Synthesis and Photophysical Properties of a Series of Novel Porphyrin Dendrimers Containing Organoiron Complexes

Abstract

This article is focused on the design and photophysical properties of a series of novel porphyrin and metalloporphyrin dendrimers containing cationic η⁶-chloroarene-η⁵-cyclopentadienyliron(II) complexes functionalized with naphthalene and capped with ferrocene. The incorporation of cationic η⁶-chloroarene-η⁵-cyclopentadienyliron moieties into the dendrimer structures enhances solubility, and facilitates nucleophilic aromatic substitution and addition reactions due to the intense electron-withdrawing ability of the iron center. Divergent approaches were employed to yield highly symmetrical branched materials, with interesting and potentially useful properties. The preparation of these dendrimers was achieved via metal-mediated nucleophilic aromatic substitutions and Steglich esterifications. These dendrimers and their precursors were characterized through nuclear magnetic resonance spectroscopy, infrared spectroscopy, and UV–Visible and fluorescence spectroscopy. Substitution on the porphyrin core with various side chains and terminal functional groups and chromophores had little impact on the porphyrin emission properties, allowing for a wide variety of fluorescent macromolecules to be prepared which retain this characteristic emission. Also, the presence or absence of the organopyrene moieties also had little effect on the emission from these compounds. Complexation of a metal cation, such as Zn²⁺, by these compounds, did however result in a significant blue shift in the observed fluorescence; this interesting result has the potential for applications in the development of fluorescent-based cation sensors.

1 Introduction

The study of dendrimers has received significant interest in recent years [1,2,3,4,5]. Dendrimers are highly symmetric and possess well-defined nanostructures. Their synthesis involves a much larger degree of control than does that of linear polymers, and as a result, dendrimers are much more monodisperse and thus have useful and controllable materials properties. As a result of their unique physical and chemical properties, dendrimers have been the focus of many recent studies and applications [4], including those involving electron-transfer processes [5].

Porphyrins and metalloporphyrins display interesting and unique photophysical, photochemical and electrochemical properties. They play essential roles in biological activities such as photosynthesis. They have useful applications in many fields such as light-energy conversion [6,7,8], photodynamic therapy [9, 10] and other areas of photomedicine [11], third-order nonlinear optical materials [12,13,14], fluorescence switches [15] and molecular wire [16,17,18,19]. Interestingly, unique porphyrin-based supramolecular architectures have also been prepared through self-assembly [20]. Most interestingly, porphyrins have been used as the core for the preparation of dendrimers; such materials are the focus of this paper.

The first synthesis of dendrimer-based on a porphyrin core was reported in 1993 by Aida and co-workers. In their approach, a porphyrin unit was integrated into the center of poly (benzyl ether) dendrimer to form a synthetic model of a hemoprotein [21]. A year later Diederich and co-workers reported on the synthesis and redox properties of a porphyrin-based poly (ether-amide) dendrimer [22]. Two years later Suslick and co-workers prepared a series of manganese porphyrin-core poly (aryl ester) dendrimers, and tested them for their oxidation properties [23]. Since that early work on dendrimers based on porphyrin cores, other interesting examples with unique properties and features have been reported, including in the past few years [24,25,26], demonstrating the utility of the porphyrin-core dendrimer motif.

Dendrimers containing ferrocene (redox-active moieties) and porphyrin or metalloporphyrin units in their molecular structure are of great interest as hosts for anion recognition [27, 28]. The facile synthesis of such dendrimers containing diverse multifunctional groups is substantially attributed to the high control in the synthesis of these hyperbranched porphyrin cores. Most of the dendrimers with porphyrin cores have been synthesized by the introduction of dendritic substituents at the meso-positions of the porphyrin by either a convergent or a divergent approach [29]. In this paper, we demonstrate the utility of incorporating cationic cyclopentadienyliron moieties into such porphyrin-core dendrimers, as the strong electron withdrawing ability of the iron in these moieties, results in easy facilitation of the nucleophilic aromatic substitution and addition reactions [30, 31].

During the past two decades, our group has reported the synthesis of numerous examples of iron-containing macromolecules, including hyperbranched polymers with ether and ester linkages and azo dye-containing dendritic species based on upper rim functionalized organoiron metallocalix [4] arenes [32,33,34,35]. Furthermore, our group has reported the preparation of star-shaped oligomers containing cationic cyclopentadienyliron complex with azo chromophores [36]. In this paper, we bring together all of these unique aspects discussed above, by reporting the first examples of dendrimers containing in a single molecule cyclopentadienyliron, ferrocene and porphyrin units. These new types of interesting and applicable macromolecules have been fully characterized, and show interesting spectroscopic and particularly emissive properties, with potentially useful applications as cationic optical sensors, as will be discussed.

2 Experimental

2.1 Materials

The synthesis of compounds 8, 9, and 13 was achieved using a previously published methodology [30, 31]. Pyrrole was redistilled before used. All of the others reagents were purchased from Sigma-Aldrich and used without further purification. All reactions and complexes containing an η⁶-dichlorobenzene-η⁵-cyclopentadienyliron(II) hexafluorophosphate moiety were kept in the dark to prevent photochemical decomposition.

