Dimeric FeFe-hydrogenase mimics bearing carboxylic acids: Synthesis and electrochemical investigation

Article abstract of DOI:10.1016/j.poly.2015.08.019

Hydrogen is expected to be important for future sustainable energy applications. Recent interest in adsorbing water splitting catalysts on a semiconductor surface for electrocatalytic hydrogen production warrants careful study of electrocatalysts with fun

Full text of DOI:10.1016/j.poly.2015.08.019

Polyhedron 103 (2016) 21–27
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Dimeric FeFe-hydrogenase mimics bearing carboxylic acids: Synthesis
and electrochemical investigation
Benjamin R. Garrett, Amneh Awad, Mingfu He, Kevin A. Click, Christopher B. Durr, Judith C. Gallucci,
⇑
Christopher M. Hadad, Yiying Wu
Department of Chemistry and Biochemistry, The Ohio State University, 100 West 18th Avenue, Columbus, Ohio 43210, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Hydrogen is expected to be important for future sustainable energy applications. Recent interest in
adsorbing water splitting catalysts on a semiconductor surface for electrocatalytic hydrogen production
warrants careful study of electrocatalysts with functional groups capable of binding to a surface. The
Received 12 June 2015
Accepted 20 August 2015
Available online 29 August 2015
monomeric complex Fe2(
dimeric complexes [{Fe2(
l
l
-S2(CH2)3)(CO)5PPh2C6H4CO2H (S2(CH2)3 = 1,3-propanedithiolate, pdt) and the
-pdt)(CO)5}2(
l, ,j
j1
1-Ph2PCH2N(Ar)CH2PPh2)] (Ar = p-CO2H-Ph, 3,5-CO2HPh,
Keywords:
p-CH2CO2HPh, m-CO2H-Ph) bearing carboxylic acid functional groups have been prepared. The first elec-
FeFe hydrogenase mimic
Electrochemical degradation mechanism
Surface immobilization
Catalysis
trochemical study of these dimeric FeFe complexes indicate that the complexes degrade upon reduction to
form the species [Fe2(
l-pdt)(CO)4(j
2-Ph2PCH2N(Ar)CH2PPh2)]. A route is proposed for the formation of
[Fe2( -pdt)(CO)4( l-pdt)(CO)5}2(l, ,j
l
j
2-Ph2PCH2N(Ar)CH2PPh2)] from [{Fe2( j1 1-Ph2PCH2N(Ar)CH2PPh2)]
and a crystal structure of an intermediate in the proposed electrochemical degradation pathway is
presented. The electrochemical study presented herein provides valuable insight into degradation
pathways of FeFe hydrogenase mimics bearing carboxylic acids amenable to surface immobilization.
Ó 2015 Elsevier Ltd. All rights reserved.
Water splitting
1. Introduction
Recently, there has been great interest in adsorbing well-stud-
ied, water splitting catalysts on a semiconductor surface. In 2012,
In 1 h the sun delivers enough energy to the earth’s surface to
sustain society’s energy demand for an entire year. Solar energy
is thus, the largest exploitable renewable energy resource [1].
The problem is energy generation from the sun is intermittent
and fairly unpredictable. Therefore, an efficient method for storing
this energy is essential for increased exploitation. Accordingly,
many research groups have focused on developing integrated solar
water splitting systems as a means to store solar energy chemi-
cally, in the form of H2 [2–10].
Ott and coworkers demonstrated light driven electron transfer
between a photosensitizer and a proton reducing catalyst using a
pentacoordinate iron complex as a functional model of the distal
iron in FeFe hydrogenases [7,18]. Using this work as motivation,
Sun and coworkers demonstrated visible light driven water split-
ting from a p-type dye-sensitized device [10]. Wu and coworkers
improved on this system by linking the dye directly to the catalyst
[2]. A carboxylic acid moiety was used as a group to bind to the
semiconducting surface in all works described above.
