Page 34 - A Practical Guide from Design Planning to Manufacturing

10 Chapter One

junctions, this device is known as the Bipolar Junction Transistor
(BJT). Sparks and Teal were able to create this structure by adding
dopants to the melt as they pulled the crystal. Pulling the crystal more
slowly created thinner layers and improved performance.
As Shockley used the lab’s team of experimentalists to pursue his
own theories of how to make a better transistor, he actively worked to
exclude John Bardeen from further work on semiconductors. Shockley
was going to make sure he was the one to make any further break-
throughs. In 1951, Bardeen, frustrated and prevented from contribut-
ing to the development of the devices he helped create, left Bell Labs to
become a professor at the University of Illinois. In the same year, AT&T
began licensing the rights to manufacture both point contact and junc-
tion transistors, and the first few products using transistors began to
appear. However, there was a serious limitation.
The first commercially available transistors were all made of germa-
nium. They tended to be very sensitive to temperature and at temperatures
above 75°C they didn’t work at all. Packed tightly with other electronics
these temperatures would not be hard to achieve. The military in particu-
lar mandated components that would function properly in a wide range of
conditions, and military applications were the most lucrative for early sales
of transistors. The problem was that the band gap of germanium was too
small. At room temperature only the carriers intentionally added by dopants
could jump the gap and conduct electricity, but at higher temperatures
many more electrons had enough energy so that germanium stopped behav-
ing like a semiconductor and began behaving like a conductor. The solu-
tion was to replace germanium with silicon.
Silicon’s band gap is almost twice as large as germanium’s, so it is far
less sensitive to temperature and has a much higher maximum operat-
ing temperature. In 1952, Gordon Teal left Bell Labs to join Texas
Instruments ® (TI) with the goal of creating a silicon transistor.
Unfortunately silicon’s melting point is also much higher than germa-
nium’s, and it reacts more easily with oxygen and other contaminants.
Creating pure crystal silicon was far more difficult, but as the man who
had pioneered growing high-quality crystals at Bell Labs, Teal was prob-
ably better suited than any man in the world to overcome these problems.
In 1954, TI produced the first silicon junction transistor. In the same year,
TI began manufacturing components for the first commercially available
transistor radio, still using germanium. The radio sold out wherever it
was offered, and the transistor revolution had truly begun.

The Integrated Circuit
Bill Shockley probably understood the enormous potential of the tran-
sistor better than any man in the world at the time, but he felt he wasn’t