an in vitro investigation of a poly(vinyl phosphonic acid) based cement

An in vitro investigation of a poly(vinyl phosphonic acid) based cementwith four conventional glass-ionomer cements. Part 1: flexural strength

and fluoride release

V.H.W. Khouw-Liu, H.M. Anstice, G.J. Pearson*

Departments of Paediatric Dentistry and Biomaterials Science, Eastman Dental Institute, University of London, 256 Gray’s Inn Road, London WC1X 8LD, UK

Received 10 March 1998

Abstract

Objective: To investigate the flexural strength and fluoride release of four conventional glass-ionomer cements: Ketac-Molar (KM), HiFi(HF), Vivaglass Fil (VF), Ketac-Fil (KF) and a newly developed glass polyphosphonate cement, Diamond Carve (DC).

Method: Disc specimens (10 mm diameter, 1 mm thick) were prepared and mould stored at 378C. After one hour, the specimens wereremoved from their mould and immersed in 20 ml of deionised water until required for testing. Biaxial flexural strength was determined at 1hour and at 1, 7, 30 and 90 days after the start of mixing. Measurements of fluoride release from the specimens were carried out at 2 hours andat 1, 3, 7, 14, 30, 60 and 90 days after the start of mixing using a fluoride ion selective electrode. The results were analysed using ANOVA andstudent ‘t’ tests.

Results: All the materials displayed different flexural strength patterns. KM and DC became stronger whilst KF and VF plateaued instrength with time. HF peaked in strength and then became weaker. At 90 days, the mean flexural strengths in decreasing order was asfollows: KM $ VF $ DC $ HF . KF. An initial fast rate of fluoride release followed by a slower but steady release of fluoride was observedin each of the materials. The mean cumulative fluoride release in decreasing order was as follows: VF. KF $ HF . DC . KM. VF releasedsignificantly higher level and KM significantly lower level of fluoride than the other materials.

Conclusions: The acid used to form the cement could not be used to predict changes in cement strength behaviour with respect to time. DCincreased in strength with time and its flexural strength at 90 days was comparable to that of HF and VF. The cumulative and rate of fluoriderelease varied for the materials. DC had a low fluoride release consistent with a fast setting material with good early resistance to water.q 1999 Elsevier Science Ltd. All rights reserved.

Keywords:Glass-ionomer cements; Glass polyalkenoate cements; Strength; Fluoride

1. Introduction

First introduced in the early 1970s by Wilson and Kent[1], glass-ionomer cement (GIC) is now established as animportant restorative material. Its inherent qualities of adhe-sion to tooth structure and anticariogenic properties throughits ability to release and uptake fluoride have encouragedmuch research and development over the years [2,3]. Likeall dental materials, glass-ionomer cement has certain draw-backs; chiefly its susceptibility to water and its low initialstrength [4,5]. Since the development of the original glass-ionomer cement, efforts have been made to improve theseproperties and changes have been made in both the glass

powder component and the polycarboxylic acid. As a result,there are significant differences in the composition andproperties of commercial materials for the various applica-tions.

Glass-ionomer cements are acid–base reaction cements[6] whereby the basic component is an acid-degradableglass and the acidic component, an aqueous solution of apolymeric acid. Setting of the cement occurs by slowneutralisation of the acidic polymer leading to the formationof a polysalt matrix. This takes place through 3 stages thatoverlap each other: dissolution, gelation and hardening. Thenature of the GIC setting reaction causes the strength of thematerial to develop with time. Glass-ionomers reach aninitial peak strength by 24 hours after mixing suggestingthat the first 2 stages of the setting reaction (dissolutionand gelation phase) have already occurred [7–9]. Anyfurther increase in physical properties is thought to be due

Journal of Dentistry 27 (1999) 351–357

Journalof

Dentistry

0300-5712/99/$ - see front matterq 1999 Elsevier Science Ltd. All rights reserved.PII: S0300-5712(98)00061-X

* Corresponding author. Tel.:1 44-171-915-1019; fax:1 44-171-915-1019.

E-mail address:[email protected] (G.J. Pearson)

to an increase in the number of ionic cross-links during thehardening stage [4] and more recently has been attributed tothe slow build-up of the silica matrix during the settingprocess [10].

