d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) e301e305

Available online at www.sciencedirect.com

ScienceDirect
journal homepage: www.intl.elsevierhealth.com/journals/dema

Adhesion of resin-modied glass-ionomer

cements may affect the integrity of tooth structure
in the open sandwich technique
Beata Czarnecka a , Anna Kruszelnicki a , Anthony Kao a ,
Marta Strykowska a , John W. Nicholson b,
a

Department of Biomaterials and Experimental Dentistry, University of Medical Sciences, ul Bukowska 70,
Poland
60-812 Poznan,
b School of Sport, Health & Applied Science, St Marys University College, Twickenham, Middlesex TW1 4SX,
United Kingdom

a r t i c l e

i n f o

a b s t r a c t

Article history:

Objective. To study the interfaces between model cavities prepared in teeth and four glass

Received 23 May 2013

ionomer cements (two conventional and two resin-modied).

Received in revised form

Methods. Ten non-cavitated molars and premolars were used and, in each, two 3 mm deep

9 October 2013

slot preparations were created on opposing sides of the tooth. The teeth were conditioned as

Accepted 21 May 2014

appropriate, then restored using the open sandwich technique, using a conventional glass
ionomer (Fuji IX, Ketac Molar) or resin modied glass ionomer (Fuji II LC or N100), followed
by completion with composite resin. The teeth were then embedded in a transparent acrylic

Keywords:

resin and cut parallel to the long axis through both restorations, using a low speed diamond

Glass-ionomer

wheel saw. Samples were evaluated using a metallographic light microscope (100). Three

Bonding

areas were assessed: the axial wall, the axial gingival line angle and the cavo-surface line

Adhesion

angle. Bonding was categorized as inadequate or adequate based on the appearance and

Tooth fracture

inadequate bonding was further studied and classied. Data were analysed statistically
using the McNamara analysis.
Results. The majority of materials failed to make adequate contact with the axial wall, and
there were also aws at the axial/gingival line angle in several samples. By contrast, the cavosurface line angle was generally soundly lled and the materials showed intimate contact
with the tooth surface in this region. The most serious inadequacy, though, was not lack of
intimate contact and/or adhesive bond, but the presence of perpendicular cracks in 30% of
the Fuji II LC samples which extended into the underlying dentin.
Signicance. The problems of placement and dentin cracking experienced with these materials demonstrate that adhesive bond strength alone cannot be used as the criterion of success
for restorative materials. In fact good adhesion can, in certain cases, promote cracking of
the dentin due to stresses within the material, an outcome which is undesirable.
2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

Corresponding author.
E-mail address: john.nicholson@smuc.ac.uk (J.W. Nicholson).

http://dx.doi.org/10.1016/j.dental.2014.05.008
0109-5641/ 2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

e302

1.

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) e301e305

Introduction

Table 1 Glass ionomer cements used.

