GB2354188A - Plasma treatment of substrate for adhesive or coating - Google Patents

Abstract

A metal substrate is prepared for reception of an organic adhesive or coating by subjecting to a high density plasma, most conveniently of density in excess of 10<SP>9</SP>cm<SP>-3</SP> for at least 10 minutes and provided by an RF DC-biassed hollow cathode discharge, to remove at least half the "normal" oxide layer and also to pit the surface with micropits, and then plasma polymerising a silica-like monolayer of at least 75nm thickness on said surface. The monolayer may be derived from hexamethyldisiloxane or hexamethyldisilazane. The substrate may be an aluminium or titanium component, particularly for the aerospace industry.

Description

2354188 Substrate Surface Preparation The present invention relates to a

method of producing a surface on a metal substrate which is receptive to an organic adhesive or coating. The invention also relates to substrates having such a surface, as well as bonded substrates and coated substrates wherein the substrates have been given such receptive surfaces prior to bonding or coating as the case may be.

In the aerospace industry particularly, but without limitation thereto, it is desirable to connect metal components by use of adhesion and to protect metal components from corrosion by coating them with protective layers. There are numerous and self-evident reasons for these. Moreover, in the case of adhesion, the stronger, more reliable and more durable that a bond can be made to be, the more applications such a mode of component connection can have. Likewise, the more cohesive, adhesive and durable a coating can be arranged to be, the less maintenance it will need during the component's lifetime.

It is well appreciated that the surface of the metal component is fundamental to the 1 5 quality of the eventual bond with the adhesive or coating, and, in the aerospace industry, there are exacting standards which components must meet before they can be employed. Nevertheless, there are several technically satisfactory methods of treating a substrate. However, all these methods involve chemical treatment and the best known method involves treatment with hexavalent chrome compositions which environmentally are problematic to dispose of after use. Thus the FPL process [1] invovles etching aluminium substrates with an aqueous sodium dichromate/sulphuric acid solution. A finiher process is a the PAA process [2] in which an anodizing potential is applied to the metal while immersed in a 10% phosphoric acid electrolyte. Finally, there is the CAA process [3], which is a chromic acid anodizing process. Venables [4] appreciated that surface morphology of a substrate might play a crucial role in subsequent adhesion of a coating or adhesive and showed that an oxide layer on the surface may be shaped. Subsequently, the oxide layer may provide a failure site when, in the case of aluminium, the oxide reacts with water to form the hydroxide, which is much less stable, and leads to degradation of the bond, as well as corrosion of the aluminium. Nevertheless, the work of that paper is in the direction of chemical modification of the oxide layer to render it more resistant to 2 hydration. Venables suggests an adsorbed monolayer of an arnino phosphonic acid to inhibit hydration. Finally, crystalline A1,03 is also found to be more resistant to corrosion than the amorphous oxides formed by the FPL or PAA process.

Other monolayers are presently preferred such as hexamethyldisiloxa-ne (HA4DSO), which are deposited by a plasma using AC (50 Hz - 5 kHz), RF (13.56 Nfflz) or microwave (2.45 Ghz) discharges, or, as suggested by Eufinger et al [5], in a DC reactor.

Likewise Wang et al [6] and Yasuda et al [7] suggest plasma treatment for the purpose of cathodically depositing a surface layer of trimethylsilane (TMS), in their case also, on steel. Lin et al [8] suggest protection of copper by a similar process.

A different approach has been taken by Turner et al [9]. Not only is deposition suggested, but also preliminary plasma etching is effected, using both DC and microwave driven plasma systems. The etching uses argon or argon/hydrogen plasmas, although no difference was discerned between surfaces pretreated by grit blasting versus plasma etching.

Plasma etching is a well understood process and is employed in a number of different fields. Indeed, the fact that surface morphology of a substrate can be altered by plasma has been appreciated, and in some cases critically [10]. However, in Moon et al [11], Vaidyanathan et al [12] and Tissington et al [13], the surface pitting of polymer and carbon fibre materials under plasma etching has been recognised and at least partially credited with improved adhesion by appropriate adhesives, perhaps by providing opportunities for mechanical interlocking.

