1 Department of Chemistry, School of Science, Tianjin University,Tianjin 300072, China
2 CAS Center for Excellence in Nanoscience, National Center for Nanoscience and Technology, CAS Key Laboratory of Standardization and Measurement for Nanotechnology, Beijing 100190, China
Shengbin Lei, Lingyu Piao
Show less
History+
Received
Revised
Published Online
2020-04-29
2020-06-13
2020-09-23
PDF
Abstract
H2O2 is widely used in the fields of chemical industry and environmental protection. The only product of its decomposition is water, which is environmentally friendly and conducive to the coordinated and sustainable development of production. The industrial synthesis of H2O2 is mainly indirectly through the anthraquinone method, which consumes a lot of energy and pollutes the environment. The preparation of H2O2 directly from the mixture of H2 and O2 has great safety risks and requires a large amount of H2. The method of converting O2 and H2O into H2O2 through photocatalytic technology avoids the direct mixing of H2 and O2, and uses endless solar energy as an energy source, which has attracted much attention in recent years. This article summarizes the research progress of photocatalytic reduction of O2 to H2O2, compares and analyzes the reaction performance and control measures of different catalytic systems, such as g-C3N4, TiO2, and other photocatalysts, and the mechanisms of photocatalytic preparation of H2O2. In the end, the development of this field is prospected.
Contents
1 Introduction
2 Detection of HO and evaluation of catalyst activity
H2O2 is an environmentally friendly oxidant with high atomic efficiency (47.1 wt% active oxygen). It has a high oxidation potential (EO=1.763V,pH=0;EO=0.878V,pH=14) over the full pH range[1]. H2O2 is mainly used in chemical synthesis, disinfection, bleaching agent, liquid propellant, Fenton reaction and so on. The by-product of the reaction is water, which is environmentally friendly, so it has attracted much attention[2⇓⇓~5][6]. There are two ways to synthesize H2O2: (1) Anthraquinone method: H2O2 is continuously produced by the oxidation of anthrahydroquinone (AHQ), which is obtained by catalytic hydrogenation of 2-ethylanthraquinone (AQ), in air to form anthraquinone. The conditions of this method are mild, but there are side reactions that lead to net consumption of anthraquinone, and complex processes such as regeneration of solution and catalyst, separation and purification of products, which lead to high energy consumption and environmental pollution[7,8]. (2) directly quantitatively mix that H2 and the O2 to synthesize the H2O2. There are two disadvantages to this process. Firstly, the H2/O2 mixture is explosive in a wide range of concentrations, requiring the addition of large amounts of N2, CO2, Ar and other gases to carefully control the ratio of H2 to O2, resulting in a significant reduction in yield. Secondly, the catalyst used to produce the H2O2 is catalytically active for the decomposition of the H2O2 into H2O and O2 and the reaction must be carried out in a suitable solvent, making the process extremely complicated[1].
In recent years, the preparation of H2O2 by photocatalytic reduction of O2 with H2O and O2 as raw materials based on semiconductor photocatalysts has attracted much attention and has a bright future. The complete photocatalytic reaction includes three basic processes: (1) the generation of photogenerated electron-hole pairs, (2) the separation and recombination of carriers, and (3) the redox reaction of adsorbates on the catalyst surface. It is generally believed that the reduction of O2 to H2O2 occurs through a proton-coupled electron transfer (PCET) process involving one-step two-electron direct reduction or two-step one-electron indirect reduction. In the two-step one-electron reduction process, the $O_{2}^{-}$ formed by the binding electrons of the O2 reacts with the H+ in the solvent to form · OOH (eq1,2). Subsequently, · OOH successively combines electrons with H+ to form H2O2(eq3,4). In addition, the O2 can also be directly reduced by eq5 in one step to obtain the H2O2. No matter which path produces H2O2, it can be expressed by the total photocatalytic reaction eq6, and the standard Gibbs free energy of the reaction is 117 kJ/mol.
$O_{2}+e^{-} \rightarrow O_{2}^{-}$
$O_{2}^{-}+H^{+} \rightarrow \cdot HOO$
$\cdot HOO+e^{-}+H^{+} \rightarrow HO_{2}^{-}$
$HO_{2}^{-}+H^{+} \rightarrow H_{2}O_{2}$
$O_{2}+2e^{-}+2H^{+} \rightarrow H_{2}O_{2}$
$O_{2}+2H_{2}O \xrightarrow{h\nu} 2H_{2}O_{2}$
However, there are two competitive reactions in the reduction of O2, namely, the one-electron reduction of O2 and the four-electron reduction of, which lead to the decrease of H2O2 yield. Due to the different reaction mechanism between different photocatalysts and O2, the two-electron reduction selectivity of O2 is quite different. For example, pure TiO2 has a selectivity of less than 2% in a H2O solvent, much lower than pure g-C3N4 with a selectivity of 90%. In addition, the generated H2O2 will be adsorbed on the surface of the photocatalyst and decomposed into H2O(eq7) or O2(eq8) by electron reduction or hole oxidation, resulting in a decrease in the total yield of H2O2.
