First Organic–Inorganic Hybrid Compounds Formed by Ge-V-O Clusters and Transition Metal Complexes of Aromatic Organic Ligands

Three compounds based on Ge-V-O clusters were hydrothermally synthesized and characterized by IR, UV-Vis, XRD, ESR, elemental analysis and X-ray crystal structural analysis. Both [Cd(phen)(en)]2[Cd2(phen)2V12O40Ge8(OH)8(H2O)]∙12.5H2O (1) and [Cd(DETA)]2[Cd(DETA)2]0.5[Cd2(phen)2V12O41Ge8(OH)7(0.5H2O)]∙7.5H2O (2) (1,10-phen = 1,10-phenanthroline, en = ethylenediamine, DETA = diethylenetriamine) are the first Ge-V-O cluster compounds containing aromatic organic ligands. Compound 1 is the first dimer of Ge-V-O clusters, which is linked by a double bridge of two [Cd(phen)(en)]2+. Compound 2 exhibits an unprecedented 1-D chain structure formed by Ge-V-O clusters and [Cd2(DETA)2]4+ transition metal complexes (TMCs). [Cd(en)3]{[Cd(η2-en)2]3[Cd(η2-en)(η2-μ2-en)(η2-en)Cd][Ge6V15O48(H2O)]}∙5.5H2O (3) is a novel 3-D structure which is constructed from [Ge6V15O48(H2O)]12− and four different types of TMCs. We also synthesized [Zn2(enMe)3][Zn(enMe)]2[Zn(enMe)2(H2O)]2[Ge6V15O48(H2O)]∙3H2O (4) and [Cd(en)2]2{H8[Cd(en)]2Ge8V12O48(H2O)}∙6H2O (5) (enMe = 1,2-propanediamine), which have been reported previously. In addition, the catalytic properties of these five compounds for styrene epoxidation have been assessed.


X-ray Crystallography
The crystal data for compound 1 were measured on a Bruker Apex II diffractometer with graphite monochromated Mo Kα (λ = 0.71073 Å) radiation. The data for compounds 2 were measured on a Rigaku R-AXIS RAPID diffractometer with graphite monochromated Mo Kα (λ = 0.71073 Å) radiation, while the data for compound 3 were measured on an Agilent Technology SuperNova Eos Dual system with a Mo Kα (λ = 0.71073 Å) microfocus source and focusing multilayer mirror optics. None of the crystals showed evidence of crystal decay during the data collections. Refinements were carried out with SHELXS-2014/7 [73] and SHELXL-2014/7 [73] using Olex 2.0 interface via the full matrix leastsquares on F2 method. In the final refinements, all atoms were refined anisotropically in compounds 1-3. The hydrogen atoms of en, phen, DETA and enMe in the three compounds were placed in calculated positions and included in the structure factor calculations but not refined. In these heavy-atom structures with reflection data from poor-quality crystals it was not possible to see clear electron-density peaks in difference maps which would correspond with acceptable locations for the various H atoms bonded to water oxygen atoms. The refinements were then completed with no allowance for these water H atoms in the models; the CCDC number: 1,525,920 for 1, 2,024,572 for 2 and 1,525,922 for 3. The reflection intensity data for compounds 4 and 5 were also measured on a Rigaku R-AXIS RAPID diffractometer with graphite monochromated Mo Kα (λ = 0.71073 Å) radiation, and the results show that the two compounds have already been reported previously [6]. A summary of the crystallographic data and structure refinements for compounds 1-3 is given in Table 1.

