Keywords

8.1 Introduction

Since the Industrial Age beginning with the invention of the steam engine in the 1760s, human beings have been exploiting fossil resources for energy and development. However, after more than 150 years’ exploitation and utilization, expected depletion of the fossil fuel reserves has emphasized the issue of energy and environmental sustainability [16]. Heterogeneous photocatalysis by semiconductors, as a green technology, has received intense attention and been widely applied to environmental remediation and solar–chemical conversion for further energy supplies, including photodegradation of organic contaminants, photocatalytic CO2 reduction for value-added chemicals, and water splitting for H2 production [1, 2, 725].

The application of photocatalysis in selective organic transformations is not common at first because semiconductor photocatalysis has long been considered as “nonselective” processes, especially in aqueous media [26]. Nevertheless, many researchers have devoted efforts to this subject and have clarified that several reactions can proceed highly efficiently and selectively in various heterogeneous photocatalytic systems [2632]. Following that, photocatalytic selective organic transformations have received growing interest and been recognized as an alternative to conventional synthetic routes for synthesis of fine chemicals through the selection of appropriate semiconductors and control of the reaction conditions [26, 27, 3337]. Compared to traditional synthesis methods [34, 3840], photocatalytic organic synthesis features several unique advantages: (i) it is driven by sunlight as a completely renewable source of energy; (ii) it can go on under milder conditions (room temperature and atmospheric pressure) and avoid environmentally detrimental heavy metal catalysts as well as strong chemical oxidants or reducing agents; (iii) it can facilitate the design of short and efficient reaction sequences, minimizing side processes and leading to high selectivity [29, 3234, 3842].

With the late but fast development, photocatalytic selective organic transformations have been considered to play a major role in promoting the evolution of twenty-first-century chemistry by replacing environmentally hazardous processes with environmental friendly and energy-efficient routes, allowing maximization of the quantity of raw material that ends up in the final product [2634, 41, 43, 44]. To date, several organic transformation reactions are promoted with high selectivity and efficiency on various photocatalytic systems [26, 2934, 3841, 4345]. This chapter will provide an overview of the basic principles and evaluation methods of photocatalytic selective organic transformations, following which recent progress in the significant examples of photocatalytic processes employed for synthetic purposes will be represented, including selective oxidation, reduction, and coupling reactions. In addition, the future research challenges will be concisely discussed.

8.2 Main Process and Basic Principles for Photocatalytic Selective Organic Transformations

In general, semiconductor-based photocatalytic process involves three main steps, as shown in Fig. 8.1, and the conversion of light into chemical energy by selective organic transformations can be described as A + D = A red + D ox. The optimization of every step plays an important role in boosting the photocatalytic efficiency.

Fig. 8.1
figure 1

Schematic illustration of semiconductor-based photocatalytic processes

Specifically, (i) upon light irradiation, the photocatalyst absorbs supra-bandgap photons (≥Eg, Eq. 8.1), and photoexcited electron (e) and hole (h+) pairs are produced in the CB and the VB, respectively (Eqs. 8.2 and 8.3). In this step, the light absorption range and coefficient of photocatalyst should be increased to maximize the light harvesting and thus to harness more photons. In addition, since UV light only accounts for about 4 % of solar spectrum, it is highly desired for a photocatalyst with a narrow bandgap to utilize visible light, which occupies about 43 % of solar light [10, 46]. (ii) The photogenerated electrons and holes are separated and migrated to catalytically active sites at semiconductor surface (or recombined). Typically, photogenerated electron–hole pairs have a recombination time on the order of 10−9 s, while the chemical interaction with adsorbed species has a longer time of 10−8–10−3 s [46]. Therefore, in this step, accelerating the separation and migration of the photoexcited electrons and holes to avoid their recombination is fundamentally important. (iii) An efficient charge separation of the electron–hole pairs allows the respective oxidation and reduction reactions on the particle surface (Eqs. 8.4 and 8.5).

However, undesirable reactions might occur. On one hand, efficient back electron transfer between the primary redox products (Eq. 8.6) will prevent in most cases successive reactions to generate the final redox products. In this regard, efficient photocatalytic systems that can inhibit charge recombination as well as photocatalysts with proper electronic band structure for visible light harvesting and redox reactions are needed.

$$ \begin{array}{l} \mathrm{Band} \mathrm{gap} (eV)={\scriptscriptstyle \frac{1240}{\lambda (nm)}}\\ {}\end{array} $$
(8.1)
$$ \mathrm{P}\mathrm{C} + \mathrm{h}\nu\ \left(\ge \mathrm{Eg}\right)\ \to\ {\mathrm{h}}^{{}^{+}} + {\mathrm{e}}^{-} $$
(8.2)
$$ {\mathrm{h}}^{+} - {\mathrm{e}}^{-}\ \to\ {\mathrm{h}}_{\mathrm{VB}}^{+} + {{\mathrm{e}}^{-}}_{\mathrm{CB}} $$
(8.3)
$$ \mathrm{A}\ \to\ {\mathrm{A}}^{\bullet -} \to {\mathrm{A}}_{\mathrm{red}} $$
(8.4)
$$ \mathrm{D}\ \to\ {\mathrm{D}}^{\bullet +}\to\ {\mathrm{D}}_{\mathrm{OX}} $$
(8.5)
$$ {\mathrm{A}}_{\mathrm{red}} + {\mathrm{D}}_{\mathrm{ox}}\ \to \mathrm{A} + \mathrm{D} $$
(8.6)
$$ {\mathrm{e}}^{-} + {\mathrm{O}}_2\ \to {\mathrm{O}}_2^{\bullet -} $$
(8.7)
$$ {\mathrm{h}}^{+} + {\mathrm{OH}}^{-}\ \left(\mathrm{or}\ \mathrm{H}\mathrm{O}\right)\ \to \bullet \mathrm{O}\mathrm{H} $$
(8.8)

On the other hand, under most circumstances, molecular oxygen (O2) usually functions as an electron acceptor by interacting with the photogenerated electrons on the CB of the photocatalyst, thus forming reactive oxygen species (ROS), superoxide radicals (O2 ·−), an oxidizing agent (Eq. 8.7). In contrast, the holes on the VB react with surface hydroxyl groups or water molecules adsorbed on the surface of the photocatalyst and produce hydroxyl radicals, a different type of ROS acting as a strong oxidizing agent (Eq. 8.8). The ROS are highly nonselective oxidants and can degrade or completely oxidize the organic compounds all the way to the deep oxidation products to CO2 and H2O. Therefore, photocatalytic selective organic transformation to the desirable products by partially oxidizations is often hard to accomplish without the use of a proper electron or hole scavenger. Thus, using appropriate substrates and employing appropriate reaction conditions are of great importance for promoting selective organic transformations. That is, the type of electron–hole scavenger, the photocatalyst, the wavelength and the intensity of light used for irradiation, and the type of reactant and solvent all should be carefully chosen to achieve desired products.

