Optical Materials xxx (2014) xxx–xxx Contents lists available at ScienceDirect Optical Materials journal homepage: www.elsevier.com/locate/optmat Synthesis and optical properties of core/shell ternary/ternary CdZnSe/ZnSeS quantum dots Nguyen Hai Yen a, Willy Daney de Marcillac b,c, Clotilde Lethiec b,c, Phan Ngoc Hong a,b,c, Catherine Schwob b,c, Agnès Mtre b,c, Nguyen Quang Liem a, Le Van Vu d, Paul Bénalloul b,c, Laurent Coolen b,c, Pham Thu Nga a,e,⇑ a Institute of Materials Science (IMS), Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet, Cau Giay, Hanoi, Viet Nam Sorbonne Universités, UPMC Univ Paris 06, UMR 7588, Institut de NanoSciences de Paris (INSP), Paris F-75005, France CNRS, UMR 7588, Institut de NanoSciences de Paris (INSP), Paris F-75005, France d Center for Materials Science, University of Natural Science, VNUH, 334 Nguyen Trai, Hanoi, Viet Nam e Duy Tân University, Danang, Viet Nam b c a r t i c l e i n f o Article history: Received 25 December 2013 Received in revised form March 2014 Accepted 16 April 2014 Available online xxxx Keywords: Ternary QDs Alloyed QDs CdZnSe/ZnSeS core/shell QDs Single-photon emitter a b s t r a c t In this paper we report on the synthesis of ternary/ternary alloyed CdZnSe/ZnSeS core/shell quantum dots (QDs) by embryonic nuclei-induced alloying process We synthesized CdZnSe core QDs emitting in the spectral range of 530–607 nm with various Cd/Zn ratios, depending on the core synthesis temperature By shelling ZnSeS on the CdZnSe core QDs, the average luminescence quantum yield is increased by a typical factor of up to 17, which we attribute to the reduction of number of non-emitting QDs The singlephoton emitter micro-photoluminescence study showed that the CdZnSe/ZnSeS core/shell QDs are good single-photon emitters and their blinking properties were improved compared to the CdZnSe core QDs Quantum yields up to 25% were measured for the core/shell samples, demonstrating the potential for high-quality ternary/ternary QDs fabrication Ó 2014 Elsevier B.V All rights reserved Introduction Semiconductor QDs have demonstrated their nice photoluminescence (PL) properties, such as tuning of the emission through the QD diameter, with applications for multiplexed biolabeling [1,2], solid-state lighting and display [3], sensors for agriculture [1,4–6] and single-photon emission [7–9] Among various QDs, CdSe QDs have been studied most extensively [10–12] By shelling CdSe core QDs with ZnS [13,14] or CdS [2,15], non-radiative decay channels related to surface states could be passivated, consequently increasing the core/shell QDs luminescence quantum yield (LQY), which reaches almost 100% in some cases [16] However, though CdSe/ZnS QDs emit strong luminescence, it also exhibit random ‘‘blinking’’ between emitting ‘‘on’’ states and non-emitting ‘‘off’’ states [17] To eliminate the blinking in a core QDs, a thick suitable shell could be applied For CdSe QDs, the thick CdS shell was used to give a quasi-suppression of the blinking [2,18–20] ⇑ Corresponding author at: Institute of Materials Science (IMS), Vietnam Academy of Science and Technology (VAST), 18 Hoang Quoc Viet, Cau Giay, Hanoi, Viet Nam Tel.: +84 90 41 20 471; fax: +84 43 73 45 895 E-mail addresses: phtnga@ims.vast.ac.vn, ngalamvn@gmail.com (P Thu Nga) Many applications of QDs such as solid-state lighting or biological labeling require tuning the emission over the whole visible spectrum This is theoretically possible with the well-controlled CdSe QDs size; however for emission in the blue–green spectral region it requires very small CdSe QDs, which are difficult to be synthesized with well-passivated surface and homogeneous size, resulting in low LQY and broad emission [21,22] Alternatively, considering alloyed QDs, e.g., various ternary CdZnSe [21,25–31], CdSeTe [24] and CdZnS [23,32] which have been synthesized, the violet–green region is accessible because the emission wavelength depends on their compositions The tunable emissions from the mentioned QDs were reported in the spectral ranges of 360–500 nm [27], 390–580 nm [26], 415–460 nm [32], 535– 620 nm [21], respectively Controlling the synthesis of ternary QDs is however a delicate question Different methods have been presented for the synthesis of alloyed CdZnSe QDs, and in particular Zhong et al distinguished two procedures [25,26]: for the first one called ‘‘embryonic nuclei-induced alloying process’’ [25], binary seeds are formed by injection of the first cationic precursor (either Cd or Zn), then the ternary QDs are grown by increase of the temperature and injection of the second cationic precursor For the second procedure called ‘‘cationic exchange process’’ [26], http://dx.doi.org/10.1016/j.optmat.2014.04.020 0925-3467/Ó 2014 Elsevier B.V All rights reserved Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx binary QDs, for instance ZnSe, are first synthesized, then the second cationic precursor Cd2+ is injected and mixed with ZnSe so that CdZnSe QDs are obtained The obtained crystalline structures were found to be hexagonal for the former synthesis [25] and zincblende for the latter [26] The growth of the ZnS or ZnSe shell on the alloyed QDs was reported, resulting in the core/shell ternary/binary structures [29,32] For CdZnSe/ZnSe QDs, a suppression of blinking of the emission at 600–650 nm was demonstrated that is attributed to a gradient of the composition of alloy [31] CdZnSe QDs have also been used for the fabrication of light-emitting diode (LED) with the spectral properties