Analysing the combined effect of crystallographic orientation and grain refinement on mechanical properties and corrosion behaviour of ECAPed ZE41 Mg alloy

S.Prithivirajan,S.Narendranath,Vijay Desai

Corrosion Engineering Lab,Department of Mechanical Engineering,National Institute of Technology Karnataka,Surathkal,Srinivasanagar,Mangalore 575025,Karnataka,India

Received 4 January 2020;received in revised form 4 August 2020;accepted 4 August 2020 Available online 6 October 2020

Abstract Two step equal channel angular pressing carried out on as cast ZE41 Mg alloy resulted in a remarkable grain refinement.As compared to grain size of 46μm in as cast sample,refinement upto 2.5μm was achieved after 8th pass equal channel angular pressing(ECAP).The combined effect of crystallographic orientation and grain refinement was investigated by analysing the mechanical properties and corrosion behaviour of ZE41 Mg alloy using electron back scattered diffraction(EBSD).The first stage comprises of 1st,2nd,3rd and 4th passes at a processing temperature of 300°C while the 5th,6th,7th and 8th passes were ECAPed at 275°C in second stage.The mechanical properties of ZE41 Mg 158 yield tensile strength(YTS),230 ultimate tensile strength(UTS)and 7% elongation in as cast condition is enhanced to 236 YTS,295 UTS and 19.76%,respectively,after first stage ECAP.The yield tensile strength deteriorated due to the effect of texture predominating grain refinement during the second stage ECAP.The corrosion resistance of ZE41 Mg was significantly enhanced by ECAP and is inferred from electrochemical impedance spectroscopy(EIS)and potentiodynamic polarisation results.The role of microstructure was minimal on corrosion behaviour of ZE41 Mg due to extra resistance when tested in 0M NaCl.However,the influence of grain refinement greatly influenced the improvement in corrosion resistance of ZE41 Mg rather than crystallographic orientation observed from EBSD.In contrast,the crystallographic orientation predominated the effect of grain refinement during ZE41 Mg corrosion in chloride containing(0.1M and 1M NaCl)solutions.From the observation of results it is found that equal channel angular pressing has the dual advantage of improving mechanical properties and corrosion resistance of ZE41 Mg alloy.

Keywords:Corrosion;EBSD;Equal channel angular pressing;Magnesium alloy.

1.Introduction

The research and development work in the automobile industry enlightens the importance of Magnesium and its alloys.Magnesium being the lightest of all structural materials remarkably reduces the fuel consumption and CO2emissions.The low density of Magnesium and its alloys also attracts the auto manufacturers to explore it as replacement for Aluminium components[1].ZE41 Mg alloy is used in aircraft gear box casing,generator housing of military helicopters,wheels of champion racing cars and vibration testing equipment[2].However,the limited ductility of Magnesium and its alloys as well as their relatively poor corrosion resistance still remains to an extent a hindrance to extend their applications in automobile industry.Recent research suggest that the mechanical properties of Magnesium and its alloys can be improved by severe plastic deformation.Equal channel angular pressing is one of the promising methods amongst them that remarkably improves yield strength,ultimate tensile strength and ductility[3].Rengen Ding et al have reported improved mechanical properties of ZE41 Mg alloy by obtaining a grain refinement of 2μm after 6 pass of ECAP at 320°C without studying the corrosion behaviour[4].The corrosion behaviour of Magnesium and its alloys is generally improved by surface coatings and different heat treatment techniques.Similarly,the enhancement of corrosion resistance of ZE41 Mg alloy is reported by few researchers by coatings,heat treatment and addition of inhibitors[5–7].Equal channel angular pressing enhanced the corrosion resistance of ZE41,ZM21,AE42 and AZ80 Mg alloys.Researchers have concluded that fine distribution of secondary phase particles resulted in such improvement[8–10].However,the adverse effects were also observed in pure Mg,AZ91D and AE21 due introduction of crystalline defects such as dislocation density and also the generation of high angle grain boundaries(HAGBs)after ECAP[11,12].The recent research suggest that crystallographic orientation has a greater influence on corrosion behaviour[13–15]and mechanical properties[34–37]of Mg alloys.From the literature it is observed that not much attention is given to analyse the combined effect of grain refinement,crystallographic orientation and the ECAP processing temperature on mechanical properties and corrosion behaviour of ZE41 Mg alloy.Hence,in the current research work an effort has been made to explore these issues.

Equal channel angular pressing was carried out at two stages.The first stage comprises of 1st,2nd,3rd and 4th passes at a processing temperature of 300°C while the 5th,6th,7th and 8th passes were ECAPed at 275°C.Also,corrosion behaviour of all ZE41 Mg samples were tested in 0M,0.1M and 1M NaCl solution to simulate conditions encountered in automobile applications[40].These process parameters were chosen in order to analyse the combined influence of grain refinement and crystallographic orientation on mechanical properties and corrosion behaviour of ZE41Mg alloy.