2.2 Characterizations

¹H and ¹³C NMR spectra were recorded at 400 and 101 MHz, respectively on a Varian Mercury Plus spectrometer equipped with a gradient field probe, with chemical shifts referenced to residual solvent peaks and coupling constants reported in Hz. A Varian Cary 100 Bio UV–Visible spectrophotometer was used to conduct UV–Visible absorption measurements and a Horiba Scientific QuantaMaster 400 Steady State Spectrometer was used to measure all fluorescence spectra, in both cases a standard 1 cm² quartz fluorescence cell was used.

2.3 Synthesis of 5,10,15,20-Tetrakis(4-Hydroxyphenyl) Porphyrin 1a (T(P-OH)PPH₂)

p-Hydroxylbenzaldehyde (40 mmol) was dissolved in 250 mL propionic acid, stirred vigorously, and heated to reflux. To this solution, 30 mL of nitrobenzene was added. Freshly distilled pyrrole (40 mmol) in 10 mL of propionic acid was slowly added drop wise, and the reaction mixture was heated to reflux for 30 min. The reaction mixture was left to cool and stirred overnight. The mixture was filtered and the solid product was washed repeatedly with a mixture of ethanol and propionic acid (1:1 v/v), then with hot water until the rinsing solution was no longer dark. The product was air dried and then dried at 100 °C to afford the purple crystalline product. All the Detailed information on the synthesis and the characterization is provided as Supporting Information (SI).

¹H NMR (400 MHz, DMSO-d₆) δ 9.96 (s, OH), 8.87 (s, 8H), 8.00 (d, J = 8.5 Hz, 8H), 7.21 (d, J = 8.5 Hz, 8H), − 2.87 (s, N–H). ¹³C NMR (101 MHz, DMSO-d₆) δ 157.65, 135.66, 132.38, 131.95, 120.44, 114.08.

2.4 Synthesis of T(P-OH)PPNi 1b and T(P-OH)PPZn 1c

T(P-OH)PPH₂ (0.1 mmol) was dissolved in 30 mL methanol and 10 mL CHCl₃ containing nickel(II) chloride or zinc(II) chloride (0.11 mmol). The reaction mixture was heated to reflux for 4 h at 70 °C, and then cooled to room temperature. Distilled water (60 mL) was added to the mixture, chloroform and methanol were evaporated under reduced pressure. The product was filtered to give either brownish-red crystals (1b) or purple crystals (1c).

1b: ¹H NMR (400 MHz, DMSO-d₆) δ 9.89 (br s, OH), 8.75 (s, 7H), 7.79 (d, J = 8.5 Hz, 8H), 7.13 (d, J = 8.5 Hz, 8H). ¹³C NMR (101 MHz, DMSO-d₆) δ 168.0, 153.5, 145.2, 142.5, 142.4, 129.6, 124.5.

1c: ¹H NMR (400 MHz, DMSO-d₆) δ 9.82 (br s, OH), 8.80 (s, 8H), 7.95 (d, J = 8.5 Hz, 8H), 7.17 (d, J = 8.5 Hz, 8H). ¹³C NMR (101 MHz, DMSO-d₆) δ 166.5, 151.2, 144.2, 14.4, 141.3, 129.2, 124.4.

2.5 Synthesis of 2a–c

Compound 1a, 1b, or 1c (0.1 mmol) was reacted with η⁶-chlorobenzene-η⁵-cyclopentadienyliron complex 9 (0.45 mmol) and K₂CO₃ (1.5 mmol) in 20 mL DMF at room temperature under N₂ in the dark for 72 h. The reaction mixture was precipitated into 250 mL 1.2 M HCl containing NH₄PF₆. The precipitate was filtered, washed with water, and dried under reduced pressure. The product was further purified through a neutral alumina column eluting with DCM, then ethyl acetate. Yield: 82%.

2a: ¹H NMR (400 MHz, DMSO-d₆) δ 9.06 (s, 8H), 8.41 (d, J = 7.8 Hz, 8H), 7.80 (d, J = 7.8 Hz, 8H), 6.70 (d, J = 6.2 Hz, 8H), 6.45 (d, J = 6.2 Hz, 8H), 5.46 (s, 20H), 2.48 (s, under DMSO peak), − 2.98 (br s, NH). ¹³C NMR (101 MHz, DMSO-d₆) δ 153.6, 138.7, 136.3, 131.4, 135.1 (br), 119.1, 119.0, 100.6, 86.9, 78.9, 77.6, 76.8, 19.4.

2b: ¹H NMR (400 MHz, DMSO-d₆) δ 8.98 (s, 8H), 8.22 (d, J = 7.8 Hz, 8H), 7.74 (d, J = 7.8 Hz, 8H), 6.69 (d, J = 6.2 Hz, 8H), 6.44 (d, J = 6.2 Hz, 8H), 5.46 (s, 20H), 2.48 (s, under DMSO peak). ¹³C NMR (101 MHz, DMSO-d₆) δ 153.4, 142.6, 138.7, 136.3, 131.4, 135.1, 133.4, 119.1, 119.0, 100.6, 86.9, 78.9, 77.6, 76.8, 19.4.

