Incorporating peptides in the outer-coordination sphere of bioinspired electrocatalysts for hydrogen production

Article abstract of DOI:10.1021/ic1025872

Four new cyclic 1,5-diaza-3,7-diphosphacyclooctane ligands have been prepared and used to synthesize [Ni(P^Ph₂N^R ₂)₂]²⁺ complexes in which R is a mono- or dipeptide. These complexes represen

Full text of DOI:10.1021/ic1025872

ARTICLE
pubs.acs.org/IC
Incorporating Peptides in the Outer-Coordination Sphere
of Bioinspired Electrocatalysts for Hydrogen Production
Avijita Jain, Sheri Lense, John C. Linehan, Simone Raugei, Herman Cho, Daniel L. DuBois, and
Wendy J. Shaw*
Chemical and Materials Sciences Division, Pacific Northwest National Laboratory, Richland, Washington 99354, United States
S Supporting Information
b
ABSTRACT: Four new cyclic 1,5-diaza-3,7-diphosphacyclooctane ligands have
Ph
R
2þ
2 2
been prepared and used to synthesize [Ni(P N ) ] complexes in which R is
2
a mono- or dipeptide. These complexes represent a first step in the development
of an outer-coordination sphere for this class of complexes that can mimic the
outer-coordination sphere of the active sites of hydrogenase enzymes. Impor-
tantly, these complexes retain the electrocatalytic activity of the parent
Ph Ph
2þ
[
Ni(P 2N ) ] complex in an acetonitrile solution with turnover frequen-
2 2
1
ꢀ
cies for hydrogen production ranging from 14 to 25 s in the presence of
p-cyanoanilinium trifluoromethanesulfonate and from 135 to 1000 s in the
ꢀ
1
presence of protonated dimethylformamide, with moderately low overpotentials,
∼
0.3 V. The addition of small amounts of water results in rate increases of 2ꢀ7
times. Unlike the parent complex, these complexes demonstrate dynamic
structural transformations in solution. These results establish a building block from which larger peptide scaffolding can be
0
R
R
2þ
2 2
added to allow the [Ni(P N ) ] molecular catalytic core to begin to mimic the multifunctional outer-coordination sphere of
2
enzymes.
’
INTRODUCTION
prepared for the electrochemical oxidation and production of
9
ꢀ18
dihydrogen.
This work demonstrated that the functional
The production of hydrogen from renewable energy sources
hydrogenase mimics with a properly positioned pendant amine,
such as solar or wind will require electrocatalysts for proton
reduction. Similarly, the efficient utilization of dihydrogen in fuel
cells requires highly active electrocatalysts for the reverse reac-
tion, hydrogen oxidation. While platinum is currently the catalyst
of choice for these reactions, naturally occurring hydrogenase
enzymes efficiently and rapidly produce and oxidize hydrogen
or proton relay, can approach the rates of [NiFe] hydrogenase
ꢀ
1
enzymes for dihydrogen production (350 and 750 s
respectively).
,
2,15
In particular, nickel catalysts with 1,5-diaza-
0
R R
3
,7-diphosphacyclooctane ligands (P N 2; 2) with the general
2
0
R
R
2þ
2 2
formula [Ni(P N ) ] exhibit rates for dihydrogen produc-
2
1
ꢀ5
tion and oxidation that are at least 2ꢀ3 orders of magnitude
without using expensive metal catalysts.
While utilizing
larger than those for analogous complexes without the positioned
hydrogenase enzymes directly in fuel cells may have technolog-
ical challenges, learning from and implementing some of the
features of these enzymes can assist in the design of small-
molecule catalysts.
11,12,14ꢀ16,19,20
relay.
Though hydrogenases have been extensively studied, the
mechanism by which they achieve fast catalytic rates and very
low overpotentials for hydrogen production is not fully
1
understood. The solving of several crystal structures in the
mid-1990s revealed bimetallic active sites consisting of abundant
4
,6,7
metals, either nickel and iron or iron and iron.
structural mimics of the bimetallic iron [FeFe] (1) or hetero-
Many
1,7
bimetallic nickelꢀiron [NiFe] complexes have been prepared
but have been found to catalyze dihydrogen production and
oxidation much more slowly than hydrogenase or with higher
Although a number of nickel-based electrocatalysts have been
reported, these synthetic catalysts still fall short of the rate of the
8
ꢀ
1 2,21
overpotentials. The pendant base in [FeFe] hydrogenases has
fastest hydrogenases, [FeFe] (10 000 s ).
The significant
been suggested to serve as a relay for the transfer of protons from
3,7
the metal to the proton conduction channels. Using this design
principle, synthetic functional analogues of hydrogenases were
Received: December 30, 2010
Published: April 01, 2011
r 2011 American Chemical Society
4073
dx.doi.org/10.1021/ic1025872
|Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
Ph DPE
rate enhancement due to the second coordination sphere in
small-molecule catalysts, as well as the structural features of the
enzyme, suggests that the outer-coordination sphere may also
contribute significantly to catalysis at the active site. As a starting
P
2
N
2
(3c; DPE = Dipeptide Ester). To a solution of glycine ethyl
ester H PO (0.168 g, 0.830 mmol) in tetrahydrofuran (25 mL) were
3
3 4
added 3a (0.244 g, 0.41 mmol), DMAP (0.109 g, 0.896 mmol), and
DCC (0.185 g, 0.896 mmol). The reaction mixture was stirred at room
temperature for 19 h. The suspended impurities were removed by
filtration through Celite. The solvent was removed under vacuum, and
point to mimic the function of the protein scaffold, we have
0
R
R
2þ
2
incorporated small peptides into [Ni(P N ) ] electrocata-
2
2
0
R
R
31
1
lysts. In this work, we report the synthesis of new P N ligands
the solid was collected. Yield: 0.282 g, 0.74 mmol, 82%. P{ H} NMR
2
2
1
containing unnatural amino acids and dipeptides on the central
nitrogen atom. These small biologically based ligands serve as a
(CD
18H, ArꢀH), 8.09 (s, 2H, NH), 4.38 (m, 4H, NCH
COOCH ), 4.07 (d, 4H, PCH N), 3.96 (d, 4H, PCH N), 2.80 (t, 4H,
2
Cl
2
): δ ꢀ50.5 (s). H NMR (CD
2
Cl
2
): δ 6.68, 7.05, 7.46, 7.63 (m,
COO), 4.14 (q, 4H,
2
Ph
R
2þ
starting point for the preparation of [Ni(P N ) ] com-
2
2
2
2
2 2
CH COO), 2.44 (t, 4H, PhCH ), 1.29 (t, 6H, CH ).
plexes with peptide functional units in the outer-coordination
sphere. They also provide a comparison of the catalytic activities
2
2
3
Ph AA
[
Ni(P
2
N
2
)
2
](BF
4
)
2
(4a). Solid 3a (0.957 g, 1.60 mmol) and
Ph Ph
2þ
[
Ni(CH
3
CN)
6
](BF
4
)
2
(0.4 g, 0.8 mmol) were stirred for 24 h in
of these complexes relative to the parent [Ni(P 2N ) ]
2 2
acetonitrile (20 mL) to give a red solution. The solution was filtered
using a medium frit to remove small amounts of suspended impurities.
The solvent was removed under vacuum, and the residue was dissolved
in 5 mL of acetonitrile. The product was then flash-precipitated by
catalyst, before incorporating larger peptides into the outer-
coordination sphere.
’
MATERIALS AND METHODS
addition to 60 mL of stirring diethyl ether. Yield: 1.05 g, 0.74 mmol, 92%.
1
H NMR (CD
3
CN): δ 7.41, 7.29, 7.17 (m, 36H, ArꢀH), 3.90 (d, 8H,
1
31
General Procedures. Solution-state H and P NMR spectra were
recorded on a Varian Inova or VnmrS spectrometer (500 MHz H
frequency). All H chemical shifts were internally calibrated to the mono-
protic solvent impurity. All P{ H} chemical shifts were externally
referenced to phosphoric acid. Solid-state P{ H} NMR spectra were
run on a Chemagnetics console using a single pulse excitation and spinning
at 11.5 kHz. Elemental analyses were done by Atlantic Microlab, Norcross,
GA, with V O as a combustion catalyst. Electrospray ionization (ESI) and
PCH N), 4.21 (d, 8H, PCH N), 2.90 (t, 8H, CH COO), 2.63 (t, 8H,
2
2
2
1
3
1
1
PhCH
2
), 9.04 (s, 4H, COOH). P{ H} NMR (CD
NiB : C, 57.17; H, 5.08; N, 3.92.
Found: C, 55.55; H, 4.98; N, 4.08. ESI MS (CH CN): m/z 1341
3
CN): δ 5.33 (s).
