ASM Metals HandBook Vol. 8 - Mechanical Testing and Evaluation

Подождите немного. Документ загружается.

result from an appropriate combination of local galvanic electrocoupling and moisture environment that is quite

common for microelectronics applications.

References cited in this section

22. A. Carre and J. Shultz, Polymer-Aluminum Adhesion. III. Effect of a Liquid Environment, J. Adhesion,

Vol 18, 1984, p 171–184

23. J.D. Venables, Review: Adhesion and Durability of Metal-Polymer Bonds, J. Mater. Sci., Vol 19, 1984,

p 2431–2453

24. H. Wu, J.T. Dickinson, and S.C. Langford, Dynamic Measurements of Humidity Attack on

Polymer/Glass Interfaces under Stress, J. Adhes. Sci., Vol 11, 1997, p 695–717

25. S.X. Mao and A.G. Evans, The Influence of Blunting on Crack Growth at Oxide/Metal Interfaces, Acta

Mater., Vol 45, 1997, p 4263–4270

26. T.S. Oh, J. Rodel, R.M. Cannon, and R.O. Ritchie, Ceramic/Metal Interfacial Crack Growth

Toughening by Controlled Microcracks and Interfacial Geometries, Acta Metall., Vol 36, 1988, p 2083–

2093

27. N.R. Moody, S.K. Venkataraman, B. Bastaez, J.E. Angelo, and W.W. Gerberich, Hydrogen Effects on

the Fracture of Thin Tantalum Nitride Films, Mater. Res. Soc. Symp. Proc., Vol 356, 1995, p 827–832

28. N.I. Tymiak, M. Li, A. Volinsky, Y. Katz, and W.W. Gerberich, The Role of Plasticity in Bi-material

Fracture with Ductile Interlayers, Metall. Trans. A, Vol 31, 2000, p 863–871

Adhesion Testing

Fundamental Aspects of Bimaterial Interfacial Adhesion

The practical work of adhesion, W

prac

, refers to the total energy necessary for separation along an interface

between two materials as affected by several energy absorption mechanisms. In terms of the critical strain

energy release rate, G

crit

, that is, elastic energy released per unit of fracture area:

prac

= G

crit

= Γ

+ Γ

+ other forms of dissipation

(Eq 4)

Here, Γ

is the work consumed by the fracture process per unit area of new surfaces. Interfacial fracture energy,

, accounts for both the thermodynamic work of adhesion, W

, and microstructural effects in the fracture

process zone such as crack bridging (Ref 29), interdiffusion, mechanical interlocking (Ref 30), and/or existing

defects along an interface. The thermodynamic work of adhesion is determined mainly by attractive interatomic

or intermolecular forces. The most common interfacial attractive forces result from van der Waals and Lewis

acid-base interactions (Ref 31). A contribution of electrostatic forces should also be considered (Ref 32).

The magnitude of W

is determined by the Dupré equation (Eq 3) (Ref 33) and is defined as:

= γ

+ γ

- γ

where γ

, γ

, and γ

are surface energies of materials along an interface and of the interface respectively. The

term Γ

in Eq 4 accounts for plastic energy dissipation rate at the interfacial crack tip. For interfaces involving

polymers, viscoelastic-plastic deformation may result in additional dissipational losses, Γ

. The stronger the

interface (higher Γ

), the more energy is dissipated through plastic and/or viscoelastic deformation (as well as

through any other dissipation process). Other factors affecting Γ

and Γ

are viscoelastic-plastic properties of

two materials forming an interface and geometry of a system. The latter implies crack configuration and

volumes of materials involved. For example, experimental results for ductile films on brittle substrates such as

copper on SiO

suggest plastic energy dissipation scaling with the copper film thickness (Ref 34). As a result,

the practical work of adhesion can increase by orders of magnitude when film thickness increases from the

nanometer to the micron scale. Among other mechanical properties determining Γ

, yield stress appears to be of

a critical concern due to its high sensitivity to microstructure, chemistry, and gradient constraints. Frictional

losses may be significant for rough interfaces with the effect increasing with the decreasing film thickness (Ref

35). For strong interfaces, additional dissipation mechanisms may be present. These include void nucleation

(Ref 36), second-phase debonding, microcracking, and shear band formation (Ref 37, 38).

