Background

Colloidal semiconductor nanocrystals or quantum dots (QD) due to their specific characteristics, intense broad absorption and narrow luminescence spectra with the size-dependent maximum position and high thermal and photostability [1, 2], find applications in various fields of modern technology, such as medical imaging and diagnostics, modern computing nanodevices, fluorescent probes for bioanalytical applications, photoelectrochemical hydrogen generation, etc. ([3,4,5,6,7] and references therein). Functionalization of the QD surface with organic molecules makes it possible to increase their solubility in water, to reduce their toxicity, and to increase their biocompatibility, preparing QD with selective affinity to desirable structures of living organisms [8]. Therefore, QD attract special interest for applications in biology [5] and medicine [6], where they could be successfully applied as fluorescent probes (FP) for fluorescence diagnostics (FD) [9] and photosensitizers (PS) for photochemotherapy (PCT) [10]. Intense absorption in a broad spectral region makes QD an effective antenna for light energy accumulation, and intense narrow luminescence band facilitates the energy transfer to corresponding PS, thus increasing the efficiency of the light energy utilization and consequently increasing the PS efficacy [7, 11]. This makes (QD+PS) pairs promising for application in FD and PCT and stimulates studies in the QD and FS interaction, especially in the transfer of energy and charge between them.

Among others, cysteine-coated (CdSe/ZnS) QD ((CdSe/ZnS)-Cys QD) and meso-tetrakis (p-sulfonato-phenyl) porphyrin (TPPS4) attract a special interest due to following reasons: a small size of cysteine-coated QD (QD-Cys) which increases its mobility and the probability to penetrate cell membrane, its high chemical stability, low nonspecific adsorption, and high luminescence quantum yield [12, 13]. On the other hand, synthetic TPPS4 porphyrin is a promising PS since it is photoactive, water soluble, and non-toxic and has already been tested in clinics in application in photodynamic therapy (PDT) demonstrating hopeful characteristics [14, 15].

Interaction between TPPS4 and QD via energy and/or charge transfer has already been documented [16]. Commonly, these processes are accompanied by reduction in the QD luminescence intensity and lifetime. One more process that causes luminescence self-quenching in QD is the self-aggregation via electrostatic interactions or hydrogen bond formation, in many cases, making the aggregation process reversible [17].

In this work, we report for the first time on the stimulation of the QD luminescence via interaction with porphyrin on the example of (CdSe/ZnS)-Cys QD and TPPS4 porphyrin.

Experimental

Preparation of (CdSe/ZnS)-Cys Quantum Dots

The (CdSe/ZnS)-Cys QD were synthesized in accordance with the method adapted from [18]. The method includes the following: (1) synthesis of colloidal hydrophobic CdSe core nanocrystals and (2) growth of an epitaxial ZnS shell around the CdSe core. To functionalize QD with cysteine, the resultant CdSe/ZnS core–shell QD (~ 3.0 mg) were purified from TOPO via threefold dispersing in chloroform (500 mL) and precipitating with methanol (800 mL). Purified QD were re-dispersed in chloroform (1.0 mL). DL-Cysteine in methanol (30 mg mL− 1, 200 mL) was added dropwise to the QD dispersion and mixed vigorously followed by centrifugation (10,000 rpm, 5 min), removing chloroform. After washing with methanol to remove the excess of DL-Cysteine through centrifugation (16,000 rpm, 10 min, 3 times), the QD precipitate was dried under vacuum and re-dispersed in Milli-Q water with 1 M NaOH (20 mL) dropwise addition and filtered with syringe filter Anotop 25 Plus (0.02 μm, Whatman).

Preparation of porphyrin + (CdSe/ZnS)-Cys QD Samples

The TPPS4 porphyrin was obtained from Midcentury Chemicals (USA) and used without additional purification. The experimental solutions were prepared in phosphate buffer (pH 7.3; 7.5 mM), using Milli-Q quality water. For luminescence measurements in (CdSe/ZnS)-Cys QD kept at 276 K for 3 months (aged QD), aliquots of a concentrated TPPS4 stock solution ([TPPS4]stock = 140 μM) were added to the (CdSe/ZnS)-Cys QD initial solution, avoiding dilution effects. For the aged QD dilution experiment, aliquots of the initial solution were replaced by the same amount of phosphate buffer. All the experiments were performed at room temperature (297 K).

