An Alternate Route of Transforming meso-Tetraarylporphyrins to

Jul 30, 2018 - ... meso-Tetraarylporphyrins to Porpholactams, and their Conversion to Amine-functionalized Imidazoloporphyrins. Michael P Luciano , Jo...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/joc

Cite This: J. Org. Chem. 2018, 83, 9619−9630

An Alternate Route of Transforming meso-Tetraarylporphyrins to Porpholactams, and Their Conversion to Amine-Functionalized Imidazoloporphyrins Michael P. Luciano,† Joshua Akhigbe,† Jiaming Ding,† Damaris Thuita,† Randy Hamchand,† Matthias Zeller,‡ and Christian Brückner*,† †

Department of Chemistry, University of Connecticut, Storrs, Connecticut 06269-3060, United States Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907-2084, United States

J. Org. Chem. 2018.83:9619-9630. Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 09/08/18. For personal use only.

‡

S Supporting Information *

ABSTRACT: A novel and efficient synthetic pathway toward known meso-tetraphenylporpholactams, also applicable to the synthesis of hitherto unknown and inaccessible meso-C6F5-substituted porpholactam, is detailed (dioxochlorin → dioxochlorin urea adduct → porpholactam). meso-Tetraphenylporpholactam was converted to an imidazoloporphyrin-α-triflate derivative that was demonstrated to be of utility for the generation of functionalized imidazoloporphyrins with a substituted amine adjacent to the outside N atom of the imidazole moiety (using pyridine, Et2NH, diethyliminodiacetic acetate, dipicolylamine (DPA), and cyclen). The DPA- and iminodiacetate-imidazoloporphyrin conjugates were structurally characterized. The chemosensing potential of the metal chelate-imidazoloporphyrin conjugates was evaluated, though their constrained metric parameters led to muted chemosensing responses to various divalent metal ions. The accessibility of the meso-arylporpholactams and the meso-tetraphenylimidazoloporphyrin triflate enables the continued exploration of porphyrin-like pyrrole-modified porphyrins that incorporate a nitrogen atom in place of a β-carbon atom in their macrocycles.

■

INTRODUCTION

Imidazoloporphyrin 1, reported by our laboratory,1 is a porphyrin analogue that combines the N4-central cavity of a porphyrin, such as tetraphenylporphyrin 2,2 with the peripheral β-N of an N-confused porphyrin,3 such as 3. The basicity of the peripheral nitrogen of 3 was demonstrated through its halochromic response under acidic conditions.1b Thus, we predicted imidazoloporphyrin 1, like other imidazoles,4 to be a competent ligand for transition metals. The established seven-step synthesis of 1 from porphyrin 2 is overall low-yielding (Scheme 1):1b Porpholactone 6, available along a number of oxidation routes from porphyrin 2 (including our preferred route via chlorin diol 5),5 is reacted with hydrazine; the resulting N-amino-lactam was reductively cleaved to provide porpholactam 9; subsequently, the iminol form of 9 was chlorinated to form chloroimidazoloporphyrin 10; it underwent hydrodehalogenation into imidazoloporphyrin 1. Shorter alternative syntheses exist, but are either even less efficient or limited to the synthesis of the Ni(II)imidazoloporphyrin complex.1a β-Azacorrins and β-aza-chlorins that are structurally related to 1 as they also contain an © 2018 American Chemical Society

Received: March 28, 2018 Published: July 30, 2018 9619

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

optical spectra of their chromophores in characteristic ways11 and is the basis for their use as metal sensors.12 The binding of a metal ion to the porphyrinic central cavity is, however, not highly selective, comparably slow, and practically irreversible for most metal ions,11,13 factors that limit their utility as chemosensors. To overcome these shortcomings, researchers have appended porphyrinoids with metal ion coordination motifs at sites remote from the chromophore,10f,14 but metal binding results in weak optical responses as these binding sites lack conjugation with the porphyrin macrocycle. These circumstances inhibit the sensitivity of these potential sensors. Unlike porphyrins such as 2, N-confused porphyrin 3 (and related N-confused porphyrinoids) can, due to the presence of the β-N atom, coordinate metals to its center as well as its periphery.3a,12,15 Since the latter coordination mode directly involves the chromophore, coordination results in strong and diagnostic changes to its optical properties. Alas, the external monodentate coordination mode is generally weaker and less selective than comparable chelation coordination modes.4 A structural compromise between the coordination modes of 2 and 3 are, for example, the Pd-porphyrin chelate 4 (and related compounds), reported by the Callot, Ruppert, Jeandon, and Richeter groups.16 Here, a metal chelating motif involving a βsubstituted tetraarylporphyrin derivative was established, achieving the coordination of the metal to heteroatoms that are, if not part of the chromophore, at least in π-conjugation with it. Similar approaches have been pursued for the realization of anion sensors.17 We report here the full details of the practical formation of porpholactams along the dione 7 → diol 8 → lactam 9 route mentioned as an incidental result in a preliminary report,8 and the preparation of its meso-pentafluorophenyl derivative inaccessible along the established route. Furthermore, we will detail the conversion of porpholactam 9 to an imidazoloporphyrin-α-triflate derivative susceptible to nucleophilic exchange with a number of amines, generating α-pyridinium- and αamino-imidazoloporphyrin derivatives. Some of these derivatives were designed to chelate transition metals while possibly involving the β-N atom of the imidazole moiety. In so doing, we are providing the synthetic methodology for the expansion of the utility of pyrrole-modified porphyrins.18 We also present some preliminary metal binding data that point toward the scope and limits of this approach toward the realization of metal sensors.

Scheme 1. Known Syntheses of Imidazoloporphyrin 1 and Porpholactam 9

imidazole building block within a porphyrinoid framework were prepared by total synthesis from imidazole derivatives.6 On the basis of the known reactivity of α-chloroimines,7 the chloroimidazole functionality of 10 is a potential synthetic handle toward the generation of a multitude of imidazoloporphyrin derivatives. When this work was first reported,1b access to porpholactam 9 was limited. However, in 2016 we discovered by fortuity a shorter and more efficient pathway toward porpholactam 9 (Scheme 1):8 meso-Tetraphenyl-2,3dioxochlorin 7, readily available along a number of complementary pathways from porphyrin 1,5a,9 formed chlorin diol adduct 8, in high yield, upon reaction with urea. Following oxidative diol cleavage, chlorin 8 fragmented along an unknown pathway to form porpholactam 9. This overall more economic pathway toward porpholactam 9 enabled the study of its conversion to (functionalized) imidazoloporphyrin derivatives presented here. Given the interest in the realization of functionalized porphyrin-based chemosensing systems,10 we tested the derivatization of imidazoloporphyrins using metal chelating moieties. meso-Tetraarylporphyrins, such as 2, coordinate metal ions only in their central cavities. This complexation changes the

■

RESULTS AND DISCUSSION An Efficient Synthesis of Porpholactam 9. The regular ketone reactivity of meso-tetraarylporphyrin diones, such as 7, was previously demonstrated,19 including their reactions with diamines to generate diimines.20 However, we found that the reaction of nonpolar, yellow-brown 7 with urea did not generate a diimine, but rather the polar, magenta-colored, nondehydrated precursor molecule that contained an imidazolidinone moiety annulated to the pyrroline β,β′-positions (8). Diagnostic for the formation of this dihydroxychlorin structure are the preservation of the 2-fold symmetry of the starting material combined with the presence of OH and NH signals in its 1H NMR spectrum. The replacement of the pyrrolindione carbonyl signal (at 188.0 ppm) in the 13C NMR spectrum of dione 7 by a signal at δ = 159.5 ppm in 8, assigned to the imidazolidinone moiety, and its regular chlorin-like UV−vis spectrum, as compared to the much broadened spectrum of dione 7 (Figure 1),9 supported our assignment of 9620

