journal of dentistry 37 (2009) 366–373

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/jden

Penetration of amalgam constituents into dentine

Johannes D. Scholtanus a,*, Mutlu Özcan a, Marie-Charlotte D.N.J.M. Huysmans b

a
University Medical Center Groningen, Center for Dentistry and Oral Hygiene, University of Groningen, Groningen, The Netherlands
b
Department of Preventive and Restorative Dentistry, Radboud University Nijmegen Medical Center, Nijmegen, The Netherlands

article info abstract

Article history: Objectives: Amalgam restorations are replaced by adhesively placed composite resin
Received 6 September 2008 restorations at an increasing rate. After the removal of amalgam dentine often shows
Received in revised form marked dark discoloration that is attributed to the penetration of corrosion products from
8 January 2009 overlying amalgams. It is questioned whether penetration of metals into dentine affects the
Accepted 17 January 2009 dentine as a substrate for adhesive procedures. This study has been performed to clarify the
origin of dark discoloration of dentine by metals from amalgam with special regards to
corrosion products.
Keywords: Methods: A review of the literature has been performed using Medline database. As key-
Amalgam words dentine and amalgam, subsequently combined with penetration, interface, crevice, inter-
Dentine action, corrosion, were used. This was followed up by extensive hand search using reference
Discoloration lists of relevant articles.
Corrosion Results: Data in the literature have been gathered from extracted amalgam filled teeth and
Penetration from artificially aged amalgam filled teeth. Corrosion studies have been performed in vivo
Metals aged teeth as well as in vitro. Sn is the main element, followed by Zn and Cu, that is
Interaction consistently found in dentine underneath amalgam, as well as in amalgam corrosion
products and in marginal seal deposits. Penetration of elements from amalgam has only
been observed in discolored and in demineralised dentine.
Conclusions: Darkly discolored dentine as found underneath amalgam restorations contains
amalgam corrosion products and is demineralised. Therefore it must be considered a
different substrate for clinical procedures than sound dentine.
# 2009 Elsevier Ltd. All rights reserved.

1. Introduction becoming more and more common in case of replacement

Since its introduction in the nineteenth century, generations
of patients have profited from amalgam, being applied
according to the classic macro-retentive principles of G.V.
Black. After a certain lifespan, however, many amalgam
restorations need to be replaced or repaired as they are
exposed to the physico-chemical challenges of intra-oral
conditions.1 In modern dentistry, the use of composite resin
materials, applied by means of adhesive techniques, is
of amalgam fillings. For the success of such adhesive
replacement restorations reliable adhesion to remaining tooth
tissue is of major importance. In this respect, some problems
arise. Firstly, the dentine that is exposed by amalgam removal,
is in fact the dentine that is left by previous excavation
procedures. Hence, some of this dentine could have under-
gone changes by the caries process (demineralisation, remi-
neralisation, sclerosis of dentinal tubuli). Depending on the
extension of the lesion and the extent of excavation

* Corresponding author at: University Medical Center Groningen, Center for Dentistry and Oral Hygiene, P.O. Box 196, 9700 AD Groningen,
The Netherlands. Tel.: +31 50 3638374; fax: +31 50 363 2696.
E-mail address: J.D.Scholtanus@med.umcg.nl (J.D. Scholtanus).
0300-5712/$ – see front matter # 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jdent.2009.01.009
 journal of dentistry 37 (2009) 366–373
procedure, this dentine is different from normal dentine. It is
known that dark dentine discoloration can be caused by the
arresting of carious lesions. The origin of this staining is not
clear yet, but it has been associated with Maillard reactions.2
Several studies have shown lower bond strengths to caries
affected and caries infected dentine.3–7
Furthermore, it is a very common observation that dentine
underneath previous amalgam restorations shows extensive
black discoloration (Fig. 1). Already in the late nineteenth and
early twentieth century, it has been suggested that dentine
staining should be caused by the penetration of Ag and Hg ions
from the overlying amalgam and deposition of metallic
sulfides.8 In terms of adhesive properties, it is not yet
sufficiently known what the effect of this discoloration is.
Only very recently Harnirattisai et al. showed that bond
strengths to dark dentine after amalgam removal are lower
than that to normal dentine.9 From a clinical point of view, it is
difficult to decide if the staining is caused by the amalgam or
by the previous caries process or both. As a consequence, it is
hard to determine whether and until what degree stained
dentine should be removed in order to get a reliable substrate
for any kind of adhesive restorative treatment.
In an attempt to throw some light on the role of products
released from amalgam in the discoloration of dentine a
review of the literature has been performed. The main goal of
the study was to determine the chemical composition and the
rate of penetration of the substances responsible for dentine
discoloration so far as they are derived from amalgam
restorations.
2. Materials and methods
A literature search has been performed using PubMed search
engine. The primary goal was finding articles on the presence
of amalgam constituents in dentine. Subsequently the search
was directed towards literature on the pathways the amalgam
ions might be available for penetration, i.e. the marginal gap
between amalgam and dentine and the corrosion products it
contains. As keywords different combinations of the words
dentine and amalgam, subsequently combined with penetration,
interface, crevice, interaction, corrosion, were used. Initially only
articles in English language have been selected based on the
Fig. 1 – Darkly discolored dentine just after the removal of
an old amalgam restoration.
367
titles and the abstracts of the articles. Subsequently, the
literature search has been extended by scanning the refer-
ences and the related links of selected articles on their
relevance.
