Synthesis, characterization and catalytic activity under homogeneous conditions of ethylene glycol substituted porphyrin manganese(III) complexes

Accepted Manuscript Synthesis, characterization and catalytic activity under homogeneous conditions of ethylene glycol substituted porphyrin manganese (III) complexes Kelly A.D.F. Castro, Fernando H.C. de Lima, Mário M.Q. Simões, M.Graça P.M.S. Neves, Filipe A. Almeida Paz, Ricardo F. Mendes, Shirley Nakagaki, José A.S. Cavaleiro PII: DOI: Reference:

S0020-1693(16)30280-8 http://dx.doi.org/10.1016/j.ica.2016.05.038 ICA 17074

To appear in:

Inorganica Chimica Acta

Received Date: Revised Date: Accepted Date:

28 February 2016 16 May 2016 17 May 2016

Please cite this article as: K.A.D. Castro, F.H.C. de Lima, M.M.Q. Simões, M.P.M. Neves, F.A. Almeida Paz, R.F. Mendes, S. Nakagaki, J.A.S. Cavaleiro, Synthesis, characterization and catalytic activity under homogeneous conditions of ethylene glycol substituted porphyrin manganese (III) complexes, Inorganica Chimica Acta (2016), doi: http://dx.doi.org/10.1016/j.ica.2016.05.038

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis, characterization and catalytic activity under homogeneous conditions of ethylene glycol substituted porphyrin manganese (III) complexes Kelly A. D. F. Castroa,b, Fernando H. C. de Limaa, Mário M. Q. Simõesb,*, M. Graça P. M. S. Nevesb, Filipe A. Almeida Pazc, Ricardo F. Mendesc, Shirley Nakagaki a,*, José A. S. Cavaleirob a

Laboratório de Química Bioinorgânica e Catálise, Universidade Federal do Paraná

(UFPR), CP 19081, CEP 81531-990, Curitiba, Paraná, Brasil b

Department of Chemistry and QOPNA, University of Aveiro, 3810-193 Aveiro,

Portugal c

Department of Chemistry and CICECO – Aveiro Institute of Materials, University of

Aveiro, 3810-193 Aveiro, Portugal

*Corresponding authors e-mail address: [email protected] (M. M. Q. Simões) e-mail address: [email protected] (S. Nakagaki)

1

ABSTRACT The catalytic efficiency and selectivity of a series of manganese(III) complexes of meso-tetrakis(pentafluorophenyl)porphyrin H2(TPFPP) derivatives bearing one to four ethylene glycol moieties are compared in the oxidation of cis-cyclooctene and cyclohexane in the presence of H2O2 at ambient temperature. The results show that the catalysts’ performance depends on the number of ethylene glycol moieties present and on the cocatalyst used; for both substrates, the most efficient catalyst in the presence of H2O2 is the Mn(III) complex bearing three ethylene glycol moieties (MnP3). Moreover, in the oxidation of cyclohexane in the presence of iodosylbenzene (PhIO) as oxidant, all the substituted Mn(III) complexes are more efficient than the parent non-substituted Mn(III) porphyrin complex. In addition, during the preparation of this series of derivatives, the tetrasubstituted free-base porphyrin (P4) was further isolated and studied in the solid state, by using single-crystal X-ray diffraction studies. Keywords: porphyrins, ethylene glycol, manganese, oxidation, hydrogen peroxide.

2

1. Introduction The use of synthetic macrocycles, such as porphyrins, as catalysts in oxidation reactions has attracted a great interest in bioinorganic chemistry [1-8], especially due to their efficiency in modelling the catalytic oxidation ability of cytochrome P-450 and lignin peroxidases [4,9]. These two heme proteins play special roles in the biosynthesis and degradation of important endogenous compounds, such as steroid hormones, fatty acid derivatives or vitamins, biodegradation of the plant cell wall constituent lignin, among others [3,10]. In the 1970s, Groves and co-workers [11] showed, for the first time, that synthetic metalloporphyrins can mimic the catalytic activity observed in biological systems, by using

the

iron(III)

complex

of

meso-tetraphenylporphyrin

(Fe(TPP)Cl)

and

iodosylbenzene (PhIO) in the epoxidation of alkenes and in the hydroxylation of alkanes. However, these early stage catalysts, the so-called first-generation porphyrin catalysts, suffered from rapid deactivation caused by oxidative degradation of the macrocycle [12]. Further studies resulted in the development of more stable and, in general, more efficient catalysts, by introducing electronegative and/or bulky auxiliaries such halogen [13-15], nitro [16] or sulfonato [4] substituents at the meso and/or βpyrrolic positions (the so-called second and third generation porphyrin catalysts) [17-19]. In fact, several studies showed how the structural modification of the macrocycle core can contribute for a better catalytic performance under homogeneous conditions [5-8,20]. This fact is usually associated with the stereo-electronic effect exerted by the substituents, which are able to increase the lifetime and the reactivity of the catalytic active species in solution. Besides, the formation of dimeric species and the oxidative self-destruction responsible for the inactivation of the catalyst are often inhibited

[17].

In

this

context,

the

use

of

the

metallocomplexes

of

the

meso-tetrakis(pentafluorophenyl)porphyrin H2(TPFPP) has proven to be a surprisingly versatile alternative presenting many possibilities of structural modification by nucleophilic substitution of its fluorine atoms [21-24]. Manganese(III) and iron(III) porphyrins are usually considered as the most important representative class of metalloporphyrins with catalytic activity in the epoxidation of alkenes [5-8,25]. For both metals, the generally accepted catalytically active species for the epoxidation reactions with different substrates and oxygen atom donors is a metal-oxo species. Although iron(III) porphyrins can be highly active in the epoxidation of alkenes, their efficiency can, sometimes, be limited by the formation of 3