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

SNAr substitution of the p-F atoms of the pentafluorophenyl groups could be used for efficient derivatization, including solubilization of the chromophore in aqueous solution.24 Perfluorinated chromophores were also reported to be more resistant toward oxidative degradation.25 Both aspects are beneficial for numerous applications. However, we found that C6F5-substituted porpholactam 9F is not available along the classic route via the well-known meso-C6F5-substituted porpholactone.5b,d,e This is because the reaction of mesoC6F5-substituted porpholactone with hydrazine generated a large number of unidentified products. A major source for the formation of the numerous products appeared to be that hydrazine induced SNAr substitutions of the aryl-F atoms, though not in a well-controlled fashion. On the other hand, the alternative reaction pathway via the urea adduct to the recently described meso-tetrakis(pentafluorophenyl)-substituted dione26 7F proved to be suitable for the generation of the target porpholactam 9F, but with a complication (Scheme 2).

Figure 1. UV−vis spectra (CH2Cl2) of the compounds indicated.

the connectivity of 8. Moreover, the composition of 8 (C45H33N6O3 for M·H+, as per ESI+ HRMS) is also consistent with this. For a reproduction of the spectra of all new products, see SI. Chlorin 8 is available in multi-100 mg batches. We previously reported multiple methods for the oxidative cleavage of the pyrroline β,β′-bond of dihydroxychlorins.18 Thus, treatment of diol 8 under classic diol cleavage reaction conditions using Pb(OAc)421 resulted in the formation of a red, nonpolar compound in good yields (69%, at 100+ mg scales). Its porphyrin-like UV−vis spectrum (Figure 2), NMR

Scheme 2. Synthesis of mesoTetrakis(pentafluorophenyl)porpholactam 9F

Figure 2. UV−vis spectra (CH2Cl2) of the compounds indicated.

spectra, composition (as per HRMS), and comparison with a genuine sample, identified it to be known porpholactam 9.1b Presumably, the expulsion of smaller fragments (technically C2HNO2 = CO2 + HCN) from the putative nonplanar pyrrolemodified intermediate containing an 8-membered heterocycle to establish a stable and planar “tetrapyrrolic” architecture drives the reaction. The mechanism of this reaction is undoubtedly complex and was not studied. These types of ring-contraction reactions that establish “tetrapyrroles” from porphyrinoids containing larger than 5-membered building blocks have been previously observed by us, and others.1a,5a,22 The fragmentation of the presumed pyrrole-expanded intermediate can be halted by N-methylation; this allows the isolation of the unusual 1,3,6-triazocine-2,4,8-trione-derived pyrrole-modified porphyrin.8 The 3-step pathway toward porpholactam 9 from meso-tetraphenylchlorin diol 5 (5 → 7 → 8 → 9) is more convenient, economic, and significantly higher yielding (overall 29% at a 100 mg scale) than the known 3-step alternative pathway (under 10% yield along 5 → 6 (two steps) → 9, at 10−20 mg scales). Synthesis of meso-Pentafluorophenylporpholactam 9F. The synthesis of the meso-C6F5-substituted porpholactam 9F is desirable as we,10c,d and others,23 have shown that an

Reaction of the meso-perfluorinated dione 7F with urea generated the expected polar chlorin diol 8F, albeit in lower yields (65%) compared to the meso-phenyl-analogue (85%). Chlorin diol 8F showed the expected spectroscopic signatures: Inter alia, a 2-fold symmetric 1H NMR spectrum with two (CO)NH and two OH moieties, a 19F NMR spectrum showing the facial differentiation of the o-F atoms on the mesoaryl group adjacent to the pyrroline, the presence of a νCO at 1695.9 cm−1 indicating the presence of the urea bislactam moiety, and a chlorin-type UV−vis spectrum (Figure 3). It is chemically much less stable than the parent meso-phenylchlorin diol 5;27 presumably on account of extensive fragmentation, we could not acquire interpretable ESI+ or ESI− MS spectra. Next 9621

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

connectivity shown. The formation of the intramolecular linkage between the alcohol(s) of a dihydroxychlorin (or bacteriochlorin) and the o-position of a flanking mesofluorophenyl group was observed before.27 Oxidation of diol 8F using Pb(OAc)4 induced the same efficient ring fragmentation/contraction reaction as observed in its phenyl congener 8, generated a single product of lower polarity than the starting materials and with the spectroscopic and analytical properties expected for the target mesotetrakis(pentafluorophenyl)porpholactam 9F (composition of C43H10F20N5O for [M·H]+, as per ESI+ HRMS). The UV−vis spectrum of the meso-perfluorinated porpholactam 9F varies to a significant degree from that of the meso-phenyl derivative 9, reflecting the electronic influence of the meso-C6F5 groups (Figure 2).5b,28 The blue-shifted Soret and Q-bands, and the relatively and absolutely more intense λmax band of the fluorinated derivative compared to its nonfluorinated analogue are also observed in other meso-C 6 H 5 vs meso-C 6 F 5 porphyrinoid pairs (see SI).5b,28 Porpholactam 9F is now also accessible in 50 mg batches for further study. The equivalent oxidation of the chromene-annulated chlorin 11F led to extensive decomposition with no identifiable major product that could be isolated. Conversion of Porpholactam 9 to Reactive Imidazoloporphyrin-α-Triflate 12. Initial experiments on the substitution of the Cl-atom in chloroimidazoloporphyrin 101b were disappointing as it exhibited low susceptibility toward nucleophilic exchange with amines. Thus, we set out to

Figure 3. UV−vis spectra (CH2Cl2) of the compounds indicated.

to the chlorin diol, the less polar product 11F was also formed, in varying yields of up to 31%. Its chlorin-type optical spectrum showed a slight bathochromic shift compared to that of chlorin diol 8F (Figure 3). Diagnostically, its 1H NMR spectrum showed no symmetry and allowed the identification of only a single OH group (at 4.63 ppm) but two nonequivalent (CO)NH moieties (at 6.79 and 7.02 ppm). Its composition (C45H12F19N6O3 for M·H+, as per ESI+ HRMS) suggested that it was derived from chlorin diol 8F by loss of HF. The loss of one o-F atom was confirmed through 19 F NMR spectroscopy (19 F atoms could be seen, 7 o-, 8 m-, and 4 p-F atoms). The νCO band of the urea adduct at 1731.6 cm−1 was nearly unchanged compared to that seen in diol 8F. Product 11F was therefore assigned the chromene-annulated

Scheme 3. Conversion of Porpholactam 9 to Imidazoloporphyrin β-Substituted Derivatives