8.3 Efficiency Evaluation of Photocatalytic Selective Organic Transformations

8.3.1 Light-Based Measures

The overall quantum yield and (apparent) quantum yield of photocatalytic process are defined in Eqs. 8.9 and 8.10, respectively:

$$ \mathrm{Overall}\ \mathrm{quantum}\ \mathrm{yield}\ \left(\%\right) = \frac{\mathrm{Number}\ \mathrm{of}\ \mathrm{reacted}\ \mathrm{electrons}}{\mathrm{Number}\ \mathrm{of}\ \mathrm{absorbed}\ \mathrm{photons}} \times 100\% $$
(8.9)
$$ \begin{array}{ll}\left(\mathrm{Apparent}\right)\ \mathrm{quantum}\ \mathrm{yield}\ \left(\%\right)=&\ \frac{\mathrm{Number}\ \mathrm{of}\ \mathrm{reacted}\ \mathrm{electrons}}{\mathrm{Number}\ \mathrm{of}\ \mathrm{incident}\ \mathrm{photons}} \hfill \\ {}& \times 100\%\hfill \end{array} $$
(8.10)

Based on the two equations, it is estimated that the apparent quantum yield should be smaller than the total quantum yield because not all the incident photons can be absorbed. However, this method is not often applied to evaluate selective organic transformations since the “number of reacted electrons” for these selective redox reactions is often difficult to be exactly determined, particularly for selective oxidation reactions, due to the relatively complex reaction process and mechanisms. For example, regarding selective oxidation reactions, both the photogenerated holes and O2 or activated oxygen (e.g., •O2 obtained by accepting the photogenerated electrons) can play the role in oxidizing the substrates.

8.3.2 Product-Based Measures

The most widely employed measure to evaluate the activity of photocatalytic selective organic transformations is the conversion, yield, and selectivity of the organic compounds transformed to target product. The definitions are as follows:

$$ \begin{array}{l}\mathrm{conversion}\left(\%\right) = \left[\left({\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{r}}\right)/{\mathrm{C}}_0\right]\times 100\\ {}\mathrm{yield}\left(\%\right)={\mathrm{C}}_{\mathrm{P}}/{\mathrm{C}}_0\times 100\\ {}\mathrm{selectivity}\left(\%\right) = \left[{\mathrm{C}}_{\mathrm{P}}\left({\mathrm{C}}_0-{\mathrm{C}}_{\mathrm{r}}\right)\right]\times 100\end{array} $$

where C0 is the initial concentration of reactant and Cr and Cp are the concentration of reactant and product at a certain time after the photocatalytic reaction, respectively.

8.4 Photooxidation of Organic Substrates

Oxidation processes play an important role in the production of a wide range of chemicals, where traditional industrial oxidation processes always need stringent reaction conditions and strong oxidants such as mineral acids, chromates, permanganate, hydrogen peroxide, etc [3, 28, 35, 47, 48]. Photooxidation can be cogitated as a possible alternative approach to overcome the harsh conditions as well as the questions of toxicity and corrosiveness.

8.4.1 Hydroxylation of Benzene

Phenol is an important industrial chemical because of its wide usage, ranging from disinfectant, precursor of phenolic resins to preservative for pharmaceutical aid [27, 4956]. Different from the industrial production of phenol from benzene by the multistep cumene process under high temperature and pressure [57], photocatalytic hydroxylation of benzene to phenol can be realized via direct electrophilic addition of hydroxyl radicals (•OH) under mild reaction conditions, since •OH can be generated through the reaction of the photogenerated holes from semiconductor photocatalysts with surface –OH groups or adsorbed H2O molecules. However, the product selectivity is often low since •OH is highly reactive and nonselective with hydroxylated phenols being further oxidized and mineralized. For example, the photocatalytic oxidation of benzene to phenol with TiO2 in aqueous media has been performed by many researchers [5154, 56]. As shown in Scheme 8.1, the phenol produced by direct hydroxylation of benzene via an electrophilic addition of •OH radical formed on the TiO2 surface can be sequentially decomposed by further reaction with •OH radicals [5860] and results in a low phenol selectivity.

Scheme 8.1
scheme 1

Photocatalytic oxidation of benzene to phenol and subsequent decomposition of phenol on TiO2 (Reproduced with permission [26]. Copyright 2008 Elsevier)

Therefore, efforts have been devoted to lower the affinity of phenol on the catalyst surface and thus reduce the possibility of further decomposition and improve the selectivity of phenol. For example, Skirmish’s group has reported a system of mesoporous TiO2 particles (mTiO2) with high phenol selectivity (>80 %) [56]. It is suggested that mesopores on the catalyst are the crucial factor for selective phenol production. As schematically shown in Fig. 8.2, benzene is a hydrophobic molecule and adsorbed well on the inner mTiO2 surface, while hydrophilic phenol is scarcely adsorbed. The •OH radicals formed inside the pores scarcely diffuse out of the pores of mTiO2 since they are deactivated rapidly in a near-diffusion controlled rate [27]. Therefore, they efficiently react with the well-adsorbed benzene, whereas less adsorbed phenol scarcely enters the pores and cannot react with them. In this way, further decomposition of phenol is effectively suppressed and high phenol selectivity is obtained.

Fig. 8.2
figure 2

Schematic representation of the reactivity difference between (a) benzene and (b) phenol on mTiO2 (Reproduced with permission [26]. Copyright 2008 Elsevier)

Similarly, Choi’s group later has reported a method for entrapping titanium oxide nanoparticles into hydrophobically modified mesocellular siliceous foam (MCF), which serves as an efficient photocatalyst for the selective hydroxylation of benzene to phenol. As shown in Fig. 8.3, the interior of the hydrophobically modified MCF (by surface organo-grafting with silylation agent) provides a hydrophobic environment where the reactant benzene molecules are preferentially attracted into mesopores, whereas the hydrophilic product phenol molecules are rapidly released out of the pores before they undergo further oxidative degradation within the MCF cages. Ultimately, the phenol selectivity is significantly enhanced [51].

