Adhesion in Microelectronics

Preface

The phenomenon of adhesion is of cardinal importance in variegated ways in the domain of microelectronics. A few eclectic examples will suffice to underscore the importance of adhesion in this industry. Adequate adhesion of thin films is sine qua non in integrated circuits; requisite adhesion of polymer and other materials is a must in packaged devices; adhesion of various materials is crucial in fabricating printed circuit boards. Many failures in the microelectronics industry can be traced to lack of proper adhesion or are imputed to sub-optimum interfaces between different materials used. This necessitates study and characterization of interfaces/interphases, devising ways to modify surfaces of materials to attain the desired level of adhesion between the mating partners; use of bonding agents and adhesion promoters. Also, the need for reliable methods of adhesion measurement is quite patent. Moreover, the reliability aspects of components and devices in the microelectronics industry is of grave concern.

Although there has been a high tempo of R&D activity, the information is scattered in a number of publication media. There is no single easily accessible source where one can find the requisite information on adhesion in microelectronics. This lacuna in the literature provided vindication for this book, which we felt was both timely and needed.

This book containing 8 chapters by subject matter experts is divided into three parts: Part 1: Adhesion: Fundamentals and Measurement; Part 2: Ways to Promote/Enhance Adhesion; and Part 3: Reliability and Failure Mechanisms. The topics covered include: application of inelastic electron tunneling spectroscopy (IETS) in understanding fundamental nature of bonding (adhesion); a number of techniques (qualitative, semi-quantitative, quantitative) for adhesion measurement of thin films and coatings; tailoring of interfaces/ interphases to promote adhesion of metal layers on polymers; application of plasma (both vacuum and atmospheric) treatments of polymer surfaces for enhanced adhesion to other materials; isotropic conductive adhesives in electronic packaging applications; role of adhesion phenomena in the reliability of electronic packaging; delamination and reliability issues in packaged devices; and mechanisms of adhesion and failure in microelectronic packages, especially at the molding compound/substrate interface.

Individuals in academia carrying out research in understanding and unraveling the fundamental aspects of adhesion of similar or dissimilar materials and those involved/interested in various aspects of adhesion in microelectronics should find this book of extreme interest. The information contained in this book should be of immense appeal to R&D, manufacturing and quality control personnel. In essence, anyone interested (peripherally or centrally) in improving adhesion in various microelectronic components and devices should find this treatise of great value. It should serve as a gateway for neophytes and a commentary on recent developments for the seasoned researcher.

This book should also be of interest to those working in other industries where understanding and control of adhesion is of paramount importance, such as thin film technology, optics, packaging of all sorts of products, adhesive bonding, aerospace, metallized plastics, encapsulation.

Acknowledgements

Now comes the pleasant task of thanking those who were instrumental in giving this book a body form. First of all, we are beholden to the authors for their interest, enthusiasm, cooperation and contributing their chapters, without which this book would not have seen the light of day. Also we very much appreciate the steadfast interest and unwavering support of Martin Scrivener (publisher) in this book project.

Kash Mittal

P.O.Box 1280

Hopewell Jct., NY 12533

E-mail: [email protected]

Tanweer Ahsan

Henkel Electronic Materials LLC

1400 Jamboree Road

Irvine, CA 92606

E-mail: [email protected]

April 14, 2014

Part 1

ADHESION: FUNDAMENTALS AND MEASUREMENT

Chapter 1

Study of Molecular Bonding or Adhesion by Inelastic Electron Tunneling Spectroscopy, with Special Reference to Microelectronics

Robert R. Mallik

Department of Physics, The University of Akron, Akron, Ohio USA

E-mail: [email protected]

Abstract

This chapter presents an outline of the principles, methods, applications, and scope of Inelastic Electron Tunneling Spectroscopy (IETS) with emphasis placed on the study of molecular adsorption on metal oxide and semiconductor surfaces. Strengths and limitations of the technique are highlighted, with particular attention being paid to applications in adhesive systems comprised of materials pertinent to microelectronics device fabrication including epoxy resins, polyimides, and silanes. A brief description of how IETS may be used to investigate adsorption and conduction mechanisms for self-assembled monolayers of molecules adsorbed on photovoltaic semiconductor materials is given as a segue into an examination of how IETS and related techniques are being developed for the study of molecules of interest in the rapidly developing field of molecular electronics.

