CURRENT TRENDS IN DENTAL COMPOSITES

J.L Ferracane
Department of Biomaterials and Biomechanics, Oregon Health Sciences University, 611 S.W. Campus Drive, Portland, Oregon 97201-3097

ABSTRACT: The clinical performance of dental composites has been significantly improved over the past decade through modifications in formulation that include: using more stable polymerization promoters for greater color stability; incorporating
high concentrations of finely ground fillers to produce adequate strength and excellent wear resistance while retaining translucency; adding radiopacifying agents for improved diagnostics; and utilizing dentin adhesives. However, there are problems
which limit the use of composites, especially in posterior teeth. The materials remain very technique-sensitive, due to the
extensive contraction which accompanies polymerization and negatively influences marginal sealing. In addition, the materials are generally considered to have inadequate mechanical properties and wear resistance in contact areas to serve as total
replacements for amalgams. Current efforts are focusing on several areas, including the development of non- or minimallyshrinking dental composites containing spiro-orthocarbonates as additives to dimethacrylates or epoxy-base resins, and the
production of alternative filler materials for ideal wear resistance and esthetics. This paper reviews the composition and characteristics of current dental composites, as well as recent areas of study.
Keywords. Dental composite, resin, restorative material, composites, resins, methacrylates, dental restoration.

(I) Introduction
"Composite" refers to a mixture. In materials science, a
composite is a mixture produced from at least two of the
different classes of materials, i.e., metals, ceramics, and
polymers. Dental composites are complex, tooth-colored
filling materials composed of synthetic polymers, particulate ceramic reinforcing fillers, molecules which promote or modify the polymerization reaction that produces the cross-linked polymer matrix from the
dimethacrylate resin monomers, and silane coupling
agents which bond the reinforcing fillers to the polymer
matrix. Each component of the composite is critical to
the success of the final dental restoration. However, the
most significant developments in the evolution of commercial composites to date have been a direct result of
modifications to the filler component.
Dental composites have been considered acceptable
restorative materials for anterior applications for many
years. Their tooth-matching ability and lack of metallic
mercury have caused them to be promoted as an adjunct
to or substitute for dental amalgam in the restoration of
the posterior dentition. Though not universally accepted
by the profession, the continued "fine tuning" of composite formulations by dental manufacturers has produced acceptable materials for more expanded use in
the posterior dentition. These alterations have mainly
involved the use of radiopaque glass fillers that are capa-

302

ble of being ground or formed into very fine particle

sizes, thus enhancing the polishability and intra-oral
abrasion resistance of the resultant composite.
Favorable results from long-term clinical trials demonstrate that when placed correctly, composites can produce esthetic posterior restorations with excellent
longevity (el-Mowafy et ai, 1994; Taylor et al, 1994).
However, there remain significant problems that limit
the usefulness of these materials in the routine practice
of dentistry. Although one of the concerns relates to what
is perceived as a limited durability when the material is
placed in occlusal contact, the most significant problems
relate to the excessive contraction accompanying intraoral polymerization of composite. It is likely that the
development of a new matrix polymer that undergoes
zero or negligible curing contraction would be a major
step toward solving the majority of the difficulties
involved in the use of dental composite.
This paper will present a summary of the composition and characteristics of current dental composites.
Each component of the composite will be discussed separately, first by describing the state-of-the-art of present
materials as presented in the literature, and subsequently by discussing recent and future areas of study. The
manner in which each component influences the use of
the material and the outcome of the final restoration will
be discussed. Recent clinical and laboratory findings
identifying the successes as well as the current problems

Crit Rev Oral Biol Med

6(4):302-318(1995)

with composites will be highlighted and used to suggest

future directions for research and development.

(II) Fillers
Fillers are often used in the plastics industry to reduce
the cost of a component, because the polymers are usually the most expensive material in the part. Fillers are
also added to provide reinforcement, e.g., fiberglass.
Fillers are used in dental composites to provide
strengthening (Ferracane et ai, 1987; Chung and Greener,
1990), increased stiffness (Braem et ai, 1989; Kim et ai,
1994), reduced dimensional change when heated and
cooled (Soderholm, 1984; Yamaguchi et ai, 1989),
reduced setting contraction (Munksgaard et ai, 1987; Iga
etai, 1991), radiopacity (van Dijkenetfl!., 1989), enhanced
esthetics, and improved handling. In general, the physical and mechanical properties of the composite are
improved in direct relation to the amount of filler added.
A variety of ceramic filler materials has been utilized, but
the selection is limited by several factors.
One of the most important considerations in the
selection of a filler is the optical characteristics of the
composite. The monomer resins used in dental composites have a refractive index of approximately 1.55. Fillers
with refractive indices which differ greatly from this value
will cause the composite to appear optically opaque, creating an esthetic and curing problem. Because glasses
can have refractive indices ranging from 1.4 to 1.9, the
selection of an appropriate filler for dental composites
must be guided by a consideration of this important
variable.
(A) PRESENT MATERIALS

Initially, fused or crystalline quartz and various borosilicate or lithium aluminosilicate glasses were used as
fillers for dental composites. The glass or quartz was
ground or milled into particles of various sizes, ranging
from approximately 0.1 |nm to 100 |xm (Phillips, 1991).
These particles were added to resin monomer at approximately 70-80 wt% (55-65 vol%) to make a paste that
could be hardened into a dental restorative material with
strength and stiffness far surpassing those of the unfilled
polymer itself. The major advantage to using quartz was
that it is readily available and has an excellent optical
match to the polymer resin. However, quartz has drawbacks in that it is not radiopaque and can be very abrasive to enamel. Another drawback to the original quartz
and glass fillers were that the particles were large and
very hard in relation to the surrounding polymer matrix.
These characteristics ensured that as the surface of the
composite was abraded, the polymer would wear away
more quickly than the fillers, leaving them raised and
exposed from the surface. This made the surface of the
restoration rough and less enamel-like, due to appreciable scattering of incident light Thus, polishability and

6(41:302-318(1995)

esthetics were compromised.

Microfills, containing amorphous silica, were developed to address the polishing requirements of anterior
restorations. These silicon dioxide particles are submicroscopic, averaging approximately 0.04 (urn in diameter, though the size varies among materials. The amorphous silica is incorporated alone and as pre-polymerized resin fillers and agglomerates. The latter are necessary to increase the volume of filler particles in the composite. The small particles have a large surface-area-tovolume ratio, thus requiring considerable amount of
monomer to wet their surface in order to produce a
homogeneous, non-sticky composite paste. This characteristic limits the filler concentration to approximately 35
wt%, which in turn limits the strength and stiffness of the
material. In order to circumvent these limitations partially and enhance filler reinforcement, a manufacturing
technique was developed to produce heavily filled polymer resin blocks that can be milled into particles approximately 25 |im in size. When these pre-polymerized resin
fillers are added to the resin with additional amorphous
silica, pastes with filler levels of 50-60 wt% (35-45 vol%)
are possible. The small size of the particles allowed these
composites to be polished without preferential abrasion,
thus producing smooth surfaces and excellent esthetics.
However, these fillers, like quartz, are not radiopaque.
Recently, radiopacity has been achieved in microfills by
the addition of small concentrations of other fillers, such
as Ytterbium trifluoride (Heliomolar RO, Ivoclar). The net
result for the microfills is excellent esthetics and polishability, but strength and stiffness lower than those of
quartz-filled or glass-filled composites that contain larger particles. The poorer mechanical properties are attributed to the use of a lower concentration of fillers in the
microfills.
Most current composites are filled with radiopaque
silicate particles based on oxides of barium, strontium,
zinc, aluminum, or zirconium (Hosoda et ai, 1990; Khan
et ai, 1992). The glasses are milled by special processes
to produce particulates of very small size. Many current
materials include few particles which equal or exceed 35 Jim in size, and most have average particle sizes of
approximately 0.6-1.0 |um. Some of the composites have
a very uniform distribution of small filler sizes, while others utilize a broader distribution of particles. These composites also contain a small amount of amorphous silica
microfiller to improve handling characteristics and
reduce stickiness.
There has not been a definitive study to prove the
superiority of any specific filler. Several studies have
identified and evaluated the elements leached from dental composites aged in vitro and have verified that the
radiopaque glasses containing zinc, barium, and strontium are more soluble than quartz or silica in aqueous
solutions (Soderholm et ai, 1984; 0ysaed and Ruyter,

Crit Rev Oral Bio! Med

303

1 0 KU

f <2 .

000

1 0M- m

0 0 0 0 0 0

Figure 1. Scanning electron micrograph of the polished surface

ofa microfill composite, Heliomolar RO, showing a very smooth
surface and the presence of the pre-polymerized resin fillers.

