Porphyrin Complexes with Highly Electronegative Metals – A New Chapter in Nucleophilic Substitution of Hydrogen?

The attempts of direct substitution of hydrogen in porphyrin macrocyclic systems, with carbanions of weak nucleophilicity, are described. Porphyrins, when converted into the corresponding metal chelates, were reacted with the above mentioned carbanions, and the coordinated central metal atom (e.g., Au III , Sn IV ), which reveals considerable electronegativity, played a role of activating group. It could be easily removed from the system after reaction. A number of attempts to substitute hydrogen by carbon nucleophiles led to various products (addition of nucleophile to porphyrin ring, ligands substitution at metal center, etc.). These investigations were successfully finalized for meso-tetraphenylporphyrin–dichlorotin(IV) complex. Further development of this idea may open a new chapter in the functionalization of porphyrins.


Introduction
The selective functionalization of porphyrins is intensively studied in recent years. [1] Earlier, we published several examples of the reactions of weak nucleophiles (carbanions) with these compounds leading to substitution of hydrogen products, however, the macrocycle was activated by the strong electronwithdrawing groups, e.g. NO 2 , which needs to be introduced to the porphyrin system before the reaction. [2] We set up the hypothesis that the same role could be played by the central metal atom when porphyrins are converted into corresponding chelates. This approach has one considerable advantage. The metal can be easily removed (often with almost quantitative yield) from the system after the Hsubstitution reaction. This metal should reveal enough high electronegativity (e). The natural candidates for this purpose are metals of e higher than 2.0; e.g., Au (e = 2.54), Sb (2. Herein, the studies on possibility of direct substitution of hydrogen in the porphyrin ring, activated by the highly electronegative central metal atom when porphyrins are converted into corresponding chelates, are described.