Keywords: Adhesion, microelectronics, molecular electronics, photovoltaics, thin-films, vibrational spectroscopy, IETS.

1.1 Introduction

Inelastic Electron Tunneling Spectroscopy (IETS) is a relatively new technique in the toolbox of surface scientists. It was discovered by Jaklevic and Lamb in 1966 while investigating the superconducting bandgap of lead in metal/insulator/superconductor tunnel junctions [1]. Specifically, these workers were recording current-voltage (I-V) curves of aluminum/ aluminum oxide/lead tunnel junctions at a temperature of 4.2 K by immersing them in liquid helium. The first metal electrode of an IET junction is usually referred to as the base electrode and the (usually) superconducting top electrode is the cover electrode. In tunneling experiments, derivatives of I-V curves (i.e., plots of quantities proportional to dI/dV and d²I/dV² with respect to bias voltage) are often recorded to reveal superconducting structure more clearly. Jaklevic and Lamb noticed that, in addition to low-bias structure present due to the superconducting energy gap, associated with the lead cover electrode (which appears at bias voltages V = ±Δ / e ≈ ±2.15mV where, e, is the electronic charge), additional fine structure was evident at higher bias voltages. This additional structure was best revealed as peaks in the second derivative plots, and the peaks appeared at bias voltages in the range of vibrational modes of molecules (approximately 0–500 mV or, equivalently, 0–4000 cm−1). It transpired that the peaks were due to the presence of minute amounts of pump oil which had inadvertently been adsorbed onto the aluminum oxide surface of the tunnel junction base electrode. The peak energies corresponded closely to IR vibrational mode energies of the hydrocarbons present in the pump oil and recognition of this fact led to the birth of IETS. Since then, IETS has been used to study a wide variety of adsorbates on metal oxide and semiconductor surfaces. The purpose of this chapter is to briefly highlight the principles of IETS in order to illustrate the strengths and limitations of the technique with particular attention given to applications relevant to adhesive bonding, particularly in the area of microelectronics and its potential in the developing area of molecular electronics. For readers wishing to learn more about the theory, experimental procedures, and scope of IETS several books on the technique are available [2] but, as a starting point, the reader is referred to the excellent review article by Hipps and Mazur [3].

IETS is a technique particularly useful for the investigation of adsorption and conduction mechanisms of ultra-thin layers. While advances in more widely used and well-established surface vibrational spectroscopies, for example multiple reflection/absorption methods in IR [4] and surface enhancement effects in Raman [5, 6], have allowed for surface-specific measurements on a variety of systems [7], the sensitivity of both of these techniques nevertheless decreases with sample layer thickness. In contrast to this, IETS sensitivity actually increases as the layer thickness decreases and becomes optimal for adsorbed layers close to monolayer coverage. The reason for this is that whilst IR and Raman require greater sample volume for increased interactions between the sample and excitation energy source (photons), quantum tunneling by its very nature is intrinsically sensitive to thin layers at length scales corresponding to nanometer thickness. This is because the wavefunction of the tunneling electrons (the excitation source) is greatly attenuated when traversing the thin layer. Indeed, the probability of electrons tunneling through a thin layer, which constitutes a potential barrier, decreases exponentially with the height and width of the barrier. For a rectangular barrier of width, d, and height, φ, the electron tunneling probability, P, is given approximately [8] by the expression P = exp(−2αd), where α² ≈ 2me φ/ħ² (where me is the electronic mass and ħ = h/2π, h being Planck’s constant). This highly sensitive dependence of the tunneling probability on barrier height and thickness is why IETS is ideally suited to probe species at, or in close proximity to, the surfaces of the tunnel barrier. Another important feature of IETS alluded to above is that, since the technique is based on measurement of thin layers I-V curves, information can be extracted regarding conduction mechanisms through said films. This information is not available via IR and Raman spectroscopies. In practice, information from all three techniques is mutually valuable. IETS can provide complimentary information, inaccessible via the other two techniques, which may lead to a more thorough characterization of the surfaces and interfaces of the system under investigation than by the use of any one of these techniques in isolation [9].

