Journal of Oral Rehabilitation 2001 28; 1106±1115
Water sorption and solubility of dental composites
and identi®cation of monomers released in an aqueous
environment
È R T E N G R E N * , H . W E L L E N D O R F ², S . K A R L S S O N * & I . E . R U Y T E R ²
U. O
Prosthetic Dentistry/Dental Materials Science, GoÈteborg University, Sweden, and ²Scandinavian Institute of Dental Materials, Haslum, Norway
SUMMARY Water sorption and solubility of six
proprietary composite resin materials were
assessed, and monomers eluted from the organic
matrix during water storage identi®ed. Water sorp-
tion and solubility tests were carried out with the
following storage times: 4 h, 24 h and 7, 60 and
180 days. After storage, water sorption and solu-
bility were determined. Eluted monomers were
analysed by high performance liquid chromatogra-
phy (HPLC). Correlation between the retention time
of the registered peak and the reference peak was
observed, and UV-spectra con®rmed the identity.
The results showed an increase in water sorption
until equilibrium for all materials with one excep-
tion. The solubility behaviour of the composite resin
materials tested revealed variations, with both mass
Introduction
Dental composite materials are today widely used not
only because of their aesthetic properties but also for
the ability to adhere to tooth substance. Dental com-
posite materials consist mainly of ®ller particles and a
polymer matrix based on different monomers, dimers
and/or oligomers of methacrylates and/or acrylates,
together with additives.
Degradation and erosion of dental composite mater-
ials will occur (Geurtzen, 1998) as a result of incomplete
polymerization and in¯uence of oral ¯uids, especially
water (Ruyter, 1995). The degree of degradation
depends on the chemical composition of the monomers,
dimers and oligomers, the degree of cross-linking in the
polymerized matrix, the degree of conversion and the
environmental impact (Ferracane, 1994).
ã 2001 Blackwell Science Ltd
*Department of
decrease and increase. The resin composition in¯u-
ences the water sorption and solubility behaviour of
composite resin materials. The HPLC analysis of
eluted components revealed that triethyleneglycol
dimethacrylate (TEGDMA) was the main monomer
released. Maximal monomer concentration in the
eluate was observed after 7 days. During the test
period, quanti®able quantities of urethanedimeth-
acrylate (UEDMA) monomer were observed,
whereas 2,2-bis[4-(2-hydroxy-3-methacryloyloxy-
propoxy)-phenyl]propane (Bis-GMA) was only
found in detectable quantities. No detectable quan-
tities of bisphenol-A were observed during the test
period.
KEYWORDS : composite resins, dental materials,
resin cements, solubility, sorption
Chemical degradation is usually caused by oxidation
processes and/or hydrolysis where water plays an
important role in the degradation process (Ruyter,
1995; Terselius, 1995). Water sorption in composite
materials is a diffusion-controlled process and occurs
mainly in the resin matrix (Braden et al., 1976; Braden
& Clarke, 1984). The time-factor and the ®ller content
of the composite material will affect the water sorption
characteristics (SoÈderholm et al., 1984; éysaed & Ruyter,
1986a). Water molecules diffuse into the material,
triggering the chemical degradation and resulting in the
formation of degradation products, which can be
released, from the material (Ruyter, 1995; Geurtzen,
1998). Water diffusion will also lead to erosion of the
composite resin material caused by release of unreacted
monomers (residual monomers) (GoÈpferich, 1996).
Erosion of ®ller particles will also take place because
1106
 DEGRADATION OF COMPOSITE RESIN MATERIALS
of the degradation processes (SoÈderholm, 1981; SoÈder-
holm et al., 1984). These erosion processes will result in
mass loss of the dental composite material.
The release of organic degradation products is depend-
ent on the size of the molecules, where smaller mole-
cules are presumed to have enhanced mobility and
should therefore be eluted faster than larger molecules.
Data elucidating this are still scarce, however (Ferracane,
1994; Geurtzen, 1998). It has been shown that residual
monomers are the main components released from
cured dental composite materials, occurring within the
®rst 7 days from placement (éysaed et al., 1988; Ruyter,
1995). Triethyleneglycol dimethacrylate (TEGDMA) has
been identi®ed as the main substance released from
dental composites (Braden & Pearson, 1981; Inoue &
Hayashi, 1982; Geurtzen, 1998; Vaubert et al., 1999).
Small amounts of other monomers such as 2,2-bis
[4-(2-hydroxy-3-methacryloyloxypropoxy)-phenyl]-
propane (Bis-GMA) and urethanedimethacrylate
(UEDMA) as well as additives may also be eluted into
water (Spahl & Budzikiewics, 1994; Spahl et al., 1998).
The main analytical methods used for the determin-
ation of substances eluted from the cured composite
material are gas chromatography (GC) and high per-
formance liquid chromatography (HPLC), if necessary
combined with mass spectrometry (MS) (Ruyter
& éysaed, 1988; Ruyter, 1995). Gas chromatography
is a useful method for analyses of low molecular weight
components with good thermal stability. Analysis of
larger components, soluble and non-volatile reactive
compounds is preferably carried out with HPLC (Ruyter
& éysaed, 1988).
Clinical relevance
Degradation of the composite material will lead to
reduced mechanical properties (éysaed & Ruyter,
1986b) and clinical drawbacks (Jokstad et al., 1994).
In addition to the loss of the mechanical properties,
some of the released organic components are potent
sensitive/irritative agents, inducing irritant and/or
delayed allergic reactions (Inoue & Hayashi, 1982;
Kanerva et al., 1995; Ruyter, 1995; Hume & Gerzina,
1996; O È rtengren et al., 1999). It is therefore important
to have an understanding of degradation and erosion
processes taking place in dental composite materials
and to identify substances released from the materials.
Previous studies have been carried out using experi-
mental composites whereas studies on composite resin
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
1107
materials for crown and bridge application, and cemen-
tation are sparse (Inoue & Hayashi, 1982; Ferracane
& Condon, 1990).
As water has an important role in the process of
releasing substances and studies on composite crown
and bridge materials and composite resin cements are
sparse, the aim of this study was to:
(i) assess the water sorption and solubility of one
restorative, two crown and bridge composite materials,
and three composite resin cements, and
(ii) identify monomers eluted from the organic matrix
during water storage.
Materials and methods
Specimen preparation
The six different composite resin materials included in
the study are presented in Table 1. They differ from each
other with respect to resin matrix, ®ller loading and
mode of curing (Table 1). The complete composition of
the materials tested were given by the manufacturers,
con®dential however, and can therefore not be accounted
for in detail. The specimens (n ˆ 200) were produced in
a stainless steel mould (diameter: 15 ‹ 1 mm, thickness:
0á5 ‹ 0á1 mm) and cured (Table 2).
Water sorption and solubility
The water sorption/solubility test was performed ac-
cording to ISO 4049 (1988) with the exception that the
specimens were weighed shortly after production (m1).
Within 3 min after curing, they were placed singly in a
polyethylene (PEL) container (Nalge Nunc Interna-
tional, USA) in 5 mL of double-distilled water (ISO
3696, 1987). The specimens of each material were
divided into ®ve groups with 5±15 specimens in each
group and stored at 37 ‹ 1 °C for 4 h, 24 h and 7, 60 or
180 days. A blank, i.e. a sample of water, was stored
under the same conditions and for the same period of
time as each group of the test samples.
After completing the storage periods, water sorption
(Wsp) and solubility (Wsl) were measured and calculated
as described below:
Wsp ˆ m2 ÿ m3 †=V
where m2 is the mass of the specimen in lg, after
immersion in water, m3 is the reconditioned mass in lg
of the specimen and V is the specimen volume in mm3.
1108 È R T E N G R E N et al.
U. O