The development of solar water splitting systems has benefited
greatly from the work provided from the years of research on
hydrogenases [11]. FeFe hydrogenases provide inspiration and
knowledge on how an efficient proton reduction catalyst for water
splitting systems may be designed [10,12]. Mimics of the FeFe
hydrogenase active site have proved to be a promising class of
molecular electrocatalysts for proton reduction in the presence of
acids in non-aqueous media [11–14]. Understanding electron
transfer chemistry of synthetic models was fundamental in guiding
the design of newer, better systems for electrocatalytic hydrogen
production for application in water splitting systems [15–17].
The interest in immobilizing proton reducing catalysts onto
semi-conductor surfaces warrants careful study of electrochemical
processes of catalysts containing moieties capable of anchoring to
surfaces. While there has been a great amount of electrochemical
mechanistic studies on simpler FeFe hydrogenase models [12,15–
17], investigations into models containing anchoring groups is
lacking. Given the initial success of integrated solar water-splitting
systems [2,7], it is important to develop and understand electro-
chemical properties of catalysts with the ability to be immobilized
on a surface.
Complexes Fe2(
l-pdt)(CO)5PPh2C6H4CO2H (2) and [{Fe2(l-pdt)
(CO)5}2(
l, ,j
j1
1-Ph2PCH2N(Ar)CH2PPh2)] (Ar = p-CO2H-Ph (3), 3,5-
CO2HPh (4), p-CH2CO2HPh (5)) bearing carboxylic acid functional
groups amenable to surface immobilization have been prepared.
⇑
Corresponding author.
E-mail address: wu@chemistry.ohio-state.edu (Y. Wu).
http://dx.doi.org/10.1016/j.poly.2015.08.019
0277-5387/Ó 2015 Elsevier Ltd. All rights reserved.

22
B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
Table 1
The synthetic strategy outlined affords a family of FeFe hydroge-
nase mimetic catalysts bearing carboxylic acid functionalities for
anchoring on a semi-conductor surface. Cyclic voltammetry in N,
N-dimethyl formamide (DMF) solutions indicate that complexes
FTIR stretching frequencies for complexes in the carbonyl region and selected
reduction potentials (À0.5 V ? À2.0 V vs. NHE).
Complexes
m
(CO)(cmÀ1) (CH2Cl2)
FeI/Fe0 V vs. NHE(DMF)
3, 4, and 5 degrade upon reduction to form [Fe2(
l-pdt)(CO)4
1
2
2–1
3
3–1
4
2071, 2032, 1998
2046, 1981, 1960, 1925, 1713
2046, 1985, 1960, 1933
2046, 1981, 1960, 1926, 1712
2045, 1981, 1960, 1925
2045, 1982, 1958, 1925, 1745, 1707
2045, 1981,1959, 1927, 1747, 1709
2021, 1947, 1892
–
À1.13
(j
2-Ph2PCH2N(Ar)CH2PPh2)]. A route is proposed for the degrada-
À1.12
tion product and a crystal structure of an intermediate in the
proposed electrochemical degradation pathway is presented.
À1.14, À1.50
À1.20, À1.49
À1.21
5
6
À1.16, À1.5
À1.45
2. Materials and methods
2.1. General considerations
2.2. X-ray structure determination
All reactions and operations were carried out under a dry argon
atmosphere with standard Schlenk techniques. Solvents for
syntheses were HPLC-grade and further purified using an alumina
filtration system. All commercially available reagents (including
Fe3(CO)12, L1, HPPh2) were of ACS grade and used without further
purification. 1H NMR spectra were recorded on 400 MHz Bruker
spectrometers and are reported relative to TMS. 31P NMR spectra
were recorded on 400 MHz Bruker spectrometers and are reported
relative to H3PO4. High-resolution mass spectra were obtained on
Bruker MicroTOF-II mass spectrometer from the mass spectrome-
try facility at The Ohio State University. IR spectra were recorded
on a Perkin-Elmer 1600 FT-IR spectrometer.
Details of X-ray structure determination and crystallographic
data are summarized in the SI.