$H_{2}O_{2}+2e^{-}+2H^{+} \rightarrow 2H_{2}O$
$2H_{2}O_{2}+2h^{+} \rightarrow O_{2}+2H^{+}$
From the thermodynamic analysis (Figure 1), in order to make the process of photocatalytic reduction of O2 to H2O2 proceed smoothly, the conduction band energy level of photocatalyst should be more negative than the two-electron reduction potential (0.68 V) of O2. For many semiconductor catalysts, this condition is not harsh. However, it should be noted that the reduction of O2 to H2O and the decomposition of H2O2 constitute two competing reactions. Therefore, in order to improve the efficiency of photocatalytic reduction of O2 to H2O2, attention should be paid to the kinetics of the process. There are many factors affecting the reaction kinetics, including the material system and the reaction environment, and the selection of suitable catalyst materials is the key.
In recent years, in addition to traditional TiO2 materials, photocatalysts that can achieve photocatalytic reduction of O2 to H2O2 include g-C3N4, BiVO4, MOFs, various graphene composites and so on[9]. At the same time, researchers have also improved the yield of H2O2 by adjusting the morphology and band gap of semiconductor materials, doping, loading promoters and other strategies[10⇓⇓⇓~14]. In this paper, several typical photocatalysts for photocatalytic reduction of O2 to prepare H2O2 were reviewed and analyzed.
2 Detection and Activity Evaluation of Heme Oxygenase
There are great differences in the efficiency of H2O2 produced by different photocatalysts in different environments. Therefore, the accurate quantification of H2O2 is very important to evaluate the performance of photocatalysts. At the same time, it is of great significance to clarify the evaluation index of the reaction for the performance comparison between different photocatalysts and the direction of improving the activity of photocatalysts. In this review, we investigate the H2O2 detection methods and activity evaluation methods of photocatalytic O2 reduction process in the literature published in the past decade, and summarize as follows.
2.1 Method for detecting H2O2
The detection of H2O2 can be traced back to 1927. With the continuous development of detection technology, the detection methods of H2O2 concentration are more abundant, which can be divided into the following categories:
(1) Redox titration, such as KMnO4 titration. The H2O2 is reductive in the KMnO4 solution and the resulting Mn2+ catalyzes the reaction (eq9). In addition, the KMnO4 itself can be used as an indicator to determine the end point of titration according to the color change of the solution. Finally, the H2O2 content is calculated from the consumed KMnO4.
(2) Spectrophotometry, such as iodometric titration, N, N-diethyl-1,4-phenylenediamine sulfate (DPD) method and o-tolidine method[15⇓⇓⇓~19]. When the H2O2 is detected by iodometric titration, the I- in the solution is oxidized to $I_{3}^{-}$(eq10) by the H2O2. Subsequently, the absorbance of the $I_{3}^{-}$ was measured spectrophotometrically at 360 nm. The linear relationship between the absorbance and the concentration of the H2O2 was obtained from the calibration curve. The method is applicable to the H2O2 in the concentration range of 0.1 ~ 500 μM.
The DPD method is based on the horseradish peroxidase catalyzing the oxidation of DPD by a H2O2 to a DPD+, which exhibits a fairly stable color with a large molar absorption coefficient (2.1×104M-1·cm-1). The absorption peak at 551 nm was detected by ultraviolet spectrophotometer, and the H2O2 was quantitatively analyzed. The o-tolidine method uses a special peroxide indicator o-toluidine to titrate the H2O2 to form a blue oxidation product, which has a typical absorption peak at 438 nm in the ultraviolet-visible spectrum and can determine the content of the H2O2.
(3) a fluorescent probe method, such as a POHPAA method. Using horseradish peroxidase as a catalyst, H2O2 reacted with p-hydroxyphenylacetic acid to form a fluorescent dimer. The dimer was quantitatively analyzed by fluorescence spectrophotometer at the excitation wavelength of 316.5 nm and the emission wavelength of 408.5 nm, and the amount of H2O2 was calculated accordingly.
2.2 Activity evaluation index
It is still a challenge to accurately and quantitatively measure the amount of H2O2 produced in the photocatalytic process due to the adsorption and decomposition of H2O2 on the surface of photocatalysts[20]. In general, the activity of a photocatalyst to produce H2O2 is expressed in terms of its rate of production. Because the reaction time, experimental environment and conditions, and the rate units provided by different literatures are different, in order to compare the performance of different photocatalysts, the rate of H2O2 generation is converted into the yield of H2O2 per unit time and per unit amount of photocatalyst, which is expressed by μmol·L·g-1·h-1 or μmol·g1·h-1.
