New York University

David B. Kriser Dental Center

College of Dentistry

Graduate School of Arts and Science                                                                                        Dec., 4, 1998

Department of Dental Materials Science

345 East 24th Street

New York, NY  10010

Tel.:  212-998-9637

Fax: 212-995-4085

Dr. Samuel Waknine

DRM Research Laboratories Inc.

29 Business Park Drive

Branford, CT  06045

Dear Dr. Waknine,

Please find enclosed a draft copy of an original research article entitled, ‘Bracket Shear Adhesion Strength (SAS) as a function of enamel etchant and bonding agent-luting cement : an In-Vitro study” to be submitted for publication to “The American Journal of Orthodontics”.  The study was undertaken by Dr. Won-Gun Chang as part of the thesis requirements for the Masters degree in Dental Materials Science under my supervision.  Additional results and analysis which do not form part of the MS thesis are also included in this manuscript.

This study explores the use of a weaker enamel conditioner specifically suited for Orthodontic purposes. The DRM Research Laboratory products, “DiamondBond”-“DiamondLink” (a dual cure adhesive resin bonding-luting cement) was evaluated along with, a visible light cure (VLC) resin adhesive and a triple cure Glass Ionomer cement; often used for direct bracket bonding.  The shear adhesive strength (SAS) values obtained for “DiamondBond-DiamondLink” combination with a Salicylic-Lactic acid organic conditioner was found to provide adequate strength for bracket bonding.  Additionally, a decreased period of etch time (< 15 seconds) with ortho-Phosphoric acid may also be used in conjunction with DiamondBond-DiamondLink combination for bracket bonding.

Please feel free to contact me for any clarifications or additional information.

We thank DRM Labs for providing the materials required for this study.

Sincerely,

T.V.V. Raghavan

T. V. Vijayaraghavan,  Ph.D.

Associate Professor

NYU Kriser Dental Center

Dental Materials Science

Bracket Shear Adhesion Strength (SAS) as a function of enamel condition and bonding agent: an In-vitro study

Won-Gun Chang, DDS, MS

T. V. Vijayaraghavan, BS, BE, MS, Ph D (*)

Associate Professor, Graduate School of Arts and Science

Department of Dental Materials

New York University Kriser Dental Center

345, East 24^th Street, Room # 1121S

New York, NY 10010

Tel:                (212) 998-9632 (Business)

                      (718) 349-6239 (Home)

Fax:               (212) 995 4889 (Business)

e-mail: tvv1@is2.nyu.edu

Introduction

Direct bonding of a metal or ceramic bracket to etched or conditioned enamel surface is achieved by the use of resin adhesives or cements[i] [ii] [iii] [iv] .The superficial enamel is cleaned with pumice and rubber cup and etch conditioned with 37% ortho-Phosphoric acid (37% (wt.%) o-H₃PO₄, pH <1, time range -15 to 30 seconds) to remove surface pellicles and debris, and provide retentive micro-spaces for interpenetration of adhesive or cement ^{4 [v]}.  The level of interpenetration depends primarily on the wetting characteristic of the bonding agent and the surface roughness of enamel.  A mechanical interlocking mechanism forms the sole basis for bond strength development with little or no chemical bonding [vi] [vii].  Under In-Vivo conditions, cement or adhesive interlocking provides for stability during initial placement of arch wires and subsequent stability under activation In-Vivo.  An inadequate bond strength may lead to a temporary suspension of treatment.  At the end of the treatment period (average of 2 years), the brackets are debonded and enamel fracture during debonding leads to loss of enamel[viii] [ix] [x].  The removal of adhesive or cement remnants in enamel after debonding may lead to surface scratches, cracking and loss of sound enamel. [xi] [xii].

Enamel fracture may be attributed to a prior enamel condition (subsurface caries) unrelated to the procedure or to the enamel conditioning method[xiii].  An acid of low pH (37 Wt.% ortho-H₃PO₄) and longer etch time may produce interconnected enamel pores below the surface, weakening subsurface enamel and resulting in enamel surface detachment or fracture during debonding[xiv] [xv].  A shorter etch time and a lower acid concentration may not provide adequate retention strength^{,[xvi] [xvii] [xviii] [xix]}.  Further, loss of sound enamel during preliminary prophylaxis and etch conditioning steps can vary from as low as 8 to 25 mm of enamel thickness, while subsurface effects from H₃PO₄ conditioning may affect enamel to a depth of 100 mm[xx].  The depth of enamel involved in bonding is often smaller than that affected by the procedure.

