Sol–gel synthesis and characterisation of nanoscopic strontium fluoride

Abstract

Strontium fluoride nanoparticles have been successfully synthesized via the fluorolytic sol–gel synthesis. The investigation by DLS, XRD, TEM and N₂ adsorption/desorption measurements (BET) revealed the existence of sol particles with a hydrodynamic diameter of approximately 6 nm, crystallite sizes below 10 nm and high surface area of 180 m² g⁻¹. The sols are highly transparent and stable for a few months, however, with time all sols tend to undergo gelation. The ageing process of these sols was investigated using WAXS. As a result, Ostwald ripening as a reason for gel formation can be excluded. By following the reaction progress fluorine-containing crystalline species were detected by XRD and solid state ¹⁹F MAS NMR spectroscopy indicating the formation of an intermediate phase during the fluorolysis reaction.

---

1 Introduction

Nanoscopic metal fluorides are characterised by extraordinary properties, e.g. high surface area and high reactivity, and are therefore nowadays of great interest in different application fields. Particularly, they exhibit exceptional properties which allow wide-ranging applicability in fields such as optics, catalysis and materials science. Due to their low refractive index, e.g. MgF₂ (n_D = 1.38) and AlF₃ (n_D = 1.35), they are potential antireflective optical coating materials as well as inorganic components embedded in an organic polymer matrix thus forming composite materials with new and innovative properties.¹ The high surface area and high Lewis acidity of e.g. HS-AlF₃^2–4 and MgF₂^5,6 allow them to perform as heterogeneous catalysts in a number of different reactions which have been reported recently.⁷ As a result of their increasing and targeted use in the optical industry, especially fluorides of alkaline earth metals gain more and more importance in current research.

Recently our group has established the fluorolytic sol–gel synthesis for the preparation of nanoscopic metal fluorides.⁸ Nanoscopic MF_n obtained via the sol–gel synthesis were thoroughly investigated in previous studies in terms of formation, stabilisation of nanoparticles and sols as well as the ageing process of sols.⁹ Although, many synthetic routes to nanoscopic alkaline earth metal fluorides (cubic structure) such as CaF₂ and BaF₂ already exist,^10–12 a major advantage of the sol–gel route is the facile access to nanoscopic particles dispersed in an organic solvent functioning as a mild, efficient and versatile method for preparation and handling.

Because of its excellent optical properties e.g. their low refractive index (n_D = 1.439 @ 589 nm), the high transmission in the infrared and ultraviolet spectral range (0.13–11 μm) and the extremely low solubility in water (120 mg L⁻¹) strontium fluoride is an ideal material for optical applications and is often used as a coating material for high-quality optical windows, prisms and lenses to reduce reflection and to increase transmission. Furthermore, it can be used as inorganic filler in dental composites and therefore can improve the mechanical and chemical properties of dental materials acting at the same time as a fluoride-releasing material, which is of great interest in caries prevention.^13,14 The desired effects are often increased by addition of nanoscopic metal fluorides.

In this study we report for the first time the sol–gel synthesis and overall characterisation of fundamental properties of nanoscopic SrF₂ starting from strontium acetate hemihydrate as precursor. The existence of SrF₂ nanoparticles within the highly transparent sol as well as in bulk material was proven by various analytical methods like dynamic light scattering (DLS) and X-ray powder diffraction (XRD). Utilising spectroscopic and crystallographic methods (solid state NMR, XRD and wide angle X-ray scattering WAXS) allowed us to gain comprehensive chemical and structural information during the fluorolysis process of strontium acetate hemihydrate while the reaction with HF proceeded stepwise. To gain a better understanding of the gelation and ageing process, WAXS measurements as well as NMR experiments were performed on SrF₂ sols over a period of several weeks and months. The results are presented and discussed in the following.

