Green and far-red-light induced electron injection from perylene bisimide to wide bandgap semiconductor nanocrystals with stepwise two-photon absorption process

Daisuke Yoshioka , Daiki Fukuda and Yoichi Kobayashi *
Department of Applied Chemistry, College of Life Sciences, Ritsumeikan University, 1-1-1 Nojihigashi, Kusatsu, Shiga 525-8577, Japan. E-mail: ykobayas@fc.ritsumei.ac.jp

Received 30th November 2020 , Accepted 6th January 2021

First published on 6th January 2021


Abstract

Stepwise two-photon absorption (2PA) processes are becoming an important technique because it can achieve high reductive photochemical reactions with visible and near infrared light and intensity-gated high spatiotemporal selectivity with much lower power thresholds than those of the simultaneous 2PA. However, excited states generated by stepwise 2PA (higher excited states and excited states of transient species) are so short-lived that the efficiency for the stepwise 2PA induced photochemical reactions is usually quite low, which limits the versatility for this technique. Here, we demonstrated that the electron of the higher excited state can be efficiently extracted in a nanohybrid of organic molecules and wide bandgap semiconductor nanocrystals (NCs). Using perylene bisimide (PBI)-coordinated CdS NCs as a model compound, we demonstrated that the electron of the higher excited state of PBI generated by stepwise 2PA can be extracted to the conduction band of CdS NCs with a quantum yield of ∼0.5–0.7. Moreover, the extracted electron survives at the conduction band of CdS NCs over nanoseconds, which is more than hundred times longer than the lifetime of the S2 state of PBI. This method can be applied to other organic molecules and larger wide bandgap semiconductors, and therefore, will expand the versatility for the photochemical reactions utilizing the short-lived excited states.


Introduction

Utilization of visible or near-infrared (NIR) light to induce high-energy photochemical reactions is an important method for site-selective photochemical reactions in condensed systems.1–4 Among several techniques to realize these reactions, there has been an increasing amount of interest in stepwise two-photon absorption (2PA) processes, where two photons are sequentially absorbed via an actual intermediate electronic state.5 While the power threshold of simultaneous 2PA is typically larger than GW cm−2, that of stepwise 2PA is much lower than that of simultaneous 2PA depending on the lifetime of the intermediate state. For example, the power threshold of the stepwise two-photon induced photochromic reaction of bis(bridged imidazole dimer), which is a nonlinear T-type photochromic compound, is at most 10 mW cm−2 and the stepwise 2PA process can be induced by continuous wave (CW) light.6 Stepwise 2PA has been widely used for more than six decades.5 For examples, 2PA processes using the S1 state as an intermediate state have been applied to photoionizations,7,8 optical switching,2,9–13 and so on,14,15 while those using the T1 state as an intermediate state have been utilized in triplet–triplet annihilation (TTA) up conversions,16 photoacid generations,17,18 reverse saturable absorbers,19 and so on.20–22 Moreover, stepwise 2PA processes using a transient photoproduct as an intermediate state have been applied to intensity-dependent photochromic materials23–27 and holographic recording materials.28,29

More recently, stepwise 2PA has been extensively studied in organic photocatalysis.4,30–33 In this research field, stepwise 2PA is often combined with intermolecular electron transfers and called “consecutive electron transfers”. For example, the excited states of photogenerated long-lived radical anions of perylene bisimide (PBI) and naphthalene diimide (NDI) have been applied to visible-light-induced photocatalysis,31,34 hydrogen evolutions,30 and the reductions of carbon dioxide.32 However, one of the serious drawbacks of the excited states of the radical anions is a very short lifetime (tens-of-ps to ∼100 ps),35–38 which makes the intermolecular photoreactions much inefficient. Moreover, molecular frameworks which generate long-lived and photostable radical anions are limited. On the other hand, high redox potentials can be also achieved by higher electronic excited states (Sn and Tn states) produced by the stepwise 2PA. However, the same serious problem emerges, i.e., the lifetimes of higher excited states are usually less than 10 ps, which is too short to apply them to intermolecular photochemical reactions. Therefore, strategies to extract the electrons or energies from the short-lived higher excited states are necessary for advancing fundamentals and applications of nonlinear photofunctional materials based on stepwise 2PA.

In this study, we propose a strategy to extract high reductive electrons of the higher excited state and extend their lifetimes by transferring the electron to the conduction band of wide bandgap semiconductor nanocrystals (NCs). As a model system, we synthesized PBI-coordinated colloidal CdS NCs (PBI-CdS, Fig. 1). PBI was selected because it has intense absorption and emission bands in the visible region and has high photostability. CdS was used because it is one of the wide bandgap semiconductors, and CdS NCs have large absorption coefficients (105–106 M−1 cm−1) in the visible region. The intense absorption band gives a huge transient signal once an electron is injected to the conduction band of CdS NCs, which can be used as an efficient probe for electron transfer.39


image file: d0nr08493j-f1.tif
Fig. 1 (a) Schematic energy diagram of the visible and far-red light induced stepwise two-photon-induced electron injection of PBI-CdS and (b) a molecular structure of a PBI. VBM and CBM indicate the valence band maximum and conduction band minimum, respectively. This scheme shows the concept to capture the electron of the higher excited state of PBI and extend the lifetime of the excited electron over nanoseconds by CdS NCs.

