High catalytic rates for hydrogen production using nickel electrocatalysts with seven-membered cyclic diphosphine ligands containing one pendant amine

Article abstract of DOI:10.1021/ja400181a

A series of Ni-based electrocatalysts, [Ni(7P^Ph ₂N^C6H4X)₂](BF₄)₂, featuring seven-membered cyclic diphosphine ligands incorporating a single amine base, 1-para-X-phenyl-3,6-triphenyl-1-aza-3,6-diphosphacycloheptane (7P ^Ph₂N^C6H4X, where X = OMe, Me, Br, Cl, or CF₃), have been synthesized and characterized. X-ray diffraction studies have established that the [Ni(7P^Ph₂N ^C6H4X)₂]²⁺ complexes have a square planar geometry, with bonds to four phosphorus atoms of the two bidentate diphosphine ligands. Each of the complexes is an efficient electrocatalyst for hydrogen production at the potential of the Ni(II/I) couple, with turnover frequencies ranging from 2400 to 27 000 s^-1 with [(DMF)H]⁺ in acetonitrile. Addition of water (up to 1.0 M) accelerates the catalysis, giving turnover frequencies ranging from 4100 to 96 000 s^-1. Computational studies carried out on the [Ni(7P^Ph₂N^C6H4X) ₂]²⁺ family indicate the catalytic rates reach a maximum when the electron-donating character of X results in the pK_a of the Ni(I) protonated pendant amine matching that of the acid used for proton delivery. Additionally, the fast catalytic rates for hydrogen production by the [Ni(7P^Ph₂N^C6H4X)₂]²⁺ family relative to the analogous [Ni(P^Ph₂N ^C6H4X₂)₂]²⁺ family are attributed to preferred formation of endo protonated isomers with respect to the metal center in the former, which is essential to attain suitable proximity to the reduced metal center to generate H₂. The results of this work highlight the importance of precise pK_a matching with the acid for proton delivery to obtain optimal rates of catalysis.

Full text of DOI:10.1021/ja400181a

Subscriber access provided by BOSTON UNIV
Article
High Catalytic Rates for Hydrogen Production Using Nickel Electrocatalysts
with Seven-Membered Diphosphine Ligands Containing One Pendent Amine
Michael P Stewart, Ming-Hsun Ho, Stefan Wiese, Mary Lou Lindstrom,
Colleen E Thogerson, Simone Raugei, R. Morris Bullock, and Monte L Helm
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja400181a • Publication Date (Web): 05 Feb 2013
Downloaded from http://pubs.acs.org on February 18, 2013
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 41
Journal of the American Chemical Society
1
2
3
4
5
6
High Catalytic Rates for Hydrogen Production Using Nickel Electrocatalysts
with Seven-Membered Diphosphine Ligands Containing One Pendent Amine
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
Michael P. Stewart, MingꢀHsun Ho, Stefan Wiese, Mary Lou Lindstrom, Colleen E. Thogerson,
Simone Raugei, R. Morris Bullock, and Monte L. Helm*
Center for Molecular Electrocatalysis, Chemical and Materials Sciences Division, Pacific
Northwest National Laboratory, P.O. Box 999, K2ꢀ57, Richland, Washington 99352
*Eꢀmail: monte.helm@pnnl.gov
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 41
1
2
Abstract: A series of Niꢀbased electrocatalysts, [Ni(7PPh2NC6H4X)2](BF4)2, featuring sevenꢀ
membered cyclic diphosphine ligands incorporating a single amine base, 1ꢀparaꢀXꢀphenylꢀ3,6ꢀ
triphenylꢀ1ꢀazaꢀ3,6ꢀdiphosphacycloheptane (7PPh2NC6H4X where X = OMe, Me, Br, Cl or CF3),
have been synthesized and characterized. Xꢀray diffraction studies have established that the
[Ni(7PPh2NC6H4X)2]2+ complexes have a square planar geometry, with bonds to four phosphorus
atoms of the two bidentate diphosphine ligands. Each of the complexes is an efficient
electrocatalyst for hydrogen production at the potential of the Ni(II/I) couple, with turnover
frequencies ranging from 2,400 to 27,000 sꢀ1 with [(DMF)H]+ in acetonitrile. Addition of water
(up to 1.0 M) accelerates the catalysis, giving turnover frequencies ranging from 4,100 – 96,000
sꢀ1. Computational studies carried out on the [Ni(7PPh2NC6H4X)2]2+ family indicate the catalytic
rates reach a maximum when the electronꢀdonating character of X results in the pKa of the
pendent amine matching that of the acid used for proton delivery. Additionally, the fast catalytic
rates for hydrogen production by the [Ni(7PPh2NC6H4X)2]2+ family relative to the analogous
3
4
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
[Ni(PPh NC6H4X2)2]2+ family are attributed to preferred formation of endo protonated isomers with
2
respect to the metal center in the former, which is essential to attain suitable proximity to the
reduced metal center to generate H2. The results of this work highlight the importance of precise
pKa matching with the acid for proton delivery to obtain optimal rates of catalysis.
2
ACS Paragon Plus Environment