Measurements of the practical work of adhesion can also be strongly dependent on the mode mixity or a

relative fraction of mode I and mode II. Mode mixity is determined by sample/applied load geometry and can

be defined as (Ref 39):

(Eq 5)

Generally, G

crit

increases as the crack tip becomes increasingly heavy under mode II conditions (Ref 40).

Possible mechanisms responsible for this trend include frictional losses (Ref 35) and/or plastic energy

dissipation rates (Ref 41, 42), which increase with increasing mode II contribution.

It is clear that even in the absence of an environment there is no single value of adhesion for a bimaterial

couple. The reasoning is that the practical work of adhesion depends on the volumes of materials involved as

well as the loading modes and the microstructural characteristics determined by a particular processing route.

References cited in this section

29. Y.Y. Wang, F.P. Chiang, and R.S. Barsoum, A Photoelastic Study of Extending Cracks in Adhesion

Joints, J. Adhes., 1995

30. A.G. Evans and J.W. Hutchinson, Effects of Non-Planarity on the Mixed Mode Fracture Resistance of

Bi-material Interfaces, Acta Met. Mater., Vol 37, 1989, p 909–916

31. F.M. Fowkes, D.W. Dwight, and D.A. Cole, Acid-Base Properties of Glass Surfaces, J. Non-Cryst.

Solids, Vol 120, 1990, p 47–60

32. J.T. Dickinson, L.C. Jensen, S. Lee, L. Scuderio, and S.C. Langford, Fracto-Emission and Electrical

Transients due to Interfacial Failure, J. Adhes. Sci. Technol., Vol 8, 1994, p 1285–1309

33. A. Dupre, Theorie Mechanique de la Chaleur, Paris, 1989

34. A. Bagchi and A.G. Evans, Measurements of the Debond Energy for Thin Metallization Lines on

Dielectrics, Thin Solid Films, Vol 286, 1996, p 203–212

35. R.G. Stringfellow and L.B. Freund, The Effect of Interfacial Friction on the Buckle-Driven Spontaneous

Delamination of a Compressed Thin Film, Int J. Solid Structures, Vol 30, 1993, p 1379–1395

36. A. Needleman, An Analysis of Tensile Decohesion along an Interface, J. Mech. Phys. Solids, Vol 38,

1990, p 289–324

37. W.D. Bascom, R.L. Cottington, R.C. Jones, and P. Peyster, Fracture of Epoxy and Elastomer Epoxy

Polymers in Bulk and as Adhesives, Organic Coat. Plast. Chem., Vol 34, 1974, p 300

38. J.N.B. Sutton, McGary, Polym. Eng. Sci., Vol 13, 1973, p 29

39. J.W. Hutchinson and Z. Suo, Adv. Appl. Mech., Vol 29, 1992, p 63

40. K.M. Liechti and Y.S. Chai, Biaxial Loading Experiments for Determining Interfacial Fracture

Toughness.J. Appl. Mech., Vol 58, 1991, p 680–687

41. V. Tvergaard V and J.W. Hutchinson, The Influence of Plasticity on Mixed Mode Interface Toughness,

J. Mech. Phys. Solids, Vol 41, 1993, p 1119–1135

42. P.A. Mataga and P. Ponte Castaneda, Stable Crack Growth along a Brittle/Ductile Interface-II Small

Scale Yielding Solutions and Interfacial Toughness, Int. J. Solids Struct., Vol 36 (No. 1), 1998, p 1–35

Adhesion Testing

Environment-Interface Interactions

An environmentally induced reduction in interfacial fracture energy may result from the following:

• Change in the thermodynamic work of adhesion

• Microstructural degradation such as microcracking

• Yield stress and/or viscoelastic property changes affecting dissipation losses

• Chemical or electrochemical reactions along an interface or in one or both materials, for example,

hydride formation

Depending on the bimaterial couple and the environment, any of the above may be prevalent.

