Corrosion behavior of stainless steel weldments in physiological solutions

The OCP of BM, HAZ and WZ in both Ringer's and Hank's solutions was determined with respect to Saturated calomel (SCE) reference electrode (+250 mV versus SHE) as given in table 5. It was observed that OCP of BM, HAZ and WZ in Ringer's solution was more negative than in Hank's solution. This was attributed to the high chloride contents in the Ringer's solution, which may possibly decrease the OCP compared to Hank's solution. At pH (7.4 ± 0.1) the stable species i.e. ${{\rm{H}}}_{2}{{{\rm{PO}}}_{4}}^{1-}$ and ${{{\rm{HPO}}}_{4}}^{2-}$ in the hanks solution have the ability to preferentially interact with the surface of stainless steel. These species could promote the stability of passive film over the surface and may limit the localized corrosion tendency of 316L stainless steel [1, 6]. The relatively negative OCP value of HAZ in Ringer's solution was considered due to the recrystallization of γ-phase and nucleation of δ-ferrite along the grain boundaries. These sites could be the active sites for electrochemical attack at the surface. Both HAZ and WZ were found to be more anodic (negative OCP) compared to the BM in Hank's solution whereas, thermodynamically the HAZ was more active than BM and WZ in the Ringer's solution.

Typical Potentiodynamic Tafel Scans for BM, HAZ and WZ were obtained (figure 7) in both simulated physiological solutions. The polarization behavior was evaluated in the potential range of (−0.5 to 1.5 V)_OCP with a scan rate of 1 mV s⁻¹. The kinetic parameters of BM, HAZ and WZ were calculated by Tafel extrapolation method using Echem Analyst software as given in table 6. The higher corrosion current density (I_corr) of HAZ and WZ in ringer solution represented higher corrosion rate compared to BM according to Butler–Volmer relation [17].

$$\begin{eqnarray}&&{{\boldsymbol{i}}}_{{\rm{net}}}={{\boldsymbol{i}}}_{{\boldsymbol{o}}}\left\{\left.\exp \left({{\boldsymbol{\beta }}}_{{\rm{a}}}\displaystyle \frac{{\boldsymbol{nF}}}{{\boldsymbol{RT}}}{{\boldsymbol{\eta }}}_{{\rm{a}}}\right)\right|-{\bf{e}}{\bf{x}}{\bf{p}}\left[-({\bf{1}}-{{\boldsymbol{\beta }}}_{{\rm{c}}})\displaystyle \frac{{\boldsymbol{nF}}}{{\boldsymbol{RT}}}{{\boldsymbol{\eta }}}_{{\rm{c}}}\right]\right\},\end{eqnarray} \tag{ 1 }$$

where '${{\boldsymbol{\beta }}}_{{\rm{a}}}$' and '${{\boldsymbol{\beta }}}_{{\rm{c}}}$' are the anodic and cathodic Tafel constants. '${{\boldsymbol{\eta }}}_{{\rm{a}}}$' and '${{\boldsymbol{\eta }}}_{{\rm{c}}}$' represent the anodic and cathodic polarization of the metal surfaces during Tafel extrapolation.

Zoom In Zoom Out Reset image size

The presence of δ-ferrite in the austenitic matrix could disrupt the uniformity of passive film. The preferential dissolution of δ-ferrite caused by the formation of local galvanic cells within the austenitic matrix could enhance the pitting tendency of surface in the chloride containing physiological solution. The higher I_corr and more negative (active) corrosion potential E_corr of HAZ than BM in the Ringer's solution was also attributed to the formation microstructural inhomogeneity during welding process. The WZ zone also presented high corrosion rate than BM in the same solution and was related with the relatively high concentration of dispersed δ-ferrite in the austenitic phase. However, under steady state condition (at or near OCP), the relative small difference in the E_corr of WZ (−275.0 mV)_SCE and BM (−228.9 mV)_SCE may be less deleterious in terms of local galvanic cell formation.

