Nanometer-scale Scanning Sensors Fabricated Using Stencil Lithography
In recent years, new forms of scanning probe microscopy have been developed which utilize small electrical devices that are scanned above samples to act as high-resolution sensors. Examples include gated scanning tunneling microscopy (g-STM), scanning single electron transistors, scanning thermal microscopy, and magnetic force microscopy (MFM). We are developing a new technique for fabricating high resolution scanning sensors at the 10-nm scale. Sensor devices are deposited directly onto conventional atomic-force microscope (AFM) tips by the evaporation of metal through holes in a stencil mask.
Alexandre Champagne, Ferdinand Kuemmeth, Aaron Couture, Dan Ralph
We are interested in fabricating high resolution scanning probe sensors on the end of AFM tips. Scanning probe techniques are limited in resolution by the size of the sensor and/or the distance of the sensor to the sample. For this reason, it is best to fabricate a sensor as small as possible directly on the apex of a tip, so that it can be scanned in close proximity to the sample. However, it is difficult to fabricate small sensors on the apex of AFM tips using traditional electron beam lithography. The problem is that high-resolution lithography requires depositing a uniform layer of a polymer electron beam resists, and this is hard to do on the surface of a tip because of the large curvature. Instead, we are developing a stencil-mask lithography technique which does not require application of resist to the tip surface .
A procedure previously used to make tunneling devices  has been adapted to make stencil masks. Freely suspended silicon nitride membranes are fabricated using photolithography and a wet KOH etch on silicon wafers coated with a thin silicon-nitride film. Nanometer sized features are written into a polymer resist layer on the silicon nitride membranes using e-beam lithography. Finally, holes are etched into the silicon nitride using reactive ion etching.
A commercial tapping mode AFM tip is positioned above the stencil mask as shown in Figure 1. The tip is used to image the surface of the silicon nitride and locate the nanometer-sized features by operating in traditional tapping mode fashion. The microscope, shown in Figure 2, was custom designed to operate in an electron-beam evaporator. The mask pattern is transferred onto the tip by evaporating metal from the source. Multiple stencils can be mounted at once on the microscope sample holder and used sequentially. The microscope is equipped with a translation system that allows positioning of the stencils under the AFM tip. Several evaporations steps can be performed on a tip in order to build complex device geometries.
Figure 1: Cartoon of pattern transfer setup
Figure 2: Picture of our homemade stencil-lithography AFM
We have been able to transfer patterns on AFM tips down to about 10nm in size and with a precise alignment. Dots and lines of Erbium about 20 nm wide were deposited (Figure 3 and 4). We have also transferred a two layer lithographic pattern consisting of 15nm Er lines intersecting with about 5nm wide Er lines. (Figure5) Those patterns were deposited on intentionally blunted AFM tip to facilitate the imaging of the deposited features. We have fabricated magnetic force microscopy (MFM) tips by depositing a single dot of Chromium-Cobalt at the apex of the tip. So far these tips have yielded a magnetic imaging resolution down to 50nm, but we expect to be able to improve this resolution by optimizing the fabrication process. We plan to build several other types of scanning sensors such as scanning single-electron electrometers  and gated scanning-tunneling microscope tips . These sensors will be used to image nm-scale properties of electronic devices and materials.
Figure 3: SEM image of an array of 5x5 Erbium dots (100nm spacing, 20nm size) at the end of an AFM tip.
Figure4: SEM image of an array of 5 Erbium lines (100nm spacing, 20nm wide) at the end of an AFM tip.
Figure 5: SEM image of a two layer lithographic pattern consisting of a 20nm wide Er line intersecting two 5nm Er lines.
- M. Deshmukh, D. C. Ralph, M. Thomas, and J. Silcox, Appl. Phys. Lett.,75, 1631 (1999).
- You et al., Science, 276, 579 (1997).
- L. Gurevich, L. Canali, and L. P. Kouwenhoven, Appl. Phys. Lett.76, 384 (2000).
- D. C. Ralph, C. T. Black, and M. Tinkham, Phys. Rev. Lett., 74, 3241 (1995).
Last updated: 2004