Re-excitation of localized electrons in SnO₂ quantum dots for enhanced water photolysis activity

Abstract

Trapping and further localization of photo-generated carriers by defects in a photocatalyst are crucial obstacles for enhancing the efficiency of hydrogen generation from water photolysis. In this paper, we reported a re-excitation mechanism of localized electrons assisted by infrared light for improving photocatalytic activity. Increasing hydrogen production is demonstrated when introducing infrared light for SnO₂ quantum dots at a grain size smaller than twice its Debye length and supported on g-C₃N₄. Comparatively, in absence of infrared light, rich oxygen vacancies in these SnO₂ quantum dots reduced the photocatalytic activity. The re-excitation of localized electrons would provide a viable approach to utilize electrons more efficiently in the photocatalytic or other catalytic fields.

---

1. Introduction

The migration of electrons governs the activity of a catalyst in a chemical reaction. As an important factor, photo-generated charge separation rates of a photocatalyst determine the photon–hydrogen conversion efficiency in water photolysis. Introducing electron trap centers into a photocatalyst is an effective manner to enhance photo-generated charge separation. Oxygen vacancy is the usual structure that acts as the electron trap (ET) in metal oxide semiconductor photocatalysts.^1,2 But in previous studies, the enhanced photo-generated charge separation caused by oxygen vacancies in photocatalysts could only enhance the photocatalytic activity for degradation of organic pollutants.^1–5 In the process of hydrogen generation from water photolysis, free electrons are required to reduce hydrogen ions into hydrogen rather than couple with adsorbed oxygen to form a superoxide radical (˙O₂) and oxidize organic matter. Therefore, oxygen vacancies usually play a negative role in hydrogen generation from water splitting, owing to its electron trapping effect. Due to self-purification,⁶ oxygen deficiencies are almost inevitable in nano-structured oxide semiconductor photocatalysts. Finding a new route to re-excite the electrons trapped by oxygen vacancies so to improve the water splitting efficiency is essential.

The interaction between lattice and electron (i.e., phonon–electron interaction) dominates the migration of electrons. Because lattice vibrations (phonons) absorb infrared light easily, the energy transfer between a phonon and a photo-generated electron could become more frequent under the irradiation of infrared light. On the other hand, photo-generated electrons are easily trapped by oxygen vacancy and become localized. These localized electrons are still lying in the excitation state, and can be more excited by external stimulation compared with a valence band electron. For example, localized electrons trapped by oxygen vacancies (usually called F-centers) can be thrown into the conduction band by lattice vibrations via complex phonon–electron interactions when the lattice interacts with infrared light.⁷ Therefore, the influence of infrared light on the photocatalytic reaction would be more evident. However, there is still lack of experimental data to reveal the impacts of infrared light on the photocatalytic activity of materials in water splitting.

Compared to homogeneous photocatalysts, heterogeneous photocatalysts have received more attention due to their further advantages and wider range of potential applications.^8–10 As a wide band gap semiconductor, tin dioxide (SnO₂) is widely applied in optical devices, catalysts, batteries, gas sensors, and biosensors.¹¹ Graphitic carbon nitride (g-C₃N₄) has been deemed a promising heterogeneous metal-free catalyst for a wide range of applications, such as solar energy utilization toward water splitting.¹² In this letter, we realized the excitation of localized electrons in SnO₂ quantum dots (QDs) supported on g-C₃N₄ (as displayed in Scheme 1) via the water photolysis test of SnO₂/g-C₃N₄ hybrids. Due to the conduction band potential difference, g-C₃N₄ could be regarded as a free electron bucket to provide electrons to SnO₂ for water splitting (Scheme 1). In the photocatalytic water splitting process, the migration of photo-generated electrons governs the reaction rate of protons to hydrogen. For the SnO₂/g-C₃N₄ hybrid, the electrons, taking part in the reaction, are derived from g-C₃N₄ due to the matching energy gap with visible light. The excitation and migration process of photo-generated charges is depicted in Scheme 1. Firstly, visible light excites a valence band electron to the conduction band in g-C₃N₄ (route A). Due to the higher conduction band bottom position of g-C₃N₄ (−1.13 eV per NHE)¹³ and the lower conduction band bottom of SnO₂ (−0.11 eV per NHE),¹⁴ the electric potential difference would drive a migration of photo-generated electrons from g-C₃N₄ to SnO₂, which seems to separate the photo-generated electrons/holes efficiently (route B). Despite the advantage of the electric potential difference in photo-generated charges separating the hybridized photocatalyst, the photo-generated electrons are easily captured by electron traps of SnO₂ and become localized. Usually, these localized electrons could not reduce protons into hydrogen (route C). As commonly viewed, introducing electron traps to photocatalysts is helpful to separate photo-generated charges more efficiently. In hydrogen evolution from water photolysis catalyzed by SnO₂/g-C₃N₄ hybrids under visible light, oxygen vacancies would localize photo-generated electrons from the donor of g-C₃N₄. The photo-generated electrons competitively migrate between the SnO₂ conduction band and electron trap center and could weaken the photocatalytic activity for hydrogen evolution.

