Luminescence properties of hydrophilic hybrid associates of colloidal CdS quantum dots and methylene blue

Journal of Luminescence 156 (2014) 212–218

Contents lists available at ScienceDirect

Journal of Luminescence journal homepage: www.elsevier.com/locate/jlumin

Luminescence properties of hydrophilic hybrid associates of colloidal CdS quantum dots and methylene blue M.S. Smirnov a,n, O.V. Ovchinnikov a, T.S. Shatskikh a, A.G. Vitukhnovsky b, S.A. Ambrozevich b, A.S. Perepelitsa a a b

Department of Optics and Spectroscopy, Voronezh State University, Voronezh 394006, Russia P.N. Lebedev Physical Institute of the Russian Academy of Sciences, Moscow 119991, Russia

art ic l e i nf o

a b s t r a c t

Article history: Received 23 February 2014 Received in revised form 8 August 2014 Accepted 11 August 2014 Available online 21 August 2014

Luminescence properties of hybrid associates of semiconductor colloidal CdS quantum dots (QDs) with an average diameter of 2.5 nm and methylene blue (MB þ ) cations were investigated. Their photoluminescence spectra, excitation spectra and luminescence decay spectra recorded using the timecorrelated single photon counting techniques were analyzed. The increase of efficiency of MB þ fluorescence at 405 nm excitation was found. It corresponds to the absorption region of CdS QDs. When ½nCdS QDs  : ½nMB þ  concentrations ratio changes from 100:1 to 1:3 for associates, then a decrease in the intensity of CdS QDs luminescence band (580 nm) and its increase for MB þ band (674 nm) were observed. At the same time it was found that the average lifetime of CdS QDs luminescence is shortened and the lifetime of MB þ luminescence becomes significantly longer. It was concluded that dynamic quenching due to nonradiative resonance energy transfer takes place. The efficiency of this process was estimated and the corresponding rate constant was calculated. It is 5:6  107 to 8:6  107 s  1 . & 2014 Elsevier B.V. All rights reserved.

Keywords: Luminescence Quantum dots CdS Methylene blue Hybrid associates Resonance energy transfer

1. Introduction Ultrasmall sizes and unique physical properties of colloidal semiconductor quantum dots (QDs) (high brightness of luminescence, photostability, wide excitation range, the size-dependent optical properties) make them ideal for various applications of science, technics and medical technology [1–9]. One of the most actively developing areas is making biological labels, sensors and sensitizers for photodynamic therapy (PDT) of cancer [1,3,6–9]. Hybrid associates of good luminescence colloidal QDs and biologically active molecules (QD–Dye) are most interesting for PDT [6–20]. Hybrid associates of QDs with organic dyes are particularly interesting. These dyes should have a high triplet yield and energy structure, appropriate for photosensitization of 3 O2 -1 O2 process (porphyrins, phthalocyanines of metal, thiazine dyes, etc. [8,9,12–20]). In particular, associates of QDs with methylene blue (MB) molecules are interesting due to unique properties of MB þ cation [3,9,19,20]. The low location of triplet state, high yield of inter-system crossing and efficiency of 3 O2 -1 O2 process photosensitization are characteristic of MB þ . Besides, the formation of QD–MB þ associates allows the dye to stay in the form of MB þ cation for a long time [3]. This is extremely important because MB þ is capable of aggregation and transformation into various proteolytic

n

Corresponding author. E-mail address: [email protected] (M.S. Smirnov).

http://dx.doi.org/10.1016/j.jlumin.2014.08.026 0022-2313/& 2014 Elsevier B.V. All rights reserved.

forms. These forms are incapable of 3 O2 -1 O2 process photosensitization [3,19]. Usually in QD–Dye hybrid associates photoexcited QDs act as donors. Dye molecules in QD–Dye hybrid associates act as acceptors of electron excitation. The main ways of excitation transfer are nonradiative (Förster) resonance energy transfer (FRET) [3,10,13–20], photoinduced electron transfer (PET) [3,12], luminescence resonance energy transfer (LRET), based on the reabsorption of donor luminescence (QDs) by the acceptor (Dye) [11]. The choice of the dominant channel for electron excitation exchange in QD–Dye hybrid associates requires a detailed justification in each case. One of the most difficult and demanding investigations is resonance nonradiative energy transfer. Universal criteria for the choice of the appropriate model are necessary for detailing the mechanism of this process in each particular case. Their formulation requires the analysis of the broadest possible spectrum of empirical regularities for different types of associates. It is necessary to note that when an associate includes QDs, having a luminescence peak with a large Stokes shift, quenching of QDs luminescence and increase in intensity of Dye luminescence are observed [17]. However, conditions of resonance for electron excitation exchange must be corrected. They should be different from the case where the associate includes QDs with dominant radiative annihilation of free excitons [3]. Thus, experimental studies of resonance energy transfer processes are relevant for a wide range of associates synthesized not

