Journal of Colloid and Interface Science 450 (2015) 339–352

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Manganese chlorins immobilized on silica as oxidation reaction catalysts Kelly A.D.F. Castro a,b, Sónia M.G. Pires b, Marcos A. Ribeiro a, Mário M.Q. Simões b, M. Graça P.M.S. Neves b, Wido H. Schreiner a, Fernando Wypych a, José A.S. Cavaleiro b,⇑, Shirley Nakagaki a,⇑ a b

Laboratório de Bioinorgânica e Catálise, Universidade Federal do Paraná (UFPR), CP 19032, CEP 81531980 Curitiba, Paraná, Brazil Department of Chemistry and QOPNA, University of Aveiro, 3810-193 Aveiro, Portugal

g r a p h i c a l a b s t r a c t

a r t i c l e

i n f o

Article history: Received 19 February 2015 Accepted 11 March 2015 Available online 20 March 2015 Keywords: Chlorins Porphyrins Immobilization Sol–gel Heterogeneous catalysis Oxidation

a b s t r a c t Synthetic strategies that comply with the principles of green chemistry represent a challenge: they will enable chemists to conduct reactions that maximize the yield of products with commercial interest while minimizing by-products formation. The search for catalysts that promote the selective oxidation of organic compounds under mild and environmentally friendly conditions constitutes one of the most important quests of organic chemistry. In this context, metalloporphyrins and analogues are excellent catalysts for oxidative transformations under mild conditions. In fact, their reduced derivatives chlorins are also able to catalyze organic compounds oxidation effectively, although they have been still little explored. In this study, we synthesized two chlorins through porphyrin cycloaddition reactions with 1.3-dipoles and prepared the corresponding manganese chlorins (MnCHL) using adequate manganese(II) salts. These MnCHL were posteriorly immobilized on silica by following the sol–gel process and the resulting solids were characterized by powder X-ray diffraction (PXRD), UVVIS spectroscopy, FTIR, XPS, and EDS. The catalytic activity of the immobilized MnCHL was investigated in the oxidation of cyclooctene, cyclohexene and cyclohexane and the results were compared with the ones obtained under homogeneous conditions. Ó 2015 Elsevier Inc. All rights reserved.

⇑ Corresponding authors. E-mail addresses: [email protected] (J.A.S. Cavaleiro), [email protected] (S. Nakagaki). http://dx.doi.org/10.1016/j.jcis.2015.03.028 0021-9797/Ó 2015 Elsevier Inc. All rights reserved.

340

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

1. Introduction Hydrocarbons are abundant in nature and represent about 50% of the petrochemical industry input. Due to the low reactivity of these compounds their use per se is limited: they are mainly applied as solvents and fuels. Developing catalysts that can promote selective hydrocarbon oxidation has long been a goal in the field of organic synthesis [1,2]. In this sense, efficient and selective naturally occurring catalytic systems have frequently inspired the rational design of new catalytic systems [3]. In recent decades, researchers have faced the challenge to understand the role that different metal complexes play in biological processes, particularly in enzymatic reactions [4]. The past 30 years have witnessed many reports on synthetic porphyrins, especially their Fe(III) and Mn(III) complexes, being described as efficient and selective catalysts of a variety of hydrocarbon oxidations. Indeed, these metallocomplexes have been found to mimic the activity of the cytochrome P450 monooxygenase enzymes bearing Fe(III) porphyrin moieties in their active sites [5–8]. Porphyrins and their structural derivatives, like chlorins, participate in key biological processes [9–14]. Iron complexes such as heme constitute the prosthetic group of important classes of enzymes and of a wide variety of proteins [8]. Chlorophylls, involved in photosynthesis, are magnesium-substituted chlorin complexes [15]. These reduced derivatives are usually green, because the loss of a double bond at the b-pyrrolic position of the macrocycle alters the molecule symmetry, shifting the absorption towards the red region of the spectrum [16,17]. The spectral properties of chlorins (absorption in the 620–680 nm wavelength range) can contribute to their potential application in photodynamic therapy (PDT) [18,19]. Additionally, synthetic MnCHLx (where x = chlorin 1, 2 or 3) appear to be promising catalysts for oxidation reactions. In fact, these complexes have furnished moderate to good oxygenated products yields during cyclohexane oxyfunctionalization reactions [7]. Fe(III) and Mn(III) complexes of synthetic chlorins can also catalyze the oxidation of sulfides as efficiently as metalloporphyrins [20]. On the basis of the efficiency and selectivity of biological systems, researchers have devoted great efforts to develop synthetic metalloporphyrins for different purposes [11,21–26], namely to obtain (i) end products or synthetic precursors for the chemical/ pharmaceutically industry (such as epoxides, alcohols, and acids), and (ii) compounds with pharmaceutical interest (such as isomers with specific pharmacological activity). In these processes, the desired catalytic selectivity towards certain products may require high degree of sophistication when modeling specific catalysts [23–27]. Apart from synthesizing novel suitable and resistant porphyrin ligands that can catalyze oxidation reactions after metallation, another strategy usually considered is the immobilization of the porphyrin complexes on rigid and inert inorganic supports, to obtain suitable catalysts for the oxidation of distinct hydrocarbons [28–32]. Catalyst immobilization (i) prevents the common catalytic species deactivation reactions that take place in homogeneous medium, avoiding catalytic efficiency loss and complex deactivation by self-oxidation [33], and (ii) offers the extra bonus of catalyst recovery from the reaction medium and reuse [30,31,33–37]. Silica is generally the support of choice when investigating the immobilization of metalloporphyrins and their analogues; silica is inexpensive and inert, and has high functionalization capacity [35,38]. Silica gel is an inorganic solid containing siloxane (SiAOASi) and silanol groups (SiAOH) [37]. Immobilizing porphyrin complexes on solids such as silica gel inhibits dimer

formation, because the matrix reduces the interactions among the molecules fixed on the support surface [30,34,38,39]. The sol–gel process usually furnishes the silica under mild conditions; it is a very promising method to obtain high-purity solids that are suitable to immobilize different complexes, including metallochlorins and metalloporphyrins (MP) [33,35]. In this context, we report here the immobilization of two easily accessible manganese chlorins (MnCHL) (Fig. 1) on silica obtained by the sol–gel process. We tested the efficiency of the resulting solids as catalysts under heterogeneous conditions in cyclooctene and cyclohexene epoxidation and in cyclohexane hydroxylation. The oxidative processes were performed using iodosylbenzene (PhIO) and hydrogen peroxide (for cyclooctene) as oxidant and for comparison the efficiency of the non-immobilized were also evaluated. 2. Experimental All the chemicals were purchased from Aldrich, Sigma, or Merck, and were of analytical grade. Iodosylbenzene (PhIO) was synthesized by hydrolysis of iodosylbenzenediacetate [40]; its purity was determined by iodometric titration [41]. The obtained solid was carefully dried under reduced pressure and kept at 5 °C. 2.1. Synthesis of manganese chlorins (MnCHL1 and MnCHL2) The manganese chlorins (MnCHL1 and MnCHL2) were prepared in three steps, according to the literature [7,42–45]: (i) synthesis of the two free-base porphyrins 5,10,15,20-tetrakis(pentafluorophenyl)porphyrin [H2(TPFPP)] (1) and 5,10,15,20-tetrakis(2,6dichlorophenyl)porphyrin [H2(TDCPP)] (2); (ii) synthesis of the respective chlorins (CHL1 and CHL2) by 1,3-dipolar cycloaddition reaction between 1 and 2 with azomethine ylide generated from N-methylglycine and paraformaldehyde; and (iii) metallation of the free-base chlorins (CHL1 and CHL2) using manganese(II) chloride in acetic acid or dimethylformamide (DMF) obtaining (Mn(III)CHL1 and Mn(III)CHL2, (Fig. 1)). The compounds were characterized by 1H and 19F NMR, UVVIS spectroscopy, and mass spectrometry; the analytical data are in accordance with the literature [7,43,44]. 2.2. Synthesis of manganese chlorin (MnCHL3) The metallation of the free-base chlorin CHL1 was also done by a modification of the conventional method of Kobayashi [46] using acetic acid as solvent, manganese(II) acetate under reflux and magnetic stirring for 6 h. After the metallation process, the solvent was removed under vacuum giving rise to a solid that was washed thoroughly with water to remove the excess of metal salt. The complex produced was dissolved in a small amount of methanol and by standing at ambient temperature for two days small green crystals in needle form (MnCHL3) were obtained. 2.3. Structure determination by single crystal X-ray diffraction X-ray data for a single crystal of MnCHL3 with approximate dimensions 0.111 mm  0.144 mm  0.560 mm were collected on an Bruker D8 Venture Photon 100 by using graphite-monochromated Cu Ka radiation (k = 1.5418 Å) at 100(2) K. Accurate unit cell dimensions and orientation matrices were determined by leastsquares refinement of the reflections obtained by h–X scans. The data were indexed and scaled with the ApexII Suite [47]. Bruker Saint and Bruker Sadabs were used to integrate and scaling of data, respectively. The structure was solved and refined using the

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

341

Fig. 1. Schematic representation for the preparation of the manganese(III) chlorins MnCHL1 and MnCHL2: (i) synthesis of the free-base porphyrins 1 and 2; (ii) synthesis of the corresponding chlorins CHL1 and CHL2; (iii) insertion of Mn(III) into the free-base chlorins CHL1 and CHL2, to obtain the corresponding Mn(III)CHL1 and Mn(III)CHL2.

