**3.1 Electrochemical investigation of the electropolymerization**

During previous electrolyses, electrodes were systematically coated by a thin colored film, ascribed to polymers. Consequently, a more detailed study of these polymers has been investigated.

Thus, in a first report, we have published the formation of a polymer of porphyrins obtained from the electrogenerated dications of ZnOEP macrocycles substituted by a bipyridinium group in *meso*-position (zinc *meso*-bipyridiniumβ-octaethylporphyrin, abbreviated ZnOEP(bpy)+) (Fig. 3) (Ruhlmann et al., 1999b, 2008).

Fig. 3. Electropolymerization scheme of the mono-substituted ZnOEP(bpy)+.

This polymer is obtained with iterative scans by cyclic voltammetry (Fig. 4). During this iterative process, the current increases progressively, showing the formation of a conducting polymer coating the electrode. Moreover, the oxidation waves of the macrocycle become irreversible, suggesting well the reactivity of the dication. Furthermore, these oxidation waves are progressively shifted to higher anodic potential values during electropolymerization. This shift is a consequence of the electron-withdrawing effect of the viologen spacers formed between porphyrins during the polymerization process. These viologen spacers are also responsible of the two reversible reduction waves which appear

of Multisubstituted Porphyrins to the Electropolymerization of the Macrocycles 59

Fig. 5. E(ECNECB)*n*E mechanism proposed for the electropolymerization of the mono-

substituted ZnOEP(bpy)+.

and grow around 0.00 and –0.65 V/SCE (peaks a and b in Fig. 4). Indeed, the two pyridiniums of the viologen spacers are reducible in two successive steps due to their mutual interaction (Schaming et al., 2011b).

Fig. 4. Cyclic voltammograms recorded during the iterative scans between –0.90 and +1.60 V/SCE of ZnOEP(bpy)+ (0.25 mM) in 1,2-C2H4Cl2 and 0.1 M NEt4PF6. Working electrode: ITO; *S* = 1 cm2; *v* = 0.2 V s–1.

#### **3.2 Mechanism of the electropolymerization process**

The mechanism explaining the formation of polymers is similar to the one proposed for the electrosynthesis of mono-substituted porphyrins, excepted for the first oxidation step. Indeed, for electropolymerization, an E(ECNECB)*n*E mechanism can be proposed (Fig. 5), where CN corresponds to a nucleophilic attack and CB to an acid-base reaction. Firstly, after formation of the dication (two steps E), the nucleophile (bipyridinium-substituted porphyrin) can react with it in *meso*-position to give an isoporphyrin (step CN). This result correlates with the previous analysis of the redox behavior of porphyrins in the presence of nucleophiles and in particular the well documented report of Hinman *et al.* (Hinman et al., 1987). This isoporphyrin corresponds in fact to a porphyrin for which aromaticity is broken, because of the simultaneous presence of a pyridinium group and a proton on the same *meso*carbon. Afterwards, this isoporphyrin can be oxidized again (step E). Finally, this spare proton is lost, leading to the re-aromatization of the macrocycle (step CB) and allowing the formation of a di-substituted porphyrin. In a similar way, this one can react again, leading gradually to the formation of a polymer. One can also note that the isoporphyrins (which have not lost removable proton) are reduced during the further cathodic scan and conduct to the irreversible wave (peak \* in Fig. 4) around +0.10 and +0.40 V/SCE, leading to the regeneration of the porphyrin.

One can notice that if the iterative sweeps are stopped in the anodic part at a potential corresponding to the formation of the radical cation of the porphyrin, no change of the cyclic voltammograms is observed (Schaming et al., 2011b). Moreover, in that case, the oxidation wave of the macrocycle remains reversible, showing that the electrogenerated radical cation does not react further. Thus, these results show that no polymerization occurs when the radical cation is formed. While the formation of the radical cation was sufficient to perform mono-substitutions onto macrocycles, the impossibility to carry out electropolymerization in this case can be explained by a kinetic problem. Indeed, the nucleophilic attack is certainly faster onto the dication than onto the radical cation. Consequently, during

58 Electropolymerization

and grow around 0.00 and –0.65 V/SCE (peaks a and b in Fig. 4). Indeed, the two pyridiniums of the viologen spacers are reducible in two successive steps due to their

Fig. 4. Cyclic voltammograms recorded during the iterative scans between –0.90 and +1.60 V/SCE of ZnOEP(bpy)+ (0.25 mM) in 1,2-C2H4Cl2 and 0.1 M NEt4PF6. Working

The mechanism explaining the formation of polymers is similar to the one proposed for the electrosynthesis of mono-substituted porphyrins, excepted for the first oxidation step. Indeed, for electropolymerization, an E(ECNECB)*n*E mechanism can be proposed (Fig. 5), where CN corresponds to a nucleophilic attack and CB to an acid-base reaction. Firstly, after formation of the dication (two steps E), the nucleophile (bipyridinium-substituted porphyrin) can react with it in *meso*-position to give an isoporphyrin (step CN). This result correlates with the previous analysis of the redox behavior of porphyrins in the presence of nucleophiles and in particular the well documented report of Hinman *et al.* (Hinman et al., 1987). This isoporphyrin corresponds in fact to a porphyrin for which aromaticity is broken, because of the simultaneous presence of a pyridinium group and a proton on the same *meso*carbon. Afterwards, this isoporphyrin can be oxidized again (step E). Finally, this spare proton is lost, leading to the re-aromatization of the macrocycle (step CB) and allowing the formation of a di-substituted porphyrin. In a similar way, this one can react again, leading gradually to the formation of a polymer. One can also note that the isoporphyrins (which have not lost removable proton) are reduced during the further cathodic scan and conduct to the irreversible wave (peak \* in Fig. 4) around +0.10 and +0.40 V/SCE, leading to the

