Photoluminescence Quantum Yield and Matrix-Induced

Apr 28, 2014 - Alexander N. Kulak , Rebecca Grimes , Yi-Yeoun Kim , Mona Semsarilar , Clara Anduix-Canto , Oscar Cespedes , Steven P. Armes , and Fion...
1 downloads 8 Views 2MB Size
Article pubs.acs.org/cm

Photoluminescence Quantum Yield and Matrix-Induced Luminescence Enhancement of Colloidal Quantum Dots Embedded in Ionic Crystals Marcus Müller,†,§ Martin Kaiser,‡,§ Gordon M. Stachowski,† Ute Resch-Genger,*,‡ Nikolai Gaponik,*,† and Alexander Eychmüller† †

Physical Chemistry, TU Dresden, Bergstr. 66b, 01062 Dresden, Germany BAM Federal Institute for Materials Research and Testing, Division Biophotonics, Richard-Willstaetter Str. 11, D-12489 Berlin, Germany



S Supporting Information *

ABSTRACT: The incorporation of colloidal quantum dots (QDs) into solid matrices, especially ionic salts, holds several advantages for industrial applications. Here, we demonstrated via absolute measurements of photoluminescence quantum yields (PL-QY) that the photoluminescence of aqueous CdTe QDs can be considerably increased upon incorporation into a salt matrix with a simple crystallization procedure. Enhancement factors of up to 2.8 and a PL-QY of 50 to 80%, both in NaCl crystals and incorporated in silicone matrices, were reached. The fact that the achievable PL enhancement factors depend strongly on PL-QY of the parent QDs can be described by the change of the dielectric surrounding and the passivation of the QD surface, modifying radiative and nonradiative rate constants. Time-resolved PL measurements revealed noncorrelating PL lifetimes and PL-QY, suggesting that weakly emissive QDs of the ensemble are more affected by the enhancement mechanism, thereby influencing PL-QY and PL lifetime in a different manner.



INTRODUCTION Nanostructures consisting of solution-synthesized colloidal quantum dots (QDs) distributed inside of polymer or inorganic matrices attract considerable attention due to the awaited increase of QD stability in the matrices and ease of processability in conjunction with comparatively low production costs.1−4 Together with a high brightness, the issue of long-term stability of photoluminescent solid composites is of special interest for their applications as optical gain media, color conversion materials, and devices for lighting, displays, and photovoltaics.5−8 Although QDs embedded in organic polymers possess a good processability and compatibility with typical packaging approaches, most of the commonly used polymer matrices like, e.g., polystyrene (PS) and poly(methyl methacrylate) (PMMA) show only limited stability under harsh conditions, i.e., ultraviolet light excitation.9 Additionally, these polymers possess relatively high oxygen diffusion coefficients (2.3 × 10−7 and 3.3 × 10−9 cm2/s respectively),10,11 which is counterproductive for the protection of QDs from photooxidation processes. Contrarily, most inorganic materials, although potentially less processable, are typically extremely robust and airtight matrices for embedded substances. By combining these beneficial properties with the tunable emission12 and high extinction coefficients13,14 of colloidal QDs, a new class of superior photonic materials can be created. Requirements for the design of improved solid QD composites include control of interparticle spacing, prevention of QD aggregation, and saturation of QD surface defects. Additionally, © 2014 American Chemical Society

the avoidance of luminescence quenching via high energy vibrational modes of the inorganic matrix is needed for optimum brightness. For some applications, matching of the refractive indices of the solid QD composites and other additional packaging materials can be beneficial to minimize scattering, as is the case for, e.g., the design of luminescent solar concentrators.15 Aiming at a simple preparation of solid QD composites with tunable and bright emission, we recently6 developed and characterized a novel type of encapsulation strategy for colloidal QDs: their embedding in ionic crystals of common salts like NaCl, KCl, KBr, etc. In addition to their strong photoluminescence and the preservation of their narrow emission bands, these new solid QD-salt composites showed a superior photochemical stability of the embedded QDs as compared to their parent QDs in solution and to QDs incorporated in various other matrices. Furthermore, we were able to demonstrate the considerable application potential of these mixed QD-salt crystals by combining them with a commercial blue emitting InGaN LED to fabricate a white emitting LED, potentially suitable for solid state lighting. Rogach et al. extended this approach only very recently to the fabrication of pure color LEDs based on CdTe QDs embedded in NaCl.16 Both ourselves6 and the authors of ref 16 observed and Received: March 13, 2014 Revised: April 25, 2014 Published: April 28, 2014 3231

