J. BIOMED. MATER. RES.
VOL. 10, PP. 789-804 (1976)
In vitro Corrosion of Dental Amalgam Phases* M. MAREK and R. F. HOCHMAN, Metallurgy Program, School of Chemical Engineering, Georgia Institute of Technology, Atlanta, Georgia, and T. OKABE, Department of Restorative Dentistry, School of Dentistry, Medical College of Georgia, Augusta, Georgia
Summary The effects of corrosion on the major phases of dental amalgam microstructure have been studied in vitro on samples prepared by electroplating mercury on AgaSn. Tests were made in synthetic saliva, and samples were examined in a scanning electron microscope before and after the exposure. Simple immersion had little effect on the phases. Contact of amalgam with gold increased the probability of deterioration of the y2 phase. By anodic polarization beyond the breakdown potential, the yz phase was selectively dissolved. Under conditions simulating a crevice, the 7 2 phase was dissolved and the y1 phase suffered morphological changes.
INTRODUCTION The major phases of a dental amalgam microstructure-y (Ag-Sn), different electrochemical properties and corrosion resistance. Although results of several investigations of the corrosion behavior of the individual phases have been reported, there is a lack of reports of direct examination of the interrelated corrosion effects. Observation of details of the corrosion-related changes in actual amalgam samples is difficult; the fine grain size requires relatively high magnification, the structural components have irregular and complicated shapes, the large variations in hardness and the presence of mercury make preparation of high-quality surfaces difficult, and structural changes often caused by preparation procedures can affect the electrochemical behavior. y1 (Ag-Hg-Sn), and y2 (Sn-Hg)-exhibit
*This paper is based in part on a presentation given to the 1975 General Session of the American Association for Dental Research in New York, N. Y. 789 @ 1976 by John Wiley & Sons, Inc.
790
MAREK, HOCHMAN, AND OKABE
I n this work, the effect of some of the variables of exposure on the three major phases of amalgam structure was investigated using direct observation of individual crystallites of the y1 and y z phases on a substrate of Ag3Sn (y phase).
MATERIALS AND METHODS The specimens were prepared by electroplating mercury on polished samples of Ag3Sn;the method had been developed for kinetic studies of amalgamation and the details are described elsewhere.' Phases 71 and y2, in polyhedral and needle form, respectively, precipitate separately on the substrate; the 01 phase can also form if the amount of mercury is relatively low. The identity of the phase is usually clear from the appearance ; however, energy-dispersive x-ray analysis was used in all experiments described here to verify the composition. The samples were examined in a scanning electron microscope (Cambridge Stereoscan Mark 11). The image of a selected area was photographed, the samples were removed and subjected t o various corrosion exposures under controlled conditions, washed in distilled water, dried, and examined again in the microscope. All the corrosion experiments were performed under potentiostatically controlled conditions for the following reasons. 1) The relative amounts of the phases prepared by plating Ag3Sn with mercury are different from their proportions in dental amalgam. To permit a n examination of the individual crystals, their number per unit area of the substrate must be limited; this results in samples with a large surface fraction of the noble Ag3Sn phase, and untypically noble free-corrosion potentials. 2) Some contamination of the surfaces by carbon deposition occurs during the examination in the electron microscope. This would also affect the free corrosion potential, making it more noble. By controlling the potential with a potentiostat, the influence of the carbon deposit is, t o a large extent, eliminated. Although this assumption could not be verified directly, i t was observed that identically prepared specimens, subjected t o same corrosion conditions but not examined in the electron microscope prior to the exposure, exhibited same general features as those described in this report. Corrosion potential of dental amalgam changes with time of exposure. To avoid an arbitrary choice of potentials for the potentio-
I N V I T R O CORROSION OF DENTAL AMALGAM PHASES
791
static control, some of the experiments were made with a potentiostat controlled by the potential of a sample of actual dental amalgam or amalgam-gold couple exposed in a separate cell. The experimental setup is illustrated in Figure 1. The conditions of the exposures are summarized in Table I. In addition, an anodic polarization curve of dental amalgam in synthetic saliva, deaerated with a nitrogen-10% carbon dioxide mixture, was recorded for correlation with the observed corrosion effects. The experimental procedure has been described elsewhere.2
RESULTS Exposure.under the Conditionof Free Corrosion of Dental Amalgam in Synthetic Saliva The corrosion potential of a sample of dental amalgam in aerated synthetic saliva, used to control the potentiostat, changed with time to more noble values, as indicated in a polarization diagram in Figure 2. Scanning electron micrographs of a sample before and after the exposure (100 hr) are shown in Figure 3. No deterioration of any of the phases was observed. After longer exposures, some pitting of the yz phase was occasionally observed, as illustrated in Figure 4.