Fig. 1a summarizes the concept of this study. PBI is excited by 520 nm green light and the S1 state is generated. It is noted that the S1 state of PBI has a characteristic absorption band at ∼705 nm corresponding to the transition of the excited electron to the higher excited state as explained later. When the S1 state absorbs another photon during the lifetime of the S1 state (∼ns), the second excited state (tentatively denoted as the S2 state) is formed. Because the S2 state rapidly relaxes to the S1 state via nonradiative relaxations (∼10 ps in this case as explained later), it is usually difficult to apply the higher excited state to induce intermolecular photochemical reactions. However, we demonstrate that the excited electron of the S2 state of PBI can be efficiently extracted by colloidal CdS NCs within ∼70 fs (instrumental response function: IRF). The extracted electron to the conduction band minimum (CBM) survives over nanoseconds and eventually a subsequent electron transfer occurs to other neutral PBI (or the radical cation of PBI) coordinated to the same NC. The ultrafast electron transfer process from the higher excited states of PBI to CdS NCs were directly observed by transient absorption spectroscopy with two excitation pulses: double pulse transient absorption spectroscopy (often called pump-repump-probe or pump-dump-probe spectroscopy).40,41 This technique is a powerful tool for revealing higher excited states and has been used to reveal charge separation in organic photovoltaic,40 hot electron transfer in semiconductor nanomaterials,41 and photochemical processes of intensity dependent photofunctional materials.6 This technique enables us to precisely evaluate hot electron transfers and even subsequent electron transfer processes between the CdS NCs and coordinated PBI.

Experimental section

Materials

All reagents were purchased from Tokyo Chemical Industry, Sigma-Aldrich, and Fujifilm Wako and were used without further purification.

Syntheses

Synthesis of the CdS NCs was based on the reported procedure by Li et al.42 For a typical example, 520 mg (4 mmol) of CdO, 13 mL of oleic acid, and 18 mL of 1-octadecene (ODE) were mixed in a three-necked flask. The reaction mixture was degassed under vacuum at 110 °C for 1 hour. The mixture solution was stirred and heated to 260 °C under nitrogen conditions, until the solution turned to clear and colourless. Then, a sulfur solution, where 69 mg (2.1 mmol) of the sulfur powder was dissolved in 6 mL ODE at 50 °C, was injected to the Cd solution at 240 °C. The temperature of the mixed solution dropped to 180 °C. The solution was stirred for 30 min (for 3 min in the case of another CdS NCs), and then cooled to room temperature, yielding a pale-yellow solution. The NCs solution was extracted by the mixed solution of methanol, hexane (1[thin space (1/6-em)]:[thin space (1/6-em)]0.7, v/v) and a few drops of butylamine. The organic fraction was collected and extracted again with methanol and hexane (1[thin space (1/6-em)]:[thin space (1/6-em)]2, v/v). The organic fraction was then precipitated using excess acetone. The precipitates of CdS NCs were redispersed in chloroform. The concentration of the CdS NCs was estimated from the first excitonic absorption peak as reported previously.42

A carboxy-group-substituted perylene bisimide derivative (denoted as PBI, Fig. 1b) was synthesized with 2 steps followed by similar procedures reported previously.43–45 The compound was purified by silica gel column chromatography and was characterized by proton nuclear magnetic resonance (1H NMR) spectroscopy and mass spectrometry. The detail is shown in the electronic ESI.

PBI-coordinated CdS NCs (PBI-CdS) were prepared by mixing the 0.5 mL of the CdS chloroform solution (4.7 × 10−6 M) and 0.3 mL of the PBI chloroform solution (7.1 × 10−5 M) at room temperature. The molar ratio of PBI per CdS NC (PBI/CdS NC) was set to 9.

Steady-state and characterization measurements

1H NMR spectra were recorded at 400 MHz by JNM-ECS 400 MHz (JEOL). Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) were recorded on Axima-CFRplus (Shimadzu). For steady-state optical measurements, UV3600 (Shimadzu) and FP-6500 (Jasco) were used for absorption and emission measurements, respectively. Ultima IV (Rigaku) was used for X-ray diffraction (XRD) measurements. TEM-2100Plus (JEOL) was used for transmission electron microscopy (TEM) measurements.