1
Elem anal. Calcd for C68
H
72
N
4
O
8
P
4
2 8
F
31
1
3
31
1
Ph DPA
þ
{
[Ni(P N2 ) ](BF )} . A cyclic voltammogram recorded in
2
þ
2
4
n
ꢀ
Bu N OTf /benzonitrile consisted of two reversible reduction waves
4
II/I
I/0
at ꢀ0.82 and ꢀ1.02 V corresponding to the Ni and Ni couples.
Ph AAE
2
5
[
Ni(P
2
N
2
2 4 2
) ](BF ) (4b). Complex 4b was prepared in a manner
chemical ionization mass spectrometry (MS) spectra were collected at the
Indiana University Mass Spectrometry Facility on a Waters/Micromass
LCT Classic using anhydrous solvents and inert-atmosphere techniques.
Synthesis and Materials. Acetonitrile was purchased from Alfa
Aesar and ethyl ether from Honeywell Burdick & Jackson, and they were
dried using an activated alumina column. 3-(4-Aminophenyl)propionic
acid and 3-(4-aminophenyl)propionic ethyl ester were purchased from
Sigma Aldrich. All reactions were performed using standard Schlenk
techniques under a dinitrogen atmosphere or in a glovebox unless
otherwise noted. Dicyclohexyl carbodiimide (DCC) and (dimethylamino)-
pyridine (DMAP) were purchased from Sigma Aldrich. PhP-
analogous to that of 4a using 3b (0.1.046 g, 1.6 mmol) and [Ni(CH
3
-
1
CN)
.41, 7.27, 7.17 (m, 36H, ArꢀH), 1.18 (t, 12H, CH
PCH N), 4.06 (q, 8H, COOCH ), 4.18 (d, 8H, PCH N), 2.93 (t, 8H,
6
](BF
4
)
2
. Yield: 0.700 g, 0.45 mmol, 91%. H NMR (CD
3
CN): δ
7
3
), 3.90 (d, 8H,
2
2
2
31
1
CH COO), 2.63 (t, 8H, PhCH ). P{ H} NMR (CD CN): δ 4.99 (s).
2
2
3
Elem anal. Calcd for C H N O P NiB F : C, 59.21; H, 5.75; N, 3.63.
7
6
88
4
8
4
2 8
Found: C, 56.42; H, 5.59; N, 3.86. ESI MS (CH
3
CN): m/z 1451
)} . A cyclic voltammogram recorded in
þ
4
N OTf /acetonitrile consisted of two reversible reduction waves
Ph AAE
þ
{
[Ni(P
2
N
2
2 4
) ](BF
n
ꢀ
Bu
at ꢀ0.83 and ꢀ1.04 V corresponding to the Ni and Ni couples.
II/I
I/0
Ph DPE
[
Ni(P
2
N
2
2 6 2
) ](PF ) (4c). Solid 3c (0.100 g, 0.125 mmol) and
(
CH OH) was prepared by reacting 1 equiv of PhPH with 2 equiv
2 2 2
1
[
Ni(CH CN) ](PF ) (0.036 g, 0.062 mmol) were stirred for 3 days in
9
3
6
6 2
of p-formaldehyde.
acetonitrile (20 mL) to give a red solution. The solution was filtered
using a medium frit to remove small amounts of suspended impurities.
The solvent was removed under vacuum, and the residue was dissolved
in 5 mL of acetonitrile. The product was then flash-precipitated by
addition to 60 mL of stirring diethyl ether. Yield: 0.093 g, 0.53 mmol,
Ph AA
2
P
N
2
(3a; AA = Amino Acid). 3a was synthesized following a
22
previously published preparation. A solution of 3-(4-aminophenyl)-
propionic acid (0.49 g, 3.0 mmol) in ethanol was added to a solution of
2 2
PhP(CH OH) (0.51 g, 3.0 mmol) in ethanol (30 mL) in a glovebox,
and the mixture was heated to 65 °C for 16 h to form a white precipitate.
The solution was cooled to room temperature, and the product was
filtered using 20 mL of ethanol to rinse it from the flask and 20 mL of
ethanol to wash the solid. The yield obtained was 0.76 g, 0.13 mmol,
3
1
1
1
8
5%. P{ H} NMR (CD
3
CN): δ 4.04 (s), 21.6 (s), ꢀ13.6 (s). H
NMR (CD CN): δ 6.88, 7.20, 7.94 (m, 36H, ArꢀH), 8.03 (s, 4H, NH),
3
4
3
.20 (m, 8H, NCH COO), 4.15 (d, 8H, PCH N), 3.86 (d, 8H, PCH N),
2 2 2
.42 (q, 8H, COOCH ), 2.92 (t, 8H, CH COO), 2.63 (t, 8H, PhCH ),
1.29 (t, 6H, CH ). ESI MS (CH CN): m/z 1739 {[Ni(P N2 )2]-
2
2
2
8
5%. The resulting product was not soluble in common organic solvents
Ph DPE
31
1
1
3
3
2
but was soluble in basic water (pH 10). P{ H} NMR: δ ꢀ50.0. H
þ
n
þ
ꢀ
(PF
6
)} . A cyclic voltammogram recorded in Bu
4
N OTf /acetonitrile
NMR (basic D
2
O): δ 7.40, 7.34, 6.87, 6.39 (m, 18H, ArꢀH), 3.96 (d,
N), 4.06 (d, 4H, PCH N), 2.52 (t, 4H, CH COO), 2.16 (t,
H, PhCH ). Elem anal. Calcd for C H N O P : C, 68.22; H, 6.06; N,
consisted of one reversible reduction wave at ꢀ0.83 V corresponding to
I/0
4
4
4
H, PCH
2
2
2
II/I
the Ni couple and one quasireversible reduction wave at ꢀ1.03 V
Ph DPA
2
34 36 2 4 2
corresponding to the Ni couple.
.68. Found: C, 68.29; H, 6.14; N, 4.73. ESI MS: m/z 621
Ph APPA
þ
[
Ni(P
2
N
2
2 6 2
) ](PF ) (4d; DPA = Dipeptide Acid). 4d was synthe-
{(P
2
N
2
)
2
] þ Na} .
Ph AAE
2
sized on a polystyrene-based Wang resin (Novabiochem, San Diego,
P
N
2
(3b; AAE = Amino Acid Ester). This ligand was prepared in
2
3
CA), using solid-phase peptide synthesis techniques. The fluorenyl-
methoxycarbonyl (FMOC) group was removed by stirring a glycine-
substituted Wang resin (0.181 g, 0.61 mmol) with 20 mL of 20%
piperidine in N-methylpyrrolidone (NMP). The solid was rinsed with
a manner analogous to that of 3a using PhP(CH OH) and 3-(4-
2
2
aminophenyl)propionic ester (0.582 g, 3 mmol) to form a pale-yellow
3
1
1
solid. Yield: 0.85 g, 1.30 mmol, 87%. P{ H} NMR (CD
2
Cl
2
): δ ꢀ50.5
1
(
s). H NMR (CD Cl ): δ 6.66, 7.04, 7.48, 7.62 (M, 18H, ArꢀH), 1.21
2
2
(
t, 6H, CH ), 4.07 (m, 4H, PCH N and 4H, CH COO), 4.41 (q, 4H,
2 2
NMP and CH Cl and collected by filtration using a medium frit. Ligand
3
2
2
COOCH
Calcd for C38
2
), 2.79 (t, 4H, CH
2
COO), 2.51 (t, 4H, PhCH
2
). Elem anal.
3a (0.182 g, 0.305 mmol), DMAP (0.075 g, 0.61 mmol), and DCC
(0.125 g, 0.61 mmol) were added to the FMOC-deprotected glycine
resin (0.181 g, 0.61 mmol) in 25 mL of dichloromethane.
44 2 4 2
H N O P : C, 69.71; H, 6.77; N, 4.28. Found: C, 69.60;
Ph AA
þ
H, 6.80; N, 4.37. ESI MS: m/z 655 {(P2 N2 ) ] þ H} .
2
4
074
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
The solution was stirred at room temperature for 19 h, followed by
filtration using a medium frit and collection of the peptidyl resin. The
resin was rinsed with 70 mL of dichloromethane to remove unreacted
increments until the catalytic current stopped increasing. Each measure-
ment was repeated multiple times for statistical purposes.
Dependence of the Catalytic Current on the Catalyst Concentra-
31
1
DMAP and DCC. Solid-state P{ H}: δ ꢀ40.8 (s), ꢀ34.4 (s). The
tion. A solution containing 20 mM p-cyanoanilinium was prepared in
þ
OTf /acetonitrile. Aliquots of a 2 mM stock solution of
4
ꢀ
resin obtained was suspended in 30 mL of acetonitrile, and [Ni(CH -
0.2 M NEt
catalyst 4a containing 20 mM p-cyanoanilinium prepared in 0.2 M
3
CN) ](PF ) (0.181 g, 0.305 mmol) was then added. The reaction
6
6 2
þ
ꢀ
mixture was stirred at room temperature for 5 days, resulting in a red
resin. The light-red solution was filtered using a medium frit, and the red
solid was collected. To cleave the complex from the resin, trifluoroacetic
acid (4.75 mL), triisopropylsilane (0.125 mL), and water (0.125 mL)
were added to the solid and the resulting solution was stirred for 4 h. The
resulting reaction mixture was filtered, and the volume of the filtrate was
reduced to 1 mL under vacuum. The product was then flash-precipitated
by addition of this solution to 60 mL of stirring diethyl ether. Yield: 0.34
NEt
catalytic current i
4
OTf /acetonitrile were added to 1.0 mL of the first solution. The
c
was measured at ꢀ0.94 V. A plot of the catalytic
current versus the catalyst concentration was used to determine the
order with respect to the catalyst.