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

separated from the parent solution, rinsed with cold water, and dried. They exhibited a variety of shapes and facets and were milled to a fine powder. Subsequently, the PL properties of the pure powders and the powders incorporated into silicone, a common matrix for the hybridization of color conversion layers with LEDs, as shown in Figure 1, were studied in comparison to those of the parent QD solutions.

reported an improvement of the photoluminescence quantum yields (PL-QY) of the QDs embedded in salts as compared to parent colloids. However, a systematic study allowing an interpretation of the observed changes in the luminescence behavior of QD colloids upon incorporation into the salt crystal was not performed. In order to understand the origin of bright emission of QDsalt composites and aiming at an efficient matrix control of their PL-QY, we studied the PL behavior of these fascinating new materials in more detail. Therefore, in the present work, we focused on differently sized parent CdTe QDs with two different thiol capping ligands, thioglycolic acid (TGA) and mercaptopropionic acid (MPA), as well as on CdTe QDs with low and high PL-QY in aqueous solution. The resulting QDsalt composites were characterized by absolute measurements of their PL-QY and studies of their PL decay kinetics. Additionally, for a better understanding of the influence of the salt matrix on the PL properties of the embedded QDs, also the PL behavior of an exemplarily chosen core−shell QD, in this case CdSe/ZnS stabilized with oleic acid (OA) and MPA, was investigated. These investigations provide the basis for the fabrication of inexpensive and efficient optical and electrooptical devices based on QD-salt composites.



Figure 1. Photograph of an exemplarily chosen QD-salt composite after milling (a) and another exemplarily chosen QD-salt composite sample after milling and incorporation into silicone (b, c). Photographs were taken under room light (b) and UV-light (a, c).

EXPERIMENTAL SECTION

Chemicals and Apparatus. All chemicals used were of analytical grade or of the highest purity available. All aqueous solutions were prepared from Milli-Q water (Millipore). The Al2Te3 lumps employed for the generation of H2Te were purchased from CERAC Inc. The synthesis of the mixed crystals was carried out according to our previous publication.6 We used either CdTe17 or CdSe/ZnS18 nanocrystals as parent QDs. Briefly, CdTe nanocrystals were synthesized by dissolving 2.305 g (5.5 mmol) of Cd(ClO4)2·6H2O in 250 mL of water and adding 7.15 mmol of the thiol stabilizing agent dropwise under stirring, followed by an adjustment to pH 12, using a 1 M solution of NaOH. The solution was placed in a three necked flask and deaerated with Ar for 30 min. H2Te gas was generated by reacting 0.4 g (0.916 mmol) of Al2Te3 lumps with 10 mL of a 0.5 M H2SO4 solution and passed through the solution with a slow argon flow under vigorous stirring. The growth of the QDs to the desired size proceeded under open air conditions upon refluxing at 100 °C with a condenser attached. Nanocrystals obtained by this well established approach possess PL-QY in the range of 10− 60%.19 The sizes of the CdTe QDs were derived from the 1s−1s transition maximum, following Rogach et al.20 CdSe/ZnS nanocrystals with an alloyed gradient shell were prepared by mixing 51.4 mg (0.4 mmol) of CdO, 733.6 mg (4 mmol) of Zn(OAc)2, 5.5 mL of OA, and 20 mL of octadecene (ODE) in a 50 mL three necked flask, using a slightly modified approach to that in ref 18. The suspension was degassed at 100 °C for 60 min. The system was filled with Ar and heated to 310 °C. Then, a mixture of 7.9 mg (0.1 mmol) of Se and 128.3 mg (4 mmol) of S dissolved in 3 mL of trioctylphosphine (TOP) was rapidly injected into the flask, followed by a temperature reduction to 300 °C. After 10 min, the reaction mixture was cooled to room temperature, and the resulting nanocrystals were purified two times by washing with 20 mL of CHCl3 and an excess of acetone. Following, the QDs were redispersed in 4 mL CHCl3 and phase transferred using an aqueous solution of MPA.21 Precisely, 30 μL of the nanocrystals were diluted with 500 μL of CHCl3 and stirred with 1 mL of 0.2 M MPA at pH 11 for 2 h. The phases were separated, and the aqueous phase was used without further purification. QD-salt crystals were prepared by melding 5 mL of a QD solution with 25 mL of a saturated NaCl solution. The solutions were stored within an oven at 30 °C to slightly increase the crystallization rate and protect them from dust. The crystallization was finished when the parent solution showed no further emission. The crystals were