Shelx2013 [48] Software Package. The refinement was performed with anisotropic full-matrix least-squares refinement on F2. 2.4. MnCHL1 or MnCHL2 immobilization on silica obtained by the sol–gel process The manganese chlorins immobilized on silica (solids labeled as MnCHLx-Si, where x = chlorin 1 or 2) were obtained by preparing the silica using the sol–gel process. To this end, TEOS was hydrolyzed in tetrahydrofuran (THF), HCl was the catalyst, and the reaction was conducted in the presence of one of the manganese complexes. The following molar ratios were employed: H2O/Si (TEOS) = 4, THF/HCl = 308, and H2O/HCl = 92; 0.009 mmol of MnCHL1 or 0.007 mmol of MnCHL2 were used in the synthesis. The reaction mixtures were stirred for 1 h at ambient temperature. After this time, no solid was formed, so the solution was aged at about 50 °C for five days. After aging and drying, a green glassy solid was obtained (the characteristic color of chlorins). The solid was exhaustively washed with dichloromethane, and the resulting supernatants were collected from the washing liquid and quantitatively analyzed by UVVIS spectroscopy, to determine the quantity of MnCHL that remained in the washed solution and the loading (mol of complex per mass of solid) for each solid. The glassy solids (designated MnCHL1-Si and MnCHL2-Si resulting from the immobilization of Mn(III)CHL1 and Mn(III)CHL2, respectively) were dried at 50 °C and analyzed by UVVIS, XRD, and EPR. A control sample of silica was prepared under the same conditions as described above, but without the addition of MnCHLx.

2.5. Catalytic oxidation reactions The oxidation of (Z)-cyclooctene (previously purified on neutral alumina column) by PhIO was accomplished in the presence of the prepared catalysts; the general following procedure was used: in a 1.5 mL reaction flask, about 10–20 mg of MnCHLx-Si (depending on the concentration (loading) of the Mn(III) complex on the support) and PhIO (at a catalyst/oxidant molar ratio of 1:50) were mixed, and the mixture deaerated with argon. (Z)-cyclooctene was added to the mixture at a catalyst/oxidant/substrate molar ratio of 1:50:5000. The reactions were performed in acetonitrile under inert atmosphere, magnetic stirring, at ambient temperature, in the dark. The reactions time varied from 15 min to 24 h. The reactions were quenched by addition of sodium sulfite (saturated solution in acetonitrile). Then, a small amount of methanol was added, to solubilize the unreacted PhIO, and the supernatant was quantitatively extracted from the reaction media using acetonitrile and transferred to a volumetric flask. The solutions containing the reaction products were analyzed by gas chromatography, and the reaction products were quantified on the basis of PhIO, using the internal standard methodology (undecane or octan-1-ol as internal standard). The oxidation of (Z)-cyclooctene was also carried out using hydrogen peroxide as oxidant. The general procedure with this oxidant is the following: in a 1.5 mL reaction flask, the catalyst (about 10–20 mg of MnCHLx-Si or 0.5 mg of MnCHLx), the co-catalyst (about 7.0 mg of ammonium acetate) were suspended in acetonitrile (500 lL) and purified (Z)-cyclooctene was added. The mixture

342

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

was maintained under magnetic stirring at ambient temperature in the absence of light. The oxidant (solution of H2O2 (30% v/v) prepared in acetonitrile (1:10)) was progressively added at intervals of 15 min. The catalyst/oxidant/substrate molar ratio was 1:140:350; 1:500:1000; 1:600:150; 1:600:2100; 1:700:1900 and reaction times were 1 h, 105 min or 2 h. After the reaction time then the supernatant was quantitatively extracted from the reaction media using acetonitrile and transferred to a volumetric flask. The solutions containing the reaction products were analyzed by GC and the results are expressed in terms of conversion of substrate, using the internal standard methodology (octan-1-ol as internal standard). The general procedure used to investigate the oxidation of cyclooctene was also used for cyclohexene (previously purified on alumina column) and cyclohexane, with PhIO as oxidant and reaction time of 1 h. After the oxidation reactions, the solid catalysts were thoroughly washed and dried for reuse. Control reactions were carried out using the same procedure described for cyclooctene oxidation. Reactions were performed using only (a) the substrate without PhIO and catalyst; (b) substrate + PhIO; and (c) substrate + PhIO + Si (without the MnCHLx). MnCHL1 and MnCHL2 in solution were also tested as catalysts under homogeneous conditions; the experimental procedure followed was similar to that described for heterogeneous catalysis, using a catalyst/oxidant molar ratio of 1:50. 2.6. Investigation of the stability of catalysts (MnCHL1 and MnCHL2) The stability of MnCHL1 and MnCHL2 in oxidation reactions was followed by UVVIS spectroscopy by considering their characteristic absorption band at about 650 nm. In any experiment it was used the catalyst in presence of solvent (acetonitrile), the substrate (cyclooctene or cyclohexane) and the oxidant (PhIO); each reaction was followed by UVVIS in different periods of time. Tests were performed also in the absence of the substrates. 2.7. Characterization techniques Qualitative and quantitative electronic spectra (UVVIS) of solution samples were recorded on a Shimadzu UV-2501PC spectrophotometer, in the 200–900 nm range, using a quartz cuvette with 1 cm optical path. Qualitative spectra of the solid samples were obtained by diffuse reflectance on a VARIAN Cary 100 spectrophotometer. The 1H and 19F NMR spectra were recorded on a Bruker Avance 300 at 300.13 and 282.38 MHz, respectively, using CDCl3 as solvent and TMS as internal reference. The mass spectrometry analyses were carried out on a 4800 Proteomics Analyzer mass spectrometer (MALDI TOF/TOF) or Thermo Fisher Scientific Inc. LTQ XL linear Ion Trap mass spectrometer (ESI-MS). FTIR spectra were acquired on a Biorad 3500 GX spectrophotometer in the 400–4000 cm 1, using KBr pellets. 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. For the X-ray diffraction (XRD) measurements, self-oriented films were placed on neutral glass sample holders. The measurements were performed in the reflection mode using a Shimadzu XRD-6000 diffractometer operating at 40 kV and 40 mA (Cu Ka radiation k = 1.5418 Å) with a dwell time of 2°/min. 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 293 or 77 K using liquid N2. X-ray Photoelectron Spectroscopy (XPS) spectra were measured on a VG ESCA 3000 with a base pressure of 2  10 10 mbar. Mg Ka (1253.6 eV) radiation was employed, and the overall energy resolution of the collected spectra was approximately 0.8 eV. The spectra were normalized to the maximum intensity after

subtraction of a constant background. Scanning Electron Microscopy (SEM) images of the samples were acquired on either a JEOL 5190 microscope operating at 15 keV or a JEOL JSM6360LV operating at 15 keV. A small amount of the sample was placed on a sample holder, which was submitted to a gold metallization process and measured by a scan mode. The crosssection was analyzed by electron dispersive X-ray spectroscopy (EDX). Mapping of the elements was carried out to investigate the distribution of these elements on the surface layers. The images were processed using the Image Pro Plus software. The catalytic oxidation reaction products were identified and quantified with a Varian 3900 gas chromatograph (FID detector) equipped with a DB-5 type capillary column (J&W Scientific, 30 m  0.25 mm i.d., 0.25 lm film thickness, using helium as the carrier gas (30 cm3 s 1) and an Agilent 6850 chromatograph (FID detector) equipped with a DB-WAX type capillary column (J&W Scientific 30 m  0.25 mm i.d., 0.25 lm film thickness), using nitrogen as the carrier gas. Quantitative analysis was made by using an internal standard.

3. Results and discussion 3.1. Synthesis of manganese chlorins (MnCHL1 and MnCHL2) Porphyrins 1 and 2 were modified at their b-pyrrolic positions by 1,3-dipolar cycloaddition reactions with the azomethine ylide, generated in situ, from paraformaldehyde and N-methylglycine in the presence of the dipolarophile (the porphyrin), as described in the literature [7,18,43,44]. The obtained compounds named CHL1 and CHL2 were metallated with manganese(III) ion according to the Adler-Longo methodology [45], which furnished the complexes MnCHL1 and MnCHL2 (Fig. 1). Cycloaddition reactions of meso-tetraarylporphyrins with 1,3dipoles, such as azomethine ylides, generate mono-chlorins and bis-addition products (bacteriochlorins and isobacteriochlorins). By controlling the reaction conditions, mono- or bis-products can be preferentially obtained [43]. Because the azomethine ylide (3) is highly reactive and unstable, it has to be produced in situ by decarboxylation of the corresponding imine, which in turn originates from the reaction of N-methylglycine with paraformaldehyde. Compound (3) then reacts with the dipolarophile (porphyrins 1 and 2) to afford the corresponding chlorin macrocycle containing the pyrrolidine moiety. The formation of the green colored chlorins CHL1 and CHL2 was monitored by TLC and UVVIS. Their UVVIS spectra displayed an intense absorption band in the 650 nm region, which is typical of these reduced derivatives [18]. The 1H and 19F NMR, MALDI, UVVIS spectroscopy, FTIR spectroscopy, and EPR techniques were used to confirm the synthesis of the free-base chlorins and metallochlorins. The 1H and 19F NMR signals proved that the synthesized solids refer to the target porphyrins and chlorins and agree with literature data [43,44]. The UVVIS spectra displayed the characteristic bands of the compounds (Supporting Information). The FTIR spectra contained the expected bands for porphyrin compounds at 3313, 3116, and 2937 cm 1 corresponding to NH, CH (phenyl), and CH (pyrrole) stretching, respectively; and at 970 and 752 cm 1, relative to dNAH (in-plane) and dNAH (out-ofplane), apart from other bands related to skeletal ring vibrations. The typical band at 1589 cm 1, assigned to C@N stretching, was also detected, which distinguishes chlorins from porphyrins and bacteriochlorins [49]. The characteristic band of the pyrrolidine group (at 2850 cm 1, attributed to NACH3 stretching) was also observed (Supporting Information – Fig. S1). The FTIR spectra of MnCHL exhibited the typical bands of chlorins. The bands in the

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

region of 3000 cm 1, ascribed to symmetric and asymmetric NH stretching, and in the region of 1600 cm 1, referring to another dNH vibrational mode [50], disappeared (figures not shown). The results are consistent with the data described in the literature for these compounds [7,18,43,44]. UVVIS and FTIR techniques, besides mass spectrometry, were used to confirm the metallation process. The manganese chlorins (MnCHL1 or MnCHL2) perpendicular microwave polarization X-band EPR spectra at 77 K temperature in solid state are very similar to each other and shows a small sign of six lines at g = 2.0, which is characteristic of Mn(II) (S = 5/2) [51,52]. These results suggest that the Mn(III) complex is contaminated with a minor amount of the corresponding Mn(II) complex. After keeping the complexes in solution for several days, no signal was observed in the EPR spectrum, confirming the full oxidation of Mn(II) to Mn(III) in the core of the complex, (figures not show). This suggests that these complexes MnCHLx contain Mn(III) metallating the macrocycle. The Mn(III) ion has four unpaired electrons and d4 configuration with S = 2, typically featuring a pronounced Jahn–Teller distortion which results in a substantial spin–orbit coupling. As a result, mononuclear Mn(III) centers typically present no signals at the X-band EPR technique when a perpendicular microwave polarization is used [51,52]. Boucher [53] showed that in solution and in the presence of air, the manganese(II) porphyrin is easily oxidized for manganese(III) porphyrin. As a result of suitable ionic radius (60 pm), shifting Mn(III) ions fit perfectly into the cavity of the porphyrin ring, resulting in the displacement of the Soret band to a longer wavelength. The UVVIS spectrum showed the shift of the Soret band region to 470 nm (Supporting Information), and the reduction in the number of Q bands due to the change of symmetry, as expected for Mn(III) chlorins [7]. The complexes were also characterized by mass spectrometry, confirming the structures shown in Fig. 1.