2 Experimental part

2.1. Synthesis of SrF₂ sols and xerogels

Strontium acetate hemihydrate (Aldrich) was dissolved in acetic acid (Carl Roth, >99.8%) under gentle warming, resulting in a clear solution. Ethanol (Carl Roth, 99.8% with 1% MEK) was added slowly to the solution giving a white precipitate of Sr(CH₃COO)₂. The ratio of HOAc/EtOH in the mixture was 1 : 3 v/v. A stoichiometric amount of HF dissolved in ethanol (14.6 M) was added to the mixture. After stirring (up to 2 days), transparent SrF₂ sols were obtained. The xerogel was obtained by removing the solvent and the resulting powder was finally dried under vacuum at temperatures up to 90 °C.

2.2. Characterisation

The X-ray powder patterns were recorded at room temperature on a XRD 3003 TT from Seiffarth using a CuKα radiation (Wavelength. CuKα1 = 1.54056 Å).

The viscosity was measured with the Anton Paar Falling ball viscometer. For determining the dynamic viscosity, glass capillaries with an inner diameter of 1.6 mm and 1.8 mm were used. The viscosity was detected at a temperature of 25 °C.

The measurements of the hydrodynamic particle diameter using DLS were carried out with a Zetasizer Nano from Malvern. Samples (concentration = 0.2 mol L⁻¹) were analysed in the Malvern supplied “size” operating procedure, the light being detected at an angle of 173° and at a temperature of 25 °C.

The high-resolution transmission electron microscopy (HRTEM) measurements were performed on a JEOL TEM/STEM 2200 FS, with an acceleration voltage of 200 kV and a field emission source. Carbon coated 300 mesh copper grids were used as sample carrier. The TEM grid was dip-coated into a 0.2 M SrF₂ sol (one week old) and dried at 50 °C for 3 h.

The C and H content were determined with a Leco CHNS-932 analyser. The thermal behaviour was studied by simultaneously coupled TA-MS measurements. A NETZSCH thermoanalyzer STA 409 C Skimmer® system, equipped with a BALZERS QMG 421, was used to record the thermoanalytical curves (T, DTA, TG, DTG) together with the ionic current (IC) curves in the multiple ion detection (MID) mode. Further experimental details were as follows: DTA-TG sample carrier system; Pt/PtRh10 thermocouples; platinum crucibles (beaker, 0.8 mL); sample mass 56 mg (measured versus empty reference crucible); constant purge gas flow of 70 mL min⁻¹ nitrogen 5.0 (AIRLIQUIDE); constant heating rate 10 K min⁻¹; raw data evaluation with manufacturer's software PROTEUS® (v. 4.3) and QUADSTAR® 422 (v. 6.02) without further data treatment, e.g. such as smoothing.

For WAXS experiments an acoustic levitator (tec5 AG, Oberursel, Germany) was used as sample holder. Applying this contact-free method, solid and liquid samples can be positioned in a gaseous environment by means of a stationary ultrasonic field. In a typical experiment, a droplet with a volume of about 4 μL was manually injected into the acoustic levitator with a common Eppendorf pipette (size 0.5–10 μL, Eppendorf, Germany). The sample remains in a fixed position during the measurement even after evaporation of the solvent. The position stability of the droplet, measured as a displacement smaller than 20 μm, allowed more than 30 minutes of data acquisition time.

The in situ X-ray diffraction experiments were performed at the synchrotron micro focus beamline μSpot (BESSY II of the Helmholtz Centre Berlin for Materials and Energy). Providing a divergence of less than 1 mrad (horizontally and vertically), the focusing scheme of the beamline is designed to provide a beam diameter of μm at a photon flux of 1 × 10⁹ s⁻¹ at a ring current of 100 mA. The experiments were carried out employing a wavelength of 1.03358 Å using a double crystal monochromator (Si 111). Scattered intensities were collected 200 mm behind the sample position with a two-dimensional X-ray detector (MarMosaic, CCD 3072 × 3072 pixel with a point spread function width of about 100 μm). The obtained scattering images were processed and converted into diagrams of scattered intensities versus scattering vector q (q is defined in terms of the scattering angle θ and the wavelength λ of the radiation, thus q = 4λ/π sin θ) employing an algorithm from the computer program FIT2D.

X-ray powder diffractometry (XRD) was carried out using a diffractometer D8 DISCOVER (Bruker-AXS, Karlsruhe, Germany) with monochromatic CuKα1-radiation and a Johansson monochromator in the incident beam. Specimens were prepared by filling the white powder into a glass capillary (diameter 0.5 mm). The diffraction patterns were collected using the following conditions: 2θ range: 5–60°, step size: 0.009°, step time: 2 s.