1986; Soderholm, 1990). The quantities of elements

leached are very low (ppm levels), and the main concern
has been related to the possibility that the filler/matrix
interface may become hydrolytically unstable due to the
solubility of the fillers, leading to a more rapid erosion of
the composite. No clinical evidence exists to support
this concern. However, a composite has been developed
in Japan specifically with this thought in mind. Graft LC
II (GC Corp.) contains glass filler particles treated with a
monomer to graft methacrylate groups covalently to the
filler surface, thus enhancing the filler/matrix interfacial
strength and its hydrophobicity (G-C product literature;
Tani, 1993).
As previously stated, the most significant changes in
commercial composites have been made through altering the filler component. These changes have prompted
the periodic development of classification systems for
dental composites based upon filler size and volume
fraction. In 1983, Lutzand Phillips published a system for
characterizing dental composites based on particle size.
Their system included traditional composites (avg. particle size = 1-15 urn), hybrids containing a mixture of
ground glass and microfill particles (avg. particle sizes =
5 nm and 0.04 |j.m, respectively), and microfills (avg. particle size = 0.04-0.1 um). The microfills were divided into
subclasses which included a characterization of the type
of pre-polymerized resin fillers incorporated, i.e., splintered, agglomerated, or spherical.
Recently, Willems et al. (1992) published a similar
classification system. The basic difference between the
two reflects the fact that the most popular non-microfill
composites have smaller mean particle sizes and fewer
large particles than composites of a decade ago. The

304

Figure 2. Scanning electron micrograph of the polished surface

or a mini-filled composite, Herculite XRV, showing a relatively
smooth surface dominated by irregularly shaped particles of less
than 1.0-|im size and few particles greater than 2 ^m.

term "hybrid" is no longer used, since nearly all dental

composites are now "hybrids" of two size ranges, because
they contain some amorphous silica to improve handling
by reducing stickiness. Instead, composites are classified
as midway-filled (< 60 vol%) and compact-filled (> 60
vol%), with classifications of ultrafine (avg. size < 3 urn)
and fine (avg. size > 3 |im) within each category. The categories of microfill composites remain virtually
unchanged, though new materials are included which are
predominantly filled with amorphous silica but also contain some macroscopic reinforcing fillers or radiopacifiers. Also, fiber-reinforced materials are classified in the
new system.
A simpler classification system has been described
by Bayne et al. (1994). In this system, the three popular
types of composites are described by the size of their
largest fillers as microfills (average particle size = 0.010.1 \im), minifills (average particle size = 0.1-1.0 urn), and
midifills (average particle size = 1.0-10.0 urn). Scanning
electron micrographs of the polished surfaces of representative composites in each class demonstrate the differences in particle size and surface smoothness (Figs. 13). This classification system is less inclusive than previous ones, but provides necessary information about the
largest added particles which significantly affect polishability. Therefore, the two most important filler considerations at present are the amount and size of the particles.
A listing of commercial products from the three different
classes of composites is presented in Table 1.
The generally accepted view is that microfill composites have the most ideal esthetic qualities, due to
their excellent polishability and capacity to retain sur-

Crit Rev Oral Bid Med

6(4):302-3I8 (1995)

TABLE 1
Classes of Composites and Representative
Commercial Products
Microfills: amorphous silica fillers (avg. size = 0.04 u.m) +
pre-polymerized resin fillers (size ranging from 5 to
50 |4.m); typical inorganic filler volume = 25-50%;
generally not radiopaque, except Helomolar RO

Figure 3. Scanning electron micrograph of the polished surface

ofa midi-filled composite, Clearfil Posterior, showing a rougher
surface dominated by irregularly shaped particles of several
micrometers' size, witn some 10 u,m or larger.
Minifills:

face smoothness over time. However, many believe that,

due to their poor mechanical properties, these materials
are contra-indicated for stress-bearing restorations such
as class lVs and moderate-to-large classes I and li
restorations in occlusal contact with opposing cusps.
Willems' classification system provides an assessment of filler volume fraction, which greatly contributes
to the physical properties of the composite. Many studies have reported a correlation between mechanical
properties and filler volume (Ferracane, 1989). However,
only a few have reported a correlation between mechanical properties and clinical performance (Tyas, 1990;
Peutzfeldt and Asmussen, 1992a). The current trend
toward minimizing filler size and maximizing filler loading is an attempt to satisfy all of the requirements for
dental composites, according to the results of these laboratory and clinical studies.
Microfill composites have been shown to have fracture resistance, stiffness, and fatigue strength lower than
those of more heavily filled composites (Drummond,
1989; Ferracane and Condon, 1992; Willems et al, 1992;
Braem et al, 1994a). Microfill composites generally show
very low wear from abrasion by food. However, clinical
studies have demonstrated that these composites
undergo greater marginal degradation and localized wear
in contact sites than do more heavily filled composites
with larger particles (Lambrechts et al., 1984; Tyas et al.,
1989; Stangel and Barolet, 1990; Bryant etai, 1992; Mazer
and Leinfelder, 1992). One microfill, Heliomolar RO
(Ivoclar), has exhibited excellent wear resistance in posterior teeth, though marginal breakdown and roughening
in contact areas were apparent (Knibbs and Smart, 1992;

6(4):302-318 (1995)

Brand

Manufacturer

Bisfil-M
Durafill
Epic TMPT
Heliomolar RO
Helioprogress
Perfection
Silux Plus
VisioDisperse

Bisco
Kulzer
Parkell
Ivoclar/Vivadent
Ivoclar/Vivadent
Den-Mat
3M
Espe-Premier

barium, strontium, or zirconia silicate fillers (avg.

size = 0.6-1.0 u.m; largest particles generally smaller than 3-4 u.m) + a small amount of amorphous
silica (avg. size = 0.04 (am), except Z100; typical
inorganic filler volume = 50-70%
Brand
Aelitefil
APH
Brilliant
Charisma
Conquest Crystal
Herculite XRV
Lite-Fil II
Palfique Estelite
Prodigy
Renamel
Tetric
TPH
Z-100

Midifills:

Crit Rev Oral Bio! Med

Manufacturer
Bisco
LD. Caulk
Coltene/Whaledent
Kulzer
Jeneric/Pentron
Kerr
Shofu
Tokuyama Soda
Kerr
Cosmedent
Ivoclar/Vivadent
LD. Caulk
3M

barium, strontium, or zirconia silicate fillers (avg.

size = 1 -5 UJTI; largest particles generally smaller
than 10-15 (im) + a small amount of amorphous
silica (avg. size = 0.04 u.m); typical inorganic filler
volume = 55-70%; 'contain quartz fillers
Brand

Manufacturer

Bisfill-P
Clearfil Photo Posterior*
Ful-Fil
Graft LC
Occlusin
Marathon
P-10
P-50
Pertac Hybrid*

Bisco
Kuraray/J. Morita
LD. Caulk
GC
GC
Den-Mat
3M
3M
Espe/Premier

305

Figure 4. Scanning electron micrograph of the fracture surfaces of two microfill composites: (a) Silux Plus, showing a rough texture with
many debonded and plucked pre-polymerized resin fillers; and (b) Epic TMPT, showing a smoother texture with cleaved pre-polymerized
resin fillers due to the enhanced bonding with the matrix.

Mazer and Leinfelder, 1992). Another microfill, Silux (3M

Dental Products), has also shown excellent abrasion
resistance after five years, but experienced an incidence
of fracture higher than that of more heavily filled materials (Tyas and Wassenaar, 1991). A current hypothesis is
that inadequate fatigue resistance is responsible for the
accelerated occlusal contact wear and marginal degradation observed in composites with filler particles that
average less than 1.0 u.m in size (Mazer et al, 1992; Braem
etai, 1994b).
A factor contributing to the degradation of the
microfill composites is the fact that the pre-polymerized
resin fillers are not bonded well to the polymer matrix.
The resin fillers are heat-cured and do not form covalent
chemical bonds with the polymerizing matrix, due to the
lack of available methacrylate groups on their surfaces.
Therefore, they become debonded and dislodged under
high stresses. A relatively new microfill composite was
developed to minimize this problem by enhancing the
bond between the matrix and the organic fillers. Epic
TMPT (Sun Medical, Japan, and Parkell, USA) uses a typical urethane dimethacrylate in conjunction with a new
monomer, trimethylol propane trimethacrylate (TMPT),
to make the organic fillers. As will be discussed shortly,
the greater availability of reactive species on the organic
filler due to the TMPT theoretically enhances the chemical reaction between the matrix polymer and the filler.
The resultant material is reported to be a highly wearresistant microfill which is also resistant to marginal
degradation when tested in vitro (Suzuki and Leinfelder,
1994). A comparison of scanning electron micrographs of
the fracture surfaces of a conventional microfill (Silux,
3M) and Epic TMPT reveals a difference in bonding, as
depicted by the high proportion of pre-polymerized resin

306

fillers debonded or protruding from the surface of the

conventional microfill (Figs. 4a-b).
The wear of composites in the oral environment is
primarily related to filler particle size and interparticle
spacing (Bayne et al, 1992). Because the matrix is softer
and less wear-resistant than the inorganic filler, it is preferentially abraded by food, toothpaste, etc. As the polymer matrix wears down, it exposes the filler particles,
allowing them to be plucked from the surrounding matrix
during the next abrasion cycle. The use of smaller particles minimizes the space between particles and the
extent of filler plucking and surface degradation during
chewing, thus reducing the rate of abrasive wear. Clinical
studies of reasonable duration, i.e., five years or more,
have confirmed the success of many composites in small
to moderate posterior occlusal cavities, showing wear
rates in contact-free areas of 10-20 urn or less per year
when the average particle size is less than 1.0 urn (Bayne
etai, 1991; Wilson etai, 1991; Mazer and Leinfelder, 1992;
Wendt and Leinfelder, 1994; Leinfelder, 1995). These
results have fueled the further development of very heavily filled small-particle composites, and several new systems have emerged.
One such material is Z100 (3M Dental Products),
which uses zirconia silica fillers produced by a synthetic
sol-gel process instead of particles made from mined
minerals or melted glasses (Fig. 5). The sol-gel process is
one in which a metal carboxylate and a metal oxide sol
are mixed and form a gel by dehydration. The gel is heattreated and then ground to produce the fillers. A similar
technology is utilized to make Palfique Estelite
(Tokuyama Soda), which contains spherical silica zirconia fillers with a particle size of 0.2-0.3 Jim (Tani, 1993).
This composite is the improved version of Palfique intro-

