RESEARCH REPORTS
Biomaterials & Bioengineering
F.R. Tay1*, S.K. Sidhu2,
T.F. Watson3, and D.H. Pashley4
1PediatricDentistry and Orthodontics, Faculty of Dentistry,
The University of Hong Kong, Prince Philip Dental
Hospital, 34 Hospital Road, Hong Kong SAR, China;
2 Restorative Dentistry, School of Dental Science,
University of Newcastle, Newcastle upon Tyne, UK;
3Department of Conservative Dentistry & Biomaterials,
Guy's, King's & St Thomas Dental Institute, King's College
London, Guy's Hospital, London, UK; and 4Department of
Oral Biology and Maxillofacial Pathology, Medical College
of Georgia, Augusta, GA, USA; *corresponding author,
kfctay@netvigator.com
J Dent Res 83(8):644-649, 2004
ABSTRACT
The function of the interfacial transition zone
(absorption layer) in resin-modified glass-ionomer
cements bonded to deep dentin remains obscure.
This study tested the hypotheses that the
absorption layer is formed only in the presence of
water derived from hydrated dentin and allows for
better bonding of resin-modified glass-ionomer
cements to dentin. Ten percent polyacrylic acid-
conditioned, hydrated, and dehydrated deep dentin
specimens were bonded with 2 resin-modified
glass-ionomer cements and sealed with resins to
prevent environmental water gain or loss. A non-
particulate absorption layer was identified over
hydrated dentin only, and was clearly discernible
from the hybrid layer when bonded interfaces
were examined with transmission electron
microscopy. This layer was relatively more
resistant to dehydration stresses, and remained
intact over the dentin surface after tensile testing.
The absorption layer mediates better bonding of
resin-modified glass-ionomer cements to deep
dentin, and functions as a stress-relieving layer to
reduce stresses induced by desiccation and
shrinkage.
KEY WORDS: RMGIC, absorption layer, hydrated
dentin, water movement.
Received May 28, 2003; Last revision May 25, 2004;
Accepted June 2, 2004
644 Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.
International and American Associations for Dental Research
Water-dependent Interfacial
Transition Zone in Resin-modified
Glass-ionomer Cement/Dentin
Interfaces
INTRODUCTION
Treported
he presence of a resin-rich, interfacial transition zone in dentin bonded
with some resin-modified glass-ionomer cements (RMGICs) was first
by Watson et al. (1994) and subsequently by Pereira et al. (1997).
This amorphous, non-particulate zone was termed the 'absorption layer'
(Sidhu and Watson, 1998; Sidhu et al., 1999) to represent its association
with water movement within the maturing RMGICs when these materials
were placed in deep, moist dentin.
Confocal scanning optical microscopy is a valuable tool for
demonstrating dynamic water fluxes between RMGICs and dentin (Watson
et al., 1998), and the use of fluorescent dyes enables the absorption layer to
be imaged (Sidhu et al., 2002). However, resolution of the hybrid layer that
is created when dentin is conditioned with polyacrylic acids (Nakanuma et
al., 1998; Abdalla, 2000) is beyond the scope of light microscopy.
Moreover, the clinical implication of the absorption layer in RMGIC/dentin
interfaces remains unknown.
Since RMGIC absorption layers were observed predominantly in deep
hydrated dentin (Sidhu and Watson, 1998), the objective of this study was to
examine the ultrastructural characteristics of the absorption layer created
with two RMGICs in polyacrylic-acid-conditioned dentin. The hypotheses
examined were that the absorption layer is formed only in the presence of
water derived from hydrated dentin, and that the absorption layer facilitates
better bonding of RMGICs to deep dentin.
MATERIALS & METHODS
Non-carious human third molars were collected after the patients' informed
consent had been obtained under a protocol reviewed and approved by the
institutional review board of the Medical College of Georgia. The occlusal
enamel and superficial dentin were removed by means of a slow-speed saw
(Isomet, Buehler Ltd., Lake Bluff, IL, USA) under water cooling. Smear layers
were created in deep coronal dentin with wet 180-grit silicon carbide paper.
Experimental Design
Bonding surfaces were treated with 10% polyacrylic acid (Dentin Conditioner,
GC Corp., Tokyo, Japan) for 20 sec, then were rinsed thoroughly (Tanumiharja
et al., 2000). The dentin substrates were divided into 2 categories: hydrated and
dehydrated dentin. For hydrated dentin, the teeth were bonded in their normal
hydrated status with their roots intact (i.e., with the pulp and dentinal tubules
filled with water). To determine if the absorption layer could form in dehydrated
dentin, we severed the roots to remove the contents of the pulp chamber. After
being acid-conditioned, these teeth were dehydrated through an ascending
ethanol series (70%, 80%, 95%, three changes in 100%) for 2 hrs each prior to
being bonded.
Two light-cured, machine-mixed RMGICs (Fuji II LC, GC Corp.; Photac-
Fil Quick Aplicap, 3M-ESPE, St. Paul, MN, USA) were investigated. They were
J Dent Res 83(8) 2004 RMGIC-Dentin Interaction 645

