JPH0337736B2 - - Google Patents

Description

[Detailed Description of the Invention] Technical Field of the Invention The present invention relates to a bipolar type transistor (quantized base transistor: QBT) which has an operating principle completely different from conventional ones and can operate at ultra high speed.
This invention relates to a semiconductor device formed by cascade-connecting.

Prior Art and Problems Generally, the principle of operation of bipolar transistors is well known, and will now be briefly explained with reference to FIG.

Figure 1 shows the energy band of a conventional npn bipolar transistor when it is in thermal equilibrium.
This is a diagram.

In the figure, E is the emitter, B is the base, C is the collector, E _F is the Fermi level, _EV is the valence band, E _C is the conduction band, and n and p are the conductivity types, respectively.

The basic operating principle of such a bipolar transistor is as follows.

Now, when a forward voltage is applied between emitter E and base B, electrons injected from emitter E to base B travel through base B by diffusion etc.
The signal has been transmitted from emitter E to collector C by reaching .

Therefore, the operating speed of this bipolar transistor ultimately depends on the emitter E due to electron diffusion, etc.
It is regulated by the time required to travel from the source to the collector C, that is, the travel time due to diffusion and the like.

OBJECTS OF THE INVENTION The present invention provides a complementary semiconductor device that is capable of high-speed operation based on a completely new operating principle and utilizes a transistor (hereinafter referred to as QBT) having unique signal transmission characteristics.

Structure of the Invention In the semiconductor device of the present invention, a base of one conductivity type (for example, p-type base B) in which sub-bands (for example, sub-bands E ₁ , E ₂ . . . ) for minority carriers can be generated; , the one conductivity is transmitted through an emitter potential barrier (for example, an emitter potential barrier PB _E ) having a larger energy band gap than the base of the one conductivity type and having a thickness sufficient to cause a tunnel effect. an emitter of the opposite conductivity type (for example, an n-type emitter E) in contact with the base of the mold, and a collector potential having a larger energy band gap than the base of the one conductivity type and having a thickness sufficient to cause a tunnel effect. a collector of an opposite conductivity type (for example, an n-type collector _C ) that is in contact with the base of one conductivity type via a barrier (for example, a collector potential barrier PB C ), and a sub-base for the minority carrier generated on the base; - A transistor in which the minority carrier undergoes resonant tunneling transition from the emitter to the collector via the band, and a sub-band having a similar structure to the transistor and having a different energy level from the sub-band are generated. The structure is characterized by cascading transistors.

The operating principle of the QBT used in the semiconductor device of the present invention is completely different from that of conventional bipolar transistors, so it will be explained next.

FIG. 2 shows an energy band diagram when QBT is in thermal equilibrium, and the same parts as those explained with respect to FIG. 1 are designated with the same symbols.

In the figure, PB _E is the emitter potential.
barrier, PB _C is the collector potential barrier, E ₁ , E ₂ ...E _o ... is the sub-band, L _B is the width (thickness) of the base B, and Z is the direction from emitter E to collector C. It shows.

Now, the base B sandwiched between the emitter potential barrier PB _E and the collector potential barrier PB _C is substantially two-dimensional, and to the extent that the movement of electrons (carriers) in the Z direction is quantized. The thickness is set to be sufficiently thin, for example, about 50 [Å].

When base B is formed thin in this way, base B becomes like a potential well, in which electrons traveling from emitter E to collector C, that is, electrons traveling in the Z direction, A state is realized in which only energy levels can be assumed. That is, along with the quantization, the base B
sub-bands E ₁ , E ₂ . . . E _{o . .} . are formed.

Approximately, the energy level E _o of the sub-band is
That is, when the potential barrier is set to infinity, E _o = πζ ² n ² /2m ^* L _B ² ζ = h/2π h: Planck's constant n = energy quantum number m ^* = effective mass of electron mass) L _B = given by base width.

Now, in such QBT, Emitsuta E
When a forward voltage is applied between B and B, the energy band diagram in that case changes as shown in FIG.

