RU2306632C1 - Thyristor triode-thyroid - Google Patents

Abstract

FIELD: semiconductor electronics; semiconductor bipolar devices using thyristor-type structure.

SUBSTANCE: proposed semiconductor bipolar device in the form of p-n-p-n thyristor triode designed for rectifying, amplifying, switching, or generating electric signals has four layers of different polarity of conductivity, p anode, intermediate n and p layers, and n cathode. Intermediate n and p layers have common lead that functions as triode emitter. Anode is connected to function as triode base and anode p-n junction is connected during electron-hole plasma injection to intermediate n and p layers. Cathode is connected to function as triode collector and cathode p-n junction functions to deplete intermediate p layer.

EFFECT: enhanced operating current of device when running as amplifier.

1 cl, 10 dwg

Description

The present invention relates to semiconductor electronics, semiconductor bipolar devices for rectifying, amplifying, switching or generating electrical signals and having a structure like a thyristor.

Semiconductor amplifiers of electrical signals are widely used in integrated microsystems due to the possibility of combining them with other components of microsystems using microelectronics methods and the creation of microminiature devices for monitoring and control in automated complexes for various purposes.

Semiconductor amplifiers of electrical signals on diodes, bipolar and MOS transistors since the invention of the semiconductor triode transistor [1] and bipolar transistor [2] have been of particular interest due to the high speed up to 100 GHz [3].

Four-layer semiconductor structures of the pnpn type as transistor switches [4] are used in power electric circuits. Contact to the outer p-layer is called the anode. Contact to the outer n-layer is called the cathode. A device with two contacts to the anode and cathode is called a diode thyristor (as well as a dynistor or a Shockley diode). A device with a third control electrode connected to the inner region closest to the cathode or anode is called a triode thyristor or simply a thyristor (as well as a semiconductor controlled valve). A detailed description of the bipolar transistor and thyristor is given in [5, 6].

The voltage to the anode and cathode of the thyristor is connected so that the middle of the pn junctions is closed and no current flows through the thyristor. Two other pn junctions are included in the forward direction. When bias voltage is applied to the control electrode, charge carriers are injected from the pn junction of the cathode or anode. Injection of carriers creates a current through the middle pn junction and the thyristor opens. Thus, the thyristor operates in a key mode.

In the patent [7], a device is proposed in which a diode thyristor structure formed by depositing conductive layers of gallium arsenide on the initial substrate is used for switching microwave signals, and then the layers are etched around the device.

In the patent [8], it is proposed to form a triode thyristor structure using ion doping to create layers of different conductivity in the initial silicon carbide substrate.

The patent [9] proposes to form a triode thyristor structure for storage devices on a silicon substrate with oxide insulation SOI using ion doping to create layers of different conductivity and using self-alignment and self-formation methods.

The closest analogue adopted by us for the prototype is the patent [10]. The integrated circuit of a semiconductor amplifier for electrical signals includes a heterojunction thyristor. The thyristor on AlGaAs has conclusions from contacts to the anode, to the cathode, to the first and second injectors. Injectors contain the first and second quantum wells, the channels of which are located between the anode and cathode contacts.

The bias elements are direct current sources, which are connected through contacts to quantum wells having n- and p-types of doping and provide the flow of current in the thyristor. Resistance is established between the source of high potential and the contact to the anode and between the source of low potential and the contact to the cathode. These resistances limit the amount of current flowing between the anode and cathode and suppress the possibility of the switch switching to the open state, i.e. thyristor operates in closed mode. The bias elements provide a mode of operation of the heterojunction thyristor with a linear voltage amplification of the electrical signals supplied to the contacts of the first or second injector and output in an amplified form at the anode and / or cathode terminal. The value of the direct current of all contacts determines the voltage amplification by the thyristor at the outputs of the anode and cathode. The gain of the electrical signal is more than 200.

The structure of the device is formed by depositing up to 10 AlGaAs layers with different A1 contents and with different doping levels with impurities that create p- and n-type conductivity in a multilayer structure on a GaAs substrate. Etching through masks provides space for contacts to all layers. Contact points are additionally alloyed with ion doping through masks to obtain ohmic contacts.