Fig. 8.3
figure 3

Titanium oxide entrapped in the cage-like mesopores of hydrophobically modified mesocellular siliceous foam (MCF) for the hydroxylation of benzene (Reprinted with permission [51]. Copyright 2011 Elsevier)

In addition to the method of enhancing the catalytic performance of photocatalysts toward hydroxylation of benzene through improving their adsorption and desorption toward reactants and products specificity, efforts in terms of visible light harvesting have been devoted. For example, Huang and co-workers have prepared M@TiO2 (M = Au, Pt, Ag) composites for direct oxidation of benzene to phenol in aqueous phenol under visible light irradiation, where Au@TiO2 with a content of 2 wt% Au exhibits a high yield (63 %) and selectivity (91 %), as shown in Fig. 8.4a. The high photocatalytic activity for the composite is attributed to the enhanced visible light absorption as well as the strongest surface plasmon resonance (SPR) effect (Fig. 8.4b) [52]. Notably, the yield and selectivity of the phenol formation are increased with initially added phenol (ranging from 0 to 18,000 ppm) in water. As schematically shown in Fig. 8.4c, it is proposed that the initially added phenol is partially ionized into phenoxy anions and protons, and the SPR effect allows the electron transfer from the Au nanoparticle (NP) to the TiO2 particle to which it is attached. Compared with benzene, the phenoxy anions are preferred to adsorb on electron-depleted Au NPs due to the opposite charges. Then, the electron-depleted Au NPs under visible light irradiation will oxidize phenoxy anions into phenoxy free radicals, which in turn oxidize benzene to phenol, thus becoming phenoxy anions again. Finally, O2 dissolved in the solution is reduced by the electrons from the conduction band minimum (CBM) of TiO2. Increasing the amount of initially added phenol will generate more phenoxy anions in water, which can lead to more phenoxy free radicals under visible light irradiation, thereby enhancing the oxidation of benzene.

Fig. 8.4
figure 4

(a) Visible light-induced catalytic oxidation of benzene with Au@TiO2-microsphere composites with 1–3 wt% of Au. (b) UV/Vis diffuse-reflectance spectra of Au@TiO2-microsphere composites. (c) Proposed mechanism for the photooxidation benzene into phenol in the presence of phenol under visible light irradiation (Reprinted with permission [52]. Copyright 2011 Royal Society of Chemistry)

Besides, Chen and co-workers have reported that Fe ion-modified porous graphitic carbon nitride (Fe-g-C3N4) is active for the direct oxidation of benzene to phenol in the presence of H2O2 at mild conditions (60 °C, 4 h) in both the presence (with a yield of 4.8 % based on benzene) and absence (with a yield of 1.8 % based on benzene) of visible light irradiation [55]. The enhanced yield of phenol with light irradiation is explained as that the surface-bound -Fe3+ can capture the photoinduced electrons, resulting in -Fe2+ to bind and reduce H2O2 to produce •OH, which is able to oxidize benzene to phenol. Moreover, when loading the Fe-g-C3N4 into the mesoporous system of SBA-15, the yield can be further improved to ~12 %, demonstrating that the activity is related to the exposure of the Fe-g-C3N4 surface structure and the increased surface active sites for benzene activation.

8.4.2 Oxidation of Alcohols

Selective oxidation of alcohols to carbonyls is a kind of rather important reactions both in industrial and laboratory synthesis since the products such as aldehydes and ketone derivatives are widely utilized in the fragrance, confectionary, and pharmaceutical industries [42, 6163]. To date, photocatalytic selective oxidation of alcohols to the corresponding aldehydes and ketones has been performed both in liquid phase and gas phase. Since gas-phase reactions need high temperature, liquid-phase reactions are more appealing in the view of green chemistry. Moreover, its selectivity can be significantly improved by the deployment of organic solvent and visible light harvesting.

By weak bonding surface complexation of heteroatom (X = O, S, or N)-containing substrates, metal oxides (such as Nb2O5 and TiO2) could straightforwardly serve as the platform for visible light-induced organic reactions. For example, Shishido and co-workers have reported that the photooxidation of 1-pentanol can proceed over Nb2O5 under light irradiation (>390 nm), as shown in Scheme 8.2a [64]. The mechanism is proposed that alcohol is adsorbed onto Nb2O5 as an alcoholate species in the dark, which can be activated by visible light and transfer electrons to the conduction band reducing Nb5+ to Nb4+, resulting in the formation of the alcoholate species and a hydroxy group. And then the alcoholate species can be photoactivated to generate an alkenyl radical, which subsequently is converted to a carbonyl compound and desorbed, with the reduced Nb4+ sites reoxidized via the reaction with molecular oxygen.

Scheme 8.2
scheme 2

(a) Photooxidation of 1-pentanol with Nb2O5 [64]. (b) Photooxidation of benzyl alcohol with various substitution groups with TiO2 in CH3CN [65, 66]

Likewise, the oxidation of alcohols to corresponding carbonyl compounds with O2 on anatase TiO2 has been achieved under visible light irradiation (Scheme 8.2b) [65, 66]. It is proposed that the surface complexes formed by the interaction of the –CH2OH group or possibly the phenyl ring of benzyl alcohol with the surface –OH group can induce absorption in the visible region [66]. Upon visible light irradiation, the surface complex is photoexcited to form holes (h+) and electrons (e), and the holes can abstract hydrogen atoms from the –CH2OH group of benzyl alcohol. Subsequently, the photoinduced benzyl alcoholic radicals may automatically release another electron to form benzaldehyde due to the current-doubling effect. Excellent conversion of 99 % and selectivity of 99 % are achieved for a series of substituted benzyl alcohols except when the substituent is –OH. This is because the substituted –OH group provides extra coordination sites for interacting with TiO2, consequently leading to the destruction of the phenyl ring [67]. This mechanism is also applicable to rutile TiO2 nanorods, which have been used for aerobic oxidation of benzyl alcohols to benzaldehydes, yielding a high selectivity of 99 % under visible light irradiation [68].

Localized surface plasmon resonance (SPR) arising from a resonant oscillation of free electrons coupled by light has been extensively studied and combined with semiconductors as promising heterogeneous photocatalysts because of their strong light absorption in the visible region. Tsukamoto et al. have prepared Au/TiO2 via the deposition–precipitation method from H4AuCl4 and TiO2 (Degussa, P25), which shows enhanced activity toward selective aerobic oxidation of alcohols relative to the reaction in the dark at room temperature [69]. It is proposed that for visible light-driven aerobic oxidation, as depicted in Scheme 8.3, plasmonic photocatalyst might oxidize a substrate (electron donor) on the Au surface, while e is consumed by the reduction of O2 (electron acceptor) on the semiconductor surface. Besides, it is found that the catalyst architecture is critical for the activity, i.e., small Au particles (dAu < 5 nm) loaded on P25 are necessary and Au particles located at the anatase/rutile interface behave as the active sites, facilitating efficient e transfer to TiO2 and successful aerobic oxidation under sunlight irradiation, as shown in Scheme 8.4a. Apart from TiO2, Kominami’s group has prepared Au/CeO2 by photochemical deposition of H4AuCl4 on CeO2 in the presence of citric acid as the reducing agent, which presents the maximum absorption around 550 nm, in agreement with the SPR of Au NPs [70]. When exposed to 530 nm LED irradiation, benzyl alcohols can be stoichiometrically oxidized to corresponding benzaldehydes with O2 in water, as displayed in Scheme 8.4b.

