Carrier-Doping by Pump-Pump-Probe Spec-troscopy in Combination

EIL and HIL stand for electron and hole injection layers, respectively; ETL and HTL stand for electron and hole transport layers, respectively. Red ar...
2 downloads 0 Views 989KB Size
Subscriber access provided by NORTH CAROLINA A&T UNIV

Spectroscopy and Photochemistry; General Theory

“Intact” Carrier-Doping by Pump-Pump-Probe Spec-troscopy in Combination with Interfacial Charge Trans-fer: A Case Study of CsPbBr3 Nanocrystals Junhui Wang, Tao Ding, Jing Leng, Shengye Jin, and Kaifeng Wu J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b01132 • Publication Date (Web): 30 May 2018 Downloaded from http://pubs.acs.org on May 30, 2018

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 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 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.

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 19 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 Letters

“Intact” Carrier-Doping by Pump-Pump-Probe Spectroscopy in Combination with Interfacial Charge Transfer: A Case Study of CsPbBr3 Nanocrystals Junhui Wang, Tao Ding, Jing Leng, Shengye Jin and Kaifeng Wu* State Key Laboratory of Molecular Reaction Dynamics and Collaborative Innovation Center of Chemistry for Energy Materials (iChEM), Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China, 116023

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 19

ABSTRACT: Carrier-doping is important for semiconductor nanocrystals (NCs) as it offers a new knob to tune their functionalities in addition to size and shape control. Also, extensive studies on NC-devices have revealed that under operating conditions NCs are often unintentionally doped with electrons or holes. Thus, it is essential to be able to controllably dope NCs and study the carrier dynamics of doped NCs. The extension of previously-reported redox-doping methods to chemically-sensitive materials, such as recently-introduced perovskite NCs, has remained challenging. Here we introduce an “intact” carrier-doping method by performing pump-pump-probe transient absorption spectroscopy on NCacceptor complexes. The first pump pulse is used to trigger charge-transfer from the NC to the acceptor, leading to NCs doped with a band edge carrier; the following pump-probe pulses measure the dynamics of carrier-doped NCs. We performed this measurement on CsPbBr3 NCs and deduced positive and negative trion lifetimes of 220±50 ps and 150±40 ps, respectively, for 10-nm-diamter NCs, both dominated by Auger recombination. It also allowed us to identify randomly-photocharged excitons in CsPbBr3 NCs as positive trions. TOC GRAPHIC

Carrier-doping has been an important subject in semiconductor science and technology. It allows for precise control of carrier density and electric conductivity in semiconductors which enables their broad applications in various electronic and optoelectronic devices. For semiconductor nanocrystals (NCs), carrier-doping is also important as it offers an additional way to tune their functionalities beyond using size and shape control.1-3 Indeed, various high-performance colloidal NC-devices,4 such as field-effect 2

ACS Paragon Plus Environment

Page 3 of 19 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 Letters

transistors, photodetectors and solar cells, have been demonstrated by using judicious carrier-doping. In addition to intentional doping, previous studies on NC-devices have indicated that NCs in operating devices are often unintentionally doped with electrons or holes.5-7 As a result, the excited states in these devices are not neutral but charged excitons (or the so-called trions). Specifically, due to the intrinsic asymmetry between electron and hole effective masses and mobilities, as well as the specific device architecture design, there usually exists an imbalance between electron and hole injection (or extraction) rates. In light-emitting diodes (Fig. 1a, left), electrons are often injected into the NCs with a higher rate than holes, resulting in negatively charged NCs for which the photogenerated excited state is a negative trion (X-).5 On the other hand, in solar cells (Fig. 1a, right), electrons are often extracted from the NCs with a higher rate than holes, thereby creating positively charged NCs whose excited state is a positive trion (X+).6 For typical NCs, the recombination of a trion is dominated by nonradiative Auger decay whereby the recombination energy of an exciton is dissipated by exciting the extra carrier.8-10 Therefore, it is essential to understand how Auger recombination competes with light emitting or conversion processes in these carrier-doped NCs and how to tweak this competition towards better device performances. Band edge carrier-doping for CdSe- or PbSe-related NCs was achieved by using chemical, photochemical or electrochemical methods,1,

8, 11-14

which has been exploited to study fundamental physics of

charged excitons, such as Auger recombination9 and photoluminescence (PL) blinking14, and to demonstrate functional NC-films with significantly improved electric conductivity3, 11. The extension of these regular doping methods to some chemically-sensitive NC materials, however, has been challenging. One of the examples is the recently-introduced, but already extensively-studied, lead halide perovskite NCs.15-20 This class of NCs can attain high photoluminescence (PL) quantum yields in the range of 50%-90% without the need of surface treatment,15-16, 21 showing strong potential as a key enabler for many optoelectronic devices.22 A recent spectro-electrochemical study on CsPbBr3 perovskites showed 3