Several groups investigating the mechanical behaviour ofGICs, have found that the strength of some materials basedon copolymers of acrylic acid deteriorates with time [9,11].A study by Cattani-Lorente and co-workers reported obser-ving a similar deterioration in strength but this onlyoccurred in a few materials and was not limited to copoly-mer based cements [12]. Cattani-Lorente et al. attributed theweakening to the plasticising effect of water on these mate-rials. However, in a recent study, the loss of strength wasreported to be as a consequence of the high cross-linkdensity found in these copolymers and not due to hydrolysisas previously suggested [13].

Another important property of the GIC is its long-termfluoride release and its ability to take up fluoride from anexternal source and subsequently release it [14,15]. Theaction of this fluoride clinically is a matter of some debate,but it is thought by many to confer a cariostatic action. Forexample in a recent study, when glass-ionomer and amal-gam restorations were compared, significantly fewer lesionsof secondary caries were found in teeth restored with GICs[16].

In the last five years, several new conventional glass-ionomer cements (e.g. HiFi, Vivaglass Fil, Ketac-Molar)have been introduced. In addition, one glass polyphospho-nate cement has been developed and has recently becomecommercially available (Diamond Carve). This cement isbased on poly(vinyl phosphonic acid), PVPA, which is thephosphorus analogue to polyacrylic acid, PAA. PVPA,being a stronger acid than PAA, is more reactive and isthought to form cements that are hydrolytically more stable[17,18] due to the rapidity of the binding of the cations to theanions. To date, no studies have been undertaken tocompare the properties of these newer GICs. This studytherefore aimed to investigate and compare some of theproperties, in this instance strength and fluoride release, ofthe more recently developed materials including the newpolyphosphonate cement.

2. Materials and methods

The materials used, including details of their composi-tion, P/L ratio, manufacturer and batch number are listedin Table 1.

For biaxial flexural strength measurements, disc specimenswere prepared using brass split rings (10 mm diameter and1 mm thick). The materials were mixed according to themanufacturer’s instructions. For the hand-mixed materials(HiFi and Diamond Carve), the powder and liquid wereweighed out to^ 0.001 g prior to mixing using a stainlesssteel spatula on a glass slab. To form the disc, the cement pastewas packed into a ring placed on a microscope slide. A secondslide was placed over the ring and light hand pressure appliedto enable the extra material to flow out of the ring through theslit. The mould assembly was then placed in a thermostaticallycontrolled oven at 378C^ 18C. After one hour, the specimenswere removed from their moulds and then immersed in 20 mlof deionised water in individual sealed plastic containersbefore being placed in the oven again at 378C until requiredfor testing.

The biaxial flexural strength of the materials was deter-mined at the following time intervals: 1 hour, 1 day, 7 days,30 days and 90 days after the start of mixing. At least sixspecimens of each material were tested at each time interval.Prior to testing, the thickness of a specimen was measured at4 points across the disc. The disc was then placed centrallyon a knife edged annulus 8 mm in diameter and loaded at arate of 1 mm/min using a ball ended indentor (4 mmdiameter) in a universal load testing machine (Hounsfield,UK) with a 200 N load cell. The force at fracture of eachspecimen was noted and the biaxial flexural strength (FS) inMPa calculated using the formula [19]:

FS� P

h2 �0:606 logeah

1 1:13� �1�where:

a is the radius of the annular (mm)h is the average thickness of specimen (mm)P is the load at fracture (N).1.13 is a constant derived from Poisson’s ratio [20].

V.H.W. Khouw-Liu et al. / Journal of Dentistry 27 (1999) 351–357352

Table 1Materials used in this study

Material Manufacturer Batch number P/L ratio[w/w] Type

Diamond Carve Associated Dental Products Ltd L:02/97 P:R985/1 3.8:1 Glass polyphosphonatecement (hand mixed)

HiFi Shofu Inc. P:069523-1 L:109427-1 4.5:1 Glass polyalkenoatecement (hand mixed)

Vivaglass Fil Ivoclar-Vivadent 808131 1:1 V:V Glass polyalkenoatecement (capsulated)

Ketac-Fil ESPE 026 12704 3.2:1 Glass polyalkenoatecement (capsulated)

Ketal-Molar ESPE 01832039 3.5:1 Glass polyalkenoatecement (capsulated)

The disc specimens used for measuring fluoride releasewere prepared using the same technique. After the 1 hourstorage period was complete, the samples were removedfrom their moulds and then weighed to an accuracy of^0.001 g. Each specimen was then stored in 20 ml ofdeionised water in a sealed plastic container at 378C. Themeasurement of fluoride release from the samples wascarried out at the following time intervals: 2 hours, 1 day,3 days, 7 days, 14 days, 30 days, 60 days and 90 days afterthe start of mixing. 7 specimens of each material wereevaluated.

