Zn(II), Cu(II), and Ni(II) 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrins (TFPs) have been synthesized and characterized. The electronic spectroscopy and cyclic voltammetry of these compounds, along with the free-base macrocycle (2H-TFP), have been determined; 2H-TFP was also structurally characterized by X-ray crystallography. The Cu(II)TFP exhibits catalytic activity for the hydrogen evolution reaction (HER). The analysis of linear sweep voltammograms shows that the HER reaction of Cu(II)TFP with benzoic acid is first-order in proton concentration with an average apparent rate constant for HER catalysis of k app = 5.79 ± 0.47 × 103 M-1 s-1.
Zn(II), Cu(II), and Ni(II) 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrins (TFPs) have been synthesized and characterized. The electronic spectroscopy and cyclic voltammetry of these compounds, along with the free-base macrocycle (2H-TFP), have been determined; 2H-TFP was also structurally characterized by X-ray crystallography. The Cu(II)TFP exhibits catalytic activity for the hydrogen evolution reaction (HER). The analysis of linear sweep voltammograms shows that the HER reaction of Cu(II)TFP with benzoic acid is first-order in proton concentration with an average apparent rate constant for HER catalysis of k app = 5.79 ± 0.47 × 103 M-1 s-1.
Porphyrin macrocycles
have found prominence in applications of
catalysis,[1−5] medicine,[6−11] bioimaging,[12,13] molecular electronics,[14−16] information storage,[17,18] and optical imaging[19] as well as energy conversion transformations.[20,21] Metalloporphyrins are especially promising as catalysts for the
activation of small molecules of energy consequence including carbon
dioxide reduction,[22−26] oxygen reduction reaction,[27−29] and hydrogen evolution reaction
(HER).[30−34] Interest in the latter reaction is driven by the potential utility
of hydrogen (H2) as a form of renewable energy storage[35] and consequently plays a key role in sustainable
fuel cycles,[36−38] especially as hydrogen fuel cells become increasingly
popular.[39−41]These diverse applications of porphyrin macrocycles,
which consist
of four pyrrole rings interconnected via methene
bridges, are derived from functionalization of their peripheries,[42] thus allowing structure–function properties
to be tuned with fidelity. In this study, we focus on metalloporphyrins
functionalized at the meso position of the macrocyclic ring and substituted
with the electron-withdrawing 4-fluoro-2,6-dimethylphenyl group, which
was chosen as fluorine provides a convenient handle for porphyrin
characterization of new metalated complexes. The free-base 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrin
(2H-TFP) platform has been metalated with Fe and shown to support
oxidative chemistry; heterolytic cleavage of bound hypochlorite produces
the compound I analogue (TFP•+)Fe(IV)(O).[43] We now report the reductive chemistry of 2H-TFP
with late metals of the first-row transition metal series. The 2H-TFP
has been structurally characterized and metalated with Ni(II), Cu(II),
and Zn(II). We posited that the late metals such as Cu(II) may exhibit
enhanced rates of HER catalytic activity. We show by cyclic voltammetry
coupled to chemical analysis that Cu(II)TFP is an HER catalyst, and
the analysis of the electrocatalytic waveform reveals an appreciable
rate of HER as compared to previous porphyrin electrocatalysts.[30,44,45]
Experimental Section
General
Considerations
Free-base meso-(4-fluoro-2,6-dimethylphenyl)porphyrin
(TFP) was purchased from
Frontier Scientific. Cu(OAc)2·H2O, Ni(OAc)2·4H2O, and anhydrous Zn(OAc)2 were
purchased from Sigma-Aldrich. Hexanes, CH2Cl2, MeOH, CHCl3, acetic acid, Na2SO4, and NaCl were used as received. Tetrabutylammonium hexafluorophosphate
(TBAPF6) was obtained from Sigma-Aldrich, recrystallized
from EtOH, dried for two days under vacuum at a Schlenk line without
heat, and stored in a nitrogen-filled glovebox. Electrolyte solutions
were stored over activated 3 Å molecular sieves before use. 1H NMR spectra were recorded at the Harvard University Department
of Chemistry and Chemical Biology NMR facility on a JEOL ECZ400S spectrometer
operating at 400 MHz. Absorption spectra were taken with a 1.0 cm
quartz cuvette on a Varian Cary 5000 UV–vis-NMR spectrophotometer.
