Electron transfer and binding affinities in an electrochemically controlled ligand transfer system containing zinc porphyrin and a meso-phenylenediamine substituent

Article abstract of DOI:10.1039/c3dt52463a

Investigations on the transfer of the ligand, imidazole (HIm), between two covalently linked redox centres-zinc porphyrin and phenylenediamine (PD)-and the influence of the length of the linker are reported. Since the binding affinity of the ligand with zinc porphyrin is different from that of the ligand with the phenylenediamine moiety, the transfer of the ligand could be electrochemically controlled by adjusting the oxidation potentials. Changes in cyclic voltammograms and absorption spectra of the complexes revealed the site of ligand binding in the various oxidation states of the modified zinc porphyrins. Binding constants of the modified zinc porphyrins in various oxidation states were also determined by photometric titration with the ligand and digital simulations. Evidence for the delocalization of the electron from the zinc porphyrin to the phenylenediamine moiety and the influence of the delocalization on them were obtained from EPR studies. The Royal Society of Chemistry 2014.

Full text of DOI:10.1039/c3dt52463a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Electron transfer and binding affinities in an
electrochemically controlled ligand transfer
system containing zinc porphyrin and a
meso-phenylenediamine substituent†
Cite this: Dalton Trans., 2014, 43,
1424
Hsu Chun Cheng,a Peter Ping Yu Chen*b and Yuhlong Oliver Su*a
Investigations on the transfer of the ligand, imidazole (HIm), between two covalently linked redox
centres – zinc porphyrin and phenylenediamine (PD) – and the influence of the length of the linker are
reported. Since the binding affinity of the ligand with zinc porphyrin is different from that of the ligand
with the phenylenediamine moiety, the transfer of the ligand could be electrochemically controlled by
adjusting the oxidation potentials. Changes in cyclic voltammograms and absorption spectra of the com-
plexes revealed the site of ligand binding in the various oxidation states of the modified zinc porphyrins.
Binding constants of the modified zinc porphyrins in various oxidation states were also determined by
photometric titration with the ligand and digital simulations. Evidence for the delocalization of the
electron from the zinc porphyrin to the phenylenediamine moiety and the influence of the delocalization
on them were obtained from EPR studies.
Received 6th September 2013,
Accepted 24th October 2013
DOI: 10.1039/c3dt52463a
www.rsc.org/dalton
(ZnTMP) covalently linked to a phenylenediamine (PD).12 The
ligation of HIm to the zinc porphyrin centre was indicated in
Introduction
Porphyrins and metalloporphyrins have widespread appli- the neutral state. When the ZnTMP-PD was partially oxidized
cations in catalysis,1,2 electron-transfer systems3–8 and photo- (−2e), the electron deficient PD became the binding site for
electric devices.5–8 For the biological process involving HIm. Since the reported formation constant (K) of oxidized
metalloporphyrins, the respiratory process occurs in the mito- species with a ligand is ∼105 M−1 13–15
we proposed that upon
,
chondria resulting in the formation of adenosine tripho- further oxidation of ZnTMP-PD, HIm would likely revert back
sphate, while haemoglobin and myoglobin act as dioxygen to the zinc porphyrin moiety because the formation constant
carriers. These critical processes are popular areas of between the cation radical of zinc porphyrin and the ligand is
research.9–11 These systems involve multiple redox centres. On ∼107 M−1
the other hand, imidazole the functional side chain of histi-
.
We have now revisited the system containing zinc porphyrin
dine in enzymes and proteins acts as a ligand in several metallo- and phenylenediamine linked to each other via a covalent
proteins and influences their behaviour.9–11 Furthermore, bridge to determine the binding constant of HIm with each
the properties of a redox centre can be influenced by the proxi- moiety at various oxidation states and to study the influence of
mity of a second redox centre. Therefore, the properties related the covalent bridge (the distance between ZnTMP and PD) on
to redox dependent ligation of a ligand and intramolecular this phenomenon. The use of phenyl groups is a straight-
electron transfer in a system with multiple redox centres and forward and basic way to separate two redox centres. The inter-
multiple sites of ligation should be further examined and clari- action between one redox centre to another could be
fied. Our group has recently reported on an imidazole (HIm) considered as that of an electron-donating or -withdrawing
shuttling system ZnTMP-PD consisting of a zinc porphyrin group based on the change in the electron cloud distribution
after an electrochemical procedure. We have now synthesized a
series of zinc porphyrins, in which the bridge length is
aDepartment of Applied Chemistry, National Chi Nan University, 1, University Road,
Puli, Nantou, 545, Taiwan. E-mail: yosu@ncnu.edu.tw; Fax: (+886)49-2917956
bDepartment of Chemistry, National Chung Hsing University, 250, Kuo Kuang Road,
elongated by phenyl groups (Chart 1). Binding affinities were
determined by photometric titrations and evaluated by digital
simulations. The ligand transfer behaviour was monitored by
cyclic voltammograms (CVs) and optically transparent thin-
layer electrode (OTTLE) spectroelectrochemical methods.
Furthermore, details of the electron delocalization in these
Taichung, 402, Taiwan. E-mail: pychen@dragon.nchu.edu.tw;
Fax: (+886)4-22840411
†Electronic supplementary information (ESI) available: 1H NMR, 13C NMR
spectra, and electrochemical analysis. See DOI: 10.1039/c3dt52463a
1424 | Dalton Trans., 2014, 43, 1424–1433
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Chart 1 Structures of phenylenediamine (PD) and PD-substituted zinc porphyrins.
systems were obtained by electron paramagnetic resonance NMR and 13C NMR spectra were obtained using a Varian Unity
(EPR) studies.
