Modification of metallic corrosion by ion implantation

Modification of metallic corrosion by ion implantation

Nuclear Instruments and Methods 182/183 (1981) 865-873 North-Holland Publishing Company 865 MODIFICATION OF METALLIC CORROSION BY ION IMPLANTATION C...

741KB Sizes 0 Downloads 5 Views

Nuclear Instruments and Methods 182/183 (1981) 865-873 North-Holland Publishing Company

865

MODIFICATION OF METALLIC CORROSION BY ION IMPLANTATION C.R. CLAYTON Department o f Materials Science and Engineering, State University of New York at Stony Brook, Stony Brook, New York 11 794, USA

This review will consider some of the properties of surface alloys, formed by ion implantation, which are effective in modifying corrosion behavior. Examples will be given of the modification of the corrosion behavior of pure metals, steels and other engineering alloys, resulting from implantation with metals and metalloids. Emphasis will be given to the modification of anodic processes produced by ion implantation since a review will be given elsewhere in the proceedings concerning the modification of cathodic processes.

1. Introduction Ion implantation has recently found application to the fields of corrosion science and corrosion protection. It has been found that by doping metallic surfaces with suitable dements the rate of the anodic and/or cathodic reactions may be reduced in order to lower corrosion rates significantly. The technique is fast developing as a research tool in the study of the corrosion mechanisms of conventional alloys. As a corrosion protection technique it has the advantage that unlike coatings no discrete interface is produced and no dimensional changes are made to the structures treated. In addition to this, unconventional alloys may be formed from elements which are normally immiscible and, therefore, the technique has opened up a new spectrum of corrosion resistant alloys. 2. Electrochemical evaluation of corrosion processes Firstly, it is important to review briefly the experimental technique which is commonly used to evaluate the effects of ion implantation upon the corrosion processes. The aqueous corrosion of metallic surfaces takes place by an electrochemical mechanism involving two simultaneous and complementary reactions, namely the anodic reaction which is oxidizing in nature, and which provides electrons for the cathodic, reduction reaction. Typically the anodic reaction is of the form: M ~ M z+ + Z e - .

The cathodic reactions will depend on the solution 00294 54x/81/0000-0000/$02.50 © North-Holland

composition as follows: 2H÷ + 2 e- ~ H2 hydrogen evolution in deaerated acidic solutions; O2+4H ++4e~2H20 oxygen reduction in acidic solutions; 02 + 2 H20 + 4 e- ~ 4 OH- oxygen reduction in basic and neutral solutions. The cathodic and anodic sites may be very close together and dynamically relocating during the corrosion process, for example in a pure, annealed single metal crystal where terraces, ledges and kinks provide sites of differing atomic bond strength and, therefore, different levels of activation energy for the anodic dissolution process. Alternatively, in polycrystalline materials, anodic and cathodic sites may be segregated to grain boundaries and/or grains of different activity according to composition, strain energy and orientation. It follows that the greater the oxidizing power of a solution, the greater will be the extent of the oxidizing process of anodic dissolution, provided that it is supported by an efficient cathodic process. Since both half-reactions are needed for corrosion to take place, corrosion can be stifled by either rate cohtrolling process. Hence, if the cathodic process is very slow, then so will be the anodic process, and the system will be under cathodic control. It is, therefore, apparent that to study the mechanism of the corrosion process, or to alter the mechanism of the corrosion process by an alloying treatment, it is essential to determine the kinetics of the cathodic and anodic reactions for the system in the range of aqueous environmental conditions (ex. pH, [0], T, etc.) of interest. VII. ION-IMPLANTEDMETALS

866

C.R. Clayton /Modification o f metallic corrosion

~o

POTENTIOSTAT °-- l WORKING AUXILIARY REFERENCE

IU,

ILIEFERI POTENTIOMETER

SALT-BRIDGEPROBE

Fig. 1. Block diagram of apparatus and circuitry associated with controlled potential measurements, ref. [1].

