J Biol Inorg Chem (2007) 12:681–690 DOI 10.1007/s00775-007-0217-y
ORIGINAL PAPER
Air-stable, heme-like water-soluble iron(II) porphyrin: in situ preparation and characterization Ro´bert Husza´nk Æ Gyo¨rgy Lendvay Æ Otto´ Horva´th
Received: 23 June 2006 / Accepted: 29 January 2007 / Published online: 28 February 2007 SBIC 2007
Abstract Preparation of the water-soluble, kinetically labile, high-spin iron(II) tetrakis(4-sulfonatophenyl)porphyrin, Fe(II)TPPS4–, has been realized in neutral or weakly acidic solutions containing acetate buffer. The buffer played a double role in these systems: it was used for both adjusting pH and, via formation of an acetato complex, trapping trace amounts of iron(III) ions, which would convert the iron(II) porphyrins to the corresponding iron(III) species. Fe(II)TPPS4– proved to be stable in these solutions even after saturation with air or oxygen. In the absence of acetate ions, however, iron(II) ions play a catalytic role in the formation of iron(III) porphyrins. While the kinetically inert iron(III) porphyrin, Fe(III)TPPS3–, is a regular one with no emission and photoredox properties, the corresponding iron(II) porphyrin displays photoinduced features which are typical of sitting-atop complexes (redshifted Soret absorption and blueshifted emission and Q absorption bands, photoinduced porphyrin ligand-to-metal charge transfer, LMCT, reaction). In the photolysis of Fe(II)TPPS4– the LMCT process is followed by detachment of the reduced metal center and an irreversible ringopening of the porphyrin ligand, resulting in the degradation of the complex. Possible oxygen-binding ability of
R. Husza´nk G. Lendvay O. Horva´th (&) Department of General and Inorganic Chemistry, University of Pannonia, P.O. Box 158, Veszpre´m 8201, Hungary e-mail:
[email protected] G. Lendvay Institute of Structural Chemistry, Hungarian Academy of Sciences, P.O. Box 17, Budapest 1525, Hungary
Fe(II)TPPS4– (as a heme model) has been studied as well. Density functional theory calculations revealed that in solutions with high acetate concentration there is very little chance for iron(II) porpyrin to bind and release O2, deviating from heme in a hydrophobic microenvironment in hemoglobin. In the presence of an iron(III)-trapping additive that is much less strongly coordinated to the iron(II) center than the acetate ion, Fe(II)TPPS4– may function as a heme model. Keywords Iron porphyrins Heme-model Sitting atop Catalysis Photochemistry
Introduction Heme proteins (containing iron porphyrin derivatives in their active sites) play an essential role in living organisms in oxygen transport and storage (hemoglobin, myoglobin) and in electron transfer processes (cytochrome c, cytochrome oxidase) [1, 2]. Hemoglobin contained within the red blood cells holds four ferrous porphyrin type groups which provide the active site for oxygen binding and are responsible for the red color of blood. Simple compounds that would bind oxygen reversibly for many cycles could also be of value as emergency blood substitutes or for fractionation of oxygen from air. There have been many attempts to model the function of hemoglobin and myoglobin (the oxygen–carbon dioxide cycle) by use of some kind of oxygen carriers as blood substitutes. In the last few decades, considerable efforts were made to find compounds that are able to mimic the hemoglobin function. Two classes of blood substitutes—namely, perfluorocarbon-type derivatives [3–8] and different human hemoglobin and bovine hemoglobin
123
682
J Biol Inorg Chem (2007) 12:681–690
preparations [9–14]—have been tested in clinical therapy. In spite of partial success, all of these compounds have numerous disadvantages and show many side effects, so they have not become widespread in medical practice. As an alternative, preparation of the heme group [hydrophobic or hydrophilic iron(II) porphyrin] synthetically as a possible blood substitute has also been investigated intensively [15–22]. These studies were focused almost exclusively on a nonaqueous environment (probably because of the hydrophobic nature of the heme group), while the desired simple oxygen carrier should be able to function in aqueous media under physiological conditions. Molecular oxygen has a triplet, paramagnetic ground state and high electronegativity. These features make it one of the most powerful oxidizing agents. Stable yet reversible coordination of this strong oxidant to the iron(II) central atom is rather difficult, because heme compounds mostly react irreversibly with dioxygen. Almost all synthetically prepared iron(II) porphyrin complexes in organic solvents were found to be very unstable in air and are rapidly oxidized to ferric porphyrin, which is biologically inactive [23]. This process is an irreversible overall redox reaction: 4FeII P þ O2 ! 2 FeIII P O FeIII P :
ð1Þ
The initial binding of molecular oxygen is followed by a bimolecular redox process, which leads to the formation of a dioxygen-bridged dimer complex. This l-peroxo dimer can be converted via the formation of an oxo-ferryl intermediate to the l-oxo dimer. This bridged species can further decompose to iron(III) porphyrin. In fact, the initial dimerization process takes place very efficiently in organic media where the complexes are not ionic, i.e., have no charge. The same reaction, however, is probably less favorable in aqueous—especially dilute—solution, because the solubility of the complex is generally ensured by ionic substituents on the porphyrin ring, and then both iron porphyrin units have large positive or negative charge and repel each other. Nevertheless, the presence of water (even a trace amount) causes autoxidation in most iron(II) porphyrins coordinating dioxygen [24]. In aqueous systems, even denatured hemoglobin proved to be extremely oxygen sensitive after losing the hydrophobic microenvironment ensured by the globin moiety for the iron(II) center. This phenomenon has been attributed to the nucleophilic attack of water molecules at the iron–oxygen coordinative bond, excluding the O•– 2 species formed in inner-sphere electron transfer from the iron(II) center. Accordingly, iron(II) porphyrins in aqueous media have been mostly studied in supramolecular systems of biological origin [24]. Quite recent investigations indicated that a possible way to prepare iron(II) porphyrins in
