Incorporation of a high-spin heptanuclear [Cu^II₆Gd] cluster into carboxyl-functionalized mesoporous silica

Abstract

We report the immobilization of heptanuclear [Cu₆Gd] high-spin magnetic clusters, namely [Cu₆Gd(aib)₆(OAc)₃(NO₃)₃(OH)₃] (where, aib = (CH₃)₂C(NH₂)COO⁻ and OAc = CH₃COO⁻), inside the channels of –COOH-functionalized mesoporous SBA-15 silica. X-ray diffraction, TEM and N₂ physisorption experiments show that the well-ordered hexagonal structure of the CA/SBA-15 hybrid material is well preserved and that the [Cu₆Gd] clusters are located inside the mesopores. The molecular structure of the incorporated [Cu₆Gd] was confirmed by elemental X-ray microanalysis, IR and UV-vis-NIR spectroscopy. Magnetic measurements indicate that the polynuclear complexes retain their magnetic properties during the impregnation process. The host–guest nanocomposite displays strong ferromagnetic interactions that result in a high-spin S = 13/2 ground state.

---

Introduction

The development of molecule-based materials with fine-tuned magnetic behavior is of great importance in materials science and nanotechnology. Such materials represent a significant dimensional reduction of magnetic grain size where single molecules act as tiny magnets. The polynuclear complexes that behave as molecular nanomagnets, referred to as single-molecule magnets (SMMs), have drawn significant attention.^1,2 These molecular species display slow relaxation of the magnetization associated with a large spin ground state, S, and high negative magnetic anisotropy, D, under cryogenic conditions. Remarkably, the magnetic hysteresis is due to discrete molecules rather than long-range magnetic ordering as observed in bulk metal and metal oxide magnets.³ Therefore, SMMs are promising for a variety of applications, including quantum computation,⁴ high-density magnetic recording⁵ and molecular-based spintronics.⁶ In addition, in the field of molecular magnetism new species have emerged that can lower the temperature of their surroundings once a magnetic field is applied due to the magnetocaloric effect, and therefore can be used as molecular coolants and refrigerants.^7–9 For both single-molecule magnets and molecular refrigerants, the main prerequisites are the low anisotropy, D, and high-spin ground state, S, of the complex.

Therefore, it becomes evident that the chemical synthesis of high-spin and/or ferromagnetic complexes is of vital importance for the employment of such species into technological applications. Consequently, the organization of magnetically interesting units in two- and three-dimensional architectures is the next step towards potential applications. For this reason, different methods of integrating molecular clusters (i.e. SMMs) into ordered arrays have been in development in recent years. These include deposition in Langmuir–Blodgett films,^10,11 incorporation in ordered nanoporous silica^12–14 and polymer matrices,¹⁵ and adsorption on self-assembled monolayers^16,17 and conducting substrates.¹⁸ Among them, the incorporation of complexes into the pore system of organically-modified silica is particularly attractive because of the benefits of ordered pore structure and chemical functionalities of the silica support. As the organic moieties attached to the silica surface allow for immobilization of inorganic complexes, the porous structure can induce fast mass transfer and molecular recognition behaviour. Nanosized arrays of these composites are expected to be of practical applications, especially in the fields of magnetic sensors and magnetic separation. However, examples concerning the synthesis of silica-supported materials are relative scarce as compared with other porous host–guest composites, mainly because the inclusion processes should take place without disturbing the structure of magnetic clusters and host phase. Moreover, the molecular complexes have to incorporate into the host structure in a robust and well-dispersed fashion, avoiding the formation of agglomerates.

Herein, we managed to overcome these difficulties and have successfully immobilized high-spin heptanuclear heterometallic copper(II)–gadolinium(III) molecules of formula [Cu₆Gd(aib)₆(OAc)₃(NO₃)₃(OH)₃] (where, aib = 2-aminoisobutyrate and OAc = acetate), referred to here as [Cu₆Gd], into a hexagonal mesoporous –COOH-functionalized silica. The [Cu₆Gd] metallic core is formed by a central {GdO₉} polyhedron that is surrounded by six Cu(II) ions in an octahedral arrangement. The Cu ions are linked by μ₃-O²⁻ bridges to each other and to the central {GdO₉} unit. Furthermore, this complex contains six peripheral monoanionic α-aminoisobutyrate ligands able to selectively interact with the anchored carboxyl groups of the silica framework. The metallic core of this complex exhibits dominant ferromagnetic interactions leading to a high-spin ground state of S = 13/2 (vide infra). The results of this work indicate that the [Cu₆Gd] polynuclear complexes are successfully embedded in the cavities of mesoporous silica, keeping their molecular structure. Also, these clusters preserve their magnetic properties, maintaining dominant ferromagnetic exchange interactions within the silica matrix.