applied to dentin in 2 two-mm-thick

layers and light-cured within 60 sec of
being mixed, to minimize the release
of the 2-hydroxyethyl methacrylate
(HEMA) component (Palmer et al.,
1999). All of the teeth that had been
bonded, including the roots or
exposed pulps, were immediately
sealed with 2 coats of light-cured
surface sealant (BisCover, Bisco,
Schaumburg, IL, USA) to prevent
either desiccation or water-sorption
via external sources (Chuang et al.,
2001). Thus, the only available water
source in hydrated specimens was of
pulpal origin, accessible via the
dentinal tubules. The specimens were
stored at 37°C and 100% relative
humidity for 24 hrs before being
processed further.
Transmission Electron
Microscopy
Five specimens were used for each of
the 4 groups investigated (i.e., 2
RMGICs bonded to either hydrated or
dehydrated dentin). Two blocks (2 x 2
x 2 mm) containing the RMGIC/dentin
interfaces were retrieved from each
tooth and sealed immediately with
BisCover resin. The specimen blocks
were embedded in epoxy resin for ease
of handling during ultramicrotomy,
Figure 1. Transmission electron micrographs and elemental analyses of the application of Fuji II LC to
according to the protocol described by hydrated dentin (A-C) and dehydrated dentin (D). Both types of substrates were conditioned with GC
Tay et al. (2001a). Unstained sections Dentin Conditioner (10% polyacrylic acid) for 20 sec before placement of the RMGIC. In dehydrated
(90 nm thick) were examined with a dentin, conditioning was performed prior to the dehydration protocol. (A) A 7- to 10-µm-thick
transmission electron microscope absorption layer (AL) can be seen between the partially demineralized dentin (i.e., hybrid layer [H];
between open arrowheads) and the RMGIC. Phase separation (arrow) can also be observed within the
(Philips EM208S, Eindhoven, The resin matrix (RM). D: mineralized dentin. (B) A high-magnification view showing electron-dense,
Netherlands) operating at 80 kV. multilocular phases (arrow) within the absorption layer (AL). A 2-µm-thick, partially demineralized
We used energy-dispersive x-ray hybrid layer (H; between open arrowheads) is present on top of the mineralized dentin (D). (C) Energy-
analysis further to determine the dispersive x-ray analysis results comparing the elemental distribution and their relative concentrations
in the absorption layer and the multilocular phases found in Fuji II LC. Cu originated from the copper
elemental composition of structural
specimen grids. (D) When bonded to dehydrated dentin, the RMGIC was in direct contact with the
phases that were initially identified, surface of the hybrid layer (H; between open arrowheads). No absorption layer could be identified.
similar to the protocol described by Multilocular phases were absent from the resin matrix (RM). D: mineralized dentin.
Hatton and Brook (1992) for
elemental analysis of glass-ionomer
cements. Grids were further coated
with carbon, and spot analyses were performed with another followed by 1 ␮m alumina (Buehler Ltd.). These surfaces were
microscope (Philips Technai 12) equipped with an x-ray analyzer brought into relief by being etched with 10% phosphoric acid (Bisco
(EDAX Inc., Mahwah, NJ, USA) at 80 kV. Inc.). They were placed on the Peltier (cooling) stage of a field-
To locate potential capillary pore spaces within RMGICs that emission/environmental-scanning electron microscpe (Philips XL-30
may harbor retained water (Yap and Lee, 1997), we immersed ESEM-FEG) and examined at 20 kV wet and without being coated.
additional 2-mm-thick slabs in 50 wt% ammoniacal silver nitrate, The temperature was fixed at 4°C, and the vapor pressure of the
according to the silver impregnation protocol reported by Tay and specimen chamber was maintained at 6.1 Torr to achieve a 100%
Pashley (2003). After reduction of the diamine silver ion relative humidity (Stokes et al., 2002).
complexes to metallic silver, the specimens were processed in the To evaluate the effects of dehydration stresses on
manner previously described. RMGIC/dentin interfaces, we further examined the wet slabs at
different humidities (95%, 90%, 85%, and 75%) by a gradual
Environmental Scanning Electron Microscopy reduction of the chamber pressure, following the method described
Another 2-mm-thick slab was produced from each bonded tooth and by Neubauer and Jennings (1996) for examination of water-based
polished under wet conditions with 1200-grit silicon carbide paper, Portland cements. This enabled us to determine specific sites along

Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.

International and American Associations for Dental Research

646 Tay et al. J Dent Res 83(8) 2004

the absorption layer (Fig. 2A) and

around the glass fillers in the resin
matrices (Fig. 2B). Unlike air voids,
these capillary pore spaces were
evident only after silver
impregnation. A siliceous hydrogel
layer, indicative of GI reaction (Tay
et al., 2001b), could also be
identified around the periphery of
the basic glass fillers (Fig. 2B).
The absorption layer could also
be clearly observed from
environmental scanning electron
microscopic images of RMGICs
bonded to hydrated dentin. Intact
interfaces were seen in specimens
Figure 2. Transmission electron micrographs showing silver deposition within the interfaces of Fuji II LC that were examined at 100%
bonded to normal hydrated dentin, after immersion in ammoniacal silver nitrate. (A) An intricate pattern relative humidity (Fig. 3A). Micro-
of water channels (water trees) (arrows) can be identified within the absorption layer (AL). H: hybrid cracks began to appear between the
layer. (B) Extension of the water trees (pointer) around the glass fillers (G) of the RMGIC. A siliceous absorption layer and the RMGICs at
hydrogel layer (open arrow) can be seen along the periphery of the glass fillers. Multilocular phases
(arrow) within the resin matrix are devoid of silver deposition. Additional unconnected silver grains are 90% relative humidity (Fig. 3B),
identified (open arrowhead). developing into larger continuous
cracks at 85% relative humidity
(not shown). Although vertical
dehydration cracks began to form
the interfaces that are susceptible to micro-crack initiation. within the absorption layer at 75%
relative humidity (Fig. 3C), integrity was still maintained
Microtensile Bond Testing and Fractographic Analysis between the absorption layer and the hybrid layer. Conversely,
Five teeth were used for each group and bonded with the continuous cracks were seen adjacent to the hybrid layer in
respective RMGICs. They were sectioned into 0.9 x 0.9 mm beams specimens bonded to dehydrated dentin, even at 100% relative
containing the interfaces. From the array of beams sectioned from humidity (Fig. 3D), indicating that these were true gaps that
each tooth, 3 beams, each located at 2 mm central to the dentino- developed autogenously before the specimens were subjected
enamel junction along the corner of the array, were selected for to dehydration stresses.
bond testing. Twenty beams from each group were subjected to Microtensile bond testing (Fig 4A) showed significant
tensile stress until failure, with the use of a universal testing differences between specimens that were bonded to hydrated
machine (Model 4440; Instron Inc., Canton, MA, USA) at a dentin and those bonded to dehydrated dentin (P < 0.001).
crosshead speed of 1 mm/min. The data were analyzed by Kruskal- Premature failure occurred in all 20 of the beams of Fuji II LC
Wallis one-way ANOVA on ranks and by Dunn's multiple- and in 19 of the beams of Photac-Fil Quick that were bonded to
comparison tests at ␣ = 0.05. dehydrated dentin. Fractographic analysis revealed the presence
The dentin sides of 4 fractured beams were air-dried and of artifactual cracks that were expected from air-dried, water-
sputter-coated with gold/palladium for examination with a containing specimens. In specimens bonded to hydrated dentin,
scanning electron microscope (Cambridge Stereoscan 360, fractured RMGICs were identified on the surfaces of
Cambridge, UK) operating at 15 kV. absorption layers (Fig. 4B). Conversely, in specimens bonded
to dehydrated dentin, remnant RMGICs were seen along the
RESULTS surface of the hybrid layer, and in specimens with the dentinal
tubules exposed by polyacrylic acid-etching (Fig. 4C).
RMGICs bonded to hydrated dentin revealed 7- to 10-µm-
thick, non-particulate absorption layers between the partially
demineralized hybrid layers and the particulate cements (Fig. DISCUSSION
1A). Both the resin matrix and the absorption layer (Fig. 1B) in Since the absorption layer was absent in dehydrated deep
Fuji II LC contained additional multilocular phases that were dentin, this supports the use of the term 'absorption layer' to
absent in Photac-Fil Quick. Elemental analyses revealed that reflect its genesis with water derived from hydrated dentin. In
calcium was present in both the absorption layer and the its original context, absorption layer formation was attributed to
multilocular phases. The relative concentrations of silicon and swelling of the poly(HEMA) component of the resin matrix via
aluminum in the multilocular phases were higher than those in water sorption (Sidhu and Watson, 1998). Such a hypothesis,
the absorption layer (Fig. 1C), but lower than those in the resin however, could not explain why it was absent from enamel, and
matrix (not shown). The absorption layer was absent when both from the finished surfaces of RMGICs that were subsequently
RMGICs were bonded to dehydrated dentin, with the cement in exposed to water. The results of this study further suggested
direct contact with the hybrid layer (Fig. 1D). that the absorption layer is formed as an integral part of the
In specimens that were immersed in ammoniacal silver RMGIC/dentin interface, when RMGICs were coupled to deep
nitrate, silver-filled water trees (Tay and Pashley, 2003) hydrated dentin. Since HEMA is a major constituent in these
representing capillary pore spaces could be identified within RMGICs, 'extractive bleeding' of HEMA may occur from the
Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.