Compared to bonded restorations the strength requirements for bracket bonding are lower and a lower depth of etch has been pursued with weaker acids or shorter time of etching.  Other acids such Nitric acid[xxi],  Polyacrylic acid (in conjunction with the GIC), Pyruvic acid[xxii] and Lactic acid[xxiii] have been suggested as alternative agents to condition enamel.  A 2.5% Nitric acid with or without a water rinse has been suggested for retaining ceramic brackets, on labial surfaces considering the small depth of etching and the ease of removal without undue enamel loss instead of 37% phosphoric acid[xxiv] [xxv].

Frequently a separate etch step using a 10% Polyacrylic acid is also used when GICs are used for bracket bonding.  These conditioning treatments and adhesive combinations lead to lower bond strength than composite resin adhesives or luting cements on 37%H₃PO₄ conditioned enamel.  However, a resin-reinforced GIC is shown to provide sufficient adhesive bond strength in one study[xxvi].

Of additional importance is the question of the In-Vivo bond strength requirements for orthodontic function.  Newman[xxvii] suggests a shear strength of 3 MPa to be sufficient to withstand orthodontic forces under clinical conditions, while Miura et al[xxviii] suggest a value of 5.1 MPa based on their clinical study over a period of 2 years.  Reynolds² suggest that an interfacial bond strength in the range of 4.9 to 7.85 MPa as adequate for most clinical orthodontic needs.  Retief ³has observed enamel failures when the bond strength exceeds 13.5 MPa.  Although the enamel can often withstand much greater forces** , enamel fracture depends on the force mode (tensile, shear or torsion) and presence of flaws.  However, enamel failures have often been observed when the conditioning acid is 37 % H₃PO₄.  Secondly, the forces imparted to the bracket /adhesive/ enamel interface under In-Vivo conditions is complex and correlation to In-Vitro conditions is not direct.  In-Vitro test results may however, be used to estimate expected behavior under clinical conditions[xxix]

An organic acid[xxx] [xxxi] may function as efficiently in producing optimum adhesive strength sufficient for bracket adhesion and debonding, without undue loss of enamel.  Previous studies have shown that the bond strength development with organic conditioners is based on a mechanism of chelation of calcium in enamel with polyfunctional acids, unlike that for inorganic acids, where the mechanism is based on preferential enamel dissolution .  This study was undertaken to evaluate whether a mixture of Salicylic acid and Lactic acid can be used to condition enamel for direct bracket bonding with existing adhesive resins and cements.

The selection of the composition of organic acid is based on the study by Georgescu^{30 31} and Lee[xxxii] on enamel and dentin surfaces.  The presence of small amounts of Salicylic acid is expected to provide some antibacterial protection while Lactic acid is known to etch and/or form products with the calcium of the Hydroxyapatite in enamel

Objective

The purpose of this study is to compare the bracket (metal) Shear adhesive strength(SAS) on enamel for the following parameters;

1) enamel conditioned by 37% o-H₃PO₄ and 2) an organic acid ( Salicylic-Lactic acid) for-:

a) a VL cure adhesive resin, b) a triple cure Glass Ionomer Cement versus c) a dual cure adhesive bonding agent coupled with a resin luting cement.

Materials and Methods

Materials

Sixty three caries free extracted premolars were selected from teeth obtained from oral surgery and orthodontic departments of New York University Dental Center and Bellevue hospital.  After extraction, the teeth were washed and immersed in saline.  The teeth were selected based on visual observation of the soundness of the coronal portion and absence of caries.  Sixty three premolar buccal brackets (Victory Series, 3M Unitek Corp., Monrovia, CA), base torqued, 022 slot, Roth prescription metal bracket were used.  The meshed bracket base is arc shaped with an area of 12.19 mm² (data obtained from manufacturer).  The bracket is machined from 17-4 PH stainless steel and bonded to 304 stainless steel base.

Three types of acids, 1) Salicylic-Lactic acid liquid (pH = 1.7, 0.22% Salicylic acid and 9% Lactic acid), S; 2) 37% ortho-Phosphoric acid gel (DRM Res. Labs. Inc., Branford, CT), P, and 3) 10% Polyacrylic acid liquid (GC America Inc., Chicago, IL), A; were used.  In addition  enamel surfaces without conditioning were also evaluated, N.