Similarly, as in the case of bulk solids, environments may be considered as chemically reactive if their reaction

with either materials at the interface leads to dissolution or compound formation. Nonreactive environments

include media that influence mechanical behavior of one or both joining materials. The other part of

nonreactive media regarded as “inert” comprises environments that only may induce changes in surface

energies of joining materials (γ

and γ

) and that of an interface (γ

). Note that specific interfacial crack-tip

conditions may change environment/material interactions. Thus, any experimental results for a bulk

material/environment may not be directly applicable to the same material along an interfacial crack. For

example, varying ionic species concentrations unique to the crack tip region may adversely affect either side of

an interface. This could involve oxygen reduction along a metal/molymer interface leading to polymer

degradation. Another example is hydrogen evolution leading to metal embrittlement. It is also necessary to

emphasize that diffusion along an interface may be much higher than in any of the two joining materials. Thus,

after the same time of environmental exposure, the effect may be much stronger for a fracture along an

interface compared to fracture in bulk materials.

Adhesion Testing

Polymer/Inorganic Interfaces

Factors Determining Polymer/Inorganic Material Adhesion. The thermodynamic work of adhesion for polymers

is determined mainly by intermolecular bonds and Lewis acid-base (donor-acceptor) interactions (Ref 31). With

a significant difference between acidities of joining materials, donor-acceptor bonds would be prevalent

compared to intermolecular bonds (Ref 43). There is experimental evidence of adhesion strength decreasing

with decreasing difference in acidities of joining surfaces (Ref 31). Environmental exposure resulting in change

of surface acidity may affect adhesion strength. A so-called weak boundary layer may exist in some cases

determining strength of an interface (Ref 44). It should be cautioned that bonding may be distinctively different

for metal-on-polymer compared to polymer-on-metal for a given couple. This would imply different response

to environment/deformation exposure.

Environmental Effects on Different Contributions of Interfacial Toughness. The considerations for fracture in

bulk polymers are relevant for interfaces involving polymers as well.

Special attention should be given to:

• Specific interactions between a polymer and an environment within a specific highly stressed near-

crack-tip region

• Liquid transport toward a crack tip

• Temperature effects

• Rate dependence (loading and/or crack growth)

Environment/inorganic material interactions should be accounted for with specific attention to the

environmentally induced surface modification. These may range from change in surface acidity to formation of

a reaction product with the chemical and/or mechanical properties distinctively different from those of the

original material.

For a particular system, environmental effects on the dominant contribution of adhesion strength would

probably be most apparent. Depending on the structure of an interface and viscoelastic-plastic properties of

joining materials, the thermodynamic work of adhesion may be small compared to dissipation and

microstructural contributions. For several polymer-on-metal systems, adhesion strength is governed by

mechanical/microstructural factors. On the other hand, viscoelastic-plastic dissipation may be dominant for

strong interfaces involving polymers that exhibit viscoelastic flow. In some cases, it is possible to identify and

evaluate environmental effects on a specific mechanism of adhesion, as discussed in the next sections.

Thermodynamic Work of Adhesion. With the peel test, Carre and Shultz (Ref 22) were able to evaluate

environmentally induced relative change in the thermodynamic work of adhesion for rubber films on aluminum

tested in various alcohols. These results agree very well with the theoretical calculations.

Microstructural/Mechanical Effects. The strongest effects may be expected here from exposure to “reactive”

environments causing structural and/or chemical changes in one or both components of an interface. For

interfaces where a dominant fraction of adhesion strength comes from a mechanical interlocking between a

polymer and underlying metal oxide, environmentally induced changes in the chemistry and morphology of an

oxide may be critical. Examples include aluminum oxide conversion to a hydroxide that is less strong and

adheres poorly to the underlying metal (Ref 23). Similarly, adhesion to Ti/TiO

substrate may deteriorate with

titanium oxide undergoing a polymorphic transformation to anatase (Ref 23). Such a transformation is highly

temperature and moisture dependent. In other systems, such as epoxy-coated phosphated steel, exposure to

humid environments results in formation of a porous, moisture-retaining metal oxide layer (Ref 45). This

facilitates adhesion strength losses. It should be noted that reactions between a solid and an environment may

be strongly affected by the presence of a joining material. For example, galvanic coupling between an exposed

metal and an adjacent polymer-coated surface results in oxygen reduction along an intact interface (Ref 46).

