- Email: [email protected]
Surface treatment of titanium for adhesive bonding to polymer composites: a review
International Journal of Adhesion & Adhesives 21 (2001) 129}136
Surface treatment of titanium for adhesive bonding to polymer composites: a review P. Molitor*, V. Barron, T. Young Department of Mechanical and Aeronautical Engineering, University of Limerick, Limerick, Ireland Accepted 3 July 2000
Abstract At present, the bonding of polymer composites to titanium is a problem, which has not been fully resolved. Previous research has shown that bond strengths can be signi"cantly improved by surface treating the adherends prior to bonding. However many of the successful surface treatments involve the use of hazardous chemicals, which have to be phased out as part of an EU directive, which paves the way for less toxic environmentally friendly methods. In this paper various methods of surface treatment including traditional treatments such as acid etch, anodisation, novel plasma spray and laser treatments for both polymer composites and titanium will be discussed. These treatments will be reviewed with respect to changes in surface tension, surface roughness, surface chemistry and how these changes a!ect bond strength and durability of polymer composite titanium adhesive joints. 2001 Elsevier Science Ltd. All rights reserved. Keywords: A. Primers; B. Composites; C. Surface treatment; Adhesive bonding; Titanium
1. Introduction In order to produce a strong and durable adhesive joint between di!erent substrates, surface treatment is necessary. Certain bonding techniques provide adequate static strength, but have little durability when exposed to hot moist environments, while others are susceptible to debonding in the presence of fuels, oils and cleaning solvents commonly encountered in aircraft applications. In addition, the nature of the surface treatment prior to bonding is found to be a major in#uence in the control of this e!ect [1]. Surface treatment increases the bond strength by altering the substrate surface in a number of ways including increasing surface tension, increasing surface roughness or changing surface chemistry. By increasing surface roughness, an increase in surface area occurs which allows the adhesive to #ow in and around the irregularities on the surface to form a mechanical bond. Changing surface chemistry may result in the formation of a chemical bond e.g. between the polymer molecules in
* Corresponding author. Tel.: #353-61-202531; fax: #353-61202944. E-mail address: [email protected] (P. Molitor). Current address: Physics Department, Trinity College Dublin, Dublin 2, Ireland.
the polymer matrix composite and the metal oxide layer on the other adherend surface layer [2}5]. As a result, the nature of the surface will also in#uence the stability of the joint. When exposed to hot/humid environmental conditions, a polymeric adhesive/polymer interface is much more stable than the equivalent polymeric adhesive/metal interface. A well-chosen polymeric adhesive/polymer interface is unlikely to fail because of the environmental-induced stress due to the nature of the bond formed. On the other hand, the durability of a polymeric adhesive/metal joint is not as stable. Early studies in the 1960s revealed that these joints did not perform well in hot/wet conditions with frequent occurrence of short-term interfacial bond failure.
2. Adherend materials 2.1. Polymer composites Both thermoset and thermoplastic composite materials have been bonded to titanium with varying degrees of success [6]. However, it is noted that thermoplastic materials are inherently more di$cult to bond [7]. Carbon-"bre-reinforced polymer composite material has been used widely as the adherend in the literature
0143-7496/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 3 - 7 4 9 6 ( 0 0 ) 0 0 0 4 4 - 0
130
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
reviewed [8]. Although glass-"bre-reinforced composite material has a much lower modulus than carbon "bre, bond integrity tends to be dominated by the characteristics of the matrix, rather than the "bre and for this reason the research conducted on carbon "bre composite is also relevant to studies involving glass "bre. Thermoset polyester or epoxy composites can have a resin rich surface layer, which can cause particular problems in bonding. The reason for this is that either most of them contain a gel coat at the surface, or if the latter has not been deliberately created, the surface layer usually contains a higher proportion of resin as opposed to the interior. Both the gel coat and the resin-rich surface layer are very brittle and can subsequently fail catastrophically when overloaded. However, proper design and correct adhesive choice may compensate for this weakness. Highly compliant adhesives are a particularly good choice, as they spread the applied load over a large area and hence reduce the stress borne by the surface of the composite. 2.2. Titanium Previous researchers [8}16] have used various di!erent titanium alloys, however Ti}6Al}4V is the most widely used in the aerospace industry [17].
3. Surface treatments of polymer composites A variety of surface treatments have been used with various degrees of success to increase surface tension, increase surface roughness, change surface chemistry and thereby increase bond strength and durability of polymer composite adhesive joints and are shown in Table 1. Wetting of the adhesive by the adhesive is critical for the formation of secondary bonds. For complete wetting the surface energy of the adhesive must be lower than the surface energy of the adhesive [18]. For thermoplastic composites the primary aim of the surface treatment is to increase the surface energy of the adherend as much as possible. Surface treatments decrease water contact angle, increase surface tension and as a result increase bond strength. 3.1. Abrasion/solvent cleaning Abrasion/solvent cleaning may be employed to degrease the surface and remove mould release agents from the adherend [19}21]. The composite sheets can be lightly abraded using 180/220-mesh alumina, then wiped clean with a solvent such as methylethylketone (MEK) and allowed to dry. Previous researchers have found a signi"cant increase in surface roughness and bond strength for thermosets. However, the same treatment for thermoplastic polymer composites did not reveal any
signi"cant increase in bond strength or durability, due to the fact that some of these composites have very smooth low surface energy surfaces [19,21,22]. 3.2. Grit blasting Alternatively, an alumina grit blast (of particle size 45 m) with three passes at a distance of 15}20 cm may be employed, followed by solvent rinse and dried in nitrogen. MEK may not be compatible with local government regulations for industrial applications and as a result may have to be phased out and be replaced by a less hazardous solvent. The DLR (German Aerospace Research Establishment) has investigated three di!erent types of cleaning agents to be used with mechanical methods of surface treatment [23]. These included acetone, frigen and ethylacetate. Of these, the latter gave the highest lap shear values. Again, as with the abrasion and solvent wipe, grit blasting gave strong and durable bond strength for thermosets but revealed very little degree of success with thermoplastic materials. 3.3. Peel-ply Peel-ply, an impregnated ply is removed immediately prior to bonding. Previous research has shown that successful joints were obtained by increasing surface roughness, degreasing and removing of mould release agents [19]. The peel-ply is released from the surface because of the non-stick nature of the substrate within which it is impregnated. These are frequently #uorine silicon or #uorine compounds, which contaminate the laminate surface. Consequently, this implies that further treatment of the surface, e.g. by abrasion, is required if the strength of the subsequent bond is not to be seriously impaired [24]. If the ply contains a release agent, the adherend must be subsequently cleaned with a solvent and dried in a stream of nitrogen. 3.4. Tear-ply Tear-ply is mainly used for thermoset composites as thermosets are reactive upon heating and hence do not require a chemical surface treatment. It is fundamentally di!erent from peel-ply and consists of a fabric, which is completely impregnated by the laminate matrix resin, and as the name implies subsequently peeled o! the moulding. Super"cially, the surfaces of the laminates treated with either peel-ply or tear-ply appear very similar as both exhibit the pattern of the ply fabric. Tearply exploits a laminate property, which is generally regarded as being a serious disadvantage: the lack of strength in the z direction (normal to the surface). Immediately prior to a bonding operation, a corner of the fabric is raised with a knife and the tear-ply is then
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
131
Table 1 A!ect of surface treatment on polymer composites Treatment type Material
Nature of treatment
Abrasion and solvent wipe Grit blasting
Remove mould release Remove mould release Etch
Peel-ply
Thermoset and thermoplastic Thermoset and thermoplastic Thermoset and thermoplastic Thermoset
Tear-ply
Thermoset
Corona discharge Plasma treatment Flame treatment Laser treatment
Thermoplastic
Acid etch
Thermoplastic
Remove mould release Remove mould release Oxidising
Thermoplastic
Ablation and/ or oxidation Oxidising
Thermoset and thermoplastic
Ablation and/ or oxidation
Surface tension
Surface roughness
Surface chemistry
Bond strength
Durability
Ref.