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

only with high-temperature organometallic synthesis, but also with significantly lower-temperature methods, such as the sol–gel method. The advantage of the latter approach is the possibility to provide low toxicity of synthesis conditions of hydrophilic QDs, their bright recombination luminescence and direct contact of QDs with Dyes. Research in this way will make steps to the formation of a unified view about the problem of resonance energy transfer in QD–Dye hybrid associates. In this paper we report the results of studies of luminescence properties of hydrophilic hybrid associates of colloidal semiconductor CdS QDs with an average diameter of 2.5 nm and MB þ cations. Presented data sufficiently detail results of the spectroscopic study of the hybrid association of CdS QDs and MB þ cations in a gelatin matrix [21,22]. They are focused on the detail of process of electronic excitation exchange in QD–MB þ associates, using, primarily, fluorescence lifetime measurements.

213

3. Results and discussion 3.1. Structure and spectral characteristics TEM and XRD pattern data agrees with preliminary studies [22]. Using the technique of synthesis allowed us to obtain colloidal CdS QDs. Analysis of TEM images (Fig. 1) showed that QDs with an average size of 2.5 70.4 nm are formed. XRD pattern shows that QDs are crystallized in the cubic phase (Fig. 2). The average diameter found from TEM images analysis was confirmed by the estimation of size using Scherrer's equation from the half-width of XRD reflexes d¼

0:89λ β cos θB

ð1Þ

d is the average diameter of nanoparticles, λ is the Kα1 Cu wavelength (1.5405 Å) radiation, β (in radians) is the full width at half maximum, θB is Bragg's angle.

2. Material and methods 3.2. UV–vis absorption spectra 2.1. Investigated samples The synthesis of colloidal CdS QDs in gelatin and their conjugation with MB þ molecules were realized within sol–gel technique. Detailed description of the preparation technique is given in Refs. [21,22]. The method consists of mixing of 1:3  10  3 M CdBr2  2:5 H2O and 1:3  10  3 M Na2S n9 H2O dissolved in 50 ml of bidistilled water into a reactor containing 200 ml of bidistilled water and 6 g of inert gelatin. Mixing was realized under continuous agitation by magnetic stirring at a speed of 300 rpm. The temperature in reactor during the synthesis was 60 1C. The QDs growth in reactor was limited by stopping chemical reaction. QD– MB þ associates were obtained by mixing solutions containing gelatin sol of QDs and MB þ with certain concentration. The hybrid association was realized during the final stage of CdS QDs interface formation. This provided direct contact of CdS QDs and MB þ molecules. The concentrations of MB þ were 10  3  10  1 moles of dye/mole of CdS QDs (mole fraction or m.f.). This MB þ concentration range provides variation of [nCdS QDs ]:[nMB þ ] concentrations ratio from 100:1 to 1:3. This ratio was unlimited by dimers and H-aggregates formation, which takes place for a large concentration of MB þ [23]. Used MB molecules had high purity and were obtained from Sigma-Aldrich. They are produced as MB hydrates (C16H18ClN3S*3H2O). For further research obtained aqueousgelatin sols of QD–MB þ associates were applied to quartz plates (2  2 cm) and dried.

Fig. 3 shows UV–vis absorption spectra of QD–MB þ mixtures and their components recorded in the region of 250–800 nm. The absorption spectrum of CdS QDs has a pronounced exciton peak at

Fig. 1. Size distribution and TEM images of CdS QDs.

2.2. Methods of investigation and equipment UV–vis absorption spectra of CdS QDs, MB and their mixtures were recorded using a Shimadzu BioSpec-mini (Japan) spectrophotometer. The photoluminescence (PL) and its lifetime were measured using Ocean Optics Maya  Pro 2000 and PicoQuant TimeHarp 100 TCSPC system. PL and its decay were measured using the PicoQuant PDL 800-B semiconductor pulsed laser with a wavelength of 405 nm and a pulse length of 75 ps and repetition 100 kHz. Luminescence excitation spectra were recorded using fluorescence spectrometer LS 45 Perkin Elmer. Transmission electron micrograph (TEM) images were obtained with an LEO 912 AB OMEGA microscope. The crystal structure of synthesized QDs was investigated with X-ray diffractometer ARL X'TRA (Switzerland) for Kα1 of copper.