Page 3 of 41
Journal of the American Chemical Society
1
2
3
4
Introduction
5
6
7
8
The expanded use of energy from intermittent renewable energy sources such as solar
and wind will require the ability to efficiently convert electricity to chemical energy for storage
as fuels. Efficient catalysts are necessary for both the production and utilization of hydrogen.
Platinum and the hydrogenase enzymes are excellent catalysts for both production and oxidation
of hydrogen. Platinum, however, is a precious metal with high cost and low abundance.1
Hydrogenase enzymes, while remarkably efficient catalysts in nature,2ꢀ4 are expensive to obtain
in large amounts and difficult to adapt to large scale commercial applications. These
considerations have spawned efforts to design molecular catalysts that employ more abundant
metals, such as nickel,5 cobalt,6ꢀ14 iron,15ꢀ28 or molybdenum29ꢀ31 as electrocatalysts for the
production and oxidation of hydrogen.32,33
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
Efforts in our laboratory have focused on developing first row metal complexes (Ni,5,34,35
Co,10,14 Fe36 and Mn37) that contain an amine base in the second coordination sphere, adjacent to
a vacant coordination site or a hydride ligand on the metal center. These mononuclear
complexes, in which the base facilitates the heterolytic cleavage/formation of the HꢀH bond and
functions as a proton relay,38 mimic an important structural feature proposed for the active site of
[FeFe]ꢀhydrogenase enzymes, the presence of an azadithiolate ligand in the dinuclear active site,
shown in structure 1.3,4,23 Nickel and cobalt diphosphine complexes that incorporate a positioned
base in
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 41
1
2
3
4
1
2
3
5
6
7
8
the ligand have been found to be active electrocatalysts for hydrogen production in acidic
acetonitrile solutions, and the Fe complexes with similar ligands exhibit evidence for heterolytic
cleavage and catalytic oxidation of H2.5,35,36
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
Using a series of [Ni(PR NR´2)2]2+ catalysts, where P2N2 stands for a 1,3ꢀR´ꢀ3,7ꢀR
2
derivative of 1,5ꢀdiazaꢀ3,7ꢀdiphosphacyclooctane, we have previously examined how the
basicity of the proton relay in cooperation with the size and electronic characteristics of the
substituent at the phosphorus atoms influence the turnover frequencies for H2 production.39ꢀ42
Under optimized conditions, [Ni(PPh2NC6H4Br2)2]2+ (structure 2, X = Br), containing electronꢀ
withdrawing bromo substituents on the aniline rings of the ligand and phenyl groups on the
phosphorus atoms, catalyzed the formation of H2 with turnover frequencies as high as 1040 sꢀ1
and an overpotential of approximately 290 mV.39
The positioned amines in the second coordination sphere play important roles in the
catalytic activity of these complexes. Our experimental and theoretical mechanistic studies of
[Ni(PR NR’ ) ]2+ catalysts suggest that the transition state for heterolytic HꢀH bond formation and
2
2 2
cleavage has structure 3.38,43ꢀ45 Further studies on reduction and protonation of closely related
analogs of 2 have shown that intermediates formed during the catalytic cycle potentially include
isomeric Ni(0) complexes in which an amine in each ligand is protonated.34,46ꢀ49 The kinetically
formed doubly protonated Ni(0) species has both protons in exo positions relative to the nickel
center, “pinched” between the two pendent amines with NꢀH • • • N bonding. Isomerization to
form endo protonated species that would be active in the catalytic cycle has been shown to
involve intermolecular deprotonation by a base followed by reprotonation at an endo position.48
These studies led us to propose that only a fraction of the doubly protonated Ni(0) intermediates
formed during hydrogen production are in the catalytically active form and that higher catalytic
4
ACS Paragon Plus Environment

Page 5 of 41
Journal of the American Chemical Society
1
2
rates might be achieved if the stabilized exoꢀprotonated structures could be avoided.39,47,48 We
now present a study of related nickel bis(diphosphine) catalysts that contain ligands that are
structurally modified, but which still retain features that allow access to similar transition state
structures.50 In this paper we report turnover frequencies up to 96,000 sꢀ1 for electrocatalytic
hydrogen production with overpotentials ranging from 550 ꢀ 640 mV using a modified nickel
catalyst containing two 7ꢀmembered cyclic diphosphine ligands 7PPh2NC6H4X (where 7PPh2NC6H4X
is 1ꢀparaꢀXꢀphenylꢀ3,6ꢀtriphenylꢀ1ꢀazaꢀ3,6ꢀdiphosphacycloheptane and X = OMe, Me, Br, Cl or
CF3). Factors that may contribute to the rates are discussed, including the effects of changing
from 8ꢀmembered to 7ꢀmembered cyclic ligands and the impact of reducing the number of amine
bases in the ligand backbones.
3
4
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
Results
Synthesis and Characterization of Ligands and Complexes. The synthesis of 1ꢀparaꢀ
Xꢀphenylꢀ3,6ꢀtriphenylꢀ1ꢀazaꢀ3,6ꢀdiphosphacycloheptane ligands (7PPh2NC6H4X, where X = OMe
(4-OMe), Me (4-Me), Br (4-Br), Cl (4-Cl) and CF3 (4-CF3)) was accomplished through the
addition of one equivalent of a pꢀsubstituted aniline to 1,2ꢀbis(hydroxymethylphenylꢀ
phosphino)ethane in CH3CN and refluxing for 12 hours, as previously reported for 4-H (X =
H)51 and shown in Equation 1. The cyclic ligands containing a 7ꢀmembered ring are formed as
(1)
X = OMe
X = Me
X = H
X = Br
X = Cl
X = CF3
4-OMe
4-Me
4-H
4-Br
4-Cl
4-CF3
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 41
1
2
approximately 1:1 mixtures of racemic and meso isomers as indicated by 31P{1H} NMR spectra
of the crude reaction mixtures. Recrystallization results in pure ligand, though as a mixture of
3
4
5
6
7
8
9
31
1
the mesoꢀ and racꢀisomers, which show one singlet each in their P{1H} NMR spectra. H and
31P{1H} NMR spectral data for each ligand are provided in the Experimental section, and these
spectra are consistent with the formulations of these ligands.
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 synthesis of the nickel complex of 4-H was reported previously, and a similar route
is followed here.50 The addition of two equivalents of 4-OMe, 4-Me, 4-Br, 4-Cl or 4-CF3 to an
CH3CN solution of [Ni(CH3CN)6](BF4)2 results in formation of [Ni(7PPh2NC6H4X)2](BF4)2 (where
X = OMe, Me, Br, Cl or CF3, 5-OMe, 5-Me, 5-Br, 5-Cl or 5-CF3), as determined by P{1H}
31
NMR spectroscopy and shown in Equation 2. In addition to the homoleptic, monoꢀ
(2)
X = OMe
X = Me
X = H
X = Br
X = Cl
X = CF3
5-OMe
5-Me
5-H
5-Br
5-Cl
5-CF3
metallic complexes, the formation of oligomeric metal complexes is suggested by broad
resonances in the P{1H} NMR spectra. Since a mixture of mesoꢀ and racꢀ isomers of 4 was
31
used, the racꢀisomers may function as bridging ligands between Ni(II) centers to form
oligomeric byproducts. The desired monometallic complexes are isolated as orange crystalline
products in yields ranging from 26ꢀ54% by concentration of the reaction mixture and
crystallization by addition of CH2Cl2 followed by Et2O. Each of the products has been fully
6
ACS Paragon Plus Environment