International and American Associations for Dental Research

J Dent Res 83(8) 2004 RMGIC-Dentin Interaction 647

developing glass-ionomer matrices,

in the presence of water derived
from the underlying dentin.
Polymerization of the diluted
aqueous HEMA solution, when the
RMGICs were light-cured, may
result in the formation of a porous,
non-particulate, poly(HEMA)-rich
hydrogel layer (Chen et al., 1993;
Chirila et al., 1993) that contains
some of the ions released by the
initial acid-base glass-ionomer
reaction between the fluoro-
aluminosilicate glass particles and
polyacids (Fig. 1C).
In this study, the RMGICs were
stored under isolated conditions,
with no possible water gain or loss
to the external environment. When
cured under similar conditions,
conventional GICs exhibited
spontaneous fracture due to the
development of internal stresses
Figure 3. Field-emission/environmental scanning electron microscopical images of Photac-Fil Quick
(Feilzer et al., 1995). Although bonded to hydrated dentin (A-C) and dehydrated dentin (D) and examined at different relative
autogenous shrinkage occurred in humidities produced by adjustment of the vapor pressure of the environmental chamber and with the
RMGICs (Watts et al., 2000), no temperature maintained at a constant temperature of 4°C. (A) Hydrated dentin at 100% relative
spontaneous fractures were humidity (6.1 Torr). A 7- to 10-µm-thick absorption layer (AL) is present between the RMGIC (C) and
the dentin hybrid layer (H). No gap is present along the entire interface. D: dentin. (B). At 90% relative
observed (Feilzer et al., 1995). In humidity (5.6 Torr), dehydration cracks begin to form between the RMGIC and the absorption layer
this study, spontaneous fractures (arrows). (C) At 75% relative humidity (4.8 Torr), apart from continuing enlargement of existing cracks,
occurred only when the RMGICs vertical cracks begin to form within the absorption layer (pointers), and between the glass filler particles
were bonded to dehydrated dentin and the resin matrix (open arrowhead). (D) Dehydrated dentin at 100% relative humidity (6.1 Torr). No
absorption layer can be identified. A gap (pointer) is present between the RM-GIC (C) and dentin.
(Fig. 3D), in the absence of the
Fractured glass filler particles (arrow) are seen adjacent to the dentin hybrid layer (H).
absorption layer (Figs. 1D), and
with direct contact of remnant
RMGICs with the surface of the
hybrid layer (Fig. 4C). This
confirmed our hypotheses that the
absorption layer is formed only in
the presence of water derived from
hydrated dentin, and is crucial for
mediating the bond between
RMGICs and dentin. The RMGIC
absorption layer has been thought to
act as a stress-breaking layer (Sidhu
et al., 2002) and may provide a
function similar to that of a dentin
adhesive layer in relieving Figure 4. (A) Microtensile bond test results of Fuji II LC and Photac-Fil Quick applied to hydrated dentin
polymerization shrinkage stresses and dehydrated dentin. Twenty beams derived from 5 third molars were used for each group (N = 20).
(Ausiello et al., 2002). Such a layer Means ± standard deviation for the 4 groups are: 17.6 ± 4.1 MPa (Fuji II LC, hydrated dentin), 0.0 ±
0.0 MPa (Fuji II LC, dehydrated dentin), 18.5 ± 4.9 MPa (Photac-Fil Quick, hydrated dentin), and 0.3 ±
was most prominent when dentinal 1.4 MPa (Photac-Fil Quick, dehydrated dentin). Groups identified by different lower-case letters are
tubules were cut 'end-on' in deep statistically different (P < 0.001). [Note: In Photac-Fil Quick, dehydrated dentin group, 19 of the 20
dentin (Fig. 1A), and was either thin specimens failed during specimen preparation. The only specimen that remained intact gave a tensile
(1-2 ␮m) or absent when hydrated bond strength of 6.25 MPa or 6.25/20 = 0.3 ± 1.4 MPa. The null bond strengths from the other 19
superficial dentin with sparsely specimens that failed prematurely were included in the statistical analysis to avoid the bias of only
measuring "survivors".] (B) Representative scanning electron microscopic image of the dentin side of a
distributed, or obliquely oriented, fractured beam in hydrated dentin bonded with Photac-Fil Quick. The non-particulate nature of the
tubules was examined (Tay and absorption layer (AL) can be clearly observed, with the presence of artifactual dehydration cracks (open
Pashley, unpublished results). arrowhead) similar to those seen in Fig. 3D. C: fractured RMGIC. Similar features were observed in
Although our use of a dehydrated debonded specimens of Fuji II LC bonded to hydrated dentin (not shown). (C) Representative scanning
electron microscopic image of the dentin side of a fractured beam in dehydrated dentin bonded with
dentin model is far-removed from Photac-Fil Quick. The absorption layer is absent, and failure occurred between the RMGIC (C) and the
clinical reality, the correlation surface of the hybrid layer (H). Note opening of the dentinal tubules. Similar features were seen in Fuji
between the absence of the II LC bonded to dehydrated dentin (not shown).

Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.