better than those of CdSe/ZnS QDs [33] To improve the spectral characteristics and optical properties, binary/ternary core/shell structure like CdSe/CdZnS QDs or core/multishell one like CdSe/(CdS/ZnCdS/ZnS) QDs have also been performed [34,35] However, to our knowledge, there has been no report on the synthesis of ternary/ternary core/shell QDs This is possibly due to the particular difficulty in controlling the synthesis conditions of ternary/ternary QDs One can expect the improvement of the optical properties from ternary/ternary core/shell QDs because of reduction of the lattice mismatch between the core and the shell that are resulted from the smooth change in the composition at the interface In this paper, we report on the synthesis of ternary/ternary CdZnSe/ZnSeS core/shell QDs The size and alloy composition, as well as the photophysical properties of the QDs obtained at different synthesis temperatures are discussed As the synthesis temperature increased from 285 °C to 310 °C, CdZnSe QDs obtained could emit strong photoluminescence (PL) ranging from 530 nm to 607 nm We then have studied the ZnSeS shell growth with different Se/S precursor ratios After shelling CdZnSe core QDs with ZnSeS, increases of the LQY by 2–16 times and values of LQY up to 25% have been determined for the CdZnSe/ZnSeS core/shell QDs The PL spectra and PL decay-times from the core/shell QDs were measured and analyzed Finally, single-photon emitter microscopy was performed showing much suppression of the PL blinking in the CdZnSe/ZnSeS core/shell QDs Experimental 2.1 Chemicals We used the following reagents (from Aldrich): cadmium acetate (Cd(Ac)2, 99.9%), zinc acetate (Zn(Ac)2, 99.9%), selenium powder (Se, 99.99%), hexamethyl disilthiane (TMS)2S, trioctylphosphine oxide (TOPO, 99%), trioctylphosphine (TOP, 90%) and hexadecylamine (HDA, 99%) All chemicals were used without further purification 2.2 Synthesis of the ternary alloyed CdZnSe core QDs For the synthesis of the ternary core QDs we used the embryonic nuclei-induced alloying procedure described in [25], with initial growth of ZnSe seeds and subsequent growth of the CdZnSe QDs Our choice of the precursors was modified with respect to Ref [25], so that the reaction conditions were adjusted accordingly All synthetic routes were carried out using standard airless procedures by filling the reaction flask by ultra-pure nitrogen gas flow In order to fabricate the cadmium stock solution, we dissolved 0.025 g of cadmium acetate in 0.54 ml TOP at 80 °C in nitrogen gas Similarly, we obtained the zinc stock solution by dissolving 0.0875 g of zinc acetate in ml TOP at 140 °C in nitrogen gas atmosphere, and we obtained the TOP–Se precursor by dissolving 0.135 g of Se in 1.665 ml of TOP at 120 °C in nitrogen gas The molar ratios of precursors were thus the same for all samples and equal to 0.2/0.8 for the Cd/Zn ratio and 1/3.3 for the (Cd + Zn)/Se ratio Briefly, 3.325 g of TOPO and 1.6625 g of HDA were poured into a three-neck reaction flask Nitrogen gas was used to remove water vapor and oxygen out of the reaction flask at room temperature for 30 min, then at 120 °C for one hour We first injected the TOP–Se precursor into the flask under vigorous stirring and heating at temperatures up to 100 °C in nitrogen gas atmosphere Kept stirring and heating the reactor up to 190 °C, and at that moment zinc stock precursor solution was injected into the reaction flask in order to form the ZnSe nanocrystallite seeds Then temperature was increased up to 280 °C, at that time we injected the cadmium stock solution into the reactor As the temperature of the liquid in the reaction flask dropped to $260 °C, the nucleation of the alloyed CdZnSe QDs started shaping quickly The alloyed CdZnSe QDs were grown for typically 20 at different temperatures of 285, 300 and 310 °C for samples A–C, respectively 2.3 Growth of the ternary alloyed ZnSeS shell on the CdZnSe core QDs The shells were grown following a modified version of the successive ion layer adsorption and reaction (SILAR) procedure from Refs [34,36], originally described by Li, Peng and colleagues [15] The Zn precursor stock solution was prepared by mixing 0.165 g of zinc acetate and 1.88 ml of TOP, then heating to $140 °C in nitrogen gas atmosphere until a clear solution was formed For the TOP–Se, we dissolved 0.026 g Se in 0.33 ml TOP, at a temperature of $120 °C in nitrogen gas (it is important that the solution should be clear after it is cooled down to room temperature) and mixed it with (TMS)2S Different values of the Se/S molar ratio (noted x/1Àx) were considered: x = 0.2, 0.4, 0.6 and 0.8 The molar ratio of the precursors for growing the ZnSeS shell was always to be Zn/(Se + S) = 1.37/1 For shelling, 0.25 mM of the as-prepared CdZnSe ternary QDs in TOPO and HDA was filled into the reaction flask with nitrogen gas protection For one hour at 50 °C the evacuation of the air and the re-fill of nitrogen gas were done several times to clean the atmosphere from the air Then, temperature was raised to 240 °C for shelling with the precursor injection in five steps, each step consisting in consecutive addition of metal then chalcogenide addition: for each step the zinc-stock solution was injected dropwise with a rate of 1–2 drops per second (a drop is of 50 ll), under vigorous stirring Then, the mixture of (TMS)2S and TOP–Se was added and the shelling temperature was kept at 240 °C for 15 under vigorous stirring The shell growth temperature was chosen sufficiently high for a good