Release of Elements due to Electrochemical Corrosion of Dental Amalgam

Journal of Dental Research http://jdr.sagepub.com/

Release of Elements due to Electrochemical Corrosion of Dental Amalgam S. Olsson, A. Berglund and M. Bergman J DENT RES 1994 73: 33 DOI: 10.1177/00220345940730010501 The online version of this article can be found at: http://jdr.sagepub.com/content/73/1/33

Published by: http://www.sagepublications.com

On behalf of: International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at: Email Alerts: http://jdr.sagepub.com/cgi/alerts Subscriptions: http://jdr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jdr.sagepub.com/content/73/1/33.refs.html

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

J Dent Res 73(l):33-43,January, 1994

Release of Elements due to Electrochemical Corrosion of Dental Amalgam S. Olsson, A. Berglund, and M. Bergman Department of Dental Materials Science, Faculty of Odontology, University of Umea, Sweden

Abstract. The corrosion pattern of dental amalgam in aqueous media was interpreted theoretically by means of log(a/arf)pe diagrams. The definitions on which the diagrams were based were given, and their features were described. All sparingly soluble compounds which were expected to be formed in reactions with the solvents considered were listed. All the corrosion products reported in the current literature were found to be formed,and the conditionsfor their formation were established. It emerged that it was necessary to exclude other sparingly soluble compounds which theoretically might be formed. Two compounds, CuSCN and AgSCN, which have not been reported previously were found to be possible corrosion products. Corrosion products containing mercury compound cannot be formed on amalgam restorations with no metallic contact with other materials. Key words. Dental Amalgam, Corrosion, Mercury.

Received February 22,1993; Accepted August 8,1993 This work was supported by the Swedish Medical Research Council (Project No. 07523). Correspondence and reprint requests should be addressed to: Dr. M. Bergman, Department of Dental Materials Science, Faculty of Odontology, University of UmeA, S-901 87 UmeA, Sweden.

Introduction During the last decade, there has been an intensification in research activity into the mechanisms involved in the release of mercury from dental amalgam. For well-founded reasons, interest has focused mainly on mercury evaporation from amalgam restorations (Svare et al., 1981; Abraham et al., 1984; Ott et al., 1984; Patterson et al., 1985; Vimy and Lorscheider, 1985a,b; Berglund et aL., 1988; Olsson et al., 1989; Berglund, 1990,1992; Bergman, 1992; Olsson and Bergman, 1992). However, since the daily uptake of mercury from inhaled mercury vapor released from dental amalgam seems to make a very small contribution to the total body burden of mercury, in comparison with what can be tolerated in the work environment (Berglund et al., 1988; Berglund, 1990; Mackert, 1991; Olsson and Bergman, 1992), it is also important to consider other routes f or mercury uptake f rom amalgam restorationsfor example, the uptake in the gastro-intestinal tract. Inorganic mercury compounds which are swallowed will be absorbed from the human gastro-intestinal tract to a level of less than 10%, on average (WHO, 1991). Apart from the dissolution in saliva of elemental mercury vapor released from amalgam restorations, further contributions to saliva mercury levels can occur as a result of both wear and electrochemical corrosion of the restorations. When electrochemical corrosion is considered, it is important to note that dental amalgams are the most complex biomaterials from a metallurgical point of view. The set amalgam is a dynamic material in which solid-state reactions occur for a long time (Marshall and Marshall, 1992). Since this takes place in the complex and variable oral environment, it is obvious that in vitro studies can reflect the real in vivo conditions to only a limited extent. Berglund (1993) studied the release of mercury vapor from different types of amalgam alloys in a combined in vitro and in vivo study. He found that the electrochemical 33

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

34

j Dent Res 73(l) 1994

Olsson et al. 10 A log

Cu(M)

151

Me(s))

-Me(

(Me Z+I

Hg(l)

Hg(l) (Hg 2+ )112 Ag(s) Ag

Cu +

02

Hg 2+

1 1 1 1 1 1 1 1 1 1 1 1 1 1

II

II II II II II II II II II II II

Zn(s) Zn 2+

-15

7

\

I I I I I I

-10

-5

-0.5916

-0.2958

0

-0.8874

0 0

0

5

-2'4

n

0.2958

10

15 Eh V

n

0.5916

0.8874

Figure 1. Log(a./a re)-pe diagram of dental amalgam in synthetic saliva according to Tani and Zucci (1967) at pH = 6.7 and 25° C. The sparingly soluble compounds which are expected to be formed in reactions with the solvent, represented by numbers, are given in Table 4 together with their pe-values and solubilities.

