Evaluation of the corrosion behaviour

a) Open Circuit Potential (OCP)

Open circuit potential (OCP) of the Mg-RE-intermetallics is shown in Figure 4.7. It reveals that the OCP of the RE element has no direct influence on the potential of Mg-RE-intermetallic e.g. it cannot be used to predict their behaviour, as the order of activity (potential) is not the same anymore. The OCP values of Mg-RE-intermetallics shifted to less noble values compared with pure RE potential values (Figure 4.1), but they are still significantly less active than pure magnesium. The intermetallic castings of Mg38Nd, Mg40Y and Mg40La show stable OCP values during the test, while Mg33Ce shows stronger potential fluctuations and the final potential shifted to a value of approximately -1576 mV (see Table 4.5). On the other hand OCP of Mg57Gd shifts towards a more noble potential values during the test compared to with the other intermetallics and reached a nobler potential at -1507 mV (see Table 4.5). The OCP results might be influence by phase mixtures especially for the Mg38Nd and Mg 33Ce castings.

Figure 4.7 Open circuit potential (OCP) vs. time of the as-cast Mg-RE-intermetallics after 30 minutes immersion in 0.5 wt.% NaCl solution

Table 4.5 Electrochemical data from open circuit potential (OCP) measurements of the as-cast Mg-RE-intermetallics in 0.5 wt% NaCl solution

Mg-RE-int.

as-cast condition E_initial

(mV vs. Ag/AgCl) E_final

(mV vs. Ag/AgCl) OCP

(mV vs. Ag/AgCl) after 30 min

Mg33Ce -1636 -1576 -1570 ± 3

Mg38Nd -1661 -1552 -1558 ± 6

Mg40La -1541 -1550 -1549 ± 0.4

Mg40Y -1568 -1538 -1551 ± 15

Mg58Gd -1550 -1507 -1529 ± 22

Mg -1704 -1600 -1629 ± 54

b) Potentiodynamic polarisation measurements

Figure 4.8 shows the potentiodynamic polarisation measurements for the Mg-RE-intermetallics.

The plot includes the response for the high purity magnesium, for comparison. These results, illustrate that combination of RE with Mg form respective intermetallic, which leads to a corrosion potential closer to Mg but which is still higher by up to 50 – 99 mV compared to pure RE and Mg (see Table 4.6). Higher corrosion rates were also observed. Mg33Ce showed the lowest corrosion rate (2.24 mm/year), while Mg40 Y has the highest corrosion rate at 15.45 mm/year. Mg38Nd, Mg40La and Mg58Gd presented intermediate values, at 7.84 mm/year, 9.93 mm/year and 10.36 mm/year, respectively.

Figure 4.8 Potentiodynamic polarisation measurements of the as-cast Mg-RE-intermetallics in 0.5 wt.% NaCl solution

Table 4.6 Electrochemical data from potentiodynamic polarisation measurements of the as-cast Mg-RE-intermetallics in 0.5 wt.% NaCl solution

Mg-RE-int.

as-cast condition

i_corr (mA/cm²)

E_corr

(mV vs. Ag/AgCl)

CR (mm/year) Mg33Ce 0.10 ± 0.06 -1536 ± 1 2.24 ± 1 Mg38Nd 0.34 ± 0.05 -1527 ± 2 7.84 ± 3 Mg40La 0.43 ± 0.03 -1519 ± 3 9.93 ± 2 Mg40Y 0.68 ± 0.02 -1532 ± 3 15.45 ± 4 Mg58Gd 0.45 ± 0.04 -1487 ± 2 10.36 ± 4 Mg 0.03 ± 0.003 -1586 ± 5 0.61 ± 0.7

4.2.2.2 Effect of the heat treatment a) Open Circuit Potential (OCP)

Figure 4.9 illustrates the open circuit potential values (OCP) of the Mg-RE-intermetallics after heat treatment at 540°C for 72 h. The OCP values of the heat treated Mg-RE-intermetallics shifted slightly to more active values compared with the as-cast intermetallics (Figure 4.7), but they are still less active than pure magnesium. The intermetallics, Mg38Nd, Mg40La and Mg40Y show unstable OCP values during the initial seconds (0 – 500 s), then their OCP values shifted to more noble potentials at -1562 mV, -1559 mV and -1540 mV, respectively. Mg33Ce shows less potential fluctuations compared with as-cast condition and the final potential shifted to a value of approximately -1584 mV (see Table 4.7). On the other hand OCP of Mg57Gd stabilised after 500 s immersion and at the end is still the noblest OCP with -1533 mV

Figure 4.9 Open circuit potential (OCP) vs. time of heat treated Mg-RE-intermetallics after 30 minutes immersion in 0.5 wt.% NaCl solution

Table 4.7 Open circuit potential (OCP) measurements of the heat treated Mg-RE-intermetallics in 0.5 wt% NaCl solution

Mg-RE-int.

