Low-density metals such as Mg and Al (and their alloys) are of high interest for lightweight engineering applications in various industries. Moisture sensitivity, poor tribology, and corrosion susceptibility limit the direct application of these light metals. Plasma electrolytic oxidation (PEO) is extensively used to passivate light metals against corrosion and enhance their mechanical properties. PEO processes in current use are often energy-intensive and use toxic electrolytes. Incorporating composite characteristics to PEO-treated surfaces typically requires modification of electrolytes with nanoparticle addition. Some applications also need post-treatment of oxidized coatings to ensure functionality. We report a versatile, environmentally friendly PEO process that uses organo-silicate electrolytes enriched with nitrogen-containing solutions. The single-step process produces ∼6 μm thick, uniform, adherent, and porous oxide coatings on AZ80 and Al6061 surfaces in 15 min. We evaluated the influence and effectiveness of in situ nitridation by comparing the coating properties with those on alloys treated in PEO electrolytes without nitrogen-containing chemicals. The two sets of coatings were porous with multilayered basalt-like topographies and were composed of metal oxides and metal silicates. Alloys treated in nitrogen-containing electrolytes exhibited the presence of oxynitrides. The use of nitrogen-containing PEO electrolytes resulted in coatings with enhanced mechanical behavior. We found that the corrosion resistance of coatings prepared using low voltages in this study was comparable to the traditional PEO-treated coatings reported in the literature. Nitridation of the coatings, however, appears to have a slightly negative influence on the coatings' corrosion resistance. Our future work will focus on improving the corrosion resistance of the mechanically resilient, nitride-containing PEO-treated coatings.
Low-density metals such as Mg and Al (and their alloys) are of high interest for lightweight engineering applications in various industries. Moisture sensitivity, poor tribology, and corrosion susceptibility limit the direct application of these light metals. Plasma electrolytic oxidation (PEO) is extensively used to passivate light metals against corrosion and enhance their mechanical properties. PEO processes in current use are often energy-intensive and use toxic electrolytes. Incorporating composite characteristics to PEO-treated surfaces typically requires modification of electrolytes with nanoparticle addition. Some applications also need post-treatment of oxidized coatings to ensure functionality. We report a versatile, environmentally friendly PEO process that uses organo-silicate electrolytes enriched with nitrogen-containing solutions. The single-step process produces ∼6 μm thick, uniform, adherent, and porous oxide coatings on AZ80 and Al6061 surfaces in 15 min. We evaluated the influence and effectiveness of in situ nitridation by comparing the coating properties with those on alloys treated in PEO electrolytes without nitrogen-containing chemicals. The two sets of coatings were porous with multilayered basalt-like topographies and were composed of metal oxides and metal silicates. Alloys treated in nitrogen-containing electrolytes exhibited the presence of oxynitrides. The use of nitrogen-containing PEO electrolytes resulted in coatings with enhanced mechanical behavior. We found that the corrosion resistance of coatings prepared using low voltages in this study was comparable to the traditional PEO-treated coatings reported in the literature. Nitridation of the coatings, however, appears to have a slightly negative influence on the coatings' corrosion resistance. Our future work will focus on improving the corrosion resistance of the mechanically resilient, nitride-containing PEO-treated coatings.
Magnesium and aluminum alloys are of high
interest for lightweight
engineering and design applications in automotive, electronics, computer,
and sporting goods industries.[1] Wear and
corrosion susceptibility of Mg and Al alloys limit their direct applicability.
Surface modification techniques such as physical vapor deposition
(PVD), chemical vapor deposition (CVD), sol–gel coating, chemical
conversion coating, electroplating, thermal spraying, and anodizing
are commonly employed to protect light metals. Wu et al. used PVD
to produce corrosion-resistant oxynitride-based coatings on Mg alloys.
The authors report that the incorporation of AlON into the protective coatings
enhanced the bio-corrosion resistance of Mg in simulated physiological
environments.[2] Liu et al. reported the
enhanced tribological behavior of Al7075 when coated with TiAlN via
PVD.[3] Noder et al. published their findings
reporting the improved tribological behavior of PVD-coated Al alloys
machined for automotive structural components.[4] Ishizaki et al. reported the use of CVD to protect Mg alloys (AZ321)
from corrosion by inducing superhydrophobic characteristics to the
surface. The authors also report the high chemical stability of the
CVD coatings in acidic, neutral, and alkaline environments.[5] The drawback of using PVD and CVD techniques
for light metal protection is the incompatibility of the mostly ceramic
coatings with the metallic substrates. Gadow et al. reported that
coatings prepared by PVD and/or CVD processes often exhibit low elasticity
and thus are susceptible to mechanical damage.[6] Economic feasibility and process scalability led to the wide use
of sol–gel for protective coatings on light metals. Shadanbaz
and Dias reported the synthesis of calcium phosphate coating using
a sol–gel technique on biocompatible Mg substrates. The study
reported some ways in which calcium phosphate coatings impeded the
corrosion rates of Mg-based bone implants.[7] Other reviews by Wang et al. and Vazirinasab et al. reported the
synthesis of superhydrophobic coatings by the sol–gel process
on Mg and Al alloys to improve the substrates’ corrosion resistance.[8,9] The negative aspect of using sol–gel coatings is their poor
adherence to metallic substrates. Chemical conversion is another common
process used to generate adherent coatings on light metal substrates.