Increase found for thermosets Increase found for thermosets Slight increase
Good for thermosets Good for thermosets Poor
[22,21]
Y
Increase
Good
[21]
Y
Increase
Good
[19]
Y
Double
[28]
Y
Increase
Y
Increase
Good (90 days) Good (90 days) * More research is necessary
[21]
Y Y Y
Y
Y Y
Y
Y Y
Increase
[23,28,21] [12,21,27]
[19,22,27,31] [8,21]
Depends on polymer matrix material.
torn o! the moulding, thus fracturing the resin interface between the tear-ply and the bulk moulding. The tear-ply must be carefully selected to enable it to be removed without any di$culty following the moulding process. Polyester resin has fracture characteristics, which permit this, and have little adhesion to the nylon "bres in the fabric. However, some resins are not compatible with this technique and hence peel-ply must be employed despite the di$culties arising from contamination. 3.5. Acid etching Acid etching has produced similar results to abrasion and grit blasting, in that an increase in bond strength is recorded for thermoset polymer composites, whereas little or no e!ect was recorded for thermoplastics [19,22,24].
surface chemistry by oxidising the polymer matrix, which results in the increase in bond strengths. 3.7. Plasma treatment The use of plasma treatment to treat polymers has been known for more than 20 years [29]. The plasma treatment involves a low-pressure plasma gas, which is electrically conductive and consists of excited atoms, ions and free radicals [30]. This allows polymer surfaces to be cleaned, etched or chemically modi"ed [19,30]. The plasma particles react not only with each other but also with the surfaces, which are exposed to the gas, giving rise to the following e!ects [19]:
3.6. Corona discharge treatment
E E E E E
Corona discharge, namely exposing the substrate surface to excited atoms, ions and free radicals at atmospheric pressure has been widely used to treat plastic surfaces for adhesive bonding [25}27]. The success of corona discharge in treating carbon-"bre}epoxide, carbon-"bre}PEEK and glass-"bre}polypropylene composites for bonding is reported, as is its use for treating polyole"n "lms to make them receptive to printing inks. Results [28] reveal that corona treatment increases surface tension and in some cases alters the
The combined e!ect of these processes results in an improvement of the adhesion properties of the surface. Plasma treatments involving various gases have been found to enhance the surface tension, oxidise the polymer matrix and increase the bond strength of PEEK composites [31]. Results obtained by Blackman [22] and Barron [7,32] echo the results obtained by corona treatment [28], in that plasma treating the PEEK composite produced an increase in surface roughness, surface tension and bond strength.
surface cleaning, degradation of the polymer chains, removal of material from the surface, formation of radicals on the surface, change of tacticity of the polymer chains.
132
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
3.8. Flame treatment Flame treatment is widely used in the surface modi"cation of polyole"ns, to improve printability and paintability, by introducing oxygen containing functional groups to the surface [25,33}36]. An oxidising #ame is very similar to a gas plasma in that it contains excited species such as atoms, ions and free radicals, which oxidise the surface of the specimens. The distance from the surface to be treated, the air/gas ratio and the dwell time are all critical parameters in producing a successful treatment. By oxidising the surface prior to bonding an increase in hydrophylicity and hence bond strength can occur. 3.9. Laser treatment Recently a novel method of surface treatment namely laser treatment [21,37,38] has been employed to increase surface roughness, surface tension and hence increase the bond strength of adhesive joints. Initial results seem promising however with respect to durability but more research is necessary. Park [39,40] has recently shown that laser treating polymer composite surfaces prior to bonding can produce high-strength adhesive bonds. Laser treatment results in increase in surface roughness due to the ablation of the polymer matrix.
#uorosilic acid under high pressure. A post treatment rinse in 5% nitric acid is required for Ti}6Al}4V titanium alloys to remove grey smut after blasting. An oxide "lm is produced which is crystalline in nature. This results in joint durability, which is reported to be slightly lower than TURCO 5578 and slightly higher than the phosphate #uoride process [12,15]. Most chemical treatments alter the titanium substrate by etching the existing oxide "lm. If the solvents are reducing agents, then a fresh thin oxide "lm is produced after the metal is removed from the solution, while in an oxidising solvent a thicker oxide "lm is produced [15]. 4.3. Acid etchants A number of acid mixtures have been used to treat the surface of titanium, including nitric-hydro#uoric acid, hydrochloric-orthophosphoric acid, hot sulphuric acid and sodium dichromate}chromic acid. Most of these treatments give adequate dry strengths but durability is reported to be poor [15]. Both stressed and unstressed peel and lap shear joints have shown to perform equal to or better than phosphate #uoride or modi"ed phosphate #uoride processes [12]. An additional problem associated with these treatments is hydrogen pick up. However, this problem can be avoided with the use of alkaline solutions as opposed to acid etchants [12].
4. Surface treatment of titanium 4.4. Alkaline peroxide etch Mechanical, chemical, electrochemical and energetic surface treatments are used to enhance the surface of titanium alloys prior to bonding. Durability studies of Ti}6Al}4V reveal that surface preparations that produce no roughness (macro or micro) yield the poorest bond durability. Those that produce signi"cant macro-roughness but little micro-roughness yield moderate to good durability. Finally those that produce signi"cant microroughness yield the best durability (Table 2). 4.1. Abrasion and grit blasting A mechanical treatment is used primarily to produce a clean macroscopically rough surface and to remove some of the existing oxide layer. The combination of a clean adherend surface with signi"cant macro-roughness improves the initial dry strength [15]. However, abrasion and grit blasting techniques as discussed in the previous section alone are not adequate methods of surface treatment, but when combined with chemical or electrochemical treatment, durable bond strengths can be obtained [19]. 4.2. VAST (vought abrasive surface treatment) In this treatment, the titanium is blasted with a slurry of "ne abrasive alumina (220 mesh) containing 2%
Initial alkaline peroxide treatments were carried out at room temperature, but this process, which lasted up to 36 h, was too lengthy for industrial applications. The immersion time can be reduced to 20 min by increasing the temperature to 50}703C [15]. Depending upon the concentration of sodium hydroxide and hydrogen peroxide, the metal is either etched or oxidised. Those concentrations, which produce grey oxides, have been found to produce adhesive wettable surfaces. An oxide layer up to 2 m thick, which is stable up to 2003C and capable of forming high bond strengths at elevated temperatures and high-humidity environments were produced using a sodium hydrogen peroxide etch [24]. Alkaline peroxide etch is reported to be satisfactory but has long heat up times and has high hydrogen peroxide consumption. Oxide formation is related to the rate of hydrogen peroxide decomposition. The rate of decomposition is increased by adding 5}10% heavy metal ions to the hydrogen peroxide solution. 4.5. Phosphate yuoride process The phosphate-#uoride-based treatments are based on combinations of trisodium phosphate, disodium tetraborate, potassium #uoride and hydro#uoric acid [41]. A process of pre-etching in 3% hydro#uoric acid and
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
133
Table 2 A!ect of surface treatment on titanium alloys Treatment type
Alloy
Nature of treatment
Surface roughness
Oxide layer (nm)
Bond strength
Durability
Ref.