Steady-state emission spectra were measured on a fluorimeter (Photon
Technology International, PTI model QM4) coupled to a 150 W Xe arc
lamp as an excitation light source. Mass spectrometry was performed
at the Harvard Center for Mass Spectrometry.
Free-base 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrin
(30 mg, 0.0376 mmol) and Ni(OAc)2·4H2O
(140 mg, 0.564 mmol) were dissolved in acetic acid (15 mL). The resulting
solution was heated at reflux for 3 h and then cooled to room temperature.
The solvent was removed in vacuo, and the residue
was subsequently redissolved in CH2Cl2 (10 mL),
treated with TEA (1 mL), washed with water and brine, and dried over
Na2SO4. The crude porphyrin solution was concentrated
to dryness, and the resulting solid was dissolved in CH2Cl2 (10 mL). The crude porphyrin solution was introduced
onto a silica column and eluted with a gradient of 0–40% CH2Cl2/hexanes. The title compound was isolated as
a dark-purple powder (25 mg, 78% isolated yield). 1H NMR
(400 MHz, CD2Cl2): δ (ppm) 8.54 (s, 8H),
7.13 (d, J = 9.6 Hz, 8H), 1.84 (s, 24H). 19F NMR (400 MHz, CD2Cl2): δ (ppm) −116.14
(s). ESI-TOF MS [(M + H)+, M = C52H40F4N4Ni] (m/z): calcd (obsd), 854.2537 (854.2534).
Free-base 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrin
(30 mg, 0.0376 mmol) was dissolved in CHCl3 (60 mL). In
a separate round-bottom flask, anhydrous Zn(OAc)2 (103
mg, 0.564 mmol) was dissolved in MeOH (10 mL). The resulting Zn(OAc)2 solution was added into the porphyrin solution, and the mixture
was heated at reflux for 3 h and then cooled to room temperature.
The solvent was removed in vacuo, and the residue
was subsequently redissolved in CH2Cl2 (10 mL).
The resulting crude porphyrin solution in CH2Cl2 was introduced onto a silica column with a syringe and eluted with
a gradient of 0–40% hexanes/CH2Cl2. The
title compound was isolated as a purple powder (28 mg, 86% isolated
yield). 1H NMR (400 MHz, CDCl3): δ (ppm)
8.69 (s, 8H), 7.20 (d, J = 9.7 Hz, 8H), 1.84 (s,
24H). 19F NMR (376 MHz, CDCl3): δ (ppm)
−116.61 (s). ESI-TOF MS [(M + H)+, M = C52H40F4N4Zn] (m/z): calcd (obsd), 861.2553 (861.2535).
Free-base 5,10,15,20-tetrakis(4-fluoro-2,6-dimethylphenyl)porphyrin
(30 mg, 0.0376 mmol) and Cu(OAc)2·H2O (112
mg, 0.564 mmol) were dissolved in acetic acid (15 mL). The solution
was heated at reflux for 3 h and cooled to room temperature. The solvent
was removed in vacuo, and the residue was subsequently
redissolved in CH2Cl2 (10 mL), treated with
TEA (1 mL), washed with water and brine, and dried over Na2SO4. The resulting CH2Cl2 solution
was introduced onto a silica column and eluted with a gradient of
10–60% hexanes/CH2Cl2. The title compound
was isolated as an orange-red powder (22 mg, 69% isolated yield).
ESI-TOF MS [(M + H)+, M = C52H40CuF4N4] (m/z): calcd
(obsd), 860.2558 (860.2534).