The potential of IETS in various, and seemingly diverse, research areas such as surface chemistry, heterogeneous catalysis, analytical chemistry, environmental pollution monitoring, adhesion science, radiation damage, biological chemistry and electronic energy level studies was recognized early in the development of the technique [10]. Of particular interest for the present chapter, as will be shown, is early work illustrating the usefulness of IETS in the study of numerous adhesive systems, for example silane coupling agents on alumina [11, 12, 13, 14, 15, 16, 17], and other adhesive, or adhesive-related, systems on alumina including epoxides [18, 19, 20], polymers [21, 22], and phenolics [23]. Since aluminum is the base electrode of choice for most IETS work, and because aluminum is a widely used material for adhesive bonding in aerospace, automotive, and packaging applications, it is perhaps not surprising that IETS was deemed particularly appealing for such adhesion studies. Since these early studies, it should be noted that IETS has also been applied to other adhesive systems for example systems of adhesion promoters on glassy substrates [24]. More recently, IETS has been used to study adsorption on materials other than the native oxide of the base electrode metal, i.e., on so-called artificial tunnel barriers. In 1989, Barner and Ruggiero reported the material and electron tunneling properties of thin radio-frequency magnetron sputtered alumina films supported on copper base electrodes [25]. Around the same time, Mazur and Cleary demonstrated the potential utility of an aluminum nitride artificial barrier deposited by reactive ion-beam sputtering [26], and two years later, it was reported that sputtered amorphous silica formed a viable artificial tunnel barrier when deposited onto gold base electrodes [27]. These and subsequent studies on artificial tunnel barriers were significant in that they demonstrated that the technique of IETS could be extended to investigate a wider range of systems than just those involving adsorption studies on alumina. IETS has now been successfully performed on other semiconductor barriers [28, 29], including photovoltaics [30], and it has also been used to the study molecular adsorption on photovoltaics [31] illustrating how the technique may be applied to materials of interest in the area of microelectronics.

1.2 Principles of IETS

1.2.1 General Overview

As outlined in the Introduction, IETS relies on the quantum mechanical phenomenon of electron tunneling between two metal electrodes through a sufficiently thin potential barrier. It allows one to measure the vibrational energies of molecular species constituting the barrier when excited by these tunneling electrons. Barriers are incorporated in metal/insulator/metal tunnel junctions fabricated sequentially on insulating substrates by conventional vacuum deposition techniques. They must be uniformly thin (of the order of 2–3 nm) and continuous if sufficient tunnel current is to flow. Monolayers of compounds of interest may be introduced onto the barriers if desired. Resulting IET spectra yield information regarding the nature of molecular bonding at the interface so formed. IR, Raman, and other modes in the barrier and metal electrodes are detectable, and IET peak intensities may be correlated to surface coverage, bond angles, and the location of bonds within the tunnel barrier [2, 32].

A description of how IETS is applied in practice, describing the above items in more detail, is given in sections 1.2.2, 1.2.3 and 1.2.4 below.

1.2.2 Key Principles of Operation

Tunneling is induced by the application of a small dc bias voltage, V, between the two metals which offsets their Fermi energies, EF1 and EF2, by an amount, eV, as indicated in Figure 1.1(a).

Figure 1.1 (a) Schematic diagram of a metal/insulator/metal IET junction. Hatched regions indicate filled electron energy states. The insulating barrier, I, may be the native oxide of the base electrode (metal 1), or an artificial barrier deposited onto metal 1 as described in section 2.4.2 of the text. Molecular species may be introduced onto the barrier if desired to create a composite tunnel barrier. (b) The effect of the bias voltage is to offset the Fermi energies of metal 1 and metal 2, EF1 and EF2 respectively, by an amount eV, causing electrons to tunnel from filled energy states in metal 1 through the insulating barrier to empty states in metal 2. Almost all electrons (approximately 99%) tunnel elastically, but a small fraction does so inelastically by interacting with vibrational modes in the barrier.

