Adhesion, the binding of different materials at an interface, is of general interest to many branches of technology, e.g., micro-electronics, tribology, manufacturing, construction, etc. However, there is a lack of fundamental understanding of such diverse interfaces. In addition, experimental techniques generally have practical objectives, such as the achievement of sufficient strength to sustain mechanical or thermal effects and/or have the proper electronic properties. In addition, the theoretical description of binding at interfaces is quite limited, and a proper data base for such theoretical analysis does not exist.

This presentation will review both experimental and theoretical aspects of adhesion in nonpolymer materials. The objective will be to delineate the critical parameters needed, governing adhesion testing along with an outline of testing objectives. A distinction will be made between practical and fundamental objectives. Examples will be given where interfacial bonding may govern experimental consideration. The present status of theory will be presented along with some recommendations for future progress and needs.

Young's modulus, tensile strength and the storage shear modulus experiments were performed on silica-filled ethylene-vinyl acetate composites and the results were correlated with polymer-filler interaction energy. The polymer-filler interaction energy was determined by wetting and infrared spectroscopy experiments. The interaction energy was characterized by the work of adhesion. The results indicate there is a strong correlation between modulus and tensile strength and the work of adhesion. There was an exponential relationship of modulus and tensile strength with the reciprocal of the work of adhesion.

Lattice statics and lattice dyanamics computer simulation methods are applied to investigate the elastic properties of grain boundaries in fcc metals. The elastic constants in the GB reigon are found to differ significantly from their bulk ideal-crystal values. Due to the broken symmetry in the GB region a pronounced anisotropy of the elastic properties is observed. Whereas Young's moduli are found to increase in the GB region, the shear moduli and Poisson's ratios generally decrease. The magnitude of these effects is shown to depend sensitively on the GB structure.

The mechanics aspects of the peel test are analyzed. In this analysis, the limitation and applicability of the peel test for thin film adhesion have been investigated. Firstly, the effect of plasticity in the peel test of thin metallic films is analyzed. The analysis has given a closed form solution for the partitioning of peel strength between the energy expenditure rate for the decohesion process and the internal work expenditure rate caused by plastic deformation (bending) of the film. The results predicted the variation of peel strength as a function of the film thickness, the film yield stress and the compliance of the substrate. This analysis shows that the peel strength is very sensitive to the thickness of the film and the yield stress of the film. As observed in the experiment, the peel strength has a peak value at a certain thickness of the film. This is explained for both L and T peel tests. Unlike the L peel test, the peak phenomenon in the T peel test (of metallic films) is due to the hardening of the adhered film. Secondly, the effect of plasticity and viscoelasticity in the peel test of thin polymer films is analyzed. Unlike the peel test of metallic films, tension and shear effect of the film deformation is very important in the analysis of the polymer peel test. An integral equation is formulated for the viscoelastic models of the adhered film. The solution of the integral equation reveals the important parameters in polymer peeling and their contribution to the peel force as well as to the viscoelastic dissipation. In addition, the meaning of the interface toughness is reexamined and the interface toughness is evaluated by the peel test. Based on these analyses, a Universal Peel Diagram has been constructed, which accounts for the hardening behavior of the film. On this diagram, objective interfacial decohesion toughness can be readily obtained using very simple peel tests. The analysis includes not only the 900° peel test but also the general angle peel test. Comparison between the theory and experiment has been made with the experimental data provided by J. Kim of IBM. The comparison shows good agreement.

A simple model is presented for the elastic behavior of a powder compact that has undergone small amounts of sintering. The analysis shows that the elastic constants are sensitive to small changes in the interparticle contact areas. The elastic constants are also shown to be dependent on factors relating to the structure of the powder compact, such as the particle coordination number. It is concluded that such elastic constant measurements are a valuable tool in studying adhesion within compacts, the mechanisms involved in the initial stages of sintering and the characterization of powders.

The problem of adhesion cannot be completely comprehended by consideration of interfacial properties alone. Other parameters such as geometry, applied and residual stresses, and the interfacial flaw population are equally important. The well-established field of fracture mechanics permits these aspects to be considered. One of the most important features of fracture mechanics is that it can provide a means of characterizing interfacial properties which does not depend upon the method of testing. It may also allow a predictive ability in determining failure and a rational explanation of what is known as “adhesive” or “cohesive” failure. These issues are addressed in conjunction with a discussion of the current limitations in the application of fracture mechanics to the topic of adhesion.

