How to differentiate conductors and semiconductors?

[y4ku24]

Hello everyone,

I have a (simple?) question: How do you differentiate a material to be a conductor or a semiconductor? It is just only by examining the material's electrical resistivity/conductivity whether it is low or high? Because I have heard that even some metallic conductors have high resistivity.

For you information, I am currently working with a material called metallic glass, with a compound of Zr55Ni5Al10Cu30. It has high resistivity at room temperature, but I am not sure whether only by that fact I can deduce it to be a semiconductor.

Thank you in advance.

y4ku24

 

[fatra2]

Hi there,

What kind of answer do you expect?

Macroscopic answer: a conductor is defined by its conductivity, or in other words, by its resistivity (the ease for current to flow).

Microscopic answer: by looking at the lattice of atoms inside a material, you will find the electrons more or less bound to the atoms. The energy needed to put these valence electrons into the free band defines the conductor type. A conductor will have pratically its valence electrons in the conduction band, which a semi-conductor will require a bit more energy from the outside.

Cheers

 

[y4ku24]

Dear fatra2,

Thanks for the info. I understand your point, but unfortunately, I do know the explanation for the macroscopic and microscopic point of view. What I actually wanted to know is, if you don't already know whether a material is a semiconductor or not, then how do you go about figuring it out?

Is it permissible to just look at the low/high value its resistivity is, or there are other ways to know for sure?

Take my sample for example (Zr-Ni-Al-CU). It composes of 4 elements, so how do you know if it is conductor/semiconductor? Is it just by logic that all the elements by itself are conductors, so that sample must also be a conductor?

On another note, if we assume that it is a semiconductor, how do I calculate the energy needed for the electrons to surpass the band gap for a multi-element material?

Thank you for your time.

 

[sokrates]

Hi y4ku24,

You'd have got more answers if you had posted this to Atomic - Solid State Physics section.

I think the most direct way would be a temperature dependent measurement. Because as you said, in RT, you might be giving the carriers enough kinetic energy to surpass the band-gap (if there is any).

In general, electrical resistivity of metals increases with temperature, while the resistivity of semiconductors decreases with increasing temperature.

So if cooling or heating is harder than the other for you, you could try the easier one.

If you cool the material, then you'd expect a very high resistivity for the semiconductor at some T (due to freeze-out of carriers), whereas the metal will probably have a lower resistivity due to the suppression of phonon scattering. (There is an exotic effect called "Kondo Effect" where the resistance of a metal diverges when T is close to 0K - but this regime becomes visible only after the temperature is really close to 0K - so you probably don't need to worry about it.)

If you heat it, the semiconductor resistivity will decrease because there will be many more free carriers thanks to thermal excitation, although phonon scattering also increases in the semiconductor, the overall resistivity will typically decrease.

This is probably the easiest way to go about it.

 

[JDRF85]

Hi there

Here's another idea, if you like.

1) Measure the work function of your sample, see http://en.wikipedia.org/wiki/Work_function#Measurement" for some techniques;

2) Get a metal with a work function greater than that of your sample, and bring the two materials together to form a junction;

3) Apply a potential difference across the junction and measure the response of the current.

4) Reverse the potential difference and see what happens.

Basically, if the junction displays diode behaviour, you have a semiconductor-metal Schottky contact, and your sample is a semiconductor. If the junction displays Ohmic behaviour, and the work function of the metal is greater than that of the sample, then you have a metal-metal junction, and your sample is metallic.

NB this analysis applies only for n-type semiconductors. You may still have a p-type semiconductor and get Ohmic behaviour. If you observe Ohmic behaviour, you should repeat the experiment using a metal with work function lower than that of your sample. If you still see Ohmic behaviour, you have a metallic sample. However,if your sample is p-type semiconducting, you will see diode (rectifying) behaviour this time.

Check out:

"[URL
[/URL]




Johnny

 

[heylouis]

Zr base metallic glasses are metal-like conductor.
You can measure the electric resistance of the sample to prove that.

 

[y4ku24]

To sokrates and heylouis:

Thank you for the response. Thank you also for the advice to post this message in solid state physics. I will try to do that!
As for the temperature dependence measurement, yes, I have done that measurement, and yes, metallic glass's resistivity decreases as temperature increases (I used the heating method). After a certain temperature, it undergoes a steep decline because of the crystallization of the material.

My theory is that at first, the material was a semiconductor, thus, the steady decline. but after crystallization, it becomes a conductor, so the electrons can move more freely than (that is why we see the abrupt decline).

But, again, the question still remains: If it shows a steady decline due to heat treatment, is it sufficient enough to conclude that it is a semiconductor? In other words, does this mean that ONLY semiconductors behave this way, and no conductors behave the same?

To JDRF85: Thanks JDRF85. Looks kinda complex, but I will look into it.

 

[heylouis]

To sokrates and heylouis:

As for the temperature dependence measurement, yes, I have done that measurement, and yes, metallic glass's resistivity decreases as temperature increases (I used the heating method). After a certain temperature, it undergoes a steep decline because of the crystallization of the material.

My theory is that at first, the material was a semiconductor, thus, the steady decline. but after crystallization, it becomes a conductor, so the electrons can move more freely than (that is why we see the abrupt decline).

From one paper I read that localized spin is the cause to this temperature related phenomenon. In my memory only the energy band structure can prove one material is conductor or semiconductor. Not sure if your experiment is strong enough.
Seems it's a hot topic in metallic glasses research now. I'm not familiar with the amorphous electrical resistivity theory, so I stop here. Why not google "electrical resistivity Zr amorphous alloy"? The search results are professional.
Good luck!

 

[SSK12345]

My theory is that at first, the material was a semiconductor, thus, the steady decline. but after crystallization, it becomes a conductor, so the electrons can move more freely than (that is why we see the abrupt decline).

But, again, the question still remains: If it shows a steady decline due to heat treatment, is it sufficient enough to conclude that it is a semiconductor? In other words, does this mean that ONLY semiconductors behave this way, and no conductors behave the same?

I would recommend looking at a phase-diagram of the system to figure out what it is that you have at the different temperatures.

As for your question, that depends on your purpose/research question. From what you've posted, it does not follow. What you could say is that your material is a semiconductor in a certain temperature range, and a conductor in a certain temperature range. (i.e. think of the property of resistivity as a spectrum ranging across insulator-semiconductor-conductor.

 

[y4ku24]

To SSK12345: I have the same assumption as what you have said, which is at normal temperature it is a semiconductor, and at high temperature it turns into a conductor. But other than the resistivity result, I am yet to find any other evidence to prove that this assumption is plausible.

Thanks for reply though.