The concentration of TPPS4 was controlled spectrophotometrically using ε515nm = 1.3 × 104 M− 1 cm− 1 [19]. The concentration of the aged (CdSe/ZnS)-Cys quantum dot was calculated using the first excitonic absorption peak at 520 nm using ε = 5857(D)2.65 according to Yu’s empirical calculation [20], where D(nm) is the diameter of the given nanocrystal. The D value was determined from the empirical fitting function of the curve as presented in [20]. For CdSe nanocrystals, this function is:

$$ D=\left(1.6122\times {10}^{-9}\right){\lambda}^4-\left(2.6575\times {10}^{-6}\right){\lambda}^3+\left(1.6242\times {10}^{-3}\right){\lambda}^2-(0.4277)\lambda +(41.57) $$
(1)

In our case, λ = 520 nm, D = 2.6 nm, and ε = 7.4 × 104 M− 1 cm− 1.

Instruments

The absorption spectra were monitored with a Beckman Coulter DU640 spectrophotometer. The steady-state luminescence measurements were made on a Hitachi F-7000 spectrofluorimeter at λex = 480 nm and λem = 558 nm. The aged QD luminescence quantum yield (QY) was determined via relative method [21] with a single point measurement, λex = 480 nm and λem = 558 nm, using 1-palmitoyl,2-(12-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]dodecanoyl)-sn-glycero-3-phosphocholine (C12-NBD-PC) as a standard (QY = 0.34 in ethanol) [22, 23] according to the equation:

$$ {\Phi}_{fl}={\Phi}_{fl0}\frac{n^2{I}_{fl}}{n_0^2{I}_{fl0}}\frac{A_0}{A} $$
(2)

where Ifl and Ifl0 are the integral fluorescence intensities of QD and C12-NBD-PC, A and A0 are their absorbances at λex = 480 nm, and n and n 0 are refractive indexes of the used solvents, respectively.

Time-resolved experiments were carried out by using an apparatus based on the time-correlated single-photon counting method. The excitation source was a Tsunami 3950 Spectra Physics titanium-sapphire laser, pumped by a Millenia X Spectra Physics solid state laser. The frequency of the laser pulse repetition was 8.0 MHz using the 3980 Spectra Physics pulse picker. The laser was tuned so that the second harmonic generator BBO crystal (GWN-23PL Spectra Physics) gave the 480 nm excitation pulses that were directed to an Edinburgh FL900 spectrometer. The spectrometer was in L-format configuration, the emission wavelength was selected by a monochromator, and the emitted photons were detected by a refrigerated Hamamatsu R3809U microchannel plate photomultiplier. The full width at half maximum (FWHM) of the instrument response function was typically 100 ps, and time resolution was 12 ps per channel. The software provided by Edinburgh Instruments and the commercial “OriginPro9” software were used to fit the experimental luminescence decay curves.

The quality of the fit was evaluated by the analysis of the statistical parameter reduced-χ2 and by inspection of the residuals distribution.

Dynamic light scattering was measured with NanoBrook 90Plus Zeta Particle Size Analyzer with excitation at 640 nm using a 40 mW HeNe laser (Brookhaven Instruments Corporation).

Results and Discussion

Freshly prepared (CdSe/ZnS)-Cys QD possess the maximum of the luminescence spectrum at 558 nm (Fig. 1, black line), as previously reported by Liu et al. [13], and quantum yield (QY) 0.75 [2, 24, 25]. The addition of TPPS4 to the fresh solution induces no changes, both in the QD luminescence intensity and the luminescence spectrum profile.