3. Results
3.1. Pathways of transport of metal ions into and along
dentine
Theoretically dentine underneath amalgam restorations can
be exposed to a variety of ions and molecules originating from
amalgam, oral fluids and pulpal liquid. Moreover, caries
lesions, such as present before the application of amalgam
restorations, inevitably have resulted in changes in the
dentine in terms of demineralisation, remineralisation and
sclerosis of dentinal tubuli.
Amalgam restorations do not chemically adhere to cavity
walls. The result is a marginal micro gap or crevice between
amalgam and cavity walls, facilitating transport of fluids, ions,
molecules, and possibly bacteria and their toxins.10 Because of
the difference in thermal expansion coefficients of amalgam
and tooth structures, intra-oral temperature changes result in
intermitting opening and closing of the gap thus creating
inward and outward transport of fluid (percolation) along the
amalgam–tooth interface.11–13 This way, dentine is easily
exposed to the components of saliva, food, drinks, and
products from bacterial metabolism.
Exposure of amalgam fillings to the aggressive oral
environment results in a complex of physical and chemical
challenges that causes alterations in structure and properties
e.g. time, temperature changes, mechanical forces, chemical
interactions with oral fluids and galvanic phenomena. All
these factors act continuously or intermittingly, and their
effect can be complex.14 One of the major outcomes of these
interactions is the release of corrosion products from the
amalgam restorations. On one hand corrosion is regarded as
an adverse phenomenon. Corrosion causes structural changes
inside the amalgam and is regarded as the main reason for
marginal fractures at the outline of amalgam restorations.15
On the other hand, corrosion is regarded as beneficial as it is
supposed to be the main factor in reducing marginal leakage
over time by deposition of amalgam corrosion products in the
marginal gap.16
In the fluid filled marginal gap amalgam constituents in
ionic form might be available for deposition of complexes
within the gap, as well as for diffusion into dentine and
formation of salts within the dentinal structure.
3.2. Penetration of amalgam elements into dentine
Several authors tried to elucidate the origin of the darkly
discolored dentine, by means of different analytical techni-
ques. As early as in 1881 Dwinelle17 reported black staining in
apparently previously soft dentine that seemed to be
‘‘fossilized’’ and he supposed this was caused by uptake of
oxides of Ag or Hg from the overlying amalgam into dentine.
Applebaum8 attributed the dentine pigmentation to the
deposition of mercuric sulfides, originating from contact of
368 journal of dentistry 37 (2009) 366–373
amalgam with S in dentinal fluid, but he could not prove this.
Within dentinal tubuli and in collateral branches black
granules could be demonstrated. This is in line with the
observation by Massler and Barber,18 who saw black clump-
like filamentous projections in dentinal tubuli under high
resolution, which could not be removed by acid treatment. In
most of the 300 amalgam filled teeth studied by Massler and
Barber,18 discoloration was visible. Both black and brown
types of discoloration could appear simultaneously in the
same teeth. It was stated that brown staining was caused by
caries, and could be easily removed by hand instruments. Dark
greyish black dentine was attributed to diffusion of amalgam
constituents into dentine; it was not necessarily soft and was
radiopaque. Brown discoloration could be bleached by hydro-
gen peroxide treatment, but the black stain remained.
Spectrographic analysis revealed large quantities of Hg and
smaller quantities of Ag, Sn, Zn and Cu. In an in vitro test, it
was made plausible that greyish black staining consisted of
metal sulfides. Typically Sn was the element that was found in
dentine in all dentine penetration studies.9,18–25 Mateer and
Reitz21 detected low Hg and Ag content in blackened dentine,
but the Sn content in dentine adjacent to amalgam was about
as high as it is in the amalgam itself. A number of authors also
demonstrated high concentrations of Sn and Zn in electron
microprobe studies. 18,19,22–25 However, in contrast to the
former findings, they did not detect Ag and Hg in dentine.
Kurosaki and Fusayama22 and Halse24 concluded that no
metal ions had penetrated into non-discolored dentine. Only
in dark discolored dentine they found penetration of Zn and
Sn. Because Sn produces black sulfides and Zn sulfides are
white, it was concluded by Kurosaki and Fusayama22 that the
black staining was caused by Sn sulfides. In a very recent study
again Sn was found in various amounts and traces of Zn and
Cu in discolored dentine.9
Table 1 summarizes the studies of dentine discoloration by
amalgam ingredients and the chemical elements that are
supposed to be responsible for discoloration.
Kurosaki and Fusayama22 demonstrated that deminera-
lised dentine is susceptible to the uptake of Sn and Zn. Sn and
Zn penetrated the softened dentine but never penetrated the
normal dentine. In an animal study, greyish discoloration was
exclusively seen in demineralised dentine. No discoloration
was found in non-demineralised dentine. Sn seemed to
Table 1 – Chronological list of publications on amalgam ingredients found in discolored dentine underneath amalgam
restorations.
Authors Year of publication
Substantial
Massler and Barber 1953 Sn, Zn, Cu, Ag, Hg
Söremark et al. 1968 Sn, Zn, Hg
Wei and Ingram 1969 Sn
Mateer and Reitz 1972 Sn
Kurosaki and Fusayama 1973 Sn,
Van der Linden and Van Aken 1973 Sn,
Halse 1975 Sn,
Hals and Halse 1975 Sn,
Harnirattisai et al. 2007 Sn,
diffuse throughout the outer demineralised layer and accu-
mulated into the inner demineralised layer over time. In
extracted human amalgam filled teeth, Sn was present in the
dentine layer immediately adjacent to the amalgam restora-
tion, so as in the outer dentine layer. The authors explain this
difference by the excavation procedure in human carious
teeth by which the outer superficial layer of softened dentine
is removed and (part of) the inner layer is left in place.
