Delayed Exciton Formation Involving Energetically Shallow Trap

Subscriber access provided by READING UNIV

Article

Delayed Exciton Formation Involving Energetically Shallow Trap States in Colloidal CsPbBr Quantum Dots 3

Yi Wang, Min Zhi, and Yinthai Chan J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b09040 • Publication Date (Web): 05 Dec 2017 Downloaded from http://pubs.acs.org on December 5, 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.

The Journal of Physical Chemistry C 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 28 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

The Journal of Physical Chemistry

Delayed Exciton Formation Involving Energetically Shallow Trap States in Colloidal CsPbBr3 Quantum Dots Yi Wang,† Min Zhi,† and Yinthai Chan*,†,‡ †Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore ‡Institute of Materials Research & Engineering, A* STAR, 2 Fusionopolis Way, Innovis # 0803, Singapore 138634, Singapore

Corresponding Author *E-mail: [email protected]

ABSTRACT: We report the occurrence of delayed exciton formation in highly emissive CsPbBr3 quantum dots, which results in anomalously slow build-up kinetics that are readily observed in a time-resolved photoluminescence trace. It is inferred from the dependence of the build-up kinetics on nanoparticle size, temperature and excitation fluence that the delayed exciton

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 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 28

formation originates from multiple carrier trapping and de-trapping events between the band edge state and energetically shallow structural disorder states. A kinetic model that incorporates these carrier pathways produces fits that are in excellent agreement with the time-resolved data. Importantly, the kinetic model allows for the determination of photoluminescence quantum yield values that closely match those obtained from integrating sphere measurements, whereas merely accounting for the pathways associated with the decay kinetics produces large discrepancies. This work highlights the crucial role played by delayed exciton formation in the photoluminescence dynamics of CsPbBr3 QDs.


2

Page 3 of 28 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


INTRODUCTION

All-inorganic perovskite (namely CsPbX3, where X = Cl, Br, I) based quantum dots (QDs) are promising as chromophores because of their spectrally narrow emission profiles (~100 meV) relative to conventional organic dyes and fluorescence wavelengths that can readily be tuned across the entire visible range (i.e. ∼400 to 700 nm) by varying their size and composition, thus providing the spectral bandwidth required by most applications. What is perhaps most striking is the fact that the as-synthesized QDs exhibit room temperature photoluminescence quantum yields (PLQYs) as high as 90% without the need for surface passivation by an epitaxial wide bandgap shell layer that is generally imperative for highly emissive II-VI, III-V or IV-VI semiconductor QDs.1 In addition to the narrow emission width, these perovskite QDs possess a sharp absorption tail, which minimizes the occurrence of interparticle energy transfer and photon re-absorption within an ensemble.2 Furthermore, it has recently been reported that the CsPbX3 nanoparticles exhibit negligible spectral diffusion and minimal inhomogeneous broadening,3 facilitating high color purity in displays. Owing to these highly desirable optical properties, a rapidly growing number of efforts to develop these QDs have been made, such as optimizing their synthetic preparation via droplet microfluidics,4 tuning their optical properties via anion exchange reactions5 and investigating the surface chemistry of the particles.6-8 To-date, the incorporation of CsPbX3 QDs in light emitting diodes with high luminance and good color saturation have been reported,9-12 alongside their use as the active material in low-threshold optically pumped lasers.13-17 It has been suggested that the high tolerance to defects and energetically shallow trap states are responsible for the outstanding PLQY of CsPbX3 QDs.18-20 However, it is also well-known


3


Page 4 of 28

that the presence of trap states, notwithstanding their energy levels, is generally detrimental to PL in perovskite nanocrystals.21,22 If a significant density of shallow trap states is indeed present, it is intriguing how and why the PLQY of colloidal CsPbX3 QDs remains high. Indeed, much of the kinetics of exciton photoluminescence in these QDs is at the present time not well understood. For example, time resolved PL (TRPL) measurements, which reflect the carrier recombination dynamics of an emissive material following optical excitation by a very short pulse of light, typically feature an extremely fast build-up of intensity followed by much slower decay. In the case of CsPbX3 QDs, however, TRPL experiments reveal an abnormally slow build-up process that was previously observed but not investigated in detail.23-25

