Previous article
, dedicated
to so-called "traditional" thin films, has defined what is meant by
a thin film, i.e. a deposit generally less than 5 µm thick, usually
obtained in a vacuum. This deposition is global, covering the entire
surface of the substrate, and the patterns are obtained by a subtractive
method, such as chemical etching.
Thin films have been used for many decades in a wide range of
applications. The oldest and still most widespread are probably optical
applications. Metallurgy and photography have also used thin films,
but the most modern applications now include chemistry, biochemistry
and medicine, around sensors of all kinds, from gases to DNA (biochips).
Interconnection, in the form of substrates equipped with deposited
conductor tracks, is also one of the preferred fields for thin films.
Although in the 1980s thick films were dethroned in this particular
field, the concept of the "multichip module" (MCM), which emerged
around 1985, has, thanks to a number of technical innovations, restored
a certain interest in thin films, which have regained the opportunity
to fully exploit their integration capacities in these applications.
But history only repeats itself. Monolithic integration is constantly
evolving and gaining ground on heterolithic integration, pushing it
to evolve in its turn. If MCMs are today's (heterolithic) answer to
the limitations engendered by (monolithic) ASICs, then (heterolithic)
SOPs will be tomorrow's answer to the (monolithic) SOCs that are trying
to take MCMs' place today. There's no doubt that thin-film electronics
will play a key role in SOPs - superhybrids that combine high-density
electrical and optical interconnections, active and passive components,
cooling devices and, why not, electromechanical microsystems.
Note :
The author would like to thank Thierry Lemoine, head of the Ceramics
and Packaging department at Thalès TRT, and head of the joint LABCOM
BGCC/TRT laboratory, and Sylvain Schmitt, CNRS/IN2P3 engineer, for
their help in writing and editing this document.