Download replacing silicon: carbon nanotube transistors

Mahboobin 10:00
L04
Disc laimer—This paper partially fulfills a writing requirement for first year (freshman) engineering students at the University
of Pittsburgh Swanson School of Engineering. This paper is a student, not a professional, paper. This paper is based on
publicly available information and may not provide complete analyses of all relevant data. If this paper is used for any purpose
than these authors’ partial fulfillment of a writing requirement for first year (freshman) engineering students at the University
of Pittsburgh Swanson School of Engineering, the user does so at his or her own risk.
REPLACING SILICON: CARBON NANOTUBE TRANSISTORS
Dillon Axarlis ([email protected])
CARBON NANOTUBES AS A
REPLACEMENT FOR SILICON
Carbon Nanotubes are small cylinders made up of
graphite with a wall width of one atom. Many uses for carbon
nanotubes are being researched, but the primary one is to
replace silicon as a transistor in computer chips. Silicon is
quickly becoming outdated and it will soon be at the limits of
its performance. Replacing silicon transistors with carbon
nanotubes will lead to faster more efficient computers that
require less energy and as a result are better for the
environment. Carbon nanotube transistors will also help
create more powerful supercomputers to help scientists
combat society’s everyday problems. However, there are
some problems with carbon nanotubes that up until recently
needed to be solved. Even after these problems are solved, a
way to mass produce carbon nanotubes will have to be found.
Despite these challenges, it is imperative that carbon
nanotubes replace silicon as a material in transistors.
THE LIMITS OF SILICON
Moore’s Law is a function that dictates the speed at
which an affordable CPU increases in power. Created by CoFounder of Intel Gordon Moore in 1965, Moore’s Law has
held out for 50 years with one revision by Moore himself in
1975 changing the time limit from one year to every two
years. In its current form, Moore’s Law states that the number
of transistors in an affordable CPU will double every two
years [1]. The problem with the computer industry is,
however, that Moore’s Law is dying. As computers keep
getting faster and faster, the limit of Moore’s Law creeps
closer and closer. In order to double the number of transistors
on a chip the size of the transistor has to decrease at an equal
rate. Current CPU manufactures are running into problems
manufacturing these transistors however. In wake of delaying
its latest chipset from 2016 too late 2017, Intel has effectively
killed Moore’s Law. Furthermore, Intel has declared that
improvements in current transistor design can only continue
for 5 more years. The problem is that current transistors are
made up of silicon [2]. Although silicon has been at the
forefront of transistor design, As transistor size decreases,
electron mobility decreases. This causes an increased power
consumption and a larger chance of faulty transistors. In
order to continue the development of CPUs a new material
University of Pittsburgh, Swanson School of Engineering 1
10.04.2016
with a higher electron mobility must be found and scientists
have less than five years to find it [3]. One of these proposed
materials is carbon nanotubes. Carbon nanotubes have many
advantages over silicon with the major benefit being size [4].
WHAT ARE CARBON NANOTUBES?
What exactly are carbon nanotubes? Carbon nanotubes
are a cylindrical fiber of pure carbon graphite. These
nanotubes are extremely small at a diameter of 0.7-50 nm.
Comparatively, the current silicon transistor is 14nm and the
projected minimum for a silicon transistor is 5nm. Carbon
nanotubes have other advantages over silicon. These include
the fact that the hexagonal pattern of bonded carbon atoms
creates a phenomenon known as electron delocalization. This
means that electrons can flow freely through carbon
nanotubes with no impedance. This solves the electron
mobility problems of silicon transistor. Carbon nanotubes
high electron mobility also means that CPUs utilizing them
will require less voltage resulting in a computer will have
much lower power consumption and generate less heat [5].
Although carbon nanotubes may sound wonderful, there are
many road blocks engineers face before they can be used in
the cheap affordable CPUs listed under Moore’s Law.