The most common technique to determine separately the anodic and cathodic kinetics of a metallic system is the polarization technique. Employing the circuit shown in fig. 1 and a polarization test cell shown in fig. 2 the test sample (working electrode) is connected via a milliameter to a potential controlling device known as a potentiostat. The other half of the cell is an auxilliary electrode which is normally

pure Pt. The sample thus becomes a half-cell, countering the Pt electrode. When the system is at open circuit the working electrode contains anodic and cathodic sites. By applying a potential across the two halfcells the sample can become polarized to a more electropositive value, whereupon the cathodic sites gradually become neutralized and the anodic process fmally dominates. Similarly the cathodic process can be studied by polarizing in the reverse direction. The current passing through the cell is, therefore, a measure of the rate of the corrosion process for a sample of a given surface area being observed at the impressed potential. In order to plot the change in current (ultimately recorded as a current density) with potentiN, the change in sample potential is measured relative to a reference electrode, which is essentially a non-polarizable electrode, calibrated by convention to a standard hydrogen electrode [1]. Fig. 3 shows a schematic of typical plots for the polarization of a stainless steel in a mildly oxidizing acid (A) and for the same acid containing C1- ions (B). Polarizing the sample from the open circuit potential to more electropositive potentials initially leads to a higher corrosion rate. Above the passivation potential ~'pp a sudden reduction in anodic current density occurs due to the formation of a film which acts as an efficient barrier to the corrodent. Thus, passivity is seen to markedly reduce corrosion to a low level indicated

NOBLE /

THERMOMETER

[7

(A)

f.-- /

r " 'AS OUTLET

@

TRA NSPASSI VE

o

ET t

i

,,z, EBr

oE-

/CON.ECT'O" \\

~ H / II

13 H )::::=::~1 I[ @

J:l.I-l/

/---AUXILIARY ~LILltII / EHLoE(g##DE

(B)

~o

!

wz

II I I

oQs

~

PASSIVE

~g

~m

l

d

2g

~)

ACTIVE-PASSIVE TRANSITION J

ACTIVE

g

CATHODIC •

ELECTRODE Fig. 2. Standard multineck polarization cell, ref. [ ] ].

ieorr ~oss

icri~

LOG CURRENTDENSITY

Fig. 3. Polarization curves of a stainless steel in a mildly oxidizing deaerated acid (A) and in the same acid containing C1ions (B).

C.R. Clayton / Modification of metallic corrosion

by ipass. The

passive trim later breaks down at higher oxidizing potentials (a condition seldom encountered in practice) resulting in a return to higher corrosion rates in the transpassive region. Following curve B of fig. 3, it can be seen that in the C1- solution transpassive breakdown is preceded by a sudden surge in current density, which denotes the localized breakdown of the film and corresponds to pit formation. The higher the potential (EB) at which breakdown occurrs, the lower is the probability that pitting will occur in a natural corrosion environment, since ha practice most corrosion environments are limited to an oxidizing power equivalent to a moderate level of anodic potential.

3. Alloying for corrosion resistance by ion implantation In the design of corrosion-resistant conventional alloys the most common approach is to promote passivity by adding elements which tend to lower the passivation potential and critical current density. This philosophy is typified by the addition of Cr, Mo and Ni to steel to form stainless steel [2]. In certain cases it is necessary to alloy a reactive metal with an inert species wich is capable of stimulating cathodic kinetics in order to promote passivity. Titanium has been alloyed with dilute concentrations of Pd to achieve self-passivating alloys by producing a galvanic couple having a corrosion potential that is more noble than the passivation potential for Ti. [2]. Where active dissolution is inevitable, i.e., where passivity cannot be maintained, elements may be added to decrease cathodic activity and thus to stifle the corrosion process [2]. Ion implantation, as a surface alloying technique, offers greater scope for alloy design for corrosion resistance, since it has been shown that in many cases solid solutions may be made from elements which are normally immiscible under equilibrium conditions [3]. This is an important advantage, since conventional alloys which are not single phase are susceptible to local galvanic attack at phase boundaries. In addition, amorphous phases may be produced by ion implantation, and should offer the same potential for ultra-high corrosion resistance as those produced by splat-quenching. In splat-quenched amorphous alloys extremely high levels of corrosion resistance can be achieved when the correct passivators are added [4]. This is because in addition to the absence of grain

867

boundaries, which produce imperfections in passive films, the alloys have a high degree of chemical homogeneity. In the case of implanted metals amorphicity leads to the removal of grain boundaries; however, if the substrate has a second phase present, then the resulting lack of chemical homogeneity may result in a slightly lower corrosion resistance than might be expected from the splat-quenched alloys. Nevertheless, with pure metals equivalent corrosion resistance should be achievable.