123
aqueous systems is in situ reduction of a water-soluble iron(III) porphyrin encapsulated in an O-methylated b-cyclodextrin dimer ensuring a hydrophobic microenvironment for the metal center [25]. This system proved to be relatively stable against autoxidation in the presence of dioxygen. However, preparation of air-stable iron(II) porphyrin in a real hydrophilic environment required a different strategy [26, 27]. In addition, the conversion of ferrous porphyrin to ferric porphyrin in aqueous solution has not yet been explained. Besides the significant biological function of ferrous porphyrin derivatives, the description of thermal and photoinduced properties of water-soluble complexes would also be important in bioinorganic chemistry. Nowadays, interest in metalloporphyrins keeps increasing, owing to their importance in many fields, such as biochemistry, medical research and catalysis. The photophysical and photochemical properties of excited metalloporphyrins have been utilized, e.g., in optical sensors [28–30], artificial photosynthetic systems (light-harvesting dendrimers) [31] and photodynamic therapy [32–34]. The main subject of this paper is a report on the preparation of iron(II) tetrakis(4-sulfonatophenyl)porphyrin, Fe(II)TPPS4–, in aqueous solution so that the complex formed is stable and is not oxidized in the presence of oxygen. Production of Fe(II)TPPS4– was realized by the complexation reaction of free-base porphyrin (H2TPPS4–) and iron(II) ions in the presence of iron(III)-trapping acetate buffer. In the following, after summarizing the techniques used, we present the preparation method, then the equilibrium, photophysical and photochemical properties of the complex, according to which the complex can be classified as a sitting-atop metal porphyrin. Finally, we discuss whether this complex can serve as a simple heme model with respect to reaction with dissolved molecular oxygen.
Materials and methods Materials All the solutions studied were made by using ionexchanged then distilled Milli-Q water. H2TPPS4– and FeSO47H2O were purchased from Aldrich. The iron(III) porphyrin [iron(III)meso-tetrakis(4-sulfonatophenyl)porphyrin] was from Frontier Scientific Europe. In order to hinder the formation of iron(III) ions, argon was bubbled and metallic iron was added into the iron(II) sulfate solution. The pH values of 6 and 7 were obtained by acetate buffers made from CH3COONa and CH3COOH in the appropriate ratio. The pH was measured with a pH meter using a glass electrode. For the exclusion of oxygen,
J Biol Inorg Chem (2007) 12:681–690
99.996% argon gas and the Schlenck technique were used where necessary. [Fe(II)TPPS4–, 4Na+] The water-soluble porphyrin was synthesized in the equilibrium reaction of 1.5 · 10–4 M Fe(II) and 3 · 10–6 or 5 · 10–6 M free-base porphyrin in aqueous, argon-saturated solution at pH 6 and 7. The solutions also contained 0.3 M acetate buffer to adjust the pH and to trap trace amounts of iron(III) ions, and, in addition, 1 M NaCl at pH 7 to increase the ionic strength. In the first step of the process, the iron(II) and the acetate solutions were mixed and then immediately degassed with argon gas (for about 30 min). At pH 7, NaCl was also added in the first step. Subsequently, the mixture was kept under argon for 4 h. Then porphyrin was added, also under oxygen-free conditions. The system was allowed to react for at least 2 days at room temperature. The spectral change of the starting system (containing H2TPPS4–) indicates the formation of the complex (details in ‘‘Formation of Fe(III)TPPS3– and Fe(II)TPPS4– in water’’). Apparatus All the absorption and emission spectra were recorded with a Specord S100 UV–vis one-way spectrophotometer and with a PerkinElmer LS50 spectrofluorimeter, respectively. Nanosecond emission lifetimes were measured using PicoQuant time-correlated single photon counting apparatus. For the assessments and data fitting, deconvolution software was used (FluoFit from PicoQuant). For continuous photolysis at different wavelengths an AMKO LTI system (consisting of a 150-W high pressure Xe–Hg arc lamp and a monochromator) was utilized. The light intensity was determined with a thermopile calibrated by ferrioxalate actinometry. Quantum yield measurements were carried out with samples of nearly 100% light absorption. The concentration of the known porphyrin species in the solutions was determined by using their molar absorption coefficients, utilizing a least-squares fitting method on the 350–500-nm region of the spectra. Theoretical methods Density functional theory (DFT) was used for the theoretical determination of the structures and energies of iron(II) porphyrin and its derivatives containing other ligands in the singlet, triplet and quintet states. We used the B3LYP [35–37] combination of functionals as coded in Gaussian98 [38], in conjunction with the LANL2DZ [39–42] basis set. This combination proved to be adequate in numerous related applications in the literature [43]. We made tests
683
using other basis sets, including 3–21G* and 6–31G*, and found that the computed structures remain essentially the same. Where comparison was possible, our results agree well with those of earlier calculations [43, 44]. The molecular geometries presented are true minima as checked by calculation of vibrational frequencies. For the comparison of the energies of complexes with various spin multiplicity, the DSCF method was used; i.e., the structures and energies were calculated using the restricted self-consistentfield (SCF) procedure (RB3LYP) for the singlet species and the unrestricted method (UB3LYP) for the open-shell molecules, and the energies were directly compared. The convergence of both the SCF iteration and the geometry optimization was very slow, so level shifting, enforcing and relaxing symmetry and change of occupation of orbitals were necessary to satisfy tight convergence criteria.
