For additional information, please see http://www.deas.harvard.edu/matsci/people/aziz/aziz.html.

MATERIAL SCIENCE:

Fabricating Nanopores and other Nanoscale Features

The ability to manipulate matter on the single-digit nanometer length scale is critical for probing biopolymers. But more generally, sculpting nanostructures using feed-back control represents an entirely new approach to crafting nanostructures that has profound implications for the development of a wide range of nanoscale devices. Understanding the physical and chemical mechanisms of ion beam nanosculpting therefore presents an exciting challenge to materials scientists. The demonstrated ability to continuously monitor changing dimensions while varying experimental parameters provides a unique opportunity to develop and test microscopic models to account for the observed materials phenomena.

The mechanisms underlying the ability of an ion beam to cause morphology changes include (i) the removal of material from the surface by the well-known ion sputter effect ("atomic sand-blasting") (1), as well as two processes that have been proposed only very recently, our understanding of which is under rapid development: (ii) surface transport caused by the creation of adsorbed species ("adatoms") on the surface that can diffuse rapidly (2), and (iii) collective motion such as viscous flow through the bulk of the film with ion-induced changes in viscosity, stress generation, and lateral swelling (3).

Our measurements of nanopore morphology evolution so far are consistent with a picture in which incoming ions not only sputter away, or remove, atoms from the surface of the film and the edge of the nanopore, but also create an adatom supersaturation (4). A balance between ion-induced creation and annihilation, as shown in Figure 1, governs the evolution of the adatom concentration.

Figure 1. The concentration of surface adatoms C(r,t), is governed by the two dimensional diffusion equation shown below, where r and t are surface position and time, D is the adatom surface diffusion coefficient, and F is the incident ion beam flux.

This model explains the evolution of nanopores under steady state irradiation by 3 keV Ar+, as shown in Figure 2. Some of the transient behavior that is observed upon ion beam pulsing is also explained by this picture. When the beam is switched off, the ion-induced annihilation mechanism (the last term in Figure 1) goes away, and more adatoms reach the nanopore edge and fill the pore than with the ion beam on. The result is an increased efficiency of pore closing per incident ion (bottom trace in Figure 2).

 

Figure 2. Ion beam sculpting, flux dependence. Pore area vs. total dose for samples exposed at different instantaneous fluxes, F, to a continuous beam (green points), or a pulsed beam (black points). The plotted orange curves are predicted from the diffusion model under steady-state conditions.

The temperature dependence of ion beam sculpting is also consistent with this picture, as shown in Figure 3 of our Nature paper (4). At low enough temperature, surface diffusion is quenched out, the nanopore's adatom collection zone is extremely small, and the process is dominated by the ion beam scraping away the edge of the pore, thereby opening it as originally envisaged. As the temperature goes up, the adatom collection zone gets bigger and bigger, eventually overwhelming the edge-scraping effect and leading to pore closure.

There are many things we don't yet know and we need to understand to design robust recipes for nanopore manufacturing with exacting tolerances.

· How can we measure, understand, and control the evolution of the nanopore's cross-sectional profile?

· Under what conditions is a circular nanopore shape stable, and under what conditions is it unstable? Some of our nanopores develop an irregular morphology for reasons we do not understand.

· How can we understand the general transient response of the nanopore when the ion flux is varying in time, and how can we use this to better manipulate materials on nanoscale dimensions?

· How are the ion beam's effects on surface and bulk defects, surface energetics, and surface diffusion and incorporation barriers different in amorphous materials (such as Si3N4 and SiO2) and crystalline materials (such as Al, which we hope to be able to sculpt into electrical leads to the nanopore)?

· How can we measure, understand, and control the formation of electrically active defects (which are deleterious to our electrical signals) during ion irradiation?

· Under what conditions should ion-stimulated viscous flow come into play? These effects have not yet been discerned in nanopore evolution, but have been seen in other aspects of ion irradiation of some materials under different ion irradiation conditions. Is different behavior to be expected when this mechanism is operating?

· At large surface slope (e.g. at the nanopore edge) and large adatom concentration, nonlinear effects are expected to come into play. How do we characterize them experimentally and understand them theoretically?

· Can we use and extend our model to understand and control the evolution of arbitrary shapes, such as slits, hillocks, and corrugations, under ion beam stimulation?

 

References

1. P. Sigmund, "A Mechanism of Surface Micro-Roughening by Ion Bombardment", J. Mater. Sci. 8, 1545 (1973).
2. J. Erlebacher, M. Aziz, E. Chason, M. Sinclair, and J. Floro, "Spontaneous Pattern Formation on Ion Bombarded Si(001)", Phys. Rev. Lett. 82, 2330 (1999).
3. M.L. Brongersma, E. Snoeks, T. van Dillen, and A. Polman, "Origin of MeV Ion Irradiation-Induced Stress Changes in SiO2", J. Appl. Phys. 88, 59 (2000).
4. J. Li, D. Stein, C. McMullan, D. Branton, M.J. Aziz, and J. Golovchenko, "Nanoscale Ion Beam Sculpting", Nature 412, 166 (2001).