Recent developments in nanostructures have brought to light exceptional electromagnetic, thermal and optical properties of a class of foam-like nanostructures formed of disordered intertwined structural units (nanowires, nanobelts, nanotubes). Such disordered assemblies are named turfs. Applications include thermal switches, flat panel displays, hard discs drives, and, chemical and biological sensors. Although the mechanical properties are usually not the primary service characteristic of turfs, they are nevertheless of paramount importance. Irrespective of application, the turfs are often subjected to mechanical loads, either as service load as in thermal switches, or, as accidental contacts. Under externally forced deformation, the nano-topology of the turf changes, which, in turn, affects all the other effective properties: electrical, thermal, optical, sensing and permeability. We will develop an integrated approach to the problem: multiscale modeling, nanomechanical experiments, and, nanostructure characterization, with the following objectives: Understanding and quantification of the behavior of turfs as materials on the basis of the physical and geometrical properties of the individual units and their collective behavior in the assembly. Development of the nanoscale characterization methods that reveal the relevant parameters of the nanostructure. Practical technological impact of the project is that the results will enable rational design of nanoturfs tailored for particular application in sensors, thermal switches and other devices. The REU component of the program is carefully structured and includes assessment methods, developed and proven at the Center for Teaching and Learning at WSU. Our pilot student mentoring program will provide graduate students with mentoring experience a skill that PhD graduates need, but is sorely missing in most graduate programs.
Carbon nanotubes (CNTs) are often grown from catalyst materials on substrates. Depending on the growth conditions the nanotubes can form structures that range from brushes with straight parallel bristles, to tortuous hairball shapes. One common structure is the turf, where arrays of tubes are nominally end-to-end close to vertical, but are locally tortuos and interconnected. The applications of such materials include thermal switches, nanoscale sensors, hard disks and flat panel displays. By design or by accident, turfs are subjected to mechanical loading/deformation, which changes their properties. If we are to design a turf with desired properties for a particular application, we must know what nano-structure to design. In this project we asked a basic question regarding a nanotube turf: Given the nano-scale structure and geometry of a turf (How dense is the turf? How curved are the tubes? How dense are the contacts between them? ), can one predict mechanical properties, and conversely, given the loading/deformation, can one predict the changes in the nano-structure? Moreover, measuring the nano-scale structure is very difficult, so we aimed at designing the methods for such measurements. At the time the project funded a model capable of predicting the mechanical properties of the turf from its nano-structure was not available. In similar bulk materials (such as foam rubber or other cellular solids) there are well established models that relate density, shapes of ligaments, and contact points to the resulting stiffness of the structure (imagine patting your hair and comparing that stiffness to the stiffness of one individual hair). However, preliminary experimental results have demonstrated that the CNT turfs do not follow these models of bulk materials, and therefore new ways of describing, measuring and predicting the properties are needed for these unique structures. CNT turfs were grown using established methods, and then the structure characterized using scanning electron and transmission electron microscopy. The mechanical response of the turf was measured using nanoindentation, where a tip is pushed into the turf while measuring the load and penetration depth. In most cases the turfs are elastic (like pressing your hand onto a cushion). Because most CNT turfs are almost 98% empty space (very low density of actual tubes, with a lot of space between each one), the properties have significant scatter from location to location. Our first significant result was to show that the scatter was uniform; measuring in one local region was not too different than measuring at a different point. This means that turf can be described as a material, rather than as a structure consisting of individual components. We then showed that if the density variation in the turf was small, the point-to-point noise in the measurement of the stiffness was greater than the density variation. However, when turfs were densified by a factor of 10 to 50, the stiffness increased by 10 to 50 times, suggesting that density and stiffness are directly related. Methods to characterize the turf using widely available electron microscopy tools were developed, and we hope they will be adopted by other researchers to quantify the structure of a turf. Model for simulations of CNT assemblies that form the turf was developed. It includes bending/buckling stiffness of the tubes as well as the adhesion between them. The analysis of simulation results and comparison to experiments enabled us to establish the connection between the nano-scale geometry and the mechanical properties (stiffness, strength). In particular, it allowed us to understand the mechanism of failure as collective layer buckling. These results were widely distributed to the science community through presentations and papers, and have helped shape the current discussion of controlling the properties of a nanomaterial structure when millions of nanotubes are used together to make useful devices, such as sensors, filters, and electrical contacts.