These findings are in agreement with those of other
studies23–25 that showed large amounts of Sn and Zn in
dentine (without Cu, Ag, Hg) which were strictly related to
discolored and apparently demineralised dentine. No Zn was
found in dentine without Sn and vica versa.23 In vitro residual
carious dentine underneath freshly packed amalgam is also
susceptible to discoloration by tea and chlorhexidine.26
Underneath amalgam restorations occasionally radiopa-
que zones are visible on radiographs.8,18,23 These radiopaque
zones appear to be soft and can easily be removed during
excavation.23 Because Ca is low, hypermineralisation is not
the explanation for the radiopacity. In this demineralised
dentine only Sn and Zn have been demonstrated. Apparently
Zn and Sn filled the places where Ca was depleted. By means of
the use of a caries-disclosing dye during excavation proce-
dures Rudolphy et al.27 demonstrated that the radiopaque
zone almost entirely consisted of irreversibly demineralised
dentine. In another study, it was shown that radiopacities on
bite-wing radiographs had not been changed in dimensions
over a period of 6 years; therefore, it was concluded that these
radiopacities correspond with non-progressing caries
lesions.28
It should be kept in mind that the studies which are
referred to in this paragraph are all performed on teeth
restored with conventional amalgam types, except the recent
one by Harnirattisai et al.9 For high-copper amalgams, no
further studies on the penetration of elements into dentine
and only data from studies on marginal seals and corrosion
phenomena of high-copper amalgams are available. From the
referred studies only qualitative and semi-quantitative data
are generated as they generally are based on electron
microprobe analysis and characteristic X-ray image analysis.
The amalgam constituents that penetrate into the dentine
can theoretically only be derived directly from the amalgam
restoration or from the corrosion products that are present in
Metals in dentine Analytical methods
Traces
Spectrographic analysis
Thermal neutron activation
+ g-ray spectrometric analysis
Electron microprobe
Ag, Hg Electron microprobe
Zn Electron microprobe
Zn Electron microprobe
Zn Electron microprobe
Zn Electron microprobe
Zn, Cu Energy dispersive spectroscopy
 journal of dentistry 37 (2009) 366–373 369

Fig. 2 – Black discolored dentine after the removal of an old amalgam restoration in tooth 17 before (left) and after excavation
procedure (right).

the marginal gap between amalgam and cavity walls. There-
fore in the next section the data on marginal seal composition
are described.
3.3. Composition of marginal seal
Frequently upon removal of old amalgam, a black granular
layer can be seen, which is rather loosely attached to the
underlying dentine (Fig. 2) and can rather easily be removed by
excavation burs under gentle pressure. Remaining hard
dentine still shows considerable black speckled pigmenta-
tions.
In Table 2 the results of studies on marginal seal products
and on tooth–amalgam interface are listed. In these stu-
dies20,24,29–36, Sn was consistently found in marginal gaps. In a
number of these studies, Zn was also identified. It was
supposed that Zn was released from Zn containing amalgams
as well as from liners and bases.23,32
Cu was detected not only in marginal seals of high and
medium-Cu amalgams, but also in some seals of low-Cu
amalgams.32 Sutow et al.35 identified SnO, Sn4(OH)6Cl and
Cu2O in conventional amalgam interfaces, and in high-Cu
amalgams, they found CuCl and Cu2O. Also Ca and P have been
detected as a component of marginal seals in some studies.
The origin of Ca and P was supposed to be the saliva
(penetrating the marginal gap), cement bases and the dentine
underlying the amalgam as a result of low pH in the marginal
gap.32
In order to prevent interaction of underlying or adjacent
structures with the measurements, Grossman et al.32 and
Witcomb et al.33 used a replica extraction technique on
marginal seal products. Apart from the already mentioned
elements, they also measured smaller amounts of Cu, Ag, P, Cl,
and S and traces of Fe, K, and Al. Another in vitro study
indicated that after 3 months and 1 year, Sn is the essential
element in marginal seals around both low- and high-copper
amalgams.36
From the above-mentioned studies, it is evident that in
marginal seal products underneath conventional amalgams,
Sn is present in most copious amounts, followed by Zn. Cu is
found in marginal seals of high-copper amalgams. The reason
why these elements were found and Ag and Hg were found
only sporadically might be explained by corrosion phenomena
of respective amalgam alloys.

Table 2 – Chronological list of publications on amalgam elements detected in marginal gaps or amalgam–tooth interfaces
and the respective chemical elements detected in vivo and in vitro.
