Two Mechanisms Determine Quantum Dot Blinking - ACS Nano (ACS

www.acsnano.org

Two Mechanisms Determine Quantum Dot Blinking Gangcheng Yuan,† Daniel E. Gómez,‡ Nicholas Kirkwood,†,§ Klaus Boldt,†,⊥ and Paul Mulvaney*,† †

ARC Centre of Excellence in Exciton Science, School of Chemistry, University of Melbourne, Parkville, Victoria 3010, Australia RMIT University, Melbourne, Victoria 3000, Australia

‡

S Supporting Information *

ABSTRACT: Many potential applications of quantum dots (QDs) can only be realized once the luminescence from single nanocrystals (NCs) is understood. These applications include the development of quantum logic devices, single-photon sources, long-life LEDs, and single-molecule biolabels. At the single-nanocrystal level, random fluctuations in the QD photoluminescence occur, a phenomenon termed blinking. There are two competing models to explain this blinking: Auger recombination and surface trap induced recombination. Here we use lifetime scaling on core−shell chalcogenide NCs to demonstrate that both types of blinking occur in the same QDs. We prove that Augerblinking can yield single-exponential on/of f times in contrast to earlier work. The surface passivation strategy determines which blinking mechanism dominates. This study summarizes earlier studies on blinking mechanisms and provides some clues that stable single QDs can be engineered for optoelectronic applications. KEYWORDS: quantum dots, photoluminescence intermittency, Auger recombination, surface states

Q

In the MRC model, multiple emission states arise due to a set of nonradiative recombination centers that are switching between activation and deactivation, and no charging is required. With progress in QD synthesis and measurement, further blinking-related behaviors have been identified.24−30 Blinking can be suppressed by passivating the traps or suppressing Auger recombination,31 and even nonblinking QDs have been realized.32−36 Despite empirical progress toward the control of blinking, its exact mechanism is still under debate.4,37−39 In addition to the Auger-blinking, a different type of blinking due to interception of hot carriers (“HC-blinking”) has been reported, which is distinct from BC-blinking.30 Since neither the charging model nor the MRC model excludes the other, a combination of both blinking models has also been tried.40 Although pioneering works point out a combination of Auger and BC blinking in QDs,41−43 more direct evidence is still needed. Here, we report unequivocal evidence that both types of blinking occur in the same QDs, thus unifying previous experiments and models. Emissive states with reduced QYs but equal radiative decay rates compared to the bright states are observed. These gray states cannot be explained by the charging model. On the contrary, they provide direct evidence that blinking can be initiated via opening and closing of nonradiative recombination centers. We compare the radiative lifetime

uantum dots (QDs) are beginning to appear in modern electronic devices such as light-emitting diodes1 and single-photon sources,2 but their performance is limited by blinking, a photoluminescence (PL) fluctuation between bright (on) and dark (of f) states.3,4 A charging model5 was first proposed to explain blinking. In this model, the PL fluctuation is due to photoionization6 and neutralization. In a neutral QD, on state emission is produced via radiative recombination. Once the QD is photoionized, fast Auger recombination quenches the emission via transfer of the exciton energy to the third carrier in the core. However, while the early charging model predicts an exponential distribution of both the on and of f durations, they are found to be power-law distributed from experiment.7,8 Modified charging models have been presented to explain the origin of power-law blinking, typically by varying the barriers into and out of multiple traps or by energetic diffusion.9−12 We will denote this type of behavior “Auger-blinking”. Over the past 10 years, it has been realized that the simple charging model is not sufficient to explain the origin of the of f state.13,14 The Auger quenching rate of a singly charged exciton (trion)15 does not explain the low quantum yield (QY) of the off state,13,16,17 unless multiply charged excitons are invoked; alternatively, it is possible that charging may not be the only reason for the off state. It has also been reported that there exist continuous emission states with varying nonradiative rates but fixed radiative rates.18−21 This behavior is referred to here as “BC-blinking” because, as we show later, such blinking is from band-edge carrier trapping. Such observations are consistent with a model without the long-lived traps22 and also with a later model using multiple recombination centers (MRC model).23 © XXXX American Chemical Society

Received: December 22, 2017 Accepted: March 26, 2018 Published: March 26, 2018 A

DOI: 10.1021/acsnano.7b09052 ACS Nano XXXX, XXX, XXX−XXX

Article

Cite This: ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 1. BC-blinking (top panels) and Auger-blinking (bottom panels). Two single QDs (QD 1 and QD 2, QD@639) were excited at 100 nW. The repetition rates are 2.5 and 10 MHz for QD 1 and QD 2, respectively. (a) BC-blinking. When the trapping channel is blocked, only the radiative rate kr exists; when the trapping channel is unblocked, the nonradiative rate knr appears. (b) Photoluminescence intensity trace of QD 1. Multiple intensity levels exist including a bright state (exciton 1, red line), a well-defined gray exciton state (exciton 2, black line), and a trion state (black dashed line). (c) PL decays of QD 1. The bright state (exciton 1, red line) and the gray state (exciton 2, black line). (d) FLID of QD 1. The white arrow indicates the trion emission. The red arrow indicates the background noise. (e) Auger-blinking. In the exciton state, only the radiative rate kr exists; in the trion state, the nonradiative Auger rate kA appears, and the radiative rate becomes 2kr. (f) Photoluminescence intensity trace of QD 2. (g) Photoluminescence time decays for the three intensity levels of QD 2. Exciton (220 counts/20 ms, blue line), trion (50 counts/20 ms, red line), and mixed (110 counts/20 ms, black line). (h) FLID of QD 2. The white line is given by eq 2.

only due to changes in the nonradiative channels, as depicted in Figure 1a. In many cases blinking was too fast to resolve gray states (our bin width was 20 ms). However, we were able to confirm the unchanged radiative lifetime in a few cases where we observed slow intensity-switching events. Such an analysis is presented in Figure 1. The traces in Figure 1b show two periods with stable PL intensities, corresponding to a bright state (red line) and a gray state (black line), which are selected for the lifetime−intensity analysis below. Within each period, the intensity is stable (I1 = 167 counts/20 ms, I2 = 100 counts/ 20 ms, and the background noise is 2 counts/20 ms), and the PL decays in Figure 1c can be fitted well by single-exponential functions (τ1 = 23.2 ns and τ2 = 14.3 ns). The radiative lifetime ratio of the two periods is close to unity as

scaling, examine correlations in QD lifetime, and analyze the PL intensity fluctuations for QDs exhibiting both BC-blinking and Auger-blinking. We show that almost all blinking behaviors in QDs can be understood via one of these two models. The passivation condition determines which blinking mechanism plays the main role in QDs. We do not include HC-blinking in the analysis because it was not observed in our systems.