POTENTIOSTAT
ISOLATION
ELECTROMETER
AMPLIFIER
CONTROLLED SAMPLE
CONTROLLING SAMPLE
Fig. 1. Experimental potentiostatic setup with potential control following the corrosion potential of another sample.
Y1Y1,YZ
Y,YI,Y?
YlYl
Y,YI,Y2
Y,YI
c1
C2
D1
D2
2
2
1 1
1 1
I 1
15 15
1.3 16.3
- .I95
100 100
- .30 - .30
- .78 to 0.0 - .78 to 0.0
- .190
- .57 t o - .29 - .57 t o - .20
Potential (V, SCE)
100 500
Test Exposure solution- period (hr)
~
HC1 t o p H 1.0,containing 2.6 M NaCl, a t 25°C. New True Dentalloy (S. S. White Co.). 24 carat.
9 10
7 8
5 6
4
3
Figureno. of the micrographs
2) Synthetic saliva acidified with
~~
Following the free corrosion potential of a commercial dental amalgamb in synthetic ~ a l i v a , 3 aerated ,~ with air-lO% COz, p H 6.6, at 25°C. Following the corrosion potential of a goldc-amalgamb couple in synthetic ~ a l i v a , ~aerated .~ with air-10% CO,, p H 6.6, a t 25°C. Gold-amalgam area ratios 1:l (sample Bl) and 2:l (sample B2). Scan a t 0 . 6 V/hr. Scan a t 0.6 V/hr followed by 15 hr a t 0.0 v. Potentiostatic. Potentiostatic.
Potential control
Test solutions: 1) Synthetic saliva3v4deaerated with Nz-lO% COz, p H 6 . 6 , a t 25°C;
Y,Y1,81jY2
BI B2
8
Y,YI,Y2 Y,YI,YS
Phases identified
A1 A2
Sample
TABLE I Summary of Testing Conditions
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
1 10”
I
I
10.’
1
793
1
10“
I
1r3 C U R R E N T DENSITY, AMPSICM’
Fig. 2. Anodic polarization curve of a commercial dental amalgam (New True Dentalloy) with indicated potential range in free corrosion and in tests with external polarization.
Exposure under the Condition of Electrical Contact between an Amalgam Sample and a Gold Electrode The electrode potential of the amalgam-gold couple, used to control the potentiostat, was found to be almost independent of time, and to become only slightly more noble when a higher gold-amalgam ratio was used. With a freshly polished amalgam sample, the mixed potential during the 100 hr exposure was about - . 2 V (SCE). In some cases, no change in the appearance of the phases was observed. In other instances, the y2 phase started to deteriorate, as shown in Figures 5 and 6.
Exposure under the Conditions of Potentiodynamic and Potentiostatic Anodic Polarization beyond the Breakdown Potential The range of polarization is indicated in a polarization diagram in Figure 2 . The scanning electron micrographs in Figure 7 show the sample surfaces before and after the test. The y2 phase has been completely corroded away; no significant changes in the crystals of the y1phase are apparent. After an extended exposure at a potential
794
MAREK, HOCHMAN, AND OKABE
(b)
Fig. 3. SEM micrographs of sample Al, exposed to synthetic saliva for 100 hr at the free corrosion potential of dental amalgam. 6000 X. (a) Before exposure; (b) after exposure.