Crit Rev Oral Bid Med

6(4):302-318 (1995)

Figure 5. Scanning electron micrograph of the polished surface

ofZlOO showing a relatively smooth surface dominated by
spherically shaped fillers with an average size of less than 1.0
mm and few particles larger than 3-4 mm.

duced in 1983 (Tani, 1983). Though a similar filler was

used in its predecessor, P-50 (developed in 1988), the
fillers in Z100 are ground differently to produce rounded
particles in a wide distribution of very small sizes (average = 0.5-0.7 u,m; largest particles = 4 u,m). This obviates
the need to add amorphous silica to improve handling.
The composite is very dense (65-70 vol% fillers) and has
good strength, handling, and abrasion resistance. The
good abrasion resistance and polishability are believed
to be due in part to the spherical nature of the fillers. At
least one other composite developed in japan, Progress
(Kanebo), also consists of a very high concentration (75
vol%) of spherical fillers which are graded at two sizes
(1.5 u,m and 0.1-0.5 urn). The high filler loading can be
achieved only by having a non-uniform distribution of
particles.
Another relatively recent addition to the composite
market is Tetric (lvoclar). The novelty of this composite is
that it contains four different types of fillers. Barium
glass is used to achieve radiopacity and is combined with
spheroidal silica particles, amorphous silica for
improved handling, and ytterbium trifluoride for its
potential therapeutic effects. Studies have confirmed
that fluoride is released from one fluoride-containing
composite, Heliomolar RO (Arends and Ruben, 1988),
but that the amount is two orders of magnitude lower
than that leaching from glass-ionomer restoratives
(Takahashi et al, 1993). Thus, the potential therapeutic
effects of the fluoride released from these composites
are questionable. The maximum particle size of Tetric is
approximately 3 u.m. The properties of this material are
comparable with those of Z100 and other ultrafine com-

6(4):302-318 (1995)

posites, such as Brilliant D.I. (Coltene/Whaledent),

Charisma (Kulzer), Herculite XRV (Kerr), and Prisma TPH
(L.D. Caulk). The average particle size for most of these
composites is 0.6-0.8 \im. Recent studies suggest that
these composites are all very wear-resistant, averaging
approximately 10-20 u.m/year due to abrasion
(Leinfelder, 1995). In addition, all can be used as heattreated inlays to enhance marginal sealing.
Perhaps the most wear-resistant composite
designed for posterior applications is Clearfil
Photoposterior (Kuraray, Japan) (Wendt and Leinfelder,
1992). This composite contains a high level (71 vol%) of
quartz fillers of approximately 3 u.m average size (Fig. 3).
It also contains some particles which equal and exceed
10 u,m. The material has excellent mechanical properties
(Willems et al, 1992), though it is difficult to polish as
smoothly as other small-particle composites due to the
large fillers. In addition, because quartz is a very abrasive
material, this composite has the potential to wear
opposing dentition at a more rapid rate than other softer glasses. It is possible that the low abrasion wear for
this material is slightly misleading, in that it may be
wearing the opposing teeth at the same time, thus minimizing its own loss.
This review of current composites emphasizes that
the best mechanical properties are achieved by the
incorporation of high concentrations of filler particles of
various sizes into the resin. The packing of particles is
improved by the use of a non-uniform size distribution
(i.e., Clearfil and Z100), and this results in greater filler
density and maximum reinforcement. Though these
materials, many of which are characterized as midifills,
are acceptable for posterior use, they are less appropriate for anterior restorations, where esthetics and polishability are of the utmost importance. The microfills
have the most ideal esthetics, and should be used in
anterior restorations where stresses are low. The
minifills, which are also highly polishable, can be used in
anterior teeth when greater strength is required due to
moderate occlusal loading. Thus, there is no ideal composite restorative, though the minifills may come the
closest to reaching this goal.
(B) RECENT AREAS OF STUDY

With the possible exception of the microfills, virtually

any dental composite may prove to be abrasive to enamel. The development of fillers that are softer than current
glasses would be clinically beneficial because they
should be less abrasive to opposing teeth. This is especially important when multiple restorations are placed in
an arch. Calcium metaphosphate (CMP) ceramics offer
one potential filler material that meets this criterion. The
major advantage to the use of the CMPs is a softer surface texture as a result of their less brittle nature. Many
calcium phosphate ceramics have been tried as fillers for

Crit Rev Oral Biol Med

307

composites, but their use has been seriously limited by

a high refractive index. Recently, Antonucci et al. (1991)
prepared calcium metaphosphates with improved refractive indices. These minerals exist in vitreous as well as
crystalline form. Composites made from the ground vitreous CMP in Bis-GMA absorbed substantial amounts of
water and were very weak. The crystalline form of the
mineral produced better composites, but they were still
considerably weaker than commercial materials, possibly due to the difficulty in coupling the CMP to the resin
matrix with silane-coupling agents.
As will be described, polymerization contraction and
its accompanying stresses are among the biggest problems facing the expanded use of composites. Because
the polymerization of the resin monomers causes the
contraction, it is obvious that maximizing the quantity of
the inert filler and thus minimizing the quantity of the
resin results in less shrinkage during curing (Iga et al,
1991). This provides a driving force for maximizing filler
levels in composites. However, modifications of the filler
may also affect the amount of contraction and its
accompanying stress. In a recent study, polymerization
contraction stress in composites was reduced through
the addition of 1 wt% of small (1 (im) hollow plastic
spheres which provide sites for stress relief during curing
(Li et al, 1993). Contraction stress could be reduced to
near zero with only a minimal reduction (i.e., 10%) in
mechanical properties. There has apparently been no
commercial application of this idea to date. Also, it is
likely that the hollow spheres would have a negative
effect on the optical characteristics of the composite.
An alternative to conventional composites has been
developed (Bowen et al, 1991). "Megafilled" composite
restorations are produced by filling the bulk of the cavity
preparation with beta-quartz glass inserts (Lee
Pharmaceuticals). The inserts are surrounded by lightcured composite, which bonds to the insert via a silanecoupling agent. The inserts are produced in a variety of
shapes and sizes to fit most cavity preps. When fitted
into the cavity, they minimize the volume of shrinking
composite and reduce curing contraction (George and
Richards, 1993). One recent in vitro study showed the
wear of a beta-quartz restoration to be similar to that of
P-50 or Herculite XR (Kawai and Leinfelder, 1993).
Other recent studies have utilized glass fibers as
reinforcing fillers for composites, showing improvements
in fracture toughness and wear resistance (Li et al, 1993).
At least one commercial composite (Restolux SP4, Lee
Pharmaceuticals) makes use of fiber reinforcement,
though polishability and abrasion resistance would
seem to be less than ideal for such materials.
The future in filler technology probably lies in the
expanded use of sol-gel processing of particles for composites with excellent radiopacity (Suzuki et al, 1991) and
abrasion resistance (Seghietal, 1993). The ability to pro-

308

duce particles of even smaller than current sizes would

be a great advantage, because it would allow manufacturers to produce higher levels of filler loading through
more efficient packing. One consideration is the use of
nanofillers, particles anywhere from 1 to 100 nm in size
(Dagani, 1992; Bayne et al, 1994). It is likely that new
composites will be formulated with some proportion of
fillers below 50 nm in size, though there are technical difficulties in producing such materials with appropriate
handling characteristics.
It is also reasonable to assume that a greater
emphasis will be placed on the use of fluoride-containing inorganic materials as fillers. As mentioned, ytterbium fluoride is used in several products (i.e., Heliomolar
RO, Helioprogress, and Tetric from Ivoclar), and substantial fluoride has been shown to be released from these
materials for up to five years (Arends and Ruben, 1988).
A recent study examining in situ demineralization and fluoride release from these composites demonstrated that
enamel demineralization decreased as fluoride release
increased (Dijkman et al, 1993). By extrapolation of the
data, the authors predicted that a composite that
released 200-300 |iig/cm2 fluoride over a one-month period would completely inhibit secondary caries under
plaque conditions. This amount is approximately 40-50
times as much as that released by Heliomolar RO.
Therefore, despite the fact that these fluoridated composites release significant amounts of fluoride, their present ability to prevent secondary caries is questionable,
and improvements in technology will be required to
increase the level of fluoride release. Dijkman et al. (1993)
showed that a composite with a fluoro-aluminum silicate
filler released an order of magnitude more fluoride than
Heliomolar, and that the fluoride release rate increased
in a linear manner with time. This was hypothesized to
be due to the greater water solubility of the silicate filler
in comparison with ytterbium fluoride. Though this may
enhance fluoride release, a filler with dramatically
increased water solubility may eventually compromise
the properties of the composite in the oral environment.