mixture of the ZnSeS alloy but 45–70 °C below the core growth temperature in order to avoid any further growth of the core QDs or diffusion between the core and shell The volume of precursor solution injected at each step was calculated in order to match the reaction stoichiometry to the area of QD surface to be covered, as described in [37] After the last step, the heater was removed and the mixture was cooled down to stop the reaction When the temperature reached 70 °C, the core/shell QDs were dispersed in organic solvent (toluene or choloroform) 2.4 Characterization of ternary CdZnSe/ZnSeS core/shell QDs The size of the core QDs and the shell thickness were determined by the transmission electron microscopy (TEM) with a JEOL Jem 1010 microscope operating at 100 kV The powder X-ray diffraction (XRD, Siemens D5005) was used to confirm the wurtzite (w) or zinc-blende (zb) crystalline structure The XRD patterns were compared with the tabulated values of bulk CdSe (JCPDS card No 19-191 (zb) and 8-459 (w)), ZnSe (JCPDS 37-1463 (zb) and 15105 (w)) and ZnS (JCPDS 5-566 (zb) and 39-1363 (w)) The energy- Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx dispersive X-ray spectroscopy (EDS) was used to check the presence of the Cd, Zn, Se and S elements in our ternary QD samples For optical characterizations, all the QD samples were diluted in toluene In the PL measurements a pulsed nitrogen laser (337 nm, pulse width 0.6 ns, repetition rate of 10 Hz) was use as the excitation source The PL from the samples was collected by an optical fiber on the same side as the excitation light, then was analyzed by a Jobin–Yvon Spectrometer HR460 and detected by a multichannel CCD detector (2000 pixel) The PL decays were analyzed with a PM Hamamatsu R5600U and a Tektronix TDS 784A scope with the time resolution of ns Finally, in order to measure the LQY, we measured for each sample the ratio of the integrated emission (under Cd–He laser 442-nm excitation) to the 442-nm absorption, and compared it with the ratio measured under the same conditions for a rhodamine 6G sample The concentration was adjusted in order to keep the absorption at 442 nm in the range 3–8% which was found adequate in our previous work [38] For individual QD measurements, the QDs were spin-coated on a glass coverslip and then covered them by a 50-nm layer of PMMA to protect them from the oxidation The sample was imaged with a microscope and a x100 oil-immersion objective with 1.45 numerical aperture On the one hand, for blinking studies (imaging tens of QDs at the same time), a portion of the sample was illuminated at 436 nm by a mercury-vapor lamp The sample was imaged by a CCD camera, with a pixel size of 6.3 lm corresponding to a resolution of 63 nm The CCD rate was of 10 frames per second On the other hand, for time-resolved studies, a single QD was excited by a 405-nm pulsed diode laser Its emission was selected by a pinhole located in the objective image plane, and detected by two avalanche photodiodes in Hanbury-Brown and Twiss configuration, connected to a Picoharp acquisition card The overall setup provided a 400-ps resolution Results for the ternary CdZnSe core QDs 3.1 Structural characterization We first characterize the CdZnSe core QDs, as obtained before the shell growth Various core growth conditions (temperature and reaction time) were used as summarized in Table We also report in Table the core diameter, ranging from 4.2 nm to nm as estimated from TEM images as shown in Fig For each sample, the measured core diameters typically displayed a ±0.5–1 nm inhomogeneity Let us point out that a significant portion of this measured inhomogeneity is introduced by the uncertainty of measuring the QD size on the TEM image, so that the actual QD size distribution is probably of 5–10% We also performed energy dispersive X-ray spectrometry (EDS) The results, reported in Table 1, show a decrease of the Zn/Cd ratio from A to C It also shows however some non-stoichiometric (Cd + Zn)/Se values Similar results were reported in [35], showing that the ligand cleaning procedure preliminary to EDS measurement can alter significantly the composition of the surface of the QDs: for CdSe QDs, Cd/Se ratios between and 3.8 were measured and attributed to a lack of Se atoms on the QD surface due to Fig TEM images of nm (sample A), 4.2 nm (B) and 4.8 nm (C) CdZnSe core QDs sample cleaning (typically two thirds of the QD atoms are on the surface) [35] Although the quantitative compositions obtained from EDS may be affected by the cleaning procedure, EDS demonstrates the presence of the three elements Cd, Zn and Se in the QDs The crystalline structure of QDs alloy can be different from the bulk alloy and is difficult to be predicted Bulk ZnSe can have a zinc-blende (zb) phase or a metastable wurtzite (w) phase; on the other hand bulk CdSe can have a wurtzite phase or a metastable zinc-blende phase [34,39] In the literature, both wurtzite [24] and zinc-blende [27] structures have been reported for CdZnSe QDs For QDs grown on ZnSe seeds, a wurtzite structure was reported in [26], although a zinc-blende structure has also been found for ZnSe QDs [25] Experimentally, the analysis of the XRD data is complicated because of the broadening of the diffraction peaks due to the absence of long-range order Fig shows the XRD patterns of the three QD samples (synthesized with the conditions indicated in Table 1, namely A and B and C) and the tabulated diffraction lines of bulk CdSe and ZnSe Similar XRD patterns are