(2) Description of the Prior Art Steel panels and other components are used extensively in the construction of motor vehicle bodies and other structures. The problem of corrosion of steel by environmental factors is well known, and much work has been carried out to provide steel with coatings to reduce corrosion. Zinc coatings are widely used in the protection of steel strip against corrosion. In the automobile industry the introduction of zinc coatings in conjunction with phosphate and/or chromate treatment processes and multiple paint layers has provided six or more years of protection. The phosphate or chromate treatment is necessary to ensure that the zinc-coated surface is sufficiently corrosion resistant and can be electrocoated with a sufficiently coherent paint layer.

In addition to the barrier protection provided by these coatings the zinc can act in a sacrificial manner to prevent rust formation if the steel is exposed by scratching or stone-chipping to the atmosphere. The zinc- coated steel is also capable of being formed to shape and welded.

In the past 15 years many attempts have been made to improve the corrosion resistance of zinc coatings through

alloying of the zinc, for example as disclosed in Japanese Examined Patent (Kokoku) number 50-29821. Electrodeposited Zn-Ni alloy has been widely used to protect steel sheet products, with an improvement in corrosion resistance compared to Zn coatings. Typically, over 12 wt% of nickel is incorporated to provide an improved coating. Other attempts at corrosion resistance improvement include the dispersion of inorganic substances in the zinc, for example as disclosed in EP 0 174 019.

It is an object of the present invention to provide a zinc-based pre-treatment coating for a steel substrate, which has improved corrosion resistance. A further object is to provide a zinc-based pre-treatment coating which is suitable for electrocoating without the need for a phosphate or chromate treatment.

SUMMARY OF THE INVENTION According to one aspect of the present invention there is provided a steel article at least a part of a surface of which is plated with a pre-treatment coating layer comprising at least 90% zinc, plus cobalt, at least one trivalent or higher-valent metal, and at least one colloidal inorganic material.

All percentages are given by weight unless otherwise specified.

In a preferred embodiment, the coating comprises : 92 to 99% zinc ; 0. 5 to 5% cobalt ; 0. 05 to 1. 5% of at least one trivalent or higher-valent metal ; and 0. 4 to 5% of at least one colloidal inorganic material.

The steel article may be steel strip suitable for use in manufacturing motor vehicle bodies.

It is preferred that the higher-valent metal is chromium or molybdenum. Suitable colloidal inorganic materials include silica, alumina, and ferric oxide. A preferred colloidal inorganic material is silica, notably silica having a particle size range of 5 to 30 nm, preferably 10 to 20 nm. For convenience hereinafter, the invention will be described with reference to preferred embodiments in which the colloidal inorganic material is silica, but it is to be understood that the invention is not limited to these embodiments.

In a preferred embodiment, the coating comprises : 92 to 99% zinc ; 0. 5 to 5% cobalt ; 0. 05 to 0. 5% chromium ; and 0. 4 to 5% silica.

In a particularly preferred embodiment, the coating comprises : 93 to 97. 9% zinc ; 1 to 5% cobalt ; 0. 1 to 0. 2% chromium ; and 1 to 3% silica.

Another aspect of the invention provides a steel article at least a part of a surface of which is plated with a corrosion-resistant coating layer consisting essentially of zinc, plus cobalt, chromium, and at least one colloidal inorganic material.

We have surprisingly also found that suitably coated steel

substrates can accept paint without the need for pretreatment by phosphates or chromates, allowing painted steel strip to be produced with fewer production steps and reduced cost.

Without in any way limiting the present invention, the following theory is postulated as a possible mechanism for the improved properties obtained by the coatings. We have found that zinc crystal nucleation and growth occurs in a manner that results in each zinc crystal (about 50 nm thick) being encased in a 4 nm thick zinc oxide film.

When zinc is co-deposited with cobalt, the deposits exhibit differential distribution of the other elements where the major portion associates with the oxide layer.

Here nickel is believed to be acting as a barrier layer which protects each zinc crystal.

The present invention makes use of the fact that zinc oxide is an n-type semiconductor. Zinc atoms in the oxide may be displaced by tri and higher valency ions, for example chromium. This will limit oxide growth and thus enhance its protection of the underlying zinc metal. To enhance the thickness and stability of the nanoscale-thick oxide, one or more colloids are incorporated, for example colloidal silica. We believe that the oxide layer forms from a colloid of zinc oxide as the pH of the plating bath becomes less acid. The additional colloid is present with the ZnO colloidal suspension in the near cathode regions and becomes occluded in the deposit.