Controlled-Potential Coulometry. A three-necked flask having
a total volume of 147 mL was used for bulk electrolysis experiments. A
cylinder of reticulated vitreous carbon (working electrode) was intro-
duced in a flask via a rubber septum. The reference and counter elec-
trodes were silver and nickelꢀchromium wires, respectively, placed in
7 mm glass tubes terminating in Vycor fritted disks and filled with an
31
1
g, 0.21 mmol, 68%. Solid-state P{ H}: δ 21.5 (s), ꢀ13.4 (s). Solution-
3
1
1
1
state P{ H} NMR (CD
NMR (CD
3
CN): δ 4.04 (s), 22.14 (s), ꢀ13.7 (s). H
CN); δ 7.39, 7.26, 7.14 (m, 36H, ArꢀH), 7.98 (s, 4H, NH),
electrolyte solution (0.2 M Et NBF in acetonitrile). The cell was
4 4
3
4
2
1
.24 (d, 8H, PCH
.90 (t, 8H, CH COO), 2.63 (t, 8H, PhCH
627 {[Ni(P
2
N), 3.76 (d, 8H, COCH
2
N), 3.98 (d, 8H, PCH
). ESI MS (CH CN): m/z
)} . Note: Dichloromethane was run
2
N),
purged with dinitrogen for 20 min prior to the experiment. A 25 mL
þ
2
2
3
solution of metal complex 4c (1 mM) and a 2 mL solution of DMFH
Ph DPA
þ
2
N
2
)
2
](PF
6
(50 mM) were added to the flask. Controlled-potential coulometry was
performed at ꢀ0.95 V versus the ferrocenium/ferrocene couple. After
17.6 C of charge was passed, a 300 μL sample of the headspace gas was
removed via a gastight syringe and analyzed by gas chromatography.
From this data, it was found that 4.6 mol of dihydrogen were produced
per 1 mol of catalyst and a current efficiency of 93% was calculated for
through basic alumina before use. If the catalysts were prepared in
reagent-grade dichloromethane, a chloride ion was incorporated because
of residual HCl, resulting in a small portion of a five-coordinate chloride
complex. This complex resulted in asymmetric resonances about the
central phosphorus resonance that were distinct from those observed for
n
þ
ꢀ
15
the dipeptides. A cyclic voltammogram recorded in Bu N OTf /
dihydrogen production, similar to other catalysts of this type. An
4
3
1
acetonitrile consisted of one reversible reduction wave at ꢀ0.84 V
I/0
aliquot of the catalytic solution was analyzed by P NMR spectroscopy,
and 80% of the original catalyst remained.
II/I
corresponding to the Ni couple and one quasi-reversible reduction
wave at ꢀ1.03 V corresponding to the Ni couple.
X-ray Crystallography. Crystals of 4a were grown by the slow
diffusion of ether into a MeCN solution of 4a at room temperature. A
Peptide Titration Studies. Metal complex 4b was titrated with
various concentrations of the ethyl ester of the glycylglycine dipeptide to
investigate interaction of the amide functionality with the nickel metal
center. Stock solutions of complex 4b (5 mL, 2.8 mM) and the ethyl
ester of the glycylglycine dipeptide (5 mL, 28 mM) were made in
methanol-d. The two solutions were mixed in order to get 1:1, 1:2, and
3
0.05 ꢁ 0.05 ꢁ 0.16 cm orange, block-shaped crystal was selected,
mounted on a MiTeGen MicroMounts pin using Paratone-N oil, and
cooled to the data collection temperature of 100(2) K. Data were
collected on a Br €u ker-AXS Kappa Apex II CCD diffractometer with
0.710 73 Å Mo KR radiation. Data were measured using ω and j scans
of 0.5°/frame for 35 s. The final resolution was 0.72 Å. Cell parameters
31
1
1
:4 dipeptide-to-metal complex ratios. P{ H} NMR spectra of the
2
4
solutions were recorded, and all chemical shifts were externally refer-
enced to phosphoric acid.
were retrieved using Br €u ker APEXII software, raw data were integrated
2
5
using SAINTPlus, and absorption correction was applied using
2
6
Electrochemical Studies. All cyclic voltammetry experiments
were carried out on a computer-aided CHI 1100 A potentiostat under
SADABS. The structure was solved using the Patterson method and
2
refined by a least-squares method on F using the SHELXTL program
n
þ
ꢀ
27
a dinitrogen atmosphere in 0.2 M Bu
4
N OTf /acetonitrile solutions
package. The structure was solved in the space group Pnn2 (No. 34) by
at a scan rate of 50 mV/s. The electrochemistry of complex 4a was
investigated in benzonitrile. The working electrode used was a glassy
carbon disk, and the counter electrode was a glassy carbon rod. A silver
wire was used as a pseudoreference electrode. Ferrocene was used as an
internal standard, and all potentials are referenced to the ferrocenium/
ferrocene couple.
analysis of systematic absences and intensity statistics.
Ph AA
2þ
The [Ni(P2 N2 ) (CH CN)] cation is centered on a special
2
3
position with the halves of the cation related by symmetry. All atoms in
the main residue that were not part of the disorder modeling were
refined anisotropically. There was considerable disorder among the
nitrogen-bound phenyl groups. The disordered phenyl group composed
of atoms C15ꢀC20 was modeled by assigning two different orientations
to the phenyl ring about the axis through the C15 and C18 atoms (C16,
C17, C19, C20 and C16A, C17A, C19A, C20A, each having 50%
occupancy). In order to prevent abnormally short and long bond
distances, atoms C15, C16, C517, C18, C7A, and C8A were constrained
to a regular hexagon, d = 1.39 Å, and the bond distances for C15ꢀC16A,
C17AꢀC18, C15ꢀC20A, and C19AꢀC20A were restrained to the
same distances. Similarly, the disordered phenyl group composed of
atoms C26ꢀC31 was modeled by assigning two different orientations to
the phenyl ring about the axis through atoms C26 and C29 (C27, C28,
C30, C31 and C27A, C28A, C30A, C31A). The disordered atoms could
not be refined anisotropically and were thus refined isotropically. A
solvent ether molecule, occupancy 50%, is also present in the structure.
The bond distances of the solvent molecule were restrained to >1.4 Å to
Dependence of the Catalytic Current on the Acid Concentration.
Metal complexes were titrated with acetonitrile solutions containing
p-cyanoanilinium trifluoromethanesulfonate (p-cyanoanilinium) or the
þ
triflate salt of protonated dimethylformamide (DMFH ) in order to
determine the order of the reaction with respect to acid. A total of 1.0 mL
of a 0.7 mM solution of each metal complex in acetonitrile containing a
n
þ
ꢀ
0
Et
.2 M electrolyte ( Bu N OTf was used with p-cyanoanilinium, and
4
þ
ꢀ
þ
4
N BF
4
was used with DMFH ) was made. The solution was
syringe-filtered using a 1.6 μm poly(tetrafluoroethylene) syringe filter.
The solution was then purged with dinitrogen for 5 min, and the initial
II/I
p
voltammogram was measured. The cathodic current (i ) for the Ni
couple was recorded. Aliquots of acid were added via microsyringe in
0ꢀ100 μL increments until the catalytic current stopped increasing.
The cathodic current (i ) was corrected for dilution, and the catalytic
1
p
current (i
c
) at ꢀ0.94 V was measured. The ratios of i
c
/i
p
versus the acid
prevent abnormally short bond distances. Crystal data for
Ph AA
concentration were plotted to determine the order with respect to acid.
[Ni(P
2
N
2
2 3 4 2
) (CH CN)](BF ) are shown in Table S1 in the Sup-
At the end of the experiment, degassed water was added in 5 μL
porting Information.
4
075
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
31
Figure 1. Solid-state PNMRofligand3d (bottom) and metal complex 4d (top) before cleavage from the resin. Spinning side bands are indicated by asterisks.
3
1
1
Computational Methods. A gas-phase model for the complex
D O solutions. A P{ H} NMR spectrum recorded in this
2
Ph
Ph
2þ
1
was built starting from the crystal structure of the [Ni(P 2N
complex. The structure was relaxed using the conjugated gradient
) ]
solvent exhibited a single resonance as expected, and the H NMR
and mass spectral data are consistent with the proposed structure.