Three types of cements; 1) a single cure(photo), VLC composite resin adhesive ‘Transbond-XL’, TX, (3M Unitek, Monrovia, CA), 2) a triple cure (photo, chemical and acid-base reaction), resin reinforced GIC, FG (Fuji LC ortho II, GC America Inc., Chicago, IL) and 3) a dual cure (photo and chemical cure) resin based bonding agent-luting cement combination; DiamondBond-DiamondLink, DL (DRM Res. Labs. Inc., Branford, CT) were used.  Table 1. lists the materials used.

Methods

The root portion of the tooth was excised using a low speed diamond saw (Buehler , Ltd., Evanston, IL).  The lingual area of the tooth was embedded in self-cure acrylic resin.  The buccal surface of each tooth was polished with pumice/water slurry for 10 seconds with a rubber cup, rinsed with water and air-dried.  The embedded tooth samples were divided into 7 groups on a random basis with at least 10 samples/group.  Table 2. lists the etch conditioning time as well the light cure time used for each of the groups.

The bonding agent was applied as per manufacturer’s instruction.  For PDL and SDL groups a universal bonding agent, DiamondBond (DRM Res. Labs. Inc., Branford, CT), base and catalyst mixture, was applied to the tooth surface and the luting cement, DL, was applied to the bracket base.  For PTX, STX and NTX group, an adhesive bonding agent (supplied with the adhesive resin with the kit) was applied to the tooth surface and resin, TX, was applied to the bracket base.  The bracket was adapted to the proper position of the buccal surface and pressed accurately with bracket placer for 5 seconds.  Once the bracket was in the correct position, the placer was removed.  The excess adhesive/cement was removed from the margin of bracket with a dental probe.  Each sample was cured using an Optilux 150 visible light curing unit (Demetron Corp., Danbury, CT) for 30 seconds via exposure to the bracket tooth interface.  The measured light output (density) of the unit was between 375 and 425 mW/cm².  For NFG and AFG groups, the cement was applied to the bracket base and positioned on the tooth as before.  All specimens were stored in distilled water for 24 hours at 37 °C prior to testing.

An “Instron” universal test machine (Model 1130, Instron Corp., Canton, MA) was used at a cross-head speed of 0.5 mm/minute (0.2”/minute) and the maximum load, P_f, at separation in shear was recorded on a chart recorder.  In order that the bracket/adhesive/tooth interface experience shear forces, the bracket bonded embedded tooth specimen was secured in a vise, such that the long axis of the tooth was parallel to the loading direction.  These adjustments were made prior to testing, to ensure maximum parallelism between the displacement direction and the interface plane.

The shear adhesive strength, SAS, was calculated in MPa from P_f, and the actual (measured) bonded area on the metallic bracket (bracket base area + excess resin/cement area attached to bracket).  The measurement of the area was carried out using a Nikon V-12 profile projector at a magnification of 10X.  A minimum of 8 test values were obtained per group.  The Adhesive Remnant Index (ARI)[xxxiii] was used to determine the resin or cement remnant on the tooth and bracket after separation.  ARI is one method to determine bond failure location and fracture mode (adhesive, cohesive, etc.). The ARI scores and bracket failure area were expressed as a percentage of the total number of teeth tested using the criteria defined in table 3..

Statistical Analysis

The mean SAS and standard deviation (SD) were obtained from the raw data.  Significant differences in mean SAS were evaluated using the t-Test and One-Way ANOVA at the confidence level of 95% (p = 0.05).  Additionally, a two parameter Weibull analysis was used to estimate the failure probability of a specimen under a given load and estimate the reliability of the test data^{29 [xxxiv]}.  The procedure involves the evaluation of characteristic strength, s₀, and Weibull modulus, m, in the Weibull equation³⁴

p_i = 1- exp{-(s_i -s_u)/s₀)^m}                                 (1)

where p_i is the probability of failure, at a given stress, s_i, s_u, is the lowest level at which p_i approaches a value of 0, and s_o, is the characteristic strength.  The raw data is ranked in ascending order and p_i is calculated using the expression;

p_i = n_i/N+1,                                                                              (2)

 where ‘n’ is the sample rank, and N the total number of samples in the population forming the group.  The constants s₀ and 'm' are obtained from linear regression of the data to the Weibull distribution function using a computer. The regression coefficient R², is a measure of the fit of the data to exhibit a 'Weibull distribution function (Equation. 1).  The strength level corresponding to a failure probability of 1%(survival probability of 99%) and 90% (survival probability of 10%) was calculated from ‘s₀‘and ‘m’.

