Dental Materials

Interfacial bond strengths of amalgam bondedto amalgam and resin composite bonded to amalgamTerry J. Fruits, DDS, MEd*/Manville G. Duncanson, Jr, DDS,Thomas L. Coury, DDS,

Objective: The effects of time, surface preparation, and use of a bonding agent on ¡he ability to bond amalgamor resin composite to amalgam were studied.Method and materials: Cylindrical amalgam specimens were prepared in a split die 132 givups of 10 each). The32 groups were divided into Two groups of ¡6 each: amalgam bonded to amaigam and resin composite bonded toamalgam. One half of each group of ¡6 was bonded within ! hour after the initial condensation (fresh) and iheother half was bonded after 21 days'storage in physiologic saline ¡old). Four honding methods (control ¡nobonding agent]. Advance. Alt-Bond 2. or Amalgambond Plus) were each utilized on one abraded group (air abra-sion with 50-/im aluminum oxide) and one unabraded group. After an additional 21 days of storage, bondstrengths were measured with a three-point flexure test. Solid amalgam cylinders were also tested.Results: Solid amalgam control cylinders exhibited mean bond strengths 10.000 psi greater than any of thebonded specimens. Among specimens of amalgam bonded to amalgam, those that were abraded after 21 daysand used no bonding iigent showed a significantly greater bond strength than did specimens utilizing any ofthe other combination.^ of the .^elected independent variables. Among specimens of resin composite bonded loamalgam, those bonded after 21 days and utilizing a bonding agent exhibited significantly greater bondstrengths than did specimens utilizing any of the other combinations of the selected independent variables.Conclusions; The results of this in vitro study suggest that the strength of a repaired amalgam is only about 40%ofthat of an unrepaired amalgam. They also indicate that the strongest repair of amalgam using additional amal-gam material may be accomplished without the use of any bonding agents utilized in this study. When adding resincomposite material to amalgam, the resin composite should be added after the amalgam has had adequate time toset, and the use of a bonding agent increases the strength of the repair. (Quintesseti ce Int 1998;29;327-334)

Key words: atnalgatn. amalgam bonding, amalgam repair, inteifacial bond strength, resin composite

Clinical relevance

The success of a repair of an existing amalgamrestoration is highly dependent on the establish-ment of an effective bond between the new and theexisting material. The results of this study offer in-formation that may help in the practitioner's selec-tion of techniques and materials for amalgam repairprocedures.

' Assistant Prufessor. DepartmcnL of Operative Dentistry, University ofOklahoma, College of Denlistry, Oklahoma City, Oklatioma.

" Professor anii Chair, Departmcnl of Denial Materials, University ofOklahoma, Cuilege of Denlislry, Oklahoma Cily. Oklahoma.* " Associate Professor and Chair, Department ni Operative Dentistry,* " Associate f roicasor and Chair, Department ni Operative Denti»tjniversily of Oklahoma, College of Dcntibtry. Oklahoma City, Ok]LU.>,..,,..Reprinl requests; Dr Terry J. Fruits, Assistant Professor, Deparlmentof Operative Denlislry, University of Oklahoma, College of Dentistry,PO Box 26901, Oklahoma City, Oklahoma 73190. E-mait: lerry-i>..;icß>in->khse.edu

The succes.sful additioti of new restorative materialsto an existing restoration may be the most con-

servative course of treatment in certain situations.Occasionally, the repair of an amalgam restoration canprovide a less invasive procedure than its complete re-moval and replacement. The addition of a resin com-posite veneer over an existing amalgam for eiitheticreasons may be more conservative than placing a looth-colored inlay. These types of repair and esthetic proce-dures may be selected in cases when it is desirable toreduce further trauma to the tooth or to avoid addi-tional stress on the patient.

