Dimensional change following setting of root canal sealer materials

Dimensional change following setting of root canal sealer materials

dental materials Dental Materials 17 (2001) 512±519 www.elsevier.com/locate/dental Dimensional change following setting of root canal sealer materia...

225KB Sizes 1 Downloads 23 Views

dental materials Dental Materials 17 (2001) 512±519

www.elsevier.com/locate/dental

Dimensional change following setting of root canal sealer materials Dag érstavik*, Inger Nordahl, John E. Tibballs NIOMÐScandinavian Institute of Dental Materials, PO Box 70, N-1305 Haslum, Norway Received 13 June 2000; revised 16 October 2000; accepted 11 January 2001

Abstract Objective: The study was designed to evaluate a method proposed for measuring dimensional changes of endodontic sealers, and to assess the dimensional changes of 11 commercial sealers after prolonged storage in water. Methods: The method for linear dimensional change described in the draft standard for endodontic sealers was applied to 11 different types of endodontic sealers. One material (Sealapex) could not be tested by the method. The other 10 materials were followed for dimensional change over 48 weeks. Results: The sealers showed markedly different dimensional properties. For most materials, the greatest dimensional changes took place within the ®rst 4 weeks. Zinc-oxide-eugenol based sealers generally showed shrinkage ranging from 0.3 to 1%, while one product (ProcoSol) exhibited expansion exceeding 6% after prolonged storage. The epoxy-based materials, AH 26 and AH 26 silverfree, exhibited a large, initial expansion of 4±5%. AH Plus expanded from 0.4% after 4 weeks up to 0.9%. Apexit, a Ca(OH)2-based material, showed only minor variation round baseline value, 20.14 to 10.19%. Roeko-Seal expanded to 0.2% within 4 weeks, but was stable thereafter. Signi®cance: The test methodology adequately assessed dimensional changes exceeding ^0.2%, but some brands of material either could not be made into adequate test specimens or showed surface changes which interfered with dimensional change measurements. Theoretical approaches to the consequences of expansion by materials of low bulk strength question the necessity of a strict requirement against expansion, whereas bacterial penetration may be a real threat from sealers shrinking as little as 1%. q 2001 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. Keywords: Endodontic sealers; Dimensional change; Test method; Standard speci®cation; Root fracture; Bacterial penetration

1. Introduction The purpose of the endodontic ®lling is to seal off the root canal and to prevent ingress of bacteria from the oral environment to the periapical tissues [1]. The physical properties necessary for this function include adaptation and adhesion to the root canal dentine surface and dimensional stability of the ®lling. It has been shown that root ®llings may be prone to bacterial penetration along their entire length [2]. Dimensional changes of root canal sealers over time may introduce gaps and channels along the sealer/dentine or sealer/guttapercha interface, channels which may be large enough to permit micro-organisms to pass along the spaces. The dimensional stability of root canal sealers is therefore relevant for appropriate function of the root ®lling. Few studies have been carried out on the dimensional stability of sealers. Kazemi et al. [3] used a volumetric technique to study dimensional changes in sealers based on epoxy resin, zinc-oxide-eugenol (ZOE) and silicone. They found that the * Corresponding author. Tel.: 147-6751-2205; fax: 147-6759-1530. E-mail address: [email protected] (D. érstavik).

largest and most rapid dimensional changes occurred with ZOE-based sealers, the least and slowest with the siliconebased material. As most endodontic sealers are of chemical formulations similar to materials used in operative or prosthodontic procedures, information may be sought from studies of such materials. Again, surprisingly few studies have been performed. Glass ionomer-based cements and resin-based composites are generally associated with a degree of shrinkage, typically in the order of 0.5±1.5% within minutes after setting [4]. Zinc-oxide-eugenol-based materials are also believed to shrink over time, but little experimental work is available to document this [5]. Dental elastomers are typically tested by standard procedures [6]. Condensation-type silicones show a relatively large setting contraction (up to 0.5% after 24 h), whereas the addition-type silicones have a high degree of dimensional stability (approximately 0.1% after 24 h) [7]. Dimensional stability has been introduced as a requirement in the draft international standard (DIS) for root sealing materials [8]. The requirements for compliance with the standard have been set at a linear expansion of not more than 0.1% or shrinkage of not more than 1%.

