EP1295326A1

EP1295326A1 - Silicon bipolar transistor, circuit arrangement and method for production of a silicon bipolar transistor

Info

Publication number: EP1295326A1
Application number: EP01949264A
Authority: EP
Inventors: Martin Franosch; Thomas Meister; Herbert Schäfer; Reinhard Stengl; Konrad Wolf
Original assignee: Infineon Technologies AG
Current assignee: Infineon Technologies AG
Priority date: 2000-06-14
Filing date: 2001-06-15
Publication date: 2003-03-26
Also published as: CN1436366A; WO2001097273A1; TW512529B; US7612430B2; JP2004503936A; KR20030028483A; US20030178700A1; CN1248296C

Abstract

The silicon bipolar transistor (100) comprises a base, with a first highly-doped base layer (105) and a second poorly-doped base layer (106) which together form the base. The emitter is completely highly-doped and mounted directly on the second base layer (106).

Description

description

Silicon bipolar transistor, circuit arrangement and method for producing a silicon bipolar transistor

The invention relates to a silicon bipolar transistor, a circuit arrangement and a method for producing a silicon bipolar transistor.

A silicon bipolar transistor, such

Circuit arrangement and such a manufacturing method are known from [1].

A common silicon bipolar transistor has an emitter, a base and a collector.

In the case of the bipolar transistor known from [1], it is known that the maximum oscillation frequency of a bipolar transistor results according to the following rule:

being with

Fmax the maximum oscillation frequency of the bipolar transistor,

F> p the transit frequency of the bipolar transistor,

R _R the base resistance of the bipolar transistor,

C _R C the base-collector capacitance of the bipolar transistor,

referred to as.

As can be seen from [1], it is therefore desirable to reduce the base resistance RR of a bipolar transistor in order to to obtain the highest possible oscillation frequency of the bipolar transistor.

The base resistance RR of a bipolar transistor is determined both by the electrical resistance of the connection region and by the sheet resistance of the base doping profile with doping atoms.

The sheet resistance, which is also referred to as pinch, is inversely proportional to the layer thickness of the base in the case of homogeneous doping of the transistor base with doping atoms.

However, increasing the layer thickness of the base of the bipolar transistor leads to an increase in the base transit time τ for the minority carriers in the bipolar transistor.

Increasing the doping of the base with doping atoms beyond a concentration of 5 × 10 cm causes the breakdown voltage of the transition between the emitter and the base of the bipolar transistor to be reduced to low values and at the same time increases the base-emitter junction capacitance.

In order to reduce the base resistance, it is provided in [1] to dope the emitter of the bipolar transistor low with doping atoms in a concentration of approximately 10 cm

The base of the bipolar transistor known from [1], on the other hand, is highly doped with doping atoms in one concentration

20 -3 of the doping atoms of approximately 10 cm

In this way it is achieved that with low doping of the emitter a high doping of the base is made possible without the blocking capability of the emitter-base junction of the bipolar transistor being lost.

To increase the transit frequency, the base of the transistor described in [1] contains germanium.

Furthermore, [2] describes the reduction in on-board diffusion by adding carbon atoms for a transistor with an epitaxial emitter.

Furthermore, a bipolar transistor is known from [3], which has a very high maximum oscillation frequency f _max of 74 GHz.

[6] describes a bipolar transistor based on - gallium arsenide, with a base layer consisting of two partial base layers, a p -doped first base layer and a. Layer on a collector layer made of n-doped gallium arsenide

+ a p -doped second base layer, each made of gallium-

Arsenide is applied.

The p -doped gallium arsenide partial base layer serves as a diffusion barrier for the zinc doping atoms. An n-doped emitter stop layer is applied to the p-doped gallium arsenide layer, which serves as a barrier layer between the emitter and the base for isolating the zinc, which is intended to ensure that the emitter is n-doped. An n-doped "graded" emitter layer sequence made of gallium arsenide is applied to the emitter stop layer.

[7] furthermore shows a silicon bipolar transistor which contains a base having two base layers, wherein on

+ an n-doped collector, a first, p -doped

Base layer is applied and on this a p-doped second base layer. On the second base layer there is an n- doped, ie weakly doped first intermediate layer applied and only then the highly doped n emitter.