At each test interval, the specimen disc was removedfrom the solution and transferred to a new container with20 ml of fresh deionised water. The new container was thenreplaced in the oven until the next test interval. The concen-tration of fluoride in the storage water from the originalcontainer was measured using a fluoride ion selective elec-trode, calibrated using standard solutions of 1, 10 and100 ppm fluoride. Recalibrations were performed every2–3 hours to compensate for any variation caused bychanges in the temperature and humidity. To measure

fluoride concentration, 0.5 ml of the sample was mixedwith 1.5 ml of 0.1 M HCl in order to decomplex anybound fluoride that might be present in the solution. Oncethoroughly mixed, the concentration of fluoride in ppm ofthe sample solution was measured. Between all readings, theelectrodes were thoroughly washed and dried before beingplaced in the next test solution.

Statistical analysis of all the results was carried out usingANOVA and student t tests.

3. Results

The results of the flexural strength measurements for allmaterials are summarised in Table 2 and Fig. 1. They showthe mean values of flexural strength (and standard deviation)in MPa for the different test materials at the different testintervals.

As can be seen from the results, there was no single trendin the change of measured strength with time. HiFiincreased in strength up to day 30 but from day 30 to day90 showed a significant decrease in flexural strength.Diamond Carve increased its strength with time. The great-est increase in flexural strength for this material occurredfrom 1 hour to 1 day after which the rate of increase wasmore gradual. The difference in flexural strength from 1hour and day 90 was highly significant but that from day30 to day 90 was not. Vivaglass Fil reached its maximumflexural strength earlier at around day 7. It was interesting tonote that the material did not show any significant change instrength from day 1 to day 90. Although there was a slightdecrease in flexural strength between day 7 and day 90, thiswas not statistically significant. Ketac-Fil demonstrated a

V.H.W. Khouw-Liu et al. / Journal of Dentistry 27 (1999) 351–357 353

Table 2Flexural strength of each material over the test period (MPa). Resultsexpressed as mean (standard deviation)n $ 6

Day HiFi Diamond Carve Vivaglass Fil Ketac-Fil Ketac-Molar

1/24 31.7(3.6) 27.7(3.1) 39.6(5.4) 26.6(3.0) 34.5(10.5)1 40.4(6.7) 36.5(2.9) 44.3(5.2) 33.0(3.6) 44.7(6.0)7 40.8(3.2) 38.7(5.0) 51.7(3.0) 37.0(2.6) 46.5(8.5)30 49.0(2.9) 43.3(10.6) 48.6(6.0) 36.5(3.7) 46.3(4.1)90 43.3(4.3) 44.8(4.4) 48.9(7.0) 37.4(5.3) 54.3(4.4)

Fig. 1. Comparison of mean flexural strengths of each material at different time intervals. Note: Error bar is^ 1 standard deviation.

rapid increase in strength initially. The flexural strengthreached its maximum measured strength at day 7 afterwhich there was no change in strength for the remainderof the experiment. Although Ketac-Molar showed a rapidearly increase in flexural strength from 1 hour to 1 day, thematerial also increased in strength throughout the testperiod, the increase in strength between day 30 and day90 was highly significant.

When the measured strengths of the materials at 1 hour werecompared, Vivaglass Fil was found to be significantly strongerthan Diamond Carve, Ketac-Fil and HiFi although there wasno statistically significant difference in strength between Viva-glass Fil and Ketac-Molar. However, Ketac-Molar had a highstandard deviation representing 30% of its mean value. By day30, the properties of the materials had changed such that the

flexural strengths of Vivaglass Fil, Ketac-Molar and HiFi werenot statistically different. Ketac Fil was significantly weakerthan all other materials except Diamond Carve although itshould be noted that Diamond Carve had a high standarddeviation (25% of the mean value) at that time interval.After 90 daysstorage, Ketac-Molar was the strongest material;Vivaglass Fil, HiFi and Diamond Carve were all similar instrength and Ketac-Fil remained significantly weaker thanall the other materials.

The cumulative fluoride release results (mg F/g cement)from the 5 glass-ionomer cements are summarised in Table3 and Fig. 2.