International and American Associations for Dental Research

648 Tay et al. J Dent Res 83(8) 2004

absorption layer and poor bonding in laboratory-dehydrated This study was supported by RCG CERG grant
dentin explains why dentin surfaces should be kept moist to 10204604/07840/08004/324/01 from the Faculty of Dentistry,
promote the bonding of RMGICs (Wilder et al., 1998). University of Hong Kong, and by grants DE 014911 and DE
The existence of capillary pore spaces (i.e., water trees) 015306 from the National Institute of Dental and Craniofacial
within the absorption layers probably provided the channels for Research (PI, David Pashley). The authors are grateful to
continuous water flux from dentin across the RMGIC during its Michelle Barnes and Zinna Pang for secretarial support.
maturation (Sidhu et al., 1997; Watson et al., 1998). Despite
the differences in setting mechanisms between RMGICs and REFERENCES
Portland cements, both are equally susceptible to moisture gain
Abdalla AI (2000). Morphological interface between hybrid ionomers
and loss, and water movement can occur within the sealed
and dentin with and without smear-layer removal. J Oral Rehabil
cements during setting (Bentz and Hansen, 2000). Portland
27:808-814.
cements with low water/cement ratios exhibit the
Ausiello P, Apicella A, Davidson CL (2002). Effect of adhesive layer
characteristics of 'self-desiccation' and 'autogenous shrinkage',
properties on stress distribution in composite restorations—a 3D
and water is critical for their hydration via capillary pore spaces
finite element analysis. Dent Mater 18:295-303.
within the cement (Bentz et al., 2001).
Bentz DP, Hansen KK (2000). Preliminary observations of water
Two distinct glass-ionomer reactions were reported in
movement in cement pastes during curing using x-ray absorption.
RMGICs (Young, 2002), with the first corresponding to the
Cem Concr Res 30:1157-1168.
consumption of intrinsic water in the RMGICs, and the
Bentz DP, Geiker MR, Hansen KK (2001). Shrinkage-reducing
second attributing to subsequent extrinsic water sorption.
admixtures and early age desiccation in cement pastes and
Compared with conventional glass-ionomer cements, the
mortars. Cem Concr Res 31:1075-1085.
water content in RMGICs is reduced. With
Chen YC, Chirila TV, Russo AV (1993). Hydrophilic sponges based
photopolymerization, the acid-base reaction is also reduced
on 2-hydroxyethyl methacrylate, II: effect of monomer mixture
during the early setting stage of RMGICs (Kakaboura et al.,
composition on the equilibrium water content and swelling
1996). Water that resided within the capillary pore spaces
behaviour. Mater Forum 17:57-65.
may eventually be used up in subsequent acid-base reaction.
Chirila TV, Chen YC, Griffin BJ, Constable IJ (1993). Hydrophilic
We speculate that, under sealed conditions, RMGICs may
sponges based on 2 hydroxyethyl methacrylate, I: effect of
self-desiccate if external water is not available to replenish
monomer mixture composition on the pore size. Polym
the consumed water, by drying out the fine capillaries' spaces
International 32:221-232.
within the maturing cement. The menisci that develop along
Chuang SF, Jin YT, Tsai PF, Wong TY (2001). Effect of various
the interface between the water-filled and air-filled capillaries
surface protections on the margin microleakage of resin-modified
can create capillary pressures that generate high tensile