This leads to the transformation of a near-interface layer of polymer into a gel-like structure with subsequent

gradual interface deterioration. In many cases involving degradation of a near-interface region in one of the

joining materials, a cohesive rather than adhesive fracture occurs.

Dissipation Losses. Environmental exposure may result in both direct and indirect effects on

plastic/viscoplastic dissipation. First, environmentally induced changes in viscoelastic plastic behavior would

influence dissipation directly. Second, degradation of the thermodynamic work of adhesion and/or

microstructural component of adhesion strength may result in lower dissipation contributions. Moisture-

affected viscoplastic deformation may be regarded as a major source of humidity-induced interfacial strength

degradation for a glass/pressure sensitive adhesive (PSA) interface (Ref 47). There was no apparent difference

between exposed and nonexposed samples prior to the “yield point” corresponding to fibril formation. In

contrast, the capacity for viscoplastic deformation associated with interfacial fracture appeared greatly

deteriorated in the presence of a humid environment. This resulted from local moisture absorption in the

vicinity of the crack tip rather than humidity-induced changes in the bulk viscoelastic behavior of the polymer.

Sometimes, moisture effects are complicated by the possible change of mechanisms at different humidity

levels. In general, moisture has been shown to decrease interfacial toughness for polyimide films on glass

substrates. For low humidity levels, adhesion strength decreased with the increasing moisture content.

However, for moisture content levels above 60%, adhesion strength increased with the further humidity

increase as shown in Fig. 3 (Ref 48). Here, formation of a boundary layer along an interface has been suggested

as the possible mechanism responsible for adhesion strength increase at relative humidities above 60%.

Fig. 3 Peel strength of 14.3 μm polyimide film with a peeling rate of 0.5 mm/min as a function of relative

humidity. Source: Ref 49

References cited in this section

22. A. Carre and J. Shultz, Polymer-Aluminum Adhesion. III. Effect of a Liquid Environment, J. Adhesion,

Vol 18, 1984, p 171–184

23. J.D. Venables, Review: Adhesion and Durability of Metal-Polymer Bonds, J. Mater. Sci., Vol 19, 1984,

p 2431–2453

31. F.M. Fowkes, D.W. Dwight, and D.A. Cole, Acid-Base Properties of Glass Surfaces, J. Non-Cryst.

Solids, Vol 120, 1990, p 47–60

43. K.W. Allen, A Review of Contemporary Views of Theories of Adhesion, J. Adhes., Vol 21, 1987, p

261–277

44. J.W. Severin, R. Hokke, H. Van der Wel, and G. de With, The Influence of Thermal Treatments on the

Adhesion of Electrolessly Deposited Ni(P) Layers on Alumina Ceramic, J. Electrochem. Soc., Vol 141,

1994, p 816–824

45. Y.M. Aravot and A. Albu-Yaron, The Effect of Environmental Moisture on the Adhesion Toughness of

Epoxy-Coated Phosphated Steels, J. Mater. Sci. Lett., Vol 12, 1993, p 1437–1438

46. A. Leng, H. Streckel, and M. Stratman, The Delamination of Polymeric Coatings from Steel, Parts 1–4,

Corros. Sci., Vol 41, 1999, p 547–620

47. H. Wu, J.T. Dickinson, and S.C. Langford, Dynamic Measurements of Humidity Attack on

Polymer/Glass Interfaces under Stress, J. Adhes. Sci. Technol., Vol 11, 1997, p 695–717

48. D.-C. Hu and H.-Ch. Chen, Humidity Effect on Polyimide Film Adhesion, J. Mater. Sci., Vol 27, 1992,

p 5262–5268

49. K. Suganuma, K. Niihara, T. Fujita, and T. Okamoto, Metal-Ceramic Joints, Vol 8, Proc. MRS Int.

Meeting (Pittsburgh, PA), Doyamma et al., Ed., Materials Research Society, 1989, p 113

Adhesion Testing

Interfaces between Inorganic Materials

As of the late 1990s, there appears to be only limited experimental data available on environmental degradation

of inorganic interfaces. As for polymer/inorganic interfaces, environment could affect the thermodynamic work

of adhesion as well as deteriorate an interfacial bond via reaction with one or both joining materials. Direct

environmental effects on the dissipation losses would not be considered here, except possibly for interfaces

where one material is susceptible to environmentally induced changes (e.g., via hydrogen charging) in plastic

behavior. In contrast, indirect changes in dissipation losses are possible via environmental effects on the

thermodynamic work of adhesion and/or microstructure.