Results Formation of Fe(III)TPPS3– and Fe(II)TPPS4– in water The formation of the porphyrin complexes was detected by spectrophotometry. Figure 1 shows the spectra of the complexes formed observed in an aqueous solution containing 3 · 10–6 M H2TPPS4– and 1.5 · 10–4 M iron(II) sulfate under two different conditions: without any buffer and with 0.3 M acetate buffer. In the absence of acetate buffer, the observed spectrum is identical with that of Fe(III)TPPS3– (which in aqueous solution exhibits peaks at 393, 528 and 651 nm [26, 45]). In the presence of acetate buffer we detected a different spectrum which is not identical with that of either iron(III) porphyrin or free-base porphyrin. The characteristics of this new spectrum are in very good agreement with those observed in the in situ reduction of iron(III) porphyrin [26]. That spectrum was assigned to iron(II) porphyrin. We are convinced this complex is formed in our reaction because there is no other metal ion coordinating to porphyrin in the solution except Fe2+. In the reaction of free-base porphyrin and any kind of iron(II) salt, ferric porphyrin was the final product in all cases reported so far. On the basis of the latter observations, it was believed that iron(II) porphyrin was formed in the first step but it was immediately oxidized to iron(III) porphyrin by molecular oxygen. However, when using the same system as in the earlier experiments, we removed even trace amounts of dissolved oxygen by the Schlenk technique (at least five degassing cycles), and the predominant part of the complex formed was iron(III) porphyrin. We conclude from this that iron(III) porphyrin cannot be the result of oxidation of iron(II) porphyrin by dissolved oxygen.
123
684
Fig. 1 Absorption spectra of the iron(II) porphyrin complex in a solution containing 0.3 M acetate buffer (red lines) compared with those of the free-base porphyrin (black lines) and the iron(III) porphyrin complex (blue lines). TPPS tetrakis(4-sulfonatophenyl) porphyrin
Since we found that in all preparation procedures of iron(III) porphyrins that have been reported in the literature the starting materials were iron(II) compounds and freebase porphyrin [26, 46, 47], we checked whether Fe3+ ions react with free-base porphyrin at all. We did not observe complexation between ferric ions and free-base porphyrin even when they were allowed to react for 1 week; hence, a reaction different from the reaction in Eq. 1, suggested in the literature, must be responsible for the formation of iron(III) porphyrin. We think that Fe2+ ions catalyze the reaction of freebase porphyrin and Fe3+ ions, trace amounts of which are always present in iron(II) solutions. It is known that formation of kinetically inert metalloporphyrins can be catalyzed by metal ions such as Hg(II), Cd(II), Pb(II) and Cu((II), via formation of kinetically labile sitting-atop complexes [48]. In the case of the relatively large ionic radius, the metal center cannot fit into the core of the porphyrin; hence, it is localized out of the ligand plane, distorting it. Such a distortion makes the porphyrin coordinatively more active. In our case, iron(II) porphyrin is formed in the first step and two new coordination positions arise on the other side of the porphyrin plane (on two diagonally situated nitrogen atoms) and become accessible for another metal ion, e.g., iron(III). Following the coordination of iron(III), it excludes the iron(II) ion bound more weakly on the other side of the porphyrin, and occupies the very center of the ligand. Scheme 1 is a simplified sketch of this latter stage in the catalytic process of the formation of iron(III) porphyrin (the distortion of the porphyrin ring is deliberately exaggerated). Since the order of magnitude for the concentration of the porphyrin in our experiments was only 10–6 M, in the absence of acetate buffer even trace amounts of iron(III), as impurities in the iron(II) starting material, proved to be
123
J Biol Inorg Chem (2007) 12:681–690
enough for the efficient formation of Fe(III)TPPS3–, which is a kinetically inert complex. In the presence of acetate buffer, however, most iron(III) ions are captured via formation of very stable polynuclear acetato complexes [49]. This is the main reason for the formation and stability of Fe(II)TPPS4– in the solutions buffered with acetate. An analogous ‘‘self-catalytic’’ effect of copper(I) was observed [50] in the case of the formation of a copper(II) porphyrin, where addition of metallic copper accelerated the synthesis by a factor of 104. In this case also a copper(I) complex formed in a synproportionation reaction plays the role of the catalyst [51]. An alternative route to the formation of iron(III) porphyrin could be an electron transfer from the iron(II) metal center (in the metalloporphyrin) to an approaching iron(III) ion from the bulk. Such a reaction cannot be excluded on the basis of the standard redox potentials, E(FeIII(H2O)n/ FeII(H2O)n) = 0.77 V [52] and E([FeIII(H2O)TPPS]3–/ [FeII(H2O)TPPS]4–) = 0.01 V [26]. However, we found that Fe(II)TPPS4– is not sensitive to oxygen in the presence of acetate ions (see later). If O2, which is a much stronger oxidizing agent than iron(III), cannot oxidize the iron(II) complex, we think that electron transfer from Fe(II)TPPS4– to Fe3+ is not favored either. The catalytic activity of larger metal ions, which cannot undergo oxidation (e.g., Hg2+, Cd2+) in the formation of regular metalloporphyrins (see above) also seems to support the substitution mechanism [48]. Accordingly, Hg2+ ion as a catalyst was successfully utilized in the preparation of the iron(III) complex of mesoBrTPPS6– [53]. It is worth noting that in the presence of a reducing agent, such as ascorbic acid, we got the same spectrum as in solutions containing acetate buffer, indicating the formation of Fe(II)TPPS4–. This result, however, does not necessarily mean that ascorbic acid reduces the iron(III) porphyrin formed; rather, it reacts with the traces of iron(III) ions,