2.Experimental procedure

2.1.Material and methods

Commercially available ZE41 Magnesium alloy(4.2% Zn,1.2% RE,0.7% Zr,balance-Mg)with dimension of 16mm diameter and 200mm length was procured in as cast condition.These specimens were initially heat treated at 400°C with holding time of 14 h in order to dissolve the secondary phase particles.Severe Plastic deformation was induced in these samples by subjecting to equal channel angular pressing.ECAP die was made in house with two 16mm diameter channels having channel angle and curvature angle of 110°and 30° respectively.The temperature of ECAP die was continuously monitored by a temperature controller.After each pass the samples were rotated to 90° in counter clockwise direction following one of the conventionally available routes BC[10].In the first stage of ECAP,samples were pressed at a temperature of 300°C upto 4 passes while the second stage involved pressing of samples upto 8 pass at a temperature of 275°C.Molybdenum disulphide was used as a lubricant during ECAP.

2.2.Microstructure and micro-texture analysis

The microstructural characterisation was performed using FEI Nova Nano FEG SEM.Prior to EBSD characterisation both the cast and ECAPed samples were cut using Struers metallographic sample cutter along the extrusion direction.The samples were polished with #5000 grit emery papers and mirror like finish was obtained by using 0.25μm diamond paste.This procedure is followed by electro polishing the sample with etchant containing 15mL Perchloric acid,41.5g sodium thiocyanate anhydrous,75g citric acid,800mL Ethanol and 100mL Proponal.After electro polishing,low angle ion milling was also carried out on the samples.Pole figures,inverse pole figures and schmid factor charts were generated from TSL OIM software for all ZE41 Mg samples to investigate the influence of equal channel angular pressing on grain refinement and crystallographic orientation.For all sample conditions at least two EBSD maps were acquired.In addition,the steps mentioned in the reference[22]was followed.X-ray diffraction was carried out to analyse the phases present in ZE41 Mg before and after ECAP.

2.3.Mechanical characterisation

The tensile tests were conducted by using a horizontal table top electronic tensometer at a strain rate of 0.5mm/min.In,addition the linear increment was controlled by DC servo motor attached with tensometer.The round tensile specimens were cut along extruded direction and machined according to ASTM E8M standard having a gauge diameter 4mm and gauge length 20mm.Triplicate tests were carried out to ensure reproducibility.After completion of tensile test,the fractured surfaces were also analysed using scanning electron microscope.

2.4.Corrosion behaviour

ACM Gill AC electrochemical corrosion analyser was used to evaluate the corrosion behaviour of ZE41 Mg samples.After obtaining mirror like surface finish the samples were inserted into corrosion kit which allowed only 1cm2area of ZE41 Mg expose to testing medium.Three types of tests were carried out and the chronological order of these tests were sequenced.The thermodynamics of corrosion was given by open circuit potential.The OCP tests lasted for a duration of 20 min.The nature of surface film and its properties like charge transfer resistanceRct,double layer capacitanceCdland solution resistanceRswere obtained from EIS test.This test started at a frequency of 10,000Hz and ended at 0.001Hz with amplitude of 10mV.The Nyquist plot obtained from EIS tests were fitted using V4 analysis software.Cyclic sweep tests were conducted by sweeping through a potential of?250mV to+250mV with respect to rest potential.The corrosion rate of ZE41 Mg samples were calculated based on the cathodic slopes.The electrochemical corrosion tests were triplicated to gain confidence in measurements.All the experiments were carried out at a temperature of 25°C.

Fig.1.EBSD micrographs of(a)As Cast(b)2 Pass(c)4 Pass(d)6 Pass(e)8 Pass ZE41 Mg alloy.

