Abstract
Electrolytic plasma process (EPP) was used as a fast annealing treatment on ball milled surfaces. With localized melting of EPP, the surface morphology and the chemical composition can be drastically changed. Annealed, ball milled and EPP treated stainless steel 304 samples were examined and EPP treatment has been shown to improve the overall corrosion and localized corrosion. Materials were characterized by X-Ray Diffraction, SEM. The 24 hours of exposure in NaCl solution, the sample open-circuit potential, potentiodynamic polarization, and electrochemical impedance spectroscopy were measured to estimate the anti corrosion properties.
Introduction
High strength and high corrosion resistance are the desired properties for modern engineered surfaces. Stainless steels are widely applied in industry owning to its good balance of strength and corrosion resistance. Recent studies indicate that the surface strength of stainless steel 304 could be significantly increased by heavy mechanical deformation [1-3]. The surface hardness increases with reduction of grain size, according to Hall-Petch effect. Stainless steel 304 surfaces also experience phase transformation from Austenite (y) to Martensite (a’) during heavy surface deformation. These phases are metastable and transformable in SS304, based on the external conditions. The martensitie with b.c.c. structure acquires a much higher hardness than that of f.c.c. austenite and hence it is an important mechanism for surface hardening on SS304. Lee’s report [4] suggested that the heavy mechanical deformation on SS304 increased the surface roughness and the open-circuit potential drops. In the previous works, the annealing process was generally adopted to obtain nano-sized austenite for better corrosion resistance and reduce residual stresses from mechanical treatments. On the other hand, the annealing process affects the bulk material and consequently leads to the overall growth of grain and possible unwanted precipitation, which degrades the strength of the bulk material.
Electrolytic plasma process (EPP) was developed from conventional contact glow discharge [5]. It’s fast heating and quenching mechanism was utilized to perform surface annealing treatment [6]. Comparing to other plasma related surface technologies, atmospheric working environment makes EPP an affordable surface cleaning and modification technique for large scale applications. In EPP treatment, a crucial step is to generate a gas layer on the metal surfaces by overcharging the target metal in an electrolyte. With large joule heating and chemical yield on metal, a continuous gas layer forms on the metal surfaces. Further increase the applied voltage, the electrons in the gas layer would be ionized and form low temperature plasma (> 2000K). Discharge occurs at a certain voltage threshold accompanied with a large current flow and heat dissipation. The surface heat rate was estimated to be around 300oC/sec during the plasma onset and discharge, but the period of plasma formation and discharge is about a few milliseconds [7]. The plasma sets in with radii in the order of microns [8]. Thus, the thermal effect on substrate material is much less, comparing to other annealing treatments. The open-circuit potential was improved by EPP upon the ball milled SS304 [6]. In current study, corrosion properties of ball milled and EPP treated SS304 surfaces were studied by long term open circuit potential scan, potentiodynamic polarization, and electrochemical impedance spectroscopy. Corrosion resistance analyses were conducted based on the experimental results from electrochemical tests and surface material characterization.
Experimental
The materials used in this study were sectioned from 3mm thick stainless steel 304 plate (wt%: 0.064C, 18.20Cr, 8.64Ni, 1.45Mn, <0.03S and Fe balance). The samples were machined into 1.5" diameter disks and initially polished by 400 grade emery paper to remove initial oxides and manufacturing marks. The polished samples were annealed at 1050oC for 30 minutes and quenched in distilled water. By such treatment, austenite was retrieved and carbide precipitation was minimized. The surfaces of the samples were lightly polished by 600/1200 grade emery paper to remove the external oxides due to quenching. Sample surfaces were subjected to ball milling treatment in a SPEX 8000 mill for 40 min. The samples were then positioned at one end of a cylindrical cup (of same material) with 1.5" diameter and 2.5" height. 40 1/4" balls of 302 stainless steel type were used inside the cylinder. Electrolytic plasma processes were performed on ball milled samples. All EPP samples were cathodically connected with a DC power source (Magna PQA 6.6kW) in NaHCO3 electrolyte (7.5wt%, equivalent electrical conductivity of 43.1±0.5 S-cm-1), the anode was a much larger stainless steel plate. By using a large anode plate, the plasma is confined to the vicinity of the cathode [9]. The effective contact area between the cathode and the electrolyte is 1±5% inch2; the contact area varied during EPP due to the dynamic instability near liquid-air boundary. The applied voltage was set at 130V; the increase rate was default set at about 10V/sec. Current rises initially and then converges to a stable value (2±0.5A) after the plasma discharge is stabilized on the cathode surface. The EPP process time was 30 minutes. All solution temperature started from room temperature. It took approximately one minute to reach the stable plasma discharge. Afterwards, all samples were cleaned in ethanol alcohol with ultrasonic bath for 5 minutes.
