Isomers of β,β-Dinitro-5,10,15,20-tetraphenylporphyrin Derivatives: Valuable Starting Materials for Further Transformations

The synthesis, chromatographic isolation, and structure elucidation of β,β-substituted isomers of dinitro-5,10,15,20-tetraphenylporphyrin complexes are described. meso-Tetraphenyl-porphyrin chelates (CuII, NiII, CoII) upon reaction wit e.g., nitric acid (yellow HNO3, d = 1.52, diluted to 25–50%) in CHCl3 formed a mixture of nitro-derivatives with combined yields of ca 80%. This nitration (under optimized conditions: 25–30% HNO3, 30–40 min, r.t.) can be carried out selectively to give mainly β,β-dinitro-compounds in yields of up to 73%. From the above mixtures of five possible regioisomers that can be formed, usually two or three of them were isolated, for which the structures were assigned on the basis of 1H NMR spectra including COSY and NOESY measure-ments. These types of products are attractive starting materials for synthesis of potential anticancer PDT agents with unique structures, being practically unavailable by any other alternative method.


Introduction
A number of porphyrin derivatives are of significant importance due to their potential use in many fields of chemistry, medicine, pharmacology, new materials, etc. [1]. The desired precursors for the synthesis of these compounds can be isolated from naturally occurring substances (e.g., chlorophyll, heme) or prepared via transformations of simple synthetic moieties (e.g., meso-tetraphenylporphyrin, TPP). Thus, nowadays the selective functionalization of porphyrins is being studied intensively.
Recently, successful nitration of TPP zinc complex 1a (Scheme 1), leading mainly to [2,7-dinitro-5,10,15,20-tetra-phenylpor-phyrinato]zinc(II) as the major product with a satisfactory yield ca 40%, has been demonstrated [2]. The respective dinitroporphyrins of this type are valuable starting materials for our ongoing projects; e.g., (a) an exhaustive substitution at all β-positions in two neighbouring pyrrole rings (which allows the synthesis of highly substituted porphyrins) [3] or (b) controlled cycloaddition reactions leading to isobacteriochlorins [4], etc. They could also serve as interesting models for electrochemical studies [5].
Thus, we have attempted to elaborate the dinitration reaction for some additional TPP chelates (Cu II , Ni II , Co II ), in search for other β,β-dinitro-isomers. Such a nitration of a copper complex has been already reported. Dahal et al. observed mixtures of two or three dinitro-compounds in various reactions [6]. However, the total yields were relatively low (ca 20%) and no information was provided about the yields of individual isolated isomers. The resulting NO 2 -disubstituted porphyrins could be very attractive intermediates for synthesis of the target moieties of higher complexity, e.g., potential anticancer agents [7][8][9][10][11].

Results and Discussion
Three TPP complexes (Cu II , Ni II , Co II ) were used for our investigations. It is worth mentioning that mononitration (with good yields) of all these systems under electrophilic conditions (HNO 3 in CHCl 3 ) was described earlier [12,13]. Additionally zinc complex gave satisfactory results [2]. Unexpectedly, we discovered lately that some free base porphyrins can react according to this scheme, as well [14]. Finally, the β-monosubstituted products, contrary to polysubstituted ones, are readily available by various other methods [15][16][17][18][19][20].