2c: ¹H NMR (400 MHz, DMSO-d₆) δ = 9.00 (s, 8H), 8.26 (d, J = 7.8 Hz, 8H), 7.76 (d, J = 7.8 Hz, 8H), 6.69 (d, J = 6.2 Hz, 8H), 6.45 (d, J = 6.2 Hz, 8H), 5.46 (s, 20H), 2.48 (s, under DMSO peak).

2.6 Synthesis of 3a–c

Compound 1a, 1b, or 1c (0.3 mmol) was reacted with η⁶-dichlorobenzene-η⁵-cyclopentadienyliron complex 8 (1.40 mmol) and K₂CO₃ (3.5 mmol) in 45 mL DMF at room temperature under N₂ in the dark for 72 h. The reaction mixture was precipitated into 300 mL 1.2 M HCl containing NH₄PF₆. The precipitate was filtered, washed with water, and dried under reduced pressure. The product was further purified through a neutral alumina column eluting with DCM, then ethyl acetate. Yield: 78%.

3a: ¹H NMR (400 MHz, DMSO-d₆) δ 9.06 (s, 8H), 8.41 (d, J = 8.2 Hz, 8H), 7.82 (d, J = 8.2 Hz, 8H), 7.00 (d, J = 6.6 Hz, 8H), 6.86 (d, J = 6.6 Hz, 8H), 5.46 (s, 20H), − 2.84 (br s, NH).

3b: ¹H NMR (400 MHz, DMSO-d₆) δ 8.94 (s, 8H), 8.20 (d, J = 8.4 Hz, 8H), 7.75 (d, J = 8.4 Hz, 8H), 6.96 (d, J = 6.8 Hz, 8H), 6.80 (d, J = 6.8 Hz, 8H), 5.43 (s, 20H).

3c: ¹H NMR (400 MHz, DMSO-d₆) δ 8.95 (s, 8H), 8.21 (d, J = 8.6 Hz, 8H), 7.75 (d, J = 8.6 Hz, 8H), 6.97 (d, J = 6.9 Hz, 8H), 6.81 (d, J = 6.9 Hz, 8H), 5.44 (s, 20H). ¹³C NMR (101 MHz, DMSO-d₆) δ 153.3, 142.3, 137.8, 135.6, 132.9, 132.0, 119.4, 118.1, 103.9, 87.0, 79.5, 76.9.

2.7 Synthesis of 5a–c

Compound 3a, 3b, or 3c (0.2 mmol) was reacted with 11 (0.88 mmol) and K₂CO₃ (2.5 mmol) in 45 mL DMF at room temperature, under N₂, in the dark for 72 h. The reaction mixture was precipitate into 10% HCl (300 mL) containing NH₄PF₆. The precipitate was filtered, washed with water, and dried under reduced pressure. The product was further purified through a neutral alumina column eluting with DCM containing 1–3% methanol. Yield 72%.

5a: ¹H NMR (400 MHz, DMSO-d₆) δ 9.01 (s, 8H), 8.39 (d, J = 8.3 Hz, 8H), 8.19 (d, J = 8.9 Hz, 4H), 8.07 (d, J = 7.9 Hz, 4H), 8.03 (d, J = 7.7 Hz, 4H), 7.95 (d, J = 2.1 Hz, 4H), 7.79 (d, J = 8.3 Hz, 8H), 7.61 (m, 12H), 6.72 (d, J = 6.8 Hz, 8H), 6.54 (d, J = 6.8 Hz, 8H), 5.44 (s, 20H), − 2.89 (br s, NH).

5b: ¹H NMR (400 MHz, DMSO-d₆) δ 8.87 (s, 8H), 8.34 (d, J = 8.3 Hz, 8H), 8.16 (d, J = 8.9 Hz, 4H), 8.04 (d, J = 7.9 Hz, 4H), 7.99 (d, J = 7.7 Hz, 4H), 7.90 (d, J = 2.1 Hz, 4H), 7.68 (d, J = 8.3 Hz, 8H), 7.60 (m, 12H), 6.72 (d, J = 6.8 Hz, 8H), 6.52 (d, J = 6.8 Hz, 8H), 5.42 (s, 20H).

5c: ¹H NMR (400 MHz, DMSO-d₆) δ 8.91 (s, 8H), 8.35 (d, J = 8.3 Hz, 8H), 8.17 (d, J = 8.9 Hz, 4H), 8.05 (d, J = 7.9 Hz, 4H), 8.00 (d, J = 7.7 Hz, 4H), 7.92 (d, J = 2.1 Hz, 4H), 7.72 (d, J = 8.3 Hz, 8H), 7.60 (m, 12H), 6.72 (d, J = 6.8 Hz, 8H), 6.53 (d, J = 6.8 Hz, 8H), 5.42 (s, 20H).

2.8 Synthesis of 7a and 7c

Compound 5a or 5c was placed in a 50 mL Pyrex tube and dissolved with acetonitrile. The Pyrex tube was purged with N₂ and placed in a photoreactor using a xenon light source for 6 h. The resulting solution was extracted in CHCl₃, washed with water, and dried under MgSO₄. The CHCl₃ was removed in vacuo to collect the compounds.