All the materials showed a measurable fluoride releasewith time. Two phases of fluoride release were observed ineach of the material, an initial rapid fluoride washout


Table 3Cumulative fluoride release of each material over the test period (mg F/g cement). Results expressed as mean (standard deviation)n � 7

Day HiFi Diamond Carve Vivaglass Fil Ketac-Fil Ketac-Molar

2.24 0.10(0.03) 0.03(0.01) 0.25(0.3) 0.09(0.01) 0.01(0.00)1 0.55(0.18) 0.16(0.05) 1.22(0.22) 0.55(0.04) 0.09(0.01)3 0.71(0.25) 0.24(0.08) 1.60(0.29) 0.88(0.07) 0.15(0.02)7 0.93(0.32) 0.34(0.12) 2.03(0.36) 1.21(0.10) 0.22(0.02)14 1.16(0.39) 0.46(0.16) 2.48(0.42) 1.57(0.13) 0.29(0.03)30 1.50(0.49) 0.65(0.19) 3.16(0.54) 2.05(0.16) 0.39(0.03)60 1.96(0.62) 0.88(0.24) 4.03(0.61) 2.61(0.21) 0.51(0.05)90 2.27(0.68) 1.05(0.27) 4.69(0.67) 3.04(0.24) 0.61(0.07)

Fig. 2. Mean cumulative fluoride release. Note: Error bar is^ 1 standard deviation.

followed by a slower steady elution of fluoride that contin-ued throughout the experiment.

Vivaglass Fil released the highest amount of fluoridethroughout the experimental period. It had the highest initialburst, released during the first 24 hours and the highestensuing rate of release. The initial rates of fluoride releasefor Ketac-Fil and HiFi were similar but significantly lessthan those for Vivaglass Fil. For HiFi, the second phase offluoride release had started by day 3 whereas for Ketac-Fil,the initial non-linear phase lasted until about day 14. Thesubsequent release rates of these two materials appeared tobe similar (Fig. 2). Diamond Carve and Ketac-Molardemonstrated similar behaviour with initial releases andthe steady state of fluoride releases substantially lowerthan those observed for the other materials.

At all time intervals, the mean cumulative fluoride releasein decreasing order was as follows:

Vivaglass Fil . Ketac-Fil $ HiFi . DiamondCarve . Ketac-Molar

At all times, Vivaglass Fil showed a significantly higherlevel of fluoride release than all other materials. The releasefrom Ketac-Molar was significantly lower than all othermaterials. At 90 days, the cumulative fluoride releasesfrom all the materials were highly significantly differentfrom one another.

4. Discussion

The flexural strengths of the materials were tested and thechanges in strength after ageing in water at simulated oraltemperature over 90 days were monitored. The resultsobtained showed that both the flexural strengths and theirvariation with time were different for the selection of mate-rials tested.

An appreciable increase in strength over the first 24 hourswas seen with all the materials tested. This has beenobserved in previous studies [4,9] and was attributed tothe completion of the first two stages of the setting reaction(dissolution and gelation). All the materials showed a subse-quent increase in strength albeit at different rates up to day7. It was after this time that they behaved differently. Theincrease in mechanical properties with time has been attrib-uted to the maturation of the cement matrix due to additionalcross-linking resulting from the formation of new calciumand aluminium polyacrylates [4]. However, as this post-hardening reaction continues, the matrix becomes morerigid and the free carboxylic groups therefore become lessavailable to form ionic bonds and the rate of reaction slowsdown. More recently, Wasson and Nicholson have attribu-ted the additional strengthening of the materials to the slowbuild-up of a silica matrix resulting from the acid degrada-tion of the aluminosilicate glass [10].

In this study, the Ketac-Molar specimens were signifi-cantly stronger than those prepared from Ketac-Fil. In fact

the values reported for Ketac Molar are higher than thoseprovided by the manufacturer in their data sheet.

This is probably due to the reduced glass particle size andincreased powder to liquid (P/L) ratio in Ketac-Molar. It isclaimed that 50% of the glass particles found in Ketac-Molar are 2.8mm or smaller; 90% of the particles have adiameter of less than 9.6mm [21]. In Ketac-Molar, some ofthe polymeric acid is incorporated in a dried form into thepowder. This permits higher molecular weight polymers tobe used, resulting in improved mechanical property andincreased cross-linkage. Both Ketac-Fil and Ketac-Molarare based on a copolymer of acrylic and maleic acid, butneither demonstrated any deterioration in flexural strengthas described by some workers previously [9,11].

