Critical Factors in Determining the Heterolytic versus Homolytic Bond

7 hours ago - Heterolytic versus homolytic cleavage of the metal bound terminal oxidant is the key for determining the nature of reactive intermediate...
10 downloads 9 Views 2MB Size
Subscriber access provided by TUFTS UNIV

Critical Factors in Determining the Heterolytic versus Homolytic Bond Cleavage of Terminal Oxidants by Iron(III) Porphyrin Complexes Sawako Yokota, and Hiroshi Fujii J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b13037 • Publication Date (Web): 25 Mar 2018 Downloaded from http://pubs.acs.org on March 25, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Critical Factors in Determining the Heterolytic versus Homolytic Bond Cleavage of Terminal Oxidants by Iron(III) Porphyrin Complexes Sawako Yokota and Hiroshi Fujii* Department of Chemistry, Graduate School of Humanities and Sciences, Nara Women’s University, Kitauoyanishi, Nara 630-8506, Japan ABSTRACT: Heterolytic versus homolytic cleavage of the metal bound terminal oxidant is the key for determining the nature of reactive intermediates in metalloenzymes and metal catalyzed oxygenation reactions. Here, we study the bond cleavage process of hypochlorite by iron(III) porphyrin complexes having 4-methoxy-2,6-dimethylphenyl (1), 2,4,6trimethylphenyl (2), 4-fluoro-2,6-dimethylphenyl (3), 2-chloro-6-methylphenyl (4), 2,6-dichlorophenyl (5), and 2,4,6trichlorophenyl (6) groups at the meso-position. Oxoiron(IV) porphyrin π-cation radical complexes (CompI) are characterized from the reactions of 1 ~ 4 with tetra-n-butylammonium hypochlorite (TBA-OCl) in dichloromethane at –80 °C while oxoiron(IV) porphyrin complexes (CompII) are characterized for 5 and 6 under the same conditions. For all of 1 ~ 6, we find the formation of an epoxidation product in good yields from the catalytic reactions with TBA-OCl, suggesting the heterolytic cleavages of the O-Cl bonds. CompI of 5 and 6 are reduced to the corresponding CompII by both chloride and hypochlorite while CompI of 1 ~ 4 are not. The reduction reactions with hypochlorite are much faster than those with chloride. These results provide a mechanism where the O–Cl bond of the iron-bound hypochlorite is cleaved heterolytically to form CompI for all of 1 ~ 6, but the subsequent reduction reaction with remaining hypochlorite affords CompII for 5 and 6. The E(OCl⋅/OCl–) is the boundary to discriminate the identity of the final product: CompI or CompII. The thermodynamic analysis based on the redox potential is successfully applied for explaining the bond cleavage processes of the hypochlorite, hydroperoxide, and t-butyl peroxide complexes.

INTRODUCTION The reactions of metal centers with terminal oxidants (XO) are key steps to generate reactive high valent oxointermediates in various metal catalyzed oxidation reactions and heme enzymes such as peroxidases, catalases and cytochromes P450.1-6 The reactions initially form terminal-oxidant adducts of metal complexes, followed by the cleavage of the O–X bonds of the metal bound terminal oxidants to produce the high-valent metal-oxo intermediates. Because of their significance, the cleavage processes of the O–X bonds have been studied by the reactions of various metal complexes with terminal oxidants such as hydrogen peroxide, alkylperoxides, peracids, and iodosylarenes.7-14 To date, there are two known types of the bond cleavage fashion (Figure 1). One is the heterolytic cleavage, in which the O–X bond is cleaved ionically to form an atomic oxygen and X–. In this case, the initial metal complex is oxidized to a high-valent metal oxospecies having two-electron equivalent more oxidized state. The other is the homolytic cleavage, in which the O–X bond is cleaved radically to form O⋅ and X⋅. In this case, the initial metal complex changes to a high-valent metal oxo-species having one-electron equivalent more oxideized state. For example, the heterolytic cleavage of the hydroperoxide adduct (X = OH) of iron(III) porphyrin is expected to form an oxoiron(IV) porphyrin π-cation

Figure 1. Heterolytic and homolytic bond cleavage processes of terminal-oxidant(O–X) adduct of iron(III) porphyrin complex.

radical complex (compound-I, CompI) and hydroxide (OH–), while the homolytic cleavage is anticipated to produce an oxoiron(IV) porphyrin complex (compound-II, CompII) and a hydroxyl radical, ⋅OH. The heterolytic cleavage would be more preferable than the homolytic cleavage in catalytic reactions for avoiding undesirable radical reactions. Previously, Groves et al. reported that the heterolytic cleavage of the O–O bond of the iron-bound mchloroperoxybenzoic acid (mCPBA) produces CompI in dichloromethane while the homolytic cleavage forms the iron(III) porphyrin N-oxide in toluene.10 The O–O bond cleavage has been also studied in aqueous and aprotic solvents. Traylor et al. proposed that the O–O bonds of H2O2 and t-BuOOH are heterolytically cleaved in methanol-dichloromethane mixture.7 On the other hand, Bruice et al. provided an evidence that the O–O bond is homolytically cleaved in aqueous and aprotic solvents and

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the generation of the heterolytic cleaved product would be due to the re-oxidation of the homolytic product by another ROOH or the radical species formed from the homolytic cleavage in solvent cage.8 The pH of the solvent has been shown to be a key factor of the O–O bond cleavage process.8,12 Nam et al. studied the bond cleavage processes of the O–O bonds on the basis of the yields of the catalytic epoxidation reactions.12 The main results are that (1) the electron-deficient iron porphyrin complex show a tendency to cleave the O–O bond more heterolytically than the less electron-deficient iron porphyrin complex, (2) the heterolytic cleavage is more preferable as the electron-withdrawing effect of the substituent of the iron-bound O–O bond is higher (mCPBA > H2O2 > tBuOOH), (3) an electron-donating axial imidazole ligand induces the homolytic cleavage more than a less electrondonating one. The results of the porphyrin and axial ligand effects, (1) and (3), seems to contradict the push-pull effect.4 On the other hand, as pointed out in their studies, one should pay attention to determine the bond cleavage fashion from the product yield of catalytic reaction because the catalytic product yield is affected not only by the bond cleavage fashion, but also by various factors for the reactive species produced in the catalytic reactions: e.g. the formation and reaction rates, the lifetime, and side reactions in the reaction mixture. This means that the high-valent metal oxo-species in the reaction mixture is not always identical to the one expected from the O–O bond cleavage fashion. Nam et al. also studied the bond cleavage process of the O–I bond of the iron-bound iodosylarene and showed that the counter anion of iron porphyrin is an essential factor to determine the bond cleavage process.14(a) This result is reasonably understood from a x-ray crystal structure of manganese(IV) salen iodosylmesitylene complex, which shows that the counter anion interacts with the manganese-bound O–I moiety and changes the property of the I–O bond.15 While these studies qualitatively clarified the factors affecting the bond cleavage processes of these oxidants, it has not been studied well for other terminal oxidants and the mechanism for determining the fashion of the bond cleavage, either the heterolytic or homolytic cleavage, has remained unclear. Hypochlorite (OCl–) has been used as a terminal oxidant in many catalytic oxidation reactions. For example, Jacobsen et al. showed highly enantioselective epoxidation reactions by using a chiral manganese(III) salen complex and sodium hypochlorite.16 The reactions have been proposed to be catalyzed by an manganese(V) oxo species generated by the heterolytic cleavage of the O–Cl bond. Meunier et al. reported manganese porphyrin catalyzed epoxidation and chlorination reactions by sodium hypochlorite.17 While there have been many studies of the catalytic oxidation reactions with hypochlorite, the bond cleavage process of the metal-bound hypochlorite has not been studied well. Recently, we succeeded in the preparation and characterization of the hypochlorite adducts of iron(III) porphyrin complexes.18 This progress allows us to study the bond cleavage process of the iron-

Page 2 of 13

bound hypochlorite spectroscopically. Here, we report the heterolytic versus homolytic bond cleavage of hypochlorite (OCl–) by iron(III) porphyrin complexes. The electron-withdrawing effect of porphyrin ligand on the bond cleavage process is studied by using iron(III) porphyrin complexes having various meso-phenyl groups, 1 ~ 6, shown in Figure 2. This study clearly shows that the O–Cl bond is cleaved heterolytically to produce CompI for all of 1 ~ 6, but the identity of the final product is changed to CompII by the subsequent reduction of the formed CompI with hypochlorite for 5 and 6. This conclusion is supported by the comparison of the redox potentials of CompI and hypochlorite. We also propose the thermodynamic analysis based on the redox potentials of CompI and a terminal oxidant (OX) to explain the bond cleavage process of hypochlorite, hydrogen peroxide, and t-butyl hydroperoxide. This analysis can be applicable to predict the heterolytic versus homolytic bond cleavage and the identity of high-valent metal oxo species characterized as the final product complex in the reaction mixture under various conditions.