3.Results and discussion

3.1.Microstructure of ZE41 Mg

The electron backscattered micrographs of as cast and 2nd Pass,4th Pass,6th Pass,8th Pass equal Channel angular pressed ZE41 Mg samples generated from TSL OIM software are presented in Fig.1(a–e)respectively.The EBSD micrographs of 1st Pass,3rd Pass,5th Pass and 7th Pass ZE41 Mg are also presented in supplementary Fig.S1 to enlighten the grain refinement phenomenon.The stereographic triangle in the lower left corner of Fig.1 represents the colour corresponding to the crystallographic orientation.As previously described,during the first step of ECAP the samples were pressed at a temperature of 300°C whilst the second step was carried out at 275°C.The grain morphology of as cast ZE41 Mg samples evinced heterogeneous grains with grain size varying from 35μm to 58μm as depicted in Fig.1(a).However,the average grain size of cast sample was calculated to be 46μm.After 2nd pass of equal channel angular pressing a bimodal grain morphology was observed with grain diameter of 25 and 12μm having area fraction of 0.109 and 0.058 respectively.Also,recrystallisation occurred only along the grain boundaries which is evident from Fig.1(b).The average grain diameter of as cast,1st Pass,2nd Pass,3rd Pass and 4th Pass ECAP samples was measured to be～46,～24,～13,～6 and～10μm respectively.Despite increasing the number of ECAP passes,grain coarsening was observed in 4th pass sample as shown in Fig.1(c).This grain growth during 4th pass is attributed to performing ECAP at same temperature of 300°C and inter pass time that the sample spent during ECAP.Also,the grain boundaries of fine grain microstructure is less stable than the coarse grain microstructure.Earlier researchers who also observed similar phenomena attributed this to fine grains possessing lower activation energy than the coarse grains.This leads to increase in grain size to attain stable state during dynamic equilibrium[16–19].In order to enlighten the mechanism of grain refinement the grain boundary maps are represented in Fig.2.They are generated from field emission scanning electron microscope(FESEM)micrographs.The Fig.2(a,c,e,g)and(b,d,f,i)represents the FESEM images and grain boundary mapping of as cast,1st pass,2nd pass and 8th pass respectively.The grain boundaries with misorientation angle less than 15°and greater than 15° are classified as low angle grain boundaries(LAGBs)and high angle grain boundaries(HAGBs)respectively.From,grain boundary mapping,the zone with dynamically recrystallised(DRXed)grains are observed.The volume fraction of LAGBs and HAGBs of all ZE41 Mg samples are represented in the Fig.3.From Figs.2 and 3,it is apparent that with increase in number of passes the area fraction of high angle grain boundaries(HAGBs)increases.Similar relative increase in HAGBs and relative decrease in LAGBs were observed when LAE442 Mg alloy was subject to ECAP[42].X-ray diffraction peaks of ZE41 Mg alloy is depicted in the Fig.4.It is observed that peaks corresponding to primary Mg Phase and T phase(Mg7Zn3RE)are clearly visible.Our previous study on ZE41 Mg also revealed the presence of Mg and secondary phase(T-phase)when observed under SEM[40].One can observe the increase in relative intensity of peaks with respect to increase in number of pass.This is correlated to the variation in texture and fraction of secondary phases or precipitates[42].ECAP is dynamic in nature where during extrusion process,heating of material coarsen the grain while simple shearing action at the intersection of equal channels contribute to grain refinement.The resulting microstructure after reaching dynamic equilibrium gives an insight into factors that dominated during each pass of ECAP.In this study,two-step ECAP process was chosen to further enhance the mechanical properties of ZE41 Mg alloy.During the second stage of ECAP process the average grain diameter was calculated to be～4.8,～7.14,～3.93 and～2.5μm after 5th Pass,6th pass,7th pass and 8th pass,respectively.There is no remarkable change in grain size during two step ECAP at 275°C which is clearly observed from Fig.1(d)and 1(e).This minimal reduction in grain size during second stage of ECAP was also observed by various researchers[16,17].The grain refinement obtained after ECAP is in good accordance with recent findings[4,40,41]and proposed grain refinement models[20].

Fig.2.FESEM images of(a)As cast(c)1st pass(e)2nd pass(g)8th pass and their corresponding grain boundary maps(b)As cast(d)1st pass(f)2nd pass(i)8th pass.

Fig.3.Grain boundary misorientation angles as function of number of ECAP pass.

Fig.4.X-ray diffraction patterns of ZE41 Mg alloy.

3.2.Micro-texture evolution

The pole figures corresponding to(0001)plane for as cast and 2nd Pass,4th Pass,6th Pass,8th Pass ECAPed ZE41 Mg samples generated from EBSD plots are illustrated in Fig.5(a–e)respectively.The pole figures of 1st pass,3rd pass,5th pass and 7th pass ECAPed samples are also presented in the supplementary Fig.S2.The as cast ZE41 Mg alloy exhibited a randomly distributed texture with a maximum texture intensity value of 7.3 as represented in the Fig.5(a).The intensity of texture increased to a maximum value of 10 after 2nd pass ECAP,this is due to formation of new grains during grain refinement which is apparent from Fig.1b.Similar kind of increase in texture intensity and formation of new grains was also observed and reported by Kim et al when AZ61 Mg alloy was subject to ECAP[21].The position of maximum texture also shifted to a new location away from extrusion direction(ED)and transverse direction(TD)after the completion of 2nd pass of ECAP as presented in Fig.5(b).Equal channel angular pressing resulted in variation of position,value of maximum texture intensity and formation of new texture elements which is attributed to rotation of ECAP billet by 90° after each pass following the conventional routeBc.From Fig.5(c)it is clear that the 4th pass ECAP resulted in weakening of basal texture element with a slight decrease in intensity to a value of 9.6.A decrease in maximum texture intensity 9.23 and 6.9 was also observed after 6th and 8th pass respectively which signifies the formation of new texture element.After completion of 6th pass and 8th pass the basal texture element aligned～65° and～55° to extrusion direction(ED)which is clearly indicated in Fig.5(d)and(e),respectively.This particular orientation of basal texture generally deteriorates yield strength of ZE41 Mg and is also the reason for insignificant reduction in grain size after 6th pass and 8th pass ECAP as observed in Fig.1(d)and(e).It is also well known that cast and extruded Mg samples have random and typical fibre texture respectively.Upon equal channel angular pressing texture softening occurs and eventually the samples ends up in formation of new texture element.The texture modification[21]and increase in texture intensity after first stage followed by sharp decrease during the second stage was also reported by various researchers during equal channel angular extrusion of AE21,AE42,LAE442,ZK60 and ZM21 Mg alloys[16,17,22].The as cast AE21,AE42 and LAE442 alloys obtained ultrafine grains within 8 pass because of random texture[16].In contrast,the extruded ZK60 and ZM21 Mg alloys relatively more number of passes in order to achieve ultrafine grains due to the typical fibre texture[17,22].In the present study,fine grains were obtained by 8th pass ECAP for as cast ZE41 Mg with random texture.Hence,it is reasonable to believe that the apart from processing temperature the initial texture or history of processing decides the number of ECAP passes to achieve grain refinement.In addition,the XRD peaks of ZE41 Mg alloy exhibits variation in intensity which indicates variation in texture as shown in Fig.4.The texture evolution based EBSD also indicates the variation in texture with respect to each ECAP condition.Thus,the XRD peaks are also in line with the texture plots generated from EBSD.The XRD and texture plots obtained in the present study is in accordance with recent findings[43].