2.3. Electrochemistry
Cycling voltammetry measurements were performed using a
Gamry Reference 600 Potentiostat. All voltammograms were
obtained in a three-electrode cell under Ar atmosphere at room
temperature. The working electrode was a glassy carbon disk
(0.071 cm2), and a platinum mesh was used as the auxiliary
Scheme 1. Reaction scheme of 1–5.

B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
23
electrode. The experimental reference electrode was Ag/AgNO3,
2.5. Ph2PCH2N(p-CO2H-Ph)CH2PPh2 (L2)
and measured potentials are reported versus NHE, as determined
from recording with an internal reference of Cp2Fe/Cp2Fe+ = 0.69 V
versus NHE in DMF. The supporting electrolyte was 0.1 M nBu4NPF6
and the scan rate was 100 mV/s.
Yield: 0.9 g (65%). 1H NMR (d6-DMSO): d 12.15 (s, 1H), 7.36 (m,
13H), 7.32 (m, 7H), 7.00 (m, 2H), 6.78 (m, 2H), 3.94 (d, 4H), 3.41 (s,
2H). 31P NMR (CDCl3): d À26.8.
2.6. Ph2PCH2N(3,5-CO2H-Ph)CH2PPh2 (L3)
2.4. Representative synthesis of L2 and L3 [19]
Yield: 1 g (65%). 1H NMR (d6-DMSO): d 12.98 (s, 2H), 7.77 (m,
1H), 7.58 (m, 2H), 7.37 (m, 12H), 7.33 (m, 8H), 3.99 (m, 4H). 31P
NMR (CDCl3): d À28.3.
To an oven dried round bottom flask was added diphenylphos-
phine (0.93 mL, 5.4 mmol) and paraformaldehyde (150 mg,
5.4 mmol) under an atmosphere of Ar. The slurry was heated for
1 h at 110 °C to afford a clear, colorless liquid. After allowing to
cool, 10 mL of MeOH (dried over CaH2) and the amine
reagent (2.68 mmol, amine reagent = 4-aminobenzoic acid or
5-aminoisophthalic acid) was added to the flask to afford a yellow
solution. The solution was heated at reflux for 2 h under Ar, during
which time a solid precipitated out of solution. The solid was
collected by filtration and washed with MeOH and Et2O to afford
crystalline solid.
2.7. Ph2PCH2N(p-CH2CO2H-Ph)CH2PPh2 (L4) [19]
To an oven dried round bottom flask was added diphenylphos-
phine (0.93 mL, 5.4 mmol) and paraformaldehyde (150 mg,
5.4 mmol) under an atmosphere of Ar. The slurry was heated for
1 h at 110 °C to afford a clear, colorless liquid. After allowing to
cool, MeOH (10 mL, dried over CaH2) and 4-aminophenylacetic
Fig. 1. X-ray crystal structures of 3 and 4. Thermal ellipsoids are shown at 50% probability and hydrogen atoms, solvent molecules and disorder at the propane dithiolate
bridge are omitted for clarity.

24
B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
acid (2.68 mmol) was added to the flask to afford a clear, orange/
yellow solution. The solution was heated at reflux for 3 h before
solvent was removed by vacuum rotary evaporation to afford a
gummy off white/yellow solid. The solid was crystallized by layer-
ing hexanes over a concentrated ethanol solution to afford an off
white crystalline solid.
2.8. Ph2PCH2N(p-CH2CO2H-Ph)CH2PPh2 (L4)
Yield: 1 g (70%). 1H NMR (d6-DMSO): d 12.15 (s, 1H), 7.36 (m,
13H), 7.32 (m, 7H), 7.00 (m, 2H), 6.78 (m, 2H), 3.94 (m, 4H), 3.41
(s, 2H). 31P NMR (CDCl3): d À27.5.
2.9. Fe2(l-pdt)(CO)6 (1)
The following procedure is adapted from previous methods
[20]. A 50 mL Schlenk flask was charged with toluene (10 mL).
Under a strong flow of Ar was added Fe3(CO)12 (1.3 g, 2.6 mmol).
