Open Access Article
David Steinebrunner
ab,
Günter Schnurpfeilc,
Mathis Kohrödea,
Alexander Eppa,
Khaetthariya Klangnogc,
Jorge Adrian Tapia Burgos
ab,
Andre Wichmanna,
Dieter Wöhrle*c and
Arne Wittstock
*ab
aInstitute of Applied and Physical Chemistry and Center for Environmental Research and Sustainable Technology, University Bremen, Leobener Str. UFT, 28359 Bremen, Germany. E-mail: awittstock@uni-bremen.de
bMAPEX Center for Materials and Processes, University Bremen, Bibliothekstr. 1, 28359 Bremen, Germany
cOrganic and Macromolecular Chemistry, University Bremen, Leobener Str. NW2, 28359 Bremen, Germany. E-mail: woehrle@uni-bremen.de
First published on 17th June 2020
Nanoporous gold powder was functionalized in a two-step approach by an azide terminated alkanethiol self-assembled monolayer (SAM) and a zinc(II) phthalocyanine (ZnPc) derivative by copper catalyzed azide-alkyne cycloaddition (CuAAC). A series of different hybrid systems with systematic variation of the alkyl chain length on both positions, the alkanethiol SAM and the peripheral substituents of the ZnPc derivative, was prepared and studied in the photooxidation of diphenylisobenzofuran (DPBF). An enhancement by nearly one order of magnitude was observed for the photosensitized singlet oxygen (1O2) generation of the hybrid systems compared to the same amount of ZnPc in solution caused by the interaction of the npAu surface plasmon resonance and the excited state of the immobilized sensitizer. This interaction was shown to be distance dependent, with decreasing activity for short SAMs with alkyl chain lengths < 6 methylene groups caused by quenching of the excited state via electron transfer as well as decreasing activity for SAMs with n > 8 methylene groups due to decreasing energy transfer for long distances. An unexpected distance dependent behaviour was observed for the variation of the peripheral alkyl chain on the photosensitizer revealing a planar orientation of the immobilized photosensitizer on the nanoporous gold surface by a penta-coordinated central zinc ion through interaction with free azide groups from the self-assembled monolayer.
The probably best understood type of hybrid system in this context is the one usually used for PDT consisting of either a metal phthalocyanine or porphyrin derivative attached to gold nanoparticles (AuNPs). Since their first report by Russell and coworkers in 2002, several studies on this type of hybrid materials were performed showing that such systems are superb 1O2 sensitizers with increased photocatalytic activities compared to the analogous complexes used in solution.15,16 The underlying photophysical interactions were investigated in detail in the group of Lemmetyinen, showing that various energy and electron transfer processes occur in such systems, which are also responsible for the observed increase in 1O2 formation.13,17
In the context of photocatalytical 1O2 based oxidation reactions, that are also of interest in industrial processes, colloidal nanoparticle systems are only of minor interest as truly heterogeneous systems are usually better in terms of lifetime and recyclability. Here, nanoporous gold (npAu, Fig. 1b) could be a valuable alternative, as it shows comparable photophysical properties of AuNPs while fulfilling the criteria of a truly heterogeneous, monolithic catalyst.3,18–20 This monolithic material consists of ligaments and pores on the order of 40 nm throughout macroscopic dimensions of up to centimeters (Fig. 1b, inset). In addition, npAu was already shown to be an efficient catalyst for oxidation reactions in both, gas and liquid phase systems.21–24 On the other hand, the same principles for functionalization of the npAu surface with organic compounds such as macrocyclic metal complexes can be applied as in the case of AuNPs.
We recently reported a comparable system where a ZnPc derivative was immobilized on different shaped npAu supports, as examples for a truly heterogeneous monolithic system, a heterogeneous powder catalyst as well as a photocatalytic coating.25–27 These new hybrid photocatalysts showed high overall 1O2 sensitization activities, where the activity was an order of magnitude above the activity of the same amount of the photosensitizer in solution.26
By selective irradiation of only the plasmon resonance of npAu (Fig. 1b) it was shown that a contribution by energy transfer to the attached ZnPc (Fig. 1a) is responsible for the enhanced activity.26 The investigation of the distance dependency of the energy transfer between both parts of the hybrid system is therefore of immense interest to optimize the system regarding its 1O2 sensitization activity. The system itself offers two different positions to achieve hybrid materials with several distances between the npAu surface and the attached photosensitizer. The variation of the alkyl chain length can be performed at the SAM alkanethiol as well as on the peripheral side chains of the ZnPc derivative (Fig. 2). The length of the alkyl chains, but also the orientation of the macrocycle on the surface contributes to the overall distance between the ZnPc and the npAu surface. Therefore, a careful investigation of both distance determining parameters in nanoscaled systems is essential as it can help to improve the design and understanding of novel hybrid systems in this recently strong and fast growing research field.