Scheme 8.3
scheme 3

Proposed mechanism for visible light-driven aerobic oxidation by Au particles supported on semiconductor particles (Reprinted with permission [69]. Copyright 2012 American Chemical Society)

Scheme 8.4
scheme 4

(a) Aerobic oxidation of alcohols with Au/TiO2 in toluene under the irradiation of natural sunlight [69]. (b) Photooxidation of alcohols with various substitution groups with Au/CeO2 [70]

Our group has found that the introduction of graphene (GR) into semiconductors leads to enhanced performance for aerobic oxidation of various benzylic alcohols with different substituents and allylic alcohols [7174]. As displayed in Table 8.1, GR-TiO2 composite has been fabricated via a two-step wet chemistry approach, using graphene oxide (GO) and TiF4 as the precursors of GR and TiO2, respectively [72]. It is proposed that the superior and easily accessible “structure-directing” role of GO, the intimate interfacial contact between GR and TiO2, and the better separation of the photogenerated carriers of GR-TiO2 play a synergistic role in leading to the enhanced photocatalytic performance. Similarly, the introduction of GR into CdS substrate can also influence the morphology and structure, enhance the visible light absorption intensity, and improve the lifetime and transfer of photogenerated electron–hole pairs over the CdS–GR, leading to enhanced photocatalytic performance toward photocatalytic selective oxidation of benzylic and allylic alcohols, as shown in Fig. 8.5a [75]. In addition, it has been found the intimate interfacial interaction between CdS and GR can inhibit the photocorrosion of CdS during the photocatalytic reactions effectively.

Table 8.1 Selective oxidation of a range of alcohols over the TiO2-5 % GR photocatalyst under the visible light irradiation (λ > 400 nm) for 20 h (Reprinted with permission [72]. Copyright 2011 American Chemical Society)
Fig. 8.5
figure 5

(a) Photocatalytic selective oxidation of benzyl alcohol to benzaldehyde under the visible light irradiation over the as-prepared samples: (a) blank CdS; (b) CdS-1 % GR; (c) CdS-5 % GR; (d) CdS-10 % GR; and (e) CdS-30 % GR nanocomposites (Note: C and Y is short for conversion and yield. Reprinted with permission [75]. Copyright 2011 American Chemical Society). (b) CdS–GR–TiO2 nanocomposites (Note: CG is short for CdS-5 % GR; GR–M–CdS nanocomposites. Reprinted with permission [74]. Copyright 2012 American Chemical Society). (c) Photocatalytic performance of blank CdS, CdS–GR, and CdS–(GR–M) (M = Ca2+, Cr3+, Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) nanocomposites with different weight addition ratios of GR for photocatalytic selective oxidation of benzyl alcohol under visible light (λ > 420 nm) for 2 h (Reprinted with permission [73]. Copyright 2014 American Chemical Society)

In addition to the dual-ingredient hybrid systems, we have also constructed ternary nanocomposites with further improved photocatalytic performances toward catalytic selective oxidation of alcohols, as shown in Fig. 8.5b [74]. Based on our previous CdS–GR composite with lamellar structure, TiO2 nanoparticles can be uniformly carpeted on the surface of the CdS–GR by an in situ growth strategy [74]. The enhanced photocatalytic performance of ternary CdS–GR–TiO2 hybrids can be ascribed to the combined interaction of larger surface area, the intimate interfacial contact among them, and two possible routes for electron transfer in the system [74]. The photogenerated electrons can transfer to the graphene nanosheets and TiO2 simultaneously, which contributes to promoting the interfacial charge transfer rate and lengthening the lifetime of photogenerated electron–hole pairs, consequently resulting in the enhanced photoactivity.

Recently, our group has reported the transfer efficiency of photogenerated charge carriers across the interface between graphene (GR) and semiconductor CdS can be further improved by introducing a small amount of metal ions (M = Ca2+, Cr3+, Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) as “mediator” into their interfacial layer matrix [73]. The photoactivity of GR–M–CdS for aerobic oxidation of alcohol is significantly improved (Fig. 8.5c) because the metal ions introduced can optimize the atomic charge carrier transfer pathway across the interface between GR and the semiconductor as well as drive a balance between the positive effect of GR on retarding the recombination of electron–hole pairs photogenerated from semiconductor and the negative “shielding effect” of GR resulting from the high weight addition of GR. Inspired by this, Pd has been selected as a typical noble metal to investigate whether it can play the similar role to metal ions as the interfacial mediator between GR and CdS. It is found that the ternary CdS–(GR–Pd) nanocomposite shows significantly enhanced visible light photocatalytic activity as compared to both blank CdS and the optimum binary CdS–GR, which verifies Pd can also play as the interfacial mediator to optimize and improve the spatial charge carrier separation and transfer across the interfacial domain between GR and CdS upon visible light irradiation [76].

Since optimizing the structure of the photocatalysts also has significant impact on the photocatalysts’ performance, our group has synthesized the Pt/CeO2 nanocomposite in an aqueous phase with tunable core–shell and yolk–shell structure via a facile and green template-free hydrothermal approach toward selective oxidation of benzyl alcohol [77]. The yield of benzaldehyde obtained over core–shell Pt/CeO2 is ca. 9, 27, and 39 times higher than that of nanosized CeO2, blank CeO2, and supported Pt/CeO2, respectively (Fig. 8.6), which can be ascribed to the advantageous core–shell structure, where Pt core can trap/store the photogenerated electrons to prolong the lifetime of charge carriers, hence improving the efficiency toward photocatalytic redox process. As for the yolk–shell structure, the interfacial contact between the Pt core and CeO2 shell is quite loose, and thus the role of Pt core has decreased markedly, resulting in lower photocatalytic activity.

Fig. 8.6
figure 6

The yield (a) and conversion (b) of photocatalytic selective oxidation of benzyl alcohol to benzaldehyde over the core–shell Pt/CeO2, yolk–shell Pt/CeO2, supported Pt/CeO2, nanosized CeO2 powder, and blank CeO2 under the irradiation of visible light (λ > 420 nm) at room temperature and ambient atmosphere (Reprinted with permission [77]. Copyright 2011 Royal Society of Chemistry)

In addition, Pd@CeO2 semiconductor nanocomposite with “plum-pudding” structure has been fabricated via a facile low-temperature hydrothermal reaction of polyvinylpyrrolidone (PVP)-capped Pd colloidal particles and cerium chloride precursor followed by a calcination process in air, as shown in Fig. 8.7a [78]. This unique nanostructure endows the Pd@CeO2 nanocomposite with enhanced activity and selectivity toward the visible light-driven oxidation of various benzylic alcohols to corresponding aldehydes using dioxygen as oxidant at room temperature and ambient pressure compared with a supported Pd/CeO2 nanocomposite and nanosized CeO2 powder (Fig. 8.7b, c), which is ascribed to the unique structure assembly of multi-Pd core@CeO2 shell nanocomposite. First, the uniform structure composition, i.e., the evenly dispersed Pd cores are spatially encapsulated by the CeO2 shell, provides a homogeneous environment for photocatalytic reaction. Second, the multi-Pd core@CeO2 shell nanostructure can significantly increase the surface area compared with the supported counterpart. Third, the “three-dimensional” intimate contact between the evenly dispersed Pd core and CeO2 shell maximizes the metal–support interaction, which facilitates the interfacial charge transfer process.

