Study of the Precipitation Hardening Behaviour and Intergranular Corrosion of Al-Mg-Si Alloys with Differing Si Contents

metals

Article Study of the Precipitation Hardening Behaviour and Intergranular Corrosion of Al-Mg-Si Alloys with Differing Si Contents

Yaya Zheng 1,2, Binghui Luo 1,*, Zhenhai Bai 1, Juan Wang 1 and Yuan Yin 1

1 College of Materials Science and Engineering, Central South University, Changsha 410083, China; [email protected] (Y.Z.); [email protected] (Z.B.); [email protected] (J.W.); [email protected] (Y.Y.) 2 Key Laboratory of Safety Design and Reliability Technology for Engineering Vehicle, Changsha University of Science and Technology, Changsha 410114, China * Correspondence: [email protected]; Tel.: +86-138-0842-5282

Received: 10 August 2017; Accepted: 17 September 2017; Published: 21 September 2017

Abstract: The effects of Si addition on the precipitation hardening behaviour and evolution of intergranular corrosion (IGC) of Al-Mg-Si alloys were investigated using hardness tests, scanning electron microscopy (SEM), potentiodynamic polarization measurements, and high-resolution transmission electron microscopy (HRTEM). With an increase of the Si content, the peak hardness of the Al-Mg-Si alloys considerably increased by enhancing the density of the β” (Mg5Si6) phase inside the grains. The microstructures affecting the IGC performance consisted of MgSi particles, Si particles, Al-Fe-Mn-Si intermetallics, and the precipitate-free zone (PFZ). The IGC susceptibility of the Al-Mg-Si alloys was mainly attributed to the high electrochemical potential difference between the MgSi particles and solute-depleted zones. Excess Si improved the IGC susceptibility of the alloys, mainly due to an increase of the grain boundary MgSi precipitates. Furthermore, the evolution of the IGC process was discussed in detail.

Keywords: intergranular corrosion; microstructure; precipitation hardening; Al-Mg-Si alloys; electrochemical potential difference

1. Introduction Al-Mg-Si alloys have been widely used to produce ocean ships, automotive bodies, and fuselages owing to their high strength-to-weight ratio, excellent formability, and corrosion resistance [1–3]. Their service properties are mainly acquired by the formation of high-density metastable strengthening phases during artiﬁcial ageing after solution heat treatment, followed by rapid water quenching to room temperature [4]. The precipitation of Mg/Si master alloys during the ageing treatment is generally believed to occur via the following sequence: GP zone (Guinier Preston zone)—β00 phase—β0 phase—β phase [5]. The main contributors to hardness and mechanical strength in Al-Mg-Si alloys are the β00 phase and GP zone. To improve the mechanical performance, a small amount of Cu and an excessive amount of Si are often added to Al-Mg-Si alloys. Cu can accelerate the formation of a coherent reinforcement phase of GP zones and MgSi precipitates, and can also form metastable strengthening Cu-containing precipitates, primarily including the θ phase (Al2Cu), S phase (Al2CuMg), or Q phase (Al4Cu2Mg8Si7)[6,7]. Increasing the Si content in Al-Mg-Si alloys usually provides more formable and higher mechanical strength properties by improving the density of MgSi precipitates with a ﬁne and uniform distribution [8]. High electrochemical potential differences between the aluminium matrix and precipitates play critical roles [9–11]. Generally, these precipitates include inhomogeneities, intermetallics, and

Metals 2017, 7, 387; doi:10.3390/met7100387 www.mdpi.com/journal/metals Metals 2017, 7, 387 2 of 12 grain boundaries, which may introduce susceptibility to intergranular corrosion (IGC) [12,13] and become the main initial locations of electrochemical corrosion [8,9]. MgSi particles represent the main strengthening precipitate, with diverse structures and sizes depending on the thermal processing parameters; neutral chloride electrolytes possess a lower corrosion potential. The MgSi phase must be considered in studies of corrosion behaviour [14,15], as this phase results in a high electrochemical potential difference between MgSi secondary phases and the cathode. In addition, intermetallic particles and anodic sites [16] lead to the preferential dissolution of more active sites, including the matrix and precipitate-free zone (PFZ) [17], as they are often present as cathodes and protected in the electrochemical corrosion process. With excessive Si and Cu addition, the formation of Si-rich or Cu-containing precipitates introduced to improve mechanical properties are not conducive to improving the corrosion performance of the alloy [18,19]. These precipitates often act as unfavourable electrochemical heterogeneities for corrosion resistance, related to pitting corrosion and IGC attack. To resolve this problem, increasing the Mg content can reduce the corrosion propagation susceptibility by altering the precipitate type and the structure containing the Si strengthening phase [8]. However, this process is simultaneously detrimental to the mechanical properties of hardness and strength [20]. Therefore, to acquire Al-Mg-Si alloys with a suitable performance, extensive works on IGC susceptibility have been conducted to investigate the corrosion behaviour of Al-Mg-Si alloys [21–23]. Mol et al. [24] studied the contribution of the MgSi structure in the IGC process, and reported that stable MgSi particles could improve the corrosion resistance compared to alloys with coherent MgSi precipitates. Li et al. [25] reported that the continuous distribution of grain boundary precipitates (GBPs) were a primary reason for the high IGC susceptibility of Al-Mg-Si alloys. Mizuno et al. [26] suggested that surface particles should be considered as initial propagation sites in pitting corrosion, where one of the main problems involved the inﬂuence of MgSi particles on the initial propagation and process of IGC. Regarding this problem, Table1 summarizes some ﬁndings from the relevant literature. As shown in Table1, different experimental conditions (e.g., the alloy composition and heat treatment conditions) complicate the analysis. Although the corrosion of Al-Mg-Si alloys is strongly affected by the chemical composition, size, and distribution of precipitates and intermetallics, the corrosion mechanism of these alloys still has no suitable explanation, which may originate from the different chemical compositions and heat treatment conditions used in the various studies.

Table 1. Summary of the corrosion behaviours of Al-Mg-Si alloys.

Alloys (Composition Heat Treatment Corrosion Form Corrosion Mechanism in wt %) Conditions 520 ◦C/6 h (ST) + Propagation of corrosion pits in the corrosive media occurs Al-6.34Mg-3.66Si IGC aged at 200 ◦C/6 h along the interface between the MgSi particles and α-Al [27]. The ratio of Mg to Corrosion initiates on the MgSi surface and PFZ. Corrosion Al-0.63Mg-0.28Si and ST + aged at 175 Si less than 1.73 develops along the grain boundary PFZ at the adjacent MgSi Al-0.63Mg-0.88Si ◦C/1 h resulted in IGC precipitates [28]. IGC caused by micro-galvanic coupling between the Al-1.31Si-0.4Mg and 540 ◦C/30 min (ST) Aged sample is cathodic Cu-containing precipitates and solute-depleted Al-0.6Si-0.52Mg-0.18Cu + aged at 185 ◦C susceptible to IGC active zone [29]. 550 ◦C/ST + aged Very strong dissolution of the Al-Mg- and Si-containing Al-0.86Mg-0.92Si - at 175 ◦C/8 h particles along the grain boundaries [30]. Al-12Si-0Mg and Al-12Si-5Mg and Corrosion primarily initiates from the Al matrix adjacent to Casting - Al-12Si-10Mg and the primary Mg2Si particles [31]. Al-12Si-20Mg 550 ◦C/1 h + aged Precipitation of MgSi at the grain boundary favours Al-0.4Mg-1.0Si IGC at 175 ◦C/16 h intergranular corrosion attack [32]. MgSi particles were anodically precipitated compared to the Al-0.6Mg-0.5Si T6 - Al matrix in acidic and neutral pH [33]. IGC: intergranular corrosion; PFZ: precipitate-free zone; ST: solution treatment. Metals 2017, 7, 387 3 of 12 Metals 2017, 7, 387 3 of 12

In this study, the effecteffect ofof thethe SiSi contentcontent onon thethe hardnesshardness behaviourbehaviour and IGC of Al-Mg-Si alloys was investigatedinvestigated using using electrochemical electrochemical and and transmission transmission electron electron microscopy microscopy (TEM) (TEM) analyses, analyses, paying payingparticular particular attention attention to the evolutionto the evolution of the IGCof the process IGC process and the and corrosion the corrosion behaviour behaviour of the of MgSi the MgSiparticles particles during during exposure exposure of an Al-Mg-Siof an Al-Mg-Si alloy in alloy 3.5 wt in %3.5 NaCl wt % solution. NaCl solution. In addition, In addition, the artiﬁcial the artificialage-hardening age-hardening behaviour behaviour of Al-Mg-Si of Al-Mg-Si alloys andalloys the and structure the structure of the of main the main precipitates precipitates were were also alsoinvestigated investigated in detail. in detail.