N₂ adsorption/desorption analysis was performed at 77 K using an ASAP 2020 Micromeritics instrument. The specific surface area was determined according the BET method.

The ¹H Liquid-NMR spectra were recorded on a Brucker AVANCE II 400. All experiments were carried out using standard conditions for NMR parameters. The spectra were obtained at 400 MHz using a C₆D₆ capillary as lock standard.

A Bruker AVANCE 400 spectrometer (Larmor frequency: ν_19F = 376.4 MHz) equipped with a 2.5 mm MAS probe and with 2.5 mm rotors made of ZrO₂ was used to record ¹⁹F MAS NMR spectra. The spectra were recorded with a π/2 pulse duration of 4.0 μs and a spectrum width of 400 kHz. Recycle delays between 5 s and 480 s were applied to test the spin lattice relaxation behaviour of different fluorine species. The spinning frequencies are given in the caption of the figures. The isotropic chemical shifts δ_iso of ¹⁹F resonances are given with respect to the CFCl₃ standard. A phase-cycled depth pulse sequence according to Cory et al.¹⁵ was taken to suppress existent background signals of ¹⁹F.

3 Results and discussion

3.1. Fluorolytic sol–gel synthesis

The fluorolytic sol–gel synthesis of nanoscopic metal fluorides based on the reaction of metal alkoxides as starting materials with a stoichiometric amount of HF gives transparent metal fluoride sols as shown before.^2,8

However, the formation of nanoscopic strontium fluoride sols via the fluorolytic sol–gel synthesis was successfully performed using the commercially available strontium acetate hemihydrate as precursor.

For the preparation of strontium fluoride sols, strontium acetate hemihydrate is first molecularly dissolved in acetic acid, followed by adding ethanol to the solution, wherein strontium acetate precipitates. To this suspension a stoichiometric amount of HF was added. After 1–2 days, a highly transparent sol is formed having a low viscosity of 1–2 mPa s depending on the concentration. The highest concentration achieved by this route is 0.8 mol L⁻¹.

3.2. Characterisation of nano-SrF₂

The hydrodynamic diameters of SrF₂ sol particles were measured via dynamic light scattering (DLS). Fig. 1 presents the size distribution by intensity of a 0.2 M SrF₂ sol. The distribution of three size classes of sol particles with hydrodynamic diameters around 6 nm, 21 nm and 164 nm was found. The area of the peak for particles around 164 nm is several times larger than the peak for the 6 nm fraction (1 : 1 000 000 ratio). It is known from Rayleigh's approximation that the intensity of scattering of a particle is proportional to the sixth power of its diameter (I = d⁶) because large particles scatter much more light than small particles.^22,23 Based on this, it can be approximated that the number of smaller particles with a hydrodynamic diameter of 6 nm in the SrF₂ sol is much larger than that of the larger particles.

Fig. 1 Intensity size distribution for a SrF₂ sol (c = 0.2 mol L⁻¹) studied by DLS.

Nevertheless, a small amount of agglomerates or larger particles exists in the sol, which overestimates the non-corrected size distribution curve. It has to be taken into account that by DLS only the hydrodynamic diameter is measured but not the particle size. When suspended particles are moving in a liquid medium, a thin electrical dipole layer is formed on its surface. As a result, the measured hydrodynamic particle diameter in the dynamic light scattering (DLS) is usually larger than the diameter measured via transmission electron microscopy (TEM).

