Yu-Fei
Song
ab,
De-Liang
Long
b and
Leroy
Cronin
*b
aState Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, BeiSanHuan East Road 15, 100029, Beijing, P. R. China
bWestCHEM, Department of Chemistry, University of Glasgow, Glasgow, G12 8QQ, UK. E-mail: L.Cronin@chem.gla.ac.uk; Web: http://www.croninlab.com Fax: (+44) 141 330 4888
First published on 27th August 2009
The assembly of polyoxometalate architectures based on Mn-Anderson clusters capped with different aromatic platforms via the ‘Tris’ linker (Tris = tris(hydroxymethyl)aminomethane), allows the formation and isolation of four new hybrid polyoxometalate clusters of 2–5 with the general composition of [n-N(C4H9)4]3[MnMo6O18{(OCH2)3CN = CH-R}2]·solv., (R = –(C6H5) 2, –(C6H5O) 3, –(C11H8N) 4 and –(C13H9) 5; solv. = MeOH or H2O). These compounds were obtained simply by the reaction of compound 1, ([n–N(C4H9)4]3[MnMo6O18{(OCH2)3CNH2}2]), with the corresponding aldehyde and the aromatic platforms were chosen to explore the design principles for covalently constructed POM aromatic–organic–inorganic hybrids.
The development of organic–inorganic hybrid POM building blocks, which comprise covalently connected cluster and organo fragments, is an important designed assembly since it allows the intimate combination of the properties of the metal-oxo and organic building blocks; this has been used to prepare polymers,9 dendrimers10 and macroporous materials.11 It is now apparent that organic components can dramatically influence the microstructures of inorganic oxides, thus providing a way for the design of novel materials.12 In this respect we are interested in the combination of highly conjugated organic molecules with POMs since this class of materials has hardly been explored, and should allow the assembly of electronically interesting systems.13–15 In addition, such organic–inorganic hybrid materials will not only combine the advantages of organic molecules such as structural fine tuning, but may also give rise to synergistic effects arising from the electronic interactions between the organic and inorganic parts.16 However, the major limitation is that the covalent functionalisation of POMs in general is not straightforward, depending critically upon the building blocks chosen.17 One route is to use a flexible synthetic strategy, in which the organic linker between the POMs and the organic site is designed to be bifunctional.18–21 Therefore in recent work we started to utilise ‘Tris’ (tris(hydroxymethyl) aminomethane (HOCH2)3NH2), in combination with a Mn-Anderson type POM21 and this was inspired by the work of Hasenknopf and Gouzerh et al., who first demonstrated the utility of this approach to construct hybrid POM-organic units.18–20
In this work we aimed to assemble a range of isostructural POM clusters, capped with a number of structurally related aromatic organic substituents, building on our recent results, see Fig. 1. Here, we use the Anderson cluster, which can be directly functionalised, both top and bottom, via the grafting of tripodal tri-alkoxides (e.g. derivatised from the ‘Tris’ ligand). This is because the number of conserved building blocks in POM-based architectures promises access to the controlled growth of nanoscale clusters with predetermined structures via a bottom-up route. Additionally, the systematic manipulation of the hybrid organic–inorganic POM-based clusters could yield a deeper insight into the self-assembly process and thereby allowing the design and construction of multifunctional materials based upon POMs.22 For example, the combination of highly delocalised, pyrene-based substituents with POM-based building blocks will allow us to examine the architectural aspects with respect to self assembly and crystallisation, see Fig. 1.21
![]() | ||
Fig. 1 A recent synthetic approach to the hybrid POM-aromatic architecture. LEFT: Representation of the [Mn-Anderson-(pyrene-tris)2]3−. RIGHT: Resulting structure of the compound produced showing the large void spaces. C – grey, O – red, H – white, N – blue, Mo – green (Mn and TBA cations not shown); H-bonds in purple. |
In the case of the [Mn-Anderson-(pyrene-tris)2]3−, when this was crystallised with TBA+ (tetrabutyl ammonium) counter ions, a crystalline material with nanoscale void spaces was formed and it could be seen that this network was built exclusively from C–H⋯O hydrogen bonds between the pyrene groups and the Anderson cluster. Further, there is the total absence of π⋯π interactions in the structure despite the eighteen pyrene moieties present in the unit cell. This is very interesting and this observation could help develop a new architectural principle if it could be shown that C–H⋯O hydrogen bonds could dominate the assembly of the supramolecular structure of the aromatic-POM hybrids.