A novel analysis is presented to explain mechanistic aspects of the fracture of low molecular weight polymers. The fracture energy is calculated as a function of the mean interpenetration distance of a polymer chain which has been derived from the solution of the Fokker-Planck equation. Experimental results with monodisperse polystyrene have been used to validate the theory.

A simple adhesion process has been developed for applying glass lenses to the active surfaces of Si and GaAs optical devices such as detectors and solar cells [1]. The method, which uses a film of fluorinated ethylene propylene (FEP) offers reliability, cost and throughput advantages over the conventional dimethyl silicone adhesive. Critical to the process is the ion bombardment of individual surfaces before assembly. The resulting enhanced adhesion suggests some modification of surface properties. Four possible mechanisms are briefly discussed. The technique is believed to have broad applicability to amorphous and crystalline semiconductor systems and devices other than those under test.

A theoretical formalism to describe interface adhesion phenomena between materials has been developed. In this model electron densities, potentials, and adhesion energies of thin metal films at metal–semiconductor and metal–insulator interfaces are derived through a partially self–consistent calculation. The theory is based on a density–functional formalism applied to a simple model of the system in which the metal is replaced by a uniform positive background and the semiconductor by a continuum with a static dielectric constant. Numerical results of the metal electron density distribution and effective potential of Au–vacuum and Au–GaAs system, and the adhesion energy at Au–GaAs interface are presented. The interface energies are then used to estimate the adhesion strength and compare with experiment.

An atomic force microscope has been used to measure the frictional force on a tungsten tip sliding across the basal plane of graphite at low loads < 10-4 N. The frictional force displays the 2.5 Å periodicity of the graphite surface.

The presence of a liquid film between contacting ultra-flat surfaces can produce high levels of adhesion. The liquid film may be a pre-existing film of lubricant or may be a capillary condensate. This paper considers the meniscus and viscous components to the adhesion. Different regimes are identified which are distinguished by differing quantities of liquid in the gap between the surfaces. The role of surface roughness is also evaluated. The force required to separate the surfaces in the tangential direction is considered. The magnitude of this force is dependent not only on the adhesion but also on the separation rate and loading conditions.

Since the discovery in 1981 that the electronic excitation produced by MeV ions can enhance thin film adhesion to a substrate, there have been many investigations to determine the origin of the effect and to apply it to problems of technological interest. As we approach the seventh anniversary of its discovery, I shall use this opportunity to review our progress toward these goals. In general, we have discovered that although the electronic excitation provided by the incident beam is responsible for the improvement in adhesion, it is certain that there are quite a number of separate effects involved. These have been shown to include: surface cracking of the substrate; the breakup of contaminant or oxide layers on the substrate by the beam; and the relaxation of surface strain in the substrate. The recent use of a variety of new techniques seems to offer the hope that we are beginning to make progress in untangling these phenomena.

We report on the use of 20 MeV C14+ ion beams for pre-irradiation of Si, SiO2 and Al2O3 surfaces with resulting alterations in the adhesion of subsequently deposited Au and Ag thin films. The effect does not follow the same pattern for Au and Ag films, nor is there a simple correlation to the previously observed adhesion effect induced by ion irradiation after film deposition. Possible mechanisms, both for pre-irradiation and post-irradiation effects, are discussed, as is the importance of chemical structures with respect to interfacial bond strength.

A systematic study of Al/metal reactions has been performed. The thin-film interactions of Al with refractory metals (Ti, V, Ta, Cr, Mo, W and Co) and near-noble metals (Ni, Pd, and Pt) have been investigated. The initial aluminide phases to grow are the Al-rich phases: TiAl3, Val3, TaAl3, Cr2Al13, MoAl12, Wal12, Co2Al9, NiAl3, Pd2Al3, and Pt2Al3 at temperatures between 225 and 525°C. With the exceptions of Val3, Pd2Al3, and Pt2Al3. these are the most Al-rich phases on the phase diagrams. Marker experiments were performed and Al was the dominant diffusing species during the growth of these phases, TiAl3, Val3, Cr2Al3, MoAl12, Co2Al2, NiAl3, Pd2Al3, and Pt2AI3. Consistent with the faster Al diffusion, which provides a greater supply of Al to the growing interface, is the growth of the most Al-rich phase initially. For the exceptions to this rule, Val3, Pd2AI3, and Pt2 Al3, the complexity of the Al-rich V, Pd, and Pt end phases may have hindered nucleation, resulting in the growth of the observed phases. The subsequent phase formation was examined in the Ni-, Pd-, and Pt-Al systems. After initial phase formation consumed all the Al or metal, subsequent phases formed in accordance with the overall stoichiometry. The results of this study, along with a brief literature review, are presented and the generalized behavior of Al/transition metal reactions discussed.