343

3.2. Synthesis of manganese chlorin (MnCHL3) The metallation of the free-base chlorin CHL1 was also made using manganese(II) acetate and the reaction was monitored by UVVIS. Interesting, in this case, the UVVIS spectra of MnCHL3 showed clearly the presence of a Soret band at 422 nm probably due to the presence of Mn(II); a broad band at 458 nm also detect is the result of a minor contamination of Mn(III) species [54,55]. It is well established that structural changes in the core of the porphyrin can stabilize different oxidation states for manganese, such as Mn(II), Mn(III) and Mn(IV) [56–60]. Assis and co-workers [61] observed a mixture of Mn(II) and Mn(III) porphyrins using H2(PFTDCCl8PP). In this case, the presence of electron-withdrawing chlorins at the periphery of the porphyrin is able to stabilize preferentially manganese(II); the same behavior was also observed by Meunier [62] and Mansuy [63] using porphyrins containing bulky substituents. Nango et al. [64] used EPR spectroscopy for the characterization of the Mn chlorin complexes. The EPR spectra for manganese chlorophyll derivatives (MnPChlide a ME and MnMPDME) showed signals in the region of g = 2.0, consisting of 6-lines, which is characteristic of Mn(II). The perpendicular microwave polarization X-band EPR analysis of the MnCHL3 was made [51,52] and presents an isotropic broad signal in the region of g = 2.0 (Supporting Information – Fig. S2). This signal can be attributed to the presence of Mn(II) species in the solid MnCHL3, since in the Mn(II) ion a system with 5 electrons in an environment of D4h symmetry is expected, being a paramagnetic system and, therefore, EPR active [51]. Additionally, MnCHL3 complex was successfully crystallized from methanol which allowed the isolation of crystals, whose structure was unveiled from single-crystal X-ray diffraction. In this structure Mn(II) lies over a special position (0.5, 0.5, 1) with 1/4 of occupancy with four crystallographic related nitrogen atoms in its

Fig. 2. View of compound MnCHL3 with four symmetric units and atomic displacements ellipsoids at the 50% level. Hydrogen atoms were omitted for sake of simplicity. (i = y, x + 1, z + 2; ii = x + 1, y + 1, z; iii = y + 1, x, z + 2).

344

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

Table 1 Crystal data and refinement final parameters. C50H28F20MnN5O2 Mr = 1165.71 Tetragonal, P4 a = 13.8948(4) Å c = 6.4631(2) Å V = 1247.80(8) Å3 Z=1 F(0 0 0) = 584 Detector resolution: 10.4167 pixels mm 1 f and w scans Absorption correction: multi-scan SADABS V2014/2 (Bruker AXS Inc.) Tmin = 0.45, Tmax = 0.70 11,460 measured reflections 2115 independent reflections Refinement on F2 Least-squares matrix: full R[F2 > 2s(F2)] = 0.103 wR(F2) = 0.313 S = 1.54 2115 reflections 203 parameters 53 restraints

Dx = 1.551 Mg m 3 Cu Ka radiation, k = 1.54178 Å Cell parameters from 563 reflections h = 3.1–35.0° l = 3.25 mm 1 T = 100 K 0.82  0.14  0.12 mm 1991 reflections with I > 2s(I) Rint = 0.035 qmax = 68.7°, qmin = 3.2° h = 16 ? 16 k = 16 ? 16 l= 7?6 Hydrogen site location: mixed H-atom parameters constrained w = 1/[s2(Fo2) + (0.2P)2] where P = (Fo2 + 2Fc2)/3 (D/s)max = 0.033 Dqmax = 3.25 e Å 3 Dqmin = 0.51 e Å 3 Absolute structure: Refined as an inversion twin Absolute structure parameter: 0.24(2)

Primary atom site location: structureinvariant direct methods

coordination. Fig. 2 shows the unit cell packing arrangement. The distance Mn1AN1 of 2.016(7) and Mn1AO1 of 2.156(10) corroborates a structural model for Mn(II) chlorin complex. Crystal data and refinement final parameters are presented in Table 1 and the complete crystallographic data can be found in CCDC 1044150. The complex MnCHL3 was also characterized by mass spectrometry (ESI-MS) (m/z): 1084.12 [M+H]+ corresponding to the formula C47H15F20MnN5 (Supporting Information – Fig. S3) confirming the expected structure for the Mn chlorin. UVVIS spectrum of the MnCHL3 crystals dissolved in dichloromethane showed the characteristic Soret band for manganese(II) and the Q-bands at 518 and 572 nm (Supporting Information – Fig. S4). The solution of MnCHL3 was slowly oxidized by air after a few hours, showing a characteristic spectrum of Mn(III) chlorin. The fact that the chlorin CHL1, after metallation with manganese acetate (compound MnCHL3), presents evidences of Mn(II) and also Mn(III) species, while the corresponding precursor porphyrin (P1) only shows characteristic spectrum of Mn(III), indicates that there are differences in the reduction potential of the porphyrin and the chlorin derivative. Smith and co-workers [65] showed that the insertion of metal ions in the chlorins’ core, especially with cadmium, magnesium and zinc ion, reduces the standard oxidation potential, which for a chlorin is already ca. 0.3 V lower than for the corresponding porphyrin complex [65]. It is well established in literature that metalloporphyrins are more stable that the corresponding metallochlorins, being oxidized more easily (by approximately 300 mV) [66] thus being possible to correlate the electrochemical data with orbitals HOMO and LUMO. Kadish [67] noted that changes in the planarity of the ring destabilize the HOMO more than the LUMO, and the LUMO is even more affected by the effect of ring substituents and consequently different redox potentials are observed. The oxidation and reduction potentials follow the energy changes of the HOMO and LUMO observed for porphyrins and chlorins. The oxidation potentials found in the literature for H2TPP is 0.95 V and 0.88 V for the chlorin derivative [68].

It was observed that the MnCHL3 in solution is slowly oxidize to Mn(III) affording MnCHL1; this fact was confirmed by UVVIS and EPR spectroscopies. The UVVIS spectrum of the MnCHL3 in dichloromethane showed the displacement of the Soret band, from 422 nm observed before exposure to air, to 458 nm as expected for Mn(III) chlorins (UVVIS (CH2Cl2) kmax, nm: 458, 536 and 642 nm). The EPR spectrum of the solution of MnCHL3 after exposure to air using perpendicular microwave polarization X-band at ambient temperature or 77 K did not show EPR signals thus suggesting that Mn(II) was fully oxidized to Mn(III) (Supporting Information – Fig. S5) [51].

3.3. MnCHL1 or MnCHL2 immobilization on silica obtained by the sol– gel process The sol–gel methodology is an effective method to obtain silica nanoparticles (SNP) and this solid is frequently used to immobilize different compounds, including charged or neutral porphyrins and their analogues [33–35,69]. The first attempts to immobilize porphyrins on silica [70] used the procedure developed by Stöber [71] and involved hydrolysis and condensation of an adequate silane in basic medium, affording non-covalently immobilized porphyrins inside SNP [70]. This co-condensation strategy furnishes monodisperse SNP that are particularly interesting to design catalytically active and biologically compatible materials [70]. The onset of the hydrolysis and polycondensation reactions occurs at numerous sites as soon as TEOS and water mix; acids or bases can catalyze the reaction. The pH strongly affects these hydrolysis and condensation reactions: acidic medium favors hydrolysis, whereas basic conditions trigger condensation before hydrolysis is complete [72,73]. In the presence of acid catalysts, the bimolecular nucleophilic substitution reactions proceed via protonation of the silica ethoxide, facilitating the nucleophilic attack by a water molecule. The intermediate formed return to a tetrahedral structure for the output of an alcohol group resulting in a silanol protonated species. In the condensation step the nucleophilic attack of a silanol group of another species containing protonated silanol or alkoxide groups leading to the formation of the silica network [71]; in basic medium, hydroxyl or silanolate anions attack Si directly. Here, the silica synthesis by the sol–gel process was made in the presence of MnCHL1 or MnCHL2 using the alkoxysilane TEOS, a mixture of THF and water as solvent and HCl as catalyst. The acid conditions were employed because this route ensures maximum hydrolysis before starting the condensation process, resulting in the improvement of MnCHLx incorporation into the SNP [33,35]. The obtained solids MnCHL1-Si and MnCHL2-Si are green, and the color intensity depends on the MnCHLx loading (Table 2). On the other hand, the control silica sample, prepared in the absence of MnCHLx, is white in color. Several factors affect the immobilization of complexes on silica obtained by the sol–gel process, such as the precursor nature, the catalysts, the solvents, the pH, the reaction temperature and the amount of water. For example, Iamamoto and co-workers [74] showed that the porphyrin immobilization degree and the