Synthesis Description
Compounds 1 and 2 are all based on Cd 2 Ge 8 V 12 and compounds 3 and 4 are based on Ge 6 V 15 . The alkalinity (pH > 9) and the stirring time of the reaction mixture are important for the formation of Ge 6 V 15 in compounds 3 and 4. We have a relatively clear grasp of the synthetic conditions of the two different clusters. The molar ratio of V 2 O 5 to GeO 2 for compounds 1 and 2 is about 1:2, and the molar ratios of NH 4 VO 3 to GeO 2 for compounds 3 and 4 are 2:1. The addition of 2, 2 -bpy is important for the preparations of compounds 2 and 3. Though it is absent in the products, 2, 2 -bpy is required for the syntheses of compounds 2 and 3. It should be noted that such a phenomenon is not unusual in hydrothermal preparations [74].  8 V 12 ), two [Cd(phen)(en)] 2+ and 12.5 water molecules. As shown in Figure 1, an unusual feature of 1 is that two [Cd(phen)] 2+ take the place of the two VO 2+ fragments located at the two opposite positions of {Ge 8 V 14 O 50 } [59], forming Cd 2 Ge 8 V 12 . The two substituted cadmiums each is coordinated by four oxygens from two {Ge 2 O 7 } units with Cd-O distances of 2.290(5)-2.366(5) Å, and two nitrogens from a phen ligand with Cd-N distances of 2.353(6)-2.427(7) Å. That is to say, two phen ligands were decorated onto the surface of Cd 2 Ge 8 V 12 . The two phen located at the two sides of Cd 2 Ge 8 V 12 are not parallel to each other. There is a dihedral angle of 36.116 • between the two phenanthroline-planes. All the bond distances in 1 are comparable to those of previously reported compounds [6,[55][56][57][58][59][60][61][62][63][64][65][66]. Bond valence sum (BVS) calculations for Ge and V indicate that both Ge and V exist in the +4 oxidation-state (Table S1). BVS calculations were also conducted for the cadmium and oxygen atoms in compound 1 to determine the locations of the hydrogen atoms in compound 1 (see supporting information and discussions in "BVS calculations to determine the locations of hydrogen atoms for compounds 1-3") [75]. The asymmetric unit of 1 consists of a di-Cd-substituted Ge-V-O [H8Cd2(phen)2Ge8V12O48(H2O)] 4− (Cd2Ge8V12), two [Cd(phen)(en)] 2+ and 12.5 wate cules. As shown in Figure 1, an unusual feature of 1 is that two [Cd(phen)] 2+ take t of the two VO 2+ fragments located at the two opposite positions of {Ge8V14O50} [59 ing Cd2Ge8V12. The two substituted cadmiums each is coordinated by four oxyge two {Ge2O7} units with Cd-O distances of 2.290(5)-2.366(5) Å, and two nitrogen phen ligand with Cd-N distances of 2.353(6)-2.427(7) Å. That is to say, two phen were decorated onto the surface of Cd2Ge8V12. The two phen located at the two Cd2Ge8V12 are not parallel to each other. There is a dihedral angle of 36.116° betw two phenanthroline-planes. All the bond distances in 1 are comparable to those o ously reported compounds [6,[55][56][57][58][59][60][61][62][63][64][65][66]. Bond valence sum (BVS) calculations for G indicate that both Ge and V exist in the +4 oxidation-state (Table S1). BVS calc were also conducted for the cadmium and oxygen atoms in compound 1 to determ locations of the hydrogen atoms in compound 1 (see supporting information and sions in "BVS calculations to determine the locations of hydrogen atoms for com 1-3") [75]. Except for [Cd(phen)] 2+ , there are two [Cd(phen)(en)] 2+ . It should be noted that the two [Cd(phen)(en)] 2+ are different from each other. Cadmium of [Cd(4)(phen)(en)] 2+ of the two is bonded to four nitrogens from a phen and an en with Cd-N distances of 2.250(8)-2.320(7) Å, a terminal oxygen from Cd 2 Ge 8 V 12 with the Cd-O distance of 2.385(5) Å and a water molecule with the Cd-O distance of 2.583(9) Å, exhibiting a cis-octahedral geometry. Therefore, the cluster acts as a monodentate inorganic ligand coordinating with Cd(4), forming a cluster supported transition metal complex (TMC). Cadmium of [Cd(3)(phen)(en)] 2+ of the two receives contributions from four nitrogens from a phen and an en with Cd-N distances of 2.290(7)-2.367(7) Å and two terminal oxygens from two  V 12 to construct a novel cluster dimer. It should be noted that there are two Cd(3) TMCs acting as a double bridge linking two Cd 2 Ge 8 V 12 . The dimer further supports two Cd(4) TMCs at the two sides of the dimer. That is to say, Cd(3) TMCs act as bridges joining Cd 2 Ge 8 V 12 , but Cd(4) TMCs terminate the connection of the clusters by the terminating water molecule.