This layer sequence has in particular the disadvantage that an n-doped intermediate layer has to be introduced between the base and the emitter, which means considerable expenditure in terms of production technology, in particular in mass production, and thus causes considerable technological difficulties. Furthermore, the mass production costs for such a transistor are very high.

[8] describes a power transistor in which there is another layer of the same conductivity type as the base layer in the base layer.

[9] describes a transistor in which a conduction zone of the base zone is located in front of the emitter zone, the

Impurity concentration is lower than that of the base zone,

16 3 but at least 10 defects per cm.

[10] describes a semiconductor device in which an emitter layer, an intrinsic base layer which surrounds the emitter layer, the surface of the emitter layer allowing exposure, external base layers and link base layers which lie between the intrinsic base layer and the external base layers, are formed on a collector layer become.

[11] describes a transistor for switching with a partially falling characteristic and a semiconductor body with the

+ +, + + Zone sequence npp n or pnn p.

[12] describes a bipolar transistor and a method for its production, the bipolar transistor having a collector region and insulation regions surrounding it. Above the collector area is a single-crystalline layer sequence and above that Isolation areas arranged a polycrystalline layer sequence, wherein a cover layer is arranged over the base layer, the cover layer being partially or completely removed in the active emitter region.

Further GaAs bipolar transistors are described in [13], [14] and [15].

The invention is therefore based on the problem of specifying a silicon bipolar transistor, a circuit arrangement and a method for producing a silicon bipolar transistor, the silicon bipolar transistor having an increased maximum oscillation frequency fmax compared to a silicon bipolar transistor according to [3].

The problem is solved by the silicon bipolar transistor, the circuit arrangement and by the method for producing a silicon bipolar transistor with the features according to the independent claims.

A silicon bipolar transistor has an emitter, a base and a collector. The entire emitter is highly doped with doping atoms, the doping atoms being of the opposite conductivity type to that for the

Doping of the base regions uses doping atoms. This means that if the base regions are n-doped, the entire emitter is highly p-doped, and if the base regions are p-doped, the entire emitter is highly n-doped. The emitter preferably has polysilicon.

The base is grouped into a first base region and into a second base region, the second base region being doped with doping atoms, for example with boron atoms, in a low concentration, ie the second base region has low doping with doping atoms. In contrast, the first base region has a high doping with the doping atoms, for example with boron atoms.

The terms “low doping” and “high doping” are to be understood in the context of this invention in such a way that the

3 The number of doping atoms per cm in the case of high doping compared to low doping is considerably larger, preferably larger by at least a factor of two.

For example, the second base region can be doped

17 19 3 of 5 x 10 to 1 x 10 doping atoms per cm

19 and the first base region has a doping of 10 to approx.

20, 3

2 x 10 doping atoms per cm.

The invention can clearly be seen in that the base of a bipolar transistor, which is usually doped as homogeneously as possible, is divided into a first region, the first base region, with high doping and a second region, the second base region, with low doping.

In this way the sheet resistance of the base, i.e. the base resistance RR is considerably reduced, it can even be reduced by more than a factor of 5.

According to one embodiment of the invention, the widths of the first base area (first base width W1) and the second base area (second base width W2) can be dimensioned according to the following regulations:

The second base width W2 preferably has a width of 10 nm to 40 nm, which is selected according to the desired blocking voltage of the emitter-base-pn junction, for example, with a blocking voltage of 2 V, a second base width W2 of 20 nm. The first base width Wl is chosen to be as thin as possible. Furthermore, the first base region is doped as high as possible, so that there is no strong widening of the profile during subsequent temperature steps. The first base width W1 can be, for example, 1 nm to 30 nm.

The first base region is located on the collector of the bipolar transistor and, as can be seen from the regulations described above, is preferably as narrow as possible, i.e. formed with the smallest possible first base width W1 and doped as high as possible with doping atoms.

In order to further reduce the diffusion of the individual doping atoms between the first base region and the second base region, it is advantageous, according to one embodiment of the invention, to supply carbon atoms to the base in order to reduce the diffusion of the boron doping atoms, for example.