Electrons tunnel from filled states close to the Fermi energy of one metal to empty ones in the other, and most do so by traversing the barrier elastically i.e., with no loss of energy; but a small fraction (<1%) does so inelastically by transferring energy to species within the barrier (see Figure 1.1(b)). The I-V relationship of an IET junction solely due to elastic tunneling is nominally linear over the voltage range of interest (0 to 500 mV), however, the onset of inelastic events produces small increases in current at specific threshold voltages corresponding to the energies of vibrational modes of barrier species, the net effect on the I-V curve being corresponding increases in slope at these threshold voltages as shown in Figure 1.2(a).

Figure 1.2 (a) Idealized I-V curves of an IET junction at zero kelvin (solid line) and at the usual experimental operating temperature, typically 4.2 kelvin (dashed line). A small increase in slope is shown corresponding to loss of energy of tunneling electrons to a particular vibrational mode in the tunnel barrier at a threshold bias voltage, V = hν/e. The slope change is actually of order 1%, but has been exaggerated here for clarity. (b) Slope increases in the I-V curve are more easily revealed as steps in the first derivative plot. (c) IET spectra are plots of quantities proportional to the second derivative, and a series of peaks are observed at threshold voltages identified with the vibrational modes of the barrier material. The figure shows a single IET peak (dashed line) which is broadened due to thermal smearing of the electron energies. Actual IET peaks are broadened further by instrumental effects resulting from measurement techniques as described in section 1.2.3.

The threshold voltages are related to the vibrational frequency, ν, of a particular mode through the relationship eV = hν. An IET spectrum is a plot of the I-V curve’s second derivative (recovered either numerically or electronically, see section 1.2.3 below) as a function of bias voltage and reveals inelastic scattering events as peaks. IETS samples must be cryogenically cooled to minimize thermal smearing of electron energies which is responsible for broadening of IETS peaks’ full-width-at-half-maximum (FWHM) by approximately 5.4 kT where k is Boltzmann’s constant and T the temperature in kelvin. Cooling may be conveniently achieved by submerging the samples in liquid helium (whose boiling point at atmospheric pressure is 4.2 K). The thermal broadening component of the peaks’ FWHM at 4.2 K is of the order of 1 meV which is commensurate with the natural line width of most vibrational modes. For comparisons to IR data note that 1 meV is equivalent to 8.065 cm−1.

Some important factors which affect IETS peak intensities must be considered when interpreting IET spectra. Firstly, as is the case for all vibrational spectroscopies, peak intensities correlate with the number of (dipole) bonds excited. Secondly, IETS peak intensities also depend on the orientation of the bonds with respect to the plane of the tunnel junction. In general, IETS peaks are stronger for bonds whose axes are aligned perpendicular to the plane of the junction and weaker for bonds whose axes lie parallel to it. This is because electrons tunneling from one metal electrode to the other through the barrier couple most strongly with bonds that are coaxial with their tunneling path. Thirdly, IETS peak intensities correlate with a bond’s location within the tunnel barrier. This is a consequence of the localized nature of the electron-mode interaction in the tunneling process and manifests itself as a bias polarity dependent peak intensity asymmetry. Forward bias in IETS is defined as the case where the cover electrode is positive with respect to the base electrode, therefore electrons tunnel from the base electrode to the cover electrode, while reverse bias is defined in the opposite sense. If one assumes that a particular mode is, for example, situated at the cover electrode side of the junction it is more probable that an electron will tunnel elastically with higher energy from the base electrode through the barrier towards cover electrode and then inelastically excite the mode (i.e., under the forward bias condition) rather than first interacting inelastically with the mode and then tunneling elastically through the remainder of the barrier towards the base electrode with a lower energy (reverse bias). Therefore, the peak associated with the mode located at the cover electrode side of the junction would appear stronger under forward bias and weaker under reverse bias. A schematic depiction of the two processes is given in Figure 1.3.