Fig. 2. XRD pattern of CdS QDs.

214

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

Fig. 3. UV–vis absorption spectra: MB þ in gelatin (4  10  5 m.f.) – 1, CdS QDs – 2, QD–MB þ ([nCdS QDs ]:[nMB þ ] ¼1:1) – 3 and luminescence excitation spectra of QD–MB þ ([nCdS QDs ]:[nMB þ ] ¼1:1) – 4. Inset: dependence of luminescence intensity on [nCdS QDs ]:[nMB þ ] concentration ratio.

Fig. 4. Normalized luminescence spectra: MB þ in gelatin (4  10  5 m.f.) – 1, CdS QDs – 2, QD–MB þ ([nCdS QDs ]:[nMB þ ] ¼ 1:1) in gelatin – 3. Inset: luminescence spectra of MB þ in gelatin (4  10  5 m.f.) – 10 , QD–MB þ ([nCdS QDs ]:[nMB þ ]¼1:1) – 30 .

380 nm. The quantum-size effect can be observed in UV–vis absorption spectra (Fig. 3). CdS QDs absorption spectrum shows blue shift relative to its position in the spectrum of the bulk CdS crystal (2.36 eV) (Fig. 3) by 0.9 eV. The average sizes of QDs were estimated using shift of this position in relation to the fundamental absorption edge of cubic CdS crystals. Estimations were made using the Kayanuma equation [24]

exciton peak in the luminescence spectrum is caused by the rapid capture of carriers by defects [31]. These features are characteristic of luminescence of QDs, synthesized at a temperature below 100 1C in aqueous solution in the presence of stabilizing polymer [32]. There is a peak at 680 nm in the photoluminescence spectrum of MB þ in gelatin. It corresponds to the fluorescence of molecule from the excited singlet state to its ground state, Sn1 -S0 [33]. Blue shift of MB þ luminescence peak from 680 nm in gelatin to 674 nm in an associate with CdS QDs also confirms the earlier conclusion about QD–MB þ association, made from absorption spectra. Changing of [nCdS QDs ]:[nMB þ ] concentrations ratio from 1:1 to 1:3 for QD–MB þ mixtures gives a pronounced decrease in the intensity of CdS QDs luminescence peak. Also a spectral dip in the region of 650–660 nm is enhanced. It is clearly manifested in the normalized luminescence spectra (Fig. 4, curve 3). At the same time, the peak intensity of luminescence at 674 nm, corresponding to MB þ emission increases. A significant increase in luminescence intensity of MB þ during association with QD (Fig. 4, inset), together with luminescence quenching of QD, suggests that the excitation of MB þ molecules is realized by transfer of electron excitation from QD. This conclusion is also confirmed by luminescence excitation spectra of QD–MB þ associates (Fig. 3, curve 4). Really, the luminescence peak at 674 nm, attributable to MB þ , is excited in two spectral regions. The first region is situated at 360– 420 nm and it corresponds to absorbance of CdS QDs in associate (peak at 380 nm). The second region is situated at 550–680 nm and has a peak at 657 nm, corresponding to absorbance of MB þ in the QD–MB þ associate. Also it was found that the redistribution in luminescence intensity of QDs and MB depends on [nMB þ ]:[nCdS QDs ] concentration ratio (Fig. 3, inset). The intensity of dye luminescence increases with the increase of this ratio. However, this increase in luminescence intensity is limited. The luminescence intensity, excited by interaction between components of associate, does not increase as soon as the number of MB þ cations in a mixture with CdS QDs exceeds the number of reactive centers on an QDs interface. At the same time, dimer peak in the absorption spectrum appears. Then H-aggregate peak appears [23]. Therefore, it was concluded that the aggregation of excess numbers of MB þ molecules in gelatin takes place.

