The Role of Steps on the Micro Structure of Titanium-Oxide Films Deposited by Pld Onto (100) MgO Surfaces

  • PDF / 3,011,303 Bytes
  • 5 Pages / 414.72 x 648 pts Page_size
  • 80 Downloads / 178 Views

DOWNLOAD

REPORT


SIMON KING AND C. BARRY CARTER Department of Chemical Engineering and Materials Science, University of Minnesota, 421, Washington Av S.E., Minneapolis, MN 55455. ABSTRACT The phenomenon of graphoepitaxy, whereby island nucleation occurs preferentially at steps, is a prime example of a manner in which the substrate morphology (and notably surface steps) can affect the microstructure of as-deposited thin-films. This paper provides a further illustration of the fundamental importance of the substrate morphology to the development of the thin-film microstructure, employing the example of TiO 2 deposited by pulsed-laser deposition (PLD) directly onto specially prepared (100) MgO TEM foils. The same areas of these foils are imaged both before and after deposition. INTRODUCTION Some examples of graphoepitaxy are the islanding of a- and y-Fe20 3 and YBCO deposited by pulsed-laser deposition (PLD) and chemical vapor deposition (CVD) onto (100) MgO and A12 0 3 foils [1,2]. In this paper, TiO 2 is deposited by PLD onto pre-characterized (100) MgO TEM foils and the presence, or absence of well-defined surface steps is seen to affect not only the microstructure of the as-deposited thin film but also the phases that are formed. It has already been shown elsewhere, that the faces of the larger surface steps on these substrates are shallowly inclined to the (100) MgO foil normal: As such they effectively present numerous high-index vicinal faces to the deposition plume [3]. Unlike most previous studies, pre-characterized TEM-ready substrates with a known and reproducible surface morphology formed after acid-washing and annealing in air, are employed. This practice both eliminates the possibility of introducing post-deposition specimen-preparation artifacts and allows the TEM imaging of identical areas before and after deposition. EXPERIMENTAL Prior to deposition, the TEM-ready substrates are prepared according to the method of Norton et al. [4] in which dimpled foils are perforated by ion-milling, cleaned in acid, with the active ingredient being aqua-regia, and heat treated in air. Results from substrates heat-treated at 1350°C for 3.5, 15.5 and 30 min. are shown in this paper. Specimens prepared in this fashion exhibit a 'known' morphology of (100) faceted terraces separated by characteristic, generally curved, surface steps. [3,5] Full details of the deposition conditions have been given elsewhere [6]: Briefly, a rutile pellet of 98% theoretical density is ablated at 10Hz with a KrF excimer laser. Prior to deposition, the MgO substrates are cleaned by baking in-situ at 800*C for 30 min. in 100mTorr of flowing 02. Two different deposition conditions are employed: 1) 4 min. (2400 pulses; figures 1 to 4) at 500*C (nominal) and 2) 38 sec. (380 pulses; figure 5) at 900'C (nominal). All depositions are carried out in 50 mTorr of flowing oxygen. The deposition rate, as calibrated by cross-sectional TEM, is 0.2 A/pulse, leading to expected film thicknesses of 50nm (1) and 8nm (TiO 2 equivalent (2)). After deposition, no further specime