by-products resulting, for example, from allylic oxidation reactions [26]. This fact is less limitative concerning the manganese(III) complexes of porphyrins. Additionally, some studies by Mansuy and other groups [20,27-29] showed that Mn-porphyrins in the presence of adequate cocatalysts are effective catalysts to activate hydrogen peroxide for the epoxidation of alkenes [30-32]. This is considered a green and environmentally friendly oxidant because the sole by-product is water [33]. We have recently studied the efficiency of the iron(III) and manganese(III) complexes of meso-tetrakis(pentafluorophenyl)porphyrin H2(TPFPP) derivatives bearing one and four ethylene glycol moieties as catalysts in the oxidation of alkenes and alkanes with PhIO as oxidant, confirming the superior efficiency of the manganese(III) complexes [20]. This fact, and our interest in developing oxidative catalytic processes under environmentally benign conditions [33], prompted us to evaluate the efficiency of the manganese(III) complexes of H2(TPFPP) derivatives bearing one to four ethylene glycol moieties (MnP1-MnP4, Figure 1) in the oxidation of cis-cyclooctene and cyclohexane with H2O2 as oxidant. The efficiency of the new Mn(III)complexes with two and three aliphatic chains was also evaluated in the presence of PhIO using the same model substrates. Moreover, the free-base porphyrin P4 was isolated in the solid state as a crystalline material and single-crystal X-ray diffraction studies were carried out. 2. Experimental Chemicals were purchased from Aldrich, Sigma, or Merck, and were of analytical grade. 2.1. Synthesis of the manganese porphyrins The five ethylene glycol manganese(III) porphyrin derivatives MnP1-MnP4 (Figure 1) were prepared from the corresponding free-bases P1-P4 using the adequate manganese(II) acetate salt according to the conventional methodology reported by Kobayashi [34]. The free-base porphyrins P1, P2a, P2b, P3 (see below for an improvement procedure of the reactions conditions) and P4 were prepared as previously described [20]. Purple crystals of porphyrin P4 were obtained by dissolving the porphyrin in a small amount of dimethyl sulfoxide and leaving the resulting mixture at ambient temperature over a period of two weeks.

4

Figure 1. Schematic representation of the manganese(III) porphyrins prepared.

i) Improved synthesis of the free-base porphyrin P3: Ethylene glycol (8.6 mmol) was dissolved in 15 mL of dry THF, and NaH (0.95 mmol) was then added. The reaction mixture was kept under magnetic stirring and nitrogen atmosphere for 30 min. A solution of H2(TPFPP) in THF (15.4 mmol/L) was added and the reaction mixture was heated at reflux temperature until the total consumption of the starting porphyrin (24 h, monitored by TLC). After cooling to ambient temperature, the mixture was diluted with dichloromethane and washed 3 times with 10 mL of water. The solvent was evaporated under reduced pressure and the crude solid was purified by preparative TLC using dichloromethane:methanol (9:1 v/v) as the eluent. The fourth fraction was identified as the porphyrin P3 (47% yield), mp > 300 °C. 1H NMR: δH ppm (CD3OD); 9.10 (broad s, 8H, H-β); 4.11 (t, J = 4.4, 6H, CH2) and 4.70 (t, J = 4.4, 6H, CH2).

19

F NMR: δF ppm (CD3OD) -188.63 (dt, J = 7.3 and 21.3, 2F, F-meta), -

182.84 (dd, J = 7.3 and 21.3, 6F, F-meta), -179.13 (t, J = 21.3, 1F, F-para), -165.75 (dd, J = 7.3 and 21.3, 6F, F-ortho), -163.73 (dd, J = 7.4 and 21.3, 2F, F-ortho). UV– VIS (CH2Cl2) λmax nm (logε): 412 (5.30), 510 (4.19), 582 (3.65) 654 (2.69). HRMS (FAB+): m/z calcd for C50H25F17N4O6: 1101.1536 (M+H)+; found: 1101.1559.

5

ii) Synthesis of the manganese porphyrins (MnP1-MnP4): The metalation of the free-base porphyrins P1-P4 with [Mn(CH3COO)2] was performed by employing a modification of the conventional methodology reported by Kobayashi, as described previously by our research groups for the ethylene glycol porphyrin derivatives MnP1 and MnP4. MnP2a (C48H18F16MnN4O 4): (95% yield); UV-VIS (CH3OH) λmax, nm (logε): 457 (5.27), 497 (3.48) and 551 (3.83). HRMS-ESI(+): m/z calcd for C48H18F18MnN4O4 1111.0421 (M+.); found: 1111.0419. MnP2b (C48H18F18MnN4O 4): (98% yield); UV-VIS (CH3OH), λmax, nm (logε): 457 (4.91), 497 (3.60) and 552 (4.04). HRMS-ESI(+): m/z calcd for C48H18F18MnN4O4 1111.0421 (M+.); found: 1111.0425. MnP3 (C50H23MnF17N4O6): (97% yield); UV-VIS (CH3OH) λmax, nm (logε): 457 (4.98), 497 (3.57) and 552 (3.90). HRMS-ESI(+): m/z calcd for C50H23MnF17N4O6 1153.0727 (M+.); found: 1153.0702 (Supporting information, Figures S1-S3). 2.2. Catalytic oxidation reactions The oxidation of cis-cyclooctene (previously purified in a neutral alumina column) or cyclohexane was carried out by using hydrogen peroxide as oxidant. The general procedure was as follows: in a 1.5 mL reaction flask, the catalyst (about 0.5 mg of MnP) and the cocatalyst (about 7.0 mg of ammonium acetate or 10 mg of imidazole) were suspended in acetonitrile (500 µL), followed by the substrate (0.27 mmol) and the solution of 30% (w/w) H2O2 diluted in acetonitrile (1:10). The mixture was maintained under magnetic stirring at ambient temperature in the absence of light. The oxidant (a solution of 30% (w/w) H2O2 diluted in acetonitrile (1:10)) was progressively added in aliquots corresponding to a half-substrate amount at intervals of 15 min. The catalyst/oxidant/substrate (cat/ox/sub) molar ratio was 1:1200:600 and the reaction time was 1 h. After the reaction time, the supernatant was transferred to a volumetric flask and the volume was completed with acetonitrile. The solutions containing the reaction products were analysed by GC-FID and the results are expressed in terms of conversion of the substrate, using the internal standard methodology (octan-1-ol was employed as the internal standard). The oxidation reactions using PhIO as oxidant were performed as described previously by us [20]. The cat/ox/sub molar ratio was 1:50:5000 and the reaction time was also 1 h.

6

2.3. Evaluation of the stability of the MnP3 and MnP4 catalysts The stability of MnP3 and MnP4 in the oxidation reaction conditions was monitored by UV-VIS spectroscopy by considering their characteristic absorption Soret band at about 460 nm. In these experiments, the catalyst stability was evaluated in the presence of solvent (acetonitrile), substrate (cis-cyclooctene) and oxidant (H2O2). Each reaction was followed by UV-VIS at different periods of time. 2.4. Characterization techniques The 1H and

