
cationꢀπ interactions and is located within the plane defined by
the pyrazolateꢀbridged dinickel dihydride core, presumably
supported by attractive K+∙∙∙hydride interactions (d(K+ꢁꢁꢁH) =
2.45(3) – 2.53(3) Å). A related situation has been observed in
a trinuclear Ni3H4 complex capped by βꢀdiketiminato ligands,
where K+ ions sandwiched between aryl rings were located
close to two bridging hydrides (2.47(3) Å).25 The distance
between K+ and the centroid of the DIPP aryl rings in K∙2 is
2.84 Å, which lies in the typical range for cationꢀπ bonding of
K+ to aromatic systems.26 In Na∙2, the smaller Na+ ion is situꢀ
ated above the pyrazolateꢀbridged dinickel dihydride core and
outside of the DIPP cleft, with close contacts to the two hyꢀ
drides, the two pyrazolateꢀN, and coordinated by two addiꢀ
tional THF ligands. Ni∙∙∙Ni distances for K∙2
(4.1584(7)/4.1636(7) Å) and Na∙2 (4.1049(5) Å) are similar,
suggesting that the alkali metal cation does not exert any maꢀ
jor structural influence on the dinickel dihydride core, and that
the arrangement of the two DIPP aryl rings is favorably suited
for accommodating a K+ cation.
N
N
H
H
H
N
N
H3L
1) nBuLi, ꢀ 780C
THF
2) NiBr2(dme), RT
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
N
N
Ni
Ni
N
Br
N
1
KHBEt3
N
NaHBEt3
N
N
N
N
H
N
H
N
H
N
Ni
Ni
Ni
N
Ni
N
N
N
H
Na
(THF)2
K
.2
K
Na.2
DB18C6
[K(DB18C6)]+
N
N
N
N
Ni
Ni
N
N
H
H
.2
[K(DB18C6)]
Scheme 3. Preparation of complexes [LNi2(µꢀBr)] (1),
K[LNi2(H)2] (Kꢁ2), Na[LNi2(H)2] (Naꢁ2), and
[LNi2(H)2][K(DB18C6)] (2[K(DB18C6)]).
Figure 2. Molecular structure (30% probability thermal ellipꢀ
soids) of 1. Solvent molecules and hydrogen atoms omitted for
clarity).
Figure 3. Molecular structures (30% probability thermal ellipꢀ
soids) of K[LNi2(H)2] (Kꢁ2; top; only one of two independent
molecules shown) and Na[LNi2(H)2] (Naꢁ2; bottom). Hydroꢀ
gen atoms except the Niꢀbound hydride omitted for clarity.
Treatment of [LNi2(µꢀBr)] with 2.5 equivalents of MHBEt3 (M
= Na and K) in THF smoothly gave the dihydride complexes
M[LNi2(H)2] (Kꢁ2 for M = K, Naꢁ2 for M = Na). Crystals
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 13
1
2
3
4
Table 1. Selected distances (Å) and angles (deg) for 1, K∙2, Na∙2, 2[K(DB18C6)], and K∙3
Kꢁ2
Naꢁ2
2[K(DB18C6)]
Kꢁ3
1
Ni–N
1.843(2) ꢀ 1.897(2)
1.878(2) ꢀ 1.911(2)
1.866(2) ꢀ 1.921(2)
1.862 (2) ꢀ 1.920(2)
1.850(2) ꢀ 1.949(2)
Ni–Br
2.3812(4) / 2.3978(4)
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
5
Ni–H
ꢀ
1.32(2) ꢀ 1.38(3)
1.37(2) / 1.40(2)
1.44(3) / 1.46(3)
Ni–C
K–C
K–H
K–N
Na–N
Na–H
NiꢁꢁꢁNi
NiꢁꢁꢁK
NiꢁꢁꢁNa
N–Ni–N
(opposite)
N–Ni–Br
(opposite)
N–Ni–H
(opposite)
N–Ni–C
(opposite)
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
1.896(3) ꢀ 1.928(2)
3.005(2) ꢀ 3.112(2)
ꢀ
2.875(2) ꢀ 3.138(2)
ꢀ
ꢀ
3.876(1) / 3.890(1)
3.454(6) ꢀ 3.786(7)
ꢀ
6
7
8
9
3.098(2) ꢀ 3.239 (2)
2.45(3) ꢀ 2.53(3)
ꢀ
ꢀ
2.567(2) / 2.705(2)
2.26(2) / 2.50(3)
4.1049(5)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
3.8066(5)
ꢀ
ꢀ
4.1584(7) / 4.1636(7)
3.7807(5) ꢀ 3.8307(5)
ꢀ
4.0636(7) / 4.1152(7)
ꢀ
ꢀ
ꢀ
2.8468(8) / 3.0084 (8)
176.93(7) / 178.19(7)
175.98(9) / 178.02(9) 176.55(7) ꢀ 178.66(7)
177.22(8) / 177.25(8)
162.04(8) ꢀ 173.60(9)
167.91(7) / 168.20(6)
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
172.5(12) ꢀ 174.7(10)
ꢀ
172.7(9) / 173.4(10)
ꢀ
171.0(13)/172.4(13)
ꢀ
ꢀ
153.62(9) ꢀ 167.12(9)
To further assess the effect of the alkali metal cations and to
prepare dinickel dihydride complex with a vacant cleft, Kꢁ2
was treated with [2,2,2]cryptand or dibenzo(18ꢀcrownꢀ6)
(DB18C6) to separate the K+ cation from the [LNi2(H)2]−
complex anion. No obvious color change is associated with
these reactions, but the ionic products become poorly soluble
in THF; these differences in solubility suggest that the alkali
metal cations remain closely associated with the [LNi2(H)2]−
anion and that contact ion pairs are present in THF solutions
of Kꢁ2 and Naꢁ2. Single crystals of [LNi2(H)2][K(DB18C6)]
(2[K(DB18C6)]) were obtained from THF/Et2O solutions and
subjected to Xꢀray diffraction. The asymmetric unit contains
two crystallographically independent molecules with crystalꢀ
lographically imposed C2 symmetry (the idealized point group
of the anion is C2v); one of the two molecules is shown in
Figure 4. The core structure of the “naked” [LNi2(H)2]− comꢀ
plex anion shows no significant differences compared to the
neutral compounds Kꢁ2 and Naꢁ2. Selected metrical parameꢀ
ters of the three dinickel dihydrido complexes are listed in
Tables 1 and 2. 2[K(DB18C6)] appears to feature longer NiꢀH
bonds and a shorter H1ꢁꢁꢁH1’ distance than Kꢁ2 and Naꢁ2 (in
fact, NiꢀH bonds in the range 1.32(2) ꢀ 1.40(2) Å determined
for Kꢁ2 and Naꢁ2 are shorter than most terminal NiꢀH bonds
reported in literature.8,15,27 However, all these data have to be
considered with caution because of inherent ambiguities of H
atom positions derived from Xꢀray crystallography. In the
absence of neutron diffraction data, NMR and IR signatures
(vide infra) can be considered more reliable for assessing
differences among the NiꢀH moieties in the three compounds.
Figure 4. Molecular structure (30% probability thermal ellipꢀ
soids) of the anion of 2[K(DB18C6)] (only one of two indeꢀ
pendent molecules shown). Most hydrogen atoms omitted for
clarity, except for the nickelꢀbound hydrides. Symmetry transꢀ
formations used to generate equivalent atoms: (') –x, y, 3/2–z.
Table 2. Selected metrical and spectroscopic data for the three
complexes K∙2, Na∙2 and 2[K(DB18C6)].
Compound
K∙2
Na∙2
2[K(DB18C6)]
4.115
4.064
Ni···Ni [Å]
4.158/4.164
4.105
1.32(2)
1.38(2)
2.33(3)
/2.33(3)
2.45(3)/
2.53(3)/
2.49(2)/
2.53(3)
ꢀ24.16
1.37(2)
1.40(2)
1.46(3)
1.44(3)
NiꢀH [Å]
NiHꢁꢁꢁHNi
[Å]
2.28(4)
2.11(5)/2.14(5)
KꢁꢁꢁH/NaꢁꢁꢁH
[Å]
2.26(2)
2.50(3)
ꢀ
δ(1H) (ppm)
ν (cmꢀ1)
ꢀ23.54
ꢀ
1961/1367
1846/1337
1907/ꢀ
(NꢀH/NꢀD) a
a ATRꢀIR spectra recorded on solid samples.