Table 1. Composite resin materials used in the study. Indicated compositions are based on information in the material safety data sheets
(MS/DS) from the manufacturers

Material (code) Batch No. Matrix based on Filler

Alcaglass (AG) 917610 Bis-GMA Silica (fumed)

Crown and Bridge 1,3-Glycerol- Alcaline glass composite material
(Ivoclar/Vivadent, Liechstenstein) dimethacrylate Ytterbiumtri¯uoride
UEDMA Ba-Al-¯uorsilikateglass
C & B Cement (CB) 069309 (catalyst) Bis-GMA Silica (fused)
Composite cement 079039 (base) dimethacrylate monomers Silica (amorphus)
(Bisco, USA) Glass frit
Sodium ¯uoride
Sono-Cem (SC) 105 34457 (catalyst) Bis-GMA Quartz
Composite cement 102 34459 (base) TCDDMA Silica
(ESPE, Germany)
Targis (TG) 910053 Bis-GMA Barium-glass
Crown and Bridge DDDMA Mixed oxides composite material
(Ivoclar/Vivadent, Liechstenstein) UEDMA Silica
TPH Spectrum (ST) 9704 000689 Bis-GMA-adduct Barium-
Composite restorative Bis-EMA Aluminum-
material TEGDMA Borosilicate-glass
(Dentsply/DeTrey, Germany) Silica
Vario-Link II (VL) 906566 (catalyst) Bis-GMA Barium-glass
Composite cement 911375 (base) UEDMA Mixed oxides
(Ivoclar/Vivadent, Liechstenstein) TEGDMA Ytterbium-tri¯uoride

Bis-GMA = Bis-phenol-A-glycidyl-dimethacrylate, UEDMA = urethane dimethacrylate, TEGDMA = triethyleneglycol dimethacrylate,

Bis-EMA = ethoxylated bisphenol-A-dimethacrylate, DDDMA = decandiol dimethacrylate, TCDDMA = bis-(methacryloyloxymethyl)
tricyclo[5á2.1á02,6] decane.

