Journal
of
Dentistry
Journal of Dentistry 26 (1998) 591–597
The uptake and release of fluoride by ion-leaching cements after exposure
to toothpaste
M. Rothwell, H.M. Anstice*, G.J. Pearson
Biomaterials Department, Eastman Dental Institute, London, UK
Received 30 January 1997; revised 16 April 1997; accepted 17 June 1997
Abstract
Objectives: The cariostatic action associated with the glass-ionomer cement (GIC) is usually attributed to its sustained release of fluoride.
However the ability of the GIC to act as a fluoride reservoir, taking it up from an external source (e.g. toothpaste, mouthwash) and
subsequently releasing it over time, may also be a contributory factor. This study investigated the reservoir effect of various recently
introduced ion-releasing cements: two resin-modified glass-ionomer cements (Fuji II LC, Vitremer), a compomer (acid-modified composite
resin) (Dyract), and a recently introduced conventional glass-ionomer (Fuji IX).
Methods: Specimens were exposed to a fluoridated toothpaste after 28 and/or 58 days. The release of fluoride into the storage water, both
before and after exposure, was monitored using a differential electrode cell.
Results: There was no significant difference in the fluoride releases from Vitremer and Fuji II LC. These materials released significantly
more fluoride than Fuji IX and Dyract. All the materials released more fluoride on the day after exposure to an external fluoride source
compared with the day before exposure. Release rates returned to baseline within 3 days. Within the time periods of the study, only the
uptake/re-release of Fuji IX was adversely affected by late exposure. All the materials showed an enhanced uptake and release on repeated
exposure to the external fluoride source.
Conclusions: All the materials under test (Dyract, Fuji II LC, Vitremer and Fuji IX) released significant amounts of fluoride and reacted
positively to exposure to an external fluoride source. q 1998 Elsevier Science Ltd. All rights reserved.
Keywords: Fluoride; Resin-modified glass-ionomer; Compomer; Release; Sorption
1. Introduction
One of the most desirable properties of a restorative
material is that it should have caries-inhibiting properties.
The properties of a glass-ionomer cement (adhesion,
minimal setting shrinkage and thermal properties similar
to tooth tissue) ensure the formation of a good seal around
the restoration. This seal will play a significant role in the
prevention of further caries, but it is widely thought that the
cariostatic action associated with the glass-ionomer [1–3] is
as a result of its sustained release of fluoride [4].
It should be noted that a recent publication has questioned
the cariostatic properties of the glass-ionomer and claims
that its performance is not superior to that of a non-fluoride
releasing composite resin [5]. The findings of this study
require further evaluation, but it should be noted that they
* Corresponding author. Biomaterials Department, Eastman Dental
Institute, 256 Grays Inn Road, London, WC1X 8LD, UK. Tel.: 0171 915
1018; fax: 0171 915 1019; e-mail: hanstice@eastman.ucl.ac.uk
0300-5712/98/$19.00 q 1998 Elsevier Science Ltd. All rights reserved.
PII: S0 30 0 -5 7 12 ( 97 ) 00 0 35 - 3
are in conflict with other evidence in the area [6–12], and in
particular a recently published longitudinal study which
found that the occurrence of caries on surfaces adjacent to
glass-ionomer restorations was significantly less than that
found adjacent to amalgam [3].
In vitro studies have shown an area of tooth resistant to
demineralisation existing around a glass-ionomer cement
restoration [13]. This property has also been observed
with the resin-modified glass-ionomer cement [14].
There is no evidence in the literature of the level of
fluoride release required to make a restorative material
cariostatic. Indeed it is perhaps not the inherent fluoride
release of a cement that is important, rather the ability of
a material to be reactivated by exposure to external fluoride
sources. Forsten [15], Hatibovic-Kofman and Koch [16] and
Seppa [17] have all demonstrated that the glass-ionomer
cement will take-up fluoride from external sources (fluoride
solutions or toothpaste). This is then released in a controlled
manner over time. For example, in the study reported by
Hatibovic-Kofman and Koch the fluoride release from a
592
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
Table 1
Materials used in this study
Material
Description
Manufacturer
Batch Number
Dyract
Fuji II LC
Acid-modified composite resin
Resin-modified glass-ionomer
DeTrey Dentsply, Konstanz, Germany
GC Corporation, Tokyo, Japan
Vitremer
Resin-modified glass-ionomer
3M Dental, Minnesota, USA
Fuji IX
Conventional glass-ionomer
GC Corporation, Tokyo, Japan
S94082252
Powder: 454
Liquid: 19940830
Powder: 280441
Liquid: 210445
Powder: 110341
Liquid: 110341
variety of glass-ionomers was boosted to values significantly above baseline for at least the 2 weeks following a
single exposure to fluoridated toothpaste. The behaviour of
these new ion-releasing materials on exposure to fluoride
toothpaste has not been established clearly.
