Glass ionomer cements: effect of strontium substitution on esthetics, radiopacity and fluoride release.

OBJECTIVE: SrO and SrF2 are widely used to replace CaO and CaF2 in ionomer glasses to produce radiopaque glass ionomer cements (GIC). The purpose of this study was to evaluate the effects of this substitution on release of ions from GIC as well as its effect on esthetics (translucency) and radiopacity. MATERIALS AND METHODS: Cements were produced from ionomer glasses with varying content of Sr, Ca and F. The cements were stored in dilute acetic acid (pH 4.0) for up to 7 days at 37°C. Thereafter, the cements were removed and the solution was tested for F(-), Sr(2+), Ca(2+), and Al(3+) release. Radiopacity and translucency were measured according to BS EN ISO 9917-1:2003. RESULTS: Ion release was linear to t(1/2) suggesting that this is a diffusion controlled mechanism rather than dissolution. The fluoride release from the cements is enhanced where some or all calcium is replaced by strontium. Radiopacity shows a strong linear correlation with Sr content. All cements were more opaque than the C0.70 0.55 standard but less opaque than the C0.70 0.90 standard which is the limit for the ISO requirement for acceptance. SIGNIFICANCE: This study shows that the replacement of calcium by strontium in a glass ionomer glass produces the expected increase in radiopacity of the cement without adverse effects on visual properties of the cement. The fluoride release from the cements is enhanced where some or all calcium is replaced by strontium.

The effect of ultrasound on the uptake of fluoride by glass ionomer cements.

Ultrasound has been shown to improve the set of glass ionomer cements (GICs) and also other cement properties. In particular, the release of fluoride is enhanced. These cements also can take up fluoride ion from liquids. The aim of this study is to investigate the effect of ultrasound on this cement property. Two commercial dental restorative GICs were used together with a modified commercial material and an experimental material based on a F-free glass. All three commercial materials came in capsules which were mixed as makers directed, the experimental material was mixed as in previous papers. Mixed cement was placed polyethylene moulds to create 3 × 2 mm thick discs. These were either allowed to standard set for 6 min or set with ultrasound for 55 s. 18 samples were made for each material/set. Three samples were placed in 4 ml of 0.2% NaF solution for 24 h at 37°C. The cylinders were removed and the F concentration of the solutions measured by ISE using TISAB decomplexant. F uptake was determined by difference from the original NaF concentration. The two conventional GICs showed reductions of 17.4 and 8.5% for ultrasound compared to standard set whereas the modified material increased by 32.3% and the experimental one by 20.6%. It is suggested that the effect of ultrasound may increase the surface area of the residual glass particles in the GIC which would increase F uptake. In GICs where considerable F ion is released into the cement matrix by the enhanced reaction caused by ultrasound this may be sufficient to reverse the former effect producing the reduced uptake observed.

The effect of ultrasound on the setting reaction of zinc polycarboxylate cements.

The set of glass ionomer cement (GIC) is accelerated by application of ultrasound. Although GIC has somewhat displaced zinc polycarboxylate cement (ZPC) in dental applications the latter is still extensively used. Like GIC, it provides direct adhesion to tooth and can provide F release, but is more radiopaque and biocompatible than GIC. The aim of this study is to examine the effect of ultrasound on the setting of ZPC using Fourier transform infra red spectroscopy and any interaction with SnF(2) addition. ZPC with and without SnF(2) addition (+/-S) at luting (L) 2:1 P/L ratio and restorative (R) 4:1 P/L ratio consistencies. Ultrasound is applied to the cement using Piezon-Master 400, EMS, Switzerland at 60 s from start of mixing for 15 s. The ratios of absorbance peak height at 1,400 cm(-1) -COO(-) to that at 1,630 cm(-1) -COOH were measured and compared those obtained for the cement not treated with US. These values were taken at the elapsed time at which no further change in spectrum [ratio] was observed at room temperature [10-20 min]. The US results are taken at 2 or 3 min. No US: R/+S (1.09), R/-S (1.2), L/+S (1.07), L/-S (1.04); US: R/+S (1.50), R/-S (1.64), L/+S (1.38), L/-S (1.05). The results show all four ZPC formulations are very sensitive to ultrasound whether with or without SnF(2). Reducing US to 10 s produces lower initial ratios but these increase up to 10 min when very high ratios (>2) are obtained. Previous studies with restorative GICs found that 40-55 s US was needed to produce the effect found with 15 s on ZPCs. ZPC powder is more basic than GIC glass; this may account for ZPC's greater sensitivity to US. Ultrasound may provide a useful adjunct to the clinical use of ZPC both as luting agent and temporary restorative.