Scheme 1 Simplified demonstration of the formation of ferric porphyrin from a temporarily existing ferrous porphyrin
J Biol Inorg Chem (2007) 12:681–690
685
forming a complex or reducing it to iron(II), and blocks them in a similar way to acetate ions. It should also be mentioned that in situ formation of iron(II) porphyrin was observed in the presence of nitrogen monoxide [54]. The absorption spectrum of Fe(II)TPPS4– in aqueous solution containing 0.3 M acetate buffer exhibits strong peaks at 421, 557 and 598 nm (for more detailed characterization see ‘‘Photophysical studies’’). This compound proved to be kinetically labile. Its formation reaction is described by Eq. 2. The apparent equilibrium constant (K¢) for this reaction can be expressed by the actual concentrations, according to the law of mass action (Eq. 3). Since constant pH was applied in all solutions studied, the effect of H+ concentration is not taken into account in this case. K
Fe2þ þ H2 TPPS4 $ Fe(II)TPPS4 þ 2Hþ K0 ¼
K [Fe(II)TPPS4 ] ¼ þ 2 [H ] [Fe2þ ][H2 TPPS4 ]
a0 ¼
[H2 TPPS4 ] 1 ¼ 4 4 [H2 TPPS ] þ [Fe(II)TPPS ] 1 þ K 0 [Fe2þ ]
ð2Þ ð3Þ
ð4Þ Thus, the partial molar ratio of the free-base porphyrin (a0) can be given as a function of Fe2+ concentration (Eq. 4). Hence, 1/a0 is in linear correlation with the Fe2+ concentration, and the slope of this straight line is K¢. Accordingly, K¢ can be estimated from the spectra of solutions containing constant porphyrin concentrations and various concentrations of iron(II), evaluating the absorbances in the range of the Soret bands (Fig. 2). Since in these solutions, in spite of the presence of the acetate buffer, also some iron(III) porphyrin was formed, during the evaluation the individual concentrations of H2TPPS4–, Fe(II)TPPS4–, Fe(III)TPPS3– and Fe2+ were determined by nonlinear
Fig. 2 Absorption spectra of solutions containing 1.5 · 10–6 M porphyrin, 0.3 M acetate buffer (pH 6) and various concentrations of iron(II) (designated by the CFe(II)/CTPPS values). Insert: 1/a0 versus Fe2+ concentration for the determination of K¢ (see the text for details)
least-squares fitting with the Solver procedure of a Microsoft Excel program, using the known individual molar absorbances at several wavelengths. Taking the actual values of the concentrations of H2TPPS4–, Fe(II)TPPS4– and Fe2+, we calculated 1/a0 for each solution and plotted it as a function of Fe2+ concentration. The slope of the line fitted to this plot gave the equilibrium constant, K¢ = 2.3±0.6 · 104 M–1. Photophysical studies The absorption spectra of the free-base ligand (H2TPPS4–) and the ferric and the ferrous porphyrins are shown in Fig. 1. In the case of porphyrins, the absorption bands with very high molar absorption coefficient can be assigned to p–p* transitions. It is known from the literature that the Q bands (first singlet excited state, S1) of the free-base porphyrins are split into Qx and Qy bands because of the symmetry reduction caused by the two protons on the imido nitrogens [55]; hence, this kind of splitting is expected to disappear in metalloporphyrins. A more detailed analysis of the absorption spectrum, however, indicates that in iron porphyrins the splitting of the Q bands does not totally disappear. This suggests that the four metal–nitrogen bonds are not completely identical. The ferric porphyrin’s B(0,0), or Soret, band (second singlet excited state, S2) is at 393 nm, the more intensive Qy bands are at 528 and 498 nm and the Qx bands are at 650 and 686 nm. The Soret and Qx bands are blueshifted, while the Qy peaks are redshifted compared with the corresponding ones of the free-base porphyrin (kB ¼ 412 nm, kQy ¼ 516; 555 nm, kQx ¼ 582, 635 nm): This regular (coplanar) paramagnetic porphyrin complex does not show any luminescence. The ferrous porphyrin displays an intensive B(0,0) absorption band at 421 nm and a weak B(1,0) band at 400 nm; the Q(0,0) and Q(1,0) bands are at 597 and 556 nm. Two further weak bands can be seen at about 520 and 480 nm which can be assigned as Qy bands. The Soret band is redshifted, while the Q bands are blueshifted compared with those of the free ligand. The absorption spectrum of Fe(II)TPPS4– is very similar to that of the ˚ deoxyhemoglobin [56], in which the central ion lies 0.4 A above the porphyrin plane [52]. Deviating from the corresponding iron(III) complex, the ferrous porphyrin shows characteristic emissions (Fig. 3). The intensive fluorescence bands (S0 ‹ S1) at 608 and 656 nm can be assigned to the transitions to different vibrational levels of the electronic ground state. Figure 4 shows the complete energy-level diagram of Fe(II)TPPS4– we propose. Since this complex does not phosphoresce, the diagram shows only estimated energy levels for the triplet excited states.
