Register to reply

Is it possible to have a magnetic diode?

by haael
Tags: magnetism, semiconductors
Share this thread:
Oct14-12, 09:34 AM
P: 438
I wonder: do there exist "magnetic semiconductors" and magnetic diodes and transistors? Normal electrical semiconductors rely on the fact that there are positive and negative charge carriers. Magnetism does not have monopole charges, so it will not be possible to construct an isolated region of a magnetic semiconductor, but maybe it is possible to have an indivisible magnetic semiconductor junction, in the same sense that you can have indivisible magnetic dipoles despite lack of the monopoles?

What is a magnetic analogue of electric potential? Is there something analoguous to electic capacity?

Do there exist some laws that forbid existence of magnetic semiconductors? If not, have they been researched or even invented?
Phys.Org News Partner Physics news on
'Squid skin' metamaterials project yields vivid color display
Team finds elusive quantum transformations near absolute zero
Scientists control surface tension to manipulate liquid metals (w/ Video)
Oct14-12, 01:57 PM
P: 12,037
Before you think about magnetic semiconductors and more advanced elements: How does a magnetic conductor look like? It conducts magnetic charges. How do magnetic charges look like? We don't know.

What is a magnetic analogue of electric potential?
Magnetic vector potential

Do there exist some laws that forbid existence of magnetic semiconductors?
I think you have to define what you mean with "magnetic semiconductors" first. "Something which can be called semiconductor but somehow uses some magnetism" is not a description.
Oct14-12, 02:19 PM
P: 1,115
Quote Quote by haael View Post
Do there exist some laws that forbid existence of magnetic semiconductors? If not, have they been researched or even invented?
First two hits on a Google search:

Nov21-12, 09:14 PM
P: 13
Is it possible to have a magnetic diode?

I doubt if this is relevant but the words bring it to mind.
While researching magnetic amplifiers I once wondered if an easily saturated toroid wound with many turns on one winding (a) and biased by another similar winding (b) with a steady DC current, would act as a diode when winding (a) is exposed to AC. So I made the device and observed the result with an oscilloscope. The answer is no. The current waveform is distorted but while there is a large current when the device is saturated, that doesn't last long and there is a long part of the cycle during which the current is small but in the opposite direction. This was a definite disappointment.
Nov22-12, 01:13 AM
P: 438
Thank you for refreshing this thread.

No, I didn't mean that. I'm not talking about an electrical semiconductor that exhibits ferromagnetism, but about some material that allows magnetic polarization in only one direction.

I was referring to the analogy between the electricity and magnetism often exploited in engineering. Magnetic field is analogous to the electric current. We can say about magnetic resistance, the analog of Kirhoff's law and so on. We can for sure have magnetic conductors, magnetic SEM, magnetic resistors and magnetic inductors (which happen to be the same thing as electric inductors, which serves as the connection between electric and magnetic circuits).

Usual hard magnets will be the analog of electrical superconductors.

My question is: does it go further? Could we have magnetic capacitors and magnetic semiconductors?

As for capacitors I think the answer is no, because of lack of magnetic charges. But what with semiconductors? Do there exist such materials that can alter magnetic field in some odd maner, for instance allowing magnetic diodes and transistors?

A magnetic diode in my view is a material with this property: if we have a magnetometer and we shield it with the "diode" on one side then when we put one pole of a magnet near that shield the magnetometer will show a usual magnetic field, but if we flip the magnet and use another pole, then the magnetometer will show zero.
So it is a one-way shield for magnetic field.

Hard magnets are like superconductors. I wonder if we could see any phenomenons similar to a superconductor-metal junction in your experiment.
Nov22-12, 03:26 AM
P: 900
no this does not work because you're thinking of 2 different things. one is rectifying a current (flow) of particles obeying Fermi-Dirac statistics. that's a diode.

magnetic field is not analogous to an electric current. electric current is exactly what it says it is: a particle flow. that's why there's *fluid analogs of electricity*. Fluids are also a flow of particles obeying Fermi-Dirac statistics.

ok, so tell me what particle makes up a magnetic field and its flow properties? What is flowing and how does it flow? Only if you have a flow can you talk about current. If you're talking about shielding from a field, that's not a current.

whats a magnetic conductor? the other name is "high impedence surface". Doesn't sound like a conductor anymore does it?
Nov22-12, 02:44 PM
P: 661
Haael, beware the similarity with electric current is very limited. You're overstretching it.

A magnetic material polarized near saturation is a magnetic rectifier. Use AlNiCo if you want both the magnetization and the saturated permeability in one material. AlNiCo won't let the induction increase easily beyond saturation, but reducing it is rather easy.

wvphysicist, you didn't see an asymmetric effect on the voltage because voltage results from the derivative of the flux, so over one flux cycle, the mean induced voltage is zero. But if measuring the induction or the flux, say with a Hall sensor, you would see an asymmetric effect, with the induction dropping more easily than increasing.

Could the asymmetry be obtained near zero induction? Not that I know. Some sensors, especially the fluxgate, rely on the assumption that the magnetic material used is perfectly symmetric.

This is a difficulty only for homogenous materials. If you combine several parts, some hard saturated in one direction, others magnetized in the other direction without hard saturation, I'm confident you can get an asymmetric behaviour near zero induction.

Less direct: take an anisotropic permeability - thin parallel lines of isotropic material if you wish. Put a strong permanent field at +45, and put a field and observe the induction at -45, the H-B curve will be non-linear.
Nov23-12, 01:12 PM
P: 13
If you are interested in materials properties you should read up on superfluids. Helium 4 and helium 3 have a host of interesting properties from vortices to second sound. Some of these phenomenon have only recently been discovered like the acoustic oscillation in a capilary between two resoviors. I believe I read about it in a recent Physics Today. Have fun.

Register to reply

Related Discussions
Simple Diode Analysis Problem - Calculating Current/Voltage in a diode: Engineering, Comp Sci, & Technology Homework 1
Inductor, switch, snubber diode(flyback diode) Electrical Engineering 2
How do I find the forward resistance of a diode using diode law? Engineering, Comp Sci, & Technology Homework 3
Help with Diode Circuit with Current/Voltage through Diode Engineering, Comp Sci, & Technology Homework 6
A Diode in a Magnetic field Introductory Physics Homework 0