Brand

Glass ionomer (GI) cements were rst reported by Wilson and

Kent in 1972 [1] and have since become widely used in clinical
dentistry [2]. They have many desirable properties, in particular the ability to form satisfactory adhesive bonds with enamel
and dentin [3], and to release of uoride in a sustained way
over a prolonged periods of time [4]. As far as bonding is concerned, they have the additional advantages that no extra
preparation of the tooth structure is required for consistent
adhesion [5], and that the bond becomes more durable with
time, due to the formation of an ion-exchange layer between
the tooth and the cement [6].
Despite the favorable qualities of conventional glass
ionomers, they have some drawbacks. In particular, the slow
setting reaction causes the material to be hydrolytically unstable in the early stages of hardening [6], making it sensitive
to early water uptake or loss, depending on circumstances.
This may affect the mechanical properties of the material and
certainly affects the appearance, which may become chalky
due to the formation of micro-cracks in the surface [2]. To
overcome these problems, the resin-modied glass ionomers
(RMGICs) have been developed [7]. These combine the components of conventional glass ionomers with the monomer
2-hydroxyethyl methacrylate (HEMA), together with appropriate polymerization initiators, so that the material sets by a
combination of acidbase reaction and addition polymerization. The majority of brands of resin-modied glass ionomers
are light-cured, which gives them the advantage that the clinician has considerable control over the setting reaction, much
of which cannot occur until the dental curing lamp is switched
on.
Resin-modied glass ionomers retain many of the advantages of conventional glass ionomers, in particular their
adhesion to the tooth surface [7] and their ability to release
uoride [8]. However, the monomer HEMA is potentially damaging and its presence reduces the biocompatibility of these
materials relative to conventional glass ionomers [9]. Unpolymerized HEMA is known to be released by these materials [10]
and HEMA has been shown to be able to penetrate the dentin to
reach the pulp [11], where it proves to be cytotoxic [12]. HEMA
monomer is also a well-known allergen [9].
Conventional and resin modied glass ionomer cements
have been employed clinically in the so-called open sandwich
technique for restoring Class II cavities [1316]. In this technique, glass ionomer cement, or RMGIC, is used for the gingival
portion of the restoration, and composite resin is used to complete the repair [15,16]. Clinical results with this technique are
good and the use of a glass ionomer (of either type) has been
found to signicantly reduce the marginal micro-leakage at
the cemento-enamel junction compared with the use of other
materials, such as composite resin [13,15].
Glass-ionomers of both types have been tested for their
adhesion to the tooth surface [17,18]. Materials have been
tested in shear bond mode [1720] and, more recently, in
microtensile mode [2123]. At relatively short times, bond
strengths have been found to be acceptable, even without
surface pre-treatment [2]. This leads to the suggestion that
forming such adhesive bonds is straightforward, and the

Manufacturer

Fuji IX
Fuji II LC
Ketac Molar
N100

GC (Japan)
GC (Japan)
3M ESPE (USA & Germany)
3M ESPE (USA & Germany)

Type
Conventional
RMGIC
Conventional
RMGIC (nano-ionomer)

expectation that the resulting interfaces should be sound and

free from aws. This hypothesis was the original one under
test in the current study.
Since the open sandwich technique was pioneered, new
versions of glass ionomers, both conventional and resinmodied, have been launched by manufacturers. The current
study was initially undertaken with a view to examining the
performance in vitro of some of these newer materials in the
open sandwich technique, with particular emphasis on their
bonding and the interfacial region with the tooth. However,
our work has led to some new observations, not previously
reported, that raise important questions about the issue of
adhesive bonding by glass-ionomers.

2.

Materials and methods

Ten non-cavitated molars and premolars, extracted for

orthodontic reasons, were used in this study. Teeth were
stored prior to use in physiological saline to which small
amounts of thymol had been added for 120 days at a temperature of 8 C. Materials used in the study are listed in Table 1,
and consisted of two conventional and two resin-modied
glass ionomers. One of the materials, N100, is described as
nanoionomer because it contains nanometer sized ller. It
is prepared and dispensed using a double tubed dispenser. Fuji
II LC is a capsulated material, and was mixed using a vibratory
mixer for 10 s prior to extrusion from the capsule.
In each of the ten teeth, two slot preparations were created
on opposing sides of the tooth (see Fig. 1), using a ssure bur
(ISO 806 314 107 524 012, Lot D-4362 at 300 000 RPM), diamond

Fig. 1 Slot preparation on opposing sides of the tooth.

