Selective Alkane Desaturation Catalyzed by Molecular Cage Copper Complexes Under Mild Conditions
At a Glance
Section titled āAt a Glanceā| Metadata | Details |
|---|---|
| Publication Date | 2024-01-27 |
| Journal | CCS Chemistry |
| Authors | Hao Sun, Xueyuan Wang, Xudong Wang, YuāFei Ao, DeāXian Wang |
| Institutions | University of Chinese Academy of Sciences, Beijing National Laboratory for Molecular Sciences |
| Citations | 4 |
Abstract
Section titled āAbstractāOpen AccessCCS ChemistryCOMMUNICATIONS23 Feb 2024Selective Alkane Desaturation Catalyzed by Molecular Cage Copper Complexes Under Mild Conditions Hao Sun, Xue-Yuan Wang, Xu-Dong Wang, Yu-Fei Ao, De-Xian Wang and Qi-Qiang Wang Hao Sun Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 University of Chinese Academy of Sciences, Beijing 100049 , Xue-Yuan Wang Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Xu-Dong Wang Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 , Yu-Fei Ao Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 University of Chinese Academy of Sciences, Beijing 100049 , De-Xian Wang Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 University of Chinese Academy of Sciences, Beijing 100049 and Qi-Qiang Wang *Corresponding author: E-mail Address: [email protected] Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Molecular Recognition and Function, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 University of Chinese Academy of Sciences, Beijing 100049 https://doi.org/10.31635/ccschem.024.202403936 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Catalytic oxidative alkane desaturation by base metals with benign oxidants under mild conditions is highly attractive, as it could enable an efficient way to produce valuable olefin products. Here we report the design and synthesis of macrocyclic and molecular cage copper complexes in catalyzing highly selective cyclohexane desaturation with H2O2/O2. The macrocycles and cages possess two to three phenanthroline units connected by versatile binary linkers or tripodal caps. These compounds readily formed bi- and tri-nuclear copper complexes with confined cavities and adaptive coordination environments, as suggested by X-ray single crystal analysis. In the catalytic oxidation of cyclohexane, a considerable yield of cyclohexene with unprecedented dehydrogenation/oxygenation selectivity up to 6.1:1 was achieved with the optimal cage compound C3. Control and mechanistic experiments suggested a pronounced cage effect and radical-mediated processes. The high selectivity was attributed to a cooperative effect between the copper centers, which facilitated the second hydrogen atom abstraction on the trapped cyclohexyl radical and the inhibition of the Fenton process within the shielding cavity environment. Download figure Download PowerPoint Introduction Catalytic dehydrogenation of alkanes to akenes has drawn great attention as it can transform the abundant, cheap alkane raw materials into olefin products, which are important industrial feedstock and versatile functionalization platforms.1-10 Many heterogeneous catalysts1-3 and homogeneous noble metal catalysts4-8 are known to be effective in alkane dehydrogenation, which usually requires either high temperatures or the use of precious metals and a sacrificial hydrogen acceptor (another olefin in most cases). On the other hand, oxidative alkane dehydrogenation catalyzed by homogeneous base metals (e.g. Cu, Fe, Mn) under mild conditionsāas inspired by an enzymeāis highly attractive, yet limited successes have been received. This is especially true when considering the use of benign oxidants such as oxygen and hydrogen peroxide.10-21 In this aspect, PĆ©rez et al.11 recently reported a seminal work in which the dehydrogenation of n-hexane and cycloalkanes could be accomplished with trispyrazolylborate-copper catalysts and hydrogen peroxide. A trace of alkene product (0.88% yield for cyclohexene) was observed with a selectivity of only 4% (compared to the major products, viz alcohol and ketone). The low