reactions occurring on the surfaces of amalgam specimens in test solutions influenced the release of mercury vapor in various ways. Mercury is the most noble of the major elements in dental amalgam, and under the conditions prevalent in the oral cavity, silver and mercury have the lowest electrochemical activity (Gross and Harrison, 1989). However, despite their limited tendency to dissolve electrochemically, these elements may form some almost-insoluble sulfides. Among the corrosion products of various kinds which have been identified on corroded in vitro specimens and in retrieved amalgam restorations are SnO, SnO2, Sn4(OH) 6C2, Cu20, CuCl2 3Cu(OH)2, and CuCl (Mateer and Reitz, 1970; Otani et al., 1973; Holland and Asgar, 1974; Sarkar et al., 1975; Espevik, 1977; Marshall and Marshall, 1980; Marshall et al., 1980, 1982,1987; Jensen, 1982; Lin et al., 1983a,b; Sutow et al., 1991). The formation of ZnSn(OH)6and CuSn(OH)6, whichdemandshigh pH values, is not likely to occur in vivo, according toJensen (1982). Since certain of these corrosion products may form protective films, thus influencing the release of mercury vapor from amalgam restorations, it was considered to be of interest to relate the experimental findings from the abovementioned studies on corrosion products to known electrochemical data. The aim of the present work was therefore to evaluate pos-

sible redox-reactions, with regard to the major elements in dental amalgam and the possible corrosion products formed in reactions with the environment.

Materials and methods The corrosion pattern of dental amalgams in the aqueous media of the oral cavity is very complex, owing to many competing corrosion reactions and the complexity of human saliva. Since the latter changes rapidly outside the oral cavity, it is usually very difficult or indeed impossible to use human saliva for in vitro corrosion studies. Therefore, so-called synthetic saliva solutions are generally used. There are several compositions reported in the literature which has been reviewed by Marek (1983). Among those, the two compositions given in Tables 1 and 2 were selected for this paper. The Fusayama solution is very often used in corrosion studies and is most suitable, since it contains sulfide. The synthetic saliva according to Tani and Zucci (1967), which contains HCO3ions, was selected so that the CO2/HCO3-buffering system which exists in human saliva could be discussed. In most corrosion studies where corrosion products were analyzed, the reactions have reached a steady state which is close to equilibrium. Real equilibrium cannot be attained, due

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

35

Electrochemical Corrosion ofDental Amalgam

j Dent Res 73(l) 1994

pH 2

4

6

8

10

-4

-6 -8 pe

-10 -12 I

I

-1.183

-0.8874

I

I

-0.5916 -0.2958

I

I

I

I

0

0.2958

0.5916

0.8874

N

Figure 2. Log(a/arf)-pe diagram of dental amalgam in Fusayama solution at pH = 5.35 and 25 C. The sparingly soluble compounds which are expected to be formed in reactions with the solvent, represented by numbers, are given in Table 5 together with their pe-values and solubilities.

to the fact that solid-state reactions change the phase composition in the amalgam fillings. However, the solid-state reactions are very slow compared with the corrosion reactions. It seems therefore reasonable to assume that the corrosion products can be treated as if chemical equilibrium exists. One way to evaluate complex redox-reaction patterns is to use log(a,/aref)-pe diagrams (Silkn, 1952; Sillen and Martell, 1964). The equations for possible redox reactions occurring in the media used and the reaction products of the corrosion of dental amalgams are given in Table 3. The basic data were selected from Stability Constants (Sillen and Martell, 1964). Since there are many reactions possible in the media used, and since only a few of the reaction products are determined at 37°C but all are determined at 25C, all the reactions in Table 3 are given at 25'C, for comparability. In Figs. 1 and 2, the corresponding log(a,/aref)-pe diagrams from the reactions given in Table 3 are shown for the solvents given in Tables 1 and 2. In Table 3, there are two kinds of reactions. The first reaction in Table 3, a redox reaction, corresponds to a line with a slope of -2, denoted Zn(s)/Zn2+ in Fig. 1. The line cuts the pe axis at -12.9. The independent coordinate, pe, of the pe-diagram is defined: pe = - logl0fe-I = E/(RT ln 10) F

(1)

-14 Figure 3. pe as a function of pH and ion concentration of In com-

pounds in synthetic saliva (Tani and Zucci, 1967) at 25°C. In(OH)3; * ln(OH)1.75C11.25

where E is the electrode potential in the cell which contains the redox couple and the standard Pt, H2IH+ reference electrode, R is the gas constant, T absolute temperature, and F Faraday constant. The value of (RT/F) ln 10 at 25°C is 0.05916 V. The pe-scale is used in order f or the redox data to be combinable with the equilibrium data. The relation between the scales given in Figs. 1 and 2 is obtained from:

Eh = pe x 0.05916

(2)

The Eh-scale is given only for orientation to the electrode potential data. The value 12.90 in the first reaction of Table 3 is denoted pe', which is defined: pe = E°/(RT ln 10) = 1 logl0K Z F

(3)

where E is the standard electrode potential, z is the number of electrons of the redox reaction, and log10K is given in Stability Constants (Sillen and Martell, 1964). The reactions 2 to 9 of Table 3 represent insoluble or sparingly soluble compounds of the Zn case. The corresponding equations are straight lines parallel with the pe axis. Their intersections with the line Zn(s)/Zn2+ are denoted by means of a horizontal line and a figure given in Tables 4 and 5 for the media used. In amalgam, the activity of the metallic components is less

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

Olsson et al.

