Volume Scaling - ACS Publications - American Chemical Society

Subscriber access provided by UNIV OF NEWCASTLE

Communication

Effect of Interfacial Alloying versus “Volume Scaling” on Auger Recombination in Compositionally Graded Semiconductor Quantum Dots Young-Shin Park, Jaehoon Lim, Nikolay S. Makarov, and Victor I. Klimov Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b02438 • Publication Date (Web): 27 Jul 2017 Downloaded from http://pubs.acs.org on July 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Effect of Interfacial Alloying versus “Volume Scaling” on Auger Recombination in Compositionally Graded Semiconductor Quantum Dots Young-Shin Park†,‡, Jaehoon Lim†, Nikolay S. Makarov†, and Victor I. Klimov*,† †

‡

Chemistry Division, Los Alamos National Laboratory, Los Alamos, New Mexico, 87545, USA Center for High Technology Materials, University of New Mexico, Albuquerque, New Mexico,

87131, USA *E-mail: [email protected] / Telephone: 505-665-8284

Abstract: Auger recombination is a nonradiative three-particle process wherein the electron-hole recombination energy dissipates as a kinetic energy of a third carrier. Auger decay is enhanced in quantum-dot (QD) forms of semiconductor materials compared to their bulk counterparts. Since this process is detrimental to many prospective applications of the QDs, the development of effective approaches for suppressing Auger recombination has been an important goal in the QD field. One such approach involves “smoothing” of the confinement potential, which suppresses the intraband transition involved in the dissipation of the electron-hole recombination energy.

1 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

The present study evaluates the effect of increasing “smoothness” of the confinement potential on Auger decay employing a series of CdSe/CdS-based QDs wherein the core and the shell are separated by an intermediate layer of a CdSexS1-x alloy comprised of 1 to 5 sublayers with a radially tuned composition. As inferred from single-dot measurements, use of the five-step grading scheme allows for strong suppression of Auger decay for both biexcitons and charged excitons. Further, due to nearly identical emissivities of neutral and charged excitons, these QDs exhibit an interesting phenomenon of lifetime blinking, for which random fluctuation of a photoluminescence lifetime occur for a nearly constant emission intensity.

Keywords: Semiconductor nanocrystal, quantum dot, suppression of Auger recombination, interfacial alloy layer, lifetime blinking, single-dot spectroscopy


Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Auger recombination is a nonradiative process wherein an energy released during electron-hole recombination is transferred to a third carrier and then dissipated as heat.1 In nanocrystal quantum dots (QDs), the Auger decay rate is greatly increased compared to parental bulk solids due to the enhancement in carrier-carrier Coulomb interactions caused by close proximity between interacting charges, effective reduction of dielectric screening, and relaxation of translation momentum conservation.2 Auger recombination leads to quenching of emission from multicarrier states in QDs and represents a performance-limiting factor in a broad range of QDbased applications including light-emitting diodes (LEDs)3 and QD lasers.4, 5 For monocomponent QDs, the Auger recombination time scales linearly with the QD volume (referred below as “V-scaling”), which is a universal trend observed for many semiconductor compositions including direct- and indirect-gap materials.6 According to this scaling, Auger lifetimes quickly increase with increasing QD size, however, even in the largest studied nanoparticles (~10 nm diameter), the Auger recombination rate is still much faster than the radiative decay rate. Thus, in standard QDs, multicarrier states are virtually nonemissive for all practically accessible QD dimensions, suggesting that approaches other than size control are required for suppressing Auger recombination.7 Among such approaches researchers have explored the reduction of electron-hole overlap in type-II and quasi-type-II heterostructures,8-16 the control of conduction- and valence-band mixing,17 and more recently, grading of the confinement potential for one or both carriers (an electron and a hole).18-23

The effect of the shape of the confinement potential on Auger

recombination was first analyzed theoretically by Cragg & Efros.20 Using a one-dimensional model, they showed that by “smoothing” the confinement potential, one could reduce Auger-


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

decay rates by orders magnitude due to suppression of the intraband transition involved in the dissipation of the electron-hole recombination energy. A strong influence of potential grading on Auger decay was first observed experimentally in studies of thick-shell (“giant”) QDs with unintentionally alloyed core/shell interfaces.24,

