Chinese Journal of Chemical Physics  2017, Vol. 30 Issue (6): 649-656

The article information

Jun-hui Wang, Gui-jie Liang, Kai-feng Wu
王俊慧, 梁桂杰, 吴凯丰
Long-lived Single Excitons, Trions, and Biexcitons in CdSe/CdTe Type-Ⅱ Colloidal Quantum Wells
CdSe/CdTe Ⅱ型胶态量子阱中长寿命的单激子、带电激子和双激子
Chinese Journal of Chemical Physics, 2017, 30(6): 649-656
化学物理学报, 2017, 30(6): 649-656
http://dx.doi.org/10.1063/1674-0068/30/cjcp1711206

Article history

Received on: November 8, 2017
Accepted on: November 22, 2017
Long-lived Single Excitons, Trions, and Biexcitons in CdSe/CdTe Type-Ⅱ Colloidal Quantum Wells
Jun-hui Wang, Gui-jie Liang, Kai-feng Wu     
Dated: Received on November 8, 2017; Accepted on November 22, 2017
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 116023, China
*Author to whom correspondence should be addressed. Kai-feng Wu, E-mail:kwu@dicp.ac.cn
Part of the special issue for "the Chinese Chemical Society's 15th National Chemical Dynamics Symposium"
Abstract: Light-harvesters with long-lived excited states are desired for efficient solar energy conversion systems. Many solar-to-fuel conversion reactions, such as H2 evolution and CO2 reduction, require multiple sequential electron transfer processes, which leads to a complicated situation that excited states involves not only excitons (electron-hole pairs) but also multi-excitons and charged excitons. While long-lived excitons can be obtained in various systems (e.g., semiconductor nanocrystals), multi-excitons and charged excitons are typically shorted-lived due to nonradiative Auger recombination pathways whereby the recombination energy of an exciton is quickly transferred to the third carrier on a few to hundreds of picoseconds timescale. In this work, we report a study of excitons, trions (an exciton plus an additional charge), and biexcitons in CdSe/CdTe colloidal quantum wells or nanoplatelets. The type-Ⅱ band alignment effectively separates electrons and holes in space, leading to a single exciton lifetime of 340 ns which is ~2 order of magnitudes longer than that in plane CdSe nanoplatelets. More importantly, the electron-hole separation also dramatically slows down Auger decay, giving rise to a trion lifetime of 70 ns and a biexciton lifetime of 11 ns, among the longest values ever reported for colloidal nanocrystals. The long-lived exciton, trion, and biexciton states, combined with the intrinsically strong light-absorption capability of two-dimensional systems, enable the CdSe/CdTe type-Ⅱ nanoplatelets as promising light harvesters for efficient solar-to-fuel conversion reactions.
Key words: Solar energy    Excited state lifetime    Exciton    Trion    Biexciton    Type-Ⅱ quantum wells    Nanoplatelets    Auger recombination    
Ⅰ. INTRODUCTION

In solar energy conversion systems, efficient charge separation across the donor/acceptor interfaces is desired, which thus requires light harvesters with long-lived excited states [1-5]. Semiconductor nanocrystals or quantum dots (QDs) are a family of such harvesters, with exciton (a photogenerated electron-hole pair) lifetime on the order of 10 ns [6, 7]. In addition, the quantum confinement effect leads to strong wavefunction overlap at the nanocrystal/acceptor interface, facilitating charge transfer from nanocrystal to acceptors [8-11]. Indeed, previous studies on various nanocrystal/acceptor systems indicate that electrons can be extracted from nanocrystals on the picoseconds to sub-ns timescale, resulting in near-unity charge separation yields [12-20]. However, with a few exceptions [21-25], the majority of previous work is focused on single-exciton dissociation, presumably because single excitons are the most commonly encountered excited states for semiconductor nanocrystals.

In reality, many solar-to-fuel conversion reactions, such as H$_2$ evolution and CO$_2$ reduction [3], involve multiple sequential electron transfer events; see the scheme in FIG. 1. Under sunlight illumination, the nanocrystal absorbs one photon and creates an exciton (FIG. 1A). Interfacial electron transfer from nanocrystal to catalyst dissociated this exciton and generated a charge separated state. As mentioned above, this step is often very efficient because single excitons typically have a lifetime of $\tau_{\rm{X}}$$\approx$10 ns. If hole transfer from the nanocrystal is not fast enough, which is indeed the case for many photocatalytic systems [26], the next photon absorption occurs not in a neutral nanocrystal but in the charge separated state, resulting in a trion state (X$^*$) [27, 28], a neutral exciton plus an additional charge, in the nanocrystal. A trion has a much shorter lifetime of $\tau_{\rm{X}^*}$ than an exciton ($\tau_{\rm{X}}$) because both radiative and nonradiative rates are enhanced in the trion [28-30]. Especially significant is the activated nonradiative Auger recombination process [6, 31-33], in which the recombination energy of an exciton is quickly transferred to the extra carrier. Because of quantum confinement enhanced coulombic carrier-carrier interactions and the relaxation of momentum conservation [32], Auger decay in nanocrystals is extremely fast (a few to hundreds of picoseconds) [21, 31, 34], thus being very difficult for the interfacial electron transfer to compete with. In the second scenario (FIG. 1B), multiple excitons are not sequentially but directly generated in nanocrystals, either through multiple photon absorption [22, 24, 25] under concentrated sunlight or through a so-called carrier multiplication process [35-38] whereby absorption of a highly energetic photon generates multiple excitons. In this case, even the first electron transfer event needs to compete with radiative and nonradiative decays in a multiexciton state, at least a biexciton state (with a lifetime of $\tau_{\rm{BX}}$), which is very challenging because Auger recombination in a biexciton is even much faster than in a trion due to more Auger relaxation pathways available in a biexciton [36]. Thus, for multi-electron solar-to-fuel conversion reactions, excited states involve not only single excitons but also multiexcitons and charged excitons. In order to achieve efficient interfacial charge separation, nanocrystals should be engineered to simultaneously attain long-lived excitons, trions, and multi-excitons.