Page 7 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
characterized by NMR spectroscopy, elemental analyses, and single crystal Xꢀray
crystallography.
7
8
9
Xꢀray quality crystals of 5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 were grown by slow
diffusion of Et2O into saturated CH3CN or CH2Cl2 solutions of the complexes. The Ni(II)
complexes 5-OMe, 5-Me, 5-Br and 5-Cl complexes crystallize with two nonꢀcoordinating
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
ꢀ
CH3CN molecules per unit cell, and no close contacts are observed between the BF4 anions and
any of the [Ni(7PPh2NC6H4X)2]2+ cations. Complex 5-CF3 crystallizes with no solvents, and shows
ꢀ
the BF4 ions interacting with the metal center resulting in a pseudoꢀoctahedral structure around
the Ni center, with Ni • • • F distances of 3.02 Å. A drawing of each cation is shown in Figure 1.
The structures confirm that the ligands coordinate as diphosphine chelates to give a distorted
square planar geometry with all four NiꢀP bond distances nearly equal, ranging from 2.20ꢀ2.23
Å. The PꢀNiꢀP bond angle for each of the diphosphine ligands span a narrow range from 79.7ꢀ
80.1°. This angle is smaller than the 82ꢀ84° typically observed for the PꢀNiꢀP bonds in
[Ni(PR NR’ ) ]2+ complexes,39,40 which form two sixꢀmembered chelate rings upon binding to the
2
2 2
metal. In contrast, for the [Ni(7PPh2NC6H4X)2]2+ complexes one fiveꢀmembered and one sixꢀ
membered ring formed upon chelation of the ligand. The smaller bite angle of 7PPh2NC6H4X
compared to the PPh2NR’ ligands results in a decrease in the steric interactions between the
2
phenyl substituents on adjacent phosphorus atoms of the two ligands39 and a more planar
structure than observed in the corresponding [Ni(PPh2NR’ ) ]2+ cations. For example, in
2 2
[Ni(PPh NC6H4Me2)2]2+ the dihedral angle between the two planes defined by the Ni atom and the
2
two phosphorus atoms of each diphosphine ligand is 24.16°,39 whereas it is 0.0° for
[Ni(7PPh2NC6H4Me)2]2+, as in all of the [Ni(7PPh2NC6H4X)2]2+ structures. Thus the replacement of
one sixꢀmembered chelate with a fiveꢀmembered ring has significant structural consequences for
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 41
1
2
3
4
these
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
(B)
(A)
(D)
(C)
(E)
Figure 1. Xꢀray crystal structure of (A) [Ni(7PPh2NC6H4OMe)2](BF4)2 ꢁ 2CH3CN, 5-OMe (B)
[Ni(7PPh2NC6H4Me)2](BF4)2 ꢁ 2CH3CN, 5-Me (C) [Ni(7PPh2NC6H4Br)2](BF4)2 ꢁ 2CH3CN 5-Br, (D)
[Ni(7PPh2NC6H4Cl)2](BF4)2 ꢁ 2CH3CN, 5-Cl and (E) [Ni(7PPh2NC6H4CF3)2](BF4)2 ꢁ 2CH3CN, 5-CF3.
ꢀ
The BF4 counterions, CH3CN solvent molecules and H atoms have been omitted for clarity.
8
ACS Paragon Plus Environment

Page 9 of 41
Journal of the American Chemical Society
1
2
3
4
Thermal ellipsoids are all shown at the 50% probability level.
5
6
7
8
9
complexes. As shown in Figure 1, the two sixꢀmembered rings containing the pendent amines
are trans to each other across the metal center, and, except for 5-CF3, adopt boat conformations.
The chair conformations of the sixꢀmembered rings in observed in 5-CF3 likely result from the
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
ꢀ
interaction of the BF4 ions with the metal center in the solid state. The nonꢀbonding Ni•••N
distances are 3.17ꢀ3.23 Å for 5-OMe, 5-Me, 5-Br and 5-Cl, where the six membered ring is in
the boat conformation, and 3.68 Å for 5-CF3 where the six membered ring is in the chair
conformation. Complete crystallographic information along with bond distances and angles for
5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 are contained in the Supporting Information.
The 31P{1H} NMR spectra of 5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 in CD2Cl2 consist of
a single peak ranging from 44.6 ppm for 5-OMe to 41.8 ppm for 5-CF3. The 1H NMR spectra of
5-OMe, 5-Me, 5-Br, 5-Cl, and CF3 in CD2Cl2 also show the expected number and intensity of
peaks consistent with the fourꢀcoordinate Xꢀray structures. The 31P{1H} NMR spectra of 5-
OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 in acetonitrileꢀd3, however, consists of two peaks, a broad
singlet (ranging from 51.8 – 48.8 ppm) and a sharp singlet (ranging from 47.2 – 45.2 ppm). In
acetonitrile it is believed that the compounds are fiveꢀcoordinate species with an acetonitrile
bound to the nickel center, as previously reported for 5-H.50 As shown in Figure 2, this results in
CH3
C
2+
2+
CH3
C
2+
XC6H4
CH3
C
C6H4X
C6H4X
N
N
N
Ph
Ph
N
N
Ph
Ph
N
P
P
Ph
Ph
P
Ni
P
Ni
P
P
Ni
P
P
Ph
P
P
Ph
Ph
P
P
Ph
Ph
N
Ph
N
N
C6H4X
C6H4X
C6H4X
A
B
C
Figure 2. Proposed isomers of 5 in CH3CN.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 41
1
2
the possibility of three distinct isomers. Low temperature 31P{1H} and H NMR data in
acetonitrileꢀd3 indicate isomers A and C are rapidly interconverting through the
dissociation/association of acetonitrile, resulting in the broad downfield singlet observed at room
temperature. Isomer B has two different ligand environments; however, dissociation/association
of the acetonitrile results in a chemically equivalent complex and therefore the sharper upfield
chemical shift.
3
4
5
6
7
8
9
1
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
Electrochemical Studies. The cyclic voltammograms of 5-OMe, 5-Me, 5-Br, 5-Cl and
5-CF3 indicate two overlapping oneꢀelectron reversible redox couples, with observed E1/2 values
(reported as the average of the potentials of the maximum cathodic and anodic current) ranging
from –1.05 to –1.14 V vs. the ferrocenium/ferrocene (Cp2Fe+/0) couple, as summarized in Table 1
(e.g. 5-OMe, Figure 3). A plot of the peak current (ip) vs. the square root of the scan rate
Figure 3. Cyclic voltammogram of 1.0 mM [Ni(7PPh2NC6H4OMe)2](BF4)2, 5-
OMe in 0.10 M [Bu4N][PF6]/CH3CN. Conditions: 1 mm glassy carbon
working electrode; scan rate 0.1 V/s at 25 °C.
shows a linear correlation in all cases, implying diffusionꢀcontrolled electrochemical events.52
The difference in the potential of cathodic and anodic peak potentials (ꢀEp) at a scan rate of 0.1
V/s for these processes is measured to be 120ꢀ82 mV (Table 1), where the ꢀEp of Cp2Fe+/0 ranges
from 68ꢀ71 mV. The observation that the ꢀEp values for the [Ni(7PPh2NC6H4X)2]2+ family are
10
ACS Paragon Plus Environment