Femtosecond transient absorption measurements

Transient absorption measurements were conducted by a homemade pump–probe system. An amplified femtosecond laser, Spirit One 1040-8 (Spectra-Physics, 1040 nm, pulse width: ∼270 fs), was split into two beams with a ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]9. The weaker beam was focused to a deuterated water placed in a 10 mm quartz cuvette to generate the white light continuum for the probe beam. The stronger beam was directed to a noncollinear optical parametric amplifier (NOPA), Spirit-NOPA-3H (Spectra-Physics) to generate the 360 and 705 nm femtosecond laser pulses for the pump beam. These pulses were used for standard transient absorption spectroscopy. In addition, these pulses were used as the second excitation pulse for double-pulse transient absorption spectroscopy. A residual 520 nm laser pulse inside NOPA was directed to another output port and was used as the first excitation pulse. The time delay between the first 520 nm pulse and the second 705 nm pulse was fixed to 19.6 ps, where no coherent events induced by the first pulse is observed and the population of the S1 state created by the first pulse does not considerably decrease. The 705 nm pump beam was chopped at 500 Hz prior to the sample for signal differencing, while the 520 nm light was not chopped (1 kHz). In this condition, the obtained ΔΔOD is described as follow:
ΔΔOD = ΔOD520+705 − ΔOD520
where ΔOD520+705 and ΔOD520 represent ΔOD under double pulse excitation and a 520 nm pulse excitation, respectively. Both pump and probe beams were focused to the sample solution placed in the 2 mm quartz cuvette. The first 520 nm excitation beam was focused to a spot with the full width at the half maximum (FWHM) of 420 μm, while those of the 705 nm and 360 nm pulses were 100 and 70 μm, respectively. The spot size was measured by knife-edge method assuming a Gaussian laser beam. The polarization between the pump and probe pulses was set at magic angle. The transmitted probe beam was detected with multichannel detection system, PK120-C-RK (UNISOKU), composed of a CMOS linear image sensor and a spectrograph. The obtained spectra were calibrated for group velocity dispersion using the data obtained by the optical Kerr signal of chloroform between the pump pulse and the white-light continuum. The IRF was shorter than approximately 70 fs for 705 nm laser pulse and ∼250 fs and 520 nm laser pulse. The sample solution was degassed and stirred with a stirrer during the experiments and the measurements were performed at room temperature.

Results and discussion

XRD measurements revealed that the CdS NCs have a zinc blende structure (Fig. S2). Homogeneous and spherical colloidal CdS NCs with an average diameter of 4.7 ± 0.3 nm were confirmed by TEM measurements (inset of Fig. 2 and Fig. S4). CdS NCs have a sharp absorption band at 441 nm assigned to the first excitonic absorption band. The molar absorption coefficient at 441 nm was estimated to be 8.30 × 105 M−1 cm−1 from the first excitonic absorption peak followed by a previous report.46 It is noted that the average diameter estimated by the absorption spectra (4.89 nm) is consistent with average diameters measured by TEM images. A synthesized PBI has almost the same absorption and emission spectra of conventional PBI derivatives, namely, sharp absorption bands at 458, 488, and 524 nm, and emission bands at 534 and 576 nm upon excitation at 480 nm. The relative emission quantum yield of PBI in chloroform is 0.88 upon excitation with 480 nm (rhodamine 6G was used as the reference47).
image file: d0nr08493j-f2.tif
Fig. 2 Steady-state absorption (solid) and emission (dashed) spectra of PBI, CdS NCs, and PBI-CdS (the molar ratio: PBI/CdS = 9) in chloroform at room temperature. The excitation wavelength was 480 nm. The inset shows a TEM image of CdS NCs.

The absorption spectrum of PBI-CdS (the molar ratio: PBI/CdS NC = 9) in chloroform was very similar to the superposition of the absorption spectra of PBI and CdS NCs (Fig. 2). However, the absorption and emission bands of the PBI-CdS were slightly shifted to the longer wavelength, and the emission intensity slightly decreased. To investigate the association process in detail, the concentration dependence on the spectral shift was measured under the fixed PBI concentration in chloroform (7.6 × 10−7 M, Fig. S8). The spectral shift was observed even at the very low concentration of the CdS NCs (2.5 × 10−9 M) and the shift ceased at ∼2.0 × 10−8 M. Because the spectral width does not change during the spectral shift and the observed transient absorption spectra of PBI-CdS (shown later) do not show any individual PBI components, it indicates that most of PBI are coordinated to the surface of CdS NCs.

Fig. 3 shows the femtosecond to nanosecond transient absorption spectra of PBI and PBI-CdS in chloroform excited with a 520 nm (29 nJ pulse−1) or a 360 nm femtosecond laser pulse (25 nJ pulse−1). ΔOD520 and ΔOD360 indicate the ΔOD induced by a 520 nm and a 360 nm laser pulses, respectively. Upon excitation of PBI with a 520 nm pulse (Fig. 3a), the S0–S1 transition is resonantly excited. The ground state bleach and stimulated emission signals instantaneously appeared at 460, 490, 532, and 580 nm, respectively. A sharp excited state absorption band and a broad absorption were also observed at 706 nm and over 750 nm, respectively. The 706 nm absorption resembles to the reported absorption band of the PBI radical anion,37 suggesting that it is assigned to the optical transition of the excited electron to a higher molecular orbital (MO) that leads to ionization. The spectral shape of the transient absorption of PBI does not change with time and gradually decays with a time scale of nanoseconds (Fig. 3d).


image file: d0nr08493j-f3.tif
Fig. 3 Femtosecond to nanosecond transient absorption spectra and dynamics of (a and d) PBI (1.1 × 10−5 M) and (b and e) PBI-CdS (PBI/CdS = 9) in chloroform excited at 520 nm (29 nJ pulse−1) at room temperature. (c and f) transient absorption spectra and dynamics of PBI-CdS (PBI/CdS = 9) in chloroform excited at 360 nm (25 nJ pulse−1) at room temperature. Inset of c is the magnification in the region of 480–950 nm.