123
686
J Biol Inorg Chem (2007) 12:681–690
Fig. 3 Uncorrected emission spectra of H2TPPS4– (broken line) and Fe(II)TPPS4– (continuous lines) in 0.3 M acetate buffer (pH 6)
Fig. 4 Energy-level diagram and electron transitions of Fe(II)TPPS4– in water. The energies of the triplet states and the ligand-to-metal charge transfer (LMCT) state are only estimated
The quantum yields of fluorescence (FS1) for Q band and Soret (B) band excitations are shown in Table 1. The S1 fluorescence rate (kS1) and the corresponding radiative lifetime (sS1) for the Fe(II)TPPS4– and H2TPPS4– species were calculated for each complex using the measured decay time (sm) and quantum yield (FS1). It is seen that while the measured fluorescence lifetime of the iron(II) porphyrin complex decreases compared with that of H2TPPS4–, the radiative lifetime of the S1 excited state increases. This means that ferrous porphyrin can release the excitation energy also in some other, nonradiative ways. A weak emission was also observed at 433 nm upon excitation at 390 nm. This band can be assigned to the S0 ‹ S2 transition. This kind of emission is scarcely detected but it was also seen for the hydrophobic Zn(II)TPP and H2TPP species (H2TPP is tetraphenylporphyrin) [59],
123
as well as for water-soluble Tl(III)TPPS3– and H2TPPS4– [57]. The quantum yield of the emission from the S2 excited state (FS2) is extremely low owing to its very short lifetime. The moderately high quantum yields for the S1 emission (fluorescence) of Fe(II)TPPS4– (FS1 = 9.5 · 10–3 and 7.0 · 10–3), in accordance with the relatively long lifetimes (in the nanosecond range), may be surprising because metalloporphyrins with open-shell metal centers are generally nonfluorescent or extremely weakly fluorescent with quantum yields below 10–4 [60, 61]. However, even paramagnetic metallaporphyrins display considerable fluorescence as the examples of manganese(II) [62] and lanthanide(III) [63] porphyrins indicate (FS1 < 10–3). The rather low fluorescence quantum yields in the case of regular in-plane metalloporphyrins are attributed to a partial overlap of the eg(dp) orbitals of the ligand with the nd orbitals of the central metal ion, leading to increased spin–orbit interaction, which decreases the lifetimes of the excited states, promoting radiationless transitions [64]. In the case of water-soluble out-of-plane metalloporphyrins, however, owing to the larger distance between the ligand and the metal center, the electrons of the latter cause only a minor perturbation and the photophysical properties of these complexes are determined essentially by the porphyrin ring’s p electrons and the distortion of the plane, affecting the energy levels of the delocalized p and p* frontier orbitals. This interpretation has been confirmed by photophysical studies of several heavy-metal porphyrins, such as Pd(II)TPPS4–, Pt(II)TPPS4– [65], Hg(II)TPPS4– [66] and Tl(I)2TPPS4– [58]. The last two complexes are known to be characterized by a sitting-atop structure, i.e., the metal atom is out of the porphyrin plane. The fluorescence quantum yields for these complexes are of the same order of magnitude as for the freebase porphyrin. These water-soluble complexes have similar absorption spectra and exactly the same type of emission spectra as Fe(II)TPPS4– (the corresponding emission bands are at about the same wavelengths),
J Biol Inorg Chem (2007) 12:681–690
687
Table 1 Dynamic photophysical parameters of H2TPPS4- and Fe(II)TPPS4Complexes H2TPPS4– Fe(II)TPPS
4–
FS1(10–2)a
sm(ns)
kS1(106s–1)b
sS1(ns)c
Fdegradation(10–4)d
FS2(10–5)e
Stokes shifts (cm–1)
7.5 [57]
9.99
7.5
133
0.06 [58]
–f
253.6
0.95, 0.7
1.97
4.83
207
9.7, 7.8
6.3
322.4
TPPS tetrakis(4-sulfonatophenyl)porphyrin a
The quantum yields of emission from the S1 state were calculated using the emission quantum yield of a known reference material such as Ru(2,2¢-bipyridyl)32+ complex. The excitation wavelengths for Fe(II)TPPS4– at both the Q band and the Soret band, in this order
b
kS1 ¼ US1 =sm
c
sS1 ¼ kS11
d
Calculated from the photoinduced disappearance of the complex at the wavelengths for Fe(II)TPPS4– at both the Q band and the Soret band, respectively
e
The quantum yields of emission from the S2 state were calculated using the quantum yield of the S1 emission (FS1) and the ratio of the two measured peaks’ integral area
f
Extremely weak S2 fluorescence for H2TPPS4– was also observed, but its quantum yield was not determined [57]
indicating the similar structure of the latter. In accordance with the interpretation above, for Tl(III)TPPS3– [57] the fluorescence quantum yield is 2 orders of magnitude less than for the previous complexes. In this case, as a consequence of the smaller size of the metal center, it is close to in-plane position, resulting in stronger spin-orbit coupling. Photolysis of Fe(II)TPPS4– Irradiation of the iron(II) porphyrin at both the Soret band (421 nm) and a Q band (556 nm) resulted in the degradation of the complex itself (Fig. 5), accompanied by the formation of ring-opened tetrapyrrole derivatives (bile-type pigments) as indicated by the new bands at 390 and 450 nm [67]. The same types of products were obtained in photolyses of other sitting-atop porphyrin complexes of Tl(I), Tl(III) and Hg(II) [58, 66, 68] as well as in the case of photobleaching of H2TPPS4– [69]. The irradiated ferrous
