JPS584980A - Semiconductor device - Google Patents

Abstract

PURPOSE:To improve the noise characteristic of a transistor and the current amplification factor of a low current region by preventing the crossing of a SiO2 film on a P-N junction determining breakdown voltage by a metallic wiring thin-film for an electrode. CONSTITUTION:A P base 12 and a P<+> auxiliary layer 11 in shape that adjoins to the base 12 and is deeper are formed to an N epitaxial layer 2 on P type Si 1. An N<+> emitter 13 and said electrode 16 extending to the layers 11, 12 are molded, and the layer 13 is not extended to the layer 12 side. In the N-P-N elements of the layers 13-12-2 and 13-11-2, the former has the high efficiency of injection and narrow base width and shows a high current amplification factor, and the current amplification factor is determined by the layers 13, 12 while using the contacting surface 21 of the layers 11, 12 as a boundary. The emitter electrode 16 is drawn out of the N<+> layer 13, and the crossing of the metallic thin-film 16 on the SiO2 5 on the junction of the layers 12, 13 is prevented. According to this constitution, the characteristics of the transistor is largely improved because the level at a low level of the interface of Si and SiO2 in the vicinity of the junction of the emitter and the base is decreased remarkably and the recrystallization layer of Si-the metal in the alloying process of the formation of the ohmic junction is separated from the dominant junction of the emitter and the base.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is directed to a semiconductor device K1 having an advantageous value in drawing out electrodes.
1.

In a conventional integrated semiconductor integrated circuit device, for example, when the lead electrode metal crosses the PN junction of a so-called Zener diode, which utilizes the reverse breakdown voltage of the diode, the electrode metal has a small amount of 81- 810. Mechanical stress is applied near the interface. As a result, a phenomenon is observed in which the reverse breakdown voltage gradually fluctuates toward higher values over time. Normally, the variation is small at ~100mV, and is not a problem in most cases, but it is necessary to suppress the temperature coefficient of the Zener diode voltage close to zero by using the diode's forward voltage, resistance, etc. Kei fR
Even if the t variation is 10 to 100 mV, it becomes a barrier.

The purpose of this JA@ is to provide 1 m of semiconductor devices that are not susceptible to electrode metals by the PN@ association.

By applying this to the construction of transistors, the noise characteristics,
The current amplification factor, etc. in the collector low current region is improved. If you use this invention zener diode Kxa, you will be happy 10
When a zener diode is obtained by diffusing impurities of s6 strength in the Pal diffusion region KNm, and when electricity is extracted from the metal thin film from the Nfi high degree region, the PH@gold in the atPH region and the NW high concentration region is Pull it out without crossing the top. For this reason, the P layer region t of $1, the JIIIP board deaf region and the inter-board region are formed as auxiliary regions with a lower doping amount than the jllPNfi region, and the J11m1g and Pml regions K ji and extend it. The electrode is pulled out from the Nil high-1 region above the asp bulge region. In this way, the metal wiring is prevented from crossing the PN junction that determines the reverse voltage of the Zener diode. This is a conventional zener diode, in which an epitaxial layer 2 of Nm is formed on a pH silicon substrate, and this layer 2 has p-groups and
The axis is divided into the insulation m* region 8 of the P board that reaches IK, and the element is divided into *X for each divided Nmmg, although not shown in the figure. An Nfi region 4 is formed in the insulating region 8 of the P lever for element isolation. Nff1 area 4 is not shown in the figure 1t41
For example, #1NPN) is simultaneously formed by phosphorus diffusion for forming an entrant in the 20 divided regions of the epitaxial layer. Silicon dioxide jIIS is formed on the epitaxial layer 2 and isolation region 8, and J
A hole is opened in the illI6, and a cathode metal lead wire 6 connected to the isolation region 8 and a negative metal lead wire 7 connected to the Nff1 area 4 are formed, respectively. The reverse breakdown voltage of the diode obtained in this way is about 1 V lower than that of the entrapped junction of the NPN transistor obtained at the same time.

In this conventional diode, as shown at 20 in FIG. 1, the cathode lead electrode always crosses at least part of the upper tube silicon oxide film of the PN junction between regions 8 and 4. As mentioned earlier, the electrode metal thin film @7 applies mechanical stress to the silicon oxide film 6 directly below it, and
i-sio, a low-level unit is formed near the interface where charges are easily captured by IjII. Zener diodes have reverse breakdown not only due to the Zener effect, but also when avalanche breakdown occurs, hot charged particles are generated and spread into the silicon oxide film, which is absorbed by the low-level units, causing air-transmission problems. redistribution occurs, and the breakdown voltage shifts with time. Anneal (600
By doing 'C°8G min) @longitudinal mechanical stress is summed, thus 5t-sio, the above-mentioned low level unit near the interface can be reduced, but not completely.

In this invention, 810 on the PNii joint that determines the breakdown voltage.
．． The metal thin film for one electrode is prevented from crossing.

Example applied to this *@tNPN) transistor 'f:lA
Among the transistor characteristics shown in 2FgK, the noise characteristics,
The low current t and current amplification factor within the region are greatly influenced by the emitter paste junction and the 81-810m interface in its vicinity. Jin's j1411 was designed to avoid this effect.
A space region 12 is formed in the epitaxial layer 2 of N]li in the iIO example. Close to this normal base region 12, this one) is also highly doped, and finally the region 1
2, that is, closer to the substrate IK (formed as a P@0 region 11Vr@auxiliary region).
A high concentration region 18 is formed. Emitter electrode 16tP
l! It is extracted from the Ng high concentration region 18 on the region 11, and is not applied to the region 12@. Therefore, two υ emitter base junctions are formed, and the area 1 and ml! trespective construction (ivy and base transistors, region l
Two transistors coexist, each serving as an emitter and a paste. #i person is #l! Since the injection efficiency is lower than that of the transistor shown in FIG. The boundary is determined by the emitter 111 and the base 12. Therefore, the emitter voltage 4ffi
16 l 111111 area] upper NlllAl11 degree i
810 above any point of the junction between regions 12 and 18, i.e.
, the metal thin m[16 does not cross over it.

Claims (1)

[Claims] A 1M'4-body region with a high degree of conductivity, a continuous 1M'4 semiconductor region with a low II & degree of conductivity, and a 1M'4 body region with a high degree of conductivity
The m8 semiconductor region and vtj1N of different conductive wrinkles formed across the i and 2 semiconductor regions, and each electrode drawn from the 2 semiconductor region and the rattan conductor region are A semiconductor device characterized in that the PNm gold formed at the boundary between the conductive region and the semiconductor region is formed so as not to pass over the PNm gold.