Table 2. Shade and curing scheme

Curing Curing Preparation Mixing
of the composite resin materials
Material Shade initiation time time (min) Mix time (s)
used in the study
AG ± Light* 40 s ´ 9
CB Universal Chemical 15 1:1 15
TG Incisal S2 Light** 20 s ´ 9
SC A3 Chemical 6 1:1 30
Light* 40 s ´ 9
ST A3 Light* 40 s ´ 9
VL Universal Chemical 6 1:1 30
Light* 40 s ´ 9

*The materials AG, SC, ST and VL were light-cured with an Optilux 400 (serial no. 4212876;
Demetron Res. Corp., Danburg, CT, USA) with an 8-mm light tip in nine overlapping sections
(ISO 4049 1988). The materials SC and VL were light-cured 6 min after the start of mixing.
**The material TG was light-cured with a XL 3000 curing light (serial no. 1029933M, St Paul,
MN, USA) through a mylar ®lm in nine overlapping sections. Thereafter the material was cured
in a light curing oven (Targis power; Ivoclar, Liechstenstein), program 1 ¢Targis-VerguÈtung-
sprogramm¢.

Wsl ˆ m1 ÿ m3 †=V Monomer leakage and HPLC

where m1 is the mass of the specimen in lg before The storage water was transferred to a refrigerator
immersion in water. immediately after the specimen had been removed

ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115

 DEGRADATION OF COMPOSITE RESIN MATERIALS
until analysis could be carried out. The analysis of
organic substances released from the matrix as well as
reference substances of known compounds in the
matrix (Table 1) was carried out by HPLC with the
following conditions:
Column: Chromsep-steel column*, 250 mm length,
4á6 mm in diameter. Octadecylsilane treated SiO2
(Chromsphere 5 C18) with a particle size of 5 lm.
Mobile phase: CH3CN/H2O. The optimal mobile
phase was used for each material/reference material.
Flow speed: 0á8 mL min±1.
Detection: UV: 205 and 225 nm.
Temperature: Constant at room temperature.
Injection: 20 lL `loop'.
The results of the analyses are presented as follows:
ND ˆ not detectable; D ˆ detectable, with the limit for
detection at S/N ˆ 2:1 (signal to noise); Q ˆ quanti®-
able, with the limit for quanti®able amounts set at
S/N ˆ 10:1 (signal to noise).
Statistics
The mean value and standard deviation were calculated
for each series. Differences in water sorption and
solubility within each material with increasing time
were tested for signi®cance using analysis of variance
(Scheffe's F-test). Differences between the various
series of materials were tested for signi®cance using
Wilcoxon's signed rank sum test. A signi®cance level of
5% was used.
Together with the above statistical methods, addi-
tional regression analysis for linear trend was per-
formed for the water sorption behaviour. For water
solubility, additional variance analyses (Bon-Ferroni
and Duncan) were carried out because of the different
solubility behaviour within and between the materials.
Neither of the additional methods changes the conclu-
sions based on the statistics.
Results
Water sorption
The results for water sorption are presented in Fig. 1.
The composite cement SC² was shown to have the
lowest water sorption during the test period, followed
*Chrompack, Berger op zoom, the Netherlands.
² ¶
ESPE, Seefeld, Germany.
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
1109
by the composites TG³ and ST§, both of which showed
higher water sorption values (just below 15 lg/mm3).
The other two composite cements used in the study,
CB¶ and VL³, showed even higher sorption values
(25 lg/mm3) and the highest sorption values (over
40 lg/mm3) were observed for AG at 180 days of water
storage. All differences in water sorption behaviour
between the materials were shown to be statistically
signi®cant (P < 0á05, Wilcoxon's signed rank sum test)
except for the difference recorded between the two
composite cements VL and CB (P ˆ 0á67, Wilcoxon's
signed rank sum test).
A statistically signi®cant increase (P < 0á05, Scheffe's
F-test) in water sorption was recorded for all materials
over the test period (4 h±180 days) with a levelling off
at the end of the storage period (60±180 days) with the
exception of the material AG, which showed increasing
water sorption, veri®ed by regression analysis.
Water solubility
The results for solubility are presented in Fig. 2. A
signi®cantly increased solubility in line with the mass
decrease in the reconditioned specimen was shown for
the composite cements SC (4 h±180 days) and CB
(4 h±7 days) (P < 0á05, Scheffe F-test). The slightly less
hydrophilic composite cement SC, showed a moder-
ately increasing solubility. The increase between 24 h
and 180 days was statistically signi®cant (P < 0á05,
ScheffeÂ's F-test).
A statistically signi®cant decrease in mass loss was
shown for the materials VL and TG over the test period
4 h±180 days (P < 0á05, Scheffe's F-test). The material
TG even showed a mass increase at 180 days.
According to the measurements of mass loss, the
material ST had approximately the same solubility at
all storage times. The material AG showed increased
solubility, in accordance with the mass decrease.
However, the changes in solubility over time were
not shown to be statistically signi®cant, either
with Scheffe's F-test or with the additional statistical
methods used.
Differences in water solubility over the test period
between the materials were shown to be statistically
signi®cant (P < 0á05, Wilcoxon's signed rank sum
test) except for the difference recorded between the
³
Ivoclar/Vivadent, Liechstenstein.
§
Dentsply/DeTrey, Konstanz, Germany.
Bisco, Schaumburg, IL, USA.
1110 È R T E N G R E N et al.
U. O