Metal Films on Ceramic Substrates. As of the late 1990s, environmental effects for metal films on ceramic

substrates have only been evaluated for very few metal/ceramic systems. All of these studies examined

humidity or hydrogen effects on interfaces between metallic films and ceramic substrates. From these studies,

the condition of a ceramic surface appeared to be very important in determining susceptibility to environmental

cracking as shown by the Al/Si

(Ref 49) system tested in water. Presence of a damaged near-surface region

in Si

seemed to increase propensity for SCC. With such a layer, a crack propagated through this layer rather

than along an interface. Removal of the damaged material yielded interfacial strength approaching that for bulk

In contrast, with the deliberately patterned SiO

surface, Cu/SiO

interfaces exhibit much higher toughness in

both dry and humid environments (Ref 26) as shown in Fig. 4. The effect was attributed to crack bridging.

Indicating similarity of mechanisms, the form of velocity-energy release rate (V-G) dependencies was

essentially the same for smooth and patterned interfaces. Likewise for bulk materials, the kinetic curves for

interfacial crack growth exhibit three characteristic regions. The strong environmental effect on both threshold

(~20% reduction) and crack growth rates (more than 3 orders of magnitude increase) is evident from

comparison of V-G curves in wet and dry nitrogen. Enhanced crack growth rates in humid environments were

attributed to absorbed-water-induced weakening of interfacial bonds, which limits the extent of plastic

stretching of the bridging segments of copper film. Namely, indirect changes in plastic energy dissipation were

present there. It should be noted that observed interfacial crack velocities exhibited higher sensitivity to stress

intensity and were more than 3 orders of magnitude greater than crack velocities for bulk soda lime glass in

water vapor (Ref 50).

Fig. 4 Crack growth rates along plain and patterned glass/copper interfaces in wet and dry gaseous

nitrogen environments

Experimental results for gold films on Al

substrates tested in laboratory air and dry nitrogen indicate a

moisture-induced reduction in adhesion strength (Ref 25). Examination of fracture surfaces indicated separation

along an interface with no crack kinking into joining materials. With no apparent changes in any joining

material, observed adhesion strength reduction would be attributed to the environmental effects on the

interfacial fracture energy. Similarly, as in the previous example, a decrease of the interfacial bond strength

resulted in less plastic deformation of the metal.

Hydrogen effects have been examined for thin 200 to 1000 nm copper film on oxidized silicon with 10 nm

titanium interlayers between copper and SiO

(Ref 51). For both charged and noncharged films, the practical

work of adhesion (W

prac

) increased with the increasing film thickness, as shown in Fig. 5. Based on previous

studies (Ref 52), this trend was attributed to plastic energy dissipation scaling with the film thickness.

Hydrogen charging induced degradation of the practical work of adhesion. The possibility of a direct decrease

in Γ

has been eliminated as there was no change in the copper film yield stress following hydrogen charging.

Based on the W

prac

values and a theoretical model (Ref 53), a hydrogen-induced decrease of Γ

from 4 to 2 J/m

independent of film thickness, had been determined. An additional investigation would be required to determine

whether this change is attributed to the decreased thermodynamic work of adhesion, microstructural effects, or

both. Examination of chemical composition and topography of fracture surfaces would clearly be necessary

here. Also, to eliminate the possibility of a contribution from humidity, charging with gaseous hydrogen could

be utilized.

