Solid-State Nanopores: Methods of Fabrication and Integration, and Feasibility Issues in DNA Sequencing - Nanopores: Sensing and Fundamental Biological Interactions - page 183

It is quite possible that the e-beam removal (or “sputtering”) of the materials is only
to thin down the membrane locally to the point where the probability of spontaneously
nucleating a pore become unity. The fact that the diameter of the TEM “drilled” is
critically sensitive to the electron beam shape (Venkat Balagurusamy and Paul
Weinger, private communications, 2010) lends support to this view.

8.2.3 Feedback Chemical Etching

Here a low-cost method in fabricating nanopores in silicon chips is discussed. The
basic concept behind this method (Park et al. [ 9 ]) is similar to that in the work of
Apel et al . [ 13 ], but the method of Park et al. [ 9 ] is more precise and is unique to
silicon processing.
The starting material is a silicon wafer. The first step is to use standard

silicon processes to fabricate an inverted pyramid with a sharp tip on one side of the
silicon wafer, using the well-known techniques of anisotropic etching of Si in
alkaline solutions such as KOH, or (CH 3 ) 4 NOH (tetramethylammonium hydroxide
or TMAH). After the first step, the silicon chip is mounted as a partition in an
electrochemical setup separating two chambers filled with KCl and KOH solutions,
with KCl on the front side with the pyramid, and KOH on the back, as shown in
Fig. 8.5 (left, inset). While KOH is etching away silicon from the back side, the
electrical current across the silicon chip is monitored using two Pt electrodes, one in
KCl and the other in KOH. Under proper bias voltages, it is possible to detect the

< 100 >

Fig. 8.5 Current vs. time during the feedback etching process [ 9 ]. Used with permission. Copyright
Small 2007