1. Introduction

Studies on noble metal nanoparticles (NPs) have evolved into a significant area of research with applications ranging from biology and medicine to photonics [1–3]. NPs are of great interest as they exhibit nanoparticle plasmon resonances (NPPRs), sometimes also called localised surface plasmon resonances. NPPRs are collective oscillations of the NPs’ conduction electrons driven by an electromagnetic field. Accordingly, a NPPR will affect the optical properties of a chromophore in the NP’s vicinity. In particular, emission from such a chromophore at a specific electronic transition might be enhanced or suppressed depending on the spectral position of this transition relative to the NPPR. Enhancement and suppression of the transition correspond to an increase and decrease in the transition rate, respectively. This effect is known in the case of far-field resonators (with dimensions of at least half the emission wavelength) as the Purcell effect [4] and has been demonstrated for dye molecules and quantum dots in a variety of cavities [5–8].

NPs are near-field resonators that overcome the size limitation inherent to the aforementioned classical far-field resonators by allowing light manipulation at the sub-wavelength level. The latter is exemplified by the modification of the transition rates of chromophores in the close vicinity of metallic nanoparticles [9–19]. Moreover, for a chromophore that exhibits multiple emission lines, the spectral shape of the emission is modulated if, for example, the plasmonic structure enhances a specific transition to the detriment of others [20–27].

In addition to the previously mentioned work, which reported on the modification of dipolar allowed singlet-singlet transitions, there have also been investigations on triplet-singlet transitions in the immediate surroundings of plasmon structures. A particular class of triplet emitters are the porphyrin and porphyrin-like molecules [28]. Porphyrins have found a wide range of applications in science just as in nature where they are the basic building blocks of both haem and chlorophyll [29]. Applications include oxygen sensing because energy transfer between the triplet states of the porphyrin and molecular oxygen is favourable. In particular, platin porphyrins, with lifetimes between 50 and 100 µs, are moderate oxygen photosensitizers and thus, suitable for oxygen sensing in the physiological range [30]. Furthermore, organometallic porphyrins have gained importance in the field of light-emitting diodes (LEDs) where triplet emitters form an integral part of phosphor-converted white LEDs [31]. Further, organometallic porphyrin-based LEDs show large light emission efficiencies as their electroluminescence stems from both singlet and triplet excitons [32–36].

Research on the emission from hybrid nanostructures of porphyrin or porphyrin-like molecules and metallic resonators has so far solely focused on the change in the overall transition rate [37–41]. Control of the spectral shape of such emission has only been achieved when the porphyrins were positioned at the tunnel junction between a thin film and a scanning tunnelling microscope tip [42–44].

In this work, we demonstrate for the first time the spectral reshaping of the phosphorescence from hybrid nanostructures of porphyrin molecules and gold nanorods (AuNRs). Hybrid nanostructures with NPPRs spanning the full phosphorescence range were studied and we observed spectral changes as the NPPR is successively tuned in and out of resonance with the different transitions that contribute to the phosphorescence. This result is of particular relevance because the NPPRs are dipolar resonances while phosphorescence emission is a triplet-singlet transition and hence (at least partially) dipolar forbidden.

2. Sample preparation

Colloids of AuNRs with differing aspect ratios were purchased from Nanopartz Inc., USA (Fig. 1(a) ). The five colloids exhibited extinction maxima at 529, 614, 654, 705 and 734 nm. The NRs are stabilised with a bilayer of cetyltrimethylammonium bromide (CTAB) to avoid aggregation. In solution, CTAB loses its bromide counter-ion and hence acquires a positive charge. The bare gold nanorods did not show any luminescence emission under irradiation with laser light at 405 nm.

 

Fig. 1 (a) Transmission electron micrograph of a gold nanorod with a short axis of 35 nm and a long axis of 65 nm. (b) Chemical structure of Pt-TSPP (c) Extinction spectrum (black solid line) of an aqueous solution of Pt-TSPP. The Soret and Q bands are labelled. Corresponding emission spectrum (red solid line) with a weak emission at 600 nm (L-Peak) and two phosphorescence bands (P1 and P2 peaks).

Download Full Size | PDF

The triplet emitter 5,10,15,20-Tetrakis-(4-Sulfonatophenyl)porphyrin-Pt(II) (Pt-TSPP) was purchased from Porphyrin Systems GbR, Germany, and used without further purification. Pt-TSPP is a sulfonate, which deprotonates in water and hence carries a net negative charge (Fig. 1(b)). Figure 1(c) shows the extinction (black solid line) and emission (red solid line) spectra of Pt-TSPP molecules in aqueous solution. The extinction features a strong peak at 395 nm termed the Soret band and two weaker peaks at 507 and 537 nm termed the Q bands [28]. The photoluminescence spectra were recorded by excitation with laser light at 405 nm, a wavelength close to the Soret maximum. The emission of Pt-TSPP in aqueous solution features three bands: a very weak band at 600 nm (L-peak) and two phosphorescence bands at 660 and 715 nm labelled P1 and P2, respectively.