HiFi, which is based on polyacrylic acid, became progres-sively stronger during the first 30 days but thereafter dete-riorated significantly in strength. Although this behaviourwhen first observed was attributed to cements based oncopolymers, Cattani-Lorente and co-workers also reporteda deterioration in strength of some homopolymer glass-ionomer materials after long-term storage in water [12].They attributed the weakening of these materials withtime to erosion and the plasticizing effect of water althoughthe cement could also have become more brittle and hencesensitive to flaws due to high crosslink density as describedby Nicholson and Abiden [13].

Vivaglass Fil had the highest initial strength and thematerial achieved its maximum strength early at day 7.This observation was rather surprising considering its softerconsistency when mixed in comparison to the other materi-als. One possible explanation could be the addition of driedpolyacrylic acid with a higher molecular weight to thepowder. This may have enhanced the flexural strength [6].

Diamond Carve had low flexural strength initially but itbecame stronger with time. The greatest increase in strengthtook place in the first 24 hours after which, the increase wasmore gradual. By 90 days, the cement was highly signifi-cantly stronger than it was at one hour and its flexuralstrength was comparable to that of HiFi and Vivaglass Fil.This slow build-up of strength was consistent with the find-ings of Anstice and Nicholson [22]. It is surprising becausePVPA is a stronger acid than PAA and would therefore reactmuch faster initially causing the bulk of the ionic matrix tobe laid down in the first 24 hours. A possible reason for thisgradual increase in strength could lie in the glass formula-tion—the use of less reactive glasses. The subsequentincrease in strength indicated a post hardening reactionwithin the set cement. The short working time andhigh P/L ratio used with Diamond Carve made mixingdifficult. This can potentially lead to specimens withmore flaws or inhomogeneities. This would explain thewide scatter of results seen at day 30. Anstice andNicholson also found that PVPA-based materials werevery sensitive to desiccation [22] and if subjected to adesiccating environment, a partially hydrated matrixwould occur resulting in an inferior cement.


Again a comparison with the manufacturers data showsthat these results at 24 hours are initially similar. However atall subsequent comparable test time the manufacturer’sresults are substantially higher. In fact there is a differenceof between 10 and 27% at the later stages (Table 4). It isinteresting to note that the manufacturer’s scatter of resultsis much greater than reported in this study.

All the materials released measurable quantities of fluor-ide during the experimental period. However, there was alarge variation in the rate and amount of fluoride release.The greatest amounts of fluoride were released during thefirst days after which the release rate fell but continued at aconstant rate until the end of the experiment. This pattern ofrelease was similar to those described previously [23,24].The rapid initial phase of fluoride elution is probably due tosurface wash off and the slower prolonged phase due todiffusion of fluoride from the bulk of the material [25].The initial high fluoride release rate may have a bacterio-static role in caries prevention whilst the continuous steadyphase may be important in the ionic exchange that occurs inthe demineralisation-mineralisation process of tooth decay[26].

The ranking of the different cements according to theamount and rate of release remained unchanged throughoutthe experimental period. However, a highly significantdifference in the amount of fluoride released from thevarious cements was observed at the end of 90 days.Other workers have also shown that fluoride releaseamong various glass-ionomer cements could be signifi-cantly different [27]. This could be attributed to the manyvariables that govern the release of fluoride such as the typeof glass/polyacid used, its composition and proportion or thesize of the glass particles used [14,28], its P/L ratio [29],mixing time and porosity [30].

Vivaglass Fil released the largest amount of fluoridethroughout the experiment. One may speculate that thisexceptionally high fluoride release could have resultedfrom the incorporation of the opacifying agent, Ytterbiumtrifluoride (YbF3) into the formulation. YbF3 has previouslybeen incorporated in some composite resins not only toconfer radiopacity to the material but also to provide somemeasure of fluoride release. Previous high releases havebeen attributed to material degradation but this seemsunlikely given the maintenance of strength from 30 to 90days.

It has been suggested that the rate of fluoride release isdependent on the maturity of the cement matrix when firstimmersed in water [31]. Davies et al. explained that whenimmature cements are immersed in water, they are perma-nently weakened, the surface area is increased by erosionand roughening which brings about a higher level of fluoriderelease [31].

