I How do point charges in a conductor move and stop?

[feynman1]

Let's speak in the classical context (non quantum). We assume that point charges move in a conductor following Newtonian mechanics. How do point charges move along the boundary of the conductor and how do they stop (equilibrium) in the end?

 

[f95toli]

Non-quantum charge transport in a conductor is not a thing; it is an inherently quantum mechanical process.
We can of course use semi-classical models such as the Drude model, but you can't (and shouldn't) use these to think about individual charges.

It is possible to create systems where you have "true" single electron transport (usually in semiconducting systems); but this is an exception and if you look into details about what is actually moving you will find that it is a wave packet; not something which can be described by Newtonian mechanics.

 

[Dale]

you can't (and shouldn't) use these to think about individual charges

Well said. Classically, the sources in Maxwell’s equations are $\rho$ and $\vec J$ which are continuous. As far as I know, all of the paradoxes and inconsistencies in classical EM are based on using Dirac delta functions as the sources.

 

[feynman1]

We can of course use semi-classical models such as the Drude model, but you can't (and shouldn't) use these to think about individual charges.

Can we only think of them as an ideal gas, thus a continuous medium? What result will go wrong when thinking about individual charges?

 

[feynman1]

all of the paradoxes and inconsistencies in classical EM are based on using Dirac delta functions as the sources.

What's the problem with using direct Delta function at its source?

 

[Dale]

What's the problem with using direct Delta function at its source?

There are several. The most obvious one is that the field energy is infinite around a point charge. This would lead to point charges being infinitely massive, an infinite amount of energy release as two opposite charges collide, etc.

Another common issue is the radiation self-reaction. This can lead to runaway acceleration.

I am sure there are others.

 

[Herman Trivilino]

How do point charges move along the boundary of the conductor and how do they stop (equilibrium) in the end?

They respond to an electric field.

 

[feynman1]

They respond to an electric field.

If there is only an electrostatic field, particles can't stop, can they

 

[Delta2]

If i understand correctly you are asking why the free electrons are not being ejected from the conductor when an external electrostatic field is applied, instead they seem to stop at the surface. This cannot be explained by classical physics, from what I know according to quantum physics the free electrons must be accelerated to a very high velocity in order to be ejected from the conductor.

 

[feynman1]

If i understand correctly you are asking why the free electrons are not being ejected from the conductor when an external electrostatic field is applied, instead they seem to stop at the surface. This cannot be explained by classical physics, from what I know according to quantum physics the free electrons must be accelerated to a very high velocity in order to be ejected from the conductor.

No, I wasn't asking this question. 2 questions below. I wonder what happens when charges hit the boundary. Do they bounce several times on the boundary or do they just stick to the boundary upon colliding it? How do charges move along the boundary and finally grind to a halt?

 

[Delta2]

Do they bounce several times on the boundary or do they just stick to the boundary upon colliding it

I think you have some sort of mechanical model in your mind (the electrons being some sort of balls that are bouncing against a solid wall-the boundary) which i don't think is entirely correct. But maybe a more knowledgeable member than me should answer here, maybe @vanhees71 .

 

[feynman1]

I think you have some sort of mechanical model in your mind (the electrons being some sort of balls that are bouncing against a solid wall-the boundary) which i don't think is entirely correct. But maybe a more knowledgeable member than me should answer here, maybe @vanhees71 .

Exactly. I need (wish) a mechanical model with electrons thought of as balls bouncing around, as suggested by the original post (classical).

 

[f95toli]

Exactly. I need (wish) a mechanical model with electrons thought of as balls bouncing around, as suggested by the original post (classical).

As was stated above: This is simply not possible, because electrons are NOT "small balls" meaning a "mechanical" model simply won't work.
For the electrons close to a surface the wiki on Friedel oscillations might be relevant.

The simplest model treats the free electrons in a Fermi gas with a plane wave-like behaviour
See wiki
https://en.wikipedia.org/wiki/Friedel_oscillations

The reason this sort-of works is because it takes the wave-like behaviour into account which means you can get interference.