9622

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry prepare the corresponding triflate derivative, expecting it to be more reactive. We reacted purple porpholactam 9 with triflic anhydride (Tf2O) in dry CH2Cl2 under an inert atmosphere at −78 °C (Scheme 3). Upon warming to ambient temperature, one major nonpolar (Rf = 0.88, silica-CH2Cl2), purple compound was obtained in good yields (76%). Its composition (C44H29F3N5SO3 for M·H+, as determined by ESI+ HRMS) agreed with the composition of the target imidazoloporphyrin triflate 12. Notably absent from its 1H NMR spectrum was the diagnostic (CO)NH peak of the porpholactam, a signal commensurate with a lactam carbonyl carbon in its 13C NMR spectrum, and a single diagnostic peak (s at −71.1 ppm) in its 19 F NMR spectrum, all as expected of this triflate moiety. Its regular porphyrin-type UV−vis spectrum (λmax = 654 nm) is very similar to that of the parent imidazoloporphyrin 10 (λmax = 655 nm)1b (see SI). Once crystallized, the compound is stable and can be stored for months in a freezer. Reaction of Imidazoloporphyrin Triflate 12 with Nitrogen Nucleophiles. Triflate imine 12 was expected to react with a range of nucleophiles. Indeed, substitution of the triflate group by pyridine, either in situ when 12 was prepared by reaction of lactam 9 with Tf2O in the presence of pyridine, or by reaction of isolated 12 with pyridine, generated pyridinium-substituted imidazoloporphyrin 13 in excellent (86%) yield. Its spectroscopic and analytical data (e.g., composition of C48H33N6 for M·H+, as per ESI+ HRMS) confirmed its assigned structure. The UV−vis spectral properties of 13 (λmax = 683 nm) are, due to the presence of the positive charge of the pyridinium substituent located on a β-position of the chromophore, significantly altered compared to those of its starting material 12 (Figure 4). The influence of a cationic charge directly attached to the porphyrinic macrocycle on the optical properties of its chromophore have been recognized.29

Reaction of triflate 12 with Et2NH (in the presence of Na2CO3) at ambient temperature smoothly afforded the diethylamine-substituted imidazoloporphyrin 14 (Scheme 3). Diagnostic for its connectivity are the appearance of the signals for the two ethyl groups in its 1H NMR spectrum; all other spectroscopic and analytical data (e.g., composition of C47H39N6 for M·H+, as per ESI+ HRMS) are also consistent with the structure shown. Interestingly, the UV−vis spectrum of this compound is significantly broadened and features a split Soret band (Figure 4). No change was observed upon addition of an excess of Et3N, but the addition of acid (TFA) led to protonation-related changes in the spectrum. The resulting spectrum is dissimilar to that of the pyridinium derivative 13 under neutral conditions, but very similar to the spectrum of 13 upon protonation. This suggests that on addition of TFA to 14, protonation had taken place at the same positions as in 13 under acidic conditions. This further implies that the β-amine and at least one other nitrogen, either at the imidazole moiety and/or the central porphyrinic cavity, is protonated in acidic solution. The structurally closest relatives to these peripherally substituted imidazoloporphyrins are the substituted Nconfused porphyrins recently introduced by the group of Furuta and utilized in a study of how the substituents changed the tautomeric equilibria in the macrocycle.30 However, their derivatives are carbon-based and the stepwise build-up from βcyano-N-confused porphyrins varied from the realization of the amine-substituted imidazoloporphyrins described here. Modification of Imidazoloporphyrin with Chelating Motifs. With the proof of concept in hand that triflate 12 can be efficiently derivatized with amines, we investigated whether amine-based chelates could be introduced. As we will detail with three examples (using an imino-diethyl acetate/-diacetic acid, a di(2-picolyl)amine, and a cyclen moiety), this is indeed possible. Reaction of triflate imidazole 12 with diethyl iminodiacetate in the presence of Na2CO3 in hot acetone over 24 h afforded a polar, brown compound. The composition of the product (C51H43N6O4 for M·H+, as per ESI+ HRMS) matched that of the target chelate 15. Observed in its 1H NMR spectrum were the peaks corresponding to the ethyl ester groups, as well as a peak at 4.4 ppm assigned to the methylene group of the iminodiacetate moiety. A single crystal of 15 could be grown to confirm the connectivity of the chromophore (Figure 5). The structure also revealed the macrocycle to be slightly nonplanar, with similarly strong but overall minor saddling and ruffling deformation modes discerned by an NSD analysis.31 In comparison, the solid state structure of the unmodified free base imidazoloporphyrin 1 showed it to be almost perfectly planar (Figure 5),1a whereby the crystal structure of chloroimidazoloporphyrin 10 shows the chloroimidazole group to be slightly slanted out of the idealized plane formed by the remainder of the macrocycle (not shown).1b The open conformation of the chelating unit is not particularly well preorganized to bind a metal, a finding that is not untypical for these types of flexible ligands.4 For the metal ion binding experiments, the diacid derivative of 15 was needed. Thus, nonpolar diester 15 was saponified (KOH, wet THF, reflux conditions, 2 h) (Scheme 3). After workup, a brown, polar product (19) was obtained that showed the expected loss of the ester side chains by HRMS. The free acid was only slightly soluble in CHCl3 and CH2Cl2, but soluble in CH2Cl2/MeOH mixtures as well as pure MeOH.

Figure 4. UV−vis spectra in CH2Cl2 (blue traces) and CH2Cl2 + 10% TFA (red traces) of the compounds indicated. 9623

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

Figure 5. Left: Stick presentations of the molecular structures of the diethyl iminodiacetate- and DPA-substituted imidazoloporphyrins 15 and 16, respectively, top (top row) and front (bottom row) views; all hydrogens on sp2-hybridized carbons, disorder, and solvents removed for clarity. Right: Comparative NSD analysis31 of the chromophore conformations in the unsubstituted, parent imidazoloporphyrin 1 and the two substituted derivatives 15 and 16.

The 1H NMR (in CD2Cl2/MeOD) of the free acid indicated a slight upfield shift but otherwise full retention of the methylene peaks (∼4 ppm) of the iminodiacetic acid moiety; all other spectral data were also as expected (see SI). The di(2-picolyl)amine (DPA)-substituted imidazoloporphyrin 16 could be prepared in good yields (76%) by treatment of 12 with DPA in CH2Cl2 under ambient conditions in the presence of Na2CO3. The 1H NMR spectrum of the DPA-derivative 16 was characterized by the presence of additional peaks in the aromatic region exhibiting the expected 1 H−1H COSY correlations of the 2-substituted pyridine moieties, along with a peak corresponding to the methylene unit of the DPA moiety (see SI). All other spectroscopic and analytical data also supported the formation of the target compound, ultimately also shown by single crystal X-ray diffractometry (Figure 5). Unlike the chromophore of the corresponding iminodiacetate derivative 15, the DPA-derivative 16 is significantly more saddled, with few other conformation modes discernible, perhaps providing a first indication that the DPA moiety is introducing some intramolecular strain into the molecule. The four potentially coordinating nitrogen atoms of the ligand and the imidazole moiety are not preorganized into a metal binding pocket. The metrics of 15 and 16 will be discussed in more detail below. Lastly, under the conditions of the formation of the DPAderivative 16, we could also prepare the corresponding cyclensubstituted imidazoloporphyrin 17 by treatment of 12 with stoichiometric excess of cyclen. In this case, however, we observed the formation of a major product 17 (62% yield) and a mixture of two difficult to separate minor products (in a combined total of 16%). A series of four peaks in the aliphatic region of the 1H NMR spectrum of the major product 17, each integrating to four protons, indicated the presence of the cyclen unit in 17. The expected composition of the target compound was also confirmed by HRMS (C51H48N9 for M·H+, as for per ESI+). Its UV−vis spectrum was, like that of the other derivatives, only slightly different from that of unsubstituted imidazoloporphyrin 1 (see SI). The minor, less polar, brown products possessed similar but slightly broadened