4. The effects of the implantation process on corrosion behavior The effectiveness of ion implantation as a means of modifying metallic corrosion is dependent not only on the original alloying concepts, but also on the following conditions: surface contamination; the nature of the surface oxide; implantant distribution; defect concentration; phase structure.

5. Surface contamination For corrosion studies of implanted metals it is common practice, in the absence of an ultra-high vacuum, for a cold trap to be used in order to reduce the level of hydrocarbons in the vicinity of the sample and thus prevent the build-up of a corrosion inhibitive layer of carbon on the sample surface resulting from hydrocarbon cracking. Covino et al. [5] have reported that surface contamination of Si, C1 and C remained after implantation of Cr into Fe. Carbon was also observed to be distributed throughout the alloy layer, perhaps due to recoil implantation of surface carbon. They report, however, no observed effect on the electrochemical behavior of the surface alloy.

6. The nature of the surface oxide Art inevitable surface contaminant, however, is the surface oxide which is formed on the freshly implanted sample during exposure to the atmosphere. Several workers [6,7] have reported that air-formed films tend to grow to a greater thickness on implanted VII. ION-IMPLANTEDMETALS

868

C.R. Clayton / Modification of metallic corrosion 1200

I

.~ I

.~

.:" ,.~. ...... //

............

~

-

350

-750

i

.v'":f/ /. .: f k

850

tO

/

i

// //¢/ (

I 't POS-GOING SWEEP ----- I s! NEG-GOING SWEEP 2ndpOs-GOING SWEEP

2-~

--

-n5o

0.01

I 0.1

I I I I0 CURRENT DENSITY, #~A/mrn p

Fig. 4. P.otentio-kinetic polarization curves of Ar implanted Fe, ref. [6].

surfaces compared to those of unimplanted metals and alloys. Two factors may be responsible for this effect, namely the high defect concentration produced by radiation damage, which would stimulate diffusion and oxidation, and the temperature to which the sample has cooled down to when it was exposed to the atmosphere. It should be noted that Okabe et al. [8] report that iron samples reached about 180°'C during implantation with Ni + and Cr + at 150 keV, operating with a beam current of about 2 gA cm-:. The nature of the air-formed film is important since it can act as a sink for the main passivator, tending to deplete the surface alloy. In using ion implantation as a research tool to understand a corrosion mechanism this elemental depletion may in some instances be deleterious to the experiment. Conversely, in the context of corrosion protection, the air-formed film can play a very important role since in solution it can often modify in composition and structure to facilitate passivation. Ashworth et al. [6] have taken account of this in their analysis by carrying out a three-sweep polarization experiment (fig. 4). The first positive-going sweep from the immersion potential considers the repair and modification of the air-formed film a~d on achieving a maximum potential the sweep is reversed whereupon film thickening occurs, until at the open circuit potential, the sweep enters the cathodic region where the hydrogen evolution reaction dominates and the film becomes reduced. The final positive sweep determines the anodic and cathodic kinetics of the cathodically cleaned surface.