RESULTS BC-Blinking. One of the reasons for blinking is opening and closing of nonradiative channels (Figure 1a). Such channels can be created by fluctuations in adsorbate binding to the semiconductor crystal surface. This model predicts a linear relationship between PL intensity and lifetime. The intensity I is proportional to the QY, and hence directly proportional to the lifetime τ, and inversely proportional to the radiative lifetime τr according to I ∝ QY =

kr τ = τk r = τr k r + k nr(t )

100 − 2(background) τr1 τ I 23.2 = 1 2 = × ≈ 0.96 τr2 τ2 I1 14.3 167 − 2(background)

From the unity radiative lifetime scaling, we infer that the gray states here are not charged or trion states, because the radiative lifetime scaling between the charged and neutral states is approximately 2,15 as shown later. Hence, there must be a mechanism other than charging that is responsible for the blinking in Figure 1b−d. Our results are partially consistent with the MRC model.23 In the MRC model, in addition to the radiative pathway, the exciton can also nonradiatively relax through multiple recombination centers. The opening and closing of each center modulates the nonradiative rate and hence the PL intensity. Although it is challenging to explain why these centers randomly switch on and off,46−48 this linear correlation of lifetime and intensity has been observed frequently in different QD systems.18−20,49 If we make an assumption that only radiative recombination exists in the

(1)

where kr is the radiative decay rate and knr(t) is the nonradiative decay rate that evolves with time. Figure 1b shows a typical trace of the PL intensity as a function of time collected from a single QD (QD 1, QD@639, graded shell CdSe/CdxZn1−xS QDs;44 see Methods; the absorption and fluorescence spectra as well as TEM images for the QDs can be found in our recently published work45). The PL jumps among a set of intensity levels. The linear correlation between the PL lifetime and intensity is confirmed by the plot of the fluorescence lifetime−intensity distribution (FLID) in Figure 1d. From the linear FLID pattern, we can infer that the radiative decay rate kr (or the radiative lifetime τr) does not change with time. Therefore, the fluctuations in the PL intensity in Figure 1b are B


Article

ACS Nano bright state,45 then the radiative lifetime is 23.2 ns and the nonradiative lifetime is 37.3 ns for the gray state. One may doubt whether the multistate emission is from single QDs or clusters. But cluster emission can be excluded using a simple argument. We consider the emission of two quantum dots as an example. There are two possibilities. First, if there are two quantum dots with different PL decay time constants, then the PL decay from the highest emission level should be biexponential, because it contains emission from each of the two quantum dots. However, we find that the PL decay from the most intense level is monoexponential. Second, if the two quantum dots have the same PL decay time constants, then each emission level should have the same PL decay rate. However, we find that the lifetime is different for each state. Hence, the emission here cannot stem from QD clusters but arises from single QDs. Interestingly, the BC-blinking here is comparable to the HC-blinking reported previously. Both are related to nonradiative recombination;30 however, the competition between radiative and nonradiative pathways is different. HC-blinking results from the activation and deactivation of the bypass channel, which may be associated with emptying and filling of the corresponding surface trap states. In HC-blinking events, some hot carriers are intercepted by surface states before cooling down to the band edge. These carriers do not contribute to the PL intensity because of nonradiative recombination following the interception. But some unintercepted hot excitons can successfully reach the band edge after cooling and contribute to the PL intensity and lifetime in the measurement. In the single-exciton regime, the PL lifetime of band-edge excitons does not change with light intensity. Conversely, in the case of BC-blinking observed here, the competition between the radiative and nonradiative recombinations occurs after the cooling of hot excitons. In the nonradiative process, hot excitons first reach the band edge; then the hole (electron) is trapped in surface states, and subsequently, it recombines nonradiatively with the core state electron (hole).14 These traps are short-lived (e.g., shallow traps), and the time scale of trapping and nonradiative recombination is close to that of radiative recombination of the band-edge exciton. The competition between the fixed radiative and fluctuating nonradiative relaxations leads to the linear correlation between lifetime and intensity. In Figure 1b, there is an intermediate level around 50 counts/20 ms, marked by a dark dashed line. This corresponds to the small island in Figure 1d, as indicated by the white arrow above the linear FLID pattern. It arises from trion emission as discussed later. Indicated by the red arrow, there is also a pattern near the horizontal axis in the FLID. This is from the background and statistical noise. As shown in the Methods section, the average lifetime is defined based on the average photon arrival time relative to the excitation pulse for each time bin. Because the average photon arrival time is a sample mean within a small time bin, it always has statistical noise associated with it. When the intensity is low, the average photon arrival time is determined by background photons, which are almost independent of the time of the excitation pulse. Thus, within this pattern where the intensity is below 30 counts/20 ms, the average lifetime ranges from 0 to 50 ns. Auger-Blinking. When the BC-blinking can be suppressed, it is very easy to observe the conventional Auger-blinking, due to exciton−trion transitions (Figure 1e), especially at high excitation powers. The Auger-blinking has features distinct from the BC-blinking. A typical exciton−trion trace from a

single QD (QD 2, QD@639) is presented in Figure 1f. Two intensity levels can clearly be seen, corresponding to exciton X and trion X*, respectively (IX = 220 counts/20 ms, and IX* = 50 counts/20 ms). As the PL intensity switches mostly between exciton and trion states, the dark state with background noise (5 counts/20 ms) is ignored in the following analysis. Figure 1g shows the monoexponential PL decays for the exciton and trion states. After single-exponential function fitting, exciton lifetime τX and trion lifetime τX* are estimated to be 29.6 and 3.1 ns, respectively. The ratio of the two radiative lifetimes, τXr and τX*r, is τX r τ I 29.6 50 − 5 = X X* = × ≈ 2.00 τX * r τX * IX 3.1 220 − 5