Authors Year of Metals in marginal gap Other elements in Analytical methods
publication marginal gap
Substantial Traces

Wei and Ingram 1969 Sn Electron microprobe

Holland and Asgar 1974 Sn P, Cl, S Electron dispersive spectroscopy
Halse 1975 Sn, Zn Electron microprobe
Sarkar et al. 1981 Sn, Zn (low-copper Ca P Electron dispersive spectroscopy
amalgams), Sn, Zn, Cu,
Hg (high-copper
amalgams)
Port and Marshall 1985 Sn Ca Electron dispersive spectroscopy
Grossman et al. 1986 Sn, Zn Cu, Ca, Fe P, Cl Electron dispersive spectroscopy
Witcomb et al. 1987 Sn, Zn Cu, Ag Fe, K, P, Cl, S, Electron dispersive spectroscopy
Al, Ca
Weiland et al. 1990 Sn, Zn Ca, Mg, Na P, O Electron dispersive spectroscopy
Sutow et al. 1991 Sn, Cu (low copper Electron dispersive spectroscopy
amalgams), Cu
(high-copper amalgams)
Grossman et al. 1995 Sn, Zn Ca P, Cl Electron dispersive spectroscopy
370 journal of dentistry 37 (2009) 366–373
3.4. Corrosion products of dental amalgam
Corrosion of amalgam is a destructive phenomenon which is
not only limited to the external surfaces or amalgam–tooth
interfaces, where it is evident, but also affects the bulk of the
restoration. Over a period of time, corrosion progresses from
interfaces towards the center of the restoration, reducing
hardness and embrittling the matrix structure by molecular
disintegration of the g2 phase.14,21,29,31,37 Chloride ions in
saliva, food and drinks play an essential role in the corrosion
process and the deposition of metallic salts.
In conventional amalgams the result of trituration of alloy
and mercury and subsequent setting is essentially a matrix
structure of Ag2Hg3 (g1 phase) and Sn8Hg (g2 phase) surround-
ing the original alloy particles containing Ag3Sn (g phase). The
g2 phase forms an interconnecting network throughout the
bulk of the amalgam. It is the most easily corroded phase in
conventional amalgams and it is also the weakest phase in
mechanical aspect.15,38,39 The Ag3Sn (g) and Ag2Hg3 (g1) phase
are electrolytically stable and do not dissociate. In the complex
interactions the g2 phase is selectively attacked by the
corrosion process, resulting in oxidation of Sn.40,41. The
protons that are released by this process are responsible for
lowering of the pH in the corrosion crevices, so corrosion will
promote in an acid milieu. Sn migrates from amalgam to the
interface to form poorly soluble deposits.
In conventional amalgam restorations retrieved from
patients, corrosion products have been identified as a variety
of Sn salts: tin-oxides, tin sulfide, tin chloride, tin oxychloride,
tin hydroxide, tin hydroxychloride.39,42–45 Also Ca was found
in corrosion products.31 The Hg that is released from the g2
phase stays in its molecular form. This metallic mercury is not
dissolved in the surrounding electrolyte, but diffuses into the
Ag3Sn phase and will further react to g1 and g2.15,41,43 This
phenomenon causes mercuroscopic expansion and plays an
important role in the origin of marginal defects.15
Because the net effect of corrosion process is decrease or
depletion of g2 phase in old amalgam restorations little or no
remaining g2 phase might be found.29,31,43 So it is not
surprising that Sn has been found as the main metal that is
released by corrosion processes in conventional amalgams.
Modern high-copper amalgams contain little or no Sn8Hg
phase, as a consequence these amalgams are also called non-
gamma-2 amalgams. Instead of g2 phase, a copper–tin product
is formed as a result of the high-copper content: Cu6Sn5, also
called h0 phase.46,47 In these amalgams, introduced by Innes
and Youdelis in 1963, the h0 phase is the most corrosion prone
phase.48 Because no Hg is released by the corrosion process in
high-copper amalgams, mercuroscopic expansion is pre-
vented, thus reducing marginal fracture potential.
Corrosion of Cu6Sn5 involves oxidation of Sn followed by
oxidation of Cu. Like in conventional amalgams oxidation of
Sn results in rather insoluble corrosion products but in smaller
amounts than in conventional amalgams.44 Subsequent
oxidation of Cu may lead to disintegration of h0 phase and
relatively soluble Cu complexes, which may leach out in the
liquid environment.37,49,50 On Cu-rich amalgams in extracted
and exfoliated teeth as well as in in vitro corrosion experi-
ments complexes of Ca–Sn–P–Cl and crystalline products
containing Sn or SnCl have been found on external surface, at
amalgam–tooth interface and within the restoration. No Cu
was shown as Cu complexes were assumed to be leached out
into the liquid environment.51 However, in contrast to the in
vivo observations, Cu containing corrosion products are
detected and classified as Cu2O and CuCl2!3Cu(OH)2 in vitro.
The latter compound has a typical green color. It was also
present on retrieved amalgam restorations that had been kept
in storing solutions over time.49,50
From Zn containing amalgams (conventional as well as high-
copper amalgams) Zn is the most easily corroded element.
Corrosion of Zn is supposed to delay corrosion of Sn8Hg (g2
phase) and Cu6Sn5 (h0 phase).52,53 Zn has been identified as the
major corrosion product released in saliva and ZnSn(OH)6 has
been detected on Zn-containing amalgams 49,50,51,54. On both
Zn-free conventional and Zn-free Cu-rich amalgams CuS-
n(OH)6 was found.54 The ZnSn(OH)6 appeared after a few days,
while CuSn(OH)6 appeared later, probably due to the initial
high dissolution rate of Zn compared to Cu. The small
amounts of Zn in the zinc-containing amalgams play a
sacrificing role in corrosion processes in dental amalgam,
thereby protecting the g2 phase as well as the h0 phase from
early attack.53 Quick dissolution of Zn ions has been found
from conventional and high-Cu amalgam specimens in lactic
acid solutions, but no Zn was detected in adherent corrosion
products. Main corrosion products in all tested amalgams
were based on Sn and Cu.55
From the above-mentioned studies, it is evident that Sn, Zn
and Cu are the main elements that are involved in amalgam
corrosion processes and might be available for penetration
into dentine. The Sn containing products are rather insoluble,
and might be integrated in the marginal seal and precipitates.
