next up previous
Next: Results Up: Title page Previous: Introduction References Figure Captions

Experimental

The films used in this study were all grown using the ALL-MBE technique, which has been described in detail previously.[5] ALL-MBE is a variant of molecular beam epitaxy in which the composition and structure of a growing layered film are controlled by sequential evaporation of the constituent metals and their oxidation. The growth is performed in a highly reactive ozone atmosphere at elevated temperatures between 650 and 700°C. Growth is monitored in situ by observing reflection high-energy electron diffraction patterns, and post-growth characterization has been performed by x-ray diffraction, Rutherford backscattering (RBS), Auger electron spectroscopy (AES), x-ray fluorescence (XRF), secondary ion mass spectroscopy (SIMS), and atomic force microscopy (AFM). The RHEED observations confirm that abrupt changes in crystalline phase can be made, namely that ALL-MBE allows alternate layers of 2212 and 2201 to be grown as a superlattice.[4] RBS, AES and EDAX measurements confirm that the stoichiometry of the films is within a few percent of the nominal value.[7] X-ray diffraction measurements show that single-phase growth is possible using the ALL-MBE technique.[8] Total film thickness is typically on the order of 1000 Å. The superconducting transitions of all films used for TEM studies were measured resistively and found to be narrow, with a transition temperature Tc similar to that reported earlier.[6] Films like those used for TEM sample preparation have been fabricated into tunnel junction devices.[5]

Preparation of high-temperature superconducting films for examination by TEM has proven to be difficult. The major difficulties are as follows: differential thinning rates between substrate, film materials and materials used to fabricate the cross-section specimen; the inherent brittle nature of substrate and film materials; and the production of specimen-preparation-induced artifacts such as ion damage and amorphization. The results of these problems are specimens with only a small transparent region, fracture of the substrate and delamination of the film. To reduce and avoid the occurrence of these specimen preparation difficulties, advanced techniques are employed. The specimen preparation procedure described here employs a number of steps to avoid mechanical damage and reduce artifacts.

The basic construction of the cross-sectioned specimen is similar to the techniques describe by Newcomb[9] and Bravman.[10] The following is a detailed description of how their procedure has been adapted for BiSrCaCuO films.

To begin with, a piece of blank Si wafer is epoxied to the SrTiO3 substrate of the superconducting thin film. The long direction of the Si strip is oriented on the surface of the film-substrate combination along the predetermined TEM viewing direction; i.e., [010] or [110]. The substrate is then trimmed away with a low speed diamond saw. The remaining Si/epoxy/film/substrate composite is laminated with epoxy between two semi-circular brass rods and inserted into a brass tube filled with epoxy. After curing, discs measuring .5 mm thick are sliced from the brass rod using a low-speed diamond saw. Disc specimens are lapped using a gravity feed holder from one side using a succession of fine diamond lapping films to a thickness of .25 mm. The specimen disc is then final polished on a vibrating polishing machine for several hours. Prior to lapping the second side, a Cu grid is epoxied to the polished surface for support. The second side is then lapped using the same diamond grit films to a final specimen thickness of 100 µm (excluding the Cu grid). The specimen is dimpled from the second polished side directly on top of the Si/SrTiO3 interface. The dimpling conditions are: 15 g load, 60 rpm, 2-4 µm diamond powders. The final dimpled specimen thickness prior to ion milling is approximately 15-20 µm. Dimpling to thinner dimensions often caused the substrate to fracture or the film to delaminate. Then a Gatan, Inc. PIPS ion mill is used to sputter the film at a low angle of 4.5°. Low-angle ion milling in combination with ion beam modulation is utilized to reduce the difference in ion milling rates between the substrate, film, epoxy and Si. The first side of the specimen (dimpled side) is ion milled for about 90 min. The specimen is turned over and ion milled until perforation. After perforation both sides are again low-angle ion-milled at 1.5 keV to reduce surface damage. The specimen is examined using low-resolution electron microscopy, and if necessary is returned for further ion milling.

In order to characterize the layering and defect microstructure of these atomically engineered films, high resolution electron microscopy (HREM) was utilized. A top entry JEOL 4000EX with a point-to-point resolution of 0.17 nm was used to image atomic structure detail in cross-sectioned specimens having film orientations of [010] and [110]. Experimental images were recorded near the scherzer defocus value (approximately -45 nm) and the first cross-over of the contrast transfer function (approximately -75 nm). Images were also recorded at other defocus values at which specific layering and atomic structure detail was best resolved. The instrumental imaging conditions were as follows: 400 keV accelerating voltage, Cs = 1.0 mm, focus spread of 8.0 nm, beam semi-divergence angle of about 0.8 mr and an objective aperture radius of 0.096 nm-1.

The atomic structure of the layering was confirmed by comparison of the experimental images to those of simulated images. Calculation of the simulated images was performed using the MacTemPas software package.[11] Published lattice constants were used as inputs for 2201 and 2212, while lattice constants for 2278 and 1278 were estimated by using the same interatomic spacings as in the equilibrium phases. Since the specimen thickness and defocus conditions are unknown for particular images, calculations were performed for a range of values of these parameters.



next up previous
Next: Results Up: Title page Previous: Introduction References Figure Captions



alison (Alison Chaiken)
Sat Jan 20 14:12:46 PST 1996