The photocatalytic activity can be evaluated by apparent quantum yield (AQY) under different experimental conditions[19,21]. Quantum yield (QY) refers to the utilization rate of photons in photochemical reactions, which is defined as the number of molecules of reactants produced per photon absorbed. However, in the photocatalytic reaction, the number of absorbed photons is difficult to measure, and the number of incident photons is usually used instead to obtain AQY. AQY can be calculated using the following formula (eq11):
Where the N(H2O2) represents the number of molecules producing the H2O2, and N (X) represents the number of incident photons.
In addition, the solar energy conversion efficiency (SCC) of the photocatalyst can be calculated using a solar simulator in the AM1.5 spectrum. Calculated by formula (eq12):
Where $\Delta G^{0}(H_{2}O_{2})$ represents the standard Gibbs free energy of formation of the H2O2, (117 kJ·mol-1),$n(H_{2}O_{2})$ represents the amount of the substance of the H2O2, $E_{input}$ represents the incident light power, and T represents the reaction time.
3 Catalyst system for photocatalytic reduction of O2;
3.1 Graphite phase carbon nitride (g-C3N4)
g-C3N4 is a kind of organic metal-free layered polymeric material composed of abundant C and N elements on the earth. This new material was predicted by Liu et al. In 1989 using an empirical model of covalent solids and ab initio calculations of bulk modulus[22]. This kind of superhard material predicted by theory was successfully prepared by Niu et al in 1993 by laser ablation of graphite target with strong atomic nitrogen source, and diffraction analysis showed that the carbonitride microcrystalline structure was $\beta-C_{3}N_{4}}$[23]. In 1996, Teter et al. Used the first principles to study the structures and properties of five different structures of carbon nitride, including β-C3N4, α-C3N4, g-C3N4, quasi-cubic -C3N4 and cubic -C3N4.The results show that the g-C3N4 is formed by stacking the planar structures ABAB …[24]. In 2009, Wang et al found that g-C3N4 had good performance in photocatalytic decomposition of water and hydrogen, which aroused great interest of researchers in g-C3N4. Since then, it has been widely used in photocatalytic decomposition of water and degradation of organic pollutants. So far, it has become a typical photocatalyst and is very popular in photocatalytic applications[25⇓⇓~28].
The g-C3N4 has the advantages of non-toxicity, stability, low cost, easy preparation and the like. It has a band gap of 2.7 eV, a valence band of Ev=1.4 V(vs.NHE,pH=7) and a conduction band of Ec=-1.3 V(vs.NHE,pH=7), and can absorb ultraviolet and visible light with a wavelength of less than 475 nm. However, the g-C3N4 also has the disadvantages of high electron-hole recombination rate and slow charge mobility[29]. It was Shiraishi et al. Who first discovered that g-C3N4 has reductive properties. In 2014, they found that g-C3N4 has a very high selectivity (about 90%) for two-electron photoreduction of O2, and elaborated the process of producing H2O2 (Figure 2). When light irradiates on the surface of g-C3N4, the photogenerated electrons are mainly concentrated at the 1 and 4 positions, and the O2 on the surface are reduced to superoxide radicals by combining with the electrons, and then rapidly combine with the electrons at the 4 position to form 1,4-internal peroxides. The formation of 1,4-internal peroxides can effectively promote the two-electron reduction of O2 and inhibit the processes of four-electron reduction to H2O and one-electron reduction to · OOH[30]. In addition, the hydration between H2O and H2O2 can inhibit the decomposition of H2O2, and the stability of H2O2 can be improved by increasing the amount of H2O[31].
Table 1 summarizes the research progress of photocatalytic reduction of O2 by g-C3N4 materials to prepare H2O2 in recent years according to the order of publication time. There are more than 20 articles on this system, mainly after 2014. Most of the original texts do not give the yield per unit time and per unit catalyst dosage. Table 1 converts the performances given in their respective works for easy unified comparison. The performance of the catalyst varies greatly due to the different conditions of use. The optimal value of H2O2 production efficiency was 1 283 μmol/ (G · H). There are few literatures giving the quantum efficiency and light conversion efficiency, in which the highest AQY is 7.2%, and the SCC is 0.27%, which is generally low. The amount of catalyst used ranges between 1~4 g·L-1, with 1 g·L-1 being used in most cases. Most studies did not use adjuvants or sacrificial agents. The light source mainly uses 420 nm visible light. It can be seen that in the process of photocatalytic research, comprehensive reaction conditions and evaluation index data should be provided as far as possible to facilitate the consistent comparison of scientific research work.
In order to enhance the absorption of visible light, improve the carrier migration rate, promote the two-electron reduction selectivity of O2 and inhibit the decomposition of H2O2, the modification strategies used in the literature for the g-C3N4 system mainly include doping, loading additives, constructing defects and controlling the morphology.