Table 2 Immobilization of MnCHLx on silica obtained by the sol–gel process. Solid

Percentage of MnCHL immobilization (%)

Loading (MnCHL per gram of silica) (mol g

MnCHL1-Si MnCHL2-Si

25.0 100.0

1

1.9  10 5.3  10

)

% (m/m) 6 6

0.20 0.60

345

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

higher MnCHL2 immobilization degree (Supporting Information, Fig. S7). The FTIR spectra present only the characteristic bands of the silica support (Supporting Information, Fig. S8): there is an intense wide band in the region of 3444 cm 1, due to the typical axial deformation of superficial SiAOH groups. A large band at 1642 cm 1, ascribed to water molecules adsorbed/absorbed on the silica, also appears. The band in the region of 1100 cm 1 is typical of the stretching of the 180 degree angle of the SiAOASi groups of the four tetrahedral SiO2. The band at 943 cm 1 corresponds to SiAOH vibration, whereas the band at 793 cm 1 refers to the distorted structure of SiO4 [50]. Vibrational bands of the MnCHLx do not appear in the FTIR spectra, as a result of the low concentration of these complexes relatively to the support (Table 2). SEM images reveal that both the pure silica and the solids MnCHLx-Si have a solid, heterogeneous surface (Supporting Information, Fig. S9). EPR studies were made to investigate the metal ion oxidation state in the solids MnCHLx-Si. The spin state of the metal ion could change, because MnCHLx immobilization on silica could distort the porphyrin core [35,57]. This distortion can be associated with the requirement of the MnCHLx in order to maximize the interactions on the silica surface or entrapment in the pores. The control silica (not shown) displays EPR-silent spectrum, as expected, and also the solids MnCHL1-Si and MnCHL2-Si did not present any EPR signal at conventional EPR analysis [75]. This absence of EPR signals at perpendicular microwave polarization X-band analysis of the solids MnCHLx-Si spectra indicates that the samples are free of contaminating paramagnetic species that could have been inserted into the silica during the synthetic process and that the only manganese species present in the prepared solids MnCHL1-Si and MnCHL2-Si are manganese(III) [51]. Diffuse Reflectance UVVIS spectra (DR UVVIS) also attest the presence of MnCHLx in the solids MnCHLx-Si (Fig. 3). The UVVIS spectra of the solids present the characteristic Soret band in the region of 400 nm and an intense band in the region around 650 nm, typical of chlorins [81]. 3.4. Catalytic oxidation reactions 3.4.1. (Z)-cyclooctene Table 3 lists the results for the (Z)-cyclooctene epoxidation by PhIO catalyzed by MnCHLx (x = 1, 2 or 3) under homogeneous and heterogeneous conditions. Comparing the results for MnCHL1 and MnCHL2 in homogeneous catalysis, the latter

638 450 644

Intensity (a.u)

morphology of the xerogel depend on the porphyrin and on the conditions of the sol–gel process. As already mentioned, the addition of catalyst during the synthesis of silica by the sol–gel process allows one to immobilize charged or neutral compounds like metalloporphyrins and analogues like metallochlorins [33,35,57,69]. The immobilization of neutral MnCHLx and metalloporphyrins can occur through interactions between the conjugated macrocycle ring p-electronic cloud and the highly hydroxylated silica surface [35]. Moreover, there is no doubt that the addition of Mn(III) chlorins during the silica preparation by the sol–gel process confines part of the catalyst like MnCHLx in the three-dimensional silica structure [35] resulting from the SiAOASi bonding. This network forms large particles and cavities that adequately immobilize complexes like metalloporphyrins and metallochlorins [69,75]. It is also expected that some catalyst quantity is not confined inside the silica structure [35] and so it is removed during the washing process of the prepared solid. Two Iron porphyrins ([FeIII(TPFPP)] – (pentafluorophenyl)porphyrin iron(III)) and [FeIII(DETC)] – (diethyldithiocarbamate) porphyrin iron(III)) were entrapped within a silica matrix by the sol–gel process. The solids were characterized by UVVIS and EPR and investigated as sensors for nitric oxide (NO) [76]. Garcia-Sanchez [77] reported the importance of hydrolysis control and condensation during the sol–gel process for avoiding leaching of the entrapped porphyrin. Ciuffi and co-workers [78] also prepared catalysts based on nitro-porphyrins entrapped into a silica matrix by the sol–gel methodology. The presence of porphyrins was confirmed by UVVIS. The materials were characterized by TGA, isotherm of nitrogen adsorption, 29Si NMR and SEM. Ciuffi and co-workers [79] studied the conditions for entrapment of iron porphyrins into porous silica using sol–gel methodology. The entrapped porphyrin was characterized by textural analysis, UVVIS and EPR. The UVVIS spectrophotometry was used to determine the concentration of MnCHLx in the silica, by measuring the amount of non-immobilized MnCHLx recovered during the washing process of the solids. Under the reaction conditions used here, MnCHL immobilization loading depends on the MnCHLx structure. In fact, MnCHL2 yields the best immobilization degree in the solid MnCHL2-Si in comparison to MnCHL1 in the solid MnCHL1-Si (1.9  10 6 mol/g and 5.3  10 6 mol/g, respectively). The bulkier chlorine substituents present in the MnCHL2 structure if compared with the fluorine ones in MnCHL1 seem to provide better MnCHL encapsulation. The solids MnCHL1-Si and MnCHL2-Si as well as the control silica (without MnCHLx) were analyzed by X-ray powder diffraction (XRD), Transmission Fourier Transform Infrared spectroscopy (FTIR), Electron Dispersive X-ray (EDX), X-ray Photoelectron Spectroscopy XPS, Electron Paramagnetic Resonance (EPR), Scanning Electron Microscopy (SEM) and UVVIS spectroscopy. The XRD patterns of all the solids (Supporting Information, Fig. S6), exhibit the same profile: a broad halo appears in the 2h region of 20° and 30°, characteristic of silica amorphous samples [35,75,80]. The XPS technique provides qualitative information about the chemical composition of the materials surface. The EDX and XPS analyses performed on both MnCHL1-Si and MnCHL2-Si reveal the typical peaks of the elements Si2p, C1s, and O1s, assigned on the basis of the characteristic binding energy of each element. The other elements (Mn, N, F, and Cl) present in the structure of the MnCHLx are not evident, probably because the MnCHLx concentration near the surface of the support is low. The surface Si/C atomic ratios of the solids MnCHL1-Si and MnCHL2-Si calculated from the XPS data are 4.5 and 2.3 respectively, confirming the

(a)

454 482 652

(b)

467 637

(c)

(d)

400

500

600

700

Wavelength (nm) Fig. 3. UVVIS spectra of the solid samples (a) MnCHL1 (b) MnCHL2 (c) MnCHL1-Si, and (d) MnCHL2-Si.

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

Table 3 (Z)-cyclooctene oxide yields obtained in the (Z)-cyclooctene epoxidation catalyzed by MnCHLx in homogeneous and heterogeneous media (MnCHLx-Si).a Catalyst

Run

Epoxide yield (%)b

MnCHL1 (homogeneous) MnCHL1-Sic MnCHL1-Si first reuse MnCHL1-Si second reuse MnCHL1-Si third reuse MnCHL2 (homogeneous) MnCHL2-Sic MnCHL2-Si first reuse MnCHL2-Si second reuse MnCHL2-Si third reuse MnCHL3 (homogeneous) Si + PhIO (control)d PhIO (control)e

1 2 3 4 5 6 7 8 9 10 11 12 13

83 58 52 54 56 99 64 65 66 45 80 12 10

a Reaction conditions: 1 h, ambient temperature, acetonitrile as solvent, inert atmosphere, MnCHL/PhIO/cyclooctene molar ratio = 1:50:5000. b The cyclooctene oxide yield was based on the amount of PhIO used in the reaction. The results represent an average of at least triplicate reactions. c The piece of glass obtained from the sol–gel process was hardly grounded, and the resulting powder was used as catalyst. d Control reaction for heterogeneous catalysis. Si = silica obtained by the sol–gel process without MnCHL. e Control reaction for homogeneous catalysis; the reaction took place using only PhIO and substrate under similar reaction conditions, without any catalyst.

MnCHL2 (run 6) gives rise to higher catalytic yield than MnCHL1 (run 1) (99% vs 83%), under the same conditions. These results indicate that MnCHL2 is more resistant to the oxidation conditions of the catalytic reaction and in consequence present higher catalytic performance. As observed for metalloporphyrins [82], bulky and/ or electronegative substituents in the ortho positions of the mesophenyl groups of the chlorin (or porphyrin) account for the excellent catalytic results. The electronic effect of the electronegative substituents (e.g., Cl and F) can increase the lifetime of the active catalytic species [83], improving catalytic efficiency (as observed for runs 1 and 6). The steric effect of the substituents (Cl vs F) can also enhance catalyst efficiency by preventing the contact between the catalytic species. The solid MnCHL3 maintained under inert atmosphere (i.e. containing Mn(II) in the chlorin core) showed similar yield like MnCHL1 (80% vs 83%); this result is expected because, in solution and in the presence of oxidant, the manganese(II) is oxidized during the catalytic reaction. The MnCHL3 was not immobilized on silica, since in solution and in the oxidation reaction conditions it behaves like MnCHL1. The catalytic activity of the Mn(III) complexes of porphyrins 1 and 2 (that originate the MnCHLx studied here) is well established in the literature. The oxidation of (Z)-cyclooctene by PhIO in the presence of these catalysts furnishes epoxide yields between 75% and 100%, depending on the reaction conditions [84–87]. Under the conditions used here, it has been observed that the Mn(III) complex of porphyrin 1 performs better than the Mn(III) complex of porphyrin 2 (100% and 87% epoxide yield, respectively) [85]. Using the Mn(III) complexes of porphyrins 1 and 2 in the oxidation of a more inert substrate (cyclohexane), Nakagaki and coworkers [58] established a relation for the activities of a series of Mn(III) porphyrins. The authors also verified that the Mn(III) complex of porphyrin 2 has equal or slightly higher catalytic activity than the Mn(III) complex of porphyrin 1. The yields obtained in [58] are similar to those achieved here for the corresponding MnCHL1 and MnCHL2. The oxidation of (Z)-cyclooctene by PhIO in the presence of the immobilized catalysts MnCHL1-Si (run 2) and MnCHL2-Si (run 7) under the same conditions referred for the homogeneous processes, affords the epoxide respectively in 58% yield and 64%. The