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
795
(b) Fig. 4. SEM micrographs of sample A2, exposed to synthetic saliva for 500 hr, at the free corrosion potential of dental amalgam. 1200 X. (a) Before exposure; (b) after exposure.
796
MAREK, HOCHMAN, AND OKABE
(b) Fig. 5. SEM micrographs of sample B1, exposed to synthetic saliva for 100 hr a t the corrosion potential of a gold-amalgam couple; gold-amalgam area ratio 1:l. 6000 X. (a) Before exposure; (b) after exposure.
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
797
(b) Fig. 6. SEM micrographs of sample B2, exposed to synthetic saliva for 100 hr at the corrosion potential of a gold-amalgam couple; gold-amalgam area ratio 2:l. 12,000 X. (a) Before exposure; (b) after exposure.
798
MAREK, HOCHMAN, AND OKABE
(b)
Fig. 7. SEM micrographs of sample C1, polarized in synthetic saliva from -0.78 to 0.0 V (SCE) at 0.6 V/hr. 2400 X. (a) Before exposure; (b) after exposure.
I N VZTRO CORROSION O F DENTAL AMALGAM PHASES
799
of 0.0 V (SCE), initiation of pits in the y1 phase can be observed, but corrosion damage is slight (Fig. 8).
Exposure Simulating the Conditions in a Tight Crevice An aggressive solution of high chloride content and low pH, based on the results of a study of crevice corrosion in dental amalgam,5 was used as the environment, and the potential was maintained potentiostatically a t a value typical for a crevice cell. The effect of this exposure is shown in Figures 9 and 10. The y2 phase disappeared completely; severe changes in the morphology of the y1 phase are also apparent.
DISCUSSION The value of the free corrosion potential of dental amalgam depends mainly on the cathodic polarization behavior and the state of anodic surfaces. The anodic polarization curve (Fig. 2 ) shows that in synthetic saliva, the anodic surfaces can be passive. When the sample is polarized to a potential in the passive range, the current a t first decreases with time, until a balance is reached between the rates of dissolution and formation of the protective film. I n free corrosion, the potential changes to more noble values as the anodic current decreases because the point of intersection of the anodic and cathodic curves moves. In the passive region the corrosion rate is extremely low, so that no damage is expected. Figure 3 shows that under these conditions no significant selective corrosion of the most active 7 2 phase takes place; this result is in agreement with previous observations in this laboratory.2v6 Depending on the cathodic conditions, such as the area of the cathodic surfaces and the redox potential of the solution, the corrosion potential may eventually reach a value noble enough for initiation of pitting. On a polarization curve (Fig. 2), the occurrence of pitting is marked by a sharp increase in current. The effect on the 7 2 phase is shown in Figure 4. The localized character of the attack is clearly noticeable. Coupling of the amalgam sample with gold provides a larger cathodic surface area, thereby increasing the total cathodic current, so that the intersection of the anodic and cathodic curves moves to a more noble potential. If this results in a breakdown of a passivity,
800
MAREK, HOCHMAN, AND OKABE
(b1 Fig. 8. SEM micrographs of sample C2, polarized in synthetic saliva from - .78 to 0.0 V (SCE) at 0.6 V/hr, followed by 15 hr exposure at 0.0 V. 6000 X. (a) Before exposure; (b) after exposure.
I N V I T R O CORROSION OF DENTAL AMALGAM PHASES
801
(b)
Fig. 9. SEM micrographs of sample D1, exposed for 15 hr to synthetic saliva acidified to pH 1, with increased chloride content, at -0.3 V (SCE). 1200 X. (a) Before exposure; (b) after exposure.
802
MAREK, HOCHMAN, AND OKABE
(b) Fig. 10. SEM micrographs of sample D2, exposed for 15 hr to synthetic saliva acidified to pH 1, with increased chloride content, at -0.3 V (SCE). 2400 X . (a) Before exposure; (b) after exposure.