 

[Vanadium 50]

Exactly. I need (wish) a mechanical model with electrons thought of as balls bouncing around, as suggested by the original post (classical).

Why would you want such a thing?

You have Classical E&M, which works well in its domain of validity. Now you want to take quantum objects, electrons, and treat them like classical billiard balls. Why would you think this would be helpful?

In 1900, there was something called the Drude model, which attempted to do this. It got Ohm's Law as an output (sort of) and pretty much got everything else wrong.

 

[Dale]

I wonder what happens when charges hit the boundary. Do they bounce several times on the boundary or do they just stick to the boundary upon colliding it? How do charges move along the boundary and finally grind to a halt?

Classically, on a conductor at DC there is a surface charge density and a volume current density. Both of these are continuous charge distributions classically, not point charges, for the reasons I pointed out above.

I wonder what happens when charges hit the boundary. Do they bounce several times on the boundary or do they just stick to the boundary upon colliding it? How do charges move along the boundary and finally grind to a halt?

The surface charge density produces an internal field such that the current density near the surface has no component perpendicular to the surface. Thus there is no “hitting the boundary” etc.

 

[vanhees71]

The classical charged point particle is a contradiction in itself. It simply doesn't make fully sense. After subtracting an infinite mass from the also infinite bare mass you get a 3rd-order equation of motion, the Lorentz-Abraham Dirac equation, which has big problems being acausal and/or exploding ("runaway and preacceleration solutions") of you make a perturbative approximation, leading to the Landau-Lifshitz equation, which has not the deficits of the LAD equation but violates (to some extent) energy conservation.

With quantum (field) theory you are much better! There you can, after mass, charge, and wave function renormalization, calculate things at any order of perturbation theory and for some quantities like (g-2) of the electron or the Lamb shift of the hydrogen atom you get a description which is correct for 12 or more significant digits.

On top in the many-body version you can also describe material properties correctly. The quantum (statistical) version of the above mentioned Drude model for electric (as well as heat) conductivity for metals was among the first quantum many body systems which have been solved in the early days of quantum theory (Sommerfeld and Bethe wrote an early textbook on it). This was the beginning of the huge and tremendously successful field of condensed-matter theory, and it's all quantum. There's no way to understand the behavior of matter (including the usual matter surrounding us in everyday life) within any classical model. It already starts with the observation that there is stable matter at all. With classical mechanics you can't even understand the most simple atom, the hydrogen atom!

 

[Herman Trivilino]

If there is only an electrostatic field, particles can't stop, can they

If the direction of the electric field is opposite to the direction of motion.

 

[feynman1]

The surface charge density produces an internal field such that the current density near the surface has no component perpendicular to the surface. Thus there is no “hitting the boundary” etc.

Isn't the work function rather than the surface charge itself that inhibits current density from going perpendicular to the surface?

 

[feynman1]

If the direction of the electric field is opposite to the direction of motion.

Even that doesn't mean charges will stop. But they can still bounce around.

 

[Dale]

Isn't the work function rather than the surface charge itself that inhibits current density from going perpendicular to the surface?

No, it is the surface charge. The work function keeps the surface charge from leaving, but it only affects the surface charge density. The surface charges, in contrast, affect the current density throughout the conductor.

 

[feynman1]

Classical physics has a lot of issues indeed. But all I want is just a classical model describing the motion of point charges in a crude way much like the classical model of the atom with electrons orbiting the nucleus.

 

[feynman1]

The surface charge density produces an internal field such that the current density near the surface has no component perpendicular to the surface. Thus there is no “hitting the boundary” etc.

Are you speaking under the context of electrostatics or electrodynamics, with current or without?

 

[Dale]

Are you speaking under the context of electrostatics or electrodynamics, with current or without?