Page 11 of 41
Journal of the American Chemical Society
1
2
neither 70 mV, as observed for the oneꢀelectron Cp2Fe+/0 couple, nor close to 30 mV, as
3
4
5
6
predicted for a twoꢀelectron process, suggest that 5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 undergo
7
8
9
two,
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
Table 1. Selected Electrochemical Data for of [Ni(7PPh NC6H4X)2](BF4)2 complexes in 0.10 M
2
[Bu4N][PF6]/CH3CN.
Overlapping Ni(II/I , I/0)
c
Complex
ꢂEp (mV)
Observed E1/2a,b (V)
[Ni(7PPh2NC6H4OMe)2](BF4)2, 5-
OMe
ꢀ1.14
120
[Ni(7PPh2NC6H4Me)2](BF4)2, 5-Me
[Ni(7PPh2NPh)2](BF4)2, 5-H
[Ni(7PPh2NC6H4Br)2](BF4)2, 5-Br
[Ni(7PPh2NC6H4Cl)2](BF4)2, 5-Cl
[Ni(7PPh2NC6H4CF3)2](BF4)2, 5-CF3
ꢀ1.13
ꢀ1.12
ꢀ1.08
ꢀ1.08
ꢀ1.05
120
105
104
120
82
aAll potentials are referenced to the Cp2Fe+/0 couple at 0 V. bCalculated as the average of
the potentials of the maximum cathodic and anodic current. cPeakꢀtoꢀpeak separation of
cathodic and anodic waves at a scan rate of 0.1 V/s.
overlapping, oneꢀelectron processes as previously observed and simulated for 5-H (the
determination of ip for the catalytic rate calculations is discussed below).50 In the analogous
[Ni(PPh NC6H4X2)2]2+ systems (e.g. 2) an increase in basicity of the pendent amine results in a
2
negative shift of the redox potential of the complex. The same trend is observed here, with 5-
OMe having the most negative redox potential at ꢀ1.14 V and 5-CF3 has the most positive redox
potential at ꢀ1.05 V.39 The observed trend in redox potentials is also in accordance with that
predicted by the “electrochemical ligand parameter” (EL) series developed by Lever which
catalogs common substituents by their electron donating/withdrawing ability.53,54
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 41
1
2
3
4
5
6
7
8
9
Electrocatalytic Production of Hydrogen. When electrochemical measurements on 5-
OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 are carried out in the presence of acid, a large increase in the
cathodic peak current is observed (e.g. 5-Br, Figure 4). Protonated dimethylformamide triflate,
[(DMF)H]OTf (pKa = 6.1 in MeCN), a crystalline solid,55 was used as the acid.56,57 Figure 5
shows that a plot of the catalytic current (icat) vs. [(DMF)H+]½ gives a linear correlation,
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 4. Cyclic voltammograms of [Ni(7PPh2NC6H4Br)2](BF4)2, 5-Br in 0.10 M
[Bu4N][PF6]/CH3CN with subsequent additions of [(DMF)H]+ and H2O.
Conditions: 1 mm glassy carbon working electrode; 25 ˚C; scan rate 10 V/s.
12
ACS Paragon Plus Environment

Page 13 of 41
Journal of the American Chemical Society
1
2
Figure 5. Plot of icat vs. [[(DMF)H]+]½ for [Ni(7PPh2NC6H4Br)2](BF4)2, 5-Br,
measured in 0.10 M [Bu4N][PF6]/CH3CN with increasing concentrations of
[(DMF)H]+. Conditions: 1 mm glassy carbon working electrode; 25 ˚C; scan
rate 10 V/s.
3
4
5
6
7
8
9
indicating the reaction is firstꢀorder with respect to acid concentration (i.e. Equation 3 where n is
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 number of electrons involved in the catalytic reaction, F is Faraday’s constant, A is the area
of the electrode, D is the diffusion coefficient, k is the rate constant, and x is the order of the
reaction with respect to acid).58ꢀ61
ꢁ
ꢅꢂx
H ꢆ
(3)
ꢃ
D
icat
ꢀ
nFAꢁcatꢂ
ꢄ
k
The observed catalytic current (icat) in cyclic voltammograms recorded at slower scan
rates (ca. < 0.5 V/s) deviated from the plateau shape expected for catalytic waves by displaying a
moderate scan rate dependence below scan rates of 1 V/s, as well as peakꢀshaped diffusionꢀ
controlled wave forms (e.g. 5-OMe, Figure 6), as previously described for 5-H.50 These wave
Figure 6. Cyclic voltammograms of 1.0 mM
[Ni(7PPh2NC6H4OMe)2](BF4)2, 5-OMe in 0.10 M
[Bu4N][PF6]/CH3CN with 0.21 M [(DMF)H]+ at
scan rates of 1, 5 and 10 V/s. Conditions: 1 mm
glassy carbon working electrode at 25 °C.
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 41
1
2
3
4
5
6
7
8
9
shapes have been attributed to significant depletion of acidic substrate at the electrode as
expected for very fast catalysts. For each catalyst, detailed studies were conducted to determine
the minimum scan rate threshold above which the observed behavior is that expected for a
catalytic wave with no significant substrate depletion, as discussed in the Supporting
Information.
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
Cyclic voltammetry studies were carried out on 5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 as a
function of acid concentration with sequential additions of [(DMF)H]+ in CH3CN at scan rates
above those that showed icat independence (i.e. Fig. 6, 5 and 10 V/s scan rates). Due to catalyst
decomposition at high acid concentrations, addition was halted once an acid concentration
greater than 0.40 M was attained, despite an apparent continuation of catalytic current
enhancement. The length of these experiments was restricted to ca. 30 min, a period for which
minimal (< 5 %) catalyst decomposition was observed under these acidic conditions, as
determined by UVꢀvis spectroscopy for each complex (see Supporting Information). The icat for
each addition of acid was measured at the potential where the current first reaches its plateau (see
Supporting Information), and the ratio of icat/ip is used in Equation 4 (υ
= scan rate in V/s).52,60ꢀ62
Assuming two electrons are passed for each H2 molecule produced (n = 2), and the acid
concentration does not change significantly during the course of the measurement (as indicated
by a current plateau), the catalytic rate constant (kobs = k[H+]x), or turnover frequency, can be
calculated using the simplified Equation 5 (where T = 298 K). Using a constant acid
concentration and measuring the icat as a function of catalyst concentration, the data also show a
firstꢀorder dependence on catalyst concentration (see Supporting Information).
14
ACS Paragon Plus Environment