Figure 2. Structures of iron porphyrin complexes, 1 ~ 6, used in this study

Results and Discussion Electron-withdrawing Effect of meso-Phenyl Group. To confirm the electron-withdrawing effect of the meso-phenyl groups of 1 ~ 6, we measured the redox potentials of CompI/CompII, E(I/II), at –40 °C.19 The CompII of 1 ~ 6, 1-CompII ~ 6-CompII, were prepared by published methods (see experimental section).20 The cyclic voltammograms and differential pulse voltammograms of 1-CompII ~ 6-CompII show reversible redox peaks corresponding to CompI/CompII redox couples (Figures S1 and S2) and the E(I/II) values are summarized in Table 1. The E(I/II) values increases in numerical order, 1 < 2 < 3 < 4 < 5 < 6.

ACS Paragon Plus Environment

Page 3 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Table 1. Redox Potentials (vs. SCE) and yields (%) of catalytic epoxidation reactions with TBA-OCl for 1 – 6 E(I/II) (V)

a

Yield (%) of cyclooctene oxb ide at 4 °C (–20 °C)

1

0.845

53 (50)

2

0.853

51

3

0.924

53

4

0.999

55 (56)

5

1.117

55

6

1.167

55 (59)

a

The E1/2 of ferrocene was 0.380 V in the present electrob chemical cell at –40 °C. The yields are based on TBA-OCl. The yields without catalyst are 7 % at 4 °C and 2 % at –20 °C. The yields are average values of three different experiments.

Reaction of 1 ~ 6 with Hypochlorite. Figure 3 shows absorption spectra of 1-Cl, 4-Cl, and 6-Cl and their spectral change upon the addition of 10 equivalents of TBAOCl in dichloromethane at –80 °C. The results for 2-Cl, 3Cl, and 5-Cl are also shown in Figure S3. The absorption spectra of the reaction products of 1-Cl ~ 3-Cl with TBAOCl, showing strong absorptions at around 670 nm, are identical to those of 1-CompI ~ 3-CompI prepared by the oxidation with ozone (Figure S4), respectively. These results indicate that the heterolytic cleavages occur for 1 ~ 3. Unique absorption spectral changes are observed for 4. With the addition of TBA-OCl, the absorption spectrum of 4-Cl initially changes to a new one with absorption peaks at 539 nm and 680 nm within a few minutes, and then slowly changes to a final spectrum with the absorption peak at 536 nm with disappearance of the peak at 680 nm over 1 hour (the red and green lines in Figure 3(b)). The spectrum of the intermediate is not identical to the spectrum of either authentic 4-CompI or 4-CompII (Figure S4 and S5). The absorption peak at 680 nm suggests the formation of CompI, but the intensity is lower than that of the authentic 4-CompI (Figure S4). When 50 equivalents of TBA-OCl were reacted with 4-Cl, the intensity of the 680 nm peak increased (Figure S6). Moreover, the spectrum of the final product was also obtained from the reaction of 4-CompI with TBA-OCl at –80 °C (Figure S7). These results indicate that 4-CompI is initially formed from the reaction with TBA-OCl, but the formed 4-CompI reacts with remaining TBA-OCl to give a decomposed complex. The decomposition process of 4 is faster than those of 1 ~ 3 because 4-CompI is more reactive than 1-CompI ~ 3-CompI due to the stronger electron-withdrawing meso-substituent.21 All of these results indicate that the heterolytic cleavage occurs even for 4. On the other hand, with addition of TBA-OCl, the absorption spectra of 5-Cl and 6-Cl initially change to new ones showing the absorption peaks at 542 and 544 nm (the blue lines in Figure 3(c) and S3(c)), followed by the final spectra with 552 and 558 nm with clear isosbestic

Figure 3. Absorption spectral change of (a) 1-Cl, (b) 4-Cl, and (c) 6-Cl after addition of 10 equivalent of TBA-OCl in dichloromethane at –80 °C. Black solid line: before reaction, pink line: immediately after addition of TBA-OCl, gray line: after addition of TBA-OCl, interval: (a) 3 s, (b) 50 s, (c) 240 s, red line: the reaction product with TBAOCl. Green line in (b): the final reaction product after 1 hour. Blue line in (c): the intermediate (the bishypochlorite complex) of 6 formed in the reaction. Inset: EPR spectra (4 K) of the reaction intermediates and products shown by the red, green, and blue lines in the absorption spectra.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

points, respectively. The absorption spectra of the intermediates are very close to that of iron(III) porphyrin bishypochlorite complex, (ClO)2, reported previously.18 The absorption spectra of the final products for 5 and 6 are almost identical to those of the authentic 5-CompII and 6-CompII, respectively (Figure S5). These results suggest that 5-CompII and 6-CompII are produced from the reactions of 5-Cl and 6-Cl with TBA-OCl via 5-(ClO)2 and 6(ClO)2. These assignments of the complexes produced from the bond cleavage processes were further confirmed by EPR spectroscopy (Figure 3 and Figure S8). The EPR spectra of the reaction products with TBA-OCl for 1 ~ 3 show strong EPR signals at g = 4.45, 3.55, and 1.98. These EPR spectra are identical to those of compound I22 and supports the heterolytic bond cleavages for 1 ~ 3. The compound-I like EPR signals are also observed for 4, but the intensity of the EPR signals is much lower than those of 1 ~ 3. In addition, the EPR signals disappear with progress in the reaction. These results support the heterolytic cleavage for 4 and the further decomposition of the formed 4CompI. The initial formations of 5-(ClO)2 and 6-(ClO)2 in the reactions of 5-Cl and 6-Cl with TBA-OCl are also confirmed by EPR spectra with small g-anisotropy (g = 2.253, 2.139, and 1.965 for 5-(ClO)2 and 2.254, 2.137, and 1.963 for 6-(ClO)2).18 Disappearance of the EPR signals with progress in the reactions is also consistent with the formation of CompII from the O–Cl bond cleavage of 5(ClO)2 and 6-(ClO)2. The very small signals at g = 2.0 observed for 5 and 6 may be due to organic radical species formed in the bond cleavage processes. Catalytic Reactions of 1 ~ 6 with Hypochlorite. The above results provide the following two possible mechanisms. One is that the bond cleavage process determines the final product; the O–Cl bonds of the iron-bound hypochlorites are cleaved heterolytically to form CompI for 1 ~ 4, but homolytically to generate CompII for 5 and 6. The other is that the reactivity of CompI to chloride or hypochlorite determines the final product; the O–Cl bonds are cleaved heterolytically for all of 1 ~ 6 to produce CompI, but the subsequent reactions of CompI with chloride or hypochlorite produce CompII as a finally characterized complexes for 5 and 6. To reveal the reaction mechanism, we examined the catalytic epoxidation reactions of cyclooctene with 1-Cl ~ 6-Cl and TBA-OCl. Previously, the heterolytic versus homolytic bond cleavages of the terminal oxidants have been studied by using yields of the epoxidation products of the catalytic reactions.7,8,12 These analyses are based on an assumption that CompI formed from the heterolysis of the O–O bond is a reactive compound to produce the epoxide from olefin, but CompII formed from the homolysis is not so reactive as to afford the epoxide. The yields of cyclooctene oxide from the catalytic reactions of 1-Cl ~ 6-Cl at 4 °C are listed in Table 1. For all of 1 ~ 6, we found good yields of cyclooctene oxide. Interestingly, the yields for 5 and 6, which afford CompII in the absence of substrate, are slightly higher than those of 1 ~ 4. Furthermore, similar results were obtained for the cat-

