Boosting Biexciton Collection Efficiency at Quantum Dot–Oxide

Letter pubs.acs.org/JPCL

Boosting Biexciton Collection Efficiency at Quantum Dot−Oxide Interfaces by Hole Localization at the Quantum Dot Shell Hai I. Wang,†,‡,§ Mischa Bonn,† and Enrique Cánovas*,† †

Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Graduate School of Material Science in Mainz, University of Mainz, Staudingerweg 9, 55128 Mainz, Germany

‡

S Supporting Information *

ABSTRACT: Harvesting multiexcitons from semiconductor quantum dots (QDs) has been proposed as a path toward photovoltaic efficiencies beyond the Shockley−Queisser limit. Although multiexciton generation efficiencies have been quantified extensively in QD structures, the challenge of actually collecting multiple excitons at electrodesa prerequisite for high-efficiency solar cell deviceshas received less attention. Here, we demonstrate that multiexciton collection (MEC) at the PbS QD/mesoporous SnO2 interface can be boosted 5fold from ∼15 to reach ∼80% quantum yield, by partial localization of holes in a QD molecular capping shell. The resulting weakened Coulombic interactions give rise to reduced Auger recombination rates within the molecularly capped QDs, so that biexciton Auger relaxation, competing with MEC, is strongly suppressed. These results not only highlight the importance of surface chemistry and energetics at QD/ligand interfaces for multiexciton extraction but also provide clear design principles for realizing the benefits of MEG in sensitized systems exploited in solar cells and fuel geometries.

I

necessary requirement for the use of MEG in efficient solar cellshas been much less studied.13 In this work, we interrogate biexciton dissociation dynamics (transfer rates and MEC efficiency) in a QD-sensitized system where the biexciton regime is reached in a controlled manner through two sequential single-photon absorption events. Our data reveal that the collection efficiency of multiexcitons at the QD−oxide interface is dictated by kinetic competition between the electron transfer (ET) rate toward the oxide electrode (KET) and the Auger recombination rate (KAug) within the QDs. We demonstrate that MEC at the QD−oxide interface can reach ∼80% quantum yield by achieving partial localization of holes in a QD molecular capping shell. The partial localization of holes in the QD capping shell reduces the Auger recombination rate within the QDs, a consequence of reduced Coulombic interactions within the QDs. PbS QDs sensitizing SnO2 mesoporous matrixes were prepared using the successive ionic layer adsorption and reaction (SILAR) method, in which the QDs are nucleated in situ onto the SnO2 electrode. The PbS QD samples used in this study were fabricated in a glovebox under N2 conditions with the recipe of 3.5 SILAR cycles. The half-cycle Pb-terminated SILAR growth recipe provides a good atomic passivation scheme for the QDs, as reported previously.16,17 As illustrated in Figure S1 (see the SI), the HRTEM histogram reveals that the samples are characterized by a relatively narrow distribution of QD sizes (10−18% dispersion) with an averaged base

n a photovoltaic device based on a single absorber, the efficiency for solar energy conversion under 1 sun illumination is thermodynamically limited to ∼33%, the socalled Shockley−Queisser (SQ) limit.1 This upper limit is determined to a large extent by losses associated with thermalization of charges generated by photons well exceeding the material’s absorption onset. As such, efficiencies beyond the SQ limit can, in principle, be achieved in solar cell designs where thermal losses in the absorber are circumvented, for example, in hot carrier solar cells2 or in solar cells exploiting multiexciton generation3 (MEG, also known as “carrier multiplication”). MEG is a phenomenon in which absorption of one photon with an energy exceeding twice the semiconductor’s bandgap (hν/Eg > 2) is capable of producing two or more excitons4−8 by impact ionization.9,10 The theoretical limit of photoconversion efficiency by exploiting MEG at 1 sun is ∼44%.4,11 As has been shown for complete solar cell devices,12 sensitization of oxide electrodes by quantum dots (QDs) is an appealing system for efficient collection of multiple excitons generated in the QD by impact ionization. The intimate contact between the QD and oxide in a sensitized system enables strong donor−acceptor coupling, thereby potentially allowing for ultrafast extraction of multiple electrons before exciton− exciton annihilation takes place within the QDs (e.g., via Auger recombination13−15). Although a vast number of experimental reports exist on MEG efficiency and dynamics as a function of photon energy and fluence in semiconducting QDs,4,6−8 the subsequent transfer of generated multiple excitons (multiple exciton collection, MEC) to an external electrodea second © XXXX American Chemical Society