For radiative recombination of excitons, one electron recombines with one hole; for trion radiative recombination, both electrons (holes) have a chance to recombine with one hole (electron). The scaling order, 2, is a useful signature of exciton−trion blinking.15 If we make an assumption that only radiative recombination exists in the bright exciton state,45 then the radiative lifetime is 29.6 ns and the nonradiative Auger lifetime is 3.5 ns. Figure 1g also presents the PL decay of an arbitrarily selected intermediate state (110 counts/20 ms) between exciton and trion states. The biexponential PL decay of this intermediate state can be fitted with a combination of trion and exciton decay constants. This implies that the intermediate state is not a real state but an artifact related to the finite time resolution of our experiment, resulting in the mixing of exciton and trion emissions. This is also confirmed again by the curvature of the FLID in Figure 1h. During a unity bin time T, a single QD stays in state 1 (intensity I1 and lifetime τ1) for time T1, and in state 2 (intensity I2 and lifetime τ2) for time T2. The average lifetime is calculated based on the average photon arrival time relative to the laser pulse. Then the average intensity I and the average lifetime τ for the whole bin time T are T = T1 + T2

I=

I1T1 + I2T2 T

τ=

I1T1τ1 + I2T2τ2 I1T1 + I2T2

Then the average lifetime τ is inversely proportional to the average intensity I as τ=

I1I2(τ2 − τ1) 1 I τ − I2τ2 + 11 I1 − I2 I I1 − I2

(2)

Although a bit different in the average lifetime definition, eq 2 is similar to the early report.30 This equation is very different from eq 1. Inputting the known PL lifetimes and intensities of both the exciton and trion states, the curvature in the FLID plot is reproduced by eq 2 (the white line in Figure 1h). Importantly, surface traps play an important role in both Auger-blinking and BC-blinking. Adsorbate mobility, surface diffusion, and desorption events generate short-lived traps (e.g., shallow traps), which lead to nonradiative-rate fluctuations and hence BC-blinking, while for the formation of the trion longlived traps are required (e.g., deep traps). Once a carrier from the first exciton is trapped, it must be relatively long-lived so C


Article

ACS Nano

Figure 2. Coexistence of BC-blinking and Auger-blinking. Two single QDs (QD@639) were excited at 400 nW, 10 MHz repetition rate: QD 3 (the top three panels) and QD 4 (the bottom three panels). (a, d) Photoluminescence intensity traces: (a) QD 3, multiple intensity levels appear, but there is no clear trion level; (d) QD 4, trion emission is discernible. (b, e) FLIDs of (b) QD 3 and (e) QD 4. Trions are labeled with red triangles. (c, f) Exciton (blue line) and trion (red line) decays of (c) QD 3 and (f) QD 4.

Figure 3. Exponential and power-law Auger-blinking. The top four panels are the exponential blinking from QD 2 (QD@639) excited at 100 nW, 10 MHz repetition rate. The bottom four panels are the power-law blinking from QD 5 (a different batch, QD@618) excited at 200 nW, 5 MHz repetition rate. Bin size is 20 ms. (a, e) Photoluminescence intensity traces: (a) QD 2 and (e) QD 5. (b, f) Histograms of the measured intensity: (b) QD 2 and (f) QD 5. The red dashed lines divide the intensity from low to high into four parts: dark state, trion state, mixed state, and exciton state. (c, g) Statistics of exciton state duration (black squares) on a semilogarithmic scale: (c) QD 2 and (g) QD 5. The distributions are fitted by a single-exponential function and a truncated power-law function (blue line), respectively. (d, h) Statistics of trion state duration (black squares) on a semilogarithmic scale: (d) QD 2 and (h) QD 5. Both distributions are fitted by single-exponential functions (blue line).

that there is enough time to generate a second exciton to form a trion. Coexistence of the Two Blinking Mechanisms. Pure BC-blinking or pure Auger- (exciton−trion) blinking occurs rarely. In most cases, both mechanisms coexist and a mixture of blinking behaviors is observed. This coexistence is illustrated by two typical single QDs, QD 3 and QD 4 in Figure 2 (both are QD@639). The emission of QD 3 jumps among a set of levels, and the trion emission level is blurred. Conversely, for QD 4, the trion state is located at an intensity level of 180 counts/20 ms. Despite a slight bending in the FLID of QD 4, both FLIDs are basically linear (Figure 2b and e). As mentioned above, this

linear correlation between the intensity and lifetime is a signature of BC-blinking. In the lower left corner of both FLIDs, there is a small feature above the main linear pattern indicated by red triangles in Figure 2b and e, which is a signature of the trion state. We can compute the radiative lifetime scaling. The brightest exciton state is selected for computation. In principle, any part of the linear pattern of FLID can be used because they all are from excitons, but each part of the distribution corresponds to a period with different nonradiative decay rates. The single-exponential PL decays of the brightest exciton state and of the trion state are plotted in D