Upon excitation of PBI-CdS with a 520 nm pulse, PBI is selectively excited (Fig. 3b). The ground state bleach and stimulated emission bands of PBI were observed as similar to those of individual PBI. A small difference between PBI and PBI-CdS is the weaker stimulated emission of PBI-CdS as compared to that of PBI (Fig. 3b). The excited state absorption band is also very similar to that of PBI. However, the excited state absorption band becomes broader and the peak was slightly shifted to the longer wavelength (715 nm). In addition, PBI-CdS has a slightly larger fast decay component (almost within IRF, ∼300 fs) as compared to PBI (Fig. S13). It may be assigned to the solvent response or a fast nonradiative relaxation from the Franck–Condon state to the ground state. The transient absorption dynamics are also very similar (Fig. 3e). These results indicate that CdS NCs have very few interactions with the S1 state of PBI.

Upon excitation of PBI-CdS with a 360 nm pulse, CdS NCs are almost selectively excited (Fig. 3c). The sharp and intense negative signal was observed at 440 nm, ascribable to the ground state bleach of the first excitonic absorption band of CdS NCs. The initial growth of the ground state bleach indicates the intraband relaxations of the photogenerated electrons and holes to the CBM and the valence band maximum (VBM). The time constant of the rise was approximately 150 fs. The recovery of the ground state bleach was tentatively fitted with the bi-exponential decay function and the time constants were 120 ps and 2.2 ns. Moreover, negative (495 and 532 nm) and positive signals (718, 804, and 873 nm) gradually appeared with time constants of 80 ps and 2.1 ns (inset of Fig. 3c). These positive signals are safely assigned to the absorption bands of the radical anion of PBI because the spectral shape is very similar to the electrochemically generated radical anion of PBI37 and a similar spectral evolution was reported in a previous report on a PBI derivative.48 The negative signals at 493 and 532 nm are assigned to the ground state bleach of PBI. It shows that the electron transfer occurs from the conduction band of CdS NCs to the lowest unoccupied molecular orbital (LUMO) of PBI. The possibility for energy transfer can be excluded because of the absence of the stimulated emission signals at 580 nm. The quantum yield of the electron transfer is calculated to be 0.69 from the amplitudes of the ground state bleach signals at 441 and 532 nm, which is proportional to the concentrations of generated excited states of CdS NCs and the produced radical anion of PBI, respectively (see ESI for details).

To directly observe dynamics of higher excited states, we conducted double pulse transient absorption measurements. Namely, the first excitation pulse was set to 520 nm (400 nJ pulse−1) to produce the S1 state of PBI. The second excitation pulse was set to 705 nm (80 nJ pulse−1) to directly excite the S1 state of PBI. The time delay between two pulses was set to 19.6 ps, where no coherent events induced by the first pulse is observed. As was described in the Experimental section, ΔΔOD was defined as the subtraction of ΔOD under the excitation with both pulses (ΔΔOD520+705) by ΔOD under the excitation with 520 nm pulse. Therefore, ΔΔOD can selectively observe the signal induced only by both pulses. We have confirmed that no signals were observed under the excitation with a single pulse excitation condition.

The bottom of Fig. 4a shows the double-pulse transient absorption spectra of PBI in chloroform at room temperature, while the top of Fig. 4a shows the transient absorption spectrum of PBI excited at 520 nm and probed at 20 ps after the excitation (single pulse excitation condition). Times denoted in the figures indicate the time delay between the second excitation pulse (705 nm) and the probe pulse. Several positive signals were observed at 535, 579, and 628 nm. Because the stimulated emission signals were observed in these spectral regions before the second pulse excitation (top of Fig. 4a), these positive ΔΔOD signals indicate that the stimulated emission signals decrease by the excitation with 705 nm. Moreover, a sharp negative ΔΔOD signal was observed at 706 nm, where the excited state absorption was observed before the second pulse excitation. These results indicate that the second pulse with 705 nm induces the bleach of the excited state absorption and successfully excites the excited electron of PBI to the higher excited states.


image file: d0nr08493j-f4.tif
Fig. 4 (Top) Transient absorption spectra of (a) PBI and (b) PBI-CdS (PBI/CdS = 9) and CdS NCs excited with a single pulse (520 nm for PBI and PBI-CdS and 360 nm for CdS NCs). (Bottom) Early time range (−1 ps to 30 ps) of double pulse transient absorption spectra of (a) PBI and (b) PBI-CdS (PBI/CdS = 9) at different time delays. The first pulse was set to 520 nm (400 nJ pulse−1), while the second pulse was set to 705 nm (80 nJ pulse−1). The time delay between the excitation pulses was fixed to 19.6 ps.