0109-5641/01/$20.00 + 0.00 q 2001 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. PII: S01 09- 5641(01)0001 1-2

D. érstavik et al. / Dental Materials 17 (2001) 512±519

513

Table 1 Sealers tested Type

Product

Manufacturer

Ingredients

Zinc oxide eugenol

Proco-Sol

DEN-TAL-EZ, Inc., Lancaster, PA, USA

Grossman's sealer a

NIOM mix

Pulp Canal Sealer

Kerr USA, Romulus, MI, USA

Tubli-Seal

Kerr Italia S.p.A., Scafati, Italy

AH 26 AH 26 silverfree

Dentsply De Trey GmbH, Konstanz, Germany Dentsply De Trey

AH Plus

Dentsply De Trey

Apexit

Vivadent Schaan, Liechtenstein

Sealapex

Kerr Italia S.p.A.

Ketac-Endo Aplicap Roeko-Seal Automix

ESPE Dental AG, Seefeld, Germany ROEKO GmbH 1 Co, Langenau, Germany

Powder: ZnO 38.6%, hydrogenated resin 28.8%, (BiO2)2CO3 14.4%, BaSO4 14.4%, Na2B4O7 3.8%; liquid: eugenol Powder: ZnO 38.6%, hydrogenated resin 28.8%, (BiO2)2CO3 14.4%, BaSO4 14.4%, Na2B4O7 3.8%; liquid: eugenol Mixed paste: ZnO 41.25%, precipitated silver 20.25%, eugenol 19.5%, thymol iodide 3.75%, various resins Mixed paste: ZnO 59%, resins 14%, BaSO4 4%, thymol iodide 3%, oil 8%, modi®ers 2%, eugenol 10% Powder: Bi2O3 60%, methenamine 25%, silver 10%; TiO2 5%; liquid: bisphenol-A-diglycidylether 100% Powder: Bi2O3 80%, methenamine 20%; liquid: bisphenol-Adiglycidylether 100% Bisphenol-A-diglycidylether, calcium tungstate, iron oxide, ZrO2, adamantane amine, diamines, silicon oil Mixed paste (approx.): Ca(OH)2 15.9%, hydrogenated colophony 15.8%, silicon dioxide 0.4%, salicylates 18.2%, Bi salts 18.2% Mixed paste: CaO2 24%, BaSO4 20%, ZnO 7%, salicylate resins, ethyl toluene sulphonamide Poly maleinate glass ionomer Addition type silicone

Epoxy polymer

Salicylate-Ca(OH)2

Glass ionomer Silicone a

Ref. [9].

The purpose of the present study was to apply the method proposed in the DIS to assess the dimensional stability of several commercially available root canal sealers, representing a variety of chemical formulations. A secondary purpose was to assess practical aspects of the method in order to judge its suitability as a standard for dimensional change measurements of this group of materials. 2. Material and methods Table 1 shows the 11 different root canal sealers used in this study. The sealers were all commercially available except Grossman's sealer, which was mixed from reagent grade components at NIOM Laboratory. 2.1. Test specimens Cylindrical test specimens, with a diameter of 6 mm and a height of 12 mm, were made in split moulds according to clause 5.6 in the DIS [8]. One product, Apexit, was also tested according to subclause 5.6.2.b, which speci®es a procedure that includes mixing water into the material before the test specimens are made. The moulds were lubricated with Luvax (5% in hexane), slightly over®lled with mixed sealer and backed by a thin polyethylene ®lm and a glass plate on each side. The mould and plates were held ®rmly together with a clamp. Five minutes after start of mixing the mould with clamp was transferred to a cabinet with 95±100% relative humidity and kept at 378C (^18C). Sealers with a setting time up to