25

More recent studies explored the effect of intentional interfacial alloying on Auger recombination by incorporating an intermediate CdSe0.5S0.5 alloy layer between the CdSe core and the CdS shell.21, 23 This single-step grading scheme resulted in a considerable increase of photoluminescence (PL) quantum yields (QYs) of charged exciton and biexcitons, as well as the improved performance of QD-LEDs22 and reduced lasing thresholds.5 While the results of the initial studies with a single interfacial alloy layer of a fix composition indicate a considerable promise of the grading approach for controlling Auger recombination, more work is required to identify optimal grading schemes and to distinguish between the effects due to the increased spatial extend of electronic wave functions versus “smoothing” of the confinement potential. To address these questions, here we explore the effect of composition of the graded layer using first a single-step grading approach, and then extend our work to more complex structures with a multi-step interfacial grading. To evaluate the degree of Auger decay suppression, we analyze a biexciton PL QYs of graded-QDs inferred from singledot photon correlation measurements and compare them to values observed for reference samples with “sharp” core/shell interfaces. These studies demonstrate that using a five-step grading scheme, we can boost the biexciton PL QY to ~60% for individual dots and ~31% on average. These values are a considerable improvement versus those in reference samples with a similar effective “excitonic volume”, where the biexciton PL QY is less than ~15%. These results provide direct proof that the primary reason for Auger decay suppression in graded


Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

structures is the engineered shape of the confinement potential but not a trivial volume-scaling effect. We start our studies using a series of structures with a single-step alloying (denoted as C/A(1)/S QDs) where the intermediate CdSexS1-x alloy layer has a fixed thickness (L = 1.5 nm) but a varied composition (x = 0, 0.25, 0.5, 0.75, and 1), which corresponds to a varied height of the potential step separating the CdSe core (radius r = 1.5 nm) and the CdS shell (thickness H = 2.0 nm). The two limiting cases (x = 0 and 1) represent QDs with “sharp” interfaces (referred to C/S QDs) of the same total radius (R = 5 nm) and two different core radii (r = 1.5 and 3 nm, respectively). Examples of transmission electron microscopy (TEM) images for samples with x = 0 and 0.5 are displayed in Figure 1a (panels at right). The PL QY of the synthesized QDs measured in the single-exciton regime (QX) is ~ 60%. More details on the synthesis, and TEM and optical characterizations of the fabricated samples can be found in Experimental Section and Supporting Information (Sections 1 and 2). An approximate shape of the conduction- and valence-band confinement potentials of these QDs for x = 0.5 is displayed in Figure 1a (panel at left). Because of the small conductionband offset between CdSe and CdS, which is further smeared out by the intermediate alloy layer, the electron wave function (Ψe) is delocalized over the entire QD. On the other hand, a large valence-band offset between CdSe and CdS confines the hole wave function (Ψh) primarily to the core region. In the C/A(1)/S structures, the degree of hole confinement can be controlled by the composition of the CdSexS1-x interlayer. Specifically, the increase in x lowers down the confinement potential and results in increasing delocalization of Ψh outside the core. Experimentally, this effect is observed as a red shift of the PL band (from ~625 to ~665 nm) closely followed by the band-edge absorption peak (Figure 1b). 5 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

To experimentally probe the degree of Auger decay suppression, we analyze biexciton PL QYs inferred from a second-order intensity correlation function (g(2)(t), t is time) measured using a pulsed version of a Hanbury Brown & Twiss experiment (pulse-to-pulse separation T = 1 µs, pulse duration ~40 ps, and excitation wavelength 488 nm); see Figure 1c and Experimental Section. As shown previously, if such measurements are conducted without spectrally discriminating between single excitons and biexcitons, the ratio of the central and side peaks (g(2)(0) and g(2)(T), respectively) measured in the limit of low excitation powers provides a direct measure of the ratio of the biexciton and single-exciton PL QYs or the relative biexciton emission efficiency qXX = QXX/QX.24, 26 All single-dot experiments were conducted using low pump intensities for which the average number of excitons injected per-pulse, per-dot did not exceed 0.2. In Figure 1d, we display an example of the g(2) function obtained for the sample with x = 0.5. According to these data, qXX is 0.18. This value can be then compared to one derived from the measured biexciton (τXX) and single-exciton (τX) lifetimes (the latter is acquired during highemissivity “ON” periods): τXX = 2.1 ns and τX = 49 ns (Figure 1e). Assuming that in the ON state, exciton dynamics is purely radiative (i.e., τr,X = τX) and applying statistical scaling for determining the biexciton radiative lifetime7 (τr,XX = τr,X/4), we obtain qXX = kr,XX/kXX = 4τXX/τX = 0.17, where kXX = 1/τXX and kr,XX = 1/τr,XX, are the total and the radiative decay rates of the biexciton state, respectively. The derived value of qXX is in excellent agreement with that determined from the photon-correlation measurements. In Figure 2a, we provide a summary of qXX measurements for 76 individual dots from the C/A(1)/S samples with a varied Se content x in the intermediate alloy layer. These data are


Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

plotted as a function of the effective QD “excitonic” volume expressed as Veff =

8π Re2 Rh2 , 3 Re + Rh

where Re and Rh are the effective localization radii of the electron and hole wave functions, respectively (see Supporting Information, Section 3). As the electron wave function is delocalization over the entire QD, we assume that Re = R independent of x. On the other hand, as discussed earlier, the hole localization radius depends on Se content in the CdSexS1-x interlayer and changes from ~1.5 nm to ~3 nm as x is varied from 0 to 1. To quantify Rh we define it as the radius of a sphere which confines a QD hole with the 90% probability. The Rh computed in this way using a numerical approach of ref 27 is displayed in the inset of Figure 2a and Figure S2c as a function of x. First, we analyze the data for the reference C/S samples with a sharp core/interface (open symbols in Figure 2a). For the C/S QDs with r = 1.5 nm, the relative biexciton PL QY averaged over individual QD measurements, , is only ~2%, which is indicative of very fast Auger recombination. Then the core radius is increased to 2.5 nm and then 3 nm, also increases, first to 11% and then 14%. As expected, these values follow the trend (gray line) predicated by the V-scaling (τA,XX ∝ Veff). The graded C/A(1)/S QDs, for which Veff is tuned through the composition of the intermediate alloy layer but not the core size, show a distinctly different behavior (solid symbols and a blue trend line). As the fraction of Se is increased to x = 0.25, increases to ~7%, which is a factor of ca. 2 greater than the value expected from the V-scaling. This trend continues for x up to 0.75, when becomes as high as ~20%, in contrast to ~10% in the C/S structures with the same effective volume. These results clearly demonstrate the effectiveness of interfacial alloying as a means for suppressing Auger decay. Especially illustrative is the reversal of the


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

trend as x is increased from 0.75 to 1 and the core/shell interface becomes again sharp. In this case, the average biexciton PL QY decreases to less than 15% despite the increase in the effective QD volume. Even more considerable suppression of Auger decay is possible by employing a finer grading strategy when the interfacial alloy layer is assembled of several sublayers with x progressively decreasing in the radial direction. To analyze the effect of the multi-step grading on Auger decay, we have synthesized a series of samples for which the number of intermediate sublayers (interlayers) is progressively increased from n = 1 to 3, and then 5 (referred to as C/A(n)/S QDs). All of these samples have a similar overall size (R = 5 – 6 nm) and PL energy (Supporting Information, Figure S1), but distinct internal compositional profiles described in Supporting Information, Section 2. The QD effective volumes computed using these profiles are plotted in Figure S2d where they are compared with the reference C/S structures shown previously in Figure 2a. These calculations indicate that when going from the C/S sample with r = 1.5 nm to the sample with a single core/shell interlayer, Veff sharply increases (by ~35%) and then remains nearly constant (within about ±10%) as the number of interlayers is increased to 3 and 5. Another sharp increase in the effective volume (by ~50%) occurs then the internal QD structure switches again to the C/S type with the increased core radius (r = 3 nm). These computed variations in Veff are corroborated by the observed changes in the emission spectra (Figure S1e). The PL peak position experiences a large redshifted (~34 nm) as the QD structure changes from the C/S to C/A(1)/S, and then, stays virtually unchanged for the structures with 3 and 5 interlayers. It again moves sharply to the red for the C/S sample with the r = 3 nm. In Figure 2b, we display single-dot qXX measurements along with for the above samples. As was discussed earlier, introduction of a single interlayer already leads to a 8 ACS Paragon Plus Environment