TEM was used to gain further information such as particle size, morphology, and crystallinity of nanoparticles. Fig. 2 shows a high-resolution TEM (HRTEM) image of SrF₂ nanoparticles. The image shows typical degrees of agglomeration of several primary particles and polydispersity obtained as a result of the drying process. Furthermore, the HRTEM image shows crystal lattice planes of nano-SrF₂ proving the crystallinity of the particles. The measured lattice spacing of 0.334 nm is consistent with the 111 lattice planes of cubic SrF₂. The contrast patterns of atomic columns of the central particle differ from patterns in other regions. A special orientation of atomic columns relative to the incident electron beam result in this special pattern. In this case, the atomic columns viewed along a 110 direction show bright lines which can be attributed to the 111 and the 200 planes of SrF₂. The lattice plane distances of 0.334 nm and 0.288 measured in HRTEM image of a single particle (Fig. 2, middle) are consistent with the SrF₂ 111 and 200 lattice plane distances. The particle size can be estimated to be in a range of about 5 nm. The determined BET surface area of the SrF₂ xerogel of S_g = 180 m² g⁻¹ is a further hint for very small particles.

Fig. 2 High-resolution TEM image of SrF₂ nanoparticles obtained via sol–gel synthesis. The TEM grid was dip-coated into a 1 week old 0.2 M SrF₂ sol and dried at 50 °C for 3 h.

After evaporation of the solvent in the sol a powder-like xerogel of SrF₂ was obtained and characterised by X-ray powder diffraction. The recorded diffraction patterns can be assigned to the crystallographic planes of pure SrF₂ which crystallises in the well-known cubic fluorite structure (Fig. 3). Due to a small crystallite size of SrF₂ nanoparticles a significant diffraction peak broadening and the corresponding low intensity of the reflections are observed. In particular, the peak broadening, being due to the nanoparticles' size, is inversely proportional to nanoparticle size.

Fig. 3 X-ray diffraction (XRD) patterns of nanoscopic SrF₂ xerogel obtained by the fluorolytic sol–gel synthesis.

According to the Scherrer equation where K = 0.89 (shape factor in the case of spherical nanoparticles, as approximated here), λ = 1.54056 Å (X-ray wavelength for CuKα1) and Δ(2θ) is line broadening at half the maximum intensity (FWHM, in radians),¹⁶ the average crystallite diameter of the nanoparticles, evaluated from the most distinct reflections broadening 111, 220 and 311 was approximated to be below 10 nm for the samples obtained via the sol–gel synthesis. These results are in good agreement with DLS experiments in Fig. 1 and TEM measurements (Fig. 2), respectively.

The C, H, N elemental analysis results of the xerogel reveal a residual carbon content of about 3%. We also investigated the thermal behaviour of SrF₂ xerogel by DTA-TG-MS measurements. The DTA-TG curves coupled with MS profiles are shown in Fig. 4. The TG curve of SrF₂ shows mass losses in two consecutive steps between 100 and 550 °C. The first step up to 300 °C, involving a mass loss of 11.2%, is mainly ascribed to the release of water and the solvent adhering to the particle surface. The last step occurs up to 500 °C with a mass loss of 5.4% and is due to the decomposition of the remaining acetate groups. The DTA curve shows no thermal effects thus being characteristic for samples obtained from the fluorolytic sol–gel route. After the thermal analysis the recorded XRD patterns show very narrow reflections of SrF₂.

Fig. 4 DTA and TG curves obtained by thermal analysis of SrF₂ xerogel with ion current curves for m/z18 (H₂O), m/z43 (CH₃O⁺), m/z44 (CO₂), m/z58 (CH₃C(O)CH₃), m/z60 (CH₃COOH).

3.3. Ageing of SrF₂ sols

Of great interest is the ageing process of the sols. Although a transparent SrF₂ sol with low viscosity was obtained after max. 2 days, after 6 months the same sol became gelous. Wide angle X-ray diffraction investigations (WAXS) were performed in order to monitor changes in the crystalline structure and the crystallite size of the sol particles over a period of 6 months. Fig. 5 shows WAXS patterns of 0.2 M SrF₂ sols at different ageing times of 3 days to 6 months after HF addition. In all cases very broad reflections are observed which can be clearly assigned to the fluorite crystal structure of SrF₂. The difference in crystallite sizes can be distinguished by the decrease in peak broadening with ageing time. 3 days after HF addition very broad reflections and low intensities of the peaks 111, 220 and 311, which are the most distinct, are observed indicating the formation of nano-SrF₂. The crystallite size can be estimated according the Scherrer equation, and can be approximated to be below 5 nm. Over the following 3 months the reflection form does not change noticeably. With increasing ageing time, the reflections increase in intensity and two other reflections of 200 and 222 appear in the diffractogram. At the same time a slightly decrease in the reflection widths is observed indicating on the one hand the increase in the crystallite size, on the other hand, might be more probably a result of a higher degree of crystallinity with increasing ageing time. However, according to the Scherrer equation the average crystallite size of the SrF₂ sol particle obtained 6 months after HF addition was approximated to be 6 nm. Hence, this indicates that obviously no change in crystallite size of the SrF₂ sol particles occurs during the first 3 months after HF addition. After 6 months a slight increase in crystallite size is observed meanwhile the sol was observed to undergo a sol–gel transition. Since the Scherrer equation is not expected to be valid for crystallite sizes <10 nm, the calculated values must be considered as an estimate.