Page 4 of 13

alytic reactions at –20 °C, at which TBA-OCl itself cannot oxidize cyclooctene to the epoxide (Table 1). These results suggest that the reactions of 1-Cl ~ 6-Cl with TBAOCl initially generate CompI by the heterolytic cleavages of the O–Cl bonds. Since the epoxidation reactions of 5CompI and 6-CompI compete with their reduction reactions to CompII and self-decomposition reactions in their catalytic reactions, the moderate yields of cyclooctene oxide suggest that the epoxidation reactions would be faster than the reduction reactions and the decomposition reactions under the catalytic reaction conditions. In addition, since 5-CompI and 6-CompI are effectively stronger oxidants than other CompI, 5-CompI and 6CompI can transfer the oxygen atom to olefin more effectively than other CompI, resulting in slightly higher yields. Reaction of CompI with Chloride. To further investigate the mechanism, we examined the reactions of 1CompI ~ 6-CompI, prepared by other oxidant, with chloride at –80 °C. We examined the preparation of 1-CompI ~ 6-CompI from the oxidation of 1-Cl ~ 6-Cl with ozone gas at –80 °C. 1-Cl ~ 4-Cl were oxidized to 1-CompI ~ 4CompI, respectively, but 5-Cl and 6-Cl could not be oxidized to the corresponding CompI species. These results suggest the reactions of 5-CompI and 6-CompI with chloride. Stable 5-CompI and 6-CompI were prepared by the oxidation of the iron(III) porphyrin nitrate complexes of 5 and 6 by ozone at –80 °C. When 50 equivalents of tetra-n-butylammonium chloride (TBA-Cl) were added to 1-CompI ~ 6-CompI at –80 °C, no reactions are observed for 1-CompI ~ 4-CompI for 1 hour except selfdecomposition reactions to 1-Cl ~ 4-Cl, respectively (Figure S9). However, 5-CompI and 6-CompI are reduced to the mixture of the corresponding Cl and CompII during 1 hour (Figure S9). The reactions followed the first-order kinetics in the presence of 50 equivalents of TBA-Cl, and the rate constants depended on the concentration of TBA-Cl (Figure S10). The second-order rate constants were estimated to be 0.14 and 0.46 M-1s-1 at –80 °C for 5 and 6, respectively. These results indicate that 1-CompI ~ 4-CompI do not react with chloride, but 5-CompI and 6CompI slowly react with chloride. Reaction of CompI with Hypochlorite. We further examined the reactions of 1-CompI ~ 6-CompI with hypochlorite at –80 °C. 1-CompI ~ 3-CompI do not react with hypochlorite after the addition of 10 equivalents of TBA-OCl (Figure 4 and S11). 4-CompI slowly reacted with TBA-OCl and changed to a new compound having absorption peak at 536 nm for 10 min (blue line in Figure S11(c)). The new compound was identical with the decomposed compound formed from the reaction of 4-Cl with TBA-OCl (Figure S7(b)). The similar results were also obtained for the reactions of 1-CompI ~ 4-CompI prepared from the iron(III) porphyrin nitrate complexes; 1-CompI ~ 3-CompI did not react with hypochlorite, but 4-CompI formed the new compound. The reaction of 4CompI from the nitrate complex is much slower than that from the chloride complex. 4-CompI prepared from the nitrate complex changed to the new compound for 1

ACS Paragon Plus Environment

Page 5 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

hour. On the other hand, 5-CompI and 6-CompI, prepared from the iron(III) porphyrin nitrate complexes, produce the corresponding CompII immediately after the addition of 10 equivalents of TBA-OCl (Figure 4 and S11). 5-CompII and 6-CompII further change to new compounds having absorption peak at 538 nm (blue lines in Figures 4(c) and S11(d)), which are very close to the final product obtained from the reaction of 4-CompI with TBA-OCl. The reaction rate constants for the formations of 5-CompII and 6-CompII could not be measured with an absorption spectrometer because of extremely fast reactions, but are expected to be larger than 103 M-1s-1 from the reactions completed within a few second. Therefore, the reduction reactions with TBA-OCl are much faster than those with TBA-Cl. Overall, 1-CompI ~ 4-CompI are not reduced to the corresponding CompII by hypochlorite, but 5-CompI and 6-CompI are immediately reduced to the corresponding CompII. Although the absorption spectral features of the final compounds observed for 4 ~ 6 are close to that of iron(IV) bismethoxide complex, we need further spectroscopic study to identify them and to clarify the formation mechanism.

Figure 4. Absorption spectral change for the reactions of (a) 1-CompI and (b) 6-CompI with TBA-OCl in dichloromethane at –80 °C. Green line: CompI, red line: immediately after addition of 10 equiv of TBA-OCl, blue line: after 1000 s.

Mechanism of O–Cl Bond Cleavage. Although 5CompII and 6-CompII are detected as the final products from the reactions of 5-Cl and 6-Cl with TBA-OCl, the good yields of cyclooctene oxide from the catalytic reac-

tions with hypochlorite clearly support the formation of CompI for all of 1 ~ 6. The detection of 5-CompII and 6CompII is due to the subsequent reduction reactions with hypochlorite because the reduction reactions with hypochlorite are much faster than those with chloride for 5 and 6. These results provide a possible mechanism that the reactions of 1-Cl ~ 6-Cl with TBA-OCl initially produce CompI by the heterolytic cleavages of the O–Cl bonds, but the formed 5-CompI and 6-CompI are immediately reduced to the corresponding CompII by hypochlorite remaining in the reaction mixture when the substrate is not present (Scheme 1). In the presence of the substrate (e.g. cyclooctene), the reactions of 5-CompI and 6-CompI, formed from the heterolysis of the O–Cl bond, with cyclooctene compete with the reduction reactions with hypochlorite, resulting in the formation of cyclooctene oxide in moderate yields (Scheme 1). Scheme 1. A proposed mechanism for the reactions of iron(III) porphyrin complexes with hypochlorite.

The formation of CompII for 5 and 6 can be explained by the redox potential of hypochlorite. The standard potential of one-electron oxidation process of hypochlorite, E°(OCl⋅/OCl–), has been reported to be 1.39 V (1.15 V vs. SCE).23 This value is close to the E(I/II) values of 5CompI and 6-CompI and is altered by the solvent. The E(OCl⋅/OCl–) would be lower than the E(I/II) values of 5CompI and 6-CompI in dichloromethane at –80 °C, resulting in the rapid reduction to the corresponding CompII. The E(OCl⋅/OCl–) value is an essential factor in determining the identity of the finally characterized complex from the reaction of iron(III) porphyrin with hypochlorite in the absence of substrate; CompI is detected for the iron porphyrin complex with the E(I/II) lower than the E(OCl⋅/OCl–) while CompII is characterized for the iron porphyrin complex with the E(I/II) higher than the E(OCl⋅/OCl–). The standard potential of one-electron oxidation process of chloride, E°(Cl⋅/Cl–), has been reported to be 2.432 V (2.190 V vs. SCE).23 Since the E°(Cl⋅/Cl–) is much higher than E(I/II) in dichloromethane, chloride cannot reduce CompI to produce CompII and chlorine radical (Cl⋅) by the one-electron redox process. Therefore, another reduction reaction by chloride should be considered. One of the possible reactions is that chloride reduces CompI to yield CompII and chlorine molecule (Cl2) because the