The corresponding ester, R = Et, was also prepared for comparison
3b). This ligand is soluble in common organic solvents and avoids
the more reactive carboxylic acid groups that might influence the
chemistry of the target nickel complexes. The P{ H} and H
NMR spectra, elemental analyses, and mass spectrometery data are
all consistent with the assigned structure of ligand 3b.
A second amino acid can be incorporated into the outer-
coordination sphere of the ligand 3a by reaction with glycine
ethyl ester, as shown in reaction (2), to form the glycine ethyl
ester derivative 3c. Glycine was chosen as the amino acid because
its lack of functional groups provides a simple platform to begin
2 2
15
2
8
scheme and the semiempirical PM6 force evaluation method, which
has been proven to reproduce with good accuracy geometries and, in
some cases, energies of metal complexes when compared to more
demanding ab initio methods (see, for instance, refs 29 and 30). All of
(
31
1
1
3
1
the calculations have been performed using the CP2K code.
’
RESULTS
Synthesis and Spectroscopic Studies. One of the major
objectives in this work was to develop a synthetic approach that
would permit the incorporation of the catalytically active
Ph Ph
2þ
[
Ni(P 2N ) ] core into a polypeptide matrix. To achieve
2 2
31
1
1
our studies. P{ H} and H NMR spectra and mass spectro-
metry data are all consistent with the assigned structure of 3c.
Similar attempts to prepare the acid from P2 N2
glycine instead of its ethyl ester, resulted in a red reaction
this objective, we used the unnatural amino acid 3-(4-aminophe-
nyl)propionic acid to prepare the cyclic diphosphine ligand 3a, as
shown in reaction (1) for R = H. The presence of the carboxylic
acid group allows the attachment of additional amino acids
Ph DPA
, using
31
1
1
mixture, and the P{ H} and H NMR spectra of the reaction
mixtures indicated a complex distribution of products.
or small peptides, while the presence of the phenyl substituent on
nitrogen should ensure the same electronic environment as the
Ph Ph
Ph Ph
2þ
P
2N ligand used in the [Ni(P 2N ) ] complex, a highly
2
15,20
2 2
active electrocatalyst for dihydrogen production.
Ligand 3a
To overcome the presence for what were apparently unwanted
side reactions and to determine if solid-phase synthesis methods
is not soluble in common organic solvents but is soluble in basic
4
076
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
Table 1. Selected Bond Distances and Angles for
Ph AA
[
Ni(P2 N2 ) (CH CN)](BF )
2 3 4 2
Bond Distance (Å)
Ni1ꢀP1
Ni1ꢀP2
Ni1ꢀN3
Ni1---N1
N3ꢀC35
2.2140(15)
2.2179(18)
2.003(9)
3.393(6)
1.168(13)
Bond Angle (deg)
P1ꢀNi1ꢀP2
P2ꢀNi1ꢀP2#1
P1ꢀNi1ꢀP1#1
N3ꢀNi1ꢀP1
N3ꢀNi1ꢀP2
Ni1ꢀN3ꢀC35
82.71(6)
130.35(12)
175.87(13)
92.06(6)
114.82(6)
180.000(2)
Figure 2. Crystal structure for 4a. The ellipsoids are drawn at 30%
probability, and the hydrogen atoms, counterions, and solvent molecules
have been omitted for clarity. Nickel is shown in green, carbon in gray,
oxygen in red, nitrogen in purple, and phosphorus in orange.
complexes 4a and 4b. The crystal structure for 4a (Figure 2) is
very similar to that for [Ni(P2 N2 ) (CH CN)](BF ) .
Ph Ph
15
2
3
4 2
commonly used in peptide synthesis would be useful, the
Selected bond distances and angles are given in Table 1. The
nickel atom is five-coordinate, bound to four phosphorus atoms
of two bidentate diphosphine ligands and an acetonitrile mole-
cule, and is best described as distorted trigonal bipyramidal. As
synthesis of ligand 3d was achieved by using glycine immobilized
3
2
23
on a Wang resin using FMOC chemistry, as shown in
31
reaction (3). The solid-state P NMR spectrum of the resin-
bound ligand (Figure 1) consists of two resonances at ꢀ34.4
and ꢀ40.8 ppm. These shifts are downfield of the fully cyclized
ligand observed for
Ph Ph
with [Ni(P2 N2 ) (CH CN)](BF ) , the two nitrogen atoms
2
3
4 2
in the rings in boat conformations are folded toward the nickel
center and thus positioned to interact with either a dihydrogen or
hydride bound to the nickel, while the other two rings are chair
conformations, with the nitrogen atom folded away from the
nickel.
Intermolecular hydrogen bonding is observed between the
carboxylic acid groups (Figure 3), with each cation hydrogen
bonding to four different cations via the four carboxylic acid
groups. Cations connected by hydrogen bonds form planes,
which interpenetrate with other hydrogen-bonding planes in the
three-dimensional structure. No intramolecular hydrogen bond-
ing was detected.
31
1
1
The P{ H} and H NMR, cyclic voltammetry, and mass
spectrometry data are consistent with the assigned structures of
31
1
the metal complexes. The P{ H} NMR of metal complexes 4a
and 4b in deuteroacetonitrile each displayed a single resonance at
5
.3 and 5.0 ppm, respectively (Figure 4), similar to the value
Ph Ph 14
reported for [Ni(P2 N2 ) (CH CN)](BF ) (4.82 ppm).
2
3
4 2
At ꢀ60 °C in deuteroacetone, two resonances were observed in
31
the P NMR spectra for both of these complexes (Figure 4).
These resonances are attributed to axial and equatorial phos-
phorus atoms expected for a five-coordinate trigonal-bipyramidal
complex with an equatorial acetonitrile ligand. At room tem-
perature, the rate of exchange of the axial and equatorial
phosphorus atoms (Berry pseudorotation) is fast, which averages
31
the P NMR resonances, resulting in a single observed reso-
nance. At reduced temperatures, the slowed exchange process
results in a static five-coordinate species with resolved chemical
shifts. This behavior has been observed previously for analogous
ligands 3aꢀ3c in solution (ꢀ50 ppm) but are consistent with the
solid-state NMR of the unbound ligand 3a (ꢀ35.0 ppm, data not
shown). The two resonances likely represent isomers of the
eight-membered ring, as observed previously for similar soluble
16
metal complexes.
31
For complex 4c, a more complex P NMR spectrum is
observed at room temperature with a broad singlet at 4.04
ppm and two broad resonances at þ21.6 and ꢀ13.6 ppm at
room temperature (Figure 5, left). The central resonance has a
chemical shift similar to the exchange-averaged peak seen in the
33ꢀ36
ligands prior to metalation.
The reaction of stoichiometric amounts of P N ligands 3a
2
2
and 3b with [Ni(CH CN) ](BF ) in acetonitrile for 24 h
3
6
4 2
resulted in the formation of red, moderately air-sensitive metal
4
077
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
Figure 3. Hydrogen bonding observed for crystals of 4a. Solvent molecules and anions have been omitted for clarity.
single amino acid modified catalysts 4a and 4b. The two addi-
tional resonances are symmetric about the central resonance.
Upon warming of the NMR solution to 60 °C, the peak at 4.0 ppm
became sharper and the peaks at 21.8 and ꢀ13.6 ppm broadened.
Using acetone as a solvent to extend the low-temperature range,
the sample temperature was reduced to ꢀ60 °C. At this tempe-
rature, the central resonance was unobservable and only the two
multiplets at þ23.5 and ꢀ12.0 ppm were observed. Additionally,
amide groups, it is possible that they are binding to the nickel as a
31
fifth ligand, causing the anomalous features in the P NMR
spectrum. To test this hypothesis, 4b was titrated with the ethyl
3
1
ester of the glycylglycine dipeptide and monitored by P NMR
(Figure 6). Complex 4b was used to avoid amide or acid
functionalities in the metal complex. Because of the limited
solubility of the dipeptide in acetonitrile, the experiment was
31
performed in methanol-d . The P NMR spectrum of complex
4
3
1
31
Pꢀ P EXSY at 20 °C (Figure S1 in the Supporting In-
4b displayed a single resonance at 6.2 ppm. Upon an increase in
the concentration of the dipeptide, two resonances at þ23.2
and ꢀ15.2 ppm grew and the central resonance at 6.2 ppm
disappeared, demonstrating that the presence of the dipeptide
can result in a similar restriction of motion. A similar experiment
was performed by adding ethyl acetate; however, in this case, no
formation) shows cross peaks between all three resonances,
indicating that the sites for the two symmetric resonances are
in exchange with the site of the central resonance as well as with
each other. This does not distinguish between intra- and inter-
molecular exchange (both would be observed), but it does
confirm that the three sites are able to exchange with each other.
Taken together, these data are most consistent with a single
complex exhibiting one or more exchange processes. These
exchange processes could include restricted Berry pseudorota-
tion or reversible dissociation of a fifth ligand at elevated tempe-
ratures and a static structure at low temperature. Although no
31
change was observed in the P NMR spectra, suggesting that the
ester carbonyl is not involved in the interaction. Modeling studies
demonstrated that, because of the rigidity and length of the
aromatic ring, neither the amide nor the carbonyl groups within
the same molecule are able to interact with the nickel center
(Figure 7), so a bimolecular interaction would be required to
bind an amide as a fifth ligand.