Fig. 1. Mean values and standard deviation for water sorption (lg/mm3) of the composite resin materials tested after 4 h, 24 h, 7, 60 and
180 days of water storage.

Fig. 2. Mean values and standard deviation for water solubility (lg/mm3) of the composite resin materials tested after 4 h, 24 h, 7, 60 and
180 days of water storage.

materials SC and ST (P ˆ 0á29, Wilcoxon's signed rank
sum test) and CB and AG (P ˆ 0á23, Wilcoxon's signed
rank sum test).
HPLC±UV analysis
Results for released monomers are given in Table 3.
Correlation between the retention time of the regis-
tered peak and the reference peak was observed and
UV-spectra con®rmed the identity.
The monomer TEGDMA was the main substance
found in the storage water and quanti®able quantities
were already registered after 4 h of storage for the
materials CB, ST and VL.
The results showed a maximum concentration of
monomers in the storage water after 7 days, exempli-
®ed by TEGDMA in Figs 3±5. Detectable quantities of
Bis-GMA were found in the storage water for the
materials CB, TG and VL. Detectable and quanti®able
quantities of UEDMA were found in the storage water for
the materials AG, TG and VL. The material AG showed
quanti®able release of 1,3-glyceroldimethacrylate after
4 and 24 h but after longer storage the monomer could
not be detected in the storage water. In this study,

ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115

 DEGRADATION OF COMPOSITE RESIN MATERIALS 1111

Table 3. Organic substances

Material Organic substance 4h 24 h 7 days 60 days 180 days
released from the composite resin
materials tested at the different AG Bis-GMA ND ND ND ND ND
storage times UEDMA D (3) D (4) ND ND ND
ND (2) ND (1)
1,3-Glycerol Q Q (2) ND ND ND
dimethacylate D (1)
ND (2)

CB Bis-GMA D D D D D
Bis-EMA ND ND ND ND ND
TEGDMA Q Q Q Q Q (2)
D (3)

TG Bis-GMA D D D D D
DDDMA ND ND ND D (1) ND
ND (4)
UEDMA D Q Q Q D

SC TCDDMA ND ND ND ND ND
Bis-GMA ND ND ND ND ND

ST Bis-GMA ND ND ND ND ND
Bis-EMA ND ND ND ND ND
TEGDMA Q Q Q ND ND

VL Bis-GMA D (2) D D (4) D (2) ND

ND (3) ND (1) ND (2)
UEDMA D Q (1) Q (2) Q (2) ND
D (4) D (3) D (2)
TEGDMA Q Q Q Q ND

ND = Not detectable, D = detectable, with the limit for detection at S/N = 2:1 (signal to noise),
Q = Quanti®able, with the limit for quanti®able amounts set at S/N = 10:1 (signal to noise).
Numbers in parantheses refer to the numbers of repetitive observations of a total of ®ve (four)
analyses.
No number indicates same result for ®ve analyses.