NiꢀH stretching vibrations for solid samples of the three comꢀ
plexes are found at 1961 cmꢀ1 (K∙2), 1846 cmꢀ1 (Na∙2) and
1907 cmꢀ1 (2[K(DB18C6)]), well within the range 1690ꢀ
ACS Paragon Plus Environment

Page 5 of 13
Journal of the American Chemical Society
2000 cmꢀ1 reported in literature for terminal NiꢀH stretching
bands.8 Band assignment has been corroborated by recording
1
2
3
4
5
6
7
8
spectra for the corresponding NiꢀD compounds (vide infra)
which show the expected isotopic shift of ν(NiꢀH)/ν(NiꢀD) ≈
1.4 (Table 2; spectra are shown in the Supplementary Inforꢀ
mation). According to a SQUID measurement, solid K∙2 is
diamagnetic over the temperature range 2 – 295 K (Figure
S47), reflecting the lowꢀspin d8 configuration of the nickel(II)
ions. In contrast, 2[K(DB18C6)] proved to be highly sensitive
and decayed when crystals were dried under reduced pressure
or kept at rt for some time. These findings will be discussed in
more detail below.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
Figure 5. H NMR spectrum (77 MHz, 298K) of Kꢁ2-D in
The SQUID results are in agreement with solution properties
of the complexes. 1H NMR spectra of K∙2 and Na∙2 in THFꢀd8
reveal high field shifted resonances at ꢀ24.16 (K∙2), ꢀ23.54
(Na∙2) assigned to the Niꢀbound hydrides (Figures S8 and S16,
respectively; Table 2). These values are within the wide range
of chemical shifts reported for terminal nickel hydrides.8 The
different alkali metal cations in K∙2 and Na∙2 obviously have
only a minor effect on the electronical shielding of the hyꢀ
drides. 2[K(DB18C6)] appears to be rather unstable in soluꢀ
tion and gives rise to broad, paramagnetically shifted resoꢀ
nances (Figure S25).
THF (bottom spectrum) and upon treatment with H2, yieldꢀ
ing Kꢁ2 and D2 (middle spectrum). HD is formed only after
long reaction times (top spectrum). Residual solvents are
marked (*).
To directly monitor the exchange processes and to confirm the
synchronous exchange of both hydrides of a single molecule
of 2−, the following NMR experiment was designed: a solution
of K[LNi2(H)2] (K∙2) was treated with a small amount of HD,
not sufficient for full conversion to K[LNi2(H)(D)] (K∙2-HD)
but adjusted to provide roughly equal peaks intensities for the
two isotopologues. The reaction mixture then contained K∙2,
H2/D2 exchange reaction. Upon addition of D2 (atmospheric
pressure) to degassed solutions of K∙2 or Na∙2 in THFꢀd8, both
nickel bound hydrogen atoms are rapidly exchanged to give
the deuterated congeners K[LNi2(D)2] (K∙2-D) and
Na[LNi2(D)2] (Na∙2-D), respectively, and these reactions are
reversed upon addition of H2 to solutions of the deuterated
complexes in THF (Scheme 4). The latter reaction from 2-D−
to 2− is most conveniently followed via 2H NMR spectroscopy,
which shows the disappearance of the signals for NiꢀD around
1
K∙2-HD, H2, and HD, all of which are detectable in the H
NMR spectrum. The hydride resonances of K∙2 and K∙2-HD
differ slightly (ꢀ24.16 vs. ꢀ24.18 ppm at rt; ꢀ24.03 vs. ꢀ24.05 at
273 K) because of a secondary isotope effect between the two
1
hydrides. Importantly, the twoꢀdimensional H EXSY specꢀ
trum of the mixture (Figure 6) revealed correlations only beꢀ
tween K∙2 and H2 as well as between K∙2-HD and HD, clearly
evidencing a synchronous (i.e., pairwise) exchange of H2 and
HD, respectively, without any scrambling. All possible exꢀ
change processes in this scenario are shown in the upper part
of Figure 6.
1
ꢀ24 ppm (concomitant for the appearance of NiꢀH in the H
NMR spectrum) and the rise of a signal at 4.57 ppm originatꢀ
ing from D2 (Figure 5). Surprisingly, no HD formation is
2
observed during the initial stages of the reaction: the H NMR
spectrum after 75 min still shows solely the presence of D2,
while signals for HD, identified by a doublet at 4.58 ppm with
1JHD = 42.7 Hz, evolve only slowly after longer reaction times
(see spectrum after 7 h in Figure 5). HD formation during later
stages of the reaction presumably occurs via secondary proꢀ
cesses that may involve trace amounts of Niꢀcontaining deꢀ
composition products. These findings indicate that 2− underꢀ
goes a pairwise H2/D2 exchange process that synchronously
involves both NiꢀH (or NiꢀD, respectively) moieties, clearly
distinct from the common monometallic mechanisms sketched
in Scheme 4. This process is observed for both compounds,
K∙2-D and Na∙2-D, showing that it does not depend on the
type of, and mode of association with, the alkali metal cation.
D2
H2
N
N
N
N
N
H
N
N
D
N
Ni
Ni
Ni
Ni
N
N
N
N
H
D
D2
H2
2-D
2
Scheme 4. Synchronous pairwise H2/D2 exchange reaction
between [LNi2(H)2]− and [LNi2(D)2]−
.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 13
exchange of H2/D2. To explore this potential mechanism, two
1
2
3
4
5
6
7
8
sets of constrained optimizations were performed, fixing the
distance between the two hydrides in K∙2 and 2−. Figure 7
shows the results at the level of theory used for the optimizaꢀ
tions (BP86ꢀD3/SVP), and also energies for the metaꢀGGA
M06L functional with a larger basis, but using the same geꢀ
ometries. For each point in the curve the singlet, the triplet and
the broken symmetry state (with the two lone electrons at the
NiI centers) were computed. Down to d(HꢀH) = 0.95 Å the
pure singlet state is the most stable. At even shorter HꢀH disꢀ
tances, the broken symmetry state is found to be lowest, with
the triplet still lying higher above in energy. The energetic
order of the states was further confirmed with the B3LYP and
PBE0 functionals at d(HꢀH) = 0.8 Å, providing the same qualiꢀ
tative picture. All results reported (including the optimized
structures) correspond to the lowest electronic state found at
each point.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. 1Hꢀ1H EXSY spectrum (400 MHz, 0.5 s mixing, 273
K) of a mixture of Kꢁ2, Kꢁ2ꢀHD, H2 and HD in THFꢀd8 (botꢀ
tom) and the possible exchange processes (top).
To provide mechanistic insight for the H2/D2 exchange process
and to explain the effect of the alkali metal ions on the stabilꢀ
ity of (and H2 release from) the dihydride complex, density
functional theory computations were performed.
DFT Calculations. Possible pathways for H2/D2 exchange
were probed through DFT calculations. For ease of discussion,
we will consider 2− as the reactant, D2 as the entering molecule
and 2-D−+H2 as the products. The alkali metal cation was
removed in this initial set of calculations to allow for more
flexibility in the exchange paths.
Figure 7. Potential energy surface plots for the recombination
of the Niꢀbound H atoms in 2− and Kꢁ2. Relaxed scans were
computed along the HꢀH distance at the BP86ꢀD3/def2ꢀSVP
level. The dotted vertical line represents the bond distance in
the H2 molecule at the same level of theory (0.767 Å). The
reference point is provided by the most stable geometry, in
both cases a distance of 2.1 Å.
The first question addressed the potential binding of D2 to the
complex. Although η2ꢀH2 coordination to transition metals is
well known, no stable minimum for such structure could be
found in the present case. Indeed, NiII dihydrogen complexes
are very rare.28,29 All attempts for sideꢀon D2 binding to the
NiII ions in 2− resulted in a weak endꢀon interaction with one
of the metal (d(HꢁꢁꢁNi) ≈ 2.6 Å)). We carried out an extensive
search of minima, placing the entering molecule close to the
Ni ions in different orientations (see SI for details), but other
structures were less stable than the one with endꢀon bound D2.
All concerted pathways for insertion of D2 starting from these
minima were particularly high in energy. The lowest connectꢀ
ed path for insertion was through an early H2 recombination in
the pocket, but such path yields a lower bound for the elecꢀ
tronic energy barrier of 136 kJ/mol. This would be much too
high in energy, so that we excluded the possibility of a conꢀ
certed mechanism in which the incoming D2 forms NiꢀD
bonds at the same time as the NiꢀH bonds are broken.
Both BP86 and M06L agree in that H2 recombination is much
more facile when the K+ cation is not present. We start by
discussing the latter curves. An estimate to the barrier for
removal of H2 is provided at distances slightly above the optiꢀ
mal HꢀH bond value for free H2 (0.767 Å) In the case of 2−,
both methods agree in a barrier of about 50 kJ/mol. Carrying
out a linear interpolation of the M06ꢀL/TZVP energy curve,
the barrier can be approximated by the energy value at the H2
equilibrium bond distance, which is 62.5 kJ/mol. Differences
between the results using the two functionals are foremost in
the shape of the curve. In the case of BP86, we found a shalꢀ
low minimum, which we were able to fully optimize and charꢀ
acterize at d(HꢀH) = 1.16 Å; this feature is not visible in the
M06L curve. The electronic structure of this minimum is
similar to the hydride complex, still keeping the NiꢀH bonds.