![]() | ||
| Fig. 2 Schematic representation of the two step npAu powder functionalization: (a) formation of the azide terminated SAM using azidothioates with various alkyl chain lengths, and (b) copper catalyzed azide-alkyne cycloaddition (“click chemistry”) to bind the peripherally alkyne-substituted ZnPc derivatives 1 and 2. (npAu modified from [ref. 20]). | ||
:
30 at%) in concentrated HNO3 and powdering of the obtained npAu disk as described previously.26 The resulting powder showed a mean particle size of around 28 μm and pore sizes in the range of 40 nm.26 The residual silver concentration was below 1 at%, which was confirmed by EDX spectroscopy and shows good agreement to previously reported results.28–32
The organic functionalization of the npAu powder was achieved following a two-step approach established in our group (Fig. 2).25–27 In a first step, the surface of the npAu powder was functionalized with an azide-terminating SAM using acetyl-protected azidoalkylthioates. For the systematic variation of the distance between the immobilized photosensitizer and the npAu surface, SAMs with alkyl chain lengths ranging from nSAM = 3 to 11 CH2 groups between the sulphur anchor and the free azide group were employed. In the following step, the respective ZnPc derivative with either a propynoxy- or hexynoxy-group with nZnPc = 1 or 4 CH2 groups as peripheral substituent was bound to the free azide groups of the SAM via copper catalyzed azide-alkyne cycloaddition (CuAAC). By this method a complete set of hybrid systems was obtained with chain lengths ranging from nSAM = 3 to 11 CH2 groups in the basic SAM for both ZnPc derivatives (Table 1).
| Hybrid | nSAM | nZnPc | Zn [μg gCatalyst−1] | ZnPcIrr. [mol] |
|---|---|---|---|---|
| H1-3 | 3 | 1 | 154.9 | 1.4 × 10−10 |
| H2-3 | 3 | 4 | 151.7 | 1.4 × 10−10 |
| H1-4 | 4 | 1 | 235.3 | 2.1 × 10−10 |
| H2-4 | 4 | 4 | 191.1 | 1.8 × 10−10 |
| H1-5 | 5 | 1 | 161.2 | 1.5 × 10−10 |
| H2-5 | 5 | 4 | 165.9 | 1.5 × 10−10 |
| H1-6 | 6 | 1 | 165.8 | 1.5 × 10−10 |
| H2-6 | 6 | 4 | 189.6 | 1.7 × 10−10 |
| H1-7 | 7 | 1 | 170.1 | 1.6 × 10−10 |
| H2-7 | 7 | 4 | 150.0 | 1.4 × 10−10 |
| H1-8 | 8 | 1 | 147.7 | 1.3 × 10−10 |
| H2-8 | 8 | 4 | 133.4 | 1.2 × 10−10 |
| H1-9 | 9 | 1 | 155.9 | 1.4 × 10−10 |
| H2-9 | 9 | 4 | 187.6 | 1.7 × 10−10 |
| H1-10 | 10 | 1 | 151.6 | 1.4 × 10−10 |
| H2-10 | 10 | 4 | 176.1 | 1.6 × 10−10 |
| H1-11 | 11 | 1 | 157.9 | 1.5 × 10−10 |
| H2-11 | 11 | 4 | 183.9 | 1.7 × 10−10 |
Each of the obtained hybrid systems was analyzed in detail regarding its chemical composition and porous structure of the support. It was shown by SEM that the porous structure of the npAu powder stayed unchanged at 40 nm during the functionalization. The energy dispersive X-ray spectroscopy (EDX) measurements of all samples showed a homogeneous zinc distribution, which was confirmed by EDX mapping and line scan experiments. This confirms previous results of selected systems as described previously by us.26,27
The amount of immobilized sensitizer in every hybrid system was determined by quantification of the amount of Zn in the system by inductively coupled plasma mass spectrometry (ICP-MS) after dissolution of the corresponding sample in ultrapure aqua regia. The obtained values as well as the light exposed fractions of the immobilized sensitizer are summarized in Table 1. The amount of light exposed sensitizer was around 1.5 × 10−10 mol for every hybrid system which was determined from the immobilized sensitizer amount and a penetration depth of 300 nm for visible light into the nanoporous structure.26
For the ZnPc derivative 1 with peripheral propynoxy-substituents the TOF values showed a clear trend with maximum activity for the hybrid H1-7 build of a SAM with seven methylene groups. In contrast, for longer and shorter distances the photocatalytic singlet oxygen sensitization activity decreases rapidly (Table 2, Fig. 3 and 4). Similar trends were observed with decreasing photocatalytic activities by irradiation of mainly the ZnPc using a 700 nm bandpass filter. A further decrease was observed for irradiation of only the plasmon resonance of npAu employing a 550 nm bandpass filter. The highest activities using the 550 nm cut-on filter clearly shows the synergistic effect of the npAu surface plasmons for the generation of singlet oxygen (Table 2).
