Biomaterials 24 (2003) 2901–2907

Reactive fibre reinforced glass ionomer cements

.
Ulrich Lohbauera,*, Jurgen Walkerb,c, Sergej Nikolaenkoa, Jan Wernerc, Alexis Clareb,
Anselm Petschelta, Peter Greilc
a
Policlinic for Operative Dentistry and Periodontology, University of Erlangen-Nuremberg, Glueckstrasse 11, Erlangen 91054, Germany
b
New York State College of Ceramics, Alfred University, One Saxon Drive, Alfred, NY 14802, USA
c
Department of Materials Science and Engineering (Glass and Ceramics), University of Erlangen-Nuremberg, Martensstrasse 5,
Erlangen 91058, Germany
Received 14 October 2002; accepted 20 February 2003

Abstract

The mechanical properties of glass ionomer cements used in restorative dentistry reinforced by chopped glass fibres were
investigated. Reactive glass fibres with a composition in the system SiO2–Al2O3–CaF2–Na3AlF6 and a thickness of 26 mm were
drawn by a bushing process. The manufacturing parameters were optimized with respect to maximum strength of the glass fibre
reinforced ionomer cements. Powder to liquid ratio, pre-treatment of the glass, grain size distribution and fibre volume fraction were
varied. Glass fibre and cement were characterized by X-ray diffraction, transmission electron microscopy and energy dispersive
spectroscopy techniques, respectively. The highest flexural strength of the reinforced cement (15.6 MPa) was found by compounding
20 vol% reactive fibres and extending the initial dry gelation period up to 30 min. Microscopic examination of the fractured cements
indicated a distinct reactive layer at the fibre surface. A pronounced fibre pull out mode gives rise to an additional work-of-fracture
contributed by pulling the fibres out of the fracture surface.
r 2003 Elsevier Science Ltd. All rights reserved.

Keywords: Glass ionomer cement; Glass fibres; Reinforcement; Fibre drawing; Mechanical strength

1. Introduction dry stage, the specimens have to be stored in water to

impede dehydration [1,2].
In the late 1970s the glass ionomer cements (GICs) GIC has several unique advantages among restorative
were developed as an outgrowth from the research into materials. The content of fluorine plays an important
dental silicate cements and zinc polycarboxylate cements role [6–8]. Fluorine disrupts the glass network and
[1,2]. They consist of a sodium aluminium calcium (or lowers the fusion temperature of the glass melt, increases
strontium) silicate glass powder and an aqueous the mechanical strength of the set cement and increases
polyacrylic or related polymeric acid [3]. The single the susceptibility of the glass to acid attack. It allows the
compounds react as soon as they are mixed together. matching of the refractive index to that of the tooth for
The polymeric acid attacks the glass and leaches out the translucency of the cement. The presence of fluorine
mainly calcium and aluminium ions. The metal ions in as a crystalline phase in the glass has been observed by
turn catalyse cross linking the polymer chains and cause many authors [9,10]. They reported the presence of a
the cement to set. GICs set with the formation of a second droplet phase that is rich in calcium and fluoride
siliceous hydrogel as a result of this acid–base reaction and assumed the existence of CaF2 crystals in the
[4,5]. During a first phase after mixing a dry storage of separated phase [11]. However, the main advantage is a
the material has to be ensured in order to provide the constant fluoride release during the lifetime of the
gelation process from water dilution. After that initial restoration which makes it cariostatic i.e. prevent from
secondary caries. Many studies have shown that the
*Corresponding author. Tel.: +49-9131-853-4236; fax: +49-9131-
fluoride release of the cements can remineralize tooth
853-3603. tissue. This allows a minimal invasive cavity preparation
E-mail address: [email protected] (U. Lohbauer). technique, preserving more sound tissue [1,12,13].