An alternative to the substitution mechanism is a nonꢀ
concerted but stepwise pathway, with H2 leaving the pocket
and subsequent D2 coordination at the vacant NiI sites. This
would also be consistent with the pairwise nonꢀscrambling
ACS Paragon Plus Environment

Page 7 of 13
Journal of the American Chemical Society
Due to the flatness of the potential in this region it was not reported for some porphyrin ruthenium monohydride comꢀ
1
2
3
4
5
6
7
8
possible to converge the transition state to the lower neighborꢀ
ing minimum. The impact of the zeroꢀpoint vibrational energy
correction on the energies cited above should be very small
since the NiꢀH harmonic vibrational frequency is roughly half
of the HꢀH stretching frequency in the hydrogen molecule.
plexes, proposed to proceed via an intermolecularly bridging
dihydrogen intermediate.30 Using a dimeric system with two
cofacial porphyrin subunits it was then possible to detect the
bridged dihydrogen complex, for which NMR evidence sugꢀ
gested that the H2 ligand is bound with the HꢀH axis perpenꢀ
dicular to the RuꢁꢁꢁRu axis.31 Steric constraints of the present
pyrazolateꢀbased ligand scaffold clearly enforce a different
geometric situation in 2−.
In the case of K∙2, the estimated activation barrier for removal
of H2 would lie between 80ꢀ105 kJ/mol (102.2 kJ/mol in the
case of M06ꢀL/TZVP). Along the relaxed surface scan the K+
remained relatively fixed in its position between the two aryl
rings, approximately in plane with the leaving H2 and thereby
raising the energy. It is likely that the reported values correꢀ
spond to an upper estimate. The cation could potentially
change its position along the faces of the aryl rings, lowering
the energy barrier for H2 formation.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The electronic structure calculations carried out quite clearly
go against any concerted (associative or associate interchange)
mechanism for substrate substitution. Instead, in a first step,
the two Niꢀbound hydrides should recombine to H2, then exit
the pocket and leave two Ni(I) centers with a vacant bimetallic
cleft. The barrier for H2 formation is quite low in the absence
of K+ (about 50 kJ/mol), but could also be lowered in the
presence of cations due to binding equilibria ins solution, viz.
if there is dynamic exchange with the solvent or the cation is
readily displaced. The calculations have not included quantum
tunneling effects which would further lower the effective
barrier.
Figure 8. χMT vs. T measurements in the temperature range of
2 – 295 K at 0.5 T for a solid sample of 2[K(DB18C6)] after
different sample preparations: (i) pristine crystals kept under
residual mother liquor in a sealed tube (red data points); (ii)
after drying in gloveꢀbox atmosphere for 1 h (blue data
points); and (iii) after prolonged drying under vacuum for 15
hours. The solid red line represents the best fit for two coupled
S = ½ spin centers with J = –69.5 cm–1 and g = 1.79 (
H2 loss from the “naked” dinickel dihydride. The experiꢀ
mentally observed instability of 2[K(DB18C6)] in combinaꢀ
tion with the above DFT results suggested that H2 release from
the dinickel dihydride core 2− should be facile in the absence
of a stabilizing alkali metal ion within the dihydride cleft.
ˆ
ˆ ˆ
H = −2JS1S 2
Hamiltonian; see SI for details); the unreasonaꢀ
g value reflects incomplete conversion of
Keeping
a
sample of freshly prepared crystalline
bly low
2[K(DB18C6)] in a closed flask under N2 at rt for 6 hours and
monitoring the headspace by GCꢀMS indeed shows the graduꢀ
al formation of H2. Reductive elimination of H2 from 2− can be
expected to yield a {LNiI2} species, and therefore a series of
challenging SQUID measurements were performed (Figure 8;
see SI for details). A sample of crystalline material of
2[K(DB18C6)] that was quickly removed from the mother
liquor without drying and kept under residual solvent in a
sealed tube showed the expected diamagnetism in the temperaꢀ
ture range 2–295 K. When crystals of 2[K(DB18C6)] were
dried in gloveꢀbox atmosphere for 1 h (mother liquor evapoꢀ
rated without applying vacuum), however, a paramagnetic
contribution arises that amounts to 0.32 cm3·mol–1·K (correꢀ
sponding to 1.6 ꢀB) at 295 K. SQUID data for a sample of
crystalline 2[K(DB18C6)] that has been thoroughly dried
under vacuum for 15 h (resulting in a powder sample) shows a
magnetic moment of 0.5 cm3·mol–1·K (corresponding to 2.0
ꢀB) at 295 K, not too far from the value expected for two S =
½ ions (2.45 ꢀB for g = 2.0). The decrease of ꢀB upon lowering
the temperature (shown as χMT vs. T plot in Figure 8) indicates
significant antiferromagnetic coupling. These experiments
provide experimental evidence for the idea that the dinickel(II)
dihydride core 2−, in the absence of any alkali metal ion within
the dihydride cleft, is prone to facile loss of H2 and can be
viewed as a masked dinickel(I) species. The presence of K+ (or
Na+) obviously stabilizes the dinickel(II) dihydride complex
and prevents H2 loss, in line with the DFT results.
2[K(DB18C6)] to the dinickel(I) species.
Kinetics of the H2/D2 exchange. 1Hꢀ1H EXSY spectra of
mixtures containing K∙2 (or Na∙2) and H2, recorded in the
temperature range from 253 to 278 K (in steps of 5 K),
showed exchange between H2 and the Niꢀbound hydrides.
While a full EXSYCALC analysis was hampered by partial
peak overlap that introduced large errors, it was possible to
use two EXSY peaks to obtain pseudoꢀfirst order rate conꢀ
stants for the transition [L(NiꢀH)2]− → H2 in the initial build
up regime (kexτmix < 0.15), which is in the temperature range
from 253 to 278 K (spectra recorded in steps of 5 K; Table S1
in the SI). Samples with different concentrations of K·2 yieldꢀ
ed essentially the same pseudoꢀfirstꢀorder rate constants (Figꢀ
ure 9), in agreement with the reaction being first order in
[L(NiꢀH)2]−. Arrhenius plotting (Figure 9) gives an apparent
activation energy Ea = (74 ± 5) kJ mol‒1 for K·2, attributed to
the rate determining loss of H2 from the dinickel(II) dihydride
core (Eyring analysis gives ꢂHǂ = (72 ± 5) kJ mol‒1, see SI;
the limited temperature range does not allow to derive a reliaꢀ
ble value for ꢂSǂ). Pseudoꢀfirst order rate constants for the
transition [L(NiꢀH)2]− → H2 in case of Naꢁ2 were found to be
significantly different (Figure 9; Ea = (39 ± 3) kJ mol‒1 for
Naꢁ2), which reveals a strong dependence on the alkali metal
cation in qualitative agreement with the computational results
(vide supra). This furthermore indicates that H2 exchange is
likely affected by the solution dissociation equilibria between
M∙2 and solvated 2− + M+. More detailed kinetic studies conꢀ
The proposed scenario of H2 release from 2− is reminiscent of
the oxidatively induced bimolecular H2 reductive elimination
ACS Paragon Plus Environment

Journal of the American Chemical Society
sidering those equilibria will thus be required for deriving a
Page 8 of 13
1
comprehensive mechanistic picture and reliable activation
2
3
4
parameters. A similar hydride exchange between K∙2-D and
2
D2 was observed in Hꢀ2H EXSY spectra (Figure S12) but not
quantitatively analyzed due to the much lower sensitivity of 2H.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Molecular structure (30% probability thermal
ellipsoids) of one of the two crystallographically independent
molecules of Kꢁ3; most hydrogen atoms and disorder omitted
for clarity.