Fig. 1
figure 1

Normalized luminescence spectra of (CdSe/ZnS)-Cys 558 quantum dots in phosphate buffer (7.5 mM) at pH 7.3: fresh prepared (black line, λmax = 558 nm), after 3 months in the refrigerator at 276 K (aged QD) without TPPS4 (red line, λmax = 556 nm), and at addition of [TPPS4] = 5.0 μM to aged QD (blue line, λmax = 559 nm), λex = 480 nm

Full size image

For (CdSe/ZnS)-Cys QD dissolved in water and kept in the refrigerator at 276 K for 3 months (aged QD), the position of the maximum of the luminescence spectrum, measured in phosphate buffer (7.5 mM) at pH 7.3, was blue-shifted for 2 nm (λmax = 556 nm), as compared with that of fresh QD. The emission band appeared widened and slightly asymmetric (Fig. 1, red line). The quantum yield of the aged QD luminescence, determined by the method described above, was 0.23 ± 0.03.

The addition of TPPS4 to the aged QD solution induced significant increase in the luminescence intensity (Fig. 2a), QY value reaching 0.75 ± 0.08 (Fig. 2a, inset), the value closed to that for fresh QD [2, 24, 25].

Fig. 2
figure 2

a Luminescence spectra and quantum yield (inset) of aged (CdSe/ZnS)-Cys 558 QD ([QD] = 570 nM, black curve) solutions as a function of the TPPS4 porphyrin concentration. b Decay kinetics of QD luminescence and the \( {I}_{0_3}/\left({I}_{0_2}+{I}_{0_3}\right) \) ratio (inset, see Eq. (3)) as a function of the TPPS4 porphyrin concentration

Full size image

Moreover, in the presence of TPPS4, the symmetrization of the luminescence band of the aged QD and the reduction of its bandwidth were observed, accompanied by maximum red shift up to λmax = 559 nm, close to the maximum of the fresh QD spectrum (Fig. 1, blue line.).

The luminescence decay curves obtained at 480 nm excitation for solutions of both fresh and aged QD were successively fitted as a sum of three exponentials:

$$ I={I}_{0_1}{e}^{-t/{\tau}_1}+{I}_{0_2}{e}^{-t/{\tau}_2}+{I}_{0_3}{e}^{-t/{\tau}_3} $$
(3)

where \( {I}_{0_i} \) and τ i are pre-exponential factor (amplitude) and lifetime of the i-th decay component, respectively.

The component lifetimes for both fresh and aged QD are independent from the porphyrin presence (Table 1). The luminescence lifetimes of fresh QD solutions are typical for (CdSe/ZnS)-Cys QD [26, 27]. For aged QD, the component lifetimes are much shorter (Table 1).

Table 1 Lifetimes of the (CdSe/ZnS)-Cys QD luminescence decay curve components in phosphate buffer (7.5 mM) at pH 7.3: fresh QD and those after 3 months in the refrigerator at 276 K (aged QD) in the absence and presence of TPPS4, λex = 480 nm, λem = 558 nm
Full size table

The values of τ1 in all cases, fresh and aged QD in the presence and absence of porphyrin, are close to time resolution of the single-photon counting equipment (≈ 100 ps) used in this study. Therefore, it should be associated with the scattered light of the exciting pulse.

It is well established [28,29,30] that the short-lived (τ2) and long-lived (τ3) components are associated with the luminescence resulted from the electron-hole annihilation in the QD core (τ2) and shell (τ3), respectively. Total intensity of these two components characterizes the whole annihilation process in the QD. In this case, the relative intensity (amplitude) of the τ3 component should demonstrate the contribution of the electron-hole annihilation in the QD shell. The relative contribution I3 of 3rd component to the decay curve was calculated as:

$$ {I}_3=\frac{I_{0_3}}{I_{0_2}+{I}_{0_3}} $$
(4)

The addition of TPPS4 to fresh QD solutions does not change significantly the relative content of the components (data not shown), while for aged QD solutions, the relative content of τ3 component I3 increases with the TPPS4 concentration (Fig. 2b, inset). The dependence of QY for aged quantum dot luminescence on the TPPS4 concentration is similar to that for I3 (Fig. 2a, b, insets), both reaching the maximum values approximately at 2.0 μM TPPS4. This means that TPPS4 interacting with aged QD affects stronger the luminescence of QD shell than that of its core. However, TPPS4 in fresh QD solutions demonstrates no effect upon the QD luminescence. Therefore, we conclude that the TPPS4 effect observed for the aged QD solution cannot be explained by the porphyrin binding to the QD surface.