Page 15 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
icat
n
RTꢈk[H+]xꢉ
ꢇ
(4)
=
ip 0.4463
Fυ
7
8
9
ꢘ
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
ꢒꢓꢔꢕ
ꢊꢋꢌꢍ ꢀ 1.94 Vꢎꢏ ∙ ꢐ ꢑ
ꢗ
(5)
ꢒꢖ
Digital simulation of cyclic voltammograms of each Ni complex, in the absence of acid,
allowed for the estimation of oneꢀelectron corrected ip values, which differ from experimental ip
values featuring two electrons by a factor largely dependent upon peak overlap (i.e., ꢀEp,
discussed further in the Supporting Information). While accounting for dilution, conservative ip
values were used to determine icat/ip ratios in order to calculate kobs for each acid addition. With
the exception of 5-OMe, plots of kobs vs. the acid concentration show firstꢀorder dependence (e.g.
5-Br, Figure 7A) on acid concentration up to either 0.43 M [(DMF)H]+ or until a 30 minute
experiment time, at which point decomposition of the catalyst becomes significant. For 5-OMe
the
(A)
(B)
Figure 7. Plot of kobs vs. [(DMF)H]+ for [Ni(7PPh2NC6H4X)2](BF4)2, (A) 5-Br and (B) 5-OMe,
measured in 0.10 M [Bu4N][PF6]/CH3CN with increasing concentrations of [(DMF)H]+.
Conditions: 1 mm glassy carbon working electrode; 25 ˚C; scan rate 10 V/s.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 41
1
2
3
4
5
6
7
8
observed rate constant becomes independent of acid concentration above 0.14 M [(DMF)H]+
(Figure 7B). Higher concentrations of acid were not studied due to decomposition of each
catalyst, resulting in unreliable measurements. From the plots of kobs vs [H+] the secondꢀorder
rate constants (first order in acid and first order in catalyst) were determined in the absence of
added water and ranged from 6,300 to 69,000 Mꢀ1 sꢀ1 (Table 2). The catalytic production of H2
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
Table 2. Electrocatalytic data for hydrogen production.
Overꢀ
kobs
k
[(DMF)H]+,
[H2O]
0.21 M,
kobs
k
Complex
[(DMF)H]+
0.15 M
potentiala
(sꢀ1)
(Mꢀ1 sꢀ1)
(sꢀ1)
(Mꢀ1 sꢀ1)
640 mV
630 mV
620 mV
580 mV
580 mV
550 mV
7,500
51,000b
22,000
ꢀꢀ
5-OMe
0.52 M
0.41 M,
0.74 M
0.42 M,
1.2 M
0.39 M
0.43 M
0.40 M
0.43 M
0.41 M
5-Me
5-Hc
27,000
33,000
69,000
77,000
96,000
240,000
250,000
106,000
0.40 M,
0.74 M
0.43 M,
1.0 M
6,100
4,500
2,400
15,000
10,000
6,300
17,000
15,000
4,100
42,000
34,000
10,000
5-Br
5-Cl
0.40 M
0.81 M
5-CF3
a As determined by the method of Evans for calculating overpotentials.63 b Obtained for the
linear region of the kobs vs. [(DMF)H]+ before the acidꢀindependent region. c As previously
reported.50
was confirmed by quantitative gas chromatographic analysis of the H2 produced during a
controlled potential electrolysis experiment with 5-Me. At a potential of –1.4 V and an
[(DMF)H]+ concentration of 0.2 M, a current efficiency of 99 ±5% for H2 production (11
turnovers) was observed.
16
ACS Paragon Plus Environment

Page 17 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
Based on previous studies in which water was shown to significantly increase catalytic
rates for [Ni(PR NR’ ) ](BF4)2 complexes, aliquots of purified H2O were added subsequent to the
2
2 2
7
8
9
completion of acid additions, resulting in a further catalytic current enhancement (e.g. 5-Br,
Figure 4).39ꢀ41 Water was added until the observed catalytic current enhancement ceased; the
data is summarized in Table 2. The secondꢀorder rate constants k for 5-Me, 5-Br, 5-Cl and 5-
CF3 were found to range from 10,000 to 240,000 Mꢀ1sꢀ1 with the addition of water. For catalyst
5-OMe, at acid concentrations above 0.15 M, the catalytic current and hence the catalytic rate
become independent of the acid concentration, with a firstꢀorder rate constant (equivalent to the
turnover frequency of the catalyst) of 22,000 sꢀ1 in the presence of water (Figure 7B).
Detailed computational studies on 5-OMe, 5-H and 5-CF3 were carried out to gain
further insights into the complex properties and the catalytic process and reported below. All
calculations were carried out using the free energies of solvation in acetonitrile, which was also
explicitly considered as fifth ligand for all of Ni(II) complexes. The results of these calculations
are discussed below.
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
Discussion
Structural Effects on Catalyst Properties. Xꢀray diffraction studies of the new
complexes 5-OMe, 5-Me, 5-Br, 5-Cl and 5-CF3 have shown that the smaller ring size of the
heterocyclic ligand results in significant structural differences in the complexes relative to the
[Ni(PR NR’ ) ]2+ series, and this in turn leads to substantial electronic differences that influence
2
2 2
the catalytic activities. An important structural feature in each of the new complexes is the small
PꢀNiꢀP bite angle resulting from the 7PPh2NC6H4X ligand (79.7ꢀ80.1°). This small bite angle
reduces the interligand steric interactions of the phenyl substituents of the phosphorus atoms of
the two diphosphine ligands compared to [Ni(PPh2NC6H4X2)2]2+ complexes which have somewhat
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 41
1
2
larger bite angles (82ꢀ84º).39,40 As a result, all of the [Ni(7PPh2NC6H4X)2]2+ cations exhibit planar
structures with dihedral angles of 0.0º between the two planes defined by the phosphorus atoms
of each diphosphine ligand and nickel. In previous studies of [M(diphosphine)2]2+ complexes
(where M = Ni, Pd, and Pt), this dihedral angle has been shown to correlate with the potentials of
M(II/I) couples.64ꢀ66 The potentials of the Ni(II/I) couples shift to more negative values as the
complexes become more planar, approaching the potentials of the M(I/0) couples. As a result,
3
4
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
the potentials of the Ni(II/I) couples of the [Ni(7PPh NC6H4X)2]2+ complexes are more negative by
2
250ꢀ300 mV than those of the corresponding [Ni(PPh2NC6H4X2)2]2+ complexes, which have larger
dihedral angles. The resulting redox waves associated with the Ni(II/I) and Ni(I/0) couples
overlap for the [Ni(7PPh2NC6H4X)2]2+ complexes due to the small difference in the Ni(II/I) and
Ni(I/0) potentials (ꢀE1/2). For example, simulations of cyclic voltammograms of 5-H indicate a
difference in redox potentials of 70 mV.50 In contrast, these potentials differ by 190 mV for
[Ni(PPh NPh ) ]2+.39
2
2 2
An extensive theoretical analysis of the structural, electrochemical and acid/base
properties of the 5-OMe, 5-H and 5-CF3 derivatives fully supports this interpretation. The
calculated reduction potentials of the complexes are reported in Table 3. Overall, the
Table 3. Calculated electrochemical potentials (V) for
[Ni(7PPh2NC6H4X)2]2+ complexes in CH3CN.
X
OMe
H
E(II/I)
ꢀ1.07
ꢀ1.04
ꢀ0.98
E(I/0)
ꢀ1.15
ꢀ1.18
ꢀ1.07
CF3
18
ACS Paragon Plus Environment