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 19

that CsPbBr3 was decomposed before the applied bias reached the conduction or valance band edges.23 Our preliminary experiments on doping CsPbBr3 NCs using a relatively mild photochemical method12 showed no evidence of band edge doping but some peak-shifting indicative of NC etching by the doping chemistry. These results highlighted the difficulty of doping carriers into the band edge of perovskite NCs without degrading their stabilities. We note that there are a few studies on CsPbBr3 NCs where species with faster PL decay rate and lower emissivity were assigned to trions21, 24-27; however, it remains unknown whether the extra carriers are doped into band edges or mid-gap states and, if it is the former case, whether the doped carriers are electrons or holes. Here, instead of using redox chemistry, we introduce an “intact” carrier-doping method by performing pump-pump-probe transient absorption (TA) spectroscopy on NC-molecular acceptor complexes. The first pump pulse is used to trigger charge transfer from the NC to the acceptor, leading to NCs doped with a band edge carrier, and the following pump-probe pulses measure the dynamics of carrier-doped NCs (Fig. 1b). As a demonstration, we performed this experiment on CsPbBr3 NC-Rhodamine B (RhB) complexes. This allowed us to deduce a positive trion (X+) lifetime of 220±50 ps and a negative trion (X-) lifetime of 150±40 ps for the 10-nm-diamter CsPbBr3 NCs, dominated by the ultrafast Auger recombination process. The emission efficiencies of X+ and X- were calculated to be only 11.9±3.0% and 8.1±2.2% that of neutral excitons, respectively. In addition, we found that the charge transfer efficiency from X+ to the attached RhB molecules (65.4±4.2%) was considerably lower than that from neutral excitons (89.8±3.6%). These trion lifetimes also allowed us to identify that randomly-photocharged excitons in CsPbBr3 NCs are dominated by positive trions.

4

ACS Paragon Plus Environment

Page 5 of 19 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 Letters

Figure 1. a) In LEDs or solar cells, NCs are often unintentionally doped due to an imbalance between electron and hole injection (extraction) rates. EIL and HIL stand for electron and hole injection layers, respectively; ETL and HTL stand for electron and hole transport layers, respectively. Red arrows represent Auger recombination of charged excitons (X- or X+). b) Scheme of doping and measuring NCs using pump-pump-probe TA spectroscopy in combination with interfacial charge transfer. The first pump pulse triggers ET from photoexcited NCs to attached acceptors with a rate of kET,n (n is the number of acceptors per NC). After a fixed delay time of T, when ET finishes, the second pump pulse excites the hole-doped NCs and generates positive trions (X+)in the NC; in this case, the total decay rate is the sum of X+ recombination rate (kX+) and ET rate to unreduced acceptors (kET,n-1). The principle of carrier-doping and measurement in perovskite NCs using pump-pump-probe TA spectroscopy is illustrated in Fig. 1b. The first pump pulse excites the NCs and triggers interfacial electron transfer (ET) from the NC to attached electron acceptors, with a rate of kET,n (n is the number of attached 5