(Ill) Resin Monomers

The hardening of a dental composite is the result of a
chemical reaction between dimethacylate resin
monomers that produces a rigid and heavily cross-linked
polymer network surrounding the inert filler particles.
The extent of this reaction, often called the degree of
cure, is very important in that it dictates many of the
physical and mechanical properties of the composite
restoration. The degree of cure is influenced by many factors, including the addition of polymerization promoters
and inhibitors (Yoshida and Greener, 1994), the chemical
structure of the monomers (Ferracane and Greener, 1986;
Beatty et al, 1993), the chemical or light energy imparted
to activate the reaction (Rueggeberg and Jordan, 1993),

Crit Rev Oral Biol Med

6(4):302-318(1995)

the filler composition of the material

TABLE 2
(Kawaguchi et al, 1994), and the shade of
the composite.
Chemical Structure of Monomers Used in Dental Composites
Accompanying the polymerization
reaction is a dimensional change that
[(CH3)2-C-[C6H6-O-CH2-CH(OH)-CH2-O-C(=O)-C(CH3)=C(H2)]2
results in shrinkage. Shrinkage is caused
Bis-GMA
by the monomers becoming covalently
bonded by the polymerization reaction,
[(CH3)2-C-[C6H6-O-CH2-CH2-O-C(=O)-C(CH3)=C(H2)]2
thus exchanging van der Waals' distances
Bis-EMA
for covalent bond distances. The magni(H2)C=(CH3)C-(O=)C-O-CH2-CH2-O-CH2-CH2-O-C(=O)"C(CH3)=C(H2)
tude of the shrinkage is dictated by the
TEGDMA
number of covalent bonds that form, i.e.,
the extent of the reaction, as well as by the
[(CH3)2-C-[CH2-CH2-NH-C(=O)-O-CH2-CH2-O-C(=O)-C(CH3)=C(H2)]2
sizes of the monomers. Therefore, the goal
UEDMA
of achieving maximum curing reaction to
enhance the properties of the polymer
matrix is at odds with the desire to minimize dimensional change, the latter being
necessary for achieving good marginal integrity of the
cosity, UDMA-based composites may cure more extensively than Bis-GMA-based composites (Ferracane et al,
restoration. One method to counteract this problem is to
1992). However, a recent study showed depth of cure to
use monomers of very large molecular weight, thus minbe less in certain UDMA-composites, due to a greater
imizing the contraction per given volume of material.
mismatch in the refractive index between monomer and
The monomer originally developed for dental comfiller (SoderholmeU/., 1993).
posites approximately 25 years ago, Bis-GMA, is still
Bis-GMA has a very high viscosity because of the
used in most products today. Though new resins have
hydrogen bonding interactions that occur between the
been developed which claim to be less hydrophilic,
hydroxy groups on the monomer molecules. Thus, Bistougher, or have less dimensional change than Bis-GMA,
GMA must be diluted with a more fluid resin in order to
there is little evidence that the clinical performance of
be useful for dental composites. TEGDMA (triethylenegcomposites has been dramatically affected by differences
lycoldimethacrylate) has excellent viscosity and copolyin resin composition. Because the polymerization conmerization characteristics, and it is most often used as
traction is the most significant problem with current
the diluent monomer for UDMA or the more viscous Bisdental composite, it is highly desirable to develop a new
GMA. Optimal properties are produced when TEGDMA is
or modified resin system that does not result in a net
used in a 1:1 ratio with Bis-GMA (Stannard et al, 1993).
shrinkage during curing. This is currently an area of
Other diluents include ethylene- and hexamethylene-glyintense research.
coldimethacrylate and benzyl methacrylate, a monofunc(A) PRESENT MATERIALS
tional monomer added to enhance polymer chain elongation and degree of cure (Ruyter and Nilsen, 1993).
As has been true for the past 30 years, 80-90% of comThe drawbacks to existing resins for composites
mercial dental composites utilize the Bis-GMA (2.2include:
excessive polymerization shrinkage, incomplete
bis(4-(2-hydroxy-3-methacryloyloxyconversion
and cross-linking, and undesirable water
propoxy)phenyl]propane) monomer developed by Dr.
sorption.
Composite
monomers cure by a free radical
Rafael Bowen as their matrix-forming resin (Ruyter and
polymerization
reaction
to form glassy, cross-linked
Oysaed, 1987). Other base monomers used in present
polymer
networks
that
are
relatively insoluble and somecommercial composites include: urethane dimethacrywhat
brittle,
but
reasonably
strong. Resin strength is
lates (UDMA), with or without Bis-GMA (used in Ivoclar
dependent
upon
monomer
composition,
being greatest
products like Helionnolar, Helioprogress, andTetric); urewhen
stiff
Bis-GMA
molecules
are
used
and
the degree of
thane tetramethacrylate (UTMA) (used in Kuraray prodconversion
(DC)
of
the
methacrylate
groups
is maximized
ucts like Clearfil Photoposterior); bis(methacryloyloxy(Ferracane,
1989).
The
minimal
flexibility
of
the Bis-GMA
methyl)tricyclodecane (used in ESPE-Premier products
molecule
enhances
the
rigidity
of
the
polymer
backbone.
like Pertac Hybrid); ethoxylated bisphenol-A-dimethacryIncreased
cure
results
in
enhanced
cross-links
between
late (BisEMA); and a linear polyurethane made from Bispolymer
chains,
thus
enhancing
the
stiffness
of
the
polyGMA and hexamethylenediisocyanate (used in L.D.
mer
network.
In
addition,
the
further
conversion
of
Caulk products like Prisma APH and TPH) (Ruyter, 1988;
monomer
to
polymer
limits
the
number
of
unreacted
Tani, 1993). The chemical structure of many of these
monomers that may serve as plasticizers in the polymer
monomers is shown in Table 2. Due to their lower vis-

6 ( 4 ) 3 0 2 - 3 1 8 (1995)

Crit Rev Oral Biol Med

309

restorations.
Shrinkage stresses can be reduced,
but not eliminated, by increasing filler
loading. Water sorption by the polymer
network also contributes to a reduction
in the stress (Feilzer et al, 1990). Though
;rnc=c
the shrinkage and build-up of internal
stress are very rapid and take place within minutes in a composite restoration
POLYMERIZATION
(Davidson and De Gee, 1984), water
uptake by resin composite takes place at
a much slower rate, requiring hours or
days to reach saturation (Ferracane and
Condon, 1990). Thus, the delayed nature
of the water sorption minimizes its effect
on stress reduction. In addition, water
sorption may cause erosion of the
filler/matrix interface and a softening of
the polymer network, which can conCROSS-LINKED POLYMER NETWORK
tribute to reductions in strength, stiffness, and wear resistance (Kawaguchi et
al, 1988; Soderholm and Roberts, 1990;
Figure 6. Schematic representation of the polymerization of dimethacrylate monomers
Ferracane et al, 1993). Water sorption
to form the cross-linked polymer network of dental composites containing small
can
be reduced by the use of more
amounts of unreacted monomers and many pendant methacrylate groups (C=C).
hydrophobic monomers, such as the
ethoxylated version of Bis-GMA (BisEMA), which do not contain unreacted hydroxyl groups
matrix (Fig. 6). In present commercial composites, it has
on the main polymer chain (Ruyter and Nilsen, 1993).
been verified by infrared spectrophotometry that 25-55%
of the methacrylate groups remain unreacted after poly(B) RECENT AREAS OF STUDY
merization (Ferracane and Condon, 1992; Ferracane,
1994). An analysis of data on the extraction of unreacted
Recent studies have investigated the effects of incorpospecies from the polymerized material suggests that less
rating different monomers as well as developing entirely
than 1 in 10 of the unreacted molecules is free and capanew monomers for dental composites. The ultimate goal
ble of being released (Ferracane, 1994). The result is that
has been to develop materials with enhanced cure and
nearly 90% of the unreacted methacrylate groups are preenhanced properties, or minimal polymerization shrinksent on pendant molecules which have reacted at one
age.
end by linking with the polymer chain (Fig. 6). These molPeutzfeldt and Asmussen (1991, 1992b) hypotheecules are therefore capable of serving as internal plastisized that cyclic acid anhydrides could serve as crosscizers for the composite.
linking agents for Bis-GMA or UDMA composites to
enhance their strength and abrasion resistance. They
Though DC is maximized by the inclusion of a high
showed that the tensile strength of composites could be
percentage (40-50%) of diluents in the resin, the cure is
increased by a mean of 20% when a 1:1 molar ratio of
accompanied by a significant polymerization shrinkage
maleic anhydride and methacrylamide was added to the
(1.5-3 vol%) for most commercial materials (De Gee et al,
base resin. Though only a few of the formulations
1993). The stress associated with the curing contraction
showed an increase (up to 10%) in flexural strength, and
is one of the most significant problems for current matemost showed a reduction in elastic modulus, modulus of
rials, because it adversely affects the seal at the cavosurresilience was increased for many. In general, the
face margin. This is of paramount importance, because
improvement in properties was greater for the UDMAthe predominant reason for replacement of composites
based resin than for the Bis-GMA-based resin, especially
is secondary caries (Qvist et al, 1990; Maclnnis et al,
when the composites were heat-treated for one hour at
1991). Recently, an experimental composite containing
125-150C after being light-cured, as is commonly done
small concentrations of an antibacterial monomer has
for composite inlays. The properties were tested after
been evaluated and shown to have adequate physical
being aged for only one week in water, however, and a
properties (Imazato and McCabe, 1994). This approach
recent study suggested that the properties of certain
seems worthy of further study in light of the current difheat-treated composites may decline with time due to
ficulty in producing excellent margins with composite

c=cc=
c=cc=c
c=c- c=c c=cc=c c = c _ c , c

c=cc=c

c=cc=c
c=cc=c
c=cc=c

c=cc=c

c=c

310

Crit Rev Oral Biol Med

6(4):302-318(1995)

water sorption (Ferracane et al, 1993).