Experimental

Synthesis of heptanuclear Cu(II)-Gd(III)complex

The [Cu₆Gd(aib)₆(OAc)₃(NO₃)₃(OH)₃] (aib = 2-aminoisobutyrate and OAc = acetate) complex was prepared following the method described in the literature.¹⁹

Synthesis of –COOH-functionalized mesoporous silica

The mesoporous silica functionalized with carboxyl groups was prepared as previously reported but with some modifications.²⁰ Briefly, 2 g of Pluronic P123 (PEO₂₀PPO₇₀PEO₂₀, where PEO = poly(ethylene oxide) and PPO = poly(propylene oxide), Aldrich) was dissolved in a solution containing 75 ml of water, 11 ml of 37% HCl (Sigma-Aldrich) and 4.5 mmol of 3-(triethoxysilyl)propionitrile (Aldrich, 98%). After stirring for 30 min, 17.5 mmol of tetraethylorthosilicate (TEOS, Merck, ≥99%) was slowly added and the resultant mixture was vigorously stirred for 24 h at 40 °C. After this, the mixture was left to react at 90 °C for 2 days under static conditions and the white powder was recovered by filtration, washed thoroughly with ethanol and acetone, and dried at 90 °C to give cyanide-functionalized silica (CN/SBA-15). To remove the surfactants inside the pores and hydrolyze the cyano groups, 1 g of dried CN/SBA-15 was dispersed in 120 ml of 48 wt% H₂SO₄ solution (Sigma-Aldrich) and the mixture was refluxed at 95 °C for 18 h. The product was filtered off, washed copiously with ethanol and acetone, and dried at 90 °C for 12 h to obtain carboxyl-functionalized silica (CA/SBA-15).

Incorporation of [Cu₆Gd] clusters into the CA/SBA-15

The impregnated [Cu₆Gd]/CA-SBA hybrid silica was prepared by suspending 55 mg of dried CA/SBA-15 in 12 ml of CH₃CN (SDS, Carlo Erba, 99.7% pure) containing 160 mg of [Cu₆Gd] compound. The slurry was refluxed at 75 °C for 4 h and the light blue solid was collected by filtration, washed thoroughly with acetonitrile and acetone until the eluent became colorless and dried under vacuum at 60 °C.

For comparative studies, the impregnated Cu₆Gd/SBA-15 sample was also prepared by suspending 50 mg of mesoporous SBA-15 silica²¹ in 4 ml of acetonitrile followed by addition of 62 mg of [Cu₆Gd]. The slurry was vigorously stirred for 30 min and the solvent was evaporated slowly at 40 °C. During this process, the compound diffused into the pores of silica by capillary action. The solid product was dried at 60 °C to yield the 55% Cu₆Gd/SBA-15 sample. The molecular dispersion of [Cu₆Gd] in the silica matrix was confirmed by powder X-ray diffraction, which indicated the absence of Bragg peaks (result not shown here).

Physical characterization

Thermogravimetric (TGA) and differential thermal (DTA) analysis experiments were performed in a Perkin-Elmer Diamond thermal analyzer by heating the sample up to 600 °C at heating rate of 5 °C min⁻¹ under nitrogen flow of ∼100 ml min⁻¹. The X-ray diffraction (XRD) patterns were collected on a PANanalytical Xpert Pro MPD X-ray diffractometer operated at 45 kV and 40 mA using Ni-filtered Cu-Kα radiation (λ = 1.5418 Å) in Bragg–Brentano geometry. Elemental microprobe analyses were performed on a JEOL Model JSM-6390LV scanning electron microscope (SEM) equipped with an Oxford INCA PentaFET-x3 energy dispersive X-ray spectroscopy (EDS) detector. The sample was ground and fixed on a carbon-taped aluminium stub for elemental analysis. Data acquisition was performed several times in different areas of sample using an accelerating voltage of 20 kV and acquisition time of 100 s. Transmission electron microscopy (TEM) was performed using a JEOL Model JEM-2100 electron microscope equipped with an EDS detector (Oxford Instruments) and operated at an accelerated voltage of 200 kV with a thermionic (LaB₆) emission source. Samples were prepared by dispersing finely ground powders in ethanol using sonication, followed by depositing a drop of solution onto a Ni grid coated with organic (Formvar) film. Nitrogen adsorption and desorption isotherms were measured at −196 °C with a Quantachrome Autosorb 1-MP volumetric analyzer. Before analysis, the samples were degassed overnight at 353 K under vacuum (<10⁻⁵ Torr). The specific surface areas were calculated using the Brumauer–Emmett–Teller (BET) method²² on the adsorption data in the 0.05–0.2 relative pressure (p/p₀) range. The pore size distributions were derived from the adsorption branch of isotherms, using the Barrett–Joyner–Halenda (BJH) method²³ with modified statistical film thickness equation (KJS method).²⁴ Infrared spectra were recorded on a Perkin-Elmer Model Spectrum-BX FT-IR spectrometer. Spectra were obtained on averaging 100 interferograms with resolution of 2 cm⁻¹. Samples were prepared as KBr pellets. Diffuse reflectance UV-vis-near IR spectra were obtained on a Perkin-Elmer Lambda 950 optical spectrophotometer equipped with an integrating sphere. BaSO₄ powder was used as a 100% reflectance standard and base material on which the powder sample was coated. The reflectance data were converted to absorption using to the Kubelka–Munk equation:²⁵α/S = (1 − R)²/(2R), where R is the reflectance and α, S are the absorption and scattering coefficients, respectively.