Chemical or electrochemical reactions between substrates and highly
toxic chromate-based treatments form conversion coatings. Some less
toxic alternates for conversion coatings on Mg alloys are phosphate-
and/or stannate-based solutions.[10] Becker
reviewed Zr/Ti-based chemical conversion coatings to protect Al alloys
for applications in aeronautics industries.[11] While conversion coatings are more adherent to Mg/Al alloy surfaces,
they are prone to cracking, thereby reducing their efficacy as corrosion
protectors.[12] Electroplating is one of
the oldest techniques employed to protect metallic substrates from
corrosion; however, most aqueous electroplating solutions are not
suitable for self-passivating (Mg or similar) alloys. Self-passivating
coatings are nonuniform and interfere with the adhesion of electrodeposited
coatings.[13] Gu et al. electrodeposited
nanocrystalline Ni coatings on AZ91D substrates to enhance their corrosion
resistance. The authors employed an electroless Ni conversion coating
to mitigate the passivation of Mg in aqueous plating baths.[14] Abbott et al. reviewed the utilization of ionic
liquids as plating baths for protecting Mg and Al alloys from corrosion.
The authors note that environmental compatibility is not a natural
consequence of using ionic liquids for electroplating.[15]Plasma electrolytic oxidation (PEO) or
micro-arc oxidation (MAO)
is a well-known surface treatment process for protecting light metallic
materials such as Be, Mg, Al, and Ti.[16] Markov et al.[17,18] introduced the original notion
of generating oxidized coatings on metals using discharge energy.
Traditional PEO processes are energy-intensive. PEO occurs by applying
high voltages between a target electrode and a stable cathode via
direct current (DC), pulsed DC, or alternating current (AC) in an
electrolyte. The high voltages generate a plasma state localized at
the arc discharge points, leading to the formation of adherent mixed
oxide coatings on the metallic substrates. The bidirectional coating
growth takes place by the formation of a porous surface layer and
extending into the bulk of the substrate. The coatings exhibit a range
of desirable properties such as strong adhesion, high hardness, thermal
and electrical insulation, and high corrosion resistance.Factors
such as processing voltages, current density, temperature,
and electrolyte composition influence the performance of PEO-treated
surfaces. The chemical composition of a PEO electrolyte plays a significant
role in the formation and performance of coatings.[19] For example, enhanced hardness of PEO-treated surfaces
can be attributed to silicate and phosphate compounds in electrolytes.[20,21] Other studies on the enhancement of the coatings’ performance
entailed the incorporation of secondary substances during PEO treatments.
Pezzato et al. used molybdate salts in an alkaline PEO electrolyte
to treat Mg alloys and improve their corrosion resistance.[22] Lee et al. added CNTs to PEO electrolytes to
enhance the densification of resulting coatings on Al alloys.[23] Studies are also available on the incorporation
of carbide and nitride content into PEO-treated coatings by adding
materials such as WC and TiN nanoparticles to the electrolytes.[19,24]Cirrus Materials aimed to design a sustainable PEO process
for
light metal surface modification in the current study. Here, we report
a PEO surface modification process that produces adherent and mechanically
robust coatings on Mg and Al alloys using low-energy and benign silicate-based
organo-alkaline electrolytes. Approaches to simultaneous nitridation
and carburization of the coatings entailed the addition of nitrogen-containing
eco-friendly organic chemicals to the PEO electrolytes. The resulting
coatings were composites of metal oxides, oxynitrides, carbonates,
and silicates. Our primary target was to devise a single-step method
to produce protective ceramic coatings on Mg and Al alloys enhanced
with silicate and nitrogen-containing compounds. We use commonly available
chemicals. The PEO baths are very stable at the laboratory level and
do not require replenishment. However, at an industrial scale, we
expect bath maintenance to be a process cost. Once prepared, the baths
can be safely stored at room temperature. This is a very cost-effective
feature of our technology. Additionally, the total energy consumed
for treating Mg or Al using this technology is less than 35 W-h/dm2, making it a sustainable surface treatment process. We have
evaluated the energy contribution of both the PEO energy and process
energy and determined that compared to the traditional PEO, the consumption
is less than 10%.
Results and Discussion
Synthesis
We used
a single-electrolyte formulation
to create PEO surfaces on Mg-A and Al-A alloys. We modified the same
bath using an aminophenolic compound to treat samples Mg-B and Al-B.
The aminophenol modification of bath chemistry was a pathway to incorporate
nitride content into the coatings to enhance their mechanical behavior.
The samples treated in PEO baths with and without aminophenols have
similar appearance. Figure shows the variation of voltage during PEO treatment of AZ80
and Al6061 alloys. The graphs show the low processing voltages required
to form oxide layers on the Mg and Al alloy surfaces. The processing
voltage for AZ80 alloys is <130 V (Figure a), and that for Al6061 alloys is <160
V (Figure b). These
values are lower than the ca. 400–700 V typically used in traditional
PEO processes.[25−27] In the case of Mg, low current densities are sufficient
to generate surface arcs required to produce the oxide coating. The
use of high current densities tends to burn the Mg oxide coating.