immobilization of MnCHLx on silica is probably responsible by the decrease in the efficiency of the oxidative heterogeneous processes when compared with the homogeneous processes since the catalytic species are less accessible to the reactants [35,57]. Comparing MnCHL1-Si and MnCHL2-Si, their catalytic efficiency is similar (58% – run 2 vs 64% – run 7, respectively), suggesting that the apparent structural advantage presented by MnCHL2 in homogeneous medium is minimized after immobilization on silica. Knowing that under heterogeneous conditions the access of the reactants (PhIO and cyclooctene) to the metal center is more difficult, longer reaction times could improve the efficiency of the catalytic system. Therefore, cyclooctene oxidation by PhIO in the presence of MnCHL1-Si and MnCHL2-Si was carried out for different reaction times (Fig. 4). After 24 h of reaction, both MnCHL1-Si and MnCHL2-Si furnished over 90% of epoxide yield, suggesting that the access of the reactants to the metal center is partially blocked by MnCHL immobilization on silica and the reaction time increase can surpass this barrier. Immobilized catalysts are advantageous: it is possible to recover and to reuse them, preferentially without catalytic efficiency loss. Silica as the catalyst support contributes to catalyst reuse since it is easy to separate from the catalysis solution by a simple procedure such as centrifugation followed by washing and drying processes. The reusability of MnCHL1-Si and MnCHL2-Si was investigated by performing cyclooctene oxidation reactions by PhIO under the conditions established for the first run. After 1 h, the catalyst was filtered, washed, dried, and subjected to another catalytic run. Both solids MnCHL1-Si and MnCHL2-Si were stable during the catalytic reactions; in fact, UVVIS spectroscopy did not detect any traces of leached MnCHL in the reaction solution after the reuse in the epoxidation reactions. The solids were recovered and reused again and their catalytic activity remained in general virtually the same for at least three runs (Table 3).

3.4.2. Cyclohexene Since the solids MnCHLx-Si showed good catalytic performance for (Z)-cyclooctene, they were also tested as catalysts in the oxidation of cyclohexene (Fig. 5). The oxidation of cyclohexene mediated by PhIO/MP (MP = metalloporphyrin) systems results from a competition between the C@C to give the epoxide, and the allylic position to produce the corresponding alcohol and/or ketone products (A + K) [35,88,89]. Usually the main products observed for the cyclohexene oxidation mediated by PhIO/MP in homogeneous systems are the

Cyclooctene oxide (Yield %)

346

100

(b)

90

(a)

80 70 60 50 40 30 20 10 0

1

2

3

4

5

6

24.0

Reaction time (hours) Fig. 4. (Z)-cyclooctene epoxidation by PhIO for different reaction times catalyzed by: (a) MnCHL1-Si (j) and (b) MnCHL2-Si (.).

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

cyclohexene oxide (the major product), and the allylic products in minor yields (cyclohex-2-en-1-ol and cyclohex-2-en-1-one). However, the allylic products can also be produced as main products by free radical routes not involving the active complex intermediate and generally involving molecular oxygen. The efficiency and selectivity of the catalytic reaction can be controlled by the reaction conditions (solvent, temperature, inert atmosphere and reactants molar ratio) and also by the catalyst [35,88,89]. Comparing the results for MnCHL1 and MnCHL2 in homogeneous catalysis in the oxidation of cyclohexene, the MnCHL1 (Fig. 5) gives rise to higher total yield (epoxide + alcohol + ketone) than MnCHL2, under the same conditions (90% vs 74%). These results are opposite to the results obtained in the oxidation of (Z)-cyclooctene. On the other hand, the epoxide yields for both catalysts were similar (70% vs 65%). Besides that, according to the oxidation products yields shown in Fig. 5, the selectivity to the epoxide can be calculated and vary between 78% for MnCHL1 (70% of epoxide vs 90% of total product yields) and 88% for MnCHL2. Although MnCHL2 showed the lower total yield, it was more selective for epoxide (65% of epoxide and 9% of alcohol + ketone) in comparison to MnCHL1. The epoxide and the allylic products were also observed in the absence of catalyst but with low epoxide yield (8% of epoxide and 50% or 11% allylic products) confirming that the MnCHLx are involved in the catalytic epoxide formation. These results suggest that the epoxide is produced by a catalytic route involving the MnCHLx and the allylic products are probably obtained in a radicalar route independent of the complex presence. The epoxide is favored in presence of PhIO and MnCHLx, because this oxidant prevents the formation of radical species [89,90]. As observed for the (Z)-cyclooctene oxidation, the MnCHLx-Si also presented good epoxidation results for cyclohexene and the solid MnCHL1-Si showed total yields (94%) higher than MnCHL2Si (72%). Again, as under homogeneous catalysis, in spite of the total yield being lower for MnCHL2-Si this presented higher epoxide yield and selectivity (64% of epoxide vs 8% of A + K) if compared to MnCHL1-Si (44% of epoxide vs 50% of A + K). The apparent catalytic similarity evidenced by MnCHL1 and MnCHL2 under homogeneous conditions is differentiated after their immobilization on silica. However, due to the major amount of allylic products for MnCHL1-Si in comparison to MnCHL2-Si, the total products yield was 94% for MnCHL1-Si vs 72% for MnCHL2-Si in heterogeneous catalysis. MnCHL x immobilization on silica makes the catalytic species less accessible to the reactants. This fact is evident for MnCHL1 (70% vs 44%) but not for MnCHL2 (65 vs 64) with cyclohexene, where other products (allylic products) are also produced. Using cyclooctene or cyclohexene as substrate, the epoxide yield drops when the process change from a homogeneous to a heterogeneous procedure. Using MnCHL1 in the oxidation of cyclooctene or cyclohexene, the epoxide yields are the following: 83% (homogeneous)/58% (heterogeneous) and 70% (homogeneous)/44% (heterogeneous).

Fig. 5. Cyclohexene oxidation by PhIO catalyzed by MnCHLx in homogeneous and heterogeneous media (MnCHLx-Si), A = alcohol and K = ketone.

347

On the other hand, for MnCHL2 the variation on the epoxide yield depends on the substrate used. This behavior is presumably related with the structures of the two porphyrins and with the arrangement of the complexes on the silica structure. The presence of the chlorine substituents in MnCHL2, which are more bulky ligands in comparison to fluorine, might justify a certain degree of difficulty for this complex to penetrate and be englobed by the silica during the synthesis by the sol–gel process. Because of this the MnCHL2 complex can be more at the surface of the silica than the MnCHL1, its catalytic behavior being less susceptible to the substrates access to the catalytic species. Thus, for this substrate, the catalytic behavior of MnCHL2 in solution or immobilized on silica is almost the same and the total yields are very similar (homogeneous 74% vs heterogeneous 72%). For the solid MnCHL1-Si a significant allylic products yield was observed, probably resulting from the non-catalytic route involving dioxygen. Controlling the presence of dioxygen is a problem inherent to heterogeneous catalysis since this gas might be present in the voids of the small pores and cavities of the silica, making it difficult to ensure its absence. The presence of dioxygen in the support was also suggested by taking into consideration the yield of allylic products obtained when only the support was used (Fig. 5). MnCHL1-Si and MnCHL2-Si are robust and reusable solid catalysts for this substrate also, since both solids have been recovered and washed for the removal of traces of products after the first use. UVVIS analysis of the solvents used in the washing process did not show any traces of MnCHLx leaching from the support.

3.4.3. Cyclohexane It is well established that iron or manganese porphyrins are efficient and selective catalysts for cyclohexane oxidation (homogeneous catalysis). High selectivity for the alcohol is also observed with this type of catalysts [1,3,23,26,82,91–94]. Cyclohexanol formation occurs via oxygen transfer from the manganese(V)-oxo porphyrin complex intermediate [95], which is the active species responsible for the hydroxylation reactions. For comparison reasons, similar cyclooctene and cyclohexene reaction conditions were used for cyclohexane oxidation in the presence of MnCHLx-Si (heterogeneous process) and MnCHLx (homogeneous process). The results obtained are shown in Fig. 6. In a previous study the precursors Mn(III) porphyrins 1 and 2 were investigated as catalysts in the oxidation of cyclohexane using iodosylbenzene as oxidant. These studies showed yields around 40% for alcohol, or for alcohol plus ketone, depending on the reaction conditions used, with slightly higher catalytic activity for the Mn(III) complex of porphyrin 1 [7,57,58,84]. The MnCHLx derivatives studied here also showed catalytic activity in the oxidation of this more inert substrate, with similar behavior in the homogeneous process for MnCHL1 (10% alcohol and 2% ketone) vs MnCHL2 (10% alcohol and 3% ketone). The yields obtained previously with Mn(III) porphyrins 1 and 2 were higher than those achieved for the corresponding MnCHL1 and MnCHL2, but the selectivity for the alcohol was maintained. Vinhado and co-workers [7] performed the reactions in the presence and absence of oxygen using iodosylbenzene with or without imidazole and MnCHL-1 as catalyst. The reactants molar ratio used by these authors was 1:10:7440 (catalyst:oxidant:substrate). In the presence of air they obtained better results to those obtained by us when using porphyrin 1 (around 40%). It was also observed an improvement of the alcohol yield when a co-catalyst was used [7]. Thus, it is possible to conclude that the difference in the yields observed here in comparison to the Vinhado’s results can be attributed to the use of a co-catalyst and higher amount of oxidant that can contribute to the deactivation of the catalyst (MnCHL1) in solution.