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

standard potential of two-electron redox process of chloride, E°(Cl2/Cl–), has been known to be 1.358 V (1.116 V vs. SCE) and close to E(I/II).23 In fact, we observed an irreversible broad redox peak at 1.02 V vs. SCE in the CV of TBA-Cl in acetonitrile (Figure S1 and S2). In addition, we previously reported the formation of Cl2 from the reaction of CompI having pentafluorophenyl group with chloride.24 The reaction mechanism of CompI with chloride is not clear at present, but Cl2 may be produced via the coordination of chloride to the heme-iron center of CompI, followed by the intermolecular coupling of the iron-bound Cl to form CompII and Cl2. In the previous report, we found the coordination of chloride to the ironcenter before the formation of CompII in the absorption spectral change of the reduction reaction.24 The reaction of CompI with chloride hardly occurs in the presence of excess hypochlorite and substrate because this reaction is much slower than the reactions with hypochlorite and substrate. In fact, the second-order reaction rate constant of 6-CompI having nitrate axial ligand with cyclooctene at –80 °C is estimated to be 5.5 × 10 M-1s-1 (Figure S12) and that of 6-CompI having chloride axial ligand is expected to be 1-order magnitude greater than this rate.25 Under the catalytic conditions where the concentration of substrate is 1 ~ 3-order magnitude greater than that of hypochlorite, the reaction rate of the epoxidation reaction would be comparable to that of the reduction reaction of hypochlorite, providing the epoxidation product from the catalytic reactions. Thermodynamic Analysis of Bond Cleavage Processes. To gain further insight into the bond cleavage process, we considered the bond cleavage reactions from thermodynamic analysis. Scheme 2 shows thermodynamic analysis of the bond cleavage processes of the ironbound terminal oxidant (OX). The bond cleavage fashion of the iron-bound terminal oxidant should be determined by the relative stability between two states formed by the heterolytic and homolytic bond cleavages. The relative stability should depend on the free energies of reactions for the heterolytic and homolytic processes. Moreover, according to the Hess’s law, the relative stability can be estimated from the difference of the ionization potentials between CompII and X–, which are rationalized by their redox potentials, E(I/II) and E(X⋅/X–), using the relationship ∆G° = –nFE°, where n is the number of electron in the redox process, F is Faraday constant. When E(I/II) is lower than E(X⋅/X–), the state formed by the heterolytic cleavage is more stable than that by the homolytic cleavage and the O–X bond should be cleaved heterolytically. However, when E(I/II) is higher than E(X⋅/X–), the O–X bond should be cleaved homolytically.

Page 6 of 13

be changed by the solvent, the E°(Cl⋅/Cl–) is much higher than the E(I/II) of 1 ~ 6, listed in Table 1. Thus, the thermodynamic analysis of the iron-bound hypochlorite predicts that the state formed by the heterolytic cleavage is more stable than that by the homolytic cleavage, resulting in the formation of CompI for all of the iron porphyrin complexes in this study. This is consistent to the results of catalytic epoxidation reactions in this study. Scheme 2. Thermodynamic analysis of the bond cleavage process of the iron-bound terminal oxidant (O–X).

The present analysis would be applicable to the bond cleavage processes of other terminal oxidants. For example, the heterolytic bond cleavage of the O–O bonds of iron(III) hydroperoxide complex (X = OH) affords CompI and hydroxide while the homolytic cleavage yields CompII and hydroxyl radical. The standard potential of one electron redox process of hydroxide, E°(HO⋅/HO–), is calculated to be +1.800 V (+1.558 V vs. SCE) from E°(HO⋅/H2O) = +2.730 V and the pKa value of water.23 Although this value would be changed by the solvent, the E°(HO⋅/HO–) is much higher than the E(I/II) values of most iron porphyrin complexes (Figure 5). Therefore, the heterolytic cleavage of the O–O bond occurs more preferentially than the homolytic cleavage and CompI is generated from the iron(III) hydroperoxide complex even in the absence of proton (Scheme 3). As observed in the reactions of hypochlorite, the subsequent reduction reaction of CompI by hydrogen peroxide Scheme 3. Reaction Scheme for the bond cleavage of hydroperoxide complex.

For an iron(III) porphyrin hypochlorite adduct (X = Cl), the heterolytic cleavage forms CompI and chloride while the homolytic cleavage produces CompII and chlorine radical. According to the above discussion, the bond cleavage fashion should be determined by E(I/II) and E(Cl⋅/Cl–). The standard redox potential of one-electron redox process of chloride is reported to be E°(Cl⋅/Cl–) = +2.432 V (+2.190 V vs. SCE). 23 Although this value would

ACS Paragon Plus Environment

Page 7 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

has to be considered for some iron porphyrins and/or under some conditions. The standard potential of oneelectron process of hydrogen peroxide, E°(HO2⋅/H2O2), has been reported to be +1.46 V (+1.218 V vs. SCE) and in the range of the E(I/II).23 Therefore, hydrogen peroxide would become a possible reductant for CompI having high E(I/II) value, such as CompI of mesopentafluorophenylporphyrin (TPFPP): E(I/II) = 1.60 V (+1.358 V vs. SCE).19 The reaction of iron(III) complex of TPFPP would initially form CompI by the heterolytic cleavage of hydroperoxide, but the subsequent reduction reaction of CompI with hydrogen peroxide may occur to produce CompII in the absence of substrate. Furthermore, this reduction reaction would compete with the oxygenation reaction in the presence of substrate (Scheme 3). Even when the E(I/II) is lower than the E°(HO2⋅/H2O2), the one electron reduction of CompI with hydrogen peroxide would become a possible reaction in neutral and basic aqueous solutions. As shown in Figure 5, the E°(HO2⋅/H2O2) and E(I/II) shift to lower potential as the pH becomes higher. In addition, since the pKa of CompII is smaller than that of hydrogen peroxide, the E(I/II) value becomes higher than the E°(HO2⋅/H2O2) value in neutral and basic aqueous solutions. In this case, although the O–O bond is heterolytically cleaved, CompII may be identified from the reactions of most iron(III) porphyrin complexes with hydrogen peroxide. This is consistent to the previous report that CompII is produced from hydrogen peroxide in basic conditions.7,8 In heme enzymes, the terminal oxygen atom of the iron-bound hydroperoxide makes hydrogen bonding interaction with amino acid residues and water molecules in the distal side.4 This interaction is called the pull-effect and has been proposed to assist the heterolytic cleavage of the O–O bond to generate CompI and water. The present analysis also provides reasonable explanation for the pull effect. As mentioned in above paragraph, E°(HO⋅/H2O) is about 900 mV higher than E°(HO⋅/HO–). This means that the state formed by the heterolytic cleavage is stabilized much more than the state formed by the homolytic cleavage by the distal hydrogen bonding interaction: the pull effect. According to the linear free energy relation, this drastic stabilization would lower the activation energy of the heterolytic bond cleavage process more than the homolytic cleavage process, accelerating the heterolytic O–O bond cleavage reaction, i.e. the formation of CompI. We further apply this analysis to the O–O bond cleavage process of the iron-bound t-butyl peroxide (Scheme 4). The redox potential of one-electron redox process of t-butoxide has been calculated to be E°(tBuO•/tBuO–) = – 0.058 V (–0.30 V vs. SCE) in acetonitrile and much lower than E(I/II).25 Therefore, the state formed by the homolytic cleavage is more stable than that by the heterolytic cleavage, and CompII would be generated from the tbutyl peroxide complex in aprotic solvent. However, the bond cleavage fashion would be altered in the presence of proton, such as the reactions of iron(III) porphyrins with t-butyl hydroperoxide and the reactions in aqueous solu-

Figure 5. Dependence of E(HO⋅/H2O), E(HO2⋅/H2O2), and E(I/II) on the pH of aqueous solution. The following parameters are used for the preparation; E(I/II) = 1.0 V, pKa(CompII) = 5.0, pKa(CompI) < 0, pKa(H2O2) = 11.7, T = 298 K. The heterolytic cleavage is expected at pH from 0 to 14 (the area drawn by pale blue). The regions drawn by green and red indicate the pH regions expected to form CompI and CompII as final products, respectively.

tion. Proton should be taken into account in the bond cleavage reaction. The bond cleavage fashion is sensitive to the pKa value of t-butanol in solvent used for the reaction. E°(tBuO•/tBuOH) can be calculated to be +2.59V (+2.35 V vs. SCE) in acetonitrile from the pKa value of 44.8 and much higher than E(I/II).26 Therefore, the state formed by the heterolytic cleavage becomes more stable than that by the homolytic cleavage, resulting in the formation of CompI from the reaction of an iron(III) porphyrin with t-butyl hydroperoxide in acetonitrile. However, E°(tBuO•/tBuOH) can be calculated to be +1.07 V (+0.828 V vs. SCE) in water from the pKa value of 19 if we assume that E°(tBuO•/tBuO–) in water is the same as that in acetonitrile.26 This value is smaller than E(I/II) of most iron porphyrins, resulting in the homolytic cleavage in water. In a water-organic solvent mixture, the pKa value of t-butanol, and thus E°(tBuO•/tBuOH), becomes higher, as the content of organic solvent is higher. When E(tBuO•/tBuOH) becomes higher than E(I/II) in a waterorganic solvent mixture, the heterolytic cleavage occurs. Moreover, the E(tBuO•/tBuOH) and E(I/II) are changed by the pH of the solvent and the pKa value of CompII, respectively, the bond cleavage fashion is also altered by these factors. These predictions deduced from the proposed mechanism are consistent with the previous experimental results.7,8 The subsequent reduction reaction of CompI to CompII with excess t-butyl hydroperoxide is also a possible reaction. The redox potential of one-electron redox process of t-butyl hydroperoxide has been estimated to be

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Scheme 4. Reaction Scheme for the bond cleavage of t-butyl peroxide complex.

