Basically the only difference between semiconductors and insulators is the width of the band gap from valence band to conduction band. In semiconductors this is typically ~1 eV and in insulators >5eV. However, the band structures are broadly speaking similar. (Iif you check the wikipedia article about semiconductors and its discussion page, you'll notice that this is not fully agreed upon)

From usage point of view, however, the key point is the ability to rather freely tune the conductivity of semiconductors by doping over range of several decades, which makes it possible to manufacture devices from it.

This material of 2,86eV seems to be somewhere in between. I'd say that if you can dope it and use it as a semiconductor, I'd call it as such. If not, i'd sat its an insulator.

Why is it important to be able to define your material as either semiconductor or insulator?

Basically the only difference between semiconductors and insulators is the width of the band gap from valence band to conduction band. In semiconductors this is typically ~1 eV and in insulators >5eV. However, the band structures are broadly speaking similar. (Iif you check the wikipedia article about semiconductors and its discussion page, you'll notice that this is not fully agreed upon)

From usage point of view, however, the key point is the ability to rather freely tune the conductivity of semiconductors by doping over range of several decades, which makes it possible to manufacture devices from it.

This material of 2,86eV seems to be somewhere in between. I'd say that if you can dope it and use it as a semiconductor, I'd call it as such. If not, i'd sat its an insulator.

Why is it important to be able to define your material as either semiconductor or insulator?

Dear Mikko Juntunen,

It's not important just I want to know.

If you come from basic point of view,

Insulator-------- >3eV

Semiconductor-------

The transition is arbitrary. I would say if you can dope it efficiently, it's a semiconductor, if not, an insulator. In practice 3eV is a good boundary.

There is another way to determine what type is your material. There are three categories for atomic bonds in solid materials. (1) Metallic bonds (like Aluminum) which have free electrons and are conductor. (2) Ionic bonding (like NaCl) which is insulator, and (3) Covalent bonds (like Silicon) which semiconductor. If your material has ionic bonds it can be insulator. Dielectric (insulator) materials have ionic bonds and their electron are not free.

It should be considered that by doping dielectric (insulator) materials they can become semiconductor.

To Amin: there are several exceptions in your categorization. For example, diamond is a covalent-bond solid and an insulator. In addition, solid organic materials has a lot more bonding/interaction types and there is no such general rule to determine conductor/semiconductor/insulator materials.

To the question: I agree with Mikko and Manuel. You have to look at your materials properties specifically to determine whether it's semiconductor or insulator, especially when the bandgap is not at the extreme (~1.5 eV for S/C and >5 for insulator).

Dear Hung, thanks a lot for your useful comment. Yes, there are several exceptions, but there are always exceptions in any kind of systems. In my point of view, regarding to Kotagiri's question, that simple definition (which has several expectations) can help him to find the answer, specially when in the literature as Kotagiri has mentioned it is dielectric (Insulator).

I would agree with Hung that saying ionic solids are insulators and covalent solids are semiconductors is misleading. Most common dielectrics (Al2O3, SiO2, Si3N4) have a certain electronegativity difference that places them somewhere in between, as do ZnO and SnO2 which are considered wide-band-gap semiconductors or even TCOs as they are readily doped. I'm fairly sure I've never heard of elemental sulphur, phosphorus and arsenic being used as semiconductors even though to my knowledge they are bonded covalently.

I agree with Manuel. I also think that the bonding type is not the issue here but the key characteristic are the ability to tune the conductivity/resistivity. The band gap is also used, but the ~3eV limit is arbitrary.

There is one more potential souce of definition. It should be noted that in pure form (which does not exist in practice) also the materials generally known as semiconductors are insulating since there are no free charge carriers left over after covalent bonding. Why then call them semiconductors? In practice, there always are impurities and crystal imperfections, bringing in extra electrons or missing electrons (=holes) that are practically free to move (=have very low ionization energy).

So,

Conductors always have plenty of free charge carriers

Insulators are insulating (in most commonly existing forms) even with naturally occurring impurities

Semiconductors then are materials which (due to impurities) always have some free charge carriers and thus some conductivity.

From a materials tetrahedron perspective ...

Nature

- Metals have metallic bonds, with some options for partial ionic character or d-orbital covalent bonds.

- Ceramics have ionic bonds, with some options for mixed covalent + ionic bonds (silicates for example).

- Polymers have covalent bonds and secondary bonds.

- Metalloids (also called semiconductors) have covalent bonds.

Properties

When the bonds are metallic, no band gap exists. When the bonds are covalent or ionic, a band gap exists. The width of the band gap determines the extent that electrons can be promoted to the valence band at any given temperature. The ability to dope the material and create additional charge carriers, either as holes or electrons, also plays a role.

- dielectrics allow polarization with breakdown

- metals allow conduction, and conductivity varies inversely with temperature

- semiconductors allow conduction, and conductivity varies exponentially with temperature

So, the term "semiconductor" crosses over from a classification by the nature of the material to a designation of the property of the material. Hence the inherent potential for confusion.

I suggest the clarity you seek should not be to classify your material based on its band gap energy (an intrinsic nature of the material), rather it should be to classify it in terms of how it responds to an applied voltage field at different temperatures (its conductivity property).