EP1891674A1

EP1891674A1 - Semiconductor protective structure for electrostatic discharge

Info

Publication number: EP1891674A1
Application number: EP06762020A
Authority: EP
Inventors: Franz Dietz; Volker Dudek; Michael Graf; Manfred Klaussner
Original assignee: Atmel Germany GmbH
Current assignee: Microchip Technology Munich GmbH
Priority date: 2005-06-14
Filing date: 2006-06-13
Publication date: 2008-02-27
Also published as: DE102005027368A1; WO2006133888A1

Abstract

Disclosed is a semiconductor protective structure (10) that is suitable for electrostatic discharge and comprises a first subarea (12) encompassing a first bipolar transistor for diverting the electrostatic discharge as well as a second subarea (14a, 14b) equipped with a second bipolar transistor. The first bipolar transistor and the second bipolar transistor are connected in parallel such that both emitter semiconductor regions (44, 44' ) and both collector semiconductor regions (16, 16') of the first bipolar transistor and the second bipolar transistor are also conductively interconnected, respectively. A first basic semiconductor region (34, 46, 42) of the first bipolar transistor borders a second basic semiconductor region (34', 46', 42') of the second bipolar transistor. The first basic semiconductor region (34, 46, 42) of the first bipolar transistor and the second basic semiconductor region (34', 46', 42') of the second bipolar transistor are provided with an identical conduction type. A first basic collector diode of the first bipolar transistor has a greater breakdown voltage than a basic collector diode of the second bipolar transistor, and/or the first bipolar transistor is provided with greater current amplification than the second bipolar transistor.

Description

Semiconductor protection structure for electrostatic discharge

The present invention relates to a semiconductor protective structure suitable for electrostatic discharge (ESD).

US Pat. No. 6,873,505 B2 discloses a semiconductor component which has a protective circuit for electrostatic discharges, which is connected to a common discharge line (CDL). In one exemplary embodiment of US Pat. No. 6,873,505 B2, the semiconductor component comprises a multiplicity of bond pads, to each of which a protective circuit is assigned. For protection, a controlled semiconductor rectifier (SCR) is composed of a pnp bipolar transistor and an npn bipolar transistor whose triggering voltage is reduced by a zener diode as the trigger element.

US Pat. No. 6,236,087 B1 discloses an ESD component for a connection pin of an integrated circuit with a unipolar characteristic.

In this case, voltages of any polarity can be limited by the ESD component. The ESD component shows a thyristor-like

Behavior, as shown in Figure 4. ESD components with a

Thyristors are disclosed in US 2004/0042143 A1 and US 6,614,061 B2. These also reveal the so-called snap-back effect due to the thyristor structure.

The object of the invention is to provide a semiconductor protective structure for an electrostatic discharge which has no acting Thyristor structure or thyristor-like structure and thus have as little or no negligible snap-back effect.

This object is achieved by the semiconductor protective structure having the features of the independent claim. Advantageous developments are the subject of dependent claims.

As a result, a semiconductor protective structure suitable for electrostatic discharge (ESD) is provided. The semiconductor protective structure has a first portion with a first bipolar transistor for dissipating the electrostatic discharge. In addition, the semiconductor structure has a second subregion with a second bipolar transistor. The bipolar transistors may be lateral or vertical bipolar transistors. The semiconductor protective structure is preferably formed exclusively by the first partial area and the second partial area. The bipolar transistors are advantageously designed as parasitic bipolar transistors of a field effect transistor structure. For example, a drain semiconductor region is the collector region, a source semiconductor region is the emitter region, and a well with a BuIk connection is the base region of the first bipolar transistor and / or of the second bipolar transistor.

The first bipolar transistor and the second bipolar transistor are connected in parallel so that a first emitter semiconductor region of the first bipolar transistor is conductively connected to a second emitter semiconductor region of the second transistor. Furthermore, a first collector semiconductor region of the first bipolar transistor is conductively connected to a second collector semiconductor region of the second bipolar transistor. Preferably, the first emitter semiconductor region directly adjoins the second emitter semiconductor region. Further preferably, the first collector semiconductor region directly adjoins the second collector semiconductor region. A first base semiconductor region of the first bipolar transistor in the first subregion of the semiconductor protective structure directly adjoins a second base semiconductor region of the second bipolar transistor of the second subregion. Therefore, neither a metal nor an insulating material is provided between the first base semiconductor region and the second base semiconductor. Due to the immediate adjacency of the first base semiconductor region to the second base semiconductor region, these are conductively connected to one another.