0142-9612/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0142-9612(03)00130-3
2902 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907

Compared to other restorative materials like amalgam powders were measured using light scattering with a
or polymer composites, however, GICs suffer from laser (Cilass 715, Cilas Corp., Marcoussis, France), and
lower strength, wear resistance and fracture toughness the surface area was determined by standard nitrogen
[3,14–16]. Reinforcement of GIC has therefore become a adsorption method (BET, ASAPs 2000 Nitrogen,
matter of increasing research efforts. In former experi- Micromeritics Corp., Norcross, USA). The crystalline
ments various alumina, zirconia, silicon carbide, carbon phase composition was determined by X-ray diffraction
fibres, or metal particles were used with encouraging using monochromated CuKa radiation (D 5000s, Sie-
results [2]. Fibre length and volume fraction are thereby mens Corp., Mannheim, Germany). The microstructure
key microstructural parameters determining the result- of the quenched glass and the fibre was examined in an
ing cement properties [17]. A strengthening effect, due to analytical transmission electron microscope (Philips CM
fibre pull-out or crack bridging during fracture process 30 T, Philips, USA) and by energy dispersive spectro-
is mainly influenced by the matrix–fibre interface scopy (EDS). The TEM work was performed using
coupling [18]. Fibre reinforcement with reactive glass 300 KV acceleration voltage. The specimens were pre-
fibres is one attempt to achieve proper fibre embedding pared by focused ion beam (FIB) technique.
in the matrix cement. Short glass fibre reinforcement Thirty grams of the glass frit were melted in a
was studied by Kawano et al. [19] and Kobayashi et al. platinum crucible with a hole in the bottom. The glass
[20] with a glass composition in the system CaO–P2O5– melt flowing through the hole with a diameter of 1.5 mm
SiO2–Al2O3. They used phosphate glasses, since im- was pulled into a fibre and, after rapid air cooling-
proved self-adhesion to human dentine is described [21]. wound onto a rotating drum with a diameter of 150 mm.
The aim of the present work is to incorporate high The rate of mass flow of the glass melt out of the
strength reactive glass fibres into the GIC matrix. bushing is dependent on geometrical factors of the
Depending on the interfacial coupling of fibre-to-matrix, bushing and the viscosity of the glass [26,27]. The two
theoretical models predict a significant increase of the parameters that were varied to adjust an optimum
work-of-fracture when continuous fibres are aligned thickness of the chopped fibre were temperature and
with the primary loading direction so that effective load drawing speed. Finally a temperature of 1120
C and a
transfer from the matrix to the fibre can be induced. In drawing speed of 3 m/s were selected. The glass frit was
the case of short fibres dispersed with random orienta- heated up to 1300
C before drawing and maintained
tion in the matrix the situation is more complex [22,23]. there for 1 h in order to homogenize the melt.
Fibres below a critical length (which is determined by The drawn glass fibre was cut into short fibres by a
the ratio of the fibre strength to the shear strength of the rotating guillotine. The critical fibre length 2xc ; which is
matrix–fibre interface) are less effective in reinforcement supposed to provide maximum load transfer by means
whereas using longer fibres nothing is gained by the of elastic strain and frictional pull out, was calculated
extra length [22,24]. For the system under investigation by [22]
a critical fibre length of approximately 550 mm was d sf
estimated from fibre and matrix properties. In this 2xc ¼ ; ð1Þ
4 sm
investigation, a model glass frit of the system SiO2–
Al2O3–CaF2–Na3AlF6 was used to prepare glass fibres where d is the fibre diameter, sf the fibre strength and
by a bushing process. The study deals with fluoride sm the shear strength at the matrix–fibre interface.
glasses, since an increased fluorine content was shown to Ashby et al. reported a correlation between interface
result in a rise of strength [25]. The influence of different shear strength and matrix yield strength:
processing conditions on the mechanical strength of the smðShearÞ E1=2smðyieldÞ [22]. The strength smðyieldÞ was
fibre reinforced glass ionomer cement (FRGIC) was derived from the flexural matrix strength of the used
analysed. base cement system. The tensile strength of the fibres
was measured in an universal testing machine (Z 2.5,
Zwick, Germany) according to ENV 1007-4 standard.
2. Experimental procedure The fibre length and diameter distributions were
measured using light microscopy, Fig. 2.
2.1. Glass fibres
2.2. Glass ionomer cements
The following glass composition was selected for fibre
processing: SiO2:33.3, Al2O3:16.7, CaO:14, NaF:3.3, The GIC matrices were prepared by mixing a glass
AlF3:3.3, Na3AlF6:16.2 [wt%]. The oxide powder powder with an aqueous polycarbonic acid. The mean
mixture was melted at 1400
C for 1 h and subsequently particle size of the glass powders are listed in Table 1.
quenched in water of room temperature. The glass frit The milled glass powder was treated in various ways to
was milled to obtain different grain sizes of 6.8 and reduce surface reactivity and to increase working time
2.4 mm, respectively. The grain size distributions of the [28,29]. A combination of acid treatment and annealing
 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907 2903

