INTRODUCTION

Quantum dots (QDs) have been found to be a very promising material for designing optoelectronics systems and lasers [1]. Their photostability and a very high photoluminescence (PL) quantum yield (QY), reaching almost 100% [2], allow this object to be considered as one of the most reliable tools in nanophotonics [3–6], nanomedicine [7, 8], and nano-biotechnology [9]. Unlike conventional molecular fluorophores, QDs can have several coexisting excited states [10]. Biexciton generation is one of unique photophysical effects in QD, which is currently of great interest to researchers and applications. Moreover, the biexciton PL QY in QDs can be sufficiently high to be observed experimentally [11]. Such a biexciton emission with a high QY could be applied to the fields whenever one-photon sources are required. For example, biexcitons can be used for generating entangled photons for solving problems of quantum cryptography [12–14]. Another example is designing a laser based on quantum dots [15], tunable throughout the optical region by varying the size and material of the QDs. Nevertheless, the biexciton QY in bare QDs is much lower than the monoexciton QY due to strong Auger interaction between two excitons in the same QD, which leads to a high nonradiative relaxation rate.

Strong and weak coupling between localized plasmons in noble metal nanoparticles and excitons in semiconductor nanoparticles can modulate the PL properties of QDs. In particular, these effects can strongly affect the luminescence lifetime of a single QD and can induce biexciton states under specific conditions [16]. It is well known that the QY of fluorophores might be increased in the presence of plasmonic nanoparticles [17] due to the shortening of the luminescence lifetime which is explained by the Purcell effect [18]. Some research groups assert that the QY of biexciton emission can also be significantly enhanced in the plasmon field [19].

In this study, we have experimentally investigated the PL properties of single QDs altered by plasmonic nanoparticles placed close to them and demonstrated the enhancement of biexciton emission.

EXPERIMENTAL

We have investigated the PL properties of thin films of a nanohybrid material based on QDs and gold NRs (Fig. 1), which was produced by spin-coating of individual components onto a glass substrate. We used colloidal QDs with a CdSe core and a ZnS/CdS/ZnS multishell capped with hexadecylamine, synthesized according to the procedure described in [18]. The QY of as-synthesized QDs in a hexane solution was 90%, which was measured using Rhodamine 6G as a reference. We have also used an aqueous solution of gold NRs which were 20 × 20 × 40 nm in size and coated with CTAB. The QDs PL band centered near 560 nm was found to be completely overlapped with the extinction spectrum of NRs (Fig. 2), which ensures resonance conditions for interaction between excitonic (QD) and plasmonic (NR) states.

Fig. 1.
figure 1

(Color online) Optical scheme of the PicoQuant MicroTime 200 experimental setup for measurements of the photoluminescent signal of a single quantum dot doped in thin films of PMMA in the presence of gold nanorods.

Full size image
Fig. 2.
figure 2

(Color online) The quantum dot luminescence spectrum (dashed line) and the extinction spectrum of gold nanorods (solid line).

Full size image

The effects of plasmonic nanoparticles on the PL properties of single QDs, as a function of the distance between them, were investigated using set-up shown in Fig. 1. Here, the bottom layer of PMMA served as a thin (3–4 nm) spacer between the glass substrate and QDs preventing their aggregation on the polar glass surface. The top PMMA layer served as a thin (3–4 nm) spacer between QDs and plasmonic nanoparticles. The thickness of PMMA layers was measured by the AFM. Each layer (except the NRs layer) has been spin-coated for 1 min at a rate of 1000 rpm. The surface density of QDs was lower than 0.3 μm–2. According to our estimations, a concentration of 10–9 M was enough to ensure a homogeneous film of NRs on the top of the nanostructure.

The optical measurements where performed with the PL confocal microscopy approach employing the PicoQuant MicroTime 200 setup using an UPLSAPO 60× PlanApochromat water-immersion objective lens with NA = 1.2. For the PL lifetime measurements, a diode excitation laser (λ = 485 nm, P ≤ 3 μW) was used operating in the pulse regime with a repetition rate of 10 MHz and a pulse duration of 50 ps. The excitation beam was focused on the region of interest of the sample by the objective lens, backward luminescence radiation was directed through a dichroic mirror and a confocal pinhole and then split between two avalanche photodiodes (Detector 1 and Detector 2, Fig. 1) working in the single-photon counting regime. For PL lifetime and time trace measurements, the backward signal went only to Detector 1, bypassing the beam spl-itter.

