Asymmetrical heterobiaryls are extremely important motifs in semiconductor, liquid crystal materials, and drug development [1, 2]. Consequently, a number of methods have been reported for the synthesis of the asymmetrical heterobiaryls [3, 4]. Particularly, in recent decades, different kinds of efficient ways to synthesize the asymmetrical heterobiaryls have been developed. The traditional ways focus on direct cross-coupling using transition-metal, such as Suzuki [5], Stille [6], Negishi [7], etc. However, these methods have many disadvantages, the use of plenty of metal organic precursors and high temperature. After those reactions, there will be a large number of metal wastes generate, which were hard to deal with and harmful to the environment. As a consequence, exploring a convenient, novel, green and sustainable approach for cross-coupling synthesis is a highly desirable and challenging task.

Recently, scientists have repeatedly found that organic transformation induced by visible-light has shown its innately continuous and green chemistry [8-14]. What is more, visible light photoredox catalyst also has received much attention in the organic synthesis [15-23]. It can be used to induce visible-lighttriggered reactions as most organic compounds do not absorb visible light. Pd NPs is a kind of catalyst which has been widely applied in the organic reactions [24-27] due to their advantages of high stability, reusability and low toxicity. However, there are only rare studies about using it as catalyst under visible light irradiation in the organic synthesis until now. Pd NPs supported catalyst (Pd/CeO₂, Pd/ZrO₂) is a new kind of photocatalyst, it has been used in our laboratory for years [28]. As we all know, under the visiblelight irradiation, the excited electron on the metal Pd(0) surface can inject into the chemical antibonding orbital directly, then induce the reaction started because of chemically adsorbed molecules [29, 30]. To our knowledge, Pd/CeO₂ induces the cross-coupling for the synthesis of asymmetrical heterobiaryls has not been reported yet. Fortunately, we disclose a mild and highly efficient, visible-light promoted protocol for the synthesis of heterobiaryls from aryl halide and bromo aromatic heterocyclic compound derivatives without using external reductants. A series of asymmetrical heterobiaryls can be conveniently generated in an efficient and scalable fashion.

The Pd/CeO₂ was prepared via a simple impregnation-reduction process using PdCl₂, lysine and NaBH₄ as starting materials (detail see Supplementary information).

The prepared 3 wt% Pd/CeO₂ was characterized by TEM, Pd nanoparticle size distribution, UV–vis, XRD and XPS. First, as can be seen from the TEM image (Fig. 1A), the Pd nanoparticles disperse equally on the surface of the CeO₂. Fig. 1B shows that the average particle diameter of the metal palladium nanoparticles is < 7 nm and dispersed on the surface of the CeO₂ evenly. Moreover, UV–vis spectra (Fig. 2A) shows that the light absorption of Pd/CeO₂ is stronger than CeO₂ both in the UV and visible range, suggesting that the enhanced light absorption arises from the dispersion of Pd nanoparticles on the CeO₂ surface. It is obvious that the diffraction peaks of the Pd/CeO₂ could correspond to the pure CeO₂ entirely from Fig. 2B, indicating that the structure of CeO₂ remained unchanged after the metal NPs were loaded, which may be account of the metal content is inferior to the detection limit and/or the crystallinity of the surface metal NPs is poor. As seen from Fig. 2C, we find that the XPS spectra of Pd 3d exhibits two signal peaks at 336.01 and 341.41 eV, which are attributed to the Pd⁰ species. From XPS analysis, it can be deduced that the Pd⁰ species are predominant in the composites. XPS analysis also shows that the Pd content on the surface of CeO₂ was about 2.75%, which is almost identical with the theoretical content of Pd (3%).

Initial investigation was conducted by employing reacts 1 with 2 in the presence of 3 wt% Pd/CeO₂ as a photoredox catalyst, K₃PO₄ as the base and DMA as the solvent under the irradiation of an incandescent light (0.79 W/cm²) (with a glass filters to cut off the irradiation below a certain value of wavelength) at 60 ℃ for 24h, resulting in 81% yield of the target product 3 (Table 1, entry 6). As shown in Table 1, the application of 2 wt% and 4 wt% Pd provided 72% and 60% yield respectively (Table 1, entries 7, 8), and 5 wt% Pd gave only a small amount (Table 1, entry 9). Pd NPs is nanoparticles of nonplasmonic transition metals, we then turn to choose 3 wt% Au/ZrO₂, which has been well demonstrated to have localized surface plasmon resonance (LSPR) [13, 14], but it only give a 40% yield of the target product (Table 1, entry 1). Then we screen kinds of other photocatalysts (Table 1, entries 2–5), when 3 wt% Pd/TiO₂ and Au/TiO₂ were used as the photoredox catalyst, we achieved quite low yields, 13%, 12% respectively (Table 1, entries 2 and 4). The use of Pd/ZrO₂ or Au/CeO₂ as the photoredox catalyst has no improvement in the transformation (Table 1, entries 3 and 5). The use of CeO₂ as a catalyst led to a 7% yield of product, this result shows that the CeO₂ support itself has a very limited contribution to the catalytic activity of the photocatalyst (Table 1, entry 10). Then, different solvents were also evaluated for this conversion (Table 1, entries 11–15), and those prove DMA was facilitated the reaction effectively. In order to improve the yield of the target product, different bases were tested, including K₂CO₃, KOAc, KOH, Na₂CO₃ and NaOAc. The results promoted us to use K₃PO₄ to conduct this reaction (Table 1, entries 16–19). It is important to note that visible light and catalyst are essential in the reaction (Table 1, entry 20).