2 h (Grossman's sealer, Ketac-Endo Aplicap, Pulp Canal Sealer, Roeko-Seal, Sealapex and Tubli-Seal) were kept in the cabinet for at least three times the measured setting time. Sealers with a setting time longer than 2 h (AH Plus, AH 26, AH 26 silverfree, Apexit and Proco-Sol) were kept in the cabinet overnight (18±26 h). After setting, the ends of the test specimens were ground ¯at on wet silicone carbide paper (Struers FEPA P#1000, 600 grit) before being removed from the mould. 2.2. Measuring equipment The apparatus for measuring the length of the test specimen consisted of a 6 V displacement transducer (DC-DC LVDT, Model No. 0244-0000, Trans-Tek Inc, Ellington, CT, USA) and a table for the test specimen. The transducer with the sliding core and the table were mounted in a holder as shown in Fig. 1. A bottom plate made of brass, with a diameter of 7 mm, was mounted at the lower end of the sliding core to ensure that the measurement pressure was uniformly distributed over the ends of the specimen. A cap of Te¯on attached on the top end of the core acted as a handle for raising the core. The sliding core including the bottom plate and Te¯on top had a total weight of approximately 3.8 g, well within the requirements of the draft standard, which speci®es an instrument with a measuring force of no more than 0.1 N. The table for the test specimen was made of stainless steel and placed at a distance of approximately 25 mm from the bottom of the transducer. On the table two concentric circular marks (5 and 7 mm diameter) for the specimen were placed in the longitudinal axis of the sliding core. The working distance between the

514

D. érstavik et al. / Dental Materials 17 (2001) 512±519

Fig. 1. Schematic illustration of equipment used for measuring the height of test specimens. a, Te¯on cap for raising the sliding rod; b, sliding core; c, housing for transducer; d, brass bottom plate; e, test specimen; f, circular base plates.

table and the bottom-plate of the sliding core was approximately 14.2 mm. The signal from the transducer was transferred to a Radiometer REC 61 servograph, at 20 mV cm 21 and recorded on paper at a chart feed of 1 cm min 21. The instrument was calibrated by means of a dial gauge with an accuracy of 1 mm and the Radiometer servograph was adjusted so that a 10-mm change in the position of the sliding core was recorded as 10 mm on the paper. The paper had 2 mm markings and the measuring accuracy of the instrument was thus 2 mm. The full range of the servograph corresponded to 250 mm. 2.3. Measuring procedure All measurements were performed at a temperature of 238C (^28C) and a relative humidity of 50% (^5%). Three cylindrical brass specimens, with heights of 11.50, 12.00 and 12.50 mm and a diameter of 8 mm, were used for calibration of the measuring device. The appropriate brass cylinder was centered on the circular marks on the table and the sliding core was carefully lowered until the bottom-plate of the sliding core came in contact with the cylinder. The height of the relevant brass specimen was recorded as `zero' on the chart paper. After setting and polishing as described above, the length of each test specimen was measured. The specimen was placed on the table inside the 7-mm circular marking, and the sliding core of the transducer was lowered until the bottom plate of the core rested on the top of the specimen. The test specimens of some of the root sealing materials seemed to be deformed elastically when subjected to the pressure of the sliding core during measurement. The test