Figure 1.3 (a) Energy diagram for an IET junction with zero applied bias. A rectangular barrier of height, φ, is a reasonable approximation to a real barrier. Located towards the metal 2 side of the barrier is a moiety with vibrational mode energy, hv, indicated by a filled circle. (b) Under the forward bias condition, electrons first tunnel elastically through the barrier, then excite the moiety through an inelastic interaction. After losing energy to the moiety, the electrons emerge on the other side of the barrier and occupy the lowest available empty energy states. (c) Under reverse bias, electrons interact inelastically first with the moiety before tunneling through the remainder of the barrier with a lower energy. Since tunneling probability decreases with electron energy, the process shown in this figure is, on average, less likely than that of Figure 1.3(b). The effect of this phenomenon is that the IETS peak associated with the moiety under reverse bias is weaker than for the forward bias case.

This phenomenon has been invoked to isolate the position of Si-H moieties in sputtered amorphous SiO films. It was demonstrated that the Si-H moieties were created on the SiO surface during the fabrication process and could be subsequently removed by exposure to a plasma discharge in vacuo [33].

1.2.3 IET Spectrometer Design and Implementation

At present, no commercial IET spectrometers are available so in-house built instruments are used in research laboratories, several of which are described in the literature [3]. They employ a widely used spectroscopic method for small signal recovery namely modulation of the input signal at a fixed frequency, ω, coupled with phase-sensitive detection of output signal harmonics generated due to non-linear sample responses [34]. Figure 1.4 shows the typical IET spectrometer design. The spectrometer applies a small modulation current, Iω, superimposed upon a slowly ramped dc bias voltage, V, across the IET junction. Taylor series analysis shows that the second harmonic voltage response, V2ω, developed across the junction is proportional to the quantity d²I/dV². (The magnitude of V2ω depends on that of the slope changes of the junction’s I-V curve and Iω). V2ω is recovered experimentally by a lock-in amplifier.

Figure 1.4 Schematic diagram of a constant resolution IET spectrometer. A digital-to-analog converter (DAC) supplies a ramped dc bias voltage, V, while an oscillator provides a small ac modulation current Iω. A mixer circuit isolates the dc and ac power sources, combines the dc bias voltage and ac modulation current, and applies them simultaneously to the junction. In response to the applied modulation current, an ac modulation voltage, Vω, and corresponding second harmonic, V2ω, develops across the junction. A lock-in amplifier (LIA) is used to recover the second harmonic voltage, while the dc bias voltage is measured by a digital multi-meter (DMM). In order to maintain constant resolution across the entire dc bias range, a regulator circuit monitors Vω across the junction in a feedback loop and adjusts the oscillator current output as necessary such that Vω remains essentially constant. Commercial software packages are normally used to control the various instruments in the spectrometer via a GPIB interface and plot the resulting IET spectra (i.e, V2ω versus bias voltage).

Since the dc bias across the junction is modulated by a corresponding amount, Vω, i.e., the modulation voltage, IET peaks are broadened by the same amount which adds to the thermal broadening described in section 2.2 above. Typically, modulation voltages of order 1 mV are employed in IETS such that thermal and modulation broadening components are of approximately the same magnitude at 4.2 K since the FWHM of IETS peaks is given by FWHM = . Some IET junctions, for example those with semiconducting or low bandgap insulator barriers, may exhibit highly non-linear I-V curves, particularly at high bias voltages. For such junctions the modulation voltage, Vω, developed across the junction decreases significantly with increasing bias which, in turn, leads to a drop-off in the second harmonic signal response V2ω. To compensate for this drop-off, elegant circuit designs have been implemented to monitor the modulation signal and boost it as necessary so as to maintain an essentially constant value over the entire spectral range. Spectrometers of this type are referred to as Normal Tunneling Intensity (NTI) or constant modulation spectrometers [35, 36].

1.2.4 IET Sample Preparation

Many reviews of IETS are available [2, 3, 8, 10] which describe fully the metal/tunnel barrier/metal sample fabrication process so only brief description is given here paying particular attention to samples with semiconducting barriers of potential importance in the area of microelectronics and photovoltaics. Sample preparation consists of sequential deposition of the metal base electrode, tunnel barrier, and cover electrode.

