I.M. Zin, S.B. Lyon, S.J. Badger*, J.D. Scantlebury, V.I. Pokhmurskii
Corrosion and Protection Centre, UMIST, P. O. Box 88, Manchester, M60 1QD, UK, *Uniscan Instruments Ltd, Sigma House, Burlow Rd, Buxton, SK17 9JB, UK.
The objective of this work was to study the mechanism of corrosion inhibition by pigments used in coil-coated cladding with focus on chromate substitutes able to provide phosphate and calcium ions. Such pigments include commercial compounds e.g.: Ca2+ ion exchanged silica and zinc phosphate/molybdate.
Keywords: inhibitive pigments, galvanised steel, impedance, cut-edge corrosion, SVET
Currently intensive investigations are underway worldwide with the aim of finding substitutes for chromate inhibitors. Phosphates are one of many possible alternatives. They are usually classified as non-oxidising anodic inhibitors in which the presence of oxygen is required for corrosion protection [1,2]. However under certain conditions, phosphates may also act as cathodic inhibitors. Their action consists in precipitation of a phosphate film on a metal surface and quality of the film (i.e. compactness and barrier properties) defines the inhibition efficiency.
Uhlig et al.  found that the enhanced efficiency of polyphosphate in reducing the corrosion rate, in conjunction with the divalent ions Ca2+ or Zn2+, was connected with the formation on the cathodic areas of a diffusion barrier layer impeding access of oxygen to the iron surface. Szklarska-Smialowska and Mankowski  found from measurement of the potentiodynamic kinetic polarisation of iron in dilute aerated orthophosphate solutions containing Ca2+, that orthophosphates inhibit the cathodic process by covering the surface with a film of sparingly soluble calcium phosphate as the result of the reaction:
2HPO42- + 2OH- + 3Ca2+ = Ca3(PO4)2 + 2H2O
The OH- ions involved in the reaction are derived from the cathodic reduction of oxygen. The film is claimed to be effective by limiting access of dissolved oxygen to the metal surface. Andrzejachek  found that the cathodic reduction of oxygen on iron in tap water is inhibited by acidified Na3PO4 solution as a result of cathodic adsorption of positively charged colloidal particles of calcium and magnesium phosphates. Thus, in this way a highly protective film can be deposited on metal surfaces in solutions containing phosphate inhibitors and calcium salts. However, it is well known that the function of soluble inhibitors in corrosive solutions differs from their behaviour when used as inhibiting pigments distributed in an organic coating. Additionally, the screening effect of a polymeric binder influences the leaching of inhibitive ions into the solution. This causes the pigment solubility to be much lower compared to solution-phase inhibitors. Thus, this work describes an investigation of the synergetic inhibitive properties of calcium ion-exchanged silica combined with zinc phosphate/molybdate in the solution phase as well as dispersed in an organic coating.
The corrosion behaviour of bare galvanised steel was investigated using electrochemical impedance spectroscopy (EIS). The measurements were carried out using a three-electrode electrochemical cell with a sample testing area of 4.5 cm2 and with a platinum auxiliary electrode as a counter and saturated calomel as reference electrode. The impedance measurements were carried out close to the corrosion potential using an ACM potentiostat and a Solartron Frequency Response Analyser 1250 in the 10000 to 0.005 Hz frequency range. The signal amplitude was 10 mV. The impedance spectra were interpreted with the Equivalent Circuit software program written by Boukamp . Polarisation measurements on the same samples were carried out using the same three-electrode electrochemical cell and an ACM potentiostat.
All electrochemical measurements were carried out in an artificial acid rain solution (at the natural concentration) . Extracts of the pigments zinc phosphate/molybdate (Actirox 106), calcium ion-exchanged silica (Shieldex CP7394), strontium chromate and the double extract of zinc phosphate/molybdate and calcium ion-exchanged silica were prepared by introducing 2 g of each pigment into 1 litre of acid rain solution, stirring for 24 hours and then filtering twice.
The ability of the various pigment extracts to reduce galvanic current between zinc and steel at a cut edge was evaluated using a model cell . This consisted of a "sandwich" of steel sheet, plastic film and zinc foil mounted in epoxy resin, Figure 1, and immersed in artificial acid rain solution saturated with each of the inhibiting pigments. The zinc and steel components of the cell were connected to a Zero Resistance Ammeter (ZRA) in order to measure the galvanic current.
Figure 1: Schematic diagram of galvanic sandwich cell
Coated samples were prepared from hot-dip galvanised (HDG) mild steel sheet, with a zinc layer of 15 m m. An inhibitor-containing polyester primer (5 m m) and polyester top-coat (20 m m) were applied using a draw-bar. Four different primer formulations were used, namely zinc phosphate/molybdate, calcium ion-exchanged silica, strontium chromate and a zinc phosphate/molybdate plus calcium ion-exchanged silica blend. Laser-ablation was used to place a 1 cm long scribe through the zinc, primer and topcoat layers. The laser spot size (and hence scribe width) was 180 m m, and the depth was 50 m m thus exposing the steel substrate.