UV−vis spectra as that of 17 (see SI) and identical compositions as per HRMS (C94H75N14 for M·H+, as per ESI+), suggesting a connectivity of two imidazoloporphyrin moieties per cyclen moiety. We thus assigned them to the cyclen-1,4- and 1,7-disubstituted dimeric species; no further separation or characterization of this minor fraction was attempted. Metal Sensing by Imidazoloporphyrin Derivatives. All three metal binding moieties utilized in the imidazoloporphyrin derivatives 16 (DPA), 17 (cyclen), and 19 (iminodiacetic acid) are excellent and versatile chelates for a range of metal ions.4 Alas, their imidazoloporphyrin conjugates did not display the strong sensing profiles hoped for. We should point out that this highlights a general design flaw of these chemosensors that inhibits the involvement of the imidazole βnitrogen in the coordination interaction, thereby greatly diminishing the sensor response. In detail: Addition of a range of M2+ ions (for example, Ca2+, Mg2+, Zn2+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+) in methanolic solution to the DPA-substituted imidazoloporphyrin 16 showed the same and almost negligible UV−vis absorption and fluorescence emission responses (less than 2% spectral changes) of the parent unsubstituted imidazoloporphyrin 1. This finding suggests that, at a minimum, the DPA moiety does not bind a metal in a way that allows a chelating interaction with the imidazoloporphyrin nitrogen. A molecular model suggests that binding of a metal ion (such as shown for 16·Zn) in a chelating fashion introduces some steric interaction between the methylene groups of the DPA moiety and the flanking meso-phenyl group. In addition, the number of atoms between the imidazole nitrogen and the DPA alkylamine implies the formation of an undesirable 4-membered metallacycle upon metal chelation, rather than a preferred 5- or 6-membered ring. Lastly, the large bite angle (120°, Figure 5) between the two nitrogen atoms makes formation of a 4membered metallacycle practically impossible. The latter design feature also affects the ability of the other derivatives to bind metal ions. 9624

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

binding site with Ni2+, no response was observed upon addition of zinc (Figure 6B). The sum of the results suggests that the iminodiacetic acid moiety on 19 does not bind to zinc, at least not in a way that affects the optical properties of the chromophore, i.e., unlikely involving the imidazoloporphyrin nitrogen. The titration of 19 with Ni2+ (and similarly with Ni2+, Cu2+, and Co2+) showed a different optical response (Figure 6C). While the Soret band also blue-shifts (by 10 nm) and increases in intensity, the subtle emergence of two Q bands at 618 and 674 nm is noted. These metal ions did not insert into the central cavity of the porphyrin, as a comparison of the final UV−vis spectrum resulting from the titration with the spectrum of the genuine Ni-complex 20 (Figure 6C, insert) clearly demonstrates. These metal ions are known to require much harsher reaction conditions to be inserted into a porphyrin.11 The observed optical changes may thus be linked to coordination to the peripheral ligand. All-in-all, however, the response is significantly weaker than hoped for. The spectrophotometric titration data of cyclen derivative 17 with Zn2+ also highlights the propensity of this conjugate to coordinate this metal ion at its periphery (Figure 6D). A Hill plot analysis of the Zn2+ titration data indicates strong (Kd = 9.0 μM) binding of zinc to 17 (Figure 6D, insert). Over time, however, we observed a change of the optical spectrum of the peripherally zinc-bound species to a species that suggested that metal insertion into the porphyrinic macrocycle had taken place. Evidently (and not surprisingly), zinc ion binding to the peripheral and more flexible N4-coordination sphere of the cyclen moiety is faster than zinc ion binding to the more rigid N4-coordination sphere of the porphyrinic macrocycle. Over

The responses of the iminodiacetic acid and cyclen derivatives 19 and 17 with respect to the addition of metal ions were better but still not great, likely the result of the metric shortcomings of their design. We thus present here only select metal binding experiments that illustrate some general findings, but also reveal a surprise. The spectrophotometric titration of iminodiacetic acid derivative 19 with a range of divalent metal ions showed two distinct modes of optical responses (Figure 6): Addition of Zn2+, for example, was marked by, over time, an increase and sharpening of the Soret band, along with a small (6 nm) hypsochromic shift, and a reduction of the number of Q-bands (Figure 6A). Thus, the dominant response is metal ion insertion into the porphyrin ring.11,13 In fact, Zn2+ is known to readily insert into porphyrins at room temperature,11 and the UV−vis spectrum of the zinc chelate of imidazoloporphyrin 1 is nearly identical to the final spectrum of the zinc-titration of 19 (Figure 6A, insert). Correspondingly, when we tested the spectrophotometric response of 20, the iminodiacetic acid derivative 19 in which we blocked the porphyrinic metal ion

Figure 6. (A) Spectrophotometric titrations of 19 titrated with Zn2+; (A insert) normalized absorption spectra of free base imidazoloporphyrin 1 and its internally metalated [imidazoloporphyrinato]zinc(II) complex 1Zn (CH2Cl2); (B) spectrophotometric titration of 20 with Zn2+; (C) spectrophotometric titration of 19 with Ni2+, and (D) spectrophotometric titrations of 17 with Zn2+, including (D insert) Hill plot model of reaction coordinate as a function of change in Soret band intensity. Conditions: [19] = 6.4 × 10−6 M, [20] = 5.7 × 10−6 M, [17] = 5.1 × 10−6 M, MeOH, 25 °C, aerated solution, addition of Zn2+/Ni2+ as 1.00 mM (A−C) or 4.39 mM (D) methanolic solutions of their chlorides, 5 min equilibration times between metal addition and recording of spectrum. 9625