I N VZTRO CORROSION OF DENTAL AMALGAM PHASES
803
an increase in the gold-amalgam surface area ratio then causes only a slight further change in the corrosion potential, since the intersection of the polarization curves now moves on the steep anodic slope beyond the breakdown potential; however, the corrosion rate increases. The cathodic effect of gold seems to be, though, relatively small; the corrosion potential of the gold-amalgam couple remained nearly stable at a value close to the pitting potential of the amalgam, and the rate of deterioration of the y2 phase seemed to be relatively low (Fig. 5 and 6). Polarization further beyond the breakdown potential brings about a rapid destruction of the y 2 phase (Fig. 7) and even some slight damage to the y1 phase (Fig. 8). Such conditions are, however, quite artificial and do not stimulate any probable situation in vivo because the currents are much higher than cathodic rates that can be reached in the mouth. All the exposures previously discussed were in nearly neutral synthetic saliva, approximating the environment of the external surfaces of a restoration under idealized conditions. Substantially different solution chemistry is likely to exist in enclosed spaces, such as crevices between the restoration and the tooth wall, and in pores. The possible changes of solution chemistry in these locations include depletion of dissolved oxygen, acidification, and an increase in chloride ion c~ncentration.~.’The presence of plaque on the amalgam surface also restricts the transport of oxygen to the metal and can result in formation of macrocells and acidification of the electrolyte at the interface. Samples exposed to a highly acidic solution with increased chloride ion concentration, at a typical potential observed in crevices, exhibited extreme changes. The y2 phase disappeared, and the topography of the y1 phase changed considerably (Fig. 9 and 10). Outlines of some of the original y1particles are still visible, but crystals cover a larger surface area of the substrate than before. A semiquantitative x-ray analysis of the new formations showed the composition to be that of the y1 phase, with a slight increase in the mercury-silver ratio from the value obtained for the original particles. It is hypothesized that under this condition of exposure, dissolution of not only tin, but also silver from the y1 phase takes place; mercury is released, and reacts with the Ag,Sn substrate to form new y1 phase. Any y z phase formed in this process is immediately dissolved. At present there is no experimental evidence that this effect plays a
804
MAREK, HOCHMAN, AND OKABE
significant role in the deterioration of dental amalgam restorations; however, the possibility cannot be overlooked. No deterioration of the y phase (Ag3Sn) was observed on any of the samples, indicating that this phase is the most stable one in this system.
CONCLUSIONS Specimens prepared by electroplating mercury on Ag3Sn provide new opportunities for studying corrosion properties of the individual phases of the dental amalgam structure. I n nearly neutral synthetic saliva, the major phases are either in the state of passivity or immunity and the corrosion rate is negligible. Some pitting of the yzphase may occur if corrosion potential becomes sufficiently noble due to a high rate of the cathodic reaction. When gold is in contact with dental amalgam, the rate of the cathodic reaction and the probability of deterioration of the y2 phase increase. In a highly acidic solution with a high concentration of chloride ions, which may exist in crevices and pores, the yz phase rapidly dissolves and the y1 phase suffers morphological changes. The exact mechanism for this process and its significance for dental amalgam deterioration arc the subject of a study which is in progress. The authors gratefully acknowledge the support of the NIDK. Grants DE 03601 and D E 03714.
References 1. T. Okabe, R . F. Hochman, and L. 0. Sims, J . Biomed. Muter. Res., 9, 221 (1975). 2. M. Marek and K. F. Hochman, paper presented to the 51st Session of the IADR, 1973, No. 192 (microfilm). 3. G. Tarii and F. Zucci, Mzneroa Stomat., 16, 710 (1967). 4. F. V. Wald and F. H. Cocks, J . Dent. Res., 50, 48 (1971). 5. M. Marek and R. F. Hochman, paper presented to the 53rd Session of the IADR, 1975, No. L166 (microfilm). 6. M . Marek and R. F. Hochman, paper submitted to the 50th Session of the IADR, 1972, No. 63 (microfilm). 7. M. Marek and R . F. Hochman, paper presented to the 51st Session of the IADR, 1973, No. 194 (microfilm).
