Page 163 - Principles and Applications of NanoMEMS Physics
P. 163

4. NANOMEMS APPLICATIONS: CIRCUITS AND SYSTEMS                151


             4.2.2 NanoMEMS SoC Building Blocks

             4.2.2.1  Interfaces

                The idea behind NanoMEMS is that of creating a system that, in order to
             accomplish a given function, avails itself of devices and techniques spanning
             the range from the micro- down to the nano-scale and beyond. In the most
             general case, the input signal to a NanoMEMS SoC will be analog, i.e., will
             exhibit continuous  amplitude and will exist at all  times,  see  Figure  3-1.
             Processing this signal, therefore, will entail deciding whether it is feasible to
             act on it as received/detected, or to transform it to a more convenient state.
             The nature of the interface sensor, in particular, its sensitivity, bandwidth,
             and dynamic range, will come into play here and will dictate the need for
             transduction, amplification, digitization, filtering, etc., thus determining the
             rest of the architecture. In  this context, the doubly-anchored  Si  beam  has
             been  considered  as  a potential mechanical sensing  element in future
             NanoMEMS SoCs, and impressive estimates for its intrinsic force sensitivity
             (S F), dynamic range (DR), mass sensitivity (M), and  bandwidth (BW) have
             been obtained by Roukes [174]. For instance, a beam of length, width, and
             thickness 0.1 x 0.01 x 0.01 microns and active mass 10ag  would  exhibit
             S  F  / 1  2 (ω 0 ) =  3× 10 − 17  N /  Hz ,  DR =  35 dB ,  M =  7 . 1 × 10 − 21 g , and
             BW =    7 . 7  GHz , assuming a temperature  of 300K  and a Q  of  10,000.

             Unfortunately, it is unclear whether the full extent of these parameters will
             be accessible due to various practical difficulties such as mass variation due
             to unpredictable adsorbates,  and the  impossibility of  realizing  a  noiseless
             read-out.  This latter  theme is also  common to electrostatic- and  optically-
             based sensing  interfaces  as well.  In the former case, which  according to
                                                              − 18
             Roukes [174] may attain a minimum capacitance of  10  F , the parasitic
             capacitance would preclude resolving it. In the latter case, the fact that the
             spot size  of  the light delivered by the  optical  fiber  used in  AFM
             displacement-sensing is much greater than nanoscale dimensions, precludes
             its resolution and, hence, proper detection.
                In systems with an electronic input signal sensing scheme, however, the
             sensor may take the form of a quantum superlattice-based analog-to-digital
             converter, Fig.  4-2  [175]. Here,  the pulsating nature of  the superlattice’s
             current-voltage  characteristic directly samples/quantizes the voltage axis.
             The resulting current is used to generate pulses that drive a counter whose
             output is a digital representation of the input voltage. For highest resolution,
             the superlattice may be realized with molecular devices.
   158   159   160   161   162   163   164   165   166   167   168