Dimensional Changes of Glass Ionomers and a Giomer during the Setting Time.

OBJECTIVES: The aim of this study was to evaluate dimensional changes of conventional glass ionomer cements, resin-modified glass-ionomer cement, and a giomer during the setting time using digital laser interferometry. Additionally, the influence of different curing modes ("high", "soft", and "low") of a light-emitting diode (LED) curing unit on dimensional changes was evaluated.
MATERIALS AND METHODS: Linear curing shrinkage of conventional glass ionomer cements (CGICs): Fuji IX Extra (F9E), Fuji IX Fast (F9F), Ketac Molar Aplicap (KM), Ketac Molar Quick Aplicap (KMQ), resin-modified glass ionomer cement (RM GIC): Fuji II LC (F2LC) and giomer: Beautifil II (B2) was analyzed. All tested materials were of shade A3, while all of the GIC were encapsulated. Discoid specimens (n=10, d=10 mm, h=0.85 mm) were prepared for each tested material and each curing mode (for light-curable materials) according to the manufacturer's instructions. Light-curable specimens were cured with LED curing unit (Bluephase G2, Ivoclar-Vivadent, and Schaan, Liechtenstein). Dimensional changes during curing were recorded in real-time. The results were analyzed by ANOVA, and Tukey post hoc test was used for multiple comparisons (α˂ 1%).
RESULTS: All tested materials showed an initial setting expansion and a subsequent setting shrinkage. KM and KMQ had significantly lower setting shrinkage than RM GIC polymerized using any of the three curing modes. B2 showed lower shrinkage compared to F2LC.
CONCLUSIONS: The extent of curing shrinkage in RM GIC measured in this study can affect longevity of restorations.

Introduction

Modern restorative dentistry is still faced with the problem of marginal gap formation. Adhesive technology, i.e. direct chemical bonding and/or micromechanical adhesion to tooth structure, did not solve the issue of the marginal gap and consequent microleakage (, ). The phenomenon of microleakage may lead to serious problems such as postoperative sensitivity, penetration of bacteria, which causes secondary caries and inflammatory reaction of the pulp, tooth discoloration, and eventually loss of the restoration ().

Dimensional changes of materials during setting change the structure of the material and may influence the adhesion thereby greatly contributing to creating the marginal gap followed by microleakage (-). Similarity in dimensional changes occurring during the setting of restorative materials and thermal expansion of the tooth may play an important role in marginal gap minimization (). According to the literature, dimensional changes are an important factor for restoration longevity (-). The stress resulting from dimensional changes is dependent on the extent of curing shrinkage and elastic modulus (, ).

A total resin composite polymerization can be divided into pre- and post-gel phases. During the pre-gel phase, monomer molecules have the ability to move within the polymer network, thus compensating polymerization stress. However, in the post-gel phase, monomers become immobilized in the polymeric matrix of rising viscosity, generating a large amount of stress in the tooth structure and at the tooth-material interface (, ). Gradual polymerization of the material is recommended to reduce the shrinkage stress and consequent microleakage ().

Glass ionomer cements (GICs) are hydrophilic materials with widespread clinical use due to the chemical adhesion to tooth structure, fluoride release and biocompatibility (). Today, there are a large numbers of commercially available GICs for clinical use with compositional variations regarding powder, liquid or both components. Different compositions of GICs are optimized for specific clinical purposes. Two main groups of GICs can be distinguished, considering the setting mechanism: conventional GIC (CGICs) (setting through an acid-base reaction) and resin-modified GIC (RM GIC) (setting through an acid-base reaction and the simultaneous polymerization of the resinous part). The setting reaction of RM GIC is more complex because both reactions occur concurrently thus influencing each other ().

Giomers are resin-based materials with the basic composition similar to that of resin composites, which contain already reacted glass ionomer component named pre-reacted glass (PRG) particles. The PRG fillers are produced from the complete or partial reaction of ion-leachable glasses and polyalkenoic acids (, ).