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry

159.5, 158.5, 152.8, 141.4, 141.1, 140.0, 135.4, 135.2, 134.12, 134.10, 134.08, 134.06, 133.2, 133.1, 128.6, 128.4, 127.5, 127.3, 127.7, 125.7, 123.1, 112.9, 93.5 ppm; UV−vis (CH2Cl2) λmax (log ε) 406 (5.41), 512 (4.30), 541 (4.30), 591 (3.96), 642 (4.53) nm; FT-IR (neat, diamond ATR) νCO = 1715.7 cm−1; HRMS (ESI+, 100% CH3CN, TOF) calcd for C45H33N6O3 (M·H+) 705.2609, found 705.2607. β,β′-Imidazolidinone-Annulated meso-Tetrakis(pentafluorophenyl)dihydroxychlorin 8F and β,β′-Imidazolidinone-Annulated, Chromene-Annulated meso(Pentafluorophenyl)dihydroxychlorin 11F. Prepared from meso(pentafluorophenyl)-derived dione 7F (30 mg, 3.0 × 10−5 mol) in pyridine (5 mL) in a 10 mL round-bottom flask equipped with a magnetic stir bar, nitrogen inlet, and reflux condenser and urea (80 mg, 1.3 × 10−3 mol, 45 equiv) as described for 8, yielding the magenta product in 65% yield (20 mg). 8F: Rf (silica-CH2Cl2/5% MeOH) = 0.26; 1H NMR (400 MHz; DMSO-d6) δ 9.21 (dt, J = 1.2, 0.6 Hz, 1H), 8.84−8.78 (m, 2H), 7.61 (s, 1H), 5.48 (s, 1H), −2.42 (s, 1H) ppm; 19F NMR (376 MHz; DMSO-d6) δ −134.0, −138.1, −139.7, −153.4, −155.1, −160.9, −162.5, −165.2 ppm; UV−vis (CH2Cl2) λmax (log ε) 400 (4.61), 501 (3.66), 594 (3.24), 585 (3.75), 647 (3.99) nm; FT-IR (neat, diamond ATR) νCO = 1695.9 cm−1; HRMS (ESI+ or ESI−, 100% CH3CN, TOF) could not be obtained. 11F: Rf (silica-CH2Cl2/5% MeOH) = 0.31; 1H NMR (400 MHz, CDCl3) δ 8.88 (s, 1H), 8.64−8.64 (m, 3H), 8.45 (s, 2H), 8.33 (s, 1H), 7.02 (s, 1H), 6.79 (s, 1H), 4.63 (s, 1H), −1.03 (br s, 1H, exchangeable with D2O), −1.26 (br s, 1H, exchangeable with D2O) ppm; 19F NMR (376 MHz, CDCl3) δ −133.2, −135.2, −136.4, −136.7, −137.1, −137.9, −150.8, −151.1, −153.0, −157.5, −160.5, −160.7, −161.1 ppm; UV−vis (CH2Cl2) λmax (log ε) 408 (4.72), 508 (3.44), 541 (3.65), 599 (3.19), 652 (3.95) nm; FT-IR (neat, diamond ATR) νCO = 1731.6 cm−1; calcd for C45H11F19N6O3 (M·H+) 1045.0662, found 1045.0636. meso-Tetraphenyporpholactam (meso-Tetraphenyl-2-aza-3-oxoporphyrin) 9. Dihydroxychlorin 8 was dissolved in dry THF (11 mL) and Et3N (5−6 drops) in a round-bottom flask equipped with a magnetic stir bar. Freshly acquired Pb(OAc)4 (32 mg, 7.1 × 10−5 mol, 1.7 equiv) was added in portions and the reaction mixture was stirred at ambient temperature. When the starting material was consumed (reaction monitored by UV−vis spectroscopy and TLC), the solvent was removed by rotary evaporation and the remaining residue was purified by column chromatography or preparative plate chromatography (silica-CH2Cl2/1% MeOH) to afford known porpholactam 9 in 69% yield (18 mg). Spectroscopic data as previously described.1b meso-Tetrakis(pentafluorophenyl)porpholactam 9F. Prepared from β,β′-imidazolidinone-annulated dihydroxychlorin 8F (37 mg, 3.5 × 10−5 mol) dissolved in dry THF (11 mL) and Et3N (5−6 drops) in a 25 mL round-bottom flask equipped with a magnetic stir bar and fresh Pb(OAc)4 (26 mg, 5.9 × 10−5 mol, 1.7 equiv) as described for its phenyl-congener 9. Purified by column or preparative plate chromatography (silica-CH2Cl2/2% MeOH) to afford porpholactam 9F in 75% yield (28 mg). 9F: 1H NMR (400 MHz; CDCl3) δ 9.98 (s, 1H, exchangeable with D2O), 8.90 (d, J = 5.0 Hz, 1H), 8.86 (d, J = 4.7 Hz, 1H), 8.81 (d, J = 4.7 Hz, 1H), 8.74 (d, J = 4.8 Hz, 1H), 8.66 (d, J = 4.6 Hz, 1H), 8.60 (d, J = 4.7 Hz, 1H), −1.98 (s, 1H, exchangeable with D2O), −2.17 (s, 1H, exchangeable with D2O) ppm; 19F NMR (376 MHz, CDCl3) δ −135.9, −136.7, −138.9, −149.3, −150.9, −152.7, −159.5, −161.1, −162.3 ppm; 13C NMR (100 MHz, CDCl3) δ 169.4, 156.1, 154.1, 148.6, 140.3, 139.7, 139.0, 138.2, 136.3, 134. 6, 133.5, 129.3, 127.8, 127.0, 126.1, 115.1, 115.0, 114.9, 112.7, 111.6, 111.6, 107.7, 104.7, 100.7, 88.7, 77.3, 77.22, 77.17, 77.1, 77.0, 76.93, 76.88, 76.76, 76.69, 29.72 ppm; UV−vis (CH2Cl2) λmax (log ε) 408 (4.72), 508 (3.44), 541 (3.65), 599 (3.19), 652 (3.95) nm; FT-IR (neat, diamond ATR) νCO = 1703.3 cm−1; HRMS (ESI+, 100% CH3CN, TOF) calcd for C43H10F20N5O (M·H+) 992.0561, found 992.0550. meso-Tetraphenyl-2-aza-3-trifluoromethylsulfonylporphyrin (12). To a flame-dried 25 mL round-bottom flask equipped with a septum was added 10 mL of dry CH2Cl2 to dissolve meso-tetraphenyl2-aza-3-oxo-porphyrin 9 (21 mg, 3.35 × 10−5 mol) under an atmosphere of dry N2. The mixture was cooled to −78 °C and stirred

time, zinc insertion into the internal cavity of the imidazoloporphyrin in (cyclen·zinc-bound) 17 was observed in the presence of a stoichiometric excess of zinc ions. This was as expected. Surprisingly, however, was the finding that the cyclen substitution had a distinct effect on the rate at which the zinc ion inserted into the imidazoloporphyrin. When both the parent unsubstituted imidazoloporphyrin 1 and cyclen derivative 17 were treated with 5 equiv of Zn2+ under identical conditions, insertion of zinc into the porphyrinic macrocycle was at least an order of magnitude faster for 17 than it was for 1 (see SI). Whether this is because cyclen-modified derivative 17 is, parallel to that of the 15 or DPA-derivative 16 (cf. to Figure 5) less planar than 1 or because of any assistance from the side chains remains unclear.

■

CONCLUSIONS In conclusion, we introduced a new and significantly more efficient (and nonobvious) synthetic pathway toward known meso-tetraphenylporpholactam 9 that is also applicable to the synthesis of hitherto unknown and inaccessible C6F5substitued porpholactam 9F. We demonstrated the conversion of meso-tetraphenylporpholactam to an imidazoloporphyrin triflate derivative 12 that proved useful in the preparation of imidazoloporphyrins with a substituted amine adjacent to the outside N atom of the imidazole moiety. When choosing amines that are part of well-known and versatile chelating units (DPA, cyclen, and iminodiacetic acid),4 a number of conjugates were prepared that were designed to chelate a metal ion while also involving the outside imidazole nitrogen atom in their coordination sphere. Alas, the potential imidazoloporphyrin conjugate chemosensors for (transition) metal ions did not display the strong sensing profiles that were hoped for. This was rationalized by their metric parameters, revealing a principle design flaw. This notwithstanding, the facile replacement of the triflate in the imidazoloporphyrin αtriflate 12 by a variety of amines demonstrates the utility of this derivative for the generation of functionalized imidazoloporphyrins. Its accessibility now enables the continued exploration of the reactivity of the porpholactams and imidazoloporphyrins.