348

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

Fig. 6. Cyclohexane oxidation by PhIO catalyzed by MnCHLx in homogeneous and heterogeneous media (MnCHLx-Si).

Furthermore, we have verified that the best reaction medium for the catalytic reactions using metalloporphyrins similar to [M(TPFPP)]Cl (M = Fe or Mn) is a mixture of solvents such as dichloromethane:acetonitrile (1:1; v/v); [57,58] this mixture solubilizes both the substrate and PhIO, while the catalyst is not always completely soluble in this medium. The use of only acetonitrile is justified for ensuring complete solubilization of the catalyst and being more appropriate from an environmental standpoint, thus allowing also a better comparison between the performance of metallochlorin catalysis in solution and immobilized on a solid support. The immobilization of both MnCHLx changed the catalytic results (Fig. 6). The total yield (A + K) for the MnCHL1-Si decreased from 12% in the homogeneous process to 6% in the heterogeneous process. This yield decrease corroborates the observation already made using the two previous substrates: after immobilization the complex MnCHL1 is entrapped in the silica structure, thus hampering the access of the reactants (PhIO and substrate) to the catalytic metal center. This observation can be enhanced by the observation that for this catalyst the alcohol yield decreased to 1% and the ketone yield increased to 5% when homogeneous and heterogeneous process are compared, changing the reaction selectivity for the ketone. The presence of more ketone strongly suggests that the alcohol produced in the catalytic reaction assisted by the manganese complex is entrapped into the silica structure in the neighborhood of the catalytic center and is overoxidized to the ketone. After the immobilization, MnCHL2 also presented a decrease of the total yield (A + K), but in this case the decrease was smaller:

2.0

only from 13% (homogeneous) to 10% (heterogeneous). On the other hand the ketone yield was almost the same (homogeneous 3% vs heterogeneous 4%). These observations also corroborate the catalytic results with the two previous substrates: probably for MnCHL2-Si the MnCHL2 is less entrapped in the silica structure, so the access of the reactants to the metal center is not so difficult as for MnCHL1 and the catalyst exhibits a similar performance to the complex in solution, and the alcohol produced in the metal center located near the surface of the solid can escape and less overoxidation is observed in comparison to MnCHL1-Si. Although the results obtained under homogeneous conditions are better than the ones under heterogeneous conditions, the great advantage of catalysts MnCHL1-Si and MnCHL2-Si is the easy recovery at the end of the reaction, the stability and the possibility to be reused.

3.4.4. Stability of the catalysts (MnCHLx) In order to better understand the catalytic performance observed in solution for the complexes MnCHLx, for investigating if the MnCHLx/PhIO molar ratio used for the oxidation reactions carried out in homogeneous medium is adequate and also ideal for a good catalyst performance, and finally to monitor if the MnCHLx are resistant in the presence of oxidant, the oxidation reactions of (Z)-cyclooctene and cyclohexane were analyzed by UVVIS at different time intervals. The MnCHLx/PhIO molar ratio used in the catalytic studies was 1:50. It is observed that in the (Z)-cyclooctene oxidation reaction, the MnCHLx degradation is minimal (by monitoring the decrease of the chlorin characteristic band at the region of 600 nm – Fig. 7), being possible to conclude that an adequate molar ratio to prevent catalyst destruction was used in the catalytic study. It was observed that MnCHL1 show only a decrease of about 25% of the characteristic band (around 650 nm) and 17% for MnCHL2 (Fig. 7). The smallest degradation of MnCHL2 was expected since it presents the best catalytic performance in the oxidation of (Z)cyclooctene (Fig. 7). For cyclohexane, the intensity of the characteristic band at 645 nm decreased about 21% for MnCHL1 in the first 5 min of reaction and 66% at the end of the reaction (Fig. 8), corroborating the assumption that MnCHL1 is destroyed in the presence of larger PhIO quantities, since better yields were observed by Vinhado [7] using a molar ratio 1:10 MnCHL/PhIO. For MnCHL2, in the first 5 min the intensity of the characteristic band at 652 nm decreased about 44%, and 72% after 1 h of reaction,

3.0

650

(b)

(a) 2.5 2.0

Absorbance

Absorbance

1.5

1.0

0.5

1.5

642

1.0 0.5 0.0

0.0 450

500

550

600

650

Wavelength (nm)

700

750

800

400

500

600

700

800

Wavelength (nm)

Fig. 7. UVVIS spectrophotometric study of the MnCHLx solution in presence of PhIO and (Z)-cyclooctene at different reaction times and 1:50:5000 M ratio for MnCHL:PhIO:substrate. (a) MnCHL1 and (b) MnCHL2.

349

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

3.0

645

(a)

(b)

479

3

Absorbance

Absorbance

2.5

2.0

1.5

2

652

1

1.0

0 0.5 500

550

600

650

700

750

800

Wavelength (nm)

450

500

550

600

650

700

750

800

Wavelength (nm)

Fig. 8. UVVIS spectrophotometric study of the MnCHLx solution in presence of PhIO and cyclohexane at different times. (a) MnCHL1 and (b) MnCHL2.

650

(b)

Intensity (a.u)

(a)

510

540

570

600

630

660

690

720

750

780

Wavelength (nm) Fig. 9. (a) UVVIS spectrophotometric study of the MnCHL2 solution in presence of PhIO and no substrate at different times. (b) MnCHL1 solution in absence and presence of substrate. From left to right: absence of substrate, with (Z)-cyclooctene and with cyclohexane.

showing that the active catalytic species was probably used to oxidize mostly the chlorin ring and, to a minor extent, the substrate. The low miscibility of cyclohexane in acetonitrile also contributes to further degradation of the catalyst. Moreover, the oxidation of a more inert substrate such as cyclohexane provides increased consumption of the catalyst. In the absence of (Z)-cyclooctene or cyclohexane, in the first 5 min of reaction, for both complexes the characteristic band of MnCHLx disappears and these results demonstrate that catalyst lifetime is improved in the presence of an excess of substrate, which prevents the degradation of the catalyst. Fig. 9a shows the UVVIS spectra of MnCHL1 in the presence of iodosylbenzene and solvent. UVVIS spectra for MnCHL2 are similar to those shown in Fig. 9a. Fig. 9b illustrates MnCHL1 profile in solution with PhIO, in the presence and absence of substrates; after 1 h of reaction the destruction of MnCHL1 in the absence of substrate can be observed. The study of the stability of MnCHLx in solution is in agreement with the catalytic results observed for both substrates. 3.4.5. Oxidation of (Z)-cyclooctene using hydrogen peroxide as oxidant Since MnCHLx are stable in the presence of (Z)-cyclooctene and iodosylbenzene, the hydrogen peroxide was used as oxidant. The

results are summarized in Table 4. When hydrogen peroxide is used as oxidizing agent, the use of a co-catalyst is frequently necessary [7,44,96]. In this study, ammonium acetate was used as co-catalyst. The function of the co-catalyst is to promote the formation of the active catalytic species; it acts both as acid–base catalyst, and favoring the axial heterolytic bond cleavage, thus favoring the formation of oxo-complexes [96]. Table 4 resumes the catalytic results obtained using hydrogen peroxide and MnCHLx (homogeneous and heterogeneous catalysis). MnCHL1 presents a percentage of conversion of (Z)-cyclooctene to the corresponding epoxide ranging from ca. 3.0% to 12% (runs 1–4, Table 4). Under higher hydrogen peroxide molar ratios (run 1 and 2) the conversion percentage is lower, this suggesting that some destruction of MnCHL1 occurs under this conditions. This situation can be overcome using higher concentration of substrate (run 3), and in the presence of an excess of substrate and low oxidant molar ratio (at 1:150:600 catalyst/oxidant/substrate molar ratio) the best catalytic performance was obtained (run 4). The increase of the reaction time is favorable for the catalytic reaction (runs 1 and 2) under similar molar ratio conditions. Taken together, these data demonstrate that the catalyst lifetime is improved in the presence of an excess of (Z)-cyclooctene, corroborating the previously discussed results for the others substrates and oxidant.

350

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

Table 4 (Z)-cyclooctene oxidation catalyzed by MnCHLx under homogeneous and heterogeneous conditions using hydrogen peroxide as oxidant. Run

Cat:Ox:Sub molar ratioa

MnCHL1

1 2 3 4

a

1:500:1000 1:500:1000c 1:600:2100b 1:150:600a

3.0 4.2 6.5 12.0

19 27 39 19

MnCHL2

5 6 7 8

1:500:1000a 1:500:1000c 1:700:1900a 1:600:2100b

25.0 27.0 19.0 44.0

130 138 115 274

MnCHL2-Si

9 10 11 12 13

1:500:1000a 1:500:1000b 1:140:350a 1:1500:150a 1:600:2100c

6.0 5.0 11.4 15.0 16.2

29 24 16 23 97

MnCHL2-Si

14

1:50:5000f

Catalyst

Conversion (%)d

0.65

TOF (h

1 e

)

30

Reaction conditions: different MnCHL:H2O2:substrate molar ratio, reaction time 1 ha, 105 minb and 2 hc. All reactions were performed at least in duplicate. dPercent conversion was calculated based on the amount of product formed. eTurnover Frequency (h 1) = number of moles of product per number of moles of catalyst per hour. Turnover Frequency was calculated from the reaction rate at 60 min. f MnCHL:PhIO:substrate molar ratio, reaction time 1 h.