Finally, hollow cathode discharge (HCD) is also a well-known phenomenon and several papers discuss the characteristics of HCD, of which a significant aspect is a high density, high energy plasma with relatively little power input [14]. Horwitz [15] employs an HCD for reactive sputter etching, but, more significantly, Korzec et al [16] employs an HCD for surface treatment of polymer foils, to improve, inter alia, adhesion properties.

While discussing some of the impacts of the cathode design, it is appreciated that, in an oxygen plasma and with an aluminium cathode, there is a constant competition between removal and growth of alurninium. oxide at the surface of the aluminium. However, there is no discussion as to any use to which this observation may be put.

3 It is an object of the present invention to benefit from these teachings, and to provide a method of preparing metal substrates for reception of adhesive coatings. It is also desirable to produce such a method which is environmentally less damaging than known chemical processes, but which offers comparable performance.

This is achieved by an appreciation that surface morphology seems significant in providing a key for subsequent adhesion. Whether this is by mechanical interlock or other effects such as increased surface area or enhanced chemical activity at edges and points is a matter of debate. It is also appreciated that oxide layers may provide at least one source of weakness, especially in subsequent humid environments. Consequently they should therefore be minimised or modified to remove weaknesses. In practical applications oxide layers cannot be eliminated entirely. It is also seen that high density, high-energy plasmas not only etch oxide layers but also they will create micro-pits in substrate surfaces. They therefore provide advantageous surface morphology. Finally such techniques can be employed for relatively large surface areas, such as aeroplane wing surfaces which has particular benefit in the aerospace industry.

In accordance with the present invention, therefore, there is provided a method of preparing a surface of a metal substrate for subsequent application of an organic adhesive or coating, the method comprising submitting the surface to a plasma of sufficient density and energy of ions that, after treatment, the surface is pitted with micropits of depth between 5 and 150 mn, diameter of between 10 and 300 rim and distribution of between I and 10 per square micron in respect of micropits larger than 100 nni in diameter, and then plasma polymerising a silica-like monolayer of at least 75nm thickness on said surface.

"Normal thickness" of an oxide layer on an aluminiurn alloy for use in the aerospace industry, and where the alloy is stored under appropriate conditions, is usually between about 5 and I Orim. By applying the method of the invention to aluminium it is found that the oxide layer may be as much as twice as thick as that. Nevertheless, it is surprisingly found that bonding of the surface with organic polymers is improved, or at least not significantly changed in comparison with traditional methods of surface preparation.

The oxide layer is preferably at least 20 nin deep, indicating that the surface, to that depth, has been reconstructed by the plasma process. It is considered that this also is 4 instrumental in improving bonding by regularising the surface and removing cracks and sites of weakness therein.

In a different aspect, the present invention provides a method of preparing a surface of a metal substrate for subsequent application of an organic adhesive or coating, the method comprising the steps of submitting the surface of the substrate to a plasma of ion density in excess of 10' cm-', preferably in excess of 10" cm-', for a period of at least 10 minutes, and then plasma polymerising a silica-like monolayer of at least 75nm thickness on said surface.

The less dense the ion concentration, the longer the duration of the submission that is required in order to achieve reasonable results.

Said substrate may be aluminium. Alternatively, it may be titanium. It may comprise an aircraft component, such as a wing.

Preferably said plasma is produced by a hollow cathode discharge arrangement. In that event, the pressure of the plasma is between 30 and 10000 mtorr and a Ubi,,, of between -200 and - 1000 V, and the duration is at least 3 0 minutes.

Said plasma is preferably argon, hydrogen or a combination of argon and hydrogen.

However, other gases may be employed, such as trifluoromethane, oxygen, chloride and/or fluoride ions, and mixtures such as hydrogen and methane.

However, the precise combination of ion identity, pressure, bias voltage and duration is a matter of design choice, depending on the final surface morphology desired of the substrate in question.