obtained for the three samples synthesized at temperatures from 285 °C to 310 °C We find experimental XRD peaks between the tabulated peaks corresponding to the wurtzite phases of CdSe and ZnSe, which would be in agreement with a wurtzite CdZnSe alloy (although, given the width of the experimental spectra, we cannot exclude a polytype of zinc-blende and wurtzite phases) The wurtzite structure would be consistent with the structure reported in [26] with a ZnSe-seeded growth process Table Growth conditions (temperature and reaction time after TOP–Cd injection) and QD diameter (from TEM images) and Cd/Zn and (Cd + Zn)/Se ratios from EDS data of three samples of ternary CdZnSeQDs Sample name Reaction temperature (°C) Reaction time (min) QD diameter (nm) Cd/Zn amount (Cd + Zn)/Se amount A B C 285 300 310 20 20 17 5.0 4.2 4.8 0.3/0.7 0.4/0.6 0.5/0.5 3.3 1.6 Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx line widths are presented in Table 2, showing a large dependence on the QDs synthesis protocol The QD emission wavelength ranging from 530 nm for sample A to 607 nm for sample C could depend on both the alloy bandgap and on the QD diameter (d) However, given the similar sizes of these samples, we expect that most contribution to the optical transition energy comes from the change in the QDs compositions (Zn/Cd ratio) One can see also the larger Stokes shifts in samples B and C as compared to that of sample A that may originate from the localization of charge carriers in the alloy, as similarly observed in In(Zn)P QDs [40] The QD emission energy EQD has been described by ECdSe,QD = ECdSe,bulk + 1.83/d1.06 in Ref [41] for a CdSe QD and EZnSe,QD = EZnSe,bulk + 2.08/d1.19 in Ref [42] for a ZnSe QD For bulk Cd1ÀxZnxSe alloys, a quadratic relation can be used to describe the bulk alloy bandgap ECdZnSe,bulk as a function of the Zn fraction x These elements suggest that the QD emission energy ECdZnSe,QD (in eV) can be described by the following empirical relation: Fig Powder XRD patterns of CdZnSe ternary QD cores with different growth temperature, from 285 °C (A) to 310 °C (C) Bulk diffraction peaks for zinc blende (zb) and wurtzite (w) ZnSe (top) and zb and w CdSe (bottom) are indexed for identification purpose (for bulk CdSe [(JCPDS card 19-191 (zb) and 8–459 (w)] and bulk ZnSe [JCPDS card 37-1463 (zb) and 15–105 (w)]) 3.2 Optical properties Fig 3(a) plots the absorption and PL spectra of the three samples A–C of CdZnSe core QDs The absorption spectra display a clear excitonic peak showing high quality of the QDs and their corresponding energies are much higher than the bandgap energy of bulk CdSe The absorption and emission wavelengths and emission 1:06 ECdZnSe;QD ẳ ECdZnSe;bulk ỵ xị1:83=d 1:19 ỵ x2:08=d 1ị Kim et al have used this relation with ECdZnSe,bulk = 1.74Á (1Àx) + 2.6ÁxÀ0.35x(1Àx) to estimate x by using the emission wavelength and the TEM value of d [21] By using the same method, we find x of the order of 0.4, 0.1 and for samples A–C, respectively The latter value is surely false as we know from the EDS data that the QDs are CdZnSe alloys There is thus probably a bias on the x value obtained from the empirical Eq (1), and possibly due to errors on the measured d or on the bulk bandgap value, as different values are reported in the literature (1.74–1.8 eV for CdSe [43]) However, we can make the qualitative estimation, from Eq (1), that the amount of Zn is decreased in QDs synthesized at higher temperature, which would be in agreement with the composition values obtained by EDS (Table 1), while addition of the Cd precursor to ZnSe seeds induces more Cd as the temperature is increased Fig 3(b) shows the plots in ln scale of the PL decay curves from the three samples The PL decays are non-exponential with a large contribution of the fast decay component of sample A, which suggests an important contribution from non-radiative channels The decay curves of samples B and C are closer to an exponential decay process A reasonable fit is obtained with the stretched exponential function: exp(-(t/t0)b), with, for the core samples A–C, the respective characteristic times t0 = 6.3, 13 and 12.6 ns and exponents b = 0.48, 0.64 and 0.64 (plotted as dotted lines on fig 2(b)): similar parameters are found for the samples B and C, with shorter characteristic time for sample A and a lower b factor, indicating a stronger deviation from exponential shape Finally, the measured LQY are reported in Table The LQY is low for sample B (11%) and very low for sample A (2%), which confirms the presence of stronger non-radiative decay channels for sample A Both samples should benefit from addition of a capping shell in order to improve their LQY As a short conclusion for CdZnSe core QDs, we have found that (i) a lower reaction temperature leads to a higher Zn concentration indicated by a shorter emission wavelength and (ii) for the lowest temperature (sample A) larger non-radiative decay channels appear The improvements of the synthesis parameters in order Table Absorption peak, PL peak wavelength and PL line width in nm of the CdZnSe QDs Fig (a) Absorption (dotted lines) and photoluminescence (full lines) spectra of the three CdZnSe core samples: A–C (norm units) (b) Photoluminescence decay curves (in ln scale) of the samples A–C (full lines) and stretched-exponential fits (dotted lines) Sample (nm) A B C Absorption wavelength Emission wavelength Emission line width 526 530 34 554 570 35 588 607 43 Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx Table