The success of such procedures tnay be directly af-fected by the strength of the bond between the old andnew restorative materials.'-'" The purpose of this invitro investigation was to measure the inierfacial bondstrengths of both resiri composite and amalgam bondedto set amalgam with different adhesive systems underselected conditions. The testable null hypothesis forthis study was that there is no statistically significantdifference in the measured interfacial bond strengths of

Quintessence International 327

bonded materialAmalgam or composite

,| Adhesive agent |

AduanoeAll-Bond 2

•Amalgambond Plus-No bond agent

Surfaoe treatment

Abraded (air abrasion)or

Unabraded (no treatment)

Bonded 21 days after condensation(Old)

onded within 1 hr of oondensation(Fresh)

J^

Fig 1 Variabies in specitnen preparation.

specimens prepared by bonding amalgam or resincomposite to existing amalgam under selected bondingconditions.

Method and materials

This investigation measured the interfacial bondstrengths between fresh resin composite and freshamalgam with set amalgam as a function of the follow-itig variables:

1. Two materials (bonded to an initial increment oí' setamalgam)• Dispersed-phase amalgam alloy [Dispersalloy,

Caulk)• Hybrid resin cotnposite (TPH, Caulk)

2. Four bonding systems (adhesive agents)• Advance (Caulk)• Amalgambond Plus (Parkell)•All-Bond 2 (Bisco)• No bond agent

3. Two surface treatments (treatment of tbe initial incre-ment of set amalgam)• Abraded—air abrasion with 5()-pm aluminum oxide• Unabraded—no surface treatment

4. Two time frames (age of the initial increment of setamalgam)

•Fresb anialgatit—additional material bondedwithin 1 hour of condensation of the initial incre-ment of amalgatn

• Old amalgam—additional material bonded 21 daysafter condensation of the initial increment ofamalgam

Figure 1 offers a diagrammatic description of thevariables involved in the specimen preparation.

Specimen preparctiion

The study required the preparation of two basic groupsof specimens. One group of 160 specimens was used toevaluate the bond strength that occurred alter amalgamwas placed on existing amalgam that had set for 1 hour.The second group of 160 specimens was utilized toevaluate the bond strengtb tbat. occurred after resincomposite was placed on existing amalgam tbat had setfor 21 days. The following is a description of the prepa-ration of each of these two basic groups of specimens.

Amalgam placed on existing amalgam. One bun-dred sixty cylinders of amalgam were fabricated in asplit two-piece die measuring 4.75 mm in diameter and19.00 mm in length. Acrylic die spacers that were 4.75mm in diameter and 9.50 mm long were inserted intothe die prior to the condensation of amalgam (Dispers-

328 Voiume 29. Number 5,

Adaea wth n 1 Urof condensât on j

J

Added to

Amalgam added lo amalgam180 total specimens

fresh amaigam

Q .•am

fD

- L -rooCL

O"rad

10

rJ—,

cabr

Q .

Q .

10

mmrad

<DQ.

DJCT

DJaa

10 10

1

3algi

Ml

3CT

a.

Q

CE

TO

c:

O "

Q .

Q .

10

LzoO"

oQ.

1

V ^ A d d e d 21 days after^ c o ^ d e n s a t o n

Sol

d

:onlro

J

Added toold amaigam

11 1 1

ajan

m

Ali-

enocM

3

lambc

a

là•rad

Q .

10

c

çQ.

Q.

±^

i_

10 10

C

c

C

C

ê|[^21

°-C

C

C

Í

10 10 10 10

1g.

10

_Zo•"ond ag

[3

b.

iO

.

c

3 -

b.

10

Fig 2 Preparation of specimens consisting ot amaigam added to amalgam.

alloy) to allow the fahrication of amalgam cylindersmeasuring 9.50 mm in iength. These initial 160 cylin-ders were designated as the prebond population.

In addition. 20 solid amalgam cylinders were fahri-cated in the same die without the placement of theacrylic spacers. These solid amalgam cylinders, whichserved as control specimens, measured 4.75 mm in di-ameter and 19.00 mm in length.

The 160 prehond cylinders were then separated intotwo groups of 80. After the amalgam cylinders had setfor 1 hour, one group (fresh) of SO was further dividedinto four groups of 20 cylinders. Each of these fourgroups of 20 cylinders was subjected to a differenthonding technique prior to the condensation of the addi-tional amalgam (Fig 2), These four honding groups con-sisted of the following four techniques; (]) Advance ad-hesive; {2} All-Bond 2 adhesive; (3) Amalgambond Plusadhesive: and {4) no bonding agent (see Fig 1 ).