Received: April 20, 2017 Accepted: May 30, 2017 Published: May 30, 2017 2654

DOI: 10.1021/acs.jpclett.7b00966 J. Phys. Chem. Lett. 2017, 8, 2654−2658

Letter

The Journal of Physical Chemistry Letters

Figure 1. ET dynamics for PbS QDs sensitizing a SnO2 mesoporous film under hν/Eg < 2 photon excitation. (A) Time-resolved photoconductivity data normalized to values at 1 ns for several 800 nm pump excitation fluences; the black line represents the best fit to single-exciton N1X oxide(t) dynamics, as discussed in the text; the inset shows single-exciton (1X) early time dynamics. (B) Inferred biexciton N2X oxide (t) dynamics; the inset shows dynamics normalized to the peak signal; the black line represents the model calculation described in the text.

Under these conditions, the fluence is sufficiently high that a substantial fraction of QDs is excited twice by the same laser pulse, and the photoconductivity dynamics accordingly reflect a mixture of ET processes taking place from QDs hosting single nX excitons (N1X QD) and multiexcitons (denoted here as NQD, n = 2, 23 3, ...), respectively. The multiexciton contribution to the photoconductivity dynamics can be obtained by subtracting the 1X appropriately scaled low-fluence,23,24 single exciton Noxide response from the measured high-fluence OPTP data, as shown in Figure 1B. The normalized, multiexciton-derived photoconductivity dynamics are indistinguishable within the analyzed range of photon fluences, as shown in the inset of Figure 1B (and Figure S3 in the SI). This similarity suggests that the inferred multiexciton contribution to the photoconductivity dynamics shown in Figure 1B stems only from the biexciton (2X) response (N2X oxide), and higher-order effects are negligible. This is confirmed by modeling the data using Poisson statistics, as discussed below. Applying the same model as for single-exciton dynamics, we find τ2X ET = 0.45 ± 0.15 ps and τ2X = 100 ± 10 ps for 2X dynamics (with β constrained to the BET same value as for 1X dynamics, 0.6 ± 0.1). Thus, both interfacial ET and BET processes are sped up substantially for atomically passivated PbS QDs populated with biexcitons when compared to those populated by single excitons. The inferred rise in OPTP signal in the nonlinear regime represents an average response for the expected sequential transfer of two electrons from QDs populated with biexcitons and trions, respectively.13 The yield of charge collection from QDs populated with trions is expected to be more effective than that from QDs populated by 2X considering that Auger lifetimes are larger for QDs populated with trions when compared with QDs populated with biexcitons.25 A discussion on the possible origin for the enhanced ET and BET rates in the 2X regime is provided in the SI. While the dynamics discussed above clearly reflect interfacial ET processes in the multiexciton regime, the key aspect is to quantify the efficiency of MEC at the QD−oxide electrode as this is the relevant parameter for device performance. The number of electrons populating the oxide electrode as a 2X function of photon flux for the N1X oxide and Noxide contributions can be directly obtained from the maximum OPTP amplitudes as a function of photon flux (solid and open squares,