2. Experimental Materials and Methods

2.1. Materials and Procedures The experimental materials used in this study were nominal Al-Mg-Si al alloysloys with various Si contents balanced with pure Al (wt(wt %). The measured chemical compositions of the alloys (shown in Table2 2)) werewere determineddetermined byby inductivelyinductively coupledcoupled plasmaplasma opticaloptical emissionemission spectrometryspectrometry (ICP-OES).(ICP-OES). ◦ The slabsslabs werewere prepared prepared by by common common chill chill casting, casting, scalped, scalped, andhomogenized and homogenized at 520 atC before520 °C solution before solutiontreatment. treatment. Samples Samples with different with different diameters diameter useds for used subsequent for subsequent characterization characterization were cutwere from cut ◦ fromthe middle the middle section section of hot of rolling hot rolling slabs (75% slabs hot (75% deformation hot deformation in four passesin four atpasses 460 C).at 460 The °C). solution The ◦ solutiontreatment treatment was conducted was conducted in an air circulationin an air circul furnaceation at 520furnaceC for at 3520 h, and°C thenfor 3 the h, samplesand then were the ◦ watersamples quenched were water in cold quenched water at in a temperaturecold water at below a temperature 30 C. Afterwards, below 30 all °C. the Afterwards, samples were all aged the ◦ samplesat 170 C were for different aged at times 170 to°C facilitate for different metastable time strengthenings to facilitate phasemetastable precipitation. strengthening The hardness phase precipitation.measurements The were hardness performed measurements on an HV-10B were machine performed (Huangqi, on Nanjing,an HV-1 China)0B machine at a load (Huangqi, of 500 g withNanjing, a dwell China) time at of a 10load s. Eachof 500 data g with point a dwell in the hardnesstime of 10 curves s. Each was data the point average in the of hardness 10 indentations. curves was the average of 10 indentations. Table 2. Chemical compositions of the examined alloys (wt %). Table 2. Chemical compositions of the examined alloys (wt %). Chemical Compositions Sample Chemical Compositions Sample Mg SiMg FeSi Fe Cr Cr Mn Mn Zr Zr Ti Ag Ti Al Ag Al A 1.91A 1.21 1.91 ≤1.210.1 ≤0.1 0.13 0.13 0.290.29 0.098 0.098 0.1 0.1 0.1 Bal. 0.1 Bal. ≤ B 1.91B 1.73 1.91 1.730.1 ≤0.1 0.12 0.12 0.290.29 0.11 0.11 0.1 0.1 0.1 Bal. 0.1 Bal. C 1.90 2.52 ≤0.1 0.14 0.28 0.10 0.1 0.1 Bal. C 1.90 2.52 ≤0.1 0.14 0.28 0.10 0.1 0.1 Bal.

The backscatteredbackscattered electronelectron images images of of the the three three alloys alloys and and corresponding corresponding energy energy dispersive dispersive X-ray X-rayspectroscopy spectroscopy (EDS) results(EDS) areresults shown are inshown Figure in1 .Figu Obviously,re 1. Obviously, a high number a high of number black areas of black identiﬁed areas identifiedas MgSi particles as MgSi were particles identiﬁed were by identified EDS. Since by MgSiEDS. particlesSince MgSi are hardparticles and are brittle, hard a portionand brittle, of the a portionparticles of fell the off particles the aluminium fell off the matrix aluminium in the grindingmatrix in process.the grinding In addition, process. elemental In addition, Si pitselemental could Sihave pits also could fell. have In addition also fell. to theIn MgSiaddition precipitates, to the MgSi the emissionprecipitates, scanning the emission electron microscopyscanning electron (SEM) microscopyimages contained (SEM) a images small number contained of visiblea small whitenumber particles, of visible as shownwhite particles, in the white as shown areas, in which the white were areas,mainly which Al-Fe-Mn-Si were intermetallics,mainly Al-Fe-Mn-Si including intermetallics, cubic (Al15(Fe,Mn) including3Si2) and cubic monoclinic (Al15(Fe,Mn) (A15FeSi)3Si2 phases.) and monoclinicThese intermetallics (A15FeSi) have phases. high These melting intermetallics points and largehave sizes,high andmelting are formedpoints and in the large process sizes, of and casting are formedand cannot in the be process completely of casting dissolved and duringcannot thebe co solutionmpletely treatment. dissolved during the solution treatment.

(a) (b) (c) 1

Figure 1. Cont.

Metals 2017, 7, 387 4 of 12 Metals 2017, 7, 387 4 of 12

(d) (e)

Figure 1. SEM images of the samples: (a) alloy A, (b) alloy B, and (c) alloy C. (d,e) Corresponding Figure 1. SEM images of the samples: (a) alloy A, (b) alloy B, and (c) alloy C. (d,e) Corresponding energy dispersivedispersive X-rayX-ray spectroscopy spectroscopy (EDS) (EDS) results results of particlesof particles from from mark mark 1 and 1 2,and respectively. 2, respectively. (ZAF: Correction(ZAF: Correction for Quantitative for Quantitative Electron-probe Electron-probe Microanalysis Microanalysis of Microparticles). of Microparticles).

2.2. Corrosion Tests IGC tests were performed according to British standard 11,846 method B [34]. The standard IGC tests were performed according to British standard 11,846 method B [34]. The standard involved the following steps: degreasing in ethanol, 5 min etching in 10 wt % NaOH, and 2 min involved the following steps: degreasing in ethanol, 5 min etching in 10 wt % NaOH, and 2 min desmutting in 30 wt % HNO3, followed by a 24 h immersion in an acidified sodium chloride solution desmutting in 30 wt % HNO3, followed by a 24 h immersion in an acidified sodium chloride solution (30 g NaCl and 10 mL HCl per litre). The surface and cross-sectional morphologies of the corroded (30 g NaCl and 10 mL HCl per litre). The surface and cross-sectional morphologies of the corroded samples were observed by an XJG-05 optical microscope (OM, Chincan, Dongguan, China) and samples were observed by an XJG-05 optical microscope (OM, Chincan, Dongguan, China) and JEOL JEOL JSM 6490 SEM instrument (JSM, Tokyo, Japan). For each condition, at least five valid parallel JSM 6490 SEM instrument (JSM, Tokyo, Japan). For each condition, at least five valid parallel tests tests were performed to obtain statistical data of the IGC depth. Different immersion times were were performed to obtain statistical data of the IGC depth. Different immersion times were applied to applied to clarify the evolution of the IGC process and the corrosion behaviour of MgSi particles clarify the evolution of the IGC process and the corrosion behaviour of MgSi particles during exposure during exposure of the Al-Mg-Si alloys in 3.5 wt % NaCl solution. Subsequently, the corroded of the Al-Mg-Si alloys in 3.5 wt % NaCl solution. Subsequently, the corroded surface appearance and surface appearance and corrosion depth on the cross-section were examined. corrosion depth on the cross-section were examined. Potentiodynamic polarization measurements were performed using a conventional Potentiodynamic polarization measurements were performed using a conventional three-electrode three-electrode electrochemical cell in 3.5 wt % NaCl solution at 25 ± 1 °C. A saturated calomel electrochemical cell in 3.5 wt % NaCl solution at 25 ± 1 ◦C. A saturated calomel electrode (SCE) was electrode (SCE) was used as the reference electrode, and a platinum sheet served as the auxiliary used as the reference electrode, and a platinum sheet served as the auxiliary electrode. Samples with a electrode. Samples with a size of 10 mm length × 10 mm width × 8 mm thickness were cut. Next, the size of 10 mm length × 10 mm width × 8 mm thickness were cut. Next, the polished face (L-T planes) polished face (L-T planes) was swept from an initial potential of −1.2 V to a final potential of 0.2 V at was swept from an initial potential of −1.2 V to a final potential of 0.2 V at a scanning rate of 2 mV/s a scanning rate of 2 mV/s using an EG&G Model 273 Galvanostat/Potentiostat (GAMRY Reference using an EG&G Model 273 Galvanostat/Potentiostat (GAMRY Reference 3000, Pennsylvania, PA, 3000, Pennsylvania, PA, USA). The software plotted the polarization curve and calculated the USA). The software plotted the polarization curve and calculated the corrosion current density (Icorr) corrosion current density (Icorr) and corrosion potential (Ecorr). and corrosion potential (Ecorr).

