A novel porphyrazine ligand tailored to homogeneous metal catalyzed transformations

A novel centrosymmetric porphyrazine (Pz) 1 decorated with pyrazino-dibenzo[ b,f ]azepine units have been prepared via Linstead macrocyclization reaction of a dinitrile precursor. Accordingly, the peripheral azepine nitrogen offers a chemical handle for subsequent functionalizations. Characterization of the metal complexes (MPzs = Metal Porphyrazines) of 1 was accomplished and good catalytic performances were achieved in the Cu(II)- and Co(II)- catalyzed cyclopropanation with ethyl diazoacetate as a test reaction.


Introduction
Phthalocyanines (Pcs) are D 4h symmetry (due to mesomeric averaging as metal complexes, dications or dianions) macrocyclic compounds with aromatic 18 π -electrons, closely related to the natural occurring porphyrins.As porphyrins they could host a variety of metal ions inside the central cavity of the system.Pcs per se and their metal Pcs (MPCs) offer a wide assortment of extensively functionalized structures with classical application as pigments and dyes and, more recently, in photodynamic therapy, as photonic materials (e.g.mesogenics, NLO, optomaterials) and functionalized solids (e.g.][3][4] In addition to the above mentioned applications, Pcs had found place as ligands in a wide number of metal catalyzed organic transformations (e.g.oxidation, synthesis of nitrogencontaining compounds, C-C bond formation, reduction and many other useful synthetic procedures). 5The application of MPcs as catalysts is an emerging strategy and has attracted broad interests due to their accessibility in terms of the cost and straightforward preparation on a large scale as well as their chemical and thermal stability.Porphyrazines (Pzs) represent a structural analogue of porphyrines in which the meso carbons of porphyrines are replaced with N atoms.This structural similarity would make Pzs a good candidate for development of functional dyes, molecular devices and biomedical applications.A peculiar styudy on porphyrazines and related metal complexes was reported by Ercolani and coworkers. 6Recently, Pzs has attracted a lot of attention for their intense electronic absorption and emission in their near infrared region.Homogeneous and heterogeneous catalysis with metal porphyrazines (MPzs) has been extensively demonstrated, also their use as biomimetic catalyst was successfully achieved. 7owever, for some of these applications, MPzs are of disappointing use due to their scarce solubility in common organic solvents and this drawback could be overcome by an appropriate decoration of the periphery of the macrocycle. 8

Figure 1
In this paper we report the synthesis of a novel centrosymmetric Pz system (Figure 1), arising from the formal substitution of four benzo subunits in a Pc by a 9H-dibenzo[b,f]pyrazino [2,3-d]azepine moiety, wherein the peripheral azepine nitrogen atoms offer a chemical handle for subsequent functionalizations (i.e., introduction of lipophilic chains in order to address the above issue as in 1b).Accordingly, cyclopropanation with ethyl diazoacetate (EDA) catalyzed by the Cu(II)-and Co(II)-metal complexes 11 and 12, respectively, of the highly lipophilic N-hexadecanoyl porphyrazine 1b is also reported as test reaction.

