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

Improved Self-Annealing of Copper Using Cobalt or Nickel Underlayer

IP.com Disclosure Number: IPCOM000032868D
Original Publication Date: 2004-Nov-16
Included in the Prior Art Database: 2004-Nov-16
Document File: 4 page(s) / 73K

Publishing Venue

Motorola

Related People

Gregor Braeckelmann: AUTHOR [+3]

Abstract

The self-annealing behaviour of copper on various substrates was investigated. It was discovered that self-annealing is strongly dependent on the barrier layer the copper is deposited on. A fast and significant change in sheet resistance at room temperature was observed when depositing copper on a cobalt or nickel layer as barrier, while a smaller change was seen for traditional tantalum based barriers. This paper describes an improved copper annealing process by utilizing a copper barrier layer that promotes self-annealing. The required anneal after copper deposition to change the microstructure of copper can thereby be reduced or omitted.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 41% of the total text.

Improved Self-Annealing of Copper Using Cobalt or Nickel Underlayer

Gregor Braeckelmann, Sam Garcia, Chris Prindle

Abstract

The self-annealing behaviour of copper on various substrates was investigated.  It was discovered that self-annealing is strongly dependent on the barrier layer the copper is deposited on.  A fast and significant change in sheet resistance at room temperature was observed when depositing copper on a cobalt or nickel layer as barrier, while a smaller change was seen for traditional tantalum based barriers. 

This paper describes an improved copper annealing process by utilizing a copper barrier layer that promotes self-annealing.  The required anneal after copper deposition to change the microstructure of copper can thereby be reduced or omitted. 

Introduction

Copper (Cu) self-annealing refers to the change in sheet resistance and intrinsic stress at room temperature corresponding to a change in texture and grain size.  This effect was observed for Cu deposited by electroplating where a significant change in microstructure (texture and grain size) was seen and studied in detail.[1]  This change typically occurs over rather long periods of time (hours to weeks).  During self-annealing the predominant change is grain growth, which results in lower sheet resistance and lower stress, also, the texture changes as only grains with a preferred orientation grow.  The additives to the plating bath, some of which get incorporated in the Cu film, play in integral part in the self-annealing behavior.[2] 

In the typical process flow for the formation of interconnects in semiconductor devices, Cu has to be annealed after every deposition process to densify the material, reducing void formation, and forming a stable Cu film with minimal variation over plating bath life or queue times.  The temperatures for this Cu anneal are typically 250 – 400°C.  The effect of annealing at elevated temperatures and self-annealing on the Cu properties are surprisingly similar, however, the control of self-annealing is challenging.

 We present here a study of the effect of the substrate, underlying the Cu film, on the self-annealing properties of Cu.  Using different substrate materials it is shown that the self-annealing can be promoted or hindered. 

Experiment

The experiments were carried out on 8” wafers using commercially available deposition systems from Applied Materials (PVD) or Semitool (electroplating).

Self-annealing of Cu films deposited on blanket oxide wafers was studied using the following barrier layers:

– 300Å TaN/Ta (ultra Ta)

– 250Å Ta

– 250Å Co

– 250Å Ni

– 250Å Ta / 50Å Co

Following the barrier deposition, a 1000Å Cu seed was deposited without breaking vacuum.  Cu was then electroplated with 3000 or 8000Å thickness.  Sheet resistance of the film stack was measured immediately after deposition and in regular intervals for ten days.  Finally, a last measurement was taken after an anneal at 250°C...