Crystallography
Crystals were grown through vapor-diffusion
of hexanes into pyridine solution of 2H-TFP at 3 °C. X-ray diffraction
data were collected on a Bruker three-circle platform goniometer equipped
with an APEX II CCD detector and an Oxford Cryosystems cryostat cooling
device using φ and ω scans. A fine-focus sealed tube Cu
Kα (1.54178 Å) X-ray source was used. The crystal was mounted
on a cryoloop using Paratone oil. The crystal is a four-component
non-merohedral twin, and the useful data in the hkl5 format were created
by using CELL NOW/Twinabs in the APEX3 program suit. The structure
was solved by intrinsic phasing using SHELXT and refined against F2 on all data by full-matrix least-squares with
SHELXL. All non-hydrogen atoms were located in the difference density
maps (obtained using Fourier coefficients) and refined anisotropically.
Hydrogen atoms were added at calculated positions and refined with
a riding model. The disordered solvent could not be located in the
Fourier map and was squeezed out by using PLATON/SQUEEZE. Crystal
data, data collection, and structure refinement details are included
in Table S1.
Electrochemical Methods
All electrochemical experiments
were performed with a CH Instruments 760D electrochemical workstation
(Austin, Texas) and CHI Version 10.03 software in a N2-filled
glovebox. For cyclic voltammetry and hydrogen electrocatalysis experiments,
a three-electrode undivided cell configuration was used with a BASi
3 mm diameter glassy carbon working electrode and a Pt wire auxiliary
electrode. The working electrode was polished between each measurement.
For bulk electrolysis experiments, a H-cell was used; the working
electrode was a glassy carbon plate of 2 cm2, and the auxiliary
electrode was a Pt mesh. The reference electrode for all electrochemical
experiments was a BASi Ag wire immersed in the 0.1 M TBAPF6 solution in CH2Cl2, and ferrocene was used
as an internal standard to define the reference potential. All experiments
were performed at room temperature (25 ± 1 °C).
Fluorescence
Lifetime Measurements
Time-resolved emission
spectra were collected with a Hamamatsu C4334 Streak Scope camera
placed perpendicular to the excitation path. The polarizer in the
excitation path was set parallel to the excitation pulse, and the
polarizer in the emission path was set to the magic angle (54.7°)
relative to the parallel one. Measurements were taken in CH2Cl2 under nitrogen and in the presence of air. All samples
were excited with 400 nm light generated using an OPerA Solo (Coherent)
femtosecond optical parametric amplifier seeded with fundamental pulses
(800 nm, ∼50 fs) from a Ti:sapphire regenerative amplifier
(Coherent Libra-HE). Measurements were made with the following streak
camera settings: a slit width of 2000, a grating of 100 g mm–1 blazed at 450 nm, and a ruling of 100. Time-resolved emission spectra
were the result of the integration of 15,000 exposures. After integration
over wavelength, the emission decay curves were fit to a single or
biexponential decay function in Logger Pro 3.14.1.
Results and Discussion
Synthesis
of Metalloporphyrins
The synthetic schemes
to incorporate the metals of Ni(II), Cu(II), and Zn(II) into 2H-TFP
are shown in Figure . Metalation reactions were conducted using the metal acetate salts
followed by workup and silica gel chromatography to obtain the target
porphyrins between 69 and 86% isolated yields. 1H NMR spectra
for Ni(II)TFP and Zn(II)TFP (Figure S1)
were consistent with expectations, where each showed a loss of the
pyrrolic N–H resonance in the 2H-TFP upon metalation, as well
as the expected chemical shifts and integrations for the porphyrin
macrocycle. The corresponding 19F NMR spectra of Ni(II)TFP
and Zn(II)TFP (Figure S2) showed the expected
singlet. Given its paramagnetism, NMR spectra were not collected for
Cu(II)TFP. The S = 1/2 ground state of Cu(II)TFP
is consistent with the effective magnetic moment of 1.55 μB as measured by the Evans method.
Figure 1
Synthesis schematic of
metal porphyrins.
Synthesis schematic of
metal porphyrins.The TFP macrocycle has
been crystallographically characterized,
as shown in Figure . The porphyrin crystallizes in the P1̅ space
group with half of the macrocycle contained in the asymmetric unit.