Hybrid nanostructures of the AuNRs and Pt-TSPP molecules were assembled with the following method. First, 250 µl of a 2 mg/ml aqueous solution of Pt-TSPP was added to 250 µl of AuNRs solution for each particular batch of AuNRs. Next, the mixture was left undisturbed overnight allowing the negatively-charged Pt-TSPP molecules to electrostatically bind to the positively-charged CTAB-capped AuNRs. Deionized water was added to the mixture and the unattached Pt-TSPP molecules were removed by centrifugation.

3. Results and discussion

Figure 2 shows the extinction spectra of colloids containing un-functionalised AuNRs (orange solid line) with increasing long axis lengths from top to bottom. These spectra feature two NPPRs: a transversal mode in the green region, which corresponds to electron oscillation along the rods' short axis and a longitudinal mode, which corresponds to electron oscillation along the rods' long axis. The latter depends strongly on the aspect ratio between long and short axis and is shifted to longer wavelengths for increasing aspect ratios. In the remainder of this letter, the short axis NPPR will no longer be considered as it has only a negligible overlap with the emission bands of the Pt-TSPP molecules.

 

Fig. 2 Comparison of the extinction spectra of un-functionalised AuNRs (orange solid line) and AuNRs functionalised with Pt-TSPP molecules (green dashed line). The short and long axis NPPRs are labelled in Fig. 2(b). There is a permanent wavelength-dependent shift of the long-axis NPPR to longer wavelengths upon functionalization. The long axis NPPR of the AuNRs without porphyrin peaking at 529, 614, 654, 705 and 734 nm, are shifted to 542, 626, 678, 717 and 758, respectively, in the case of the hybrids. Inset: Sketch of the AuNRs with increasing aspect ratios from top to bottom.

Download Full Size | PDF

Upon the formation of the hybrid nanostructures, the extinction properties were altered. Figure 2 compares the extinction of the bare AuNRs (orange solid line) to the extinction of the corresponding hybrid nanostructures (green dashed line). In Fig. 2(a), the NPPR at 529 nm overlaps with the extinction peaks of the Pt-TSPP molecules at 507 and 537 nm. The resulting extinction spectrum of the hybrid system peaks at 514 and 542 nm. In Figs. 2(b)-2(e), the hybrid nanostructures have NPPRs that are shifted to longer wavelengths compared to the NPPRs of the corresponding un-functionalised nanorods. The immediate surroundings of the NPs have changed through the binding of the Pt-TSPP molecules to the AuNRs via the CTAB resulting in an increase of the effective refractive index and consequently, a redshift of the plasmon resonance is observed. Remarkably, the extent of the redshift depends on the particular AuNRs solution that was employed. Colloids with the NPPR at 614 and 705 nm showed a redshift of ~10 nm upon functionalization with Pt-TSPP molecules, while colloids with the NPPR at 654 and 734 nm showed a substantial redshift of ~25 nm. We note that the larger shifts correspond to the two cases where the NPPR overlaps strongly with one of the triplet-singlet transitions P1 or P2. This is an indication of strong coupling between porphyrin molecules and metallic nanoparticles similar to what has been reported in the case of singlet-singlet transitions [45, 46]. Haes et al. [45] have shown that the NPPR shift becomes wavelength dependent and reaches its maximum when the NPPR is near the molecular resonance. Pt-TSPP has two molecular resonances, which can explain the two excessive redshifts in Fig. 2(c) and Fig. 2(e). Further reports on strong coupling include anticrossing behaviour between porphyrin excitons and surface plasmon polaritons [47] and between singlet emitters and NPPRs [48, 49].

So far, studies on the coupling between porphyrin molecules and gold nanoresonators have focused on the extinction properties. In the remainder of this paper, we investigate for the first time the effect such coupling has on the emission properties of the hybrid system. Figure 3 compares the photoluminescence of porphyrin-gold nanostructures (red solid line) to the corresponding extinction spectrum (green dashed line). The top graph in Fig. 3 presents the results for the AuNRs solution with the NPPR at 529 nm. It is often the case that the emission from a dye in solution differs from the emission of a dye on a substrate [50]. In our case, the substrates are the CTAB-capped AuNRs. Hence, the difference of the luminescence spectra of Figs. 1(a) and 3(c) can be easily explained and in fact, the luminescence spectrum in Fig. 3(a) should be considered as the “out of NPPR resonance” reference spectrum of Pt-TSPP rather than the spectrum given in Fig. 1(c). The second phosphorescence peak that appears only as a shoulder in the case of the Pt-TSPP in aqueous solution (Fig. 1(c)) is now better spectrally resolved from the main peak owing to a diminution of the fwhm of both peaks.