7a: ¹H NMR (400 MHz, DMSO-d₆) δ 8.93 (s, 8H), 8.23 (d, J = 8.5 Hz, 8H), 8.01 (d, J = 8.9 Hz, 4H), 7.93 (d, J = 8.0 Hz, 4H), 7.86 (d, J = 8.0 Hz, 4H), 7.55–7.41 (m, 28H), 7.38 (dd, J = 8.9, 2.5 Hz, 4H), 7.30 (d, J = 9.0 Hz, 8H).

7c: ¹H NMR (400 MHz, acetone-d₆) δ 8.86 (m, 8H), 8.07 (d, J = 8.2 Hz, 8H), 7.99 (m, 8H), 7.92 (d, J = 8.1 Hz, 4H), 7.83 (d, J = 8.2 Hz, 4H), 7.50 (m, 4H), 7.39–7.45 (m, 16H), 7.36 (m, 4H), 7.28 (d, J = 8.1 Hz, 8H), 7.23 (m, 4H).

2.9 Synthesis of 4a–c

Compound 3a, 3b, or 3c (0.3 mmol) was reacted with 10 (1.40 mmol) and K₂CO₃ (3.5 mmol) in 45 mL DMF at room temperature under N₂ in the dark for 72 h. The reaction mixture was precipitated into 300 mL 1.2 M HCl containing NH₄PF₆. The precipitate was filtered, washed with water, and dried under reduced pressure. The product was further purified through a neutral alumina column eluting with DCM, then ethyl acetate. Yield: 76%.

4a: ¹H NMR (400 MHz, DMSO-d₆) δ 9.01 (s, 8H), 8.37 (d, J = 7.0 Hz, 8H), 7.78 (d, J = 7.0 Hz, 8H), 7.55 (d, J = 7.6 Hz, 8H), 7.38 (d, J = 7.6 Hz, 8H), 6.70 (d, J = 5.9 Hz, 8H), 6.41 (d, J = 5.9 Hz, 8H), 5.41 (s, 20H), 4.60 (s, 8H), − 2.90 (br s, NH).

4b: ¹H NMR (400 MHz, DMSO-d₆) δ 8.90 (s, 8H), 8.18 (d, J = 8.0 Hz, 8H), 7.71 (d, J = 8.4 Hz, 8H), 7.54 (d, J = 8.0 Hz, 8H), 7.36 (d, J = 8.4 Hz, 8H), 6.64 (d, J = 7.0 Hz, 8H), 6.38 (d, J = 7.0 Hz, 8H), 5.37 (s, 20H), 4.59 (d, J = 5.5 Hz, 8H). ¹³C NMR (101 MHz, dmso) δ = 153.93, 151.88, 142.27, 140.84, 137.42, 135.50, 132.77, 130.78, 129.83, 128.67, 120.37, 119.02, 118.08, 78.01, 75.79, 74.76, 62.18.

4c: ¹H NMR (400 MHz, DMSO-d₆) δ 8.95 (s, 8H), 8.21 (d, J = 8.0 Hz, 8H), 7.74 (d, J = 8.4 Hz, 8H), 7.54 (d, J = 8.0 Hz, 8H), 7.37 (d, J = 8.4 Hz, 8H), 6.66 (d, J = 7.0 Hz, 8H), 6.40 (d, J = 7.0 Hz, 8H), 5.39 (s, 20H), 4.59 (d, J = 5.5 Hz, 8H).

2.10 Synthesis of 4d

Compounds 4a was placed in a 50 mL Pyrex tube and dissolved with acetonitrile. The Pyrex tube was purged with N₂ and placed in a photoreactor for 6 h. The resulting solution was extracted in CHCl₃, washed with water, and dried under MgSO₄. The CHCl₃ was removed in vacuo to collect the compounds.

¹H NMR (400 MHz, DMSO-d₆) δ = 8.91 (s, 8H), 8.21 (d, J = 8.16 Hz, 8H), 7.43 (d, J = 8.81 Hz, 16H), 7.36 (d, J = 8.81 Hz, 8H), 7.18 (d, J = 8.54 Hz, 8H), 7.06 (d, J = 8.54 Hz, 8H), 4.49 (d, J = 5.5 Hz, 8H).

2.11 Synthesis of 6a and 6c

Compound 4a or 4c (0.3 mmol), carboxylic ferrocene 12 (1.5 mmol), DCC (1.8 mmol), and DMAP (1.8 mmol) were stirred with 15 mL DCM and 5 mL DMF under N₂ for 24 h. The mixture was placed in a freezer to precipitate DCU. The reaction mixture was poured into 1.2 M HCl containing NH₄PF₆ and extracted into DCM. The organic layer was washed with water and dried over MgSO₄. The product was filtered and the solvent removed in vacuo. The product residue was dissolved in acetone and filtered to remove any remaining DCU. The product was precipitated into basic water to remove any excess carboxylic ferrocene. The precipitate was collected and dried under reduced pressure. Yield: 74%.