The molecular structure of K∙3 shows the phenylacetylene
substrate bound within the bimetallic pocket, having replaced
the two hydrides. The two nickel ions are again found in disꢀ
torted squareꢀplanar coordination environment, suggesting
they are in the +II oxidation state in accordance with the diaꢀ
magnetic nature of K∙3. In line with this, the length of the
central CꢀC bond (C41–C42) of 1.358(3) Å is indicative of a
double bond, consistent with twofold reduction of the pheꢀ
nylacetylene substrate. The bridging unit thus is best described
as a dianionic styreneꢀ1,2ꢀdiyl group. The potassium cation in
K∙3 is associated with the πꢀsystem of the bridging C=C unit
and is further ligated by three THF molecules. The Ni∙∙∙Ni
separation in K∙3 is in the common range (3.88 Å), but beꢀ
cause of the steric requirements of the substrate phenyl group,
the bimetallic scaffold is significantly more distorted and the
DIPP substituent on the Ni2 side is bent out of the plane of the
pyrazolateꢀbased dinickel core. The resulting C1 symmetry of
the entire complex is reflected in the NMR spectra, which
could be fully assigned (see SI for details). Furthermore, temꢀ
peratureꢀdependent NMR spectra revealed hindered rotation of
the phenyl group around the C42ꢀC43 bond. Lineshape analyꢀ
sis yielded activation parameters ꢁHǂ = (52 ± 1) kJ molꢀ1, ꢁSǂ =
(ꢀ18 ± 4) J molꢀ1 Kꢀ1, and ꢁGǂ298 = (57 ± 1) kJ molꢀ1 (see SI for
details), which is in the range of the rotational barriers measꢀ
ured for some phenylallyl alkali metal salts (Figures S27ꢀ29
and Tables S3 and S4).32
Figure 9. Arrhenius plot for the M[L(NiꢀH)2] → H2 exchange
for Kꢁ2 (0.1, 0.2, 0.3 and 0.4 M) and Naꢁ2 (0.2 M) with data
points determined in 5 K steps between 278 K and 253 K.
Rate constants k/sꢀ1 were extracted from the integral ratios of
exchange and diagonal peaks in 2D H EXSY spectra at the
various temperatures.
1
H2/D2 substitution by substrate. Removal of H2 from 2− (or
D2 from 2-D−) formally leaves two reducing equivalents on the
{LNiI2} core available for reductive substrate binding. To
probe whether the synchronous replacement of H2 from 2− (or
D2 from 2-D−) occurs with substrates other than D2 (or H2,
respectively), reactions with phenylacetylene have been studꢀ
ied. These experiments also provided further evidence that
substrate binding indeed proceeds via formation of H2 (or D2)
without scrambling. Treatment of a THF solution of 2− or 2-D−
(either K∙2-D or Na∙2-D) with one equivalent of phenylacetyꢀ
lene gives in a clean reaction the product [LNi2(µꢀη1 η1
: ꢀ
CHCPh)]− (3−) (Scheme 5). When starting from K∙2, the prodꢀ
uct K∙3 could be isolated in crystalline form and analyzed by
Xꢀray diffraction (Figure 10).
H
Ph
N
N
N
N
N
D
N
N N
Ni
Ni
Ni
Ni
When K∙2-D was treated with phenylacetylene and the reacꢀ
tion monitored by 2H NMR spectroscopy, the formation of D2
was clearly detected (Figure 11). No evidence could be found
for the formation of HD, and no D incorporation into the styꢀ
reneꢀ1,2ꢀdiyl bridge was observed, neither at the vinylic CꢀH
group nor at the phenyl group. This indicates that alkenyl or
vinylidene intermediates are not involved, and it is in agreeꢀ
ment with intramolecular D2 formation prior to binding of the
incoming phenylacetylene substrate, akin to the H2 (or D2)
formation from the two adjacent terminal NiꢀH (NiꢀD) as
described above.
N
N
N
N
D
H
Ph
D2
3
2-D
Scheme 5. Preparation of K[LNi2(µ-η1 η1
: -CHCPh)].
ACS Paragon Plus Environment

Page 9 of 13
Journal of the American Chemical Society
valent metal species or thermodynamically unfavorable oneꢀ
1
2
3
4
5
6
7
8
9
electron reduced intermediates during the reductive activation
of small molecules. The present bimetallic system based on
the new pyrazolateꢀbridged bis(nacnac) ligand appears particuꢀ
larly well suited for exploiting this concept. We are currently
studying the activation and transformation of a range of rather
inert substrates using the new dinickel dihydride complex.
Figure 11. 2H NMR spectrum (77 MHz, 298 K) of Kꢁ2 in THF
(bottom) and after reaction with phenylacetylene, showing
formation of D2 from the two NiꢀD (top). Residual solvent
signals are marked (*).
EXPERIMENTAL SECTION
Materials and methods. Manipulations involving airꢀ and
moisture sensitive compounds were conducted under an atꢀ
mosphere of dried (phosphorous pentoxide on solid support
[Sicapent®, Merck]) dinitrogen or argon using standard
Schlenk techniques, or in a glovebox (O2 < 0.5 ppm and H2O
< 0.5 ppm) filled with dinitrogen atmosphere. THF, THFꢀd8,
nꢀhexane and pentane were dried over sodium in the presence
of benzophenone. Hydrogen gas was purchased from Messer
and deuterium gas from Linde. Chemicals used were either
present in the working group or were purchased from comꢀ
mercial sources, or their synthesis is described below or in the
supporting information. NMR samples for the H2/D2 exchange
reactions were prepared under an inert atmosphere.
M[LNi2(H)2] (M = Na, K) was dissolved in THFꢀd8 and transꢀ
ferred into a J. Young NMR tube. The solution was degassed
by means of pumpꢀfreezeꢀthaw procedure with liquid N2. The
process was repeated three times. Because the dihydride comꢀ
plexes are extremely sensitive to traces of moisture (ultimately
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SUMMARY AND CONCLUSIONS
In this contribution we present a new dinucleating ligand
scaffold comprising two nacnac compartments spanned by a
central pyrazolateꢀbridge. Dinickel complexes of this ligand
scaffold having bulky substituents at the peripheral nacnacꢀN
(DIPP substituents in the present case) create a bimetallic cleft
into which the remaining accessible coordination sites are
directed; these sites are available for binding exogenous ligꢀ
ands or substrates within the clamp of the two metal ions. This
way we have prepared a series of dinickel(II) dihydride comꢀ
plexes in which the two NiꢀH units are oriented such that the
hydrides are located in rather close proximity in the bimetallic
pocket. H2/D2 exchange was found to proceed via an unpreceꢀ
dented pairwise H2/D2 exchange that synchronously involves
both NiꢀH units of [LNiII (H)2]−, without H/D scrambling.
2
giving the hydroxidoꢀbridged dinickel(II) complex [LNiII (ꢀꢀ
DFT calculations provided a mechanistic picture of this proꢀ
cess and revealed that the exchange mechanism should not be
concerted, meaning that both NiꢀH bonds are broken and H2 is
formed in a late transition state scenario before the incoming
D2 forms new NiꢀD bonds. In a first step, H2 can form within
the cleft with a moderate, mostly enthalpic activation barrier
of around 62 kJ/mol and will then exit the complex, which
formally leaves two reducing equivalents on the {LNi2} core
for substrate binding, such as the bimetallic, homolytic splitꢀ
2
OH)]), gases HD, D2 and H2 were rigorously dried by storing
them over conc. H2SO4 for around 1 d; this method proved
superior over transfer through cooling traps (liquid N2) folꢀ
lowed by passage over phosphorus pentoxide (Sicapent).
1
2
Instrumentation. H NMR, H NMR and 13C NMR spectra
were recorded on Bruker Avance spectrometers (300 MHz,
1
400 MHz, 500 MHz for H) at room temperature unless othꢀ
ting of D2. The [LNiII (H)2]− complex (2−) can thus be viewed
erwise noted. Chemical shifts are reported in parts per million
relative to residual proton and carbon signals of the solvent
(CDCl3, δH = 7.26, δC = 77.16 ppm; THFꢀd8, δH = 1.73 and
2
as a masked form of a reactive, antiferromagnetically coupled
[LNiI2]− species. In the present work this concept, demonstratꢀ
ed for the H2/D2 exchange reaction, has been further exploited
for the reaction of [LNiII (H)2]− or [LNiII (D)2]− with phenylaꢀ
1
3.59; δC = 25.31 and 67.21). Hꢀ1H NOESY/EXSY spectra
were recorded at 400 MHz using a mixing time of 0.5 s. Elecꢀ
tron ionization (EI) and electrospray ionization (ESI) mass
spectra were collected on an Applied Biosystems API 2000
device or on a Bruker HCTultra instrument. Moisture or oxyꢀ
2
2
cetylene. Treating [LNiII (D)2]− with phenylacetylene leads to
2
D2 formation and twofold reduction of the substrate, giving a
product complex with unusual styreneꢀ1,2ꢀdiyl bridging unit in
the bimetallic pocket (3−). Alkali metal ions Na+ and K+ were
found to be closely associated with the dinickel dihydride core
in 2− and also with the olefinic C=C bond in the case of 3−
(shown for K+). While these interactions have only a minor
influence on the structural and spectroscopic properties of the
gen sensitive samples were prepared in
a glovebox
(MBRAUN UNIlab) under an N2 atmosphere and injected into
the Bruker HTCultra instrument via a direct Peek™ tubing
connection. Elemental analyses were carried out with an Eleꢀ
mentar 4.1 vario EL 3 instrument. IR spectra of solid samples
were measured with a Cary 630 FTIR spectrometer equipped
with a DialPath and Diamond ATR accessory (Agilent) placed
in a glovebox (MBRAUN UNIlab, argon atmosphere). IR
bands were labeled according to their relative intensities with
vs (very strong), s (strong), m (medium), w (weak) and very
weak (vw). The detection of molecular hydrogen was carried
out by gas chromatography using a GCꢀ2014 gas chroꢀ
matrographjy with ShimAdzu, Shincarbon column (4.0 m ×
2.00 mm, oven temperature 100°C, carrier gas Ar, 180 KPa).