A feature of the present invention when employing a hollow cathode is the fact that relatively high pressures are permissable, indeed preferable, to achieve high energy, high density plasmas. This could be important, particularly in the aerospace industry where it may be desired to treat large wing sections in a single chamber, and in which it is impractical to reduce pressures to those normally employed in plasma discharge apparatus.

Said surface may be mechanically pretreated, for example by grit-blasting. It is preferably also previously degreased with, for example, butanone.

Said monolayer is preferably derived from hexamethy1disiloxane (HMDSO) monomer. The monomer may be in an oxygen environment and possibly also in an acrylic acid (AcAc) environment. Alternatively, the monomer may comprise hexamethyldisilazane (EA4DZN).

In another aspect the invention provides a metal component in which the surface of the component is pitted with micropits of depth between 5 and 150 Mn, diameter of between 10 and 300 run and distribution of between I and 10 per square micron in respect of micropits larger than 100 mn in diameter, a silica-like monolayer of at least 75 run thickness covering and protecting the surface.

An oxide layer on the surface is preferably at least 10 run deep.

In yet another aspect, the invention provides components as hereinbefore defined adhered together by an organic adhesive, or coated with an organic paint or varnish.

The invention is further described hereinafter with reference to the accompanying drawings, in which:- Figures I a, b and c are: a side section through a cooled hollow cathode arrangement to give effect to the present invention; a section on the line B-B in Figure I a; and a section on the line D-C in Figure 1 a, respectively; Figures 2a and b are schematic diagrams of a) a plane parallel plate geometry cathode discharge, and b) a hollow cathode discharge, each together with a respective diagram of averaged distribution of potential between the facing plates; Figure 3 shows graphs of typical electron energy distributions for a parallel plate and hollow cathode; Figures 4a, b, c and d are electron micrographs of the surface of smooth aluminiurn samples at two different magnifications (after parallel plate plasma pretreatment), and the surface of aluminium samples after grit blasting, also at two different magnifications; 6 Figures 5a and b are electron micrographs of the surface of aluminium samples after treatment in accordance with the present invention in a hollow cathode; Figures 6a and b are electron micrographs of the surface of aluminium samples after treatment in accordance with the present invention; Figures 7a and b are electron micrographs of the surface of aluminium samples after treatment in accordance with the present invention; Figures 8a to f are electron micrographs of titanium alloy samples at different magnifications and angles of view, a and b each being untreated samples, c and d, and e and f being hollow cathode treated samples in accordance with the present invention; Figure 9 shows comparative results in wedge cleavage tests showing fracture energies for a sample in accordance with the invention and control samples; and Figure 10a and b are schematic illustrations of potential continuous processing of strip substrate.

In Figure 1, apparatus 50 comprises two cathode plates 1,2 which are interconnected, physically and electrically, by pillars 3. Plate 1 is adapted to seat on a power source (not shown). Samples 32,33 are clamped to plate 2 by clamping straps 27A,B and 29. They are in electrical contact therewith. Plate 2 is a composite plate defining a cooling chamber 7 closed by a cover plate 5 sealed to the plate 2 by O-ring 6. Pipes 35,36 communicate with chamber 7 and convey cooling water or other suitable medium through the chamber. Bars 9,10 define a serpentine flow path to avoid hot spots in the plate 2. However, while cooling of the substrate is desirable in many applications, for example to prevent annealing of some materials, the substrate could be mounted on uncooled plate 1. Surrounding, but separated from, plates 2,5, is a grounded anode plate or shield 12,13.

The whole arrangement is disposed in a vacuum chamber capable of reducing pressure to at least 50 mtorrs, although this is not a particularly challenging pressure reduction and is capable of being achieved in large volume chambers with mechanical pumps. Not shown in the drawings is a gas input to introduce a gas into the chamber which is to be ionised and rendered a plasma by the apparatus 50.