The first base semiconductor region of the first bipolar transistor and the second base semiconductor region of the second bipolar transistor in this case have the same conductivity type. In a first, preferred embodiment variant, the first base semiconductor region and the second base semiconductor region are p-doped. On the other hand, in a second embodiment variant, the first base half conductor region and the second base semiconductor region are n-doped.

A first variant of the proposed problem provides that a first collector-base diode of the first bipolar transistor has a larger breakdown voltage than a collector-base diode of the second bipolar transistor. The breakdown voltage is preferably influenced predominantly by the Zener effect. The different breakdown voltages of the first collector-base diode and the second collector-base diode preferably cause an at least temporary base cross-current, which flows from the second base semiconductor region into the first base semiconductor region, by forming the semiconductor protective structure.

A second solution variant of the stated task provides that the first bipolar transistor of the first subregion has a greater current amplification than the second bipolar transistor of the second subregion. The difference in the current gain is in a first sub-variant by a difference in the base width between the first bipolar transistor - A -

and the second bipolar transistor. A second sub-variant combinable with the first sub-variant provides a difference in the dopant concentration from the first base half conductor region to the second base semiconductor region for a different current amplification. It is also possible to form the dopant concentration in the emitter semiconductor region of the first bipolar transistor and of the second bipolar transistor differently. This third sub-variant can also be combined with the first sub-variant and / or with the second sub-variant.

The variants of the structure advantageously have the effect that, at least temporarily, a base transverse current flows from the second base half conductor region into the first base half conductor region and opens the first bipolar transistor in order to dissipate the electrostatic discharge. For this purpose, it can furthermore be advantageously provided that a first base-emitter junction of the first bipolar transistor has a lower resistance than a second base-emitter junction of the second bipolar transistor. This can also be effected by virtue of a charge carrier generation in a first collector-base diode of the first bipolar transistor being smaller than a charge carrier generation in a second collector-base diode of the second bipolar transistor.

In a preferred embodiment, it is provided that the particular parasitic bipolar transistors are formed as a closed structure, in particular as a ring structure. Round or oval structures are advantageous as ring structures. These allow the connection which is critical for electrostatic discharges, in the case of pnp bipolar transistors, to arrange the cathode connection of the semiconductor protection structure in the interior region of the semiconductor protection structure. Furthermore, the closed structure reduces edge effects at interfaces, in particular to dielectrics, or even completely excludes them. In one embodiment, a pn junction is provided along the outer region of the closed structure, which completely replicates the closed structure in its shape. The pn junction is formed by the semiconductor protective structure itself. This is advantageously the collector-base pn junction of the first bipolar transistor and the second bipolar transistor.

According to a preferred development, it is provided that within the closed structure, the second partial area is at least partially a curved area. The first subarea is preferably characterized exclusively by straight structural sections.

In a preferred embodiment, the second base semiconductor region has a lower dopant concentration than the first base semiconductor region. If the base region is subdivided into a plurality of dopant profiles, for example, the difference in the dopant concentration is advantageously introduced at least in one of these subdivisions. For example, a (separate) implantation or diffusion of dopants into a region of the base semiconductor region can be reduced by masking the second base semiconductor region.

In an advantageous development, it is provided that a first Zener breakdown voltage of a first collector base diode of the first bipolar transistor of the first partial region is greater than a second Zener breakdown voltage of a second collector base diode of the second bipolar transistor of the second partial region. The zener breakdown voltage is dependent, for example, on a distance between a highly p-doped collector semiconductor region and a highly n-doped base semiconductor region of a pnp transistor. The same applies to an npn bipolar transistor. Furthermore, the Zener breakdown voltage is dependent on a dopant concentration in the base semiconductor region. For a different zener breakdown voltage, therefore, at least one of the parameters of the distance and the dopant concentration is advantageously designed differently. Preferably, the second Zener breakdown voltage is lower than the first Zener breakdown voltage in the first partial area due to a greater distance between a heavily doped base terminal semiconductor region of the second partial region and a highly doped base connection semiconductor region of the second partial region than a first partial region first base semiconductor region of the first partial region and a highly doped collector semiconductor region of the first partial region.

In an advantageous refinement, a connection of a collector semiconductor region to a connection of an emitter semiconductor region of the first partial region and / or of the second partial region to a surface of the semiconductor protective structure is separated by a dielectric. An advantageous extension of this dielectric into the depth of the adjacent semiconductor layer also makes it possible to increase a zener breakdown voltage by extending the distance between the highly p-doped semiconductor region and the highly n-doped semiconductor region into the depth.