3
1
31
Pꢀ P COSY crosspeaks are observed at either ꢀ40 or þ30 °C
between the central resonance and the symmetric resonances (cross
peaks were observed between the two symmetric resonances), this is
not surprising because of the lack of resolved splitting.
Another origin of the restricted motion in 4c could be due to
hydrogen bonding. The modeling discussed above indicates that
intramolecular hydrogen bonding between peptide chains is
possible (Figure 7). The crystal structure of 4a provides evidence
that intermolecular hydrogen bonding between peptide chains is
also possible (Figure 3). Both of these methods provide support
for the interpretation that hydrogen bonding between peptide
chains could be present in these catalysts. Consequently, two
The variable-temperature NMR data suggest that 4c, the
dipeptide ester ligand, is undergoing exchange processes similar
to 4a and 4b, the amino acid ligands, but at higher temperatures.
This could be due to intra- or intermolecular hydrogen bonding
restricting Berry pseudorotation or the fact that the fifth ligand in
3
1
4
c is not acetonitrile, as it is for 4a and 4b.
interpretations describing the restricted motion observed by P
In previous studies, we have observed that amides can act as
NMR are consistent with the collective data: an amide functional
group from a second nickel complex could occupy the fifth-
coordination site in 4c, or intra- or intermolecular hydrogen
bonding between peptide chains could result in restricted Berry
pseudorotation.
2
þ
good ligands to these Ni complexes. For example, dissolving
Ph PhCF
3
[
Ni(P 2N
) ](BF ) (where PhCF is 4-CF C H ) in
2 2 4 2 3 3 6 4
dimethylformamide can result in displacement of the diphosphine
37
ligand by dimethylformamide. Because our complexes have
4
078
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
31
1
Figure 4. P{ H} NMR spectra of 4a (left) and 4b (right) at room temperature (bottom) and ꢀ60 °C (top) in acetone-d.
31
1
Figure 5. P{ H} NMR spectra of complexes 4c (left) and 4d (right) at room temperature (bottom), 60 °C in CD
acetone-d.
3
CN (middle), and ꢀ60 °C (top) in
Figure 7. Complex 4d generated by molecular modeling demonstrates
that an intramolecular fifth-coordination site of the amide or carbonyl
groups is not possible, but intramolecular hydrogen bonding between
peptide chains could occur.
3
1
1
Figure 6. Room temperature P{ H} NMR spectra of complex 4b,
showing the effect of an increase in the concentration of the glycylglycine
ethyl ester dipeptide. The ratio of the metal complex to the dipeptide is
shown on the left-hand side of the spectra (MC = metal complex; DP =
dipeptide glycylglycine ethyl ester).
symmetrical resonances also become broadened at 60 °C, in-
dicating an exchange process similar to that observed for 4c.
Electrochemical Studies. Complex 4b displays two reversible
one-electron reductions at ꢀ0.83 and ꢀ1.04 V versus the
ferrocenium/ferrocene couple at a scan rate of 50 mV/s in
acetonitrile (Figure 8, left side). These couples are assigned to
The metal complex 4d was prepared by adding [Ni(CH3-
CN) ](BF ) to the resin-immobilized ligand. The solid-state
6
4 2
NMR spectrum shows nearly complete incorporation of the
metal, as demonstrated by the two equivalent downfield reso-
nances (Figure 1, top). Upon cleavage from the resin, the
solution-state behavior is very similar to that of 4c. The P{ H}
NMR spectrum of 4d exhibits a broad singlet at 4.0 ppm at room
temperature and two small peaks symmetrically placed about the
central resonance at þ21.6 and ꢀ13.6 ppm (Figure 5). The two
II/I
I/0
the Ni and Ni reductions, respectively. The electrochem-
istry of this complex is consistent with that of the previously
31
1
Ph Ph
2þ
13,14
reported parent [Ni(P 2N ) ] complex.
The cyclic
2 2
II/I
I/0
voltammogram of 4a displays similar reversible Ni and Ni
couples at ꢀ0.82 and ꢀ1.02 V in benzonitrile. In acetonitrile, two
4
079
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
n
þ
ꢀ
Figure 8. Cyclic voltammograms of 0.65 M solutions of 4b (left) and 4c (right) in acetonitrile containing a 0.2 M Bu N OTf electrolyte using a scan
4
rate of 50 mV/s. Ferrocene was used as an internal standard.
irreversible one-electron waves at ꢀ0.82 and ꢀ0.98 V are
observed for 4aꢀ4d, respectively (Table 2), in the presence of
p-cyanoanilinium.
II/I
I/0
observed corresponding to Ni and Ni reductions, respec-
II/I
I/0
þ
39
tively. The irreversibility of the Ni and Ni couples of 4a can
Acetonitrile solutions of DMFH (pK 6.1) gave higher
a
0
be attributed to the insolubility of the Ni complex in acetonitrile,
turnover frequencies (450, 140, 1000, and 340 for 4aꢀ4d,
0
which precipitates on the electrode surface upon reduction to Ni
respectively) than those obtained with p-cyanoanilinium. The
and retards electron transfer during the oxidative sweep. The
turnover frequencies of these complexes are comparable to the
II/I
Ph Ph
cyclic voltammograms of 4c and 4d display reversible Ni
turnover frequency for the parent complex [Ni(P N2 )2]-
2
I/0
15,16,37
couples, but in contrast to 4a and 4b, quasi-reversible Ni
(BF ) measured under similar conditions.
Previous stud-
4
2
reductions are observed (Figure 8, right side, 4c).
ies have shown that catalytic rates of this class of molecules are
15
Catalytic Studies. Figure 9 shows successive cyclic voltam-
mograms of metal complexes 4aꢀ4d recorded in acetonitrile
first-order in catalyst, and we confirmed this for one of the
catalysts in this study (4a; see Figure S5 in the Supporting
Information).
38
with increasing concentrations of p-cyanoanilinium (pK 7.0).
a
A catalytic wave for the reduction of protons to dihydrogen is
observed with a half-wave potential near the Ni couples of the
The overpotential for the catalytic reduction of the protons
using p-cyanoanilinium was calculated from the half-wave po-
II/I
respective catalysts. At low acid concentrations, the catalytic
current increases as the acid concentration increases, until acid
concentrations above approximately 0.15 M, where the current is
independent of the acid concentration. Repeating selected
experiments in the absence of ferrocene yielded identical results.
tential of the catalyst and the pK value of p-cyanoanilinium
a
38
(7.0) using the method suggested by Evans and co-workers,
4
0
with E° = ꢀ140 mV. Similar results (within experimental
Hþ
error) were obtained using the method of Fourmond and co-
workers. The estimated overpotentials for catalytic reduction
41
Plotting the ratio of the observed catalytic current (i ) to the
of the protons of p-cyanoanilinium by 4aꢀ4d are 0.27, 0.27, 0.28,
c
II/I
þ
current of the Ni reduction in the absence of acid (i ) versus
and 0.28 V, respectively (Table 2), and using DMFH , they are
p
the concentration of acid results in a linear relationship at low
acid concentrations (Figure 10 and S2ꢀS4 in the Supporting
Information), suggesting that the reaction is second-order in acid
0.25, 0.25, 0.33, and 0.26 V, respectively. The calculated over-
0
R
R
2þ
2 2
potentials and catalytic rates for peptide-coupled [Ni(P N ) ]
2
complexes are comparable to the overpotential and catalytic rate of
13,15,17,19
Ph Ph 2þ
(
eq 4) below ∼0.04 M p-cyanoanilinium.
the parent [Ni(P2 N2 ) ] complex measured under similar
2
conditions.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
þ 2
Electrocatalytic Activity in the Presence of Water. Previous
studies in our group suggest an order of magnitude increase in
the rates as a function of small amounts of added water using
ic
2
RTk½H ꢂ
¼
ð4Þ
ip
0:4463
Fv
37
anilinium salts as the acid. The role of water on electrocatalysis
of these complexes is intriguing because the fixed proton relay
plays such an important role; additional relays, such as closely
associated water molecules, may also contribute to efficient
proton transfer. The influence of water on the rate of catalysis
was determined by recording cyclic voltammograms of metal
complexes in acetonitrile in the presence of p-cyanoanilinium or
At higher acid concentrations, no acid dependence was
observed, consistent with previous dihydrogen production cat-
1
3,14,20
alysts of this type.
The dependence on the acid concentra-
tion has previously been compared to MichaelisꢀMenten satura-
1
3
tion kinetics, where catalysis becomes independent of the
substrate under conditions of a large substrate excess (see the
Supporting Information, eqs S1ꢀS5). In this region, the rate is
þ
DMFH and varying the concentrations of water. Large in-
creases in the turnover frequencies were observed upon the
16,20
equivalent to the release of hydrogen, the rate-limiting step.
