What is new on the experimental front? There are good things and
bad things. Good things first; otherwise there is a risk of losing the
audience.
We are not giving up after perhaps 30 measurements. Every one of
them shows that there is a net charge on plastics A and B. [I am
referring to previously described experiments in which two sheets
of plastic are squeezed between two blocks of aluminum and 1000
volts is applied for couple of minutes. We are now doing this in a
well grounded box, as suggested by Bob Sciamanda; the results are
qualitatively the same.]
Typical charges are several hundreds of nF (the capacitance of our
electrometer turned out to be 2.5 pF, nearly an order of magnitude
less that the "educated guess"). The bad news is that magnitudes
of Q are not more reproducible than before. Thinking about this
we invented an explanation, a model if you wish. The outcome
depends on the "intimacy of the contact". This phrase, if I recall
correctly, was introduced here by Dave Marx, in connection with
friction.
If the amount of the net charge crossing the boundary depends on
a small number of random contacts, for example, only 5 or 10, then
statistical fluctuations are to be expected. An analogy would be a
Geiger counter recording numbers of clicks in short time intervals.
For the mean value of 10 counts per second the individual readings
are often as small as 5 and as large 15. This is "the irreproducibility"
by the factor of 3.
Armed with this hypothesis we decided to see what happens when
the intimacy is improved by putting more and more weight on the
aluminum electrodes. The data are in the table below. Charges are
in the number of divisions, plastic B is in contact with the positive
aluminum block, plastic A is in contact with the negative block.
The negative is grounded.
load (kg) Q on B Q on A comment
0.5 +14 -4
1.0 + 22 - 22 OK
2.0 + 40 - 30 good
4.0 + 37 -3 hmm
5.0 + 60 +15 yes both positive
1.0 -25 +25 yes, a reversal
1.0 + 50 +45 What is going on?
What can I say? I wish somebody else was also playing this
electrostatic game (to compare observations, etc.). We were under
the impression that data could already be taken seriously but now
we are less sure.
Here are some additional speculations. We have three contacts:
plastic_plastic, +AL_plastic and -Al_plastic. We do not know where
net charges are. Electrons can be transferred at three different places
independently, sometimes more here and sometimes more there.
This is likely to be controlled by Q.M. (gaps in the band structure of
energy levels, acceptors, donors, etc., as in semiconductors. In fact,
the only difference between semiconductors and dielectrics, from
that perspective, is the width of the forbidden energy gap (about
10 eV versus about 1 eV).
The impurity of surfaces may play an important role. Our samples
are "dirty"; we discharge pieces before each experiment (verifying
their neutrality with the electometer) by squizzing them between
two hands and touching the ground. Sometimes this has to be done
two or three times. Who know how many levels are created by
dorors and acceptors transfered to surfaces?
Keep in mind, however, that not a single experiment, was in
agreement with what would be expected from the textbook model
of dielectrics. That model leads to a definite prediction, net
macroscopic charges must be zero on each piece, clean or dirty. This
contradiction alone is a good reason for experimenting. How can a
science teacher, a professional modeler, tolerate this?
Physics is an experimental science.
Ludwik Kowalski