Fig. 8.7
figure 7

Overall flowchart for fabrication of the multi-Pd core@CeO2 shell semiconductor nanocomposite (a); time-online photocatalytic selective oxidation of benzyl alcohol to benzaldehyde over the multi-Pd core@ CeO2 shell nanocomposite, supported Pd/CeO2, and commercial CeO2 nanosized powder under the irradiation of visible light (λ > 420 nm) under ambient conditions, selectivity (b); conversion (c) (Reprinted with permission [78]. Copyright 2011 American Chemical Society)

Notably, aiming to give validity to targeted reactions for a fast screening of catalysts for photocatalytic transformations, Colmenares and co-workers have synthesized 24 different titania-based systems (either alone or modified with metals) through the sol–gel process varying the precursor (titanium isopropoxide or tetrachloride) and the aging (magnetic stirring, ultrasounds, microwave, or reflux) conditions. They are tested for liquid-phase selective photooxidation of 2-butenol (crotyl alcohol) to 2-butenal (crotonaldehyde) and gas-phase selective photooxidation of 2-propanol to acetone [61]. It is found that both test reactions (despite having very different reactant/catalyst ratio and contact times) show quite similar results in terms of influence of the precursor and the metals. That is, taking titanium isopropoxide as the precursor of titanium leads to better results than titanium tetrachloride and the presence of iron, palladium, or zinc is detrimental to activity, whereas zirconium and especially gold can improve the results as compared to pure titania. Especially, for 2-propanol transformation into acetone, platinum-containing catalyst shows quite high selectivity values to acetone (in the 78–80 % range at 22–28 % conversion), suggesting that this test reaction is more sensitive to platinum, as shown in Fig. 8.8.

Fig. 8.8
figure 8

Dependence of selectivity to acetone on 2-propanol conversion for all the titania-based systems used in the present study. Highlighted values correspond to platinum-containing titania (Reprinted with permission [61]. Copyright 2012 Elsevier)

8.4.3 Oxidation of Saturated Primary C–H Bonds

The selective oxidation of stable alkyl aromatics such as toluene to commercial chemicals, including benzaldehyde, benzyl alcohol, benzoic acid, and benzyl benzoate with molecular oxygen, has important applications in fine chemicals and pharmaceutical production [79]. However, selective oxidation of C–H bonds with environmentally benign oxygen is often very difficult to control due to the abundance and inertness of C–H bonds in organic substrates. On the other hand, in thermal heterogeneous catalysis, the activation of saturated sp3 C–H bonds often requires transition metal nanoparticles as a catalyst and relatively harsh reaction conditions. The rapid progress in selective transformation by heterogeneous photocatalysis presents the possibility of the selective activation of saturated sp3 C–H bonds using molecular oxygen as a benign oxidant and visible light under ambient conditions [80, 81].

Zhu and co-workers have devised a new class of photocatalysts, metal hydroxide nanoparticles grafted with alcohols, which can efficiently oxidize alkyl aromatic compounds with O2 using visible or ultraviolet light or even sunlight to generate the corresponding aldehydes, alcohols, and acids at ambient temperatures and give very little overoxidation [81]. For example, toluene can be oxidized with a 23 % conversion after a 48 h exposure to sunlight with 85 % of the product being benzaldehyde and only a trace of CO2. A tentative free radical mechanism is proposed, as shown in Scheme 8.5. The surface complexes catalyze the selective oxidation by an efficient mechanism that employs light absorption by the complexes to yield highly reactive surface radicals (−O–C–H–R), and these surface radicals initiate the aerobic oxidation of the organic molecules in contact with them. Moreover, the oxidation ability of the photocatalysts can be tuned through selection of the metal element and the alcohols grafted.

Scheme 8.5
scheme 5

Proposed pathways of the reaction mechanism for the selective oxidization of alkyl aromatics (Reprinted with permission [81]. Copyright 2012 Royal Society of Chemistry)

Our group has reported a cubic phase CdS semiconductor with specific sheet structure morphology synthesized by a simple room temperature method, which is able to be used as a visible light-driven photocatalyst for the selective oxidation of saturated primary C–H bonds in alkyl aromatics with high activity and selectivity using molecular oxygen as a benign oxidant and benzotrifluoride as the solvent under ambient conditions [82]. The superior photocatalytic performance of CdS is attributed to its unique structure assembly of specific sheet morphology with cubic phase, high surface area, and efficient separation of photogenerated charge carriers upon visible light irradiation. This work strongly suggests that tuning the microscopic structure and composition in diverse respects, e.g., size, shape, phase, crystallinity, or morphology, continues to be a very versatile, feasible, and general strategy to adjust the photocatalytic performance of a specific semiconductor photocatalyst with tunable photoactivity.

Considering the fact that there is still a need to improve the photocatalytic performance of CdS for selective oxidation of C–H bonds, we later have designed and synthesized ternary GR–CdS–TiO2 composites with an intimate spatial integration and sheetlike structure, which is afforded by assembling two co-catalysts, graphene and TiO2, into the semiconductor CdS matrix with specific morphology as a visible light harvester [83]. As shown in Fig. 8.9, under visible light irradiation of 10 h, 53 % conversion of toluene is achieved over the optimal 5 % GR–CdS composite, which is much higher than 33 % conversion over the blank-CdS photocatalyst with the selectivity still maintained as high as 99 % (Fig. 8.9a). With the addition of second co-catalyst TiO2, the photoactivity toward aerobic oxidation of C–H in toluene is further enhanced as compared to the binary 5% GR–CdS composite (Fig. 8.9b). It is proposed that the co-catalysts are able to cooperate with the light harvester to facilitate the charge separation/transfer and lengthen the lifetime of photogenerated electron–hole pairs, thereby resulting in an enhanced overall photocatalytic performance. This work demonstrates a wide, promising scope of adopting co-catalyst strategy to design more efficient semiconductor-based photocatalyst toward selective activation of C–H bonds using solar light and molecular oxygen.