19

F NMR spectra were recorded on a Bruker AVANCE 300 at

300.13 and 282.38 MHz, respectively, using CDCl3 as solvent and TMS as the internal reference. Mass spectrometry analyses were carried out on a Mass spectrometer – Micromass Q-TOF-2, and HRMS were recorded on VG AutoSpec Q and M mass spectrometers. FTIR spectra were recorded on a Biorad 3500 GX spectrophotometer in the 400-4000 cm-1 spectral range using KBr (Aldrich, 99%) pellets. In a typical experiment, KBr was crushed with a small amount of the solid samples, and the spectra were collected with a resolution of 4 cm-1 and accumulation of 32 scans. Qualitative and quantitative electronic spectra of the solution samples were recorded on a Cary-Varian spectrophotometer, in the 200-800 nm range using a quartz cuvette with 1 cm optical path. Electron paramagnetic resonance (EPR) spectra of the solid materials were recorded on an EPR Bruker EMX micro X spectrometer operating in the X band (approximately 9.5 GHz) at 77 K using liquid N2. The catalytic oxidation reaction products were identified and quantified using gas chromatographs Agilent 6850 and Bruker Scion 436 (FID detector) equipped with DB-WAX type capillary columns (30 m × 0.25 mm i.d. × 0.25 µm film thickness). Quantitative analyses were performed by using an internal standard. 2.5. Single-Crystal X-ray Diffraction Studies of P4·2(DMSO) Single crystals of the tetrasubstituted free base porphyrin P4·2(DMSO) were manually harvested from the crystallization vial and immersed in highly viscous FOMBLIN Y perfluoropolyether vacuum oil (LVAC 140/13, Sigma-Aldrich) to avoid degradation caused by the evaporation of the solvent [35]. Crystals were mounted on Hampton Research CryoLoops with the help of a Stemi 2000 stereomicroscope equipped with Carl Zeiss lenses. X-ray diffraction data were collected at 150(2) K on a Bruker D8 QUEST equipped with Mo Kα sealed tube (λ = 0.71073 Å), a multilayer 7

TRIUMPH X-ray mirror, a PHOTON 100 CMOS detector, controlled by the APEX2 software package [36], and an Oxford Instruments Cryostrem 700+ Series low temperature device. Diffraction images were processed using the software package SAINT+ [37], and data were corrected for absorption by the multiscan semi-empirical method implemented in SADABS [38]. The structure was solved using the algorithm implemented in SHELXT-2014 [39], which allowed the immediate location of almost all of the heaviest atoms composing the asymmetric unit. The remaining missing and misplaced non-hydrogen atoms were located from difference Fourier maps calculated from successive full-matrix least-squares refinement cycles on F2 using the latest SHELXL from the 2014 release [40,41]. All structural refinements were performed using the graphical interface ShelXle [42]. Hydrogen atoms bound to carbon were placed at their idealized positions using appropriate HFIX instructions in SHELXL: 43 (aromatic carbon atoms), 23 (for the – CH2– moieties of the substituent groups) or 137 (for the terminal methyl groups of the DMSO solvent of crystallization). These hydrogen atoms were included in subsequent refinement cycles with isotropic thermal displacements parameters (Uiso) fixed at 1.2 (for the two former family of hydrogen atoms) or 1.5×Ueq (solely for those associated with the methyl groups) of the parent carbon atoms. The hydrogen atoms associated with the terminal hydroxyl groups were placed in geometrical positions using the HFIX 147 instruction in SHELXL in order to maximize the hydrogen bonding interactions in which these groups are involved. The isotropic thermal displacement parameter (Uiso) of these hydrogen atoms were fixed at 1.5×Ueq of the parent oxygen atoms. Crystal data for P4·2(DMSO): C56H42F16N4O10S2, M = 1299.05, monoclinic, space

group

P21/c,

Z = 2,

a = 15.148(5) Å,

b = 9.699(3) Å,

c = 18.233(5) Å,

β = 97.067(4)º, V = 2658.3(13) Å3, µ(Mo-Kα) = 0.223 mm-1, Dc = 1.623 g cm-3, red plates with crystal size of 0.19×0.16×0.03 mm3. Of a total of 34859 reflections collected, 4843 were independent (Rint = 0.0683). Final R1 = 0.0634 [I > 2σ(I)] and wR2 = 0.1960 (all data). Data completeness to theta = 25.24°, 99.6%. The last difference Fourier map synthesis showed the highest peak (0.617 eÅ-3) and the deepest hole (-0.586 eÅ-3) located at 1.67 and 0.69 Å from H26B and S1,

8

respectively. Structural drawings have been created using the software package Crystal Impact Diamond [43]. Crystallographic data (including structure factors) for the crystal structure of compound P4·2(DMSO) have been deposited with the Cambridge Crystallographic Data Centre as supplementary publication No. CCDC-1438587. Copies of the data can be obtained free of charge on application to CCDC, 12 Union Road, Cambridge CB2 2EZ, U.K. FAX: (+44) 1223 336033. E-mail: [email protected]. 3. Results and Discussion 3.1. Synthesis and characterization of the manganese(III) porphyrins The synthetic strategy to obtain the manganese complexes MnP1-MnP4 bearing one to four ethylene glycol units is outlined in Figure 1 and involved the previous preparation of the corresponding free-bases P1-P4, followed by metalation with [Mn(AcO)2]. The free-base derivatives were obtained from the reaction of meso-tetrakis(pentafluorophenyl)porphyrin with ethylene glycol in the presence of NaH following the experimental procedure previously described by our research groups [20]. Depending on the reaction conditions employed, the formation of a derivative can be preferentially favoured. In this study we were able to slightly improve the yield of P3 from 40% to 47% just by acting on the reaction time. The structures of all the porphyrins P1-P4 are in accordance with the previously analytical data (1H and

19

F NMR, HRMS, UV-VIS and FTIR). Additionally, P4 was successfully

crystallized from dimethyl sulfoxide that allowed the isolation of crystals, whose structure was unveiled from single-crystal X-ray diffraction (see below). The metalation process of the free-base porphyrins P1, P2a, P2b, P3 and P4 with [Mn(AcO)2] was followed by recording the electronic spectra at different reaction times. The structure of the metalloporphyrins was confirmed by UV-VIS, FTIR techniques and ESI-MS, with the collected analytical data of MnP1 and MnP4 being in accordance with that already available in the literature [20]. The UV-VIS spectra of the new complexes MnP2a, MnP2b and MnP3 reveal the characteristic Soret band at 457 nm, showing a bathochromic shift in comparison to the Soret band of the free-base porphyrins (408 nm) and two Q-bands at 497 and 551 nm (see Supporting information, Figure S4 for MnP2a), resulting from the alteration in the micro symmetry of the porphyrin macrocycle due to the presence of the Mn(III) ion in the porphyrin core. Bands at 365 and 390 nm are assigned to an 9