selectivity for alkene products in these systems is usually due to the competing, dominant hydroxylation pathway on the produced alkyl radical active intermediates, as well as nonselective processes associated with the occurrence of highly-active free hydroxyl radical (Fenton-like pathways).22-26 The development of supramolecular chemistry has enabled abundant tools and approaches to mediate chemical reactivity and boost efficient catalysis, leading to the emergence of supramolecular catalysis.27-33 Especially, macrocycle and cage compounds possessing functional cavity and binding sites have been used to mimic the characteristics of enzymes that promote selective catalysis.34-47 In nature, desaturases desaturate long alkyl chain substrates with specific dehydrogenation selectivity.48-50 These enzymes utilize surrounding protein cavity environments to exert precise control on the formation and stabilization of high-valent metal oxo active site and high-active alkyl radical intermediates exclusively distinguishing the pathway of dehydrogenation from oxygenation that operates for mechanistic-related oxygenases.51-55 Inspired by these outcomes, we envisioned taking advantage of macrocycle and cage cavities to embed suitable base metal sites for catalytic alkane dehydrogenation. By choosing suitable ligand and linker units, metal centers could be installed precisely with a confined cavity microenvironment. In doing so, the produced metal oxo active species and alkyl radical intermediates could be stabilized and shielded from the bulk solution, thereby adopting a preferred dehydrogenation pathway (Figure 1). Following this idea, phenanthroline56-58 was chosen as ligand units, as it provides well-defined chelating coordination and semi-open site for many metals, especially for copper, and has shown good performance in the catalysis of many C-H transformations.59-63 Figure 1 | Molecular cage copper complexes for catalyzing selective alkane desaturation. Download figure Download PowerPoint Results and Discussion In previous work, we synthesized a series of bis-phenanthroline macrocycles with different linker lengths and studied their Cu(II) complexation properties64 (Figure 2, top). These macrocycles readily form binuclear complexes with Cu(II) with adjustable Cu-Cu distances. To construct a more enclosed cavity, cage compounds C1- C3 containing three phenanthroline units were synthesized (Figure 2, bottom). Three different tripodal caps, triphenylbenzene (flatten), triphenyltriazine (proximately planar), and triphenylethane (conical), were employed to build cages with distinct cavity shapes and inner environments. The one-pot SNAr reactions between 3 equiv 2,9-dichloro-1,10-phenanthroline and 2 equiv corresponding triphenols heated in dimethylformamide (DMF) in the presence of Cs2CO3 afforded the desirable cage products in 14-32% yields. All the cage compounds were fully characterized, as revealed by nuclear magnetic resonance (NMR), mass spectrometry (MS), and elemental analysis data (see Supporting Information). Figure 2 | Structures of macrocycles and molecular cages containing phenanthroline units. Download figure Download PowerPoint To our delight, crystal structures for the free cage C2, the trinuclear Cu(I) complex of cage C3, and the binuclear Cu(I) complex of macrocycle M5 were obtained (Figure 3; for details, see Supporting Information Figures S1-S3 and Tables S1-S3). The free cage C2 adopts a twisted, propeller-like conformation with approximate D3 symmetry (Figure 3c). The two triazine rings stack onto each other with the phenyl units staggeringly arrayed. The three phenanthroline units are skewed with all nitrogen coordination sites pointing toward the inner of the cage. By contrast, in the trinuclear Cu(I) complex of C3, the cage adopted an expanded conformation (Figure 3d). The three phenanthroline units became more upright and an enclosed cavity was formed by the strutting of the two conical triphenylethane caps. Each phenanthroline coordinated with one copper, with Cu-Cu distances of 