2.3. Characterization The compositions and morphologies of the microstructuresmicrostructures in the corroded samples were measured by an SEM instrument (JSM, Tokyo, Japan) equipped with EDS functionality. The detailed characteristics ofof thethe microstructuresmicrostructures of of the the three three investigated investigated alloys alloys were were further further observed observed by anby FEIan TecnaiFEI Tecnai G2 F20 G2 TEMF20 TEM instrument instrument (FEI, (FEI, Hillsboro, Hillsboro, OR, USA)OR, USA) operating operating at 200 at kV. 200 The kV. samples The samples for TEM for characterizationTEM characterization were preparedwere prepared by mechanically by mechanically grinding grinding to 75 µtom–95 75 μm–95µm, then μm, punchingthen punching into discs. into discs. The discs were perforated in an electrolyte consisting of 30% HNO3 and 70% methanol by The discs were perforated in an electrolyte consisting of 30% HNO3 and 70% methanol by Twin-Jet electropolishing.Twin-Jet electropolishing. The electropolishing The electropolishing conditions conditions were T = were−25 ◦ TC = and −25V °C= 22and V. V = 22 V.

3. Results

◦ 3.1. Hardness Evolution During Ageing at 170 °CC The hardness evolutions for the three investigatedinvestigated Al-Mg-Si alloys with different Si contents ◦ treated byby artificial artificial ageing ageing at 170at 170C immediately°C immediately after quenchingafter quenching are shown are shown in Figure in2 .Figure The hardness 2. The curveshardness of thecurves three of alloys the three displayed alloys similar displayed profiles, similar featuring profiles, a sharp featuring rise in a hardness sharp rise within in hardness the first threewithin hours, the first followed three byhours, a slow followed increase by during a slow the in nextcrease three during hours, the and next finally three reaching hours, and maximum finally reaching maximum peak-aged hardness values of 123 HV, 127 HV, and 129 HV for alloys A, B, and C, respectively. The hardness remained at a relatively constant state for a long time, and with an

Metals 2017, 7, 387 5 of 12 peak-aged hardness values of 123 HV, 127 HV, and 129 HV for alloys A, B, and C, respectively. The hardness remained at a relatively constant state for a long time, and with an increase in the ageing time, the hardness value gradually decreased. However, the required ageing time to obtain the maximum peak-aged hardness value was different for each alloy, and the ageing kinetics of the Metals 2017, 7, 387 5 of 12 three alloys were different, as the increased Si content accelerated age-hardening. The age-hardening speed of alloyincrease A in with the ageing 1.21 wt time, % the Si washardness the value lowest gradually amongst decreased. the three However, alloys, the required reaching ageing a peak value time to obtain the maximum peak-aged hardness value was different for each alloy, and the ageing of 123 HVMetals after 2017 ageing, 7, 387 for 4.5 h. By increasing the Si content to 1.73 wt %, the age-hardening5 of 12 rate kinetics of the three alloys were different, as the increased Si content accelerated age-hardening. The and peak hardnessincreaseage-hardening in the value ageingspeed were of time, alloy both the A withhardness enhanced; 1.21 wtvalue % after Si gr wasadually 4 the h oflowest decreased. ageing, amongst However, alloy the Bthree reached the alloys, required areaching peak ageing hardnessa of 127 HV. However,timepeak tovalue obtain increasing of the 123 maximum HV the after Si peak-aged contentageing for hardness to 4.5 2.52 h. wtvaluBy %increasinge was different not the beneﬁcial Sifor content each alloy, to to further 1.73and thewt enhance ageing%, the the peak hardness value,kineticsage-hardening of as the the three rate hardness andalloys peak were of hardness thedifferent, sample value as thwere wase increased bo notth enhanced; signiﬁcantly Si content after accelerated 4 greaterh of ageing, age-hardening. than alloy alloy B reached B.The In contrast, age-hardeninga peak hardness speed of 127 of alloyHV. However,A with 1.21 increasing wt % Si was the the Si lowestcontent amongst to 2.52 wtthe % three was alloys, not beneficial reaching to a the hardness plateau became less and indistinct. From previous studies described in References [35,36], peakfurther value enhance of 123 the peakHV after hardness ageing value, for as4.5 the h. ha Byrdness increasing of the samplethe Si wascontent not significantlyto 1.73 wt greater%, the 00 peak hardnessage-hardeningthan alloy was B. shown In rate contrast, and to peak be the mainly hardness hardness due value plateau to were the became bo presenceth enhanced; less and of high-density afterindistinct. 4 h of ageing,From (Mg previous alloy + Si)B reached studies clusters and β phase (Mgadescribed5 peakSi6) precipitates,hardness in References of 127 and HV. [35,36], anHowever, increased peak increasinghardness Si content was the Sishown content increased to beto 2.52mainly the wt number %due was to notthe of beneficial thesepresence strengthening toof phases. Tofurtherhigh-density further enhance investigate (Mg the + peak Si) clustershardness strengthening and value, β′′ asphase the phase ha(Mgrdness5Si microstructures6) ofprecipitates, the sample andwas and annot increased significantly the difference Si contentgreater among the thanincreased alloy theB. Innumber contrast, of thethese hardness strengthening plateau phasbecamees. To less further and indistinct. investigate From strengthening previous studies phase peak-aged hardness values, the microstructures of the three alloys after peak-ageing treatment were describedmicrostructures in References and the difference[35,36], peak among hardness the peak-a wasged shown hardness to be values, mainly the due microstructures to the presence of theof characterized,high-densitythree alloys as shown after (Mg peak-ageing below.+ Si) clusters treatment and β′′ werephase characterized, (Mg5Si6) precipitates, as shown below.and an increased Si content increased the number of these strengthening phases. To further investigate strengthening phase microstructures and the difference among the peak-aged hardness values, the microstructures of the 130 three alloys after peak-ageing treatment were characterized, as shown below.

120 130 110 A 120 B C

Hardness(HV) 100 110 A 90 B 0 200 400 600 800 C

Hardness(HV) 100 Aging Time(min)

Figure 2. Vickers hardness evolution90 of the three alloys with ageing time at 170 °C immediately◦ after 0 200 400 600 800 Figure 2. Vickersquenching hardness (error limit evolution ±2%). of the three alloys with ageing time at 170 C immediately after quenching (error limit ±2%). Aging Time(min) 3.2. MicrostructureFigure 2. Vickers hardness evolution of the three alloys with ageing time at 170 °C immediately after quenching (error limit ±2%). 3.2. MicrostructureFor the studied Al-Mg-Si alloys, the peak-aged state persisted for a long time. To acquire distinct microstructure views of the peak-aged states of the three alloys, all alloys were treated at 170 3.2. Microstructure For the°C studiedfor 5 h (at Al-Mg-Si the peak-aged alloys, state). the Figure peak-aged 3 shows statethe bright-field persisted TEM for images a long of time. the three To acquirealloys distinct ◦ microstructureafterFor being views the peak-aged studied of the Al-Mg-Siat peak-aged 170 °C foralloys, 5 statesh takenthe peak-aged ofwith the the three electron state alloys,persisted beam allparallel for alloys a tolong the were time.[001] treated AlTo zone acquire axis. at170 C for 5 h (at thedistinctA peak-aged high-density microstructure state). of rod-shaped Figure views of3 theshowsprecipitates peak-aged the bright-ﬁeld (marstatesked of thewith three TEM the alloys, black images all arrow) alloys of the wereand three cross-sectionstreated alloys at 170 after being °C(marked for 5 hwith (at thethe peak-aged white arrow) state). we Figurere dispersed 3 shows in th alle bright-fieldthe TEM images. TEM images As clearly of the observed, three alloys an peak-aged at 170 ◦C for 5 h taken with the electron beam parallel to the [001] zone axis. A high-density afterincreased being Si peak-aged content resulted at 170 °C in foran increased5 h taken withnumb theer andelectron density beam of theseparallel precipitates, to Althe [001] andAl zone therefore axis. of rod-shapedAthe high-density hardness precipitates and of strength rod-shaped (marked of the withalloysprecipitates theaccordingl black (mary arrow)ked improved, with and the which cross-sectionsblack is consistentarrow) and with (marked cross-sections the previous with the white arrow) were(markedliterature dispersed with[37,38]. the in white all the arrow) TEM we images.re dispersed As clearly in all the observed, TEM images. an increasedAs clearly Siobserved, content an resulted in an increasedincreased number Si content and densityresulted in of an these increased precipitates, number and and density therefore of thesethe precipitates, hardness and and therefore strength of the the hardness and strength of the alloys accordingly improved, which is consistent with the previous alloys accordinglyliterature [37,38]. improved, which is consistent with the previous literature [37,38].