Absolute Measurement of Photoluminescence Quantum Yield. The photoluminescence quantum yield of the QDs in solution, the QD-salt crystals, and the QD-salt crystal silicone composites were determined absolutely with a custom-designed integrating sphere setup.22 Sample excitation was at 480 nm using a xenon lamp coupled into a single monochromator. Sample and blank were center mounted in a 6 in. Spectraflect-coated integrating sphere (Labsphere GmbH). The emitted and scattered light was collected with a quartz fiber coupled to the integrating sphere, attached to an imaging spectrograph (Shamrock303i, Andor Inc.) and detected with a Peltier-cooled thinned back side illuminated deep depletion charge coupled device (CCD array). The spectral responsivity of the detection channel was determined with a calibrated tungsten lamp (integrating sphere-type spectral radiance transfer standard from Gigahertz of known spectral radiance), thus providing traceability to the spectral radiance scale.23−25 QDs in solution were investigated in 10 × 4 mm quartz cuvettes in order to minimize reabsorption. Powders of milled QD-salt crystals were distributed uniformly in a 200 μm thick round cuvette (diameter 15 mm) or embedded into a silicon pellet (see Figure 1). The powder samples were tilted with respect to the entrance port of the sphere in order to collect most of the scattered excitation light to ensure minimal measurement uncertainties. We used either a solvent-filled cuvette, a round cuvette filled with the pure salt used as a QD matrix or a silicone pellet doped with QD-free pure salt as blanks for the determination of the number of absorbed photons from the transmitted and reflected incident spectral radiant flux and the blank correction of the emission spectra.23,26 More details regarding the setup, calibration, and measurement protocols can be found in ref 25. Determining the Optimum Concentration of QD-Salt within the Silicone. To optimize the concentration of the QD-salt crystals embedded in silicone for the PL studies, three QD-salt silicone composites from an exemplarily chosen QD-salt batch were prepared, varying the amount of the QD-salt powder to determine a concentration range suitable for accurate PL-QY measurements with the minimum contribution from reabsorption. The concentrations were chosen to be 5, 10, and 20 mg of powder per 500 μL of silicone as follows from Table 1. Spectroscopic measurements of these samples revealed matching absorption and emission spectra, an increase in 3232

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

influence of the QD−NaCl interface, we additionally assessed salt composites of CdSe/ZnS QDs possessing Zn atoms on their surface as well as CdTe QDs incorporated into a halogenfree borax (NaB4O7) matrix. An overview of all samples discussed in the following sections can be found in the Supporting Information (SI), including the size of the QDs, the stabilizing agent, representative emission spectra, as well as PL-QY data and decay behavior in solution, in the salt matrices and the corresponding enhancement factors. Influence of PL-QY of the Parent QD on the PL Properties of QD-Salt Composites. To gain a better understanding of the influence of the PL-QY of the parent QDs on the PL-QY of the resulting QD-salt composites, the PL properties of salt crystals made from CdTe QDs with high (ca. 50% in solution) and low (ca. 10% in solution) PL-QY were compared. The PL-QYs of four different CdTe QDs in solution and in a NaCl matrix are shown in Figure 2a. The two samples originating from QDs with low PL-QY in solution showed a very promising PL enhancement of more than 2.5 times, thereby considerably exceeding the PL increase reported in ref 16 for CdTe QDs with comparable PL-QY. The composites derived from already in solution stronger emitting parent QDs revealed only an increase in PL by a factor up to 1.3 after incorporation into NaCl crystals. Nevertheless, on average, the PL-QYs of our CdTe-salt crystal composites were on the order over 30% (Figure 2a) and up to 80% (Table SI1). This underlines the considerable potential of our simple strategy to obtain highly emissive solids of QDs. PL-LT measurements performed with our samples reveal multiexponential decay kinetics for both the CdTe solution and the CdTe-salt composites, as to be expected for such rather heterogeneous systems, see Figure 2b. The corresponding lifetimes are summarized in Table SI1 together with the PL-QY data. As shown in Figure 2b, despite the considerable PL enhancement, the PL-LT and the decay kinetics of the QDs incorporated into the NaCl salt crystals change only slightly compared to the parent QD solutions. Generally, changes on the surface of QDs, in the QD shell, or at the core−shell interface can affect both radiative and nonradiative recombination rates of charge carriers. For a better understanding of the observed PL-QY enhancement, two effects need to be considered: the interplay between changes in PL-QY and PL-LT and the influence of the refractive index of the matrix on both quantities. The influence of the refractive index on PL-QY and PL-LT follows from the Strickler−Berg

Table 1. PL-QY of a Set of Silicone Pellets Containing Different Amounts of the Same QD-Salt Batch concentration