E(tBuO2•/tBuO2H) = +1.47 V vs NHE (+1.228 V vs. SCE) in water at pH = 0.27 E(tBuO2•/tBuO2H) would also be changed by the content of organic solvent and the pH of solvent, as discussed for E°(tBuO•/tBuOH). The reduction of CompI to CompII with t-butyl hydroperoxide would be a possible reaction for CompI having high E(I/II) value and in neutral and basic aqueous solution.7,8,12 After all, the bond cleavage fashion of t-butyl hydroperoxide and the identity of the final product are changed by the solvent and pH conditions. We show one example of the present thermodynamic analysis for the prediction of the bond cleavage process of t-butyl hydroperoxide in a water-organic solvent mixture (Figure 6). Under the conditions described in Figure 6,

t

t

Figure 6. Dependence of E( BuO•/ BuOH), t t E( BuO2•/ BuO2H), and E(I/II) on the pH in a waterorganic solvent mixture. The following parameters are t t used for the preparation; E°( BuO•/ BuOH) is 1.50 V, t t E( BuO2•/ BuO2H) = +1.47 V, E(I/II) = 1.40 V, pKa(CompII) t = 4.0, pKa(CompI) < 0, pKa( BuO2H) > 14, T = 298 K. The area drawn by pale blue and pale yellow indicate the pH ranges expected the heterolytic and homolytic cleavages, respectively. The regions drawn by green and red indicate the pH regions expected to form CompI and CompII as final products, respectively.

Page 8 of 13

the line of the dependence of the E(I/II) intersects that of the E(tBuO•/tBuOH) at pH = ~ 5.5. Therefore, the heterolytic cleavage is expected at pH < 5.5 (the area shown by pale purple) because the E(tBuO•/tBuOH) is larger than E(I/II) while the homolytic cleavage is expected at pH > 5.5 (the area shown by pale yellow) because the E(I/II) is larger than the E(tBuO•/tBuOH). Moreover, the identity of the final compound in the absence of substrate may be switched from CompI to CompII at pH = ~5.0 because the line of the dependence of the E(I/II) intersects that of the E(tBuO2•/tBuO2H) at pH = ~ 5.0. In the pH range from 5.0 to 5.5, the O–O bond of t-butyl hydroperoxide is heterolytically cleaved to form CompI, but CompII may be characterized from the reaction because of the subsequent reduction reaction of CompI with t-butyl hydroperoxide. In addition, under the catalytic reaction conditions where large excess of t-butyl hydroperoxide and substrate are present, the oxygenation reaction of CompI with substrate would compete with the reduction reaction in this range. The parameters used here are reasonable values for an iron porphyrin complex having the odifluorophenyl or o-dichlorophenyl group at the meso position.19,28 These analyses are consistent with the previous result that the high valent oxo-species generated by the bond cleavage of the O–O bond of 2-methyl-1phenylpropan-2-yl hydroperoxide is switched from CompI to CompII at pH = 4 ~ 5 in a solvent mixture of watermethanol-acetonitrile (5:2:3).12 To further confirm these discussions, we investigated the reactions of iron(III) TPFPP complex with t-butyl hydroperoxide in organic solvents. From the thermodynamic analysis, we expect that the reaction of iron(III) TPFPP complex with t-butyl hydroperoxide leads to the homolytic cleavage of the O–O bond to form CompII in the presence of a base (in the absence of proton), but the heterolytic cleavage to form CompI, followed by the reduction to CompII by the subsequent reaction with tbutyl hydroperoxide in the presence of proton (methanol). Moreover, we expect that the catalytic epoxidation reaction with cyclooctene affords cyclooctene oxide in the presence of proton because of the heterolysis of the O–O bond, but does not produce the epoxide in the presence of base because of the hemolysis of the O–O bond. The absorption spectra of the reactions in the presence of tetra-n-butylammonium hydroxide (TBA-OH) and in methanol indicate the immediate formation of CompII at –40 °C (Figure S13). The catalytic reaction with cyclooctene in the presence of TBA-OH at room temperature does not afford cyclooctene oxide (yield: 0 % based on t-butyl hydroperoxide), but that in the presence of methanol produces cyclooctene oxide in good yield (yield: 51 % based on t-butyl hydroperoxide). These results are consistent with the predictions and support our proposed mechanism. It should be pointed out here that the observation of CompII from the reaction of iron porphyrin with a terminal oxidant does not necessarily indicate the homolytic cleavage of the O-X bond because it can be produced by the subsequent reduction reaction of CompI formed from

ACS Paragon Plus Environment

Page 9 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

the heterolytic cleavage of the O–X bond. In addition, the subsequent reduction reaction of CompI usually competes with the oxygenation reaction in the presence of substrate when the E(I/II) is higher than the redox potential of the terminal oxidant. Therefore, while the high yield of the reaction product suggests the heterolytic cleavage of the O-X bond (the formation of CompI), a low yield of the reaction product does not necessarily indicate the homolytic cleavage. Table 2. Apparent Rate Constants for the Reactions of 1 ~ 6 with TBA-OCl (1 mM) at –80 °C. Cl → (OCl)2

Cl → CompI -1

1

(OCl)2 → CompI

-1

(s )

-1

(s )

(s )

-1

(1.14 ± 0.05) × 10

2 (1.11 ± 0.04) × 10-1 -

3 (3.76 ± 0.28) × -2 10

(1.92 ± 0.07) × 10

1

-

4

(9.51 ± 0.42) × 10

3

5

-

(7.66 ± 0.37) ×10

2

6

-

(1.46 ± 0.09) ×10

2

-3

(3.26 ± 0.11) × 10

-

(1.40 ± 0.09) × 10

3

Kinetics of the O–Cl Bond Cleavage. We examined the kinetics of the bond cleavage processes of hypochlorite by iron(III) porphyrins. The time courses of the absorption for the bond cleavage processes are summarized in Figure S14. The absorption spectral changes with clear isosbestic points indicate that no intermediates, such as mono-hypochlorite complex, OCl, and bishypochlorite complex, (OCl)2, are accumulated in the reactions of 1-Cl and 2-Cl with TBA-OCl (Figures 3 and S3). The time courses of the absorbance for the reactions can be fitted with a single exponential functions, providing apparent reaction rate constants from 1-Cl and 2-Cl to 1-CompI and 2-CompI, respectively. The reaction of 3-Cl with TBA-OCl also affords 3-CompI, but the spectral change does not afford clear isosbestic points (Figure S3(b)). The time course of the absorbance cannot be simulated well with a single exponential function, but with a double exponential function, suggesting two pathways from 3-Cl to 3-CompI. The time of the absorbance at 680 nm for 4 can be simulated with a single-exponential function. The absorption spectral change for the reactions of 5-Cl and 6-Cl with TBA-OCl indicate the formation of 5(ClO)2 and 6-(ClO)2, followed by the formation of 5CompII and 6-CompII, respectively (Figures 3(c). and S3(c)). The time courses of the absorbance for the consecutive reactions can be simulated well with a double exponential function, providing the reaction rates from 5Cl and 6-Cl to 5-(ClO)2 and 6-(ClO)2, and from 5-(ClO)2 and 6-(ClO)2 to 5-CompII and 6-CompII, respectively.

