1. History of single ion implantation(ref.1)
Ion microprobe system(IMP) is a modified version of Rutherford backscattering system(RBS) with a function of both depth and areal resolution and useful for three dimensional element analysis. Increasing ion current for better statistics, however, had a disadvantage of serious damage in a target, so we decided to extremely reduce the ion current and finally to single out each ion from the ion beam. The single ion microprobe(SIMP) developed in this way was used for the first time to study site dependence of radiation hardness in semiconductor devices such as MOSFETs, pn junctions and Schottky diodes.
In the middle 1990s, size reduction of semiconductor devices went beyond the reach of the SIMP because of the insufficient areal resolution. In the course of our study to seek for much better areal resolution of the single ion technology, we found that, in the era of nm size devices, the distance between dopant atoms became comparable to the device size, and that the fluctuation in device functions due to fluctuation in dopant number could not be negligible. Finally in 1994 we developed a single ion implantation (SII) as a tool to suppress device function fluctuation induced by fluctuation in discrete dopant number. The first SI implanter was born in 1996 by adding some new features to an existing focused ion beam (FIB) system. The new features were extraction of single ions, a singularity of extracted ions, detection of single ion incidence and high aiming precision. By improving these technological factors step by step, we have finally realized the controllability of number and position of dopant ions.
2. Novel applications of SII
A Single atom devices(ref.2, ref.3, ref4)
We fabricated silicon transistors containing two, four and six arsenic ions implanted in one dimensional array along the channel using SII method. The quantum transport was measured through the D0 and D- states of the arsenic ions at low temperature in the subthreshold region. Two different transport mechanisms contributed to the deterministically doped device: the Coulomb blockade and the Hubbard band formation. In case of the two arsenic donor sample, the Coulomb blockade was dominant, and each current peak was isolated. In case of the six arsenic donor sample, the Hubbard band formed, and the current peaks overlapped. These results indicate that our deterministic single-ion doping method is more effective and reliable for single-atom device development and pave the way towards single atom electronics for extended CMOS applications.
B SII to living cells(ref.5)
We performed gold atom doping into live cells by using the FIB implantation method. We evaluated the viability of the gold-implanted cells by measuring the concentration of adenosine triphosphate (ATP), which is an intracellular energy source produced in the mitochondrial membrane. The viability of the implanted cells was found to be 20% higher than that of the unimplanted control cells. The implanted atoms might promote the energy generating processes within the mitochondrion. These results suggest that the viability of live cells can be modulated by accurately controlling the dopant atom numbers. Our ion implantation technique may be considered as a more accurate tool to quantitatively elucidate the dose-dependent effects of dopants than the conventional methods.
References
1. I. Ohdomari, J. Phys. D. 41, 043001 (2008).
2. E. Prati, M. Hori and T. Shinada et al., Nature Nano. 7, 443-447 (2012).
3. T. Shinada et al., IEDM Tech. Dig., 697-700 (2011).
4. M. Hori et al., APEX 4, 046501 (2011).
5. T. Shianda et al., Biotechnology and Bioengineering 108, 222-225 (2011).