Abrasion and solvent wipe Grit blasting
Ti}6Al}4V
Macro
*
Poor
Poor
[15]
Macro
*
Increase
Adequate
[12,15]
VAST
Ti}6Al}4V
Macro
*
Good
Poor
[12,15]
Acid etch Alkaline etch Phosphate}#uoride Modi"ed Phosphate}Fluoride
Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V
Remove mould release Remove mould release Remove mould release Etch Etch Etch Etchant and oxidation
Micro Micro None None
* 60}200 20 8
Adequate poor Good Adequate Adequate
[15,45] [15,24,45] [15,45] [15,45]
TURCO DAPCOtreat Pasajell Chromic acid anodisation NaOH anodisation Cathodically deposited Al O Plasma spray
Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V
Oxidising Oxidising Oxidising
Macro Macro Macro Micro
17.5 6 10}20 40}140
Adequate Increase Adequate High
Poor Good Poor Better than phosphate} #uoride Adequate Good Adequate Excellent
Ti}6Al}4V *
Oxidising Oxidising
Micro *
80}90 *
High Adequate
Excellent Adequate
[45] [15,45]
Ti}6Al}4V
Micro
130
High
Excellent
[15,45]
Sol gel
Ti}6Al}4V
*
*
High
Good
[9]
Laser treatment
Ti}6Al}4V
Ablation and oxidation Coupling and oxidation Ablation and oxidation
Macro
*
High
Poor
[16,49]
Ti}Al}4V
15% nitric acid followed by a 2 min dip in the 5% trisodium phosphate, 2% potassium #uoride and 2.6% hydro#uoric acid mixture has become known as the phosphate #uoride process (US Patent 2 864 732). The rinsing procedure after this treatment is critical to remove excess chemicals and a 15 min soak in de-ionised water at 603C has been recommended [15]. 4.6. Modixed phosphate yuoride process It has been shown that on exposure to warm moist environments, the anatase oxide layer produced by the phosphate, #uoride process slowly reverts to rutile [15]. As a result, there is a decrease in volume of about 8%, which results in the development of stresses at the adhesive/oxide layer interface. By stabilising the anatase structure the joint durability is enhanced. This is achieved by adding 0.75% anhydrous sodium sulphate into the etch solution [12,15]. However, both of these treatments have been out-performed by treatments such as TURCO 5578, sodium hydroxide anodisation and chromic acid anodisation [15]. 4.7. TURCO 5578 TURCO 5578 is an alkaline-based etchant, which contains caustic soda, sodium metasilicate and pyro-
[15,45] [45] [15,45] [45]
phosphate. This treatment produces a large amount of macroroughness with little or no microroughness. The oxide "lm produced is &17.5 nm thick with a macroroughness of 3.4 m peak to valley [42]. The durability of this treatment is much better than that of the phosphate #uoride treatments. In durability trials, TURCO 5578 is only out-performed by chromic acid anodisation [12]. However, the added advantage of the TURCO 5578 treatment is that there is no hydrogen embrittlement as observed in the acid etching process. 4.8. DAPCOtreat DAPCOtreat is similar to TURCO 5578 in that it results in an increase in macroroughness with little or no microroughness. DAPCOtreat produces a thin 6 nm oxide layer on the titanium substrate [12]. 4.9. Pasajell 107 treatment Pasajell 107 is recommended as a prebond treatment for titanium [5,12,14,15,43]. The chemical composition of pasajell 107 is 40% nitric acid, 10% combined #uorides, 10% chromic acid, 1% couplers and the balance is water. It is available as a thixotropic paste for brush applications or as an immersion solution for tank treatment. A recommended treatment time is 10}15 min.
134
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
Pasajell 107 is applied in combination with a pre-treatment of MEK wipe, blast with 320 grit non-silicone sand, followed by non-chlorinated solvent wipe [14]. After applying the pasajell for 10}15 min, the coated area is rinsed with deionised water. The process produces an amorphous looking oxide, which has anatase structure, which is stable up to 1753C and converts to rutile at 3503C. Comparing this treatment to the alkaline peroxide etch, the bond durability of the pasajell treatment is reported to greatly exceed that of the alkaline peroxide etch at elevated temperatures and humid environments [14]. 4.10. Electrochemical reactions Electrochemical reactions have been used for cleaning, etching and oxidising metal surfaces, where the metal acts as an anode or cathode. Anodisation creates an oxide surface that is dependent upon the electrolyte, anodisation voltage, time and temperature [5]. There are various anodisation techniques employed for titanium alloy surface treatment including chromic acid anodisation and sodium hydroxide anodisation. 4.11. Chromic acid anodisation Chromic acid anodisation produces a surface with signi"cant microroughness and an oxide thickness of 40 and 80 nm for the 5 and 10 V treatments, respectively [12]. In addition, chromic acid anodisation oxides are reported to exhibit remarkable bond durability and provide the target bond strengths and durability for all other treatments. 4.12. Sodium hydroxide anodisation Previous researchers investigated the use of sodium hydroxide}hydrogen peroxide anodisation and peroxide free sodium hydroxide anodisation as methods of surface treating titanium for adhesive bonding [11]. It was found that specimens prepared by these methods produced high-strength adhesive bonds and exhibited good durability when exposed to conditions of heat, moisture and stress. The presence of hydrogen peroxide was found to be neither necessary nor bene"cial and in some cases resulted in a decrease in bond strength. In addition, the sodium hydroxide anodising solution could be reused and was considered to present very low operational and environmental hazards. 4.13. Cathodically deposited aluminium oxide Cathodic depositions of metal oxides from alcohol solutions containing inorganic nitrates have shown good wettability and environmental resistance in hot and humid conditions. A recommended solution is based on
10 g of hydrated aluminium nitrate dissolved in 1 l of isopropyl alcohol [15]. By making titanium, the cathode of an electrolytic cell at 30 V the oxide is deposited on the surface. Results have shown joint strengths and durability to be better than the VAST and TURCO 5578 processes [15]. 4.14. Plasma treatment Both glow discharges and plasma sprays have been examined as methods of surface treatment for titanium alloys. Aronsson [44] investigated the use of glow discharge. Results obtained revealed that this method of plasma treatment produced clean surfaces and reproducible results, depending upon the degree of treatment. Furthermore, oxidation in pure oxygen resulted in uniform and stoichiometric TiO surface oxide layers with reproducible composition and thickness. Plasma spraying involves rapidly heating either, TiO TiSi , MgO or SiO powder to a molten or semi-molten state and then spraying it onto the substrate at high velocity. This is another method of surface treatment employed for adhesive bonding that involves no hazardous chemicals or pollutants. Other advantages of plasma spraying over chemical treatments include [13]: E E E E E
#exibility to design coatings for speci"c applications, insensitivity to surface contamination, inde"nite shelf life prior to bonding, suitability for repair, low processing costs.
Successful plasma treatments of titanium have been reported, whereby bond strengths equivalent to the best chemical treatments were obtained [13]. At high temperature, the plasma-sprayed titanium joints produced superior results. Ramani [45] employed a similar technique to plasma spray called silicon sputtering, where a 200 As thick silicon layer was sputtered onto titanium. An increase in bond strength and durability was recorded and attributed to an increase in surface tension and surface roughness, which allowed the molten polymer adhesive to #ow in and around the increased surface area and interact with the silicon and oxides. 4.15. Sol/gel methods Since many chemical treatments have hazardous pollutants and/or working environments, alternative methods of surface treatment have been investigated. The sol/gel formation is water based; hence contains no hazardous pollutants. This recently developed system is based on the principle of a hybrid organic/inorganic coating providing a gradient between titanium and the adhesive [8]. This system has covalent bonding through the gradient coating, thereby reducing dependence of
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
Lewis acid base and hydrogen bonding interactions for adhesive bonding, which in turn increases bond durability in hot wet environments. Researchers have used an acid catalysed sol consisting of zirconium alkoxide and glycidtrimethoxysilane coupling agent in water [9]. In addition, a pre-bond treatment of solvent wipe and/or grit blasting prior to the sol/gel treatment was used. Drench and spray techniques as well as immersion are employed for the application of the sol/gel. Results obtained compared very favourably with chromic acid anodisation results after 2000 h of hot/wet exposure [9]. 4.16. Primers Primers may be applied to substrate surfaces for one or more of the following reasons: E to protect the substrate surface until bonding is carried out, E to increase surface wettability, E to block pores of porous surface thereby preventing adhesive escaping, E as a vehicle for corrosion inhibition, E as a coupling agent capable of forming chemical bonds with the adherend and adhesive. Coupling agents are believed to form covalent bonds between the adhesive and adherend, thereby producing strong and durable joints. Silanes are widely used coupling agents. Silane coupling agents have the following structure R}Si(OR) where R is the functional group that chemically reacts with the adhesive. R is usually an ethyl or methyl group. The main advantage of silane coupling agents is to improve durability of adhesive bonds in the presence of water or water vapour. 4.17. - APS -APS (aminopropyltriethoxysilane) is a common primer used on titanium and titanium alloys [10,12,46]. The -APS primer is applied by immersing the titanium substrate in 1% -APS for 15 min followed by blowing the excess o! in a stream of nitrogen [47]. The reported increase in the wet strength between Ti}6Al}4V and epoxy was 50% [46]. The -APS and -GPS (glycidoxypropylmethoxysilane) have been shown to improve the level of durability of grit-blasted specimens to that of sodium hydroxide and catalytic hydrogen peroxide etch [12]. In addition a primer called BR-127 is recommended by CYTEC the manufacturers of the adhesive FM-300, while BR-127, EC-3960 and EA-9223 are suitable primers for bonding titanium to composites [14]. However, the problem with BR-127 is that it contains hazardous chromates. Alternative primers
135
recommended include the non-chromated CYTEC XBR 6757. 4.18. Laser In the past surface treatment of titanium substrates has been based on either chemical or electrochemical processes. Although these treatments have been very successful, many contain hazardous chemicals such as chromates. As a result, new methods have been developed which are environmentally friendly. Recent studies have shown that excimer laser treatments result in an increase in surface roughness and the formation of a thin oxide layer [16]. The parameters that a!ect the degree of surface treatment include wavelength, polarisation and intensity. Results obtained [16] reveal no signi"cant difference in single lap shear strength results obtained using laser treatment compared to chromic acid anodisation which augers well for an environmentally friendly surface treatment. Broad [48] has revealed outstanding durability results using the patented CLP (CIBA Laser Pretreatment) laser treatment, whereby laser treated adhesive joints have not failed after 1400 days.