Fig.5.Pole Figures corresponding to(0001)Plane(a)As Cast(b)2 Pass(c)4 Pass(d)6 Pass(e)8 Pass of ZE41 Mg alloy.

3.3.Influence of grain refinement and micro texture on mechanical properties of ZE41 Mg alloy

Fig.6.Mechanical properties of ZE41 Mg alloy as function of number of ECAP passes.

The mechanical properties of ZE41 Mg alloy viz.,yield strength(YS),ultimate tensile strength(UTS)and% elongation are depicted in Fig.6 as a function of ECAP passes.It is well established that various mechanisms such as solid solution,precipitate and grain boundary strengthening influence mechanical properties of ECAPed Mg samples.From XRD peaks represented in Fig.4,it is observed that there is no significant changes in the lattice parameters of ZE41 Mg alloy.This indicates that solid solution strengthening is not the reason for improvement in mechanical properties.Our previous study on ZE41 Mg indicated the precipitates are fragmented after ECAP processing.The size of the precipitate was in the range of 1.5 to 5μm[40].It is well known that the dispersion strengthening mechanism is possible only if the particle size is about 100nm.Also,if precipitation hardening is the reason for strengthening,the mechanical properties would have increased chronologically with increase in number of passes[4].From Fig.6,it is evident that the mechanical properties did not show increasing trend with increase in number of passes.Hence,it is reasonable to conclude that precipitation strengthening did not occur during ECAP.The XRD peaks(present study)and the range of precipitate size[40]after ECAP are in accordance with findings of Ding et al.[4].They also reported that solid solution and precipitate strengthening did not influence the mechanical properties of Mg alloy.In addition other studies on ECAP of AE21,AE42,LAE442,ZK60,AX41 and ZM21 Mg alloys revealed that mechanical properties are dependant on grain boundary and texture[16,17,21,22,41,42,43].Hence,in the present study the grain boundary strengthening and texture are taken into consideration.According to Hall–Petch relation,the yield strength of ZE41 Mg alloy should be improved due to grain refinement after ECAP.The yield strength of ZE41 Mg was 158MPa in as cast condition and gradually increased to 176,190,192 and 236MPa after 1st pass,2nd pass,3rd pass and 4th pass ECAP respectively.From Fig.6 it is evident that remarkable increase in yield tensile strength was observed with increasing ECAP passes during the first stage of ECAP.This increase in yield strength resulted from the grain refinement after ECAP observed from Fig.1(b)and(c).R Ding et al also investigated the mechanical properties of ZE41 Mg before and after performing 6 pass ECAP at 320°C.The yield strength of all ECAPed samples were greater than as received.However,the yield strength of 1st,2nd and 6th pass ECAP samples were similar and measured to be 230MPa[4].In single step process all ECAP passes is carried out same temperature.In this procedure grain growth is experienced by fine grains because they are not stable at higher temperature.To control the grain growth and further enhance the mechanical properties recent studies performed ECAPed at two stages[16,17].Hence,in the present study second stage of ECAP was carried at relatively lower temperature of 275°C.In contrast,the second stage of ECAP witnessed a sharp decrease in yield strength(YS)values of 204,197,181 and 172MPa after 5th pass 6th pass 7th pass and 8th pass,respectively.This deviation could be explained by the fact that other metallurgical factors like texture influence the yield strength of ZE41 Mg alloy[4].The formation of new texture element and their alignment with ECAP shear plane as represented in Fig.5(d)and(e)decreased the yield strength of ZE41 Mg despite the grain refinement depicted in Fig.1(d)and(e).In short,during the second stage of ECAP processing texture softening predominates the effect of grain refinement.Similar drop in yield strength of ECAPed Mg alloys at the second stage was also reported by various investigators[16,17,21,22].Also,the mechanical behaviour of Mg alloys at room temperature and relatively higher temperatures are typically affected by basal slip.Hence,the schmid factor values for basal slip(0001)11–20are generated from EBSD plots for all ZE41 Mg samples.The Schmid factor values for basal slip was computed to be 0.21,0.29,0.30,0.34 and 0.33 for as cast,2nd pass,4th pass,6th pass and 8th pass,respectively.The schmid factor values in the current study are also in good agreements with recent findings[16,17,22].During the first stage of ECAP,elongation was 7% in as cast condition which increased to～20% after 4th ECAP pass.However the maximum elongation of 23% was observed during second stage after 6th pass ECAP.The basal texture orientation of 6th pass sample as shown in Fig.5(d)along with the maximum schmid factor value of 0.34 favours dislocation glide on slip plane thereby improving the ductility.The fractured surface morphology of for as cast and 2nd Pass,4th Pass,6th Pass,8th Pass ECAPed ZE41 Mg samples are depicted in the Fig.7(a–e)respectively.Tear ridges and micro cracks were observed on the fractured surface of as cast ZE41 Mg.This signifies brittle mode of fracture and hence the as cast sample exhibited the lowest% elongation of 7% as depicted in Fig.7(a).The fractured surface morphology of 2nd pass ECAP sample exhibited cleavages despite increase in% elongation(refer Fig.7(b)).This might be due to the bimodal grain distribution of 2nd pass ECAP in which fine grains contribute to increase in% elongation upto 21%.In the 4th pass ECAP sample dimples were observed as shown in Fig.7(c).But the% elongation declined relatively to value of 20%.This is attributed to grain growth experienced while ECAP as represented in Fig.1(c).In 6th and 8th pass ECAPed samples dimples were observed but their size is relatively lower than 4th pass sample as depicted in Fig.7(d)and(e)respectively.In addition micro-cracks were observed in both 6th and 8th pass ECAP sample due to processing ECAP at a relatively lower temperature of 275°C.The attenuation in elongation upto 15% evinced after 8th pass.Probably,it is related to the incomplete dissolution of orthorhombic T-phase which is incompatible with HCP structure of magnesium[23,24].The trend of UTS followed a similar pattern as that of yield strength with respect to ECAP pass number which is also in accordance with recent research findings.In summary,the mechanical properties of ZE41 Mg alloy subject to ECAP has correlation to microstructure especially grain refinement,texture and processing temperature.R Ding et al also reported that global or bulk texture generated from XRD affects the mechanical properties of ZE41 Mg alloy[4].But,the current study involves micro texture analysis of ECAPed ZE41 Mg samples generated from EBSD results.Minarik et al proved that global texture of Magnesium alloys are in accordance with the micro texture[22].The present study also reveal that the mechanical behaviour of ZE41 Mg are affected by both micro texture and grain refinement.