Page 19 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
experimentally measured (overlapping) potential for the Ni(II/I, I/0) processes and the calculated
potentials for the Ni(II/I) and Ni(I/0) couples are in good agreement, with the Ni(I/0) couples
occurring negative of the Ni(II/I) couples by 0.14 V or less. In particular, calculations semiꢀ
quantitatively reproduce the shift toward more positive values for the Ni(II/I) couple from 5-
OMe to 5-CF3.
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 more negative potentials of the Ni(II/I) couples of the [Ni(7PPh2NC6H4X)2]2+
derivatives compared to the [Ni(PPh2NC6H4X2)2]2+ analogues also correlates with the hydride
donor abilities of their Ni(II) hydrides (Eq. 6). On the basis of previous correlations, the free
energy
for
the
CH3CN
[Ni(7PPh2NPh)2]2+
H- ꢀG˚H- = 54.9 kcal/mol
[HNi(7PPh2NPh)2]+
+
(6)
5-H
heterolytic cleavage of the NiꢀH bond in acetonitrile to form Hꢀ, ꢀG°Hꢀ, is estimated to be 54.9
kcal/mol for [HNi(7PPh2NPh)2]+ (5-H), while that for [HNi(PPh2NPh ) ]+ is 59.0 kcal/mol.39,67,68
2 2
For the [HNi(7PPh2NC6H4X)2]+ family, the ꢀG˚Hꢀ values range from 54.5 for 5-OMe to 56.0
kcal/mol for 5-CF3.
Mechanistic Studies. A likely ECEC mechanism for the formation of H2 for the
[Ni(7PPh2NC6H4X)2]2+ complexes is proposed in Figure 8.
The complexes are first
electrochemically reduced (moving clockwise around Figure 8) to the Ni(I) species (ECEC),
followed by protonation of a pendent amine (ECEC), and the second electron transfer (ECEC) to
form the Ni(0) complex [Ni(7PPh2NC6H4XH)(PPh NC6H4X)]+. This step is likely followed by a
2
intramolecular proton transfer to form a Ni(II) hydride, [HNi(7PPh2NC6H4X)2]+. Finally, the
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 41
1
2
3
4
5
6
second protonation step (ECEC) occurs, followed by HꢀH bond formation, hydrogen elimination
and regeneration of the original catalyst.
7
8
9
Computational and experimental data support electron transfer followed by protonation
of the Ni(I) species as the first steps in the catalytic process (Fig. 8, steps 1ꢀ2). The calculated
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
redox potentials for singly protonated Ni(II) complexes, [Ni(7PPh2NC6H4XH)(PPh NC6H4X)]3+,
2
Figure 8. Proposed mechanism for catalytic H2 formation.
resulting from protonation of 5-OMe, 5-H and 5-CF3, are significantly more positive than those
of the nonꢀprotonated species (Table 4). For instance, the potential was calculated to be at ꢀ0.41
V for the Ni(II/I) couple of the singly endoꢀprotonated complex derived from 5-H (Table 4).
When using [(DMF)H]+ as the proton source (pKa = 6.1 in CH3CN),57 the experimental data
20
ACS Paragon Plus Environment

Page 21 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
show catalysis occurring near the Ni(II/I)/(I/0) overlapping potentials (i.e. ꢀ1.1 V), supporting
electron transfer as the first step in the mechanism. Additionally, the computationally derived
pKa values of singly protonated Ni(II) complexes (Table 4) are greater than 6 pKa units more
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
acidic
than
Table 4. Calculated electrochemical potentials (V) and pKa values in
CH3CN for monoꢀprotonated Ni(II) complexes,
[Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]3+.
2
X
OMe
H
E(II/I)
ꢀ0.42
ꢀ0.41
ꢀ0.30
E(I/0)
ꢀ0.49
ꢀ0.43
ꢀ0.35
pKa
ꢀ1.8
ꢀ3.0
ꢀ5.3
CF3
[(DMF)H]+, indicating protonation at nitrogen prior to reduction of Ni(II) to Ni(I) is unlikely
under catalytic conditions.
We calculated the relative free energies and pKa values of species that can form by
reduction followed by protonation. The pKa values for the lowest freeꢀenergy isomers of the
singly protonated Ni(I) isomers, [Ni(7PPh2NC6H4XH)(PPh2NC6H4X)]2+, are reported in Table 5.
Table 5. Calculated pKa values in CH3CN for endo and exo isomers of
[Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]2+ and endo isomer of [HNi(7PPh2NC6H4XH)(7PPh2NC6H4X)]2+
2
(e/NiH).
X
OMe
H
e(I)
7.7
5.9
3.7
x(I)
4.0
e/NiH
10.3
8.9
1.4
ꢀ0.2
6.8
CF3
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 41
1
2
3
4
5
6
7
8
Protonation on the nitrogen atoms can occur endo to form e(I) (e indicating endo, I indicating the
oxidation state of the metal, Fig. 8, step 2endo) or exo to form x(I) (x indicating exo protonation,
Fig. 8, step 2exo), with respect to the metal center, resulting in the possibility of branching away
from the main catalytic pathway. Calculations show that singly protonated Ni(I) exo isomers,
x(I), are systematically more acidic than the endo isomers e(I) by 3.7 – 4.5 pKa units (Table 5).
Consequently, e(I) isomers are more stable than x(I) isomers (e.g. 5-H, Fig. 9A). This relative
energy contrasts with the [Ni(PPh2NC6H4X2)2]2+ catalysts, where the x(I) protonated species is
stabilized by a NH•••N hydrogen bond from the second pendent amine in the ligand,69 resulting
in stabilization of the x(I) species relative to the e(I) species (e.g. X = H, Fig. 9B). As such, the
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
(B)
(A)
Figure 9. Ranking of the lowestꢀfree energy isomers of the singly protonated Ni(I) complexes
of (A) [Ni(7PPh2NPhH)(7PPh NPh)]2+ and (B) [Ni(PPh2NPh H)(PPh NPh )]2+.
2
2
2
2
22
ACS Paragon Plus Environment