A gain in the transistor current which possibly decreases with increasing base charge can advantageously be compensated for by adding germanium atoms in accordance with one embodiment of the invention. Furthermore, the transit frequency of the bipolar transistor and thus also the maximum oscillation frequency are further increased by adding Ge.

According to a further embodiment of the invention, the second base region is provided with a concentration

18-3 of approx. 5 x 10 cm and the first base area with

19 -3 3 x 10 cm doping atoms.

As described in [3], the emitter window, i.e. the area in which the emitter is to be formed is opened in a sandwich structure by means of dry etching. The

Viewed from bottom to top, the sandwich structure has: • p + polysilicon, • TEOS,

• nitride.

The side wall of the emitter window is formed by a nitride spacer.

The collector, which is initially still covered by an oxide layer, is exposed by an isotropic wet etching. This results in a polysilicon overhang by undercutting the overlying polysilicon, as described in [4].

Furthermore, it is provided according to a preferred embodiment of the invention to use aluminum atoms or also gallium atoms as doping atoms instead of boron atoms.

However, the use of boron atoms has the advantage that the boron atoms are usually smaller than the other known doping atoms, which of course can be used alternatively

Diffusion speed has, which is particularly advantageous for producing the two base regions with very different doping.

A circuit arrangement with at least one

Bipolar transistor is particularly suitable for use in high-frequency applications, for example in the mobile radio sector or in general for high-speed processors.

In a method for producing a bipolar transistor, after the insulation, that is to say after the collector has been formed, a first base layer which forms the first base region is grown on the collector, preferably by means of gas phase epitaxy using a first partial pressure, for example using diborane (B2Hg). as doping gas. It should be noted that the incorporation of the doping atoms is, in a first approximation, linear to the partial pressure used during the gas phase epitaxy.

A second base layer, which forms the second base region, is grown on the first base layer using a second partial pressure by means of gas phase epitaxy, the second partial pressure being considerably lower than the first partial pressure and also using diborane as doping gas in the context of the gas phase epitaxy of the second base layer the second base layer and thus the second base region have a significantly lower doping than the first base layer, ie the first base area.

With the formation of the base, germanium is added in accordance with an embodiment of the invention in accordance with the procedure described in [5], it being ensured according to the invention that the step profile formed by the first base area and the second base area accordingly, as in the following in connection with FIG. 3 described, is formed. The emitter is applied directly to the second base area. The entire emitter is heavily doped with doping atoms.

Embodiments of the invention are shown in the figures and are explained in more detail below.

Show it

1 shows a cross section through a bipolar transistor according to an embodiment of the invention;

Figures 2a to 2c cross sections through the structure of the bipolar transistor to different

Production times; and FIG. 3 shows a sketch of the doping profile of the bipolar transistor from FIG. 1.

1 shows a bipolar transistor 100 with a base connection 101, an emitter connection 102 and a collector connection 103.

The base connection 101 is coupled via a p-doped polysilicon layer 104 to two base regions which form the base.

A first base region 105 has a high doping of

19 3 x 10 doping atoms, according to this embodiment of boron atoms.

A second base layer 106 has been grown on the first base region 105 as a second base region by means of gas phase epitaxy, the doping of the second base region

18 3 106 is approximately 5 x 10 doping atoms per cm.

The base, in particular the first base region 105, has been grown on a collector layer 107 by means of gas phase epitaxy, as will be explained in more detail below. There is an n -doped layer 108 in the collector layer

+ embedded (n -Buried Layer).

Over the n -doped layer 108, the collector 107, i.e. the collector layer 107, coupled to the collector connection 103.

The method for producing the bipolar transistor described below, as shown in FIGS. 2a to 2c, corresponds essentially to the production method as described in [3] for a Bipolar transistor with a homogeneously doped base has been described.

However, a difference in the manufacturing process is envisaged in the formation of the base layer.

Starting from a silicon-containing collector layer 107, as described in [3], the emitter region, i.e. the area in which the emitter is to be formed at the end of the production process is defined by means of a sandwich structure which is formed on an oxide layer formed by means of a gas phase deposition process (CVD process).

When viewed from bottom to top, the sandwich structure has the following layers:

P-i — polysilicon,

• TEOS,

• nitride.

The side wall of the emitter window is formed by a nitride spacer.

17 -3 The collector layer 107 is 2 x 10 cm

Doping atoms n-doped.

