🚀 go-pugleaf

Hi.  I am looking for Gary Knight who worked at MCC a few years ago.
He was interested in hearing about proposals to do something in the
way of a nanotech business, and I might have found one.  You can 
pass the encluded item around privately to as many as you like, but
I don't want it "published" yet (for patent reasons.)  Keith
----
Proposed Method for making STM Tips.

I recently visited a local company which makes ion milling equipment 
for doing engineering mods on ICs.  The ion source is a gallium-wet 
needle which they operate at a modest positive voltage.  Electrostatic 
forces acting against the liquid metal surface tension shape up the 
tip to a one atom point--from which ions leave.  Slick gadget.  There 
may be additional markets for this thing in the interface between 
current micro and the nanotech world.  I can send a contact to anyone 
with a need. 

But that is not what this posting is about.

A friend of mine (Mark Voelker) has been complaining about the 
difficulty of making good STM tips for quite a while.  (It is quite a 
black art).  A week or so after I learned about the ion source, it 
occurred to me that *any* liquid metal under a strong field should 
take on the same shape--and might keep the shape if you cooled it to 
the solid state.  Gallium is *way* too soft even frozen to be of 
interest as an STM tip, but a modest pulse of energy from a laser beam 
or a pulse of electrical energy in the form of a spark or a burst of 
electrons should melt the end of a tungsten or platinum STM tip.  If 
the tip cooled with enough voltage to shape it up to an ion emitter, 
it should end in a single atom. 

According to Mark, typical STM tips are 20 mil of (mostly) platinum 
alloy wire.  Converting to metric, that would be 20/1000*2.54 or about 
.05 cm.  Melting the end of the tip would involve a volume of roughly 
(.05) exp 3, around 1*10 exp -4 cm cube.  This times a density of 
about 20 gives a mass of about 2/10000 of a gram.  The delta T 
involved in round number is 2k degrees, the specific heat of Pt in the 
range of interest is about .04, and the heat of fusion is about 30.  
So, 2k*2/10000*.04 is .016 cal plus 30*2/10000 or about .02 cal, or .1 
joule.  I.e., in the range of tenth of a watt-second.  (Assuming I did 
not drop a decimal or two.) 

Since the energy stored in a capacitor in watt-seconds is 1/2 CV exp 2,
0.1 watt-second would take a modest 20 uF charged to 100 volts. 

The tip should be heated fast compared to its cooling time.  The area 
of this speck (considered as a cube) is 6*.0025, or about .01 square 
cm, or 1 * 10 exp -6 square meters.  A square meter at 2k deg absolute 
would radiate (2k) exp 4 * 5.67 * 10 exp -8 watts per square meter, or 
about 1 * 10 exp 6 watts.  A cooling time of a second is implied.  
(Ignoring conduction.)  This is not out of line with the time it takes 
ordinary incandescent lamps of similar wire size and temperature to 
cool, implying that a heating time in the low tens of milliseconds 
would be ok.  With the time set at 1/100 second, and the voltage at 
100 volts, the required current is 100 mA.  

Tip heating could be accomplished by laser, a spark in an inert 
atmosphere, or electrons in a vacuum.  The current is well within the 
output of medium sized vacuum tube cathodes.  After the heating pulse, 
the tip would have to be maintained at a high enough positive voltage 
to emit ions until it cooled. 

For all I know, this may be a well known method for forming tips, but 
if so, Mark was not aware of it being used and there was nobody at the 
recent nanotech conference (not even the Japanese representative) who 
knew of it being used.  I am posting this note to get a widespread 
time stamp of when this method was conceived, and a request to the 
community to let me know if this is already in use.  It may be 
patentable, so such rights are reserved. 

Keith Henson 11/14/92