Table 1
Test specifications and results of the cement optimizing tests

Glass charge Grain size [mm] BET surface area Acid washing/heat Working time [sec] Powder/liquid Compressive strength
(d10 =d50 =d90 ) [m2/g] treatment ratio [MPa] (S.D.)

G1 1.7/12.2/46.0 2.0 Non n.m. 2.0 112 (34)

G2 1.7/12.2/46.0 2.0 Non 45 2.8 64 (24)
G3 1.6/6.8/28.7 4.1 7 h, 3% HCl/6 h 268 2.0 70 (4)
360
C
G4 1.6/6.8/28.7 4.1 7 h, 3% HCl/6 h 268 2.8 96 (3)
360
C
G5 n.m. n.m. 5 h, 360
C 49 2.8 n.m.
G6 n.m. n.m. 5 h, 3% HCl 111 2.8 n.m.
G7 1.6/6.8/28.7 4.1 7 h, 3% HCl/6 h 268 2.8 n.m.
360
C
G8 1.2/3.0/21.1 6.1 Non n.m. 1.8 138 (31)
G9 0.9/2.4/7.3 10.4 7 h, 3% HCl/ 6 h n.m. 1.5 170 (18)
360
C

Glass charge Fibre fraction BET surface area Base glass Working time [sec] Powder/liquid Compressive strength
[vol.%] [m2/g]a ratio [MPa] (S.D.)
F1 20 1.624 G4 n.m. 4 134 (4)
F2 40 1.249 G4 n.m. 5.4 98 (4)
F3 60 0.873 G4 n.m. 5.4 82 (6)
a
Calculated cylindrical fibre surface area: 0.1225 m2/g.

procedure was applied, as shown in Table 1 (G5–G7). ing tests. The 4-point-bending-test was applied on GICs
The optimum powder to liquid ratio was found for the and FRGICs with optimized parameters. The data were
processing conditions which finally resulted in the GIC statistically treated by an ANOVA test with a Bonfer-
with the highest compressive strength, Table 1 (G1–G4). roni post-hoc routine (po0:05). For fractographic
The medium grain size was decreased for G8 and G9. analysis the fractured FRGIC specimens were examined
For all cements an aqueous solution of 40–45 wt% under a SEM (Leitz ISIs SR50, Akashi, Japan).
polyacrylic–maleic acid solution containing 8–10 wt%
tartaric acid was used as suspension liquid. The
FRGICs were prepared by adding a percentage of fibres
to the glass powder. The optimum fibre loading was 3. Results and discussion
determined by measurements of the compressive
strength of the composite materials. The GIC and 3.1. Reactive glass fibre
FRGIC were all hand mixed and manufactured in a
suitable mould according to ISO standard. The speci- Due to a 1 h homogenisation of the glass melt and due
mens for compressive strength testing had a diameter of to the knowledge of fluorine as a volatile, the loss of
4 mm and a height of 6 mm (ISO 9917-1:2002). The fluorine was determined from previous work to 53.4%.
bending bars were produced with the dimensions Based on a calculated content of 1.26 g fluorine in 10 g
25  2  2 mm3 (ISO 4049). The specimens were all oxide powder prior to melting, a remaining content of
stored for 24 h in distilled water at 37
C after an initial 0.67 g in 10 g glass fibres was detected, using the ion-
dry gelation period. Table 1 summarizes the specimen selective electrode (ISE) method [30].
compositions, test specifications and results of strength However, the XRD patterns of the glass still show a
measurements. distinct, crystalline peak of calcium fluoride (CaF2) as
the only crystalline phase. The broadness of the peak
suggests the presence of small CaF2 crystals.
2.3. Strength measurements and fractography Fig. 1 shows a TEM image of the glass. TEM and
EDS examination exhibit a droplet phase (A) in the
Fracture strength under compressive and bending glass that is rich in fluorine and calcium. The droplet
loading of the GICs and FRGICs was determined by phase can also be seen to contain nanoscale crystals (B).
measuring the critical fracture strength of at least 15 Other authors assumed the presence of CaF2 crystals in
specimens each set. All specimens were loaded in an the droplet phase [9,10]. Small crystals (B) may also be
universal testing machine (Z 2.5, Zwick, Germany) with seen in the glass. EDS spectra (A–C) were taken and the
a crosshead speed of 0.75 mm/min. The compressive same elements, calcium, fluorine, aluminium, silicon and
strength measurement was used for the cement optimiz- sodium were detected in all regions. The amorphous
2904 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907