The 20 cylinders in each of these four different bond-ing groups were divided into two groups of 10. The 10cylinders in one of these groups were abraded on oneend prior to the condensation of additional atnalgam.This was accomplished with a Model er Microetcher(Danville Engineering) and a 50-(jm aluminum oxideabrasive powder. The prehond cylinder was removedfrom the die and one end of the cylinder was abraded.The specimen was rinsed with an air-water syringe,

dried, and placed back into the die. The 10 cylinders inthe other group were not abraded.

The manufacturer's instructions for the various desig-nated honding techniques were followed, and the appro-priate agent was applied to the end of the cylinder thatwas to receive the additional amalgam material. Amal-gam was then mixed and condensed against the end of theprebond cylinder to create a finished bonded specimenthat was 4.75 mm in diameter and 19.00 mm in length.

After the new amalgam had set for 24 hours, thespecimens were carefully removed from the die andstored in an intercellular fluid substitute solution(Normosol-R. Abbott Laboratories) at 37°C for 21 daysprior to testing. The 20 unbonded control specimensalso remained stored in Normosol-R for the same21-day period.

The remaining 80 (old) prebond cylinders were re-moved from the dies and stored in an intercellular fluidsuhstitute at 37°C for 21 days. After storage for thistime period, these old cylinders were suhdivided intofour groups to receive additional amalgam in an identi-cal fashion as the previously described groups of freshspecimens.

Resin composite placed on existing amalgam. Onehundred sixty cylinders of amalgam were fabricated asdescribed in tbe previous section. Tbese initial 160 speci-mens were designated as the prebond population. No

Quintessence International 329

Added w

Res

til n Ihrol condensât on ^ ^

1Q .

1 ^

I t

F

rarl

Q,

J

1

10 10

Added toresh amalgam

11 l _IA

ÍI-

cooa.

a 3ID UQ. 3.

C

3

to"OJ

oa

•

rad T

10 10 10 10

n composite added to amalgam160 total specimens

^ ^ ^ V Added 21 days after^ V condensât on

1oCTOQ .

IO

m

Trad

•0

a

Added tooid amaigam

11 1

CL

(D

F>i

hc13

•-

Q .

a

[71̂

enoO-

Tc

abr

01Q ,

mQ.

3Si.

CO

CT•on

Q.

^ ^

X

rad 01

3a.(DQ.

OD"^O.

— ^

F

rad

a.

G

5

10 10 10 10 10 10 10 10 10 10

Fig 3 Preparation ol specimens consisting of resin composite added to amalgam.

Fig 4 Testing apparatus tor tiie tiiree-point flexure test

solid control specimens were made because the datafrom the existing solid control specimens were used(Fig 3),

The prebond cylinders were divided into the samebonding groups as in the aforementioned protocol. Fol-lowing the appropriate surface treatments (air abrasionor no air abrasion) and the application of any designatedbond agent, a hybrid resin composite (TPH) was in-

jected onto the surface of the prebond specimen untilthe die was filled with the re,sin composite. The resincomposite was injected in layers of no more than 2.0mm, and each layer was cured for 60 seconds with avisible ligbt-curing unit (EFOS), The specimens werestored prior to testing as described previously.

Following storage, all specimens were subjected to trans-verse loading witb a three-point flexure te.st to determinethe maximum load required for fracture. Specimens werecentered on the testing apparatus of an Instron UniversalTesting Instrument (Model No, 4468] and subjected toincreasing loads at the bonded interface (crosshead speedof 1,25 mm/min) until fracture occurred (Fig 4). For thesolid control specimens, transverse loading was appliedat the specimen midpoint.

Peak load values at fracture were obtained fromgrapbic and digital readouts. Bond strengths were ex-pressed as maximum tlexural load per cross-sectionalarea of the bonded interface (MPa), Statistical analysisof the data involved analysis of variance and Duncan'sNew Multiple Range Test {P = ,05),"'^ Following frac-ture of the specimen, the surfaces exposed by the frac-ture were observed under x60 magnification.