On the other hand, the observed increase of the aged QD luminescence intensity at interaction with TPPS4 cannot be explained via reverse energy transfer from TPPS4 to QD, since the TPPS4 fluorescence spectrum is localized in the range λ > 600 nm where QD absorption is weak (Additional file 1: Figure S3). Therefore, the energy transfer via Förster resonance energy transfer (FRET) mechanism is low probable. Moreover, the QD luminescence was excited at 460 or 480 nm, where TPPS4 optical absorption is negligible. In addition, the absorption spectra of TPPS4 remained unchanged in the mixed solutions, demonstrating the absence of charge transfer between QD and TPPS4 (Additional file 1: Figure S4b, c).

The ability of quantum dots to aggregate via formation of non-covalent NH···H hydrogen bonds between the QD surface groups was already documented [13, 17]. Aggregation reduces the QD luminescence, quenching most effectively the component attributed to the QD shell [13, 17]. The reduction of the QD luminescence intensity and lifetime was observed for CdSe-QD in solid films due to formation of 3D aggregates [31]. The authors proposed a model, where this reduction is associated with energy transfer between individual QD in the aggregate [32].

On the bases of this evidence, we believe that while in the refrigerator, QD do aggregate, which reduces the luminescence intensity and lifetimes. Therefore, we associate the observed increase in the QD luminescence intensity and lifetimes in the presence of TPPS4 with QD disaggregation, stimulated by TPPS4 at its binding with the aggregate. The similar effect was observed for the emission of aggregated QD at their interaction with fluorine ions [17].

The observed changes in the luminescence band profile for aged QD (Fig. 1) can be explained by QD aggregation, as well, its asymmetry being associated with existence of different types of aggregates. Interaction with TPPS4 reduces aggregation and makes the luminescence band profile similar to that for non-aggregated QD, observed in fresh solutions.

At neutral pH, the QD-Cys surface possesses negative net charge due to deprotonation of terminal amino groups on its surface [17, 33, 34]. At this pH, TPPS4 has net charge (4-) due to four negatively charged sulfonate phenyl groups in its structure ([35, 36] and references therein). Therefore, interaction between QD cysteine groups and TPPS4 molecules is low probable because of electrostatic repulsion. However, high affinity of the porphyrin π-conjugated system for metal surfaces is well documented [37]. This affinity should be responsible for TPPS4 binding on the quantum dots surface, despite of the electrostatic repulsion between the QD and porphyrin side groups. The interaction between QD surface and π-conjugated system of the bound porphyrin could explain weak broadening of the porphyrin fluorescence spectrum (Figs. 1, 3, and Additional file 1: Figure S3a, inset) and observed changes in the fluorescence excitation spectrum (Additional file 1: Figure S5b, inset) [38].

Fig. 3
figure 3

Normalized luminescence emission spectra of TPPS4 in phosphate buffer (7.5 mM, pH 7.3) for various TPPS4 concentrations in the presence of aged (CdSe/ZnS)-Cys 558 quantum dot (570 nM), λex = 460 nm

Full size image

Binding of some porphyrin molecules on the QD surface increases the QD surface negative charge, thus increasing electrostatic repulsion between particles and inducing their disaggregation (Scheme 1) [39].

Scheme 1
scheme 1

The scheme of the interaction between aged (CdSe/ZnS)-Cys 558 QD and TPPS4 porphyrin at neutral pH. The porphyrin molecules adsorb on the QD surface due to the high affinity of the porphyrin π-conjugated system for metal surfaces increasing the net negative charge on the QD surface, thus increasing electrostatic repulsion between particles and inducing their disaggregation

Full size image

The QD surface area AQD ≈ 145 nm2 is sufficient to adsorb several TPPS4 molecules (ATPPS4 ≈ 1.8 nm2 per unit) [40], as it was observed for porphyrins interacting with magnetic and gold nanoparticles [41, 42].