ΔE ¼

ℏ2 π 2

2μR

2



1:78e2  0:248EnRy εR

ð2Þ

ΔE ¼ ℏω  Eg , ℏω ¼ 3:26 eV, Eg ¼ 2:36 eV [25], R is the QD radius,

mne ¼ 0:205m0 and mnh ¼ 1:6m0 are effective masses of electron and hole, respectively [25,26], ε ¼ 5:5 [26,27], EnRy ¼ e4 =2ε2 ℏ2 ðmen  1 þ mhn  1 Þ is the effective Rydberg energy. Produced estimations gave a value of average diameter close to those obtained by XRD pattern and TEM images. It is 2.8 70.2 nm. UV–vis absorption spectrum of MB þ cation in gelatin has two absorption peaks. A long-wavelength peak is located at 664 nm. It corresponds to the allowed S0 -Sn1 transition of MB þ of π-πn type [23]. In addition, the peak at 290 nm and the feature at 340 nm also correspond to the allowed S0 -Sn2 transitions apparently of σ-σn type [28]. For QD–MB þ mixtures the resulting absorption spectrum is the sum of QDs and MB þ absorption spectra. It has three distinct peaks. They are λ1 ¼290 nm, λ2 ¼380 nm, λ3 ¼657 nm. Blue shift of MB þ absorption peak to 657 nm compared to the dye spectrum in gelatin indicates the interaction and participation of MB þ π-electrons in it. On the contrary, there is no shift of the absorption peak at 290 nm. Apparently, this is due to localization of MB þ σ-electrons on those bonds groups in the molecule, which are not involved in QD–MB þ associates formation. 3.3. Luminescence spectra Fig. 4 shows normalized luminescence spectra of QD–MB þ associates, for which were shown above absorption spectra. QDs are characterized by a broad luminescence band with a peak at 580 nm. It has a significant full width at half maximum and large Stokes shift (200 nm) relative to the exciton absorption peak. Comparison with the available empirical data on the nature of recombination luminescence of CdS nanocrystals [29] gives evidence that the observed emission is attributed to donor–acceptor recombination according to the Williams–Prener mechanism [30]. QDs size distribution makes a significant contribution to the width of luminescence peak. In our case it is about 20%. The absence of

3.4. Luminescence decay Fig. 5 shows normalized curves of luminescence decay of CdS QDs and QD–MB þ associates for different ratios between components at

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

215

Fig. 5. Luminescence decay curves: (a) fixed amount of CdS QDs and various concentrations of MB molecules for a wavelength luminescence of 580 nm; (b) fixed amount of MB molecules and various concentrations of CdS QDs for a wavelength luminescence of 674 nm.

Table 1 Parameters of fitting of luminescence decay, estimated using Eqs. (3) and (5). The sum of five exp. functions

Stretched exponential function τn , ns

CdS QDs

2.197 0.18

β

0.277 0.01

〈τ〉, ns

kst ; ns  1

34.0 7 0.8 0.086 7 0.005

CdS QDs–MB þ

1.0770.12

0.317 0.01

8.6 7 0.3

two wavelengths. The first wavelength is 580 nm. It is a peak of QDs luminescence. The second is a peak of MB þ fluorescence (and edge of QDs luminescence). It is seen that when [nMB þ ] fraction increases in [nCdS QDs ]:[nMB þ ] concentration ratio the luminescence lifetime of QDs decreases. These changes are especially significant when 1–3 MB þ molecules correspond to one QD in the associate. The decrease in the luminescence lifetime of donor (in this case it is QDs) is typical of systems with energy transfer to acceptor (Dye) [34]. This conclusion is also confirmed by the data presented in Fig. 5. Indeed, when [nCdS QDs ] increases in [nCdS QDs ]:[nMB þ ] concentration ratio the luminescence lifetime of acceptors increases. It reaches its maximum when [nCdS QDs ]:[nMB þ ]¼ 10:1. For a detailed analysis of experimental curves in Fig. 5a we used some of the most widespread fitting functions. The first is stretched exponential function (Williams–Watts function) [35–37] " IðtÞ ¼ exp 


β # t

τn



 τ 1

β

Γ

β

τi , ns

0.05 7 0.005 0.096 7 0.005 0.08 7 0.006 0.2647 0.015 0.517 0.003 0.025 7 0.001 0.059 7 0.001 0.0627 0.002 0.282 7 0.017 0.5757 0.027

232 7 16 897 7.9 24.87 2.1 2.7 7 0.3 0.6 7 0.03 171.6 7 12.3 61.3 74.7 16.2 7 1.6 2.377 0.24 0.55 7 0.03

〈τ〉, ns

ksum ; ns  1

23.2 7 0.8 0.058 7 0.004

9.9 7 0.5

luminescence lifetime is 34.07 0.8 ns. For QD–MB þ associates this value is 8.6 70.3 ns. Also the fitting was realized using superposition of several exponential functions [38] 5

IðtÞ ¼ ∑ ai  exp½  t=τi  i¼1

ð5Þ

The fitting of luminescence decay by the sum of five exponential functions for CdS QDs and QD–MB þ associates gives a high veracity of approximation. χ2 parameters are 98.7% and 98.1% for CdS QDs and QD–MB þ associates, respectively. The average lifetime was estimated using discourse given by Tachiya et al. [39], based on the well-known equation for the mean value: Z 1 〈τ〉 ¼ t  f ðtÞ  dt ð6Þ 0