To this green solution was added 1,3-propanedithiol (0.29 mL,
2.9 mmol). The green solution was heated at reflux for 3 h, turning
red after 20 min. Solvent was removed by vacuum rotary evapora-
tion and the crude product was subjected to silica gel chromatog-
raphy (hexanes) to afford a red crystalline solid. Yield: 1.0 g (80%).
1H NMR (CDCl3): d 2.14 (br t, 4H, SCH2), 1.80 (m, CH2). IR (CH2Cl2):
m
CO = 2071, 2032, 1998 cmÀ1
.
Fig. 2. Cyclic voltammograms of
dichloromethane and N,N-dimethylformamide under an Ar atmosphere.
2
and 2–1 (1 mM) in 0.1 M nBu4NPF6 in
2.10. Representative synthesis of complexes 2–5
To round bottom flask was added 1 (100 mg, 0.259 mmol,
1.1 eq (2.2 eq for 3–5)) followed by trimethylamine N-oxide
hydrate (26.2 mg, 0.236 mmol (1:1 (CH3)3NO:Phosphine) and dry
CH3CN (10 mL). The flask was evacuated and refilled (Ar) and the
resulting dark red solution was allowed to stir for 5 min at RT
before PPh2C6H4CO2H (72.1 mg, 0.236 mmol) was added. The flask
was again evacuated and refilled (Ar) and the mixture was then
allowed to stir for 3 h at RT. The resulting suspension was diluted
with DCM, washed with 1 M HCl, water, and brine. The organic
layer was dried over Na2SO4, filtered and solvent removed by
vacuum rotary evaporation to afford a red solid. The solid was
dissolved in DCM and the crude product was subjected to silica
gel chromatography, eluting first with DCM to remove 1 and then
DCM:MeOH (9:1) to afford a red crystalline solid.
2.14. [{Fe2(
CH2PPh2)] (5)
l-pdt)(CO)5}2(l, , j
j1 1- Ph2PCH2N(p-CH2CO2HPh)
Yield: 110 mg (70%). 31P NMR (CDCl3): d 58.87. IR (CH2Cl2):
CO = 2044, 1978, 1922, 1710 cmÀ1. ESI-MS: m/z 1261.85 [MÀH]À.
m
2.15. [{Fe2(
l-pdt)(CO)5}2(j
2-(Ph2PCH2)2N(Ph)] (6)
31P NMR (CDCl3): d 53.07. IR (CH2Cl2):
mCO = 2021, 1947,
1892 cmÀ1
.
3. Results and discussion
3.1. Synthesis and characterization
Yield(2): 100 mg (65%). 31P NMR (CDCl3): d 65.7. IR (CH2Cl2):
m
CO = 2046, 1985, 1934, 1735 cmÀ1. ESI-MS: m/z 662.9 [MÀH]À.
Previous work has shown when (Ph2PCH2)2N(Ph) was used for
the CO-displacement of Fe2( -pdt)(CO)6 (1) in the presence of Me3-
NOÁ2H2O in CH3CN at room temperature, the dimeric complex
[{Fe2( -pdt)(CO)5}2(
1- Ph2PCH2N(Ph)CH2PPh2)] (6) was
Yield(2–1): 100 mg (60%). 31P NMR (CDCl3): d 64.9. IR (CH2Cl2):
l
m
CO = 2046, 1985, 1934, 1735 cmÀ1
.
l
l, , j
j1
2.11. [{Fe2(l-pdt)(CO)5}2(
l,
j1
,
j
1- Ph2PCH2N(p-CO2H-Ph)CH2PPh2)]
the sole product [21]. Furthermore, a family of ligands (L2–L4,
Scheme 1) bearing carboxylic acid functional groups has also be
described [19]. Combining these two synthetic strategies, two
equivalents of 1 in the presence of CO-removing reagent Me3-
NOÁ2H2O afforded the monosubstituted complexes 3, 4, and 5
(65–90%) (Scheme 1). All new complexes presented herein were
characterized by MS, IR and 31P NMR spectra. Furthermore, X-ray
crystal structures were obtained for complexes 3 and 4. The results
of the mass spectra for all complexes are in good agreement with
the previously reported diiron complexes [21].
