Stocks and News
Home | Week in Review Process | Terms of Use | About UsContact Us
   Articles Go Fund Me All-Species List Hot Spots Go Fund Me
Week in Review   |  Bar Chat    |  Hot Spots    |   Dr. Bortrum    |   Wall St. History
Stock and News: Hot Spots
  Search Our Archives: 
 

 

Dr. Bortrum

 

AddThis Feed Button

http://www.gofundme.com/s3h2w8

 

   

03/28/2007

Melting and Freezing

Melting and freezing have at times been uppermost on people’s
minds in our area of New Jersey this past couple of months.
Many have fallen on the treacherous wintry mix of snow and ice
that has finally melted in the relative warmth of spring. I had a
few scary moments myself when I thought I was going to join
the ranks of the fallen on our ice-covered driveway. Hopefully,
golf is just around the corner. On the other hand, some years ago
I managed to fall and break my leg on a golf course in August.
(As longtime readers know, I take any chance to segue into
noting that I broke the leg on the same hole on which I earlier
had a hole-in-one.)

Back to melting, I’m an authority on the subject. Actually, that’s
not true. However, my first publication at Bell Labs was a paper
in 1955 in the Journal of Physical Chemistry titled “On the
Melting Point of Germanium”. I’ve probably mentioned this
work before but there is reason for bringing it up again. In 1955,
there were wide discrepancies in the reported values of
germanium’s melting point. In a period of over four decades,
there were at least 14 different melting points ranging from 925
to 975 degrees Celsius, a remarkably large range of 50 degrees
uncertainty for an element that in the 1950s could be made
extremely pure.

In our work, Francis Hassion, Carl Thurmond and I measured
both the melting and freezing points of germanium in
experiments that involved either the cooling of molten or the
heating of solid germanium. My contribution was pretty simple-
minded. I heated small pieces of germanium in sealed silica
(SiO2) tubes in a furnace, taking the samples out of the furnace
periodically as I raised the temperature to see if they had melted.
This bracketed the melting point to within a degree or so. My
melting point agreed with the freezing point measurements of my
coworkers. Our results confirmed more precise measurements
made by Bell Labs colleagues E. Greiner and P. Breidt, who
reported a value of 937.2 degrees Celsius with an uncertainty of
plus or minus 0.5 degree.

Why bring up this old work here? Last fall, I clipped a brief item
on melting germanium that I found shocking and disturbing.
This week, I found a detailed article on this work by Paul Preuss
titled “Nanocrystals Are Hot” on the Web site of the Lawrence
Berkeley National Laboratory. The work was published in
Physical Review Letters and had 13 authors from the University
of California at Berkeley and one from Australian National
University in Canberra. What the researchers did was to measure
the melting point of extremely tiny nanocrystals of germanium
embedded in silica (SiO2) glass.

What was expected? Let’s look at what happens when a crystal
melts. In a crystal of germanium, atoms are held in place by
strong bonds linking them to other atoms. The atoms on the
surface, however, are more loosely bound because they don’t
have any atoms above them to bond to. As we heat the crystal,
these surface atoms are freer to move around and they form a
liquid layer before the rest of the crystal melts. So, what might
happen if, instead of a relatively large crystal, we measure the
melting point of a very tiny crystal? The smaller the crystal, the
bigger the fraction of the atoms on the surface. With relatively
more surface atoms, we should not be surprised if the tiny crystal
melts well below the normal bulk melting point.

Take the case of gold. If you make particles of gold small
enough, nanoparticles standing free, you can melt gold at room
temperature! This is just one example that indicates why there’s
so much excitement about the world of nanomaterials. Things
can behave completely differently in the nanoworld. Germanium
is a semiconductor and nanocrystals of semiconductors with only
hundreds or a few thousand atoms can melt 300 degrees or more
below the normal melting point of 937. What the Berkeley
workers did was to embed crystals of germanium in silica glass,
the crystals averaging only 2.5 nanometers in diameter.
(Remember, a nanometer is a billionth of a meter, a millionth of
a millimeter.)

I can see why there were 13 authors on the paper. To make the
silica glass, they first oxidized a very thin silicon wafer in steam
to form an SiO2 film. Next they implanted germanium ions in
the thin silica. This, I assume, means they had to use a machine
capable of shooting ions into the glass film, only 500 nanometers
thick, about a hundredth of the thickness of a human hair. Then,
to get crystals of germanium they annealed the sample at 900
degrees to get the germanium atoms to clump together and form
crystals. How do you tell whether you have crystals? You thin
the film down to 300 nanometers so you can see through it with
an electron beam in a transmission electron microscope (TEM).

