The proposed work will experimentally test many theoretical predictions in nucleation science, and provide data on the evolution of nanoscale nuclei in glassy phase change materials under various conditions. The nucleation process is ubiquitous in nature, and the knowledge obtained from this work will be applicable to essentially any material systems involving phase transformation. One example is phase change memory technology, since nucleation and crystallization are the key processes that govern fast data writing and long-term retention. A pump-probe laser technique, atomic force microscopy, and membrane electrical thermometry will be combined for understanding nucleation in glassy phase change materials such as AgInSbTe, Ge2Sb2Te5, and GeSb. The experimental results will be correlated with theoretical model calculations. First, some predictions of the general nucleation theory will be tested experimentally. The temperature-dependence of steady-state subcritical nuclei distribution and the dependence of nucleation rate on pre-existing subcritical nuclei will be explored. Second, the effects of interfaces, compositional variation, and thermal history on the evolution of nuclei in phase change materials will be investigated. Since the distribution of nuclei is one of the most important factors that influence the speed and stability of a phase change memory, FTEM will be used to determine the concentration of nuclei in actual memory devices. Third, as a relatively exploratory work, electrical thermometry will be employed to characterize the glass transition and nucleation at very high heating rates. Important thermal parameters are dependent on temperature and heating rate, but conventional techniques can only provide heating rates many orders of magnitude lower than the rates involved in actual memory operations. The data from this study will provide crucial information for technology, and serve as fundamentally illustrative examples of transient nucleation. The principal task of the proposed work is the investigation of the fundamental science of nucleation.
The project addresses basic research issues in a topical area of materials science with technological relevance, and is expected to provide unique opportunities for graduate and undergraduate training in an interdisciplinary field. This research project is also expected to have broader impacts through the training of women and men leaders in this research field, through the wide dissemination of the findings of this research through publications. One or two undergraduates will carry out portions of this work as their senior theses. Notably, the experimental results can be used in textbooks and lectures as excellent examples to illustrate the nucleation theory, which is an important subject of materials science education. Qualified students will be identified and advised, in particular through the SURGE (Support of Under-represented Groups in Engineering) program at the University of Illinois.
The seminal contribution of our research has been the demonstration that fluctuation transmission electron microscopy (FTEM) can detect the presence of ordered nuclei (crystallites) on the 1 - 3 nm scale in amorphous materials. We have applied FTEM to the study of phase change materials that can be rapidly and reversibly switched between their amorphous and crystalline phases. These phase change materials have enabled the technology of rewriteable optical discs (CDs, DVDs) and the development of next-generation, non-volatile random access memories. Phase transformation generally begins with nucleation, in which a small number of atoms organize into a new structural symmetry. Although nucleation theory is widely used in physics, chemistry, and materials science, many of its predictions have not been verified experimentally because it has been difficult to detect and analyze nanometer-scale nuclei. Now, however, FTEM enables the study of nanoscale order. Time-resolved reflectivity measurements during pulsed laser crystallization reveal the rates of solid-phase transformation, while FTEM detects the nanoscale order in the amorphous phase prior to crystallization. The evolution of subcritical nuclei in the amorphous phase change material Ge2Sb2Te5 (GST) has been analyzed as a function of nitrogen alloying and thermal annealing. The results reveal that nitrogen doping alters the crystallization behavior of amorphous GST via a change in the ability of the material to generate subcritical nuclei that can rapidly coarsen into nucleation centers when the material is heated. We investigated the effect of low temperature annealing on the subsequent nucleation behavior of amorphous AgIn-incorporated Sb2Te, another phase change memory material. The effect of annealing is observed to saturate: there is no further reduction in nucleation time or increase in nanoscale order for annealing at 100 °C beyond 3 hours. This result supports the general prediction of classical nucleation theory that the size distribution of subcritical nuclei increases from the as-deposited state to a quasi-equilibrium. The evolution of nanoscale order in amorphous Ge xSe1−x alloys has been analyzed using FTEM. Two distinct structural signatures that were detected behave independently as a function of composition. A strong signature of nanoscale order that appears in Ge-rich alloys (x > 0.40) diminishes rapidly in Se-rich compositions. However, a different signature of order appears only for compositions in the middle range x = 0.30–0.53. We infer that two distinct forms of structural ordering occur in amorphous Ge xSe1−x, one form among pure Ge tetrahedra in Ge-rich compositions, and a second form among GeSe4 tetrahedra in the Se-rich compositions. We report that the FTEM covariance, computed at two non-degenerate Bragg reflections, is able to distinguish different regimes of size vs. volume fraction of order. We use a Monte-Carlo approach to compute different regimes of covariance, based on the probability of exciting multiple Bragg reflections when a STEM nanobeam interacts with a volume containing ordered regions in an amorphous matrix. We performed experimental analysis on several sputtered amorphous thin films including a-Si, nitrogen-alloyed GeTe and Ge2Sb2Te5. The samples contain a wide variety of ordered states. Comparison of experimental data with the covariance simulation reveals different regimes of nanoscale topological order. This approach also provided evidence for a remarkable preferred orientation of nanoscale order at the surface of amorphous Ge2Sb2Te5 films. The nanoscale crystal nuclei in an amorphous Ge2Sb2Te5 bit in a phase change memory device were evaluated by FTEM. The quench time in the device (∼10 ns) afforded more and larger nuclei in the melt-quenched state than in the as-deposited state. However, nuclei were even more numerous and larger in a test structure with a longer quench time (∼100 ns), verifying the prediction of nucleation theory that slower cooling produces more nuclei. It also demonstrates that the thermal design of devices will strongly influence the population of nuclei. Broader Impacts: Glass formation is still one of the unsolved mysteries in physics. We now expect that the fundamental investigation of the glass-forming ability in the germanium-selenium chalcogenide system will add a new chapter to the field of glass science. FTEM studies of a broad range of good glass-forming and poor glass-forming compositions will provide valuable insights. Two graduate students completed their Ph.D. degrees with the support of this grant. They have implemented FTEM on two generations of electron beam microscopes in the Materials Research Laboratory on the UIUC campus. They developed software to control the data acquisition of nanodiffraction patterns obtained in a scanning transmission electron microscope and to analyze these data. These resources are part of the infrastructure of the MRL. Our fundamental FTEM studies of phase change materials and devices supplied to us by our IBM collaborators is likely to benefit them in their search for improved phase change device designs. The potential impact on society will be the energy saving, time saving, and convenience which will be afforded by the development of commercially viable, high speed, nonvolatile phase change random access memory devices.