5. Conclusions While certain bonding techniques may provide adequate static strength, they may have little durability when exposed to hot moist environments. Others may be susceptible to debonding when exposed to harsh environmental conditions. Structural adhesive bonding of joints is achieved either by mechanical interlocking of the polymer with the adherend surface or chemical bonding of the polymer molecules with the metal oxide. To improve bond strength and bond durability surface treatments are a necessary pre-treatment prior to adhesive bonding. Titanium surface treatments include traditional methods such as chromic acid anodisation and sodium hydroxide anodisation as well as laser treatment. Typical composite surface treatments include traditional abrasion/solvent cleaning techniques for thermoset composites, while thermoplastic composites require surface chemistry and surface topographical changes to ensure strong and durable bond strengths. However, one fact remains, by increasing surface tension, increasing surface roughness and changing surface chemistry, a more intimate bond can be formed, which allows for increase in strength and durability. At present, the traditional treatments, which provide excellent durability, have to be phased out and as a result novel, cleaner greener methods have to be investigated. For titanium the laser treatment developed by CIBA has produced some very durable joints comparable to the traditional hazardous methods, while treatments such as plasma etching for polymer composites seems to produce durable joints.
136
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
References [1] Sha!er DK, Clear"eld HM, Ahearn JS. In: Minford JD, editor. Treatise on adhesion and adhesives, vol. 7. New York: Marcel Dekker, 1991. p. 437. [2] Clear"eld HM, McNamara DK, Davis GD. Adhesive bonding. New York: Plenum Press, 1990. [3] Venables JD. J Mater Sci 1984;19:2431. [4] Ahearn JS, Davis GD. Proceeding Adhesion 1987: Third International Conference, York, UK: Plastics and Rubber Institute, 1987. [5] Brockmann W, Henneman OD, Kollek H, Matz C. Int J Adhesion Adhesives 1986;6:115. [6] Stein BA, Hodges WT, Tyeryar JR. Proceedings of the 26th Structures, Structural Dynamics and Materials Conference AIAA, 1985, Paper No. 85-0750-CP. [7] Barron V. Ph.D. thesis, University of Limerick, 1998. [8] Arnold JR, Sanders CD, Bellevou DL, Martinelli AA, Gaskin GB. Proceedings of the 29th International SAMPE Technical Conference, vol.1, 1997. p. 345. [9] Blohowiak KY, Osborne JH, Krienke KA, Sekits DF. Proceedings of the 28th International SAMPE Technical Conference, 1996. p. 440. [10] Boerio FJ, Dillingham RG. In: Mittal KL, Editor. Adhesive joints formation, characteristics, and testing, New York: Plenum Press, 1991. p. 541. [11] British Aerospace Aircraft Group Weybridge-Bristol Division, 1984, Report No. AL/MAT/3864. [12] Critchlow GW, Brewis DM. Int J Adhesion Adhesives 1995;15(3):161. [13] Davis GD, Gro! GB, Zatorski RA. Surf Interface Anal 1997;25:366. [14] Lockheed-California Company. Prepared for National Air Development Centre, 1983, Report No. NADC-84124-60. [15] Mahoon A. In: Kinloch AJ, editor. Durability of structural adhesives. London: Applied Science Publishers, 1983. p. 255. [16] Sancaktar E, Zhang E. Reliability, stress analysis and failure prevention aspects of composite and active materials. ASME 1994;79:65. [17] Allen MR. NASA contractor. 1995. Report 198193. [18] Gutowski TG. The mechanics of composite deformation during the manufacturing processes. First Conference on Composite Materials, Dayton, Ohio: American Society for Composites, Oct 7}9, 1986. [19] Kinloch AJ. Adhesion and adhesives. London: Chapman & Hall, 1987. p.123}127. [20] Yoon TA, McGrath JE. Proceedings of the 36th International SAMPE Symposium, San Diego CA, USA, 1991. p.428. [21] Wing"eld J. Int J Adhesion Adhesives 1993;13(3):151.
[22] Blackman, Kinloch, Watts. Composites 1994;25(5). [23] Kempe G, Krauss H. Adhesion bonding techniques for highly loaded parts of continuous carbon-"bre reinforced polyetherketone (CF}PEEK/APC02). ICAS-92-6.9.1, 1992. [24] Cotter JL, Mahoon A. Int J Adhesion Adhesives 1982;2:47. [25] Hill JM, Karbashewski E, Lin A. J Adhesion Sci Technol 1995;9(12):1575. [26] Strobel M, Walzak MJ, Hill JM. J Adhesion Sci Technol 1995;9(3):365}83. [27] Kodokian GKA, Kinloch AJ. J Mater Sci Lett 1988;7:627. [28] Comyn J, Mascia Xiao. Int J Adhesion Adhesives 1996;16:301}4. [29] Schut. J Plastics Technol 1992;64. [30] Yasuda HK. Plasma polymerisation. USA: Academic Press, 1985. [31] Baalmann A, Vissing KD and Gross A. J Adhesion 1994;46. [32] Barron V. Buggy M. Proceedings of the 23rd Annual Adhesion Society Conference Myrtle Beach SC, 2000. p. 540. [33] Potts MW. Proceedings of the Polymer, Laminations and Coatings Conference, 1993. p. 444. [34] DiGiacomo J. Proceedings Finishing &91 Conference Cincinnati, Ohio, 1991. p. 376. [35] Demblon F. Vide Couches Minces 272 (Suppl) 1994;5:394. [36] Briggs D. J Mater Sci 1979:14. [37] Wurzberg E, Buchman A, Zylerstein E, Holdengraber Y, Dodiuk H. Int J Adhesion Adhesives 1987;10:337. [38] Buchman A, Dodiuk H, Rotal M, Zahavi J. Int J Adhesion Adhesives 1991;11:3. [39] Park JK, Mukherjee K. Mater Manufactr Process 1998;13(3):359. [40] Park JK, Howard JD, Chen CW, Mukherjee K. Proceedings of the American Society for Composites 14th Technical Conference, Dayton, OH, 1999. p. 661. [41] Miller PD, Je!reys RA, Pray HA. Metal Progress Report, 1974. p. 61. [42] Filbey JA, Wightman JP. In: Allen KW, editor. Adhesion, 12 ed. London: Applied Science Publishers, 1987. p. 17. [43] Pasajell Data sheets SEMCO Division Products Research and Chemical Corporation 5454 San Francisco Road Glendale CA. 1997. [44] Aronsson BO, Lausmaa J, Kasemo B. J Biomed Mater Res 1997;35(1):49. [45] Ramani K, Weidner WJ, Kumari G. Int J Adhesion Adhesives 1998;18(6):401. [46] Boerio FJ, Gosselin CA, Williams JW, Dillingham RG, Burkstrand JM. Proceedings of the Symposium on Polymer Composites, Seattle Washington, USA, 1983. p. 171. [47] Lees WA. Proceedings of the Bonding and Repair of Composites Seminar, Birmingham, UK, 1989. p. 17. [48] Broad R, French J, Sauer J. Int J Adhesion Adhesives 1999;19(2}3):193.