Magnetic measurements

Variable-temperature, solid-state direct current (dc) magnetic susceptibility data down to 5 K were collected on a Quantum Design MPMS-XL SQUID magnetometer equipped with a 7 T dc magnet for complexes. Diamagnetic corrections were applied to the observed paramagnetic susceptibilities using Pascal's constants.

Results and discussion

Synthesis and chemical composition

The synthesis of a hybrid CA/SBA-15 material, i.e. –COOH-functionalized mesoporous silica, was performed by a two-step grafting and chemical transformation of cyano groups onto the silica matrix. Namely, we utilized direct co-condensation of ethylcyano silane ((EtO)₃Si(CH₂)₂CN) and tetraethyl orthosilicate (Si(OEt)₄) monomers in the presence of non-ionic surfactant, followed by chemical treatment with H₂SO₄ to remove the template and hydrolyze the –CN groups into –COOH functionalities. The carboxylic groups at the surface provide a suitable medium through which the [Cu₆Gd] complexes can be anchored via strong chemical bonds. Inclusion of [Cu₆Gd] within the silica host was carried out by refluxing a mixture of CA/SBA-15 and [Cu₆Cd] in acetonitrile for 4 h (Scheme 1). The reaction conditions were optimized taking into account the solubility and stability of [Cu₆Cd] complex. Prolonging the reaction time does not further increase the loading amount of [Cu₆Gd], while other solvents like DMSO, DMF and methanol cause progressive disruption of the molecular structure.

Scheme 1 Schematic representation showing the synthesis of Cu₆Gd/CA-SBA hybrid material.

The chemical composition of CA/SBA-15 and Cu₆Gd/CA-SBA hybrid materials was determined with thermogravimetric/differential thermogravimetric (TG/DTG) and differential thermal (DTA) analysis and energy dispersive X-ray spectroscopy (EDS). The TG curve of CA/SBA-15 shows an appreciable weight loss of ∼11.1 wt% below 320 °C due to release of physisorbed or chemisorbed solvent left in the silica network and possible decomposition of the residual template (Fig. 1a). The rapid weight loss of ∼16.1 wt% observed between 320 and 595 °C, which is accompanied by a broad exothermic peak centered at ∼430 °C in DTA (Fig. 1b), could be attributed mainly to the decomposition of anchored organic groups. This result indicates a concentration of about 2.2 mmol of carboxyl groups per gram of sample, which is considered very high as compared to other values in the literature.^26,27 Note that small fluctuations may also contribute to the weight loss in this temperature range due to the hydroxylation of silanol species, but these can be considered minimal.

Fig. 1 (a) TG/DTG and (b) DTA profile of mesoporous CA/SBA-15 hybrid material.

Elemental energy dispersive X-ray (EDS) microanalysis of the Cu₆Gd/CA-SBA composite showed the presence of Cu and Cd elements with the expected atomic ratio of Cu : Gd ≈ 6 : 1. This indicates that the chemical identity of the inserted [Cu₆Gd] clusters is retained. The EDS analysis along with the TG/DTA results reveal a ∼55 wt% of [Cu₆Gd] content in Cu₆Gd/CA-SBA, which corresponds to about one [Cu₆Gd] unit per ∼14 carboxyl groups.