Zn is easily dissolved and tends to migrate into the environ-
ment. Cu is found in soluble complexes (chlorides and
sulfides). Still Sn, Zn and Cu might be available in ionized
forms in the liquid surroundings of the amalgam restorations,
depending on the composition of the amalgam. No Ag and Hg
could be demonstrated in corrosion products.
4. Discussion
Different aspects of corrosion of dental amalgams have been
studied. The results of these studies cannot readily be
compared, because of differences in analytical methods,
objectives, duration of observation periods and number of
specimens studied. In some studies, elements have been
searched for selectively, thereby excluding other ele-
ments.20,29,34
For detection of amalgam constituents, different analytical
methods have been used. With the exception of two earlier
ones, the reviewed studies of elements in dentine and in
marginal gaps used electron microprobe and electron dis-
persive spectroscopy (EDS) for their elemental analysis. Both
techniques are based on the detection of specific X-rays that
are emitted by excitation of samples by bombardment with
electron beams. Electron microprobe renders X-ray images
and graphic representations, demonstrating the presence and
the relative distribution of elements within a solid material.
Chemical composition is determined by comparing the
intensities of characteristic X-rays from the sample material
 journal of dentistry 37 (2009) 366–373
with intensities from known composition. A limitation is the
recording technique. Trace elements cannot be detected.
Results of electron microprobe are qualitative or semi-
quantitative.56 Energy dispersive spectroscopic methods
produce more accurate data, however, in a semi-quantitative
way. EDS detects characteristic X-rays emitted by elements
and produces graphs with peaks of specific energies of
elements. This technique is suitable for determining the
relative composition of solid materials Limitations of this
method are detector resolution and possible overlaps of peaks
causing difficulties in interpretation.56 Because above men-
tioned techniques are applied under vacuum, samples need to
be dried, causing loss of the water component in dentine
samples. In dentine samples the inherent uneven density and
structural heterogeneity might produce unfavourable inter-
ferences in analysis. New technologies may result in more
accurate results. However in this review, the variations
between studies perhaps not for the substantial elements
but for the trace elements noted could be attributed to the
methodological differences as well as their accuracy.
It is obvious that corrosion of amalgam plays a major role in
the efflux of metal ions. It is agreed that the g2 phase (Sn8Hg) in
conventional amalgams and the h0 phase (Cu6Sn5) in high-
copper amalgams are the phases that are most easily
corroded. Corrosive attack of Sn8Hg renders Sn that is oxidized
to form oxides and hydroxides. In contrast to Sn, the Hg is
released in metallic form, and it is supposed to react further
with remaining g phase (Ag3Sn) to new reaction products
inside the amalgam. This might explain that Hg is not found in
corrosion products of amalgam or in dentine.
From the Cu6Sn5 in high-copper amalgams again Sn is
released, but at a smaller rate because of the higher corrosion
resistance compared to the g2 phase. So it is not surprising that
the same Sn oxides and hydroxides are found as on
conventional amalgams in smaller quantities. Cu is available
to form relatively soluble complexes of chlorides and sulfides
on the amalgam. Zn, however, is the most easily corroded
element from Zn-containing amalgams. The Zn forms soluble
complexes on amalgam and it is easily dissolved in the liquid
environment. Zn has also been shown to be released from
bases and liners.
Most authors studied the characteristics of corrosion
products on retrieved amalgam restorations from extracted
teeth. In many of these studies it was not known how long the
fillings had survived intra-orally. In other studies fresh
amalgam restorations were subjected to artificial corrosive
attacks in vitro. In both types of studies the same corrosion
products were found, essentially products based on Sn, thus
confirming that g2 and h0 phases are prone to selective
chemical attack.
It is evident that Sn, Zn and Cu are the essential elements
that are related to corrosion processes in dental amalgam.
After insertion in cavities, amalgam is surrounded by irregular
micro crevices or marginal gaps that are filled with fluid,
originating from saliva and dentinal liquid from dentinal
tubuli. In this liquid, environment corrosion products are
formed resulting in obturation of marginal gaps.
Deposits in marginal gaps contain considerable amounts of
products formed by Sn, Zn and Cu, in accordance with what
might be expected from corrosion studies. In some earlier
371
studies also Hg and Ag have been detected, but these findings
might be attributed to methodological imperfections produ-
cing interferences by the composition of structures adjacent or
underlying the products in the shallow gap. Grossman et al.32
and Witcomb et al.33 tried to overcome this problem by the use
of a replica extraction technique for studying the composition
of the marginal seal.
Also Ca and P have been detected in considerable amounts
in marginal gap deposits. It is suggested that these elements
are released from tooth structure by the acidity in the marginal
gap. Low pH might have been caused by oxidation of Sn and/or
by acids from metabolism of microorganisms in gaps or
secondary carious lesions.