The introduction of tri or higher valency elements in the coating is not a simple step as there are limits to the solubility of such elements in the bath. The introduction of silica and the presence of certain levels of divalent cobalt appear to enhance the occlusion of the trivalent

element to be absorbed into the zinc oxide or the zinc metal-oxide interface. This means that both the silica, the divalent metal (Co) and the higher-valent metal are not evenly dispersed in the coating ; they exist as concentrated layers (about 3 to 10 nm thick) surrounding each zinc crystal. These additions modify the size and shape of the zinc crystals. It is believed that this helps to produce a surface profile into which the paint layer can interlock and form an effective bond.

To form the coating, the steel article is electroplated in an aqueous solution of the appropriate metal ions, containing a dispersion of the colloidal inorganic material.

Accordingly, a further aspect of the invention provides a process for applying a corrosion-resistant coating comprising at least 90% zinc to a steel substrate, the process comprising electroplating the steel substrate in an acidic solution containing : zinc ions having a concentration in the range 0. 2 to 2. 5 mol/1 ; divalent cobalt ions having a concentration in the range 0. 10 to 1. 0 mol/1 ; ions of one or more trivalent or higher-valent metals having a total concentration in the range 0. 004 to 0. 05 mol/1 ; and a dispersion of a colloidal inorganic material having a concentration in the range 0. 02 to 0. 2 mol/1.

It is particularly preferred that the plating solution has components in the following concentration ranges : zinc, 0. 5 to 0. 8 mol/1 ; cobalt, 0. 1 to 0. 3 mol/1 ; trivalent and/or higher valent metal, 0. 01 to 0. 03 mol/1 ;

colloidal inorganic material, 0. 05 to 0. 1 mol/1.

The process may be carried out as a continuous process on, for example, strip steel, or as a batch process.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be further described, by way of example, with reference to the following experimental results, and the accompanying drawing in which : Figure 1 is a graph showing comparative corrosion performances of various coatings ; Figure 2 shows the arrangement of scribe cuts used in testing for electrocoat film adhesion ; Figure 3 shows SEM pictures and schematic surface profile drawings of conventional zinc and zinc alloy coatings, and of a coating in accordance with one aspect of the present invention ; Figure 4 shows SEM analysis results for a conventional zinc coating and for a coating in accordance with the present invention ; and Figure 5 shows the breakaway of crystallites from a coating in accordance with the invention under the exposure of a TEM electron beam, and EDX spectra corresponding to the crystallites and the remaining region.

DETAILED DESCRIPTION OF THE EXAMPLES Example 1

A plating solution having the composition set forth in Table 1 was prepared : Component Weight/litre mol/litre ZnSO4-7H2O 165 g 0. 57 <BR> <BR> <BR> CoS°4-7H2° 56 g 0.20 <BR> <BR> <BR> <BR> <BR> Cr2 (S04) 3.xH2O 4 g 0.02* Na(CH3COO).3H2O 14 g Na2SO4 142 g silica** 5 g 0. 08 * concentration of chromium ion **colloidal silica (10 to 20 nm size) from Brent Europe Ltd.

Table 1 Bake hardening steel panels approximately 10 x 10 mm in size were plated using the solution in Table 1, in a static plating bath. Operating conditions were : temperature : 50 to 60°C ; Current density : 120 mA/cm2; pH 2.

Plating was carried out for 90 seconds, to produce a 5 ym coating.

Example 2 A plating solution having the composition set forth in Table 2 was prepared :

Component Weight/litre mol/litre ZnS04. 7H20 550 g 1. 9 CoSO4.7H2O 120 g 0. 44 (NH4)2SO4 50 g 0. 2 (NH4)6Mo7O24.4H2O 0.75 g 0.004* Silica** 5 g 0. 08 * concentration of Mo ion ; ** colloidal silica (10 to 20 nm in size) from Brent Europe Ltd.

Table 2 Bake hardening steel panels of 10 x 10 mm in size were plated using the solution in Table 2, in a static plating bath. Operating conditions were : - Temperature : - 50 °C ; Current Density : 100 mA/cm²; pH : 1. 5 Plating was carried out for 90 and 130 seconds, to produce 5 Am and 7 Am coatings respectively.

