Measurements in thin films

There are several methods that have been implemented for measurements of thin film piezoelectric coefficients. In looking at the relative accuracy of the techniques, it is important to consider the differences between thin film and bulk samples. Firstly, thin films are, by definition, thin. Since the maximum strain that can typically be generated is 1 %, the film displacements, themselves are quite small. The second, rather obvious, difference is that thin films are not amenable to fabrication of unsupported samples. A limited amount of information is available on the piezoelectric coefficients of thin films released by dissolving away the 

Far more frequently, however, measurements are made with a relatively thin film (often 5 /rm) on a comparatively massive substrate (often 400 //,m in thickness). In principle, the piezoelectric and electromechanical properties of the films can be extracted by fitting conventional resonance measurements. In practice, multiple resonances due to the film/substrate system are measured it is then necessary to deconvolute the data in order to extract information on the films itself. In practice, this is difficult to do unambiguously [21,22], 

As a result of the presence of the supporting substrate, the piezoelectric coefficients reported for films are usually effective numbers, rather than the free piezoelectric coefficients typically reported for bulk samples [23], The two coefficients most widely measured are the effective d j coefficient and the e j coefficients given in Equation (2.11). 

Here the subscript / denotes the property measured for the film. It is clear from an inspection of the equations that d33i/ is lower than the value which would be obtained from an unconstrained d33, due to the substrate clamping. To illustrate this, consider a thin film electroded 

In many microelectromechanical systems (mems) based on piezoelectric thin films, flexure is deliberately used to amplify the available displacements (or alternatively to increase the sensitivity of a sensor). For simplicity (and to keep poling and actuation voltages low), films are often poled and driven by electrodes at the top and bottom surfaces. As a result, the critical piezoelectric coefficient is often e31 j, rather than d33j [24], For the direct effect, the effective film coefficient, e3ij can be defined by 

---

For the usual dc measurement the constant dc current source should be capable of providing currents in the range 0.1-10 mA for a typical bar of 1 mm square cross-section, 1 cm length, and a resistivity at 100 K of 50 pOhm-cm the voltage measured for a 1 mA current source would be 1 / V. Since even for a typical low value of the critical current density, 100 A/cm2, the measurement current would be 1000 times less and thus have essentially no effect on the measurement. However, the measurement of 1 / V to a precision of 1% already requires care to assure that noise and thermal voltages are reduced well below this value. Currents of similar value are used for measurements in thin films. 

LIG 94] LIGEN Y., HAILIN S., KEWEI X., JIAWEN H., A correction of the Seemann Bohlin method for stress measurements in thin films , J. Appl. Cryst, vol. 27, p. 863-867,1994. 

There are no experimental gas-phase or thin-film photoelectron spectra (PES) for these two carbon nanostructures. Ultraviolet PES measured in thin films of C% were published [44, 45]. No absolute values were given for the peak positions, instead, the spectra were depicted over the energy scale relative to the Fermi energy level. The low-energy part of this spectrum contains an intense peak followed by a jagged plateau of lower intensity. At least three small peaks are visible on that plateau. The energy difference between the main peak position and the first small peak on the plateau is about 0.5 eV. 

Wang X., Xu L., Li D., Liu L., Wang W. Themio-optic properties of sol-gel-fabricated organic-inorganic hybrid waveguides. J. Appl. Phys. 2003 94 4228-4230 Weber H.P., Dunn F.A., Leibolt W.N. Loss measurements in thin-film optical waveguides. Appl. Opt. 1973 12 755-757... 

From the comparison of UPS measurements in thin films and in the gas phase of monothiophene it can be concluded that the molecular properties are conserved in the solid state. There are several calculations (and UPS studies) on the molecular orbitals of these systems, which come to similar results [219-224]. Ab initio as well as semi-empirical calculations attribute the HOMO of the monomer to the (la2) TT-orbital with contributions from the carbon Pz orbitals. This 7r-orbital splits into n orbitals for a-nT and forms a wide 7r-band for poly thiophene. The HOMO-1 (3bi) orbital of the monomer has no Pz contributions and forms in an-T a non-bonding narrow 7r-band with little dispersion. At around 4 eV a broad band results from nonbonding and bonding 7r-orbitals, whereas the broad emissions at 8eV and ll.SeV result from a orbitals. The UPS data on longer oligomers ( > 6) exhibit mainly the features of polythiophene [222, 223]. 

Picard F, Buffeteau T, Desbat B, Auger M and Pezolet M (1999) Quantitative orientation measurements in thin films by attenuated total reflection spectroscopy. Biophysical Journal 16 539-551. 

The often-cited Amontons law [101. 102] describes friction in tenns of a friction coefiBcient, which is, a priori, a material constant, independent of contact area or dynamic parameters, such as sliding velocity, temperature or load. We know today that all of these parameters can have a significant influence on the magnitude of the measured friction force, especially in thin-film and boundary-lubricated systems. 

Theron C C, Mars J A, Churms C L, Farmer J and Pretorius R 1998 In situ, real-time RBS measurement of solid state reaction in thin films Nuol. Instrum. Methods B 139 213... 

Thomas and Rice [/. Appl. Mech., 40, 321-325 (1973)] applied the hydrogen-bubble technique for velocity measurements in thin hquid films. DureUi and Norgard [Exp. Mech., 12,169-177 (1972)] compare the flow birefringence and hydrogen-bubble techniques. 

