Browse Prior Art Database

Ozone Oxidant for Chemical Vapor Deposition of High Dielectric Constant Insulator Films

IP.com Disclosure Number: IPCOM000104130D
Original Publication Date: 1993-Mar-01
Included in the Prior Art Database: 2005-Mar-18
Document File: 2 page(s) / 97K

Publishing Venue

IBM

Related People

Rubloff, GW: AUTHOR

Abstract

Disclosed is a technique for growth of high quality, high dielectric constant insulator thin films which uses ozone (O&sub3.) as a direct oxidation reactant in a chemical vapor deposition process.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Ozone Oxidant for Chemical Vapor Deposition of High Dielectric Constant Insulator Films

      Disclosed is a technique for growth of high quality, high
dielectric constant insulator thin films which uses ozone (O&sub3.)
as a direct oxidation reactant in a chemical vapor deposition
process.

      Two applications are emerging for high dielectric constant
(&epsilon.) insulators in mainline chip technology - (1) the DRAM
storage node dielectric, and (2) decoupling capacitors.  Higher
&epsilon.  would allow higher capacitance per unit area and/or larger
thickness (broader process window) for the same capacitance per unit
area.  Materials like transition metal oxides and ferroelectrics are
of primary interest.

      By far the most attention has been focussed on Ta&sub2.O&sub5.
(&epsilon.~25, a large value as compared to the "standard"
dielectric SiO&sub2.  with &epsilon.=3.9).  There has already been a
notable amount of published work on Ta&sub2.O&sub5.  At the same
time, deposition/growth of the transition metal oxides must really be
carried out using chemical vapor deposition in order to achieve
extremely high conformality consistent with the large aspect ratios
of projected 3-D structures (especially stacked and trench
capacitors); thermal CVD is expected to show more promise than plasma
CVD due to its higher typical conformality, and the former has indeed
been the subject of considerably more research.  Thermal CVD has been
carried out using halogen precursors (TaF&sub5.  and TaCl&sub5.),
organometallic precursors (primarily Ta(OC&sub2.H&sub5.)&sub5.), and
in some cases enhancement by ultraviolet light (UV).  While work
carried out on Si-containing electrode surfaces has resulted in an
SiO&sub2.  interlayer to suppress leakage current, it is highly
desirable to avoid this layer for extendibility of the technology.

      The major problem in achieving usable CVD Ta&sub2.O&sub5.
films has been to reduce their leakage current, which we believe it
typically too high due to incorrect local bonding of the atoms in the
material.  This is really an issue of how best to achieve the correct
oxidation state of the Ta atoms.  The most successful work so far is
that carried out at Hitachi [1-3], in which two annealing steps were
employed after CVD growth to reduce leakage - first an anneal at
300ºC in O&sub2.  in the presence of UV, and second an anneal at
800ºC in dry O&sub2..  The paper suggests a model in which the
UV/O&sub2.  anneal at moderate temperature repairs defective bonding
sites generally in the film, while the high temperature O&sub2.
anneal corrects local defects.  The learning from this work and other
[4]  work is that oxygen, and particularly excited oxygen in the form
of ozone (O&sub3.)  formed by UV irradiation, is effective in
repairing defective chemical bonds and thereby reducing leakage...