Katsuhiko Nishiguchi Nanodevice's group
Advanced Applied Physical Science Laboratory
NTT Basic Research Laboratories


 WhySingle electrons?
   All the world, we can access a huge amount of data and take advantage of them using various kinds of information technology (IT) supported by a lot of electronics products, e.g., personal computers, cellular phones, data servers, and so on. Since an IT society provides us with more useful, comfortable and friendly life, performance of electronics products has been continuously improved and numbers of both electronics products and their users have been increasing. On the other hand, such a trend gives rise to the serious issue: power consumption to use IT has been also increasing drastically. The goal of our research is reduction in power consumption of electronics products. For example, in a transistor, which is one of the well-known and widely-used devices in electronics products, one information data is represented by a lot of electrons, which are source of electric power. What a waste of electrons or power. Our approach to reduce power consumption is the ultimate reduction of electros representing one information data: SINGLE ELECTRON. 

How to useSingle electrons
   Single electrons can be treated by various kinds of nanoscale devices called "single-electron devices (SEDs)". Before the introduction of such devices, let me explain a metal-oxide-semiconductor field-effect transistor (MOSFET) which is one of the most used devices and integrated into an electrical circuit, e.g., CPUs and micro- processors. The MOSFET has two conductive electrodes, i.e., source and drain, semi- conductor area (channel) between the electrodes, and gate. There is an insulator layer between the channel and gate, which means a capacitor. Since the semiconductor channel is insulative for electrons, there is no current between the source and drain, which corresponds to an “OFF” state of the MOSFET. When positive voltage is applied to the gate, negative charges, i.e., electrons, are induced from the source and the channel becomes conductive. As a result, electrons can travel from the source and drain, that’s, current flows through the channel, which corresponds to an “ON” state of the MOSFET. As easily expected from the case of capacitor, at the ON state, a lot of electrons are induced randomly and widely in the channel, which prevent us from dealing with single electrons.

 Such difficulty of control of a single electron can be available in the SEDs. One of the SEDs is a single-electron box (SEB), which is the most basic one. The SEB has a quite small conductive island, source, and gate. What’s important here is that the island is extremely small so that charging energy caused by one electron injected into the island is larger than thermal energy. Another important thing is that the island is located close to the source so that electrons can travel between the island and source owing to tunnel event. These features prevent electrons from entering the island by thermal energy. This phenomenon is called “Coulomb blockade” (Proc. IEEE, vol. 87, p. 606-632, 1999). When positive voltage corresponding to the charging energy is applied to the gate, just one electron can enter the island from the source. The number of electrons in the island as a function of the gate voltage shows a staircase pattern. The size of the island in which Coulomb blockade is active at room temperature is <10 nm.The SEB can be used for a memory device in which one stored bit is represented by one electron (e.g., Appl. Phys. Lett.v. 68, p. 1377,1996).

  Another well-known SED is a single-electron transistor (SET), which has a source, drain, gate, and island. Although there are gaps between the drain/source and island, electrons can travel between the source and drain due to tunneling event. In this case, Coulomb blockade causes interesting electron tunneling characteristics: when voltage applied to the gate is changed, current flowing through the SET oscillates uniformly against voltage, which is called Coulomb blockade oscillation. When voltage is small, no electron go through the island, that’s, no current flow. When voltage becomes a value corresponding to the charging energy, single electrons can go through the island, that’s, current flows. When voltage becomes larger, just one electron stay at the island and other electrons cannot go through the island due to Coulomb blockade. These phenomena cause oscillating current characteristics, which gives new functionality to SET-based circuit (e.g., Tech. Dig. -Int. Electron Devices Meet. (2001) p. 147). The difficulty of realization of SET circuits is to form an extremely small island (<10 nm) with high quality and controllability. Our group was succeeded in a development of a unique fabrication method originating from that of silicon MOSFETs (Tech. Dig. -Int. Electron Devices Meet. (1994) p. 938) and in various kinds of circuit demonstrations.

 In the SETs, single electrons go through the island one after another with random interval. In other wards, it is impossible to transfer one electron with intended timing. One-by-one electron transfer with precise timing can be achieved by using single-electron turnstile (Phys. Rev. Lett. V. 64, p. 2691 1990) and pump (Physica B v. 169, p. 573, 1991). Especially, the turnstile composed of an electrically formed island instead of a physically formed one makes it easy to realize the one-by-one single-electron transfer with precise timing (Phys. Rev. Lett. v. 67, p. 1626, 1991). This can be achieved by connecting two MOSFETs in series (Appl. Phys. Lett. v. 84, p. 323, 2004). When a MOSFET connected to a source is ON, electrons can flow into the MOSFET. Then, when two MOSFETs become OFF, one electron can stay at the center between two MOSFETs due to Coulomb blockade. Next, when the other MOSFET becomes ON, the electron go into the drain. By repeating such alternative switching of two MOSFET, single electrons can be transferred with precise timing, which allows a data information circuit using one electron as one bit of information (e.g., Appl. Phys. Lett. v. 88, p. 183101, 2006).