Using the same current density on Al would produce a lower arc density,
thus requiring a longer duration to breach the barrier layer and generate
a coating. Also, Al can be processed at high current densities without
burning the coatings. Thus, we selected 1 A/dm2 for Mg
and 4 A/dm2 for Al.
Figure 1
Voltage vs time graphs showing the low
energies required to produce
oxide coatings on (a) Mg AZ80 alloys and (b) Al6061 alloys with and
without nitrogen-containing PEO electrolytes.
Voltage vs time graphs showing the low
energies required to produce
oxide coatings on (a) Mg AZ80 alloys and (b) Al6061 alloys with and
without nitrogen-containing PEO electrolytes.There are some reported studies on using either low PEO processing
voltages or nontoxic electrolytic chemicals. However, we found little
research on the combination of low processing voltages and benign
electrolytic chemicals for PEO surface treatment. Cai et al. report
using an ∼140 V processing voltage in a relatively toxic bath
for surface treating AZ91D Mg alloys and incorporated cerium oxide
into the surface oxide layer during the micro-arc oxidation (MAO)
treatment to enhance the mechanical properties of the coating.[28] Dong et al. have also reported using a low-voltage
MAO process to produce oxide film on AM60 Mg alloy in fluoride-based
electrolytic baths.[29] The lowest processing
voltage reported for aluminum alloys is in the range of 350–500
V. Lee et al. used a 375 V processing voltage to produce an oxide
layer on Al7075 alloy surfaces.[23] Zhang
et al. prepared PEO coatings on pure aluminum at an ∼500 V
processing voltage and using varying current densities.[30] Compared to the studies presented by other research
groups, the PEO treatment process reported in this paper is significantly
less energy-intensive and uses environmentally friendly electrolytes.
Surface Morphology
Macroscopically, Mg and Al alloys
treated in PEO electrolytes with and without nitrogen-containing compounds
possess a similar appearance. Scanning electron microscope (SEM) images
(Figure ) of the surface-treated
alloys showed that the addition of aminophenols to the PEO electrolyte
has significantly changed the surface morphologies of the coatings.
Mg-A and Al-A, PEO-treated in an organo-silicate electrolyte, exhibit
pores and surface cracks typical of the ceramic oxide coating.[26] Mg-A coating (Figure a) exhibits plateau-like features and pores
of varying dimensions. The surface cracks appear to propagate into
the coating by a few nanometers. Mg-B coatings (Figure b) exhibit a basalt-like morphology with
uniformly distributed pores on the surface. The pore dimensions range
from 10 nm to around 2.00 μm. Al-A coating (Figure c) shows sporadically distributed
pores of <50 nm and a bilayer surface morphology. The coating also
exhibits surface cracks on the surface that appear to propagate in
the coating. Al-B coatings (Figure d) show increased porosity and a spongelike morphology.
The average pore dimensions are larger compared to Al-A and range
from 10 nm to ∼1.5 μm. We do not see any prominent cracks
on Al-B. The observed uniformities in surface morphologies of Mg-B
and Al-B suggest the presence of an aminophenol in PEO electrolytes
supplying diffusion pathways for an even distribution of generated
arc energies.
Figure 2
(a) Surface morphology of Mg-A. The porous oxide layer
on AZ80
alloy treated in PEO bath is shown. The cracks and varied pore dimensions
are typical of a surface treated using PEO. (b) Surface morphology
of Mg-B. The morphology appears flatter and more uniform compared
to Mg-A coatings. We also see an improved uniformity in the pore distribution
for Mg-B surfaces. (c) Surface morphology of Al-A. The porous oxide
layer formed because the PEO surface treatment of Al6061 alloys shows
a nonuniform morphology with prominent surface cracks. (d) Surface
morphology of Al-B. Treating Al6061 alloys with nitrogen-containing
PEO bath altered the appearance of the resultant porous oxide coating.
The pore dimensions are larger on the Al-B surface, and there are
no prominent cracks.
(a) Surface morphology of Mg-A. The porous oxide layer
on AZ80
alloy treated in PEO bath is shown. The cracks and varied pore dimensions
are typical of a surface treated using PEO. (b) Surface morphology
of Mg-B. The morphology appears flatter and more uniform compared
to Mg-A coatings. We also see an improved uniformity in the pore distribution
for Mg-B surfaces. (c) Surface morphology of Al-A. The porous oxide
layer formed because the PEO surface treatment of Al6061 alloys shows
a nonuniform morphology with prominent surface cracks. (d) Surface
morphology of Al-B. Treating Al6061 alloys with nitrogen-containing
PEO bath altered the appearance of the resultant porous oxide coating.
The pore dimensions are larger on the Al-B surface, and there are
no prominent cracks.The cross sections (Figure ) of the PEO-treated
Mg and Al alloys analyzed using SEM showed
interesting variations. Mg-A coating, in Figure a, is 6 μm thick and exhibits nonuniform
porosity with an ∼20 nm thick barrier layer between the coating
and the alloy. The image illustrates that the surface cracks observed
in Figure a extend
the entire depth of the film and propagate through the pores. PEO
treatment of Mg alloys in the nitrogen-containing electrolyte produces
a thinner oxide coating. The coating (Figure b) is ∼3.5 μm thick and porous.