It is noted that ΔΔOD signals originated from the ground state bleach (460–520 nm region) were quite small as compared to those to the stimulated emission. Because the 705 nm pulse selectively excites the excited electron of the S1 state, this result indicates that the ground state bleach signal mostly reflects the ground state bleach (the decrease in the population of the ground state) and the contribution of the state filling of LUMO is small. The transient absorption bands assigned to the S2 state quickly decay with a lifetime of 15 ps. The spectral shape at 30 ps is very similar to the inversion of the transient absorption spectrum excited at 520 nm (top of Fig. 4a). The slower time range (>30 ps) will be discussed later (Fig. 6).

The double-pulse transient absorption spectra of PBI-CdS are substantially different from those of individual PBI (bottom of Fig. 4b). In addition to the signals observed in PBI (bottom of Fig. 4a), other negative and positive sharp signals were observed at 440 and 465 nm instantaneously after the second pulse excitation, respectively. The negative signal is ascribable to the ground state bleach of the first excitonic absorption of CdS NCs as shown in the top of Fig. 4b. Moreover, it is known that the initial quick decay of the excited state absorption of CdS NCs (463 nm) at the early time scale (0.3 ps) shown in the top of Fig. 4b indicates that the intraband relaxation of carriers (mainly electron in CdS) from the higher excited states.49,50 The positive signal of PBI-CdS at 463 nm quickly decayed within hundreds of fs. This result suggests that the electron at the S2 state of the PBI was firstly transferred to the higher excited state of CdS NCs instantaneously, and then quickly relaxes to the CBM within hundreds of fs. We have confirmed that no signals were observed upon excitation with only a 520 nm pulse or a 705 nm pulse, indicating that the bleach signal is not due to the simultaneous 2PA process. In addition to the sharp negative and positive signals, a broad positive absorption band appears to be superposed at 480–600 nm. This signal is also ascribable to the charge transfer state generated by the hot electron transfer.

To further confirm that the observed signals are generated by stepwise 2PA process, we conducted the excitation intensity dependence of both pulses (Fig. S9 and S10). These figures show that both ΔΔOD signals at 440 and 535 nm linearly depend on both excitation intensities. It shows that the observed signals are generated by the absorption of the 520 nm and 705 nm pulses in a stepwise manner.

Fig. 5 shows the double-pulse transient absorption dynamics of PBI-CdS probed at 440 and 540 nm. After the second 705 nm pulse excitation, the bleach signal of CdS NCs at 440 nm instantaneously generated after the nonlinear coherent signals, indicating that the electron transfer from the higher excited state of PBI to the conduction band minimum of CdS NCs occurs within the IRF (<70 fs, inset of Fig. 5). The beach signal at 440 nm slowly decays with a time scale of nanoseconds. On the other hand, the signal at 540 nm has a fast decay with a time constant of 9.8 ps, which corresponds to the relaxation from the S2 to the S1 state (15 ps for individual PBI). Then, the signal at 540 nm gradually decays with the same time constant as that at 440 nm, which indicates that the slower decay process is correlated with the excited electron at the CBM of CdS NCs superposed at 480–600 nm.


image file: d0nr08493j-f5.tif
Fig. 5 Double pulse transient absorption dynamics of PBI-CdS (PBI/CdS = 9) in chloroform probed at 440 and 540 nm. Thick lines are fitting lines by global analyses. Inset shows the time profile at 440 nm at early time scales.

Fig. 6a shows the slower time range (30 ps to 2500 ps) of the double-pulse transient absorption spectra of PBI. After the decay with a time scale of tens of ps, the long-lived residual components remained over nanoseconds. This decay process is too long to be assigned to the relaxation process from higher excited states. The quantum yield for the residual component was calculated to approximately 0.09 from the decay at 535 nm the residual component. On the other hands, the transient absorption and steady-state absorption signals did not degrade at all under repeated intense double pulse excitations over several hours. Therefore, it indicates that the residual signal is not due to irreversible biproducts. One of the possibilities for this component may be due to the nonradiative relaxation from the S2 state to the S0 state. Another possibility might be due to the stimulated emission by the second excitation pulse (705 nm), although it is less conceivable because the wavelength of the second excitation pulse is different from the spectral range of the emission spectrum.


image file: d0nr08493j-f6.tif
Fig. 6 (Top) Transient absorption spectra of (a) PBI and (b) PBI-CdS (PBI/CdS = 9) and CdS NCs excited with a single pulse (520 nm for PBI and PBI-CdS and 360 nm for CdS NCs). (Bottom) Slow time range (30 ps to 2500 ps) of double pulse transient absorption spectra of (a) PBI and (b) PBI-CdS (PBI/CdS = 9) upon excitation with the first 520 nm (400 nJ pulse−1) and the second 705 nm (80 nJ pulse−1) pulses. The time delay between the excitation pulses was fixed to 19.6 ps.

