126
The Body Electric
Our confusion also helped the scientific establishment accept the "triv-
ial" electrical stimulation of bone by considering it something unique to
bone. The relationship between our experiment and regeneration proper
was lost.
The Inner Electronics of Bone
Charlie Bachman and I decided to investigate the electrical properties of
bone in more detail and try to figure out how Wolff's law worked. We
put together a hypothesis based on my experiments with Bassett. We
postulated that the bone matrix was a biphasic (two-part) semiconduc-
tor. That is, either apatite or collagen was an N-type semiconductor; the
other, a P type. Their connected surfaces would thus form a natural PN
junction diode that would rectify any current in the bone. We further
theorized that only one of the materials was piezoelectric. On the com-
pressed side of a stressed bone, we expected the positive pulses to be
filtered out, leaving a negative signal to stimulate periosteal cells to
grow new bone.
We made several pairs of sample blocks, cut side by side from pieces
of bone removed from patients for medical reasons. From one member of
each pair we chemically removed the apatite. The other we treated with
a compound that dissolved the collagen. The resulting pure collagen was
yellowish and slightly rubbery, and the apatite pure white and brittle,
but otherwise both blocks still looked like bone. Our first step was to
test bone's component materials separately for semiconduction and
piezoelectricity. Collagen turned out to be an N-type semiconductor and
apatite a P type.
Then we tested our samples for piezoelectricity in the same way that
Bassett and I had previously tested whole bone. We expected that apa-
tite would be the only one to show an effect, since it was a crystal.
However, collagen turned out to be a piezoelectric generator, while apa-
tite was not. We now had the makings of a PN junction—two semicon-
ductors, one an N type, the other a P type, joined together in a highly
organized fashion.
Now came the crucial part of our hypothesis. We had to figure out a
way to test for rectification at the PN junction. It was an important
crossroads.
Here we rn up against what's known in the trade as a technical
problem. To test for rectification we had to
put one electrode
on
the
collagen and one on
the
apatite as they appeared in whole bone.
Unfor-