Scheme 1 Photo-generated electron migration in SnO₂/g-C₃N₄ hybrids under visible light. g-C₃N₄ acts as a free electron source for hydrogen evolution in the water photolysis reaction. The location and distribution of electron trap centers caused by oxygen vacancies in SnO₂ were marked by referring to the model proposed in ref. 15.

Herein, the heat isolate filter was used to investigate the thermal effect on localized electron behavior via the water photolysis test of SnO₂/g-C₃N₄ hybrids. By tailoring the SnO₂ mean grain size, these hybrids had varied concentrations of oxygen vacancies. A low-lying hydrogen evolution was observed by tailoring the mean grain size of SnO₂. We present a charge separation mechanism involving the re-excitation of trapped electrons in SnO₂ quantum dots for water photolysis. The process enhanced the photo-generated charge utilization in the photolysis of water when compared with pure g-C₃N₄. From the mechanism, a new understanding about electron migration properties under quantum size effects was obtained, and the re-excitation of localized electrons would provide a viable approach to utilize electrons more efficiently in the photocatalytic or other catalytic fields.

2. Experimental

2.1 Materials

Melamine (C₃H₆N₆), ethanol, methanol, SnCl₄·5H₂O, ethylenediamine, urea, and triethanolamine were of a chemically pure grade. All of the materials were purchased from Sinopharm Chemical Reagent Corp (P. R. China) and used without further purification.

2.2 Chemicals for materials preparation

2.2.1 Preparation of g-C₃N₄. 10 g of melamine was heated at 520 °C in a muffle furnace for 4 h in a static area with a heating rate of 10 K min⁻¹. The resultant yellow solids were milled into a powder in an agate mortar.

2.2.2 Preparation of SnO₂. SnO₂ quantum dots (QDs) were prepared according to the procedures described in Preparation of SnO₂ uncapped QDs in ref. 16. In detail, 5.83 mmol SnCl₄·5H₂O was dissolved in a mixed solvent (18.50 mL methanol and 9.28 mL ethylenediamine) under constant magnetic stirring for 1 h. Urea (58.3 mmol) was dissolved in 18.50 mL deionized (DI) H₂O and then was added to the reaction mixture. The resultant solution was stirred for 30 min. The obtained white slurry was then transferred to a 100 mL Teflon-lined stainless steel chamber and heated at 90 °C for 8 h. After cooling down to room temperature, the product was centrifuged. The as-collected product was washed with DI H₂O and ethanol several times to remove impurities and dried overnight in vacuum. After that, the as-prepared SnO₂ QDs were milled into powders and annealed at different temperatures (280 °C, 320 °C, 400 °C, 450 °C, 500 °C, 600 °C, and 800 °C) for 80 minutes, to remove residual ethylenediamine (boiling point of 117 °C at constant pressure), and SnO₂ with different mean grain sizes or crystallinities were obtained. To express this more clearly, the obtained SnO₂ were named SnO₂-X, where X is the mean grain size of SnO₂. For example, SnO₂-3.0 is the SnO₂ with the mean grain size of 3.0 nm.