TEM analysis of the structure of these coatings revealed a nanostructure pattern in which Zn crystallites form hexagonal platelets of 100 to 200 nm diameter. The platelets are encapsulated by a thin ZnO layer (up to 10 nm thick), where the Mo and Co and silica are enriched.

The detailed technique for TEM analysis of microstructure is described in a paper by H. Yan, J. Downes, P. J. Boden & S. J. Harris, Philosophical Magazine A, Vol. 70 No. 2, 373-389 (1994).

The resultant coated panels from Examples 1 and 2 were tested using the Salt Spray method (ASTM B117 specification). Comparative results with other coatings prepared in a similar manner are given in Table 3, and shown graphically in Figure 1. All of the coatings in Figure 1 are 5 Am thick.

Comparative Corrosion Performance Type of Coating Time to 5% red rust (hrs) Zn 62 Zn-Ni (Ni : 13%) 192 Zn-Co (Co : 1. 5%) 96 Zn-Co-Cr (Co : 1. 5% ; Cr : 0. 1%) 168 Zn-Co-Cr-SiO2 (Co:2.6%; Cr:0.2%; Si02:1.9%) 600 Zn-Co-Mo-SiO2 (Co:2.0%Mo:0.5%; SiO2 :2.0%) 288 Table 3 Both of the coatings in accordance with the invention show improved corrosion performance compared to conventional coatings. The time to corrosion of 600 hours for the 10 x 10 mm panel with Zn-Co-Cr-SiO2 is extremely, and surprisingly, high. This value is a mean value from several plated panels, the spread of results varying from 540 to 656 hours. Carrying out the same test on a panel of 100 x 50 mm size, coated with Zn-Co-Cr-SiO2 (Co:2.4%;Cr:0.2%;SiO2:1.9%) gave a time to corrosion result of 240 hours (spread 220 to 248 hours), which is lower, but still significantly better than the comparative known coatings.

The Zn-Co-Mo-SiO2coated panel with a coating thickness of

7 µm had a time to 5% red rust of 504 hours.

Adhesion to Electropaint Layer Four steel panels (approximately 50 x 150 mm) were plated with a coating in accordance with the invention. The coating thickness was measured using a Fischer Permascope Model M10, and determined to vary between 9 and 12 microns across the four panels. The coating on each panel had the following approximate % composition : <BR> <BR> <BR> <BR> Zn : 96. 7 <BR> <BR> <BR> Co : 1. 1 <BR> <BR> <BR> Cr : 0. 2 <BR> <BR> <BR> SiO2 2.0 Two of the panels were given a standard phosphate treatment prior to electrocoating, and the other two panels were untreated prior to electrocoating, so that the paint was applied directly to the coating layer.

A cathodic electrocoat bath was made up using a commercial coating formulation comprising an epoxy resin and a lead silicate anti-corrosion pigment paste. This was used to electrodeposit a paint coating on the coated panels.

The panel depositions used, and the results obtained, are given below.

All panels were cured for 15 minutes at effective metal temperature (EMT) of 175°C.

Panels without Phosphate Pre-treatment a) 100 ohms series resistance. Wind up to 260 V with

an initial current of 0. 43 A (max). Bath temperature 29. 5°C ; total deposition time 135 s.

Current passed : 10. 6 coulombs. Precoating film thickness : 12 m. This produced a smooth looking film with an average film build of 24 m. No pinholing defects were noted. b) 100 ohms series resistance. Wind up to 260 V with an initial current of 0. 42 A (max). Bath temperature of 29. 5°C ; total deposition time 135 seconds. Current passed : 9. 7 coulombs. Precoating film thickness : 9 m. This produced a smooth looking film with an average film build of 24 m.

No pinholing defects were noted.

Panels with Phosphate Pre-treatment c) 100 ohms series resistance. Wind up to 280 V with an initial current of 0. 36 A (max). Bath temperature of 29. SOC ; total deposition time 135 seconds. Current passed : 10. 0 coulombs. Precoating film thickness : 12 m. This produced a smooth looking film with an average film build of 22 ym.

A small area of moderate to bad pinholing was noted along one edge of the panel. Pinhole defects at 280 V are not uncommon, and defects similar to this are usually seen when depositing electrocoat over Galvannealed precoated steels. d) 100 ohms series resistance. Wind up to 260 V with an initial current of 0. 33 A (max). Bath temperature of 31. 5°C ; total deposition time 135 seconds. Current passed : 10. 4 coulombs. Precoating film thickness : 11 m. This produced a smooth

looking film with an average film build of 24 tm.