7. I m p l a n t a n t

distribution

The depth distribution of the implantant can be altered by radiation enhanced diffusion, which when combined with sputter damage can result in considerable deviation from the distribution calculated from the commonly applied theory of Lind_hard, Scharff and Schiott (LSS) [9]. An example of this is given in fig. 5 which compares the LSS derived theoretical profiles (corrected for sputtering) for the distribution of Ni implanted into 430 stainless steel with the distribution determined by XPS and Ar ÷ milling [10]. The nickel is seen to be enriched in the outer regions of the surface alloy. This contrasts to the redistribution of Cr shown in fig. 6, where it is seen that chromium is depleted from the outer region and enriched at a greater depth [10]. This cannot be explained by selective elemental sputtering. The most satisfactory explanation of this form of elemental redistribution has been given by Okamoto and Wiedersich [11] who found a similar redistribution of Ni and Cr in a 316 stainless steel implanted with Ni at 3.25 MeV. Their model suggests a flow of radiation4nduced interstitins, which in turn causes a flow of solute elements CALCULATED FINAL SURFACE

ORIGINAL SURFACE

40 ~

I

I

I

I

::.36--

LSS :32-28--

o 24--

ESCAN \

z 0 F- 2 0 - -

\

t..r,

z~ 16

w tj z



~z

q

8 4 0

I 200

I 400

I 600

8OO

DISTANCE (~)

Fig. 5. Comparison of calculated (LSS) and observed (ESCA) distributions of Ni4mplanted 430 stainless steel (150 keV, 5 × 1017 ions cm-2) ref. [10].

C.R. Clayton/Modification o f metallic corrosion O,4O

I

I

I

i

l

iO-I

I

I

0.35 i0-z

f

I

Fe-6.6(s)Cr Fe- 6.6 (s) Ni

.... -- -----Fe-

>.

v

I

I ~,,,~/~" .~,~. ~'~" " / ~ -

6 . 6 [s) AI

..~/~ S /"

- - - - Fe-13.3(s) Or ..... Fe

0.30

/.Q.Q.Q.Q.Q~.Q.Q./Q.~I

/ i

J

I 0 -3

F-

0.25

,,=, ~- 10-4

"-. 0.20

0.15

/"-".

Z .Z¢//////

iO-5 A 30 at. % Ni IMPLANTED o 15 a t % Ni IMPLANTED o 5 Q1.% Ni iMPLANTED

0.10

10. 6 -0,6

0.05 t 0

869

0

[

I

I0

20

~

I

I

30 40 50 ETCH TIME (rain)

I 60

70

Fig. 6. Relative concentrations of C~ and Fe in the region of implantation u 5 at.% Ni implanted, o }5 at.% Ni implanted, zx 30 at.% Ni implanted, ref. [10].

in different directions according to atomic misfit. The undersized Ni atoms are considered to favor interstitial sites (ex., in the dumbell configuration) whilst the oversized solute atoms such as Cr diffuse away from sinks by vacancy migration. Clearly the redistribution of the two passivators Ni and Cr will radically alter the intended composition of the surface alloy and consequently the corrosion behavior. In the case of chromium4mplanted Fe, surface segregation can occur to such an extent that the resulting surface peak formed can be sputter removed, leading to a net loss of the implantant. This has been shown by XPS [121 and SIMS [81 studies.

8. Defect concentration Ashworth et al. [6] and SartweU et al. [13] have considered the role of defects on the modification of metallic corrosion. Ashworth [6] found that Ar implantation in Fe resulted in a thickening in the airformed film, which altered the immersion potential and improved the corrosion resistance of Fe polarized in pH 8.4 solution. Removal of the oxide layer by cathodic reduction caused the sample to revert back to the original corrosion behavior of pure Fe.

-0.4

-0,2 0 0.2. 0.4 ELECTRODE POTENTIAL (volts)vs N.H.E

0.6

Fig. 7. Anodic polarization curves for several different surface alloys, ref. [13].