FIG. 1 Competition between excited state decay and interfacial electron transfer in photocatalysts based on nanocrystal/catalyst complexes. (A) In the case of sequential photon absorption, interfacial electron transfer (ET) first competes with single exciton radiative and/or nonradiative decays (with a lifetime of $\tau_{\rm{X}}$). Afterwards, photon absorption ($h\nu$) of the charge separate state creates a trion state (X$^*$), a neutral exciton plus an additional charge, in the nanocrystal; consequently, electron transfer needs to compete with the trion radiative and nonradiative decays (with a lifetime of $\tau_{\rm{X}^*}$), with the nonradiative decay dominated by ultrafast (typically ps to sub-ns) Auger recombination process. (B) In the presence of concentrated sunlight or carrier multiplication, a biexciton state (BX) is directly generated in the nanocrystal. In this case, the first electron transfer process already needs to compete with biexciton radiative and nonradiative Auger decays (with a lifetime of $\tau_{\rm{BX}}$) and the second electron transfer process competes with trion decays ($\tau_{\rm{X}^*}$). The Auger recombination in a biexciton is even much faster than in a trion due to more relaxation pathways available in a biexciton.

Recent developments in shape control of colloidal nanocrystals have led to colloidal nanoplatelets (NPLs) with atomically precise thickness of 1-2 nm [39-44]. These are essentially the colloidal form of two-dimensional (2D) semiconductor quantum wells (QWs) which have accomplished tremendous success in many modern technologies [45]. As extensively demonstrated [46, 47], QWs are promising light harvesters as they exhibit strong light-matter interaction, giving rise to large absorption cross sections. In addition, the restoration of momentum conservation in two dimensions in QWs may slow down Auger recombination as compared to QDs. On the other hand, however, the single exciton lifetime in QWs is much shorter than in QDs due to a so-called "giant oscillator strength transition" phenomenon [41, 48, 49]. To prolong the single exciton lifetime as well as further suppress Auger recombination of trions and biexcitons, the strategy of wavefunction engineering using heterostructures can be applied [8, 28, 50]. For example, it has been reported that both single exciton and multiexciton lifetimes are significantly lengthened in quasi-type Ⅱ CdSe/CdS core/shell QDs [51] and type-Ⅱ CdS/ZnSe or CdTe/CdSe core/shell QDs [50, 52]. These structures feature spatially separated electron-hole pairs and, hence, strongly reduced electron-hole wavefunction overlap that is essential for both radiative and Auger recombination pathways [53]. Herein, we apply this strategy to colloidal NPLs to obtain hetero-NPLs with simultaneously long-lived single excitons, trions, and biexcitons. By laterally growing CdTe crowns on CdSe core NPLs to form type-Ⅱ CdSe/CdTe NPLs, we observe a single exciton lifetime of 340 ns, a trion lifetime of 70 ns, and a biexciton lifetime of 11 ns, all of which are among the longest values ever reported for colloidal nanocrystals [31, 32, 51].

Ⅱ. MORPHOLOGY AND STRUCTURE OF CdSe/CdTe NPLS

Colloidal CdSe/CdTe NPLs were synthesized by lateral epitaxial growth of CdTe on pre-formed CdSe NPLs [54]. CdSe NPLs were prepared following well-established literature methods [39, 41] (see supplementary materials for details). The tunability in the thicknesses and thus absorption and emission spectra of CdSe NPLs are demonstrated in FIG. S1 (supplementary materials). Following the convention in the literature, zinc-blende NPLs with n monolayers (ML) have n+1 Cd atomic layers and n Se atomic layers [40]. Therefore, the thicknesses of the 3, 4, and 5 ML CdSe NPLs in FIG. S1 (supplementary materials) are 1.06, 1.36, and 1.66 nm, respectively. The atomically uniform thickness of the NPLs gives rises to the sharp absorption and photoluminescence (PL) bands (full width at half maximum ~35 meV) in FIG. S1. We use the 4 ML CdSe NPLs as the "seeds" for growing CdSe/CdTe NPLs [54, 55] (see supplementary materials for details). CdTe only laterally extends on CdSe NPLs and therefore the resulting CdSe/CdTe NPLs are also 4 ML but their sizes are increased to 36 nm$\times$14 nm (rectangular-shaped) from 19 nm$\times$8 nm for CdSe NPLs (FIG. 2A, 2B). This core/crown structure is schematically depicted in FIG. 2C. Due to a type Ⅱ band alignment between CdSe and CdTe, the lowest conduction band (CB) and valence band (VB) levels are situated in CdSe and CdTe domains, respectively. Therefore, in addition to the electronic transitions in CdSe and CdTe domains, there exists the lowest energy transition from the VB of CdTe to CB of CdSe [54], which is called charge transfer transition (CT) and is an indication of strong electronic coupling at the CdSe/CdTe interface.