porphyrin was transformed almost completely during a 40min illumination period. The quantum yields for the degradation are summarized in Table 1. The photochemistry of this complex is characterized by a porphyrin ligand-to-metal charge transfer (LMCT) reaction, similarly to the photoinduced reactions in the case of sitting-atop metalloporphyrins with a metal center other than iron(II) (Tl+, Hg2+) [57, 66]. Irradiation of H2TPPS4– under the same conditions results in degradation, the quantum yield of which is 2 orders of magnitude lower than that observed for the ferrous porphyrin (Table 1). In the absence of a metal center, no LMCT process can occur. In the case of the sitting-atop complexes, the LMCT process takes place in an indirect way, because appropriate absorption bands do not appear in the absorption spectra. Thus, the excited electron must come from one of the orbitals of the porphyrin ligand to the iron center. Since the rate of the phototransformation does not decrease significantly in the presence of an efficient triplet
Fig. 5 Photodegradation of Fe(II)TPPS4– (3 · 10–6 M) upon irradiation in 0.3 M acetate buffer in a period of 40 min
123
688
J Biol Inorg Chem (2007) 12:681–690
Scheme 2 Simplified demonstration of the mechanism for the photoinduced degradation of iron(II) porphyrin. For the sake of simplicity, the ionic substituents are designated by dashed lines)
quencher (dissolved molecular oxygen), the LMCT process must occur through the S1 and S2 singlet excited states (Fig. 4). In contrast, irradiation of the iron(III) porphyrin, under the same conditions, caused no permanent chemical change in the system, even if a similar LMCT reaction may take place in the primary photochemical step. The reason for this is that the regular in-plane metalloporphyrins can be reversibly oxidized [70], because the temporarily formed p cation is very stable and the electron promoted to the metal center has enough time to be transferred back to the porphyrin ligand. In contrast, in the case of sitting-atop porphyrins, after the LMCT reaction, the reduced metal ion, [Fe(I) in our case] can be easily detached because of the kinetic lability of this complex, and, thus, the electron cannot be transferred back. Scheme 2 demonstrates the assumed key step of the photoinduced reaction of the iron(II) porphyrin. Fe(II)TPPS4– as a possible heme model For Fe(II)TPPS4– to be useful as a water-soluble heme analog, it should be able to bind and release molecular oxygen. To see if this is possible, after the Fe(II)TPPS4– complex had been formed, the solution was saturated with oxygen. The ferrous species proved to be stable against oxidation; no conversion to the ferric form was observed even over weeks. Besides, after the saturation of the solution with oxygen, the spectra of the iron(II) porphyrin complex did not change. This means that either molecular oxygen does not coordinate to the ferrous porphyrin under
Table 2 The dissociation energy (kcal mol–1) obtained in density functional theory calculations for binary complexes of iron(II) porphyrin (FeP) with H2O, CH3COO– (Ac) and 3O2 as well as the energy needed to remove a H2O or 3O2 ligand from their ternary complexes containing Ac. See the text for details of the calculations FeP–H2O
FeP–Ac
FeP–O2
FeP–Ac–H2O
FeP–Ac–O2
S
26.6
45.3
11.4
13.7
–9.2
T
19.7
31.5
8.6
8.9
7.2
Q
14.9
36.3
12.4
11.2
0.1
123
such conditions, or this coordination does not cause spectral changes. A possible reason why O2 is not coordinated to Fe(II)TPPS4– is that the latter is more strongly coordinated by some other ligands present in the solution used for the synthesis of Fe(II)TPPS4–, so molecular oxygen cannot be connected to the iron ion. In order to see how strongly the most abundant ligands in the solution are bound to Fe(II)TPPS4–, we performed DFT calculations using iron(II) porphyrin as a model compound, denoted in the following as FeP. We explored in some cases how much the energetics is influenced by the presence of the four phenyl groups of TPP, and found that the changes are small and we can rely on the data obtained with the bare porphyrin ring. We investigated the binding energy of H2O, O2 and CH3COO– (denoted as Ac) to FeP in the singlet, triplet and quintet states. As the iron(II) porphyrin molecule is assumed to have two unoccupied coordination sites, not only its 1:1 complexes with each ligand were studied, but also all possible combinations of them. We found that for all complexes except FeP and FeP–Ac–O2 the quintet state is the ground state. Selected binding energies are listed in Table 2. In most cases the binding energies in ‘‘binary’’ FeP–L complexes are found to change monotonously with the spin multiplicity of the system. The acetate ion was found to form the strongest bond with