Material characterization and electrochemical based corrosion tests were performed on three groups of samples, including annealed, ball milled (40minutes) and ball milled (40 minutes) followed by 30 minutes EPP. Annealed, BM and BM+EPP are respectively referred for these groups in the following discussion. Surface microstructures were characterized by XRD (Rigaku, MiniFlex, Cu Ka, A=0.154nm) and analyzed with JADE 5. The surface morphologies were studies by field emission scanning electron microscope (FE-SEM) and surface roughnesses were measured by TENCOR P11 surface profiler. The electrochemical studies were conducted with a CHInstruments electrochemical workstation 600C. Open circuit potential was measured over 24 hours for different groups. The potentiodynamic polarization were performed after 1-hour exposure of corrosion medium (3.5% NaCl) and electrochemical impedance spectroscopy (EIS) measurements were conducted at 0.5 hour, 1 hour, 2 hour, 5hour and 10 hour of immersion in order to examine the corrosion property evolution over time. The 3.5% NaCl solution for all the tests were deaerated by bubbling N2 gas for 1 hour. The testing procedures were also guided by ASTM G5-94 and G3-89.
Results and analysis
Material characterization (XRD)
The annealed, ball milled and EPP treated samples were inspected with XRD and the diffraction patterns are shown in Fig 1. Assuming austenite (y) and martensite (a’) are the only two phases in sample, the volumetric fraction of martensite could estimated by a direct comparison method [10]. The calculation for the volume fraction of martensite (fa) is conducted through the following equation.
Where Ij is the relative intensity jth phase and could be measured from the integrated area of the relevant peak. Cj is the factor related with Bragg angle of diffraction. hkl is the miller indices for the plane direction vector. In this calculation, only the following peaks were used (y (200), y (220), a’ (200), a’ (211)) in order to avoid the overlapping of peaks. Annealing treatment provided austenitic (y) initial state (fa=0.15%). Ball milling effectively converted surface microstructure into large amount of martensite (fa=40.2%). Also, very much widened peaks (Fig 1(b), y (200)) suggest ultrafine grain size. With further 30 minutes EPP treatment on ball milled samples, the martensite volume fraction reduced to 6.01%.
Fig 1 X-ray diffraction patterns: (a) annealed; (b) ball milled 40 minutes; (c) ball mill 40min + Epp 30min
Open Circuit potential over time
All the samples were exposed in 3.5wt% NaCl solution for 24 hours and the open-circuit-potentials (OCP) were recorded (Fig 2). All groups had 2 characteristic passivation cycle, indicated by arrows in Fig 2. The annealed sample had very quick passivation after immersion and OCP reached -0.15 V at about 2 hours, then its OCP dropped at a rate of 0.0125 V/hr for 8 hours. The reduction of OCP on annealed sample indicated a depassivation process and such processes were attributed to the penetration of Cl- through initial chromium oxide layer. Following the slow depassivation, an accelerated breakdown of passivation film indicated by a steeper drop (0.04V/hr) in OCP proceeded until it reached a minimum level, -0.58V. The repassivation produced a second plateau (-0.52V). The second passivation period took 4 hours to decay. The BM sample had a similar pattern of evolution in OCP as the annealed one, but its first passivation kept at a much lower potential level (-0.23V) and sustained for less time (3 hours). It’s believed that the surface roughness played an important role. It was shown in table 1 that the surface roughness of BM sample was five times larger than the annealed ones. The second passivation on BM sample also took place and kept at roughly the same level as the annealed one. The OCP for BM+EPP sample initially dropped at a very high rate. But, after surfing through a small repassivation-depassivation cycle (1 hour), it restored quickly after 2 hours. Although EPP generated the highest surface roughness level, BM+EPP samples manifested a sustainable passivation from 6th hour and high OCP was kept till the end of the test. The integrity and good sealing of protective layer on EPP treated sample are related to its special composition resulted from localized high temperature treatment.