specimens were therefore as a rule kept in the measuring apparatus for 3 min before the reading was done. For the most elastic material (Roeko-Seal Automix) the reading was done after 5 min, at which time the curve on the paper had reached a stable plateau. Between measurements, the specimens were stored separately in distilled water at 378C (^18C) in small glass containers with lids of polyethylene. The test specimens were re-measured every 4th week for up to 48 weeks. Before each measuring procedure the test specimens were conditioned to room temperature resting in the distilled water in their containers. The distilled water of the containers was not changed during the testing period. For each measurement, the conditioned test specimen was removed from the distilled water and top and bottom ends were blotted with paper tissue. The test specimen was placed in the measuring apparatus and the length recorded as described above. The change in length was calculated as a percentage of the original length. Three specimens of each material were tested, and the mean (^ standard deviation) per cent change in length was recorded as the dimensional change. For each set of three specimens, the statistical signi®cance of changes in dimension over time was calculated using the Scheffe test. Pilot tests showed that the surface of test specimens of Ketac-Endo Aplicap cracked rapidly when kept in air following the conditioning period. Therefore, a 9 mm wide belt of moist ®lter paper was applied immediately around these specimens after removal from the container to protect them from drying. This moist belt remained in place on the test specimen for the duration of the measurement. The silicone-based material (Roeko-Seal) was also tested according to the test method for linear dimensional change described in the international standard for elastomeric impression materials [6]. This method was used for specimens stored both dry and wet in distilled water between measurements every 4th week. 2.4. Measures of uncertainty 2.4.1. Repeatability testing Whereas the draft standard speci®es a measurement accuracy of 1 mm, the actual recording is to be done to the nearest 10 mm. The LVDT used in our set-up could perform measurements to 2 mm, which would be satisfactory for the measuring requirements of the standard. The high sensitivity of the measuring device made it necessary to apply some methodological statistics. For repeatability measurements, one specimen of each material was re-measured 10 consecutive times with instrument calibration performed each time. The results (Table 2) showed a very high repeatability of individual measurements, with standard deviations ranging from 2 to 7 mm and the range of each set of 10 measurements was from 4 to 24 mm. Two standard errors of the mean divided by the mean was calculated as a

D. érstavik et al. / Dental Materials 17 (2001) 512±519

515

Table 2 Repeatability measurements (n ˆ 10) Product

Mean

Max

Min

Range

SD

SE of mean

Uncertainty a

AH 26 AH 26 silverfree AH Plus Apexit Grossman's sealer Ketac-Endo Aplicap Proco-Sol Pulp Canal Sealer Roeko-Seal Automix Tubli-Seal

12.504 12.562 12.149 11.988 11.948 11.849 12.725 11.895 11.949 11.970

12.510 12.572 12.158 12.004 11.950 11.854 12.734 11.904 11.960 11.980

12.498 12.552 12.142 11.980 11.946 11.846 12.712 11.890 11.940 11.964

0.012 0.020 0.016 0.024 0.004 0.008 0.022 0.014 0.020 0.016

0.0043 0.0053 0.0053 0.0074 0.0018 0.0023 0.0069 0.0040 0.0063 0.0056

0.0013 0.0017 0.0017 0.0023 0.0006 0.0007 0.0022 0.0013 0.0020 0.0018

0.02 0.03 0.03 0.04 0.01 0.01 0.03 0.02 0.03 0.03

0.0156 0.024 0.004

0.00493 0.0074 0.0018

0.00156 0.0023 0.0006

0.026 0.04 0.01

Mean Max Min a

Uncertainty ˆ [(2 £ SE)/Mean] £ 100.

percentage and presented as the measurement uncertainty. Table 2 shows that the uncertainty ranged from 0.01 to 0.04%. 2.4.2. Measurement variability All measurements were done on three separate specimens of each material. This provided means, ranges, standard deviations and standard errors of the mean for each measured value. The maximum range of measured values was 150 mm, and the minimum 0. The standard deviation of the means of three measurements ranged from 0 to 38 mm and the 95% con®dence interval ranged from 9 to 45 mm (see Table 3). When the data were analyzed as per cent change of each specimen vs. its own base-line, the three measurements had ranges from 0 to 1.4%, with an apparent but weak tendency for the range to increase with time (Fig. 2). 3. Results The Sealapex sealer could not be made into satisfactory test specimens. The setting process was inconsistent Table 3 Ranges of measurements of three parallel specimens (mm) Sealer