Description of Crystal Structures
Distances between the central water molecule of the Cd 2 Ge 8 V 12 and Cd (3) and Cd(4) are 7.886-7.888 Å, and the angle of Cd (3) The dimer of clusters was reported by our group in 2002 [76] and very recently [77]; the first one was based on the Mo 8 V 6 cluster, and the second one was based on the V 15 O 36 cluster. However, compound 1 here is the most complex one of the three, which is the first example of dimer of substituted clusters. The other two reported compounds are both based on traditional clusters but not the substituted one.
The building block [H 7 Cd 2 (phen) 2 Ge 8 V 12 O 48 (0.5H 2 O)] 5− (Cd 2 Ge 8 V 12 ) of 2 is almost identical to that of 1, which is also a cadmium di-substituted Ge-V-O cluster; each substituted cadmium is also coordinated by a phen ligand. The main difference between the building blocks of compounds 2 and 1 is the number of the attached hydrogen atoms. There are only slight differences between the bond lengths and angles in compounds 2 and 1. Bond valence sum calculations for Ge and V also indicate that Ge and V are in the +4 oxidation-state (Table S1).  [54]. However, Yang's cluster is based on aliphatic organic ligands but not aromatic organic ones. Secondly, Yang's coordination fragment is formed by en ligands rather than DETA ligands. Finally, the 1-D chain of Yang's compound is sinusoidal, but the one here is linear.
substituted Ge-V-O cluster of aromatic organic ligands. Yang et. al. also reported a 1-D chain structure formed by similar substituted Ge-V-O clusters and coordination frag ments [54]. However, Yang's cluster is based on aliphatic organic ligands but not aromati organic ones. Secondly, Yang's coordination fragment is formed by en ligands rather than DETA ligands. Finally, the 1-D chain of Yang's compound is sinusoidal, but the one here is linear.  (Table S1).
In conclusion, there are five types of TMCs in compound 3. To the best of our knowledge, compound 3 contains the largest number of TMC types.
The TMCs and the Ge 6 [6]. However, there are several significant differences between our compound and Yang's compound. Firstly, and most importantly, the Ge-V-O cluster of Yang's compound is Ge 4 V 16 , but the corresponding cluster of our compound is Ge 6 V 15 . Secondly, Yang's compound is based on diethylenetriamine ligands but not en in our compound. Finally, Yang's compound did not exhibit various channels that were found in our compound.

BVS Calculations to Determine the Locations of Hydrogen Atoms of Compounds 1-3
Single crystal X-ray diffraction cannot exactly determine the positions of the hydrogen atoms from the Fourier maps. For further verifying the correctness of the formula of the three compounds, BVS calculations [75] were carried out to determine the positions of the hydrogen atoms for all the three compounds. As for compound 1, the oxygens can be classified into eight groups: (1) seven Ge-O t terminal oxygens; (2) one Ge-O t -Cd µ 2 -oxygen; (3) eleven V-O t terminal oxygens; (4) one V-O t -Cd µ 2 -oxygen; (5) eight µ 3 -oxygens located between two vanadiums and one germanium; (6) eight µ 3 -oxygens located between three vanadiums; (7) eight µ 3 -oxygens between a vanadium, cadmium and germanium; and (8) four µ 2 -oxygens between two germaniums. All the atoms of the eight groups except groups (1) and (2) can be assigned to the −2 valence state, with BVS calculation results in the range of 1. 56-2.16. With respect to the group (1) oxygens, all seven oxygens exist in the -1 valence state, with BVS results ranging from 1.01-1.04, indicating that all seven terminal Ge-O t oxygens are mono-protonated. The BVS value of the group (2) oxygen is 1.38, meaning that although this oxygen is coordinated by both one cadmium and one germanium, it exists in the −1 valence state. Therefore, the cluster in compound 1 is attached by eight hydrogens, and all eight hydrogens are attached on the eight Ge-O terminal oxygens.
As for compound 2, the oxygens can also be divided into eight groups. Seven of the eight groups are similar to the corresponding groups in compound 1. Only the eighth one is not found in compound 1: it is a µ 3 -oxygen between two cadmiums and a germanium. This µ 3 -oxygen is a terminal oxygen from a {Ge 2 O 7 } simultaneously interacting with two cadmiums and one germanium. Therefore, its valence state is not −1 but −2, with the BVS result of 1.85. In conclusion, only six of the eight terminal Ge-O t oxygens are monoprotonated. Thus, there is still one hydrogen atom whose position cannot be determined. We think this hydrogen should be disorderedly distributed on the surface of the cluster.
There are also seven groups of oxygens in compound 3. However, only five of the seven have corresponding groups in compound 1. The five groups are: (1) V-O t terminal oxygens; (2) µ 3 -oxygens between two vanadiums and one germanium; (3) µ 3 -oxygens between three vanadiums; (4) µ 2 -oxygens between two germaniums; and (5) µ 2 -oxygen between one terminal vanadium and one cadmium. The remaining two groups are: (6) µ 3oxygen between two cadmiums and one germanium, which has the corresponding group in compound 2; and (7) µ 2 -oxygen between one cadmium and one germanium, which is only observed in compound 3. Compound 3 did not contain Ge-O t terminal oxygens, and all the Ge-O t terminal oxygens simultaneously interact with one or two cadmiums and finally form the group (6) and (7) oxygens. For the contributions of the cadmiums of group (6) and (7) oxygens, the BVS values of these oxygens are in the range of 1.56-2.01, indicating that there are no hydrogens attached on the cluster in compound 3.