After thin nitride spacers 203 have been formed on the sandwich structure 201, the oxide layer 202 is formed

+ Wet etching below the p -doped polysilicon layer

204 undercut, so that about 0.1 μm wide contact areas 205 arise. In a further step (cf. FIG. 2b), the base layer 206 is grown using gas phase epitaxy at a temperature of 650 ° C. to 900 ° C. and a pressure of 1 to 100 Torr, formed by the first base layer 207 and the second base layer 208.

According to this exemplary embodiment, the gases used in the gas phase epitaxy are a hydrogen carrier gas with 10 to 50 slm, the following gases for incorporating carbon atoms and germanium atoms to maintain the properties of the bipolar transistor and to reduce the diffusion of the doping atoms described below between the first base area and the second base area contains: • dichlorosilane (SiH2Cl ₂ ),

Chlorine-hydrogen (HC1),

Germanium hydrogen (GeH / j),

• Methylsilane (SiH3CH ₃ ), which gases in the context of the gas phase epitaxy at a

-4 -2 partial pressure of 10 - 10 of the total pressure used in the gas phase epitaxy.

Diborane (B2Hg) is used as the doping gas when the first is formed

-5 base layer 207 used with a partial pressure of 10 of the total pressure. Germanium hydrogen is added

-4 added a partial pressure of 10 of the first base layer so that the first base layer has a concentration

19 of germanium atoms of approximately 20% for 3 x 10 boron atoms, so that, for example, the doping profile for germanium described in [5] results, as shown in FIG.

As a result of the dependence of the concentration of the doping in a first approximation, linearly to the partial pressure, doping atoms are doped in the first base layer 207 which is at least a factor 2 larger than that Concentration, ie the doping with doping atoms in the second base layer 208.

The first base width Wl is chosen to be as thin as possible. Furthermore, the first base region is doped as high as possible, so that the profile is not greatly modified during subsequent temperature steps. According to this exemplary embodiment, the first base width W1 is 1 nm to 30 nm.

On the first base layer 207, using

Diboran (B2Hg) the second base layer 208 with a

-6 partial pressure formed from 10 of total pressure.

Again, germanium hydrogen is generated according to the profile shown in Figure 3 during the formation of the second

-5 base layer 208 added, at a partial pressure of 10 of the total pressure.

The second base layer 208 with one is thus formed

18 doping concentration of approximately 5 x 10 boron

3 doping atoms per cm and approximately 5% germanium atoms,

According to this exemplary embodiment, the second base width W2 has a width of 10 nm to 40 nm, which is selected according to the desired blocking voltage of the emitter-base-pn junction, with a blocking voltage of 2 V a second base width W2 of 20 nm.

In a further step (see Fig. 2c) the nitride

Spacer 203 removed using phosphoric acid and according to the method described in detail in [3], the n-doped polysilicon emitter 209 is formed on further spacers 210, which in turn are grown on the second base layer 208. According to this procedure, the doping profile 300 shown in FIG. 3 for the bipolar transistor 100 from FIG. 1 results.

Along the ordinate 301, which describes the local orientation against the growth direction of the individual layers within the bipolar transistor 100, the respective concentration of the doping atoms in the respective layer is represented by means of the abscissa 302.

It is advantageous to make the first base region as narrow as possible and to dope it high in order to keep the transit time of the electrons over the base low.

Starting from an emitter doping curve 303, which represents the doping concentration of the doping atoms in the emitter layer 209, a second basic doping curve 304 is then used along the second

Base width W2 the course of the doping of boron atoms in the

18 3 second base layer 208 of 5 x 10 boron atoms per cm, which is step-shaped, i.e. essentially jumps into a high doping in the first base region, i.e. the first base layer 207 with the first

Base width Wl, in which a doping of boron atoms from

19 3 3 x 10 boron atoms per cm (symbolized by the first

Basic doping curve 305).

The corresponding concentration curve for germanium atoms in the first base layer 207 and second base layer 208 in the base is shown with dashed lines 306. The plateau area, i.e. the second base region contains approximately 5% germanium atoms. The collector side area, i.e. the first base region contains approximately 20% germanium atoms.