The heterojunction thyristor electrodes are located on the surface in the following order: in the middle there is a contact to the p-anode, left and right contacts to the quantum well layer with n-type doping, then left and right contacts to the quantum well layer with p-type doping, further away from center contacts to the n-type conductivity substrate.

The device can sense, detect, amplify optical signals and belongs to the class of patents according to the international classification H01L 31/0328 - semiconductor devices that are sensitive to radiation.

The main disadvantage of this device is that the device operates in the closed state of the thyristor, i.e. at low currents.

The purpose of the invention is to increase the operating currents of the thyristor when operating in gain mode.

The essence of the invention is to create a thyristor-triode structure in which a thyristor of four layers of different conductivity types has an electrical connection of two middle layers, so that the device has three external electrodes and is a triode. To distinguish this structure of the thyristor triode, the name TIROD is introduced. Changes in the structure and operation mode of the thyristor with a shorted central junction, forward bias of the anodic pn junction and reverse bias of the cathodic pn junction provide current flow due to injection of an electron-hole plasma, i.e. thyristor operation in open mode.

Figure 1 shows a cross section of a semiconductor device of a thyristor triode-TIROD. In figure 2. The presentation in the form of diodes and injection connections between them and the symbol of TIROD are given. Figure 3 shows the static characteristics of the device. Figure 4 shows the frequency dependence of the current transfer coefficient. Figure 5 shows the concentration distribution of injected electrons and holes in the volume of the semiconductor. Figure 6 shows a section and distribution of impurities for TIROD formed by ion doping. Figure 7 shows a section of the structure of TIROD when performing a sequence of operations of ion doping. On Fig shows the topology of the planar structure of the device and the sequence of masks used. Figure 9 presents the sequence of masks used in the formation of a p-emitter by the method of self-formation. Figure 10 shows the frequency dependence of the transfer coefficient and conductivity for all conclusions TIROD formed using ion doping.

Figure 1 shows a cross section of a semiconductor device of a thyristor triode-TIROD, where the device consists of a layer of the first type of conductivity (1); the contact area of the first type of conductivity (2) to the layer of the first type of conductivity (1); areas of the second type of conductivity - thyristor anode (3); the contact of the TIROD base (4) to the thyristor anode - the region of the second conductivity type (3); the region of the second contact of the first type of conductivity (5) to the layer of the first type of conductivity (1); a layer of the second type of conductivity (6); the contact area of the second type of conductivity (7), to the layer of the second type of conductivity (6); areas of the first type of conductivity - thyristor cathode (8); the contact of the TIROD collector (9) to the thyristor cathode — the region of the first conductivity type (8); the second contact region of the second type of conductivity (10) to the layer of the second type of conductivity (6); the contact of the TIROD emitter (11) to the thyristor control electrodes - contacts to the regions (2), (5), (7), (10).

Figure 2 shows the representation of a thyristor with a general conclusion from the middle layers in the form of diodes and injection links between them and the TIROD symbol, which reflects the presence of two types of pnp and npn transistors in the thyristor structure, their injection coupling and the presence of a field effect in the operation of the device. The conclusions are indicated: to the collector contact (9) - K, to the base contact (4) - B, to the emitter contact (11) - E.

Figure 3 shows the dependence on the bias voltage U _BE between the base (4) and the emitter (11) of the currents passing through the collector contacts (9) - designation c5, base (4) - designation c2, emitter (11) - designation of current through contacts (2), (5) - с1 and through contacts (7), (10) - с4, with a collector voltage (9) of 1.5 V.

Figure 4 shows the frequency dependences of the current transfer coefficient h21 = abs (Y21 / Y11) in the TIROD connection circuit with input 1 to the base (4) and output 2 from the collector (11) for several bias voltages U _BE from 0.3 up to 1.2 V at a collector voltage of 1.5 V.

Figure 5 shows the concentration distribution of the injected electrons of Fig. 5a and the holes of Fig. 5b in the semiconductor volume of the TIROD structure.

The isolines indicate the concentrations of electrons or holes in the injected electron-hole plasma at a voltage of 1.2 V on the basis of (3), 1.5 V on the collector (8), potential on the emitter (2), (5), (7), (10) equal to zero.