Received August 24, 1975 Revised December 1, 1975
VOL. 10, PP. 789-804 (1976)
In vitro Corrosion of Dental Amalgam Phases* M. MAREK and R. F. HOCHMAN, Metallurgy Program, School of Chemical Engineering, Georgia Institute of Technology, Atlanta, Georgia, and T. OKABE, Department of Restorative Dentistry, School of Dentistry, Medical College of Georgia, Augusta, Georgia
Summary The effects of corrosion on the major phases of dental amalgam microstructure have been studied in vitro on samples prepared by electroplating mercury on AgaSn. Tests were made in synthetic saliva, and samples were examined in a scanning electron microscope before and after the exposure. Simple immersion had little effect on the phases. Contact of amalgam with gold increased the probability of deterioration of the y2 phase. By anodic polarization beyond the breakdown potential, the yz phase was selectively dissolved. Under conditions simulating a crevice, the 7 2 phase was dissolved and the y1 phase suffered morphological changes.
INTRODUCTION The major phases of a dental amalgam microstructure-y (Ag-Sn), different electrochemical properties and corrosion resistance. Although results of several investigations of the corrosion behavior of the individual phases have been reported, there is a lack of reports of direct examination of the interrelated corrosion effects. Observation of details of the corrosion-related changes in actual amalgam samples is difficult; the fine grain size requires relatively high magnification, the structural components have irregular and complicated shapes, the large variations in hardness and the presence of mercury make preparation of high-quality surfaces difficult, and structural changes often caused by preparation procedures can affect the electrochemical behavior. y1 (Ag-Hg-Sn), and y2 (Sn-Hg)-exhibit
*This paper is based in part on a presentation given to the 1975 General Session of the American Association for Dental Research in New York, N. Y. 789 @ 1976 by John Wiley & Sons, Inc.
790
MAREK, HOCHMAN, AND OKABE
I n this work, the effect of some of the variables of exposure on the three major phases of amalgam structure was investigated using direct observation of individual crystallites of the y1 and y z phases on a substrate of Ag3Sn (y phase).
MATERIALS AND METHODS The specimens were prepared by electroplating mercury on polished samples of Ag3Sn;the method had been developed for kinetic studies of amalgamation and the details are described elsewhere.' Phases 71 and y2, in polyhedral and needle form, respectively, precipitate separately on the substrate; the 01 phase can also form if the amount of mercury is relatively low. The identity of the phase is usually clear from the appearance ; however, energy-dispersive x-ray analysis was used in all experiments described here to verify the composition. The samples were examined in a scanning electron microscope (Cambridge Stereoscan Mark 11). The image of a selected area was photographed, the samples were removed and subjected t o various corrosion exposures under controlled conditions, washed in distilled water, dried, and examined again in the microscope. All the corrosion experiments were performed under potentiostatically controlled conditions for the following reasons. 1) The relative amounts of the phases prepared by plating Ag3Sn with mercury are different from their proportions in dental amalgam. To permit a n examination of the individual crystals, their number per unit area of the substrate must be limited; this results in samples with a large surface fraction of the noble Ag3Sn phase, and untypically noble free-corrosion potentials. 2) Some contamination of the surfaces by carbon deposition occurs during the examination in the electron microscope. This would also affect the free corrosion potential, making it more noble. By controlling the potential with a potentiostat, the influence of the carbon deposit is, t o a large extent, eliminated. Although this assumption could not be verified directly, i t was observed that identically prepared specimens, subjected t o same corrosion conditions but not examined in the electron microscope prior to the exposure, exhibited same general features as those described in this report. Corrosion potential of dental amalgam changes with time of exposure. To avoid an arbitrary choice of potentials for the potentio-
I N V I T R O CORROSION OF DENTAL AMALGAM PHASES
791
static control, some of the experiments were made with a potentiostat controlled by the potential of a sample of actual dental amalgam or amalgam-gold couple exposed in a separate cell. The experimental setup is illustrated in Figure 1. The conditions of the exposures are summarized in Table I. In addition, an anodic polarization curve of dental amalgam in synthetic saliva, deaerated with a nitrogen-10% carbon dioxide mixture, was recorded for correlation with the observed corrosion effects. The experimental procedure has been described elsewhere.2
RESULTS Exposure.under the Conditionof Free Corrosion of Dental Amalgam in Synthetic Saliva The corrosion potential of a sample of dental amalgam in aerated synthetic saliva, used to control the potentiostat, changed with time to more noble values, as indicated in a polarization diagram in Figure 2. Scanning electron micrographs of a sample before and after the exposure (100 hr) are shown in Figure 3. No deterioration of any of the phases was observed. After longer exposures, some pitting of the yz phase was occasionally observed, as illustrated in Figure 4.