Potentiodynamic polarisation curves of the Mg-RE-intermetallics after heat treatment at 540°C for 72 h are shown in Figure 4.10. After the heat treatment the corrosion potential values (Ecorr) of the intermetallics and Mg were divided in three categories. Lower E_corr differences compared to Mg and the as-cast condition were observed for Mg33Ce, Mg38Nd and Mg40La with around 18 mV, 35 mV and 50 mV respectively. Mg40Y showed nearly no change and the E_corr difference remained at 53 mV. Mg58Gd showed the highest Ecorr difference around 135 mV which is even higher than in the as-cast condition (see Table 4.8). The current densities were also affected by the heat treatment changing the corrosion rates compared with the values of the as-cast intermetallics (see Table 4.6). The corrosion rate increased dramatically for Mg38Nd (14.41 mm/year). The opposite was observed for Mg58Gd with a corrosion rate of 4.75 mm/year, indicating a reduction of 50% of the as-cast value. Mg33Ce, Mg40La and Mg40Y did not show remarkable changes of their corrosion rates (see Table 4.6 and Table 4.8). The Mg addition tends to produce less noble values of Ecorr for the intermetallics, as the Mg additions are less noble than pure RE [15]. Corrosion potential values (E_corr) for the Mg-RE-intermetallics are clearly nobler than corrosion potential of Mg. However icorr of Mg-RE-intermetallics are considerably higher than icorr of pure Mg generating high corrosion rates. In general the intermetallic phases degraded faster than pure magnesium

Table 4.8 Electrochemical data from potentiodynamic polarisation measurements of the heat treated Mg-RE-intermetallics in 0.5 wt. % NaCl solution

4.2.2.3 Galvanic coupling

The galvanic current was measured between the magnesium matrix and the Mg-RE-intermetallics in the as-cast condition and following the heat treatment. The measurements were performed using an exposed area of 0.5 cm² for both electrodes. The galvanic coupling reveals the interaction between the intermetallic phases and the matrix when in contact with a corrosive medium. Figure 4.11 shows the resulting current density vs. time for the coupling of pure Mg with the Mg-RE-intermetallics in the as-cast condition (a) and after the heat treatment (b) in 0.5 wt.% NaCl solution. For the as- cast intermetallics the coupling between Mg and Mg57Gd show the highest current density of around 0.375 mA/cm², followed by Mg and Mg38Nd at 0.166

Mg-RE-int.

after heat treatment

i_corr (mA/cm²)

E_corr (mV vs. Ag/AgCl)

CR (mm/year) Mg33Ce 0.11 ± 0.01 -1567 ± 4.92 2.57 ± 0.19 Mg38Nd 0.63 ± 0.03 -1551 ± 1.13 14.41 ± 0.09 Mg40La 0.47 ± 0.20 -1536 ± 2.76 10.64 ± 4.16 Mg40Y 0.67 ± 0.03 -1533 ± 1.65 15.36 ± 0.97 Mg58Gd 0.21 ± 0.01 -1451 ± 2.51 4.75± 0.55

Mg 0.03 ± 0.003 -1586 ± 5.43 0.61 ± 0.7 Figure 4.10 Potentiodynamic polarisation measurements of heat treated

Mg-RE-intermetallics in 0.5 wt.% NaCl solution

mA/cm², Mg and Mg33Ce with 0.107 mA/cm² . Mg and Mg40Y revealed large current density fluctuations during the test but the final current density was 0.032 mA/cm². Mg and Mg40La showed the lowest current density at 0.011 mA/cm². The coupling experiment following the heat treatment show changes to the current density of the galvanic couples, some show an increase while others show a decrease of current densities compared with those as-cast samples. Mg and Mg33Ce showed a slight increase in the current density (0.144 mA/cm²), while for Mg-Mg38Nd and Mg-Mg40Y current density doubled (0.30 mA/cm²) and 0.086 mA/cm² respectively.

Mg and Mg40La showed the strongest relative increase six times more in the final value of the current density with 0.05 mA/cm².

Mg and Mg57Gd showed a 50% reduction of the current density (0.185 mA/cm²) in the heat treated condition compared to the as-cast samples. The uniform distribution of elements, and the reduced amount of additional phases detected after the heat treatment resulted in the increase of the current density of the galvanic coupling of all intermetallics with Mg. Only the Mg57Gd showed an opposite trend. A possible explanation is the reduction of Į-Mg in the intermetallic castings after heat treatment. This has removed or reduced a local self-protection effect of the cathode by incorporated sacrificial Mg anode areas reducing the external current between Mg anode and intermetallic cathodes.

The intermetallic containing Nd was the exception, as even after the heat treatment two different intermetallics phases were observed and it is difficult to judge which one of the phases is responsible for the properties.

At this point only three Mg-RE systems are of further interest: Mg-La because of the lowest exchange currents, Mg-Ce because of the lowest practical potential difference and the lowest corrosion rate of its intermetallic phase and Mg-Gd because of its high solubility offering the chance to prevent critical intermetallic precipitates. Furthermore the intermetallics in this system it present seems to be the more detrimental from the point of galvanic coupling.

The other two systems (Mg-Nd and Mg-Y) were omitted, because they did show only average properties. For the other systems binary alloy were casted with different amounts of intermetallics in the microstructure and the detailed analysis of microstructure and properties will be addressed in the following chapter.

a)

b)

Figure 4.11 Current density vs time of the galvanic coupling Mg-Intermetallics:

a) as-cast condition and b) after heat treatment