3.1.1 Doping
The most commonly used modification strategy of g-C3N4 system is to dope organic, inorganic non-metallic elements and metal ions. Doping can change the band gap of g-C3N4, enhance the absorption of visible light, and adjust the band position. When the bottom of the conduction band of the g-C3N4 is located between the one-electron reduction potential of the O2 (E = -0.33 V) and the two-electron reduction potential of the O2 (E = 0.68 V), the reduction ability of the g-C3N4 is not enough to reduce the O2 by one electron, so that the competitive reaction can be inhibited.
(1) Organic doping
Shiraishi et al doped pyromellitic diimide (PDI), biphenyl diimide (BDI) and trimellitic diimide (MTI) into g-C3N4 respectively to obtain corresponding g-C3N4/PDI, g-C3N4/BDI and g-C3N4/MTI polymers[17][47][45]. With the increase of doping amount, the valence band value of the g-C3N4 is more positive, and the oxidation ability is enhanced, which is conducive to the occurrence of the H2O oxidation process in the final step. But that catalytic activity of the above system is still low[17]. Unlike the g-C3N4/PDI in which both electrons and holes are located in the melem unit, the electrons and holes formed by the g-C3N4/BDI after photoexcitation are located in the melem and BDI units, respectively, and this spatial separation of charges inhibits the recombination of carriers. Under visible light, the amount of H2O2 produced by g-C3N4/BDI is 1.3 times that of g-C3N4/PDI(30μmol)[47]. On this basis, the performance of the g-C3N4 can be further optimized by changing the doping structure. After doping the three-branch MTI, the density of the g-C3N4/MTI polymer structure is much higher than that of the two-branch g-C3N4/PDI, which is beneficial to the charge transfer in the layer. At the same time, the high-density structure is beneficial to the π-π overlap between melem and MTI layers, which enhances the charge transfer between layers. The yield of H2O2 was 27. 5 μmol and the transformation efficiency of SCC was 0. 18%, which were higher than those of g-C3N4/PDI and g-C3N4/BDI[45]. At present, the performance of organic-doped photocatalysts is generally low, and the efficiency of H2O2 generation is not significantly improved compared with pure g-C3N4. However, it provides a feasible way to improve the performance of g-C3N4. Moreover, we attribute the enhanced photocatalyst performance to the fact that the high electron affinity shifts the oxidation and reduction potentials in a positive direction, and promotes the carrier separation and H2O oxidation processes.
(2) Inorganic dopant
Organic doping has higher requirements for its structure, and there are few such studies at present. In addition to the above organic doping, the single element or multi-element co-doping of inorganic metal ions and non-metallic elements is more studied. Among them, C, O, N, P and other non-metallic elements are mostly doped; Metal elements are mostly K, Na and so on.
Xiang et al. Studied the g-C3N4 with different amounts of C doping, and found that C doping had no significant effect on the band gap of the g-C3N4, but could change the position of the conduction band and valence band. When the doping amount is g-C3N4-Cx(x=0.006), both the conduction band and the valence band are positively shifted. At this point, the conduction band potential is still far less than 0. 68 V, which is enough for two-electron O2 reduction. With the increase of the oxidation potential, the ability of the g-C3N4 to oxidize the H2O, which is the rate-determining step, is enhanced without losing the ability to reduce the O2, and the yield of H2O2 is increased to 317 μmol/ (G · H). After five cycles, the activity of the g-C3N4 is slightly decreased (< 3%) compared with that of the fresh sample[15]. Srivastava et al. Obtained C, O co-doped g-C3N4 by copolymerization of urea, thiourea, 2-methylimidazole and other raw materials. Under the condition of not loading cocatalyst and not adding sacrificial agent, the H2O2 yield under visible light is 2.01 μmol · L/ (G · H), and the stability is good. After five cycles, the activity does not decrease significantly[32]. Hu et al. Studied the effect of different O doping amounts on the g-C3N4. O doping can enhance the interlayer interaction of the g-C3N4 and promote the carrier separation. In addition, O doping can shift the conduction and valence band positions from the g-C3N4 of -1. 06 V and 1. 62 V to 0. 37 V and 2. 15 V, respectively, which promotes the two-electron reduction of oxygen and the transformation of OH- into hydroxyl radicals, and further recombines into H2O2[38].