FIG. 2 Morphology and structure of colloidal nanoplatelets (NPLs). (A, B) Representative transmission electron microscopy (TEM) images of CdSe NPLs seeds (A) and CdSe/CdTe core/crown NPLs. (C) Top, a scheme of the core/crown structure, with CdTe crown laterally growing on CdSe core; bottom, the type Ⅱ band alignment in this structure, with the lowest conduction band (CB) and valence band (CB) levels situated in CdSe and CdTe domains, respectively. Therefore, the lowest energy transition in this structure is a charge transfer transition (CT) associated with an electron in CdSe and a hole in CdTe.
Ⅲ. SPECTRAL PROPERTIES OF CdSe/CdTe NPLS

FIG. 3A shows the absorption and photoluminescence (PL) spectra of as-prepared CdSe/CdTe NPLs dispersed in hexane; the PL is collected under 400 nm $cw$ (continuous wave) excitation. In addition to the absorption peaks at 510 and 550 nm that can be assigned to the band edge excitonic bands in CdSe and CdTe, respectively, there is an absorption tail extending to 650 nm (FIG. 3A). This is the CT band typically observed in type-Ⅱ heterostructures, as depicted in FIG. 2C. The peak position of the CT band, measured using transient absorption which can select the lowest energy CT band from many overlapping transitions, is ~620 nm (FIG. 3A). The PL is also dominated by the emission from the CT band (with a quantum yield of ~40%), with a peak at ~650 nm. The large Stokes shift ($\Delta_{\rm{s}}$$\approx$100 meV) is in stark contrast to vanishingly small Stokes shift (a few meVs) observed in plane CdSe NPLs (FIG. S1 in supplementary materials) [41]. In CdSe NPLs, because of weak electron-hole exchange interaction and weak electron-phonon coupling, the Stokes shift is much smaller than the one observed in CdSe QDs (~20 meV) [56, 57]. In principle, the electron-hole exchange interaction in CdSe/CdTe NPLs is even weaker than in CdTe NPLs due to the reduced electron-hole wavefunction overlap. Hence, the large Stokes shift is most likely because of enhanced electron-phonon coupling in CdSe/CdTe NPLs [58]. Indeed, it has been reported that the spatially separated electron-hole pairs in type Ⅱ structures are more polarizable and can couple more effectively with longitudal optical (LO) phonons through Frolich interactions [59-61].

FIG. 3 Spectral properties of CdSe/CdTe type Ⅱ NPLs. (A) Absorption (green solid line) and photoluminescence (PL, brown dashed line) spectra of CdSe/CdTe NPLs. Different bands, including CdSe, CdTe, and charge transfer (CT) transitions, are labeled on the absorption spectrum. To better visualize the featureless CT band, we selectively excite this band using transient absorption spectroscopy ($-\Delta\alpha$, blue dashed line). (B) Photoluminescence excitation (PLE, solid lines) spectra of the CdSe/CdTe NPLs obtained by monitoring various positions (see the spikes) along the PL spectra (brown dashed line).

In addition to the large Stokes shift, yet another interesting observation is the large line-width of the CT emission (with a full with at half maximum of ~130 meV}), which is again in stark contrast to the very narrow emission (~35 meV) observed in plane CdSe NPLs (FIG. S1 in supplementary materials). Since the CdTe laterally grows on CdSe with a precision down to the atomic monolayer level, inhomogeneous broadening mechanisms should be excluded. To support this speculation, we collect the PL excitation (PLE) spectra of the CdSe/CdTe NPLs by monitoring various positions on the PL band (FIG. 3B; the spikes are due to scattered light and correspond to the monitored emission wavelengths). Indeed, these PLE spectra are essentially the same, suggesting that the ~130 meV emission line-width is homogeneously broadened and most likely results from the strong electron-phonon coupling [61]. Thus, both the large emission line-width and large Stokes shift are indicative of every effective separation of electrons and holes in space.

Ⅳ. DYNAMICAL PROPERTIES OF CdSe/CdTe NPLS A. Single excitons in NPLs

We measure the excited state lifetime of NPLs using time correlated single photon counting (TCSPC); the excitation source is a 150 fs, 3.1 eV femtosecond laser and the time resolution of the measurement is ~200 ps (see supplementary materials for details). FIG. 4 shows the time-resolved PL kinetics of CdSe and CdSe/CdTe NPLs, measured at pump fluences corresponding to an average exciton number per NPL of ~0.02, thus excluding the complication from multi-exciton dynamics [31]. For the CdSe NPLs (FIG. 4A), the single exciton recombination kinetics is highly heterogeneous. We fit the kinetics using a tri-exponential decay function, with the fitted time constants (and relative amplitudes) of 0.65 ns (27%), 3.8 ns (68%), and 17 ns (5%). Note that this fitting is not sufficient to describe some even longer-lived components on the hundreds of ns timescale (FIG. 4A, black dashed line) which can be ascribed to delayed PL from surface trapping induced charge separated states [62, 63]. Given the reasonably high PL quantum yield (~50%) of these NPLs, the 3.8 ns} component is most likely associated with the radiative time constant of the single exciton state, which is consistent with previous reports on CdSe NPLs [41] (see FIG. S2 in supplementary materials for a detailed discussion on single exciton lifetime). For the CdSe/CdTe NPLs (FIG. 4B), its kinetics is also fitted with a tri-exponential decay function, with time constants (and relative amplitudes) of 102 ns (35%), 340 ns (58%), and 5400 ns (7%). Again, we assign the major component time constant (340 ns) to the radiative decay of the single exciton state. Thus, by spatially separating electron-hole pairs, the type-Ⅱ electronic structure lengthens single-exciton lifetime by two orders of magnitude. This long lifetime is highly desired for achieving high charge separation yields in single-electron transfer processes.