Vivaglass Fil and HiFi had relatively high initial rates offluoride release which were similar to that of Ketac-Fil.Both materials were found to set very slowly. This supportsthe hypothesis that high fluoride release during the initialphase may be due to the immaturity of the cement when itwas first immersed in water.

Ketac-Fil showed a long non-linear phase of releaselasting until day 14 suggesting a significant surfacewashout phase. Despite a short setting time, Ketac-Filhad a prolonged high initial rate of fluoride release.This was not consistent with findings from other studies[32–33]. Ketac-Molar released considerably less fluoridecompared to all other materials. This may be explainedby its low solubility. The manufacturers had reported a24-hour total solubility of 0.05% (Ketac-Fil 0.2%) whensamples of Ketac-Molar were placed in water after anhour [21]. Furthermore, Ketac-Molar had a higher P/Lratio (3.5:1) and a longer mixing time (15 seconds)when compared to Ketac-Fil (P/L ratio of 3.2:1 andmixing time: 10 seconds). This may explain the differ-ence in their patterns of fluoride release despite beingmade by the same manufacturer.

Diamond Carve released a relatively low level offluoride throughout the experiment. This can again beexplained by its rapid setting nature and its low solubil-ity as well as its stiff consistency that was related to itsvery high P/L ratio.

5. Conclusions

Only Ketac-Molar and Diamond Carve showed a contin-uous increase in strength over the 90 days. The other mate-rials showed either a plateauing effect (Ketac-Fil, VivaglassFil) or a decrease in strength (HiFi) after a certain period oftime. The acids used did not predict the behaviour patternsof the materials.

The pattern of fluoride release supported the two-phasediffusion theory postulated by previous authors, an earlyrapid surface elution followed by bulk diffusion from thecore of the cement. The cumulative fluoride release and therate of fluoride release varied for the materials. Vivaglass Filreleased the highest amount of fluoride throughout theexperimental period. Conversely, Ketac-Molar releasedsignificantly lower amounts of fluoride than the rest of thematerials at all time intervals. Diamond Carve, based onpoly(vinyl phosphonic acid), had a low fluoride releaseconsistent with a fast setting material with good earlyresistance to water.


Table 4Comparison of results for Biaxial Flexure Strength of Diamond Carve usingstudy and manufacturers results Manufacturer’s results from technical datasheet provided by Associated Dental Products

Manufacturers Results Current study results % difference

1 day 34.5[5.8] 36.5[3.1] 1 57 days 43.3[6.2] 38.7[5.0] 2 101 month 59.7[16.2] 43.3[10.6] 2 273 months 52.2[9.0] 44.8[4.4] 2 15

Acknowledgements

The authors would like to thank the manufacturers: ESPE(Germany), Shofu Inc. (UK), Ivoclar (Lichtenstein) andAssociated Dental Products Ltd (UK) for supplying thematerials used in this study.

References

[1] Wilson AD, Kent BE. A translucent cement for dentistry: the glass-ionomer cement. British Dental Journal 1972;132:133–135.

[2] Wilson AD. Developments in glass-ionomer cements. The Interna-tional Journal of Prosthodontics 1989;2:438–446.

[3] Smith DC. Evolution and development of glass ionomer cementsystems. First European Union Conference on Glass-ionomers.Biomaterials 1998;19:467–478.

[4] Crisp S, Lewis BG, Wilson AD. Characterisation of glass-ionomercements. 1. Long term hardness and compressive strength. Journal ofDentistry 1976;4:162–166.

[5] Mount GJ, Makinson OF. Glass-ionomer restorative cements: clinicalimplications of the setting reaction. Operative Dentistry 1982;7:134–141.

[6] Wilson AD, McLean JW. In: Glass-ionomer cement. Chicago, IL:Quintessence, 1988.

[7] Crisp S, Wilson AD. Reactions in glass-ionomer cements. I. Decom-position of the powder. Journal of Dental Research 1974;53:1408–1413.

[8] Crisp S, Wilson AD. Reactions in glass-ionomer cements. III Theprecipitation reaction. Journal of Dental Research 1974;53:1420–1424.

[9] Pearson GJ, Atkinson AS. Long-term flexural strength of glass iono-mer cements. Biomaterials 1991;12:658–660.

[10] Wasson EA, Nicholson JW. New aspects of the setting of glass-iono-mer cements. Journal of Dental Research 1992;72:481–483.