Furthermore, these assignments were confirmed by the dependence of the rate constant on the concentration of TBA-OCl; the rate from 6-Cl to 6-(ClO)2 depends on the concentration of TBA-OCl, but that from 6-(ClO)2 to 6CompII does not (Figure S15). Based on the present results, we discuss the reaction mechanism (Scheme 5). The reaction of ferric chloride complex with hypochlorite initially forms an undetectable mono-hypochlorite complex (OCl). The absorption spectral change with clear isosbestic points and the pseudofirst order kinetics for 1 and 2 indicate that the ratelimiting step is the binding of hypochlorite to the iron center: Cl → OCl. Therefore, 1-CompI and 2-CompI would be formed via 1-OCl and 2-OCl, respectively. With an increase in the electron-withdrawing effect of the meso-substituent, the reaction pathway via the bishypochlorite complex would be more dominant. 3CompI would be produced via both 3-OCl and 3-(OCl)2, resulting in the disappearance of the isosbestic points and biphasic kinetics, as shown in Figure S3(b) and S14(c). The fast reaction process would be assignable to the process via 3-(OCl)2 because the strong axial ligand effect has been know to accelerate the formation of CompI.11 For more electron-deficient iron porphyrin (4 ~ 6), the reactions via the bis-hypochlorite complexes are main reaction pathways. 5-(OCl)2 and 6-(OCl)2 are observed as metastable intermediates in the reactions. Although 4(OCl)2 cannot be detected in the reaction of 4-Cl with TBA-OCl, but 4-(OCl)2 is characterized as an initial intermediate in the reaction of iron(III) porphyrin complex having a weakly binding axial anion such as nitrate (Figure S16). The heterolytic cleavage occurs for 4-(OCl)2 ~ 6(OCl)2, but 5-CompII and 6-CompII are characterized as the final products. Since 5-CompI and 6-CompI are not detected in their absorption spectral change, the bond cleavage step of 5-(OCl)2 and 6-(OCl)2 are the ratelimiting steps. The estimated reaction rate constants are summarized in Table 2. As shown in Table 2, both of the binding step of hypochlorite to the iron-center and the heterolytic bond cleavage step become slower as the electron-withdrawing effect of the meso-substituent is stronger. Scheme 5. Proposed reaction mechanism for the reaction of chloroiron(III) Porphyrin complex with hypochlorite

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In summary, we studied here the bond cleavage processes of hypochlorite by iron(III) porphyrin complexes of 1 ~ 6. This study showed that the O–Cl bonds of the ironbound hypochlorites are cleaved heterolytically to generate CompI for all of 1 ~ 6, but CompII is produced as a final product by the subsequent reduction of CompI by hypochlorite for 5 and 6. The essential factor in determining the final reaction compound, CompI or CompII, is whether the E(I/II) is larger or smaller than E(OCl⋅/OCl–). We propose a new thermodynamic analysis based on the comparison between the redox potentials of CompI and the terminal oxidant for predicting of the bond cleavage of terminal oxidant. The bond cleavage processes of hypochlorite, hydroperoxide, and t-butyl peroxide are successfully explained by the thermodynamic analysis. The heterolytic versus homolytic cleavage of the metal-bound terminal oxidant (XO) can be predicted by the comparison of E(I/II) with E(X⋅/X). In addition, this analysis predicts that, in spite of the heterolyic cleavage, CompII may be identified as the final product, due to the reduction of the formed CompI with a terminal oxidant, when the E(I/II) is higher than the one-electron oxidation potential of the terminal oxidant, E(OX⋅/OX).

Experimental Section Instrumentation. UV-visible absorption spectra were recorded on an Agilent 8453 spectrometer (Agilent Technologies) equipped with a USP-203 low-temperature chamber (UNISOKU). 1H NMR spectra were measured on a Lambda-400 spectrometer (JEOL). The chemical shifts were referenced to the residual peaks of the deutrrated solvents: dichloromethane (5.32 ppm) and chloroform (7.24 ppm). The concentrations of NMR samples were 1 ~ 3 mM. EPR spectra were measured on EMX Plus continuous-wave X-band spectrometer (Bruker). For EPR measurements at 4 K, an ESR910 helium-flow cryostat (Oxford Instruments) was used. The following parameters were commonly used for EPR measurements, microwave frequency: 9.46 GHz, modulation frequency: 100 kHz, modulation amplitude: 10 gauss, microwave power: 1 mW, time constant: 81.9 ms, receiver gain: 30 dB, sample concentration: ~1mM. The cyclic voltammograms and differential pulse voltammograms were measured with an ALS612A electrochemical analyzer (BAS) in degassed acetonitrile containing 0.1M tetra-n-butylammonium perchlorate (TBAP) as a supporting electrolyte. A platinum electrode was used as the working electrode and a platinum-wire electrode was employed as the counter-electrode. The potentials were recorded with respect to a saturated calomel electrode (SCE) as the reference electrode. The reaction product analyses were performed using a gas chromatograph mass-spectrometer GCMS QP-2010 SE (Shimadzu). Ozone gas was generated by the UV irradiation of oxygen gas (99.999%) with an ozone generator PR-1300 (Clear Water) and used without further purification. Materials. Anhydrous organic solvents were obtained commercially and stored in a glove box. Dichloromethane was purified by passing through alumina column just before use in the glove box. Sodium hypochlorite pentahydrate (NaOCl・5H2O) was purchased from Wako (Japan) and used without further purification. Tetra-n-butylammonium chloride (≥ 99.0 %) was obtained from Aldrich. t-Buthyl hydroperoxide (5 M solution in decane) was purchased from Aldrich. The concentration of tButhyl hydroperoxide was determined by the iodometric titration.7a Other chemicals were purchased commercially and used

