In vitro determination of the mechanical and chemical properties of a ﬁ bre orthodontic retainer

SUMMARY (cid:3) The aim of this study was to analyse, in vitro , the chemical and mechanical properties of a new ﬁ bre retainer , Everstick, comparing its characteristics with the requirements for an orthodontic retainer. Chemical analysis was used to examine seven ﬁ bre bundles exposed to a photocuring lamp and then to different acids and resistance to corrosion by artiﬁ cial saliva fortiﬁ ed with plaque acids. The mechanical properties examined were tensile strength and resistance to ﬂ exural force. Ten ﬁ bre samples were tested for each mechanical analysis and the mean value and standard deviation were calculated. Wilcoxon signed rank test was used to e valuate change in weight after treatment in each group. To determine change s over time between the groups for each acid considered separately , both repeated measures analysis of variance ( ANOVA ) on original data and on rank transformed data were used. If the results were different , ANOVA on rank-transformed data was considered.


Introduction
recognized that after orthodontic treatment , the teeth tend to return to their original positions because the periodontal tissues have not had suf cient time to reorganize themselves. During this reorganization period , the teeth must be maintained in the position achieved orthodontically, by means of removable or  xed retention systems. When relapse is expected, a splint is often used on completion of orthodontic treatment to resist the dislocating forces acting on the teeth. These forces ( Littlewood et al. , 2006 ;Edman Tynelius et al., 2010 ) are mainly occlusal, those exerted by the masticatory muscles or tongue, and elastic rebound forces are exerted by the periodontal tissues. The resultant force is transmitted to the teeth and then to the splint, principally as  exural forces. Retention is consequently important and must continue until periodontal reorganization has been fully achieved. Stainless steel (SS) orthodontic sectional archwires are often used in the anterior lingual segment as  xed retention. These , however, have some drawbacks due to the excessive stiffness of the wire, which can delay tooth reorientation and stabilization after orthodontic treatment. SS and chromium -cobalt alloys ( Rucker and Kusy, 2002 ;Zachrisson, 2007 ), which are currently used for retainers, have a coef cient of elasticity that is 10 times higher than that of bone. Glass  bres have been proposed as orthodontic retainers ( Karaman et al. , 2002 ) to overcome the drawbacks associated with the use of metal wires for orthodontic retention.
New invisible glass  bres that are bonded with  owable composite have recently been introduced for use as retainers. Failure has been reported to occur with glass- bre retainers ( Tacken et al. , 2010 ), chie y due to fracture or detachment of the retainer from the tooth surface, allowing uncontrolled tooth movements.
The substructure of the  ller composite resin has also been studied to determine its compressive fatigue limits ( Kurtulmus et al. , 2010 ), the importance of the type of resin and of the location of the  bre reinforcements; these factors signi cantly in uence  exural strength ( Narva et al. , 2005 ) and the ability to impregnate the  bres with polymer matrix 694 A. SILVESTRINI-BIAVATI ET AL. A. SILVESTRINI-BIAVATI ET AL.
2 of 5 ( Vallittu, 1999 ). van Heumen et al. (2008) found that  bre architecture (woven versus unidirectional) is more important than the type of  bre in determining  exural strength and  exural modulus. With regard to failure due to undesired debonding of  bre-reinforced composite (FRC), Lassila et al. (2007) reported that the shear bond strength values were highest when the  bres were orientated perpendicular to the bonding surface. Tezvergil et al. (2003) found that the bond strength of FRC did not differ from that of particulate  ller composite and that the addition of  owable composite did not improve bond strength values.
It has , however , also been reported ( Foek et al. , 2009 ) that SS orthodontic bonded retainers deliver higher bond strengths than  bre retainers, the difference being statistically signi cant. Brauchli et al. (2009) , who investigated  ve  owable composites to test whether composite properties are important for the long-term stability of retainers, concluded that all were equally effective. Littlewood et al. (2006) in a stud y of randomized controlled trials on children and adults concluded that there was at present insuf cient data on which to base clinical practice. Rose et al. (2002) found that in terms of reliability for permanently  xed orthodontic retention from canine to canine, direct-bonded multistranded wire was superior to  bre-plus-resin ribbon retainers.
All types of retainer are at risk of attack by acids produced by dental plaque that consist of microbial  ora comprising numerous aerobic and anaerobic bacteria, together with salivary components ( Mashimo et al. , 1981 ). The production of acid is triggered by the ingestion of sugary substances and causes pH variations in the oral environment. The Stephan curve ( Higham and Edgar, 1989 ) subdivides the phases of acid generation in the plaque into three areas: acid production , minimal pH , and acid elimination. In the  rst two phases, the pH decreases from the normal value (mean pH = 6.8) to a critical pH of approximately 5.5 or less .
The aim of the present study was to analyse the mechanical and chemical properties of the EverStick®Ortho  bre bundle (Stick Tech Ltd, Turku, Finland) and to compare its intrinsic qualities with the optimal requirements for an orthodontic retainer: resistance to occlusal forces and acid attack, minimal dimensions, and  exibility. Resistance to corrosion in the oral environment was analysed chemically, considering the action of plaque acids and pH variations after consumption of sugary foods , and resistance to normal biting forces (tension and bending) was examined mechanically.