(C)
(A)
(A)

(B)

(B) 500 µm
(C)
Fig. 2. Short glass fibres cut in length of 580 mm (F1–F6).

200 nm Untreated glass of G1 showed the highest compressive

strength of 112 MPa at a powder to liquid ratio of 2.0.
The low strength of 64 MPa (G2) at a higher powder to
Fig. 1. TEM image of the glass with different regions. A: separated
droplet phase; B: CaF2 crystals; C: matrix glass.
liquid ratio is caused by poor mixing properties of the
samples, due to a rapid setting of the cement. In general,
the higher the powder to liquid ratio for a cement system
droplet phase (A) was rich in fluorine, while in the other
might be selected, the stronger the resulting cement will
regions the presence of fluorine is not very pronounced.
be formed. The maximum powder to liquid ratio that is
The small crystals in region B were found to be CaF2
attainable, however, is limited by the increase in
which was confirmed by XRD. In the glass (C) the most
viscosity of the cement [31]. G9 cement showed the
prominent peaks are aluminium and silicon, with
highest compressive strength of 170 MPa at a ratio of 1.5
smaller peaks of calcium, fluorine and sodium.
and was finally chosen to serve as the base glass for the
According to Wood and Hill an amorphous phase
fibre reinforced composites (F4–F6). For a clinical
separation in ionomer glasses is effected by fluorine in
acceptability, the minimum compressive strength is
the glass through a nucleation decomposition process
limited to 100 MPa, according to ISO 9917-1:2002
[10]. They stated that the droplet phase contains CaF2
standard. However, modern commercial GICs exhibit
and is much more reactive than the matrix phase. It is
a compressive strength of 200 MPa or more [3]. Since the
the droplet phase that provides the cations for the cross
system under investigation was created for fibre draw-
linking of the polyacrylic molecules, while the matrix
ing, a simplified model glass composition was intended,
phase provides a stable interface to which the poly-
suffering from a reduced strength performance.
acrylic chains can bond [9].
A low scatter in strength within the treated groups is
Fibres were drawn from the melted glass frit and
most likely the result of the prolonged working time,
subsequently cut into short fibres. The tensile strength of
allowing for better homogenisation of the cement paste.
the drawn fibres attained 420798 MPa. The fibre
Comparable results were published by other authors
diameter was determined to 26 mm. The shear strength
who investigated the influence of optimum powder to
of the matrix–fibre interface was estimated to be half the
liquid ratio and of the viscosity of the cement paste on
yield strength of the unreinforced GIC [22]. The flexural
the final GIC strength [31–33]. The treatment of the
matrix strength smðyieldÞ of the used base cement G9
glass with a combination of acid washing and annealing
attained 10 MPa. Following Eq. (1) a resulting critical
procedure led to reasonable working time. Heat treat-
fibre length 2xc of 546 mm, e.g. a fibre pull-out length xc
ment, acid washing or the application of complex
of 273 mm was calculated. Fig. 2 shows the fibres which
fluoride salts might be promising approaches to prolong
were cut by a rotating guillotine to yield an average
the working time of the cement system. A deactivation
length of 5807160 mm.
of the reactive glass surface by an ion leaching process
or heat treatment that relieves residual stresses from the
3.2. Glass ionomer cement quenched specimen was reported to be successful in
extending the working time [28,29,34].
The optimum powder to liquid ratio for the various The particle size and distribution of the glass frit is of
glass compositions are shown in Table 1 (G1–G4). great importance to control the working and setting
 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907 2905

characteristics of the cement [35]. The grain size 423

distributions of the glasses before and after acid
treatment show a pronounced difference. The reactive 420