Herein, we report the successful extension of this synthetic and architectural design strategy in the synthesis and self assembly of four new architectures with different end-capped aromatic rings grafted onto Mn-Anderson cluster: [n-N(C4H9)4]3[MnMo6O18{(OCH2)3CNCH–(C6H5)}2]·4H2O (2), [n–N(C4H9)4]3[MnMo6O18–{(OCH2)3CN
CH–(C6H5O)}2]· 3CH3OH (3), [n-N(C4H9)4]3[MnMo6O18{(OCH2)3CN
CH–(C11H8N)}2]·2CH3OH (4) and [n–N(C4H9)4]3-[MnMo6O18 {(OCH2)3CN
CH–(C13H9)}2] (5). The results presented here show that systematic manipulation of POM assemblies could be realized by applying ‘POM-linker’ strategy via covalent functionalization approach to a [Mn-Anderson-(Tris)2]3− building block.
![]() | ||
Scheme 1 Syntheses of new POM architectures of 2–5 by grafting various aromatic rings to the Mn-Anderson cluster of 1via covalent bonds. |
The reaction of 1 with benzaldehyde resulted in the formation of compound 2, which has been isolated in 63% yield as orange block crystals. Compound 2 crystallises in a tetragonal space group I41cd with Z = 16; a summary of crystal structure parameters and refinement details for this compound is given in Table 1. Detailed structural analysis (see Fig. 2a) revealed that the structure of compound 2 retains the Anderson structure of the parent compound 1 and the Mo–O bond distances and the O–Mo–O bond angles are all in the normal range for such Anderson clusters. The central Mn3+ ions are chelated by six oxygen atoms by two Tris moieties with the bond distances in the range of 1.918(9) and 2.090(8) Å. The supramolecular structure is defined by the intermolecular interactions that exist between terminal Mo–O and C–H of the benzene rings with the C⋯O bond distances ranging from 3.51(3) to 3.61(2) Å and C–H⋯O bond angles from 157.0(10) to 162.3(10)°, respectively. Further, it is these hydrogen bonds formed by the tips of the benzene rings (i.e. the hydrogen atoms in the 4 position) which connect the Anderson-based cluster spacer into a 1-D hydrogen bonded chain that is ‘coated’ with TBA cations, see Fig. 2. The TBA coating interacts with these 1-D chains via C–H⋯O hydrogen bonded interactions, which define the majority of the intermolecular interactions between the molecules in the cell (Fig. 2b). The intermolecular interactions can be observed between both bridging Mo–O–Mo and terminal Mo–O oxo ligands present in the Anderson-unit and C–H moieties of the TBA+ cations with C⋯O bond distances falling within the range of 3.50(5) to 3.60(3) Å and bond angles of the C–H⋯O interactions range 112.8(10) to 152.5(10)°, respectively.