In similar studies, Peutzfeldt and Asmussen
(1992c,d) showed that adding 20% of bifunctional
ketones, such as diacetylacetone, to UDMA and Bis-GMA
composites caused a significant increase in diametral
tensile strength, flexural strength, and modulus of rupture, presumably by increasing cross-linking. Similar
results were obtained when aldehydes, such as propanal,
were added to these resins (Peutzfeldt and Asmussen,
1992e). Subsequent studies verified that the degree of
cure was increased for these resin systems, though it was
not possible to state whether the improvement in properties was a result of enhanced polymerization or greater
cross-linking (Peutzfeldt, 1994a,b). Likewise, Antonucci et
al (1992) reported some improvement in the flexural
strength of experimental composites by adding benzaldehyde to increase DC by chain-transfer reactions.
Other attempts to increase DC have had the additional aim of reducing polymerization shrinkage.
Stansbury (1990) reported on the synthesis of a series of
cyclopolymerizable monomers described as "oxy bismethacrylates". These dimethacrylate monomers could
be reacted to a DC of 90%, suggesting that they undergo
cyclopolymerization with a 40% reduction in curing contraction compared with TEGDMA resins with similar DC.
When mixed with Bis-GMA and an appropriate filler,
composites with tensile strength comparable with that of
conventional composites were produced. A series of
difunctional oxybismethacrylates, such as the oligomer
of ethoxylated bis-phenol A diacrylate (OEPBA)
(Stansbury, 1992a), has been synthesized and used to
produce light-cured composites with properties similar
to those of ethoxylated Bis-GMA composites. The benefit of these new resin composite systems is that they
show enhanced toughness and less curing shrinkage
than conventional Bis-GMA-based composites. However,
the differences are probably too small to be reflected in
a clinical improvement.
Stansbury and Antonucci (1992) have also synthesized methylene-butyrolactone (MBL), the cyclic analog
of methyl methacrylate, in order to increase the DC and
improve the stability of composites. Because MBL is
more reactive than MMA, the addition of 10 wt% to a BisGMA/TEGDMA resin produced composites with slightly
improved strength and a significant increase in DC.
Furthermore, IR spectroscopy showed that the MBL was
almost completely reacted in the polymerization
process. These results warrant the study of other multifunctional lactones.
In another study, 50-100% increases in elastic modulus were produced in unfilled resins made from copolymers of Bis-GMA and tetrahydrofurfuryl methacrylate
(Davy and Braden, 1991). The authors hypothesized that
the increase in stiffness could be attributed to the smaller diluent molecules occupying free volume spaces with-

6 ( 4 ) : 3 0 2 - 3 l 8 (1995)

in the main polymer network, thus enhancing the molecular interactions within the polymer.
Other studies have addressed the water sorption
concerns of composites. Fluorinated monomers and
oligomers which are more hydrophobic than existing
base monomers have been synthesized (Antonucci et al,
1993). Composites made from fluorinated monomers
have been shown to have greater strength than conventional Bis-GMA-based materials after storage in water,
due to the hydrophobic nature of the fluoride additive
(Kawaguchi et al, 1989). These resins and composites are
not as heavily plasticized by water as a result of the
reduced water sorption. A recently synthesized fluorinated monomer has also been shown to produce composites with improved creep resistance, presumably a result
of its reduced water uptake (Culbertson et al, 1993).
In the past few years, another monomer system has
been developed and commercialized for dental composites. The material, Conquest Crystal (leneric/Pentron), is
based on what is described as a methacrylic ester of an
oligocarbonate polymer that produces a semi-crystalline
polymer matrix (Waknine, 1991). The supposed benefits
of this material are an enhanced toughness and lower
polymerization shrinkage compared with those of BisGMA-based or UDMA-based composites. The effect of
this different composition has not been clinically determined to date.
Because the mechanical properties and clinical success of current materials appear to be dictated more by
the filler component than by the resin matrix, it is questionable whether the monomers developed in these
studies will have a dramatic effect on the clinical performance of dental composites. More than likely, the greatest advances in resin technology will be made in the area
of minimizing polymerization shrinkage and its accompanying stress. Earlier efforts were not successful, but
several recent studies have been more encouraging.
Stansbury (1992b) has synthesized spiro-orthocarbonate monomers (SOCs) which expand during polymerization through a double-ring-opening process. These
monomers contain methylene groups capable of free
radical polymerization, making them useful as additives
to dimethacrylates (Fig. 7). The task is to cause sufficient
double-ring-opening to produce enough expansion to
counter the free radical polymerization shrinkage of the
dimethacrylates, all within a reasonable time frame for
clinical application. This has not occurred to date.
Recently, Stansbury (1992c) reported that spiro orthocarbonate-substituted methacrylates showed nearly complete ring-opening of the SOC when polymerized in
dilute solutions. Less ring-opening was obtained when
the resin was cured in bulk and the composites had
about 1% shrinkage. Miyazaki et al (1994) reported on the
synthesis of acrylates and methacrylates containing
spiro ortho esters that were capable of being polymer-

Crit Rev Oral Bid Ued

31

CH,
CH,

Figure 7. Chemical structure of two of the SOC monomers under

study for non-shrinking dental composites.

ized by heat, ionic and free radical initiators. Though all

common methods of curing were possible, the resultant
polymers were weak, and the reduction in polymerization
shrinkage was not enough to be clinically significant.
Other ring-opening monomers based on vinyl cyclic
acetals have been co-polymerized with Bis-EMA to give
composites with good strength, but shrinkage has not
been reported (Reed et ai, 1992).
Because the reaction of vinyl functional groups (containing C=C) ensures polymerization contraction, others
have taken a different approach with SOCs. Byerley et al.
(1992) and Eicket ai (1992) recently reported on the synthesis of new SOCs polymerized with epoxy resins via
cationic UV photo-initiation. These alicyclic SOCs contain four rings attached to a central spiro carbon, and
expansion is again achieved by a double-ring-opening
mechanism (Fig. 7). The cationic initiator used is (4-octyloxyphenyl)phenyliodonium hexafluoroantimonate, with
chlorothioxanthone as a sensitizer. The mixing of 5% of
the SOC in an epoxy base produced a resin with substantial tensile strength and modulus, acceptable water
sorption and solubility, and a slight expansion. Increased
concentrations of the SOC produced greater expansion
and slightly stronger polymers, but water sorption and
solubility were high due to incomplete reaction of the
SOC. However, the results of these studies are encouraging, and work is continuing.
The development of a dental composite with zero
net dimensional change would be the single most signif312

icant advancement in these materials since their development. Though composites do not have ideal strength,
wear resistance, or durability, significant improvements
in their formulation have produced materials capable of
being used in a wide variety of dental applications.
However, the difficulty encountered in placing the materials, most of which is a direct result of the curing contraction, severely limits their use in many dental practices. The clinician must continually rely on an adhesive
material as an intermediate between the composite and
tooth structure to provide a durable marginal seal.
Though materials have been developed which exhibit
excellent adhesive strength to both enamel and dentin,
the ability for this adhesion to be achieved routinely in
the moist oral cavity is questionable and highly doubtful,
due to the significant stress imposed on the
tooth/restoration bond by the contracting composite filling material. The development of a non-shrinking composite would obviate the need for a very strong adhesive,
one that is capable of resisting the substantial force of
polymerization contraction. For this reason, it is well
worth the effort expended in the development of this
new material. However, it is realistic to expect that the
marketing of commercial dental composites with adequate strength and abrasion resistance and zero shrinkage through the use of expanding monomers remains
several years away.

(IV) Polymerization Promoters and Modifiers

The polymerization of the monomers in a dental composite paste can be accomplished in several ways.
Appropriate chemical species are added to the composite paste to facilitate this process. The reaction must be
activated, however, by an external stimulus. In self-cured
or auto-cured materials, this stimulus is the mixing of
two pastes, one of which contains a chemical activator
and the other of which contains a chemical initiator. For
heat-cured materials, temperatures of 100C or more
provide the stimulus which activates the initiator. For
light-cured systems, the visible-light energy provides the
stimulus to activate the special initiator in the paste.
Each of these mechanisms is efficient in producing a
high degree of cure under the appropriate conditions.
However, several factors directly related to the polymerization promoters affect the rate and extent of the polymerization.
(A) PRESENT MATERIALS

Self- or auto-curing composites were the first types

developed. The chemistry was based on traditional
acrylic systems. Curing is initiated by mixing two pastes,
which brings together the initiator, benzoyl peroxide, and
the activator, an amine such as dihydroxyethyl-p-toluidine (DHEPT), in order to start the polymerization reaction. Self-cure composites lessened in popularity when