Properties of the CdZnSe/ZnSeS QDs, obtained from TEM images and ensemble optical measurements, as a function of the CdZnSe core sample (A and B) and the Se/S shell nominal ratio (1–4 corresponding respectively to a Se/S ratio of 0.2/0.8, 0.4/0.6, 0.6/0.4 and 0.8/0.2) The decay characteristics t0 (in ns) and b are obtained by fitting the decay curves with a stretched exponential: exp(À(t/t0)b) The LQY (in%) was obtained by comparison with a rhodamine 6G Ref [38] Sample Se/S precursor ratio Diameter (nm) Absorption wavelength (nm) Emission wavelength (nm) Emission line width (nm) t0(ns)/b decay factors LQY (%) A A1 A2 A3 A4 B B1 B2 B3 B4 – 0.2/0.8 0.4/0.6 0.6/0.4 0.8/0.2 – 0.2/0.8 0.4/0.6 0.6/0.4 0.8/0.2 5.0 6.0 5.5 5.9 6.3 4.2 5.1 5.1 5.3 4.4 526 536 – 536 – 554 556 553 554 555 530 560 589 562 578 570 572 567 569 571 to achieve both emission in the blue region and low non-radiative channels are under progress In the following we describe the synthesis of the core/shell QDs, focusing on core samples A and B which showed the shortest emission wavelengths and shelling with ZnSeS by different Se/S ratios Results for the ternary/ternary CdZnSe/ZnSeS core/shell QDs 4.1 Structural characterization Samples A and B were covered by a ZnSeS shell with, for each sample, different concentrations of the Se and S precursors The obtained core/shell samples are labeled 1–4 corresponding to the Se/S ratio of 0.2/0.8, 0.4/0.6, 0.6/0.4 and 0.8/0.2 from the precursors, respectively The diameters of QDs including core and shell are extracted from the TEM images and shown in Table The mean QD diameter is the same for all core/shell samples of the same series (the differences between samples of a same series are not significant given the 0.5–1 nm distribution of measured size in a given sample): about 6.0 nm for the samples series A and 5.2 nm for the samples series B, corresponding to a 1-nm shell for both series Sample B4 is an exception as the shell thickness measured was only 0.2 nm (the shell thickness is then lower than the measurement uncertainty, but EDS and XRD pattern have confirmed the presence of the ZnSeS shell) We have checked the structure of all samples prepared by using XRD method However, in Fig we represent only the XRD pat- Fig Powder XRD patterns of CdZnSe ternary QD cores (sample B) and CdZnSe/ ZnSeS core/shell (samples B1–B4) Bulk diffraction peaks for wurtzite ZnS (top), ZnSe (middle) and CdSe (bottom) are shown 34 46 63 44 61 35 37 38 37 36 6.3/0.48 3.3/0.50 4.4/0.55 3.9/0.64 5.4/0.6 13/0.64 11/0.6 14/0.62 10/0.59 8/0.62 14 26 4 11 18 21 25 terns of sample B (the CdZnSe core QDs) and their core/shell structures named as B1–B4 For a 4.2-nm core QDs with a 1-nm shell, we estimate by geometric considerations that about half of the atoms are located in the QD core and half in the QD shell The three peaks at 43°, 47° and 51° are characteristic to the wurtzite structure (the zinc-blende structure creates only two peaks) For the core, as discussed previously we find a polytype zinc-blende/ wurtzite structure (two peaks and a smaller one) For the core/shell samples, the structure is clearly wurtzite, indicating that the shell grows in wurtzite phase The peak positions for the core/shell QDs are shifted to higher angles as compared to the core sample because of the ZnSeS shell contribution with smaller lattice constant In this cases, some compression might happen in the core QDs Although the samples B1 – B4 are synthesized with very different Se/S precursor ratios, the XRD patterns indicate to be similar, suggesting that Se and S concentrations in the shell did not clearly influence to the structure of the shell 4.2 Optical properties of QDs ensemble The absorption and PL spectra, and the PL decay curves of all A, B core QDs and their core/shell structures are plotted in Fig The main spectral characteristics (peak positions and width, decay factors t0 and b obtained by a stretched-exponential fit) are reported in Table For the series A, the PL spectra of the core/shell QDs are broad and differently red-shifted as compared to the core QDs, namely sample A This is possibly due to the less stable structure of the core QDs synthesized at 285 °C We have observed similar effect with our CdZnSe/ZnS QDs (data not shown here), with a 16-nm red-shift by addition of the 0.5-nm ZnS shell and a 20-nm red-shift by addition of the 0.8-nm or 1.4-nm ZnS shell, because of the tunnel penetration of the charges into the shell which decreases the confinement energy The charges penetration inside the shell is expected to be larger for the ZnSeS shell than for the ZnS shell as ZnSeS has a smaller bandgap that explains the larger red-shift Indeed the redshift for samples A1–A4 (ZnSeS shell) ranges from 30 to 59 nm, larger than the 20-nm red-shift for the ZnS shell Optimizing the shell composition is necessary to compromise between the red-shift and the lattice mismatch For the samples with core B, the emission band from all samples is at the same wavelength (within 2–3 nm) and the emission line width is always 37 nm, indicating the stability of the core and good reproducibility of the shell The emission wavelength is exactly the same as for the CdZnSe core, which is in contrast with sample series A The presence of a significant ZnSeS shell in samples B1–B4 has been demonstrated with EDS, TEM and XRD One may suggest that the alloy composition is not uniform inside the core B, with a Zn-richer surface so that the electron–hole pair is better confined and the shell has less