Article

ACS Nano

Figure 4. Effect of excitation power on exponential Auger-blinking. QD 6 (top four panels) and QD 7 (bottom four panels) are from QD@ 639. The laser repetition is 5 MHz. (a, e) Linear dependence of exciton (black solid squares) and trion (red open squares) intensities on excitation power: (a) QD 6 and (e) QD 7. (b, f) Dependence of charging (black solid squares) and discharging (red open squares) rates on excitation power: (b) QD 6 and (f) QD 7. The charging and discharging rates are extracted from (c), (d), (g), and (h) with the 95% confidence bounds smaller than the marker size. (c, g) Exponential distributions of exciton state duration: (c) QD 6 and (g) QD 7. (d, h) Exponential distributions of trion state duration: (d) QD 6 and (h) QD 7. Each distribution is labeled with excitation power.

behavior implies that the charging and discharging rates do not change much with time. Multiple traps with varying barriers are used to explain the power-law blinking.10 The exponential blinking indicates that the distribution of trap barriers is narrow in QD 3.57 In comparison, a different single QD with a smaller core size, QD 5, from the other batch of QDs (QD@618) shows different exciton−trion blinking. While the same exponential distribution is observed for the trion state duration in Figure 3h, the exciton state duration is power-law governed in Figure 3g. The QD 5 switches rapidly between the neutral and charged states in the first 25 s. The blinking rate is substantially reduced in the following 25 s (Figure 3e). This implies that the trapping channel can be blocked and unblocked during the Auger-blinking as well as in the BC-blinking. This can explain why the neutral exciton state of QD 5 obeys a power law. The fast and slow charging here can also be well fitted using a biexponential function with two charging rates (Figure S2). For the BC-blinking with a continuous distribution of emission states, we are not able to analyze the statistics of duration, because it is not proper anymore to use a threshold line to divide on and off states. The change-point analysis20,58 and the power spectral density estimation12,59 are possible methods for BC-blinking. Since the BC-blinking always exists, it may distort the experimental duration histograms of exciton and trion states. Effect of Excitation Power on Exponential AugerBlinking. The effects of excitation power on the exponential charging−blinking for two single QDs are presented in Figure 4 and in Figures S8 and S9. The intensities of both the exciton and trion states are plotted as a function of excitation power in Figure 4a and e, and the rates of charging and discharging are plotted in Figure 4b and f. The charging rate, kc, and the discharging rate, kd, are extracted from the exponential histograms of exciton and trion state durations at different excitation powers in Figure 4c, d, g, and h. The duration

Figure 2c and f. Again the radiative lifetime scaling order is around 2. τX r τ I 24.2 218 − 2 = X X* = × ≈ 1.99 for QD3 τX * r τX * IX 4.4 600 − 2 τX r τ I 33.3 180 − 2 = X X* = × ≈ 1.97 for QD4 τX * r τX * IX 4.5 670 − 2

For the trion emission of QD 1 in Figure 1, we also have this relation: τX r τ I 23.2 50 − 2 = X X* = × ≈ 1.93 for QD1 τX * r τX * IX 3.5 167 − 2

From the above experiments, it is evident that there are two coexisting mechanisms of PL intermittency: BC-blinking and Auger-blinking. These two types of blinking can also be found in the other batch of QDs (QD@618) with a smaller core size. The PL of QDs is sensitive to the condition of the QD surface and the environment.16,26,50−56 In the Supporting Information, we compare the blinking of QDs (QD@618) under N2 and air environments. We find that, in a N2 environment, most QDs exhibit fast or noise blinking, which is typical of BC blinking. Once the sample is in contact with air, however, more QDs with neat and slow blinking are found. Representative QDs are shown in Figure S3. Two possible explanations are given in the Supporting Information. It reveals that the passivation condition determines which blinking mechanism plays the dominant role in QD blinking. Exponential and Power-Law Auger-Blinking. The blinking statistics for a single QD can be quasi-exponential or exponential. For example, the exciton−trion blinking of QD 2 is shown again in Figure 3a. The PL intensity is divided from low to high into four parts: dark, trion, mixed, and exciton states in Figure 3b. The histograms of exciton and trion state durations are described by an exponential distribution and not by a power law in Figure 3c and d. Similar exponential distributions have been reported previously.16 The exponential E


Article

ACS Nano

Figure 5. Simplified kinetics of blinking. (a) BC-blinking. When traps are deactivated, the QD is in the bright state with the zero nonradiative rate; when traps are activated, in addition to the radiative decay, the exciton can also nonradiatively decay via traps. (b) Auger-blinking (exciton−trion blinking). In the first row, the traps are deactivated and the QD is in the bright state. In the second row, the long-lived traps are activated but no electron is trapped due to the slow trapping rate. In the third row, a hole is trapped due to exciton ionization and biexciton ionization. The core is negatively charged. Because of the long-lived trapped hole, the second exciton is generated by further excitation, and then the nonradiative Auger decay competes with the radiative decay. The charging state can be neutral again via spontaneous detrapping of the trapped hole or ejecting an electron from the core to the surface.

distributions of exciton and trion states, P(t), are dependent on kc and kd as

not rely too much on the excitation power, which means that only spontaneous neutralization exists. For QD 6, however, the discharging process is assisted by the light.16,66 Under illumination, the negative trion will be formed when a second exciton is generated in a negatively charged core. If the hole of the negative trion is captured by surface traps and two electrons are left in the core, then the QD core is doubly charged. This will convert the QD from the normal gray state into a much less emissive or even dark state. One electron of the negative trion can accept the energy from the Auger recombination and overcome the barrier. Once this hot electron reaches the surface traps or recombines with the surface trapped hole, the QD core becomes neutral. Both multiple charging and neutralization are possible. However, multiple charging events can be neglected, because we have shown that the majority of PL jumps occur between bright (exciton) and gray (trion) states (see the blinking traces in Figures S8 and S9).

