Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Soft Error Reducing Shield

IP.com Disclosure Number: IPCOM000051212D
Original Publication Date: 1982-Aug-01
Included in the Prior Art Database: 2005-Feb-10
Document File: 3 page(s) / 48K

Publishing Venue

IBM

Related People

Sai Halasz, GA: AUTHOR [+2]

Abstract

A problem with circuits of small dimensions caused by ionizing radiation is referred to as the "soft error" problem. An effective scheme for dealing with the soft error problem is the use of a buried n shield composed to a blanket implanted p barrier with a lattice or grid of n regions disposed therein, which operates by collecting and dispersing the carriers drifting toward the sensitive nodes.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 3

Soft Error Reducing Shield

A problem with circuits of small dimensions caused by ionizing radiation is referred to as the "soft error" problem. An effective scheme for dealing with the soft error problem is the use of a buried n shield composed to a blanket implanted p barrier with a lattice or grid of n regions disposed therein, which operates by collecting and dispersing the carriers drifting toward the sensitive nodes.

Experiments and simulations show that the shield does not permit any of the charges to cross through it, and the only ones collected on the surface are those created between the shield and the surface. Simulations also show that the effectiveness of the shield does not depend on the exact line-up of the n regions with the collecting nodes on the surface. The combination of n disperser with the p barrier acts together. The p barrier holds back the carriers long enough to be collected in the n layer. This picture holds until the smallest lateral separation between parts of the n layer does not increase beyond approximately 15-20 micrometers. The fact that a precise positional relation between the n layer and the surface nodes is not required should afford a significant simplification in implementing the shield. This article proposes the introduction of the n layer with only very crude masking which has to show no alignment relative to the surface structures.

Simulation shows that a shield of long (order of hundreds of microns) closely spaced n strips with an appropriately placed p barrier is effective. The distance between the long n lines should be no more than 10-15 micrometers. The width of the n layers is determined mainly from the point of view of minimizing their impact on the circuit parameters. A width comparable to their separation is adequate.

A combination of n and p layers can be implemented as follows. The p layer is a blanket implant' The desired depth and dose will be discussed later. The main point is the implementation of the long n lines, which is done through a mechanical mask for a whole wafer. In the mechanical mask about 10 15 micron wide cuts have to be made that are approximately the same distance apart. The walls of the cuts must be vertical, and the aspect ratio has to be high. The mask has to be made of a material of sufficient strength to support such a relatively fragile structure (Figs. 1a and 1B). Such a structure can be manufactured from a high strength metal or silicon. It is known that, in (110) Si, anisotropic etch...