8.4, 8.7, and 9.8 Ć . The uneven distances were consistent with some distorted conformation of the cage. The remaining semi-open sites of the copper centers were exposed toward the cavity and coordinated by acetonitrile molecules. Cu1 was tetrahedrally coordinated, and Cu2, Cu3 formed nearly planar trigonal coordination. The distinct coordination geometries could reflect the cavity constraint upon filling the many coordinated acetonitrile molecules. The binuclear Cu(I) complex of M5 adopted an expanded conformation with a long Cu-Cu distance of 11.1 Ć (Figure 3b). In this circumstance, both copper centers were allowed to form tetrahedral coordination with two acetonitriles. One dichloromethane molecule was nicely filled within the center of the macrocyclic cavity. Such expanded conformation was distinct from the twisted, compressed conformation observed for the free macrocycle (Figure 3a), suggesting the conformational adaption ability of these kinds of macrocycles and cages. Figure 3 | Crystal structures of (a) M5,64 (b) Cu2(M5)(CH3CN)42Ā·CH2Cl2, (c) C2, and (d) Cu3(C3)(CH3CN)43Ā·CHCl3Ā·(CH3CN)2. Counteranions, nonessential solvent molecules, and hydrogen atoms are omitted for clarity. Download figure Download PowerPoint With the defined coordination sites and cavity environments for multinuclear copper complex formation, these macrocycles and cages were applied to catalyze the oxidation of cyclohexane, one of the most common nonactivated alkane substrates using the benign oxidant hydrogen peroxide (Table 1). The Cu(I) salt, Cu(CH3CN)4PF6, was initially used to establish the reaction conditions (Table 1, entries 1-8). After some initial screening, the reactions were performed in acetonitrile, also called methyl cyanide/dichloromethane (MeCN/DCM; 1:2) at room temperature with Cu/L (6 mol % Cu salt, 3 mol % M1- M5, and 2 mol % C1- C3), H2O2 (27.1% in H2O, 1 equiv) and excess of cyclohexane. To our delight, in addition to the major oxygenation products cyclohexanol and cyclohexanone, certain amounts of dehydrogenation product cyclohexene were detected. The total yields for these products were around 10% (based on the limiting reagent H2O2; the reactions were stopped when a maximum yield of cyclohexene was reached). Interestingly, from macrocycles to cages, the yield and selectivity for cyclohexene generally increased, with the best yield (3.8%) and selectivity (1:2.0) observed for cage C3 (Table 1, entry 8). With these exciting results, the reactions were further optimized: Among the different copper salts tested, copper(II) triflate [Cu(OTf)2] was the best (Table 1, entries 9-12). The reaction temperature had a pronounced effect (Table 1, entries 13-14). Better yield and selectivity for cyclohexene were observed, but were opposite for cyclohexanol and cyclohexanone production when the reaction was performed at 0 Ā°C. Finally, when the reaction was carried out under an O2 atmosphere, the yield for cyclohexene was further increased to 9.2%, and an unprecedented selectivity of 6.1:1 was observed (Table 1, entry 15). Table 1 | Catalytic Oxidation of Cyclohexane by Cu Complexes with Macrocycles and Molecular Cagesa Entry Cu L Time Yield (%)b Selectivity 2/( 3+ 4) 2 3 4 1 Cu(CH3CN)4PF6 M1 4 d 1.8 8.9 2.3 1:6.2 2 Cu(CH3CN)4PF6 M2 2 d 0.9 6.2 3.4 1:10.7 3 Cu(CH3CN)4PF6 M3 1 d 1.7 6.1 2.2 1:4.9 4 Cu(CH3CN)4PF6 M4 4 d 2.3 5.2 2.8 1:3.5 5 Cu(CH3CN)4PF6 M5 1 d 2.3 6.8 1.3 1:3.5 6 Cu(CH3CN)4PF6 C1 1 d 2.3 4.5 1.7 1:2.7 7 Cu(CH3CN)4PF6 C2 2 d 2.1 5.0 0.5 1:2.6 8 Cu(CH3CN)4PF6 C3 2 d 3.8 5.2 2.3 1:2.0 9 Cu(CH3CN)4BF4 C3 2 d 3.4 3.4 2.2 1:1.6 10 CuOTf C3 1 d 3.4 4.0 2.2 1:1.8 11 Cu(BF4)2Ā·6H2O C3 1 d 4.3 3.3 0.2 1.2:1 12 Cu(OTf)2 C3 1 d 5.4 4.8 0.2 1.1:1 13c Cu(OTf)2 C3 1 d 2.6 10.1 0.8 1:4.2 14d Cu(OTf)2 C3 4 d 5.6 Trace Trace - 15 d, e Cu(OTf)2 C3 4 d 9.2 0.7 0.8 6.1:1 aReaction conditions: cyclohexane (7.5 mmol), Cu salts (7.8 Ī¼mol), M1- M5 (3.9 Ī¼mol) or C1- C3 (2.6 Ī¼mol), 27.1% aqueous H2O2 (131 Ī¼mol), acetonitrile/dichloromethane (MeCN/DCM; 1:2 v/v, 3 mL), room temperature. After the reaction, the