Turnover frequencies were calculated from the acid-independent
addition of water (Figure 11) in the presence of p-cyanoanili-
1
3
region using eq 5.
þ
nium (5ꢀ7-fold) compared to DMFH (2-fold).
rffiffiffiffiffiffiffiffi
ic
2
RTk
¼
ð5Þ
’
DISCUSSION
ip
0:4463 Fv
Enzymes are capable of shuttling gases, protons, and electrons
42,43
In this acid-independent region, turnover frequencies of 14, 15,
3, and 25 mol of dihydrogen per 1 mol of catalyst per second are
with great speed and precision.
Enzymes are also capable of
2
very specifically controlling the local environment around the
4
080
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
electronic character. Furthermore, enzymes are dynamic mol-
ecules that can control the desired environment by subtle or
significant changes in structure. Like other enzymes, hydroge-
nase enzymes use many of these outer-coordination-sphere
þ
1
features to very efficiently convert H to H and back again.
2
0
R
R
The Ni(P N )2 functional mimics of hydrogenase achieve
2
2
rates that approach those of NiFe hydrogenase but are still an
1
5,16
order of magnitude slower than those of FeFe hydrogenase.
0
Mimicking the enzymes placement of a pendant amine in the
second-coordination sphere was essential in achieving the cur-
20
rent rates. Meeting or exceeding the rates of the enzyme may
require mimicking of not only the second-coordination sphere
but also the outer-coordination sphere of the enzyme. By
building on these active functional hydrogenase mimics, we want
to investigate several of these aspects (proton channels, the local
environment, and dynamics) of the outer-coordination sphere in
a stepwise manner to develop an understanding of their con-
tribution to catalysis, as well as their interdependence upon one
another.
Previously, we have attached peptides and polymers to rho-
dium metal centers to capture the regulatory features of
44,45
enzymes.
This effort was aimed at understanding how the
reactivity of the metal can be controlled by applying a stimulus to
special stimulus-responsive ligands, mimicking the dynamic and
regulatory role that outer-coordination spheres of enzymes play.
Other groups have also pursued gaining the advantage of some of
4
6ꢀ49
50ꢀ56
the features of enzymes.
Chirality and selectivity
and
57,58
active-site pockets have been a common focus, but there have
been other studies investigating allostery,
mers for structural and functional control
50,59
the use of dendri-
and catalyst reco-
60,61
62ꢀ64
very.
These approaches often, but not always, utilize amino
acids and peptides in their design because of their inherent
chirality and the flexibility with the functional group and struc-
tural design. More complex approaches have modified existing
6
5ꢀ69
70,71
enzymes
talated
or designed enzymes de novo,
both nonme-
7
2ꢀ74
75ꢀ77
and metalated.
In contrast, our work focuses
on the development of new redox-active catalysts that oxidize
and produce dihydrogen, utilizing the attributes of the enzyme
scaffold as necessary but specifically avoiding reconstruction of
the entire enzyme.
The addition of amino acids and small peptides to the
0
R
R
2þ
2 2
[
Ni(P N ) ] catalyst demonstrates two very important
2
advancements in our understanding: (1) Amino acids and peptides
can be incorporated into this catalyst core, allowing a foundation
for larger peptides that could begin to take advantage of the larger
framework used by enzymes, allowing the inclusion of more
sophisticated and controlled structures around the active site. (2)
The active site is sensitive to very subtle changes far from the
active site, suggesting that the outer-coordination sphere can
have a large impact on the catalytic rates. In the design of mole-
cular catalysts, it is now well recognized that changes in the first-
and second-coordination spheres can have dramatic impacts on
20
catalysis. The results presented in this paper clearly demon-
strate that modifications at even greater distances (at least in
terms of the number of bonds) can also have significant impacts
on the catalytic rates. In this sense, these small molecules are
reproducing effects that have been proposed to be important in
enzymes.
To build hydrogenase mimics that incorporate features of the
outer-coordination sphere, the development of a framework into
which peptides can be incorporated was a necessary first step.
The current work has demonstrated this capability by utilizing a
Figure 9. Successive cyclic voltammograms of complexes 4aꢀ4d shown from
top to bottom as a function of increasing concentrations of p-cyanoanilinium.
active site. Using the large range of functional groups available
with the 20 naturally occurring amino acids, and the structural
control of the protein architecture, precise placement of hydro-
phobic or hydrophilic groups can be achieved to control
the active-site hydrophobicity, charge, steric accessibility, or
4
081
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
Table 2. Turnover Frequencies and Overpotentials of Metal Complexes 4aꢀ4d in Acetonitrile
ꢀ1
turnover frequency (s ; (25%; acid,
þ
ꢀ
-1
4
-CNAnH OTf ; electrolyte,
turnover frequency (s ; (10%; acid,
n
þ
ꢀ
þ
ꢀ
þ
4ꢀ
Bu
4
N OTf )
DMFH OTf ; electrolyte, Et
4
N BF
)
metal complex
w/o water
w/water ([H
2
O])
overpotential (V; (70 mV) w/o water w/water ([H
2
O]) overpotential (V; (70 mV)
4
4
4
4
a
14
15
23
25
28
71 (0.86 M)
62 (0.43 M)
130 (1.4 M)
180 (1.4 M)
72 (1.1 M)
0. 27
0.27
0.28
0.28
0.30
450
140
900 (0.35 M)
200 (0.50 M)
1400 (0.65 M)
510 (0.50 M)
720 (0.03 M)
0.25
0.25
0.33
0.26
0.31
b
c
1000
340
d
Ph Ph
2þ
[
Ni(P
2
N
2
)
2
]
590
37
acid type. For instance, when using p-cyanoanilinium as the
ꢀ
1
acid, rates for the peptide catalysts are 14ꢀ25 s . These rates
þ
increase an order of magnitude or more when using DMFH ,
ꢀ1
where the observed rates are 140ꢀ1000 s . Catalytic rates are
also observed to show significant enhancement (1.4ꢀ7 times
faster) when water is added to the acidic solution. The difference
in rates with different acids has been addressed extensively in a
separate paper, and is thought to be related to the substrate
size; DMFH is a smaller acid than p-cyanoanilinium and,
37
þ
consequently, has better access to the crowded area around the
3
7
protonated amine. The effect of acid pK was also investigated
a
and found not to correlate with rate trends in a meaningful way.
The argument of the acid size can also explain the increased rates
observed in the presence of water. Note that, in the presence of
p-cyanoanilium, metal complexes display a much larger rate
c
Figure 10. Plot of the ratio of the catalytic current (i ) to the peak
II/I
current of the Ni couple in the absence of the acid (i ) versus the
p
concentration of p-cyanoanilinium in acetonitrile. Conditions: 50 mV/s
n
þ
ꢀ
scan rate, 0.2 M Bu
4
N OTf /acetonitrile, 0.7 mM complex 4a, glassy
þ
enhancement with added water than in DMFH . It was pro-
carbon electrode.
posed that water facilitates the movement of the proton from the
exo to the endo position on the amine, the position required
for dihydrogen elimination. The effect is more enhanced for
p-cyanoanilium because its larger size decreases accessibility to
non-natural amino acid as a base ligand in the cyclooctane ring
reaction (1)], the carboxylic acid terminus of which allows
[
þ
condensation reactions with other amino acids and peptides. The
attachment of additional groups was demonstrated by the
addition of a second amino acid [reactions (2) and (3)]. This
was achieved using a homogeneous solution-state route in the case
of a protected carboxylic acid. However, to attach an unpro-
tected amino acid, solid-phase synthesis (SPS) was required to
avoid unwanted side reactions. This also allows easy separation of
reactants and products, as has previously been observed when using
the amine, a problem not as pronounced for the smaller DMFH .
Studies are ongoing to further develop the mechanism explaining
the catalytic impact that the acid size and water have on catalysis.
While the catalysts are similar to the parent complex, there are
clear effects of the outer-coordination sphere, even with the
relatively simple modifications introduced by the mono- and
dipeptides. Of note is that 4c is one of the fastest reported
þ
catalysts of this type using DMFH , in both the presence and
78
37
SPS of organometallic systems. The SPS route provided a clean
nickel complex and is the planned synthetic route for the construc-
tion of larger peptide-based catalysts. This approach not only avoids
the unwanted reaction of the carboxyl terminus but also avoids
reactions with functional groups of the side chains, groups that
are protected when bound to the resin and deprotected during
the cleavage step. Overall, the SPS approach results in a cleanly
produced metal complex and eliminates unwanted side reactions.