Mg-B does not exhibit any cracks, suggesting that the coating could
produce superior mechanical behavior compared to Mg-A. We also observe
that the barrier layer between the Mg-B coating and the substrate
is thinner than that of Mg-A. It is possible that the difference in
the coating thickness could influence the mechanical performance of
the coating and affect the corrosion protection properties. The coatings
on aluminum alloys (Figure c,d) are both ∼6 μm thick but exhibit different
porous characteristics. Figure c confirms that Al-A coating possesses a trilayer morphology
with pores distributed sporadically through the depth of the coating.
The surface cracks (Figure c) appear only in the top layer. The image also shows the
presence of a barrier layer between the aluminum oxide coating and
the Al6061 alloy. The layer appears to be ∼50 nm thick and
nonuniform along the coating–substrate interface. Al-B coating
(Figure d) does not
exhibit the surface layer seen in Al-A. The figure also shows that
the cross-sectional porosity of aluminum alloys treated in nitrogen-containing
PEO baths is different from the porosity observed on the surface (Figure d) of the coating.
We observe cracks in the Al-B coating that are not evident from SEM
images. The interfacial layer between the Al-B coating and substrate
is more uniform compared to the Al-A and substrate. Our hypothesis
is that the improved uniformity is due to the presence of nitrogen-containing
polymeric compounds in the PEO baths. Epoxy contamination during sample
preparation obscures the porosity.
Figure 3
Cross-sectional morphologies of (a) Mg-A
coating on PEO-treated
AZ80 alloy, (b) Mg-B coating on PEO-treated AZ80 alloy with a nitrogen-containing
compound in the electrolytic bath, (c) Al-A coating on PEO-treated
Al6061 alloy, and (d) Al-B coating on Al6061 alloy treated in nitrogen-containing
PEO electrolyte.
Cross-sectional morphologies of (a) Mg-A
coating on PEO-treated
AZ80 alloy, (b) Mg-B coating on PEO-treated AZ80 alloy with a nitrogen-containing
compound in the electrolytic bath, (c) Al-A coating on PEO-treated
Al6061 alloy, and (d) Al-B coating on Al6061 alloy treated in nitrogen-containing
PEO electrolyte.PEO-produced micro-plasma,
at localized discharge points, creates
high thermal and pneumatic energies at the alloy–electrolyte
interface. The high thermal energies both melt the substrate and ionize
the electrolyte surrounding the arc discharge point. The pressure
variation between the generated plasma and molten substrate surface
causes gases from the substrate bulk to bubble to the surface, while
silicates and oxides are diffused into the surface developing a porous
crystalline oxide composite coating.[31] The
highly porous interfacial layer between the coating and the substrate
is a characteristic property of PEO-treated surfaces.[32] The addition of a nitrogen-containing polymer to the PEO
bath appears to suitably distribute the energy fluxes along the substrate
surface, which could explain the high degree of uniformity observed
on Mg-B and Al-B coatings.
Composition
Crystallographic Analysis
Figure shows the
X-ray diffraction (XRD) patterns
collected for the PEO-treated AZ80 and Al6061 alloys. The insets are
at a higher graphical magnification to highlight the low-intensity
peaks, which are not obvious from the overall spectra. The Mg alloy
surfaces exhibited different compositions and crystallographic signatures
for the samples treated in PEO baths with and without nitrogen-containing
compounds. Mg-A coating (Figure a) is composed of oxides, hydroxides, and silicates
of magnesium. The XRD patterns also show the presence of alumina and
silica in the Mg-A coating. Additionally, we detected the presence
of α-Mg and magnesium aluminide (Mg17Al12) in the Mg-A coating composition. The highest intensity peaks in
Mg-A XRD suggest that MgO dominates the crystallographic signature
of the coatings. The intensities of peaks at 2θ = 32.52°
and 2θ = 34.74° indicate that Al2O3 and Mg(OH)2 also influence the degree of crystallinity
of Mg-A coatings but to a lesser extent than MgO. The XRD pattern
of Mg-B (Figure b)
shows the presence of oxides, hydroxides, and silicates of Mg; oxides
of aluminum; and oxides of silicon in the coatings. The patterns also
indicate the presence of Mg3N2 and MgON in the coating.
The peak intensities corresponding to MgO appear to be higher and
sharper for Mg-B coatings even at higher 2θ values (48.3 and
63.7°). These results suggest that the addition of nitrogen-containing
compounds to the bath chemistry improved the degree of crystallinity
of magnesium oxide in the coating. We hypothesize that the arc energies
generated during aminophenol-enhanced PEO treatment of AZ80 alloys
are sufficiently high for the formation of Mg3N2 in the coating chemistry.[33] The appearance
of magnesium oxynitride in the coating chemistry could suggest that
the localized arc energies are capable of nitriding MgO.
Figure 4
XRD patterns
for PEO-treated AZ80 and Al6061 alloys. (a) Mg-A coatings
on AZ80 alloys prepared in PEO electrolyte without nitrogen-containing
polymers. The pattern shows the presence of Mg, Mg-Al, MgO, and magnesium
silicates. (b) Mg-B coatings on AZ80 alloys prepared in aminophenol-containing
PEO bath. The pattern exhibits the presence of all of the compounds
detected in Mg-A along with nitrides of Mg. (c) Al-A and Al-B coatings
exhibit similar crystallographic composition consisting of aluminum
oxides and silicates.