Crit Rev Oral Biol Med

6(4):302-318(1995)

light-curing was developed. Light-curing offered a controlled working time and the elimination of the mixing
procedure that took time and incorporated porosity into
the material. In addition, clinical studies have shown
that self-cure composites undergo more darkening than
light-cured composites overtime (Tyas, 1992). However,
self-cure composites may be experiencing a re-emergence.
A technique that was introduced years ago in Japan
has recently been described where the proximal box of a
class II restoration is filled with a self-cured composite
and the occlusal portion is filled with a light-cured composite (Bertolotti, 1991). The rationale is that certain
dentin adhesives may initiate the cure of the self-cure
composite at the interface, causing the polymerization
contraction to occur toward instead of away from the
warm tooth surface. The use of the light-cured composite on the surface provides the benefits of maximum
working time and fewer bubbles in the surface material.
No controlled clinical studies have been presented to
support this procedure to date, but at least one dental
manufacturer has produced a self-curing version of their
composite for this purpose (Bis-fil II, Bisco, Inc).
Self-curing polymerization is also used in dual-cure
composites used as cements for composite and ceramic
inlays. The rationale behind these materials is that any
portion not exposed to the light source will self-cure on
its own. However, the self-cure reaction does not produce as high a DC as the light activation (Caughman and
Rueggeberg, 1992). This may be because the cure of the
monomers is significantly inhibited by the use of low levels of amine/peroxide or high levels of inhibitor (such as
butylated hydroxytoluene) that are incorporated to maximize working time. The clinical significance of this difference in cure is not yet known.
The most popular mode of causing the curing of dental composites is light activation. The initiator in lightactivated dental composites is camphoroquinone (CQ),
which is sensitive to blue light in the 470-nm region of
the electromagnetic spectrum. The reactivity of CQ is
enhanced by the addition of an amine-reducing agent,
such as dimethylamino ethylmethacrylate (DMAEM),
ethyl-4-dimethylaminobenzoate (EDMAB), or N,N-cyanoethyl-methylaniline (CEMA). CQ and amine concentrations vary in commercial composites from 0.2-1.2 wt%
(Taira et al, 1988). Recently, Yoshida and Greener (1994)
systematically varied the ratios of CQ and DMAEM in
unfilled resins and found that maximum cure was produced when the two were used in a 1:1 ratio, as long as
CQ was present at 1.0 mol% or more. However, the incorporation of small fillers of a size that is equivalent to the
wavelength of the curing light (approximately 0.5 jim)
produces composites with poorer light transmission
(Kawaguchi et al, 1994). This has prompted proprietary
modifications in the types and amounts of additives

6 ( 4 ) 3 0 2 - 3 1 8 (1995)

used to enhance overall curing depth. At the same time,

these alterations in composition have produced composites which are more sensitive to the operatory and
dental unit lights. The clinician must be careful to minimize exposure to light until he or she is ready to cure the
material; otherwise, premature polymerization will
begin, and the working time will be reduced to an unacceptable level.
Another common method for curing composites
extra-orally is through the application of heat, either
alone or in conjunction with light-curing. This procedure
is commonly used to cure composite inlays and onlays,
either in the laboratory or at chairside. Many composites
used for direct placement can also be cured with light
and heat or in light-curing ovens to produce these
restorations. Concept (Ivoclar) is a microfilled composite
specifically designed as an inlay/onlay material. It is
cured by being heated to 120C in water under approximately 80 psi of pressure. The pressure produces a dense
material, and the heat activates the benzoyl peroxide catalyst present in the paste. Heat-curing or a post-lightcure heat treatment improves the properties of composites by enhancing cure and/or cross-linking of the polymer (Wendt, 1989; Ferracane and Condon, 1992). Clinical
evaluations have shown improvement in wear resistance
(Leinfelder and Broome, 1994) and resistance to marginal degradation (Wendt and Leinfelder, 1990) as a result of
the heat treatment.
(B)

RECENT AREAS OF STUDY

Though current activity in this area appears to be minimal, the development of new monomer systems will certainly stimulate work to produce agents for maximum
curing efficiency. One such area of study involves the
development of non-shrinking polymers which are polymerized via cationic initiation (Byerley et al, 1992). The
question that arises deals with the compatibility of both
the resins cured via cationic polymerization and those
used in dental adhesives which polymerize via free-radical mechanisms. A recent study by Stansbury et al. (1993)
showed that a monomer with ring-opening potential
could be co-polymerized with Bis-EMA by a free radical
mechanism or a combination of a free radical and cationic mechanism, but did not co-polymerize well under
cationic conditions alone. Extensive studies of the biocompatibility of the cationic initiators will also be
required to prove their safety.
Another emerging area involves the use of lasers to
cure dental composites to a greater extent and to a
greater depth of cure using shorter illumination times.
Theoretically, shorter curing times, even at higher intensity, will also produce less heat build-up in teeth, though
this has never been proven to be a significant clinical
problem. Generally, studies show that composites cured
with an argon laser have slightly greater surface hardness

Crit Rev Oral Biol Med

313

Silane Coupling Agent

C=C-

filler

|-Si

O-Si-(CH)-C=C C=CC=C-

silane

C=CFigure 8. Simplified schematic representation of silane bonding between the fillers

and the polymer matrix in dental composites.

organic polymer matrix (Bowen, 1963).

The linking of these two phases is
brought about by coating the fillers
with a coupling agent that has characteristics of both the filler and the
matrix. The usual coupling agent for
dental composites is a molecule which
has silanol (Si-OH) groups on one end
and methacrylate groups (containing
C=C) on the other end. These molecules are capable of forming covalent
bonds to both the silicon-oxygen
groups in the silica-based fillers and
the methacrylate groups in the resin
matrix (Fig. 8). Studies have verified the
importance of this coupling in dictating
the mechanical properties of the resultant composite (Nishiyama et al, 1991;
Mohsen and Craig, 1995a,b).
(A) PRESENT MATERIALS

and depth of cure than composites cured with a visiblelight source (Blankenau et al, 1991). However, these differences have been less apparent when adequate curing
times of 40-60 seconds are used with efficient visiblelight-curing units (Kurokawa, 1990).
In addition, though peak temperatures can be
reduced with low-power laser curing, considerable heat
is generated when lasers are used at the higher powers,
where they are most efficient for curing (Burtscher, 1991).
Therefore, though some minimal improvements may be
obtained by the use of an argon laser, the results to date
do not demonstrate a rationale for the use of this expensive device for routine curing of composites. In addition,
another factor which sheds doubt on the usefulness of
this rapid-curing technique is the stress produced on the
adhesive bond. It has been suggested that composites
that undergo a slow cure may have enhanced marginal
integrity, because strong bonds to tooth structure can be
secured by the adhesive before the forces of polymerization contraction from the composite reach significant
levels (Unoand Asmussen, 1991). Faster curing produces
a more rapid elevation in stress within the composite as
well as at the interface, which may disrupt the bonds as
they are forming, thus compromising marginal sealing
(LoscheeUL, 1994).

(V) Filler/Matrix-coupling Agents

During the initial development of dental composites, it
was shown that the acquisition of good properties in the
material was dependent upon the formation of a strong
bond between the inorganic filler particles and the

314

Most commercial composites contain

silica-based fillers, and therefore use
silanes with functional methacrylate
groups as filler/matrix-coupling agents. The most common silane is 3-methacryloxypropyl trimethoxysilane
(MPS). Other agents have been experimented with, such
as other silanes, 4-META, and various titanates and zirconates, but none has been as successful as MPS. For
example, MPS was recently compared with APS (3-acryloxypropyltrimethoxysilane) and shown to produce
stronger and more stable composites with quartz or zirconia silicate (Mohsen and Craig, 1995a,b).
(B) RECENT AREAS OF STUDY

Craig and Dootz (1993) reported that a mixture of silanes

may be used to improve the properties of composites.
They added a fluoro-alkyl silane to MPS (1:3) for Zr-silicate fillers. The resultant composites had 20% greater
tensile strength and 60% less water sorption than composites with MPS alone. Recently, Mohsen and Craig
(1995b) demonstrated that silanation at three times the
amount necessary to provide uniform coverage of a zirconia-silica filler gave the best properties.
Further developments in fillers for composites may
necessitate the continued development of coupling
agents. However, based on the tremendous history of the
use of silanes in industrial composites and their reasonably successful use in dental composites to date, it seems
likely that future studies will involve only further refinements of existing silanes in terms of optimal amounts
and modes of coverage (Soderholm and Shang, 1993).

(VI) Summary
It is apparent that current composites have significantly

Crit Rev Oral Biol Ued

6(4):302-318 (1995)

improved clinical performance compared with their predecessors, especially in posterior teeth and in stressbearing areas on anterior teeth. However, it is also apparent that current formulations remain less than ideal in
terms of dimensional change during curing and longterm fatigue and fracture resistance.
Future research efforts will remain committed to the
development of non-shrinking polymer systems which
can be mixed with appropriate curing modifiers and
fillers to produce restorative materials with excellent
qualities. Considerable effort will be expended in the
laboratory and clinic to ensure that the new materials
handle and perform at and above the level of existing
materials. Weaknesses will be identified, and further
refinements in formulations will then take place to maximize performance. The extent of these tests will depend
upon the nature of the candidate monomers, because if
methacrylate systems are used as base monomers with
free-radical catalysts, the efforts will necessarily be less
extensive than if cationic polymerization composites are
produced from non-methacrylate base resins. A full complement of biocompatibility testing of all of the components in the new materials will be required. In the interim, further changes in dental composite formulation will
be minor, probably in the areas of new fillers and grinding processes which produce the necessary sizes of particles in high yields. These particles will ideally be very
fracture-resistant and only minimally abrasive to dentin
and enamel. In addition, their refractive index will be
accurately matched to existing as well as newly developed polymers. Finally, continued development of fillers
and polymers containing therapeutic agents, such as fluoride, is likely to produce composites which may have
cariostatic properties like glass ionomers.

Acknowledgments
The author thanks Mr. }ohn Condon for his efforts in producing the
scanning electron micrographs. A portion of this work was presented at
the Second International Congress on Dental Materials, held in
Honolulu, Hawaii, on November 1-4, 1993. The Congress was jointly
sponsored by the Academy of Dental Materials and The Japanese
Society for Dental Materials and Devices.
REFERENCES

Antonucci ]M, Fowler BO, Venz S (1991). Filler systems

based on calcium metaphosphates. Dent Mater 7:124129.
Antonucci JM, Stansbury JW, Keeny SM, Matsukawa S
(1992). Effect of aldehydes on the mechanical
strength of dental composites (abstract). I Dent Res
72:598.
Antonucci )M, Liu D-W, Stansbury JW (1993). Synthesis of
hydrophobic oligomeric monomers for dental appli-

6(4):302-318 (1995)

cations (abstract). I Dent Res 72(Spec Iss):369.