IR Spectrophotometry
The IR spectra of compounds 1-4 were recorded in the regions between 4000 and 200 cm −1 ( Figure S1, Supporting Information). The strong peak at 984 cm −1 of compound 1 can be attributed to the stretching vibration of V=O. The patterns of the bands in the region characteristic of ν(V=O t ) indicate the presence of V IV sites: clusters which contain exclusively V IV generally possess ν(V=O t ) bands in the range of 970-1000 cm −1 , while bands in the region 940-960 cm −1 are characteristic of V V . The observation of a strong absorbance in the 970-1000 cm −1 region provides a useful diagnostic for the presence of V 4+ centers [78]. The strong peaks at 793 and 821 cm −1 of compound 1 may be due to asymmetric Ge-O stretching vibrations of {GeO 4 }. The infrared spectrum of compound 2 is very similar to that of compound 1. It also shows characteristic peaks at 983 cm −1 and 788 cm −1 , which should be ascribed to V=O t and Ge-O vibrations in compound 2.
Compounds 3 and 4 are based on Ge 6 V 15 , which is different from that of compounds 1 and 2. However, it should be noted that Ge 6 V 15 is also formed by {GeO 4 } and {V IV O 5 }; thus, the IR spectra of compounds 3 and 4 are very similar to those of compounds 1 and 2.
The IR spectra of compounds 3 and 4 present characteristic peaks at 979, 801 cm −1 and 982, 800 cm −1 , respectively, which correspond to V=O t and Ge-O vibrations in compounds 3 and 4. The main difference between the IR spectra of compounds 1 and 2 and 3 and 4 is that the bands at 667 and 660 cm −1 of compounds 1 and 2 are weak, but the corresponding bands at 691 and 692 cm −1 for compounds 3 and 4 are much stronger. Bands of 667-692 cm −1 can be ascribed to V-O-V vibrations.

XRD Powder Diffractometer
The powder X-ray diffraction patterns for compounds 1-4 are all in good agreement with the ones simulated based on the data of the single-crystal structures, indicating the purity of the as-synthesized products ( Figure S2). The differences in the reflection intensity are probably due to preferred orientations in the powder samples of compounds 1-4.

UV-Vis Spectrophotometry
The UV-vis spectra of compounds 1-4, in the range of 250-600 nm, are presented in Figure S3. The UV-Vis spectrum of compound 1 displays an intense absorption sharp peak centered at about 266 nm, a shoulder peak at 294 nm and a peak tailing to the longer wavelength side (to about 450 nm), which can be assigned to O→V charge transfer, n→π* transitions of phen ligands and d→d transitions of complexes in compound 1. The UV-Vis spectrum of compounds 2 displays an intense absorption peak at about 265 nm assigned to the O→V charge transfer in the polyoxoanion structure of compound 2. The peak corresponding to the n→π* transitions of phen ligands was overlapped by the O→V charge transfer and cannot be separated.
The UV spectra of compounds 3 and 4 are similar to each other, but are different from those of compounds 1 and 2, which exhibit absorption peaks at about 254 and 255 nm due to the O→V charge transfer in compounds 3 and 4. The difference in the UV-Vis spectra between compounds 3-4 and compounds 1-2 may be due to the difference in their clusters.

ESR Spectrophotometry
The ESR spectra of compounds 1-4 were studied at room temperature ( Figure S4). The ESR spectra of compounds 1-4 are very similar to one another, which show Lorentzian shapes accompanied by signals at g = 1.968, 1.968, 1.912 and 1.941, respectively, indicating that the vanadium atoms in compounds 1-4 are in a +4 oxidation-state. The ESR spectra further confirm the results of the bond valence sum calculations for compounds 1-4.