A = axial wall, B = axial/gingival line angle, C = cavo-surface
line angle.

e303

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) e301e305

round bur (ISO 806 314 001 524 016, Lot D-5113 at 280 000 RPM)
and a high speed rose bur (ISO 500 314 001 001 016, Lot D434040 at 280 000 RPM). The depth of each preparation was
3 mm, (the gingival wall of the box is designed so it extends
1 mm below the cervical line of the tooth, into the root). Following cavity preparation, the teeth were stored in saline solution
at a temperature of 8 C for 7 days.
The ten teeth were divided, at random, into two groups:
Group A (prepared using the two 3M ESPE products) and Group
B (the GC products). In each group, one slot in a particular tooth was lled with conventional glass ionomer cement
(Ketac Molar or Fuji IX) and the other slot in the same tooth
was lled with resin modied glass ionomer (N100 or Fuji II
LC). Materials used are listed in Table 1.
Having prepared the tooth, the material was placed in a
thin layer along the axial wall and gingival oor of the cavity
preparation, after which the lling of the cavity was completed
by placement of composite material. In this way, four different
sample groups were compared for their ability to bind to tooth
structure.
The cavities designated for conventional glass ionomer
samples were conditioned for 10 s prior to restoration with 20%
polyacrylic acid, followed by a 10 s water rinse and drying with
cotton pellets. The glass ionomer cements were mixed according to the manufacturers instructions, in a 1:1 powder to liquid
ratio and condensed into the cavity to a depth of 1 mm on both
axial and gingival walls. The cement was allowed to set for
5 min after which the cavity was etched, primed, bonded then
restored with composite resin (Filtek Z250, ex 3M; shade A2).
The resin modied glass ionomer slot for Group A (N100)
was primed with Ketac N100 primer for 15 s, air thinned for
10 s and light cured for 10 s. The resin modied glass ionomer
was dispensed using the Ketac Clicker dispenser and mixed for
20 s, followed by placement into the cavity (using the dispenser
provided by 3M), along axial and pulpal walls to a thickness
of 1 mm, followed by light curing. For group B, the dentin
was conditioned with 20% polyacrylic acid for 10 s, rinsed
with water for 10 s and cotton dried. The resin-modied glass
ionomer (Fuji II LC) was mixed on a glass slab and placed using
a plugger, followed by light curing for 20 s. Lastly, the cavity
was etched, prime and bonded and restored with composite
resin as for Group A.
Following restoration, each group was stored in saline solution in a tightly sealed container for 1 week, at a temperature
of 37 C. The teeth were then embedded in a transparent
acrylic resin (Duracryl) and were cut parallel to the long axis
of the tooth, through both restorations, using a low speed diamond wheel saw (model 650). Each tooth yielded two slices,
which were polished prior to microscopic evaluation. Samples
were then evaluated using a metallographic light microscope
(Nikon Eclipse LV 100) under a magnication of 100. In each
tooth, there were three areas of observation: the axial wall (A),
the axial gingival line angle (B) and the cavo-surface line angle
(C) (see Fig. 1).
In the initial assessment, bonding was categorized as inadequate or adequate based on the appearance of the region
between the dentin and glass ionomer cement. Inadequate
bonding was further studied, and details of the inadequacies
recorded. Data were further analysed statistically using the
McNamara analysis.

Table 2 A comparison of adequate (A) and inadequate

(I) bonding sites of the four materials tested.
Material

Adequate
(out of 30)

Inadequate
(out of 30)

Fuji IX GIC
Fuji II RMGIC
Ketac Molar GIC
N100 RMGIC

18/30 = 60%
21/30 = 70%
14/30 = 47%
15/30 = 50%

12/30 = 40%
9/30 = 30%
16/30 = 53%
15/30 = 50%

3.