Arends J, Ruben J (1988). Fluoride release from a composite resin. Quintessence \nt 19:513-514.
Bayne SC, Taylor DF, Wilder AD, Heymann HO, Tangen
CM (1991). Clinical longevity of ten posterior composite materials based on wear (abstract). ) Dent Res
70:344.
Bayne SC, Taylor DF, Heymann HO (1992). Protection
hypothesis for composite wear. Dent Mater 8:305-309.
Bayne SC, Heymann HO, Swift El (1994). Update on dental composite restorations. I Am Dent Assoc 125:687701.
Beatty MW, Swartz ML, Moore BK, Phillips RW, Roberts
TA (1993). Effect of crosslinking agent content,
monomer functionality, and repeat unit chemistry on
properties of unfilled resins. 1 Biorned Mater Res 27:403413.
Bertolotti R (1991). Posterior composite technique utilizing directed polymerization shrinkage and a novel
matrix. Pract Perio Aesth Dent 3:53-58.

Blankenau RJ, Kelsey WP, Powell GL, Shearer GO,

BarkmeierWW, Cavel WT (1991). Degree of composite
resin polymerization with visible light and argon
laser. Aust Dent J 4:40-42.
Bowen R (1963). Properties of silica reinforced polymer
for dental restoration. 1 Am Dent Assoc 66:57-64.
Bowen RL, Eichmiller FC, Marjenhoff WA (1991). Glassceramic inserts anticipated for "megafilled" composite restorations. ) Am Dent Assoc 122:71-75.
Braem M, Ringer W, Van Doren VE, Lambrechts P,
Vanherle G (1989). Mechanical properties and filler
fraction of dental composites. Dent Mater 5:346-349.
Braem MJA, Davidson CL, Lambrechts P, Vanherle G
(1994a). \n vitro flexural fatigue limits of dental composites. I Biomed Mater Res 28:1397-1402.
Braem M, Lambrechts P, Vanherle G (1994b). Clinical relevance of laboratory fatigue studies. I Dent 22:97-102.
Bryant RW, Marzbani N, Hodge KV (1992). Occlusal margin defects around different types of composite resin
restorations in posterior teeth. OperDent 17:215-221.
Burtscher P (1991). Curing of composites with an argon
laser (abstract). ] Dent Res 70(Spec lss):526.
Byerley T), Eick ID, Chen GP, Chappelow CC, Millich F
(1992). Synthesis and polymerization of new expanding dental monomers. Dent Mater 8:345-350.
Caughman W, Rueggeberg F (1992). Monomer conversion
in dual cure resin adhesives (abstract). I Dent Res
71 (Spec Iss): 160.
Chung K, Greener EH (1990). Correlation between degree
of conversion, filler concentration and mechanical
properties of posterior composite resins. I Oral Rehabil
17:487-494.
Craig RG, Dootz ER (1993). Effect of mixed silanes on
strength and sorption of composites (abstract). I Dent
Res 72(Spec Iss):383.

Crit Rev Oral Biol Med

315

Culbertson BM, Sang J, Brantley WAr Lo SK (1993). New

dental resins with improved creep resistance
(abstract). ] Dent Res 72(Spec Iss):178.
Dagani R (1992). Nanostructured materials promise to
advance range of technologies. Chem Eng News 70:1824.
Davidson CL, De Gee AJ (1984). Relaxation of polymerization contraction stresses by flow in dental composites. J Dent Res 63:146-148.
Davy KW, Braden M (1991). Study of polymeric systems
based
on
2,2-bis-4(2-hydroxy-3-methacryloyloxypropoxy)phenylpropane. Biomater 12:406-410.
De Gee AJ, Feilzer AJ, Davidson CL (1993). True linear
polymerization shrinkage of unfilled resins and composites determined with a linometer. Dent Mater 9:1114.
Dijkman GEHM, de Vries J, Lodding A, Arends J (1993).
Long-term fluoride release of visible light-activated
composites in vitro: a correlation with in situ demineralization data. Caries Res 27:117-123.
Drummond JL (1989). Cyclic fatigue of composite
restorative materials. J Oral Rehabil 16:509-520.
Eick JD, Byerley TJ, Chappell RP, Chen GP, Bowles CQ,
Chappelow CC (1992). Properties of expanding
SOC/epoxy copolymers for dental use in dental composites. Dent Mater 9:123-127.
el-Mowafy OM, Lewis DWF Benmergui C, Levinton C
(1994). Meta-analysis on long-term clinical performance of posterior composite restorations. J Dent
22:33-43.
Feilzer AJ, deGee AJ, Davidson CL (1990). Relaxation of
polymerization contraction shear stress by hygroscopic expansion. J Dent Res 69:36-39.
Ferracane JL (1989). In vitro evaluation of composite
resins.
Structure-property
relationships.
Development of assessment criteria. Trans head Dent
Mater 2:6-35.
Ferracane JL (1994). Elution of teachable components
from composites. J Oral Rehabil 21:441-452.
Ferracane JL, Antonio RC, Matsumoto H (1987). Variables
affecting the fracture toughness of dental composites. J Dent Res 66:1140-1145.
Ferracane JL, Condon JR (1990). Rate of elution of leachable components from composite. Dent Mater 6:282287.
Ferracane JL, Condon JR (1992). Post-cure heat treatments for composites: properties and fractography.
Dent Mater 8:290-295.
Ferracane JL, Greener EH (1986). The effect of resin formulation on the degree of conversion and mechanical
properties of dental restorative resins. J Biomed Mater
Res 20:121-131.
Ferracane JL, Condon JR, Suh B (1992). Effect of filler on
degree of conversion (DC) of resins (abstract). J Dent
Res 72(Spec Iss):598.

316

Ferracane JL, Hopkin JK, Condon JR (1993). The properties of heat-treated composites after aging (abstract).
I Dent Res 72(Spec lss):135.
George LA, Richards ND (1993). Polymerization shrinkage in a composite restoration involving a glassceramic insert (abstract). J Dent Res 72(Spec Iss).-351.
Hosoda H, Yamada T, Inokoshi S (1990). SEM and elemental analysis of composite resins. J Prosthet Dent
64:669-676.
Iga M, Takeshige F, Ui T, Torii M, Tsuchitani Y (1991). The
relationship between polymerization shrinkage measured by a modified dilatometer and the inorganic
filler content of light-cured composites. Dent Mater ]
10:38-45.
Imazato S, McCabe JF (1994). Influence of incorporation
of antibacterial monomer on curing behavior of a
dental composite. I Dent Res 73:1641-1645.
Kawaguchi M, Fukushima T, Horibe T (1988). Effect of
monomer structure on the mechanical properties of
light-cured unfilled resin. Dent Mater] 7:174-181.
Kawaguchi M, Fukushima T, Horibe T (1989). Effect of
monomer structure on the mechanical properties of
light-cured composite resins. Dent Mater] 8:40-45.
Kawaguchi M, Fukushima T, Miyazaki K (1994). The relationship between cure depth and transmission coefficient of visible-light-activated resin composites. J
Dent Res 73:516-521.
Kawai K, Leinfelder KF (1993). Effect of glass inserts on
wear of composite resins (abstract). J Dent Res 72(Spec
Iss):114.
Khan AM, Suzuki H, Normura Y, Taira M, Wakasa K,
Shintani H, et al (1992). Characterization of inorganic
fillers in visible-light-cured dental composite resins. I
OralRehabil 19:361-370.
Kim KH, Park JH, Imai Y, Kishi T (1994). Microfracture
mechanisms of dental resin composites containing
spherically-shaped filler particles. J Dent Res 73:499504.
Knibbs PJ, Smart ER (1992). The clinical performance of a
posterior composite resin restorative material,
Heliomolar R.O.: 3-year report. J Oral Rehabil 19:231237.
Kurokawa M (1990). Study on light-cured composite
resins: consideration of the polymerization characteristics of composites cured by argon ion laser. J Nihon
Univ Sch Dent 32:48.

Lambrechts P, Vanherle G, Vuylsteke M, Davidson CL

(1984). Quantitative evaluation of the wear resistance
of posterior dental restorations: a new three-dimensional measuring technique. J Dent 12:252-267.
Leinfelder KF (1995). Posterior composite resins: the
materials and their clinical performance. J Am Dent
Assoc 126:663-676.
Leinfelder KF, Broome JC (1994). ]n vitro and in vivo evaluation of a new universal composite resin. J Esthet Dent

Crit Rev Oral Biol Ued

6(4):302-318 (1995)

6:177-183.
Li X, Edwards K, Yee AF (1993). Reducing cure stress and
improving wear resistance in dental composites
(abstract). I Dent Res 72(Spec Iss): 113.
Losche AC, Losche GM, Roulet JF (1994). Effects of
increased light intensity on the marginal quality of
class II composite fillings. Dtsch Zahnarztl Z 49:590-594.
Lutz F, Phillips RW (1983). A classification and evaluation
of composite resin systems. I Prosthet Dent 50:480-488.
Maclnnis WA, Ismail A, Brogan H (1991). Placement and
replacement of restorations in a military population.
Can DentAssoc] 57:227-230.
Mazer RB, Leinfelder KF (1992). Evaluating a microfill
posterior composite resin: a five-year study.) Am Dent
Assoc 123:32-37.
Mazer RB, Leinfelder KF, Russell CM (1992). Degradation
of microfilled posterior composite. Dent Mater 8:185189.
Miyazaki K, Takata T, Endo T, Inanaga A (1994). Thermaland photo-polymerization of (meth)acrylates containing a spiro ortho ester moiety and the properties
of poly|(meth)acrylate]s. Dent Mater] 13:9-18.
Mohsen NM, Craig RG (1995a). Effect of silanation of
fillers on their dispersability by monomer systems. I
OralRehabil 22:183-189.
Mohsen NM, Craig RG (1995b). Hydrolytic stability of
silanated zirconium silica-urethane dimethacrylate
composites. I Oral Rehabil 22:213-220.
Munksgaard EC, Hansen EK, Kato H (1987). Wall-to-wall
polymerization contraction of composite resins versus
filler content. Scand ] Dent Res 95:526-531.
Nishiyama N, Ishizaki T, Horie K, Tomari M, Someya M
(1991). Novel polyfunctional silanes for improved
hydrolytic stability at the polymer-silica interface. J
Biomed Mater Y<es 25:213-221.