In the optimal conditions, we turned to study the scope and generality of this method and all the products were characterized by ¹H NMR. As shown in Table 2, we can see that the most yields in the dark of target products were < 10%. All of those show that the visible-light acts a conclusive role. From Table 2 we can see that varies substituent group, like electron-donating groups and electron-withdrawing groups on the substrates 1 and 2 worked well in the reaction to give the corresponding products with moderate to good yields. This implied that the electronic effect had little influence on the reaction. Regardless of differences in the electronic and steric properties of substrates 1, they afford the desired product in good yields range from 60% to 82% with the reactant 2 unchanged (Table 2, 3a–3h). We also choose three kinds of different bromo-aromatic heterocyclic (pyridine, furan, thiophene) to explore the practicality of our reaction conditions with the reactant 1 unchanged. To our delight we receive quite satisfactory consequence, as shown in Table 2 (3i–3u). Compared to furan and thiophene, pyridine has a lower yield. But we can still obtain the yield above 55%. From those, we can conclude that the reaction conditions bear different aromatic heterocyclic and their substrates. What is more, from Table 2, we can see the reaction bears the chemical reactivity of aryl iodine better than bromine.

As shown in Fig. S1a (Supporting information) we can know that at the early stage of the reaction process, the yield of 3 has a big increase. And, the yield increased slowly from the twentieth hour. When the reaction time ranged from 24 h to 36 h, the same target product was obtained and slight increase. So we choose 24 h as the best choice reaction time finally.

The advantages of heterogeneous catalysts exist in their easy separation, good stability and fantastic recyclability. As expected, the recycled photocatalyst showed good reusability and only slight decrease in its activity after five cycles, which was reveled on Fig. S1b (Supporting information).

According to the previous literatures [28, 31, 32], under visible light irradiation, Pd nanoparticles can absorb the energy of incident light to make the ground state bound electrons on the surface of the metal reach its excitation, which could occur energy level transition to transferred to a higher energy level orbit become the exciting electrons, named photo-induced electrons. It can enhance the catalytic performance of photocatalyst obviously. What is more, Pd (0) d¹⁰ complex has a long-life time excited state (three-line excited state) in solution state [33-35], and the increase of charge density on the metal palladium surface also makes the metal interact more closely with the reactants in the reaction system. Moreover, as to our reaction system, the key step is the process of dehalogenation, and our photocatalyst due to the photoredox electron shuttling, it could avoid using external reductants, offering a promising bond-forming strategy [36]. Based on the results above, we accounted out a photoredox cycle and the C—C cross-coupling events product was strongly favored due to the Fischer–Ingold persistent radical effect (PRE) [37, 38]. The PRE is a general principle that explains the highly specific formation of the cross-coupling product (R¹–R²) between two radicals R¹ and R², when one species is persistent (long lived) and the other is transient. During the initial step, the concentrations of both radicals are very low (usually < 10^-7 mol/L), which enables the homo-coupling of transient radical and the cross-coupling of two different radicals is unfavorable. After this period, the increasing of two different radicals concentration steers the reaction subsequently to follow a single pathway, that is, the cross reaction [39].

A plausible mechanism of the photocatalytic cross-coupling is illustrated in Fig. 3. Under the visible light irradiation, the concentrated energetic electrons can transfer to the halogen atom from the Pd atoms to facilitate the cleavage of C—X bonds to form an Ar—Pd—X complex A and Het-Pd-Br complex B. Next, the intermediates Ar-Pd-Het C generates in situ from the complexs A and B on the surface of Pd NPs. The final product D can be achieved by reductive elimination, and the Pd active sites can be returned to the initial states.

In summary, we have developed a novel and environmental benign process for the synthesis of asymmetrical heterobiaryls through cross-coupling by using the visible-light-mediated catalyst Pd/CeO₂. In contrast to previous work, the tolerances of substrates are good. Moreover, further studies on the application of recycle photochemical based on earth abundant metals for other organic transformations in progress in our laboratory

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.cclet.2018.01.002.