The interfacial adhesion of Ti films on SiO2 substrate was studied by room temperature argon ion irradiation. Range of ion energy was chosen from 100 to 300 keV. Adhesive strength was measured by scratching test. For Ti/SiO2 pair irradiated by 100 keV argon ions, the adhesion was easier to enhance and much greater strength was obtained than that irradiated by 300 keV argon ions. The threshold dose also increased with the increasing of ion energy. The adhesive strength and the threshold dose increased when the metallizing temperatures were higher. Rutherford backscattering spectra(RBS) showed that a transition layer of about 10 nm thick was formed in Ti/SiO2 interface region after irradiation to a dose of 5X1016 Ar/cm2, indicating some chemical reaction has probably taken place. The experimental results are discussed in terms of thermodynamics of solids.

We have examined the effects of irradiation with 200 keV helium ions on the adhesion and optical transmission properties of 220 nm thick antireflective, optical coatings that consisted of layers of spherical, 20 nm diameter silica particles. In the as-deposited state these sol-gel coatings have very low adhesion to the fused silica subtrates. Results for adhesion and optical transmission have been obtained for doses of 1013–1017 He/cm2. Significant improvement in adhesion was found for doses exceeding about 2×1014 He/cm2. Optical transmission measurements for wavelengths of 200–1200 nm showed increasing absorption with dose. We have evidence that the helium ions decompose various contaminants in the coating into two types of degradation products. One is volatile and the other remains in the coating as optically absorbing species. UV/oxidative-gas treatment effectively removes the absorbing species.

X-ray Photoelectron Spectroscopy and Rutherford Backscattering Spectroscopy with channeling are employed to study surface and interface changes resulting from irradiation of thin Al films on Si-SiO2 substrates using < 6eV visible photons. Results indicati that surface oxidation and bonding rearrangements at the Al-SiO2-Si interface can take place at room temperature under photon bombardment. These changes are correlated with enhanced adhesion and modification of film etch properties which are also a result of photon irradiation.

A thin Cu film(400 Å) was deposited on a smooth polyimide(PI) substrate. Ar+ ion implantation onto the Cu/PI film has been shown to mix Cu and PI and to modify the impact-abrasive wear property. Ar+ ions with energies of 200 keV, and dosage between 1015 to 4×1016 ions/cm2 were used. The surface analyses were carried out with RBS, X-ray and Optical Microscope. The wear properties of the Cu-PI system were determined by a newly constructed Impact-Abrasive Wear Tester. An X-ray diffraction study shows an increase in Cu( 111 )peak with Ar+ ion dosage. The wear property was found to be a function of ion energy, ion dosage, crystallinity of Cu, the amount of mixing of Cu and PI, and the damage of PI substrate due to ion implantation.

Atomistically deposited films may form with high residual stresses which may be either tensile or compressive in nature. These film stresses represent stored strain energy which may affect the adhesion of the film-substrate couple and in the limit may cause spontaneous fracture at or near the film-substrate interface (loss of adhesion). In the post cathode magnetron sputter deposition of molybdenum films, we have found that the intrinsic film stresses are generally anisotropic and may easily exceed the fracture or adhesive strength of the film-substrate couple. The residual stress anisotropy in the film is dependent on the orientation with respect to the post cathode and the magnitude and nature of the stresses are very dependent on the deposition conditions, particularly gas pressure during sputtering. By using a pressure-cycling technique, we have deposited thick (5 microns) films of molybdenum which have little residual stress or stress anisotropy.

The influence of a number of substrate conditions and interfacial contaminants on the adhesion strength of thin Au films (10 to 100 nm thick) to GaAs substrates was studied. The GaAs surfaces were prepared by cleaving on the <110> plane under controlled atmospheres and conditions prior to depositing the Au film. Adhesion strengths were measured using the pinpull test. The study included standard characterization techniques to examine the substrate surface and the Au-GaAs interfaces. A direct correlation was found between cases of poor adhesion and the presence of hydroxyl groups on the GaAs surface. The influence of post evaporative treatments to improve the adhesion of specific cases was also studied.