Fig 2 Open-circuit-potential over 24 hours of immersion in 3.5% NaCl solution for anneal, ball milled and EPP treated samples
Table 1 Surface roughness for different treatment groups. Scan speed: 400um/sec, sampling rate: 100Hz
 |
Annealed |
BM |
BM+EPP |
|
Ra |
0.385 |
1.669 |
1.888 |
|
Rq |
0.472 |
2.085 |
2.345 |
|
Rt |
2.565 |
13.071 |
13.272 |
Potentiodynamic polarization test
The potentiodynamic polarization was utilized to examine the corrosion potential and protection potential (Ep). Tests were conducted after one hour of insertion into the solution. Figure 3 is a selected plot of current density vs. EOCP. The averaged results of potentiodynamic polarization tests are listed in Figure 4. The EPP treated sample exhibited highest averaged EOCP and Ep values, although its initial roughness is the highest. Such performance could be related to the high temperature treatment from EPP. On the other hand, BM samples yield the lowest EOCP and Ep values due to the larger surface roughness and possible galvanic effect between austenite and martensite phases [11]. But BM samples have the lowest corrosion current on average. The annealed sample did not have significant passivation but the protective layer could sustain till higher potential (Fig 3).
Fig 3 Potentiodynamic polarization. Scanning range: -0.6 to 1.6 Volts; scan rate: 1mV/sec
Since the protection potential could be used as a measurement of pitting corrosion, the initial and corroded sample surfaces from potentiodynamic polarization tests were examined by SEM. Figure 5(a), (c), (e) shows the initial surface morphology of annealed, BM and BM+EPP samples. Annealed samples remain clear polishing marks and BM sample surfaces have cracks due to heavy impingement of steel balls. EPP samples contain typical spheroidal extrusion and craters from melting and quenching process. Figure 5(b), (d), (f) are corroded surfaces. Low magnifications were used to view the whole picture of corroded morphologies. The annealed samples have most severe pitting after the polarization cycle. Exfoliation of the surface crust was clear on Fig 5(b). The amount of pits on BM sample is less than the annealed ones, the ball milling process effectively refined the surface grains and the increase of grain boundary density would promote the diffusion of Cr [12] and possibly insist the continuous formation of Cr oxide protective layer. Nevertheless, cracks were also observed on BM samples. On EPP samples, the pits were smaller but more in amount. It was reported that the light annealing or heat treatment on sandblasted steel could lead to increase of OCP [2], which was proved by this study (Fig 4). The EPP induces much higher temperature than conventional annealing temperature on the steel surfaces. Such high temperature leads to the precipitation and depletion of Cr during melting of surface materials. Tables 2 are the statistical atomic percentage distribution of surface chemical element for different groups. ED AX was used for this measurement. The BM sample has very similar element distribution as the annealed one, except chromium and oxygen percentage is slightly higher and matches the reported values [2]. Three selected measurements from different spots BM+EPP samples were listed. The bulk result of EPP sample shows there is a 50% increase in carbon and 18% decrease in chromium, compare to the annealed sample. On the spheroids, which are the solidified molten material, the depletion of chromium is very clear (-61%). In the valleys (Fig 5 (e)), the atomic percentage of chromium recovered to the similar level as annealed sample’s. From this observation, the pits on EPP treated samples could be considered to initiate from those chromium depleted molten spots.