7 Indeed, apparatus 50 is supported on pivotal arm 34 through pivot 25, so that the cathode can be manoeuvred inside the vacuum chamber. Rod 34 is capable of being raised and lowered from outside the chamber so that plate I may be pressed against a cathode power source. Of course, while this arrangement is suitable for experimental purposes, it 5 is probably not appropriate for industrial applications. However, that is a matter of scaling UP- In Figure 10a, a potential industrial application for treating strip material 78 is schematically illustrated, comprising wind-on and -off drums 76,75, and vacuum chamber 72. Exits 74 are differentially pumped to maintain the vacuum in chamber 72. A cooled grounded anode "anvil" 73 supports the strip 78, a hollow cathode 71 facing the exposed surface of the strip. A gas supply 77 maintains an appropriate supply of gas.

As will become clear below, the present invention involves an essentially two stage process of pretreatment and monolayer deposition. A potentially suitable arrangement is shown in Figure 10b in which the strip 88 passes through two chambers 82ab, each with its own hollow cathode 81a,b and anvil 83ab. The first chamber 82a is dedicated for pretreatment, while the second is for monolayer deposition. The conditions operating in each chamber is adjusted so that the rate of draw through of the strip 88 is appropriate for each stage of treatment.

Figure 2a shows a traditional cathode discharge arrangement in which a cathode plate is disposed directly opposite, and parallel to, a grounded anode plate 62.)When the cathode is DC biassed to say 500 V and subjected to a radio frequency fluctuation, electrons are emitted from the plate and travel to the anode 62. In colliding with atoms/molecules on the way, a plasma of ions is created having a specific density of ions dependent on a number of factors including the energy distribution of the electrons, the density of the gas, pressure etc. The distribution potential of a parallel plate arrangement is obviously unbalanced as shown.

Figure 2b shows a hollow cathode arrangement where two cathode plates 64a, b face one another and an anode plate 66 is disposed at one side. With the same potential applied, firstly there is an even distribution of potential between the cathode plates, and secondly, just as many electrons leave the cathode plate. But instead of simply immediately striking 8 a facing anode and being absorbed therein, electrons are more likely to be reflected off each cathode and bounce back and forth numerous times before finally reaching the anode. Thus a much greater concentration of electrons gathers in the space between the cathodes 64a,b and consequently many more atoms/molecules are ionised and a far denser plasma, 5 in terms of the concentration of ions, results.

Indeed, as shown in Figure 3, not only are there more electrons, but also their energy distribution is different.

Turning to Figures 4 to 8, these show scanning electron microscope (SEM) micrographs of various sample surfaces. In Figures 4a and b two magnifications show the substantially smooth and featureless surface of an aluminium substrate treated in a parallel plate cathode chamber for 41 minutes at a voltage bias of -800V in an Argon and Hydrogen atmosphere of 80 mtorT pressure. The striations are present in the sample prior to any treatment. In Figure 4c, the severely altered surface of a grit-blasted sample is shown, which surface is broken and jagged with a roughness up to 5 microns.

By comparison, Figures 5, 6 and 7 show samples having been subject to impaction by plasma ions in a hollow cathode apparatus as described above with reference to Figure 1. Here, the samples are covered with micropits of diameters varying from about 10 mn to about 300 rim. Depths are about 5 nm to 150 nm. The surface striations mentioned above are still eveident and unaffected. Although in some samples the pits seem to form randomly and fairly evenly over the substrate surface, in Figure 7a,b, it is clear that not only can they follow lines of weakness or crystal boundaries of the sample, but also pits can overlap. It is not the purpose of this document, however, to explain their distribution. On the other hand, since the normal oxide layer of these samples is about 10 m-n usually and following pretreatment it is found to have grown to a depth of about 20 nm, it is speculated that perhaps the entire surface to that sort of depth is removed and re-deposited by the plasma during treatment. Along with the micropitting, this may explain the good adhesion results explained below in that a surface with greater integrity and cohesion is created.