In a TEM, the electron beam forms diffraction rings if it passes
through a crystal. So, now you heat the sample up to above 937
degrees and nothing happens! The rings are still there showing
the germanium is still crystalline. You go up to a hundred
degrees above that and still no melting! You are understandably
shocked but finally at about 200 degrees above 937 the rings
disappear; the germanium has melted. Getting over the shock,
you cool down the sample below 937 and nothing happens until
you get about 200 degrees below that. Finally the rings come
back and the germanium has frozen. This supercooling would
not have surprised me. At Bell Labs, I did some experiments
where I was cooling down silica tubes containing tiny droplets of
molten germanium that cooled well below red heat and they
would light up like twinkling lights on freezing due to the heat
generated when they froze – there was lots of supercooling.

The researchers were surprised the embedded nanocrystals
melted way above, instead of way below, the bulk melting point.
Back in the 1950s, there was a fellow at GE, David Turnbull,
who was a leading light in the theory of crystal growth. Daryl
Chrzan, one of the Berkeley team, took Turnbull’s theory and
modified it. With Chrzan’s modifications, Turnbull’s theory
predicts the high melting point and the equally low freezing point
for the nanocrystals. The explanation has to do with the solid-
glass interface between the silica glass and the germanium
nanocrystal. In essence, the silica constrains the formation of a
nucleus of liquid on melting or a solid on freezing.

As I started to write this final paragraph, I looked back through
my early papers and found that I had completely forgotten a very
short discussion I had written on a paper by Richard Oriani. My
discussion appeared in 1956 in the Journal of The
Electrochemical Society and contained a reference to a paper in
1950 by David Turnbull and a colleague. They reported that
small samples of certain metals could be supercooled hundreds
of degrees below their melting points! My brilliant idea was to
make a cell in which one electrode was supercooled liquid and
the other solid metal. I even remarked in the discussion that an
attempt was underway to perform this experiment with
germanium. I didn’t succeed. Maybe I should try again with
nanomaterials? On the other hand, maybe not!

Allen F. Bortrum



AddThis Feed Button

 

-03/28/2007-      
Web Epoch NJ Web Design  |  (c) Copyright 2016 StocksandNews.com, LLC.

Dr. Bortrum

03/28/2007

Melting and Freezing

Melting and freezing have at times been uppermost on people’s
minds in our area of New Jersey this past couple of months.
Many have fallen on the treacherous wintry mix of snow and ice
that has finally melted in the relative warmth of spring. I had a
few scary moments myself when I thought I was going to join
the ranks of the fallen on our ice-covered driveway. Hopefully,
golf is just around the corner. On the other hand, some years ago
I managed to fall and break my leg on a golf course in August.
(As longtime readers know, I take any chance to segue into
noting that I broke the leg on the same hole on which I earlier
had a hole-in-one.)

Back to melting, I’m an authority on the subject. Actually, that’s
not true. However, my first publication at Bell Labs was a paper
in 1955 in the Journal of Physical Chemistry titled “On the
Melting Point of Germanium”. I’ve probably mentioned this
work before but there is reason for bringing it up again. In 1955,
there were wide discrepancies in the reported values of
germanium’s melting point. In a period of over four decades,
there were at least 14 different melting points ranging from 925
to 975 degrees Celsius, a remarkably large range of 50 degrees
uncertainty for an element that in the 1950s could be made
extremely pure.

In our work, Francis Hassion, Carl Thurmond and I measured
both the melting and freezing points of germanium in
experiments that involved either the cooling of molten or the
heating of solid germanium. My contribution was pretty simple-
minded. I heated small pieces of germanium in sealed silica
(SiO2) tubes in a furnace, taking the samples out of the furnace
periodically as I raised the temperature to see if they had melted.
This bracketed the melting point to within a degree or so. My
melting point agreed with the freezing point measurements of my
coworkers. Our results confirmed more precise measurements
made by Bell Labs colleagues E. Greiner and P. Breidt, who
reported a value of 937.2 degrees Celsius with an uncertainty of
plus or minus 0.5 degree.