ð3Þ

where β is the nonexponential parameter, 0 o β o 1. The fitting by this function gives a low veracity for experimental data. χ2 parameter is 68.8% for CdS QDs, and it is 76.8% for QD–MB þ associates. The average luminescence lifetime for this approximation was estimated using Eq. (32) in [35]: 〈τ〉 ¼

ai

ð4Þ

Parameters, obtained for the case of stretched exponential function, are presented in Table 1. For CdS QDs at 580 nm the average

where f ðtÞ  dt is the possibility that a QD excited at time zero will decay between t and t þ dt. Turning to the measured luminescence decay of QDs, and QD-MB þ associates, we obtain
  dIðtÞ=dt  dt ð7Þ f ðtÞ  dt ¼ Ið0Þ In view of these arguments Eq. (2.9) in [39] is  Z 1
1 dIðtÞ 〈τ〉 ¼  dt t  Ið0Þ 0 dt   Z 1 1 ½ tIðtÞ1 ¼ IðtÞ dt 0 þ Ið0Þ 0 Z 1 1 ¼ IðtÞ dt Ið0Þ 0

ð8Þ

216

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

When going from the first line to the second line of Eq. (8), integration by parts is used. And for the case of fitting by Eq. (5) we obtain 〈τ〉 ¼

∑5i ¼ 1 ai  τi ∑5i ¼ 1 ai

ð9Þ

For the case of the sum of five exponential functions parameters of fitting are also presented in Table 1. For CdS QDs at 580 nm the average luminescence lifetime is 23:2 70:8 ns. For QD–MB þ associates this value is 9:9 7 0:5 ns. 3.5. Discussion of obtained results Experimental features of dynamic quenching of CdS QDs luminescence in QD–MB þ hybrid associates were found as a result of studies. This conclusion is confirmed by quenching of QDs luminescence, excited by 405 nm. Simultaneously the intensity of luminescence peak of MB þ cations in QD–MB þ associates increases (Fig. 3, inset). It is increased when the MB þ concentration in [nCdS QDs ]:[nMB þ ] ratio increases. If CdS QDs luminescence quenching during forming of hybrid associates is static nature then there should be no increase in the intensity of MB þ luminescence [34]. Pronounced spectral dip at 640–660 nm, which becomes stronger when [nMB þ ] concentration in QD–MB þ associates increases, is also caused by dynamic quenching. In this case CdS QDs act as donors of electron excitation, and MB þ molecules act as acceptors. The spectral dip is also due to energy transfer in components of QD–MB þ associates. It is necessary to note that according to the estimations made in Ref. [29] the broadening of CdS QDs luminescence spectrum is mainly due to their size dispersion in studied ensembles. Then the maximum overlap integral of absorption spectrum of (MB þ ) acceptors and donor emission spectrum (QDs) corresponds to the largest values of QDs size (2.8–3.8 nm) in the ensemble. They have luminescence at 580–700 nm and most efficiently energy transfer to MB þ . In contrast, QDs with smaller size produce luminescence in more short-wave part (450–580 nm) of total luminescence spectrum with a peak at 580 nm. The energy transfer from this part of ensemble is less effective. The most important arguments in favor of conclusion about resonance energy transfer in QD–MB þ associates are pronounced speed-up of QDs luminescence decay and its slowing for MB þ . These regularities are enhanced when [nMB þ ] concentration in QD–MB þ associates increases (Fig. 5). Time-resolved experiments clearly show that LRET for investigated simples is not observed. This conclusion is confirmed by analysis of luminescence excitation spectra of MB þ in QD–MB þ associates (Fig. 3). Thus, obtained experimental data allow us to conclude in favor of resonance nonradiative energy transfer from donor (CdS QD) to acceptor (MB þ ) (Fig. 6). It should be noted that in our case downright donor is the center of recombination luminescence. The radiative recombination on this center occurs according to the donor–acceptor mechanism. In this case donor and acceptor are defects of crystal lattice. And the resonance condition is nearness of QD recombination energy to energy of light absorption by the MB þ cation. This feature is fundamentally different from cases discussed in Refs. [3,10,13–20]. However, the overall characteristics of observed spectral and kinetic regularities are similar. Also it should be noted that the available experimental data suggest a low probability of CdS QDs luminescence quenching according to PET mechanism. At the same time, the relatively low position of fundamental electron state of MB þ (HOMO) could contribute to this. Indeed, the potential of MB þ photoionization is close to 5.9–6.0 eV [3]. The electron affinity energy for CdS is also