Furthermore, a buried dielectric for isolating the collector semiconductor region, the base semiconductor region or the emitter semiconductor region from a substrate is advantageously provided (SOI structure). For this purpose, alternatively or in an advantageous combination, the semiconductor protective structure has a trench structure insulated with dielectric for isolation from an adjacent semiconductor structure.

In the following the invention will be explained in more detail by an exemplary embodiment with reference to drawings. Show it

1 is a schematic plan view of a semiconductor protective structure which is suitable for an electrostatic discharge,

FIG. 2 shows a schematic sectional view through the semiconductor protective structure of FIG. 1, FIG.

3 shows a further schematic sectional view through the semiconductor protective structure of FIG. 1,

Fig. 4 is a schematic first current-voltage characteristic, and

Fig. 5 is a schematic second current-voltage characteristic, the

Semiconductor protective structure of Fig. 1 is assigned.

A schematic plan view of a semiconductor protective structure 10 is shown in FIG. 1. This is suitable for electrostatic discharges and is also referred to as an ESD (electro-static-isharge) structure. This semiconductor protection structure 10 is connected to an integrated circuit 32, which is to be protected against electrostatic discharges by the semiconductor protection structure 10. The integrated circuit 32 is connected to terminals 24 and 30 to a supply voltage source 26. In addition, electrostatic discharges can act via these connections 24 and 30. For protection, the semiconductor protection structure 10 is also connected to these terminals 24 and 30. Advantageously, the semiconductor protection structure 10 is integrated with the integrated circuit 32 to be protected on a semiconductor chip together and manufactured in the same manufacturing process. In contrast to the semiconductor protective structure 10, the plan view of which is shown in FIG. 1, the integrated circuit 32 and the terminals 30, 24 are shown only functionally as a block. The semiconductor protective structure 10 has a first partial area 12 and a second partial area 14a and 14b. While the first partial region 12 has exclusively straight sections within the semiconductor protective structure 10, the second partial region has two curved regions 14a and 14b. In this case, the first partial area 12 and the second partial area 14a, 14b adjoin one another, so that they form a closed structure in the form of an oval. This is also called a ring structure. An anode region 16 is connected to the terminal 24 by means of an anode metallization 22. The anode region 16 is adapted to the shape of the oval structure. The anode region 16 also forms an outer boundary of the oval structure. On the other hand, a cathode region 18 is arranged inside the oval structure. The cathode region 18 is connected to the terminal 30 by means of a cathode metallization 28. The anode region 16 and the cathode region 18 are separated on the surface of the semiconductor protective structure 10 by a dielectric 20 of width a.

FIG. 2 schematically shows a section along the line II-II through the semiconductor protective structure 10. The sectional view is of the

Anodenmetallisierung 22 limited to the cathode metallization 38. With

In this case, with the exception of the anode metallization 22, the semiconductor protective structure 10 is formed symmetrically with respect to the axis of the oval. The

Metallization 22 of the anode and the metallization 28 of the cathode are conductively connected, for example via a silicide layer, not shown, with underlying semiconductor regions 16, 42 and 44.

The semiconductor protection structure 10 for a pnp bipolar transistor will be explained in more detail below. The semiconductor region 16, which is conductively connected to the anode metallization 22, is highly p-doped and forms a collector semiconductor region 16 of the pnp bipolar transistor. Also highly p-doped is the semiconductor region 44, which forms the emitter semiconductor region 44. By contrast, a base semiconductor region 34, 46, 42 is n-doped. Highly n-doped is a connection region 42 of the base semiconductor region (34, 46, 42) which effects a low-resistance connection to the cathode metallization 28. The connection region 42 of the base and the emitter semiconductor region 44 are short-circuited via the cathode metallization 28 and thus always have the same potential. The regions 46, 34 and 40, however, are lower n-doped. In this case, the region 34 again has a lower dopant concentration than the regions 46 and 40. The, a current amplification of the pnp transistor influencing dopant concentration within the base is significantly determined by at least one of the dopant concentrations in the n-doped semiconductor regions 46 and 34. Another influencing factor is the base width W, which is determined by the distance between the two p-doped semiconductor regions 16 and 44 of the collector or of the emitter.