330 Voiume 29, Numbers, 199g

Fruits et al

TABLE 1 Mean (SD) bond strengths of amalgam added to amalgatn (MPa)

Adhesive agent

Fresh amalgam(bonded within 1 hour)

Old amalgam(bonded after 21 days]

Abraded LJnabraded Abraded Un abraded

Mvanœ 29,2 (n,7¡ 30,8 (8.5) 33.5 (1t.2) 27.6 (10.1)'^I'-Bo'^'í2 32,9 (8,6) 26,1 (101) 25,0 (9.0] 30,5 (8,2]Amalgambond Plus 21,4(10.1) 30,6 (8,8) 28,8 (6.3] 37.4 (17 5]

' ^ " ^ g ^ " ' 23-3 (6.8) 45,3(20,5) 51,7(48,0) 19,3(116)

Solid amalgam control specimens had a mean bond strength of 121,7 (44.5) MPa.

TABLE 2 Mean (SD) bond strengths of resin composite added to amalqam(MPa)

Adhesive agent

Fresh amalgam(bonded within 1 hour)

Old amalgam(bonded after 21 days)

Abraded Unabraded Abraded LJnabraded

Advance 5.3 (5,4] 7,0 (3.8] 49.8 (23.3) 43.1 (13,4]All-Bond2 9,7 (7,0) 14.6 (13.2] 35.2 (6 7] 37.6 (14.3]Amalgambond Plus 5.4 (4.3) 5.5 (5,0) 35.5 (15.8] 26.1 (13.0)Noagent 13.0 (0.9) 0.0 (0.0) 16.6 (10.0) 20,9 (10.7)

Solid amalgarn control specimens had a mean bond strength o( 121.7(44.5] MPa

Results

The meati bond stretigth of the solid amalgatn controlspecimens was 121,7 (44.5) MPa. Table 1 presents themean bond strengths and standard deviations of thespecimens that involved the honding of atnalgam toamalgam. Table 2 presents the mean bond strengths andstandard deviations of the specimens that involved thebonding of resin cotnposite to amalgam.

Figures 5 and 6 present a graphic comparison of thedata. The two techniques that produced the greatestbond strengths when amalgam was added to existingamalgam did no! utilize a bonding agent. When com-posite was added to existing amalgam, however, the useof a honding agent seemed to increase the bondstrength. The addition of composite material within Ihour of amalgam condensation did not produce accept-able bond strengths.

Discussion

A successful technique for the repair of existing amal-gam restorations would certainly be advantageous tothe elinician interested in conservative dental treat-ment. Past studies have indicated that a repaired amal-gam exhibits a strength that ranges from 12% to 79%of that of nonrepaired amalgam,-' *•"-'' Much of the pastresearch seems to indicate that the strength of a re-paired amalgam may fall tnto a range of only 50% orless of that of a nonrepaired amalgam. This reductionin .strength in the repaired amalgam is one of the mainreasons that amalgam repairs are not freqtiently at-tempted by clinicians.

This study utilized some new materials and methodsto reinvestigate this problem. The results of this studycorroborate findings in previous studies cf the relativebond strengths of bonded (repaired) amalgam and non-

Ouintessence International 331

Old amalgamadditional amalgam placed after 21 days)

^ Old amalgam | UnabraûedFresh amalgamadditional ama/gam placed wittiin 1 hr¡S Abraded nUnabraded

ADVAB2ADVAM+AB2ADVAM+ADVAB2AB? NA AM+MA

Fig 5 tvlean bond strengths ol amalgam added lo amaigam. (AB2) All-Bond 2; (ADV)Advance: {AM+) Amalgambond Plus: (NA) no adhesive agent.

Old amalgam

(resin composite placed alter 21 days)

^ Abraded HunabradedFresh amalgam(resin composite placed within 1 hr)

^ Abraded QUnabraded

ADV ADW AB2 AM+ AB2 AM+ WA NA AB2 NA AB2 ADV AM+ AM+ ADV MA

Fig 6 Mean bond strengths of resin composite added to amalgam. {AB2) All-Bond 2, (ADV)Advance; (AM+) Amalgambond Plus; (NA) no adhesive agent.

bonded amalgam. The mean bond strength of the speci-mens in the most successful of the experimental bondedgroups of this study exhibited only 42% of the meanbond strength of the solid (nonbonded) specimens. Inaddition, all of the specimens appeared to fail along thebond interface.