PL-QY

Abs

2.5 mg/mL 5 mg/mL 10 mg/mL

0.28 0.30 0.23

8.8% 27% 49%

absorption with increasing sample concentration, and PL-QY values of 28%, 30%, and 23%, respectively. The decrease in quantum yield found for the highest QD-salt concentration was attributed to reabsorption. Therefore, we used only 10 mg of powder for the subsequently prepared samples. These first measurements also showed the considerable challenge of the PL-QY determination of these strongly scattering, unsymmetrically faceted materials with acceptable uncertainties. PL Lifetime (PL-LT) Measurements. Luminescence decay curves were determined with an Edinburgh Instrument lifetime spectrometer (FLS 920) equipped with a supercontinuum laser (SC400-PP, Fianium) as a pulsed excitation light source (pulse frequencies of 0.5 or 1 MHz and a pulse width of 150 ps), a double monochromator, a MCP-PMT (R3809U-50, Hamamatsu), and a time-correlated singlephoton-counting (TCSPC) module (TCC 900) for signal detection. The instrument response function was measured with a nonluminescent scatterer at the excitation wavelength. The luminescence decays of the QDs in solution, in the crystal matrices, and in the silicon pellets were measured with identical instrument settings to obtain comparable values, e.g., excitation at 480 nm, detection at the emission maximum, identical detection/excitation bandpasses, and repetition rates (0.5 or 1 MHz). For data evaluation, the PL-LTs of these multiexponential decays were set equal to the time when the intensity corresponds to 1/e of the initial intensity. In addition, we determined the areas under the decay curves for comparison purposes.27 Both methods of data evaluation yielded comparative trends, i.e., relative changes in decay behavior (deviation of maximum 10%).



RESULTS AND DISCUSSIONS To gain a better understanding of the PL behavior of our QDsalt crystal composites, confirm the occurrence of matrixinduced PL enhancement, and derive prerequisites for a PL enhancement,6,16 the PL-QY and PL-LT of varying CdTe QD systems (parent QD, QD-salt crystal, and QD-salt crystal in silicone) were studied. To ensure representative results, we systematically studied various CdTe QDs sample series including QDs of different sizes, QDs stabilized with either TGA or MPA, as well as QDs possessing relatively low and relatively high PL-QY in a parent solution. To clarify the

Figure 2. (a) PL-QY of four exemplarily chosen CdTe QD samples with relatively low (low1, 2) and high (high 1, 2) PL QYs of parent solutions. The black squares show the PL-QY of the QDs in solution and the red ones the PL-QY of the QDs in the NaCl crystals. The numbers within the panel provide the salt-induced PL-QY enhancement factor of the QDs as compared to their solution values. (b) PL-LT decay curve of the CdTe sample “low 1” in solution (black squares, amplitude-averaged lifetime of 16.9 ns) and within the salt crystals (red squares, amplitude-averaged lifetime of 14.6 ns). A nonexponential version of the first 200 ns of part b can be found in Figure SI2. 3233

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

equation28 at least for molecular systems, which describes the correlation between the radiative rate constant and the square root of the refractive index of the surrounding material. Considering the refractive indices (n) of water and NaCl to be respectively 1.3334 and 1.544 (at 20 °C and 589 nm),29 one can expect an increase in the radiative rate constant of 34% according to the Strickler−Berg equation when comparing QDs in salt vs QDs in water. Furthermore, recently described cavity models can be used to take the influence of the local refractive index on the radiative transition probability into account using an effective medium approximation.30 As an estimate for such effects, we calculated the influence of the refractive index on the radiative rate constant and, hence, PL-QY (Φ = kr/(kr + knr)). Assuming an isotropic medium, these calculations were done using the simplified Strickler−Berg (SB) formula,31 the virtual cavity (VC) model, the empty cavity (EC) model, and the fully microscopic (FM) model, see eq 1, as reported by Pillonnet et al.30 In eq 1, kr represents the radiative rate constant, n̅ is the average refractive index, and krv is the radiative rate constant for n̅ = 1. We approximated n̅ with the refractive index of the solvent which is, according to Pillonnet et al., suitable for high index volume fractions below 1%.30,31 k r(n)̅ = n ̅ 2k rv Strickler − Berg equation

⎛ n2 + k r(n ̅ ) = ⎜ ̅ ⎝ 3

2⎞ ⎟ nrv̅ virtual cavity model ⎠

Figure 3. Enhancement of PL-QY values arising from the incorporation of thiol-stabilized CdTe into a NaCl matrix for differently sized CdTe (radii of 2−3 nm) QDs. The factors for the PL enhancement originating from a change of the radiative rate constant were calculated with the Strickler−Berg equation31 (eq 1) and for the virtual cavity, empty cavity, and finite microscopic model30 given in eqs 2−4. For better orientation, enhancement factors allowing a gain of QYs =1 are shown as a solid line.