Results

Each restored slice was observed at three points, giving a total

of thirty observations for each material used. Bonding was
characterized initially as either adequate or inadequate and
the results show that Fuji II LC had the highest proportion
of adequate bonded specimens, whereas Ketac Molar had the
least (see Table 2).
Further evaluation of the inadequately bonded samples
is given in Table 3 and several distinct types of inadequate
bonding were identied. These were: complete detachment of
cement from dentin, perpendicular crack through the glass
ionomer only, detachment plus perpendicular crack into the
glass ionomer and nally crack through the glass ionomer into
the dentin.
In evaluating the inadequate bonding associated with Fuji
IX, it was found that over half (58%) of the inadequate samples showed detachment from the dentin. With Fuji II LC,
the inadequacies tended to consist of cracks into dentin.
These comprised 56% of inadequate samples. Additionally,
33% of inadequate samples showed transverse cracks within
the material. In these specimens, no detachment failures were
observed. For Ketac Molar, the most common inadequacies
observed were detachment plus transverse cracking of the
cement (63% of inadequate samples). In no sample did the
crack penetrate into the dentin, meaning that the cracks did
not exceed a maximum of 2 mm in length, and were typically
shorter than this. Finally, with N100, most of the inadequacies were observed as detachments of the cement from the
dentin (53%). Again, no cracks penetrated the dentin with this
material.
In terms of the geometry of the restoration, most of the
inadequate bonding occurred on the axial wall (Table 4),
whereas at the axial/gingival line angle and the cavo-surface
line angle, bonding appeared more satisfactory and was generally classied as adequate.

4.

Discussion

The fact that glass ionomer cements generally bond well to

both dentin and enamel has been recognized for many years
[17]. Indeed, this adhesion was considered so important that
when resin-modied versions of these materials were rst
launched, almost immediately a paper appeared conrming
their adhesion to dentin [18]. In these early papers, bonding was typically determined by measuring bond strength of
cement samples loaded in shear [1719]. It still is widely used
as a test method in reporting the bonding properties of modern
cement formulations [20].

e304

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) e301e305

Table 3 Inadequate bonding of material to tooth structure according to location and type of inadequacy.
Material

Detachment
(interfacial failure)

Fuji IX
Fuji II LC
Ketac Molar
N100

7/12 = 58%
0
6/16 = 38%
8/15 = 53%

Transverse crack
within glass ionomer

Detach + T-crack

Crack into dentin

3/12 = 25%
3/9 = 33%
0
4/15 = 27%

2/12 = 17%
1/9 = 11%
10/16 = 63%
3/15 = 20%

0
5/9 = 56%
0
0

Table 4 Number of adequate (A) and inadequate (I) bonding sites according to location within the cavity preparation.
Location

Axial wall
Axial/gingival line-angle
Cavo-surface line-angle

Fuji IX

Fuji II LC

Ketac Molar

N100

4
5
9

6
5
1

4
7
10

6
3
0

1
4
9

9
6
1

3
6
6

7
4
4

More recently, studies have been reported in which the test

regime has been changed to the determination of microtensile bond strength [2123]. The principal advantage of this
mode of testing is that the surface area of attachment is much
smaller than in conventional shear bond evaluation. Results
are therefore less likely to be affected by the presence of aws
and irregularities at the cementdentin interface. Using microsized bonding areas reduces the likelihood of such weakening
occurring.
However, none of these testing regimes examines a realistic bonding situation. Real glass ionomer restorations ll
prepared cavities in teeth. The material therefore has to be
able to be forced fully into the cavity, occupying the whole
space and coming into intimate contact with at surfaces in
three dimensions. Shear or microtensile bond tests only evaluate the latter aspect: the ability to come into contact with
relatively at surfaces and there to form a strong adhesive
bond. In the present study we have considered a model that
more accurately reects the conditions in which bonds must
form in clinical application. When we have done that, we nd
that problems of inadequate bonding are apparent mainly at
the axial wall. These must occur because of difculties in
fully condensing the cement, meaning that it is not pushed
properly to the back of the cavity. This problem has not been
reported previously, and appears to arise from problems in
technique of placement of a relatively thin layer (1 mm) of
the glass ionomer cement, even less than would be placed
as a base. There is also a difculty caused by adhesion of the
unset cement to the instruments being used. By contrast, published accounts of bonding of the material Fuji IX have shown
that, in the hands of experienced clinicians, it can be placed
in intimate contact with the complete surface of a prepared
cavity in a carious tooth [6]. However, most of these studies
employed glass ionomer cement as a full restoration, not as a
base. In the present study, the thinness of the base was necessary the cavities were relatively small and space was necessary
for placement of resin composite. This situation is common in
class V cavities (for example, abrasion and erosion) at the neck
of the tooth. In our clinical experience, it is easier to place the
cement in one bulk portion, lling the whole cavity, than to
place a thin layer of glass ionomer at the base of the cavity
and leaving the remainder of the cavity material free [24].
By contrast with the problems at the axial wall, bonding at
the cavo-surface angle was found to be good. This shows that