I started that post with “on a conductor at DC”. I didn’t want to deal with the messy details of electromagnetic waves and skin depth etc. But of course electrostatics is just DC with 0 current density, so what I said applies there too.

 

[Delta2]

The surface charge density produces an internal field such that the current density near the surface has no component perpendicular to the surface. Thus there is no “hitting the boundary” etc

Since the surface charge density interacts with the free electrons, so free electrons apply forces to the surface charges, what keeps the surface charges at their place, that is at the surface. Why they do not move to the outside of the conductor or to the interior?

 

[feynman1]

Since the surface charge density interacts with the free electrons, so free electrons apply forces to the surface charges, what keeps the surface charges at their place, that is at the surface. Why they do not move to the outside of the conductor or to the interior?

I know why they don't move outside of the conductor. That is due to a work function. But I don't know why they can't move back to the interior.

 

[Dale]

Since the surface charge density interacts with the free electrons, so free electrons apply forces to the surface charges, what keeps the surface charges at their place, that is at the surface. Why they do not move to the outside of the conductor or to the interior?

The bulk of the wire is neutral so there is no force on the surface charge density from the current density.

 

[sophiecentaur]

No, it is the surface charge. The work function keeps the surface charge from leaving, but it only affects the surface charge density. The surface charges, in contrast, affect the current density throughout the conductor.

I know why they don't move outside of the conductor. That is due to a work function. But I don't know why they can't move back to the interior.

It may be a bit confusing to try to use Energy and Field in the same argument. You can do it all with Energy by saying the energy to move from place to place within the conductor is a lot lower than the energy to remove charge from it. Water moves easily from place to place in a bowl but it needs a lot of energy to climb up the sides. You can tilt the bowl slightly to displace the water so it can flow into another bowl through a hole just above the surface, which is like a connection to another conductor (If you really want a mechanical approach / analogy.)

 

[feynman1]

It may be a bit confusing to try to use Energy and Field in the same argument. You can do it all with Energy by saying the energy to move from place to place within the conductor is a lot lower than the energy to remove charge from it. Water moves easily from place to place in a bowl but it needs a lot of energy to climb up the sides. You can tilt the bowl slightly to displace the water so it can flow into another bowl through a hole just above the surface, which is like a connection to another conductor (If you really want a mechanical approach / analogy.)

who mentioned energy?

 

[sophiecentaur]

who mentioned energy?

Work Function?

 

[Lord Jestocost]

How do point charges move along the boundary of the conductor and how do they stop (equilibrium) in the end?

Equilibrium doesn’t mean that mobile electrons ever “stop” in a conductor in a literal sense. In case a dynamical equilibrium is established – the electrochemical potential of the electrons becomes constant throughout the conductor –, currents in one direction are – so to speak –“cancelled” by currents in the opposite direction.

 

[feynman1]

Work Function?

that's not a kind of energy

 

[f95toli]

that's not a kind of energy

Yes it is, it is the minimum amount of energy needed to move an electron from a solid to the vacuum,

 

[feynman1]

Yes it is, it is the minimum amount of energy needed to move an electron from a solid to the vacuum,

Yes, sorry I didn't mean to deny it being an energy. I just meant that wasn't 'work' in the usual sense of introductory physics.

 

[sophiecentaur]

Yes, sorry I didn't mean to deny it being an energy. I just meant that wasn't 'work' in the usual sense of introductory physics.

The clue is surely in the Units involved, rather than a "sense". In the case of Work Function, the name itself associates the Kinetic Energy and the Photon Energy so it's actually quite hard to be confused in the context of Photoelectricity. What we are discussing is not exactly introductory Physics.

 

[feynman1]

What kind of model are those point charges on physics textbooks? Are they valence electrons or ions? If those point charges can move, can they not be ions?

 

[sophiecentaur]

Only in a fluid.

 

[feynman1]

Only in a fluid.

can't get what you mean

 

[sophiecentaur]

can't get what you mean

In a solid, the ions are locked in place; in a fluid they can move.