36

J Dent Res 73(1)1994

Table 1. Composition of the modified Fusayama solution

Compound

g/L

NaCl

0.4 0.4 0.795 1.0 0.78 0.005

KCI CaCl2 Urea

NaH2PO4 H20 Na2S x H20 (X = 7 - 9) pH = 5.2 - 5.5 for freshly made solution a

b c

mol/L

log(conc)a

0.0176b

log[CI-I = -1.75

0.067 0.005 0.000021

log[H2PO4 I = -2.3 log[H2ScI = -4.7

Real concentration at the pH used. Total chloride concentration. H2S concentration not stabilized.

Table 2. Composition of synthetic saliva (Tani and Zucci, 1967)

Compound

g/L

mol/L

log(conc)a

KSCN NaHCO3

0.517 1.253 1.471 0.1878 0.90

0.00532 0.0149 0.0197 0.00136 0.010

log[SCNi = -2.27 log[HCO3 b1= -2.02 log[Cl-I = -1.70 log[H2P04-I = -2.98 log[HLI = -4.84

KCI NaH2PO4 H20 Lactic acid (HL) pH = 6.7 a

b

Real concentration at the pH used. HCO3- concentration kept constant by means of 10% CO2 in the gas mixture (Wald and Cocks, 1971).

than unity. The activity of the pure metal in the redox reaction can be replaced with:

{Me(s)) = [amalgam]f.,

(4)

where [amalgam] denotes the amalgam used in the specimens and still has the activity equal to unity, andfMe is the activity factor of the least noble metallic component of the corroding phase of the amalgam used. When Eq. (4) is introduced into the equation of the redox couple, Eq. (5) is obtained:

log [amalgam] = - z [pe - pe + 1 logfe]

tMez+l

(5)

Z

When fMe decreases, the line representing the redox couple Me(s)/Mez+ is shifted to the right in the pe-diagram. Since the activity of the solid metal Me(s) or the amalgam and that of the insoluble solid compounds are unity, the same concentration of Mez+ is obtained when in Table 3, for instance, the first equation and one of the equations of the solid compounds of the same redox couple are put equal. Then, pe corresponding to the solubility of, for instance, Zn3(PO4)2(s) is obtained from: pe = -11.73 - 2/6 logfH2PO4-I - 4/6 pH

In order for correct pe-values to be attained, the anion concentrations corresponding to the pH-value used must be calculated from the pKa of the corresponding acids and the total concentrations given in Tables 1 and 2. The pe-values are measures of the reactivity of the redox couple at the solubility level of the solid compound formed. The pe-values and the solubility, given by the concentration of Mez+ in the medium used, are given in Tables 4 and 5 in order of increasing pevalues. The more negative the pe-value, the more corrosionprone the metal. Therefore, the corrosion will start with the least noble component of the amalgam and with formation of the corrosion product which has the most negative pe-value. The corrosion reaction will proceed until the activity f actor of the corroding component of the amalgam has decreased so much that the next pe-value has been reached [cf Eq. (5)] or until the corrosion is strongly reduced due to passivation. If the passivation is not efficient enough to stop the corrosion, the latter will continue with the second noblest metallic component and will form the compound which has the second most negative pe-value. Equations equivalent to Eq. (6) also give the possibility for pe to be calculated due to variations in the concentration of anions and pH in the aqueous solution used. When pe is plotted against pH, it is possible for pe-values of several possible reaction products of the same redox couple to be compared,

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

Electrochemical Corrosion ofDental Amalgam

J Dent Res 73(l) 1994

37

and the corrosion products found in experimental studies were identified.

6

Discussion

4

2

pe 0

N mA

4

6

8

10

-2 -4

pH Figure 4. pe as a function of pH and ion concentrations of Cu compounds in synthetic saliva (Tani and Zucci, 1967) at 25 C. -Orepresents CuSCN; 0, Cu 0 A Cu2(OH)2CO3; A, Cu3(OH)2(CO3)2; 0 * CuO; X, CuCI; X AgSCN. Cu (PO4)2; 0Cu(OH) 5C1

and a decision made as to which of the products might be formed. Since the comparisons are made between compounds belonging to the same corroding metallic component, it is not necessary to study the decreasing activity of the corroding component simultaneously. The extension of the curves was found by means of calculations of AG vs. the saturated calomel electrode (SCE, 0.2446 V) for the processes where the solid metal was directly electrochemically converted to the corrosion products in synthetic saliva (Tani and Zucci, 1967).