P(t ) ∝ exp( −kcordt ) Many reports have studied the dependence of the ionization rate on the power36,60−66 and excitation pulse repetition rate.66−68 It has been postulated that hot carriers from Auger recombination of biexcitons may contribute to the ionization events and that the excitation power dependence under low excitation conditions is quadratic.5,60−62 However, a linear dependence has also been reported, and this indicates that traps can capture carriers directly from single excitons.65,66,69 Here, we observe a linear dependence of the charging rate on the power for QD 6 (Figure 4b) and a quasi-quadratic dependence (a power law of which the exponent is 2.5) for QD 7 (Figure 4f). As the excitation power increases, the emission intensities of exciton and trion states increase linearly without saturation (Figure 4a and e). This demonstrates that the QDs are still in the low-excitation regime. In Figure 4b, the charging rate increases linearly with the excitation power, and a zero intercept is inferred. The zero intercept of the charging rate is reasonable because it is impossible to photocharge QDs without light. For QD 7, a zero intercept of the charging rate can also be obtained using a quasi-quadratic function. The discharging rate characterizes how fast a QD transitions from the trion state to the neutral state. We do not have direct evidence to show whether it is a positive or negative trion here, but we take a negative trion for example. The significant, nonzero intercept of the discharging rate in Figure 4b and f means that the trapped hole at the surface can return spontaneously to the negatively charged core in the dark, making the core neutral.66 For QD 7, the discharging rate does

CONCLUSIONS We have shown experimentally that two types of PL blinking can exist in the same single QD: the first is due to fluctuations in the nonradiative recombination rate alone, while the second type involves both radiative and nonradiative rate-jumps due to charging. The two kinds of blinking are depicted graphically in Figure 5. In BC-blinking, the nonradiative rate fluctuates because of the activation and deactivation of short-lived surface traps. Competition occurs between the fixed radiative and fluctuating nonradiative decay rates for band-edge excitons. This leads to a linear correlation between the PL lifetime and the PL intensity; this is quite distinct from HC-blinking, where the PL lifetime is constant. In Auger-blinking, both radiative and nonradiative rates change when the QD switches between F


Article

ACS Nano Present Addresses

the neutral and charged states. The Auger-blinking is also affected by opening and closing of the trapping channel, leading to conventional power-law distributed on and off durations. However, if the traps are always active, the blinking exhibits a quasi-exponential distribution. We propose further that BCblinking is due to fast trapping and then nonradiative recombination via short-lived traps, while long-lived traps are involved in the exciton−trion transition.

§

Optoelectronic Materials Section, Faculty of Applied Sciences, Delft University of Technology, Van der Maasweg 9, 2629HZ Delft, The Netherlands. ⊥ Department of Chemistry and Zukunftskolleg, University of Konstanz, 78457 Konstanz, Germany. Author Contributions

G.Y. performed all the spectroscopic work and carried out the data analysis. D.G. helped build the instrumentation and assisted with data analysis. P.M. conceived the experiments and directed the research. K.B. and N.K. designed and synthesized the quantum dots used in the experiments. All the authors contributed to the writing and editing of the final manuscript.

METHODS CdSe/CdxZn1−xS QD Synthesis. The QDs were synthesized by following our alloyed shelling method (details can be found in ref 44). Two batches of QDs were used in this work. The synthesis of the small QDs (QD@618, fluorescence peak at 618 nm) started from CdSe cores with a diameter of 4.2 nm, and these were overcoated with two shells of nominally 4 monolayers (MLs) of CdS and 2 MLs of ZnS; the synthesis of the large QDs (QD@639, fluorescence peak at 639 nm) started from 5.4 nm diameter CdSe cores with shells of nominally 4 MLs of CdS and 2 MLs of ZnS. The QYs of QD@618 and QD@639 in hexane are 91% and 70%, respectively, measured relative to a standard (Rhodamine 101). The ensemble absorption, fluorescence spectra, and TEM are available from ref 45. QD Measurement. A small drop of dilute QD hexane solution was cast on coverslips. Single NCs were distributed sparsely after spincoating. These single QDs were then overcoated with a 100 nm thick PMMA film by spin-coating. Then the coverslips were left in air for measurement. A custom-built confocal microscope based on an Olympus IX71 was used to do single-QD measurement. A 466 nm pulsed laser diode (PicoQuant, LDH-P-C-470, tunable repetition rate from 2.5 to 20 MHz) was used to excite QDs. The emission was collected through an oil-immersion objective (Olympus, PlanApo NA 1.4) and detected by avalanche photodiodes (PerkinElmer, SPCM-AQR-14). The PL trace and decay measurements were carried out with a time-correlated single photon counting card (PicoQuant, TimeHarp 200) in time-tagged time-resolved mode. The average lifetime τ is calculated based on the average photon arrival time t ̅ relative to the laser pulse. The relationship is t̅ =