Nitration of Cobalt(II) and Copper(II) Complexes
Initially, in the reactions of cobalt complex 1b (under various conditions: 1-5% HNO 3 , 5 min, 2 h, 0-20 • C, in CHCl 3 , under argon) the formation of dinitro-compounds with yields in the 30-64% range was observed. However, the products formed included partial decomplexation derivatives, yielding practically inseparable mixtures. Thus, these experiments were temporarily suspended.
We found in our laboratory that copper complexes were the most convenient substrates for these transformations (the total nitration yield is high and the chromatographic isolation of the products is relatively satisfactory). Thus, the selected TPP copper(II) porphyrinate 1c was subjected to the reaction with nitric acid in CHCl 3 , applying the method based on our previously reported procedures, elaborated for mononitration [12,13]. Initially, this electrophilic reaction with the use HNO 3 of higher concentration (35-55%, freshly prepared from fuming yellow nitric acid) usually gave a mixture of dinitro-compounds, in 30-43% yield. These mixtures contained three (or more) β,β-dinitro-isomers: 2,7-(one spot on TLC) and 2,8-/3,7-(another spot on TLC) in a ca 1:1 ratio. We found that from these mixtures the two major dinitro-isomers-2,7-dinitro-(3ca) and 3,10,15, (3cb)-could be isolated in considerable amounts by precise column chromatography followed by preparative TLC (see Scheme 1 and Experimental). In the remaining chromatographic fractions some amounts of these isomers were still present. It should be also mentioned that in these reactions formation of the mononitro-product 2c was observed, with variable yields (ca 10%).
In the case of this complex (1c), partial optimization of the reaction conditions allowed us to obtain the best yield of β,β-dinitro-isomers when using 25% HNO 3 . Thus, treatment of 1c with the above nitric acid in CHCl 3 at room temperature (reaction time-30-40 min) gave moderate amounts of monosubstituted product (2c, 28%) and a mixture of the desired dinitrated derivatives (56% combined, confirmed by MS measurements) (Scheme 1). The above fractions were isolated by column chromatography, however separation of the major dinitro-isomers required precise TLC conditions (CHCl 3 /n-hexane-1:1). In this reaction five different dinitro-isomers may form (Scheme 1). The investigations reported herein allowed us to identify the three main products.
The corresponding structures of the products 3ca-3cc couldn't be assigned directly on the basis of NMR due to the paramagnetic properties of the copper cation, thus some of them were identified by comparison with authentic compounds obtained earlier in our laboratory as by-products of other reactions, or selected compounds that were demetallated. Removal of the paramagnetic copper cation allowed us to record 1 H NMR spectra (see Supplementary Materials) and fully characterize the products as the free base porphyrins 4a-c. This problem is discussed in details in Section 2.4. The demetallation reactions were carried out in a mixture of acids (CF 3 CO 2 H/H 2 SO 4 ) at room temperature (15-30 min), leading to the metal-free derivatives in high yields (ca 90%, see Experimental).

Nitration of Nickel(II) Complex
In the case of nitration of the nickel complex 1d (20-30% HNO 3 , 10-40 min) we also observed the formation of considerable amounts of a mixture of dinitro-derivatives (yield 18-64%), always accompanied by mono-substituted product. Partial optimization of the reaction conditions (a large excess of 30% HNO 3 /CHCl 3 , r.t., 30 min, under argon) allowed us to increase the dinitration product yield of up to 73% (mononitro-derivative: 6.5%, known compound 2d [12]). It is worth mentioning that the yield was calculated for two steps (complexation of TPP and nitration) as the nickel complex 1d is only moderately soluble; its purification by column chromatography is also somewhat troublesome, thus decreasing the isolated yield. The crude (meso-tetra-phenylpor-phyrinato)nickel(II) (1d) when directly subjected to the nitration reaction gave a mixture of several products from which only the 2-nitro-porphyrin 2d was isolated easily by column chromatography, while a mixture of dinitro-compounds had to be separated additionally by preparative TLC (CHCl 3 /n-hexane-2:1, developed four times). From this mixture of dinitro-moieties two major isomers were isolated: (2,7-dinitro-meso-tetra-phenylpor-phyrinato)nickel(II) (3da, 22%) and (3,7-dinitro-meso-tetra-phenylpor-phyrinato)-nickel(II) (3db, 22%). 1 H NMR, MS, HR-MS and UV-vis analyses confirmed their structures. In the 1 H NMR of the 2,7-dinitro-compound 3da there are two diagnostic downfield-shifted singlets of H β -protons neighbouring NO 2 groups (8.89 ppm and 8.81 ppm) and two AB systems (in the region 8.56-8.69 ppm; from the remaining four β-protons). This is the only unsymmetrical β,β-dinitro-substituted product. Thus, its identification was very simple, contrary to assignment of the spectrum of compound 3db. The latter case was not a trivial problem, and is discussed below (Section 2.4). excess of 30% HNO3/CHCl3, r.t., 30 min, under argon) allowed us to increase the dinitration product yield of up to 73% (mononitro-derivative: 6.5%, known compound 2d [12]). It is worth mentioning that the yield was calculated for two steps (complexation of TPP and nitration) as the nickel complex 1d is only moderately soluble; its purification by column chromatography is also somewhat trouble-   2.3. Nitration of 5,10,15,porphyrin-Copper(II) Complex (5) The above substitution orientation and the yields concerning β,β-dinitration of TPP complexes should also be observed for other porphyrin derivatives. Indeed, this was the case. We selected [5,10,15,20-tetrakis(3-methylphenyl)porphyrinato]copper(II) (5) for this experiment and its treatment with the above nitric acid in CHCl 3 gave similar results. Interestingly, all the possible dinitro-isomers 6a-10a were successfully isolated (see Figure 1).  (5) The above substitution orientation and the yields concerning β,β-dinitration of TPP complexes should also be observed for other porphyrin derivatives. Indeed, this was the case. We selected [5,10,15,20-tetrakis(3-methylphenyl)porphyrinato]copper(II) (5) for this experiment and its treatment with the above nitric acid in CHCl3 gave similar results. Interestingly, all the possible dinitro-isomers 6a-10a were successfully isolated (see Figure 1).