Fig. 5 WAXS diffractograms of 0.2 M SrF₂ sols at different ageing times of 3 days to 6 months after HF addition.

In general it is assumed that the polymerization of an inorganic network e.g. during the sol–gel process of silica proceeds as follows: (i) formation of small inorganic (primary) particles (sol), (ii) particle growth followed by a subsequent linking of the large particles forming an inorganic network (gelation).¹⁷ Usually the particle growth proceeds via Oswald ripening.^18,19 Due to their higher solubility and surface energy the smallest particles dissolve and re-crystallize in favour of larger particles. As a result, a significant decrease in reflection width would be detected. Since this is not observed, Oswald ripening is less probable to be the dominating process in the gelation of the SrF₂ sols. Hence, other mechanisms have to be considered being relevant during gelation of SrF₂ nanoparticles. From polymers and supramolecular chemistry it is known, that hydrogen bonding and van der Waals interactions due to polar solvents are expected to be the key factor of gelation.²⁰ In our system, gelation has to be considered due to strong hydrogen bonding since the SrF₂ sol particles are obtained in polar protic solvents such as acetic acid and ethanol. Additionally, the SrF₂ particles are highly distorted and coordinatively unsaturated, thus providing polar surface sites for interactions with solvent molecules. Moreover, enhanced Lewis acidity due to huge amount of coordinatively unsaturated surface sites in nano metal fluorides have been evidenced e.g. for AlF₃ and MgF₂, respectively,^2,6 indicating the high ability of these materials to interact with electron donating solvent molecules. In order to investigate the influence of the organic solvent within the sol–gel process of SrF₂¹H NMR spectra were recorded over the period of several weeks and months corresponding to the changes in viscosity. For simplification, only an extract of the ¹H NMR spectra of a SrF₂ sol is shown in Fig. 6. Shortly after the fluorolysis, only acetic acid and ethanol can be observed in the spectra. Because of the rapid proton exchange between ethanol and acetic acid molecules, the NMR spectrum shows just a single signal at an average chemical shift δ = 7.53 ppm. With increasing time, this signal shifts slightly towards high-field indicating the progressive esterification of acetic acid with ethanol to ethyl acetate and water. At the same time as this signal shifts towards high-field, an increased signal intensity for ethyl acetate can be observed. With increasing time the average signal shifts further to high-field, since increasingly more water is formed due to progressive esterification (cf. δ (H₂O) = 4.9 ppm). All SrF₂ gels appeared highly transparent and were stable for several months.

Fig. 6 ¹H NMR spectra of a 0.2 M SrF₂ sol taken at different ageing times.

3.4. The reaction progress of SrF₂

To get a deeper insight into the reaction progress during the formation of SrF₂ sol particles and to gain more structural information on the SrF₂-crystallites formed via the fluorolysis of strontium acetate hemihydrate, X-ray powder diffractograms were recorded for samples with increasing F stoichiometry. Fig. 7 displays powder diffraction patterns for samples with different F : Sr ratios. For a F : Sr ratio of 0.3 only very intense and narrow reflections are observed. These can be clearly assigned to Sr(OAc)₂·½H₂O. The precursor crystallises in the triclinic structure as distrontium tetra-acetate monohydrate. Upon further increase of the F : Sr ratio the reflections for Sr(OAc)₂·½H₂O disappear and other narrow reflections at different 2θ values occur. Samples with an HF amount of 1.0 eq. only show reflections that can be assigned neither to Sr(OAc)₂·½H₂O nor to SrF₂ indicating the formation of a new crystalline intermediate species. Since for samples with a F : Sr ratio of 1.3 and above the XRD reflections shift to the expected values; the position and broad peak shape indicate the formation of nanocrystalline SrF₂ particles.

