Off-axis electron holography is a powerful method for providing quantitative micromagnetic structure to nanometer-scale resolution. One of the main challenges in determining magnetic induction from reconstructed phase images is the separation of phase shifts due to electrostatic and thickness effects from magnetostatic effects. We have used in-situ magnetization reversal to remove electrostatic and thickness effects which would otherwise prevent quantification of the induction. The Philips CM200-FEG microscope which was used for the holography described here is equipped with a powerful mini-lens below the specimen enabling 2nm spatial resolution (∼5nm phase resolution) and only a small residual field at the sample. In addition, the objective lens field versus current has been calibrated so that external vertical fields over a wide range of values may be applied to the sample. Tilting the sample in the presence of the vertical field has the effect of ‘applying an inplane component. This combination of high spatial resolution, quantitative analysis and in situ capabilities allows for the study of individual defects and provides experimental input for comparison with simulations.

The recent increased interest in Lorentz Microscopy methods is due to two important factors: (1) the advent of digital detection systems (CCD cameras) and advanced computer controlled transmission electron microscopes, and (2) the continually decreasing length scale of magnetic recording media and related magnetic materials. Along with the increased experimental resolution and detection sensitivity one should ask the question: how small a magnetic moment can one detect with conventional Lorentz or electron holography techniques? And, perhaps more importantly, which technique should one use to obtain the best spatial resolution? To address these questions we identified a simple model system :the uniformly magnetized sphere.

The phase of the electron wave after passing through a region with a non-zero magnetic vector potential and an internal electrostatic potential Vscan be computed from the Aharonov-Bohm trajectory integral. For a spherical particle with a uniform magnetization and a constant mean inner potential (see geometry in Fig. la), this phase shift can be analytically computed.

The shape anisotropy of a magnetic thin film competes with the easy-axis anisotropy, which gives rise to different domain structures compared to the bulk magnetic materials. Recent Lorentz microscope and electron holography and differential-phase-contrast experiments show complex domain structures with features not found in bulk materials seen by magneto-optic method.

In this report we use micromagnetic simulation techniques to study the effect of thickness on the domain structure of magnetic thin films with various orientations relative to the easy axis and compare with those in the bulk materials. The results of the simulations can be used to compare with the data obtained using electron holography experiments to give us quantitative understanding of the domain structure in hard magnetic materials.

We model the magnetic films with an array of dipoles of size 16×l6×Lz, where Lz the thickness of the sample.

Magnetic imaging can be performed both on conventional (projection) and scanning TEM (CTEM/STEM). The CTEM-based magnetic imaging, classified into Fresnel, Foucault, holography, and other modes, has proven successful by now in observing various magnetic objects, and has accordingly been well documented elsewhere. The STEMbased one has emerged relatively recently prompted by the rapid growth of current STEM technology.

The most ubiquitous implementation of Lorentz microscopy on a STEM is the differential phase contrast (DPC) mode. Each scanning signal obtained in the DPC mode is linear to Lorentz deflection, thus reflects the magnitude of an in-plane component of magnetic induction integrated along the optical axis(Fig.l). The signals therefore suffice to generate integrated in-plane magnetic induction maps(Fig.2). Furthermore, by making use of the fact that the Lorentz deflection distribution across the image plane forms an irrotational (vortex-free) 2D vector field, a pertinent numerical image processing yields a function equivalent to the electron phase function representing magnetically distorted electron wavefront.

Mulitlayers consisting of various combinations of 1-10 nm thick ferromagnetic, antiferromagnetic and nonmagnetic films have numerous applications in magnetic recording, non-volatile data storage, and sensor technologies. One key aspect to understanding how these devices work is being able to image their magnetic structure. SEMPA is well suited for this job, because of its surface sensitivity and compatibility with thin film growth and surface analysis techniques. This talk will review the SEMPA technique and describe thep use of SEMPA in two examples; one in which the magnetization is imaged while growing a multilayer, and another which examines the magnetization while depth profiling a multilayer by ion milling.

A SEMPA investigation of the magnetic exchange coupling in an Fe/Cr/Fe sandwich is shown in Figs. 1 and 2. A wedge shaped Cr interlayer was grown as shown in Fig. 1 in order to measure the Cr thickness dependence of the exchange coupling.

The magnetic properties of permanent magnets are sensitive to their microstructure. In particular, for the family of Nd(Pr)-Fe-B magnets a very different coercivity and energy products may be obtained by several processing techniques. It was experimentally found that a small excess of Nd over the exact phase composition of Nd2Fe14B plays an important role in obtaining high-energy products during the die-upset processing of the anisotropic hard magnets. However the specific role of the Nd excess on both magnetic structure and microstructure of these die-upset magnets is unclear and controversial. Answers to these questions may help to correctly address some major issues in materials science, e.g. how microstructure is related to magnetic structure of hard magnets, and how to optimize the performance of hard magnets.

In-situ TEM magnetizing experiments combined with Lorentz magnetic microscopy in Fresnel-Foucault modes were used to characterize the magnetic structure of die-upset, high energy-product hard magnets Nd13.75Fe80.25B6 and Pr13.75Fe80.25B6.