Between the emitter semiconductor region 44 and the collector semiconductor region 16, the dielectric 20 of width a extends. The dielectric 20 can be introduced into a recess STI (shallow trench isolation). As a dielectric, for example, a PECVD oxide can be used. The base width W is dependent on the depth of this filled with the dielectric 20 well. Another dielectric 36 isolates the semiconductor region 34 of the base from the semiconductor substrate 38, so that a so-called SOI structure (semiconductor on insulator) is formed. Furthermore, the semiconductor protective structure 10 may be insulated from adjacent semiconductor structures by isolated trench structures (not shown in FIG. 2). In addition, the semiconductor regions 16, 42, 44, etc. adjacent to the surface are covered by a layer 48 of boron-phosphorus-silicate glass (BPSG).

Due to the short circuit of the base semiconductor lead region 42 and the emitter semiconductor region 44 by the cathode metallization 28, FIG Semiconductor protection structure 10 also a zener behavior. The Zener effect between the collector semiconductor region 16 and the base semiconductor region 42, 46, 34 is thereby likewise governed by a set of parameters, such as the dopant concentration in the semiconductor regions 34 and 46 and by the distance between the highly p-doped collector semiconductor region 16 and the highly n-doped Base semiconductor lead region 42 is affected. If the space charge zone reaches the highly n-doped base semiconductor connection region 42, the Zener breakdown occurs. In addition to the Zener effect, the avalanche effect also produces a charge carrier which can cause the pnp bipolar transistor to switch on.

In order to switch on the pnp bipolar transistor shown in section II-II in the case of an electrostatic discharge earlier, a trigger element is connected in parallel to the collector base diode, which is formed at least by the semiconductor regions 16, 34, 46 and 42. This triggering element causes even at lower voltages a current flow parallel to the collector base diode in the base semiconductor region 34, 46, 42. In the semiconductor protection structure 10, the trigger element is formed by the second portion 14a, 14b. The second portion 14a, 14b is arranged within at least one curvature of the oval structure.

The triggering element is also a further pnp bipolar transistor in the exemplary embodiment, which is shown in FIG. 3 as a schematic sectional view along the section line I-I. A base semiconductor region 34 ', 46', 42 'of the further bipolar transistor directly adjoins the base semiconductor region 34, 46, 42 of the pnp bipolar transistor shown in section in FIG. Both base semiconductor regions 34, 46, 42 and 34 ', 46', 42 'have the same conductivity type. In this embodiment of FIG. 1, a collector semiconductor region 16 'of the further bipolar transistor likewise adjoin the collector semiconductor region 16 of the pnp bipolar transistor shown in section in FIG. 2. In this embodiment of FIG. 1, an emitter semiconductor region 44 'of the further bipolar element also borders. transistor to the emitter semiconductor region 44 of the pnp bipolar transistor shown in section in FIG. The pnp bipolar transistor and the further pnp bipolar transistor are different in terms of their electrical properties.

In order to achieve the different properties, the further pnp bipolar transistor has a different dopant concentration in at least one of the semiconductor regions 16 ', 40', 34 ', 46', 42 'or 44' relative to the respective associated semiconductor region 16, 40, 34, 46 , 42 or 44 on. Another possibility for achieving different properties is a geometrical difference, wherein, for example, a width a 'of the dielectric 20 of the further pnp bipolar transistor of the width a and thus, for example, a base width W of the further pnp bipolar transistor is different from the base width W.

In order to make the current amplification in the further pnp bipolar transistor in the second subregion 14a, 14b smaller than in the first subregion 12, for example, a 'is greater than a. Another possibility for a lower current gain in the second subarea 14a, 14b is to reduce the dopant concentration in the base semiconductor regions 34 'and / or 46' relative to the base semiconductor regions 34 and 46, respectively. Since the second subregion 14a, 14b functions as a triggering element, the charge carrier generation is preferably higher in this second subregion. This can be achieved, for example, by using the Zener effect in the second subregion 14a, 14b at lower voltages.

For smaller zener voltages, the distance between the collector semiconductor region 16 'and the base terminal semiconductor region 42' of the further pnp bipolar transistor of the second portion 14a, 14b is opposite the distance between the collector semiconductor region 16 and the Base terminal semiconductor region 42 of the PNP bipolar transistor of the first portion 12 shortened.

Another possibility, which can also be combined therewith, is that the dopant concentration in at least one of the base semiconductor regions 34 'or 46' of the second partial region 14a, 14b is reduced compared to the corresponding dopant concentration of the base semiconductor regions 34, 46 of the first partial region 12, so that a collector Base space charge zone reaches the n + doping of the base terminal semiconductor region 42 'at lower voltages.