Figure 1. A schematic illustration of TRPL decay profiles with (blue line) or without (red line) a slow build-up component, which may be due to: (1) radiative exciton recombination (red arrow); (2) charge diffusion followed by radiative exciton recombination (middle blue arrow); (3) Carrier de-trapping followed by radiative exciton recombination (right blue arrow).


4

Page 5 of 28 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


Intuitively, the slow build-up in the TRPL profile of a semiconductor can be ascribed to the delayed formation of excitons. For TRPL experiments in which a dilute solution of colloidal CsPbX3 QDs is used, the delay in exciton formation may be due to carrier diffusion and/or trapping and de-trapping events. Figure 1 provides a cartoon illustration of the TRPL profiles expected for a colloidal QD in the presence and absence of delayed exciton generation. Such markedly slow build-up kinetics (on the order of 100 ns) have previously been observed in organolead halide perovskite nanowires and nanoplates and was attributed to prolonged carrier diffusion prior to exciton formation.26 For CsPbX3 QDs, small exciton binding energies (e.g. ∼40 meV for a ∼7 nm CsPbBr3 QD1) are expected to lead to significant charge separation at room temperature. This can lead to a diffusion limited formation of excitons, albeit at much faster time scales due to the small particle volumes of the QDs.27 Another possible reason for delayed exciton formation is the occurrence of multiple carrier trapping and de-trapping events, as has been widely observed in bulk perovskites28,29 and hybrid organic-inorganic perovskite nanocrystals.30 In this work, we investigate and seek to clarify the primary mechanism for delayed exciton generation in colloidal CsPbBr3 QDs, and argue that it is likely due to multiple carrier trapping and de-trapping events from energetically shallow trap states. Our experimental findings suggest that the trap states arise from structural disorder rather than surface defects. Importantly, we demonstrate that taking the build-up time into account is critical for accurately determining the PLQY of these QDs based on TRPL measurements, which is generally not necessary for prototypical semiconductor QDs such as CdSe nanocrystals.


5


Page 6 of 28

EXPERIMENTAL SECTION Synthesis of CsPbBr3 QDs. The synthesis of CsPbBr3 quantum dots in this work is based on a previously reported procedure1 with slight modifications. Typically, in a three-neck round bottom flask, 136 mg of lead bromide, 2 mL of oleic acid, 2 mL of oleylamine and 10 mL of 1octadecene (ODE) were degassed under vacuum at 80 oC for 1 h. The temperature was then increased to 185 oC under N2, whereupon 0.78 mL of Cs-oleate (0.13 M in ODE) was swiftly injected into the mixture. This initiated the nucleation and growth of the perovskite nanocrystals. After 5 min, the reaction mixture was cooled down to room temperature and 5 mL of hexane was added to the reaction flask. This yields a bimodal distribution of particle sizes. To recover the larger of the two sizes, 2 mL of reaction solution was centrifuged for 5 min at 5000 rpm, and the precipitate (13.0 nm edge length QDs) was dispersed in toluene. The smaller of the two sizes (9.2 nm edge length QDs) in the supernatant was recovered by precipitating them using 1 mL of 2-propanol and centrifuging for 5 min at 5000 rpm. The recovered QDs were then dispersed in toluene for further use. The CsPbBr3 nanocrystals with the smallest size (5.7 nm edge length QDs) were obtained by following the same procedure as above with the exception that 0.39 mL of Csoleate was injected at 165 oC and letting the reaction continue for ~10 sec before cooling down to room temperature. 5 mL of hexane was added to the reaction flask and the QDs were precipitated by adding 1 mL of 2-propanol. The mixture was then centrifuged for 5 min at 5000 rpm and the supernatant discarded. The processed QDs were then dispersed in toluene for further use. Transmission electron microscopy and X-ray powder diffraction. Transmission electron microscopy (TEM) images of CsPbBr3 nanocrystals were obtained using a JEOL JEM-1220F microscope with a 100 kV accelerating voltage. For TEM sample preparation, a dilute drop of CsPbBr3 QDs was placed onto a 300 mesh copper grid. Excess solution was removed by an