The dentine along the amalgam–tooth interface is exposed
to corrosion products in several ways: the fluid in the marginal
gap being a transport medium released by corrosion of
amalgam at the interface can form complexes but might
diffuse also into the fluid filled marginal gaps and subse-
quently into the underlying hydrophilic dentine.39 Addition-
ally, marginal seal products might dissolve depending on the
pH changes in the marginal gap. Within dentinal tubuli new
complexes can be formed. Applebaum8 and Massler and
Barber18 found clump-like black structures within tubuli
visible with light microscopy. All but one study on penetration
of elements in dentine concern conventional amalgams.
Contrary to what Dwinelle17 and Applebaum8 suggested, Sn
and Zn were found in dentine adjacent to amalgam consis-
tently. No Ag and Hg were detected in in vivo studies on
retrieved amalgam filled teeth.
Surprisingly, Söremark19 found considerable amounts of
Hg in dentine tubuli, besides Sn and Zn, within short
evaluation periods in artificially corroded amalgam restora-
tions. In animal studies Hörsted-Bindslev et al.57 and Akyüz
and Çaglar58 demonstrated penetration of Hg from amalgam
into dentine and pulp tissue. The former authors suggested
that base material can prevent Hg uptake into dentine, the
latter showed significant reduction of Hg uptake but bases
could not prevent this. Ferracane et al.59 and Okabe et al.60
demonstrated Hg release from fresh amalgam into solutions.
Release was influenced by pH of the fluids. Maybe these
phenomena can be explained by free Hg that is released from
amalgam while hardening reactions have not been completed.
It is still unclear why no Hg could be demonstrated in
dentine in other studies.
Application of cavity varnishes has been a regular treat-
ment over time. The rationale was to reduce the initial
microleakage and to prevent discoloration of tooth structures
by prevention of ion penetration. Tveit and Hals61 demon-
strated that copal varnish is no barrier for Zn ions.
Liners do not seem to be able to block electric currents and
prevent transport of Sn ions.39
Dentine cavity walls underneath amalgam restorations
must have undergone structural histological and chemical
changes due to past caries attacks. Subsequent operative
treatment is aimed at excavating soft caries infected dentine,
obturation of the cavity and remineralisation of the remaining
hard caries affected dentine. If the cavity preparation
principles of Black are applied, some parts of the dentine
are prepared until apparently sound dentine, resulting in
trauma of tubuli and their content. In central parts of the
372 journal of dentistry 37 (2009) 366–373
prepared cavity discolored caries affected dentine is left if it
feels hard when touched with an explorer tip. In very deep
carious lesions, even soft dentine can be left in place locally if
properly capped by restorative material.62
From different studies, it is apparent that demineralised
dentine is susceptible to penetration of Sn and Zn. No metal
ions were detected in non-demineralised dentine.22,23,27
Moreover Sn and Zn ions only could be shown in black
discolored dentine. So it is suggested that there is a relation-
ship between ion penetration, demineralisation and disco-
loration. This susceptibility of demineralised dentine to Sn is
supported by a study by Wei et al.63 They reported uptake of Sn
from stannous fluoride solutions in such quantities that
radiopacity was highly increased and it was suggested that Sn
ions had replaced missing calcium ions in the demineralised
layers.
Caution is needed in interpretation of the findings of Zn in
dentine underneath amalgam fillings, because Zn in dentine
does not necessarily originate from the amalgam as indicated
by Takuma et al.64 In primary carious lesions in dentine they
demonstrated an increase of Zn, especially in remineralised
surface layers. Except amalgam components also external
pigments can cause discoloration of dentine. Under freshly
packed amalgam restorations residual carious dentine can be
stained if exposed to tea and chlorhexidine in vitro.26 This
demonstrates that external pigments can penetrate via
marginal gaps into demineralised (caries affected) dentine.
Interpretation of darkly discolored dentine is complicated
by the physiological feature of dark discoloration of carious
dentine in slowly progressing and arrested caries lesions.65
The exact origin of this darkening has not been elucidated, but
it has been associated with chemical changes in dentine
comparable to Maillard reactions.2 Underneath amalgam
fillings both types of dentine discoloration can be found, from
the changes caused by the caries process as well as from the
penetration of amalgam constituents.
5. Conclusions
On the basis of the literature, it is concluded that discoloration
of dentine underneath amalgam is an indicator for the
presence of amalgam constituents. As demineralised dentine
appears susceptible to uptake of metals from amalgam, this
discoloration can be regarded as an indicator for deminer-
alised dentine. It is likely that this demineralised dentine, in
fact, is the layer of affected (i.e. reversibly demineralised)
dentine that remained after cavity preparation before applica-
tion of the amalgam restoration.
It is not known which are the effects of elements from
amalgam in dentine in relation to adhesive restorative
procedures. As several authors have shown that caries
affected dentine renders lower bond strengths to adhesive
restorations,3–7 caution should be exercised in applying
adhesive techniques on darkened dentine after amalgam
removal. Indeed Harnirattisai et al.9 found lower bond
strengths to darkly discolored dentine after amalgam removal.
Further studies will be required for developing clinical
guidelines for adhesive restorative procedures on discolored
dentine after amalgam removal.
references
1. Mjör IA, Gordan VV. Failure, repair, refurbishing and
longevity of restorations. Operative Dentistry 2002;27:528–34.
2. Kleter GA. Discoloration of dental carious lesions (a review).
Archives of Oral Biology 1998;43:629–32.
