MOLECULAR COMPUTING: INTEGRATION OF MOLECULES FOR NANOCOMPUTING - Bio-Inspired and Nanoscale Integrated Computing - page 314

11.5. MOLECULAR GRAFTING FOR SILICON COMPUTING

We have described the intensive research endeavor in building switching and
memory using molecules. Direct integration of molecular switches and memory
units suffer from the uncertainty in chemical self-assembly. The works described in
this chapter so far attempt to alleviate this problem through fabrication and
design innovations. Nevertheless, all these solutions endeavor to replace semi-
conductor devices, at least partially, with molecular bundles. While lab demon-
strations have show promise, these solutions remain ambitious and not
immediately implementable in the industry, which has a considerable vested
interest in semiconductor. In this last section, we examine a very different
application of molecular computing: Control the properties of semiconductor
for better computing with molecules, instead of the direct use of molecules or
molecular bundles for switching and memory. The application is called molecular
grafting. We showed that a monolayer of molecules can change the threshold
voltage (V T ) of conventional silicon MOSFETs. Molecular grating thus provides
a new mechanism to control process variations [84] and leakage current [85],
which have been two of the most important challenges to modern CMOS-based
circuits. We also showed that the mobility of intrinsic silicon nanowires can be
dramatically changed through fluoride ion treatment. Molecule grafting can thus
considerably improve the performance of silicon nanowires-based computing.

11.5.1. Controlled Modulation of Conductance in Silicon Devices
While many alternatives, including molecules, have been proposed to implement
computing in the nanometer era, silicon remains the stalwart of the electronics
industry. Generally, the behavior of silicon is controlled by changing the
composition of the active region by impurity doping, while changing the surface
(interface) states is also possible [86-89]. As scaling to the sub-20 nm-size region is
pursued, routine impurity doping becomes problematic due to its resultant
uncertainty of distribution [90, 91]. Doping uncertainty is a major contributor
to process variations that have challenged deep-submicron silicon MOS circuit
design [92]. Provided that back-end processing of future devices could be held to
molecularly permissive temperatures (300-350 C) [93], it is attractive to seek

controllable modulation of device performance through surface modifications,
taking advantage of the dramatic increase in the surface-area-to-volume-ratios of
nanoscale features.
Several techniques have been used to covalently attach molecules directly onto
silicon surfaces [19, 20, 94, 95]. The Si-C bond formed using these methods is both
thermodynamically and kinetically stable due to its high bond strength (3.5 eV)
and low polarity [95, 96]. The majority of research in this area has focused on the
grafting methods or the influence on the surface (or interface) properties of bulk
semiconductors. So far little research has been conducted showing controlled
modulation of semiconductor devices by grafting molecular layers onto oxide-free
active device areas, and particularly via silicon—sp 2 -hybridized-carbon bonds.

1