influence Alloy gradient effects have been reported previously for CdZnSe QDs [24] Another possibility Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx a maximum value of 26% This could be explained by an increase of the radiative rate; however an x1.5 increase of the total decay rate and a x17 quantum yield increase would correspond to a x25 increase in radiative decay rate, which would be quite unrealistic We will discuss this further in the next section and propose an explanation by the decrease of the non-radiative rate 4.3 Blinking properties from individual QDs To further characterize the photophysical properties, the QDs from various CdZnSe core and CdZnSe/ZnSeS core/shell samples of series A were spin-coated onto a glass substrate and imaged by a microscope onto a CCD camera With a proper dilution, the QD were separated by a few microns and could be observed individually Typical intensity fluctuations for QDs of sample A1 are plotted in Fig 6, with a 100-ms time resolution On these curves, the PL dynamics exhibit switching between a stable fluorescent ‘‘on’’ state and a non-fluorescent ‘‘off’’ state or the so-called blinking This blinking of the PL is well-known for CdSe/ZnS QDs [17] and has been attributed to the charge exchanges between Fig (a) Absorption (dotted lines) and PL (full lines) spectra of two CdZnSe samples and eight CdZnSe/ZnSeS samples (norm units) (b) PL decay curves of these samples (in ln scale; full lines) and fit by a stretched exponential (dotted lines) would be that the red-shift due to electron–hole penetration into the shell is balanced by a blue-shift caused by a diffusion of S into the CdZnSe core Such a blue-shift upon shell addition has been observed for CdZnS/ZnS QDs [32] Further characterization would be necessary to discriminate between these interpretations The decay curves of samples B1–B3 are similar to the decay curve of the core sample B (with B3 slightly slower) The decay is slightly faster for B4, suggesting an increase in defect contributions to the non-radiative decay, which may be explained by its thinner shell (0.2 nm instead of nm for the other samples, as measured by TEM) Values of the LQY are given in Table For series B, the quantum yield is improved by a factor of as compared to the core (with the exception of sample B4, which confirms that the quality of the shell is poorer for this sample) Quantum yields up to 25% are obtained On the other hand, the decay curves are not modified for B1–B3 as compared to the core, indicating that the total (radiative + non-radiative) decay rate is not modified, and suggesting that both the radiative and non-radiative rates are not modified The LQY being the ratio of the radiative to total decay rates, should then not be modified either However, it should be kept in mind that the QDs ‘‘blink’’ The decay curves originate only from the emitting ‘‘on’’ QDs, so they show that the decay rates are not modified for the emitting QDs The quantum yield measurement, on the other hand, is an average over all (‘‘on’’ emitting and ‘‘off’’ nonemitting) QDs, as it is the ratio of emission to absorption and all (‘‘on’’ and ‘‘off’’) QDs absorb The increase of the quantum yield upon shell addition may thus be attributed to a reduced blinking (less ‘‘off’’ QDs), without significant modification of the radiative and non-radiative decay rates of the ‘‘on’’ QDs For the samples with core A, all decays are about 1.5 times faster than for the core sample The LQY of the core/shell QDs is increased by a factor 2.6–17 with respect to the core sample, with Fig (a) PL dynamics of four typical A1 QDs; (b) distribution of the fraction of ‘‘on’’ times for sample A1; (c) distribution of the durations of each ‘‘on’’ (blue) and ‘‘off’’ (green) state duration (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx the QD core and trap states in its shell of environment The clear switching between two well-defined ‘‘on’’ and ‘‘off’’ states constitutes a reasonable photostability, as is usually observed for the high quality CdSe/ZnS QDs For each nanocrystal, we can define a threshold between ‘‘on’’ and ‘‘off’’ states so that the duration of the successive ‘‘on’’ and ‘‘off’’ states can be estimated For a collection of about 50 QDs for each sample, we determined the portion of the total acquisition time that was spent by each QD in the ‘‘on’’ state Average portions of time spent in the ‘‘on’’ emitting state for the three samples A, A1 and A4 have been determined to be 25%, 35% and 40%, respectively, showing that the shell has reduced the QD blinking by increasing the duration time in the ‘‘on’’ state This increase alone cannot account for the x2–x10 increase in LQY obtained by ensemble luminescence measurements (Table 3), but it has been shown that, in a typical QD sample, a large portion of the QDs are ‘‘dead’’ [39]: they never emit light because their non-radiative channels are always too high We can thus assume that shell addition results in a reduction of the portion of ‘‘dead’’ QDs If, for samples A, A1 and A4 the LQY in solution is respectively 2%, 14% and 4% while the QDs spent resp 25%, 35% and 40% of their time in the ‘‘on’’ state, we deduce that roughly resp 8%, 40% and 10% of QDs are not ‘‘dead’’, showing that shell addition can increase the number of QDs which are not ‘‘dead’’ up to times For sample A1, we also plot in Fig 6(c) the distribution of the ‘‘on’’ and ‘‘off’’ states duration (collected over 50 QDs) These plots, in log–log scale, agree reasonably with a power-law