AlF complexes are formed when fluoride is taken up by a free-fluoride cement.
Both fluoride and potassium were uptaken by the cement.
The retained fluoride was partially re-released in water after 24 h.
Most of the amount of fluoride absorbed was confined in the cement surface.
The aim of the present study was to determine the chemical species formed inside glass-ionomer cements after fluoride uptake and to investigate the depth of penetration of fluoride ions within the cement matrix.
An experimental fluoride-free glass with composition 2SiO 2 –AlO 3 –CaO was produced. The glass powder was mixed with aqueous poly(acrylic acid) (PAA), and allowed to set. The resulting specimens were stored in 20 ml KF solution with 1000 ppm fluorine for 24 h and then placed into the same amount of water as for 24 h. A fluoride selective electrode was used to give the F concentration of the respective solutions. 19 F MAS-NMR spectra were recorded on powdered cement specimens using a Bruker AVANCE-NEO 600 spectrometer. In addition, SEM observation and EDX chemical analysis were conducted on the cross-section of a carefully fractured specimen.
Fluoride was shown to be mainly present in the surface layers of the specimen after placement in the KF solution, and only a small fraction was re-released into water. 19 F NMR spectroscopy showed that AlF complexes were formed within the cement.
The fluoride taken up by a free-fluoride glass ionomer cement mostly occupies surface layers and is retained because it bonds to aluminum within the matrix. This finding explains why the majority of fluoride taken up by conventional glass ionomer cements is retained.
Fluoride is known to have anti-caries properties [ ]. As a consequence, there is considerable interest in using dental restorative materials in contemporary dentistry that are capable of releasing fluoride [ ]. There are essentially two types of materials that are able to do this, fluoridated composite resins and glass-ionomer dental cements. The fluoridated composites are formulated to contain a fluoride compound, typically ytterbium fluoride, though there are other substances used as well. The glass-ionomers, both conventional and resin-modified, include fluoride as an inherent part of the material [ ]. Fluoride is included in ionomer glasses for a number of reasons. It reduces the fusion temperature of the glass, making the manufacture less energy intensive and reducing manufacturing costs [ ]. It also strengthens the resulting cement [ ], though the mechanism of this is not entirely clear. Some of the fluoride in the glass is transported to the matrix of the cement during setting, where it has been shown to reduce flexibility at the atomic level [ ]. This increased rigidity may be important in improving the final strength of the cement.
Fluoride that transfers to the matrix is capable of being released from the cement. This release involves two mechanisms [ ]. The first is a relatively rapid one that is sometimes referred to as “early wash-out”, and takes place within a few days [ ]. The second one is a slower stage that is based on diffusion, and is capable of being maintained for at least five years [ ]. As well as releasing fluoride, glass-ionomer can take it up [ ]. This uptake occurs as the result of exposure to fluoride-containing solutions, typically either aqueous sodium or potassium fluoride at reasonable concentrations. Some of the newly absorbed fluoride can be released again into water, leading to enhanced levels of release compared with specimens that had not been exposed to fluoride solution. However, most of the fluoride taken up, usually over 90%, is not released but becomes firmly embedded in the cement in some sort of insoluble form [ , ]. Studies using dynamic SIMS have shown that most of the fluoride taken up remains in the surface, and only small amounts diffuse further into the cement specimens [ ]. Other work, using X-ray photo-electron spectroscopy, suggested that the fluoride taken up becomes associated with calcium in Ca-based cements [ ]. A similar association may occur in strontium-based cements, because strontium has very similar chemical effects to calcium. For example, it makes almost no difference to the structure of the glass when it is substituted for calcium [ ] and it has a similar insolubility [ ].
Despite these clues about the nature of fluoride that has been taken up, the mode of uptake and the reason it becomes bound within the cement are still far from understood. We have considered this problem by carrying out an experimental study using solid-state Magic Angle Spinning (MAS) 19 F NMR spectroscopy to examine fluoride species taken up by a fluoride-free glass-ionomer cement. Such a cement was used to avoid the signals from the newly absorbed fluoride being overwhelmed by signals from fluoride already present in the cement. Fluoride-free glass-ionomers have previously been shown to be able to take up fluoride from aqueous solutions of simple fluoride salts and later to release at least some of this fluoride into water [ ]. By using this experimental cement, we have been able to gain insight into the fluoride species that forms within the cement on uptake. This also gives information on why such a high proportion is usually retained on subsequent exposure to water.
MAS NMR has been used in other studies to determine the nature of fluoride in specific solids. These have included alumino–silicate glasses [ , ] and the so-called “glass carbomer” restorative material [ ]. The technique has the advantage of giving detailed information on the environments occupied by the element of interest, in this case fluoride, in terms of the elements to which it is attached. This ought to allow information to be obtained about the speciation of fluoride following uptake into a glass-ionomer. The aim of the present study was to determine the chemical species formed within the cement matrix after fluoride uptake by an experimental fluoride-free glass.