Importantly, this work demonstrates that the high electro-
catalytic activity of the parent complex is maintained in the
presence of bulky groups with multiple functionalities (Figure 9
and Table 2), a critical baseline to establish before the attachment
of larger peptides. For example, complexes 4a and 4d with
absence of water. At the same time, there are trends that are
difficult to explain. Electrocatalysis using p-cyanoanilinium as the
acid shows that the peptide catalysts have rates (∼0.5ꢀ1 times)
similar to that of the parent phenyl-substituted complex
(Table 2) in the absence of water and the rates increase with
increasing peptide size. However, the electrochemistry of the
þ
same catalysts using DMFH as the acid shows no clear
correlation between the ligand size (one or two amino acids)
or the functional group (acid or ester termination). Under these
conditions, the peptides range from ∼2 times faster (4d) than
the parent complex to ∼4 times slower (4b). The larger size due
to incorporation of a glycine unit for the two ester-terminated
catalysts 4b and 4c results in a 4ꢀ8-fold increase in the rate. This
is a surprising increase for a relatively simple modification and
is not observed for the corresponding acid-terminated catalysts
4a and 4d, which have rates very similar to each other and
demonstrate the opposite trend with respect to the additional
glycine unit; i.e., 4d is slightly slower than 4a when compared to
the ester derivatives.
ꢀ
1
catalytic rates of 450 and 340 s , respectively, are quite com-
Ph Ph
2þ
ꢀ1
parable to [Ni(P 2N ) ] with 590 s . Similar overpoten-
2 2
tials are also observed, indicating that, on the whole, the peptide
catalysts behave similarly to the parent complex.
Another similarity between the peptide catalysts and the
parent catalyst is the strong dependence on water and on the
4
082
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
’
SUMMARY
In order to more fully understand the contribution of the
outer-coordination sphere to the catalytic activity of hydrogenase
enzymes, functional mimics containing short peptides in the
outer-coordination sphere have been synthesized, characterized,
and studied. Mono- and dipeptides were introduced to a
Ph Ph
2þ
previously reported complex [Ni(P2 N2 ) ] , which has
2
ꢀ
1
been shown to display a high catalytic rate (590ꢀ720 s ) and
þ
a moderate overpotential (310 mV) using DMFH as the acid.
These peptide-containing complexes are moderately air-stable
and can be prepared in good yields. All four complexes are active
catalysts for the reduction of protons to dihydrogen, ranging
from ∼4 times slower to ∼2 times faster than the parent catalyst.
The observed differences in the catalytic rates for these com-
plexes with the introduction of minor modifications in the outer-
coordination sphere indicate that the outer-coordination sphere
can significantly enhance or decrease the rate of catalysis compa-
red to the parent complex. Of particular note, complex 4c
exhibits rates 4ꢀ8 times that of 4b, which differ only by the
presence of a glycine unit, with a dramatic change in rate resulting
from a small structural change. In addition, the introduction of
short peptides in the nickel-based hydrogen production electro-
catalysts establishes a framework for extension of this work to
larger peptides to more closely mimic enzyme structures.
’
ASSOCIATED CONTENT
Supporting Information.
31
31
S
Pꢀ P EXSY data, plots of
b
i /i versus acid concentration for 4bꢀ4d, derivation of Michae-
c
p
lisꢀMenten kinetics, linear dependence on the catalyst concen-
tration, and a CIF file for the crystal structure of 4a. This material
is available free of charge via the Internet at http://pubs.acs.org.
’
AUTHOR INFORMATION
Corresponding Author
*
E-mail: wendy.shaw@pnl.gov.
’
ACKNOWLEDGMENT
We thank Dr. John Roberts for useful discussion. This work
was supported by the U.S. Department of Energy Basic Energy
Sciences, Chemical Sciences, Geosciences & Biosciences Divi-
sion. The Pacific Northwest National Laboratory is operated by
Battelle for the U.S. Department of Energy.
Figure 11. Cyclic voltammograms of 4aꢀ4d in the presence of
p-cyanoanilinium (orange trace) and water (green trace).
’
REFERENCES
Restricted motional processes for the dipeptide metal com-
plexes, 4c and 4d, appear to arise from intermolecular interac-
tions of the dipeptide functional groups. However, the observed
restricted motion does not appear to be correlated with electro-
(
1) Fontecilla-Camps, J. C.; Volbeda, A.; Cavazza, C.; Nicolet, Y.
Chem. Rev. 2007, 107, 4273–4303.
2) Frey, M. ChemBioChem 2002, 3, 153–160.
(
(3) Nicolet, Y.; de Lacey, A. L.; Vernede, X.; Fernandez, V. M.;
Hatchikian, E. C.; Fontecilla-Camps, J. C. J. Am. Chem. Soc. 2001,
123, 1596–1601.
3
1
catalysis. For instance, 4c and 4d both exhibit the P NMR
spectrum, indicating restricted motion; however, one is faster
and one slower than the single amino acid analogues, where
restricted motion is not observed. It is likely that, as the outer-
coordination sphere includes more complex side chains (polar,
charged, acidic, and basic) and gets larger, additional interactions
will be observed and the development of a mechanism will be
possible. Even within the limited set of data presented in this
manuscript, it is clear that apparently subtle changes in the outer-
coordination sphere can significantly modulate the catalytic
activity.
(
4) Peters, J. W.; Lanzilotta, W. N.; Lemon, B. J.; Seefeldt, L. C.
Science 1998, 282, 1853–1858.
5) Shima, S.; Pilak, O.; Vogt, S.; Schick, M.; Stagni, M. S.;
Meyer-Klaucke, W.; Warkentin, W.; Thauer, R. K.; Ermler, U. Science
008, 321, 572–575.
6) Volbeda, A.; Charon, M.-H.; Piras, C.; Hatchikian, E. C.; Frey,
M.; Fontecilla-Camps, J. C. Nature 1995, 373, 580–587.
7) Volbeda, A.; Garcin, E.; Piras, C.; de Lacey, A. L.; Fernandez,
(
2
(
(
V. M.; Hatchikian, E. C.; Frey, M.; Fontecilla-Camps, J. C. J. Am. Chem.
Soc. 1996, 118, 12989–12996.
4
083
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
(
8) Gloaguen, F.; Rauchfuss, T. B. Chem. Soc. Rev. 2009, 38,
00–108.
9) Henry, R. M.; Shoemaker, R. K.; DuBois, D. L.; Rakowski
DuBois, M. J. Am. Chem. Soc. 2006, 128, 3002–3010.
10) Henry, R. M.; Shoemaker, R. K.; Newell, R. H.; Jacobsen, G. M.;
DuBois, D. L.; Rakowski DuBois, M. Organometallics 2005, 24,
481–2491.
11) Jacobsen, G. M.; Shoemaker, R. K.; McNevin, M. J.; Rakowski
DuBois, M.; DuBois, D. L. Organometallics 2007, 26, 5003–5009.
12) Jacobsen, G. M.; Yang, J. Y.; Twamley, B.; Wilson, A. D.;
Bullock, M.; Rakowski DuBois, M.; DuBois, D. L. Energy Environ. Sci.
(38) Appel, A. M.; Lee, S.-J.; Franz, J. A.; DuBois, D. L.; Rakowski
DuBois, M.; Twamley, B. Organometallics 2009, 28, 749–754.
(39) Izutsu, K. AcidꢀBase Dissociation Constants in Dipolar Aprotic
Solvents; Blackwell Scientific Publications: Oxford, U.K., 1990.
(40) Felton, G. A. N.; Glass, R. S.; Lichtenberger, D. L.; Evans, D. H.
Inorg. Chem. 2006, 45, 9181–9184.
(41) Fourmond, V.; Jacques, P.-A.; Fontecave, M.; Artero, V. Inorg.
Chem. 2010, 49, 10338–10347.
(42) Creighton, T. E. Proteins; W. H. Freeman and Company: New
York, 1996.
(43) Ragsdale, S. W. Chem. Rev. 2006, 106, 3317–3337.
(44) Shaw, W. J.; Chen, Y.; Fulton, J.; Linehan, J.; Gutowska, A.;
Bitterwolf, T. J. Organomet. Chem. 2008, 693, 2111–2118.
(45) Shaw, W. J.; Linehan, J. C.; Gutowska, A.; Newell, D.; Bitterwolf,
T.; Fulton, J. L.; Chen, Y.; Windisch, C. F. Inorg. Chem. Commun. 2005,
8, 894–896.
(46) Loganson, A. A. Russ. Chem. Rev. 1985, 54, 277–292.
(47) Wohrle, D. Angew. Chem., Int. Ed. 2005, 44, 7500–7502.
(48) Severin, K.; Bergs, R.; Beck, W. Angew. Chem., Int. Ed. 1998,
37, 1634–1654.
(49) Ogawa, M. Y.; Gretchikhine, A. B.; Soni, S.-D.; Davis, S. M.
Inorg. Chem. 1995, 34, 6423–6424.
(50) Gianneschi, N. C.; Bertin, P. A.; Nguyen, S. T.; Mirkin, C. A.;
Zakharov, L. N.; Rheingold, A. L. J. Am. Chem. Soc. 2003, 125,
10508–10509.
1
(
(
2
(
(
2
008, 1, 167–174.
13) Pool, D. H.; DuBois, D. L. J. Organomet. Chem. 2009,
94, 2858–2865.
