The structural investigation of the catalytically active aggregates takes place via atomic force microscopy (AFM) and small angle neutron scattering (SANS).
Under strong acidic conditions (pH < 2) the two inner nitrogen bases of a metal-free porphyrin ring become protonated leading to a hexavalent species in the case of cationic tetraphenylporphyrins accompanied by a strong colour change from red to green [
83]. The valency increase from four to six is accompanied by an increase in symmetry from D
2H to D
4H. Therefore, the planarity of the macrocycle of TMPyP and TAPP disappears and the metal-free porphyrins have the same symmetry as for example Zn-TMPyP. In the case of TMPyP, both, a pentavalent and a hexavalent porphyrin, can be received depending on pH. For the TMPyP monoacid pH 1.7 is sufficient whereas the diaicd is generated by dissolving the porphyrin directly in 1 M hydrochloric acid,
i.e., at pH = 0. Going from the tetravalent free-base porphyrin to the pentavalent TMPyP monoacid leads to a decrease of symmetry from D
2H to C
2V. This one additional charge has a tremendous effect on the catalytic activity as will be discussed in the following.
3.1. Atomic Force Microscopy (AFM)
In the AFM images shown in
Figure 1, a loading ratio
l = 0.4 was chosen for comparison, as this turned out to be the ratio with the highest catalytic activity under neutral conditions [
31]. Structural differences are evident for the aggregates formed by the different porphyrin species with PSS brush.
Figure 1a,b displays network-like structures formed by TAPP diacid and PSS brush, which were already found under neutral conditions [
26,
30]. Under acidic conditions these networks exhibit smaller dimensions in length (up to 645 nm) and even smaller dimensions in height (4 to 5.5 nm) as compared to former studies under neutral conditions that showed TAPP-PSS brush assemblies several μm in length and up to 15 nm in height [
26].
To investigate the influence of each additional charge in the case of TMPyP systematically, the structure of TMPyP-PSS brush samples under neutral conditions was investigated (
Figure 1g,h). Under neutral conditions, TMPyP-PSS brush forms well-defined, network-like structures which are 21 nm in height and exhibit also slightly larger meshes. In contrast, TMPyP monoacid (
Figure 1e,f) forms rather undefined networks with several μm size which are up to 64 nm high. Hence, one additional charge causes a transition from well-defined network-like structures to rather undefined networks. Increasing the number of charges further so that the TMPyP diacid (
Figure 1c,d) is present, again well-defined network-like structures result. They exhibit a height of 32 nm. Considering the PSS brush structures in the obtained aggregates more closely, one can see that the diameter of the PSS brush is up to twice as large in TMPyP diacid-PSS brush aggregates (
Figure 1d) than under neutral conditions (
Figure 1h). For the TAPP-PSS brush, one can see that there is nearly no difference in PSS brush width in the two considered pH regions. Furthermore,
Figure 1i,j displays network-like structures for Zn-TMPyP and PSS brush under neutral conditions which are 17 nm in height and several hundreds of nm in size and which are therefore very similar to structures formed with TAPP and TMPyP under equivalent conditions (
Figure 1g,h).
Several reasons could be responsible for the formation of well-defined or rather undefined structures with the different TMPyP species. The even or odd number of charges may be one reason for this because the PSS brush molecules can distribute less regularly with only one charge in the porphyrin ring. Increasing the number of charges from five to six, networks with TMPyP diacid (
Figure 1d) (32 nm) are half as high as those with TMPyP monoacid (64 nm) and much more defined. Structural differences between TMPyP monoacid and TMPyP diacid can derive from their different symmetry (C
2V D
4H). In addition, differences of the porphyrins TMPyP and TAPP can originate from the different ionic strengths. pK
a values for TAPP diacid and for TMPyP diacid are 3.6 and 1.4, respectively, indicating that TMPyP is the most acidic and the ionic strength is much larger in the TMPyP diacid sample than in the TAPP diacid sample. Usually, high ionic strength leads to a decrease in porphyrin-porphyrin charge repulsion due to screening and therefore the π-systems of two porphyrin macrocycles are more prone to interact intermolecularly. One might expect that the networks of TMPyP diacid-PSS brush exhibit tighter meshes due to the higher ionic charge, but it is in fact the opposite. The reason for this is not directly evident. High ionic strength was also investigated for TAPP diacid-PSS brush by dissolving both also directly in 1 M hydrochloric acid. The sample precipitates immediately and investigation with AFM was not possible. At higher ionic strength, screening of the electrostatic forces takes place leading to a stronger aggregation tendency. The structural investigation by AFM showed that TAPP diacid in combination with PSS brush also forms network-like structures similar to those under neutral conditions. In contrast, TMPyP diacid assembles into “huge” broader meshed networks and TMPyP monoacid makes larger undefined structures with PSS brush.
