Metal-oxide-semiconductor diodes containing C60 fullerenes for non-volatile memory applications

Abstract

For non-volatile memories, silicon-oxide-nitride-oxide-silicon or floating gate structures are used to store information by charging and discharging electronic states reversibly. In this article, we propose to replace the floating gate by C60 molecules. This would allow more defined programming voltages because of the discrete molecular energy levels and a higher resistance to tunneling oxide defects because of the weak electrical connection between the single molecules. Such C60 MOS diode structures are produced and their electrical properties are analyzed regarding current transport and charging mechanism of the molecules. To create the MOS structures, C60 molecules (5% of a monolayer) are evaporated onto a part of a clean silicon wafer and covered by amorphous silicon in situ in an ultra high vacuum system. Then the wafer is oxidized in wet atmosphere at just 710 °C through the C60 layer. The goal is to produce a clean oxide above and under the molecules without destroying them. Aluminum gate contacts are defined on top of these layers to perform complementary capacitance voltage (CV) and current voltage (IV) measurements. First, the gate voltage is swept to analyze the injection current, then CV measurements are performed after each sweep to analyze the charge state of the C60 layer and the oxide quality. Reference diodes without C60 on the same wafer show an identical Fowler-Nordheim (FN) tunneling behavior for currents injected from silicon or from aluminum, respectively. In the CV curves, no pronounced flatband voltage shift is observable. In diodes with C60, for negative gate voltages, a classical FN tunneling is observed and compared to theory. The electron injection from silicon shows a different tunneling current behavior. It starts at a lower electric field and has a smaller slope then a FN current would have. It is identified as a trap-assisted tunneling (TAT) current caused by oxidation-induced traps under the C60 layer. It is modeled by an established analytic TAT model which reproduces the data with a trap energy of 1.8 eV below the oxide conduction band. In the CV measurements the negative voltage IV sweeps result in a shift of the flatband voltages to more negative values. Positive voltage IV sweeps shift the CV curves back onto the starting curves. This proves positive charge trapping in the oxide which results in a non-volatile memory behavior for the diodes with C60.

Using a functional C84 monolayer to improve the mechanical properties and alter substrate deformation

“…Therefore, this monolayer fullerene on Si thin lm is proposed to scale down the size of current electronic device in a nano device, in contrast to the SiC substrate. In addition, the dramatic change in properties makes possible not only in the application of electronic device but also in that of optoelectronic, 16 batteries, 17 new generation photo-resists, 18,19 nonvolatile memories, 20 and cutting tool material.…”

Non-volatile memory devices with redox-active diruthenium molecular compound

“…Redox molecules for this application hold promise and have been demonstrated with fullerenes, 9, 12, 10 ferrocenes, 8, 1113 and porphyrins. 8 The redox-active molecule diruthenium(II,III)tetrakis(2-anilinopyridinate) (now referred to as Ru 2 , see Figure 1c) offers accessibility to multiple redox states 14, 15 and can be potentially exploited for multilevel programmability.…”

“…To scale the charge trapping layer to nanometer dimensions, it is appealing to use a discrete charge storage layer that is based on nanocrystals [2][3][4], nanoparticles [5], nanographene [6], or organic molecules [7][8][9][10][11].…”

“…Molecules that undergo reduction-oxidation (redox) reactions are attractive as the charge trapping sites because they can access additional charged states when the system's potential is altered. Redox molecules for this application hold promise and have been demonstrated with fullerenes [9,10,12], ferrocenes [8,11,13], and porphyrins [8]. The redox-active molecule diruthenium(II,III)tetrakis(2-anilinopyridinate) (now referred to as Ru 2 , see figure 1(c)) offers accessibility to multiple redox states [14,15] and can be potentially exploited for multilevel programmability.…”