Results In Fig. 1 it can be seen that the most probable corrosion products in Zn-containing amalgams are those of Zn and Sn in conventional amalgam and those of Zn, Sn, and Cu in highCu amalgam. It can also be seen that corrosion of Ag and Hg is less probable. Furthermore, it was found that indium in Incontaining amalgams is the most reactive metal component after Zn. If the medium contains sulfide, the most probable corrosion products will be metal sulfides (Fig. 2). In Figs. 3 and 4, the pe-values of the expected corrosion products were plotted against pH within the region where AG vs. SCE for the corresponding corrosion process is negative. The conditions for the formation of the compounds are shown in the Figs.,

An alloy surface which containsdifferent phasesconsisting of metal components with different electrode potentials will form a corrosion cell if the surface comes in contact with an aqueous medium. General as well as local corrosion-for example, pitting-can appear. At the anodic site, the metal ions released as a result of corrosion react with the medium. In certain cases, the ions are to a large extent hydrolyzed and a more acid solution is generated, causing a local rise in the passivation potential with increased pe-values. In addition, an increased chloride ion concentration corresponding to decreased pe-values may occur in a pit or crevice as a consequence of the chloride ions migrating with the corrosion current, i.e., due to the charge transport in the solution. The decrease in pH increases the dissolution of the passive layer, and the increase in chloride concentration increases the corrosion, i.e., both factors further corrosion and prevent or render it difficult for repassivation to be obtained. The corrosion of amalgam is very complex, but by means of log(a/aref)-pe diagrams, it is possible to obtain a survey of the corrosion pattern. From Fig. 1, it is seen that in synthetic saliva, according to Tani and Zucci (1967), the reactivity of the metal components is ranked Zn, In, Sn, Cu, Ag, and Hg. In the case of tin, the composition of corrosion products suggested has been discussed. The composition of basic Sn(II) chloride was originally assumed to be Sn(OH)ClH2O (Sarkar et al., 1975), but was later found to be Sn4(OH)6C12 (Marshall and Marshall, 1980). Old solubility data for Sn(OH)ClH2O are given in Stability Constants (Sillen and Martell, 1964), but there are no solubility data f or Sn4(OH)6C12. Furthermore, there are no reliable data for the Sn(s)/Sn4+ system. It is therefore not C2 and Sn02. Fortupossible to give the pe-values for Sn the composition of studied al. (1963) et nately, Donaldson from that all precipitates and found basic Sn(II) chlorides the had 1.4-4.5 composition oxygen-free solutions of pH Sn4(OH)6C12 and that above pH 4.5, Sn4(OH)6C12 appeared together with increasing content of hydrous Sn(II) oxide, with increasing pH. The system was sensitive to oxygen, and if oxygen was present, hydrous Sn(IV) oxide was also formed. These results are in agreement with clinical experience. In pits and deep crevices, where pH may be low and the chloride concentration may be increased due to ion migration, Sn4(OH)6C12 can be formed from the y2-phase (Marshall and Marshall, 1980; Moberg and Oden, 1985). The formation of hydrous SnO2 may be reduced due to lack of oxygen in the crevices. Above pH 4.5, SnO and SnO2 can also be formed. When the activity of Sn in the amalgam surf ace decreases, the line in Fig. 1 denoted Sn(s)/Sn2+ will be shifted to the right [Eq. (5)]. The pe-value for SnO (No. 10, Table 4 and Fig. 1) must then be shifted to that for CuSCN (No. 12, Table 4 and Fig. 1). Then, the line according to Eq. (5) is to be shifted from -8.20

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

38

J Dent Res 73(1)1994

Olsson et al.

Table 3. Reactions on which (Log(ai/aref)-pe)-diagrams are based [data from stability constants (Sillen and Martell, 1964)1 Reaction at 25 C Redox couple Zn

Zn2+

1 = -2 [pe + 12.901 log{( fZn2+1

log Zn(OH)2(s)1 = -11.5 + 2 pH

{Zn241 log Zn(°OH')5C(l.(s

=

-7.60 + 0.5 logiCI-1 + 1.5 pH

= -8.20 + 0.4 logiCllogfZn(OH)½tClo4(s)1 Zn

+1

+

1.6 pH

log{ZnCO3(s)1 = 0.37 + log{HC03 la + pH

logIZn(OH)1j8(CO3)0.3s)1 {Zn ~1 loglZn(Lb)2(s)1 fZn2-rI

=

=

-7.22 + 0.36 logIHCO3ila + 1.64 pH

-3.97 + 2 log{HLtI + 2 pH

lOgiZn3(P4)2(S) "3 = -2.34 +2/3 log {H2PO4-) + 4/3 pH IZn2+I

log In

In3+

2S) I ZnI

=

4.91 + log IH2SIC + 2 pH

= -3 [pe + 5.681 logIIn(s)i {In3+1

log in(OH)3 s)1 = -8.65 + 3 pH

{ln3+)

loglIn(OH)u[In +IC.2sl(

= 0.70 + 1.25 ICI-l + 1.75 pH

log1In2SPS)+I 1/2 = 3.37 + 1.5 logfH2SIC + 3 pH fin

Sn

Sn2+

log[

(Sn2+1

= -2 [pe + 2.381

log 21 = -1.76 + 2 pH (Sn2~ = xxx + 0.5 logIClI) + 1.5 pH log{Sn(OH)5CJ0.5U(s) [Sn +1

log[S2) [Sn2~ Cu Cu+

= 4.98 + log [H2SI + 2 pH

logfCu(s)1 --1 [pe - 8.801

[GuI+

= 6.50 + log(ClI} log[CuCl(s)) {Cu+}

lOg{Cu20(s)/2 = 0.84 + pH log[CuSCN(s)1 = 14.32 + log(SCN[Cu+1 /2 = 13.31 + 0.5 1ogfH2S)C + pH 1ogKCu2S(s)l ICU+) Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