crude reaction mixture was treated with PPh3 (1 mmol) for analysis. bDetermined by gas chromatography (GC) using an internal standard. All yields were calculated based on H2O2. An average value of three repeated reaction runs was reported. cThe reaction was taken at 40 Ā°C. dThe reaction was taken at 0 Ā°C. eThe reaction was taken under an O2 atmosphere (with a balloon). A series of comparison and control experiments were further carried out to explore the cage effect (Table 2). As expected, the use of phenanthroline fragment (1,10-phenanthroline, āPhenā) or cage C3 in the absence of copper salt gave no conversions (Table 2, entries 1-2). On the other hand, the use of copper salt Cu(OTf)2 alone led to sluggish conversion with only a trace of detectable cyclohexene (Table 2, entry 3). Among the different ligands used, including Phen and the two representative macrocycles M1, M5, the cages performed much better (Table 2, entries 4-6 vs entries 7-9), suggesting a significant cage effect. Among these, C3 was obviously superior to C1 and C2, indicating that a more enclosed cavity and appropriate Cu-Cu distances were essential. By changing the Cu/ C3 ratio from 3:1 to 1:1 (Table 2, entries 9-11), the yield of cyclohexene dropped, and more importantly, the selectivity varied, with the best selectivity (8.3:1) observable at Cu/ C3 = 2:1. This finding indicated that a cooperative effect between the different copper sites could be involved and had a significant influence on the selectivity. The best cooperation observed for two Cu sites could reflect a suitable steric demand within the crowed cage cavity. To further explore the cage effect, a non-cage tripodal analogue ( TA; for the structure of this compound, see Supporting Information) was synthesized and subjected to the catalytic reaction (Table 2, entry 12). The apparent decreased yield and reduced selectivity suggested once again that the simultaneous presence of Cu-Cu cooperation and an enclosed cage environment was most beneficial. Finally, in the presence of only O2 but without H2O2, the reaction still proceeded with a small conversion (Table 2, entry 13). This indicated that O2 activation could independently operate in this system, but the coexistence of the two oxidants was more beneficial. Replacing the O2 atmosphere with N2 also caused a diminished conversion (Table 2, entry 14). Table 2 | Comparison and Control Reactionsa Entry Cu L Time Yield (%) Selectivity 2/( 3+ 4) 2 3 4 1 Phen 4 d ā ā ā ā 2 C3 4 d ā ā ā ā 3 Cu(OTf)2 4 d 0.6 2.3 Trace 1:3.8 4 Cu(OTf)2 Phen 4 d 2.8 2.5 Trace 1.1:1 5 Cu(OTf)2 M1 4 d 2.9 3.6 1.5 1:1.8 6 Cu(OTf)2 M5 4 d 4.3 2.8 1.7 1:1 7 Cu(OTf)2 C1 2 d 6.0 1.0 1.3 2.6:1 8 Cu(OTf)2 C2 1 d 3.7 1.1 1.2 1.6:1 9 Cu(OTf)2 C3 4 d 9.2 0.7 0.8 6.1:1 10b Cu(OTf)2 C3 2 d 2.5 0.3 Trace 8.3:1 11c Cu(OTf)2 C3 1 d 1.3 0.4 Trace 3.3:1 12 Cu(OTf)2 TA 2 d 5.6 1.0 Trace 5.6:1 13d Cu(OTf)2 C3 4 d 1.4 Trace Trace ā 14e Cu(OTf)2 C3 4 d 4.9 2.3 Trace 2.1:1 aReaction conditions: cyclohexane (7.5 mmol), Cu(OTf)2 (7.8 Ī¼mol), C1- C3 or TA (2.6 Ī¼mol), M1 or M5 (3.9 Ī¼mol), Phen (7.8 Ī¼mol), 27.1% aqueous H2O2 (131 Ī¼mol), acetonitrile/dichloromethane (MeCN/DCM; 1:2 v/v, 3 mL), 0 Ā°C, O2 atmosphere (with a balloon). Standard workup and gas chromatography (GC) analysis procedures were applied. bCu(OTf)2 (5.2 Ī¼mol) was used. cCu(OTf)2 (2.6 Ī¼mol) was used. dThe reaction was taken in the absence of H2O2. eThe reaction was taken under the N2 atmosphere. To shed more light on the reaction mechanism, especially the possible radical processes, a set of experiments was performed (Scheme 1). First, oxone, which contained KHSO5 as the oxidant and did not produce hydroxyl radical during oxidation,65 was applied instead of H2O2 and O2 (Scheme 1b). The reaction gave cyclohexene as the sole product without the detection of cyclohexanol and cyclohexanone. The similarity in yields of cyclohexene indicated that in both reactions the cyclohexene formation might have been mediated through a similar, non-hydroxyl radical mechanism; however, relative to cyclohexanol