Surface treatment of titanium for adhesive bonding to polymer composites: a review P. Molitor*, V. Barron, T. Young Department of Mechanical and Aeronautical Engineering, University of Limerick, Limerick, Ireland Accepted 3 July 2000
Abstract At present, the bonding of polymer composites to titanium is a problem, which has not been fully resolved. Previous research has shown that bond strengths can be signi"cantly improved by surface treating the adherends prior to bonding. However many of the successful surface treatments involve the use of hazardous chemicals, which have to be phased out as part of an EU directive, which paves the way for less toxic environmentally friendly methods. In this paper various methods of surface treatment including traditional treatments such as acid etch, anodisation, novel plasma spray and laser treatments for both polymer composites and titanium will be discussed. These treatments will be reviewed with respect to changes in surface tension, surface roughness, surface chemistry and how these changes a!ect bond strength and durability of polymer composite titanium adhesive joints. 2001 Elsevier Science Ltd. All rights reserved. Keywords: A. Primers; B. Composites; C. Surface treatment; Adhesive bonding; Titanium
1. Introduction In order to produce a strong and durable adhesive joint between di!erent substrates, surface treatment is necessary. Certain bonding techniques provide adequate static strength, but have little durability when exposed to hot moist environments, while others are susceptible to debonding in the presence of fuels, oils and cleaning solvents commonly encountered in aircraft applications. In addition, the nature of the surface treatment prior to bonding is found to be a major in#uence in the control of this e!ect [1]. Surface treatment increases the bond strength by altering the substrate surface in a number of ways including increasing surface tension, increasing surface roughness or changing surface chemistry. By increasing surface roughness, an increase in surface area occurs which allows the adhesive to #ow in and around the irregularities on the surface to form a mechanical bond. Changing surface chemistry may result in the formation of a chemical bond e.g. between the polymer molecules in
* Corresponding author. Tel.: #353-61-202531; fax: #353-61202944. E-mail address: [email protected] (P. Molitor). Current address: Physics Department, Trinity College Dublin, Dublin 2, Ireland.
the polymer matrix composite and the metal oxide layer on the other adherend surface layer [2}5]. As a result, the nature of the surface will also in#uence the stability of the joint. When exposed to hot/humid environmental conditions, a polymeric adhesive/polymer interface is much more stable than the equivalent polymeric adhesive/metal interface. A well-chosen polymeric adhesive/polymer interface is unlikely to fail because of the environmental-induced stress due to the nature of the bond formed. On the other hand, the durability of a polymeric adhesive/metal joint is not as stable. Early studies in the 1960s revealed that these joints did not perform well in hot/wet conditions with frequent occurrence of short-term interfacial bond failure.
2. Adherend materials 2.1. Polymer composites Both thermoset and thermoplastic composite materials have been bonded to titanium with varying degrees of success [6]. However, it is noted that thermoplastic materials are inherently more di$cult to bond [7]. Carbon-"bre-reinforced polymer composite material has been used widely as the adherend in the literature
0143-7496/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 3 - 7 4 9 6 ( 0 0 ) 0 0 0 4 4 - 0
130
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
reviewed [8]. Although glass-"bre-reinforced composite material has a much lower modulus than carbon "bre, bond integrity tends to be dominated by the characteristics of the matrix, rather than the "bre and for this reason the research conducted on carbon "bre composite is also relevant to studies involving glass "bre. Thermoset polyester or epoxy composites can have a resin rich surface layer, which can cause particular problems in bonding. The reason for this is that either most of them contain a gel coat at the surface, or if the latter has not been deliberately created, the surface layer usually contains a higher proportion of resin as opposed to the interior. Both the gel coat and the resin-rich surface layer are very brittle and can subsequently fail catastrophically when overloaded. However, proper design and correct adhesive choice may compensate for this weakness. Highly compliant adhesives are a particularly good choice, as they spread the applied load over a large area and hence reduce the stress borne by the surface of the composite. 2.2. Titanium Previous researchers [8}16] have used various di!erent titanium alloys, however Ti}6Al}4V is the most widely used in the aerospace industry [17].
3. Surface treatments of polymer composites A variety of surface treatments have been used with various degrees of success to increase surface tension, increase surface roughness, change surface chemistry and thereby increase bond strength and durability of polymer composite adhesive joints and are shown in Table 1. Wetting of the adhesive by the adhesive is critical for the formation of secondary bonds. For complete wetting the surface energy of the adhesive must be lower than the surface energy of the adhesive [18]. For thermoplastic composites the primary aim of the surface treatment is to increase the surface energy of the adherend as much as possible. Surface treatments decrease water contact angle, increase surface tension and as a result increase bond strength. 3.1. Abrasion/solvent cleaning Abrasion/solvent cleaning may be employed to degrease the surface and remove mould release agents from the adherend [19}21]. The composite sheets can be lightly abraded using 180/220-mesh alumina, then wiped clean with a solvent such as methylethylketone (MEK) and allowed to dry. Previous researchers have found a signi"cant increase in surface roughness and bond strength for thermosets. However, the same treatment for thermoplastic polymer composites did not reveal any
signi"cant increase in bond strength or durability, due to the fact that some of these composites have very smooth low surface energy surfaces [19,21,22]. 3.2. Grit blasting Alternatively, an alumina grit blast (of particle size 45 m) with three passes at a distance of 15}20 cm may be employed, followed by solvent rinse and dried in nitrogen. MEK may not be compatible with local government regulations for industrial applications and as a result may have to be phased out and be replaced by a less hazardous solvent. The DLR (German Aerospace Research Establishment) has investigated three di!erent types of cleaning agents to be used with mechanical methods of surface treatment [23]. These included acetone, frigen and ethylacetate. Of these, the latter gave the highest lap shear values. Again, as with the abrasion and solvent wipe, grit blasting gave strong and durable bond strength for thermosets but revealed very little degree of success with thermoplastic materials. 3.3. Peel-ply Peel-ply, an impregnated ply is removed immediately prior to bonding. Previous research has shown that successful joints were obtained by increasing surface roughness, degreasing and removing of mould release agents [19]. The peel-ply is released from the surface because of the non-stick nature of the substrate within which it is impregnated. These are frequently #uorine silicon or #uorine compounds, which contaminate the laminate surface. Consequently, this implies that further treatment of the surface, e.g. by abrasion, is required if the strength of the subsequent bond is not to be seriously impaired [24]. If the ply contains a release agent, the adherend must be subsequently cleaned with a solvent and dried in a stream of nitrogen. 3.4. Tear-ply Tear-ply is mainly used for thermoset composites as thermosets are reactive upon heating and hence do not require a chemical surface treatment. It is fundamentally di!erent from peel-ply and consists of a fabric, which is completely impregnated by the laminate matrix resin, and as the name implies subsequently peeled o! the moulding. Super"cially, the surfaces of the laminates treated with either peel-ply or tear-ply appear very similar as both exhibit the pattern of the ply fabric. Tearply exploits a laminate property, which is generally regarded as being a serious disadvantage: the lack of strength in the z direction (normal to the surface). Immediately prior to a bonding operation, a corner of the fabric is raised with a knife and the tear-ply is then
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
131
Table 1 A!ect of surface treatment on polymer composites Treatment type Material
Nature of treatment
Abrasion and solvent wipe Grit blasting
Remove mould release Remove mould release Etch
Peel-ply
Thermoset and thermoplastic Thermoset and thermoplastic Thermoset and thermoplastic Thermoset
Tear-ply
Thermoset
Corona discharge Plasma treatment Flame treatment Laser treatment
Thermoplastic
Acid etch
Thermoplastic
Remove mould release Remove mould release Oxidising
Thermoplastic
Ablation and/ or oxidation Oxidising
Thermoset and thermoplastic
Ablation and/ or oxidation
Surface tension
Surface roughness
Surface chemistry
Bond strength
Durability
Ref.