(Table continues next page)

Electrochemical Corrosion of Dental Amalgam

J Dent Res 73(1)19 94 Cu

Cu2+

39

log CU(S)1 = -2 [pe - 5.701

tCu2+i

logCUO(s)1= -8.37 + 2 pH {Cu2+1

logICu(OH)1.5ClO.5(s)) = -3.65 + 0.5 log{ClI} + 1.5 pH (Cu2+1

log{Cu3(OH)2(C0322(s)jll/3 = -0.90 + 2/3 log{HCO3- a+ 4/3 pH {Cu2+1

log1CU2(OH)2,C3(s)j1'/2 (Cu +}

=

-2.28 + 0.5 logfHCO3 la + 1.5 pH

logICU3(po )2(s))L/3 = -0.72 + 2/3 logIH2PO4 I + 4/3 pH ICu

}

log Cu(SCN),(s)I = 3.65 + 2 logISCN-I

ICu2+,

log(Cu(Lb)2(s)) =-2.88 + 2 logtHLb ICu2+I

= 15.53 logC2uS(s)i fcu2+

Ag

Ag+

+

+2 pH

logIH2SIc + 2 pH

log Ag(s)I = -1 [pe -13.511

{Ag+I)

ioglAgCl(s)} = 9.75 + logICli (Ag+i logAg2(s1/2 = -6.31 + pH JAg+I logg2CO3(s)_2 = -0.54 + 0.5 logIHCO3 la + 0.5 pH (Ag+I 1OglAg3PO4(s)'/3 =-0.56 + 1/3 log(H2PO4 1+ 2/3 pH {Ag+}

log(AgSCN(s)I = 11.94 + log{SCN-I {Ag+)

logLAg2S(S)/2 = 14.64 + 0.5 logIH2S)c + pH {Ag+I

H g-

(Hg22+)1/2

log H

= -1

IHg22+I1/2

[pe -13.401

logtg2Cl2(S)1I/2 logf g2CO3S)I

=

8.87 + logICl

= 2.86 + 0.5

log(HCO3 }a + 0.5 pH

(s)1'/2 = 9.76 + log(SCN-I logtg2(SCN) {Hg22+ii/2 Hp

Hg2+

a

means of 10% CO2 in the gas mixture (Wald

log{Hg(l), = -2 [pe - 14.381 {Hg2+} b

log{HgO(s)I = -2.54 + 2 pH IHg2+I g(HgS(s)} = 30.77 + log{H2S1C.+ 2 pH

HCO3- concentration kept constant by

c d e

and Cocks, 1971). Lactic acid (HL), lactate (L). H2S concentration not stabilized. Azurite.

Malachite.

{Hg2+)

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

40

Olsson et al.

J Dent Res 73(1)1994

Table 4. pe-values and solubilities of compounds in synthetic saliva solution (Tani and Zucci, 1967) No.

Compound

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

ZnCO3

to -3.25, i.e., -1/2

Concentration of Mez+ (mol/L)

10-5.05 10-4.60 10-3 04

pe-value

Zn(OH)2

10-1.9

Zn(OH)1.6Cl0.4 Zn(OH)1.5C10.5

10-1.84

Zn lactate

1o+0.25 lo-11.5

-15.4 -15.2 -14.4 -13.9 -13.8 -13.7 -12.8 -9.49

lo-10.3 10-11.6

-9.11 -8.20

CuSCN Cu2O

lo-12.1 10-7.54

CU2(OH)2CO3 CU3(OH)2(CO3)2

Cu3(PO4)2

10-6.76 10-6.68 10-6.23

Cu(OH)1.5C10.5

10-5 55

-3.25 +1.26 +232 +2.36 +2.59 +2.93 +3.19 +3.84 +4.00 +5.28 +5.46 +5.91 +6.15 +6.23 +8.20 +8.95 +10.6 +10.6 +13.1 +13.9

Zn3(PO4)2

Zn(OH)1.28(CO3)0.36

In(OH)3

In(OH)1.75C11.25 SnO

Sn(OH)1.5C10.5

CuO AgSCN CuCl Cu(II)lactate AgCl Hg2(SCN)2 Cu(SCN)2 Hg2C12 Hg2CO3 HgO Ag2CO3 Ag3PO4 Ag2O 02 (Activity = 0.2)

logfs. = 4.95 andfs. = 10-99, and thus theoreti-

cally the concentration of Sn in the amalgam surface must be

very low in order for the corrosion of Sn and Cu to be exchanged. Since the amalgam surface of a _y2-amalgam probably never loses all the -y2-phase in the surface or in the pores, it seems to be less probable that copper will corrode in this

particular case. In dental amalgam containing indium, two corrosion products are possible (Nos. 8 and 9, Fig. 1 and Table 4). In Fig. 3, their pe-functions were plotted against pH. It is seen that ln(OH)3 is formed above pH = 5.8 and that In(OH)175Cl125 is the corrosion product at pH below 5.8. Since these compounds are sensitive

to variations in pH (Table 3), their ability to cause passivation is probably reduced. This is supported by experimental results obtained by Berglund (1993), who found, in isotonic saline solution at pH = 6, that breakthroughs in the layer of corrosion products had occurred, resulting in mercury vapor release

lo-1.60

10-5.03 10-9.67 10-480 10-0.84 10-8.05 10-7.49 10+0 90 10-717 10-5.20 10-10.9 10-2.87 10-2.91