Inova 300 WB spectrometer (Varian Assoc., Palo Alto, CA, USA).
EPR spectra were obtained using a Bruker Model EMX-10/12
spectrometer (Bruker Optic GmbH, Karlsruhe, Germany).
Experimental section
Synthesis
Chemicals for synthesis were obtained commercially from Compounds 1, 3, and 4 were synthesized in 53%, 50% and
ACROS (Geel, Belgium). Organic solvents were dried before 73% yields, respectively, following methods reported in the
use. Analytical grade tetra-n-butylammonium perchlorate literature.12,16
(TBAP) for use in electrochemical experiments was also
obtained from ACROS (Geel, Belgium) and recrystallized twice (3%–5% mol) of tetrakis-triphenylphosphine palladium and
from ethyl acetate and then dried in a vacuum prior to use. 3.0 mL of aqueous 2 M Na2CO3 were added to a solution of
4′-Bromo-4-triphenylbenzaldehyde (2). A catalytic amount
Electrochemistry was performed using a CHI Model 760 4-bromo-4′-iodobiphenyl (2.39 g, 6.67 mmol) in 20 mL of
series electroanalytical workstation (CH Instruments, Inc., toluene. A solution of 4-formylbenzeneboronic acid (1.00 g,
Texas, USA). Cyclic voltammetry was conducted with the use of 6.67 mmol) in 20 mL of ethanol was subsequently added, and
a three-electrode cell in which a BAS glassy carbon electrode the mixture was heated to reflux for 3 hours in an argon atmos-
(area = 0.07 cm2) was used as the working electrode. The glassy phere. After cooling, the mixture was extracted with dichloro-
carbon electrode was polished with 0.05 μm alumina on methane (three times) and the combined organic phase was
Buehler felt pads and was sonicated for 2 min in an ultra- washed with water and brine, dried, and evaporated under
sound bath to remove the alumina residue. The platinum wire vacuum. The residue was purified by chromatography to give 2
1
served as an auxiliary electrode and the reference electrode in 50% yield (1.12 g, white solid). H NMR (300 MHz, CD2Cl2)
was a home-made Ag/AgCl, KCl(sat’d) reference electrode. The δ(ppm) = 7.58 (4H, q), 7.74 (4H, q), 7.83 (2H, d), 7.97 (2H, d)
spectroelectrochemical cell was composed of a 1 mm cuvette, and 10.06 (1H, s); 13C NMR (75 MHz, CDCl3) δ(ppm) = 127.5,
a thin layer carbon gauze as a working electrode, a platinum 127.8, 128.1, 128.2, 128.4, 129.0, 129.2, 129.4, 129.5, 130.6,
wire as an auxiliary electrode, and an Ag/AgCl, KCl(sat’d) 132.5, 135.9, 140.3, 148.6. Anal. Calc. for C19H13BrO: C 67.67;
reference electrode. The environment of cyclic voltammetry H 3.89%; found: C 67.52; H 3.95% (Fig. S11 and S12, ESI†).16
and spectroelectrochemical experiment is CH2Cl2 containing
Zn(Ph2-Br)(mesityl)3P (5). Pyrrole (0.67 mL, 10.0 mmol),
the supporting electrolyte, tetrabutylammonium perchloride 4′-bromobiphenyl-4-carbaldehyde (0.65 g, 2.5 mmol), and mesityl-
(TBAP). The recorded CV was simulated by DigiSim (v 3.03, aldehyde (1.34 g, 10.0 mmol) were mixed in 1000 mL CHCl3
Bioanalytical Systems, Inc., West Lafayette, IN, USA). Absorp- under a nitrogen atmosphere for 10 min. BF3·OEt2 (0.85 mL,
tion spectra were measured using a HP-8453 UV-Vis spectro- 6.77 mmol) was added to this mixture. After stirring for about
photometer (Agilent Technologies, Santa Clara, CA, USA). 1H an hour, 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ, 1.75 g,
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 1424–1433 | 1425

View Article Online
Paper
Dalton Transactions
7.55 mmol) was added for 1 hour. Finally, triethylamine temperature for 15 min. The two solutions were mixed and
(0.46 mL, 3.30 mmol) was added for 30 min. The resulting pre- refluxed at 110–115 °C for 48 hours under a nitrogen
cipitate was filtered and washed with methanol, dissolved in a atmosphere. The reaction solution was filtered and then
minimum quantity of dichloromethane, and chromatographed chromatographed on silica gel (eluting with a solution of
on silica gel (eluting with a solution of dichloromethane) to dichloromethane–hexane = 1 : 3) to afford ZnTMP-Ph-PD in
1
afford a purple solid. This solid was a mixture of compounds 48% yield (45 mg, purple solid); H NMR (300 MHz, CD2Cl2) δ
that were difficult to separate over a column of silica. We (ppm) = 1.84 (18H, s), 2.62 (9H, s), 5.98 (1H, s), 6.98 (2H, t),
inserted zinc into the porphyrin product to enhance the 7.15 (2H, d), 7.23 (12H, q), 7.84 (2H, d), 7.97 (2H, d), 8.27 (2H,
polarity difference between the desired product and other by- d), 8.70 (4H, s), 8.75 (2H, d), and 8.97 (2H, d) (Fig. S10†); 13C
products and thereby assist in its purification. Complexation NMR (75 MHz, CDCl3) δ(ppm) = 21.9, 117.5, 118.0, 118.7,
with zinc was performed by mixing a solution of crude pro- 119.0, 120.0, 120.2, 122.3, 127.9, 128.4, 129.4, 130.8, 131.3,
ducts in CH2Cl2 with zinc acetate (1.0 g, 5.45 mmol) in MeOH 132.4, 133.2, 135.1, 137.6, 138.7, 139.2, 139.3, 139.5, 139.8,
for 3 hours, and the solution was monitored by UV-Vis spectro- 141.4, 143.4, 148.3, 149.9. Anal. Calc. for C77H64N6Zn: C 81.21;
scopy till the initial spectrum had completely changed. The H 5.66; N 7.38%; found: C 81.36; H 5.71; N 7.32%. HRMS
solution was extracted with water (3 times) and the concen- calcd for [C77H65N6Zn]+
= 1136.4512, found: 1136.4533
trated organic layer was purified by chromatography over silica (Fig. S17 and S18, ESI†).
