Orientationally ordered glasses via controlled deposition

The free surface of a liquid or solid is not just the bulk phase, truncated. The inhomogeneity at the interface generates a local state quite distinct from the bulk phase: one, typically, with its own distinct structure. Just how different the interface and bulk structures are has been a subject of extensive study, at least since Gibbs (1) described surface segregation in mixtures. Surface reconstruction (2), metal–insulator transitions (3), and interfacial molecular alignment (4), along with segregation, are examples of the variety of interfacial organization that has been observed (Fig. 1). Traditionally, this structural variety has not been regarded as a source of additional types of order for bulk material design since it is restricted to an interfacial layer typically no more than 2 to 3 molecules in thickness. But what if it was possible to assemble bulk phases using interfacial structures, for example, by the sequential addition of surface layers, one on top of another, to build up a bulk phase (Fig. 1) while retaining the distinctive arrangements of molecules that had been stabilized by the surface? This would represent an extraordinary expansion in the routes to material design and fabrication. The whole spectrum of interfacial structures would then represent an untapped library of structural and, hence, material property options.

Fig. 1.

The accumulation of various types of surface organization via vapor deposition. Three different types of surface-associated structures are depicted: surface segregation, molecular alignment, and surface reconstruction. Also sketched is an equilibrium liquid with a surface and the deposited film. In both cases, the structured material is indicated with shading.

The immediate obstacle to this ambitious program is how to retain a microscopic structure, stabilized by virtue of being at a surface, once the material is no longer at a surface but incorporated into the bulk of the material. In PNAS, Bishop et al. (5) report an experimental demonstration of just how this remarkable trick can be pulled off. An elongated organic molecule, posaconazole, was vapor deposited onto a silicon surface at a temperature of 324 K, just below the molecule’s glass transition temperature T_g, to produce an amorphous film. Using grazing-incidence wide-angle X-ray scattering and optical birefringence (6), Bishop et al. (5) establish that the posaconazole molecules were oriented with their long axes ∼33° from the substrate normal and were organized into layers parallel to the substrate surface. This structure resembles that found in smectic liquid crystal phases but posaconazole does not have any liquid crystal phases. It does, however, exhibit a similar smectic-like structure at its equilibrium liquid surface, as established in ref. 5. While this ordering at the equilibrium surface is expected to extend a few molecular layers into the liquid (where a molecular layer is ∼3 nm), ordered films have been deposited up to 1 μm in thickness (5), i.e., macroscopic length scales. By this measure, a short-range equilibrium correlation is converted into long-range order through the agency of the nonequilibrium fabrication process. This is the essence of the results of Bishop et al. (5).

The explanation of the structure of the deposited films is based on a previously proposed surface equilibration mechanism (7–9). The key idea is that the mobility at the surface of the organic glass can be much higher than that in the bulk (10). This means that a deposited molecule will be able to equilibrate to the (metastable) equilibrium structure of the liquid surface (i.e., a smectic-like ordering) until it is buried by subsequent depositions. Then, the mobility will abruptly decrease, trapping the smectic ordering by kinetic arrest and preventing it from relaxing to the isotropic structure of the bulk. This is the “trick” discovered by Bishop et al. (5)—the delicate adjustment of substrate temperature and deposition rate to ensure that enough time is allowed for equilibration and, thus, ordering, at the surface while not allowing so much mobility and/or time that this order is subsequently relaxed once it is covered by subsequent deposition. This mechanism implies that decreasing the deposition rate or increasing the deposition temperature (at least up to some sub-T_g threshold) will allow for greater equilibration of the surface and hence an increased degree of ordering, with the upper bound on this order set by the equilibrium interface structure. In this way, it is possible to “dial up” the degree of smectic order desired through small variations in the deposition conditions. These general predictions are confirmed experimentally in ref. 5.

Along with its exciting promise of materials with continuously controllable properties, ref. 5 also points to some ongoing challenges. An important issue is the thermal stability of the deposited film. Since the stability of the order is ensured only by slow kinetics, we should expect the onset of rapid and irreversible relaxation of the smectic order when heating above T_g. This is indeed what is observed (5). What is not yet clear are the magnitude and temperature dependence of the structural relaxation in the film at or below T_g. Sub-T_g relaxation, referred to as “aging” in the glass literature (11), has been extensively studied in isotropic materials (12), but there is still much to learn about how the presence of order influences the process. This issue is of considerable importance in establishing the working range of the materials generated via the surface equilibration mechanism. A second challenge is the time and cost required to fabricate a film via vapor deposition. At a deposition rate of 0.2 nm/s it would take over 1 h of continuous deposition to produce a 1-μm-thick sample and require a vacuum chamber, pumps, and an oven to vaporize the depositing molecule. While there are applications of amorphous films that require thicknesses only on the order of tens of nanometers, we simply note that vapor deposition can be a slow and expensive mode of fabrication. One possible resolution of these logistic difficulties would be to switch from vapor deposition to solution deposition, i.e., precipitation. This possibility hinges on whether similar structural selection can be achieved through precipitation. The results of a recent modeling study (13) on the precipitation of a glass former provide some cause for optimism in this regard. The modeling indicated that suitably controlled precipitation can indeed produce the very low-enthalpy glasses previously obtained by vapor deposition (8, 14, 15).

Outstanding technical challenges aside, the study by Bishop et al. is an elegant demonstration of a strategy for identifying and manipulating surface-sourced material structures.

Outstanding technical challenges aside, the study by Bishop et al. is an elegant demonstration of a strategy for identifying and manipulating surface-sourced material structures. In addition to fabricating molecularly aligned glasses, it raises the possibility of incorporating other types of surface order (i.e., compositional segregation, surface reconstruction, etc.) into deposited films. Ref. 5 highlights the merit of seeing in the structure of amorphous materials not the absence of crystallinity by which it is traditionally characterized, but, instead, the presence of a surfeit of different types of order. For over a decade, Mark Ediger (15) and his coworkers have pioneered the careful mapping of how the enhanced kinetics and, more recently, equilibrium surface organization, of a free surface can be employed to access a subset of this library of potential structures, a subset unobtainable by standard bulk cooling protocols. In this current paper (5), they take another important step in putting together a guidebook for material design based on the sequential accumulation and kinetic stabilization of surface-generated structures.