Scheme 1
With ample supply of oxcarbazepine 2 available, our intention was to effect a one-pot tandem oxidation/N-decarbamoylation to yield diketone 3 required for the elaboration of porphyrazine ring system (Scheme 2).We initially considered application of a report by Heckendorn describing the transformation of 2 into 3 by a multi-step procedure, entailing α-carbonyl bromination of 2, nucleophilic substitution with KOAc followed by aerial oxidation. 10In our hands, however, yields of 3 remain unsatisfactory, notwithstanding strict adherence to conditions prescribed by Heckendorn.Therefore, we attempted to design a method that would enhance the yields and, possibly, through a shorter synthetic pathway.Several methods currently exist to obtain α-diketones by oxidation of α-methylene ketones.Among these are the use of venerable SeO 2 (Riley method), 11 seleninic acids and anhydrides, pyridinium chlorochromate (PCC) or KMnO 4 as oxidant agents.Multi-step approaches have also been reported like Kornblum oxidation of α-bromo ketones, nitrosation, microwave promoted oxidation or CuBr 2 adsorbed on alumina. 12ach method has its advantage and drawbacks in any given situation.For the sake of expediency, we opted for an oxidant able to gain 3 from 2 in only one step.Thus, a survey of oxidants was undertaken and on treatment of 2 with freshly sublimed SeO 2 (2.5 equiv) in refluxing 1,4-dioxane for 7 h, clean tandem oxidation/N-decarbamoylation pleasingly ensued to produce 3 in 90% yield. 13,14This compound precipitates directly from the reaction mixture as orange solid so that this process proved useful enough to prepare multigram quantities of 3. The mechanism of N-decarbamoylation is not clear at this time, but it seems plausible that reaction may commence with nucleophilic attack of amide carbonyl (as carboximidic acid tautomer) on SeO 2 , followed by proton transfer with formation of a N-cyano intermediate and seleninic acid.The subsequent hydrolysis of the former would result in 3. 15,16  In the approach to 3, alternatives to Scheme 2 were examined which proved less successful or less direct.Chen and coworkers reported a three-step procedure from iminostilbene (5H-dibenzo[b,f]azepine) concerning a progressive oxidation of the bridged double bond within an acceptable overall yield of 43%. 17We checked the possibility to obtain 3 from commercially available 10,11-dihydro-5H-dibenzo[b,f]azepine 5. Unfortunately, our initial efforts on this proved to be less straightforward than expected.For instance, use of either SeO 2 or benzeneseleninic anhydride under conditions identical to those used with 3 proved fruitless because of various side products including the aldehyde 6 (formed by tandem benzylic rearrangement/ oxidative decarboxylation), presumably as a consequence of the unprotected NH.9][20] Thus, reaction of 5 with refluxing Ac 2 O (3 h) gave 7 21 which was converted to 3 in 67% overall yield, upon exposure to benzeneseleninic anhydride (BSA) (2 equiv) in chlorobenzene at 120 °C. 22In stark contrast, access to 3 from 5 with SeO 2 in refluxing dioxane was messy and resulted in a complex mixture of compounds from which the isatin derivative 8 23 was isolated, albeit in abysmally low yield.The structure of 8 has been established by singlecrystal X-ray crystallography (Figure 2).At this point, to get ready for Linstead macrocyclization step, 24 we only needed to react the 1,2-diketone 3 with diaminomaleonitrile as 1,4-bis(nucleophile) to form the required odinitrile 4 (Scheme 3), an operation which succeeded readily by refluxing the mixture in AcOH (1 h) (65% yield).
En route to porphyrazine, we were forced to postpone the tetramerization process of 4 owing to its poor solubility, which has hampered the purification.Thus, attempts were made to improve the solubility by incorporation of a long alkyl chain at the azepine N atom.After sampling a variety of derivatives (amides and tertiary N-alkyl amines), the hexadecanoyl group (i.e., 9) appeared to be optimal in terms of yield and lipophilicity.Acylation of 4 was performed in anhydrous pyridine at 100 °C with hexadecanoyl chloride (1.1 equiv) in the presence of a catalytic amount of 4-N,N-dimethylaminopyridine (DMAP) and resulted in the isolation of 9 in 77% yield (Scheme 3).Finally, compound 9 was converted into its porphyrazine magnesium complex 10 via Linstead's Mg template cyclization.Therefore, Mg(n-BuO) 2 was prepared by heating Mg in 1-butanol for 6 h after which the dinitrile 9 was added and the reaction was allowed to run under reflux for 8 h.After silica gel chromatography the magnesium complex 10 was isolated in 32% yield as a dark green powder.With 10 in hand, the conversion into the respective Cu(II) and Co(II) complexes, 11 and 12, was undertaken by well-known transmetallation procedure. 9Treatment of 10 with a variety of Bronsted acids (e.g., TFA, AcOH, H 2 SO 4 ) followed by subsequent remetallation with either Cu(OAc) 2 or Co(OAc) 2 uniformly resulted in formation of a reaction mixture in which only small quantities of the expected 11 and 12, respectively, could be isolated.At this juncture, we turned our attention to Linstead's metal template strategy.Gratifyingly, when 9 was irradiated for 10 min in a microwave oven at 175 °C/350 W in freshly distilled 2-(dimethylamino)ethanol (DMAE), 25 in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and anhydrous Cu(OAc) 2 or Co(OAc) 2 , the respective Cu and Co porphyrazines 11 and 12 were isolated in 67% and 72% yields, respectively.These two complexes show a good solubility in different classic organic solvents (e.g.CH 2 Cl 2 , CHCl 3 , THF, Et 2 O, MeOH).