The slower time range (30 ps to 2500 ps) of the double-pulse transient absorption spectra of PBI-CdS are quite different from those of PBI (bottom of Fig. 6b). After the partial decay of the transient absorption bands within tens of ps, other sharp negative signals gradually appeared at 493 and 529 nm, and several sharp positive peaks also appeared at 723, 804, and 872 nm with a time scale of nanoseconds. The generated sharp bands are very similar to the transient absorption spectrum of PBI-CdS at 2500 ps after the excitation with a 360 nm pulse (top of Fig. 6b). Especially, characteristic peaks at ∼715 and 804 nm indicate that the radical anion of PBI was formed by the subsequent electron transfer from the conduction band of CdS NCs to other neutral PBI. The reason for the slight shift of the peak at 715 nm to the longer wavelength (∼725 nm) is most probably due to the superposition of the spectra of the radical anion and S1 states of PBI. The reason why the geminate charge recombination is the minor process is most probably because the number of the neutral PBI is much larger than the radical cation of PBI at the surface of NCs. The quantum efficiency for the electron transfer from the S2 state of PBI to CdS NCs is estimated to be ∼0.5–0.7 by considering the absorption coefficients of the neutral and the radical anion of PBI and the quantum yield of the electron transfer from the conduction band of CdS NCs to the LUMO of PBI (0.69, see ESI for detail). This value is actually much larger than that estimated by the bleach signal of CdS NCs. The bleach signal at 440 nm appears to be much smaller than that excited at 360 nm (Fig. 3c). One possibility may be due to the carrier-induced Stark effect by the generated charge transfer state, which shifts the absorption band depending of the electric fields induced by generated carriers, and the apparent bleach signal may decrease by the superposition of these signals.

We conducted the same experiments with smaller CdS NCs, whose first excitonic absorption peak is 425 nm and the average diameter is 4.3 ± 0.3 nm (Fig S2a). The hot electron transfer from the S2 state was similarly observed in the smaller CdS NCs systems (Fig S18).

Moreover, we also investigated the effect of the molar ratio of PBI and CdS NC on the secondary formation process of the radical anion of PBI. The subsequent electron transfer dynamics from the conduction band of CdS NCs to other neutral PBI become faster with the increase in PBI/CdS NC (Fig. 7). These dynamics were tentatively fitted with bi-exponential rise functions and the time constants are tabulated in Table 1. The analysed time range was set to 30–2800 ps (end of the time delay), so that the initial fast decays due to the relaxation from the S2 to S1 (∼10 ps) can be excluded from the analyses. It clearly shows that the formation process of the radical anion of PBI is decelerated with the decrease in PBI/CdS NC. The fact that the formation of the radical anion of PBI takes nanoseconds indicates that the excited electron in the conduction band survives over nanoseconds. Therefore, this result shows that the fewer PBI/CdS NC elongates the lifetime of the excited electron in the conduction band of CdS NCs. It is expected that the larger CdS nanostructures could further elongate the lifetime of the excited electrons because the larger NC size decreases the surface volume ratio.


image file: d0nr08493j-f7.tif
Fig. 7 Normalized double pulse transient absorption dynamics of PBI-CdS in chloroform excited at 520 and 705 nm (time delay = 19.6 ps) and probed at 725 nm in different molar ratios of PBI and CdS NC. Thick lines are fitting lines. The formation kinetics indicate the electron transfer from the conduction band of CdS NCs to neutral PBI.
Table 1 Time constants of the subsequent electron transfer from the conduction band to PBI of different molar ratios of PBI-CdS
PBI/CdS NC 25 9 2
τ 1 80 ps (42%) 180 ps (16%) 410 ps (13%)
τ 2 780 ps (58%) 2.9 ns (84%) 6.9 ns (87%)


Conclusions

The whole photochemical reaction processes of PBI-CdS by stepwise 2PA is summarized in Scheme 1. It is noted that PBI-CdS is denoted as PBI-CdS-PBI in the scheme because two PBIs coordinated to NCs are involved in the reaction. In the first step, the S1 state of PBI (PBI*) is formed by the excitation with the 520 nm pulse (1). Secondly, the S2 state (PBI**) is formed by the excitation with the 705 nm pulse (2). The electron transfer from the S2 state of PBI to the conduction band of CdS NCs instantaneously occurs within IRF (<70 fs) (3), and the quantum yield for the electron transfer was estimated to be ∼0.5–0.7. The excited electron of the conduction band of CdS NCs survives over nanoseconds, and finally the subsequent electron transfer occurs to produce a radical anion of PBI with the quantum yield of 0.69 (4). While the lifetime of the higher excited state of PBI is at most ∼10 ps, the lifetime of the extracted electron at the conduction band of CdS NCs was over nanoseconds. The NC part can be replaced with other NCs whose conduction bands are higher than CdS such as ZnS, indicating that the higher reductive electrons can be stored in NCs with this strategy. Moreover, the delayed formation process of the radical anion of PBI by the stepwise electron transfer process can be used as “an indirect probe” for the electron transfer from the higher excited states of organic molecules to NCs. The method to extract the electron from a short-lived excited state in this study can be also applied to the excited states of radical anions, which is also short-lived, and therefore, the present strategy will be important to expand the versatility for higher excited states and the short-lived excited states.
image file: d0nr08493j-s1.tif
Scheme 1 The whole photochemical reaction processes of PBI-CdS by stepwise 2PA. It is noted that PBI-CdS was denoted as PBI-CdS-PBI because two PBIs are involved in the reaction.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported partly by JSPS KAKENHI Grant Number JP18H05263, and Nippon Sheet Glass Foundation for Materials Science and Engineering. The authors acknowledge Dr Daiki Fujioka for assisting TEM measurements.