the iron(II) center (its binding energy is 45–36 kcal mol–1 to the singlet and the quintet iron porphyrin, respectively). The binding energy of the first H2O ligand is much smaller, between 27 and 15 kcal mol–1, depending on the multiplicity. The binding energy of 3O2 to FeP is around 10 kcal mol–1. The second ligand is always connected to the opposite side of the porphyrin plane compared with the first ligand, and is generally much less strongly connected to the iron(II) center than in the absence of the other ligand. For example, the binding energy of the acetate ion in FeP–H2O.Ac is about 13 kcal mol–1 smaller than in FeP–Ac, in all three electronic states. In the binary FeP–H2O complex in all electronic states and in the singlet and triplet states of the ternary FeP–Ac–H2O complex, the H2O molecule, as expected, is connected to the iron through one of the lone pairs of the oxygen atom. In contrast, in the quintet state of
J Biol Inorg Chem (2007) 12:681–690
FeP–Ac–H2O the hydrogen atoms of the water molecule are closer to the FeP unit than its oxygen atom, and point towards two neighboring nitrogen atoms of the porphyrin frame, forming two O–H–N hydrogen bonds instead of a donor–acceptor bond with the iron. This indicates that in this state the iron(II) center is coordinatively saturated by the acetate ligand. The most stable binary complex is definitely FeP–Ac. We can assume that FeP is in the quintet ground state in solution. The coordination of H2O and acetate to FeP will very probably not change the spin multiplicity of the FeP complex, and they are bound to FeP by 15 and 36 kcal mol–1, respectively. The oxygen molecule is in the triplet state in solution, and it will not necessarily form the thermodynamically most stable complex (the quintet state). If 3O2 forms a triplet complex with FeP, its binding energy will be about 9 kcal mol–1; if a quintet complex is formed, about 13 kcal mol–1 will be released. This suggests that most FeP will be complexed by acetate, and the least probable first ligand is the oxygen molecule. Should ternary complexes be formed, H2O as a ligand will be preferred to O2, because it is much more strongly bound to FeP–Ac, especially in the quintet state. This means that in solutions with high acetate concentration—as in our case where it is used to mask the Fe3+ ions—there is very little chance that FeP, and very probably also Fe(II)TPPS4–, will be able to bind and release O2 as in heme. Fe(II)TPPS4– may be used as an oxygen carrier if one is able to mask Fe3+ with an additive that is much less strongly coordinated to the iron(II) center in Fe(II)TPPS4– than the acetate ion. Conclusion Preparation of a water-soluble, kinetically labile iron(II) porphyrin has been realized in neutral or weakly acidic solutions. The role of this complex in the synthesis of the corresponding iron(III) porphyrin has also been elucidated. Besides photophysical and photochemical characterization of Fe(II)TPPS4–, its possible oxygen-binding ability (as a heme model) has been studied as well. In the reaction of free-base porphyrin and iron(II) ions, ferric porphyrin was the final product even in deoxygenated solutions because from the temporarily formed, kinetically labile iron(II) porphyrins the Fe(II) centers were excluded by iron(III) ions existing or formed in these systems. Moreover, this is the only way for formation of iron(III) porphyrin because no direct reaction between freebase porphyrin and iron(III) occurs. From this one can also infer that iron(II) ions catalyze the formation of iron(III) porphyrins. The synthesis of the Fe(II)TPPS4– complex could only be realized in the presence of acetate buffer
689
capturing iron(III) ions. In buffered solutions, Fe(II)TPPS4– proved to be stable even in the presence of molecular oxygen. The photophysical properties of the iron(II) porphyrin complex consist of typical ‘‘sitting-atop’’ behavior (redshifted Soret absorption and blueshifted emission and Q absorption bands, irreversible photoinduced porphyrin LMCT reaction), while the corresponding iron(III) porphyrin is a regular one with no emission and photoredox properties. DFT calculations indicated that in solutions with high acetate concentration—as in our case where it is used to mask the Fe3+ ions—there is very little chance for iron(II) porphyrin to bind and release O2 as heme does in a hydrophobic microenvironment in hemoglobin. Fe(II)TPPS4– can probably be used to model heme in the presence of an iron(III)-trapping additive that is much less strongly coordinated to the iron(II) center than the acetate ion. Acknowledgement This work was supported by the Hungarian Research Fund (OTKA K63494, T049257).