Results and discussion 2.1. Gold complexes.
Electronegativity of gold is relatively high, e = 2.54. Thus, at the beginning, we tested 5,10,15,20-tetraphenylporphyrin-gold(III) chloride (meso-tetraphenylporphyrin-gold(III) chloride) (1) which can be easily prepared according to literature prescriptions. [3] This porphyrinate was purified by column chromatography and the yield was even higher (76%) as compared to that described previously. [3a] However, its reaction with carbanion of ClCH 2 SO 2 Tol, which usually allows nucleophilic substitution of hydrogen in electrophilic aromatic compounds according to vicarious nucleophilic substitution mechanism (VNS), [4] herein, in t-BuOK/THF system, led to a complicated mixture of several products. They were probably an effect of ligand-exchange processes. Some modifications of the procedure and the reaction conditions did not give better results (see Experimental). Next, we have undertaken some attempts to enhance the electrophilicity of the parent system by exchanging Clligand for CN -. However, in this case, CNinstead of exchange chloride Clanion entered immediately the addition to meso-carbon atom, thus giving phlorin moiety 3 (80%; Scheme 1). Similar reactions (with OH -) for gold(III) and for antimony(V) porphyrin complexes were observed by Segawa [5] and Knör. 4; having unsubstituted meso-positions). In this case, the synthesis of the substrate was somewhat troublesome. When the reaction was carried-out in CHCl 3 /AcOH mixture (OEP+KAuCl 4 +AcONa, 19 h, reflux), unexpectedly the only product obtained (with very small yield; 7.7%) was the acetoxy-derivative 5 ( Figure 1). It was probably formed via tandem chlorination / Cl -→AcOexchange processes (source of chlorine: KAuCl 4 ). At higher temperature (130°C) in DMF/CHCl 3 , after shortening the reaction time to 13 h, the chlorinated product 6 was isolated in poor yield (7.8%), along with a large amount of the starting octaethylporphyrin (63%). Its meso-Cl structure was elucidated by HR-MS and 1 H NMR measurements. Finally, in the reaction carried-out in chloroform/methanol mixture (60°C, 20 h) both the above products were identified. In all the experiments the conversion rate was rather low and a lot of substrate OEP was recovered. In some reactions we also observed a pink, very polar spot on TLC which after longer period of time of refluxing disappeared. Analysis of the reaction mixture (after 3 h of heating in CHCl 3 /AcOH) allowed us to identify this new compound(s The attempts to obtain the desired complex 4 directly by the macrocyclization, in which from the beginning of the reaction an excess of KAuCl 4 was added to cause a template effect, also failed. Due to the above problems with the synthesis of the desired gold substrate the investigations were temporarily suspended. It is worth mentioning that Jamin and Iwamoto observed similar difficulties when trying complexation of etioporphyrin with KAuCl 4 . [7] 2.2. Tin complexes: meso-tetraphenylporphyrindichlorotin(IV) and octaethylporphyrin-dichlorotin(IV) systems. We tried to verify the above concept using another relatively easily available porphyrin complexes (dichlorotin derivatives; e Sn = 1.96): octaethylporphyrin-dichlorotin(IV) (8) and mesotetraphenylporphyrin-dichlorotin(IV) (9). They were synthesized on the basis of two literature reports. [8,9] Nevertheless, some modifications were introduced (sulfolane as a solvent, 170-200°C, ca 1 h, 68-100%; see Experimental). Confirmation of their structures was not a trivial problem. In MS spectrum of product 8 (ESI(+), in MeOH) instead of molecular and pseudomolecular ions M + or (M+H) + , we observed another ions, probably formed during the measurements (m/z = 665 and m/z = 679; see Figure 2). Their formation can be explained easily by the ligand-exchange and ligand-losing processes. This is rather characteristic in the chemistry of such labile chelates and we observed it earlier. [10] In MS-FD spectrum the only observed peaks also were originating from the fragmentation ions and multicharged ions. Thus, the structure cannot be confirmed definitively on the basis of these data.
Some verifications came from the 1 H NMR studies. The spectrum was in agreement with the structure.  The complex obtained 8 was reacted, in the presence of base, with -CH(Br)SO 2 Tol carbanion, and we expected nucleophilic substitution of hydrogen [4] in meso-position. It could be an exceptional example of direct substitution of hydrogen with a carbanion nucleophile in porphyrinoid system. However, the reaction failed to afford the desired product. When analyzing the post-reaction mixture by MS method (APPI-photospray(+) in AcOEt), we observed only the fragmentation ions as a result of ligand-exchange and ligand-losing process in the substrate (m/z = 683, OEPSnCl + , and m/z = 707, OEPSn(OAc) + ). There were no ions originated from the substitution of hydrogen product. The outcome of the reaction with carbanion of para-chlorophenoxyacetonitrile ( -CH(CN)OC 6 H 4 Cl p ), which is stronger nucleophile as compared to -CH(Br)SO 2 Tol, was similar (degradation of the reagents occurred).
Finally, in the reaction of 5,10,15,20-tetraphenylporphyrin-dichlorotin(IV) (9) with halomethyl paratolyl sulphone carbanion ( -CH(Cl)SO 2 Tol; t-BuOK/ DMSO, r.t.) we observed the formation of new product (TLC monitoring). One can suppose it was a VNS product, formed due to substitution of hydrogen at the -position. Initially, we couldn't confirm its molecular formula directly by MS method; however, we found in the spectrum (ESI(+), in CH 3 OH) an ion peak m/z = 927 originating from the product 10 (see Scheme 2 and Experimental). This ion is a result of ligand-exchange/ligand-losing process which is possible during the MS measurement. We supported this hypothesis when the sample of product was dissolved and measured in ethanol. The formation of the analogous ion m/z = 941 was observed. It seems these ions are rather strong evidence for the structure of the desired product because such an easy spontaneous conversion of dichlorotin porphyrin complexes into dialkoxy-and diphenoxy-tin moieties was reported earlier by Arnold. [11] Similar observations were made in our previous studies. [10] Nevertheless, finally we also detected the molecular ion of the examined compound by MS-FD method (m/z = 966; C 52 H 36 N 4 O 2 SCl 2 Sn). Independently, we confirmed its structure by 1 H NMR. All the diagnostic signals were found in the spectrum: 2.21 (s, 3H, CH 3 -Tol), 4.62 (s, 2H, CH 2 ), 6.89/7.18 (2×d, 4 H, J = 8.2 Hz, H-Tol), and 9.05-9.24 (m, 7H β ).

Scheme 2
Decomplexation of the product obtained with lithium in ethylenediamine (reflux, 3 h) [12] leads to the free base porphyrin moiety substituted with CH 2 SO 2 Tol group at the β-position (11, m/z = 782, M + , C 52 H 38 N 4 O 2 S 1 ); and this is one more proof for the structure 10.

