Several levéis of structural information are of interest to the thin-film scientist and technologist in research, process development, and reliability and failure analysis activities. The first broadly deals with the geometry of patterned films where issues of lateral or depth dimensions and tolerances, uniformity of thickness and coverage, completeness of etching, etc.. are of concern. Beyond this, the film surface topography and morphology, including grain size and shape, existence of compounds, presence of hillocks or whiskers, evidence of film voids, microcracking or lack of adhesión, formation of textured surfaces, etc., are of concern. Somewhat more difficult to obtain, but crucial to microelectronic device fabrication and optical coating technology, are the cross-sectional views of multilayer structures where interfacial regions, sub-strate interactions, and geometry and perfection of electronic devices with associated conducting and insulating layers may be observed.
Lastly, and most complex of all, are diffraction patterns, the crystallo-graphic information they convey, and the high-resolution lattice images of both plain-view and transverse film sections. Among the applications here are defect structures in films and devices, structure of grain boundaries, identification of phases, and a host of issues related to epitaxial structures—e.g., the crystallo-graphic orientations involved, direct imaging of atoms at interfaces, interfacial quality and defects, perfection of quantum well and strained-layer superlat-ticcs. The transmission electrón microscope (TEM) is required for these applications, whereas those of the previous paragraph are normally addressed by the scanning electrón microscope (SEM) and, occasionally, by the reflec-tion metallurgical microscope. There is an interesting distinction between the TEM and SEM. The former is a true microscope in that all image information is acquired simultaneously or in parallel. In the SEM, however, only a small portion of the total image is probed at any instant, and the image builds up serially by scanning the probé. Strictly speaking, the SEM is more like the scanning Auger electrón and SIMS microprobes than a traditional microscope. In this section we treat only electrón microscopy, a subject dealt with at length in the recommended references (Refs. 9, 10). We start with the SEM.
Scanning Electrón Microscopy
Because seeing is believing and understanding, the SEM is perhaps the most widely employed thin-film and coating characterization instrument. A schematic of the typical SEM is shown in Fig. 6-7. Electrons thermionically emitted from a tungsten or LaB6 cathode filament are drawn to an anode, focused by two successive condenser lenses into a beam with a very fine spot size (~ 50 A). Pairs of scanning coils located at the objective lens deflect the beam either linearly or in ráster fashion over a rectangular área of the specimen surface. Electrón beams having energies ranging from a few thousand to 50 keV, with 30 keV a common valué, are utilized. Upon impinging on the specimen, the primary electrons decelérate and in losing energy transfer it inelastically to other atomic electrons and to the lattice. Through continuous random scattering events, the primary beam effeetively spreads and filis a teardrop-shaped interaction volume (Fig. 6-8a) with a multitude of electronic excitations. The result is a distribution of electrons that manage to leave the specimen with an energy spectrum shown schematically in Fig. 6-8b. In addition, target X-rays are emitted, and other signáis such as light, heat, and specimen current are produced, and the sources of their origin can be imaged with appropriate detectors.
The various SEM techniques are differentiated on the basis of what is subsequently detected and imaged.
Secondary Electrons. The most common imaging mode relies on detection of this very lowest portion of the emitted energy distribution. Their very low energy means they origínate from a subsurface depth of no larger than several angstroms. The signal is captured by a detector consisting of a scintillator-photomultiplier combination, and the output serves to modulate the intensity of a CRT, which is rastered in synchronism with the raster-scanned primary beam. The image magnification is then simply the ratio of sean lengths on the CRT to that on the specimen. Resolution specifications quoted onresearch quality SEMs are ~ 50 A. Great depth of focus enables images of beautiful three-dimensional quality to be obtained from nonplanar surfaces. The contrast variation observed can be understood with reference to Fig. 6-8c. Sloping surfaces produce a greater secondary electrón yield because the portion of the interaction volume projected on the emission región is larger than on a fíat surface. Similarly, edges will appear even brighter. Many examples of secondary electrón SEM images have been reproduced in various places throughout the book.
Electron-Beam-Induced Current (EBIC). The EBIC mode is ap-plicable to semiconductor devices. When the primary electrón beam strikes the surface, electron-hole pairs are generated and the resulting current is collected to modulate the intensity of the CRT image. The technique is useful in spatially locating subsurface defects and failure sites within a junction región.