Temperatureꢀdependent magnetic susceptibilities were measꢀ
ured using a QuantumꢀDesign MPMS XLꢀ5 SQUID magneꢀ
tometer equipped.
[LNiII (H)2]− core, the alkali metal ions significantly stabilizes
2
the dihydride complex against H2 release. This points to a
potentially important effect of Lewis acids on metalꢀhydride
reactivity, which will be further explored in future work.
Storing two reducing equivalents in adjacent metal hydrides
that evolve H2 upon substrate binding is reminiscent of the
proposed N2 binding step at the FeMo cofactor of nitrogenase.
However, recent electro catalytic and DFT studies have shown
that the rateꢀlimiting step for H2 formation in the FeMo protein
involves proton transfer from sulfur to a bridging iron hydride,
highlighting the important role of the sulfide ligands in FeMoꢀ
co33 Appreciating the possibility of different mechanistic traꢀ
jectories for reductive H2 release, using metal hydrides is
increasingly recognized as a means of avoiding unstable lowꢀ
Single-Crystal X-Ray structure determinations. Crystal
data and details of the data collections are given in Table S5.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 13
Xꢀray data were collected on a STOE IPDS II diffractometer
138.10(CHAr), 122.80 (CHAr), 101.20 (CHPz), 94.52 (CH),
40.06 (CH2), 28.08 (CHiPr), 23.76 (CH3iPr), 22.77 (CH3iPr),
21.62 (CH3), 19.17 (CH3). MS (ESI(+), MeCN) m/z: [M+H]+
609.36. IR: 3190(br)(NH), 3130 (NH), 3104 (NH), 3060 (w),
3020 (w), 2960(m), 2923 (w), 2867 (w), 1621 (vs), 1551 (vs),
1501 (w), 1454 (m), 1432 (m), 1377 (m), 1361 (m), 1292 (m),
1284 (m), 1268 (m), 1226 (m), 1179 (m), 1159 (m), 1090 (m),
1049 (m), 1020 (m), 1005 (w), 934 (w), 919 (w), 879 (w), 819
(w), 804 (w), 784 (s), 758 (s), 728 (s), 695 (m), 664 (w), 626
(w), 607 (w), 582 (w), 519 (w), 497 (w). Elemental analysis
(%) calc. for C39H56N6·(C4H8O) (681.00 g/mol) = C 72.84 H
9.47 N 12.34; Found C 72.92 H 9.20 N 12.48.
1
2
3
4
5
6
7
8
(graphite monochromated MoꢀKα radiation, λ = 0.71073 Å) by
use of scans at –140 °C. The structures were solved by
SHELXT,34 and refined on F2 using all reflections with
SHELXLꢀ2013/14/16.35 Nonꢀhydrogen atoms were refined
anisotropically. Most hydrogen atoms were placed in calculatꢀ
ed positions and assigned to an isotropic displacement parameꢀ
ter of 1.2/1.5 Ueq(C). The nickel bound hydrogen atoms in K∙2
and Na∙2 were refined freely. In case of 2[K(DB18C6)] a fixed
isotropic displacement parameter of 0.08 Å2 was applied. Two
9
i
disordered Pr groups in K∙2 (occupancy factors = 0.766(14)/
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.234(14) and 0.731(5)/0.269(5)) have been refined using
SIMU, DELU, ISOR, BUMP, SAME and SADI restraints and
EADP constraints. Disordered potassium coordinating solꢀ
vents (thf (only C atoms): occupancy factors = 0.616(9) /
0.384(9); thf/Et2O: occupancy factors = 0.712(7) / 0.288(7))
and nonꢀcoordinating Et2O (occupancy factors = 0.504(11) /
0.496(11)) in 2[K(DB18C6)] have been refined using SIMU,
DELU, ISOR, DFIX, SAME and SADI restraints and EADP
constraints. In case of the coordinating solvents thf and Et2O
occupy the same coordination site of the potassium atom. In
K∙3 one carbon atom of a coordinating thf was found to be
disordered about 2 positions (occupancy factors = 0.55(2) /
0.45(2)) and was refined using SADI restraints and EADP
constraints. Faceꢀindexed absorption corrections were perꢀ
formed numerically with the program XꢀRED.36
[LNi2(µ-Br)] (1). H3L (505 mg, 0.83 mmol, 1.0 equiv) was
dissolved in dry THF (4 mL) and the clear yellow solution
cooled to ꢀ78 °C. nꢀBuLi (2.6 M in toluene) was added slowly
via a syringe to give a clear orange solution. Under stirring the
reaction mixture was warmed to room temperature over 15
minutes to yield a dark red, clear solution. NiBr2(DME) (512
mg, 1.66 mmol, 2.0 equiv) was then slowly added. The resultꢀ
ing darkꢀbrown mixture was stirred overnight at room temperꢀ
ature. The suspension was separated by a centrifugal device
and the dark brown THF solution discarded. The solid material
was washed with acetone for several times until the solvent
remained colorless. After removing residual solvent under
reduced pressure, a beige colored fine powder was obtained
(506 mg, 0.63 mmol, 76 %). Slow evaporation of a concenꢀ
trated solution in CH2Cl2 or CHCl3 yielded small greenꢀbrown,
cubeꢀshaped crystals of 1. 1H NMR (300 MHz, CDCl3) = 6.91ꢀ
6.80 (m, 2 H, CHAr), 6.73ꢀ6.75 (d, JHꢀH = 6 Hz, 4 H,
CHAr),5.46 (s, 1 H, CHPz), 4.68 (s, 2 H, CH), 4.07 (s, 4 H,
CH2), 3.30ꢀ3.20 (sept, 4 H, CHiPr), 1.95 (s, 6 H, CH3), 1.39 (d,
JHꢀH = 6 Hz, 12 H, CH3iPr), 1.26 (s, 6 H, CH3), 0.95 (d, JHꢀH = 6
Hz, 12 H, CH3iPr). 1H NMR (300 MHz, THFꢀd8): = 6.79 ꢀ 6.95
(m, 6H, CHAr), 5.53 (s, 1 H, CHPz), 4.77(s, 2 H, CH), 4.15(s, 4
H, CH2), 3.33 ꢀ 3.40 (sept, 4 H, CHiPr), 2.01 (s, 6 H, CH3), 1.47
(d, JHꢀH = 6 Hz, 12 H, CH3iPr), 1.30 (s, 6 H, CH3), 1.02 (d, JHꢀH
= 6 Hz, 12 H, CH3iPr). 13C NMR (75 MHz, CDCl3) = 159.74
(CqꢀMe), 153.24 (CqꢀMe), 147.66 (CPz), 141.50 (CAr) 125.43
(CHAr), 123.30 (CHAr), 97.24 (CH), 91.51 (CHPz), 54.44 (CH2)
28.13 (CH3iPr), 24.84 (CH3iPr), 24.12 (CH3iPr), 23.37 (CH3),
21.53 (CH3). HRMS (ESI(+), MeOH/DCM) m/z: calc [M+H]+
DFT calculations. Geometry optimizations and frequency
calculations were carried out with the BP86 functional37 and
the def2ꢀSVP basis set,38 referred in the text as SVP. Disperꢀ
sion corrections were included as proposed by Grimme and
coworkers, with a BeckeꢀJohnson type damping.39 All results
were computed at this level of theroy unless otherwise noted.
