Because without diffusion, there would be no significant charge transport.
Charge transport is due to drift, where an electric field imparts force on charges, and diffusion, where density gradients determine charge motion.
Diffusion is one method of charge transport.
Never said otherwise.
Yes, but drift is an insignificant part of a forward-biased diode. You can see how little it is by reverse-biasing a diode, which cuts off the diffusion current, and noting how little the drift current is.
Sorry, that does not compute. A forward biased diode has its charge carriers subjected to an E field from an external source. That is drift, plain and simple. Where do you see any reputable text saying otherwise?
The flow of particles from a higher concentration to a lower concentration, provided the particles are mobile.
Agreed, classic diffusion mechanism.
Here is where you show your misunderstanding. When a P and N slab are put together during manufacture, the excess electrons in the N slab and the excess holes in the P slab will meet and annihilate each other to form a depletion zone. After a time, the uncovered charges will form a barrier voltage that will limit any further diffusive action. Applying a forward-voltage is applied across the diode lower the barrier voltage and allow more diffusion to occur until a new equilibrium is reached. The voltage across the PN junction controls the diffusion and gives the diode its exponential voltage-current curve. What little drift current occurs is insignificant compared to the diffusion current.
Drift is the reason the carriers cross the junction in the first place. They diffuse as well per the density gradient. The barrier voltage does not prevent drift. If the external source providing the E field has sufficient potential the drift continues right through the junction barrier. Take a 12 volt supply and a 10 kohm resistor in series with a p-n diode. When first energized, current exits the battery, through the resistor, into the diode. The original barrier voltage is just 25.7 mV. But carriers cross junction, and increase depletion zone charge and barrier potential. A 0.70 volt barrier is reached and the current is simply (12.0 - 0.70)/10kohm = 1.13 mA.
No, drift current is insignificant to diffusion in a forward-biased diode.
They both are active. W/o drift, how can charge carriers move in a uniform direction?
Why not? The interior of the device is where its operation is determined.
But external fields influence what happens on interior. Otherwise how do charges move? They need external stimulus. Really Ratch, where do you get this stuff?
A BJT is a diffusion device. To say that diffusion does do not matter is just plain wrong.
I never said diffusion does not matter, quite the contrary. It is one important mechanism in understanding semicon behavior, but it is not the only thing going on. Drift and diffusion both must be included in any serious discussion on this subject.
How can a BJT operate without diffusion. How can you explain its exponential voltage-current curve without diffusion?
Again, I have always stated that diffusion is important, but so is drift. You need both to fully understand the topic.
No, pure silicon is not a good conductor and does not readily release its valence electrons. N-type silicon is doped with a pentavalent element that binds to the silicon atoms with four of its valance electrons and holds its fifth electron very loosely. The loose electrons wander around the N-type material and readily move into P-type material, which is looking for another electron to fill its bonding structure. This process is explained extensively in elementary books on semiconductors.
But remember that the Si atoms greatly outnumber the donor/acceptor atoms. Take the n type material. The donor atoms, pentavalent, like arsenic, As, shed the 5th electron with little energy needed, and the key is that NO HOLE is left behind. The As atom bonds with 3 Si atoms in a tetrahedron and 4 electrons are covalent. Hence all 4 bonds are complete and a free electron does not see a hole here.
But remember that thermal vibrations due indeed knock electrons out of Si valence bands as well. It takes more energy to liberate a host atom's electron (Si), vs. a donor atom electron (As). If the 2 atom types had equal density, then the majority of electrons available for conduction would originate from donor atoms. But please remember that the host Si atoms greatly outnumber donor (As) atoms.
Electrons liberated from the Si host atoms are also part of the conduction process. The doping of the Si lattice with donor atoms provides an excess of electrons not the case in intrinsic Si. The electrons liberated from host Si leaves behind a hole, which a free electron may combine with. Not so with As donor atoms. The donors create an excess of electrons. But current conduction consists of electrons from donors, and host, and external source. All participate in conduction.
I said thermal vibrations generate electron-hole pairs in "Si", maybe "Si lattice" is a better choice of words. The thermal energy associated w/ lattice vibrations liberates electrons from both donor and host atoms, but much less energy is needed to liberate electrons from donors. But host atoms are the overwhelming majority.
Yes, what else is new?
No need to be hostile.
When no voltage is applied across the junction, the drift and diffusion currents are the same for a net total of zero. When the junction is forward-biased, the drift current hardly changes and the diffusion current increases exponentially.
With no junction voltage, how can drift happen? No external E field, so how do charges drift?
Wrong! Diffusion is the primary cause of charge motion. Drift is insignificant in a forward-biased diode. That is why they call a BJT a diffusion device. A FET works differently and is not called a diffusion device. Electrons and holes in N and P slabs diffuse when they are first put into contact without any external voltage.
Diffusion is one of two TYPES OF charge motion. Something has to give rise to charge motion, i.e. an E field. Diffusion happens at room temp with a device sitting unconnected on the table. A little diffusion, and no drift.
They both can cause charge motion independently of each other. Who said charge motion is the cause of itself?
Both can exist independently of each other. But w/o an external E field, how can you have drift?
OK, how can you claim that drift current is significant in a forward-biased diode when it is known to be otherwise?
Just how is it "known to be otherwise"? In all my course work, BE, MS, and Ph.D., drift in forward diodes is very significant. I'll post quotes later.
Because I read what is in books written by knowledgeable authors.
So do I, and many others. But taking courses is the best way. On our own we can pick up misconceptions and bad theories that would be quickly corrected in a class and with homework. Self study can get you started, but it leaves large gaps in your data base.
Please do.
I will, promise.
Ratch