In FIG. 3, the same parts as those explained in connection with FIG. 2 are indicated by the same symbols.

In the figure, V _EB is the emitter-base voltage,
V _CB is the collector-base voltage, DB is the DeBloglie wave, and the emitter potential barrier PB _E and the collector
The thickness of each potential barrier PB _C is selected to allow electron tunneling.

When voltage is applied as shown,
For example, when electrons with the same energy as energy level _E1 are injected from emitter E in the Z direction,
The electrons reach the collector C with a transmittance of 1, that is, complete transmission, as seen in the de Broglie wave DB.

The process in which these electrons reach collector C from emitter E does not rely on conventional travel, but rather a transition that relies on so-called resonant tunneling, in which they pass through two potential barriers due to the tunnel effect. It is extremely fast.

When operating this QBT as a transistor,
Depending on the emitter-base voltage V _EB, the energy
All you need to do is achieve band alignment.
In principle, the base current is not essential for operation.

Therefore, although the input voltage is zero in direct current terms, a displacement current flows to charge the capacitance between the emitter and base, so this becomes a constant current through the circuit, and an ohmic current flows through the emitter and base. AC power is generated due to the resulting Joule loss.

In general, the output power is the collector current and the collector current.
Since this is the product of the base-to-base voltage V _CB , this QBT has the function of an amplifier.

In addition, if the base-emitter voltage V _EB is increased so that the conduction band E _C of the emitter is aligned between the energy levels E ₁ and E ₂ , the electron transmittance becomes zero. That is, it becomes a state of complete reflection, and the collector current also becomes zero.

Furthermore, when the base-emitter voltage V _EB is increased and, for example, the energy level E ₂ and the conduction band E _C are aligned, complete transmission becomes possible again and a collector current is generated.

In this way, in addition to its rare high speed, QBT has strong nonlinearity and a variety of functions not found in conventional transistors.

FIG. 4 is a diagram showing the transfer characteristics in QBT necessary to further clarify the above point.

In the figure, the vertical axis shows the collector current I _C and the horizontal axis shows the emitter-base voltage V _EB .
Here, we consider only the base sub-band as a sub-band, that is, only the case where n=1 in the energy order E _o in the sub-band, and also,
The base material is GaAs (m ^* = 0.068m ₀ m ₀ :
(the mass of an electron in a vacuum).

A in the figure shows a QBT with a base width L _B of 60 [Å] and an energy level E ₁ of 44 [meV], and B shows a QBT with a base width L _B of 40 [Å] and an energy level E1 of 44 [meV]. Each characteristic line of QBT where E ₁ is 99 [meV], V ₁ and V ₂ are the conduction band E _C and the base width L _B
The emitter-base voltage V EB for alignment to the energy level E ₁ in the QBT where the base width L B is = 60 [Å] and the energy level E ₁ in the QBT where the base width L _B is = 40 [Å _]. are shown respectively.

As can be seen from the figure, in QBT, a pulsed collector current I _C is generated only when the emitter-base voltage V _EB is V ₁ and V ₂ , and this collector current I _C is the emitter-base voltage V EB. It shows extremely nonlinear dependence on _EB .

The present invention has this characteristic transfer characteristic.
By combining QBTs, an excellent complementary semiconductor device is constructed.

Incidentally, in a conventional bipolar transistor, the collector current I _C only increases monotonically with respect to the emitter-base voltage V _EB .

Embodiment of the Invention FIG. 5 shows a cutaway side view of essential parts of a QBT used in an embodiment of the invention.

In the figure, 1 is a semi-insulating GaAs substrate, 2 is a collector, 3 is a collector potential barrier,
4 is a base, 5 is an emitter/potential barrier, 6 is an emitter, 7 is a gold/germanium/gold (Au/Ge/Au) ohmic collector electrode, 8
9 indicates the base electrode of the gold/zinc/gold (Au/Zn/Au) ohmic, and 9 indicates the emitter electrode of the Au/Ge/Au ohmic, respectively.