6

Page 7 of 28 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


adsorbent paper and the sample was dried under ambient conditions. X-ray diffraction (XRD) data were obtained with a diffractometer (Bruker AXS, GADDS) using Cu-K α radiation (λ = 1.540598Å) in the range of 10˚ to 50˚. Thin film samples were prepared by drop-casting 20 µL of a concentrated CsPbBr3 QD solution onto a clean silicon wafer and drying in a glovebox under N2. This was repeated several times until a thin film of CsPbBr3 QDs was formed on the silicon substrate. Steady-state spectroscopy. The steady-state UV-Vis absorption spectra for colloidal CsPbBr3 QDs were recorded on an Agilent-8453 (Agilent Technologies) absorption spectrometer using a quartz cuvette with a path length of 1 cm. The same sample and cuvette were also used to measure the steady-state photoluminescence (PL) spectra and PL quantum yield by a Fluorolog-3 spectrofluorometer (Horiba) equipped with a Quanta-ϕ integrating sphere (Horiba). To minimize the effect of reabsorption on emission spectra measurements, the nanoparticle solution was diluted to ensure the absorbance at the excitation wavelength (~405 nm) was less than 5×10−2. Time-resolved photoluminescence measurement. The colloidal CsPbBr3 solution in a 2 mm-thickness quartz cuvette was excited by 100 fs pulses at 400 nm, which is the second harmonic of the 800 nm fundamental output from a 1 kHz repetition rate regenerative Ti:Sapphire amplifier (Coherent, Libra-F-1K-HE-230). The excitation beam was focused by a cylindrical lens (focal length = 20 cm) into a stripe with a size of ~0.8 × 0.04 cm2 and the transient PL signal was recorded by a time-correlated single photon counting (TCSPC) module (PicoQuant, PicoHarp 300). The same excitation source was also used to measure the fluencedependent PL spectra, and the PL signals were recorded by a triple grating imaging spectrometer (Princeton Instruments, Acton SpectraPro SP-2500). The excitation fluence was modulated by a series of neutral-density filters.


7


Page 8 of 28

RESULTS AND DISCUSSION

Figure 2. (a) TEM image of typical CsPbBr3 QDs. The inset is a photograph of a toluene solution of QDs under UV lamp excitation (λ=365 nm). (b) The size histogram counted from the TEM image. (c) XRD spectrum of CsPbBr3 QDs (orange) and the reference standard for cubic CsPbBr3 (green). (d) The UV-vis absorption (solid black line) and PL emission (dashed red line) spectra of CsPbBr3 QDs in toluene. (e) Semi-log plot of band edge absorption spectrum. (f) Tauc plot of CsPbBr3 QDs. (g) Representative TRPL profile in CsPbBr3 QDs. Inset: build-up profile at short delay times. The orange, blue and green lines are the experimental data, the fit to the decay component and the instrument response function (IRF) respectively.