Kinetics of fluoride ion release from dental restorative glass ionomer cements: the influence of ultrasound, radiant heat and glass composition.

To compare the effect of ultrasonic setting with self curing on fluoride release from conventional and experimental dental glass ionomer cements. To compare hand mixed and capsule mixing and the effect of replacing some of the reactive glass with zirconia. In a novel material which advocated using radiant heat to cure it, to compare the effect of this with ultrasound. To evaluate the effect of ultrasound on a glass ionomer with fluoride in the water but not in the glass. 10 samples of each cement were ultrasonically set for 55 s; 10 controls self cured for 6 min. Each was placed in 10 ml of deionised water which was changed at 1, 3, 7, 14, 21, 28 days. The solution fluoride content was measured using a selective ion electrode. All ultrasound samples released more fluoride than the controls. Release patterns were similar; after a few days, cumulative fluoride was linear with respect to t(1/2). Slope and intercept of linear regression plots increased with ultrasound. With radiant heat the cement released less fluoride than controls. The effect of ultrasound on cement with F in water increased only slope not intercept. Zirconia addition enhances fluoride release although the cement fluorine content is reduced. Comparison of capsule and hand mixing showed no consistent effect on fluoride release. Ultrasound enhances fluoride release from GICs. As heat has an opposite effect the heat from ultrasound is not its only action. The lesser effect on cement with fluoride only in the water indicates that of ultrasound enhances fluoride release from glass.

The role of glass composition in the behaviour of glass acetic acid and glass lactic acid cements.

Cements have recently been described, made from glass ionomer glass reacted with acetic and lactic acid instead of polymeric carboxylic acid. From their behaviour a theory relating to a possible secondary setting mechanism of glass ionomer has been adduced. However, only one glass (G338) was used throughout. In this study a much simpler glass ionomer glass (MP4) was compared with G338. This produced very different results. With acetic acid G338 formed cement which became resistant to water over a period of hours, as previously reported, MP4 formed cement which was never stable to water. With lactic acid G338 behaved similarly to G338 with acetic acid, again as reported, but MP4 produced a cement which was completely resistant to water at early exposure and unusually became slightly less resistant if exposure was delayed for 6 h or more. These findings indicate that the theories relating to secondary setting in glass ionomer maturation may need revision.

Uptake of fluoride ions by the glass component of glass ionomer cement.

OBJECTIVES: The uptake of F(-) ions by glass ionomer cement (GIC) has been extensively studied but the precise location of the F(-) ion in the GIC structure has not been reported. The aim of this study is to use model materials to elucidate the possible locations. GIC consists of residual particles of aluminosilicate glass surrounded by depleted glass in an ionically crosslinked polymeric matrix. This study evaluates uptake by both unreacted glass particles (RAW) and particles acid-treated to produce a depleted glass surface (DEPLETED). METHOD: Three glasses, previously studied as cements, were tested at the particle size used in GICs. LG30 contained Al, Ca, O, P, and Si; LG26 also contained F and AH2 contained Na as well. To produce depleted surfaces they were immersed in acetic acid washed and dried at room temperature. Test samples (N=5) were immersed in KF solution (900ppm F). Control solutions without glass were used. Both were stored at 37C for 48h. F concentrations were measured using ISE with TISAB IV. RESULTS: Uptake was Control-test expressed in micromol/g glass. RAW LG30=56 (16); LG26=35(17); AH2=17(31). DEPLETED LG30=285(41); LG26=431(42); AH2=286(50). The levels of F uptake by DEPLETED were comparable to those found with GICs formed with these glasses. CONCLUSIONS: The glass particles of GIC, and particularly the depleted zones surrounding them, can account for an appreciable amount of the F uptake observed with cements but do not show the same relation between F content and F uptake previously observed with cements.

The effects of lubrication on the temperature rise and surface finish of amalgam and composite resin.

