S.H. Zhang, S.B. Lyon
Corrosion & Protection Centre, UMIST,
P.O.Box 88, Manchester, M60 1QD, UK.
Samples of iron were exposed during 1993 in two batches; firstly over the summer and early autumn, and also during the winter and early spring periods. In each case, and as intended, a large majority of the samples were covered in thick adherent corrosion product "rust" after the end of the exposure period. However, and unexpectedly, a few samples from each batch remained visually uncorroded retaining their original bright appearance. This observation is presented as an interesting insight into, and as supporting evidence for, mechanisms of atmospheric corrosion of iron involving passivation.
Recently, there has been increased interest in the detailed mechanism of atmospheric corrosion of iron. This has been driven particularly by the application of novel techniques of investigation from the group of Stratmann in Germany [1-5]. Results from this work have re-emphasised the importance of redox cycles, which occur during wetting and drying episodes, first identified by the earlier work of Evans . Thus, it is now relatively well-established that atmospheric corrosion of iron is intermittent, proceeding when the surface is wet (with a corresponding drop in potential into the active state) and stifling by surface passivation when drying (with a rise in potential of up to or, more than, 0 mV SCE) [1,7,8]. This, then, reveals a picture of a metal in which the normal, passive, state is, from time-to-time interrupted by bursts of corrosion during wet episodes.
The original aim of the atmospheric exposures described herein were to prepare a natural rust film in order to investigate its electrochemical behaviour in-situ. When several electrodes refused to corrode after 4 months we repeated the exposure to confirm the effect. This short communication reports the observations made and provides a comment on likely reasons.
High purity iron (99.99%) was obtained from Goodfellow Metals in the form of wire 0.5 mm in diameter. Short lengths of this wire were then used to prepare semi-micro electrodes, initially for atmosphere exposure but ultimately for electrochemical measurements, as follows. Each cut section was initially coated over its length using a thin layer of epoxy adhesive; this was done to avoid subsequent crevicing of the electrode. After curing for 48 hours at 25°C, the wires were then mounted in standard metallographic moulds using an epoxy resin. After full curing, the electrodes were polished to 800 grit on SiC paper, dried and stored in a desiccator prior to exposure.
An initial batch of ten samples was prepared and exposed for 4 months commencing on June 1st, 1993. Subsequently, a further batch of sixteen samples was prepared and placed outdoors for 3 months commencing February 20th, 1993. Samples were thus exposed outdoors, unsheltered and in a garden environment, over the summer and early autumn periods as well as (separately) during the winter and early spring. The weather was typical of Manchester for the time of year; i.e. generally cool and damp. After exposure, the samples were removed from the exposure site on a dry day, dusted with a soft artists' brush, and stored in a desiccator over dry silica gel prior to photography.
Examination of the first batch, i.e. that exposed for 4 months during the summer and early autumn, indicated that one of the samples was visually uncorroded (bright). All the other samples in this batch were corroded to a similar degree. As this was such an unexpected occurance, this first observation was initially discounted and the specific sample discarded. Nevertheless, a second batch of samples was exposed and examined at more frequent intervals in case the phenomena repeated and, if so, to obtain an estimate of the onset of corrosion. Thus, after 3 months, two (out of sixteen) samples remained apparently uncorroded (bright) with the balance corroded as expected.
This information is summarised below:
Exposure time No. corroded No. uncorroded
1 weeks 4 12 4 weeks 13 3 8 weeks 14 2 12 weeks 14 2
Figure 1 and
Figure 2 show, respectively, low magnification images of a corroded (i.e. rusty) and an apparently uncorroded (i.e. bright) sample after the 4 month exposure period.
Figure 3 shows the other bright sample at higher magnification. The bright samples all show scratches from the initial grinding (it should be remembered that the samples were not polished). On close inspection it can be seen that the "bright" samples do, in fact, have evidence of corrosion product on their surfaces. Thus, in Figure 3, very small, darker areas on be seen although, to the naked eye, they could not be discerned. About 20-30% of the surfaces of the two bright samples were slightly darkened in this way, with 70-80% completely of the surface completely bright.
That bare, untreated iron remains largely uncorroded and passive for several months in a Northern temperate climate is a remarkable result. The reasons for the observations can only be speculative at this stage. In particular, the small size of the exposed area is clearly important as a macroscopic surface could be normally expected to develop a visible patina of rust in a few hours or days depending on the weather. Another factor is the relatively high purity of the iron which limits the presence of second phase particles, especially sulphides. A third reason may well be the wire form of the metal as heavy drawing will tend to reduce the dimensions of secondary phase particles and string them out.
Whatever the reason, this observation is strong evidence that the corrosion on iron and steel in the atmosphere proceeds by disruption of an initially passive film and that the film is normally repairable under conditions of atmospheric exposure.
About 10-14% of high purity iron wire samples, exposed to the atmosphere for periods of up to 4 months, remain uncorroded. This retained passivity is evidence that atmospheric corrosion on iron proceeds by disruption of a passive film and, under normal conditions, the film is self-repairing.