Dimensional changes due to curing contraction can be described as a linear and volumetric contraction (). Most common methods for determining the volumetric curing contraction are based on measurements using water or mercury dilatometers (, , -). Watts and Cash used indirect method, the “deflecting disc” method, which ensures reproducible measurements of polymerization shrinkage. They calculated the volumetric contraction from the post-gel linear displacement of a deflecting disc resting on a brass ring (). Grajower and Guelmann determined the dimensional change of GIC by means of linear displacement transducer (). In a large number of previous studies, the total dimensional changes that occur during the setting were measured (, , ). In this study, for the first time, the digital laser interferometry method was used to determine dimensional changes of glass ionomer cements and giomer in real time during the setting period, as recommended by respective manufacturers. This method has previously been shown to be reliable for monitoring dynamic processes such as shrinkage of resin composites (-).

The aim of this study was to evaluate linear dimensional changes of CGICs a RM GIC, and a giomer during the setting time and to assess the influence of three different curing modes of a LED curing unit on dimensional changes of RM GIC and giomer.

Research hypotheses tested were:

Dimensional changes will be affected by the material type (giomer and RM GIC are expected to show a similar degree of shrinkage, while CGICs are expected to show a lower degree of shrinkage than RM GICs and giomer).

Dimensional changes of light-curable materials will be influenced by curing modes.

Materials and methods

Materials selection

This study evaluated four conventional GIC (CGICs): Fuji IX Extra (F9E), Fuji IX Fast (F9F), Ketac Molar Aplicap (KM), and Ketac Molar Quick (KMQ), a light-curable resin-modified GIC (RM GIC): Fuji II LC (F2LC) and a giomer: Beautifil II (B2). All tested materials, listed in Table 1, were of shade A3, while all of the GICs were encapsulated.

Table 1

Materials investigated in the study.

Material	Code	Type of material	Manufacturer	LOT	Aplication
Fuji IX EXTRA	F9E	CGIC	GC Corporation (Tokyo, Japan)	1009205	Restorative
Fuji IX FAST	F9F	CGIC	GC Corporation (Tokyo, Japan)	1005251	Restorative
Ketac Molar Aplicap	KM	CGIC	3M ESPE (Seefeld, Germany)	425688	Restorative
Ketac Molar Quick Aplicap	KMQ	CGIC	3M ESPE (Seefeld, Germany)	416807	Restorative
Fuji II LC a	F2LC	RM GIC	GC Corporation (Tokyo, Japan)	1009277	Restorative
Beautifil II a	B2	Giomer	Shofu Inc. (Kyoto, Japan)	041008	Restorative

a Light-curable materials.

Specimen preparation

10 discoid specimens (d=10 mm, h=0.85 mm) were prepared for each tested material (and curing mode, for light-curable materials), 10 discoid specimens (d=10 mm, h=0.85 mm) were prepared. The encapsulated materials (F9E, F9F, KM, KMQ and F2LC) were mixed according to the manufacturer's instructions. An adequate amount of material was extruded onto a polyethylene terephthalate (PET) film in a stainless steel ring mold of 1 mm height, covered with another sheet of PET film and compressed using a flat stainless steel plate. The total thickness of the specimen (0.85 mm) was the result of subtracting the thickness of two PET sheets from the thickness of the ring mold (1 mm). The non-encapsulated material (B2) was applied onto the PET film using a spatula. Specimen preparation was performed in the dark room with red light to avoid curing effect of ambient light on light-curable materials. Light-curable materials were cured using LED curing unit (Bluephase G2, Ivoclar-Vivadent, Schaan, Liechtenstein) for 20 s with one of the following light-curing modes: “high” (1100 mW/cm2), “soft” (650-1100 mW/cm2) and “low” (650 mW/cm2).