2.2.3 Hybridization of g-C₃N₄ and SnO₂. The as-prepared SnO₂ were milled with g-C₃N₄ at a mass ratio of SnO₂/g-C₃N₄ = 2 : 3. These mixtures were mixed with an appropriate amount of ethanol and then ultrasonic for 20 min. Next, these slurries were stirred in a fume hood to evaporate the ethanol. The dry mixture was roasted at 400 °C in the air for 1 h. The roasted products were milled into a powder for characterization of structure, morphology, valence, etc. To express this more clearly, the obtained hybrids were named SCTX, where X is the mean grain size of SnO₂. For example, SCT3.0 is the hybridized sample with 3.0 nm mean grain size of SnO₂.

2.3 Analytical and testing instruments

The phase structure and crystallinity of samples was determined by X-ray diffraction patterns that were recorded on a Rigaku MiniFlex II benchtop X-ray diffractometer (XRD) with Cu Kα irradiation (λ = 1.5418 Å). The mean grain sizes of SnO₂ were obtained by fitting XRD patterns with MDI Jade 6.0. Thermal gravimetric (TG) data of the samples were recorded on a Netzsch STA449F3 thermal analyzer at a heating rate of 10 K min⁻¹ in air. The particle morphologies of the samples were performed via transmission electron microscopy (TEM, JEM-2010 produced by JEOL). The chemical states of the samples were analyzed using X-ray photoelectron spectroscopy (XPS, monochromatic Al Kα X-ray source operating at 150 W), and binding energies were referenced to the C 1s peak at 284.6 eV that arises from adventitious carbon. The XPS spectra were fitted by the Thermo Avantage software. Electron paramagnetic resonance (EPR) measurements for the samples were performed under ultraviolet light via a Bruker-BioSpin, E500 spectrometer to confirm the existence of oxygen vacancies. A frequency of ca. 9.866 GHz was used for a dual-purpose cavity operation. The magnetic field of 0.1 mT was modulated at 100 kHz. Ultraviolet-visible (UV-vis) absorbance spectra of the samples were measured by a Varian Cary-500 spectrophotometer under ambient conditions. The specific surface area of these samples was characterized from the nitrogen absorption data measured at liquid nitrogen temperature on a Micromeritics Tristar 3000. Before N₂-sorption analysis, these samples were pretreated under vacuum at 120 °C for 1 h.

2.4 Photocatalytic activity measurements

Photocatalytic hydrogen evolution was conducted in an online photocatalytic hydrogen generation system (Labsolar-IIIAG, Perfectlight Corp, Beijing, P. R. China), and keep at a temperature of 5 °C. A 300 W Xe lamp equipped with a UV-cutoff filter (λ > 420 nm) and heat insulation filter was used as the light source for different investigation processes, while a light-cutoff filter (λ > 260 nm) was used alone in the hydrogen generation test of SnO₂-8.0 to ensure the photocatalytic activity in hydrogen generation was from water photolysis. In detail, 100 mg samples and 90 mL deionized water were added into the reactor. 10 mL triethanolamine, which acted as a sacrificial agent, was added. H₂PtCl₆ solution (1.0 wt% for Pt) was used for Pt deposition, which would act as a co-catalyst to improve the photocatalytic performance of all samples. Prior to the reaction, the whole reaction system was degassed to remove O₂ and CO₂. Gas evolution was analyzed by an on-line gas chromatograph (Fuli 9790II) with a TCD detector purchased from Fuli Analytical Instrument Corp, Zhejiang, P. R. China. The hydrogen evolution value was calculated based on 100 mg catalyst. Before the formal test, 30 minutes were spent for the light deposition of Pt from H₂PtCl₆, and then the system was evacuated again to start the reaction. The Pt deposition process can be described as follows:

H₂PtCl₆ → 2H⁺ + PtCl₆²⁻ (1)

PtCl₆²⁻ + 4e⁻ → Pt + 6Cl⁻ (2)

3. Results and discussion

3.1 XRD characterization

Fig. 1 illustrates the variation of mean grain size with the annealing temperature from 280 to 800 °C. Based on the Scherrer formula, the mean grain sizes of SnO₂ samples range from 3.0 nm to 27.0 nm. The diffraction patterns full width at half maximum (FWHM) of SnO₂ are narrowed with the SnO₂ grain size increase, which suggested the crystallinity of SnO₂ were enhanced with increasing grain size. Table 1 shows the mean grain size of the as-prepared SnO₂ via the annealing process. To express more clearly, the obtained SnO₂ were named as SnO₂-X, here X is the mean grain size of SnO₂. For example, SnO₂-3.0 is the SnO₂ with the mean grain size of 3.0 nm.