1 H NMR Spectra and Structure Elucidation of the Isomers
The correct assignment of the corresponding structures to the obtained products was not trivial. It was mainly accomplished on the basis of 1 H NMR, however as the copper complexes used herein are paramagnetic their spectra are not recordable. On the other hand, Cu-chelates were the most convenient models for the studies (separation of the products was relatively satisfactory, yields were good, and three isomers were isolated). Thus, we solved this problem indirectly. From the mixture of dinitrated porphyrins three of them were isolated as pure individual compounds by preparative TLC. Their demetallation using H2SO4/CF3CO2H gave copper-free isomers, which were examined spectroscopically in details.
In the case of nickel products the isomers were analyzed as chelates. The identification of 3da (2,7-dinitro-) was simple (see above) contrary to the assignment of the structure of the next isomer. In the spectrum of the latter we found a singlet δ = 8.88 ppm originating from two β-protons, and an AB-like system (4H) at 8.63 ppm and 8.66 ppm (with J = 5.2 Hz). Unfortunately, these data are in agreement with the three structures (3db, 3dc, 3dd; Figure 2). Structure 3de was not taken under consideration because as a highly symmetrical one it should give in this region only three singlets (3 × 2H). On the basis of two-dimensional COSY and NOESY measurements structure of 3db was assigned. There is a diagnostic correlation between the signals at 8.06-8.02 ppm (2H) and the triplet at 7.58 ppm (J = 7.6 Hz, 2H) (see Figures 3 and 4). These signals are outside the two multi-H multiplets; thus, the corresponding protons must be under the strong influence of NO2 groups. We therefore ascribed them to the ortho-and meta-protons of one phenyl ring situated between two nitro groups. This is in agreement with the proposed 3,7-dinitro-structure 3db. Moreover, in this analysis step the isomer 3dd can be definitively excluded from the list of potential products because the integration ratio of all the ortho-protons of the meso-phenyl rings should be equal to 4:4 in this case.

1 H NMR Spectra and Structure Elucidation of the Isomers
The correct assignment of the corresponding structures to the obtained products was not trivial. It was mainly accomplished on the basis of 1 H NMR, however as the copper complexes used herein are paramagnetic their spectra are not recordable. On the other hand, Cu-chelates were the most convenient models for the studies (separation of the products was relatively satisfactory, yields were good, and three isomers were isolated). Thus, we solved this problem indirectly. From the mixture of dinitrated porphyrins three of them were isolated as pure individual compounds by preparative TLC. Their demetallation using H 2 SO 4 /CF 3 CO 2 H gave copper-free isomers, which were examined spectroscopically in details.
In the case of nickel products the isomers were analyzed as chelates. The identification of 3da (2,7-dinitro-) was simple (see above) contrary to the assignment of the structure of the next isomer. In the spectrum of the latter we found a singlet δ = 8.88 ppm originating from two β-protons, and an AB-like system (4H) at 8.63 ppm and 8.66 ppm (with J = 5.2 Hz). Unfortunately, these data are in agreement with the three structures (3db, 3dc, 3dd; Figure 2). Structure 3de was not taken under consideration because as a highly symmetrical one it should give in this region only three singlets (3 × 2H). On the basis of two-dimensional COSY and NOESY measurements structure of 3db was assigned. There is a diagnostic correlation between the signals at 8.06-8.02 ppm (2H) and the triplet at 7.58 ppm (J = 7.6 Hz, 2H) (see Figures 3 and 4). These signals are outside the two multi-H multiplets; thus, the corresponding protons must be under the strong influence of NO 2 groups. We therefore ascribed them to the orthoand meta-protons of one phenyl ring situated between two nitro groups. This is in agreement with the proposed 3,7-dinitro-structure 3db. Moreover, in this analysis step the isomer 3dd can be definitively excluded from the list of potential products because the integration ratio of all the ortho-protons of the meso-phenyl rings should be equal to 4:4 in this case. 2.3. Nitration of 5,10,15,porphyrin-Copper(II) Complex (5) The above substitution orientation and the yields concerning β,β-dinitration of TPP complexes should also be observed for other porphyrin derivatives. Indeed, this was the case. We selected [5,10,15,20-tetrakis(3-methylphenyl)porphyrinato]copper(II) (5) for this experiment and its treatment with the above nitric acid in CHCl3 gave similar results. Interestingly, all the possible dinitro-isomers 6a-10a were successfully isolated (see Figure 1).