3.4.Corrosion behaviour of ZE41 Mg

3.4.1.Open circuit potential

The OCP curves of ZE41 Mg studied in three different concentration of sodium chloride solution is represented in Fig.S3.The potential steadily increased for all ZE41 Mg samples in interaction with 0M NaCl solution indicating a stable surface film formation.The as cast sample evinced lowest potential value whilst the potential of ECAPed samples moved towards nobler direction and the overall potential value varied within the range of 1500±60mV.In contrast when tested in 0.1M NaCl solution,all the ZE41 Mg samples attained a steady state at 1550±10mV electrode potential.The OCP curves of ZE41 Mg exhibited drastically different behaviour in interaction with 1M NaCl solution.The 3rd pass and 4th pass sample showed initial increase and attained steady state at 1600mV while the other samples did not show change in potential as a function of time.It is interesting to observe that the 4th pass ECAPed ZE41 Mg exhibited the highest positive potential in chloride containing environments.Similar trends in OCP curves was observed in ZE41 Mg when alkyl carboxylates were used as inhibitors for corrosion[7]and when influence of pH and chloride ion concentration was varied[25].In contrast,the current study investigates the combined influence of grain refinement and crystallographic orientation on ZE41 Mg corrosion after ECAP.8th pass sample exhibited highest potential values in 0M NaCl indicating better corrosion resistance.In contrast,4th pass sample evinced higher potential values in comparison with other samples when tested in chloride containing environments.However,one cannot conclude the improvement in corrosion resistance based on OCP curves because it describes only the thermodynamics of corrosion.Hence,the kinetics of corrosion of ZE41 is evaluated by EIS and potentiodynamic polarisation studies.