electronic transition ligand metal charge transfer (LMCT), a1u(π),a2u(π)
eg(dπ). The weak bands at 676 and 767 nm were assigned by Boucher [44] to an electronic transition LMCT, a1u, a2u
eg (dπ), typical of manganese porphyrins [44, 45]. The manganese(III) porphyrins exhibited the typical FTIR bands of porphyrins (Figures not shown). The bands in the region of 3000 cm-1, ascribed to symmetric and asymmetric NH stretching, and in the region of 1600 cm-1, referring to another δNH vibrational mode [46] have disappeared. The bands at 1404 and 1589 cm-1 in the complex are assigned to the acetate groups, attributed to symmetric and asymmetric stretching of the C=O bond. The difference in the positioning of these bands (∆ = 185 cm-1) suggests, according to Nakamoto [46], that acetate is present as the counter-ion. The new metalloporphyrins were also characterized by EPR in order to identify/confirm the oxidation state of the metal present, as well as its symmetry and possible distortions. The perpendicular microwave polarization X-band EPR spectra of MnP2a, MnP2b and MnP3 recorded at ambient temperature and 77 K are very similar and no signals were detected (Figures not shown). These results are in accordance with the presence of Mn(III) in the porphyrin core [47]. 3.2. Single-Crystal X-ray Diffraction Studies of P4·2(DMSO) Porphyrin P4 was isolated in the solid state as a crystalline material cocrystallising with two molecules of DMSO per molecular unit. Single-crystal X-ray diffraction studies show that the molecular unit adopts in P4·2(DMSO) a centrosymmetric geometry as depicted in Figure 2. Remarkably, contrary to that usually observed for other porphyrinic derivatives with pendant alkylic substituent chains, the two crystallographically independent ethylene glycol residues do not show any structural disorder (Figure 2). We believe that this structural feature arises mainly from the presence of the strong [dD···A = 2.747(5) Å] and somewhat directional [<(DHA) of 145º] O–H···O intermolecular hydrogen bonding involving the two independent terminal hydroxyl groups. This sole supramolecular interaction holds together neighbouring individual P4 porphyrins as shown in Figure S5 (see Supporting information).

10

Figure 2. Schematic representation of the centrosymmetric porphyrinic molecular unit present in compound P4·2(DMSO). Non-hydrogen atoms are represented as thermal ellipsoids drawn at the 50% probability level. Hydrogen atoms are represented as small spheres with arbitrary radii. Please note: the asymmetric unit of compound P4·2(DMSO) is composed of half of the porphyrin molecule and one DMSO solvent molecule of crystallization.

The close packing in the crystal structure of P4·2(DMSO) is mainly driven by the need to closely pack individual porphyrin molecules while fulfilling the aforementioned, structure directing O–H···O intermolecular hydrogen bonds. It is, however, important to emphasize that a number of weak C–H···O and C–H···F also exist in the crystal structure globally contributing to structural robustness. The close packing of individual P4 molecules leads to the presence of structural voids that are filled by the DMSO solvent molecules as depicted in Figure 3.

11

Figure 3. Schematic representation of the crystal packing of compound P4·2(DMSO) viewed in perspective along the [010] direction of the unit cell. Porphyrin molecules are represented in ball-and-stick mode while the DMSO solvent molecules of crystallization in space filling mode.

3.3. Investigation of the catalytic activity We have recently reported that the porphyrin complexes MnP1 and MnP4 are able, in the presence of iodosylbenzene and under homogeneous conditions, to very efficiently epoxidize cis-cyclooctene (88% vs 96%) and also to hydroxylate cyclohexane (52% vs 64%) [20]. In the present study we intend to verify if that efficiency is maintained in the Mn complexes of H2(TPFPP) bearing two and three ethylene glycol moieties (MnP2a, MnP2b and MnP3). Additionally, it is important to verify if the distribution of the aliphatic chains in the disubstituted complexes affects (or not) the catalytic efficiency. As a preliminary experiment, in order to evaluate if the new complexes are catalytic active and stable under the reaction conditions, iodosylbenzene was used as oxidant and cis-cyclooctene and cyclohexane were studied as model compounds. Figure 4 summarizes the catalytic results obtained in the epoxidation of ciscyclooctene in the presence of MnP2a, MnP2b and MnP3 together with the previous

12

results already reported for MnP1 and MnP4 [20]. Figure 4 also compiles the catalytic results

obtained in the hydroxylation of cyclohexane with this series of

manganese(III) porphyrins. The catalytic tests confirmed the efficiency of the glycol substituted manganese(III) porphyrins in the oxidation of cis-cyclooctene. The epoxide yields decreased in the following order: MnP3 > MnP4 > MnP2b > MnP1 > MnP2a. The epoxide yield obtained for MnP2a, MnP2b and MnP1 was very similar, only MnP2a yield was slightly lower than the other two. These results suggest that the substitution of two of the electronegative p-fluorine atoms by ethylene glycol substituents did not significantly affect the stability of the porphyrin core. The better result observed for MnP3 (98%) and MnP4 (96%) can be justified by the stability of the catalytic species found for this two complexes, as expected for this family of second generation porphyrins [20]. Cyclooctene oxide 100 90

Cyclohexanol

Cyclohexanone 98

88

96

89

85

Yield (%)

80 70 60

59

64

52

50

45 35

40 30 20

10

10 0 MnP1

MnP2a

MnP2b

MnP3

MnP4

Figure 4. Cyclooctene and cyclohexane oxidation by PhIO catalysed by MnP1-MnP4 under homogeneous conditions. Yields were calculated based on the amount of PhIO added. Results represent reactions performed in duplicate or triplicate.

Cyclohexane is also a very useful model substrate to study catalytic systems involving MnP/PhIO system. Besides the catalytic efficiency, some information about selectivity can also be obtained by using this substrate, because the alcohol and the ketone, the main products frequently observed, can be formed. If compared to ciscyclooctene, cyclohexane is obviously harder to oxidize. In fact, manganese(III) porphyrins are known to be efficient and selective catalysts for cyclohexane oxidation under homogeneous conditions with high selectivity to the alcohol. The proposed

13

mechanism for the oxidation of cyclohexane involves the abstraction of a hydrogen atom from the substrate by a metal-oxo catalytic species, which generates an alkyl radical and a hydroxo-metalloporphyrin in the solvent cage (Scheme 1). Oxygen rebound preferably occurs within the cage, forming the alcohol [48-50]. Figure 4 shows that all the MnP compounds studied in this manuscript were selective to the alcohol, its yields decreasing in the following order: MnP4 > MnP3 > MnP1 > MnP2b > MnP2a. The tetra substitution of the p-fluorine atoms improved the efficiency of MnP4 if compared with the other complexes, affording alcohol with higher selectivity (64%). For MnP3 the selectivity for the alcohol is close to that observed for MnP4 (59% vs 64%, respectively) suggesting, in a resembling way when using cis-cyclooctene as substrate, similar stability of the catalytic species. In general, the here described modified manganese(III) porphyrins afforded higher product yields than its parent Mn(TPFPP)Cl (34% alcohol and 6% ketone) under similar conditions [20]. The inclusion of the –OR groups at the p-positions of the meso-aryl groups of the macrocycle increases the catalytic efficiency (MnP4 > MnP3 > MnP1 > MnP2b), except for MnP2a that exhibits lower selectivity (78%) for cyclohexanol relatively to Mn(TPFPP)Cl (85% selectivity). Such behaviour could also be attributed to the stability of the catalytic active species.