(1)

behavior of the strongest luminescent QDs, with no or very little contribution from weakly emitting (or dark) QDs. Although PL-QY and PL-LT are straight correlated for molecular emitters showing commonly monoexponential decay kinetics in a homogeneous environment,33 there has not yet been a straightforward and stringent correlation reported for PL-QY and the decay behavior of QD ensembles. For such systems, only similar trends of both parameters can be usually observed, with changes in PL-QY often exceeding the corresponding changes in the decay behavior.34,35 A possible explanation was given by Bawendi et al., showing with single particle measurements that fluctuations of the decay rates can be caused by changes in emission intensities of single QDs.36 Also Rogach et al., who similarly reported multiexponential decay kinetics for their materials, did not find a clear correlation between both quantities for their QD-salt systems, especially when considering the area under the decay curves calculated from the data reported for their fits of the decay behavior of their CdTe-salt composites.16 One explanation for the stronger changes in PL-QY compared to those in the decay behavior, even when considering the respective changes in radiative rate constants previously calculated (see eqs 1-4), could be that an improved surface passivation of the QDs in the salt crystal enhances especially the PL-QY of weakly emissive nanocrystals, causing their PL-LT to reach similar values as observed for already stronger emissive QDs. In summary, the observed PL changes seem to arise from both changes in refractive index and changes in the radiative and nonradiative rate constants. Influence of Surface-CdClx on PL-QY. The increase in PL-QY upon incorporation of CdTe into NaCl crystals also agrees with the findings of Sargent and co-workers for the influence of halide ions on the emission intensity of PbS QDs.37,38 This enhancement was attributed to the formation of a thin PbClx or CdClx layer, as depicted in Figure 4, which seems to passivate dangling bonds at the QD surface. As a result, the nonradiative relaxation routes are reduced and the PL-QY increases. Comparing the measured PL-QY enhancement factors for the samples shown in Figure 3 with their corresponding change in PL emission maxima, a clear correlation was found. As shown in Table SI2, all samples

2

(2)

⎛ 3 n 2 ⎞2 k r(n ̅ ) = ⎜ 2 ̅ ⎟ nrv̅ empty cavity model ⎝ 2n ̅ + 1 ⎠

(3)

n2 + 2 k r(n ̅ ) = ̅ k rv finite microscopic model 3

(4)

For a complete understanding of the observed PL enhancement, it would be crucial to know which cavity model provides the best description. To the best of our knowledge, there has not yet been undoubted proof for one of the models. Therefore, the calculations were made for all four models, keeping critically in mind the assumption of a nonaffected nonradiative rate constant. As follows from Figure 3, the enhancement of PL-QY can be explained with the cavity models for about half of our samples. For these samples, a refractive index-induced change in kr (and not in knr) can be responsible for the enhancement. For a correlation of PL-QY and PL-LT, it must be considered additionally that the presented data arise from ensemble measurements of QDs and from QDs which reveal multiexponential decay kinetics. These ensemble data reflect most likely a broad distribution in PL-QY, the distribution of trap states at the QD surface or core−shell interface, and the heterogeneous QD environment. Moreover, energy transfer from smaller to larger QDs can also occur, as suggested by the red shift in emission observed for nearly all nanocrystal-salt systems. Our PL-QY ensemble data equal the number of emitted per absorbed photons of the QD ensemble studied, with the sample-specific fraction of dark QDs contributing only to the absorption measurement. A correlation between the PLQY and the ratio of bright to dark QDs was, e.g., shown by Ebenstein et al.32 for CdSe/ZnS nanocrystals by single particle measurements of QD ensembles of varying PL-QY. PL-LT data of QD ensembles are commonly dominated by the decay 3234

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

Influence of QD Size on the Luminescence Behavior of CdTe-NaCl Composites. We subsequently studied the influence of the parent QD size (and QD size-related PL-QY) on the luminescence properties of the resulting CdTe-salt composites. In the left part of Figure 6, the evolution of PL-QY Figure 4. Formation of a thin CdClx-layer on the surface of a CdTeNP after mixing of the QD solution with the saturated NaCl solution.

exceeding the model-predicted PL-QY enhancement factors show a slightly larger red-shift in their emission maximum than the samples which are within the predicted range of the previously discussed models. These findings indicate that the formation of a CdClx-shell might be a reason for the PL-QY enhancement reaching beyond the model predictions. Indeed, proper shell formation usually causes red shifts in the emissions of QDs, which was also observed by Sargent et al. for QDs with CdClx shells.38 To test our hypothesis of surface defect curing related to the formation of a chloride layer and the binding of chloride to surface Cd atoms, we performed comparative studies of CdSe/ ZnS QDs embedded in NaCl crystals and CdTe QDs in NaB4O7 (sodium tetraborate or borax), respectively. In both cases, CdClx cannot be formed on the QD surface. As follows from Figure 5a, CdSe/ZnS QDs show a typical drop in PL-QY