 

Fig. 3 Influence of the NPPRs (green dashed line) of Pt-TSPP-AuNRs on the emission (red solid line) from these hybrid nanostructures. The extinction spectra are normalised at the long axis NPPR and the emission spectra are normalised at the P1 emission peak.

Download Full Size | PDF

Next, there is strong evidence for the manifestation of the plasmon induced Purcell effect. The Pt-TSPP molecules have two phosphorescence bands that are competing radiative channels and as each of them comes successively into resonance with the NPPR of increasing AuNR aspect ratio, the emission spectrum is modified. The peak to peak amplitude (PP) ratio between the two phosphorescence bands is determined in order to quantify the change in the emission spectrum. A high PP ratio indicates an enhancement of the first phosphorescence band P1 compared to the second band P2 while a low PP ratio indicates the opposite trend. In the case of Fig. 3(a) where the NPPR is at 529 nm, the PP ratio is 1.6. There is no overlap here between the NPPR and the P1 and P2 bands and the NPPR should have no influence on the emission. In Fig. 3(b), the long axis NPPR is shifted to longer wavelengths and has a moderate overlap with the first phosphorescence band P1 and a weak overlap with the second phosphorescence band P2. The PP ratio increases to 1.8 as the triplet-singlet transitions couple to the NPPR. The P1 band experiences a stronger coupling and hence a greater enhancement than the P2 band due to its better overlap with the NPPR.

In Fig. 3(b), the NPPR at 626 nm overlaps well with the L band at 600 nm. Only in this case, the 600 nm emission is visible, but in the subsequent cases where the NPPR is at longer wavelengths, it does not appear in the spectrum. When in resonance with the NPPR, the 600 nm emission is enhanced and is seen in the spectrum. When out of resonance, it is not enhanced, cannot compete with the P1 and P2 bands and is therefore not seen in the spectrum.

Next, we discuss the hybrid nanostructures comprising the AuNRs with the long axis NPPR peak at 678 nm (Fig. 3(c)). This resonance strongly overlaps with the first phosphorescence band P1 but only moderately with the second phosphorescence band P2. In this case, there is a drastic increase of the PP ratio to 2.4 and the second emission band only appears as a shoulder to the main peak. The transition rate of Pt-TSPP at the first phosphorescence channel P1 has increased stronger compared to the second phosphorescence channel P2. The fourth colloid suspension has its NPPR at 717 nm, which strongly overlaps with the second phosphorescence band P2 (Fig. 3(d)) but only weakly with P1. Compared to the previous case, the roles are reversed, the second phosphorescence band shoots up and here the PP ratio drops to 1.4. The transition rate of Pt-TSPP at the second phosphorescence channel has increased stronger compared to the first phosphorescence channel. Up to this point, the phosphorescence bands have kept the same spectral positions with changes in their respective PP ratio only. This is no longer the case for the last batch of hybrid nanostructures with a NPPR at 758 nm (Fig. 3(e)). The phosphorescence bands are pulled towards the red by a few nm for the first band and ~10 nm for the second band, which is spectrally closer to the NPPR. Additionally, the PP ratio has decreased even further; it is now down to 1.2. This might appear surprising at first as the overlap between the long axis NPPR and the P2 band is better in Fig. 3(d) than it is in Fig. 3(e). However, the PP ratio depends on the overlap of each of the two emission bands with the long axis NPPR. While in Fig. 3(d), the overlap between the NPPR and the P1-peak is non negligible, it has become negligible in Fig. 3(e). On the other hand, the overlap between the NPPR and the P2-peak is good in both Fig. 3(d) and 3(e). These two enhancement factors combine to give a lower PP ratio in Fig. 3(e) compared to Fig. 3(d).

Finally, the findings presented in Fig. 3, are summarised in Fig. 4 . The PP ratio is largest when the NPPR is close to the P1 band and smallest when close to the P2 band. This change of the PP ratio as a function of the NPPR, in particular the lower value of the PP ratio for Fig. 3(e) compared to Fig. 3(d) and the disappearance of the 600 nm emission when out of resonance are strong indicators of the plasmon induced Purcell effect.

 

Fig. 4 The PP ratio ( = P1/P2, see Fig. 3) as a function of the long axis NPPR of the hybrid nanostructures. It reaches its maximum for colloids with the NPPR at 678 nm and has its lowest value for colloids with the NPPR at 758 nm.

Download Full Size | PDF

4. Conclusions

In conclusion, we have shown that the phosphorescence emission from ensembles of hybrid porphyrin molecules and gold nanorods follows the long axis plasmon resonance of the nanorods, a signature of the plasmon induced Purcell effect. When scanning the emission range of the triplet emitters with gold nanorods of increasing aspect ratios, the relative rates of the transitions contributing to the phosphorescence are successively changed with the effect of modulating the emission spectrum. These results should help to further increase the emission efficiency of porphyrin-based light emitting diodes and allow for colour tunability. Moreover, the water solubility of these porphyrin molecules makes the hybrid nanostructures suitable for biosensing applications.