Materials and methods
The tested product EverStick ® Ortho ( Tezvergil et al. , 2003 ) is a semi-manufactured product for direct tooth retention, comprising glass  bre strands plus a polymerresin gel matrix (PMMA + BIS-GMA), that is used to reinforce dental acrylic polymers and composites. Each bundle contains 1000 individual glass  bres, providing an effective diameter of 0.75 mm and a cross-section of approximately 0.5 mm 2 . The  bres are unidirectional, increasing the strength of the stick and its stiffness perpendicular to the  bre direction. The  bre bundles are marketed in foil packages to protect them from light; each package contains 2 × 12 cm of pre-impregnated glass  bre. The  bre bundles were removed from the package and exposed for 40 seconds/cm 2 to a light-emitting diode photocuring lamp (Starlight Spa, Mectron, Carasco, Italy) with a luminous radiation wavelength of 470 nm.

Chemical analysis
To test the effect of corrosion by different acids, seven  bres (length: 5 cm) were weighed to four decimal places using a scale ( AE 240 Mettler Toledo, Novate Milanese, Italy) and placed in a numbered beaker, containing 10 ml of a different acid. Pure acids (Carlo Erba s.p.a., Milan, Italy) were employed in racemic mixtures, at a concentration of 100 per cent ( Table 1 ). Lactic, formic, acetic, propionic , and butylic acids are produced by bacteria normally present in plaque; lactic and formic acids require the presence of sugars, whereas acetic, propionic , and butylic acids are produced in the absence of sugars. Phenylacetic acid is produced by Porphyromonas g i ngivalis , in subjects with periodontal disease ( Mashimo et al. , 1981 ), and hydro uoric acid is formed in the mouth under particular conditions ( Lindhe et al. , 2008 ).
The beakers containing the  bre bundles were covered with plastic  lm to prevent contamination and placed in storage at 37°C for 14 days. The bundles were then removed and washed in deionized water, ethyl alcohol , and diethylic ether to evaporate the absorbed liquids. The  bres were left to dry at 40°C for 2 days and then re weighed.
Resistance to corrosion was tested in arti cial saliva forti ed with plaque acids, reproducing as closely as possible in vitro the biochemical characteristics of the oral environment. Arti cial saliva, with a pH of 7.6, was prepared as follows ( Kurtulmus et al. , 2010 ): 1.47 g KCl, 1.25 g NaHCO 3 , 0.517 g KSNC, 0.188 g KH 2 PO 4 , deionized H 2 O to 1 l.
Four beakers were  lled with 100 ml of arti cial saliva. In two of these , the pH was decreased to a value of 6.8 by adding acetic and propionic acid, respectively; in the other two , it was reduced to 5.5 using lactic and formic acid, respectively.
pH values were determined with a pH meter (Model 336, Amel, Milan, Italy), an electronic instrument with glass electrode sensitive to H + ion concentrations calibrated at pH values of 4 and 9 using standard solutions to cover the study range.
The  bres were weighed and placed in the beakers containing the arti cial saliva solutions for 14 days, after which they were washed and dried as described above.