Figure 6. PL-QY of thiol-stabilized CdTe QDs before (black) and after (red or blue) incorporation into the salt matrix with the respective enhancement factors being given as numbers. An increase in particle size leads to a higher PL-QY in solution and a smaller PL-QY enhancement factor for both thiol ligands. The two TGA-stabilized samples with a radius of 2.7 nm differ slightly in size and, thus, also slightly in the position of their absorption and emission maxima.

for a series of differently sized MPA-stabilized CdTe QDs is shown, revealing an increase in PL-QY with increasing particle size (from 2.1 to 3.0 nm) in solution. This is due to the improved crystallinity and, hence, diminished number of surface defects evoking through a longer growing time as well as a reduced surface-to-volume ratio.20 After incorporation of the MPA-stabilized QDs into the NaCl matrix, the PL-QY increased for all sizes, yet in a particle size-specific manner, thus yielding size-dependent enhancement factors ranging from 1.1 to 2.8. This is consistent with the behavior of the samples shown in Figure 2. A similar trend can be observed for the TGA-stabilized samples shown in the right part of Figure 6. For this ligand, the differences in the enhancement factors of the differently sized QDs (2.3−2.7 nm) are less pronounced, ranging from 1.1 to 1.6. Nevertheless, it can be concluded that the trend of size specific PL-QY enhancement is not dependent on the QDs stabilizing agent. Figure 6 also underlines the stronger enhancement of the PL-QY found by us as compared to the effects reported by the authors of ref 16 (see also Table SI1 and ref 16). These favorable effects can be attributed to the different synthetic methods (i.e., in the ratios of the precursors) applied for producing the CdTe QDs by Rogach40 and us.17 Furthermore, the observed variations might be also ascribed to different amounts of QDs incorporated into the salt matrix, yielding different ratios of reabsorption. Influence of the MatrixIncorporation into Silicone. As detailed in the Introduction, many applications of solid QDs like the fabrication of LEDs require incorporation of the QDsalt composites in an additional matrix like silicone. Hence, mixed crystals from the same or comparable batches were studied with and without silicone encapsulation. The overall goal was to ensure that the incorporation into silicone does not affect the beneficial optical properties of our solid QDs. The

Figure 5. (a) Relative PL-QY of CdSe/ZnS and the respective QD-salt mixtures. OA-capped CdSe/ZnS QDs were phase transferred from CHCl3 to H2O by ligand exchange using MPA as a new stabilizing ligand, resulting in a PL-QY drop, whereas subsequent incorporation into NaCl causes only a negligible change in PL-QY. (b) CdTe QDs from one batch were incorporated into either NaB4O7 or NaCl crystals, yielding almost no change for the former, but a strong increase by a factor of 1.5 in the case of the NaCl host.

during ligand-exchange mediated phase-transfer from organic media (OA-capped) to H2O (MPA-capped)21,39 which remains nearly unaffected by incorporation into the NaCl host. In the case of borax-encapsulated CdTe, the PL-QY of the QDs increases only slightly upon incorporation in the salt matrix (see Figure 5b and Table SI1). Changes of PL-QY, which fall below the expected refractive index-induced changes, point to an increase of nonradiative processes. This suggests that the formation of CdClx at the QD surface, which can occur for CdTe in NaCl, possibly contributes to the salt-crystal induced PL enhancement. Also PL-QY values with other QD-NaCl systems shown in the SI underline our assumption of saltinduced healing of Cd atom-related surface defects. 3235

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

Author Contributions

matching PL-QY of a CdTe-NaCl mixed crystal batch, measured in silicone and without silicone encapsulation (size of ∼3 nm) and the very similar PL decay behavior (see Table SI1) underline the negligible influence of such an additional polymer matrix. Obviously, using micrometer-sized QD-salt composites, the immediate QD environment is not affected by silicone.

§

These authors equally contributed to this work

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We acknowledge the expert assistance of Christin Rengers, André Wolf, and Dr. Markus Grabolle. The work was partly financially supported by the German Research Foundation (DFG) projects EY16/14-1 and RE1203/12-1 as well as the EU FP7 Network of Excellence “Nanophotonics for Energy Efficiency.” M.K. gratefully acknowledges financial support from the Federal Ministry of Economics and Technology (BMWI-14/09).