XRD patterns
for PEO-treated AZ80 and Al6061 alloys. (a) Mg-A coatings
on AZ80 alloys prepared in PEO electrolyte without nitrogen-containing
polymers. The pattern shows the presence of Mg, Mg-Al, MgO, and magnesium
silicates. (b) Mg-B coatings on AZ80 alloys prepared in aminophenol-containing
PEO bath. The pattern exhibits the presence of all of the compounds
detected in Mg-A along with nitrides of Mg. (c) Al-A and Al-B coatings
exhibit similar crystallographic composition consisting of aluminum
oxides and silicates.The XRD patterns of Al-A
and Al-B coatings (Figure c) both exhibited similar crystallographic
signatures. Al-A and Al-B coatings are composed of alumina, aluminum
hydroxide, and aluminum silicates. From the XRD peak intensities for
Al-A and Al-B coatings, we observe that alumina exhibits the highest
degree of crystallinity. The peak at 2θ = 24.2° is higher
for Al-A compared to Al-B coatings. The results for PEO-treated Al6061
alloys indicate that nitrogen-containing electrolytic bath does not
significantly influence the crystallographic composition of the Al-based
oxide layers.
Surface Chemistry
We obtained a
detailed analysis of
the surface compositions on the PEO-treated Mg and Al alloys using
X-ray photoelectron spectroscopy (XPS). The results determine the
binding states of various elements in the composite coatings. Figure shows the core-level
spectra collected from the surface of Mg-A coating. C 1s spectral
(Figure a) deconvolution
resulted in three distinct binding states of carbon for the PEO-treated
surface. Accumulated adventitious carbon is responsible for the C–C
peak at 284.8 eV, C–O–C peak at 285.8 eV, and O–C=O
at ∼289 eV. The C–O–C signature due to the carbonate
content in the coating is indistinguishable from adventitious contamination.
The spectrum also indicates the presence of some polymeric carbon
at 290.2 eV, assumed to be the result of citrates in the PEO bath
chemistry. There are no observed metal–carbon bonds from the
deconvoluted C 1s spectrum. Figure b shows the deconvoluted O 1s spectrum for Mg-A coating.
The peaks at 532 and 533.2 eV are the signals from C–O bonds
of the adventitious carbon content. The peak at 530.9 eV is from the
metal oxide content of the coating. We also observe the Na auger peak
at ∼536 eV. We attribute the presence of sodium in the coating
to remnant contamination from the PEO electrolyte. Figure c provides the peak deconvolution
for the Mg 1s spectrum. Mg 1s constituent peaks indicate the presence
of MgO (1303.5 eV)[34] and magnesium silicates
(1304.3 eV)[35] in Mg-A coating. The peak
at 1305.5 eV indicates the presence of MgCO3. The spectral
deconvolution of Si 2p resulted in two peaks. The majority of the
signature is due to organic silicon (102.3 eV) content in Mg-A. The
peak at 105.1 eV corresponds to Si–O bonds in the coating.
Figure 5
XPS core-level
scans for Mg-A coatings. (a) Spectrum showing deconvoluted
C 1s peaks with C–C aligned at 284.8 eV; (b) deconvoluted spectrum
for O 1s at ∼530 eV; (c) deconvoluted spectrum for Mg 1s at
∼1303 eV; and (d) deconvoluted spectrum for Si 2p at ∼102
eV.
XPS core-level
scans for Mg-A coatings. (a) Spectrum showing deconvoluted
C 1s peaks with C–C aligned at 284.8 eV; (b) deconvoluted spectrum
for O 1s at ∼530 eV; (c) deconvoluted spectrum for Mg 1s at
∼1303 eV; and (d) deconvoluted spectrum for Si 2p at ∼102
eV.Figure provides
the core-level spectra collected for C, N, O, Mg, and Si content in
Mg-B coatings. The results show that the binding states for C, O,
Mg, and Si elements in Mg-B coatings are almost identical to their
counterparts in Mg-A coatings. The spectra were analyzed by aligning
the C–C peak in the C 1s spectrum at 284.8 eV. The N 1s peak
at ∼400 eV (Figure c) indicates overlapping signatures from C–NH2 bonds and silicon oxynitrides in the coating. The Mg 1s peak deconvolution
(Figure d) shows that
we did not detect any signature from magnesium carbonates in Mg-B
coatings. This result only indicates that MgCO3 might not
be present in the top ∼20 nm of the coating formed by aminophenol-enhanced
PEO treatment. The Si 2p signature at 102.3 eV (Figure e) on the surface is from the organic silicon
content.
Figure 6
XPS core-level scans for Mg-B coating on AZ80 alloy. (a) Deconvoluted
C 1s spectrum with C–C peak aligned at 284.8 eV; (b) N 1s spectrum
at ∼399 eV; (c) deconvoluted O 1s spectrum at ∼533 eV;
(d) deconvoluted Mg 1s spectrum at ∼1303 eV; and (e) spectrum
for Si 2p at ∼102 eV.
