A Short Overview of Commercial Inorganic Nanoelectronics

1
A Short Overview of Commercial Inorganic
Nanoelectronics

• Robert J. Davis, Director
• Ohio MicroMD Laboratory
• The Ohio State University
• 04 April 2006

2
Overview

• Current and near-future update on cutting-edge
  silicon technology
• Lesser-known near-future nanotechnology thin
  film magnetic heads
• Parallels between MEMS and nanotechnology
• An attempt at some conclusions

3
Silicon Nanoelectronics Photolithography

• Cutting-edge silicon is currently at the 90 nm
  technology node, using extensions of 193 nm
  photolithography to get there
• AMD began shipping 90 nm devices in late 2004
• Intel claims it entered the nanotechnology
  arena late in 2000, when it began shipping 0.12
  micron parts that had physical gate lengths of
  lt100 nm
• This technology uses steppers (or scanners) to
  print features into UV-sensitive resists, using
  mostly off-axis illumination and phase-shift
  masking
• Initial capital outlay is high and mask costs are
  also very high

Right schematic of modern deep UV Nikon
stepper/scanner (from Nikon web pages)
4
Silicon Nanoelectronics Photolithography (cont.)
Above illustration of the use of phase-shift
masking (PSM) and optical proximity correction
(OPC), together with off-axis illumination, to
achieve a final desired pattern using optical
lithography (borrowed from F.M. Schellenberg,
Mentor Graphics) some of the reasons that mask
costs are so high
5
Silicon Nanoelectronics Photolithography

• What comes next will be interesting
• One strategy is to extend 193 nm photolithography
  to smaller effective wavelengths using
  immersion technology
• This technology uses flavored liquids in the
  exposure chamber
• ? becomes ? / n, where n is the refractive index
  of the liquid
• An alternative (and more radical) strategy uses
  Extreme Ultraviolet (EUV) with 13.5 nm radiation
  from plasma sources in combination with
  reflective optics
• This technology has been in the research phase
  for years but since 1997 it has received
  considerable attention from Intel

Right Intels Extreme Ultraviolet MET tool
(micro exposure tool)
6
Silicon Nanoelectronics Photolithography

• Schism? Recent industry reports suggest the use
  of immersion lithography at the planned 45 nm
  node by IBM and others, but the projected use of
  dry EUV by Intel at the 32 nm technology node
• All of this technology will continue to evolve,
  but will it be cost-effective for anyone but
  cutting-edge silicon?

Right Reflective mask technology for EUV
lithography (Intel)
7
Magnetic Thin Film Heads A lesser-known
nanotechnology

• Meanwhile, thin film heads for magnetic disk
  drives have been steadily been decreasing in size
• An IBM / Hitachi forecast (below) a few years ago
  predicted that CDs (critical dimensions) for
  cutting-edge thin film heads will become smaller
  than those of silicon devices later this decade

8
Magnetic Thin Film Heads A lesser-known
nanotechnology

• In combination with perpendicular recording
  technologies that are in development this
  technology will enable disk drives of 100s of GB
  in the coming years, and the length scales will
  be firmly in the nanotechnology regime

9
Alternative Affordable Nanolithography
Solutions For the rest of us

• Electron beam direct write example Leica
  EBPG-5000 tool at Ohio State
• Nanoimprint lithography (NIL)

10
MEMS versus Nanotechnology What can we learn
from the past?

• Like the current state of nanotechnology, MEMS
  (microelectromechanical systems) technology was
  in its very early stages in the late 1980s and
  early 1990s
• When automotive airbag controllers based on MEMS
  technology became available (Analog Devices
  etc.), for example, it still took several years
  for those devices to displace the existing
  technology in cars
• Now, MEMS accelerometers, gyroscopes, pressure
  sensors, and video chips (e.g, DLP) are
  commercially available with high performance and
  at low relative cost
• MEMS devices are produced and sold because they
  solve specific problems for customers at
  acceptable cost
• Big companies have not always been winners in
  commercialization

Right Image courtesy of Sandia National Labs,
SUMMIT Technologies, www.mems.sandia.gov
11
An attempt at some conclusions

• Technology in the semiconductor and thin film
  head industries are or soon will be in the
  sub-100 nm regime however, their solutions for
  nanolithography are extremely expensive
• Nanotechnology lithography needs that cannot
  afford DUV / EUV solutions will most likely use
  other technologies such as nanoimprint
  lithography or direct-write electron beam
  lithography
• If MEMS technology is a guide, then commercial
  successes in other areas of nanotechnology will
  occur when and if the technology delivers
  solutions for customers at acceptable cost