Increase found for thermosets Increase found for thermosets Slight increase
Good for thermosets Good for thermosets Poor
[22,21]
Y
Increase
Good
[21]
Y
Increase
Good
[19]
Y
Double
[28]
Y
Increase
Y
Increase
Good (90 days) Good (90 days) * More research is necessary
[21]
Y Y Y
Y
Y Y
Y
Y Y
Increase
[23,28,21] [12,21,27]
[19,22,27,31] [8,21]
Depends on polymer matrix material.
torn o! the moulding, thus fracturing the resin interface between the tear-ply and the bulk moulding. The tear-ply must be carefully selected to enable it to be removed without any di$culty following the moulding process. Polyester resin has fracture characteristics, which permit this, and have little adhesion to the nylon "bres in the fabric. However, some resins are not compatible with this technique and hence peel-ply must be employed despite the di$culties arising from contamination. 3.5. Acid etching Acid etching has produced similar results to abrasion and grit blasting, in that an increase in bond strength is recorded for thermoset polymer composites, whereas little or no e!ect was recorded for thermoplastics [19,22,24].
surface chemistry by oxidising the polymer matrix, which results in the increase in bond strengths. 3.7. Plasma treatment The use of plasma treatment to treat polymers has been known for more than 20 years [29]. The plasma treatment involves a low-pressure plasma gas, which is electrically conductive and consists of excited atoms, ions and free radicals [30]. This allows polymer surfaces to be cleaned, etched or chemically modi"ed [19,30]. The plasma particles react not only with each other but also with the surfaces, which are exposed to the gas, giving rise to the following e!ects [19]:
3.6. Corona discharge treatment
E E E E E
Corona discharge, namely exposing the substrate surface to excited atoms, ions and free radicals at atmospheric pressure has been widely used to treat plastic surfaces for adhesive bonding [25}27]. The success of corona discharge in treating carbon-"bre}epoxide, carbon-"bre}PEEK and glass-"bre}polypropylene composites for bonding is reported, as is its use for treating polyole"n "lms to make them receptive to printing inks. Results [28] reveal that corona treatment increases surface tension and in some cases alters the
The combined e!ect of these processes results in an improvement of the adhesion properties of the surface. Plasma treatments involving various gases have been found to enhance the surface tension, oxidise the polymer matrix and increase the bond strength of PEEK composites [31]. Results obtained by Blackman [22] and Barron [7,32] echo the results obtained by corona treatment [28], in that plasma treating the PEEK composite produced an increase in surface roughness, surface tension and bond strength.
surface cleaning, degradation of the polymer chains, removal of material from the surface, formation of radicals on the surface, change of tacticity of the polymer chains.
132
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
3.8. Flame treatment Flame treatment is widely used in the surface modi"cation of polyole"ns, to improve printability and paintability, by introducing oxygen containing functional groups to the surface [25,33}36]. An oxidising #ame is very similar to a gas plasma in that it contains excited species such as atoms, ions and free radicals, which oxidise the surface of the specimens. The distance from the surface to be treated, the air/gas ratio and the dwell time are all critical parameters in producing a successful treatment. By oxidising the surface prior to bonding an increase in hydrophylicity and hence bond strength can occur. 3.9. Laser treatment Recently a novel method of surface treatment namely laser treatment [21,37,38] has been employed to increase surface roughness, surface tension and hence increase the bond strength of adhesive joints. Initial results seem promising however with respect to durability but more research is necessary. Park [39,40] has recently shown that laser treating polymer composite surfaces prior to bonding can produce high-strength adhesive bonds. Laser treatment results in increase in surface roughness due to the ablation of the polymer matrix.
#uorosilic acid under high pressure. A post treatment rinse in 5% nitric acid is required for Ti}6Al}4V titanium alloys to remove grey smut after blasting. An oxide "lm is produced which is crystalline in nature. This results in joint durability, which is reported to be slightly lower than TURCO 5578 and slightly higher than the phosphate #uoride process [12,15]. Most chemical treatments alter the titanium substrate by etching the existing oxide "lm. If the solvents are reducing agents, then a fresh thin oxide "lm is produced after the metal is removed from the solution, while in an oxidising solvent a thicker oxide "lm is produced [15]. 4.3. Acid etchants A number of acid mixtures have been used to treat the surface of titanium, including nitric-hydro#uoric acid, hydrochloric-orthophosphoric acid, hot sulphuric acid and sodium dichromate}chromic acid. Most of these treatments give adequate dry strengths but durability is reported to be poor [15]. Both stressed and unstressed peel and lap shear joints have shown to perform equal to or better than phosphate #uoride or modi"ed phosphate #uoride processes [12]. An additional problem associated with these treatments is hydrogen pick up. However, this problem can be avoided with the use of alkaline solutions as opposed to acid etchants [12].
4. Surface treatment of titanium 4.4. Alkaline peroxide etch Mechanical, chemical, electrochemical and energetic surface treatments are used to enhance the surface of titanium alloys prior to bonding. Durability studies of Ti}6Al}4V reveal that surface preparations that produce no roughness (macro or micro) yield the poorest bond durability. Those that produce signi"cant macro-roughness but little micro-roughness yield moderate to good durability. Finally those that produce signi"cant microroughness yield the best durability (Table 2). 4.1. Abrasion and grit blasting A mechanical treatment is used primarily to produce a clean macroscopically rough surface and to remove some of the existing oxide layer. The combination of a clean adherend surface with signi"cant macro-roughness improves the initial dry strength [15]. However, abrasion and grit blasting techniques as discussed in the previous section alone are not adequate methods of surface treatment, but when combined with chemical or electrochemical treatment, durable bond strengths can be obtained [19]. 4.2. VAST (vought abrasive surface treatment) In this treatment, the titanium is blasted with a slurry of "ne abrasive alumina (220 mesh) containing 2%
Initial alkaline peroxide treatments were carried out at room temperature, but this process, which lasted up to 36 h, was too lengthy for industrial applications. The immersion time can be reduced to 20 min by increasing the temperature to 50}703C [15]. Depending upon the concentration of sodium hydroxide and hydrogen peroxide, the metal is either etched or oxidised. Those concentrations, which produce grey oxides, have been found to produce adhesive wettable surfaces. An oxide layer up to 2 m thick, which is stable up to 2003C and capable of forming high bond strengths at elevated temperatures and high-humidity environments were produced using a sodium hydrogen peroxide etch [24]. Alkaline peroxide etch is reported to be satisfactory but has long heat up times and has high hydrogen peroxide consumption. Oxide formation is related to the rate of hydrogen peroxide decomposition. The rate of decomposition is increased by adding 5}10% heavy metal ions to the hydrogen peroxide solution. 4.5. Phosphate yuoride process The phosphate-#uoride-based treatments are based on combinations of trisodium phosphate, disodium tetraborate, potassium #uoride and hydro#uoric acid [41]. A process of pre-etching in 3% hydro#uoric acid and
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
133
Table 2 A!ect of surface treatment on titanium alloys Treatment type
Alloy
Nature of treatment
Surface roughness
Oxide layer (nm)
Bond strength
Durability
Ref.
