An Errant Dip of the Pen
The history of science is filled with examples of discoveries that
resulted from accidents or mistakes. Brian Trumbore called my
attention to an article in the September 3 Wall Street Journal
titled “In Physics, Big Errors Can get the Ball Rolling Toward
Big Discoveries”. The article, by Sharon Begley, quotes
University of Chicago physicist Michael Turner’s view that big
mistakes, if bold enough, can prompt others to charge into arena
and sometimes emerge with profoundly significant results.
The Journal article cited a paper Albert Einstein co-authored in
1935 with two colleagues in which they pooh poohed the idea in
quantum mechanics that measuring a property of a particle in a
lab can determine the property of a related particle far away,
even trillions of miles away. Since then, researchers have
actually demonstrated that this “spooky action at a distance” is
indeed real. I have to admit that I still can’t believe it.
Another example of a big bold mistake is one made by famed
wheelchair-bound physicist Stephen Hawking and the bet that he
lost to a fellow physicist. You may have seen media reports of
Hawking paying off the bet with a baseball encyclopedia.
Almost 30 years ago, he published a paper stating that all
information is lost when it enters a black hole. The media put it
in more colorfully by saying that, if you stumbled upon a black
hole, you would never be able to tell whether it was made of
Swiss cheese or a collapsed star. Hawking’s paper spurred other
physicists to check this lost information. They concluded he was
wrong. Hawking now has his own mathematical proof that he
was mistaken and hence the lost bet. I’m amused that some now
question whether Hawking’s proof of his mistake is a mistake!
The above are big time mistakes only high-powered physicists
can recognize. I came across a simple mistake that is totally
understandable and that led to a process that had profound
consequences for all of us. This mistake is mentioned in an
article by Pawel Tomaszewski and Robert Cahn in the May 2004
issue of the Materials Research Society (MRS) Bulletin. The
man who made the mistake was Jan Czochralski, who was born
in Poland in 1885. Jan worked in a drugstore in his hometown of
Kcynia and in his teens moved to Berlin, where he also worked
in a drugstore while studying part time in a technical university.
One day in 1916, Czochralski was experimenting with tin and
had melted some of the metal in a crucible. He set the crucible
aside while he made some notes. (Younger readers may have to
be reminded that, until relatively recently, we actually had to use
something called pen and ink.) Well, Czochralski was sitting
there writing his notes and moved to dip his pen in the inkwell.
Mistakenly, he dipped the pen in the crucible of molten tin.
When he pulled the pen out of the crucible, there was a thin
filament of solid tin hanging from the tip of the pen.
When an accident or mistake happens, a prepared mind can
sometimes take the mistake and turn it into gold. Czochralski
was prepared. He wrote a paper, published in 1918 (publication
delayed for two years due to World War I). The paper described
a method for pulling single crystals of metals such as tin, lead
and zinc from a bath of molten metal. Czochralski dipped a fine
glass tube into the molten metal. The glass tube had a hook on
top and was attached to a clockwork mechanism to pull the tube
upward at a constant rate. By controlling the rate of pulling he
could make the crystal wider or thinner.
Normally, a rod of metal such as tin is composed of a whole
bunch of crystals of various orientations. If you etch the surface
of such a rod, you see a pattern of “grains”, patches of single
crystals. If you narrow down the rod thin enough, the chances
are good that only one grain will be exposed to the molten metal
and this single grain or crystal will then propagate as you keep
pulling the rod from the melt. Czochralski went on to form a
research laboratory for a German company that became a leading
metallurgical firm; he also helped found a leading metallurgical
journal and a metallurgical society. He prospered and eventually
returned to his native Poland.
During World War II, he seems to have played a game of
cooperating with the Germans by helping with certain
manufacturing problems and at the same time helping a
resistance movement against the German occupiers. After the
war, he was accused of being a collaborator and was dismissed
from his university job. He was a bitter man and retired to his
hometown of Kcynia, where he founded a drugs and cosmetics
company. He died in 1953, a year after I arrived at Bell Labs.
Over the years, his reputation has been restored and recently a
monument dedicated to him was erected in Kcynia.
When I interviewed at Bell Labs in 1952, one of my interviews
was with Gordon Teal and Carl Thurmond and I accepted the
offer to work in their area. I had never heard of Czochralski nor
was I aware that Teal had taken Czochralski’s crystal pulling to
new heights that would materially influence the solid-state
revolution. I was surprised to read in the MRS Bulletin article
that Teal had to struggle with his Bell Labs colleagues and
management to let him use the crystal pulling method to grow
single crystals of germanium. For details, I referred to the book
“Crystal Fire. The Birth of the Information Age” by Michael
Riordan and Lillian Hoddeson. Teal did have a rough time.
Teal had worked with germanium compounds in the 1920s as a
graduate student and was using a “pyrolytic” method to make
germanium and silicon samples that were not single crystals.
When the transistor was invented in late 1947, Teal pleaded with
Shockley and others to allow him to look into the growth of
single crystals. However, Shockley said that single crystals were
not needed and, if they wanted a single crystal, they could cut out
one of the grains from the pyrolytic germanium. But Teal was
stubborn and, in September of 1948, he encountered a
mechanical engineer, John Little, when both were on their way to
catch the bus that I was to ride during my years at Bell Labs.
Teal knew of Czochralski’s work and by the end of the bus ride,
they had sketched out an apparatus that would use an induction
heating coil that Little had in his lab and a clockwork mechanism
to pull the crystal. Working on their own time without approval,
they quickly built the sketched apparatus with a bell jar and the
heating coil and pulled a scraggly germanium crystal under a
hydrogen atmosphere. They still could not get support for their
Undaunted, they built a 7-foot tall crystal pulling apparatus that
they put on wheels so it could be rolled into a storage closet and
brought out after normal working hours. Spending late nights,
stretching sometimes to 2 or 3 AM, they grew larger and more
uniform crystals. The crystals were also much purer than the
pyrolytic stuff, with lower levels of unwanted impurities and the
“Teal-Little” single crystals proved to have superior electrical
properties that were uniform from crystal to crystal.
Word got around and, by late 1949, Shockley admitted his
mistake. Teal and Little got their own lab and an assistant, Ernie
Buehler, who turned out to be one great crystal grower. No more
wheeling the puller in and out of the closet! When I arrived in
November of 1952, there was a group devoted to crystal pulling.
Teal left Bell Labs the next month. He had seen an ad in the
New York Times. A company known as Texas Instruments
wanted someone to head up their research effort.
Teal and Ernie Buehler had grown single crystals of silicon in
1951 and Teal pushed this work at Texas Instruments. Silicon
was a difficult material and it took until 1954 to make a silicon
transistor worthy of putting into production. When Teal
announced at a meeting in May of that year that he had silicon
transistors in his pocket and that they were in production, the
response was electric. Silicon was on its way. Four years later,
Jack Kilby at Texas Instruments would invent the integrated
circuit and the silicon chip would soon appear.
For several years, I had my own Czochralski-Teal-Little crystal
puller and was happy to grow crystal of various materials the size
of my thumb. Today’s crystal pullers produce silicon crystals
one foot (300 millimeters) in diameter and several feet in length.
Quite a contrast to that first tin filament pulled from an “inkwell”
of molten tin!
Allen F. Bortrum