UNLABELLED: It was thought that when finishing and polishing direct filling materials lubrication would affect the surface roughness and temperature rise in samples of amalgam and composite. OBJECT: Previous work by the authors has shown that there is an optimum load, speed and time that produced the smoothest surface when finishing amalgam and composite resin using each of four grades of a disc system. This work was undertaken to examine the effects on temperature rise in samples of amalgam and composite resin of finishing dry compared to finishing with different lubricants. The experiments all used these optimum loads, speeds and times. It also compares the surface finish produced using different lubricants. MATERIALS AND METHODS: A high copper amalgam and a hybrid composite resin were finished using the four grades of abrasive discs. Samples produced were 25 mm long by 6 mm wide by 2 mm deep. A thermocouple was inserted 1 mm into the base of the samples. The thermocouple was connected via an electronic thermometer to a computer that permitted the display and recording of temperature against time. After roughening, the samples were finished and polished in a specially constructed jig that mimicked oral finishing. The pre-determined optimum loads, speeds and times were used sequentially for each of the four grades of disc. Five samples were tested for each method of finishing. Firstly, run dry, then in turn lubricated with water, walnut oil and petroleum jelly. After the use of each abrasive disc the surface roughness was measured. One of the five samples was selected at random and prepared for examination in the scanning electron microscope. All results were subjected to non-parametric statistically analyses. RESULTS: With both materials the temperature rise was greatest when run dry, followed by petroleum jelly, walnut oil and the least was when lubricated with water. With these two materials the surface roughness correlates negatively with the temperature rise. The smoothest surface being achieved when finished dry. CONCLUSIONS: To obtain the smoothest surface finish amalgam and composite should be finished dry but further work is needed to assess the effect of the temperature rise found in the materials on the pulp.

Ion processes in glass ionomer cements.

Ion processes are involved in many aspects of glass-ionomer cements. The ions released from the glass take part in the formation of the cement matrix. Although this process has been investigated, particularly using model cement systems, no study provides a complete matrix composition. Combining results from different studies enables an approximate composition to be derived. The importance of Phosphorous in controlling ion release from the glass surface has been identified in a number of studies. The release of ions from the set cement into water (and other aqueous liquids) has been much reported, particularly for fluoride. Over most of the release periods studied (i.e. from >7 days up to 3 years), release of F ion is related to t1/2 indicating a diffusion-controlled process. Other ions, except possibly Na+ also show this relationship. The amount of cumulative F release whilst maintaining this relationship indicates that more F than is in the matrix is involved. Ion chromatography would probably elucidate the precise form of the ionic species released. Glass-ionomer cements take up ions from solutions in which they are immersed. The levels are much higher than required to produce as internal/external equilibrium. Studies using dynamic SIMS and XPS give some information on ion location and elemental association. It is suggested that ToF SIMS would elucidate these further. Re-release of uptaken ions can vary considerably for different cements and ion species. Surface disruption of glass ionomers is caused by both F ion and monofluorophosphate ion and occurs much more readily in F containing cements than in F free ones. The mechanism of this process has not been elucidated. Analysis of the ions released from the cement as disruption occurs should provide an indication of the site of attack.

The effects of lubrication on the temperature rise and surface finish of glass-ionomer cements.

OBJECT: Previous work [Jones CS. Factors influencing the finishing of direct filling materials. PhD Thesis, University of London; 2002] has shown that there is an optimum load, speed and time that produced the smoothest surface when finishing glass-ionomer cement using each of four grades of a disc system. This study looks at the effects of lubrication on the temperature produced in samples of GIC when finished dry and with different lubricants using these optimal loads, speeds and times. It also compares the surface finish produced using different lubricants. MATERIALS AND METHODS: A thermocouple connected so that it permitted the display and recording of temperature against time was inserted 1mm into the base of samples of a glass-ionomer cement. The samples were finished and polished using each of the grades of a disc system in a specially constructed jig that mimicked oral finishing. After roughening, the pre-determined optimum loads, speeds and times were used sequentially for each of the four grades of disc. Five samples were tested for each method of finishing. Firstly run dry, then in turn lubricated with water, walnut oil and petroleum jelly. After the use of each abrasive disc the surface roughness was measured using a profilometer. One of the five samples was selected at random and prepared for examination in the scanning electron microscope. All results were subjected to non-parametric statistically analyses. RESULTS: Walnut oil and petroleum jelly produced significant temperature increases compared to both dry and with water finishing. Lubricated with water significantly reduced the temperature rise compared to dry. The Ra values of 0.5 microm was obtained for the coarse and a value of 0.3 microm for the medium discs run without lubrication. With lubrication the Ra increased although there was little difference between the lubricants. However the photomicrographs showed that walnut oil and petroleum jelly caused gross morphological changes indicating major surface destruction. CONCLUSIONS: The practice of finishing GICs using petroleum jelly or similar lubricant appears to be detrimental. Further experimental work needs to be done to advise practitioners on finishing GICs to produce the smoothest surface possible.