No pinholing defects were noted.

Film Adhesion Each panel was tested for film adhesion of the deposited electrocoat film using Ford Laboratory Test Method B1 106- 01 Method B (Paint Adhesion Test). Each film was subjected to a three way scribe cut (using a carbide tipped scriber) at right angles and diagonally in one direction, as shown in Figure 2. The parallel scribed lines are 3 mm apart. To the scribed area was applied 3M No. 898 adhesive tape, with firm pressure. Within 90 +/- 30 seconds of application, the tape was pulled off rapidly (not jerked) back upon itself at as close an angle of 180° to the panel surface as possible.

All four panels produced very good results. All had excellent adhesion with no removal of any paint.

Further 5 m coatings were plated using the formulation given in Table 1, under different temperatures and currents. The variables and the resulting coating compositions are given in Table 3 below. Plating Variables Composition (%) Zn Co Cr SiO "at120'45°C97.31.40.21.1 mA/cm² 50°C 95.6 2.4 0.2 1.8 5 µm 60°C 94.2 4.2 0.1 1.5 at 50°C 90 mA/cm² 96.6 1.7 0.1 1.6 5 ml20mA/cm² 95.6 2.4 0.2 1.8 150 mA/cm 96.3 1.7 0.1 1.9

Table 4 By varying the temperature at which plating is carried out, it was found possible to vary the proportions of each component in the coating. The concentration of cobalt was particularly temperature sensitive, increasing with increasing temperature. All of these coatings showed improved corrosion resistance compared to conventional coatings.

Further corrosion test results are given below in Table 5, for selected coatings from Table 4, and for other coatings obtained using double the concentrations of zinc and cobalt in the coating solution given in Table 1. Sample substrate Current Density used size/mm 90 mA/cm² 120 mA/cm² 200 mA/cm² 300 mA/cm² TA)10x10'450'600 lOOx50 218 240 (B) 100x50 92 140 211

Corrosion Test Results (ASTM B117) Table 5 (A) = coating solution of Table 1 (B) = as coating solution of Table 1, but double concentration of zinc sulphate and cobalt sulphate.

All values are for static plating. A person skilled in the art will readily be able to determine suitable plating conditions for use in a continuous plating process.

Referring now to Figure 3, the pictures on the left hand side are SEM microstructures of conventional zinc coatings (top) and the Zn-Co-Cr-SiO2 coating from Table 3 ("New Zn Coating"). The pictures in the middle are schematic microstructures from TEM examination of the coatings, and the graphs on the right hand side show EDX analysis results from the TEM examinations. The conventional zinc coating has a laminated structure consisting of Zn and ZnO. The new coating shows a similar structure but with finer Zn crystals and the enrichment of Co, Cr and SiO2 additions along the ZnO layers. It is surmised that this confers on the new coating the property of superior corrosion resistance which is observed.

Based on TEM analysis, the new coatings have a microstructure in which Zn crystallites form hexagonal

platelets of diameter 20 to 200 nm and thickness of 20 to 50 nm. They are encapsulated by a thin ZnO layer of 2 to 20 nm thickness, in which the components other than Zn are enriched. The surface morphology of the coatings shows surface roughness which is apparently caused by stacks of platelets forming peaks or promontories as discussed below.

Figure 4 shows SEM micrographs of surface structures of conventional Zn and Zn alloy coatings with and without phosphating (left hand side), and of the new Zn-Co-Cr-SiO2 coating from Table 3. Conventional Zn and Zn alloys require phosphating or similar surface treatment to enable them to receive a sufficiently coherent paint layer. The new coatings can bond directly to paint, and it is believed that this property is due to the micro-rough surface profile (represented at the top right of Figure 4) which has peaks or promontories to which the paint can key.

Further evidence for the microstructure of the new coatings in accordance with the invention is given in Figure 5. The top left TEM picture shows the breakaway of Zn crystallites from a Zn-2.4Co-0.lCr-2.1SiO2 coating under exposure of a TEM beam due to a thermal fracture along the Zn/ZnO laminae. The bottom left TEM picture shows the remaining region after the breakaway of the Zn crystallites. The corresponding EDX spectra shows the enrichment of Co, Cr, Si together with the ZnO in the remaining region.