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 19

electron acceptors per NC). After a fixed delay of T, when ET finishes and all the photoexcited NCs are left with a hole, the second pump pulse arrives and excites these hole-doped NCs, generating a positive trion (X+) in the NC. In this case, the total decay of the system (ktot) is the sum of two competing processes-Auger dominated recombination of X+ (with a rate of kX+) and interfacial ET to the unreduced electron acceptors (with a rate of kET,n-1). If we ignore the effect of charging energy induced by the extra hole remaining the NC on electron transfer rate28 (see SI for our justifications for the CsPbBr3 NC-RhB complexes), simple statistical considerations would give: kET,n-1 = (n-1) kET,n /n.29 Hence, the trion decay rate in NCs can be deduced using: kX+ = ktot - (n-1) kET,n /n. This, however, is an over-simplified treatment; the real situation is complicated by a Poisson distribution of the number of adsorbed acceptors on NCs,30 resulting in heterogeneous electron transfer rates. Nonetheless, as we show later, by attaching many acceptors per NC (i.e., using a large value), kET,n-1 ≈ kET,n, and hence, kX+ ≈ ktot - kET,n. CsPbBr3 NCs were synthesized according to a literature method;16 details are provided in the Supporting Information (SI). These NCs exhibit typical cube-like morphologies (Fig. 2a),16 with an average edge length of 10.1±1.0 nm (see a size-distribution histogram in Fig. S1). Because the size of these NCs is slightly larger than the estimated bulk Bohr exciton diameter (~7.4 nm),16 they are weakly-confined. As a result, their absorption edge is at ~520 nm (Fig. 2b), similar to bulk CsPbBr3. RhB molecules were attached to CsPbBr3 NCs using a simple sonication procedure; see SI for details. The absorption spectrum of CsPbBr3 NC-RhB complexes contains features from both CsPbBr3 in the 100 ps) to 1. As shown in Fig. 2d, with increasing pump fluence, fast decay components appear on the tens of picoseconds timescale, a typical signature of multiexciton Auger recombination.10 By fitting the saturation behavior of the long-lived tails at ~ 200 ps as a function of pump fluence, we can calculate the average exciton number at each pump fluence (Fig. S4). The biexciton recombination kinetics was then obtained using a wellestablished subtracting procedure. Fig. 2d inset shows the result for a subtraction between kinetic traces with = 0.57 and = 0.19, for which the contribution from higher-order multiexcitons beyond biexciton is negligible. A single-exponential fit to the subtracted kinetics yields a biexciton lifetime (τXX) of 46.0±3.6 ps, consistent with previous reports on CsPbBr3 NCs of similar volume.26 Fig. 3a shows the TA spectra of CsPbBr3 NC-RhB complexes measured with regular pp-TA using a 400 pump pulse that selectively excites the NC part in the complex. The pump fluence was the same as for Fig. 2c ( ~ 0.13). The XB feature of NCs recovers much faster than that of free NCs, as shown by the kinetics comparison in Fig. 3b. In addition, the decay of XB is accompanied by the formation of a bleach feature centered at ~560 nm that can be assigned to the ground state bleach (GSB) of RhB (Fig. 8

ACS Paragon Plus Environment

Page 9 of 19 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 Letters

3a inset). The gradual instead of instantaneous formation of the GSB feature confirms that RhB molecules are not excited by the 400 pump pulse. Although the energy level alignment between CsPbBr3 NCs33 and RhB29 should in principle allow electron, hole, and energy transfer from photoexcited CsPbBr3 NCs to ground state RhB, a careful spectroscopic analysis of CsPbBr3 NC-RhB complexes using PL and TA measurements indicates that the interaction mechanism is dominated by interfacial ET from photoexcited NCs to attached RhBs (See Fig. S6 for details), which is also consistent with a previous report on CsPbBr3 NC-RhB complexes31. Based on the signal amplitude of the XB before ET and the maximum signal amplitude of the GSB after ET, we estimate that the average number of RhBs attached per NC is as many as 52 ( = 52); see SI for details. Fig. 3b also shows that after the fast ET process (within ~600 ps), the slowly-decaying components of the XB and RhB agree well with each other, which correspond to the regeneration of ground state CsPbBr3 NC-RhB complexes through charge recombination. Simultaneous fitting of the XB and RhB kinetics reveals a two exponential ET process with time constants (and relative amplitudes) of 74.3±3.2 ps (47.9%) and 472.6±5.3 ps (34.4%), and a charge recombination process with time constants >8 ns (17.7%); see SI for details of the fitting model. The none-single-exponential charge transfer kinetics likely results from the distribution in the number of acceptors per NC and from various possible binding geometries of the acceptors on the NC surfaces.34 The charge transfer efficiency is calculated to be 89.8% by comparing XB kinetics of NCs with and without RhB attached.

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 19

Figure 3. a) TA spectra of CsPbBr3 NC-RhB complexes probed at indicated time delays following 400 nm ( ~ 0.13). Inset is the zoom-in of RhB bleach signals. b) TA kinetics probed at the XB of NC (dark red squares) and at the bleach of RhB (light red circles). They are rescaled to match their slowlydecaying components. Black solid lines are fits to these kinetics. The XB kinetics of pure NCs (green dashed line) is also plotted for comparison. c) XB kinetics in NC-RhB complexes measured using pump-probe (light red) and pump-pump-probe (dark red) TAs. They are rescaled to match their slowlydecaying components; inset shows the original kinetic traces. d) TA kinetics that contains contribution from both trion recombination and interfacial ET (light red squares), as illustrated in the inset, obtained by performing a subtraction between the kinetic traces in (c). The black solid line is a fit to the kinetics. 10