Tian et al synthesized K and P doped g-C3N4 catalysts. Doping of K and P can not only enhance the absorption of ultraviolet light, but also make the absorption edge shift to the visible region. After doping, the conduction band moves to the negative direction to increase the reduction potential, which is conducive to the reduction of two electrons in the O2 to H2O2. In addition, the stability of the photocatalyst was improved by doping, and the activity was almost unchanged after 4 cycles[33]. Zhao et al. Prepared g-C3N4-SiW11 catalysts by covalently binding heteropolyacid salts to g-C3N4 through organic linkers. Under the irradiation of simulated sunlight (AM1.5), the catalyst has good catalytic activity (15.2μmol·h-1)[42]. Hu et al. Doped Ni into g-C3N4 nanofibers, Ni and N atoms were coordinated into the interstitial sites, and the formed Ni (I) -N bond could be used as a chemisorption site to activate O2 molecules, and as a "bridge of electron transfer" to promote electron transfer from the catalyst to the adsorbed O2 molecules, so as to promote their combination with H+ to produce H2O2[44]. Choi et al. Formed a g-C3N4 doped with K, P and O by adding phosphate. The doping of phosphate enhances the absorption of light, prolongs the lifetime of carriers, and forms charge trapping sites, which effectively promotes the transfer of interface electrons. At the same time, the decomposition of H2O2 on the surface of the catalyst was inhibited, and the yield of H2O2 was finally improved, and after four times of 48 H recycling, the activity was close to the initial activity[29].
To sum up, doping can regulate the performance of photocatalyst mainly by selecting the appropriate dopant and the appropriate doping amount. The yield of the H2O2 is improved by changing the energy band position to promote the reaction kinetic process. However, it should be noted that doping can also be used as a carrier recombination center, and the doping amount needs to be accurately controlled. At present, due to the great difference in experimental conditions, the performance of various photocatalysts is only roughly compared. It can still be seen that the improvement of g-C3N4 performance by inorganic doping is much higher than that by organic doping, and the design idea of dopants is simpler. However, it is undeniable that organic doping is a very promising method for photocatalyst modification.
3.1.2 Defect Regulation
Vacancy defects in semiconductor photocatalysts can trap photogenerated electrons and inhibit carrier recombination. In addition, it can also be used as a reactive center to enhance the adsorption and activation of gas molecules, promote the transfer of electrons from the catalyst to the O2, and carry out reduction reactions.
Zhang et al prepared g-C3N4 containing N vacancies by thermal reduction method. The N vacancy is located at the tertiary N, and the formation of the N vacancy can introduce an intermediate energy level and broaden the spectral absorption range. At the same time, the formation of C C N functional groups at the triazine heterocycle leads to the narrowing of the band gap and the positive shift of the conduction and valence bands. C C N is a strong electron-withdrawing group, which attracts electrons to migrate to C N functional groups, so that oxidation and reduction reactions occur on melem and C N functional groups, respectively, and the photocatalytic activity is enhanced. However, after the second cycle stability test, the activity of the photocatalyst decreased to 90% of the initial activity[41]. Wang et al. And Zhang et al. Introduced C and N vacancies into the g-C3N4, respectively. Both C and N vacancies can act as active sites to promote the transfer of electrons from the catalyst surface to the adsorbed O2, followed by the reduction reaction to improve the yield of H2O2[46][19]. It is worth noting that defects can also act as carrier recombination centers and reduce the performance of photocatalysts.
3.1.3 Morphology control
Because of its structural characteristics, g-C3N4 has a high selectivity for the two-electron reduction of O2 to produce H2O2. Therefore, it is necessary to improve the catalytic activity while maintaining its high selectivity in the modification design. It is generally believed that the larger the specific surface area of the catalyst, the higher the reaction activity. At present, there are many ways to increase the specific surface area of catalysts by morphology control, such as the synthesis of nanorods, nanofibers, mesoporous g-C3N4 and so on. Mesoporous $g-C_{3}N_{4}$ with different specific surface areas were synthesized by Shiraishi et al[30]. The mesoporous structure can reflect light for many times, enhance the absorption of light, and increase the specific surface area of the photocatalyst. When the specific surface area of the catalyst increased to 56 and 160 m2·g-1, the catalyst showed higher catalytic performance. However, when the temperature is further increased to the 228 m2·g-1, the selectivity of the generated H2O2 is reduced, and the catalytic performance is reduced. The selectivity of g-C3N4(10), g-C3N4(56), g-C3N4(160) and g-C3N4(228) with different specific surface areas to H2O2 is 92%, 72%, 58% and 42% respectively. In addition, the increase of specific surface area can increase the number of primary amines on the surface, which can promote the four-electron reduction competition reaction of. At the same time, as a basic active site, primary amine has a strong interaction with acidic H2O2, which makes the H2O2 decompose on the surface and reduces the yield of H2O2[30]. We attribute the nature of morphology control to the efficient formation of carriers and their timely reaction with the surface O2, which avoids recombination and improves the H2O2 yield. At present, there are few studies on the effect of morphology control on photocatalytic H2O2 production, which may be limited by the current methods of controlling the morphology of photocatalysts. Therefore, the controllable synthesis of photocatalysts with different morphologies will be beneficial to the progress in this field.