Abrasion and solvent wipe Grit blasting
Ti}6Al}4V
Macro
*
Poor
Poor
[15]
Macro
*
Increase
Adequate
[12,15]
VAST
Ti}6Al}4V
Macro
*
Good
Poor
[12,15]
Acid etch Alkaline etch Phosphate}#uoride Modi"ed Phosphate}Fluoride
Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V
Remove mould release Remove mould release Remove mould release Etch Etch Etch Etchant and oxidation
Micro Micro None None
* 60}200 20 8
Adequate poor Good Adequate Adequate
[15,45] [15,24,45] [15,45] [15,45]
TURCO DAPCOtreat Pasajell Chromic acid anodisation NaOH anodisation Cathodically deposited Al O Plasma spray
Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V Ti}6Al}4V
Oxidising Oxidising Oxidising
Macro Macro Macro Micro
17.5 6 10}20 40}140
Adequate Increase Adequate High
Poor Good Poor Better than phosphate} #uoride Adequate Good Adequate Excellent
Ti}6Al}4V *
Oxidising Oxidising
Micro *
80}90 *
High Adequate
Excellent Adequate
[45] [15,45]
Ti}6Al}4V
Micro
130
High
Excellent
[15,45]
Sol gel
Ti}6Al}4V
*
*
High
Good
[9]
Laser treatment
Ti}6Al}4V
Ablation and oxidation Coupling and oxidation Ablation and oxidation
Macro
*
High
Poor
[16,49]
Ti}Al}4V
15% nitric acid followed by a 2 min dip in the 5% trisodium phosphate, 2% potassium #uoride and 2.6% hydro#uoric acid mixture has become known as the phosphate #uoride process (US Patent 2 864 732). The rinsing procedure after this treatment is critical to remove excess chemicals and a 15 min soak in de-ionised water at 603C has been recommended [15]. 4.6. Modixed phosphate yuoride process It has been shown that on exposure to warm moist environments, the anatase oxide layer produced by the phosphate, #uoride process slowly reverts to rutile [15]. As a result, there is a decrease in volume of about 8%, which results in the development of stresses at the adhesive/oxide layer interface. By stabilising the anatase structure the joint durability is enhanced. This is achieved by adding 0.75% anhydrous sodium sulphate into the etch solution [12,15]. However, both of these treatments have been out-performed by treatments such as TURCO 5578, sodium hydroxide anodisation and chromic acid anodisation [15]. 4.7. TURCO 5578 TURCO 5578 is an alkaline-based etchant, which contains caustic soda, sodium metasilicate and pyro-
[15,45] [45] [15,45] [45]
phosphate. This treatment produces a large amount of macroroughness with little or no microroughness. The oxide "lm produced is &17.5 nm thick with a macroroughness of 3.4 m peak to valley [42]. The durability of this treatment is much better than that of the phosphate #uoride treatments. In durability trials, TURCO 5578 is only out-performed by chromic acid anodisation [12]. However, the added advantage of the TURCO 5578 treatment is that there is no hydrogen embrittlement as observed in the acid etching process. 4.8. DAPCOtreat DAPCOtreat is similar to TURCO 5578 in that it results in an increase in macroroughness with little or no microroughness. DAPCOtreat produces a thin 6 nm oxide layer on the titanium substrate [12]. 4.9. Pasajell 107 treatment Pasajell 107 is recommended as a prebond treatment for titanium [5,12,14,15,43]. The chemical composition of pasajell 107 is 40% nitric acid, 10% combined #uorides, 10% chromic acid, 1% couplers and the balance is water. It is available as a thixotropic paste for brush applications or as an immersion solution for tank treatment. A recommended treatment time is 10}15 min.
134
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
Pasajell 107 is applied in combination with a pre-treatment of MEK wipe, blast with 320 grit non-silicone sand, followed by non-chlorinated solvent wipe [14]. After applying the pasajell for 10}15 min, the coated area is rinsed with deionised water. The process produces an amorphous looking oxide, which has anatase structure, which is stable up to 1753C and converts to rutile at 3503C. Comparing this treatment to the alkaline peroxide etch, the bond durability of the pasajell treatment is reported to greatly exceed that of the alkaline peroxide etch at elevated temperatures and humid environments [14]. 4.10. Electrochemical reactions Electrochemical reactions have been used for cleaning, etching and oxidising metal surfaces, where the metal acts as an anode or cathode. Anodisation creates an oxide surface that is dependent upon the electrolyte, anodisation voltage, time and temperature [5]. There are various anodisation techniques employed for titanium alloy surface treatment including chromic acid anodisation and sodium hydroxide anodisation. 4.11. Chromic acid anodisation Chromic acid anodisation produces a surface with signi"cant microroughness and an oxide thickness of 40 and 80 nm for the 5 and 10 V treatments, respectively [12]. In addition, chromic acid anodisation oxides are reported to exhibit remarkable bond durability and provide the target bond strengths and durability for all other treatments. 4.12. Sodium hydroxide anodisation Previous researchers investigated the use of sodium hydroxide}hydrogen peroxide anodisation and peroxide free sodium hydroxide anodisation as methods of surface treating titanium for adhesive bonding [11]. It was found that specimens prepared by these methods produced high-strength adhesive bonds and exhibited good durability when exposed to conditions of heat, moisture and stress. The presence of hydrogen peroxide was found to be neither necessary nor bene"cial and in some cases resulted in a decrease in bond strength. In addition, the sodium hydroxide anodising solution could be reused and was considered to present very low operational and environmental hazards. 4.13. Cathodically deposited aluminium oxide Cathodic depositions of metal oxides from alcohol solutions containing inorganic nitrates have shown good wettability and environmental resistance in hot and humid conditions. A recommended solution is based on
10 g of hydrated aluminium nitrate dissolved in 1 l of isopropyl alcohol [15]. By making titanium, the cathode of an electrolytic cell at 30 V the oxide is deposited on the surface. Results have shown joint strengths and durability to be better than the VAST and TURCO 5578 processes [15]. 4.14. Plasma treatment Both glow discharges and plasma sprays have been examined as methods of surface treatment for titanium alloys. Aronsson [44] investigated the use of glow discharge. Results obtained revealed that this method of plasma treatment produced clean surfaces and reproducible results, depending upon the degree of treatment. Furthermore, oxidation in pure oxygen resulted in uniform and stoichiometric TiO surface oxide layers with reproducible composition and thickness. Plasma spraying involves rapidly heating either, TiO TiSi , MgO or SiO powder to a molten or semi-molten state and then spraying it onto the substrate at high velocity. This is another method of surface treatment employed for adhesive bonding that involves no hazardous chemicals or pollutants. Other advantages of plasma spraying over chemical treatments include [13]: E E E E E
#exibility to design coatings for speci"c applications, insensitivity to surface contamination, inde"nite shelf life prior to bonding, suitability for repair, low processing costs.
Successful plasma treatments of titanium have been reported, whereby bond strengths equivalent to the best chemical treatments were obtained [13]. At high temperature, the plasma-sprayed titanium joints produced superior results. Ramani [45] employed a similar technique to plasma spray called silicon sputtering, where a 200 As thick silicon layer was sputtered onto titanium. An increase in bond strength and durability was recorded and attributed to an increase in surface tension and surface roughness, which allowed the molten polymer adhesive to #ow in and around the increased surface area and interact with the silicon and oxides. 4.15. Sol/gel methods Since many chemical treatments have hazardous pollutants and/or working environments, alternative methods of surface treatment have been investigated. The sol/gel formation is water based; hence contains no hazardous pollutants. This recently developed system is based on the principle of a hybrid organic/inorganic coating providing a gradient between titanium and the adhesive [8]. This system has covalent bonding through the gradient coating, thereby reducing dependence of
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
Lewis acid base and hydrogen bonding interactions for adhesive bonding, which in turn increases bond durability in hot wet environments. Researchers have used an acid catalysed sol consisting of zirconium alkoxide and glycidtrimethoxysilane coupling agent in water [9]. In addition, a pre-bond treatment of solvent wipe and/or grit blasting prior to the sol/gel treatment was used. Drench and spray techniques as well as immersion are employed for the application of the sol/gel. Results obtained compared very favourably with chromic acid anodisation results after 2000 h of hot/wet exposure [9]. 4.16. Primers Primers may be applied to substrate surfaces for one or more of the following reasons: E to protect the substrate surface until bonding is carried out, E to increase surface wettability, E to block pores of porous surface thereby preventing adhesive escaping, E as a vehicle for corrosion inhibition, E as a coupling agent capable of forming chemical bonds with the adherend and adhesive. Coupling agents are believed to form covalent bonds between the adhesive and adherend, thereby producing strong and durable joints. Silanes are widely used coupling agents. Silane coupling agents have the following structure R}Si(OR) where R is the functional group that chemically reacts with the adhesive. R is usually an ethyl or methyl group. The main advantage of silane coupling agents is to improve durability of adhesive bonds in the presence of water or water vapour. 4.17. - APS -APS (aminopropyltriethoxysilane) is a common primer used on titanium and titanium alloys [10,12,46]. The -APS primer is applied by immersing the titanium substrate in 1% -APS for 15 min followed by blowing the excess o! in a stream of nitrogen [47]. The reported increase in the wet strength between Ti}6Al}4V and epoxy was 50% [46]. The -APS and -GPS (glycidoxypropylmethoxysilane) have been shown to improve the level of durability of grit-blasted specimens to that of sodium hydroxide and catalytic hydrogen peroxide etch [12]. In addition a primer called BR-127 is recommended by CYTEC the manufacturers of the adhesive FM-300, while BR-127, EC-3960 and EA-9223 are suitable primers for bonding titanium to composites [14]. However, the problem with BR-127 is that it contains hazardous chromates. Alternative primers
135
recommended include the non-chromated CYTEC XBR 6757. 4.18. Laser In the past surface treatment of titanium substrates has been based on either chemical or electrochemical processes. Although these treatments have been very successful, many contain hazardous chemicals such as chromates. As a result, new methods have been developed which are environmentally friendly. Recent studies have shown that excimer laser treatments result in an increase in surface roughness and the formation of a thin oxide layer [16]. The parameters that a!ect the degree of surface treatment include wavelength, polarisation and intensity. Results obtained [16] reveal no signi"cant difference in single lap shear strength results obtained using laser treatment compared to chromic acid anodisation which augers well for an environmentally friendly surface treatment. Broad [48] has revealed outstanding durability results using the patented CLP (CIBA Laser Pretreatment) laser treatment, whereby laser treated adhesive joints have not failed after 1400 days.