diameter of 2r = 2.7 ± 0.5 nm (and aspect ratio of h/r = 0.84 ± 0.08). The QDs are defined by a bandgap of ∼0.95 ± 0.1 eV as inferred from optical absorption measurements. Interfacial donor-to-acceptor carrier dynamics were interrogated by optical pump−THz probe (OPTP) spectroscopy (see the SI for a description of the setup). OPTP is able to time-resolve the photoconductivity of a sample in a contactless fashion and with subps resolution and has been proven to be a powerful approach for monitoring interfacial carrier dynamics in QD-sensitized systems.16−21 As long as charge carriers remain confined within the QD, the real photoconductivity is zero and a null OPTP response is resolved.22 As ET from the QD to the oxide proceeds, the real photoconductivity Re[σ] becomes finite and increases as a function of pump−probe delay. As such, the increasing OPTP signal is a direct readout of an increasing number of free carriers in the oxide conduction band Noxide at different times following QD photoexcitation (Re[σ](t) ∝ Noxide(t)).16 If multiple carriers per photon are collected, that is, MEC occurs, the conductivity will increase accordingly. Figure 1A shows the time-dependent, pump-induced real conductivity as a function of 800 nm photon flux (between ∼3.9 × 1014 and ∼2.2 × 1016 photons/cm2), normalized to the conductivity observed at 1 ns. The rise of the real conductivity directly reflects ET taking place from the QDs to the oxide. Similarly, the subsequent decay component represents the back-ET processes (BET) from the oxide conduction band to the QDs and/or to oxide surface states.16 As evident from Figure 1A (and its inset) for fluences ≤ 5 × 1015 photons/cm2, the scaled photoconductivity traces are identical. This indicates that the photoconductivity originates primarily from ET in the single exciton (1X) regime, with an average number of photogenerated excitons/QD, ⟨N⟩, well below 1 (specifically ⟨N⟩ ≤ 0.15 demonstrated below). Given the large separation of time scales, the dynamics associated with ET and BET can be well-described by a stretched-exponential ingrowth and an exponential decay,16 respectively, so that Re[(t)] = A(1 − e−(t/τET)β)e−t/τBET. The black solid line fit in Figure 1A provides, for the single exciton (1X) ET dynamics: τ1X ET = 4.2 ± 0.2 ps and τ1X BET = 14.7 ± 0.3 ns respectively, with β = 0.6 ± 0.1. For excitation fluences exceeding 5 × 1015 photons/cm2, the photoconductivity becomes distinctly fluence-dependent. 2655

DOI: 10.1021/acs.jpclett.7b00966 J. Phys. Chem. Lett. 2017, 8, 2654−2658

Letter

The Journal of Physical Chemistry Letters

Figure 2. Multiexciton collection efficiency for PbS QDs sensitizing a SnO2 mesoporous film under hν/Eg < 2 photon excitation. (A) Black and 2X nX white squares represent 800 nm fluence-dependent N1X oxide and Noxide electron populations resolved in the oxide by OPTP (Noxide ∝ Re(σ)max) for Pb1X passivated QDs. The solid blue lines represent modeling following Poisson statistics of the number of electrons in NQD and N2X QD states populating QDs after excitation. The difference between biexcitons photogenerated in the QDs and transferred into the oxide represents losses in the QDs via Auger recombination (red area). The highlighted green area shows the additional electrons transferred by photogenerated biexcitons in the oxide electrode. (B) The same plot as panel (A) but for a sample where QDs sensitizing SnO2 are passivated by 4-mercaptobenzoic acid (4-MBA), clearly leading to enhancement of the biexciton collection efficiency.