FIG. 4 Dynamical properties of single exciton states in CdSe and CdSe/CdTe NPLs. (A) Time-resolved PL kinetics of CdSe NPLs (gray dashed line) measured with time-correlated single photon counting (TCSPC). The black solid line is a tri-exponential fit to the kinetics; the black dashed line indicates long-lived components which are not included in the fit. (B) Time-resolved PL kinetics of CdSe/CdTe NPLs (gray dashed line). The black solid line is a tri-exponential fit to the kinetics.
B. Trions in NPLs

Next, we examine the effect of electron-hole separation on the rate of trion recombination. Previous studies on colloidal nanocrystals have established that nanocrystals can be automatically charged under long-term photo illumination, without the need of introducing any redox species. The cause of this photocharging phenomenon is often attributed to Auger-assisted ionization [29, 64, 65] and/or hot-carrier trapping from nanocrystals [30, 66-68], essentially creating a charged separated state with one charge on the nanocrystal surface and the other inside the nanocrystal volume. Excitation of this charged separated state then leads to a trion state inside the nanocrystal.

To investigate the photocharging in the NPLs, we adopt a previously-established method: comparing PL kinetics of a solution sample measured under either static or vigorously stirred conditions [69-71]. Following a laser pulse excitation, some of the photogenerated charge separated states may have a long enough lifetime that they persist until the next laser pulse arrives. In the case of a static solution these charged separated states can be re-excited, generating trion species, whereas in the stirred sample the excitation volume is continuously replenished with fresh, charge-neutral nanocrystals. As shown in FIG. 5A and 5B, for both CdSe and CdSe/CdTe NPLs, the static solution has a higher initial amplitude and a faster decay than the stirred solution, which is a conclusive signature of trion species [71]. The radiative rate of a trion ($r_{\rm{X^*, r}}$) is higher than that of an exciton ($r_{\rm{X}}$) and in the framework of statistic scaling, they can be related by [36]: $r_{\rm{X^*, r}}$=2$r_{\rm{X}}$, which explains the increased initial PL amplitude and partially accounts for the shortening of the PL lifetime. The major contribution to the lifetime shortening is from the nonradiative Auger recombination, which is typically on the ps to sub-ns timescale in nanocrystals. By normalizing the static and stirred kinetics at the long-lived tail and performing a subtraction between them, we can isolate the recombination kinetics of the trion states, which are displayed in FIG. 5C and 5D for CdSe and CdSe/CdTe NPLs, respectively. By fitting these kinetics using single-exponential decays, we find trion lifetimes of 0.92 ns for CdSe NPLs and 70 ns for CdSe/CdTe NPLs. To the best of our knowledge, the 70 ns trion lifetime is the longest value ever reported for all kinds of colloidal nanocrystals at room temperature [28, 34, 71, 72]. To extract the Auger lifetime of the trion ($\tau_{\rm{X^*, A}}$) from the measured time constants, we use the following relationship [36], 1/$\tau_{\rm{X^*, A}}$=1/$\tau_{\rm{X^*}}$-1/$\tau_{\rm{X^*, r}}$, where $\tau_{\rm{X^*, r}}$ is the radiative lifetime of the trion (=1/$\tau_{\rm{X^*, r}}$). The calculated values of $\tau_{\rm{X^*, A}}$ are 1.8 and 119 ns, respectively. Hence, the Auger recombination of trions in CdSe/CdTe NPLs is suppressed by 66-fold as compared to CdSe NPLs. Recognizing that Auger recombination is a three-particle process involving a interband transition between an electron-hole pair and a intraband transition of the third carrier, there are two possible effects accounting for the Auger suppression in CdSe/CdTe NPLs. First, the type Ⅱ band alignment effectively separates electrons and holes in space and thus suppresses the interband transition. Secondly, the strong coulombic repulsion between the two carriers of the same sign located in the same domain force them to spatially avoid each other, leading to a trion configuration with the extra charge located far away from the electron-hole pair (FIG. 5D inset), which suppresses the intraband transition involved in Auger recombination.

FIG. 5 Dynamical properties of trion states in CdSe and CdSe/CdTe NPLs. (A, B) Time-resolved PL kinetics of CdSe NPLs (A) and CdSe/CdTe NPLs (B), measured under both static (red solid line) and vigorously stirred (black solid line) situations. (C, D) Trion recombination kinetics (gray dashed line) in CdSe NPLs (C) and CdSe/CdTe NPLs (D) measured by subtracting the stirred kinetics from static kinetics after normalizing them to the long lived tail. The black solid lines are single-exponential fits to the trion recombination kinetics. The insets show possible configurations of trions in NPLs.

We note that the sign of the trion (positive or negative) remains unknown for our NPL samples, but based on previous studies, photocharging of Ⅱ-Ⅵ nanocrystals usually results negatively charged trions because the hole is more susceptible to surface trapping for this family of materials [29, 70].