Mesoporous structure and surface area

The mesoscopic order of CN/SBA-15, CA/SBA-15 and Cu₆Gd/CA-SBA samples was investigated with powder X-ray diffraction (XRD) and transmission electron microscopy (TEM). The diffraction patterns of CN/SBA-15 and CA/SBA-15 show an intense diffraction peak accompanied by a weak feature in the 2θ range of 0.7–2.2°, which according to TEM could be attributed respectively to the (100) and (200) reflections of hexagonal P6mm structure (Fig. 2a). The observation of these peaks constitutes a clear evidence of the long-range hexagonal order in these materials. The XRD pattern of Cu₆Gd/CA-SBA resembles that of the pristine CA/SBA-15 solid, indicating that the hexagonal array of the silica host is preserved undisrupted even after [Cu₆Cd]-functionalization. The small decrease in intensity of the (100) peak and absence of a high-order (200) reflection for the [Cu₆Cd]-loaded sample is attributed to the reduction in X-ray scattering contrast between the silica walls and the pore-filling cluster.²⁸ The mesoporous structures show similar lattice parameters (a₀) of ∼11.6–11.8 nm, as calculated from the angular position of the (100) diffraction peak. In wide-angle XRD (2θ > 20°) of Cu₆Gd/CA-SBA, the essential absence of Bragg diffractions excludes the possibility of large [Gu₆Gd] ensembles within the silica matrix (see Fig. 2b). This means that the [Cu₆Gd] clusters are well distributed in the silica mesopores, which is a fascinating property if we consider the very high [Cu₆Gd] loading (ca. 55 wt%) of this material.

Fig. 2 (a) Low-angle XRD patterns of (i) CN/SBA-15, (ii) CA/SBA-15 and (iii) Cu₆Gd/CA-SBA hybrid materials. (b) Wide-angle XRD patterns of (i) mesoporous Cu₆Gd/CA-SBA and (ii) polycrystalline [Cu₆Gd] compound.

Fig. 3, displays representative TEM images and the corresponding fast Fourier transform (FFT) patterns of the mesoporous Cu₆Gd/CA-SBA. The TEM images, recorded along the [001] and [110] directions, reveal large domains of hexagonally arranged pore channels. These results along with low-angle XRD data confirm that the long-range order of the pristine CA/SBA-15 material is preserved to a considerable extent after immobilization of the [Cu₆Gd] complexes. A detailed analysis of TEM images indicates an average pore-to-pore distance of about 11.2 nm and a pore size of about 7.1 nm, consistent with XRD and N₂ physisorption results (see below). Although the [Cu₆Gd] clusters inside the mesopore channels are difficult to observe by TEM imaging, their presence was evident by selective area X-ray microanalysis (TEM-EDS). The TEM-EDS spectra acquired across various local areas of mesoporous structure display strong signals from silicon, copper and gadolinium elements with atomic ratios of Si/Cu/Gd ∼ 67.1 : 28.0 : 4.8 (Fig. 3d). These values correspond to ∼0.38 mmol of clusters per gram of host material, which is very close to the compositional estimate from scanning electron microscopy (SEM) EDS analysis (∼0.35 mmol g⁻¹). Noteworthy, no visible Gd-rich aggregates were observed on the surface grains, confirming that the molecular clusters are exclusively hosted inside the mesopores.

Fig. 3 (a–c) Typical TEM images and corresponding FFT patterns (insets) of the mesoporous Cu₆Gd/CA-SBA taken along the [001] (a) and [110] (b and c) zone axes of hexagonal P6mm symmetry. (d) EDS spectrum obtained on TEM for mesoporous Cu₆Gd/CA-SBA. The nickel signal comes from the Ni TEM grid.