Digital laser interferometry

Dimensional changes were calculated from experimental measurements obtained by using digital laser interferometry device schematically shown in Figure 1. The discoid specimens were covered on both sides with PET foils and sandwiched between two glass plates. The upper glass plate had a reflective surface and acted as a mirror. The specimen carrier was designed to allow photo polymerization of the specimen from one side and simultaneous measurements of thickness variation of the specimen from the other side. A micrometer was mounted to enable manual compensation of the displacement of the upper surface of the specimen, which was covered by a mirror. The micrometer was used at the end of each measurement to verify the final value for linear shrinkage of the specimen.

Figure 1

Illustration of the interferometric experimental setup. M, mirror; L, lens; R, reference mirror; O, object mirror; CL, curing light.

The beam emerging from the He-Ne laser (Spectra Physics, power = 25 mW, wavelength = 632.8 nm) was first expanded and collimated and then steered onto the specimen carrier plate. The plate reflected incident light by means of two flat mirrors, one placed on the upper surface of the specimen and the other outside of the specimen, thus forming two beams, the object beam and the reference beam. Subsequently, both beams were directed at a small angle to the CCD sensor with 752×582 pixels allowing the recording rate of 25 frames per second. The resulting interference fringe patterns carried the specimen phase information. For the time-dependent dimensional changes of the specimen, the location of fringes was also time-dependent and the required values were evaluated numerically. For GICs, all of the measurements started approximately 45 s after the end of mixing until the end of the setting time specified by respective manufacturers (for F9F, F9E, KM, KMQ up to 6 min and for F2LC up to 5 min). In the case of giomer (B2), measurements started concurrently with the initiation of light curing and lasted for 5 min. The interferograms recorded during the aforementioned time were analyzed using a custom-made computer program prepared in LabVIEW 2011 (National Instruments, Austin, Texas, USA) according to the previously described procedure ().

Statistical Analysis

The linear dimensional changes obtained using digital laser interferometry were analyzed by analysis of variance (ANOVA), and Tukey post hoc test for multiple comparisons (α = 0.01).

Results

All of the tested CGICs showed an initial setting expansion, followed by shrinkage (Figure 2). KM showed significantly lower shrinkage than F9E and F9F (p˂0.01), while KMQ showed significantly lower shrinkage than F9F (p˂0.01).

Figure 2

Dimensional changes of different CGICs during the setting and dimensional changes of RM GIC/giomer polymerized with different curing modes of Bluephase G2 LED curing unit.

The largest curing shrinkage had F9F followed by F9E, KMQ and KM (Table 2).

Table 2

Mean values and standard deviations of linear setting shrinkage of CGICs. Same lowercase letters denote statistically homogeneous groups within a column.

	shrinkage/%
material	mean	st dev.
F9E	1.22a,c	0.38
F9F	1.32 a	0.57
KM	0.47 b	0.13
KMQ	0.62 b,c	0.33

RM GIC and giomer showed an initial setting expansion and a subsequent setting shrinkage in all three curing modes (Figure 2). The differences among curing modes were statistically significant only in B2 between the “high” and “low” curing mode. In these cases, the linear dimensional changes were significantly higher for the “high” curing mode (B2 1.15±0.23%) compared to the “low” mode (B2 0.80±0.17%), (p˂ 0.01), (Table 3).

Table 3

Mean values and standard deviations of linear setting shrinkage of B2 and F2LC illuminated using different curing modes. Same lowercase letters within a row represent statistically homogeneous groups for the comparison among curing modes. Same uppercase letters denote statistically homogeneous groups within a column.

Material	shrinkage/ %
	high		soft		low
	mean	st.dev.	mean	st.dev.	mean	st.dev.
B2	1.15 a, A	0.23	0.99 a,b,A	0.23	0.80 b,A	0.17
F2LC	1.59a, A	0.40	1.65 a,A	0.34	1.62 a,B	0.35

B2 showed lower shrinkage than F2LC, but significantly lower shrinkage B2 compared to F2LC was measured for “low” curing mode (p˂0.01) (Table 3). If we compare dimensional changes of CGICs and RM GIC, the significance was only demonstrated by comparing KM and KMQ with F2LC: KM and KMQ had a significantly lower setting shrinkage than F2LC polymerized with any of the three curing modes (Figure 2).

