Andrej Atrens,Zhiming Shi,Syed U.Mehreen,Sen Johnston,Gung-Ling Song,Xinhu Chen,Fusheng Pn

aThe University of Queensland,School of Mechanical and Mining Engineering,Centre for Advanced Materials Processing and Manufacturing(AMPAM),St Lucia,Qld 4072,Australia

b Xiamen University,College of Materials,Center for Marine Materials Corrosion & Protection,State Key Laboratory of Physical Chemistry of Solid Surfaces,422 Siming Rd.,Xiamen 361005,China

cChongqing University,College of Materials Science and Engineering,Chongqing,400045,China

d Chongqing University,National Engineering Research Center for Magnesium Alloys,Chongqing 400045,China

Received 11 April 2020;received in revised form 3 August 2020;accepted 7 August 2020 Available online 28 September 2020

Abstract A review of the literature confirmed that the intrinsic corrosion rate of high-purity Mg as measured by weight-loss is 0.3mm/y in a concentrated chloride solution.Atmospheric corrosion of Mg alloys has produced corrosion rates of Mg-Al alloys an order of magnitude lower than the intrinsic corrosion rate of Mg in a concentrated chloride solution of 0.3mm/y.The only successful strategy to produce a Mg alloy with a corrosion rate in a concentrated chloride solution substantially less than the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y has been to improve the protectiveness of the corrosion product film.

Keywords:A,magnesium;B,weight loss;Polarization;EIS.

1.Introduction

Much research has been expended to understand Mg corrosion[1–6]and to produce Mg alloys with low corrosion rates,particularly in aggressive concentrated chloride solutions.Mg alloys are sought with corrosion rates as measured by weigh loss lower than the intrinsic corrosion rate of highpurity Mg of 0.3mm/y in a concentrated chloride solution.This benchmark identifies the intrinsic corrosion rate as measured by weight loss of Mg in concentrated chloride solutions as the lowest reliable corrosion rate for Mg in these solutions[7–10].It is of importance to understand the Mg alloy development strategies that have produced Mg alloys with low corrosion rates.

The aims of this review are:

(i)to review Mg corrosion rates,

(ii)to confirm that the intrinsic corrosion rate of high-purity Mg as measured by weight loss is 0.3mm/y in a concentrated chloride solution,

(iii)to understand the Mg alloy development strategies that have produced Mg alloys with low corrosion rates in such solutions,and

(iv)to examine the recent claims of a new highly-corrosionresistant Mg alloy[11,12].

To explore these aims,the recent literature of Mg corrosion was reviewed,with an emphasis on corrosion rates measured using weight loss,PW(mm/y).Weight loss corrosion rate measurements were preferred,because there are issues with the other common methods for measuring the corrosion rate of Mg alloys[1,3,13].The Mg corrosion rate is also evaluated(i)from the evolved hydrogen,PH[14],(ii)by Tafel extrapolation of polarisation curves to givePi,and(iii)from electrochemical impedance spectroscopy to give,Pi/EIS,[15].

Fig.1.Values of corrosion rate from Table 1 as measured by electrochemical techniques by Tafel extrapolation of polarisation curves,Pi or from electrochemical impedance spectra,Pi,EIS,plotted against the corrosion rate measured from the evolved hydrogen,PH or measured by weight loss,PW.Note that there is usually good agreement between PH and PW.The line drawn on the figure is a guide to the eye and shows the condition for equality of the two measurements of the corrosion rates.

The corrosion rate of Mg measured using hydrogen evolution in a chloride solution,PH,is typically in good agreement with the corrosion rate measured by weight loss,provided the corrosion rate is substantial[14,16,17].This provides confidence that both techniques provide a reliable measurement of the Mg alloy steady-state corrosion rate.In contrast,experimental evidence[2,13]indicates that electrochemical measurements of the corrosion rates of Mg alloys have not provided a good measurement of the steady-state corrosion rate of Mg alloys.This is confirmed by the data in Table 1 and by Fig.1.

The results of the review of Mg corrosion rates are presented in Table 1 and plotted in Figs.1 and 2.In particular,Table 1 and Fig.2 provide corrosion rates for Mg alloys in chloride solutions with an emphasis on the recent literature.

2.Electrochemical corrosion rate measurements

Fig.1 shows the values of corrosion rate from Table 1 as measured by electrochemical techniques by Tafel extrapolation of polarisation curves,Pi,or from electrochemical impedance spectra,Pi,EIS,plotted against the corrosion rate measured from the evolved hydrogen,PHor measured by weight loss,PW.Note that there is usually good agreement betweenPHandPW.The plotted data do not fall on the line drawn.Thus,PiandPi,EISconsistently underestimated the steady-state corrosion rates for Mg alloys in chloride solutions as measured fromPHandPW.Fig.1 provides reinforcement that electrochemical methods have not been reliable for the measurement of the steady-state corrosion rate of Mg alloys in concentrated chloride solutions[4,5,13].

