Trapping of transient processes in aluminium oxide thin films in a voltage pulse experiment Achim Walter Hassel a, * , Detlef Diesing b a Max-Planck-Institut f€ ur Eisenforschung GmbH, Max-Planck-Strasse 1, 40237 D€ usseldorf, Germany b Institut f€ ur Physik der kondensierten Materie, Heinrich-Heine-Universit€ atD€ usseldorf, Universit€ atsstr. 1, 40225 D€ usseldorf, Germany Received 13 September 2001; received in revised form 10 October 2001; accepted 10 October 2001 Abstract An experimental setup is presented that allows the trapping of transient states in potentiostatic and potentiodynamic experi- ments. The setup is suitable for electrochemical experiments as well as for dielectric investigations. The system stops an experiment by triggering at a predefined current level after a minimum time of the voltage pulse. The advantage of this device is demonstrated by means of a voltage pulse annealing procedure for a metal–insulator–metal (MIM) contact with an anodically prepared aluminium oxide film as insulator. The setup significantly increases the stability against a breakdown of the anodic oxide film. Ó 2002 Elsevier Science B.V. All rights reserved. Keywords: Current transients; Dielectric films; Aluminium; Tunnelling; Electrochemistry; MIM 1. Introduction In electrochemistry and dielectrics, the trapping of transient states is of interest. This allows the investiga- tion of a system in a well-defined state, using methods that are otherwise too slow. Transient states may proceed the breakdown of thin oxidefilms[1,2].Breakdownsarecharacterisedbycurrent levels that are significantly higher than the steady-state currents at that time [3]. Nevertheless, the current might be only a local maximum since the absolute current value can be higher directly after a voltage pulse due to capac- itive charging. This current does not attack the film at all [4]. One needs a device that limits the current in a certain time range, but allows the current to exceed this limit during the capacitive charging. 2. Experiments and results Fig. 1 shows the current transient of a cathodic breakdown in a double-logarithmic presentation on a high purity aluminium sample (99.95%) covered by an 4.5 nm thick aluminium oxide film in pH 6.0 acetate buffer. Four regions can be distinguished in this transient. After capacitive charging (A) the current decreases down to a local minimum (B) due to re- laxation processes [5,6]. The current then increases again during the cathodic breakdown (C) before it reaches the constant value of stationary hydrogen evolution (D). The time to breakdown (taken as the point of in- flection in section (C) of the current transient) varies significantly if the experiment is repeated under the exact same conditions [7]. This indicates a statistical variation of the breakdown process which was also discussed by Boksiner and Leath [8]. Thereby, the experiment cannot simply be switched off after a predefined time. To trap the experiment at a certain stage the switch off must be triggered by the current level. Problems arise experimentally when current measurements at short times (Fig. 1, sections (A) and (B)) activate the switching off procedure too early. Hence, a more sophisticated setup has to be used, that prevents the system to enter the prohibited experimental conditions which are indicated by the dashed rectangle in Fig. 1. Electrochemistry Communications 4 (2002) 1–4 www.elsevier.com/locate/elecom * Corresponding author. Tel.: +49211-6792-464; fax: +49211-6792- 218. E-mail address: hassel@elchem.de (A.W. Hassel). 1388-2481/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved. PII:S1388-2481(01)00260-0