[11] Williams JA, Billington RW. Changes in compressive strength ofglass ionomer restorative materials with respect to time periods of24 hours to 4 months. Journal of Oral Rehabilitation 1991;18:163–168.

[12] Cattani-Lorente MA, Godin C, Meyer JM. Mechanical behaviour ofglass ionomer cements affected by long-term storage in water. DentalMaterials 1994;10:37–44.

[13] Nicholson JW, Abiden F. Changes in compressive strength on ageingin glass polyalkenoate (glass-ionomer) cements prepared fromacrylic/maleic acid copolymers. Biomaterials 1997;18:59–62.

[14] Forsten L. Short- and long-term fluoride release from glass-ionomersand other fluoride-containing filling materials in vitro. ScandinavianJournal of Dental Research 1990;98:179–185.

[15] Koch G, Hatibovic-Kofman S. Glass ionomer cements as a fluoriderelease system in vivo. Swedish Dental Journal 1990;14:267–273.

[16] Qvist V, Laurberg L, Poulsen A, Teglers PT. Longevity and cario-static effects of every day conventional glass-ionomer and amalgamrestorations in primary teeth: Three-year results. Journal of DentalResearch 1997;76:1387–1396.

[17] Wilson AD, Ellis J. Poly-vinylphosphonic acid and metal oxide orcermet or glass-ionomer cements. UK Patent Application No: 2, 219,289A, 1989.

[18] Ellis J, Wilson AD. Polyphosphate cements: A new class of dentalcements. Journal of Materials Science Letters 1990;9:1058–1060.

[19] Timoshenko S, Woinowsky-Krieger S. In: Theory of plates and shells,2nd ed. Singapore: McGraw-Hill, 1959.

[20] Akinmade AO, Nicholson JW. Poisson’s ratio of glass polyalkenoate(glass-ionomer) cements determined by an ultrasonic method. Journalof Materials Science: Materials in Medicine 1995;6:483–485.

[21] Product Literature from ESPE supplied for Ketac-Molar. Ketac-Molar Aplicap: Remarkable plasticity! 1996.

[22] Anstice HM, Nicholson JW. Investigation of the post-hardening reac-tion in glass-ionomer cements based on poly(vinyl phosphonic acid).Journal of Materials Science: Materials in Medicine 1995;6:420–425.

[23] Forsten L. Fluoride release and uptake by glass ionomers. Scandina-vian Journal of Dental Research 1991;99:241–245.

[24] Miller BH, Komatsu H, Nakajima H, Okabe T. Effect of glass iono-mer manipulation on early fluoride release. American Journal ofDentistry 1995;8:82–186.

[25] Tay WM, Braden M. Fluoride ion diffusion from polyalkenoate(glass-ionomer) cements. Biomaterials 1988;9:454–456.

[26] Norman RD, Mehra RV, Swartz ML, Phillips RW. Effects of restora-tive materials on plaque composition. Journal of Dental Research1972;51:1596–1601.

[27] Meryon SD, Smith AJ. A comparison of fluoride release from threeglass-ionomer cements and a polycarboxylate cement. InternationalEndodontic Journal 1984;17:16–24.

[28] De Maeyer EAP, De Moor RJG, Verbeeck RMH. Water sensitivity ofthe fluoride release of conventional restorative glass-ionomer cement.First European Union Conference on Glass-ionomers. Warwick, UK,May 1996:65–66 (Abstract).

[29] Perrin C, Persin M, Sarrazin J. A comparison of fluoride from fourglass-ionomer cements. Quintessence International 1994;25:603–608.

[30] Smith DC. Composition and characteristics of glass-ionomercements. Journal of the American Dental Association 1990;120:20–22.

[31] Davies EH, Sefton J, Wilson AD. Preliminary study of factors affect-ing the fluoride release from glass-ionomer cements. Biomaterials1993;14:636–639.

[32] DeSchepper EJ, Berry III EA, Cailleteau JG, Tate WH. A compara-tive study of fluoride release from glass-ionomer cements. Quintes-sence International 1991;22:215–220.

[33] Diaz-Arnold AM, Holmes DC, Wilstrom DW, Swift EJ. Short-termfluoride release/uptake of glass ionomer restoratives. Dental Materials1995;11:96–101.


an in vitro investigation of a poly(vinyl phosphonic acid) based cement

Documents