3.4.2.Electrochemical impedance spectroscopy

Nyquist plot of as cast and 1st,2nd,3rd and 4th pass ZE41 Mg samples ECAPed at 300°C are depicted in Fig.8(a–c).The supplementary Fig.S4 presents the Nyquist plot for 5th,6th 7th and 8th pass ZE41 Mg samples ECAPed at 275°C.The characteristic features of Mg samples viz.,high and intermediate frequency capacitive loops as well as low frequency inductive loop(LFID)was observed in 0M,0.1M and 1M NaCl solutions.High and intermediate frequency capacitive loops are corroborated to charge transfer resistance(Rct)and ingress of testing medium into the corrosion product layer respectively.The occurrence of LFID is generally ascribed to adsorption of species like Mg+and Mg(OH)+[7,26,27].Fig.9(a–c)presents the parameters of simple Randles circuit derived from Nyquist plots with respect to number of ECAP passes.The fitting of Nyquist plot is carried out by using v4 analysis software as represented in supplementary Fig.S7.Charge transfer resistance is the most important parameter influencing the corrosion resistance of material.However,solution resistance(Rsol)and double layer capacitance(Cdl)also influence the corrosion process and are attributed to amount of corrosion products and thickness of Helmholtz double layer respectively.All the ECAPed samples exhibited increased resistance to charge transfer in comparison with their as cast counter parts in 0M,0.1M and 1M NaCl corrosive medium.From Fig.1(a–e)it is clear that equal channel angular pressing resulted in a remarkable grain refinement.But,the effect of grain refinement and crystallographic orientation was marginal on charge transfer resistance(Rct)in 0M NaCl as shown in Fig.9(a).This is related to less aggressive nature of 0M NaCl solution and its minimal influence on microstructure.Similar results were obtained when high purity Mg,ZE41,AZ91,Mg2Zn0.2Mn Mg alloys were tested in less aggressive corrosion medium[28].While the values of charge transfer resistance(Rct)exhibited substantial influence on microstructure(refer Fig.1)which is evident also from Fig.9(a).In 0M NaCl solution,the values of solution resistance(Rsol)varied apparently with respect to number of ECAP passes.In contrast,negligible variation of solution resistance(Rsol)values was observed when tested the impedance value is collectively influenced by summation of charge transfer resistance(Rct),solution resistance(Rs)and double layer capacitance(Cdl)[29].Higher value of impedance signifies good corrosion resistance of material with interacting environment.The Bode impedance plot of as cast and 1st,2nd,3rd and 4th pass ZE41 Mg samples ECAPed at 300°C is depicted in Fig.10(a–c).From the trend of Fig.10(a–c),it is apparent that at higher frequency(10kHz?100Hz)and moderate frequency(100Hz-0.1Hz)minimal and moderate changes were observed on impedance values respectively.In contrast,the variation of impedance values became significant at very low frequency(0.1Hz to 0.001Hz).The inflection points on impedance curves corresponds to the time constants observed in the Nyquist plots(refer Fig.8(a–c).The Bode plots obtained in the current study are in reasonable agreement with models proposed by Juttner[29]and recent findings[7,27].The Bode impedance plots of as cast and 5th,6th,7th,8th pass ZE41 Mg samples ECAPed at 275°C also show the same trends and is represented in the supplementary Fig.S5.It is evident from Fig.10(a)that the decrease in 0.1M and 1M NaCl solution.Also the solution resistance(Rsol)decreased with increasing the chloride ion concentration which is clear from Fig.9(b).This phenomenon is ascribed to the increasing corrosion products i.e.release of Mg2+ions into corrosive medium with increase in chloride ion concentration.The thickness of surface film which is inversely proportional to double layer capacitance(Cdl)also did not have significant influence on the ECAP pass number in 0M,0.1M and 1M NaCl solution as shown in Fig.9(c).In summary,the values of charge transfer resistance(Rct),solution resistance(Rsol)and double layer capacitance(Cdl)neither increased nor decreased steadily with respect to ECAP pass number as clearly depicted in Fig.9(a–c)respectively.The diminution in values of charge transfer resistance(Rct)and solution resistance(Rsol)with increase in NaCl concentration is related to destructive nature of chloride ions on surface film of ZE41 Mg alloy which is evident from Fig.9(a)and(b).Similar drop in charge transfer resistance(Rct)and solution resistance(Rsol)in the presence of chloride ions was reported by various researchers[25,27].However,in impedance was observed at very low frequency of 0.01Hz which is attributed to occurrence of pitting in 0M NaCl.The impedance values diminished because of gradual increase in chloride ions in 0.1M and 1M NaCl which is evident from Fig.10(b)and(c)respectively.The decrease in impedance values with increasing chloride ion concentration is accordance with the decrease in electrode potential observed in OCP curves(refer Fig.S3)as well as reduction in values of charge transfer resistance(Rct)and solution resistance(Rs)(refer Fig.9(a and b)).Similar kind of results were reported by King et al.when pure Magnesium was tested in NaCl solutions[27].

Fig.9.Parameters of simple Randle’s circuit as a function of number of ECAP Passes(a)Rct(b)Rs(c)Cdl.

3.4.3.Potentiodynamic polarisation

Fig.11(a–c)illustrates the Potentiodynamic polarisation plots of ZE41 Mg samples ECAPed at 300°C(1st 2nd 3rd and 4th)tested in 0M,0.1M and 1M NaCl respectively.The Potentiodynamic polarisation plots of samples processed at 275°C(5th 6th 7th and 8th)during the second stage ECAP evinced similar trends as first stage and are presented in the Fig.S6.From the Fig.11(a)the following observations were made(i)theEcorrvalues are in good agreement with the OCP curves(refer Fig.S3)(ii)theIcorrvalues calculated from extrapolation of cathodic slope evinced insignificant change with advancement of ECAP passes.These results are in good accordance with Nyquist plots obtained from EIS presented in Fig.8(a).From Fig.11(b)and(c),it is apparent that theIcorrvalues of all ECAPed samples are lesser than the as cast counterpart.It can also be observed that the 3rd and 4th pass samples exhibited significant shift ofEcorrtowards noble direction and less values ofIcorrin both 0.1M and 1M NaCl solution.This is attributed to higher double layer film thickness of 3rd and 4th pass ECAPed ZE41 Mg samples as represented in the Fig.9(c).In general,the Tafel plots shifted towards direction of increasing current density and lower potential values with increase in chloride ion concentration due to its destructive nature.This behaviour is in good agreement with charge transfer resistance(Rct)and solution resistance(Rsol)values evident from Fig.9(a and b).The nature of Potentiodynamic polarisation plots obtained in the present study is in complete agreement with research findings of King et al.who reported the corrosion behaviour of high purity Mg in 0.1M,1M and 5M NaCl solution[27].In contrast,the current study proposes equal channel angular pressing as a method to enhance corrosion resistance and enlightens the factors dominating corrosion of ZE41 Mg alloy.

Fig.10.Bode Impedance plots of ECAP samples processed at 300°C.