(3)
Yield: 135 mg (90%). 31P NMR (CDCl3): d 59.2. IR (CH2Cl2):
CO = 2045, 1981, 1925, 1712 cmÀ1. ESI-MS: m/z 1247.84 [MÀH]À.
m
2.12. [{Fe2(l-pdt)(CO)5}2(
l, , j
j1 1- Ph2PCH2N(Ph)CH2PPh2)] (3–1)
Yield: 135 mg (90%). 31P NMR (CDCl3): d 59.2. IR (CH2Cl2):
m
CO = 2045, 1981, 1925, 1712 cmÀ1
.
Each complex displays three mCO stretches comparable to those
reported for analogous diiron complexes, along with an additional
2.13. [{Fe2(
(4)
l
-pdt)(CO)5}2(
l
,
j1
,
j
1- Ph2PCH2N(3,5-CO2HPh)CH2PPh2)]
stretch at ꢀ1700 cmÀ1 for the carboxylic acid moiety [21]. The
mCO
stretches of the monomeric complex 2 display a red shift from
25 cmÀ1 to 64 cmÀ1 in comparison with the all-CO diiron complex
1 indicating the strong donating character of the phosphine
Yield: 150 mg (80%). 31P NMR (CDCl3): d 58.6. IR (CH2Cl2):
CO = 2044, 1980, 1928, 1732 cmÀ1. ESI-MS: m/z 1291.82 [MÀH]À.
m
(L2). The mCO stretches of 2 shows very little difference with the

B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
25
Fig. 3. Cyclic voltammograms of complexes 3, 3–1, and 5, (1 mM) in 0.1 M nBu4-
NPF6 in DMF under an Ar atmosphere.
Fig. 5. Cyclic voltammograms of 3–1 and 6 (1 mM) in 0.1 M nBu4NPF6 in DMF under
an Ar atmosphere.
Fe2(
environment of the iron centers with PPh3 or PPh2(p-C6H4CO2H)
(L2, 2) as ligands. Similar changes in the CO bands as in 2 and
2–1 were observed for complexes 3–5. The stretching frequencies
of all complexes are summarized in Table 1, and the spectra are
presented in the SI.
l-pdt)(CO)5PPh3 (2–1) indicating little effect in electronic
3.2. Electrochemistry
m
The redox behavior of the diiron complexes in solution was
investigated using cyclic voltammetry (CV). CV data were collected
for 2–5 for comparison with the analogous unfunctionalized com-
plexes, 2–1 and 3–1. Redox potentials are summarized in Table 1.
Initial CVs were recorded in dichloromethane (DCM), as is typical
for CV measurements of FeFe hydrogenase mimics [23]. Due to
the low solubility of the carboxylic acid derivatives in DCM, follow
up experiments were carried out in DMF. A restricted region of the
CV was used in order to simplify the analysis of the reducing region
(À0.5 V ? À2.0 V versus NHE). When a wider range was used, CVs
matched other previously reported monosubstituted FeFe hydro-
genase mimics, and the full range CVs are presented in the SI [23].
CVs of the 2 and 2–1 in DCM and DMF show an irreversible
reductive process at À1.15 V versus NHE indicative of the reduc-
tion of a monophosphinated Fe center (Fig. 2) [23]. CVs of 3–5 in
The molecular structures of 3 and 4 were determined by X-ray
diffraction studies of single crystals, which are shown in Fig. 1.