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

considerable increase in qXX (up to ~14% on average; blue circles) versus the C/A sample (black diamonds). A finer grading using initially three (orange circles) and then five (green circles) interlayers leads to a further increase in qXX which is boosted up ~60% in some of the C/A(5)/S dots and is ~31% on average based on the measurements of 32 individual particles; for 16% of the measured dots, qXX is ≥ 50%. In contrast, the biexciton PL QY expected for the similarly sized structures with the sharp core-shell interface would be less than 10%. Interestingly, the qXX values sharply drops for the large-core C/S sample (down to = 14%) despite a considerable increase in the QD volume (by ~50%). These results again provide a convincing illustration of the strong effect of compositional grading (and associated “smoothing” of the confinement potential) on the rate of Auger recombination. A significant promise of this approach as a practical means for controlling Auger decay is further indicated by the comparison of our present study of finely graded QDs to literature results for “giant” CdSe/CdS dots with unintentional alloying24 and QDs with intentional single-step interface grading21 (see the summary of present and literature data in Figure S3). While the applied grading procedures clearly allow for suppression of Auger decay at both single-dot and ensemble levels, they are still imperfect, as indicated by a considerable spread in the qXX values within the QD ensemble, which becomes especially pronounced for the large number of interlayers. This problem is likely a consequence of variations in the composition/structure of the intermediate alloy layer and, in principle, it could be addressed by improving the control of chemistry employed for the growth of the compositionally graded interlayers. One more manifestation of considerable suppression of Auger decay in QDs with a finely graded core/shell interface is the observation of an interesting phenomenon of “lifetime blinking” 9 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

manifested in the fluctuating PL lifetime for a nearly constant PL intensity. It was first reported for the CdSe/CdS giant QDs with an especially thick (~5 nm) shell28 and ascribed to nearly complete suppression of Auger decay for negatively charged excitons (negative trions). In this case, random charging/discharging of the QDs does not cause fluctuations in the PL intensity, but instead only leads to the fluctuating PL lifetime which switches between values characteristic of neutral excitons and trions. Some of our dots with the five-step grading exhibit a similar blinking behavior, as illustrated in Figure 3. First, we establish a single-dot character of these measurements by analyzing the second order intensity correlation function. The g(2) trace obtained in a standard way by recording all pairs of photons independent on the time of their arrival versus a pump pulse (black trace in Figure 3a) shows weak antibunching (g(2)(0) = 0.57, Figure 3c), which might be indicative of emission from a QD cluster but not a single dot. However, the time-gated experiment conducted by collecting only photons arriving at t ≥ 50 ns (i.e., past biexciton decay) shows complete antibunching (Figure 3d). This proves that the measurements are, in fact, conducted in the singledot regime, and further indicate that the biexciton PL QY of the measured dot is very high (qXX = 57%) due to strong suppression of Auger decay. Next, we analyze the single-dot PL intensity (IPL) and lifetime (τPL) trajectories. The PL intensity seems to show usual fluctuations with the ON-time fraction of ~65%. However, when analyzed together with the PL lifetime variations, the observed blinking pattern appears to be unusual. Specifically, the PL dynamics measured during the ON-periods (Figure 3e) exhibits double-exponential decay suggesting the contributions not from one but two states with similar emissivities but distinct lifetimes (64 and 25 ns). While these states are indistinguishable in the PL-intensity histogram (Figure 3b), they can be still visualized using fluorescence-lifetime10 ACS Paragon Plus Environment

Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

intensity-distribution (FLID) plots (Figure 3f), wherein a false-color representation is used to show the probability of the dot to exhibit a certain PL intensity IPL simultaneously with a certain lifetime τPL. The FLID plot in Figure 3f clearly shows two distinct “bright” states characterized by nearly identical intensity levels but different lifetimes; in addition, this plot reveals the third “dark” state (XD) with a considerably lower emissivity and a shorter lifetime. As in the study of ref