5. Conclusions While certain bonding techniques may provide adequate static strength, they may have little durability when exposed to hot moist environments. Others may be susceptible to debonding when exposed to harsh environmental conditions. Structural adhesive bonding of joints is achieved either by mechanical interlocking of the polymer with the adherend surface or chemical bonding of the polymer molecules with the metal oxide. To improve bond strength and bond durability surface treatments are a necessary pre-treatment prior to adhesive bonding. Titanium surface treatments include traditional methods such as chromic acid anodisation and sodium hydroxide anodisation as well as laser treatment. Typical composite surface treatments include traditional abrasion/solvent cleaning techniques for thermoset composites, while thermoplastic composites require surface chemistry and surface topographical changes to ensure strong and durable bond strengths. However, one fact remains, by increasing surface tension, increasing surface roughness and changing surface chemistry, a more intimate bond can be formed, which allows for increase in strength and durability. At present, the traditional treatments, which provide excellent durability, have to be phased out and as a result novel, cleaner greener methods have to be investigated. For titanium the laser treatment developed by CIBA has produced some very durable joints comparable to the traditional hazardous methods, while treatments such as plasma etching for polymer composites seems to produce durable joints.
136
P. Molitor et al. / International Journal of Adhesion & Adhesives 21 (2001) 129}136
References [1] Sha!er DK, Clear"eld HM, Ahearn JS. In: Minford JD, editor. Treatise on adhesion and adhesives, vol. 7. New York: Marcel Dekker, 1991. p. 437. [2] Clear"eld HM, McNamara DK, Davis GD. Adhesive bonding. New York: Plenum Press, 1990. [3] Venables JD. J Mater Sci 1984;19:2431. [4] Ahearn JS, Davis GD. Proceeding Adhesion 1987: Third International Conference, York, UK: Plastics and Rubber Institute, 1987. [5] Brockmann W, Henneman OD, Kollek H, Matz C. Int J Adhesion Adhesives 1986;6:115. [6] Stein BA, Hodges WT, Tyeryar JR. Proceedings of the 26th Structures, Structural Dynamics and Materials Conference AIAA, 1985, Paper No. 85-0750-CP. [7] Barron V. Ph.D. thesis, University of Limerick, 1998. [8] Arnold JR, Sanders CD, Bellevou DL, Martinelli AA, Gaskin GB. Proceedings of the 29th International SAMPE Technical Conference, vol.1, 1997. p. 345. [9] Blohowiak KY, Osborne JH, Krienke KA, Sekits DF. Proceedings of the 28th International SAMPE Technical Conference, 1996. p. 440. [10] Boerio FJ, Dillingham RG. In: Mittal KL, Editor. Adhesive joints formation, characteristics, and testing, New York: Plenum Press, 1991. p. 541. [11] British Aerospace Aircraft Group Weybridge-Bristol Division, 1984, Report No. AL/MAT/3864. [12] Critchlow GW, Brewis DM. Int J Adhesion Adhesives 1995;15(3):161. [13] Davis GD, Gro! GB, Zatorski RA. Surf Interface Anal 1997;25:366. [14] Lockheed-California Company. Prepared for National Air Development Centre, 1983, Report No. NADC-84124-60. [15] Mahoon A. In: Kinloch AJ, editor. Durability of structural adhesives. London: Applied Science Publishers, 1983. p. 255. [16] Sancaktar E, Zhang E. Reliability, stress analysis and failure prevention aspects of composite and active materials. ASME 1994;79:65. [17] Allen MR. NASA contractor. 1995. Report 198193. [18] Gutowski TG. The mechanics of composite deformation during the manufacturing processes. First Conference on Composite Materials, Dayton, Ohio: American Society for Composites, Oct 7}9, 1986. [19] Kinloch AJ. Adhesion and adhesives. London: Chapman & Hall, 1987. p.123}127. [20] Yoon TA, McGrath JE. Proceedings of the 36th International SAMPE Symposium, San Diego CA, USA, 1991. p.428. [21] Wing"eld J. Int J Adhesion Adhesives 1993;13(3):151.
[22] Blackman, Kinloch, Watts. Composites 1994;25(5). [23] Kempe G, Krauss H. Adhesion bonding techniques for highly loaded parts of continuous carbon-"bre reinforced polyetherketone (CF}PEEK/APC02). ICAS-92-6.9.1, 1992. [24] Cotter JL, Mahoon A. Int J Adhesion Adhesives 1982;2:47. [25] Hill JM, Karbashewski E, Lin A. J Adhesion Sci Technol 1995;9(12):1575. [26] Strobel M, Walzak MJ, Hill JM. J Adhesion Sci Technol 1995;9(3):365}83. [27] Kodokian GKA, Kinloch AJ. J Mater Sci Lett 1988;7:627. [28] Comyn J, Mascia Xiao. Int J Adhesion Adhesives 1996;16:301}4. [29] Schut. J Plastics Technol 1992;64. [30] Yasuda HK. Plasma polymerisation. USA: Academic Press, 1985. [31] Baalmann A, Vissing KD and Gross A. J Adhesion 1994;46. [32] Barron V. Buggy M. Proceedings of the 23rd Annual Adhesion Society Conference Myrtle Beach SC, 2000. p. 540. [33] Potts MW. Proceedings of the Polymer, Laminations and Coatings Conference, 1993. p. 444. [34] DiGiacomo J. Proceedings Finishing &91 Conference Cincinnati, Ohio, 1991. p. 376. [35] Demblon F. Vide Couches Minces 272 (Suppl) 1994;5:394. [36] Briggs D. J Mater Sci 1979:14. [37] Wurzberg E, Buchman A, Zylerstein E, Holdengraber Y, Dodiuk H. Int J Adhesion Adhesives 1987;10:337. [38] Buchman A, Dodiuk H, Rotal M, Zahavi J. Int J Adhesion Adhesives 1991;11:3. [39] Park JK, Mukherjee K. Mater Manufactr Process 1998;13(3):359. [40] Park JK, Howard JD, Chen CW, Mukherjee K. Proceedings of the American Society for Composites 14th Technical Conference, Dayton, OH, 1999. p. 661. [41] Miller PD, Je!reys RA, Pray HA. Metal Progress Report, 1974. p. 61. [42] Filbey JA, Wightman JP. In: Allen KW, editor. Adhesion, 12 ed. London: Applied Science Publishers, 1987. p. 17. [43] Pasajell Data sheets SEMCO Division Products Research and Chemical Corporation 5454 San Francisco Road Glendale CA. 1997. [44] Aronsson BO, Lausmaa J, Kasemo B. J Biomed Mater Res 1997;35(1):49. [45] Ramani K, Weidner WJ, Kumari G. Int J Adhesion Adhesives 1998;18(6):401. [46] Boerio FJ, Gosselin CA, Williams JW, Dillingham RG, Burkstrand JM. Proceedings of the Symposium on Polymer Composites, Seattle Washington, USA, 1983. p. 171. [47] Lees WA. Proceedings of the Bonding and Repair of Composites Seminar, Birmingham, UK, 1989. p. 17. [48] Broad R, French J, Sauer J. Int J Adhesion Adhesives 1999;19(2}3):193.