1 H NMR Spectra and Structure Elucidation of the Isomers
The correct assignment of the corresponding structures to the obtained products was not trivial. It was mainly accomplished on the basis of 1 H NMR, however as the copper complexes used herein are paramagnetic their spectra are not recordable. On the other hand, Cu-chelates were the most convenient models for the studies (separation of the products was relatively satisfactory, yields were good, and three isomers were isolated). Thus, we solved this problem indirectly. From the mixture of dinitrated porphyrins three of them were isolated as pure individual compounds by preparative TLC. Their demetallation using H2SO4/CF3CO2H gave copper-free isomers, which were examined spectroscopically in details.
In the case of nickel products the isomers were analyzed as chelates. The identification of 3da (2,7-dinitro-) was simple (see above) contrary to the assignment of the structure of the next isomer. In the spectrum of the latter we found a singlet δ = 8.88 ppm originating from two β-protons, and an AB-like system (4H) at 8.63 ppm and 8.66 ppm (with J = 5.2 Hz). Unfortunately, these data are in agreement with the three structures (3db, 3dc, 3dd; Figure 2). Structure 3de was not taken under consideration because as a highly symmetrical one it should give in this region only three singlets (3 × 2H). On the basis of two-dimensional COSY and NOESY measurements structure of 3db was assigned. There is a diagnostic correlation between the signals at 8.06-8.02 ppm (2H) and the triplet at 7.58 ppm (J = 7.6 Hz, 2H) (see Figures 3 and 4). These signals are outside the two multi-H multiplets; thus, the corresponding protons must be under the strong influence of NO2 groups. We therefore ascribed them to the ortho-and meta-protons of one phenyl ring situated between two nitro groups. This is in agreement with the proposed 3,7-dinitro-structure 3db. Moreover, in this analysis step the isomer 3dd can be definitively excluded from the list of potential products because the integration ratio of all the ortho-protons of the meso-phenyl rings should be equal to 4:4 in this case. Theoretically, the 2,8-dinitro-isomer 3dc could also give such a correlation pattern. Thus, the structure 3db was definitively confirmed by the NOESY technique. In the spectrum several correlations between protons due to their spatial proximity (Figure 4; black arches) were observed. The most diagnostic one is the correlation of the ortho-phenyl protons, δ = ca 8.00 ppm, with the two protons of the AB-like system ('the red protons', at 8.66 ppm) from one side, and the two β-protons neighbouring the NO 2 groups ('the green protons', singlet at δ = 8.88 ppm), from the other side. Additionally, the β-protons of the remaining part of the AB-like system (2H, 8.63 ppm, 'the violet protons') correlate with the ortho-protons of the last phenyl ring (δ = ca 7.98 ppm; in the 8.06-7.96 ppm multiplet). Finally, no correlation between the ortho-protons at 8.06-8.02 ppm with any β-protons was observed. This is an additional unambiguous evidence for the structure 3db. Moreover, demetallation of 3db and comparison of the 1 H NMR spectrum of the free base porphyrin obtained with the data observed for 4b (in the light of the information given below), allowed us to find the same 3,7-disubstitution pattern (the products were the same). By this way, we proved that the analyzed product, obtained in the nitration of TPP(Ni) (1d), has structure 3db.
Molecules 2018, 23, x FOR PEER REVIEW 5 of 11 Theoretically, the 2,8-dinitro-isomer 3dc could also give such a correlation pattern. Thus, the structure 3db was definitively confirmed by the NOESY technique. In the spectrum several correlations between protons due to their spatial proximity (Figure 4; black arches) were observed. The most diagnostic one is the correlation of the ortho-phenyl protons, δ = ca 8.00 ppm, with the two protons of the AB-like system ('the red protons', at 8.66 ppm) from one side, and the two β-protons neighbouring the NO2 groups ('the green protons', singlet at δ = 8.88 ppm), from the other side. Additionally, the β-protons of the remaining part of the AB-like system (2H, 8.63 ppm, 'the violet protons') correlate with the ortho-protons of the last phenyl ring (δ = ca 7.98 ppm; in the 8.06-7.96 ppm multiplet). Finally, no correlation between the ortho-protons at 8.06-8.02 ppm with any β-protons was observed. This is an additional unambiguous evidence for the structure 3db. Moreover, demetallation of 3db and comparison of the 1 H NMR spectrum of the free base porphyrin obtained with the data observed for 4b (in the light of the information given below), allowed us to find the same 3,7disubstitution pattern (the products were the same). By this way, we proved that the analyzed product, obtained in the nitration of TPP(Ni) (1d), has structure 3db.  