2. Materials and methods
The materials used in this study were an encapsulated
acid-modified composite resin (compomer), two handmixed resin-modified glass-ionomers and a hand-mixed
conventional glass-ionomer. Details of trade name,
manufacturer and batch number are given in Table 1.
Disc specimens of the materials (10 mm diameter, 1 mm
thick) were prepared using brass split ring moulds. The
hand-mixed materials were dispensed and then mixed
according to the manufacturers’ instructions. For all the
materials the cement paste was placed in the mould,
which was positioned on a polythene separating sheet
supported on a glass microscope slide. The mould also
contained the end of a 15 cm length of unwaxed dental
floss inserted into the mould through its split. The floss
was positioned so its end was at the centre of the mould.
Once sufficient cement paste had been placed in the mould,
a second polythene sheet and slide were placed over the
filled mould and any excess cement paste was extruded
through the split in the mould using hand pressure.
The resin-modified glass-ionomer and acid-modified
composite specimens were then light-cured using a standard
dental light-curing unit (Euromax, DeTrey Dentsply)
following their manufacturers’ instructions. A patch-curing
technique was used to ensure that all parts of the specimen
were irradiated for at least the time recommended by the
various manufacturers. The specimens, still sealed in their
moulds, were then placed in an incubator at 378C. After 1 h
the specimens were removed from the incubator, removed
from their moulds and then weighed ( 6 0.0001 g). The
specimens were suspended, using the floss, in 100 ml of
deionised water in individual sealable plastic containers
and stored at 378C for the duration of the experiment.
The specimens were divided into two test groups. The
first group (comprising of five specimens of each material)
was exposed to an external fluoride source after 28 days.
The exposure was then repeated after a further 28 days
(early first exposure followed by second exposure). The
second group (also comprising five specimens of each
material) was left for 58 days before exposure to the external fluoride source (late first exposure).
The external fluoride source used in this study was a
toothpaste containing 0.32% sodium fluoride (Colgate
Total, Colgate Palmolive, UK). The procedure used for
exposure was that the specimen was removed from its
storage solution, wiped clean with a tissue to remove any
surface water and then totally immersed in 10 ml of toothpaste for 1 h. After immersion the specimen was wiped
clean and then thoroughly rinsed by exposure of each side
of the specimen to a stream of deionised water for 15 s. The
specimen was wiped clean with a tissue and then placed in
100 ml of fresh deionised water. The fluoride release from
the materials was measured by monitoring the fluoride
concentration of the storage solution for each specimen.
Fluoride concentrations were determined using the
differential electrode cell technique comprising a fluoride
sensitive electrode and a combination pH electrode. This
method gives a fast response using only 0.5 ml of test
solution [18]. For each determination 0.5 ml of the sample
and 1.5 ml of hydrochloric acid (0.1 M) were placed in a
plastic microsample dish with a small stirring rod. Hydrochloric acid acted as a decomplexing agent during the
fluoride measurements. Prior to each use the apparatus
was calibrated using standard fluoride solutions of known
concentration and was recalibrated every 2 h to compensate
for local temperature and humidity changes. Once the
fluoride concentration in the storage solution was known,
the fluoride release per gram of cement could be calculated.
Measurements were made both before and after exposure
to the external fluoride source. The initial release profile was
determined by frequent measurements over the first week
after preparation of the samples, weekly measurements were
then made up until the time of exposure of the specimen.
Regular measurements (days 1, 3 and 7) were made in the
week following exposure to the external source and then
weekly measurements were made up until 4 weeks after
exposure. Cumulative fluoride release results were plotted
against the square root of time to ascertain whether the
process was diffusion-controlled. The best fit straight lines
were generated using Excel 97 (Microsoft Corporation).
Where appropriate the results were analysed using Mann–
Whitney non-parametric statistics.