surface area of G8 increased from 6.1 to 10.4 m2/g in G9 Glass fibre

Flexural tensile strength σ [MP]

while the grain size decreased from 3.0 to 2.4 mm by ion 417

leaching with 3% hydrochloric acid for 7 h. The 2.7 2.9

compressive strength could be increased from 138 to 15

170 MPa by decreasing the glass powder grain size from

12
3.0 to 2.4 mm (G8–G9). The powder to liquid ratio was FRGIC composite
thereby reduced from 1.8 to 1.5 with a decreased particle 9
size distribution, due to a higher reactive surface area of GIC matrix
10.4 m2/g. 6

A fibre loading of 20 vol% (F1–F3) finally resulted in

3
a fibre reinforced cement with the highest compressive
strength. The powder to liquid ratio was individually 0
adjusted to account for the decreased surface area due to 0.0 0.2 0.4 0.6 0.8 1.0

fibre loading. The cements showed an increase in Strain ε [mm]

compressive strength of approximately 30% after Fig. 3. Stress–strain curve of FRGIC.

reinforcement e.g. the compressive strength of 96 MPa
for GIC (G4) increased to 134 MPa for FRGIC (F1).
The fibre loading is limited by a critical fibre volume Compared to the unreinforced GIC matrix the
that has to be exceeded to result in an increased strength FRGIC clearly exhibits a higher fracture strength and
[22] and by a maximum fibre loading above which large a non-catastrophic failure. The large area under the
microstructural defects are formed [16,17]. Kawano et al. stress–strain curve indicates a significant increase of the
worked with 40 vol% as the highest fraction of work-of-fracture in the FRGIC compared to the brittle
reinforcing fibres whereas Kobayashi et al. reported a GIC matrix material. However, compared to strength
maximum fibre loading of 60 vol%, respectively [19,20]. optimized current commercial materials, a fibre reinfor-
The results of flexural strength testing are listed in cement of those materials may lead to a further
Table 2. The duration of the dry hardening period, fibre improvement [16].
compounding and optional pre-treatment were set as An increased flexural strength for the FRGIC
variables. All cements were mixed with their optimum materials could only be determined by extending the
powder–liquid ratio. Non-reinforced control groups initial dry gelation period from 5 min to 30 min (F5),
were included, G4 and G9. The optimized fibre since extended neutralisation of the acid-based gelation
reinforced cements based on G9 glass powder. Group process takes place and the material might set more
F5 attained the highest flexural strength of 15.6 MPa. A sufficiently [4,5]. The ion leaching of the glass fibre
typical stress–strain curve for FRGIC (F5) and for the surface (F6) counteracts an increase in strength since a
GIC matrix (G9) in case of bending loading is shown in reduced liberation of the metal ions impedes the
Fig. 3. The stress–strain curve of the fibre is given for neutralisation process. As a consequence, the gelation
tension. process on the glass fibre surface is delayed [34]. Both
findings suggest that the reaction kinetics at the interface
matrix–fibre might be different compared to the reaction
Table 2 at the interface matrix–glass powder. The reactivity of
Test specifications for producing GIC and FRGIC and their resulting the fibres might also be affected by fibre processing [36].
flexural strengths
Fig. 1 shows a fully amorphous surface layer, free of
Material Specification P/l ratio Flexural strength nanocrystalline precipitations. It was caused by rapid
[MPa] (S.D.) cooling during fibre drawing. Since in phase separated
G4 d50 ¼ 6:8 mm 2.8 7.8 (1.2)a ionomer glasses the droplet phase will be highly
G9 d50 ¼ 2:4 mm 1.5 9.6 (1.3)b responsible for its reactivity [10] the proportion of
F4 20 vol% fibres, 5 min 1.8 8.9 (1.8)a,b phase separation is reduced in the surface layer of the
initial dry gelation period
fibres.
F5 20 vol% fibres, 30 min 1.8 15.6 (1.6)
initial dry gelation period The analysis of the fracture surfaces gave indications
F6 20 vol% treated fibres, 1.8 10.7 (2.2)b of the failure origin in the cement. Fig. 4 shows a
30 min initial dry gelation fractured surface of a FRGIC (F5). The FRGIC surface
period shows matrix cracks generated by water evaporation.
Data with same superscript letters are not significantly different The cracks propagated along the interface of the initial
(ANOVA, Bonferroni, po0:05). glass particles and fibres.
2906 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907