3. Nakajima M, Sano H, Burrow MF, Tagami J, Yoshiyama M,
Ebisu S, et al. Tensile bond strength and SEM evaluation of
caries-affected dentin using dentin adhesives. Journal of
Dental Research 1995;74:1679–88.
4. Nakajima M, Sano H, Urabe I, Tagami J, Pashley DH. Bond
strengths of single-bottle dentin adhesives to caries-
affected dentin. Operative Dentistry 2000;25:2–10.
5. Yoshiyama M, Tay FR, Doi J, Nishitani Y, Yamada T, Itou K,
et al. Bonding of self-etch and total-etch adhesives to
carious dentin. Journal of Dental Research 2002;81:556–60.
6. Pereira PN, Nunes MF, Miguez PA, Swift Jr EJ. Bond strengths
of a 1-step self-etching system to caries-affected and
normal dentin. Operative Dentistry 2006;31:677–81.
7. Omar H, El-Badrawy W, El-Mowafy O, Atta O, Saleem B.
Microtensile bond strength of resin composite bonded to
caries-affected dentin with three adhesives. Operative
Dentistry 2007;32:24–30.
8. Applebaum E Lymph. Channels in dentine and enamel
stained by amalgam. Journal of Dental Research 1929;9:487–
502.
9. Harnirattisai C, Senawongse P, Tagami J. Microtensile
bondstrengths of two adhesive resins to discolored dentin
after amalgam removal. Journal of Dental Research
2007;86:232–6.
10. Ben-Amar A, Cardash HS. The fluid-filed gap under
amalgam and resin composite restorations. American Journal
of Dentistry 1991;4:226–30.
11. Jodaikin A. Experimental microleakage around ageing
dental amalgam restorations: a review. Journal of Oral
Rehabilitation 1981;8:517–26.
12. Pickard HM, Gayford JJ. Leakage at the margins of amalgam
restorations. British Dental Journal 1965;119:69–77.
13. Kidd EAM. Microleakage: a review. Journal of Dentistry
1976;4:199–206.
14. Marek M. Interactions between dental amalgams and the
oral environment. Advances in Dental Research 1992;6:100–9.
15. Jörgensen KD. The mechanism of marginal fracture of
amalgam fillings. Acta Odontologica Scandinavica 1965;23:347–
89.
16. Ben-Amar A, Cardash HS, Judes H. The sealing of the tooth/
amalgam interface by corrosion products. Journal of Oral
Rehabilitation 1995;22:101–4.
17. Dwinelle WH. Meeting of the New York Odontological
Society. Dental Cosmos 1881;23:427.
18. Massler M, Barber TK. Action of amalgam on dentin. Journal
of the American Dental Association 1953;47:415–22.
19. Söremark R, Wing K, Olsson K, Goldin J. Penetration of
metallic ions from restorations into teeth. Journal of
Prosthetic Dentistry 1968;20:531–40.
20. Wei SHY, Ingram MJ. Analyses of the amalgam–tooth
interface using the electron microprobe. Journal of Dental
Research 1969;48:317–20.
21. Mateer RS, Reitz C. Galvanic degradation of amalgam
restorations. Journal of Dental Research 1972;51:1546–51.
22. Kurosaki N, Fusayama T. Penetration of elements from
amalgam into dentin. Journal of Dental Research 1973;52:309–17.
23. Van der Linden LWJ, Van Aken J. The origin of localized
increased radiopacity in the dentin. Oral Surgery
1973;35:862–71.
24. Halse A. Metals in tubules beneath amalgam fillings in
human teeth. Archives of Oral Biology 1975;20:87–8.
 journal of dentistry 37 (2009) 366–373
25. Hals E, Halse A. Electron probe microanalysis of secondary
caries lesions associated with silver amalgam fillings. Acta
Odontolgica Scandinavica 1975;33:149–60.
26. Kidd EAM, Joyston-Bechal S, Smith MM. Staining of residual
caries under freshly-packed amalgam restorations exposed
to tea/chlorhxidine in vitro. International Dental Journal
1990;40:219–24.
27. Rudolphy MP, Van Amerongen JP, Ten Cate JM.
Radiopacities in dentin under amalgam restorations. Caries
Research 1994;28:240–5.
28. Rudolphy MP, Gorter Y, Van Loveren C, Poorterman JHG,
Van Amerongen JP. Progression of radiopacities and
radiolucencies under amalgam restorations on bite-wing
radiographs. Caries Research 1997;31:19–23.
29. Holland GA, Asgar A. Some effects on the phases of
amalgam induced by corrosion. Journal of Dental Research
1974;53:1245–54.
30. Sarkar NK, Fuys RA, Schönfeld CM, Armstrong LM, Stanford
JW. The tooth–amalgam interaction. Journal of Oral
Rehabilitation 1981;8:401–11.
31. Port RM, Marshall GW. Characteristics of amalgam
restorations with variable clinical appearance. Journal of the
American Dental Association 1985;110:491–5.
32. Grossman ES, Wittcomb MJ, Jodaikin A. Elements in
marginal seals at amalgam–tooth interfaces. Journal of Dental
Research 1986;65:998–1000.
33. Witcomb MJ, Jodaikin A, Grossman ES. The use of extraction
replicas in scanning electron microscopy for the elemental
analysis of marginal seals. Journal of Biomedical Materials
Research 1987;21:955–64.
34. Weiland M, Kaiser A, Borrmann S, Klimm W. Qualitative
Randspaltuntersuchungen am Amalgamfüllungen. Zahn-
Mund-Kieferheilkunde 1990;78:19–24.