Figure 3. TEM micrographs of the three alloys at peak-ageing conditions: (a) alloy A, (b) alloy B, and (c) alloy C.

Figure 3. TEM micrographs of the three alloys at peak-ageing conditions: (a) alloy A, (b) alloy B, and Figure 3. TEM(c) alloy micrographs C. of the three alloys at peak-ageing conditions: (a) alloy A, (b) alloy B, and (c) alloy C. Metals 2017, 7, 387 6 of 12

Metals 2017, 7, 387 6 of 12 To further analyse and conﬁrm the structure of the rod-shaped precipitates in the three peak-aged To further analyse and confirm the structure of the rod-shaped precipitates in the three alloys, a high-resolution TEM (HRTEM) image and corresponding fast Fourier ﬁltering transform peak-aged alloys, a high-resolution TEM (HRTEM) image and corresponding fast Fourier filtering (FFT) patterntransform of (FFT) the cross-section pattern of the cross-section of a rod-shaped of a rod-shaped precipitate precipitate were obtained,were obtained, as shown as shown in in Figure 4. The precipitateFigureMetals 2017 4. was,The 7, 387 precipitate polygonal was and polygonal belonged and to belonged a monoclinic to a monoclinic system withsystem lattice with lattice parameters parameters6a of= 12 1.51 nm, c = 0.67a = nm, 1.51 nm, and c β= 0.67= 105.3nm, and◦ (Figure β = 105.3°4a). (Figure The 4a). aforementioned The aforementioned analysis analysis results results show show that that the To further analyse and confirm the structure of the rod-shaped precipitates in the three the precipitate 00was the β′′ phase through comparison with the published database [39] and the precipitatepeak-aged was the alloys,β phase a high-resolution through comparison TEM (HRTEM) with image the publishedand corresponding database fast [39 Fourier] and thefiltering orientation orientation relationships of (200)β′′ parallel to (301)Al. The main precipitate contributing to the peak relationshipstransform of (200)(FFT)β pattern00 parallel of the to cross-section (301)Al. The of maina rod-shaped precipitate precipitate contributing were obtained, to the as peak shown hardness in of hardness of the alloys was the β′′ precipitate elongated along the <001>Al axis with a rounded the alloysFigure was 4. the Theβ precipitate00 precipitate was polygonal elongated and along belonged the <001>to a monoclinicaxis withsystem a roundedwith lattice cross-section. parameters cross-section. Al a = 1.51 nm, c = 0.67 nm, and β = 105.3° (Figure 4a). The aforementioned analysis results show that the precipitate was the β′′ phase through comparison with the published database [39] and the orientation relationships of (200)β′′ parallel to (301)Al. The main precipitate contributing to the peak hardness of the alloys was the β′′ precipitate elongated along the <001>Al axis with a rounded cross-section.

(a) (b)

Figure 4. High-resolution TEM (HRTEM) images of a representative β′′ precipitate imaged parallel to Figure 4. High-resolution TEM (HRTEM) images of a representative β00 precipitate imaged parallel to the <010>Al axis of the precipitates in the peak-aged (170 °C/5 h) Al-Mg-Si alloy: (a) HRTEM image the <010> axis of the precipitates in the peak-aged (170 ◦C/5 h) Al-Mg-Si alloy: (a) HRTEM image andAl (b) the corresponding fast Fourier filtering transform (FFT) pattern. and (b) the corresponding fast Fourier(a) filtering transform (FFT) pattern.(b) The composition and distribution of grain boundary particles in aluminium alloys play importantFigure roles 4. High-resolution in obtaining TEM good (HRTEM) corrosion images resistan of ace representative properties. β′′The precipitate influence imaged of the parallel Si content to in The composition and distribution of grain boundary particles in aluminium alloys play important the Al-Mg-Sithe <010> Alalloys axis ofon the the precipitates grain boundary in the peak microstructure-aged (170 °C/5 is shownh) Al-Mg-Si in Figure alloy: 5.(a )Alloy HRTEM A (Figure image 5a) roles indid obtaining notand exhibit (b) the good obviouscorresponding corrosion grain fastboundary resistance Fourier precipitates,filtering properties. transform and The the(FFT) influencegrain pattern. boundaries of the of Si alloys content B (Figure in the 5b) Al-Mg-Si alloys onand the C grain(Figure boundary 5c) were decorated microstructure with a mass is shown of particles, in Figure but the5. AlloyPFZs of A all (Figure the alloys5a) didwere not not exhibit The composition and distribution of grain boundary particles in aluminium alloys play obviousobvious grain boundary near the grain precipitates, boundaries. and These the grain grain boun boundariesdary precipitates of alloys observed B (Figure in the5b) TEM and images C (Figure 5c) important roles in obtaining good corrosion resistance properties. The influence of the Si content in consisted of two types of particles: (1) particles larger than 500 nm and (2) particles less than 100 nm, were decoratedthe Al-Mg-Si with alloys a mass on the of particles,grain boundary but themicrostructure PFZs of all is the shown alloys in Figure were not5. Alloy obvious A (Figure near 5a) the grain but only a few particles larger than 500 nm were found on the grain boundaries in these alloys. The boundaries.did not These exhibit grain obvious boundary grain boundary precipitates precipitates, observed and the in thegrain TEM boundaries images ofconsisted alloys B (Figure of two 5b) types of large particles observed in all the alloys were Al-Fe-Si (Mn) dispersoids [40]. The numerous small and C (Figure 5c) were decorated with a mass of particles, but the PFZs of all the alloys were not particles:rod-shaped (1) particles particles larger on the than grain 500 boundaries nm and (2) in particlesall three alloys less thanwere 100MgSi nm, precipitates but only with a few sizes particles obvious near the grain boundaries. These grain boundary precipitates observed in the TEM images larger thanranging 500 from nm were0.1 μm–0.15 found μ onm [40]. the grainHowever, boundaries the structure in these of thes alloys.e MgSi Theprecipitates large particles is not studied observed in consisted of two types of particles: (1) particles larger than 500 nm and (2) particles less than 100 nm, all the alloysin this paper; were Al-Fe-Sithis will be (Mn) further dispersoids investigated [ and40]. discussed The numerous in future small research. rod-shaped With an increased particles Si on the but only a few particles larger than 500 nm were found on the grain boundaries in these alloys. The content, the MgSi particle density on the boundaries increased, resulting in a more continuousµ µ grain boundarieslarge particles inall observed three alloys in all the were alloys MgSi were precipitates Al-Fe-Si (Mn) with dispersoids sizes ranging [40]. The from numerous 0.1 m–0.15 small m [40]. distribution. In addition, solute-depleted zones of Mg and Si inevitably existed near the MgSi However,rod-shaped the structure particles of on these the grain MgSi boundaries precipitates in all is three not alloys studied were in MgSi this precipitates paper; this with will sizes be further particles, although the PFZs could not be identified in these TEM images. investigatedranging and from discussed 0.1 μm–0.15 in futureμm [40]. research. However, With the structure an increased of these Si MgSi content, precipitates the MgSi is not particle studied density on the boundariesin this paper; increased, this will be resultingfurther investigated in a more and continuous discussed in distribution. future research. In With addition, an increased solute-depleted Si zones ofcontent, Mg and the Si MgSi inevitably particle existed density near on thethe MgSiboundaries particles, increased, although resulting the PFZsin a more could continuous not be identified distribution. In addition, solute-depleted zones of Mg and Si inevitably existed near the MgSi in these TEM images. particles, although the PFZs could not be identified in these TEM images.