Fig. 6. Model of nonradiative resonance energy transfer from donor (QD) to acceptor (MB þ ) for QD–MB þ hybrid associates.

close to 5.9 eV [40]. The quantum size effect for holes of CdS QDs (2.5 nm) is located in the range of 0.1 eV. Also it should be taken into account that the acceptor level of luminescence center lies higher than CdS QDs state valence band of 0.7–0.8 eV. Therefore, the injection only from donor level is possible. But in this case, the reduction of MB þ leads to MB0 formation. The absorption and luminescence peaks of this MB form are situated in another spectral region [12]. The observed spectral dip at 640–660 nm is not situated in the MB0 absorption region. The process of electron– hole recombination is possible only when injection of charge carriers from quantum levels of QDs to MB þ electron states takes place. However, this process should be faster than 100 ps. This is the time, when exciton destruction occurs in used CdS QDs, due to the capture of charge carries on levels of structural-impurity defects, including luminescence centers. This conclusion was made using available data, obtained by the pump-probe [31]. Characteristic times of luminescence decay, showing exchange of electronic excitation in QD–MB þ hybrid associates, are 1–1500 ns. When there is PET from excited CdS QDs to MB þ cation then QDs luminescence will also be extinguished. But the spectral dip should be absent for PET. Thus, only one of the possible competing processes is realized when CdS QDs are excited. Using data on the luminescence decay of CdS QDs in CdS–MB þ associates we estimated the efficiency of nonradiative energy transfer from donor to acceptor (Fig. 6). The parameters of electron excitation energy transfer are estimated by several ways for comparison. First, we use the equation for efficiency of nonradiative transfer of electron excitation energy from donor to acceptor within the FRET framework [34], taking into account the average luminescence lifetime (see Table 1) for both fitting functions:

ϕkin ET ¼ 1 

〈τDA 〉 〈τD 〉

ð10Þ

where 〈τDA 〉 is the average luminescence lifetime for associate, 〈τD 〉 is the average luminescence lifetime for “pure” QDs. For the case of stretched exponential function the transfer efficiency is 75%, for the sum of five exponential functions it is 57%. Both these values show significant efficiency of electron excitation energy transfer from CdS QDs to MB þ molecules. Substantial progress has been made in the approach developed by Tachiya [41–43] for the description of the luminescence decay by taking into account that the quenching of an excited QD by intrinsic traps and added ligands which are both distributed among QDs according to Poisson distributions. In our case a complex non-exponential luminescence decay curve can be explained on the basis of the Tachiya model by taking into account two types of intrinsic traps

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

217

Table 2 Approximation parameters for CdS QDs and their associates with MB. Simple CdS QDs CdS QDs-MB

þ

I n ð0Þ

τ; nsðk0 ; ns  1 Þ

mt

0.98

136.9 (0.0073) 136.9 (0.0073)

0.98

mt

0.680

0.043

1.27

0.68

–

–

0.9989

0.680

0.043

1.27

0.68

1.05

0.056 7 0.004

0.9826

and added ligand MB þ . In the absence of added ligand an excited QD is quenched by two types of intrinsic traps and the decay of luminescence is described by Iðt; mÞ ¼ I n ð0Þ  exp½ k0 t mt1  ð1  exp½  kqt1  tÞ  mt2  ð1  exp½  kqt2  tÞ

ð11Þ

where k0 is the unimolecular decay rate constant of an excited QD, kqti (i ¼ 1; 2) is the quenching rate constant of an excited QD by an intrinsic trap of type i ði ¼ 1; 2Þ, mti ði ¼ 1; 2Þ is the mean number of intrinsic traps of type i ði ¼ 1; 2Þ in a QD. In the presence of added ligand MB þ the decay of luminescence is described by Iðt; mÞ ¼ I n ð0Þ  exp½ k0 t  mt1  ð1  exp½  kqt1  tÞ  mt2  ð1  exp½  kqt2  tÞ  mMB þ  ð1 exp½  kMB þ  tÞ

ð12Þ

where kMB þ is the energy transfer rate constant from an excited QD to a ligand MB þ , and mMB þ is the mean number of ligands attached to a QD. The values of the parameters obtained by fitting Eqs. (11) and (12) to the observed decay curves are presented in Table 2. It should be noted that the use of two types of intrinsic traps in “pure” QDs is an assumption. It will be tested in future our research. However, now we can say that several types of quenchers for “pure” QDs are justified by the presence of several types of defects [31]. Under the above considerations it was found that the value of transfer rate constant (kMB), obtained using the Tachiya model, is in good agreement with data, calculated using average luminescence lifetime, according to the following equation: 1 1 ¼ þ ktransf 〈τDA 〉 〈τD 〉