Fig. 7 XRD patterns of samples with varying F : Sr ratio. The values on the right represent the HF equivalents in the sample.

Solid state NMR spectroscopy is well suited to study chemical environment and the chemical composition in the coordination sphere of solid compounds. In order to gain further chemical information during the formation of SrF₂¹⁹F MAS NMR spectra were recorded perfectly complementing the XRD data (Fig. 8).

Fig. 8 Solid state ¹⁹F MAS NMR spectra of samples with different stoichiometric F : Sr ratio recorded at 25 kHz. The values on the left represent the HF equivalents in the sample.

The spectra of samples with a F : Sr ratio below 1 : 1 exhibit a signal with a narrow line width of approx. 450 Hz at a chemical shift of −79 ppm indicating the formation of a fluorine-containing crystalline species as the only fluorination product. With increasing HF amount the intensity of this peak decreases while another two broad and slightly overlapping signals at −83 and −88 ppm appear in the spectra. Adding the stoichiometric amount of HF for adjusting a F : Sr ratio of 2 : 1 results in the appearance of a very broad peak with a chemical shift of −88 ppm that can be clearly assigned to fluoride in the cubic structure of SrF₂. Both, the large peak width of approximately 1.5 kHz and the short spin-lattice relaxation time T₁ = 5 s, indicate a highly disturbed structure with a very low degree of crystallinity. Upon closer examination of the signal shape, a slight asymmetry and broadening of the signal is observed. Additionally the signal has a shoulder that is shifted low-field by a quite large amount. As shown in previous work,²¹ this asymmetry and broadening of the peak shape towards low-field is characteristic of fluorine, in this case in a cubic structure of SrF₂, with a varying chemical environment of oxygen and fluorine in the second coordination sphere.

3.5. Discussion

For the first time the formation of nanoscopic strontium fluoride sols via the fluorolytic sol–gel synthesis was successfully performed using commercially available strontium acetate hemihydrate as precursor. Thus, stable sols were obtained with a concentration up to 0.8 mol L⁻¹. Since the sol appears transparent and shows low viscosity the particle size can be approximated to be in the low nanometer range. The particle size distribution determined by DLS shows the existence of three particle size classes. Since the overall size distribution obtained from DLS is intensity based, the smallest particles with a hydrodynamic diameter of 6 nm dominate the particle population, meaning particles with larger hydrodynamic diameters around 21 and 164 nm are overestimated and probably represent the minority of either larger particles or agglomerates of smaller particles.

The XRD patterns provided direct evidence that the reaction of Sr(OAc)₂ in an acetic acid–ethanol mixture with a stoichiometric amount of HF leads to the formation of SrF₂ nanoparticles with crystallite sizes below 5 nm as determined via Scherrer equation. As the sol appears transparent only after 1 or 2 days of stirring one can assume that the formation of nanosized particles is a process which might not only be due to deagglomeration process of large particles leading to smaller particles but, compared to previously published results,⁹ also due to the formation of intermediate species during the conversion of Sr(OAc)₂ to SrF₂.

