Charge Transfer from n-doped Nanocrystals: Mimicking Intermediate Events in Multielectron Photocatalysis

Transcription

1 Supporting Information for: Charge Transfer from n-doped Nanocrystals: Mimicking Intermediate Events in Multielectron Photocatalysis Junhui Wang, Tao Ding 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 * Corresponding Author: kwu@dicp.ac.cn Content list: Figures S1-S9 Sample preparations TA experiment set-ups Estimation of the number of attached PTZ molecules per QD Kinetics fitting models S1

2 Figure S1. (a) TEM image, (b) high-resolution TEM image, and (c) size distribution histogram of the CdS QDs used in this study. S2

3 Figure S2. TA spectra of CdS QDs probed at indicated time delays following various excitation fluences from 140 to 7000 μj/cm 2. S3

4 Abs (a.u.) XB Signal (Normalized) TA Signal a ps fits w Flux (J/cm 2 ) 0 b Delay (ps) 0.1 subtration fit Delay (ps) Figure S3. a) TA kinetics probed at the XB of CdS QDs excited with various pump fluences from 28 μj/cm 2 to 7000 μj/cm 2. They are rescaled to set the saturated signal amplitude at 800 ps to 1. Inset shows the plot of the rescaled signal amplitude as a function of the excitation fluence (blue squares) and its fit to a Poisson statistics model (black solid line). According to this model, the photon absorption events when using above band-gap excitation follows a Poisson distribution, 1 and hence, the XB signal at a delay time of 800 ps can be expressed as: I 1 P(0) = 1 e -<N>, where <N> is the average number of photons absorbed per QD or the average exciton number per QD prior to Auger recombination. <N> is proportional to the pump laser fluence: <N> = Cj. Therefore, the XB signal at 800 ps can be fitted with: I 1 S4

5 e -Cj, with C as the only fitting parameter. Using the fitted C value, the average exciton number <N> at each pump fluence can be calculated (blue solid line in the inset). b) Biexciton recombination kinetics (blue open circles) obtained by performing a subtraction between XB kinetics with <N> ~ 0.03 and <N> ~ The black solid line is a single-exponential fit with a time constant of 224±18 ps, which is ~6 fold faster than the X - Auger lifetime (Fig. 2b). According to a previously-established superposition principle (statistical arguments) 2-4, the Auger lifetime of XX is related to those of X - and X + via: 1 τ XX,A = 2 τ X+,A + 2 τ X,A. Also, it has been shown that for II-VI group QDs, because of the asymmetry between conduction and valence band structures, Auger recombination of the X- channel is considerably slower than that of the X + channel. 4 On the basis of these considerations, X - Auger lifetime is at least 4-fold longer than the XX Auger lifetime, which is consistent with our measured results. S5

6 Figure S4. a) Absorption spectra of CdS QDs (blue) and CdS QD-BQ complexes (green) dispersed in hexane. b) TA spectra of CdS QD-BQ complexes probed at indicated delays following the excitation by a 400 pump pulse. S6

7 Figure S5. TA spectra of (a) n-doped-3 and (b) neutral CdS QDs probed at indicated delays following the excitation by a 400 pump pulse. The neutral sample was prepared by exposing the n-doped sample to the air, which released the doped electrons in a few seconds. S7

8 Figure S6. TA kinetics probed at the XB feature of n-doped-2 and 3 samples (light blue) and neutral (blue) CdS QDs. Insets show that the negative trion (X - ) recombination kinetics (blue squares) obtained by performing a subtraction between the two kinetics in the main panels, and their single-exponential fits (black solid line). The fits in (a) and (b) give single-exponential time constants of 1.34±0.14 ns and 1.31±0.07 ns, repectively. S8

9 Figure S7. a) Absorption spectra of CdS QDs (blue), n-doped (light red) and neutral (red) CdS QD-PTZ complexes dispersed in hexane. The doping process did not affect the absorption of PTZ except for introducing ~9% bleach of the 1S absorption of the CdS QDs, indicating that ~18% of QDs are doped with one electron and the rest QDs remain undoped. The doping level is controlled to be low such that the influence from doubly or even more heavily doped QDs can be excluded. b) TA spectra of neutral CdS QD-PTZ complexes probed at indicated delays following the excitation by a 400 pump pulse. The sample was prepared by exposing the n-doped sample to the air. Inset shows the formation of PTZ + signal as a result of interfacial hole transfer from CdS to PTZ. S9