ð13Þ

For the case, described by Eq. (3), the transfer rate constant kst ¼ 0:086 ns  1 . For a sum of exponential functions (5) kst ¼ 0:058 ns  1 . Thus, all three relations for energy transfer from QDs to MB þ give almost the same constant value. It indicates the high veracity of chosen models. It should also be noted that the average number of dye molecules nMB, obtained by fitting of luminescence decay curves using the expression (12) for QD– MB þ associates, gave a value of 1.05. It is very close to the real ratio of samples ½nCdS QDs  : ½nMB þ  ¼ 1 : 1, for which estimation was realized.

4. Conclusion QD–MB þ hybrid associates of semiconductor colloidal CdS QDs with an average diameter of 2.5 nm with methylene blue molecules were investigated. It was shown that effective excitation of MB þ from the QDs absorption region due to nonradiative resonance energy transfer is possible for these systems. When [nMB þ ] concentration in QD–MB þ associates increases, the enhancement of luminescence quenching of QDs and increase of MB þ luminescence are observed in the emission spectrum. The average luminescence lifetime was estimated using two fitting functions. A decrease in luminescence lifetime of QDs at 580 nm and an

2

kqt2 ; ns  1

nMB

kMB ; ns  1

χ2

kqt1 ; ns  1

1

increase in luminescence lifetime of MB þ at 674 nm were shown for associate. According to these data the rate constant of nonradiative energy transfer from QD to MB þ in associate was estimated. It is 5.8–8.6  107 s  1. Using the Tachiya model for description of nonexponential luminescence decay curves of QDs the energy transfer rate constant was obtained. It is 5:6  107 s  1 . Thus, all three approaches for the description of luminescence decay give similar values of the rate constant of nonradiative energy transfer.