Page 10 of 13

without further purification. 4-Methoxy-2,6dimethylbenzaldehyde was synthesized by the Vilsmeier reaction of 3,5-dimethylanisole, followed by the purification by the 4-fluoro-2,6recrystallization in hexane.29 dimethylbenzaldehyde was prepared by the bromination of 5fluoro-1,3-dimethylbenzene, followed by the formylation with dimethylamide and n-butyl litium.30 TBA-OCl in dichloromethane. Excess amount of solid NaOCl ・5H2O was added to 0.1 M dichloromethane solution of tetra-nbutylammonium chloride at room temperature and the mixture was stirred vigorously. The progress of the exchange reaction was monitored by the absorption spectroscopy. TBA-OCl formed from the anion exchange reaction showed the absorption peak at 300 nm. Concentration of TBA-OCl was calculated from the intensity of the absorbance at 300 nm with ε = 300 M-1cm-1. After decantation, the TBA-OCl solution was cooled with ice. The TBA-OCl solution was prepared every time, just before use, because it gradually decomposed in dichloromethane. meso-Tetraarylporphyrins. meso-Tetraarylporphyrins (free base of iron porphyrins 1 ~ 6) were prepared from pyrrole and the corresponding benzaldehydes according to a previously published procedure.31 There porphyrins were purified by silica gel column with dichloromethane as an eluent. Spectroscopic data. UV-vis (nm) in dichloromethane. 1: 419, 515, 548, 591, 647. 2: 418, 515, 547, 592, 650. 3: 417, 513, 545, 590, 646. 4: 417, 513, 544, 589, 645. 5: 418, 513, 589, 650, 704. 6: 419, 513, 589, 651, 705. 1H NMR (400 MHz, ppm from TMS) in CDCl3 at 298 K: 1: 8.62 (py-H), 1.85 (o-CH3), 6.99 (m-H), 4.04 (p-OCH3), -2.54 (NH). 2: 8.38 (pyH), 1.85 (o-CH3), 7.30 (m-H), 2.60 (p-CH3), -2.51 (NH). 3: 8.60 (py-H), 1.85 (o-CH3), 7.18, 7.15 (m-H), -2.56 (NH). 4: 8.62 (py-H), 7.68, 7.66, 7.63, 7.62 and 7.60 (m-H), 7.53 and 7.51 (p-H), 1.94 and 1.92 (o-CH3), -2.52 (NH). 5: 8.64 (py-H), 7.76 (m-H), 7.78 (p-H), 2.56 (NH). 6: 8.66 (py-H), 7.82 (m-H), -2.64 (NH). Ferric chloride complexes. Ferric chloride complexes of 1 ~ 6 were prepared by insertion of iron into free base porphyrins of 1 ~ 6 with FeCl2 and sodium acetate in acetic acid, respectively, and purified with a silica gel column.21 Spectroscopic data. UVvis (nm) in dichloromethane. 1: 377, 422, 508, 578, 662, 696. 2: 376, 419, 509, 577, 661, 696. 3: 374, 417, 506, 579, 655, 690. 4: 372, 418, 506, 581, 653, 683. 5: 374, 419, 508, 582, 648. 6: 360, 418, 507, 580, 645. 1H NMR (400 MHz, ppm from TMS) in CDCl3 at 298 K: 1: 79.8 (py-H), 15.2 and 13.7 (m-H), 4.86 (p-OCH3), 6.45 and 3.66 (o-C H3). 2: 79.5 (py-H), 15.9 and 14.3 (m-CH3), 4.14 (p- CH3), 6.46 and 3.91 (o- CH3). 3: 80.4 (py-H), 14.9 and 13.5 (m-H). 4: 81.0 (py-H), 15.2, 14.3, 13.7 and 13.0 (m-H), 8.22 (p-H). 5: 79.7 (py-H), 13.9 and 12.6 (m-H), 8.17 (p-H). 6: 80.1 (py-H), 13.8 and 12.6 (mH). Ferric nitrate complexes. Ferric nitrate complexes. Ferric nitrate complexes of 1 ~ 6 were synthesized from reaction of ferric chloride complexes with 2 equivalent of silver(I) nitrate, respectively, and purified by recrystallization from dichlorometanehexane.32 Spectroscopic data. UV-vis (nm) in dichloromethane. 1: 333(sh), 414, 515, 578, 660. 2: 333(sh), 412, 514, 578, 656. 3: 338(sh), 512, 579, 655. 4: 339(sh), 412, 512, 578, 650. 5: 339(sh), 412, 512, 580, 645. 6: 340(sh), 411, 510, 575, 642. 1H NMR (400 MHz, ppm from TMS) in CD2Cl2 at 298 K: 1: 72.6 (py-H), 15.5 and 14.5 (mH), 4.92 (p-OCH3), 6.13 and 3.95 (o-CH3). 2: 73.0 (py-H), 16.4 and 15.2 (m-H), 4.28 (p-CH3), 6.14 and 3.99 (o-CH3). 3: 74.6 (py-H), 15.3 and 14.3 (m-H), 6.04 and 3.95(o-CH3). 4: 76.8, 75.4, 74.0 and 72.5 (py-H), 15.5, 14.6 and 13.5 (m-H), 8.24 (p-H). 5: 76.7 (py-H), 14.3 and 13.4 (m-H), 8.14 (p-H). 6: 77.1 (py-H), 14.1 and 13.3 (m-H). Reactions of Ferric Porphyrin with Hypochlorite. Chloroiron(III) porphyrin complex (100 μM) in dichloromethane was placed in a 1 cm quartz cuvette in a low-temperature chamber set on a UV-visible absorption spectrometer. After stabilization of solution temperature at –80 °C, 10 equivalent TBA-OCl

ACS Paragon Plus Environment

Page 11 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

was then added to the solution with stirring, and the reactions were monitored by the absorption spectral change at constant time intervals. The reaction rate constants were estimated from the simulations of time courses of the absorbance at 670 nm for 1, 666 nm for 2, 670 nm for 3, 680 nm for 4, 552 nm for 5, and 558 nm for 6 by commercial software, Igor Pro version 6.37 (WaveMetrics), with single- or double-exponential function. Catalytic Epoxidation Reactions of Cyclooctene. Dichloromethane solution containing iron(III) porphyrin chloride complex (1-Cl ~ 6-Cl) was cooled to 4 °C . Cyclooctene was dded to the solution. The catalytic reaction was initiated by addition of TBA-OCl. The initial concentrations of iron porphyrin, cyclooctene, and TBA-OCl were 10 µM, 1.0 M, and 3.0 mM, respectively. The reaction mixture was stirred for 10 min. The reaction was quitted by addition of tetra-n-butylammonium iodide (30 mM). The reaction mixture was passed through silica gel short column. After addition of standard sample (n-dodecane) to the eluent, the products and their yields were analyzed by GC-MS. The catalytic reactions at –20 °C were carried out in a temperature controlled cooling bath under the following conditions, iron porphyrin: 10 µM, TBA-OCl: 3.0 mM, cyclooctene: 10 mM. Reactions of CompI with Chloride. Iron(III) porphyrin chloride or nitrate complex (0.1 mM) in dichloromethane was placed in a 1 cm quartz cell. The cell was set on the lowtemperature chamber and cooled to –80 °C. After stabilization of temperature, ozone gas was bubbled in the solution. The formation of CompI was monitored by the absorption spectroscopy. After formation of CompI, excess ozone gas was removed by bubbling argon gas. After reconfirming the absorption spectrum of CompI, 50 equivalent of TBA-Cl in dichloromethane was added to the solution. After addition, the absorption spectral change was monitored at constant interval. Reactions of CompI with Hypochlorite. Iron(III) porphyrin chloride (for 1 ~ 4) or nitrate complex (for 5 and 6) (0.1 mM) in dichloromethane was placed in a 1 cm quartz cell. The cell was set on the low-temperature chamber and cooled to –80 °C. After stabilization of temperature, ozone gas was bubbled in the solution. The formation of CompI was monitored by the absorption spectroscopy. After formation of CompI, excess ozone gas was removed by bubbling argon gas. After reconfirming the absorption spectrum of CompI, 10 equivalent of TBA-OCl in dichloromethane was added to the solution. After addition, the absorption spectral change was monitored at constant interval. Catalytic Epoxidation Reactions of Cyclooctene with tbutyl hydroperoxide. Iron(III) TPFPP trifluoromethanesulfonate complex, TPFPP-Tf, was prepared by the reaction of iron(III) TPFPP chloride complex with silver trifluoromethanesulfonate in dichloromethane and purified by recrystallization.33 The catalytic reaction in the presence of TBA-OH was carried out in acetonitrile-dichloromethane mixture (3:1). Cyclooctene was dissolved in dichloromthane. TPFPP-Tf, TBAOH•30H2O, and t-butyl hydroperoxide were dissolved in acetonitrile. The total volume of the reaction solution was 2.0 mL and the final concentration of heme, cyclooctene, TBA-OH, and t-butyl hydroperoxide are 0.5 mM, 500 mM, 50 mM, and 50 mM, respectively. The reaction mixture was stirred for 30 ~ 120 min at room temperature. The products were analyzed by GC-MS after addition of a standard sample (n-tetradecane) to the reaction solution. The catalytic reaction in the presence of methanol was carried out in dichloromthane-methanol mixture (7:3). The total volume of the reaction solution was 2.0 mL and the final concentration of heme, cyclooctene, and t-butyl hydroperoxide are 0.5 mM, 500 mM, and 50 mM, respectively. The reaction mixture was stirred for 30 ~ 120 min at room temperature. The products were analyzed by GC-MS after addition of a standard sample (ntetradecane) to the reaction solution. The product yields, based

on t-butyl hydroperoxide, were averages of three independent experiments and hardly changed by the reaction time (30 ~ 120 min). Electrochemistry. Oxoiron(IV) porphyrin complexes of 1 ~ 3 were prepared by published method4 and electrochemical measurements were carried out in acetonitrile containing 0.1 M tetran-butylammonium perchlorate at -40 ºC. Oxoiron(IV) porphyrin complexes of 4 ~ 6 were prepared by ozone oxidation of the corresponding ferric hydroxide complexes of 4 ~ 6 in acetonitrile containing 0.1 M tetra-n-butylammonium perchlortate at –40 °C. Excess ozone gas was removed by bubbling Ar gas. The working and counter electrodes were directly immersed in the sample solution and the reference electrode (SCE) was immersed in the solution connecting the sample solution through the ion permeability porous glass. The electrochemical measurements were carried out before and after the ozone oxidation at –40 °C.

ASSOCIATED CONTENT Supporting Information Figure S1 – S16 and experimental details. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author

[email protected] Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This study was supported by grants from JSPS (Grant Nos. 25288032 & 17H03032), CREST, Nara Women’s University, The Naito Foundation, and The Uehara Memorial Foundation. We thank IMS for assistance of EPR measurements.