CONCLUSION We demonstratedvia absolute quantum yield measurementsthat the PL-QY of CdTe QDs can be increased upon incorporation of QDs into a salt matrix using a simple crystallization procedure. The resulting PL enhancement factors strongly depend on the change of the refractive index of the surrounding matrix and on the PL-QY of the parent QD, with the incorporation of QDs with small PL-QY and, thus, a higher number of surface defects, favoring nonradiative decay processes, yielding higher enhancement factors. In summary, enhancement factors of up to 3.0 and PL-QY of 30 to 80% could be achieved for both, QDs in NaCl crystals and QD-salt composites incorporated in a silicone matrix. Time-resolved PL measurements revealed that the PL decay behavior of the QDsalt crystals and the silicone composites do not correlate with the size of the changes in PL-QY of the parent QDs upon salt incorporation. We attribute the observed changes to contributions from refractive-index related changes of the radiative rate constants and to QDs, which are initially weakly emitting (or even dark) in solution and turn bright upon crystal incorporation, thereby influencing PL-QY and the PL decay kinetics in a different manner. The observed PL enhancement is ascribed to the curing of surface defects, most likely due to the formation of a thin passivation layer of CdClx. A hint for this hypothesis could be derived from studies with CdSe/ZnS owing a Cd-free surface incorporated into NaCl as well as CdTe QDs incorporated into borax, where the crystal-induced PL-QY increase fell below the values expected for the respective change of the refractive index. The favorable effect on PL-QY in conjunction with the previously reported excellent stability encourages the use of QDs embedded in salt matrices for the design of new photonic materials with tunable optical properties, e.g., for color conversion. The eventually desired derivatization of structure−property relationships for the design of such fascinating materials requires more systematic studies. These should involve different host materials like KCl, NaBr, and KBr as well as QDs of various and well-known surface chemistries, thereby addressing our hypothesis of surface defect curing and influence of refractive index of the matrix. Moreover, silicones or other processable polymers with matching refractive indices to control the scattering characteristics of the QD-salt crystalpolymer composites should be investigated in further studies.





ASSOCIATED CONTENT

S Supporting Information *

Information about the used QDs, including the composition and stabilizers, size, PL-QY data, enhancement factors, PL-LT data, and spectroscopic data. This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

(1) Lee, J.; Sundar, V. C.; Heine, J. R.; Bawendi, M. G.; Jensen, K. F. Adv. Mater. 2000, 12, 1102−1105. (2) Gaponik, N.; Rogach, A. L. Phys. Chem. Chem. Phys. 2010, 12, 8685−8693. (3) Talapin, D. V.; Poznyak, S. K.; Gaponik, N. P.; Rogach, A. L.; Eychmüller, A. Phys. E (Amsterdam, Neth.) 2002, 14, 237−241. (4) Zhang, H.; Cui, Z.; Wang, Y.; Zhang, K.; Ji, X.; Lü, C.; Yang, B.; Gao, M. Adv. Mater. 2003, 15, 777−780. (5) Schubert, E. F. Light-Emitting Diodes, 2nd ed.; Cambridge University Press: Cambridge, U. K., 2006; p 434. (6) Otto, T.; Müller, M.; Mundra, P.; Lesnyak, V.; Demir, H. V.; Gaponik, N.; Eychmüller, A. Nano Lett. 2012, 12, 5348−5354. (7) Müller-Mach, R.; Müller, G. O.; Krames, M. R.; Trottier, T. IEEE J. Sel. Top. Quantum Electron. 2002, 8, 339−345. (8) Demir, H. V.; Nizamoglu, S.; Erdem, T.; Mutlugun, E.; Gaponik, N.; Eychmüller, A. Nano Today 2011, 6, 632−647. (9) Fox, R. B.; Isaacs, L. G.; Stokes, S. J. Polym. Sci., Part A: Gen. Pap. 1963, 1, 1079−1086. (10) Hormats, E. I.; Unterleitner, F. C. J. Phys. Chem. 1965, 69, 3677−3681. (11) Gao, Y.; Ogilby, P. R. Macromolecules 1992, 25, 4962−4966. (12) Gaponik, N. J. Mater. Chem. 2010, 20, 5174−5181. (13) Yu, W. W.; Qu, L.; Guo, W.; Peng, X. Chem. Mater. 2003, 15, 2854−2860. (14) Smith, A. M.; Nie, S. Acc. Chem. Res. 2010, 43, 190−200. (15) Purcell-Milton, F.; Gun’ko, Y. K. J. Mater. Chem. 2012, 22, 16687−16697. (16) Kalytchuk, S.; Zhovtiuk, O.; Rogach, A. L. Appl. Phys. Lett. 2013, 103, 103105. (17) Shavel, A.; Gaponik, N.; Eychmüller, A. J. Phys. Chem. B 2006, 110, 19280−19284. (18) Bae, W. K.; Char, K.; Hur, H.; Lee, S. Chem. Mater. 2008, 20, 531−539. (19) Grabolle, M.; Spieles, M.; Lesnyak, V.; Gaponik, N.; Eychmüller, A.; Resch-Genger, U. Anal. Chem. 2009, 81, 6285−6294. (20) Rogach, A. L.; Franzl, T.; Klar, T. A.; Feldmann, J.; Gaponik, N.; Lesnyak, V.; Shavel, A.; Eychmüller, A.; Rakovich, Y. P.; Donegan, J. F. J. Phys. Chem. C 2007, 111, 14628−14637. (21) Tamang, S.; Beaune, G.; Texier, I.; Reiss, P. ACS Nano 2011, 5, 9392−9402. (22) Würth, C.; González, M. G.; Niessner, R.; Panne, U.; Haisch, C.; Genger, U. R. Talanta 2012, 90, 30−7. (23) Würth, C.; Pauli, J.; Lochmann, C.; Spieles, M.; Resch-Genger, U. Anal. Chem. 2012, 84, 1345−1352. (24) Resch-Genger, U.; Bremser, W.; Pfeifer, D.; Spieles, M.; Hoffmann, A.; DeRose, P. C.; Zwinkels, J. C.; Gauthier, F.; Ebert, B.; Taubert, R. D.; Monte, C.; Voigt, J.; Hollandt, J.; Macdonald, R. Anal. Chem. 2012, 84, 3889−3898. (25) Würth, C.; Grabolle, M.; Pauli, J.; Spieles, M.; Resch-Genger, U. Nat. Protoc. 2013, 8, 1535−50.