Results

Table 4 lists the mean value, standard deviation (SD), and range of SAS as a function of enamel condition and adhesive.  PTX and PDL groups showed higher SAS than other groups.  The lowest value of SAS was obtained for the unconditioned enamel groups, NTX and NFG.

Table 5 shows the result of statistical comparison between group means.  There was no significant difference(p<0.05) in mean SAS values between either PTX and PDL group or between STX and SDL groups.  Mean SAS values were significantly higher(p<0.05) for the AFG group in comparison to the NFG group.

Table 6 shows the results of Weibull regression analysis.  The characteristic strength, s₀, the Weibull modulus, ‘m’, the regression coefficient, R², and strength at a failure probability of 1% and 90% are shown in the table.  The low values of ‘m’ for the NTX group indicate the widest spread and least reliability.  The ‘m’ values for the other groups are in the expected range for cements.

Table 7 lists ARI values, which rates the bond failure as a percentage of remnant adhesive on enamel.  STX and SDL groups with less than 10% resin remnant on enamel shows a predominantly adhesive separation at the enamel resin interface.  PTX and PDL group show bracket adhesion failure at enamel/adhesive or cement  interface and cohesive in resin site.  Enamel fracture was observed for only 2 out of 77 tests (2 samples in PDL group).

Discussion

Enamel fracture was noted in two instances (only in sample group PDL) with SAS greater than 13.5 MPa.  However, additional samples with SAS values greater than 13.5 MPa did not show enamel fracture.  SAS may not necessarily reflect or be used as a monitor/precondition for the possibility of enamel fracture.

The resin adhesive TX (single cure, VLC), and cement FG (triple cure) are commonly used for orthodontic purposes.  The resin adhesive cement DL(a dual cure adhesive cement resin), is often used for bonding-luting composite restorations and indirect prosthesis (inlays, inlays, veneers and crown & bridge)in conjunction with enamel etch conditioning using 37% o-H₃PO₄, on enamel/dentin (30/15 seconds etch duration).  In the present study a 15 second conditioning was deemed ideal for enamel based on the practice in orthodontics.

There was no significant difference in mean SAS values(p.>0.05) between these two groups PTX and PDL, or SDL and STX, at constant enamel conditioning.  The mean SAS for the H₃PO₄ groups, was significantly higher(p<0.05) than that of Salicylic-Lactic acid groups, at constant resin composition (PDL > SDL; PTX > STX).  The lower SAS and the interfacial separation (resin-enamel adhesive failure) associated with the Salicylic-Lactic acid conditioner, indicates an advantage particularly during debonding of brackets.  This may be related to physical factors relating to extent of penetration of the bonding resin/or the adhesive/cement in the interstices of the etched enamel (physical interlocking).  SEM photographs of the conditioned surfaces show a higher surface roughness associated with the Phosphoric acid group in comparison to other groups.

The significantly lower strength(p<0.05), obtained for the unattached condition NFG, indicates that physical penetration of the GIC is greater when the enamel is etched, AFG; an increase in strength from 5.0 to 8.2 MPa.  Unlike single or dual cure adhesive cements, the resin based GIC is identified to be triple cure; an acid-base reaction  in addition to the chemical and photo cure polymerization reactions.  The absence of any significant difference between the SAS values obtained for a) PTX and PDL b) STX and SDL, and c) STX, SDL and AFG groups, table 5, suggest that multiple cure mechanisms in the luting cement do not have a major effect on SAS.  This may indicate that the conditioning method and the extent of penetration determine SAS.  This is supported by the SAS for TX groups.  In the absence of any conditioning, NTX, the mean SAS values drop to a low of 1.9 MPa.

Weibull analysis

The interface region formed between a bracket and enamel is not a homogeneous; and is composed of etched interior region, insoluble reaction products from conditioning, bonding resin, adhesive resin/cement, bracket mesh, etc.  Further, the incorporation of porosities (enamel not completely wetted by bonding agent), porosities from setting reaction  and presence of insoluble reaction product have an effect on the SAS.  The large standard deviation (table 4.) observed in shear testing of these interfaces indicate the uncertainty in an expected strength value under In-Vivo conditions.  At constant experimental conditions, the spread in data is associated with technique and strength limiting factors which create a discontinuity in the interface structure.