Sartwell et al. [13] compared the corrosion behavior of Fe implanted with Cr and Ni which have similar masses and, therefore, are likely to form the same level of defect damage. The results given in fig. 7 show that a distinct difference is seen between the two surface alloys attributed to the intrinsic chemical properties of the implantant. A1 implantation was also compared with that of Cr and Ni since it has a lower mass and is likely to produce a lower defect concentration. Again improvements in the corrosion behavior were observed (fig. 7) suggesting that radiation damage alone is not strongly influencing the corrosion mechanism. From the preceding discussion on radiation enhanced diffusion it is expected that defects may affect the implantant distribution. Indeed it is arguable that volume diffusion during selective dissolution may be enhanced by the high defect concentrations of implanted metals. Whereas the kinetics are very slow, nevertheless such diffusion has been observed in several systems [4,15]. More recently Hubler et at. have reported that in studies of Pd implanted Ti RBS showed the possible diffusion of Pd into Ti during the dissolution of the surface alloy in boiling H~SO4 [16]. In another study, Zamanzedeh et at. [17] have found similar evidence by RBS of the diffusion of Pt into Fe during the selective dissolution of Fe in H2SO4 under potential control. Since the misfit of solute atoms appears to be an important factor in addition to defect concentration, it would appear that inert gas implantation into pure metals does not VII. ION-IMPLANTED METALS

870

C.R. Clayton/Modification of metallic corrosion

adequately test the effect of defects on the corrosion mechanism.

I

I

I

I

I

I ------

E I I

9. Phase

structure

I

I

I

I

3 0 4 S.S. 5 0 4 S.S. P IMPLANTED 1N HzSO 4 + 2 % N e C [

>.. I0 I-2~

One of the major advantages of ion implantation is that by choosing suitable implantation parameters solid solutions may in many cases be made between two or more species which are virtually insoluble under equilibrium conditions. Ashworth et al. [8] have for example shown that Ta4mplanted Fe (20 keV, 5 X 1016 and 2 X 1017 ions cm -2) behaved as a single phase solid solution. Improvements in the general corrosion behavior were found in pH 7.2 soclium acetate/acetic acid buffer solution comparable to that found for a conventional 4.9 at.% C r - F e alloy. No confirmation of the phase structure of this alloy was reported. This work also showed that the surface alloy retained a considerable level of corrosion resistance even after dissolving the surface of the sample to a depth equivalent to the thickness of the surface alloy. Grazing angle RBS analysis revealed the retention of Ta resulting from the selective removal of Fe. The tendency for implanted surface aUoys to produce solid solutions enabled Wolf et al. [22] to implant the elements Ne, At, Cu, Pb and Au into Fe, all having different levels of solubility in Fe. These workers investigated the possibility of lowering the corrosion rate of Fe in acid sulphate solute by inhibiting the cathodic hydrogen evolution reaction. It was found that generally Au increased the corrosion rate by increasing the rate of cathodic kinetics, whilst Cu and, to a greater extent, Pb inhibited the hydrogen evolution rate. Some discrepancies were, however, observed in the weighted exchange current densities and several causes of this were discussed. Inert species were found to lead to a slight increase in corrosion rate which was attributed to the formation of defects according to the interpretation of Lorentz [23]. In addition to the conventional approach to alloy design for corrosion resistance, ion implantation also offers the opportunity to experiment with unconventional alloyNg additions such as metalloids. In a recent study [19] phosphorus was implanted into 304 stainless steel in order to form a glassy or amorphous type of surface alloy, following the work of Whitton et al. [20]. Implantation at 40 keV and a fluence of 1017 ions cm -2 resulted in the suppression of phosphide formation, as indicated by both TEM

I

t I

102 I I

I ! v

I0:

I -400

I -200

I 0

I

J

I

[

J

i

200

400

600

800

I000

1200

ELECTRODE POTENTIAL (mV) vs SCE

Fig. 8. Anodic polarization for unimplanted and P-implanted 304 stainless steel, ref. [211. and XPS analysis. However, some evidence of microcrystallinity was apparent from TEM at a magnification of 180000X. The system revealed a marked reduction in the critical current density in deaerated 1 M H2NO4 solution. XPS revealed that the passive £flm formed on the unimplanted 304 steel was based on a chromium enriched oxy-hydroxide structure, whereas the implanted steel formed a mixed Fe, Cr hydroxide film. Initially breakdown in a 2% NaC11 M H2SO4 solution revealed a lower breakdown potential than for the unimplanted steel. However, improvements in the heat sinking of the samples resulted in an improvement in both the general and localized corrosion resistance as shown in fig. 8 [21]. This would indicate that heat sinking has insured complete amorphicity.