28

mentioned earlier, the observed behavior can be explained by

fluctuations between neutral and charged exciton states, where the latter is nearly Auger-decay free. Indeed, for purely radiative recombination, the trion lifetime is half of the neutral-exciton lifetime. If we assign the longer-lived “bright” state to a neutral exciton, then the expected radiative trion lifetime would be 32 ns. The lifetime of the second “bright” state (25 ns) is only slightly shorter than this value, which can be ascribed to the contribution from slow, residual Auger decay with the time constant of ~114 ns. Based on this analysis, the trion PL QY is 78% (25 ns divided by 32 ns), indicating a nearly complete elimination of Auger decay, which again confirm the effectiveness of compositional grading in suppressing Auger recombination. Now, we address the question on the sign of the observed “bright” trion. Previous spectroelectrochemical29 and magneto-optical30 studies have suggested that in the case of CdSe/CdS QDs, photoionization produces primarily negatively charged species. More recent single-dot measurements have also revealed blinking events due to positive trions (X+) manifested as “dark” states with emissivities below those of negative trions (X-).31 This difference in emissivity levels indicates that the X+ Auger decay is faster than that of the Xspecies. The asymmetry between X+ and X- Auger decay channels in II-VI QDs has been also observed in direct measurements of charged exciton dynamics and attributed to the higher spectral density of valence- versus conduction-band states, which favors Auger processes 11 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

involving re-excitation of a hole. As discussed earlier, the Auger lifetime of the trion responsible for the short-lived bright state in our QDs is 114 ns. To determine the lifetime of the oppositely charged trion, we apply a superposition principle,32 according to which the biexciton Auger decay can be described in terms of superposition of negative- and positive trion pathways,31 and hence, the corresponding lifetimes are connected by 1/τA,XX = 2(1/τΑ,X+ + 1/τΑ,X-). Based on qXX = 0.57 (Fig. 3c) and τX = 64 ns (Fig. 3e), we find that the total biexciton lifetime is 9.1 ns, which yields 21 ns for its Auger lifetime (determined from 1/τA,XX = 1/τXX – 1/τr,XX). Using this value along with the “bright” trion Auger lifetime of 114 ns, we obtain from the superposition principle that the Auger time constant of the oppositely charged trion is 67 ns. Given that τA,X+ is expected to be shorter than τA,X-, the time constants of 114 ns and 67 ns can be attributed to negative and positive trions, respectively. Interestingly, while in standard core-only and non-graded core/shell QDs, positive trions are considerably dimmer than negative trions,31 in our graded QDs both trions have comparable (and high) emission efficiencies. Indeed, based on the derived positive trion Auger lifetime, the X+ PL QY is 68%, which is only ~13% lower (in relative terms) than that of the negative trion. Given this small difference, we cannot exclude that in addition to the X- species, the faster “bright” state is our QDs is also contributed by the X+ species, which might be responsible for slight “diagonal” broadening of the X* feature apparent in the FLID diagram in Figure 3f. This analysis also indicates that the low-emissivity “dark” state in our QDs (PL QY of ~3%; see Figure 3b &3f) is not due to trions (either negative or positive) but rather is associated with some non-Auger-decay-related nonradiative process as discussed in, e.g., refs 33, 34. To summarize, we have synthesized a series of multilayered CdSe/CdS QDs wherein the core and the shell are separated by 1 to 5 radially graded CdSexS1-x alloy layers. We have used 12 ACS Paragon Plus Environment

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

these samples to examine the role of QD-effective-volume scaling versus interfacial alloying in suppression of Auger recombination. To quantify the degree of this suppression, we analyze biexciton emission efficiencies inferred from single-dot second-order intensity correlation measurements. We demonstrate that using a five-step grading scheme, we can boost single-dot biexciton PL QYs to ~60% and ~31% on average; the latter is a ca. three-fold improvement compared to non-graded structures with a similar effective volume. Even stronger suppression of Auger decay is observed for charged excitons. In fact, in some of the measured QDs, emissivities of charged excitons and neutral exciton are nearly identical, which leads to an interesting phenomenon of lifetime blinking during which PL lifetime fluctuations are not accompanied by fluctuations of the PL intensity. These experimental findings confirm the effectiveness of confinement-potential “smoothing” in core/shell structures with alloyed interfaces as a practical means for considerable suppression of Auger decay. The demonstrated strategy might help improve the performance of QD-based devices and especially LEDs and lasers as both of these applications would benefit from highly efficient emission from charged and neutral multicarrier states. Synthesis of C/S and C/A(n)/S QDs. In order to prepare C/S and C/A(n)/S QDs we used a modified version of the synthetic procedure by Bae et al22. Briefly, to fabricate samples with a single-step alloying (C/A(1)/S QDs), we slowly added a mixture of Se-(n-trioctylphosphine) and 1-dodecanethiol (0.3 M in 1-octadecene), and cadmium oleate (0.3 M in 1-octadecene) separately to 0.7 µmol CdSe QDs (radius r = 1.0 nm or 1.5 nm) and 0.1 mmol cadmium oleate in 20 mL 1-octadecene at 300 °C. The chemical composition and the thickness of the alloyed layer was controlled by a molar ratio of Se and 1-dodecanethiol in the precursor mixture (xSe,p = [Se]/([Se] + [S])) and the amount of precursors. In the case of the C/A(n)/S QDs with n = 3 and