■

EXPERIMENTAL SECTION

Materials. All solvents and reagents were used as received. Diols 521b and 5F5e and diones 79 and 7F26 were prepared as described in the literature. Fluorescence Measurements. The fluorescence quantum yields (ϕ) were determined relative to that of meso-tetraphenylporphyrin (2) (0.13 in CH2Cl2,);32 λexcitiation = λSoret. Preparation of Compounds. β,β′-Imidazolidinone-Annulated meso-Tetraphenyldihydroxychlorin 8. Porphyrin dione 7 (118 mg, 1.83 × 10−4 mol) was dissolved in pyridine (25 mL) in a roundbottom flask equipped with a magnetic stir bar, reflux condenser, and nitrogen inlet. Urea (245 mg, 4.1 × 10−3 mol, 22 equiv) was added and the mixture was heated to reflux for 30 min under a N2 atmosphere. The solvent was evaporated in vacuo. The remaining residue was taken up in CHCl3 and filtered through a glass frit (M). The filtrate was washed with 5 × 25 mL distilled water and dried over Na2SO4. The dried residue was separated by column chromatography (CH2Cl2−5% MeOH) recovering dione 7 in 8% yield (9 mg), followed by the magenta product 8 in 85% yield (110 mg): Rf (silicaCH2Cl2/5% MeOH) = 0.36; 1H NMR (400 MHz, CDCl3) δ 8.64 (d, 3 J = 4.7 Hz, H), 8.49 (s, 1H), 8.21 (s, 1H), 8.14 (d, 3J = 5.8 Hz, 3H), 7.86 (d, 3J = 6.8 Hz, 1H), 7.72 (m, 6H), 5.34 (br s, 1H, exchangeable with D2O), 4.26 (br s, 1H, exchangeable with D2O), −2.03 (br s, 1H, exchangeable with D2O) ppm; 13C NMR (100 MHz, DMSO-d6) δ 9626

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

Article

The Journal of Organic Chemistry for 20 min. Triflic anhydride (300 μL) was added dropwise via syringe to the reaction flask. The mixture was stirred for 20 min at −78 °C (acetone/dry ice bath), then warmed to 0 °C (ice/water bath) and stirred for an additional 20 min. The reaction mixture was allowed to warm to ambient temperature and stirred for 15 h until the starting material was consumed (reaction monitored by UV−vis spectroscopy and TLC). The reaction was quenched by addition of a saturated aq solution of sodium bicarbonate (NaHCO3, 5 mL) and CH2Cl2 (5 mL). The biphasic mixture was poured rapidly into a separatory funnel and the organic layer was separated from the aqueous layer. The aqueous layer was extracted with CH2Cl2 (2 × 5 mL), and the combined organic layers were dried over anhyd Na2SO4. The crude material was purified by preparative TLC (CH2Cl2/20% petroleum ether 30−60) to furnish 12 as a purple solid in 76% (20 mg) yield. 12: Rf (silica-CH2Cl2) = 0.88; 1H NMR (400 MHz, CDCl3) δ 9.17 (d, 3J = 5.2 Hz, 1H), 8.96 (d, 3J = 5.0 Hz, 1H), 8.90 and 8.88 (two overlapping d, 3J = 5.0 Hz, 2H) 8.74 (s, 2H), 8.27 (d, 3J = 7.7 Hz, 2H), 8.20 (t, 3J = 7.7 Hz, 4H), 8.11 (d, 3J = 7.0 Hz, 2H) 7.85−7.77 (m, 12H) − 2.46 (s, 1H, exchangeable with D2O) ppm; 19F NMR (376 MHz, CDCl3, external reference CFCl3 set to −71.1 ppm); 13C NMR (100 MHz, CDCl3) δ 163.6, 157.5, 157.2, 153.0, 141.6, 140.6, 140.3, 140.1, 139.3, 139.0, 138.8, 136.2, 135.7, 135.6, 134.8, 134.7, 134.5, 133.6, 130.2, 129.6, 129.0, 128.8, 128.3, 127.9, 127.8, 127.2, 127.1, 126.9, 121.6, 121.4, 121.1, 120.6, 119.2, 117.4 ppm; UV−vis (CH2Cl2) λmax (log ε) 422 (5.80), 524 (4.40), 562 (3.94), 598 (3.90), 654 (4.42) nm; UV−vis (CH2Cl2 + 10% TFA) λmax (log ε) 447 (5.62), 680 (4.76) nm; HRMS (ESI+, cone voltage = 30 V, 100% CH3CN, TOF) calcd for C44H29F3N5SO3 (M·H+) 764.1938, found 764.1943. meso-Tetraphenyl-2-aza-3-(N-pyridinium)porphyrin chloride (13). Compound 13 was prepared similarly to 12, except that pyridine (1 mL) was also present in the reaction flask. The crude material was purified by preparative TLC (CH2Cl2/5% MeOH) to furnish 13 as a purple solid in 86% yield (15 mg). 13: Rf (silicaCH2Cl2/10% MeOH) = 0.38; 1H NMR (400 MHz, CDCl3) δ 9.12 (d, 3J = 5.4 Hz, 2H), 9.05 (d, 3J = 4.9 Hz, 1H), 8.90 (d, 3J = 4.9 Hz, 1H), 8.85 (s, 2H), 8.68 (s, 2H), 8.38 (t, 3J = 7.5 Hz, 1H) 8.25 (dd, 3J = 5.8, 1.6 Hz, 2H), 8.19−8.17 (m, 4H), 7.99 (br d, 3J = 4.8 Hz, 4H), 7.83−7.76 (m, 6H) 7.61 (br d, 3J = 5.1 Hz, 3H), 7.46−7.43 (m, 3H), −1.99 (s, 1H, exchangeable with D2O) ppm; 13C NMR (100 MHz, CDCl3) δ 159.3, 158.9, 152.8, 147.5, 145.9, 142.7, 142.2, 141.0, 140.9, 140.4, 139.5, 139.3, 137.8, 136.5, 136.4, 135.8, 135.1, 134.9, 134.8, 131.5, 130.9, 129.8, 129.5, 128.6, 128.5, 128.4, 128.3, 127.4, 127.3, 127.2, 123.5, 121.7, 121.6, 119.4 ppm; UV−vis (CH2Cl2) λmax (log ε) 431 (5.07), 538 (3.93), 528 (3.45), 630 (3.51), 683 (4.07) nm; UV− vis (CH2Cl2 + 10% TFA) λmax (log ε) 463 (5.14), 728 (4.37) nm; HRMS (ESI+, cone voltage = 30 V, 100% CH3CN, TOF) calcd for C48H33N6 ([M − Cl−)+ 693.2761, found 693.2770. meso-Tetraphenyl-2-aza-3-diethylaminoporphyrin (14). To a 25 mL round-bottom flask equipped with magnetic stir bar was added triflate 12 (40 mg, 5.25 × 10−5 mol), Na2CO3 (14 mg, 0.13 × 10−3 mol), diethylamine (0.10 mL) and CH2Cl2 (5 mL). The mixture was stirred at room temperature for 20 h. After the starting material was consumed, the solvent/excess diethylamine was evaporated and the product was purified by column chromatography (silica-CH2Cl2/1% MeOH) to obtain 14 as a yellow-brown solid in 87% yield (32 mg). 14: Rf (silica-CH2Cl2/5% MeOH) = 0.75; 1H NMR (500 MHz, CDCl3) δ 8.91 (d, 3J = 4.9 Hz, 1H), 8.78 (d, 3J = 4.9 Hz, 1H), 8.75 (d, 3J = 4.9 Hz, 1H), 8.70 (d, 3J = 4.9 Hz, 1H), 8.62 (d, 3J = 4.56 Hz, 1H), 8.61 (d, 3J = 4.56 Hz, 1H), 8.34−8.31 (m, 4H), 8.23−8.20 (m, 4H), 7.79−7.73 (m, 10H), 7.71−7.68 (m, 2H), 3.48 (q, 3J = 7.1 Hz, 4H), 1.01 (t, 3J = 7.1 Hz, 6H), −1.90 (broad s, 2H, exchangeable with D2O) ppm; 13C NMR (125 MHz, CDCl3) δ 171.6, 160.8, 155.8, 154.1, 142.4, 142.3, 142.0, 140.6, 140.2, 139.6, 138.6, 137.6, 136.1, 135.7, 134.9, 134.8, 134.7, 134.1, 133.4, 128.29, 128.25, 128.2, 128.9, 127.8, 127.7, 127.04, 126.97, 126.90, 126.86, 126.81, 126.7, 121.7, 119.6, 117.9, 116.3, 77.2, 46.2, 12.5 ppm; UV−vis (CH2Cl2) λmax (log ε) 406 (4.89), 449 (4.95, 538 (4.19), 615 (3.64) 673 (3.16) nm; Fl (DMF, λexcitation = λSoret) λmax 655, 712 nm, ϕ (DMF) = 0.02; FT-IR (neat, diamond ATR) νCO = 1522.4 cm−1; HRMS (ESI+, 100%