A comparison between XRD data and ¹⁹F MAS NMR results provide a deeper insight in the reaction process of the fluorolysis of strontium acetate hemihydrate yielding nanoscopic SrF₂ particles. Due to the extremely low solubility of Sr(OAc)₂ in ethanol the starting material for the synthesis appears as a very turbid suspension. The preparation of samples with a F : Sr ratio below 0.9 leads to turbid suspensions of Sr(OAc)₂ containing a further – so far unknown – crystalline species. The existence of both can be shown by XRD data. With further fluorination (1.0 eq. HF) the sols turn to slightly opaque resulting, after just 1 or 2 days, in gelation while sedimentation can be observed in the following. The addition of 1.0 to 1.1 eq. HF to the strontium acetate suspension leads to the formation of a highly crystalline fluorine-containing species which can be detected by ¹⁹F MAS NMR. Since no reflections neither of Sr(OAc)₂ nor SrF₂ appear in the XRD diffraction patterns, it is likely that Sr(OAc)₂ reacts virtually completely to yield a F-containing intermediate phase whereas SrF₂ particles are not formed at this stage. Thus, the so far unknown species is the main fluorination product within the stoichiometric ratio F : Sr of approx. 1 : 1. With further addition of HF the reaction system turns into forming transparent sols, thus indicating the formation and existence of nano-sized particles. Upon a F : Sr ratio of 1.3 only very broad reflections with low intensities at the expected 2θ values for SrF₂ can be observed in the powder diffraction patterns. Additionally, ¹⁹F MAS NMR spectra evidence the formation of SrF₂ particles which starts at a stoichiometric F : Sr ratio of about 1.3. Since the signals show an asymmetric decay and a shoulder towards low-field as well as a very broad line width, the existence of highly disordered SrF₂ particles in the lower nanometer range with a varying but low amount of oxygen in the second coordination sphere can be assumed.

At a certain but constant concentration and temperature the SrF₂ sols were stable for several months but then turned to form gels. With lower concentration gelation time is postponed. Although the gelation of SrF₂ sols proceeds several months after SrF₂ formation, X-ray patterns of SrF₂ sols at different ageing time do not show any significant change in crystallite size according to Scherrer equation. Hence, one can assume that SrF₂ nanoparticles, alike MgF₂,⁹ do not undergo Ostwald ripening, as described e.g. for the sol–gel process of SiO₂.¹⁷ Here, gelation seems to be rather a phenomenon caused by additional water formed in the esterification reaction of acetic acid with ethanol during the ageing of SrF₂ sols. Therefore, the reaction progress of water and ethyl acetate was monitored by ¹H NMR spectroscopy until the ester equilibrium had been reached after approx. 5 months. As additional amount of water is formed with increasing aging time, obviously this formed water enforces the formation of hydrogen bonds between particles and solvent molecules. Since the esterification reaction takes several months to approach the equilibrium even the sol gelation is an equally slow process. Obviously, the quantity of water in the sol–gel reaction strongly influences the kinetics of gel formation. The sol particles seem to be stable towards a small amount of water, above a crucial amount of water the sol–gel transition of SrF₂ sols becomes relevant. Additionally, to verify this hypothesis, a few amounts of water were added to the stable SrF₂ sol of low viscosity resulting in gelation that occurred within a few hours, thus evidencing the role of water as major cause of gel formation. Since the size of SrF₂ nanoparticles is in the low nanometer range, the high number of coordinatively unsaturated sites at the surface of the nanoparticles can lead to strong hydrogen bonds between SrF₂ surfaces and solvent molecules. Since both, the sol and the gel are transparent, water, and therefore the hydrogen bonds in the sol, are supposed to cause a soft agglomeration of the particles into larger units, which results in gelation rather than in sedimentation. These soft agglomerates are likely to form a weak integrated network of particles held together by weak surface forces such as van der Waals interactions and hydrogen bonding. Therefore, we assume the gel to be a network with a rather irregular structure of small particles and soft agglomerates. An agglomeration of primary particles into very large units (hard agglomerates) due to strong chemical bonds between the particle surfaces will lead to a reduction in the concentration of the SrF₂ sol accompanied with a decrease in viscosity. Hence, hard agglomerates will more probably result in sedimentation rather than in gelation.

4 Conclusion

In this study we were able to develop for the first time a synthesis route for nanoscopic SrF₂ forming transparent sols. Stable SrF₂ sols and gels with high concentrations were obtained via the fluorolytic sol–gel synthesis using strontium acetate hemihydrate as precursor and non-aqueous HF solution as fluorinating agent. Both, the hydrodynamic particle diameter as well as crystallite sizes of such SrF₂ nanoparticles, determined via DLS, TEM and XRD, were found to be below 10 nm. Furthermore, utilising solid state MAS NMR experiments and WAXS measurements the formation of a crystalline fluorine-containing species as an intermediate phase during the fluorolysis reaction at F : Sr ratio of 1 : 1 was evidenced. It was shown that the sol–gel transition is not caused by Ostwald ripening, but is rather due to additional water that is formed as a result of a side reaction in the applied solvent system, that is the esterification reaction between ethanol and acetic acid formed from the strontium precursor. One can therefore conclude that the solvents used, based on their properties (e.g. polarity, hydrogen donor and acceptor ability) and their side reactions (e.g. esterification) are important factors that determine the sol–gel reaction process as well as the long-time behaviour of SrF₂ sols.