Mean

Max

Min

SD

95-CI

n

AH 26 AH 26 silverfree AH Plus Apexit Grossman's sealer Ketac-Endo Aplicap Proco-Sol Pulp Canal Sealer Roeko-Seal Automix Tubli-Seal

0.117 0.047 0.060 0.045 0.029 0.012 0.048 0.066 0.013 0.016

0.150 0.070 0.070 0.060 0.040 0.030 0.070 0.100 0.020 0.030

0.020 0.040 0.010 0.030 0.010 0.000 0.010 0.010 0.000 0.000

0.038 0.009 0.017 0.011 0.010 0.010 0.018 0.023 0.008 0.009

0.075 0.019 0.034 0.021 0.019 0.020 0.036 0.046 0.015 0.017

13 13 13 13 13 13 13 13 13 13

and failed to produce stable cylinders of the speci®ed dimension. Attempts to incorporate water during mixing led to specimens which initially appeared workable, but which within days after setting swelled and thereafter disintegrated. No data could be obtained for this material. The results for the other 10 materials are shown in Fig. 3. For most materials, the greatest dimensional changes took place within the ®rst 4 weeks. However, some materials displayed changes up to 48 weeks of storage in water. The zinc-oxide-eugenol based sealers Tubli-Seal, Pulp Canal Sealer, and Grossman's sealer showed shrinkage ranging from some 0.3% (Tubli-Seal) to 1% (Pulp Canal Sealer). Contrarily, the ZnO-eugenol sealer Proco-Sol showed pronounced expansion of 4.5% after 4 weeks rising to over 6% during subsequent storage. Epoxy-based materials all showed some expansion. The older formulations, AH 26 and AH 26 silver free, both exhibited a large, initial expansion after 4 weeks of 4±5%, but were stable afterwards. AH Plus expanded only 0.4% after 4 weeks, with a slight but consistent continuous expansion up to 1.2% over the next months. Apexit, the Ca(OH)2-based material, showed a small contraction after 4 weeks (0.2%) followed by a gradual increase up to baseline at 24 weeks. When mixed with water to facilitate the setting process, this material appeared stable for the ®rst 4 weeks followed by a shrinkage of some 0.1% during the next 4-week period. Roeko-Seal showed a small expansion of some 0.2% within 4 weeks, and was stable thereafter. The results were virtually identical when this material was tested after wet storage with the equipment described in the international standard for elastomeric impression materials [6]. Specimens stored dry and tested by this method showed a small contraction initially and were thereafter stable.

516

D. érstavik et al. / Dental Materials 17 (2001) 512±519

Fig. 2. Ranges of three parallel measurements of dimensional change as a percentage of baseline for the different materials and times.

4. Discussion 4.1. The sensitivity and accuracy of the measuring device The selected transducer had a capability slightly less than that speci®ed in the draft international standard. However, since the procedure outlined in the draft does not take advantage of the accuracy (1 mm) speci®ed for the transducer, the selected accuracy of 2 mm was still adequate for the purpose of the test. Repeated measurements of the same specimen typically showed a range of 10±20 mm (min 4, max 24 mm), indicating that far greater uncertainty was introduced by the handling and positioning of the specimens than by the sensitivity of the instrument. Based on these ®ndings (Table 2) the use of 10 mm as the lowest unit of recording seemed appropriate. However, the discrimination of materials that exhibit an expansion of only 0.1% becomes problematic. That represents a 12 mm difference between specimens, while with an average 95% con®dence interval of 23 mm, the method only discriminates a 0.2% expansion. 4.2. Physical properties of root canal sealers Root canal sealers typically vary greatly as regards physical properties. Clinical success may be achieved with e.g. hard-setting as well as non-setting products; as a consequence, materials are proposed and marketed which have a very wide range of physical properties. This complicates the construction and operation of mechanical tests designed for quality assessment of this group of materials. While Sealapex has a good clinical [10] and biological [11] record as a root canal sealer, we found it impossible to produce specimens that could be used for testing of dimensional change according to the protocol. Moreover, although the draft standard allows water to be included when mixing this