10 Figure S8. a,b) Expanded view of the spectral range within nm which shows the PTZ + signal and the photoinduced absorption (PIA) signal for n-doped (a) and neutral (b) CdS QD-PTZ complexes. c,d) Kinetics plotted at selected wavelengths of the PIA signal for n-doped (c) and neutral (d) CdS QD-PTZ complexes. It can be seen from (a) and (b) that the PIA signal amplitude is relatively constant in a very broad wavelength range and from (c) and (d) that the PIA kinetics is wavelength-independent. Therefore, to obtain pure PTZ + kinetics, we assume that the PIA kinetics overlapped with PTZ + is the same as those in the longer wavelength range and simply subtract the PIA kinetics averaged between nm (scaled by a factor to ensure that the subtraction yields zero background before time zero) from the kinetics probed within nm. S10

11 Figure S9. Comparison between the knitics shown in the insets of Fig. 2b and Fig. 3c in the main text. The former reflects electron decay due to the recombination of X - whereas the latter reflects not only X - recombination kinetics but also the hole transfer kinetics. S11

12 Sample preparations Synthesis of CdS QDs g Cadmium Oxide (CdO), 7.5 ml Oleic acid (OA) and 15 ml 1-octadecene (ODE) were degassed under vacuum for 10 min at 90 C. The solution was then heated to 270 C in 25 min under argon until complete dissolution of CdO. After the temperature was stabilized, 3 ml of sulfur stock solution (0.1 M of S in ODE) was swiftly injected. The reaction was stopped after 30 s by injecting 7 ml room-temperature ODE and removing the heating mantle. The CdS QDs were purified several times by using ethanol and finally dispersed in toluene for use. The as-prepared CdS QDs have the first exciton peak at 450 nm. Preparation of CdS-BQ and CdS-PTZ complexes. Benzoquinone (BQ) and phenothiazine (PTZ) molecules were purchased from Aldrich and were used without further purification. The CdSe-BQ complex and CdSe-PTZ complex were separately prepared by adding BQ and PTZ powders, respectively, to CdSe QD solution in hexane followed by sonication and filtration to remove undissolved BQ and PTZ. Photochemical n-doping of QDs. In the photochemical doping experiments, LiEt 3 BH (1 M solution in tetrahydrofuran) was diluted to 0.01 M with toluene. The CdS QDs solution had an optical density of ~0.86 at the band-edge peak. This solution was transferred into a glove box with a N 2 atmosphere (oxygen level <0.1 ppm). The diluted LiEt 3 BH solution was added into the CdS-QD solution under vigorous stirring and under room light. The solution was then illuminated by a UV lamp to accelerate the reaction until it reached an equilibrium charged state. For successful photochemical doping, only 10-to-100 equivalents of LiEt 3 BH per CdS was required, consistent with previous reports. The degree of n-doping could be S12

13 controlled by varying the amount of LiEt 3 BH used in the reaction. To measure the negative trion lifetime, we chose an n-doped QD solution that yielded a single-exponential component for the charged species using the procedure introduced in Fig. 2b. The n-doped QD solution was sealed in a custom-made airtight cuvette (optical path 1 mm) and transferred out of the glove box for all optical measurements. The neutralized solution was obtained by exposing the n-doped QD solution to the air for a few minutes. For doping the CdS-PTZ complex, the whole procedure was exactly the same except for the illumination light (Vis-405 nm lamp), which could avoid the excitation of PTZ molecules. TA experiment set-ups The femtosecond pump-probe TA measurements were performed using a regenerative amplified Ti:sapphire laser system (Coherent; 800 nm, 70 fs, 6 mj/pulse, and 1 khz repetition rate) as the laser source and a femto-ta100 spectrometer (Time-Tech Spectra). Briefly, the 800 nm output pulse from the regenerative amplifier was split in two parts with a 50% beam splitter. The transmitted part was used to pump a TOPAS Optical Parametric Amplifier (OPA) which generated a wavelength-tunable laser pulse from 250 nm to 2.5 μm as pump beam. The reflected 800 nm beam was split again into two parts. One part with less than 10% was attenuated with a neutral density filter and focused into a 2 mm thick sapphire window to generate a white light continuum (WLC) used for probe beam. The probe beam was focused with an Al parabolic reflector onto the sample. After the sample, the probe beam was collimated and then focused into a fiber-coupled spectrometer with CMOS sensors and detected at a frequency of 1 KHz. The intensity of the pump pulse used in the experiment was S13