6a: ¹H NMR (400 MHz, DMSO-d₆) δ 8.96 (s, 8H), 8.33 (d, J = 8.1 Hz, 8H), 7.79 (m, 16H), 7.46 (d, J = 8.3 Hz, 8H), 6.64 (d, J = 6.6 Hz, 8H), 6.40 (d, J = 6.6 Hz, 8H), 5.31 (s, 20H), 4.79 (t, J = 2.0 Hz, 8H), 4.50 (t, J = 2.0 Hz, 4H), 4.16 (s, 20H).

6c: ¹H NMR (400 MHz, DMSO-d₆) δ 8.87 (s, 8H), 8.15 (d, J = 8.1 Hz, 8H), 7.68 (m, 16H), 7.43 (d, J = 8.3 Hz, 8H), 6.60 (d, J = 6.6 Hz, 8H), 6.39 (d, J = 6.6 Hz, 8H), 5.32 (s, 20H), 4.79 (t, J = 2.0 Hz, 8H), 4.50 (t, J = 2.0 Hz, 4H), 4.16 (s, 20H).

2.12 Synthesis of 14a

Compound 4a (0.4 mmol), valeric bimetallic complex 13 (1.7 mmol), DCC (2.0 mmol), and DMAP (2.0 mmol) were stirred with 20 mL DCM and 10 mL DMF under N₂ for 24 h. The mixture was placed in a freezer to precipitate DCU. The reaction mixture was filtered and poured into 1.2 M HCl containing NH₄PF₆. The precipitated product was filtered and the residue was dissolved in acetone and filtered to remove any remaining DCU. The product was precipitated again into water, collected, and dried under reduced pressure. Yield: 72%.

¹H NMR (400 MHz, DMSO-d₆) δ 8.40 (br s, 8H), 7.78 (br m, 8H), 7.60 (br m, 8H), 7.43, (br m, 8H), 7.40 (br m, 24H), 7.28 (br m, 16H), 6.76(br m, 16H), 6.70 (br m, 8H), 6.43 (br m, 24H), 5.28 (br s, 20H), 5.19 (br s, 40H), 5.13 (br s, 8H), 2.51 (br m, 8H), 2.27 (br m, 8H), 1.71 (br s, 12H), − 2.87 (br s, NH). ¹³C NMR (101 MHz, DMSO-d₆) δ 26.9, 29.7, 35.9, 45.0, 64.9, 75.3, 75.9, 76.3, 77.9, 79.3, 86.8, 103.6, 118.7, 119.1, 119.9, 120.1, 120.5, 129.2, 129.3, 129.9, 130.2, 130.5, 131.9, 134.1, 136.3, 138.7, 146.0, 146.2, 151.1, 153.2.

3 Results and Discussion

η⁶-Dichlorobenzene-η⁵-cyclopentadienyliron complexes have been shown to readily undergo nucleophile aromatic substitution reactions under mild conditions [30, 31]. This is the motivation for the current incorporation of these types of organoiron complexes into the porphyrin-based frameworks, as this allows for a versatile method for producing functionalized porphyrins, analogous to that of other multibranched polymers and polymer stars based on other cores [32,33,34,35,36].

The preparation of porphyrins and metalloporphyrins based on η⁶-dichlorobenzene-η⁵-cyclopentadienyliron is interesting as the presence of the terminal chloro groups on the organoiron complex allows for the nucleophilic aromatic substitution reaction to continue to further functionalize the prepared porphyrin, allowing for the construction of dendrimers of varying generation. The presence of the phenolic groups on the porphyrin and metalloporphyrins rings further allows for the addition of cationic organoiron groups to the molecule. The η⁶-monochlorobenzene-η⁵-cyclopentadienyliron complex 9 was reacted with free-base porphyrin 1a (5,10,15,20-tetrakis(4-hydroxyphenyl) porphyrin) or with metalloporphyrins 1b or 1c in DMF, with an excess amount of potassium carbonate as a base to give porphyrins 2a–c containing four cationic organoiron moieties (Scheme 1). This reaction was carried out at room temperature, yielding deep purple colored materials. All of the detailed information on the synthesis and the characterization is provided as Supporting Information (SI). Under the same reaction conditions porphyrins 3a–c were synthesized to produce the first step in making the target porphyrin-based dendrimers containing cationic iron moieties, ferrocene, and naphthalene (Scheme 1). The ¹H NMR spectrum shows that the organoiron unit was successfully attached to the porphyrin core, as evidenced by the appearance of the cyclopentadienyl ring (Cp) peak at 5.46 ppm as well as all downfield shift of the phenyl and pyrrole protons (Figure S6 and S8) (SI).

Scheme 1

Synthesis of free-base porphyrin and metalloporphyrins 2a–c and 3a–c

The HSQC spectrum of 2a (Fig. 1) further confirms the substitution of the iron complex onto the porphyrin core. The protons of the methyl group overlap with the DMSO solvent peak at 2.50 ppm, but nonetheless correlate well to the methyl carbon peak at 19.4 ppm. The phenyl carbons resonate at 119.1-138.7 ppm (ortho- and meta-C-Ph) and correlate to protons that resonate at 7.80–8.41 ppm respectively. The Cp protons show a correlation to a carbon which resonates at 77.6 ppm.