distribution, and they can be fit by a tÀ1.5 function for the ‘‘on’’ times and a tÀ1.4 function for the ‘‘off’’ times These power-law distributions (Levy laws) with the 1.5 coefficient are in very good agreement with previous results on the blinking statistics of CdSe/ZnS QDs [44] The blinking properties of the CdZnSe/ZnSeS QDs are thus qualitatively of same nature as for the CdSe/ZnS nanocrystals Various mechanisms have been proposed to explain this power-law distribution, involving either a distribution of trapping states with different trapping/detrapping rates [45] or the random-walk fluctuations of the emission wavelength [46] 4.4 Single-photon emitter time-resolved properties The PL properties of sample A1 were characterized with better (400 ps) time resolution by exciting a single nanocrystal with a pulsed 400-nm laser and detecting with a photon-counting avalanche photodiode Fig plots the detected intensity for a typical QD (in photon counts/s), with a 45-ms resolution We find an ‘‘off’’ state intensity of about 1000 counts/s, which corresponds to the electronic and optical background The ‘‘on’’ states typically range between and 14 kcps, with a large distribution of intensity values Fig 7(b) shows the different intensity levels corresponding to different decay rates We select the photons detected when the emission intensity was between 0.6 and 1.5 kcps (blue), 2.5 and 4.5 kcps (green), and above 6.5 kcps (red) and we plot the three corresponding decay curves in Fig 7(b) The decay curve is almost a single exponential (with 18 ns decay time) for the highest emission intensities, and faster and non-exponential (1/e decay in 1.5 ns) for the lowest intensities This indicates that the intensity fluctuations are caused by fluctuating non-radiative decay channels, as was observed previously for CdSe/ZnS [47] and CdSe/CdS [48] QDs: the faster decay and lower intensities are correlated because they correspond to higher non-radiative rates Finally we plot in Fig 7(c) the histogram of the values of the delay s between one photon detected on photodiode and one photon detected on photodiode This curve constitutes the intensity autocorrelation function g(2)(s) and characterizes the light emission quantum properties We used here a pulsed laser with Fig (a) PL dynamics of a single QD (sample A1) (b) Decay curves (in ln scale) measured when the QD intensity was between 0.6 and 1.5 kcps (blue), 2.5 and 4.5 kcps (green) and above 6.5 kcps (red) (c) Intensity autocorrelation function g(2)(s), plotted as the histogram of the time delays between one photon on photodiode and one photon on photodiode (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) a 400-ns period between the pulses, so that when a photon is detected it is more likely that another photon is detected after a delay which is a multiple of 400 ns This appears on the g(2)(s) curve as peaks for values of s multiples of 400 ns However, at s = 0, there is no peak (or very weak peak due to background fluorescence from the substrate) This corresponds to the single-photon emission property (antibunching): only one photon can be emitted after one laser pulse and never are two photons detected at the same time This property has been demonstrated for CdSe/ ZnS QDs [7,49] and recently for CdZnSe/ZnSe QDs emitting at 600 nm [31] Conclusion In summary, we have successfully synthesized high-quality CdZnSe ternary core and ternary/ternary CdZnSe/ZnSeS core/shell QDs using embryonic nuclei-induced alloying process Depending on the Cd/Zn ratios and the synthesis temperature, CdZnSe core QDs could strongly luminescence in the spectral range of 530– 607 nm This corresponds to a decrease of the Zn concentration in the core QDs as the reaction temperature increased By shelling Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 H.Y Nguyen et al / Optical Materials xxx (2014) xxx–xxx ZnSeS on the CdZnSe core QDs, the average LQY is increased by a typical factor of up to 17 and up to LQY values of 25%, showing the potential for high-quality ternary/ternary QD synthesis We attribute this increase in LQY to the reductions of both the average ‘‘off’’ time and the percentage of ‘‘dead’’ non-emitting QDs By using single-photon emitter spectroscopy, we have demonstrated for the CdZnSe/ZnSeS core/shell QDs a complete photon antibunching We confirmed the reduction of the average ‘‘off’’ time by shelling the core QDs and showed that blinking property was improved comparing to that of the CdZnSe core QDs We are currently optimizing the synthesis parameters in order to take advantage of the core and shell concentrations to tune the emission further to the blue region and to reduce blinking and non-radiative channels Acknowledgments This work was funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED, Project 103.062011.03), the PICS cooperation project between CNRS and VAST (Project Number 5724) and by the Centre de Compétences C’Nano – Ile de France (NanoCrisPho and NanoPlasmAA projects) and the Agence Nationale de la Recherche (Delight project) The authors thank the National Key Laboratory for Electronic Materials and Devices – IMS for the use of its facilities and Pr Le Van Vu for his advice on XRD analysis Appendix A Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.optmat.2014 04.020 References [1] Igor L Medintz, H Tetsuo Uyeda, Ellen R Goldman, Hedi Mattoussi, Nat Mater (2005) 435–446 [2] Ou Chen, Jing Zhao, Vikash P Chauhan, Jian Cui, Cliff Wong, Daniel K Harris, He Wei, Hee-Sun Han, Dai Fukumura, Rakesh