Ultrathin membranes are ubiquitous in nature, serving as both a means to encapsulate and protect living cells from their environment and to regulate the myriad of biochemical processes necessary to sustain life. Ultrathin membranes also have important industrial applications including gas separation, sensors and fuels cells. However, although organic membranes can be readily grown in a test tube, their inability to withstand even modest temperatures beyond 100ºC limits their applications.

Several methods have been developed to grow thermally stable membranes from inorganic materials such as carbon or silicon. But most involve multiple time‑consuming fabrication steps such as photolithography, or result in membranes that are mechanically fragile and difficult to handle. Izumi Ichinose and colleagues at the National Institute for Materials Science in Tsukuba, Japan describe a simple technique exploiting self‑assembled organic membranes as templates for the growth mechanically robust inorganic membranes.1

Fig. 1: A copper grid immersed in a solution of surfactant molecules and then removed to dry, results in the self-assembly of a reversed bilayer membrane in each hole (left). An ultrathin film of inorganic material, such as platinum, can then be grown onto of this bilayer, which is subsequently removed by washing in water (right).

The authors’ procedure begins by dipping a copper grid of square 7 μm wide holes into a solution of surfactant molecules and water, to capture a small volume within the holes of the grid. As the water molecules in this solution evaporate, the surfactant molecules gradually self–assemble to form a thin ‘reversed’ bilayer membrane— similar in structure to that of a conventional biological membrane but with its constituent molecules oriented with their hydrophilic end in the centre rather than at the surface of the film—known as dried foam film (DFF). The DFF is then used as a support for the growth of an inorganic film by conventional physical vapour deposition (PVD). In this way the authors grew pure, free‑standing membranes of carbon, silicon, platinum, iron and cadmium selenide with thicknesses ranging from a few to hundred nanometres.

The remarkable feature of this approach is not its versatility for producing membranes of almost any material grown by PVD, but that it works at all.

“Initially, we were afraid that our DFFs would be thermally unstable. However, we found that some of them were stable at above 150ºC,” says Ichinose. “We now realize that the properties of DFFs are more like those of polymers and not soft like conventional bilayer membranes. This is quite surprising, although the origin of their stability is still unclear. But, they are stable.”

The authors are searching for better and cheaper surfactant molecules for making DFF films in addition to studying their potential for growth of porous diamond‑like carbon membranes for applications that include gas separation and zinc oxide membranes for fabricating novel optical devices.