3.1.4 Promoter
Loading noble metal Pt, Au, Ag and other additives is an effective way to improve the carrier separation efficiency. It is generally believed that the semiconductor photocatalyst and the noble metal form a Schottky barrier at the interface, which promotes the separation of electron-hole pairs. When only 0.01 wt% Au is loaded on the surface of the g-C3N4, Meng et al., the H2O2 yield can reach 16.89 μmol/ (G · H)[35]. Highly dispersed Au can directly reduce O2 to H2O2. Au ensure that stability of the H2O2 relative to the decomposition of the H2O2 by Pt. Although noble metal promoters show high activity, they are limited due to their high cost. Researchers are trying to find low-cost transition metals (such as Co, Ni, Cu) or non-metals as substitutes. Zhang et al. Successfully synthesized NiCoP/g-C3N4 composite catalysts by continuous in-situ growth and phosphating method. NiCoP has better conductivity and electronic structure than Ni2P or CoP[37]. It can not only enhance the absorption of visible light and inhibit the recombination of electrons and holes, but also provide more active sites. In addition, the stability of NiCoP was better, and the activity did not decrease after 6 cycles. Shiraishi et al. Used non-metallic materials such as rGO and BN as co-catalysts to compound g-C3N4/PDI with rGO or g-C3N4/PDI with rGO and BN, and the SCC conversion efficiency was as high as 0.2% and 0.27%, respectively[48][21]. The photogenerated electrons formed on the g-C3N4 or the g-C3N4/PDI migrate to the rGO and selectively reduce the O2 to form the H2O2 by forming 1,4-endoperoxide. The valence band of BN (2. 10 V vs. RHE) is more negative than that of V vs. RHE (2. 46 V vs. RHE) after recombination with V vs. RHE. Therefore, electrons and holes migrate to rGO and BN, respectively, thereby promoting the separation of electrons and holes and improving the catalyst activity. To sum up, we attribute the role of the promoter to effectively promoting the separation of electrons and holes and providing reduction sites. In addition, we believe that the development of low-cost, pollution-free, non-noble metal additives that can effectively inhibit the decomposition of H2O2 is a very attractive research direction.
3.2 TiO2 system
In 1972, Fujishima and Honda first discovered the phenomenon of water decomposition on TiO2 electrodes under near ultraviolet light, making TiO2 materials the most classical materials in the field of photocatalysis[49]. With the help of theoretical calculation and in-situ characterization technology, the mechanism of photocatalytic reduction of O2 by TiO2 to produce H2O2 was studied in depth, and the mechanism was suggested as follows[50]. The Ti3+ sites on the surface of the TiO2 adsorb O2 and reduce them to superoxide radicals. Subsequently, the hydrogen peroxide formed by the combination of superoxide radicals with electrons is protonated to obtain H2O2(eq13~16). Because hydroperoxide is unstable on the surface of TiO2, it is easy to combine electrons and decompose into inactive OH-(eq17,18).
In addition, when there is a sacrificial agent in the TiO2 system, the mechanism is as shown in the figure. After the 3,TiO2 is excited by light, the electrons are transferred from the lattice oxygen to the adjacent Ti4+, forming a charge-separated state. The electron-hole pair removes the alpha-hydrogen of the alcohol, producing the alcoholate. The beta hydrogen of the species is then removed by photoexcitation to produce carbon radicals. The electrons on the Ti3+ reduce the O2 and produce superoxide radicals. In the case of benzyl alcohol, the superoxide radical binds to an adjacent carbon radical and forms a O2 bridged complex. Heterolytic cleavage of the complex produces aldehydes with exchanged O atoms and laterally coordinated peroxides. Finally, the side peroxide protonates to produce the H2O2 and complete the photocatalytic cycle. In contrast, for fatty alcohols, the combination of carbon radicals and superoxide radicals does not occur efficiently, and the release of aldehydes does not exchange O atoms. The superoxide radical remaining on the surface of the TiO2 is reduced to -OOH by electrons. Through further electron reduction, it is decomposed into OH-. The above mechanism indicate that that production of peroxides bridge by O2 via heterolytic cleavage promotes efficient and selective production of H2O2[50,51].
Table 2 summarizes the research progress of photocatalytic O2 reduction of TiO2 materials to prepare H2O2, totaling 9 papers. The research of this system is mainly concentrated in the past 20 years. The photocatalytic activity of the photocatalyst ranged from 0. 96 μmol/ (G · H) to 2650 μmol/ (G · H), which varied greatly with different conditions. Most of the original text does not give the output per unit time and per unit catalyst dosage. Table 2 converts the performance and is easy to compare. The amount of catalyst to be use is in that range of 0.5 g·L-1 to 1 g·L-1, with 0.5 g·L-1 being use more frequently. Unlike the C3N4 system, the TiO2 system uses more additives and sacrificial agents. The light source mainly uses ultraviolet light. There are few literatures giving the quantum efficiency and light conversion efficiency, and the highest value of AQY in the report is 17%, which is significantly improved compared with the C3N4 system.