Compounds | 2 | 3 | 4 | 5 |
---|---|---|---|---|
Empirical formula | C70H140 MnMo6N5O28 | C73H144MnMo6N5O29 | C83H150MnMo6N7O27 | C84H140MnMo6N5O24 |
M r/g mol−1 | 2130.45 | 2186.51 | 2308.68 | 2234.59 |
a/Å | 31.8004(6) | 24.516(3) | 37.5576(15) | 14.8970(4) |
b/Å | 31.8004(6) | 15.4509(17) | 13.7547(6) | 13.9137(4) |
c/Å | 36.9032(16) | 24.2330(3) | 25.1345(10) | 24.7439(7) |
α/° | 90 | 90 | 90 | 90 |
β/° | 90 | 94.577(5) | 128.073(2) | 97.349(1) |
γ/° | 90 | 90 | 90 | 90 |
V/Å3 | 37318.9(19) | 9150.1(18) | 10221.6(7) | 5086.6(2) |
Z | 16 | 4 | 4 | 4 |
Crystal system | Tetragonal | Monoclinic | Monoclinic | Monoclinic |
Space group | I41cd | P21/c | C2/c | P21/c |
Crystal color | orange | orange | orange | brown |
T (K) | 100 | 100 | 100 | 150 |
hkl | −33 ≤ h ≤ 33, −26 ≤ k ≤ 32, −38 ≤ l ≤ 32 | −29 ≤ h ≤ 29, −18 ≤ k ≤ 18, −28 ≤ l ≤ 26 | −44 ≤ h ≤ 44, −15 ≤ k ≤ 16, −28 ≤ l ≤ 29 | −18 ≤ h ≤ 14, −16 ≤ k ≤ 15, −29 ≤ l ≤ 29 |
No. of data collected | 54876 | 135771 | 76149 | 30648 |
No. of unique data | 10663 | 16069 | 9053 | 9434 |
R1 (I > 2σ(I) | 0.0570 | 0.0359 | 0.0496 | 0.0672 |
wR2 (all data) | 0.1524 | 0.0860 | 0.1635 | 0.2066 |
![]() | ||
Fig. 2 (a) Crystal structure of compound 2 showing the cluster with the appended benzene moiety. The purple arrows show the direction of the hydrogen-bonded chain. (b) Illustration of the hydrogen-bonding interaction between each Anderson cluster in the unit cell including the interactions with the TBA counterions (shown in yellow). Mn orange, Mo ink blue, C grey, N yellow. TBA shown in yellow (c) Packing projection onto the ab plane with the TBA counter ions omitted. The hydrogen-bonded interactions are shown in purple. |
Compound 3 was synthesized by reaction of 1 and salicyaldehyde and this compound was obtained as orange block crystals in a yield of 67%. Structural analysis of 3 by single-crystal X-ray diffraction shows that the phenol groups have been successfully capped on the Anderson cluster and it crystallizes in a monoclinic space group P21/c with a Z = 4. Compound 3, like compound 2, also comprises the tris-Anderson building block and can be formulated as [n-N(C4H9)4]3[MnMo6O18{(OCH2)3CNCH (C6H5O)}2]·3CH3OH. Similar to compound 2, extensive hydrogen bonded interactions can be seen in the structure however there are two different Anderson clusters in the asymmetric unit. One cluster-hybrid forms hydrogen bonded 1-D chains (Fig. 3c-red), via phenyl C–H interactions with the adjacent cluster unit whilst the other accepts C–H bonds into the metal oxide framework (Fig 2c-blue). The 1-D chains form between the C–H moieties of the appended phenyl groups and two terminal Mo = O groups with C⋯O distances ranging 3.338(6)–3.343(6) Å and C–H⋯O angles of 127.4(3)–166.1(3)°. The remainder of the intermolecular interactions are either solvent based or from the TBA cations; Anderson cluster interactions via C–H⋯O interactions involving both terminal and bridging oxo-ligands with the C⋯O bond distances and C–H⋯O bond angles range from 3.270(5) to 3.641(8) Å and from 133(1) to 146.1(3)° (see Fig 2c). By grafting phenol groups onto the Anderson cluster, the arrangement of the capped aromatic rings in the resulting structure of compound 3 shows differences from that in compound 2, for example, the dihedral angle of the neighbouring benzene rings in 2 is around 88(1)°, close to being perpendicular to each other; whereas the dihedral angle of two phenol rings in 3 is around 12(2)°. It should be noted that the molecular packing of 3 shows an ordered, 2-D corrugated framework whereby the TBA ‘glues’ the 1-D chains together to form a 2-D network (Fig. 3d).