sealer, our experience that this procedure resulted in complete disintegration of specimens is telling evidence that the materials should not be modi®ed to ®t the test. The other Ca(OH)2-based sealer, Apexit, had specimens produced both with and without water in the mix. This resulted in specimens with similar, but not identical dimensional properties: the slight initial shrinkage was delayed by several weeks when water was introduced during mixing. It would appear unjusti®ed to add water indiscriminately to setting sealers in order to produce physically acceptable test specimens. The clinical application of sealers limits their exposure to signi®cant temperature or humidity changes. The present test, however, involves several severe changes in ambient temperature and humidity. Brie¯y, the specimens set for hours at 95±100% relative humidity, but not wet, at 378C; they are then stored wet at 378C; but they are tested at 258C and 50% relative humidity. This imposes several types of stress on the specimens, as re¯ected by a pronounced tendency for materials to crack on the surface when equilibrated to the testing situation (ZnO-sealers and, particularly, Ketac-Endo Aplicap). For Ketac-Endo Aplicap, this was so pronounced that a means of preserving 100% humidity during testing was applied throughout, using a water-soaked ®lter paper covering the specimens at all times when they were out of the water bath for measurements. Moreover, some test specimens of Apexit exhibited a surface softening that continued and increased during the months of testing. Many specimens of different chemical types gradually revealed blisters, cracks or holes (Proco-Sol, AH Plus, Grossman's sealer, AH 26) indicating gas and bubble formation and release on the surface of the specimens. However, and again based on the repeatability and reproducibility of the method, these phenomena had apparently little effect on the actual linear dimensional change recordings.

D. érstavik et al. / Dental Materials 17 (2001) 512±519

517

Fig. 3. (a±d) Linear dimensional change of 10 endodontic sealers. (a) Epoxy resins; (b) Zinc-oxide-eugenol-based; (c) silicone-based (RS 4823wet and RS 4823dry are Roeko-Seal Automix tested according to the standard for elastomeric impression materials); (d) Ca(OH)2- and glass ionomer-based materials.

518

D. érstavik et al. / Dental Materials 17 (2001) 512±519

Fig. 4. Theoretical tooth root segment with calculations of pressure and hoop (tangential) stress. Note: The anisotropy in the elastic properties of dentine found by Palamara et al. [17] was reconciled with the isotropic model proposed by Kinney et al. [16] by assigning the following elastic stiffness constants (the third axis is parallel to the dentine tubuli): c11 ˆ c22 ˆ c33 ˆ 14.5 GPa; c12 ˆ c23 ˆ c31 ˆ 5.9 GPa; c44 ˆ c55 ˆ 5.1 GPa; c66 ˆ 8.6 GPa.

4.3. Dimensional changes of the various materials Four of the 10 materials tested actually complied with the requirements of the draft standard. After 1 month, the target time in the speci®cations, Apexit, Grossman's sealer, Tubli-Seal and Pulp Canal Sealer, while all shrinking, were within the speci®ed limit of 1.0%. Five expanded more than the requirement of max 0.1% (the AH series, Proco-Sol, Roeko-Seal). AH Plus appeared to shrink during setting prior to the ®rst measurement, yielding specimens with concave surfaces. Roeko-Seal showed values that were close to and by statistical criteria not signi®cantly different from the limit. One material showed shrinkage more than the acceptable 1.0% (Ketac-Endo Aplicap). With few exceptions, the materials appeared stable after 1 month of storage. The exceptions were AH Plus, which expanded consistently throughout the observation period, and Grossman's sealer and Pulp Canal Sealer, which continued to shrink. Previous studies on the dimensional stability