In the literature, surface passivation, cocatalyst loading and morphology control were mainly used to improve the activity of O2 reduction to H2O2 in TiO2 system.
3.2.1 Surface passivation
The hydroxyl group (Ti-OH) on the surface of the TiO2 is easily oxidized by the H2O2. Coordination of anions, cations or polyacids on the surface of TiO2 to inhibit the formation of Ti-OOH is the earliest method used to improve the yield of H2O2. Maurino et al. Found that the Zn2+ had a great impact on the formation of H2O2 when they studied the anatase/water system. When the pH value is higher, the Zn2+ is almost completely coordinated with -Ti-OH, so that the surface of -Ti-OH is blocked, the further reaction of the generated H2O2 with the surface is inhibited, and the yield of the H2O2 is improved[56]. Similarly, the addition of Cu2+[58,59] or surface fluorination can also inhibit the decomposition of H2O2 and improve the yield of H2O2[57]. Chen et al. Found that the presence of polybasic acids can change the reaction path of O2 reduction. One carboxyl group of the polybasic acid is tightly adsorbed on the surface of the TiO2, while the remaining carboxyl groups are suspended on the surface to form directional hydrogen bonds to provide protons, thus bypassing the one-electron reduction process. At the same time, the adsorption of polybasic acids can block the further decomposition of H2O2 on the surface of TiO2[60]. In a word, surface blocking is a simple and effective way to inhibit the decomposition of H2O2.
3.2.2 Promoter
The action mechanism of noble metal promoter and photocatalyst is shown in Figure 4. Due to the formation of Schottky barrier, the conduction band electrons of the TiO2 flow to Au, which can inhibit the reduction of Ti-OOH on the surface of the TiO2 by electrons (Ti-OOH+H++e-→Ti-OH+OH-) and improve the yield of H2O2. Au can act as a reduction site for O2 and improve that selectivity of the two-electron reduction. At the same time, Au has a strong adsorption on the H2O2, so that the H2O2 is reduced to hydroxyl and OH-.
Shiraishi et al. formed a heterojunction by loading Au-Ag bimetallic alloy nanoparticles on the surface of TiO2, which effectively promoted the separation of electrons and holes. At the same time, the adsorption energy between Ag and H2O2 in the alloy is more negative (-17.4 kcal·mol-1), which is conducive to the adsorption of H2O2 on the Ag surface, while the electrons needed for photocatalytic reduction reaction are mainly concentrated in Au, thus spatially separating the decomposition sites and generation sites of H2O2 and avoiding the reduction of generated H2O2. In addition, the stability test showed that the activity remained unchanged after 3 cycles[10]. Kim et al. Proposed a new method, as shown in Figure 5, to improve the electron density of Pd by coordinating the organic ligand fixed on the surface with Pd. The Pd on the TiO2 has higher affinity for the O2, the O2 is combined with a single electron to form the superoxide, and the Pd is negatively charged after being regulated by the electron of the organic ligand,It has a strong affinity for H+ and can attract H+ to gather on the surface of TiO2, thus promoting the H2O2 of two-electron reduction of O2. The H2O2 yield is 150μmol·h-1 under AM 1.5 light source[52].
Non-noble metal carbon materials have attracted great attention because of their low price, high conductivity, large specific surface area and non-toxicity. Long et al. Used S and N co-doped graphene quantum dots (SN-GQDs) to regulate the TiO2. After S and N doping, the distribution of electrons on the graphene quantum dots is rearranged, which provides conditions for the reduction of two-electron O2. The O2 is adsorbed on the S and Cl atoms with higher positive charge density in the SN-GQDs, and then the photoexcited electrons in the TiO2 captured by the SN-GQDs react with the adsorbed O2 on the S atom to form · OOH, and finally the · OOH combines with an electron and a proton to form a H2O2[55]. To sum up, the loading of promoters can not only capture photogenerated electrons and act as reduction sites, but also change the reaction pathway and promote the two-electron reduction of O2.
3.3 Other catalytic system
After decades of research, the preparation methods of classical photocatalysts such as TiO2 are relatively mature, and the catalytic mechanism is relatively in-depth. It is widely used because of its good photocatalytic activity and high stability, but new photocatalytic materials (BiVO4, BiOCl, MOFs, etc.) Have incomparable advantages over TiO2. For example, the monoclinic BiVO4 has a suitable band gap (E = 2.4 eV), absorbs visible light, and has good stability in aqueous solution [68]. Shiraishi et al. reported the synthesis of H2O2 using Au/BiVO4 as a photocatalyst. The conduction band bottom potential (0. 02 V vs. NHE, pH = 0) of the V vs. NHE is between the one-electron reduction potential (-0. 13 V) and the two-electron reduction potential (0. 68 V), which is beneficial to the two-electron reduction of the O2 and thus more effective in generating H2O2[62].