10. Applications of ion implantation to engineering materials In this section a few examples will be given in which ion implantation has been carried out on pure metals and engineering alloys. Preliminary studies of this nature are important in determining whether ion implantation can be used to produce surface alloys which can form the basis for studies of the mechanism of pitting corrosion in conventional alloys. This work also serves to determine the potential in applying ion implantation as a corrosion protection treatment for specialized systems. Perhaps the most critical test of ion implantation as a viable surface treatment for corrosion protection is the performance of the implanted surface in solutions which promote pitting. The pitting of surface

871

C.R. Clayton / Modification of metallic corrosion

alloys will inevitably penetrate the surface alloy whereupon pit growth will continue by galvanic corrosion in the "interfacial" region of the surface alloy and substrate. Covino et al. [5] have carried out anodic polarization of cathodically treated ion implanted pure Fe and F e - N i alloys in deaerated pH 8.5 buffered solutions containing 2400 ppm C1-. Anodic polarization of Fe implanted with Ni ÷ (25 keV, 8.5 X 10 is ions cm -=) and Cr+ (25 keV, l ~ X 1016 ions cm -2) resulted in improvement in the active passive transition as well as pitting resistance. In a further study of Covino et al. [24], tertiary surface alloys containing 15 at.% Ni (25 keV, 2.05 X 10 a6 ions cm -2) and 10 at.% Cr ÷ (25 keV, 2.08 X 1016 ions cm -2) were compared with 316 stainless steel and Vascomax 250 (V-250) maraging steel implanted with Cr÷ (25 keV, 2.08 X 1016 ions cm-2). The general and localized corrosion resistance of pure Fe and V-250 were improved by the implantation treatment. A1-Saffar et al. [25] have reported the effect of molybdenum implantation into pure AI and a high strength aluminum alloy (7075-T6). The general corrosion resistance was studied in a pH 7 Na2SO4 solution, and pitting work was carried out in the same solution containing 1000 ppm CI-. Implantation with Mo + (20 keV, 1017 ions cm -2) resulted in significant improvements in both the general corrosion and pitting resistance of pure A1 and 7075-T6 alloy. Wang et at. [12] have carried out pitting and general corrosion studies of ion implanted M50 bearing steel, as part of a project aimed at improving the corrosion resistance of bearings commonly used in turbo-jet engines. These studies considered separate implantations of the following: Cr ÷ (150 keV, 1.5 X 1017 ions cm-2); Mo + (100 keV, 5 X 1016 ions cm-2); Ti+ (55 keV, 2 X 1017 ions cm-2); Cr + and Mo + double implantation was also carried out. Polarization in deaerated 1 N tt2SO4 showed that the Cr and Mo implant resulted in the best general corrosion behavior. Fig. 9 shows the marked reduction in the critical current density. A very substantial improvement in pitting resistance was found for the Cr + Mo implant. Polarization in a pH6 buffered 0.1 M NaC1 solution (fig. 10) shows a marked enoblement of the breakdown potential for each of the implanted samples. The improvement in breakdown potential ranged from 2 5 0 - 1 5 0 0 mV s. These results have recently been compared with long term corro-

1750

~

I

I

1500 -1250 I000 co 7 5 0 .-?

I

Cr, Mo

m

!

500

/ ,,;.J\

,.z 2 5 0

bo. 0 -250

m

--

- 500 --

I

-750

I

102

I0

I

I

103 104 CURRENT DENSITY ( p A / c m 2)

105

106

Fig. 9. Potentio-kinetic polarization curves produced in hydrogen-saturated 1 N H2SO4 for M50 steel and for M50 steel implanted with Ti, Cr, Mo and Cr + Mo, ief. [ 12]. sion tests which simulate the corrosion conditions existing in turbo-jet lubricants. The electrochemical tests were shown to predict accurately the long term resistance of the surface alloys to localized corrosion. In addition to applying ion implantation for the

1750

m

1500 --

Cr, Mo ( I )

1250 Cr, Mo (11)

1000

Cr

8 • ¢n 7 5 0

y,/

/ //

/

500

..J 1 ~z 2 5 0 LU

a.