35. Sutow EJ, Jones DW, Hall GC, Owen CG. Crevice corrosion of
dental amalgam. Journal of Dental Research 1991;70:1082–7.
36. Grossman ES, Witcomb MJ, Matejka JM. Influence of
amalgam, bases, and varnish on seal composition at
restoration tooth interfaces. Journal of Prosthetic Dentistry
1995;73:290–8.
37. McTigue D, Brice C, Nanda CR, Sarkar NK. The in vivo
corrosion of dispersalloy. Journal of Oral Rehabilitation
1984;11:351–9.
38. Guthrow CE, Johnson LB, Lawless KR. Corrosion of amalgam
and its phases. Journal of Dental Research 1967;46:1372–81.
39. Mateer RS, Reitz C. Corrosion of amalgam restorations.
Journal of Dental Research 1970;49:399–407.
40. Jörgensen KD. Amalgame in der Zahnheilkunde. Karl
Hanser Verlag, Munchen; 1977, pp. 64–73.
41. Marek MI. Alterations of dental amalgam. In: Eliades G,
Eliades T, Brantley WA, Watts DC, editors. Dental materials in
vivo ageing and related phenomena. Chicago: Quintessence
Publishing; 2003. p. 61–77.
42. Sarkar NK, Greener EH. Corrosion behaviour of the tin-
containing silver-mercury phase of dental amalgam. Journal
of Dental Research 1974;53:925–32.
43. Sarkar NK, Marshall GW, Moser JB, Greener EH. In vivo and
in vitro corrosion products of dental amalgam. Journal of
Dental Research 1975;54:1031–8.
44. Marshall SJ, Marshall GW. Sn4(OH)6Cl2 and SnO corrosion
products of amalgams. Journal of Dental Research 1980;59:820–3.
373
45. Marshall GW, Jackson BL, Marshall SJ. Copper-rich and
conventional restorations after clinical use. Journal of the
American Dental Association 1980;100:43–7.
46. Mahler DB, Adey JD, Van Eysden J. Quantitative microprobe
analysis of amalgam. Journal of Dental Research 1975;54:218–
26.
47. Takatsu T, Iwaku M, Fusayama T. Structure and effects of
non-gamma-2 amalgam. Journal of Dental Research
1977;56:40–5.
48. Innes DBK, Youdelis WV. Dispersion strengthened
amalgams. Journal of the Canadian Dental Association
1963;29:587–93.
49. Marshall SJ, Lin JH, Marshall GW. Cu2O and CuCl2!3Cu(OH)2
corrosion products on copper rich dental amalgams. Journal
of Biomedical Materials Research 1982;16:81–5.
50. Lin JH, Marshall GW, Marshall SJ. Corrosion product
formation sequence on Cu-rich amalgams in various
solutions. Journal of Biomedical Research 1983;17:913–20.
51. Brune D. Corrosion of amalgams. Scandinavian Journal of
Dental Research 1981;89:506–14.
52. Johnson LB, Paffenbarger GC. The role of zinc in dental
amalgams. Journal of Dental Research 1980;59:1412–9.
53. Letzel H, Van’t Hof MA, Marshall GW, Marshall SJ. The
influence of the amalgam alloy on the survival of amalgam
restorations: a secondary analysis of multiple controlled
clinical trials. Journal of Dental Research 1997;76:1787–98.
54. Jensen SJ. Corrosion products of dental amalgam.
Scandinavian Journal of Dental Research 1982;90:239–42.
55. Johansson B, Dérand T. Corrosion of amalgams with special
regard to zinc. Scandinavian Journal of Dental Research
1983;91:320–4.
56. Hals E, Tveit AS, Tøtdal B. X-ray microanalysis of dentin: a
review. Scanning Microscopy 1988;2:357–69.
57. Hörsted-Bindslev P, Danscher GH, Hansen JC. Dentinal and
pulpal uptake of mercury from lined and unlined amalgam
restorations in minipigs. European Journal of Oral Sciences
1997;105:338–43.
58. Akyüz S, Çaglar E. Pulpal uptake of mercury from lined
amalgam restorations in guinea pigs. European Journal of Oral
Sciences 2002;110:460–3.
59. Ferracane JL, Mafian P, Cooper C, Okabe T. Time-dependent
dissolution of amalgams into saline solution. Journal of
Dental Research 1987;66:1331–5.
60. Okabe T, Elvebak B, Carrsco L, Ferracane JL, Keanini RG,
Nakajima H. Mercury release from dental amalgams into
continuously replenished liquids. Dental Materials
2003;19:38–45.
61. Tveit AB, Hals E. Penetration of ions from silicate cement
restorations into Copalite1-covered cavity walls. Acta
Odontologica Scandinavica 1978;36:15–24.
62. Massler M. Pulpal reactions to dental caries. International
Dental Journal 1967;17:441–60.
63. Wei SHY, Kaqueler JC, Massler M. Remineralization of
carious dentin. Journal of Dental Research 1968;47:381–91.
64. Takuma S, Ogiwara H, Suzuki H. Electron-probe and
electron micrsosope studies of carious dentinal lesions with
a remineralized surface layer. Caries Research 1975;9:278–85.
65. Fusayama T, Okuse K, Hosoda H. Relationship between
hardness, discoloration and microbial invasion in carious
dentin. Journal of Dental Research 1966;45:1033–46.