Update your browser to view this website correctly. Update my browser now
We report the synthesis of monolithic porous hybrid materials from tetramethoxysilane (TMOS), the bifunctional silane bis(trimethoxysilyl)propylamine (BTMSPA), and the ionic liquid 1-butyl-3-methylimidazolium tetrachloridoferrate(III), [Bmim][FeCl4], via a one-pot reaction. The reaction products are meso- and macroporous organosilica monoliths, the iron is incorporated within the silica matrix, and the materials are efficient catalysts for H2O2 decomposition.
Ionic liquids (ILs) have attracted interest for many different reasons and they have been investigated for a multitude of applications. A rather recent development is the use of ILs as precursors for (inorganic) materials. Among others, inorganic nano- and microparticles such as CuCl, Ag, or Au have been made by using ILs as precursors, where the IL provides at least one component of the final material. Accordingly, these ILs have been termed ionic liquid precursors (ILPs) and an analogous synthesis strategy has also been developed for (doped) carbon materials, Ni and NiO nanomaterials, CuS, Fe3C, In2O3, ZnO, CuO. Other studies have focused on the synthesis using fluorine-containing anions as fluoride precursors. Alternatively, deep eutectic solvent precursors (DESPs) have also been introduced for the synthesis of such materials. As a result, proper choice of an ILP (or DESP) will lead to materials with interesting and useful properties via synthesis strategies such as carbonization or precipitation, whereby the IL is at the same time the solvent, template, and the precursor for the (inorganic) material.
So far, however, most studies concentrate on the synthesis of rather simple, often binary, compounds from ILPs. The current work focuses on the question whether [Bmim][FeCl4] can act as an ILP for the synthesis of a transition metal containing mesoporous silica material. It has also been demonstrated that ILs are efficient tools for the design of mesoporous silica. As a result, we wondered if a suitable IL may (i) catalyze the hydrolysis of appropriate silanes to form a mesoporous matrix and (ii) at the same time serve as the metal source for the formation of catalytically active metal species within the SiO2 network.
Determine if the Lewis acidic IL [Bmim][FeCl4] is an all-in-one catalyst-template-precursor for the synthesis of iron-doped mesoporous silica monoliths.
Synthesis of the monolithic materials was achieved using a published synthesis protocol. In short, the combination of tetramethoxysilane (TMOS), the bifunctional silane bistrimethoxysilylpropylamine (BTMSPA), and the ionic liquid (IL) butylmethylimidazolium tetrachloridoferrate(III), [Bmim][FeCl4], yields stable macroscopic monoliths via a one-pot reaction (Fig. 1A and B). We have prepared a library of monoliths with varying fractions of TMOS and BTMSPA as well as different amounts of [Bmim][FeCl4] present in the reaction mixture (see experimental part for details and sample nomenclature).
The monoliths are white to dark yellow and exhibit large pores. The color is likely related to the presence of some iron species but the exact nature of these iron species (see below) could not be determined. Generally, the higher the fraction of BTMSPA, the softer the materials are. The large pores appear to occur mainly at low fractions of BTMSPA. At higher fractions of BTMSPA, the (outer) surfaces of the monoliths appear much more homogeneous and smoother.
Closer inspection of the aerogels with scanning electron microscopy (SEM) reveals that all materials are macroporous and consist of a silica matrix containing smaller, roughly spherical nanoparticle like features on their surface (Fig. 1C and D). We assign the smaller features to the presence of iron-based nanoparticles that have formed during the reaction by decomposition of some of the IL. While energy dispersive X-ray spectroscopy (EDXS) shows the presence of 1–2% of iron, complementary electron paramagnetic resonance (EPR, data not shown) spectroscopy data suggest that the iron may be present as Fe3+ rather than Fe2+. As the EPR spectra are, however, only very poorly resolved, a final assignment of the iron species cannot be given.
Computer X-ray microtomography further confirms that the entire monoliths are (macro)porous and that the macropores are homogeneously distributed throughout the entire aerogel. Complementary nitrogen sorption experiments (Fig. 1H and I) show that the 100A aerogels (i.e. silica aerogels produced with 100 mg of IL) produce type IV isotherms with their characteristic H2 hysteresis loop. A type IV isotherm is typical for mesoporous adsorbents. The H2 hysteresis loop is associated with capillary condensation taking place in mesopores and may be attributed to a difference in mechanism between condensation and evaporation processes occurring in pores with narrow necks and wide cavities (often referred to as 'ink bottle' pores). However, also the role of network effects must be taken into account, e.g., network adjustment to adsorbent.