In this work, we report on the observation of two coexisting patterns in Magnetic Force Microscopy (MFM) images of shape memory alloys. The MFM signal of both patterns presents similar behavior with tip-surface separation. An investigation on the origin of these patterns presents strong evidence that both are of magnetic nature only and, furthermore, can be assigned as bulk and surface-related, respectively.

Shape memory alloys are a class of materials that exhibit martensitic transformation [1]. The shape memory effect, which may be defined as the property of recovering its original shape during a thermal cycle after a material has been apparently deformed in a permanent way, is found in several materials, including the Fe-Mn-Cr-Ni-Si alloy, studied in this work [1]. A standard metalographic process was employed to reveal ferritic-phase islands surrounded by the austenitic-phase matrix, which constitute the structure of this alloy. The images were acquired with a MultiMode microscope from Digital Instruments.

For quantitative magnetic image analysis it is important that the value of the magnetic field applied to the sample is known. Calibration of the magnetic field inside an electron microscope is not a simple procedure, however, because it requires the use of a special sample holder which contains a Hall probe sensor. We report here on a simple and effective procedure for calibration of the magnetic field at the sample position of the JEOL 3000 FEG transmission electron microscope (equipped with ultra-high resolution pole pieces URP-32F, point resolution of 1.6 Å and tilt angle ±30°) without the use of a special sample holder. The technique is based upon the fact that the magnetic flux present in a magnetized material in a field-free environment, known as the remanence Br, is a sensitive function of the largest applied magnetic field. The inverse of this principle — that the value of an applied field may be determined from measurements of the remanence, — was used to quantify the magnitude of the field present at the sample position.

The physical and chemical characterization of defects and particles on wafer surfaces gives important information on their origin. An effective approach for root-cause determination of wafer surface particle problems is to perform both high-resolution images and chemical characterization on individual defects. The traditional root-cause analysis technique used in the semiconductor industry is Scanning Electron Microscopy combined with Energy Dispersive X-Ray Spectroscopy (SEM/EDX). This technique provides high-resolution images and the atomic composition of the defects. However, this is sometime insufficient to fully characterize them and to determine their origin. In this case, the identification of the molecular composition of the defects can help determine their origin.

Optical spectroscopy, such as Infrared and Raman spectroscopy, is a powerful method for determining the molecular composition of many substances. Micro-FT-IR spectroscopy can characterize semiconductor particles only down to 10 μm because of the IR wavelength diffraction limit. Micro-Raman spectroscopy, instead, can provide useful molecular information on many organic as well as inorganic species with 1 μm spatial resolution.

μ-Raman spectroscopy has been used to characterize two types of biomedical materials: a multilayer, silicone elastomer used in implants, and a thin, polyethylene layer used in a medical device. Raman spectroscopy is the collection of light inelastically scattered by a material or compound. The technique is based on the Raman effect, which involves the interaction of light and matter. When light strikes a material, the light is inelastically scattered and is frequency shifted according to the vibrations of the chemical-functional groups and/or macrostructure of the material. The result is a Raman spectrum of the material that can be interpreted to determine the characteristics of the material, including identity, macrostructure, and quantity of a specific material within a matrix.The application of Raman spectroscopy in the characterization of polymers has been well established. The technique has been used to determine the chemical composition and morphology of polymers. Raman spectroscopy could therefore be a powerful tool for characterizing polymeric biomaterials.

The silicone elastomer characterized in this work consisted of three layers: polydimethylsiloxane, polydimethyl/polydiphenyl siloxane, and polydimethylsiloxane.

With the introduction of high efficiency short acquisition time Raman microscopes at the beginning of the 1990's Raman mapping experiments became more realistic for many different materials. Previous to this time Raman maps of any useful size would take far too long for all but the strongest of Raman scatterers. A typical Raman spectrum used to take 30 minutes or more, a relatively small map of 50μ × 50μ using a step size of 1μ, would take over 50 days to acquire. Today the same map can be obtained in less than 4 hours. The use of an xyz mapping stage in conjunction with a Raman microscope is now common. The user defines an area on the sample that requires mapping and selects a step size and spectral acquisition time. Once the experiment is started the mapping stage drives the sample to the start position and takes the first spectrum, once this is completed the sample is moved by the requested step size and the next spectrum is measured.

In order to remain competitive in the rapidly growing inkjet printer industry, higher and higher demands are being made on print speed and print/image quality. As a result, the critical dimensions of thermal inkjet cartridge printheads are being constantly reduced in order to yield smaller ink droplets at higher frequencies. In order to enhance our laboratory’s ability to acquire vibrational spectral data from the increasingly small sample sizes encountered in failure analysis problems, a Raman microscope was recently added to our infrared microscopy laboratory.

The addition of a Raman microscope has increased our analysis capabilities in several ways. While infrared microscopy is diffraction limited at 10 μm, Raman microscopy can be used to study spot sizes as small as 1 μm in diameter. This has greatly expanded the range of particles which can be identified in our work. The Raman spectrum acquired of an adhesive particle in a printhead bore is shown in Fig.1.