(a)(b)(c)

Figure 5. TEM images showing the distributions of grain boundary particles in (a) alloy A, (b) alloy B, and (c) alloy C.

(a)(b)(c)

Figure 5. TEM images showing the distributions of grain boundary particles in (a) alloy A, (b) alloy Figure 5. TEM images showing the distributions of grain boundary particles in (a) alloy A, (b) alloy B, B, and (c) alloy C. and (c) alloy C. Metals 2017, 7, 387 7 of 12

3.3. Corrosion Behaviour Figure 6 displays the corroded morphologies of the cross-sections and surfaces of the peak-aged (170 °C/5 h) Al-Mg-Si alloys after immersion in 3.5 wt % NaCl solution for 12 h, as per British standard 11,846 method B. All the specimens displayed distinctly different degrees of IGC. The corrosion level was determined by the maximum IGC depth [34], and the values of the IGC depths and corrosion levels are presented in Table 3. Clearly, the typical IGC layers of all the samples were homogeneously distributed, but the degrees of corrosion varied. The IGC depth of alloy A with 1.21 wt % Si was 65 μm, and the entire sample presented superficial IGC (Figure 6a,d). With an increased Si content to Metals1.73 wt2017 %,, 7, the 387 IGC depth of alloy B increased, and some particles were stripped from the Al matrix7 of 12 (Figure 6b,e). Serious corrosion etched alloy C with a Si content of 2.52 wt %, displaying an IGC depth of 96 μm (Figure 6c,f). These different IGC depths demonstrated that alloy C presented the 3.3. Corrosion Behaviour highest IGC susceptibility among the three alloys, as the increased Si content weakened the IGC corrosionFigure 6resistance displays theof corrodedthe Al-Mg-Si morphologies alloys, wh ofich the may cross-sections be correlated and surfaces to the ofdensity the peak-aged of MgSi (170precipitates◦C/5 h) Al-Mg-Sion the grain alloys boundaries. after immersion Alloys in B 3.5 and wt %C NaCldisplayed solution similar for 12 sizes, h, as perand Britishmaximum standard IGC 11,846depths method (Table 3) B. were determined at the same IGC level.

(a) (b)(c)

(d) (e)(f)

FigureFigure 6.6.Typical Typical IGC IGC cross-sections cross-sections and correspondingand corresponding surface surface morphologies morphologies of the corroded of the Al-Mg-Sicorroded alloysAl-Mg-Si with alloys different with Si different contents: Si ( acontents:,d) alloy ( A,a,d (b) alloy,e) alloy A, ( B,b, ande) alloy (c,f )B, alloy and C.(c,f) alloy C.

Table 3. Maximum IGC depths for the Al-Mg-Si alloys at peak-aged conditions. All the specimens displayed distinctly different degrees of IGC. The corrosion level was determined by the maximumAlloy IGC Maximum depth [ 34Corrosion], and the Depth values (μ ofm) the IGC depthsLevel and corrosion levels are presented in Table3.A Clearly, the typical IGC65 layers of all the samples 3 were homogeneously distributed, but the degreesB of corrosion varied.89 The IGC depth of alloy 3 A with 1.21 wt % Si was 65 µm, and the entire sampleC presented superﬁcial96 IGC (Figure6a,d). With 4 an increased Si content to 1.73 wt %, the IGC depth of alloy B increased, and some particles were stripped from the Al matrix (Figure6b,e). Serious corrosion etched alloy C with a Si content of 2.52 wt %, displaying an IGC depth of 96 µm (Figure6c,f). These different IGC depths demonstrated that alloy C presented the highest IGC susceptibility among the three alloys, as the increased Si content weakened the IGC corrosion resistance of the Al-Mg-Si alloys, which may be correlated to the density of MgSi precipitates on the grain boundaries. Alloys B and C displayed similar sizes, and maximum IGC depths (Table3) were determined at the same IGC level.

Table 3. Maximum IGC depths for the Al-Mg-Si alloys at peak-aged conditions.

Alloy Maximum Corrosion Depth (µm) IGC Level A 65 3 B 89 3 C 96 4 Metals 2017, 7, 387 8 of 12 Metals 2017, 7, 387 8 of 12

3.4. Potentiodynamic Polarization Tests 3.4. Potentiodynamic Polarization Tests The potentiodynamic polarization curves of the three peak-aged samples are shown in Figure 7, The potentiodynamic polarization curves of the three peak-aged samples are shown in Figure7, and the polarization features (e.g., corrosion voltages, Ecorr, and corrosion current densities, Icorr) of and the polarization features (e.g., corrosion voltages, Ecorr, and corrosion current densities, Icorr) of the three studied alloys calculated from potentiodynamic polarization curves are presented in Table the three studied alloys calculated from potentiodynamic polarization curves are presented in Table4. 4. The shapes of the polarization curves were largely unchanged with an increasing Si content. The The shapes of the polarization curves were largely unchanged with an increasing Si content. The anodic polarization behaviour was dictated by activated anodic dissolution with a sharp increase in anodic polarization behaviour was dictated by activated anodic dissolution with a sharp increase in current density at potentials higher than the pitting potential (Epit). As shown in Table 4, Icorr current density at potentials higher than the pitting potential (Epit). As shown in Table4, Icorr increased increased with an increasing Si content, which is consistent with the IGC test results. with an increasing Si content, which is consistent with the IGC test results.

0.0

-0.2

-0.4

-0.6

-0.8 Al-1.91Mg-1.21Si Al-1.91Mg-1.73Si -1.0

Potential vs. SCE (V) Al-1.90Mg-2.52Si

-1.2 -8 -6 -4 -2 0 2 Log current density (A/cm )

FigureFigure 7. 7. Polarization curves ofof thethe peak-aged peak-aged Al-Mg-Si Al-Mg-Si alloys alloys with with different different Si contentsSi contents tested tested in IGC in IGCsolution. solution. SCE: SCE: saturated satura calomelted calomel electrode. electrode.

TableTable 4. EEcorrcorr andand IcorrIcorr valuesvalues of of the the three three alloys alloys from from the the po polarizationlarization curves curves determined determined by by the the Tafel Tafel extrapolationextrapolation method. method.

CorrosionCorrosion Potential Potential,, CorrosionCorrosion Current Current Density, Pitting Potential, AlloyAlloy 2 Pitting Potential, Epit (V) EEcorrcorr (mV)(mV) Density,IcorrI corr(μA/cm(µA/cm2) ) Epit (V) AA −532532 ± 66 0.47 0.47± ±0.02 0.02 –0.408 –0.408 BB −612612 ± 44 0.55 0.55± ±0.06 0.06 –0.440 –0.440 CC −606606 ± 66 0.57 0.57± ±0.04 0.04 –0.453 –0.453