3.5.Combined effect of grain refinement and crystallographic orientation on ZE41 Mg corrosion

The impedance values and corrosion rate of all ZE41 Mg samples were obtained from Bode impedance plots and Tafel plots respectively.The corrosion rate of all ZE41 Mg samples is calculated from the Eq.(1)[25].

The values of impedance and corrosion rate are represented in the Fig.12(a)and(b)respectively with respect to ECAP pass number.The corrosion rate of ZE41 Mg samples tested in 0.1M NaCl are slightly higher than that of 0M NaCl.Surprisingly,the corrosion rate of 3rd and 4th pass ZE41 Mg tested in 0M and 0.1M NaCl are almost comparable as observed from Fig.11(b).This is related to the better surface film formation of ECAPed ZE41 Mg in 0.1M NaCl than 0M NaCl as shown in Fig.9(c).The double layer capacitance(Cdl)is directly proportional to local dielectric constant(ε)and inversely proportional to thickness of double layer film(d)as given by Helmholtz relation and represented in the Eq.(2).The lower the value of double layer capacitance(Cdl)higher the robustness of double layer film(d)[7].

Fig.11.Potentiodynamic polarisation plots of ECAP samples processed at 300°C.

Fig.12.(a)Impedance(b)Corrosion rate as a function of number of ECAP Passes.

Electrochemical impedance spectroscopy is non-destructive in nature and corrosion resistance is measured from impedance value of surface film.In contrast,Potentiodynamic polarisation calculates the corrosion rate by destruction of sample surface.However,similar trends in corrosion resistance was observed for ZE41 Mg which is evident from the Fig.12(a)and(b).The role of pH and chloride ion concentration(0M,0.1M and 1M NaCl)of ZE41 Mg was reported earlier.The corrosion resistance obtained in the present study is relatively higher than prior study.However,the present study cannot be compared to prior work because of the following difference in parameters(i)they added HCl and NaOH to adjust the pH(ii)the grain size is constant[25].Recent reports on corrosion of high purity Mg also conclude that corrosion resistance measured from both EIS and potentiodynamic polarisation techniques exhibited same trends in results[27].All ECAPed samples exhibited better corrosion resistance than as cast sample.However,on careful observation one could note from Fig.12(a)and(b)that the corrosion resistance of ZE41 Mg alloys neither increased nor decreased with chronological increase in ECAP pass number.If grain refinement is the only reason for increase in corrosion resistance,the corrosion resistance would have increased with respect to increase in number of ECAP passes in linear manner due to fine distribution of secondary phase particles.Generally,ECAP converts the regime of corrosion from localised to uniform mode by reducing the size and number of secondary phase particles[8–10,38].But in the present study such trend was not observed.Hence,other factors such as crystallographic orientation also influenced the corrosion behaviour of ZE41 Mg alloy.The corrosion behaviour of Mg alloys are influenced by microstructure i.e.grain refinement[10,12,39]and crystallographic orientation[13,30,31].Previous studies on ECAP of magnesium alloys reported that fine distribution of cathodic particle(secondary phases)resulted in enhanced corrosion resistance[8–10,38],while the adverse effects were attributed to crystalline defects[32]without considering the role of crystallographic orientation.In this study,Inverse pole figures(contours)were generated from EBSD micrographs in order to enlighten the effect of crystallographic orientation on corrosion.The Inverse pole figures of all ZE41 Mg samples are illustrated in the Fig.13(a–e)with a sample plot denoting(0001),(10–10)and(2–1–10)plane orientation.It is worth mentioning that the basal plane in HCP crystals is the most closely packed plane than the prism planes.The higher planar atomic density of basal planes provides better corrosion resistance than the prism planes and their values are mentioned in Table 1[31,33].When the samples are tested in less aggressive solutions like 0M NaCl there is a minimal influence on microstructure i.e.no significant change was observed from impedance value presented in Fig.12(a)and corrosion rate Fig.12(b)with respect the microstructural evolution depicted in Fig.1(a–e).However,the as cast sample and 8th pass sample exhibited major of grain orientation towards(0001)and(2–1–10)plane respectively which is evident from Fig.13(a)and(i)respectively.Also,the 8th pass sample with a grain size of 2.5μm exhibited the highest corrosion resistance in 0M NaCl solution.The corrosion morphology of as received,4th pass and 8th pass sample in three different testing conditions are represented in Fig.14.(a–c),(d–f)and(g–i)respectively.When tested in 0M NaCl,the grain boundaries of as received sample were corroded due to micro galvanic corrosion between bulk Mg and T-Phase.In contrast the secondary phase particles are finely distributed after ECAP.Hence,4th pass sample tested in 0M condition the corrosion is less serve when compared to as received one.However,in the 8th pass samples grain boundaries were not corroded.The corrosion morphology represented in Fig.14 is in accordance with corrosion rate depicted in Fig.12(b).The observation clearly indicates that the corrosion resistance of ZE41 Mg in 0M NaCl solution is majorly influenced by grain refinement than crystallographic orientation.Majority of the grains were oriented in between(0001)/(2–1–10)in 4th pass ECAP sample as observed in Fig.13(e).The 4th pass ECAPed sample evinced highest corrosion resistance when tested in 0.1M and 1M NaCl solution.From Fig.13 it is observed that all the samples viz.,Fig.13(a–d,f–i)exhibited maximum intensity at two crystallographic orientations in exception to 4th pass(refer Fig.13(e)).Interestingly,it is reported that micro-galvanic cells are formed between planes with different orientations deteriorate the corrosion resistance[14].In summary,the highest corrosion resistance was exhibited by the ECAPed samples at 4th pass on interaction with 0.1M NaCl and 1M NaCl rather than ultra-fine grain 8th pass sample.In 0.1M condition,the corrosion morphology of as received and 8th pass sample evinced pits and relatively severe corrosion when compared to the 4th pass sample.In contrast,when tested in 1M NaCl condition the surface appears to be severely corroded than 0.1M condition.Summarised,the corrosion morphology depicted in Fig.14 is in accordance with the corrosion rate presented in Fig.12(b).Thus,crystallographic orientation has predominated the effect of grain refinement during ZE41 Mg corrosion when tested in 0.1M and 1M NaCl which is clearly depicted in Fig.1(a–e)and 13(a–i)and OCP curves Fig.S1.In summary,the corrosion of ZE41 Magnesium alloy is influenced by microstructure i.e.combined effect of both grain refinement and crystallographic orientation.The anodic partial reaction at bulk magnesium occurs by the dissolution of magnesium and release of an electron as mentioned in Eq.(4).This electron is taken up at Mg7Zn3RE cathodes resulting in hydrogen evolution as shown in Eq.(3).In addition to electrochemical mode of corrosion,dissolution of magnesium occurs by chemical reaction which is referred as anodic hydrogen evolution mentioned in Eq.(5).