References

  1. Y. Tsuboi, R. Shimizu, T. Shoji and N. Kitamura, J. Am. Chem. Soc., 2009, 131, 12623–12627 CrossRef CAS.
  2. K. Mori, Y. Ishibashi, H. Matsuda, S. Ito, Y. Nagasawa, H. Nakagawa, K. Uchida, S. Yokojima, S. Nakamura, M. Irie and H. Miyasaka, J. Am. Chem. Soc., 2011, 133, 2621–2625 CrossRef CAS.
  3. I. Ghosh, L. Marzo, A. Das, R. Shaikh and B. König, Acc. Chem. Res., 2016, 49, 1566–1577 CrossRef CAS.
  4. F. Glaser, C. Kerzig and O. S. Wenger, Angew. Chem., 2020, 59, 2–21 CrossRef.
  5. Y. Kobayashi, K. Mutoh and J. Abe, J. Photochem. Photobiol., C, 2018, 34, 2–28 CrossRef CAS.
  6. K. Mutoh, Y. Nakagawa, A. Sakamoto, Y. Kobayashi and J. Abe, J. Am. Chem. Soc., 2015, 137, 5674–5677 CrossRef CAS.
  7. M. Koga, Y. Yoneda, H. Sotome and H. Miyasaka, Phys. Chem. Chem. Phys., 2019, 21, 2889–2898 RSC.
  8. T. Kawakami, M. Koga, H. Sotome and H. Miyasaka, Phys. Chem. Chem. Phys., 2020, 22, 17472–17481 RSC.
  9. M. P. O′neil, M. P. Niemczyk, W. A. Svec, D. Gosztola, G. L. Gaines and M. R. Wasielewski, Science, 1992, 257, 63–65 CrossRef.
  10. H. Miyasaka, M. Murakami, A. Itaya, D. Guillaumont, S. Nakamura and M. Irie, J. Am. Chem. Soc., 2001, 123, 753–754 CrossRef CAS.
  11. M. Murakami, H. Miyasaka, T. Okada, S. Kobatake and M. Irie, J. Am. Chem. Soc., 2004, 126, 14764–14772 CrossRef CAS.
  12. C. L. Ward and C. G. Elles, J. Phys. Chem. Lett., 2012, 3, 2995–3000 CrossRef CAS.
  13. C. L. Ward and C. G. Elles, J. Phys. Chem. A, 2014, 118, 10011–10019 CrossRef CAS.
  14. H. Stiel, K. Teuchner, A. Paul, W. Freyer and D. Leupold, J. Photochem. Photobiol., A, 1994, 80, 289–298 CrossRef CAS.
  15. M. Fujitsuka and T. Majima, J. Photochem. Photobiol., C, 2018, 35, 25–37 CrossRef CAS.
  16. N. Yanai and N. Kimizuka, Chem. Commun., 2016, 52, 5354–5370 RSC.
  17. M. A. Tasdelen, V. Kumbaraci, S. Jockusch, N. J. Turro, N. Talinli and Y. Yagci, Macromolecules, 2008, 41, 295–297 CrossRef CAS.
  18. N. A. O&rsquo;Connor, A. J. Berro, J. R. Lancaster, G. Xinyu, S. Jockusch, T. Nagai, T. Ogata, S. Lee, P. Zimmerman, C. G. Willson and N. J. Turro, Chem. Mater., 2008, 20, 7374–7376 Search PubMed.
  19. S. Hirata, K. Totani, T. Yamashita, C. Adachi and M. Vacha, Nat. Mater., 2014, 13, 938–946 CrossRef CAS.
  20. J. C. Scaiano, L. J. Johnston, W. G. McGimpsey and D. Weir, Acc. Chem. Res., 1988, 21, 22–29 CrossRef CAS.
  21. X. Cai, M. Sakamoto, M. Hara, S. Inomata, M. Yamaji, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka and T. Majima, Chem. Phys. Lett., 2005, 407, 402–406 CrossRef CAS.
  22. M. Shimizu, M. Schelper, K. Mochida, T. Hiyama, M. Adachi, Y. Sasaki, S. Akiyama, S. Maeda, H. Kanbara, Y. Mori and T. Kurihara, Adv. Mater., 2007, 19, 1826–1829 CrossRef CAS.
  23. I. Yonekawa, K. Mutoh, Y. Kobayashi and J. Abe, J. Am. Chem. Soc., 2018, 140, 1091–1097 CrossRef CAS.
  24. K. Mutoh, Y. Kobayashi, T. Yamane, T. Ikezawa and J. Abe, J. Am. Chem. Soc., 2017, 139, 4452–4461 CrossRef CAS.
  25. Y. Kobayashi, T. Katayama, T. Yamane, K. Setoura, S. Ito, H. Miyasaka and J. Abe, J. Am. Chem. Soc., 2016, 138, 5930–5938 CrossRef CAS.