Specimens of amalgam placed on existing amalgam

In general, there was little difference between the meanbond strengths of the groups that had additional amal-gam added within 1 hour of condensation and tho.se that

had additional amalgam added 21 days after conden-sation. The exceptions to this were found in the groups ofspecimens that utilized no adhesive bonding agent. Thegroup of specimens that was abraded and had additionalamalgam added after 21 days without the use of a hond-ing agent exhibited a mean bond strength that was muchgreater than that of the group of specimens that wasabraded and had additional amalgam added within 1 hourwithout the use of a bonding agent. On the other hand,the group of specimens that had additional amalgamadded within 1 hour of condensation without air abra-sion or a bonding agent had a much higher mean hond

332 Volume 29. Number 5, 1993

strength than did the group that had additional amalgamadded after 21 days witbout air abrasion or a bondingagent.

One explanation for tbese results may be ibat thealuminum oxide abrasive particles may have embed-ded in the softer fresb amalgatn. The embedded parti-cles may tben have interfered with the ability of thefresh amalgam to adhere to the existing amalgam. Theharder surface of the amalgam that had set for 21 dayswould be less susceptible to embedding of the particlesthus leaving a roughened, uncontaminated amalgamsurface.

The use of an adhesive agent did not significantly in-crease the mean bond strengths of any of the specimenswith the exception of the groups that were unabradedand had additionai amalgam added after 21 days. Thespecimens tbat were abraded and tben had additionalamalgam added without any adbesive bonding agent 21days after condensation exhibited the highest bondstrength, and tbe specimens that simply had addifionalamalgam added within 1 hour of condensation withoutthe use of an adhesive bonding agent or air abrasionsbowed tbe next bigbest mean bond strengtb. Tbe statis-tical analysis revealed tbat both of these groups weresignificantly different from the other groups of amal-gam-amalgam specimens. Among the 12 groups that didutilize an adhesive bonding agent, tbere was no signifi-cant difference in mean bond strengths.

In general, tbe use of air abrasion to prepare tbe sur-face of tbe amalgam prior to tbe addition of new amal-gam did not seem to increase tbe bond strengths exceptin the groups that did not utilize an adhesive bondingagent.

Specimens of resin composite bonded ta cinialgam

There was a considerable difference in the mean bondstrength of the groups in which resin composite wasadded to amalgam within 1 hour of the condensation ofthe amalgam and those in which the resin compositewas bonded to amalgam 21 days after condensation oTthe amalgam. All of the specimens in which resin com-posite was added within I hour of the condensationof the amalgam had mean bond strengths tbat were sta-tistically significantly lower than tho.se of tbe speci-mens in whicb the resin composite was added to amal-gam 21 days after condensation. Tbe bond strengths ofall of the specimens that had resin composite addedwithin 1 hour of condensation were extremely v/eakand failed readily.

Air abrasion of the existing amalgam surface did notseem to have a great effect on the bond strengths of anyof the specimens in which resin composite was added

to amalgam. There was no significant diTterence be-tween tbe abraded and nonabraded groups in wbicb allother variables were the same.

The use of a bonding agent had a significant effecton the bond strength of the specimens in which resincomposite was added to the amalgam after 21 days. Oftbese, the groups of specimens that were abraded andutilized an adhesive bonding agent had significantlyhigher mean bond strengths rban did the two groups inwhich an adhesive agent was not used.

Of the specimens in which resin composite wasadded with an adhesive bonding agent after 21 days,only the specimens prepared with the re sin-modifiedglass-ionomer adbesive were found to be significantlystronger. The manufacturers of this material do not listthis as one oftbe material's uses; bowever, it seemed tobe fairly effective as an adbesive agent for bondingresin composite to amalgam.