and cyclohexanone formation, hydroxyl radical might have been involved. Second, when cyclohexanol or cyclohexanone was subjected to the reaction, no production of cyclohexene was detected (Scheme 1c). This suggested that cyclohexene was not converted through the oxygenated products; thus, the dehydrogenation and oxygenation should have occurred through two different pathways. Finally, radical scavengers were applied to probe the possible radical species. Since the commonly used radical inhibitors such as 2,6-ditertbutyl-4-methylphenol (BHT) and 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO) could be oxidized by hydrogen peroxide, CCl4 and BrCCl3 were applied.11 The results showed that the production of cyclohexene was reduced or completely inhibited by the two scavengers, and corresponding halo-cyclohexane products were detected (Scheme 1d,e). These experiments suggested that cyclohexyl radical was the key intermediate, from which cyclohexene and cyclohexanol were further converted. Scheme 1 | Mechanistic study experiments depicting alkane desaturation catalyzed by molecular cage copper complexes. Standard reaction conditions and work-up procedures were applied. For details see Supporting Information. Download figure Download PowerPoint Based on the above experimental results and relevant literature reports,11 a catalytic mechanism involving the high-valent copper oxo active species and sequential radical transformations was proposed (Figure 4). The initial activation of H2O2 and O2 by the trinuclear copper complex of cage C3 could lead to the formation of the copper-oxo active sites. This was further supported by the in situ electrospray ionization mass spectrometry (ESI-MS) analysis of the reaction mixture, from which a series of cage-copper complex species, especially the [cage-Cu=O]+ species could be assigned (see Supporting Information Figure S11). It should be noted that the formation of dicopper-oxygen species could be precluded due to the large Cu-Cu distances and relatively rigid ligand environments.66 The high activity of the copper-oxo species was suggested based on the faster reaction kinetics observed for the cage than for Phen (see Supporting Information Figure S10). The following hydrogen atom abstraction on cyclohexane by copper-oxo site led to the key cyclohexyl radical intermediate, which could be stabilized and confined within the enclosed molecular cage cavity. From this step, the cyclohexene formation could be accomplished through the fast, second hydrogen atom abstraction on the vicinal CāH bond by the formed copper-hydroxo species or another Cu=O site.48-50 The existence of the second neighboring copper site could facilitate this hydrogen atom abstraction process and afford the olefin product selectively. This could be supported by the observed cooperative effect. On the other hand, due to the existence of a shielding and low-polar cavity environment, the competing Fenton process could be inhibited to a limited level inside the cage.67,68 Figure 4 | A plausible mechanism for the catalytic oxidation of cyclohexane to cyclohexene by the Cu complex with molecular cage C3. Download figure Download PowerPoint Conclusion Molecular cages containing three phenanthroline units and different tripodal caps were synthesized efficiently. These cages and their macrocyclic analogues possessed adaptive conformational structures and well-defined metal coordination sites. Multi-semi-open copper centers could be readily installed at adjustable distances and geometries. As such an enzyme-mimic cavity was built around the embedded copper sites, especially for cage C3 where the conical multiphenyl caps constituted a relatively enclosed and shielding pocket. These unique copper complexes were applied in the catalysis of challenging alkane desaturation reactions. In the oxidation of cyclohexane with H2O2 and O2 under mild conditions, a considerable yield for the production of cyclohexene with unprecedented dehydrogenation/oxygenation selectivity up to 6.1:1 was achieved with