| Hybrid | TOFco [min−1] | TOF700 [min−1] | TOF550 [min−1] | dSAM [nm] |
|---|---|---|---|---|
| H1-3 | 988.9 | 639.7 | 99.2 | 0.96 (s) |
| H2-3 | 719.6 | 428.3 | 60.6 | 0.96 (l) |
| H1-4 | 2347.8 | 1729.1 | 186.8 | 1.07 (s) |
| H2-4 | 1905.7 | 975.8 | 127.8 | 1.07 (l) |
| H1-5 | 2596.6 | 1979.4 | 243.9 | 1.21 (s) |
| H2-5 | 1984.8 | 1070.4 | 136.9 | 1.21 (l) |
| H1-6 | 2961.9 | 2164.9 | 294.4 | 1.32 (s) |
| H2-6 | 2400.9 | 1308.8 | 171.2 | 1.32 (l) |
| H1-7 | 3670.8 | 2481.8 | 304.3 | 1.46 (s) |
| H2-7 | 3125.1 | 1826.1 | 260.7 | 1.46 (l) |
| H1-8 | 2934.9 | 1872.9 | 254.1 | 1.58 (s) |
| H2-8 | 2329.8 | 1503.2 | 187.2 | 1.58 (l) |
| H1-9 | 2634.1 | 1624.7 | 238.3 | 1.71 (s) |
| H2-9 | 1807.9 | 1316.7 | 149.4 | 1.71 (l) |
| H1-10 | 2452.6 | 1601.6 | 171.2 | 1.83 (s) |
| H2-10 | 1497.8 | 1026.8 | 119.7 | 1.83 (l) |
| H1-11 | 1619.4 | 949.1 | 119.5 | 1.96 (s) |
| H2-11 | 835.9 | 552.3 | 53.5 | 1.96 (l) |
For the photophysical interactions, two different pathways are possible in such hybrid materials, which were shown spectroscopically in detail on AuNP systems. Lemmetyinen and coworkers observed energy transfer from both, the excited ZnPc to the AuNPs and vice versa.13,17 Also, electron transfer from the ZnPc to the AuNPs resulting in the corresponding phthalocyanine radical cation was observed.13 In general, energy transfer is a long ranging interaction with decreasing efficiency from 1 to 10 nm distance, whereas electron transfer is usually limited by a distance between donor and acceptor of around 1–2 nm in alkanethiol-bridged systems.33,34 The decreasing singlet oxygen sensitization activity for the hybrid systems H1-3 to H1-6 is therefore a result of the competition between energy and electron transfer, where electron transfer for shorter distances gets predominant. This process acts as quenching of the excited state of the ZnPc photosensitizer with negative influence on the singlet oxygen formation. For the hybrid systems H1-7 to H1-11 the distance between donor and acceptor is beyond the electron transfer regime, and therefore only energy transfer takes place with decreasing activity towards larger distances.
The series of the ZnPc derivative 2 with peripheral hexynoxy-substituents surprisingly showed a similar trend as the series of the ZnPc derivative 1 with the maximum activity for the hybrid H2-7. The activities for all other hybrids of this series showed the same electron and energy transfer behaviour for longer and shorter chain lengths of the SAM as already discussed for the first series. This observation is quite unexpected as, for an example, H1-6 and H2-3 should have the same distance between the immobilized sensitizer and the npAu surface due to the same number of CH2 groups if the linking unit is fully linearly extended. Therefore H1-6 and H2-3 should show a comparable photocatalytic activity. But in fact, the observed activities exhibit a completely contrary result as H2-3 shows the lowest activity of all hybrid systems whereas H1-6 is one of the most active hybrids.