Laboratory study of the loads, speeds and times to finish and polish direct restorative materials.

This study investigated the loads, speeds and times required to achieve the smoothest surface on samples of amalgam, composite resin and glass-ionomer cement using the four grades of a disc polishing system. The tests were conducted on a specially constructed jig. The restorations were abraded with the appropriate disc using a range of loads, speeds and times based on a previous practitioner study. The surface roughness of each material was measured before and after abrading. The load was varied first, followed by speed and finally time. Photomicrographs were taken of a sample finished at the optimum load, speed and time and of loads above and below the optimum that produced the smoothest surface finish. In addition the fractured edge of a sample finished at the optimum values was examined. There was an optimum load, speed and time that produced the smoothest surface for each restorative material. For amalgam and composite the surface roughness value decreased as the discs became finer. There was a reduction in the roughness value for glass-ionomer cement using the two roughest discs. The values obtained were substantially in contrast from the earlier practitioner study. There were specific values for load, speed and time that produced a smoother surface for each material.

The advantages and disadvantages of running a clinical trial in general practices.

This paper reviews the experiences encountered in running a clinical trial on the use of a metal reinforced glass-ionomer cement in general dental practices. The practitioners were asked to place both the test material and amalgam in the same patient and to take impressions both at placement and at three recall intervals. Subsequently plaster casts were produced from these impressions. These were then assessed by three independent observers to provide evaluation of the relative wear of the two restoratives under evaluation. A commercial laboratory manufactured the models on which the three independent observers carried out the assessment of wear. Although the practitioners indicated they would be able to provide the number of restorations required in a relatively short period these expected numbers were never achieved. Despite recruitment of additional participants the trial never did achieve the number of restorations required. The performance of the metal reinforced glass ionomer appeared to vary dependent on the practitioner placing the filling. Questioning of the participants found that some participants were finishing the material using a method specifically contraindicated in the protocol for the trial, the directions supplied with the product, and in briefing sessions held prior to the trial. This implies that there can be major problems in undertaking clinical trials of this nature in the general dental service and has serious clinical implications for those contemplating this type of evaluation and for manufacturers introducing new products.

Chlorhexidine release from an experimental glass ionomer cement.

Glass ionomer cements (GIC) can potentially be used as matrices for the slow release of active species, as has been shown previously for fluoride ions. This study investigated the use of an experimental GIC as a carrier for the release of chlorhexidine acetate (CHA) at included concentrations ranging from 0.5% to 13.0% of CHA by weight. Release into water was examined using high-performance liquid chromatography. All measurable chlorhexidine was released within 22 h1/2, however this was less than 10% of the total mass incorporated in the specimens. An increased percentage of CHA incorporated into the powder gave an increased release into the surrounding water. The bulk of the CHA was retained within the cement. For comparison, the surface chemistry of a CHA-containing GIC was examined using X-ray photoelectron spectroscopy before and after prolonged immersion in water. This confirmed retention of a large amount of CHA. Spectra after leaching appeared very similar to those from a CHA-free GIC after immersion in a CHA solution. In order to explore the effect of CHA-inclusion on the cement properties, compressive strengths, working and setting times were also measured. In general, compressive strengths were found to be decreased in direct proportion to the quantity of CHA added, while working and setting times increased.

Effects of variation in particle size on biaxial flexural strength of two conventional glass-ionomer cements.

Conventional glass-ionomer cements (GICs) have a slow maturation time. Reduction in time of maturation may be achieved by acceleration of the setting reaction. One factor that assists this is the reduction in glass particle size producing a larger surface area for reaction. The resulting rapid set and more rapid maturation should potentially lead to less long-term degradation. Biaxial strength measurements were made with respect to time for two GICs of similar compositions but with differing particle size distributions at different time intervals after immersion in both water and artificial saliva. There was little difference between the strength of the two materials over periods up to 12 weeks. A theoretical estimation of the relative surface areas of glasses showed that, despite there being twice the surface area available for reaction for one glass, there was little difference in strengths values between the two materials at any of the times tested here. The similarity in strength values despite this substantial difference suggests that the larger particles may have a greater influence in the cement forming process.

Glass ionomer cement: evidence pointing to fluorine release in the form of monofluorophosphate in addition to fluoride ion.