the 2X QD population is highlighted in Figure 2A as the green area). While Poisson modeling reveals the total number of populated biexciton states within the QDs, the MEC efficiency monitored by THz represents the percentage of electrons (not states) injected into oxide. In this respect, a 100% MEC figure will indicate that two electrons are collected in the oxide electrode after the absorption of two photons in a QD; the 58% value inferred from the fit in Figure 2 implies that, on average, QDs populated with biexcitons in the atomically passivated PbS QDs are only able to transfer to the oxide 1.16 ± 0.06 electrons or, equivalently, that 84 ± 6% of the biexcitons populating QDs recombine efficiently via Auger-related processes (recombination losses in the QDs are highlighted in Figure 2A as the red area). It has been recognized previously that total or partial hole localization in the QD capping shell, for example, by exploiting core−shell type-II or quasi-type-II band alignment, allows for enhanced Auger lifetimes in the QDs.28−30 Accordingly, many reports have demonstrated that the QD hole wave function preferentially localizes in the capping layer in thiol-based, molecularly capped QDs.31−34 Following those reports, we prepared PbS/SnO2 samples for which the QDs were passivated with 4-mercaptobenzoic acid (4-MBA) ligands. Before analyzing the MEC efficiency at the QD−oxide interface, we experimentally verified that the molecular capping treatment did not affect the ET and BET rates in the 1X regime in comparison with atomically passivated QDs, that is, the treatment does not introduce new recombination pathways in the system (see Figure S5 in the SI) nor does it affect the interfacial coupling and or energetics of the system.17 1X 2X Figure 2B summarizes the measured Noxide and Noxide components as a function of 800 nm fluence for samples where QDs are passivated by 4-MBA molecules. For this data 2X set, the modeled N1X QD and NQD contributions to MEC as a function of ⟨N⟩ lead to an estimated electron extraction efficiency from biexciton states for the 4-MBA-passivated QDs equaling η2X ET = 88 ± 4%, substantially larger than the 58 ± 3% measured for Pb-capped QDs. This figure implies that on

respectively, in Figure 2A). On the other hand, the number of 2X electrons in nX states populating the QDs (e.g., N1X QD and NQD) 5,26,27 after excitation can be modeled using Poisson statistics. The number of electrons NnX QD in nX states populating the QDs before ET takes place can be inferred from nX NQD = ⟨N ⟩

Pn X =

n∗Pn X n →+∞ ∑n = 0 (n∗Pn X) n −⟨N ⟩

⟨N ⟩ e n!

where

n = 0, 1, 2, ...

(1)

Here, ⟨N⟩ represents the average number of electrons per QD, which we assume to be proportional to the incident photon flux.6 Describing the N1X oxide versus fluence obtained from the maximum amplitude of the real conductivity OPTP signal using 1X eq 1 (i.e., simply equating N1X QD to Noxide), we obtain a correlation between the experimentally employed photon flux (Figure 2 bottom x axis) and the average number of electrons photogenerated within the QDs ⟨N⟩ (Figure 2 upper x axis). From this estimate of ⟨N⟩, higher-order multiexciton contributions NnX QD as a function of photon flux can be obtained directly from eq 1. The blue solid lines in Figure 2A represent 2X the inferred N1X QD and NQD contributions to MEC as a function of ⟨N⟩. Within this model, the number of electrons in triexciton states N3X QD accounts for at most 10% of photogenerated electrons at the highest fluence analyzed here (see Figure S4 in the SI). Accordingly, the triexciton contribution is negligible in the normalized 2X dynamics, as is apparent from the inset of Figure 1B, and will therefore be neglected in the following. The biexciton collection efficiency toward the oxide can then easily be calculated from the ratio between the experimentally determined number of electrons transferred toward the oxide (N2X oxide, white open squares in Figure 2A) and the theoretically expected (maximum) number of biexcitons generated in the QDs as a function of photon flux (N2X QD, solid blue line in Figure 2A). This leads to an estimated electron extraction efficiency from biexciton states in QDs of η2X ET = 58 ± 3% for the atomically passivated PbS QDs sensitizing SnO2 (the gain from 2656