Fig. 3. (A) Reference chromatogram of TEGDMA, (B) Release
from cured VL after 4 h' water storage. Peak 1 represents an
impurity in water, which is also observed in the blank
solution. Peak 2 represents quanti®able TEGDMA. Peak 3
represents detectable UEDMA. The chromatograms are registered
at 205 nm with a mobile phase of: 70% CH3CN/30% H2O.
Fig. 4. (A) Release from cured VL after 1-day water storage. Peak 1
represents an impurity in water, which is also observed in the
blank solution. Peak 2 represents quanti®able TEGDMA. Peak 3
represents detectable UEDMA. (B) Reference chromatogram of
UEDMA. (C) Reference chromatogram of Bis-GMA. The chroma-
tograms are registered at 205 nm with a mobile phase of:
70% CH3CN/30% H2O.

ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115

1112 È R T E N G R E N et al.
U. O
Fig. 5. (A) Release from cured VL after 7-day water storage. Peak 1
represents an impurity in water, which is also observed in the
blank solution. Peak 2 represents quanti®able TEGDMA. Peak 3
represents detectable UEDMA. Peak 4 represents detectable
Bis-GMA. (B) Release from cured VL after 60-day water storage.
Peak 1 represents an impurity in water, which is also observed in
the blank solution. Peak 2 represents quanti®able TEGDMA. Peak
3 represents quanti®able UEDMA. Peak 4 represents detectable
Bis-GMA. (C). Release from cured VL after 180-days water
storage. Peak 1 represents an impurity in water, which is also
observed in the blank solution. No peaks were detected for
TEGDMA, UEDMA or Bis-GMA. The chromatograms are regis-
tered at 205 nm with a mobile phase of: 70% CH3CN/30% H2O.
neither any detectable nor any quanti®able quantities
of monomers were registered in the storage water for
the material SC, in spite of the moderate decrease in
mass during the test period.
No detectable quantities of bisphenol-A were
observed in the eluates during the test period.
Discussion
Water has an important role in the chemical degrada-
tion of composite materials, resulting in both hydro-
lytic reactions and swelling of the material (Ruyter,
1995; Terselius, 1995; Geurtzen, 1998). In that respect,
the water sorption and solubility behaviour of compo-
site materials is of great interest. Water uptake in
composite materials mainly takes place in the resin
matrix and is a diffusion-controlled process, the
diffusion coef®cient decreasing with the concentration
of water in the matrix (Braden et al., 1976; Braden
& Clarke, 1984). Therefore, a reduction in the rate of
water uptake is to be expected with increasing storage
time.
In the present study, the expected rate reduction in
water uptake at increased storage times was shown for
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
the majority of the dental composite materials tested
(Fig. 1). However, the materials showed a wide range
in the magnitude of water sorption at equilibrium.
These differences in water sorption between the
different materials could be explained by differences
in matrix composition (Table 1). The material SC,
which is a composite resin cement, showed less water
sorption than the other materials even after 180 days.
This could be explained by the matrix formulation of
Bis-GMA and the hydrophobic monomer bis(metha-
cryloyloxyethyl)-tricyclo[5á2.1á02,6]decane.Polymerized
materials from mixtures with a high concentration of
hydrophobic monomers will show less water sorption
than materials polymerized from more hydrophilic
monomers within the matrix (éysaed & Ruyter,
1986a). The latter could explain why AG still showed
an increase in water sorption after 180 days of water
storage. The matrix in AG is based on the monomers
UEDMA, Bis-GMA and 1,3-glyceroldimethacrylate. The
latter two have hydroxyl groups, increasing the water
sorption. The polar nature of such a polymer matrix
and the presence of ether linkages are of importance for
water sorption and hygroscopic expansion of composite
resin materials (Peutzfeldt, 1997). Results presented by
Pearson and Longman (1989) showed less water
sorption from a composite based on UEDMA (less
hydrophilic monomer) compared with composites
based on Bis-GMA. Braden (1984) found that compo-
sites based on TEGDMA alone exhibited higher diffu-
sion coef®cients than composites based on a blend of
UEDMA and TEGDMA. The difference was explained
by a higher degree of crosslinking in the latter case. As
crosslinking of the matrix is often correlated to the
degree of conversion, it would also be expected that the
water sorption will increase with a decreased degree of
conversion of the material. However, Rueggeberg and