1.2.4.1 Electrode Deposition

Base and cover electrodes are nearly always deposited by thermal evaporation onto glass microscope slides through shadow masks to define the electrode geometry. Resistively heated evaporation sources are employed and the procedure is performed in a conventional high vacuum chamber at pressures of the order of 10−5 to 10−7 Torr. The IETS base and cover electrodes are usually aluminium and lead, respectively.

1.2.4.2 Barrier Preparation

Tunnel barriers may be formed in a variety of ways. Native oxides of the base electrode metal are by far the most common barriers and are formed either by exposing the electrodes to a low-pressure oxygen plasma (of order 100 mTorr) while in the vacuum chamber to form a plasma oxide or simply by venting the chamber to atmosphere to form a thermal oxide. In both cases, the oxides created are typically 2–3 nm thick. Artificial barriers (insulators or semiconductors) may also be deposited directly onto the (un-oxidized) base metal electrodes by radio-frequency magnetron sputtering in an inert (e.g. argon) or reactive (e.g., oxygen) plasma. Such artificial barriers must be formed under low-power conditions to ensure a sufficiently low deposition rate which is vital to ensure that the barriers are sufficiently thin and continuous. Typical powers and deposition rates are approximately 5 W, and 0.01 nm/s respectively [27]. Figure 1.5 shows an AFM image of a CdTe film deposited in this way.

Figure 1.5 AFM image of an ultra-thin film CdTe artificial tunnel barrier prepared by radio-frequency magnetron sputtering. X and y length scales are the same. To simulate experimental conditions in IETS, the film was grown on an underlying aluminum thin film (approximately 300 nm thick) vacuum evaporated previously onto a clean glass microscope slide consistent with standard practice for IETS base electrodes as described in section 1.2.4. The CdTe film was deposited in a 50 mTorr background of Ar, by employing a very low power and deposition rate (~5W rms, and ~0.02 nm/s respectively). These conditions are required such that extremely small clusters of the CdTe target material are ejected during the sputtering process to grow a film which is sufficiently uniform and thin enough for IETS purposes. Films of this type display a granular structure; this particular film has an rms roughness of 0.82 nm, and is 16.5 nm thick.

If desired, molecular species may be deposited onto the barriers. This is normally achieved by spin coating the barrier with a very dilute solution of the compound in question (typical concentrations are of the order 0.1% w/v) or by exposing the barrier to a vapour of the compound preferably in a fume hood. The former method is normally referred to as liquid phase doping and the latter vapor phase doping. For both, the solution concentration and exposure times are varied by trial-and-error with the goal of achieving near-monolayer coverage of adsorbate.

1.3 Application of IETS in Microelectronics

Adhesive bonding plays a major role in the manufacture of microelectronic devices, assemblies, and packaging as described in a thorough review article by Yacobi and co-workers [37]. During the device fabrication process, it is often necessary to bond dissimilar materials such as metals, semiconductors, and polymers. The goal, when bonding such materials, is to minimize any possible adverse effects that may arise due to the bonding process, for example the creation of stresses which may lead to a reduction in the structural integrity of the bonded layers which may cause unwanted changes in the electrical properties of the microelectronic devices. Judicious choice of the adhesive system to be employed for the particular materials in the microelectronic device in question is essential for achieving this goal. A pivotal factor in choosing an adhesive system is a detailed knowledge of the physical and chemical adhesion mechanisms at the molecular level which can be provided by surface-sensitive spectroscopic methods.