C. Biexcitons in NPLs

To study recombination dynamics of biexcitons of our NPL samples, we measure time-resolved PL kinetics as a function of laser pulse fluence (FIG. 6, traces are normalized to the long-lived tails). To avoid the introduction of trion species, samples were vigorously stirred in these measurements. As shown in FIG. 6A, 6B, With increasing laser pulse fluence (j, quantified in terms of number of photons per cm$^2$ per pulse), fast decay components on the sub-ns and 10 ns timescales appear for the CdSe (FIG. 6A) and CdSe/CdTe (FIG. 6B) NPLs, respectively, which is a typical signature of multiexciton Auger recombination [31]. The measured traces can be normalized such as to match their long-lived tails at longer delay times after all multiexcitons have already decayed. At this stage, the PL kinetics is exclusively contributed by single-exciton recombination and, hence, the signal magnitude (I) is proportional to the total number of photoexcited QDs in the ensemble. If we define P(0) as the fraction of unexcited QDs in the ensemble, then we have: I$\propto$1-P(0). Assuming Poissonian statistics for photon absorption typically observed for nanocrystals when using the above band-gap excitation [31], P(0) =e$^{-\langle N\rangle}$, where $\langle N\rangle$ is the average number of photons absorbed per NPL and hence is also the average exciton number per NPL immediately following photoexcitation. $\langle N\rangle$ is related to the laser pulse fluence by $\langle N\rangle=\sigma j$ with $\sigma$ being the QD absorption cross-section for the pump photons (3.1 eV). Therefore, the PL signals at 1 ns for CdSe NPLs (FIG. 6C) and at 40 ns for CdSe/CdTe NPLs (FIG. 6D) can be fitted with the following expression: I$\propto$1-e$^{-\sigma j}$, with $\sigma$ as the only adjustable fitting parameter. The near perfect fitting indicates that the photon absorption events under our experimental conditions indeed follow the Poissonian statistics. It also gives the $\langle N\rangle$ value for each corresponding pump fluence j.

FIG. 6 Dynamical properties of biexciton states in CdSe and CdSe/CdTe NPLs. (A, B) Time-resolved PL kinetics of CdSe NPLs (A) and CdSe/CdTe NPLs (B) measured under various per-pulse pump fluences from a few to thousands of nJ/cm$^2$. Traces are normalized at t=1 ns in (A) and 40 ns in (B). (C, D) PL intensities (red triangles) at indicated times as a function of per-NPL exciton number, $\langle N\rangle$, for the CdSe NPLs (C) and CdSe/CdTe NPLs (D). The blue solid lines are fitted saturation curves according to Poissonian statistics. (E, F) Biexciton recombination kinetics (gray dashed lines) for CdSe NPLs (E) and CdSe/CdTe NPLs (F) obtained by taking the difference between PL traces corresponding to $\langle N\rangle$=0.1 (red solid line) and 0.05 (blue solid line). The black solid lines are single-exponential fits to the biexciton recombination kinetics.

The biexciton recombination kinetics can be extracted from the pump fluence dependent PL traces by performing a subtraction between two traces for which $\langle N\rangle$ is small enough so that the contribution of higher-order multiexcitons beyond biexciton is negligible. FIG. 6E and 6F show the subtraction between kinetic traces with $\langle N\rangle$=0.1 and $\langle N\rangle$=0.05. The subtraction indeed results in single-exponential decay components that can be assigned to biexciton recombination processes. By fitting them, we obtain biexciton lifetimes ($\tau_{\rm{BX}}$) of 0.35 and 11 ns for the CdSe (FIG. 6E) and CdSe/CdTe (FIG. 6F) NPLs, respectively. Again, with the exception of some reports on complete biexciton Auger suppression in bulk-like giant core/shell nanocrystals [73, 74], the biexciton lifetime of 11 ns is among the highest value reported for quantum confined colloidal nanocrystals. As in the case of trions, to extract Auger lifetimes of biexcitons ($\tau_{\rm{BX, A}}$) from the measured biexciton lifetimes, we use the following expression [36], 1/$\tau_{\rm{BX, A}}$=1/$\tau_{\rm{BX}}$-1/$\tau_{\rm{BX, r}}$, where $\tau_{\rm{BX, r}}$ is the radiative lifetime of the biexciton. Using standard statistical scaling of radiative rates, $\tau_{\rm{BX, r}}$ is related to the single exciton radiative lifetime ($\tau_{\rm{X}}$) by $\tau_{\rm{BX, r}}=\tau_{\rm{X}}$/4 [36] The calculated values of $\tau_{\rm{BX, A}}$ are 0.55 ns for CdSe NPLs and 12.6 ns for CdSe/CdTe NPLs. Therefore, the Auger recombination in CdSe/CdTe NPLs is suppressed by 23-fold as compared to CdSe NPLs. Similar to our rationale for Auger suppression in trions, we believe that both interband and intraband transitions involved in the biexciton Auger recombination are suppressed, due to electron-hole separation and spatially avoided biexcitons (FIG. 6F inset), respectively.

Ⅴ. CONCLUSION

In order to obtain light harvesters with simultaneously long-lived single excitons, trions, and biexcitons for multi-electron solar-to-fuel conversion reactions, we synthesized type-Ⅱ CdSe/CdTe colloidal nanoplatelets (NPLs), or quantum wells, by laterally growing CdTe crowns on CdSe core NPLs with thickness being control led down to the atomic monolayer level. Investigation of the spectral properties of the type-Ⅱ NPLs revealed a giant Stokes shift and a large emission line-width that were radically different from the starting CdSe NPLs, which was attributed to very effective separation of electrons and holes in space in the lowest energy transition. This electron-hole spatial separation led to dramatically prolonged lifetimes of single excitons, trions, and biexcitons (340, 70, and 11 ns, respectively) in CdSe/CdTe NPLs as compared to those (38, 0.92, and 0.35 ns, respectively) in CdSe NPLs. All of these lifetime constants for CdSe/CdTe NPLs are among the longest values ever reported for their counterparts in colloidal nanocrystals. These long excited state lifetimes should leave ample time window for efficient carrier extraction through interfacial charge transfer, a key enabler for high-performance light harvesters for multi-electron solar-to-fuel conversion systems.

Supplementary materials: FIG. S1, S2, and sample synthesis details and ultrafast spectroscopy set-ups are given.