gel in which the elution solution was dichloromethane–hexane
ZnTMP-Ph2-PD. A 30 mL toluene solution of compound 6
1
= 1 : 5 to afford 5 in 6.3% yield (152 mg, red-purple solid); H (50.00 mg, 0.048 mmol) and t-BuONa (23.20 mg, 0.241 mmol)
NMR (300 MHz, CD2Cl2) δ(ppm) = 1.84 (18H, s), 2.62 (9H, s), and another 30 mL toluene solution containing TPA-NH2
7.29 (6H, s), 7.72 (2H, d), 7.83 (2H, d), 7.97 (2H, d), 8.31 (2H, (62.87 mg, 0.242 mmol), Pd(OAc)2 (3.82 mg, 0.004 mmol), and
d), 8.71 (4H, s), 8.76 (2H, d), and 8.94 (2H, d); 13C NMR dppf (2.31 mg, 0.004 mmol) were stirred separately at room
(75 MHz, CDCl3) δ(ppm) = 21.9, 118.9, 119.1, 119.6, 122.0, temperature for 15 min. Under a nitrogen atmosphere, two
125.2, 127.8, 129.2, 130.9, 131.3, 132.3, 132.3, 135.2, 137.6, solutions were mixed and refluxed at 110–115 °C for 48 hours.
139.0, 139.2, 139.3, 139.5, 139.7, 140.1, 142.7, 149.9, 150.0, The reaction solution was filtered, and chromatographed on
150.1. Anal. Calc. for C59H49BrN4Zn: C 73.87; H 5.15; N 5.84%; silica gel (eluting with a solution of dichloromethane–hexane
found: C 74.12; H 5.23; N 5.88% (Fig. S13 and S14, ESI†).
= 1 : 3) to afford ZnTMP-Ph2-PD in 41% yield (24 mg, purple
Zn(Ph3-Br)(mesityl)3P (6). Pyrrole (0.67 mL, 10.0 mmol), solid); 1H NMR (300 MHz, CD2Cl2) δ (ppm) = 1.84 (18H, s),
4′-bromo-4-triphenylbenzaldehyde (0.65 g, 2.5 mmol), and mesi- 2.62 (9H, s), 5.98 (1H, s), 6.98 (2H, t), 7.15 (2H, d), 7.23 (12H,
tylaldehyde (1.34 g, 10.0 mmol) were mixed in 1000 mL CHCl3 q), 7.84 (2H, d), 7.97 (2H, d), 8.27 (2H, d), 8.70 (4H, s), 8.75
solution under a nitrogen atmosphere for 10 min. BF3·OEt2 (2H, d), and 8.97 (2H, d) (Fig. S10†); 13C NMR (75 MHz, CDCl3)
(0.85 mL, 6.77 mmol) was then added to this mixture. After δ (ppm) = 21.9, 117.6, 119.4, 120.3, 121.4, 122.6, 123.7, 125.5,
stirring for about an hour, 2,3-dichloro-5,6-dicyanobenzo- 126.0, 126.2, 127.2, 127.4, 128.2, 128.3, 129.7, 131.1, 131.6,
quinone (DDQ, 1.75 g, 7.55 mmol) was added for 1 hour. 132.7, 135.5, 138.1, 139.7, 140.1, 140.4, 142.5, 143.9, 148.6,
Finally, triethylamine (0.46 mL, 3.30 mmol) was added for 150.4. Anal. Calc. for C83H68N6Zn: C 82.06; H 5.64; N 6.92%;
30 min. The resulting precipitate was filtered and washed with found: C 82.32; H 5.71; N 6.98%. HRMS calcd [C77H65N6Zn]+ =
methanol, taken into a minimum of dichloromethane, and 1136.4512, found: 1136.4533 (Fig. S19 and S20, ESI†).