In contrast, silica aerogels made with 500 or 1000 mg of IL show type II isotherms with a slight contribution from H3 hysteresis. The reversible type II isotherm is typically obtained with a non-porous or macroporous adsorbent; it represents unrestricted monolayer-multilayer adsorption. The beginning of the almost linear middle section of the isotherm is often taken to indicate the stage at which monolayer coverage is complete and multilayer adsorption is about to begin. The H3 type loop, which does not exhibit any limiting adsorption at high p/p0, is observed with aggregates, that is, loose assemblies, of plate-like particles giving rise to slit-shaped pores. The surface areas of the 500A and 1000A samples are too small to be meaningful. They are in the range of the outer surface of a fine powder.
Table 1 (Suppl. Info.) shows the exact values of the active surface and the pore volume from BET analysis. Only the 100A samples show higher surfaces areas from 270 to 390 m2/g due to their mesoporous nature. They look and behave like amorphous silica materials with high surface areas, pore sizes between 6 and 12 nm, and porosities between 50 and 60 %. The surface area increase correlates with a decreasing amount of the organo-silica precursor BTMSPA. The mesoporous structure disappears for higher amounts of IL and the material turns into a macroporous structure with lower surface areas (<30 m2/g).
Infrared (IR) spectra (Suppl. Info.) of all samples are essentially identical and no signals from the IL are visible. All spectra are dominated by the bands of the silica matrix. A broad band around 3000–3400 cm-1 can be assigned to Si-OH and H2O vibrations. Bands at 1110, 1040, 910–940, 790, and 570 cm-1 can be attributed to the asymmetric Si-O-Si stretching vibration, Si-OH vibration, and the symmetric Si-O-Si stretching vibration and rocking modes, respectively. Bands around 1440–1480 cm-1 can be assigned to the scissor vibration of the CH2 groups of the precursor BTMSPA pretty much as the signals at 2800 and 2950 cm-1 which can be attributed to N-H and C-H stretching vibrations of BTMSPA.
The signal intensity depends on the ratio of the two precursors TMOS and BTMSPA in the preparation process. Lower fractions of BMTSPA decrease the intensities of the bands at 1440, 2800, and 2950 cm-1 but increases the intensity of the band at 570 cm-1 consistent with the above assignments.
Thermogravimetric analysis (TGA, data not shown) of all samples reveals a multistep decomposition process of all the hybrid materials. A first weight loss of 3–5% can be assigned to the desorption of surface-adsorbed water molecules. Up to 300°C the system loses just about 10% of its mass. This mass loss is assigned to condensation of remaining silanol groups. Between 300 and 600°C the most significant mass loss due to decomposition of the organic linkers and further silica condensation is observed. Above 600°C the mass of the materials is essentially constant. Generally, higher fractions of the organosilane precursor BTMSPA results in a higher mass losses from the material.
As the materials, especially the 100A materials, combine a set of interesting properties, namely (1) mesoporosity, (2) resonable surface areas, (3) up to ca. 2% of an iron species in the form of nanoparticles, and (4) macroscopic integrity and mechanical stability, the question of catalytic activity of these materials arose during the course of the project. To evaluate the catalytic activity of some of the materials, we chose the catalytic decomposition of hydrogen peroxide. Hydrogen peroxide is thermodynamically unstable and decomposes to water and oxygen at room temperature. Many transition metals catalyze the decomposition and especially Fe2+ can lead to a decomposition pathway including radical species like HO. and HOO..
To evaluate the effect of the aerogels, we prepared a sealed reaction vessel containing 15 mL of a 10% aqueous hydrogen peroxide solution and recorded the percentage of oxygen in relation to the air volume inside the vessel (Fig. 1L). The control measurements done without the addition of one of the aerogels show a roughly linear increase of the oxygen concentration in the gas phase above the liquid. This is due to the well known decomposition of H2O2 even without any catalyst being present. The same experiment was then done with adding the catalyst to the solution after 6 min. Clearly, addition of the catalyst results in a rapid oxygen release. However, after ca. 20 min the catalyst seems to become inactive and the slope in the oxygen evolution curves of both reactions, with and without catalyst, are quite similar again.
The IL [Bmim][FeCl4] is a viable all-in-one catalyst-template-precursor for the synthesis of iron-doped mesoporous silica monoliths. Moreover, the fact that up to 2% of iron is firmly incorporated into the silica aerogel matrix suggests applications in catalysis.
Currently, there are two main limitations to these materials: (1) the nature of the iron species is unclear and could not be resolved so far, including analysis with EPR spectroscopy, X-ray diffraction, or (high resolution) transmission electron microscopy. Moreover (2), the catalyst appears to be active for only ca. 20 min; pathways to catalyst regeneration and the exact reason for deactivation are currently unclear. This is closely related to the fact that the exact nature of the iron species is unknown.