An example of the material composition etc. in this QBT is as follows.

Emitter semiconductor: n-type GaAs Dopant concentration: 4×10 ¹⁸ [cm ^-3 ] Dopant: Si Thickness: 1 [μm] Emitter potential barrier semiconductor: p-type Al _0.3 Ga _0.7 As Dopant concentration: 2×10 ¹⁹ [cm ^-3 ] Dopant: Be Thickness: 20 [Å] Base semiconductor: p-type GaAs Dopant concentration: 2×10 ¹⁹ [cm ^-3 ] Dopant: Be Thickness: 50 [Å] Collector potential barrier semiconductor: p-type Al _0.3 Ga _0.7 As Dopant concentration: 2×10 ¹⁹ [cm ^-3 ] Dopant: Be Thickness: 20 [Å] Collector semiconductor: n-type GaAs Dopant concentration: 4×10 ¹⁷ [cm ^-3 ] Dopant: Si Thickness: 1 [μm] ] Note that the emitter potential barrier and collector potential barrier are p-type, but this is a consideration to prevent the thin base from being filled with a depletion layer extending from the pn junction. For example, increase the dopant concentration of the emitter by 2×
10 ¹⁷ [cm ^-3 ], and if the collector dopant concentration is set to about 2×10 ¹⁵ [cm ^-3 ], the depletion layer will extend to the emitter side or collector side, so even if an i-layer is used, good.

An outline of the process applied when manufacturing this QBT will be explained.

(a) First, by applying, for example, the molecular beam epitaxy (MBE) method, n is grown on a semi-insulating GaAs substrate 1.
A type GaAs collector 2, a p-type Al _0.3 Ga _0.7 As collector potential barrier 3, a p-type GaAs base 4, a p-type Al _0.3 Ga _0.7 As emitter potential barrier 5, and an n-type GaAs emitter 6 are grown in this order. The n-type GaAs emitter 6 is formed to be sufficiently thick, for example, approximately 1 [μm], so as not to cause penetration when the emitter electrode is later heat-treated for alloying.
If the GaAs emitter 6 is made to have a sufficiently high concentration, the heat treatment described above becomes unnecessary, and therefore it is possible to make it thin.

(b) Perform mesa etching for isolation between elements using, for example, a hydrofluoric acid (HF)-based etching solution. This mesa etching is a semi-insulating GaAs
Continue until substrate 1 is reached.

(c) Form a base pattern photoresist film (not shown), use the photoresist film as a mask, and use a hydrofluoric acid etching solution to perform mesa etching until it reaches the n-type GaAs collector 2. The collector potential barrier 3, the base 4, the emitter potential barrier 5, and the emitter 6 are selectively removed.

(d) Form a photoresist film (not shown) having openings where the emitter electrode and collector electrode are to be formed.

(e) By applying the vapor deposition method, Au/Ge/
An Au film is formed, and then, in order to pattern it by lift-off, the photoresist film is dissolved and removed, and then an alloying process is performed to form a collector electrode 7 and an emitter electrode 9. .

(f) forming a photo resistor film (not shown) having a pattern for forming the base electrode 8;
The emitter 6 is selectively removed by dry etching using a CCl ₂ F ₂ based etchant.

(g) By applying the vapor deposition method, Au・Zn/Au
In order to form a film and then pattern it by lift-off, the photoresist film formed in step (f) is dissolved and removed, and then an alloying process is performed to form the base electrode 8. Form and complete.

The QBT explained above is
It goes without saying that it operates in accordance with the operating principle explained in .

By the way, in the above-mentioned QBT, the essential items are: (1) Formation of a sub-band for the minority carrier in the base (2) Emitter potential barrier and collector potential thick enough to cause a tunnel effect - This is the formation of a barrier, so the material composition etc. may be any as long as the above (1) and (2) are satisfied, and are not limited to those exemplified above. Also, the base does not need to be p-type like the QBT mentioned above, but n
It may be a type. Of course, in that case, the conductivity type of the emitter and collector will be p-type.