Ⅵ. ACKNOWLEDGEMENTS

This work was supported by the start-up funding from Dalian Institute of Chemical Physics, Chinese Academy of Sciences, and the Collaborative Innovation Center of Chemistry for Energy Materials (iChEM-2011).

Reference
[1] N. S. Lewis, and D. G. Nocera, Proc. Natl. Acad. Sci. USA 103 , 15729 (2006). DOI:10.1073/pnas.0603395103
[2] D. Gust, T. A. Moore, and A. L. Moore, Acc. Chem. Res. 34 , 40 (2001). DOI:10.1021/ar9801301
[3] X. B. Chen, C. Li, M. Grätzel, R. Kostecki, and S. S. Mao, Chem. Soc. Rev. 41 , 7909 (2012). DOI:10.1039/c2cs35230c
[4] J. H. Alstrum-Acevedo, M. K. Brennaman, and T. J. Meyer, Inorg. Chem. 44 , 6802 (2005). DOI:10.1021/ic050904r
[5] A. J. Nozik, Annu. Rev. Phys. Chem. 29 , 189 (1978). DOI:10.1146/annurev.pc.29.100178.001201
[6] V. I. Klimov, Annu. Rev. Phys. Chem. 58 , 635 (2007). DOI:10.1146/annurev.physchem.58.032806.104537
[7] D. J. Norris, A. L. Efros, M. Rosen, and M. G. Bawendi, Phys. Rev. B 53 , 16347 (1996). DOI:10.1103/PhysRevB.53.16347
[8] H. M. Zhu, and T. Q. Lian, Energy Environ. Sci. 5 , 9406 (2012). DOI:10.1039/c2ee22679k
[9] H. M. Zhu, Y. Yang, K. F. Wu, and T. Q. Lian, Annu. Rev. Phys. Chem. 67 , 259 (2016). DOI:10.1146/annurev-physchem-040215-112128
[10] W. A. Tisdale, and X. Y. Zhu, Proc. Natl. Acad. Sci. USA 108 , 965 (2011). DOI:10.1073/pnas.1006665107
[11] W. A. Tisdale, K. J. Williams, B. A. Timp, D. J. Norris, E. S. Aydil, and X. Y. Zhu, Science 328 , 1543 (2010). DOI:10.1126/science.1185509
[12] C. Burda, T. C. Green, and S. Link, M. A. El-Sayed, J. Phys. Chem. B 103 , 1783 (1999). DOI:10.1021/jp9843050
[13] M. Sykora, M. A. Petruska, J. Alstrum-Acevedo, I. Bezel, T. J. Meyer, and V. I. Klimov, J. Am. Chem. Soc. 128 , 9984 (2006). DOI:10.1021/ja061556a
[14] A. Boulesbaa, A. Issac, D. Stockwell, Z. Q. Huang, J. R. Huang, J. C. Guo, and T. Q. Lian, J. Am. Chem. Soc. 129 , 15132 (2007). DOI:10.1021/ja0773406
[15] J. R. Huang, D. Stockwell, Z. Q. Huang, D. L. Mohler, and T. Q. Lian, J. Am. Chem. Soc. 130 , 5632 (2008). DOI:10.1021/ja8003683
[16] J. E. Huang, Z. Q. Huang, S. Y. Jin, and T. Q. Lian, J. Phys. Chem. C 112 , 19734 (2008). DOI:10.1021/jp808291u
[17] H. M. Zhu, N. H. Song, and T. Q. Lian, J. Am. Chem. Soc. 132 , 15038 (2010). DOI:10.1021/ja106710m
[18] Y. Yang, W. E. Rodríguez-Córdoba, and T. Q. Lian, J. Am. Chem. Soc. 133 , 9246 (2011). DOI:10.1021/ja2033348
[19] K. F. Wu, Z. Liu, H. M. Zhu, and T. Q. Lian, J. Phys. Chem. A 117 , 6362 (2013). DOI:10.1021/jp402511m
[20] K. F. Wu, N. H. Song, Z. Liu, H. M. Zhu, W. RodríguezCórdoba, and T. Q. Lian, J. Phys. Chem. A 117 , 7561 (2013). DOI:10.1021/jp402425w
[21] V. V. Matylitsky, L. Dworak, V. V. Breus, T. Basché, and J. Wachtveitl, J. Am. Chem. Soc. 131 , 2424 (2009). DOI:10.1021/ja808084y
[22] J. R. Huang, Z. Q. Huang, Y. Yang, H. M. Zhu, and T. Q. Lian, J. Am. Chem. Soc. 132 , 4858 (2010). DOI:10.1021/ja100106z
[23] Y. Yang, W. Rodríguez-Córdoba, and T. Q. Lian, Nano Lett. 12 , 4235 (2012). DOI:10.1021/nl301847r
[24] H. M. Zhu, and T. Q. Lian, J. Am. Chem. Soc. 134 , 11289 (2012). DOI:10.1021/ja304724u
[25] H. M. Zhu, N. H. Song, W. Rodríguez-Córdoba, and T. Q. Lian, J. Am. Chem. Soc. 134 , 4250 (2012). DOI:10.1021/ja210312s
[26] K. F. Wu, Z. Y. Chen, H. J. Lv, H. M. Zhu, C. L. Hill, and T. Q. Lian, J. Am. Chem. Soc. 136 , 7708 (2014). DOI:10.1021/ja5023893
[27] M. Shim, P. Guyot-Sionnest, Nature 407 , 981 (2000).