593
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
Table 2
Cumulative fluoride release (mg/g cement) over the first 28 days (all specimens)
Day
Dyract
Fuji II LC
Vitremer
Fuji IX
1
2
3
4
7
14
21
28
0.16 (0.04)
0.18 (0.04)
0.19 (0.02)
0.20 (0.03)
0.27 (0.04)
0.43 (0.09)
0.54 (0.13)
0.66 (0.16)
0.49 (0.27)
0.53 (0.23)
0.61 (0.24)
0.68 (0.25)
0.85 (0.33)
1.20 (0.40)
1.30 (0.38)
1.56 (0.51)
0.57 (0.18)
0.67 (0.23)
0.85 (0.29)
0.91 (0.34)
1.11 (0.42)
1.35 (0.40)
1.40 (0.43)
1.80 (0.46)
0.23 (0.05)
0.27 (0.07)
0.29 (0.08)
0.34 (0.08)
0.39 (0.10)
0.52 (0.14)
0.60 (0.15)
0.71 (0.20)
Note: results expressed as mean (s.d.) (n ¼ 10).
3. Results
The results of these experiments on fluoride release and
exposure to external fluoride are summarised in Table 2 and
Figs 1–5. The results are presented as either cumulative
fluoride release or as daily release rates both in mg of
fluoride released per gram of cement.
All the materials under test released measurable amounts
of fluoride throughout the period of the study. Fig. 1 shows
the cumulative fluoride release curves for all the groups of
test materials for the first 28 days of the experiment. Fig. 2
shows the linear dependence of the fluoride release from the
materials on the square root of time. R 2 was greater than, or
equal to, 0.97 for all the generated best-fit lines.
The mean cumulative release at day 4 in descending order
was Vitremer . Fuji II LC . Fuji IX . Dyract. Statistically
there was no significant difference between the release of
Fuji II LC and that for Vitremer. However, all the other
differences were statistically significant (p , 0.05) at this
time.
Figs. 3 and 4 show the effect of fluoride exposure on the
daily fluoride release of the specimens. All the materials
included in the test reacted positively to exposure to fluoride
and showed an increase in release for approximately 3 days
after exposure, after which time the levels returned to baseline.
The values for fluoride release in the 24 h postexposure
are shown in Fig. 5. There was no significant difference in
the amount of fluoride released in the 24 h after early (day
28) or late (day 58) exposure for Fuji II LC, Dyract or
Vitremer. However, Fuji IX released significantly less
fluoride after late first exposure. A second exposure to
fluoride caused a significantly higher fluoride release than
that after either early or late first exposure for all the
materials under test.
4. Discussion
The patterns of release for the conventional glassionomer (Fuji IX) and the resin-modified glass-ionomers
(Vitremer and Fuji II LC) with an initial rapid elution
followed by a slow prolonged release are similar to those
reported in other studies [19,20]. In Fig. 2 it can be seen that
Fig. 1. Cumulative fluoride release over first 28 days plotted against the square root of time (all specimens).
594
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
Fig. 2. Calculated best fit lines for the diffusion phase of fluoride release. Note: Non-zero ordinate intercept for Vitremer, Fuji II Lc and Fuji IX indicate an
initial washout phase.
the calculated best fit lines for the diffusion phase of the
release profile of the resin-modified glass-ionomers and
conventional glass-ionomers have a positive intercept with
the y-axis. This indicates an initial washout phase and is
consistent with the description of fluoride release as a
surface washout followed by bulk diffusion as described
by Tay and Braden [21]. The glass-ionomer standard release
profile was not seen for Dyract, which showed no rapid
elution phase. Instead, Dyract showed only the
diffusion-controlled fluoride release (the calculated best fit
line was linear through the origin) similar to that of a
fluoride-releasing composite resin [22]. This is consistent
with what is known about the chemistry of this material.
Although Dyract includes a fluoride containing
acid-degradable glass and an acidic species capable of
reacting with glass, there is no water present in the material
to facilitate the acid–base reaction. If the reaction does
occur, it is due to the diffusion-controlled uptake of water
by the cement from the surroundings. The batch number of
the Dyract used in this study is prefixed by the letter ‘S’,
which indicates that the material used was the reformulated
Dyract, where a fluoride salt has been incorporated into
the formulation. The manufacturer claims that the
refomulated material has a higher fluoride release than
that measured for the original Dyract. Consequently the
fluoride release measured in this study is likely to be predominantly due to the diffusion-controlled release of that
fluoride salt.
Fig. 3. Daily fluoride release rates for specimens exposed twice: After days 28 and 56. Note: Initial washout phase excluded.
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
595
Fig. 4. Daily fluoride release rates for specimens exposed after day 56. Note: Initial washout phase excluded.
The release of the resin-modified glass-ionomers was
always higher than that of the conventional glass-ionomer.