of ®lling materials of similar composition have usually followed changes during the actual setting process [4], and there seems to be few data on changes occurring during storage after setting [12,13]. The method used here records as baseline the height of the test specimen after complete setting, so that most or all of the setting-related volumetric changes occurred before the actual test. Even so, most materials exhibited signi®cant changes in the period after setting up to 1 month in storage. However, the test conditions change dramatically at the time when the ®rst measurement is performed: the specimens are transferred from a humid atmosphere to complete immersion in water. For several of the materials, this may lead to reactions with water, with possible consequences for their volumetric behavior. The mould used to produce the specimens measured 12.00 mm, however, the starting height of the ®nished specimens ranged from 11.94 to 12.06. This is in part a re¯ection of initial setting contraction/expansion, which is not addressed by the proposed test. For example, 12 of 18 test specimens of AH Plus were concave at their top and bottom, suggesting an initial setting contraction. Specimens of Roeko-Seal Automix were routinely measured to approximately 11.95 mm after setting, similarly suggesting an initial contraction. 4.4. Relevance of sealers' dimensional change It is generally held that physical blockage of the root canal from invading bacteria originating in the oral cavity is the primary function of the root canal ®lling. From this perspective sealers and other components of the ®lling should ideally be either volumetrically stable or increase slightly. If the ®lling material expands, there is a risk that the root may fracture. Setting and storage expansion of root canal sealers induces radial pressure on the pulpal aspect of the dentine. The risk of fracture is due to the associated tangential (hoop) strain, the magnitude of which is governed by the elastic moduli of dentine and the ®lling, the percentage expansion of the ®lling and tensile strength of dentine, e.g. [14]. Sealers with high bulk moduli (stiff and hard

Fig. 5. Calculation of maximum hoop stress in dentine. Data for President Light Body and Z100 published by Chabrier et al. [18] represent elastomer and composite, respectively.

D. érstavik et al. / Dental Materials 17 (2001) 512±519

materials) pose a greater threat than those with low moduli. Epoxy-based sealers exhibit high moduli while elastomerbase materials are expected to induce little strain in the dentine. We have used reported elastic moduli for dentine [15±17] and representative ®lling materials [18] to derive the stress pro®le in the dentine wall of a 3 mm diameter cylinder with a 1 mm diameter root canal. The maximum hoop stress, developed at the pulpal aspect, is proportional to the expansion and to the bulk modulus of the sealant (Fig. 4). Even for the stiffest composites (bulk modulus 16.5 GPa [18]) a total expansion of 0.5% leaves the hoop stress a safety factor of 2 with respect to the measured tensile strength of dentine (59 ^ 7 MPa [17]) (Fig. 5). Therefore, the consequences of a limited expansion by the sealers are highly material-dependent. In addition, the placement of gutta-percha, in itself a low-modulus material reduces the volume of the root canal affected by expansion by the sealer and can be expected to absorb some of the stress generated. Contraction appears to be a less desirable property in these materials. With a requirement of 1% maximum shrinkage it only takes a material thickness of 100 mm to produce a void of 1 mm. This is theoretically suf®cient for a multitude of micro-organisms to occupy and penetrate. Given the additional problems associated with completely ®lling the root canal and the organic material frequently left after instrumentation, an insuf®cient and shrinking root ®lling material may be more likely to threaten success [19] than a slightly expanding ®lling. There are no inherent requirements for strength of a root canal sealer other than those associated with preparation for a root canal post. Rather, they should show some elasticity to follow the strain of the root subjected to masticatory or other stresses. Sealers with a low bulk modulus and a slight expansion, such as materials based on elastomers, may offer advantages in this regard. 5. Conclusions The test methodology adequately assessed dimensional changes exceeding ^0.2% of 12-mm specimens. The accuracy of the method may not suf®ce for detection of dimensional changes of 0.1%, which has been suggested as a maximum limit for expansion. Some materials either could not be made into adequate test specimens (Sealapex) or showed surface changes which could interfere with dimensional change measurements. Some materials clearly exceeded the proposed maximum expansion limit for this group of materials (AH 26, AH 26 silverfree, Proco-Sol). One material (Ketac-Endo Aplicap) reached the proposed limit for contraction (1%). Theoretical approaches to the