fl'-

- -~'"

~¢'"

-~ f lk -

~'-- M- 5 0

0 -250 -500

I0

I

I

I

I

102

I03

104

IO 5

106

CURRENT DENSITY (/.LA/crn2)

Fig. 10. Potentio-kinetic polarization curves produced in a buffer solution of pH6 containing 0.1 M NaC1 for M50 steel and for M50 steel implanted with Ti, Cr, Mo and Cr + Mo, ref. [12]. VII. ION-IMPLANTEDMETALS

872

C.R. Clayton / Modification of metallic corrosion

improvement of the corrosion resistance, studies have been carried out in which the mechanism of corrosion of a commercial alloy have been determined by testing surface alloys. A1-Saffar et al. [26] have recently investigated the mechanism by which sacrificial anodes made from a commercial A l - Z n - H g alloy maintain high anodic currents by the depassivation effects of the Fig component. Sacrificial anodes are commonly attached to steel structures in sea water since the steel becomes cathodically polarized and hence immune to corrosion attack when connected to the more active alloy. Since Hg is insoluble in A1 and tends to form grain boundary precipitates, the exact mechanism by which Fig prevents passivation (i.e., enoblement) of the AI anode is not clear. The authors examined the extent to which Hg implanted A1, which forms a solid solution, can prevent passivation. Carrying out potentiokinetic analysis of the implanted A1 and by studying galvanic corrosion of a Fig implanted A1 sample connected to a mild steel plate, they found that corrosion of the implanted AI sample was uniform and that this behavior was sustained to a depth deeper than the depth of implantation. The authors have postulated, on the basis of these experiments, a new model, involving surface diffusion, for the mechanism by which Hg prevents passivation in the commercial A 1 - Z n - H g anodes. Finally, there appears to be several reports concerning the success with which ion implantation has been used to improve surface mechanical properties of metals and alloys [27]. Indeed one research group has recently improved the cavitation (corrosionerosion) resistance of 1018 steel by implantation with N [28]. It is therefore surprising that no published reports, to the author's knowledge, have been made of the successfull use of ion implantation in improving the resistance of metallic systems to corrosion fatigue or stress corrosion. Ion implantation would appear to be a promising technique for improving resistance to both of these forms of conjoint attack, and indeed for the study of the mechanism of breakdown.

face alloys formed in order to modify corrosion behavior. It was also intended to illustrate the great potential which this surface treatment offers as a research tool in evaluating corrosion mechanisms, as well as a future technique for the corrosion protection of engineering materials. Some general conclusions can be drawn from the studies reviewed, and they may be stated as follows: Whereas a few studies have considered the effect of the ion implantation process upon corrosion behavior, there is certainly considerable scope for further work. In particular the role played by defects in modifying corrosion behavior is an important issue and deserves more consideration than has been given to it so far. It has been shown that ion implantation can be used to control corrosion processes systematically. The technique, therefo.re, appears to have considerable potential as a research tool for more definitive studies of passivity, localized breakdown of passivity, and for studies concerning selective dissolution. In considering ion implantation as a suitable corrosion protection treatment it is clear that more basic studies have to be carried out before the full potential of this surface alloying technique can be realized. In addition to the studies of the effects of the implantation process upon corrosion behavior, studies must be carried out wkich aim at developing the mechanical as well as chemical properties of surface alloys in order to produce surface alloys capable of withstanding stress and wear as well as strictly chemical attack. In this context the development of high strength highly corrosion resistant amorphous surface alloys is worth consideration. Finally, since passivation plays a role in stress corrosion and corrosion fatigue, it would appear that ion implantation may have a similar, as yet unconfirmed, potential as a research tool for studies of these forms of corrosion. The author wishes to thank the Office of Naval Research for financial support of ion implantation studies carried out at SUNY and to Philip A. Clarkin for his encouragement.

11. Summary and conclusions This brief review has not attempted to discuss all papers which have been written on this subject. It was instead intended to highlight some of the important aspects of the ion implantation process and the resulting properties which must be controlled of the sur-

References [1] 1974 Annuai Book of ASTM Standards, Part 10, American Society for Testing and Materials Philadelphia, Penn. Pa. (1974).