Similar to that of previously reported complexes [21], the central
[2Fe2S] structure of all complexes has a butterfly framework, and
each iron atom is coordinated with a pseudosquare-pyramidal
geometry. The main framework of 3 and 4 are similar to each other
and to previously reported dimeric complexes [21]. Complexes 3
and 4 are of the type Fe2(SR)2(CO)5(PR3) with the formal oxidation
state of the Fe centers as +1. Complexes of this type were reported
as diamagnetic as early as 1973 [22]. The metal–metal bonding
was used to explain the diamagnetism and how to satisfy the 18-
electron rule on each iron. Furthermore, the Fe–Fe bond distances
in complexes 3 and 4 in our work were 2.502 Å and 2.511 Å respec-
tively, which is similar to the Fe–Fe bond distance in similar com-
plexes [12,21]. Therefore, given the bond distances and oxidation
state in our complexes, an Fe–Fe bond is shown in each of
our molecular structures. Crystallographic details of each complex
are outlined in the SI.
DMF also show
a similar irreversible reductive process at
À1.15 V versus NHE. However, a second, semi-reversible reductive
process at À1.48 V occurs in 3–5 (Fig. 3). Therefore, there is an
unexpected difference in redox chemistry between the complexes
2 and 3–5. An early study of Fe2(l-pdt)(CO)6 type complexes
demonstrated that CO substitution with P(OMe)3 occurs upon
reduction [24]. Furthermore, the large amount of literature on
-
-
S
S
S
S
CO
CO
OC
OC
OC
CO
Fe
Fe
Fe
Fe
N
OC
OC
Ligand Dissociation
CO
CO
S
S
PPh2 Ph2P
N
OC
OC
PPh2
PPh2
Fe
CO
Fe
OC
3-1
6
Fig. 4. Proposed reduced species observed in a selected region of the CV (À0.5 V ? À2.0 V vs. NHE).

26
B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
Fig. 6. Select regions of the IR spectra of 3–1 and 6. The spectra show the clear shift
of 6 relative to 3–1.
Fig. 8. X-ray crystal structures of I. Thermal ellipsoids are shown at 50% probability
and hydrogen atoms are omitted for clarity. The crystal structure of I is partially
oxidized at atom P(2), where oxygen atom O(8) is present 14% of the time. As a
result, this crystal consists of a mixture of oxidized (14%) and non-oxidized (86%) Fe
dimer molecules. Both structures are presented in the SI.
the electrochemistry of Fe2(pdt)(CO)5L and Fe2(pdt)(CO)4L2
(L = PX3–PPh3 for example) allowed us to hypothesize that the sec-
ond reduction peak may be the di-substituted complex 6 (Fig. 4)
[23]. We therefore synthesized 6 according to the reported proce-
dure and a CV was recorded in DMF [25]. When the CV of 3–1 and 6
are compared (Fig. 5), one can see the excellent agreement in the
second reductive process which indicates that the reductive degra-
dation of 3–1 to form 6 is underway.
showing that there is no 6 present in the starting material.
Therefore 6 is being formed in-situ.
A complex where the Ph2PCH2N(Ar)CH2PPh2 ligand binds in a
P–Fe–P (6 for example) fashion could be present in the starting
material. The presence of such a complex may result in the second
reduction wave at À1.45 V versus NHE, without going through a
degradative pathway as shown in Fig. 4. The IR stretching frequen-
cies of 6 show a significant red shift relative to IR stretching fre-
quencies of 2–5 (Table 1). No indicative peaks of bidentate
complexes (such as 6) are found in any of the dimeric structures.
As an example, Fig. 6 shows an IR spectra of 3–1 and 6, clearly
3.3. Degradative route
We hypothesized that the reductive degradation pathway from
3–1 to 6 may proceed through a route outlined in Fig. 7. Upon
reduction of an Fe center, dissociation of phosphorous on one of
the FeFe centers occurs to afford intermediate (I). Association of
the free phosphorous to the remaining FeFe center followed by
dissociation of a CO ligand affords 6, which is then reduced in solu-
tion (Fig. 7).