The fluoride ion released from glass ionomer cements into water is reportedly, in part, complexed with other elements present in the cement. When measured using ion selective electrode potentiometry (ISE) a decomplexant TISAB IV (T) is used to convert all fluoride to F- ion which the ISE can detect. In this study, an additional decomplexing procedure (H) designed to hydrolyse fluorine covalently bonded to phosphorus in the monofluorophosphate (MFP) ion into F- was also used. The soluble products from three glass ionomers were analysed by both techniques (H & T). Five 1 x 10 mm discs were each immersed in 10 ml of de-ionised water. This was changed and 4 ml analysed by T and 4 ml by H at 1, 2, 3, 6, 10, 13, 17, 21, 24, 28, and 31 days. H was greater than T for 161 of the 165 pairs ( chi2=74.7, p=<0.001 ). The total cumulative F release H (in micromol/g cement) at 31 days for AH2 was 122.3, s.d. 30.8; LG26 44.0, s.d. 1.55; LG30 10.0, s.d. 3.15 as compared T results of 100.1, s.d. 31.1; 30.3, s.d. 1.92; 3.7, s.d. 1.36, respectively. In all three cases the H was significantly greater than T (matched pair 't' test with p=0.01 or less). H-T was show to have a very strong associative relationship with t1/2 (R2=0.98 or greater p<0.001 ). Evaluating the ratio of P:F in the cements in comparison with the ratio of additional F measured by H to that measured by T produced a relationship log[(H-T)/T]=0.28 x log[P/F]-0.45 with R2=0.999. It is concluded that glass ionomers release more fluorine than is detected by ISE using TISAB IV. If this F is in the form of MFP this may be released more completely into saliva than F as F-, release of which is substantially reduced by Ca2+, since calcium monofluorophosphate is more soluble than CaF2.

The in vivo perception of roughness of restorations.

OBJECTIVE: To determine a threshold of detection value for surface roughness of restorations by patients using their tongue. METHOD: Samples of composite resin were finished with differing grades of abrasive. The surface roughness (Ra) was measured and representative scanning electro-micrographs taken. These were compared with labial enamel. Twenty-five volunteers were asked to rank them in order of perceived roughness using the tip of their tongue. RESULTS: These showed that the 60% of volunteers were able to rank the specimens correctly, and were able to distinguish differences in roughness values from between 0.25 and 0.50 microm. This range encompasses that of natural enamel. CONCLUSIONS: The subjects were able to distinguish lower roughness values than have previously been reported. It is concluded that when finishing restorations the surface should have a maximum roughness 0.50 microm if it is not to be detected by the patient.

The effects of adding fluoride compounds to a fluoride-free glass ionomer cement on subsequent fluoride and sodium release.

Studies have shown that ions in a glass ionomer matrix are 1-10% of the amounts present in the original glass. To measure more precisely the release from a cement matrix, known amounts of ions were added to LG30 glass which was fluoride and sodium-free. Cement without additions acted as the control. 1.4-1.6% of each of sodium, calcium and aluminum fluorides were added to three portions of control blend. The sodium and fluoride release into deionised water from five discs of each cements blend was measured for 8 months. This represented complete release for sodium but not for fluoride. Traces of fluoride and sodium in the glass produced low but measurable amounts indicating about a third of the fluoride and substantially all sodium present in LG30 was released. The addition of calcium fluoride had no significant effect on sodium or fluoride release and aluminium fluoride minimal effects. Adding sodium fluoride significantly enhanced release of both ions although fluoride release was less than from a glass containing 5% fluoride. Only small proportions of the additions, 2-5% of the fluoride and 13% of sodium, were released. Sodium and fluoride appeared to be released independently. For LG30 cements additives were poor at supplying extra ions.

Fluoride uptake by glass ionomer cements: a surface analysis approach.

Despite extensive research, the mechanism by which glass ionomer cements take up fluoride ions from solution remains unclear. To date, the majority of studies have concentrated on measuring the removal of ions from solution. In this study, we demonstrate the application of X-ray photoemission spectroscopy and secondary ion mass spectrometry to the surface analysis of the cements, after the introduction of fluoride either by doping or by immersion. Fluoride ion uptake from potassium fluoride solution is correlated with the formation of a surface layer which is rich in calcium as well as fluoride.

The glass ionomer cement: the sources of soluble fluoride.