Conclusions
We reported herein the attempts of direct nucleophilic substitution of hydrogen in porphyrin systems, activated by the coordinated central metal atom (of increased electronegativity), when porphyrins are converted into the corresponding chelates. Complexes of Au(III) and Sn(IV), and their reactions with carbanions of weak nucleophilicity, were examined. These investigations were successfully finalized for meso-tetraphenylporphyrin-dichlorotin(IV) complex. The above mentioned metal atom played a role of activating moiety, as well as a labile protective group  for the inner NH-protons, and after reaction can be easily removed from the system, if needed. We believe that our concept involving complexation/decomplexation procedure and new type of activation for nucleophilic attack will receive future attention in the area of porphyrin skeleton functionalizations. Molecular formulas of new compounds were confirmed by elemental analysis, HR-MS (ESI and FD), and by comparing the isotope molecular patterns (theoretical and experimental).

Reaction of 1 with carbanion of ClCH 2 SO 2 Tol in t-BuOK/THF system
Procedure A: In a round-bottomed flask t-BuOK (50 mg, 0.45 mmol) was stirred in anhydrous THF (7 mL; under argon) at room temperature for ca 5 min. To this solution a mixture of porphyrin-gold(III) chloride (1; 50 mg, 0.059 mmol) and chloromethyl para-tolyl sulphone (98 mg, 0.479 mmol) in THF (5 mL) was added dropwise via syringe. After 30 min of stirring, the mixture was poured into 3% HCl containing ice (10 mL) and extracted with CHCl 3 (315 mL). The combined organic layers were washed with water (350 mL) and dried with anhydrous MgSO 4 . Several products were observed (TLC monitoring). After evaporation of the solvent, the column chromatography was performed using gradient mixture as eluent (from CHCl 3 to CHCl 3 /MeOH, 20:1). None of the defined products were isolated.
Procedure B: In a round-bottomed flask chloromethyl para-tolyl sulphone (87 mg, 0.425 mmol) was stirred in anhydrous THF (3 mL; under argon) at room temperature for ca 2 min. To this solution t-BuOK (50 mg, 0.45 mmol) in THF (3 mL) was added and the stirring was continued for the next 5 min. Then, porphyrin-gold(III) chloride (1; 45 mg, 0.053 mmol) in THF (5 mL) was added dropwise via syringe. After additional 30 min of stirring, the reaction mixture was concentrated in vacuo to 13 of the initial volume, 20 mL of CHCl 3 was added, and washed with water (330 mL). The aqueous layers were extracted with CHCl 3 (330 mL) and all the combined organic layers were dried with anhydrous MgSO 4 . After evaporating the solvent, the residue was recrystallized from CHCl 3 and various CCl 4 /CH 2 Cl 2 mixtures. Neither the VNS product nor the addition of carbanion product, or moiety with exchanged Cl ligand for carbon ligand, has been isolated. Always, the mixture of several products was crystallized.

Attempts of complexation of octaethylporphyrin with gold(III) cation
Procedure A: In a round-bottomed flask, equipped with a reflux condenser, KAuCl 4 (35 mg, 0.093 mmol) and AcONa (46 mg, 0.56 mmol) were heated to reflux (under argon) in acetic acid (6 mL) for ca 15 min. Then, octaethylporphyrin (40 mg, 0.075 mmol) in CHCl 3 (10 mL) was added and the reaction was continued at reflux for the next 19 h. To this mixture (cooled to room temperature), CHCl 3 (10 mL) was added and it was washed with water (330 mL). The organic layer was dried with anhydrous MgSO 4 . After evaporating the solvent, the residue was subjected to column chromatography (eluent: CHCl 3 /n-hexane, 2:1); 3.4 mg of product 5 was isolated (7.7%), along with the recovery of 5.4 mg of octaethylporphyrin (13.5%).
When the reaction was carried out according to Procedure A and the time was shortened to 3 h (heating at reflux) a mixture of products 4 and 7 was isolated (by preparative TLC, eluent: CHCl 3 , five times developed); yield below 5%.

Decomplexation of VNS product 10
A crude sample of product 10 (6.4 mg) in ethylenediamine (4 mL) was preheated to reflux (under argon) in a round-bottomed flask equipped with a reflux condenser. After 1 h, 34 mg of lithium was added (as thin wires) and the reaction was continued at reflux for the next 3 h. Then, the post-reaction mixture was poured into water (30 mL) and extracted with CHCl 3 (320 mL). The combined organic layers were washed with water (330 mL) and dried with anhydrous MgSO 4 . After evaporation of the solvent, the residue was analyzed by MS method. The molecular ion originating from the desired product 11 was observed. MS (APPI-photospray(+), AcOEt/CH 2 Cl 2 );