S
S
S
S
CO
CO
OC
OC
OC
CO
Fe
Fe
Fe
Fe
N
OC
OC
CO
CO
S
S
PPh
2 Ph2P
OC
PPh2
Fe PPh2
Fe
N
OC
OC
CO
e-
AS
(DMF or THF)
SLOW
n
6
F
T
o
i
t
a
i
c
o
3-1
s
s
A
d
n
-
a
g
i
L
)
1
O
C
-
)
2
S
S
S
S
S
S
CO
CO
OC
OC
OC
CO
CO
OC
Ligand Dissociation
CO
Fe
Fe
Fe
Fe
Fe
Fe
OC
OC
OC
OC
CO
CO
PPh2Ph2P
N
PPh2 PPh2
N
I
Fig. 7. Proposed degradative route of 3–1 to 6. The route is believed to be operating during the reduction of 3–5.

B.R. Garrett et al. / Polyhedron 103 (2016) 21–27
27
that the intermolecular bridged complexes degrade upon reduc-
tion to form [Fe2(CO)4(
2-Ph2PCH2N(Ar)CH2PPh2)]. A route is
j
proposed for this degradation product and a crystal structure of
an intermediate in the proposed electrochemical degradation path-
way is presented. The complexes presented in this work may
potentially be used in the design of heterogeneous hydrogen
evolution catalysts. Therefore, understanding the electrochemical
processes is important. Thus, the electrochemical degradation
pathway presented here may provide valuable insight in the future
designs of heterogeneous hydrogen evolution catalysts.
Acknowledgments
We acknowledge exceptional support, discussion and mentor-
ship from Zhongjie Huang, and Damian Beauchamp. We acknowl-
edge support from the U.S. Department of Energy, Office of Basic
Energy Sciences, Division of Materials Science and Engineering
under award DE-FG02-07ER46427.
Fig. 9. Cyclic voltammograms of 3–1 and 4 (1 mM) in 0.1 M nBu4NPF6 in DMF under
an Ar atmosphere.
Appendix A. Supplementary data
CCDC 1404584, 1404585, and 1404586 contains the supple-
mentary crystallographic data for complexes 3, 4, and I. These data
can be obtained free of charge via http://www.ccdc.cam.ac.
uk/conts/retrieving.html, or from the Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44
1223 336 033; or e-mail: deposit@ccdc.cam.ac.uk. Supplementary
data associated with this article can be found, in the online version,
at http://dx.doi.org/10.1016/j.poly.2015.08.019.
Evidence for a similar pathway, but without an applied potential
was found when growing crystals for these structures. We found
that when crystals were grown from a tetrahydrofuran (THF) solu-
tion (coordinating environment), two distinct crystals structures
were apparent. Red, square platelets and red needles were
obtained. The needle structures found were determined to be 3
(Fig. 1). The platelets were determined to be [Fe2(l-pdt)(CO)5
(j
1-Ph2PCH2N(p-CO2H-Ph)CH2PPh2)] (intermediate I, Fig. 8).
31P NMR confirmed that I was not present in the starting material.
The 31P NMR of all complexes synthesized may be found in the SI.
Again, no indicative peaks of I were found in any of the dimeric
structures.
References
[1] N.S. Lewis, D.G. Nocera, PNAS 103 (2006) 15729.
[2] Z. Ji, M. He, Z. Huang, U. Ozkan, Y. Wu, J. Am. Chem. Soc. 135 (2013) 11696.
[3] B. Kumar, M. Beyler, C.P. Kubiak, S. Ott, Chem. Eur. J. 18 (2012) 1295.
[4] S. Pullen, H. Fei, A. Orthaber, S.M. Cohen, S. Ott, J. Am. Chem. Soc. 135 (2013)
16997.
[5] F. Wen, X. Wang, L. Huang, G. Ma, J. Yang, C. Li, ChemSusChem 5 (2012) 849.
[6] F. Wang, W.-G. Wang, X.-J. Wang, H.-Y. Wang, C.-H. Tung, L.-Z. Wu, Angew.
Chem., Int. Ed. Engl. 50 (2011) 3193.
[7] J.M. Gardner, M. Beyler, M. Karnahl, S. Tschierlei, S. Ott, L. Hammarström, J. Am.
Chem. Soc. 134 (2012) 19322.
[8] G.F. Moore, I.D. Sharp, J. Phys. Chem. Lett. 4 (2013) 568.