Tetramethyl orthosilicate (TMOS) was purchased from Acros Organics, Bis(trimethoxysilylpropyl)amine (BTMSPA) from ABCR, 1-Butyl-3-methylimidazolium chloride, [Bmim][Cl] from Iolitec, FeCl*6H2O from Roanal, acetone from VWR, and n-propanol from Roth. All chemicals were used as received without further purification.
Preparation of iron-containing ionic liquid [Bmim][FeCl4]
1-Butyl-3-methylimidazolium tetrachloridoferrate(III) was synthesized in a one-step reaction. For this purpose, [Bmim][Cl] (18.5 mmol, 3.231 g) and FeCl3*6H2O (18.5 mmol, 5 g) were separately dissolved in 50 mL of n-propanol and then the two solutions were mixed. Subsequently, the solvent was removed via azeotropic distillation to obtain the dark brown liquid [Bmim][FeCl4]. The IL was freeze-dried overnight to remove water and was stored in a Schlenck flask under argon gas until use.
Preparation of iron-containing silica materials
Silica monoliths were synthesized via sol-gel reaction using the co-precursor BTMSPA as a basic catalyst following a published protocol. The total silicon concentration in the solution was 16.5 mmol from TMOS and BTMSPA combined. Notice that BTMSPA contributes 2 moles of silicon for every mole of silane. The amount of IL was 100 mg, 500 mg, or 1000 mg, respectively. For every reaction, the amount of water was 41.5 mmol (0.743 mL) and the amount of acetone was 100 mmol (5.4 mL).
In a typical experiment TMOS (0.244 mL), BTMSPA (1.127 mL), acetone (5.4 mL), and [Bmim][FeCl4] (1.0 g) were mixed in a 50 mL plastic tube. All reactants are liquid and completely miscible at room temperature. After shaking, the clear, brown liquid mixture was poured into a polypropylene mold, sealed with a septum and covered with parafilm. After 10 min gelation occurred and an off-white to yellow, solid silica monolith formed. The resulting mixture of brown IL and yellow sponge-like silica was allowed to sit without agitation for 24 h for further condensation of the silica. The resulting material was washed extensively by soxhlet extraction with methanol and dried under vacuum producing macroscopic aerogels with an off-white to light brown color.
The resulting aerogels (A) are denoted as xAy:z were x = 100, 500, 1000 (amount of IL in mg used in the synthesis), y:z = 1:9, 2:8, 3:7, 4:6, 5:5, 6:4, and 7:3 (molar ratio of the silica precursors TMOS:BTMSPA).
Infrared (IR) spectroscopy was done in ATR mode on a Thermo Nicolet FT-IR Nexus 470 with ATR equipment. Spectra were taken from 400 to 4000 cm-1 with a resolution of 2 cm-1 and 32 scans.
TGA was done on a TGA 4000 thermal analyzer in air from 30 to 900°C with a heating rate of 10°C.min-1. The samples were placed in ceramic crucibles.
Nitrogen adsorption was done on a NOVA-4000e instrument within a partial pressure range of 10-6 to 1.0. Before measurements, the samples were degassed at 120°C for 2 h. The Brunauer-Emmett-Teller (BET) surface areas and pore volumes of each sample were measured by N2 adsorption experiments performed at -196°C (77 K). The pore volume was taken by a single point method at p/p0≈0.99.
Scanning electron microscopy (SEM) was done on a Phenom desktop electron microscope (FEI, Eindhoven, Netherlands) operated at 5 kV and on a JEOL JSM-6510 (JEOL GmbH, Eching, Germany) operated at 15 kV with a tungsten filament. EDXS measurements were done with an Oxford Instrument INCAx-act X-ray detector mounted on the JEOL SEM. Prior to measurements the samples were coated with a 100 nm carbon layer using a POLARON CC7650 Carbon Coater.
Micro X-ray computed tomography
All scans were obtained with a Skyscan 1172 high resolution micro CT at 59 kV and 165 μA, without filter, and set at 5 μm image pixel size, 360° rotation, 3 frames averaging, a rotation step of 0.02° and ring artefact correction set at 20, beam hardening correction set to 55%. The reconstruction software (NRecon v.22.214.171.124, SkyScan N.V.) was used to reconstruct cross-section images from tomography projection images. Quantitative parameters were measured by using the microCT analysis software (CTAn., v. 126.96.36.199+, SkyScan N.V.), normalization and histogram were done in Excel.
O2 Evolution measurements
Measurements of the oxygen content evolved over the course of the H2O2 decomposition were done with a fiber-optical method following a protocol published eleswhere.
The University of Potsdam and the DFG projects TA571/2-1, TA571/3-1, and TA571/13-1 are thanked for financial support.
We thank Dr. C. Günter for access to the EDX instrument, and Dr. A. Winter and Prof. P. Strauch for EPR measurements.