chromatographed on silica gel (eluting with a solution of
dichloromethane) to afford a purple solid which was a mixture
of products and other compounds. Complexation with zinc
and subsequent purification over silica gel conducted in
Results and discussion
Cyclic voltammograms
accordance with the procedure described in the purification of
1
compound 5 gave 6 in 7% yield (192 mg, red-purple solid); H To elucidate the electrochemistry of PD-substituted zinc por-
NMR (300 MHz, CD2Cl2) δ(ppm) = 1.86 (18H, s), 2.62 (9H, s), phyrins, the cyclic voltammetry of PD and ZnTMP were first
7.30 (6H, s), 7.62 (2H, s), 7.78 (2H, d), 8.01 (4H, t), 8.33 (2H, d), studied separately. Cyclic voltammograms (CVs) of both PD
8.73 (4H, s), 8.78 (2H, d), and 8.99 (2H, d); 13C NMR (75 MHz, and ZnTMP exhibited two well-defined one-electron reversible
CDCl3) δ(ppm) = 21.9, 118.8, 119.1, 119.8, 119.9, 121.9, 125.3, redox couples. The half potentials (E1/2) for PD were +0.59 and
127.3, 127.7, 127.9, 128.0, 128.9, 129.1, 129.2, 130.9, 131.4, +1.01 V while those for ZnTMP were +0.79 and +1.11 V.12,17
132.2, 132.3, 135.2, 137.6, 139.3, 139.5, 139.8, 140.5, 142.5, When PD and ZnTMP were titrated separately with HIm, their
149.9, 150.1. Anal. Calc. for C65H53BrN4Zn: C 75.40; H 5.16; N CVs responded differently. In the case of PD (Fig. 1(A)), the
5.41%; found: C 75.28; H 5.26; N 5.34% (Fig. S15 and S16, current of the first redox couple with an E1/2 of +0.59 V
ESI†).
increased at the cost of the second redox couple, which gradu-
ZnTMP-Ph-PD. A 30 mL toluene solution of compound 5 ally disappeared. This one-step two electron oxidation is inter-
(80.00 mg, 0.083 mmol) and t-BuONa (26.45 mg, 0.275 mmol) preted to result from the electrochemically induced hydrogen
and another 30 mL toluene solution containing TPA-NH2 bond formation between oxidized PD and HIm.17 Whereas,
(65.13 mg, 0.250 mmol), Pd(OAc)2 (3.82 mg, 0.004 mmol), and when ZnTMP (Fig. 2) was titrated with HIm, the first redox
dppf (2.31 mg, 0.004 mmol) were stirred separately at room couple exhibited a cathodic shift (+0.60 V) while the second
1426 | Dalton Trans., 2014, 43, 1424–1433
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 1 (A) Experimental and (B) simulated cyclic voltammetry of 1.0 ×
10−3 M PD in CH2Cl2 containing 0.1 M TBAP in the presence of [HIm] =
Fig. 3 Cyclic voltammetry of 1.0 × 10−3 M ZnTMP-Ph-PD in CH2Cl2
containing 0.1 M TBAP in the presence of (A) [HIm] = 0.00–1.00
equivalent and (B) 1.00–2.00 equivalents. Working electrode: glassy
0.00–1.00 equiv. Working electrode: glassy carbon. Scan rate: 0.1 V s−1
.
carbon. Scan rate: 0.1 V s−1
.
in the quantity of HIm to one equivalent of ZnTMP-Ph-PD,
there was a decrease in the current of the third redox couple
and an increase in current of the first redox couple (with E1/2
of +0.56, +0.82, and +1.14 V). Further increase in the quantity
of ligand to 2.0 equivalents of ZnTMP-Ph-PD led to a cathodic
shift in the second redox couple and an anodic shift in the
third (with E1/2 of +0.56, +0.61, and +1.21 V). The results from
the voltammetry of ZnTMP-Ph2-PD (Fig. S1, ESI†) were similar
to those observed for ZnTMP-Ph-PD, with the three redox
couples at E1/2 values of +0.56, +0.80, and +1.10 V. Titration
with HIm also brought about the same changes in the CV of
ZnTMP-Ph2-PD as observed for ZnTMP-Ph-PD (1.0 equiv. with
E1/2 of +0.56, +0.80, and +1.14 V; 2.0 equiv. with E1/2 of +0.56,
+0.62, and +1.24 V).
Fig. 2 Cyclic voltammetry of 1.0 × 10−3 M ZnTMP in CH2Cl2 containing
0.1 M TBAP in the presence of [HIm] = 0.00–1.00 equiv. Working
A comparison of the half-wave potentials observed in
the CVs of PD, ZnTMP, ZnTMP-PD, ZnTMP-Ph-PD, and
ZnTMP-Ph2-PD and the changes therein upon HIm titration
allows us to identify the nature of the electrochemical reac-
electrode: glassy carbon. Scan rate: 0.1 V s−1
.
redox showed an anodic shift (+1.21 V).18,19 When the compo- tions occurring in ZnTMP-Ph-PD and ZnTMP-Ph2-PD.17 The
site molecule ZnTMP-Ph-PD (Fig. 3) was scanned, its CV three reversible redox couples in the CVs of ZnTMP-Ph-PD,
exhibited three redox couples with E1/2 of +0.56, +0.82, and and ZnTMP-Ph2-PD correspond to the loss of one electron on
+1.12 V. The effect of various concentrations of HIm on E1/2 of the PD moiety ([ZnP-PD]+•), the following loss of one electron
ZnTMP-Ph-PD was investigated. Concurrent with the increase on zinc porphyrin moieties ([ZnP-PD]2+), and the overlap of
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 1424–1433 | 1427

View Article Online
Paper
Dalton Transactions
Table 1 Oxidation potentials of PD, ZnTMP, ZnTMP-PD, ZnTMP-Ph-PD, and ZnTMP-Ph2-PD in the absence or presence of HIm
0.00 equiv. HIm 1.00 equiv. HIm 2.00 equiv. HIm
Eox1/V
Eox2/V
Eox3/V
Eox1/V
Eox2/V
Eox3/V
Eox1/V
Eox2/V
Eox3/V
PD
ZnTMP
ZnTMP-PD
ZnTMP-Ph-PD
ZnTMP-Ph2-PD
+0.59
+0.79
+0.58
+0.56
+0.56
+1.01
+1.11
+0.81
+0.82
+0.80
+0.62a
+0.60
+0.57
+0.56a
+0.56a
+1.21
+0.70
+0.82
+0.80
+1.13a
+1.12a
+1.10a
+1.21
+1.14
+1.14
+0.57
+0.70
+0.61
+0.62
+1.21
+1.21
+1.24
+0.56a
+0.56a
a Two-electron transfer involved in a redox couple.
respective loss of one electron on both zinc porphyrins and
the PD moiety ([ZnP-PD]4+), respectively. The oxidation poten-
tials of all species in various oxidation states are shown in
Table 1.