To estimate the QY of the biexciton emission, we used the methods of cross-correlation spectroscopy. The second order correlation function was measured according to Eq. (1),

$${{g}^{{\left( 2 \right)}}}{\text{(}}\tau {\text{)}} = \frac{{{{I}_{1}}(t){{I}_{2}}(t + \tau )}}{{{{I}_{1}}(t)\langle {{I}_{2}}(t + \tau )\rangle }},$$
((1))

where \({{I}_{1}}\), \({{I}_{2}}\) are the PL signal intensities at Detector 1 and 2, respectively, and \(\tau \) is the delay time between the two detectors. Equation (1) describes the probability of photon detection by each of the two detectors. For a single fluorophore, a typical value is \({{g}^{{\left( 2 \right)}}}\left( 0 \right) \approx 0\), which means the absence of simultaneous events at the detectors. In the case of biexciton emission, the central peak of cross-correlation function is non-zero. It is known that the ratio of the area under the central peak to the area under the neighboring ones equals to the ratio of the QY of biexciton emission (QYbx) to the exciton QY (QYexc)9:

$$\frac{{Q{{Y}_{{bx}}}}}{{Q{{Y}_{{{\text{exc}}}}}}} = \frac{{area(central)}}{{area(side)}}.$$
((2))

Therefore, we used this equation to estimate the biexciton QY. All the obtained fluorescence decay curves were approximated by the equation

$$I(t) = \mathop \sum \limits_i {{A}_{i}}{{e}^{{ - t/{{\tau }_{i}}}}},$$
((3))

where I is intensity and t is time, and Ai is an amplitude of the “i” component. To compare the temporal characteristics of luminescence before and after gold NRs deposition, we used the parameter τawl, amplitude-weighted average lifetime, expressed as

$${{{\tau\text{}}}_{{{\text{awl}}}}} = \frac{{{{\Sigma }_{i}}{{A}_{i}}{{\tau }_{i}}}}{{{{\Sigma }_{i}}{{A}_{i}}}}.$$
((4))

RESULTS AND DISCUSSION

The cross-correlation function was measured for a single QD in the absence and in the presence of gold NRs. One can see that \({{g}^{{\left( 2 \right)}}}(0)\) is very close to 0 for an isolated QD (Fig. 3a); this is the so-called antibunching effect, which means that there are very few simultaneous events at the detectors. It is a typical situation for a single QD, when an exciton undergoes radiative recombination by emitting one photon, which can be detected by only one detector at a time. In addition, the fluorescence time trace shows a clear blinking effect (Fig. 4a, black line), and the fluorescence lifetime of the QD, τ = 18 ns (Fig. 4b, black line). In the case of close location of plasmonic nanoparticles, the QD luminescence properties underwent dramatic changes. The area of the central peak of the cross-correlation function grows by a factor of five and becomes comparable to the neighboring ones (Fig. 3b).

Fig. 3.
figure 3

Cross-correlation function \({{g}^{{(2)}}}(\tau )\) of the quantum dot photoluminescence signal in the absence (a) and in the presence (b) of gold NRs.

Full size image
Fig. 4.
figure 4

(Color online) Fluorescence time trace and photoluminescence decay kinetics for a single quantum dot (QD) without closely positioned gold nanorods (NRs) and in the presence of closely positioned NRs. Panel (a) shows fluorescence time trace for a single QD without (black—1) and with (gray—2) gold NRs, with the background noise shown in light gray (3). Panel (b) shows photoluminescence decay kinetics of a single QD without (black—1) and with (gray—2) gold NRs.

Full size image

Also strong quenching of QD luminescence was observed in the presence of gold NRs. The amplitude of luminescence intensity decreased by a factor of five (Fig. 4a, red line), but still significantly exceeded the background noise level (Fig. 4a, green line). At the same time, the blinking effect disappeared. The luminescence lifetime of QDs strongly decreased to τ = 1.5 ns (Fig. 4b), presumably due to the Purcell effect. According to the results of Nair G. et al. [10], the above-mentioned effects indicate a significant increase in the QY of biexciton emission. Note that the prevailing fluorescence lifetime was τ ≈ 300 ps, which can be also attributed to biexciton luminescence em-ission.

CONCLUSION

Experimental data on the PL properties of the hybrid material based on CdSe/ZnS/CdS/ZnS core/multishell QD and gold NRs, separated by the PMMA films of different thicknesses, are presented. The cross-correlation function \({{g}^{{\left( 2 \right)}}}\), value of which is indicative of one- or multi-exciton emission, has been measured for the same QD in the absence and in the presence of gold NRs. In the case of isolated QD, the cross-correlation function \({{g}^{{\left( 2 \right)}}}(0) \approx 0\), which can be evidence of achieving the single-photon emission regime. At the same time, biexciton PL signal emission was found to be negligible. The fluorescence lifetime of a single QD, before gold NRs deposition, is 18 ns. The presence of NRs causes at least an order-of-magnitude decrease in the PL lifetime of single QD, down to ~1.5 ns. The central peak \({{g}^{{\left( 2 \right)}}}(0)\) of the cross-correlation function becomes comparable to the neighbor ones, which indicates that the biexciton QY in QDs became equal to the single-exciton QY.

The results demonstrate the possibility of biexciton formation and emission from single QD, in a QD-NR hybrid material, and the possibility of designing photostable sources of biexcitons emission using only a single QD.