BiOCl has attracted attention due to its layered structure and effective internal electric field which can promote charge separation. Using BiOCl as a photocatalyst and HCOOH as a sacrificial agent, Wang et al. Proposed a new H2O2 generation path, in which the holes generated in the valence band of BiOCl combined with HCOO- to generate HCOO ·, and HCOO · further reacted with OH- to generate OH · as the source of H2O2. In addition to HCOOH, the O2 also plays an important role, which not only promotes the circulation between Bi and Bi3+, but also effectively reduces the recombination of carriers, so that the process has a high H2O2 yield (1020 μmol/H). During the stability test, because the surface of the used BiOCl was carbonized, the generated CO2[63].
Metal-organic frameworks (MOFs) are a kind of crystalline porous materials with periodic network structure, which are formed by the self-assembly of inorganic metal centers (metal ions or metal clusters) and bridged organic ligands. One of the attractive features of MOFs is that their structures can be tailored to achieve desired physical and chemical properties. Yamashita et al. report that application of a photoactive metal-organic framework (MOF)MIL-125-NH2 in the photocatalytic reduction of O2 to H2O2 in a benzyl alcohol/water two-phase system. As shown in the figure, under the irradiation of visible light, the ligand transfers charge to the cluster, reduces the TiIV in the Ti8O8(OH)4 to TiIII, and generates holes in the organic linker. By replacing the amino group in the MIL-125-NH2 with an alkyl chain, the MOF is rendered hydrophobic and transferred to the organic phase. The O2 combines with electrons in the organic phase to generate superoxide anion, and then transfers to the aqueous phase to combine with hydrogen ions to generate H2O2. The method skillfully utilize that hydrophilic and hydrophobic characteristics of substance, effectively avoids the decomposition of the H2O2 on the surface of the catalyst, and improves the yield of the H2O2[64].
4 Conclusion and prospect
Photocatalytic reduction of O2 to produce H2O2 is an environmentally friendly and promising technology as an alternative to the current industrial production methods with high energy consumption and pollution. This review focuses on the basic principles of photocatalytic reduction of O2 to H2O2 by g-C3N4 and TiO2, and various modification methods to improve the efficiency, and briefly introduces several other new photocatalysts. So far, a lot of research has been done on the classical photocatalysts, and the modification strategies mainly include controlling the morphology, loading promoters, doping, forming defects, surface blocking and so on, in order to enhance the light absorption of photocatalysts, carrier separation and inhibit the decomposition of H2O2. Although encouraging results have been achieved, there is still a certain distance to achieve the goal of industrialization and H2O2.
At present, the overall yield of photocatalytic reduction of O2 to prepare H2O2 is still low, the mechanism is not comprehensive enough, and the stability is rarely investigated. It can be expanded and improved from the following aspects:
(1) the utilization rate of sunlight by the catalyst is improved. The absorption of light by TiO2 and g-C3N4 is concentrated in the ultraviolet region, which accounts for only 5% of solar radiation. By adjusting the band gap, the absorption wavelength can be extended to the visible region, but part of the redox ability will be lost. The method of constructing a Z-shaped heterojunction can be considered to enhance the light absorption while ensuring the redox capability. Second, the ability to collect light is enhanced by controlling the morphology of the photocatalyst, such as the formation of mesopores. In addition, the photocatalyst can be improved by learning from the absorption of light by green plants in nature.
And (2) that carry separation efficiency is improved. Supported promoters are an effective method to promote electron-hole separation, and noble metal promoters are mainly used at present, but there are problems of high cost and adsorption and decomposition of H2O2, which need to be further improved, such as developing cheap non-noble metal promoters to replace Pt, Au and other precious metal promoters and provide higher catalytic activity.
(3) Selection of light source. It is recommended to use natural light instead of various laboratory light sources as the final evaluation condition. Whether xenon lamp or mercury lamp, most of the literatures do not give the exact light density and intensity to reach the reaction system, which makes it difficult to compare the results of different work.
(4) promote that production of H2O2 and inhibit its decomposition. The reduction ability of photocatalyst is enough to drive the reduction of two electrons in O2, which is the premise of producing H2O2. By adjusting the band gap of the photocatalyst to an appropriate potential, the two-electron reduction of the O2 and the oxidation of the H2O can be simultaneously enhanced to generate the H2O2. In addition, we should constantly look for new angles to ensure that H2O2 are desorbed in time after formation and are not decomposed.
(5) The mechanism of photocatalytic reduction of O2 to H2O2 was studied by theoretical calculation and in situ characterization, which can help to understand the dynamic behavior of photocatalyst surface, improve the strategy and design new and efficient photocatalysts. In addition, due to different reaction environments, the evaluation methods of catalyst performance are not uniform, so the establishment of reasonable performance evaluation criteria is helpful for researchers to compare the performance of different photocatalysts. In view of the above problems, it is expected that future research will continue to deepen and make greater breakthroughs.
PDF