3.2. Structural Investigation by Small Angle Neutron Scattering (SANS)
To gain further insight into the structure and the shape of the aggregates in solution, small angle neutron scattering (SANS) measurements for each system were performed at polyelectrolyte concentrations of c (PSS brush) = 1 g·L
−1, except for the TMPyP diacid sample where the concentration was c (PSS brush) = 0.05 g·L
−1. Data are reported in
Figure 2a. Scattering curves (
Figure 2a) give evidence of the cylindrical shape of the PSS brush as the slope in a log/log representation for the intermediate
q-range is −1.09, according to the scaling of the form factor
P(q) with
q−1 for long rods. From the first point or from the point where a plateau can be seen, the minimum length or approximate length of the cylinder is found via
l = 2π/
qmin which is approximately 250 nm. To gain more information, the curves have been analyzed by Guinier analysis (
Figure 2b). The linearity of a cross-section Guinier plot confirms the cylinder shape. From the slope, the cross-section radius of gyration
RGc, can be obtained, which for the PSS brush is
RGc = 4.8 nm. Assuming a homogeneous structure, this
RGc can be converted into a cross-section radius and consequently into a diameter, which is 13.8 nm, in good agreement with former studies [
26,
30]. To see how the porphyrin influences the PSS brush in the aggregates, samples with the same loading ratio
l were investigated for each porphyrin.
In each case, cylindrical nanoassemblies were found.
Table 1 shows the cylinder lengths,
RGc and radii. On the basis of this first analysis, the experimental curves have been fitted according to structural models to obtain the particle shape and dimensions. Results are given in
Figure 2a (solid lines) and
Table 1. It can be seen that all the determined radii are smaller than the one of the PSS brush alone, while lying all in the same range. TMPyP diacid-PSS brush differs, as with 12 nm the radius is nearly twice as large as that of the PSS brush. The result fits well with the observations during sample preparation as precipitation occurs immediately for polyelectrolyte concentrations of
c (PSS brush) = 1 g·L
−1, which is why a distinct smaller concentration has to be used.
Hence, SANS results showed that cylindrical aggregates are formed, which is consistent with AFM where individual strands of the networks exhibit a cylindrical shape. The lengths of these cylinders range from 93 to 2700 nm. Differences in length are also evident from the AFM images: for the TMPyP diacid-PSS brush sample, large network-like structures were found, the single-strand dimensions of which are given in comparison are the longest and widest. The largest radius for the cylinders formed by TMPyP diacid and PSS brush is in good agreement to the corresponding AFM. An exception is the TMPyP monoacid, where structures found in AFM are quite undefined. All the diameters are smaller than that of TMPyP diacid and also the determined lengths are all distinctly smaller. Hence, overall SANS and AFM results are in very good agreement.
The observation that, with the exception of TMPyP daicid, the radii are all smaller than those of PSS brush in a broader view agrees very well with observations from former studies, where SANS measurements under neutral conditions with and without salt were performed [
62,
66]. It was observed that PSS brush had the largest radius followed by TAPP-PSS brush and TMPyP-PSS brush. Thus the shrinkage herein can derive from the larger number of charges of the porphyrin, which causes the porphyrin to enter more into the inside of the PSS brush, as the PSS brush has more power to bind it due to the additional two charges and therefore smaller diameters result. Former studies pointed out that side-chain interconnections seem to be responsible for the smaller diameters of the porphyrin-PSS brush assemblies under neutral conditions, [
30] which also is the case for TAPP diacid and TMPyP monoacid. The behavior of TMPyP diacid, in contrast, appears to be different and is not understood yet.
3.3. Spectroscopic Investigation
The difference in valency becomes evident spectrochemically as can be seen in
Figure 3. Here, samples with and without polyelectrolyte under neutral and acidic conditions are investigated for the three different porphyrins.