Raman Microprobe Spectroscopy has been used in research laboratories to model the growth of thin film silicide layers and the re-crystallization of α-silicon, however little progress has been made towards using these techniques as either at line or on line tools.

At line and on line instrumental techniques such as ellipsometry and X-Ray Fluorescence have been used to characterize the physical properties of thin films in the semiconductor industry, which has lead to improvements in device yields. These techniques, however, do not provide the specificity to determine thin film phase changes which result in improved device performance, whereas Raman Microprobe Spectroscopy can monitor these changes.

In order to perform at line or on line measurements, the spectrometer must be moved to the process line floor. The recent improvements in spectrometer design which has produced systems robust enough to endure high technology manufacturing environments will be briefly discussed. Two case studies of semiconductor process monitoring via Raman Microprobe analysis will be presented.

Attention has been given in the recent past to the various issues that arise from the general lack of standardized procedures for the calibration of Raman spectral intensity. With the increasing acceptance of analytical Raman methods and the availability of diverse types of commercial Raman systems, it becomes increasingly important that uniform calibration methods be developed that provide some means for the correction of the Raman instrument response function. As matters stand now, Raman spectra are typically not corrected for instrument-dependent variations in spectrometer sensitivity across the Raman range. Especially for measurements at different excitation wavelengths, this leads to often severe distortion of relative peak intensities which impedes calibration transfer and affects spectral matching in library search routines. These issues have been brought forth by the McCreery group to the point where national standards organizations (e.g., ASTM, NIST), federal regulatory agencies, the Raman instrument manufacturers, and the analytical Raman community, have taken note.

Optical microscopy has been the workhorse pathological technique, qualitatively differentiating tissue sections by characterizing morphological variations. Vibrational spectroscopic techniques provide quantitative as well as qualitative analytical information that reflects a sample’s biochemical composition and molecular structure. The utility of infrared and Raman techniques for biological characterization has been demonstrated for a variety of applications.[1-4]

FTIR imaging microscopy is a newly developed technique that incorporates the imaging capabilities required for histological procedures with the chemical discrimination of IR spectroscopy.[5-7] The ability to maintain spatial integrity while accessing precise spectroscopic data intrinsic to the sample represents a powerful combination. This technique is much more amenable to analysis by a pathologist than conventional spectroscopy because the data can be presented as images. These images provide direct visualization of a sample’s biochemical heterogeneity.

Vibrational spectroscopic imaging techniques provide excellent sample statistics for the accurate classification of individual spectral signatures because tens of thousands of independent spectra from different spatial locations within the sample are simultaneously recorded.

Four major components comprise an inkjet cartridge (Fig.lA): (1) the pen body which contains the ink and positions the cartridge in the printer; (2) an ink delivery system which supplies ink at the correct backpressure and flow rate; (3) a flex circuit which provides the electrical interconnect to the printer; and (4) the printhead which generates and directs the drops. Because of its complexity and small critical dimensions, the printhead poses the greatest analytical challenge for many failure analysis problems. Each printhead may contain as many as 300 firing chambers (Fig. IB). Each chamber consists of a resistor surrounded by a polymer-based barrier material which forms a cavity into which ink flows from the ink delivery system. The chamber is capped by a metal or polymer orifice-containing plate. Printing occurs when the resistor is heated to form an ink vapor bubble which subsequently ejects a droplet of ink though the orifice onto the paper.

A forensic examination of automobile paint may involve either a comparative analysis of two paint samples or an attempt to identify a recovered paint chip to determine the make, model, and year of the vehicle from which it originated. Along with microscopy, the main tool which most U.S. forensic laboratories use for such analyses is infrared spectroscopy. Paint binders are normally identified by this means together with some inorganic pigments. Until recently, however, there were no published reports describing the identification of organic pigments in the cured paint samples normally analyzed in the forensic laboratory. The main reasons for this stem from the difficulties in isolating these pigments from the inert paint matrices and the insolubilities of the pigments. Using infrared microspectroscopy and spectral subtractions, several organic pigments used in automotive finishes have now been identified in situ based on spectra of intact paint samples.

Infrared spectra were obtained of U.S. automobile original (OEM) finiśhes from the Reference Collection of Automotive Paints.

Over the past several years many developments have taken place in the field of molecular spectroscopy. For Raman spectroscopy many of the improvements have arisen from technological innovations that include diode-based lasers, holographic notch filters and charged coupled detectors. In contrast, a majority of the developments in infrared spectroscopy have been in the area of new sampling accessories. A major emphasis has been placed on attenuated total internal reflection (ATR) accessories. The devices are allowing infrared spectroscopy to be employed in process control environments and quality control laboratories where the method is not only robust but has the advantages of limited sample preparation and/or in situ analysis.

In the realm of microspectroscopy, ATR accessories have the added advantages of providing better spatial resolution, equal to or higher S/N for equivalent sample size compared to transmission measurements and most importantly the ability to collect spectra of small samples without the adverse effect of diffraction. One accessory which was developed several years ago is known as the Split-Pea.