 

[feynman1]

In a solid, the ions are locked in place; in a fluid they can move.

Very well! In a physics textbook with point charges (+ and -), should we always regard + to be ions and - to be electrons? Then if the context is a solid, + point charges can't move, but there's no such restriction on physics textbooks, why?

 

[sophiecentaur]

Very well! In a physics textbook with point charges (+ and -), should we always regard + to be ions and - to be electrons? Then if the context is a solid, + point charges can't move, but there's no such restriction on physics textbooks, why?

The Physics books that I have read all tell us that charge is carried in metals (and other solids) by negative charges (electrons). Can you find anywhere that the "restriction" is relaxed? By definition, if the positive ion cores in solids could move then why wouldn't the 'solid' flow and be a liquid? Or are you just trying to wind me up?

PS If you consider the +holes in a semiconductor to be charge carriers, they only exist because of the motion of electrons from atom and the + ions do not actually move; holes are just a way of thinking about it.

 

[feynman1]

The Physics books that I have read all tell us that charge is carried in metals (and other solids) by negative charges (electrons). Can you find anywhere that the "restriction" is relaxed? By definition, if the positive ion cores in solids could move then why wouldn't the 'solid' flow and be a liquid? Or are you just trying to wind me up?

A pair of + and + charges could be initially in the interior of a metal. Afterwards both charges will be repelled and move in the metal. If this appears in a textbook, that'll look quite normal and none doubts about whether + charges are supposed to move?

 

[sophiecentaur]

If this appears in a textbook,

I challenge you to find this idea in a textbook. Ions (+ and -) are held in place in a solid and that is what defines the solid phase. If they could move around then would the solid be solid? In a solution (e.g. water) the theory describes the simple situation of a Salt (NaCl) Solution as mobile Na⁺ and Cl^- having been separated from each other (in their solid state).

Taking this one stage further, it does happen that impurity ions can slowly make their way below the surface of a 'pure' solid material but, unless the substance is near its melting point, the time scale is many years. Alloys can form on the surface of metals when a melted solder (+ metal ions) can penetrate the faces of two copper wires and form a bond. But I would say that none of this is regarded as a flow of electric current.

 

[feynman1]

I challenge you to find this idea in a textbook.

Do you mean this pic in a textbook can only be correct if it implies that this happens in a fluid or free space?

 

[Lord Jestocost]

This illustration depicts the magnetic force when considering the "convential current" in a conductor.
http://hyperphysics.phy-astr.gsu.edu/hbase/magnetic/magfor.html

 

[vanhees71]

I read the picture as describing the magnetic part of the Lorentz force on a positively charged point particle. That holds in free space as well as in the medium. In the latter case, of course, there are other forces by interaction with the medium, which can be described macroscopically by friction. The relevant transport coefficient is electric conductivity.

In a usual metallic conductor, what's moving are the conduction electrons which are quasifreely moving quasiparticles.

In semiconductors the charge carriers can be either negatively or positively charged quasiparticles, depending on the doping of the material. The positively charged quasiparticles are microscopically "holes", i.e., unoccupied electron states.

Which electric charge the relevant quasiparticles make up a current within a solid can be measured by making use of the Hall effect, which is based on the above pictured force of a charged particle in a magnetic field.

 

[sophiecentaur]

Learning by telling never works as well as learning by reading and listening.

 

[feynman1]

Does electrostatic equilibrium in a metal mean that electrons don't move at all or they do move but their electric effects like electric field remain constant in some sense?

 

[feynman1]

From non Newtonian mechanics, what will happen to electrons when they hit a conductor wall?

 

[Vanadium 50]

what will happen to electrons when they hit a conductor wall

You are (still) describing a classical path. Electrons behave quantum-mechanically, not classically.

 

[feynman1]

You are (still) describing a classical path. Electrons behave quantum-mechanically, not classically.

What's the correct way of describing this without using 'hit'?