Scheme 1. Proposed mechanism for cyclohexane hydroxylation catalysed by Mn(III) porphyrins (MnP).

For MnP2a the formation of cyclohexanone was also observed (35% cyclohexanol and 10% cyclohexanone) because the produced alcohol probably

14

underwent catalytic overoxidation to ketone or, at the stage of oxygen rebound mechanism, there may be a leak of the alkyl radical causing it to react with other species in solution, ultimately resulting in several other products [50]. The two MnP2 complexes gave a total yield of 45%, depending on the position of the ethylene glycol substituent. The opposite or trans-isomer MnP2b was, however, more selective for the alcohol than the adjacent or cis-isomer MnP2a. This interesting behaviour may suggest that it is possible to modulate the selectivity just by modifying the macrocycle structure. Considering the stability and the efficiency of this series of catalysts in the epoxidation of cis-cyclooctene with PhIO, the following studies were focused on its catalytic behaviour when hydrogen peroxide was employed as oxidant. The catalytic activity using hydrogen peroxide for the compounds reported in this manuscript is unprecedented. It is well-known that when hydrogen peroxide is used as oxidant in the presence of Mn porphyrins the presence of a cocatalyst is required [13,51]. In our study, we have decided to perform the oxidation reactions in the presence of two well-known cocatalysts, ammonium acetate and imidazole. The function of the cocatalyst is to promote the formation of the active catalytic species; it acts as an acid-base catalyst to facilitate the axial heterolytic cleavage of the O-O bond and also to transfer one oxygen atom to the metal, in order to generate the metal-oxo active species [12]. Figure 5 summarizes the results obtained in the oxidation of ciscyclooctene in the presence of the different catalysts prepared in this work. In these experiments all the reactions were performed under the same cat/ox/sub molar ratio conditions (1:1200:600) and the oxidant was progressively added at intervals of 15 min. Reactions were stopped after 1 h (the same reaction time when using PhIO as oxidant). The gradual addition of the oxidant is important to minimize the possible catalase-like activity of the catalyst that promotes hydrogen peroxide dismutation and, in this way, to maximize the formation of the desired oxidation product, the epoxide. In reactions catalysed by Mn(III) porphyrins in the presence of an excess of hydrogen peroxide, the active catalytic species can also catalyse the conversion of hydrogen peroxide to molecular oxygen and water, ultimately leading to low yields [4,13,15,20]. As mentioned above, the Mn(III) porphyrins are able to promote hydrocarbon oxidation with hydrogen peroxide mainly in the presence of a cocatalyst, in order to promote the heterolytic cleavage of the O-O bond of hydrogen peroxide. 15

The Mn(III) porphyrins frequently used are robust catalysts (from the second and third generation of porphyrins) and the cocatalysts can be a nitrogen base (e.g., imidazole or pyridine), acetic acid, benzoic acid or ammonium acetate. It is well established that the presence of the cocatalyst favours the formation of a Mn(V)-oxo porphyrin species that is considered as the active oxidative species. 60

Epoxide yield (%)

50 40

52.0

Mn(TPFPP)Cl MnP1 37.7

MnP2a

32.8

MnP2b 30 20

MnP3 MnP4

20.5 10.6

10 0

2.7

0.2 0.4 0.1 0.1 0.4 0.5

H2O2

0.5

3.1

13.6 8.2

5.0 0.4

H2O2 + Imidazole

H2O2 + Ammonium acetate

Figure 5. Cyclooctene oxidation by H2O2 aqueous solution catalysed by MnP1-MnP4 and Mn(TPFPP)Cl under homogeneous conditions. Yields were calculated based on the amount of substrate. Results represent reactions performed in duplicate or triplicate.

In Figure 5, the very low conversions obtained in the absence of a cocatalyst (0.1% - 0.5%) corroborate its importance when hydrogen peroxide is used as oxidant. The cocatalyst is involved in several reaction steps, mainly in the removal of the proton of hydrogen peroxide in order to facilitate the heterolytic cleavage of the O-O bond of the intermediate; it also acts as axial ligand to Mn, thus favouring the formation of a leaving group (Scheme 2). However, for all the ethylene glycol substituted catalysts (MnP1-MnP4), low conversions were also observed when imidazole was used as cocatalyst. The best result was obtained for the MnP3 complex, showing 5.0% of conversion towards the epoxide. A different situation was observed using ammonium acetate as cocatalyst. All the catalysts were able to promote the epoxidation process, though with different overall efficiency. The most efficient catalyst was MnP3 (52% yield) followed by Mn(TPFPP)Cl (37.7% yield) and MnP4 (32.8% yield). The remaining catalysts MnP1,

16

MnP2a, and MnP2b showed epoxide yields ranging from 8.2% to 13.6%. The parent non-substituted complex Mn(TPFPP)Cl showed to be more efficient than the ethylene glycol substituted ones in the presence of imidazole (20.5% epoxide yield). The low conversion obtained with imidazole is probably linked to the different stages in which the cocatalyst participates. Johnstone and co-workers [52] studied the effect of a number of cocatalysts, including imidazole, by employing Mn(III) porphyrins as catalysts and hydrogen peroxide as oxidant. The authors concluded that one of the problems was centred on the amount of oxidant used, the destruction/oxidation of imidazole. Furthermore, when acetonitrile is used as solvent, there is a strong possibility of destruction of imidazole, forming peroxymidic acid. For the reactions performed using PhIO as oxidant and an excess of substrate, low conversions were registered if compared to those with hydrogen peroxide. A larger excess of oxidant is, however, used in the latter conditions. The large excess of oxidant may, thus, be responsible for the destruction of the catalyst in homogeneous catalysis, eventually resulting in lower overall efficiency. L ROH

H2O2

MnIII RH

L

L

MnIII

MnV O

O H

OH

L L H2O

MnIII O OH L

H

Scheme 2. Different roles of the cocatalyst in oxidation reactions using H2O2 and Mn(III) porphyrins as catalysts.