ACS Paragon Plus Environment

Page 11 of 19 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 Letters

With the charge separation and recombination dynamics in CsPbBr3 NC-RhB complexes measured, we then use pump-pump-probe TA (ppp-TA) to study recombination and charge transfer of trions in CsPbBr3 NCs by adding another 400 nm pump pulse to the regular pp-TA setup. The delay between the two pump pulses is fixed at T = 600 ps to ensure that almost all the photoexcited complexes have turned into charge separated states (NC+-RhB-); see SI for experimental details. We would like to note that the fist pump pulse is not chopped and the second is chopped in our experiments; therefore, the ppp-TA only reflects spectral changes induced by the second pulse while the first one is used to create charge separated states that correspond to the role of ground states in regular pp-TA. The fluences of both pump pulses are 19.6 µJ/cm2, corresponding to ~ 0.13. This low fluence was chosen because it allowed for the exclusion of biexciton (XX) contribution to the positive trion (X+) kinetics; see Fig. S7 for details. The TA spectra and kinetics of CsPbBr3 NC-RhB complexes measured using ppp-TA are presented in Fig. S5 and they are very similar to the data in Fig. 3a and 3b. A careful comparison of the XB kinetics measured using pp-TA and ppp-TA reveals faster decay in the latter, which can be better visualized by normalizing them to the long-lived tail (Fig. 3c). Note that the original pp-TA and ppp-TA kinetics have the same initial signal amplitude (Fig. 3c inset) because, as we emphasized above, only the second, chopped pulse contributes to the TA signal in ppp-TA measurement. Since both pump pulses create ~ 0.13 in the ppp-TA measurement, there exists a subset of NC-RhB complexes that are unexcited by the first pulse but excited by the second one. In this case, the ppp-TA is measuring the complexes excited by only one pulse, which is essentially the same as regular pp-TA. Thus, by normalizing ppp-TA and pp-TA kinetics to their slowly-decaying component (as in Fig. 3c) and performing a subtraction between them, we can exclude the contribution from pp-TA kinetics and obtain the decay kinetics for the photoexcited NC+-RhB- species. The latter is plotted in Fig. 3d and it contains contribution from both X+ recombination and interfacial ET to unreduced RhBs (Fig. 3d inset), as discussed earlier. Because the average number of RhBs attached per NC is ~52, it satisfies the condition for: kET,n-1 ≈ kET,n. Hence, when 11

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 19

fitting the kinetics in Fig. 3d, we fixed the rate constants and relative amplitudes for the exponential components used to fit electron transfer kinetics in pp-TA and simply added a trion recombination constant (kX+) to the above constants; see SI for details of the fitting model. This highly-constrained model yields a very good fit the kinetics by using 0.0046±0.0012 ps-1 for kX+. The other set of ppp-TA experiment performed using a higher fluence of 28 µJ/cm2 for both pump pulses, corresponding to ~ 0.20, gives almost the same result (Fig. S6). Thus, by using ppp-TA in combination with interfacial ET experiments, we successfully doped CsPbBr3 NCs with an extra hole and determined the positive trion lifetime (τX+) in 10-nm-diamter NCs to be 220±50 ps, which is dominated by Auger recombination as the trion radiative life in these NCs is ~1.85 ns (τX/2) according to statistical scaling of radiative rates35. Because of this ultrafast Auger recombination, the charge transfer efficiency from positively charged NCs is lowered to 65.4±4.2%, as compared to 89.8±3.6% for neutral NCs (See SI for details of the calculation). This model study highlights the difficulty of charge extraction from perovskite NCs if they are unintentionally charged in solar cells. Using the measured positive trion lifetime (τX+) and biexciton lifetime (τXX), the negative trion lifetime (τX-) is estimated to be 150±40 ps, by applying a well-established superposition principle for Auger recombination; see SI for details of the calculation. The radiative emission efficiencies of the positive and negative trions, relative to neutral excitons, are estimated to be 11.9±3.0% and 8.1±2.2%, respectively (SI). Therefore, the presence of an extra charge would significantly impair the efficiency of perovskite NC based light-emitting devices. The conducted studies also allow us to determine the nature of doping in the randomly-photocharged species reported in previous studies. The carrier dynamics in these species can be extracted by comparing the TA kinetics of a NC solution measured under static and vigorously-stirred conditions; in the former case, photocharged species accumulate in the excitation volume and hence can be probed by TA 12

ACS Paragon Plus Environment

Page 13 of 19 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 Letters

measurements, whereas in the latter case the excitation volume is continuously replenished with fresh NCs. As shown in Fig. 4a, after rescaling the XB kinetics of CsPbBr3 NCs measured under static (dark red) and vigorously-stirred conditions to match their slowly-decaying components, the former exhibit a much faster decay in the