Fig. 1

HSQC spectrum of porphyrin 2a

The presence of the cyclopentadienyliron moiety on the complexed arene allowed for the chlorine group to readily undergo further nucleophilic aromatic substitution reactions in these relatively mild conditions. The incorporation of naphthalene into a free-base porphyrin or a metalloporphyrin results in the preparation of photoactive materials with wide potential applications in different fields such as organic light-emitting diodes (OLEDs), light harvesting, or switches. Chloro-terminated free-base porphyrin 3a and metalloporphyrin complexes 3b and 3c were reacted with 2-hydroxynaphthalene (11) in DMF in the presence of a weak base like potassium carbonate to give porphyrin dendrimers 5a–c containing four cationic organoiron moieties (Scheme 2). The relatively weak metal-aryl coordination bonds between the cyclopentadienyliron and the arene ring in the organoiron cation groups can be readily cleaved using UV light at a wavelength of 300 nm in a strongly coordinating solvent such as acetonitrile. Solutions of dendrimers 5a and 5c in CH₃CN were exposed to 300 nm light for 2 h to produce metal-free dendrimers 7a and 5c (Scheme 2). After cleavage of the cationic cyclopentadienyliron moieties, the solubility of the dendrimers in non-polar organic solvents was found to be drastically increased. The ¹H NMR data of dendrimer 7a shows that the resonance at 5.44 ppm from the cationic cyclopentadienyliron moiety has completely disappeared, confirming the removal of the iron arene moiety.

Scheme 2

Synthesis of free-base porphyrin and metalloporphyrins 4a–c, 5a–c, 6a, 6c, 7a and 7c

The reaction of the previously prepared free-base porphyrin and metalloporphyrin dendrimers 3a–c with 4-hydroxybenzyl alcohol (10) in the presence of a weak base gave dendrimers that possess terminal primary aliphatic alcohols 4a–c (Scheme 2).

The ¹H NMR spectrum of 4b (Fig. 2) clearly shows the successful incorporation of the 4-hydroxybenzyl alcohol, as evidenced by the presence of the benzyl methylene group doublet at 4.59 ppm. To study the effect of cyclopentadienyliron moieties on the fluorescence and absorption spectra of these compounds, solutions of dendrimer 4a in CH₃CN were exposed to 300 nm light for 6 h to give the de-metallated dendrimer 4d, i.e., the corresponding dendrimer without the presence of the cationic cyclopentadienyliron moieties.

Fig. 2

400 MHz ¹H NMR spectrum of metalloporphyrin dendrimer 4b

The terminal aliphatic alcohol groups of both the free-base porphyrin and the metalloporphyrin dendrimers 4a and 4c were esterified using ferrocene carboxylic acid. Carboxylic acid ferrocene is readily synthesized by Friedel–Crafts acetylation, followed by oxidation with iodine. The ease and afforability of its synthesis make carboxylic acid ferrocene a valuable compound, and its use allows for the incorporation of ferrocene into both free-base porphyrins and metalloporphyrins. The reaction of ferrocene carboxylic acid (12) with the terminal aliphatic alcohol groups of free-base porphyrin and metalloporphyrin dendrimers 4a and 4c in the presence of DCC/DMAP resulted in the preparation of the ferrocene-capped free-base porphyrin and metalloporphyrin dendrimers 6a and 6c (Scheme 2).

The ¹H NMR data confirmed the successful synthesis of the ferrocene-capped porphyrin and metalloporphyrin dendrimers 6a and 6c, as indicated by the presence of a new Cp proton resonance at 4.16 ppm arising from non-functionalized Cp ring of ferrocene. There are also two resonances from the functionalized Cp ring at 4.79 ppm and 4.50 ppm. The benzylic CH₂ resonance was also shifted close to the cationic iron Cp at 5.32 ppm due to the increased electron-withdrawing nature of the ester, while the cationic iron Cp resonated at 5.34 ppm. The dendrimers 6a and 6c were orange in colour and were highly soluble in organic solvents as a result of the incorporation of the ferrocene moiety into the dendrimer structure.

Additional metallic moieties can be attached to the dendritic porphyrins in order to increase the number of metal species in the molecule. For this purpose, the free-base porphyrin 4a was reacted with four equivalents of the valeric bimetallic complex (13) to yield the free-base porphyrin dendrimer 14a containing twelve metal centres (Scheme 3). The formation of the second generation of the metallic unit was confirmed by the presence of two different Cp resonances at 5.28 and 5.19 ppm which, when integrated, show a 2:1 ratio, as there are forty and twenty cyclopentadienyliron protons in these two different electronic environments.

Scheme 3

Synthesis of free-base porphyrin and metalloporphyrins 14a

3.1 Absorption and Fluorescence Spectroscopy

The absorption spectra obtained for the free-base porphyrin 2a, Ni-porphyrin complex 2b, and Zn-porphyrin complex 2c (Table 1; Fig. 3) showed significant similarity to the absorption spectrum of the corresponding free or metal-complexed porphyrin cores 1a, 1b, and 1c. However, a new absorption band appeared at 309 nm, attributed to the incorporation of the cyclopentadienyliron complex onto the porphyrin core. The metal ions in the metalloporphyrins behave as Lewis acids and accept lone pairs of electrons from the dianionic porphyrin ligand. Unlike most transition metal complexes, their color is due to electronic transitions within the porphyrin ligand involving the excitation of electrons from π to π* porphyrin ring orbitals. The change in the spectrum, specifically the fewer peaks in the case of metalloporphyrins, is a result of the increased symmetry of the metallated porphyrin compared to the free-base porphyrin. The two amino hydrogens on present in the free-base porphyrin reduce the ring symmetry from square, in case of metalloporphyrins, to rectangular.