POTENTIOSTAT
ISOLATION
ELECTROMETER
AMPLIFIER
CONTROLLED SAMPLE
CONTROLLING SAMPLE
Fig. 1. Experimental potentiostatic setup with potential control following the corrosion potential of another sample.
Y1Y1,YZ
Y,YI,Y?
YlYl
Y,YI,Y2
Y,YI
c1
C2
D1
D2
2
2
1 1
1 1
I 1
15 15
1.3 16.3
- .I95
100 100
- .30 - .30
- .78 to 0.0 - .78 to 0.0
- .190
- .57 t o - .29 - .57 t o - .20
Potential (V, SCE)
100 500
Test Exposure solution- period (hr)
~
HC1 t o p H 1.0,containing 2.6 M NaCl, a t 25°C. New True Dentalloy (S. S. White Co.). 24 carat.
9 10
7 8
5 6
4
3
Figureno. of the micrographs
2) Synthetic saliva acidified with
~~
Following the free corrosion potential of a commercial dental amalgamb in synthetic ~ a l i v a , 3 aerated ,~ with air-lO% COz, p H 6.6, at 25°C. Following the corrosion potential of a goldc-amalgamb couple in synthetic ~ a l i v a , ~aerated .~ with air-10% CO,, p H 6.6, a t 25°C. Gold-amalgam area ratios 1:l (sample Bl) and 2:l (sample B2). Scan a t 0 . 6 V/hr. Scan a t 0.6 V/hr followed by 15 hr a t 0.0 v. Potentiostatic. Potentiostatic.
Potential control
Test solutions: 1) Synthetic saliva3v4deaerated with Nz-lO% COz, p H 6 . 6 , a t 25°C;
Y,Y1,81jY2
BI B2
8
Y,YI,Y2 Y,YI,YS
Phases identified
A1 A2
Sample
TABLE I Summary of Testing Conditions
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
1 10”
I
I
10.’
1
793
1
10“
I
1r3 C U R R E N T DENSITY, AMPSICM’
Fig. 2. Anodic polarization curve of a commercial dental amalgam (New True Dentalloy) with indicated potential range in free corrosion and in tests with external polarization.
Exposure under the Condition of Electrical Contact between an Amalgam Sample and a Gold Electrode The electrode potential of the amalgam-gold couple, used to control the potentiostat, was found to be almost independent of time, and to become only slightly more noble when a higher gold-amalgam ratio was used. With a freshly polished amalgam sample, the mixed potential during the 100 hr exposure was about - . 2 V (SCE). In some cases, no change in the appearance of the phases was observed. In other instances, the y2 phase started to deteriorate, as shown in Figures 5 and 6.
Exposure under the Conditions of Potentiodynamic and Potentiostatic Anodic Polarization beyond the Breakdown Potential The range of polarization is indicated in a polarization diagram in Figure 2 . The scanning electron micrographs in Figure 7 show the sample surfaces before and after the test. The y2 phase has been completely corroded away; no significant changes in the crystals of the y1phase are apparent. After an extended exposure at a potential
794
MAREK, HOCHMAN, AND OKABE
(b)
Fig. 3. SEM micrographs of sample Al, exposed to synthetic saliva for 100 hr at the free corrosion potential of dental amalgam. 6000 X. (a) Before exposure; (b) after exposure.