This study aimed to investigate certain processes of fluoride production which enable glass ionomer cements to leach fluoride. Two fluoroaluminosilicate glasses, G338 and LG26 were used. The free and total fluoride which could be dissolved from the glasses was measured, before and after acetic acid washing. Both glasses contained appreciable amounts of soluble fluoride prior to any acid treatment. The latter process reduced the amount to some 75% of the original levels. Replacing the customary polymeric acid with propionic acid produced a cement which disintegrated in water allowing the amount of fluoride generated by the cement forming process to be measured. Cement production increased soluble fluoride by a further 3%. Both glasses behaved similarly when undergoing the various processes. G338 produced significantly greater quantities of fluoride, of the order of 10, compared with LG26 although containing only three times the amount of fluoride in the glass formula. A substantial proportion, over half, of the total fluoride was complexed especially after contact with cement and when G338 was used. During the period of the experiment, 21 days, total fluoride release did not seem to depend on the square root of time.

Interactions between glass ionomer cement and alkali metal fluoride solutions: the effect of different cations.

This study examines the effect of different cations in equimolar alkali metal fluoride solutions on their interactions with glass ionomer cements. Uptake of both fluoride and cation were measured together with change in solution pH and morphological changes in the cement surface. Two cements were used; AH2, a dental restorative cement containing both fluorine and alkali metal (Na) as glass components and LG30, which contained neither. Discs of cement 1 x 10 mm were set in moulds at 37 degrees C for 1 h then, stored in water for 3 days at 37 degrees C. Discs in each test group (N = 5) were immersed in 10 ml of solutions of either NaF, KF, or RbF, all containing 900ppm F, control discs were stored in water, all at 37 degrees C for 24h. Solutions were analysed for F- by ISE potentiometry, Na+ by the same technique and K+ and Rb+ were analysed by atomic absorption spectrometry. Uptake was obtained by difference between solution used for immersion and the control solution. Solution pH was measured potentiometrically. The surface roughness of the discs was measured by linear stylus profilometry. Fluoride ion uptakes for AH2 were 451 micromol/g NaF, 378 KF, and 318 RbF. The comparable figures for LG30 were 202, 161, and 159. Differences between cements were all statistically significant and also between solutions pairings except for the KF/LG30 vs. RbF/LG30. Uptake of cations was equimolar for AH2/ NaF, AH2/RbF and LG30/KF but M+:F- ratios were significantly above unity for AH2/KF and LG30/NaF and significantly below unity for LG30/RbF. The pH changes were all positive and were significantly higher for AH2 than LG30 and for RbF compared to the other fluoride solutions for each cement (probably because of its lower initial pH). The final pH of all solutions were less than I pH unit from neutral (pH7). The AH2 cement test discs all showed significant increase in roughness (Ra) compared to control discs stored in water whereas the LG30 discs showed no such difference. Regression analysis showed a significant positive correlation between fluoride uptake and Ra. It was concluded that changing the alkali metal cation influenced all four variables examined (F- uptake, M+ uptake, pH change and extent of cement surface roughening).

Kinetics of fluoride release from zinc oxide-based cements.

Considerable attention has been given to the release of the cariostatic fluoride ion from glass-based dental cements (dental silicate and glass ionomer). In these, the total available fluoride content is not precisely known since fluorine is distributed between the cross-linked aqueous salt matrix, partially dissolved glass, and undissolved glass. In analogous cements based on zinc oxide the fluoride is added as highly soluble SnF2. The object of this study is to compare the F- ion release profiles of commercial zinc polycarboxylate and zinc phosphate containing 4.4 and 3.6% SnF2, respectively. Mixed cements were clamped in split ring moulds to produce discs of 10 mm x 1 mm after storage at 37 degrees C for 1 h. Each was weighed and immersed in 10 ml of deionised water. When this changed, at 13 time intervals up to 98 days, the fluoride content was measured using an ion selective electrode. The mean (N = 3) values obtained were expressed cumulatively [F] in micromol F ion/g cement. The total [F] released was 111 for the zinc polycarboxylate and 286 for zinc phosphate compared with total F in the cements of 561 and 464, respectively. When the cumulative [F] was plotted versus t(1/2) close associations were found for both cements. For the polycarboxylate the regression line [F] = 10.6t(1/2) + 9.9 fitted well over the whole 98 days (R = 0.997). For the phosphate a better fit regression line was obtained using results up to 32 days only; [F] = 36.8t(1/2) - 8.4 (R = 0.999). For t > 32 days results increasingly deviated from this line. These results fitted a regression line of the form [F] = 81.7log(e) t - 87.3 (R = 0.9997). Comparisons are made with data from previous authors both for zinc phosphate cement and glass-based cements and with diffusion theory of F ion release. It is concluded that zinc-based cements provide some indications of how glass-based cements may behave over long periods of release and that zinc phosphate is the material of clinical choice for orthodontic cementation if maximal fluoride release is the prime criterion.