XPS core-level scans for Mg-B coating on AZ80 alloy. (a) Deconvoluted
C 1s spectrum with C–C peak aligned at 284.8 eV; (b) N 1s spectrum
at ∼399 eV; (c) deconvoluted O 1s spectrum at ∼533 eV;
(d) deconvoluted Mg 1s spectrum at ∼1303 eV; and (e) spectrum
for Si 2p at ∼102 eV.Figure illustrates
the XPS core-level spectra collected for Al-A coatings. The C 1s peak
deconvolution, seen in Figure a, exhibits the presence of C–C bonds aligned at 284.8
eV, C–O–C bonds at 286.3 eV, and O–C=O
bonds at 288.7 eV. The carbon signature is primarily due to adventitious
carbon accumulated on the coating surface. The O–C=O
peak suggests the presence of carbonyl and carboxyl functional groups
in the coating. However, these are indistinguishable from the accumulated
adventitious carbon. The O 1s peak deconvolution (Figure b) indicates that the coating
contains Al2O3 (531.1 eV), C=O-based
compounds or metal carbonates (532 eV), and C–O-based compounds
(533.1 eV). C–O and C=O bonds correspond to the adventitious
contamination on the surface. The presence of metal carbonates is
due to the incorporation of carbonyl and carboxyl groups into Al-A
coating during PEO surface treatment. We observe a Na auger peak (∼535
eV) in the deconvoluted O 1s spectrum. Al 2p spectral deconvolution
(Figure c) resulted
in two peaks—one at 74.6 eV indicating the presence of Al–O–Si
bonded compounds and the other at 76.2 eV indicating the presence
of alumina in the coatings. Si 2p peaks (Figure d) indicate the presence of organic silicon
(101.9 eV) and silica (103.3 eV) in Al-A coatings.
Figure 7
Al-A core-level scans
obtained from XPS. (a) Deconvoluted C 1s
spectrum with C–C peak aligned at 284.8 eV; (b) deconvoluted
O 1s spectrum at ∼533 eV; (c) deconvoluted Al 2p spectrum at
∼74 eV; and (d) deconvoluted Si 2p peaks at ∼102 eV.
Al-A core-level scans
obtained from XPS. (a) Deconvoluted C 1s
spectrum with C–C peak aligned at 284.8 eV; (b) deconvoluted
O 1s spectrum at ∼533 eV; (c) deconvoluted Al 2p spectrum at
∼74 eV; and (d) deconvoluted Si 2p peaks at ∼102 eV.Figure illustrates
the XPS core-level spectra for aluminum alloys treated in aminophenol-modified
PEO electrolytes. The C 1s spectrum (Figure a) exhibited four deconvoluted speaks. The
C–C peak at 284.8 eV and the C–O–C peak at 286.6
eV correspond to the adventitious contamination accumulated on the
surface. We attribute the N 1s peak at 400.5 eV to overlapping signatures
for C–NH2 and silicon (Figure b). The O 1s spectra (Figure c) and Al 2p (Figure d) spectra are similar for Al-A and Al-B
coatings. The Si 2p spectrum (Figure e) resolved into three peaks—at 102.7 eV corresponding
to aluminosilicate, at 103.7 eV corresponding to organic silicon,
and at 104.6 eV corresponding to silica.
Figure 8
Al-B core-level scans
obtained from XPS. (a) Deconvoluted C 1s
spectrum with a C–C peak aligned at 284.8 eV; (b) an N 1s spectral
peak at ∼399 eV; (c) a deconvoluted O 1s spectrum at ∼533
eV; (d) a deconvoluted Al 2p spectrum at ∼74 eV; and (e) a
deconvoluted Si 2p peak at ∼102 eV.
Al-B core-level scans
obtained from XPS. (a) Deconvoluted C 1s
spectrum with a C–C peak aligned at 284.8 eV; (b) an N 1s spectral
peak at ∼399 eV; (c) a deconvoluted O 1s spectrum at ∼533
eV; (d) a deconvoluted Al 2p spectrum at ∼74 eV; and (e) a
deconvoluted Si 2p peak at ∼102 eV.The analysis of XRD patterns and XPS spectra for Mg-A, Mg-B, Al-A,
and Al-B coatings indicates that the oxide content in the coatings
is due to the PEO treatment of the alloys. The decomposition of PEO
electrolytes during treatment incorporates metal silicates and carbonates
into the coatings. The addition of nitrogen-containing compounds to
the electrolytic bath is responsible for the nitride content observed
in Mg-B and the silicon oxynitride content in Mg-B and Al-B coatings.
Localized elevated temperatures (>3000 °C) obtained during
the
arc discharge processes could contribute to the formation of nitrides
in the coatings.
Mechanical Behavior
Nanoindentation
tests, performed
on the coating cross sections, evaluated the mechanical behavior of
Mg-A, Mg-B, Al-A, and Al-B. The PEO process reported in this paper
improved the coatings’ hardness by at least 7 times for Mg
alloys and by at least 3 times for Al alloys. Untreated AZ80 alloy
exhibits a nanohardness of 1.06 GPa. Mg-A coatings exhibit 7.66 GPa,
and Mg-B coatings exhibit 8.58 GPa. Untreated Al6061 alloy exhibits
1.81 GPa. Al-A coatings exhibit 5.73 GPa and Al-B coatings exhibit
7.03 GPa. The incorporation of nitrides and oxynitrides into the coating
composition improved the hardness of Mg-B by 12% and Al-B by 22%.