FIGS. 6 and 7 are diagrams showing the electrical characteristics obtained by the QBT having the semiconductor heterojunction described above.

Figure 6 shows the base-grounded collector characteristics of QBT.

In the figure, the vertical axis represents the collector current I _C , the horizontal axis represents the collector-base voltage V _CB , and the parameter is the emitter current I _E. Incidentally, the temperature T at the time this data was obtained was 77 [K].

Figure 7 shows the transfer characteristics of QBT.

In the figure, the vertical axis shows the collector current I _C and the horizontal axis shows the emitter-base voltage V _EB .
The temperature T in this case was also 77 [K].

The dimensions of the QBT from which these data were obtained are as follows: Emitter electrode width: 1 [μm], emitter electrode length: 50 [μm], emitter-base electrode spacing:
It was 1 [μm].

As can be seen from Figure 6, if the emitter current I _E is 0, the collector current I _C is also almost 0, and it is understood that the emitter current I _E ≒ collector current I _C , and the base current is extremely small. Ru.

This means that the current amplification factor β is large.

Next, we will discuss various QBTs different from the QBTs exemplified above.
present.

A Case of other QBT with semiconductor heterojunction Part 1 Emitter: Ge Emitter potential barrier: GaAs Base: Ge Collector potential barrier: GaAs Collector: Ge Part 2 Emitter: Si _{1-X Ge} Barrier: Si Base: Si _{1-X Ge X} Collector Potential Barrier: Si _Collector : Si _{1 -X Ge}
Ga _1-y As Base: Al _z Ga _1-z As Collector Potential Barrier: Al _v
Ga _1-v As Collector: Al _u Ga _1-u As Part 4 Emitter: InSb Emitter potential barrier: CdTe Base: InSb Collector potential barrier: CdTe Collector: InSb Part 5 Emitter: InAs Emitter potential barrier: GaSb base: InAs Collector potential barrier: GaSb Collector: InAs In addition to the materials presented above, it is possible to appropriately select materials that satisfy the conditions that there is a difference in energy gap and that the lattice constants are similar. can.

QBTs can be constructed not only by junctions using semiconductors as described above, but also by semiconductors and insulators, or by pure metals as long as the emitter can supply a carrier. .

Next, let us show and explain various such QBTs.

B For semiconductor/insulator junction system QBT Emitter semiconductor: n-type Si Dopant concentration: 10 ¹⁹ [cm ^-3 ] Dopant: As Thickness: 1 [μm] Emitter potential barrier material: SiO ₂ Thickness: 20 [Å] ] Base semiconductor: p-type Si Dopant concentration: 4×10 ¹⁹ [cm ^-3 ] Dopant: B Thickness: 50 [Å] Collector potential barrier material: SiO ₂ Thickness: 20 [Å] Collector semiconductor: n-type Si dopant Concentration: 5×10 ¹⁸ [cm ^-3 ] Dopant: As Thickness: 1 [μm] To manufacture QBT of this semiconductor/insulator system, (a) By applying the MBE method, n Form a type Si collector.

(b) A collector potential barrier made of SiO ₂ is formed by transferring the Si substrate to a plasma oxidation chamber and performing plasma oxidation without taking it out into the air.

In this case, the pressure inside the oxidation chamber is 10 ^-3
[Torr], the energy may be set to 100 [W].

During the formation of the collector potential barrier made of SiO ₂ by oxidation, for example, by applying the ellipsometry method used to measure the thickness of the oxide film when manufacturing Josephson devices, it is possible to The thickness of the collector potential barrier is measured (in-situ), and oxidation is performed to the set value.

(c) By applying the MBE method again, p-type
Forms a Si base.

In this case, Si becomes polycrystalline or amorphous, but it is easy to convert it into a single crystal using techniques such as electron beam annealing and laser annealing.

(d) After this, the emitter potential barrier and emitter are formed through the same process as above.