Page 23 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
relative pKa values of the protonated pendent amines with respect to the pKa of the proton source
(i.e. the exogenous acid, protonated DMF in our studies) will play an important role in
determining which protonation pathway is dominant in both cases (Fig. 8, step 2endo or step 2exo).
In the [Ni(7PPh2NC6H4X)2]2+ family, free energy calculations indicate protonation in the e(I)
position will be favored in the case of 5-CF3 and 5-H (Fig. 8, step 2endo), as the pKa of x(I)
position lies 4ꢀ5 pKa units more acidic than that of the proton source, [(DMF)H]+. As the
basicity of the pendent amine increases, however, exo protonation may start to play a role in the
catalytic pathway. For example, in the case of 5-OMe the calculated pKa values of both the e(I)
and x(I) complexes are only 1ꢀ2 pKa units different than that of [(DMF)H]+, so protonation to
form the x(I) species is thermodynamically favorable (Fig. 8, step 2exo). The resulting effect may
be an observed decrease in the catalytic rate of 5-OMe, as protonation to generate x(I) becomes
competitive and removes active catalyst from the cycle. As previously reported for the
[Ni(PR NR’ ) ]2+ catalysts, exo protonation is thought to result in slower rates of catalytic
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
2 2
hydrogen production, and that also appears to be the case for 5-OMe.48
After protonation to form the Ni(I) species, [Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]2+, rapid
2
reduction to Ni(0) likely occurs. As shown in Table 4, the calculated Ni(I/0) redox potentials for
the mono endoꢀprotonated [Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]2+ species occur between ꢀ0.49 and ꢀ
2
0.35 V (Fig. 8, step 3endo), and is ꢀ0.71 V when protonated on exo site (Figure 8, step 3exo), well
positive of the Ni(II/I) redox couple for [Ni(7PPh2NC6H4X)2]2+ complexes where catalysis occurs.
As such, reduction and protonation to form the [Ni(7PPh NC6H4XH)(7PPh NC6H4X)]2+ species is
2
2
expected
to
be
rapidly
followed
by
a
second
reduction
to
form
[Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]+, which are Ni(0) complexes that are either endo e(0) or exo
2
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 41
1
2
3
4
x(0) protonated at the pendant amine. Previous experimental and computational studies on
5
6
[Ni(PR NR’ ) ]2+ have shown that endo protonated Ni(0) species rapidly isomerize to form Ni(II)
2
2 2
7
8
9
hydride species.47,70 Similarly, once reduction to form the Nꢀprotonated Ni(0) species,
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
[Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]+, occurs rapid intramolecular proton transfer likely follows
2
(Fig. 8, step 4), forming the Ni(II) hydride species [HNi(7PPh2NC6H4X)2]+ (NiH). No such
intramolecular isomerization is available for the x(0) species.
Protonation of the Ni(II) hydride species (NiH) can generate either an endoꢀprotonated
hydride (e/NiH, Fig 8, step 5) or an exoꢀprotonated hydride (x/NiH) (see Figure 10).38,47,48
Protonation of x(0) can occur in an endo position, relative to the nickel, yielding an e/x(0)
species (which would undergo a rapid intramolecular isomerization to the exo protonated Ni(II)
hydride, x/NiH, Fig. 10) or exo, forming x/x(0) (Fig. 10), a doubly protonated Ni(0) species. All
of the doublyꢀprotonated species with exo positioned protons (x/NiH or x/x(0)), however, are
calculated to be systematically higher in free energy than their endo counterpart (e/NiH). As
shown in Figure 10, the energy differences between endo and exo species are even more
pronounced for the second protonation than the first (Figure 9). Given that the calculated pKa
24
ACS Paragon Plus Environment

Page 25 of 41
Journal of the American Chemical Society
1
2
3
4
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
(B)
(A)
Figure 10. Ranking of the lowestꢀfree energy isomers of the doubly protonated complexes of
(A) [HNi(7PPh2NC6H4XH)(7PPh2NC6H4X)]2+/[Ni(7PPh NPhH)2]2+ and (B)
2
[HNi(PPh NPh H)(PPh NPh )]2+/[Ni(PPh NPh H)2]2+.
2
2
2
2
2
2
values of the lowest energy intermediate (e/NiH) for 5-CF3, 5-H, and 5-OMe (Table 5) are
greater than that of [(DMF)H]+, whereas the calculated pKa values of the e(I) intermediates are
comparable to that of [(DMF)H]+, it is postulated that the first protonation step, Ni(I) +
[(DMF)H]+ → e(I) + DMF, is the rateꢀdetermining step rather than the second protonation step,
NiH + [(DMF)H]+ → e/NiH + DMF.
In stark contrast to the [Ni(7PPh2NC6H4X)2]+ family, protonation studies of the
[Ni(PR NR’ ) ]2+ family of compounds, complemented by extensive theoretical modeling,
2
2 2
indicate that the energy difference between endo and exo protonation is much smaller, and all
isomers are likely populated under ambient conditions (Figure 9).38,46ꢀ48 Indeed, the presence of
the second pendent amine on each phosphine ligand allows the formation of exo hydrogenꢀ
bonded structures where the proton is “pinched” between two pendent amines, NH•••N. In the
case of [Ni(PPh2NPh ) ]2+ this hydrogen bond stabilizes the x/x(0) and x/NiH species, making
2 2
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 41
1
2
their relative energies comparable (Figure 10).38 The stability of the exo protonated species in
3
4
5
6
the [Ni(PPh2NPh ) ]2+ family is thought to be a key factor in limiting their rate of catalysis, and
2 2
7
8
9
one that the [Ni(7PPh2NC6H4X)2]+ family has overcome.
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 evolution of hydrogen from the e/NiH species (Fig. 8, step 6) is thought to proceed
in a heterolytic fashion through a transient dihydrogen Ni(II) complex, [Ni(PPh2NPh ) (H2)]2+.38
2 2
As shown Figure 11, computational studies carried out on 5-CF3, 5-H, and 5-OMe show
formation of the Ni(II) dihydrogen complex through a heterolytic transition state from the e/NiH
intermediate. The activation barriers calculated for this transition state range from 6.0ꢀ7.4
kcal/mol, far smaller than the barrier (> 10 kcal/mol) inferred from the observed turnover
frequency, further supporting the proposal that protonation of the Ni(I) species is the rateꢀ
determining step. The barrier for the subsequent H2 elimination is believed to be low (smaller
than those reported above), resulting in facile regeneration of the original Ni(II) catalysts.38
Figure 11. Lowest free energy H2 elimination pathway for 5-OMe, 5-H and 5-CF3.
26
ACS Paragon Plus Environment