Fig. 5 Hydrogen effects on strain energy release rates for Cu/Ti/SiO

interfaces

Ceramic Films on Metallic Substrates. Studies of environmental stability of ceramic coatings on metallic

substrates have shown interfacial strength degradation during long-term aging without applied stress. For

example, a substantial adhesion loss was observed for flame sprayed Al

coatings on 316L stainless steel

aged in aerated Ringers solution (Ref 54). Ferber and Brown (Ref 55) evaluated alumina coatings on 316L

stainless steel and titanium alloy Ti-6Al-4V in distilled water under applied four-point bending. With the lower

“inert” strength, alumina/stainless-steel interface exhibited higher susceptibility to environmental stress

cracking. For both systems, the extent of slow crack growth was much greater than that reported for porous

alumina (Ref 56). With increasing temperature, adhesion strength degradation increased for both types of

coatings. Reaction between water and strained crack-tip bonds was suggested as a governing mechanism for

environmentally assisted interfacial crack growth. At ambient temperature, fracture was always 100% adhesive

for the alumina/stainless-steel interface. In contrast, for the alumina/titanium-alloy interface, fraction of

adhesive fracture decreased with increasing fatigue life. This indicated that slow crack growth in a coating

became the dominant mechanism at lower stress levels where longer environmental exposure was involved. For

alumina/stainless steel, an observed environmental effect may be attributed to interfacial strength degradation.

The situation is more complex for the alumina/titanium-alloy interface. Here, cohesive energy reduction for

alumina becomes more significant with increasing time and, eventually, becomes less than the adhesive energy.

For the weaker alumina/stainless-steel interface, the cohesive energy still exceeds the adhesive energy of an

interface.

Ceramic-Ceramic Interfaces. Similarly, as for metal/ceramic interfaces, only very few systems have been

evaluated. Only humidity or hydrogen effects have been investigated in these studies. Several studies (Ref 57,

58) address subcritical crack growth along SiO

/TiN interfaces exposed to aqueous environments. Here

environmental effects were evaluated both ex situ and in situ. For the ex situ experiments, crack growth

velocities have been found to be highly sensitive to humidity exposure time (Ref 57). With the above, it was

suggested that subcritical crack growth involved the diffusion of environmental species along the interface

ahead of a crack. In fact, ion mass spectroscopy (SIMS) measurements (Ref 59) revealed that moisture

diffusion along an interface was about 10

times faster than diffusion in the bulk SiO

. In situ experiments (Ref

58) were conducted under controlled humidity (40% RH) and temperature. It was suggested that mechanisms

responsible for SCC in bulk glasses may also be involved in the environmentally assisted interfacial fracture

(Ref 58).

Moody et al. (Ref 60) evaluated deuterium effects on adhesion between 600 nm thick tantalum nitride films and

sapphire. As revealed by a microscratch test, deuterium exposure resulted in an interfacial toughness decrease

from 24.5 to 9.1 MPa . At the same time, indentation testing indicated no changes in the mechanical

properties of TaN films. Thus, direct deuterium effects on plastic energy dissipation could be eliminated.

Therefore, decreasing interfacial toughness was attributed to an interfacial fracture energy decrease. Available

experimental evidence supported hydrogen-induced decohesion. However, deuterium could also affect adhesion

strength by affecting the interfacial microstructure. Thus, the necessity of a more precise microstructural

evaluation was suggested.

References cited in this section

25. S.X. Mao and A.G. Evans, The Influence of Blunting on Crack Growth at Oxide/Metal Interfaces, Acta

Mater., Vol 45, 1997, p 4263–4270

26. T.S. Oh, J. Rodel, R.M. Cannon, and R.O. Ritchie, Ceramic/Metal Interfacial Crack Growth

Toughening by Controlled Microcracks and Interfacial Geometries, Acta Metall., Vol 36, 1988, p 2083–

2093

49. K. Suganuma, K. Niihara, T. Fujita, and T. Okamoto, Metal-Ceramic Joints, Vol 8, Proc. MRS Int.

Meeting (Pittsburgh, PA), Doyamma et al., Ed., Materials Research Society, 1989, p 113

50. J.M. Howe, Bonding, Structure, and Properties of Metal/Ceramic Interface: Part 2, Interface Fracture

Behavior and Property Measurement, Int. Mater. Rev., Vol 38, 1993, p 257–271

51. N.I. Tymiak, M. Li, A.A. Volinsky, Y. Katz, and W.W. Gerberich, Environmental Effects on Cu/SiO

and Cu/Ti/SiO

Thin Film Adhesion, Reliability in Microelectronics MRS Proc., 1999

52. W.W. Gerberich, D.E. Kramer, N.I. Tymiak, A.A. Volinsky, D.F. Bahr, and M.D. Kriese,

Nanoindentation-Induced Defect-Interface Interactions: Phenomena Methods and Limitations Acta