Fig.2.Plot of the corrosion rate,PW or PH,against total alloying content for the Mg alloys in Table 1,with a horizontal line drawn at the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y as shown by high-purity(HP)Mg and a second horizontal line drawn at a corrosion rate of 1mm/y.These lines are drawn as a guide to the eye.

Reasons[1,3,4]why electrochemical measurements give corrosion rates for Mg alloys less than the steady-state corrosion rates include:

(1)The corrosion rate soon after specimen immersion can be orders of magnitude smaller than the steady state corrosion rate.Electrochemical measurements are typically carried out soon after specimen immersion in the solution,before there is steady state corrosion behaviour[93,98].

(2)The evolving hydrogen(from the cathodic partial reaction)can isolate part of the specimen,so that this self-corrosion cannot be detected by any electrochemical measurement[17].

3.Mg corrosion rates

The Mg corrosion rates from Table 1 are presented in Fig.2.Fig.2 plots the corrosion rate,PWandPH,against total alloying content(TA).A horizontal line is at the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y.A second horizontal line is at a corrosion rate of 1mm/y.These lines are a guide to the eye.

Fig.2 provides confirmation that the intrinsic corrosion rate as measured by weight loss for Mg is～0.3mm/y in concentrated chloride solutions as the lowest reliable corrosion rate in these solutions for high purity Mg.Research that has measured corrosion rates between 0.2 and 0.4 for HP and UHP Mg includes the following:Hanawalt,Nelson,Peloubet[7]measuredPW=0.3mm/y for alternate immersion in 3 wt% NaCl solution for 16 weeks(the specimen was dipped in the solution for 30s,followed by 2min in air during which the specimens did not completely dry);Cao et al.[8]measuredPW=0.25mm/y for immersion in 3.5 wt% NaCl saturated with Mg(OH)2for 14 days;Yang et al.[9]measuredPW=0.22 and 0.33mm/y for immersion in 3.5 wt% NaCl solution for 2 days;Liu et al.[10]measuredPW=0.2mm/y for immersion in Hanks’solution for 3 days and 15 days(0.8 wt% NaCl,0.14M NaCl);Gao et al.[28]measuredPW=0.27mm/y for immersion in Hanks’solution for 10 days;Johnston[36,50]measuredPW=0.4mm/y for immersion in Hanks’solution for 1 week;Jia et al.[38]measuredPW=0.4mm/y for immersion in Dulbecco’s modified eagle medium(DMEM)+10vol.% foetal bovine serum,100 units/mL penicillin+100 units/mL Streptomycin,at 37 C with 5 vol% CO2atmosphere for 56 days(DMEM has a chloride concentration similar to that of Hanks’solution);and Shi and Atrens[96]measuredPW=0.4mm/y for immersion in 3.5 wt% NaCl saturated with Mg(OH)2for 3 to 20 days.Taltavull et al.[77]showed that the high-purity Mg corrosion rate is nearly independent of chloride concentration,provided that there is no micro-galvanic corrosion.

(continued on next page)

(continued on next page)

Table 1(continued)

(continued on next page)

In this context it is worth repeating that the high-purity Mg that shows the intrinsic corrosion rate in chloride solutions as measured by weight loss of 0.3mm/y contains no second phase particles and is typically in ingot form.Heat treatment of ingot Mg can cause precipitation of Fe-rich particles and much higher corrosion rates[98].Heat treatments are associated with processing steps such as extrusion to form rod material.For such heat-treated material,the Fe tolerance limit may be less than～2ppm[98].

Table 1(continued)

(continued on next page)

Fig.2 presents data(as green diamonds)for the corrosion rate for atmospheric corrosion or simulated atmospheric corrosion of HP Mg and Mg-Al alloys.There was a significant decrease in the atmospheric corrosion rate with increasing Al content,attributed to a more protective surface film,with increasing Al content.This indicates that stable protective films form during atmospheric corrosion,whereas the films that form on similar Mg alloys during solution exposure are not protective.A key factor seems to be the dry-ing which occurs periodically during atmospheric exposure[105,106].

Fig.2 shows that Mg alloys typically have corrosion rates in chloride solutions greater than the intrinsic corrosion rate of Mg of 0.3mm/y,and most corrosion rates are above 1mm/y,as shown by the full black squares in 2.Mg is the most active engineering metal with a high tendency to corrode,and little protection is provided by corrosion product films.Mg alloys typically have phases in addition to the alpha-Mg matrix.These second phases typically accelerate corrosion by microgalvanic interaction with the matrix,so that the typical corrosion rates in chloride solutions are greater than 1mm/y[1–5].With the two exceptions presented below,despite claims to the contrary,there is no Mg alloy with a corrosion rate(measured by weight loss)in a concentrated chloride solution much lower than the intrinsic Mg corrosion rate of 0.3mm/y in a concentrated chloride solution.This is evidenced by the data presented in Table 1 and Fig.2.