0ysaed H, Ruyter IE (1986). Water sorption and filler

characteristics of composites for use in posterior
teeth. J Dent Res 65:1315-1318.
Peutzfeldt A (1994a). Quantity of remaining double
bonds of diacetyl-containing resins. I Dent Res 73:511515.
Peutzfeldt A (1994b). Quantity of remaining double
bonds of propanal-containing resins. I Dent Res
73:1657-1662.
Peutzfeldt A, Asmussen E (1991). Influence of carboxylic
anhydrides on selected mechanical properties of
heat-cured resin composites. I Dent Res 70:1537-1541.
Peutzfeldt A, Asmussen E (1992a). Modulus of resilience
as predictor for clinical wear of restorative resins. Dent
Mater 8:146-148.
Peutzfeldt A, Asmussen E (1992b). Effect of temperature
and duration of post-cure on selected mechanical
properties of resin composites containing carboxylic
anhydrides. Scand I Dent Res 100:296-298.
Peutzfeldt A, Asmussen E (1992c). Influence of ketones

6(41:302-318(1995)

on selected mechanical properties of resin composites. I Dent Res 71:1847-1850.

Peutzfeldt A, Asmussen E (1992d). Ketones in resin composites. Effect of ketone content and monomer composition on selected mechanical properties. Acta
Odontol Scand 50:253-258.

Peutzfeldt A, Asmussen E (1992e). Influence of aldehydes

on selected mechanical properties of resin composites. I Dent Res 71:1522-1524.
Phillips RW, Editor (1991). Skinner's science of dental
materials. 9th ed. Philadelphia, PA: W.B. Saunders
Co., pp. 215-248.
Qvist V, Qvist I, Mjor IA (1990). Placement and longevity
of tooth-colored restorations in Denmark. Acta Odontol
Scand 48:305-311.
Reed B, Stansbury J, Antonucci I (1992). Ring-opening
dental resin systems based on cyclic acetals
(abstract). ) Dent Res 71 (Spec Iss):276.
Rueggeberg FA, Jordan DM (1993). Effect of light-tip distance on polymerization of resin composite. \nt J
Prosthodont 6:364-370.

Ruyter IE (1988). Composites-characterization of composite filling materials: reactor response. Adv Dent Res
2:122-129.
Ruyter IE, Nilsen J (1993). Chemical characterization of
six posterior composites (abstract). 1 Dent Res 72(Spec
Iss):177.
Ruyter IE, 0ysaed H (1987). Composites for use in posterior teeth: Composition and conversion. J Biomed Mater
Res 21:11-23.
Seghi RR, Sang J, Hayes D, Culbertson BM (1993).
Relative abrasion rates of sol-gel derived dual matrix
composites. ] Dent Res 72(Spec Iss): 113, Abstr. No. 80.
Soderholm K-JM (1984). Influence of silant treatment and
filler fraction on thermal expansion of composite
resins. J Dent Res 63:1321-1326.
Soderholm K-JM (1990). Filler teachability during water
storage of six composite materials. Scand ) Dent Res
98:82-88.
Soderholm K-JM, Achanta S, Olsson S (1993). Variables
affecting the depth of cure of composites (abstract). J
Dent Res 72(Spec Iss): 138.
Soderholm KM, Roberts MJ (1990). Influence of water
exposure on the tensile strength of composites. J Dent
Res 69:1812-1816.
Soderholm K-JM, Shang S-W (1993). Molecular orientation of silane at the surface of colloidal silica. J Dent
Res 72:1050-1054.
Soderholm K-JM, Zigan M, Ragan M, Fischlschweiger W,
Bergman M (1984). Hydrolytic degradation of dental
composites. I Dent Res 63:1248-1254.
Stangel I, Barolet RY (1990). Clinical evaluation of two
posterior composite resins: two-year results. J Oral
Rehabil 17:257-268.
Stannard JG, Sornkul E, Collier R (1993). Mechanical

Crit Rev Oral Biol Med

317

properties of composite resin co-polymers (abstract).

) Dent Res 72(SpecIss):135.
Stansbury JW (1990). Cyclopolymerizable monomers for
use in dental resin composites. I Dent Res 69:84-848.
Stansbury JW (1992a). Synthesis and evaluation of novel
multifunctional oligomers for dentistry. J Dent Res
71:434-437.
Stansbury JW (1992b). Spiro orthocarbonate-substituted
methacrylates: New monomers for ring-opening polymerization (abstract). I Dent Res 71:239.
Stansbury JW (1992c). Synthesis and evaluation of new
oxaspiro monomers for double ring-opening polymerization. J Dent Res 71:1408-1412.
Stansbury JW, Antonucci JM (1992). Evaluation of methylene lactone monomers in dental resins. Dent Mater
8:270-273.
Stansbury JWr Antonucci JM, Reed BB (1993). Initiator
effects on the tensile strength of novel dental composites (abstract). J Dent Res 72(Spec Iss):385.
Suzuki H, Taira M, Wakasa K, Yamaki M (1991).
Refractive-index-adjustable fillers for visible-lightcured dental resin composites: preparation of TiO2SiO2 glass powder by the sol-gel process. J Dent Res
70:883-888.
Suzuki S, Leinfelder KF (1994). An in vitro evaluation of a
copolymerizable type of microfilled composite resin.
Quintessence \nt 25:59-64.
Taira M, Urabe H, Hirose T, Wakasa K, Yamaki M (1988).
Analysis of photo-initiators in visible-light-cured dental composite resins. J Dent Res 67:24-28.
Takahashi K, Emilson CG, Birkhed D (1993). Fluoride
release in vitro from various glass ionomer cements
and resin composites after exposure to NaF solutions. Dent Mater 9:350-354.
Tani Y (1983). A new composite resin "Palfique"characteristics and clinical use. Dent Outlook 62:887-896.
Tani Y (1993). New technology of composite resins developed in Japan. Trans Second \nt Cong Dent Mater, pp. 5461.
Taylor DF, Bayne SC, Leinfelder KF, Davis S, Koch GG
(1994). Pooling of long term clinical wear data for
posterior composites. Am J Dent 7.167-174.
Tyas MJ (1990). Correlation between fracture properties

318

and clinical performance of composite resins in class

IV cavities, hustr Dent J 35:46-49.
Tyas MJ (1992). Colour stability of composite resins: a
clinical comparison, hustr Dent J 37:88-90.
Tyas JM, Wassenaar P (1991). Clinical evaluation of four
composite resins in posterior teeth. Five-year results.
hustr Vent] 36:369-373.
Tyas MJ, Truong VT, Goldman M, Beech DR (1989).
Clinical evaluation of six composite resins in posterior teeth, hustr Dent J 34:147-153.
Uno S, Asmussen E (1991). Marginal adaptation of a
restorative resin polymerized at reduced rate. Scand J
Dent Res 99:440-444.
van Dijken JWV, Wing KR, Ruyter IE (1989). An evaluation
of the radiopacity of composite restorative materials
used in class I and class II cavities, hcta Odontol Scand
47:401-407.
Waknine S (1991). Conquest DFC: a novel universal dental composite restorative system. Esthet Dent Update
2:70-72.
Wendt SL (1989). Time as a factor in the heat curing of
composite resins. Quintessence \nt 20:259-263.
Wendt SL, Leinfelder KF (1990). The clinical evaluation of
heat-treated composite resin inlays. J Am Dent hssoc
120:177-181.
Wendt SL, Leinfelder KF (1992). Clinical evaluation of
Clearfil photoposterior and bonding system: 3-year
results. Am J Dent 5:121-125.
Wendt SL, Leinfelder KF (1994). Clinical evaluation of a
posterior resin composite: 3-year results. Am J Dent
7:207-211.
Willems G, Lambrechts P, Braem M, Celis JP, Vanherle G
(1992). A classification of dental composites according to their morphological and mechanical characteristics. Dent Mater 8:310-319.
Wilson NHF, Wilson MA, Wastel DG, Smith GA (1991).
Performance of Occlusin in butt-joint and bevel-edged
preparations: five-year results. Dent Mater 7:92-98.
Yamaguchi R, Powers JM, Dennison JB (1989). Thermal
expansion of visible-light-cured composite resins.
OperDent 14:64-67.
Yoshida K, Greener EH (1994). Effect of photoinitiator on
degree of conversion of unfilled light-cured resin. J

Crit Rev Oral Biol Ued

6(4):302-318(1995)