THE CHALLENGES WITH CARBON
NANOTUBES
In order to make carbon nanotubes a viable replacement
for silicon the challenges with transforming the nanotubes to
transistors must be overcome. Development on carbon
nanotubes began over 25 years ago with a paper written by
Sumio IIjima in 1991. If carbon nanotubes have been out for
so long than why haven’t they replaced silicon yet? This boils
down to three problems with making a carbon nanotube a
transistor. These problems are a mixture problem involving
the impurities of carbon nanotubes, an electrical resistance
problem that occurs when trying to interface carbon
nanotubes with regular conductors, and an alignment
problem when trying to place an arrangement of carbon
nanotubes on a computer chip [6].
The Mixture Problem
When carbon nanotubes are manufactured two types of
carbon nanotubes are produced in a seemingly unavoidable
mixture. The problem with this is that one of the two types of
carbon nanotubes created is not suitable for the production of
transistors. The two types of carbon nanotubes created are a
metallic carbon nanotube and a semi conductive carbon
nanotube. Fortunately, as of April of 2015, a small team at
the University of Illinois has devised a way of separating the
metallic carbon nanotubes from the semi conductive ones.
The process laying out carbon nanotubes of a sheet of metal.
Than the carbon nanotubes were covered with an organic
material. Because the metallic carbon nanotubes conduct
electricity, when the sheet was charged with an electric
current, the metallic carbon nanotubes heated up and burned
of the organic material. This caused only the metallic carbon
nanotubes to be exposed to an etching compound that
destroyed the metallic nanotubes leaving only pure
semiconducting carbon nanotubes behind. The method
devised by the University of Illinois is relatively inexpensive
making it perfect for the large scale manufacturing of carbon
nanotube transistors [6].
The Electrical Resistance Problem
With the problem of separating carbon nanotubes solved
scientists moved onto the problem of getting carbon
nanotubes to interface with other materials. It would be the
researchers at IBM who would figure out how to solve this
problem. A problem with too much electrical resistance and
carbon nanotubes may seem confusing. Nanotubes by
themselves have amazing electrical properties. However, it is
when the metallic contacts required to connect the transistors
are added to a large electrical resistance occurs. This
resistance makes traditional carbon nanotube transistors
slower than their silicon counterparts. IBM found a solution
to this problem by changing the orientation of a nanotube
transistor. Traditionally the contacts for a nanotube transistor
are attached to the top of the nanotube. By attaching the
contacts to the ends rather than the ends of the carbon
nanotube, IBM found that they can negate the electrical
resistance issue found on nanotube transistors smaller than
10nm. This method allows transistors to be made as small as
the technology limits without having to worry about the
contact size. It also allows IBM to make many transistors out
of one carbon nanotube further decreasing chip size. With
this problem solved, IBM believes that it is on track to deliver
carbon nanotube based computers by 2020 [7]. However, in
order to mass produce carbon nanotube based CPUs the
transistors must be able to be efficiently arranged on a
computer chip.
The Arrangement Problem
With the ability to make carbon nanotube transistors
made up of pure semiconducting nanotubes, scientists and
engineers focused on arranging the nanotube transistors onto
a single wafer. In order to do this the University of Wisconsin
developed a process called dose-controlled, floating
evaporative self-assembly [6]. This process involves placing
thin strips of carbon nanotubes on a substrate and submerging
it is a solvent. The substrate is the vertically removed from
the solvent and the substrate is baked to evaporate of any
remaining solvent. This process results in a densely packed
group of carbon nanotubes [8]. University of Wisconsin’s
carbon nanotube transistors performed flawlessly and was
able to achieve current that was 1.9 times that of its silicon
cousin. For the first time, carbon nanotubes outperformed
silicon and the had done it by an impressive margin.
However, the process that the University of Wisconsin used
to arrange the transistors is not ready for mass production.
Until a way to mass produce these carbon nanotube
transistors for use in integrated circuits and CPUs, carbon
nanotubes will not be available for both commercial and
consumer use [9].
WHAY ARE CARBON NANOTUBE
TRANSISTORS IMPORTANT?