The conditions of such pretreatment is given for various samples in the table below, but generally the conditions are similar to that given in parallel plate mode, the essential 9 difference being the use of the hollow cathode.

Finally, having been subjected to plasma etching in accordance with the present invention, a silica-like monolayer coating is applied. Again the conditions are summarised in the table below and the essential difference between the samples is the use or otherwise of a hollow cathode.

Once treated, samples were put to test by adhering them together using a proprietary adhesive (FM 73 adhesive, which is an aerospace epoxy film adhesive suitable for structural applications and which is cured at 121'C). Later, wedges are driven between the samples and the energy required to develop cracks between them at fracture, over a period of time, is monitored. Control samples for comparison purposes were also prepared using traditional techniques, as explained further below. The results are shown in Figure 9.

There are three control samples. The first was subjected to the CAA process described above (ie chromic-sulphuric acid etching according to Bae Airbus specification

8-2297 followed by anodising in chromic acid and 40-50 V acid at 20'C, for 15 minutes.

Secondly, there was a DG+GB sample, which was simply degreased in butanone and grit blasted using 50 im alumina grit. The third is a silane treated sample.

Samples G59p and G60p used only parallel plate treatment. The fracture energies for these samples are considerably less than for the traditional CAA process sample, although better than the DG+GB and silane treated samples. Nevertheless, there is no indication thatsuch a process may be satisfactory for, for example, aerospace applications.

Sample G64p however showed remarkable results in the cleavage tests. Some of that remarkable result will have been due to some annealing of the sample during pretreatment which would result in plastic deformation of the sample giving excessive results.

Nevertheless, it is believed that subjecting a substrate to hollow cathode plasmas, or at least plasmas of sufficient density and energy such as provided by a hollow cathode, poses a credible alternative to chemical treatment, while still promising satisfactory bonding characteristics to the substrate surface..

Since the method according to the present invention avoids the use of toxic chemicals it will meet demands for a reduction in the use of such chemicals in aerospace and other fabrication industries. Indeed, application of the invention is not limited to aluminium but probably equally has effect with other metals such as titanium, steel or copper.

Indeed, Figures 8 a to f are SEM micrographs of titanium alloy (Ti90/Al6N4) samples. Figures 8a and b are unprocessed showing the jagged surface, even at a micron level. On the other hand Figures c to f show two samples G 111 and G 117 subjected to plasma pretreatment in a hollow cathode in which the surface is still very rough at a micron level but is much more globular and lobular and suggests reprocessing of the surface by continuous erosion and deposition of the surface layer. The lobes seen on the surface probably consist of titanium nitride (G1 I I) or oxide (G117), both of which are generally tougher materials than the original alloy.

Returning to the samples referred to in Figure 9, each comprised (as did the controls) a strip of aluminium 2024-T3 Alclad alloy of thickness 3.2 mm, width 25 mm and length 150 mm. The samples were subjected to the pretreatments and plasma polymerisation as shown in the table below, which also gives details of treatments of samples shown in the micrographs in Figures 5 to 8.

I 11 No exp. Deposition Pretreatment Deposition Coating substrate thickness nm G59 2024-T3 Alclad Parallel 12late mode. Parallel Dlate mode. 195 Ar - 13 min, p=80 mtorr, HMDSO+O,, 16 min, P=190 W. pF,= 32.1 rntorr, Ubj. =-(703-732) V p(HMDSO)=2.1 rntorr, P=12 W, ullj' =-190 V G60 2024-T3 Alclad Parallel plate mode. Parallel Rlate mode. 202 Ar - 13 min, p=80 mtorr, HMDSO+02, 18 min, P=190 W, pF,= 32.4 mtorr, Ubi. =-(710-724) V p(HMDSO)=2.4 rntorr, P=12 W, Ubj' =-190 V G64 2024-T3 Alclad Hollow cathode mode. Hollow cathode mode -(130-150) Gas flow (sccrn): Ar-7.8, H2-25; HMDSO+02, 8 Min, P=150 W, pF,= 32.8 mtorr, a) p(H,+Ar)=30 rntorr, 10 min, p(HMDSO)=2.8 rntorr, Ubj. =-(231-254) V; P=12 W, b) P(H2+Ar)=50 rntorr, 29 min, Ubi. =-159 V Ubi. =-(237-228) V; c) p(Ar)=160 mtorr, 13 niin U,_ =-(210-213) V G68 2024-T3 Alclad Hollow cathode mode. No deposition stage Ar+ 0 2 - 61.5 rnin, p=50 rntorr, Gas flow (sccrn): Ar16.7, 02-5; P=350 W, U,i_ (312-268) V G82 2024-T3 Alclad Hollow cathode mode. No deposition stage Ar+H2 - 47 min, p=900 ratorr, Gas flow (sccrn): Ar-14, H2-30; P=460W, Ubj, (162_205) V GIII Ti90/AI6/V4 Hollow cathode mode. Hollow cathode mode.