Figure 3a–c exhibit the complete spectrum, whereas
Figure 3d–f focus on the enlarged Q-bands. Under neutral conditions, the Soret band of pure TAPP (
Figure 3a) is red-shifted for TAPP-PSS brush aggregates with
l = 0.4, indicating the formation of J-aggregates and a head-to-tail interaction of the transition dipole moments. Changing the pH from neutral to acidic, the TAPP Soret band undergoes a bathochromic shift indicating the formation of J-aggregates. This bathochromic shift with 9 nm under neutral conditions is slightly larger than under acidic conditions with 6 nm indicating larger interactions between the TAPP and the PSS brush or the formation of larger J-aggregates under neutral conditions. This pH change is, as already mentioned, accompanied by an increase of symmetry, which again is accompanied by a reduction of the number of Q-bands from four to two. The presence of PSS brush leads to a small further red-shift of the TAPP diacid Soret band (
Figure 3b). The Q-bands undergo no spectral shift upon combination with the polyelectrolyte under neutral conditions, while this is different under acidic conditions. Here, a slight bathochromic shift of the Q-bands in the presence of PSS brush can be seen. Spectral data from
Figure 3 are summarized in
Table 2. In the case of TMPyP (
Figure 3c), no band shifts of the Soret band from the neutral TMPyP to TMPyP monoacid can be observed. The presence of PSS brush in neutral conditions leads to a small change of the Soret band characteristics but not to a band shift. For the TMPyP diacid, a clear bathochromic shift of the Soret band can be observed indicating the formation of J-aggregates and a head-to-tail interaction of the transition dipole moments as for TAPP diacid, which in the presence of PSS brush is slightly more expressed. For TMPyP monoacid, the presence of PSS brush leads to a red-shift of the Soret band and for the illustrated charge ratio to a band splitting, indicating that more than one dominant species exists. In addition, in the Q-region, some spectral shifts occur (
Figure 3d). For the neutral TMPyP samples, a slight red-shift of the Q-bands with polyelectrolyte addition can be seen. This is also the case for the TMPyP diacid. For Zn-TMPyP (
Figure 3e) less spectral changes are expected, because the metal center prevents the increase of charge and also symmetry changes are not expected. Thus, only for pure Zn-TMPyP Soret band can a slight blue shift by changing the milieu from neutral to acidic be observed, thereby indicating the formation of H-aggregates and a face-to-face interaction. However, the corresponding Q-bands show shifts (
Figure 3f).
For the pure Zn-TMPyP in solution, the first Q-band rises at 519 nm, the second at 563 nm and, additionally, a weak shoulder at higher wavelengths is present. In the presence of the PSS brush under acidic conditions, four Q-bands can be seen, which are almost at the same wavelengths as TMPyP under neutral conditions, indicating that Zn-TMPyP becomes demetallated under such acidic conditions.
The spectroscopic investigation therefore showed that the structure formation of porphyrin and polyelectrolyte always leads to a red-shift indicating the formation of J-aggregates. For the TMPyP monoacid, additionally a band splitting can be observed indicating the presence of more than one species, which fits very well to the rather undefined structures in AFM.
3.4. Catalysis
Recently we found that at pH = 7 porphyrin-polyelectrolyte assemblies catalyze the light-induced oxidation of iodide in aqueous solution more effectively than the unassociated porphyrin [
67]. Here, we performed a study of the catalysis of different porphyrins under strongly acidic conditions. The free-base porphyrins are two-fold protonated, resulting in two additional charges,
i.e., hexavalent porphyrins. Due to the fact that electrostatic interactions between the positively charged porphyrins and the negatively charged polyelectrolytes are responsible for structure formation and the structure formation itself under neutral conditions has caused a higher catalytic performance, two additional charges are expected to have a substantial influence on the catalytic activity. Therefore, porphyrin diacid-polyelectrolyte assemblies are promising for catalysis especially under strong acidic conditions where a variety of systems cannot be used. For the main part of the catalysis study, three different porphyrins, TAPP, TMPyP and Zn-TMPyP, are each combined with a cylindrical PSS brush and the catalytic activity of these aggregates was compared with the one of porphyrin only in acidic solution. In addition, measurements were also done with linear PSS to identify the role of the polyelectrolyte architecture.
The chosen model reaction is the light-induced oxidation of iodide into triiodide, because I
−/I
3− is used, for example, in solar cell applications. The generation of triiodide can be monitored through two characteristic absorption bands at 287 and 353 nm in the UV/Vis spectrum. The development of the triiodide absorption as a function of time for the different systems is plotted in
Figure 4. Triiodide concentrations are summarized in
Table 3. The system of TAPP diacid and PSS brush is exemplarily chosen and the influence of the loading ratio
l on the catalytic activity was investigated over a large
l regime 0.01 ≤
l ≤ 2. As can be seen from
Figure 4a, the catalytic activity of TAPP diacid-PSS brush assemblies increases successively with an increasing amount of polyelectrolyte: that is, the concentration of triiodide increases from
l = 2 successive up to
l = 0.03 where the highest catalytic activity can be observed. Further increase of polyelectrolyte leads to a smaller catalytic activity, which shows that
l = 0.03 is the optimum loading ratio which is necessary for an improvement of the catalytic activity. At
l = 0.01, the amount of TAPP is too small to allow for building sufficient aggregates. Between 0.3 ≤
l ≤ 0.6, no differences can be seen. Similar observations can be made for the TMPyP monoacid-PSS brush system, where the considered regime was 0.1 ≤
l ≤ 0.5. As can be seen in
Figure 4b, a distinct difference in the catalytic activity with and without polyelectrolyte can be observed. The highest catalytic activity was found for
l = 0.1. For the Zn-TMPyP-PSS brush system, the results of which are illustrated in
Figure 4c, samples with 0.1 ≤
l ≤ 0.8 were investigated. Similar to the results for the TMPyP monoacid-PSS brush system, a distinct increase in catalytic activity through the polyelectrolyte can be seen.