Fig. 1 XRD patterns of the SnO₂ with different grain sizes.

Table 1 Experimental and calculated energy band gaps (E_g) of the selected SnO₂ nanocrystals at different grain sizes

Annealing temperature (°C)	280	320	400	450	500	600	800
Mean grain size (nm)	3.0	4.0	5.2	8.0	13.0	18.0	27.0
Calculated Eg (eV)	4.2	3.8	—	3.7	3.6	—	—
Experimental Eg (eV)	4.0	3.8	—	3.6	3.5	—	—

---

3.2 UV-vis spectra and quantum confinement of SnO₂

The variation of SnO₂ grain size resulted in different band gaps as well as oxygen vacancy concentrations. As shown in Fig. 2A, the UV-vis absorption peak changed from 290 nm to 310 nm as the mean grain size of SnO₂ increased from 3.0 nm to 18.0 nm, and their absorbance edges also exhibited a clear red shift (Fig. 3). The energy band gaps could be estimated from the Kubelka–Munk function,¹⁷

(αhν)ⁿ = K(hν − E_g) (3)

where α and E_g refer to the absorption coefficient and bandgap energy of the semiconductors, respectively. K is a constant, hν represents the photon energy and n is 2 for SnO₂, a direct band gap semiconductor. The energy band gaps were 4.0, 3.8, and 3.6 for SnO₂-3.0, SnO₂-5.2, and SnO₂-8.0, respectively (Fig. 2B). Comparatively, a lower gap value of 3.5 eV was obtained for the as-prepared SnO₂-18.0, which was somewhat lower than the reported bulk value of 3.6 eV. Size-dependent optical properties of QDs were widely explained by the effective mass model. When the SnO₂ grain radius is below its Bohr radius (2.7 nm), the electron–hole motion is not correlated. In that case, the effective band gap energy is given by the following equation,¹⁶where r is the mean grain size, μ is the effective reduced mass (0.27m_e for SnO₂), E_g refers to the bulk band gap energy (3.6 eV), and E^eff_g represents the effective band gap energy. The experimental and calculated band gap energies of SnO₂-3.0, SnO₂-5.2, SnO₂-8.0, and SnO₂-18.0 are compared in Table 1. The experimental values were slightly smaller than, or equal to, the calculated ones. The variation of band gap proved the presence of evident quantum confinement in SnO₂-X. Moreover, the quantum confinement effect not only reduced the energy state density of the electrons, but also resulted in a distinct phonon bottleneck effect, which could dramatically slow down electron–phonon relaxation in nanoclusters.¹⁸

Fig. 2 (A) UV-vis spectra and (B) the energy band gaps of SnO₂-X.

Fig. 3 EPR spectra of the samples SnO₂-3.0, SnO₂-8.0, SnO₂-13.0, and SnO₂-18.0.

3.3 EPR analysis for SnO₂-X

In order to reveal the defect variation of SnO₂-X, EPR tests under ultraviolet light were performed. Isolated Sn⁴⁺ and Sn²⁺ ions are EPR silent due to their electron configuration of [Kr]4d¹⁰. However, when oxygen vacancies are introduced into SnO₂-X nanocrystals, single electrons are easily trapped in these oxygen vacancies to form single charge centers and to become EPR active. As shown in Fig. 3, one signal at g = 2.004 is due to the single electron trapped in oxygen vacancies (V_O) at the surface of SnO₂.¹⁹ These ionized oxygen vacancies in SnO₂ form shallow trap levels with an energy of ∼0.03 eV and ∼0.15 eV below the bottom of the conduction band (CBM).¹⁵ The other two signals at g = 2.058 and g = 2.061 are attributed to the superoxide ions (O₂⁻)²⁰ that are caused by the combination of adsorbed oxygen with the trapped electron in the oxygen vacancy. The formation mechanism of O₂⁻ can be described by the following reaction,¹⁹