Page 27 of 41
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Interestingly, the overall free energy for H2 elimination from the e/NiH intermediate (ꢀ
ꢀG˚H2) obtained from our quantum chemical calculations is more favorable for [Ni(7PPh2NPh)2]+
(5-H, ꢀꢀGo = ꢀ10 kcal/mol) than for [Ni(PPh2NPh2)2]2+ (ꢀꢀGo = ꢀ7.5 kcal/mol). As discussed
H2
H2
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
previously,71 ꢀGo is the result of a fine balance between the free energy for hydride transfer,
H2
ꢀG°Hꢀ, and the pKa of the protonated pendent amine of e/NiH. This is evident from the
thermodynamic cycle shown in Fig. 12 (e.g. 5-H).
The more favorable
−ꢀG˚H-
[Ni(7PPh2NPh)2]2+
[HNi(7PPh2NPh)2]+
[HNi(7PPh2NPh)2]+
H-
H+
H2
+
+
[HNi(7PPh2NPhH)(7PPh2NPh)]2+
H- H+
ꢀG˚ = -1.37(pKa)
+
ꢀG˚ = 76 kcal/mol (in CH3CN)
[HNi(7PPh2NPhH)(7PPh2NPh)]2+
ꢀG˚H2
[Ni(7PPh2NPh)2]2+
H2
+
Figure 12. Themochemical cycle for the determination of the driving force for hydrogen
addition, ꢀG˚H2.
ꢀꢀGo for the [Ni(7PPh2NC6H4X)2]+ complexes is a consequence of the greater hydride donor
H2
ability of [HNi(7PPh2NC6H4X)2]+ discussed above, which dominates over the increased basicity of
the pendent amine.
Of particular note from the kinetic studies of the [Ni(7PPh2NC6H4X)2]2+ catalysts is the
observation that they all show a firstꢀorder dependence on the concentration of acid. This
indicates that although two protonation steps must occur during the catalytic cycle, only one
protonation step appears to be rateꢀdetermining for these catalysts. Evidence that the first
protonation event may be rateꢀdetermining comes from analysis of the secondꢀorder rate
constants derived from the observed rate constants for [Ni(7PPh NC6H4X)2]2+ catalysts (column 5
2
of Table 2) and the pKa values of proposed intermediates. The calculated pKa values of the
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 41
1
2
3
4
5
6
7
8
9
e/NiH intermediates for 5-CF3, 5-H, and 5-OMe (Table 5) are greater than the pKa of the
[(DMF)H]+ proton source (6.1), whereas the calculated pKa values of the e(I) intermediates are
comparable to or less than that of [(DMF)H]+. As a result, it is reasonable to assume that the first
protonation step, Ni(I) + [(DMF)H]+ → e(I) + DMF (Fig. 8, step 2), is the rateꢀdetermining step
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
rather than the second protonation step, NiH + [(DMF)H]+ → e/NiH + DMF (Fig. 8, step 5).
A plot of the secondꢀorder rate constants observed for the [Ni(7PPh2NC6H4X)2]2+ catalysts
vs. the pKa values of the protonated amine of e(I), ([Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]2+), shows
2
the maximum rate constant occurring with the pendent amine that best matches the pKa of the
[(DMF)H]+ proton source (Figure 13). The pKa values of e(I) species derived from more acidic
pendent amines than 5-H (5-CF3, 5-Cl and 5-Br) are below that of [(DMF)H]+, resulting in slow
(or incomplete) protonation. In complexes where the pendent amine substituents have a greater
electron donating ability than the H atom in 5-H, the pendent amine is basic enough to give rise
to competitive exo protonation (Fig. 8, step 2exo), and the competitive formation of x(I) becomes
likely. This competition is particularly pronounced in the case of the catalytic rate for 5-OMe,
which shows acid independence beyond 0.14 M [(DMF)H]+, indicating the equilibrium between
e(I) protonation/deprotonation plays a key role in limiting the overall catalytic process. Figure
13 illustrates the importance of pKa matching between the pendent base and the exogenous base
for relaying protons from the substrate to the metal via the pendent amines. Additionally, Figure
13 supports the proposed ECEC mechanism. An EECC mechanism would proceed through
doubly protonated Ni(0) species that rapidly interconverts to the protonated Ni(II) hydride (Fig.
8, e/NiH). Table 5 shows the pKa values for the [HNi(7PPh2NC6H4XH)(7PPh NC6H4X)]2+ (e/NiH)
2
species are all greater than that of [(DMF)H]+. Hence, if an EECC mechanism were occurring,
the expected trend in Figure 13 would not be observed.
28
ACS Paragon Plus Environment

Page 29 of 41
Journal of the American Chemical Society
1
2
3
4
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 13.
Graph of calculated pKa values of the endo protonated, e(I)
[Ni(7PPh2NC6H4XH)(7PPh NC6H4X)]2+ complexes vs. their secondꢀorder rate constants for
2
hydrogen production. Black line between data points was added to illustrate the trend.
Effect of Water on Catalytic Rates. Addition of water (0.52 – 1.2 M) to reaction
mixtures containing [Ni(7PPh2NC6H4X)2]2+ complexes and [(DMF)H]+ results in significant
catalytic rate enhancements, as previously observed for the [Ni(PR NR’ ) ]2+ family of
2
2 2
complexes.39ꢀ41 Computational and experimental studies on the doubly protonated
[Ni(PCy2NBn2H)2]2+ compounds suggest the rates of protonation and deprotonation of the pendent
amines are hindered through steric interaction between phosphine substituents and the
approaching substrate.48 Indeed, the use of smaller substrates for proton delivery ([(DMF)H]+
vs. 2,5ꢀdichloroanilinium) in the [Ni(PPh2NC6H4X2)2]2+ family of catalysts resulted in higher rates.
Additionally, in the family of [Ni(PR NPh ) ]2+ catalysts that contain bulky substituents at the
2
2 2
phosphorus atoms, large increases in the rate of hydrogen production upon addition of water are
attributed to more facile proton delivery through water acting as an intermolecular proton relay
29
ACS Paragon Plus Environment