3.5.3.5. Corrosion Corrosion Process Process TheThe principle principle of of IGC IGC in in age-hardened age-hardened Al Al alloys alloys is is closely closely related related to to electrochemical electrochemical corrosion, corrosion, resultingresulting from from composition composition segregation segregation and and electrochemical electrochemical microcouples microcouples continuously continuously distributed distributed alongalong the the grain grain boundaries. boundaries. As As previously previously stated, stated, we we concluded concluded that that the the Si Si content content of of the the Al-Mg-Si Al-Mg-Si alloysalloys was was a a predominant predominant factor factor in in the the corrosion corrosion sp speedeed and and level. level. Despite Despite the the different different IGC IGC levels levels of theof thethree three alloys, alloys, their their IGC IGC mechanisms mechanisms are arecons consistent.istent. To uncover To uncover the thecommon common IGC IGC mechanism mechanism of theof theAl-Mg-Si Al-Mg-Si alloys alloys with with high high Si Sicontents, contents, we we selected selected alloy alloy B Bas as a arepresentative representative sample sample with with excellentexcellent integrated integrated performance performance to to elaborate elaborate and and analyse analyse the the corrosion corrosion process. process. SEM SEM images images of of a a freshlyfreshly polished polished alloy alloy B Bsurface surface aged aged at at170 170 °C◦ Cfor for 5 h 5 and h and subsequently subsequently exposed exposed to a to solution a solution of 3.5 of wt3.5 % wt NaCl % NaCl for forvarious various times times are are shown shown in inFigure Figure 8.8 The. The corroded corroded alloy alloy morphologies morphologies with with different different exposureexposure times times showed showed the the IGC IGC process process via via the the corrosion corrosion morphology morphology and and provided provided an anoverview overview of theof thecorrosion corrosion evolution evolution on the on thelocal local surface surface and andcross-section cross-section morphology morphology of the of Al-Mg-Si the Al-Mg-Si alloy. alloy. At theAt beginning the beginning of the of corrosion the corrosion process, process, corrosion corrosion occurred occurred at and at around and around the MgSi the surface, MgSi surface, and some and corrosionsome corrosion trenching trenching appeared appeared around around the particles. the particles. With With an anincreasing increasing corrosion corrosion time, time, the the alloy alloy presentedpresented obvious obvious IGC, IGC, and and the the internal internal gr grainsains also also displayed displayed many corroded pits.

Metals 2017, 7, 387 9 of 12 Metals 2017, 7, 387 9 of 12

Figure 8. Corrosion evolution of the Al-Mg-Si alloy after immersion in 3.5 wt % NaCl solution. Figure 8. Corrosion evolution of the Al-Mg-Si alloy after immersion in 3.5 wt % NaCl solution.

4. Discussion 4. Discussion To further clarify the IGC susceptibility and corrosion process of the high-Si-content alloy B at To further clarify the IGC susceptibility and corrosion process of the high-Si-content alloy B prolonged immersion time, two typical microstructural aspects must first be satisfied: (i) the grain at prolonged immersion time, two typical microstructural aspects must first be satisfied: (i) the boundary microstructures must contain a large number of discontinuously distributed MgSi grain boundary microstructures must contain a large number of discontinuously distributed MgSi precipitates and Al-Fe-Mn-Si intermetallics, and indistinguishable pure Si particles must exist; (ii) precipitates and Al-Fe-Mn-Si intermetallics, and indistinguishable pure Si particles must exist; (ii) large large differences must exist between the corrosion potentials of the grain boundary microstructures. differences must exist between the corrosion potentials of the grain boundary microstructures. Generally, the typical grain boundary microstructures in peak-aged Al-Mg-Si alloys mainly consist Generally, the typical grain boundary microstructures in peak-aged Al-Mg-Si alloys mainly consist of of MgSi precipitates, Al-Fe-Mn-Si intermetallics, pure Si, an Al matrix, and PFZ; however, the PFZ MgSi precipitates, Al-Fe-Mn-Si intermetallics, pure Si, an Al matrix, and PFZ; however, the PFZ was was not identified by TEM in this study. The concentration of Si and Mg atoms in the PFZ was lower not identified by TEM in this study. The concentration of Si and Mg atoms in the PFZ was lower than than the aluminium matrix and precipitates, resulting in a lower corrosion potential at these the aluminium matrix and precipitates, resulting in a lower corrosion potential at these solute-depleted solute-depleted zones than at the adjacent precipitates or Al matrix, which become the starting zones than at the adjacent precipitates or Al matrix, which become the starting points of corrosion. points of corrosion. Figure9 shows a schematic diagram of the IGC evolution of Al-Mg-Si alloys. When the peak-aged Figure 9 shows a schematic diagram of the IGC evolution of Al-Mg-Si alloys. When the samples were immersed in the corrosive solution, the corrosion priority initiated from the grain peak-aged samples were immersed in the corrosive solution, the corrosion priority initiated from the boundary PFZ adjacent to the intermetallic or precipitate peripheries, and the grain boundary PFZ grain boundary PFZ adjacent to the intermetallic or precipitate peripheries, and the grain boundary acted as an anode, resulting in anodic dissolution and the presence of a corrosion ring surrounding PFZ acted as an anode, resulting in anodic dissolution and the presence of a corrosion ring these Al-Fe-Mn-Si intermetallics and MgSi particles. Meanwhile, corrosion could also start at some surrounding these Al-Fe-Mn-Si intermetallics and MgSi particles. Meanwhile, corrosion could also intermetallics and precipitates due to preferential dissolution. After 4 h of immersion, these two main start at some intermetallics and precipitates due to preferential dissolution. After 4 h of immersion, types of particles corroded, and considerable trenches surrounding the MgSi particles and Al-Fe-Mn-Si these two main types of particles corroded, and considerable trenches surrounding the MgSi intermetallics were observed. During this corrosion process, the Si phase acted as a cathodic zone with particles and Al-Fe-Mn-Si intermetallics were observed. During this corrosion process, the Si phase strong polarization; however, the existence of pure Si in the grain boundary microstructures was not acted as a cathodic zone with strong polarization; however, the existence of pure Si in the grain found, and corrosion dynamic conversion existed in the MgSi precipitate particles. During corrosion boundary microstructures was not found, and corrosion dynamic conversion existed in the MgSi initiation, the MgSi precipitates underwent self-corrosion on their surface, resulting from the selective precipitate particles. During corrosion initiation, the MgSi precipitates underwent self-corrosion on preferential dissolution of Mg in chloride-containing environments. As the corrosion time extended, their surface, resulting from the selective preferential dissolution of Mg in chloride-containing the IGC continued into the corrosive MgSi phase, which led to Si gradually enriching the remnant environments. As the corrosion time extended, the IGC continued into the corrosive MgSi phase, MgSi particles, thus transforming polarity between the MgSi particles and solute-depleted zones and which led to Si gradually enriching the remnant MgSi particles, thus transforming polarity between resulting in the MgSi particles becoming cathodic to the PFZ. In other words, the electrochemical the MgSi particles and solute-depleted zones and resulting in the MgSi particles becoming cathodic driving force for the anodic dissolution of the solute-depleted zones increased, resulting in a larger to the PFZ. In other words, the electrochemical driving force for the anodic dissolution of the dissolution velocity of the alloy due to larger differences in the corrosion potential between the solute-depleted zones increased, resulting in a larger dissolution velocity of the alloy due to larger de-alloyed MgSi particles and anode. This conversion was also found in some particles containing differences in the corrosion potential between the de-alloyed MgSi particles and anode. This low corrosion potential elements. After 20 h of exposure to the corrosion solution, the corrosion conversion was also found in some particles containing low corrosion potential elements. After 20 h trenching width around the particles was extensive—even forming a continuous corrosion channel, of exposure to the corrosion solution, the corrosion trenching width around the particles was despite the fact that the MgSi precipitates were discontinuously distributed at the grain boundaries. extensive—even forming a continuous corrosion channel, despite the fact that the MgSi precipitates In addition, some intergranular particle remnants were removed. After 40 h of immersion, many were discontinuously distributed at the grain boundaries. In addition, some intergranular particle corroded small pits were found in the internal grains (submicro/nanometre range). These small pits remnants were removed. After 40 h of immersion, many corroded small pits were found in the internal grains (submicro/nanometre range). These small pits may have been dissolved in the MgSi strengthening phase, and were not detectable before etching the micro-pits in the SEM images

Metals 2017, 7, 387 10 of 12

Metals 2017, 7, 387 10 of 12 may have been dissolved in the MgSi strengthening phase, and were not detectable before etching thebecause micro-pits of their in infinitesimal the SEM images size. because As the ofcorrosi theiron inﬁnitesimal time extended, size. Ascorrosion the corrosion constantly time proceeded extended, corrosionalong the constantly corrosion proceededfront pathway. along theTherefore, corrosion we front uncovered pathway. a Therefore, common wecorrosion uncovered evolution a common by corrosionanalysing evolutionthe abovementioned by analysing process, the abovementioned which we hope process, proves which advantageous we hope proves in improving advantageous alloy inperformances. improving alloy performances.

Figure 9. Schematic diagram of the corrosion evolution of a cross-section of the Al-Mg-Si alloy. Figure 9. Schematic diagram of the corrosion evolution of a cross-section of the Al-Mg-Si alloy.