17

Besides the porphyrin structure resistance that frequently justifies best catalytic results [15,17,20,53-57], the choice of the solvent is also important. The low catalytic activity observed for MnP1, MnP2a and MnP2b if compared with MnP3 and MnP4 can be partly explained by the partial solubility of the catalyst in acetonitrile under the conditions studied, since the inclusion of the ethylene glycol moieties increases the overall polarity of the catalyst. In addition, Mn(III) porphyrins may have pronounced catalase-like activity, thereby reducing epoxidation yields, and the chemical structure of the Mn(III) porphyrins could also affect this catalase activity. Knowing that the electronegative groups generally increase the lifetime of the catalytically active metaloxo species, the absence of these groups should conversely reduce the catalytic activity, explaining the lower activity observed for the MnP1 and MnP2 complexes, if compared with the parent Mn(TPFPP)Cl under the same reaction conditions. However, the epoxide yield decrease observed for the mono and the disubstituted complexes was not registered for the trisubstituted one, and less drastic than for the tetrasubstituted Mn(III) complex. Recently, we have reported that the presence of bulkier substituents on the porphyrin macrocycle can affect the ring symmetry and, consequently, influence the formation and stabilization of the catalytic active oxo-species [20,33]. For this ethylene glycol substituted manganese porphyrin series, the MnP3 catalytic activity was better than MnP4 (52% vs 32.8% of epoxide yield, respectively), suggesting that the MnP3 catalytic species is more stable than the manganese tetrasubstituted MnP4 under this experimental conditions. The amount of hydrogen peroxide employed may be responsible for the destruction of the catalyst in homogeneous catalysis, which may explain the lower efficiency of MnP4 relatively to MnP3. In order to explain the results obtained for MnP3 and MnP4, the stability of these catalysts in homogeneous catalysis was further investigated. Figure 6 illustrates the MnP3 and MnP4 degradation during the oxidation reaction by monitoring the decrease of the manganese(III) porphyrin characteristic band at 457 nm in the presence of cis-cyclooctene and hydrogen peroxide. Throughout the reaction, the destruction of both MnP3 and MnP4 can be observed, since a decrease in the Soret band is registered, but in the first 5 minutes the characteristic Soret band has the greatest overall reduction for both complexes. Figure 6 shows the percentage of the remaining compound over time. These results justify the dissimilar catalytic activity observed for each of the complexes under the reaction conditions 18

investigated. The smallest degradation of MnP3 was expected since it presents the best catalytic performance in the oxidation of cis-cyclooctene (Figure 5). Figure 7 summarizes the results obtained in the oxidation of cyclohexane in the presence of the different catalysts with and without a cocatalyst. The data reveal high selectivity for cyclohexanol (ol) for most of the catalysts, in line with that reported in the literature [52]. 100 90

MnP3

80

MnP4

% MnP

70 60 50 40 30 20 10 0 0

5

10

15

20

25

30

35

40

45

50

55

60

Time (min)

Figure 6. UV-VIS spectrophotometric study of the MnP3 or MnP4 solutions in the presence of peroxide hydrogen and cis-cyclooctene at different reaction times and 1:1200:600 cat/ox/sub molar ratio. The Soret band intensity was monitored at 457 nm. When hydrogen peroxide was used as the oxidant, a similar behaviour to that registered for the oxidation of cis-cyclooctene is observed. In the absence of a cocatalyst no product formation was observed. For this substrate neither MnP3 nor MnP4 were able to efficiently produce cyclohexanol using imidazole as cocatalyst. The MnP1-MnP4 derivatives studied in this manuscript showed catalytic activity in the oxidation of cyclohexane in the presence of ammonium acetate only. As observed with PhIO, the inclusion of ethylene glycol groups increases the catalytic activity in the following order: MnP3 > MnP4 > MnP2b > MnP2a > MnP1 (16.1%, 15.6%, 11.7%, 7.4% and 1.3%, respectively). MnP3 and MnP4 show a better catalytic performance than the parent Mn(TPFPP)Cl (16.1% and 15.6% vs 12%) suggesting that both catalysts are more stable than Mn(TPFPP)Cl. In addition, higher selectivity for the alcohol is observed for MnP3. These results showed that the

19

selectivity can be modulated by the position and number of –OR groups: while MnP3 showed the best performance concerning selectivity for the alcohol, MnP2a was instead more selective for the ketone. 17.6

Products (%)

15.6 13.6 11.6 9.6

16.1% ol

Mn(TPFPP)Cl MnP1 MnP2a

12% ol

MnP2b MnP3

11.2% ol + 4.4% one

9.7% ol 6.7% ol + + 2.3% one 5.0% one

3.0 % ol + 4.4% one

MnP4

7.6 5.6 3.6

1.3% ol

1.6 -0.4

H2O2

H2O2 + Imidazole

H2O2 + Ammonium acetate

Figure 7. Cyclohexane oxidation by H2O2 catalysed by MP1-MnP4 and Mn(TPFPP)Cl in homogeneous conditions. Conversions were calculated based on the amount of substrate. Results represent reactions performed in duplicate or triplicate. Cyclohexanol (ol) and cyclohexanone (one). Under these conditions the conversion values were in the range of ± 0.5%

4. Conclusions The effects of the inclusion of one to four –OR moieties at the meso-aryl ppositions of H2(TPFPP) in the catalytic efficiency of their manganese(III) complexes was studied for the oxidation of cis-cyclooctene and cyclohexane using PhIO and hydrogen peroxide. It was found that the inclusion of an appropriate number of those bulky groups into the parent H2(TPFPP), in association with the remaining electronegative fluorine atoms, resulted in efficient and selective homogeneous catalysts for alkenes epoxidation and alkanes hydroxylation with both oxidants. For both substrates, the most efficient catalyst in the presence of H2O2 is the Mn(III) complex bearing three ethylene glycol moieties (MnP3). It was possible to modulate catalyst selectivity and efficiency by choosing the number of ethylene glycol moieties.