Nitrogen physisorption measurements were performed to assess the textural properties of the CN/SBA-15, CA/SBA-15 and Cu₆Gd/CA-SBA materials. All the samples display typical type-IV adsorption–desorption isotherms with a pronounced H₃ hysteresis loop, according to the IUPAC classification, characteristic for mesoporous solids with ink-bottle pore structure (Fig. 4a).²⁹ The sharp increase in the adsorption branch at relative pressure range p/p₀ = 0.6–0.7 is related to the capillary condensation of nitrogen in narrow-sized mesopores. This suggests that despite the presence of organosilanes and [Cu₆Gd] clusters onto the silica walls, a significant portion of mesopores are filled with nitrogen, leading to an inflection point in the adsorption curve. As expected, the CN/SBA-15 hybrid sample exhibits a low nitrogen adsorption due to the filling of mesopores with template, while the template-free CA/SBA-15 demonstrates a significant increase in nitrogen uptake which is reflected in the higher surface area and pore volume. Analysis of absorption data with the Brunauer–Emmett–Teller (BET) method indicates that mesoporous CN/SBA-15 and CA/SBA-15 have a surface area of 136 and 578 m² g⁻¹ and a pore volume of 0.23 and 0.96 cm³ g⁻¹, respectively. After insertion of [Cu₆Gd] into the pores of CA/SBA-15, the surface area and pore volume drop to 140 m² g⁻¹ and 0.21 cm³ g⁻¹, respectively, as a consequence of the pore clogging and the heavier structure due to the added [Cu₆Gd] clusters. Nevertheless, the Cu₆Gd/CA-SBA composite still possesses a noteworthy surface area and porosity which is accessible on adsorption. All the textural parameters of the CN/SBA-15, CA/SBA-15 and Cu₆Gd/CA-SBA hybrid materials are given in Table 1.

Fig. 4 (a) N₂ adsorption–desorption isotherms at 77 K and (b) corresponding BJH pore size distribution calculated from the adsorption branch of mesoporous materials: (i) CN/SBA-15, (ii) CA/SBA-15 and (iii) Cu₆Gd/CA-SBA.

Table 1 Textural properties of mesoporous CN/SBA-15, CA/SBA-15 and Cu₆Gd/CA-SBA materials

Sample	Surface area (m^2 g^−1)	Pore volume^a (cm^3 g^−1)	Pore size (nm)	d 100 spacing (nm)	Unit cell^b (nm)
a Cumulative pore volume at p/p0 = 0.95. b The size of unit cell is given by a0 = (2/√3)d100.
CN/SBA-15	136	0.23	5.8	10.06	11.6
CA/SBA-15	578	0.96	7.4	10.25	11.8
Cu6Gd/CA-SBA	140	0.21	6.9	10.02	11.6

---

The pore size distributions were calculated from the adsorption branches of the isotherms using the modified Barrett–Joyner–Halenda (BJH) method. All the samples exhibit a quite narrow distribution of pore size with a peak maximum being varied with the treatment process (Fig. 4b). Specifically, after acid treatment of CN/SBA-15, the pore size was significantly increased from ∼5.8 to ∼7.4 nm, owing to the elimination of surfactants. While incorporation of [Cu₆Gd] in CA/SBA-15 mesopores led to a reduction of the pore width (∼6.9 nm) due to the pore occupancy by the immobilized clusters.

Spectroscopy

The organic functionalities as well as the molecular structure of [Cu₆Gd] complexes on the silica surface were characterized by Fourier transform infrared (FT-IR) and diffuse reflectance ultraviolet-visible-near IR (UV-vis-NIR) spectroscopy. The IR spectra of mesoporous CN/SBA-15, CA/SBA-15 and Cu₆Gd/CA-SBA materials and the [Cu₆Gd] compound are shown in Fig. 5. The hybrid materials show two intense absorption peaks at ∼974 cm⁻¹ and ∼1045–1215 cm⁻¹ due to the Si–OH and Si–O–Si vibration bands of the silica matrix, respectively.³⁰ The CN/SBA-15 shows a characteristic absorption peak at ∼2257 cm⁻¹, corresponding to the stretching vibration of the cyano (C≡N) groups. The weak absorption peaks at ∼2878 and ∼2932 cm⁻¹ are ascribed to the C–H stretching vibration modes of the ethyl chain. After treatment with sulfuric acid, an intense peak appears at ca. 1723 cm⁻¹ that is related to the carbonyl (C=O) stretching mode of the carboxylic species. This absorption band is significantly bathochromic/red-shifted compared to the monomeric propanoic acid (ca. 1783 cm⁻¹),³¹ indicating strong hydrogen-bonding interactions of –COOH groups with other neighboring –COOH groups and/or silanol (SiOH) species at the surface.³² Indeed, the broad absorption in the 3000–3600 cm⁻¹ range is associated with the hydrogen-bonded hydroxyl species. These results together with the strong reduction of the cyano peak imply a complete transformation of the –CN groups into –COOH functionalities.

Fig. 5 Infrared spectra of hybrid (i) CN/SBA-15, (ii) CA/SBA-15 and (iii) Cu₆Gd/CA-SBA materials and (iv) [Cu₆Gd] compound.