Based on straightforward access to nano-SrF₂ sols as provided by this work, further activities are in progress to use this new material in different fields of applications like medicine and hybrid polymers.

Acknowledgements

The authors like to thank Gudrun Scholz and Detlef Heidemann for performing the solid state NMR measurements and Michael Feist for the thermal analysis. L. Schmidt is a member of the graduate school GRK 1582 “Fluorine as a Key Element” of DFG (Deutsche Forschungsgemeinschaft).

Notes and references

1. J. Noack, L. Schmidt, H. G. Gläsel, M. Bauer and E. Kemnitz, Nanoscale, 2011, 3, 4774–4779.
2. E. Kemnitz, U. Groß, St. Rüdiger and S. C. Shakar, Angew. Chem., Int. Ed., 2003, 42, 4251–4254.
3. St. Rüdiger, U. Groß, M. Feist, H. A. Prescott, S. C. Shakar, S. I. Troyanov and E. Kemnitz, J. Mater. Chem., 2005, 15, 588–597.
4. St. Rüdiger, U. Groß and E. Kemnitz, J. Fluorine Chem., 2007, 128, 353–368.
5. M. Nickkho-Amiry and J. M. Winfield, J. Fluorine Chem., 2007, 128, 344–352.
6. M. Nickkho-Amiry, G. Eltanany, St. Wuttke, St. Rüdiger, E. Kemnitz and J. M. Winfield, J. Fluorine Chem., 2008, 129, 366–375.
7. E. Kemnitz, S. Wuttke and S. M. Coman, Eur. J. Inorg. Chem., 2011, 4773–4794.
8. St. Rüdiger and E. Kemnitz, Dalton Trans., 2008, 1117–1127.
9. J. Noack, F. Emmerling, H. Kirmse and E. Kemnitz, J. Mater. Chem., 2011, 21, 15015–15021.
10. R. N. Grass and W. J. Stark, Chem. Commun., 2005, 1767–1769.
11. L. Sun, L. C. Chow, J. E. Bonevich, T. Wang and J. W. Mitchell, J. Biomed. Mater. Res., Part B, 2011, 98(2), 223–229.
12. B. C. Hong and K. Kawano, Jpn. J. Appl. Phys., 2007, 46, 6319–6323.
13. L. Ling, X. Xu, G.-Y. Choi, D. Billodeaux, G. Guo and R. M. Diwan, J. Dent. Res., 2009, 88(1), 83–88.
14. H. H. K. Xu, J. L. Moreau, L. Sun and L. C. Chow, J. Dent. Res., 2010, 89(7), 739–745.
15. D. G. Cory and W. M. Ritchey, J. Magn. Reson., 1988, 80, 128–132.
16. A. L. Patterson, Phys. Rev., 1939, 56, 978–982.
17. C. J. Brinker and G. W. Scherer, Sol-Gel-Science, The Physics and Chemistry of Sol-Gel-Science, Academic Press, San Diego, 1990.
18. W. Ostwald, Z. Phys. Chem., 1897, 22, 289.
19. W. Ostwald, Z. Phys. Chem., 1900, 34, 495.
20. D. Ka, M. Seo, H. Choi, E. H. Kim and S. Y. Kim, Chem. Commun., 2010, 46, 5722–5724.
21. G. Scholz, C. Stosiek, J. Noack and E. Kemnitz, J. Fluorine Chem., 2011, 132, 1079–1085.
22. A. M. Keene and K. M. Tyner, J. Nanopart. Res., 2011, 13, 3465–3481.
23. J. M. Pettibone and N. R. Reardon, Nanoscale, 2012, 4, 5593–5596.

---