10-0.39

with peaks in mercury vapor measurements recorded during a ten-day interval. The next pe-value after those of Sn is that of CuSCN (No. 12, Table 4 and Fig. 1). In Fig. 4, the pe-values of CuSCN are plotted against pH. It is seen that the pe-f unction of CuSCN is below that of the other Cu-compounds. Since CuSCN is insoluble in hydrochloric acid (Partington, 1947) and has a very low solubility in the solution used (Table 4), it should be stable in the oral cavity and also in crevices. Therefore, if thiocyanate is present in the solution, CuSCN can be expected to be present among the corrosion products of dental amalgams. The reason for the fact that CuSCN has never been identified as a corrosion product is, perhaps, that no one has ever looked for it. The corrosion of Cu has been observed in dental amalgam. The determining pe-value in solutions which do not contain thiocyanate is seen to be that of Cu20 (No. 13, Table 4 and Fig.

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

j Dent Res 73(l) 1994

Electrochemical Corrosion of Dental Amalgam

41

Table 5. pe-values and solubilities of compounds in Fusayama solution No.

Compound

1 2

Zn3(PO4)2

3

4 5 6 7 8 9 10 11 12

13

Concentration of MeZ+ (mol/L)

10-10.9 10-3.26

ZnS

-18.4 -14.5

Zn(OH)1.6C10.4 Zn(OH)1.5C10.5

i0+° 34

-12.7

-12.7

Zn(OH)2

10+0.45 10+0.80

In2S3

lo-124

In(OH)1.75CI125

-9.81 -8.30

SnS

10-788 10-7.40 10110

Cu2S

lo-16.3

SnO

10-8.94

-8.14 -7.88 -7.52 -6.85

lo 21.5

-5.08

-17.6

-4.14 -4.02

Inl.OH)3

Sn(OH)1.5C10.5 uO5

14

A2S

15

HgS

10 368

16 17

Cu2O

10-619

18 19

CU3jPO4)2

10-4.8

Cu(OH)1.5C105

10-3580

AgCl Hg2C12 HgO

10-8.00 10-712

20 21 22 23 24 25 26

pe-value

10337 CuCn 10 233 CuO 10 8.16 10 -224

Ag3PO4 Ag2O

10+0.96

02 (Activity = 0.2)

1). This compound is stable in contact with Cu and has been reported as a corrosion product on copper-rich dental amalgams (Marshall et a L, 1982). In Fig. 4, the pe-values [cf Eq. (6)] of all the copper compounds except that of Cu(II) lactate given in Table 4 are plotted against pH. It is seen that CU20, AG < 0 above pH = 4, has the lowest pe-values from pH 10 to 4, and that below pH = 4, CuCl determines the corrosion of copper (AG < 0 for formation of CuCl when the chloride concentration is higher than 0.015 mol/L, and the Cu activity is equal to unity). The formation of CuCl below pH = 4 is in agreement with the results of Sutow et al (1991), who found CuCl at pH = 1 and a chloride concentration not less than 0.068 in crevices. This means that the pe-values of Cu2O and CuCl, i.e., the redox couple Cu(s)/Cu+, determines the corrosion of copper. From Fig. lit is seen that there are five Cu(II) compounds (Nos.14 to 18, Table 4 and Fig. 1) within a small range of pevalues. (The curves are shown in the region where AG < 0 and the Cu activity is equal to unity.) In Fig. 4 it is seen that they are so close together that the composition of the solvent will probably be the factor determining which one of them will be formed. In synthetic saliva (Tani and Zucci, 1967), basic Cu(II) carbonate should be formed after oxidation of Cu2O. Of the

-12.5

+2.61 +3.26

+3.95 +4.05 +4.53 +5.51 +6.28 +10.3 +11.3 +14.5 +15.3

Cu(II)compoundsshownin Fig. 4,onlyCuCl2 3Cu(OH)2, which canalsobewrittenCu(OH)15 5 (No. 17, Table 4 and Fig. 1), has been found in Ringer's solution at 37TC by Marshall et al. (1982), who found that green CuCI23Cu(OH)2 covered the layer of red Cu20 on the surfaces of copper-rich dental amalgam specimens. Sutow et a L. (1991) alsofound that CuCl, formed in crevice corrosion at pH < 4 at high chloride concentration, could be oxidized to CuCl2 3Cu(OH)2 in moist air. But when the

specimens were properly stored and analyzed immediately after the experiment was ended, only CuCl was formed. These results are in agreement with the present calculations of AG demonstrating that it is always possible to obtain all the proposed Cu(II) compounds (Fig. 4) from both CuCl and CU20 by means of oxidation by moist air. In Fig. 4, the comparison of Cu(I) and Cu(II) compounds was carried out with the Cu activity equal to unity. This was done due to the lack of knowledge of the values of the activity factor at the start of the corrosion. When the amalgam surface is depleted of Cu, 'cu [Eq. (5)] decreases, and, since the shift of the line Cu(s)/Cu+ is not equal to that of Cu(s)/Cu2+, pe of Cu20 will be shifted closer to pe of, e.g., Cu(OH)1 5C105 (No. 17, Table 4 and Fig. 1). But before equal pe-

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

42

Olsson et al.