In Hank's solution the corrosion rate of WZ was relatively high and followed the trend (corrosion rate of BM < HAZ < WZ). In comparison, the corrosion rate of AISI 316L weldment (of all BM, HAZ and WZ) in the Ringer's solution was higher than in Hank's solution. The lower corrosion rate in Hank's solution may be related with the adsorption of stable species (${{\rm{H}}}_{2}{{{\rm{PO}}}_{4}}^{1-}$ ${{{\rm{HPO}}}_{4}}^{2-}$) at the local anodic sites developed at the surface under applied conditions. These adsorbed anions may restrict the de-polarization of surface and hence may decrease the ingress of other aggressive ions towards the surface. In relation with adsorption, the electrochemical stability of the passive film was enhanced and therefore decreased the corrosion tendency of weldment in Hank's solution. But the relatively high concentration of δ-ferrite in the WZ within and along the grain boundaries could disrupt the uniformity of the passive film, which could be the possible reason of high corrosion rate in both solutions. The reported results about the electrochemical behavior of are in support and comparable to other electrochemical investigations on literature austenitic steel welded joints [12, 18–20]. To evaluate the passivation tendency of different zones in the Ringer's solution, the each sample was polarized from its OCP to +1.5 V (versus OCP) with a scan rate of 5 mV s⁻¹. The passive current density (i_p) of HAZ was 35.93 μA cm⁻² compared to 13.80 and 15.05 μA cm⁻² for BM and WZ, respectively. Similarly in the Hank's solution, the i_p for HAZ (11.03 μA cm⁻²) was slightly higher than BM (9.59 μA cm⁻²) but WZ presented no limiting current value (representing passive behavior). The continuous increase in the current density during anodic polarization was associated with formation of relatively unstable passive film due to presence of large concentration δ-ferrite in the γ matrix.

During CP the surface of all samples was scanned between −0.5 mV/_OCP to an apex potential (1.00 V versus OCP). The pitting resistance and protection tendency was measured by reversing the potential from the apex potential at a scan rate of 2.5 mV s⁻¹. The abrupt increase in the current during forward (anodic) scan belongs to the transpassive region corresponding to the breakdown of passive film, designated as pitting potential, E_pit. The intersection of reverse polarization curve with the anodic polarization line resulted in the positive hysteresis loop, represented the pitting tendency and is designated as protection potentials, E_prot. In other words, the pitting may happen above E_prot and can be used as a diagnostic to estimate complete protection of surface from localized attack under applied conditions. There is always a possibility for the formation of metastable pits above this potential which are difficult to analyze by other physical characterization means. However, the potential larger than E_pit represents the formation of stable pits at the surface.

The CP scans of all the samples are shown in figure 8 and the values of corrosion potential, (E_corr) pitting potential (E_pit) and protection potential (E_prot) are given in table 7.

Zoom In Zoom Out Reset image size

The stability of passive film in the Ringer's and Hank's solution depends on the microstructure and electrochemical species present in the electrolytes. Pit propagation kinetics was measured by analyzing the hysteresis loop in the CP trends [13, 21, 22].

The anodic polarization depicted the rapid increases in current, which reflected the pit initiation and growth due to breakdown of passive film from the localized sites. In Ringer's solution, the small current fluctuation within the passive region of HAZ and WZ was observed which confirmed to the presence of localized active sites within the microstructure, corresponding to the metastable pitting tendency. These localized corrosion reactions at the surface were attributed to the presence δ-ferrite in the austenitic (γ) matrix.

The lower pitting resistance and protection tendency of HAZ and WZ than BM in both solutions are represented in figure 9. In comparison to the solution types, the presence of ${{\rm{H}}}_{2}{{{\rm{PO}}}_{4}}^{1-}$ and ${{{\rm{HPO}}}_{4}}^{2-}$ in Hank's solution behaved as buffer to modify the passive film and may adsorb at the local active sites to restrict the initiation and formation of pits at the surface [23].

Zoom In Zoom Out Reset image size

The higher pitting resistance of BM than HAZ and WZ demonstrated the direct relation with the microstructural inhomogeneity produced during weld thermal cycle. The decreasing trend in the pitting resistance from BM to WZ indicated the active localized dissolution of HAZ and WZ. The pitting resistance and protection tendency of welded samples in Ringer's solution was lower than in Hank's solution which could be associated with the high activity of Cl⁻ ions species in the former electrolyte. Also, the ${{\rm{H}}}_{2}{{{\rm{PO}}}_{4}}^{-1}$ and ${{{\rm{HPO}}}_{4}}^{-2}$ species present in the Hank's solution could preferentially adsorb at the active sites and may possibly restrict the dissolution tendency. The variation in the δ-ferrite concentration and other phases in the microstructure of HAZ and WZ could deteriorate the uniformity of passive film. It has been evaluated that the heterogeneity in the microstructure during welding could deleteriously affect the corrosion properties of AISI 316L and may accelerate localized dissolution of WZ in the simulated physiological solution.