The Weibull analysis results indicate that for the major part, the SAS data fits the Weibull distribution function at a level of correlation greater than 0.9.  The value of ‘m’ for the NTX was found to be less than 1 indicating a non-reliable condition with a wide scatter in data.  This group was evaluated as a negative control for the resin based materials.  For all other groups the ‘m’ value (range 2.9-5.2.

The probability of exceeding an SAS level of 3.0 MPa is at least 99% for all the test groups except NTX and NFG.  At the higher end, the value of the stress above which a 10% failure is expected, s_(90%), is below the stress level at which enamel failure is expected.  This value can be treated as the attainable value for each group and is below the value of 14.0 MPa for all groups except for PDL.  The range of values obtained for the organic acid conditioners indicate the possibility of their application in bracket bonding.

A value of 13.5 MPa ⁴ may be considered an upper limit beyond which enamel failure or fracture is a possibility.  An SAS value less  than 13.5 MPa, however, may be used as a limit requirement to prevent enamel failure during removal of brackets, since enamel fracture was noted only when the SAS exceeded this value.  This is not a necessary condition, since enamel fracture was not observed when the value of SAS exceeded 13.5 MPa for other samples in PDL and PTX groups.  In addition, the effect of prior enamel condition and orientation of loads on the interface region may also have an effect on enamel fracture, under In-Vitro and In-Vivo conditions..

Adhesive Remnant Index

For PTX, PDL and AFG groups, the ARI indicates enamel/adhesive and some cohesive failure in the adhesive on the fracture plane.  The enamel-adhesive interface is, in general, considered the most desirable location for bond failure provided the level of penetration is small.  For a larger depth of interpenetration, the risk for enamel damage during debonding may be higher and resin remnant removal from the enamel surface may lead to additional enamel loss due to grinding and polishing steps.

For all other groups, shear separation occurs at the interfacial region between enamel and adhesive with a major portion of the adhesive on the bracket.  This may be attributed to the mechanism of conditioning and the lower level of interpenetration.  Visual observation of tooth surfaces revealed a clean surface with minimal resin remnant when the organic conditioner was used.  Further, the appearance of the enamel surface after Salicylic-Lactic acid conditioning show a smoother surface than that for Phosphoric acid.  The feasibility of an organic acid enamel conditioner for direct bracket bonding needs further clinical trials.

Based on the above, a value of SAS between 3 and 14 MPa may be ideal for orthodontic adhesive/luting purposes, that may provide the mechanical stability during the treatment period and  finally during debonding, with minimal enamel loss and physical alterations.

Conclusions

1.      There was no significant difference shear adhesive strength between Transbond XL and Diamond Link.(p>0.05) at constant enamel conditioning.

2.      Enamel conditioned by 37% o-H₃PO₄ showed significantly higher SAS(p<0.05) compared to those conditioned by Salicylic-Lactic acid, irrespective of the adhesive or bonding resin.

3.      Enamel surface morphology after conditioning was different for inorganic acid compared to organic acid.  Phosphoric acid showed a typical etch pattern, while Salicylic-Lactic acid showed a smoother surface.

4.      The SAS values were significantly lower(p<0.05) for unetched enamel surfaces bonded with GIC than any other condition.

5.      The SAS value obtained for the organic acid conditioner and resin based adhesive/cement combination were not significantly different than the Polyacrylic acid conditioner and GIC(p>0.05)

6.      The ARI and failure mode for organic acid conditioners indicate an adhesive separation at the enamel resin interface with resin remnants predominantly on the bracket at separation..

7.      The use of organic acid for conditioning enamel is a viable methodology for both of the resin adhesives.  Clinical trials are required to evaluate In-Vivo implications.

Table 1. List of materials

Table 2. Test Groups - Conditioner / adhesive or cement and time.

Table 3. Definition of Adhesive/cement remnant (index) , ARI.

Table 4. Mean, SD and range  of bracket SAS values as a function of test group

Table 5. Results of Statistical analysis

x- Significant difference between mean value pairs(p<0.05, Scheffe test)

Table 6. Weibull analysis results of regression fit of raw data using Eq.1

s₀ - Characteristic strength, m - Weibull modulus, SE - Standard error, s_(10%), s_(90%) - The stress at which the failure probability is 10% and 90%(Survival probability of 90% and 10%), 

Table 7. Adhesive remnant index, ARI, expressed

as a percentage of total number of teeth in each range.