The inner pyrrolic H-atoms are disordered about the two N-atoms in
the asymmetric unit and modeled with an occupancy of 0.5, engendering D4 molecular symmetry. Consistent
with this molecular symmetry, the distance between each pair of nitrogen
atoms was equal (d[N(1)···N(1)] =
4.11(2) Å and d[N(2)···N(2)]
= 4.11(2) Å). The ring is nearly planar with a deviation of 0.03(2)
Å for the 4 – N least-squares plane.
Figure 2
Crystal
structure of TFP. Hydrogen atoms have been omitted for
clarity.
Crystal
structure of TFP. Hydrogen atoms have been omitted for
clarity.
Electronic Spectroscopy
The UV–vis spectra for
2H-TFP, Ni(II)TFP, Cu(II)TFP, and Zn(II)TFP, shown in Figure , are dominated by the Soret
and Q band absorptions of the four-state Gouterman model.[46] As typically observed, the high D4 symmetry of the metalated porphyrins
leads to a single Q-band as compared to 2H-TFP. The red-shift of the
Soret and Q bands across the period from Ni to Zn TFP is consistent
with previous observations of metalloporphyrins.[47]
Figure 3
UV–vis absorption spectra of (A) 2H-TFP, (B) Ni(II)TFP,
(C) Zn(II)TFP, and (D) Cu(II)TFP. Solutions of 2H-TFP and Zn(II)TFP
also exhibited emission, the spectra for which are shown by the dashed
line [λexc = 430 nm for 2H-TFP and 432 nm for Zn(II)TFP].
The absorption and emission spectra are in arbitrary units (A.U.).
The spectra were recorded in CH2Cl2.
UV–vis absorption spectra of (A) 2H-TFP, (B) Ni(II)TFP,
(C) Zn(II)TFP, and (D) Cu(II)TFP. Solutions of 2H-TFP and Zn(II)TFP
also exhibited emission, the spectra for which are shown by the dashed
line [λexc = 430 nm for 2H-TFP and 432 nm for Zn(II)TFP].
The absorption and emission spectra are in arbitrary units (A.U.).
The spectra were recorded in CH2Cl2.As expected for metalloporphyrins with partially occupied
d-orbitals,
the Ni(II) and Cu(II)TFPs exhibit no appreciable emission owing to
fast non-radiative decay from the porphyrin π-aromatic chromophore via the d-orbital state manifold. In the absence of this
non-radiative pathway, the 2H-TFP and the d10 Zn(II)TFP
porphyrins are highly emissive (Figure A,C). Laser excitation of solutions of Zn(II)TFP under
N2 yields emission with a monotonic lifetime decay of 2.27
ns; the emission of 2H-TFP exhibits a biphasic decay of 1.26/8.72
ns under N2 (Figure S3). Consistent
with the assignment of the emission to a fluorescent-excited state,
the excited state decay lifetimes are unaffected by the presence of
oxygen (τo = 2.39 ns for Zn(II)TFP in air and 0.81/9.06
ns for 2H-TFP in air).
Electrochemistry
The cyclic voltammograms
(CVs) of
the TFP series of porphyrins are shown in Figure . The 2H-TFP shows reversible anodic and
cathodic waves (E1/2 = 0.60 and −1.80
V, respectively) in the potential window scanned, whereas the metalated
porphyrins all show two reversible anodic waves and one reversible
cathodic wave, except for Ni(II)TFP for which the reduction wave is
irreversible. This irreversibility is indicative of a reaction of
DCM with the reduced porphyrin. Consistent with this contention, the
CV of Ni(II)TFP becomes reversible in non-halogenated solvents. For
instance, although only slightly soluble in acetonitrile, a reversible
reduction wave is observed (Figure S4).