Hence, the activity increases with increasing amount of polyelectrolyte so that for
l = 0.1 the highest amount of triiodide was found. For the TMPyP diacid-PSS brush system, the observations are different as can be seen in
Figure 4d. The pure TMPyP diacid solution is the catalytically most active one of the investigated porphyrin only solutions. With polyelectrolyte, the catalytic activity is the same as without polyelectrolyte. Among the considered loading ratios,
l = 0.1 shows the highest catalytic activity, but the activity does not continuously increase with increasing amount of polyelectrolyte. Among the considered systems, TMPyP diacid-PSS brush assemblies with c (I
3−) = 4.2 × 10
−4 mol·L
−1 generate the highest concentration of triiodide followed by TAPP diacid with c (I
3−) = 1.5 × 10
4 mol·L
−1, Zn-TMPyP with c (I
3−) = 8.7 × 10
−5 mol·L
−1 and TMPyP monoacid with c (I
3−) = 7.8 × 10
−5 mol·L
−1, as can be seen in
Table 3.
These observations are also evident from the corresponding turnover numbers (TON) and the turnover frequency (TOF), which are summarized in
Table 4. TON describes the amount of triiodide which can be generated with the chosen porphyrin concentration. TOF is the turnover per time. The catalytic activity of TMPyP diacid-PSS brush assemblies is evident with a TON of 109 and a TOF of 1.82, which are higher than those of pure TMPyP diacid and essentially larger than those of the other systems. Thus, the catalytic activity of TAPP diacid, TMPyP monoacid and Zn-TMPyP can be obviously enhanced with regard to triiodide generation in the presence of polyelectrolyte.
From the results it can be seen that one additional charge in the case of TMPyP enhances the catalytic activity tremendously. The effect of the porphyrin diacids is much more significant than with the porphyrin under neutral pH conditions [
31]. Under acidic conditions, the triiodide generation is up to 13.6 times larger than in the pH 7 case in the TAPP diacid-PSS brush and up to 4.1 times larger in the case of TMPyP monoacid-PSS brush, and finally up to 22.1 times larger in the case of TMPyP diacid-PSS brush.
To identify if a certain polyelectrolyte architecture is necessary to increase the catalytic activity, measurements with linear PSS as polyelectrolyte at one chosen loading ratio
l = 0.4 were performed. As shown in
Figure 5, also with the linear polyelectrolyte, the catalytic activity becomes enhanced whereas in the case of TMPyP diacid, the catalytic activity without linear PSS is slightly higher than that with linear PSS. The concentrations of generated triioide are slightly higher for the TMPyP diacid with linear PSS as with the PSS brush, while the other porphyrins lie in the same range as with PSS brush, as given in
Table 5. This is in contrast to neutral conditions. From
Figure 5 and
Table 5, it is evident that most triiodide is generated with the TMPyP diacid system, which is consistent with the results of the PSS brush. For selected samples, the influence of long-time irradiation on the catalytic performance of porphyrin diacid-PSS brush aggregates was investigated. Samples were irradiated up to five hours. A more extended irradiation interval was not possible for the TMPyP diacid samples due to absorption limits reached at the chosen concentration. Again, the extinction coefficients are plotted versus the irradiation time in
Figure 6. It becomes evident that the further increase of the catalytic activity due to the four-hour longer irradiation is not that significant for the majority of the porphyrin-polyelectrolyte samples. Only for the TMPyP diacid system can a clear enhancement of the catalytic activity be observed. Consistent with the results for one-hour irradiation, the highest concentration of generated triiodide is found for TMPyP diacid-PSS brush aggregates, as summarized in
Table 5. Yet, still no increase of the iodide concentration due to the polyelectrolyte can be seen. The activity of both samples is nearly the same as after one-hour irradiation. For the remaining three porphyrin-PSS brush systems, a clear increase of the catalytic activity caused by the polyelectrolyte is evident. Again, the observations can be underlined with the corresponding TON and TOF, which are summarized in
Table 6. The maximum increase with a 7.5-times higher concentration of generated triiodide was found for TAPP diacid-PSS brush.
The increase of the catalytic activity can additionally be seen from the color of the investigated samples in
Figure 6. Samples with porphyrin and polyelectrolyte evidently are more intensively yellow-colored.
The reusability of the porphyrin diacid-PSS brush-assemblies was also investigated. For this, a sample of TAPP diaicd-PSS brush and a TAPP diacid solution, which were already irradiated and used as catalyst, were irradiated on the next day again. The measurements showed that TAPP diacid-PSS brush assemblies are still catalytically active.