Fig. 6 shows a cross-sectional view of a planar structure and impurity distribution for a TIROD, for example, realized using technology using ion doping of layers in a silicon substrate and with contact regions (7), (2), (4), (9) obtained by applying polysilicon. The same figure shows the distribution of the concentration of N impurities of boron and phosphorus in the cross section of the TIROD structure with ion implantation of layers and the concentration of electrons and holes in the regions of the pnpn structure forming the thyristor.

Figure 7 shows a section of the structure of TIROD when performing a sequence of operations of ion-doping (IL) of silicon wafers of p-type conductivity (pSi) coated with silicon oxide SiO ₂ 300 Å thick through photoresist masks (Ф / Р). For the formation of contacts (7), (2), (9) and the anode (4), polysilicon of different conductivity types pSi *, nSi * is used. Polysilicon films are applied after opening the windows in the oxide.

Figure 8 shows the topology of the planar structure of TIROD with ion implantation of n- and p-layers, n-cathode and the sequence of photo masks used to form all areas of the device, including polysilicon contacts to n- and p-layers, n-cathode and p- anode.

Figure 9 shows the sequence of photomasks used in the formation of a narrow p-emitter by the method of self-formation, when the polysilicon region of the p-anode overlaps an additional layer of silicon nitride by a small amount.

Figure 10 shows the frequency dependences of the current transfer coefficient h21 and the squares of the conductivities of the contact (7) to the p-layer - Y2 (4.4), of the contact (2) to the n-layer - Y2 (1,1), the contact (4) to the p-anode - Y2 (2,2), the contact (9) to the n-cathode - Y2 (5,5), between the contacts to the anode and the cathode - Y2 (5,2) TIRODE formed with using ion doping.

The structure of the TIROD shown in Fig. 1 in the form of a cross section differs primarily in that a short circuit is introduced between the two control contacts of the thyristor and there is one external output from the two control contacts, i.e. in this case, an external short circuit of the middle pn junction is performed, which is indicated in the drawing in the form of a bold line. The second important feature of the TIROD structure is the location of the anode, cathode, and contacts to the middle layers in such a way that cross-flow of the electron flux flowing to the cathode and the hole flux flowing from the anode takes place.

The four-layer pnpn structure appears to be composed of two pnp and npn transistors. The pnp transistor consists of a p-emitter (3), an n-layer of the base (1) with two heavily doped contact areas (2), (5), a lightly doped p-layer (6) of the collector with two heavily doped contacts (7), (10 ) The n-p-n transistor consists of the n-layer of the emitter layer (1) with the contacts (2), (5), the base p-layer (6) with the contacts (7), (10) and the collector n-region (8). The same zero potentials are applied to contacts (2), (5), (7), (10).

With this connection, the p-n-p transistor has the same potential at the collector and base, i.e. enabled in open diode mode. The emitter and the base are electrically connected in the n-p-n transistor, therefore, the transition between the emitter n-layer and the p-layer of the base is shorted and is not injective, i.e. the transistor is turned on in a closed diode mode. A positive potential is applied to the p-emitter (3) with respect to the zero potential at contacts (2), (5), so that its p-n junction is switched in the forward direction and injects holes. A positive potential is applied to the n-collector (8) with respect to the zero potential at contacts (7), (10), so that its p-n junction is switched in the opposite direction.

In this device, the flow of carrier streams of two signs has the following features. In the passive n-base of the p-n-p transistor between the contacts (3) and (2), (5), the fluxes of impurity conduction electrons and injected holes flow in counter flows. In the active n base, the flows of holes and electrons flow in one direction due to diffusion of the injected holes and compensation of their charge by electrons, as a result of which an electron-hole plasma is formed. In the region of the p-layer between the n-collector (8) and the p-emitter (3), the plasma flow is separated. Holes from the plasma go to contacts (7), (10). Electrons from the plasma fall from the p-layer of the base of the n-p-n transistor into the n-collector (8). Thus, despite the absence of an external bias voltage at the emitter junction of the n-p-n transistor, the injection of electrons into the p-base of the n-p-n transistor occurs using plasma.

In diode pin structures, the double injection of charge carriers from heavily doped n- and p-regions into the i-region with almost intrinsic conductivity leads to the formation of an electron-hole plasma in the presence of recombination of charge carriers and a pulling electric field in the region of plasma existence, which determines the drift mechanism current flow.