Sn⁴⁺ + V_O + O₂ → Sn⁴⁺–O₂⁻ (5)

This process is very similar to the reaction of O₂ with subsurface oxygen vacancies on the (101) plane of anatase TiO₂.²¹ The EPR signal of O₂⁻ is widely used to decide oxygen vacancy concentrations. From Fig. 2, one can see that the intensity of the EPR signal decreases with and increase of SnO₂-X grain size. The sample SnO₂-3.0 exhibited the strongest O₂⁻ EPR signal, while the signal for sample SnO₂-13.0 was hardly distinguishable and the signal for SnO₂-18.0 disappeared completely. This observation demonstrated a reducing of oxygen vacancies with SnO₂ grain size increases, and the concentration of oxygen vacancies should be extremely low in SnO₂ particles at a grain size beyond 18.0 nm.

3.4 O 1s XPS analysis of SnO₂/g-C₃N₄ hybrids

The as-prepared SnO₂-X were milled with g-C₃N₄ at a fixed mass ratio of 2 : 3 via the calcination method at 400 °C for 1 h, which led to the final hybrids. To express more clearly, the obtained hybrids were named as SCTX, here X is the grain size of SnO₂. The XRD patterns of the hybrids are shown in Fig. S1,† and the mass ratios of the two components were confirmed by TG testing (Fig. S2†). The variation of oxygen vacancies in SnO₂/g-C₃N₄ hybrids was further evaluated by the O 1s XPS spectra based on the fact that the oxygen vacancy can affect the chemical and physical adsorption behavior of a compound (Fig. 4).

Fig. 4 O 1s XPS spectra of SCT3.0, SCT5.2, SCT13.0, and SCT18.0 hybrids.

The selected O 1s XPS spectra for SCT3.0, SCT8.0, SCT13.0, and SCT18.0 are shown in Fig. 3. All of O 1s core levels could be de-convoluted into 3 sub-peaks using the Thermo Avantage software. The first one at the lowest binding energy of 530.2 eV could be identified to the lattice oxygen species O²⁻ coordinated with Sn⁴⁺. The other two sub-peaks with binding energies of 531.5 eV and 532.8 eV were attributed to the absorbed oxygen species.²² Notably, the peak at 531.5 eV, denoted as OH⁻, was governed by the adsorbed oxygen species caused by surface oxygen vacancies.²³ The integral area ratios of absorbed oxygen species to the lattice oxygen in Table S1,† revealed that the concentration of absorbed oxygen species was reduced with the increase of SnO₂ grain size in hybrids. That is to say, oxygen vacancy concentrations in SCTX decreased with the grain size increase of SnO₂. The TEM morphology observations for selected SCTX also confirmed the surface defects were reduced with increases of SnO₂ grain size (Fig. S4†).

4. Photocatalytic activities in hydrogen generation

As commonly viewed, introducing electron traps to a photocatalyst is helpful to separate photo-generated charges more efficiently. In hydrogen evolution from water photolysis catalyzed by SnO₂/g-C₃N₄ hybrids under visible light, the oxygen vacancy would localize photo-generated electrons from the donor, g-C₃N₄. The photo-generated electrons competitively migrate between the SnO₂ conduction band and the electron trap center, which could weaken the photocatalytic activity in hydrogen evolution from water photolysis.

For the as-prepared SCTX, the hydrogen evolution test from water photolysis was carried out for 3 h under visible light irradiation and the temperature of 5 °C was kept constant. During the first test, the light source was a 300 W Xe lamp equipped with a UV-cutoff filter (λ > 420 nm) and heat insulation filter (the heat insulation filter can isolate most of the infrared light). As shown in Fig. 5, the SCT27.0 exhibited the highest hydrogen yield of 11.3 μmol. Further decreasing of SnO₂ grain size was accompanied with decreasing hydrogen yield. That is, in presence of a heat insulation filter, SnO₂ nanocrystals with small grain sizes in the hybrid exhibited poor photocatalytic activity. It should be also emphasized that the activity of the as-prepared SnO₂-8.0 in hydrogen evolution was really poor and yielded hydrogen of 3.0 μmol (under full light irradiation, λ > 260 nm) and 0 μmol (under visible light irradiation, λ > 420 nm) only after irradiation for 5 h (Fig. S6†). As illustrated in Fig. 6A, the pure g-C₃N₄ showed a hydrogen yield of 6.0 μmol, which is less than the 7.5 μmol of SCT3.0. The EPR and XPS data has proved that the oxygen vacancy concentrations in SnO₂ increased with the decrease of SnO₂ grain size. Therefore, we concluded that the photocatalytic activity of SnO₂/g-C₃N₄ hybrids in hydrogen evolution under visible light was weakened by oxygen vacancies.