XRD offers unparalleled accuracy in the measurement of atomic spacings and is the technique of choice for determining strain states in thin films. XRD is noncontact and nondestructive, which makes it ideal for in situ studies. The intensities measured with XRD can provide quantitative, accurate information on the atomic arrangements at interfaces (e.g., in multilayers). Materials composed of any element can be successfully studied with XRD, but XRD is most sensitive to high-Z elements, since the diffracted intensity from these is much lar r than from low-Z elements. As a consequence, the sensitivity of XRD depends on the material of interest. With lab-based equipment, surface sensitivities down to a thickness of -50 A are achievable, but synchrotron radiation (because of its higher intensity)... 

Stress in crystalline solids produces small shifts, typically a few wavenumbers, in the Raman lines that sometimes are accompanied by a small amount of line broadening. Measurement of a series of Raman spectra in high-pressure equipment under static or uniaxial pressure allows the line shifts to be calibrated in terms of stress level. This information can be used to characterize built-in stress in thin films, along grain boundaries, and in thermally stressed materials. Microfocus spectra can be obtained from crack tips in ceramic material and by a careful spatial mapping along and across the crack estimates can be obtained of the stress fields around the crack. ... 

The analytical techniques covered in this chapter are typically used to measure trace-level elemental or molecular contaminants or dopants on surfaces, in thin films or bulk materials, or at interfaces. Several are also capable of providing quantitative measurements of major and minor components, though other analytical techniques, such as XRF, RBS, and EPMA, are more commonly used because of their better accuracy and reproducibility. Eight of the analytical techniques covered in this chapter use mass spectrometry to detect the trace-level components, while the ninth uses optical emission. All the techniques are destructive, involving the removal of some material from the sample, but many different methods are employed to remove material and introduce it into the analyzer. 

Although the majority of NAA applications have been in the area of bulk analysis, some specialized uses need to be mentioned. One such unique application is the measurement of phosphorus in thin films (about 5000 A) of phosphosilicate (PSG) or borophosphosilicate (BPSG) glasses used in VLSI device fabrication. In this case,... 

NAA cannot be used for some important elements, such as aluminum (in a Si or Si02 matrix) and boron. The radioactivity produced from silicon directly interferes with that ftom aluminum, while boron does not produce any radioisotope following neutron irradiation. (However, an in-beam neutron method known as neutron depth profiling C3J be used to obtain boron depth profiles in thin films. ) Another limitation of NAA is the long turn-around time necessary to complete the experiment. A typical survey measurement of all impurities in a sample may take 2-4 weeks. 

Pharr, G.M., Harding, D.S. and Oliver, W.C., Measurement of fracture toughness in thin films and small volumes using nanoindentation methods. MRS, 1995, pp. 663-675. 

Absolute photoluminesccnce efficiency measurements in thin solid films are quite difficult, since light-trapping, waveguiding effects and, possibly, distributions in the emission dipole moments of individual chromophorcs modify the angular distribution of the emission. Dc Mello el al. ]126] have described an improved mclh-... 

One of the more important uses of OM is the study of crystallization growth rates. K. Cermak constructed an interference microscope with which measurements can be taken to 50° (Ref 31). This app allows for study of the decompn of the solution concentrated in close proximity to the growing crystal of material such as Amm nitrate or K chlorate. In connection with this technique, Stein and Powers (Ref 30) derived equations for growth rate data which allow for correct prediction of the effects of surface nucleation, surface truncation in thin films, and truncation by neighboring spherulites... 

Some years later a more thorough discussion of the motion of pairs of electrons in a metal was given by Cooper,7 as well as by Abrikosov8 and Gor kov,9 who emphasized that the effective charge in superconductivity is 2e, rather than e. The quantization of flux in units hc/2e in superconducting metals has been verified by direct experimental measurement of the magnetic moments induced in thin films.10 Cooper s discussion of the motion of electron pairs in interaction with phonons led to the development of the Bardeen-Cooper-Schrieffer (BCS) theory, which has introduced great clarification in the field of superconductivity.2... 

More systematic study of the dynamics of radicals in solution should now be possible using CIDNP. Investigations so far reported have indicated that the rates of very rapid chemical reactions and other dynamic processes undergone by radicals can be measured in a crade way greater refinement should be possible. Special effects have been predicted for reactions in thin films (Deutch, 1972). Moreover the time-scale of polarization is such that the technique may prove capable of throwing new light on the dynamics of excited states. 

Another remarkable feature of thin film rheology to be discussed here is the quantized" property of molecularly thin films. It has been reported [8,24] that measured normal forces between two mica surfaces across molecularly thin films exhibit oscillations between attraction and repulsion with an amplitude in exponential growth and a periodicity approximately equal to the dimension of the confined molecules. Thus, the normal force is quantized, depending on the thickness of the confined films. The quantized property in normal force results from an ordering structure of the confined liquid, known as the layering, that molecules are packed in thin films layer by layer, as revealed by computer simulations (see Fig. 12 in Section 3.4). The quantized property appears also in friction measurements. Friction forces between smooth mica surfaces separated by three layers of the liquid octamethylcyclotetrasiloxane (OMCTS), for example, were measured as a function of time [24]. Results show that friction increased to higher values in a quantized way when the number of layers falls from n = 3 to n = 2 and then to M = 1. 

Heavy-Ion ERDA for Measurement of Cr Dopant in Thin Films of / -FeSi2... 

---