Spectroelectrochemistry
The variations in the absorption spectra of ZnTMP-Ph-PD as a
function of the applied potentials (Eappl.) were investigated by
the OTTLE spectroelectrochemical method. With an increase
in the applied potential from −0.10 to +0.61 V (Fig. 4(A)), the
absorption peak of ZnTMP-Ph-PD shows a decrease in the
intensities at 316, 322 and 422 nm, in addition to the appear-
ance of a new broad band (600–1000 nm) which can be attribu-
ted to the absorption by PD+• (Fig. S2, ESI†). Notice that the
slight decrease observed in the intensity of the Soret band
(422 nm) is indicative of the delocalization of the electron
cloud from zinc porphyrin to PD+•. When the applied potential
is further increased from +0.70 to +0.94 V (Fig. 4(B)), the peaks
at 422 and 550 nm gradually disappear, while the broad band
(500–700 nm) corresponding to the absorption by ZnTMP+•
grows in intensity (Fig. S3, ESI†). Therefore, we could attribute
the second oxidation to the zinc porphyrin. To observe the
ligand transfer behaviour of the modified zinc porphyrins, we
added 0.75 equivalent of HIm to the zinc porphyrin solution.
The observed spectra then results from the mixture of two
species, i.e. ZnTMP-Ph-PD (λSoret = 422 nm) and (HIm)-
ZnTMP-Ph-PD (shoulder band at 432 nm) (Fig. S4, ESI†). In
the potential range from −0.10 to +0.53 V (Fig. 5(A)), there is a
decrease in the intensities at 316 and 322 nm along with the
formation of a broad band (600–1000 nm). These trends
observed in the spectra of the complex are the same as those
observed in the absence of HIm, but for the disappearance of
band at 432 nm which is the evidence for the (HIm)-
Fig. 4 Potential dependent spectral changes of 4.0
ZnTMP-Ph-PD in the absence of HIm in CH2Cl2 containing 0.1 M TBAP.
× M
10−5
ZnTMP-Ph-PD to ZnTMP-Ph-(PD+•)(HIm) transition (enlarged 0.00 V), the recovery of ZnTMP-Ph-PD (53%) is lower than that
inset in Fig. 5(A)). The likely reason for the observed transfer of ZnTMP-PD (73%), showing irreversibility. The absence of
of the ligand from one moiety to other is the post-oxidation protection at the meso position leads to easier decomposition
change in the relative binding constants. The binding constant of the oxidized porphyrin and results in a decrease of intensity
for HIm with an oxidized PD is larger than that for HIm with in each absorption band.18 Therefore, we propose that the
the neutral zinc porphyrin. Nevertheless, from our previous decomposition of zinc porphyrin along with the longer bridge
studies with ZnTMP-PD (Fig. S5, ESI†),12 the transition is the likely reason for the ambiguous spectral change repre-
between four and five coordinated zinc porphyrin was unclear senting the transition of HIm from (HIm)ZnTMP-Ph-PD to
at the increasing absorption spectra of four coordinated zinc ZnTMP-Ph-(PD+•)(HIm). A potential range of +0.56 to +0.68 V
porphyrin (λSoret = 422 nm). In the recovery experiment (after (Fig. 5(B)) is not typically positive enough to oxidize the zinc
the initial application of +0.53 V, the potential was reverted to porphyrin moiety, however, the observed slight decrease in the
1428 | Dalton Trans., 2014, 43, 1424–1433
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
In the case of ZnTMP-Ph2-PD, the patterns of change in
spectra are similar to those of ZnTMP-Ph-PD (Fig. S6 and S7,
ESI†). However, a more severe decomposition of oxidized zinc
porphyrin competes with the transition of ZnTMP-Ph2-PD to
(HIm)ZnTMP-Ph2-PD (recovery of cation radical is ∼40%), and
presents difficulties in the spectral analysis (enlarged picture
in Fig. S7†). On the basis of the decrease in the intensity in
the absorption spectra at 422 nm after the transition of
ZnTMP-Ph2-PD to (HIm)ZnTMP-Ph2-PD, we infer that the
decomposition continued at the applied oxidative state. These
data present spectral evidence for the transfer of the HIm
ligand from zinc porphyrin to the PD moiety, even when the
distance between the two redox centres has changed.
Binding constants between zinc porphyrin and HIm
To gain further insights into the transfer of HIm between PD
and zinc porphyrin moieties, the binding constants of HIm
with zinc porphyrin in various oxidation states were further
surveyed. To determine the binding constants of the neutral
and cation radical states of the zinc porphyrin moiety, we inde-
pendently titrated ZnTMP-Ph-PD and ZnTMP-Ph2-PD with
HIm and monitored the resulting UV-Vis spectra. With increas-
ing concentration of HIm the absorption band (Q band) of
ZnTMP-Ph-PD decreases at 550 nm, and increases at 567 and
607 nm. Furthermore, three well-defined isosbestic points
appear at 558, 586, and 593 nm, indicating a conversion of
ZnTMP-Ph-PD to (HIm)ZnTMP-Ph-PD (Fig. S8 ESI†).19 The
absorption spectra of ZnTMP-Ph2-PD exhibit a similar pattern
(Fig. S9, ESI†). The reported binding constant of ZnTMP-Ph-
PD with HIm is 4.84 × 104 M−1, and that of ZnTMP-Ph2-PD is
1.87 × 105 M−1 20
.