(e) To pattern each of the semiconductors or insulators, and to form and complete the required electrodes, a normal bipolar transistor manufacturing process may be applied.

C For QBT with metal emitter Emitter material: Al Thickness: 1 [μm] Emitter potential barrier material: SiO ₂ thickness: 20 [Å] Base semiconductor: p-type Si Dopant concentration: 4×10 ¹⁹ [cm ^-3 ] Dopant: B Thickness: 50 [Å] Collector potential barrier material: SiO ₂ Thickness: 20 [Å] Collector semiconductor: n-type Si Dopant concentration: 5×10 ¹⁸ [cm ^-3 ] Dopant: As Thickness: 1 [ [mu]m] FIG. 8 shows an energy band diagram of a QBT with metal emitters, in which the same parts as those described with respect to FIGS. 1 to 3 are designated with the same symbols.

Note that for the sake of simplicity, the bending of the band is omitted from the illustration.

In the figure, emitter potential barrier
PB _E and collector potential barrier PB _C are
Needless to say, it is composed of SiO ₂ .

This QBT structure has the advantage that the series resistance of the emitter can be reduced and it is easy to form fine emitters.

In addition, the manufacturing process of this QBT is
It is almost the same as when manufacturing "QBT" which is a semiconductor/insulator system junction.

Now, in the present invention, QBT as explained above is used.
A semiconductor device, that is, an inverter is constructed by cascade-connecting the QBTs, which have different sub-band energy levels.

FIG. 9 is a circuit diagram showing one embodiment of the present invention.

In the figure, 11 is a QBT with characteristic A seen in Figure 4, 12 is a QBT with characteristic B seen in Figure 4, R is a resistor, V _io is an input signal, V _put is an output signal, V _cc indicates the positive power supply level.

The operation of the embodiment shown in FIG. 9 will be explained as follows.

Consider V ₁ (low level) and V ₂ (high level) as input signals V _io .

First, if the input signal V _io = V ₁ , QBT
Only 11 is turned on, and the output signal V _put = V _cc −I _C・
R>0 (high level) (I _C :collector current).

Then, if the input signal V _io = V ₂ , then QBT
Since QBT12 is turned on and QBT11 is turned off, the output signal V _put is at ground level (low level).
become.

As can be seen from the above description, when a signal is input to the semiconductor device, a signal obtained by inverting the input signal is output, so the semiconductor device functions as an inverter.

In order to enable the circuit configuration shown in Figure 9, the energy level E ₁ of the sub-band in QBT12 must be one level higher than the energy level E ₁ of the base sub-band in QBT11. It is sufficient that the energy level of the sub-band is smaller than E ₂ .

FIG. 10 is a diagram showing the relationship between energy and base width in QBTs 11 and 12.

In the figure, the energy E _Y is plotted on the vertical axis, and the base width L _B is plotted on the horizontal axis, with n=1 and n=2
The characteristic lines marked with are the energy levels.
Corresponding to n=1 and n=2 in E _o , and
L _B1 is the base width corresponding to the energy level E ₁ of the fundamental sub-band of QBT11, and L _B2 is the base width of QBT11.
The base width corresponding to the energy level E ₂ of the second sub-band is shown. As you can see from the figure,
The base width L _B in QBT11 is L _B2 <L _B <L _B1
It is necessary to set it as follows. That is, for example,
If the base width L _B of QBT11 is taken as indicated by L _BX in the diagram, the energy level E ₁ in the corresponding base sub-band will be obviously lower than that in QBT12, and The energy level E ₁ in the fundamental sub-band is QBT1
It is lower than the sub-band energy level E ₂ in the exciting state of 1. If this is not done, there is a risk that QBT 11 and 12 will be turned on at the same time, and they will not function as a normal inverter.

This inverter does not consume direct current power, and can also be configured only with npn type.

The manufacturing method of this example is extremely simple, and the point is that QBTs with different base widths L _B are fabricated by the MBE method, etc.
It is best to grow it in batches.

