I can add a little here, Harry. As you say, the energy of the photon is
closely related to the bandgap, which depends on the composition. Thats
why red LEDs drop about 1.7 V, while blue is up near 3.5 or 4 V.
AIUI, when the junction is forward biased, the electrons and holes
recombine and release the energy they gained falling through the junction
as photons, just as you say.
I don't know of any semiconductors available in 1927 that would have
enough bandgap (>1.7V) to emit visible light. IIRC, you were using Ga
with In doping. The N saturated substrate doping might have made a GaInN
or GaInAs junction, which could explain your observation.
For white (or any color) light, you can, of course, simply combine red,
green and blue LEDs in a single package, and those are available, but not
cheap.
The trick behind the common low cost white LEDs is to use a single blue
LED to excite red and green fluorescent materials to get the primary
colors. For me, flashlight bulbs are history.
It would be nice to see more cordial, informative posts.
Bill Ward- Hide quoted text -
- Show quoted text -
Bill, it is refreshing to find someone who actually knows something
about physics here in sci.physics.
You posted:
"I don't know of any semiconductors available in 1927 that would have
enough bandgap (>1.7V) to emit visible light. IIRC, you were using
Ga
with In doping. The N saturated substrate doping might have made a
GaInN
or GaInAs junction, which could explain your observation."
Just some details for whatever value that they might add to the
thread....
Our work began using pulled crystals of n-doped germanium (presumably
the doping agent employed was arsenic). [The assingment required to
have monthly blood tests conducted for arsenic, even though we never
directly handled arsenic reagents.] The junctions were formed by
flowing molten mixes of indium and gallium over a thin slice taken
from the mother crystal, conducted in a hydrogen atmosphere.
It never crossed my mind at that time that this could result in a
GalnAs junction, particularly since we were working with degenerately
doped germanium crystals. [For the benefit of non-physicists, the
definition of "degerate doping" refers to the fact that the doping
agent in the semiconductor medium is (in this case arsenic) is much
greater than that found in conventional transistor components. In
simple terms, it means that you have dumped too much salt or sugar
into the cookie mix,]
Just to set the stage, Bill in 1959 it was well known how to
manufacture a transistor, but the then ongoing research focused on the
"what if" things. One of these avenues of research was since we know
pretty well how intrinsic semiconductors perform when slightly doped,
what would happen if we degenerately dope them. The tunnel diode was
an early result from this avenue of research, and since I moved on to
other areas since, I'm not sure what other results followed, although
I strongly suspect that the LED was one.
Bill, it is interesting how careers can change over the years. I
originally started working in particle accelerator design, not
semiconductors. Later moved to the design of earth satellites, then
railway signaling, and finally radar systems and retirement. Life
offers incredible experiences if you give it a chance and live long
and healthy enough. In my case, and I don't suggest this for
everyone, two shots of Jim Beam and several beers each night helps,
along with having a good wife for 50 years and 3 now adult kids.
On a far more serious note, I would seriously urge every young person
that reads this to attend college and complete a degree. That degree
assures potental employers that you have at least studied the material
present to you, passed exams on the same, and helps to show that you
now sit on the shoulders of those who went before you, and hopefully
see farther.
Harry C.