CH3CN, TOF) calcd for C47H39N6 (M·H+) 687.3236, found 687.3242. meso-Tetraphenyl-2-aza-3-(diethyliminodiacetate)porphyrin (15). To a 25 mL round-bottom flask equipped with magnetic stir bar was added triflate 12 (49.0 mg, 6.4 × 10−5 mol), sodium carbonate (15 mg, 1.4 × 10−4 mol, 2.2 equiv), diethyl iminodiacetate (250 μL, 1.4 mmol, 21 equiv) and acetone (8 mL). The mixture was heated to reflux overnight. After 23−27 h, the reaction mixture was evaporated to dryness. Flash column chromatograpy (silica-CH2Cl2/5% MeOH) afforded recovered 12 (10−13 mg, 20−27%) and 15 as a yellow/ brown solid (36−39 mg, 70−76%). 15: Rf (silica-CH2Cl2/2% MeOH) = 0.63; 1H NMR (500 MHz, CDCl3) δ 8.93 (d, 3J = 5.0 Hz, 1H), 8.83 (d, 3J = 4.9 Hz, 1H), 8.75 (two overlapping doublets, 3J = 4.5 Hz, 2H), 8.65 (s, 2H), 8.43 (d, 3J = 7.1 Hz, 2H), 8.24−8.20 (m, 6H), 7.82−7.70 (m, 13H), 4.39 (s, 4H), 4.03 (q, 3J = 7.1 Hz, 4H), 1.16 (t, 3J = 7.1 Hz, 6H), −2.11 (br s, 2H, exchangeable with D2O) ppm; 13C NMR (100 MHz, CDCl3) δ 170.0, 158.6, 156.1, 154.6, 142.2, 141.8, 140.1, 139.9, 139.9, 139.6, 138.9, 138.0, 136.2, 134.9, 134.5, 134.4, 133.8, 133.6, 128.6, 128.5, 128.3, 128.3, 128.0, 127.9, 127.9, 127.7, 127.0, 126.97, 126.9, 126.6, 121.7, 119.9, 118.2, 117.2, 77.4, 77.2, 76.9, 60.7, 54.4, 30.0, 14.2 ppm; UV−vis (CH2Cl2) λmax (log ε) 439 (5.07), 532 (4.06), 568 (sh), 610 (3.5), 664 (3.1) nm; Fl (DMF, λexcitation = λSoret) λmax 682, 745 nm, ϕ (DMF) = 0.02; FT-IR (neat, diamond ATR) ν = 1734.4, 1258.1, 1014.6, 792.5, 698.39 cm−1; HRMS (ESI+, 100% CH3CN, TOF) calcd for C51H43N6O4 (M· H+) 803.3346, found 803.3368. meso-Tetraphenyl-2-aza-3-(iminodiacetic acid)porphyrin 19. In a round-bottom flask equipped with a magnetic stir bar, diester 15 (35.0 mg, 4.3 × 10−5 mol) was dissolved in THF (10 mL) at ambient conditions. Next, 1 M aqueous KOH (10 mL) was added and the reaction mixture was heated to reflux for 2 h. After the starting material was consumed (reaction monitored by TLC), the reaction mixture was allowed to cool and as much THF as possible was removed by rotary evaporation. 1 M Aqueous HCl (∼10.0 mL) was added to neutralize the aqueous phase (between pH 6 and 8), and the aqueous phase was extracted with CH2Cl2 (4 × 20 mL). The organic extracts were combined and evaporated to dryness. The remaining residue was purified by column chromatography (silicia-CH2Cl2/10% MeOH) to afford 19 as a yellow/brown solid in 80% yield (26 mg). 19: 1H NMR (400 MHz, CD2Cl2/10% MeOD) δ 8.92 (d, 3J = 5.0 Hz, 1H), 8.90 (d, 3J = 5.0 Hz, 1H), 8.87 (d, 3J = 5.0 Hz, 1H), 8.80 (d, 3 J = 5.0 Hz, 1H), 8.65 (s, 2H), 8.32 (d, 3J = 7.1 Hz, 2H), 8.21−8.17 (m, 6H), 7.88 (t, 3J = 7.4 Hz, 2H), 7.83−7.77 (m, 10H), 3.97 (s, 4H) ppm; 13C NMR (125 MHz, CD2Cl2/10% MeOD) δ 171.9, 169.2, 157.0, 155.4, 153.8, 141.5, 141.0, 140.2, 139.41, 139.32, 138.8, 137.6, 135.5, 134.9, 134.67, 134.57, 134.2, 128.97, 128.79, 128.3, 128.02, 127.92, 127.3, 127.0, 126.9, 122.2, 120.4, 119.5, 115.7, 56.7 ppm; UV−vis (MeOH) λmax (log ε) 442 (4.86), 534 (3.92), 571 (sh), 612 (3.51), 676 (3.0) nm; Fl (DMF, λexcitation = λSoret) λmax 655, 712 nm, ϕ (DMF) = 0.03; FT-IR (neat, diamond ATR) ν = 1603.5, 1216.0, 1151.7, 796.0, 752.3, 698.5 cm−1; MS (ESI+, 100% CH3CN, TOF) m/z calcd for C47H35N6O4 (M·H+) 747.2720, found 747.2701. DPA-Substituted Imidazoloporphyrin (16). To a round-bottom flask equipped with a magnetic stir bar under N2 was added triflateimidazole 12 (31 mg, 4.1 × 10−5 mol), di(2-picolyl)amine (365 μL, 2 × 10−3 mol, 50 equiv), Na2CO3 (17.0 mg, 1.6 × 10−4, 4 equiv) and CH2Cl2 (10 mL). The reaction mixture was stirred for 24 h. When the starting material was consumed (reaction monitored by TLC), the reaction mixture was filtered through a glass frit and the filtrate washed with water (3 × 20.0 mL). The organic layer was dried over Na2SO4, filtered, and the residue purified by preparative TLC (silicaCH2Cl2/10% MeOH) to afford DPA-substituted imidazoloporphyrin 16 as a yellow-brown solid in 76% yield (25 mg): 16: Rf (silicaCH2Cl2/10% MeOH) = 0.41; 1H NMR (400 MHz, CDCl3) δ 8.91 (d, 3J = 5.0 Hz, 1H), 8.80 (d, 3J = 5.0 Hz, 1H), 8.72 (d, 3J = 4.7 Hz, 1H), 8.69 (d, 3J = 4.9 Hz, 1H), 8.62−8.60 (m, 2H), 8.43 (s, 2H), 8.26 (d, 3J = 6.7 Hz, 2H), 8.19 (dd, 3J = 7.2, 4J = 2.0 Hz, 6H), 7.76−7.67 (m, 12H), 7.38 (td, 3J = 7.7, 4J = 1.6 Hz, 2H), 7.12 (d, 3J = 7.8 Hz, 2H), 7.03 (dd, 3J = 7.0, 4J = 5.3 Hz, 2H), 4.77 (s, 4H), −2.03 (s, 2H) ppm; 13C NMR (125 MHz, CDCl3) δ 171.3, 158.1, 156.1, 154.5, 9627