Scanning Vibrating Electrode measurements were performed using an EG&G Instruments SVP100. This instrument was used to make measurements of the local potential gradients (and hence the localised corrosion activity) at scratches in pigmented polyester coatings applied onto galvanised steel. The instrument incorporates a platinum microelectrode, which is electrochemically sharpened to 5 m m diameter at the tip and also platinised in order to reduce the interfacial impedance at the electrode/solution interface. This microelectrode is vibrated in a plane perpendicular to the sample surface, at an amplitude of 30 m m, and the potential is measured at the upper and lower limits of the amplitude via an electrometer and lock-in amplifier. All experiments were carried out at the free corrosion potential.
Impedance of galvanised steel in pigment extracts.
Electrochemical impedance spectroscopy data of galvanised steel in all single extracts of the pigments and in the zinc phosphate/molybdate plus calcium ion-exchanged silica blend show a depressed semicircle, Figure 2. In case of both chromate and the blend the full semicircle cannot be observed, due to the very effective corrosion inhibition.
Figure 2: Impedance of galvanised steel after 1 day immersion in acid rain solutions saturated by: -¨ -zinc phosphate/molybdate; -n -; calcium ion-exchanged silica; -D - zinc phosphate/molybdate + calcium ion-exchanged silica; -*- Strontium Chromate
Calculation of the impedance components was carried out with the equivalent circuit Re(QdlRt), where Re is the resistance of the electrolyte, Rt the metal transfer resistance, and Qdl the constant phase element. The replacement of conventional double layer capacitance by the Qdl in the equivalent circuit gives a good simulation of the experimental data with a minimal error.
It was established, Figure 3, that the initial charge transfer resistance is about 10 times higher for galvanised steel in strontium chromate and the zinc phosphate/molybdate plus calcium ion-exchanged silica mixture compared to zinc phosphate/molybdate or calcium ion-exchanged silica alone. This clearly shows a considerable synergetic effect due to the inhibitor combination, because the sum of impedance values observed in the single pigment extracts is about 5-10 times less compared to the impedance in the blend extract. One possible explanation of the effect is the formation of a highly protective surface film on galvanised steel caused by an interaction of phosphate and calcium ions, released from silica after ion exchange, and zinc surface. Changes in the admittance Yo and the component n of Qdl confirm the film development.
Figure 3: The variation of Rct with time for galvanized steel in acid rain solution saturated by: : -¨ -zinc phosphate/molybdate; -n -; calcium ion-exchanged silica; -D - zinc phosphate/molybdate + calcium ion-exchanged silica; -*- Strontium chromate
Inhibition of cut edge corrosion of coil-coated galvanised steel in acid rain by chromate-free pigments.
All saturated solutions to some extent decrease the galvanic current in the model cell compared with the blank solution, Figure 4.
Figure 4 The variation of galvanic current of the model cell in acid rain saturated by different pigments (click for a full-size image).
The strontium chromate solution has the best inhibitive properties with the current decreasing to 0.5 mA after only 100 minutes exposure. Other pigments (zinc phosphate/molybdate and calcium ion-exchanged silica) during the test reached only a level of 4 mA and their inhibitive efficiency is thus about 50%-60% compared to the corrosion of the metal in the blank acid solution. The acid rain solution, saturated with both zinc phosphate/molybdate and calcium ion-exchanged silica, is close in efficiency to chromate solution with a minimal current of 0.8-1.0 mA. However, this solution differs from the chromate solution by the longer period which is needed to reach the minimal steady galvanic current.
The role of silica in the inhibiting system zinc phosphate/molybdate plus calcium ion-exchanged silica on the kinetics of electrochemical reactions is not so clear. Amaral and Muller  have found that addition of silica (SiO2) to alkaline solution increased film resistance, decreased its capacitance and affected the kinetics of interfacial electrochemical reactions, leading to an increase in the charge transfer resistance. Armstrong and Zhou  concluded that calcium ion-exchanged silica pigment could act through formation of Ca2+ and polysilicate ions in solution. The polysilicate anions adsorb on the iron surface and inhibit oxygen reduction and iron dissolution.
SVET study of organic coatings containing inhibitive pigments.
It was interesting to investigate the effectiveness of these chromate-free pigments when they are incorporated into an organic coating. The protective action of pigments starts after coating damaging, therefore coatings with artificial defects were studied. The corrosion of coated metallic samples with through-coating defects takes place on a local scale and clear separation of anodic and cathodic parts must exist. The Scanning Vibrating Electrode Technique (SVET) is a promising method to study localised corrosion processes [15-18], and therefore was employed to measure the localised corrosion activity in the vicinity of laser-ablated artificial defects.