In a bipolar transistor with weak doping of the collector in the diode inclusion, in which the base and collector have the same potential, the charge carriers injected from the emitter pass into the base and into the collector and change the carrier concentration there, and according to the electroneutrality condition, the carrier concentration of both electrons and holes.

Usually in a bipolar transistor, it is believed that the concentration of impurities and, accordingly, the concentration of charge carriers in the base and collector is higher than the concentration of injected carriers. However, with a high level of injection and low doping of the base and collector regions, the concentration of injected charge carriers and charge carriers of a different sign compensating them can exceed the concentration of carriers determined by the impurity. An electron-hole plasma is formed in the base and collector of the p-n-p transistor, which determines the operation of the transistor.

The bias voltage at the collector-base pn junction creates an electric field that helps the carriers injected from the emitter through the pn junction and prevents the transfer of carriers of a different sign. With equal potentials on the base and collector, the electric field due to the voltage applied externally is absent. But in the space charge region of the pn junction, there is a built-in field due to the contact potential difference between p and n regions, which also separates the charge carriers, like a field with a reverse bias of the junction.

The plasma concentration can exceed the concentration of the space charge in the pn junction, and then the separation of carriers does not limit the penetration of plasma into the collector due to diffusion transfer. Bipolar injection of charge carriers of two signs occurs due to the formation of an electron-hole plasma in a lightly doped region of the collector of a p-n-p transistor.

Response threshold, i.e. The voltage or current of the emitter junction, at which the injected plasma concentration is sufficient to open the collector pn junction, is an important characteristic of a transistor with a closed collector-base junction when operating in the bipolar injection mode of an electron-hole plasma.

Figure 2 shows the representation of TIROD in the form of diodes and injection bonds between them due to the formation of an electron-hole plasma. 2 also shows the TIROD symbol, reflecting the presence of two types of transistors in the thyristor and the relationship between them.

Static characteristics of TIROD, i.e. the dependences of the currents of all electrodes on the bias voltage U _BE on the basis of (4) at U (9) = 1 V, are presented in figure 3. The TIROD structure has the following parameters. The thickness of the n-layer of 0.2 μm, the concentration of the impurity Nd = 10 ¹⁷ cm ^-3 . The p-layer thickness is 0.8 μm, and the impurity concentration is Na = 10 ¹⁶ cm ^-3 . The sizes of p + and n + of the contact areas are 3 μm, the gap between them is 2 μm, and the impurity concentration is N = 10 ²⁰ cm ^-3 . In practice, the collector currents Ic5 and the n-layer Id up to a bias voltage of 0.75 V are independent of U _BE , and the base currents Ic2 and the contacts to the p-layer Ic4 grow exponentially. The currents of the contacts Id and Ic3 add up to the current of the contact Ic5, and the currents of the contacts Ic4 and Ic6 add up to the current of the contact Ic2. The main current flows between contacts c1, c3 and c5. At low bias voltages, the ratio of the current of contact c5 to the current of contact c2 varies from 10 ⁸ to 10. If we determine the current gain by the ratio of the currents of contact c5 and contact c2, then this is a very high current transfer coefficient.

However, the flow of electronic currents in n-p-n and hole currents in p-n-p transistors is almost independent. The bias voltage at the junction between the n-layer with contacts (2), (5) and the p-region with contact (3) determines the level of injection from the p-emitter of the p-n-p transistor. The reason for the large current of the n-collector is the puncture mode of the base, which occurs at the selected doping level of the p-layer due to the expansion of the space charge region of the p-layer and n-collector transitions to the n-layer emitter of the n-p-n transistor. The electric field in the junction repels the holes and the hole flux is blocked by the space charge. Electrons are pulled into the space charge region and create an electrode current c5, which determines the contribution of the field effect to the operation of the TIROD.

At bias voltages greater than 1.2 V, the current transfer coefficient is less than unity, because contact current c2 exceeds contact current c5. With a large bias, the currents of contacts c1, c3 in total give the current of contact c2, therefore, the main current flows through the n-layer between contact c2 and contacts c1, c3 as a result of the double injection mode from p-emitter c4 and contacts c1, c3 to n- layer.