The nature of all stationary points mentioned in the text was
confirmed through vibrational frequency analysis. The refineꢀ
ment of electronic energies for the approximate reaction pathꢀ
way of hydrogen abstraction was carried out with the M06L
functional40 and the def2ꢀTZVP orbital basis set, referred in
the text as TZVP. The default def2 Coulomb fitting basis were
used.41 All calculations were carried out with the Orca 4.0
program package.42
H3L. 2ꢀ[(2,6ꢀdiisopropylphenyl)imido]ꢀpentꢀ2ꢀenꢀ4ꢀone (16.4
g, 63 mmol, 2 equiv.) was dissolved in dry dichloromethane
(100 mL) to give a colorless solution. To this solution triꢀ
ethyloxonium tetrafluoroborate (12.7 g, 67 mmo, 2.1 equiv.),
dissolved in dry dichloromethane (20 mL) was added slowly
and the mixture stirred under inert atmosphere over night at
room temperature. Triethylamine (9.3 mL) was added to the
yellowish reaction mixture resulting in a color change to red
after 20 minutes. This solution was then slowly added to a
suspension of bis(3,5ꢀaminomethyl)pyrazole (4.0 g, 32 mmol,
1 equiv) in dry triethylamine. The mixture was stirred for 3
days under inert atmosphere at room temperature. After reꢀ
moval of the solvents under reduced pressure the orange resiꢀ
due was extracted with toluene. The solvent was removed and
the crude material purified by Kugelrohr distillation (10ꢀ5
mbar, 110 °C). After recrystallization from ethanol the ligand
was obtained as yellow solid material (9.7 g, 16 mmol, 43 %).
1H NMR (300 MHz, CDCl3) = 7.11ꢀ7.02 (m, 6 H, CHAr), 6.02
(s, 1 H, CHPz), 4.73 (s, 2 H, CH), 4.40 (s, 4 H, CH2), 2.90ꢀ2.85
(sept, 4 H, CHiPr), 1.95 (s, 6 H, CH3), 1.65 (s, 6 H, CH3), 1.15
(d, 12 H, CH3iPr), 1.06 (d, 12 H, CH3iPr). 13C NMR (75 MHz,
CDCl3) = 166.22 (3.5ꢀCpz), 155.58(CHAr), 146.28(CHAr),
803.2273; exp [M+H]+ 803.2258. IR (ATR)
ν
/ cmꢀ1 = 3058
(w), 2959 (m), 2923 (m), 2862 (m), 1555 (m), 1532 (vs), 1462
(vs), 1435 (s), 1399 (s), 1381 (s), 1369 (vs), 1313 (s), 1279 (s),
1252 (s), 1236 (m), 1186 (m), 1175 (s), 1093 (s), 1052 (s),
1032 (m), 1012 (m), 957 (m), 935 (m), 795 (vs), 759 (vs), 745
(vs), 542 (m). Elemental analysis (%) calc. for
C39H53N6Ni2Br·(CH2Cl2)1.5 (926.15 g/mol) =: C 52.48 H 6.10
N 9.07; Found C 52.77 H 6.45 N 9.33.
K[LNi2(H)2] (K∙2). [LNi2(µꢀBr)] (300 mg, 0.37 mmol, 1.0
equiv) was suspended in dry THF (4 mL). After slow addition
of a 1 M solution of KHBEt3 in THF (0.93 ml, 0.93 mmol, 2.5
equiv) to the beige suspension, the mixture was stirred at room
temperature under inert conditions for 30 minutes. The formed
dark brownish suspension was filtered over a filter medium to
yield a clear dark yellowꢀbrownish solutuion. A slow diffusion
of pentane vapors into the filtrate yielded the product as large,
cubeꢀ, rodꢀ or needle shaped crystals (191 mg, 0.25 mmol, 67
%) that were washed with pentane. 1H NMR (300 MHz, THFꢀ
d8) = 6.77ꢀ6.88 (m, 6 H, CHAr), 5.56 (s, 1 H, CHPz), 4.56 (s, 2
H, CH), 4.23 (s, 4 H, CH2), 3.48ꢀ3.43 (sept, 4 H, C CHiPr),
1.83 (s, 6 H, CH3), 1.22 (s, 6 H, CH3), 1.12 (d, JHꢀH = 6 Hz, 12
H, CH3iPr), 1.03 (d, JHꢀH = 6 Hz, 12 H, CH3iPr), ꢀ24.17 (s, 2 H,
ACS Paragon Plus Environment

Page 11 of 13
Journal of the American Chemical Society
3052 (w), 2982 (m), 2955(m), 2924 (m), 2863 (m), 1913 (Niꢀ
NiH). 13C NMR (100 MHz, THFꢀd8) = 158.64 (CqꢀMe),
157.57 (CqꢀMe), 157.13 (CqꢀMe), 156.60 (CqꢀMe), 140.34
(CPz), 123.91 (CAr), 123.18 (CAr), 97.03 (CH), 92.25 (CHPz),
53.34 (CH2), 28.33 (CH3iPr), 24.46 (CH3iPr), 23.89 (CH3iPr),
22.66 (CH3). MS (ESI(+), THF/MeCN) m/z: [M+H]+ 764.41.
H) (m), 1595 (m), 1503 (vs), 1452 (s), 1436 (s), 1426 (s), 1396
(s), 1366 (s), 1356 (m), 1320 (s), 1298 (w), 1270 (w), 1246
(vs), 1211 (vs), 1191 (w), 1118 (vs), 1094 (m), 1055 (s), 1020
(w), 943 (s), 902 (m), 848 (w), 796 (m), 778 (m), 756 (s), 741
(vs), 726 (vs), 715 (vs), 646 (w), 629 (w), 600 (m), 560 (w),
522 (w).
1
2
3
4
IR (ATR)
ν
/ cmꢀ1 = 3054 (w), 2955 (m), 2858 (m), 1961 (Niꢀ
5
H) (m), 1562 (s), 1518 (vs), 1461 (s), 1431 (vs), 1398 (vs),
1355 (m), 1315 (s), 1272 (s), 1254 (m), 1230 (m), 1196 (m),
1159 (m), 1097 (m), 1057 (w), 1028 (m), 935 (w), 861 (m),
804 (m), 773 (m), 731 (m), 720 (m), 644 (m), 546 (m), 516
(m), 458 (m), 422 (m). Elemental Analysis calculated for
C39H55KN6Ni2 (%): C 61.28, H 7.25, N 10.99. Found: C 60.73,
H 6.95, N 10.84.
6
7
8
9
(2[K(DB18C6)]-D). In a Young tube, a solution of
[LNi2(H)2)][K(Dibenzo(18ꢀcrownꢀ6))] (2[K(DB18C6)]) in
THF (0.5 mL) was freezeꢀthaw degassed under vacuum three
times. Then dry D2 (ca. 1 atm) was introduced into the head
space of the flask at rt.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
K[LNi2(µ-η1 η1-CHCPh)] (K∙3). K[LNi2(H)2] (70 mg, 92
:
K[LNi2(D)2] (K∙2-D). In a Young tube, a solution of
K[LNi2(H)2] in THF (0.5 mL) was freezeꢀthaw degassed under
vacuum three times. Then dry D2 (ca. 1 atm) was introduced to
the head space of the flask at room temperature.
µmol, 1.0 equiv) was dissolved in dry THF (3 mL) to give a
clear, dark yellow solution. To this solution phenyl acetylene
(12 µL, 110 µmol, 1.2 equiv) was added, resulting in a dark
green, clear solution which was cooled (ꢀ30 °C) and layered
with hexanes. Slow diffusion of the solvents at ꢀ30 °C yielded
the product as dark green, rod shaped crystals suitable for Xꢀ
ray diffraction (Yield: 54%). 1H NMR (500 MHz, THFꢀd8, ꢀ35
°C) = 10.14 (d, 3JHH = 7.8 Hz, 1 H, CHPh), 7.01ꢀ6.99 (dd, 3JHH
= 7.7 Hz, 4JHH = 1.2 Hz, 2 H, CHPh), 6.97ꢀ6.95 (dd, 3JHH = 7.7
K[LNi2(H)(D)] (K∙2-HD). In a Young tube, a solution of
K[LNi2(H)2] in THF (0.5 mL) was freezeꢀthaw degassed under
vacuum three times. Then dry HD (ca. 1 atm) was introduced
to the head space of the flask at room temperature.