This is consistent with the results of other studies [23,24]
and can be explained in terms of the slower acid–base
reaction in the RMGICs. The presence of the organic
species and subsequent polymeric matrix slows down the
acid–base reaction in the resin-modified glass-ionomer [25]
and consequently the ionic matrix is less mature than a
conventional material of the same age and the measured
fluoride release is higher [26]. The exposure to water of
an immature ionic matrix is potentially hazardous, because
any matrix ions still in a soluble form could be lost.
The conventional material used in this study had a
particularly low measured fluoride release. Fuji IX was
specifically designed for use in the Atraumatic Restorative
Technique (ART), a technique which would require it to be
less moisture-sensitive than other conventional GICs [27].
This could be achieved by using a fast-maturing material or
a glass of low solubility. Low solubility glasses can be
achieved by reducing the alkali metal content of the glass
and the manufacturer of Fuji IX holds a patent covering such
glasses [28]. Alkali metal ions (due to their monovalency)
cannot crosslink the ionic matrix and thus act as disruptions
opening the matrix. A glass containing far fewer monovalent ions would be more closely bound with the di- or
trivalent ions crosslinking the polymer chains holding
Fig. 5. Fluoride release in the 24 h postexposure.
596
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
them close together. It is possible that where the polymer
chains in the matrix are held closely together the water
transport through the matrix required to facilitate fluoride
release would be impaired and therefore reduced.
All the materials under test reacted positively to exposure
to the external fluoride source. This is in contrast to work by
Forsten [29], who indicated that acid-modified composites
would not act as a fluoride sink. However, the experimental
technique used in this study was very different to that used
in Forsten’s study. Forsten stored the specimens in running
water to mimic the replacement of saliva. To measure
fluoride release at a specific time the specimens were
removed from the running water and placed into storage
pots containing a known volume of distilled, deionised
water. Thus this technique does not measure the fluoride
that is immediately washed out of the specimen, but instead
only studies the long-term effects of fluoride exposure.
The exposure did not have any effect on the underlying
fluoride release from the cement, release rates having
returned to baseline within a week of exposure. This
suggests that the fluoride uptake may be more of a surface
rather than a bulk diffusion effect. The dip in the daily
fluoride release rates shortly after exposure, as seen in
Fig. 3, could possibly be an experimental artifact, but
requires further investigation.
It had been expected that the maturity of the cement
would affect its fluoride uptake because as the cement
ages the ionic matrix matures and becomes more crosslinked. A more crosslinked matrix would impede the
required diffusion processes. However, this was only the
case for the conventional glass-ionomer, which showed a
significantly lower release after late first exposure. A closely
crosslinked matrix containing a reduced number of alkali
metal ions would be particularly affected. The lack of
difference between the measured releases after early or
late first exposure for the RMGICs could be an indication
that the presence of the polymeric matrix impedes the
long-term acid–base reaction.
The release data after the second exposure indicates that
fluoride release is enhanced by a repeated exposure. This
implies that the first exposure has modified the cement,
making it more susceptible to external fluoride. The
mechanism by which this occurs is not understood, but
could involve disruption of the cement matrix on the first
exposure, thus opening pathways for fluoride uptake on the
second exposure. Further investigation of this phenomenon
is required. In particular, the effect of repeated exposure to
fluoride over an extended time period on the fluoride release
of these materials should also be determined, since a previous study has indicated that the reservoir effect only
occurs a limited number of times [30]. This study showed
that the measured fluoride release from all the ion-releasing
materials under test was enhanced by exposure to fluoridecontaining toothpaste. With the exception of Fuji IX, that
late exposure, within the timescale of this study, did not
affect the uptake and release of fluoride by the cement.
Also, a second exposure to fluoride enhanced the uptake
and release for all the materials under test.
Acknowledgements
The authors are grateful to DeTrey Dentsply, 3M and the
GC Corporation for supplying the materials used in this
study.
References
[1] Tyas MJ Cariostatic effect of glass ionomer cement: a five-year
clinical study. Aust Dent J, 1991;36:236–239.
[2] Svanberg M Class II amalgam restorations, glass-ionomer tunnel
restorations and caries development on adjacent tooth surfaces: A 3
year clinical study. Caries Res, 1992;26:315–318.
[3] Qvist V, Laurberg L, Poulsen A, Teglers PT. Glass ionomer and
amalgam restorations in primary teeth. 4-4.5 year results. J Dent
Res 1997;76(SI), 41 (Abstract 224).
[4] Forsten L. Fluoride release of glass ionomers. In: Glass Ionomers: The
Next Generation. Proceedings of the 2nd International Symposium on
Glass Ionomers, June 1994, Philadelphia, Pennsylvania, 1974:47-56.