1 τ

∫0

1/ f

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS P.M. thanks the ARC for support through DP130102134 and CE170100026. K.B. acknowledges support from the Humboldt Foundation through a Feodor Lynen-Fellowship (2012-2015) and from the Fonds der Chemischen Industrie (FCI) through a Liebig Fellowship (since 2015). G.Y. thanks the University of Melbourne for MIFR and MIEA scholarships. D.E.G. thanks the ARC for funding through a Future Fellowship (FT140100514). REFERENCES (1) Dai, X.; Deng, Y.; Peng, X.; Jin, Y. Quantum-Dot Light-Emitting Diodes for Large-Area Displays: Towards the Dawn of Commercialization. Adv. Mater. 2017, 29, 1607022. (2) Aharonovich, I.; Englund, D.; Toth, M. Solid-State Single-Photon Emitters. Nat. Photonics 2016, 10, 631−641. (3) Nirmal, M.; Dabbousi, B.; Bawendi, M.; Macklin, J.; Trautman, J.; Harris, T.; Brus, L. Fluorescence Intermittency in Single Cadmium Selenide Nanocrystals. Nature 1996, 383, 802−804. (4) Efros, A. L.; Nesbitt, D. J. Origin and Control of Blinking in Quantum Dots. Nat. Nanotechnol. 2016, 11, 661−671. (5) Efros, A. L.; Rosen, M. Random Telegraph Signal in the Photoluminescence Intensity of a Single Quantum Dot. Phys. Rev. Lett. 1997, 78, 1110−1113. (6) Krauss, T. D.; Brus, L. E. Charge, Polarizability, and Photoionization of Single Semiconductor Nanocrystals. Phys. Rev. Lett. 1999, 83, 4840−4843. (7) Kuno, M.; Fromm, D.; Hamann, H.; Gallagher, A.; Nesbitt, D. Nonexponential “Blinking” Kinetics of Single CdSe Quantum Dots: A Universal Power Law Behavior. J. Chem. Phys. 2000, 112, 3117−3120. (8) Gómez, D. E.; van Embden, J.; Jasieniak, J.; Smith, T. A.; Mulvaney, P. Blinking and Surface Chemistry of Single CdSe Nanocrystals. Small 2006, 2, 204−208. (9) Verberk, R.; van Oijen, A. M.; Orrit, M. Simple Model for the Power-Law Blinking of Single Semiconductor Nanocrystals. Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 66, 233202. (10) Kuno, M.; Fromm, D. P.; Johnson, S. T.; Gallagher, A.; Nesbitt, D. J. Modeling Distributed Kinetics in Isolated Semiconductor Quantum Dots. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 67, 125304. (11) Tang, J.; Marcus, R. A. Diffusion-Controlled Electron Transfer Processes and Power-Law Statistics of Fluorescence Intermittency of Nanoparticles. Phys. Rev. Lett. 2005, 95, 107401. (12) Pelton, M.; Smith, G.; Scherer, N. F.; Marcus, R. A. Evidence for a Diffusion-Controlled Mechanism for Fluorescence Blinking of Colloidal Quantum Dots. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 14249−14254. (13) Zhao, J.; Nair, G.; Fisher, B. R.; Bawendi, M. G. Challenge to the Charging Model of Semiconductor-Nanocrystal Fluorescence Inter-

⎛ x⎞ x exp⎜ − ⎟ dx ⎝ τ⎠

where f is the laser repetition rate. Therefore,

⎛ 1⎞ 1 ⎛ 1⎞ t ̅ = τ − τ exp⎜ − ⎟ + exp⎜ − ⎟ f ⎝ fτ ⎠ ⎝ fτ ⎠ As long as the repetition rate is low, we will have τ = t.̅

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b09052. BC-blinking and Auger-blinking for QD@618, environmental effects on QD blinking, details for the effect of excitation power in Figure 4, and intensity histograms for QD 1 and QD 2 (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. ORCID

Daniel E. Gómez: 0000-0002-7871-4068 Klaus Boldt: 0000-0002-0035-2490 Paul Mulvaney: 0000-0002-8007-3247 G


Article

ACS Nano mittency from Off-State Quantum Yields and Multiexciton Blinking. Phys. Rev. Lett. 2010, 104, 157403. (14) Rosen, S.; Schwartz, O.; Oron, D. Transient Fluorescence of the Off State in Blinking CdSe/CdS/ZnS Semiconductor Nanocrystals Is Not Governed by Auger Recombination. Phys. Rev. Lett. 2010, 104, 157404. (15) Jha, P. P.; Guyot-Sionnest, P. Trion Decay in Colloidal Quantum Dots. ACS Nano 2009, 3, 1011−1015. (16) Gómez, D. E.; van Embden, J.; Mulvaney, P.; Fernee, M. J.; Rubinsztein-Dunlop, H. Exciton-Trion Transitions in Single CdSeCdS Core-Shell Nanocrystals. ACS Nano 2009, 3, 2281−2287. (17) Spinicelli, P.; Buil, S.; Quélin, X.; Mahler, B.; Dubertret, B.; Hermier, J. P. Bright and Grey States in CdSe-CdS Nanocrystals Exhibiting Strongly Reduced Blinking. Phys. Rev. Lett. 2009, 102, 136801. (18) Zhang, K.; Chang, H.; Fu, A.; Alivisatos, A. P.; Yang, H. Continuous Distribution of Emission States from Single CdSe/ZnS Quantum Dots. Nano Lett. 2006, 6, 843−847. (19) Fisher, B. R.; Eisler, H.-J.; Stott, N. E.; Bawendi, M. G. Emission Intensity Dependence and Single-Exponential Behavior in Single Colloidal Quantum Dot Fluorescence Lifetimes. J. Phys. Chem. B 2004, 108, 143−148. (20) Schmidt, R.; Krasselt, C.; Göhler, C.; von Borczyskowski, C. The Fluorescence Intermittency for Quantum Dots Is Not Power-Law Distributed: A Luminescence Intensity Resolved Approach. ACS Nano 2014, 8, 3506−3521. (21) Schlegel, G.; Bohnenberger, J.; Potapova, I.; Mews, A. Fluorescence Decay Time of Single Semiconductor Nanocrystals. Phys. Rev. Lett. 2002, 88, 137401. (22) Frantsuzov, P. A.; Marcus, R. A. Explanation of Quantum Dot Blinking without the Long-Lived Trap Hypothesis. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 155321. (23) Frantsuzov, P.; Volkán-Kacsó, S.; Jankó, B. Model of Fluorescence Intermittency of Single Colloidal Semiconductor Quantum Dots Using Multiple Recombination Centers. Phys. Rev. Lett. 2009, 103, 207402. (24) Rabouw, F. T.; Kamp, M.; van Dijk-Moes, R. J.; Gamelin, D. R.; Koenderink, A. F.; Meijerink, A.; Vanmaekelbergh, D. Delayed Exciton Emission and Its Relation to Blinking in CdSe Quantum Dots. Nano Lett. 2015, 15, 7718−7725. (25) Sampat, S.; Karan, N. S.; Guo, T.; Htoon, H.; Hollingsworth, J. A.; Malko, A. V. Multistate Blinking and Scaling of Recombination Rates in Individual Silica-Coated CdSe/CdS Nanocrystals. ACS Photonics 2015, 2, 1505−1512. (26) Javaux, C.; Mahler, B.; Dubertret, B.; Shabaev, A.; Rodina, A. V.; Efros, A. L.; Yakovlev, D. R.; Liu, F.; Bayer, M.; Camps, G.; Biadala, L.; Buil, S.; Quelin, X.; Hermier, J. P. Thermal Activation of NonRadiative Auger Recombination in Charged Colloidal Nanocrystals. Nat. Nanotechnol. 2013, 8, 206−212. (27) Kambhampati, P. Hot Exciton Relaxation Dynamics in Semiconductor Quantum Dots: Radiationless Transitions on the Nanoscale. J. Phys. Chem. C 2011, 115, 22089−22109. (28) Galland, C.; Ghosh, Y.; Steinbruck, A.; Hollingsworth, J. A.; Htoon, H.; Klimov, V. I. Lifetime Blinking in Nonblinking Nanocrystal Quantum Dots. Nat. Commun. 2012, 3, 908. (29) Qin, W.; Guyot-Sionnest, P. Evidence for the Role of Holes in Blinking: Negative and Oxidized CdSe/CdS Dots. ACS Nano 2012, 6, 9125−9132. (30) Galland, C.; Ghosh, Y.; Steinbruck, A.; Sykora, M.; Hollingsworth, J. A.; Klimov, V. I.; Htoon, H. Two Types of Luminescence Blinking Revealed by Spectroelectrochemistry of Single Quantum Dots. Nature 2011, 479, 203−207. (31) Cragg, G. E.; Efros, A. L. Suppression of Auger Processes in Confined Structures. Nano Lett. 2010, 10, 313−317. (32) Chen, Y.; Vela, J.; Htoon, H.; Casson, J. L.; Werder, D. J.; Bussian, D. A.; Klimov, V. I.; Hollingsworth, J. A. Giant” Multishell CdSe Nanocrystal Quantum Dots with Suppressed Blinking. J. Am. Chem. Soc. 2008, 130, 5026−5027.