Na[LNi2(H)2] (Na∙2). A solution of NaHBEt3 in THF (1.0 M)
(0.75 mL, 0.75 mmol, 3 equiv) was added dropwise to a
stirred brown solution of [LNi2(µꢀBr)] (200 mg, 0.25 mmol, 1
equiv) in THF (2 mL) at room temperature. After stirring the
resulting redꢀbrown solution for 2 hours and filter, all volatiles
were removed in vacuo. The redꢀbrown residue was washed
twice with hexane (40 mL). After concentration in vacuo the
product was obtained as a red powder (130 mg, 0.17 mmol, 68
%). Recrystallization from pentane diffusion into THF at room
temperature yielded orange block crystals of Na[LNi2(H)2]. 1H
NMR (400 MHz, THFꢀd8) = 6.75ꢀ6.86 (m, 6 H, CHAr), 5.61 (s,
1 H, CHPz), 4.58 (s, 2 H, CHCCH3), 4.27 (s, 4 H, CH2Pz), 3.41
ꢀ 3.49 (m, 4 H, (CH3)2CHPh), 1.86 (s, 6 H, CH3CCH), 1.23 (s,
6 H, CH3CCH), 1.04 (dd, 24 H, JHꢀH = 6 Hz, (CH3)2CHPh), ꢀ
23.54 (s, 2 H. NiꢀH). 13C NMR (100 MHz, THFꢀd8) = 157.91
(CqꢀMe), 156.99 (CqꢀMe), 155.16 (CHAr), 139.26 (CHAr),
123.15(CHAr), 122.21 (CHAr), 95.95 (CHPz), 91.91 (CH2),
51.35 (CH2), 27.25 (CH3iPr), 25.39 (CH3iPr), 23.51 (CH3iPr),
4
3
Hz, JHH = 1.2 Hz, 2 H, CHPh), 6.89ꢀ6.85 (dt, JHH = 7.7 Hz, 2
H CHPh), 6.76ꢀ6.72 (dd, 3JHH = 7.7 Hz, 2 H, CHAr), 6.62 (m, 2
H CHAr), 6.24 (d, JHH = 7.5 Hz, 1 H, CHPh), 5.63 (s, 1 H,
3
CHPz), 4.87 (s, 1 H, CHvinyl), 4.62 (s, 1 H, CH), 4.63 (s, 1 H,
2
CH), 4.33ꢀ3.91 (dd, JHH = 17.5 Hz, 2 H, CH2), 4.26ꢀ4.07 (dd,
2 H, CH2), 3.76ꢀ3.65 (sept, JHH = 7.1 Hz, 1 H, CHiPr), 3.58ꢀ
3
3.48 (sept, JHH = 7.1 Hz, 1 H, CHiPr), 3.29ꢀ3.19 (sept, JHH
=
3
3
7.0 Hz, 1 H, CHiPr), 1.92 (s, 3 H, CH3), 1.84 (s, 3 H, CH3),
1.79 (d, 3JHH = 6.5 Hz, 3 H, CH3iPr), 1.39ꢀ1.29 (sept, 3JHH = 7.1
Hz, 1 H, CHiPr), 1.18 (d, JHH = 6.8 Hz, 3 H, CH3iPr), 1.02 (d,
3
3JHH = 6.5 Hz, 3 H, CH3iPr), 0.93 (d, JHH = 6.8 Hz, 3 H,
3
CH3iPr), 0.91 (d, 3JHH = 6.8 Hz, 3 H, CH3iPr), 0.79 (d, 3JHH = 6.8
Hz, 3 H, CH3iPr), 0.51 (d, JHH = 7.0 Hz, 3 H, CH3iPr), 0.23 (d,
3
3JHH = 6.5 Hz, 3 H, CH3iPr). 13C{1H} NMR (126 MHz, THFꢀd8,
ꢀ35 °C):
δ
/ ppm = 160.95 (CqꢀMe), 159.19 (CqꢀMe), 158.46
(CHvinyl), 158.36 CqꢀMe), 157.40 (CqꢀMe), 155.15 (CPz),
155.14 (CPz), 152.19 (CAr), 150.56 (CAr), 150.17 (CAr), 144.43
(CAr), 143.34 (CAr), 143.14 (CAr), 142.47 (Cvinyl), 140.14 (CPh),
128.34 (CHPh), 127.18 (CHPh), 126.15 (CHPh), 125.89 (CHAr),
124.92 (CHAr), 124.38 (CHAr), 124.10 (CHAr), 123.81 (CHAr),
123.35 (CHPh), 121.63 (CHPh, CHAr), 98.55 (CH), 96.43 (CH),
92.73 (CHPz), 52.76 (CH2), 51.58 (CH2), 29.71 (CH3iPr), 28.50
(CHiPr), 28.14 (CHiPr), 27.96 (\(CHiPr), 26.76 (CHiPr), 26.59
(CH3iPr), 25.93 (CH3), 24.78 (CH3iPr), 24.76 (CH3iPr), 24.33
(CH3iPr), 24.27 (CH3iPr), 24.14 (CH3iPr), 23.72 (CH3), 23.03
(CH3iPr), 21.43 (CH3), 20.55 (CH3). MS (ESI(+), THF/MeCN)
22.94 (CH3iPr), 21.53 (CH3), 19.61 (CH3). IR (ATR)
ν
/ cmꢀ1 =
3052 (w), 2953 (m), 2962 (m), 1846 (NiꢀH) (m), 1554 (m),
1521 (s), 1511 (s), 1459 (vs), 1373 (vs), 1396 (s), 1313 (m),
1271 (m), 1251 (m), 1231 (m), 1189 (m), 1100 (m), 1049 (m),
933 (m), 891 (m), 796 (m), 756 (m), 725 (m), 716 (m), 644
(m), 575 (m), 544 (w). Elemental analysis (%) calc. for
C39H55N6NiNa (C4H8O)2 (892.50 g/mol) =: C 63.34 H 8.04 N
9.44; Found C 63.47 H 8.37 N 9.17.
m/z: [M+H]+ 823.42. IR (ATR)
ν
/ cmꢀ1 = 3278 (vw), 3119
Na[LNi2(D)2] (Na∙2-D). In a Young tube, a solution of
Na[LNi2(H)2] in THF (0.5 mL) was freezeꢀthaw degassed
under vacuum three times. Then dry D2 (ca. 1 atm) was then
introduced to the head space of the flask at room temperature.
(vw), 3051 (w), 2955 (m), 2923 (m), 2863 (m), 1553 (m),
1521 (s), 1503 (m), 1430 (s), 1397 (vs), 1314 (m), 1306 (m),
1271 (m), 1250 (m), 1229 (w), 1190 (w), 1175 (m), 1091 (m),
1054 (s), 1028 (m), 1017 (m), 892 (m), 799 (m), 752 (s), 726
(m), 704 (s), 645 (w), 624 (w), 595 (w), 554 (w), 545 (w), 523
(w), 447 (w), 417 (m).
[K(Dibenzo(18-crown-6))][LNi2(H)2)]
(2[K(DB18C6)]).
Dibenzo(18ꢀcrownꢀ6) (7.2 mg, 0.02 mmol, 1 equiv) was added
into a solution of KLNi2(H)2 (K·2) (15.2 mg, 0.02 mmol, 1
equiv) in THF (2 mL) at room temperature. After stirring the
resulting red solution for 2 hours, all volatiles were removed
in vacuo. The redꢀbrown residue was washed twice with hexꢀ
anes (10 mL). The crude powder was recrystallized from a
solvent mixture of hexanes and diethyl ether layered onto the
THF solution at –30°C, yielding orange diamond crystals of
ASSOCIATED CONTENT
Supporting Information.
Detailed synthetic procedures, crystallographic information,
NMR, and IR spectra; DFT calculations; The Supporting Inforꢀ
mation is available free of charge on the ACS Publications webꢀ
site at xxx.
the product 2[K(DB18C6)] (Yield: 90%). IR (ATR)
ν
/ cmꢀ1 =
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 13
[19] Bellows, S. M.; Arnet, N. A.; Gurubasavaraj, P. M.; Brennessel,
W. W.; Bill, E.; Cundari, T. R.; Holland, P. L. J. Am. Chem. Soc.
2016, 138, 12112–12123.
AUTHOR INFORMATION
Corresponding Author
1
2
3
4
5
6
7
8
9
[20] (a) Lee, Y.; Anderton, K. J.; Sloane, F. T.; Abboud, K. A.; Garꢀ
ciaꢀSerres, R.; Murray, L. J. J. Am. Chem. Soc. 2015, 137, 10610–
10617. (b) Ermert, D. M.; Murray, L. J. Dalton Trans. 2016, 45,
14499–14507.
*Email: franc.meyer@chemie.uniꢀgoettingen.de
*Email: rmata@gwdg.de
Notes
The authors declare no competing financial interest.
[21] Anderton, K. J.; Knight, B. J.; Rheingold, A. L.; Abboud, K. A.;
GarciaꢀSerres, R.; Murray, L. J. Chem. Sci. 2017, 8, 4123–4129.
ACKNOWLEDGMENT
[22] Dugan, T. R.; Bill, E.; MacLeod, K. C.; Brennessel, W. W.;
Holland, P. L. Inorg. Chem. 2014, 53, 2370–2380.
Support from the China Scholarship Council (PhD fellowship to
PengꢀCheng Duan) is gratefully acknowledged. We thank Dr.
Anne Schober (University of Göttingen) for the crystallographic
analysis of compound K∙2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[23] (a) Jessop, P. G.; Morris, R. H. Coord. Chem. Rev. 1992, 121,
155–284; (b) Heinekey, D. M.; Oldham, W. J. J. Chem. Rev. 1993, 93,
913–926; (c) SaboꢀEtienne, S.; Chaudret, B. Chem. Rev. 1998, 98,
2077–2091; (d) Maseras, F.; Lledos, A.; Clot, E.; Eisenstein, O.