519

consequences of expansion by materials of low bulk strength question the necessity of a strict requirement against expansion, whereas bacterial penetration may be a real threat from sealers shrinking as little as 1%. References [1] Sundqvist G, Figdor D. Endodontic treatment of apical periodontitis. In: érstavik D, Pitt Ford TR, editors. Essential endodontology: prevention and treatment of apical periodontitis, Blackwell Science, 1998. p. 242±77. [2] Barthel CR, Shuping GC, Moshonov J, érstavik D. Bacterial leakage compared to dye leakage in obturated root canals. Int Endod J 1999;32:370±5. [3] Kazemi RB, Safavi KE, Spangberg LS. Dimensional changes of endodontic sealers. Oral Surg Oral Med Oral Pathol 1993;76:766±71. [4] Kanchanavasita W, Pearson GJ, Anstice HM. In¯uence of humidity on dimensional stability of a range of ion-leachable cements. Biomaterials 1995;16:921±9. [5] Gilles JA, Huget EF, Stone RC. Dimensional stability of temporary restoratives. Oral Surg Oral Med Oral Pathol 1975;40:796±800. [6] International Standard ISO 4823:1992 Dental elastomeric impression materials, 1992. [7] Fano V, Gennari PU, Ortalli I. Dimensional stability of silicone-based impression materials. Dent Mater 1992;8:105±9. [8] Draft international standard ISO/DIS 6876.2:1999 Dental root canal sealing materials, 1999. International Organization for Standardization, Geneva 1999. [9] Grossman LI. Endodontic practice. Philadelphia: Lea & Febiger, 1978. p. 294. [10] Boiesen J, Eriksen HM, érstavik D. Clinical performance of endodontic sealers with Ca(OH)2. J Dent Res (special issue) 1994;73:382. [11] Leonardo MR, Silva LA, Utrilla LS, Assed S, Ether SS. Calcium hydroxide root canal sealers±histopathologic evaluation of apical and periapical repair after endodontic treatment. J Endod 1997;23:428±32. [12] Attin T, Buchella W, Kielbassa AU, Helwig E. Curing shrinkage and volumetric changes of resin-modi®ed glass ionomer restorative materials. Dent Mater 1995;11:359±62. [13] Sidhu SK, Sherriff M, Watson TF. The effects of maturity and dehydration shrinkage on resin-modi®ed glass-ionomer restorations. J Dent Res 1997;76:1495±501. [14] Sokolnikoff IS. Mathematical theory of elasticity. 2nd ed. New York: McGraw-Hill, 1956. p. 300. [15] Sim TPC, Knowles JC, Ng Y-L, Shelton J, Gulabivala K. Effect of sodium hypochlorite on mechanical properties of dentine and tooth surface strain. Int Endod J 2001;34:120±32. [16] Kinney JH, Balooch M, Marshall GW, Marshall SJ. A micromechanics model of the elastic properties of human dentine. Arch Oral Biol 1999;44:813±22. [17] Palamara JEA, Wilson PR, Thomas CDL, Messer HH. A new imaging technique for measuring the surface strains applied to dentine. J Dent 2000;28:141±6. [18] Chabrier F, Lloyd CH, Scrimgeour SN. Measurements at low strain rates of the elastic properties of dental polymeric materials. Dent Mater 1999;15:33±8. [19] Kirkevang L-L, érstavik D, Hùrsted-Bindslev P, Wenzel A. Periapical status and coronal restorations in a Danish population. Submitted for publication, 2000.