Summary

Tbis in vitro study investigated tbe effects of certainvariables, including time, surface treatment, and tbeuse of adhesive agents, on the mean bond strength ofspecimens prepared by adding amalgam or resin com-posite materia! to an existing amalgam cylinder. Tbebonding techniques utilized in this study producedmean bond strengths that were less than 50% of that ofsolid amalgam control specimens. For specimens inwhich additional amalgam was added to tbe existingamalgam cyhnder, the existing amalgam cylinders thatwere abraded after having set for 21 days and bad nobonding agent applied showed a significantly strongerbond. For specimens in wbich resin composite wasadded to tbe existing amalgam cylinder, the existingamalgam cylinders that bad set for 21 days and hadbonding agent applied prior to tbe placement of tberesin composite exhibited significantly stronger bonds.Tbe specimens in which the existing amalgam cylinderhad set only 1 hour prior to the addition of resin com-posite exhibited very weak bond strengtbs.

Acknowledgments

This research wus supported by grant No. At)OO5769 from theuniversity of Oklahoma, College of Dentistry.

References

1. Barütieri LN. Monteiro Jtinior S, de Andrada MA. Amalgam re-pair: A aise report. Quintessence Int I992;23:527-53I.

2. Carr-Hosie MA, Miranda FJ, Collard EW. Durcanson MG. Theeffect of Amatgambond on the flesural bond strength of dentalamalgam. Okia Dent Assoc J 1992;82:20-24,

Quintessence International 333

3. Erkes EO. Burgess JO, Hombeck DD. Amiilgani repair: An invitro evíiliíalion of bond integrity. Gen Denl I990;38:203-205.

4. Gouriey JM, Ambrose ER. Veneering Limalgam restorations.J Can Denl Asscc I983;48:49-5O.

5. Hadavi F, Hey JH, Czech D, Ambrose ER. Tensile bund slrenglhof repaired amalgam. J Proslhel Denl 1992:67:313-317.

6. Hibler JA, Foor JL, Miranda FJ, Diincanscn MG Jr. Bondslrenglh coniparisoiis of repaired dental amalgams. QiiinlessenceInt l9SS;iy:411^l5.

7. Lacy AM, Rupprecht R, Walanabc L. Use of self-curing compos-ite resin to lacilitule amalgam repair. Quintessence Inl 1992:23:53-59.

8. Nelson JA, The resin composite-amalgam windoiv preparation.Praci Peiiodonl Aeslhel Denl 1992:4:21-26.

9. Plasmans PJ, Reukers EA. Esthetic veneering of amalgam restor-ations with resin composite—combinmg the best of both worlds?Oper Dent I993;18:66-71.

10. Roda RS, Zwicker PF. Ttie comhined resin composite and amal-gam restoration for posterior leetli: A clinical report. Quin-tessence Int 1992:23:9-13.

11. Li Jerume CR. StalisLcal Inference, vol. I, Ann Arhor, Ml:Edwards Brothers, 1969:270-237.

12. SAS/STAT User's Guide, vol. I. The ANOVA Procedure. Cary,NC: SAS Institute, 1989:209-244.

13. Terkla LG. Mahler DB. Mitchem JC. Bond strength of repairedamalgam. J Proslhet Dent 1961;] 1:942-944.

t4. Scolt GL, Grisius RJ Bond strength at the interface of new andold spherical amalgam. tJS Navy Med Newsletter 1969;54:34.

15. Consani S. Ruhnke LA, Stolf WL. Infiltration of a radioactivesolution into joined silver amaigam. J Priisthet Dent 1977;37:158-163.

16. Berge M. Flexural strength ol'joined and inlact amalgam. ActaOdontol Seand 1982;4O:313-3I7,

17. Walker AC Jr, Reese SB. Bond strength of amalgam to amalgamin a high copper amalgam. Opcr Dent 1983:8:99-102.

18. Hadavi F, Hey JH.Amhro.se FR. Shear bond .strength of compos-ite resin to amalgam: An experijtient in vitro using différentbonding systems. Oper Dent 1991; 15:2-5.

19. Brown K, et al. Flexural strength of repaired high copper amal-gam. Oper Dent 1986;! 1:131-133.

Answers to Ql 2/98 Questions

1. A2. A3. A4. D

5. D6, B7, A8. A

9,10.11.12.

BECA and E

13.14.15,16.

BB

CA and E

334 Voiume 29, Number 5, iggg