8

Page 9 of 28 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


To investigate the origins of the slow build-up time in the TRPL profile, we first synthesized colloidal CsPbBr3 QDs based on a previously reported method1 (see details in the Experimental Section). The CsPbBr3 QDs were then dispersed in toluene for further characterization. Analysis by TEM, as presented in Figure 2a, revealed cube-like nanoparticles with an average edge length of about 13.0 ± 1.1 nm (Figure 2b). This length is approximately 2 times of the Bohr exciton diameter (∼7 nm), which suggests that quantum confinement effects should be relatively weak. Structural characterization by XRD, as illustrated in Figure 2c, shows that the synthesized nanocubes are cubic which is consistent with previous reports on CsPbBr3 QDs. Figure 2d shows the UV-vis absorption and PL emission spectra of the 13.0 nm edge length CsPbBr3 QDs dispersed in toluene. The sample features a small Stokes shift (∼20 meV, see Figure S1 of Supporting Information, SI), as is characteristic of most direct band-gap semiconductors; the excitonic absorption peak is nearly featureless, which is indicative of weak quantum confinement. Concomitantly, the PL peak is symmetric with a relatively narrow full width at half-maximum (FWHM) of ∼80 meV, which is typical of large (>10 nm edge length) CsPbBr3 particles. A semi-log plot of the absorption spectra, as illustrated in Figure 2e, shows an exponential relation between absorbance and photon energy below the band-gap, which can be assigned to an Urbach tail.31 Analysis of the semi-log plots yielded Urbach energy (EUr) and band-gap energy (Eg) values of 17.4 meV and 2.40 eV respectively. The value of Eg could be further verified by the Tauc plot shown in Figure 2f. The small value of EUr, which is lower than the thermal energy kBTR of ∼25.7 meV at room temperature, suggests a small degree of structural


9


Page 10 of 28

disorder and a narrow distribution of shallow intra-gap states near the band edge.32,33 The room temperature TRPL profile of the CsPbBr3 QDs is given in Figure 2g, where the PL decay kinetics are fitted by a multi-exponential function yielding an average decay lifetime () of ~3.3 ns. The multi-exponential decay kinetics in CsPbBr3 QDs has commonly been observed,1,34,35 and has been suggested to arise from imperfect surface passivation.36 Zooming into the first 500 ps of the TRPL trace (inset of Figure 2g), a steep rise within the first 100 ps followed by a gentler one that extends to 500 ps may be observed. The FWHM of the instrument response function (IRF) is ~50 ps, therefore the more gradual rise component, which we will refer to as the slow build-up, is within the temporal resolution of the detector and cannot be dismissed as a measurement artefact.


10

Page 11 of 28 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


Figure 3. (a) TRPL measurements of a sample of 13.0 nm edge length CsPbBr3 QDs at 180 K and 280 K. Inset shows the TRPL data for a longer time window. (b) The TRPL trace recorded at different detection wavelengths in 13.0 nm edge length CsPbBr3 QDs. (c) The TRPL kinetics of CsPbBr3 QDs with average edge lengths of 13.0 nm, 9.2 nm and 5.7 nm. (d) The TRPL kinetics for a 13.0 nm edge length CsPbBr3 QD sample at an average exciton occupancy of 0.02, 0.51 and 0.72. Note that for all the figures, the open circles/triangles are experimental data while the dashed curves are a guide to the eye.


11


Page 12 of 28

Given that the slow build-up time is premised on the facile thermal dissociation of excitons under ambient conditions, low temperature should disfavor exciton dissociation, thereby resulting in a faster build-up time. We investigated the temperature dependence of the TRPL build-up profiles at two distinct temperatures as depicted in Figure 3a. To avoid freezing of the solvent at low temperature, we employed a thin random-packed film of particles for the TRPL measurements. We found only slight differences in the TRPL kinetics between the thin film and a solution of particles at room temperature (see Figure S2), which gave confidence to our thin film measurements. As seen from Figure 3a, the build-up time at lower temperature is substantially reduced, consistent with the notion of a suppressed thermal dissociation of excitons. It is also observed that the decay component is faster (inset of Figure 3a), which cannot be attributed to the opening of a fast non-radiative channel since the quantum yield of the film did not appear to decrease at lower temperature. A more plausible explanation is the reduced dissociation of excitons and a concurrent increase in the radiative rate at low temperature.37 However, while the reduced build-up time at low temperature can be explained by a lowered rate of thermally-induced exciton dissociation, why the build-up time is anomalously slow at room temperature remains unanswered. Although the build-up time in semiconductor nanocrystals is often attributed to hot-exciton relaxation and at times interparticle energy transfer, neither of them can be used to explicate the slow build-up observed in Figure 2g. The time constant of hot-exciton relaxation in CsPbBr3 QD (typically less than 5 ps38,39) is much shorter than that of the slow build-up (>100 ps); the effect of interparticle energy transfer (e.g. from smaller to larger particles) should be negligible in a dilute solution of particles, which can be verified by the fact that their PL lifetimes are independent of the detection wavelength. This is seen in the TRPL spectrum in Figure 3b, where