  26. Y. Kobayashi, K. Mutoh and J. Abe, J. Phys. Chem. Lett., 2016, 7, 3666–3675 CrossRef CAS.
  27. K. Mutoh, Y. Nakagawa, A. Sakamoto, Y. Kobayashi and J. Abe, J. Am. Chem. Soc., 2015, 137, 5674–5677 CrossRef CAS.
  28. D. M. Burland, Acc. Chem. Res., 1983, 16, 218–224 CrossRef CAS.
  29. V. Gerbig, R. K. Grygier, D. M. Burland and G. Sincerbox, Opt. Lett., 1983, 8, 404–406 CrossRef CAS.
  30. Y. Xu, J. Zheng, J. O. Lindner, X. Wen, N. Jiang, Z. Hu, L. Liu, F. Huang, F. Würthner and Z. Xie, Angew. Chem., 2020, 59, 10363–10367 CrossRef CAS.
  31. I. Ghosh, T. Ghosh, J. I. Bardagi and B. Konig, Science, 2014, 346, 725–728 CrossRef CAS.
  32. N. T. La Porte, J. F. Martinez, S. Hedström, B. Rudshteyn, B. T. Phelan, C. M. Mauck, R. M. Young, V. S. Batista and M. R. Wasielewski, Chem. Sci., 2017, 8, 3821–3831 RSC.
  33. F. Brandl, S. Bergwinkl, C. Allacher and B. Dick, Chem. – Eur. J., 2020, 26, 7946–7954 CrossRef CAS.
  34. C. J. Zeman, S. Kim, F. Zhang and K. S. Schanze, J. Am. Chem. Soc., 2020, 142, 2204–2207 CrossRef CAS.
  35. D. T. Breslin and M. A. Fox, J. Phys. Chem., 1994, 98, 408–411 CrossRef CAS.
  36. J. C. Gumy and E. Vauthey, J. Phys. Chem. A, 1997, 101, 8575–8580 CrossRef CAS.
  37. D. Gosztola, M. P. Niemczyk, W. Svec, A. S. Lukas and M. R. Wasielewski, J. Phys. Chem. A, 2000, 104, 6545–6551 CrossRef CAS.
  38. J. A. Christensen, B. T. Phelan, S. Chaudhuri, A. Acharya, V. S. Batista and M. R. Wasielewski, J. Am. Chem. Soc., 2018, 140, 5290–5299 CrossRef CAS.
  39. T. Uno, M. Koga, H. Sotome, H. Miyasaka, N. Tamai and Y. Kobayashi, J. Phys. Chem. Lett., 2018, 9, 7098–7104 CrossRef CAS.
  40. T. W. Kee, J. Phys. Chem. Lett., 2014, 5, 3231–3240 CrossRef CAS.
  41. S. S. Lim, D. Giovanni, Q. Zhang, A. Solanki, N. F. Jamaludin, J. Wei, M. Lim, N. Mathews, S. Mhaisalkar, M. S. Pshenichnikov and T. C. Sum, Sci. Adv., 2019, 5, 1–6 Search PubMed.
  42. Z. Li, Y. Ji, R. Xie, S. Y. Grisham and X. Peng, J. Am. Chem. Soc., 2011, 133, 17248–17256 CrossRef CAS.
  43. U. Avcıbaşı, H. Dinçalp, T. Ünak, Y. Yıldırım, N. Avcibaşi, Y. Duman and S. Içli, J. Radioanal. Nucl. Chem., 2007, 273, 669–675 CrossRef.
  44. M. E. Bhosale and K. Krishnamoorthy, Chem. Mater., 2015, 27, 2121–2126 CrossRef CAS.
  45. N. Rehmat, A. Toffoletti, Z. Mahmood, X. Zhang, J. Zhao and A. Barbon, J. Mater. Chem. C, 2020, 8, 4701–4712 RSC.
  46. W. W. Yu, L. Qu, W. Guo and X. Peng, Chem. Mater., 2003, 15, 2854–2860 CrossRef CAS.
  47. R. F. Kubin and A. N. Fletcher, J. Lumin., 1982, 27, 455–462 CrossRef.
  48. A. Nowak-Król, B. Fimmel, M. Son, D. Kim and F. Würthner, Faraday Discuss., 2015, 185, 507–527 RSC.
  49. V. I. Klimov, J. Phys. Chem. B, 2000, 104, 6112–6123 CrossRef CAS.
  50. P. Kambhampati, J. Phys. Chem. C, 2011, 115, 22089–22109 CrossRef CAS.

Footnote

Electronic supplementary information (ESI) available: Molecular synthesis, material characterizations, and transient absorption spectra. See DOI: 10.1039/d0nr08493j

This journal is © The Royal Society of Chemistry 2021