Expansion of CGICs lasted longer than the expansion of RM GIC. The KM expanded for 4.0±0.7 min, KMQ expanded 3.4±0.8 min, F9F expanded through 2.0±0.4 min and F9E 0.8±0.3 min. F2LC expanded through the initial 0.1±0.06 min and B2 expanded during 0.1±0.03 min of light curing.

Discussion

A large range of contemporary GICs and their diverse and growing spectrum of clinical indications, considered in light of unavoidable setting shrinkage with all possible consequences, point to the importance of expanding the knowledge about their dimensional changes. The CGICs can be considered as low-shrinkage materials (, ). However, some researchers recognized the GICs as materials capable of reaching significant shrinkage values, comparable even to those of resin composite materials ().

In the literature, various methods of measuring dimensional changes of GIC have been described. Despite some specific differences that are pertinent to each method, the evaluation of the total dimensional changes that occur during the setting can be considered common to most of them (, , -).

The measurements in the digital laser interferometry device started immediately the specimen preparation and lasted throughout the setting time. Specimen preparation lasted approximately 45 s after the capsule mixing had been completed. In the initial phase of the setting, expansion was identified in all materials (Figure 2). It is thus possible to assume that materials also expanded slightly during the specimen preparation, which lasted for 45 s before the interferometric measurements started. KM expanded during 4.0±0.7 min, while KMQ expanded for 3.4±0.8 min; however, both materials showed lower shrinkage values compared to those of the other GICs. The remaining two CGICs (F9F and F9E) expanded through 2.0±0.4 min and 0.8±0.3 min, respectively. The RM GIC and giomer (F2LC and B2) also exhibited a slight expansion during illumination. From the curves in Figure 2, one can observe that CGICs expanded between first 0.8±0.3 min to 4.0±0.7 min, while RM GIC and giomers expanded through the initial 10 seconds of light curing. It can be assumed that the initial expansion in materials containing the resinous component is limited by monomer polymerization and crosslinking.

It would be beneficial if the expansion, occurring early in the curing period, would last longer because it may compensate for the negative effects of shrinkage and reduce the shrinkage time (). As mentioned earlier, the CGICs had a higher expansion which lasted longer compared to that of RM GIC. The initial expansion of F2LC and B2 was lower compared to the following shrinkage (Figure 2). Despite these findings, it was not possible to precisely quantify the expansion because of high variation in the expansion of specimens in the early stage of measuring. This early expansion was also noticed in the Bryant and Mahler study (), but was not assessed because of the material instability in early stage of the setting process ().

A number of previous studies have confirmed that dimensional changes of materials during setting depended upon both the chemical composition and experimental conditions (). Because of the resin-free structure of the original GICs, they can be considered as slow-setting materials that produce less stress at the tooth-restoration interface (). However, the addition of the resin changes the way of the setting reaction in the RM GICs and consequently results in a larger shrinkage compared to that of CGICs (, -).

In this study, the values of linear curing shrinkage of the CGICs were lower than those of the light-polymerized materials were and ranged between 0.47% and 1.32%, while RM GIC showed linear shrinkage between 1.11% and 1.65%, and giomer had shrinkage values between 0.80% and 1.15%. KM had significantly lower shrinkage compared with F9E and F9F. In addition, KMQ showed lower shrinkage compared to F9E and F9F, but statistical significance was demonstrated only in the comparison of KMQ with F9F. Lower shrinkage of KM and KMQ compared to F9E and F9F may be attributed to the powder glass particles size. The mean particle size in KM and KMQ is lower (4.7 µm and 6.6 µm, respectively) compared to that in F9E and F9F (14.9 µm and 7.13 µm, respectively) (, ). Small particles imply lower shrinkage due to a smaller distance between the particles and a smaller volume ratio of the matrix (). Some authors have recognized that a higher powder/liquid ratio leads to lower shrinkage because the part of the particles remains unreacted in the high-viscosity matrix (). This is contrary to our results because F9E and F9F, which were formulated with a greater powder/liquid ratio (F9F 3.6 g: 1 g, F9E 4.0 g: 1.2 g) (), had higher values of shrinkage than KM and KMQ with lower powder/liquid ratio (KM 3.4 g: 1g and KMQ 3.4 g: 1 g). However, comparing KM and KMQ, which both had the same powder/liquid ratio (3.4 g: 1g), but different particle size (KM 4.7 µm and KMQ 6.6 µm), it can be concluded that greater impact on curing shrinkage was exerted by the particles size than by the powder/liquid ratio.