One can notice that if the iterative sweeps are stopped in the anodic part at a potential corresponding to the formation of the radical cation of the porphyrin, no change of the cyclic voltammograms is observed (Schaming et al., 2011b). Moreover, in that case, the oxidation wave of the macrocycle remains reversible, showing that the electrogenerated radical cation does not react further. Thus, these results show that no polymerization occurs when the radical cation is formed. While the formation of the radical cation was sufficient to perform mono-substitutions onto macrocycles, the impossibility to carry out electropolymerization in this case can be explained by a kinetic problem. Indeed, the nucleophilic attack is certainly faster onto the dication than onto the radical cation. Consequently, during

mutual interaction (Schaming et al., 2011b).

electrode: ITO; *S* = 1 cm2; *v* = 0.2 V s–1.

regeneration of the porphyrin.

**3.2 Mechanism of the electropolymerization process** 

Fig. 5. E(ECNECB)*n*E mechanism proposed for the electropolymerization of the monosubstituted ZnOEP(bpy)+.

**4.1 A first example** 

of Multisubstituted Porphyrins to the Electropolymerization of the Macrocycles 61

The way of electropolymerization of porphyrins presented above requires a first step consisting in the synthesis of the starting monomeric subunit (ZnOEP(bpy)+). The synthesis of this monomer can be performed by the electrochemical process described before, (reaction of 4,4'-bipyridine with the electrogenerated radical cation of the ZnOEP, see part 2). Even if this synthesis appears easy, the further purification of the obtained compound is more difficult, due to the possibility of multi-substitutions onto the macrocycle. In order to avoid the synthesis of this monomeric subunit, we have recently proposed an alternative method consisting in the direct use of commercial and non-substituted ZnOEP, with the

**4. A new process of Easy Polymerization Of Porphyrins (EPOP process)** 

presence of free 4,4'-bipyridine (bpy) in the solution (Fig. 7) (Giraudeau et al., 2010).

Fig. 7. Electropolymerization scheme of the non-substituted ZnOEP with free bpy.

As previously, electropolymerization can be performed with iterative scans by cyclic voltammetry (Fig. 8). The current increases again during the iterative sweeps. Compared to the cyclic voltammograms obtained in the previous case (Fig. 4), an additional reversible wave is observed at –0.30 V/SCE (peak c in Fig. 8). This additional wave can be attributed to the reduction of the bipyridinium substituents (Schaming et al., 2011b). Indeed, in this case, many bipyridinium groups can be grafted on the macrocycles, without further polymeric chains growth, because of the too important hindrance between the macrocycles. Consequently, the polymer obtained from non-substituted ZnOEP and free bpy can have many bipyridinium groups substituted onto the porphyrins (represented in blue in Fig. 7), leading to the appearance of this new reversible wave corresponding to the reduction of these bipyridinium substituents. On the contrary, when ZnOEP(bpy)+ is used to carry out

electropolymerization, the rate of the sweeps is too fast to perform nucleophilic attacks onto the radical cation, the reverse scan in the cathodic part (leading to its reduction) being too fast. Nevertheless, if an electrolysis is carried out at a potential corresponding to the formation of the radical cation, the electrode is coated, showing also the formation of oligomers and/or polymers. However, this polymeric film formation may also be due to the disproportionation of the radical cation (slow kinetic) leading to the non-oxidized ZnOEP and the porphyrin dication which can be attacked more rapidly by the nucleophile. Nevertheless, the advantage to perform electropolymerization by iterative scans instead of electrolysis at a fixed potential is double: the protons released during the polymerization can be reduced during the cathodic scans, and the iterative scans allow a better homogeneity of the polymeric film deposited onto the electrode, therefore enhancing its conductivity. Another point which can explain the need to apply a higher potential for electropolymerization concerns the degree of substitution of macrocycles. Indeed, into the polymers, porphyrins are at least substituted twice by positively charged groups. Moreover, when the chain of polymer grows, the quantity of positively charged groups increases. Consequently, porphyrins are more and more difficult to oxidize, needing an increase of the oxidative potential to perform the polymerization. This explanation is also supported by the fact that a higher potential has always been required to perform electrosyntheses of multisubstituted porphyrins by electrolysis. As a matter of fact, it can also be noted that the higher the limit potential during the anodic sweeps is, the longer the polymeric chains could be obtained. Otherwise, if the limit potential is too low, only small oligomers can be obtained.

#### **3.3 Control of the geometry of the polymer**

The use of ZnOEP(bpy)+ as monomer allows multi-substitutions onto macrocycles because three *meso*-positions remain free (positions 10, 15 and 20, Fig. 3). Thus, nucleophilic attacks can occur on each of them, and consequently that can lead to the formation of zig-zag polymers and eventually of hyper-branched polymers. In order to permit a better control of the geometry of the obtained polymers, porphyrins substituted by two protecting groups as chlorides (zinc 5-bipyridinium-10,20-dichloroβ-octaethylporphyrin, abbreviated ZnOEP(Cl)2(bpy)+) can be used (Ruhlmann et al., 1999b). In this case, linear wires of polymers are obtained because only one *meso*-position remains free. Transmission electron micrographs of this polymer show long wires having length of several micrometers and diameter of about 20 Å (Fig. 6). This is in agreement with the 19 Å molecular width of a ZnOEP molecule.

Fig. 6. (a) Transmission electron micrograph of a linear fiber obtained by electropolymerization of ZnOEP(Cl)2(bpy)+. (b) Scheme of the linear chain obtained.