As an alternative to two pnp bipolar transistors, it is also possible to use two npn bipolar transistors. In this case, the conductivity types of the individual semiconductor regions are to be inverted. Preferably, the cathode are arranged inside the oval structure and the anode in the outer region of the oval structure.

The condition caused by a high Zener voltage, so-called snap-back behavior, is shown schematically in FIG. The characteristic 54 shown across collector current Ic and collector voltage UCE of a bipolar transistor shows a snap-back voltage 56 at about 28 volts. By contrast, the characteristic curves 56 of FIG. 5 are achieved by means of the semiconductor protective structure of FIGS. 1 to 3, the semiconductor protective structure having no snap-back effect. The release voltage of about 8 volts can be influenced by electrical properties of the trigger element in the second portion 14a, 14b.

Claims

claims

1. Semiconductor Protection Structure (10) Suitable for Electrostatic Discharge (ESD)

- Having a first portion (12) having a first bipolar transistor for discharging the electrostatic discharge, and - with a second portion (14 a, 14 b), the second

Having bipolar transistor,

wherein the first bipolar transistor and the second bipolar transistor are connected in parallel, so that both emitter semiconductor regions (44, 44 ') and both collector semiconductor regions (16, 16') of the first bipolar transistor and of the second bipolar transistor are each conductively connected to one another,

wherein a first base semiconductor region (34, 46, 42) of the first bipolar transistor borders on a second base semiconductor region (34 ', 46', 42 ') of the second bipolar transistor, wherein the first base semiconductor region (34, 46, 42) of the first bipolar transistor

Bipolartransistors and the second base semiconductor region (34 ', 46', 42 ') of the second bipolar transistor have a same conductivity type, wherein

- a first collector-base diode of the first bipolar transistor has a higher breakdown voltage than a collector-base diode of the second bipolar transistor and / or ^■ - the first bipolar transistor having a greater current gain than the second bipolar transistor.

2. A semiconductor structure (10) suitable for electrostatic discharge (ESD) according to claim 1, wherein both collector semiconductor regions (16, 16 ') of the first bipolar transistor and the second bipolar transistor adjoin one another.

3. semiconductor protective structure (10) suitable for an electrostatic discharge (ESD), according to one of the preceding claims, in which the particular parasitic bipolar transistors are formed as a closed structure, in particular as a ring structure.

A semiconductor protective structure (10) suitable for electrostatic discharge (ESD) according to claim 3, wherein a pn junction completely orbits the closed structure.

5. Semiconductor protection structure (10) suitable for an electrostatic discharge (ESD) according to one of claims 3 or 4, wherein within the closed structure the second subregion (14a, 14b) is at least partially a curvature region and / or the first subregion (12 ) exclusively has straight structural sections.

A semiconductor protective structure (10) suitable for electrostatic discharge (ESD) according to any one of the preceding claims, wherein said second base semiconductor region (34 ', 46', 42 ') has a lower dopant concentration than said first base semiconductor region (34, 46, 42 ).

7. A semiconductor protective structure suitable for electrostatic discharge (ESD) according to claim 1, wherein a first Zener breakdown voltage of a first collector base diode of the first partial region is greater than a second Zener breakdown voltage of a second collector base diode of the second partial region. 14b).

A semiconductor protective structure (10) suitable for electrostatic discharge (ESD) according to claim 7 wherein the second Zener breakdown voltage is lower than the first Zener breakdown voltage due to a lower dopant concentration in the second base semiconductor region (34 ', 46', 42 ').

A semiconductor protective structure suitable for electrostatic discharge (ESD) according to any one of claims 7 or 8, wherein the second zener breakdown voltage is smaller due to a smaller distance between a heavily doped one than the first portion (12)

Base terminal semiconductor region (42 ') of the second base semiconductor region of the second partial region (14a, 14b) and a highly doped collector semiconductor region (16') of the second partial region (14a, 14b) is lower than the first zener breakdown voltage in the first partial region (12).

10. A semiconductor protective structure suitable for an electrostatic discharge (ESD) according to one of the preceding claims, wherein a terminal of a collector semiconductor region (16,16 ') of a terminal of an emitter semiconductor region (44, 44') of the first portion (12) and / or of the second portion (14a, 14b) are separated at a surface of the semiconductor protective structure by a dielectric (20).

11. Semiconductor protective structure suitable for electrostatic discharge (ESD) according to one of the preceding claims, which is a buried

Dielectric (36) for isolation from a substrate (38) and / or a dielectric insulated trench structure for isolation from an adjacent semiconductor structure.