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

5, the alloyed layer was prepared through a stepwise addition of precursor solutions using a set of specific conditions defined in terms of precursor volume (mL)/xSe,p/injection rate (mL/hr). The growth of the QDs with the 3- and the 5-step grading was conducted using, respectively, the sequence of the following reaction steps: (1) 0.6/0.59/1.6, (2) 0.8/0.41/1.4, (3) 1.0/0.25/1.2, and (1) 0.23/0.81/1.5, (2) 0.46/0.59/1.5, (3) 0.8/0.41/1.4, (4) 1.2/0.25/1.3, and (5) 1.6/0.09/1.3. After the alloyed layer was grown, the resulting C/A(n) QDs were capped with the CdS shell using cadmium oleate and 1-dodecanethiol at 300 °C as precursors. The growth was terminated when the overall radius of the final C/A(n)/S QD sample had reached 10 nm. Single-dot spectroscopy. Single-dot studies were performed using a home-built confocal fluorescence microscope system. A dilute QD sample in hexane was drop-cast onto a glass cover slip to produce a sub-monolayer film with a dot-to-dot separation of ~10 µm or greater. A pulsed laser (488 nm wavelength, ~40 ps pulse width, and 1 MHz repetition rate) was used for photoexcitation, and a 100X objective lens (numerical aperture 0.9) was employed for both focusing the laser beam onto the sample and collecting emitted PL. Pump power corresponded to a low-intensity excitation regime when the average per-dot excitonic occupancy did not exceed 0.2. Time-correlated photon counting was conducted using a standard Hanbury Brown and Twiss configuration with two matched, single-photon avalanche photodiodes (350 ps time resolution). A time-tagged, time-resolved mode was used to record both a macro-time (referred to the beginning of a measurement) and a micro-time (referred to each laser pulse) for each arriving photon in both detection channels. All measurements were carried out at room temperatures under ambient condition.


Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 1. (a) Left: Schematic illustration of a C/A(1)/S QD which contains a single intermediate layer of a CdSexS1-x alloy of a varied composition between a CdSe core and a CdS shell (in this example x = 0.5). Black and red lines show, respectively, calculated electron (Ψe) and hole (Ψh) wave functions. Right: TEM images of the reference C/S QDs and the C/A(1)/QDs with x = 0.5. Both samples have the same mean total QD radius of 5 nm. (b) Absorption and PL spectra of a series of similarly sized C/A(1)/S QDs (r = 1.5 nm, L = 1.5 nm, H = 2 nm) with a varied content of Se in the alloyed layer (x = 0, 0.25, 0.5, 0.75, and 1). (c) A Hanbury Brown and Twiss setup used in measurements of the second-order intensity correlation function g(2); APD stands for “avalanche photodiode.” (d) An example of a g(2) function obtained for a single C/A(1)/S QD with x = 0.5. The ratio between the areas of the central and the side peaks measured in the low pump-intensity limit yields the relative biexciton PL QY (qXX = QXX/QX). (e) Single-exciton (black circles) and biexciton (red squares) PL dynamics measured for the same QD. Lines are single-exponential fits. 15 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

Figure 2. (a) Relative biexciton PL QYs of the C/S and C/A(1)/S samples inferred from singledot measurements of the second-order intensity correlation functions as a function of QD effective “excitonic volume.” The overall radius of QDs in all samples is 5 nm. For the C/S samples, the core radii are 1.5 nm, 2.5 nm, and 3 nm, while the corresponding shell thicknesses are 3.5 nm, 2.5 nm, and 2 nm; these dimensions are indicated in the plot as r(nm)/H(nm). All C/A/S samples have the same dimensions of the core (r = 1.5 nm), the intermediate alloyed layer (L = 1.5 nm), and the shell (H = 2 nm). The varied parameter is the alloy composition in the 16 ACS Paragon Plus Environment

Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

intermediate layer (indicated in the figure in terms of the percentage of Se). Schematics depict the changes in the electron and hole confinement potentials for the C/S (right) and C/A(1)/S (left) samples. The inset is the plot of the effective hole confinement radius (Rh) versus Se content. Red circles with error bars are single-dot averages. Thick lines (blue and gray) are the guide for the eye. (b) Single-dot relative biexciton PL QYs of a series of samples with increasing number of compositionally graded alloyed interlayers as a function of QD effective volume. The displayed samples include C/S QDs with the core radii 1.5 nm (black diamonds) and 3 nm (gray diamonds), the C/A(1)/S QDs with x = 0.50 (blue circles), the C/A(3)/S QDs (orange circles), and the C/A(5)/S QDs (green circles); see the text for structural/compositional details of these samples. Insets illustrate the approximate shapes of the electron and hole confinement potentials in the C/S and C/A/S QDs (the same color-coding as for experimental data points).


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 22

Figure 3. (a) The PL intensity trajectory and (b) the corresponding PL intensity histogram for a single C/A(5)/S QD. X, X* and XD denote a neutral exciton, a charged exciton, and a lowemissivity “dark” state, respectively. The bin time is 50 ms. (c) The g(2)(t) functions obtained by collecting all photon pairs independent on the time of their time delay (td) versus the pump pulse. (d) Same but based on statistics of photons arriving at t ≥ 50 ns. (e) PL dynamics of the highemissivity, “bright” state for two emission levels shown in panel ‘a’ by orange and green shadings; PL traces have the same color coding (symbols for measurements and lines for doubleexponential fits). The dynamics of the low-emissivity “dark” state (purple diamonds) is fit to single exponential decay (purple line); the corresponding PL intensity level is shown in panel ‘a’ by purple shading. (f) The FLID plot revels that the “bright state” is due to superposition of at least two states of similar emissivities but different lifetimes. 18 ACS Paragon Plus Environment

Page 19 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

ASSOCIATED CONTENT Supporting Information. The Supporting Information is available free of charge on the ACS Publications website. TEM images, and absorption and PL spectra of C/S and C/A/S QDs, simulation of effective volumes of C/S and C/A(n)/S QDs, single-dot biexciton emission efficiencies (comparison to literature results) (PDF)

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] Author Contributions Y.-S. P. and J. L. made equal contributions to this work. ACKNOWLEDGMENT Y.-S.P, N.S.M., and V.I.K were supported by the Chemical Sciences, Biosciences and Geosciences Division, Office of Basic Energy Sciences, Office of Science, U.S. Department of Energy. J.L. was supported by the Laboratory Directed Research and Development program at Los Alamos National Laboratory. REFERENCES (1) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Science 2000, 287, 1011-1013. (2) Pietryga, J. M.; Zhuravlev, K. K.; Whitehead, M.; Klimov, V. I.; Schaller, R. D. Phys. Rev. Lett. 2008, 101, 217401. (3) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354-357. 19 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 22

