Thin-Film Active Materials

Department of Mechanical and Aerospace Engineering University of California, Los Angeles Los Angeles, California 

In the Active Materials Laboratoiy at the University of California, Los Angeles (UCLA), research is being conducted on a wide range of active materials, 

Although shape-memory materials have unique attributes, the bandwidth (1 Hz) of bulk materials limits their apphcability in many situations. The bandwidth limitation is due to relatively slow coohng processes related to bulk materials. However, thin-films have very large surface-to-volume ratios so their cooling times can be orders of magnitude higher. 

Recently at UCLA, thin-film shape-memory alloy bandwidths higher than 100 Hz have been demonstrated (Shin et al., 2004) theoretical predictions approach the kHz regime. These large bandwidths, along with associated large stress and strain output, provide films with enormous values of power-per-unit mass (e.g., 40 kW/kg). The relatively large specific power (e g., compared to 100 W/kg for other milli- or micro-motors) provides unique opportunities for small-scale apphcations (e.g., miniature motors and heart valves). 

One pump design with dimensions in centimeters is used to articulate the nose cone of a small missile system (Shin et al., in press). Another application is a pumping motor that can move small amounts of fluids for various laboratory (e.g., laboratory-on-a-chip) and biomedical applications. For example, the pump can be used as an embeddable drug-dehvery system. Coupled with appropriate sensors, this miniature (submillimeter) pump could someday replace a human 

---

INVESTIGATION OF THIN-FILM ELECTRODE MATERIALS AS CATHODIC ACTIVES FOR POWER SOURCES... 

Thin films of materials play an important role in modern electronics. Langmuir-Blodgett films (Mort, 1980) are prepared from surface-active molecules that adsorb at the surface of an aqueous solution. The surface is gently compressed to produce a close-packed monolayer, which can be coated on appropriate substrates by dipping at constant surface pressure. The layers so produced are of a precise thickness, and repeated dipping gives multilayers. For example, diacetylenes and other polymerizable... 

Tungsten trioxide thin films activated by gold layers have also been used for NOj. detection.115 This material possesses excellent sensitivity towards nitric oxide and nitrogen dioxide. An automobile exhaust gas NO sensor that uses a gold-platinum alloy electrode has also been reported.116... 

WO3 thin films activated by An layers have been used [625]. This type of material possesses excellent sensitivity towards NO and NO2 gases. An exhaust gas NOj sensor that uses a platinum-gold alloy electrode to selectively remove oxygen but not NO was also reported [626]. 

Table 3.1 Availability of materials for thin-film active layers. Adapted from Keshner and Arya, 2004 with permission from NREL... 

A belief that solid interfaces are easier to understand than liquid ones shifted emphasis to the former but the subjects are not really separable, and the advances in the one are giving impetus to the other. There is increasing interest in films of biological and of liquid crystalline materials because of the importance of thin films in microcircuitry (computer chips ), there has been in recent years a surge of activity in the study of deposited mono- and multilayers. These Langmuir-Blodgett films are discussed in Section XV-7. 

Amorphous Silicon. Amorphous alloys made of thin films of hydrogenated siUcon (a-Si H) are an alternative to crystalline siUcon devices. Amorphous siUcon ahoy devices have demonstrated smah-area laboratory device efficiencies above 13%, but a-Si H materials exhibit an inherent dynamic effect cahed the Staebler-Wronski effect in which electron—hole recombination, via photogeneration or junction currents, creates electricahy active defects that reduce the light-to-electricity efficiency of a-Si H devices. Quasi-steady-state efficiencies are typicahy reached outdoors after a few weeks of exposure as photoinduced defect generation is balanced by thermally activated defect annihilation. Commercial single-junction devices have initial efficiencies of ca 7.5%, photoinduced losses of ca 20 rel %, and stabilized efficiencies of ca 6%. These stabilized efficiencies are approximately half those of commercial crystalline shicon PV modules. In the future, initial module efficiencies up to 12.5% and photoinduced losses of ca 10 rel % are projected, suggesting stabilized module aperture-area efficiencies above 11%. 

In most cases, CVD reactions are activated thermally, but in some cases, notably in exothermic chemical transport reactions, the substrate temperature is held below that of the feed material to obtain deposition. Other means of activation are available (7), eg, deposition at lower substrate temperatures is obtained by electric-discharge plasma activation. In some cases, unique materials are produced by plasma-assisted CVD (PACVD), such as amorphous siHcon from silane where 10—35 mol % hydrogen remains bonded in the soHd deposit. Except for the problem of large amounts of energy consumption in its formation, this material is of interest for thin-film solar cells. Passivating films of Si02 or Si02 Si N deposited by PACVD are of interest in the semiconductor industry (see Semiconductors). 

Electroless reactions must be autocatalytic. Some metals are autocatalytic, such as iron, in electroless nickel. The initial deposition site on other surfaces serves as a catalyst, usually palladium on noncatalytic metals or a palladium—tin mixture on dielectrics, which is a good hydrogenation catalyst (20,21). The catalyst is quickly covered by a monolayer of electroless metal film which as a fresh, continuously renewed clean metal surface continues to function as a dehydrogenation catalyst. Silver is a borderline material, being so weakly catalytic that only very thin films form unless the surface is repeatedly cataly2ed newly developed baths are truly autocatalytic (22). In contrast, electroless copper is relatively easy to maintain in an active state commercial film thicknesses vary from <0.25 to 35 p.m or more. 

Usually they are employed as porous pellets in a packed bed. Some exceptions are platinum for the oxidation of ammonia, which is in the form of several layers of fine-mesh wire gauze, and catalysts deposited on membranes. Pore surfaces can be several hundred mVg and pore diameters of the order of 100 A. The entire structure may be or catalytic material (silica or alumina, for instance, sometimes exert catalytic properties) or an active ingredient may be deposited on a porous refractory carrier as a thin film. In such cases the mass of expensive catalytic material, such as Pt or Pd, may be only a fraction of 1 percent. 

Yet another alternative is the thin-film solar cell. This cannot use silicon, because the transmission of solar radiation through silicon is high enough to require relatively thick silicon layers. One current favourite is the Cu(Ga, InjSci thin-film solar cell, with an efficiency up to 17% in small experimental cells. This material has a very high light absorption and the total thickness of the active layer (on a glass substrate) is only 2 pm. 

---