[9] A. Krawicz, D. Cedeno, G.F. Moore, Phys. Chem. Chem. Phys. 16 (2014) 15818.
[10] L. Li, L. Duan, F. Wen, C. Li, M. Wang, A. Hagfeldt, L. Sun, Chem. Commun.
(Camb.) 48 (2012) 988.
The presence of I in a coordinating solvent suggests that the
first step of the degradative pathway to form I does not require
reduction to proceed. However, it has been demonstrated
previously that when the Fe center is reduced, the process of ligand
dissociation is more facile [24]. When crystals were grown from a
DCM:Ethanol mixture (50:1), where only a small portion of the sol-
vent is coordinating, only one crystal structure was apparent – red
needles, complex 3.
[11] M. Wang, L. Chen, X. Li, L. Sun, Dalton Trans. 40 (2011) 12793.
[12] C. Tard, C.J. Pickett, Chem. Rev. 109 (2009) 2245.
3.4. Improved stability of 4
[13] F. Gloaguen, J.D. Lawrence, T.B. Rauchfuss, J. Am. Chem. Soc. 123 (2001) 9476.
[14] J.R. McKone, S.C. Marinescu, B.S. Brunschwig, J.R. Winkler, H.B. Gray, Chem. Sci.
5 (2014) 865.
[15] I.A. De Carcer, A. DiPasquale, A.L. Rheingold, D.M. Heinekey, Inorg. Chem. 45
(2006) 8000.
[16] S.P. Best, S.J. Borg, J.M. White, M. Razavet, C.J. Pickett, Chem. Commun. (Camb.)
(2007) 4348.
[17] S.J. Borg, T. Behrsing, S.P. Best, M. Razavet, X. Liu, C.J. Pickett, J. Am. Chem. Soc.
126 (2004) 16988.
[18] M. Beyler, S. Ezzaher, M. Karnahl, M.-P. Santoni, R. Lomoth, S. Ott, Chem.
Commun. 47 (2011) 11662.
[19] S.E. Durran, M.R.J. Elsegood, N. Hawkins, M.B. Smith, S. Talib, Tetrahedron Lett.
44 (2003) 5255.
[20] D. Seyferth, G.B. Womack, M.K. Gallagher, M. Cowie, B.W. Hames,
Organometallics 6 (1987) 283.
[21] N. Wang, M. Wang, J. Liu, K. Jin, L. Chen, L. Sun, Inorg. Chem. 48 (2009) 11551.
[22] P.C. Ellgen, J.N. Gerlach, Inorg. Chem. 12 (1973) 2526.
[23] G.A.N. Felton, C.A. Mebi, B.J. Petro, A.K. Vannucci, D.H. Evans, R.S. Glass, D.L.
Lichtenberger, J. Organomet. Chem. 694 (2009) 2681.
Complexes 2, 3 and 5 all showed similar CV responses in the
reductive region À1.15 V and À1.48 V versus NHE (Fig. 3). Complex
4 does show a slight peak at À1.48 V, however the reductive peak
is significantly smaller (Fig. 9). The smaller second reductive peak
indicates that 4 may be more stable to ligand dissociation relative
to the other complexes. The improved stability may be because the
carboxylic acid moieties in the 3,5 positions may protect the iron
centers from the coordinating solvents, which is thought to initiate
the degradation [24]. The carboxylic acids may also contribute an
inductive effect, where the electron-withdrawing nature of the car-
boxylic acids contributes to better P–Fe interactions.
4. Conclusion
[24] A. Darchen, H. Mousser, H. Patin, J. Chem. Soc., Chem. Commun. (1988) 968.
[25] S. Lounissi, G. Zampella, J.F. Capon, L. De Gioia, F. Matoussi, S. Mahfoudhi, F.Y.
Pétillon, P. Schollhammer, J. Talarmin, Chem. A Eur. J. 18 (2012) 11123.
Complexes 2–5 bearing carboxylic acid functional groups have
been prepared. Cyclic voltammetry in a DMF solution indicates