[28] P. P. Jha, P. Guyot-Sionnest, ACS Nano 3 , 1011 (2009). DOI:10.1021/nn9001177
[29] C. Galland, Y. Ghosh, A. Steinbrück, J. A. Hollingsworth, H. Htoon, and V. I. Klimov, Nat. Commun. 3 , 908 (2012). DOI:10.1038/ncomms1916
[30] C. Galland, Y. Ghosh, A. Steinbrück, M. Sykora, J. A. Hollingsworth, V. I. Klimov, and H. Htoon, Nature 479 , 203 (2011). DOI:10.1038/nature10569
[31] V. I. Klimov, A. A. Mikhailovsky, D. W. McBranch, C. A. Leatherdale, and M. G. Bawendi, Science 287 , 1011 (2000). DOI:10.1126/science.287.5455.1011
[32] I. Robel, R. Gresback, U. Kortshagen, R. D. Schaller, and V. I. Klimov, Phys. Rev. Lett. 102 , 177404 (2009). DOI:10.1103/PhysRevLett.102.177404
[33] A. Pandey, and P. Guyot-Sionnest, J. Chem. Phys. 127 , 111104 (2007). DOI:10.1063/1.2786068
[34] A. W. Cohn, J. D. Rinehart, A. M. Schimpf, A. L. Weaver, and D. R. Gamelin, Nano Lett. 14 , 353 (2014). DOI:10.1021/nl4041675
[35] R. D. Schaller, and V. I. Klimov, Phys. Rev. Lett. 92 , 186601 (2004). DOI:10.1103/PhysRevLett.92.186601
[36] V. I. Klimov, Ann. Rev. Condens. Matter. Phys. 5 , 285 (2014). DOI:10.1146/annurev-conmatphys-031113-133900
[37] R. J. Ellingson, M. C. Beard, J. C. Johnson, P. R. Yu, O. I. Micic, A. J. Nozik, A. Shabaev, and A. L. Efros, Nano Lett. 5 , 865 (2005). DOI:10.1021/nl0502672
[38] O. E. Semonin, J. M. Luther, S. Choi, H. Y. Chen, J. B. Gao, A. J. Nozik, and M. C. Beard, Science 334 , 1530 (2011). DOI:10.1126/science.1209845
[39] S. Ithurria, and B. Dubertret, J. Am. Chem. Soc. 130 , 16504 (2008). DOI:10.1021/ja807724e
[40] S. Ithurria, G. Bousquet, and B. Dubertret, J. Am. Chem. Soc. 133 , 3070 (2011). DOI:10.1021/ja110046d
[41] S. Ithurria, M. D. Tessier, B. Mahler, R. P. S. M. Lobo, B. Dubertret, and A. L. Efros, Nat. Mater. 10 , 936 (2011). DOI:10.1038/nmat3145
[42] J. Joo, J. S. Son, S. G. Kwon, J. H. Yu, and T. Hyeon, J. Am. Chem. Soc. 128 , 5632 (2006). DOI:10.1021/ja0601686
[43] J. S. Son, X. D. Wen, J. Joo, J. Chae, S. I. Baek, K. Park, J. H. Kim, K. An, J. H. Yu, S. G. Kwon, S. H. Choi, Z. W. Wang, Y. W. Kim, Y. Kuk, R. Hoffmann, and T. Hyeon, Angew. Chem. Int. Ed. Engl. 48 , 6861 (2009). DOI:10.1002/anie.v48:37
[44] J. S. Son, J. H. Yu, S. G. Kwon, J. Lee, and J. Joo, Adv. Mater. 23 , 3214 (2011). DOI:10.1002/adma.v23.28
[45] Y. Arakawa, and H. Sakaki, Appl. Phys. Lett. 40 , 939 (1982). DOI:10.1063/1.92959
[46] W. H. Knox, R. L. Fork, M. C. Downer, D. A. B. Miller, D. S. Chemla, C. V. Shank, A. C. Gossard, and W. Wiegmann, Phys. Rev. Lett. 54 , 1306 (1985). DOI:10.1103/PhysRevLett.54.1306
[47] J. Feldmann, G. Peter, E. O. Göbel, P. Dawson, K. Moore, C. Foxon, and R. J. Elliott, Phys. Rev. Lett. 59 , 2337 (1987). DOI:10.1103/PhysRevLett.59.2337
[48] L. C. Andreani, F. Tassone, and F. Bassani, Solid State Commun. 77 , 641 (1991). DOI:10.1016/0038-1098(91)90761-J
[49] B. Deveaud, F. Clérot, N. Roy, K. Satzke, B. Sermage, and D. S. Katzer, Phys. Rev. Lett. 67 , 2355 (1991). DOI:10.1103/PhysRevLett.67.2355
[50] H. M. Zhu, N. H. Song, and T. Q. Lian, J. Am. Chem. Soc. 133 , 8762 (2011). DOI:10.1021/ja202752s
[51] F. García-Santamaría, Y. F. Chen, J. Vela, R. D. Schaller, J. A. Hollingsworth, and V. I. Klimov, Nano Lett. 9 , 3482 (2009). DOI:10.1021/nl901681d
[52] V. I. Klimov, S. A. Ivanov, J. Nanda, M. Achermann, I. Bezel, J. A. McGuire, and A. Piryatinski, Nature 447 , 441 (2007). DOI:10.1038/nature05839