Why bring up this old work here? Last fall, I clipped a brief item
on melting germanium that I found shocking and disturbing.
This week, I found a detailed article on this work by Paul Preuss
titled “Nanocrystals Are Hot” on the Web site of the Lawrence
Berkeley National Laboratory. The work was published in
Physical Review Letters and had 13 authors from the University
of California at Berkeley and one from Australian National
University in Canberra. What the researchers did was to measure
the melting point of extremely tiny nanocrystals of germanium
embedded in silica (SiO2) glass.

What was expected? Let’s look at what happens when a crystal
melts. In a crystal of germanium, atoms are held in place by
strong bonds linking them to other atoms. The atoms on the
surface, however, are more loosely bound because they don’t
have any atoms above them to bond to. As we heat the crystal,
these surface atoms are freer to move around and they form a
liquid layer before the rest of the crystal melts. So, what might
happen if, instead of a relatively large crystal, we measure the
melting point of a very tiny crystal? The smaller the crystal, the
bigger the fraction of the atoms on the surface. With relatively
more surface atoms, we should not be surprised if the tiny crystal
melts well below the normal bulk melting point.

Take the case of gold. If you make particles of gold small
enough, nanoparticles standing free, you can melt gold at room
temperature! This is just one example that indicates why there’s
so much excitement about the world of nanomaterials. Things
can behave completely differently in the nanoworld. Germanium
is a semiconductor and nanocrystals of semiconductors with only
hundreds or a few thousand atoms can melt 300 degrees or more
below the normal melting point of 937. What the Berkeley
workers did was to embed crystals of germanium in silica glass,
the crystals averaging only 2.5 nanometers in diameter.
(Remember, a nanometer is a billionth of a meter, a millionth of
a millimeter.)

I can see why there were 13 authors on the paper. To make the
silica glass, they first oxidized a very thin silicon wafer in steam
to form an SiO2 film. Next they implanted germanium ions in
the thin silica. This, I assume, means they had to use a machine
capable of shooting ions into the glass film, only 500 nanometers
thick, about a hundredth of the thickness of a human hair. Then,
to get crystals of germanium they annealed the sample at 900
degrees to get the germanium atoms to clump together and form
crystals. How do you tell whether you have crystals? You thin
the film down to 300 nanometers so you can see through it with
an electron beam in a transmission electron microscope (TEM).

In a TEM, the electron beam forms diffraction rings if it passes
through a crystal. So, now you heat the sample up to above 937
degrees and nothing happens! The rings are still there showing
the germanium is still crystalline. You go up to a hundred
degrees above that and still no melting! You are understandably
shocked but finally at about 200 degrees above 937 the rings
disappear; the germanium has melted. Getting over the shock,
you cool down the sample below 937 and nothing happens until
you get about 200 degrees below that. Finally the rings come
back and the germanium has frozen. This supercooling would
not have surprised me. At Bell Labs, I did some experiments
where I was cooling down silica tubes containing tiny droplets of
molten germanium that cooled well below red heat and they
would light up like twinkling lights on freezing due to the heat
generated when they froze – there was lots of supercooling.

The researchers were surprised the embedded nanocrystals
melted way above, instead of way below, the bulk melting point.
Back in the 1950s, there was a fellow at GE, David Turnbull,
who was a leading light in the theory of crystal growth. Daryl
Chrzan, one of the Berkeley team, took Turnbull’s theory and
modified it. With Chrzan’s modifications, Turnbull’s theory
predicts the high melting point and the equally low freezing point
for the nanocrystals. The explanation has to do with the solid-
glass interface between the silica glass and the germanium
nanocrystal. In essence, the silica constrains the formation of a
nucleus of liquid on melting or a solid on freezing.

As I started to write this final paragraph, I looked back through
my early papers and found that I had completely forgotten a very
short discussion I had written on a paper by Richard Oriani. My
discussion appeared in 1956 in the Journal of The
Electrochemical Society and contained a reference to a paper in
1950 by David Turnbull and a colleague. They reported that
small samples of certain metals could be supercooled hundreds
of degrees below their melting points! My brilliant idea was to
make a cell in which one electrode was supercooled liquid and
the other solid metal. I even remarked in the discussion that an
attempt was underway to perform this experiment with
germanium. I didn’t succeed. Maybe I should try again with
nanomaterials? On the other hand, maybe not!

Allen F. Bortrum