The measured distributions of potential around the scratches, upon initial exposure and after 7 days, are given as 2-dimensional area maps for each inhibitor. As indicated by the shaded scale below each map, relatively dark areas on the map represent local anodic sites, whereas the lighter areas represent cathodic sites. The map for initial exposure of the zinc phosphate/molybdate/calcium ion-exchanged silica mixture, shown in Figure 5 (left picture), shows intense corrosion activity at the scribe, with an uneven distribution of relatively negative (anodic) and positive (cathodic) sites along the scribe length. Isaacs and co-workers have previously observed these non-symmetrical distributions at mechanically produced artificial scribes on painted zinc and zinc alloy coated steels . The effect was attributed to the localised film breakdown at certain areas on the mostly passive zinc surface, which subsequently lead to intense pitting corrosion. Figure 5 (right picture) shows the area map of solution potential after 7 days, with a clear reduction in the measured activity.
Figure 5: SVET maps of local solution potentials for the zinc phosphate/molybdate plus calcium ion-exchanged silica pigment combination: left picture initial exposure, right picture after 7 days (click for the full-size image).
The initial SVET potential map for strontium chromate also shows a number of local anodic and cathodic sites along the scribe length, though the initial intensity of these active sites is less than that for the zinc phosphate/molybdate plus calcium ion-exchanged silica combination. After 7 days immersion, the corrosion activity at the scribe has visually decreased to a level comparable with that for the zinc phosphate/molybdate plus calcium ion-exchanged silica combination. Similar trends are seen when comparing the data for zinc phosphate/molybdate and calcium ion-exchanged silica separately. However, in order to make a quantitative assessment of activity, the total anodic potential measured at every data point was summed for each map and is displayed in Figure 6 which shows the total anodic activity measured daily for all 6 samples. The results for the entire 7-day immersion period indicate that all samples reach approximately the same level of activity after 72 hours. However, closer inspection of the results between 72 and 168 hours gives a clearer indication of the comparative performance. Though the initial summed potential upon immersion for the zinc phosphate/molybdate plus calcium ion-exchanged silica combination was the highest of all the samples, Figure 6 (b) illustrates that the corrosion activity in this sample decreased rapidly up to 96 hours after which it was comparable to that of strontium chromate.
Figure 6: Summed local anodic potential versus time for the period:
1. M. Pryor, M. Cohen, J. Electrochem. Soc. 98 (1951) 263.
2. U.R. Evans, Corrosion and Oxidation of Metals, pub. Arnold, London (1960).
3. H.H. Uhlig, D.N. Triadis, M. Stern, J. Electrochem. Soc.102 (1955) 59.
4. Z. Szklarska-Smialowska, J. Mankowski. Centre Belge D'etude Et De Documentation Des Eaux. 20, N288 (1967) 474.
5. B.J. Andrzejaczek. Br. Corros. J., 14 (1979) 176.
6. B.A. Boukamp, Equivalent Circuit. Version 3.97, Techn. U. of Twente, May 1989.
7. J.B. Johnson, S. Haneef, G.E. Thompson, G.C. Wood, European Cultural Heritage Newsletter, 2, 4, (1988) 13.
8. R.L. Howard, S.B. Lyon, J.D. Scantlebury, Proceedings 13th ICC, Melbourne, Australia, 26-30 November 1996, Paper 022.
9. G.N. Bhar, N.C. Debnath, S. Roy, Surf. and Coat. Technol., 35 (1988) 171.
10. M. Kamrath, P. Mrozek, A. Wieckowski, Langmuir. 9 (1993) 1016.
11. Y. Tanizawa, T. Suzuki, J. Chem. Soc., Faraday Trans. 91, (19) (1995) 3499.
12. B.P.F. Goldie, J Oil Col. Chem. Assoc. N9 (1988) 257.
13. S.T. Amaral, I.L. Muller, Corrosion. 55, N 1 (1999) 17.
14. R.D. Armstrong, S. Zhou, Corros. Sci. 28 , N12 (1988) 1177.
15. H. S. Isaacs, J. Electrochem. Soc., 135, N9 (1988) 2180.
16. F. Zou, C. Barreau, R. Hellouin, D. Quantin, D. Thierry, Galvatech'95, (1995) 857.
17. D. A. Worsley, H. N. McMurray, A. Belghazi, Chem. Commun., (1997) 2369.
18. H. S. Isaacs, A. J. Aldykiewicz Jr., D. Thierry, T. C. Simpson, Corrosion, 52, (1996) 163.