N2 +H2 - 61 min, HMDSO+O,, 10 min, Gas flow (sccrn): N2-10, H2-30; p(HUDSO)=3.3 rntorr, P = 150 Mtorr; pZ= 33.3 mtorr, P=300 W, P=12 W, Ubj, =-(474-592) V Ubj, =-(105-92) V GI 17 Ti90/Al6N4 Hollow cathode mode. Hollow cathode mode.

Ar + H,. - 61 rnin; HMDSO+02, 10 Min, Gas flow (sccm): Ar-41.7, H,-10; p(HMDSO)=2.7 rntorr, p = 150 rntorr; pF,= 32.7 mtorr, P=30OW, P=12 W, Ubj. =-(401-540) V Ubi,, (101-82) V 12 References [I] H W Eichner and W E S chowalter, Forest Products Laboratory, Madison, WI, US, Report No 1813 (1950) [2] G S Kabayashis and D J Donnelly, Boeing Co, Seatlle, WA, US, Report No DG 41517 (Feb 1974) [3] Fokker-VFW, Amsterdam, NL, Process Specification TH 6.785 (Aug 1978) [4] J D Venables, Adhesion and durability of metal-polymer bonds, JnI of Materials Science 19 (1984) 2431-2453 [5] S Eufinger, W J van Ooij and K D Conners, DC-Plasma Polyrnerisation of HDMS Part II Surface and Interface Characterisation of Films Deposited on Stainless-steel Substrates, Surface and Interface Analysis, Vol 24, 841- 855 (1996) [6] Tinghao F Wang, T J Lin, D J Yang, J A Antonelli and H K Yasuda, Corrosion protection of cold-rolled steel by low temperature plasma interface engineering I Enhancement of E-coat adhesion, Progress in Organic Coatings 28 (1996) 291-297 [7] H K Yasuda, Tinghao F Wang, D L Cho, T J Lin, and J A Antonelli, Corrosion protection of cold-rolled steel by low temperature plasma interface engineering II Effects of oxides on corrosion performance of E-coated steels, Progress in Organic Coatings 30 (1997) 31-38 [8] Y Lin and H Yasuda, Effect of Plasma Polymer Deposition Methods on Copper Corrosion Protection, JnI of Applied Polymer Science, Vol 60, 543-555 (1996) [9] R. H Turner, I Segall, F J Boerio and G D Davis, Effect of Plasma- Polymerised Primers on the Durbility of Aluminium/Epoxy Adhesive Bonds, Jnl of Adhesion, Vol 62, 1-21(1997) [10] D M Manos and D L Flamm, Plasma Etching, An Introduction, Academic press,