14 controlled by a variable neutral-density filter wheel. The delay between the pump and probe pulses was controlled by a motorized delay stage. The pump pulses were chopped by a synchronized chopper at 500 Hz and the absorbance change was calculated with two adjacent probe pulses (pump-blocked and pump-unblocked). The samples were placed in 1 mm airtight cuvettes in a N 2 -filled glove box and measured under ambient conditions. Samples were vigorously stirred in all the measurements. Estimation of the number of attached PTZ molecules per QD According to the absorption spectra in Fig. S6a, the absorbance of CdS QDs at its first excitonic peak (~450 nm) is ~1.4. Using an extinction coefficient of ~ reported for CdS QDs of this size, the concentration of QDs in the hexane solution is ~ M. The absorbance of PTZ in CdS QD-PTZ complexes can be obtained by subtraction of QD absorption from that of the complexes, which results in an optical density of ~0.7 at 350 nm. Using the PTZ extinction coefficient of ~1900 M -1 cm -1 at 350 nm, the concentration of PTZ can be calculated to be M. Therefore, the mole ratio between CdS QDs and PTZ in the solution is ~1:2470. However, as PTZ is slightly soluble in hexane, we need to account for the equilibrium between PTZ molecules dissolved in hexane and those bound to the QD surfaces. Based on previous studies of QD-molecule systems, the adsorption of PTZ molecules onto CdS QDs should follow the Langmuir adsorption isotherm. 5 θ = m N sites = θ max K a [PTZ] 1+K a [PTZ] (S1), where θ stands for the mean fractional surface coverage of PTZ on QDs, m is the number of S14

15 PTZ molecules adsorbed per QD, N sites is the total number of available adsorption sites per QD, θ max is the maximum fractional surface coverage of PTZ on QDs, K a is the binding constant of PTZ on QDs, and [PTZ] is the concentration of PTZ. According to a previous reports on CdS nanorod (NR)-PTZ complexes in hexane, θ max is ~1.2% and K a is ~250 M -1. Also, N sites for a CdS NR with rod length of ~22 nm and diameter of ~3.6 nm is ~3700. Assuming that the number of binding sites is proportional to the surface area of the nanocrystals, N sites is estimated to be ~1000 for our QDs. Using these parameter values, we estimate that m, the number of PTZ molecules adsorbed per QD in our sample, is ~5.8. Kinetics fitting models According to the scheme in Fig. 3a, the photoexcited, n-doped CdS-PTZ complexes (QD -* -PTZ) can decay either via Auger-dominated trion recombination to from ground state n-doped CdS-PTZ complexes (QD - -PTZ), with a rate of k AR or via hole transfer (HT) to from charge separated states with two electrons in QD and one hole on PTZ (QD 2- -PTZ + ), with a rate of k HT ; QD 2- -PTZ + eventually decays through interfacial charge recombination to from QD - -PTZ, with a rate of k CR. We write the following kinetic equations to account for these processes: * d QD PTZ dt * kar kht QD PTZ (S2), d QD PTZ * 2 k AR QD PTZ k CR QD PTZ dt (S3), 2 d QD PTZ * 2 k HT QD PTZ k CR QD PTZ dt (S4), where [QD -* -PTZ], [QD - -PTZ], and [QD 2- -PTZ + ] stand for the concentrations of QD -* -PTZ, S15

16 QD - -PTZ, and QD 2- -PTZ + species. The solution for this set of equations is: * kar kht t QD PTZ C0 e (S5), kar kar kht t kht kcrt QD PTZ C 0 1 e 1 e kar k HT kar kht k 2 HT kcrt kar kht t QD PTZ C0 e e kar kht (S6), (S7), where C 0 is the initial concentration of the photoexcited QDs. The XB spectral feature of CdS QDs are contributed by both QD -* -PTZ and QD 2- -PTZ + species, and hence, the kinetics of XB can be expressed as: k S t e e e XB, QD k k k t kar kht t HT CR kar kht t AR HT (S8). The PTZ + spectral feature is only contributed by QD 2- -PTZ + species and its kinetics can be expressed as: k HT kcrt kar kht t S t e e PTZ, QD kar k HT (S9). For neutral CdS QD-PTZ complexes, the three species involved in the kinetic processes are photoexcited CdS QD-PTZ complexes (QD * -PTZ), ground state complexes (QD-PTZ), and charge separated states (QD - -PTZ + ). The time-dependent concentrations of these species can be obtained by replacing the trion recombination rate k AR with a single exciton recombination rate k X : * kx kht t QD PTZ C0 e (S10), k k k k k k X k QD PTZ X k HT t HT k CR t C e e X HT X HT k X HT QD PTZ C e e HT kcrt k k t 0 kx k HT (S11), (S12). S16

