Probe technologies

From the earliest days of scanning probe techniques, it has been recognized that scanning probes are not only useful for imaging surfaces at the atomic scale but also for modifying surfaces at a similar scale. This feature has triggered the consideration of scanning probes in the development of ultrahigh areal-density storage devices. Probe-based storage technologies may be regarded as natural candidates for extending the physical limits that are being approached by conventional magnetic and optical storage as well as semiconductor nonvolatile memories.

Scanning probe-based recording methods rely on physical bit sizes determined by the interaction of a sharp probe tip with a storage medium. There is a wide variety of methods for creating nanometer scale marks on surfaces. One recording method which has been successfully shown to achieve very high densities is thermomechanical recording, which uses sharp heated tips to create nm-scale indentations on thin polymer films [1]. Another promising concept is based on so-called phase-change materials, the local structure of which may be reversibly switched between a highly conductive and a highly resistive state by a rise in temperature; the local temperature change may be induced by either a hot tip or the injection of current through a conductive tip.

We are investigating the application of scanning probes in data storage, in particular archival storage. Key requirements for this application are write-once-read-many (WORM) media, long retention times, high storage density and high reliability / long lifetime. We are engaged in the development of recording/readback techniques for low medium and tip wear [3], high data rates and high density [2], in modeling of the recording and readback processes [8,10,11], and in the design of coding and signal processing techniques for maximum reliability. We collaborate closely with other IBM Research - Zurich groups that develop WORM media based on polymers, and high-performance cantilever/tips for high data rates.

In addition to archival storage applications, we are also investigating alternative data storage concepts based on local probe techniques. In particular, we have been involved in the first-ever demonstration of a complete storage system prototype with an array of scanning probes [5]; the system comprises

• an array of thermomechanical AFM cantilevers,
• a MEMS microscanner for positioning the polymer storage medium,
• a navigation system based on thermal position sensors and media-derived servo information [3,4,6],
• an integrated analog front-end, and
• a dataflow engine with a micro-controller, host interface and ECC.

As this prototype demonstrated, the inherent parallelism, the ultrahigh areal densities and the small form factor that probe storage techniques offer may open up new perspectives and opportunities in areas beyond those envisaged today.

Figure 1. Tip endurance: SEM image of tip after 140 m of travel; diameter of tip apex stays below 18 nm by using intermittent-contact mode sensing.

Figure 2. Modeling and coding/signal processing for probe storage: Combination of proper modeling of the probe storage channel and soft detection/decoding techniques improves reliability in the presence of heavy noise and jitter distortion.

Figure 3. Areal density: Demonstration of reliable data storage at 1.2 Tb/in² using thermomechanical probe recording; Symbol pitch is 13.3 nm, track pitch is 26.6 nm, and bit-error rate is 8×10^-5.

Figure 4. Areal density: Line scan at 1.2 Tb/in².