[53] D. Oron, M. Kazes, and U. Banin, Phys. Rev. B 75 , 035330 (2007). DOI:10.1103/PhysRevB.75.035330
[54] K. F. Wu, Q. Y. Li, Y. Y. Jia, J. R. McBride, Z. X. Xie, and T. Q. Lian, ACS Nano 9 , 961 (2015). DOI:10.1021/nn506796m
[55] S. Pedetti, S. Ithurria, H. Heuclin, G. Patriarche, and B. Dubertret, J. Am. Chem. Soc. 136 , 16430 (2014). DOI:10.1021/ja509307m
[56] D. J. Norris, and M. G. Bawendi, Phys. Rev. B 53 , 16338 (1996). DOI:10.1103/PhysRevB.53.16338
[57] C. B. Murray, D. J. Norris, and M. G. Bawendi, J. Am. Chem. Soc. 115 , 8706 (1993). DOI:10.1021/ja00072a025
[58] J. Cui, A. P. Beyler, I. Coropceanu, L. Cleary, T. R. Avila, Y. Chen, J. M. Cordero, S. L. Heathcote, D. K. Harris, O. Chen, J. S. Cao, and M. G. Bawendi, Nano Lett. 16 , 289 (2016). DOI:10.1021/acs.nanolett.5b03790
[59] C. Lin, K. Gong, D. F. Kelley, and A. M. Kelley, ACS Nano 9 , 8131 (2015). DOI:10.1021/acsnano.5b02230
[60] A. M. Kelley, J. Phys. Chem. Lett. 1 , 1296 (2010). DOI:10.1021/jz100123b
[61] E. Groeneveld, C. de Mello Donegá, J. Phys. Chem. C 116 , 16240 (2012). DOI:10.1021/jp303216w
[62] F. T. Rabouw, M. Kamp, R. J. A. van Dijk-Moes, D. R. Gamelin, A. F. Koenderink, A. Meijerink, and D. Vanmaekelbergh, Nano Lett. 15 , 7718 (2015). DOI:10.1021/acs.nanolett.5b03818
[63] F. T. Rabouw, J. C. van der Bok, P. Spinicelli, B. Mahler, M. Nasilowski, S. Pedetti, B. Dubertret, and D. Vanmaekelbergh, Nano Lett. 16 , 2047 (2016). DOI:10.1021/acs.nanolett.6b00053
[64] S. G. Cloutier, P. A. Kossyrev, and J. Xu, Nat. Mater. 4 , 887 (2005). DOI:10.1038/nmat1530
[65] J. I. Fujisawa, and T. Ishihara, Phys. Rev. B 70 , 205330 (2004). DOI:10.1103/PhysRevB.70.205330
[66] J. A. McGuire, M. Sykora, I. Robel, L. A. Padilha, J. Joo, J. M. Pietryga, and V. I. Klimov, ACS Nano 4 , 6087 (2010). DOI:10.1021/nn1016296
[67] T. Franzl, J. Mller, T. A. Klar, A. L. Rogach, and J. Feldmann, J. Phys. Chem. C 111 , 2974 (2007). DOI:10.1021/jp067166s
[68] T. D. Krauss, S. O'Brien, and L. E. Brus, J. Phys. Chem. B 105 , 1725 (2001). DOI:10.1021/jp0023415
[69] J. A. McGuire, M. Sykora, J. Joo, J. M. Pietryga, and V. I. Klimov, Nano Lett. 10 , 2049 (2010). DOI:10.1021/nl100177c
[70] W. K. Bae, L. A. Padilha, Y. S. Park, H. McDaniel, I. Robel, J. M. Pietryga, and V. I. Klimov, ACS Nano 7 , 3411 (2013). DOI:10.1021/nn4002825
[71] K. F. Wu, J. Lim, and V. I. Klimov, ACS Nano 11 , 8437 (2017). DOI:10.1021/acsnano.7b04079
[72] Y. S. Park, W. K. Bae, J. M. Pietryga, and V. I. Klimov, ACS Nano 8 , 7288 (2014). DOI:10.1021/nn5023473
[73] M. Nasilowski, P. Spinicelli, G. Patriarche, and B. Dubertret, Nano Lett. 15 , 3953 (2015). DOI:10.1021/acs.nanolett.5b00838
[74] C. Javaux, B. Mahler, B. Dubertret, A. Shabaev, A. V. Rodina, A. L. Efros, D. R. Yakovlev, F. Liu, M. Bayer, G. Camps, L. Biadala, S. Buil, X. Quelin, and J. P. Hermier, Nat. Nanotechnol. 8 , 206 (2013). DOI:10.1038/nnano.2012.260
CdSe/CdTe Ⅱ型胶态量子阱中长寿命的单激子、带电激子和双激子
王俊慧, 梁桂杰, 吴凯丰     
中国科学院大连化学物理研究所, 分子反应动力学国家重点实验室, 能源材料化学协同创新中心, 大连 116023
摘要: 在本文中,我们研究CdSe/CdTe胶态量子阱或者纳米片里面的单激子、带电激子和双激子态.我们发现,CdSe/CdTe纳米片中Ⅱ型的能带排布能够在空间上有效地分离电子和空穴,导致单激子寿命长达340 ns.这比单纯的CdSe纳米片中的单激子寿命长将近两个数量级.更重要的是,电子空穴的分离也显著地减慢了俄歇衰退的速率,导致带电激子寿命长达70 ns和双激子寿命长达11 ns,高于以往报导的所有胶态纳米晶材料.这些长寿命的单激子、带电激子和双激子态,再加上二维材料本征上很强的吸光能力,使得本文报导的CdSe/CdTe Ⅱ型纳米片成为太阳能到燃料转换反应中极具前景的捕光材料.
关键词: 太阳能    激发态寿命    激子    带电激子    双激子    Ⅱ型量子阱    纳米片    俄歇复合