Craig (1988) found that water sorption was a poor
predictor for the degree of conversion. Therefore,
crosslinking of the matrix seems to be more important
for water sorption kinetics (diffusion coef®cient) but
has a limited in¯uence on water sorption per volume.
As shown in the present study, water sorption seems to
be in¯uenced by the af®nity to water depending on the
quantity of hydroxyl groups within the matrix, creating
hydrogen bonds with water. The monomer TEGDMA
does not contain any hydroxyl groups but still has some
af®nity to water because of the water compatible ether-
linkage structure within the molecule. Braden and
Davy (1986) reported that matrix compositions of
 DEGRADATION OF COMPOSITE RESIN MATERIALS
ethylene glycol dimethacrylates (constituting ether-
linkages) increased water sorption in composite resin
matrices.
Braden and Clarke (1984) and éysaed and Ruyter
(1986a) reported higher water uptake in composite
materials with low ®ller content (i.e. high resin
content) compared with materials with higher ®ller
content. The latter may also partly explain why SC, in
spite of the hydrophilic mer-unit Bis-GMA within the
matrix, still has less water sorption than the other
materials.
The solubility behaviour of composite resin materials
besides the previous factors described, will also be
affected by the type of ®ller used, the treatment of
the ®ller (i.e. silane treatment) (SoÈderholm, 1981;
SoÈderholm et al., 1984) as well as air voids within
the composite resin material (éysaed & Ruyter, 1986a),
often leading to a decrease of mass of the material.
In the present study, all materials showed different
water solubility values. Two materials, TG and VL,
showed a decreasing loss of mass during the test period
and TG showed a mass increase after 180 days in spite
of monomer leakage from the material (Fig. 2 and
Table 3). One possible explanation for the increased
mass of TG is chemical reactions with water within the
composite. The glass ®ller and metal oxides in the
composite material can hydrolyse with the formation
of metal hydroxides as reaction products, i.e. addition
of water (SoÈderholm et al., 1984; Kildahl & Ruyter,
1997).
The HPLC analyses showed the highest concentration
of monomers in the storage water after 7 days, in line
with other studies (éysaed et al., 1988; Ruyter, 1995).
Ferracane (1994) has stated some factors of importance
for the release of unbond substances from polymerized
dental composites:
(i) The monomer±polymer conversion determine
the quantity of leachable components,
(ii) The composition and solubility parameters of
the solvent which is used for extraction in¯uence the
kinetics and mechanism of elution processes, and
(iii) Size as well as chemical characteristics of the
leachable substances determine the diffusion through
the polymer network.
Progressive degradation changes the microstructure of
the bulk through the formation of pores, via which
oligomers, residual monomers, degradation products
and additives can be released. Along with this, the pH
inside the pores may begin to be affected by degradation
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
1113
products, which typically have some acid±base func-
tionality (GoÈpferich, 1996; Geurtzen, 1998).
Ferracane (1994) reported that 5±10% of unbonded
substances in composite material was released into an
aqueous solution (2 wt% of the resin matrix), and
Tanaka et al. (1991) found that small monomers were
extracted in considerably higher quantities than the
large monomers. Small molecules have enhanced
mobility and will be eluted faster than larger molecules
(Gedde, 1995). In the present study, TEGDMA, one of
the smallest molecules in the composite resin materials
tested, was the major monomer eluted (three of the six
materials). This result is in line with other studies
(Braden & Pearson, 1981; Inoue & Hayashi, 1982;
Geurtzen, 1998; Vaubert et al., 1999). MuÈller et al.
(1997) found that twice as much TEGDMA leached
out from an experimental composite based on Bis-GMA/
TEGDMA as from a UEDMA/TEGDMA-based composite
resin material. This can be explained by a higher degree
of curing because of enhanced mobility of the UEDMA
and TEGDMA molecules. The detectable quantities of
the high-molecular mass substances Bis-GMA and
UEDMA released from the materials CB, TG and VL,
might also be explained by the degree of curing of the
materials (i.e. less degree of curing), depending on the
mobility of the molecules as well as the blend monomer
formulation (MuÈller et al., 1997), and the initiator/
activator, inhibitor system used (Asmussen, 1982).