REFERENCES (1) Liu, W.; Groves, J. T. Acc. Chem. Res. 2015, 48, 1727-1735. (2) Nam, W. Acc. Chem. Res. 2007, 40, 522-531. (3) Meunier, B.; Bernadou, J. In Metal-Oxo and Metal-Peroxo Species in Catalytic Oxidations; Meunier, B. Ed.; Structure and Bonding; Springer: Berlin, 2000, 97, pp 1−35. (4) Poulos, T. L. Chem. Rev. 2014, 114, 3919-3962. (5) Fujii, H. In Heme Peroxidases; Raven, E. and Dunford, B. Eds.; RSC Metallobiology Series; The Royal Society of Chemistry: London, 2016, pp 183-217. (6) Groves, J. T. J. Inorg. Biochem. 2006, 100, 434-447. (7) (a) Traylor, T. G.; Xu, F. J. Am. Chem. Soc. 1990, 112, 178-186. (b) Traylor, T. G.; Xu, F. J. Am. Chem. Soc. 1987, 109, 62016203. (c) Traylor, T. G.; Fann, W.-P.; Bandyopadhay, D. J. Am. Chem. Soc. 1989, 111, 8009-8010. (d) Traylor T. G.; Ciccone, J. P. J. Am. Chem. Soc. 1989, 111, 8413-8420. (e) Traylor T. G.; Tsuchoya, S.; Byun, Y.-S.; Kim, C. J. Am. Chem. Soc. 1993, 115, 2775-2781. (f) Traylor, T. G.; Kim, C.; Richards, J. L.; Xu, F.; Perrin, C. L. J. Am. Chem. Soc. 1995, 117, 3468-3474. (8) (a) Bruice, T. A. Acc. Chem. Res. 1991, 24, 243-249. (b) Almarsson, Ö; Bruice, T. C. J. Am. Chem. Soc. 1995, 117, 45334544. (c) He, G.-X.; Bruice, T. C. J. Am. Chem. Soc. 1991, 113, 2747-2753.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(9) Allentoff, A. L.; Bolton, J. L.; Wilks, A.; Thompson, J. A.; Oritz de Montellano, P. R. J. Am. Chem. Soc. 1992, 114, 97449749. (10) (a) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1986, 108, 7834-7836. (b) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1986, 108, 7836-7837. (11) Yamaguchi, K.; Watanabe, Y.; Morishima, I. J. Am. Chem. Soc. 1993, 115, 4058-4065. (12) (a) Nam, W.; Choi, H. J.; Han, H. J.; Cho, S. H.; Lee, H. J.; Han, S.-Y. Chem. Commun., 1999, 387-388. (b) Nam, W.; Han, H. J.; Oh, S.-Y.; Lee, Y. J.; Choi, M.-H.; Han, S.-Y.; Kim, C.; Woo, S. K.; Shin, W. J. Am. Chem. Soc. 2000, 122, 86778684. (13) Van Heuvelen, K. M.; Fielder, A. T.; Shan, X.; DeHont, R. F.; Meier, K. K.; Bominaar, E. L.; Münck, E.; Que, Jr. L. Proc. Natl. Acad. Sci. 2012, 109, 11933-11938. (14) (a) Nam, W.; Lim, M. H.; Oh, S.-Y. Inorg. Chem. 2000, 39, 5572-5575. (b) Guo, M.; Dong, H.; Li, J.; Cheng, B.; Huang, Y. Q.; Feng, Y. Q.; Lei, A. Nat. Commun. 2012, 3, 1190. (c) Collman, J. P.; Zeng, L.; Brauman, J. I. Inorg. Chem. 2004, 43, 2672-2679. (15) (a) Wang, C.; Kurahashi, T.; Fujii, H. Angew. Chem. Int. Ed. 2012, 51, 7809-7811. (b) Wang, C.; Kurahashi, T.; Inomata, K.; Hada, M.; Fujii, H. Inorg. Chem. 2013, 52, 9557-9566. (16) (a) Jacobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J. R.; Deng, L. J. Am. Chem. Soc. 1991, 113, 7063-7064. (b) Palucki, M.; Finney, N. S.; Pospisil, P. J.; Güler, M. L.; Ishida, T.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120,948-954. (17) Meunier, B.; Guilmet, E.; Caralho, M. E. D.; Poilblanc, R. J. Am. Chem. Soc. 1984, 106, 6668-6676. (18) Cong, Z.; Yanagisawa, S.; Kurahashi, T.; Ogura, T.; Nakashima, S.; Fujii, H. J. Am. Chem. Soc. 2012, 134, 2061720620. (19) Takahashi, A.; Kurahashi, T.; Fujii, H. Inorg. Chem. 2011, 50, 6922-6928. (20) (a) Groves, J. T.; Gross, Z.; Stern, M. K. Inorg. Chem. 1994, 33, 5065-5072. (b) Jeong, Y. J.; Kang, Y.; Han, A. R.; Lee, Y. M.; Kotani, H.; Fukuzumi, S.; Nam, W. Angew. Chem. Int. Ed. 2008, 47, 7321-7324. (c) Pan, Z.; Newcomb, M. Inorg. Chem. 2007, 46, 6767-6774. (21) Fujii, H. J. Am. Chem. Soc. 1993, 115, 4641-4648. (22) Fujii, H.; Yoshimura, T.; Kamada, H. Inorg. Chem. 1996, 35, 2373-2377. (23) Armstrong, D. A.; Huie, R. E.; Lymar, S.; Koppenol, W. H.; Merényi, G.; Neta, P.; Stanbury, D. M.; Steenken, S.; Wardman, P. Bioinorg. React. Mech. 2013, 9, 59-61. (24) Cong, Z.; Kurahashi, T.; Fujii, H. Angew. Chem. Int. Ed. 2011, 50, 9935-9939. (25) Takahashi, A.; Kurahashi, T.; Fujii, H. Inorg. Chem. 2009, 48, 2614-2625. (26) (a) Workentin, M. S.; Maran, F.; Wayner, D. D. M. J. Am. Chem. Soc. 1995, 117, 2120-2121. (b) Bietti, M.; DiLabio, G. A.; Lanzalunga, O.; Salamone, M.; J. Org. Chem. 2010, 75, 58755881. (c) Donkers, R. L.; Maran, F.; Wayner, D. D. M.; Workentin, M. S. J. Am. Chem. Soc. 1999, 121, 7239-7248. (d) Wayner, D, D. M.; Parker, V. D. Acc. Chem. Res. 1993, 26,287-294. (27) Das, T. N.; Dhanasekaran, T.; Alfassi, Z. B.; Naeta, P. J. Phys. Chem. 1998, 102, 280-284. (28) Boaz, N. C.; Bell, S. R.; Groves, J. T. J. Am. Chem. Soc. 2015, 137, 2875-2885. (29) Kang, Y.-F.; Yan, W.-J.; Zhou, T.-W.; Dai, F.; Li, X.-Z.; Bao, X.-Z.; Du, Y.-T.; Yuan, C.-H.; Wang, H.-B.; Re, X.-R.; Liu, Q.; Jin, X.-L. Zhou, B.; Zhang, J. Chem. Eur. J. 2014, 20, 89048908.

Page 12 of 13

(30) (a) Hu, Y.; Ishikawa, Y.; Hirai, K.; Tomioka, H. Bull. Chem. Soc. Jpn. 2001, 74, 2207-2218. (b) Gruenberg, M. F.; Goossen, L. J. Chem. Eur. J. 2013, 19, 7334-7337. (31) Lindsey, J. S.; Wagner, R. W. J. Org. Chem. 1989, 54, 828836. (32) Wyllie, G. R. A.; Munro, O. Q.; Schulz, C. E.; Sheidet, W. R. Polyhedron, 2007, 26, 4664-4672. (33) Gismelseed, A.; Bominaar, E. L.; Bill, E.; Trautwein, A. X.; Winkler, H.; Nasri, H.; Doppelt, P.; Mandon, D.; Fischer, J.; Weiss, R. Inorg. Chem. 1990, 29, 2741.

ACS Paragon Plus Environment

Page 13 of 13 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Critical Factors in Determining the Heterolytic versus Homolytic Bond Cleavage of Terminal Oxidants by Iron(III) Porphyrin Complexes Sawako Yokota and Hiroshi Fujii*

ACS Paragon Plus Environment

13