Catalytic Activity
Epoxidation is an important industrial reaction, and epoxides are key intermediates in the manufacture of a wide variety of valuable products [79][80][81]. The epoxidation of styrene to styrene oxide with aqueous tertbutyl hydroperoxide (TBHP) using compound 1, 2, 3, 4 or 5 as the catalyst was carried out in a batch reactor. In a typical run, the catalyst (compound 1 (2 mg, 0.57 µmol), compound 2 (2 mg, 0.60 µmol), compound 3 (2 mg, 0.58 µmol), compound 4 (2 mg, 0.62 µmol), compound 5 (2 mg, 0.70 µmol), 0.114 mL (1 mmol) of styrene and 2 mL of CH 3 CN were added to a 10-mL two-neck flask equipped with a stirrer and a reflux condenser. The mixture was heated to 80 • C and then 2 mmol of TBHP was injected into the solution to start the reaction. The liquid organic products were quantified using a gas chromatograph (Shimadzu, GC-8A, Beijing, China) equipped with a flame detector and an HP-5 capillary column and identified by comparison with authentic samples and GC-MS coupling. In a blank experiment carried out in the absence of catalyst, no products were observed. Also, the styrene epoxidation reactions in the presence of GeO 2 (2 mg, 19.1 µmol) and V 2 O 5 (2 mg, 11.0 µmol) were carried out respectively, and the activities are 24.8% and 71.2%, respectively, after 8 h. Table 2 shows the catalytic reaction results of TBHP oxidation of styrene over various catalysts. As expected, all the catalysts are active for the oxidation of styrene. Compound 1 as a catalyst shows a performance with 50.1% conversion and 62.8% selectivity to styrene oxide after 8 h. Compound 2 shows the highest activity among the five with 96.3% conversion and 71.6% selectivity to styrene oxide. Compound 3 shows a catalytic performance with 81.4% conversion and 63.0% selectivity. The performance of compound 4 is similar to that of compound 3 with 84.1% conversion and 55.5% selectivity. The activity and selectivity of compound 5 are 41.7% and 67.1%, respectively. Compounds 3 and 4 are based on Ge 6 V 15 , group 12 metals (Cd and Zn) and similar organic ligands (en and enMe), and both exhibit extended framework structures (3-D and 2-D). Therefore, the catalytic activities of the two are similar. The structures of compounds 2 and 5 are more similar to each other. Compounds 2 and 5 are based on similar Cd 2 Ge 8 V 12 clusters and similar cadmium complexes, and both exhibit similar 1-D extended structures. The significant difference between compounds 2 and 5 is that compound 2 contains aromatic organic ligands but compound 5 dose not; however, the catalytic activities of the two are thoroughly different from each other. To further understand the catalytic mechanism, we still need not only more Ge-V-O crystals but also more catalytic experimental results of the synthesized crystals. Although there have been no investigations on Ge-V-O metal-oxo-clusters as catalysts, there are some similar catalysis studies using catalysts formed by other POMs. The comparisons of the catalytic oxidation of styrene for compounds 1-5 and other reported POMs have been summarized in Table S2. The recyclability and reusability of compound 3, including the conversion and catalyst recovery in three cycles, were studied ( Table 3). The same experimental conditions were used. Generally, when using soluble heteropolyacid (e.g., H 3 [PW 12 O 40 ]) as the catalyst, the used catalyst was recovered by precipitation and ion exchange [82]. In comparison, it was easy to separate (centrifugation) and recycle compound 3. The process of recovery possibly resulted in the loss of approximately 40 wt.% after each cycle. The conversion dropped from 81.4% to 44.0% after three cycles. Recovery experiments showed that compound 3 suffered significant activity losses after three cycles. However, the residual catalyst of compound 3 and the as-synthesized crystals used for X-ray analysis can still be considered homogeneous ( Figure S5). The FT-IR spectra of compound 3 after the three cycles also remain identical to the one before the reaction ( Figure S6).

Conclusions
The synthesis of Ge-V-O clusters, especially secondary metal substituted Ge-V-O clusters is still a great challenge for chemists. In this manuscript, we synthesized compounds 1 and 2, which are the first examples formed by Ge-V-O clusters and transition metal complexes of aromatic organic ligands. Compounds 1 and 2 are also the first secondary metal substituted Ge-V-O clusters of aromatic organic ligands. Compound 3 is a novel 3-D framework with interesting channel structure. The catalytic properties of these compounds and two previously reported compounds have been investigated. We plan to apply these compounds in other oxidation catalytic reactions and hope to find applications of them in electrochemistry as well.