DOI: 10.1021/acs.joc.8b00790 J. Org. Chem. 2018, 83, 9619−9630

The Journal of Organic Chemistry

■

149.0, 142.1, 141.84, 141.78, 140.3, 139.9, 139.7, 138.8, 137.9, 136.2, 136.0, 135.3, 134.9, 134.7, 134.4, 133.7, 128.5, 128.4, 128.3, 128.3, 128.0, 127.9, 127.5, 127.1, 127.02, 126.97, 126.9, 126.7, 123.6, 122.0, 121.8, 119.8, 118.2, 116.7, 77.4, 77.2, 76.9, 58.8 ppm; UV−vis (CH2Cl2) λmax (log ε) 403 (sh), 449 (5.17), 542 (4.07), 588 (3.77), 618 (3.71), 681 (3.66) nm, UV−vis (CH2Cl2 + 10% TFA) λmax (log ε) 473 (5.15), 629 (sh), 680 (4.40) nm; Fl (DMF, λexcitation = λSoret) λmax 693, 750 nm, ϕ (DMF) = 0.01; FT-IR (neat, diamond ATR); HRMS (ESI+, 100% CH3CN, TOF) calcd for C55H41N8 (M·H+) 813.3454, found 813.3485. Cyclen-Substituted Imidazoloporphyrin 17 and Bis-CyclenSubstituted Imidazoloporphyrins 18. To a round-bottom flask equipped with a magnetic stir bar was added cyclen (7.0 mg, 4.1 × 10−5 mol, 1.6 equiv), Na2CO3 (11 mg, 1.0 × 10−4 mol, 4 equiv) and CH2Cl2 (3 mL). A solution of triflate-imidazole 12 (20.0 mg, 2.6 × 10−5 mol) in CH2Cl2 was added to the stirred solution under N2. The reaction mixture was stirred at r.t. for 2 h. The reaction mixture was filtered through a glass frit and the filtrate was washed with H2O (2 × 10 mL), brine, dried over Na2SO4, filtered, and reduced to dryness by rotary evaporation. The remaining residue was purified by preparative TLC (silica-CH2Cl2/10% MeOH) to recover 12 (2.0 mg, 10%), afford 17 (13.0 mg, 62%) and the mixture of the two dimers species 18 (6 mg, 16%). 17: Rf (silica-CH2Cl2/10% MeOH) = 0.09; 1H NMR (500 MHz, CDCl3) δ 8.91 (d, 3J = 5.0 Hz, 1H), 8.86 (two overlapping d, 3J = 4.1 Hz, 2H), 8.78 (d, 3J = 4.9 Hz, 1H), 8.65 (d, 3J = 4.6 Hz, 1H), 8.63 (d, 3J = 4.6 Hz, 1H), 8.27 (two overlapping d, 3J = 7.5, 7.0 Hz, 4H), 8.21−8.17 (m, 4H), 7.84 (td, 3J = 7.4, 4J = 3.4 Hz, 4H), 7.80−7.75 (m, 8H), 3.47 (s, 4H), 2.80 (s, 4H), 2.46 (s, 4H), 2.22 (s, 4H), −2.14 (s, 1H, exchangeable with D2O), −2.24 (s, 1H, exchangeable with D2O) ppm; 13C NMR (125 MHz, CDCl3) δ 170.5, 156.9, 155.4, 142.3, 141.8, 141.4, 140.24, 140.16, 139.5, 139.4, 139.3, 138.7, 135.4, 135.14, 135.07, 134.9, 134.8, 134.72, 134.68, 134.6, 134.3, 129.0, 128.8, 128.73, 128.66, 128.6, 128.42, 128.35, 128.28, 128.2, 128.1, 127.9, 127.5, 127.1, 127.03, 126.94, 122.4, 121.9, 120.3, 119.4, 119.3, 116.0, 51.0, 47.3, 44.0 ppm; UV−vis (CH2Cl2) λmax (log ε) 438 (5.0), 533 (4.0), 570 (sh), 610 (3.54), 670 (3.58) nm; UV−vis (CH2Cl2 + 10% TFA) λmax (log ε) 470 (4.9), 624 (sh), 682 (4.27) nm; Fl (DMF, λexcitation = λSoret) λmax 678, 739 nm, ϕ (DMF) = 0.02; FT-IR (neat, diamond ATR); HRMS (ESI+, 100% CH3CN, TOF) m/ z cald for C51H48N9 ([M·H]+) 786.4033, found 786.4061. 18: Rf (silica-CH2Cl2) = 0.45; 1H NMR (400 MHz, CDCl3) δ 8.82 (d, 3J = 4.9 Hz, 1H), 8.79 (d, 3J = 4.9 Hz, 1H), 8.70 (d, 3J = 4.6 Hz, 1H), 8.67 (d, 3J = 4.6 Hz, 1H), 8.42 (d, 3J = 5.0 Hz, 1H), 8.24 (d, 3J = 6.4 Hz, 2H), 8.15 (t, 3J = 8.6 Hz, 5H), 7.85−7.74 (m, 8H), 7.59−7.54 (m, 3H), 6.40 (s, 1H), 5.88 (s, 1H), 3.36 (s, 4H), 2.61 (s, 4H), −1.92 (s, 1H, exchangeable with D2O), −2.36 (s, 1H, exchangeable with D2O) ppm; 13C NMR (125 MHz, CDCl3) δ 170.2, 157.6, 156.8, 155.0, 142.0, 141.5, 140.7, 140.4, 139.4, 139.0, 138.7, 138.4, 135.04, 134.98, 134.7, 134.4, 134.2, 132.9, 129.0, 128.7, 128.5, 128.4, 128.2, 128.1, 128.03, 127.94, 127.4, 127.3, 127.2, 127.13, 127.08, 127.0, 126.9, 125.9, 125.13, 125.05, 122.3, 119.8, 118.6, 115.6, 77.2, 50.4, 48.3 ppm; UV−vis (CH2Cl2) λmax (log ε) 404 (sh), 438 (5.14), 570 (sh), 534 (4.29), 608 (3.86), 665 (3.58) nm; UV−vis (CH2Cl2 + 10% TFA) λmax (log ε) 470 (5.18), 495 (sh), 570 (sh), 622 (sh), 679 (4.59) nm; Fl (DMF, λexcitation = λSoret) λmax 684, 743 nm, ϕ (DMF) = 0.01; HRMS (ESI+, 100% CH3CN, TOF) calcd for C94H75N14 (M· H+) 1400.6330, found 1400.6337. UV−Vis and Fluorescence M2+ Binding Titrations. For a typical UV−vis titration, 3 mL of a 6.43 μM imidazoloporphyrin solution in MeOH in 1 × 1 cm glass cuvettes were titrated with a M2+ chloride solution. The methanolic M2+ chloride solutions (1.00−2.20 mM) were added in increments of 0.2 equiv (1.75−3.86 μL), up to a maximum of 3 equiv of M2+. The overall change in volume was thus negligible (