It has been found experimentally that such a bipolar transistor with the doping profile shown above a sheet resistance of the base of usually 7 kΩ halved by using the base profile to 3.5 kΩ, the transit time τ of the bipolar transistor increasing only insignificantly from 1.5 ps with a homogeneous base to a transit time with a divided base with different doping to 1.6 ps ,

The following publications are cited in this document:

[1] B. Heinemann et al., Influence of low doped emitter and collector regions on high-frequency performance of SiGe-Base HBTs, Solid State Electronics, Vol. 38, No. 6, pp. 1183-1189, 1995

[2] D. Knoll et al, Si / SiGe: C heterojunction bipolar

Transistors in an Epi-Free Well, Single-Polysilicon Technology, IEDM 98, pp. 703-706, 1998

[3] TF Meister et al. , SiGe Base Bipolar Technology with 74 GHz f _max and 11 ps Gate Delay, IEDM, pp. 739 - 740, 1995

[4] US 5 326 71!

[5] Niu Guofu et al, Noise Parameter Modeling and SiGe

Profile Design Tradeoffs for RF Applications, Proc. Of the 2nd Topical Meeting on Silicon Monolithic Integrated Circuits in RF Systems, pp. 9 - 14, 2000

[6] US 5 132 764

[7] US 5,177,583

[8] DD 230 677 A3

[9] DE OS 151 48 48

[10] DE 42 40 205 AI

[11] DE AS 1 089 073

[12] DE 198 45 789 AI

[13] US 4,593,305 [14] JP 03-280 546

[15] JP 03-192 727

LIST OF REFERENCE NUMBERS

100 bipolar transistor

101 basic connection

102 Emitter connection

103 Collector connection

104 p-doped polysilicon layer

105 First base layer

106 Second base layer

107 collector layer

108 n-doped layer

201 sandwich structure

202 collector layer

203 nitride spacer

204 p-doped polysilicon layer

205 underetched contact area

206 base

207 First base layer

208 Second base layer

209 emitter layer

210 spacers

300 doping profile

301 ordinate

302 abscissa

303 emitter doping

304 doping curve second base region

305 Doping course of the first base region

306 Concentration course base with germanium

Claims

claims

Silicon bipolar transistor, with a base, an emitter that is fully doped with doping atoms, a collector, the base having a first base region and a second base region, the first base region being highly doped with doping atoms, the second base region being doped with atoms is low doped, and wherein the emitter is applied directly to the second base region.

2. Silicon bipolar transistor according to claim 1, wherein the first base region is arranged closer to the collector than the second base region.

3. Silicon bipolar transistor according to claim 1 or 2, wherein the first base region is doped by at least a factor two higher with doping atoms than the second base region.

4. Silicon bipolar transistor according to claim 3, wherein the second base region has a doping of approximately

18 3

Has 5 x 10 doping atoms per cm.

5. Silicon bipolar transistor according to claim 3 or 4, wherein the first base region has a doping of approximately

19 3

3 x 10 doping atoms per cm.

6. Silicon bipolar transistor according to one of claims 1 to 5, in which the second base region has a base width between 10 nm and 40 nm.

7. silicon bipolar transistor according to one of claims 1 to 6, wherein the first base region has a base width between 1 nm and 30 nm.

8. Silicon bipolar transistor according to one of claims 1 to 7, in which the base additionally has further doping atoms with which an essentially abrupt transition between the first base region and the second base region is supported.

9. silicon bipolar transistor according to one of claims 1 to 8, wherein the doping atoms contain boron atoms.

10. Silicon bipolar transistor according to one of claims 1 to 9, in which in the base further carbon atoms for diffusion reduction of the doping atoms.

11. Silicon bipolar transistor according to one of claims 1 to 10, wherein the base contains germanium atoms.

12. Silicon bipolar transistor according to one of claims 1 to

11, in which the emitter contains polysilicon.

13. Circuit arrangement with at least one silicon bipolar transistor according to one of claims 1 to 12.

14. Method for producing a silicon bipolar transistor,

• in which a collector is formed,

• in which a base is formed, In which a first base region is heavily doped with doping atoms,

In which a second base region is doped with doping atoms,

In which an emitter is formed directly on the second base region, and

• in which the entire emitter is heavily doped with doping atoms.