M echanical analysis
The tensile properties of the  bres were examined with dynamometer, model 5565 ( Instron Norwood, Maryland, USA), with a 5KN working head interfaced with a linear variable displacement transducer data processor , (model 24DCDT, Hewlett-Packard, Palo Alto, California, USA) ( Dieter, 1986 ;Vallittu et al. , 1998 ). The machine was set up to test cylindrical samples, operating at a crosshead speed of 20 mm/minutes, at a  eld speed of 20 pt/s, at 27°C and 70 per cent humidity, with a distance between the application clamps of 1.5 cm.
The following parameters were determined: Young ' s modulus of crystalline solids ( Vallittu et al. , 1998 ) deformation to rupture and stress to rupture.
Resistance to  exural forces was determined with a dynamo tacograph machine (Roell Korthaus Lic. Haan, Germany), based on counter-rotating screws, moved by a direct current engine with adjustable excitation, variable voltage, two-speed electromagnetic engagement gear ( Vallittu, 1994 ;Vallittu et al. , 1998 ); and a loading cell type U1 (Hottinger, Baldwin Messtechnik, Darmstadt, Germany) with strain gauges, class 0.03 with central power type MG 350B, class 0.01. The machine was set to perform  exural tests on  bres whose section had been measured using a digital gauge ( model 500 ; Mitutoyo Italia, Lainate, Italy). The tests were performed on 10 samples of  bres, their two extremities being placed at a distance of 10.4 mm. A force was applied to the mid-point of the  bre. A sensor that reproduced data graphically was attached to the machine. Maximum resistance to stress and  exural amplitude before fracture were determined.

Statistical analysis
A Wilcoxon signed rank test was used to evaluate change in weight after treatment for each group. To determine changes over time between the groups for each acid considered separately, both repeated measures ANOVA on original data and on rank transformed data were used. If the results were different, ANOVA on rank-transformed data was considered.
A P value of 0.05 was considered statistically signi cant. The Statistical Package for Social Sciences version 18 (SPSS Inc., Chicago, Illinois, USA) was used for computation.

Results
The results of the chemical analyses are shown in Table 2 (effects of corrosion by different acids) and Table 3 (resistance to corrosion tested in arti cial saliva with the addition of plaque acids). The  bres were tested in the most signi cantly corrosive plaque acids present in the oral cavity: the results are listed in order of the degree of damage caused by each acid ( Table 4 ). Acetic acid was the most corrosive, whether used pure or at a pH of 6.8 (salivary pH), and caused the greatest substance loss. Lactic acid, when tested pure, showed a slight tendency to corrode the  bre bundle; it was more active at pH 5.5, causing 8.18 per cent substance loss over the 14 day period. The same was found for propionic and formic acids, although their mean oral concentrations were lower than those of lactic and acetic acids, and their activity on the  bres was also less ( Table 3 ). Phenylacetic acid, when used pure, caused a 21.12 per cent corrosion of the  bre bundle. However, this acid is present only in the vicinity of pockets of periodontally damaged teeth and in very low concentrations in salivary