References 1. Mathews CK, van Holde KE, Ahern KG (eds) (2000) Biochemistry. Addison Wesley Longman, San Francisco 2. Garrett RH, Grisham CM (eds) (1999) Biochemistry. Saunders College Publishing, Fort Worth 3. Reiss JG, LeBlanc M (1982) Pure Appl Chem 54:2383 4. Hammerschmidt DE, Vercellotti GM (1989) In: Chang TM, Geyer RP (eds) Blood substitutes. Dekker, New York, p 431 5. Lowe KC (1991) Vox Sang 60:129 6. Biro GP (1993) Transfus Med Rev 7:84 7. Faithful NS (1992) Adv Exp Med Biol 317:55 8. Faithful NS (1994) Biomater Artif Cells Immobil Biotechnol 22:181 9. Kaplan HR, Murthy VS (1975) Fed Proc 34:1461 10. Feola M, Gonzalez H (1983) Surg Gynecol Obstet 157:399 11. Savitsky JP, Doczi J (1978) Clin Pharmacol Ther 23:73 12. Bonhard K (1975) Fed Proc 34:1466 13. Rosen A, Sehgal L, Gould S (1986) Physiologist 29:161 14. Feola M, Simoni J (1990) Biomater Artif Cells Artif Organs 18:233 15. Chang CK, Traylor TG (1973) Proc Natl Acad Sci USA 70:2647 16. Leal O, Anderson DL, Bowman RG, Basolo F, Burwell RL (1975) J Am Chem Soc 97:5125 17. Collman JP (1977) Acc Chem Res 10:265 18. Basolo F, Hoffman BM, Ibers JA (1975) Acc Chem Res 8:384 19. Morgan B, Dolphin D (1987) Struct Bonding 64:115 20. Baldwin JE, Permutter P (1984) Top Curr Chem 121:181 21. Wuenschell GE, Tetreau C, Lavalette D, Reed CA (1992) J Am Chem Soc 114:3346 22. Collman JP, Herrman CP, Boitrel B, Zhang X, Eberspacher TA, Fu L (1994) J Am Chem Soc 116:9783 23. Shikama K (1988) Coord Chem Rev 83:73 24. Shikama K (1998) Chem Rev 98:1357 25. Kano K, Kitagishi H, Dagallier C, Kodera M, Matsuo T, Hayashi T, Hisaeda Y, Hirota S (2006) Inorg Chem 45:4448
123
690 26. Barley MH, Takeuchi KJ, Meyer TJ (1986) J Am Chem Soc 108:5876 27. Husza´nk R, Horva´th O (2005) J Chem Soc Chem Commun 224 28. Mitchell-Koch JT, Pietrzak M, Malinowska E, Meyerhoff ME (2006) Electroanalysis 18:551 29. Amao Y, Komori T, Tabuchi Y, Yamashita Y, Kimura K (2005) Sensor Lett 3:168 30. Balaji T, Sasidharan M, Matsunaga H (2005) Analyst 130:1162 31. Harriman A, Sauvage JP (1996) Chem Soc Rev 25:41 32. Scalise I, Durantini EN (2004) J Photochem Photobiol A 162:105 33. Nyman ES, Hynninen PH (2004) J Photochem Photobiol B 73:1 34. Chandrashekar TK, Ganesan S (2003) Photochem Photobiol 78:487 35. Becke AD (1993) J Chem Phys 98:5648 36. Lee C, Yang W, Parr RG (1988) Phys Rev B 37:785 37. Becke AD (1988) Phys Rev A 38:3098 38. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Zakrzewski VG, Montgomery JA, Stratmann RE, Burant JC, Dapprich S, Millam JM, Daniels AD, Kudin KN, Strain MC, Farkas O, Tomasi J, Barone V, Cossi M, Cammi R, Mennucci B, Pomelli C, Adamo C, Clifford S, Ochterski J, Petersson GA, Ayala PY, Cui Q, Morokuma K, Rega N, Salvador P, Dannenberg JJ, Malick DK, Rabuck AD, Raghavachari K, Foresman JB, Cioslowski J, Ortiz JV, Baboul AG, Stefanov BB, Liu G, Liashenko A, Piskorz P, Komaromi I, Gomperts R, Martin RL, Fox DJ, Keith T, Al-Laham MA, Peng CY, Nanayakkara A, Challacombe M, Gill PMW, Johnson B, Chen W, Wong MW, Andres JL, Gonzalez C, Head-Gordon M, Replogle ES, Pople JA (2002) Gaussian 98, revision A.11.4. Gaussian, Pittsburgh 39. Dunning TH, Hay PJ (1976) In: Schaefer HF III (ed) Modern theoretical chemistry, vol 3. Plenum, New York, p 1 40. Hay PJ, Wadt WR (1985) J Chem Phys 82:270 41. Wadt WR, Hay PJ (1985) J Chem Phys 82:284 42. Hay PJ, Wadt WR (1985) J Chem Phys 82:299 43. Dunietz BD, Dreuw A, Head-Gordon M (2003) J Phys Chem B 107:5623 44. Rovira C, Kunc K, Hutter J, Ballone P, Parrinello M (1997) J Phys Chem A 101:8914 45. Hoshino M, Ozawa K, Seki H, Ford PC (1993) J Am Chem Soc 115:9568
123
J Biol Inorg Chem (2007) 12:681–690 46. Fleischer EB, Palmer JM, Srivastava TS, Chatterjee A (1971) J Am Chem Soc 93:3162 47. Taniguchi VT (1978) PhD thesis, University of California, Irvine 48. Tabata M, Tanaka M (1985) J Chem Soc Chem Commun 42 49. Ciavatta L, Nunziata G, Sillen LG (1969) Acta Chem Scand 42:1637 50. Tabata M, Babasaki M (1992) Inorg Chem 2:5268 51. Tabata M, Miyata W, Nahar N (1995) Inorg Chem 34:6492 52. Shriver DF, Atkins PW (eds) (2002) Inorganic chemistry. Oxford University Press, Oxford, p 651 53. Xu X, Zhang H, Zhang C, Cheng J (1991) Anal Chem 63:2532 54. Fernandez BO, Lorkovic IM, Ford PC (2003) Inorg Chem 42:2 55. Rest A (1982) In: Coyle JD, Hill RR, Roberts DR (eds) Light, chemical change and life. The Open University Press, Walton Hall 56. Weissbluth M (1967) Struct Bonding 2:1 57. Valicsek Z, Horva´th O (2007) J Photochem Photobiol A 186:1 58. Valicsek Z, Horva´th O, Stevenson KL (2004) Photochem Photobiol Sci 3:669 59. Yu HZ, Baskin JS, Zewail AH (2002) J Phys Chem A 106:9845 60. Antipas A, Buchler JW, Gouterman M, Smith PD (1977) J Am Chem Soc 100:3015 61. Harriman A (1980) J Chem Soc Faraday Trans I 77:369 62. Harriman A (1980) J Chem Soc Faraday Trans I 76:1978 63. Tsvirk MP, Stelmakh GF, Pyatosin VE Solovyov KN, Kachura FF, Piskarsas AS, Gadonas RA (1986) Chem Phys 106:467 64. Kalyanasundaram K (1992) Photochemistry of porphyrin complexes. Academic, London 65. Kuba´t P, Mosinger J (1996) J Photochem Photobiol A 96:93 66. Horva´th O, Valicsek Z, Vogler A (2004) Inorg Chem Commun 7:854 67. Bensasson RV, Land EJ, Truscott TG (eds) (1983) Flash photolysis and pulse radiolysis. Pergamon, Oxford, p 47 68. Smith KM, Lai J-J (1980) Tetrahedron Lett 21:433 69. Rotomskis R, Streckyte G, Bagdonas S (1997) J Photochem Photobiol B 39:167 70. Felton RH (1978) In: Dolphin D (ed) The porphyrins. Academic, New York, p 53