20

Acknowledgments The authors are grateful to Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Universidade Federal do Paraná (UFPR) for financial support. The authors also thank Fundação para a Ciência e a Tecnologia (FCT, Portugal), QREN, FEDER and COMPETE founds for the QOPNA (PEst-C/QUI/UI0062/2013; FCOMP-01-0124-FEDER-037296), and Universidade de Aveiro for financial support. Part of this work was developed within the scope of the project CICECO-Aveiro Institute

of Materials, POCI-01-0145-FEDER-007679

(FCT Ref. UID /CTM

/50011/2013), financed by national funds through the FCT/MEC and when appropriate co-financed by FEDER under the PT2020 Partnership Agreement. Authors further wish to thank CICECO for specific funding towards the purchase of the single-crystal X-ray diffractometer. Kelly A. D. F. Castro also thanks CNPq for the granted post-doctoral scholarship (Process: 151512/2013-2 and 201107/2014-7). References [1] A.J. Appleton, S. Evans, J.R.L. Smith, J. Chem. Soc. Perkin Trans. 2 (1995) 281. [2] J.T. Groves, T.E. Nemo, J. Am. Chem. Soc. 105 (1983) 6243. [3] D. Mansuy, C.R. Chimie 10 (2007) 392. [4] D. Mansuy, P. Battioni, Dioxygen activation at heme centers in enzymes and synthetic analogs, in: J. Reedijk, E. Bouwman (Eds.), Bioinorganic Catalysis, Marcel Dekker, Second Edition, New York, 1999, pp 323-354. [5] R.A. Sheldon, Metalloporphyrins in Catalytic Oxidations, Marcel Dekker, New York, 1994, pp 390-483. [6] F. Montanari, L. Casella, Metalloporphyrins Catalyzed Oxidations, Kluwer Academic, Dordrecht, The Netherlands, 1994, pp 49-86. [7] J. Bernadou, B. Meunier, Adv. Synth. Catal. 346 (2004) 171-184. [8] M.M.Q. Simões, C.M.B. Neves, S.M.G. Pires, M.G.P.M.S. Neves, J.A.S. Cavaleiro, Pure Appl. Chem. 85 (2013) 1671. [9] P. Zucca, A. Rescigno, A.C. Rinaldi, E. Sanjust, J. Mol. Catal. A: Chem. 388–389 (2014) 2. [10] D. Mansuy, Comp. Biochem. Physiol., C: Pharmacol. Toxicol. Endocrinol. 121 (1998) 5. [11] J.T. Groves, T.E. Nemo, R.S. Myers, J. Am. Chem. Soc.101 (1979) 1032. 21

[12] S. Nakagaki, G.K.B. Ferreira, G.M. Ucoski, K.A.D.F. Castro, Molecules 18 (2013) 7279. [13] F.S. Vinhado, M.E.F. Gardini, Y. Iamamoto, A.M.G. Silva, M.M.Q. Simões, M. G.P.M.S. Neves, A.C. Tomé, S.L. Rebelo, A.M.V.M. Pereira, J.A.S. Cavaleiro, J. Mol. Catal. A: Chem. 239 (2005) 138. [14] E.A. Vidoto, M.S.M. Moreira, F.S. Vinhado, K.J. Ciuffi, O.R. Nascimento, Y. Iamamoto, J. Non-Cryst. Solids 304 (2002) 151. [15] E.C. Zampronio, M.C.A.F. Gotardo, M.D. Assis, H.P. Oliveira, Catal. Lett. 104 (2005) 53. [16] M.A. Schiavon, Y. Iamamoto, O.R. Nascimento, M.D. Assis, J. Mol. Catal. A: Chem. 174 (2001) 213. [17] D. Dolphin, T.G. Traylor, L.Y. Xie, Acc. Chem. Res. 30 (1997) 251. [18] B. Meunier, Chem. Rev. 92 (1992) 1411. [19] A. Gonsalves, M.M. Pereira, J. Mol. Catal. A: Chem. 113 (1996) 209. [20] K.A.D.F. Castro, M.M.Q. Simões, M.G.P.M.S. Neves, J.A.S. Cavaleiro, F. Wypych, S. Nakagaki, Catal. Sci. Technol. 4 (2014) 129. [21] J.I.T. Costa, A.C. Tomé, M.G.P.M.S. Neves, J. Cavaleiro, J. Porphyrins Phthalocyanines 15 (2011) 1116. [22] J. Králova, T. Bríza, I. Moserova, B. Dolensky, P. Vasek, P. Pouckova, Z. Kejik, R. Klaplánek, P. Martázek, M. Dvorak, V. Král, J. Med. Chem. 51 (2008) 5964. [23] S. Silva, P.M.R. Pereira, P. Silva, F.A.A. Paz, M.A.F. Faustino, J.A.S. Cavaleiro, J. Tomé, Chem. Commun. 48 (2012) 3608. [24] K.J. Ciuffi, H.C. Sacco, J.C. Biazzotto, E.A. Vidoto, O.R. Nascimento, C.A.P. Leite, O.A. Serra, Y. Iamamoto, J. Non-Cryst. Sol. 273 (2000) 100. [25] G.G. A. Balavoine, Y.V. Geletii, D. Bejan, Biol. Chem. 1 (1997) 507. [26] B.S. Lane, K. Burgess, Chem. Rev. 103 (2003) 2457. [27] D. Mansuy, Coord. Chem. Rev. 125 (1993) 129. [28] F.S. Vinhado, C.M.C. Prado-Manso, H.C. Sacco, Y. Iamamoto, J. Mol. Catal. A: Chem. 174 (2001) 279. [29] R. De Paula, I.C.M.S. Santos, M.M.Q. Simões, M.G.P.M.S. Neves, J.A.S. Cavaleiro, J. Mol. Catal. A: Chem. 404 (2015) 156. [30] W. Nam, I. Kim, M.H. Lim, H.J. Choi, J.S. Lee, G.H. Jang, Chem. Eur. 8 (2002) 2067.

22

[31] P. Battioni, J.P. Renaud, J.F. Bartoli, M. Reina-Artiles, M. Fort, D. Mansuy, J. Am. Chem. Soc. 110 (1988) 8462. [32] J.P. Renaud, P. Battioni, J.F. Bartoli, D. Mansuy, J. Chem. Soc., Chem. Commun. (1985) 888. [33] K.A.D.F. Castro, S.M. Pires, M.A. Ribeiro, M.M.Q. Simões, M.G.P.M.S. Neves, W.H. Schreiner, F. Wypych, J.A.S. Cavaleiro, S. Nakagaki, J. Colloid Interf. Sci 450 (2015) 339. [34] H. Kobayashi, T. Higuchi, Y. Kaizu, H. Osada, M. Aoki, Chem. Soc. Jpn. 48 (1975) 3137. [35] T. Kottke, D. Stalke, J. Appl. Crystallogr. 26 (1993) 615. [36] APEX2 Data Collection Software Version 2.1-RC13, Bruker AXS, Delft, The Netherlands 2006. [37] SAINT+ Data Integration Engine v. 8.27b© 1997-2012, Bruker AXS. [38] G.M. Sheldrick, SADABS 2012/1, Bruker AXS Area Detector Scaling and Absorption Correction 2012, Bruker AXS. [39] G.M. Sheldrick, SHELXT-2014, Program for Crystal Structure Solution, University of Göttingen 2014. [40] G.M. Sheldrick, Acta Cryst. A 64 (2008) 112. [41] G.M. Sheldrick, SHELXL Version 2014, Program for Crystal Structure Refinement, University of Göttingen 2014. [42] C.B. Hübschle, G.M. Sheldrick, B. Dittric, J. Appl. Crystallogr. 44 (2011) 1281. [43] K. Brandenburg, DIAMOND, Version 3.2f. Crystal Impact GbR, Bonn, Germany 1997-2010. [44] L.J. Boucher, J. Am. Chem. Soc. 92 (1970) 2725. [45] L. Ruhlmann, A. Nakamura, J.G. Vos, J.H, Fuhrhop, Inorg. Chem. 37 (1998) 6052. [46] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordenation Compounds parts A and B, 5th edition , Wiley Interscience Publication, 1997. [47] M.T. Caudle, C.K. Mobley, L.M. Bafaro, R. Lobrutto, G.T. Yee, T.L. Groy, Inorg. Chem. 43 (2004) 506. [48] F.S. Vinhado, M.E.F. Gardini, Y. Iamamoto, A.M.G. Silva, M.M. Q. Simões, M.G.P.M.S. Neves, A.C. Tomé, S.L. Rebelo, A.M.V.M. Pereira, J.A.S. Cavaleiro, J. Mol. Catal. A: Chem. 239 (2005) 138.