AUTHOR INFORMATION

Corresponding Authors

*E-Mail: [email protected]. *E-Mail: [email protected]. 3236

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237

Chemistry of Materials

Article

(26) Würth, C.; Grabolle, M.; Pauli, J.; Spieles, M.; Resch-Genger, U. Anal. Chem. 2011, 83, 3431−3439. (27) Sillen, A.; Engelborghs, Y. Photochem. Photobiol. 1998, 67, 475− 486. (28) Strickler, S. J.; Berg, R. a. J. Chem. Phys. 1962, 37, 814−822. (29) Lide, D. R. Handbook of Chemistry and Physics, 89th ed.; Taylor & Francis Group: Oxfordshire, U. K., 2008; p 2736. (30) Pillonnet, A.; Fleury, P.; Chizhik, A. I.; Chizhik, A. M.; Amans, D.; Ledoux, G.; Kulzer, F.; Meixner, A. J.; Dujardin, C. Opt. Express 2012, 20, 3200−3208. (31) Suhling, K.; French, P. M. W.; Phillips, D. Photochem. Photobiol. Sci. 2005, 4, 13−22. (32) Ebenstein, Y.; Mokari, T.; Banin, U. Appl. Phys. Lett. 2002, 80, 4033. (33) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, 2009; p 954. (34) Wadhavane, P. D.; Galian, R. E.; Izquierdo, M. A.; AguileraSigalat, J.; Galindo, F.; Schmidt, L.; Burguete, M. I.; Pérez-Prieto, J.; Luis, S. V. J. Am. Chem. Soc. 2012, 134, 20554−20563. (35) Hoy, J.; Morrison, P. J.; Steinberg, L. K.; Buhro, W. E.; Loomis, R. A. J. Phys. Chem. Lett. 2013, 4, 2053−2060. (36) Fisher, B. R.; Eisler, H.-J.; Stott, N. E.; Bawendi, M. G. J. Phys. Chem. B 2004, 108, 143−148. (37) Tang, J.; Kemp, K. W.; Hoogland, S.; Jeong, K. S.; Liu, H.; Levina, L.; Furukawa, M.; Wang, X.; Debnath, R.; Cha, D.; Chou, K. W.; Fischer, A.; Amassian, A.; Asbury, J. B.; Sargent, E. H. Nat. Mater. 2011, 10, 765−771. (38) Ning, Z.; Ren, Y.; Hoogland, S.; Voznyy, O.; Levina, L.; Stadler, P.; Lan, X.; Zhitomirsky, D.; Sargent, E. H. Adv. Mater. 2012, 24, 6295−6299. (39) Breus, V. V.; Heyes, C. D.; Nienhaus, G. U. J. Phys. Chem. C 2007, 111, 18589−18594. (40) Wu, S.; Dou, J.; Zhang, J.; Zhang, S. J. Mater. Chem. 2012, 22, 14573−14578.

3237

dx.doi.org/10.1021/cm5009043 | Chem. Mater. 2014, 26, 3231−3237