DOI: 10.1021/acs.jpclett.7b00966 J. Phys. Chem. Lett. 2017, 8, 2654−2658

Letter

The Journal of Physical Chemistry Letters average 1.76 ± 0.08 electrons are harvested in the oxide electrode from QDs populated by biexcitons, as opposed to 1.16 ± 0.06 in the absence of molecular passivation. Therefore, recombination losses associated with Auger recombination in the molecularly passivated QDs have been reduced, leading to an almost 5-fold increase in biexciton MEC after the capping of QDs by 4-MBA. On the basis of these numbers, we calculate KAug = 7 × 1011 s−1 (τAug ≈ 1.5 ps) from KET/(KAug + KET) = 0.76. This estimate is consistent with the value τAug ≈ 3 ps inferred from colloidal QDs and the nanocrystal volume (see the SI). The slight difference between the Auger rate calculated from the measurements (1.5 ps) and that calculated from the QD volume (3 ps) can be traced to the nature of our samples and/or the sequential nature of the biexciton collection process, as discussed in detail in the SI. Our study highlights the important role of the hybrid organic−inorganic shell−core interfacial chemistry and energetics for controlling the collection yield from multiexciton states in QDs. Furthermore, our results support the notion that a simple kinetic competition between the QD-to-oxide transfer rate (KET) and the Auger recombination rate (KAug) within the QDs determines the MEC efficiency at the QD−oxide interface. To efficiently harvest biexcitons in QDs sensitizing a mesoporous electrode, ultrafast interfacial ET rates must effectively compete with Auger recombination processes taking place in the QDs: KET ≫ KAug.13,14,26 As a result, reducing Auger rates, for example, by localization of holes in the outer QD shell, via a convenient surface capping modulation, allows a substantial boosting to the interfacial MEC efficiency in sensitized systems (see the SI for a discussion relating the expected impact of our findings for a complete QD-sensitized solar cell device, that is, including a hole conductor).

■

Present Address §

H.I.W.: Institute of Physics, Johannes Gutenberg-University Mainz, Staudingerweg 9, 55128 Mainz, Germany. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to Sapun Parekh for his proofreading and constructive feedback of our manuscript. This work has been financially supported by the Max Planck Society. H.I.W. is a recipient of a fellowship of the Graduate School Materials Science in Mainz (MAINZ) funded through the German Research Foundation in the Excellence Initiative (GSC 266). E.C. acknowledges financial support from the Max Planck Graduate Center (MPGC).

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.7b00966. QD-sensitized SnO2 sample preparation and characterization (TEM and absorbance); description of the THzTDS setup and measurements; discussion of the boost in 2X ET rates when compared with those in the 1X regime; discussion of the boost in 2X BET rates when compared with those in the 1X regime; estimate of multiexciton contributions to the TRTS signal in percentage from Poisson statistics; OPTP ET dynamics and estimate of passivation efficiency for atomically and molecularly passivated QDs; early-time OPTP 2X dynamics as a function of QD capping; estimation of Auger lifetimes, comparison to literature values and implications toward MEC efficiency estimates from Poisson statistics; and expected impact of our findings for a complete QD-sensitized solar cell device (PDF)