Fig. 8.9
figure 9

Photocatalytic selective oxidation of toluene to benzaldehyde under visible light irradiation of 10 h over the as-prepared GR–CdS composites and blank CdS (a) and over GR–CdS–TiO2 composites (b) (Reprinted with permission [83]. Copyright 2013 Nature Publishing Group)

8.4.4 Epoxidation of Alkenes

The epoxidation of alkene is another type of reaction of industrial importance. However, since alkenes lack the suitable coordination sites present in heteroatom substrates, the weak adsorbed alkene substrates are unable to interact with TiO2 directly to induce visible light absorbance. Thus, the deployment of H2O2 as the terminal oxidant is needed to form active surface complex through the interplay between H2O2 and TiO2. As early as 2001, it is reported that 1-decene can be converted to 1,2-epoxydecane on TiO2 powder using molecular oxygen as the oxygen source under UV light irradiation [84]. For rutile TiO2 powders, the rate of epoxide generation can be significantly increased by addition of hydrogen peroxide (Scheme 8.6). In this case, the reaction can occur under visible light as well as UV light. The selectivity of the production of 1, 2-epoxydecane is higher under visible light than under UV light, due to the formation of Ti-η2-peroxide species on the surface of rutile TiO2 upon treatment with H2O2, which can induce visible light absorbance [84].

Scheme 8.6
scheme 6

Epoxidation of 1-decen to 1,2-epoxydecane with TiO2 in CH3CN/C3H7CN [84]

With regard to the plasmonic photocatalysts, Linic and co-workers have found that Ag nanocubes of ~60 nm edge length supported on α-Al2O3 can be used for selective oxidation of ethylene to ethylene oxide (Scheme 8.7a) [85]. At 450 K, with the introduction of visible light into the system, the steady-state oxidation rate shows a fourfold increase relative to that driven by thermal energy only. And it is verified that the dissociation of molecular O2 on silver to form adsorbed atomic oxygen controls the reaction rates. Later, they reported that Cu nanoparticles (NPs) with an average size of 41 nm supported on the inert SiO2 (Cu/SiO2) can tune the selectivity for the light-driven epoxidation of propylene to propylene oxide (Scheme 8.7b) [86]. The selectivity enhancement by Xe lamp irradiation is attributed to the localized SPR of Cu NPs which weakens the Cu–O bond, thereby prompting the reduction of Cu2O to Cu0.

Scheme 8.7
scheme 7

Epoxidation of ethylene to ethylene oxide (a) [85] and propylene to propylene oxide (b) [86]

Our group has reported an organic dye-like macromolecular “photosensitizer” role of graphene (GR) in wide-bandgap ZnS semiconductors for aerobic epoxidation of alkenes, including styrene, cyclohexene, and cyclooctene by visible light [87]. In this study, the assembly of nanosized ZnS particles on the two-dimensional platform of GR with an intimate interfacial contact has been prepared by a facile two-step wet chemistry process. Although the introduction of GR into the substrate of ZnS is not able to narrow the bandgap of ZnS to visible light region, the as-prepared ZnS–GR exhibits visible light photoactivity toward visible light irradiation (λ > 420 nm). Different from previous studies on GR–semiconductor photocatalysts, where GR is claimed to behave as an electron reservoir to capture and shuttle the electrons photogenerated from the semiconductor, it is proposed that the GR in the ZnS–GR nanocomposites is serving as an organic dye-like macromolecular “photosensitizer.” Under visible light irradiation, GR is excited from the ground state to the excited state GR* and then injects electrons into the conduction band of ZnS, ultimately leading to visible light photoactivity.

8.4.5 Sulfoxidation of Thioethers

The sulfoxidation of sulfides is important organic transformations in many fields, including pharmaceuticals, fossil fuel desulfurization, industrial wastewater treatment, and chemical warfare agent disposal [47, 88]. However, the over-sulfoxidation of sulfides into sulfone is the main side reaction, and efforts should be devoted to achieve selective sulfoxidation [89]. Photocatalytic sulfoxidation using molecular oxygen as the terminal oxidant turns out to be more advantageous toward achieving sustainable chemistry. Although the mechanism of sulfoxidation by molecular oxygen via photochemical irradiation has been investigated in physical chemistry [9094], the applications are just beginning to be reported recently [48, 9598].

Yao’s group has reported a metal-free aerobic selective sulfoxidation photosensitized by Rose Bengal (an organic dye) or solid-supported Rose Bengal, utilizing visible light as the driving force and molecular oxygen as the oxidant (Scheme 8.8) [97]. With the assistance of a catalytic amount of hydrochloric acid, Rose Bengal appears to be the most efficient and selective catalyst toward the selective sulfoxidation of thioanisole, compared to the metal complex photoredox catalysts. A series of other thioethers have been subjected to the optimized reaction conditions, verifying the generality of this reaction. In addition, Rose Bengal is readily accessible and inexpensive, and with the use of a solid-support catalyst, the workup procedure can be significantly improved. Moreover, a 50 mmol scale reaction has been carried out in an Erlenmeyer flask under sunlight on the roof of the chemistry building and demonstrated that the reaction is practical and scalable.

Scheme 8.8
scheme 8

Selective sulfoxidation of thioanisole under sunlight with Rose Bengal (Reprinted with permission [97]. Copyright 2013 Royal Society of Chemistry)

Apart from various organic photocatalysts, our group has reported a spatially branched hierarchical system composed of inorganic photocatalysts, CdS/ZnO nanocomposites (CZ), and its photoactivity toward sulfoxidation of thioanisole has been tested (Fig. 8.10) [99]. The conversion for thioanisole and yield for methylsulfinylbenzene are about 60 % and 52 % over the branched hierarchical CZ, respectively, which are much higher than the values obtained over CdS NWs (conversion 36 % and yield 33 %) and blank ZnO (conversion 3.5 % and yield 3 %). The enhanced photoactivities can be ascribed to the branched hierarchical structure of CZ nanocomposites, which can not only increase the light harvesting efficiency but also boost charge separation and faster charge transport and collection. Additionally, under light irradiation, the branched hierarchical CZ nanocomposites are thought to generate a “Z-scheme” system [7] to drive the photocatalytic processes.

Fig. 8.10
figure 10

(a) Schematic illustration for synthesis of branched hierarchical CZ nanocomposites and (b) photocatalytic performance of CdS NWs, blank ZnO, and branched hierarchical CZ nanocomposites for selective oxidation of thioanisole (Reprinted with permission [99]. Copyright 2014 Royal Society of Chemistry)

Recently, Zhao and co-workers have achieved the selective oxidation of thioanisole with TiO2 photocatalyst under visible light irradiation by combining the aerobic oxidation of sulfide and the aerobic oxidative formylation of amine with methanol through the synergistic interplay of reactants and catalyst (Scheme 8.9) [95]. In contrast, attempts to perform these two reactions individually are not successful. The mechanism for the selective synergistic aerobic oxidation of thioanisole 1 and benzylamine 2 has been proposed and demonstrated in Scheme 8.10. The adsorption of benzylamine on TiO2 leads to the formation of surface complex a, which shows activity under visible light irradiation, facilitating electron transfer from the adsorbed benzylamine to TiO2, and thus a series of reactions are initiated. During the whole stage, the addition of redox player, CH3OH, is required to associate the two reactions. It is expected that more synergistic redox reactions can be achieved by judiciously selecting a pair of substrates and an appropriate solvent.

