All entries for Sarah Paydavosi

• Print This Page
• Download as PDF

Integration of Small Organic Molecules in Flash Memory Devices

• Authors: S. Paydavosi, K. Aidala, P. Brown, G. J. Supran, P. Hashemi, T. P. Osedach, J. L. Hoyt, V. Bulović
• Sponsorship: SRC/FCRP MSD

As demands for high storage density, high chip memory capacity, and decreasing process costs continue to mount, conventional flash memory has found it challenging to continue scaling because of the minimum tunnel oxide thickness and poor charge retention due to defects in the tunneling oxide, necessitating modification in the implementation of the flash memory technology ^{[1] [2]} .

Molecular organic materials exhibit fascinating electronic properties that motivate their hybridization with traditional silicon-based memory devices in order to continue memory scaling. A floating gate consisting of a thin film of molecules would provide the advantage of a uniform set of identical nanostructured charge storage elements with high molecular area densities (e.g., 8 × 10¹³ cm^-2, which can result in a several-fold higher density of charge-storage sites as compared to QD memory and even SONOS devices) ^{[3] [4]} . Additionally, the discrete charge storage in such nano-segmented floating gate designs limits the impact of any tunnel oxide defects to the charge stored in the proximity of the defect site.

In order to study the memory behavior of organic molecules, we inject electrons/holes into the molecules by applying negative/positive bias to a conductive atomic force microscopy (AFM) tip (Pt probe tip with 10-nm radius) in contact with the organic layer (Figure 1). During the charge injection phase, the tip is brought into contact with the sample surface by reducing the amplitude setpoint to 0.5 V. The stored charges within molecules can be detected from surface potential mapping of the sample by Kelvin force microscopy (KFM). Figure 1 shows the KFM image of the charged spots by applying 9V tip bias. The minimal temporal decay of injected charges and their corresponding lateral spreading indicate highly localized charge distribution, suggesting potential use of small organic molecules in multi level trap based molecular flash memory cells with high storage capacity. The calculated stored charge density within molecules is shown in Figure 2.



Figure 1: The schematic of the prepared sample and KFM image of the written charge dots into a 5-nm-thick molecular layer on a 4-nm-thick oxide layer.



Figure 2: Calculated stored charge density from the measured surface potential assuming uniform distribution of charge across the thickness of the trapping layer.




1. P. Pavan, R. Bez, P. Olivo, and E. Zanoni, “Flash memory cells-An overview,” Proc. IEEE., vol. 85, no. 8, pp. 1248-1271,1997. [↩]
2. International Technology Roadmap for Semiconductors, ITRS. (2007). [Online]. Available: http://www.itrs.net. [↩]
3. P. K. Singh, R. Hofmann, K. K. Singh, N. Krishna, and S. Mahapatra “Performance and reliability of Au and Pt single-layer metal nanocrystal flash memory under NAND (FN/FN) operation,” IEEE Trans. Electron Devices, vol. 56, no. 9, Sep. 2009. [↩]
4. S. Paydavosi, H. Abdu, G. J. Supran, V. Bulović, “Performance Comparison of Different Organic Molecular Floating Gate Memories,” IEEE Trans. Nanotechnology, vol. 10, no. 3, May 2011. [↩]

A Low-loss Squeezable Micro-Electro-Mechanical Switch

• Authors: S. Paydavosi, F. Yaul, A. Murarka, T. L. Andrew, A. Wang, J. H. Lang, V. Bulović
• Sponsorship: National Science Foundation

It has been known for several decades that polymers doped with conducting particles, for example silicone nickel nano-particles, will exhibit a dramatically decreasing resistivity as the polymer is compressed. It is possible that the conductivity will vary by 12 orders of magnitude over a 40% strain ^[1] . Such composites conduct via tunneling from particle to particle, and the tunneling currents grow exponentially as the particles become closer together. These composites have already been used in applications from tactile sensors to fuses.

In this study we use the composites as the active element in an electronically-controlled switch. The Squeezable electronically controlled switch, referred to here as a “squitch,” is shown in Figure 1. In this embodiment, the squitch is a three-terminal device, with its terminals labeled as per the comparable terminals in a MOSFET. The central component of the device is the doped polymer labeled “Squitch Material” connected to drain and source electrodes. As fabricated, the squitch material would be a poor conductor, permitting little if any electron current to flow from the source to drain. That is, the resistance of this conduction path would be very large, putting the squitch in an off state. By applying voltage to the gate electrode, either positive or negative, an electric field is developed between the gate and the source. This electric field causes the gate to be attracted to the source, thereby compressing the squitch material. As the squitch material is compressed in the vertical direction, it begins to conduct, putting the squitch in an on state.

Thus, the squitch is a voltage controlled conductor, much the same as a FET or a BJT but with very large on-to-off conduction ratio and subthreshold swing (S) < 60 mV/dec, which allows for more aggressive supply voltage scaling and improvement in the energy efficiency.



Figure 1: SEM image of nickel-silicone composite and cross section schematic of a three-terminal squitch. More than 7 orders of magnitude change in the conductivity of the composite was observed over a 20% strain.



Figure 2: The initial demonstration of the squitch, showing 4 orders of magnitude change in the current by applying voltage to the gate terminal.




1. D. Bloor, K. Donnelly, P. J. Hands, P. Laughlin, and D. Lussey, “A metal–polymer composite with unusual properties,” J. of Phys. D: Appl. Phys., 38 (2005) 2851–2860. [↩]