The IR spectrum of Cu₆Gd/CA-SBA indicates the presence of [Cu₆Gd] compounds, showing a series of absorption peaks in the region of 620–850 cm⁻¹ due to the overlap of N–H and C–H bending vibration bands of α-aminoisobutyrate ligands. The vibrational feature at ∼578 cm⁻¹ is ascribed to the stretching mode of Gd–O bonds in the cluster. The absence of any shift for the Gd–O absorption suggests that the cluster core is fully retained within the mesoporous host structure. Compared to the pristine CA/SBA-15, the Cu₆Gd/CA-SBA exhibits a remarkable bathochromic shift (by over 60 cm⁻¹) of the carbonyl stretching band. This suggests that the [Cu₆Gd] complexes are strongly anchored to the terminal carbonyl groups possibly via –CO–N amide bonds.

Fig. 6 shows the diffuse reflectance UV-vis-NIR spectra of mesoporous CA/SBA-15 and Cu₆Gd/CA-SBA materials and [Cu₆Gd] compound. The CA/SBA-15 exhibits a strong absorption at ∼226 nm accompanied by a weak shoulder at ∼273 nm, which are attributed to the π → π* and n → π* electron transitions in the C=O bond of carboxylic acids, respectively.³³ The analogous absorption bands in the spectrum of the [Cu₆Gd]-loaded sample are hard to observe because of the high concentration of complex. The Cu₆Gd/CA-SBA exhibits an intense absorption at ∼270 nm, corresponding mainly to the O(−II) → Cu(II) charge transfer transition in Cu–O bonds and a broad absorption peak at low-energy regime (∼700–800 nm), corresponding to the internal d–d electron transition of Cu(II) ions of the polynuclear [Cu₆Gd] complex.^34,35 The series of absorption peaks in the range of 1400–2000 nm are assigned to the intra-atomic d–d electron transitions of tetrahedral Cu(II) species.³⁶ These absorption bands are slightly bathochromic shifted (about 5 nm) compared to the pure [Cu₆Gd] solid, possible due to the binding of [Cu₆Gd] clusters to the carboxylic acids.

Fig. 6 Diffuse reflectance UV-vis-NIR spectra of mesoporous CA/SBA-15 and Cu₆Gd/CA-SBA materials and [Cu₆Gd] compound.

To investigate the stability of the incorporated clusters in the host material, Cu₆Gd/CA-SBA was dispersed in acetonitrile under vigorous stirring for 1 h and the copper and gadolinium content remaining in sample were determined by SEM-EDS analysis. For comparison we also prepared a mesoporous Cu₆Gd/SBA-15 composite by impregnating 55 wt% of [Cu₆Gd] on the surface of mesoporous SBA-15 silica and this sample was examined for [Cu₆Gd] leaching under similar conditions. The analytical data, in Table 2, indicated no appreciable loss of [Cu₆Gd] clusters (ca. 3%) from Cu₆Gd/CA-SBA, suggesting strong binding of [Cu₆Gd] units to the –COOH functional groups. This is in line with the results from IR and UV-vis-NIR spectroscopy. On the contrary, however, the impregnated Cu₆Gd/SBA-15 sample suffered a remarkable depletion of [Cu₆Gd] (ca. 21%), which could be explained by the weak interactions between the complexes and the surface silanol species. These results demonstrate that the modification of silica surface by carboxyl groups is essential for the strong anchoring of [Cu₆Gd] compounds.

Table 2 Analytical data of mesoporous Cu₆Gd/CA-SBA and Cu₆Gd/SBA-15 composites after treatment in acetonitrile

Sample	Atomic ratio^a Si : Cu : Gd	[Cu6Gd] loading (wt%)	[Cu6Gd] loss^b (%)
a SEM-EDS elemental data normalized to the six-atom Cu cluster. b Percent of clusters lost determined by reference to the initial [Cu6Gd] content in composite samples (55 wt%).
Cu6Gd/CA-SBA	16.1 : 6.0 : 1.0	53	3.2
Cu6Gd/SBA-15	34.3 : 6.0 : 1.0	43	21.3