Table 1 UV–Visible data of the compounds in DMF (1.5 × 10⁻⁵ M)

Fig. 3

Absorption spectra of compounds 2a–c in DMF

There is significant metal to ligand π-backbonding occurring via the π* orbital of the porphyrin in the Ni-porphyrin complex, which increases the porphyrin π* orbital energy, resulting in an increased porphyrin π–π* energy gap. This causes the electronic absorptions to undergo a slight blue shift relative to the free base porphyrin. In contrast, the Zn-porphyrin complex contains metal d–π orbitals that are relatively low in energy, having very little effect on the porphyrin π–π* energy gap in the electronic spectrum of the porphyrin. As a result, the electronic absorptions of the Zn-porphyrin complex occur at lower energies and therefore the absorption spectrum undergoes red shifts relative to the free base porphyrin.

Fluorescence measurements were performed on both free-base porphyrin 2a and Zn-porphyrin complex 2c. All of the fluorescence bands were blue-shifted compared to that of porphyrin cores 1a and 1c (Table 2; Figs. 4, 5). The most likely explanation for this observation is that the incorporation of the iron complex decreases the electron density in the porphyrin macrocycle, which raises the energy gap for electronic transition, resulting in a blue shift in the fluorescence bands. More significantly, there is a very large blue shift in the emission band of this dendrimer upon Zn²⁺ complexation, from 653 to 606 nm.

Table 2 Fluorescence data of the compounds in DMF (1.5 × 10⁻⁵ M)

Fig. 4

Fluorescence spectra of free-base porphyrin 2a excited at 420 nm

Fig. 5

Fluorescence spectrum of Zn-porphyrin 2c excited at 420 nm

The incorporation of the dichlorobenzene-cyclopentadienyliron moiety 8 into the porphyrin core, which is indicated in the absorption spectrum by bands in the range of 203–268 nm, led to free-base porphyrin 3a and porphyrin dendrimers containing nickel (3b) and zinc (3c), all of these exhibiting similar absorption spectra to the corresponding complexes 1a, 1b, and 1c. The introduction of electron withdrawing groups (complex 9 or 8) to the para-phenyl positions has been shown to lower the energy of the b1 orbital by decreasing the amount of electron density in this orbital, leading to an increase in the energy of transition and hence blue shifted bands compared to the porphyrin cores (Table 1; Fig. 6). However, the presence of the Cl versus CH₃ substituent had only a minor effect on the emission spectrum, with just a slight blue shift of 2 nm for 3a and 3c compared to 2a and 2c. Again, there is a very large blue shift in the emission of this compound upon Zn²⁺ complexation, from 651 to 604 nm (Fig. 7).

Fig. 6

Absorption spectra of compounds 3a–c in DMF

Fig. 7

Fluorescence spectrum of Zn-porphyrin 3c excited at 420 nm

The incorporation of naphthalene moieties (11) into the porphyrin complexes 3a and 3c led to free-base porphyrin dendrimer 5a and Zn-porphyrin dendrimer 5c. These compounds were also UV-photolyzed in order to cleave the cyclopentadienyliron moiety and produce the corresponding organoiron-free free-base porphyrin dendrimer 7a and Zn-porphyrin dendrimer 7c, both with similar absorption spectra to those of 3a and 3c, respectively (Table 1; Fig. 8). Thus, the presence or absence of the organoiron moieties has no measurable effect on the fluorescence properties of these compounds, and it is clear therefore that the iron does not quench the porphyrin emission. One difference between 7 and 3 however is the clear presence of the absorption band of naphthalene, in the range of 290–350 nm, which can be clearly seen in Fig. 8. Again, as was seen in the case of compounds 2 and 3, there is a large blue shift in the emission of this compound upon Zn²⁺ complexation, in this case from 652 nm for compound 7a to 604 nm in the case of compound 7b.

Fig. 8

Absorption spectra of compounds 7a and 7c in DMF

The presence of the naphthalene moieties in the case of the free-base dendrimer 7a and its zinc counterpart 7c has very little impact on the fluorescence spectrum when the compounds are excited at 420 nm (porphyrin absorption), compared to the spectra of 3a and 3c. There is no evidence of naphthalene emission using this porphyrin-based excitation wavelength. Similar to the absorption spectra, the change in the energies of the fluorescence bands was observed for free-base dendrimer 7a. The free-base dendrimer fluoresced at 652 nm and 713 nm (Table 2; Fig. 9). The Zn-complexed dendrimer 7c however showed a strong, new band at 610 nm, as shown in Fig. 9, which was not observed in the case of the free base 7a, or in any of the other compounds for that matter. This is a potentially useful result, as this compound 7c could find applicability as a metal cation sensor, with the appearance of this strong fluorescence at 610 nm upon complexation. This could be achieved by monitoring the fluorescence intensity of a solution of dendrimer 7c; the intensity of the signal would be correlated to the presence and concentration of Zn²⁺ in the solution. Recent reports of other examples of such metal-cation optical sensors have demonstrated their utility in various applications [37, 38]. Again, as in the case of the previous compounds as described above, there is a very large blue shift in the emission of this compound upon Zn²⁺ complexation, in this case from 652 to 604 nm.