Transmission Electrón Microscopy
As the ñame implies, the transmission electrón microscope is used to obtain structural information from specimens that are thin enough to transmit electrons. Thin films are, therefore, ideal for study, but they must be removed from electrón-impenetrable substrates prior to insertion in the TEM. The two basic modes of TEM operation are differentiated by the schematic ray.Electrons thermionically emitted from the gun are acceler-ated to 100 keV or higher (1 MeV in some microscopes) and first projected onto the specimen by means of the condenser lens system. The scattering processes experienced by electrons during their passage through the specimen determine the kind of information obtained. Elastic scattenng, ínvolving no energy loss when electrons interact with the potential field of the ion cores, gives rise to diffraction patterns. Inelastic interactions between beam and matrix electrons at heterogeneities such as grain boundaries, dislocations,second-phase particles, defects, density variations, etc., cause complex absorp-tion and scattenng effects, leading to a spatial variation in the intensity of the transmitted beam. The generation of characteristic X-rays and Auger electrons also occurs, but these by-products are not usually collected.
The emergent primary and diffracted electrón beams are now made to pass through a series of post-specimen lenses. The objective lens produces the first image of the object and is, therefore, required to be the most perfect of the lenses. Depending on how the beams reaching the back focal plañe of the objective lens are subsequently processed distinguishes the operational modes. Basically, either magnified images are formed or diffraction patterns are obtained as shown in Fig. 6-10. A discussion of the analysis of diffraction effects is well beyond the scope of this book. Some notion of the correlation between structure and diffraction patterns can be gained from the model of thin films presented in Section 5.7.3.
Images can be formed in a number of ways. The bright-field image is obtained by intentionally excluding all diffracted beams and only allowing the central beam through. This is done by placing suitably sized apertures in the back focal plañe of the objective lens. Intermedíate and projection lenses then magnify this central beam. Dark-field images are also formed by magnifying a single beam; this time one of the diffracted beams is chosen by means of an aperture that blocks the central beam and the other diffracted beams. The micrograph of alternating 40 A-wide films in the GaAs-Al05Ga05As superlat-tice structure shown in Fig. 6-11 is a dark-field image employing the 200 diffracted beam. In both of these cases we speak of amplitude contrast because diffracted beams with their phase relationships are excluded from the imaging sequence. In a third method of imaging, the primary transmitted and one or more of the diffracted beams are made to recombine, thus preserving both their amplitudes and phases. This is the technique employed in high-resolution lattice imaging, enabling diffracting planes and arrays of individual atoms to be distinguished.
The ability to prepare thin vertical sections of integrated circuits for TEM observation is one of the most important recent advances in technique. If one can imagine the plañe of this page to be the surface thinned for conventional TEM work, transverse imaging requires head-on thinning and viewing of the ~ 75-/¿m-thick page edge. What is involved is the transverse cleavage of a number of wafer specimens, bonding these in an epoxy button, and thinning them by mechanical grinding and polishing. Finally, the resulting thin disk is mounted in an ion-milling machine where the specimen is further sputter thinned by ion bombardment until a hole appears. In VLSI applications, many specimens must be simultaneously mounted to enhance the probability of capturing images from desired circuit features.
X-Ray Diffraction
X-ray diffraction is a very important experimental technique that has long been used to address all issues related to the crystal structure of bulk solids, including lattice constants and geometry, identification of unknown materials, orientation of single crystals, and preferred orientation of polycrystals, defects, stresses, etc. Extensión of X-ray diffraction methods to thin films has not been pursued with vigor for two main reasons: First the great penetrating power of X-rays means that with typical incident angles, their path length through films is too short to produce diffracted beams of sufficient intensity. Under such conditions the substrate, rather than the film, dominates the scattered X-ray signal; thus, diffraction peaks from films require long counting times. Second, the TEM provides similar diffraction information with the added capability of performing analysis over very small selected áreas. Nevertheless, X-ray methods have advantages because they are nondestructive and do not require elabórate sample preparation or film removal from the substrate.
No hay comentarios:
Publicar un comentario