In the present study, the release of the degradation
product, bisphenol-A, could not be detected in the
aqueous eluates from the materials tested. In a study,
where bisphenol-A was observed, dimethacrylate
monomers containing the bisphenol-A unit were
treated with an alkaline or acidic aqueous solution at
100 °C (Olea et al., 1996). Under these vigorous
conditions, hydrolysis of hydrolysable substances such
as Bis-GMA results in formation of the individual
structural components, i.e. bisphenol-A, methacrylic
acid and glycerol. It has been established that Bis-GMA,
used in composite restorative materials can be resolved
into several components (Ruyter & Sjùvik, 1981). The
Bis-GMA samples contain several isomers such as linear
and a single-branched Bis-GMA in a ratio of approxi-
mately 3:1. In addition, a double-branched isomer
(» 2%) will be present, and the monomethacrylate
derivative of Bis-GMA has been observed as an
impurity (Ruyter & éysaed, 1988). The Bis-GMA
derivatives with higher molecular masses have also
been observed (Ruyter & éysaed, 1987), together with
1114 È R T E N G R E N et al.
U. O
several impurities (not identi®ed). However, Bis-GMA of
poor quality containing impurities such as bisphenol-A
can result in release of bisphenol-A from the polymer-
ized material. Bisphenol-A may therefore be regarded
as an impurity rather than a degradation product from
Bis-GMA when a composite is exposed to the oral
conditions.
Conclusions
Within the limitations of this study, it is concluded that
the resin matrix composition is important for the water
sorption and solubility behaviour of composite resin
materials. Materials made from hydrophilic monomers
will have a high water sorption. Chemical reactions
between ®ller particles and water can result in a mass
increase of the dental composite material. The HPLC
analyses of eluted components revealed that triethylene-
glycol dimethacrylate (TEGDMA) was the main
released monomer. A maximum monomer concentra-
tion in the eluate was observed after seven days. The
molecular mass and the degree of curing seem to be
important for the elution rate. During the test period,
quanti®able quantities of (UEDMA) monomer were
observed, whereas Bis-GMA was only found in detect-
able quantities. No detectable quantities of bisphenol-A
could be observed under the conditions and with the
materials used in this investigation.
Acknowledgments
This study was supported by grants from the Swedish
Council for Work Life Research (96-0503, 97-1335) and
the Swedish Foundation for Health Care Sciences and
Allergy Research (A 96015).
We thank Bo Eriksson (Nordic School of Public
Health) for his advice and help with the statistical
calculations, John Gulliver for grammatical corrections
and Dentsply/DeTrey (Germany), Ivoclar/Vivadent
(Liechtenstein), ESPE (Germany) and Bisco (USA) for
the materials.
References
ASMUSSEN , E. (1982) Factors affecting the quantity of remaining
double bonds in restorative resin polymers. Scandinavian Journal
of Dental Research, 90, 490.
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
BRADEN , M. (1984) Water absorption characteristics of dental
micro®ne composite ®lling materials. II Experimental materials.
Biomaterials, 5, 373.
BRADEN , M., CAUSTON , E.E. & CLARKE , R.L. (1976) Diffusion of
water in composite ®lling materials. Journal of Dental Research,
55, 738.
BRADEN , M. & CLARKE , R.L. (1984) Water absorption character-
istics of dental micro®ne composite ®lling materials. I Proprie-
tary Materials. Biomaterials, 5, 369.
BRADEN , M. & DAVY , K.W.M. (1986) Water absorption character-
istics of some un®lled resins. Biomaterials, 7, 474.
BRADEN , M. & PEARSON , G.J. (1981) Analysis of aqueous extracts
from ®lled resins. Journal of Dental Research, 9, 141.
FERRACANE , J.L. (1994) Elution of leachable components from
composites. Journal of Oral Rehabilitation, 21, 441.
FERRACANE , J.L. & CONDON , J.R. (1990) Rate of elution of
leachable components from composite. Dental Materials, 6, 282.
GEDDE , U. (1995) Polymer Physics, pp. 55±75. Chapman & Hall,
London.
GEURTZEN , W. (1998) Substances released from dental resin
composites and glass ionomer cements. European Journal of Oral
Science, 106, 687.
GoÈPFERICH , A. (1996) Mechanisms of polymer degradation and
erosion. Biomaterials, 17, 103.
HUME , W.R. & GERZINA , T.M. (1996) Bioavailability of compo-
nents of resin-based materials which are applied to teeth. Critical
Review in Oral Biological Medicine, 7, 172.
INOUE , K. & HAYASHI , I. (1982) Residual monomer (Bis-GMA) of