C.R. Clayton / Modification of metallic corrosion [2] M.G. Fontana and N.D. Greene, Corrosion Engineering, 2nd ed. (McGraw-Hill, New York, 1978) p. 325. [3] J.A. Borders and J.M. Poate, Phys. Rev. B13 (1976) 969. [4] K. Asami, T. Hashimoto, T. Masumoto and S. Shimodaira, Corr. Sci. 16 (1976) 909. [5] B.S. Covino, B.D. Sartwell and P.B. Needham, J. Electrochem. Soc. 125 (1978) 370. [6] V. Ashworth, W.A. Grant, R.P.M. Procter and T.C. Wellington, Corr. Sci. 16 (1976) 393. [7] S.B. Agarwal, PhD dissertation, State University of New York at Stony Brook (1979). [8] Y. Okabe, M. Iwaki, K. Takahashi, H. Hayashi, S. Namba and K. Yoshida, Surf. Sci. 80 (1979) 257. [9] J. Lindhard, M. Scharff and H.E. Schiott, Dan. Vial. Selsk. Mat. Fys. Medd. 33 (1963) 3. [10] S.B. Agarwal, Y.F. Wang, C.R. Clayton, H. Herman and J.K. Hirvonen, Thin Solid Films 63 (1979) 19. [11] P.R. Okamoto and H. Wiedersich, J. Nucl. Mat. 53 (1976) 336. [12] Y-F Wang, C.R. Clayton, G.K. Hubler, W.H. Lucke and J.K. Hirvonen, Thin Solid Films 63 (1979) 11. [13] B.D. Sartwell, A.B. Campbell and P.B. Needham, Proc. Fifth lnt. Conf. on Ion Implantation in Semiconductors and Other Materials, 1976, Boulder, Colo., USA. [14] H.W. Pickering and C. Wagner, J. Electrochem. Soc. 114 (1967) 678. [15] H.W. Picketing, J. Electrochem. Soc. 117 (1970) 8. [16] G.K. Hubler and E. McCafferty, Corr. Sci. 20 (1980) 103.

873

[17] M. Zamanzadeh, A. Allam, H.W. Picketing and G.K. Hubler, private communication. [18] V. Ashworth, D. Baxter, W.A. Grant and R.P.M. Proctor, Corr. Sci. 17 (1977) 947. [19] C.R. Clayton, K.G.K. Doss, H. Herman, S. Prasad, Y-F Wang, J.K. Hirvonen and G.K. Hubler, Proc. Materials Research Society Boston 1979. [20] J.L. Whitton, W.A. Grant and J.S. Williams, Proc. Int. Conf. on Ion Beam Modification of Materials, Budapest, Hungary (1978). [21] C.R. Clayton, K.G.K. Doss, H. Herman and J.K. I-Iirvonen, unpublished work. [22] H. Ferber, A. Kasten, E.K. Wolf, W.J. Lorenz, H. Schweickert and H. Folger, Corr. Sci. 20 (1980) 117. [23] F. Hilbert, Y. Miyoshi, G. Eichkorn and W. Lorenz, J. Electrochem. Soc. 118 (1971) 1919 [24] G.S. Covino, B.D. Sartwell and P.B. Needham, J. Electrochem Soc. 125 (1978) 370. [25] A.H. A1-Saffar, V. Ashworth, A.K.O. Bairamov, D.J. Chivers, W.A. Grant, R.P.M. Procter, Corr. Sci. 20 (1980) 127. [261 A.H. A1-Saffar, V. Ashworth, W.A. Grant and R.P.M. Procter, Corr. Sci. 18 (1978) 687. [27] Proc. Materials Research Society Symp. on Surface Modification of Materials by Ion Implantation, Cambridge, Mass., USA, Nov. 1979. [28] W.-W. Hu, C.R. Clayton, H. Herman and J.K. Hirvonen, J. Mat. Sci. and Eng. 45 (1980) 263.

VII. ION-IMPLANTED METALS