Inc, 1989 (ISBN 0-12-469370-9) [I I] S I Moon and J S Jang, Interfacial adhesion improvement of oxygen plasma treated UHMPE fiber/vinylester composites using different plasma output power, Korea Polymer Jnl, Vol 5, No 1, 26-32 (1997) [12] N P Vaidyanathan, V N Kabadi, R Vaidyanathan and R L Sadler, Surface- treatment of carbon-fibres using low-temperature plasma, Jn1 of Adhesion, Vol 48, No 1-4, 1-24,(1995) [ 13] B Tissington, G Pollard and I M Ward, A study of the effects of oxygen plasma treatment on the adhesion behaviour of polyethylene fibres, Composites Science and tyechnology, Vol 44, No 3, 185-195 (1992) [ 14] M E Pillow, A critical review of spectral and related physical properties of the hollow cathode discharge, xxxxxchimica Acta, Vol 36B, No 8 821-843 (198 1) [ 15] C M Horwitz, Hollow cathode reactive sputter etching - A new high- rate process, Applied Physics Letters, Vol 43, No 10 977-979, (1983) [16] D Korzec, M Schott and J Engemann, Radio frequency hollow cathode discharge for large-are double-sided foil processing, Jn1 Vacuum Science and Technology, Vol A 13, No 3 843-848 (1995) 14

Claims (19)

Claims

1. A method of preparing a surface of a metal substrate for subsequent application of an organic adhesive or coating, the method comprising submitting the surface to a plasma of sufficient density and energy of ions that, after treatment, the surface is pitted with micropits of average depth of more than 2 nm, average diameter of between 10 and 200 nm and distribution of between 0. 1 and 0. 2 per square micron, and then plasma polymerising a silica-like monolayer of at least 75nm thickness on said surface.

2. A method of preparing a surface of a metal substrate for subsequent application of an organic adhesive or coating, the method comprising the steps of submitting the surface of the substrate to a plasma of ion density in excess of 10' cm-', preferably in excess of 10" cm-', for a period of at least 10 minutes, and then plasma polymerising a silica-like monolayer of at least 75nni thickness on said surface.

3. A method as claimed in claim I or 2, in which said substrate is aluminium or titanium.

4. A method as claimed in claim 3, in which said substrate comprises an aircraft component.

5. A method as claimed in any preceding claim, in which said plasma is produced by a hollow cathode discharge arrangement.

6. A method as claimed in claim 5, in which the pressure of the plasma is between 5 0 and 100 mtorr.

7. A method as claimed in claim 5 or 6, in which the voltage bias of the cathode (Ujj.) is between -200 and -1000 V.

8. A method as claimed in claim 5, 6 or 7, in which the duration of said submission is at least 30 minutes.

9. A method as claimed in any preceding claim, in which said plasma is argon, hydrogen or a combination of argon and hydrogen.

is

10. A method as claimed in any preceding claim, in which said surface is mechanically pretreated, for example by grit-blasting, before submission to said plasma

11. A method as claimed in any preceding claim, in which said surface is degreased with, for example, butanone before submission to said plasma.

12. A method as claimed in any preceding claim, in which said monolayer is derived from hexamethyldisiloxane (BlvMSO) or hexamethyldisilazane (BNDZN) monomer.

13. A method as claimed in claim 12, in which the monomer is in an oxygen environment and prefereably, also in an acrylic acid (AcAc) environment.

14. A method as claimed in claim 1, or in any of claims 3 to 13 when dependent on claim 1, in which said micropits have an average depth of more than 10 tun and preferably many have an average depth more than 20 nm.

15. A method as claimed in claim 14, in which an oxide layer on the surface of the substrate prior to said plasma polymerisation, is as least as deep as most pits in the surface, preferably at least 10 nin thick.

16. A metal component having an oxide layer at its surface which is less than 50% of the normal thickness (as hereiribefore defined) and in which the surface of the component is pitted with micropits of average depth of more than 2 nm, average diameter of between 10 and 200 nm and distribution of between 0.1 and 0.2 per square micron, a silica-like monolayer of at least 75nm thickness covering and protecting said oxide layer.

17. A metal component as claimed in claim 16, made by a method as claimed in any of claims 1 to 15.

18. A metal component as claimed in claim 16 or 17, ftirther comprising an organic paint or varnish applied to said surface.

19. An assembly of two or more components as claimed in claim 16 or 17, adhered together by an organic adhesive.