Fig.13.Inverse pole figures(contour)of ZE41 Mg alloys for different conditions(a)As Cast(b)1st pass(c)2nd pass(d)3rd pass(e)4th pass(f)5th pass(g)6th pass(h)7th pass(i)8th pass.

Table 1Atomic densities of planes in different crystallographic orientation.

Fig.14.SEM images of[(a)0M,(b)0.1M,(c)1M of as received][(d)0M,(e)0.1M,(f)1M of 4th pass][(g)0M,(h)0.1M(i)1M of 8th pass]after corrosion.

The overall corrosion reaction and product formation represented in the Eqs.(6)and(7)respectively.Based on the above mentioned reaction,grain refinement,distribution of secondary phase particles and crystallographic orientation possible corrosion mechanism is proposed.In chloride containing environments(0.1M and 1M NaCl),the corrosion initially occurs by electrochemical reaction i.e.micro galvanic corrosion between bulk Mg and T Phase.The process is then accompanied by occurrence of corrosion due to chemical reaction from chloride ions.This combination of corrosion due electrochemical and chemical leads to undermining of secondary phase particles as observed from corrosion morphology(refer Fig.14).The deviation from usual Tafel behaviour of magnesium alloys also signifies the occurrence of anodic hydrogen evolution due to chemical corrosion(5)as depicted in Fig.11.In addition to corrosion reaction,the sample having fine grains and distribution of secondary phase particles with preferred orientation i.e.4th pass sample exhibited the highest corrosion resistance.

4.Conclusion

The mechanical properties and corrosion behaviour of ZE41 Mg alloy before and after two step ECAP processing were studied using EBSD technique.First stage of ECAP involves the 1st 2nd 3rd and 4th ZE41 Mg pressed at 300°C while the second stage involved additional 4 passes(5th 6th 7th and 8th)ECAPed at 275°C.The corrosion behaviour of all ZE41 Mg was tested in 0M,0.1M and 1M NaCl solution to simulate conditions encountered in automobile applications.The following conclusion were drawn from this study.

(i)The first stage of ECAP enhanced the yield strength of ZE41 Mg alloy from 158MPa in as cast condition to 236MPa after 4th pass due to grain refinement.However,during the second stage of ECAP yield strength decreased to 172MPa after 8th pass attributed to texture softening effect.

(ii)The as cast sample exhibited a percentage elongation of 7.A remarkable improvement in percentage of elongation was observed during 6th pass ECAP related to the highest schmid factor value of 0.34 for basal slip.

(iii)The corrosion rate of 0.0036mm/yr in as cast condition decreased to a value of 0.027mm/yr at 8th pass ECAP when tested in 0M NaCl.This marginal improvement in corrosion resistance signified the minimal influence of ZE41 Mg on microstructure which is also evident from EIS and Potentiodynamic polarisation results.

(iv)4th pass ECAP sample with 10μm average grain diameter exhibited the corrosion rate of 0.034 and 0.708mm/yr when tested in 0.1M and 1M NaCl respectively.These values of corrosion rate were found to be relatively lower than of 8th pass ECAP sample with grain size of 2.5μm i.e.0.548 and 0.735mm/yr.Thus,the effect of grain refinement on corrosion of ZE41Mg is predominated by crystallographic orientation.

(v)ECAP was successful in improving both ductility and corrosion resistance of ZE41 Mg.

Declaration of Competing Interest

There is no conflict of interest.

Acknowledgements

The authors acknowledge support from Defence Research and Development Organisation–Naval Research Board,Government of India.Project Grant number NRB/4003/PG/366.

Supplementary materials

Supplementary material associated with this article can be found,in the online version,at doi:10.1016/j.jma.2020.08.015.