At a high doping level of the p-layer, for example, 2 × 10 ¹⁶ cm ^{-3, the} space charge pn of the junction of the p-layer, the n-collector does not reach the n-layer. The p-base protocol does not occur. The injected holes penetrate the p-layer and create a large contact current c4.

At average bias voltages between 0.75 and 1.2 V, the electron-hole plasma penetrates the p-layer and a relationship appears between currents through contacts c2 and c5.

The small-signal analysis of the circuit shown in Fig. 4 shows the presence of amplification of the variable signal H21 = h52 = abs (Y52 / Y22). In this circuit, input 1 is the base electrode (4), output 2 is the collector electrode (9). The cutoff frequency depends on the bias voltage and reaches 0.8 GHz at 0.8 V. Therefore, the possibility of high-frequency application of TIROD should be considered.

Figure 5 shows the concentration distribution of injected electrons and holes in the volume of the semiconductor structure TIROD with the parameters indicated below.

The n-layer (1) has a thickness of 0.2 μm and a concentration of phosphorus impurities equal to N = 2 · 10 ¹⁸ cm ^-3 .

The p-layer (6) has a thickness of 0.5 μm and a boron impurity concentration of N = 5 · 10 ¹⁵ cm ^-3 .

The distance between p + (3), (7), (10) and n + (2), (5), (8) polysilicon regions is 0.7 μm. The length of p + and n + regions of polysilicon is 0.7 μm. The concentration of boron impurities in p + regions of polysilicon is N = 5 · 10 ²⁰ cm ^-3 . The concentration of phosphorus impurities in n + polysilicon regions is N = 1 · 10 ²⁰ cm ^-3

The isolines indicate the electron concentration in the injected electron-hole plasma at a voltage of 1.2 V on the basis of (3), 1.5 V on the collector (8), potential on the emitter (2), (5), (7), (10) equal to zero.

As can be seen in the drawing, the concentration of injected holes in the n-layer (1) and, accordingly, the electron concentration near the p-emitter equal to the condition of electroneutrality is 8 · 10 ¹⁸ cm ^-3 , and exceeds the electron concentration of 2 · 10 ¹⁸ cm ^-3 , determined by the level of doping of the n-layer with an admixture of phosphorus. In the p-layer, the electron concentration of 1 · 10 ¹⁸ cm ^-3 at the interface between the n- and p-layers is less than the concentration of 3 · 10 ¹⁸ cm ^-3 in the n-layer exactly by the value of the impurity concentration. Further from the transition, the concentration decreases due to spreading of the electron flux and their recombination with holes and amounts to 3 · 10 ¹⁷ cm ^-3 near the n-collector at an impurity hole concentration in the p-layer of 5 · 10 ¹⁵ cm ^-3 . The concentrations of injected holes and electrons are almost equal at all points of the existence of an electron-hole plasma.

Fig. 6 shows a cross-sectional view of a planar structure and impurity distribution for a TIROD realized using a technology using ion doping of layers in a silicon substrate and with contact areas (7), (2), (3), (8) obtained by applying polysilicon . The same figure shows the distribution of the concentration of N impurities of boron and phosphorus in the cross section of the TIROD structure with ion implantation of layers and the concentration of electrons and holes in the regions of the pnpn structure forming the thyristor.

The sequence of operations during the technological implementation of TIROD was carried out in a structure with three-layer ion implantation and polysilicon regions of the p-emitter and heavily doped contacts to the regions of n- and p-layers along the following technological route.

A p-silicon substrate with a boron concentration of 1 · 10 ¹⁵ cm ^-3 .

Oxidation of 30 nm in oxygen at a temperature of 1000 ° C.

Photolithography 1, Implantation: P, E = 350 keV, D = 1.5e14 cm ^-2 .

Photolithography 2, Implantation: P, E = 50 keV, D = 2e13 cm ^-2 .

Photolithography 3, Implantation: B, E = 40 keV, D = 1.3e13 cm ^-2 .

Annealing at a temperature of 850 ° С in a nitrogen medium for 10 min.

The formation of polysilicon contact areas and the anode.