17 The kinetics of XB and PTZ + features are: k SXB, QD t e e e k k k t kx kht t HT CR kx kht t k HT kcrt kx kht t S t e e PTZ, QD kx k HT X HT (S13), (S14). In reality, however, all the kinetics processes described above are not single-exponential but rather multi-exponential. Yet another complication is from the existence of a subset of undoped QDs in the doped sample. To simply the analysis, we scale the XB kinetic traces in the n-doped and neutral samples to the long-lived tail (see Fig. 3c) which represents the contribution of undoped QDs in the n-doped sample. Thus, a subtraction between these two traces should yield the XB kinetics for pure n-doped QDs, which can be fitted using a modified eq. S8 which accounts for the heterogeneity of hole transfer process: 3 3 kar kht, i t kht, i kcrt kar kht, i t S t A XB, QD ie Ai e e k k (S15), i1 i1 AR HT, i where k HT,i and A i are the hole transfer rate constants and their corresponding amplitudes. Note that the charge recombination process should also be heterogeneous and hence be described by multi-exponential functions, but we simply used a single-exponential function because the slow processes are beyond the time window of our measurements (8 ns). As for the PTZ + kinetics, the scaling procedure used for the XB kinetics does not apply as the charge recombination in the n-doped sample is significantly faster than the undoped sample. To fit the kinetics of PTZ + in the n-doped sample, we need to account for the contribution from both doped and un-doped QDs. Base on the absorption spectra in Fig. S6a, the 1S exciton absorption is bleached by 9% in the n-doped sample. If we use a bimodal S17

18 distribution for the doped electrons in the QD ensemble, this means 18% of the QDs are doped with one electron and 82% of the QDs are undoped. The latter is exactly the factor used to scale the XB kinetics in the neutral sample for it to match that in the n-doped sample (Fig. 3a), confirming that the doped electrons indeed follow a bimodal distribution. A previous work reported a Poisson distribution of doped electrons in thick-shell type-i QDs. This difference can be rationalized considering the behavior of inter-dot charge transfer in the QD solution. For our core-only QDs, this charge transfer should readily occur and therefore results in bimodal-distributed electrons. For the thick-shell type-i QDs, inter-dot charge transfer might be difficult, and hence, the distribution of electrons maintains its initial state which is most likely a Poissonian. With the percentages of doped and undoped QDs in the n-doped sample resolved, we can fit its PTZ + kinetics using a linear combination of eq. S9 and eq. S14: 1 2 S t C S t C S t (S16), PTZ PTZ, QD PTZ, QD where C1 and C2 are the percentages of undoped and doped QDs in the sample. Note that the expressions for S t and S t PTZ, QD PTZ, QD are modified from eq. S9 and eq. S14 by using three-exponential functions to represent the hole transfer process, as used in eq. S15. With the kinetic models established, the kinetics in the inset of Fig. 3c is fitted using eq. S15 and the kinetics for the neutral and n-doped samples in Fig. 3d are fitted using eq. S14 (also modified by using three-exponentials) and eq. S16, respectively. The fitting parameters are tabulated in Table S1; note that the rate constants in the above equations are converted to time constants in the table. The averaged hole transfer time (τ HT,ave ) are calculated using the following expression: S18

19 HT, ave A k 3 i (S17), i1 HT, i The charge separation yields (Φ CS ) for the neutral and n-doped CdS-PTZ complexes can be calculated using the following expressions: CS, QD CS, QD k 3 HT, i Ai (S18), i1 k X k HT, i k 3 HT, i Ai (S19). i1 k AR k HT, i Table S1. Charge separation and recombination time constants and yields. τ HT,1 (ps) τ HT,2 (ps) τ HT,3 (ps) τ HT,ave (ps) τ CR (ns) Φ CS (%) A 1 (%) A 2 (%) A 3 (%) neutral QDs (32.5) (25.5) (41.2) n-doped QDs (32.5) (25.5) (41.2) References for SI: 1. 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, Wu, K.; Lim, J.; Klimov, V. I. Superposition Principle in Auger Recombination of Charged and Neutral Multicarrier States in Semiconductor Quantum Dots. ACS Nano 2017, S19

20 11, Park, Y.-S.; Bae, W. K.; Pietryga, J. M.; Klimov, V. I. Auger Recombination of Biexcitons and Negative and Positive Trions in Individual Quantum Dots. ACS Nano 2014, 8, Klimov, V. I. Multicarrier Interactions in Semiconductor Nanocrystals in Relation to the Phenomena of Auger Recombination and Carrier Multiplication. Annual Review of Condensed Matter Physics 2014, 5, Morris-Cohen, A. J.; Frederick, M. T.; Cass, L. C.; Weiss, E. A. Simultaneous Determination of the Adsorption Constant and the Photoinduced Electron Transfer Rate for a Cds Quantum Dot Viologen Complex. J. Am. Chem. Soc. 2011, 133, S20