![]() | ||
Fig. 3 (a) X-ray single crystal structure of compound 3 and the central Mn-Anderson part with the organic appendages. (b) Shows the 1-D chain formed by three units with associated TBA cations. (c) Shows the alternating 1-D chain form by the ‘donor’ Anderson moieties (red) and acceptor groups (blue). (d) shows a representation of the molecular packing viewed from ‘c’ direction; The TBA+ cations are shown in yellow in (b) and omitted in (a), (c) and (d) for clarity (Mn orange, Mo ink blue, O red and C grey). |
Compound 4 was obtained as orange square crystals by reaction of 1 with 2-pyridine-benzaldehyde in a yield of 65%. This compound crystallises in a monoclinic group C2/c with a Z = 4. The crystal structure of 4 shows that the organic groups have been attached on both sides of the Mn-Anderson cluster via the grafted imine moieties. It is noted that the pyridine rings are fully co-planar with the benzene rings, which are almost perpendicular to the MnMo6 plane (Fig. 4a). The extended aromatic platforms therefore serve as multiple H-bond donors where as the Anderson cluster accepts multiple H-bonds. The packing in compound 4 reveals the formation of layers (Fig. 4b) whereby each unit can interact via C–H⋯O (C⋯O range 3.19(1)–3.39(1) Å and 127.8(6)–169.6(5)°) and π⋯π interactions (3.32(2) Å). Once again, the TBA cations are located in the ‘pores’, see Fig. 4c, and also interact via C–H⋯O interactions.
![]() | ||
Fig. 4 (a) Crystal structure of compound 4 showing the central Anderson cluster and the appended organic groups. (Mn orange, Mo ink blue, O red and C grey). (b) Illustration of intermolecular interactions in compound 4. (c) Crystal packing shows the pores; solvents and TBA+ are omitted for clarity. The H-bonds are shown in purple. |
Compound 5 was obtained with a planar, aromatic fluorene rings capped at both sides of Mn-Anderson cluster (Fig 5a). The crystal structure of 5 crystallises in a monoclinic space group P21/c. Structural analysis of 5 shows that intermolecular interactions between the TBA+ cations and the terminal Mo–O atoms with the bond distances and bond angles, but more importantly the fluorene group is able to hydrogen bond to two terminal MoO units present in the Anderson cluster and these C⋯O interactions range 3.28(1) to 3.45(1) Å with C–H⋯O bond angles that range 141.0(6) to 173.5(6)°, respectively. As in compound 4, the extended aromatic fluorene moiety multiple (in this case 2 2) C–H hydrogen bonds fluorene which interact with two terminal oxo ligands on the Anderson cluster framework (Fig. 3b).
![]() | ||
Fig. 5 (a) Representation of crystal structure of 5 showing the central Anderson and the two organic groups. (Mn orange, Mo ink blue, N orange and C grey). (b) Illustration of the intermolecular hydrogen bonding interactions the cluster and the C–H groups of the fluorene and the terminal Mo–O of the Anderson cluster, respectively; (c) Illustration of 5 looking down the b axis, which shows the direction of the 1-D chains. |
Given the weak interactions that are responsible for the assembly of the supramolecular structures of compounds 2–5, we decided to investigate the stability of the compounds using TGA and DSC. Interestingly all the compounds gave almost identical results, as shown in Fig. 6. The TGA results, done on air dried samples, show that the solvent of crystallisation is lost by the air drying and the first weight loss is at ca. 290 °C which corresponds to the decomposition and loss of three TBA+ cations in each of the samples 2–5 (see Fig. 6. and ESI†). The DSC measurements show a very sharp endothermic absorption at between 250–270 °C for compounds 2–5 (see ESI†) which is consistent with a melting process. This is interesting since despite the weak interactions that define the structures, they appear to be robust until at least 250 °C.
![]() | ||
Fig. 6 (a) Representative TGA and (b) DSC curves of 2. The TGA curve shows weight loss due to the decomposition of the TBA cations, which are observed before decomposition of the cluster unit. The DSC curve shows a very rapid melting and recrystallisation process at ca. 270 °C. |
![]() | ||
Fig. 7 Representation of the building blocks found in compounds 2–5. The hydrogen bonded interactions are shown in purple (donor) and yellow (acceptor). |
Preparation of compound 3, 4 and 5 were carried out in a similar way except that benzaldehyde were replaced with salicylaldehyde, 4-(2-pyridyl)benzaldehyde and fluorene-2-carboxaldehyde, respectively. Characterization of these compounds is listed below.
Footnote |
† Electronic supplementary information (ESI) available: TGA and DSC of compounds 3–5. Structures of 2–5 as cif files are available. CCDC reference numbers 733962–733965. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/b910434h |
This journal is © The Royal Society of Chemistry 2010 |