The electrochemistry shown in Figure is typical of metalloporphyrins, with the oxidations
assigned to the generation of the π-cation radical and dication,
and the reduction to the π-radical anion of the porphyrin ring.[48] As originally reported by Wolberg and Manassen,
the Cu and Zn porphyrins within a homologous series are more easily
oxidized than the Ni (as well as Fe and Co) analogue.[49] Moreover, the redox chemistry follows the trend originally
identified by Furhop, Kadish, and Davis that the ΔE between the radical anion and cation is more or less constant (ΔE = 2.44 ± 0.05), indicative of a similar HOMO–LUMO
gap across a homologous series.[50] However,
for the reversible couples, there is an inverse relation for the ease
of metalloporphyrin ring oxidation and the ease of metalloporphyrin
ring reduction, which has been attributed[50] to be a result of electrostatic and electronegativity effects of
the divalent metal ion residing within the macrocycle cavity.
Figure 4
Cyclic voltammograms
of (A) 2H-TFP, (B) Ni(II)TFP, (C) Zn(II)TFP,
and (D) Cu(II)TFP in CH2Cl2 with a 0.1 M TBAPF6-supporting electrolyte. All CVs were recorded in a three-compartment
cell with a glassy carbon working electrode, a Pt auxiliary electrode,
and a Ag wire dipped in 0.1 M TBAPF6 solution and ferrocene
as an internal reference electrode at a scan rate of ν = 100
mV s–1.
Cyclic voltammograms
of (A) 2H-TFP, (B) Ni(II)TFP, (C) Zn(II)TFP,
and (D) Cu(II)TFP in CH2Cl2 with a 0.1 M TBAPF6-supporting electrolyte. All CVs were recorded in a three-compartment
cell with a glassy carbon working electrode, a Pt auxiliary electrode,
and a Ag wire dipped in 0.1 M TBAPF6 solution and ferrocene
as an internal reference electrode at a scan rate of ν = 100
mV s–1.
HER Catalysis
The CV of Cu(II)TFP in the presence of
an acid is indicative of HER catalysis. CVs of 0.6 mM Cu(II)TFP were
recorded in the presence of 0–8 equiv of benzoic acid (BA)
(Figure A). BA is
a good proton source as its comparative mild acidity (pKa = 20.7 in acetonitrile[51])
does not induce demetalation of Cu(II)TFP. In agreement with catalytic
HER behavior, the cathodic wave increases in current with the addition
of BA, with a concomitant loss of reversibility (Figure A). The irreversibility of
the reduction wave with acid addition is consistent with an ECEC mechanism
where the reduced porphyrin is protonated, followed by further reduction
and protonation to produce H2.[52−55] The analysis of the gaseous headspace
by gas chromatography of bulk-electrolyzed solutions of 1.12 mM Cu(II)TFP
in the presence of 15.4 mM BA confirms the production of H2, which is linearly produced and will consume current (Figure S5). At the completion of bulk electrolysis,
the solution was analyzed by TLC and there was no evidence of demetalated
porphyrin. Additionally, the UV–visible absorption spectra
of electrolyzed solutions match those of the original solution. At
low BA concentrations, the peak-shaped CV is characteristic of mass-transported
limitations on the substrate. However, as the proton concentration
is increased, the peak-shaped CV is less pronounced and approaches
the canonical S-shape where catalysis is governed by pure kinetic
control of the reaction without the consumption of substrate becoming
limiting.[56] The S-shape is evident from
the linear scan voltammograms (LSVs) shown in Figure B, although we note that at large negative
potentials, the plateau region is convoluted with background H2 generation from the glassy carbon electrode.
Figure 5
(A) CVs of Cu(II)TFP
in CH2Cl2 with a 0.1
M TBAPF6-supporting electrolyte with the addition of varying
equivalents of BA. Upon the addition of acid, the electrode was polished
between scans except for the 1 equiv BA scan. (B) LSV (solid line)
and the calculated current trace (dashed line) using eq . LSV traces were corrected for
the capacitive current at −1.0 V vs Fc+/Fc. All CVs were recorded with a glassy carbon button (3
mm diameter) working electrode, a Pt auxiliary electrode, and a Ag
wire reference electrode at a scan rate of ν = 100 mV s–1.