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
795
(b) Fig. 4. SEM micrographs of sample A2, exposed to synthetic saliva for 500 hr, at the free corrosion potential of dental amalgam. 1200 X. (a) Before exposure; (b) after exposure.
796
MAREK, HOCHMAN, AND OKABE
(b) Fig. 5. SEM micrographs of sample B1, exposed to synthetic saliva for 100 hr a t the corrosion potential of a gold-amalgam couple; gold-amalgam area ratio 1:l. 6000 X. (a) Before exposure; (b) after exposure.
I N VITRO CORROSION OF DENTAL AMALGAM PHASES
797
(b) Fig. 6. SEM micrographs of sample B2, exposed to synthetic saliva for 100 hr at the corrosion potential of a gold-amalgam couple; gold-amalgam area ratio 2:l. 12,000 X. (a) Before exposure; (b) after exposure.
798
MAREK, HOCHMAN, AND OKABE
(b)
Fig. 7. SEM micrographs of sample C1, polarized in synthetic saliva from -0.78 to 0.0 V (SCE) at 0.6 V/hr. 2400 X. (a) Before exposure; (b) after exposure.
I N VZTRO CORROSION O F DENTAL AMALGAM PHASES
799
of 0.0 V (SCE), initiation of pits in the y1 phase can be observed, but corrosion damage is slight (Fig. 8).
Exposure Simulating the Conditions in a Tight Crevice An aggressive solution of high chloride content and low pH, based on the results of a study of crevice corrosion in dental amalgam,5 was used as the environment, and the potential was maintained potentiostatically a t a value typical for a crevice cell. The effect of this exposure is shown in Figures 9 and 10. The y2 phase disappeared completely; severe changes in the morphology of the y1 phase are also apparent.
DISCUSSION The value of the free corrosion potential of dental amalgam depends mainly on the cathodic polarization behavior and the state of anodic surfaces. The anodic polarization curve (Fig. 2 ) shows that in synthetic saliva, the anodic surfaces can be passive. When the sample is polarized to a potential in the passive range, the current a t first decreases with time, until a balance is reached between the rates of dissolution and formation of the protective film. I n free corrosion, the potential changes to more noble values as the anodic current decreases because the point of intersection of the anodic and cathodic curves moves. In the passive region the corrosion rate is extremely low, so that no damage is expected. Figure 3 shows that under these conditions no significant selective corrosion of the most active 7 2 phase takes place; this result is in agreement with previous observations in this laboratory.2v6 Depending on the cathodic conditions, such as the area of the cathodic surfaces and the redox potential of the solution, the corrosion potential may eventually reach a value noble enough for initiation of pitting. On a polarization curve (Fig. 2), the occurrence of pitting is marked by a sharp increase in current. The effect on the 7 2 phase is shown in Figure 4. The localized character of the attack is clearly noticeable. Coupling of the amalgam sample with gold provides a larger cathodic surface area, thereby increasing the total cathodic current, so that the intersection of the anodic and cathodic curves moves to a more noble potential. If this results in a breakdown of a passivity,
800
MAREK, HOCHMAN, AND OKABE
(b1 Fig. 8. SEM micrographs of sample C2, polarized in synthetic saliva from - .78 to 0.0 V (SCE) at 0.6 V/hr, followed by 15 hr exposure at 0.0 V. 6000 X. (a) Before exposure; (b) after exposure.
I N V I T R O CORROSION OF DENTAL AMALGAM PHASES
801
(b)
Fig. 9. SEM micrographs of sample D1, exposed for 15 hr to synthetic saliva acidified to pH 1, with increased chloride content, at -0.3 V (SCE). 1200 X. (a) Before exposure; (b) after exposure.
802
MAREK, HOCHMAN, AND OKABE
(b) Fig. 10. SEM micrographs of sample D2, exposed for 15 hr to synthetic saliva acidified to pH 1, with increased chloride content, at -0.3 V (SCE). 2400 X . (a) Before exposure; (b) after exposure.