Although carbon nanotubes are not yet able to be mass
produced, carbon nanotubes are important for the future of
society. On the small scale, carbon nanotubes lead to
improvements of our everyday lives. Carbon nanotubes
promise faster more energy efficient computers, cell phones
and electronic devices. Carbon nanotube transistors are a
huge leap forward in the green initiative. As the human race
becomes more dependent on computers, more and more
energy is used to support these devices. By creating
electronic devices with longer battery lives and less power
consumption we reduce our dependence on power plants that
burn fossil fuels. More efficient computers also help extend
the battery life on computer heavy electric cars such as the
Tesla Model S [9]. Another strong benefit of Carbon
nanotubes is their use in the high power computing industry.
Some of the main uses of supercomputers are to understand
climate change, develop better batteries and super
conductors, and improve drugs and medications. With the
current technology, the power of supercomputers barley
increases as the years go by. With carbon nanotubes more
powerful super computers can be made to help increase the
speed at which scientists can discover new ways to combat
climate change [2]. With few alternatives to silicon there is a
distinct need for carbon nanotubes to take its place in circuit
chips. The time on the clock is ticking as silicon is expected
to reach its limit by 2019 [6]. Fortunately, IBM thinks that it
can commercially produce carbon nanotube transistors by
2020 [7].
CARBON NANOTUBES WILL CHANGE
THE FUTURE
Silicon is becoming outdated and is an inefficient way
to tackle the problems of the future. Carbon nanotubes are its
most prominent replacement. The challenges that go along
with replacing silicon with carbon nanotubes are being
overcome as we speak and it is just a matter of developing a
way to mass produce carbon nanotube transistors. By
developing carbon nanotubes, we create faster more efficient
computers that help us solve modern problems and lead to a
better cleaner earth.
SOURCES
[1] “Moore’s Law.” MemeBridge Accessed 10.30.16
http://www.mooreslaw.org/
[2] T. Simonite. “Moore’s Law Is Dead. Now What?” MIT
Technology Review. 5.13.16. Accessed 10.30.16.
https://www.technologyreview.com/s/601441/moores-lawis-dead-now-what/
[3] S. Anthony “7nm, 5nm, 3nm: The new materials and
transistors that will take us to the limits of Moore’s law.”
ExtremeTech.
7.26.13
Accessed
10.30.16
https://www.extremetech.com/computing/162376-7nm5nm-3nm-the-new-materials-and-transistors-that-will-takeus-to-the-limits-of-moores-law
[4] L. Wood. “Carbon nanotubes in a race against time to
replace silicon.” ComputerWorld 11.10.15. Accessed
10.30.16.
http://www.computerworld.com/article/3002260/emergingtechnology/carbon-nanotubes-in-a-race-against-time-toreplace-silicon.html
[5] “What Are Carbon Nanotubes?” Nanocomp
Technologies,
INC.
Accessed
10.30.16
http://www.nanocomptech.com/what-are-carbon-nanotubes
[6] K. Murane. “Carbon Nanotubes Are Getting Closer To
Making Our Electronic Devices Obsolete.” Forbes 9.8.16
Accessed
10.30.16.
http://www.forbes.com/sites/kevinmurnane/2016/09/08/carb
on-nanotubes-are-getting-closer-to-making-our-electronicdevices-obsolete/#26a2d40943bb
[7] M. Orcutt. “IBM Reports Breakthrough on Carbon
Nanotube Transistors.” MIT Technology Review. 10.1.15
Accessed
10.30.16
https://www.technologyreview.com/s/541921/ibm-reportsbreakthrough-on-carbon-nanotube-transistors/
[8] Y. Joo, G. Brady, M. Arnold, et al. “Dose-Controlled,
Floating Evaporative Self-assembly and Alighnment of
Semiconducting Carbon Nanotubes from Organic Solvents.”
ACS Publications 2014 pp. 1-7
[9] J. Fimgas. “Carbon nanotube transistors promise faster,
leaner processors.” Engadget. 9.5.16 Accessed 10.30.16
ACKNOWLEDGMENTS
This paper is inspired by a man who I only know as
Demonseed Elite who first introduced me to the topic of
carbon nanotubes and helped facilitate my interest in the
world of modern computing solutions.