all of the materials examined can occupy this type of conned

space, either by owing in on application, or as the result of
condensing into the corners during placement.
The emphasis on bond strength in the published literature has overshadowed consideration of other aspects of the
performance of these materials. The nding that Fuji II LC is
capable of causing cracks to develop in the dentin in a proportion of the samples (30%) is extremely important. In no
instance was this material found to detach from the dentin
surface in the course of these experiments. Its adhesion is
excellent. Indeed, this is partly the cause of the problems
observed. When stresses develop, either due to polymerization contraction or swelling due to water sorption, both of
which are known to occur with this material [25], rather than
the adhesive bond failing, or cracks developing in the material,
these stresses can cause cracks in dentin. This is of considerable clinical importance, and requires further study. To date,
clinical results with Fuji II LC have been good [26,27], and it
may be that our experimental study exaggerates the problem of dentinal cracking, which may not occur so frequently
under clinical conditions. Nonetheless, the fact that excellent
adhesion can lead to this type of serious failure is important
and should be noted by the research community. Our ndings show that these materials cannot be evaluated safely on
the basis of adhesive bond strength alone. Other factors (polymerization contraction, swelling in moisture, mechanisms of
stress transfer) must be taken into consideration, and it is
essential that materials do not have such good bond strengths
that their dimensional stresses cause tooth fracture in vivo.

5.

Conclusions

This study has shown that the use of glass ionomer materials,
both conventional and resin-modied, for the repair of model
cavities using sandwich technique is very satisfactory in the
majority of cases. However it has been shown to be capable
of leading to problems of adhesion. In a number of instances,
there were problems in establishing contact between the
material and the furthest axial wall of the cavity, which may
arise from problems in handling the placement of a thin layer
of unset cement. In addition, for one of the resin-modied
glass ionomers (Fuji II LC), in a small but important proportion
of the samples (16%), the material caused a perpendicular

d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) e301e305

crack to develop in the adjacent dentin. This has serious

implications for clinical application and demonstrates that
adhesive bond strength alone cannot be used as the criterion
of success for an adhesive restorative material. Other factors
such as ease of condensation and dimensional stability are
important and need to be considered as well.

references

[1] Wilson AD, Kent BE. A new translucent cement for dentistry:
the glass ionomer cement. Br Dent J 1972;15:1335.
[2] Mount GJ. Color atlas of glass ionomer cements. London:
Dunitz; 2002.
[3] Akinmade AO, Nicholson JW. Glass-ionomer cements as
adhesives. Part I. Fundamental aspects and their clinical
relevance. J Mater Sci Mater Med 1993;4:93101.
[4] Forsten L. Short and long term uoride release from glass
ionomers and other materials. Scand J Dent Res
1991;99:2415.
[5] Mount GJ. Glass-ionomer cements; past, present and future.
Oper Dent 1994;19:8290.
[6] Ngo H, Mount GJ, Peters MCRB. A study of glass-ionomer
cement and its interface with enamel and dentin using a
low-temperature, high-resolution scanning electron
microscopic technique. Quintessence Int 1997;28:639.
[7] Mitra SB. Adhesion to dentine and physical properties of a
light-cured glass-ionomer liner/base. J Dent Res
1991;70:704.
[8] Forss H. The release of uoride and other elements from
light cured glass ionomers in neutral and acidic conditions. J
Dent Res 1993;72:125762.
[9] Czarnecka B, Nicholson JW. The biocompatibility of
resin-modied glass-ionomer cements for dentistry. Dent
Mater 2008;24:17028.
[10] Palmer G, Anstice HM, Pearson GJ. The effect of curing
regime on the release of hydroxyethyl methacrylate (HEMA)
from resin-modied glass-ionomer cements. J Dent
1999;27:30311.
[11] Hamid A, Hume WR. Diffusion of resin monomers through
human carious dentin in vitro. Endod Dent Traumatol
1997;13:15.
[12] Kan KC, Messer LB, Messer HH. Variability in cytotoxicity
and uoride release of resin-modied glass-ionomer
cements. J Dent Res 1997;76:15027.
[13] Hembree JH. Microleakage at the gingival margin of class II
composite restorations with glass-ionomer liner. J Prosthet
Dent 1989;61:2830.