We attribute the higher deviation in Al-B results to the irregularities
in the coating–substrate interfacial layer and the intermittently
formed secondary top layer.
Corrosion Behavior
We extracted
the corrosion potentials
of Mg-A, Mg-B, Al-A, and Al-B coatings from the Tafel plots in Figure . Table provides the corrosion potentials
and associated current densities. We compared the corrosion performance
of Al-A and Al-B with the performance of untreated Al alloys. We obtained
the corrosion potentials for untreated Mg alloys from Naik et al.
The data extracted from Figure a show that the corrosion potentials for Mg-A and Mg-B are
similar but the corrosion current for Mg-B is higher. A lower corrosion
current implies a lower corrosion rate of Mg-A compared to Mg-B. Our
hypothesis: this is due to the higher porosity and the lower thickness
of the Mg-B coatings. Figure b illustrates the higher corrosion resistance exhibited by
Al-A and Al-B compared to untreated Al6061 alloys. The absence of
a surface layer in Al-B coatings could contribute to their slightly
reduced corrosion resistance. Ma et al. and Shanaghi et al. suggest
that nitridation of coatings on Mg and Al alloys has a positive influence
on their corrosion resistance.[33,36] Thus, we attribute
the slightly reduced corrosion resistance of Mg-B and Al-B alloys
to their physical characteristics.
Figure 9
Tafel plots for (a) Mg-A and Mg-B coatings
formed on AZ80 alloys
and (b) Al-A and Al-B coatings formed on Al6061 alloys by PEO surface
treatment without and with aminophenol in the bath chemistry.
Table 1
Corrosion Potential and Current Densities
Extracted from Tafel Plots in Figure for PEO-Treated Mg and Al Alloysa
untreated
AZ80 alloy
Mg-A
Mg-B
Ecorr (mV)
–1507[37,38]
–1349.5
–1350.1
icorr (× 10–7 A/cm2)
27.72[37]
8.06
29.49
The data for untreated Mg and Al
alloys are available for comparison.
Tafel plots for (a) Mg-A and Mg-B coatings
formed on AZ80 alloys
and (b) Al-A and Al-B coatings formed on Al6061 alloys by PEO surface
treatment without and with aminophenol in the bath chemistry.The data for untreated Mg and Al
alloys are available for comparison.Figure demonstrates
the impedance behavior of PEO-treated aluminum alloys vs the untreated
substrates. We modeled the electrochemical impedance spectroscopic
(EIS) response of untreated and PEO-treated substrates using the equivalent
electrical circuits (EECs) in Figure d and Figure e. Rsubs, Rcoating, and Relec represent the
substrate, coating, and electrolyte resistances, respectively. Q represents the admittances of constant phase elements
for substrates (CPEsubs) and coatings (CPEcoating). ZWarburg represents the Warburg impedance
owing to the diffusion of chemical reactants in the solution. Rcoating encompasses the resistances of coatings
(Al-A/Al-B), pores, and defects and is in parallel with CPEcoating. We calculated the capacitance of the coatings according to eq 1.
Figure 10
Electrochemical impedance spectroscopy of Al-A and Al-B coatings
in comparison with EIS results for bare Al6061 substrate. (a) Nyquist
plots; (b, c) Bode plots for Al6061, Al-A, and Al-B; (d) EEC for untreated
Al6061 substrate; and (e) EEC modeled for Al-A and Al-B.
Electrochemical impedance spectroscopy of Al-A and Al-B coatings
in comparison with EIS results for bare Al6061 substrate. (a) Nyquist
plots; (b, c) Bode plots for Al6061, Al-A, and Al-B; (d) EEC for untreated
Al6061 substrate; and (e) EEC modeled for Al-A and Al-B.Rsubs is in series with the Warburg
element and parallel with CPEsubs. The dotted lines in
Nyquist plots (Figure a) and Bode plots (Figure b,c) represent the impedance response, and the solid lines
represent the fitted data. Table lists the respective resistance values. EIS modeling
shows that the polarization resistance of Al-A coating is higher than
that of Al-B. The capacitance of Al-A is lower than that of Al-B.
The results suggest that currently the nitridation of coatings is
likely to be detrimental to corrosion behavior. The higher porosity
of Al-B coatings potentially contributes to the increased capacitance
observed in the coatings. The reduction in ZWarburg also suggests that higher porosity lowers the resistance
of the coatings to diffusion of chemical reactants from the solution
to the substrate.
Table 2
Fitting Parameters of EIS EECs Obtained
for Untreated Al6061, Al-A, and Al-B Surfaces
element
bare Al-6061
Al-A
Al-B
Relec (Ω/cm–2)
1.324
3.132
1.428
Rcoating (Ω/cm–2)
1911
1374
CPEcoating (mS/sn·cm2)
0.95
1.71
ncoating
0.91
0.92
Ccoating(mF)
1.01
1.84
Rsubstrate (Ω/cm–2)
4399
3690
1845
CPEsubstrate (μS/sn·cm2)
25.22
6.86
9.88
diffusion (ms)
0.4
158.5
3.5
ZWarburg (Ω/cm–2)
1.23
545.4
42.59
nWarburg
0.42
0.33
0.30
Conclusions
In conclusion, we have
demonstrated that voltages as low as 150
V can reliably produce thick, adherent, and corrosion-resistant coatings
on Mg and Al alloys. The coatings exhibit crystalline properties and
are composed of metal oxides, silicates, and oxynitrides (in cases
where we used aminophenol-containing electrolyte for surface treatments).
