The recent discovery that with certain dopants hafnium oxide can be crystallized in a form that exhibits ferroelectricity has opened up a range of device possibilities(1). We have chosen to work within the Hf/Zr system for our initial studies. The advantages of the Hf/Zr doping system include the similarity in precursors which can enable several running modes for the ALD process including a co-pulsing scenario, as well a cycle ratio scenario for controlling the Zr doping. Because Zr and Hf are so similar structurally as well as chemically, the isologous precursors we employ have essentially identical delivery characteristics, including vapor pressure, relative sticking coefficient and relative decomposition point. Thus we expect the process to be capable of maintaining a highly consistent composition over a broad range of topology. In addition, the relative range of doping for which a strong remnant polarization is observed is much broader for the Hf/Zr system than for other dopants such as Al, Si, etc. which means this system should be more tolerant of small changes in composition than the other doping schemes. Generally, the ferroelectric phase of doped Hf oxide is thought to be a non-centrosymmetric orthorhombic phase that is co-crystallized with some other observable phases. In our studies we have found a strong correlation between the polarizability we observe and signals attributed to a tetragonal phase by grazing incidence X-ray diffraction using synchrotron radiation(2).

In terms of electrical performance, we have shown reasonable polarization characteristics for films as thin as 5nm, and demonstrated the feasibility of using these films for multiple potential applications. We have explored the possibility of incorporating these films as gate dielectrics for a ferroelectric memory and have also demonstrated that we can fabricate a transistor exhibiting a steep subthreshold slope (<60mV/dec.), due to a negative capacitance effect observed as the transistor switches and sweeps through the subthreshold region simultaneously(3). Other potential uses for these films include the possibility to create multi-domain ferroelectric FETs that can be used within a neuromorphic device to mimic synapse behavior. This talk will review and update our work to date on these films and their potential uses.

References:

1. T. S. Böescke, J. Müller, D. Bräuhaus, U. Schröder and U. Böttger, in Technical Digest of the International Electron Devices Meeting, p. 24.5.1 (2011).
2. S. Dey, K. Tapily, S. Consiglio, R. D. Clark, C. S. Wajda, G. J. Leusink, A. R. Woll, P. Sharma, S. Datta and A. Diebold, in Frontiers of Characterization and Metrology for Nanoelectronics, E. M. Secula and D. G. Seiler Editors, p. 223, Monterey, CA (2017).
3. P. Sharma, K. Tapily, A. K. Saha, J. Zhang, A. Shaughnessy, A. Aziz, G. L. Snider, S. Gupta, R. D. Clark and S. Datta, in 2017 Symposium on VLSI Technology, p. T154 (2017).