Fig. 5 Comparisons of hydrogen yield of SnO₂/g-C₃N₄ hybrids (SCTX) with different SnO₂ grain sizes under visible light (λ > 420 nm) irradiation with and without the heat insulation filter.

Fig. 6 Catalytic hydrogen yields from water splitting under visible light (420 nm) irradiation for 3 hours with 100 mg of (A) the as-prepared g-C₃N₄ and SCT3.0 with heat isolation filter; and (B) the as-prepared g-C₃N₄ and SCT8.0 without the heat isolation filter.

The hydrogen evolution test from water photolysis was also carried out without the heat insulation filter, i.e. in presence of infrared irradiation or thermal light effects in addition to excitation of electrons. As described above, in the absence of a heat insulation filter, the hydrogen yield of the hybrids increased with the grain size of SnO₂ nanocrystals, and g-C₃N₄ gave the least hydrogen yield. Comparatively, under visible light irradiation but without heat insulation filter, g-C₃N₄ exhibited hydrogen yields as high as 10.0 μmol, but still less than the minimum of 10.2 μmol observed for SCT8.0 (Fig. 6B). Different from the monotonous decreasing tendency of photocatalytic performance in the presence of the heat insulation filter, hydrogen evolution exhibited a valley at the SnO₂ grain size of 8.0 nm in the absence of the heat insulation filter. More specifically, for hybrids where the grain size of the SnO₂ nanocrystal was larger than 8.0 nm, the equipping with or without the heat insulation filter just altered the hydrogen yield, but did not change the decreasing tendency with reduced mean grain size of SnO₂ nanocrystals. However, when the mean grain size of SnO₂ was smaller than 8.0 nm, removing the heat insulation filter changed the hydrogen evolution rate of hybrids from a decrease to an increase. Moreover, hydrogen evolution reached 21.0 μmol for SCT3.0, very close to the 21.1 μmol of SCT27.0. On the other hand, the hydrogen evolution valley for sample SCT8.0 was very close to that of pure g-C₃N₄.

It is well known that the bigger specific area of catalyst usually leads to more active sites for a catalytic reaction, thus enhancing the photocatalytic activity.¹² However, based on the specific area data of the samples (SCT-18.0, SCT-13.0, and SCT-4.0 (Table S2†)) one can see that even the specific area for SCT-13.0 was bigger than the specific area for SCT-18.0 and the former sample shows a lower photocatalytic activity. Certainly, the hybrids' photocatalytic activities obey the principle when infrared light was filtered, but not when the infrared light was allowed.

Hence, two important points were revealed from the comparison of hydrogen evolutions under visible light with and without a heat insulation filter for SnO₂/g-C₃N₄ hybrids: (i) the infrared light altered the photocatalytic activity tendency of SnO₂/g-C₃N₄ hybrids when increasing the SnO₂ grain size from 3.0 nm to 27.0 nm; (ii) oxygen vacancies could reduce the photocatalytic activity of hybrids in water photolysis in the absence of infrared irradiation, and the grain size of SnO₂ in the hybrids was a key parameter that determined the activity valley position.

