SPI-MODULE ™ Sputter/Carbon Coater System
III. General Comments About Sputter Coaters
The design of the cathodes used for the SPI Sputter Coaters has changed
continuously with the advent of newer technology. The original design in
the mid-1970s was a diode geometry, which was quite acceptable, except
considerable sample heating occurred. Sample heating came from three
sources:
a) Electrons bombarding the surface being coated.
b) Absorption of kinetic energy of sputtered atoms.
c) Radiant heat from the glow of the plasma just underneath the cathode.
The triode design was a definite improvement and was featured in the
original SPI Sputter™ Coater. The magnetic field existing during
operation tended to deflect electrons away from the sample. One of the
three sources of sample heating was thereby eliminated. Since SPI was
one of the very first manufacturers to take advantage of this then newest
design, the SPI units started earning their reputation for depositing
under conditions of "cool" sputtering.
The original SPI Sputter "magnetic head" design, which also incorporated
the main "triode" features, allowed for very rapid coating (10-15 seconds
vs. 45 to 4 to 5 minutes for the older designs). For the first time, very
heat
sensitive samples could be coated (such as a broken piece of chocolate)
because radiant heat effects were essentially "eliminated"). However,
there was a slight penalty paid and that was a slightly larger grain size
(however, it was so slight that unless operation of the SEM was going to
be over 60,000x, one would never detect the difference!).
Today, the SPI Module Sputter Coaters represent the very latest state of
the art design. While still employing the best features of the "magnetic
head" design, the cathode design has reverted more to a diode geometry.
However, because of the unique SPI-designed spatial arrangement of the
magnetic field within this new diode cathode, one can still obtain the
rapid deposition rates (15-20 seconds in this instance) but with only a
slight increase in grain size, and one that would be observed only at
magnifications higher than "60,000x".
One often times hears about "water cooled" cathodes. There is no
demonstrated evidence that water cooled cathodes have any efficacy for
reducing specimen damage from heating and this is because the heating
effects are from the "glow" of the plasma, not from any radiant heat
actually emanating from the cathode. Because of the extremely fast
deposition times, at least that heating due to electron bombardment
reduced essentially to zero.
But the important point is that to whatever degree there are still some
heating effects, there is only marginal evidence that water- cooled
stages have any such efficacy for reducing such effects anyhow. We are
philosophically and adamantly opposed to offering any "gimmick"
attachment that accomplishes nothing for the researcher, yet greatly
increases the cost of the sputter coater.
The SPI-MODULE sputter/carbon coater does have built-in, "standard", the
provision for water cooling the sample stage in the event the user
does feel it desirable. However, this is done in an innovative way that
enables it to be done without increasing in any noticeable way, the
costs of production. So while we ourselves question whether any water
cooled stage offers particular real advantage during the preparation of heat
sensitive samples for SEM, to whatever degree it does offer a benefit in
any other commercially available coater on the market today, one would
realize the same identical benefit via the SPI-MODULE coater system.
After all, the same laws of physics apply no matter what unit is being used!
There is some marginal evidence that argon (or other inert gasses) can
have beneficial effects when present during sputtering, and one can bleed
in Ar (or any other inert gas) into the SPI-MODULE Sputter/Carbon Coater
units
as easily as in any other currently and commercially available sputter
coater. However, generally speaking, for most users, the effect of argon
is far less than expected. As will be explained later, the use of an
inert gas is far less important in the SPI-MODULE system anyhow,
because the sputtering rate , inherently, is so much faster.
All low voltage sputter coater systems, irrespective of manufacturer, as
already pointed out, obviously operate under the same laws of physics.
Since the operating
parameters can be varied widely on all commercially available systems
(vacuum, sample to target distance, ionization current), they will all,
more or less, perform the same way. Similarly, sputtering systems all
"evaporate" carbon but the resulting coating is of such poor quality (it
does not even seem to be conductive), one can not really apply carbon by
simply putting in a "carbon cathode". Indeed that is why carbon must be
deposited via an altogether different process, that is, by a "flash
evaporation" in a specially designed
carbon coater module.
The carbon source can be either
carbon "fiber" or
carbon rods, a
carbon fiber head coming as "standard" with an SPI-MODULE Carbon Coater
system or module and an optional
carbon rod head
being available for use
with carbon rods
(rods give a slightly higher quality carbon coating than fiber, again,
this is a function of the physics of the situation, not the result of one
unit being "better" than another).
Note: We would not recommend anyone's carbon coater system for making
carbon films for TEM. One really does need the higher vacuum normally
found in a vacuum evaporator, using a diffusion pumped (or better)
system. This again is as true of the SPI carbon coater system as it is
for anyone else's carbon coater system using a mechanical or rotary vane
pumped system.
In recent years, certain manufacturers offering competing sputter and
carbon coaters have introduced several "features" which, in our opinion,
for the typical user, add nothing to the capability of the unit and
unnecessarily raises its
cost. More significantly, from the manufacturer's standpoint, they serve
as exclusionary barriers for the submission of "responsive" bids.
Several examples of such "features" are as follows:
a) Stainless steel shutters between specimen and cathode to prevent
redepositing of etched material. We have never seen published a
side-by-side comparison of a sample etched with vs. without such a
shutter that, in fact, shows a difference. All such attempts to
demonstrate the efficacy of a shutter, to our knowledge, have been
unsuccessful.
And as if this was not enough reason, so few real systems actually
can be etched by sputter etching anyhow, that for the typical
laboratory it would be quite extraordinary for this situation to ever
even arise. Of course, the need to etch does arise, but far more often
than not, the species to be etched, can not be etched by sputter etching
and what is really needed of a reactive plasma etcher such as the
SPI Plasma Prep II system.
b) Automatic argon flush has no demonstrated benefits and even
sputtering with argon as an inert atmosphere has limited benefits. In
any case, the percentage of users routinely use
argon for any reason is extremely small.
c) Argon to be admitted above specimen and pumped out below. Using
argon under any geometry, at best, offers only marginal benefits to those
using the SPI system because with the
unique SPI magnetic design for the cathode, the sputtering times are already
very short. This particular modification adds significantly to cost
while offering virtually nothing beneficial in return, at least nothing
that has been published in peer reviewed scientific literature.
d) Power supplies in excess of 1300 volts have no benefit for the
preparation of specimens for SEM examination. Some competing units offer
power supplies to 2800 volts (or more), without any benefit to the end
user. There is an exception to this statement and that would be for the
sputtering of certain other elements such a chromium. But we are now
talking about systems costing 5-10 times more than the SPI units and in
any case, unless one has a FEG or other high performance SEM, they would
not realize the benefit of chromium unless operating on the order of
"100,000x" or higher.
e) Rotating and/or tilting tables are not needed in a sputter coater
(unlike a vacuum evaporator) because the operation is in a "soft"
mechanical pump (rotary vane) vacuum. The species being deposited do not
travel in straight lines because of their constant collisions with
gaseous molecules in the chamber. This phenomenon makes possible the
continuous coating of even convoluted surfaces. A rotating/tilting
table is not needed in the SPI-MODULE Sputter Coater. One can always
install one, but it adds appreciably to the selling price while producing
no tangible benefit to the user.
To Ask a Question or Make a Comment
To Place an Order or Request a Quote
Return to:
Sunday July 06, 2008
© Copyright 1997 - 2008. By Structure Probe, Inc.
Contacting Structure Probe, Inc.
All rights reserved.
All trademarks and trade names are the property of their respective owners.
Privacy Policy
Worldwide Distributors, Representatives, and Agents