J Dent Res 73(l) 1994

values of these two compounds were obtained (pe = 4.6 atfcu = 0.0005), a pe-value of 3.84 f or AgSCN (No. 19, Table 4 and Fig. 1) will be reached. (The activity shift of the line Ag(s)/ Ag+ is negligible.) Therefore, the redox couple Cu(s)/Cu+ determines the corrosion of Cu and Cu(OH)15C105 can be formed only through air oxidation of Cu2O or CuCl in solutions such as saliva. Fig. 1 shows that the pe-value of AgSCN (No. 19, Table 4 and Fig. 1) is very close to that of CuCl (No. 20, Table 4 and Fig. 1). From Table 4 it can be seen that the difference in their pevalues is only 0.16. Fig. 4 shows the pe-f unction of AgSCN, and it is obvious that it is very close to that of CuCl. Therefore, below pH = 4, it is expected that a mixture of CuCl and AgSCN will be formed. AgSCN is a white precipitate insoluble in nitric acid (Partington, 1947). It is therefore stable even in crevices and should be formed in solutions containing thiocyanate. However, AgSCN has not been identified among the corrosion

products of dental amalgam. In amalgam containing Zn, the Zn content is low, and the corrosion of Zn is of minor importance. As can be seen in Fig. l and Table 4, ZnCO3 (No. 1) and Zn3(PO4)2 (No. 2) and, in Fig. 2 and Table 5, ZnS (No. 1) can be expected to be formed. The corrosion of dental amalgam in the Fusayama solution is shown in Fig. 2. It can be seen that the sulfide content of the solution changes the corrosion pattern considerably compared with that in Fig. 1. Of special interest is the region around the pevalues of the Sn compounds, where all the sulfides except that of Zn are present. All pe-functions of the sulfides vary with pH in the same way, and their curves are therefore parallel to each other. This means that the sulfide with the most negative pevalue determines the corrosion of amalgam at all values of pH. In dental amalgams containing indium, the corrosion leads to the formation of In S2 (No. 6, Table 5 and Fig. 2). In conventional dental amalgam, the corrosion is determined by SnS or Cu2S (Nos. 9 and 10, Table S and Fig. 2). Since the activity shift of the pe-values [Eq. (5)] is different for Sn(II) and Cu(I), and since the difference in their pe-values is only 0.36, the activity of Sn and Cu in the amalgam surface at the moment of corrosion is the determining factor for the formation of the corrosion product, i.e., either SnS or Cu2S can be formed. In high-Cu amalgam, Cu2S (No. 10, Table S and Fig. 2) is most likely to be formed. It is also seen that the formation of CuS, Ag2S, and HgS (Nos. 13, 14, and 15, Table 5 and Fig. 2) is not possible when amalgam without any electron-conducting contact with other metallic materials is concerned. In conclusion, it was found that, by means of log(a,/aedf)-pe diagrams, it is possible to interpret the corrosion pattern of dental amalgam. All the compounds reported in experimental studies were confirmed, and the conditionsfor their formation were established. Other sparingly soluble compounds which might be expected to form can now be excluded. It was also found that CuSCN and, at pH < 4, AgSCN could not be excluded, and therefore these compounds may exist, although they have not been identified. Most important, however, corrosion products containing mercury cannot be formed on

amalgam restorations with no metallic contact with other materials. This is in agreement with the opinions expressed by

Gross and Harrison (1989).

Acknowledgment We are grateful to Professor PhD Staffan Sjoberg, Department of Inorganic Chemistry, University of Umea, for valuable advice and fruitful discussions.

References AbrahamJE, Svare CW, Frank CW (1984). The effect of dental amalgam restorations on blood mercury levels.] Dent Res 63:71-73.

Berglund A, Pohl L, Olsson S, Bergman M (1988). Determination of the rate of release of intra-oral mercury vapor from amalgam.J Dent

Res 67:1235-1242.