The two exceptions of Mg alloys with a corrosion rate(measured by weight loss)in a chloride solution much lower than the intrinsic Mg corrosion rate of 0.3mm/y are the Mg alloys with weight loss corrosion rates of 0.1mm/y:(i)Mg-1.5Sr[30],and(ii)Mg-4Y-3Nd[90].Argade,Panigrahi and Mishra[90]reported a corrosion rate of 0.1mm/y for a Mg-4Y-3Nd alloy with a grain size of 0.0017mm,see Table 1 and Fig.1.This corrosion rate below the intrinsic Mg corrosion rate seems to have been produced by a more protective surface film.In addition,this research provided clear evidence of a significant decrease of corrosion rate caused by a decrease in grain size,which could be related to the increased impurity tolerance after grain refinement[107].Similarly,the Mg-1.5Sr of Dong et al.[30]had a corrosion rate below the intrinsic Mg corrosion rate.Thus,the only successful Mg alloy development strategy relates to Mg alloys that have produced more protective surface films on immersion in solution.

This Mg alloy development strategy was suggested by Song and Atrens[4]in 2003;a stainless Mg alloy might be possible if a Mg alloy could be produced that spontaneously produces a much more protective corrosion product film.This suggestion was based on the analogy with stainless steels which can be considered as Fe-Cr alloys,and which typically spontaneously form a passive surface oxide(approximately Cr2O3)for a Cr content greater than 10.5wt.%.

Fig.2 does also indicate that there are a significant number of alloys[9,16,17,20,39,41,49,54,82,90,94,96,104],including commercial alloys or modified commercial alloys[22,37,59,63,79,85,89,92,97,99,101–103],with corrosion rates in chloride solutions comparable to the intrinsic corrosion rate of Mg of 0.3mm/y in a concentrated chloride solution[7–9].These Mg alloys are plotted in Fig.2 as red circles for corrosion in chloride solutions,and as full blue circles for corrosion in various synthetic body fluids,(SBFs).

Thus,it is demonstrably not accurate or reasonable to claim that an alloy with a corrosion rate as measured by weight loss above 0.3mm/y in a concentrated chloride solution has a corrosion rate much lower than any existing Mg alloy[11].The lowest corrosion rate of the Mg-Li alloy described by Xu et al.[11]wasPW=0.8mm/y(see Table 1)and was substantially above the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y.This Mg alloy was claimed to have a corrosion rate much lower than any existing Mg alloy[11,12].For example:“Here we design an ultralow density(1.4g cm?3)Mg–Li-based alloy that is strong,ductile,and more corrosion resistant than Mg-based alloys reported so far”[11];“Ferry and colleagues[11]describe the development of a new alloy with a combined improvement in strength,ductility,and corrosion resistance,compared with other Mg alloys”[12],and“Emergence of“stainless”Mg alloy”as a landmark in the scientific and technical development of magnesium corrosion research[6].These claims were not supported by the experimental evidence of Table 1 and Fig.2.

It was also suggested that low corrosion rates in chloride solutions could be produced by decreasing the cathodic partial reaction of hydrogen evolution[41,42,108,109].In each casePW≥0.8mm/y.This corrosion rate was substantially above the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y,and so this approach has not produced a Mg alloy with a corrosion rate substantially less than the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y.

4.Conclusions

1.A review of the literature confirmed that the intrinsic corrosion rate as measured by weight loss of high-purity Mg is 0.3mm/y in a concentrated chloride solution.

2.Atmospheric corrosion of Mg alloys has produced corrosion rates of Mg-Al alloys an order of magnitude lower than the intrinsic corrosion rate of Mg in a concentrated chloride solution of 0.3mm/y,

3.Two Mg alloys,(namely Mg-1.5Sr,and Mg-4Y-3Nd)were identified with corrosion rates as measured by weight loss less than the intrinsic corrosion rate of high-purity Mg of 0.3mm/y in a concentrated chloride solution,

4.The only successful strategy to produce a Mg alloy with a corrosion rate as measured by weight loss substantially less than the intrinsic corrosion rate as measured by weight loss of Mg of 0.3mm/y has been to improve the protectiveness of the corrosion product film,

5.Corrosion rates for Mg alloys measured by electrochemical methods are typically lower than the steady-state corrosion rates measured by weight loss,often by orders of magnitude,

6.The recent claims that new Mg alloys have been produced that are more corrosion resistant than Mg-based alloys reported so far are not supported by the literature.

Acknowledgements

This work was supported by the Australian Research Council Discovery Project DP170102557.

Data availability

The data required to reproduce these findings are from the published literature and are available within the paper