Despite the large amount of published data about polymerization shrinkage, it is difficult to compare the results of different studies because of some variations in methodology, different testing conditions, different GIC materials investigated, and curing using different light intensity. Several studies indicated that volumetric shrinkage of different GICs and resin composites varied from 2% to 5% (, , , ). This volumetric shrinkage corresponds to the linear shrinkage of 0.75% to 1.72%, which coincides with our results of shrinkage in the range of 0.47% to 1.65%. However, the concrete numerical values obtained in our study cannot be directly compared to those from other reports due to unique specimen geometry and measurement conditions. The main limitation of our setup is that positioning the discoid specimens with a high diameter / height ratio between the PET foils, to which the materials partially adhered, might have acted as external constraint on the setting shrinkage. Moreover, the mere weight of the mirror positioned on the unset specimen surface might have interfered with the extent and direction of the setting shrinkage.

An additional important factor in polymerization shrinkage of light-curable materials is the curing mode applied. From the Figure 2, it is apparent that the curing shrinkage of B2 and F2LC depended on the light intensity. B2 showed significantly less shrinkage when polymerized using the “low” curing mode compared to the “high” curing mode. The influence of the curing mode was also observable in F2LC but the difference was not statistically significant. During the light curing of B2 and F2LC, the “soft” curing mode led to the lower shrinkage in the stage of exposure to the lower intensity light and increased with the increase in light intensity. For the giomer, the “low” curing mode produced significantly lower curing shrinkage than the “high” curing mode. Some authors concluded that reduced light intensity leads to slower curing and be beneficial for reducing the setting shrinkage (). In this study, F2LC showed a more extensive shrinkage than B2. It is possible that the acid-base reaction occurring during the manufacturing of giomer resulted in large pre-polymerized S-PRG fillers that produced less curing shrinkage (). Other authors also reported higher shrinkage of RM GIC than composite materials, which concurs with our findings of a higher shrinkage in F2LC compared to that in giomer (, ).

It should be emphasized that test conditions in our study corresponded to the clinical conditions of restoration placement. Measurement lasted a little bit longer than the manufacturer recommended finishing time. We measured shrinkage of material prior to their exposure to saliva or aqueous media. In some previous studies, the influence of different test condition has already been proven (). Due to the hydrophilic nature and the structure of glass ionomer cement, water sorption leads to partial compensation of the shrinkage (). However, the material passes through a certain extent of dimensional changes before any contact with the saliva. Thus, the setting shrinkage takes place while the cement is still protected against water exposure.

Conclusion

The results of this study show that dimensional changes during setting depended on the material type: the CGICs showed a lower degree of shrinkage than the RM GIC and the giomer. The first hypothesis was partially accepted; considering the comparison between KM and KMQ with F2LC polymerized with all of the three curing modes, as well as the comparison between KM with B2 polymerized with the “high” curing mode.

The differences among curing modes were statistically significant only in B2 between the “high” and the “low” curing mode. There was no significant difference in F2LC between the three curing modes.

The material expansion during the initial setting time, which reduces the net amount of curing shrinkage, was higher and lasted longer in the CGICs compared to the RM GIC and the giomer.