Catalytic cyclopropanation.
As benchmark catalysis test we opted for transition metal complex-mediated cyclopropanation of alkenes with diazo compounds, an efficient and selective method for access to synthetically and biologically interesting cyclopropanes. 261][32] Some of us reported recently studies in cyclopropanation reactions by using transition metal complexes with nitrogen ligands. 33,34.Yields and diastereomeric ratio were obtained by GC-MS and 1 H NMR using examethylbenzene as internal standard.
Our aim was to test the efficacy of the metal-porphyrazine complexes 11 and 12 in activating the diazo precursor (i.e., EDA), via the metal carbenoid specie, to the cyclopropanation reaction of aromatic, aliphatic and carbonyl conjugated alkenes.In this contest we decide to run the catalytic tests (Table 1) with EDA as limitating reagent (ratio catalyst:EDA:substrate 1:1000:2500), the reactions were considered terminated when no trace of EDA was detected (IR) in the mixture.The reaction was monitored by GC-MS and 1 H NMR, it was related to maximum % of product formation respect to the quantity of diazo compound loaded into the reactor.The stereochemical trans:cis ratios of the cyclopropanes formed were also studied and reported in Table 1.
Good performances were obtained either from Cu(II) complex (11) (yield range 57-96 %) and Co(II) complex ( 12) (yield range 53-95 %), even if the latter revealed a slower turnover frequency, in accordance with the data on phthalocyanine-catalyzed cyclopropanation previously reported by Zhou and coworkers. 30The reactions were run only once and the reported results are therefore unoptimized.The yields and the stereochemical outcomes of the reaction had proved to be substrate dependent, in fact the best result in cyclopropanation products were obtained with styrene (Table 1, entries 1 and 6) and the best trans selectivity emerged from the cyclopropanation of cyclohexenone (Table 1, entries 4 and 9).

Experimental Section
General.All reactions were performed using standard glassware and IKA num heating plates.Reactions utilizing air-and moisture-sensitive reagents were performed in dried glassware under an argon atmosphere.Solvents were used as received without further purifications, unless stated otherwise.All reagents, if not otherwise specified, were used as received and, if necessary, stored under inert gas.Oxcarbazepine was supplied by Trifarma SpA, Ceriano Laghetto, Italy.Silica thin-layer chromatography (TLC) was done on E. Merck plastic or aluminum-backed plates (silica gel 60, F 254 , 0.2 mm).Quantitative analyses of products were performed on a Shimadzu GC-17A gas chromatograph coupled with a QP5000 mass detector.Column chromatography was performed on E. Merck silica gel 60 (70-230 mesh).Microwave heating was performed on a CEM Discover SP instrument with a single-mode microwave cavity providing continuous irradiation at 2.45 GHz and power up to 300W.Melting point determinations were performed by using a Gallenkamp melting point apparatus. 1H NMR (400 MHz) and 13 C NMR (100 MHz) were recorded on Bruker AV400 spectrometer; chemical shifts (δ) are expressed in parts per million (ppm) downfield from TMS (δ 0.0) and relative to the signal of CDCl 3 (i.e., δ 7.26 (singlet) and δ 77.0 (triplet) for 1 H NMR and 13 C NMR, respectively.Coupling constants are given in hertz (Hz).Multiplicities were given as: s (singlet), d (doublet), t (triplet), dt (doublet of triplet), m (multiplet), br (broad).Infrared spectra were recorded in a Shimadzu Prestige 21 FTIR.UV-Vis spectra were obtained on a Thermo Scientific Evolution 220 instrument.MALDI-TOF-MS were recorded on Buker Multiflex LT spectrometer using α-cyano-4-hydroxycinnamic acid (CHCA) as the matrix.Chemical ionization mass spectra ( + ve mode) (CI + -MS) were performed on a Finnigan-MAT TSQ70 with isobutane as the reactant gas.Elemental analyses were performed on a Perkin Elmer Series II CHNS/O Analyzer 2400; the samples were kept under vacuum (0.001 hPa) at 60°C for 3 h before subsequent elemental analysis.
General procedure for catalytic cyclopropanation.A solution of ethyl diazoacetate (EDA) (5.0 mmol in 10 mL of CH 2 Cl 2 ) was added slowly under stirring to a refluxing CH 2 Cl 2 solution (0.2 M) consisting of alkene (12.5 mmol) and 11 or 12 (0.005 mmol).The reaction was halted when no trace of EDA [ν N=N 2110 cm -1 ] was detected in the IR spectrum of the mixture.Yields and diastereomeric ratio were obtained by GC-MS and 1 H NMR using examethylbenzene as internal standard.

Figure 2 .
Figure 2. ORTEP drawing of the molecular structure of compound 8, with thermal ellipsoids drawn at the 30% probability level.For sake of clarity, H atoms have been drawn as circles of arbitrary radius.

Table 1 .
Catalytic cyclopropanation of alkenes mediated by 11 and 12 in CH 2 Cl 2 at 40 °C.Yields and ratio were obtained by GC-MS and 1 H NMR using examethylbenzene as internal standard General procedure for catalytic cyclopropanation.