Low-energy PEO surface modification can form coatings with enhanced
mechanical properties. Mg-A and Al-A coatings exhibit good corrosion
resistance behavior. Nitridation of the coatings has a positive influence
on their mechanical properties, which are desirable for automotive
and semiconductor industry applications. Cirrus Materials Science
is currently exploring the use of AC-based PEO treatment to potentially
improve the corrosion resistance of nitride/oxynitride-containing
coatings.
Materials and Methods
Substrate Preparation and Pretreatment
The current
study used magnesium AZ80 alloys and aluminum Al6061 alloys. Prior
to PEO processing, the sample surfaces were prepared by mechanically
roughening the substrate using emery paper followed by cleaning in
a commercially available 80 °C alkaline bath, for 15 min. The
pretreatment process ensured the removal of organic contamination
from machined substrates. The bath comprises 20 g/L NaCO3, 20 g/L Na2PO4, 20 g/L Na2SiO3, and 3 g/L OP-10 surfactant. Finally, we used DI water to
rinse the substrate. We designed the cleaning process to prevent any
buildup of the native oxide layer on the substrate.
PEO Process
We treated two sets each of Mg and Al alloys
in a 25 °C PEO bath comprising 70 g/L NaOH, 60 g/L Na2SiO3, 10g/L Na3C6H5O7, 6 mL/L H2O2, and 0.05 mmol/L sodium
dodecyl sulfate (SDS). We prepared two samples sets, “A”
and “B”, and we added 4.9 mL/L aminophenol to the electrolyte
for sample set “B”. Alkaline bath chemistry modified
with H2O2 ensured the generation of oxide coatings
on metal surfaces. We used a silicate compound to incorporate silicate
content into the coating and enhance the conductivity of the bath.
Citrate compounds aided in the uniform distribution of surface arc
generation. We used SDS to moderate the physical properties of PEO
bath to enhance the removal of gas bubbles generated during processing.
We conducted the treatment process for 15 minutes using a stainless-steel
counter electrode. A variable-voltage DC power system supplied a constant
current of 1 A/dm2 for Mg and 4 A/dm2 for Al
resulting in an average processing voltage of <160 V.
Morphological
Characterization
We used an FEI XL30
SEM equipped with a 30 kV field emission gun to analyze the surface
morphologies and composition of the PEO-treated light metal alloys.
Prior to imaging, we sputtered the samples with Pt using a Quorum
Tech Q150T turbomolecular pumped coater to improve their electron
conductivity for SEM imaging.
X-ray Diffraction (XRD)
We used a Rigaku XtaLAB Synergy-s
single-crystal X-ray diffractometer equipped with a Cu Kα source
(λ = 1.54184 Å, 2θ = 20–80°, 0.02°
step size) to collect phase and composition data on the oxidized sample
surfaces. We analyzed the XRD patterns using the Materials Explorer
application on Materials Project open database.[39]
X-ray Photoelectron Spectroscopy (XPS)
We employed
a Kratos AXIS DLD X-ray photoelectron spectrometer with a hemispherical
electron energy analyzer for analyzing the oxide surfaces on Mg and
Al alloys. We obtained the spectra using monochromatic Al Ka X-rays (1486 eV) operated at 150 W and maintained the analysis chamber
at 1 × 10–9 Torr for data collection. We used
survey scans from −5 to 1350 eV (160 eV pass energy) to determine
the material composition. We collected core-level data (20 eV pass
energy) for C, O, and Si from all of the samples, for N from the samples
treated in electrolytic bath modified with nitrogen-containing compounds,
and for Mg and Al from the respective alloys. We analyzed the data
using Casa XPS 2.3.14 after aligning the C 1s peak at 284.8 eV. Thermo
Scientific XPS database was used to analyze the deconvoluted spectral
peaks.[40]We analyzed the mechanical behavior
of the ceramic oxide coatings using a Hysitron TI 950 tribometer equipped
with a Berkovich tip. We mounted PEO-treated magnesium and aluminum
sample cross sections in epoxy for the nanoindentation tests to eliminate
any interference from the underlying substrate. We applied a maximum
load of 1000 mN for 2 s with 5 s preloading and unloading times. Nanoindentation
tests on nonoxidized control substrates aided in evaluating the mechanical
enhancement provided by the ceramic coating.
Corrosion Resistance
To evaluate the electrochemical
performance of the PEO-treated coatings, we used a CH Instruments
three-cell electrochemical workstation equipped with a Metek designed
K0235 flat cell kit. We immersed the samples in 3.5 wt % NaCl solution
for 30 min, prior to testing. The tests employed a Ag/AgCl reference
electrode and a Pt counter electrode. We used a freshly prepared 3.5
wt % NaCl solution for Tafel and EIS tests. We scanned the open-circuit
potential (OCP) of the coatings from −0.3 to 0.3 V for 5 min
for Tafel tests. We also measured the EIS of the ceramic coatings
between 0.1 Hz and 10 kHz frequencies with a perturbation amplitude
of 10 mV.