of 5
 uid, where its concentration increases following scaling and root planing ( Mouton, 1997 ). Hydro uoric acid was shown to be the most corrosive ( Table 2 ). For all acids analysed in both groups (lactic, formic, acetic, propionic) , changes after treatment were statistically different between two groups ( P < 0.001 for lactic, acetic, and propionic; P = 0.004 for formic acid). The mean results concerning  bre tensile properties were: Young ' s modulus 78 793.84 MPa (SD 12 063.35); deformation to rupture 3.780 per cent (SD 0.47); stress to rupture 1842.40 MPa (SD 278.83). Three of the 10 samples were not included when calculating the mean value because they differed too greatly from the median values. The stress -strain curve ( Figure 1 ) shows that the physical behaviour of the  bres under strain can be divided into three phases.
Young ' s modulus is a numerical expression of the elastic potential of a material; the value of the modulus is inversely related to its elasticity. The Young ' s modulus value showed the tested  bres to be superior to intertwined 0.0215 inch steel wire, whose Young ' s modulus, ≈ 210 000 MPa ( Rucker and Kusy, 2002 ;Zachrisson, 2007, van Heumen et al. , 2008, is approximately three times higher, indicating three times less elasticity. The deformation the  bre bundle can withstand before separating into its constituent elements (glass  bre and composite) and losing its mechanical properties, calculated for a length of 1.5 cm, was found to be 3.9 per cent. This corresponds to a linear measurement of 1.16 mm: the  bre bundle can thus withstand deformation of up to 1.16 mm before rupture.
This range allows safe clinical application. The stress to rupture of  bres was 1546 MPa, which equals 157.67 Kg/ mm 2 (where MPa = N /mm 2 ; N = Kg × g ; g = 9.81).
The mean results of the  exural tests were: stress to rupture 534.53 MPa (SD 1.401). The  bres showed a resistance to bending of 534 MPa, or 5448 Kg/mm 2 . The de ection produced was 1.4 mm over a length of 12 mm.

Discussion
The results of the chemical analyses showed that acetic acid was the most corrosive, whether used pure or at a pH value of 6.8 (salivary pH) as it caused the greatest substance loss. This was signi cant for the acid ' s activity at salivary pH, where corrosion of 10.8 per cent in 14 days was observed, rather than for its maximun -concentration activity. Phenylacetic acid, when used pure, caused 21.12 per cent corrosion of the  bre bundle; this acid was included in the study in order to determine whether periodontal disease is a contraindication for the use of glass - bre retention systems.
If plaque is well developed, acetic acid is produced in the absence of sugary substrate and is therefore constantly present in the oral cavity in such conditions. Hydro uoric acid was shown to be the most damaging acid and may induce modi cations to the glass  bre.
The results of the  exural tests showed that the  bres possess a resistance to bending of 534 MPa, or 5448 Kg/mm 2 , which is more than suf cient to oppose the occlusal forces that are transmitted to the retainer through the teeth. The occlusal force during normal activity (mainly during mastication) is Figure 1 Stress-strain curve of the physical behaviour of the  bres under strain: in the  rst phase,  bres and resin are most elastic to mechanical stimuli (almost horizontal portion of the curve); in the second phase (steepest part of the curve), the  bres begin to separate from the surrounding resin and show a lower tendency to deformation, until they break (peak of the curve); in the third part (descending portion), they completely lose their physical characteristics. The different lines represent the physical behaviour of the seven  bres that were tested under strain .  ( Zachrisson, 2007 ), while under forced contraction (as in case of parafunctional activity) , it can reach ≈ 100 N or 10 Kg: these values are within the resistance capability of the tested  bre bundle. The de ection produced was 1.4 mm over a length of 12 mm, suggesting that the  bre bundle is capable of resisting tooth displacements such as buccal movements, rotation and extrusion, or other movements that may occur during post-treatment relapse.

Conclusions
The results of the study show that the mechanical properties of the  bre correspond to the requirements of an orthodontic retainer. It appears that the tested  bres may appropriately be used for holding closed diastemas or post-extraction spaces and for positioning derotated teeth since the forces that may cause a recurrence are between 10 and 100 times lower than the resistance capability of the tested  bres.
With regard to the forces necessary to cause  bre rupture, the  ndings demonstrated that the strength of the  bre was between 10 and 100 times in excess of clinical requirements; the  bre bundle was also shown to be suf ciently strong to oppose occlusal forces.
With regard to chemical properties, the  bre bundle was attacked by acids that are potentially present in the oral cavity. The degree of aggressiveness, depending on the concentration of these acids, may affect the mechanical properties of the  bre bundles. Thus , it is clear that in order to preserve the  bre bundle in the long term, oral hygiene is important both in normal post-orthodontic patients and, especially, in those with gingival in ammation or periodontal disease. Careful plaque control is necessary to increase the life of the retainer. However, clinical and longitudinal studies will be needed to con rm these in vitro results .