To cover the whole area of QD by porphyrins, 80 porphyrin molecules per individual QD are necessary. However, the saturation of the luminescence QY and I3 values in 570 nM QD solution was observed at approximately [TPPS4] = 2.0 μM (Fig. 2), which demonstrates that the binding of four porphyrin molecules per QD is sufficient for QD disaggregation. This could be explained by larger charge density on the porphyrin molecule as compared with that of QD (Additional file 1: Figure S6) which produces stronger electrostatic repulsion between QD with bound porphyrins. Indeed, Zeta-potential for the aged QD (ζQD) is − 36.1 mV and that for TPPS4 molecule (ζTPPS4) is − 37.6 mV. Average charge density, calculated as σ = ζ /AQD, for an individual aged QD is

σQD = − 36.1 mV/145 nm2 = − 0.25 mV/nm2.

At the same time, for an individual aged QD bound with four TPPS4 molecules, the average charge density (σQD+TPPS4) is

σQD+TPPS4 = − (36.1 + 37.6 × 4) mV/145 nm2 = − 1.29 mV/nm2.

Thus, the binding of four TPPS4 molecules with an individual aged QD increases its σ more than 5 times, increasing the force of electrostatic repulsion more than 25 times and inducing the aged QD disaggregation.

In accordance with the QD aggregation hypothesis, a similar effect to that induced by TPPS4 addition should be observed at dilution of aged QD solutions. Really, we have observed the increase in the QY of QD luminescence at dilution of its buffer solution (Fig. 4a, inset), which demonstrates that the self-quenching of QD luminescence in aged QD-Cys solutions depends on QD concentration [17]. Simultaneously, the I3 value in the QD luminescence kinetics increases with dilution, as well (Fig. 4b, inset).

Fig. 4
figure 4

a Luminescence spectra and quantum yield (inset) of aged (CdSe/ZnS)-Cys 558 QD solutions in function of its concentration. b Decay kinetics of the QD luminescence and the I3 value (inset, see Eq. (3)) in function of its concentration

Full size image

Moreover, the dynamic light scattering experiments show that Dhd of scattering particles in the QD solutions after aging was (330 ± 170) nm, which is much larger than that of fresh QD. Dilution reduces Dhd down to (25 ± 6) nm, thus demonstrating directly the QD disaggregation (Additional file 1: Table S1).

There exists one more interesting aspect of the problem: can addition of TPPS4 to the solution of fresh QD prevent their aggregation during storage at low temperature, thus stabilizing their luminescence characteristics? However, clarification of this problem needs an independent and detailed study using various experimental methods and varying experimental conditions, such as reagent concentrations, temperature, duration of the solution storage (several months), etc. We plan to realize this profound study in the nearest future.

Conclusions

Basing on the obtained data, we can assert that the long storage of CdSe/ZnS-Cys QD in aqueous solutions even at low temperatures induces their aggregation, which reduces the luminescence quantum yield and lifetimes. The addition of TPPS4 porphyrin stimulates disaggregation of aged CdSe/ZnS-Cys QD which is pronounced via increase of the QD luminescence quantum yield and the contribution of electron-hole annihilation in the QD shell in the total QD luminescence. The disaggregation, stimulated by porphyrin, takes place due to the increase of electrostatic repulsion between aggregated QD at their binding with negatively charged porphyrin molecules. Disaggregation has been observed at the dilution of QD solution, as well.

The obtained results demonstrate the way to repair the aged QD by adding some molecules or ions to solutions, stimulating QD disaggregation and restoring their luminescence characteristics, which could be important for QD biomedical applications, such as bioimaging and fluorescence diagnostics. On the other hand, disaggregation is important for QD applications in biology and medicine since it reduces the size of the particles facilitating their internalization into living cells across the cell membrane.