inforcement of GICs to be applied as a dental

restorative material with improved mechanical proper-
ties.
Compression strength for the GIC material was
increased from 64 to 170 MPa. This was reached by
refined glass particle sizes and by pre-treatment of the
glass surface. A maximum flexural strength of 15.6 MPa
was achieved in the FRGIC compared to 8.9 MPa of the
GIC matrix. This increase was attained by using a fibre
fraction of 20 vol% and by extending the initial dry
gelation period of the GIC matrix. Fractographic
examination showed a pronounced fibre pull-out mode
which gives rise to a resulting increased fracture
toughness and work-of-fracture, respectively.
Fig. 4. Fracture surface of a FRGIC showing pull-out of fibres (F5).
Application of FRGIC materials with improved
mechanical strength and toughness properties might
lead to extended clinical indications, especially in stress
bearing areas.

Acknowledgements

The authors wish to thank the 3 M ESPE Corporation

(Seefeld, Germany) for financially supporting this work.
Special thanks to Dr. G. Rackelmann and S. Hoescheler
for the fruitful discussions on that project.

References

Fig. 5. Reactive surface layer at the fibre–matrix interface (F5). [1] Wilson AD, Kent BE, Clinton D, Miller RP. The formation
and microstructure of dental silicate cements. J Mater Sci
1972;7:220–38.
At the interface between matrix and fibre, a distinct [2] Wilson AD, McLean JW. Glass-ionomer cement. Berlin: Quin-
reactive layer has been formed during the setting tessence, 1988.
[3] Saito S, Tosaki S, Hirota K. Characteristics of glass-ionomer
process. Wilson stated in his work [2] that the interface cements. In: Davidson CL, MjoR. IA, editors. Advances in glass-
between the silica gel layer around the unreacted glass ionomer cements. Berlin: Quintessence Publishing Co, 1999.
core of glass particles in cements tends to be weak. On [4] Wasson EA, Nicholson JW. New aspects of the setting of glass-
this site fracture was observed as illustrated in Fig. 5. ionomer cements. J Dent Res 1993;72:481–3.
[5] Wasson EA, Nicholson JW. Studies on the setting chemistry of
The fracture surface of the fibre reinforced cements
glass-ionomer cements. Clin Mater 1991;7:289–93.
showed clear evidence of extended fibre pull-out. The [6] Forsten L. Fluoride release and uptake by glass-ionomers
absorbed energy due to pull-out gives rise for an and related materials and its clinical effect. Biomaterials
increase in fracture toughness. A further increase in 1998;19:503–8.
toughness and failure strength of the FRGIC can be [7] Griffin SG, Hill RG. Influence of glass composition on the
expected by embedding strong fibres in a weak matrix properties of glass polyalkenoate cements Part IV: influence of
fluorine content. Biomaterials 2000;21:693–8.
material. Defined interface properties of matrix–fibre [8] Williams JA, Billington RW, Pearson GJ. The glass ionomer
lead to a high profit in work-of-fracture and high cement: the sources of soluble fluoride. Biomaterials
toughness, respectively [17,37]. 2002;23:2191–200.
[9] Barry TI, Clinton DJ, Wilson AD. The structure of a glass-
ionomer cement and its relationship to the setting process. J Dent
Res 1979;58:1072–9.
4. Conclusions [10] Wood D, Hill R. Structure–property relationships in ionomer
glasses. Clin Mater 1991;7:301–12.
[11] Vogel W. Glaschemie. Berlin: Springer, 1992.
Short reactive glass fibres with a length of 580 mm [12] De Maeyer EAP, Verbeek RMH, Vercruysee CWJ. Reactivity of
were prepared from a glass frit of the system SiO2– fluoride-containing calcium aluminosilicate glasses used in dental
Al2O3–CaF2–Na3AlF6. The fibres were used for re- glass-ionomer cement. J Dent Res 1998;77:2005–11.
 U. Lohbauer et al. / Biomaterials 24 (2003) 2901–2907 2907