14) Wilson, A. D.; Fraze, K.; Twamley, B.; Miller, S. M.; DuBois,
D. L.; Rakowski DuBois, M. J. Am. Chem. Soc. 2008, 130, 1061–1068.
15) Wilson, A. D.; Newell, R. H.; McNevin, M. J.; Muckerman, J. T.;
Rakowski DuBois, M.; DuBois, D. L. J. Am. Chem. Soc. 2006,
28, 358–366.
16) Wilson, A. D.; Shoemaker, R. K.; Miedaner, A.; Muckerman,
J. T.; DuBois, D. L.; DuBois, M. R. Proc. Natl. Acad. Sci. U.S.A. 2007,
04, 6951–6956.
17) Yang, J. Y.; Bullock, R. M.; Shaw, W. J.; Twamley, B.; Fraze, K.;
Rakowski DuBois, M.; DuBois, D. L. J. Am. Chem. Soc. 2009,
31, 5935–5945.
18) Curtis, C. J.; Miedaner, A.; Ciancanelli, R.; Ellis, W. W.; Noll, B. C.;
Rakowski DuBois, M.; DuBois, D. L. Inorg. Chem. 2003, 42, 216–227.
19) Fraze, K.; Wilson, A. D.; Appel, A. M.; Rakowski DuBois, M.;
DuBois, D. L. Organometallics 2007, 26, 3918–3924.
20) Rakowski DuBois, M.; DuBois, D. L. Chem. Soc. Rev. 2009,
8, 62–72.
21) Leger, C.; Jones, A. K.; Roseboom, W.; Albracht, S. P. J.;
Armstrong, F. A. Biochemistry 2002, 41, 15736–15746.
22) Le Goff, A.; Artero, V.; Jousselme, B.; Tran, p. D.; Guiller, N.;
Metaye, R.; Fihri, A.; Palacin, S.; Fontecave, M. Science 2009,
26, 1384–1387.
23) Meienhofer, J.; Waki, M.; Heimer, E. P.; Lambros, T. J.;
Makofske, R. C.; Chang, C.-D. J. Pept. Protein Res. 19779, No. 13, 35–42.
24) Br €u ker APEX 2, version 2010.3-0; Br €u ker AXS Inc.: Madison,
WI, 2009.
(
6
(
(
1
(
1
(
1
(
(51) Knor, G. ChemBioChem 2001, 7/8, 593–596.
(52) Breit, B.; Seiche, W. Angew. Chem., Int. Ed. 2005, 44, 1640–1643.
(53) Lordi, V.; Yao, N.; Wei, J. Chem. Mater. 2001, 13, 733–737.
(54) Miller, S. J. Acc. Chem. Res. 2004, 37, 601–610.
(55) Roelfes, G.; Feringa, B. L. Angew. Chem., Int. Ed. 2005,
44, 3230–3232.
(
(
3
(
(56) Erickson, L. E.; Bailey, P. D.; Kimball, T. L.; Morgan, B. R. Inorg.
Chim. Acta 2003, 346, 169–180.
(
(57) Slagt, V. F.; Reek, J. N. H.; Kamer, P. C. J.; Leeuwen, P. W. N.
M. V. Angew. Chem., Int. Ed. 2001, 40, 4271–4274.
(58) Kersting, B. Z. Anorg. Allg. Chem. 2004, 630, 765–780.
(59) Berg, T. Angew. Chem., Int. Ed. 2009, 28, 2–5.
(60) Liang, C.; Frechet, J. M. J. Prog. Polym. Sci. 2005, 30, 385–402.
(61) Scott, R. W. J.; Datye, A. K.; Crooks, R. M. J. Am. Chem. Soc.
2003, 125, 3708–3709.
3
(
(
(
25) SAINTPlus: Data Reduction and Correction Program, version
(62) Bergbreiter, D. E. Chem. Rev. 2002, 102, 3345–3384.
(63) Dickerson, T. J.; Reed, N. N.; Janda, K. D. Chem. Rev. 2002,
102, 3325–3344.
(64) Madhavan, N.; Jones, C. W.; Weck, M. Acc. Chem. Res. 2008,
41, 1153–1165.
(65) Nakajima, H.; Ichikawa, Y.; Satake, Y.; Takatani, N.; Manna,
S. K.; Rajbongshi, J.; Mazumdar, S.; Watanabe, Y. ChemBioChem 2008,
9, 2954–2957.
7
2
23a; Bruker AXS Inc.: Madison, WI, 2004.
(
26) SADABS: an Empirical Absorption Correction Program, version
001/1; Bruker AXS Inc.: Madison, WI, 2001.
27) Sheldrick, G. M. SHELXTL: Structure Determination Software
Suite, version 6.14; Bruker AXS Inc.: Madison, WI, 2004.
(
(28) Stewart, J. J. Mol. Model. 2007, 13, 1173–1213.
(29) Correa, A.; Ragone, F.; Cavallo, L. Green metathesis chemistry:
great challenges in synthesis, catalysis and nanotechnology; Springer:
Amsterdam, The Netherlands, 2010.
(66) Park, H.-S.; Nam, S.-H.; Lee, J. K.; Yoon, C. N.; Mannervik, B.;
Benkovic, S. J.; Kim, H.-S. Science 2006, 311, 535–538.
(67) Qi, D.; Tann, C.-M.; Haring, D.; Distefano, M. D. Chem. Rev.
2001, 101, 3081–3111.
(68) Thomas, C. M.; Ward, T. R. Chem. Soc. Rev. 2005, 34, 337–346.
(69) Walter, K. U.; Vamvaca, K.; Hilvert, D. J. Biol. Chem. 2005,
280, 37742–37746.
(
30) Sorkin, D.; Truhlar, G.; Amin, E. A. J. Chem. Theory Comput.
009, 5, 1254–1265.
31) Vande Vondele, J. K. M.; Mohamed, F.; Parrinello, M.;
Chassaing, T.; Hutter, J. Comput. Phys. Commun. 2005, 167, 103.
2
(
(32) Wang, S.-S. J. Am. Chem. Soc. 1973, 95, 1328–1333.
(
33) Weidner, E.; Yang, J. Y.; Dougherty, W. G.; Kassel, W. S.;
(70) Nanda, V.; Koder, R. L. Nat. Chem. 2010, 2, 15–24.
(71) Lu, Y.; Yeung, N.; Sieracki, N.; Marshall, N. M. Nature 2009,
460, 855–862.
(72) Jiang, L.; Althoff, E. A.; Clemente, F. R.; Doyle, L.; Rothlisberger,
D.; Zahghellini, A.; Gallaher, J. L.; Betker, J. L.; Tanaka, F.; Barbas, C. F.,
III; Hilvert, D.; Houk, K. N.; Stoddard, B. L.; Baker, D. Science 2008,
319, 1387–1391.
(73) Rothlisberger, D.; Khersonsky, O.; Wollacott, A. M.; Jiang, L.;
DeChancie, J.; Betker, J.; Gallaher, J. L.; Althoff, E. A.; Zanghellini, A.;
Dym, O.; Albeck, S.; Houk, K. N.; Tawfik, D. S.; Baker, D. Nature 2008,
453, 190–197.
(74) Dwyer, M. A.; Looger, L. L.; Hellinga, H. W. Science 2004,
304, 1967–1971.
Bullock, R. M.; Rakowski DuBois, M.; DuBois, D. L. Organometallics
2
1
1
2
010, 29, 5390–5401.
34) Albrand, J. P. G. D.; Robert, J. B. Chem. Commun.
968, 1469–1470.
35) Albrand, J. P. G. D.; Picard, M.; Robert, J. B. Tetrahedron Lett.
970, 11, 4593–4596.
36) M €a rkl, V. G. J. G. Y.; Schoerner, C. Tetrahedron Lett. 1980,
1, 1409–1412.
37) Kilgore, U. J.; Roberts, J. A. S.; Pool, D. H.; Appel, A. M.;
(
(
(
(
Stewart, M. P.; Rakowski DuBois, M.; Dougherty, W. G.; Kassel, W. S.;
Bullock, R. M.; DuBois, D. L. J. Am. Chem. Soc. 2011, No. 133, XX,
accepted for publication.
4
084
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085

Inorganic Chemistry
ARTICLE
(
75) Kaplan, J.; DeGrado, W. F. Proc. Natl. Acad. Sci. U.S.A. 2004,
01, 11566–11570.
76) Ghadiri, M. R.; Soares, C.; Choi, C. J. Am. Chem. Soc. 1992,
14, 4000–4002.
77) Salgado, E. N.; Radford, R. J.; Tezcan, F. A. Acc. Chem. Res.
010, 43, 661–672.
78) Heinze, K.; Beckmann, M.; Hempel, L. Chem.—Eur. J. 2008,
4, 9468–9480.
1
1
2
1
(
(
(
4
085
dx.doi.org/10.1021/ic1025872 |Inorg. Chem. 2011, 50, 4073–4085