Skip to main content

Ion Beam Applications to High-density Memory Archives

Final Report Summary - IBAHMA (Ion Beam Applications to High-density Memory Archives)

The IBAHMA project was concerned with a new approach to providing ultra-stable (>50 years), ultra-high density (> 1Tbit/sq.in.) data storage for archival applications. We used ion-implantation to write nanoscale data into hydrogenated amorphous silicon carbide (a-SiC:H) films. The role of implantation conditions and post-implantation treatments on the achievable data density, readout contrast and data longevity was investigated, and optimised conditions determined. The likely limitations for practical application of this potentially important new approach to data storage have also been assessed.

The archival sector is becoming increasingly important, due in part to the introduction of new legal requirements governing the storage of governmental and commercial data, but also as a consequence of the ever-increasing amount of digital data generated by all aspects of our everyday life. Indeed, it has been estimated [1] that the total archival capacity required world-wide will exceed 60 ExaBytes by 2013, generating market values of in excess of $30 billion. For archival applications reliability, data integrity and media longevity feature much more prominently than in other storage sectors. The main recognised archival storage media currently in use are magnetic tape and optical disks, although it is estimated that around 50% of the archive data for commercial organisations is in fact held on magnetic hard disk drives. While optical disks for professional archiving are usually 'guaranteed' a lifetime of at least 50 years, magnetic tapes and disks have in general a much shorter predicted life-span. It is clear therefore that while archival storage is an exceedingly important application, conventional archival storage media are quite limited in terms of lifetime and in terms of storage density (bits per square inch). It is in this context that the IBAHMA project aim was to develop novel, ultra-stable and ultra-high density storage media for archival applications. Indeed, the approach of ion-implantation in SiC can potentially lead to extremely long (many hundreds of years) data lifetimes, which may find uses in the long-term preservation of important scientific and cultural assets.

To realise our goal of ‘permanent’ high-density storage we have used amorphous hydrogenated silicon carbide (a-SiC:H) as a storage medium, due to its promising material properties (e.g. high transparency in the visible region due to its wide optical band gap from 1.8 to 3.0 eV, as well as mechanical durability and chemical inertness [2]). The relative immunity of SiC to environmentally-induced degradation, in particular its high thermal stability (stable to temperatures in excess of 1500 ºC) make it attractive for data storage applications, and promising results have been achieved by using ion-implantation to write micro- and nanoscale optical marks in SiC films; see Fig S1.

Ion-beam implantation is a standard technique for controlling the bandgap, and hence the electrical and optical properties of semiconductors [3]. The development of computer-controlled focusedion-beam (FIB) systems have enabled the fabrication of sophisticated ion-implanted structures [4-6]. The focused-ion-beam diameter can be less than 10 nm, allowing the modification of the dielectric and optical properties of materials on this same nanoscale (10nm bit sizes is roughly equivalent to a data density of 10Tbit/sq.in.). Such nanoscale property changes in optical transmission and reflection offer ultra-high density optical data archives, the readout of which require a super-resolution (sub-diffraction) optical technique, e.g. scanning near-field optical microscopy (SNOM) [7] (see Fig.1 (a) and (b)). Of course imaging by SNOM, which is suitable for research purposes, is slow (and expensive), but for practical applications other super-resolution readout techniques, such as the use of solid-immersion lens (SIL) optics [8], the so-called super-RENS technique [9] or arrays of scanning near-field apertures might ultimately be used to achieve the necessary readout data rates (and required system cost).