23

[49] J.R. Lindsay-Smith, Y. Iamamoto, F.S. Vinhado, J. Mol. Catal. A: Chem. 252 (2006) 23. [50] J. T. Groves, D. V. Adhyam, J. Am. Chem. Soc. 106 (1984) 2177. [51] S.M.G. Pires, R. De Paula, M.M.Q. Simões, M.G.P.M.S. Neves, I.C.M.S. Santos, A.C. Tomé, J.A.S. Cavaleiro, Catal. Commun. 11 (2009) 24. [52] A.M.A.R. Gonsalves, R.A.W. Johnstone, M.M. Pereira, J. Shaw, J. Chem. Soc. Perkin Trans. 1 (1991) 645. [53] B. Meunier, J. Bernadou, Top. Catal. 21 (2002) 47. [54] J. Haber, L. Matachowski, K. Pamin, J. Poltowicz, J. Mol. Catal. A: Chem. 198 (2003) 215. [55] S. Rayati, F. Nejabat, S. Zakavi, Inorg. Chem. Commun. 40 (2014), 82. [56] K.A.D.F. Castro, M.M.Q. Simões, M.G.P.M.S. Neves, J.A.S. Cavaleiro, R.R. Ribeiro, F. Wypych, S. Nakagaki, Appl. Catal. A: Gen. 503 (2015) 9. [57] K.A.D.F. Castro, S. Silva, P.M.R. Pereira, M.M.Q. Simões, M.G.P.M.S. Neves, J.A.S. Cavaleiro, F. Wypych, J.P.C. Tomé, S. Nakagaki, Inorg. Chem. 54 (2015) 4382.

24

SUPPORTING INFORMATION Synthesis, characterization and catalytic activity under homogeneous conditions of ethylene glycol substituted porphyrin manganese (III) complexes Kelly A. D. F. Castroa,b, Fernando H. C. de Limaa, Mário M. Q. Simões b, M. Graça P. M. S. Neves b, Filipe A. Almeida Pazc, Ricardo F. Mendesc, Shirley Nakagakia, José A. S. Cavaleirob a

Laboratório de Química Bioinorgânica e Catálise; Universidade Federal do Paraná

(UFPR), CP 19081, CEP 81531-990, Curitiba, Paraná, Brazil. b

Department of Chemistry and QOPNA, University of Aveiro, 3810-193 Aveiro,

Portugal. c

Department of Chemistry, CICECO – Aveiro Institute of Materials, University of Aveiro,

3810-193 Aveiro, Portugal

Contents Figure S1. HRMS-ESI(+): m/z spectrum of MnP2a………………………………...............2 Figure S2. HRMS-ESI(+): m/z spectrum of MnP2b………………………...………….……2 Figure S3. HRMS-ESI(+): m/z spectrum of MnP3……………………………...……...…....3 Figure S4. UV-VIS spectrum of MnP2a in methanol………………………………....…..…4 Figure S5. Schematic representation of the sole O–H···O hydrogen bonding interactions (via the pendant ethylene glycol moieties of the porphyrin molecules) present in the crystal structure of compound P4·2(DMSO)……………………………………….….……..5

1

1111.04189 C 48 H 18 O 4 N 4 F 18 Mn 0.29313 ppm

100 95 90 85 80 75 70

Relative Abundance

65 60 55 50 45 40 35 30 25 20 1113.04697

15 1109.04763

10

1110.05071

5 0

1114.05006

1101.75749 1103.80516 1105.70973 1107.05530 1102

1104

1106

1108

1110

1112

1114

1117.11614 1116

1118

1120.08109 1120

1122

1124

m/z

Figure S1. HRMS-ESI(+): m/z spectrum of MnP2a. 1111.04255 C 48 H 18 O 4 N 4 F 18 Mn 0.88964 ppm

100 95 90 85 80 75 70

Relative Abundance

65 60 55 50 45 40 35 30 25 20 1113.04773

15 10

1109.04867 1108.03900

5 0

1101.83171 1102

1105.03746 1104

1110.05163

1114.05080

1107.03408

1106

1108

1110

1112

1114

1117.11635 1116

1118

1120

1122

1124

m/z

Figure S2. HRMS-ESI(+): m/z spectrum of MnP2b.

2

Figure S3. HRMS-ESI(+): m/z spectrum of MnP3.

3

457

Intensity (a.u.)

551 497 365 390 767

676 480

350

400

450

500

510

540

550

570

600

630

600

660

650

690

720

700

750

780

750

800

Wavelength (nm) Figure S4. UV-VIS spectrum of MnP2a in methanol.

4

Figure S5. Schematic representation of the sole O–H···O hydrogen bonding interactions (via the pendant ethylene glycol moieties of the porphyrin molecules) present in the crystal structure of compound P4·2(DMSO). Hydrogen bond geometrical details: O4–H4···O3i with dD···A = 2.747(5) Å and <(DHA) of 145º. Symmetry transformation used to generate equivalent atoms: (i) 1+x, 0.5+y, 1.5-z. Color scheme as in Figure 1. Hydrogen bonding interactions depicted as dashed green lines.

5

Highlights • Synthesis and characterization of ethylene glycol substituted Mn(III) porphyrins • Catalysis with ethylene glycol substituted Mn(III)porphyrins at ambient temperature • Oxidation of cis-cyclooctene and cyclohexane in the presence of hydrogen peroxide • Catalysts’ performance depends on the number of ethylene glycol moieties present • Tetrasubstituted free-base porphyrin is studied by single-crystal X-ray diffraction

7