■

REFERENCES

(1) Shockley, W.; Queisser, H. J. Detailed Balance Limit of Efficiency of P-N Junction Solar Cells. J. Appl. Phys. 1961, 32, 510−519. (2) Ross, R. T.; Nozik, A. J. Efficiency of Hot-Carrier Solar-Energy Converters. J. Appl. Phys. 1982, 53, 3813−3818. (3) Kolodinski, S.; Werner, J. H.; Wittchen, T.; Queisser, H. J. Quantum Efficiencies Exceeding Unity Due to Impact Ionization in Silicon Solar-Cells. Appl. Phys. Lett. 1993, 63, 2405−2407. (4) Schaller, R. D.; Klimov, V. I. High Efficiency Carrier Multiplication in PbSe Nanocrystals: Implications for Solar Energy Conversion. Phys. Rev. Lett. 2004, 92, 186601. (5) Klimov, V. I. Spectral and Dynamical Properties of Multilexcitons in Semiconductor Nanocrystals. Annu. Rev. Phys. Chem. 2007, 58, 635−673. (6) Trinh, M. T.; Houtepen, A. J.; Schins, J. M.; Hanrath, T.; Piris, J.; Knulst, W.; Goossens, A. P. L. M.; Siebbeles, L. D. A. In Spite of Recent Doubts Carrier Multiplication Does Occur in PbSe Nanocrystals. Nano Lett. 2008, 8, 1713−1718. (7) ten Cate, S.; Sandeep, C. S. S.; Liu, Y.; Law, M.; Kinge, S.; Houtepen, A. J.; Schins, J. M.; Siebbeles, L. D. A. Generating Free Charges by Carrier Multiplication in Quantum Dots for Highly Efficient Photovoltaics. Acc. Chem. Res. 2015, 48 (2), 174−181. (8) Sandeep, C. S. S.; ten Cate, S.; Schins, J. M.; Savenije, T. J.; Liu, Y.; Law, M.; Kinge, S.; Houtepen, A. J.; Siebbeles, L. D. A. High Charge-Carrier Mobility Enables Exploitation of Carrier Multiplication in Quantum-Dot Films. Nat. Commun. 2013, 4, 2360. (9) Franceschetti, A.; An, J. M.; Zunger, A. Impact Ionization Can Explain Carrier Multiplication in PbSe Quantum Dots. Nano Lett. 2006, 6, 2191−2195. (10) Allan, G.; Delerue, C. Role of Impact Ionization in Multiple Exciton Generation in PbSe Nanocrystals. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 73, 205423. (11) Hanna, M. C.; Nozik, A. J. Solar Conversion Efficiency of Photovoltaic and Photoelectrolysis Cells with Carrier Multiplication Absorbers. J. Appl. Phys. 2006, 100, 074510. (12) Sambur, J. B.; Novet, T.; Parkinson, B. A. Multiple Exciton Collection in a Sensitized Photovoltaic System. Science 2010, 330, 63− 66. (13) Zidek, K.; Zheng, K. B.; Abdellah, M.; Lenngren, N.; Chabera, P.; Pullerits, T. Ultrafast Dynamics of Multiple Exciton Harvesting in the Cdse-ZnO System: Electron Injection Versus Auger Recombination. Nano Lett. 2012, 12, 6393−6399. (14) Yang, Y.; Lian, T. Q. Multiple Exciton Dissociation and Hot Electron Extraction by Ultrafast Interfacial Electron Transfer from PbS QDs. Coord. Chem. Rev. 2014, 263-264, 229−238. (15) Huang, J.; Huang, Z. Q.; Yang, Y.; Zhu, H. M.; Lian, T. Q. Multiple Exciton Dissociation in Cdse Quantum Dots by Ultrafast Electron Transfer to Adsorbed Methylene Blue. J. Am. Chem. Soc. 2010, 132, 4858−4864. (16) Wang, H.; Barcelo, I.; Lana-Villarreal, T.; Gomez, R.; Bonn, M.; Canovas, E. Interplay between Structure, Stoichiometry and Electron

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Hai I. Wang: 0000-0003-0940-3984 Mischa Bonn: 0000-0001-6851-8453 Enrique Cánovas: 0000-0003-1021-4929 2657