Materials and methods
Preparation of glasses
The glass 2SiO 2 –AlO 3 –CaO used in this study was manufactured via the melt quench route. A 200 g batch of analytical grade SiO 2 (Prince Minerals Ltd., Stoke-on-Trent, UK), Al 2 O 3 , CaCO 3 , (Sigma–Aldrich, Gillingham, UK) were mixed and transferred into a platinum–rhodium (80/20) crucible for melting in an electric furnace (EHF 17/3, Lenton, Hope Valley, UK) at 1500 °C for 1 h. The melted glass was then quenched quickly in distilled water. After melting at 1500 °C temperature for 1 h in a melt quench route and grinding thoroughly by a Gyro Mill (Glen Creston, London, UK) for 7 min, the resultant powder was sieved twice through a 38 μm sieve, to obtain consistently fine and uniform particles in accordance with the requirement of <45 um for particle size of the final glass powder for a restorative indication [ ].
For preparation of the specimens, the cement was mixed as previously reported as a high-powder:liquid composition [ ] i.e. 0.6 g of glass was mixed with 0.2 g of poly(acrylic acid) and subsequently combined with 0.2 g of (+) tartaric acid solution at a concentration of 10% m/m. Polytetrafluoroethylene (PTFE) moulds of 15 mm diameter × 0.5 mm thickness were used in order to ensure flat and thin specimens, with a large relative surface area to detect the fluoride compounds formed within the cement matrix. Once placed inside the mould, the cement was covered by a polyester strip at the top and bottom surfaces and pressed by a glass plate on both sides with the help of a screw clamp. They were then placed in an incubator at 37 °C for 1 h and then placed into plastic tubes containing 20 ml of 0.275% KF solution (representing 1000 ppm F and 1850 ppm K) for 24 h. Some specimens were removed from the KF solution and after rinsing placed into the same amount (20 ml) of deionised water for 24 h. A fluoride selective-electrode and associated meter (Orion 9609BNWP with Orion pH/ISE meter 7 Fisher Scientific Loughborough UK) were used in order to obtain fluoride concentrations, where TISAB IV solution was added in a 1:1 ratio of volumes with the original amount of the storage solution. The fluoride electrode was previously calibrated twice (with and without TISAB) with five standard solutions before the measurements.
19 F magic-angle spinning nuclear magnetic resonance analysis (MAS NMR)
The 19 F MAS-NMR spectra were obtained both on the disc samples ion exchanged as described previously and on powder ground from discs for 7 min in a Gyro Mill before being placed into the KF solution. The ion exchange was conducted with cement powder since the ion exchange of the disc resulted in a spectrum with a low signal to noise because of the fact that only small amounts of fluoride were taken up into the surface. After drying the obtained cement powder, from both experiments the 19F MAS-NMR data were collected at 564.686 MHz using a Bruker AVANCE-NEO 600 spectrometer (Bruker Coventry UK). The powder was packed into a regular-walled, bottom-closed, 2.5 mm zirconia rotor. The samples were spun at the magic angle at 22 kHz and recycle times were 30 s for all spectra. The chemical shift was referenced with a signal from 1 M aqueous solution of NaF at −120 ppm.
Scanning electron microscope (SEM) and Energy dispersive X-ray analysis (EDX)
After placement in the KF solution for 24 h, the thin cement disc was carefully broken in half in order to perform the analysis of the external cross-section by the Scanning Electron Microscope (Zeiss Evo LS15, Oberkochen, Germany). The external surface was also analysed. Chemical analysis was processed by the software Inca EDX System (INCAx-act Oxford Instruments, Abingdon, United Kingdom). To prevent electrical charging of the samples, they were first sputter coated with gold for 30 s.
19 F-MAS-NMR spectroscopy of the fluoride-free glass powder showed no peak ( Fig. 1 ), confirming that there was no fluoride present in the original cement. Table 1 presents the fluoride concentration determined by the Ion Selective Electrode before and after 24 h immersion in the KF solution, the resulting difference and finally the amount of fluoride re-released after 24 h in water. As shown in Fig. 2 19 F-MAS-NMR spectroscopy of the ion exchanged cement powder exhibited a peak at −150 ppm, that corresponds to AlF complexes formed within the cement matrix after the fluoride was taken up. The 19 F MAS-NMR spectrum of the cement disc after ion exchange in a spectrum identical to Fig. 2 but with a relatively high signal to noise. In addition, there was a small sharp peak at −124 ppm corresponding to free fluoride that may correspond to fluoride in solution possibly within pores within the disc. The presence of one single peak in the spectrum of the ion exchanged cement powder shows that all bound fluorine atoms are in closely similar chemical environments. Spinning side bands are also seen and denoted in the spectrum by asterisks. Fig. 3 presents SEM images of the specimen with uniformity of particles.