Scheme 8.9
scheme 9

The synergistic aerobic oxidation of two substrates, sulfide 1 and amine 2, occurs simultaneously on the surface of TiO2 under visible light irradiation (Reprinted with permission [95]. Copyright 2015 Royal Society of Chemistry)

Scheme 8.10
scheme 10

Proposed mechanism for the synergistic selective oxidation of thioanisole 1 and benzylamine 2 with O2 on TiO2 under visible light irradiation (Reprinted with permission [95]. Copyright 2015 Royal Society of Chemistry)

8.5 Photoreduction of Nitroaromatics

Photocatalytic reduction has also been studied as extensively as oxidation reactions since it is generally safer than the conventional methods, which always employ dangerous reducing agents such as hydrogen and carbon monoxide [28, 38]. Among the photocatalytic reductions, nitroaromatics reduction has been studied most extensively since amino compounds have been widely applied in the industrial synthesis of dyes, biologically active compounds, pharmaceuticals, rubber chemicals, and photographic and agricultural chemicals [73, 100102]. In the transformation process, the removal of O2 from the reaction mixture by purging N2 and adding hole scavenger acting as electron donors can effectively improve the reaction rates and selectivities.

As early as 1997, Brezova et al. have reported that 4-nitrophenol can be successfully reduced to 4-aminophenol in various alcohols (methanol, ethanol, 1-propanol, 2-propanol, 1-butanol, and 2-butanol) with TiO2 (Degussa P25) under UV irradiation [103]. The best yield of 92 % has been obtained after 20 min irradiation (λ = 365 nm) of 4-nitrophenol in methanol suspensions. It is proposed that the photoreduction rate is significantly affected by the solvent parameters, such as viscosity, polarity, and polarizability. For example, increase in the polarity parameter leads to better stabilization of the charged intermediate and accelerates the photocatalytic reduction.

Xu et al. have prepared GR-modified TiO2 hybrids by electrostatic assembly, and its photocatalytic reduction of nitroaromatics to the corresponding amino-aromatics is examined under UV light irradiation [104]. When the dosage of GO is about 1.0 %, almost all of the 4-nitrophenol is converted, and the yield rate of 4-aminophenol achieves around 95 % with H2C2O4 as hole scavenger under N2 purge. The results indicate that the addition of GR can effectively minimize the recombination of photogenerated charge carriers derived from the irradiated TiO2 and better encourage these separated electrons to participate in the reactions. In addition, it is found that different hole scavengers lead to discrepant reduction efficiency. Therefore, to adjust and optimize such photocatalytic reduction reactions is of great importance to improve its reduction ability.

Our group has fabricated various GR-based nanocomposites such as CdS nanowire/GR [100], CdS nanosphere/GR [105], In2S3/GR [102], ZnIn2S4/GR [106], TiO2/CdS nanowire/GR [107], GR–M–CdS (M = Ca2+, Cr3+, Mn2+, Fe2+, Co2+, Ni2+, Cu2+, and Zn2+) [73], GR–Pd–CdS [76], and hierarchical CdS–ZnO–GR hybrids [108]. It is found that their photoactivities for reduction of nitro compounds to amino compounds with ammonium formate (HCOONH4) for hole scavenger in N2 atmosphere (Scheme 8.11) under visible light irradiation are all remarkably enhanced as compared to the blank semiconductors [73, 100102, 106, 107]. It has been concluded that the selective reduction efficiency can be driven by appropriate introduction of GR into the matrix of pure semiconductor, which can boost the transfer and prolong the lifetime of the electrons photoexcited from the semiconductor due to the tighter connection between GR and the semiconductor, as well as the optimization of the atomic charge carrier transfer pathway across the interface between GR and the semiconductor.

Scheme 8.11
scheme 11

Photocatalytic reduction of nitro compounds to amino compounds in water with the addition of ammonium formate (HCOONH4) for quenching photogenerated holes under N2 atmosphere

In addition, our group has designed a Pd/CeO2 hollow core–shell nanocomposite (Pd@hCeO2) composed of tiny Pd nanoparticle (NP) cores encapsulated within CeO2 hollow shells, as shown in Fig. 8.11a [109]. As compared to supported Pd/CeO2 and commercial CeO2, the as-prepared Pd@hCeO2 demonstrates improved photoactivity toward selective reduction of aromatic nitro compounds under visible light irradiation with the addition of ammonium oxalate as quencher for photogenerated holes and N2 purge at room temperature (Table 8.2). It is recognized that the CeO2 shell acts as the primary active component to give photogenerated electron–hole pairs, while the Pd nanoparticle cores with low-lying Fermi level serve as an electron reservoir to prolong the lifetime of the charge carriers. The three-dimensional interfacial contact between Pd cores and hollow CeO2 shells facilitates the efficient charge carrier transfer, thereby leading to the enhanced fate of photogenerated electron–hole pairs from CeO2. In particular, the core–shell strategy efficiently prevents the aggregation of Pd NPs in the high-temperature calcination process and the leaching of Pd NPs for the catalytic reaction in a liquid phase, which is not able to be achieved for traditional supported Pd/CeO2 catalyst. As shown in Fig. 8.11b, the core–shell Pd@hCeO2 almost does not have the loss of photoactivity during the recycled activity testing on reused samples, whereas the significant loss of photoactivity is clearly observed for supported Pd/CeO2, due to the significant leaching of Pd nanoparticles in supported Pd/CeO2.

Fig. 8.11
figure 11

(a) Typical SEM images of Pd@hCeO2 core–shell nanocomposite. (b) Stability testing of photocatalytic activity of supported Pd/CeO2 and Pd@hCeO2 core–shell nanocomposites for reduction of 4-nitrophenol under visible light irradiation (λ > 420 nm) for 6 h (Reprinted with permission [109]. Copyright 2013 American Chemical Society)

Table 8.2 Photocatalytic reduction of substituted aromatic nitro compounds over Pd@hCeO2 core–shell nanocomposite, supported Pd/CeO2, and commercial CeO2 aqueous suspension under visible light irradiation (λ > 420 nm) with the addition of ammonium oxalate as quencher for photogenerated holes and N2 purge at room temperature (Reprinted with permission [109]. Copyright 2013 American Chemical Society)

8.6 Coupling Reactions

The photoinduced charge separation occurring on the surface of photocatalysts creates both a reduction center and an oxidation center. This unique feature allows multistep reactions on a single photocatalyst: intermediates generated from one reaction center could be the substrates at another center. The integrated use of both reaction centers could therefore complete a sophisticated multistep synthesis in “one-pot” reaction, i.e., the coupling reactions.