For MnCHL2 a similar behavior was observed; the best conversion was obtained using high excess of substrate and lower concentration of hydrogen peroxide (run 8). With this catalyst the percentage of conversion of (Z)-cyclooctene into the epoxide ranges from ca. 19.0% to 45% (runs 5–8, Table 4) and these are higher than the values observed with MnCHL1, similarly to the results with iodosylbenzene. In the presence of an excess of hydrogen peroxide a catalaselike reaction may be occurring in which the catalytic active species formed can be involved in the transformation of hydrogen peroxide into molecular oxygen and water through the well described OAO homolysis, resulting in lower epoxide yield and dioxygen evolution [97]. This behavior is frequently observed using catalysts immobilized on solid supports [98,61]. Nevertheless, the good results obtained for MnCHL2 motivated us to study the solid MnCHL2-Si in heterogeneous catalysis with this oxidant. As expected, when MnCHL2-Si was used as catalyst, in the presence of hydrogen peroxide, lower results than those obtained in homogeneous catalysis were observed (Table 4, run 9–13). The lower results obtained in heterogeneous catalysis can be also due to the dismutation of H2O2. However, the reactions carried out using hydrogen peroxide showed good turnover frequency [Table 4, TOF values from 16 (run 11) to 29 (run 9)], generally slightly lower than those observed in the reactions using PhIO as oxidant (TOF = 30, run 14), except for run 13 (TOF = 97). Also in this case, like in homogeneous catalysis, better results are obtained when an excess of substrate is used (run 13). The excess of substrate can minimize hydrogen peroxide dismutation. No substrate conversion was observed with hydrogen peroxide in the absence of MnCHL2-Si. Despite the lower performance evidenced by MnCHL2-Si in comparison to MnCHL2 in homogeneous catalysis, MnCHL2-Si has the great advantage of not leaching the immobilized complex, since no trace of the complex was observed in solution after reaction, using UVVIS spectroscopy; thus, as observed with iodosylbenzene, MnCHL2-Si can be recovered and reused with this oxidant also.

resulting materials (MnCHLx) was investigated for (Z)-cyclooctene and cyclohexene epoxidation, and in the cyclohexane hydroxylation. The manganese(III) chlorins exhibit excellent catalytic activity in homogeneous catalysis, similar to that observed for the corresponding manganese(III) porphyrins 1 and 2. Although metalloporphyrins and metallochlorins present related structures and chemical properties, the latter have been poorly studied as oxidation catalysts. It was also found that the solids resulting from manganese(III) chlorin immobilization on silica catalyze the oxidation reaction in similar ways, although in lower yields than those obtained for the same metallochlorins, if used in homogeneous processes under similar reaction conditions. Performing the corresponding reactions under heterogeneous conditions, for longer reaction times, products yields similar to those in homogeneous medium are obtained. This suggests that the catalyst immobilization makes more difficult the access of the reactants to the metal center. On the other hand, the behavior of MnCHL1 was slightly different from the catalytic behavior of MnCHL2; this difference can bring some insights to the way the immobilization of the complexes on the silica support takes place. The MnCHL2 might be more on the surface of the silica and MnCHL1 is more entrapped in the structure of such support. This immobilization way is difficult to explain but it can result from the difference in the substituents in the meso-phenyl groups of each chlorin ligand. The solid catalysts are very stable and no leaching was observed by UVVIS spectroscopy. So they could be recovered and reused at least for 3 times. In the oxidation of cyclohexene the epoxide preferential formation was observed. MnCHL2-Si showed similar results to those obtained in homogeneous catalysis. Although MnCHL1-Si and MnCHL2-Si showed lower results in the oxidation of cyclohexane, both catalysts were selective for the alcohol formation. The stability of the catalysts MnCHLx in homogeneous catalysis was investigated in the absence and presence of (Z)-cyclooctene or cyclohexane. The results obtained justify the catalytic activity observed for each of the complexes. The oxidation of (Z)-cyclooctene using H2O2, an environmentally friendly reagent, was also investigated. In homogeneous catalysis higher conversions of the substrates were observed using iodosylbenzene as oxidant. MnCHL2 is considered the best catalyst. Its heterogeneous derivative MnCHL2-Si, although giving rise to lower yields if comparated MnCHL2, has the advantage derived from its reuse possibility. These results point to further investigation into different metallochlorins as oxidation catalysts. We are currently developing and evaluating other related chlorin macrocycles. Acknowledgments The authors are grateful to Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Fundação Araucária, Fundação da Universidade Federal do Paraná (FUNPAR), and Universidade Federal do Paraná (UFPR) for financial support. The authors also thank Fundação para a Ciência e a Tecnologia (FCT, Portugal), European Union, QREN, FEDER and COMPETE for funding the QOPNA research unit (Project PEst-C/ QUI/UI0062/2013) and the Portuguese National NMR Network, also supported by funds from FCT. Kelly A.D.F. Castro also thanks CAPES for her PhD sandwich scholarship (Process 6883-10-9).

4. Conclusions Appendix A. Supplementary material Two manganese(III) chlorin complexes were prepared and characterized by several techniques. After their immobilization on silica by the sol–gel process, the catalytic activity of the

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2015.03.028.

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

References [1] D. Mansuy, Coord. Chem. Rev. 125 (1993) 129–142. [2] A. Rezaeifard, M. Jafarpour, H. Kavousi, M. Alipour, H. Stoeckli-Evans, Polyedron 30 (2011) 2303–2309. [3] M. Costas, Coord. Chem. Rev. 255 (2011) 2912–2932. [4] M.M.Q. Simões, R. Paula, M.G.P.M.S. Neves, J.A.S. Cavaleiro, J. Porphyrins Phthalocyanines 13 (2009) 589–596. [5] A.E. Shilov, G.B. Shulpin, Chem. Rev. 97 (1997) 2879–2932. [6] S.K. Mohapatra, P. Selvam, Topic. Catal. 22 (2003) 17–22. [7] 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) 138–143. [8] J.T. Groves, J. Inorg. Biochem. 100 (2006) 434–447. [9] K.M. Smith (Ed.), Porphyrins and Metalloporphyrins, Elsevier, Amsterdam, 1975. [10] M. Boulton, M. Rózanowska, B. Rózanowski, J. Photochem. Photobiol., B: Biol. 64 (2001) 144–161. [11] K.S. Suslick, in: K. Kadish, K. Smith, R. Guillard (Eds.), The Porphyrin Handbook, Academic Press, New York, 1999. [12] T.P. Wijesekera, D. Dolphin, in: R.A. Sheldon (Ed.), Synthetic Aspects of Porphyrin and Metalloporphyrin Chemistry, Marcel Dekker Inc., New York, 1994, p. 193 (Chapter 7). [13] R.A. Sheldon, Metalloporphyrins in Catalytic Oxidations, Marcel Dekker, New York, 1994. [14] J.T. Groves, Y.Z. Han, in: P.R.O. Montellano (Ed.), Cytochrome P-450: Structure, Mechanism and Biochemistry, second ed., Plenum Press, New York, 1995. [15] H. Scheer, Synthesis and Stereochemistry of Hydroporphyrins in the Porphyrins, Academic Press, New York, 1978. [16] R. Bonnett, Chem. Soc. Rev. 24 (1995) 19–33. [17] H.-J. Kim, J.S. Lindsey, J. Org. Chem. 70 (2005) 5475–5486. [18] N.M.M. Moura, F. Giuntini, M.A.F. Faustino, M.G.P.M.S. Neves, A.C. Tomé, A.M.S. Silva, E.M. Rakib, A. Hannioui, S. Abouricha, B. Röder, J.A.S. Cavaleiro, Arkivoc (2010) 24–33. [19] W.M. Sharman, C.M. Allen, J.E. Lier, Drug Discovery Today 4 (1999) 507–517. [20] R. Rahimi, A.A. Tehrani, M.A. Fard, B.M.M. Sadegh, Catal. Commun. 11 (2009) 232–235. [21] K. Suslick, P. Bhyrappa, J.-H. Chou, M.E. Kosal, S. Nakagaki, D.M. Smithhenry, S.R. Wilson, Acc. Chem. Res. 38 (2006) 283–291. [22] R. Ricoux, Q. Raffy, J.P. Mahy, C. R. Chim. 10 (2007) 684–702. [23] D. Mansuy, C. R. Chim. 10 (2007) 392–413. [24] J. Haber, L. Matachowski, K. Pamin, J. Poltowicz, J. Mol. Catal. A: Chem. 198 (2003) 215–221. [25] A. Fuert, A. Corma, M. Iglesias, E. Morales, F. Sánchez, J. Mol. Catal. A: Chem. 246 (2006) 109–117. [26] B. Meunier, Chem. Rev. 92 (1992) 1411–1456. [27] A.J. Appleton, S. Evans, J.R.L. Smith, J. Chem. Soc., Perkin Trans. 2 (1995) 281– 285. [28] C. Crestini, A. Pastorini, P. Tagliatesta, J. Mol. Catal. A: Chem. 208 (2004) 195– 202. [29] E.A. Vidoto, M.S.M. Moreira, F.S. Vinhado, K.J. Ciuffi, O.R. Nascimento, Y. Iamamoto, J. Non-Cryst. Solids 304 (2002) 151–159. [30] M. Halma, K.A.D.F. Castro, C. Taviot-Gueho, V. Prévot, C. Forano, F. Wypych, S. Nakagaki, J. Catal. 257 (2008) 233–243. [31] A.M. Machado, F. Wypych, S.M. Drechsel, S. Nakagaki, J. Colloid Interface Sci. 254 (2002) 158–164. [32] A.L. Faria, T.O.C. Mac. Leod, V.P. Barros, M.D. Assis, J. Braz. Chem. Soc. 20 (2009) 895–906. [33] G.M. Ucoski, K.A.D.F. Castro, K.J. Ciuffi, G.P. Ricci, J.A. Marques, F.S. Nunes, S. Nakagaki, Appl. Catal., A: Gen. 404 (2011) 120–128. [34] M. Halma, A. Bail, F. Wypych, S. Nakagaki, J. Mol. Catal. A: Chem. 243 (2006) 44–51. [35] K.A.D.F. Castro, M. Halma, G.S. Machado, G.P. Ricci, G.M. Ucoski, K.J. Ciuffi, S. Nakagaki, J. Braz. Chem. Soc. 21 (2010) 1329–1340. [36] F. Wypych, A. Bail, M. Halma, S. Nakagaki, J. Catal. 234 (2005) 431–437. [37] K.A.D.F. Castro, A. Bail, P.B. Groszewicz, G.S. Machado, W. Schreiner, F. Wypych, Appl. Catal., A: Gen 386 (2010) 51–59. [38] A. Bail, V.C. Santos, M.R. Freitas, L.P. Ramos, W.H. Schreiner, G.P. Ricci, K.J. Ciuffi, S. Nakagaki, Appl. Catal., B: Environ. 130–131 (2013) 314–324. [39] Y. Iamamoto, I.M. Idemori, S. Nakagaki, J. Mol. Catal. A: Chem. 99 (1995) 187– 193. [40] J.C. Sharefkin, H. Saltzman, Org. Synth. 43 (1960) 60. [41] A.I. Vogel, Análise Química Quantitativa, 5ª ed., Editora Guanabara Koogan, 1992. [42] A.M.A. Rocha Gonsalves, J.M.T.B. Varejão, M.M. Pereira, J. Heterocycl. Chem. 28 (1991) 635–640. [43] A.M.G. Silva, A.C. Tomé, M.G.P.M.S. Neves, A.M.S. Silva, J.A.S. Cavaleiro, Chem. Commun. (1999) 1767–1768. [44] 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) 24–28. [45] A.D. Adler, F.R. Longo, F. Kampas, J. Kim, J. Inorg. Nucl. Chem. 32 (1970) 2443– 2445. [46] H. Kobayashi, T. Higuchi, Y. Kaizu, H. Osada, M. Aoki, Chem. Soc. Jpn. 48 (1975) 3137–3141.