In contrast to ZnTMP-PD,17 the order of the
binding constants are ZnTMP-Ph2-PD > ZnTMP-Ph-PD
>
ZnTMP-PD. The binding constants between HIm and the oxi-
dized zinc porphyrin moiety were also calculated as described
previously based on the potential shift of CVs (Table 1).21 The
binding constants of [ZnTMP-Ph-PD]+• and [ZnTMP-Ph2-PD]+•
are 2.60 × 107 M−1 and 1.00 × 108 M−1, respectively (Table 2).
The details of calculations for determining the binding con-
stants of neutral and cation radical states of the zinc porphyrin
moiety are shown in page S11, ESI† (Scheme 1).
Table 2 Binding constants for HIm with PD, ZnTMP, ZnTMP-PD,
ZnTMP-Ph-PD, and ZnTMP-Ph2-PD in various oxidation states
Fig. 5 Potential dependent spectral changes of 4.0 × 10−5 M ZnTMP-Ph-PD
in the presence of 0.75 equivalents HIm in CH2Cl2 containing 0.1 M TBAP.
Molecule
Binding constant (K) (M−1
)
Reference
a
a
PD+•
4500
PD2+
2.76 × 1012
1.92 × 104
5.13 × 107
9.19 × 101
2.32 × 104
1.24 × 107
4.84 × 104
2.60 × 107
1.87 × 105
1.00 × 108
intensity of the peak at 422 nm in the absorption spectra is a
characteristic of oxidized zinc porphyrin (Fig. S3, ESI†). It indi-
cates a likelihood of interaction between two redox centres
that is caused by the delocalization of the electron density
from the zinc porphyrin ring to PD+•. With the increase in
applied potential from +0.70 to +0.94 V (Fig. 5(C)), the absorp-
tion at 422 nm gradually disappears, while a broad band
(500–700 nm) grows, indicating the formation of cation radical
of zinc porphyrin, ZnTMP+•-Ph-PD+•.
ZnTMP
17
17
17
12
ZnTMP+•
ZnTMP2+
ZnTMP-PD
ZnTMP-PD+•
ZnTMP-Ph-PD
ZnTMP-Ph-PD+•
ZnTMP-Ph2-PD
ZnTMP-Ph2-PD+•
12
a
a
a
a
a This work.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 1424–1433 | 1429

View Article Online
Paper
Dalton Transactions
Scheme 1 Synthesis of ZnTMP-(Ph)n-PD.
As for the binding constant of HIm with oxidized PD, the
best way to obtain quantitative information on electrochemi-
cally induced binding systems is the use of digital simula-
tions to obtain
a fit for the experimentally acquired
voltammograms.22–26 The reaction mechanism consisting of
only electron transfers and hydrogen bonding equilibria is rep-
resented by the six-membered square scheme in Scheme 2.
DigiSim (v 3.03) was used to carry out the digital simulations
of experimental CVs. The values for oxidation potentials of PD
(Eo1x and Eo2x), the diffusion coefficient (D), and the uncompen-
sated resistance (Ru) were determined first by fitting the CVs of
PD recorded in the absence of imidazole. The magnitude of
Scheme 2 Equilibria in the square reaction mechanism for the redox-
dependent binding of HIm to PD.
current was fitted by adjusting the value of D. For all un- fit was 290 Ω. The determined Eo1x′, E2ox′, D, and Ru were then used
complexed PD, D is assumed to have the same value of 1.0 × to fit the complete sets of CVs recorded during titration. The
10−5 cm2 s−1. The peak-to-peak separation was fitted by adjust- value of K1 depends both on the shift in Eo1x and the binding
ing the values of D and Ru. The value of Ru resulting from the constant associated with the neutral state, while K2 is related
1430 | Dalton Trans., 2014, 43, 1424–1433
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Table 3 Parameters for the simulation of the CV for the PD/HIm systema
Electrochemical reactions
E0
ks (cm s−1
)
Eo1x
Eo2x
Eo1x′
Eo2x′
(PD)+ + e = (PD)
0.61 V
1.03 V
0.61 V
0.51 V
1
0.03
10
0.005
(PD)2+ + e = (PD)+
(PD)(HIm)+ + e = (PD)(HIm)
(PD)(HIm)2+ + e = (PD)(HIm)+
Chemical reactions
K
kf
K1
K2
(PD)+ + (HIm) = (PD)(HIm)+
(PD)2+ + (HIm) = (PD)(HIm)2+
4500 M−1
1.0 × 108 M−1 s−1
1.0 × 1010 M−1 s−1
2.76 × 1012 M−1
a Diffusion coefficient (D) for PD, PD+, PD2+, (PD)(HIm), (PD)(HIm)+, and (PD)(HIm)2+ were set to be 1.0 × 10−5 cm2 s−1; D for (HIm) = 5.0 × 10−5
cm2 s−1; Ru = 290 Ω; A = 0.07 cm2; all transfer coefficients (α) = 0.5; scan rate = 0.1 V s−1
.