Acknowledgments The work was supported by RFBR (Project 14-02-31646-Mol_a, 12-02-90827-mol_RF_nr, and 13-02-90750-mol_RF_nr), the Russian Ministry of jobs within the public institutions of higher education in the field of scientific work on the 2014–2016 years (Project 1230). References [1] M. Bruchez Jr., M. Moronne, P. Gin, Sh. Weiss, A.P. Alivisatos, Science 281 (2010) 2013. [2] S. Coe, W.-K. Woo, M. Bawendi, V. Bulovic, Nature 420 (2002) 800. [3] A. Rakovich, D. Savateeva, T. Rakovich, J.F. Donegan, Y.P. Rakovich, V. Kelly, V. Lesnyak, A. Eychmuller, Nanoscale Res. Lett. 5 (2010) 753. [4] X. Ren, Z. Chen, X. Chen, J. Liu, F. Tang, J. Lumin. 145 (2014) 330. [5] C. Wang, H. Zhang, J. Zhang, M. Li, H. Sun, B. Yang, J. Phys. Chem. C 111 (2007) 2465. [6] P. Juzenas, W. Chen, Y.-P. Sun, M.A.N. Coelho, R. Generalov, N. Generalova, I.L. Christensen, Adv. Drug Deliv. Rev. 60 (2008) 1600. [7] D.E. Prasuhn, A. Feltz, J.B. Blanco-Canosa, K. Sosumu, M.H. Stewart, B.C. Mei, A.V. Yakovlev, Ch. Loukou, J.-M. Mallet, M. Ohem, Ph.E. Dawson, I.L. Medintz, ACS Nano 4 (2010) 5487. [8] K.-L. Chou, H. Meng, Y.C.L. Li, J.-Y. Chen, J. Nanopart. Res. 15 (2013) 1348. [9] J.P. Tardivo, A.D. Giglio, C.S. de Oliveira, D.S. Gabrielli, H.C. Junqueira, D.B. Tada, D. Severino, R.de F. Turchiello, M.S. Baptista, Photodiagn. Photodyn. Therapy 2 (2005) 175. [10] M. Idowu, T. Nyokong, J. Lumin. 129 (2009) 356. [11] M. Nyk, K. Palewska, L. Kepinski, K.A. Wilk, W. Strek, M. Samoc, J. Lumin. 130 (2010) 2487. [12] B. Burfeindt, T. Hannappel, W. Storck, F. Willig, J. Phys. Chem. 100 (1996) 16463. [13] M.-X. Zhao, H. Su, Z.-W. Mao, L.-N. Ji, J. Lumin. 132 (2012) 16. [14] O. Adegoke, T. Nyokong, J. Lumin. 146 (2014) 275. [15] A. Baride, D. Engebretson, M.T. Berry, S. May, J. Lumin. 141 (2013) 99. [16] X. Fei, Y. Gu, J. Wang, G. Jia, Zh. Liu, J. Lumin. 131 (2011) 291. [17] M.A. Shivkumar, K.S. Adarsh, S.R. Inamdar, J. Lumin. 143 (2013) 680. [18] M.F. Frasco, V. Vamvakaki, N. Chaniotakis, J. Nanopart. Res. 12 (2010) 1449. [19] Ch. Huang, Ch. Wu, Y. Zhao, J. Nanopart. Res. 12 (2010) 2153. [20] L. Pecci, G. Montefoschi, A. Antonucci, M. Costa, D. Cavallini, Biochem. Biophys. Res. Commun. 301 (2003) 411; J. Chen, Th.C. Cesario, P.M. Rentzepis, J. Phys. Chem. A 115 (2011) 2702. [21] O.V. Ovchinnikov, M.S. Smirnov, B.I. Shapiro, A.N. Latyshev, T.S. Shatskikh, A. N. Latyshev, Pham Thi Hai Mien, V.Yu. Khokhlov, Opt. Spectrosc. 115 (2013) 340. [22] O.V. Ovchinnikov, M.S. Smirnov, T.S. Shatskikh, V.Yu. Khokhlov, B.I. Shapiro, A. G. Vitukhnovsky, S.A. Ambrozevich, J. Nanopart. Res. 16 (2014) 2286. [23] K. Bergmann, C.T. O'Konski, J. Phys. Chem. 67 (1963) 2169. [24] Y. Kayanuma, Phys. Rev. B 38 (1988) 9797. [25] Y.-M. Yu, K.-M. Kim, S.-K. Lee, Y.D. Choe, P.Y. Yu, J. Appl. Phys. 92 (2002) 1162. [26] S.I. Pokutnyi, Phys. Status Solidi B 173 (1992) 607. [27] Y. Wang, N. Herron, Phys. Rev. B 42 (1990) 7253. [28] G.N. Lewis, J. Begeleisen, J. Am. Chem. Soc. 65 (1943) 1144. [29] A.I. Ekimov, I.A. Kudryavtsev, M.G. Ivanov, Al.L. Efros, J. Lumin. 46 (1990) 83. [30] J.S. Prener, F.E. Williams, J. Phys. Rev. 101 (1956) 1427. [31] M.S. Smirnov, D.I. Stasel'ko, O.V. Ovchinnikov, A.N. Latyshev, O.V. Buganov, S.A. Tikhomirov, A.S. Perepelitsa, Opt. Spectrosc. 115 (2013) 651.

218

M.S. Smirnov et al. / Journal of Luminescence 156 (2014) 212–218

[32] O.Ye. Rayevska, G.Ya. Grodzyuk, V.M. Dzhagan, O.L. Stroyuk, S.Ya. Kuchmiy, V.F. Plyusnin, V.P. Grivin, M.Ya. Valakh, J. Phys. Chem. C 114 (2010) 22478. [33] B.S. Fujimoto, J.B. Clendenning, J.J. Delrow, P.J. Heath, M. Schurr, J. Phys. Chem. 98 (1994) 6633. [34] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, 3rd ed., Springer, New York, 2006. [35] C.P. Lindsey, G.D. Patterson, J. Chem. Phys. 73 (1980) 3348. [36] L.E. Shea-Rohwer, J.E. Martin, D.F. Kelley, J. Electrochem. Soc. 157 (2010) J1.

[37] [38] [39] [40]

T. Uchihara, H. Kato, E. Miyagi, J. Photochem. Photobiol. A 181 (2006) 86. L. Spanhel, M.A. Anderson, J. Am. Chem. Soc. 112 (1990) 2278. H. Sano, M. Tachiya, J. Chem. Phys. 75 (1975) 2870. X.-Q. Dai, W.-Ch. Lu, Z.-X. Yang, Sh.-Y. Wei, T. Zang, Commun. Theor. Phys. 26 (1996) 257. [41] M. Tachiya, Chem. Phys. Lett. 33 (1975) 289. [42] M. Tachiya, J. Chem. Phys. 76 (1982) 340. [43] S. Sadhu, M. Tachiya, A. Patra, J. Phys. Chem. C. 113 (2009) 19488.