(A) CVs of Cu(II)TFP
in CH2Cl2 with a 0.1
M TBAPF6-supporting electrolyte with the addition of varying
equivalents of BA. Upon the addition of acid, the electrode was polished
between scans except for the 1 equiv BA scan. (B) LSV (solid line)
and the calculated current trace (dashed line) using eq . LSV traces were corrected for
the capacitive current at −1.0 V vs Fc+/Fc. All CVs were recorded with a glassy carbon button (3
mm diameter) working electrode, a Pt auxiliary electrode, and a Ag
wire reference electrode at a scan rate of ν = 100 mV s–1.The catalytic HER activity
may be quantified by normalizing the
observed current under catalytic conditions, i, to
the current of the one-electron wave associated with the reduction
of Cu(II)TFP, ip0. In this
way, the magnitude of i/ip0 establishes whether the experimental current response
is catalytic; in the case of HER, the catalytic regime for the two-electron
reduction of 2H+ to H2 is attained when i/ip0 > 2. The i/ip0 ratio of >2
for BA concentrations >2 equiv is indicative of HER catalytic activity.From the quasi-plateau current, the HER efficacy of Cu(II)TFP may
be ascertained by fitting the canonical S-shaped catalytic CVs to[52,57]where i is the theoretical
current response, ipl is the experimentally
derived plateau current, F is Faraday’s constant, R is the gas constant, T is the temperature, E is the applied potential, and E1/2 is the experimentally derived half-wave potential. Figure B shows the fits to eq for Cu(II)TFP. We note
that the inflection point of the plateau current prior to the increase
in background current at large negative potentials is modeled well
by eq .The maximum
turnover frequencies, TOFmax, at each concentration
of BA are provided from[57]where S is the surface area
of the electrode, CCu0 is the
concentration of the Cu(II)TFP catalyst, and DCu is the diffusion constant of the Cu(II)TFP catalyst based
on that measured for typical porphyrins (D = 5 ×
10–6 cm2 s–1[58]). The catalytic rate constant, which is the
determinant of the plateau current, is related to the TOFmax byTable S2 lists the kapp obtained from eqs and 3 for each concentration
of BA, yielding an average apparent rate constant for HER catalysis
of kapp = 5.79 ± 0.47 × 103 M–1 s–1. Additionally,
the log(TOFmax) values for Cu(II)TFP at every log[BA] concentration
are shown in Figure S6. The slope of the
line furnishes a reaction order of unity with respect to the acid
across the range of [BA], indicating that the HER rate is first order
in proton concentration, consistent with the rate-determining step
involving protonation of the reduced porphyrin in the ECEC process.
We note that the half-wave potential (potential when the current is
half of the plateau current) is slightly shifted from the standard
potential of the catalyst, as observed from the anodic shifts of the
experimental LSV from the calculated LSV in Figure B. This shift may be indicative that the
rate-determining step is not the first irreversible protonation step
after reduction but is rather the second chemical step to produce
H2.
Conclusions
We report the synthesis
and characterization of Ni(II)TFP, Cu(II)TFP,
and Zn(II)TFP and the structural characterization of the free-base
parent porphyrin 2H-TFP. The M(II)TFP series possesses prototypical
spectral and redox properties of metalloporphyrins. In the case of
Cu(II)TFP, its CVs indicate HER catalytic activity. Hydrogen generation
is consistent with an ECEC mechanism that is first order in proton
concentration with an apparent HER rate constant for catalysis of
5.79 ± 0.47 × 103 M–1 s–1, as determined from the analysis of the catalytic
traces of LSVs. These studies expand upon the growing body of literature
on using metalloporphyrins as catalysts for energy conversion transformations.
Authors: Barbara Pucelik; Robert Paczyński; Grzegorz Dubin; Mariette M Pereira; Luis G Arnaut; Janusz M Dąbrowski Journal: PLoS One Date: 2017-10-10 Impact factor: 3.240