Chem. Rev. 2000, 100, 601–636; (e) McGrady, G. S.; Guilera, G.
Chem. Soc. Rev. 2003, 32, 383–392; (f) Kubas, G. J. Chem. Rev.
2007, 107, 4152–4205; (g) Morris, R. H. Coord. Chem. Rev. 2008,
252, 2381–2394.
REFERENCES
[1] (a) Kaesz, H. D.; Saillant, R. B. Chem. Rev. 1972, 72, 231–281.
(b) Moore, D. S.; Robinson, S. D. Chem. Soc. Rev. 1983, 12, 415–
452.
[24] Tye, J. W.; Darensbourg, M. Y.; Hall, M. B. in Activation of
Small Molecules: Organometallic and Bioinorganic Perspectives;
Tolman, W. N., Ed.; WileyꢀVCH: Weinheim, 2006, pp 121–158.
[2] (a) Keim, W. Angew. Chem, Int. Ed. 2013, 52, 12492–12496. (b)
Robinson, S. J. C.; Heinekey, D. M. Chem. Commun. 2017, 53, 669–
676.
[25] Pfirrmann, S.; Limberg, C.; Herwig, C.; Knispel, C.; Braun, B.;
Bill, E.; Stösser, R. J. Am. Chem. Soc. 2010, 132, 13684–13691.
[3] (a) Weller, A. S., McIndoe, J. S. Eur. J. Inorg. Chem. 2007, 4411–
4423. (b) Kubas, G. J. J. Organomet. Chem. 2009, 694, 2648–2653.
(c) DuBois, D. L. Inorg. Chem. 2014, 53, 3935–3960. (d) Bullock, R.
M.; Helm, M. L. Acc. Chem. Res. 2015, 48, 2017−2026.
[26] (a) Ruan, C.; Rodgers, M. T. J. Am. Chem. Soc. 2004, 126,
14600ꢀ14610. (b) Abraham, S. A.; Jose, D.; Datta, A. Chemꢀ
PhysChem. 2012, 13, 695–698.
[4] (a) Lubitz, W.; Ogata, H.; Rüdiger, O.; Reijerse, E. Chem. Rev.
2014, 114, 4081–4148. (b) Schilter, D.; Camara, J. M.; Huynh, M. T.;
HammerꢀSchiffer, S.; Rauchfuss, T. B. Chem. Rev. 2016, 116, 8693–
8749.
[27] (a) Boro, B. J.; Duesler, E. N.; Goldberg, K. I.; Kemp, R. A.
Inorg. Chem. 2009, 48, 5081–5087. (b) Dong, Q.; Zhao, Y.; Su, Y.;
Su, J.ꢀH.; Wu, B.; Yang, X.ꢀJ. Inorg. Chem. 2012, 51, 13162–13170.
[28] (a) Tsay, C.; Peters, J. C. Chem. Sci. 2012, 3, 1313–1318; (b)
Connelly, S. J.; Zimmerman, A. C.; Kaminsky, W.; Heinekey, D. M.
Chem.ꢀEur. J. 2012, 18, 15932–15934.
[29] Ni0ꢀH2 adducts are more common; for recent examples see: (a)
W. H. Harman, T, –P. Lin, J. C. Peters, Angew. Chem. Int. Ed. 2014,
53, 1081–1086; (b) C. Lu, R. C. Cammarota, C. C. Lu. J. Am. Chem.
Soc. 2015, 137, 12486–12489.
[5] Hideaki, O.; Wolfgang, L.; Yoshiki, H. J. Biochem, 2016, 160,
251–258.
[6] Shaw, W. J.; Helm, M. L.; DuBois, D. L. Biochim Biophys Acta,
2013, 1827, 1123–1139.
[7] (a) Xu, T.; Chen, D.; Hu, X. Coord. Chem. Rev. 2015, 303, 32–41.
(b) Brazzolotto, D.; Gennari, M.; Queyriaux, N.; Simmons, T. R.;
Pecaut, J.; Demeshko, S.; Meyer, F.; Orio, M.; Artero, V.; Duboc, C.
Nat. Chem. 2016, 8, 1054–1060. (c) Ogo, S. Coord. Chem. Rev. 2017,
334, 43–53.
[30] (a) Collman, J. P.; Hutchinson, J. E.; Wagenknecht, P. S.; Lewis,
N. S.; Lopez, M. A.; Guilard, R. J. Am. Chem. Soc. 1990, 112, 8206–
8208; (b) Collman, J. P.; Wagenknecht, P. S.; Lewis, N. S. J. Am.
Chem. Soc. 1992, 114, 5665–5673.
[8] Eberhardt, N. A.; Guan, H. Chem. Rev. 2016, 116, 8373–8426.
[9] (a) Coric, I.; Holland, P. L. J. Am. Chem. Soc. 2016, 138, 7200–
7211. (a) Djurdjevic, I.; Einsle, O.; Decamps, L. Chem. Asian J. 2017,
12, 1447–1455.
[31] Collman, J. P.; Wagenknecht, P. S.; Hutchinson, J. E.; Lewis, N.
S.; Lopez, M. A.; Guilard, R.; L’Her, M.; BothnerꢀBy, A. A.; Mishra,
P. K. J. Am. Chem. Soc. 1992, 114, 5654–5664.
[10] Thorneley, R. N. F.; Lowe, D. J. in Molybdenum Enzymes;
Spiro, T. G., Ed.; Wiley: New York, 1985; pp 221−284.
[32] Deacon, G. B.; Felder, P. W. J. Am. Chem. Soc. 1968, 90, 495–
497.
[11] (a) Hoffman, B. M.; Lukoyanov, D.; Yang, Z. Y.; Dean, D. R.;
Seefeldt, L. C. Chem. Rev. 2014, 114, 4041–4062. (b) Lukoyanov, D.;
Khadka, N.; Yang, Z. Y.; Dean, D. R.; Seefeldt, L. C.; Hoffman, B.
M. J. Am. Chem. Soc. 2016, 138, 10674–10683.
[33] Khadka, N.; Milton, R. D.; Shaw, S.; Lukoyanov, D.; Dean, D.
R.; Minteer, S. D.; Raugei, S.; Hoffman, B. M.; Seefeldt, L. J. Am.
Chem. Soc. 2017, 139, 13518–13524.
[34] Sheldrick, G. M. Acta Cryst. 2015, A71, 3–8.
[35] Sheldrick, G. M. Acta Cryst. 2015, C71, 3–8.
[36] XꢀRED; STOE & CIE GmbH, Darmstadt, Germany, 2002.
[12] Lukoyanov, D.; Khadka, N.; Dean, D. R.; Raugei, S.; Seefeldt, L.
C.; Ho man, B. M. Inorg. Chem. 2017, 56, 2233–2240.
[13] Ballmann, J.; Munhá, R. F.; Fryzuk, M. D. Chem. Commun.
2010, 46, 1013–1025.
[37] (a) Perdew, J. P. Phys. Rev. B: Condens. Matter Mater. Phys.
1986, 33, 8822–8824. (b) Becke, A. D. Phys. Rev. A: At., Mol., Opt.
Phys. 1988, 38, 3098–3100.
[14] Smith, J. M.; Lachicotte, R. J.; Holland, P. L. J. Am. Chem. Soc.
2003, 125, 15752–15753.
[38] Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
3297–3305.
[15] Pfirrmann, S.; Limberg, C.; Ziemer, B. Dalton Trans. 2008,
6689–6691.
[39] (a) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys.
2010, 132, 154104. (b) Grimme, S.; Ehrlich, S.; Goerigk, L. J.
Comput. Chem. 2011, 32, 1456–1465.
[16] Yu, Y.; Sadique, A. R.; Smith, J. M.; Dugan, T. R.; Cowley, R.
E; Brennessel, W. W; Flaschenriem, C. J; Bill, E.; Cundari, T. R.;
Holland, P. L. J. Am. Chem. Soc. 2008, 130, 6624–6638.
[40] Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
[41] Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057–1065.
[42] Neese, F. WIREs Comput. Mol. Sci. 2012, 2, 73–78.
[17] Pfirrmann, S.; Yao, S.; Ziemer, B.; Stosser, R.; Driess, M.;
Limberg, C. Organometallics. 2009, 28, 6855–6860.
[18] Gehring, H.; Metzinger, R.; Herwig, C.; Intemann, J.; Harder, S.;
Limberg, C. Chem.ꢀEur. J. 2013, 19, 1629–1636.
ACS Paragon Plus Environment

Page 13 of 13
Journal of the American Chemical Society
1
2
3
4
Table of Contents artwork
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment