JPH0823090A - Electron wave directional coupling device - Google Patents

Abstract

PURPOSE:To attain a high speed operation by reducing the coupling length of an electron wave directional coupling device. CONSTITUTION:The title device has a pair of quantum thin wire channels 2A and 2B contacted through a quantum well point contact 1 having a size greater than the de Broglie's wavelength of electron and lower the mean free path of electron and a gate electrode 3 for modulating the Fermi energy of electron in the well point contact.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electron wave directional coupling device, and more particularly to a device structure capable of reducing the size of the device.

[0002]

2. Description of the Related Art FIG. 3 is a structural diagram of an electron wave directional coupling device according to the prior art. Such an electron wave directional coupling element is disclosed in, for example, JA Del Alam.
o) et al., Appl. Phys. Lett. 56, 78.
Page, 1990 and Applied Physics Vol. 62, No. 127-13
3 (1993). In the figure, 1
0 is a semi-insulating (SI) semiconductor substrate, 31 is a tunnel barrier region, and 32A and 32B are quantum wire channels. The quantum wire channels 32A and 32B are in contact with each other via a tunnel barrier region 31, and a gate electrode 33 is provided in contact with the tunnel barrier region. Further, a source electrode 34A and a drain electrode 35A are formed at both ends of the quantum wire channel 32A, and a source electrode 34B and a drain electrode 35B are similarly formed at both ends of the quantum wire channel 32B. In such an electron-wave directional coupling element, the potential barrier height Vb is modulated by the gate potential to change the coupling state of the electron wave function in the quantum wires 32A and 32B, and the electron transition probability between the quantum wire channels. Modulate (T _AB ).

[0003]

Here, in order to improve the switching efficiency and switch the electron transition probability T _AB from the quantum wires 32A to 32B from 0 to 1, the contact areas of the quantum wire channels 32A and 32B are changed. The length Lt needs to be set according to the transfer length defined by the following equation.

Lt = π / 2κ (Vb) (1) In the equation, π is a circular constant, and κ is a quantum wire channel 32A.
And the coupling constant of the electron wave function at 32B. Lt is a function of the potential barrier height Vb,
According to del Alamo et al., Lt = 1 μm when Vb = 5 meV and Lt = 5 μm when Vb = 15 meV.
It is estimated that Here, the thickness of the tunnel barrier 31 is 30 nm. Here, in order to obtain the interference effect of the electron wave, the transfer length Lt must be shorter than the phase interference length Lφ of the electron wave, but the condition of Lφ = 1 to 5 μm is that the quantum wire has a highly pure selective doping structure. Formed of high-purity GaAs having an absolute temperature T = 10K
It can be realized by cooling to a certain degree. However, in such an electron wave directional coupling element,
Since it is difficult to reduce the coupling length Lt to several μm, it is difficult to reduce the parasitic capacitance and the like, and it is also difficult to reduce the intrinsic delay time (electron transit time), making it difficult to operate at high speed. Was.

[0005]

According to the present invention, there are provided first and second quantum wire channels having ohmic electrodes at both ends, the first and second quantum wire channels being mutually electron An electron-wave directional coupling device is obtained in which a quantum well point contact having a size not less than the de Broglie wavelength and not more than the mean free path is contacted, and a gate electrode is provided on the quantum well point contact. .

[0006]

1 is a structural view of an embodiment of an electron wave directional coupling element according to the present invention. FIG. 1 (a) is a bird's eye view and FIG.
(B) shows a plan view seen from a direction perpendicular to the substrate. In the figure, 1 is a quantum well point contact, 2A, 2
B is a quantum wire channel, 3 is a gate electrode, 4A and 4B are source electrodes, 5A and 5B are drain electrodes. Also,
The interfaces of the quantum point contact 1 that come into contact with the quantum wires 2A and 2B are referred to as end faces 6A and 6B, respectively. The feature of this embodiment is that the tunnel barrier region having a length of about the transfer length Lt in the prior art is replaced with a quantum well point contact having a size of about the mean free path of electrons λe or less.

[0007] Such an element is manufactured as follows. Non-doped GaAs buffer layer: 1 μm, non-doped Al _0.2 Ga _0.8 As spacer layer: 40 nm, n-type Al _0.2 Ga _0.8 on a semi-insulating (SI) GaAs substrate 10 by, for example, molecular beam epitaxy (MBE). A selective doping structure consisting of an As layer (impurity concentration 3 × 10 ¹⁸ / cm ³ )... 20 nm and an n-type GaAs cap layer (impurity concentration 3 × 10 ¹⁸ / cm ³ ) 3 nm is grown. Here, the GaAs buffer layer at the interface with the spacer layer is a channel layer.

Next, an X-type quantum wire channel region is formed by partially etching up to the buffer layer having the selective doping structure. Here, the width of the quantum wire channel is about 50 nm. The size of the intersection of the quantum wire channels is Lcl = Lct, where Lcl is the length in the source / drain direction and Lct is the distance between the quantum wires 2A and 2B.
= About 50 nm, and this intersection region forms the quantum well point contact 1. After vapor deposition of the source electrodes 4A and 4B and the drain electrodes 5A and 5B, ohmic contact is made by a normal alloying process, and a resist pattern formed by, for example, an electron beam (EB) exposure method is formed at the intersection of the quantum wire channels. The Schottky electrode 3 is formed by depositing a gate metal as a mask. In this way, the electron wave directional coupling element as shown in FIG. 1 is manufactured.

In this embodiment, the incident angles of the quantum wires 2A and 2B with respect to the quantum well point contact 1 are given by θ1
= Θ2 = 45 °, but generally θ1 ≠ θ2 or an angle other than 45 °. Alternatively, a non-doped InGaAs strained layer may be formed between the GaAs buffer layer and the AlGaAs spacer layer as a channel layer, and this layer may be used as a channel layer.

FIG. 2 shows a traveling path of electrons in the vicinity of the quantum well point contact 1 of this embodiment. In this embodiment, the quantum wire 2A injects electrons into the quantum well point contact 1 at the incident angle θ ₁ at the end face 6A, and
Electrons emitted from the end face 6A are collected. Similarly, the quantum wire 2B injects electrons into the quantum well point contact 1 at the incident angle θ ₂ at the end face 6B, and collects electrons emitted from the end face 6B. Quantum point contact 1
The size of Lcl = Lct = 50nm and electron mean free path (λe = 100nm in the case of high-purity GaAs)
Since the following, the electrons travel in the point contact without scattering. In this case, quantum well point contact 1
When the transmission angle of the electrons inside is θ _r, and the Fermi energies of the electrons at the quantum wire 2A and the quantum well point contact 1 are E ₁ and E _r , respectively, the following relationship holds.

Sin θ ₁ / sin θ _r = (E _r / E ₁ ) ^1/2 (2) This equation is described by J. Spector et al. In Applied Physics Letters (Appl.
Phys. Lett. Vol. 56, p. 1290, 199
Reported in year 0. Therefore, the transmission angle θ _r can be modulated by changing the Fermi energy E _r of the quantum well point contact 1 by applying a gate voltage. When E _r = E ₁ , θ ₁ = θ _r and total transmission occurs (electron transition probability T _AB = 1). Also, θ _r
= A condition for a 90 ° When E _r = E _c (sinθ ₁
= (E _c / E ₁ ) ^1/2 ), and when E _r <E _c , total reflection occurs (T _AB = 0). Thus, the transition probability T _AB of electrons from the quantum wire 2A to the quantum wire 2B can be modulated in the range of 0 to 1 according to the gate voltage. Similarly, the electron transition probability T _BA from the quantum wire 2B to the quantum wire 2A can be modulated, and according to the present invention, the switching of the electron transition between the quantum wires can be performed very efficiently.

Here, in order to realize good switching, it is necessary for the electrons to travel in the quantum point contact 1 without scattering, and for that purpose, the size (Lcl, Lct) of the point contact is the mean free path of the electron. (In the case of high-purity GaAs or InGaAs, λe = 100 nm)
The following is acceptable. Further, (Lcl, Lct) is a limit length at which electrons penetrate a point contact due to a tunnel effect, that is, de Broglie of electrons.
It is necessary to be longer than the wavelength (approximately λ = 10 nm). Therefore, the point contact size (Lcl, Lc
t) needs to be not less than λ and not more than λe. Therefore,
The coupling length of the electron wave directional coupling device according to the present invention is 100 nm.
Since it can be reduced to about several tens of nanometers, the intrinsic delay time (electron transit time) can be reduced, and the parasitic capacitance and the like can be reduced, so that higher speed operation can be realized.

[0013]

As is apparent from the above detailed description,
According to the present invention, an electron directional coupling element can be realized by using a pair of quantum wires that are in contact with each other through a quantum point contact having a size of the electron mean free path as a channel, and the electron directional coupling can be realized. The size of the element can be reduced and the operation speed can be improved.

[Brief description of drawings]

FIG. 1 is an element structure diagram of an example of an electron wave directional coupling element according to the present invention.

FIG. 2 is a diagram showing an operating principle of an electron wave directional coupling element according to the present invention.

FIG. 3 is an element structure diagram of a conventional electron-wave directional coupling element.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Quantum well point contact 2A, 2B, 32A, 32B Quantum wire channel 3, 33 Gate electrode 4A, 4B, 5A, 5B, 34A, 34B, 35A, 3
5B Ohmic electrode 6A, 6B End face 10 Semiconductor substrate 31 Tunnel barrier

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location H01L 29/80

Claims (2)

[Claims] The present invention has first and second quantum wire channels having ohmic electrodes at both ends, wherein the first and second quantum wire channels are equal to or longer than the de Broglie wavelength of electrons and averaged by electrons. An electron wave directional coupling element, wherein said elements are in contact with each other via a quantum well point contact having a size equal to or smaller than a free path, and a gate electrode is provided on said quantum well point contact. 2. The electron according to claim 1, wherein the channel layer of the quantum wire channel is formed of GaAs or InGaAs, and the size of the quantum well point contact is 10 nm or more and 100 nm or less. Wave directional coupling element.