(33) Omogo, B.; Gao, F.; Bajwa, P.; Kaneko, M.; Heyes, C. D. Reducing Blinking in Small Core-Multishell Quantum Dots by Carefully Balancing Confinement Potential and Induced Lattice Strain: The ″Goldilocks″ Effect. ACS Nano 2016, 10, 4072−4082. (34) Chen, O.; Zhao, J.; Chauhan, V. P.; Cui, J.; Wong, C.; Harris, D. K.; Wei, H.; Han, H. S.; Fukumura, D.; Jain, R. K.; Bawendi, M. G. Compact High-Quality CdSe-CdS Core-Shell Nanocrystals with Narrow Emission Linewidths and Suppressed Blinking. Nat. Mater. 2013, 12, 445−451. (35) Mahler, B.; Spinicelli, P.; Buil, S.; Quelin, X.; Hermier, J.-P.; Dubertret, B. Towards Non-Blinking Colloidal Quantum Dots. Nat. Mater. 2008, 7, 659−664. (36) Qin, H.; Niu, Y.; Meng, R.; Lin, X.; Lai, R.; Fang, W.; Peng, X. Single-Dot Spectroscopy of Zinc-Blende CdSe/CdS Core/Shell Nanocrystals: Nonblinking and Correlation with Ensemble Measurements. J. Am. Chem. Soc. 2014, 136, 179−187. (37) Stefani, F. D.; Hoogenboom, J. P.; Barkai, E. Beyond Quantum Jumps: Blinking Nanoscale Light Emitters. Phys. Today 2009, 62, 34− 39. (38) Frantsuzov, P.; Kuno, M.; Janko, B.; Marcus, R. A. Universal Emission Intermittency in Quantum Dots, Nanorods and Nanowires. Nat. Phys. 2008, 4, 519−522. (39) Cordones, A. A.; Leone, S. R. Mechanisms for Charge Trapping in Single Semiconductor Nanocrystals Probed by Fluorescence Blinking. Chem. Soc. Rev. 2013, 42, 3209−3221. (40) Osad’ko, I. S.; Eremchev, I. Y.; Naumov, A. V. Two Mechanisms of Fluorescence Intermittency in Single Core/Shell Quantum Dot. J. Phys. Chem. C 2015, 119, 22646−22652. (41) Tenne, R.; Teitelboim, A.; Rukenstein, P.; Dyshel, M.; Mokari, T.; Oron, D. Studying Quantum Dot Blinking through the Addition of an Engineered Inorganic Hole Trap. ACS Nano 2013, 7, 5084−5090. (42) Park, Y.-S.; Lim, J.; Makarov, N. S.; Klimov, V. I. Effect of Interfacial Alloying versus “Volume Scaling” on Auger Recombination in Compositionally Graded Semiconductor Quantum Dots. Nano Lett. 2017, 17, 5607−5613. (43) Schwartz, O.; Oron, D. A Present Understanding of Colloidal Quantum Dot Blinking. Isr. J. Chem. 2012, 52, 992−1001. (44) Boldt, K.; Kirkwood, N.; Beane, G. A.; Mulvaney, P. Synthesis of Highly Luminescent and Photo-Stable, Graded Shell CdSe/ CdXZn1−XS Nanoparticles by In Situ Alloying. Chem. Mater. 2013, 25, 4731−4738. (45) Yuan, G.; Gómez, D.; Kirkwood, N.; Mulvaney, P. Tuning Single Quantum Dot Emission with a Micromirror. Nano Lett. 2018, 18, 1010. (46) Voznyy, O. Mobile Surface Traps in CdSe Nanocrystals with Carboxylic Acid Ligands. J. Phys. Chem. C 2011, 115, 15927−15932. (47) Houtepen, A. J.; Hens, Z.; Owen, J. S.; Infante, I. On the Origin of Surface Traps in Colloidal II−VI Semiconductor Nanocrystals. Chem. Mater. 2017, 29, 752−761. (48) Gómez-Campos, F. M.; Califano, M. Hole Surface Trapping in CdSe Nanocrystals: Dynamics, Rate Fluctuations, and Implications for Blinking. Nano Lett. 2012, 12, 4508−4517. (49) Yuan, G.; Ritchie, C.; Ritter, M.; Murphy, S.; Gómez, D. E.; Mulvaney, P. The Degradation and Blinking of Single CsPbI3 Perovskite Quantum Dots. J. Phys. Chem. C 2017, DOI: 10.1021/ acs.jpcc.7b11168. (50) Pechstedt, K.; Whittle, T.; Baumberg, J.; Melvin, T. Photoluminescence of Colloidal CdSe/ZnS Quantum Dots: The Critical Effect of Water Molecules. J. Phys. Chem. C 2010, 114, 12069−12077. (51) Müller, J.; Lupton, J. M.; Rogach, A. L.; Feldmann, J.; Talapin, D. V.; Weller, H. Air-Induced Fluorescence Bursts from Single Semiconductor Nanocrystals. Appl. Phys. Lett. 2004, 85, 381−383. (52) Ito, Y.; Matsuda, K.; Kanemitsu, Y. Photoluminescence Intermittency in Single CdSe Nanoparticles: Environment Dependence. J. Lumin. 2008, 128, 868−870. (53) Koberling, F.; Mews, A.; Basché, T. Oxygen-Induced Blinking of Single CdSe Nanocrystals. Adv. Mater. 2001, 13, 672−676. H