(4) Klimov, V. I.; Mikhailovsky, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leatherdale, C. A.; Eisler, H.-J.; Bawendi, M. G. Science 2000, 290, 314-317. (5) Park, Y.-S.; Bae, W. K.; Baker, T.; Lim, J.; Klimov, V. I. Nano Lett. 2015, 15, 73197328. (6) Robel, I.; Gresback, R.; Kortshagen, U.; Schaller, R. D.; Klimov, V. I. Phys. Rev. Lett. 2009, 102, 177404. (7) Klimov, V. I. Annu. Rev. Condens. Matter Phys. 2014, 5, 285-316. (8) Oron, D.; Kazes, M.; Banin, U. Phys. Rev. B 2007, 75, 035330. (9) Klimov, V. I.; Ivanov, S. A.; Nanda, J.; Achermann, M.; Bezel, I.; McGuire, J. A.; Piryatinski, A. Nature 2007, 447, 441-446. (10) Ivanov, S. A.; Piryatinski, A.; Nanda, J.; Tretiak, S.; Zavadil, K. R.; Wallace, W. O.; Werder, D.; Klimov, V. I. J. Am. Chem. Soc. 2007, 129, 11708-11719. (11) Mahler, B.; Spinicelli, P.; Buil, S.; Quelin, X.; Hermier, J.-P.; Dubertret, B. Nat. Mater. 2008, 7, 659-664. (12) Chen, Y.; Vela, J.; Htoon, H.; Casson, J. L.; Werder, D. J.; Bussian, D. A.; Klimov, V. I.; Hollingsworth, J. A. J. Am. Chem. Soc. 2008, 130, 5026-5027. (13) Rabouw, F. T.; Lunnemann, P.; van Dijk-Moes, R. J. A.; Frimmer, M.; Pietra, F.; Koenderink, A. F.; Vanmaekelbergh, D. Nano Lett. 2013, 13, 4884-4892. (14) Zavelani-Rossi, M.; Lupo, M. G.; Tassone, F.; Manna, L.; Lanzani, G. Nano Lett. 2010, 10, 3142-3150. (15) Sitt, A.; Sala, F. D.; Menagen, G.; Banin, U. Nano Lett. 2009, 9, 3470-3476. (16) Vaxenburg, R.; Rodina, A.; Lifshitz, E.; L. Efros, A. Nano Lett. 2016, 16, 2503-2511. (17) Qin, W.; Liu, H.; Guyot-Sionnest, P. ACS Nano 2013, 8, 283-291. (18) Nasilowski, M.; Spinicelli, P.; Patriarche, G.; Dubertret, B. Nano Lett. 2015, 15, 39533958. (19) Climente, J. I.; Movilla, J. L.; Planelles, J. Small 2012, 8, 754-759. (20) Cragg, G. E.; Efros, A. L. Nano Lett. 2010, 10, 313-317. (21) Park, Y.-S.; Bae, W. K.; Padilha, L. A.; Pietryga, J. M.; Klimov, V. I. Nano Lett. 2014, 14, 396-402. (22) Bae, W. K.; Park, Y.-S.; Lim, J.; Lee, D.; Padilha, L. A.; McDaniel, H.; Robel, I.; Lee, C.; Pietryga, J. M.; Klimov, V. I. Nat. Commun. 2013, 4, 3661. (23) Bae, W. K.; Padilha, L. A.; Park, Y.-S.; McDaniel, H.; Robel, I.; Pietryga, J. M.; Klimov, V. I. ACS Nano 2013, 7, 3411-3419. (24) Park, Y.-S.; Malko, A. V.; Vela, J.; Chen, Y.; Ghosh, Y.; Garcia-Santamar, F.; Hollingsworth, J. A.; Klimov, V. I.; Htoon, H. Phys. Rev. Lett. 2011, 106, 187401. (25) García-Santamaría, F.; Chen, Y.; Vela, J.; Schaller, R. D.; Hollingsworth, J. A.; Klimov, V. I. Nano Lett. 2009, 9, 3482-3488. (26) Nair, G.; Zhao, J.; Bawendi, M. G. Nano Lett. 2011, 11, 1136-1140. (27) Lin, Q.; Makarov, N. S.; Koh, W.-k.; Velizhanin, K. A.; Cirloganu, C. M.; Luo, H.; Klimov, V. I.; Pietryga, J. M. ACS Nano 2015, 9, 539-547. (28) Galland, C.; Ghosh, Y.; Steinbrück, A.; Hollingsworth, J. A.; Htoon, H.; Klimov, V. I. Nat. Commun. 2012, 3, 908. (29) Galland, C.; Ghosh, Y.; Steinbruck, A.; Sykora, M.; Hollingsworth, J. A.; Klimov, V. I.; Htoon, H. Nature 2011, 479, 203-207.


Page 21 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(30) Javaux C; Mahler B; Dubertret B; Shabaev A; Rodina, A. V.; EfrosAl, L.; Yakovlev, D. R.; LiuF; BayerM; CampsG; BiadalaL; BuilS; QuelinX; Hermier, J. P. Nat. Nanotechnol. 2013, 8, 206-212. (31) Park, Y.-S.; Bae, W. K.; Pietryga, J. M.; Klimov, V. I. ACS Nano 2014, 8, 7288-7296. (32) Wu, Kaifeng; Lim, Jaehoo; Klimov, V. I. ACS Nano 2017, DOI: 10.1021/ acsnano.7b04079. (33) Frantsuzov, P. A.; Marcus, R. A. Phys. Rev. B 2005, 72, 155321. (34) Frantsuzov, P. A.; Volkán-Kacsó, S.; Jankó, B. Phys. Rev. Lett. 2009, 103, 207402.


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 22

For TOC only


Volume Scaling - ACS Publications - American Chemical Society

Recommend Documents