e305

[14] Prati C. Early marginal microleakage in class II resin

composite restorations. Dent Mater 1989;5:3928.
[15] Dietrich T, Losche AC, Losche GM, Roulet JF. Marginal
adaptation of direct composite and sandwich restorations in
Class II cavities with cervical margins in dentin. J Dent
1999;27:11928.
[16] Arora V, Kundabala M, Parolia A, Thomas MS, Pai V.
Comparison of the shear bond strength of RMGIC to a resin
composite using different adhesive systems: an in vitro
study. J Conserv Dent 2010;13:803.
[17] Powis DR, Folleras T, Merson SA, Wilson AD. Improved
adhesion of a glass ionomer cement to dentin and enamel. J
Dent Res 1982;61:141622.
[18] Mitra SB. Adhesion to dentin and physical properties of a
light-cured glass-ionomer liner/base. J Dent Res
1991;70:724.
[19] Lin A, McIntyre NS, Davidson RD. Studies on the adhesion of
glass-ionomer cements to dentin. J Dent Res
1992;71:183641.
[20] Carvalho T-S, van Ameronga W-E, de Gee A, Bonecker M,
Sampaio F-C. Shear bond strengths of three glass ionomer
cements to enamel and dentine. Med Oral Patol Oral Cir
Bucal 2011;161:e40610.
[21] Tanumiharja M, Burrow MF, Tyas MJ. Microtensile bond
strengths of glass ionomer (polyalkenoate) cements to
dentine using four conditioners. J Dent 2000;28:3616.
[22] Jordehi AY, Ghasemi A, Zadeh MM, Fekrazad R. Evaluation of
microtensile bond strength of glass ionomer cements to
dentin after conditioning with Er, Cr:YSGG laser. Photomed
Laser Surg 2007;25:4056.
[23] Fragkou S, Nikolaidis A, Tsiantou D, Achilias D, Kotsanos N.
Tensile bond characteristics between composite resin and
resin-modied glass-ionomer restoratives used in the
open-sandwich technique. Eur Arch Paediatr Dent
2013;14:23945.
[24] Czarnecka B, Limanowska-Shaw H, Nicholson JW.
Microscopic evaluation of the interface between
glass-ionomer cements and tooth structures prepared using
conventional instruments and the atraumatic restorative
treatment (ART) technique. Quintessence Int 2006;37:55764.
[25] Attin T, Buchalla W, Klelbassa AM, Helwig E. Curing
shrinkage and volumetric changes of resin-modied glass
ionomer restorative materials. Dent Mater 1995;11:35962.
[26] Daou MH, Tavernier B, Meyer JM. Two-year clinical
evaluation of three restorative materials in primary molars. J
Clin Pediatr Dent 2009;34:548.
[27] Perdigao J, Dutra-Correa M, Saraceni SHC, Ciaramicoli MT,
Kiyan VH. Randomized clinical trial of two resin-modied
glass-ionomer materials: 1 year results. Oper Dent
2012;37:591601.