5. Conclusions 5. Conclusions In this study, the hardness behaviours and IGC susceptibilities of the three alloys after ageing at In this study, the hardness behaviours and IGC susceptibilities of the three alloys after ageing 170 °C for different times were investigated by hardness tests, electrochemical experiments, SEM, at 170 ◦C for different times were investigated by hardness tests, electrochemical experiments, SEM, and TEM. The precipitate microstructures and IGC evolutions of the peak-aged Al-Mg-Si alloys and TEM. The precipitate microstructures and IGC evolutions of the peak-aged Al-Mg-Si alloys were were uncovered, and the main conclusions can be summarized as follows: uncovered, and the main conclusions can be summarized as follows:

1.1. The hardness of the Al-Mg-Si alloys withwith peak-ageingpeak-ageing treatmentstreatments mainlymainly originatedoriginated fromfrom contribution from the ββ′′00 phase. With With an an increased Si content, the age-hardeningage-hardening responseresponse improved,improved, and the hardness value also increased by enhancing the quantity and density of thethe β′′β00 strengtheningstrengthening phase. phase. 2. The microstructures affecting the IGC performances the Al-Mg-Si alloys consisted of MgSi 2. The microstructures affecting the IGC performances the Al-Mg-Si alloys consisted of MgSi particles, Al-Fe-Mn-Si intermetallics, and the PFZ. The IGC susceptibility of the Al-Mg-Si alloys particles, Al-Fe-Mn-Si intermetallics, and the PFZ. The IGC susceptibility of the Al-Mg-Si alloys was mainly attributed to the high electrochemical potential difference between the MgSi was mainly attributed to the high electrochemical potential difference between the MgSi particles particles and solute-depleted zones. and solute-depleted zones. 3. Corrosion priority initiated from the grain boundary PFZ adjacent to the Al-Fe-Mn-Si 3. Corrosion priority initiated from the grain boundary PFZ adjacent to the Al-Fe-Mn-Si intermetallics or from the MgSi precipitate peripheries forming a trench around the particles. intermetallics or from the MgSi precipitate peripheries forming a trench around the particles. Meanwhile, some intermetallics and precipitates displayed self-corrosion until dislodging after Meanwhile, some intermetallics and precipitates displayed self-corrosion until dislodging after forming continuous corrosion channels. With an extended corrosion time, IGC constantly forming continuous corrosion channels. With an extended corrosion time, IGC constantly proceeded along the corrosion front pathway. proceeded along the corrosion front pathway. Acknowledgments: This study is supported by the Open Fund of Hunan Province Key Laboratory of Safety Acknowledgments:Design and ReliabilityThis Technology study is supportedfor Engineering by the Vehicl Opene No. Fund KF1604 of Hunan and the Province National Key Defense Laboratory Foundation of Safety of DesignChina No. and 2011-006. Reliability Technology for Engineering Vehicle No. KF1604 and the National Defense Foundation of China No. 2011-006. AuthorAuthor Contributions:Contributions: YayaYaya ZhengZheng andand BinghuiBinghui LuoLuo conceivedconceived andand designeddesigned thethe experiments;experiments; Yaya ZhengZheng performedperformed thethe experiments; experiments; Yaya Yaya Zheng Zheng and Zhenhaiand Zhenha Bai analysedi Bai analysed the data; the Juan data; Wang Juan and Wang Yuan Yinand contributed Yuan Yin reagents/materials/analysiscontributed reagents/materials/analysis tools; Yaya Zhengtools; Yaya wrote Zheng the paper. wrote the paper.

ConﬂictsConflicts ofof Interest:Interest: TheThe authorsauthors declaredeclare nono conﬂictconflictof ofinterest. interest.

ReferencesReferences

1.1 Hua, D.R.;D.R.; Yang,Yang, W.S.; W.S.; Yu, Yu, Z.H.; Z.H.; Wu, Wu, P.; Hussain,P.; Hussain, M.; Jiang,M.; Jiang, L.T.; Wu,L.T.; G.H.Wu, AgingG.H. Aging behavior behavior of 6061Al of 6061Al matrix compositematrix composite reinforced reinforced with high with content high content SiC nanowires. SiC nanowires.J. Alloys J. Compd. Alloys Compd2015, 649. 2015, 1037–1042., 649, 1037–1042. 2.2 Lin, C.W.;C.W.; Hung, Hung, F.Y.; F.Y.; Lui, T.S.;Lui, Chen, T.S.; L.H. Chen, Microstructure L.H. Microstructure Evolution and Evolution High-Temperature and High-Temperature Compressibility ofCompressibility Modiﬁed Two-Step of Modified Strain-Induced Two-Step Melt Strain-Induced Activation-Processed Melt Activation-Processed Al-Mg-Si Aluminum Al-Mg-Si Alloy. Metals Aluminum2016, 6, 113.Alloy. [CrossRef Metals 2016] , 6, 113. 3.3 Kaseem, M.; Choi, K.; Ko, Y.G. AA highlyhighly compact coatingcoating responsible for enhancing corrosion properties ofof Al-Mg-SiAl-Mg-Si alloy.alloy. Mater.Mater. Lett.Lett. 20172017,, 196196,, 316–319.316–319. [CrossRef] 4 Li, J.R.; Kai, W.; Jiang, Q.T.; Sun, H.Y.; Li, Y.T.; Hou, B.R.; Li, W.Z.; Liu, M. Corrosion and Discharge Behaviors of Mg-Al-Zn and Mg-Al-Zn-In Alloys as Anode Materials. Metals 2016, 6, 65.