12

Page 13 of 28 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


the decay lifetimes detected at 500 nm, 517 nm and 530 nm are 3.5 ± 0.2 ns, 3.6 ± 0.1 ns and 3.8 ± 0.1 ns respectively. In view of their large exciton dissociation rates and long carrier diffusion lengths,27 one possible explanation for the slow build-up time in CsPbBr3 QDs is delayed exciton formation by prolonged carrier diffusion. This supposes that the exciton is formed after carrier photogeneration and diffusion within the nanoparticle. However, although a slow build-up TRPL profile induced by carrier diffusion has been observed in micron-sized organolead halide perovskite nanowires and nanoplates,26 the time scale associated with carrier diffusion in our CsPbBr3 QDs is estimated to be within the IRF of our TRPL setup given their small size (~13.0 nm edge length) and high carrier mobility (∼4500 cm2 V−1 s−1)27. Subsequently, carrier diffusion cannot be the basis for the observed slow build-up. It is also possible that delayed exciton formation results from the transition between free carriers and excitons. In such a scenario, different particle sizes spanning various levels of quantum confinement should yield dissimilar build-up profiles since the energy difference between the band edge state (where free carriers reside) and the first excitonic state, i.e. the exciton binding energy, will be distinct for each particle size. To elucidate the effect of particle size on the build-up time, CsPbBr3 QDs with different average edge lengths of 13.0 ± 1.1 nm, 9.2 ± 0.6 nm and 5.7 ± 0.7 nm were synthesized (see Figure S3). Figure 3c shows the TRPL kinetics for the as-synthesized QDs, where the average exciton number per particle () is kept the same across the different samples by tuning the pump fluence. Details on the determination of are given in the SI (Figure S4). It is interesting to see from Figure 3c that the build-up time remains comparable across all 3 samples. While the PLQY of all three samples may be different (its effects on build-up time are discussed later), the data strongly indicates that the rate


13


Page 14 of 28

determining process in the slow build-up cannot be attributed solely to the delayed exciton formation from free carriers. We also investigated the interplay between the slow build-up of PL and multiexcitonic Auger recombination, which is an extremely fast non-radiative recombination pathway that can dominate the early TRPL profile at sufficiently high pump fluence. We used 13.0 nm edge length QD samples and optically excited them at 400 nm by 100 fs pulses from a regeneratively amplified Ti:Sapph with a repetition rate of 1 kHz. As seen from Figure 3d, the slow build-up becomes less pronounced with the increase of pump fluence, with varying from 0.02 to 0.72. This is surprising, given that the 13.0 nm edge length CsPbBr3 QD is expected to exhibit weak quantum confinement and its TRPL profile should not be heavily influenced by Auger recombination effects. We find, however, that this is not the case. At a pump fluence corresponding to = 0.72, the proportion of biexcitons in the 13 nm edge length QDs is estimated to be ~ 12%. Given a fast Auger dominated biexciton lifetime of ~340 ps as determined from a known subtraction procedure,34 it is conceivable that this relatively low proportion of biexcitons (as well as higher order multiexcitons) is sufficient to mask the slow build-up profile. It is therefore important to employ pump fluences that correspond to

Delayed Exciton Formation Involving Energetically Shallow Trap

Recommend Documents