I N VZTRO CORROSION OF DENTAL AMALGAM PHASES
803
an increase in the gold-amalgam surface area ratio then causes only a slight further change in the corrosion potential, since the intersection of the polarization curves now moves on the steep anodic slope beyond the breakdown potential; however, the corrosion rate increases. The cathodic effect of gold seems to be, though, relatively small; the corrosion potential of the gold-amalgam couple remained nearly stable at a value close to the pitting potential of the amalgam, and the rate of deterioration of the y2 phase seemed to be relatively low (Fig. 5 and 6). Polarization further beyond the breakdown potential brings about a rapid destruction of the y 2 phase (Fig. 7) and even some slight damage to the y1 phase (Fig. 8). Such conditions are, however, quite artificial and do not stimulate any probable situation in vivo because the currents are much higher than cathodic rates that can be reached in the mouth. All the exposures previously discussed were in nearly neutral synthetic saliva, approximating the environment of the external surfaces of a restoration under idealized conditions. Substantially different solution chemistry is likely to exist in enclosed spaces, such as crevices between the restoration and the tooth wall, and in pores. The possible changes of solution chemistry in these locations include depletion of dissolved oxygen, acidification, and an increase in chloride ion c~ncentration.~.’The presence of plaque on the amalgam surface also restricts the transport of oxygen to the metal and can result in formation of macrocells and acidification of the electrolyte at the interface. Samples exposed to a highly acidic solution with increased chloride ion concentration, at a typical potential observed in crevices, exhibited extreme changes. The y2 phase disappeared, and the topography of the y1 phase changed considerably (Fig. 9 and 10). Outlines of some of the original y1particles are still visible, but crystals cover a larger surface area of the substrate than before. A semiquantitative x-ray analysis of the new formations showed the composition to be that of the y1 phase, with a slight increase in the mercury-silver ratio from the value obtained for the original particles. It is hypothesized that under this condition of exposure, dissolution of not only tin, but also silver from the y1 phase takes place; mercury is released, and reacts with the Ag,Sn substrate to form new y1 phase. Any y z phase formed in this process is immediately dissolved. At present there is no experimental evidence that this effect plays a
804
MAREK, HOCHMAN, AND OKABE
significant role in the deterioration of dental amalgam restorations; however, the possibility cannot be overlooked. No deterioration of the y phase (Ag3Sn) was observed on any of the samples, indicating that this phase is the most stable one in this system.
CONCLUSIONS Specimens prepared by electroplating mercury on Ag3Sn provide new opportunities for studying corrosion properties of the individual phases of the dental amalgam structure. I n nearly neutral synthetic saliva, the major phases are either in the state of passivity or immunity and the corrosion rate is negligible. Some pitting of the yzphase may occur if corrosion potential becomes sufficiently noble due to a high rate of the cathodic reaction. When gold is in contact with dental amalgam, the rate of the cathodic reaction and the probability of deterioration of the y2 phase increase. In a highly acidic solution with a high concentration of chloride ions, which may exist in crevices and pores, the yz phase rapidly dissolves and the y1 phase suffers morphological changes. The exact mechanism for this process and its significance for dental amalgam deterioration arc the subject of a study which is in progress. The authors gratefully acknowledge the support of the NIDK. Grants DE 03601 and D E 03714.
References 1. T. Okabe, R . F. Hochman, and L. 0. Sims, J . Biomed. Muter. Res., 9, 221 (1975). 2. M. Marek and K. F. Hochman, paper presented to the 51st Session of the IADR, 1973, No. 192 (microfilm). 3. G. Tarii and F. Zucci, Mzneroa Stomat., 16, 710 (1967). 4. F. V. Wald and F. H. Cocks, J . Dent. Res., 50, 48 (1971). 5. M. Marek and R. F. Hochman, paper presented to the 53rd Session of the IADR, 1975, No. L166 (microfilm). 6. M . Marek and R. F. Hochman, paper submitted to the 50th Session of the IADR, 1972, No. 63 (microfilm). 7. M. Marek and R . F. Hochman, paper presented to the 51st Session of the IADR, 1973, No. 194 (microfilm).
Received August 24, 1975 Revised December 1, 1975