---

Magnetic properties

Direct current magnetic susceptibility measurements were performed on samples of Cu₆Gd/CA-SBA material and pure [Cu₆Gd] compound in the 5–300 K range under an applied field of 0.1 T. The results are plotted as the χ_MT product vs. T in Fig. 7. For the pure complex, the χ_MT value of 10.69 cm³ mol⁻¹ K at 300 K is as expected for six non-interacting Cu(II) ions (g = 2.24) and one Gd(III) centre (g_j = 2.0). Upon cooling, the χ_MT value increases to a maximum value of 17.31 cm³ mol⁻¹ K at 5 K, indicating the presence of dominant ferromagnetic interactions within the metallic cluster, stabilizing a high-spin ground-state of S = 13/2 for the complex. The same situation is observed for the Cu₆Gd/CA-SBA material; at room temperature the χ_MT value is 10.71 cm³ mol⁻¹ K, and upon cooling the χ_MT value starts increasing as in the case of the pure complex. In the 20–300 K temperature range the two species display identical magnetic behaviour, strongly suggesting that once immobilized within the silica host the [Cu₆Gd] unit retains both its structural and magnetic integrity. Below ∼20 K, the magnetic behaviour of the immobilized cluster deviates from the pure complex, most probably due to intermolecular magnetic interactions between neighbouring clusters within the mesoporous material.

Fig. 7 Plot of χ_MT vs. T for Cu₆Gd/CA-SBA material and pure [Cu₆Gd] complex under an applied dc field of 1000 G.

Conclusion

In summary, we have reported the encapsulation of the high-spin heptanuclear copper(II)–gadolinium(III) molecular compound [Cu₆Gd] within –COOH-functionalized mesoporous silica. The ordered mesoscopic structure of the hybrid CA/SBA-15 material, as assessed by powder X-ray diffraction and transmission electron microscopy, is maintained undisturbed after incorporation of the [Cu₆Gd] compounds. The organic functionalities and molecular structure of incorporated [Cu₆Gd] clusters were confirmed by IR and diffuse reflectance UV-vis-NIR spectroscopy. Nitrogen physisorption indicated that the [Cu₆Cd]-loaded material preserves the mesoporosity of the pristine CA/SBA-15 sample to a certain extent, but with surface area (from 578 to 140 m² g⁻¹) and pore size (from 7.4 to 6.9 nm) reduction due to the occupancy of mesopores by the immobilized clusters. The Cu₆Gd/CA-SBA nanocomposite exhibits magnetic properties that derive from the incorporated clusters. Magnetic measurements showed that this material has similar magnetic properties to that of pure [Cu₆Gd], presenting dominant ferromagnetic interactions within the metallic cluster with high-spin S = 13/2 ground state. We believe that this approach can be extended to other magnetic molecular clusters, allowing the preparation of a diverse set of porous magnetic nanocomposites.

Acknowledgements

Financial support by the European Union under the Marie Curie International Reintegration Grant (No. 230868), the Greek Ministry of Education under “THALES” project (MIS 377064) and the University of Crete–Special Account for Research (KA 3475) is kindly acknowledged. The authors also thank P. Trikalitis (University of Crete, Heraklion) for nitrogen physisorption measurements.