K Jain, Moungi G Bawendi, Nat Mater 12 (2013) 445–451 [3] J Lim, S Jun, E Jang, H Baik, H Kim, J Cho, Materials 19 (2007) 1927–1932 [4] Yu Tao, Jiang-Shan Shen, Hai-Hong Bai, Lei Guo, Ji-Jun Tang, Yun-Bao Jiang, Jian-Wei Xie, Analyst 134 (2009) 2153–2157 [5] K Zhang, Q Mei, G Guan, B Liu, S Wang, Z Zhang, Anal Chem 82 (2010) 9579–9586 [6] Thi Kim Chi Tran, Duc Chinh Vu, Thi Dieu Thuy Ung, Hai Yen Nguyen, Ngoc Hai Nguyen, Tran Cao Dao, Thu Nga Pham, Quang Liem Nguyen, Adv Nat Sci.: Nanosci Nanotechnol (2012) 035008 [7] P Michler, A Kiraz, C Becher, W.V Schoenfeld, P.M Petroff, Lidong Zhang, E Hu, A Imamoglu, Science 290 (2000) 2282–2285 [8] G Messin, J.P Hermier, E Giacobino, P Desbiolles, M Dahan, Opt Lett 26 (2001) 1891–1893 [9] B Lounis, H.A Bechtel, D Gerion, P Alivisatos, W.E Moerner, Chem Phys Lett 329 (2000) 399–404 [10] I Mekis, D.V Talapin, A Kornowski, M Haase, H Weller, J Phys Chem B 107 (2003) 7454–7462 [11] Nguyen Quang Liem, Le Quang Phuong, Ung Thi Dieu Thuy, Tran Thi Kim Chi, Do Xuan Thanh, J Korean Phys Soc 53 (2008) 1570–1574 [12] Ung Thi Dieu Thuy, Nguyen Quang Liem, Do Xuan Nguyen, M Protière, P Reiss, Appl Phys Lett 91 (2007) 241908 [13] M.A Hines, P Guyot-Sionnest, J Phys Chem 100 (1996) 468–471 [14] B.O Dabbousi, J Rodriguez-Viejo, F.V Mikulec, J.R Heine, H Mattoussi, R Ober, K.F Jensen, M.G Bawendi, J Phys Chem B 101 (1997) 9463–9475 [15] J.J Li, Y.A Wang, W Guo, J.C Keay, T.D Mishima, M.B Johnson, X Peng, J Am Chem Soc 125 (2003) 12567–12575 [16] X Brokmann, L Coolen, M Dahan, J.P Hermier, Phys Rev Lett 93 (2004) 107403 [17] M Nirmal, B.O Dabbousi, M.G Bawendi, J.J Macklin, J.K Trautman, T.D Harris, L.E Brus, Nature 383 (1996) 802–804 [18] Todd D Krauss, Jeffrey J Peterson, Nat Mater 11 (2012) 14–16 [19] B Mahler, P Spinicelli, S Buil, X Quelin, J.-P Hermier, B Dubertret, Nat Mater (2008) 659–664 [20] Y Chen, J Vela, H Htoon, J.L Casson, D.J Werder, D.A Bussian, V.I Klimov, J.A Hollingsworth, J Am Chem Soc 130 (2008) 5026 [21] Jong-Uk Kim, Jong-Jin Lee, Ho Seong Jang, Duk Young Jeon, Heesun Yang, J Nanosci Nanotechnol (2011) 725–729 [22] Michelle D Regulacio, Ming Yonghan, Acc Chem Res 43 (5) (2010) 621–630 [23] X.H Zhong, Y.Y Feng, W Knoll, M.Y Han, J Am Chem Soc 125 (2003) 13559– 13563 [24] R.E Bailey, S.M Nie, J Am Chem Soc 125 (2003) 7100 [25] X.H Zhong, Z.H Zhang, S.H Liu, M.Y Han, W Knoll, J Phys Chem B 108 (2004) 15552–15559 [26] X.H Zhong, Y.Y Feng, Y.L Zhang, Z.Y Gu, L Zou, Nanotechnology 18 (2007) 385606–385611 [27] Y.G Zheng, Z.C Yang, J.Y Ying, Adv Mater 19 (2007) 1475–1479 [28] J.Y Ouyang, C.I Ratcliffe, D Kingston, B Wilkinson, J Kuijper, X.H Wu, J.A Ripmeester, K Yu, Gradiently J Phys Chem C 112 (2008) 4908–4919 [29] M Protiere, P Reiss, Small (2007) 399–403 [30] F.C Liu, T.L Cheng, C.C Shen, W.L Tseng, M.Y Chiang, Langmuir 24 (2008) 2162–2167 [31] X Wang, X Ren, K Kahen, M.A Hahn, M Rajeswaran, S.M Zacher, J Silcox, G.E Cragg, A.L Efros, T.D Krauss, Nature 459 (2009) 686–689 [32] W.K Bae, M.K Nam, K Char, S Lee, Chem Mater 20 (2008) 5307 [33] J.M Caruge, J.E Halpert, V Wood, V Bulovic, M.G Bawendi, Nat Photonics (2008) 247–250 [34] Huiguang Zhu, Arjun Prakash, Denise N Benoit, Christopher J Jones, Vicki L Colvin, Nanotechnology 21 (2010) 255604 [35] Adam J Morris-Cohen, Martin D Donakowski, Kathryn E Knowles, Emily A Weiss, J Phys Chem C 114 (2010) 897–906 [36] P.T Nga, V.D Chinh, V.T.H Hanh, N.X Nghia, P.T Dzung, C Barthou, P Benalloul, J Laverdant, A Mtre, Int J Nanotechnol (2011) 347–359 [37] Tran Thi Quynh Hoa, Le Thi Thanh Binh, Le Van Vu, Nguyen Ngoc Long, Vu Thi Hong Hanh, Vu Duc Chinh, Pham Thu Nga, Opt Mater 35 (2012) 136–140 [38] J Laverdant, W Daney de Marcillac, C Barthou, C Schwob, L Coolen, P Benalloul, P.T Nga, A Mtre, Materials (2011) 1182 [39] P Kukura, M Celebrano, A Renn, V Sandoghdar, Nano Lett (2009) 926–929 [40] Ung Thi Dieu Thuy, Peter Reiss, Nguyen Quang Liem, Appl Phys Lett 97 (2010) 193104 [41] Steven A Santangelo, Eric A Hinds, Vladimir A Vlaskin, Paul I Archer, Daniel R Gamelin, J Am Chem Soc 129 (2007) 3973–3978 [42] C.A Smith, H.W.H Lee, V.J Leppert, S.H Risbud, Appl Phys Lett 75 (1999) 1688 [43] R.C Weast, M.J Astle, CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton, Florida, 1982 [44] M Kuno, D.P Fromm, H.F Hamann, A Gallagher, D.J Nesbitt, J Chem Phys 112 (2000) 3117 [45] M Kuno, D.P Fromm, S.T Johnson, A Gallagher, D.J Nesbitt, Phys Rev B 67 (2003) 125304 [46] P.A Frantsuzov, R.A Marcus, Phys Rev B 72 (2005) 155321 [47] B.R Fisher, PhD thesis at the Massachusetts Institute of Technology, Massachusetts, 2005 [48] C Galland, Y Ghosh, A Steinbrück, M Sykora, J.A Hollingsworth, V.I Klimov, H Htoon, Nature 479 (2011) 203 [49] P.T Nga, V.D Chinh, P.T Linh, K.C Cuong, V.T.H Hanh, N.X Nghia, C Barthou, A Mtre, Adv Nat Sci 10 (2009) 167–174 Please cite this article in press as: H.Y Nguyen et al., Opt Mater (2014), http://dx.doi.org/10.1016/j.optmat.2014.04.020 ... high-quality CdZnSe ternary core and ternary/ ternary CdZnSe/ ZnSeS core/ shell QDs using embryonic nuclei-induced alloying process Depending on the Cd/Zn ratios and the synthesis temperature, CdZnSe core. .. wavelengths and shelling with ZnSeS by different Se/S ratios Results for the ternary/ ternary CdZnSe/ ZnSeS core/ shell QDs 4.1 Structural characterization Samples A and B were covered by a ZnSeS shell. .. core, indicating that the total (radiative + non-radiative) decay rate is not modified, and suggesting that both the radiative and non-radiative rates are not modified The LQY being the ratio of