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

Methodology to model spatial variation of work-function and other physical properties in FinFETs

IP.com Disclosure Number: IPCOM000237596D
Publication Date: 2014-Jun-26
Document File: 6 page(s) / 170K

Publishing Venue

The IP.com Prior Art Database

Abstract

A methodology to model spatial variation of work-function and other physical properties in FinFETs, by assigning randomly shaped contiguous regions in space similar physical properties.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 38% of the total text.

Page 01 of 6

Methodology to model spatial variation of work -function and other physical properties in FinFETs

Background

Disclosed is a methodology to model spatial variation of physical properties by assigning randomly shaped contiguous regions in space similar properties. This methodology can be applied to, but is not limited to local work-function variation in the metal gate of a FinFET.

Device scaling is necessary to ensures that upcoming CMOS technology node offers lower cost

per function for a chip. However dimensional scaling increases the importance of the spatial variation of the metal gate work-function, since statistical variation of any device metric is inversely proportional to square root of device gate area. This results in device electrical

performance being more strongly dependent on any statistical fluctuations in the spatial variation of work-function (Idlin, Vt, Ieff). Spatial variation of other physical properties can also be modeled using the proposed approach. For example contact resistance, schottky barrier modulation etc.

Current methodologies for local variation of work-function are unable to treat random spatial extents. Current methodologies to model statistical fluctuation of work-function variation assume a shape or pattern that is repeated throughout the domain of interest. Maintaining randomness in deciding spatial extent is crucial to modeling effect correctly.

Several sources of work-function variation have been identified.

    - METAL GRAIN GRANULARITY : Each grain orientation has a unique work-function leading to spatial variation.

    - Interface effects at high-K/metal, IL/high-K and metal/metal interfaces in the HKMG stack including a) O vacancies at high-K/metal and IL/high-K interfaces, b) Thin dielectric layer at high-K/metal or IL/high-K interface, and c) High-K/metal interface stoichometry variation.

    - Point defects due to varying process conditions, mainly in the metal or high-K bulk or the interfaces. (Al, La, N, F, O , H)


- Stress on the metal gate (insitu doping) .

Other physical properties like contact/gate resistance can also be modeled. Sources of spatially varying contact resistance/gate resistance include charge transfer across interface leading to dipoles.

Spatial variation of work-function: It is well understood in literature that different grain orientations of a given material exhibit different work-functions [1]. The inputs like the

probability of finding a grain orientation and the work-function it corresponds (Fig.1) to are required to generate a spatially varying work-function landscape. In addition to that an estimate of the average grain size is also required [2].

1


Page 02 of 6

Fig.1

Spatially varying contact resistance : In the high-K/metal gate stack, different dipole moments because of charge transfer, result in a spatially varying resistance (gate resistance) as shown in Fig.2. In the source/drain contacts, charge transfer at the metal/semiconductor interface results in a spatially...