The same distance dependent behaviour for both ZnPc series shows unambiguously that the overall distance between the two active centers is mainly determined by the chain length of the employed azidoalkylthioate used for the SAM formation. This fact is indeed contrary to a fully linearly extended orientation of the immobilized sensitizer and therefore a more detailed discussion about the orientation of the ZnPc macrocycle on the npAu surface is essential to understand the observed photocatalytic activities.
The nature of the tetra peripheral substituted ZnPc derivative indeed can easily adopt such a planar configuration by undergoing multiple CuAAC reactions with up to four binding sites. There are some studies about the orientation of phthalocyanines and closely related macrocycles like porphyrins and porphyrazines on Au surfaces reported in literature, but all of them were prepared employing thiol-containing derivatives as precursors for SAM formation and none of those systems was prepared using the CuAAC reaction.35–40 Therefore, although for some systems a planar orientation is reported, the hybrid system reported here may vary in this respect due to the different preparation method. In addition, all of these experiments were performed on single crystal surfaces and therefore those results, but also most of the analysis techniques, are not directly transferable to a nanoporous surface.
The UV-Vis spectra of H1-7 and H2-7 when prepared on a light penetrable, 100 nm thick npAu foil exhibit an interesting feature where the absorption of the ZnPc Q-band is red-shifted compared to the absorption of the same sensitizer in solution, both times measured in a DMF environment (Fig. 5). Such a redshift of the Q-band of phthalocyanines and porphyrins is known to be caused by axial coordination of a nitrogen atom from an organic amine.41–44 The binding of an additional ligand in axial position is a common feature of zinc phthalocyanines and porphyrins derivatives, even forming stable crystal structures, supramolecular dimers and trimers as well as hybrid systems directly immobilized on amine-terminated SAMs.45–53 In addition, the penta-coordinating ability of the Zn central ion of zinc porphyrins is used frequently as template in many supramolecular synthesis approaches.54–57
The observed redshift after immobilization of the ZnPc onto the npAu surface suggests a similar penta-coordinated central ion, which is possible by interactions with free electron pairs of the azide groups from the SAM, representing the main pathway for photoinduced interactions in the hybrid systems (Fig. 6). Such an expected interaction might be even forced when the tetrasubstituted photosensitizer undergoes two or more triazole-based connections to the SAM, bringing the Zn central ion in a template-like direction to the free azide groups of the SAM. Such a coordination mode would result in a distance between the immobilized ZnPc and the npAu surface, which is determined by the chain length of the SAM as well as the axial coordination strength and can explain the observed distance dependent photocatalytic behaviour of the hybrid systems. The slightly higher activity that was found for the series of the ZnPc derivative bearing propynoxy-substituents can be explained by a stronger interaction of the Zn ion with the azide groups compared to the hexynoxy-substituted ZnPc derivative due to the shorter peripheral substituent as indicated by the different shifts of the Q-band after immobilization (Fig. 5).
Another interesting point is the position of the maximum photocatalytic activity for the hybrid systems with nSAM = 7 methylene groups in the SAM. By comparison of the photocatalytic activities and the calculated distances between the Zn center and the npAu surface (Table 2), the maximum was found for a distance around 1.5 nm. This finding is also important to that extend that a distance slightly above 1 nm is often described as the border until electron transfer is possible through alkanethiol-based bridges and only energy transfer is observed for further increasing distances.33,34 In hybrid systems of pyrene-functionalized AuNPs with different alkylthiol linkers, photoinduced electron transfer to the AuNP was observed in 5 methylene containing bridges but not for similar systems containing bridges of 8 methylene groups.33 Therefore, the observed maximum activity around 1.5 nm also supports our suggested planar configuration and our hypothesis of an interaction with the free azide groups.
In previous studies, the coverage of the npAu surface was found to be around 1% of a monolayer, calculated on the assumption that a complete monolayer is based on a 1
:
1 ratio of immobilized photosensitizer molecules to available Au surface atoms. Considering the planar orientation of the photosensitizer on the surface as well as the van der Waals radius of gold, it is obvious that this assumption cannot represent the steric demand of a ZnPc molecule. As a single ZnPc core without any attached peripheral substituents already covers an area of at least 25 gold atoms when adopting a planar orientation on the Au surface, a value of 1% of the previous assumption would correspond to a value of at least 25% of a ZnPc monolayer on the Au surface.