Article

ACS Nano (54) Oda, M.; Hasegawa, A.; Iwami, N.; Nishiura, K.; Ando, N.; Nishiyama, A.; Horiuchi, H.; Tani, T. Reversible Photobluing of CdSe/ZnS/TOPO Nanocrystals. Colloids Surf., B 2007, 56, 241−245. (55) van Sark, W. G.; Frederix, P. L.; Bol, A. A.; Gerritsen, H. C.; Meijerink, A. Blueing, Bleaching, and Blinking of Single CdSe/ZnS Quantum Dots. ChemPhysChem 2002, 3, 871−879. (56) Cordero, S.; Carson, P.; Estabrook, R.; Strouse, G.; Buratto, S. Photo-Activated Luminescence of CdSe Quantum Dot Monolayers. J. Phys. Chem. B 2000, 104, 12137−12142. (57) Gómez, D. E.; Califano, M.; Mulvaney, P. Optical Properties of Single Semiconductor Nanocrystals. Phys. Chem. Chem. Phys. 2006, 8, 4989−5011. (58) Watkins, L. P.; Yang, H. Detection of Intensity Change Points in Time-Resolved Single-Molecule Measurements. J. Phys. Chem. B 2005, 109, 617−628. (59) Frantsuzov, P. A.; Volkan-Kacso, S.; Janko, B. Universality of the Fluorescence Intermittency in Nanoscale Systems: Experiment and Theory. Nano Lett. 2013, 13, 402−408. (60) Eremchev, I. Y.; Osad’ko, I. S.; Naumov, A. V. Auger Ionization and Tunneling Neutralization of Single CdSe/ZnS Nanocrystals Revealed by Excitation Intensity Variation. J. Phys. Chem. C 2016, 120, 22004−22011. (61) Cordones, A. A.; Bixby, T. J.; Leone, S. R. Evidence for Multiple Trapping Mechanisms in Single CdSe/ZnS Quantum Dots from Fluorescence Intermittency Measurements over a Wide Range of Excitation Intensities. J. Phys. Chem. C 2011, 115, 6341−6349. (62) Peterson, J. J.; Nesbitt, D. J. Modified Power Law Behavior in Quantum Dot Blinking: A Novel Role for Biexcitons and Auger Ionization. Nano Lett. 2008, 9, 338−345. (63) Bruhn, B.; Qejvanaj, F.; Sychugov, I.; Linnros, J. Blinking Statistics and Excitation-Dependent Luminescence Yield in Si and CdSe Nanocrystals. J. Phys. Chem. C 2014, 118, 2202−2208. (64) Knappenberger, K. L.; Wong, D. B.; Romanyuk, Y. E.; Leone, S. R. Excitation Wavelength Dependence of Fluorescence Intermittency in CdSe/ZnS Core/Shell Quantum Dots. Nano Lett. 2007, 7, 3869− 3874. (65) Stefani, F. D.; Knoll, W.; Kreiter, M.; Zhong, X.; Han, M. Y. Quantification of Photoinduced and Spontaneous Quantum-Dot Luminescence Blinking. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 125304. (66) Meng, R.; Qin, H.; Niu, Y.; Fang, W.; Yang, S.; Lin, X.; Cao, H.; Ma, J.; Lin, W.; Tong, L.; Peng, X. Charging and Discharging Channels in Photoluminescence Intermittency of Single Colloidal CdSe/CdS Core/Shell Quantum Dot. J. Phys. Chem. Lett. 2016, 7, 5176−5182. (67) Schwartz, O.; Tenne, R.; Levitt, J. M.; Deutsch, Z.; Itzhakov, S.; Oron, D. Colloidal Quantum Dots as Saturable Fluorophores. ACS Nano 2012, 6, 8778−8782. (68) Saba, M.; Aresti, M.; Quochi, F.; Marceddu, M.; Loi, M. A.; Huang, J.; Talapin, D. V.; Mura, A.; Bongiovanni, G. Light-Induced Charged and Trap States in Colloidal Nanocrystals Detected by Variable Pulse Rate Photoluminescence Spectroscopy. ACS Nano 2013, 7, 229−238. (69) Banin, U.; Bruchez, M.; Alivisatos, A. P.; Ha, T.; Weiss, S.; Chemla, D. S. Evidence for a Thermal Contribution to Emission Intermittency in Single CdSe/CdS Core/Shell Nanocrystals. J. Chem. Phys. 1999, 110, 1195−1201.

I


Two Mechanisms Determine Quantum Dot Blinking - ACS Nano (ACS

Recommend Documents