glass ionomer cements. J Prosthet Dent 86:309-314.
stresses and strains, leading to autogenous shrinkage of the
Feilzer AJ, Kakaboura AI, de Gee AJ, Davidson CL (1995). The
RMGIC (i.e., shrinkage apart from those generated by free
influence of water sorption on the development of setting
radical polymerization or dehydration shrinkage resulting
shrinkage stress in traditional and resin-modified glass ionomer
from environmental water loss). In the absence of a stress-
cements. Dent Mater 11:186-190.
relieving mechanism, these strains may concentrate along the
Hamid A, Okamoto A, Iwaku M, Hume WR (1998). Component
junction between the RMGIC and dentin, due to a stiffness-
release from light-activated glass ionomer and compomer
toughness mismatch between the restorative material and
cements. J Oral Rehabil 25:94-99.
dentin. The observed spontaneous fractures resulted in the
Hatton PV, Brook IM (1992). Characterisation of the ultrastructure of
recorded null bond strengths in both dehydrated groups,
glass-ionomer (poly-alkenoate) cement. Br Dent J 173:275-277.
despite the presence of remnant glass-ionomer inclusions on
Kakaboura A, Eliades G, Palaghias G (1996). An FTIR study on the
the surface of the hybrid layer and within the dentinal tubules
setting mechanism of resin-modified glass ionomer restoratives.
(Fig. 4C). Conversely, autogenous shrinkage may be
Dent Mater 12:173-178.
minimized when an absorption layer functions as a stress-
Nakanuma K, Hayakawa T, Tomita T, Yamazaki M (1998). Effect of
relieving, semi-permeable membrane that permits water to
the application of dentin primers and a dentin bonding agent on
enter the RMGIC matrix, under an osmotic gradient created
the adhesion between the resin-modified glass-ionomer cement
by the unpolymerized HEMA (Hamid et al., 1998) and
and dentin. Dent Mater 14:281-286.
hydrophilic salts within the matrix. Theoretically, continuous
Neubauer CM, Jennings HM (1996). The role of the environmental
water flux should occur until there is a balance of osmotic
scanning electron microscope in the investigation of cement-based
pressure. This hypothesis should be further examined with the
materials. Scanning 18:515-521.
complementary use of confocal microscopy and
Palmer G, Anstice HM, Pearson GJ (1999). The effect of curing
environmental scanning electron microscopy.
regime on the release of hydroxyethyl methacrylate (HEMA) from
resin-modified glass-ionomer cements. J Dent 27:303-311.
ACKNOWLEDGMENTS Pereira PN, Yamada T, Tei R, Tagami J (1997). Bond strength and
We thank Amy Wong of the Electron Microscopy Unit, the interface micromorphology of an improved resin-modified glass
University of Hong Kong, for technical assistance, and ionomer cement. Am J Dent 10:128-132.
Michael Chiang of the Department of Biology and Chemistry, Sidhu SK, Watson TF (1998). Interfacial characteristics of resin-
City University of Hong Kong, for technical assistance with modified glass-ionomer materials: a study on fluid permeability
the field emission-environmental scanning electron using confocal fluorescence microscopy. J Dent Res 77:1749-1759.
microscope. The resin-modified glass-ionomer cements Sidhu SK, Sherriff M, Watson TF (1997). The effects of maturity and
examined were generous gifts from 3M-ESPE and GC Corp. dehydration shrinkage on resin-modified glass-ionomer
Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.