composite resins. Journal of Oral Rehabilitation, 9, 493.
ISO 3696 (1987) Water for Analytical Use ± Speci®cation and Test
Methods. International Organization for Standardisation,
Switzerland.
ISO 4049 (1988) Dentistry ± Resin-based Filling Materials. Interna-
tional Organization for Standardisation, Switzerland.
JOKSTAD , A., MJéR , I.A. & QVIST , V. (1994) The age of restorations
in situ. Acta Odontologica Scandinavica, 52, 234.
KANERVA , L., ESTLANDER , T. & JOLANKI , R. (1995) Dental
problems. In: Handbook of Contact Dermatitis (ed. J.D. Guin),
pp. 397±432. McGraw-Hill, New York.
KILDAHL , K.K. & RUYTER , I.E. (1997) How different curing
methods affect mechanical properties of composites for inlays
when tested in dry and wet conditions. European Journal of Oral
Science, 105, 353.
MU È LLER , H., OLSSON , S. & So
ULLER È DERHOLM , K.J. (1997) The effect of
comonomer composition, silane heating and ®ller type on
aqueous TEGDMA lechability in model resin composites.
European Journal of Oral Science, 105, 362.
OLEA , N., PULGAR , R., PEEREZ
 REZ , P., OLEA -SERRANO , F., RIVAS , A.,
NOVILLO -FERTRELL , A., PEDRAZA , V., SOTO , A. & SONNENSCHEIN , C.
(1996) Estrogenicity of resin-based composites and sealants
used in dentistry. Environmental Health Perspectives, 104, 298.
OÈ RTENGREN , U., ANDREASSON , H., KARLSSON , S., MEDING , B.
& BARREGA
ARREGARDÊ RD , L. (1999) Prevalence of self-reported skin
symptoms among Swedish dentists. European Journal of Oral
Science, 106, 496.
 DEGRADATION OF COMPOSITE RESIN MATERIALS
éYSAED , H. & RUYTER , I.E. (1986a) Water sorption and ®ller
characteristics of composites for use in posterior teeth. Journal of
Dental Research, 65, 1315.
éYSAED , H. & RUYTER , I.E. (1986b) Composites for use in posterior
teeth. Mechanical properties tested under dry and wet condi-
tions. Journal of Biomedical Materials Research, 20, 261.
éYSAED , H., RUYTER , I.E. & SJéVIK -KLEVEN , I.J. (1988) Release of
formaldehyde from dental composites. Journal of Dental Research,
67, 1289.
PEARSON , G.J. & LONGMAN , C.M. (1989) Water sorption and
solubility of resin-based materials following inadequate poly-
merization by visible-light curing system. Journal of Oral
Rehabilitation, 16, 57.
PEUTZFELDT , A. (1997) Resin composites in dentistry: the monomer
system. European Journal of Oral Science, 105, 97.
RUEGGEBERG , F.A. & CRAIG , R.G. (1988) Correlation of parameters
used to estimate monomer conversion in a light curing
composite. Journal of Dental Research, 67, 932.
RUYTER , I.E. (1995) Physical and chemical aspects related to
substances released from polymer materials in an aqueous
environment. Advanced Dental Research, 9, 344.
RUYTER , I.E. & éYSAED , H. (1987) Composites for use in posterior
teeth: Composition and conversion. Journal of Biomedical
Materials Research, 21, 11.
RUYTER , I.E. & éYSAED , H. (1988) Analysis and characterization of
dental polymers. CRC Critical Review in Biocompatibility, 4, 247.
RUYTER , I.E. & SJéVIK , I.J. (1981) Composition of dental resin and
composite materials. Acta Odontologica Scandinavica, 39, 133.
ã 2001 Blackwell Science Ltd, Journal of Oral Rehabilitation 28; 1106±1115
1115
SoÈDERHOLM , K.J. (1981) Degradation of glass ®ller in experi-
mental composites. Journal of Dental Research, 60, 1867.
SoÈDERHOLM , K.J., ZIGAN , M., RAGAN , M., FISCHLSCHWEIGER , W.
& BERGMAN , M. (1984) Hydrolytic degradation of dental
composites. Journal of Dental Research, 63, 1248.
SPAHL , W. & BUDZIKIEWICS , H. (1994) Qualitative analysis
of dental resin composites by gas and liquid chromatography/
mass spectroscopy. Fresenius Journal of Analytic Chemistry, 350,
684.
SPAHL , W., BUDZIKIEWICS , H. & GEURTZEN , W. (1998) Determina-
tion of leachable components from four commercial dental
composites by gas and liquid chromatography/mass spectro-
scopy. Journal of Dentistry, 26, 137.
TANAKA , K., TAIRA , M., SHINTANI , H., WASAKA , K. & YAMAKI , M.
(1991) Residual monomers (TEG-DMA and Bis-GMA) of set
visible-light-cured dental resin composite when immersed in
water. Journal of Oral Rehabilitation, 18, 353.
TERSELIUS , B. (1995) In: Polymerteknologi. 2 eds. Chapter 5, KTH,
Stockholm. (in Swedish).
VAUBERT , V.M., MOON , P.C. & LOVE , B.J. (1999) Extractable free
monomers from self-cured dental sealants resulting from
dispensing errors. Journal of Biomedical Materials Research (Applied
Biomaterials), 48, 5.
Correspondence: Ulf O È rtengren, Department of Prosthetic Dentistry/
Dental Materials Science, Faculty of Odontology, GoÈteborg University,
Box 450, SE-405 30 GoÈteborg, Sweden.
E-mail: ortengren@odontologi.gu.se