5. Re-excitation of localized electrons leading valley activity in hydrogen generation

The EPR and XPS detection has shown that the oxygen vacancies reduced with the increase of SnO₂ grain size. The decrease of oxygen vacancy, i.e. the electron traps, would reduce the possibility of capturing photo-generated electrons. Therefore, it could be well understood that the photocatalytic activities increased with the SnO₂ mean grain size under visible light irradiation with the utilization of a heat insulation filter. Obviously, the variation of oxygen vacancy concentration could not explain the valley activity observed at 8.0 nm in Fig. 5 with the presence of infrared irradiation. The enhanced photocatalytic activities for the hybrids with smaller SnO₂ (with the mean grain size below 8.0 nm) could be explained by the re-excitation of the localized electron in SnO₂ under infrared light irradiation. The photo-generated electrons could be localized in the oxygen vacancy. It is widely accepted that these localized electrons would interact with the SnO₂ lattice (phonon–electron interaction) until they are annihilated. On the other hand, the localized electron could also be regarded as an F-center electron that can be excited by lattice vibrations in a light absorbing process, as reported by K. Huang et al.⁷ This mechanism is very like the phonon-assisted light absorption by free carriers in n-SnO₂.²⁴ Even so, this theory still cannot accurately explain why near-infrared light alters the photocatalytic activity tendency of hybrids when the SnO₂ grain size increases from 3.0 nm to 27.0 nm. To well understand the reason for the valley activity in presence of infrared irradiation, we must mention two important features of small sized nanocrystals, i.e., the Debye length and phonon bottleneck effect. Firstly, the Debye length of SnO₂ nanocrystals is about 3 nm.²⁵ When the size of the component SnO₂ grain is smaller than twice of its Debye length, the electrons trapped by oxygen vacancies cannot be deeply shielded, i.e., the lattice vibration could throw electrons out of the trap more easily. Secondly, the bottleneck effect for small sized crystals, especially QDs, could further reduce the electron–phonon coupling. Therefore, the weakly localized electrons have a high probability to return back to the conduction band due to the bottleneck effect in small sized nanocrystals driven by an external energy such as absorbing infrared light energy. Thus, more free electrons might participate in protons reduction.

Then for SnO₂/g-C₃N₄ hybrids, when the size of the component SnO₂ nanocrystals is larger than 8.0 nm, the strong charge shielding would lead to the trapped electrons in the defect level of the oxygen vacancy be in a deeply localizing state. Infrared light irradiation, as the only driving external force, could not de-localize these electrons. Therefore, most of the localized electrons could not participate in the protons reduction, and the oxygen vacancy in large sized SnO₂ nanocrystals just act as an electron trap center. Comparatively, as the grain size of component SnO₂ nanocrystals becomes smaller than 8.0 nm, near to double the Debye length of SnO₂, the weakened shielding effect assists these captured electrons in de-localizing under infrared light irradiation. This re-excitation for localized electrons usually accompanies multiple phonons generating and annihilating. Further decreasing the size to 3.0 nm (sample SCT3.0), the component SnO₂ nanocrystals could be regarded as QDs that give off an obvious phonon bottleneck effect for successfully de-localizing the captured electrons in an oxygen vacancy. In this situation, the oxygen vacancy would act as an electron collection center that not only separates photo-generated charges more effectively, but also re-utilizes the localized electron into a useful reduction (Scheme 2).

Scheme 2 Re-excitation diagrams of localized electrons in SnO₂/g-C₃N₄ water for hydrogen generation from water photolysis by introducing infrared light. The location and distribution of electron trap centers caused by oxygen vacancies in SnO₂ were marked by referring to the model proposed in ref. 15.

6. Conclusions

The synergistic effects among infrared light irradiation, oxygen vacancy, and size effect on the migration of photo-generated electrons were investigated via the water photolysis test of SnO₂/g-C₃N₄ hybrids. The size of component SnO₂ nanocrystals was tailored from 3.0 nm to 27.0 nm by annealing SnO₂ quantum dots at different temperatures. The oxygen vacancies were demonstrated to increase as the SnO₂ grain sizes reduced. When infrared light was filtered by an insulation filter, the photocatalytic activity for hydrogen generation reduced as the oxygen vacancies content increased. When the infrared light was not filtered, a low-lying photocatalytic activity was observed, suggested a re-excitation of the localized electrons by infrared light. The influence of re-excitation on the electron migration became obviously when the grain size of SnO₂ was smaller than twice its Debye length. The present study allows one to the design a new type hybridized photocatalyst as well as utilize electrons more efficiently in photocatalytic or other catalytic reactions.

Acknowledgements

This work was financially supported by NSFC (21025104, 21271171, and 91022018).