cage C3. Control and mechanistic experiments suggested a pronounced cage effect and copper oxo initialized radical processes. The existence of the enclosed cavity environment stabilized the in-situ produced cyclohexyl radical intermediate and promoted its following selective dehydrogenation over hydroxylation via the cooperativity of different copper sites. Meanwhile, the Fenton-like process could be maximally inhibited in the shielding cavity. These results demonstrate the great potential of tailoring a defined coordination environment and catalytic pocket on metal active centers for conquering important and challenging transformations. Supporting Information Supporting Information is available and includes synthesis and characterization of compounds, crystallographic data, catalysis studies, and copies of NMR spectra of all compounds. The Cambridge Crystallographic Data Center (CCDC) deposition number (CCDC 2296531, CCDC 2296532, CCDC 2296533) contains the supplementary crystallographic data for this paper, viz Cu2( M5)(CH3CN)42Ā·CH2Cl2, C2, and Cu3( C3)(CH3CN)43Ā·CHCl3Ā·(CH3CN)2 (CIF). Conflict of Interest There is no conflict of interest to report. Funding Information This research was made possible as a result of a generous grant from the National Key R&D Program of China (grant no. 2021YFA1501600) and the National Natural Science Foundation of China (grant nos. 22022112 and 21521002). References 1. Li C.; Wang G.Dehydrogenation of Light Alkanes to Mono-Olefins.Chem. Soc. Rev.2021, 50, 4359-4381. Google Scholar 2. Dai Y.; Gao X.; Wang Q.; Wan X.; Zhou C.; Yang Y.Recent Progress in Heterogeneous Metal and Metal Oxide Catalysts for Direct of and Soc. Rev.2021, 50, Google Scholar Alkane Catalysts via a Soc. Rev.2021, 50, Google Scholar as a in Google Scholar and Catalyzed by Google Scholar C.; by and Google Scholar Y.; Alkane Google Scholar of Alkanes and by Metal Google Scholar Catalytic for Google Scholar C.; Li of Alkanes Metal Google Scholar PĆ©rez Copper as for Alkane Google Scholar of Alkanes to via Google Scholar for Direct of from and Google Scholar of with and via Following Google Scholar C.; Y.; and Desaturation by a Google Scholar by of the of a Google Scholar Wang that and 1, Google Scholar X.; for an to the in by Google Scholar of with The of Catalytic by Google Scholar C.; Catalysts for CāH An the that the between and Desaturation Google Scholar Wang Y.; Wang of Metal and in from a 50, Google Scholar C.; Alkane with H2O2 Catalyzed by an Google Scholar Copper Complexes as Catalysts for the Oxidation of Alkanes under Mild Google Scholar Y.; and of Alkanes by a 2, Google Scholar Oxidation of Alkanes with H2O2 Under a Fenton-like Google Scholar Y.; by a and H2O2 to Google Scholar Molecular and within Google Scholar 2, Google Scholar as a for and Soc. Google Scholar Soc. Google Scholar in Google Scholar Q.; the Google Scholar and Google Scholar Catalyzed by and Google Scholar From to and Soc. Google Scholar Y.; of with Google Scholar of From to for Molecular and Google Scholar Y.; of in Google Scholar of Complexes to Control in Catalytic Google Scholar as for Google Scholar Wang In of Google Scholar Zhou Ao Wang Wang of Catalysts Google Scholar Zhou Ao Wang Wang for and Google Scholar on in Google Scholar Ao Wang Wang for and with Molecular Google Scholar Sun Y.; Wang Wang C.; C.; Y.; Cage with Light Google Scholar Zhou Sun X.; Sun of Metal in for and Google Scholar the Google Scholar of the Google Scholar the an Google Scholar by Google Scholar of Oxidation Catalyzed by Google Scholar in Google Scholar by and Google Scholar and Google Scholar A Soc. Google Scholar for and Metal Soc. Google Scholar A for the of for Google Scholar Oxidation of to and An Google Scholar Li and of Google Scholar C.; of to Google Scholar Metal in a with Metal Google Scholar C.; 2, Google Scholar Wang Ao Wang Wang and Cu(II) of Google Scholar Y.; of and of a by Google Scholar Complexes and Google Scholar for of Google Scholar Y.; of in the of and China Google Scholar Information Chinese desaturation