DOI: 10.1021/acs.jpclett.7b00966 J. Phys. Chem. Lett. 2017, 8, 2654−2658

Letter

The Journal of Physical Chemistry Letters Transfer Dynamics in Silar-Based Quantum Dot-Sensitized Oxides. Nano Lett. 2014, 14, 5780−5786. (17) Wang, H. I.; Lu, H.; Nagata, Y.; Bonn, M.; Cánovas, E. Dipolar Molecular Capping in Quantum Dot-Sensitized Oxides: Fermi Level Pinning Precludes Tuning Donor−Acceptor Energetics. ACS Nano 2017, 11 (5), 4760−4767. (18) Wang, H.; McNellis, E. R.; Kinge, S.; Bonn, M.; Cánovas, E. Tuning Electron Transfer Rates through Molecular Bridges in Quantum Dot Sensitized Oxides. Nano Lett. 2013, 13, 5311−5315. (19) Cánovas, E.; Wang, H.; Karakus, M.; Bonn, M. Hot Electron Transfer from PbSe Quantum Dots Molecularly Bridged to Mesoporous Tin and Titanium Oxide Films. Chem. Phys. 2016, 471, 54−58. (20) Cánovas, E.; Moll, P.; Jensen, S. A.; Gao, Y. A.; Houtepen, A. J.; Siebbeles, L. D. A.; Kinge, S.; Bonn, M. Size-Dependent Electron Transfer from PbSe Quantum Dots to SnO2 Monitored by Picosecond Terahertz Spectroscopy. Nano Lett. 2011, 11, 5234−5239. (21) Zidek, K.; Zheng, K. B.; Ponseca, C. S.; Messing, M. E.; Wallenberg, L. R.; Chabera, P.; Abdellah, M.; Sundstrom, V.; Pullerits, T. Electron Transfer in Quantum-Dot-Sensitized ZnO Nanowires: Ultrafast Time-Resolved Absorption and Terahertz Study. J. Am. Chem. Soc. 2012, 134, 12110−12117. (22) Ulbricht, R.; Hendry, E.; Shan, J.; Heinz, T. F.; Bonn, M. Carrier Dynamics in Semiconductors Studied with Time-Resolved Terahertz Spectroscopy. Rev. Mod. Phys. 2011, 83, 543−586. (23) Zhu, H. M.; Lian, T. Q. Enhanced Multiple Exciton Dissociation from CdSe Quantum Rods: The Effect of Nanocrystal Shape. J. Am. Chem. Soc. 2012, 134 (27), 11289−11297. (24) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Quantization of Multiparticle Auger Rates in Semiconductor Quantum Dots. Science 2000, 287, 1011−1013. (25) Jha, P. P.; Guyot-Sionnest, P. Trion Decay in Colloidal Quantum Dots. ACS Nano 2009, 3 (4), 1011−1015. (26) Zhu, H. M.; Yang, Y.; Lian, T. Q. Multiexciton Annihilation and Dissociation in Quantum Confined Semiconductor Nanocrystals. Acc. Chem. Res. 2013, 46, 1270−1279. (27) Klimov, V. I. Optical Nonlinearities and Ultrafast Carrier Dynamics in Semiconductor Nanocrystals. J. Phys. Chem. B 2000, 104, 6112−6123. (28) Oron, D.; Kazes, M.; Banin, U. Multiexcitons in Type-II Colloidal Semiconductor Quantum Dots. Phys. Rev. B 2007, 75, 035330. (29) Klimov, V. I.; Ivanov, S. A.; Nanda, J.; Achermann, M.; Bezel, I.; McGuire, J. A.; Piryatinski, A. Single-Exciton Optical Gain in Semiconductor Nanocrystals. Nature 2007, 447, 441−446. (30) Nanda, J.; Ivanov, S. A.; Achermann, M.; Bezel, I.; Piryatinski, A.; Klimov, V. I. Light Amplification in the Single-Exciton Regime Using Exciton-Exciton Repulsion in Type-II Nanocrystal Quantum Dots. J. Phys. Chem. C 2007, 111, 15382−15390. (31) Frederick, M. T.; Amin, V. A.; Swenson, N. K.; Ho, A. Y.; Weiss, E. A. Control of Exciton Confinement in Quantum Dot-Organic Complexes through Energetic Alignment of Interfacial Orbitals. Nano Lett. 2013, 13, 287−292. (32) Wuister, S. F.; de Mello Donega, C.; Meijerink, A. Influence of Thiol Capping on the Exciton Luminescence and Decay Kinetics of CdTe and CdSe Quantum. J. Phys. Chem. B 2004, 108, 17393−17397. (33) Aldana, J.; Wang, Y. A.; Peng, X. G. Photochemical Instability of CdSe Nanocrystals Coated by Hydrophilic Thiols. J. Am. Chem. Soc. 2001, 123, 8844−8850. (34) Knowles, K. E.; Frederick, M. T.; Tice, D. B.; Morris-Cohen, A. J.; Weiss, E. A. Colloidal Quantum Dots: Think Outside the (Particlein-a-)Box. J. Phys. Chem. Lett. 2012, 3, 18−26.

2658

DOI: 10.1021/acs.jpclett.7b00966 J. Phys. Chem. Lett. 2017, 8, 2654−2658