As mentioned in Section 1.1 above, IETS offers unique capabilities and features which can complement Reflection Absorption IR Spectroscopy (RAIRS) and Raman data. This has practical importance in the study of adhesion at the molecular level. Most often, IETS data are compared with IR, but care must be taken when interpreting the differences. Sondag and coworkers studied monolayers of several aromatic and non-aromatic carboxylic acids chemisorbed on alumina and found that intrinsic differences in sensitivity exist between IETS and conventional RAIRS [38]. They observed that CH vibrations are stronger in IETS while vibrations involving carbon and oxygen atoms, and particularly carbonyl modes, are stronger in RAIRS. The reasons for the differences were not clear but they pointed out that the selection rules were well-established for IR but not so for IETS. In order to elucidate these observations, Devdas and Mallik performed a series of investigations to study specifically chosen carboxylic acids chemisorbed on alumina using IETS and RAIRS in tandem. To ensure similar molecular environments, they used similarly prepared samples with very thin lead cover films for both spectroscopies (normally, RAIRS samples do not have a lead cover film). They found that the cover film, which was sufficiently thin such that it transmitted IR radiation, had a significant effect on carbonyl mode intensities. In fact, they observed that the proximity of the cover film [39], location of particular bonds within molecules [40] and fractional surface coverage of the molecules [41] all have an effect on peak intensities in both IETS and RAIRS. To illustrate this phenomenon, Figure 1.6 shows a comparison of the IET and RAIR spectra for 5-oxoazelaic acid adsorbed on alumina with a surface coverage close to one monolayer.

Figure 1.6 (a) RAIR spectrum of 5-oxoazelaic acid (molecular structure as inset) spin-coated from a dilute solution in ethanol onto aluminum oxide. Carboxylic acids are known to chemisorb on alumina via acid-base reactions of the carboxylic acid group with the amphoteric surface. For the solution concentration used here, the surface coverage of 5-oxoazelaic acid is nominally one monolayer and the molecules are closely packed and adsorbed on the surface via one of the carboxylic acid groups. The strong peak at 1736 cm−1 is due to unreacted carboxylic acid carbonyls and the shoulder at 1703 cm−1 is assigned to the 5-oxo-substituted carbonyls. (b) RAIR spectrum recorded for a sample prepared identically to the one of Figure 6(a) except that it was capped by a very thin (~20 nm) lead film evaporated on top of the 5-oxoazelaic acid monolayer in order to simulate conditions used in IETS. The effect of the lead film is to suppress the unreacted carboxylic acid carbonyl peak, while the 5-oxo-substituted carbonyl peak is essentially unaffected and appears at 1705 cm−1. (c) IET spectrum recorded for a sample prepared identically to that of Figure 6(b) except a thicker lead film was deposited (~100nm). The 5-oxo-substituted carbonyl peak appears at 1704 cm−1. Note that hydrocarbon stretching modes (ca ~2900 cm−1) are much stronger in the IET spectrum than in RAIR spectrum as is generally the case for most compounds.

As can be seen, if a lead cover film is deposited onto the adsorbed monolayer to cap the structure, vibrations of the unreacted carboxylic acid carbonyl group in the molecule are suppressed in both IET and RAIR spectra. The findings taken as a whole illustrate the benefit of using both IETS and RAIRS in conjunction when investigating adsorbed monolayers which should be taken into consideration when interpreting data for systems of adhesives and in general.

IETS has been used to study several adhesive-related systems including phenolic adhesives [23], phosphorous acids [42] and silane coupling agents [11–17] all of which are important for improving adhesion for components in printed circuit board manufacture, specifically those related to encapsulation and packaging purposes.

It is well known that insulating polymer films are of significant technological importance in the fabrication of microelectronic devices. IETS has enjoyed success for some years in the investigation of adhesives [19, 43] and may be applied to adhesive systems of interest in microelectronics. The primary reason for this is that IETS is well suited to systems consisting of thin insulating layers (often polymeric in nature) adsorbed on metal oxides. For example, one family of compounds, epoxy resins, which is widely used in microelectronics for bonding, packaging, and dielectrics layers in printed circuit board assemblies has been studied by IETS. Figure 1.7 shows the IET spectrum of a model compound synthesized to simulate an amine cured epoxy resin obtained by Affrossman and coworkers [44]. The compound (the adduct of reaction of the diglycidylether of bisphenol A with excess diethylamine) was used in order to avoid difficulties associated with preparing ultra-thin resin layers necessary for IETS, or extrapolating information from the study of separate components of cured resins as the compound itself already has a structure similar to a fragment of cured resin. The compound was introduced onto alumina by spin-coating from a dilute solution in benzene followed by repeated degassing under