8.6.1 C–N Coupling

Various kinds of organic compounds can be synthesized through C–N coupling reactions [110112]. For example, Ohtani and co-workers have found that photoirradiation to an aqueous solution containing primary amines in the presence of a powdered mixture of TiO2 with Pt black (Pt/TiO2 catalyst) gives rise to the corresponding secondary amines via C–N coupling reaction (Scheme 8.12a) [113]. l-Pipecolinic acid has been achieved by the cyclization reaction through photoirradiation of a deaerated aqueous suspension of TiO2 containing l-lysine (Scheme 8.12b). This reaction involves a single-electron transfer from the terminal amino group of l-pipecolinic acid to h+ on the TiO2 surface and subsequent removal of the amino group, leading to intramolecular C–N coupling to form an imine, which is then reduced to form the l-pipecolinic acid. The highest selectivity (77 %) and conversion of l-lysine (90 %) have been achieved [110].

Scheme 8.12
scheme 12

(a) Photocatalytic transformation of primary amines to secondary amines via C–N coupling reaction on Pt/TiO2 [113]. (b) Photocatalytic transformation of l-lysine to l-pipecolinic acid via C–N coupling reaction on TiO2 [110]

The C–N coupling reaction also can proceed between amines and alcohols. For example, photoirradiation to an alcohol solution containing primary or secondary amines with Pt/TiO2 can produce the corresponding secondary or tertiary amines [114, 115]. In these reactions, alcohols are oxidized by h+ formed on the catalyst surface to the corresponding aldehydes or ketones, which then could couple with amine to form an imine. In addition, imines can also be produced by reacting alkynes with amines to hydro-amination products with Au nanoparticles supported on nitrogen-doped TiO2 (Au/TiO2–N), which serves as the visible light plasmonic photocatalyst [116]. The presence of Ti3+ of TiO2, arising from nitrogen doping, provides more coordination sites for the alkyne, thereby prompting a better performance than that of pure TiO2 (Scheme 8.13a). Through a similar reaction, propargyl-amines have been produced by a one-pot synthesis procedure with the introduction of aldehyde to the amine and alkyne mixture, on the catalyst Au/ZnO under 530 nm LED irradiation (Scheme 8.13b) [117]. The high yields of propargyl-amines result from the interaction of alkyne with the ZnO support. Therefore, the use of plasmonic photocatalysts can also be employed to manipulate the oxidative condensation of amine with aldehyde under visible light irradiation.

Scheme 8.13
scheme 13

Photocatalytic C–N coupling reactions between alkynes with amines (Reprinted with permission [33]. Copyright 2014 Royal Society of Chemistry)

8.6.2 C–C Coupling

Heterocyclic bases can easily be functionalized via C–C coupling by the reaction with amides or ethers. For example, Caronna and co-workers have found that with the addition of H2SO4 and H2O2, heterocyclic bases can react with amide to produce the corresponding amide-functionalized heterocycles (Scheme 8.14) [118]. It is considered that during these reactions, amide first reacts with •OH radical formed on the TiO2 surface to form amide radical via hydrogen abstraction, which then attacks the heterocyclic bases and produces the C–C coupling products. The additional H2SO4 and H2O2 can accelerate the amide radical formation.

Scheme 8.14
scheme 14

Photocatalytic reactions of heterocyclic bases and formamide in aqueous solution with TiO2 (Reprinted with permission [118]. Copyright 2003 Royal Society of Chemistry)

Subsequently, it is reported by the same group that the reactions between various heterocyclic bases and ethers can be induced by sunlight, and various kinds of heterocycle-ether conjugates are formed in liquid–solid heterogeneous system in the presence of TiO2 (Scheme 8.15) [119]. The derivatives obtained with trioxane may give an easy entry to heterocyclic aldehydes. It is considered that in these reactions, the α-oxyalkyl radical is generated by hydrogen abstraction from the α-carbon of the ethers or by electron transfer from the oxygen atom of the ethers, which then is claimed to attack the heterocyclic bases to form the heterocycle-ether conjugates.

Scheme 8.15
scheme 15

Photocatalytic functionalization of heterocyclic base with ethers by TiO2 (Reprinted with permission [26]. Copyright 2008 Elsevier)

In addition, other nucleophiles have also been incorporated into the activated substrates adjacent to N-atom, enabling even more complicated functionalization under visible light irradiation on P25 TiO2. Using P25 as a photocatalyst, new C–C bonds were successfully constructed under the mild visible light irradiation of an 11 W fluorescent lamp [120]. As shown in Scheme 8.16, both CH3NO2 and KCN are able to act as the nucleophile to facilitate the formation of different C–C bonds to afford different tetrahydroisoquinoline derivatives in good to excellent yields [120].

Scheme 8.16
scheme 16

C–C coupling of a variety of N-aryltetrahydroisoquinolines with CH3NO2 and KCN over TiO2 (Reprinted with permission [33]. Copyright 2014 Royal Society of Chemistry)

Besides, Möhlmann and co-workers have operated mpg-C3N4 as a visible light photocatalyst to functionalize the benzylic C–H adjacent to N-atoms with a suitable nucleophile, using 0.1 MPa of O2 as the oxidant, without the assistance of an additional additive. Under the visible light irradiation, a variety of N-aryltetrahydroisoquinolines could be connected with nitroalkanes and dimethyl malonate to construct new C–C bonds, as shown in Scheme 8.17 [121].

Scheme 8.17
scheme 17

C–C coupling of a variety of N-aryltetrahydroisoquinolines with nitroalkanes and dimethyl malonate over mpg-C3N4 (Reprinted with permission [33]. Copyright 2014 Royal Society of Chemistry)

8.7 Conclusion

Selective organic transformation plays an important role in the synthesis of various commodity chemicals closely linked to our daily lives. In comparison to conventional methods, photocatalytic organic transformations possess several inherent advantages, including benign environmental impacts and employment of mild reaction conditions. It is offering an alternative “green” route for the production of organics. However, despite significant progress made in this area of research as reviewed above, typical challenges still exist and need to be resolved before it is more viable for large-scale practical applications. In this regard, more exciting discoveries of efficient photocatalysts and more precise deployment of the reaction conditions are required to be conceived in the pursuit of higher conversion and higher selectivity for photocatalytic organic transformations in a practical scale-up manner. More investigations are needed to deepen our understanding of the photoinduced interfacial charge carrier transfer processes and optimized photocatalytic reactor design, which can help effectively promote the development and implementation of rapid global spread of organic synthesis by semiconductor-based photocatalysis. Hopefully, because many organic reactions are thermodynamic downhill exothermic reactions, thus we believe that the prospect of photocatalytic organic transformations in practical application is more available.