Metals 2017, 7, 387 11 of 12

4. Li, J.R.; Kai, W.; Jiang, Q.T.; Sun, H.Y.; Li, Y.T.; Hou, B.R.; Li, W.Z.; Liu, M. Corrosion and Discharge Behaviors of Mg-Al-Zn and Mg-Al-Zn-In Alloys as Anode Materials. Metals 2016, 6, 65. [CrossRef] 5. Yin, D.; Xiao, Q.; Chen, Y.Q.; Liu, H.Q.; Yi, D.Q.; Wang, B.; Pan, S.P. Effect of natural ageing and pre-straining on the hardening behavior and microstructural response during artificial ageing of an Al-Mg-Si-Cu alloy. Mater. Des. 2016, 95, 329–339. [CrossRef] 6. Farlkoosh, A.R.; Pekguleryuz, M. Enhanced mechanical properties of an Al-Si-Cu-Mg alloy at 300 ◦C: Effects of Mg and the Q-precipitate phase. Mater. Sci. Eng. A 2015, 621, 277–286. [CrossRef] 7. Yin, M.J.; Chen, J.H.; Wang, S.B.; Liu, Z.R.; Cha, L.M.; Duan, S.Y.; Wu, C.L. Anisotropic and temperature-dependent growth mechanism of S-phase precipitates in Al-Cu-Mg alloy in relation with GPB zones. Trans. Nonferrous Met. Soc. China 2016, 26, 1–11. [CrossRef] 8. Xu, X.X.; Yang, Z.; Ye, Y.L.; Wang, G.X.; He, X.L. Effects of various Mg/Si ratios on microstructure and performance property of Al-Mg-Si alloy cables. Mater. Charact. 2016, 119, 114–119. [CrossRef] 9. EcKermannae, F.; Sutera, T.; Uggowitzerv, P.J.; Afseth, A.; Schmutz, P. The influence of MgSi particle reactivity and dissolution processes on corrosion in Al-Mg-Si alloys. Electrochim. Acta 2008, 54, 844–855. [CrossRef] 10. Chrominski, W.; Lewandowska, M. Precipitation phenomena in ultrafine grained Al-Mg-Si alloy with heterogeneous microstructure. Acta Mater. 2016, 703, 547–557. [CrossRef] 11. Nassef, A.; El-Garaihy, W.H.; El-Hadek, M. Mechanical and Corrosion Behavior of Al-Zn-Cr Family Alloys. Metals 2017, 7, 171. [CrossRef] 12. Hai, L.; Mao, Q.Z.; Wang, Z.X.; Miao, F.F.; Fang, B.J.; Song, R.G.; Zheng, Z.Q. Enhancing mechanical properties of Al-Mg-Si-Cu sheets by solution treatment substituting for recrystallization annealing before the final cold-rolling. Mater. Sci. Eng. A 2014, 620, 204–212. 13. Ding, L.P.; Jia, Z.H.; Zhang, Z.Q.; Sanders, R.E.; Liu, Q.; Yang, G. The natural aging and precipitation hardening behavior of Al-Mg-Si-Cu alloys with different Mg/Si ratios and Cu additions. Mater. Sci. Eng. A 2015, 627, 119–126. [CrossRef] 14. Cao, C.; Zhang, D.; Wang, X.; Ma, Q.B.; Zhuang, L.Z.; Zhang, J.S. Effects of Cu addition on the precipitation hardening response and intergranular corrosion of Al-5.2Mg-2.0Zn (wt %) alloy. Mater. Charact. 2016, 122, 177–182. [CrossRef] 15. Jiang, J.T.; Xiao, W.Q.; Yang, L.; Shao, W.Z.; Yuan, S.J.; Zhen, L. Ageing behavior and stress corrosion cracking resistance of a non-isothermally aged Al-Zn-Mg-Cu alloy. Mater. Sci. Eng. A 2014, 605, 167–175. [CrossRef] 16. Song, F.X.; Zhang, X.M.; Liu, S.D.; Tan, Q.; Li, D.F. The effect of quench rate and overaging temper on the corrosion behavior of AA7050. Corros. Sci. 2014, 78, 276–286. [CrossRef] 17. Zhang, X.X.; Zhou, X.R.; Hashimoto, T.; Liu, B. Localized corrosion in AA2024-T351 aluminum alloy: Transition from intergranular corrosion to crystallographic pitting. Mater. Charact. 2017, 130, 230–236. [CrossRef] 18. Bonfils-Lahovary, M.L.; Laffont, L.; Blanc, C. Characterization of intergranular corrosion defects in a 2024 T351 aluminum alloy. Corros. Sci. 2017, 119, 60–67. [CrossRef] 19. Bonzom, R.; Oltra, R. Droplet cell investigation of intergranular corrosion on AA2024. Electrochem. Commun. 2017, 81, 84–87. [CrossRef] 20. Tang, Y.; Zhang, L.J.; Du, Y. Diffusivities in liquid and fcc Al-Mg-Si alloys and their application to the simulation of solidification and dissolution processes. Calphad 2015, 49, 58–66. [CrossRef] 21. Ferguson, J.B.; Lopez, H.E.; Cho, K.; Kim, C.S.; Farlkoosh, A.R.; Pekguleryuz, M. Temperature Effects on the Tensile Properties of Precipitation-Hardened AI-Mg-Cu-Si Alloys. Metals 2016, 6, 43. [CrossRef] 22. Kaseem, M.; Min, J.H.; Ko, Y.G. Corrosion behavior of Al-1 wt % Mg-0.85 wt % Si alloy coated by

micro-arc-oxidation using TiO2 and Na2MoO4 additives: Role of current density. J. Alloys Compd. 2017, 723, 448–455. [CrossRef] 23. Svenningsen, G.; Larsen, M.H.; Nordlien, J.H. Nisancioglu, K. Effect of high temperature heat treatment on intergranular corrosion of AlMgSi(Cu) model alloy. Corros. Sci. 2006, 48, 258–272. [CrossRef] 24. Mol, J.M.C.; Langkruis, J.; Wit, J.H.W.; Zwaag, S. An integrated study on the effect of pre-and post-extrusion heat treatments and surface treatment on the ﬁliform corrosion properties of an aluminum extrusion alloy. Corros. Sci. 2005, 47, 2711–2730. [CrossRef]

25. Li, J.F.; Maier, B.; Frankel, G.S. Corrosion of an Al-Mg-Si alloy under MgCl2 solution droplets. Corros. Sci. 2011, 53, 2142–2151. [CrossRef] Metals 2017, 7, 387 12 of 12

26. Mizuno, K.; Nylund, A.; Olefjord, I. Surface reactions during pickling of an aluminum-magnesium-silicon alloy in phosphoric acid. Corros. Sci. 2001, 43, 381–396. [CrossRef] 27. Li, C.; Sun, J.Y.; Li, Z.D.; Gao, Z.M.; Liu, Y.C.; Yu, L.M.; Li, H.J. Microstructure and corrosion behavior of Al-10%Mg2Si cast alloy after heat treatment. Mater. Charact. 2016, 122, 142–147. [CrossRef] 28. Zeng, L.F.; Wei, Z.L.; Li, J.F.; Li, C.X.; Xing, T.; Zhao, Z.; Zheng, Z.Q. Corrosion mechanism associated with

Mg2Si and Si particles in Al-Mg-Si alloys. Trans. Nonferrous Met. Soc. China 2011, 21, 2559–2567. [CrossRef] 29. Li, C.X.; Li, J.F.; Birbilis, N.; Jia, Z.Q.; Zheng, Z.Q. Synergetic effect of Mg2Si and Si particles on intergranular corrosion of Al-Mg-Si alloys through multi-electrode coupling system. J. Chin. Soc. Corros. Prot. 2010, 30, 107–113. 30. Blanc, C.; Roques, Y.; Mankowski, G. Application of phase shifting interferometric microscopy to studies of the behavior of coarse intermetallic particles in 6056 aluminum alloys. Corros. Sci. 1998, 40, 1019–1035. [CrossRef]

31. Ahlatci, H. Production and corrosion behaviors of the Al-12Si-XMg alloys containing in situ Mg2Si particles. J. Alloys Compd. 2010, 503, 122–126. [CrossRef] 32. Bhattamishra, A.K.; Lal, K. Microstructural studies on the effect of Si and Cr on the intergranular corrosion in Al-Mg-Si alloys. Mater. Des. 1997, 27, 25–28. [CrossRef] 33. Liang, W.J.; Rometsch, P.A.; Cao, L.F.; Birbilis, N. General aspects related to the corrosion of 6xxx series aluminum alloys: Exploring the inﬂuence of Mg/Si ratio and Cu. Corros. Sci. 2013, 76, 119–128. [CrossRef] 34. Wang, Z.X.; Li, H.; Miao, F.F.; Sun, W.Z.; Fang, B.F.; Song, R.G.; Zheng, Z.Q. Improving the intergranular corrosion resistance of Al-Mg-Si-Cu alloys without strength loss by a two-step aging treatment. Mater. Sci. Eng. A 2014, 590, 267–273. [CrossRef] 35. Masuda, T.; Takaki, Y.; Sakurai, T.; Hirosawa, S. Combined effect of pre-straining and pre-aging on bake-hardening behavior of an A1-0.6 massy Mg-1.0 massy Si alloy. Mater. Trans. 2010, 51, 325–332. [CrossRef] 36. Yassar, R.S.; Cai, M.; Field, D.P.; Chen, X.; Asay, J. The effect of shock-loading aging behavior of an Al-Mg-Si alloy. J. Mater. Sci. 2006, 41, 1711–1720. [CrossRef] 37. Buchanan, K.; Colas, K.; Ribis, J.; Lopez, A.; Gamier, J. Analysis of the metastable precipitates in peak-hardness aged Al-Mg-Si(-Cu) alloys with differing Si contents. Acta Mater. 2017, 132, 209–221. [CrossRef] 38. Ding, X.P.; Cui, H.; Zhang, J.X.; Li, H.X.; Guo, M.X.; Lin, Z.; Zhuang, L.Z.; Zhang, J.S. The effect of Zn on the age hardening response in an Al-Mg-Si alloy. Mater. Charact. 2015, 65, 1229–1235. [CrossRef] 39. Chen, J.H.; Costan, E.; Van-huis, M.A.; Xu, Q.; Zandbergen, H.W. Atomic Pillar-Based Nanoprecipitates Strengthen AlMgSi Alloys. Science 2006, 312, 416–420. [CrossRef][PubMed] 40. Zou, Y.; Qing, L.; Jia, Z.H.; Xing, Y.; Ding, L.P.; Wang, X.L. The intergranular corrosion behavior of 6000-series alloys with different Mg/Si and Cu content. Appl. Surf. Sci. 2017, 405, 489–496. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).