A similar analysis using 2D NMR techniques for a copper-free porphyrin 4b, obtained from 3cb, was performed earlier by Wyrębek [21]. After two-step transformation he isolated this product in 6%   Theoretically, the 2,8-dinitro-isomer 3dc could also give such a correlation pattern. Thus, the structure 3db was definitively confirmed by the NOESY technique. In the spectrum several correlations between protons due to their spatial proximity (Figure 4; black arches) were observed. The most diagnostic one is the correlation of the ortho-phenyl protons, δ = ca 8.00 ppm, with the two protons of the AB-like system ('the red protons', at 8.66 ppm) from one side, and the two β-protons neighbouring the NO2 groups ('the green protons', singlet at δ = 8.88 ppm), from the other side. Additionally, the β-protons of the remaining part of the AB-like system (2H, 8.63 ppm, 'the violet protons') correlate with the ortho-protons of the last phenyl ring (δ = ca 7.98 ppm; in the 8.06-7.96 ppm multiplet). Finally, no correlation between the ortho-protons at 8.06-8.02 ppm with any β-protons was observed. This is an additional unambiguous evidence for the structure 3db. Moreover, demetallation of 3db and comparison of the 1 H NMR spectrum of the free base porphyrin obtained with the data observed for 4b (in the light of the information given below), allowed us to find the same 3,7disubstitution pattern (the products were the same). By this way, we proved that the analyzed product, obtained in the nitration of TPP(Ni) (1d), has structure 3db.  A similar analysis using 2D NMR techniques for a copper-free porphyrin 4b, obtained from 3cb, was performed earlier by Wyrębek [21]. After two-step transformation he isolated this product in 6%  A similar analysis using 2D NMR techniques for a copper-free porphyrin 4b, obtained from 3cb, was performed earlier by Wyrębek [21]. After two-step transformation he isolated this product in 6% yield and its spectroscopic data were in good agreement with that described herein (see Experimental). In the case of the isomer 3ca (obtained in our experiments), the identification based on the 1 H NMR analysis of the demetallated porphyrin 4a was similar to 3da. In this series of compounds, again it was the only unsymmetrical β,β-dinitro-substituted product.
We also isolated the partially contaminated 2,10,15,. In 1 H NMR spectrum of its free base form 4c a very similar signal pattern was found. The key difference is the appearance of a signal for the four β-protons (H-12, H-13, H-17, H-18) as a singlet. Thus, also the structure of 3cc was indirectly proved (via demetallation).  Taiwan), UV-3600 (Shimadzu, Chiyoda-ku, Tokyo, Japan), and V-730 (Jasco, Hachioji, Tokyo, Japan) spectrophotometers. TLC analysis was performed on aluminium foil plates pre-coated with silica gel (60 F-254, Merck AG, Darmstadt, Germany). All the products were isolated by column chromatography (silica gel, 230-400 mesh; Merck AG). Some dinitro-isomers were additionally rechromatographed using preparative TLC plates (silica gel, 60 F-254, 2 mm and 0.5 mm; Merck AG). Molecular formulas of new compounds were confirmed by HR-MS (ESI, EI, and FD). Starting porphyrinates were obtained according to known procedures described in the previous literature [12,13]. Some dinitro-derivatives were also reported. Their 1 H NMR, UV-vis, and MS spectra are in agreement with the spectra described herein. These spectroscopic data are given below for more detailed and accurate characterization of the products.

Nitration of Copper(II) Complex of
In some experiments traces of trinitro-isomers were detected (ca 1%, identified by MS). The structures of the above compounds could not be elucidated due to the paramagnetic copper cation inside the core ring, thus their 1 H NMR spectra are not readable. Additionally, the compounds 8a,9a were isolated as an inseparable mixture, so to determine the structures, all the compounds obtained and the mixture 8a/9a were demetallated [12,13,22] to give free base porphyrins 6b-10b via column chromatography (eluent: CHCl 3 /n-hexane). The spectral data of the products 6a-10a are given below. Also the porphyrins 6b-10b were fully characterized. The yield of 8a and 9a was calculated on the basis of the ratio of the separated 8b and 9b.