In general, it should be mentioned that the “click-methodology”, namely the CuAAC reaction, is a convenient and versatile method often used for the functionalization of nanostructured surfaces. The attachment of photosensitizers, especially of phthalocyanines or porphyrins derivatives, employing this method is a quite common synthesis strategy for the preparation of various hybrid materials.58–62 Although used frequently, the characterization of as prepared hybrid materials is often challenging, which might be the reason that little is known about the orientation of the macromolecules in such hybrid systems. However, especially when investigating photoinduced interactions, the orientation can have a crucial impact on the overall donor–acceptor distance and therefore possible interactions between immobilized sensitizers and free functional groups of the SAM need to be taken into consideration and discussed properly in the context of new hybrid materials.
The distance dependent photocatalytic activities, the low concentration of the immobilized photosensitizer as well as the redshift of the Q-band all supports the suggested planar orientation of the phthalocyanine ring with respect to the npAu surface, also providing evidence for a penta-coordinated Zn central ion interacting with the terminal azide groups of the SAM. The orientation of the sensitizer as consequence of the chosen preparation strategy using the established CuAAC reaction is an important step towards a better understanding of photoinduced interactions in hybrid systems as well as a template tool for the design and synthesis of novel hybrid-based materials.
:
30 at% were prepared within our group according to a previously published procedure.63 American white gold (12 karat, Ag/Au 50
:
50 wt%) as starting alloy for npAu foils was bought at Noris Blattgold GmbH, Schwabach, Germany. For the CuAAC, the catalyst Cu(MeCN)4PF6 (97%) from Aldrich, hydroquinone from Merck and the co-catalyst tris(benzyl-triazolylmethyl)amine (TBTA), which was obtained by a literature procedure,64,65 were used. The chemical 1O2 quencher 1,3-diphenylisobenzofuran (DPBF, >95%) was received from TCI. All solvents including ethanol (abs., reagent grade), THF (reagent grade, ≥99.0%), DMF (analytical reagent grade, ≥99.5%), HNO3 (analytical reagent grade, 65% and NORMATOM, ultra-pure for trace analysis, 67%) and HCl (NORMATOM, ultra-pure for trace analysis, 34%) were bought at VWR and were used as received without further purification.
:
30 at%, 5 mm diameter, 200 μm thickness) in concentrated HNO3 for 72 h and cleaning with H2O. Subsequently after dealloying the npAu disks were powdered by fragmentation with a metal pincers.26 Obtained npAu samples exhibit a mean pore size of 40 nm as determined by SEM measurements with a specific surface area of around 5 m2 g−1 as calculated according to a literature procedure and also shown by BET measurements on comparable npAu samples.27,84,85
npAu foils with comparable pore sizes were prepared by floating American white gold leaf (Ag/Au 50
:
50 wt%, 1.5 × 1.5 cm, 100 nm thickness) on concentrated HNO3 for 8 h, followed by floating on H2O according to a literature procedure.27
:
1 v%) for another 72 h. Afterwards, the as prepared hybrids were repeatedly washed with THF in order to remove any physisorbed material.
Hybrid systems for the UV-Vis measurements on npAu foils were prepared following the same procedure with slightly modified conditions as described recently for the preparation of photocatalytic coatings.27
The pore size of the npAu support was determined acquiring micrographs of the samples on a scanning electron microscope (SEM, Supra 40, Zeiss, Germany) operated at 15.0 kV acceleration voltage, 300 pA probe current and 10 mm working distance. Characterization was performed by measuring the diameter of at least 250 pores in the obtained SEM images using the program ImageJ.
The zinc distribution on the surface was determined by energy dispersive X-ray spectroscopy (EDX, Bruker XFlash 6/30).
All UV-Vis spectra were recorded using a UV-1600PC UV-Vis spectrometer from VWR.
000 L mol−1 cm−1 in DMF).86 Irradiation was carried out employing a 550 nm cut-on, a 700 nm bandpass or a 550 nm bandpass filter respectively. Photocatalytic activities were determined according to literature for immobilized sensitizers by determination of the turnover number (TON, [converted DPBF] (mol)/[irradiated ZnPc] (mol)) and turnover frequency (TOF, slope of the plot of TON versus reaction time (min−1)) for every hybrid system.87
Footnote |
| † Electronic supplementary information (ESI) available. See DOI: 10.1039/d0ra03891a |
| This journal is © The Royal Society of Chemistry 2020 |