[5] Mjör IA GIC restorations and secondary caries. A preliminary report.
Quintessence Int, 1996;27:171–174.
[6] McLean JW, Wilson AD The clinical development of the glassionomer cement. III The erosion lesion. Aust Dent J,
1977;22:190–195.
[7] Hunt PR A modified class II cavity preparation for glass ionomer
restorative materials. Quintessence Int, 1984;15:1011–1018.
[8] Brandau HE, Ziemiecki TL, Charbeneau GT Restoration of cervical
contours on non-prepared teeth using glass-ionomer cement. A four
and one half year report. J Am Dent Assoc, 1984;108:782–783.
[9] Osborne JW, Berry TG Clinical assessment of glass ionomer cements
as class III restorations: A one-year report. Dent Mater, 1986;2:147–
150.
[10] Ngo H, Earl MSA, Mount W Glass ionomer cements: A 12-month
evaluation. J Prosthet Dent, 1986;55:203–205.
[11] Mount GJ The longevity of glass ionomer cements. J Prosthet Dent,
1986;55:682–685.
[12] Knibbs PJ, Plant CG, Pearson GJ The use of a glass ionomer cement to
restore class Ill cavities. Rest Dent, 1986;2:42–48.
[13] Kidd EAM, Toffenetti F, Mjör IA Secondary caries. Int Dent J,
1992;42:127–138.
[14] Souto M, Donly KJ Caries inhibition of glass ionomers. Am J Dent,
1994;7:122–124.
[15] Forsten L Fluoride release and uptake by glass ionomers. Scand J Dent
Res, 1991;99:241–245.
[16] Hatibovic-Kofman S, Koch G Fluoride release from glass ionomer
cement in vivo and in vitro. Swedish Dent, 1991;15:253–258.
[17] Seppa L, Forss H, Øgaard B The effect of fluoride application on
fluoride release and antibacterial action of glass ionomers. J Dent
Res, 1993;72:1310–1314.
[18] Tyler JE, Poole DFG The rapid measurement of fluoride concentrations in stored human saliva by means of a differential electrode cell.
Arch Oral Biol, 1989;34:995–998.
[19] Forsten L Short and long-term fluoride release from glass ionomer and
other fluoride containing materials in vitro. Scand J Dent Res,
1990;98:179–185.
[20] Musa A, Pearson GJ, Gelbier M In vitro investigation of fluoride ion
release from four resin-modified glass polyalkenoate cements.
Biomaterials, 1996;17:1019–1023.
M. Rothwell et al./Journal of Dentistry 26 (1998) 591–597
[21] Tay WM, Braden M Fluoride ion diffusion from polyalkenoate
(glassionomer) cements. Biomaterials, 1988;9:454–456.
[22] Arends J, van der Zee Y Fluoride uptake in bovine enamel and dentine
from a fluoride releasing composite resin. Quintessence Int.,
1990;21:541–544.
[23] Creanor SL, Carruthers LMC, Saunders WP, Strang R, Foye RH
Fluoride uptake and release. Characteristics of glass ionomer cements.
Caries Research, 1994;28:322–328.
[24] Aboush YEY, Torabzadeh H, Lee AR. One year fluoride release from
fluoride-containing restorative materials. J Dent Res 1995:74, 881
(abstract 478).
[25] Anstice HM, Nicholson JW Studies on the setting of polyelectrolyte
cements, part 2: The effect of organic compounds on a glass
poly(alkenoate) cement. J Mater Sci, Materials in Medicine,
1994;5:299–302.
597
[26] Davies EH, Sefton J, Wilson AD Preliminary study of factors affecting the fluoride release from glass ionomer cements. Biomaterials,
1993;14:636–639.
[27] Frencken JE, Songpaisan Y, Phantumvanit P, Pilot T A atraumatice
restorative treatment (ART) technique; evaluation after one year. Int
Dent J, 1994;44:460–464.
[28] Akahane S, Tosaki S, Hirota K, Tomioka K. Glass powder for dental
glass ionomer cements. UK Patent 2 202 221 B.
[29] Forsten L Resin-modified glass ionomer cements: fluoride release and
uptake. Acta Odontol Scand, 1995;53:222–225.
[30] Williams JA, The glass ionomer: Infinite source and reservoir,
sponge or black hole? In: The Abstract Book of the 1st European
Union Conference on Glass-Ionomers. London, 1996. ISBN
0 9528232 0 9.