International and American Associations for Dental Research

J Dent Res 83(8) 2004
restorations. J Dent Res 76:1495-1501.
Sidhu SK, Sherriff M, Watson TF (1999). Failure of resin-modified
glass-ionomers subjected to shear loading. J Dent 27:373-381.
Sidhu SK, Pilecki P, Cheng PC, Watson TF (2002). The morphology
and stability of resin-modified glass-ionomer adhesive at the
dentin/resin-based composite interface. Am J Dent 15:129-136.
Stokes DJ, Rea SM, Porter AE, Best SM, Bonfield W (2002).
Characterization of biomedical materials, cells & interfaces using
environmental SEM (ESEM). Mater Res Soc Symp Proc 711:113-
118.
Tanumiharja M, Burrow MF, Tyas MJ (2000). Microtensile bond
strengths of glass ionomer (polyalkenoate) cements to dentine
using four conditioners. J Dent 28:363-366.
Tay FR, Pashley DH (2003). Water treeing—a potential mechanism
for degradation of dentin adhesives. Am J Dent 16:6-12.
Tay FR, Pashley EL, Huang C, Hashimoto M, Sano H, Smales RJ, et
al. (2001a). The glass-ionomer phase in resin-based restorative
materials. J Dent Res 80:1808-1812.
Tay FR, Smales RJ, Ngo H, Wei SH, Pashley DH (2001b). Effect of
different conditioning protocols on adhesion of a GIC to dentin. J
Downloaded from jdr.sagepub.com at UNIV NEBRASKA LIBRARIES on April 10, 2015 For personal use only. No other uses without permission.
International and American Associations for Dental Research
RMGIC-Dentin Interaction 649
Adhes Dent 3:153-167.
Watson TF, Sidhu SK, Griffiths BM (1994). Ionomers vs. composites
at the tooth interface. In: Proceedings of the Second International
Symposium on Glass Ionomer Cements. Hunt P, editor.
Philadelphia: International Symposia in Dentistry, pp. 123-130.
Watson TF, Pagliari D, Sidhu SK, Naasan MA (1998). Confocal
microscopic observation of structural changes in glass-ionomer
cements and tooth interfaces. Biomaterials 19:581-588.
Watts DC, Kisumbi BK, Toworfe GK (2000). Dimensional changes of
resin/ionomer restoratives in aqueous and neutral media. Dent
Mater 16:89-96.
Wilder AD, Boghosian AA, Bayne SC, Heymann HO, Sturdevant JR,
Roberson TM (1998). Effect of powder/liquid ratio on the clinical
and laboratory performance of resin-modified glass-ionomers. J
Dent 26:369-377.
Yap A, Lee CM (1997). Water sorption and solubility of resin-
modified polyalkenoate cements. J Oral Rehabil 24:310-314.
Young AM (2002). FTIR investigation of polymerisation and polyacid
neutralisation kinetics in resin-modified glass-ionomer dental
cements. Biomaterials 23:3289-3295.