References

1. J. Wang, P. Liu, X. Fu, Z. Li, W. Han and X. Wang, Langmuir, 2009, 25, 1218–1223.
2. Y. Tang, H. Zhou, K. Zhang, J. Ding, T. Fan and D. Zhang, Chem. Eng. J., 2015, 262, 260–267.
3. Z. Wei, Y. Liu, J. Wang, R. Zong, W. Yao, J. Wang and Y. Zhu, Nanoscale, 2015, 7, 13943–13950.
4. H. Tan, Z. Zhao, W. b. Zhu, E. N. Coker, B. Li, M. Zheng, W. Yu, H. Fan and Z. Sun, ACS Appl. Mater. Interfaces, 2014, 6, 19184–19190.
5. T. Xia, Y. Zhang, J. Murowchick and X. Chen, Catal. Today, 2014, 225, 2–9.
6. D. J. Norris, A. L. Efros and S. C. Erwin, Science, 2008, 319, 1776–1779.
7. K. Huang and A. Rhys, Proc. R. Soc. London, Ser. A, 1950, 204(1078), 406–423.
8. X. Li, J. Yu, J. Low, Y. Fang, J. Xiao and X. Chen, J. Mater. Chem. A, 2015, 3(6), 2485–2534.
9. P. Y. Kuang, Y. Z. Su, G. F. Chen, Z. Luo, S. Y. Xing and N. Z. Q. Liu, Appl. Surf. Sci., 2015, 358, 296–303.
10. X. Li, J. Yu and M. Jaroniec, Chem. Soc. Rev., 2016, 45(9), 2603–2636.
11. L. Z. Liu, J. Q. Xu, X. L. Wu, T. H. Li, J. C. Shen and P. K. Chu, Appl. Phys. Lett., 2013, 102(3), 031916.
12. Y. P. Zhu, T. Z. Ren and Z. Y. Yuan, ACS Appl. Mater. Interfaces, 2015, 7(30), 16850–16856.
13. X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8(1), 76–80.
14. R. Yin, Q. Luo, D. Wang, H. Sun, Y. Li, X. Li and J. An, J. Mater. Sci., 2014, 49(17), 6067–6073.
15. J. D. Prades, J. Arbiol, A. Cirera, J. R. Morante, M. Avella, L. Zanotti, E. Comini, G. Faglia and G. Sberveglieri, Sens. Actuators, B, 2007, 126(1), 6–12.
16. S. Ghosh, K. Das, K. Chakrabarti and S. K. De, Dalton Trans., 2013, 42(10), 3434–3446.
17. R. Kobayashi, S. Tanigawa, T. Takashima, B. Ohtani and H. Irie, J. Phys. Chem. C, 2014, 118(39), 22450–22456.
18. S. V. Kilina, A. J. Neukirch, B. F. Habenicht, D. S. Kilin and O. V. Prezhdo, Phys. Rev. Lett., 2013, 110(18), 180404.
19. C. Canevali, N. Chiodini, P. DiNola, F. Morazzoni, R. Scotti and C. L. Bianchi, J. Mater. Chem., 1997, 7(6), 997–1002.
20. J. Dugas and C. Rey, J. Phys. Chem., 1977, 81, 1417–1419.
21. M. Setvin, U. Aschauer, P. Scheiber, Y.-F. Li, W. Hou, M. Schmid, A. Selloni and U. Diebold, Science, 2013, 341(6149), 988–991.
22. C. Fettkenhauer, G. Clavel, K. Kailasam, M. Antonietti and D. Dontsova, Green Chem., 2015, 17(6), 3350–3361.
23. Z. Shi, S. Zhang, B. Wu, X. Cai, J. Zhang, W. Yin, H. Wang, J. Wang, X. Xia, X. Dong, B. Zhang and G. Du, Appl. Surf. Sci., 2012, 258(22), 8673–8677.
24. H. Peelaers, E. Kioupakis and C. G. Van de Walle, Appl. Phys. Lett., 2012, 100(1), 011914.
25. A. Das, V. Bonu, A. K. Prasad, D. Panda, S. Dhara and A. K. Tyagi, J. Mater. Chem. C, 2014, 2(1), 164–171.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra15932j

---