Figure 7 shows a conditional section of the TIROD structure, which is obtained after performing a sequence of ion-doping (IL) operations of p-type silicon wafers (pSi) coated with silicon oxide SiO ₂ 300 Å thick through photoresist masks (Ф / Р). For the formation of contacts (7), (2), (8) and the anode (3), polysilicon of different conductivity types pSi *, nSi * is used. Polysilicon films are applied after opening the windows in the oxide to create contacts to the p and n layers.

Ion doping is carried out at different ion energies and dose values, which determines their depth distribution. The use of masks from photoresist provides positioning of the layers.

Fig. 8 shows the topology of the planar structure of the TIROD with ion implantation of n- and p-layers, n-cathode and the sequence of photo masks used to form all areas of the device, including polysilicon contacts to n- and p-layers, n-cathode and p- anode.

To obtain high-frequency characteristics, it is necessary to reduce the width of the p-emitter. This is possible when applying the method of self-formation.

Figure 9 shows the sequence of photomasks used in the formation of a narrow p-emitter by the method of self-formation, when the polysilicon region of the p-anode overlaps an additional layer of silicon nitride by a small amount.

Using a photomask, an area is formed on an additional layer of silicon nitride film, which, when opening windows in an oxide, limits a window in an oxide of size l ₁ . The deposition of boron-doped polysilicon pSi * to form a p-anode of width l ₂ occurs on silicon in a window of width l ₁ . This determines the possibility of obtaining a narrow p-emitter.

To ensure the connection of the device with an external electrical network, a dielectric layer of silicon oxide is applied to the surface of the device, contact windows to all areas and a metal wiring are formed.

For the planar structure of a TIROD with ion implantation of n- and p-layers, an n-cathode and with the above sequence of photo masks, Fig. 10 shows the frequency dependences of the current transfer coefficient h21 and the squares of the conductivities of contact (7) to the p-layer - Y2 ( 4.4), through contact (2) to the n-layer - Y2 (1.1), through contact (4) to the p-anode - Y2 (2.2), through contact (9) to the i-cathode - Y2 (5.5), between the contacts to the anode and cathode - Y2 (5.2).

The cutoff frequency is greater than 1 GHz.

TIROD can be manufactured using molecular beam epitaxy of Si-Ge-C layers for depositing thyristor-forming layers and provide an increase in the cutoff frequency of the device.

High speed and small size allow us to assume that TIROD will find a fairly wide application in integrated microelectronics.

Information sources

1. Bardeen J., Brattain W.H. The Transistor, A Semiconductor Triode, Phys. Rev. 74, 230 (1948).

2. Shockley W. The Theory of PN Junctions in Semiconductors and PN Junctions Transistors, Bell Syst. Tech. J., 28, 435 (1949).

3. B. Heinemann, D. Knoll, R. Barth, D. Bolze, K. Blum, J. Drews, K.-E. Ehwald, GG Fischer, K. Köpke, D. Krüger, R. Kurps, H. Rücker, P. Schley, W. Winkler, H.-E. Wulf. Cost-Effective High-Performance High-Voltage SiGe: C HBTs with 100 GHz fT and BVCEO × fT Products Exceeding 220 VGHz, IEEE International Electron Devices Meeting (2001).

4. Moli J. L., Tanenbaum M., Goldey I. M., Holonyak N, p-n-p-n Transistor Switches, Proc. IRE., 44, 1174 (1956).

5. S.Z. Physics of semiconductor devices. M .: World, 1984

6. I.M.Vikulin, V.I. Stafeev. Physics of semiconductor devices // M .: Radio and communications, 1990, p.225-230.

7. US patent 6700140 B2.

8. US patent 6787816 B1.

9. US patent 6767770 B1.

10. US patent 6841806 B1 - prototype.

Claims (1)

A semiconductor bipolar thyristor triode device designed to rectify, amplify, switch, or generate electrical signals in the form of a pnpn thyristor containing four layers of different types of conductivity, p-anode, middle layers of n- and p-type conductivity, n cathode, characterized in that the middle layers of n- and p-type conductivity have a common output, which is the emitter of the triode, the anode is turned on as the base of the triode, the cathode is turned on as the collector of the triode, the cathode pn junction is turned on in the depletion mode of the p-type middle layer conducted Thus, the anodic pn junction is switched on in the mode of injection of an electron-hole plasma into the middle layers of n- and p-type conductivity.

Images