Mater., Vol 47, 1999, p 4115–4123

53. Y. Katz, N.I. Tymiak, and W.W. Gerberich, Nano-mechanical Probes as New Approaches to

Hydrogen/Deformation Interaction Studies, An Invited Paper for Recent Advances in the Engineering

Aspects of Hydrogen Embrittlement, Eng. Fract. Mech., 2000

54. C.M. Baldwin and J.D. Mackenzie, Flame Sprayed Alumina on Stainless Steel for Possible Prosthetic

Application, J. Biomed. Mater. Res., Vol 10, 1976, p 445–453

55. M.K. Ferber and S.D. Brown, Delayed Failure of Plasma-Sprayed Al

Applied to Metallic Substrates,

J. Am. Ceram. Soc., Vol 64, 1981, p 737–741

56. D. Avigdor and S.D. Brown, Delayed Failure of a Porous Alumina, J. Am. Ceram. Soc., Vol 61, 1978, p

97–99

57. G. Xu, M.Y. He, and D.R. Clarke, The Effect of Moisture on the Fracture Energy of TiN/SiO

Interfaces in Multi-layer Thin Films, Acta Mater., Vol 47, 1999, p 4131–4141

58. M. Lane and R.H. Dauskardt, Subcritical Debonding of Multilayer Interconnect Structures: Temperature

and Humidity Effects, Reliability in Microelectronics, MRS Proc., Spring meeting, 1999

59. G. Xu, T. Mates, and D.R. Clarke, Moisture Diffusivity Measurement along SiO

/TiN Interface Using

Secondary Ion Mass Spectroscopy (SIMS), Reliability in Microelectronics, MRS Proc., 1999

60. N.R. Moody, S.K. Venkataraman, B. Bastaez, J.E. Angelo, and W.W. Gerberich, Hydrogen Effects on

the Fracture of Thin Tantalum Nitride Films, Mater. Res. Soc. Symp. Proc., Vol 356, 1995, p 827–832

Adhesion Testing

Testing Methods

Testing techniques for in situ environmental testing of thin-film adhesion are briefly reviewed. The focus is on

quantitative methods where a characteristic of interfacial toughness is extracted from experimental results

utilizing an appropriate model. In most cases, modeling is based on elasticity solutions. It is necessary to

emphasize that these are applicable only under small-scale-yielding crack-tip conditions with the remaining

material being elastic. However, extensive plastic deformation or even general yielding may develop in a

ductile counterpart of an interface. Lai and Dillard (Ref 61) determined general yielding conditions for the most

common test configurations and suggest appropriate types of test and sample dimensions for interfaces with

different elastic mismatch and residual stresses. Even under conditions where general yielding is not expected,

the local crack-tip region may experience extensive plastic deformation and associated dissipation losses.

Several studies evaluate plastic energy dissipation associated with interfacial fracture (Ref 62, 63, 64). Results

of these studies may be applied to practical specimen configurations that approximate situations treated in the

above analyses. As for specific testing methods, either finite element analysis (FEM) analysis (Ref 65) or

analytical approximation may be available (Ref 66) in some cases.

Methods suitable for in situ environmental testing of thin-film adhesion are briefly described below and are

summarized in Tables 1 and 2 (Ref 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84,

85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95). For ease of evaluation and to obtain useful quantitative data, the

following criteria are suggested for these tests:

• Simple sample preparation with no or minimum effect on the properties of interfaces and/or joining

materials

• A simple testing procedure with a possibility for in situ environmental exposure and/or testing shortly

after an environmental exposure

• Precision and reliability in measurements of loading and crack growth characteristics

• Availability of an adequate model for a quantitative analysis. Ideally, not only a characteristic value of

interfacial toughness should be determined, but also corresponding mode mixity and possible

contributions of various dissipation processes

• Close approximation of service conditions (geometry and environment)

Examination of fracture surfaces may also provide useful information.

Table 1 Summary of adhesion testing methods