 |
 |
Negatively Charged Plasma
Name: John
Status: student
Age: N/A
Location: N/A
Country: N/A
Date: N/A
Question:
Can plasma be negatively charged? If it can be negatively
charged, what are the usual residence times before the electrons are
stripped away?
Replies:
John-
We need to distinguish between "plasma" the collective term,
and "ions" or "particles", the individual items.
A hole in a forest is different than a hole in a tree...
A plasma is a gas that has at least some charged particles amongst it.
Usually there are some neutral atoms and/or molecules, which we call
"neutrals" or "neutral species",
and some that have gained or lost an electron, which we call "ions",
and also a few "free electrons" free in space, not presently captured by
any atom.
The relative percentages of each can vary widely depending on the
situation.
But usually the net charge of the whole plasma is not very far from
neutral,
because unbalanced space-charge has a rather high energy
and the charged particles can travel rather freely.
Especially the free electrons.
All atoms or molecules are capable of being positive ions, losing one or
more electrons.
Many molecules which lose an electron are prone to break up shortly
thereafter.
An H2 molecule losing two electrons has no more electrons, so cannot be a
molecule,
and is immediately by definition two H+ ions.
Most atoms and molecules are capable of being negative ions,
of weakly or strongly capturing an extra electron that drifts nearby.
Some atoms are notably incapable of this;
those elements have no "electron affinity" number in the reference books,
and are said to have a "negative electron affinity".
I think Titanium is an example.
The electron affinity number is the amount of energy released when a free
electron falls onto a neutral atom and is captured.
It is a positive energy release, so the residence time of an electron on an
atom with high electron affinity
(Fluorine, Chlorine, Bromine are strong) can be indefinite.
But you have to ask what rate of provocations exist in your particular
plasma.
Since the whole plasma is nearly neutral, positive ions must be around
somewhere,
and the electron can easily be stripped away by passing close to one of
them.
Transference of a captured electron of a passing neutral of the same
element is surprisingly frequent.
And even a mild impact between a negative ion and a free electron can
re-free the captured electron,
if the impact energy is higher than the electron affinity energy.
The actual residence times range over many orders of magnitude.
Scientists collect numbers called "cross-sections" to help predict them.
A neutral-transfer cross-section can be an area tens of atoms wide, (i.e.,
snatching from a distance)
and the typical speed of an atom is about the speed of sound, which is a
function of temperature and particle mass.
The kinetic energy of the electrons is often different, faster, than that
of massive species (atoms, ions, molecules),
so the term "electron temperature" gets used in opposition to the 'thermal
temperature".
The bulk of a plasma can have some voltage with respect to a conductive
surface touching it.
These voltages are proportional to some small percentage charge imbalances
in the plasma.
In common IC-industry sputter-deposition machines the plasma between two
metal plates
is typically hundreds or thousands of volts more positive than the negative
plate,
but only tens of volts more negative than the positive plate.
Very asymmetrical.
This is because negative free electrons in the plasma are tens of thousands
of times lighter than the positive ions,
and if the plasma is negative it takes little voltage difference to make a
big current of electrons flowing onto the metal.
On the other side, electrons cannot just jump out of the metal plate into a
vacuum (they are trapped in the solid),
so the current at that interface consists partly of positive ions falling
onto the negative plate,
and partly of secondary electrons freed from the metal plate by the impact,
then travelling the opposite direction.
The positive ions fall with high energy, so some atoms from the metal are
kicked off and may coat a nearby IC chip with that metal.
On the other hand, a plasma in Bromine vapor (Br2) can be dominated by positive
and negative ions
rather than positive ions and free electrons...
Electronegativity of course helps electron affinity, but larger particle
size helps too.
SF6 molecules are very good at it. Buckyballs (C60) are probably pretty
good too.
These will have the longest practical single-particle residence times.
In the larger picture, there is no sharp boundary between
the electron affinity of a large molecule and
the static charge-holding ability of a metal particle of any size.
The residence time of an electron on a macroscopic metal object, of course,
can be infinite.
Usually for a plasma, one models a steady-state situation.
Think species population densities per cm3, volts, charge and time, current
flows per cm2....
Sorry, I am out of touch with specific numbers,
but if you find and read a few articles on "sputtering" or "ion milling"
you will see some right away.
Jim Swenson
Click here to return to the Physics Archives
| |
Update: June 2012
|
|