Effects of the Invention The semiconductor device of the present invention includes a base of one conductivity type in which a sub-band for a minority carrier can be generated, and a base having a larger energy band gap than the base of one conductivity type, which can cause a tunnel effect. an emitter of the opposite conductivity type that is in contact with the base of the one conductivity type through an emitter potential barrier having a thickness of approximately a collector of the opposite conductivity type that is in contact with the base of the one conductivity type through a collector potential barrier having a thickness such that the minority
A transistor having a structure similar to the transistor and in which the minority carrier undergoes a resonant tunneling transition from the emitter to the collector via a sub-band for the carrier;
The configuration is characterized by cascade-connecting transistors that generate sub-bands having different energy levels from the band.

QBT used in the semiconductor device of the present invention
When the carrier is an electron, when the energy level of the electron injected from the emitter to the base is aligned with the energy level of the sub-band in the base, the electron reaches the collector with complete transmission. Furthermore, if the alignment described above cannot be achieved, the electrons will be completely reflected and will not reach the collector. When the alignment is achieved, the electrons move through the emitter potential barrier and the collector potential barrier due to the tunnel effect, so-called.
Since the transition is based on resonant tunneling, it is extremely fast, unlike the movement of electrons in conventional bipolar semiconductor devices. Furthermore, as is clear from the above, the collector current is turned on or off depending on whether the alignment is achieved or not. Since the alignment depends on the base-emitter voltage, the collector current will turn on or off depending on the voltage.
Furthermore, alignment can be achieved not at one point but at multiple points, so if base-emitter voltages with different values are applied in various pulses, the collector current will increase in pulses each time alignment is achieved. It has a unique flow characteristic.

The semiconductor device of the present invention has a configuration in which such QBTs are connected vertically, and no direct current power is consumed.
Since it has low power consumption like a metal oxide semiconductor (metal oxide semiconductor) and can be configured only with npn type, combined with the high speed of QBT, it can function as an extremely high speed inverter.

[Brief explanation of drawings]

Figure 1 is an energy band diagram of a conventional npn bipolar semiconductor device, Figure 2 is an energy band diagram when QBT is in thermal equilibrium, and Figure 3 is an energy band diagram when QBT is in operation. energy band diagram,
Fig. 4 is a diagram showing the transfer characteristics of QBT, Fig. 5 is a cutaway side view of the main part of QBT, Fig. 6 is a diagram showing the base-ground collector characteristics of QBT, Fig. 7 is a diagram showing QBT
Figure 8 is a diagram showing the transfer characteristics of QBT when the other embodiments are in thermal equilibrium.
A band diagram, FIG. 9 is a circuit diagram of an embodiment of the present invention, and FIG. 10 is a diagram showing the relationship between energy and base width. In the figure, 1 is a semi-insulating GaAs substrate, 2 is a collector, 3 is a collector potential barrier,
4 is a base, 5 is an emitter potential barrier, 6 is an emitter, 7 is a collector electrode, 8 is a base electrode, 9 is an emitter electrode, 11 and 12 are
QBT and R represent resistances, V _io represents an input signal, V _put represents an output signal, and V _cc represents a positive power supply level.

Claims (1)

[Scope of Claims] 1. A base of one conductivity type that can generate a sub-band for a minority carrier, and a thickness that has a larger energy band gap than the base of one conductivity type and that can cause a tunnel effect. an emitter of an opposite conductivity type that is in contact with the base of the one conductivity type through an emitter potential barrier having a a collector of an opposite conductivity type that is in contact with the base of one conductivity type through a collector potential barrier having a thickness, and a collector of the opposite conductivity type that is in contact with the base of the one conductivity type through a collector potential barrier having a thickness, and from the emitter to the collector through a sub-band for a minority carrier generated in the base. A transistor in which a minority carrier undergoes resonant tunneling transition, and a transistor in which a sub-band having a structure similar to that of the transistor and having a different energy level from the sub-band is generated are connected in cascade. semiconductor device.