Fig 4 Averaged corrosion potentials and current densities from Potentiodynamic polarization tests
Fig 5 SEM for initial and corroded surface: (a-b) annealed; (c-d) BM; (e-f) BM +EPP
Table 2 EDAX results for element distribution on initial surfaces
|
Element |
Annealed |
BM |
BM+EPP |
BM+EPP |
BM+EPP |
|
|
(bulk) |
(bulk) |
(bulk) |
(spheroid) |
(valley) |
|
C |
07.98 |
06.63 |
10.14 |
22.29 |
05.85 |
|
O |
01.25 |
02.66 |
01.05 |
00.59 |
01.04 |
|
Cr |
16.47 |
16.64 |
13.48 |
06.38 |
16.34 |
|
Mn |
01.33 |
01.68 |
01.28 |
00.52 |
01.40 |
|
Fe |
65.36 |
63.85 |
65.28 |
57.95 |
66.10 |
|
Ni |
07.60 |
08.54 |
08.77 |
12.28 |
09.28 |
Electrochemical Impedance Spectroscope (EIS)
The electrochemical impedance test was performed at OCP ±0.002V for each indicated time point. Nyquist plots (Fig 6) are used to illustrate the evolution of corrosion properties. At initial exposure (0.5h) stage, annealed sample has a large capacitive loop (radius of 3.5 x103 Q/cm2), which indicates a thicker protective layer. The loop radius reduced to half of its initial level after 2 hours. The reduction of corrosion resistance was less significant after 5 hour immersion. BM sample exhibited larger corrosion resistance (radius of 2.5×104 Q/cm2), one order of magnitude large than the annealed sample. This behavior agrees with the suggestion of faster chromium diffusion towards surface after heavy mechanical impingement [12]. A small inductive loop was observed at 1h for BM sample. This indicates a fast breakdown of passive film. It could be explained that the refined surface provide larger amount of corrosion path than annealed case, once the initial chromium oxide film was penetrated, the consequent deterioration is very fast. It could also be seen from Fig 2, where the OCP decays much faster at the initial few hours of exposure. The corrosion resistance reaches minimum after 5 hour of exposure and the corrosion resistance restored at 10 h, which could also be confirmed from Fig 2. A second passivation starts about 10 hour. EPP treated samples had a very different behavior from the previous two. The initial corrosion resistance (0.5h) was the lowest (radius of 1.3 x103 Q/cm2). It is because of the larger surface roughness and the chromium depletion at some spots like spheroids from melting. The corrosion resistance experienced a rise after 1 hour, and then the loop had a slight shrinkage after 2 hour, which corresponds to the first small peak of EPP curve on Fig 2. Significant rise of corrosion resistance after 5 hour could be seen from Nyquist plot, which could also be checked in Fig 2. The ball milled surface experience fast melting and quenching through EPP. It could be observed that the surface crust of EPP treated surface did not contain large cracks under anodic loading (Fig 5(f)). This implies good structural integrity of the top layer and secondary layer could be assumed to acquire a similar structure as BM sample. The good structural integrity of the top layer effectively sealed the chromium oxides underneath it. And hence, provides a sustainable protection against corrosion.
Fig 6 Nyquist plots for annealed, BM and BM+EPP samples
Conclusion
In this study, the electrolytic plasma process on SS304 was proved to have longer sustaining corrosion resistance in 3.5% NaCl solution. EPP annealing could effectively convert surface martensite back to austenite. Although the EPP treatment yields higher surface roughness, its sealing mechanism on top surface effectively retain the substrate protective layer and hence provides a good corrosion resistance. EPP treatment also shows better pitting corrosion resistance than the annealed and BM cases because no significant cracking and exfoliation are found. An explanation for this improvement of corrosion resistance in NaCl solution was proposed. The special structure from ball milling and consequent EPP treatment provides a sustainable corrosion barrier on 304 stainless steel surface. The promising corrosion application of this surface modification treatment is worth for further study.