Berglund A (1990). Estimation by a 24-hour study of the daily dose of intra-oral mercury vapor inhaled after release from dental amalgam.J Dent Res 69:1646-1651. Berglund A (1992). Release of mercury vapor from dental amalgam (dissertation). Swed DentJ (Suppl 85):1-95. Berglund A (1993). An in vitro and in vivo study of the release of mercury vapor from different types of amalgam alloys.JDent Res 72:939-946. Bergman M (1992). Release and uptake of mercury from dental amalgam. Proceedings from a State of the art conference on potential biological consequences of mercury released f rom dental amalgam, Apr 9-10,1992. Stockholm: Swedish Medical Research Council. Donaldson JD, Moser W, Simpson WB (1963). Basic tin(II) chloride.J Chem Soc (London):1727-1731. Espevik S (1977). In vitro corrosion of dental amalgams with different Cu content. ScandJ Dent Res 85:631-636. Gross MJ, HarrisonjA (1989).Someelectrochemicalfeaturesof the in vivo corrosion of dental amalgam.JAppl Electrochem 19301-310. Holland GA, Asgar K (1974). Some effects on the phases of amalgam induced by corrosion.J Dent Res 53:1245-1254. Jensen SJ (1982). Corrosion products of dental amalgam. ScandJ Dent Res 90:239-242. Lin J-HC, Marshall GW, Marshall SJ (1983a). Microstructures of Curich amalgams after corrosion.J Dent Res 62:112-115. Lin J-HC, Marshall GW Jr, Marshall SJ (1983b). Corrosion product formation sequence on Cu-rich amalgams in various solutions.J Biomed Mater Res 17:913-920. Mackert JR Jr (1991). Dental amalgam and mercury.J Am Dent Assoc

122(August):54-61. Marek M (1983). The corrosion of dental materials. In: Treatise on materials science and technology. Vol. 23. Scully JC, editor. London: Academic Press, pp. 331-394. Marshall SJ, Marshall GW Jr (1980). Sn4(OH) 6C2 and SnO corrosion products of amalgams.J Dent Res 59:820-823. Marshall GW Jr, Jackson BL, Marshall SJ (1980). Copper-rich and conventional amalgam restorations after clinical use.J Am Dent Assoc 100:43-47.

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.

j Dent Res 73(l) 1994

Electrochemical Corrosion of Dental Amalgam

MarshallSJ,LinJ-HC,MarshallGWJr(1982).Cu2OandCuCl2 3Cu(OH)2 corrosion products on copper rich dental amalgams. J Biomed Mater Res 16:81-85. Marshall GW Jr, Marshall SJ, Letzel H, Vrijhoef MMA (1987). Microstructures of Cu-rich amalgam restorations with moderate clinical deterioration. Dent Mater 3:135-143. Marshall SJ, Marshall GW Jr (1992). Dental amalgam. The materials. Adv Dent Res 6:94-99. Mateer RS, Reitz CD (1970). Corrosion of amalgam restorations.J Dent Res 49:399-407. Moberg L-E, Oden A (1985). The microstructure of corroded amalgams. Acta Odontol Scand 43:179-190. Olsson S, Berglund A, Pohl L, Bergman M (1989). Model of mercury vapor transport from amalgam restorations in the oral cavity.J Dent Res 68:504-508. Olsson S, Bergman M (1992). Daily dose calculations from measurements of intra-oral mercury vapor.J Dent Res 71:414-423. Otani H, Jesser WA, Wilsdorf HGF (1973). The in vivo and in vitro corrosion products of dental amalgams.J Biomed MaterRes 7:523539. Ott KHR, Loh F, Kroncke A, Schaller K-H, Valentin H, Weltle D (1984). Zur Quecksilberbelastung durch Amalgamfullungen. Dtsch Zahndrztl Z 39:199-205. PartingtonJR (1947). General and inorganic chemistry for university

students. London: MacMillan and Co., Limited.

43

Patterson JE, Weissberg BG, Dennison PJ (1985). Mercury in human

breath from dental amalgams. Bull Envir Contam Toxicol 34:459468.

Sarkar NK, Marshall GW, Moser JB, Greener EH (1975). In vivo and in vitrocorrosionproductsof dentalamalgamjDentRes54:1031-1038. Sillen LG (1952). Redox diagrams.J Chem Ed 29:600-608. Sillen LG, Martell AE (1964). Stability constants of metal-ion complexes. Spec pub No. 17. London: The Chemical Society. Sutow EJ, Jones DW, Hall GC, Owen CG (1991). Crevice corrosion products of dental amalgam.J Dent Res 70:1082-1087. Svare CW, Peterson LC, ReinhardtJW, Boyer DB, Frank CW, Gay DD, Cox RD (1981). The effect of dental amalgams on mercury levels in expired air.J Dent Res 60:1668-1671. Tani G, Zucci F (1967). Electrochemical valuation of the corrosion resistance of commonly used metals in dental prosthesis. Minerva Stomatol 16:710-713. Vimy MJ, Lorscheider FL (1985a). Intra-oral air mercury released from dental amalgam.J Dent Res 64:1069-1071. Vimy MJ, Lorscheider FL (1985b). Serial measurements of intra-oral air mercury: Estimation of daily dose from dental amalgam.JDent Res 64:1072-1075. Wald FV, Cocks FH (1971). Investigation of copper-manganese-nickel alloys for dental purposes.J Dent Res 50:48-59. WHO (1991). Environmental health criteria 118. Inorganic mercury. Geneva: World Health Organization.

Downloaded from jdr.sagepub.com by guest on July 12, 2011 For personal use only. No other uses without permission.