[13] Eichmiller FC, Marjenhoff WA. Fluoride releasing dental [26] Burgman JA, Hunia EM. The effect of fibre diameter, environ-
restorative materials. Oper Dent 1999;23:218–28. mental moisture, and cooling time during fibre formation
[14] Mount GJ. Glass ionomers: a review on their current status. Oper on the strength of E glass fibres. Glass Technol 1970;11:
Dent 1999;24:114–24. 147–52.
[15] Pelka M, Ebert J, Schneider H, Kr.amer N, Petschelt A. [27] Loewenstein KL. The manufacturing technology of continuous
Comparision of two- and three-body wear of glass-ionomers glass fibres. Amsterdam: Elsevier Science Publishers, 1993.
and composites. Eur J Oral Sci 1996;104:132–7. [28] Neve AD, Piddock V, Combe EC. The effect of glass heat
[16] Xie D, Brantley WA, Culbertson BM, Wang G. Mechanical treatment on the properties of a novel polyalkenoate cement. Clin
properties and microstructures of glass-ionomer cements. Dent Mater 1993;12:113–5.
Mater 2000;16:129–38. [29] Schmitt W, Purmann R, Jochum P, Gasser O. US Patent A.
[17] Xu HHK, Eichmiller FC, Barndt PR. Effects of fiber length and 1,376.835; 1983.
volume fraction on the reinforcement of calcium phosphate [30] McCabe JF, Carrick TE, Sidhu SK. Determining low levels of
cement. J Mater Sci Mater Med 2001;12:57–65. fluoride released from resin based dental materials. Eur J Oral Sci
[18] Evans AG, Zwez D. The physics and mechanics of fibre- 2002;110:380–4.
reinforced brittle matrix composites. J Mater Sci 1994;29:3857–96. [31] Billington RW, Williams JA, Pearson GJ. Variation in powder/
[19] Kawano F, Kon M, Kobayashi M, Miyai K. Reinforcement effect liquid ratio of a restorative glass-ionomer cement used in dental
of short glass fibers with CaO–P2O5–SiO2–Al2O3 glass on strength practice. Br Dent J 1990;22:164–7.
of glass ionomer cements. J Dent 2001;29:377–80. [32] DeBarra E, Hill R. Influence of poly (acrylic acid) content on the
[20] Kobayashi M, Kon M, Miyai K, Asaoka K. Strengthening fracture behaviour of glass polyalkenoate cements. J Mater Sci
of glass-ionomer cement by compounding short fibres with 1998;33:5487–97.
CaO–P2O5–SiO2–Al2O3 glass. Biomaterials 2000;21:2051–8. [33] Griffin SG, Hill RG. Influence of poly (acrylic acid) molar mass
[21] Shimada Y, Kondo Y, Inokoshi S, Tagami J, Antonucci M. on the fracture properties of glass polyalkenoate cements. J Mater
Demineralizing effect of dental cements on human dentin. Sci 1998;33:5383–96.
Quintessence Int 1999;30:267–73. [34] Gasser O. Evolution of the glass system. Taken from: Glass
[22] Ashby MF, Jones DRH. Engineering materials 2, 2nd ed. Oxford: ionomers: the next generation. Proceedings of the Second
Butterworth-Heinemann, 1998. International Symposium on Glass Ionomers. Philadelphia,
[23] Chou TW. Fiber reinforcements, general theory of composites of 1994.
comprehensive composite materials, vol. 1. Dundee: Pergamon [35] Takigawa T. Study on glass ionomer cement, influence of powder.
Press, 2000. J Conserv Dent 1982;25:812–3.
[24] Xu HHK, Eichmiller FC, Barndt PR. Effects of fiber length and [36] McKinnis CL, Sutton JW. The glass melting process: II, glass
volume fraction on the reinforcement of calcium phosphate structure and the effect of melting history on glass properties.
cement. J Mater Sci Mater Med 2001;12:57–65. J Am Ceram Soc 1959;42:250–3.
[25] Griffin S, Hill RG. Glass composition influence on glass [37] Lucksanasombool P, Higgs WAJ, Higgs RJED, Swain MV.
polyalkenoate cement mechanical properties. J Non-Cryst Sol Toughness of glass fibres reinforced glass-ionomer cements.
1996;196:255–9. J Mater Sci 2002;37:101–8.