Former artificial saliva in different compositions (± mucin,

studies clearly examined the corrosion behavior in artificial saliva in
different compositions (± mucin, ±
urea) and Mg-0.6Ca-0.8/1.8 wt.% Zn 7,
10, 23. Here a lower
corrosion rate was found in 0.8 wt. % Zn Alloys, as well as especially the
combination of mucin and urea served as corrosion inhibitors by the forming of
a thin, homogenous corrosion layer with protective qualities 10. The presented study focused on 0.8 wt. % Zn Alloys,
as these clearly show lower and more promising corrosion rates. Maxillofacial
implants require a slow and homogenous corrosion rate in order to prevent early
implant failure and negative side effect.


The initial steep
corrosion rate, which was confirmed in previous publications could not been
observed in our experiment. It is described to occur in the first 6 hours of
the corrosion until it reaches the nobler secondary phases, which would
decelerate it in case of a small grained microstructure with a regular and not
agglomerated distribution of b-phases 10,
24, 25. In the first 5 days
of the immersion experiment, the rate of the two different electrolytes keeps
constantly stable and nearly similar until the time of electrolyte change (Fig.
2). Previously published immersion experiments with the same alloy mixture in
Hank’s balanced solution are in accordance with our results 7. Nevertheless our results for 5 day immersion
experiments in artificial saliva differ from earlier trials, which proclaimed
them to be lower than in Hank’s balanced solution, with a rate of 0.010 ±
0.004 mg/(cm2 × day) and therefore in
the range of the previously estimated tolerable corrosion rate suggested by Song
et al 10,
17. Yet our results
indicate much higher corrosion rates and thus similar to our Hank’s solution
corrosion rate. Our data were collected from three simultaneously performed
experiments. Therefore we expect them to be more accurate, since previously
published results were determined in independent experiments. Nevertheless
Hank’s solution was predicted frequently in earlier literature to exhibit a
higher corrosion, because a higher chloride concentration (>30mmol/l) is
able dissolve the upper Mg(OH)2 protective corrosion layer and
therefore lead to further corrosion enhancement 1,
3, 7, 15, 26-28. This effect can be
observed at the time of the electrolyte change, when the corrosion rate in
Hank’s solution clearly exceeds the rate in artificial saliva and AS/HS (Fig.
2) and is confirmed by our electron pictures (Fig. 3a), which show corrosion
products ripping deeply into the magnesium layer. The corrosion rate of AS/HS
exhibits the smallest increase after the electrolyte change which is in
accordance with the electron microscopy picture (Fig. 3c) showing a homogenous
corrosion process. Therefore especially strong local inward ripping corrosion
leads to rate accelerations as seen in Fig. 3a,b. for sole artificial saliva
and Hank’s balanced solution exposure.

Previous experiments with
this alloy were finished after 5 days, therefore corrosion behavior at later
time points was unknown 7. We could prove that corrosion behavior after 5 days
in Hank’s balanced solutions strongly deviates from earlier results in the first
5 days (Fig. 2). It outgoes the former rate strongly (0.6028 vs. 0.06580 mg/(cm2×day)).
Similarly the corrosion rate of artificial saliva increases, but less
distinctly after the electrolyte change (0.3000 vs. 0.1206 mg/(cm2×day)).
Whether this effect is due to the electrolyte change or a general phenomenon
needs to be further investigated, as the electrolyte change is unavoidable due
to cloudy liquid changes and thus inadequate measurements conditions. Anyway
strong corrosion acceleration has to be avoided, because it correlates with the
occurrence of negative side effects.

Interestingly previous artificial saliva incubation
seems to have a protective effect on the alloy. Thus transposition into Hank’s balanced
solutions leads to the lowest corrosion rate of all three electrolytes (0.2108 ±
0.09249) and is therefore significantly lower than in alloys, which were
exposed by Hank’s solution only. This in accordance with our electron pictures,
which show less corrosion products and fare more homogenous corrosion for this
electrolyte (Fig. 3c) in comparison to sole Hank’s solution exposure and
artificial saliva exposure (Fig. 3a,b).


The corrosion
mechanism of our results follows previous publication 7,
29. Figure 4d) proves
existence of Mg2Ca (point 1) and Mg6Ca2Zn3
(point 2) are strongly decreased in surface related areas, which are certainly
the most in Hank’s solution (Fig. 3a-c). We confirm the previously claimed
hypothesis in which electrochemical potentials follow Mg6Ca2Zn3
> a-Mg > Mg2Ca 7. Indeed Mg2Ca is dissolved in first stage
and contributes to high dissolution rates in the beginning, while being mainly
presented in grain boundaries. Regarding surface related areas, different
layers mainly consisting of O, P, Mg, Ca were observed. Those could indicate
different calcium phosphate salts e.g. CaHPO4(2H2O),
octacalcium phosphate (OCP), tricalcium phosphate (TCP) and hydroxyapatite (HA),
which stand for progressive corrosion stages. On the other hand areas close to
the matrix show decreased Ca, P contents, while being mainly consistent of Mg
and O. This could indicate rather early stage corrosion with Magnesium
hydroxide and its oxidation as main parts due to atmospheric contact. This
gradient is persistent throughout all three electrolyte. Additionally all
pictures show surface related voluminous cracks, which could be either the
result of drying or distinctive hydrogen evolution 7,
29. Since only Mg-0.6Ca-0.8 wt. % Zn Alloys
were used for all three different electrolytes and no large difference could be
observed concerning their EDX analysis, we state that all three follow the same
corrosion mechanism. This is in accordance with the overall 10 days corrosion
rate, which show no significant difference between the groups. The different
rates therefore have to be the results of locally occurring corrosion acceleration.
Anyhow these immersions experiments were performed for 10 days. Further research
needs to be done to investigate on the exact mechanism behind the upper
described phenomena and corrosion rate behavior in longer trials. Also the
general rate has to be decrease to reach the desired rate by Song et al. So far
efforts undertaken to tailor corrosion rate already show decreased hydrogen
evolution in vitro and in vivo 26,


wound closure and thus serum like environmental conditions preincubation in
artificial saliva protects from increasing corrosion rates, which could have been
observed in sole Hank’s solution and sole artificial saliva exposure and thus
facilitates homogenous corrosion. Therefore less negative side effects through
bubble forming, alkalization, local necrosis or implant failure occurs. This favors
the use of Mg-0.6Ca-0.8
wt. % Zn Alloys in maxillofacial surgery in comparison to other applications,
where a sole serum like environment (e.g. orthopedic field) may contribute to
higher corrosion rates after 5 days.