Fig. 9

Fluorescence spectra of free-base porphyrin 7a and zinc porphyrin dendrimer 7c

The incorporation of 4-hydroxybenzyl alcohol (10) moieties into the previously prepared porphyrin complexes 3a and 3c led to the preparation of the free-base porphyrin dendrimer 4a and its corresponding zinc-based metalloporphyrin dendrimer 4c, which exhibited slightly red-shifted fluorescence spectra (again, just on the order of 2 nm) compared to those of 3a and 3c (Tables 1, 2, Figs. 10, 11). This slight red shift in the absorption and fluorescence bands after the introduction of 4-hydroxybenzyl alcohol moieties can be explained as being a result of the hydroxyl group donating electron density to the aromatic ring, strengthening the resulting electronic density on the phenyl ring. Therefore, the phenyl ring in this case is conjugated with the porphyrin macrocycle, which results in a reduced electron transition energy of the porphyrin macrocycle, resulting in the observed red-shifted peaks in the absorption spectrum. As in the case of the previous compounds, there is a large blue shift in the emission of this compound upon Zn²⁺ complexation, in this specific case from 653 to 609 nm.

Fig. 10

Absorption spectra of compounds 4a and 4c in DMF

Fig. 11

Fluorescence spectra of compounds 4a and 4c

To investigate the effect of the cyclopentadienyliron moieties on the fluorescence and absorption spectroscopy of these compounds, the fluorescence and absorption measurements of free-base porphyrin dendrimer 4d were performed. Interestingly, the fluorescence and absorption spectra were exactly the same as those for the corresponding compound containing the cyclopentadienyliron groups, indicating that the electron transition energy of the porphyrin macrocycle is not measurably affected by the demetallation process (Figs. 12, 13).

Fig. 12

Absorption of 4a and demetalated porphyrin 4d in DMF

Fig. 13

Fluorescence spectra of porphyrin dendrimers 4a and 4d

Finally, the incorporation of the valeric bimetallic complex 13 and the carboxylic acid ferrocene 12 into the previously prepared porphyrin complex 4a led to the isolation of free-base porphyrin dendrimers 6a and 14a, with no observable shift in the absorption and fluorescence spectra (Tables 1, 2).

4 Conclusions

Novel free-base porphyrins, nickel and zinc porphyrin dendrimers containing cationic η⁶-chlorobenzene-η⁵-cyclopentadienyliron(II) complexes and cationic η⁶-methylbenzene-η⁵-cyclopentadienyliron(II) complexes functionalized with naphthalene and capped with ferrocene were synthesized and characterized. The incorporation of cationic η⁶-dichloroarene-η⁵-cyclopentadienyliron moieties into these dendrimeric structures enhanced the solubility of the dendrimer and facilitated further nucleophilic aromatic substitution and addition reactions, as a result of the strong electron-withdrawing ability of the iron center. However, the results found here show that the presence of the organoiron groups has no significant effect on the fluorescence properties of these compounds, and it is concluded that the iron cations do not quench the porphyrin emission. This is important for the potential application of these compounds as fluorescent sensors. Divergent approaches were also employed to give highly symmetrical branched materials.

Photophysical studies of these dendrimers showed that the incorporation of various terminal methyl group cyclopentadienyliron complexes led to a small blue shift in the absorption spectra on the order of ~ 3 nm, while the incorporation of the terminal chloro group cyclopentadienyliron complex led to a slightly larger blue shift in the absorption spectra on the order of ~ 6 nm. This shift is most likely due to the incorporation of the iron complexes decreasing the electron density in the porphyrin macrocycle, which thereby raised the energy of electron transition, leading to a blue shift in the absorption and fluorescence spectra. On the other hand, the presence of the electron-donating group, naphthalene, resulted in no significant shift of either the absorption or fluorescence spectrum (≤ 1 nm), indicating the negligible impact of the naphthalene moiety on the electronic properties of the porphyrin macrocycle.

In all cases, addition of various chains and terminal groups to the porphyrin core did not significantly change the overall fluorescence properties of these series of compounds, with only slight spectral shifts on the order of 6 nm or less, as described above. In fact, in the case of the wide range of porphyrin dendrimers synthesized and characterized, the emission maximum of the free base form was found to be in the range of 651–653 nm in all cases. This opens up the synthetic possibilities for these types of compounds, allowing for the synthesis of a wide range of porphyrin-core dendrimers which retain the desired fluorescence properties. Most importantly, all of these compounds showed a significantly large blue shift of the emission (on the order of close to 50 nm!) upon complexation of Zn²⁺ cations, opening up the possibility of the utilization of these compounds as optical sensors for metal cations.