References

1. R. Sessoli, D. Gatteschi, A. Caneschi and M. A. Novak, Nature, 1993, 365, 141–143.
2. R. Sessoli, H. L. Tsai, A. R. Schake, S. Y. Wang, J. B. Vincent, K. Folting, D. Gatteschi, G. Christou and D. N. Hendrickson, J. Am. Chem. Soc., 1993, 115, 1804–1816.
3. H. J. Eppley, H. L. Tsai, N. Devries, K. Folting, G. Christou and D. N. Hendrickson, J. Am. Chem. Soc., 1995, 117, 301–317.
4. M. N. Leuenberger and D. Loss, Nature, 2001, 410, 789–793.
5. L. Krusin-Elbaum, T. Shibauchi, B. Argyle, L. Gignac and D. Weller, Nature, 2001, 410, 444–446.
6. M. H. Jo, J. E. Grose, K. Baheti, M. M. Deshmukh, J. J. Sokol, E. M. Rumberger, D. N. Hendrickson, J. R. Long, H. Park and D. C. Ralph, Nano Lett., 2006, 6, 2014–2020.
7. G. Karotsis, S. J. Teat, W. Wernsdorfer, S. Piligkos, S. J. Dalgarno and E. K. Brechin, Angew. Chem., Int. Ed., 2009, 48, 8285–8288.
8. M. Evangelisti and E. K. Brechin, Dalton Trans., 2010, 39, 4672–4676.
9. M. Manoli, A. Collins, S. Parsons, A. Candini, M. Evangelisti and E. K. Brechin, J. Am. Chem. Soc., 2008, 130, 11129–11139.
10. M. Clemente-Leon, H. Soyer, E. Coronado, C. Mingotaud, C. J. Gomez-Garcia and P. Delhaes, Angew. Chem., Int. Ed., 1998, 37, 2842–2845.
11. L. Bogani, L. Cavigli, M. Gurioli, R. L. Novak, M. Mannini, A. Caneschi, F. Pineider, R. Sessoli, M. Clemente-Leon, E. Coronado, A. Cornia and D. Gatteschi, Adv. Mater., 2007, 19, 3906.
12. M. Clemente-Leon, E. Coronado, A. Forment-Aliaga, P. Amoros, J. Ramirez-Castellanos and J. M. Gonzalez-Calbet, J. Mater. Chem., 2003, 13, 3089–3095.
13. S. Willemin, G. Arrachart, L. Lecren, J. Larionova, T. Coradin, R. Clerac, T. Mallah, C. Guerin and M. Sanchez, New J. Chem., 2003, 27, 1533–1539.
14. T. Coradin, J. Larionova, A. A. Smith, G. Rogez, R. Clerac, C. Guerin, G. Blondin, R. E. P. Winpenny, C. Sanchez and T. Mallah, Adv. Mater., 2002, 14, 896–898.
15. D. Ruiz-Molina, M. Mas-Torrent, J. Gomez, A. I. Balana, N. Domingo, J. Tejada, M. T. Martinez, C. Rovira and J. Veciana, Adv. Mater., 2003, 15, 42.
16. G. E. Poirier, Chem. Rev., 1997, 97, 1117–1127.
17. S. A. Levi, P. Guatteri, F. van Veggel, G. J. Vancso, E. Dalcanale and D. N. Reinhoudt, Angew. Chem., Int. Ed., 2001, 40, 1892–1896.
18. A. Cornia, A. C. Fabretti, M. Pacchioni, L. Zobbi, D. Bonacchi, A. Caneschi, D. Gatteschi, R. Biagi, U. Del Pennino, V. De Renzi, L. Gurevich and H. S. J. Van der Zant, Angew. Chem., Int. Ed., 2003, 42, 1645–1648.
19. G. J. Sopasis, A. B. Canaj, A. Philippidis, M. Siczek, T. Lis, J. R. O'Brien, M. M. Antonakis, S. A. Pergantis and C. J. Milios, Inorg. Chem., 2012, 51, 5911–5918.
20. C. M. Yang, B. Zibrowius and F. Schuth, Chem. Commun., 2003, 1772–1773.
21. D. Y. Zhao, Q. S. Huo, J. L. Feng, B. F. Chmelka and G. D. Stucky, J. Am. Chem. Soc., 1998, 120, 6024–6036.
22. S. Brunauer, L. S. Deming, W. E. Deming and E. Teller, J. Am. Chem. Soc., 1940, 62, 1723–1732.
23. E. P. Barrett, L. G. Joyner and P. H. Halenda, J. Am. Chem. Soc., 1951, 73, 373–380.
24. M. Kruk, M. Jaroniec and A. Sayari, J. Phys. Chem. B, 1997, 101, 583–589.
25. P. Kubelka, J. Opt. Soc. Am., 1948, 38, 448.
26. S. Fiorilli, B. Camarota, D. Perrachon, M. C. Bruzzoniti, E. Garrone and B. Onida, Chem. Commun., 2009, 4402–4404.
27. S. Fiorilli, B. Onida, B. Bonelli and E. Garrone, J. Phys. Chem. B, 2005, 109, 16725–16729.
28. W. Hammond, E. Prouzet, S. D. Mahanti and T. J. Pinnavaia, Microporous Mesoporous Mater., 1999, 27, 19–25.
29. S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and Porosity, 2nd edn, Academic Press, London, UK, 1982.
30. R. C. Schroden, C. F. Blanford, B. J. Melde, B. J. S. Johnson and A. Stein, Chem. Mater., 2001, 13, 1074–1081.
31. V. Sablinskas, M. Pucetaite, J. Ceponkus and L. Kimtys, J. Mol. Struct., 2010, 976, 263–269.
32. K. Takei, R. Takahashi and T. Noguchi, J. Phys. Chem. B, 2008, 112, 6725–6731.
33. C. N. R. Rao, Ultra-violet and visible spectroscopy: chemical applications, Butterworths, 1961.
34. M. C. Marion, E. Garbowski and M. Primet, J. Chem. Soc., Faraday Trans., 1990, 86, 3027–3032.
35. L. Bonneviot, O. Legendre, M. Kermarec, D. Olivier and M. Che, J. Colloid Interface Sci., 1990, 134, 534–547.
36. A. Le Nestour, M. Gaudon, G. Villeneuve, R. Andriessen and A. Demourgues, Inorg. Chem., 2007, 46, 2645–2658.

---