to the change in Eo2x. The determination of the K1 is particu-
larly important because its value largely determines the behav-
iour in the second oxidation, which would exhibit two
different types, two resolved waves or shifted waves in the
electrochemically induced ligand binding system.22 Based on
our previous studies with electrochemically induced hydrogen
bonding between oxidized PD and pyridines (pKa = 0.67–7.48),
reasonable starting values for K1 could be assumed for the
simulations.17 Since the pKa of HIm is 6.92, the binding con-
stant of cation radical, PD+• with HIm (K1) was estimated to be
4500 M−1 and the one of dication, PD2+ and HIm (K2) was calcu-
Scheme 3 Electrochemically controlled ligation of HIm with
ZnTMP-Ph-PD (black letters) and ZnTMP-Ph2-PD (blue letters).
lated to be 2.76 × 1012
M
−1. In addition, for the shift in E1/2
to be observed, the forward rate of hydrogen bond formation
should be fast. Therefore, with values for kinetic constants kf,1
and kf,2 in the 108–1010 M−1 s−1 range, we could obtain the
best fit for the CVs as shown in Fig. 1(B). Other parameters are
listed in Table 3.
The binding constants of PD, ZnTMP, ZnTMP-PD,
ZnTMP-Ph-PD and ZnTMP-Ph2-PD in various states of oxi-
dation (Table 2) explain the ligand transfer behaviour of this
electrochemically controlled system. In the initial neutral
state, HIm ligates to the centre of the zinc porphyrin moiety
(K = 104–105 M−1). After the electrode potential induced oxi-
dation of the PD moiety, the binding site of HIm shifts to the
PD moiety (since K = ∼1012 M−1). As the oxidation potential
becomes more positive, the zinc porphyrin is further oxidized.
However, with a binding constant in the 107–108 M−1 range for
HIm with electron deficient zinc porphyrins, the reversal of
the HIm ligand from the oxidized PD moiety to the oxidized
zinc porphyrin is unlikely and forbidden. Scheme 3 shows the
oxidation dependent changes in ligation of HIm in the PD-sub-
stituted zinc porphyrin system. This ligand transfer system,
then, is an electrochemically controlled switch rather than a
shuttle.
Fig. 6 EPR spectra of (A) PD+•
, , ,
(B) ZnTMP+• (C) ZnTMP-PD+•
(D) ZnTMP-Ph-PD+•, and (E) ZnTMP-Ph2-PD+• at 298 K.
centred at g = 2.0030 is different from the isoelectronic N2+•
with a five-line spectrum because a phenyl group is replaced
by a proton with hyperfine coupling interaction of ∼4.6 G, con-
tributing two nitrogen ions with ∼5.7 G for each, as shown in
Fig. 6(A) (Fig. S10, ESI†). The EPR spectrum of ZnTMP+• (Fig. 6
EPR studies on electron transfer in modified zinc porphyrins
To observe the electronic interactions between zinc porphyrin (B)) is a representative of the typical pattern observed for
and the PD moiety, cation radicals PD+•, ZnTMP+•, ZnTMP porphyrin macrocycles oxidized at the a2u orbital with
ZnTMP-PD+•, ZnTMP-Ph-PD+•, and ZnTMP-Ph2-PD+• were g = 2.0030. In the spectrum of ZnTMP-Ph-PD+•, the splitting
generated in situ at 298 K in CH2Cl2 and were found to be due to hyperfine coupling ascribed to ZnTMP+• is found to be
stable at room temperature. Those systems were separated into superimposed on the primary pattern of PD+• (Fig. 6(D) and
two portions, ZnTMP and PD. The six-line spectrum of PD+• the enlarged inset). This well-resolved spectrum demonstrates
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 1424–1433 | 1431

View Article Online
Paper
Dalton Transactions
that the radical is delocalized between the porphyrin ring and two moieties critically influences the delocalization of
the PD moiety. In contrast, for the phenyl-linker lacking electrons from the PD to the ZnTMP moiety. A shorter bridge
ZnTMP-PD+•, the EPR splitting appears to be similar to that is beneficial for better electron transfer. It is likely that the
obtained from a mixture of ZnTMP+• and PD+•. In contrast, electron density of each moiety can be controlled by varying
when the linker is composed of two phenyl groups, the EPR the degree of delocalization of the bridge in this progressive
spectrum is similar to that of ZnTMP-Ph-PD+•, but with switch (Chart 2). The EPR spectra further substantiated the
smaller hyperfine couplings (Fig. 6(C) and (E)). The EPR spec- proposed model for this phenomenon.
trum of ZnTMP-PD+• has an obscure cleavage, and a broad line
width which is similar to the pattern observed for PD+•. The
interaction between the two redox centres is proportional to
the hyperfine coupling constants. The larger hyperfine coup-
Acknowledgements
ling constant of ZnTMP+•-PD+• was not split easily on the
The authors thank the National Science Council of Taiwan,
shape. It seems like a merger of both hyperfine couplings
Republic of China, for support of this work.
from zinc porphyrin and the PD moiety occurs. In contrast
with the EPR spectrum of ZnTMP+•-Ph2-PD+• had the same
broad line width, but the cleavage was easy to distinguish
because of the smaller hyperfine coupling constant.
Notes and references
1 J. T. Groves, K. Shalyaev and J. Lee, in The Porphyrin Hand-
book, ed. K. M. Kadish, K. M. Smith, R. Guilard, Academic
Press, New York, 2000, vol. 4, p. 17.
Conclusions
2 N. A. Stephenson and A. T. Bell, J. Mol. Catal. A: Chem.,
2007, 272, 108.
