Non-contact friction can be artificially enhanced. Usually for two
bodies in relative motion to feel friction the respective surface atoms
have to be in contact. There is a type of friction, however, which can
act between two surfaces not actually in contact. This dilute friction
is attributed to the van der Waals force, a common but weak attractive
force which arises when an atom or molecule spontaneously develops a
dipole moment (that is, although it is neutral, a small region of net
negative charge can develop, offset slightly from a comparable positive
region) owing to a thermal fluctuation (related to the random motion
of the electrons and ions) or a quantum fluctuation (the very positions
of the particles varies from moment to moment owing to the uncertainty
relations built into quantum reality). This short-lived polarity can
in turn induce a dipole moment in a neighboring atom or molecule, some
distance away. A new study of van der Waals friction by Alexander
Volokitin and Bo Persson at the Institut fur Festkorperforschung
(Julich, Germany) accounts for recent odd friction experiments conducted
with STM probes. The theory holds that van der Waals friction can be
greatly enhanced (by up to a factor of ten million at a separation of
10 angstroms in comparison with the case of good conductors with clean
surfaces) by adsorbing certain molecules onto one or both of the surfaces.
This increases the resonant electromagnetic force (which can be viewed
as the tunneling of photons) between the objects, especially if they
are made of the same material. The adsorbate atoms can be thought of
as tiny antennas, one acting as an emitter and one as a receiver; when
the two antennas are in tune the electromagnetic interaction between
them will be greatly enhanced (see
figure).
A better understanding of this kind of non-contact friction
will, at the fundamental level, help physicists to study the quantum
behavior of atoms at surfaces and, at the level of applications, to
prepare "brakes" for micromachines where large friction is
not needed. (Volokitin
and Persson, Physical Review Letters, 5 September 2003)