351

[47] APEX2 v2014.1-1 (Bruker AXS), AINT V8.34A (Bruker AXS Inc., 2013). [48] G.M. Sheldrick, Acta Crystallogr. 71 (2015) 3–8. [49] T. Mashiko, D. Dolphin, in: G. Wilkinson, R.D. Gillard, J.A. McCleverty (Eds.), Porphyrins, Hydroporphyrins, Azaporphyrins, Phthalocyanins, Corroles, Corrins and Related Macrocycles, Comprehensive Coordination Chemistry, vol. 2, Pergamon Press, 1987. [50] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordenation Compounds Parts A and B, fifth ed., Wiley Interscience Publication, 1997. [51] M.T. Caudle, C.K. Mobley, L.M. Bafaro, R. Lobrutto, G.T. Yee, T.L. Groy, Inorg. Chem. 43 (2004) 506–514. [52] D.P. Goldberg, J. Telser, J. Krzystek, A.G. Montalban, L.C. Brunel, A.G.M. Barrett, B.M. Hoffman, J. Am. Chem. Soc. 119 (1997) 8722–8723. [53] L.J. Boucher, J. Am. Chem. Soc. 90 (1968) 6640–6645. [54] P.A. Loach, M. Calvin, Biochemistry 2 (1963) 361–371. [55] L.J. Boucher, J. Am. Chem. Soc. 92 (1970) 2725–2730. [56] V.S. Silva, L.I. Teixeira, E. Nascimento, Y.M. Idemori, G.F. Silva, Appl. Catal., A: Gen. 469 (2014) 124–131. [57] 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) 129–141. [58] G.R. Friedermann, M. Halma, K.A.D.F. Castro, F.L. Benedito, F.G. Doro, S.M. Drechsel, A.S. Mangrich, M.D. Assis, S. Nakagaki, Appl. Catal., A: Gen. 308 (2006) 172–181. [59] J.T. Groves, M.K. Stern, J. Am. Chem. Soc. 26 (1988) 8628–8638. [60] B. Girek, W. Sliwa, J. Porphyrins Phthalocyanines 17 (2013) 1139–1156. [61] F.G. Doro, J.R.L. Smith, A.G. Ferreira, M.D. Assis, J. Mol. Catal. A: Chem. 164 (2000) 97–108. [62] P. Hoffmann, A. Robert, B. Meunier, Bull. Soc. Chim. Fr. 129 (1992) 85–97. [63] K. Ozette, P. Battioni, P. Leduc, J.F. Bartoli, D. Mansuy, Inorg. Chim. Acta 272 (1998) 4–6. [64] M. Kondo, T. Mitsui, S. Ito, Y. Kondo, S. Ishigure, T. Dewa, K. Yamashita, J. Nakamura, R. Oura, M. Nango, Langmuir 26 (2010) 7774–7782. [65] K.M. Smith, D.A. Goff, D.J. Simpson, J. Am. Chem. Soc. 107 (1985) 4946–4954. [66] M.G.H. Vicente, in: K. Kadish, K. Smith, R. Guillard (Eds.), The Porphyrin Handbook, vol. 1, Academic Press, New York, 1999, p. 154. [67] D. Chang, T. Malinski, A. Ulman, K.M. Kadish, Inorg. Chem. 33 (1984) 4186– 4188. [68] L.G. Arnaut, in: R.V. Eldik, G. Stochel (Eds.), Advances in Inorganic Chemistry: Inorganic Photochemistry, vol. 63, 2011, pp. 188–229. [69] A.T. Papacídero, L.A. Rocha, B.L. Caetano, E. Molina, H.C. Sacco, E.J. Nassar, Y. Martinelli, C. Mello, S. Nakagaki, K.J. Ciuffi, Colloids Surf., A: Physicochem. Eng. Asp. 275 (2005) 27–35. [70] F. Figueira, J.A.S. Cavaleiro, J.P.C. Tomé, J. Porphyrins Phthalocyanines 15 (2011) 517–533. [71] W. Stöber, A. Fink, E. Bohn, J. Colloid Interface Sci. 26 (1968) 62–69. [72] H. Tanaka, T. Yamada, S. Sugiyama, H. Shiratori, R. Hino, J. Colloid Interface Sci. 286 (2005) 812–815. [73] E. Fagadar-Cosma, C. Enache, D. Vlascici, G. Fagadar-Cosma, M. Vasile, G. Bazylak, Mater. Res. Bull. 44 (2009) 2186–2193. [74] H.C. Sacco, K.J. Ciuffi, E.A. Vidoto, J.C. Biazzotto, C.A. Mello, D.C. Oliveira, O.A. Serra, O.R. Nascimento, Y. Iamamoto, J. Non-Cryst. Solids 284 (2001) 174–182. [75] S.T. Castaman, S. Nakagaki, R.R. Ribeiro, K.J. Ciuffi, S.M. Drechsel, J. Mol. Catal. A: Chem. 300 (2009) 89–97. [76] J.C. Biazzotto, J.F. Borin, R.M. Faria, C.F.O. Graeff, Mater. Res. Soc. Symp. Proc. 723 (2002) 161–166. [77] M.A. Garcia-Sanchez, A. Campero, J. Non-Cryst. Solids 333 (2) (2004) 226–230. [78] M.A. Silva, D.C. Oliveira, A.T. Papacidero, C. Mello, E.J. Nassar, K.J. Ciuffi, H.C. Sacco, J. Sol–Gel. Sci. Technol. 26 (2003) 329–334. [79] K.J. Ciuffi, D.C. Oliveira, H.C. Sacco, O.R. Nascimento, J. Non-Cryst. Solids 284 (1–3) (2001) 27–33. [80] G.S. Machado, K.A.D.F. Castro, O.J. de Lima, E.J. Nassar, K.J. Ciuffi, S. Nakagaki, Colloids Surf., A: Phys. Eng. Asp. 349 (2009) 162–169. [81] M. Erbacher, F.-P. Montforts, J. Porphyrins Phthalocyanines 15 (2011) 1070– 1077. [82] D. Dolphin, T.G. Traylor, L.Y. Xie, Acc. Chem. Res. 30 (1997) 251–259. [83] K.S. Suslick, Comprehensive Supramolecular Chemistry, Bioinorganic Systems, vol. 5, Elsevier, London, 1996. [84] E.C. Zampronio, M.C.A.F. Gotardo, M.D. Assis, H.P. Oliveira, Catal. Lett. 104 (2005) 53–56. [85] H.C. Sacco, Y. Iamamoto, J.R.L. Smith, J. Chem. Soc., Perkin Trans. 2 (2001) 181– 190. [86] 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. Solids 273 (2000) 100–108. [87] O.J. Lima, D.P. Aguirre, D.C. Oliveira, M.A. Silva, C. Mello, C.A.P. Leite, H.C. Sacco, K. Ciuffi, J. Mater. Chem. 11 (2001) 2476–2481. [88] A.J. Appleton, S. Evans, J.R. Lindsay-Smith, J. Chem. Soc., Perkin Trans. 2 (1995) 281–285. [89] J.R. Lindsay-Smith, P.R. Sleath, J. Chem. Soc., Perkin Trans. 2 (1982) 1009–1015. [90] E. Baciocchi, F. d’Acunzo, C. Galli, M. Loele, J. Chem. Soc., Chem. Commun. (1995) 429–430. [91] J.T. Groves, T.E. Nemo, J. Am. Chem. Soc. 105 (1983) 6243–6248. [92] P. Battioni, J.F. Bartoli, D. Mansuy, Y.S. Byun, T.G. Traylor, J. Chem. Soc. Chem. Commun. (1992) 1051–1053. [93] P. Battioni, J.P. Renaud, J.F. Bartoli, M. ReinaArtiles, M. Fort, D. Mansuy, J. Am. Chem. Soc. 110 (1988) 8462–8470.

352

K.A.D.F. Castro et al. / Journal of Colloid and Interface Science 450 (2015) 339–352

[94] P. Inchley, J.R. Lindsay-Smith, R.J. Lower, New J. Chem. 13 (1989) 669–676. [95] J.T. Groves, W.J. Kruper, R.C. Haushalter, J. Am. Chem. Soc. 102 (1980) 6375– 6377. [96] D. Mansuy, Pure Appl. Chem. 59 (1987) 759–770.

[97] M.A. Martinez-Lorente, P. Battioni, W. Kleemiss, J.F. Bartoli, D. Mansuy, J. Mol. Catal. A: Chem. 113 (1996) 343–353. [98] C. Piovezan, K.A.D.F. Castro, S.M. Drechsel, S. Nakagaki, Appl. Catal., A: Gen. 293 (2005) 97–104.