3 C. Monnereau, J. Gomez, E. Blart and F. Odobel, Inorg.
Chem., 2005, 44, 4806.
4 A. Kira, T. Umeyama, Y. Matano, K. Yoshida, S. Isoda,
J. K. Park, D. Kim and H. Imahori, J. Am. Chem. Soc., 2009,
131, 3198.
5 C.-Y. Liu and A. J. Bard, Acc. Chem. Res., 1999, 32, 235–
245.
6 F. Fungo, L. Otero, C. D. Borsarelli, E. N. Durantini,
J. J. Silber and L. Sereno, J. Phys. Chem. B, 2002, 106,
4070.
7 M. Gervaldo, F. Fungo, E. N. Durantini, J. J. Silber,
L. Sereno and L. Otero, J. Phys. Chem. B, 2005, 109, 20953.
8 C. M. Drain, A. Varotto and I. Radivojevic, Chem. Rev., 2009,
109, 1630.
9 S. F. Miller and G. T. Babcock, Chem. Rev., 1996, 96, 2889.
10 G. T. Babcock and M. Wikstrom, Nature, 1992, 356, 301.
11 J. P. Abrahams, A. G. W. Leslie, R. Lutter and J. E. Walker,
Nature, 1994, 370, 621.
12 K. Y. Chiu and Y. O. Su, Chem. Commun., 2009, 2884.
13 J. Bu, N. D. Lilienthal, J. E. Woods, C. E. Nohrden,
K. T. Hoang, D. Truong and D. K. Smith, J. Am. Chem. Soc.,
2005, 127, 6423.
We report here that both ZnTMP-Ph-PD and ZnTMP-Ph2-PD
exhibit an ability to switch the binding sites for the ligand
HIm. In the neutral state, HIm is coordinated as an axial
ligand to form a five-coordinated zinc porphyrin. In the pres-
ence of HIm, the PD moiety loses two electrons on oxidation
due to the electrochemically induced formation of hydrogen
bond. Since the affinity of HIm for PD2+ is greater than its
affinity for the neutral zinc porphyrin, HIm is repositioned to
bind with PD2+. Further application of a more positive poten-
tial leads to the formation of a cation radical of the zinc por-
phyrin moiety; however, the binding constant of HIm with the
oxidized zinc porphyrin is smaller than that for HIm and the
PD2+ moiety. Therefore, the shuttling behaviour could not be
seen in this system. A switch from PD back to the zinc por-
phyrin moiety could, however, be observed on returning to the
neutral state. Furthermore, the length of the covalent bridge
affects the binding constant of the five-coordinated zinc por-
phyrin with the following trend in the magnitude: ZnTMP-Ph2-
PD > ZnTMP-Ph-PD > ZnTMP-PD.
For studies related to electron transfer behaviour, it was
observed that the length of the covalent bridge that links the
14 Y. Ge, R. R. Lilienthal and D. K. Smith, J. Am. Chem. Soc.,
1996, 118, 3976.
15 Y. Hirao, A. Ito and K. Tanaka, J. Phys. Chem. A, 2007, 111,
2951.
16 D. Simoni, G. Giannini, M. Roberti, R. Rondanin,
R. Baruchello, M. Rossi, G. Grisolia, F. P. Invidiata,
S. Aiello, S. Marino, S. Cavallini, A. Siniscalchi, N. Gebbia,
L. Crosta, S. Grimaudo, V. Abbadessa, A. D. Cristina and
M. Tolomeo, J. Med. Chem., 2005, 48, 4293.
17 Y. C. Chung, Y. J. Tu, S. H. Lu, W. C. Hsu, K. Y. Chiu and
Y. O. Su, Org. Lett., 2011, 13, 2826.
Chart 2 Degree of delocalization in the ZnTMP-bridge-PD system.
1432 | Dalton Trans., 2014, 43, 1424–1433
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
18 (a) R. Guilard, N. Jagerovic, A. Tabard, C. Naillon and 22 S. R. Miller, D. A. Gustowski, Z. H. Chen, G. W. Gokel,
K. M. Kadish, J. Chem. Soc., Dalton Trans., 1992, 1957;
(b) Y. O. Su, D. Kim and T. G. Spiro, Electroanal. Chem.,
L. Echegoyen and A. E. Kaifer, Anal. Chem., 1988, 60,
2021.
1987, 234, 145; (c) C. K. Chang, G. Aviles and N. Bag, J. Am. 23 D. H. Evans, Chem. Rev., 2008, 108, 2113.
Chem. Soc., 1994, 116, 12127.
24 J. Bu, N. D. Lilienthal, J. E. Woods, C. E. Nohrden,
19 C. L. Lin, M. Y. Fang and S. H. Cheng, J. Electroanal. Chem.,
2002, 531, 155.
K. T. Hoang, D. Truong and D. K. Smith, J. Am. Chem. Soc.,
2005, 127, 6423.
20 G. S. Marbury, C. Brewer and G. Brewer, J. Chem. Soc., 25 C. C. Leonor, S. L. Martin and D. K. Smith, J. Org. Chem.,
Dalton Trans., 1991, 1377. 2005, 70, 10817.
21 K. M. Kadish, L. R. Shiue, R. K. Rhodes and 26 H. C. Cheng, K. Y. Chiu, Y. J. Tu, T. F. Yang and Y. O. Su,
L. A. Bottomley, Inorg. Chem., 1981, 20, 1274.
Org. Lett., 2013, 15, 3868.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 1424–1433 | 1433